ABSTRACT 
EUTROPH ICATION, HYPOXIA AND TROPH IC 
Ti tle of Dissertation: TRANSFER EFF ICIENCY IN CHESAPEAKE BAY 
James Dixon Hagy III , Doctor of Philosophy, 2002 
Dissertation directed by: Professor Walter R. Boynton 
University of Maryland 
Center for Environmental Science 
Coastal eutrophication is a global problem that has contributed to loss of 
estuarine habitats and potentially decreased fisheries production. Hypoxia is often 
observed in eutrophic estuaries where it can be an important cause of habitat loss. 
This study utilized a suite of empirical analyses to examine key linkages relating 
coastal eutrophication to hypoxia, trophic structure, and trophic transfer efficiency in 
Che apeake Bay (CB), USA. A salt- and water-balance model , or "box" model, was 
developed to quanti fy large-scale physica l transport for CB, an input to many 
subsequent analyses. Historical ( 1950-1999) di ssolved oxygen (DO) data for CB 
1
showed that moderate hypoxia (D0<2.0 mg r ) increased - 3-fold, modulated by 
1
spring river flow . Severe hypoxia (D0<0.7 mg r ) occurred only in high flow years 
during 1950-1967, but was present annually since 1968. 
Analysis using tree-structured regression showed that hypoxia was the most 
important factor determining patterns of macrobenthic biomass in Chesapeake Bay. 
Carbon budgets showed that, where habitat quality was poor, macrobenthic biomass 
was much less than could be supported by the organic carbon supply. rn  these cases, 
even dramatic reductions in carbon supply would not be expected to limit benthic 
production and by extension, trophic transfers to upper trophic levels via the benthos. 
The effect of eutrophication and hypoxia on trophic structure and trophic 
transfer efficiency were examined by estimating trophic flow networks for three 
regions of CB during summer. In addition, a series of "rules" were described and used 
to infer the trophic flow network for a "restored" middle CB from historical data, 
comparative ecological relationships and mass balance constraints. Excessive carbon 
now through bacteria was the most pronounced symptom of eutrophication in the 
modern mid Bay. The microbial food web transferred organic matter to trophic levels 
comparable to large piscivorous predators, maintaining average trophic transfer 
efficiency, even as the fraction of primary production transferred to top predators 
decreased. In the restored Bay, increased macrobenthic production shifted metabolic 
activity away from the microbial food web, increasing the potential trophic transfer to 
fi sh by 7-fold, even as total primary production decreased to 63% of the current 
average. 
EUTRO PHICATION, HYPOXIA AND TROPH IC TRANSFER 
EFFICIENCY IN CHESAPEAKE BAY 
by 
James Dixon Hagy 11 I 
Dissertation submitted to the Faculty of the Graduate School of the 
University of Maryland, Co ll ege Park in partial fulfillm ent 
of the requirements for the degree of 
Doctor of Philosophy 
2002 
Advisory Committee: 
Professor Wa lter R. Boynton, Chair 
Professor W. Michael Kemp 
Professor Thomas J. Mi ller 
Professor Estelle Russek-Cohen 
Professor Robert E. Ulanowicz 
DEDJCATION 
To Melissa and Lau ren 
II 
ACKNOWLEDGMENTS 
I am grateful to Dr. Wa lter Boynton, my advisor, for support, advice, 
encouragement , understandi ng and good humor over the past IO years. Walter's 
unique blend or enthusia m, hard work, optimism and rea l ism , combined with his 
c lear po l icy or putt ing fam il y and per ona l matters first helped bring me to thi s point 
in one piece. Wa lter is a true fri end . It is hard to contemp late working for anyone 
else. 
I am also indebted to W. Michae l Kemp. I particularly thank Mike for 
challenging me to clarify my thin king and writing, elevating the quality o f my work. 
Working a few feet from M ike 's desk th rough the last and most challenging year of 
my dissertati on work, I frequentl y tapped him for papers, numbers, opinions, and 
adv ice. I thank him for hi s open door, and li ke W alter, hi s genuine friendshi p. 
Each of my other committee members contri buted substanti all y to my graduate 
career. I thank Bob Ulanow icz for introducing me to network analysis, providing hi s 
NETWRK software, and fo r many helpful discuss ions on speci fyi ng the networks and 
interpretin g the pages or output. A lthough any stati sti ca l erro rs are obviously my own, 
I thank stell e Russek-Cohen fo r her excellent stati sti ca l instruct ion, for introduci ng 
me to tree-s tructured data analys is, and fo r her rev iew of stati sti ca l methods in my 
di ssertati on. Tom Mi ller chall enged me at every point of my di ssertati on work , from 
the comprehensi ve exam to the di ssertati on defense. I espec iall y thank him for hi s 
deta iled rev iew of my di ssertati on, which was undoubtedl y a time-consuming 
undertaking. 
111 
While completing thi s research , I located my desk at three laboratories, 
UMCES Chesapeake Biological Laboratory, Virginia Institute of Marine Science, and 
UMCES Horn Point Environmental Laboratory. I thank the administration and staff 
of each of the e institutions for their hospitality and support. Circumnavigating the 
Bay in my graduate career, I had the pleasure and benefit of interacting with many 
talented researchers in the Bay community. This was invaluable considering the 
breadth of subject matter encompassed by my disse1tation research. For data, papers, 
or other technical guidance, I thank Cath y Lascara, LaJTy Sanford, Ed Houde, 
Sukgeun Jung, Se-Jong Ju, Liz Canuel, Rebecca Dickhut, Larry Harding, Jenny 
Purcell , Bill Boicourt, Larry Sanford, Rodger Harvey, Mary Beth Decker, Jeff 
Cornwe ll , Rodger Newell, Vic Kennedy, Erik Smith, Linda Schaffner, Mike Roman, 
and Diane Stoecker. This di ssertation could not have been completed without the 
combined talents of these people. 
A number of individuals he lped with the field component of this di ssertation. 
Ned Burger, Dave Jasinski and Jen Harman-Fetcho contributed to collections of 
ediment chlorophyll-a samples. The hundreds of successful sediment grabs collected 
for Chapter 3 would not have been possible without the dedication of Paul Moylan 
who tuned and operated the Smith-Macintyre corer. My role in the sediment 
chlorophyll -a method comparison described in Chapter 3 was limited to sampling 
design and data analysis. The field and laboratory work was sk.illfull y and generously 
completed by Bob Stankeli s, Jerry Frank, and Nancy Kaumeyer. I am deeply 
greateful for their generosity. I also thank the fish-head crew members involved in the 
Trophic Interactions in Estuarine Systems (TIES) cruises for their collections of fish 
IV 
used in Chapter 6 and for their assistance in collecting the water, plankton and other 
amples. All of the field co llections were conducted aboard the RIV Cape Henlopen 
and I am indebted to all her crew members who worked hard to make these research 
crui ses successful and , more often than not, enjoyable. 
1 am also grateful for the generous assistance of Jeri Phari s and Frances 
Rohland who committed themselves for years to helping me as needed for countless 
tasks. My Bay-encircling li festy le would have been vastly more challenging without 
their help . 
Financial support fo r my graduate studies was provided through a Chesapeake 
Biologica l Laboratory (CBL) Graduate Fellowship , a CBL Teaching Assistantship, 
through multiple contracts with the EPA Chesapeake Bay Program and by the NSF-
Funded Chesapeake Bay Land Margin Ecosystem Research Program (Trophic 
Interactions in Estuarine Systems, DEB-9412113) awarded to W.R. Boynton, W. C. 
Boicourt, E. D . Houde, W. M. Kemp and M. R. Roman . In the final months of my 
di ssertation work, fundin g was provided by the EPA-funded Multiscale Expe1imental 
Environmental Research Center at the UMCES Hom Point Environmental Laboratory. 
I am also grateful for the generous travel support provided by Walter Boynton and by 
the CBL Graduate Education Committee which enabled me to present my work at 
several Estuarine Research Federation Conferences . 
I am grateful for the unwavering support of my parents whose pledge to always 
love and support me no matter what has always enboldened me to pursue my 
ambitions. In 1997, I was fortunate to gain Wayne and Frances Ederington as parents-
in-law. I thank them for their love, thoughts and prayers. 
V 
Finally, l thank my wife and daughter, Melissa Ederi ngton Hagy and Lauren 
Hagy, for their love and support and fo r enduring my absence, particularly in the final 
months when I was "absent" even when I was home. I dedicate this volume to them. 
YI 
TABLE OF CONTENTS 
LIST OF TABLES ....... ........... ..... ... .. ....... .. ...... .. ........ ... ... ................................. ...... ix 
LIST OF FIGURES .... ............ ............ ........... .............. ... .......... ..... .. ... ... ............... xiv 
I !APTER 1. Background and Introduction ............ ....... ...... ...... .. .................... ....... 1 
Literature Cited ............................................................................... ..... ........ 7 
Figures .................... ..... ..... ........... ............... ...... ............. ......... .... ... .......... .... 9 
CH APT R 2. A Box Model Approach for Estimating Physical Transports 
and Exchanges for the Chesapeake Bay ................................................................. 11 
Abstract. .. .......... ............... ... ....... ................... ...... ..... ..... ....... ...... .. ....... ... .. .. I I 
Introduction ............... ...... .. ........ .... ... ........... ..... .... ... ... .. ..... ......................... 12 
Methods .............. .. ..................................................... ............................. ... 17 
Results and Discussion ............ ...................................................... .. ....... .... 29 
Conclusions .. ...... .......... ........ ................. ... ............... ... ... ...................... ....... 37 
Literature Cited ....... .. .... ........ .................................................. ... .............. .. 38 
Tables .... ....... ..... .. ................. .. .. ............................... ....... .. .................. .... ... 42 
Figures ... ... ........... .. .... .. .... ................................... ... .. .................... ............. . 48 
CHAPTER 3. Hypoxia in Chesapeake Bay, 1950- 1999: Progressive 
Development and River Flow Effects ........... ................. ........ .... .... ......................... 61 
Abstract. ....... ................... ..... .. ............. .... .... .... ... .... .... ...... .... ........ .... ... ....... 61 
Introduction ... .. ... .. ... ............. ......... ...... ....... ......... ... ........ ..... ........ ............ .. . 62 
Methods .. ... .. ................... .. ........................... ..... ... .. ............ .. ...... .. .... .......... 66 
Results ........... .... .. ..................................... .. ...... ..... .. ........ ............. ...... ... .... 71 
Discussion ....... .. .......... ...... ... ....... .... ... ... ..................................................... 78 
Literature Ci ted .................................................... .... ........ .. ..... .......... .... ..... 87 
Tables ..... .... ......... .......... .. ......................................................... .... .. .... .... ... 91 
Figures ....... .. ....... ........ ..... ........ .. .................. ... ....................... ........ ... .... .. ... 96 
CHAPTER 4. Phytoplankton Deposition to Chesapeake Bay Sediments 
during Winter-Spring ... ......... ........ .... ...... .... ............... .. ................ ... ............ ....... .. 106 
Abstract. ................... ...... ....... .. .. ........ ........... ...... .... .... ..... ... ..... ... ........ ...... I 06 
Introduction ....... ................................................................. ... .. ................. I 07 
Methods .................................................................................... ............... l 11 
Results and Discussion ... ......... ... ..... ........... .. .. ........ .. ................... ......... .... 1 14 
Conclusions ................................ ... ......................... ......... .. .. ..... ................ 130 
Literature Cited ............ ..... .. ....... ............ .. .. ......... ..... .......... ...... .... ............ 131 
Tables ........... ..... ............. ... ....... .......... .... ...... ........... ... ... .. .. ... ... ... ...... ... .... 136 
Figures ............. ........ .... .... .................. ..... ......... .... .... ...... ......... .... .. ........... 141 
VII 
Cl IA PTER 5. Patterns or Macrobenthic Biomass and Community 
Bioenergetics in Chesapeake Bay during Summer in Relation to 
1l abitat Qua lity and Organic Carbon Supply ...... ........... ...... ... ........... ...... .... .. ... .... 150 
Abstract. ..... ......... ... .... ... .... .... ........ ... ... ...... ... .. ..... ..... .. .... .. ... .. ............... .... 150 
Introduction ... .... ... .... ....... .... .. ....... ... .......... ... .. ......... .. ..... .............. ........ .... 151 
Methods ............. ..... ...... ...... .... .. ........... ... ....... ... .......... ... ........... ... .... ... ... .. 156 
Results and Discussion ...... .. .. ...... ..... .... ........ ..... .... ........ ............ ... ..... ...... . 166 
Summary and Conclusions .............................. ..... ... .. .. ........... ..... ........ ..... l 86 
Literature Cited .......... ..... ..... ... ...... ........ ........ ..... .... ......... ..... ..... .......... ..... 189 
Tables ....... ........ .. .... .... .......... ....... ........ ... ...... ... ...... .. .... ... ... .. ... ............ ..... 196 
Figure ... .... ...... ......... .... ........... ...... .... ..... .. .. .... ... ..... .. ..... ... ........ .... ........... 205 
CHAPTER 6. A Network Analys is or Mainstem hesapeake Bay Food 
Webs during Summer: Eutrophicat ion Effects on Carbon 
Transfer Effi ciency to Fish ... ..... .... .. ..... ...... .. ...... ................... ..... .... .................. .... 2 16 
Abstract. ... ............ .... ..... ...... .. .... ... .. .. ... ... ..... ... ...... ..... ........ .. ... ... .... .. ...... .. . 216 
Introduction .. ..... ... .......... .. ...... .... ...... .... ... ....... ..... .... ... ......... .. ...... ......... ... . 2 17 
Methods .. ... .. .. ........... .......... .. .... ...... ..... ...... .... ........ .. ..... ......... .. .... ......... ... 223 
Results and Discussion ..... ....... ........... .... ... .. .... .. ... ...... .... ...... ...... ... ...... .... . 234 
Literature Cited ... ... ...... ..... ... ... ..... ........... ...... .. .......... .... .. .... ... ... .... .. ..... .. .. 264 
-rabies ... ......... ... ...... .... ......... ... .... ..... ..... ...... ... .. .... ... ....... ............ ...... ... .... . 280 
Figures ...... ... ..... ... ..... ....... ..... ... .... ... ......... ... ...... .... ........ ....... ... ....... .. ....... . 299 
CHAPTER 7. Validation of Trophic Level Estimates for Chesapeake 
Bay Food Web Using 15N/ 14N ..... ......... ....................... ........... ...... ... ...... ..... ....... ... 311 
Abstract. ... ............ .. ........... ...... ... ..... .... ... ... ......... ........ ..... .. ....... ...... .. ........ 311 
Introduction ... ...... .. .. .... ....... ..... ... ..... ... .... ..... .... ............ .... ..... ..... .... .. .. .. ..... 31 l 
Methods .. .. ... ... .... ........... .. .. .... .. .. .. .... ........ ... ....... .. .... ... ... .... ... ...... ......... .... 314 
Results and Discussion .... ... .. ..... ...... ..... .. .. ... ...... ......... .. .. ... .... ... ... .. ..... ... ... 315 
Literature Cited ....... .... .. ........ ... .. ... .... ... ..... .. .. .... ...... .... .... ... ... ... .... ..... ... .... 320 
Tables .. ...... ...... .. ....... ....... .... ... ... ... .. .. ... ..... ..... ... ....... .... ... ..... ...... .. .. ..... .. ... 323 
Figures ....... ... ..... ... .... ...... ...... ....... .... ..... ... ..... ... .......... ... ...... ... ......... .. ....... 326 
APPENDIX A. Description and Discussion of Procedures for Estimating 
Network Models for Chesapeake Bay .. .. ... .. .... ....... ... ..... ..... .... .... ............... ... .... .. . 328 
Organic Storage, Primary Production and Bacteria ... ... .... .... ........ .... .. ... ... . 328 
The Plankton Community .. .... ........ ... ......... ... .......... ...... ........ ... ..... .. ...... ... . 343 
The Benthic Community ..... ... .. ..... ..... ..... .... .. .. .. ... ....... ...... ...... ... .... ..... .. .. . 362 
Finfish and Crabs .. .. .......... ....... .. .. ... ............ ... .... ... ... ....... .... ......... ..... .. .. ... 372 
Literature Cited ..... ..... ....... .. ... .. .... .. ... ..... ...... .... .... .... ...... ... ... ... .. ...... ........ . 41 2 
COMBlNED LITERATURE CJTED ... .... ...... ....... ..... ............ ......... ...... .... ....... .... 444 
VIII 
UST OF TABLES 
Tab le 2- 1. The boundaries of the box model segments in channel kilometers 
from the mo uth of the Bay and the depth of the pycnocl ine dividing the surface 
and botton1 layers . ..... .................. .... ..... .................................................................. 42 
Tab le 2-2. The volume of mainstem Chesapeake Bay and the tributari es and 
embayments with which latera l sa lt exchanges occur and the stati ons used to 
estimate mean sa linity th roughout the Chesapeake Bay estuarine system ............... 43 
Tab le 2-3. The fall line gauges, USGS ID and drainage area upstream of the 
fo ll li ne gaug ing station on the maj or tributaries of Chesapeake Bay .... .......... ...... .. 44 
Tab le 2-4. The dimensions, including the length , pycnocline area, 
cross-sectional area (cross-sec), and vo lume of the surface layer (SL) and 
bottom layer (BL) boxes in segments 1 through 9 of the box model. .... ..... .... ........ .45 
Table 2-5. Box model estimates of the long-term ( 1986- 1998) average monthly 
p 11 ys1.c a I tran sport (m J s - I ) for Chesapeake B ay ...... .......... .......... .. ....................... .. .. 46 
Tab le 3- 1. Data types, spatia l reso lution, tempora l resolution and sources of 
data . ... .. ............ ..... .... ....... ... ........... .... ..... .............. ........... ..... .... ........... .. .. ... ... ....... . 9 1 
Tab le 3-2. Data sources, cru ise dates, and ca lculated hypoxic vo lumes 
for Chesapeake Bay from 1950-1999 ... ....... .... ..................... .... ....... ..... ...... .. .......... 92 
Table 3-3. Estimated parameters of the non-linear multiple regression, 
ln (V + I) = /Jo + {3 1 (T - 1949 J + f3 2Q + E, relating hypoxic vo lume 
( 109 m3) to time (year) and January-May average Susquehanna River 
now ( 111 3 s- 1) ... . . .. .................. . .. . ....... . ...... . .... . ... .. . ................................ . . ......... . .... . ... 93 
10 3 
Table 3-4. Regress ions re lating hypoxic-volume-days (units: 10 m -days) 
fo r Chesapeake Bay for 1985 -1 999 to January-May average Susquehanna 
Ri ver flow ......... ............... .. .. .......................... ... .... .. .... ... .... .. ... ... .... .. .. ............. ..... .. 94 
Table 3-5 . Estimated rates of spring DO decline in bottom waters at station 
845G for 1964-1977 and at station CB4.3C for 1985 -1998 .............. ... .... .... ........ .. .. 95 
Table 4- 1. The most abundant phytopl ankton taxa ( excluding picoplankton) 
in three regions of Chesapeake Bay during spring and the average fraction 
of tota l phytoplankton carbon contributed by diatoms . ... ..... ......... .. ...... ... ........... .. 136 
Table 4-2 . C ruise dates for sediment chlorophyll-a surveys and the number 
of sediment cores collected in each region of the Bay .. ... .. ............. .. ....... ........ .. .... 13 7 
IX 
Tabl e 4-3 . Re ults of a method compari son experiment used to evaluate the 
effect of three sonicati on trea tments and si ngle vs . double ex tracti on on the 
amount (mean?se, % change from contro l) of chi-a (?gig) extracted from 
15 aliquot of homogeni zed Che apeake Bay sedim ents .. .. .. ........ .... .. .. .. ...... .. ....... 137 
Tabl e 4-4. Regional/annual mean sediment total chl orophyll-a inventories. 
These were computed from 0-1 cm chi-a inventories by adjusting for mixing 
to below I cm on short time scales (i.e. days-weeks) ... .. .. .. ........ .. .... .. ........ .... .... .. . 138 
Tabl e 4-5. Minimum, maximum and modal va lues used to specify triangular 
di s t n' b ut1. ons for parameters m.  eq. 3 . .. .. ........ .. .. .... ............. .. ......... ... .. .. .. .... .. ........ . 139 
Tabl e 4-6 . Estimated average (?standard deviation) chi -a deposition rates 
(mg m?2 ct ?1) and total winter-spring chi-a deposition (mg m?2) for 
winter?-spring in the upper, mid and lower Chesapeake Bay 
during J 993-2000 ..... .. ......... .. .... .. .. ....... .. .. .... .... .. .. ..... ... .. .. .. ........... .... ... ......... .... ... I 40 
Tabl e 5-1. Parameter estimates and F-table for a multiple regression 
2 
(r2~0.34)_fr.edictin~ log _ash-free ?iomass per m (recoded by adding 
0.0 g m ) at locallons 111 the mamstem Chesapeake Bay ............................ .. .. .. .. . 196 
Table 5-2. The importance of water quality and habitat variables as factors 
in a CART regression-tree model predicting biomass of mainstem 
Chesapeake Bay macrobenthic communities ............................... ... ................ .. .. .. 197 
Table 5-3. Node detail for the regression-tree model shown in Fig. 5-4 ..... .... ...... 198 
Table 5-4. Regional area and summer (June-August) average biomass, 
daily production and daily P/B ratio, as estimated using the Edgar (1990) 
n1odel. ........ .. .......................... .. ................. .... ....................... .. ...... ....................... . 200 
Table 5-5 . Estimated ranges for summer (June-August) average 
consumption (C), re piration (R), and excretion plus egestion (U) by 
macrobenthic communities in Chesapeake Bay ... .. ............................................... 20 I 
Table 5-6. The parameters describing mixing characteristics of 5 segments of 
Chesapeake Bay, leading to estimates of the fraction of the water volume 
available to suspension feeders each day .............................................................. 202 
X 
Table 5-7. Computati ons based on the models of Gerritsen et al (1994) 
showing the relationship between potential ratio available via suspension 
feeding and the range of estimated carbon requirements of the suspension 
feed ing macrobenthos. ???????????? ?? ????????????????? ???? ??????????? ?????????? ????????? ??? ??????????? ?????? ???? 203 
Tab le 5-8. Average total benthic metabolic rates for three regions of 
hesapeake Bay during June-August. .... ...... .. .... ........... ........ ............................... 204 
Tab le 6- 1. Definitions of parameters in equations ............ ... .. ..... ....... ................... 280 
Tab le 6-2. Sources of published and unpublished data for defining the trophic 
flow networks ....................... ... .. ...... ... ...... .. .... .. ............ ....... ............... .... ............ . 28 1 
Table 6-3. Scientific names of species or groups of species referred to in the 
tcxt. .................... ........... .... ..... .. .................................. ... ........... .. ............ ......... ..... 284 
Tab le 6-4. Estimated biomass (mgC m-2) for each node in the summer trophic 
flow networks for the upper, mid and lower Chesapeake Bay ............. .................. 285 
Tab le 6-5. Es timated trophic level (TL), production (P, mgC m-2 d- 1) and 
ccotroph ic effic iency (EE) for each node in each region of the Bay . .. ... ....... ..... .. .. 287 
Table 6-6a. The matrix of organic carbon flows (T;j, mgC m-2 d- 1) among the 
nodes of the upper Chesapeake Bay summer trophic flow network ... .. .... ........ .. ... 289 
Table 6-6b. The matrix of organic carbon flows (TiJ) among the nodes of the 
mid Chesapeake Bay trophic flow network ........................................................ .. 29 l 
Table 6-6c. The matrix of organic carbon flows (TiJ) among the nodes of the 
lower Chesapeake Bay trophic flow network ........................................................ 293 
Table 6-7. Respiration (mgC m-2 d- 1) for each node in the three summer 
trophic fl ow networks (UB=upper Bay, MB=mid Bay, LB=lower Bay) and 
the fraction of total respiration contributed by each of 5 major groups .................. 295 
Table 6-8. Estimated summer fisheries landings in each of three regions of the 
mainstem Bay ................................................................................................ ...... 297 
Table 6-9. Estimated trophic level (TL), biomass (B, mgC m?2), production 
(P, mgC m-2 d-1), ecotrophic efficiency (EE), and respiration (R, rngC m-2 c1-1) 
for the restored mid Chesapeake Bay trophic flow network . ............................... .. 298 
Table 7-1. The date of the sampling cruises during which organic matter 
samples were collected for stable isotope analysis .. .... .......... ..... .. .............. ......... .. 323 
XI 
Tab le 7-2. Mean (?standard error) 8 15N in for each region/season 
combinati on ........ ............ .. ......... ..... .. ........... ........ ... .. ... .......... .... ............. .. ........... 323 
Tabl e 7-3. Estim ated trophic level (?standard error) for mesozooplankton 
based on PN as the reference sample ....................... .. ...... .. ... .... .. ............ .............. 324 
Tab le 7-4. Est imated reference trophic leve l for mesozooplankton (used in 
eq. 2) and the resulting trophic leve l estimates for mesozooplankton based 
on 15N/ 14N . For compari son, the estimated trophic level for mesozooplankton 
computed from trophic fl ow networks (Chapter 6) is also shown . .. ........... ........... 324 
Tab le 7-5. Trophic level estimates, referenced to mesozooplankton, for 27 
tax a or groups of taxa, arranged in order of increasing trophic leve l estimate .. ... .. 325 
Tab le A- 1. torage and physical input and export of detrital dissolved organic 
carbon (DOC) and detrital particul ate organic carbon (POC) . ... .. ................. ..... ... .427 
Table A-2. Estimates of summer average primary production rates deri ved from 
I lard ing et a l. (2001) and Kemp et al. ( J 997) .. ......... ............ .. ... .......... ..... ...... ...... 427 
Table A-3. Summer phytoplankton biomass, gross phytoplankton production 
(G PP) , algal respiration (R), extracellular release of DOC (ER), and net 
PO production by phytoplankton (PP) .. ........ .. .......... ..... .... ... ..... .......... .... ......... .428 
Table A-4 . Biomass, gross primary production, algal respiration, extracellular 
DOC release and net production by microphytobenthos in three regions of 
Chesapeake Bay during summer. .... ..... ........ .... ........... ........ ... .... ... ..... .... ... ........ ... 429 
Table A-5. Coverage of submersed aquatic vegetation (SAV) in Chesapeake 
Bay by region ................ .... ......... .. ..... .... ......... .... ..... ....... .. ....... ..... .. ..................... 429 
Table A-6. Biomass, production, consumption and respiration of free-living 
and parti c le attached bacteria in Chesapeake Bay during summer. ...... ... .... .......... .430 
Table A-7. Estimates of biomass (B), production (P), resp iration (C), 
consumption (C) and and egestion plus excretion (U) for the zooplankton 
in the upper, mid and lower Bay ...... .... .... .. ........... .. ... .... ....... .................... ... .... .... .43 l 
Table A-8. Biomass, consumption , respiration and egestion for copepod 
nauplii and mesozooplankton ...... .. .......... .. ..... ......... ... .. ...... ........ ...... ......... .... ... .. .. 432 
2
Table A-9. Geometric mean biovolumes (ml m- ) of the ctenophore 
Mnemiopsis /eidy i for summer in three regions of Chesapeake Bay ......... ... ....... .. .432 
XII 
Table /\-10. Geometric mea n biovolumes (m l 2111 ? ) of Chrysaora 
qui11quecirrha for summer in three reg ions of Chesapeake Bay ............................ 433 
Table A-11. Active sediment carbon pool size, and rates of sediment 
bacterial metaboli sm ............................................................... ................ .... ... ... ... 433 
Table A-12. Biomass and bioenergetic rates for benthic bacteria , meiofauna, 
suspension feeders and deposit feeders ............................ ............ .. .... ...... .. ........... 434 
Table A- 13. 1:stimates of biomass (B, mgC 111?2), specific growth rate 
(P/8 , ct?'), gross growth efficiency (GGE=P/C), net growth efficiency (NGE), 
assimi lation e fficiency (AE), biomass accumu lation (BA), net migrations 
(immigration-emigrat ion), and removals by recreational fishing (RF) and 
commercial fi shing (CF) ..... ... ... ..... ...... ... ................ ..... ......... ..... ... ..... ..... . ............ 435 
Table A-14. (A) The fraction of al l blue crab habitat located in the 
'hcsapeake Bay and its tributaries that is located in each region of the 
main tern Bay. (B) The total blue crab biomass in each region. (C) The 
total urface area of each region (includes non blue crab habitat area). 
(D) The blue crab biomass per unit of bottom area ... ........................................... . 437 
Table A-15 . Biomass (MT wet-wt) of adult Bay anchovy in Chesapeake 
Bay during sum mer of 1993 ................................................................................. 437 
Table A-16. Preliminary estimates of biomass (tons wet-wt) of adult Bay 
anchovy in Chesapeake Bay during summer of 1995-1999 .................................. .438 
Table A-17 . Age-structured carbon balance for bay anchovy in Chesapeake 
Bay during sum mer. ...... ........................... .. .......................... ...... .... .. ............ ........ 439 
Table A-18. Diet composition(%) by weight of adult bay anchovy during 
J . . 
une 111 mid-Chesapeake Bay ......................................... ....................................... 440 
Table A-19. Specific consumption and total consumption for bay anchovy ..... ... .441 
Table A-20. Consumption for the combined bay anchovy cohorts with diet 
a llocated to diet components according to diet composition in Table A-18 .......... .442 
Table A-21. Monthly and summer total respiration (mgC m?2) by bay 
anchovy. Rates were converted from 0 2 consumption using respiratory 
quotient = 1.0 ............. .... .............................................................................. ...... ... 443 
Table A-22. Adjustments to initial bioenergetic estimates for bay anchovy to 
achieve carbon balance for the summer. ............................................................... 443 
X111 
LIST OF FIGU RES 
Fig. 1- 1. Hypothesized functional relati onships between total organic input 
(p rim ary prod uction ex ternal inputs) and fi sheries production . ... ... ... .. .. ... .... .. .. ... ..... 9 
Fi g. 1-2. A conceptual diagram illustra ting mediation of the transfer of 
organi c carbon from inputs and primary producers to fi sheri es yields ..... .. .. ....... .. .. . I 0 
Fig. 2- 1. A map of Che apeake Bay showing the hathymetry, the boundaries 
o f" the segment of the box model, the embayments that were included in the 
salt budget, and the fa ll lines for whi ch freshwater inputs data were obtained ...... ... 48 
Fig. 2-2. (A) The simplest poss ible box model for an estuary, where 
Qi is freshwater input , Q is advection out, and E is di spersive exchange 
at the seaward margin . (8) An illustration of the possible physical exchanges 
fo r a surface layer box (box "m") in a verti ca ll y structured, branching model 
such as the Chesapeake Bay box model. .... ....... .... ... ... ... ... .. ..... .. .. .... ...... ......... .. ...... 50 
Fig. 2-3 . A schematic diagram of box model for Chesapeake Bay, drawn 
overl ying the maximum depth profile of the estuary .. .. .. ....... ...... ....... .... .. ......... .. .... 51 
Fig. 2-4. An illustration of the definition ofmax(N2), a measure ofpycnocline 
strength ba ed on the squared Brundt-Vaisala frequency (N2). ..... .. ... ....... ... ...... .. .. .. 53 
ig. 2-5. Monthly average down-estuary advection computed using the box 
model and expressed as a multiple of the total landward freshwater input (R) .. .. ... . 54 
Fi g. 2-6 . A compari son of cross-section average current velocity computed by 
the box model for surface layer and bottom layer in segment 5 with residual 
current velocity obtained by averaging 1995-1999 current meter data at the 
mid-Bay mid-channel buoy of the Chesapeake Bay Observing System .......... ..... ... 55 
Fig. 2-7. Seasonal and spatial patterns in average upwelling current velocity 
(cm d-1) in Chesapeake Bay as computed by the box rnodel.. .. .......... ...... ...... ..... ..... 56 
Fi g. 2-8. A profile of the Brundt-Vaisala frequency in the region of 
Chesapeake Bay just to the south of the Potomac River, collected in 
April 1998 by the Chesapeake Bay LMER (TIES) program using a towed 
undul ating CTD device ... .. ..... ... .................. ... .... .... .... .. ... ............ ...... .. ...... .... .. .... ... 57 
Fig. 2-9. Seasonal and spatial patterns in average non-advective exchange 
ve locity (cm d-1) in Chesapeake Bay as computed by the box model ......... ... .. .. ... .. 58 
XIV 
Pig. 2- 10. Upper Panel - The relationship between January-May average 
' usquchanna River !low and April-September average water column 
stratifica tion index , max(N2), in the mesohaline Chesapeake Bay. 
Lower Panel - The relationship between April-September median vertical 
di ffu ivity (DJ and January-May average river flow .... .... ........ ....... ..... ................. . 59 
Fig. 2- 1 I. A veragc up-estuary advection within the mid Bay region related 
to .J anuary-May average usquehanna River flow (m3 s? 1) ?.??.?.?.?. ... ..???..??..??.??..?...? 60 
I? ig. 3- 1. /\ map of Chesapeake Bay showing the locations and identifications 
of Chesapeake Bay Monitoring Program Water Quality Monitoring stations 
used for thi s ana lysis ....... ........ ... .......... .. ... ... .... ........................ .... .... .. .. ........ ..... ..... 96 
Fig. 3-2 . The time-series of hypoxic vo lume (DO< 1.0 mg 1-J) in Chesapeake 
Bay in 1993 . The shaded area is the integrated hypox ic-volume-days for 1993, 
computed by summing the areas of the indicated trapezoids ..... ......... ............ .... .... . 97 
Fig 3-3. A schematic diagram of the physical transport of dissolved oxygen 
into and out of three segments within the mesohaline Chesapeake Bay .... ........... ... 98 
Fig 3-4. alculated summertime hypoxic vo lumes for Chesapeake Bay 
during 1950-2000 as reported in Table 3-2 ...................................... ....................... 99 
Fig. 3-5. Predictions and asymmetric confidence bands(? standard error of 
mean) for July hypoxic volume (at each of three definitions) as a function of 
time and January-May average Susquehanna River flow based on the 
non-linear regressions fitted to data in Fig. 3-4 ............................... .............. ... ..... JOO 
Fig 3-6. Summer dissolved oxygen profiles in Chesapeake Bay during four 
years with near average January-May Susquehanna river flow ....... ........ .... ....... .. . 10 l 
Fig. 3-7. The relationship between January-May average Susquehanna River 
flow and time-integrated hypoxia .......... .............. .. ............. ... .................. .... ...... .. . 103 
Fig. 3-8 . The relationship between January-May average Susquehanna River 
flow, March bottom water temperature, and the date of onset of anoxia in 
bottom water at station CB4.3C ... .... .................. ... ...... .... ...... ... ...... ....... ............ ... I 04 
fig. 3-9. Average vertical diffusive, net horizontal advective, and total 
disso lved oxygen inputs to the mid Bay region of the Chesapeake Bay 
during 1950- 1999 ..... ... .. ..... ...... ... .. .. .. ..... ....... .... .......... ....... ....... ..... .. ....... .. ... .. ...... 105 
Fig 4- 1. Average seasona l distribution of water column integrated 
ch lorophyll -a (mg m-2) in Chesapeake Bay (1984-1999) ............... .. ................... 141 
xv 
I? ig. 4-2. A map or Chesapeake Bay indicating regional boundaries and 
the di stribution of ediment types a computed from the Chesapeake Bay 142 
Monitoring Program Benthic Data .............. .... ...... .... .... .................... .. ............ ..... 
1-ig. 4-3. The distribution of total chlorophyll-a in the top I cm of 
'hesapcake Bay sediment during late spring in I 993-2000 .... ... .......................... I 43 
f-=' ig 4-4. Regi onal and overall mean sediment total ch lorophyll-a inventories 
in late spring related to winter-spring (Jan-Apr) average 
? usq uehanna River flow .. .. .... .... ...... ...... .... .... ...................................................... 145 
Fi g 4-5 . January-Apri l average water column integrated chi-a in the lower 
'hcsapeakc Bay during 1993-2000 related to Jan-Apr average Susquehanna 
River tl o\v .... ..... .... .......... .. ... ...... .. ... .. ...... ........... ..... ......... ..... .. ........ .................. ... 146 
Fig 4-6. The relationship between January-April average water column 
integrated ch lorophyll-a and sediment chlorophyll -a in each of three 
reg ion of 'hesa peake Bay ..................... ......................... .......... ............ .. .... .. .... .. 147 
Fig. 4-7. Water co lumn integrated ch lorophyll-a concentra ti ons in Chesapeake 
Bay averaged by region. Verti ca l dotted lines indicate the dates of sediment 
ch i0 1?o  p I1 y 11 -a n1?a pp1? ng stu ct?1 es . ... ............ ............ .... ...... .. ?. ? .. ? ? ? .. ? ? ? ... .. . ? .. .. ... ? ? ? ? ? ... .. 148 
F ig. 4-8. Monthly mean and standard errors of particulate carbon (PC) 
sinking fluxes measured using sediment traps just below the pycnocline 
in the mid hesapeake Bay during l 984- 1992 ..... .. .. ...... .. ..................................... 149 
Fig. 5- 1. The loca tions where cores were co ll ected by the Chesapeake Bay 
Benthic Monitoring Program in the mainstem Chesapeake Bay 
fro 1n 1984- 1999 . ..... ... .. ......... .... ........ .... ....... .. ... .... ......... .... .... .. .... ......... .. ..... ........ 205 
Fig. 5-2. 'I he partitioning of the mid Chesapeake Bay into regions for 
computati on of potential filtrati on by suspension feeders using the approach 
of Gerritsen et al. ( I 994) .. .. .. .... ........... .. ....... ... .......... .. ...... ... ... ...... ... ... ..... ..... ....... 206 
Fig. 5-3. Uni va ri ate pl ots relating macrobenthic biomass in Chesapeake 
Bay to key va riables expected to affect macrobenthic biomass . ................... .. ..... .. 207 
Fig. 5-4. A regression-tree predicting macrobenthic biomass in Chesapeake 
Bay .. .... ........ ... .... ..... ........ ......... ...... ... .. ... .. ... ...................................................... .. 209 
Fig. 5-5. The distribution of macrobenthic biomass among observations 
c lassified into the 27 terminal nodes of the regression tree in Fig. 5-3 ....... ......... .. 210 
XVI 
Fig. 5-6. Observed vs. predicted macrobenthic biomass at sites in mainstem 
hesapeake Bay.??? ?????? ??????? ??? ???? ???? ??????????? ????? ??? ?????? ?? ?????????? ?? ???? ?????? ??? ????????? ?? ???????? 2 11 
Fig. 5-7. A cross-section o f the mainstem Chesapeake Bay indicating the 
maximum depth profile up to 35 m. Depth and axial distance boundaries 
are shown for reg ions identified by a regression tree-model predicting 
212 
macrobenthic biomass from latitude and depth ... .. ...... .. ..... ..... .... ???????? ?????? ?? ?? ?? ??? ??? 
Fig. 5-8. A comparison of three tati tical models predicting 
production/biomas ratios for macrobenthic animals ... ... ...... .. .................... ...... ?? ?? 213 
rig. 5-9. Mean body s ize of macrobenthos in 9 regions of Chesapeake 
Bay computed from the C hesapeake Bay Benthic Monitoring Program data .. .... .. 214 
Fig. 5- 10. The relationship between annua l phytoplankton production 
and macrobenthic biomass is estuaries and coastal systems as shown in a 
review by Herman et al. ( 1999). Three observations have been added 
renecting macrobenthic biomass and annual phytoplankton production 
in three regions of Chesapeake Bay ..... .. .............. ....... ..... .... .... ....... .. ...... .. ..... ...... . 215 
Fig. 6-1. A map of Che apeake Bay showing the boundaries of the 
main stem Bay regions for which trophic flow networks were computed .. ...... .. .. . 299 
Fig. 6-2. A diagram of the basic model structure for the Chesapeake Bay 
trophic flovv networks .... .. ............... .. ....... ..... .... .. .......... ...... ....... ... ... ........... .. ... ... . 300 
Fig.6-3. stimates of organic matter inputs to the summer food webs in the 
three regions of hesapeake Bay .. .... ...... ........ ........ .. .. ..... .... ... .... ...... .................... 302 
Fig. 6-4. The fraction of total respiration due to algae (including 
microphytobenthos), bacteria (including benthic bacteria), meiobenthos and 
macrobenthos, and fish and crabs ....................................... ....... ........................... 303 
Fig. 6-5a. The trophic flow networks for the upper and mid Chesapeake 
Bay projected into a linear chain of canonical trophic levels ........ ... ..... ... .. ........... 304 
Fig. 6-5b. Linear chains of canonical flows for the lower Chesapeake Bay food web 
during sumn1er. .. .... ...... ... ......... ...... .......... .. ...... .... .... ............. .. ..... ........ ...... .... .... 305 
Fig. 6-6. An illustration, for the mid Bay, of how the canonical flows in 
Figs 6-5a,b can be collapsed into a Lindeman Spine ... ........... ...... .. ...... ...... ... ... .... 306 
Fig. 6-7. The organic carbon flow to each trophic successive canonical 
trophic level in the upper, mid and lower Bay food web ..... .. ........ ............ .... ... ..... 307 
XVII 
fig: 6-8. The bottom area present in I m depth increments in the mid Bay 
region and the the penetration of light at current average li ght attenuation 
as we ll as at the c timated reduced light attenuation in the restored mid Bay ... .. .. 308 
Fig 6-9. Trophic transfer effic iencies for canonical trophic levels computed 
from the trophi c now networks for the modern mid Bay and the restored 
Bay .. .... ............................ ..... ... ... ... ..... ......... .... ... .. ........ ........ ..... .......... .. ...... ..... ... 309 
f.i g 6- 10. Total contribution coefficients, indica ting the fraction of net 
production from the indicated producer eventually reaching (a) striped 
bass or (b) benthic bacteria over all pathways .. ...... .. .. ..... ... .... .... .. .... ........ ... ......... 3 I 0 
Fi g. 7- 1. The three general regions of the Chesapeake Bay within which 
co ll ection of organi sms and organic matter were completed for stable isotope 
analys is ...... .... ...... ......... .... ..... .......................... ... ............ ...... ....... ........... ... ...... ... . 326 
Fig. 7-2. Average (?standard error) trophic level (TL) estimates for the 
0-200 ?m plankton size fraction , ctenophores (Mnemiopsis !eidyi) , sea 
4
~ctt_l e ( h,:l'saora quinquecirrha), and bay anchovy (Anchoa milchilii) 327 
unng SJ)nng summer a11d c.a ll ... ........................ .. .......... ................ 
, , Ii ........... . ...... .. 
XV III 
Chapter 1: 
BACKGROUND AND INTRODUCTION 
Coasla l eulrophication is a global-scale problem that has contributed to loss of 
es luarine habitat and possibly to declines in important fish populations. 
Eutrophicalion has been strictly defined as an increase in the rate of supply of organic 
ma rter (Nixon 1995) and can resull from either external inputs of organic matter, or 
more lypica lly, from increased primary production. Eutrophication has sometimes 
been more broadly defined to include both the increase in organic matter supply and 
the myriad ecological changes that result (Cloem 2001). Here the stricter definition is 
preferred, while the associated ecological changes are termed "consequences of 
eut rophicalion" or "eutrophication effects." 
One of the most alarming of these consequences of coastal eutrophication, and 
a major cau e of habitat loss, is the growing incidence of hypoxia and anoxia (Diaz 
and Rosenburg 1995). Hypoxia is defined here in the broadest possible way as a 
conditi on of depressed dissolved oxygen (DO) concentration sufficient to cause any 
adverse ecologica l effect. The severety or intensity of hypoxia refers to the degree to 
Which DO is depressed. Many sessile benthic or epibenthic species are well-adapted 
1 
to moderately depressed DO levels, such as 2.0 mg r (Diaz and Rosenburg 1995). 
1
Hypox 1? a-t o Ie rant be nt ho s can survi?ve D0<2 ?O  mg r , but severe hypoxia (i.e. DO<O ?5  
mo 1-1) . only eliminates all metazoan life (Diaz and 
b or complete anox1a comm 
Ros en b urg . ? uch as fish and crabs can often (but not always) 1995). Motile species s 
. ti less well adapted to hypoxia. Thus, subtle 
escape hypoxic habitats and are appaien Y 
1 
behavioral effects can sometimes be observed at DO levels only slightly below 
ve s easily 
saturation (Breitburg et al. 1997), whi le acute effects can occur at DO le I . 
1 exposure 
adequate for hypoxia-tolerant taxa. Aside from direct effects associated w?th 
to hypoxia, biota may be adversely affected by the loss of use of the affected habitat 
and the prey that may have been present within it. The consequences may be 
increased if the degraded habitat was prefeffed because it provided refuge from 
predation , opti mal temperature, or other benefits. 
The possible loss of hab itat and prey resources associated with hypoxia is a 
maJor reason that one may associate eutrophication and hypoxia with the potential for 
decreased fisheries production and yield. Another related mechanism, not emphasized 
in this study, is loss of vegetated habitats associated with eutrophication (Kemp et al. 
19 8 3). Lacking effects of thi s nature, there is no a priori basis for assuming that 
eutrophication will have any adverse effect on fisheries yield. Three hypothesized 
relationships between total organic input and combined fisheries yield illustrate some 
poss i bi Ii ti es (Fig. 1-1). The type 'a' response en tails a Ii near increase in secondary 
production with primary production, implying constant or nearly constant trophic 
efficiency. This may be the most commonly reported relationship in the literature. 
Ni xon (1982) observed a positive relationship between primary production and 
fi sheiies yie ld, where primary production varied from -50 gC m?' i' to nearly 1000 
gC m?' i'. Iverson (1990) observed a very strong correlation (r' ~0.96) between 
phytoplankton production and fish plus squid yield in JO open-ocean and coastal 
ecosystems with primary production between 50 and 250 gC m?' /'. Finally, 
as . h . ? an indicator of the primary productivity of lakes Hanson 
suming total phosp 01 us JS ? 
2 
and Leggett ( J 982) also fo und that fi sh yield in lakes was positi vely correlated with 
primary production . There is ample reason to believe that type 'b ' and type 'c' 
functional relationships can and do occur (Fig. 1-1). The initial decrease in the slope 
of lines 'b ' and 'c ' indicate that beyond some level of primary production or organic 
input, one or more more of many possible alternative controls begin to become more 
important (Fi g. 1-2). Examples inc lude changes in food quality, limiting food supply 
at some other point in time (i.e. "bottlenecks"), availability of suitable habitat or 
substrate, and control by predators . A decrease in the slope of 'c ' to less than zero 
( ig. 1-1) implies that further increases in organic supply rate decrease the production 
potential of the ecosystem. The best documented examples of this type of response 
have been related to macrobenthic "succession" along a gradient in time or space of 
organic enrichment (Pearson and Rosenburg 1978). The species-abuundance-biomass 
(SAB) model of Pearson and Rosenburg (1978) describes the general nature of 
changes in these three properties of benthic communities with distance from a source 
of enrichment. Closest to the source, where emichment is greatest, all three properties 
decline to zero, ostensibly due to oxygen stress. 
Resolving eutrophication effects on fisheries production in whole systems may 
be hard to be resolve because fishe1ies production has sometimes, or perhaps more 
often than not, been degraded by overfishing before the most dramatic effects of 
eutrophication became evident (e.g. Jackson et al. 2001, but see Boesch et. al. 2001). 
Regardless of the primacy of fishing versus eutrophication effects on fisheries, fishing 
pressure can dramatically affect fish populations, giving managers good reason to 
account for, if not emphasize, population dynamics of targeted populations. 
3 
Nonetheless, there is growi ng recognition that fish populations and the 
fi sheries they support must be managed in a way that accounts for the 
interconnectedness of li ving systems. This idea underlies the growing emphasis on 
"eco yslem-based fisheries management." In the same vein, it has often been im p J1' e d , 
if not clearly staled, that protection and restoration of fisheries is an important 
objective of waler quality and ecosystem restoration. A good example is the recently 
signed agreement to continue restoration of Chesapeake Bay, known as "Chesapeake 
2000" (EPA 2000). This new emphasis increases the need for scientific models and 
data effectively characterizing linkages between primary production, water quality and 
secondary production. 
The environmental history of Chesapeake Bay over the past 100 years has seen 
dramatic changes in both fishe1ies (e.g. Houde et al. 1999) and water quality (e.g. 
Harding and Perry 1997), and may be a classic illustration of the need for integrated 
ana lysis of ecological change. This dissertation research developed along those Jines. 
Chapter 2 exami nes the underlying physical transport regime which is central to the 
ecology of estuaiies in general and Chesapeake Bay in particular. The resulting 
e ti mates of physical transport characteristics, while useful in their own right, enter 
ana lyses in each of the subsequent chapters. 
Chapter 3 examines the 50-year record of dissolved oxygen data for the Bay, 
establi shing the extent of hypoxia today and the nature of its increase since 1950. 
Further, the physical and biological controls on hypoxia are examined, suggesting 
prospects for restoration. 
4 
hapter 4 quantifies the deposition of organic matter resulting from the spring 
phytop lankton bloom, a major ecological feature of the Bay. This process has long 
been connected to the formation and maintenance of hypoxia in summer, as well as the 
sustenance of macrobenthic production. The relationship between freshwater inputs 
during spring and deposition of phytoplankton to sedi ments is explored and quantified, 
d irect ly exam ining a process implic it in the river now-hypoxia relationships identified 
1n hapter3. 
hapter 5 foc uses on the macrobenthic communiti es of the Chesapeake Bay. 
These communities may have changed more than any other in the Bay, first due to 
ma sive removal of oyster stocks by fishing and secondly by the two-fold insult of 
excess ive sedimentation and hypoxia. The patterns of macrobenthi c biomass in the 
Bay are expl ored in relation to habitat factors utili zing a novel tree-structured data 
ana lysis approach. Tree-structured data analysis is effective for uncovering hidden 
relationships in large data sets for which responses to a large number of explanatory 
variables a re hi ghly non-linear or discontinuous and where interactions among 
vari ab les are complex. The application to Chesapeake Bay illustrates where and when 
hypox ia affects the benthos, as well as the effect of other properties such as sediment 
type. Chapter 5 also begins to explore the connections between primary and 
secondary production in Chesapeake Bay. A question first raised by simulation 
mode ling results, namely whether decreased primary production could actually reduce 
secondary production in the benthos , is examined using a bioenergetic budgeting 
approach. These computations, as well as those from all previous chapters were 
utili zed in Chapter 6 . 
5 
hapter 6 presents the results of a trophi c network modeling acti vity which 
examines how, or if, secondary production in Chesapeake Bay could be sustained if 
primary production is reduced substantiall y through environmental management. 
Here questi ons re lated to trophic transfer efficiency, alternative carbon fl ow path ways, 
the effect of gel atinous zooplankton blooms, and bacterial domination of the 
ecosystem are brought together. Hi storical data, comparative ecological model s, and 
mass balance arguments are combined to envision how carbon flow in a "restored" 
middle Che apeake Bay might be different from today, and as importantly, how much 
fi sh production it might be able to sustain. 
Chapter 7 serves to some extent as an epilogue to Chapter 6, examining the 
trophic level estimates derived from the trophic network models using analysis of 
stable nitrogen isotopes. Chapter 8 provides a brief synopsis of the major results of all 
the chapters and suggests future directions for related research. 
In total , the results of this research connect nutrient-driven eutrophication 
e ffects , mediated by physical processes, to trophic transfer efficiency and potential 
fi sh production both today and in a restored Chesapeake Bay. The results suggest that 
nutrient loading rate reductions should lead to improved water quality, specifically a 
decrease in the extent and intensity of summer hypoxia and anoxia. These changes 
wi II result from a substantial decrease in phytoplankton production. Total organic 
input wou ld likely decrease by a smaller fraction due to an offsetting increase in 
benthic primary production. Improved habitat quality should lead to increased 
biomass and production of macrobenthos in the mid Bay region. This and other 
ecological changes that can be expected should lead to decreased carbon flow through 
6 
both planktonic and benthic bacteria. Although it is suggested that the restored Bay 
wi ll have lower levels of primary production than the modern Bay, more, not less, fish 
production could be supported by the restored Bay. 
The objectives of ecosystem management and restoration (e.g. EPA 2000) 
require that ecosystem scientists begin to establi sh linkages across the trophic levels of 
the ecosystem and across the traditi ona ll y separate disciplines of science and 
management. Drawing on decades of research that makes such work possible, this 
dissertation makes a step in that direction. 
Literature Cited 
Boesch D ., E. Burreson, W. Denni son, E. Houde,~? M. Kemp, V. Kennedy, R. 
Newell, K. Paynter, R. Orth , and R. Ulanow1cz . . 2001. Factors in the decline of 
coastal ecosystems. Science. 293 (5535): 1589-1590. 
Breitburg, D . L., T. Loher, C. A. Pace~, ~nd A. ?erSt~in. 1997. Varying effects of 
low di ssolved oxygen on trophic mteract1ons m an estuarine food web. Ecol. 
Monogr. 67(4): 489-507. 
Cloern, J.E. 2001. Our evolving conceptual model of the coastal eutrophication 
problem. Mar. Ecol. Prog. Ser. 210: 223-253. 
EPA. 2000. Chesapeake Bay 2000 Agreement. US EPA Chesapeake Bay Program, 
Annapolis, MD. 
Harding, L. w., Jr. and E. S. Perry. 1997. Long-term increase of phytoplankto5n2 
biomass in Chesapeake Bay, 1950-94. Mar. Ecol. Prog. Ser. 157: 39_ . 
Hanson, J.M. and w. c. Leggett. )982. Empirical prediction of fish biomass and 
yield. Can. J. Fish. Aquat. Sci. 39: 257-263. 
7 
I Toudc, E. D., S. Jukic-PeJad ic, S. B. Brandt and S. D. Leach. 1999. Fisheries: Trends 
in ca tches, abundance and management, pp. 341-366. In : T. C. Malone, A. 
Malej , L. W. Harding, Jr., .N. Smodl.aka, and R. E. Turner (eds .) Ecosystems at 
the Land-Sea Margin, D ra mage Bas m to Coasta l Sea. American Geophysical 
Union, Washington , DC. 
Iverson, R. L. J 990. Control of marine fish production. limnol. Oceanogr. 35(7): 
1593-1604. 
Jackson J . B . c. , M. X. Kirby, W. H. Berger, K. A. Bjorndal, L. W. Botsford, B. J . 
Bourque, R. H. Bradbury, R. Co_oke, J. Erlandson,~- A. Estes, T. P . Hughes , S. 
Kidwell , C. B. Lange, H. S. Lemhan , J . M. Pandolfi , C.H. Peterson, R. S. 
Steneck, M. J. Tegner and R.R . Warner _- 2001. Historical overfishing and the 
recent collapse of coastal ecosys tems. Sczence. 293 (5530): 629-638. 
Nixon , s. w. 1982. Nutrient dynamics, primary production and fisheries yields of 
lagoons. Oceanol. acta. SP: 357-37 l 
Nixon, s. w. 1995. Coastal marine eutrophication: a definition, social causes and 
future concerns. Ophelia. 41: 199-219. 
Pearson, T. H. and R. Rosenburg. 1978. Macrobenthic succession in relation to 
organic enrichment and pollution of the marine environment. Ocean. Mar. 
Biol. Ann. Rev. 16: 229-311. 
8 
b 
C 
Total Organic Input 
Fig. 1-l. Hypothesized functional relationships between 
total organic input (primary production+external inputs) and 
fisheries production. For a type 'a' response, fish produc-
tion is a constant fraction of organic input. A type 'b' 
response involves saturating kinetics and would could be 
expected when secondary producers are unable to utilize 
additional inputs, but are not adversely affected by excess 
inputs. A type 'c' response indicates that excess organic 
inputs caus adverse eco_Iogical cha~ges such as habitat deg-
radation that decrease fish product10n. 
9 
Eutrophication 
Organic Inputs Effects I 
("Food Supply") r---------: 
Food Quality 
Competition and 
Predation 
Trophic Habitat Quality 
Transfer Efficiency 
Recruitment and 
Mortality 
Other factors ... 
Fisheries Yield 
Effects of Fishing 
Figure 1-2. A conceptual diagram illustrating mediation of the trans-
fer of organic carbon from inputs and primary producers to fisheries 
yields. Solid lines indicate carbon flow. Dashed lines indicate influ-
ence or " information" flow. The li st of effects is not intended to be 
comprehensive. 
10 
Chapter 2: 
A BOX MODEL APPROACH FOR ESTIMATING PHYSICAL TRANSPORTS 
AND EXCHANGES FOR THE CHESAPEAKE BAY 
Abstract 
A sa lt and water balance model , or "box" model , was developed to estimate the 
average hydraulic transport within Chesapeake Bay at a monthly time interval. The 
C hesapeake Bay system is a large complex of estuari es and river basins. Strong 
seasonal patterns in freshwater inputs lead to seasonal changes in salinity in the Bay 
itse lf, as well as its large tributary estuaries and f1inging embayments. Therefore, 
implementing a box model for thi s system required that multiple freshwater sources 
and latera l sa lt exchanges be included in the model. The latter has not previously been 
considered in a box model. Estimated down-estuary transport was l to 3 times the 
discharge of the Susquehanna River into the oligohaline region of the Bay. In the 
mesohaline region , transport was 4 to 6 times the sum of up-stream freshw ater inputs. 
In the pol yhaline region of the Bay, down-estuary transport was 6 to 10 times total 
freshwater inputs. The distribution of vertical advection featured an area where 
seasonal downwe lling occurs, in agreement with recent hydrodynamic modeling 
results and fi e ld observations. Vertical exchange velocity, a measure of non-advective 
vertical exchange, was lowest in the mesohaline region of the Bay, especially during 
summer when exchange velocity was 10-20 cm ct?' versus 60-90 cm ct?' in late fall. An 
index of water column stratification, based on the Brundt-Vaisala frequency , revealed 
a strong positive con-elation between the average strength of water column 
stratification and the average spring discharge of the Susquehanna River (r2=0.92). 
11 
stimates of summer average vertical diffusion decreased with freshwater input and 
strati fication, but other factors, potentially wind-driven lateral seiching, contributed 
significantly to vertical mixing. Landward advection into the mid Bay bottom layer 
increased wi th piing river flow, but this effect was attenuated in the up estuary 
di rection. 
This box model approach appears to be an effective and re latively simple tool 
for estimating time-averaged transport rates among regions of the Bay and between 
the surfa ce and bottom layers, enabling many potential applications in ecology, 
environmental chemistry, and environmental management. 
Introduction 
Phy ical tran port processes are a central feature of estuarine ecosystem 
dynamics. In fac t, many biological and biogeochemical processes in estuaries would 
be di fficult to understand except in the context of the physical transport regime. While 
the periodic ebb and flow of the tides is the most obvious feature of the circulation to 
the casual observer, much less obvious features may be as important. For example, 
net horizontal transport, whether the result of tidal dispersion or residual (sub-tidal) 
advection is responsible for transport of sediments, nutrients, organic matter, and 
plankton in both the up-estuary and down-estuary direction. In stratified estua1ies, 
rates of cross-pycnocline exchange, whether vi a upwelling or turbulent diffusion, 
determine nutrient and oxygen transport across the pycnocline, affecting 
phytoplankton dynamics in the surface layer and the potential for hypoxia in the lower 
water column . 
12 
A variety of approaches are available fo r estimating transport within estuari es 
and for eva luati ng the possib le ro le of physical transport in determining ecologica l or 
biogeochemica l observations. These include mixing di agrams (e.g. Fi sher et al. 1998), 
hydrodynamic simu lation models (Johnson et al. 199 1, Hood et al. 1999, Wang and 
Joh nson 2000) , direct cuITent measurements and drifters (Va lle-Levinson and Lwiza 
1995 , Janzen and Wong 1998), and box models (Pritchard 1969, Officer 1980, Hagy et 
a l. 2000) . D ue to the potential for substantial changes in cun ent regimes over short 
periods of ti me, as we ll as the hi gh degree of spatial vari abi lity in cunent velocity (e.g. 
E lli ott 1978), cuJTent meters are not li kely to be an effective means of resolving 
average residual cuITents at the whole estuary scale. Hydrodynamic simulations 
provide the most mechani stic, spatial and temporal detail. However, thi s comes at the 
cost of enormous data requi rements and computational complexity. Moreover, an 
addi tional consideration applies if the objecti ve is to estimate sub-tidal vertical and 
horizonta l cuJTent ve locity and vertical di ffusive transport at aggregated temporal and 
spatial sca les . The hi gh-frequency (i.e. minutes to hours) vari ations in tidal current 
ve loc ities are the most certain predicti ons of hydrodynamic models, while residual 
cuITent velocities are the least certain. This results from the fac t that the tidal cuITents 
can largely be predicted by a barotropic model (i.e. verticall y integrated), while the 
res idual currents require a much more complicated baroclinic model. Accurate 
predictions of residual circul ation therefore depend on careful calibration of diffusivity 
parameters until the mode l reasonably replicates observed salinity. Although thi s is a 
tractable problem, significant errors are possible and may be obscured by the 
otherwise impress ive reproduction of tidal circulation . Careful work by skill ed 
13 
investi gator is needed to obtain good estimates. Therefore, another approach for 
estimating resi dual ci rcu lation, particularly a relatively simple approach, may also be 
useful. 
Pritchard (1969) introduced box models as an approach for estimating residual 
c irculation in estuaries using salinity as a natural conservative tracer of physical 
transport. Further e laborated upon by Officer (1980) and later by Hagy et a l. (2000), 
the method involves dividing the estuary into a seri es of finite elements, or "boxes," 
which are assumed to be well-mixed. Two equations can be written for each box . 
One describes the balance of water flo w into and out of the box , while the other 
desc ribes the balance of salt transport into and out of the box. Provided that a few 
simplifying ass umptions can be accepted (see "Methods"), a unique simultaneous 
so lution to this system of linear equations can be obtained, providing estimates of the 
physical transport coefficients (Hagy et al. 2000). These estimates can then be readily 
utili zed for ecological investigation . For example, Taft et al. (1978) used a box model 
based on the Officer (1980) methodology to infer the importance of ammonium 
recycling in the lower water column for the nitrogen requirements of the summer 
phytoplankton community in Chesapeake Bay. 
Hagy et al. (2000) described two simple elaborations of the basic box modeling 
methodology (i.e. described by Officer 1980) that enabled these relatively simple 
model s to reproduce surprisingly realistic features of the monthly average physical 
transport regime of Patuxent River, MD. Specifically, these elaborations involved 
relaxing two key model assumptions . Whereas Officer (1980) assumed that all 
freshwater entered the estuary at a single point, Hagy et al. (2000) generalized the 
14 
approach uch that any quantified input of freshwater could be accommodated. The 
e ffec t of thi s genera li zation on model estimates for Patuxent River estuary was not 
particularly large since Patuxent River is the only major freshwater source. However, 
the Susquehanna River account for on ly about 50% of freshwater inputs to 
Chesapeake Bay (Bue 1968), with the remainder coming from several other major 
tributary rivers. Thus, this was expected to be important for a Chesapeake Bay box 
model. The second, and perhaps more important, generalization described by Hagy et 
al. (2000) wa the provision for seasonal changes in salinity. This proved to be very 
important for estimating residence times for Patuxent River estuary. Specifically, if 
sa linity was increasing on a seasonal basis, a model neglecting the increase 
substantially overestimated residence time. The difference reflects the Jack of direct 
coITespondence at seasonal time scales between freshwater inflow and seawater inflow 
due to grav itational circulation. Since salinity also changes seasonally in Chesapeake 
Bay, this generalization is probably also important for a Chesapeake Bay box model. 
This paper describes a 9-segment, 2-dimensional box model developed to 
estimate hydraulic transpo11 in the mainstem of Chesapeake Bay. The model 
incorporates the features described by Hagy et al. (2000), plus additional elaboration 
that was included to improve applicability to Chesapeake Bay, a large, physically 
complex estuary adjoined by substantial tributary estua1ies. The intended use of this 
model was to compute circulation in a manner sufficient to estimate regional and 
seasonal scale input-export budgets of major dissolved and particulate materials. 
These include, but are not limited to, organic carbon, dissolved oxygen, nitrogen, 
phosphorus, and silicate (Hagy 1996). The highest possible spatial and temporal 
15 
arti cu lati on was utilized to minimize the etTor that can result from overly aggregated 
model of this type (Webster et al. 2000). 
Study Area 
C hesapeake Bay (the Bay) is a large, partially-stratified, estuary. In addition, 
the Bay is part of a major complex of ri ver basins and associated estuaries and 
cmbayments (Fig. 2- 1). There are several features of the Bay and associated 
tributaiies that are important for the construction of box models. The circulation is 
mostly two-layer, with net circulation flowing landward in the bottom layer and 
eaward in the suiface later (Pritchard 1952). The depth of no net motion intersects 
the bottom in the oligohaline region of the estuary (Wang and Chao 1996). Landward 
of thi s point, there is a one-layer circulation with net seaward flow at all depths. In the 
polyhaline region of the Bay, there are stronger lateral (i.e. normal to main Bay ax is) 
gradients in salinity and cuITent velocity compared to elsewhere in the Bay (P1itchard 
1952, Valle-Levinson and Lwiza 1995). 
Over the length of the estuary, 7 major tributary estuaries discharge freshwater 
into Chesapeake Bay. These include the Susquehanna, Choptank, Patuxent, Potomac, 
Rappahannock, York and James Rivers (Fig. 2-1). As important is the fact that, as 
illustrated for Patuxent River by Hagy et al. (2000), these estuaries exchange salt with 
the mainstem Bay during seasonal cycles of salt accumulation and depletion . Several 
large embayments are also present, including Tangier Sound, Pocomoke Sound and 
Eastern Bay. W hile these do not have substantial sources of freshwater input 
compared to the major rivers , they do exchange salt. The combined volumes of the 
16 
mainstem Bay, major t1ibutary ri vers, and major embayments inc lude 91%  of the total 
volume of the Chesapeake Bay system. The remaining water volume is contained 
within many smaller embayments and tributary rivers (Cronin and Pritchard 1975). 
Methods 
This tudy uses the salt and water balance-based box model approach , 
originally proposed by Pritchard (1969), elaborated upon by Officer (1980), and 
further deve loped by Hagy et al. (2000). In applying a box model , the estuary is 
divided into a series of discrete regions or "boxes," each of which is assumed to be 
well -mixed. In the simplest example of a box model for an estuary, a quantified 
freshwater flow, Q1; enters a well -mixed basin of volume, V (Fig. 2-2, upper panel). 
Estuarine water of known average salinity, s1,,, flows out of the basin into the sea at 
rate Q, while the estuary exchanges water at rate Evia non-advective dispersion with 
the coastal zone where the known salinity is Sout (Fig 2-2). For this simple case, the 
water balance equation is 
dV 
-=Q1 - Q (1) 
dt 
The change in estuarine water volume, V, over time is dV /dt, which , averaged over 
many tidal cycles, is assumed to be zero. Therefore, the seaward advection is simply 
equal to the freshwater input. 
The salt balance equation is 
V-dsi-11= E ( ) s o111 -sin - Qsi11 (2) 
dt 
17 
Substituting Q1 for Q ba ed on the water balance equation , one can so lve for E in 
terms of known values, giving 
V clsi11 , 
+Q.sill 
E = cit (3) 
(so111 - sin) 
Although simple , this basic formulation is not entire ly unrealistic, and can 
yie ld u eful results (e.g. Gordon et al. 1996). For a horizontally and vertically 
articulated box model, the approach described by Hagy et al. (2000) applies. Namely, 
the water budget for a surface layer box "m" is 
clV 
cl/ = Qlll - l + QV/11 + Q /111 - Q/11 (4) 
where Q11, is the seaward advection out of the box, Q111-1 is the seaward advection into 
the box , Q11111 is the vertical advection into the box , and QJ111 is the freshwater input to 
the box. The salt budget is 
V ds Ill --..c.:.i.t.! ll. = Q111 _,s111 - I + Q\111/sllI/  - QI/I sI ll + E 
(, ) 
VIII s Ill -s I ll + (5) 
[ ?111 - 1.111 (s111 -1- sm )+ ?111.111 +1( s111 +1 - Sm)] 
where the Q te rms are as above and V111 is the box volume, s111 is the average salinity in 
the box, s,,,_, is the average salini ty in the adjacent landward box , Sm+! is the average 
sa linity in the adjacent seaward box, and s; is the average salinity in the bottom layer 
11 
box in the same model segment. The terms ? 111 _,. 11 , and ? 111,111+1 are the horizontal 
dispersive exchange with the adjacent landward and seaward boxes, respectively, 
whi le Ev,,, is the vertical non-advective exchange across the pycnocline. In order to 
obtain a fully determined linear system of equations, where a unique solution can be 
18 
obtained, it is necessary to have no more than two unknown terms per box. Hagy et 
a l. (2000) reasoned usi ng arguments based on the estimated mass-transfer Peclet 
number that, for Patuxent Ri ver estuary, horizontal salt-transport due to ho1izontal 
di spersion would be quantitatively insignificant compared to ho1i zontal advection. 
The assumption was justifi ed in the same way for Chesapeake Bay. Therefore, the 
bracketed terms in eq . (4) were neglected. Using the approach of Hagy et a l. (2000), a 
c losed fo rm soluti on to the linear system of equation can be obtained by combining 
the sa lt and water balance equations. Namely, the water bal ance for the boundary 
between segment m and segment m+ 1 is 
Ill 
Q:11 +1 = Q111 - LQJJ (6) 
j = I 
Ill 
where L Q JJ is the sum of all freshwater inputs landward of and including box m. 
j = I 
The con esponding salt balance equation is 
Ill V d Ill d' ~ Sm ~ , Sm , , Q 
~ m d + ~ Vm d = Q11, + IS111+l - (7) 111 S 111 
j = I t j = I t 
where the summation terms refer to the combined change in salt storage in a ll surface 
layer and bottom layer boxes landward of segment m. Substituting the right hand side 
of eq. (6) for Q; + 1 in eq. (7) and solving for Q111, one obtains 11
(8) 
19 
wh ich is the same as eq. (7) in Hagy et al. (2000). The general similaiity to eq. (3) 
should a lso be apparent, except that ho1izontal advection replaces horizontal diffusive 
exchange and aggregated freshwater input and salt storage replaces the single-box 
values. Solving eq . (8) for all Q111 , then substituting into eq. (6), one can compute 
landward advection at each segment. Substituting Q111 and Q:,1i nto eq. (4) gives 
upwelling advection. Finally, given all the estimates of advection, one can compute 
E v111 from eq. (5). 
The box model divides Chesapeake Bay into 9 segments of unequal length 
(Fig. 2- L, Fig. 2-3. Table 2-1). The most landward segment includes only net seaward 
nows and is assumed to be vertically well mixed. All other model segments are 
ve11ically divided into a slllface layer and a bottom layer, each assumed to be well 
mixed vertically and horizontally. Net advection in the surface layer is down-estuary, 
while the bottom layer has net up-estuary advection. Vertical advection (i.e. 
upwelling or downwelling) exchanges water between the layers. The transition from 
one to two layers occurs at the boundary between segment l and 2. In segment 1, 
landward-flowing water from the bottom layer of segment 2 mixes with freshwater 
from the Susquehanna River, then enters the seaward-directed flow from segment l to 
segment 2. This seaward flow enters only the surface layer box in segment 2. 
Freshwater inputs enter each surface layer box as determined by the water budget. 
Additional terms were needed to account for changes in salt storage in the 
major tributary estua1ies and in the larger embayments flanking the mainstem Bay 
(Fig. 2-1, Fig. 2-2). These water bodies exchange water and salt with the mainstem via 
tidal and other mechanisms. Most rivers have either a shallow sill (<10 m) or a vary 
20 
narrow, bul s li ghtl y deeper, access channel (Fig. 2-1) . An exception is the Potomac 
River, which has a wider and deeper access channel. However, this channel is only 
12- 15 m deep, simi lar lo the pycnocline depth in the mainstem Bay. It was therefore 
ass umed that salt and water exchanges with tributaries involve only the surface layer 
of the mainstem Bay, regardless of the transport regime at the tributary mouth. 
Therefore , these exchanges do not affect the water balance equations for the Bay 
model except as a net water outflow equal to the freshwater input to the t,ibutary. 
Such freshwater inputs were already accommodated in the model formulation of Hagy 
et a l. (2000). In contrast, salt exchanges with tributary rivers and embayments appear 
as add itional sa lt storage terms added to eq. (5), giving 
V ds,111 + V -ds11-1 = Q1 /111 -- 11 - IS111 - I + Q s' - Q s + E (s' - s ) + VIII Ill Ill Ill VIII Ill Ill 
dt (9) di Ill 
[E,,, _,.111 (s111 -1- s111 )+E111.111+1(s111 +1- sm )] 
where y1111 is the volume of the tributaries or embayments adjacent to box m up to the 
limit of sa lt intrusion and ds1111 /dt is the change in the average salinity in that volume. 
The e terms can also be collected as in eq (7) giving 
,,, ds II/ d I 
~ V _!!!J_ + ,,, ds ~ , s111 - Q' , ~ V,11 ~ + L.J V,11 -d - m+IS111+I - Q,,, s111 (10) L.J /Ill di ~ dt ?= 1 t 
J= I ; = I J 
which can be combined with eq (6) to update eq. (8). The result is 
s:,, +1[fQJJ +Qr]+[tv1 -: -/ + tv;-:-/ + !V,,,, d:"] 
j I ; - 1 J 
Q,,, = L _ _b_~'.___ __~ __!~-s-:;:,:--,+-1-_--::s-111------------=.L (11) 
21 
The tributary basins and embayments that were included in the model are 
Eastern Bay (embayment), Choptank River, Patuxent River, Potomac River, 
Rappahannock River, Tangier Sound (embayment), Pocomoke Sound (embayment), 
York River and James River (Table 2-3). The mainstem Bay accounts for 67% of the 
total volume of Chesapeake Bay and ttibutaries. The additional tributaries and 
embayments account for an additional 24%. Thus, 91 % of the Chesapeake Bay 
system vo lume is inc luded (Table 3). Smaller tributaries and embayments comprise 
the remaining 9% of the volume and were neglected. 
Ve11ical non-advective exchanges (Ev111) can be calculated equivalently from 
eithe r surface layer or bottom layer salt-balance equations once all the other exchanges 
are known. In this case, it was simpler to compute Ev111 from the salt-balance equations 
for the bottom layer boxes. The equations were w1itten to be conditional upon the 
s ign of Qv,,, due to the known occun-ence of downwelling advection (i.e. Qv111<0) in 
Chesapeake Bay in some places at some times of the year (Chao and Paluszkiewicz 
1991). For the case where Qv111>0 the salt balance equation can be solved for Ev111 to 
obtain 
/ dds;,t,  Q' / Q I Q' I - VIII + 111 +1S 111+l - v111S111 - 111S111 
E v111 = _,,__ _____ _, _________ _,__ (12) 
Sm -Sm 
When Qv111<0 (i.e. the downwelling case), the salt balance equations changed such that 
vertical advection becomes a source of salt to the bottom layer rather than a loss. The 
revised sa lt balance is 
V, dds ;t11  = Q, , Q Q, , E ( , ) 111 111+1S111+1 - v111s111 - 111s111 - v111 s111 - s 111 (13) 
22 
which can be so lved for Ev111 to obtain 
-V / ds ;/I + QI I Q QI I -- 111 + 1S111+ 1 - v111s111 - 111s111 
111 dt (14) 
EV/II = I 
S Iii - S111 
Since the sign of Qvm changes when downwelling replaces upwelling, it was 
unnecessary to change the sign for the Qv111 term in the salt balance equation (eq. 3). 
The onl y difference between eq. 12 and eq. 14 is that Sm replaces s;,, in the term 
representing the verti ca l advecti ve sa lt flux due to the fact that downwelling water has 
the salt content of the surface layer water, while upwelling water has the salt content 
of bottom layer water. 
Sources of Data and Computation of Inputs 
Salinity data were obtained from the Chesapeake Bay Water Quality 
Monitoring Program, which collected water quality data at a series of 20 stations down 
the central axis of the Chesapeake Bay (Fig. 2-1, Table 2-2). Cruises were conducted 
on a bi -weekly basi s, except in winter when cruises were conducted monthly. 
Sampling began in 1984 and is ongoing at present. The vertical resolution of the data 
is l m. Average salinity for the tributaries was computed using the available 
Chesapeake Bay Water Quality Monitoring Program data at stations indicated in Table 
2-2. Both mainstem and tributary salinity data were interpolated to a grid transecting 
the estuary at a vertical resolution of one meter and a horizontal resolution of one 
nautical mile. Grid cells corresponded to tabulated cross-sectional volumes (Cronin 
23 
and Pritchard 1975), which were used to compute volume-weighted average salinity 
(Hagy e t al. 2000). 
T he major t1ibutary rivers are the dominant sources of freshwater to 
Chesapeake Bay (Table 2-3). These include (north to south) the Susquehanna, 
Choptank , Patuxent, Potomac, Rappahannock, York and James. These rivers are all 
gauged at their fa ll lines (Fig 2-1) and data discharge totals are available from the 
USGS . Data for 1984-1998 were obtained and averaged by month for this study. T he 
York Ri ver includes the combined flows of the Mattaponi and Pamunkey Rivers. T he 
tidal James River combines the James and Appomattox Rivers (Fig. 2-1). Together, 
the drainage basins of the gauges at the fa ll lines of these rivers compri se 130,798 
2
km , or 78% of the 166,717 km2 drainage area of Chesapeake Bay (EPA 1998, Table 
2-3) . A lthough there are many small watersheds that are gauged, none were 
s ignifi cant in size and most have only short or incomplete flow records. Thus, 
ungauged portions of the watersheds of the major tributaries were estimated on the 
bas is of the flow/area of the associated tributary. The ungauged areas of the watershed 
of the major tributaries tota led 16,992 km2 (Table 2-3) . Other ungauged areas were 
combined into four groups, the Eastern Shore in Pennsylvania, Delaware, Maryland 
and Virginia, the western shore in Pennsylvania and Maryland, the western shore in 
Virginia, and open water (Table 2-3). It was assumed that direct precipitation to the 
water surface was approximately bal anced by evaporation . Therefore, open water was 
assumed to contribute no net freshwater input (Bue 1968, Hagy et al. 2000). The flow 
for the other areas was based on the flow/area from nearby gauged basins. The basins 
that were used for this purpose included the Choptank River (Eastern Shore), Patuxent 
24 
River (MD we tern shore), and Mattaponi River (VA western shore). These ungauged 
areas totaled 16,866 km2. In total the water budget accounted for 99% of the entire 
watershed (Table 2-3). While estimates of flow for ungauged areas based on 
flow/area in nearby ba ins are not as certain as direct observations, etTors would have 
to be large to s ignificantly affect the results because such a large fraction of the 
di scharge from the watershed (78%) is directly gauged. 
At the large sca le of this analysis, human diversions of freshwater for purposes 
such as munic ipal or industri a l use would have to be very large to be of any 
significance in the overall water budget. Moreover, because of the size of Chesapeake 
Bay, w ithdrawals are balanced by returns to nearby areas of the estuary (Bue 1968). 
In the case of smaller estuari es such as Tomales Bay (Smith et al. 1991) and Patuxent 
Ri ver (Hagy et a l. 2000), these a lterations must receive careful attention. A possible 
exception is the estimated 27 m3 s-1 diversion of freshwater from the Chesapeake Bay 
to the De laware Bay due to net transport through the Chesapeake and Delaware Canal 
(Bue 1968, W ang and Johnson 1999). This flow was subtracted from the freshwater 
inputs to segment 1. 
Scale-Independent Transport Estimates 
Each of the estimated bulk transport estimates (e.g. units = m3 s-1) generated by 
box models can be expressed as a scaled quantity that can be more readily interpreted 
and compared to other data. In particular, such scaled quantities are useful for 
comparing estimates for diffe rent regions of the Bay. For example, the estimated 
advection (Q111; m3 s-1) divided by the average cross sectional area (m2) of the mode l 
25 
segment yields a quantity that approximates the spatially averaged residual (i .e. net 
1
non -tidal) CUJTent velocity (m s- ) over the cross section. Similarly, the upwelling 
transport (Q 11111 ) divided by the pycnocline area estimates spatia ll y averaged upwelling 
ve locity. Vertica l non -advective exchange (Ev,,i), scaled by the pycnocline area c? 
i.e. 
Ev111 /A), gives a quantity termed the non-advective exchange velocity. Multiplying 
by the he ight of the pycnocline region gives an estimate of vertical eddy diffusivity 
(DJ. This conversion can be illustrated by writing the analogous express ions for the 
sa lt flux across the pycnocline in terms of both the box model coefficient, vertical 
diffusivity, and a di screte approximation based on vertical diffusivity. This is 
I ) D as - AD (s;,, - SIil ) 
Ell/II ( SIii - SIii = A z az - z ,1z (15) 
where A = the pycnocline area and L'.lz =t he mean height between box centers. 
Rean ang ing gives the conversion from the non-advective exchange coefficient to 
diffusivity or D = '1zE 11111 A ? In the mesohaline Chesapeake, the mean distance fro , z m 
surface and bottom layer box centers (L'.lz) is - 9 m. Table 4 provides the physical 
dimensions of the boxes needed to compute physical transport rates in these 
a lternative units. 
Validation Data Sets 
Two data sets were used to provide independent validation of box model 
results . The Chesapeake Bay Observing System (www.cbos.org) is an an-ay of 
permanent moori.n gs located in Chesapeake Bay that provide telemetered physica, l 
26 
oceanographic measurements both in real time and archived data sets. A multi-year 
record of surface layer (2.5 m) and bottom layer (18.5 m) current velocity collected at 
a frequency of 12 h( 1 was obtained and used to compute average residual current 
velocity in that region of the mid Bay. 
The other validation data set was obtained from the National Science 
Foundation-Funded Chesapeake Bay Land-Margin Ecosystem Research Program, 
which collected physical oceanographic data along transects of the mainstem 
Chesapeake Bay at extremely hi gh spatial resolution (8 Hz) using a towed, undulating 
CTD device (SCANFISH). Temperature and salinity data collected by SCANFISH 
along a transect between the Potomac and Rappahannock Rivers were used to 
compute and map the Brundt-Vaisala frequency (N2), a measure of water column 
stability, which can be computed from the ve1tical density gradient via 
2 1 acr 
N =g?-?az- (16) a 
where g is the gravitational constant (9.81 m f 2) and eris the density of water (kg m-3 ; 
Massei 1999). This quantity , N, is a "frequency" because in eq. 16 the units of mass 
(kg) and di stance (m) cancel, leaving only units of inverse time, namely s- 1 or Hertz 
(Hz). A reason that high buoyancy frequency indicates high water column stability is 
analogous to the dependence of the pitch (frequency) of a plucked violin string on the 
tensioning of the stiing. 
27 
Tnterannual Variability and the Effect of Freshwater Inflow 
The effect of seasonal to annual scale changes in freshwater discharge rates on 
physical transport regimes was examined using both the box models and a model-free 
(i.e . does not depend on the box model) measure of vertical water column structure 
(i.e. stratification strength). For the former, physical transport rates were computed 
for each of the 156 months during the period 1986-1998 using the box model. Rather 
than using seasonal average inputs, the box model computation utilized the observed 
monthly data. Seasonal average rates of horizontal advection and vertical diffusion 
were compared to freshwater inflow rates using regression. 
For the latter, it was necessary to develop an index of stratification appropiiate 
for the seasonal/whole-system scale of interest. The index was based on the Brundt-
Vaisala frequency, which is a measure of water column stability as described above 
(Massei 1999). In this case, the vertical density gradient, da/dz, was calculated at a 
1-m depth interval using a 2-m moving window. Maximum N 2 (units: s-2) typically 
occurs within the pycnocline, and with lower values within the surface and bottom 
mixed layers (Fig. 2-4). The strength of the pycnocline was quantified as the 
maximum value of N2 (Fig. 2-4), hereafter referred to using the notation "max(N2)." 
2
For each sampling date between 1984 and 1999, max(N ) was calculated at 1.85 km (l 
NM) intervals down the entire axis of the Bay. An index of stratification strength for 
summer in the mid-Chesapeake Bay was obtained by averaging all observations of 
max(N2) between 102 and 222 km from the Bay mouth for April-September, the 
period encompassing the beginning of oxygen depletion in spring and ending with the 
re-oxygenation of the lower mixed layer in fall (Chapter 3). 
28 
Results and Discussion 
Water Budget 
The computed 1986-1998 average freshwater input to Chesapeake Bay was 
2,215 m3 s-1, of which the Susquehanna flow accounted for 48% (Table 2-5a). All 
other inputs were combined by model segment. Thus , segment 6, which includes the 
Potomac River, accounted for 22% of the total input. Segment 9, which includes the 
James and Appomattox Rivers, accounted for 14% of the total. Combined inputs 
between the Susquehanna River and the Potomac River accounted for only 5.2% of 
the total freshwater input. 
The estimated freshwater inputs can be aggregated to compare with the 
estimates of the United States Geological Survey, which uses the method of Bue 
(1968) to calculate monthly streamflow inputs above 5 cross sections of Chesapeake 
Bay. The present estimates were in very close agreement with the USGS estimates, 
departing by less than 4% at each of 5 reference points during 1996 and 1997, the two 
years for which direct comparisons were made. The extremely close agreement 
between the USGS estimates and this study reflect the fact that large areas of the 
Chesapeake Bay watershed are gauged and that watershed areas needed to estimate the 
remaining areas have been tabulated and published. 
Seaward Advection 
Down-estuary advection increased in the seaward direction, reflecting the 
inclusion of the upwelling flows from the bottom layer into the seaward flows. This 
29 
can be expressed as a multiple of the total landward freshwater inputs (R,)  , g?1 v.m g an 
indication of the magnitude of the landward transport in the bottom layer (Table _ b 
2 5
' 
Fig. 2-5). This multiple increased from l .5R, at 240 km (Patapsco R.) to a value 
between 3R, and 7 R, at the Bay mouth (Fig. 2-5). Expressed as a multiple of the 
Susquehanna River flow only (R), advection varied between l.5R at 240 km to a 
seasonally varying range of 4.5R to llR near the Bay mouth (Table 2-5b). The 
seasonal pattern in thi s multiple does not reflect a seasonal intensification of 
gravitational ci rculation , as was found for the Patuxent River (Hagy et al. 2000). In 
this case, it simply reflects persistence of the gravitational circulation despite the 
seasonal decrease in river flow (Table 2-5b). The magnitude of up-estuary advection 
in each segment is not reported, but can be easily computed from seaward advection 
minus the cumulative freshwater inputs (Table 5a,b), satisfying the water balance 
expressed in eq. 6. 
Average net non-tidal current velocities along the main axis of the estuary 
were estimated by dividing the advective transport rate by the respective suzface layer 
or bottom layer cross-sectional area. Although the advective transport rates increased 
down-Bay (Table 2-5b), the residual current velocities were relatively constant due to 
the increasing cross-sectional area of the estuary. Estimated net non-tidal current 
velocity in the surface layer averaged 4.8 cm s-' in the upper Bay (box 1) with higher 
velocity (B-9 cm s-') associated with spring freshwater discharge. In the mid-Bay (box 
5) average net non-tidal current velocity was 3.7 cm s-' with peak velocity (-5 cm s-') 
in late spring and a seasonal minimum (2 cm s-') in July. Using these velocities and 
the length of each model segment, the estimated transit time in the surface layer from 
30 
the Susquehanna River to the Atlantic Ocean varied between a minimum of 57 days in 
March and a maximum of 141 days in August. 
The mid-Bay net seaward velocity was smaller than average velocities 
estimated using a current meter located at 2.5 m depth on the mid-Bay Chesapeake 
Bay Observing System (CBOS) buoy. The current meter measured an average 
veloc ity of 5-8 cm s-1 compared to 2-5 cm s-1 from the box model (Fig. 2-6). The 
hi gher rates observed by the CBOS buoy most likely reflect higher current velocities 
in the mid-channel relative to the cross-section average, the latter of which the box 
model estimates. Net non-tidal landward current velocity in the bottom layer averaged 
7.7 cm s-1 in the mid-Bay with higher velocity (10-11 cm s-1) in winter-spring and 
lower velocity in summer (4 cm s-\ These rates compared much better to the CBOS 
current meter estimates, which are for a cuITent meter at 18.5 m depth (Fig. 2-6). The 
box model estimates agreed especially well with the current meter data during July-
October, while the box mode l estimates were hi gher earlier in the year. The closer 
agreement may refl ect lower lateral variation in current velocity in the naJTower lower 
layer channel, which would tend to reduce differences between the mid-channel 
estimate and the cross-sectional average. Interestingly, agreement in the bottom layer 
implies that the box model is correct in the surface layer as well provided that the 
freshwater inputs are reasonably correct (see eq. 6). In general, the agreement that 
was observed between the box model and current meter data was very encouraging. 
31 
Vertical Advection 
Upwelling occurs along the entire salinity gradient of the Bay du1ing most of 
the year. Since the pycnocline area varies among the model segments, the volumetric 
rate of upwelling (Table 2-5c) was scaled by the pycnocline cross-sectional area 
(Table 2-4) to estimate average upwelling velocity (Fig. 2-7). North of the Choptank 
Ri ver ( 180 km), upwell ing velocity reached seasonall y maximal rates duri ng October 
through April , coincident with peak river di scharge. Peak velocity was approximately 
90-100 cm d-1? During summer, upwelling was slower, 17-40 cm d-1? In the 
mesohaline Bay, between the Rappahannock Ri ver and the Choptank Ri ver, seasonal 
variations in upwelling were small and average upwelling was general ly lower, about 
24 cm d-1? Further to the south , in the region near the mouth of the Potomac Ri ver 
(120 km), upwelling was more intense, - 40-70 cm d-1? A small component of thi s 
additional upwelling could be attributable to upwelling in the Potomac Ri ver, although 
it does not seem like ly that thi s is sufficient to account for the di fference in upwelling 
between segment 6 and segment 5 (Fig. 2-7). Interestingly, the box model estimated 
weak downwelling (0 to -20 cm d-1) in segment 7 during October-Ap1il. During the 
remainder of the year, there was only weak upwelling ( <10 cm d-'). 
Both the region of enhanced upwelling near the Potomac Ri ver and the region 
of weak upwelling or downwelling may be related to effects of the shoaling of the 
central channe l of the Chesapeake Bay in segment 7 between the Potomac and 
Rappahannock Ri vers (Fig. 2-1). During periods of high river flow, the surface layer 
outflow is topographically impeded, leading to downwelling landward of the shoal. In 
response, enhanced upwelling is expected landward from the downwelling region 
32 
(Chao and Paluszkiewicz 1991). Chao and Paluszkiewicz (1991) investigated this so-
called region of "hydraulic control" theoretically via model simulations. Direct 
observations also suggest the presence of this feature . A near-synoptic profile of the 
Brundt-Vais~ila frequency in the region during a high 1iver flow period in Ap1il 1998 
showed that the pycnocline is displaced downward landward of the shoal, then 
rebounds further landward (Fig. 2-8). Although the scale of this feature is smaller 
than suggested by the upwe lling and downwelling patterns in Fig. 2-7, the difference 
may be related to the spatial resolution of the box model. Thus, the locations of the 
regions of upwelling and downwelling in Fig 2-7 are only approximate. 
Vertical Non-Advective Exchange 
Upward diffusion of salt across the pycnocline occurs along the salinity 
gradient of Chesapeake Bay and causes the bottom layer salinity to decline in the 
landward direction. DUJing spring, vertical exchange was lowest (-20-60 cm ct?') in 
the lower mesohaline Bay, near the mouth of the Potomac River. Minimum non-
advecti ve exchange velocity was observed du1ing May through October between the 
Potomac River and Eastern Bay (Fig. 2-9). Values were typically 10-20 cm d-1? 
Summertime vertical exchange velocity was higher in the upper Bay (-70-90 cm d-1) 
and much higher (-90-130 cm d-1) in the lower Bay. In fall, exchange velocity in the 
middle Bay approximately doubled in September compared to August, then doubled 
again in October. Thus, fall non-advective exchange was -60-90 cm ct?' in the middle 
Bay. Fall exchange in the lower Bay was comparable to spring, while in the upper 
Bay, higher exchange was observed in the fall, - 110 cm ct?'. 
33 
Compared to Patuxent River, vertical non-advective exchange velocity in 
Chesapeake Bay was relatively greater. For example, in the middle region of both 
estuaries, where non-advective exchange during summer was minimal, Patuxent River 
1 
had exchange velocity -5 cm d- (result derived from Hagy et al. 2000). Chesapeake 
1
Bay had values - 10-20 cm d- ? Near the estuarine turbidity maximum of each 
1 
estuary, exchange velocity was - 27 cm d- for Patuxent River versus 65-120 cm d-1 in 
Chesapeake Bay. However, because Chesapeake Bay has a greater average depth than 
Patuxent River, the turnover time for the lower layer volume based on the non-
advective exchange alone was higher in Chesapeake Bay. During summer in the 
middle reach of Chesapeake Bay, this turnover time was -50 days, approximately 5 
times the turnover time for Patuxent River (-8-10 days), perhaps contributing to the 
greater severity of hypoxia and anoxia in Chesapeake Bay compared to Patuxent River 
(Hagy and Boynton 2000). 
The average summertime non-advective exchange velocities (Ew,IA) for the 
1
mesohaline Bay, 10-20 cm d- , are equivalent to vertical diffusivities (Dz) of 0. IO-o.21 
c m2 s-1 (eq. 5), a value similar to or slightly lower than previously reported estimates 
for Chesapeake Bay (Kemp et al. 1992). 
Physical Transport Variability Associated with Freshwater Inflow 
It has been previously suggested that spring average river flow affects summer 
water column stratification and vertical mixing by depressing average salinity 
throughout the Bay (Boicourt 1992). The horizontal salinity gradient between the 
estuary and the adjacent coastal water thereby serves as a "buoyancy reserve" from 
34 
which vertical salinity structure can be regenerated following any mixing event 
through gravitational circulation (Boicourt 1992). This basic model was supported by 
the results of this study. Summer water column stratification dwing 1984-1999, 
quantified using an index based on the Brundt-Vaisala frequency, was strongly related 
to Susquehanna River flow averaged over the period January-May (r2=0.90; p<O.Ol; 
Fig. 2-10), increasing about 50% over the range of river flow. In contrast, other less 
integrative measures of river flow (i .e. individual months, two-month periods, etc.) 
were inferior predictors of stratification. Regardless of river flow, summer 
stratification in the mid Chesapeake Bay is strong. For example, the average squared 
2 2 
buoyancy frequency (N in Chesapeake Bay is about 10- s-2) , 500 times greater than 
the comparable value for the Atlantic Ocean (Massei 1999 reported a period of 4 
minutes, or 2 ( 1/240) = 7 x l 0 -s s -2) . 
Vertical diffusivity (Dz) is expected to be inversely related to water column 
stability, a result supported here with qualification (Fig. 2-10). April-September 
median Dz for the middle segment of the mesohaline Bay (segment 5) varied between 
2 
0.24 cm2 s-1 in 1992 and 0.09 cm s-' in 1998. The seasonal median was used to obtain 
a good indication of central tendency due to the presence of one or more extreme 
values in most years. In general, median Dz varied such that higher values were 
associated with low river flow (Fig. 9, lower panel). However, there was substantial 
uncertainty within this overall pattern. To quantify the relationship, while not 
allowing any single observation excessive leverage (in the statistical sense, see Sokal 
and Rohlf 1995, pg. 531), and without arbitrarily excluding any observations, the 
2
relationships between Dz and max(N ) and Dz and January-May average 1iver flow 
35 
were quantified using iteratively-reweighted regression (SAS Institute 1990). For 
1986-1998, April -September median Dz varied was related to max(N2) via 
D = 0 .26 - ll.8max(N 2 ) (17) 
2 
(s.e. of slope = 6.5) and to January-May average Susquehanna River flow (m3 s- 1) via 
D:. = 0.20 - 3.3(10- 5 )Q (18) 
(s.e . s lope= 1.9 10-5). According to the latter model , the expected Dz at low, average 
and hi gh 1iver flow is -0.17 , 0.15 and 0.12, respectively, a range of 40% relative to the 
minimum. This is slightly less variability than was observed for max(N2) over the 
same range of sp1ing river flow . Both models explained only -22% of the va1iability 
in D, and gave admittedly imprecise estimates of Dz , indicating that factors affected 
Dz in addition to water column stratification in the central channel of the Bay. This 
could include wind- and tide-driven seiching events (Breitburg 1990, Sanford et al. 
1990, Boicourt 1992). The salt exchange resulting from such events would be 
integrated in the box-model based measures of vertical exchange, even though the 
effect on stratification in the central channel of the Bay might be minimal. 
Interestingly, residuals in the lower panel of Fig. 2-10 were related to departures from 
the relationships between hypoxia and river flow (e.g. 1992, 1995, 1998, Chapter 3). 
Negative residuals for 1995 and 1998 (Fig. 2-10) were related to unexpectly high 
hypoxic volume (Chapter 3), while positive residuals for 1992 and 1996 were 
associated with lower than expected hypoxia (Chapter 3). 
While it has long been recognized that interannual changes in river flow could 
affect vertical mixing in Chesapeake Bay, the effect of river flow on horizontal 
advection is both more obvious intuitively and less clear in reality. On the one hand, 
36 
river fl ow contributes directly to seaward advection , and should be corre lated, as 
imp lied by F ig. 2-5. On the other hand, seasonal variations in the flow multiples 
( Q/Q,. , i.e. 2R, 3R, 4R in Fig. 4) are associ ated with the seasonal pattern of freshwater 
input such that in spring, where river flow is hi gh, the mu ltiple is low. In summer and 
fa ll , when river flow is low, the mul tip le is high (Fig. 4). This mitigates agai nst the 
s imple effect o f freshwater inflow on R. The same effect appears to operate 
interannuall y (Fig. 2- 11 ), except that the increase in ri ver fl ow overwhelms the 
dec reas ing multiple such that landward advecti on did in fact increase with fres hwater 
3 
inflow . At low average spring ri ver fl ow (1000 m s-'), landward advection into the 
mid Bay during summer was 3.3-fold greater than spring average freshwater input, 
however thi s decreased to l.9R at high fl ow (Fig. 2- 11). Summer average landward 
3 1
advection into the mid Bay (m f ) was predicted by 
(19) 
Q7 == 2386 ? 511 + 0.94 ? 0.3Q,. 
3 1 2
where Q,. is the January-May average Susquehanna River flow (m s- , r =0.68 , 
p<0 .05). F urther up-estuary, the multiple was lower (simil ar to Fi g. 2-5) and there 
was no c lear re lationship between spring average freshwater inflow and landward 
advection. 
Conclusions 
The results suggest that thi s rel atively simple box model is an effective tool for 
estimating net transport in Chesapeake Bay at seasonal time scales. Given the 
complexiti es of implementing a hydrodynamic circulation model for thi s purpose 
alone, or for interpreting the results of such a simul ation , thi s approach may be a 
37 
useful alternative for ecologists and other environmental scientists seeking to evaluate 
large-scale transport issues for this system. In addition , as has been described to some 
extent here, comparisons among estuaries are possible, indicating how differences in 
physical transport could affect ecological processes in those ecosystems. Many such 
app lications are possible. For example, it is possible to develop regionally and 
seasonally resolved oxygen budgets for the lower water column (see Chapter 2). 
These transport estimates could also be used to examine nutrient, carbon, or even 
dissolved toxicant transpo1t within the estuary. Some of these applications have been 
approached in the past usi ng similar methodologies (e.g. Taft et al. 1978, Officer et al. 
1984). However, the model presented here utilized important improvements to the 
box model as well as a much more extensive water quality and freshwater input 
record. 
Literature Cited 
Boicourt, w. c. 1992. Influences of circulation processes on dissolved oxygen in the 
Chesapeake Bay. In: Smith, D. E., M. Leffler, and G. Mackiernan (eds.). 
Oxygen Dynamics in the Chesapeake Bay. A synthesis of recent research. 
Maryland Sea Grant College, College Park, MD. 
Breitburg, D. L. 1990. Nearshore hypoxia in the Chesapeake Bay: Patterns and 
relationships among physical factors. Est. Coast. Shelf Sci .. 30(6): 593-609. 
Bue, c. D. 1968. Monthly surface-w~ter inflow to Chesapeake Bay. U.S. Geological 
Survey Open-File Report. Arlington, VA. October 1968. 
Chao, s. Y. and T . Paluszkiewicz. 1991. The hydraulics of density currents over 
estuarine sills. J. Geophys. Res. 96:7065-7076. 
Chesapeake Bay Water Quality Monitoring Program. EPA Chesapeake Bay Program. 
Annapolis, MD. http://www.chesapeakebay.net. 
38 
Cronin, W. B . and D. W. Pritchard . 1975. Additi onal stati sti cs on the dimensions of 
Chesapeake Bay and its tributaries: Cross section widths and segment volumes 
per meter depth. Chesapeake Bay Insti tute . Reference 75-3 . Special Report 
42. 
E lliott, A. J. 1978. Observations of the meteorologically induced circul ati on in the 
Potomac Estuary. Est. Coast. Shelf Sci. 6: 285-299. 
EPA 1998. 1985 Reference Scenario. Phase IV Chesapeake Bay W atershed Mode l. 
07 January 1998. Chesapeake Bay Program Office, Annapoli s, MD. 
Fi sher, T. R. , J . D . Hagy and Emma Rochelle-Newall. 1998. Dissolved and particulate 
organic carbon in Chesapeake Bay. Estuaries. 21(2): 215-229. 
Gordon , D . C. , Jr. , P.R. Boudreau, K. H. Mann , J .-E. Ong, W. L. Silvert, S. V. Smith, 
G . Wattayakorn, F. Wulff, and T. Tanagi . 1996. LOICZ Biogeochemical 
Modell ing Guidelines . LOICZ/R&S/95-5, vi +96 pp. LOICZ, Texel , The 
Netherlands . 
Hagy, J . D. 1996. Residence Times and Net Ecosystem Processes in Patuxent River 
Estuary. M.S. Thesis. University of Maryland at College Park. 
Hagy, J. D. and W. R. Boynton. 2000. Controls on hypoxia in Chesapeake Bay and 
its major tributaries. pp. 91-102. In : Kemp, W. M., R. Bartleson , S. 
Blumenshine, J. D. Hagy and W.R. Boynton. Ecosystem Models of the 
Chesapeake Bay Relating Nutrient Loadings , Environmental Conditions, and 
Living Resources . Final Report. US EPA Chesapeake Bay Program, 
Annapolis , MD. 
Hagy, J. D ., L. P . Sanford, and W.R. Boynton. 2000. Estimation of net physical 
transport and hydraulic residence times for a coastal plain estuary using box 
models. Estuaries. 23(3): 328-340. 
Hood R.R. , H. V. Wang, J.E. Purcell , E. D. Houde, and L. W. Harding, Jr. 1999. 
Modeling particles and pelagic organisms in Chesapeake Bay: Convergent 
features control plankton distributions. J. Geophys. Res. I04(C2): 3289-3290. 
Hopkins, K. , B. Brown, L. C. Linker, and R. L. Mader, Jr. 2000. Chesapeake Bay 
watershed model land uses and linkages to the airshed and estuarine models. 
US EPA Chesapeake Bay Program, Annapolis, MD. 
Janzen , C. D . and K.-C. Wong. 1998. On the Low-Frequency Transport Processes in 
a Shallow Coastal Lagoon. Estuaries 21(48): 754-766. 
39 
Johnson, B. , R. Heath , K. Kim and H. Butler. 1991. Development and verification of 
a 3-D numeri ca l hydrodynamic salinity and temperature model of Chesapeake 
Bay. Tech. Rept. HL-91-7. Army Corps of Engineers, Waterways 
Experimentation Station, Vicksburg, MS . 192 pp. 
Kemp, W . M. , P .A. Sampou, J . Garber, J. Tuttle, and W.R. Boynton . 1992. 
Seasonal depletion of oxygen from bottom waters of C hesapeake Bay: Roles 
of benthic and planktonic respiration and physical exchange processes. Mar. 
Ecol. Prag. Ser. 85(1-2): 137-152. 
Massei , S. R. 1999. Hydrodynamics of Coastal Zones. E lsevier Science, New York. 
Officer, C. B . 1980. Box model s rev isited. In P . H amilton and R. B. Macdonald 
(eds.) . Estuarine and Wetland Processes. Marine Sciences Seri es . Vol. 11. 
New York: Plenum Press . 
Officer, C. B. , R. B. Bi ggs, J. L. Taft, L. E. Cronin, M.A. Tyler, and W.R. Boynton . 
1984. Chesapeake Bay anoxia: Origin , development and significance. 
Science. 233 : 22-27. 
Pritchard, D . W . 1952. Salinity distribution and circulation in the Chesapeake 
estuarine system. J. Mar. Res .. 11(2): 106-123. 
Pritchard, D. W . 1969. Dispersion and flushing of pollutants in estuaries. American 
Society of Civil Engineers Journal of the Hydraulics Division . 95(HYI) : 115-
124. 
SAS Institute, Inc. 1990. SAS/STAT User's Guide, Version 6 , 41h Edition. Volume 2. 
p 1165. 
Sanford, L. P. , K. G. Sellner, and D . L. Breitburg. 1990. Covariability of dissolved 
oxygen with physical processes in the summertime Chesapeake Bay. J. Mar. 
Res. 48: 567-590. 
Smith, S. V. , J. T. Hollibaugh, S. J. Dollar, and S. Vink. 1991. Tamales Bay 
metabolism: C-N-P stoichiometry and ecosystem heterotrophy at the land-sea 
interface. Est. Coast. Shelf Sci. 33: 223-257. 
Sokal , R.R. and F. J. Rohlf. 1995. Biometry. Third Edition . Freeman and 
Company, New York. 
Taft, J. L. , A. J. Elliott, and W.R. Taylor. 1978. Box model analysis of Chesapeake 
Bay ammonium and nitrate fluxes. In Wiley, M. L. (ed.), Estuarine 
Interactions . New York: Academic Press. 
40 
Valle-Levinson, A. and K.M. M. Lwiza. 1995. The effects of channels and shoals on 
exchange between the Chesapeake Bay and the adjacent ocean. J. Geophys. 
Res. 100(C9): 18,551-18,563. 
Wang, H. V. and B. H. Johnson. 2000. Validation and application of the second 
generation three-dimensional hydrodynamic model of Chesapeake Bay. 
Submitted to Journal of Water Quality and Ecosystem Modeling. In review. 
Wang, H. v. c. and S.-Y. Chao. 1996. Intensification of subtidal surface cuffents 
over a deep channel in the upper Chesapeake Bay. Est. Coast. Shelf Sci. 42: 
771-785. 
Webster, I. T., Parlow, J. S., and S. V. Smith. 2000. Implications of spatial and 
temporal variation for biogeochemical budgets of estuaries. Estuaries. 23(3): 
341-350. 
41 
Table 2- J. The boundaries of the box model segments in channel 
kilometers from the mouth of the Bay and the depth of the pycnocline 
dividing the surface and bottom layers. 
Segment South North Pycnocline Depth 
Number Boundary (km) Boundary (m) 
(km) 
9 l 19 6 
8 19 46 7 
7 46 93 8 
6 93 139 12 
5 139 185 12 
4 185 222 12 
3 222 241 7 
2 241 259 5 
l 259 287 NIA 
42 
Table 2-2. The volume of mainstem Chesapeake Bay and the tributaries and 
embayments with which lateral sa lt exchanges occur and the stations used to estimate 
mean salinity throughout the Chesapeake Bay estuarine system. 
Basin Volume % of CBMP Stations 
(106 m3) Total 
Chesapeake Bay 51 ,027 67.0 CBl.l, CB2.l, CB3.l, CB3.2, CB3.3C, 
CB4.1C, CB4.2C, CB4.3C, CB4.4, 
CBS.l, CBS.2, CBS.2, CBS.3, CBS.4, 
CB.5.5, CB6. l CB6.2, CB6.3, CB6.4, 
CB7.3 , CB7.4 
Eastern Bay 721 0.9 EEl.l 
Choptank River 1,324 1.7 EE2. l 
Patuxent River 430 0.6 TFl.5, TFl.6, TFl.7, RETl.l, LEl. l , 
LEl.2, LEl.3, LEl.4, CBS.lW 
Potomac River 7,141 9.4 TF4.2, TF2.0, TF2.l, TF2.2, TF2.3, 
TF2.4, RET2.2, RET2.4, LE2.2, LE2.3 
Rappahannock River 1,640 2.2 TF3.2, TF3.3, RET3.l , RET3.2, LE3.l, 
LE3.2, LE3.4, LE3.6 
Tangier Sound 3,169 4.2 EE3.4, EE3.5 
Pocomoke Sound 848 1.1 EE3.3 
York River 917 1.2 TF4.2, RET4.l , RET4.3, LE4.l, LE4.2, 
LE4.3 , WE4.2 
James River 2,166 2.8 TFS.6, RETS.2, LES.l, LES.2, LES.3, 
LES.4, LES.5 
Total, Tributaries 18,356 24 
and Embayments 
Total Volume 69,383 91 
Chesapeake Bay Water Quality Monitoring Program stations from which data was 
used to calculate the salinity of the respective water body. 
43 
Table 2-3. The fall line gauges, USGS ID and drainage area upstream of the fa ll I' 
gauging station on the major tributari es of Chesapeake Bay. Freshwater flows ent~:e 
the segment or segments indicated. 
USGS ID Gauged Ungauged Box Model 
Bas in Drainage Drainage Segment 
Area Area (krn2) 
(km2) 
Major Tributary Watersheds: 293 n/a 1 5 
Choptank River near Greensboro, MD 01491000 
01646500 29,940 6,699 6 
Potomac River near Washington , DC 
01594440 901 550 5 
Patuxent River near Bowie, MD 70,189 l 
Susquehanna River at Conowingo, MD 01578310 
01668000 4,134 2,462 7 
Rappahannock River near 
Fredericksburg, VA 01673000 2,800 
Pamunkey River near Hanover, VA 2 1,557 
Mattaponi River near Beulahville, VA 01674500 
4,357 2,323 8 
York River (Mattaponi+Pamunkey) 
02041650 3,481 
Appomattox River near Matoaca, VA 
02037500 17,503 
James River near Richmond, VA 20,984 4,958 9 
James River (James+Appomattox) 130,798 16,992 
SUB-TOTAL 
Smaller Watershed Areas 
3 4 9,086 1,3-7 
PA/DE/MD Eastern Shore '
34 831 7 
VA Eastern Shore ? 
3 5 3,059 2,3,4 
PA/MD Western Shore '
3 6 1,812 6-9 
VA Western Shore ? 2,078 
Open Water 16,866 
SUB-TOTAL 130,798 33,858 
TOTAL AREA 
Below fall line areas of the Choptank River were included within the PA/DE/MD 
1 
ETashteer Mn aSthtaopreo.n i gauges did not function during late 1987 through early 1989. 
2 
Month ly average flow of the Mattaponi during that period was estimated via a 
regression model (r2>0.9) from the monthly average flow of the Pamunkey River. 
3 Areas determined from Phase IV Chesapeake Bay Watershed Model segments (EPA 
1998; Hopkjns et al. (2000). 
Flow/area based on gauged port!on of Choptank ~iver (USGS gauge 01491000). 
4 
s Flow/area based on gauged port'.on of Patuxent ~1~er (USGS gauge 01594440) 
6 Flow/area based on gauged port10n of Mattaponi River (USGS gauge 01594440) 
44 
Table 2-4. The dimensions, including the length, pycnocline area, cross-sectional area 
(cross-sec), and volume of the surface layer (SL) and bottom layer (BL) boxes in 
segments l through 9 of the box model. 
Box l Box 2 Box 3 Box 4 Box 5 Box 6 Box 7 Box 8 Box 9 
Length 1 28 18 29 37 46 46 47 27 18 
Surfa ce Area2 217 255 169 489 655 1025 1425 782 454 
Pycnocline 80 92 144 249 278 904 510 289 
Area 2 
Cross-Sec, SL3 28 60 52 107 123 185 212 178 139 
Cross-Sec, BL3 13 31 27 35 37 85 83 
Volume SL4  
57 
773 1077 995 3961 5675 8502 9977 4796 2500 
Volume ' BL4 242 593 988 1612 1684 4015 2246 
1 2 2 3 3 2 4 1018 units=km; units=km ; units= l0 111 ; units=106 3 111
45 
Tabl e 2-5. Box mode l estimates of the long-term (1986-1998) average 
monthl y phys ical transport (m3 s-1) for Chesapeake Bay. Estimates include (a) 
F reshwater inputs to each model segment, (b) seaward advection in the 
surface layer of each segment, (c) upwelling in each two layer segment, and 
(d) non-advecti ve verti cal exchange in each two layer segment. A ll transport 
rates have units of m3 s-1? 
(a.) 
Month Q19 Total 
Jan 1300 28 38 25 21 78 722 184 133 482 3011 
Feb 1269 36 35 28 23 84 731 181 158 493 3038 
M ar 1923 55 49 37 31 11 3 1075 237 179 620 4318 
Apr 2167 28 37 25 21 78 799 169 144 501 3969 
May 1298 18 35 21 18 67 675 147 98 380 2757 
Jun 681 1 23 13 12 42 317 98 53 252 1490 
Jul 501 -12 17 7 7 25 199 60 36 156 996 
Aug 389 - 11 15 8 7 25 179 48 27 109 797 
Sep 389 -17 15 6 6 20 219 63 31 178 909 
Oct 61 2 -13 18 7 7 25 203 65 36 161 11 21 
Nov 1020 -9 26 10 10 34 310 95 56 206 1758 
Dec 1257 10 32 18 16 57 487 130 99 311 241 8 
Average 1067 9 28 17 15 54 493 123 88 321 2215 
% input 48 0.4 1.3 0.8 0.7 2 22 6 4 14 
(b.) 
Month 
Jan 1300 1647 2529 3623 5235 6002 9207 8075 8504 16915 
Feb 1269 1625 2525 3625 5091 5431 8520 6611 7373 14032 
Mar 1923 2222 3040 4183 5975 6644 9837 9190 10417 21008 
Apr 2167 2447 3064 3850 5342 6132 8987 9026 9145 15880 
M ay 1298 1529 2081 2617 3544 4179 7288 8403 8598 13430 
Jun 681 814 1293 1612 2197 2881 5116 5974 6671 11680 
Jul 501 673 1070 12611921 24884274 4693 5481 9301 
Aug 389 590 1020 1212 1938 2657 4592 5305 5808 9475 
Sep 389 604 1180 1518 2220 3035 5362 6487 6922 13348 
Oct 61 2 884 1499 2168 281 2 3454 5254 5593 5539 9966 
Nov 1020 1284 1989 2941 4569 5159 6822 5969 6276 11656 
Dec 1257 1533 2415 3471 5109 5955 8633 8213 7738 15970 
Average 1067 1321 1975 2673 3829 4501 6991 6962 7373 13555 
Ratios : 1 
Q,,/tQp,, 1.0 1.2 1.8 2.4 3.4 3.8 4.2 3.9 3.9 6.1 
Q,11/QJO 1.0 1.2 1.9 2.5 3.6 4.2 6.6 6.5 6.9 12.7 
calculated from the annual average. 
46 
(c.) 
Qv3 Qv4 Qv5 Q.,6 Qv7 Qvs 
Qv9 
Month Qv2 
1068 1592 689 2482 -1315 296 7929 Jan 844 
865 1073 1442 256 
2359 -2090 604 6167 
Feb 
1106 1761 557 2118 -884 
1049 9971 
Mar 768 
580 761 1471 713 
2055 -129 -26 6235 
Apr 
516 515 908 569 
2433 968 97 4452 
May 
306 574 642 1918 761 
644 4757 
Jun 456 652 542 1587 359 752 3664 
Jul 380 184 
414 184 719 694 
1756 665 475 3558 
Aug 
333 696 796 2107 
1062 404 6248 
Sep 561 
662 637 618 1596 
275 -91 4266 
Oct 597 1618 556 1353 -949 251 5174 
Nov 679 943 
1037 1623 788 2191 
-549 -574 7921 
Dec 851 
(d.) 
Ev4 Ev5 Ev6 Ev7 Evs Ev9 Month Ev2 Ev3 
984 1398 1054 6740 13642 10620 
Jan 641 849 802 3298 7816 8647 
Feb 818 1109 
863 1107 
875 1073 1193 
2341 5801 8089 8173 
Mar 619 766 692 1267 6865 8820 5038 
Apr 801 757 
670 619 401 
430 454 4423 8090 5037 
May 392 451 630 3807 7580 3752 
Jun 728 724 
475 194..  415 
253 3226 5389 3285 
Ju l 765 -31 397 224 3617 
5916 3778 
Aug 915 520 310 804 1170 5430 11750 6694 
Sep 1148 892 1333 1982 7688 11059 6643 
Oct 1036 929 
1236 
1468 2081 3288 9394 15775 8879 
Nov 944 927 2375 11448 31371 15948 
Dec 681 609 
799 1562 
* This estimate has been included for completeness, but is invalid and cannot be 
interpreted. Most likely, Ev4 for August was smal l, but comparable to adjacent 
months. 
47 
Fig. 2-1. A map of Chesapeake Bay showing the bathymetry (<10, 10-lSm 
>15m), the boundaries of the segments of the box model, the embayments ' 
that were included in the salt budget (e.g. TS, YRK), and the fall Jines for 
which freshwater inputs data were obtained (denoted by "k> Tangier (TS) 
and Pocomoke Sounds (POC) were assumed to exchange primarily with box 
7 because the deep channel runs in the North South direction into Box 7. The 
passes between the islands to box 6 are much narrower and shallower. Dis-
tances indicated on the segment boundaries are channel kilometers from the 
Bay mouth. The location of the mid-Bay Ch~sapeake Bay Observing System 
(CBOS) buoy, from which current meter readings were obtained is also indi-
cated. 
48 
O Water Quality Monitoring 
Station 
Mattaponi 
49 
A. 
B. 
Down-estuary 
Up-estuary 
Lateral 
Exchange 
Vertical 
Exchange 
Fig. 2-2. (A) The simplest possible box model for an estuary, where Q is 
freshwater input, Q is advection ~ut, and Eis di~persive exchange at th~ 
seaward margin. (B) A~,ilJ.~s~rat10n o: the possible physical exchanges for 
a surface layer box (box m ) rn a vertically structured, branching model 
such as the Chesapeake Bay box model. The exchanges correspond to 
terms in eg. (9). For all the model segments, Em,m+J and ? _ ,,,, were 
111 1
assumed to be negligible. For bottom layer boxes, Q1,,, and L1S are not 111 
present and the horizontal advection terms are directed up-estuary. 
50 
Fig. 2-3. A schematic diagram of ?ox model for Chesapeake Bay, drawn 
overlying the maximum depth profile of the estuary. The approximate 
boundaries of the major salinity zones of the estuary are shown as heav 
lines . Bi-directional airnws indicate salt exc~anges_, while uni-directio:aJ 
arrows indicate advective flows. Exchanges rnvoJvrng Eastern Bay T.a . 
, ng1-
er Sound and Pocomoke Sound are lateral salt exchanges. Exchanges with 
the Choptank, Patuxent, Potomac, Rappahannock,_ York, and James Rivers 
involve salt exchanges and exports of fresh water into the indicated 
regions. Freshwater inputs to box 1 through 3 include only diffuse fresh-
water inputs from small watersheds. 
51 
~----
Oligohaline Mesohaline Polyhaline 
Region Region Tangier Sound, Region 
Eastern Choptank R., Pocomoke Sound. 
Bay Patuxent R . Potomac R . Rappahannock R . York R . James R. 
Q f3 Q f4 Q fS 6.stS Qf6 6.stS Q17 6.sts Q 17 6.sts Q17 6.sts 
0 -.~ ~~~- { ? L J_ ? ~-4-? - -}{--1- i-l \{ { 
Qfl Q~ Q4 Q 5~ Q 6 Q7~Q17Q~Q~ 
l 
5 \ Ev7 
t 
~ 
~ 
10 Ev4 Evs 1 Ev6 
V\ 
N I 15 r 
.....c..  r 
0.. I-
(l) 
0 20 ~ Box2 Box 9 '" 
Box 5 Box6 
25 - Box 7 
r 
rt-30  
Box3 
Box 8 
35 
296 259 222 185 148 111 74 37 0 
Channel Distance from Atlantic Ocean (km) 
or,-----~--=------
maxN2 
5 
~ 
10 
,--.. 
..E.__ , 
N 
15 
..r..:..,  
0. 
Cl) 
0 
20 
dC5 
2-- 1 25 N g (5 dz 
30 +-_,-__,_-,---.---r-.---,---.---,---
0.0 0.5 1.0 1.5 2.0 
2 4 6 
8 10 
N2, s-2 
Density, cr1 
2
Fig. 2-4. An illustration of the definition of max(N ), a measure 
of pycnoc Ii ne strength based on the squared B ru ndt-Viii siilii fre-
quency (N'). The data are from the Chesapeake _Bay Water Quality 
Monitoring Program and were collected at a station in the mid 
Chesapeake Bay (CB4.3C) on July 7, 1998. In the fonnula for N' 
g is the gravitational acceleration (Massei 1999). ' 
53 
28 0 0 0 0 0 0 0 0 
0 0 0 
0 0 0 
0 -- 0 1.5R -~ 
0 0 --2.R ---- - PTP 
24 0 - . o- 1.?R - f5R ~ 
.,-... 
E 
.~__ , 
C 200 
cu CHP 
Q) <Sf? 
(.) 
0 
0 0 0 
(.) PXT 
:;::; 0 ~ 0 
C 
cu 160 
:;:; 
<!'. 
E ~Q;-
-0 oR I- 0 0 Q) 120 
(.) 0 POT 
C 
..c.u..  
(/) 
0 
Q) 0 V 0 RAP C 80 0 0 0 C 
cu 
.c 
0 
0 a IE a 
40 0 3of YRK f?$-~ ~ ~ 
-~~ >-)~-~- 02} 0 JAM 1-,-
0 1- 4 5 6 7 8 
9 10 11 12 
1 2 3 Month 
Fig. z-5. Monthly average down-estuary advection computed using the 
1 
box model and expressed as a multiple of the total landward freshwate . 
input (R). Contours show the change in relative advection along the 
length of the estuary and through the seasons of the year. Shaded circl 
indicate the time/location combinations obtained from the box model es 
and used to generate the contours. Codes on the right indicate the loca-
tions of major tributaries. JAM==James R., York==York R., 
Rap=Rappahannock R., POT=Potomac R., PXT=Patuxent R., 
CHP==Choptank R. , PTP==Patapsco R. 
54 
12 
Seaward 
10 
8 
6 
.-.-I 
Cf) 4 
E ---.----? -
...0.. .... 
..>..... . 2 
?o 
0 
Cl) 
> 0 ....... 
C 
.Cl) .........  -2 
:::J ~.l~ 
0 
cu -4 
:::J ----- ? 
"O 
Cf) 
Cl) -6 
0::: 
-8 ?0  Surface Layer Avg. 
~J 
Channel (2.5 m) 
-10 ? ---? --? / -?-Bottom Layer Avg. 
I Landward 1- -1 u Channe-l1  (18.5 m) -12 1 -, - - T - -T - ?,-- .,. 1- i 
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 
Fig. 2-6. A comparison of cross-section average cuJTent velocity computed by the 
box model for surface layer and bottom layer in segment 5 with residual cuJTent 
ve locity obtained by averaging 1995-1999 cuJTent meter data at the mid-Bay mid-
channel buoy of the Chesapeake Bay Observing System (CBOS, W. Boicourt, 
unpublished data). CurTent meters were located at 2.5 m and 18.5 m depth and 
recorded the cuITent velocity at approx. 5 minute intervals . 
55 
280 
PTP 
240 0 
,--... 
E 0 
.~.__ , 
C 200 80 
co CHP 
Q.) 
() 
0 0 
() PXT 
:;::; 
C 
co 160 
:;:; 
<( 
E 
.0.. . 
'+-
Q.) 120 POT 
() 
C 
co 
~ 
(/) 
0 
RAP 
Q.) 
C 80 
C 
co 
L: 
0 
0 0 0 
0 
0 0 
0 
40 YRK 
0 JAM 
- l -
-,--r T - 1 I r 
0 I - 1 8 9 10 1 1 12 
2 3 4 
5 6 7 
1 Month 
Fig. 2-7. Seasonal and spatial patterns in average upwelling current 
velocity (cm d? l) in Chesapeake Bay as computed by the box model. 
Shaded circles indicate the time/location combinations obtained from the 
box model and used to generate the contours. Codes on the right refer to 
the locations of the major tributary rivers, and are described on Fig. 2-5. 
Modest negative values during November-_Ap_nl near 80 km (between 
Rappahannock River and Potomac River) indicate downwelling of sur-
face water into the lower layer. 
56 
0 
:)Ii,, 
5 
10 
...--.. 
._E_.. . 15 
......c..  
Q. 
Q) 
0 20 
1111( 
25 
30 
35 I 
10 15 20 25 30 
Transect Distance (km) 
Fig.2-8 . A profile of the Brundt-Vaisala frequency in the region of Chesapeake 
Bay just to the south of the Potomac River, collected in April 1998 by the Ches-
apeake Bay LMER (TIES) program using a towed undulating CTD device. 
Darker colors indicate increased density gradients, indicating the boundaries 
between water masses . The down welling region is indicated by" l" while the 
enhanced upwelling region is indicated by "2." 
57 
______ _/ J 
240 
'c,0 PTP 
-e 0 
200 
CHP 
.....c..  
::::J PXT 
0 160 
~ 
>, 
co 
ro 
E 
0 
~ 120 POT 
Q) 
(.) 
C 
..c..o.  
Cf) 
0 
Q) 
C 
C 
co RAP 
..c 
u 
0 
YRK 
JAM 
0 
0 - ---i-- ,- ,----i--,---- - ,--,-
1 2 3 4 5 6 7 8 9 10 11 12 
Month 
Fig. 2-9. Seasonal and spatial patterns in average non-advective 
exchange velocity (cm d- 1) in Chesapeake Bay as computed by the box 
model. Shaded circles indicate the time/location combinations obtained 
from the box model and used to generate the contours. Codes on the right 
refer to the locations of the major ttibutary livers, and are described on 
Fig. 2-5. This quantity is the non-advective exchange coefficient (Ev ) 
111
scaled by the pycnocline area of the respective model segment. 
58 
T 
13 
I ye 
N 12! 
'(/) 
N 
-a I 
0 
,c._u  11 1 
(") 
0 I 0 ..- 101 
N-z --.__ - 0 , 
>cu<  9 
E 
8 
- L 
7 r -' 
0.26 ' 
0 1992 
0.24 
0.22 
...... 0.20 0 
'e n 0 
N 0.18 
E 
(.) 
N 0.16 
0 -------- 0 1996 0 
0.14 0 0 0 
0.12 1995 
0.10 1998 
0.08 J- _L ___ .L ---'- _ _.! --'- - L--' 
800 1o oo 1200 1400 1600 1800 2000 2200 2400 2600 
January-May Average River Flow (m3 5-1) 
Fig. z. JO. Upper Panel . The relationship between January-May average 
Susquehanna Ri ver fl ow and Apnl-September average water column stratifica-
tion index, max(Nz), in the mesohalme Chesapeake Bay. The indicated line is 
the least-squares regression line (r2=0.90). Lower Panel - The relationshi 
z - ay aver-
between April -September median verti cal diffusivity (D ) and January Mp 
age ri ver fl ow. The indicated line is the interatively-reweighted (robust) least 
squares regression line (SAS Institute 1990). 
59 I 
5500 
Q=2386(?511 )+0.94(?0.3)R 0 
5000 r2=0.68, p<0.05 
4500 
4000 
0 
3500 
3000 3.31P 
....... 0 Box 7 to Box 6 .,.... 
I 
(/) 2500 
C') 
-E- OQ) 2400 0 C 
0 
:;:; 
(.) 
Q) 2200 
> 
"'O Co.ss 
<( 2000 
",'O._  0 
ro 0 0 0 0 
3:: 1800 1.SR "'O 
C 0 
ro Box 6 to Box 5 
_J 
1600 0 
Oo 
1800 
0 
1600 0 
0 
1400 
1.3~ CO.SR 0 
1200 0 
0 0 
Box 5 to Box 4 
1000 
1000 1500 2000 2500 
Jan-May Avg Susquehanna Flow (m3 s-1) 
Fig. 2-11. Average up-estuary advection within the mid Bay region 
related to January-May average Susquehanna River flow (m3 s-1). 
Multiples (i.e. 2.5R) refer to the relative magnitude of the landward 
advective flow compared to the freshwater input. Multiples are D.Qt 
comparable to those in Fig. 4 (they are smaller) since these compare 
summer advection to spring freshwater inflow. 
60 
Chapter 3: 
HYPOXIA IN CHESAPEAKE BAY, 1950-2000: PROGRESSIVE 
DEVELOPMENT AND RIVER FLOW EFFECTS 
Abstract 
A 50-year record of dissolved oxygen in Chesapeake Bay was ana lyzed to 
describe the progressive development of hypoxia and anox ia in Chesapeake Bay 
pelagic and benthic habitats since 1950. The effect of Susquehanna River flow on 
hypoxia was exami ned to separate changes in hypoxia associated with river flow from 
changes due to other effects that may have varied over time. From 1950-2000, mid-
summer hypoxic volume increased substantially and at an accelerating rate. At 
1
average river flow , near anoxic volume (D0<0.2 mg r ) increased from zero in 
1950 
to 3.7xl09 m3 in 2000. Moderate hypoxia increased from 1.7xl09 m3 to 6.5xl09 mJ 
over the same period, while mild hypox ia (D0<2.0 mgr') increased from 3.6x 109 to 
9.7x 1Q9 m3. Hypoxic volume was positively correlated with January-May average 
Susquehanna River flow throughout the 50-year peiiod, but this correlation did not 
explain the Jong term trend. The area of benthic habitat affected by hypoxia was 
estimated to have increased nearly in proportion to the hypoxic volume. 
Ana lysis of river flow-related effects showed that anoxia occun-ed earlier in 
years with higher spring river flow and warmer early-spring water temperatures. This 
severely narrowed the window of time during which temperature and dissolved 
oxygen (DO) were simultaneously adequate for significant macrobenthic ac tivity. 
Vertica l diffusive DO inputs to mid Bay bottom water increased over 1950-1999, but 
horizontal advective DO inputs decreased due to increased depletion of DO in the 
61 
lower Bay in recent years. The result was a net decrease in physical DO tra 
nsport to 
the lower water column of the mid Bay during summer, exacerbating possible 
increases in metabolic rates over the same period. The unstable configuration of 
positive feedbacks contributes to the difficulty of predicting hypoxic volumes, even 
when external forcing varies substantially. The implications for management are 
clear. Nutiient loading rate reductions should lead to improvements in hypoxia, but 
patience and persistence will be needed in the face of inevitable natural variability. 
Introduction 
Depletion of dissolved oxygen (DO) from deep waters is a common feature in 
estuaries and other coastal systems where seasonally or permanent stratification if the 
water column restricts re-aeration of bottom waters by the atmosphere (e.g. 
Chesapeake Bay-Officer et al. (1984); Long Island Sound - Welsh et al. 1994; Black 
Sea_ Zaitsev 1992; Baltic Sea - de Jonge et al. 1994; Gulf of Mexico- Rabalais et al. 
1999). Hypoxia here refers to the existence of DO concentrations sufficiently low to 
harm biota directly or adversely affect normal ecological interactions. The exact DO 
concentration that defines hypoxia is nearly arbitrary since a continuous spectrum of 
effects has been observed as DO declines from a level not far below saturation. 
Chronic non-lethal effects on ecological interactions have been observed when DO 
' 
was on ly moderately depressed (- 4.0 mg r1 or about 75% saturation at Chesapeake 
Bay summer temperatures, Breitburg et al. 1997). Ritter and Montagna (1999) 
identified 3 mg r1 as a critical threshold for macrobenthos in Corpus Christi Bay, TX. 
In contrast, some hypoxia-tolerant species survive DO as low as 1.0 mg r1 (Diaz and 
62 
Rosenburg 1995). However, even hypoxia-tolerant benthic communities ai? d 
ea versely 
1
affected when DO is less than 1 mg r , with reduction to 0.6 mg r 1 being paiticuJarJy 
harmful (Diaz and Rosenburg 1995). 
In Chesapeake Bay, hypoxia or anoxia presently affects much to all of the 
below-pycnocline waters between 37? 44' N and 39? 04 ' N (Fig. l) for most or all of 
the summer. Events in which a combination of wind- and tide-driven tilting of the 
pycnocline brings hypoxic below-pycnocline waters into shallower areas , or even to 
the surface, are common in the mesohaline Bay (Malone et al. 1986; Sanford et al. 
1990; Breitburg 1990). Consequently, hypoxia may also adversely affect many 
shallower habitats in the mesoha/ine Bay at some time each summer. Many of the 
major tributaries of Chesapeake Bay are also affected (Hagy and Boynton 2000). 
Management efforts directed toward reducing hypoxia by reducing nutrient loading (N 
and P) to the estuary are being implemented on the basis of a scientific consensus that 
nutrient enrichment has led to the present status of hypoxia in Chesapeake Bay (EPA 
2000). However, it is not clear how much improvement in hypoxia should be 
expected from achievable reductions in nitrogen and phosphorus loading rates. 
Regrettably, a present Jack of estimates of nutrient loading to Chesapeake Bay piior to 
1978 precludes approaching this question by regressing hypoxia on nutrient loading 
rate despite the fact that data sufficient to characte1ize mid-summer hypoxia in 
Chesapeake Bay are available for many years back to 1950. As an alternative, if one 
is willing to assume that nutiient loading rates increased dwing this period, the trend 
over time suggests the possible consequences of increased nutrient loading, provided 
that other possible causes can be ruled out. Two studies have evaluated the available 
63 
data describing hypox ia in Chesapeake Bay between 1950 and the early 1980's 
(Flemer et al. 1983; Seliger and Boggs 1988). Unfortunately, these studies did not 
agree that hypoxia had increased, indicating the need further analysis to resolve this 
issue. 
The substanti al interannual va1i abili ty in hypoxia has greatly complicated 
assessment of long-term change and is probably a major reason for the discordant 
conc lusions that have been reached. Seliger and Boggs (1988) found that interannual 
differences in hypox ic volume were related to mean springtime freshwater discharge 
into the estuary during sp1ing. In their study, accounting for these 1iver flow effects 
not onl y explained the interannual differences , but also the apparent long-term 
increase in hypoxia (Seliger and Boggs 1988). In contrast, Flemer et al. (1983) found 
no re lationship between freshw ater inflow and concluded that hypoxi a increased 
dramaticall y during 1950-1980, probably in response to increased nutrient loading 
rates. The difference may have resulted from each study utilizing slightly different 
subsets of the available data or from their slightly different definitions of hypoxia 
(<0.7 mgr' VS. 1.0 mg r1) . 
The nature of river flow effects on hypoxia in Chesapeake Bay was also a 
major focus of this study. Previous research suggests that summer hypoxic volume in 
Chesapeake Bay is more extensive in years with high river flow (Seliger and Boggs 
1988, Hagy and Boynton 2000). The effect of average river flow on hypoxia is one 
among many effects of 1iver flow on the ecosystem. These effects have generally 
been related to two highly con-elated factors, water column stratification and nutrient 
loading (Boicourt 1992, USGS RIMP, Boynton and Kemp 2000, Chapter 2). The 
64 
extent of the coffelation between 1iver flow and nutrient loading rates to the Bay since 
1978, when direct measurements of nutrient inputs began, is so great that these two 
effects are stati sticall y confounded. Consequently, a purely coffelative approach using 
onl y data for 1978-1999 provides no insight into the relative roles of physical factors 
(not controllable) and nutrient loadi ng-related factors (more controllable) in 
determining the ex tent of hypoxia. However, accounting for river flow effects while 
evaluating trends in hypoxi a for the longer record, 1950-1999, provides a means of 
separating river flow effects from other factors that may have varied during that time, 
such as nutrient loading rates. The observed 2-3 fold increase in annual average 
nitrate (N03) concentrations in the tidal fresh Chesapeake Bay between 1965 (-0.3 mg 
N r1) and 1980 (- 1.0 mg N i- 1; Flemer et al. 1983) provides at least some direct 
indication that loading rates increased substantially during that period. 
Another factor that could have affected the development of hypoxi a over time 
is the interaction between physical transport regimes and changing DO distributions 
within the Bay. Kemp et al. (1992) suggested that decreased oxygen concentrations in 
the lower water column may have increased the physical flux of DO to the bottom 
layer by intensifying DO gradients. Conversely, an increase in DO due to decreased 
DO demand by biota would reverse this effect, dampening further increases in DO 
below the pycnocline. At the same time, Kuo and Neilsen (1987) and Kuo et al. 
(1991) illustrated the role of ho1izontal DO fluxes in determining the development of 
hypoxia. 
This chapter describes the results of a detailed analysis of the DO record for 
Chesapeake Bay since 1950. The extensive measurements of DO collected by the 
65 
Chesapeake Bay Institute (1950-1980) and the Chesapeake Bay Monito1ing Program 
(1984-1999) were combined to estimate the long-term change in hypoxia and the 
effect of river flow as a modulator of that trend, resolving and identifying the causes 
of discordance in previous studies and expanding on the results using data for recent 
years. Hypoxia was defi ned using three different thresholds, 0.2, 1.0 and 2.0 mg r1, 
with the results obtai ned for each being contrasted. For recent years, the extent of 
hypox ia was also eva luated using a time-space integrating approach. Mechansisms 
affec ting the development and maintenance of hypoxia in the mid Bay were explored 
by examining the effect of river flow on the rate of spring-time oxygen depletion (e.g. 
Officer et a l. 1984) and using box model based estimates of physical transport rates 
(Chapter 2). 
Methods 
Sources of Data and Analytical Methods 
This study was based on previously collected water quality and 1iver discharge 
data. The available data and sources are described in Table 3-1. The Chesapeake Bay 
Institute (CBI) di ssolved oxygen data (1950-1980) were collected by pumping water 
from depth and measu1ing DO using a modified Winkler titration (Carpenter 1965a), 
which has an estimated accuracy of0.1 % (Carpenter 1965b). The Chesapeake Bay 
Water Quality Monitoring Program (1985-2000) measured water temperature, salinity 
and dissolved oxygen using an instrument lowered through the water column. Water 
temperature, salinity and dissolved oxygen were measured using a thermocouple, 
conductivity cell, and polarographic oxygen probe, respectively. The exact equipment 
66 
varied by laboratory and over time but was not specified by the monitoiing program's 
data dictionary (EPA 1993). The precision of the oxygen measurements most likely 
depended somewhat upon the instrument used; however, precision probably depended 
to a greater extent on daily calibration and use. Winkler titration was used daily for a 
subset of measurements to ensure accuracy to within ?0.5 mg r 1 (EPA 1993). 
Experience with similar instruments suggests that accuracy was probably better than 
?0.5 mg r1. 
Calculation of Hypoxic Volumes for 1950-1980 
Summertime hypoxic volumes were calculated for each year during 1950_1980 
for which data was available and for each of three definitions of hypoxia, <0.2 mg r1 , 
<l.O mg r 1 and <2.0 mg r 1? The middle level corresponds to the definition utilized by 
Seliger and Boggs (1988). The first level was calculated as an estimate of anoxic or 
near anoxic water volume. Typically there was one cruise per summer in the CBI 
data, usually during July. A July cruise was always used where available. In 1958, 
1960 and 1972, data from August was used. In 1978, data from September was used 
(Table 2). If more than one cruise was available, the highest volume was not selected 
arbitrarily because this would introduce an artifact whereas higher hypoxic volume 
would tend to be observed when more data was availabile. To calculate hypoxic 
volume, the available mid-channel oxygen data were interpolated to a 2-D grid with 
grid ce ll dimensions of 1 m vertical by 1 nautical mile horizontal. The grid 
corresponds to tabulated cross-sectional volumes for Chesapeake Bay (Cronin and 
67 
Pritchard 1975), which were summed for cells with DO less than the specified 
maximum to compute hypoxic volume. 
The interpolation procedure involved a two step process intended to account 
for the vastly greater influence of depth on DO distributions as compared to horizontal 
gradients. First, linear interpolation was used to extrapolate the vertical DO profile at 
each station to a 1 m interval. Subsequently, linear interpolation was used to 
extrapolate horizontally at constant depth. The interpolation procedure prevents any 
extrapolation across unequal depths except within a single vertical profile, a process 
which due to extreme anisotropy almost always produced unreasonable results e 
, ven 
using data with high spatial resolution. In contrast, this algorithm always produced 
good results when sampling was thorough, as was always the case with the 
Chesapeake Bay Monitoring Program data (1984-2000). Sometimes good results were 
obtained with more sparse data; however some paiticular data gaps led to interpolation 
problems. These were identified using contour plots generated for each year p,ior 
1985. To con-ect these problems, adjustments were made to the interpolated profiles in 
the following specific situations: 
(1) When DO was not measured at the lower extent of the pycnocline, but was 
measured within the mixed layer, the concentration within the lower mixed layer 
was assumed to extend upward to the lower extent of the pycnocline. 
(2) When DO was not measured within the bottom mixed layer, the DO profile within 
the bottom mixed layer was interpolated from adjacent up-Bay and down-Bay 
stations, rather than interpolating vertically through the pycnocline. 
68 
(3) Unless data were present to indicate otherwise, the DO contours near the northern 
extent of the channel trench were assumed to extend unmodified at constant depth 
to the up-estuary extent of the trench. 
These modifications resulted in estimated DO distributions that were structurally 
consistent with DO distributions observed when sampli ng was thorough (i.e. 1985-
1999). In addition , the estimated distributions were plausible given the known water 
column structure and approximate physical transport regime. 
Calculation of Hypoxic Volumes for 1984-1999 
The interpolation algorithm that was used to interpolate 1950-1980 DO 
profiles was also used for 1984-1999, except that adjustments for incomplete data 
were never needed because the sampling regime of the monitoring program was 
thorough and consistent. In addition, because two July cruises and co1Tesponding 
hypoxic volume estimates were available for each year, the average of the two values 
was used to determine an estimate of July hypoxic volume for each year. 
For 1985-1999 there was a sufficient number of sampling dates each year to 
characterize the seasonal time-series of hypoxic volume within each year. This 
increased temporal resolution allowed a time-space integrated measure of hypoxia to 
be computed by calculating trapezoidal areas underlying a linear-spline fitted time-
series of hypoxic volume (Fig 3-2). 
Oxygen Input Budgets 
69 
Oxygen input budgets were developed for the lower water column in three 
segments of the mesohaline Chesapeake Bay (Fig. 3-1; Fig 3-3). These bud t 
ge s were 
used to compute and compare vertical non-advective, net horizontal advectiv d 
e,an 
total DO inputs to the lower water column in July during 1950-1999 July rat 
? es were 
computed because most of the early data was for mid summer In addition 
? , 0 xygen 
concentrations typically approached steady-state in July. Therefore, DO inputs to the 
lower layer in summer approximate DO consumption rates . 
For each segment, m, vertical non-advective DO input was calculated using 
Ev111 (c:,, - c111 ), where E11111 is the non-advective vertical exchange coefficient and c' 
/1/ 
and c,11 are the bottom and surface layer mean DO concentrations, respectively. 
Advective DO transport was calculated as the product of advective transport and the 
DO concentration in the originating model segment (Fig. 3-3). Since the water 
balance equation Q:,,+l = Q;,, + Qvm holds for each segment, the net advective DO 
input (input minus outputs) can be computed as Q;,1+1 (c:,1+1 - c;,,). Total DO input 
for each segment is the sum of the vertical non-advective DO input and the net 
advective DO input. All inputs were expressed per unit of pycnocline area (i.e. m-2). 
River flow dependent estimates of Evm and Q;,, were computed using regression 
models (Chapter 2). 
Statistical Methods 
The long-term change in hypoxic volume as well as the effect of iiver flow 
was evaluated using multiple regression. The characteristics of the data required 
special consideration. First, the variability in hypoxic volumes increased with the 
70 
... 
mean. The was addressed via log-transformation. Secondly, the long-term change in 
hypoxic volume was not constant over time. This was addressed by addition of an 
exponent to the time-term of the model, effectively a shape-parameter. Finally, 
because hypoxic volume was zero in some years, recoding was needed to permit log-
transformation. The resulting model was 
V' = {30 + {31 (T -1949)a + /J2;Q (5) 
where V' = \n(V + 1) and Vis the hypoxic volume. Asymmetric confidence bands for 
the predicted means were generated by computing confidence limits on the log-scale, 
then back transforming. 
Results 
Hypoxia in Chesapeake Bay, 1950-2000 
A plot of the hypoxic volumes shows that summer hypoxic volumes increased 
substantially during 1950-2000 (Fig. 3-4). Regression analysis was used to quantify 
the rate of increase of hypoxia through time while simultaneously investigating and 
controlling for suspected effects of variations in average spring river flow on hypoxia 
(Seliger and Boggs 1988, Boicourt 1992). The optimal period of time over which to 
average river flow was evaluated by examining several periods. For example, the 
April-May average river flow used by Seliger and Boggs (1988) was considered, as 
was the January-March average flow . The January-May average Susquehanna River 
flow was included in the fina l model to remove the effect of 1iver flow from the trend 
estimate because it explained more of the flow dependence than did either of the other 
two averages. Incorporating several averages in the model by estimating several 
71 
model coefficients was also infe1ior to the longer averaging period. Log-
transformation resulted in approximately homoscedastic and normally distributed 
residuals. Residuals were also not autocon-elated. 
According to these regression models, mid-summer hypoxic volume increased 
significantly and dramatically du1ing 1950-2000, while the level of hypoxia in any 
3 4
given year was also significantly dependent on spring river flow (Table 3-3, Fig. _ , 
Fig. 3-5). The river flow effect was most pronounced for severe hypoxia and least 
pronounced for moderate hypoxia. At average river flow levels (1,500 m's'), the 
estimated vo Iu me of near-anox ic water increased from zero in 19 50 to 3. 7 x 10' m, in 
2000 (Fig 4). The estimated volume of severe IY  hypoxic water increased 4-foI d from 
l.7x 10 m to 6.5xl09 m3. The estimated volume of moderately hypoxic water 9 3 
9 3
increased 2.7-fold from 3.6xl09 m3 to 9.7x10 m . Correlation between estimates of 
the exponent, a, and the slope fJ1 inflated the confidence intervals for these model 
parameters. Therefore, the significance of the trend overall was evaluated by 
computing confidence intervals for model predictions (means?sd) at several levels of 
river flow (Fig. 3-5). In addition, 95% confidence limits were computed. As an 
example, the 95% confidence interval for moderate hypoxia in 2000 does not overlap 
with the corresponding intervals for all years up to 1980. Similarly, the 95% 
confidence interval for near anoxia in 2000 does not overlap with the con-esponding 
3 5
interval for years up to 1985. The graph of the predicted hypoxic volumes (Fig. _ ) 
also illustrates the increasing rate of expansion of hypoxia volumes. Due to the log-
transformation and non-linear functional form, this is difficult to describe. The rate of 
increase for near-anoxia was greater than for moderate hypoxia in the early portion of 
72 
, e annual 
the record, but by the end of the record the opposite was true. For example th 
ypoxia. In 
increase in near-anoxia in 1960 was 2.8 times more than that for moderate h . 
or near-
contrast, the annual increase in moderate hypoxia in 1990 was 2.1 times that f 
anoxia. 
The spatial pattern of increase in hypoxic volumes was examined by 
0 
contrasting DO profiles in Chesapeake Bay for years with approximately averaoe 
spring flow (Fig 3-6). As part of this study, a complete set of similar profiles was 
5
produced for all years between 1950 and 1980 and for selected years during 198 -
140 
1999. In July 1959, the 1 mgr' DO contour intersected the channel bottom at km 
3
from the Atlantic Ocean, between Point Lookout, MD and Patuxent River (Fig. _1 , 
Fig. 3-6). There were no DO observations less than 0.2 mgr' . The July 1970 DO 
profile was particularly different from the 1959 profile in that near anoxic or anoxic 
water was present. In J980, the 1 mgr' contour moved down-estuary to - JOO km 
from the Atlantic Ocean (Smith Point, VA), while the 2 mgr' contour had changed 
little from 1959. The 0.2 mgr' contour was greatly expanded such that in most places 
the 0.2 mg r' contour was very near to the J.0 mgr' contour. In 1991 , both the l.O 
and 0.2 mg r' contours extended to a shallower depth in the mesohaline region as 
1 0 
com pared to 9 . particularly of note was the extension of the 1. 0 mg r and 2. mg 
1 80
r' contours to include areas further to the south, -60 km from the Atlantic Ocean near 
the mouth of the Rappahannock River (Fig 3-1, Fig. 3-6). The most recent years in the 
record, and 99, inc Iu ded a particu Ia rl y dramatic expansion of the hypoxic zone 
199 8 19
to the south. In , the 0.2 mgr' contour extended to between 60 and 240 km from 
1998
the Atlantic Ocean, or from the Bay Bridge to as far south as the Rappahannock River, 
73 
and included a ll the waters up to - 10 m depth. In 1999, a year with low spring 
Susquehanna Ri ver fl ow, near-anoxic water in the mesohaline was greatly reduced, 
1 
but a region wi th D0<2.0 mg r extended to within 30 km of the Atlantic Oce 
an, 
separated from the mesohaline Bay by a region of normoxic (i.e. D0>2.0 mg ri) 
water. 
Over 1950- 1999 the ratio between the calculated hypoxic volume and h . 
ypox1c 
benthic area, essentially the mean depth of the hypoxic volume, varied between m 
5 
and 7 m, with no significant trend over time. It can be infeITed that the trend in 
hypox ic benthic area, to the extent that it can be addressed with thi s data , was the 
same as the trend in hypoxic volume. The estimate of hypoxic benthic area may be a 
minimum es timate because the benthic environment can become hypoxic due to strong 
local DO demand, even as the overlying water remains normoxic. 
~patia l and Tempora l Dimensions of Hypoxia 
The water quality data avai lable for 1985-1999 were sufficient to quantify both 
the spatial and temporal extent of hypoxia in Chesapeake Bay for that period by 
integrating hypoxic volume through the year (Fig. 3-2, Fig. 3-7). These time-space 
integrals were calcu lated using the same three DO thresholds previously discussed, 
1 
D0<0.2 mg ri, DO<l.O mg r and D0<2.0 mg r1, and were then related to iiver flow 
and time. Un like July hypoxic volume for 1950-1999, this shorter (15 year) time 
series was re lated to spring average river flow but there was no significant increase or 
decrease over time. As a result, the time component was dropped from the final model 
3 1
(Table 3-4). At average flow (1500 m s- ), the expected amount of hypoxia (DO<l.O 
74 
r1 10 mg ) was 43 .5xl0 m3-days, while this increased to 56.8xl0 10 m3-days with flow 
equal to 2200 m3 s?1. Time-space integrals of hypoxic volume have been criticized as 
management tools because, as a somewhat abstract concept, they are not particularly 
engaging fo r environ mental managers or the public. However, the temporal 
dimension of hypox ia that they incorporate is important. In particular, increases in the 
temporal dimension of hypoxia tend to occur via earlier onset of hypoxia in late 
spring, just when new recruits in the benthic community is beginning to respond to 
increased water temperatures. 
Taft et al. (1980) calculated the rate of dec line of bottom water DO at a 
mesohaline station from winter levels to the minimum concentration during summer 
for 1964-1971 and 1977. The dec line was typically nearly linear and proceeded at 
rates between 0.04 to 0.15 mg r1 d-1, with an average of 0.106 mg r1 d-1 (Table 5). In 
this study, the series was expanded to include 1985-1999 using bottom water DO data 
from station CB4.3C, also in the central mesohaline Bay (Fig. 1). The rate of DO 
dec line was estimated by fitting a linear regression to the 3-5 observations describing 
the spring decline in DO. The date of onset of anoxia was determined with greater 
precision than the bi -weekly sampling program would otherwise permit by 
2
extrapolating this line to zero DO. Typically, these regressions had r >0.95. 
1
Regression slopes averaged -0.119 mg r1 d- , which is a slightly more rapid decline 
than was observed for 1964-1977 (Table 3-5). Assuming the typical late winter DO 
1 
concentration of 9.0 mg ,-1, this 0.013 mg r1 d- increase in the rate of DO depletion 
leads to onset of anoxia -9 days earlier during 1985-1999 than during the late 1960 's. 
Compared to the variation within the recent and older data, however, this difference 
75 
may be regarded as insignificant. The rate of DO depletion was not significantly 
correlated with spring average river flow during either pe1iod, or in both periods 
combined. 
In contrast to the above results, the date of onset of bottom water anoxia was 
significantly related to spring average Susquehanna river flow (p<0.01; Fig 3-8). 
Du,ing 1985-1999 this date varied between May 2 and July 5 (Table 3-5). January to 
May average river flow was the best predictor of the date of first anoxia. January-
March or April-May average flow did not predict or were substantially inferior 
predictors, respectively. Assuming that it was correctly determined that DO did not 
decline faster at high river flow, but did reach hypoxic levels earlier, one or both of 
two possibilities are suggested: the DO decline either began earlier or started at a 
lower initial DO, or both. It was hypothesized that earlier declines and/or lower initial 
DO might be related to warmer water temperatures in early sp1ing. To examine this, 
March bottom water DO and March bottom water temperature were added to river 
flow as explanatory variables in a multiple regression predicting the date of onset of 
hypoxia. 
Water temperature (p<0.05) along with spring river flow (p<0.01) explained -55% of 
the variance via the multiple regression 
D = 241- 0.028Q - 7.26T (6) 
Where Dis the calendar date on which anoxia first occurred, Q is the January-May 
average  n?  ver fl ow ( m 3 s -1) an d T 1?s the March bottom water temperature at CB4.3C 
76 
(Fig 3-8). March bottom water DO was negatively corTelated with bottom water 
temperature, indicating that the effect of temperature was to lower the initial DO level. 
Linear decline of DO did not tend to begin as early as March , perhaps due to the 
strength of storms at that time of year. According to eq . 6, an increase in average 
spring river flow of 1000 m3 s-1 results in anoxia occun-ing 28 days earlier. Similarly, 
the observed range in March bottom water temperature (3.1 - 6.8 ?C) was associated 
with a 27 day change in the date of onset of anoxia. Because of the rapid increase in 
water temperature during spring, the date of onset of anoxia dramatically affected the 
range of water temperatures avai lable to the benthos prior to the onset of anoxia. As 
an example, eq . 6 predicts that bottom water reaches anoxia on 6/15 when Q=l200 m3 
1 
s? and T=5.5?C. Bottom water temperature on 6/ 15 is - l8 ?C and has exceeded J5 ?C, 
sufficient for increased activity by the benthos, for about 22 days (Fig. 3-8). In 
contrast, with Q=2200 m3 s?1 and T=5.5 ?C, anoxia occurs on 5/18, by which time 
bottom water temperature is only -14?C, living no window at all for growth ofbenthic 
opportunists. The effect of water temperature shifts on the water temperature at 
anox ia was relatively smaller, especially considering that March temperature 
anomalies tended to persist throughout spring. 
Qxygen Input Budgets 
0 xygen r? npu t s b ud gets were used to examine the relative importance of the 
two sources of DO to the lower water column, horizontal advection and vertical 
diffusion. Of particular interest was how these sources may have changed over time 
as DO distributions in the Bay changed. The computations showed that in mid 
77 
summer, for the mid Bay as a whole, the input of DO due to horizontal advection of 
water from the lower Bay (Jess DO advected out of the lower water column) was 2-
3 
fold g reater than vertical diffusive inputs (Fig. 3-9). Vertical diffusive inputs 
increased slightly over 1950-1999. In recent years, the effect of river flow on vertical 
diffusivity was particularly evident in vertical diffusive DO inputs , reflecting both 
large interannual va1iations in freshwater inflow and substantial differences between 
DO above and below the pycnocline. Net advective DO inputs decreased over 1950-
1999. Although landward advection increased with spring freshwater input rate, net 
advective DO inputs reflected the gradient in DO concentration at least as much as the 
difference in advection. The lowest advective input rate occuJTed in 1999. In this 
year, freshwater inflow was at a record low in Maryland, but was not as low in 
Virginia. Hypoxia extended well outside the mesohaline region of the Bay. By 
depleting the DO concentration in the inflowing water, the advective input was 
decreased greatly. 
Discussion 
htterns and Trends in Hypoxia, 1950-2000 
The results of this study show that the volume of water affected by hypoxia in 
mid-summer in Chesapeake Bay increased dramatically over 1950-2000 and that this 
long-term trend, although modulated by average sp1ingtime 1iver flow, was not caused 
by an increase in river flow. Moreover, the increase appears to have accelerated, 
especiall y in the case of milder hypoxia (D0<2.0 mg r1) as compared to near-anoxia 
(D0<0.2 mg r1). Whereas the principal change during 1950-1980 was that anoxia, 
78 
which was previously not a regular feature of Chesapeake Bay, became an annual 
certai nty, the principal change during 1980-2000 was that milder hypoxia expanded 
dramatically to the south as compared to the region previously affected. The results of 
earlier s tudies (i.e. F lemer et al. 1983; Seliger and Boggs 1988) may have depended 
upon particular issues of interpolation methods, inclusion or exc lusion of data from 
certai n years, and the means by which 1i ver flow effects were included in the analysis . 
In contrast, the basic conclusion of this study that hypoxia has increased si nce 1950 
' 
independentl y from river flow, is insensiti ve to the choice of any reasonable 
interpolation method and is robust to inclusion or exclusion of observations from any 
particularly year. 
The trend in hypoxia du1ing 1985-2000 was less obvious only because the 
analyses based only on data for thi s period indicated no significant time trend. In 
contrast, within the context of the longer time series it appears that the upward trend in 
hypoxia was accelerating, rather than stabilizing, in recent years. The large 
interannual variability, which could not be overwhelmingly related to river flow, may 
have weakened the power to resolve time trends from the shorter time se,ies. If the 
volume affected has in fact increased since 1985, it appears that much of this increase 
may be in the form of milder hypoxia at the southern extent of the mesohaline region , 
or even within the polyhaline region of the Bay (Fig. 3-4, Fig. 3-6). Perhaps not 
coincidentally, this pattern of expansion to the south corresponds directly to the 
pattern of increase in phytoplankton biomass in the southern Bay, observed for 1950-
1994 by Harding and Perry (1997). Both patterns may be related to a long-term 
increase in nutrient inputs, while the increased phytoplankton production documented 
79 
by Harding and Perry (1997) may be the major contributor to increased oxygen 
depletion in the lower Bay. 
There are few systematic records of hypoxia for other estua,ies in the world 
With comparable detai l and duration. One example is the Northwest Shelf (NWS) area 
of the Black Sea in which nitrogen and phosphorus loading increased 56% and 400% 
respectively between the 1960 's and 1980's (Zaitsev 1992). Seasonal hypoxia in 
bottom waters on the NWS was first recorded in the late 1960 's and increased -5-fold 
by the 1980 's, affecting waters between 7-8 and 35-40 m. Zaitsev (1992) also noted 
that in some recent years, meteorological conditions led to hypoxia persisting for only 
a few days. In contrast, hypoxia persisted in some places for most or all of the 
summer in Chesapeake Bay. Other shallow water areas are also becoming subject to 
increased hypoxia. For example, Rabalais et al. (1999) has shown that the hypoxic 
zone beneath the Miss issippi River plume in the Gulf of Mexico has increased in 
recent years. Shallower areas of the Baltic Sea region have also begun to experience 
hypoxia (Baden et al. 1990 cited by Diaz and Rosenburg 1995). 
A huge area of the central Black Sea, which exceeds to 2000 m depth, is 
permanently and probably naturally anoxic. Permanent anoxia is also present in a 
large and growing area of the Baltic Sea at >65 m depth (de Jonge et al. 1994) and 
certain areas of the coastal ocean (e.g. Arabian Sea, Peruvian Shelf; Diaz and 
Rosenburg 1995). Such persistent hypoxia and anoxia is not uncommon worldwide 
and appears to be increasing due to anthropogenic effects (Diaz and Rosenburg 1995). 
However, hypoxia and anoxia in shallow waters may be of particular importance 
because of the profoundly negative consequences on these otherwise productive 
80 
habitats which are tightly coupled to heavily exploited and valuable demersal and 
pe lagic fi sheries. For example, Zaitsev (1992) concluded that the effect of hypoxia in 
the NWS of the Black Sea was to eliminate 100-200 tons wet weight of benthic 
macrofauna and demersal fishes per km2 (20-40 g AFDW m-2, assuming 0.2 g 
AFDW/g WW). Sturgeon and turbot catches in the NWS region decreased to 22% 
and 10% of earlier levels, respectively (Zaitsev 1992). 
Sionificance of Increased Hypoxia 
t:> 
There are no studies that have shown that declines in commercial fishe1ies as 
' 
mentioned above for the Black Sea, have occurred in Chesapeake Bay as a direct 
result of increased hypoxia and anoxia. However, there are several other ways that the 
significance of the increase in Chesapeake Bay hypoxia since 1950 could be 
evaluated. The fraction of total water volume and benthic habitat that is affected is 
one indicator of significance of the observed increase in Chesapeake Bay hypoxia. In 
1950, severe hypoxia (DO<l.0 mg i- 1) affected about 20% of the volume and sediment 
area below 10 m depth in the mesohaline Bay. Compared to the overall volume of the 
mesohaline region or to the entire Bay, the hypoxia-affected fraction was small, 
approximately 6% and 2.5%, respectively. For a year with average river flow in the 
\ate 1990's, however, hypoxia affected a volume and area equal to the entire 
mesohaline region below 10 m and equal to 46% of the total area and 26% of the total 
volume in the mesohaline Bay. Compared to the entire mainstem Chesapeake Bay, the 
areas affected by moderate hypoxia constituted a significant fraction of both the 
sediment area (18%) and total water volume (12%). 
81 
In Chesapeake Bay, the results of thi s study suggest that benthic and pelagic 
habitats at all depths may have been productive as late as the 1950 's, even if the deeper 
benthic habitats were limited to hypoxia-tolerant species and subject to peiiodic 
hypoxia induced mortality in high flow years (e.g. 1952). However, the increase in 
near-anoxia in Chesapeake Bay since 1950 probably degraded deeper habitats by 
l 980. The presence of near-anoxic water near the pycnocline since 1980 (Fig. 3-6) 
could al so have degraded shallower habitats by increasing the effect of apeiiodic 
events that bring below-pycnocline waters onto shallower flanks of the Bay (Malone 
et al. 1986, Breitburg 1990, Sanford et al. 1990). While benthic communities can 
potentially survive moderate hypoxia until normoxic conditions return by using a 
variety of adaptive strategies (Diaz et al. 1992), mortality tends to be more rapid under 
more severe hypoxia and in the presence of H2S (Diaz and Rosenburg 1995). Thus, 
While aperiodic events bringing below-pycnocline waters into shallower habitats have 
probably always occurred, the detrimental effects of these events may now be much 
greater due to anoxia. 
The increase in anoxia has probably affected the biogeochemistry of the Bay, 
perhaps accounting in part for the accelerating increase in hypoxic volume. Based on 
the estimated effects of anoxia on N and P efflux from sediments, and the estimated 
increase in the sediment area affected by anoxia, the effect of increased anoxia on 
nutrient dynamics in Chesapeake Bay is probably significant. The P efflux from 
anoxic rnesohaline sediments in mesohaline Chesapeake Bay is 50-150 ?mo! m-2 h-', 
While p efflux from oxic sediments is apparently negligible (Cowan and Boynton 
8 2
1996). For the estimated anoxic sediment area in the 1990's, - 5.7(10 ) m , this 
82 
4 1
amounts to an additional 2. 1-6.3(10 ) kg P d- , - 3-10 times recent annual average p 
loading rates from Susquehanna Ri ver to Chesapeake Bay. Similarly, anoxia prevents 
coupled nitrification-denitrification, which occurs at significant rates in warm, organic 
ri ch, but oxic sediments (Seitzinger 1988). Summertime~+ efflux from the 
mesohaline Bay sediments in recent years were substantially greater than elsewhere in 
the Bay (Cowan and Boynton 1996), perhaps reflecting thi s limitation of coupled 
nitrification-denitrification by anoxia. Assuming a reasonable potential for 
2 1 
summertime denitrification is 50-100 ?mol m- h- (Seitzinger 1988), the additional 
NH,/ 3efflux associated with anoxia amounts to 9.6-19.2(10 ) kg N d-1, or-10% of the 
annual average and 11-23 % of summer average Susquehanna River total N loading 
rates for 1984-1996. The N efflux may be especially important due to the close spatial 
proximity to the N-limited phytoplankton community in the surface layer during 
summer (Fisher et al. 1992). 
l&ng-Term and River Flow Dependent Changes in Oxygen Transport 
It has long been recognized that DO patterns reflect a combination of several 
biological and physical processes. Kuo et al. (1991) used a Lagrangian model to 
examine DO distributions in the Rappahannock River. One of their major conclusions 
Was that residual advective velocity was an important factor determining the 
maximum extent of DO depletion, principally because it determined the residence 
time of a parcel of water in the lower water column. Kemp et al. (1992) used an 
entire! Y dI" f"i" erent approac h , 1?n "en-ing physical transport rates from the balance of 1, 
measu 1-e d b1' 0 1o g1? ca 1 rates. They concluded that both transport processes were 
83 
important, but that vertical diffusion was 4-fold more important in two of three 
summer months. In contrast, this study found that horizontal advective inputs were 
usually the larger flux, although not by a large margin (Fig 3-9). Amazingly, the 
residual current velocity Kemp et al. (1992) infeJTed from biological rates was similar 
to the rate estimated for this study using a box model, about 5 cm s- 1 (Chapter 2). 
Therefore, it is not surpri sing that their estimated horizontal DO transport rates were 
not substanti ally different from the present estimates. However, their estimate of 
vertical diffusivity, about 0.2 cm2 s- 1, was at the high end of the flow-dependent range 
of rates esti mated for this study (0.09-0.24 cm2 s-', Chapter 2). Due to lower estimates 
of Dz the vertical DO flux was estimated to be smaller, lending relatively larger 
importance to the ho,izontal advection of DO. The results of this study also reveal, 
however, that these processes vary either inversely or independently, but they are not 
PositiiveJy con-elated. Therefore, their relative importance varies. Location within the 
Bay was also very important. At the north end of the mesohaline region of the Bay, 
Veitical diffusion was the only source of DO when the advecting source water was 
ano xi?c  or nearly anoxic. 
Of greater interest here was how the relative importance and total magnitude of 
these DO fluxes changed during 1950-1999. Here, physical transport rates were a 
noi sy , but essenti.a  11 y stati.o na.r y ba c kground , i ?e  ? neither vertical diffusivity nor the 
residual cuiTent velocity changed directionally over this period of time. Directional 
changes ? t he  ph ys1. cal transpo1 .t  of  DO (Fig ? 3-9) were therefore due to changes in the in 
spatial d. . . . . . h B y The increase in hypoxia over time increased 
1stnbut1on of DO w1thrn t e a ? 
Vertical ct1?f f usI.O n of DO pred1. cta bl Y( F.i g. 3-9) ? However, the average surface layer 
84 
DO concentrations decreased somewhat, despite high DO in the top several meters due 
to high rates of phytoplankton production (not shown). This was likely due to 
increased DO demand at the pycnocline, due to both downward DO diffusion , and, 
under the worse anoxic conditions, scouring of DO within the pycnocline by hydrogen 
sulfide (Kemp et al. 1992). As a result, vertical di ffusion of DO increased only 
slightl y during 1950- 1999, and with great variability. In contrast, horizontal advective 
inputs into the mid Bay decreased more substantially (Fig. 3-9), reflecting the 
previously described southerly expansion of hypox ia (Fig. 3-6). The net effect of 
changes in the spati al di stribution of DO during 1950-1999 has been to exacerbate the 
depletion of DO in the mid Bay by decreasi ng the DO flux due to advection. At the 
sca le of the mid Bay, thi s observation counters the observation of Kemp et al. (1992) 
that the magnitude of DO response to changing DO inputs would be buffered by 
coupled biological -physical interactions, leading to initially rapid, but proportionately 
decreasing improvements in water quality. Rather, these results suggest that the 
biologically-mediated physical transport effects act together with the positive 
feedbacks due entirely to biological processes (e.g. sediment NH4+ , Pol-and H2S 
efflux). The combined positive feedbacks likely contribute to the increasing rate of 
expansion of the near anoxic zone that has been described (Fig. 3-4). At the larger 
scale of the whole Bay, the DO flux to the bottom layer must decrease as average DO 
levels increase. However, the inputs to the mid Bay, the only place where severe 
anoxia and associated effects have occun-ed, are the most relevant. The implications 
for management are favorable. Since all the feedback processes appear to be positive, 
Water quality changes associated with decreased nutrient loading should bring about 
85 
significant beneficial effects , reversing the unfavorable trend followed to date. The 
dynamically unstable configuration of feedbacks (i.e. all feedbacks are positive) 
suggests why the unexplained variability in hypoxic volumes is so large. Prediction is 
possible only because of the very large range in ecosystem forcing in recent years 
(river flow and flow-dependent nutrient loading) and over the long-term (river flow 
and nutrient loading both dependent on and independent of flow). 
?.rospects for Restoration 
While this study does not give estimates of nutrient loading rate targets needed 
to reach hypoxia-related restoration goals, the detailed characterization of the temporal 
pattern in hypoxia will be useful as estimates of historical loading rates become 
available. An effort to obtain these estimates is underway. The results strongly 
suggest that some decrease in hypoxia should be expected from any decrease in 
nutrient loading and that any improvement in water quality will promote additional 
improvement. That said, interannual variability can be expected to remain large. 
Therefore, the restoration effort would need to proceed with persistence, rejecting 
suggestions that current levels of hypoxia are a natural feature of Chesapeake Bay 
ecology or that hypoxia do not have significant detremental effects on living resources 
(Chapter 5, Chapter 6). 
In upcoming years, restoration efforts will face mounting challenges regarding 
management of hypoxia. Not only do increasing human pressures on the landscape 
make it difficult to limit nut,ient loading rates, climate change may also affect hypoxia 
in Chesapeake Bay. Najjar (1999) concluded that annual discharge from the 
86 
Susquehanna River could be expected to increase 24% in response to an expected 
doubling of atmospheric CO2 and associated global warming. This study showed that 
both increased river flow and warmer winter water temperatures hasten the formation 
of anox ia in Chesapeake Bay during late spring and promote the maintenance of 
hypoxi a during sum mer through both biological and biologically-mediated physical 
transport effects . 
Literature Cited 
Baden , S. P., L. Pihl , and R. Rosenburg. 1990. Effects of eutrophication on benthic 
communities including fish - Swedish west coast. Ambio 19: 113-122. 
Boicourt, W. C. 1992. Influences of circulation processes on dissolved oxygen in the 
Chesapeake Bay. In: Smith, D. E., M. Leffler, and G. Mackiernan (eds.). 
Oxygen Dynamics in the Chesapeake Bay. A synthesis of recent research. 
Maryland Sea Grant College, College Park, MD. 
Boynton, W. R. and w. M. Kemp. 2000. Influence of river flow and nutrient loads on 
selected ecosystem processes. A synthesis of Chesapeake Bay data. pp. 269-
298 In : J.E. Hobbie (ed.). Estuarine Science: A Synthetic Approach to 
Research and Practice. Island Press, Washington, DC. 
Breitburg, D. L. 1990. Nearshore hypoxia in the Chesapeake Bay: Patterns and 
relationships among physical factors. Est. Coast .. Shelf Sci. 30(6): 593-609. 
Breitburg, D. L., T. Loher, c. A. Pacey, and A. Gerstein. 1997. Varying effects of 
low dissolved oxygen on trophic interactions in an estuarine food web. Ecol. 
Monogr. 67(4): 489-507. 
Carpenter, J. H. 1965. The Chesapeake Bay Institute technique for the Winkler 
titration. Limnol. Oceanogr. 10(1): 141-143. 
Carpenter, J. H. 1965b. The accuracy of the Winkler method for dissolved oxygen 
analysis. Limnol. Oceanogr. 10(1): 135-140. 
Cowan, J. L. w. and w. R. Boynton. 1996. Sediment-water oxygen and nutrient 
exchanges along the longitudinal axis?~ Chesapeake B_ay: Seasonal patterns, 
control ling factors and ecological signrf1cance. Estuanes. 19(3): 562-580. 
.... 87 
Cronin, W. B. and D. W. Pritchard . 1975. Additi onal statistics on the dimensions of 
Chesapeake Bay and its tributaries: Cross section widths and segment volume 
per meter depth. Chesapeake Bay Institute. Reference 75-3. Special Report s 
42. 
de Ionge, V. N ., W. Boynton, C. F. DEii a, R. Elmgren, and B. L. Welsh. 1994. 
Responses to developments in eutrophication in four different North Atlantic 
estuarine systems. In : Dyer, K. R. and R. J . Orth. (eds.). Changes In Fluxes In 
Estuaries: Implications From Science To Management. Olsen & Olsen, 
Fredensborg, Denmark. pp. 179-196. 
Diaz, R. J ., R. J. Neubauer, L. C. Schaffner, L. Pihl , and S. P. Baden. 1992. 
Continuous monitoring of di ssolved oxygen in an estuary experiencing 
periodic hypoxia and the effect of hypoxia on macrobenthos and fish. Sci. Tot. 
Env., Supplement 1992. pp. 1055-1067. 
Diaz, R. J. and R. Rosenberg. 1995. Marine benthic hypoxia: A review of its 
ecological effects and the behavioral responses of benthic macrofauna. 
Oceanogr. Mar. Bio. Ann. Rev. 33: 245-303. 
EPA 1993. Guide to using the Chesapeake ~ay Program_W ater Quality Monitoring 
Data. Chesapeake Bay Program Office, Annapolis, MD. 
EPA 2000. Chesapeake 2000 Bay Agreement. US EPA Chesapeake Bay Program, 
Annapolis , MD. 
Fisher, T. R. , E. R. Peele, J. W. Ammerman, and L. W. Harding, Jr. 1992. Nutrient 
limitation of phytoplankton in Chesapeake Bay. Mar. Ecol. Prag. Ser. 82: 51-
63. 
Flemer, D. A., G. B. Mackieman, W. Nehlsen, V. K. Tippie, R. B. Biggs, D . Blaylock 
N. H. Burger, L. C. Davidson, D. Haberman, K. S. Price and J. L. Taft. 1983. ' 
Chesapeake Bay: A Profile of environmental change. U.S. EPA Chesapeake 
Bay Program Report. Annapolis, MD. 
Hagy, J . D . and w. R. Boynton . 2000. Controls on hypoxia in Chesapeake Bay and 
its major tributaries, pp. 91-101. In: Kemp, W. M., S. Blumenshine, J. D. 
Hagy and w. R. Boynton. Ecosystem ~~dels of the _C~esapeake Bay Relating 
Nutrient Loadings, Environmental Cond1t1ons, and Lrvmg Resources. EPA 
Chesapeake Bay Program, Annapolis, MD. 
Harding, L. w. and E. s. Pen-y. 1997. Long-term increase of phytoplankton biomass 
in Chesapeake Bay, 1950-1994. Mar. Ecol. Prag. Ser. 157: 39-52. 
88 
Kemp, W. M ., P.A. Sampou, J. Garber, J. Tuttle, and W . R. Boynton . 1992. 
Seasonal depletion of oxygen frombottom waters of Chesapeake Bay: roles of 
benthic and planktonic respiration and physical exchange processes. Mar. 
Ecol. Prag. Ser. 85: 137-152. 
Kuo , A. Y ., K. Park, and M. Z. Moustafa. 1991. Spatial and temporal variabi lities of 
hypoxia in the Rappahannock River, Virginia . Estuaries. 14(2): 11 3-121. 
Malone, T. C., W. M. Kemp, H. W. Ducklow, W.R. Boynton, J. H. Tuttle and R. B. 
Jonas. 1986. Lateral va,iation in the production and fate of phytoplankton in a 
partially stratifi ed estuary. Mar. Ecol. Prag. Ser. 32(2-3): 149-160. 
M alone, T. C. 1992. Effects of Water Column Processes on Di ssolved Oxygen, 
Nutri ents, Phytoplankton and Zooplankton. In : Smith, D. E., M. Leffler, and 
G . Mac kiernan (eds.). Oxygen Dynamics in the Chesapeake Bay. A synthesis 
of recent research. Maryland Sea Grant College, College Park, MD. 
Najjar, R. 1999. The water balance of the Susquehanna River Basin and its response to 
c limate change. J. Hydro!. 2 19:7-19. 
Officer, C. B. , R. B. Biggs, J. L. Taft, L. E. Cronin, M.A. Tyler, and W.R. Boynton. 
1984. Chesapeake Bay anoxia: O,igin, development, and significance. 
Science. 233: 22-27. 
Rabalais, N. N ., R . E. Turner, D. Justic, Q. Dortch and W. J. Wiseman, Jr. 1999. 
Characterization of Hypoxia. Topic l Report for the Integrated Assessment on 
Hypoxia in the Gulf of Mexico. NOAA Coastal Ocean Program Decision 
Analysis Series No. 15. National Oceanic and Atmospheric Administration. 
Silver Spring, MD. 
Ritter, C. and P.A. Montagna. 1999. Seasonal hypoxia and models of benthic 
response in a Texas bay. Estuaries 22(1): 7-20. 
Sanford, L. P., K. G . Sellner, and D. L. Breitburg. 1990. Covariability of dissolved 
oxygen with physical processes in the summertime Chesapeake Bay. J. Mar. 
Res . 48: 567-590. 
Se itzinger, S. P . 1988. Denitrification in freshwater and coastal marine ecosystems: 
Ecological and geochemical significance. Limnol. Oceanogr. 33(4, part 2): 
702-724. 
Seliger, H . H. and J . A. Boggs. 1988. Long term patterns of anoxia in the Chesapeake 
Bay. In: M . Lynch and E. C. Krome (eds.). Understanding the Estuary: 
Advances in Chesapeake Bay Research . Publication Number 129. 
Chesapeake Research Consortium. Solomons, MD. 
89 
Taft, J. L. , E. O. Hartwig, and R. Loftus. 1980. Seasonal oxygen depletion in 
Chesapeake Bay. Estuaries. 3: 242-247. 
Welsh, B. L., R. I. Welsh, and M. L. DiGiacomo-Cohen. 1994. Quantifying hypoxia 
and anoxia in Long Island Sound. In: Dyer, K. R. and R. J. Orth. (eds.). 
Changes In Fluxes In Estuaries: Implications From Science To Management. 
Olsen & Olsen, Fredensborg, Denmark. pp. 131-137. 
Zaitsev, Y.P. 1992. Recent changes in the trophic structure of the Black Sea. Fish. 
Oceanogr. 1(2): 180-189. 
90 
Table 3-1. Data types, spatial resolution, temporal resolution and sources of data. 
Data Type (Period Spatial Temporal Source 
of Record) Resolution Resolution 
Chesapeake Bay -10 stations on 1-4 surveys per Collected Chesapeake 
Institute DO Data Bay axis, 2-4 m year, 1-2 per Bay Institute Data 
(1950-1980) vertical resolution summer Reports, and Special 
with gaps. Report. See Table 2. 
Chesapeake Bay -20 stations on -20 suveys per Chesapeake Bay 
Monitoring Bay axis, l m year, bi-weekly Program web page 
Program DO, continuous during summer 
Salinity, vertical resolution 
Temperature Data 
(1984-1999) 
River Flow Inputs, One Location. Daily United States 
Conowingo Dam Geological Survey, 
or Harrisburg, PA 1 National Water 
(1900-1999) Info1mation System 
Web Site 
Monthly mean Susquehanna River flow at Conowingo, MD prior to 1968 
Was estimated from the flow at HaITisburg, PA using a regression model 
based on the overlappping records since 1968. The Harrisburg flow includes 
>90% of the flow at Conowingo. This model relating these flows had 
2
r >0.99. 
91 
Table 3-2. Data sources, cruise dates, and calculated hypoxic volumes for 
Chesapeake Bay from 1950-1999. 
9 3 
Hypoxic Volume, 10 m
D0<0.2 DO<l .O D0<2.10  Comment Year Data Dates 1 
Source mB r1 m~ r mg r
0.00 0.98 2.17 
1950 1 7/14-7/19 
7/15-8/6 1.93 3.47 
5.78 
1952 2 
0.51 1.84 2.86 2 
1957 3 7/23-7/26 
4 8/6-8/22 0.41 
4.76 7.37 
1958 
7/6-7/17 0.00 2.61 
4.85 
1959 5 
6 8/22-9/9 0.24 
3.70 4.20 3 
1960 
0.02 4.64 6.35 
1961 7 7/19-7/31 
7/24-8/5 0.00 1.41 
3.90 
1962 8 
9 7/30-8/1 5 0.03 
2.67 3.50 5 
1963 
0.00 0.08 2.34 2 
1965 10 7/1/-7/6 
0.75 3.46 4.78 4 
1968 11 7/8-7/10 
12 7/7-7/10 0.74 
1.67 2.44 
1969 
0.37 1.36 2.96 
1970 12 7/9-7/12 
3.70 5.12 7.64 3 
1972 13 8/27-8/31 
6/25-6/29 0.83 2.76 
4.93 
1973 13 2.87 4.69 
1978 14 9/18-9/20 
1.84 
2.88 
1979 14 7/9-7/12 
1.28 1.79 
2.20 3.52 5.03 
1980 14 7/23-8/2 
2.84 4.46 5.49 1 
1984 15 July Avg. 4.42 1 
1985 15 July Avg. 
0.54 2.47 
4.23 7.52 11.00 1 
1986 15 July Avg. 1.57 5.19 9.30 1 
1987 15 July Avg. 3.61 5.02 1 
1988 15 July Avg. 
2.03 
4.21 7.25 11.16 1 
1989 15 July Avg. 
July Avg. 2.02 3.45 
5.86 1 
1990 15 7.35 9.26 1 
1991 15 July Avg. 
2.42 
2.52 4.3 1 6.52 1 
1992 15 July Avg. 9.08 11.95 1 
1993 15 July Avg. 
5.87 
5.70 7.68 1 
1994 15 July Avg. 
3.30 
3.36 5.83 9.86 
1 
1995 15 July Avg. 4.54 5.84 1 
1996 15 July Avg. 
2.84 
4.37 5.71 8.15 
1 
1997 15 July Avg. 1 6.25 9.61 12.07 
1998 15 July Avg. 1.20 5.20 9.33 
1 
1999 15 July Avg. 5.25 7.40 1 
2000 15 Jul}:'. Avg. 
3.05 
92 
Table 3-2: 
Data Sources: (1) Ches. Bay Institute Data Report #4 (1950); (2) 
Chesapeake Bay Institute Data Report #18 (1954); (3) Chesapeake 
Bay Institute Data Report #33 (1962); (4) Chesapeake Bay Institute 
Data Report #38 (1962); (5) Chesapeake Bay Institute Data Report 
#41 (1962); (6) Chesapeake Bay Institute Data Report #44 (1962); 
(7) Chesapeake Bay Institute Data Report #47 (1962); (8) 
Chesapeake Bay Institute Data Report #48 (1962); (9) Chesapeake 
Bay Institute Data Report #50 (1963); (IO) Whaley et al. (1966); (11) 
Chesapeake Bay Program Database; (12) Taylor and Cronin (1974); 
(13) Chesapeake Bay Program Database; (14) Cronin et al. (1982); 
(15) Chesapeake Bay Water Quality Monitoring Program, 
Annapolis, MD. 
Comments: (1) Average of two July cruises; (2) No data south of 
Potomac River, but available data fully delineates the l mg r1 and 
lower contours; (3) Cruise dates later than likely period of maximum 
hypoxic volume; (4) DO concentration in lower pycnocline and 
upper portion of lower layer poorly described. Pycnocline depth 
considered when interpolating DO profile; (5) No data south of 
Potomac River. Data do not fu lly delineate DO contours greater than 
0.2 mg r1? Therefore, hypoxic volumes for 1.0 mg r1 and 2.0 mg r1 
are lower bounds. 
93 
Table 3-3. Estimated parameters of the non-linear multiple regression, 
9 
ln(V + 1) = /Jo + f3 (T -1949t  + j32Q +?,relating hypoxic volume (10 m3) to time 1 
3 1
(year) and January-May average Susquehanna River flow (m s- ) . Significance levels 
for /31 and a are not shown. Due to the fact that the estimates of these parameters are 
strongly and negatively correlated (r<-0.98), their confidence intervals are 
meaniningless except in a bivariate context. The significance of the trend over time 
can be evaluated via confidence bands for the predicted values (Fig. 3-5). 
DO Threshold Intercept, Bo Ti me Trend, B, Exponent River Flow-
(model r2) P(Ho: Bo=O) P(Ho: B,=0) 
a Dependence, Bz 
P(Ho: B2-0) 
2 4 
D0<0.2 mg r 1 -0.67 l.76(10Y 1.15 4.05(10Y
(r2=0.69) (J2<0.05) 3 
D0<0.7 mgr' 0.58 2.ll(lOY 1.57 2.75(1or
4 
(r2=0.67) c12~0.052 4 4 
D0<2.0 mg r 1 1.30 2.33(lOY 2.08 l.47(10Y
(r2=0.62) c12~0.os2 
10 3
Table 3-4. Regressions relating hypoxic-volume-days (units: 10 m -days) for 
Chesapeake Bay for 1985-1999 to January-May average Susquehanna River flow 
(~nits: m3 s-'). The three regressions reflect hyp~~ic volume calculated using the three 
different DO thresholds, D0<0.2, 0.7 or 1.0 mg I ? 
DO Threshold Interce t (?S.E.) 
Slo e (?S.E.) 
D0<0.2 mgr -6.57 (?4.51) 
0.017 (?0.003), r =0.76, n= l5 
p<0.01 
0.019 (?0.004), r 2= 0.59, n=l5 
DO<l.O mgr' 14.95 (?7.27) 
p<0.01 
0.015 (?0.005), r 2= 0.38, n=l5 
D0<2.0 mgr' 47.5 (?9.14) 
=0.01 
94 
Table 3-5. Estimated rates of spring DO decline in bottom waters 
at station 8450 (Fig. 4) for 1964-1977 (Taft et al. 1980) and at 
station CB4.3C for 1985-1998 (this study). Rates for 1964-1977 
1 1
were transformed via 1.0 ml r 1 d-1 = 1.4 mg r d- ? Taft et al. 
(1980) did not estimate the date of first onset of anoxia. 
Rate of DO Onset of Anoxia Station Year 
Decline (Date) 
(m r1 d-1) 
8450 1964 0.14 
1965 0.11 
1966 0.08 
1969 0.15 
1970 0.08 
1971 0.04 
1977 0.13 
CB4.3C 1985 0.103 
6/20 
1986 0.096 
6/22 
1987 0.146 
5/28 
1988 0.182 
5/26 
7/5 
1989 0.093 
1990 0.119 
5/20 
1991 0.089 
6/6 
1992 0.095 
7/3 
1993 0.184 
5/21 
0.100 6/3 1994 
1995 0.075 
7/3 
1996 0.107 
6/10 
0.085 6/28 1997 5/2 
1998 0.196 
1999 0.114 
6/14 
95 
39? 30' 
39? 00' 
38? 30' 
38? 00' 
37? 30' 
37? 00' Atlantic 
Ocean 
- - - 1- - -, 
76? 00' 75? 30' 
Fig. 3-1. A map of Chesapeake Bay showing the locations and identifications of 
Chesapeake Bay Monito,ing Program Water Quality Monito,ing stations used for 
this analysis. Note that the aspect ratio of the map has been distorted to provide 
more ho,izontal space to label stations. The boundaries of the mesohaline region, 
between 102 and 222 km from the Atlantic Ocean, are indicated. The additional 
boundaries at 139 and 185 km define the box model segments used for the DO 
input budgets. 
96 
, , 
13 
12 
11 
...--... 10 
(") 
E 9 
0) 
0 
..-- 8 
"-' 
Q) 
E 7 
:J 
0 6 
> 
(.) 5 
?x 
0 4 
a. 
>. 
I 3 
2 
1 L?-? ..L _ _  U-.-L---'J..-'--'....i.~ <-~ 1- ;t.-L- A L ? - . I --e--1? 0 ~?- i e  ~ ~ ~ ~ ~ ~ ~ 
"'"' cp"'q, n}"f, / ,:,"' 'd"' ,.,_,.:!$ ~,$ ,i~  ~ ~ ~ ~"0"' ""'"' "i'' ,/'J  
1
Fig 3-2. The time-se1ies of hypoxic volume (DO<l.0 mg 1- ) in Chesapeake 
Bay in 1993. The shaded area is the integrated hypoxic-volume-days for 
1993 , computed by summing the areas of the indicated trapezoids. 
97 
Upper Mesohaline, Middle Mesohaline, Lower Mesohaline, 
Model Seg 4 Model Seg 5 Model Seg 6 
Surface Layer 
C5 I C5 
C4 
Evic\-c4) Ev5( c\-c5) \ Ey5(C'5-C5) 
a 
12m 
Oy4C'4 ~ OvsC's I Oy5C'5 
Q' c' 
I 4 
c' 4 4 <Ii' IO 'sc's e's ~Q' c' "'\ 0 '6c'6 7 7 
c' 
\0 
00 Bottom Layer 6 
222 km 185 km 139 km 93 km 
Fig 3-3. A schematic diagram of the physical transport of dissolved oxygen 
into and out of three segments within the mesohaline Chesapeake Bay. The 
location of the segment boundaries is indicated in kilometers from the mouth 
of Chesapeake Bay. The pycnocline is located at 12 m depth. Vertical non-
advective exchange is designated Evm' where mis the segment number. Land-
ward flowing advection is designated Qm, while vertical advection (upwelling) 
is designated Qvm? All the vertical transport estimates were computed using 
the box model in Chapter 2. Dissolved oxygen concentrations are designated 
cm (surface layer) or c'm (bottom layer). 
14 
"O 5 
(l) co 0 
12 2: 0 
(l) 
Cl) 
C', 10 ..0 
E 0 
(j) 8 
0 
r-
6 0 
~~ ~oo
4 ~ -3 0 0 00 
0::: 
2 0 5 
Predicted 
0 10 ,----~~ 
Ou 
"O 
(l) 
12 0 0 ~ 5 0 0 9:,8 c5l o 
10 Cl) QJOOQ ..0 CO(b 0 
C', 
E 0 io o 8 - 0 ...__.~J--------1 
(j) 
0 
r- 6 ~~~o 
Cl) 6 0 
4 (l) -
0::: 
2 0 5 10 
Predicted 
0 
cP 
-o 10 o0 (l) 0 0 
12 2: oO OO o 
10 ~ 5 ..0 t~ 0 
C', 0 
E 8 
(j) 
0 
r- 6 &:Q~ ~_O  c;i:,  6 o ~en o 4 
2 5 10 
Predicted 
0 
1960 1970 1980 1990 2000 
1950 
Fig 3-4. Calculated summertime hypoxic volumes for Chesapeake Bay during 
1950-2000 as reported in Table 3-2. Nonlinear regression lines show predictions 
of multiple regression on time and January-May average Susquehanna river flow 
(Table 3-3). The three lines in each family of curves indicate predictions at 1100 
m3 s- (low flow), 1500 m3 s-1 (average flow, unlabelled line) or 2200 m3 s-1 1 
(high flow). The figures to the right provide a sense of the goodness of fit. 
99 
Low Flow (1000 m3 s-1) High Flow (2000 m3 5-1) 
12 r----------,----------
D0<0.2 mg 1-1 
10 
("') 8 
E 
'b 6 
4 
2 
0 ~==:~-:=:~=~=--==-=-=-==-=-=--::~~==-=-=-==-=-==-=-===== 
D0<1.0 mg 1-1 
10 
/ 
8 / ("') 
E / / 
0) 6 / 
/ 
/ / 
0 
....... / / / / 
/ / / / 
4 / / / ----
---- ----
2 ---- ---- --
--
0 
D0<2.0 mg 1-1 I 
I I 10 
I // 
/ /// 8 
("') / / / / 
E / / / / / 
0) 6 / / / 
0 / ....... / ----
4 ----
/ 
---- ------
2 
0 1960 1970 1980 1990 2000 
1950 1960 1970 1980 
1990 2000 
Fig. 3-5. Predictions and asymmetric confidence bands (? standard en-or of mean) 
for July hypoxic volume (at each of three definitions) as a function of time and 
January-May average Susquehanna River flow based on the non-linear regressions 
fitted to data in Fi g. 3-4. Two levels of river flow are indicated con-esponding to 
low flow and high flow conditions. 
100 
Fig 3-6. Summer dissolved oxygen profiles in Chesapeake Bay dming four years 
With near average January-May Susquehanna 1iver flow (approx 1,600 m3 s-1). 
Individual values indicate the locations of observations used in the contoming. 
Data for 1959 are from the Chesapeake Bay Institute Data Report #41 (1962). 
Data for 1970 are from Taylor and Cronin (1974). Data for 1980 are from Cronin 
et al. (1982). Data for 1991 were obtained from the Chesapeake Bay Water 
Quality Monitoring Program. 
101 
0 
10 
-E 20 
.-.-c--n. 30 
Q) 
0 
40 
July 1959 
50 
- -1- -i--
I mg 1-1 
0 ?:1I \J:L/-
 -l1i 
7.0 
10 
- 6.0 E 20 
-..-nc
--
. 30 5.0 
Q) 
0 
40 4.0 
July 1970 
50 3.0 
0 2.0 
(:s1 ~SA  
'[) 1.0 
0.7 
.....c..  
Q_ 3) 
Q) 0.2 
0 
4) 
July1980 
0 
10 
-E 20 
.-.-c-- 
n.30 
Q) 
0 
40 July 1991 
50 --- i-----,-----, ----i-----,-----
237 187 137 87 37 
287 Channel Distance from Atlantic Ocean (km) 
102 
120 -, ,-
-- v 00<2.0 (f) >-, 
cu 100 ? 00<1.0 
-0 0 00<0.2 
(') 
E v 1999 
-0. -
0 80 -. -
'--' 
(f) v v ? >-, 
cu 
0 60 v) 
I v ~ ? 
Q) 
E ? 
::, ? ? 
0 40 ?? ? 0 
> 
(.) 
?x ? 
0 
Q. 20 
>-, 0 
I 
0 
0 _J - _J__ .L___ __ , ___L _ 
800 1000 1200 1400 1600 1800 2000 2200 2400 2600 
January-May Average Susquehanna River Flow (m3 5-1) 
Fig. 3-7. The relationship between January-May average Susquehanna River 
flow and time-integrated hypoxia (annual hypoxic-volume-days), where hypoxia 
is defined as either dissolved oxygen less than 0.2, 1.0 or 2.0 mg 1-1. The peiiod 
encompassed is 1985-1999 the period covered by the Chesapeake Bay Water 
Quality Monitoring Program, from which the data were computed. 
103 
Bottom Water Temperature(? C) 
5 10 15 20 25 
l 
7/10 
0 .<._ 
0 ? 
~ 
?cxu  6/26 Q 
0 
C 
-<( ..0... . 6/12 
Q) 0 
Cf) 
C 
-0 0 5/29 
..Q...).  
cu 
0 0 
5/15 
I o 3.1? 
5/1 0 6.80 
I -- ,- - I r ---, 
' --r 
800 1000 1200 1400 1600 1800 2000 2200 2400 2600 
January-May Average River Flow (m3 s-1) 
Fig. 3-8. The relationship between January-May average Susquehanna River flow 
March bottom water temperature, and the date of onset of anoxia in bottom water ' 
at station CB4.3C (Fig. 3-2), located in the mesohaline Chesapeake Bay, as deter-
mined via multiple regression. Bubble size indicates the March bottom water 
temperature at station CB4.3C. Right sloping lines indicate the family of date (D) 
vs. flow (Q) relationships associated with different March bottom water tempera-
tures (T) as expresed by eq. 6. The upward sloping line (D vs. T) indicates the 
seasonal increase (corresponding to dates on y-axis) in bottom water temperature 
(top axi s) at CB4.3C (r2=0.96). The water temperature at the time of onset of 
hypoxi a can be determined by chasing flow and temperature and determining the 
date f onset of hyp xia from the family of lines, then referring to the date vs. 
temperature line. All the water quality data was obtained from the Chesapeake 
Bay Water Quali ty Monitoring Program. River flow averages were computed 
fr m daily di scharge values reported by the United States Geologica l Survey. 
104 
1.0 ,-------------------- ---~ 
Vertical Diffusive Input 
..-
-' 0 
N 0.8 
' 
E 
N 
0 
Eo.6c 
0.4 -r-------,r--------.-----i-------,r-------1 
3.0 ,----------- ---- ---------, 
2.5 Net Advective Input 
..-
-'o  2.0 
N 
' 
E 
N 1.5 
<;,10 ( 
E . 
0.5 
0.0 --i--------,------i--------,-- - - - .--- - ~ 
3.5 ,-------------- ----------
3.0 Total DO Input 
..-
i'  2.5 
'E  
('\J2.0 
0 
E1 .5 
1.0 
0.5 -r-- ---.-------.-----.-- ----,--- ----l 
1950 1960 1970 1980 1990 2000 
Year 
Fig. 3-9. Average vertical diffusive, net horizontal advective, and total dissolved 
oxygen (DO) inputs to the mid Bay region of the Chesapeake Bay during 1950-1999. 
Physical transport rates and their dependence on spring average river flow was com-
puted using a box model (Chapter 2). Advective and diffusive DO transport rates 
were computed as the product of historical average DO concentrations (see Table 3-2 
for data sources) and corresponding transport rates (see. Fig. 3-3). The size of the 
plot symbol indicates the magnitude of January to May average Susquehanna River 
discharge, whereas large symbols indicate high flow. Dark symbols correspond to 
the pe1iod covered by the more detailed Cheapeake Bay Water Quality Monitoring 
Program data. 
105 
Chapter 4: 
PHYTOPLANKTON DEPOSITION TO CHESAPEAKE BAY SEDIMENTS 
DURING WINTER-SPRING 
Abstract 
The often rapid deposition of phytoplankton to sediments at the conclusion of 
the spring phytoplankton bloom is an important component of benthic-pelagic 
coupling in temperate and high latitude estuaries and other aquatic systems. However, 
quantifying the flux is difficult, particularly in large and spatially heterogeneous 
environments. Surficial sediment chlorophyll-a (chi-a), which can be measured 
quickly at many locations, has been used effectively by previous studies as a 
biomarker indicating deposition of phytoplankton to estuarine sediments. In this 
study, surficial sediment chi-a was mapped in late spring at 20-50 locations 
throughout Chesapeake Bay during 8 years (1993-2000). A model was developed to 
estimate chi-a and carbon deposition using these measurements, while coITecting for 
chi-a degradation during the time between deposition and sampling. 
Bay-wide, the springtime accumulation of chi-a on sediments by late spring 
averaged 171 mg m-2, from which the chi-a and carbon sinking fluxes , respectively, 
2
Were estimated to be 353 mg m-2 and 26.5 g C m- . These deposition estimates were 
-50% of estimates based on a sediment trap study in the mid-Bay. Dming 1993-2000, 
the highest average chi-a flux was in the mid-Bay (248 mg m-2), while the lowest was 
in the lower-Bay (191 mg m-2). Winter-spring average river flow was positively 
COtTelated with increased phytoplankton biomass in the lower Bay water column, 
increased chi -a deposition to sediments and down-Bay translation of chi-a deposition . 
106 
In the 8 years of observation , estimates in several years di verged strongly from the 
overall pattern. A comparison of the C flux associated with the deposition of the 
spring bloom with annual benthic carbon budgets indicated that the spring bloom did 
not contribute a disproportionately large fraction of annual C inputs to sediments. 
Regional patterns in chi -a deposition did not correspond with the strong regional 
patterns that have been found for net plankton metabolism during sp1ing. 
Introduction 
The spring increase in phytoplankton production and biomass is a well-known 
feature of the phytoplankton dynamics of Chesapeake Bay and other temperate and 
high latitude aquatic ecosystems. In Chesapeake Bay, the spring increase in 
phytoplankton biomass typically begins in March . Subsequently biomass begins to 
decline some time in April and the bloom concludes by the end of May (Fig. 4-1). 
Decline of spring phytoplankton blooms in Chesapeake Bay has been attributed to 
phosphorus and dissolved silica limitation (Conley and Malone 1992, Malone et al. 
1996). Nutrient limitation promotes a physiological response leading to increased 
sedimentation of diatoms (Conley and Malone 1992), which are a major component of 
the winter-spring phytoplankton assemblage in Chesapeake Bay (Marshall and Nesius 
1996). Excluding picoplankton, diatoms in winter-spring accounted for -80% of 
phytoplankton cells in the lower Bay, 67% of cells in the mid-Bay (above pycnocline), 
and 56% of cells in the upper Bay (Chesapeake Bay Monitoring Program, unpublished 
data). Diatoms accounted for similar proportions of phytoplankton carbon (R. 
Lacouture, personal communication, Table 4-1). Physiological responses to bloom 
107 
senescence, such as formation of large aggregates, also enhance sedimentation and are 
an important aspect of the life cycle of di atoms (Smetacek 1985). Consequently, 
sinking is a quantitatively important fate of diatom blooms. In a mesocosm 
experiment simulating a spring bloom in Na1i-agansett Bay, Keller and Riebesell 
(1989) estimated that sedimentation accounted for 14-65% of gross production. 
Rapid sedimentation of intact phytoplankton to sediments is an important 
pul sed input of organic matter to benthic communities in many marine systems (e.g. 
St. Lawrence River, Townsend and Cammen 1988; Chesapeake Bay, Malone 1992). 
The importance arises not only from the quantity of the input, but from the high 
nutritional quality of the input, which has been shown to stimulate rapid increases in 
macrobenthic production (Graf et al. 1982, Marsh and Tenore 1990), microbial 
processes , and nutrient regeneration (Jensen et al. 1990). Townsend and Cammen 
(1988) suggested that the large spring flux of organic matter to sediments could play a 
role in recruitment success of juvenile demersal fishes. Spring bloom phytoplankton 
deposition has been identified as a key annual event in Chesapeake Bay, linking 
ecosystem processes in the winter-spring period to subsequent summer conditions, 
including hypoxia and summer phytoplankton blooms (Malone 1992, Chapter 6). 
Because of the potential importance of spring bloom deposition to ecosystem 
processes, quantifying the flux is of particular interest. Unfortunately, this is 
technically challenging, a fact reflected in the paucity of flux estimates. Sediment 
traps have been used to quantify vertical fluxes of particles in various aquatic systems 
(e.g., Smetacek et. al. 1978), including in Chesapeake Bay (Boynton et al. 1993). 
Although effective and useful, there are significant complications associated with the 
108 
design and use of sediment traps (Blomqvist and Hakanson 1981, Knauer et al. 1984, 
Butman 1986, Butman et a l. 1986, Asper 1987). Among other problems, the effort 
and expense required to deploy and maintain sediment traps severely limits the 
number of traps that can be deployed. In spatially heterogeneous environments such 
as estuaries, this means that the small number of sediment traps likely to be employed 
may not adeq uately characterize the vertical particle flux. For example, if 
phytoplankton production is locali zed outside the vicinity of the trap, the measurement 
will underesti mate the flux. Alternatively, an overestimate could result from 
phytoplankton bei ng locali zed in the area sun-ounding the sediment trap. Therefore, 
an approach that can estimate the flux at many locations is preferable. 
Previous studies have demonstrated that chlorophyll-a (chi-a) and other 
phytoplankton pigments are effective indicators of fresh phytoplankton inputs to 
sediments (Sun et a l. 1991 , Josefson and Conley 1997). Where sediments are not 
euphotic, production of chi-a on sediments will be small compared to deposition of 
phytoplankton. This study used sediment chi-a measured in late spring as an indicator 
of deposition to sediments of phytoplankton originating from the spring bloom. Since 
spring bloom sedimentation appears to occur rapidly and in relatively cold water, 
degradation rates are likely to be small relative to sedimentation rates. This suggests 
that chi-a accumulation in late spring, with an appropriate cotTection for degradation , 
could be used to estimate recent deposition of phytoplankton to sediments. To obtain 
adequate spatial resolution, sediment chi-a was mapped on a regular grid throughout 
the estuary. Recognizing the significant interannual variability, particularly in 
assoc iation with freshwater input rates (Boynton and Kemp 2000), sediment chi-a was 
109 
mapped annuall y for 8 consecuti ve years . Interpretation was supported by comparison 
with contemporaneous estimates of phytopl ankton biomass in the water col umn , 
sedimentation estimates from a sediment trap study, estimated phytopl ankton sinking 
rates, and by compa1ison with net plankton community production estimates. 
Study Site 
Chesapeake Bay is a large, parti ally stratifi ed estuary that extends 300 km from 
the mouth of the Susquehanna River in Maryl and to the Atlantic Ocean between Cape 
Henry and Cape Charles, VA. The oligohaline upper Bay has a mean depth of 4.5 m 
with a deeper (- 10 m) channel near the eastern margin. The mesohaline mid-Bay has 
a deep central channel, 20-50 m, flanked by shallower shoal areas to the east and west, 
giving it a deeper mean depth of 10.3 m. The polyhaline lower Bay is broader with a 
wide central channel region averaging -15 m depth as well as broad shoal areas on the 
flanks of the channel. The mean depth is 9.2 m. 
The physical transport regime throughout most of the estuary is best 
characterized by 2-layer gravitational circulation in which net up-estuary advection 
occurs below the pycnocline and net down-estuary advection occurs in the surface 
layer (Pritchard 1952). In the upper Bay, the circulation is initially down-estuary at all 
depths and at some point down-estuary makes a transition to the two-layer circulation. 
Sediment-types vary throughout the estuary. North of Patuxent River and in 
the western half of the Bay south of Patuxent River (Fig. 4-2), sediments are >80% 
silt-clay except in shallow waters. In these shallow waters, and in deeper areas of the 
110 
eastern half of the south Bay, more porous sandy sediments (>80% sand) predominate 
(Kerhin et al. 1983, Chesapeake Bay Benthic Monitoring Program, unpubli shed data). 
Methods 
.Eie ld Methods 
Sediment cores were obtained throughout the Bay during mid to late April in 
each year during 1993-2000 (Table 4-2, Fig. 4-3). Sampling crui ses were conducted 
aboard the RIV Cape Henlopen and were part of a multi-disciplinary research project 
(Chesapeake Bay Land Margin Ecosystem Research Program). 
Cores with an undisturbed sediment-water interface were obtained using a 0.25 
2 
m Smith-Macintyre coring device at 20-50 locations usually located along hoiizontal 
transects spaced - 20 km apart. In 1993-1995, when the highest numbers of stations 
were sampled, additional stations were occupied between transects. Cores were 
obtained in waters from the deepest portions of the Bay to as shallow as 8 m. 
Shallower depths were not sampled due to draft limitations of the research vessel. 
Once onboard, a sub-core was obtained using a modified 60 cc plastic syringe. 
This provided a sample of precise cross-sectional area and l cm depth, which was 
frozen immediately in a plastic centiifuge tube. In 1993, the top 2 mm from 2 sub-
cores was combined in a single centrifuge tube, rather than 1 cm from a single sub-
core. In 1994-1995, two samples were obtained at each station. One sample included 
the top 1 cm from a single sub-core, while the other included the top 2 mm from 2 
sub-cores, as in 1993. This provided a means for comparing the two types of samples. 
111 
The reasons for these changes in fi eld methods were unrelated to thi s study, but 
provided a limited means to examine the vertical distribution of chi-a in Chesapeake 
Bay sediments. 
Pigment Analysis 
Frozen sediment samples were briefly thawed at room temperature, then 40 ml 
of 90% acetone was added. Samples were extracted for 12 hours in a dark 
refri gerator, shaking 2-3 times during the course of the extraction, then centrifuged at 
-1760 rpm for 5 minutes before decanting into a cuvette. Total chi-a , active chi-a and 
phaopi gment concentrations in the acetone extracts were determined fluorometiically 
using the acidification method described in Strickland and Parsons (1972) and Parsons 
et al. (1984). Only the total chi-a and phaopigment data were examined in this study. 
The laboratory utilized a Turner Designs Model TD700 fluorometer calibrated against a 
spectrophotometer using pure chlorophyll-a from spinach (Sigma Chemical Company, 
C 5753), or liquid standards from Turner Designs, #10-850. 
The extraction method that was used was later found to be different from that 
used by some published studies (e.g. Sun et al. 1991). Specifically, sediments were 
not sonicated prior to extraction, and on ly a single extraction was used. Therefore, a 
method comparison study was undertaken to determine whether the results would have 
differed significantly by use of sonication and/or an additional extraction. In this 
experiment surficial sediments were obtained from box cores col lected on Patuxent 
' 
River, then processed in the field as described above. In the laboratory, the samples 
were thawed, then homogenized. Fifteen equal size aliquots from the continuously 
112 
stirred mud-sluJTy were extracted as described above after one of three sonication 
treatments. The treatments were: (1) no sonication (control) ; (2) microsonication for 3 
minutes; and (3) sonication in a sonicator bath. The extracted pigments were decanted 
and analyzed as above. A second extraction of each sample was also analyzed as 
above, with the sum of the first and second extractions being recorded as the value for 
double-ex traction. No sonication was performed prior to the second extractions. 
Although this design resulted in 30 values desc1ibing each of 6 treatment 
combinations, there were only 15 independent observations. Therefore, statistical 
signifi cance was evaluated using repeated measures ANOVA. A comparison of single 
vs. double extraction (without any sonication) was also done on 7 non-homogenized 
samples from different locations in Patuxent River. 
1.Dterpolation Methods 
Sediment chlorophyll-a data were interpolated to a regular gzid for the purpose 
of contouring and computing regional means using the kriging procedure of Surfer 
software (Golden Software, Inc., Golden, CO). A quadrant search algorithm was 
selected such that a maximum of four observations were selected from each of 4 
quadrants divided by north -south and east-west oriented axes . 
.Water Column Chlorophyll -a 
Water column chlorophyll-a data were obtained from the Chesapeake Bay 
Water Quality Monitoring Program. Chlorophyll-a concentration was determined 
spectrophotometrically from acetone extractions of ground filters (EPA 1993). 
113 
Seasonal and regional integrated chlorophyll-a distributions as we!J as regional mean 
integrated chlorophyll-a concentrations were computed from interpolated distributions 
based on data from a series of approximately 20 stations located down the axis of the 
estuary (Fig. 4-1). Integrated chlorophyll-a was computed from vertical profiles of 
chlorophyll -a weighted by cross-sectional volumes per meter depth (Cronin and 
Pritchard 1975). Integrated chlorophyll-a was extrapolated in time and horizontal 
space using linear interpolation with quandrant searching on north-south and time 
axes. 
Results and Discussion 
Pigment Analysis Method Comparison. The results of a method comparison 
experi ment showed that sonication and multiple extraction of sediment samples (e.g., 
Sun et al. 1991) could be expected to give sediment chi-a measurements 15.6% higher 
that those obtained using the method used in this study (Table 4-3). The difference 
was found to be a nearly constant proportion of chl-a as measured using a single 
extraction and without sonication, allowing a correction to be applied. Compared to 
the control (no sonication, single extraction), 3% more chi-a was extracted after use of 
a sonicator bath and 4.6% more chi-a was extracted after microsonication (p<0.01, 
Table 4-3). The second extraction removed 9.9-11.3% additional chl-a (p<0.01), 
depending on the sonication treatment (p<0.01). A larger amount was extracted on the 
second extraction if microsonication was used prior to the first extraction. Sediment 
chi-a measured in 7 non-homogenized sediment samples from Patuxent River using a 
single extraction and no sonication varied between 77 and 148 mg chi-a m-2. A 
114 
second extraction obtained l 1.0?0.5% (mean?std error) additional chl-a, a proportion 
comparable to that obtained for the coITesponding treatments using homogenized 
samples (Table 4-3). This indicated that a proportional (15.6%) correction could be 
applied to the 1993-2000 Chesapeake Bay samples, which were previously processed 
With a single extraction and without sonication, with a high degree of confidence. 
Although this coITection is not large compared to other possible sources of 
uncertainty, it was applied in the interest of beginning the analysis with the most 
accurate measurements that could be obtained. 
Computing Sediment Total Chlorophyll-a Inventories. Due to vertical mixing 
of sediments on short time scales (days to weeks), the total inventory (i.e., vertically 
integrated concentration) of recently deposited chl-a may not have been accurately 
represented by the inventory within the top 0-10 mm of sediments. This leads to an 
Underestimate of chi-a deposition, and, to the extent that sediment mixing could differ 
regionally, could affect comparisons among regions. The simultaneous collection of 
0-2 mm and 0-10 mm sediment samples dwing 1994-1995 provided an opportunity to 
examine this issue and compute the sediment chi-a inventory. Bay-wide, the ratio of 
the 0-10 mm to 0-2 mm chl-a inventories was estimated to be 2.75. As the ratio is 
substantially less than 5, this indicates a decline in chi-a concentration with depth in the 
sediments. Because of poor compliance with the assumptions of ANOVA, a non-
parametric ANOVA was used to test for differences in the mean ratio in the three 
regions. The ratio was found to differ significantly among regions of the Bay 
(Kruskal-Walli s Test; p<0.05). The respective median ratios for the upper-, mid- and 
lower-Bay regions were 2.7, 2.3, and 3.1. Using more detailed vertical profiling of 
115 
sediment chi -a in the top 10 cm of Long Island Sound sediments, Sun et al. (1994) 
observed an exponential decrease in chl-a with depth be low the sediment-water 
interface. Thi s model has been assumed to apply to Chesapeake Bay as well. 
According ly, the chi -a in ventory (C111 ) integrated to a depth h, is 
z=h 
c _ f e - k: -_ C _k_!_  (l _e  - kh ) int - c rnax max (1) 
:=0 
where C111ax is the maximum (surface) concentration and k is the rate of decrease with 
depth. Using thi s expression , the ratio, R, between the 0-10 mm concentration and the 
l - IOk 
0-2 mm concentration is R = -e , , which gives k=0.19, 0.25, and 0.14 for the 
l -e - 2k 
upper-, mid- and lower-Bay. Using these estimates of k, the top 10 mm was estimated 
to include (in same order) 85%, 92% and 76% of the total chi-a inventory(:::::: 0-10 cm 
integrated chi-a). These factors were used to compute the chi-a inventory from the 
measured concentrations. 
The regional differences in vertical chl-a distribution (i.e., ink) may reflect 
differences in sediment properties and/or mixing processes. The mid-Bay is 
characterized by fine, silty sediments (Fig. 4-2), deep and seasonally anoxic water, and 
lower physical energy (i.e. waves and currents) than other areas of the Bay. This 
would be expected to minimize physical and biological mixing of sediments. In 
contrast, the lower Bay is shallower and has an increased prevalence of sandy 
7
sediments. During winter-spring, the penetration depth of Be (half-life= 53 d) in the 
lower Bay was 3-5 cm, with physical mixing due to tidal current and wave action 
being dominant (Dellapenna et al. 1998). This may explain the deeper mixing of 
116 
deposited ch i-a in the lower Bay, although no comparable data are available for the 
mid Bay or upper Bay. An Apri l minimum in sediment mixing in the lower Bay, prior 
to a summer increase associated with bioturbation (De ll apenna et al. 1998), suggests 
that macrobenthi c activity was suppressed by low water temperature (5-15 ?C) up until 
the time that surficial sediment samples were collected. Thi s is important to the 
overall approach of thi s study because losses of chi-a due to macrobenthic activity 
could be more variable and therefore difficult to quantify. 
Phaopigments. Phaopigments are a product of the earl y breakdown of chi-a. 
As such, they are also an indicator of deposition of phytoplankton to sediments. 
Macrobenthic biomass has been found to be positively associated with phaopigments 
CJosefson and Conley 1997); however, this may be due to the presence of 
phaopigments in feces of plankton and macrobenthos and the slow degradation rate of 
phaopi gments as compared to chi-a (Furlong and Carpenter 1988). With this in mind, 
a high ratio of chi-a to chl-a+phaopigments suggests that substantial deposition of chl-
a occun-ed recently. For the 84 measurements Bay-wide in 1996-2000 in which 
sediment phaopigment concentration was measured, chi-a accounted for an average of 
47% (range=38-58%) of chl-a+phaopigments. This is regarded as a high ratio 
(Josefson and Conley 1997), suggesting that a significant amount of phytoplankton 
was deposited to Chesapeake Bay sediments each sp1ing. 
Distributions of Sediment Total Chlorophyll-a. The computed sediment total 
chi-a inventory averaged 175 mg m-2 over 272 observations Bay-wide during 1993-
2000. The median value was 164 mg m-2 (interquartile range= 88-234 mg m-2). 
117 
Regional and overa ll mean chi-a was calculated for each year from interpolated 
di stribu tions to account fo r the non-random dist,ibution of observations. Computed in 
thi s way, the long-term overall mean was 171 mg m-2, very close to the unweighted 
average of all observations. However, regional means, particularly the upper Bay 
mean, were s li ghtl y more sensiti ve to the averaging procedure. Therefore, the 
interpolated fi elds were used to compute means rather than the raw data. The highest 
average chi-a in ventory, 195 mg m-2, was found in the mid-Bay. Lower chi-a was 
fo und in the lower Bay (148 mg m-2) and the upper Bay (172 mg m-2, Table 4-4). 
Interannuall y, the highest Bay-wide mean chi-a inventory was 244 mg m-2 in 1999, 
While the lowest was 117 mg m-2 in 1995, a >2-fold range (Table 4-4). The 
di stribution of raw observations illustrates the patterns and magnitude of spatial and 
interannual vari ability in sediment chi-a (Fig. 4-3). 
Both the regional distribution of chi-a within the Bay and the Bay-wide 
average sediment chi-a concentration were related to the magnitude of winter-spring 
(Jan-Apr) total di scharge of freshwater into Chesapeake Bay, which over the past 15-
years was highly coITelated (r2=0.91) with the Jan-Apr discharge from the 
Susquehanna River, the largest tributary of the Bay (Fig. 4-4, Chapter 2). Bay-wide, 
the average sediment chi-a increased with spring river flow, a pattern also observed 
for the lower Bay and less obviously for the mid-Bay. This positive association likely 
reflects a nutrient enrichment mechanism operating at a seasonal/whole-estuary scale. 
Since nutrient loading is positively and strongly correlated with river flow (Boynton 
and Kemp 2000), increased river flow can be expected to increase phytoplankton 
biomass and production when nutrient limitation is important. Since nut,ient 
118 
limitati on, principally by phosphorus and silicate (for diatoms) , is well known near the 
end o f spring in the mid- and lower-Bay (Conley and Malone 1992; Fisher et al. 1992; 
Fi sher et a l. 1999), it is not surpri sing that a strong positi ve correlation between river 
flow and spring average water column chi-a was observed for the lower Bay (Fi g. 4-
5). Thi s increased ph ytoplankton biomass apparently contributed to increased chi-a 
export to sediments (Fi g. 4-6). Physical transport processes associated with hi gh fl ow 
may enhance nutrient enrichment in the lower Bay by decreasing water residence 
times in the upper Bay (Hagy et al. 2000) , thereby increasing down-Bay nutrient 
transport. Whether by a river flow-dependent mechanism (thi s study) or in association 
with a flow-independent long-term nutrient loading rate increase (Harding and Perry 
1997), the effects of increased nutrient loading on primary production appear most 
dramatic in the lower Bay. 
In contrast to the mid and lower Bay, sediment chi-a in the upper Bay tended 
to decrease with increasing 1iver flow (Fig. 4-4) . One possible explanation is that high 
river flow decreased water residence time and increased turbidity. This would be 
expected to translate phytoplankton biomass and primary production down-estuary, 
precluding deposition to sediments in the upper Bay region. If this were a sufficient 
explanation for decreased sediment chi-a in high flow years , then one would expect to 
make two observations: (1) water column chi-a and river flow would be strongly and 
negatively corre lated and (2) water co lumn chi-a would be positively coITelated with 
sediment chi-a. The first expectation did not hold . Although a broad negative 
association was present between water column chi-a and 1iver flow , the correlation 
was not strong. In fact, the negative coJTelation between river flow and sediment chi-a 
119 .. 
(Fi g. 4-4) was stronger. The second expectati on also did not hold. Instead, in 6 of 8 
years, sediment chi -a was lower when water column chi -a in the upper Bay was higher 
(Fi g. 4-6). A speculative expl anation fo r thi s general observation is that high 
phytoplankton deposition to sediments may cause low phytoplankton biomass in the 
water column due to low primary production and negative net plankton production 
(Smith and Kemp 1995). Conversely, in the presence of low rates of primary 
production , hi gh phytoplankton biomass can generally only occur in the absence of 
high deposition rates. Of course, as 01iginally hypothesized, simultaneously low 
phytoplankton deposition to sediments and low phytoplankton biomass in the water 
column could result from very high flushin g rates. This "wash-out" could have 
occurred during the record floods of spring 1993, explaining the low sediment chi -a 
that year (Fi g. 4-6, upper panel). On the other hand, sediment chi-a in the upper Bay 
in 1999 was consistent with expectation based on low river flow in that year (Fig. 4-4), 
but water column chi-a remained high. This made 1999 an outlier in the water column 
vs. sediment chi-a relationship (Fig. 4-6) . The explanation for this departure from the 
general pattern observed in other years is not known . 
The species composition of the winter-spring phytoplankton assemblage in 
1993-2000 was examined in an effort to explain more of the variability that was 
observed in sediment chi-a those years (Chesapeake Bay Phytoplankton Monitoring 
Program, unpublished data). It was hypothesized that higher sedimentation in some 
years was due in part to a greater relative abundance of diatoms, whose tendency 
toward sinking has been noted (Smetacek 1985). Some suggestive results were 
obtained. In the lower Bay, diatom counts largely paralleled average chi-a due to the 
120 
dominance of diatoms in the winter-spring phytoplankton community. However, 
diatom counts did not predict sediment chi-a as well as water column chi -a did , 
probably due to larger random variability in diatom counts due to less frequent 
sampling. Sediment chi-a in the upper Bay appeared to increase with average diatom 
counts, apparently contradicting the results based on chi-a. However, there were two 
substantial outliers and a weak relationship among the remaining observations, 
suggesting that the appearance of any relationship was due to chance. Thus, the 
analysis of phytoplankton species data neither contradicted nor supported the 
hypothesis, in significant part because the temporal resolution of these labor-intensive 
data collections was too low to adequately characterize the highly variable 
phytoplankton community during the winter-spring period. 
As the above discussion illustrates, a vaiiety of ecosystem processes can affect 
relationships among river flow, water column chi-a and sediment chi-a. These clearly 
have the potential to cause dramatic departures from relationships expressed by 
empirical models. However, the predictable ecosystem responses that were observed 
among many observations indicates that, on a region-specific basis, certain 
mechanisms appear to maintain first-order importance and drive large (2-3 fold) 
responses. In some cases, outliers illustrate that an implicit assumption of the model 
Was not met. For example, the conceptual basis for Figs. 4-6 implicitly assumes a Jan-
Apr time domain for forcing and response. The 1997 outlier, in which water column 
chi-a in the mid- and lower-Bay was much higher than expected, may reflect the 
unseasonably high flow that occutTed dwing fall 1996. This river flow substantially 
increased January nutrient concentrations (N, P, Si) in surface waters at a station in the 
121 
lower Bay from the ir long-term (1984-1999) averages. Total N increased from 27 to 
41 ?M, total P from 0.88 to 1.16 ?Mand di ssolved Si from 5.5 to 11.6 ?M 
(Chesapeake Bay Monitoring Program, unpubli shed data). These high January 
nutrient concentrations reduced the importance of a normal spring freshet for 
supplying the 1997 spring phytoplankton community with nutrients. 
Another possible source of uncertainty in the observed relationships (Fig 4, 
Fig. 4-6) is the variable timing of the ecosystem response, specifically the dates of 
maximum phytoplankton biomass accumulation and bloom collapse relative to the 
dates of the sedi ment chi-a surveys (Fig. 4-7). For example, peak water column chi-a 
in 1996 occurred on 5/14/96 in both the mid- and lower Bay, one week after sediment 
sampling was conc luded. This may explain why Bay-wide average sediment chi-a in 
1996 was lower than expected from the level of spring river flow. In contrast, peak 
water column biomass in the mid and lower Bay in 1997 occurred on 4/3, -2 weeks 
prior to sediment sampling (Fig. 4-7). This probably contributed to the high sediment 
chi -a observed in that year. Peak phytoplankton biomass also occurred very close to 
the sediment sampling dates in 1998 and 2000. In 1999, the highest sediment chi-a 
deposition was observed despite the lack of any large accumulation in the water 
column before or after sampling. Imp01tantly, 1996, 1998 and 2000 were not 
substanti a l outliers in the analysis (Fig. 4-4, Fig. 4-6) as was 1997 in Fig. 4-5. This 
suggests that deposition of chi-a to sediments occurred at least in part in a relatively 
steady-state process whereas some fraction of production was continuously deposited 
to sediments. Deposition could not be explained simply by a pattern of biomass 
accumulation in the water column followed by mass deposition to sediments. 
122 
futi mates of Chlorophyll -a Deposition. A simple diagenetic model was used to 
estimate the deposition of phytoplankton to sediments during sp1ing in each year using 
the observed accu mul ation of sediment ch i-a. A few simpli fying assumptions were 
needed due to data limitations. It was assumed that the input to sediments occurred at 
a constant rate,/ (mg m-2 d-1) over a period oft days, during which time depos ited chi -
1
a decayed at a first-order decay rate, k (d- ). The net acc umulation rate of chi-a on the 
sediment surface can be described by dC/ dt = I - kC. Solving under the boundary 
condition that when t=O, C=Co yields 
C I ( I] -k1 
1 =;;+ C0 -;; e (2) 
Solving for I gives 
k l - kl) I = ~C1 - C0e (3) 
l -e -kl 
Although not immediately obvious, it can be shown using L' Hopital's rule that 
I1. m I = C 1 -Co  . Thus if the degradation rate 1? s very sma 11 , an d mr?n 1? mal chi-a was 
k~o t ' 
present prior to the period of interest (C0;::::0), to tal deposition equals the observed 
accumulation (C,) and is insensitive tot. In contrast, the deposition rate depends 
inversely on t. As the degradation rate (k) increases relative to the deposition rate([) , 
a steady state model as suggested by Sun et al. (1991) may be more appropriate. In 
Chesapeake Bay, the time period, t, during which most spring bloom phytoplankton 
deposition occurs probably varies from year to year (Fig. 4-7), but it was assumed that 
123 
most deposition occu1Ted between mid to late February and the time of the sediment 
surveys, a period of -60 days. Based on Fig. 4-7, a range of 30-75 days was 
considered possible. A few measurements of sediment ch i-a in Chesapeake Bay in 
early January-February were avai lable for a number of years during the 1980's (Garber 
et al. 1989). These values varied between 37-83 mg m-2 and averaged 58 mg m-2, 
providing a base case and range of variability for Co. Estimates for the first-order chi-
a decay rate (k) were obtained by considering the work of Sun et al. (1993a) and other 
studies by Sun and colleagues (Sun et al. 1991, Sun et al. 1993b, Sun et al. 1994). 
These studi es provide a good assessment of chi-a degradation under a variety of 
conditions. The rates most applicable for this study appear to be those obtained for 
Un -frozen, oxic sediments (Sun et al. 1993a), although it is possible that sediments at 1 
cm depth were completely anoxic in some places. Oxic degradation of chi-a is highly 
temperature-dependent, with the first-order decay constant for free chi-a (kd) 
increasing 4-fold between 5 ?c and 25 ?C (Sun et al. 1993a). The first-order rate for 
release of chi-a from a particle-bound state to a free state (kr), which was required for 
most chi-a degradation, also increases more than 6-fold over the same temperature 
range (Sun et al. 1993). Over 5-25 ?c, kr was 30-50 times greater than kd; therefore, 
only the smaller rate is relevant here. During the period from mid-March through May 
1, bottom water temperature increased from 4 to 15 ?c in the upper and lower Bay and 
from 4 to 13 ?c in the mid-Bay. The average in all regions during March-May was 
1 
- 7-9 ?c. In this temperature range, kd was 0.028 d-'. Thus, 0.028 d- was used as a 
base case estimate fork in eq. (1) and eq. (2), with values between 0.02 and 0.04 
considered as a reasonable range of variability. 
124 
Estimates of chi -a deposition (?standard deviation) were computed for each 
region and year using Monte-Carlo simulations (Table 4-5, Table 4-6). In these 
simulations, the parameters Co, land k were chosen randomly from triangular 
distributions specified using the estimated min , max and mode, which is equal to the 
base case estimate for each parameter (Table 4-5). For each value of C, (i.e. each 
4
region , year), many (10 ) estimates of the average daily deposition rate([) and total 
deposition (/t) were computed using eq. 3. Means and standard deviations were then 
computed (Table 4-6). The 1993-2000 average chi-a deposition rate was estimated to 
range from 5.08 mg m-2 d-1 in the lower Bay to 6.81 mg m-2 d-1 in the mid-Bay. 
Average cumulative winter-spring chi-a deposition varied from 277 mg m-2 in the 
lower Bay to 371 mg m-2 in the mid-Bay. Estimated coefficients of vmiation for chi-a 
deposition rate and cumulative deposition estimates averaged 12% and 16%, 
respectively . Chi-a deposition rate and cumulative deposition were not directly 
proportional to the late-spring chi-a inventory ( C,) because Co was not equal to zero 
(see eq. 3). However, because C, was typically much greater than Co, the ratios 1/C, 
and Jt/C, were much less variable than C,. For example, I/C, ranged from 0.032 to 
0.036 d-1? The ratio It/C1 ranged from 1.76 to 1.95. In other words, the cumulative 
winter-spring deposition of chi-a was slightly less than two times the observed 
sediment chi-a inventory near the end of April. Therefore, regional and interannual 
patterns of chi-a deposition rates were comparable to conesponding patterns in late 
spring chi-a inventories (Table 4-4, Fig. 4-4, Fig. 4-6) . 
Boynton et al. (1993) estimated chi-a deposition in the mid-Bay during 1985-
1992 using consecutive short-term (-1 week) deployments of sediment traps at one 
125 
station. In most years, ch i-a deposition measured just below the pycnocline was -5-10 
2 
mg m? d-1 in late February, then increased to 10-20 mg m?2 d-1 in April. From the 
earliest trap deployments in early February until early May, integrated chi-a 
deposition as measured by the traps was 600 to 1200 mg m?2 with an average of 789 
2
mg m? . The comparable mid-Bay estimate from this study is 371 mg m?2, or about 
50% of the sediment trap estimate. This study estimated the average chi-a deposition 
rate in the mid-Bay to be 6.81 mg m?2 d-1, 71 % of the 9.6 mg m?2 d-1 estimated using 
the sediment traps. Because of the limitations of sediment traps (Blomqvist and 
Hakanson 1981 , Knauer et al. 1984, Asper 1987) one cannot assume that sediment 
traps provided a more accurate estimate. It is possible that the sediment traps retained 
particles more effectively than the sediment surface, leading to an overestimate of the 
flux to sediments. This "resuspension" effect clearly affected the deeper sediment 
traps for which the flux measuremen ts were often several times larger than the mid-
cup fluxes (Boynton et al. 1993). 
Another check on the chi-a deposition estimates can be made by using the 
estimated ch i-a deposition rate and estimates of water column chi-a concentrations to 
estimate an effective si nking velocity for phytoplankton cells. This velocity can then 
be compared with measurements from the published literature. This approach requires 
that one assume a uniform vertical chi-a distribution in the water column, which may 
be appropriate in late winter and early spring in Chesapeake Bay. The effective 
sinking rate (v,) can be estimated from the integrated water column chi-a 
126 
concentration (C;, 11) , the mean depth ( z ) and the rate of chi -a deposition to sediments 
(F) using 
v = -F- w he re C we = -C_int- 
z c,,vc ' z 
Given C;,11= 50-100 mg m-2 (Fig. 4-2) , z::::: 8 m, and F=6.0 mg m-2 d-1, this gives 
Vz=0.5-l.O m d-1? Mean upwelling velocities in the range of 0.5 m d-1 would affect 
cells sinking through the water column (Hagy 2001). Thus, the actual sinking rate 
may be l.0- l.5 m d-1, approximately the same as the 1.1-1.5 m d-1 estimated for larger 
cells (8-53 ?m) within a whole phytoplankton assemblage in an experimental 
enclosure (Bienfang 1981). This estimate exceeds the minimum sinking rates 
estimated for Skeletonema costatum, the most abundant species in lower Chesapeake 
Bay in winter-spring (Table 4-1), but approximates the maximal sinking rates for the 
same species (Smayda 1970). Thus, the observed deposition probably represents 
sinking of senescent and/or nutrient limited cells, consistent with observations of 
Smetacek (1985). 
The analysis of average sinking rate noted above is intended to show only that 
the estimated chi-a deposition is consistent with reported sinking rates and observed 
chi-a concentrations in the water column. It is not known, however, if the deposition 
actually occurred at this average sinking velocity. Formation of large "floes" can lead 
to settling rates of 10-100 m d-1 (Smetacek 1978), sufficient to deposit an entire 
senescent phytoplankton bloom to Chesapeake Bay sediments within one day. 
127 
?.arbon Flux to Sediments. Given an estimate of C:chl-a, the estimated spring chi-a 
flux to sedi ments described above can be used to estimate the carbon flux associated 
With spring bloom phytoplankton deposition. The long-term January-April average 
C:chl-a in mid-Bay and lower-Bay bottom water is -100 (Chesapeake Bay Monitoring 
Program Data). In the upper-Bay, average C:chl-a during winter-spring was >250. 
These values are higher than those typical of phytoplankton, indicating a large non-
phytoplankton (i.e. det,itus) component within the POC. When chi-a increased 
quickly and substantially (e.g. Fig. 4-7), C:chl -a decreased to an asymptotic value of 
-50, with typical values between 50-100 when ch1-a>20 1?g r ? Following a bloom , 
C:chl-a was found to increase quickly as chi-a decreased. This suggests that the 
Phytoplankton community had C:chl-a z 50-100 ru1d that particles lost at the 
termination of the bloom, possibly due to sinking, also had a C:chl-a in the range of 
50-lOO. This was supported by sediment trap data (Boynton et al. 1993), which 
showed that the ratio of carbon to chi-a sinking flux in March-April was -75 when the 
chi-a flux was >8 mg m?2 d-1. 
Using C:chl-a=75 and an average total chi-a deposition of 277-371 mg m?2 
(Table 4-6) the carbon flux to sediments associated with spring bloom phytoplankton 
2
deposition is estimated to have been 21-28 gC m? ? Similarly converted to C, the chl-a 
1
deposition rate was equivalent to 0.51 gC m?2 d- , 71 % of the C flux computed from 
chi-a fluxes to sediment traps (also assuming C:Chl-a=75), but only 36% of the 
directly measured PC fluxes to the same sediment traps (Fig 8, Boynton et al. 1993). 
The large r d., span? t yo bs erve d between directly measured PC fluxes and estimates from 
128 
this study reflects periods in which the sediment traps received particles with C:chl-a 
hi gher than 75, potentially due to resuspension effects on the sediment traps. 
These carbon flux estimates for the spring bloom in Chesapeake Bay are 
substanti a ll y higher than reported carbon fluxes associated with spring phytoplankton 
blooms in some other systems. For example, a 34-day bloom in the Baltic Sea 
deposited 6.2 g C m-2 to sedi ments (Smetacek et al. 1978, cited in Keller and Riebesell 
1989). A 25-day bloom in the Kiel Bight deposited 11.5 g C m-2 (Peinert et a l. 1982, 
c ited in Keller and Riebesell 1989). The estimated C flux rate for Chesapeake Bay is 
s imilar to that of the Kiel Bight bloom, but persi sted for a longer period of time, 
leading to a larger cumulative C flux (Table 4-6). This seems reasonable given the 
eutrophi c condition of Chesapeake Bay. 
The estimated carbon flux associated with the spring bloom (21-28 gC m-2) 
sediments accounts for 10-14% of annua l benthic respiration (163 g C m-2 i') plus 
carbon burial (39 g c m-2 / , Kemp et al. 1997), s li ghtly less than proportional to the 
fraction of the year encompassed (60/365 days= 16% ). That the spring bloom 
deposition did not support a larger fraction of annual metabolic C demand was 
surpri sing considering the c lear seasonality of phytoplankton biomass (Fig. 4-1) and 
net plankton metabolism (Kemp et al. 1997), and the importance generally ascribed to 
this annual ecosystem event. Assuming that the spring bloom deposition was not 
larger than estimated, but that it was important to the macrobenthic community as has 
been suggested, one may conclude that the importance arises from food quality rather 
than quantity (e.g. Marsh and Tenore 1990). 
129 
Another surpri sing result is that the sp1ing phytopl ankton deposition differed 
only s lightly by Bay region and that the regional va,i ati on did not parallel regional 
differences in net plankton metabolism (NPM, Smith and Kemp 1995), which 
increased strongly down-estuary. For the mid-Bay, the estimated carbon flu x (0.51 g 
2 1
C m- d- ) is slightly greater than NPM (=0.41 gC m-2 d-1) estimated by Smith and 
Kemp (1995 ; converted from 0 2 flux using g C=0.375 g 0 2). In contrast, the 
estimated winter-spring carbon deposition to sediments in the lower Bay (0.38 gC m-2 
1
d- ) was only 24% of the much higher estimate of NPM for the lower Bay (l.6 g c m-2 
1
d- , Smith and Kemp 1995). The fate of the apparent surplus production in the lower 
Bay is unknown , but may include export to the mid-Bay via the landward advection in 
the lower water column, or possibly export to the coastal ocean . The presence of 
significant chi-a fluxes to sediments in the upper Bay, despite negative NPM may 
indicate that allochthonous C inputs supported plankton respiration and reduced NPM 
' 
While autochthonous phytoplankton production supported vertical C fluxes to 
sediments. 
Conclusions 
Suificial sediment chlorophyll-a can be used effectively as a biomarker for 
spring bloom phytoplankton deposition to sediments. These deposition estimates 
obtained are the only known Bay-wide estimates for Chesapeake Bay. Deposition was 
2-4 times greater than estimated spring bloom deposition from some other estuarine 
and coastal systems, illustrating the intense piimary production associated with spiing 
phytoplankton blooms in Chesapeake Bay. Deposition increased with river flow and 
130 
was trans lated down-Bay in high flow years, suggesting the importance of both 
nutrient enrichment and physical transport processes in determining phytoplankton 
deposition to sediments during spring. The estimated deposition, although large, did 
not account for a larger than proportional fraction of annual benthic metabolic 
requirements. A lack of regional coJTespondence between net plankton production 
and deposition to sediments leaves important questions about this important benthic-
pelagic coupling mechanism. 
Literature Cited 
Asper, V. L. 1987. A review of sediment trap technique. Mar. Tech. Soc. J. 21(2): 
18-25. 
Bienfang, P. K. 1981. Sinking rates of heterogeneous, temperate phytoplankton 
populations. J. Plank. Res. . 3(2): 235-253. 
Blomqvist, S . and L. Hakanson. 1981. A review on sediment traps in aquatic 
environments. Arch. Fur Hydrob. 91(1): 101-132. 
Boynton, W.R. , W .M. Kemp, J.M. Barnes, L.L. Matteson, F.M. Rohland, D .A. 
Jasinski and H.L. Kimble. 1993. Ecosystem Processes Component Level l 
Interpretive Report No. 10. Chesapeake Biological Laboratory (CBL), 
University of Maryland System, Solomons, MD 20688-0038. Ref. 
No.[UMCEES]CBL 93-030a. 
Boynton, W.R. and W . M. Kemp. 2000. Influence of river flow and nutrient loads on 
selected ecosystem processes: A synthesis of Chesapeake Bay data. Pp. 269-
298 In: J.E. Hobbie (ed.). Estuarine Science: A Synthetic Approach to 
Research and Practice. Island Press, Washington, D .C. 
Butman, C. A. 1986. Sediment trap biases in turbulent flow: Results from a 
laboratory flume study. J. Mar. Res. 44: 645-693. 
Butman, C. A., W. D . Grant, and K. D. Stolzenbach. 1986. Predictions of sediment 
trap biases in turbulent flows: A theoretical analysis based on observations 
from the literature. J. Mar. Res. 44: 601 -644. 
131 
Chesapeake Bay Benthic Monitoring Program. US EPA Chesapeake Bay Program 
Annapolis, MD. ' 
Chesapeake Bay Phytoplankton Monitoring Program. US EPA Chesapeake Bay 
Program, Annapolis, MD. 
Chesapeake Bay Water Quality Monitoring Program. US EPA Chesapeake Bay 
Program, Annapolis, MD. 
Conley, D . J. and T. C. Malone. 1992. Annual cycle of dissolved silicate in 
Chesapeake Bay: implications for the production and fate of phytoplankton 
biomass. Mar. Ecol. Prog. Ser. 81: 121-128. 
Cronin, W. B. and D . W. P,itchard. 1975. Additional statistics on the dimensions of 
Chesapeake Bay and its tributaries: Cross section widths and segment volumes 
per meter depth . Chesapeake Bay Institute. Reference 75-3. Special Repoit 
42. 
Dellapenna, T. M ., S . A. Kuehl and L. C. Schaffner. 1998. Sea-bed mjxing and 
particle residence times in biologically and physically domjnated estuarine 
systems: a comparison of lower Chesapeake Bay and the York River 
subestuary. Est. Coat. Shelf Sci. 46: 777-795. 
EPA 1993. Guide to using the Chesapeake ~ay Program_ Water Quality Monitoring 
Data. Chesapeake Bay Program Office, Annapolis, MD. 
Fisher, T . R., E. R. Peele, J. W. Ammerman, and L. W. Harding. 1992. Nutiient 
limitation of phytoplankton in Chesapeake Bay. Mar. Ecol. Prog. Ser. 82 (l): 
51-63. 
Fisher, T. R., A. B. Gustafson, K. Sellner, R. Lacouture, L. W. Haas, R. L. Wetzel, R. 
Magnien , D. Everitt, B. Michaels, and R. KaITh. Spatial and temporal 
variation of resource limitation in Chesapeake Bay. Mar. Biol. 133(4): 763-
778. 
Furlong, E. T. and R. Carpenter. 1988. Pigment preservation and remineralization in 
oxic coastal marine sediments. Geochim. Cosmochim. Acta. 52: 87-99. 
Garber, J . H., w. R. Boynton, J.M. Barnes, L. L. Matteson, J. L. Watts, and S. 
Stamme,john. 1989. Ecosystem Processes Component (EPC) and Benthic 
Exchange and Sediment Transformation (BEST). M~ryland J?epartment of the 
Environment, Maryland Chesapeake Bay Water Quality Monitoring Program. 
Chesapeake Biological Laboratory (CBL), Solomons, MD 20688 
Graf, G., W. Bengtsson, u. Diesner, R. Schultz, and H. Theede. 1982. Benthic 
response to sedimentation of a sp1ing phytoplankton bloom: process and 
budget. Mar. Biol. 67: 201-208. 
132 
Hagy, J. D ., L. P. Sanford, and W. R. Boynton. 2000. Estimation of net phys ical 
transport and hydraulic residence times for a coastal plai n estuary using box 
models. Estuaries. 23(3): 328-340 
Harding, L. W., Jr. and E. S. PeITy. 1997. Long-term increase of phytoplankton 
biomass in Chesapeake Bay, 1950-94. Mar. Ecol. Prog. Ser. 157: 39-52 1997 
Jensen, M . H. E. Lomstein , and J. S0renson. 1990. Benthic NRi + and No - flux 
3
following sedimentation of a spring phytoplankton bloom in Aarhus bight, 
Denmark. Mar. Ecol. Prog. Ser. 61: 87-96. 
Josefson , A. B. and D. J. Conley. 1997. Benthic response to a pelagic front. Mar. 
Ecol. Prog. Ser. 147: 49-62. 
Keller, A. A. and U. Riebesell. 1989. Phytoplankton carbon dynamics during a 
winter-spring diatom bloom in an enclosed marine ecosystem: piimary 
production, biomass and loss rates . Mar. Biol. 103: 131-142. 
Knauer, G. A. , D . M. Karl, J. H. Ma11in , and C. N. Hunter. 1984. In situ effects of 
selected preservatives on total carbon, nitrogen and metals coJJected in 
sediment traps. J. Mar. Res. 42: 445-462. 
Kemp, W. M. , E . M . Smith, M. Marvin-DiPasquale, and W.R. Boynton. 1997. 
Organic carbon balance and net ecosystem metabolism in Chesapeake Bay. 
Mar. Ecol. Prog. Ser. 150: 229-248. 
Kerhin , R. T., J . P. Halka, E. L. Hennessee, P. J. Blakeslee, D. V. Wells, N. Zoltan 
and R. H. Cuthbertson. 1983. Physical ch~racteristics and sediments budg;t 
for bottom sediments in the Maryland Portion of Chesapeake Bay. United 
States Environmental Protection Agency, Washington, D. C. 
Lacouture, R. Philadelphia Academy of Natural Science, Estuarine Research Center, 
St. Leonard, MD. 
Malone, T . C. 1992. Effects of Water Column Processes on Dissolved Oxygen , 
Nutrients, Phytoplankton and Zooplankton. In: Smith, D. E. , M. Leffler, and 
G. Mackiernan (eds.). Oxygen Dynamics in the Chesapeake Bay. A synthesis 
of recent research. Maryland Sea Grant College, College Park, MD 
Malone, T. c., D. J. Conley, T. R. Fisher, P. M. Glibert, and L. W. Harding. 1996. 
Scales of nutiient-limited phytoplankton productivity in Chesapeake Bay. 
Estuaries 19(2B): 371-385. 
Marsh, A. G. and K. R. Tenore. 1990. The role of nut1ients in regulating the 
population dynamics of opportunistic, surface deposit feeders in a mesohaline 
community. Limnol. Oceanogr. 35(3): 710-724. 
133 
Marshall , H. G . and K. K. Nesius. 1996. Phytoplankton composition in relation to 
primary production in Chesapeake Bay. Mar. Biol. 125 611-617. 
Parsons, T.R., Y. Maita and C.M. Lalli. 1984. Dete1mination of chlorophylls and 
total carotenoids: Spectrophotometric method. pp. 101-112 In: Parsons , T.R., 
Y. Maita and C.M. Lalli. A manual of chemical and biological methods for 
seawater analysis. Pergamon Press, Oxford. 
Peinert, R., A. Saure, P. Stegman, C. Steinen, H. Haardt and V. Smetacek. 1982. 
Dyn amics of primary production and sedimentation in a coastal ecosystem. 
Neth. J. Sea Res. 16: 276-289. 
Pritchard, D . W. 1952. Salinity distribution and ci rculation in the Chesapeake 
estuarine system. J. Mar. Res. 11 (2): 106-123. 
Smayda, T. The suspension and sinking of phytoplankton in the sea. 1970. 
Oceanogr. Mar. Biol. Ann. Rev. 8: 353-414. 
Smetacek, V. 1985. Role of sinking in di atom life-history cycles: ecological, 
evolutionary and geological significance. Mar. Biol. 84: 239-251. 
Smetacek, V. , K. v. Brockel, B. Zeitzscel and W. Zenk. 1978. Sedimentation of 
plankton diatoms during a phytoplankton spring bloom in relation to the 
hydrographic regime. Mar. Biol. 47: 211-226. 
Smith, E. M. and W. M . Kemp. 1995. Seasonal and Regional variations in plankton 
community production and respiration for Chesapeake Bay. Mar. Ecol. Prag. 
Ser. 116: 217-231. 
Strickland, J.D.H. and T.R. Parsons. 1972. A practical handbook of seawater 
analysis . Bulletin 167 (2nd ed.). J. Fish. Res. Bd. Can. Ottawa, Canada. 
Sun , M., R. c. Aller and C. Lee. 1991. Early diiagenesis of chlorophyll-a in Long 
Island Sound sediments: A measure of carbon flux and pa11icle reworking. J. 
Mar. Res. 49: 279-401. 
Sun , M.-Y. , c. Lee and R. C. Aller. 1993a. Laboratory studies of oxic and anoxic 
degradation of chlorophyll-a in Long Island Sound sediments. Geochim 
Comochim Acta. 57: 147-157. 
Sun, M.-Y., c. Lee and R. C. Aller. 1993b. Anoxic and oxic degradation of 14c_ 
14
labeled chloropigments and a C-labeled diatom in Long Island Sound 
sediments. Limnol. Oceanogr. 38(7): 1438-1451. 
Sun, M.-Y. , R. c. Aller and C. Lee. 1994. Spatial and temporal dist1ibutions of 
sedimentary chloropigments as indicators of benthic processes in Long Island 
Sound. J. Mar. Res. 52: 149-176. 
134 
Townsend, D . W. and L. M . Cammen. 1988. Potential importance of the timing of 
pring plankton blooms to benthic-pelagic coupling and recruitment of juvenile 
demersa l fishes . Biol. Oceanogr. 5: 215-229. 
135 
Table 4-1. The most abundant phytoplankton taxa (excluding picoplankton) in three 
regions of Chesapeake Bay during spring and the average fraction of total 
phytoplankton carbon contributed by diatoms. Phytoplankton species counts from 
unpblished Chesapeake Bay water quality monito1ing program data (available from 
USEPA Chesapeake Bay Program web site) . Unpublished carbon composition data 
provided by R. Lacouture (personal communication). 
Region Most Abundant Phytoplankton Taxa % of Total % of Total 
(excluding picoplankton) Phytoplankton Phytoplankton 
during Jan-Apr. Counts Carbon 
(% of cells) 
Upper Bay unclassified centric diatoms 1 (23%), 56% 59% 
Katodinium rotundatum,2(12%), 
Skeletonema costatum 1(12%), 
Crytomonas spp. 3(12%), Cyclotella 
spp. 1(8% ), Skeleton.ema 
potamos 1(7%). 
Mid Bay unclassified centric diatoms 1(17% ), 67% (above 69% (above 
Katodin.ium rotun.datum2(15%), pycnocline) pycnocline) 
Crytomon.as spp.3(15%), Cyclotella 
1
spp. (12%), Cerataulin.a 
pelag ica 1( 9% ), Skeleton.ema 
costatum 1( 9% ), Chaetoceros 
spp. 1(5%) 
Lower Bay Skeleton.ema costatum 1(20%), 83% (above n/a 
unc lassified centric diatoms 1(18%), pycnocline), 
Cerataulin.a pelaf ica 1( 9% ), 84% (below 
Crytomon.as spp. (9% ), unclassified pycnocline) 
pennate diatoms\9% ), Nitzschia 
pungen.s\8%), Rhizosolen.ia 
fragilissima4( 4% ), Rhizosolenia 
delicatula\3% ). 
1 2
centric diatoms, dinoflagell ates, 3crytomonads, 4pennate diatoms, 
136 
Table 4-2. Cruise dates for sediment chlorophyll -a surveys and the number 
of sediment cores collected in each region of the Bay. 
Cruise Dates Number of Cores 
Year Begin End Upper 
- Mid Lower Bal'. Ba~ Bal'. 1993 5/8/93 5112/93 8 25 23 
1994 Mid May 9 21 33 
1995 4/28/95 5/3/95 7 25 31 
1996 4/27/96 5/7/96 8 13 10 
1997 4/20/97 4/24/97 7 17 7 
1998 4/11/98 4/ 15/98 6 13 15 
1999 4/19/99 4/23/99 6 11 14 
_ 2000 4/29/00 512/00 6 8 6 
Table 4-3 . Results of a method comparison experiment used to 
evaluate the effect of three sonication treatments and single vs. double 
extraction on the amount (mean?se, % change from control) of chi-a 
(?gig) extracted from 15 aliquots of homogenized Chesapeake Bay 
sediments. Each sonication treatment was replicated 5 times. All 
effects (sonication,extraction and interaction effect) were statistically 
significant (Repeated-measures ANOVA, p<0.01). 
_ Single Extraction, ?gig Double Extraction, ?gig 
No Sonication 9.35 (0.02, 0%)* 10.32 (0.02, +10.4%)* 
Microsonication 9.78 (0.05, +4.6%) 10.84 (0.06, +15.9%) 
~nicator Bath 9.63 (0.03, +3.0%) 10.56 (0.04, +12.9%) 
* these treatments were also compared using 7 non-homogenized 
samples. The mean difference in those samples was 11.0?0.5% 
137 
Table 4-4. Regional/annual mean sediment total chlorophyll-a inventories. 
These were computed from 0-1 cm chi-a inventories by adjusting for mixing to 
below 1 cm on short time scales (i.e. days-weeks). Calculation of the overall 
mean accounts for differences in the area of the respective regions and is 
therefore not the mean of the regional means. 
Year Upper Bay Mid Bay Lower Bay Overall Jan-Apr 
Flow (m s-1) 
1993 84 182 161 155 2989 
1994 147 217 187 191 2624 
1995 139 130 98 117 1206 
1996 107 132 146 134 2383 
1997 223 243 231 235 1403 
1998 195 169 122 153 2471 
1999 315 323 150 244 1392 
2000 162 162 92 137 1739 
Average 172 195 148 171 2026 
138 
Table 4-5. Minimum, maximum and modal values used to 
specify triangular distributions for parameters in eq. 3. 
Parameter values were randomly drawn from these distributions 
and used in Monte Carlo simulations to estimate the mean and 
standard deviation of chi-a deposition in each region and year. 
Parameter Min Mode Max 
Initial chi-a concentration, Co, mg m-2 30 58 80 
First-order decay coefficient, k, d-1 0.02 0.028 0.04 
Period of Bloom deposition, t, days. 30 60 75 
139 
Table 4-6. _Esti mat~d average (?st~~dard devi~Jion) ch_l-a depo~iti~n rates (mg m -2 d-1) 
and total wmter-spnng chi -a depos1t1 on (mg m ) for wmter-sprmg m the upper, mid 
and lower Chesapeake Bay during 1993-2000. 
(a) Average deposition ra te during win ter-spring (mg m-2 d-1) 
_ Year Upper Bay Mid Bay Lower Bay 
1993 2.69?0.34 6.34?0. 73 5.56?0.64 
1994 5.04?0.58 7.64?0.88 6.52?0.74 
1995 4.74?0.55 4.40?0.52 3.21?0.39 
1996 3.55?0.43 4.48?0.52 4.99?0.59 
1997 7.86?0.90 8.61?0.98 8.16?0.94 
1998 6.81?0.79 5.85?0.67 4.10?0.49 
1999 11.29?1.28 l l.57?1.33 5.14?0.59 
2000 5.59+0.65 5.59?0.64 2.99?0.37 
Average 5.95 6.81 5.08 
(b) Winter-spring ch i-a deposition (mg m-2) 
-Year Upper Bay Mid 
1993 148?28 345?54 304?49 
1994 275?45 415?63 355?55 
1995 258?42 241?41 175?32 
1996 193?34 244?41 272?44 
1997 429?65 470?70 444?67 
1998 371?57 319?51 224?38 
1999 613?90 631?92 280?45 
2000 305+49 305?48 163?30 
~erage 324 371 277 
140 
.--... 
E 
.~.__ . 
C 
co 120 
Q) 
(.) 
0 110 
(.) 
:;:; 100 so 
C 
co 
+:::. 90 CHP <( 
E 
-0 I... PXT Q) 
(.) 
C 
..c.o..  
-~ 
0 ~ POT 
Q) ~ 
C 
C 
co 40 
..c 
u 30 40i~p 
20 
10 30 ~ YRK 
0 I ~ 
1 i 2 3 l 
d:JAM 
4 r T 5 l 6 7 1 8 9 10 11 12 
Month 
Fig 4- I . Average seasonal di stti buti on of water co Iu mn integrated 
2
chlorophyll-a (mg m- ) in Chesapeake Bay (1984-1999). Arrows 
on the right(~ ) indicate the locations along the central axis of the 
Bay of the Chesapeake Bay Water Quality Monitoring Program 
stations used to generate the plot. Three letter codes refer to the 
major tributary rivers and are as follows : JAM==James River, 
YRK==York River, RAP==Rappahannock River, POT==Potomac 
River, PXT==Patuxent River, CHP=: Choptank River, 
PTP==Patapsco River. The rectangle indicates the time period dur-
ing which surficial sediment sampling was usually conducted. 
Exact dates are in Table 1. 
141 
L 
39.5 
39.0 
38.5 II >80% silt/clay 
D 50-80% silt/clay 
l- J <50% silt/clay 
38.0 
37.5 
37.0 
_j 
-76.0 
Fig. 4-2. A map of Chesapeake Bay 
indicating regional boundaries and the 
distribution of sediment types as 
computed from the Chesapeake Bay 
Monitoring Program Benthic Data (data 
available from US EPA Chesapeake Bay 
Program Web Site). Distribution of 
sediment types is comparable to Kerhin et 
al. (1983). 
142 
Fig. 4-3. The distribution of total chlorophyll-a in the top 1 cm of Chesa-
peake Bay sediments during late spring in 1993-2000. The center of each 
circle indicates the location at which the core was collected, while the 
size of the circle indicated total ch lorophyll-a values. The values for 
1993 are estimated from total chlorophyll-a in the top 2 mm. The three 
letter codes adjacent to the 1993 map identify the major tributaries refer-
enced in Fig. 4-l. 
143 
mg m-2 
0 0 
0 102 
0 203 
0 305 
0406 
144 
350 
? 1999 Upper Bay 300 
250 
200 ? ? 
150 ? ? ? 
100 1995 e 1993 
? 
350 
..--. e 1999 Mid Bay 
~ 300 
E 
CJ) 
E 250 .1997 .__.. 
cu 200 
I ? 
>, 
..c 150 ? ? ? Fig 4-4. Regional and overall mean 0.. 
,0._  ? ? sediment total chlorophyll-a invento-0 100 ries in late spring related to winter-
..c 
0 sp1ing (Jan-Apr) average Susquehan-
_, 
C 350 na River flow. Sediment inventories 
Q) 
Lower were computed from 0-1 cm cores. E Bay 
-0 300 The whole Bay mean reflects differ-
Q) 
Cf) ences in the size of the respective 
cu 250 e regions. Note the y-axis scales , which ?u 1999 vary to emphasize within-region pat-
t 
::J 200 ? tern rather than among region differ-Cf) ences. 
_cu,  150 ? 
~ ? ? 
100 ? ? 
350 ? 
Whole Bay 
300 
250 I 1999 
1997 
200 ? 
150 ? ? ? 
100 ? ?1 996 
50 
1.0 1.5 2.0 2.5 3.0 3.5 
River Flow (103 m3 s-1) 
145 
120 - ---
(j c?) 
I 
::2 1997 u 100 
1.0 1.5 2.0 2.5 3.0 3.5 
Jan-May Avg. Susquehanna 
River Flow (103 m3 s- 1) 
Fig 4-5. January-April average water column 
integrated chi-a in the lower Chesapeake Bay 
during 1993-2000 related to Jan-Apr average 
Susquehanna River flow . A second-order poly-
nomial explains 97% of the variation, excluding 
the 1997 observation. 
146 
Upper Bay 
~~8 J ce) 
1999 i 
250 
.1997 
200 1998 
150 ? ? 
100 1993 r2=0.73 ce) 
0 50 100 
150 
~ 
E 350 J~Mi-d Bay- -?. .------1 Fig 4-6. The relationship between Jan-
OJ 300 1999 ,Y uary-Apiil average water column inte-
'?E;  250 .1997 grated chlorophyll-a and sediment 
chlorophyll-a in each of three regions 
I 
>.. ? of Chesapeake Bay. There is a signifi--?_ 200 ? cant c01Telation in the lower Bay; the 2 1998. 0 ? indicated line is the model II regres-..c u 150 sion line. For the upper Bay, the trend 
....... ? ? line indicates the model II regression C 
Q) line excluding the 1993 and 1999 
E 100 observations. 
"O 
Q) 
CJ) 0 50 100 
150 
Lower Bay 
300 
1997 
250 ? 
200 ? 
1999 
? ? 150 ? r2100 =0.52 
0 50 100 
150 
Water-Column Integrated Chi-a 
(mg m-2) 
147 
400 
Upper Bay 
300 
N 
I 
E 
200 4/9 0) 
E 2/17 5/9 
I 
100 
0 r r ,- r I - -. 
400 
4/6 
Mid Bay 
4/3 
300 5/14 
C";J 
E 200 
0) 
E 
100 
0 
400 
Lower Bay 
300 
4/27 
I 5/ 14 4/) 1 
C";J 
E 200 
0) 
E 
100 
0 
1/93 1/94 1/95 1/96 1/97 1/98 1/99 1/00 1/01 
Fig. 4-7. Water column integrated chlorophyll-a concentrations in Chesa-
peake Bay averaged by region . Vertical dotted lines indicate the dates of sedi-
ment chlorophyll-a mapping studies. Dates indicate the date of the adjacent 
water column chi-a observation, which can be compared to the sediment chi-a 
mapping dates indicated in Table 2. 
148 
2500 
C::J Direct Measurement 
Computed from 
2250 Chi-a Flux 
2000 
1750 
~ 
"'O  1500 
~ C 
E 
0 
O> 1250 
E 
1000 
750 
500 
250 8 9 10 11 
3 4 5 
6 7 
2 
Month 
Fig. 4-8. Monthly means and standard eITors of particulate carbon (PC) sinking 
fluxes measured using sediment traps just below the pycnocline in the mid 
Chesapeake Bay during 1984-1992. Sediment trap data from Boynton et al. 
(1993) and re lated unpublished data. Stippled bars indicate vertical PC fluxes 
computed from chi-a fluxes. Reference lines indicate: A=spring average PC 
deposition (510 mgC m-2 d-1, this study); B=March-April average PC deposition 2 1
computed from chi-a flux to sediment traps (720 mgC m- d- ); C=March-Apiil 
average deposition computed directly from PC flux to sediment traps (C:chl-
a=75 in all cases) 
149 
Chapter 5: 
PATTERNS OF MACROBENTHIC BIOMASS AND COMMUNITY 
BIOENERGETICS IN CHESAPEAKE BAY DURING SUMMER IN 
RELATION TO HABITAT QUALITY AND ORGANIC CARBON SUPPLY 
Abstract 
Macrobenthic biomass and bioenergetic rates in the mainstem of Chesapeake 
Bay were related to bottom water quality and sediment characteristics and estimates of 
carbon supply. The source of data for this analysis was a 15-year monitoring record 
describing biomass, abundance and species composition for 1664 sampling events . 
Independent variables, collected concurrently with the macrobenthic data, included 
water temperature, salinity, dissolved oxygen (DO), sediment % silt-clay, % nitrogen 
and % carbon. Although univariate and multiple linear regression models were poor 
predictors of biomass (r2<0.35), a tree-structured regression model predicted biomass 
2
with r =0.59. In the tree-structured model, DO was the most important independent 
variable, followed by salinity and % silt-clay. Other variables were less important. 
Daily macrobenthic production was estimated via a temperature-dependent allometric 
relationship. Body size was greater in the upper Bay and lower Bay compared to the 
mid Bay region, leading to higher estimates of production/biomass (P/B) ratios for the 
mid Bay, especially for the deep water benthos. However, regional mean summertime 
P/B ratios varied only -2-fold, while regional mean biomass varied -100-fold. Thus, 
summertime macrobenthic production was determined primarily by the biomass 
distribution, not P/B. 
150 
Bioenergetic rates for the benthos during summer were estimated using a range 
of values for net growth efficiency and assimilation efficiency. Daily carbon 
requirements were compared to net plankton production and the ration potentially 
avai lable via suspension feed ing, two indicators of organic carbon availabi li ty. Actual 
and potential suspension feeding was estimated using a simple particle mixing model 
and c learance rates estimated via an allometric relationship. Organic carbon available 
to the benthos greatly exceeded the daily ration in the mid Bay (>IO-fold) and was 
adequate elsewhere. Overall , the results suggest that degraded benthic habitat, food 
quality, predation , or other factors, but not organic carbon supply, limited biomass and 
productivity of the mid Bay benthos. The results also suggested that organic carbon 
limitation of the upper and lower Bay macrobenthos was unlikely, but the carbon 
surplus was smaller than in the mid Bay. 
Introduction 
Macrobenthic communities are an important component of many aquatic food 
Webs because they can consume significant amounts of organic matter and also 
provide an important trophic link between primary producers and detritus and larger 
consumer organisms (Diaz and Schaffner 1990). Compared to other marine systems, 
the benthos in coastal and estuarine environments can be pai1icularly important 
because relatively shallow water depth increases the interaction between the pelagic 
and benthic communities (Kemp and Boynton 1992). Suspension feeding benthos 
may be able to access the entire water column in shallower habitats, increasing their 
production potential and allowing them to exert significant grazing pressure on the 
151 
phytoplankton community (e .g. San Francisco Bay, Cloem 1982; Chesapeake Bay, 
Gerritsen et al. 1994; Thompson 2000; Great Lakes, Budd et al. 2001). The passive 
organic flux to sediments may also be increased in shallow water because respiratory 
losses in the water column are reduced. Even in deeper waters, rapid deposition of 
intact phytoplankton following the conclusion of spring phytoplankton blooms can 
stimulate benthic production (Graf et al. 1982, Kitazato et al. 2000). 
Despite these genera/ observations, estuarine and coastal benthic communities 
do not always have high biomass, nor are they always highly productive. Even among 
highly productive estuaries, benthic communities range from extremely productive 
(Pfitzenmeyer and Drobeck 1963, Phelps 1994) to mostly afaunal (Holland et al. 
1977). The universe of possible factors that can adversely affect the benthos is large 
and may include predation, physical-transport regimes, physical disturbance, food 
quantity and quality, sediment characteristics, salinity, water temperature, dissolved 
oxygen, and toxic contamination (reviewed by Diaz and Schaffner 1990, Herman et al. 
1999, Diaz and Rosenburg 1995). The large number of factors and the likelihood that 
effects interact and are non-linear or discontinuous (e.g. thresholds), complicate the 
task of formulating statistical relationships that predict benthic biomass. Even if one 
restricts the analysis to readily observed variables such as water quality and sediment 
characteristics, the number of factors and interactions remains large. The challenge of 
fitting an adequate statistical mode/ may lead investigators to find only weak effects, 
even for variables known anecdotally or via experimental evidence to be important 
(e.g. Holland et al. 1987). 
152 
In the face of these challenges, describing patterns of abundance and 
production of organisms, and developing an ability to predict and explain the patterns 
quantitatively is fundamentally important for understanding estuarine benthic 
communities (Constable 1999). Moreover, this type of information can help resolve 
and predict long-term trends in naturally varying environments. Without highly 
resolved models, larger changes are needed before detection is possible. With this in 
mind, an objective of this study was to investigate the habitat characteristics that 
influence biomass and production of Chesapeake Bay benthic communities using tree-
structured data analysis (Breiman et al. 1984), a new statistical procedure that is 
particularly well-suited to the task of formulating predictive models when there exist 
many possible explanatory variables, complex interactions, and non-linear and/or 
discontinuous responses. 
Another objective of this study was to examine the hypothesis that food 
resources (e.g. primary production) do not limit benthic communities in Chesapeake 
Bay. This hypothesis is suggested by the substantial increase in phytoplankton 
biomass and primary production that occurred in the Bay over the past 50 years 
(Harding and Perry 1997), which has not been accompanied by a similar increase in 
macrobenthic biomass and production. Rather, a decline has been observed in some 
benthic infauna) communities (Holland et al. 1987). In general, this hypothesis runs 
counter to studies of the macrobenthos, which often find that biomass and production 
are enhanced by additions of organic matter, particularly organic matter with 
appropriate nutritional composition (Graf et al. 1982, Marsh and Tenore 1990, 
Josefson and Conley 1997, Hansen 1999). Comparative ecological studies show that 
153 
increased macrobenthic biomass is assocated with elevated primary production or 
algal biomass in both lakes and marine systems (Hanson and Peters 1983, Kemp and 
Boynton 1992, Herman et a l. 1999). However, these observations can potentially be 
reconciled with the hypothesized lack of food limitation via the observation of Pearson 
and Rosenburg (1978) that organic enrichment often leads to habitat degradation , 
principally via oxygen stress. This degradation promotes successional changes in the 
benthic community that include a decline in biomass , ultimately to zero. 
The potentially competi ng effects of organ ic enri chment on the Chesapeake 
Bay benthos have been examined by at least two studies. Blumenshine and Kemp 
(1999) foc used on the mid Bay, developing monthly organic carbon budgets for the 
macrobenthos and relating likely carbon requirements to esti mates of the carbon 
avail able . Their analysis, most appropriate for a community of deposit feeders, 
assumed that the organic carbon resource available to the benthos was determined by 
net plankton metaboli sm. They also rel ated benthic biomass in the mid Bay to an 
index of oxygen stress, directly addressing the potenti al for control of the benthos by 
habitat quality. In contrast, Ge1Titsen et al. (1994) focused on the mid- and upper Bay 
suspension-feeding communities, estimating the carbon that could be obtained via a 
particle mixing model and estimates of clearance rates. Both studies concluded that, 
in contrast to general trends observed in comparative studies, organic carbon was 
rarely limiting to the benthos in the regions of Chesapeake Bay that they examined. 
This study adds significantly to the contributions of the former studies by expanding 
the results to include all regions of Chesapeake Bay and by implementing the 
154 
computations using the large quantity of data accumulated by the Chesapeake Bay 
Benthic Monitoring Program. 
~Judy Sites 
Chesapeake Bay and its tributaries form a large estuarine complex within the 
middle Atlantic region , USA (Fig. 5-1). The salinity gradient (0-32 ppt) of the 
mainstem Bay extends over 300 km. The estuary is commonly construed to consist of 
three somewhat distinct ecological zones, characterized by differing physical 
circulation regimes, salinity-dependent differences in biota, and a gradient in 
community metabolism (e.g. Kemp et al. 1997). Sediments in the Bay range from 
sand to organic-rich silt/clay; silt-clay content generally decreases southward in the 
Bay, but all sediment types are present in each region (unpublished data, Chesapeake 
Bay Benthic Monitoring Program, Chapter 4). 
The mid-Bay region is characterized by strong summertime water column 
stratification at -10 m depth. Combined with high metabolic rates, this promotes 
seasonal hypoxia and anoxia that first develops in late spring in deep water near 39?N 
and expands southward to - 37?40', often extending up to or into the pycnocline (see 
Chapters 2 and 3). Periodic hypoxic events have been noted in shallower waters along 
the mid-Bay mainstem (Malone et al. 1986, Breitburg 1990). Although less prevalent, 
hypoxia also occurs in other regions of the Bay and in the major tributa,ies (Hagy and 
Boynton 2000). 
155 
Methods 
!@.ta Sources 
The data source for this study was a 15-year record of species-level 
macrobenthic biomass and abundance for Chesapeake Bay and tributaries collected 
between 1985 and 1999 by the Chesapeake Bay Benthic Monitoring Program 
(CBBMP). The database desc,ibes the benthic macrofauna (abundance, biomass, and 
taxonomic composition) collected in more than 6,000 cores collected in aJI seasons 
throughout Chesapeake Bay and tributaries (EPA2000a,b,c). Concurrently collected 
data desczibes water quality (water temperature, salinity, dissolved oxygen) at the time 
of collection as well as sediment properties (%sand, %si lt-clay, C-H-N) at most sites. 
The sampling protocol changed over the period of record and includes a mixture of 
sites with fixed locations and sites that were randomly located within spatial strata. As 
this study was not designed to investigate temporal trends at fixed sites, both random 
and fixed sites were considered together in order to achieve the largest possible 
number of observations. 
Survey design, as well as sampling and analytical methods are described in the 
documentation for the database (EPA 2000a, b, c). Sampling gears varied over time 
and according to the total depth at the site. In shallow water, a post hole digger or 
Petite Ponar Grab was used. In deeper waters, a Ponar grab, a WildCo or other box 
corer, or a Young-modified Van Veen grab was used (e.g. Holland et al. 1987). For 
2 
most samples, the gears sampled 175 or 250 cm of sediment surface and penetrated 
-25 cm into the sediments. Some investigators have reported that gear limitations 
may have prevented the CBBMP from properly sampling key larger, deep-dwelling 
156 
animals, particularly the polychaete Chaetopterus variopedatus, a dominant organisms 
in the lower Bay (Thompson 2000, L. C. Schaffner, pers. comm.). Separate estimates 
of biomass and production for C. variopedatus in the lower Bay were therefore 
obtained from Thompson (2000). All samples collected by the CBBMP were sieved 
to 0.5 mm and fixed in buffered formalin. Methods for ash-free dry weight (AFDW) 
determinations vaiied over time. In some years, all samples were dried and ashed 
' 
While in other years AFDW determinations were made only for certain abundant 
species using regressions on morphometric measurements. In these cases, the 
abundant taxa accounted for most of the total biomass. 
Tree-Structured Regression Models 
Measures of community biomass were related to possible habitat factors (e.g. 
sediment percent silt-clay, water temperature, salinity, dissolved oxygen, etc.) using 
CART (Classification and Regression Trees; Breiman et al. 1984, see also Hagy 
2000), a tree-structured data analysis algorithm used by CART Software for Windows 
(Salford Systems, Inc. , Steinburg and Colla 1995). Hereafter, I use the term "CART" 
to refer to both the CART algorithm and it's the implementation by CART software. 
CART can be used for both regression and classification; however, this study only 
used CART for regression. Therefore, all subsequent discussion refers to regression-
tree models. CART employs a recursive partitioning algorithm that finds a binary 
question, or splitting rule (e.g. Is salinity~ 18.5?), that divides a group of observations 
(parent node) into two groups of observations (child nodes) such that the variance of 
the target variable is minimized within the child nodes. The best splitting rule is 
157 
determined by examining all possible rules. The algorithm then treats the child nodes 
as parent nodes, recursively identifying splitting rules until each child nodes contains 
too few observations to be split further. The resu lting tree is called the maximal tree. 
While the maximal tree always has the minimal residual eITor, a smaller tree 
often has the minimal eITor when tested agai nst data not used in model construction. 
For this reason , a cross-validation procedure was used to determine eITor rates. For 
this study, CART was confi gured to use 10-fold cross-validation (the recommended 
default option; Breiman et a l. 1984, Steinberg and Colla 1995). In this procedure, all 
the data, 10% at a time, are set aside from model development as validation 
observations. T he results are then pooled such that all the data are used both in model 
formation and in cross-va lidation . 
CART computes the cross-validation eITor rate for the max imal tree, then 
considers smaller trees through a process called "pruning." Pruning involves 
collapsing child nodes back into parent nodes, then recomputing the cross-validation 
eITor rate. The nodes that account for the smallest decrease (or an increase) in eITor 
rate are pruned first, followed by other nodes. CART initially selects the tree that has 
the lowest cross-validation eITor rate, but the investigator may choose among models 
along a continuum of complexity vs. error rate . Experience indicates that there is 
often a "break" point where additional reduction in model complexity results in a large 
increase in the error rate, making the model selection relatively clear. 
Because CART selects sp litting rules by minimizing within-node variance, 
regression trees are sensitive to heterscedasticity, which can lead to spurious splits of 
parent nodes with high mean and variance and undersplitting of low mean and 
158 
variance nodes (Steinburg and Colla 1995). To minimize this effect, biomass values 
were re-coded and transformed using X' = log(X +0.01). Back-transfo,med 
predicted values give the geometric means, which are less than the arithmetic means. 
The latter, used to preserve mass for bioenergetic computations, were computed from 
untransformed observations c lassified to each node. These were simi lar to predicted 
2
va lues back-transformed using X = w x'+a- !2 -0.01, where cr2 is the node variance. 
Community Biomass and Production 
Community biomass was determined by summing the ash-free biomass of all 
the taxa in each replicate sample. Thus , all reported biomasses are ash-free dry 
weights unless otherwise indicated. Average community biomass for sites with 
repli cation were computed by averaging the replicates. 
Daily production was computed for each taxa in each sediment core in the 
benthic database using the empirical model of Edgar (1990), which was de1ived from 
a meta-analysis of production estimates for benthic macrofaunal populations (Edgar 
1990). This model relates daily production (P, ?g d-1) for a single macrobenthic 
animal to body size (B, ?g AFDW) and water temperature (T, ?C) with r2=0.94. Edgar 
(1990) formulated different models for different taxonomic groups (e.g. bivalves, 
crustaceans, polychaetes) but found them to be indistinguishable from the general 
model. Therefore, thi s study used the general model: 
p = 0 .0049Bo.soyo.s9 (1) 
159 
The Edgar (1990) model was derived from data describing daily growth of 41 
invertebrate species varying in size from 10-5 g to l g and was reported to be 
applicable for water temperatures from 5-30 ?C. It was assumed for this study that the 
model is also applicable to temperatures below 5?C (10% of cases) since it predicts 
minimal production , the expected result for a temperate estuary during winter. To 
apply the model to the Chesapeake Bay benthos, bottom water temperature was 
obtained from the concu1Tently collected water quality data. Where water temperature 
Was less than zero (0.3% of cases), a temperature of zero was assumed. If bottom 
Water temperature was missing (0.6% of cases), a seasonal mean was used. Body size 
Was computed by dividing the total biomass of each taxa by the reported abundance. 
If either the abundance or biomass was missing (10% of cases), the missing value was 
computed using the median body size for that taxa, which was computed from the 
remainder of observations. Because the effect of body size is non-linear (eq. 1), 
community production was computed as the sum of production for each taxa, rather 
than as the production for the average size organism in the community. Thus, 
community production (P, ?g m-2 d-1) was computed as 
P= ~ N?P. (2) 
I I 
Where pi is the production per individual of the /h species and Ni is the abundance of 
that species (m-2). 
160 
~arbon Requirements for the Benthos 
Carbon budgets were formulated for Chesapeake Bay benthic communities for 
the purpose of comparing likely carbon demand to the organic matter supply rate. The 
scope of this study was limited to summer (June-August). These budgets have the 
form C = P + R + U , where C=consumption, P=production , R=respiration and 
D==excretion+egestion (feces production). Ash-free dry weights for all organisms 
Were converted to carbon using 500 mgC = 1 g AFDW (J0rgensen 1979). Rand u 
Were computed from production via ratios based on published results of both empiiical 
and theoretical studies (Schroeder 1981, Bayne and Newell 1983, Peters 1983, 
Schwinghamer et al. 1986). Respiration was computed via the net growth 
efficiency, NGE = P/(P + R) . Consumption was computed via the gross growth 
efficiency (GGE) , which is equal to the product of NGE and assimilation efficiency 
(AE), where GGE = P/C = NGE. AE. Both the technical challenges of measuring 
these ratios and real variability make precise estimates for these ratios for all the taxa 
unattainable. However, most estimates fall within a range of uncertainty, the extremes 
of which were considered as alternative cases. NGE was assumed to vary between 
30% and 56%, while AE was assumed to vary between 40-60% (Banse 1979, 
Schroeder 1981, Bayne and Newell 1983). Resulting estimates of GGE are 12-34%, 
similar to the range utilized by Blumenshine and Kemp (2000). Genitsen et al. (1994) 
assumed GGE=l0%, most likely yielding a generous estimate of carbon demand. 
161 
?.arbon Flux to the Benthos 
The carbon available to the benthos was computed in two ways. In the first 
approach , the carbon supply was estimated as net plankton production plus the net 
subsidy of carbon to each region due to physical transport of organic matter from 
adjacent regions of the Bay. By mass balance, this approach effectively provided an 
estimate of the carbon actually delivered to the benthic community. Net plankton 
production was computed for different regions of the Bay using estimates of gross 
primary production and plankton community respiration computed from data reported 
by Kemp et al. (1997), Smith (2001) and Harding et al. (2001) (see also Appendix). 
NPP was computed using NPM = GPP-Rpz, where GPP=gross plankton production 
(g 3 0 2 m-2 ct-\ plankton respiration (g 0 2 m- d-') and z=  mean depth (m). For the 
mid-Bay, Rp was reduced by 62-100% in deep water (Kemp et al. 1997) and these 
lower rates were used for depths > 10 m. Because metabolic rates based on net o2 
evolution include both production and respiration, these were converted to carbon 
assuming the photosynthetic quotient (PQ) = respiratory quotient (RQ) =1.0 (Stokes 
1996). 
Net physical carbon inputs were estimated for the three major regions, but not 
for sub-regions and depth zones. The Susquehanna River was assumed to be the only 
significant allochthonous source of total organic carbon (TOC) to the upper Bay 
region (Kemp et al. 1997). Monthly mean TOC loading from the Susquehanna River 
at Conowingo Dam was computed from data obtained from the United States 
Geological Survey River Input Monitoring Program (described by Langland et al. 
162 
200 l). Net physical particulate organic carbon (POC) transport among the regions and 
between the lower Bay and the Atlantic Ocean was assumed to be negligible, 
consistent with the strong particle trapping properties that have been observed for the 
Bay in general and the upper Bay in particular (Schubel and Carter 1977, Hobbs et al. 
1992). Although these studies refer to sediments in general, it was assumed to apply 
to POC as well . Dissolved organic carbon transport among regions was estimated as 
the product of monthly surface layer and bottom layer average DOC concentration and 
corresponding average advective transport rate (Chapter 2, also see Hagy 1996). 
A second approach to computing the carbon supplied to the benthos involved 
using a random walk (Brownian motion) model to examine the volume of water that 
could be and actually is filtered by the suspension feeding benthos during summer in 
three regions of Chesapeake Bay (Gerritsen et al. 1994). Genitsen et al. (1994) 
developed and applied the model for the upper and mid Bay, while this study extended 
its application to the lower Bay as well. Gerritsen et al. (1994) derived the model and 
described it in significant detail. Therefore, only an abbreviated description follows . 
The estuarine cross section was divided into 4 compartments (Fig. 5-2). These include 
two littoral zones flanking a pelagic surface mixed layer (SML) on its eastern and 
Western margins. Suspension-feeding was assumed to occur only in the littoral zone. 
The SML overlies the deep mixed layer ("profundal zone"), which in summer was 
assumed to be completely isolated from the SML by the pycnocline. The random 
Walk model computed the mean time required for a randomly mixing parcel of water 
to reach the bottom Ctv, days) as 
163 
z2 
t =-- (3) 
" 3D 
" 
2 
where Z is the mean depth (m) and Dv is the vertical dispersion coefficient (m d-1). A 
value 10 cm2 s-' was assumed for D11 (Gerritsen et al. 1994). Accordingly, the 
probability that a parcel of water approaches the bottom in 24 hours is 
P,, = 1- exp(- 1/t,,) (4) 
The mean time for probability that a parcel of water in the SML reaches the boundary 
with either (east or west) littoral zone (t1i , days) was computed as 
xz 
t =-- (5) 
II 12D 
II 
where X (meters) is the average width of the SML and D11 is the lateral dispersion 
coefficient. A value of 10 m2 s? ' was assumed for DH, much greater than D 11 
(Gerritsen et al. 1994). Analagous to eq. (4), the probability that a water parcel in the 
SML crosses into either littoral zone (PH) was computed as 
(6) 
Since the exchange of water into and out of the littoral zones is continuous, some 
water crossing into the littoral zone is not "new" water. The fraction of "new" water 
from the SML approaching the bottom of the littoral zone (P1-1v) was computed as 
P11v = P11 {1- t ., [1- exp(-1/tv) ]} (7) 
164 
Using these expressions, the volume of water approaching the littoral zone bottom 
each day (Vs) was computed as 
(8) 
where VL is the combined volume of the littoral zones and Vs is the volume of the 
SML. Thus, the fraction of the combined volumes of littoral zones and SML reaching 
the bottom each day is VB /(VL + V5 ). 
The Gerritsen et al. (1994) mixing model was modified for the lower Bay by 
assuming that there is not a central pelagic channel in the lower Bay, isolated from the 
benthos (see also Thompson 2000). Thus, the volume of water approaching the 
benthos (Vs) in this "one box" formulation (Fig. 5-2, lower panel) was simply VPv 
where Vis the total volume. Although Gerritsen et al. (1994) pattitioned the upper 
Bay in the same way as the mid Bay (Fig. 5-2, upper panel), their value of X was 
sufficiently small that eq. (8) effectively reduced to VB = V~, . Therefore, this study 
simplified the model for the upper Bay, assuming a "one box" model. 
The clearance rates of suspension feeders (m3 animar' d-1) was computed 
using C = 0.120W 0 75 ? where Wis body size (g DW). Genitsen et al. (1994) applied a 
temperature correction to this equation. Because this study was limited to summer, 
the temperature co1Tection factor was always 1.0 (Gen-itsen et al. 1994). A species-
specific clearance rate model, C = (24/1000)(22.17W +0.048), where W=0.025 g 
AFDW, was used for the abundant suspension feeding polychate Chaetopterus 
variopedatus (Thompson 2000). 
165 
The probability that a parcel of water approaching the bottom that will be filtered by a 
suspension feeder (PF) was computed as 
(9) 
and the vo lume actually filtered is VF = V8 PF . The daily ration (R, mg c m-2 d-1) 
assoc iated with the clearance estimate was computed using 
(10) 
Where Pi s net primary production (H14C03- assimilation, mgC m-2 d-1), VF = daily 
Volume filtered, Vr= total segment volume, Ar= total segment surface area, and AL== 
the area of the littoral zone. In the case of the upper Bay and lower Bay Ar== AL. This 
approach ass umes complete mixing within the littoral segments and within the pelagic 
SML. Estimates of summer primary production in each region of the Bay were 
obtained from Harding et al. (2001). 
Results and Discussion 
~rns of Macrobenthic Biomass vs. Habitat Factors 
2
Macrobenthic biomass varied between zero and 717 g m- , with a median and 
2
mean among 1664 observations of 1.4 and 10.6 g m- , respectively. Plots of 
macrobenthic biomass vs. depth, water temperature, salinity, dissolved oxygen, and 
Percent s ilt-clay in sediments illustrated how this overall range of values was 
di stiibuted with respect to these variables (Fig. 5-3). While these plots showed that 
none of these variables was a strong predictor of biomass, several obvious features 
Were apparent. For example, the highest biomass was found consistently at sites Jess 
than -5 m in depth and very low or zero biomass was rarely observed at those depths. 
166 
There was not, however, a clear overall relationship between biomass and depth . 
Similarly, while temperature was not a good predictor of biomass, the lowest biomass 
values were found only at summer water temperatures. Biomass also followed 
di stinctive patterns with respect to salinity. Nearly or completely afaunal sites were 
found almost exclusively at mesohaline sites. In contrast, polyhaline sites had 
moderate to high biomass with little va1iability and biomass was more variable among 
oligohaline sites. Hypoxic sites contained many, but not all, of the low biomass values 
and none of the highest biomass values. Finally, sediments at most sites were either 
muddy(> 75% silt-clay) or sandy (<5% silt-clay) and both the lowest and highest 
biomass was found at muddy sites (Fig. 5-3). 
From the perspective of statistical modeling, these relationships posed several 
challenges. All of the univariate relationships are weak and most are not monotonic 
(DO is an exception), either in terms of mean or variance. Some of the patterns can be 
explained in terms of water quality patterns characteristic of the estuary. For example, 
hypoxia and anoxia occurs reliably only in deep waters of the mesohaline zone of the 
Bay during summer. In these areas, sediments tend to be muddy. This is a likely 
explanation why many of the afaunal observations occurred where depth >lOm, water 
temperature is between 23-27 ?C, salinity is between 10-24 ppt, and silt-clay content is 
>75%. It was of interest, however, to predict biomass for less obvious circumstances. 
A multiple linear regression model was used to relate macrobenthic biomass to 
Water quality and sediment characteristics (Table 5-1). The model included DO, 
salinity, water temperature, %silt-clay in sediments and all two-way interactions 
among those variables. In addition, the model included season (Dec-Feb, Mar-May, 
167 
June-Aug, Sept-Nov) as a categorical variable. The model explained 35% of the 
variance and included many significant sources of variation. Dissolved oxygen and its 
interactions with water temperature and % silt-clay content in sediment were most 
important. Such a model could likely be improved by adding additional terms to 
account for more of the apparent non-linearity. For example, DO might be recoded as 
a classification variable indicating "hypoxic" or "normoxic." However, that approach 
would involve imposing arbitrary thresholds determined by the investigator, even if 
th0se threshold were loosely based on graphic observation, and would likely lead to 
simple confirmation of patterns already known to exist. As an alternative, tree-
structured regression (CART) was used to identify optimal rather than arbitrary 
thresholds and to uncover where possible any less obvious patterns of biomass in 
relation to the explanatory variables. 
The CART model included the same explanatory variables as the multiple 
regression model, except that sediment % carbon and% nitrogen were also added due 
to the fact that CART does not exclude observations that have missing values for 
explanatory variables. Seasonality was modeled via the month of the year rather than 
a seasonal classification variable because CART does not assume a linear response. 
Therefore, CART can determine seasonal definitions without a priori specification. 
The CART regression-tree with minimum error on cross-validation had 27 terminal 
nodes (Fig. 5-4). Variable importance statistics computed by CART indicated that 
dissolved oxygen was the most important variable, giving it a relative importance 
score of 100 (Table 5_2). The other variables, in order of decreasing importance were 
salinity (71), percent silt-clay (63), total depth (58), month (39), water temperature 
168 
0 8), percent carbon (8) and percent nitrogen (5). Vari able importance stati sti cs are 
computed by CART and are based not only on the appearance of a vari able as a 
primary splitting rule, an obvious indicator of importance, but also as competitors. 
For example, if a splitting rule based on variable X2 was nearly as predictive as the 
primary splitting rule, based on X1, then X1 and X2 would have similar importance 
scores, even though X1 was the primary splitting variable. One way this can occur is if 
X1 and X2 are (locally) correlated, although this is not the only way. Considering only 
primary splitting rules reduced the relative importance of salinity, percent silt-clay, 
month and water temperature, and decreased the importance of percent carbon and 
nitrogen to zero (Table 5-2). 
Box plots of the distribution of biomass among observations classified into 
each of the 27 terminal nodes showed that the regression tree was able to predict 
benthic biomass, but that the 27 nodes resolved some different types of habitats that 
had similar levels of biomass (Fig. 5-5). Many of the terminal nodes had substantial 
numbers of observations (25-75), but three nodes (12, 15 and 23) had particularly 
large number of observations. These types of sites were particularly well studied due 
to the relatively constant nature of the monitoring program. Predicted values were 
COJTelated with observations with r2=0.59 (Fig. 5-6), indicating that this model 
achieved the goal of improved prediction compared to multiple regression. As a 
cautionary note, while certain aspects of this model may be generally applicable, the 
scope of inference for this model is limited to the mainstem Chesapeake Bay. A more 
generally applicable model could be constructed using CART, but would require data 
from other estuaries. 
169 
Observations based on the regression tree (Fig. 5-4, Table 5-3) are described 
below. Dissolved oxygen (DO), the most important predictor of macrobenthic 
biomass, initially divided the sample (N=l664) into two groups, one with DO:::; 2.34 
mgr' (N=242). Among these low DO observations , the highest biomass was found 
2
among observations from May or June (l.5 g m- ; Table 5-3 node Tl) , when seasonal 
hypoxia had developed only recently (e.g. Chapter 2). Among observations from July 
or later, depth was important, whereas observations from above the usual depth of the 
2 
PYcnocline (<10.4 m) had an average biomass of 1.1 g m- and those from deeper 
Waters had negligable biomass, <0.1 g m-2. The higher biomass among shallower 
hypoxic sites suggests that the observed hypoxia may have been an intermittent event 
' 
either due to advective intrusion of hypoxic water (e.g. Breitburg 1990) or temporary 
Do depletion. Macrobenthos may survive temporary exposure to hypoxia, but are less 
likely to survive the persistent hypoxia that occurs at deeper depths (Diaz and 
Rosenburg 1995). At all depths severe hypoxia was associated with lower 
macrobenthic biomass than moderate hypoxia at comparable depth, indicating more 
rapid and extensive mortality when DO was very low (Diaz and Rosenburg 1995). 
At most sites for which benthic data were collected, DO was >2.34 mgr' 
(N::::1422). Although it is evident that deep habitats known to suffer persistent hypoxia 
or anoxia were sampled Jess frequently, this also reflects the fact that hypoxic 
conditions are a relatively restricted phenomenon, occun-ing mostly in deeper water 
during summer. However, latent effects of hypoxia may appear where DO is in the 
normoxic range because effects of seasonal or periodic hypoxia may persist after 
normoxic conditions are restored (e.g. Pihl et al. 1991). In this dataset, DO itself 
170 
Usually did not predict macrobenthic biomass when DO was observed to be greater 
than 2 ?3 4 mg 1-1 (F ,.g . 5-4, Table 5-3). E xcept,.o ns were at nodes 12 and 22, for which 
J"1 k  
ely explanations cannot be identified. Although the relative unimpo11ance of DO at 
higher levels does not discount chronic sublethal effects (see Breitburg et al. 1997), it 
suggests that other factors were more important at the large scale. The most 
successfu l prec t1?c tor of macrobenth1.c  bi.o mass among normox1.c  si.t es was total depth, 
Whereas sites with depth >11 m (N=45l) had a mean biomass of 1.8 g m-2 and 
shallower s ites (N=97 l) had a mean biomass of 6.6 g m-2. Among the shallower sites, 
th0
se in the oligohaline region (salinity::;"8.5) had the highest mean biomass (=27.l g 
-2 
rn 'N==234), while sites in the mesohaline region (8.5<salinity$20.4, N=646) had 
Intermediate biomass (=3.0 g m-2) and polyhaline sites (salinity>20.4) had the lowest 
biomass (==0.1 g m-2, N=91). The appropriate explanation for low biomass in shallow 
Polyhaline sites is unclear, but could be partly due to an inadequate number of samples 
collected (N==26, Table 5-3). 
Among the oligohaline sites, higher percent silt-clay was associated with 
higher biomass. When % silt-clay was >59%, mean biomass was 44.6 g m-2 (N=l53), 
While sandier sites had a mean biomass of only 5.8 g m-2 (N=81). Among these 
sandier sites, nearly freshwater habitats (salinity.$0.37 ppt) had lower biomass than 
Oligohaline habitats with slightly higher salinity. Among the siltier sites, 44 sites with 
Do.::::.7.2 mgr' had much lower biomass (12.4 g m-2) than the 26 sites with D0.$7.2 
rng r' (95. l g m-2) . Variables such as season, water temperature and depth were 
Included in the model but were found to be less effective predictors than DO in this 
' 
171 
context. The coJTect explanation may involve a reversal of cause and effect, whereas 
high biomass causes decreased DO. 
Macrobenthic biomass distributions among shallow(< 11 m) no1moxic (DO> 
2-34 mg r1) mesohaline sites (node 14, N=646, Fig. 5-4) were explained by total 
depth , salinity, percent silt-clay, and water temperature. Biomass was highest at the 
shallowest sites (node 5:.7.65 m, node 15), but the 224 sandy sites (%silt-clay 5:. 3.6) 
2
among these had lower biomass (node T12, 3.2 g m- ) than the sites with higher silt-
clay content (node 17, 15.4 g m-2). Finally, as long as the sediments were not sandy, 
the shallowest mesohaline sites (depth5:.4.6m, node Tl3) had much higher biomass 
2
C31.0 g m-2) than slightly deeper sites (5-llm, 10.4 g m- , node Tl4). Mid-depth 
2
mesohaline habitats (7.6-llm) had lower biomass (1.3 g m- , node 18, N=84) than the 
shallower habitats, especially at higher water temperatures characteristic of late 
summer (0.1 g m-2, node 19). Interestingly, the analysis revealed that at both extremes 
of silt-clay contents (sand, mud), relatively more mixed sediment types were 
associated with higher macrobenthic biomass. 
1
Deep (>11 m) normoxic (DO> 2.34 mg r ) sites (node 20, N=451) generally 
included only mesohaline and polyhaline sites. Percent silt-clay explained explained 
the most variation among these sites. When % silt-clay> 58%, biomass averaged 0.9 
2 
g m- (node 23), while biomass at Jess silty sites averaged 7.8 g m-2_ Among the less 
2
silty sites, biomass was higher among polyhaline (salinity>22 ppt) sites (10.7 g m- ) 
2
than other sites (2.9 g m-2). At siltier sites, average biomass was lower (0.4 g m- ) 
2
du,ing warmer seasons (>21?C) as compared to other times (1.0 g m- ) 
172 
The results of the CART analysis (Fig. 5-4, Table 5-3) strongly implicated 
several habitat factors as important determinants of macrobenthic biomass. The extent 
to which the observed coITelations can be used to infer causation bears some 
consi deration. Nearly all of the habitat factors that were identified as important are 
known to vary among regions of Chesapeake Bay (e.g. salinity , silt-clay content), 
while other factors also known to vary among regions were not included in the 
analysis. For example, water column stratification is often higher in the mid-Bay than 
e lsewhere (e.g. Chapter 2), while physical disturbance during winter st01ms is 
probably more impo1tant in the lower Bay (Schaffner 2000). Net plankton production , 
an indicator of food resources potentially available to the benthos is much higher in 
the south Bay then elsewhere, while the upper Bay receives considerable 
allochthonous carbon inputs (Kemp et al. 1997). Thus, in another CART analysis, 
regional differences were investigated directly by adding latitude to the list of 
variables submitted. 
Regional Characterizations Macrobenthic Biomass, Composition and Bioenergetics 
A CART model including latitude as an explanatory variable confirmed the 
strong regional differences in macrobenthic biomass that have been previously 
reported (e.g. Diaz and Schaffner 1990, Weisburg et al. 1997) and suggested a 
regiona l segmentation scheme for Chesapeake Bay based solely on total macrobenthic 
biomass (Fig. 5-7). Theree main regions of the Bay, the upper Bay, mid Bay and 
lower Bay, were identified. The mid Bay was further divided into north and south 
halves and three depth zones (Fig. 5-7). Several analysis were completed for each of 
173 
the regions of the Bay and are described below. First, the computed species 
composition , biomass and production of the macrobenthos in each region are 
described. This is followed by an examination of regional differences in average 
body size. Finally, computations of the organic carbon requirements of the 
macrobenthos and the supply rate of organic carbon are desc,ibed, addressing the 
potential for food limitation in each region. The discussion is prefaced by a brief 
discussion of the macrobenthic production models that were considered and the 
justification for choosing the model of Edgar (1990). 
Macrobenthic Production Models. Because of the difficulty of measuring the 
production rates of diverse assemblages of macrobenthic organisms, several studies 
have suggested size-based approaches. This study compared the structure and 
predictions of three models, each of which are based on a meta-analysis of published 
production estimates (Banse and Mosher 1980, Edgar 1990, Tumbiolo and Downing 
1994). The Banse and Mosher (1980) mode/ is a univariate model relating annual P/B 
to body size. The Edgar (1990) mode/ predicts daily production as a function of body 
size and wa te r temperature. The T umbiolo and Downing (1994) model predicts 
annu / 
a macrobenth,.c  prod  uct1.0 n J:'.. m depth and annual mean biomass and water i1 0 
tempe,?atu,?e. Compari.s on of  th e se models reveals the primary importance of body size 
d ? declined more rapidly with body size according ao temperature (Fig 5-8). ProductJOn 
t0  h t ? relationship, a result disputed by Peters t e Banse and Mosher (1980) allome nc 
(19 ). 0 ther models are small compared to other 83 The differences between the two 
f ? articularly variations in biomass. 
actors that may affect estimates of ProductJOn, p 
T ? 15 and 25?C and body sizes between his 1?s  especially true for tempera tures between 
174 
? 
10-4 -2 
and 10 g (Fig 5-7), which encompasses much of the Chesapeake Bay 
macrobenthos. Blumenshine and Kemp (2000) utilized the Tumbiolo and Downing 
0 994) model in an analysis of benthic production in the middle Chesapeake Bay, 
citing closer agreement with a study that directly estimated benthic production in the 
mid-Chesapeake Bay (Holland et al. 1988). However, considering the close 
agreement between the two models, the Edgar (1990) model was selected for this 
study on the basis of its more appropriate scope of inference (e.g. seasonal production 
rather than annual mean production). 
Upper Bay Biomass and Species Composition. Summer average benthic 
biomass was highest (59 g m-2) in the upper Bay region, over 90% of which was due 
to the suspension-feeding bivalve, Rangia cuneata. Production in this region was the 
highest in the Bay, 0.49 g m-2 d-1, or -45 g m-2 for the summer. This estimate is 
somewhat higher than suggested for a healthy oligohaline estuarine habitat by 
Weisburg et al. (1997), possibly because it includes only the upper Chesapeake Bay, 
rather than oligohaline habitats throughout the Chesapeake Bay region. For example, 
intense tidally-driven resuspension of sediments causes reduced macrobenthic biomass 
in the ol igohal i ne y ork River (Schaffner et al. 2000). Perhaps for si mi Jar reasons, the 
estimated total summer production in the upper Chesapeake Bay exceeds the annual 
Production estimate of Diaz and Schaffner (1990) for oligohaline habitats in the 
Chesapeake Bay region. 
Mid Bay Biomass and Species Composition. Shallow water habitats at the 
northern end of the mid Bay region had by far the highest summer biomass among mid 
Bay sites, about 12 gm? This biomass was dominated by two suspension-feeding 
175 
bi va lves , Tagelus plebeius and Mya arenaria, plus Macoma balthica, a facu ltati ve 
suspension-feeder. Shallow water habitats further to the south had much lower 
average biomass, 2.8 g m-2. The assemblage there was more diverse, with the most 
abundant tax a being the small opportuni sti c bivalve Mulinia lateral is (20% ). Mid-
depth habitats in the mid-Bay were divided into two regions based on latitude. In the 
north region , nearly all the mid-depth habitat consi sts of a shelf flanking the main 
channel on the western shore between Cove Point and the South River. This region 
had an average biomass of 2.1 g m-2, dominated by M. balthica (68%) and polychaetes 
of the genus Leitoscoloplos. Mid-depth habitats in the lower mid-Bay were located on 
both western and eastern flanks and had very low average biomass (0.64 g m-2). This 
may have been due to the very low silt-clay content of the sediments ( <5 % ). 
However, this habitat was also poorly sampled, with only 5 observations during June-
August. The deep water habitat of the northern mid-Bay also had low average 
summer biomass (0 .62 g m-2) and was dominated by M. lateralis (55%), M. balthica 
(19%), and two polychaetes, Nereis succinea (10%) and Paraprionospio pinnata 
(7%). Biomass exceeded this average early in the summer, prior to the near-inevitable 
onset of complete anoxia by mid summer (Chapter 2) . In contrast, the southern mid-
Bay deep waters had an average biomass of 3.5 g m-2 and included many species more 
typical of the lower Bay, suggesting it is a transition habitat. Macrobenthic biomass in 
the mid-Bay as a who le averaged 2.9 g m-2, approximately in the same range as 
suggested for degraded sites by Weisburg et al. (1997). Macrobenthic production in 
the mid-Bay overall was only 0.04 g m-2 d-1 2 1, or 3.68 g m- summe( ? 
176 
--
Lower Bay Biomass and Production. In the northern portion of the lower Bay, 
macrobenthic biomass computed from the Benthic Monitoring Program data was 
estimated to be 10 g m-2, slightly lower than in the vicinity of the Bay mouth (14.2 g 
2
m- ). However, Thompson (2000) estimated the biomass of the large, tube-dwelling 
polychaete Chaetopterus variopedatus to be -10 g m-2, making it the dominant species 
rn terms of biomas in the lower Bay. Since this taxa accounted for only 3% of lower 
Bay biomass in the Chesapeake Bay Benthic Monitoring Program database, it seems 
likely that the biomass of this polychaete may have been greatly underestimated. Such 
an Underestimate could have resulted from the sampling gears utilized by the 
Chesapeake Bay Benthic Monitoring Program, either due to insufficient penetration 
depth or cross-sectional area (Schaffner, pers. comm.). That possibility is supported 
by the fact that the mean size of C. variopedatus collected by the monitoring program 
Was computed to be 0.068 g, the size of juveniles rather than the much larger and 
deeper-dwelling adults which are 0.2-0.3 g (Thompson 2000). Thus, the computed 
macrobenthic biomass for the lower Bay was amended by 10 g m-2 to account for the 
biomass of this polychaete, giving a total biomass estimate of 22 g m-2 for the lower 
Bay. In contrast to the middle and upper Bay, where bivalves were the largest 
contributors to macrobenthic biomass, polychaetes dominated the benthos in the lower 
Bay. Aside from c. variopedatus, there was a very diverse assemblage of 
Polychaetes, with the head-down deposit feeder Macroclymene zonalis the most 
abundant taxa (see also Schaffner 1990). 
Community production in the lower Bay was estimated to be 0.31 g m-2 d-1? 
This includes a production estimate for Chaetopterus variopedatus of 0.18 g m-2 d-' 
177 
derived from Thompson (2000), which includes production byJ?uveni?le h 
co 011s, net 
somatic growth, reproductive effort by adu lts, and tube formation At a 
? n average 
2
biomass of 10 g m- , P/B for this polychaete is -0.01 8 d- 1. Although this is higher 
than suggested by the Edgar (1990) model for an organism as large as c. 
variopedatus, it reflects a substanti al contribution from juvenile cohorts and tube 
formation, which may not be refl ected well by Edgar (1990). 
Body Size and Production/Biomass Ratios. The average body size of macrobenthos 
varied among the different regions of the Bay (Fig. 5-8, p<0.05). Differences among 
the means were tested using the Tukey-Kramer method due to the unequal numbers of 
observations in each group (Sokal and Rohlf 1995). Average body size was highest in 
the upper Bay where the benthic community was overwhelmingly dominated by the 
bivalve Rangia cuneata. The smallest average body size was found in the deep habitat 
at the north mid-Bay region where the most abundant taxa in summer was the small 
opportunistic biva lve Mulinia lateralis. The mid-Bay shallow water habitat and the 
lower Bay habitats had similar average body size. The south mid-Bay deep water 
habitat had slightly lower body size, but was not statistically distinguishable from 
several other regions (Fig. 5-8). Mean body size was lower in the south mid-Bay mid-
depth habitat than anywhere other than the north mid-Bay deep habitat; however, due 
to a small number of observations in that region , the mean body size was not 
stati sti cally different from any region other than the upper Bay. 
The community production/biomass ratio (hereafter, P/B) was computed for 
Chesapeake Bay benthic communities as the quotient of total community production 
178 
and community biomass. Since production per indi vidual was estimated as a fun cti on 
of Water temperature and body size (eq. 1), PIB ranged from zero (e.g. winter) to 0.032 
th 
d-J (95 percentile), which represents a summer community dominated by very small 
individuals . This range includes the range of P/B as reviewed by Ulanowicz and 
Baird (1986) for a variety of taxa present in the Bay (e.g. Nereis succinea, 
fleteromastus filiformes and other polychaetes, 0.027 d-1; Macoma balthica, 0.016 d-1. 
' 
Rangia cuneata, 0.011 d-1). Average P/B during summer varied among the regions of 
1 
the Bay (Table 5-4), from a low of 0.008 d- in the upper Bay to a high of 0.017 d-1 in 
the south mid-Bay deep water. Although the smallest average body size was found in 
the north mid-Bay deep water, biomass declined to negligible levels in this region by 
mid summer or earlier. Thus, much of the production was limited to early summer 
When water temperature was lower. In contrast, rapid summer growth of 
Chaetopterus variopedatus newly recruited to the lower Bay contributed to higher P!B 
in that region (Thompson 2000). While summer average P/B varied two-fold, summer 
average biomass vaiied among the regions by two orders of magnitude. Despite the 
recognized potential that relatively depauperate benthic communities could have 
higher than expected production due to the predominance of opportunists, these results 
suggested that in Chesapeake Bay production patterns largely reflected biomass 
Patterns. These regional patterns are discussed in more detail below. 
Potential Food Limitation and the Role of the Chesapeake Bay Benthos. 
Bioenergetic budgets were developed for Chesapeake Bay benthic communities dwing 
summer, a major objective being to compare the carbon requirements of regional 
179 
benthic communities with avai lable carbon sources (Tables 5-7). The results are 
presented first by region , then an overall evaluation and summary is provided. 
The plankton community in the upper Bay is net heterotrophic (i.e. respiration 
exceeds photosynthesis) on an annual basis and modest ly autotrophic in summer. 
Estimates of gross primary production vary among published studies but are between 
680 mgC m-2 d-1 (based on Kemp et al. 1997) and 1236 mgC m-2 d-1 based on Harding 
et al. (2001 ) as modified according to suggestions of Smith (2000, see Chapter 5 
Appendix). Summer average plankton respiration is 586 mgC m-2 d-1 based on Kemp 
et al. (1997) and assuming z= 4.5. Thus, net plankton production was estimated to be 
94-650 mgC m-2 d-1 (Table 5-5). The upper Bay receives a daily input of organic 
carbon from the Susquehanna River, but exports a substantial fraction of the input as 
DOC to the mid Bay. The net advective influx of organic carbon, was computed to be 
90 mgC m-2 d-1? Thus, the total carbon potentially available to the benthos (Table 5) 
amounts to 184-740 mgC m-2 d-1, probably sufficient to support the estimated 
respiration of the upper Bay macrobenthos (194-596), but not the consumption (294-
1235). Thus, these estimates suggest that the upper Bay macrobenthos can only be 
sustained through recycling of unrespired (i.e. egesta, ungrazed production) organic 
carbon back into the food web (Table 5-5, see Chapter 5). Such recycling is known to 
be important in Chesapeake Bay (Baird and Ulanowicz 1989). 
An analyses based on the suspension feeding computations described by 
Gen-itsen et a l. (1994) provided an alternative approach to estimating the carbon 
resources availab le to the upper Bay suspension feeding macrobenthos, which in the 
case of the upper Bay accounts for nearly all the biomass (Tables 6 and 7) . According 
180 
to these computations, vertica l mixing in the upper Bay is adequate to bring the entire 
vo lume of the upper Bay within reach of benthic suspension feeders one or more times 
each day (Table 5-6). The combined clearance of suspension feeders was estimated to 
be 11 .5 m3 m-2 d-1, c learing 92 % of the upper Bay volume each day (Table 5-7). 
Given the concentration of particulate organic carbon in the upper Bay, the ration 
consumed would be unrealistically hi gh. Genitsen et al. (1994) approached this 
proble m by assuming that at steady state actual clearance would be limited by the rate 
at which particles could be replaced in the water column. In other words, they 
assumed that the filtered fraction (VF/VT) estimates the fraction of daily primary 
production consumed, rather than the fraction of suspended particles consumed. 
Using this same assumption here, the ration available via suspension feeding is 747 
mgC m-2 d-1, sufficient to support consumption by suspension feeders toward the 
lower end of the estimated range (Table 5-7). By effectively neglecting the possibility 
that suspension-feeders could consume organic detritus, this is a conservative estimate 
of the possible organic matter flux to the suspension-feeding benthos. 
On the other hand, the is the potential that Rangia refiltered water in areas 
where the bivalves were most densely populated. While eq. (9) addresses refiltration 
at the large scale, it does not address the possibility than a single water parcel could be 
filtered several times during a single "approach" to the bottom. At dense populations 
refiltration rates may be as high as 48 % (ORiordan et al. 1995), effectively halving 
the clearance rate of the suspension-feeding population. In this case, the computed 
cleared ration based on primary production alone would be insufficient to meet the 
minimum estimate of organic carbon demand. Thus, it would be required that some 
181 
allochthonous carbon or other suspended particles be included in the filtered ration to 
obtain the ration estimated to be required. Given the results of both the suspension-
feeding and carbon budget computations, it appears likely that the organic carbon 
resources may be just sufficient to support the community, indicating the potential for 
food limitation. 
Comparison of the estimated bioenergetic rates for the upper Bay 
rnacrobenthos with severa l aspects of the carbon budget suggests that consumption 
(and therefore respiration and egestion) by the macrobenthos must be toward the lower 
end of the esti mated range. The range of macrobenthic respiration estimates for the 
upper Bay (Table 5-5) was 88-262% of sediment 0 2 consumption (SOC) estimated by 
rncub atr?n g benthic cores (Cowan and Boynton 1996, Table 5-8). Sm. ee metabolic 
rates for meiobenthos and benthic microbiota are not likely to be negligible (see 
Chapter 6), this implies that either the lower estimates of macrobenthic respiration are 
rnost appropriate, that macrobenthic metaboli sm was under-represented in SOC, or 
both. Although macrobenthic-rich cores were not avoided when making SOC (W. R. 
Boynton , pers. comm.), the relative ra1ity of high biomass sites makes it probable, or 
even like ly, that the macrobenthos were under-represented. As with consumption, the 
respiratory demand of the macrobenthos appears to be a significant component of 
carbon flow in the upper Bay benthos. 
Net plankton production (NPP) in the mid Bay was sufficiently positive during 
su rnrner to support even the most generous esti. mate o f consumpt1.0 n b y t h e 
rnacrobenthic community (Table 5-5, Kemp et al. 1997). Assuming that vertically 
integrated rates of both gross production and plankton respiration per unit volume are 
182 
independent of depth (Kemp et al. 1997), NPP in the shallow water habitats of the mid 
Bay is much greater than the regional average (Table 5-5). Thus, even though 
macrobenthic biomass and production in these areas was greater than in deeper waters, 
NPP exceeds the estimated consumption of the shallow water benthos by a very large 
margin (see also Blumenshine and Kemp 2000) . 
Computations based on the particulate diffusion model of Genitsen et al. 
( 1994) show that horizontal and vertical exchange is sufficient to biing all of the 
littoral zone production and 16% of the pelagic surface mixed layer production to the 
littoral zone benthos of the mid Bay each day. From this perspective, the available 
primary production was estimated to be a minimum of 7-fold greater than the 
computed requirements of the suspension-feeding benthos. Actua1 clearance by 
suspension feeders was estimated to be <5% of the available volume, but was 
sufficient to support the requirements of suspension feeders (Table 5-7). 
Although the available historical data is not extensive, data examined by 
Holland et al. (1987) suggested that stable populations of larger M. arenaria and 
Macoma spp. disappeared after the early 1970's. An abrupt decline in Maryland 
landings of M. arenaria from -3000 tons y{ 1 prior to 1972 to <1000 tons y{ 1 
thereafter suggests that the abundance of this species declined following floods due to 
tropical storm Agnes (June 1972) and never fully recovered (NOAA Landings Data). 
The bioenergetic results suggest that cutTent primary production greatly exceeds the 
requirements of the macrobenthos, while the limited histmical data point to higher 
suspension feeder populations in the past, when phytoplankton biomass (Harding and 
Perry 1997) and presumably production were lower. These results imply that 
183 
reductions in primary production as a result of nutrient loading reductions would 
probably not lead to food limitation of the shallow water benthos. 
Unlike the shallow water zones , net plankton community production reaching 
the benthos at the average depths of the mid-depth and deeper mid Bay habitats was 
estimated to be negative and therefore was not sufficient to directly support the 
metabolic requirements of the benthos (Table 5-5). However, the high rates of aerobic 
and anaerobic metabolism in deep water sediments imply that a large quantity of 
organic matter is transported to these habitats (Cowan and Boynton 1996, Marvin-
DiPasquale and Capone 1998). Lateral imports of carbon from shallower habitats are 
a likely source of carbon for the deeper areas, an explanation also invoked by 
Blumenshine and Kemp (2000) for the mid Chesapeake Bay and by Pearson and 
Rosenburg (1992) to explain a carbon imbalance in the Kattegat. 
Deposition of fresh phytoplankton to mjd Bay sediments at the conclusion of 
the spring bloom is a possible source of high quality organic matter, particularly for 
the early summer benthos. The average deposition of organic matter to mid Bay 
2 
sediments during spring has been estimated to be -27 gC m- (Chapter 4). Distributed 
over the duration of summer, this flux alone is sufficient to support at least a IO-fold 
increase in benthic respiration (Table 5-5), which would bring mid Bay biomass into 
the same range as elsewhere in the Bay (Table 5-4). Since summer benthic microbial 
2 
metabolism in the mid Bay is on the order of 1200 mgC m- d-' (Kemp et al. 1997), it 
is clear that there is a continual flux of organic matter across the pycnocline during 
summer, not just the residual biomass imported from spring. A sediment trap study in 
the mid Bay found that, despite the Jack of excess net plankton production in the 
184 
d eeper waters of the mid Bay, the vertical carbon flux infen-ed from the vertical ch\ -a 
flu x at the pycnoc\ine was about 500 mgC m-2 d-1. The directly measured vertical 
c arbon flux was - lOOO mgC m-2 d-1 (Boynton et a\. l993). The \ower number, which 
m ay represent a more \abi\e fraction of the sedimenting POC, and is \ess \ike\y to 
included resuspended POC, would a\so be more than sufficient to support the 
macrobenthos . The higher number would be required to balance the tota\ benthic 
metabolism (Tab\e 5-8) . 
Like the sha\\ow mid Bay, net plankton community production in the \ower 
Bay was substantia\\y positive. Despite the estimated net advective export of DOC 
onto the continent she\f computed from a box mode\ (Tab\e 5-5), one sti\\ obtains a 
surplus of organic matter production equal to 948 mgC m-2 d-1 (=l l48-200, Tab\e 5-5), 
sufficient to support macrobenthic consumption in the middle of the estimated range 
(Table 5-5). Thus, whi\e the avai\ab\e carbon estimated via NPP was sufficient to 
support the lower Bay benthos, it was not vastly more than sufficient as in the mid 
Bay. 
Suspension-feeding benthos accounted for s\ight\y more than 50% of 
production (Table 5-4). According to computations based on the Gerritsen et a\. 
(l994), physical mixing makes nearly a\\ of the volume of the lower Bay avai\ab\e to 
the benthos each day, enabling suspension-feeders to have a substantial grazing 
impact. The lower Bay suspension-feeders were estimated to filter about half of the 
volume lower Bay volume each day (Table 5-7). Assuming that this means that they 
could consume the same fraction of daily primary production (see discussion for upper 
Bay), the ration obtained by suspension feeding could be 756 and 680 mg m-2 d-1 in 
l85 
the north and south divi sions of the lower Bay, respectively. These values are near the 
maximum that the suspension-feeding benthos in the lower Bay may require (Table 5-
7), indicating food limitation is not likely. 
Summary and Conclusions 
Previous studies and reviews of benthic ecology have shown that the biomass 
and production of estuarine and marine macrobenthos can depend on many interacting 
factors , hypoxia being among them. Although benthic biomass and production within 
estuaries is known to be highly variable, this study developed empirical models that 
resolved the complex dependence of the Chesapeake Bay benthos on multiple habitat 
factors, providing a validated predictive capability and identifying which among the 
factors that were examined were the most predictive in which environmental context. 
Among these, hypoxia was found to be the most important factor affecting the 
biomass of the macrobenthos. Consistent with the findings of Diaz and Rosenburg 
(1995), benthic biomass was found to be reduced when dissolved oxygen (DO) was 
less than -2 mg r1, with even lower biomass associated with lower DO (Fig 4 , Table 
5-2). However, depth and interactions between depth and salinity appeared as 
predictors of benthic biomass in a way that suggested hidden effects of periodic or 
episodic hypoxia in habitats that were normoxic at the time they were sampled. Most 
obvious was the depression of macrobenthic biomass at all depths in the mesohaline 
Chesapeake Bay relative to the upper and lower Bay. These patterns are put into a 
comparative context in Fig. 5-9, which reproduces the relationship between 
macrobenthic biomass and annual phytoplankton production reported by Herman et al. 
186 
(1999). Estimates from this study were added, showing that macrobenthic biomass in 
the lower Bay was consistent with expectations based on primary production. In 
contrast, benthic biomass was greater than expected in the upper Bay, perhaps due to 
either allochthonous organic matter sources or to reduced predation on the dominant 
clam, Rangia cuneata. Finally, the mid Bay, despite very high primary production, 
had a very depauperate macrobenthos compared to coastal systems with similar 
primary production. Based on this relationship, the macrobenthic biomass in the mid 
Chesapeake Bay could be expected to be as high as 40 gC m-2, a 14-fold increase over 
the present average. A IO-fold increase would still be reasonable even if primary 
production was reduced significantly. 
Aside from the possibility that episodic hypoxia is to blame, the reason for the 
depression of macrobenthic biomass and production in shallower habitats of the mid 
Bay was not resolved and would be a good area for further investigation. One 
hypothesis is that demersal predators may be concentrated into the areas where 
adequate benthos is present and thereby exert an exceptionally strong effect. This 
could be exacerbated by low oxygen events, which force burrowed animals to the 
surface where they are more vulnerable to predation (Diaz et al. 1992). 
The results of this study suggest that the primary effect on the benthos of a 
reduction in organic matter inputs, as might be achieved through nut1ient loading rate 
reductions, would be to benefit the community by reducing oxygen stress. This is 
particularly evident in the mid Bay. In contrast, reduced organic matter inputs to the 
benthos would most likely not limit benthic production, since these populations do not 
appear strongly food limited, if food limited at all. Moreover, the models of 
187 
suspension feeding indicate that the Chesapeake Bay benthos, except for in deeper 
waters of the mid Bay, may not rely on surplus organic matter from the plankton (i.e. 
positive net plankton production). Rather they obtain their ration from net 
phytoplankton production, not net plankton production, as needed, possibly to the 
detriment of the plankton. In this way, one may hypothesize that net plankton 
production results in part from benthic suspension-feeding and that, lacking significant 
suspension-feeding, the plankton will tend toward P/R=l. Although this argument 
would not apply directly to deposit-feeders, suspension-feeding can contribute to 
deposit feeding via biodeposition, extending this coupling mechanism to all the 
macrobenthic taxa. Considering the massive, but now absent, biomass of oysters that 
once filtered the waters of Chesapeake Bay (e.g. Newell 1988), it seems likely that 
such coupling was once an important aspect of the ecology of Chesapeake Bay. One 
may conclude that a management policy that avoids reducing primary production in 
order to maintain secondary production in the benthos would be misguided. Such a 
policy would neglect the demonstrable effects of habitat quality effects in favor of a 
hypothetical and, by this analysis, improbable food limitation effect, and would not be 
consistent with inferences that can be based on hist01ical precedent and comparative 
ecology. 
Because patterns in benthic biomass appeared to reflect known seasonal and 
spatial patterns in the frequency and intensity of hypoxia, benthic monitoring is 
supported as a means of assessing progress toward ecosystem restoration. Two 
regions in the mid Bay would be particularly valuable places to target benthic 
monitoring effort in order to develop indicators of ecosystem change in Chesapeake 
188 
Bay. Benthic communities in the mid depth zone (7.5-lOm) in the mid Bay may be an 
effective integrator of periodic oxygen stress which results from upwelling of hypoxic 
or anox.ic water from the channel. Similarly, variability in the ex.tent of hypoxia is 
often manifest most dramatically as changes in the southern limit of hypoxia in the 
Bay, which changes both interannual\y and on the biweekly time scale resolved by the 
Chesapeake Bay Monitoring Program (Chapter 2). Therefore, the benthic community 
in deep waters at the southern limit of the mid Bay may also be a useful and sensitive 
indicator that could be targeted by monitoring efforts seeking to assess the progress of 
restoration. 
Literature Cited 
Baird, D. and R. E. Ulanowicz. 1989. The seasonal dynamics of the Chesapeake Bay 
Ecosystem. Ecol. Monogr. 59(4): 329-364. 
Banse, K. 1979. On weight dependence of net growth efficienct and specific 
respiration rates among field populations of invertebrates. Oecologia. 38: 
111-126. 
Banse, K. and S. Mosher. 1980. Adult body mass and annual production/biomass 
relationships of field populations. Ecol. Monogr. 50(3): 355-379. 
Bayne, B. L. and R. C. Newel\ . 1983. Physiological energetics of marine molluscs. 
pp. 407-515. In: Saleuddin , A. S. M. and K. L. Wilbur, eds. The Mollusca, 
Volume 4. Physiology, Part l. New York: Academic Press. 
Blumenshine, S. and W . M. Kemp. 2000. Food versus habitat limiation for benthic 
macrofauna in mesohaline regions of Chesapeake Bay . In: Kemp, W. M., R . 
Bartleson, S. Blumenshine, J. D . Hagy and W . R. Boynton. Ecosystem models 
of the Chesapeake Bay relating nutrient loadings, environmental conditions, 
and living resources. US EPA Chesapeake Bay Program, Annapolis, MD. 
189 
Boynton, W.R., W.M. Kemp, J.M. Barnes, LL. Matteson, F.M. Rohland, D.A. 
Jasinski and H.L. Kimble. 1993. Ecosystem Processes Component Level l 
Int~rpretive Report No. 10. Chesapeake Biological Laboratory (CBL), 
University of Maryland System, Solomons, MD 20688-0038. Ref. 
No.[UMCEESJCBL 93-030a. 
Boynton, W.R., personal communication. UMCES Chesapeake Biological 
Laboratory, Solomons, MD. 
Breiman, L., J. Friedman, R. Olshen, and C. Stone. 1984. Classification and 
Regression Trees. Pacific Grove: Wadsworth. 
Breitburg, D. L. 1990. Nearshore hypoxia in the Chesapeake Bay: Patterns and 
relationships among physical factors. Est. Coast. Shelf Sci. 30(6): 593-609. 
Breitburg, D. L., T. Loher, C. A. Pacey, and A. Gerstein. 1997. Varying effects of 
low dissolved oxygen on trophic interactions in an estuarine food web. Ecol. 
Monogr. 67(4): 489-507. 
Budd, J. D., T. D. Drummer, T. F. Nalepa, and G. L. Fhnenstiel. 2001. Remote 
sensing of biotic effects: Zebra mussels (Dreissena polymorpha) influence on 
water clarity in Saginaw Bay, Lake Huron. Limnol. Oceanogr. 46(2): 213-
223. 
Cloern, J.E. 1982. Does the benthos control phytoplankton biomas in south San 
Francisco Bay (USA)? Mar. Ecol. Prag. Ser. 9: 191-202. 
Constable, A. J. 1999. Ecology of benthic macro-invertebrates in soft-sediment 
environments: A review of progress towards quantitative models and 
predictions. Aust. J. Ecol. 24(4): 452-476. 
Cowan, J. L. and w. R. Boynton. 1996. Sediment-water oxygen and nutrient 
exchanges along the longitudinal axis of Chesapeake Bay: Seasonal patterns, 
controlling factors and ecological significance. Estuaries 19(3):562-580. 
Diaz, R. J. and L. c. Schaffner. 1990. The functional role of the estuarine benthos. 
pp. 25-56 In: in Haire, M. and E. C. Krome: (eds.). Perspectives on the 
Chesapeake Bay: Advances in Estuarine Science. Chesapeake Research 
Consortium, Solomons, MD. 
Diaz, R. J., R. J. Neubauer, L. C. Schaffner, L. Pihl, and S. P. Baden. 1992. 
Continuous monitoring of dissolved oxygen in an estuary experiencing 
periodic hypoxia and the effect of hypoxia on macrobenthos and fish. Sci. Tot. 
Env. (Supplement 1992): 1055-1068 
190 
Diaz, R. J. and R. Rosenberg. 1995. Marine benthic hypoxia: A review of its 
ecological effects and the behavioral responses of benthic macrofaun a. 
Oceanogr. Mar. Biol. Ann. Rev. 33 : 245-303. 
Edgar, G. 1990. The use of the size structure of benthic macrofaunal communities to 
estimate fauna! biomass and secondary production. J. Exp. Mar. Bio. Ecol. 
137: 195-214. 
EPA 2000a. The 2000 User 's Guide to Chesapeake Bay Program Biological and 
Living Resources Monitoring Data. US Environmental Protection Agency, 
Chesapeake Bay Program, Annapolis, MD. 
EPA 2000b. Maryland Chesapeake Bay Benthic Monitoring Data Dictionary. US 
Environmental Protection Agency, Chesapeake Bay Program, Annapolis, MD. 
EPA 2000c. Virginia Chesapeake Bay Benthic Monito1ing Data Dictionary. US 
Environmental Protection Agency, Chesapeake Bay Program, Annapolis, MD. 
Gerritsen, J., A. F. Holland, and D . E. Irvine. Suspension-feeding bivalvles and the 
fate of primary production: An estuarine model applied to Chesapeake Bay. 
Estuaries. 17(2): 403-416. 
Graf, G. , W. Bengtsson , U. Diesner, R. Schultz, and H. Theede. 1982. Benthic 
response to sedimentation of a spring phytoplankton bloom: process and 
budget. Mar. Biol. 67: 201-208. 
Hagy, J. D. 1996. Residence times and net ecosystem processes in Patuxent River 
estuary, MD. M. S. Thesis. University of Maryland at College Park. College 
Park, MD. 
Hagy, J. D . and w. R. Boynton. 2000. Controls on hypoxia in Chesapeake Bay and its 
major tributaries. In: Kemp, W. M., R. Bartleson, S. Blumenshine, J. D. Hagy 
and W. R. Boynton. Ecosystem models of the Chesapeake Bay relating 
nut1ient loadings, environmental conditions, and living resources. US EPA 
Chesapeake Bay Program, Annapolis, MD. 
Hagy, J. D. 2000. Tree-structured models of water quality and submersed aquatic 
vegetation in Chesapeake Bay. In: Kemp, W. M., R. Bartleson, S. 
Blumenshine, J. D. Hagy, and W. R. Boynton. Ecosystem models of the 
Chesapeake Bay relating Nutrient Loadings, Environmental Conditions and 
Living Resources. US EPA Chespeake Bay Program, Annapolis, MD. 
Hansen, B. w. 1999. Cohort growth of planktotrophic polychaete larvae - are they 
food limi ted? Mar. Ecol. Prog. Ser. 178: 109-119. 
191 
Hanson , J. M. and R.H. Peters. 1983. Empirical prediction of crustacean 
zooplankton biomass and profundal macrobenthos biomass in lakes. Can. J. 
Fish. Aquat. Sci. 41:439-445. 
Hardin~, L. W. and E. S. Perry. 1997. Long-term increase of phytoplankton biomass 
m Chesapeake Bay, 1950-1994. Mar. Ecol. Prog. Ser. 157: 39-52. 
Harding, L. W., Jr. , M. E. Mallonee, and E. S. Pen-y. 2001. Toward a predictive 
understanding of primary productivity in a temperate, pa11ially stratified 
estuary. Est. Coast. Shelf Sci. in press. 
Hartman K. J. , and S. B. Brandt. 1995. Trophic resource partitioning, diets, and 
growth of sympatric estuarine predators. Trans. Am. Fish. Soc. 124 (4): 520-
537. 
Herman, P. M . J. , J. J. Middelburg, J. Van De Koppel and C.H. R. Heip. 1999. 
Ecology of estuarine macrobenthos. Adv. Ecol. Res .. 15(2). 
Hobbs, C.H., III, J.P. Halka, R. T. Kerhin, and m. J. Carron. 1992. Chesapeake Bay 
Sediment Budget. J. Coast. Res. 8(2): 292-300. 
Holland, A. F., N. K. Mountford and J. A. Mihursky. 1977. Temporal variation in 
upper Bay mesohaline benthic communities: I. The 9-m mud habitat. Ches. 
Sci. 18(4): 370-378. 
Holland, A. F. , A. T. Shaughnessy, and M. H. Riegel. 1987. Long-term variation in 
mesohaline Chesapeake Bay macrobenthos: Spatial and temporal patterns. 
Estuaries 10(3):227-245. 
Holland, A. F., A. T. Shaughnessy, L. C. Scott, V. A. Dickens, J. Gerritsen, and J. A. 
Ranasinghe. 1988. Long-term benthic monitoring for the Maryland portion of 
Chesapeake Bay. Rep. to MD OEP, Baltimore, Maryland. Versar, Inc. 
Columbia, Maryland. 
Hollis, E. H. 1952. Variations in the feeding habits of the st1iped bass, Roccus 
saxatilis (Walbaum), in Chesapeake Bay. Bull. Bing. Oceanogr. Col., Yale 
Univ. 14(1):111-131. 
l0rgensen, s. E. 1979. Handbook of Environmental Data and Ecological Parameters. 
International Society for Ecological Modelling. Copenhagen. 
losefson, A. B. and D . J. Conley. 1997. Benthic response to a pelagic front. Mar. 
Ecol. Prog. Ser. . 147: 49-62. 
Karakassis, I, E . Hatziyanni, M. Tsapakis, and w_. Plaiti. 1999. Benthic recovery 
following cessation of fish farming: a sen es of successes and catastrophes. 
Mar. Ecol. Prog. Ser. 184: 205-218. 
192 
Kemp, W. M. and W.R. Boynton. 1992. Benthic-Pelagic Coupling. In : Smith, D. E. , 
M. Leffler, and G. Mackiernan (eds.). Oxygen Dynamics in the Chesapeake 
Bay. A synthesis of recent research. Maryland Sea Grant College, College 
Park, MD 
Kemp, W. M., E. M. Smith , M. Marvin-DiPasquale, and W. R. Boynton . 1997. 
Organic carbon balance and net ecosystem metabolism in Chesapeake Bay. 
Mar. Ecol. Prog. Ser. 150: 229-248. 
Kitazato H, Y. Shirayama, T. Nakatsuka, S. Fujiwara, M. Shimanaga, Y. Kato, Y. 
Okada, J . Kanda , A. Yamaoka, T. Masuzawa, and K. Suzuki. 2000. Seasonal 
phytodetritus deposition and responses of bathyal benthic foraminiferal 
populations in Sagami Bay, Japan: preliminary results from "Project Sagami 
1996-1999." Mar. Micropaleontol. 40 (3): 135-149. 
Langland, M.J. , R. E. Edwards, L.A. Sprague and S. E. Yochum. 2001. Summary of 
Trends and Status Analysis for Flow, Nut1ients, and Sediments at Selected 
Nontidal Sites, Chesapeake Bay Basin, 1985-99: U.S. Geological Survey Open 
File Report 01 -73 , 49 p. 
Lardicci , C. , S. Como, S. Corti , and F. Rossi. 2001. Recovery of the 
macrozoobenthic community after severe dystrophic crises in a Meditenanean 
coasta l lagoon (Orbetello, Italy) . Mar. Poll. Bull. 42(3): 202-214. 
Malone, T. C., W. M. Kemp, H. W. Ducklow, W.R. Boynton, J. H. Tuttle and R. B. 
Jonas . 1986. Lateral variation in the production and fate of phytoplankton in a 
partially stratified estuary. Mar. Ecol. Prog. Ser. 32(2-3): 149-160. 
Marsh, A.G. and K. R. Tenore. 1990. The role of nutrient in regulating the 
population dynamics of opportunistic, surface deposit feeders in a mesohaline 
community. Limnol. Oceanogr. 35(3): 710-724. 
Marvin-Dipasquale, M. C. and D. G. Capone. 1998. Benthic sulfate reduction along 
the Chesapeake Bay central channel. I. Spatial trends and controls . Mar. 
Ecol. Prog. Ser. 168: 213-228 . 
Newell , R. I.E. 1988. Ecological changes in the Chesapeake Bay: Are they the 
result of overharvesting the American oyster, Crassostrea virginia. pp. 536-
546 In: M. P. Lynch and E. C. Krome (eds.). Understanding the Estuary: 
Advances in Chesapeake Bay Research. Proceedings of a Conference 29-31 
March 1988. Baltimore, MD. Chesapeake Research Consortium, Publication 
Number 129. 
NOAA Landings Data. National Oceanographic and Atmorphe1ic Administration, 
National Marine Fisheries Service Landings Data. Fisheries Statistics and 
Economics Division, Silver Spring, MD. (data distributed via Internet site). 
193 
Newell, R. I.E. 1988. Ecological changes in Chesapeake Bay: Are they the result of 
overharvesting the American oyster, Crassostrea virginica? pp. 536-546. In : 
M. C. Lynch and E. C. Krome, eds. Understanding the Estuary: Advances in 
Chesapeake Bay Research. Chesapeake Research Consortium. Solomons, 
MD. 
Pearson, T. H. and R. Rosenburg. 1978. Macrobenthic succession in relation to 
organic eniichment and pollution of the ma1ine environment. Oceanogr. Mar. 
Biol. Ann. Rev. 16: 229-311. 
Peters, R.H. 1983. The ecological implications of body size. New York: Camb1idge 
University Press. 
Pearson, T. H. and R. Rosenberg. 1992. Energy flow through the SE Kattegat: A 
comparative examination of the eutrophication of a coastal marine ecosystem. 
Neth. J. Sea Res. 28(4): 317-334. 
Pfitzenmeyer, H. T. and K. G. Drobeck. 1963. Benthic survey for populations of soft-
shelled clams, Mya arenaria in the lower Potomac River, Maryland. Ches. Sci . 
4(2): 67-74. 
Phelps. H. L. 1994. The Asiantic clam (Corbiculafluminea) invasion and system 
level ecological change in the Potomac river estuary near Washington, DC. 
Estuaries 17(3): 614-621. 
Pihl , L., S. P. Baden, and R. J. Diaz. 1991. Effects of petiodic hypoxia on 
distribution of demersal fish and crustaceans. Mar. Biol. 108: 349-360. 
Schaffner, L. C. personal communication. Virginia Institute of Marine Science, 
Gloucester Point, VA. 
Schaffner, L. C. , T. M . Dellapenna, E. K. Hinchey, C. T. Friedrichs, M. Thomson-
Neubauer, M. E . Smith, and S. A. Kuehl. 2000. Physical energy regimes, 
seabed dynamics, and organism-sediment interactions along an estuarine 
gradient. In: Aller, J. Y., S. A. Woodin, and R. C. Aller. (eds.) Organism-
Sediment Interactions. Universiity of South Carolina Press, Columbia, SC. 
Schroeder, L.A. 1981. Consumer growth efficiencies: their limits and relationships 
to ecological energetics. J. Theor. Biol. 93: 805-828. 
Schubel, J. R. and H. H. Carter. 1977. Suspended sediment budget for Chesapeake 
Bay. In: M. Wiley, ed. Estuarine Processes, Vol II. Circulation, Sediments, 
and Transfer of Material in the Estuary. Academic Press, New York. 
Schwinghamer, P., B. Hargrave, D. Peer and C. M. Hawkins. Partitioning of 
production and respiration among size groups of organisms in an intertidal 
benthic community. Mar. Ecol. Prog. Ser. 31: 131-142. 
194 
Steinburg, D. and P. Colla. 1995. CART: Tree-Structured Non-Parametric Data 
Analysis. San Diego, CA. Salford Systems. 
Stokes, C. 1996. Influences of environmental variables on photosynthetic and 
respiratory quotients in Chesapeake Bay. M.S. Thesis . University of 
Maryland at College Park. College Park, MD. 
Smith , E .M. 2000. Factors regulating respiration and its coupling to primary 
production in a coastal planktonic community. Ph.D. Dissertation. University 
of Maryland at College Park. College Park, MD. 
Sokal , R. R. and F. J. Rohlf. 1995. Biometry. Third Edition . W. H. Freeman and 
Company. New York. 
Thompson , M. L. 2000. Demographics, production and benthic-pelagic coupling by 
the suspension-feeding polychaete Chaetopterus pergamentaceus in the lower 
Chesapeake Bay. Ph.D. Disse11ation. Virginia Institute of Marine Science, 
Gloucester Point, VA. 
Tumbiolo, M. L. and J. A. Downing. 1994. An empirical model for the prediction of 
secondary production in marine benthic inve,tebrate populations. Mar. Ecol. 
Prog. Ser. 114: 165-174. 
Ulanowicz, R. E. and D. Baird. 1986. A network analysis of the Chesapeake Bay 
Ecosystem. MD Department of Natural Resources, Tidewater Administrator. 
Annapolis, MD. Ref#: UMCES[CBL] 86-79. 
Weisburg, s. B., J. A. Ranasinghe, L. C. Schaffner, R. J. Diaz, D. M. Dauer, and J.B. 
Frithsen. 1997. An estuarine benthic index of biotic integ1ity (B-IBI) for 
Chesapeake Bay. Estuaries. 20(1): 149-158. 
Zajac, R. N. and R. B. Whitlatch . 2001. Response of macrobenthic communities to 
restoration efforts in a New England estuary. Estuaries. 24(2): 167-183. 
195 
Table 5-1. Parameter estimates and F-table for a multiple regression (r2=0.34) 
predicting log ash-free biomass per m2 (recoded by adding 0.01 g m-2) at locations in 
the mainstem Chesapeake Bay. Explanatory va1iables include bottom water dissolved 
oxygen (DISOXY), bottom water salinity (SALINITY), bottom water water 
temperature (WTEMP), % sediment silt-clay content (SILTCLA Y), and all two-way 
interactions. A categorical variable representing seasons (Winter=December-February, 
Spring=March-May, Summer=June-August, Fall=September-November) was also 
included in the model. n. s. = not statisticall y significant (p>0.10). 
Source Estimate DF T~ee III SS Si ~nificance 
DISOXY -0.2020 l 17.02 <0.01 
SALINITY 0.0301 l 0.84 n.s. 
WTEMP -0.1056 l 20.48 <0.01 
SILTCLAY -0.0022 l 0.20 n.s. 
DISOXY*SALINITY 0.0016 l 0.65 n.s. 
DISOXY*WTEMP 0.0107 1 66.73 <0.01 
DISOXY*SILTCLA Y 0.0013 l 16.41 <0.01 
SALINITY*WTEMP 0.0002 1 0.06 n.s. 
SALINITY*SILTCLA Y -0.0011 l 68.14 <0.01 
WTEMP*SILTCLA Y 0.0005 l 11.03 <0.01 
SEASON (W, Sp, Su, F) 0.28, 0.24, 3 21.34 <0.01 
0.30, 0.00 
Intercept 1.5977 <0.01 
Model (r2=0.34) 13 567.63 <0.01 
Error 1571 1060.06 
Con-ected Total 1584 1607.69 
196 
Table 5-2. The importance of water quality and habitat variables as factors in a CART 
regression-tree model predicting biomass of mainstem Chesapeake Bay macrobenthic 
communities. Relative importance scores are included in the CART software output 
and reflect the total contribution of the variable to model predictions. The variable 
With the highest importance score is arbitra1ily assigned a value of 100. When 
importance statistics are based on primary splitting rules only, other indicators of 
importance, such as appearance as a surrogate are discounted (Steinburg and Colla 
1995). 
Relative Impo11ance 
_ Variable Name Relative Importance (primary splitting rules only) 
Dissolved Oxygen 100 100 
Salinity 71 49 
Percent Silt-Clay 63 48 
Total Depth 58 54 
Month 39 24 
Water Temperature 38 15 
Percent Carbon 8 0 
__Eercent Nitrogen 5 0 
197 
Table 5-3. Node detail for the regression-tree model shown in Fig. 5-4. DO=bottom 
dissolved oxygen (mgr'), MONTH=month of the year (1-12), TDEPTH=total depth 
(m), SALINITY=bottom salinity (ppt), SILTCLA Y=% silt-clay content of sediments. 
For ~on-terminal nodes, an observation is directed to the left child node if the splitting 
rule 1s true. Log Mean biomass refers to the ~ean of macro?enthic biomass2 after log 
transformation by log(biomass+0.01) where biomass has umts g AFDW m- . Mean 
b"1 omass was computed by back-transformi.n g usi.n g B = 10 ~~ - 0.02 , where 
2 
B=mean biomass, B' =log mean biomass, and a is the variance of the log mean 
observations. The term a 2 /2 corrects for the bias due to log-transformation. 
Std. Mean N Child 
Node Splitting Rule Log 
Mean Dev. Biomass 
Nodes 
(g m-2) (Left, 
Biomass 
Ri~ht) 
1.017 4.1 1664 2,6 
1 0.095 D0<2.34 1.008 0.3 242 Tl, 3 
2 MONTH:s;6.5 -0.955 0.864 0.1 167 4,5 
3 -1.351 TDEPTH:s; 10.4 
-0.511 1.049 
1.1 25 T2, T3 
4 oo:s;1.1 
-1.499 0.734 
0.0 142 T4,T5 
5 oo:s;o.875 1.5 75 
Tl -0.075 
0.703 
0.1 8 
T2 -1.291 
0.917 
-0.145 0.895 
1.8 17 
T3 0.0 92 
T4 -1.659 
0.60 
0.857 0.1 50 
TS -1.204 0.904 4.8 1422 7,20 0.273 
6 TDEPTH:s;l 1.05 6.6 971 8, 13 
7 0.470 
0.835 
SALINITY:s;8.545 0.998 27.l 234 9, 11 0.936 
8 SILTCLA y:s;58.78 0.912 5.8 81 10, T8 0.351 
9 SALINITY~0.370 
-0.092 0.958 
2.3 31 T6, T7 
10 TDEPTH:s;2.95 0.898 44.6 153 12, Tl l 1.246 
11 SILTCLA y:s;89.95 34.5 70 T9, TlO 
1.000 1.037 
12 DO:s;7.l 7 0.2 11 
-0.778 0.515 
T6 5.3 20 
0.285 0.936 
T7 
0.625 0.762 
8.2 50 
T8 0.844 95.1 26 1.622 
T9 0.961 12.4 44 0.632 
TlO 49.8 83 
1.454 0.698 
TU 
198 
Table 5-3. Continued. 
Std. Mean N Child 
Node Splitting Rule Log 
Mean Dev. Biomass 
Nodes 
(g m-2) (Left, 
Biomass 
Ri~ht) 
0.716 3.8 737 14, Tl9 
13 SAL1NITY~20.4 0.322 
0.238 0.705 3.0 
646 15, 18 
14 TDEPTfr:;7.65 0.665 3.3 562 16, Tl5 
15 SAL1NITY~l4.025 0.303 
0.483 0.691 5.2 
309 Tl2, 17 
16 SILTCLA Y~3.622 
0.884 0.780 15.4 85 
Tl3, Tl4 
17 TDEPTH~4.65 0.807 1.3 84 Tl6, 19 
18 WTEMP~24.880 -0.193 
-0.825 0.900 0.4 
26 Tl 7, Tl8 
19 SILTCLAY~86.68 0.585 3.2 224 
Tl2 0.330 0.585 31.0 24 
Tl3 1.321 
0.712 0.780 10.4 
61 
Tl4 
0.083 0.560 1.7 
253 
Tl5 
0.090 0.567 1.8 
58 
Tl6 14 
Tl7 -0.156 
0.645 1.1 
0.371 0.0 12 
Tl8 -1.606 91 
Tl9 -0.916 
0.462 0.1 
0.900 1.8 451 21,23 
20 SILTCLAY~58.44 -0.150 103 22, T22 
0.644 0.706 7.8 
21 SALINITY~21.95 0.788 2.9 38 T20, T21 0.158 
22 D0~4.385 
-0.385 0.813 
0.9 348 24,26 
23 WTEMP~20.60 1.0 248 T23 , 25 
-0.181 0.612 
24 TDEPTH~22.85 
-0.529 0.634 0.4 
69 T24, T25 
25 TDEPTH~31.5 100 T26, T27 
-0.891 1.005 0.4 
26 SALINITY~l8.80 0.869 0.5 8 -0.674 
T20 0.379 0.594 3.6 
30 
T21 10.7 65 0.928 0.456 
T22 1.2 179 
-0.047 0.547 
T23 
-0.787 0.517 
0.2 45 
T24 1.2 24 -0.045 0.543 
T25 0.1 58 -1.323 0.729 
T26 -0.295 1.029 
1.7 42 
_T27 
199 
Table 5-4. Regional area and summer (June-August) average biomass, daily 
production and daily P/B ratio, as estimated using the Edgar (1990) model. Means for 
combined regions (e.g. upper Bay, mid Bay, lower Bay, whole Bay) are weighted by 
the area of component regions. Biomass and production for the lower Bay include 
estimates for the pol ychaete Chaetopterus variopedatus, which were derived 
separately from Thompson (2000). 
Biomass Production % Susp 
Region ( gm -2) (g rn-2 d-1) Feeding 
59.22 0.49 92 0.008 
Upper Bay 353 11.72 0.13 35 0.011 
N Mid Bay, <7.6 rn 303 2.77 0.03 29 0.012 
S Mid Bay, <7 .6 rn 890 2.07 0.03 8 0.013 
N Mid Bay, 7.6-10 m 172 
408 0.64 0.01 
10 0.017 
S Mid Bay, 7.6-10 rn 0.63 0.01 50 0.014 
North Mid Bay, >10 rn 783 3.51 0.04 17 0.012 
South Mid Bay, >10 m 532 20.05 0.30 67 0.015 
N Lower Bay 1009 24.18 0.33 60 0.013 
S Lower Bay 1011 59.2 0.49 92 0.008 
Upper Bay 353 2.9 0.04 29 0.012 
Mid Bay 3089 22.l 0.31 63 0.014 
Lower Bay 2019 0.17 46 0.012 
Whole Bay 5461 
13.7 
200 
Table 5-5. Estimated ranges for summer (June-August) average consumption (C), 
respiration (R), and excretion plus egestion (U) by macrobenthic communities in 
Chesapeake Bay, computing using net growth efficiency=0.3 or 0.56 and 
assimilation efficiency=0.4 or 0.6, resulting in gross growth efficiency=0.12 or 
0.34. Bioenergetic rates are compared to summer (June-August) average net 
plankton production , recomputed for each region using data from Kemp et al. 
(1997), and the net external TOC exchange, computed using physical transport 
estimates computed using a box model (Chapter 2). Net input of TOC is available 
only for regions as a whole. n.e. = not estimated. All units are mgC m-2 d-1? 
Region C R u NPP Ext. 
Src. 
Upper Bay 735-2058 194-576 294-1235 94-650 90 
North Mid Bay - Shallow 193-539 51-151 77-324 2279 n.e 
South Mid Bay - Shallow 51-142 13-40 20-85 2298 n.e 
North Mid Bay - Middle 39-110 10-31 16-66 -91 n.e 
South Mid Bay - Middle 16-45 4-13 6-27 -50 n.e 
North Mid Bay - Deep 14-38 4-11 5-23 -532 n.e 
South Mid Bay - Deep 63-178 17-50 25-107 -484 n.e 
North Lower Bay 446-1249 118-350 178-750 907 n.e 
South Lower Ba}'. 485-1359 128-380 194-815 1388 n.e 
Upper Bay 735-2058 194-576 294-1235 94-650 90 
Mid Bay 52-146 14-41 21-88 656 -10 
Lower Bay 466-1304 123-365 186-782 1148 -200 
TOTAL 249-698 66-195 100-419 801 -80 
201 
Table 5-6. The parameters desc1ibing mixing characte1istics of 5 segments 
of Chesapeake Bay, leading to estimates of the fraction of the water 
volume available to suspension feeders each day. tv =the mean time for a 
water parcel to encounter the benthos, Pv=the probability that a water 
parce l in the littora l zone wi ll encounter the bottom within l day. 
Region Mean Depth tv Pv Fraction 
(m) (hours) Available 
Upper Bay 4.50 1.9 1.00 1.00 
North Mid Bay 5.69 2.9 1.00 0.43 1 
South Mid Bay 5.33 2.6 1.00 0.51 1 
North Lower Bay 9.90 9.1 0.93 0.93 
South Lower Bay 8.19 6.2 0.98 0.98 
1 refers to the fraction of li ttoral volume plus pelagic SML vo lume 
avai lab le, computed using eqs. 4, 5, 6 and 7 from GeITitsen et al. 1994. 
202 
Table 5-7. Computations based on the models of GeITitsen et al (1994) 
showing the relationship between potential ratio available via suspension 
feeding and the range of estimated carbon requirements of the suspension 
feeding macrobenthos. V8=volume of water approaching the benthos, 
V1~volume of water in the littoral zone and adjacent pelagic surface mixed 
layer. VF=volume filtered by suspension feeders . PP=Summer Net 14C-
Primary Production (Harding et al. 2001). C demand differs from the values in 
Table 5-5 because these values are for suspension feeders only. 
Region Clearance Vs/VT VF/VT pp Filt. Ration C Demand 
m3 m- 2 d-1 mg C m- 2d-1 mg C m- 2 d-1 mgC m-2 d-1 
Upper Bay 11.46 100% 92% 811 747 667-1889 
N Mid Bay 1.02 36% 4% 2300 226 101-287 
S Mid Bay 0.25 48% 1% 1945 71 20-57 
N Lower Bay 6.78 93% 48% 1571 756 295-835 
S Lower Ba:}'. 6.65 98% 55% 1240 680 288-816 
203 
Table 5-8. Average total benthic metabolic rates for three regions of 
Chesapeake Bay during June-August. Sediment oxygen consumption 
(SOC) estimates were obtained from Cowan and Boynton (1996). Benthic 
sulfate reduction (SR) estimates were obtained from Marvin-DiPasquale 
and Capone (1998). Total benthic respiration (Rb) was computed as SR 
plus half of SOC (after Kemp et al. 1997), which reflects the likelyhood 
that some SOC resulted from chemical oxygen demand due to SR. SOC 
Was converted to C assuming RQ=l. SR was converted to C by assuming 
S:Oi=:2 and RQ=l. 
- soc Total Rb (g 0 2 m-2 d-1) (g C m-2 d-1) Upper Bay 0.58 4.64 0.22 
Mid Bay 0.18 50.2 1.24 
~er Bay 0.72 26.4 0.77 
204 
Fig. 5-1. The locations 
where cores were collected 
by the Chesapeake Bay 
Benthic Monito1ing Pro-
gram in the mainstem 
Chesapeake Bay from 
1984-1999. Samples south 
of Patuxent River (Px. R) 
accounted for 22% of the 
total, more than suggested 
here, but reflecting less 
intensive sampling effort 
in the southern portion of 
the Bay. Greater use of 
random station locations in 
the middle and upper Bay 
reduced overplotting of 
repeatedly sampled loca-
tions, exaggerating the dif-
ference. 
205 
A. 
Littoral Zone Littoral Zone 
8. 
Fig. 5-2. (A) The partitioning of the mid Chesapeake Bay into regions for 
computation of potential filtration by suspension feeders using the 
approach of Gerritsen et al. (1994). (B) A collapsed model in the case 
where it is assumed that there is not a pelagic zone, efffective/y isolated 
from the bottom by water column stratification. This scheme was used for 
the upper and lower regions of the Bay, while the scheme in (A) was used 
for the mid Bay. 
206 
Fig 5-3. Univariate plots relating macrobenthic biomass in Chesapeake Bay to 
key variab les expected to affect macrobenthic biomass. Macrobenthic bio-
mass (g AFDW m-2) was recoded by adding 0.01 g m-2 then log-transformed. 
Biomass was computed from data reported by the Chesapeake Bay Benthic 
Monitoring Program. Water quality was concurrently measured in bottom 
water. Percent si lt-clay content of sediment was also measured concurrently. 
Water quality and sediment properties were both collected by the Chesapeake 
Bay Benthic Monito1ing Program. 
207 
4 
Salinity (ppt) 
3 
2 
1 f."~{:l1i1J~It~~;iD:Jt)::t-
0 .. ?~ :. ?' ?,. .. ;(~i?lA-:- ~~""'u~:?. :?_;:~~- ? ? l:? . .. ,?: . ?. . ...... ! .... t.?,,. !'.: ?:i; ..._ :,z, .. ?'? ?? 
-1 t??. 4 = :, <21t:l'.Jt;z Depth (m) 
-2 3 '. 
0 5 10 15 20 25 30 35 2 
-- 4 .7.tt/.?.\. ?: ?I ?1 ,-,.,,_., ? I? o I ..-- Dissolved Oxygen (mg 1-1 -- 1 1; J,::'?or./; I ., t ? ? : ! I - ?;,zt,-.: ? \. ? ~1:J?,1 I ? ? ,: ? ~ 3 ..-- .....: ,~?-_: ??~" 'J.. h\ .... .. . . . . . ? I' 0 ~?4?~. .... ...~.. ~-~~ . ...  0 0 ! .. + 2 0 
N 
I . ..,.-\:=.:.:lf~ tttJ1,!:(i/: + -1 ?::\' :> -~-:{ if
!t ~  \.. .: ': ,?:?? . ..: .-:- ? ,;... .' ..) { ! ~ ! . . I O 1 
N E 1 N 
0 
00 s .. -.,?:1 . .- .... :? !?:~~: ???~'.;,:~~?i;~?~.-? 'if',. . -..:1,; ..... ,.~ ~ .. 'E 
-2 
0 t.::Y':"~~J. :-:~ ?-~t .. -t:t..}tlrt}:~~f ....- ~ ..r.~.?! ?.::$ ... r~? ?\"\:i;,.? ..... 15 0 0 5 10 20 25 30 35 40 u.. s <( -1 t!'7:f\. '.?.<?: 0 4 u.. Water Temperature (?C) 
..O....>.,  -2 <( 3 
O> 0 2 4 6 8 10 12 14 16 18 ..O....>.,  0 
O> 2 
0 
4 \ % Silt-Clay! 1 ..- ?. ?<?:J:?;:) ~-:_:~ :??.:.- ~-~-- .:J~~~-?>i:-:{~i1f }{l~t-.?::"  . 
3 .. ~-?:i: r..~. ....? ?;?. =.::?,? ~.:::.::~?--:?:~'?:'.;.o\:~:;J~a;~i??:: . 
... ~- 0 -~.- ??/.? .:: tt;.~.::-::>,~(;?~~f ... ~~-:~- :::'?:,:;1:~;.~ ...~ ~=-~::.1-,j:\!'<}":?: 
2 _..  -~i? ,.~ t,: . : ?. .~  ... ? .. rr.?:.;> .. ...1. . . ?=? ?: ?: , .. :- ~?.: ?~.- :?:.."-;?,.?.1~.-~J:r.---?: .? 
1 f.'itf ~:;\;.ti;_,: 't~:\:,.:(.-;;/;Jij -1 .-?.-=??.:.--'.:-?/~~,,:?? .... ~~ ?. .... ...-~-- .. , . ?:?:.~?~lF -2 i." ,:?J:1,.i 
0 :t,t-. :.  .. -1 . ,??.,  ~? .  ? .. . . ?. ?..  . . -. ,?  ~..I ? ?}"',-t'ff:i"l'.Yf,}  0 5 10 15 20 25 30 
-2 
0 25 50 75 100 
Hypoxic Normoxic 
1 
Shallow Deep 
Oligohaline 
g AFDW m-2 
N 0.1 
0 - o.5 
\0 - o.5 
- 5.0 
- 10.0 
- 20.0 
Fig. 5-4. A regression-tree predicting macrobenthic biomass in Chesapeake Bay. Splitting rules and node detail are provid-
ed in Table 5-2. Descriptive terms indicate major regions of the tree-structure defined on the basis of primary splitting 
rules. Terminal nodes , denoted by "T" before the terminal node number, are shaded according to the geometric mean bio-
mass for observations classified to the node. The distribution of values in the terminal nodes is also shown in Fig. 5-5. 
103 ~ - - - - - - ------- --- ------------, 
A ? 
? ' ?.
? 
  : ? 
....- ? ? ? 
0 
0 ? ? I I 
? ? 
+ 101 ? 
~ : ? I 
E I ? 
~ ? 
0 
LL 100 I 
? 
<( ? I ? ?I  O'l ?
? ? ? ?
  ? 
? I ? ? 
? ? ? ? I ? 
? 
41826 2 5 24 6 202717 3 123251516 7 12218101419221311 9 
~ 225 
0 B ..... . 
ro 150 
2: 
Q) 
Cf) 75 
..0 
0 
41826 2 5 24 6 202717 3 123251516 71221 8101419221311 9 
Node Number 
Fig. 5-5. (A) The distribution of macrobenthic biomass among observations 
classified into the 27 terminal nodes of the regression tree in Fig. 5-3 . To facili -
tate plotting on the log scale, 0.01 g m-2 was added to all values; therefore, 10-2 
g m-2 is equal to zero biomass. Boxes indicate the quartiles and median, while 
the whiskers indicate the 5th and 95th percentiles. Observations outside the 5th 
and 95th percentiles are plotted individually (B) The number of observations 
classified into each node. 
210 
3 A 0 ? 
,--... 
...--- 0 0 
0 0 [l 
0 0 C) 
+ 2 0 0 ,--... g 
N t, 8 
E g 0 
s 1 0 0 8 e 0 0 og 8g 0 
LL 
<{ 
-0--- 0 0 0 ') 
0 0 00 
---- e 0 
"O O 0 
Q) 
2: 
Q) -1 0 
(/) 
..0 
0 
() 2
() 0 r =0.59, N=1664 -2 
3 ~-------------------, 
B g 
2 8
0 
0 8~ 8 0 0 
I 
cu 1 f r 
~ 
"O 
(/) 0 
Q) 
0 
0::: -1 8 8 
-2 Censored ~ 0 
0 
-3 ..l.__,.---.----,--- -=r-----.----r___J 
-2 -1 0 1 2 3 
Predicted Biomass (log(AFDW m2)+0.01) 
F ig. 5-6 . (A) Observed vs. predicted macrobenthic biomass at sites in 
mainstem Chesapeake Bay. Predictions result from a regression tree 
mode l re lating macrobenthic biomass to habitat factors (Fig. 5-3). 
Biomass values were recoded by adding 0.01 g AFDW m-2 then log 
transformed . This resulted in left-censoring of the data at -2 on the 
log scale. (B) A plot of predicted vs . re iduals (observed-predicted). 
Because of le ft-censoring o f observed values at -2, the sum of the pre-
dic ted value and associated res idualcannot be less than -2. 
2 11 
(.,.) (.,.) (.,.) 
(!) (!) 0) 
N ->. ~ 
CJ] ~I  q --- ----5 - UB NMB- Shallow I SMB-Shallow 7_6 m ,~ ~ 
- 10 - NMB- Mid I SMB-M1d 1o  m N NLB _,_ SLB ~ CJ] (.,.) (.,.) a 
.S 0) (.,.) 15 - ~ NMB- Deep N 0) SMB f"\ ..c q 0 I o_ 20 - q Deep I~\ (\ 
Q) 
0 l ,\ 
\ ..! I I 
25 - L__J ,.J 
(, 
.N... . 30 -
~_/1 
\ ,, !J \' ,-
N 35 - ',; 
250 200 150 100 50 0 
Channel Distance from Bay Mouth (km) 
Fig. 5-7. A cross-section of the mainstem Chesapeake Bay indicating the maximum depth pro-
file up to 35 m. Depth and axial distance boundaries are shown for regions identified by a 
regression tree-model predicting macrobenthic biomass from latitude and depth . The regres-
sion-tree algorithm finds regions within which the variance of biomass is minimized. 
- -------
0.100 
0.080 
0.060 -- Edgar(1990) 
0.050 - - Tumbiolo and Downing (1994) 
0.040 Banse and Mosher (1980) 
0.030 
0.020 
..--
"' O 0.010 
-OJ 0.008 Q_ 
0.006 
0.005 
0.004 T=25? C 
0.003 
0.002 
T=5? C 
0.001 f------..----.-----,---~--'-< 
10-s 10-4 10-3 10-1 
Body Size (g AFDW) 
Fig. 5-8. A comparison of three statistical models predict-
ing production/biomass ratios for macrobenthic animals. 
For the Tumbiolo and Downing (1994) model, a depth of 
10 m has been assumed. 
213 
10? -,----------------------------, 
a b be C bed d C be b 
10-I 
?? 
I 
 ? ? ? 
.-
s ? ~ 10-2 ? I 
0 ? ? 
u. 
<( I ? 
0) 
'-" 
Q.) 
N 10-3 
Cl) ? 
>, 
"O 
0 
(0 ~ 
10-4 ? I I I
?  
? ? 
I ? I ? ? 
10-5 ~----.-------.----.?--- -----.----.----.----,----.---.-~ 
Upper N Mid SMid NMid S Mid NMid S Mid N Lower S Lower 
Bay Bay, Bay, Bay, Bay, Bay, Bay, Bay Bay 
Shallow Shallow Mid Mid Deep Deep 
Fi g. 5-9. Mean body size of macrobenthos in 9 regions of Chesapeake Bay comput-
ed from the Chesapeake Bay Benthic Monitoring Program data. Regions not shar-
ing the same letter at the top of the graph have significantly different mean body size 
(Tu key-Kramer adjusted p<0.05). "C. v." indicates the approximate size of the large 
polychaete Chaetopterus variopedatus , whose abundance is under-represented in 
the Benthic Monitoring data and is estimated to account for -50% of the biomass in 
the lower Bay (Thompson 2000). 
214 
70 
"-:' 1 Ythan estuary 
E 
s: 60 UCBD .1 
2 Grevelingen 
0 2. 
3 Oosterschelde 
~B 
u.. 4 Balgzand ( 1980s) 
<{ 50 (+ext. input) 5 Veerse Meer 
-3 ?3 6 Ems Estuary 
(J) 
(J) 40 (Wadden) 
ro 
E 7 Lynher Estuary 
0 ? 4 
en 30 8 Long Island Snd 
.5:2 6 9 San Francisco Bay 
..c 20 10 9 ? ; DLCB 10 Balgzand ( 1970s) c 
Q) 11 ? ?..s,  11 Bay of Fundy ..0 
0 ? 12 Westerschelde 
'- 10 ?12 
() 
ro 13 Ems Estuary 14 ?13 
:':! OMCB (inner) 
0 14 Columbia R. 
0 100 200 300 400 500 600 700 estuary 
Annual Phytoplankton Production (gC m-2 y-1) 
Fig. 5-10. The relationship between annual phytoplankton production and 
macrobenthic biomass is estua1ies and coastal systems as shown in a review by 
Herman et al. (1999). Three observations have been added reflecting 
macrobenthic biomass and annual phytoplankton production in three regions of 
Chesapeake Bay. Macrobenthic biomass estimates are from this study. 
Phytoplankton production estimates were computed from data in Harding et al. 
(2001) and are means of spring, summer and fall regional mean production. 
UCB=Upper Chesapeake Bay, MCB=middle Chesapeake Bay, LCB=Lower 
Chesapeake Bay. The upper Bay observation is translated to the right by adding 
the estimated 266 mgC m-2 ct- 1, which is the approximate annual subsidy due to 
allochthonous organic matter inputs to the region. 
215 
Chapter 6 : 
A NETWORK ANALYSIS OF MAINS'l'EM CHESAPEAKE BAY FOOD 
WEBS DURING SUMMER: EUTROPH[CATION EFFECTS ON CARBON 
TRANSFER EFFICIENCY TO FISH 
Abstract 
Time-invariant trophic-flow networks were constructed for three regions of the 
Chesapeake Bay during summer. The trophi c flow networks nominally represent the 
period from 1985-1999, for whi ch most o r the data were obtained. A trophic flow 
network was also described to characte ri ze the mid Bay following successful nutiient 
reductions sufficient to reduce phytoplankton biomass and production to levels 
observed in the 1950 's and early l960 's . The network parameters other than for 
phytoplankton were characterized using a combination of comparative ecological 
studies (bacteria and benthos) and the mass ba lance constraints of network analysis. 
Rather than attempting to estimate change in the biomass and production of fish, 
potential changes in fish production were cva l uated in term of the available production 
of prey items. 
The trophic flow networks were eva luated to determine how trophic transfer 
efficiency and major patterns of carbon flow differed among the regions and between 
the modern and restored mid Bay networks . The most dramatic differences pertained 
to the relative role of bacterial processing of organic matter, which was much greater 
in the mid Bay than elsewhere in the Bay as well as in other coastal systems. The mid 
Bay macrobenthos was found to be degraded as compared to the other regions of the 
2 16 
Bay and similarly productive coastal systems. Benthic production was sufficient to 
support the demands of demersal predators, but with little surplus available. The 
macrobenthos outside the mid Bay and in the restored mid Bay was sufficient to 
support a much larger biomass of demersal predators than was present, possibly 
indicating overfi shi ng. 
Trophic transfer efficiency was estimated to average 30%, but was not a 
sensiti ve indicator of efficient trophic transfers to fish . Gross elemental transfer 
efficiencies and total contribution coefficients, analogous measures of transfer along 
particular trophic pathways, were more informative. A combination of informational 
indi ces that has been proposed as a quantitative definition of eutrophication also 
proved to be insensitive in practice to indicators of change and may be overly sensitive 
to investi gator bias. Effective food web connectance is proposed as a useful indicator 
of a healthy estuarine ecosystem. 
Introduction 
Eutrophication, defined here as an increase in the rate of organic input (Nixon 
1995), is a growing global problem in estuaries and coastal systems. One frequently 
observed consequence of eutrophication is habitat degradation and loss . In 
Chesapeake Bay, as in many other estuaries, loss of vegetated habitats (01th and 
Moore 1984) and increased frequency, extent and impact of hypoxia and anoxia (e.g. 
Officer et al. 1984, Chapter 3, Chapter 5) are two major consequences of 
eutrophication. The combined effects of organic enrichment and habitat loss on the 
ecosystem are likely to be complex. For example, eutrophication could have direct 
217 
beneficia l effects, whereas increased organic inputs increase production by food 
limi ted consumer organi sms (i.e. secondary production). On the other hand, habitat 
degradat ion coul d change the abso lute and relative magnitudes of the trophic transfers 
that collecti vely desc ribe the "food web" of the estuari ne ecosystem, potentiall y 
negating any direct beneficial effects of enri chment. An understanding of these food 
web effects is an important e lement of our understanding of eco logica l changes 
assoc iated with eutrophi cati on, parti cularl y when we seek to describe the effects of 
eutrophicati on on fi sh and shellfi sh resources in estua1i es. 
Producti on of fi sh and shellfi sh is one of the important functions of estuarine 
ecosystems. In part due to their hi gh primary productivity, estuari es contribute a much 
greater frac ti on of global fi sh production than would be expected from their surface 
area (Houde and Rutherford 1993). While Chesapeake Bay is no exception , 
degradation of criti cal habitats (e.g. SA V) and water quality raise the concern that fi sh 
production in the Bay could be threatened. Interestingly, even though some fi sheri es 
(e.g. American oyster, Crassostrea virginica and American shad Alosa sapidissima) 
have dec lined dramatically, total landings from Chesapeake Bay remain hi gh, buoyed 
by substanti al landings of blue crab and menhaden (Houde et al. 1999). Whether 
eutrophication has enabled total landings to be maintained in the face of large declines 
in landings of several hi stori cally important species is an important question . A 
simil ar question is whether a reversal of eutrophication vi a decreases in nutrient 
loading rates could have deleterious effects on fi sheri es in the Bay by decreasing prey 
abundance. 
218 
The concept of trophic transfer effi ciency (TIE) is central to the questions 
introduced above. TIE is defined here as the fraction of organic matter consumed at 
one trophic level (gross production for primary producers) that is consumed at the nex t 
higher trophic leve l. For the / 11 trophic level, this is TTEi = Ci+! /Ci . Consumption at 
trophic level n can be expressed as (C2/C1 )(C3/C2 ) .. . (C,z/C11 _1 ) , or equivalently as 
lTE1 ? TI'E2 .. . TTE ? The average TTE for a series of trophic transfers is the 11
geometric mean of the efficiencies at each trophic transfer, namely 
- [ DII  Jl/11 lTE = TI'Ei . Trophic transfers over pathways of length n steps can therefore 
11 
be computed as c. = c . . TTE which is the same basic expression commonly used 
t + II I ' 
to predict potential fish yields for marine systems (e.g. Ryther 1969, Iverson 1990, 
Boude and Rutherford 1993). Accordingly, it is clear that a decrease in consumption 
at upper trophic levels, concuJTent with an increase in organic inputs, entails either a 
decrease in average TIE or an increase in the number of trophic transfers. 
Conversely, maintaining or increasing secondary production while organic inputs 
decline requires either increased TIE or a lower average number of trophic levels. 
Recognizing that this latter possibility, a desired outcome of ecosystem restoration and 
management efforts (e.g. EPA 2000), implies changes in either food web structure (i.e. 
number of transfers) and/or increased average TTE, it is important to understand if 
such changes are likely, and if so, what specific changes are most likely involved. 
This study examined TIE and food web structure by constructing and 
analyzing time-invarient, quantitative trophic flow networks. A time-invruient 
219 
modeling approach was selected rather than dynamic simulation models because the 
add iti onal information provided by dynamic simulations was not needed to address the 
specific questions of interest. Equally important is the fact that the time-invarient 
modeling approach makes maximal use of the unprecedented data resources now 
available to examine Chesapeake Bay food webs. Baird and Ulanowicz (1989), 
introducing a landmark application of time-invarient trophic flow models to 
Chesapeake Bay, suggested that "more emphasis should be placed on new methods of 
interpreting the mass of data at hand." However, their work, which was based on data 
collected through the mid 1980's, largely preceded the 15-year accumulation of 
detailed biological and physical data by the Chesapeake Bay Monitoring Program. 
Their research effort also preceded 12 years of intensive investigation of the 
Chesapeake Bay ecosystem under the auspices of two 6-year NSF-sponsored Land-
Margin Research Program (LMER) programs. The first LMER program, "Processes 
of Recycling, Organic Transformations and Exchanges between the Upland and the 
Sea" or PROTEUS, investigated primary production, nutrient cycling and major 
organic matter transformations, primarily involving the plankton and benthic 
microbiota. The second LMER program, "Trophic Interactions in Estuaiine Systems," 
or TIES, focused on processes leading to secondary production in the plankton and 
nekton. Thus, the comments of Baird and Ulanowicz (1989) are even more apropos 
today. 
Specifying and analyzing such time-invarient trophic flow networks has 
commonly been referred to as "network analysis." Network analysis provides a 
valuable framework for data synthesis and, once networks have been quantified, a 
220 
poweiful suite of algorithms for analyzing them. The algorithms were principally 
developed and described by R. E. Ulanowicz and colleagues (Ulanowicz and Kemp 
1979, Ulanowicz 1986, Szyrmer and Ulanowicz 1987, Ulanowicz and Puccia 1990) 
and Were later embodied in software packages available from Ulanowicz (NETWRK , 
Ulanowicz 1999). Another software package, ECOPATH with ECOSIM, 
(Chiistensen and Pauly 1992, Christensen et al. 2000) was built on the basic model of 
Polovina (1984) and later incorporated some of the algorithms of NETWRK, plus 
rnany additional functions. The power of both the network analysis approach and the 
breadth of questions that can be addressed are underscored by the substantial impact of 
the first study to investigate Chesapeake Bay food webs using network analysis (Baird 
and Ulanowicz 1989). Although the goals of Baird and Ulanowicz (1989) differed 
frorn those of most interest here, the model structure that they described and their 
initial parameter estimates were a valuable point of departure. 
Using the methods introduced above (see "Methods" for detailed description), 
th is study examined change in TIE and food web structure associated with 
eutrophication. The ideal manner in which to do this would be to independently 
desc1ibe trophic flow networks as they changed with eutrophication over time. 
Bowever, it was realistically possible to specify only a time-averaged network 
addressing the food web in recent years. This problem was addressed in two separate 
Ways. First, because eutrophication has affected different regions of Chesapeake Bay 
to differing degrees and in different ways (i.e. phytoplankton biomass, Harding and 
Pen?y 1997; hypoxia, Chapter 3; macrobenthic biomass, Chapter 5) food webs in three 
regions of Chesapeake Bay were contrasted. Since the mid Bay is most dramatically 
221 
affected by increased hypoxia, this study contrasted the mid Bay food web with that of 
other regions of the Bay, hypothesizing that TTE would be lower in the mid Bay than 
in the other regions of the Bay. A second approach that was used to examine 
eutrophication-related changes in the food web wm to develop an empi1ically-based 
description of a food web for a restored Chesapeake Bay. For brevity, this analysis 
was limited to the mid Bay. The analysis was based on avail able historical data and a 
series of specific "rules" derived from comparative ecological studies and mass 
balance and other constraints afforded by network analysis. As with the comparative 
approach, the hypothesis here was that the restored Bay network would show that 
despite lower primary production, secondary production could be maintained through 
hi gher average TTE. 
Study Area and Scope of the Analysis 
Chesapeake Bay is a large , partially mixed estuary on the mid-Atlantic coast of 
the United States. The spatial scope of resea rch was limited to the mainstem Bay, 
excluding the major tributaries and major embaym~nts such as Eastern Bay and 
Tangier Sound (Fig. 6-1). This choice refl ects known ecological differences between 
the mainstem and coITesponding sa lini ty zones in these adjacent environments. 
Extending -300 km from the head of the estuary at the Susquehanna River in 
Maryland to its mouth between the capes of Virginia, the mainstem Chesapeake Bay is 
commonly divided into three functionally distinct regions, the upper Bay, mid Bay and 
lower Bay. For the purposes of thi s study, the boundaries were determined by the 
segmentation scheme of the box model used to estimate physical transport among the 
222 
regions (Chapter 2). Surface a reas of the respective regions are 472, 2338, and 2661 
km 2 w ith mean depths of 4.4, 10, and 9 m. Bathymetric values were computed from 
data reported by Cronin and Pritchard (1975). 
The summer season was exa mined because this is the time of the year when 
hypoxia, one of the major effects of eu tro phication in Chesapeake Bay, is most 
deve loped (Chapte r 2) . Summer was defin ed as June l to August 31 , which has a 
durati on of 92 days. Carbon was se lec ted as the currency for the model due to the fact 
that more information is availab le to desc ribcexchanges involving carbon relative to 
other e lements. Depending on the ques ti ons at hand, investi gation of trophic 
exchanges of other e lements such as nitrogen or phosphorus may also be appropriate 
(e.g. Ducklow 1983). One could a lso ex am ine flows of energy rather than elemental 
flow s. From this perspective, ca rbo n flow approximates energy flow more than some 
other currencies (i.e. wet weight) . C hanges in energy content per unit carbon depend 
on the relative importance of carbohydrates , proteins and lipids. These changes were 
implied throught the carbon-specific assimi la:ion and growth efficiencies. 
Methods 
Use of trophic network mode ls requires explicit identification of the elements 
or nodes that comprise the food web . For thi s study, the food webs of the upper, mid 
and lower Bay were conceived to consist of 34 nodes (Fig. 6-2). These include 3 
detrital pools, 4 types of primary produce rs, 9 planktonic consumers, 5 benthic 
consumers, and 13 nektonic consumers. 
223 
Network ana lysis requires estimates of all the flows between the nodes , the 
dis-; ipation of carbon via resp iration, and the net input minus export. The latter can 
occ ur equ ivalently in time (i.e. biomass accumulation or depletion), space 
(i1n111 igration or emigration), or due to harvest. For the / 11 node these flows are 
rn 11 strained by the requ irements that 
C- = P. - R-- U . 
I I I I (1) 
EE- = Y. + E ? + BA + M 2 . ? B-I I I I I < l 
I - (2) p. 
I 
whc1e the terms of which are defi ned in Table 6-1. The flow of carbon fro m node i to 
11odcj can be represented as TiJ. If the network has n nodes, Then x n matrix of TiJ 
dcsu ibes all the internal fl ows in the network and is called the diet matri x. Network 
;111 ;11 )Sis also requires an estimate of the external fl ows, which can be specified as a 
i11 put and outputs vectors, each with n elements. One approach to specifying the 
11 c1w ork, called in verse analysis, is to estimate some of the fl ows, then solve for the 
1\.: rn ;1ming subject to constraints (eq. 1, 2). Some elements of the diet matrix can be 
est i 111ated using inverse analys is (e.g. Vezina 1989). It is also poss ible to estimate all 
1h e 11:)ws independentl y via the so-called a priori method (Field et al. 1989), in which 
al I or nearly all the estimates are computed directly from fi eld data or the publi shed 
lit cr;11ure. Thi s study used thi s latter approach. 
224 
Qvantifying the Networks 
The network models were quantifi ed using a large volume of unpubli shed data 
' 
including all the data that comprise the different elements of the Chesapeake Bay 
Monito1ing Program. In addition , -180 published studies were utilized. Every effor1 
was made to utilize research and data that pertain directly to the biota and ecosystem 
processes of Chesapeake Bay. As a result, all the unpublished data sources and 56% 
of the published studies pertained directly to Chesapeake Bay. The basic approach for 
computing the model inputs is described below. In the interest of brevity, the 
description omits many important details, including appropriate citations of data 
sources. Instead, all the citations for data sources are summarized in a Table (Table 6-
2). A complete discussion is provided as a substantial appendix which makes 
appropriate reference to the sources of data, describes how they were used, what 
assumptions were made, and on what basis the assumptions were justified. 
Publications from the widely available peer reviewed literature were used 
preferentially over other sources and account for the majority of references cited. 
Many of the unpublished data that were used were freely available to the public, often 
Published via the Internet. For example, data were obtained from the web sites of the 
US Environmental Protection Agency's Chesapeake Bay Program, the US Geological 
Survey, and the National Oceanographic and Atmospheric Administration. 
Particularly useful for many computations related to finfish was the FishBase database 
(Froese and Pauly 2001). 
Three detrital compartments in each region were dissolved organic carbon 
(DOC), suspended particulate organic carbon (POC) and sediment organic carbon 
225 
(Fig. 6-2). The abundance of DOC and POC was computed from Chesapeake Bay 
Water Quality Monitoring Program data. Allochthonous input of POC into the 
mainstem Bay was assumed to come only from the Susquehanna River and was 
computed from loading rates reported by the USGS. Inputs of DOC were computed 
from USGS estimates, plus contributions from major tributmies, which were 
computed as the product of average surface water concentration and net freshwater 
outflows. Transport of POC among the regions was assumed to be negligible, while 
transport of DOC was computed as the product of concentration and box model-
esti mated flows (Chapter 2). The flux of organic matter from the water column to 
sediments was not computed directly but was estimated as the unutilized fraction of 
aggregate flows to POC. Phytoplankton and bacterial biomass was subtracted from 
measured POC pool to compute the non-living detrial fraction of POC. Similarly, the 
biomass of bacteria and meiobenthos was subtracted from the total sediment organic 
carbon to estimate the non-living carbonbiomass in sediments. 
Regional average rates of gross phytoplankton production were computed from 
studies that employed 14C uptake and net oxygen evolution methods and were 
partitioned into net plankton and picoplankton as reported by size-fractionated 
production studies. Production of SAV  was computed as the product of SA V 
coverage area and production rates of Chesapeake Bay SA V beds. SAV  production 
was not utilized directly, but instead was directed entirely through sediment POC. 
Benthic microalgal production was computed from published estimates of light 
saturated benthic microalgal production and photosynthesis-iITadiance relations. 
226 
Extrapolation to the whole system scale was accomplished via available estimates of 
li ght attenuation and bathymetry in each region. 
The food web included three bacterial nodes , free-living, particle-attached, and 
benthic bacteria (Fig. 6-2). Biomass and production of free-living and particle 
attached bacteria in the plankton were obtained from published studies employing 
acridine-orange direct counts and radiolabeled thymidine uptake techniques. Benthic 
bacterial metaboli sm was computed by subtracting the metabolic rate of meiofauna 
from the total sediment flux measurement, which incorporated 0 2 consumption and 
sulfate reduction. The biomass of benthic bacteria was estimated from respiration via 
assumed growth efficiency and biomass turnover rates. 
The zooplankton community was divided into three components, the 
microplankton , mesozooplankton and gelatinous plankton. The microplankton 
community included heterotrophic microflagellates, ciliates, rotifers and meroplankton 
(Fig. 6-2). Biomass for these nodes was obtained primarily from literature sources , 
but also from the Chesapeake Bay Microzooplankton Monitoring Program when the 
methods employed were deemed adequate. Energetic rates were derived entirely from 
literature sources. The mesozooplankton, which were defined to include all life stages 
of copepods (nauplii , copepodid, adult) were described using data from the Micro- and 
Mesozooplankton components of the Chesapeake Bay Monitoring Program. 
Bioenergetic rates were derived from literature sources as reconciled with field data 
via an age-structured zooplankton production model. The gelatinous zooplankton 
community was represented as two nodes , ctenophores and sea nettles. The biomass 
of these organisms was derived from the Chesapeake Bay Monitoring Program 
227 
(ctenophores only) and from unpublished data collected by the TIES program 
(ctenophores and nettles). Diet and energetics were derived using literature data and 
the relevant in situ conditions for Chesapeake Bay (e.g. prey density). 
The zoobenthos, other than bacte1ia, were represented by five nodes, the 
meiofauna, deposit-feeding benthos, suspension-feeding benthos, oysters and blue 
crab (Fig. 6-2). The biomass and energetics of meiofauna were estimated from 
published sources, except that biomass was assumed to be reduced in the anoxic 
regions of the mid Bay due to the exclusive presence of anaerobic meiofauna. The 
biomass and energetics of deposit-feeding and suspension-feeding benthos were 
derived from the Chesapeake Bay Benthic Monitoring Program data, except for the 
polychaete Chaetopterus as described in Chapter 5. After examining the available 
data, the biomass of oysters in Chesapeake Bay was found to be insignificant except in 
the tributaries. Blue crab biomass and bioenergetics were derived from published 
sources, which included a stock assessment based on fisheries methods. Fisheries 
removals from the mainstem Bay were estimated via fisheries landings data for the 
crab pot fishery only and migration of biomass, principally from the tributaries to 
fishing areas in the mainstem, was inferred by mass balance. 
The fish community, like the zoobenthos, included a diverse assemblage and 
was generally represented as a collection of nodes for individual species, following the 
practice of most applications of network analysis (Fig. 6-2). An exception was the 
herrings and shads, an assemblage of anadromous species. The fish taxa that were 
included are bay anchovy, menhaden, herrings and shads, spot, croaker, hogchoker, 
American eel, catfish, white perch, striped bass, bluefish, and weakfish (Table 6-3). 
228 
The procedure for estimating the abundance, energetics and trophic relationships 
vaiied by species. The most detailed information was available for bay anchovy, the 
abundance and bioenergetics of which was computed via an age structured approach 
based on published studies. Adult abundance was obtained from published estimates 
based on egg production models. The biomass of menhaden was estimated from a 
published vi1tual population analysis and from Chesapeake Bay landings by the purse-
seine fleet. The biomass and production of other species of fish was estimated via 
age-structured computations based on growth rates, natural and fishing mortality rates, 
predator demand, and commercial and recreational landings. Where possible, both the 
Weight and size structure of landings were used to infer the age-structured exploitation 
rate, from which biomass was computed. The diets and bioenergetics of striped bass, 
bluefish, and weakfish were based on the published results of a detailed age-structured 
investigation for these species in the mid Bay. The estimated biomass and ecological 
role of fish in the Bay is most likely a minimum estimate due to the fact that an 
Unknown number of Jess abundant or little studied species were not included in the 
model. 
~ncing the Networks 
The systematic estimation of biomass, bioenergetic rates and approximate diet 
composition provided a first estimate of the biomasses and flows comprising the 
trophic flow networks. Although these initial estimates were remarkably 
commensurable, some adjustments were needed to satisfy eq. (1) and eq. (2). The 
229 
i n i t ia\ estimates were entered into ECOPATH with ECOSIM software (Christensen et 
a \ . 2 000), which provided an interactive platform for making the needed adjustments . 
In most cases, the adjustments that were needed were to the diet composition. 
Whi \e the basic diet composition was often known a priori, the proportions 
c ontributed by each diet were usua\ly not known exactly, if at all . Thus, these were 
adjusted in accordance with severa\ conditions. First, the diet compositions must 
re su\t in EE; consistent with eq. (2). If EE;> l, then one can infer that one or more 
species must obtain a sma\\er fraction of its diet from the 11i ' node than originally 
estimated. Thus, diet compositions were modified in small increments from initial 
va\ues, but within reasonable limits suggested by the available a priori data, until eq 
(2) was true for each node. In making adjustments, it was also assumed that EE; c\ose 
to zero was probably incorrect in most cases (the sea nettle is a counter-ex.ample). A 
transformed version of the Chesson (1983, cited in Ch1istensen et a\. 2000) forage 
ratio was a\so used to assist in making these adjustments. The forage ratio ex.presses 
the consumption of a particular prey relative to its availability in the environment. 
These values were examined to ensure that indicated prey preferences were consistent 
with expectation based on published accounts. 
In a few cases (e.g. spot), eq. (2) was also used to derive minimum biomass 
estimates when other means for making such estimates were not sufficient. By 
assuming EE;=l, BA;=O, and E;=O, and rearranging, one obtains P; = Y; + M2; ? B;. 
The quantity M 2 ; ?B; is the sum of a\\ flows to predators , while P; can be ex.pressed as 
the product of the specific growth rate and biomass. Rearranging gives an estimate of 
230 
th
e minimum biomass required to balance combined fishing removals and predation 
losses. 
Equation (2) also applies to detritus. Specifically, internal flows to detritus , 
plus external inputs and any net decrease in biomass, must equal or exceed utilization 
of detritus. Total flows to detritus associated with each node can be represented as 
~ (1- EE;)+ U;. Depending on the species in question and the means by which 
inputs were estimated, these flows can represent many different processes. For 
example, these could include egestion, excretion, partial consumption of prey (i.e. 
"sloppy-feeding", Moller and Nielsen 2001), pseudo-feces production, and mortality. 
The fraction of detritus production from each node allocated to each of the three 
detritus pools was specified based on an assessment of the most likely processes 
involved. As a defau lt, it was assumed that unutilized DOC was exported, that 
unutilized POC was deposited to sediment POC, and that unutilized sediment POC 
Was exported via burial. In practice these assumptions were not always consistent 
With mass balance constraints. 
~Restored Bay" Scenario 
A trophic flow network was constructed to examine likely differences in 
carbon flow between the modern mid Bay and the mid Bay if it were restored to 
conditions approximating 1950-1965 by adjusting primary production rates. To 
ensure that the network was not arbitrary, it was constructed by following a series of 
consistently applied "rules" and assumptions: 
231 
l . Phytoplankton production is 50% less than present as suggested by historical 
chlorophyll -a concentrations (Harding and Pen-y 1997). 
2. Submersed aquatic vegetation extends to the 2 m depth contour. Light attenuation 
was assumed to be reduced to permit this dist1ibution based on known SAV/light 
relationships (Batiuk et al. 1992). 
3. Production by microphytobenthos is a function of bathymetry and light 
penetration. 
4. Macrobenthic biomass is consistent with estuaries having similar trophic status 
(i.e. organic inputs, Herman et al. 1999, Chapter 5). 
5. Ratio of benthic to non-benthic metabolism consistent with other estua1ies (Nixon 
1982). 
6. Planktonic bacterial uptake/net phytoplankton production consistent with lower 
Bay and other studies (e.g. Cole et al. 1988) 
7. Subsequent changes were the minimum needed to satisfy eq. (1) and eq. (2), i.e. 
balance the network. 
The same approach that was used to balance the trophic flow networks for the 
three regions of the modern Bay were used to balance the restored mid Bay network. 
This included maintaining diet compositions that are consistent with the literature and 
ensuring that diet electivities based on the Chesson (1983) forage ratio were 
reasonable. 
232 
Network Analysis 
Balanced trophic flow networks fo r each of the three regions of the Bay were 
exported from ECOPATH into SCOR format (Ulanowicz 1999). For each network, 
the suite of outputs available in ECOPATH andNETWRK software were used to 
make inferences about the trophic fl ow networks. These specifically inc luded 
apportionment of activity o f omnivores to integer trophic levels (Ulanowicz and Kemp 
1979, Ulanowicz 1995), quantificati on o f mixed trophic impacts (Ulanowicz and 
Puccia 1990), evaluation of tota l trophic dependency (Szyrmer and Ulanowicz 1987), 
and computation of indices based on in fo rmati on theory, which describe the flow 
network overall (Ulanowicz 1986). De criptiom of many of the relevant procedures 
and equations and the ir appropriate inte rpretati ons are provided by Christensen et al. 
(2000), Kay et a l. (1989), Baird and Ulanowicz [1989), and Ulanowicz (1986, 1997, 
1999). 
Numerical Precision and Rounding Errors 
Some propagation of eJTor due to rounding occurred as the result of 
conversions among rates, biomass specific rates. efficiencies, daily rates versus 
seasonal total rates, etc., as we ll as the exchange of data between ECOPATH and 
NETWRK. Rounding eITors are reveal din mall inconsistencies between some 
numbers that are presented. 
233 
Results and Discussion 
The major res ults obtained directl y by the process of parameteri zing the 
networks include biomass and producti on es Li mates for each node (Table 6-4, T able 6-
5) , trophic flow (Tu) matrices (T ables 6a,b ,c ), respiration (Table 6-7), and summer 
fi shery landings (Table 6-8). Estimates or average trophic level and ecotrophic 
efficiency were obtained through network analysis (Table 6-5). A general description 
of the food webs in each of the three region s follows. 
Upper Bay 
The upper Bay region had a low rat e of gross primary production (1487 mgC 
m?2 d.1) compared to the other regions of' the Bay (Table 6-5). Heterotrophic activity 
was subsidized by allochthonous inputs of organic matter, principally from the 
Susquehanna River, plus an import of sediment POC via depletion of detritus 
accumulated during spring, which was the larger of the two detrital sources. Primary 
production plus inputs of detritus tota led J 854 mgC m?2 d.1 (Fig. 6-3). 
A substantial biomass of suspension-feeders, principally the introduced bivalve 
Rangia cuneata, consumed the largest fraction of both phytoplankton and particulate 
detritus; however, mesozooplankton , ci liates , rnd menhaden were also important 
herbivores, as in other regions of the Bay (Table 6-6a) . Free-living bacteria consumed 
a large quantity of dissolved organic carbon (DOC) and were consumed themselves by 
heterotrophic microflagell ates and microphagous ciliates. Ciliates consumed 67% of 
heterotrophic microflagell ates (Tab le 6-5, Tab le 6-6a) . The largest fraction of ciliate 
production was consumed by the mesozoop lankton; however meroplankton, rotifers, 
234 
and even young-of-the-year bay anchovy also consumed ciliates. The average trophic 
level (TL) for mesozooplankton was 2.6, despite the fact that they obtained 63% of 
their diet at TL 2. The high TL estimate reflects consumption at TL 3, 4 and even 5 
(22%, 7%, 7%, respectively) due to predation on ciliates and rotifers (Table 6-5, Table 
6-6a). Bay anchovy consumed 70% of mesozooplankton production in the upper Bay 
and were the most important link between the zooplankton and nekton (Table 6-6a). 
Also important as links from the zooplankton to the nekton were the menhaden, and 
heni ngs and shads, which consumed an additional 17% of the mesozooplankton 
production. Menhaden, which were primarily phytoplanktivorous, were the largest 
and most efficient link from plankton to nekton, producing 53 mgcm-2d-1 (Table 6-5), 
about 2-fo ld greater than bay anchovy. Because of the zooplankton diet of bay 
anchovy, however, they required the equivalent of 6-fold more organic matter inputs. 
Ctenophores were a minor presence in the upper Bay, consuming only a small fraction 
of the mesozooplankton . 
The suspension feeding benthos in the upper Bay consumed more organic 
matter than any other node. Principally through egestion, they also accounted for 
about half of the organic matter fluxed to upper Bay sediment POC. Accounting for 
the large remova l of seston by benthic suspension-feeders required a significant flux 
of sediment POC back to suspended POC, which was assumed to occur via 
resuspension of sediments. Resuspension has been identified as a key mechanism in 
the sediment dynamics of the upper Bay in general and the estuarine turbidity 
maximum zone in particular (Sanford 1994). The upper Bay was the only region of 
the Bay where resuspension appeared to be important as a source of suspended POC. 
235 
The microphytobenthos contributed 6.3% of total primary production in the upper Bay 
and was most important for the meiofauna and the deposit feeding benthos. The upper 
Bay macrobenthos was consumed by an array of demersal feeding predators, which 
included hogchokers, blue crab and spot, which accounted for a combined 87% of 
predatory consumption of the benthos. Despite the high benthic production in the 
upper Bay, demersal feeding accounted for only 8% of trophic transfers to the nekton 
(20% excluding menhaden). This small fraction is reflected in the low ecotrophic 
efficiencies for the macrobenthos (Table 6-5), and may reflect the apparent resistance 
of larger individuals of the biomass dominant, Rangia cuneata, to predation by the 
blue crab and potentially other predators (Ebersole and Kennedy 1994). Overfishing 
may have reduced the biomass of American eels, which may at one time have exerted 
greater predation pressure on the upper Bay benthos. If not controlled by predation, 
the biomass of Rangia may be controlled in part by periodic mass mortality. For 
example, a major decrease in biomass occurred early in 1996, possibly associated with 
the record January flood, and biomass increased slowly in subsequent years 
(Chesapeake Bay Benthic Monitoring Program data). 
Thirteen fish species (or groups), plus the blue crab, were included in the upper 
2 1 
Bay network. Total fish production was 89 mgC m- d- (Table 6-5). Menhaden had 
the highest estimated biomass and production. After menhaden, bay anchovy were 
most productive, but the relatively slow-growing catfish had higher biomass (Table 6-
4, Table 6-5). The most productive demersal feeder was the rapidly growing 
hogchoker. Of the three major piscivores included in the network, weakfish were 
estimate  d t o b e most pro d uc t1? v e , although the biomass of long-lived striped bass was 
236 
greatest. Commercial fishery landings in the upper Bay exceeded the recreational 
catch due to blue crab landings, which accounted for 83% of all summer landings 
(Table 6-8). Many finfishes are landed primarily in spring and fall rather than summer. 
Recreational fishing accounted for 64% of summer finfish landings. The dominant 
species in the catch were bluefish and catfish (Table 6-8). 
Total respiration by the upper Bay food web averaged 1803 mgC m-2 d-1 (Table 
6-7, Fig. 6-4), exceeding gross production by 318 mgC m-2 d-1, or 21 % (Table 6-7). 
Summertime net heterotrophy exceeded the estimated advective input of detritus from 
the Susquehanna River and was supported to a larger extent by sediment POC 
accumulated during sp1ing and stored in sediments. Organic matter exports were 
dominated by emigrating menhaden (48 mgC m-2 d-1). Other studies have estimated 
that, on an annual basis, a substantial quantity of organic carbon is buried in the upper 
Bay (Kemp et al. 1997). While this observation is not disputed here, this study found 
that the summer food web completely utilized the available carbon resources, leaving 
little surplus for burial. Other than to assume that carbon was not buried in summer, 
this discrepancy cannot be resolved here. 
~ 
Gross primary production in the mid Bay was the highest among the three 
regions, averaging 3515 mgC m-2 d-1 (Table 6-5). Phytoplankton production 
accounted for 93% of total production (Table 6-5). The accumulated sediment POC 
2 
from spring (Fig. 6-3), which was estimated to be 27.8 gC m- (302 mgC m-2 d-1), 
accounting for about half of total detritus inputs (Fig. 6-3). The remaining 298 mgC 
237 
-2 - l 
m d was required to balance the carbon demands of the food web and was input to 
sediment POC from unknown sources (Fig. 6-3). Inputs at upper trophic levels were a 
small fraction of total inputs, about 9 mgC m-2 d-1, and were obtained principally 
th rough depletion of macrobenthic biomass (i.e. input from sp,ing). Such depletion of 
the benthos was not indicated outside the mid Bay, where seasonal changes in 
abundance were less pronounced (Chesapeake Bay Benthic Monitoring Program 
Data). 
The most important herbivores were the ciliates and mesozooplankton, which 
each consumed about 17% of net primary production (Table 6-6b). As in the upper 
Bay, the mesozooplankton were mostly herbivores, obtaining 73 % directly from 
phytoplankton (Table 6-6b). Only 12% of carbon flows to the mid Bay 
mesozooplankton passed through DOC (i.e. via the microbial loop), with the 
remainder obtained from phytoplankton through ciliates. The menhaden were the only 
substantial herbivore other than the zooplankton. About half of gross phytoplankton 
production in the mid Bay passed to detritus, either via excretion of DOC or as 
ungrazed net production (Table 6-5, Table 6-6b) . 
Despite the abundance of bay anchovy and ctenophores, the major consumers 
of mesozooplankton, 32% of the zooplankton production was not consumed by any 
predator (i.e. ecotrophic efficiency= 0.68, Table 6-5). To the extent that carbon flow 
alone supports such inference, this suggests that mesozooplankton were neither 
controlled by predators nor by food-limitation. An alternative hypothesis, consistent 
With the "foraging-arena theory" (Walters 2000) is that zooplankton occupied the 
lower water column to the extent permitted by hypoxia (Roman et al. 1993), perhaps 
238 
to avoid predation by bay anchovy, a visual predator (Luo et al. 1996). In doing so, 
th
ey may have been food-limited, even as they left a considerable fraction of primary 
production ungrazed. Spatial and temporal heterogeneity may play a role in 
preventing effective herbivore control of primary production as well. For example, if 
a phytoplankton bloom developed on time scales shorter than generation times for the 
principal consumers, it might be underexploited. The existence of excessively large or 
otherwise unpalatable phytoplankton species could also result in ungrazed primary 
production. 
Ctenophores were the second largest consumer of mesozooplankton in the mid 
Bay, behind bay anchovy. They consumed a quantity equal to 42% of consumption by 
bay anchovy (Table 6-6b). In this capacity they may divert a substantial flow of 
organic carbon away from piscivores production. Ctenophores were both a direct 
competitor for bay anchovy and a minor predator, consuming a small fraction of eggs. 
As a result, the adverse trophic impact on bay anchovy (Ulanowicz and Puccia 1990) 
Was the largest effect arising from ctenophores and the second largest affecting bay 
anchovy (behind menhaden). Surprisingly, the reciprocal competitive effect of bay 
anchovy on ctenophores was approximately equal in magnitude to the effect of sea 
nettle predation. Thus, while sea nettles have high clearance rates for bay anchovy 
eggs and larvae (see Appendix), they were overall a beneficial predator (Ulanowicz 
and Puccia 1990) on bay anchovy. In their limited abundance in the mainstem Bay, 
they may serve a useful ecological function. Evidence for sea nettles as a beneficial 
ecological presence was provided by recent observations. Low salinity due to high 
liver flow in 1996 limited reproduction of sea nettles (Purcell et al. 1999). 
239 
Concun-ently, a massive bloom of ctenophores (6-fold more abundant than average; J. 
Purcell , unpublished data) developed in that summer; bay anchovy recruitment levels 
in that summer were the lowest recorded (North 2001 , S. Jung, unpublished data). 
The benthos of the mid Bay was depauperate compared to both the upper Bay 
and lower Bay, where biomass was >10-fold greater (Chapter 4). Despite low 
Production by the benthos , production was sufficient to support the requirements of 
the demersal feeding fish and crabs. However, demersal predators consumed -80% of 
benthic production, a much larger fraction than elsewhere in the Bay. This high 
ecotrophic efficiency suggests that predation may have a greater detremental effect on 
the benthos in the mid Bay, exacerbating the effects due to water quality (e.g. 
hypoxia). In fact , poor water quality may interact with predation pressure to enhance 
the vulnerability of the benthos to predation (Diaz et al. 1992). 
The fish community of the mid Bay was similar to that of the upper Bay, 
except that biomass of low salinity species such as catfish, white perch and eel was 
reduced. Total fish production in the mid Bay was slightly higher than in the upper 
Bay, 93 mgC m-2 d-1? As elsewhere, menhaden and bay anchovy accounted for most 
trophic transfers from plankton to the fish community, as well as most fish production. 
Other than these forage species, spot and croaker had the highest biomass and 
Production, followed by blue crab. Blue crab accounted for 86% of summer landings 
from the mid Bay, reflecting the aggressive crab pot fishery (Table 6-8). Bluefish was 
second most important followed by weakfish, spot and croaker. A significant 
component of the already small landings of Menhaden by the pound net fishery in 
Maryland were assumed to be contributed by tributary landings. Therefore, these 
240 
landings were neglected. The striped bass fishery, an important fishery in the mid 
Bay, was closed during much of the study period, and remained closed during summer 
at least through 1994 (EPA 1994). Although the recreational fishery is now open in 
summer, striped bass production was assumed to accumulate during summer with no 
landings. 
Total respiration for the mid Bay food web averaged 4049 mgC m-2 d-1, the 
highest of the three regions (Fig. 6-4), exceeding gross primary production by 538 
2 
mgc m- d-1 (15%). The mid Bay was more heterotrophic than the upper Bay in 
absolute terms, but the deficit was smaller as a fraction of primary production. 
~ 
Lower bay primary production averaged 2,956 mgC m-2 d-1 (Table 6-5) of 
Which 89% was due to phytoplankton production (Fig. 6-3). The lower Bay food web 
heavily utilized the available organic carbon, including the accumulated sediment 
Poe left from spring (21 gC m-2) and an approximately equal quantity of DOC which 
Was required from unknown sources to balance the food web. 
The most important herbivore in the lower Bay was the mesozooplankton, 
Which consumed 14% of gross primary production (GPP). Ciliates consumed an 
additional 9% of GPP. As in the upper Bay, the suspension feeding benthos also 
consumed a substantial fraction, about 10% of GPP. Only 39% of GPP was not 
consumed directly by herbivores, which, accounting for algal respiration (-20%, Fig. 
6-4) and excretion of DOC indicates that phytoplankton production in the lower Bay 
Was nearly fully exploited. 
241 
Mesozooplankton biomass was higher in the lower Bay than elsewhere, 
although mesozooplankton P/B was estimated to be lower. Thus, both production of 
mesozooplankton and consumption by mesozooplankton was comparable to rates in 
the mid Bay. The average trophic level of mesozooplankton in the mid Bay was 
estimated to be 2.4, slightly lower than in the mid Bay and reflecting only a slightly 
decreased consumption of protozooplankton. As in the mid Bay, mesozooplankton 
production was not fully exploited by predators (Table 6-5). Bay anchovy consumed 
35% of zooplankton production and was by far the largest consumer (Table 6-5, Table 
6-6c). Ctenophores, which were less abundant in the lower Bay than in the mid Bay, 
consumed 13% of production, while cannibalism by mesozooplankton accounted for a 
similar fraction. Ecotrophic efficiency for lower Bay mesozooplankton was 62%, 
leaving a significant fraction of production unexploited (Table 6-5). 
The macrobenthos of the lower Bay includes a diverse assemblage of taxa, 
dominated by polychaetes. The deep-burrowing tubicolous polychaete Chaetopterus 
variopedatus is the biomass-dominant among the suspension-feeders and accounts for 
about half of the macrobenthos (Chapter 4). While total macrobenthic biomass in the 
lower Bay was only 37% of upper Bay biomass, benthic production in the lower Bay 
was 65% of upper Bay rates due to the presence of more rapidly growing taxa in the 
lower Bay (Table 6-5). Predators consumed about 50% and 25% of deposit-feeder 
and suspension-feeder production in the lower Bay, respectively (Table 6-5). 
Although larger Chaetopterus are resistant to predation due to their deep burrows (L. 
Schaffner, personal communication), this low ecotrophic efficiency suggests that 
increased predation could be supported. 
242 
Fish production was estimated to be higher in the lower Bay than in the other 
regions. Eight of the most abundant fish species were included. As in the other 
regions of the Bay, menhaden and bay anchovy accounted for most (84%) of the total 
fish production. The purse-seine fishery in the lower Bay accounted for most of the 
fi sheries landings in the lower Bay, and Bay-wide as well (Table 6-8). Spot, croaker 
and hogchokers were also productive, accounting for 75% of the remaining fish 
production (Table 6-5). The number of fish species represented in the lower Bay 
network is less than in the other regions, probably an artifact of data limitations. 
Therefore, the estimates of the role of the fish community here are minimum estimates 
(S. Jung, unpublished data) 
2 1
Total respiration in the lower Bay was 3204 mgC m- d- , exceeding gross 
primary production by 247 mgC m-2 d-1 or 8% (Fig. 6-4). Of the three Bay regions, 
the lower Bay was the least heterotrophic during summer, in both absolute terms and 
as a fraction of primary production. Net heterotrophy was approximately balanced by 
detrital input due to the spring accumulation of sediment POC (Fig. 6-3). However, an 
2 1 
additional net detrital input (DOC) of 61 mgC m- d- was needed from unknown 
sources to balance fisheries landings and biomass accumulation in fish. 
~ mining Eutrophication Effects 
Given the primary description provided above, several means of analyzing the 
trophic flow networks were utilized to examine possible effects due to eutrophication. 
A hypothesis of this study was that compared to elsewhere in the Bay, the mid Bay 
trophic network would exhibit more substantial eutrophication effects. Specifically, it 
243 
was hypothesized that degradation of critical habitats via eutrophication decreased 
trophic transfer efficiencies, thereby limiting secondary production despite high rates 
of primary production. Trophic transfer efficiency is defined as the fraction of organic 
matter consumed at one trophic level (TL) that is subsequently consumed at the next 
TL. These efficiencies are difficult to examine in a food web, but can be readily 
computed by projecting the food web into a linear chain of canonical flows (Fig. 6-
Sa,b; Ulanowicz and Kemp 1979, Ulanowicz 1995, see also Baird and Ulanowicz 
1989) and then combining primary production and detrital flows to produce a 
"Lindeman Spine" (Fig. 6-6). In order to combine these flows , it was necessary to 
extract the cyclic flow from primary producers directly to detritus (Fig. 6-6). Uti lizing 
the algorithms of NETWRK, inflows due to immigrating biota were distributed over 
the canonical food chain according to appo11ioned activity at each trophic level. A 
similar allocation was applied for emigration, exploitation, other exports and 
respiration (Fig. 6-5a,b). 
Trophic transfer efficiencies for TL l were -85% in all three regions. This 
high efficiency reflects little more than the Joss due to algal respiration from combined 
gross production and detritus flow. For trophic levels 2-5, efficiencies were 20-30%, 
not differing systematically with increasing trophic level or among the regions. 
Trophic transfer efficiency declined precipitously for trophic levels greater than 6, as 
illustrated by the retreat of flows at these levels below the log-linear decline in flows 
to trophic levels 1-5 (Fig. 6-7). Although one might assume that this reflects 
decreased growth efficiency and biomass of top predators (see Appendix, Table A-13), 
this is incorrect. The carbon flow at these high concatenations was principally due to 
244 
ctenophores, bay anchovy and sea nettles, which obtain a fraction of their diet via the 
microbial food web. 
Average trophic efficiency over n trophic levels can be computed as the 
1/11 
geometric mean (see Methods), namely IT?,, = gI TE; , where ITE; is the 
[ ]
trophic transfer efficiency for the/'' trophic level. While n can be set to the highest 
canonical trophic level , which was 10 in the mid Bay, this appears to be arbitrary if 
presented as a property of the system. At n=8, the "average efficiency" is Jess than the 
true mean for almost all the trophic transfers that occur (Fig. 6-7). For the Chesapeake 
Bay, n==5 may be most relevant, since none of the taxa obtained more than 5% of their 
diet above this level. By this measure, the upper Bay was slightly more efficient than 
the other regions, with ITE5=0.30 versus 0.27 for both the mid Bay and lower Bay. 
This efficiency is higher than the rule-of-thumb figure of 10% and is comparable to 
the higher rates suggested by Iverson (1990). While ITE is usually computed on the 
basis of internal flows only, this decreases ITE artificially when fisheries remove a 
large fraction of the production. Eliminating this artifact increased TTE the most for 
trophic levels higher than 5 where other flows were small, but did not have a 
significant effect otherwise. 
Because the lower Bay fish community was probably under-represented, an 
alternative scenario was examined to see if this may have affected the trophic transfer 
efficiency estimates for the lower Bay. Fish biomass was artificially increased in three 
feeding guilds: p lanktivores, benthivores, and piscivores. Surprisingly, the effect on 
trophic transfer efficiency of even a dramatic increase in fish biomass was small , even 
245 
though the effect of this change on some other parameters was more substantial. For 
example, the ecotrophic efficiencies for mesozooplankton and the benthos were 
increased. These results suggest that far from a sensitive indicator of ecosystem 
performance, overall trophic transfer efficiency may be a biologically-regulated 
property that is more likely to be conserved, even across systems in which the 
structure of trophic flow differs. 
Alternate Pathways of Trophic Transfer 
In view of the apparent inconclusiveness of average trophic efficiency as an 
indicator of ecosystem perfo,mance and degradation, the trophic flow networks were 
used to contrast differences in the major pathways of trophic transfer among the 
regions. For such flows, the efficiency of transfer to a particular endpoint can be 
examined via the gross carbon transfer efficiency, which is simply the ratio of 
production at a particular endpoint to gross carbon input (i.e. gross production plus 
detrital imports). Here, substantial differences were found. 
The most dramatic difference was the prevalence of bacterial processing of 
organic matter in the mid Bay. Bacte1ial respiration (i.e. dissipation by bacteria) in the 
mid Bay was 4-fold greater than in the upper Bay and 27% greater than in the lower 
Bay. Bacteria accounted for 76% of consumption at trophic level II in the mid Bay, 
compared to 33% and 61 % in upper and lower Bay, respectively. Due to the high net 
growth efficiency of bacteria compared to most herbivores and detritivores, trophic 
transfer efficiency at TL 2 was 30% in the mid Bay compared to 20% elsewhere (Fig. 
6-7). However,  this did not translate to increased efficiency beyond TL 2 due to 
246 
-----------------------------~ 
inefficient transfer to mesozooplankton via the heterotrophic flagellates and 
microphagous ciliates (Table 6-5). Despite the enormous bacterial processing of 
organic matter in the mid Bay, the mesozooplankton there were estimated, using the 
total dependency matrix (Szyrmer and Ulanowicz 1987), to obtain only 12% of their 
diet via the microbial food web. The same analytical procedure reveals that the top 
predators such as striped bass , bluefish and weakfish obtained about 9% of their diets 
by pathways involving DOC and bacteria. While these values may seem large if one 
is accustomed to thinking of the microbial loop as primarily a "sink," they still 
represent an enormous amount of energetic dissipation . Moreover, energetic 
dissipation via the microbial loop occurs as high as the fourth trophic level, explaining 
why low average TTE was not a symptom of eutrophication in the mid Bay. As a 
final note, despite the smaller contribution of bacteria system activity in the upper and 
lower Bay, their contribution to upper trophic levels was similar or greater. In the 
upper Bay, pathways involving DOC accounted for 15% of carbon flow to 
mesozooplankton, more than in the mid Bay, while 12% of carbon flow to 
mesozooplankton in the lower Bay passed through DOC. 
Trophic flow through the gelatinous zooplankton is another feature that, 
though not unique to the mid Bay, was most pronounced there and has been 
hypothesized to divert organic matter from pathways leading to fisheries production 
(Berdnikov et al. 1999). Ctenophores were the second largest consumer of 
mesozooplankton in the mid Bay, and their effect on their competitors, predators and 
prey has already been discussed. The relative rarity of ctenophores in the upper Bay 
most likely reflects a salinity effect, rather than a eutrophication effect. Because 
247 
ctenophore abundance in the lower Bay was 85% of the mid Bay abundance, their 
trophic impact there was estimated to be comparable. Thus, while it is clear that 
ctenophores divert carbon flow from bay anchovy and, to a lesser degree, from their 
predators , it is not clear that ctenophore abundance is enhanced significantly by 
hypoxia. Moreover, since mesozooplankton production was not fully consumed 
(Table 6-5) and Klebasco (1991) has noted that bay anchovy were not usually food 
limited, it is difficult to envision how the activity of ctenophores, at their average 
abundance, could substantially limit fish production . However, blooms of ctenophores 
have been identified around the world as a possible consequence of eutrophication. 
Their increase may reflect their enormous capacity to increase consumption when prey 
is very abundant, while their absence in less eutrophic environments could reflect their 
high prey density requirements (>8 r1, Kremer and Reeve 1989). Chesapeake Bay 
may have been spared the extreme Mnemiopsis abundances (300 m-3) that have 
occurred in the Black Sea (Zaitsev 1992) because of the presence of sea nettles. 
~on Flow Through The Benthos 
The most dramatic difference between the regional food webs is in the role of 
the macrobenthos. Compared to the mid Bay, the upper and lower Bay macrobenthos 
consumed 30-fold and 10-fold more organic matter, respectively. Despite lower rates 
of primary production in these regions compared to the mid Bay, this level of 
rnacrobenthic production was easily accommodated and was in fact comparable to 
simi larly productive coastal systems elsewhere (Herman et al. 1999, Chapter 4). 
Except for menhaden, demersal feeding species contribute the bulk of fisheries 
248 
landings from Chesapeake Bay during summer (Table 6-8). In the upper Bay, blue 
crab was the largest fishery and along with American eel and catfish accounted for 
67
% of summer landings. Spot and croaker also contributed to fish production in the 
mid Bay. Along with the other demersal feeding species, they contributed 69% of mid 
Bay landings in summer. In the lower Bay, demersal feeding fish accounted for 78% 
of landings other than menhaden. The role of the benthos in supporting these 
demersal feeders is clear. For example, 78% of the diet of blue crab in the upper Bay 
passes along pathways involving the deposit and suspension feeding macrobenthos 
(Table 6-6a). Similar fractions apply for the other demersal feeders. Hartman and 
Brandt (1995a) noted that the major piscivores sustai n much of their growth during 
fa ll , when they are voracious feeders on planktivores, mainly menhaden. In this 
regard, the major piscivores depend mainly on plankton. However, their summer diets 
include substantial fractions of spot and croaker (Hartman and Brandt 1995a), giving 
them a strong indirect dependence on the benthos. Their dependence on the benthos is 
especially apparent in June-July, before their diets begin to switch to more menhaden 
(Bartman and Brandt 1995a). As a result, 29%, 50% and 34% of trophic flows to 
striped bass , bluefi sh, and weakfish, respectively, passed through the macrobenthos. 
Given the substantial dependence of fisheries on the highly degraded benthic 
community, it was surprising that the gross carbon transfer efficiency to fish in the 
mict Bay was only marginally lower there (0.019) compared to the other regions (0.047 
anct 0.028, Table 6-5). The small size of this difference reflects at least three factors. 
First, the Bay-wide abundance of bay anchovy and menhaden provides a link to the 
nekton that does not involve the benthos. Secondly, the population dynamics of 
249 
demersal feeders may not depend on the mid Bay environment because individuals 
migrate among regions as well as offshore. Finally, the diet of the demersal feeders is 
supported by the mid Bay benthos, but not with significant unused production. In fact, 
to maintain ecotrophic efficiency below unity, it was necessary to account for the 
decline in macrobenthic biomass during summer, a decline which was not evident 
outside the mid Bay region (Chesapeake Bay Benthic Monitoring Program Data). An 
increase in the abundance of demersal feeders, perhaps due to a particularly successful 
recruitment, would likely result in food limitation. Demersal feeders have already 
been estimated to be more abundant outside the mid Bay than within (Table 6-4). 
Based on the current degree of utilization of benthic production, one may infer that 
this pattern would be exacerbated to accommodate the demands of the diet if total 
biomass of demersal feeders increased. 
The observation that degraded benthos may be a trophic bottleneck in the mid 
Bay does not hold elsewhere. In the upper Bay, ecotrophic efficiency for the deposit 
feeding benthos was only 34% (Table 6-5). Only 8% of suspension feeder production 
was consumed by predators, although some fraction may be invulnerable to predation . 
The same was true for the lower Bay. In order to exploit the lower Bay macrobenthos 
fully in a hypothetical scenario, it was necessary to increase the biomass of demersal 
feeding fish (assuming growth and energetics similar to spot and croaker) by 86%. 
Some of this additional fish biomass may be present, but was overlooked. 
Nonetheless, these results seem to suggest that while fish production in the mid Bay is 
limited by poor habitat quality and degraded macrobenthic communities, overfishing 
is the more likely culp1it elsewhere in the Bay. 
250 
A&stored Middle Chesapeake Bay 
In restoring an ecosystem, it is both informative and motivating to examine 
What the restored system might "look" like. In a limited way, we know exactly. For 
example, a series of aerial photographs of the area around Solomons Island MD 
' ' 
dating as far back as the 1930's shows submerged aquatic vegetation (SA V) growing 
in an extensive bed out to several meters depth. These compelling photographs 
suggest that not long ago Chesapeake Bay was probably very different from the 
present. Hints of what the Bay was like at the time Europeans first arrived are 
provided by historical accounts, but it was assumed here that the so-called "John 
Smith" Bay was not only impossible to imagine with any certainty, but also irrelevant 
from the perspective of restoration. Rather, an attempt was made to describe the 
carbon flow networks for the Bay based on certain documented features of the Bay in 
the 1950s and early 1960's (i.e. SA V, algal biomass, hypoxia), but otherwise looking 
forward to ask not what the Bay was Jike, but what it might become. The questions 
are related, but the latter is both more relevant and, fortunately, less challenging. 
The parameterization of the restored Bay network began with the the historical 
distribution of SA v which has been infen-ed from pictures and accounts (e.g. Orth 
' 
and Moore 1987). The true extent of historical SA V coverage is not known, but a 
conservative estimate is that at some time in past 50-years, SA V probably extended to 
2 m depth. This encompasses about 10% of the mid Bay bottom area (Cronin and 
Pritchard 1975) and 7_foJd more area than present SA V coverage. Since SA V requires 
about 20% of incident light to survive (Batiuk et al. 1992), one may infer that water 
251 
clarity in the Bay was once much greater than at present. To obtain 20% light at 2 rn 
depth requires that light attenuation (kd) in nearshore waters be 0.8 d-1, or Jess. Since 
kc1 in nearshore waters tends to be about 0.4 m- 1 greater than in open water (W. M. 
Kemp, personal communication), kd for open water might have been 0.4 m-1, about 
SO% of the present mid Bay average (Harding et al. 2001). Given kd=0.4, the 
maximum depth of the euphotic zone (1 % of Io) wou ld be 11.5 m, sufficient to 
illu ? 
minate 60% of the mid Bay bottom. Perhaps as important, the pycnocline in much 
of the Bay would be within the euphotic zone, possibly altering the vertical 
di stribution of phytoplankton (and oxygen) in the water column. 
Historical data show that lower surface chlorophyll-a concentrations accounted 
for part of the apparently greater water clarity (Harding and Perry 1997). In the 1960's, 
the seasonal pattern of chlorophyll-a in the mid Bay featured a substantial spring 
Phytoplankton bloom (as today), followed by summer chlorophyll-a reduced by about 
50% compared to the present (Harding and Perry 1997). Phytoplankton production 
Was probably also lower, although this need not be the case. As an initial suggestion, 
Phytoplankton production might have been lower by 50% as well. An effort to apply a 
more sophisticated approach to estimating historical rates of primary production is 
Planned (L. Harding, personal communication). 
At the same time that historical accounts imply that Bay waters were clearer 
and that SA y beds were larger, fisheries outputs were evidently as large or larger than 
Present (Houde et al. 1999). Fisheries outputs included oysters and soft-shelled clams, 
the landings of which have declined to relative insignificance in recent years (NOAA 
landings d ata, a Is o see A ppen dI' x ) . The optimist might imagine a restored Bay 
252 
teeming with fish, similar to the exuberant descriptions of the early European arrivals. 
Realistically, this cannot be assured since it is difficult to separate pollution impacts 
from the effects of fishing and overfishing (Boreman 1997). Controversially, Jackson 
et al. (2001) noted that the first human impacts on ecosystems, and some of the most 
devastating, have been due to ove,fishing. The approach of this study was to examine 
how ecological changes associated with reversal of eutrophication might affect the 
prey resource and what possible fish production that might support. The question of 
Whether the fish community would actually fully exploit the resource was left 
unanswered. 
Specification of a trophic flow network for a "restored" mid Bay began with 
the estimated 50% reduction in phytoplankton biomass and production, noted above, 
and the expansion of SA V coverage to the 2 m depth contour. As the area of SA v 
expands to the 2 m depth contour, it would cover about 10% of the bottom. Therefore, 
2 
the average biomass would be 13500 mgC m- and average production would be 122 
mgc m-2 d-', a 7-fold increase from the present. While SAV may not grow in some 
areas due to factors other than light (Koch 2001, see also Orth and Moore 1984), this 
Was neglected, assuming that SA V also grows to greater depth in some more favorable 
areas. 
1 
A change in light attenuation, as discussed above, from 0.8 m- to 0.4 m-1 
Would increase the light-dependent production rate of microphytobenthos. Asssuming 
Unchanged light-saturated gross production, the average production would increase 
1
With the light field (Fig. 6-8) to 447 mgC m-2 d- , a 91 % increase. Because benthic 
microalgae grow as epiphytes on SA V, the vegetated area was not assumed to lack 
253 
microphytobenthos. Th us, tota I gross pn?m ary prod  uc t?1 0n m? th  e "r estored" mid Bay 
Would be 2212 mgC m?2 d-1, 63% of the present day estimate. Benthic microalgal 
production would account for 20% of production, up from 6% in the modern mid Bay. 
If the level of phytoplankton production approximates levels in the 1950's to 
60 's, one may suspect that hypoxia will be reduced as well (Chapter 3) allowing the 
benthos to increase to levels consistent with comparative ecological relationships 
(Herman et al. 1999, Chapter 5). Based on organic inputs, total macrobenthic 
biomass should be -20 g AFDW m?2, equivalent to 10,000 mgC m?2 or a 7-fold 
increase over the present. In most of the modern Bay, suspension-feeding biomass is 
2-fold greater than deposit feeding biomass and it was assumed that this ratio would 
remain unchanged. 
Without further changes in the food web, there was clearly insufficient organic 
matter to support consumption. Mass balance constraints, therefore, dictate that other 
flows would likely change as the Bay is restored. Again these changes were made in 
accordance with the predetermined procedure (see Methods). In the lower 
Chesapeake Bay, free-living bacteria consume 89% of net phytoplankton production, 
lower than the modern mid-Bay, but still higher than the average fraction consumed 
by bacteria in marine systems (Cole et al. 1988). Assuming that this fraction would 
apply to the restored mid Bay in summer, but that bacterial growth effiiciency also 
decreases from 0.5 to 0.4 (del Giorgio and Cole 1998), bacterial consumption of DOC 
2 1
Would decrease to 50% of the present level, or 1425 mgC m? d? ? Decreased free-
bacterial production requires commensurate reduction in the production of 
heterotrophic flagellates, the primary bactivores. Assuming that ecotrophic efficiency 
254 
.,, 
for free-bacteria is 90%, one can compute that the biomass of heterotrophic flagellates 
must decrease to 55 mgC m?2 d-1? Although production by heterotrophic flagellates 
must decrease, the production remains sufficient to support the consumption by 
ciliates. 
Decreased phytoplankton production also limits herbivory (i.e. notjust 
detritivory). If one assumes that menhaden consumption is unchanged and that 
suspension-feeding increases as stipulated, consumption by other herbivores must be 
decreased. Meroplankton might be expected to increase in abundance, rather than 
decrease, due to the increased abundance of macrobenthos (Table 6-9). Of the two 
remaining planktonic herbivores, it was assumed that ciliate biomass would decrease. 
Accordingly the mesozoop lankton must shift their diet slightly from ciliates to 
phytoplankton. These adjustments achieve balanced carbon flows within the plankton. 
While the exact changes are somewhat speculative, it is clear that a decrease in 
phytoplankton production and the associated decrease in bactetial productivity require 
a cascade of reductions in the role of the microplankton (Cole et al. 1988). 
The benthos can also be expected to change as a result of decreased ptimary 
production , as is commonly seen on an interannual basis in response to 1iver flow 
variations (Boynton and Kemp 2000). Nixon (1982) observed that across systems 
with a wide range of total organic input, total benthic mineralization consumed a 
nearly constant proportion of the input, about 24%. Nixon's (1982) relationship 
approximates the modern condition the upper and lower Chesapeake Bay with 
remarkable, perhaps serendipitous, precision (within 5% enor). However, the benthos 
in the present-day mid Bay consumes about 25% less than predicted (Table 6-5, Table 
255 
6-7). This may reflect the minimal suspension-feeding activity in the mid Bay 
compared to other regions of the Bay, perhaps due to anoxia and/or the near 
elimination of oysters by overfishing and parasitic diseases. In a restored mid Bay, 
respiration due to meiofauna and macrobenthos (based on Herman et al. 1999 model) 
was estimated to total 254 mgC m-2 d-1, slightly less than half of the total metabolism 
predicted by the Nixon (1982) model. Thus, benthic bacterial metabolism might be as 
low as 277 mgC m-2 d-1? Although this is 35% of the present rate, it may not be 
unreasonably low, as bacterial metabolism in the modem upper Bay was even less 
than this reduced value. 
Remarkably, no additional changes were needed to b1ing the carbon budget of 
the mid Bay into approximate balance, despite the significant decrease in 
phytoplankton production (Table 6-9). The exception was that the supply of dissolved 
organic carbon was deficient by 12%, even as some sediment POC was exported by 
burial (see later discussion). This does not imply that no other changes would or could 
occur. However, this exercise showed how it is possible to support the upper trophic 
levels of the mid Chesapeake Bay food web with a significantly decreased level of 
primary production, without departing from widely observed relationships among the 
components of the lower food web. Moreover, decreased primary production was 
clearly commensurable with even a dramatic increase in the biomass and production 
of macrobenthos. 
The increased macrobenthos production provides an increased prey resource 
for demersal fish. Since the macrobenthic biomass increased 7-fold, the prey resource 
and biomass that can be supported in the mid Bay increases commensurately. As 
256 
previously noted, it is di fficult to imagine what form such an increase might take, or if 
it would occur at all. however, since it is contrary to the propensities of ecosystems to 
leave a substanti al and avai lable prey resource unutilized, production of demersal 
feed ing fish such as spot and croaker could increase. In addition, or perhaps 
alternative ly , production of non-targeted competitors such as hogchokers or skates and 
rays (e.g. cownose ray, Rhinoptera bonasus), already present in the Bay, could 
increase. 
Because the "restored" Bay is a hypothetical construct, caution is needed in 
interpreting analyses of the network. Without resorting to complex computations, 
some major conclusions can be reached simply because it has been shown that the 
network can ex ist in balance, without resorting to unreasonable assumptions. For 
example, if fish production does not decrease, but primary production decreases, then 
the gross transfer efficiency to fi sh is greater. If benthic macrofauna increase in 
biomass and production, at the expense of plankton and benthic bacteria, then the role 
of bacteria is reduced. Oddly, but consistent with regional differences noted before, 
decreased bacte1ial production in the water column did not decrease the relative 
importance of the microbial loop as a link to upper trophic levels . The fraction of 
carbon flow passing through DOC to the mesozooplankton increased from 0.12 in the 
modern mid Bay, to 0.15 in the restored Bay. This increase reflects an assumed shift 
in the diet of ciliates, due to decreased phytoplankton abundance, toward more 
consumption of heterotrophic flagell ates. Thus, the accuracy of conclusions regarding 
the role of the microbial loop hinges on the validity of that assumption. 
257 
The average trophic transfer efficiency, computed from the "Lindeman Spine" 
for the restored mid Bay, was greater than for the modern mid Bay (Fig. 6-9). For 
example, average efficiency for trophic levels (TL) 1-5 was 27% in the modern mid 
Bay and 32% in the restored mid Bay (Fig. 6-9). The increased efficiency was a 
surprising result considering the apparent lack of differences in average trophic 
transfer efficiency among the upper, mid and lower Bay. The differences could be 
traced mainly to a few changes in the flows among the plankton. The trophic transfer 
efficiency for TL 2 was lower in the restored Bay, reflecting the decreased activity of 
the highly efficient bacteria. On the other hand, the efficiencies for TL 2-5 increased 
due to increased ecotrophic efficiencies for heterotrophic microflagellates and ciliates 
(Table 6-9). These changes were responsible for both the increased dependency of the 
mesozooplankton on DOC and the increased average efficiency for the canonical food 
chain (Fig. 6-9). Nonetheless, because of the relative uncertainty of the trophic 
transfers among the microplankton, other measures of the response to eutrophication 
may be more compelling. 
As with the inter-regional comparisons, the fraction of primary production 
passed to desired endpoints in the food web revealed the most substantial differences, 
as qualitatively described above. The matrix of total contribution coefficiency, a basic 
output of NETWRK, describes this quantitatively. In the modern mid Bay food web, 
0.125% of the production by net phytoplankton was passed to st1iped bass, while in 
the restored network, 0.237% eventually reached striped bass (Fig. 6-10). Although 
picoplankton production reached striped bass with lower efficiency than for net 
phytoplankton, the efficiency of cont1ibution to st1iped bass increased similarly in the 
258 
restored Bay network. Conversely, benthic bacte1ia (for example), consumed a 
smaller fraction of production from all sources. Interestingly, the fraction of SA v 
production reaching striped bass increased, while the fraction of mjcrophytobenthos 
reaching striped bass decreased. This reflects increased meiobenthic consumption of 
microphytobenthos in the restored Bay, which effectively added an intermediate 
trophic transfer slightly and decreased the overall efficiency. Despite decreased 
efficiency, the total flow from microphytobenthos to striped bass along all paths 
increased from 24 to 39 mgC summe(1? 
This analysis shows that reasonable adjustments to the modern food web of the 
mid Chesapeake Bay permit sustained or even increased fish production, even as 
Phytoplankton production decreases by 50%. Although the network description for 
the restored Bay was not based entirely on direct observation, it was based on sound 
relationships which have been observed to hold for coastal ecosystems around the 
World. The networks also conform to the mass balance constraints imposed by a 
network analysis approach (e.g. Nixon 1982, Cole et al. 1988). In retrospect, it mjght 
have been far more challenging to correctly envision what might happen to a system 
like the restored Bay if phytoplankton production were doubled and the system 
departed from the norm. Indeed, the food web of the modern mid Bay, in devoting 
such a large fraction of production to bacteria, has become extraordinary among its 
peers. 
259 
[nformational Indicators of Eutrophication 
Interest has been growing in recent years of developing objective indicators of 
ecosystem status and health . Toward that end, Ulanowicz (1997) defined 
eutrophication in terms of a series of indices, each based on info1mation theory, which 
describe aspects of a trophic flow network in its entirety (Ulanowicz 1986). Total 
system throughput (TST) is the sum of all the flows and effectively measures 
ecosystem size. Average mutual information (AMI) defines the "determinacy" of the 
flows . To use an analogy from baseball, the flow of base runners has a high AMI. In 
contrast, should the runners choose to run among the bases not in order, but at random, 
AMI would be low. The product of AMI and TST defines "ascendency." (Ulanowicz 
1986, 1997). Ulanowicz (1997) defined eutrophication in terms of these information 
indices: an increase in system ascendancy, due to a rise in total systems throughput 
that more than compensates for a concomitant fall in average mutual information. 
Analysis of the four networks constructed for Chesapeake Bay provided a 
natural test for this definition. TST was indeed found to be much greater in the 
modern mid Bay (1.5 106 mgC m-2 summe( 1) compared to either the restored mid Bay 
(l.l 106 mgC m-2 summer- 1) or other regions of the Bay. Similarly, AMI was lower in 
the modern mjd Bay (2.06) compared to the restored network (2.08). While the index 
technically passes the test, the difference is not compelling. The similarity of the 
estimates may reflect the largely unchanged topology of the two networks, which may 
be biased by the perceptions, knowledge or even goals of the investigator. For 
example, Monaco and Ulanowicz (1997) compared Chesapeake Bay (annual, not 
summer), Delaware Bay and Na1Tagansett Bay, finding the AMI for these three 
260 
systems was 1.2, 1.3 and 1.3, respectively. Johnson et al. (1995) compared coral 
dominated and algae dominated regions of the Great Banier Reef, Australia, observing 
th
at AMI was l.877 and 1.878. Even when systems as different as a coral-dominated 
vs. alga l-dominated reefs and Chesapeake Bay vs. Narragansett Bay are included 
' 
most differences in AMI can be related to the investigator. 
Given the constraints of investigator bias, connectance may be a more useful 
indicator of change. Topologically connectance refers to the average number of 
connections involving a node. Social network scientists refer to the number of nodes 
flowin g into a node as the "in degree" and the number of emerging connections as the 
" 
out degree." All the Chesapeake Bay networks were similar, having effective 
topological connectance of -4 connections per node. Effective connectance decreases 
relative to topological connectance as a smaller number of flows come to dominate 
th roughput. "Food web connectance," which examines only flows involving living 
compartments, varied in Chesapeake Bay from 1.89 in the mjd Bay to 2.34 in the 
restored mid Bay. The upper Bay (2.26) and lower Bay (2.31) food webs varied 
between the two mid Bay scenarios. The differences most likely reflect the greater 
degree of balance among the different sources of organic matter as well as the fates. 
Whereas 68% of throughput involving living nodes in the modern mid Bay could be 
attributed to phytoplankton and bacteria, this figure was reduced to 53% in the 
restored mid Bay. Although connectance does not reveal changes in TST, which are 
clearly ch aracten.s t1.c  o f eutrop h1' c ati?on , it may be more sensitive than AMI to 
quantitat1? ve c h anges m.  fl ows w h e n the network topology is fixed, by investigator bias 
or Otherwise. 
261 
? 
Research Issues and Needs 
One of the often noted ancillary benefits of synthesis research is that it may 
reveal weaknes ses or m? cons1? stenc1?e s m?  t he  ba se o f sc1?e nt1? f?1 c k nowledge, suggesting 
directio c 
ns ,or future research. In assembling the trophic networks for Chesapeake 
Bay, quantifying the role of the free-living bacteria and especially the sources of the 
Doc that they require was a constant source of consternation. Consistent wi th the 
nd
fi ings of Baines and Pace (1991), extracellular release of DOC by phytoplankton 
could n . . 
ot reasonably support the DOC requirements of the bactena. The advective 
flux of DO . . 
C through the estuary was very large, and recent research indicates that this 
Pool is not entireely recalcitrant (Raymond and Bauer 2000). However, advective 
inputs cannot reasonably provide the DOC required by the mainstem food web 
becaus eth  e estuary appears to export as much DOC, or more, than 1? t i?m ports 
(Appendix A). To account for the required input to bacteria, DOC was assumed to be 
released as part of the egesta from almost every trop h1 ? c r?n teract1?o n. Al t ho ugh this 
rnay in fact be reasonable the sources and fate of detritus within the network was a 
' 
lllajor source of unceitainty. While organic matter cycling was not a major focus of 
th
is study, the nature of cycling was computed by NETWRK and was found here to be 
Profoundly different than seggested by Baird and Ulanowiocz (1989). This difference 
rnay have resulted primarily from the fate of detritus, including flows among the 
detrital pools. Thus, continued study of bacterial metabolism and the sources and fate 
of detritus, in its various forms, would be warranted and helpful. 
262 
Trophic interactions within the microplankton were another source of 
significant uncertainty. While studies addressing trophic interactions involving the 
microplankton of Chesapeake Bay were not lacking, the emergent message was that 
many trophic interactions are possible and that, moreover, heretofore unappreciated 
interactions may not only occur, they may be important. These studies would be more 
useful for constructing trophic flow models if, in addition to addressing novel 
interactions, they also addressed where and under what circumstances different types 
of interactions were more or less prevalent. Such results would be useful for future 
food web models. 
The data available to characterize the fish community, though more plentiful 
than expected, were often frustratingly disconnected, perhaps reflecting the multiple 
objectives of studies addressing fisheries. Frustratingly, it often appeared that the data 
needed for network analysis were present somewhere, perhaps collected for tactical 
fisheries management, but were never organized sufficiently to be utilized for research 
purposes. Therefore, enhanced integration of fisheries data into a broadly accessible 
databases would be very valuable. Fish may also be valuable integrators of ecosystem 
function. This study relied on many published studies of fish growth and diet, often 
accessed via FishBase (Froese and Pauly 2001). While extremely valuable, local 
studies would be valuable in assessing how fish utilize the resources within the 
ecosystem of interest. Given the difficulty of estimating absolute abundance of fish, 
diet requirements of better known and exploited predators proved to be a useful means 
of estimating the abundances of their prey. 
263 
Fina ll y, additional work is needed to address seasonal and interannual 
variation. This study examined a long term average for summer only. It would be 
very valuable to expand this analysis to include coupled seasonal and regional models. 
Seasonal compari sons should be made simil ar to the analyses of Baird and Ulanowicz 
(1989), but the models should explicitly account for seasonal changes in biomass such 
that biomass is at steady state only on an annual basis . Significant interannual 
variations are known to occur in primary production as well as fish biomass and 
production . This food web described by this study is composi te of many years, 
providing a useful baseline scenario based on the best data available. However, 
additional and possibly important insights could probably be gained by addressing 
alternative food web configurations associated with natural variability. 
Literature Cited 
Baines, S. B. and M . L. Pace. 1991. The production of di ssolved organic matter by 
phytoplankton and its importance to bacteria: Patterns across marine and 
freshwater systems. Limnol. Oceanogr. 36(6): 1078-1090. 
Baird, D. and R. E. Ulanowicz. 1989. The seasonal dynamics of the Chesapeake Bay 
Ecosystem. Ecol. Monogr. 59(4): 329-364. 
Banse, K. and S. Mosher. 1980. Adult body mass and annual production/biomass 
relationships of field populations. Ecol. Monogr. 50(3): 355-379. 
Barger, L.E. 1990. Age and growth of bluefish Pomatomus saltatrix from the northern 
Gulf of Mexico and U.S. south Atlantic coast. Fish. Bull., U.S. 88:805-809. 
Batiuk, R.R. Orth, K. Moore, W . Dennison, J. Stevenson, L. Staver, V. Carter, N. 
Rybicki , R. Hickman, S. Kollar, S. Bieber, and P. Heasly. 1992. Chesapeake 
Bay Submerged Aquatic Vegetation Habitat Requirements and Restoration 
Targets: A Technical Synthesis. EPA Chesapeake Bay Program, Annapolis, 
MD. 
264 
Berdnikov, S. V., V. V. Selyutin, V. V. Vasilchenko and J . F. Caddy. 1999. 
Trophodynamic model of the Black and Azov Sea pelagic ecosystem: 
consequences of the comb jelly, Mnemiopsis leidyi, invasion. Fish. Res. 
42(3): 261-289. 
Berggreen, U. , B. Hansen and T. Ki?rboe. 1988. Food size spectra, ingestion and 
growth of the copepod Acartia tonsa during development: implications for 
determination of copepod production. Mar. Biol. 99: 341-352. 
Bogdan, K. G. and J. J. Gilbert. 1982. Seasonal patterns of feeding by natural 
populations of Keratella, Polyarthra, and Bosmina: Clearance rates, 
selectivities and contributions to community grazing. Limnol. Oceanogr. 27: 
918-934. 
Bogdan, K. G. and J . J. Gilbert. 1984. Body size and food size in freshwater 
zooplankton. Proceedings of the National Academy of Sciences (USA) 81: 
6427-6431. 
Boreman, J . 1997. Methods for comparing the impacts of pollution and fishing on fish 
populations. Trans. Am. Fish. Soc. 126: 506-513. 
B?rsheim, K. Y. and G. Bratbak. 1987. Cell volume to cell carbon conversion factors 
for a bacterivorous Monas sp. emiched from seawater. Mar Ecol. Prag. Ser. 
36: 171-175. 
Boynton, W.R. and W. M . Kemp. 2000. Influence of river flow and nutrient loads on 
selected ecosystem processes. A synthesis of Chesapeake Bay data. pp. 269-
298 In: J.E. Hobbie (ed.). Estuarine Science: A Synthetic Approach to 
Research and Practice. Island Press, Washington, DC. 
Boynton, W.R., R. M . Stankelis, F. M. Rohland, J. D. Hagy and J.M. Frank. 1999. 
Maryland Chesapeake Bay Water Quality Monitoring Program. Ecosystem 
Processes Component. Level One Interpretive Repo1t No. 16. Chesapeake 
Biological Laboratory. University of Maryland Center for Environmental 
Science. Solomons, MD. 20688-0038. Ref. No. [UMCES]CBL 99-0070. 
Boynton, W . R., personal communication. UMCES Chesapeake Biological 
Laboratory, Solomons, MD. 
Butowski, N. H. 1994. Chesapeake Bay Oyster Fishery Management Plan. 
Agreement Commitment Report Revision. US EPA Chesapeake Bay Program, 
Annapolis, MD. 
Cebrian, J. 1999. Patterns in the fate of production in plant communities. Am. Nat . 
154(4): 449-468. 
265 
Chesso~, J. !983. The estimation and analysis of preference and its relationship to 
foraging models. Ecology 64(5): 1297-1304. 
Choi, J. W., B. F. Shen-, and E. B. Sherr. 1999. Dead or alive? A large fraction of 
ETS-inactive marine bacterioplankton cells, as assessed by reduction of CTC 
can become ETS-active with incubation and substrate addition. Aquat. ' 
Microb. Ecol. 18 (2): 105-115. 
Ch? 
nstensen , V. and D. Pauly. 1992. ECO~ATH II - a software for balancing steady-
state ecosystem models and calculating network characteristics. Ecol. 
Modelling. 61: 169-185. 
Chiistensen, V. and D. Pauly. 1993. Flow characteristics of a~uatic ecosystems, p. 
338-352. In V. Christensen and D. Pauly (eds.) Trophic models of aquatic 
ecosystems. ICLARM Conference Proceedings No. 26. 
Ch ? 
nstensen, V., C.J. Walters, and D. Pauly. 2000. Ecopath with Ecosim: A user's 
guide. October 2000 edition. Fisheries Centre, Univ. of British Columbia 
Vancouver, Canada and International Center for Living and Aquatic Reso~rces 
Management,, Penang, Malaysia. 
Cole, J. J. S. Findlay, and M. L. Pace. 1988. Bacterial production in fresh and 
saltwater ecosystems: a cross-system overview. Mar. Ecol. Prog. Ser. 43: 1_ 
10. 
Cowan, J. L. and W.R. Boynton. 1996. Sediment-water oxygen and nutrient 
exchanges along the longitudinal axis of Chesapeake Bay: Seasonal patterns, 
controlling factors and ecological significance. Estuaries 19(3):562-580. 
Cronin, W. B. and D . w. Pritchard. 1975. Additional statistics on the dimensions of 
Chesapeake Bay and its t1ibutaries: Cross section widths and segment volumes 
per meter depth. Chesapeake Bay Institute. Reference 75-3. Special Report 
42. 
de! Giorgio, P. A. and J. J. Cole. Bacterial growth efficiency in natural aquatic 
systems. Ann. Rev. Ecol. Syst. 29: 503-41. 
Derrick, P.A. and v. s. Kennedy. 1997. Prey selection by the hogchoker, Trinectes 
maculatus (Pisces: Soleidae), along summer salinity gradients in Chesapeake 
Bay, USA. Mar. Biol. 129: 699-711. 
Detwyler, R. and E. D. Houde. 1970. Food selection by laboratory-rearer larvae of 
the scaled sardine Harengula pensacolae (Pisces: Clupeidae) and the bay 
anchovy Anchoa mitchilli (Pisces: Engraulidae). Mar. Biol. 7: 214-222. 
266 
D? 
iaz, R. J., R. J. Neubauer, L. C. Schaffner, L. Pihl , and S. P. Baden. 1992. 
Co~ti~uous monitoring of dissolved oxygen in an estuary experiencing 
penod,c hypoxia and the effect of hypoxia on macrobenthos and fish. Sci. Tot. 
Env. (Supplement 1992): 1055-1067. 
Dolan, J. R. 199 la. Guilds of ciliate microzooplankton in the Chesapeake Bay. Est. 
Coast. Shelf Sci. 33: 137-152. 
Dolan, J. R. 1991b. Ciliate growth rates and consumption by copepods. Mar. Biol. 
111: 303-309. 
Dolan, J. R. and D. W. Coats. 1990. Seasonal abundances of planktonic ciliates and 
microflagellates in mesohaline Chesapeake Bay waters. Est. Coast. Shelf Sci. 
31: 157-175. 
Dolan, J. R. and C. L. Gallegos. 1991. Trophic coupling of rotifers, microflagellates, 
and bacteria during fall months in the Rhode River estuary. Mar. Ecol. Prog. 
Ser. 77: 147-156. 
Dovel , W. L., J. A. Mihursky and A. J. McErlean. 1969. Life history aspects of the 
hogchoker, Trinectes maculatus, in the Patuxent River estuary, Maryland. 
Ches. Sci. 10(2): 104-119. 
Duarte, C. M. and C. L. Chiscano. 1999. Seagrass biomass and production: a 
reassessment. Aquat. Bot. 65: 159-174. 
Ducklow, H. W. and F. Shiah. 1993. Bacterial production in estuaries. pp 261 -287. 
In: Ford, T. E. (ed.). Aquatic Microbiology: An Ecological Approach. 
Blackwell Scientific Publications, Cambiidge, MA. 518 pp. 
Durbin, E.G. and A.G. Durbin. 1981. Asssimilation efficiency and nitrogen 
excretion of a filter-feeding planktivore, the Atlantic menhaden, Brevoortia 
tyrannus (Pisces: Clupeidae). Fish. Bull. 79(4): 601-616. 
Ebersole, E. L. and v. s. Kennedy. 1994. Size selection of Atlantic Rangia clams, 
Rangia cuneata, by blue crabs, Callinectes sapidus. Estuaries 17(3):668-673. 
EPA 1989. Chesapeake Bay Alosid Management Plan. US EPA Chesapeake Bay 
Program, Annapolis, MD. 
EPA 1990a. Chesapeake Bay Bluefish Fishery Management Plan. Chesapeake Bay 
Program Agreement Commitment Report. US EPA Chesapeake Bay Program, 
Annapolis, MD. 
EPA 1990b. Chesapeake Bay Weakfish and Spotted Seatr?ut Fishery Management 
Plan. Chesapeake Bay Program Agreement Commitment Report. Chesapeake 
Bay Program, Annapolis, MD. 
267 
EPA 1991a. Chesapeake Bay Atlantic Croaker and Spot Fi shery Management Plan. 
Chesapeake Bay Program, Annapolis, MD. 
EPA 1991b. Chesapeake Bay American Eel Fi shery Management Plan. Chesapeake 
Bay Program, Annapolis, MD. 
EPA 1995. Chesapeake Bay Striped Bass Fishery Management Plan. Annual 
Progress Report. 1994. EPA Chesapeake Bay Program, Annapoli s, MD. 
EPA 1997. 1997 Chesapeake Bay Blue Crab Fishery Management Plan. June 1997. 
US EPA Chesapeake Bay Program, Annapolis, MD. 
EPA 2000. The 2000 User's Guide to Chesapeake Bay Program Biological and Living 
Resources Monitoring Data. US Environmental Protection Agency, 
Chesapeake Bay Program, Annapolis , MD. 
EPA 2000b. Chesapeake 2000, A Watershed Partnership. EPA Chesapeake Bay 
Program, Annapolis, MD. 
Fenchel, T . 1978. The ecology of micro- and meiobenthos. Ann. Rev. Ecol. Syst. 9: 
99-121. 
Fenchel, T. 1982. Ecology of heterotrophic microflagellates. II. Bioenergetics and 
growth. Mar Ecol. Prag. Ser. 8: 225-231. 
Field, J . G ., F. Wulff, and K. H. Mann. 1989. The need to analyze ecological 
networks. In: F. Wulff, J. G. Field, and K. H. Mann. (eds.) Network Analysis 
in Marine Ecology. Methods and Applications. Springer-Verlag, New York. 
Froese, R. and D. Pauly (eds.) 2001. FishBase. World Wide Web electronic 
publication. http://www.fishbase.org. 
Flach, E., J. Vanaverbeke and C. Heip. 1999. The meiofauna: macrofauna ratio 
across the continental slope of the Goban Spur (north-east Atlantic). J. Exp. 
Mar. Biol. Ecol. 79(2): 233-241. 
Friedland, K. D. , D. W. Ahrenholz, and J. F. Guthrie. 1989. Influence of plankton on 
dist1ibution patterns of the filter-feeder Brevoortia tyrannus (Pisces: 
Clupeidae). Mar Ecol. Prag. Ser. 54: 1-11. 
Gaedke, U. and D . Straile. 1994. Seasonal changes of the quantitative importance of 
protozoans in a large lake. An ecosystem approach using mass-balanced 
carbon flow diagrams. Mar. Microb. Food Webs. 8(1-2): 163-188. 
Gerlach, S. A. 1971. On the impo1tance of marine meiofauna for benthic 
communities. Oecologia. 6: 176-190. 
268 
Gifford, D. J. and D. A. Caron. 2000. Sampling preservation, enumeration and 
biomass of marine protozooplankton. In: R. Harris, P. Wiebe, J. Lenz, H. R. 
Rune Skjoldal and M. Huntley, (eds.) ICES Zooplankton Methodology 
Manual. Academic Press, New York. 
Barding, L. W. , Jr. and E. S. PeITy. 1997. Long-term increase of phytoplankton 
biomass in Chesapeake Bay, 1950-94. Mar. Ecol. Prag. Ser. 157: 39-52. 
Harding, L. W., Jr. , M. E. Mallonee, and E. S. Perry. 2001. Toward a predictive 
understanding of primary productivity in a temperate, partially-stratified 
estuary. Est. Coast. Shelf Sci. (in press). 
Hargis, W. J ., Jr. and D.S. Haven. 1994. The preca1ious state of the Chesapeake 
public oyster resource. In : Hill, P. and S. Nelson, (eds.). Toward a Sustainable 
Watershed: The Chesapeake Experiment. Proceedings of the 1994 
Chesapeake Research Conference. Chesapeake Research Consortium, 
Publication Number 149. 
Bartman K. J. , and S. B. Brandt. 1995a. Trophic resource partitioning, diets, and 
growth of sympatric estuarine predators. Trans. Am. Fish. Soc. 124 (4): 520-
537. 
Bartman, K. J. ands. B. Brandt. 1995b. Comparative energetics and the development 
of bioenergetics models for sympatric estuarine predators. Can. J. Fish. Aquat. 
Sci. 52: 1647-1666. 
Bartman, K. J. ands. B. Brandt. 1995c. Estimating energy density of fish. Trans. 
Am. Fish. Soc. 124: 347-355. 
Beinbokel, J. F., D. w. Coats, K. W. Henserson, and M.A. Tyler. 1988. 
Reprodoction and secondary production of three species of the rotifer genus 
Synchaeta in the estuarine Potomac River. J. Plank. Res. 10(4): 659-674. 
Heinle, D.R. 1966. Production of a calanoid copepod, Acartia tonsa, in the Patuxent 
River estuary. Ches. Sci. 7(2): 59-74. 
Berman, P. M. J., J. J. Middelburg, J. Van De Koppel and C. H. R. Heip. 1999. 
Ecology of estuarine macrobenthos. Adv. Ecol. Res. . 15(2). 
liines, A.H., A. M. Haddon, and L.A. Wiechert. 1990. Guild structure and foraging 
impact of blue crabs and epibenthic fish in a subestuary of Chesapeake Bay. 
Mar. Ecol. Prag. Ser. 67: 105-126. 
Boenig, J. 1983. Empirical use of longevity data to estimate mortality rates . Fish. 
Bull. 82(1): 899-903. 
269 
Hobbs, C.1:., III, J.P. Halka, R. T. Kerhin, and m. J. Can-on. 1992. Chesapeake Ba 
Sediment Budget. J. Coast. Res. 8(2): 292-300. y 
Holland, A. F., A. T. Shaughnessy, and M. H. Riegel. 1987. Long-term variation in 
mesohaline Chesapeake Bay macrobenthos: Spatial and temporal patterns. 
Estuaries 10(3):227-245. 
Hollis, E. H. 1952. Variations in the feeding habits of the striped bass, Roccus 
saxatilis (Walbaum), in Chesapeake Bay. Bull. Bing. Ocean. Coll. 14: 111-
131. 
Houde, E. D. and E. S. Rutherford. 1993. Recent trends in estuarine fisheries: 
Prediction of fish production and yield. Estuaries: 16(2): 161-176. 
Houde, E. D., the late J. _c. Gamble, S. E. _Dorsey, and J. H. Cowan, Jr. 1994. Diifting 
mesocosms: the influence of gelatinous zooplankton on mortality of bay 
anchovy, Anchoa mitchilli, eggs and yolk-sac larvae. ICES J. Mar. Sci. 51: 
383-394. 
Houde, E. D., S. Jukic-Peladic, S. B. Brandt and S. D. Leach. 1999. Fisheries: Trends 
in catches, abundance and management, pp. 341-366. In: T. C. Malone, A. 
Malej , L. W. Harding, Jr. , N. Smodlaka, and R. E. Turner (eds.) Ecosystems at 
the Land-Sea Margin, Drainage Basin to Coastal Sea. American Geophysical 
Union, Washington, DC. 
Iverson, R. L. 1990. Control of fish production. Limnol. Oceanogr. 35(7): 1593-
1604. 
Jackson J. B. C., M. x. Kirby, W. H. Berger, K. A. Bjorndal, L. W. Botsford, B. J. 
Bourque, R. H. Bradbury, R. Cooke, J. Erlandson, J. A. Estes, T. P. Hughes, S. 
Kidwell, C. B. Lange, H. S. Lenihan, J.M. Pandolfi, C.H. Peterson, R. S. 
Steneck, M. J. Tegner and R.R. Warner. 2001. Historical overfishing and the 
recent collapse of coastal ecosystems. Science. 293 (5530): 629-638. 
Jassby, A. D. and T. Platt. 1976. Mathematical formulation of the relationship 
between photosynthesis and light for phytoplankton. Limnol. Oceanogr. 21: 
540-547. 
Johnson, c., D. Klumpp, J. Field, and R. Bradbury. Carbon flux on coral reefs: 
effects of large shifts in community structure. Mar. Ecol. Prog. Ser. 126: 123-
143. 
Jordan, S. J., G. E. Krant, and c. Judy. 1994. Marylan? Chesape~ke Bay oyster 
recovery areas: Summary of technical information. In: H1JJ, P. ~nd S. Nelson, 
(eds.). Toward a Sustainable Watershed: The Chesapeake Expenment. 
Proceedings of the 1994 Chesapeake Research Conference. Chesapeake 
Research Consortium, Publication Number 149. 
270 
Ju, S.-J., D. H._ Se~or, and H. R. Harvey. 1999. Use of extractible Jipofuscin for age 
determination of blue crab Callinectes sapidus. Mar. Ecol. Prag. Ser. 185? 
171-179. . 
Jung, S. and E. D. Houde. 2000. Scale and pattern in recruitment processes of bay 
anchovy in Chesapeake Bay. ICES CM2000/M: 13. 
Jung, S. , unpublished data. Preliminary analysis and presentation of mid-water trawl 
data from the Chesapeake Bay LMER (TIES) project. 
Karlsen, A. W., T. M. Cronin, S. E. Ishman, D. A. Willard, C. W. Holmes, M. Marot 
and R. Kerhin. 2000. Historical trends_i n Chesapeake Bay dissolved oxygen' 
based on benthic foraminifera from sediment cores. Estuaries. 23(4): 488-
508. 
Kay, J . J., L.A. Graham, and R. E. Ulanowicz. 1989. A detailed guide to network 
~nalysis. In: Wulff, F., J. G. Field, an~ K._H. Mann: (eds.). Network Analysis 
in Marine Ecology. Methods and Applications. Spnnger-Verlag, New York. 
Kemp, W.M., W.R. Boynton, R.R. Twilley, J.C. Stevenson, and J.C. Means. 1983. 
The decline of submerged vascular plants in Chesapeake Bay: summary of 
results concerning possible causes. Mar. Tech. Soc. J. 17(2):78-89. 
Kemp, W. M ., w. R. Boynton, R.R. Twilley, J . C. Stevenson, and L. G. Ward. 1984. 
Influences of submersed vascular plants on ecological processes in upper 
Chesapeake Bay. In: V. S. Kennedy (ed). The Estuary as a Filter. Academic 
Press, New York, NY. 
Kemp, W. M., E. M. Smith, M. Marvin-DiPasquale, and W. R. Boynton. 1997. 
Organic carbon balance and net ecosystem metabolism in Chesapeake Bay. 
Mar. Ecol. Prag. Ser. 150: 229-248. 
Kemp, W. M., J. Faganeli, s. Puskaris, E. M. Smith _and W.R. Boynton. 1999. 
Pelagic-Benthic Coupling and Nutrient Cyclrne. In : Malone, T. C., L. Malej , 
L. W. Harding, Jr., N. Smodlaka, and~- E. T~mer (eds). Ecosystems at the 
Land-Sea Margin . American Geophysical Union, Washington, DC. 
Ki0rboe, T. and T. G. Nielsen. 1994. Regulation of zooplankton biomass and 
production in a temperate, coastal ecosystem. I. Copepods. Limnol. 
Oceanogr. 39(3): 493-507. 
Ki0rboe , T . , F . M""'' hl en be 1 .g  and K . Hamburger..  1985. Bioenerg. etics of the . 
planktonic copepod Acartia tonsa: relat1~~ between_f eed1_ng, egg production 
and respiration, and composition of spec1f1c dynamic action. Mar. Ecol. Prag. 
Ser. 26: 85-97. 
271 
Kimmerer, W. J. 1987. The theory of secondary production calculations for 
continuously reproducing populations. Limnol. Oceanogr. 32(1): 1- 13. 
Kimura R, D. H. Secor, E. D. Houde ED and P. M. Piccoli. 2000. Up-estuary 
dispersal of young-of-the-year bay anchovy Anchoa mitchilli in the 
Chesapeake Bay: inferences from microprobe analysis of strontium in otoliths. 
Mar. Ecol. Prag. Ser. 208: 217-227 
Klesbasco, M. J. 1991. Feeding ecology and daily ratio of bay anchovy, Anchoa 
mitchilli, in the mid-Chesapeake Bay. M. S. Thesis . University of Maryl and 
College Park. College Park, MD. 
Koch , E. W. 2001. Beyond Light: Physical , Geological , and Geochemical Parameters 
as Possible Submersed Aquatic Vegetation Habitat Requirements. Estuaries. 
24(1): 1-17. 
Kremer, P. and M. R. Reeve. 1989. Growth dynamics of a ctenophore (Mnemiopsis) 
in relation to variable food supply. II. Carbon budgets and growth model. J. 
Plank. Res. 11(3): 553-574. 
Langland, M.J., R. E. Edwards, L.A. Sprague and S. E. Yochum. 2001. Summary of 
Trends and Status Analysis for Flow, Nutrients, and Sediments at Selected 
Nontidal Sites, Chesapeake Bay Basin, 1985-99: U.S . Geological Survey Open 
File Report 01-73, 49 p. 
Larson, R. J. 1986. The feeding and growth of the sea nettle, Chrysaora 
quinquecirrha (DeSor), in the laboratory. Estuaries. 9: 376-379. 
Laughlin, R. A. 1982. Feeding habits of the blue crab, Callinectes sapidus Rathbun, 
in the Apalachicola estuary, Florida. Bull. Mar. Sci. 32(4): 807-822. 
Linley, E. A. S., R. C. Newell, and M. I. Lucas. 1983. Quantitative relationships 
between phytoplankton, bacteria and heterotrophic microflagellates in shelf 
waters . Mar. Ecol. Prag. Ser. 12: 77-89. 
Loftus, M. E., D. V. Subba Rao, and H. H. Seliger. 1972. Growth and dissipation of 
phytoplankton in Chesapeake Bay I. Response to a large pulse of rainfall. 
Ches. Sci. 13(4):282-299. 
Luo, J. and S. B. Brandt. 1993. Bay anchovy Anchoa mitchilli production and 
consumption in mid-Chesapeake Bay based on a bioenergetics model and 
acoustic measures of fish abundance. Mar. Ecol. Prag. Ser. 98: 223-236. 
Luo, J., S. B. Brandt, and M. J. Klebasco. 1996. Virtual reality of planktivores: A 
fish's perspective of prey size selection. Mar. Ecol. Prag. Ser. 140 (1-3): 271-
283 
272 
Luo, 1., K. J. Hartman, S. B. Brandt, C. F. Cereo, and T. H. Rippetoe. 2001. A 
spatially-explicit approach for estimating carrying capacity: An application for 
the Atlantic Menhaden (Brevoortia tyrannus) in Chesapeake Bay. Estuaries 
24(4): 545-556. . 
MacIntyre, H. L., R. J. Geider, and D. C. Miller. 1996. Microphytobenthos: The 
ecological role of the "Secret Garden" of unvegetated, shallow-water marine 
habitats. I. Distribution , abundance, and primary production. Estuaries. 
19(2A): 186-201. 
Malone, T. C., H. W. Ducklow, E. R. Peele, and S. E. Pike. 1991. Picoplankton 
carbon flux in Chesapeake Bay. Mar. Ecol. Prag. Ser. 78:11-22. 
Mansueti, R. J. 1961. Age, growth and movements of the striped bass, Roccus 
sa.xatilis, taken in size selective gear. Ches. Sci. 2(1-2):9-36. 
Mansueti, R. J. 1961. 1961. Movements, reproduction, and mortality of the white 
perch, Roccus americanus, in the Patuxent estuary, Maryland. Ches. Sci. 2(3-
4 ): 142-205. 
Marvin-Dipasquale, M. C. and D. G. Capone. 1998. Benthic sulfate reduction along 
the Chesapeake Bay central channel. I. Spatial trends and controls. Mar. 
Ecol. Prag. Ser. 168: 213-228. 
Miller, D. C., R. J. Geider, and H. L. MacIntyre. 1996. Microphytobenthos: The 
ecological role of the "Secret Garden" of unvegetated, shallow-water marine 
habitats. II. Role in sediment stability and shallow-water food webs. 
Estuaries. l9(2A): 202-212. 
Miller, R. E., s. D . Sulkin, and R. L. Lippson. 1975. Composition and seasonal 
abundance of the blue crab, Callinectes sapidus Rathbun, in the Chesapeake 
and Delaware Canal and Adjacent Waters. Ches. Sci. . 16(1): 27-31. 
Moller, E. F. and T. G. Nielsen. 2001. Production of bacterial substrate by marine 
copepods: Effect of phytoplankton biomass and cell size. J. Plank. Res. 
23(5): 527-536. 
Monaco, M. E. and R. E. UJanowicz. 1997. Comparative ecosystem trophic structure 
of three u. s. mid-Atlantic estuaries. Mar. Ecol. Prag. Ser. 161: 239-254. 
Montagna, P.A., G. F. Blanchard, and A. Dinet. 1995. Effect of producion and 
biomass of intertidal microphytobenthos on meiofaunal grazing rates. J. Exp. 
Mar. Biol. Ecol. 185(2): 149-165. 
Namezie, D . A. 1991. Ammonium excretion by gelatinous zooplankton and its role 
in the regeneration of nitrogen in Chesapeake Bay. M. S. Thesis, University of 
Maryland College Park, Maryland. 
273 
Newell , R. I.E. 1988. Ecological changes in the Chesapeake Bay: Are they the 
result of overharvesting the American oyster, Crassostrea virginia. pp. 536-
546 In: M . P. Lynch and E. C. Krome (eds.). Understanding the Estuary: 
Advances in Chesapeake Bay Research . Proceedings of a Conference 29-31 
March 1988. Baltimore, MD. Chesapeake Research Consortium, Publication 
Number 129. 
Nixon, S. W. 1988. Physical energy inputs and the comparative ecology of lake and 
marine ecosystems. Limnol. Oceangr. 33(4, Part 2): 1005-1025. 
North , E . W. 2001. Transport and retention of fish early-life stages in Chesapeake 
Bay: mechani sms and implications for recruitment. Ph.D. Di ssertation . 
University of Maryland, College Park. 306 p. 
Orth , R. J. and K. Moore. 1984. Distribution and abundance of submerged aquatic 
vegetation in Chesapeake Bay. Estuaries 7(4B): 531-540. 
Orth , R. J. and K. A. Moore. 1986. Seasonal and year-to-year variations in the 
growth of Zostera marina (L.) (Eelgrass) in the lower Chesapeake Bay. Aquat. 
Bot. 24: 225-341. 
Orth , R. J., J. F. Nowak, D . J. Wilcox, J. R. Whiting and L. S. Nagey. 1997. 
Distribution of submerged aquatic vegetation in the Chesapeake Bay and 
tributaries and the coastal bays- 1996. US EPA Chesapeake Bay Program, 
Annapolis , MD. 
Pacheco, A. L. 1962. Age and growth of spot in lower Chesapeake Bay, with notes 
on distribution and abundance of juveniles in the York River system. Ches. Sci. 
3(1): 18-28. 
Peters, D.S. and M. T . Boyd. 1972. The effect of temperature, salinity, and 
availability of food on the feeding and growth of the hogchoker, Trinectes 
maculatus (Block & Schneider). J. Exp. Mar. Bio. Ecol. 7: 201-207. 
Poe, T. P ., H . Hansel , S . Vigg, D. E. Palmer, and L.A. Predergast. Feeding of 
predaceous fishes on out-migrating juvenile salmonids in John Day Reservoir, 
Columbia River. Trans. Am. Fish. Soc. 120(4): 405-420. 
Polovina, J. J. 1984. An overview of the ECOPATH model. Fishbyte 2(2): 5-7. 
Prager, M. H. 1996. A simple model of the blue crab, Callinectes sapidus, spawning 
migration in Chesapeake Bay. Bull. Mar. Sci. 58(2): 421-428. 
Purcell, J . E. 1992. Effects of predation by the scyphomedusan Chrysaora 
quinquecirrha on zooplankton populations in Chesapeake Bay, USA. Marine 
Ecology Progress Series. 87: 65-76. 
274 
Purcell , JE. 1997. Pelagic cnidarians and ctenophores as predators: Selective 
predation, feeding rates, and effects on prey populations. Ann. de l Institut 
Oceanogr. 73(2): 125-137. 
Purcell, J.E. , J. R. White, and M. R. Roman. 1994. Predation by gelatinous 
zooplankton and resource limitation as potential controls of Acartia tonsa 
copepod populations in Chesapeake Bay. Limnol. Oceanogr. 39(2): 263-278. 
Purcell, J .E. and J. H. Cowan, Jr. 1995. Predation by the scyphomedusan Chrysaora 
quinquicirrha on Mnemiopsis leidyi ctenophores. Mar. Ecol. Prog. Ser. 129: 
63-70. 
Purcell, IE and Nemazie, DA. 1992. Quantitative feeding ecology of the 
hydromedusan Nemopsis bachei in Chesapeake Bay. Mar. Biol. 113(2): 305-
311. 
Purcell, JE, White, JR, Nemazie, DA and W~ght, DA. 1999. Temperature, salinity 
and food effects on asexual reproduction and abundance of the scyphozoan 
Chrysaora quinquecirrha. Mar. Ecol. Prog. Ser. 180: 187-196 
Purcell , J.E. , T. A. Shiganova, M. B. Decker, and E. D. Houde. 2001. The 
ctenophore Mnemiopsis in native and exotic habitats: U. S. estuaries versus 
the Black Sea basin. Hydrobiol., in press. 
Purcell, J.E. personal communication. UMCES Hom Point Laboratory, Cambridge, 
MD. 
Raymond, P. A. and J. E. Bauer. 2000. Bacterial consumption of DOC during 
transport through a temperate estuary. Mar Ecol. Prog. Ser. 22: 1-12. 
Reeve, M. R., M.A. Syms, and P . .Kremer. 1989. Growth dynamics of a ctenophore 
(Mnemiopsis) in relation to variable food supply. I. Carbon biomass, feeding, 
egg production, growth and assimilation efficiency. J. Plank. Res. 11(3): 535_ 
552. 
Ricciardi, A. and E. Bourget. 1998. Weight-to-weight conversion factors for marine 
benthic macroinvertibrates. Mar. Ecol. Prog. Ser. 163: 245-251. 
Rilling, G. c. and E. D . Houde. 1999a. Regional and temporal variability in 
distribution and abundance of bay anchovy (Anchoa mitchilli) eggs, larvae and 
adult biomass in the Chesapeake Bay. Estuaries. 22(4): 1096-1109. 
Rilling, G. c. and E. D. Houde 1999b. Regional and temporal variability in growth 
and mortality of bay anchovy, Anchoa mitchilli, larvae in Chesapeake Bay. 
Fish. Bull. 97: 555-569. 
275 
Rizzo, W. M. a~d R_. L. Wetzel. 1985. Intertida! and shoal benthic community 
metabolism m a temperate estuary: Studies of spatial and temporal scales of 
variability. Estuaries. 8(4): 342-351. 
Roman, M. R., A. L. Gauzens, W. K. Rhinehart, and J. R. White. 1993. Effects of 
low oxygen waters on Chesapeake Bay zooplankton. Limnol. Oceanogr. 
38(8): 1603-1614. 
Roman, M. R. , D. V. Holliday and L. P. Sanford. 2001. Temporal and spatial patterns 
of zooplankton in the Chesapeake Bay turbidity maximum. Mar. Ecol. Prag. 
Ser. 213: 215-227 
Ross, S. W . 1988. Age, growth, and Mortality of Atlantic croaker in north Carolina 
with Comments on Population Dynamics. 1988. Trans. Am. Fish. Soc. 111'. 
461-473. 
Rugolo, L. J. , K. S. Knotts, A. M. Lange, and V. A. Crecco. 1998. Stock assessment 
of Chesapeake Bay blue crab (Callinectes sapidus Rathbun). J. Shell.fish Res. 
17(2): 493-517. 
Sadzinski, R. , P. Piavis, and A. Jarzynski. 1999. Stock Assessment of Aelected Adult 
Resident Migratory Finfish in Maryland's Chesapeake Bay. US Fish and 
Wildlife Service, Federal Aid Project Annual Report. MD Department of 
Natural Resources, Annapolis, MD. 
Sagasti, A., L. C. Schaffner and J.E. Duffy. 2000. Epifaunal communities thrive in 
an estuary with hypoxic episodes. Estuaries. 23(4): 474-487. 
Sanford, L. P. 1994. Wave-forced resuspension of upper Chesapeake Bay muds. 
Estuaries l 7(1B): 148-165. 
Schroeder, L. A. Consumer growth efficiencies: Their limits and relationships to 
ecological energetics. J. Theor. Biol. 93: 805-828. 
Schubel, J. R. and H. H. Carter. 1977. Suspended sediment budget for Chesapeake 
Bay. In: M. Wiley, ed. Estuarine Processes, Vol !I. Circulation, Sediments, 
and Transfer of Material in the Estuary. Acaderruc Press, New York. 
Shiah, F.-K. and H. w. Ducklow. 1994. Temperature regulation of heterotrophic 
bacterioplankton abundance, production and specific growth rate in 
Chesapeake Bay. Limnol. Oceanogr. 39(6): 1243-1258. 
Smith, E. M. 1998_ Coherence of microbial respiration rate a~d cell-specific bacterial 
activity in a coastal planktonic community. Aquat. Microb. Ecol. 16 (1): 27-35 
276 
Smith, E .M. 2000. Factors regulating respiration and its coupling to primary 
production in a coastal planktonic community. Ph.D. Dissertation. University 
of Maryland at College Park. College Park, MD. 
Smith, E. M. and W. M. Kemp. 2001. Size structure and the production/respiration 
balance ina coastal plankton community. Limnol. Oceanogr. 46(3): 473-485. 
Smith, G. F. 2001. A characterization of oyster habitat in mesohaline Chesapeake 
Bay. Ph.D. Dissertation. University of Maryland at College Park, College 
Park, MD. 
Smol, N., K. A. Williams, J.C. R. Govaere, and A. J. J. Sndee. 1994. Composition 
distribution and biomass of meiobenthos in the Oosterschelde estuary (SW ' 
Netherlands). Hydrobiologia 283: 197-217. 
Sternberger, R. S. and J. J. Gilbert. Body size, food concentration, and poopulation 
growth in planktonic rotifers. Ecology. 66(4): 1151-1159. 
Stokes, C. 1996. Influences of environmental variables on photosynthetic and 
respiratory quotients in Chesapeake Bay. M.S. Thesis. University of 
Maryland at College Park. College Park, MD. 
Sundback, K., P. Nilsson, C. Nilsson, and B. Jonsson. 1996. Balance between 
autotrophic and heterotrophic components and processes in microbenthic 
communities of sandy sediments: A field study. Est. Coast. Shelf Sci. 43: 
689-706. 
St. Pien-e, R. A. and J. Davis. 1972. Age, growth, and mortality of the white perch, 
Morone americana, in the James and York Rivers, Virginia. Ches. Sci. 13(4): 
272-281. 
Szyrmer, J. and R. E. UJanowicz. 1987. Tootal flows in ecosystems. Ecol. Modeling. 
53:123-136. 
Tang, E. P. Y. and R.H. Peters. 1995. The allometry of algal respiration. J. Plank. 
Res. 17(2): 303-315. 
Tarnowski, M. L., M. Homer, and L. Baylis. 1994. The effects of the 1993 freshet on 
Chesapeake Bay oyster populations. In: Hill, P. a.nd S. Nelson, (eds.). Toward 
a Sustainable Watershed: The Chesapeake Expenment. Proceedings of the 
1994 Chesapeake Research Conference. Chesapeake Research Consortium, 
Publication Number 149. 
Tucker, J. w., Jr. 1989. Energy utilization in bay anchovy, Anchoa mitchilli, and 
black sea bass, Centropages striata striata, eggs and larvae. Fish. Bull. U.S. 
78: 279-293. 
277 
----- --
Ulanowicz'. R. E. 1986. Growth and Development: Ecosystems Phenomenology. 
Spnnger-Verlag. 
Ulanowicz, R. E. 1995. Ecosystem trophic foundations: Lindeman exonerata. pp. 549_ 
560 in B .C. Patten and S. Jorgensen (eds.) Complex Ecology: The Part-Whole 
Relation in Ecosystems. Prentice-Hall, Englewood Cliffs, New Jersey. 
Ulanowicz, R. E. 1997. Ecology, The Ascendent Perspective. Columbia University 
Press, New York. 
Ulanowicz, R. E. 1999. NETWRK 4.2a: A Package of Computer Algorithms to 
Analyze Ecological Flow Networks. University of Maryland Center for 
Environmental Science. Chesapeake Biological Laboratory. UMCES Ref No. 
82- 7 CBL. 
Ulanowicz, R.E. and W.M. Kemp. 1979. Toward canonical trophic aggregations. 
Amer. Nat. 114:871-883. 
Ulanowicz, R. E. and D. Baird. 1986. A network analysis of the Chesapeake Bay 
ecosystem. Maryland Department of Natural Resources, Annapolis, MD. 
University of Maryland Center for Environmental Science. REF [UMCEESJ 
CBL 86-79. 
Ulanowicz, R. E. and C. J. Puccia. 1990. Mixed trophic impacts in ecosystems. 
Coenoses 5(1): 7-16. 
Vaughn, D. s., J. w. Smith, E. H. Williams, and M. H. Prager. 2001b. Analysis on 
the status of the Atlantic menhaden stock. NOAA Center for Cooastal 
Fisheries and Habitat Research, Beaufort, NC. Report to the Atlantic States 
Marine Fisheries Commission Menhaden Technical Committee. June 7, 2001. 
Verity, P. G. 1985. Grazing, respiration, excretion and growth rates of tintinnids. 
Limnol. Oceanogr. 30(6): 1268-1282. 
Ve,ity, P. G. 1991. Measurement and simulation of prey uptake by marine planktonic 
ciliates fed plastidic and aplastidic nanoplankton. Limnol. Oceanogr. 36(4): 
729-750. 
Vezina, A. F. Construction of flow networks using inverse methods. In: F. Wulff, J. 
G. Field, and K. H. Mann. (eds.) Network Analysis in Marine Ecology. 
Methods and Applications. Springer-Verlag, New York. 
Virnstein, R. W. 1977. The importance of predation by crabs and fishes on benthic 
infauna in Chesapeake Bay. Ecology. 58: 1199-1217? 
278 
Walters , C. 2000. Natural selection for predation avoidance tactics: implications for 
marine population and community dynamics. Mar. Ecol. Prog. Ser. 208: 309-
313. 
Wang, S.-B. and E. D. Houde. 1994. Energy storage and dynamics in bay anchovy 
Anchoa mitchilli. Mar. Biol. 121: 219-227. 
Wang, S.-B. and E. D. Houde. 1995. Distribution, relative abundance, biomass and 
production of bay anchovy Anchoa mitchilli in the Chesapeake Bay. Mar. 
Ecol. Prog. Ser. 121: 27-38. 
Webb, E. III, and E. Zlokovitz. 1999. Assessment of the 1998 recreational finfish 
harvests in the Choptank and Chester Rivers and summer head boat surveys. 
US Fish and Wildlife Service, Federal Aid Project Annual Report. MD 
Depa11ment of Natural Resources, Annapolis, MD. 
Wenner, C. A. and J. A. Musick. 1975. Food habits and seasonal abundance of the 
American eel, Anguilla rostrata, from the lower Chesapeake Bay. Ches. Sci. 
16(1): 62-66. 
Wetzel, R. L. and P. A. Penhale. 1983. Production ecology of seagrass communities 
in the lower Chesapeake Bay. Mar. Tech. Soc. J. 17(2): 22-31. 
Widbom, B. 1984. Determination of average in~ividua~ dry weights and ash-free dry 
weights in different sieve fractions of manne me1ofauna. Mar. Biol. 84: 101-
108. 
White, J. R. and M. R. Roman . 1992. Egg production by the calanoid copepod 
Acartia tonsa in the mesohaline Chesapeake Bay: the importance of food 
resources and temperature. Mar. Ecol. Prog. Ser. 86: 239-249. 
Zaitsev, Y. P. Recent changes in the trophic structure of the Black Sea. Fish. 
Oceangr. 1(2): 180-189. 
279 
Table 6-1. Definitions of parameters in equations. 
Parameter Definition Uni ts 
BI?  Average biomass mgC m-2 
c, Consumption mgC m-2 d-1 
P, Production mgC m-2 d-1 
R, Respiration mgC m-2 d- 1 
u, Unassimilation = Egestion plus rngC rn-2 d-1 
Excretion 
r . Trophic trans fe r f rom ?lh to J? lh no de  rngC m-2 d-1 lj L 
J.I  Immigration or importation of mgC m-2 d-1 
biomass 
E; Net Emigration minus Immigration, mgC m-2 d-1 
or net export of biomass vi a 
advection or other process. 
y. 
I Fisheries removals of biomass from mgC m-2 d-1 
11 
/ node 
BA1 Accumulation or depletion of mgC m-
2 d-1 
biomass 
M21 Biomass specific mortality rate due d-1 
to predation. 
EE1 Ecotrophic efficiency (defined in unitless 
eq. 2) 
280 
Table 6-2. Sources of published and unpublished data for defining the trophic flow 
networks. 
Node References 
DOC and POC Schubel and Carter (1977), Biggs and Howell (1984), 
Hobbs et al. (1992), Chapter 1, Chapter 3, Chapter 4 
Ph ytop Ia nkton Tang and Peters (1995), Baines and Pace (1991), Malone 
et al. (1991), Kemp et al. (1997), Smith (2000), Harding et 
al. (2001), Smith and Kemp (2001) 
Microphytobenthos Jassby and Platt (1976), Rizzo and Wetzel (1985), 
Macintyre et al. ( 1996), Mi lier et al. ( 1996), Boynton et al. 
(1999), Kemp et al. (1999), Harding et al. (2001) 
Vascular Plants Wetzel and Penhale (1983), Kemp et al. (1984), Kemp et 
al. (1986), Orth and Moore (1986), Orth et al. (1997), 
Duarte and Chiscano (1999). 
Bacteria Ducklow and Shiah (1993), Shiah and Ducklow (1994), 
de! Giorgio and Cole (1998), Smith (1998), Choi et al. 
(1999), Smith (2000), Raymond and Bauer (2000) 
Protozoa and Loftus et al. (1972), Fenchel (1982), Linley et al. (1983), 
Microzooplankton Sternberger and Gilbe1t (1985), Verity (1985, 1991), 
B?rsheim and Bratbak (1987), Heibokel et al. (1988), 
Dolan and Coats (1990), Dolan (199la,b), Dolan and 
Gallegos (1991) , Ducklow (1991), McManus (1991), 
Gaedke and Straile (1994), EPA (2000), Gifford and 
Caron (2000), EPA Chesapeake Bay Monitoring Program 
Data. 
Mesozooplankton Heinle (1966), Ki?rboe et al. (1985), Kimmerer (1987), 
Berggreen et al. (1988), White and Roman (1992), 
Ki?rboe and Nielsen (1994), Roman et al. (2001), 
Chesaepake Bay Microzooplankton and Mesozooplankton 
Monitoring Programs 
281 
Table 6-2 (continued) . 
Node References 
Gelatinous Larson (1986), Kremer and Reeve (1989) , Reeve et al. 
Zooplanton (1989), Nemazie (1991) , Purcell (1992), Purcell and 
Nemazie (1992) , Purcell et al. (1994, 1999, 2001), Houde 
et al. (1994), Purcell and Cowan (1995) , J. Purcell , 
personal communication, EPA Chesapeake Bay 
Zooplankton Monit01ing Program. 
Benthic Bacteria Cowan and Boynton (1996) , Kemp et al. (1997), Marvin-
DiPasquale and Capone (1998) 
Meiobenthos Gerlach (1971) , Fenchel (1978) , Banse and Mosher 
(1980), Schroeder (1981), Widbom (1984), Smol et al. 
(1994), Montagna et al. (1995), Sundback et al. (1996) , 
Flach et al. (1999) , Carlsen et al. (2000). 
Macrobenthos Sagasti et al. (2000), Chesapeake Bay Benthic 
Monitoring Program Data, Chapter 3. 
Oysters Newell (1988), Hargis and Haven (1994) , Jordan et al. 
(1994), Rothschild et al. (1994), Tarnowski et al. (1994), 
Smith (2001) 
Blue crab Miller (1975), Vimstein (1977) , Schroeder (1981), 
Laughlin (1982), Hines et al. (1990), Ebersole and 
Kennedy (1994), Prager (1996) , EPA (1997), Ricciardi 
and Bourget (1998), Rugolo et al. (1998), Ju et al. 
(1999), NOAA Landings Data 
Bay anchovy Schroeder (1981), Tucker (1989), Klebasco (1991), Luo 
and Brandt (1993), Wang and Houde (1994), Wang and 
Houde (1995), Rilling and Houde (1999a, b) , Jung and 
Houde (2000), Kimura et al. (2000). 
Menhaden Durbin and Durbin (1981), Friedland et al. (1989), Luo et 
al. (2001), Vaughn et al. (2001) , NOAA Landings Data 
HeITings and Shads Schroeder (1981), Hoenig (1983) , EPA (1989), Froese 
and Pauly (2001), S. Jung, unpublished data, NOAA 
Landings Data 
282 
Table 6-2 (continued). 
_ Node References 
Striped Bass Hollis (1952), Mansueti (1961), EPA (1994, 1995), 
Hartman and Brandt (1995a, b,c), Froese and Pauly 
(2001), NOAA Landings Data 
Bluefish EPA (1990), Barger (1990), Hartman and Brandt 
(1995a, b,c), Froese and Pauly (2001), NOAA Landings 
Data 
Weakfi sh Hoenig (1983), EPA (1990b), , Hartman and Brandt 
(l995a,b,c), S. Jung, unpublished data, NOAA 
Landings Data 
Atlantic Croaker and Pacheco (1962), Schroeder (1981), Hoenig (1983), Ross 
Spot (1988), EPA (1991), Froese and Pauly (2001), S. Jung, 
unpublished data, NOAA Landings Data 
White Perch Mansueti (1961), St. Pierre and Davis (1972), 
Schroeder (1981), Sadzinski et al. (1999), Webb and 
Zlokovitz (1999), Froese and Pauly (2001) 
Catfish Hoenig (1983), Poe et al. (1991), Sadzinski et al. 
(1999), Froese and Pauly (2001), S. Jung, personal 
communcation, NOAA Landings Data 
fiogchoker Dovel et al. (1969), Peters and Boyd (1972), Denick 
and Kennedy (1997), S. Jung, unpublished data 
American eel Wenner and Musick (1975), EPA (1991b), Froese and 
_________P ~a~u~ly~(2~0~0~1)~,. :::.S:.....:J?: ...u-=::n.:Q..:g,:....::u_n..L..p_ub_J_is_he_d_d_a_ta_ _____  
283 
Table 6-3. Scientific names of species or groups of species referred to in the text. 
UB=upper Bay, MB=mid Bay, LB=lower Bay 
Common Name or Scientific Name(s) of Taxa or Most Abundant Taxa 
Description 
Mesozooplankton Acartia tonsa, Eurytemora affinis(UB, MB), Bosmina 
longirostris (UB) 
Suspension-feeders Rangia cuneata (UB), Mya arenaria (soft-shelled 
clam, MB), Mulinia lateralis (MB), Macoma balthica 
(UB,MB), Chaetopterus variopedatus (LB) 
Deposit feeders Nereis succinea, Macroclymene zanalis (LB) 
Ctenophores Mneniiopsis leidyi, Beroe ovata (LB) 
Sea nettle Chrysaora quinquecirrha 
Oyster Crassostrea virginica 
Blue crab Callinectes sapidus 
Bay anchovy Anchoa mitchilli 
Menhaden Brevoortia tyrannus 
Herrings and Shads Alosa spp., Dorosoma spp. 
Spot Leiostomus xanthurus 
Croaker Micropogonias undulatus 
Hogchoker Trinectes maculatus 
American eel Anguilla rostrata 
White perch Marone americana 
Channel catfish Ictalurus puctatus 
Striped bass Marone saxatilis 
Bluefish Pomatomus saltatrix 
Weakfish Cynoscion regalis 
284 
Table 6-4. Estimated biomass (mgC m-2) for each node in the summer 
trophic flow networks for the upper, mid and lower Chesapeake Bay. 
Below, biomass composition summaii zed by major groupings. The table 
is continued on the next page. 
Node Descrietion Ueeer Ba~ Mid Ba~ Lower Ba~ 
1 Net Phytoplankton 1356 3326 2112 
2 Picoplankton 239 587 373 
3 Free Bacteria 649 2415 1258 
4 PA Bacteria 36 73 39 
5 Heteroflagellates 30 149 43 
6 Ciliates 66 147 87 
7 Rotifers 14 23 0 
8 Meroplankton 3.2 18 38 
9 Mesozooplankton 282 526 1073 
10 Ctenophores 17 126 108 
11 Chrysaora 0 6.3 0 
12 ?-phytobenthos 293 265 293 
13 SAY 2086 1952 1986 
14 Benthic Bacteria 30 298 220 
15 Meiobenthos 700 494 700 
16 DF Benthos 2368 1030 4089 
17 SF Benthos 27232 421 6962 
18 Oyster 0 0 0 
285 
Table 6-4. (Continued) 
Node Descri12tion U1212er Ba)'. Mid Ba)'. Lower Ba)'. 
19 Blue Crab 610 390 342 
20 Menhaden 2136 2136 2136 
21 Bay anchovy 287 381 381 
22 Herrings/Shads 212 0 0 
23 White Perch 282 29 0 
24 Spot 195 222 570 
25 Croaker 50 226 236 
26 Hogchoker 100 50 100 
27 American eel 35 9.0 0 
28 Catfish 450 45 0 
29 Striped Bass 172 172 172 
30 Bluefish 68 68 68 
31 Weakfish 67 67 211 
32 DOC 12504 28207 26915 
33 Sediment POC 201670 201607 76080 
34 POC 5249 10324 8309 
Living Biomass 40065 15651 23597 
susp POC+DOC 17753 38531 35224 
Primary Producers 3974 6130 4764 
Bacteria 715 2786 1517 
Zooplankton 412 995 1349 
Benthos 30300 1945 11751 
Fish and Crabs 4664 3795 4216 
286 
- ..... -
2 1
Table 6-5. Estimated trophic level (TL), production (P, mgC m- d- ) and 
ecotrophic efficiency (EE) for each node in each region of the Bay. 
Pi~oduction by primary producers is net primary production (i .e. production 
~i~us respiration). Total gross primary production for each region is 
indicated at the bottom of the table. The ratio of total production in several 
groupings of nodes to total organic input (O.I.) to each region is shown at 
the bottom. The table is continued on the next page. 
U_gper Bay Mid Bay Lower Bay 
TL P EETL P EETL PEE 
Desc1iption 1.0 719 99 1.0 1962 0.47 1.0 1704 0.61 
l Net Phytoplank 1.0 186 63 1.0 511 0.36 1.0 426 0.36 
2 Picoplankton 2.0 350 100 2.0 1425 0.94 2.0 742 0.66 
3 Free Bacteria 2.0 28 80 2.0 82 0.02 2.0 41 0.60 
4 PA Bacte1ia 3.0 99 67 3.0 417 0.35 3.0 142 0.61 
5 Heteroflagel !ates 2.6 165 59 2.6 366 0.48 2.6 218 0.79 
6 Ciliates 2.3 4.2 54 2.4 6.9 0.56 2.4 0.0 0.00 
7 Rotifers 2.6 1.6 64 2.5 9.0 0.89 2.4 19.0 0.73 
8 Meroplankton 2.6 107 99 2.5 263 0.68 2.4 268 0.62 
9 Mesozoop Ia n kton 3.6 2.4 0 3.5 16.4 0.22 3.4 14.0 0.00 
10 Ctenophores 4.6 0 4.4 0 0.00 0.00 
11 Chrysaora 1.0 176 67 1.0 159 0.48 1.0 234 0.66 
12 ?phytobenthos 1.0 17 0 1.0 18 0.00 1.0 18 0.00 
13 SAV 2.0 34 72 2.0 337 0.12 2.0 249 0.37 
14 Benthic Bacteria 2.1 49 63 2.3 35 0.70 2.3 49 0.95 
15 Meiobenthos 2.4 19 34 2.5 14 0.76 2.3 57 0.49 
16 DF Benthos 2.0 218 8 2.1 6 0.86 2.1 97 0.24 
17 SF Benthos O 0.00 0.00 
18 Oyster 3.1 2.4 78 3.2 1.6 0.98 3.0 1.4 0.85 
19 Blue Crab 
287 
Table 6-5. (Continued). 
U1212er Bay Mid Bay Lower Bay 
Descrietion TL p EE TL p EE TL p EE 
20 Menhaden 2.1 53 97 2.1 53 0.97 2.1 53 0.74 
21 Bay anchovy 3.6 23 41 3.4 31 0.34 3.4 31 0.32 
22 HaJTingslShads 3.6 2.1 88 0.00 0.00 
23 White Perch 3.4 0.6 88 3.4 0.1 0.17 0.00 
24 Spot 3.1 2.0 96 3.2 2.2 1.00 3.2 5.7 0.98 
25 Croaker 3.1 0.5 48 3.2 2.3 1.00 3.2 2.4 1.00 
26 Hogchoker 3.1 2.5 13 3.2 1.3 0.00 3.2 4.0 0.00 
27 American eel 3.1 0.1 99 3.3 0.0 0.99 
28 Catfish 3.2 0.5 37 3.3 0.0 0.38 0.00 
29 Striped Bass 3.6 0.3 78 3.6 0.3 0.78 3.5 0.3 0.78 
30 Bluefish 4.1 0.5 44 4.2 0.5 0.44 4.2 0.5 0.49 
31 Weakfish 4.0 0.6 78 3.9 0.6 0.78 3.6 1.9 0.78 
32 DOC 1.0 99 1.0 1.00 1.0 1.00 
33 Sediment POC 1.0 49 1.0 0.99 1.0 0.91 
34 POC 1.0 101 1.0 0.17 1.0 0.65 
Gross Primary 1487 3515 2956 
Prod. 
Det1itus In12ut 367 600 426 
Bacterial PIO.I. .22 .37 .29 
Protozoa/OJ. .14 .16 .10 
Mesozoo PIO.I. .057 .053 .074 
Benthic Pl 0. I. .13 .004 .043 
Fish Pl O.I. .047 .019 .028 
288 
'Table 6-6a. The matrix. of organic carbon flows (T;j, mgC m-2 d-1) among the nodes of 
the upper Chesapeake Bay summer trophic flow network. The matrix. is pa1titioned 
into pl ankton (node 3-lO), benthos (node 14-19), nekton (node 20-31) and detritus 
( node 32-34). S=Sum. 
No de.,  
1 
3 4 s 6 7 8 9 10 14 15 16 l7 19 20 21 
l 165 ll 2 .6 ll9 223 192 
2 33 l.4 .53 26 56 
3 283 66 
4 22 
s 66 .70 
6 l.4 l.9 90 4.2 
7 .OS l.3 .84 
8 .l2 .84 
9 7.7 5.5 ll 75 
10 
12 98 19 
13 
14 20 s 
15 29 .73 
16 .73 
17 4.4 
19 
20 
2 1 .12 .73 
22 
23 
24 
25 
26 
27 
28 
29 
30 
31 
32 811 
33 113 78 44 .73 
34 66 .27 13 815 11 2.5 
s 811 66 283 330 14 5.3 257 5.8\ 113 196 97 1117 7 .3\ 214 84 
289 
Table 6-6a. (continued) 
l No de .1  
22 23 24 25 26 27 28 29 30 31 32 33 34 Total 
l 4.6 .57 l 719 
2 56 14 186 
3 .78 1 350 
4 2.9 3 28 
5 76 76 218 
6 .4 68 68 234 
7 .08 3.8 4 9.8 
8 .08 1.4 1.4 3.7 
9 7 .10 58 58 223 
10 2.3 2.3 4.6 
12 29 29 176 
13 17 17 
14 9.4 34 
15 .20 .70 .18 116 147 
16 .41 2.8 .72 6.0 .19 .22 .36 94 10 116 
17 .61 2.8 .72 7.5 .19 1.2 .22 .36 379 379 776 
19 .22 1.8 1.8 3.8 
20 2.15 1.0 .7 15 15 33 
21 .51 .36 .65 1.4 1.7 20 20 45 
22 .43 1.2 .23 1.3 1.3 4.6 
23 .13 .23 .4 .36 1.1 
24 .22 1.0 .68 1.2 1.2 4.2 
25 .24 .4 .42 1.1 
26 .09 .23 3.5 3.5 7.3 
27 .06 .06 0.12 
28 .4 .4 .9 
29 .41 .41 .8 
30 .55 .55 1.1 
31 .46 .46 .9 
32 811 
33 .18 95 155 487 
34 .23 .20 .70 .18 1.5 910 
s 7.6 2.0 7.0 1.8 15 0.39 1.8 4.3 4.8 4.6 822 487 906 5862 
290 
Table 6-6b. The matrix of organic carbon flows (Tu) among the nodes of the mid 
Chesapeake Bay trophic flow network. The matrix is partitioned into plankton (node 
3-10), benthos (node 14-19), nekton (node 20-31) and detJitus (node 32-34). Nodes 
with zero biomass are omitted. S=Sum. 
Nod e , 
l 
6 7 8 9 10 
11 14 15 16 17 19 20 21 
3 4 5 16 192 
l 293 16 18 379 1.5 
2 147 1.2 1.5 32 
3 1192 147 1.5 
4 
5 147 1.2 6 
6 4.6 7.5 159 1 
7 .15 3.2 .8 7 
8 40 .20 11 94 32 
9 1.5 3.5 
10 
11 69 7.1 
12 
13 35 7.1 
14 21 .47 
15 1.4 
16 1.9 
17 .47 
19 
20 .83 .20 
21 
23 
24 
25 
26 
27 
28 
29 
30 
_31 
32 2850 673 35 36 .47 
33 10 11 3.3 
1.5 32 
_)4 165 73 138 71 29 4.7 2 14 111 
23 30 636 
42 3.9 6 
- S 2850 165 1192 734 
291 
Table 6-6b (Continued) 
Node i 
31 I 32 33 34 Total 
23 24 25 27 26 28 29 
30 
- 524 524 1962 
1 262 66 511 
2 43 43 1425 
3 57 24 82 
4 539 231 918 
5 172 172 520 
6 6.1 6.1 16 
7 6.5 6.5 21 
8 186 186 549 
9 14.3 14 32 
10 1.8 1.8 3.6 
11 41 41 159 
12 18 18 
13 148 148 337 
14 80 104 
15 .80 .81 .75 .43 .36 7.8 31 
55 
16 .06 4.8 4.9 .05 4.5 .05 3.1 12 23 
17 .06 1.6 1.6 .05 1.5 .07 
.22 .36 
.22 2.0 
2.7 
19 .39 1.1 5.9 9.6 2.2 
20 27 59 
.02 .86 1.5 1.7 
27 
21 .05 .06 .06 .11 
23 
.02 .43 .72 1.0 
1.3 1.3 4.7 
24 2.2 1.3 1.3 4.8 
25 .00 
26 .016 .016 .032 
27 .04 .04 .09 
28 .41 .41 .81 
29 .55 .55 1.1 
30 .46 .46 .92 
31 2850 
32 746 
33 .8 .8 .02 .02 .8 812 271 1305 
_34 4.6 28 54 910 1305 12023 4.3 4.8 
__? .21 8.0 8.1 0.1 7.5 
. 18 
292 
Table 6-6c. The matrix of organic carbon flows (T;j) among the nodes of the lower 
Chesapeake Bay trophic flow network. The mat1ix is partitioned into plankton (node 
3-10), benthos (node 14-19), nekton (node 20-31) and detritus (node 32-34). Nodes 
with zero biomass are omitted. S=Sum. 
I Node i 
3 4 5 6 8 9 10 14 15 16 17 19 
l 174 38 360 278 
2 87 4.4 46 15 
3 405 87 
4 24 
5 87 
6 15 151 
8 .7 
9 33 35 
10 
12 98 57 
13 
14 49 43 
15 43 .21 
16 2.1 
17 1.0 
19 
20 
21 
24 
25 
26 
28 
29 
30 
31 1900 
32 829 49 143 .82 
33 105 6.3 65 171 
s 1900 105 405 435 63 655 36 829 196 286 487 4.1 
293 
Table 6-6c. (Continued) 
l Nod e 1.  
20 21 24 25 26 28 29 30 31 32 33 Total 
1 192 530 132 1704 
2 219 55 426 
3 200 50 742 
4 13 3.3 41 
5 180 45 312 
6 5.5 111 28 309 
8 5.3 7.8 15 15 44 
9 5.3 94 198 198 563 
10 14 14 28 
12 39 39 234 
13 18 18 
14 78 78 249 
15 1.0 .43 2.1 100 147 
16 12 5.1 7.4 .43 .72 34 138 200 
17 7.2 3.0 12 .22 .72 63 254 341 
19 .22 1.9 2.1 
20 2.15 .39 6.5 8.3 25 8.3 51 
21 .86 1.5 2.2 52 13 70 
24 .43 .72 4.3 3.3 3.3 12 
25 2.2 1.4 1.4 4.9 
26 5.4 5.4 1.7 
28 .41 .41 .81 
29 .54 .54 1.1 
30 1.4 1.4 2.9 
31 1900 
32 1022 
33 11 3.3 213 574 
s 214 111 21 8.5 21 4.3 4.8 14 1916 1146 574 9435 
294 
Table 6-7. Respiration (mgC m-2 d-1) for each node in 
the three summer trophic flow networks (UB=upper Bay, 
MB=mid Bay, LB=lower Bay) and the fraction of total 
respiration contributed by each of 5 major groups. 
Mi ssing entries indicate nodes with zero biomass. The 
table is continued on the next page. 
Node Desc1ietion UB MB LB 
1 Net Phytoplankton 271 665 422 
2 Picoplankton 60 147 93 
3 Free Bacteria 461 1425 1157 
4 Particle Attached Bacteria 37 82 64 
5 Heterofl a gel \ates 65 274 93 
6 Ciliates 96 213 126 
7 Rotifers 4.2 6.9 
8 Meroplankton 1.6 9.0 19 
9 Mesozooplankton 34 87 92 
10 Ctenophores 1.2 9.4 8.1 
11 Chrysaora 0.27 
12 Microphytobenthos 59 53 59 
13 SAV 
14 Benthic Bacteria 79 787 581 
15 Meiofauna 49 35 49 
16 DF Benthos 50 21 86 
17 SF Benthos 340 8.6 146 
18 Oysters 
19 Blue Crab 1.8 1.2 1.0 
295 
Table 6-7. (Continued). 
Node Descrietion VB MB LB 
20 Menhaden 132 156 132 
21 Bay anchovy 35 46 36 
22 Herrings and Shads 3.07 
23 White Perch 0.82 0.08 
24 Spot 2.82 3.22 8.25 
25 Croaker 0.72 3.27 3.42 
26 Hogchoker 7.70 3.85 10.28 
27 American eel 0.16 0.04 
28 Catfish 0.71 0.07 
29 Striped Bass 3.25 3.25 3.25 
30 Bluefish 3.53 3.53 3.53 
_)l Weakfish 3.18 3.18 10.02 
TOTAL 1803 4047 3203 
Algae/Total 0.22 0.21 0.18 
Bacteria/Total 0.32 0.57 0.56 
Zooplankton/Total 0.11 0.15 0.11 
Macrobenthos/Total 0.24 0.02 0.09 
- Fish/Total 0.11 0.06 0.06 
296 
Table 6-8. Estimated summer fisheries landings in each of three regions 
of the mainstem Bay. Landings exclude estimated landings in tributaries 
as Well as landings outside of summer. R=Recreational Landings, 
C==Commercial Landings. BC=blue crab, MH=menhaden, SP=spot, 
CR==croaker, AE=American eel, CF=catfish, BF=bluefish, 
WF==weakfish. Tot=Total. 
U[>[>er Ba}'. Mid Ba}'. Lower Ba_}'. 
- R C Total R C Total R C Total Total 
mgC m?2d-' 
BC 1.82 1.82 2.09 2.09 1.09 1.09 
MR 30.6 30.6 
SP .021 .018 .039 .089 .163 .252 
CR .02 .012 .032 .05 0.08 0.13 
AE .126 .126 .032 .032 
CF .146 .038 .184 .015 .004 .019 
BF .197 .009 .206 .197 .009 .206 .21 .02 0.23 
WF 
- .036 .002 .038 .036 .002 .038 .359 .059 0.418 Tot 0.38 2.00 2.37 0.29 2.17 2.46 0.71 32.01 32.72 -1 tons WW summer 
BC 988 988 5619 5619 3336 3336 9943 
MH 74912 74912 74912 
SP 45 39 84 218 399 617 701 
CR 43 26 69 122 196 318 387 
AE 55 55 69 69 124 
CF 63 17 80 32 9 41 121 
BF 57 3 60 282 13 295 343 33 375 730 
-WF 13 l 13 62 3 65 703 116 819 897 Tot 133 1062 1195 465 5778 6243 1386 78991 80377 87815 
297 
Table 6-9. Estimated trophic level (TL), biomass (B, mgC m-2) , production (P, 
mgC m-2 d-1), ecotrophic efficiency (EE), and respiration (R, mgC m-2 d-1) for the 
restored mid Chesapeake Bay trophic flow network. Gross production for 
primary producers is the sum of the indicated P and R. 
Node Descrietion TL B p EE R 
1 Net Phytoplankton 1.0 1663 981 0.855 333 
2 Picoplankton 1.0 294 256 0.659 74 
3 Free Bacteria 2.0 503 568 0.874 855 
4 PA Bacteria 2.0 37 42 0.55 42 
5 Heterofl a gel Ia tes 3.0 55 154 0.746 101 
6 Ciliates 3.0 57 142 0.893 83 
7 Rotifers 2.5 23 6.90 0.503 6.9 
8 Meroplankton 2.5 120 60 0.568 60 
9 Mesozooplankton 2.5 316 158 0.96 52 
10 Ctenophores 3.5 126 16 0.215 9.4 
11 Chrysaora 4.5 6.3 0.126 0 0.265 
12 Microphytobenthos 1.0 559 335 0.429 112 
13 SAY 1.0 13500 122 0 0 
14 Benthic Bacte1ia 2.0 163 184 0.391 277 
15 Meiofauna 2.3 700 49 0.306 49 
16 DF Benthos 2.2 3333 47 0.443 68 
17 SF Benthos 2.1 6666 93 0.093 137 
18 Oysters 2.0 0 
19 Blue Crab 3.1 390 1.56 0.976 1.15 
20 Menhaden 2.1 2176 54 0.95 135 
21 Bay anchovy 3.5 381 31 0.342 46 
22 Herrings and Shads 3.5 0 
23 White Perch 3.4 29 0.058 0.172 0.084 
24 Spot 3.0 316 3.16 0.7 4.58 
25 Croaker 3.0 322 3.22 0.7 4.66 
26 Hogchoker 3.0 50 1.25 0 3.85 
27 American eel 3.0 9 0.027 0 0.04 
28 Catfish 3.2 45 0.045 0.376 0.07 
29 St1iped Bass 3.5 172 0.344 0.775 3.25 
30 Bluefish 4.1 68 0.476 0.439 3.53 
31 Weakfish 3.8 67 0.603 0.777 3.18 
32 DOC 1.0 28207 1.001 
33 Sediment POC 1.0 201607 0.939 
34 POC 1.0 5162 0.739 
Gross Primary Prod. 2213 2465 
Net Ecosys. -253 
Metabolism 
Detrital In2uts 350 
298 
L L 
39 30' 
39 00' 
Fig. 6-1. A map of 
Chesapeake Bay showing 
38 30' the boundaries of the 
mainstem Bay regions for 
which trophic flow 
networks were computed. 
Depth contours indicate 
the distribution of shoal 
(<7m), transition (7-lOm) 
and deep water (>!Om) 
38 00' habitats in each region. 
37 30'-
37 00'-
299 
Fig. 6-2. A diagram of the basic model structure for the Chesa-
peake Bay trophic flow networks. All the nodes are present in the 
upper Bay network, while some nodes have zero biomass in the 
mid Bay and lower Bay networks. Non-Jiving storages of organic 
matter, p1imary producers, and consumers are indicated via separate 
symbols. As indicated on the key, large arrows indicate trophic 
transfers. Labelled small arrows indicate flows to detritus, which 
are represented as collected flows returning to the respective detri -
tal pools. Labelled small arrows also indicate biomass imports and 
exports due to net changes in biomass, immigration or emigration, 
physical processes (principally advection and burial), and biomass 
exports due to fisheries harvests . Respiration is indicated by elec-
trical "ground" symbols. Capital letters are used to simplify the 
representation of flows from a large number of prey species to the 
three major piscivores. For simplicity, mesozoooplankton and 
meroplankton are represented as one node, even though they are 
separate in the network models. Oysters were found to have insig-
nificant biomass and are therefore not included in the diagram. 
Also for simplicity, some very small flows were not included in the 
diagram. All the flows are accurately and quantitatively represent-
ed in the diet matrices (Tables 6-6a,b,c). 
300 
;; Piscivores 
~ 
A,B,C,D,E.G,H ba 
I ~ ), A doc. 
poc 
I \ / ), B ,,c,c-?J:" 
T doc, 
~ poc 
r ~sed 
Net_PP l.h : ), C 
l;.l 
,0_ _ Miao-
phytobenthos "''""-BT doc, p?oc  
Net PP 
-;:;;. 
doc, POC 
sed 
I > D 
Macrobenthic 
Community 
poc 
doc, Flow to dissolved organic carbon, + ba Biomass accumulation 
+ poc, particulate organic carbon, or depletion. D Primary Producer 
sed sediment organic carbon. -r Respiration 0 Detritus storage 
+ i,e Immigration or Emigration 
+ h Harvest ~ Trophic Transfer 0 Consumer 
c::z=i Net phytoplankton 
5000 ~ Picoplankton 
- Microphytobenth. SAV 4115 4000 c:::J Spring Dep 
es2Sl Other Detritus 
..--
'"' O 
N 3000 
E 
0 
O') 
E 2000 1854 
Upper Bay Mid Bay Lower Bay 
Fig.6-3. Estimates of organic matter inputs to the summer food webs in the three 
regions of Chesapeake Bay from net phytoplankton (>3 ?m) , autotrophic 
picoplankton ( <3 ?m) , microphytobenthos, submersed aquatic vegetation, 
phytoplankton deposited to sediments during winter-spring, and other det1itus. In 
the mid Bay, there was an additional input of 9 mgC m-2 d-1 at trophic levels >l 
which was largely due to summer depletion of macrobenthic biomass accumulated 
du1ing spring. In the other regions, this input was <l mgC m-2 d-1. 
302 
Algae C=:J Plankton 
C::=J Bacteria r:::z=:l Meio/Macrobenthos i;:::::s:J Fish/Crabs 
1803 4049 3204 
0.8 
0.6 
ro 
;? 
'+--
0 
cf:. 0.4 
0.2 
0.0 
Upper Bay Mid Bay Lower Bay 
Fig. 6-4. The fraction of total respiration due to algae (including 
microphytobenthos), bacte1ia (including benthic bacteria), meiobenthos and 
macrobenthos, and fi sh and crabs. Total respiration (mgC m-2 d-1) is indicated at 
the top of each bar. 
303 
Upper Bay Canonical Food Chain 
1.2 4230 10.7 552 1.6 189 0.3 60 0.1 45 0.1 4 0 0 
34100 > I I 0 
Mid Bay Canonical Food Chain 
513 4230 256 686 56 185 2.9 57 0 39 0 1.6 0 0.01 
VJ 
0 
~ 
86% 
79600 
D 
55200 > I~-~ 
Fig. 6-5a. The trophic flow networks for the upper and mjd Chesapeake Bay projected into a linear chain of canoni-
cal trophic levels indicated by Roman numberals . The aggregated detritus pool is indicated by "D." The projection 
was computed using the algorithm of Ulanowicz and Kemp (1979) as modified by Ulanowicz (1995) and imple-
mented in NETWRK software. Arrows not originating froma box, or not pointing to a box indicate exogenous 
inputs and exports, respectively. All flows were converted to total summer (92 day) flows, with units mgC m-2 sum-
mer-1. 
Lower Bay Canonical Food Chain 
2.58 2710 17 644 4.1 186 0.5 52 0 32 0 1.6 
272000 .i 
95100 
39200 ::J111 I 
w 
0 
Vl 
Fig. 6-5b. Linear chains of canonical flows for the lower Chesapeake Bay food web during 
summer. For a full description see Fig. 6-5a. 
513 4230 
l+D 151000 
To/From 
Upper 
79600 136000 248000 Trophic 
Levels 
Fig. 6-6. An illustration, for the mid Bay, of how the 
canonical flows in Figs 6-5a,b can be collapsed into a 
Lindeman Spine by combining flow due to primary pro-
duction (I) and detritus (D). The description in the legend 
for Fig. 6-5a applies to this figure. 
306 
- _._ Upper Bay 
106 c:=i ---0- Mid Bay 
c:=i __,, Lower Bay 
105 
~L 
(1) 104 
E 
E 
:::J 
3 
NC/) 10
E 
(_) 
oi 102 
s 
Ol 
-~ 101 
r..o..  
C) 
100 
10-1 7 8 4 5 6 
1 2 3 Trophic Level 
Fig. 6-7. The organic carbon flow to each trophic successive canonical tro-
phic level in the upper, mid and lower Bay fo~d web. Flows are shown up 
to trophic level 8, although a very small quantity of flow exceeds this level 
for the mid Bay network. Reference lines drawn from the initial input to 
TL 1 for the mid Bay at constant efficiencies of 10, 20 and 30% per tro-
phic level reveal the average trophic transfer efficiency (TIE) over n 
exchanges. Bars show the TTE at each canonical trophic level in each of 
the three regions of the Bay. 
307 
lz (?E m-2 s-1) 
0.1 1 10 100 1000 
0 +---_ _._ ___ .._ ___ __J.___----1 
5 
10 
15 
20 +----1-L- ~ ----------1 
0 50 100 150 200 
Bottom Area (106 m2) 
Fig, 6-8. The bottom area present in l m depth increments in 
the mid Bay region and the the penetration of light at cun-ent 
average light attenuation (kd=0.8 m- 1) as well as at the estimated 
reduced light attenuation in the restored mid Bay (kd=0.4 m-1). 
Horizontal lines indicate the depth of 1 % light penetration (z 1%) 
for each light attenuation level. 
308 
1.0 
_._ - Modern Mid Bay 
--o- c:J Restored Mid Bay 
0.8 
>, 
(.) 
C 
Q.) 
?c::5 
:E 
w 0.6 
I.... 
Q.) 
'+-
Cf) 
C 
ro 
I.... 
I- 0.4 
(.) 
..c 
0.. 
0 
I.... 
I- 0.2 
0.0 
1 2 3 4 5 6 7 8 
Fig 6-9. Trophic transfer efficiencies for canonical trophic levels computed from 
the trophic flow networks for the modern mjd Bay and the restored Bay (bars). 
The lines indicate the geometric mean trophic transfer efficiencies for trophic 
transfers up to the indicated trophic level in the modern mid Bay and the restored 
mid Bay. 
309 
N- - 0.25r,-~-----==----------------
.b..- - - Modern Bay ...__.. 
...., (a) c=J Restored Bay 
C 
(l) 0.20 
?o 
:i= 
(l) 0.15 
u0  
C 0 .10 
0 
~ 
:::J 
..a 
?.;..:.:,  0.05 
C 
0 o.oo~--__.J 
cu  1.00 , -------------------.--------
-~ 0.80 (b) I 
:i"= 0.60 
<~;;  0.40 I 0.20 n n 
2 0.10 
:::J 
..a 
?;:: 
C o.os 
u0  
0.00 ~--__.J 
Net Phytoplankton Microphytobenthos 
Picoplankton SAV 
Fig 6-10. Total contribution coefficients, indicating the fraction of net produc-
llon from the indicated producer eventually reaching (a) striped bass or (b) 
benthic bacteria over all pathways. 
310 
Chapter 6: 
y A LIDAT lO N OF TROPHIC LEVEL ESTIMATES FOR CHESAPEAKE BA y 
FOOD WEB USING 15N/14N. 
Abstract 
Ana\ysis of the stab\e nitrogen isotopic composition (815N) of Chesapeake Bay 
biota was used to estimate average trophic \eve\ (TL) for 28 taxa or groups of taxa in 
e ach of three regions of the Bay during spring, summer and fa\\. Estimates of TL 
based on 815N (TLs1) were compared to corresponding estimates derived from trophic 
network ana\ysis (TLNA), Average TLs1 for mesozoop\ankton was found to be higher 
in the upper Bay compared to e\sewhere in the Bay. Although this pattern was a\so 
observed for TLNA, the regiona\ differences were sma\\er. On average for the 28 taxa, 
TLsi was 0.53?0. tO greater than TLNA? A\though analysis of 815N provided a usefu\ 
15
independent validation of TLNA, natura\ variability in 8 N and the \arge variety of 
15
biogeochemica\ processes potentia\\y affecting 0 N undermined the effectiveness of 
81sN for this purpose. The sma\\est differences between TLNA and TLs1, and the 
narrowest confidence \imits for TLs1, were obtained when 815N was repeatedly 
measured. This suggests that repeated measurement of 815N wou\d be an effective 
means of obtaining more certain resu\ts. 
Introduction 
Although the biota that comprise food webs feed on varied diets, the concept 
of trophic \eve\ remains useful, rescued by the introduction of fractional trophic \eve\s 
(Odum and Heald 1975). Soon after, after a concise numerical procedure was 
developed to estimate fractiona\ trophic \eve\s from quantitative trophic flow networks 
(U\anowicz and Kemp 1979, U\anowicz 1995). Trophic \eve\ estimates prove usefu\ 
in a number of ways. For examp\e, the concept of trophic \eve\ is essential to the fish 
production computations of Ryther (1969) and Iverson (1990) and estimated trophic 
\eve\ of the harvest is needed if one wishes to estimate the possible harvest. Changes 
3ll 
in the average trophic level of fish targeted by fishing fleets have been cited as 
evidence for unsustainable fishing at the global scale (Pauly et al. 1998). 
Estimates of average trophic level can now be readily obtained as standard 
output from software packages designed for the analysis of quantitative trophic 
network models (NETWRK, Ulanowicz 1999, ECOPATH with ECOSIM, Christensen 
et al. 2000). The data requirements for the specification of such models are 
substantial, however. Diet composition for each taxa or group of related taxa in the 
food web must generally be known a priori, even if it is possible to later improve or 
further constrain the estimates through mass balance constraints (i.e. inverse analysis 
and related methods). Generally, the available diet studies are few, especially if one 
limits the search to studies conducted in the ecosystem of interest. More likely, diet 
data come from a potentially similar ecosystem. By obtaining diet information in this 
way, differences among systems in the trophic levels of species may not be revealed 
through trophic network models. Even a detailed seasonal study of diet within the 
ecosystem of interest may not provide all the info1mation that it needed. For example, 
Hartman and Brandt (1995) reported age-structured and seasonally resolved diets for 
three piscivorous predators in the mid Chesapeake Bay in the early 1990's. However, 
the diet of striped bass (Marone saxatilis) in the upper Chesapeake Bay was different, 
at least in the 1930's, including more prey items that were present in low salinity 
waters (Hollis 1952). The possible seasonal, spatial and interannual differences in diet 
most likely preclude consistent resolution of diet variations through gut contents 
analysis. Examining the diets of very small organisms such as the mesozooplankton, 
microzooplankton, or even protozoa poses an even larger challenge, since gut contents 
analysis is not feasible. For these animals, carefully constructed grazing experiments 
can reveal diet preferences (Stoecker and Cupuzzo 1990), but this may differ from the 
in situ diet. 
Analysis of the natural abundance of stable isotopes in tissues, particularly 
15N/ 14N, has become a well known method for examining trophic relationships and 
sources of organic matter to food webs (Peterson and Fry 1987, e.g. Fry 1988). The 
suitability of stable isotopes for this purpose depends on two empirically observed 
312 
properties. First, the stable isotopic composition of tissues reflects the stable isotopic 
composition of the diet. Secondly, and as important, is the fact that whi le the different 
isotopes function similarly in chemical reactions, reaction rates for the heavier isotope 
(i.e. 1sN) proceed slightly more slowly. The result is a process called isotopic 
fractionation. Empirically, the ratio isN/14N increases with each trophic transfer due 
to excretion of isN depleted nitrogen (Peterson and Fry 1987). Because of the very 
small differences that are observed, standard notation ("de!" notation) is commonly 
used in which the isN/ 14N ratio for the sample (Kample) is expressed relative to the 
same ratio (Rs,a11darc1) for a standard according to 
515 N = (R sample / R s tandard - l) x 1000 (1) 
with units %0, or "per mil." The standard for 81sN is isN/14N for dinitrogen gas (N2) in 
air. The difference between 81sN for the animal and its diet, ti(8animal -8dier ), varies 
between 2%o and 6%0 but has been reported to be less vaiiable than that suggests, with 
a mean value of about 3.4%0 (Montoya et al. 1990). The application to the estimation 
of trophic level is relatively straightforward, namely , 
= (815 + 615 TL TLref N .rnmple - N reference )/3 .4 (2) 
where the reference organic matter pool may be particulate nitrogen (PN), or perhaps 
the mesozooplankton. Kline and Pauly (1998) estimated trophic levels using this 
approach to cross-validate trophic level estimates obtained using network analysis, 
apparently the first application of stable isotope methods for that purpose. Compared 
to the high cost of labor intensive grazing experiments and gut contents analysis, 
stable isotope methods can be applied inexpensively. For this reason, the approach is 
suitable for examining seasonal and regional differences in trophic relationships (e.g. 
Montoya et al. 1990). Such differences can also be compared to seasonal or regional 
differences resolved using trophic network models (Baird and Ulanowicz 1989, 
Chapter 5), similar to the approach of Kline and Pauly (1998). 
In this study, the stable nitrogen isotopic composition of many components of 
the Chesapeake Bay food web was examined for three distinct regions of the Bay 
(upper Bay, mid Bay, lower Bay) in each of three seasons (Spring, Summer, Fall) 
313 
during fall 1995 to fall 1998. Average trophic level for each food web component was 
15
derived from 8 N and compared to summer trophic level estimates obtained through 
analysis of trophic flow networks (Chapter 5). 
Methods 
Samples were collected from RIV Cape Henlopen within 3 general areas 
(upper Bay, mid Bay, lower Bay) duiing three week long cruises each year beginning 
in fall 1995 and concluding in fall 1998 (Fig. 7-1 , Table 7-1). During 1996, samples 
Were also collected in an additional area at the southern limit of the upper Bay region , 
the Gibson Island transect (Fig. 7-1). A total of 274 samples were collected for stable 
isotope analysis . These organic matter sources that were sampled included surficial 
sediments, seston (PN), microplankton , mesozooplankton, gelatinous plankton, and 
fish . Undisturbed surficial sediments were obtained with a Smith-Macintyre grab 
sampler. A 30 ml sample of the top 2 mm was obtained from the core and 
immediately frozen. PN was obtained from water samples collected by Niskin bottle 
from near bottom, pycnocline and surface and combined in equal amounts. The water 
sample was passed through a 40 ?m screen, then filtered onto pre-combusted GF!F 
filters (450 ?C for 5 hr.) until clogged. The filters were wrapped in foil and 
immediately frozen. The procedures for sampling microzooplankton and 
mesozooplankton changed over the course of the study and can be described as 
follows. 
~Mesozooplankton were obtained via a 5-minute oblique Tucker Trawl 
tow with a 280 ?m net. The trawl captured only larger copepods and fish larvae. The 
sample was sieved dry, rinsed in deionized water and frozen immediately. 
~: A 40 ?m plankton net with a non-filtering cod-end was repeatedly towed 
vertically through the water column to obtain a concentrated sample of the plankton 
(>40 ?m). The sample was gently passed serially through 200 ?m and 40 ?m screens. 
The two 1a rger 1.c. rac t?1 0ns wer e concentrated then added to 0.2 ?m filtered seawater in a 
314 
4
0 ?m mesh-bottomed incubator. The plankton were left for 2 hours to clear their 
guts. Plankton that did not settle to the bottom after 2 hours were removed with a 
turkey baster, filtered onto precombusted GF/F filters and frozen. 
Ctenophores (Mnemiopsis leidyi) and sea nettles (Chrysaora quinquecirrha 
and Cyanea capillata) were obtained in Tucker Trawls and mid-water trawls. Samples 
Were sorted by hand, rinsed in deionized water and frozen in acid-washed glass jars. 
Finfish were obtained frozen as surplus collections from mid water trawl samples. 
All samples were dried at 60? for 24 hours or until completely dry. Piior to 
drying, sediment samples and mesozooplankton not collected on filter pads were 
partially thawed, then scraped into foil pans. Gelatinous zooplankton were pa,tially 
thawed and transfe1Ted to large aluminum pans for drying. Fish were partially thawed 
and muscle tissue was separated from organs, skin and bones. Only muscle tissue was 
reserved for drying. Dried sediments, mesozooplankton (except on filters), gelatinous 
zooplankton, and fish tissue were ground to a fine powder with a mo11ar and pestle. 
Measurement of 8' 5N via mass spectrometry was performed by the Stable Isotope 
Laboratory at the Marine Biological Laboratory, Woods Hole, MA. Analytical 
replicates for 8' 5N were generally within 0.2%0. Measurements of 813C, obtained from 
the same samples, were previously described by Hagy and Boynton (1998). 
Results and Discussion 
15
Seasonal and regional patterns in 8 N for seston depend in complex ways on 
biogeochemjcal processes. For example, isotope fraction associated with uptake of 
dissolved inorganic nitrogen (DIN) by phytoplankton averages -4 to -6%0 under N-
replete (i.e. DIN not JimWng) conditions but proceeds without fractionation when N 
15
limits the phytoplankton (Peterson and Fry 1987). Moreover, 8 N in the remaining 
15
DIN pool increases because 14N uptake exceeds N uptake (Cifuentes et al. 1988). 
The dynamic nature of 15N/'4N for DIN and phytoplankton in Delaware Bay that was 
observed by Cifuentes et al. (1998), it was anticipated the data collected in this study 
could not resolve the equivalent dynamics in Chesapeake Bay. Nonetheless, the 
patterns observed for seston 8'sN in Chesapeake Bay generally followed the patterns 
315 
t hal mi ght be expected. To increase the ability to resolve regional and seasonal 
pallerns, samples from several years were combined (Table 7-2). Values were lowest 
in the upper Bay during spring (-6%0) and increased both down Bay and from spring 
lo summer. Statistica\\y, 815N was lower in spring than in summer and fa\\ and higher 
in lhc mid Bay than in the upper and lower Bay (Table 7-2) 
15
Seasonal and regional patterns in 8 N for mesozoop\ankton loosely reflected 
lhc patterns observed in the particulate nitrogen (PN , Tab\e 7-3). Entering 815N for 
m esozoop\ankton and PN into eq. (2) and assuming that TL,.e.Fl (for PN), gave initial 
eslimates of trophic \eve\ for mesozooplankton (Table 7-3). TL of mesozoop\ankton 
was si gnificantly higher (p<0.05) in the upper Bay compared to the other regions, for 
which TL was statistica\\y equa\. TL did not vary significantly among seasons, except 
for in the upper Bay, where TL was significantly greater in summer than in fa\\. In 
oenerating these estimates it was recognized that PN includes components of the 
b 
microp\ankton for which TL,.e.r>l.0. In addition, it includes a large detrital PN 
component for which 815N can be either \ess than or greater than 815N for 
phytoplankton . If the detritus poo\ was recently plankton-derived, 815N may be 
similar to that of the plankton. Additiona\\y, isotopic fractionation during 
decomposition can increase 815N (Cifuentes et a\. 1988). On the other hand, 815N for 
15
sutficial sediment was slightly \ess (0.9%o to l.4%o) than 8 N for PN throughout the 
Bay, a pattern also observed in the Delaware estuary by Cifuentes et a\. (1988). If the 
15
sediment PN poo\, with depleted 8 N, was mixed with suspended PN via 
15
resuspension this would decrease 8 N for the detrita\ component of PN. Variations in 
8' 5N for detritus that is not consumed, relative to phytoplankton and microp\ankton 
that are consumed could affect trophic \eve\ estimates based on PN as the reference. 
The consequence of uncertainty as to the origin and fate of detritus can be 
evaluated in terms of uncertainty in the value of TLref for PN. This was examined 
using estimates of the relative contribution of phytoplankton, bacteria, microp\ankton 
and detritus to PN, as we\\ as estimates of trophic \eve\ for each of these components 
based on network models (Chapter 6). Using this approach, it was estimated that TL,.e.r 
for PN was l. 11 , 1.18 and 1.12 in the upper, mid and lower Bay, respectively. 
316 
Relative to assuming that TL,.e.f=l, this increased the estimated TL for 
mesozooplankton (Table 7-4). The estimated trophic level for mesozooplankton in the 
upper Bay was higher than was estimated using the network model, but lower in the 
mid Bay and lower Bay. As was suggested above, however, 0 15N for detritus may 
have been reduced in the upper Bay due to resuspended sediment and increased in the 
lower Bay due to the preponderance of plankton derived organic matter in the PN 
(Canuel and Zimmerman 1999). These effects would tend to reduce the differences 
between the trophic level estimates derived from network analysis and those obtained 
here. Although uncertainty in estimates of Trefresults in imprecise estimates of TL for 
mesozooplankton, any estimate is bounded by 2.0 at the lower limit. By increasing 
consumption of protozoa to the highest extent permitted by mass balance constraints 
in a trophic network model (Chapter 6), the likely upper limit for the trophic level of 
mesozooplankton is about 3.0. The estimates based on 015N suggest that the average 
TL for mesozooplankton was towards the upper end of the range in the upper Bay and 
the lower to middle end of the range in the mid Bay. Differences in the diet 
composition of the mesozooplankton may have been under-represented in the network 
models (Chapter 5). 
Trophic level estimates for biota other than mesozooplankton were computed 
using eq. 2 with mesozooplankton and the reference, similar to the approach of Kline 
and Pauly (1998). This approach takes advantage of the fact that mesozooplankton can 
be definitively isolated for isotopic analysis. Four different biota, the 
microzooplankton ( 40-200 ?m size fraction of the plankton), ctenophores, sea nettles 
and bay anchovy, were collected in most regions of the Bay and in most seasons, 
enabling trophic levels (?standard en-or) to be computed on a regional and seasonal 
basis (Fig. 7-2). Trophic level estimates for the microplankton were lowest in spring. 
Consistent with observation in the field, this suggests that larger phytoplankton cells 
(or chains of cells) were retained on the sieve. TL estimates for summer and fall 
remained less than estimates for mesozooplankton. Considering that estimated trophic 
levels for ciliates and rotifers based on network analysis were 2.6 and 2.3, respectively 
(Chapter 6), it seems likely the 40-200 ?m size fraction samples included 
317 
phytoplankton or detritus at all times of the year. Particles smaller than 40 ?m were 
likely retained on the sieve due to clogging. 
Ctenophores have been reported to consume a diet consisting almost 
exclusively of mesozooplankton , but also including a small quantity of 
ichthyoplankton (Purcell et al. 2001). As a result, it was expected that their TL would 
be -1 integer step greater than mesozooplankton, or possibly slightly higher. This was 
observed almost exactly in 5 of 8 season/region combinations for which data were 
available (Table 7-2). Ctenophore trophic level was slightly lower than expected in 
th
e upper Bay in spring and the lower Bay in summer. In contrast, ctenophores were 
more than 2 trophic levels higher than mesozooplankton in the upper Bay fall (Fig. _ 
7
2). Although one cannot rule out the possibility that ctenophores consumed a diet at 
such a high trophic level, an alternate explanation seems more likely. As have been 
noted before, isotope fractionation results from preferential excretion of the lighter 
15
isotope. Therefore, maintaining a constant 8 N depends on a dynamic balance 
between consumption and excretion. Under conditions of starvation, excretion 
de I 15P  14 etes N from the animal more rapi? dly t ha n N , w h"I l e J?t  i?s  not replaced via the 
diet. Therefore, 8' 5N increases. Although this process has not been described for 
ctenophores, Rosen (1994) demonstrated this process experimentally for the sea nettle 
Chrysaora quinquecirrha. Thus, it was not surprising to observe that Chrysaora also 
exhibited elevated 8' 5N on several occasions (Fig. 7-2). Contrary to these more 
extreme observations, estimates for the mid Bay in summer, where both ctenophores 
and sea nettles were most abundant, were more consistent with expectation. 
Ctenophores were slightly more than one trophic level higher than mesozooplankton , 
consistent with a diet of mesozooplankton plus a small fraction of ichthyoplankton. 
The estimated trophic level for ctenophores was also in reasonable agreement with 
estimates based on network analysis. Similarly, the average trophic level of 
Chrysaora was about one level above ctenophores, reflecting their expected diet of 
ctenophores. The trophic level of Chrysaora in the mid Bay was slightly higher than 
Would be suggested by an exclusive diet of ctenophores or mesozooplankton (Fig. 7_ 
2) ? A lth ough  Ch rysaora 1? s known to consume Bay anchovy eggs and larvae, this 
318 
15
would not increase its 815N as compared to an exclusive ctenophore diet because 8 N 
for bay anchovy was similar to that of ctenophores (Fig. 7-2) . As elsewhere in the 
Bay, elevated 8 15N for Chrysaora simply indicate differences in isotope fraction, due 
to starvation or other causes. The observed 815N for Chrysaora was hi gher than 
values reported by Montoya et al. (1990). The estimates from both studies were 
difficult to fully explain in te1ms of trophic interactions. 
Trophic level estimates for bay anchovy were highly conserved throughout the 
Bay, averaging almost exactly one trophic level higher than mesozooplankton, their 
principle diet (Klebasco 1991). During summer in the mid Bay, trophic level 
estimates for bay anchovy were the same as was predicted by network analysis (Fig. 7-
2.). The exception , as with ctenophores and sea nettles was the upper Bay in fall. 
Here, bay anchovy were -1 .5 trophic steps higher than mesozooplankton. The reasons 
for this are uncertain since there is little evidence that anchovy consume alternate 
diets , other than small quantities of ciliates. 
Average trophic level estimates were computed for 27 taxa or groups of taxa 
for which samples were obtained an irregular basis (Table 7-5, regularly sampled biota 
also included). In all cases, trophic level was computed with mesozooplankton as the 
reference ( eq. 2). Average trophic level was found to vary between -2 and 5 .0 and on 
average was 0.5 trophic steps higher than estimates derived from trophic flow 
networks (Table 7-5). The suspension feeding polychaetes Chaetopterus variopedatus 
(collected in lower Bay) and Pectinaria gouldi (collected in mid Bay) were both very 
close to TL 2, consistent with the predicted TL. However, Mulinia Lateralis and 
Macoma balthica (which can also deposit-feed) had a higher trophic level (3.0) than 
expected (2.1). Since these bivalves were collected in the upper Bay during fall , 
explanations of inflated TL similar to those for mesozooplankton in the upper Bay 
may apply here as well. Estimated trophic level for Atlantic menhaden was much 
greater than the prediction from network analysis, which was based on the assumption 
that these fish were primarily herbivorous . However, it is known that menhaden 
consume zooplankton as juveniles and the results obtained here were more consistent 
with that than with herbivory (Table 7-5). TL estimates for blue crab, spot and 
319 
Atlantic croaker, all demersal feeders, were between 3.57 and 3.88 and were 
consistently a fractional trophic level higher than the 3.2-3.4 estimates obtained from 
the network analysis (Table 7-5). Average trophic level for weakfish was 3.92, while 
the average for white perch and striped bass was slightly higher. Estimated TL for 
these taxa based on 815N was nearly a full trophic level higher than the estimates based 
on network analysis, but was within reasonable statistical uncertainty. Interestingly, 
the estimated trophic level for Cyathura, an isopod that was found in the gills of fish, 
had a 815N suggesting that it had obtained its diet at a very high trophic level. 
Although Cyathura could conceivably have consumed tissues from its host, that would 
have resulted in higher 815N and TL higher than the host, contrary to observation. An 
alternative possibility is that Cyathura was a cleaning symbiont. 
Analysis of stable nitrogen isotopic composition in Chesapeake Bay food webs 
suggested that the average TL for mesozooplankton (TLs1) in the upper Bay was 
higher than in other regions of the Bay. Although this pattern also appeared in the 
estimates based on network analysis (TLNA), it was much more pronounced here. 
Over a large number of taxa, TLNA was 0.53?0.10 (?se) less than TLs1. TLs1 was most 
similar to TLNA when the number of 815N observations was large. Stable isotope 
analysis appeared to provide an interesting independent test of TLNA, but often the 
likely uncertainty in TLs1w as as wide as the plausible range for TLNA unless there is 
no a priori knowledge of diet composition. In addition, in a few instances where TLs1 
was substantially different from TLNA, complex biogeochemical explanations could 
sometimes be found, undermining the utility of TLs1 as a definitive test of the network 
model. 
Literature Cited 
Canuel, E. A. and A. R. Zimmerman. 1999. Composition of particulate organic matter 
in the southern Chesapeake Bay: Sources and reactivity. Estuaries. 22(4), p. 
980-994. 
Christensen, V., C.J. Walters, and D. Pauly. 2000. Ecopath with Ecosim: A user's 
guide. October 2000 edition. Fisheries Centre, Univ. of British Columbia, 
Vancouver, Canada and International Center for Living and Aquatic Resources 
Management,, Penang, Malaysia. 
320 
Cifuentes, L. A., J. H. Sharp, and M. L. Fogel. 1988. Stable carbon and nitrogen 
isotope biogeochemistry in the Delaware estuary. Limnol. Oceanogr. 33(5): 
1102-1115. 
Fry, B. 1988. Food web structure on Georges Bank from stable C, N, and S isotopic 
compositions. Limnol. Oceanogr. 33(5): 1182-1190. 
Hagy, J. D . and W . R. Boynton. 1998. Seasonal and spatial patterns in stable isotopic 
composition of Chesapeake Bay food web components. Newsletter of the Land 
Margin Ecosystem Research Program. Land Margin Ecosystems Research 
Coordination Office, The Ecosystems Center, 7 MBL Street, Marine Biological 
Laboratory, Woods Hole, MA 02543. 
Hartman K. J., and S. B . Brandt. 1995. Trophic resource partitioning, diets, and 
growth of sympatric estuarine predators. Trans. Am. Fish. Soc. 124 (4): 520-
537. 
Hollis , E. H. 1952. Vaiiations in the feeding habits of the striped bass, Roccus 
saxatilis (Walbaum), in Chesapeake Bay. Bull. Bing. Ocean. Coll. 14: 111 -
131. 
Iverson, R . L. 1990. Control of fish production. Limnol. Oceanogr. 35(7): 1593-
1604. 
Klesbasco, M. J . 1991. Feeding ecology and daily ratio of bay anchovy, Anchoa 
mitchilli , in the mid-Chesapeake Bay. M. S. Thesis. University of Maryland 
College Park. College Park, MD. 
Kline, T. C., Jr. and D. Pauly. 1998. Cross-validation of trophic level estimates from 
a mass-balance model of P1ince William Sound using 15N/14N data. pp. 693-
702. In: Funk, F., T. J. Quinn II, J. Heifetz, J. N. Ianelli, J . E. Powers, JF. 
Schweigert, P. J . Sullivan, and C.-I. Zhang (eds.) Fishery Stock Assessment 
Models. Alaska Sea Grant College Program Report No. AK-SG-98-01. 
University of Alaska, Fairbanks. 
Montoya, J.P. , S. G. Horrigan, and J. J . McCarthy. 1990. Natural abundance of 15N 
in particulate nitrogen and zooplankton in the Chesaepake Bay. Mar Ecol. 
Prag. Ser. 54: 35-61. 
Odum, W. E. and E. J . Heald. 1975. The detritus-based food web of an estuarine 
mangrove community. pp. 265-286. In: L. E. Cronin (ed.) Estuarine Research, 
Vol l. Academic Press, New York. 
321 
Pauly, D. , V. Christensen, J . Dalsgaard, R. Froese, and F. Torres, Jr. 1998. Fishing 
down marine food webs. Science. 279: 860-863 . 
Peterson, B. J. and B. Fry. 1987. Stable isotopes in ecosystem studies. Ann. Rev. 
Ecol. Syst. 18: 293-320. 
Purcell, J.E., T . A. Shiganova, M. B. Decker, and E. D. Houde. 2001. The 
ctenophore Mnemiopsis in native and exotic habitats: U. S. estuaries versus 
the Black Sea basin. Hydrobiol., in press. 
Rosen , R. L. 1994. Shrinkage of Chrysaora quinquecirrha: isotopic fractionation as 
evidence for starvation. Honors Thesis. Harvard University, Cambridge, MA. 
Ryther, J. H. 1969. Photosynthesis and fish production in the sea. Science. 166: 72-
76. 
Ulanowicz, R. E. 1995. Ecosystem trophic foundations: Lindeman exonerata. pp. 549-
560 in B.C. Patten and S. Jorgensen (eds.) Complex Ecology: The Part-Whole 
Relation in Ecosystems. Prentice-Hall, Englewood Cliffs, New Jersey. 
Ulanowicz, R. E. 1999. NETWRK 4.2a: A Package of Computer Algorithms to 
Analyze Ecological Flow Networks. University of Maryland Center for 
Environmental Science. Chesapeake Biological Laboratory. UMCES Ref No. 
82- 7 CBL. 
Ulanowicz, R.E. and W.M. Kemp. 1979. Toward canonical trophic aggregations. 
Amer. Nat. 114:871-883. 
322 
Table 7-l. The date of the sampling cruises during which organic matter 
samples were collected for stable isotope analysi s. 
Season (Cruise Id.) Dates 
Fall 1995 (95-4) 10/27-11/3/1995 
Spring 1996 (96-1) 4/27-5/7 / 1996 
Summer 1996 7/  17-7/26/1996 
Fall 1996 10/22-11/ 1/ 1996 
Spring 1997 4/20-4/24/ 1997 
Summer 1997 7/  11-7/15/1997 
Fall 1997 10/29-11/5/ 1997 
Spring 1998 4/ 11-4/ 19/ 1998 
Summer 1998 8/4-8/12/ l 998 
Fall 1998 l 0/ 19-10/23/ l 998 
Table 7-2. Mean (?standard error) 815N in for each region/season 
combination. Letters indicated in parentheses show statistical 
significance of differences between main effect means. Classes not 
sharin the same letter statistical ! different ( <0.05). 
Seston 
Upper Bay (a) 6.6 (0.42) 8.0 (0.48) 7.9 (0.79) 
Mid Bay (b) 9.6 (0.97) 10.4 (0.31) 10.9 (0.24) 
Lower Bay (a) 7.3 (0.56) 8.9 (0.58) 7.9 (1.3) 
Mesozooplankton 
Upper Bay (a) 12.0 (0.15) 13.8 (0.66) 12.33 (0.58) 
Mid Bay (b) 12.3 (0.73) 13.8 (0.79) 14.2 (0.50) 
Lower Bay (a) 11.7 (0.13) 12.5 (0.53) 12.6 (0.24) 
323 
Table 7-3. Estimated trophic level (?standard error) for 
mesozooplankton based on PN as the reference sample. 
Region Spring Summer Fall 
(A) 
Upper Bay 2.64 (0.14) 2.74(0.11) 2.20 (0.18) 
Mid Bay 1.84 (0.09) 2.02 (0.25) 2.05 (0.15) 
Lower Bay 2.29 (0.13) 1.91 (0.50) 2.04 (0.09) 
Table 7-4. Estimated reference trophic level for mesozooplankton (used in eq. 2) and 
the resulting trophic level estimates for mesozooplankton based on 15N/14N . For 
comparison, the estimated trophic level for mesozooplankton computing from trophic 
flow networks (Chapter 6) are also shown. Values are for summer (June-August) 
only. 
Region TLre( TLs, TLNA 
Upper Bay 1.11 2.85 2.6 
Mid Bay 1.18 2.20 2.5 
Lower Bay 1.12 2.03 2.5 
324 
----
Table 7-5. Trophic level estimates, referenced to mesozooplankton, for 27 taxa or 
groups of taxa, arranged in order of increasing trophic level estimate. Trophic level 
estimates for the most appropriate node of a trophic network model are presented 
for comparison. YOY=young of the year age class, 1+  = 1+  age class. 
N TLs, SD TLNA TLNr 
Variable TLs1 
16 1.82 0.40 
Microplankton 1.88 2.1 0.22 l 
Chaetopterus variopedatus (SF benthos) 2 1.98 0.02 2.1 
0.12 
Pectinaria gouldi (SF Benthos) 2 2.91 0.00 2.1 
-0.81 
Amphipods (unc lassified) 2.91 2.1 -0.81 1 
Mysids (unclassified) 3 3.00 1.10 2.1 
-0.90 
Mulinia lateralis 2 3.05 0.67 3.4 0.35 
White Perch (YOY) 6 3.05 0.42 3.6 
0.55 
Herrings and Shads 2 3.08 0.07 2.1 
-0.98 
Macoma balthica 16 3.55 0.65 3.5 
-0.05 
Mnemiopsis leidyi l 3.55 3.2 
-0.35 
Atlantic Croaker (l+) 4 3.56 0.73 2.1 -1.46 
Atlantic Menhaden 4 3.57 0.80 3.2 
-0.37 
Blue crab -0.40 2 3.60 1.14 3.2 
Spot 3.69 0.43 3.4 -0.29 17 
Bay Anchovy -0.65 1 3.85 3.2 
Atlantic Croaker 3.88 3.2 -0.68 l 
Atlantic Croaker (YOY) 7 3.92 0.74 
3.9 -0.03 
Weakfish (YOY) l 4.14 
Cyathura spp. (parasitic isopod) 4.16 0.15 2.5 -1.66 2 
Polychaetes (unclassified) 4.29 3.4 -0.89 1 
4.35 0.49 3.6 -0.75 White Perch 3 
4.46 0.97 3.6 -0.86 Striped Bass (YOY) 3 
3.4 -1.06 
Striped Bass (1 +) l 4.46 
White Perch (l+) l 4.52 
Striped Anchovy (Anchoa hepsetus) 4.88 4.4 -0.48 1 
-0.62 
Cyan.ea capillata (nettle) 8 5.02 0.81 
4.4 
Chr saora uin uecirrha (nettle) 
325 
39 30' 
39 00' 
38 30' -
38 00' 
37 30' -
37 00' -
76 30' 76 00' 
Fig. 7-1. The three general regions of the 
Chesapeake Bay within which collections 
of organisms and organic matter were com-
pleted for stable isotope analysis. 
326 
-
f",,,J 40-200 ?m size fraction Spring 
6 I I Ctenophor
c s 
I JSea nett le 
5 D Bay ancho vy Cl) 
> 
j 4 
u ... ~. - ~ 
-g_ 3 8-
,0._  
f- 2 
~ -L -- ..I... 
- - -
1 
Estimates based Summer 
? on network model 
6 
Cl) 5 
> 
Cl) 
_J 4 
u 
.c. 
,g._- 3 
f-
2 
1 
Fall 
6 
Cl) 5 
> 
Cl) 
_J 4 
u 
.c. 
,g._- 3 
f-
2 
1 
Upper Mid Lower 
Bay Bay Bay 
Fig. 7-2. Average (?standard error) Lro1:hic level (TL) estimates for the 40-200 
?m plankton size fracti on, ctenophores (Mnemiopsis leidyi) , sea nettles 
(Chrysaora quinquecirrha), and bay anchovy (Anchoa mitchilli) during spring, 
summer and fa ll. Heavy references lines indicate the TL estimate for 
mesozooplankton , which was used as a reference (eq . 2). Dotted lines indicate 
integer trophic steps from the TL or mcsozooplankton. Summer TL estimates 
based on network analysis were obtained from Chapter 6. 
327 
...... 
Appendix : 
PARAMETERIZING THE SUMMER TROPHIC FLOW NETWORKS 
FOR CHESAPEAKE BAY 
This appendi x provides supporti ng methodological detail s and discussion 
relating to the estimation of the parameters of the trophic flow networks presented in 
Chapter 6. T he appendi x is organized into sections based on sub-sections of the 
trophic fl ow networks (e.g. primary production , zoop lankton , benthos, nekton). 
Organic Storage, Primary Production and Bacteria 
The major sources of organic matter for Chesapeake Bay food webs include 
ex ternal inputs of parti culate and di sso lved organi c carbon from rivers and 
phytopl ankton production. Kemp et a l. (1997) estimated that these sources account 
for 98.7% of their 3 .96 10 12 gC estimate of the annual organic matter supply, with 
92% of the total coming from phytoplankton production . The remainder was 
estimated to come from submersed and emergent vascul ar pl ants (1.3 %) and 
atmospheric inputs (<l %). For thi s study, microphytobenthos was also considered as 
a possible source of organic matter, which increases the total input estimate sli ghtly. 
Total phytoplankton production was divided into net phytoplankton and autotrophic 
picoplankton. Because all ca lcu lations were made to pertain only to summer, net 
changes in organic matter storage is also possible. A net decrease in organic matter 
328 
st
orage (e.g. in sedi ments) amounts to a detritus subsidy of the summer food web by 
the spring food web. 
f.OC and DOC Storage and Physical Transport 
Kemp et a l. (1997) describe an annual organic carbon budget for Chesapeake 
Bay. Organic matter inputs come from the Susquehanna Ri ver, other t1ibutary rivers, 
below fa ll-line sources, and atmospheric deposits. Because the present study was 
limited to summer, new estimates were made using these estimates as a general guide. 
Summer average total organic carbon loads from the Susquehanna River at 
Conowingo Dam were obtained from the USGS River Input Monitoring Program 
(Langland et al. 2001) and apportioned into particulate and dissolved fractions 
according to the ratio of summer average particulate organic carbon (POC) and 
di ssolved organic carbon (DOC) at the Chesapeake Bay Water Quality Monitoring 
Program station located adjacent to the Susquehanna Ri ver mouth (station CBl.l ). 
From 1992 to 1995 POC and DOC at this station averaged (?sd) 2.8?0.42 and 
' 
0.95?0.23 mg r' , respectively, giving an -3:1 ratio. The summer average loading of 
Toe to the upper Bay from the Susquehanna Ri ver during 1985-1999 was 165?49 
tons d-1 (tons=l06 g). Therefore, POC and DOC inputs were estimated to be 41 and 
124 tons d-1, respectively. The estimated summer TOC input rate is 40% of the rate 
reported by Kemp et al. (J 997), reflecting TOC input from the Susquehanna River 
below the annual average. 
Transport of organic carbon between the regions and between the lower Bay 
and the coastal ocean is difficult to calculate reliably, especially for particulates for 
329 
which transport depends not onl y on patterns of water transport, but also on sinking, 
depositi on to sediments, and resuspension. In part, these dynamics explain the strong 
tendancy fo r estuari es to retain particles, most notably the upper Bay, within which it 
has been suggested that nearl y all parti cul ates entering from the Susquehanna river are 
retained (e.g. Schubel and Carter 1977, Hobbs et al. 1992). Recent studies also show 
that in the upper Chesapeake Bay thi s retenti on mechanism can extend to biological 
parti c les such as zoopl ankton (Roman et al. 200 1) and fish eggs and larvae (North 
2001 ). It was assumed here that th roughout the ti dal Chesapeake Bay system, all 
seaward POC transport during summer was balanced by an equal landward POC 
transport. Thi s inc ludes transport between the regions and between the mainstem 
Chesapeake B ay and the tidal trbutari es. 
DOC can be assumed to remain relati vely well mixed within the mixed layers 
of the water column . T herefore, the DOC transport was estimated as the product of 
average advecti on rates between the regions, estimated using a box model (Chapter 1), 
and average DOC concentrations in the advected water. DOC export from major 
tributari es was estimated as the product of surface water DOC concentration and the 
net outflow of water. For the upper Bay, summer advection from the upper Bay into 
the mid Bay averaged 1127 3 f 1 111 during 1986-1998. Landward advection from the 
mid Bay accounted fo r 593 3 1111 s? , while freshwater inputs accounted for the 
remainder, 98% of which came from the Susquehanna River. The outflowing water 
had a summer average DOC concentration of 2.84 mg rt , while the landward 
advectin g water had average concentration of 2.57 mg rt . Therefore, the net DOC 
transport to the mid B ay amounts to 135 tons d-1? 
330 
The average surface layer advection from the mid Bay to the lower Bay duting 
summer was estimated to be 4661 m3 s- 1, while the landward advection in the bottom 
layer was 3846 m3 s-1. Net advection from the upper Bay plus the Potomac River 
discharge (above and below fall line, Chapter 1) account for 91 % of the difference, the 
remainder coming from smaller tributaries. The seaward and landward flowing water 
had average DOC concentrations of 2.88 mgr' and 3.06 mgr', respectively, giving a 
net DOC transport to the lower Bay of 199 t d-1? Freshwater discharge from the 
Potomac River averaged 207 m3 s-1 in summer, while the DOC concentration in 
outflowing surface waters in summer averaged 2.77?0.34 mg r1. Therefore the net 
advective output from the Potomac River estuary was 49.5 tons d-1? Combined with 
the DOC exported from the upper Bay, this accounted for 93% of the export. The 
remaining external DOC sources were not estimated and may be smaller or greater 
than the remaining 7%. 
Advection from the lower Bay to the coastal ocean was estimated to average 
10152 m3 s-1 du1ing summer. Landward advection was 9058 m3 f I and the net 
advection from the mid Bay was 815 m3 f 1, accounting for 97% of the seaward flow. 
The combined discharge of the James and Appomattox Rivers was 170 m3 f 1, 60% of 
the remainder. The average DOC concentration in the waters advecting into the 
coastal ocean was 2.2 mgr' , while landward flowing water had a DOC concentration 
of 1.77 mgr', giving a net DOC export of 586 tons d-1? This summer rate is 25% less 
than the annual average carbon export rate which can be computed from Table 5 in 
Kemp et al. (1997). Since summer export is probably lower than the annual average, 
this is taken to indicate reasonable agreement between these two studies. The average 
331 
-
DOC concentration in the seaward flowing waters of the James River du1ing summer 
was 3.26?0.38 mg r1, giving an average DOC export of 48 tons d-1? Combined with 
the net seaward transport from the mid Bay, the total net inputs amount to 247 tons d-1, 
only 42% of the net export. This implies that the lower Bay ecosystem was either a 
substanti al net source DOC during summer, that a significant source has been omitted, 
or that the export has been signifi cantl y overestimated. 
The storage of POC and DOC in the three regions was computed from the 
Chesapeake Bay Monitoring Program water quality data at a series of 20 stations 
located down the ax is of Chesapeake Bay (Table A- 1). Phytoplankton biomass 
accounted for 30%, 38%, and 30% of POC in the upper, mid and lower Bay 
respecti vely and was subtrac ted from the POC to obtain estimates of detrital POC 
storage (Table A-2) . Bacterial bi omass was also a significant component of the 
suspended POC (Table A-6) , but could not anambiguously be subtracted from the 
POC because the GF!F filters used to measure POC cannot be expected to retain all 
free-livin g bacteria. Therefore, 50% of free-living bacteri al biomass was subtracted 
from POC while the re mainder was subtracted from DOC, a miniscule adjustment to 
the DOC pool (Table A-1). Regional differences in detrital POC and DOC storage 
mostly refl ect diffe rences in mean depth . 
Phytoplankton 
Phytoplankton production (PP) in Chesapeake Bay has been studied in 
extraordinary detail , with at least two studies resolving spatial , seasonal and 
interannual variability (Kemp et al. 1997, Smith 2000, Harding et al. 2001) using a 
332 
large number of observations, each of which was derived from carefully executed field 
assays. Seasonal/regional estimates of primary production rates determined using 14C-
uptake vs. 0 2-evolution methods (Kemp et al. 1997, Harding et al. 2001) do not 
diverge exceedingly, providing some confidence that large errors in specifying the 
average PP can be avoided. However, methodological concerns persist. In particular, 
14C-based methods may underestimate gross PP due inter alia to excretion of 
photosynthate as DOC and phytoplanktonic respiration. Correction for excretion is 
complicated by the potential for DOC to reenter the particulate phase through bacterial 
uptake. In applying 14C-uptake assays, shorter incubations lead to estimates closer to 
gross PP, while longer incubations appproximate net PP. Harding et al. (2001) 
estimated gross PP using a series of 4-6 hr incubations and net PP using a single 24-
hour incubation, finding that gross 14C-PP was -23% higher than net 14C-PP. Using 
concurrent estimates of net and gross PP measured as 0 2-evolution (e.g. Smith 2000), 
Harding et al. concluded that the photosynthetic quotient (=02 produced/CO2 fixed) 
was as high as 1.48 for gross PP, higher than the estimate of Stokes (1996), suggesting 
that the gross PP estimates of Kemp et al. (1997) should be revised downward. Due to 
the ambiguity inherent in using 14C-uptake assays to estimate gross PP, it was assumed 
here that the estimate of Stokes (1996) for PQ (=1.2) was conect. Therefore, 
estimates of gross PP were derived from the summer average net 14C-PP values of 
Harding et al. (2001) by converting to gross 14C-PP (x 1.229), then applying the 
further correction implied by the discrepancy in PQ estimates (1.48/1.2=1.233). Thus, 
gross PP=(l .233)(1.229)(net 14C-PP). The resulting estimates of gross PP should be 
comparable to the summer gross 0 2-PP estimates computed from data in Kemp et al. 
333 
0 997). While there was no obvious bias , it was found that differences among the 
regions persisted, which was interpreted here as natural variability resulting from 
different years being included in the two studies. The final estimates of gross pp 
(hereafter GPP) adopted for this study were those in column (B) of Table A-2. Based 
on these estimates, phytoplankton GPP/B was estimated to be 0.77, 0.84 and 0.86 d-1 
14
for the upper, mid and lower Bay, respectively. Net C-PP per unit biomass provided 
an estimate of regional average phytoplankton growth rates, which were 0.51, 0.56. 
0.57 d-1 for the same regions. 
The fraction of GPP that contributed to phytoplankton growth (i.e. POC 
production) is reduced by extracellular release of photosynthate as DOC and by algal 
respiration. Algal respiration is temperature and size dependent and may depend on 
other factors as well , such as irradiance and taxonomic differences (Tang and Peters 
1995). Algal respiration estimates were obtained from the temperature-dependent 
allometric relationship in Table III of Tang and Peters (1995) and converted to a daily 
rate by multiplying by a 14-hour daily photoperiod and assuming algal respiration is 
related to production, with minimal maintenance respiration at night. Thus, specific 
respiration for microalgae at summer surface water temperatures in Chesapeake Bay 
1
(27? C) was estimated to be 0.20 to 0.25 d- , depending on cell size (see below). 
Baines and Pace (1991) reviewed estimates of extracellular release (ER) of DOC by 
phytoplankton from 16 studies providing a total of 225 observations. Importantly, 
they compared ER to particulate primary production, not GPP. They concluded that 
ER accounted for 12% of particulate production. In marine systems, this fraction was 
independent of the primary production rate and of the phytoplankton biomass (Baines 
334 
and Pace 1991). Malone et al. (1991) aITived at a similar estimate (15%) fo r 
Chesapeake Bay. Expressing particulate production (PP) as PP=GPP-R-ER and 
substituting 0.12PP for ER, it can be shown that 
ER = (0.12/J. I2)(GPP - R) = l0.7(GPP - R) . Assuming a specific respiration equal 
to 0 .2 d -' , extracellular excretion amounts to - 7% of OPP in all the regions of the B ay. 
The resulting estimate of phytoplankton growth (i.e. particulate production) is similar 
14
to net C-PP (Table A-1, Table A-2). 
The relative fraction of production contributed by picoplankton (<3 ?m 
equivalent spherical diameter) varies in Chesapeake Bay from <2 % from February to 
May and increases to a summer maxi mum. Malone et al. (1991) reported that 
picoplankton production accounted for a maximum of 20% of total production in 
summer, while Smith (2000) and Smith and Kemp (2001) estimated that thi s figure 
may be as high as 45%. The difference may be methodological, since Smith (2000) 
used a 3 ?m filter while Malone et al. (1991) used a l ~tm filter. Smith (2000) also 
noted that picoplankton accounted for a smaller fraction of phytoplankton biomass and 
production at shallower stations compared to those in the central channel of the 
estuary. Thus , the fraction of total production contributed by picopJankton adopted for 
this study was toward the lower end of estimates (20%), reflecting the large area of 
shallower waters in the Chesapeake Bay. The fraction of biomass contributed by 
picoplankton during summer was assumed to be 15% of the total, based on 
picoplankton vs. total chlorophyll-a concentrations reported by Malone et al. (199 l) . 
1 
Specific respiration (R) for picoplankton was assumed to be 0.4 d- due to smaller cell 
size (Tang and Peters 1995). Extracellular release of DOC (ER) was estimated using 
335 
the procedure described above, a lthough thi s resulted in slightly lower estimates of ER 
for p icoplankton as com pared to GPP because of the higher estimate of R. All rates 
estimated fo r net p lankton and autotrophi c p icoplankton are summarized in Table A-3 . 
M icrophytobenthos 
T he microphytobenthos are the microscopic a lgae and cyanobacteria that exist 
o n and w ithin the sedime nts . A lthough not as well -studi ed as phytoplankton , recent 
rev iews by MacIntyre et a l. (1996) and M iller et a l. (1996) provide a pe rspective on 
research to date, as well as part of the bas is for estimati ng the production and fate of 
the microph ytobe nthos in the mainstem of C hesapeake Bay duri ng summer. In most 
cases, bio mass (chloroph y ll -a) and li ght avail abili ty were the most important fac tors 
affecting benthi c primary producti on (MacIntyre et al. 1996). Unfo rtunate ly, very few 
measure ments have been m ade in C hesapeake Bay, while Ri zzo and W etzel (1985), 
who examined the microph ytobenthos in the lower C hesapeake B ay, and M ac intyre et 
a l. (1996) both e mphasized th at due to sma ll -scale temporal and spati al vari abili ty , a 
large number of measure ments are needed to characteri ze biomass with precision. 
Therefore, fo r thi s study, emphasis was placed on estimating the avail able li ght fi e ld 
and the di stributio n of sha llow wate r sediments . T he daytime average li ght-saturated 
1 
gross producti on rate (Pmax) during summer was reported to be -80 mg C m-2 h-
(Ri zzo and W etzel 1985), whi ch was converted to a dail y integrated rate o f 1120 mg 
2 1 
m- d- by assuming that thi s average rate was sustained over an average photoperi od 
of 14 hours . Va lues of 1.4, 0.8 and 0.6 m-1 were adopted for the summer li ght 
attenuati o n coeffi c ie nt (kct) in the upper, midd le and lower Bay, respecti vely (Hardin g 
336 
et a l. 2001), a lthough it has been shown that nearshore light attenuation in Chesapeake 
Bay can be very different from open water (Boynton et al. 1999). Daytime average 
incident PAR during summer was assumed to be 800 ?E m-2 d-1 and Pmax was adjusted 
for li ght limitation using the hyperbo li c tangent P-1 function of Jassby and Platt 
( 1976). The light saturation parameter was assumed to be 100 ?E m-2 s-1, which is at 
the low end of the range suggested by Macintyre et al. (1996) but is consistent with the 
results presented by Kemp et al. (1999) . Thus, daily production by microphytobenthos 
in each region was computed as L AzPmaxTANH(I 2 / Ek ) where A2 is the sediment 
z 
area within lm depth intervals and / = / e - k,,z . This model assumes that Pmax does 
2 0
not vary with depth. 
Average production for each region was computed by dividing the integral 
production by the total surface area of the region, which includes a large area of 
aphotic sediments . Thus, the estimated average production rate for each region is 
much lower than the production rate for euphotic sediments. Regional mean gross 
production by microphytobenthos during summer was estimated to be 234, 212, and 
234 mgC m-2 d-1 in the upper, middle and lower Bay respectively. This amount of 
production amounts only 6- 19% of gross phytoplankton production , indicating that the 
microphytobenthos are an important carbon source. As indicated by Kemp et al. 
(1999), the benthic microa lgae may be an even more important source of carbon for 
shallow water habitats (e.g. <4rn) where benthic production is greater and 
phytoplankton production possibly lower. 
337 
The biomass of benthic microalgae is very difficult to quantify directly, for 
reasons outlined by Macintyre et al. (1996) . Therefore, the biomass of benthic 
microalgae was infe rred from the estimate of their respiration, by assuming a mass-
specific respiration rate of 0.2 d-1. Respiration and extracellular DOC release were 
assumed to account for 25% and 7% of gross production, respectively, consistent with 
the rates estimated fo r net phytoplankton (Table A-3) . Thus, the biomass of benthic 
microalgae was estimated to average 265-293 mg C m-2 (Table A-4). As with 
production, the biomass of benthic microalgae is higher if one considers only shallow 
waters. According to this computation, the average biomass in sediments 0-4 m deep, 
converted to chi -a (C:chl-a=SO) averaged 14, 23 and 26 mg chi-a m-2 in the upper, 
middle and lower Bay, respecti vely. The lower Bay estimate is consistent with the 20 
mg m-2 estimate of Rizzo and Wetzel (1985) for a shallow site in the lower 
Chesapeake Bay. Many estimates for the biomass of microphytobenthos are much 
hi gher (MacIntyre et al. 1996). However, this was not considered cause for revision. 
Many of the estimates were for intertidal mudflats which can achieve much higher 
biomass and production due to hi gh li ght levels when exposed at low tide. 
Vascular Plants 
Submersed aquatic vegetation (SAV ) is believed to have been much more 
extensive in Chesapeake Bay in the past than at present. Based on the 1995-1996 
average, approximately 9600 ha in the mainstem Bay were vegetated, accounting for 
-5% of total surface area in the upper Bay, and 1.5% of total surface area in the 
middle and lower Bay (Table A-5, Orth et al. 1997). Since the potental SAV habitat is 
338 
probably much less than the entire bottom, this overstates the possible extent of 
degradation, but this small fraction illustrates the limited contribution of SAV to total 
primary production in Chesapeake Bay. That said, SAV  communities are among the 
most productive plant communities in the world (Duarte and Chiscano 1999), 
justifying some further consideration. 
The SAV communities in mainstem Chesapeake Bay include a mixed 
assemblage of freshwater and low salinity species (e.g. Vallisneria americana, 
Myriophyllu111 spicatum) in the upper Bay. The mid Bay community included Ruppia 
maritima and Zostera rnarina, but for this study was assumed to be dominated by Z. 
marina since most of mid Bay SAV was al the southern extreme of the region (Orth et 
al. 1997). The lower Bay community was clearly dominated by Zostera marina. For 
the upper Bay, the biomass in vegetated areas was estimated to average 120 g dry wt. 
m-2, of which 20% is below-ground biomass (Kemp et al. 1984). For the lower Bay, 
the biomass density in vegetated areas was estimated to average 300 g dry wt. m-2 , of 
which 33% is below-ground biomass. This biomass estimate is at the upper end of 
estimates for Z. ,narina communities in Chesapeake Bay reported by Wetzel and 
Penhale (1983), well within the range reported by Orth and Moore (1986) and at the 
low end for Z. marina wordwide (Duarte and Chiscano 1999). The carbon content of 
SAV was assumed to be 45% of dry weight. 
Kemp et al. (1986) estimated a maximum summer SAV production rate of 1.3 
2 1 
g 0 2 m- d- for upper Bay SAV communities. This was converted to carbon units 
assuming PQ= l.2 to obtain a rate of 406 mgC m-2 d-1. An estimate of production for 
the middle and lower Bay was derived from Fig. l in Duarte and Chiscano (1999) by 
339 
assuming that Z. marina beds in Chesapeake Bay were slightly less productive than 
most (i.e . the mode). T hus , above and belowground production was estimated to total 
2.8 g dry wt. m-2 d-1, or 1260 1mgC m-2 d-1. The turnover rate amounts to -0.01 d- , 
consistent with rates reviewed by Cebri an (1999). Although this production rate is 
comparable to that of phytop lankton, the contribution to total primary production is 
minimal si nce SA V communi ties account for onl y a small fraction of the total area of 
each region (Tabl e A-5). 
P lanktonic Bacteri a 
T he bacterioplankton and protozoa are an important part of the food web, 
providing a trophic pathway between dissolved organic carbon (DOC) and metazoans 
as well as di ss ipating considerable quantities of energy from the ecosystem. Bacterial 
biomass is commonl y estimated using acridine orange direct counts (AODC), whi ch 
prov ides an estimate of bacteri al abundance. D ucklow and Shi ah (1993) and Shi ah 
and Ducklow (1994) estimated bacterial abundance in surface and bottom waters at a 
series of stations along the ax is of Chesapeake Bay and reported averages for the 
upper, middle and lower Bay. Observations were co llected thorughout the year during 
1990 and 199 1. Ducklow and Shi ah (1993) converted bacteri al counts to carbon 
biomass using the factor 2 x 10-14 gC celr 1 and thi s was adopted here as well. An 
assay utili zing a fluorogenic tetrazoliurn dye (CTC) has been used to show that much 
of the bacte ri al community in Chesapeake Bay is not active, a result which has been 
repeatedl y observed elsewhere as well (S mith 1998). The largest fraction of CTC-
active bacteri a (27%) was fo und in summer, while in other seasons an even larger 
340 
fractio n was inactive (Smith 1998). These resu lts indicate that metabolic rates for the 
bacteriopl ankton as a whole should not be based on cell -specific rates. This 
suggesti on was heeded, as rates were based on direct measurements. However, 
estimates of bacteria l biomass utili zed for thi s study were not adjusted to reflect on ly 
CTC-acti ve ce ll s. CTC-inacti ve cells were considered li vi ng biomass (rather than 
detri tus) because at least one study has demonstrated that these cell s can become CTC-
active under appropri ate environmenta l conditi ons (Choi et al. 1999). 
Bacterial production is commonl y quantified by measu1ing uptake of tritiated 
(3H) thymidine (TdR) or leucine during short incubations. Rates based on TdR 
obtained from Shi ah and Duckl ow (1994) and Ducklow and Shi ah (1993) were 
utili zed fo r thi s study. TdR assays were perfo rmed at the same locations and times 
that AODCs fo r bacteri a l abundance were perfo rmed. For thi s study, a factor of 2 x 
18 
10 cell s produced per mole of th ymidine incorporati on (Duckl ow and Shi ah 1993), 
combined wi th the above bi omass per cell , was used to convert TdR to carbon 
producti on rates. Hourly rates were extrapo lated to dail y rates by assuming constant 
bacteri a l production. 
Summer bacteri al bi omass was much hi gher in the mid Bay (2710 mgC m-2) 
than in the other regions of the Bay, despite the fac t that bacteri al abundance was 
reduced in bottom waters in the mid Bay, but not e lsewhere (Table A-6, Shiah and 
Duckl ow 1994). Simil ar measurements reported by Duckow and Shi ah (1993) for the 
1980's indicated sli ghtl y lower, but still high bacteri al biomass (2059 mgC nf\ 
Summer average bacteri al producti on in the mid Bay was 1633 mgC m-2 d-1, giving an 
average turnover time of 1.25 d. Because many ce ll s are inactive (e.g. Smith 1998), 
341 
turnover times for actively growing cell s was likely much higher. Bacterial biomass 
in the upper Bay averaged 709 mgC m-2, while production was 390 mgC m-2 d- 1, 
giving an average turnover time of 1.8 d, somewhat slower than in the mid Bay. 
Bacterial biomass and production in the lower Bay was 924 mgC m-2 and 557 mgC m-
1
2 d- , indicati ng that biomass turned over every 1.7 d (Table A-6) . The spatial pattern 
of bacterial biomass and production mirrored that of phytoplankton production, 
perhaps refl ecting the coupling between plankton production and respiration (e.g. 
Smith 2000). 
Bacteri al consumption and respiration was estimated via an assumed bacterial 
growth efficiency (GE). de] Giorgio and Cole (1998) found in a review that this value 
increased with bacterial production rates from a minimum value toward an asymptotic 
value of - 0.5 when bacterial production reached 5 ?gC r 1 h-1 according to 
0.037 + 0.65P 
GE =---+- --. Using surface water bacteri al production rates in this equation, 1.8 p 
GE for Chesapeake Bay during summer was to be 0.43, 0.52, and 0.39 in the upper, 
mid and lower Bay, respectively. These rates are slightly higher than the average 
(0.30) estimated for the York River, VA on a summer cruise (Raymond and Bauer 
2000), but were within the range of rates observed (8-56%). Smith (2000) estimated 
that growth effi ciency was 24-40% during summer, somewhat lower than suggested 
here and indicating that a downward revision of the above estimates would be 
reasonable. Bacterial consumption and respiration was derived from estimated 
production using these growth efficiencies (Table A-6). The same growth efficiencies 
were used for both particle-attached and free-li ving bacteria. 
342 
The Plankton Community 
The plankton community has been conceived to consist of 5 living 
components: phytoplankton , bacte1ia (free and attached), cili ates, heterotrophic 
microflagellates, rotifers, meroplankton, mesozooplankton, ctenophores (principally 
Mnemiopsis leidyi ) and the sea nettle , Chrysaora quinquecirrha. The biomass and 
bioenergetics of phytoplankton and bactetia in Chesapeake Bay have already been 
described. The remaining components, consisting of the heterotrophic eukaryotic 
protozoa and metazoans, encompass an approximate size range of 2 ?m up to a meter 
or more in length (Chrysaora). The literature desc1ibing mixotrophy in Chesapeake 
B ay was considered; however, it was decided that the biology and particularly the 
quantitative importance of thi s process was too poorly known to consider in this study. 
Heterotrophic microflagellates are typically 2-10 ~lm. The ciliates include small 
oligotrichs, scuticociliates, the larger tintinnids. The microzooplankton, - 20-200 ?m , 
include a community of micrometazoa such as rotifers, copepod nauplii , meroplankton 
and smaller copepodids. Of these, the rotifers and meroplankton were considered 
separately, while the nauplii and copepodids were evaluated as a component of the 
mesozooplankton via a stage-structured analysis. Including nauplii , copepodids and 
adult copepods in a single node was necessary to avoid an error in network analysis 
procedures for estimating trophic levels that would be caused by improperly 
representing non-trophic mass-transfers among nodes as predation . 
The Chesapeake Bay Monitoring Program conducted monthly surveys of 
microzooplankton abundance at a series of stations down the axis of Chesapeake Bay 
since 1984 in Maryland and since 1993 in Virginia. The Maryland and Virginia 
343 
components of the program used different sampling protocols, affecting the suitability 
of the data for use here. W hereas the Virginia monitoring program enumerated all the 
organi sms that passed through a 202 ~lm mesh by settling the preserved whole water 
sample, the Mary land moni toring program enumerated all the organisms that passed 
through a 202 ?m mesh but were retained on a 44 ?m mesh (EPA 2000). The latter 
approach is not adequate for sampli ng the protozooplankton because smaller and 
re lative ly fragile organi sms are either destroyed or wi ll pass through the larger mesh 
size utili zed (Dolan 199 l a, G iffo rd and Caron 2000). Therefore, the Maryland 
Chesapeake Bay Moni tori ng Program data were used onl y to quantify abundances of 
metazoopl ankton, while the Virgini a data were also used for estimating ci liate 
abundance. Data fro m one station in each region of the Bay were analyzed to estimate 
summer average abundances, which were converted to biomass using the most taxon-
speci fi c fac tors avail able. Above- and below-pycnocl ine averages in each region were 
combined using appropri ate volumes to compute integrated water column biomass . 
Average abundances fo r the other groups were obtained from literature values. All 
biomasses and estimated rates for the zoopl ankton community are summari zed in 
Table A-7 . 
Heterotrophic F lagell ates 
The above- and below-pycnocline abundance of heterotrophic fl agell ates (H-
fl ag) in the mid Bay was estimated to average 4 .5 and 0 .8 109 m-3, respectively, during 
summer based on monthly data reported by Dolan and Coats (1 990). Using 
appropri ate mi xed layer volumes , depth -i ntegrated H- fl ag abundance was 43 109 m-2 . 
344 
Based on the assumption that these organisms have an equivalent spherical diameter 
of 2- l O ~un, the biovolume per individual was estimated to be 4-104 ?m3, which was 
converted to carbon using the factor 100 fgC 3 ~u11 reported by B~rshei m and Bratbak 
( I 987). A mid-range diameter of 5-6 ?m gave a biomass estimate of 3.5 pg C celr1? 
Thus, the summer biomass of H-flag was estimated to be 149 mgC m-2 (Table A-7) . 
Fenchel (1982) found that H-flag growth increased with bacterial abundance, 
reaching maximal rates of 0.15-0.25 h-1 at saturating at abundances of 50-100 109 r1. 
Thus, the hi ghest bacterial abundances in Chesapeake Bay are sub-saturating, 
suggesting sub-maximal growth rates for H-flag. Moreover, Dolan and Coats (1990) 
reported that in natural assemblages H-flag was related to bacterial production, rather 
than bacterial abundance. For thi s study, it was assumed that in situ growth rates were 
0.08 h-1, less than the maximal rates. This is equivalent to a doubling time of -9 h or a 
daily specific production of 2.8 d-1? Consumption , respiration, and egesta (Table A-7) 
were computed using estimates of gross and net growth efficiency of 0 .35 and 0.60 
(Fenchel 1982). 
No direct estimates of H-flag abundances were available for the upper and 
lower Bay. However, Dolan and Coats (1990) reported that H-flag abundance 
generally con-elates with bacterial production. The present estimates of H-flag 
consumption in the mid Bay accounts for 73% of estimated bacterial production, 
suggesting that H-flag are the major bactivores. H-fl ag consumption for the upper and 
lower Bay was computed by assuming that the same fraction of bacterial production 
was consumed. H -flag biomass and corresponding bioenergetic rates were computed 
using the same gross and net growth efficiencies (Table A-7). In all three regions of 
345 
the Bay, H-flag biomass was equal to - 5% of bacteri al biomass, at the low end of the 
range reported by Linley et al. (1983) for coastal waters. 
Ciliates 
The above- and below-pycnocline biomass of c iliates in the mid Bay was 
estimated to average 14 and 8 mgC m-3, respecti vely, during summer based on 
monthly data reported by Dolan and Coats (1990). Using appropriate mixed layer 
volumes, depth-integrated ciliate abundance was 147 mg C m-2. Ciliate abundance in 
the upper Bay is poorly known compared to elsewhere in the Bay and has been 
assumed here to be equal to that of the mid-Bay. For the lower Bay, the Chesapeake 
Bay Monitoring Program data give a summer average biomass of 13 mg C m-2 or 87 
2
mgC m- , of which equal fractions of the biomass were contributed by tintinnids and 
oligot1ichs (Table A-7). 
Dolan (1991b) estimated summer growth rates of ciliates in mesohaline 
Chesapeake Bay to be -0.09 h-1, from which a daily production rate of 3.1 d-1 can be 
derived. Verity (1985) reported the food and temperature dependence of tintinnid 
growth , suggesting maximal growth rates of 2.9 d-1 at 25?C. Verity (1991) gave 
slightly lower rates (0.8 to 2.3 d-1 depending on di et) for expe1iments conducted at 
20?C. For this study a growth estimate of 2.5 d-1 was utilized for ciliates throughout 
Chesapeake Bay, reflecting the high water temperature in Chesapeake Bay during 
summer. Gross growth efficiencies for both tintinnid and oligotrich ciliates have been 
estimated to be -50% (Verity 1985 , 1991). Verity (1985) estimated the carbon 
assimi lation efficiency for tintinnids to be 63 % and thi s value was applied to all 
346 
cili ates. Consumpti on, respiration and excretion plus egestion was computed from 
production usi ng these va lues (Table A-7). 
The diet composition of cili ates depends criticall y on the type of cili ates 
in volved and is therefore di fficul t to reso lve quantitatively with the data available. 
Dolan (199 1a) identified three guil ds of cili ates. Microphagous cili ates consume 
picopl ankton sized prey, including bacteri a, and are mostl y small oli gotrichs and 
scuticocili ates . Macrophages consume larger ce ll s and mostly consist of tin tinnids and 
large oligotrichs. The third guild , predatory cili ates, consumes other ciliates and 
usually cont1ibutes a small frac ti on (3 .5%) of tota l cili ate bi omass (Dolan 199 l a). 
Macrophages are usuall y present onl y in the surface layer, while microphages are 
present throughout the water column. Microphages dominate the bottom layer 
assemblage and can thri ve in anoxic conditions by consuming bacteri a in the absence 
of predation by copepods (Dolan 199 1b). By comparing clearance rates for a number 
of bactivores, Dolan (199 l a) estimated that cili ates account for - 15% of total 
bacteri ovory. Assuming that all bacterial producti on is grazed, consumption of 
bacteri a by ciliates is 59, 244 and 84 mgC m-2 d-1 in the upper, mid and lower Bay 
respectively. For the mid and lower Bay, this is 33% and 39% of ciliate consumption, 
respectively, while for the upper Bay, thi s amounts to only 18% of cili ate 
consumpti on. One may speculate that total consumption of bacteri a by cili ates 
exceeds bacteri al production in the upper Bay due to allochthonous inputs of bacteri al 
cells. Therefore, it has been assumed here that cili ate consumption of bacteri a also 
accounts for 33% of cili ate consumption in the upper Bay. The Chesapeake Bay 
Monitoring program data showed that the biomass of oli gotri chs and tintinnids was 
347 
simil ar in the lower B ay during summer. Assuming that o li gotrichs were microphages 
and consumed a mi xture of bacteri a and picoplankton , wh il e the tintinnids were 
macrophages and did not consume bacteria, the bacteria may be expected to account 
for about 25 % (0 -50%) of the cili ate di et, in the same general range as the estimate 
above. 
Rotifers 
Roti fers, an important component of fres hwater pl ankton (e.g. Gaedke and 
Straile 1994), were found by the Chesapeake Bay Monitoring Program to be present 
during summer in the upper and mid Bay, but not in hi gher salinity areas. Water-
co lumn integrated roti fe r biomass was computed to be 14 and 25 mgC m-2 in the upper 
and mid Bay, respective ly (Table A-7). Adundances were up to - 100 r1 in surface 
waters and much lower in mid Bay bottom waters. Two recent studies examined the 
ecology of the rotifer genus Synchaeta , the bi omass dominant in Chesapeake Bay, 
during earl y spring and late fa ll (Heinboke l et al. 1988 , Dolan and Gallegos 1991 ). 
These studies found roti fe r abundance to be hi gher than estimated here for summer, 
often by 10-fold or more. Loftus et. al. (1 972) measured roti fe r abundances in 
summer phytopl ankton blooms in the mid Bay during 1971 and also found much 
hi gher abundances than were found here. Based on the descriptions of roti fer ecology 
(e .g. Ste rnberge r and Gi lbert 1985), it is specul ated that the hi gh abundances that were 
observed occun-ed under conditions of hi gher phytoplankton abundance than is 
characteri sti c of Chesapeake Bay during summer. Morover grazing pressure on 
rotifers may have been reduced (i.e. lower mesozoopl ankton abundance). As there is 
348 
no reason to believe that the sampling protocol employed by the Monitoring Program 
Was not sufficient to accurately sample rotifers, the abundances computed from 
Monitoring Program data were util ized without modification (Table A-7). 
Specific growth ra tes for two species of rotife rs of the genus Synchaeta 
growing at l 9?C were 0-0.8 and 0-0.2 d-' and depended on food concentration 
(Sternberger and Gilbert 1985). The same study estimated food concentrations at half-
satura tion for these species to be 0.29 and 1.04 ?g mr' alga l dry weight, which is 
approximately equal to 3 and JO ~tg r' chi-a. Thus, growth rates may be -50% of 
max ima/ rates . Egg development times decrease 30% as temperature increases fro m 
19-25?C, suggesting similarly higher growth rates at summer temperatures in 
Chesapeake Bay. A rate of 0.3 d-' was adopted here based on the temperature-
con-ec ted average of these two species growing at half- max imal rates. In the absence 
of specific data to indicate otherwise, net growth effi ciency and assimilati on effi ciency 
Were ass umed to be 0.5 and 0.6, respecti vely, similar to ra tes adopted for cili ates 
(Table A-7). 
The diet of rotifers depends to some ex tent on taxonomic differences in 
feeding mode. The genus Synchaeta accounted fo r 81 % and 91 % of roti fer biomass in 
the upper and mid Bay. It employs a rapid sucking action to capture parti cles and is 
only effecti ve at ingesting larger algal cell s. Therefore, it 's di et has been ass umed to 
consist entirely of net plankton. The genus Brachionus, which contributes the 
remaining portion of the biomass, uses cili ary currents to entra in food particles and 
can ingest a wider vari ety of parti cle sizes (Bogdan and Gilbert 1982, 1984, cited in 
349 
Stemburger and Gilbert 1985). Its diet was assumed to consist of a mixture of 
heterotrophic flagellates and autotrophic picoplankton . 
1kroplankton 
The meroplankton consist of the plankton dwelling larval forms of benthic 
animals. The Chesapeake Bay Monitoring Program enumerated the meroplankton in 
the 44-202 ? 111 size-fraction as well as in the >202 ?m size fraction. Average summer 
abundance in both size fractions was greatest in the lower Bay (13 r') and least in the 
upper Bay (7 r'). Integrated biomass followed the same pattern with biomass 
approximately 3, 19, and 31 2 mgC m- in the upper, mid and lower Bay, respectively 
(Table A-7). In the absence of more information on these organisms, the specific 
1
growth rate for these organisms was assumed to be 0.5 d- , with assimilation and net 
growth efficiencies equal to 0.6 and 0.5, respectively. The diets of the organisms is 
uncertain and their sizes are likely to be variable. It has been assumed that the diet 
consists of larger phytoplankton cells and microplankton, including ciliates and 
rotifers. 
Mesozooplankton 
The biomass and abundance of mesozooplankton was estimated from 
Chesapeake Bay Monitoring Program data , which utilized both 44 ?m 
(Microzooplankton Monitoring Component) and 202 ~m1 nets (Mesozooplankton 
Monitoring Component). Data from one station centrally located in each region of the 
Bay was utilized to compute average abundances of the organisms in this size class. 
350 
For the 44 ?m fraction, above and below pycnocline abundances were combined in an 
average weighted by the respective mixed layer volumes in each region. The 202 ?m 
data samples were composited for the entire water column based on oblique tows. 
The 44-202 ?m fraction sampled the abundance of copepod nauplii which are 
considered here to be mesozooplankton since production by nauplii contributed to 
mesozooplankton biomass without trophic transfer. Mean summer abundance of 
nauplii in the surface /ayer was estimated to be 84, 108 and 131 r 1 in the upper, mid 
and lower Bay, respectively. Integrated biomass of nauplii, which utilized the bottom 
layer abundances as well was 31, 83 and 107 mgC m-2 (Table A-8). 
Abundances in the 202 ?m size fraction consistent mainly of calanoid 
copepods with a small additional contribution by cJadocerans. In the upper Bay, three 
taxa, Acartia tonsa (63%), Eurytemora a/finis (20%) and Bosmina longirostris (4%) 
dominated the summer assemblage. In the mid Bay, Acartia tonsa accounted for 94 % 
of total abundance. In the lower Bay, Acartia tonsa and A. hudsonica combined to 
account for 51 % of abundance, with much of the remainder contributed by 
cladocerans of the genus Evadne and Podon. Because of the relatively larger size of 
Acartia spp., however, Acartia accounted for 85% of mesozooplankton biomass in the 
lower Bay. In mid Bay, Acartia accounted for 98 % of biomass, while Acartia and 
Eurytemora combined to account for 95% of biomass in the upper Bay. The biomass 
of Eurytemora in the upper Bay may be underestimated by these data because recent 
research has shown that abundance is often much greater near the bottom in the 
vicinity of the estuarine turbidity maximum than in the water column as a whole 
351 
(Roman e t a l. 2001). However, because it is not clear at present how to properly 
adjust abundance to account for th is, no adj ustment was made. 
Total mesozooplankton biomass in the >202 ?m size fraction was estimated to 
be 57, 43 and 105 mgC m-3 in the three regions of the Bay, which when in tegrated 
o ver mean depth amounts to 251,443 and 966 mgC m-2 (Table A-8). Because 
substanti a l dominance of biomass by Acartia spp. , energetic rates and diet in formation 
for this taxa were applied to the entire mesozooplankton biomass. The abundance of 
copepod nauplii has already been considered based on the 44-202 ?m size frac tion. 
To examine the production of Acartia, the >202 ?m data were used to examine the 
re lative abundance of copepodites and adults, which have substantially di ffe rent 
patterns of production (Heinle 1966, Kimmerer 1987, White and Roman 1992). 
Heinle (1966) estimated stage durations, mean weight at each stage, stage-specific 
mortality, and total production of Acartia tonsa in Patuxent River, MD. White and 
Roman (1992) estimated egg producti on rates fo r Acartia in mesohaline Chesapeake 
Bay. The stage-s tructured modeling approach suggested by Kimmerer (1987) was 
used to reconcile abundance estimates fo r nauplii , copepodites and adults and compute 
specific production rates. This compu ta tion suggested that the estimated rates and 
abundances were essentially compatible, except for the abundance of copepodids, 
Which was improbably low considering the abundance of nauplii and adults. One 
hypothesis is that given the mesh size used (202 ?m ), the smaller copepodite stages 
Were undersampled. Including these stages would likely increase copepodite numbers 
and biomass _2_fo/d. An alternati ve hypothes is, however, is mortality rates for 
copepodids was higher th an suggested by Heinle (1 966), due to predation or 
352 
starvation . These two possibilities cannot be resolved here. Based on the stage 
durations and weight at stage obtained from Heinle (1966) , specific production of 
nauplii was estimated to be -1 d-1, while that of copepodites was -0.5 d-1? Production 
by adult copepods was assumed to consist solely of egg production, for which a 
summer average carbon-specific rate of 0.14 d-1 was determined for the mid Bay from 
White and Roman (1992 , avg. 1.37 ~LgC female d-1 1, - 5 ~LgC female- , o :~=0.5). 
Overall population P/B for the mid Bay was -0.5 d-1, which was consistent with food-
replete specific growth rates estimated by Berggreen et al. (1988). This rate was 
therefore applied to the combined bi omass of nauplii plus mesozooplankton in the mid 
Bay (Table A-8). 
Adult copepod abundance in the lower Bay was much hi gher than in the mid 
Bay, perhaps due to reduced predation pressure. While the abundance of nauplii was 
also slightly higher than in the mid Bay, the ratio of nauplii abundance to adult 
abundance was 6.0 , -41 % of the mid Bay value. Thi s implies that Acartia egg 
production in the lower Bay was lower than in the mid Bay, perhaps due to food 
limitation (White and Roman 1992, Ki!Zlrboe et. al. 1985, Ki!Zlrboe and Nielsen 1994). 
Berggreen et al. (1988) showed that under food limited conditions, mortality rates, 
stage duration and weight at stage change for copepodids, but not nauplii. 
Accordingly, egg production is an index for overall population growth (Berggreen et 
al. 1988). Thus , it was assumed that production by copepods in the lower Bay was 
-50% of mid Bay rates, or 0.25 d-1? The ratio of nauplii abundance to adult abundance 
in the upper Bay was between the respective values for the mid and lower Bay. 
353 
Accordingly, production in the upper Bay was ass umed to be 0.37 d-1, an intermediate 
value (Table A-8). 
The bioenergetics of Acartia ronsa were examined in considerable detail by 
K0rboe et a l. (1985), who showed that ass imilat ion efficiency and net g rowth 
effic iency changed with food concentration. At high food concentrations, both 
assimilation and net growth efficiencies were reduced. In all regions of Chesapeake 
Bay, es tima ted concentrations of phytop lankton and cili ate carbon are sufficient to 
ensure that the lower efficiencies are applicable. Accordingly, carbon ass imilation 
efficiency has been assumed to be 0.55, while net growth effic iency is 0.75 (K0rboe et 
al. 1985). Identical efficiencies were assumed for nauplii , copepodites and adults 
(Table A-8). 
Gelatinous Zooplankton 
There are 5 major taxa of gelatinous zooplankton in Chesapeake Bay. These 
include the scyphomedusae Chrysaora quinquicirrha and Cyanea capillata, the 
ctenophores Mnemiopsis leidyi and Beroe ovata and the hydromedusan Nemopsis 
bachei. Nemopsis is the most important gelatinous predator in late spring, but 
abundance declines into summer, while other species become more abundant (Purcell 
and Nemazie 1992). Sampling diffi culties also impede quantifi cation of Nemosis 
abundance when large quantities of Mnemiopsis are present (J. Purcell , personal 
communication). Cyanea and Beroe, although not rare, are present in insig nificant 
numbers during summer relative to the other species . Beroe was present in <4% of 
samples collected during summer in the mid-Bay by the Chesapeake Bay Monitoring 
354 
Program (CBMP, unpublished data). Beroe abundances are typica\\y hi ghest in the 
hi gh salini ty waters of the \ower Chesapeake Bay during fa\\ (Purce\\ et a\. 2001). 
Cyanea typically disappears from Chesapeake Bay by May. Thus, the ge1atinous 
zoop\ankton community, as represented in the trophic network models for summer, 
includes only Mn.emiopsis le icly i and C/1rysaora quinquecirrha . 
Abundance of Mnerniopsis and Chrysaora was estimated usin g data from the 
Chesapeake Bay LMER (TlES) progran1 and from the Chesapeake Bay Monitoring 
Program (CBMP). D ata from TIES included -50 stations throughout the Bay sampled 
once per summer usin g two 2-minute Tucker Traw\ tows per station, each with a 1 m2 
net opening and a swept volume of - 11 5 m3 . Data were obtained for the summer 
crui ses in 1995 , 96, 97 and 2000. Data from the CBMP included 4 stations sampled 
three times per summer in each year from 1984-1998. One station (CB2.2) was in the 
upper Bay, whi\e three stations (CB3 .3C, CB4.3C, CB5.2) were in the mid-Bay. 
Oblique tows with bongo nets were used, filtering-15 m3 tow?1? 
Biovolumes of Mn.emiopsi.s varied regionally within the Bay and interannua\\y 
(Table A-9). Mnemiopsis biovo\ume was minima\ within the upper Bay, hi ghest in the 
mid-Bay, and s\ight\y 1ower in the lower Bay. lnterannua\ variations were likely 
related to temperature and freshwater inputs, which affect reproduction of both 
Mnem.iopsis and Chrysaora (Purce\\ et al. 1999). The estimated summer average 
biomass for Mnemiopsis leiclyi in the mid-Bay region is 132 mgC m?2, based on the 
converion 1 m\ = 1.012 mg WW = 0.485 mgC (Nemazie 199 1). Mnemiopsis biomass 
for the upper B ay and \ower Bay is estimated to be 18 and l 15 gC m?2, respectively. 
355 
Abundance estimates fo r Cl,rysaora quinquecirrlza were obtained only from 
th
e TIES data (J. E. Purce ll , unpu bli shed data) because the Chesapeake Bay 
M onitoring Program sampling technique did not sweep a sufficient volume to obtain 
an es timate of Chrysaora abundance. Unfortunately, at average abundance levels 
' 
even the much higher swept volume of the TIES tucker tra wls was marginall y 
sufficient fo r sampling Chrysaora abundance excep t when abundance was high. For 
example, the high abundances observed in the mid-Bay in 1995 were 0.03 m?3, or 3_5 
1 
to w? at 11 5 m 3 tow?1? A t the much lower average abundance in the south Bay in 19
95 
(i e. 0.001 m?\ one expects to catch a single medusa once every JO tows. A t such a 
low expected catch, these tows provide an uncertain estimate of Chrysaora abundance 
at best. 
A s with Mnemiopsis, summer Chrysaora abundance varied regionally and 
interannually (Table A-10). The highes t abundances were always in the mid-Bay 
region ; abundance was usually negligable outside the mid-Bay. Increased abundance 
occun-ed in years with Jow freshwater input (e.g. 1995, 1997). In the mid-Bay, 
biomass averaged - 6.3 mgCm-2 in the mid-Bay, computed using 1 ml biovolume = 
l.75 mgC (Purcell 1992; Purcell , J. E., personal communica ti on). 
Procedures for es timating consumption and growth of gelatinous zooplankton 
are described below. Egestion was es timated from the carbon balance equation and 
the other estimates. The bes t es timates of the growth dynamics of Mnemiopsis are 
based on laboratory experiments reported by Reeve et a l. (1 989) fo r Mnemiopsis 
mccradyi. References to thi s study by P urcell e t al. (2001 ) and o thers imply th at 
experts in the fie ld be lieve it reasonable to also apply these growth dynamics to 
356 
Mnemiopsis leidyi. Specific growth for Mnemiopsis increases with prey density and 
decreases with s ize of the ctenophore. Assuming zoop lankton concentrations of _ r 
20 
I 
, specifi c growth for small (15mm) Mnemiopsis is -0.4 d-1, while spec ific growth for 
larger individuals (50mm) is -0.2 d-' . At higher prey density, specific growth of 
1
larger ctenophores was only slightly higher (0.3 d- ) , while growth for smaller 
individuals was much greater, 0.8 d-' (Reeve et al. 1989). At lower prey density-IO r 
I 
, consis tent with average abundance in middle Chesapeake Bay during summer, 
specific growth dec lined significantly to -0.1 d-'. Mnemiopsis could not grow at 
zooplankton abundances lower than 8 r' (Kremer and Reeve 1989). At the lower 
specific growth rate (0.1 d-1) average Mnemiopsis growth in the mid-Bay is 13 mgC m-
2 d-' , while at the higher, food-replete rate, growth is 38 mgC m-2 d-1 (B=1 26 mgC m-
2). The food-replete growth estimate is approximately the same as the es timated 
consumption (see below), which implies an improbably high gross growth efficiency 
1
(GGE) near 100%. The lower growth estimate (0.1 d- ) and the estimated 
consumption gives GGE=31 %, only slighly lower than the 40% estimated by Reeve et 
2 
al. (1989). Assuming GGE=40% and consumption=42 mgC 111 - (see be low) implies 
2 
an intermediate rate of growth rate of 17 mgC m- d-' and spec ific growth rate of 0.13 
d-1, the estimate adopted for the mesohaline Bay. Ctenophore production for the upper 
and lower Bay was computed using the same specific growth ra te. 
Clearance rates of Mnemiopsis depend on predator size, water temperature, and 
prey type, but do not depend not significantly on prey density. Therefore, 
Mnemiopsis feeding does not saturate at high prey densities (Purcell e t al. 2001). For 
a moderate-sized individual (biovolume = 14 ml) feeding on Acartia tonsa, clearance 
357 
rates in the field are -50 I d.1, a volume-speci fie clearance rate of 3.57 / mr 1 d? '. 
1
Assuming a summer copepod density of 15 r , a copepod carbon content of 3 ?gC ind? 
I 
and summer average Mnemiopsis biovolume of 264 ml m?2 in the mesohaline Bay, 
the consumption rate is estimated to be 42 mgC m?2 d?1. Specific consumption is 
therefore 0.33 (g g? 1 ) d -1? Consumption by ctenophores in the upper and lower Bay 
was computed ass uming the same specific consumption rate (Tab le A-7). 
Respiration rates for M11emiopsis increase linearly with size (Kremer and 
Reeve 1989). Specific respiration (R.,, d.1) increases hyperbolica lly with prey 
abundance (i.e. with ingestion) according to the function 
1 1
where Rb=0.04, R111ax=0.ll , K.,=30 r and P=prey abundance (r ) (Kremer and Reeve 
1989). Thus, at P=l5 r 1, R.,=0.08 1d- ? The mid-summer Mnemiopsis population 
2
respiration rate is therefore R=RsB=0.08(1 26 mgcm? )=l0 mgC m.Z. Ctenophore 
respiration in the upper and lower Bay was computed ass uming the same specific 
respiration rate. Excretion of DOC is estimated to be 50% of respiration (Kremer and 
Reeve 1989). 
Egestion was calculated by difference, giving a mid-Bay esti mate of 11 mgC 
m-2_ At 26% of consumption this implies an ass imilation efficiency (AE) of 74%. 
Gross growth efficiency is 40%, while net growth efficiency is 51 %. The estimated 
efficiencies depend on the average size of the ctenophore and prey (copepod) density, 
1 1
which were assumed to be 14 ml ind? and 15 r , respectively. Consumption by 
ctenophores increases proportionate /y  with prey density si nee c learance rate is 
358 
constant. Specific growth increases rapidl y with prey density for small individuals, 
but only slowly for large indi viduals. Respiration rates increase toward an asymptote. 
Thus, in an alternative scenario where prey (copepod) density is 100 i- 1, specific 
growth is 0.5 d-1, specific respirati on 0.15 d-1, and specific consumption 2.22 d-1, GGE 
decreases to 23%, driven significantly by AE decreasing to 31 % (Kremer and Reeve 
1989). Such high prey abundances are not uncommon in Chesapeake Bay, but exceed 
the average significantly. This computation illustrates that under these conditions , 
Mnemiopsis grows rapidly and also shunts large amounts of zooplankton production to 
detritus. As previously noted, below average prey abundance results in mortality of 
Mnemiopsis due to starvation. 
The bioenergetics of Chrysaora quinquecirrha are not as well-studied as that 
of Mneniiopsis. Some investi gators have assumed that growth dynamics of Chrysaora 
are similar to that of Mnemiopsis (e.g. Baird and Ulanowicz 1989); however, 
published sources indicate that this is not appropriate, especially for the mainstem 
Chesapeake Bay where individual medusae tend to be large. Such large individuals 
(e.g. 80 mm bell diameter) have very low specific growth (-0.02 d-1; Larson 1986) 
which contrasts with higher specific growth rates for Mn emiopsis. Such low growth 
rates can be reconciled with the apparently high abundance of medusae in two ways. 
First, high abundances of smaller individual s occurs in the tributaries , and these areas 
may serve as a source for larger individuals in the mainstem of the Bay. Secondly, 
because medusae have no significant predators in Chesapeake Bay, even slow growth 
rates can lead to significant biomass accumulation over time. 
359 
Consumption by medusae is the only well-studied aspect of Chrysaora 
bioenergetics. Like Mnemiopsis, diet composition and consumption rates for medusae 
depend on abundances of suiTable A-prey items. Clearance rates vary significanctly 
according to prey type, with clearance of Mnemiopsis being by far the highest. For a 
large individual (80 mm diam.), Mnemiopsis clearance is esti mated to be -2 mJ d-1 
(Purce ll and Cowan 1995). At an average mesohaline Bay ctenophore density of 3 m-3 
(6.7 mgC ind-1) ctenophore inges tion by a single medusa is estimated to be -40 mgC 
1
d- . At the average medusa density this amounts to 3.5 mgC m-2 d-1. The high 
clearance rate for ctenophores makes Mnemiopsis the largest diet component (90%) at 
ctenophore abundances typical of Chesapeake Bay. However, ctenophores are often 
nearly eliminated from waters where Chrysaora is present in significant abundance 
' 
making other diet components more significant at those times. In a scenario where 
Mnemiopsis is absent or nearly so, copepods, anchovy eggs and anchovy larvae are 
estimated to account for 46%, 42% and 12% of the much smaller diet, respectively. 
Volume-specific clearance rates for medusae feeding on copepods is much less 
than for ctenophores and depends on medusa size, water temperature and prey density 
(Purcell 1992). An 80 mm medusa (biovolume=41 ml) at 27?C is estimated to clear 
1
copepods at density=15 r' at a rate of 48 l d- , comparable to an average ctenophore 
(biovolume=l4 ml). At this clearance rate, daily copepod consumption is 720 d-', or 
2.2 mgC d-1. At the average abundance of medusae, thi s is 0.2 mgcm?2d-', or-5% of 
the diet (Fig 2). 
Clearance rates fo r Clirysaora feeding on bay anchovy eggs and larvae, the 
most abundant fish eggs and larvae in the mid-Bay during summer, were estimated by 
360 
-
Purcell et al. (1994) (see also Houde et al. 1994). For an 80 mm diam. medusa, 
clearance rates for eggs and larvae are approximately 2.5 and 1.5 m3 d-1, respectively. 
These estimates are less certain than estimates for other diet components, particularly 
in the case of larvae. At average summer prey abundances, ingestion of eggs and 
larvae is esti mated to be 250 and 75 predato( 1 da/ , or 1.9 and 0.6 mgC d-1? For the 
population, this is 0.17 and 0 .05 mgC m-2 d-1, a combined 5.5% of the diet. Note that 
although copepods are 150 times more abundant than bay anchovy eggs, consumption 
of eggs by Chrysaora is the similar due to the much hi gher clearance rate for fish eggs 
and larvae. 
Total consumption by the average Chrysaora population in the mid-Bay is 
3.94 mgCm-2 d-1, a specific consumption rate of 0.85 d-1 (Table A-7). Assuming 2% 
growth per day , gross growth efficiency is 2.4% and assimilation efficiency (AE) is 
10%. If Mnemiopsis were absent, however, specific consumption declines to 0.09 d-1, 
barely exceeding respiratory demands. To allow for a realistic AE, growth under these 
conditions must be zero or negative. The ability of Chrysaora to survive while losing 
weight during starvation has been demonstrated in the laboratory and can also be 
justified by an observed decline in abundance of larger medusae in late summer (J.E. 
Purcell and R . Rosen, unpublished data). 
The estimated consumption of Mn.emiopsis by Chrysaora in the mid-Bay is 
only 22% of estimated production by Mnemiopsis. For Chrysaora to consume all 
Mneniiopsis production would require increasing the biomass of medusae by 4 times. 
Although this abundance (-16 ml m-2) has been observed and is not implausible, it is 
not consistent with the observed average abundances in the mid-Bay. lt seems likely 
36 1 
that food limitation of Mnemiopsis may be a factor at times . However, coffelative 
evidence suggests that Cluysaora can frequently control clenophore abundances in the 
mid-Bay. The diets of M11emiopsis and Cllrysaora are reasonably well studied. Diet 
composition vari es substanti ally according lo avai lable prey, indicating that these 
predators are not hi ghly selective (Purcell et al. 2001). However, the gelati nous 
zoop/ankton are clearly predators , capturing certain types of prey according to their 
physiologica l capabilities rather than filtering any type of biological particle from the 
water (Purcell 1997). Copepods, copepod nauplii , barnacle nauplii , and bivalve 
ve/igers comprise 95-93 % of the diet of adult Mnemiopsis. Other prey items include a 
variety of larvae plus ichthyoplankton (Purcell et al. 2001), although Purcell et al 
(1994) found in a study in Chesapeake Bay th at Mnemiopsis consumption of bay 
anchovy eggs was negliga ble. Although Mnemiopsis larvae have been shown to 
consume phytoplankton and cili ates (Purcell et al. 2001), detritus and attached bacte,ia 
and phytoplankton cells are not components of the adult ctenophore di et. 
The diet of Chrysaora includes Mnemiopsis, adult copepods and fish eggs and 
larvae (Purcell 1992; Purcell et al. 1994; Purcell and Cowan 1995). Copepod nauplii 
are not as readily captured by Cluysaora as are copepodites and adults, and are 
therefore not a significant part of the diet. 
The Benthic Community 
The benthos has been characterized as including two types of primary 
producers (submerged vascular plants and microphytobenthos), sediment-assoc iated 
bacte n.a , me1.0  f auna, deposit -feedinoo  macrobenthos, suspension-feeding macrobenthos 
362 
and reef-dwelling suspension feeders (principally the oyster Crassostrea virginica). 
Other organisms are clearly present, and may include motile and sessile epifauna (e.g. 
Sagasti et al. 2000). Some of these organisms may be included in the Chesapeake Bay 
Benthic Monitoring data, however, adequate representation, particularly of motile 
species, is not assured. Thus, the role of these animals was neglected in this analysis . 
Demersal fi shes and blue crabs, which are associated with the benthos, were 
considered separately as part of the fish community. The benthos also includes a 
stored pool of organic carbon. This is conceived here to be a single pool, although it 
may include a combination of particulate carbon and dissolved organic carbon 
confined in sediment pore waters . 
.S..ediment Particulate Organic Carbon and Bacteria 
The size of the stored organic pool is not easily quantified or characterized, 
particularly as the boundary between the stored but available carbon and the 
permanently buried pool is somewhat arbitrary. The di stinction is certainly time-scale 
dependent as episodic erosion of sediments can return buried ca rbon to the active pool. 
Fortunately, the indended analyses were not expected to depend on this value; it was 
estimated mainly for completeness. It was assumed here that the active pool extends 
to 5 cm depth , that the sediments contain 80% water, and that the density of dry 
sediment is 1.1 g cm? Accordingly, the active pool of dry sediment is 1.1 (104) g m-2_ 
The average carbon content of sediments is approximately 31.5, 18.4, and 7 mgC (g 
dry sedimentY1 in the upper, mid and lower Bay (Kemp et al. 1997). Therefore, the 
active sediment organic carbon pool is these regions is 346500, 202400, and 77000 
363 
mgC m-2 (Table A- 11). The biomass of mciofauna and sediment bacteria , which is 
methodologically included in these fi gures , accounts for <l % of biomass. 
M etabolic rates due to sediment bacteria in Chesapeake Bay have been 
estimated by Cowan and Boynton (1996) and Marvin-DiPasquale and Capone (1998). 
Cowan and Boynton ( 1996) described rates of sediment oxygen consumption (SOC) 
while Marvin-DiPasquale and Capone (1998) measured rates of benthic su lfate 
reduction (SR). Kemp et al. (1997) assumed that half of SOC was due to SR, and 
therefore estimate total sediment metabo li sm SR+0.5(SOC) . Furthermore, they 
assumed that due to substrate limitation, sediment metaboli sm was reduced by half in 
shallower waters as compared to the deeper channel regions were the benthic rates 
were measured more often . This same approach was adopted here. Rates mi ght have 
been obtained directly from Kemp et al. (1997) , except that summer rates were 
required, while they provided annual estimates. Macrobenthic respirati on is 
theoretically not excluded from the benthic flu x measurements. However, it was 
assumed here that the macrobenthos were patchily di stributed and not adequately 
reflected in the small cores uti Ii zed used to measure sediment fluxes. This is known to 
not be sttictly true as macrobenthos, espec iall y smaller animals, were sometimes if not 
often retained in cores (W . R. Boynton, pers. comm.). The production of bacteria was 
estimated from benthic respiration assuming bacterial growth efficiency= 0.30. 
Biomass was estimated from production assumin g P/B=l.13 d-1 (Table A-12). 
364 
Meiobenthos 
T he benthic meiofauna are commonly defined operationally as those animals 
retained on an - 50 ?m sieve but passing through the 500 ?m sieve typically used to 
o ups, 
retain the macrobenthos. The meiobenthos are often divided into two main oro 
harpacticoid copepods and nematodes, which usually constitute a major fraction of the 
biomass (Fenchel )978). Other important biota include benthic foraminiferans, 
ostracods and ju ven i\e pol ych aetes (S undbiick et a I. I 996). Production of mei ben thos 
has been related to production of microphytobenthos (Montagna et al. 1995), 
especially in intertidal and \otic environments where the meiobenthos have been 
studied most extensively. The biomass of meiobenthos has also been estimated via 
empirically determined ratios between the biomass of meiofauna and that of deposit 
feeding macrobenthos. Jt has been suggested that the biomass of meiobenthos 
amounts to 3-13% of the biomass of deposit feeders (U\anowicz and Baird 1986 and 
references cited therein), however, the ratio can be even more vari able than that (Flach 
et al. 1999). Thus, it was assumed here that meiofaunal biomass in Chesapeake Bay 
was comparable biomass in other mesotrophic to eutrophic temperate estuaries, 
without consideration of the macrobenthic biomass in Chesapeake Bay. Smol et al. 
(1994) examined the summer biomass of meiobenthos in the Oostersche\de estuary 
2
during summer and observed a total biomass equal to 400-4200 mgC m? ? Further, 
they noted that higher numbers were associated with higher silt content of sediments. 
However, the hi ghest biomass was found only on intertidal fl ats, which are not present 
in Chesapeake B?Y? sundb3ck et al. (1996) examined the meiobenthos at sandy sites 
in the Skagerrak (Swedish west coast), observing summer biomass at 4 m depth to be 
365 
m , consistent wit t 1e ower num ers suggested by Smol et al. (1994). 
- 700 mgC -2 ? ? h I I b 
Sundback et al. ( 1996) also estimated the benthic microalgae at their sites to have a 
summer biomass of - 1000 mgC m-2, which suggests a similarity between these sites 
and shallow Chesapeake Bay sites . Accordingly, the meiobenthic biomass in the 
upper and lower Bay, and at above-pycnocline sites in the mid Bay was ass umed to be 
700 mgC m?' (Tab le A-12). Below the pycnocl ine in the mid Bay, meiobenthic 
biomass was ass umed to be greatl y reduced due to the effect of anox ia. A fi gure of 70 
mgC m?' was adopted by assuming that onl y hypoxia tolerant fo ramini ferans were 
present in these areas (SundbOck et al. ) 996, see also Karlsen et al. 2000), and biomass 
was computed as a weighted mean of above- and below-pycnocline biomass (Table A-
12). 
Production of meiobenthos was computed using the allometii c relati on of 
037 
Banse and Mosher ( J9 80), which is P/ B = 0.65M ; , where M., is the mass 
equivalent in kcal (I g AFDW = 5.5 kcal). They also noted that production fo r 
me iofauna may be 3-5 times lower. Ash-free dry weights fo r typical me iobenthos 
are: nematodes ( . J 5-0 .5 9 ?g), [oraminifera ( 0 .5 ?g), ha rpactico ids ( I . 2 ?g), ostracods 
0 7 
(7.5 ?g) (Widbom I 984). Thus, these organisms are in the size range of 10? to 10?' g 
AFDW, with most being at the smaller end of thi s range. Thus, appropriate P/B for 
mei fauna may be . to o, 07 d. 1_  Accounting for the dominance of the s mailer 
O01 
O 1 
animals and high summer temperatu res, a value of O. D7 d ? was adopted fo r 
Ch k ? 1?i - t the valueofO.Jd- suggestedby Gerl ach (l97J ). 
esapea eBay, s1m1 at o 
C . . _ . were computed by setting ass imilation effi ciency and net 
onsumpt10n and resp11 auon 
- . . m hich is appropri ate for a class of aquatic animals 
gtowth efficiency to 507 0, w 
366 
---
n g a m 1x  tu ,e of algae and detritus (Sch roede1 19 81). This com pu tati on gives 
consumi ? 
an estimated me iobenthic respiration fo r the upper Bay that is a greater, possibly 
imp aus i b Iy  greater, fraction of tota I sediment metabo Ii s m than was observed for the 
1 
ol er regions of the Bay. The is partl y due to the high rates of sulfate reduction which 
h
were only Found outside the upper Bay. Altern ati vely, meiobenthic biomass and 
production may in fac t be \ower in the upper Bay than suggested here. The diet 
composition of meiofauna mainly includes benthic algae and sediment bacteri a 
(Sundbiick et al. 1996), but maY also include other meiofauna (Table A-12). 
Macrobenthos 
T he macrobenthos are operationall y defined as those benthic animals which 
are reta ined on a 500 ?m seive. The biomass of macrobenthos was estimated from the 
Chesapeake Bay Monitoring Program, which has quantifi ed the biomass and 
abundance of the macrobenthos at manY sites in Chesapeake Bay on a seasonal basis 
since 1984. A detailed analysis of macrobenthic biomass and production in 
Chesapeake Bay was presented in Chapter 4. f or thi s study, the macrobenthos were 
divided into two feeding gui\ds, suspension feeders and deposit feeders. Biomass 
estimates for each region of the Bay for each of these guilds was obtained directly 
from Chapter 4. Average production/biomass for each region was used for both 
guilds, except for in the upper Bay, where p/B For deposit feeders was not assumed to 
be equal to average for the macrobenthos in the region, due to the dominence of 
slower growing Rangia. Rather, a p/B typical of other deposit feeding communities 
(0 .014 d?') was assumed (fable A, 12). Respiration, consumption and egestion were 
367 
-=---
generated by assuming that nel growth efficiency was 0.4 and assimilation efficiency 
was 0.5, intermediate values within lhe range considered in Chapter 4 (Table A-J 2). 
The diet of the suspension-feeding macrobenthos was assumed to consist primaiiJy of 
a mixture of particulate organic carbon and net phytoplankton cells. 
The macrobenthos is grazed heavily by demersal-feeding fish and crabs. 
However, harvest by fisheries can also be a significant fate of production. Other than 
oysters and blue crabs, which are considered separately, the soft-shell clam, Mya 
arenaria, is the only macrobenthic species supporting a significant fishery in 
Chesapeake B ay (NOAA landings data). Maryland landings reached a peak of _3500 
tons y-1 in the 1960s but declined significantly in 1972 and later. The averaae 
b 
Maryland landing during 1985-1999 was 634 tons i'. The catalyst for this change 
was most likely hurri cane Agnes, in June 1972. Landings of Mya occur throughout 
the Chesapeake Bay region and it is likely that a large fraction came from outside of 
the mainstem Bay. In addition, it is also likely that some of the harvest resulted from 
production that occurred outside of summer. However, for the purpose of marking a 
conservative estimate, it was assumed here that al/ of the landings resulted from 
summer production in the mesohaline portion of the mainstem Bay (area==2,338 km\ 
2
Accordingly, the production is equal to 0.27 g wet weight m- , or 0.29 mgC m-2 d-1 
(assuming AFDW==0.2 wet weight, 500 mgC=l g AFDW, summer==92 d). This is 
equal to 5.3% of the summer suspension feeder production in mesohaline Chesapeake 
Bay. Consideiing the extremely generous assumptions, it can safely be concluded that 
the estimated production of suspension feeders is consistent with the fisheries 
368 
landings. Furthermore, the landings do not consitute a significant removal of 
suspension-feed ing production from the system and can be ignored. 
O yster 
T he biomass and production of the oyster (Crassostrea virginica) has been 
considered separately from the macrobenthos because these animals are known to 
occup y di screte reef areas and were not sampled by the Chesapeake Bay Benthic 
Monitoring Program . A separate analysis also permi ts direct eva luation of the trophic 
role of thi s hi stori call y important species, which was an impo1tant goal of thi s 
analysis. 
It is well known that the abundance of oysters in Chesapeake Bay and its 
tributaries has decreased enormously since the earl y part of the century (Newell 1988, 
Rothschild et al. 1994). Newell (1988) estimated that by 1988 total bi omass had 
decreased by >99% from pre-1900 levels. An examinati on of commercial landings 
stati stics shows that oyste r landings in both Maryland and Virgini a dec lined in weight 
and value from most important prior during 1950-1985 to relati ve obscurity after 
1985 . 
The oyster population that remains at present has been severely affected by di seases 
(MSX, Dermo) that arrest growth before market size is reached, leading to substanti al 
mortality. Commercal oyster harvest appears to be concentrated into areas where 
salinity is low enough to reduce di sease prevalence, but not low enough to kill the 
oysters. Thi s makes these popul ations vulnerable to morta lity due to low salinity 
followin g flood events (Tarnowski et al. 1994). Appropri ate habitats are o ften short 
369 
reaches of the tidal tributaries, which is where repletion efforts have been concentrated 
(Jordan et. al. 1994). The lower reaches of the same tributaries (e.g. Choptank R. , 
Chester R.) have not supported commercial production due to excessive disease 
mortality (Jordan et. al. 1994), suggesting that the same problems may befall 
populations in adjacent regions of the mainstem. Oyster abundance in the Virginia 
mainstem appears to be below commerciall y viable levels as well. Hargis and Haven 
(1994) reported that 94% of 5,484 Va. bushels harvested in the 1993-94 season came 
from the James River. From this perspective, one may conclude that the commercial 
harvest does not provide useful evidence for average biomass or production of oysters 
per unit area in any region of the present-day Chesapeake Bay mainstem, other than to 
suggest that abundance is be low levels in the commercially harvested tributaries. 
If one assumes that the abundance of oysters in the mainstem Bay cannot be 
related to commercial landings, average biomass must be estimated as the product of 
the area of oyster bottom and the biomass density on those bottoms. The area of shell 
or "cultch" bottom was surveyed in 1976-1983 by the Maryland Bay Bottom Survey 
(MBBS) using patent tongs and/or a dragged microphone (Smith 2001). However, a 
re-examination of a subset of these areas using a sophisticated acoustic seabed 
classification syste m, underwater videography, and diver surveys showed that the 
MBBS could not di stingui sh between clean shell , suitable as oyster habitat, and shell 
covered with silt and mud, which is not suitable oyster habitat (Smith 2001). 
Moreover, of the mainstem oyster bars examined, 80-100% of the bottom was sand or 
mud, with no clean shell found. This suggests that a computation of the area of oyster 
bottom in the mainstem Bay based on the MBBS would be erronously high, perhaps 
370 
dramatically so. In fact , it seems to suggest that there is no significant area of oyster 
bottom in the mainstem Bay. Th is conclusion is also supported by examining the 
factors that have been associated with healthy oyster beds. For example, oyster bars 
have been associated w ith strong currents and high relief bottoms, which deter 
accumulations of si lt. Some areas in the Maryland mainstem are assoc iated with 
points of land and may be expected to have these characteristics. However, Smith 
(2001 ) examined some of these si tes and fo und that none of them supported oyster 
popul ations. Although one cannot rule out the existence of oysters in the mainstem 
Bay, the area of suitab le habitat may be minimal. Even if the oyster biomass at the 
remaining areas was substanti al, which it may not be, overrall biomass wou ld be be 
very low at the whole estuary scale. Compared to the near ubiquitous, even if 
depauperate, mesohaline infauna! communiti es, the oyster is probably not impo1tant in 
terms of carbon flow in the mainstem Chesapeake Bay food web. This constrasts with 
the wide perception of its present and past eco logical importance. Where the oyster is 
present in greater abundances in the tributaries, different conclusions may easily be 
reached. However, this is beyond the scope of thi s study . 
It is worth noting here that estimates of the biomass of oysters in the 
mesohaline Bay have been suggested in the past. Baird and U lanowicz (1989), who 
described trophic networks for Chesapeake Bay during summer suggested an average 
mid Bay oyster biomass of 1100 mgC m-2 The estimate was derived (apparently with 
some incorrect mass-mass convers ions) from the biomass estimate of a well-known 
expett (S. Jordan, MD DNR), who suggested that the oyster biomass for the entire 
mesohaline oyster popul ati on in 1984 was equal to 4 106 bushels. Newell (1988) 
37 1 
cited an update for 1988 from the same expert, giving an estimated biomass of 1.1 106 
kg dry weight, an -50% decrease from the 1984 estimate (1 Md. bushel= 2.64 kg wet 
weight, dry weight= 20% of wet weight) . Assuming hypothetica lly that all of the 
oyster biomass was in the mesohaline mainstem Bay (area=2338 km2) and that 500 
mgC = 1 g dry weight, the average biomass of oysters in the mainstem Bay is 235 
mgC m-2 . T hi s is 16% of the biomass estimate for the already degraded mid Bay 
in fauna! community and (fo r compari son) 2% of the biomass estimate for the lower 
Bay infauna. In con trast to the hypotheti ca l assumpti on, much of the biomass must be 
located in the tributaries rather than the mainstem, since thi s is where the commercial 
fisheries are concentrated. Thus, if one accepts these estimates of total biomass, the 
mainstem biomass must be very low. Moreover, commercial harvests in Maryland 
have dec lined furthe r in recent years, to 50% of 1988 leve ls and biomass may have 
declined commensurately . From these results, it has been concluded that the oyster 
should be represented as functionally ex tinct in the mainstem Bay, even as this may 
not be strictly true, and even as this is arguably not the case in the tributaries, which 
support substantial ongoing harvests in association with rep letion efforts there. 
Finfish and Crabs 
The C hesapeake Bay supports signifi cant commercial and recreational 
fisheri es. Despite some notable popul at ion co ll apses, total fi sheri es output has not 
declined over time indicating that the community continues to be productive (Houde et 
al. 1999). Species not supporting fi sheri es are likely to be ecologically important both 
as predators and prey, although thi s has not been demonstrated to the extent one might 
372 
expect. Good examples of such studies are available for bay anchovy (e.g. Luo and 
Brandt 1993). For th is study, the fish community was analyzed as either individual 
species or taxonomically-based aggregations (e.g. alosids), primarily because this is 
the way that data have generally been collected and reported. The objective of this 
analysis was to characterize the biomass, bioenergetics, and diet of the most important 
species of fi sh and crabs in the ecosystem. l mportance was determined primari ly by 
relati ve summer biomass on a regional basis. Feeding guild (e.g. planktivore, 
benthivore, piscivore) was also considered to ensure representation of each guild . A 
few commercially and recreationally important species were included in the analysis 
specifically because of the fishery, although these species are probably ecologically 
important as well (e.g. stri ped bass) . 
To the extent possible, published data sources were utilized. However, 
additional fi shery-dependent and fishery-independent data from unpublished sources 
were needed. Annual and monthly fisheries landings are reported by state on a 
National Oceanographic and Atmospheric Administration (NOAA) web site 
(http://www.st.nmfs.gov/stl/) . These data are cited as "NOAA F ishery Statistics." 
T he field component of a comprehensive multi-year assessment of finfish 
communities in Chesapeake Bay was recentl y concluded as part of the Chesapeake 
Bay Land-M argin Ecosystem Research Program (TIES). The resulting data, although 
still largely unpubli shed, have been compiled and summarized extensively by S. Jung, 
of Chesapeake Biological Laboratory. Results of these priliminary analyses were 
utilized for this study (S. Jung, unpublished data) . These data were collected using 20-
minute mid-water trawl tows conducted at night. The tows were stepped from the 
373 
-
surface to the bottom over the duration of the tow, scraping the bottom during the last 
sever I ? . 
a mrnutes of the low. The results suggested size- and species-dependent gear 
selectivity. Larger piscivores (e.g. bluefish and st1iped bass) were caught rarely, 
suggesting net avoidance. Menhaden are known to form tight schools which also 
appeared to be able to avoid the net. ll is also possible that demersa/ fishes were 
undersamp/ed. Other studies such as stock assessments were used to supplement the 
TIES data. 
The 5 most abundant species accounted for 86% of the average abundance 
estimated via TIES trawl data. The most abundant fish captured in these trawls was 
the bay anchovy (Anclwa mitclulli), followed (descending biomass) by white perch 
(Marone americana), Atlantic croaker (Micropogonias undulatus), weakfish 
(Cynoscion regalis), and spot (Leioslomus xanthurus). Addition of the heITings and 
shads (e.g. blueback herring, alewife, gizzard shad) accounts for an additional 4% of 
the total. Other important fin fish include the Atlantic menhaden (Brevoortia 
tyrannus), which supports the largest commercial fishery in the Bay, and the striped 
bass (Marone sa.xatilis) and bluefish (Pomatomus saltatrix), which support significant 
recreational fishing . As noted, the relative biomass of these threee species may be 
Underestimated by the TIES data due to net avoidance. The bottom-dwelling 
hogchoker (Trinectes maculatus) and the striped anchovy (Anchoa hepsetus) were also 
captured in significant biomass. The channel catfish (lctalurus punctatus) and 
America nee1  (A ngut.1.1 a  rost.?r a la) were captured in significant numbers in the upper 
Bay,  but Ia ss so e 1s ew1 1 e1.e  1?1 1 the Bay?  The blue crab (Callinectes sapidus) was not 
captured r?n  g reat a b un d ance 1?n  1111?d water trawls, but is known to be very abundant 
374 
throughout Chesapeake Bay. The discussion below is organized by species. Several 
species were examined in much greater detai I than others, either because they are 
particularly abundant in the Bay, or because the quantity of relevant research data is 
greater, or both. All the estimated biomasses and bioenergetic parameters for the fish 
community are summarized in Table A-13. 
Blue Crab 
In recent years, the blue crab has supported the largest and most valuable 
fi shery in Maryland. In Virginia, its fishery has been the third most important in 
Virginia in terms of weight and value (NOAA Fishery Statistics). The size of the 
fishery can be assumed to reflect the substantial biomass and production of the blue 
crab in Chesapeake Bay. The blue crab is also of great ecological importance via its 
substantial role as an epibenthic predator (e.g. Virnstein 1977, Hines et al. 1990 and 
references cited therein) . 
Because the blue crab is highl y valued and exploited, ample fishery-
independent and fishery-dependent abundance data have been collected. Rugolo et al. 
(1998) describe a detailed stock asessment, giving annual estimates of absolute 
abundance of the recruited stock and providing much of the basis for the population 
estimates described below. T hey estimated the 1985-1995 average recruited 
abundance to be -500 106 , with the age l+ c lass 75% recruited and all o lder age 
classes fully recruited. Mean proportions at age were obtained from the theoretical 
stock described by Rugolo et al. (1998), assuming the total mortality rate Z=l.3 
(Z=0 .9, M =0.375) and aggregati ng the 4+ age group to include all older age classes 
375 
(Table A-13). Accordingly, the recruited stock accounts for 42% of total abundance, 
giving a total abundance of 1200 l06 for the population as a whole. Although aging 
techniques have been developed recently for blue crabs (Ju et al. 1999), stock 
assessments have used length-based approaches, whereas mean carapace width at age 
are approximated. This study used the values suggested by Rugolo et al. (1998). No 
published, length-weight re lationships could be found for blue crabs. However, 
relationships of the form W=aLbwere derived here from a data set containing 493 
observations on carapace width (mm) and wet weight (g) of male and female blue 
crabs collected in the Virginia winter dredge survey. For females, a=0.00672 and 
b=2.45. For males, which at larger sizes achieved slightly higher weight at a given 
length, a=0.000281 and b=2.66. The sex rati o was assumed to be 50:50, although the 
difference in weight at length was sufficiently small that any violation of this 
assumptio n would have a minimal effect. Summing the products of numbers and 
weights at age, the total biomass of the population was computed to be 150,000 tons, 
of which 125,000 tons was recruited lo the fi shery. The 1985-95 average landings, in 
which an unquantified recreational catch was assumed to equal 25% of the 
commercial landings (Rugolo et al. 1998), was estimated to be 48000 tons. The 
natural mortality rate is widely estimated to be 0.375, which results in an annual 
mortality equal to (150,000)(0.375)=56250 tons. At steady state, annual production 
equals natural plus fishing mortality, therefore, total production is approximately 
104,250 tons and annual Pj B = 0 .7 . It is not c lear how much of annual production 
occurs during summer. A value of 50% was assumed. 
376 
The blue crab is highly motile and ex hibits large-scale seasonal miorntions 
b ' 
particularly for females (e.g. Prager 1996). Conseq uently, assigning its ecological role 
on a regional and seasonal basis presents a problem. At this point, there does not 
appear to be a certain means of resolving thi s difficulty, except to examine the habitats 
prefen-ed and/or required by the blue crab, and to assume that the blue crab does not 
occupy unsuitable habitats. In this regard, several observations can be made. Blue 
crabs are known to prefer well oxygenated water (>5 mg r1). Therefore, they are 
likely to avoid deeper habitats in summer, where oxygen is often depleted. Blue crabs 
have a wide sa linity tolerance; studies have not shown strong population density 
gradients with salinity, except that males occupy lower salinity waters than do females 
(Miller et al. 1975). Thus, the area of the Bay in summer that might be excluded as 
habitat on the basis of salinity alone is minimal. It has been suggested that 
abundances of blue crabs are higher in the tributaries in summer than in the mainstem 
(references cited in Ulanowicz and Baird 1986). However, the crab pot fishery, which 
by law operates exclusively in the mainstem Bay in Maryland, accounts for 60% of the 
Maryland hard crab landings (EPA 1997), while the trotline fishery accounts for 40% 
of landings. While the pot fishery most likely intercepts migrating crabs which have 
utilized tributary habitats, blue crabs obviously do utilize mainstem Bay habitats as 
Well. Lacking more certain data, it was assumed that during summer crabs have equal 
densities within all the suitable habitats. Unsuitable habitats were based on depth, 
whereas waters greater than 8 m in depth are more likely to have oxygen conditions 
below prefeJTed levels for blue crabs. The surface area of the Chesapeake Bay and its 
2 
tidal tributaries is -18,200 km 2. Of thi s, 3,339 km in the mainstem and 6 major 
377 
2 
tributaries are at >8m depth . This leaves 14,861 km of blue crab habitat. The upper 
Bay accou nts for 3.6%, whi le the mid and lower Bay each account for 7.6% of the 
total habitat in the Bay and tributaries (Table A-14). In total , 18% of suitable blue 
crab habitat is in the mainstem Bay. T he tora l biomass in each region was computed 
from the biomass of the enti re population using these fractions. The average blue crab 
biomass in each region of the Bay was computed using total surface area rather than 
the surface area of suitable habitat. Accordingly, regional differences in average crab 
biomass reflect only differences in the proportion of habitat shallower than 8 m. Wet 
Weight biomass was converted to carbon ass uming ash-free dry weight (AFDW) is 
16% of wet weight (Ricciardi and Bourget 1998) and 500 mgC=l g AFDW. Estimates 
of net growth efficiency (NGE) and ass imilation effiency (AE) are as rare for blue 
crabs as for other taxa. Schroeder (1981) reported average values of 0.56 and 0.58, 
respectively, for aquatic invertebrate carni vores, and these values were adopted here. 
Blue crab are known to be aggressive predators that can strong ly affect prey 
populations (Virnstein 1977, Hines et al. 1990). While their diets can be varied, they 
are known to feed in Chesapeake Bay on bivalves such as Mya arenaria, and Macoma 
balthica and smaller Rangia cuneata. Large size provides a predation reguge for 
Rangia (Ebersole and Kennedy 1994). In summer, bivalve abundance in Chesapeake 
Bay can be reduced re lative to spring, espec ially the numbers of vulnerable new 
recruits. In response to decreased bivalve ab undance, blue crabs in Apalachicola Bay, 
FL, consumed a more varied di et in summer, increasing quantities of detritus, mysids, 
fishes, and small crabs (Laughlin 1982). Only larger blue crabs maintained a di et of 
bivalves, possibly because of their increased ability to pursue deeply buried prey. 
378 
Larger crabs also obtained as much as I 0% of their diet via cannibalism (Laughlin 
1982). 
Larger blue crabs to not have many natural predators. Smaller crabs are 
consumed by striped bass (Marone sa.xatilis) and by larger blue crabs. Fisheries 
account for the majority of morta lity in larger blue crabs. In Maryland, 60% of fishing 
resul ts from the pot fishery in the mainstem (EPA 1997) and 55% of annual landings 
occur during summer (NOAA Fishery Statistics). Therefore, of the average annual 
Maryland landings of 20,000 tons, summer landings from the mai nstem are 6,600 
tons. The mesohaline mainstem accounts for 70% of the crab habitat in the Maryland 
mainstem, however, it was assumed that 85% of the Maryland mainstem landings 
were taken fro m thi s region because of the prevalence of crab potting near the mouths 
of the major tri butaries of the mesohaline region. Based on the respective areas of the 
upper and mid Bay, landings amout to 1.82 and 2.09 mgC m-2 (Table A-13) . Landings 
in excess of producti on suggest that the pot fi shery intercepted crabs migrating from 
tributari es. In Virgin ia, summer landings account fo r 40% of the annual 16,700 tons 
of landings, refl ecting the winter dredge fishery in Virginia. Since the Virginia pot 
fi shery is operated th roughout the tidal waters, landings were allocated according to 
the fraction of sui table crab habitat in Virginia, - 50% of which is in the mainstem 
Bay. Therefore, the summer landings from the mainstem were estimated to be 3340 
tons, which was equal to 1.09 mgC m?2 summe( 1? Recreational crabbing was assumed 
to be restricted to the tri butaries and was therefore neglected. Natural mortality (0.375 
/ ) was assumed to be constant over the year, amounting to 0.001 d-1, or 0.63, 0.40 
and 0.35 rngC m-2 d-1 during summer. The net migration was computed as the 
379 
difference between total mortali ty and production, neglecting changes in biomass over 
the course of summer (Tab le A-13). While the interpretation is not unambiauous ?t ? 
b , I IS 
suspected that production in the Maryland tributaries as well as northward migrations 
subsidize the blue crab fi shery in the mesoha/ine and upper Bay mainstem during 
summer. 
~ anchovy 
Bay anchovy (Anchoa mitchifli) is the most abundant fish in Chesapeake Bay 
(by numbers), contributing as much as 40-50% of mid-water trawl catches Bay wide 
(S . Jung, unpubli shed data). Larva/ bay anchovy contribute an even larger fraction 
(67-88%) of the summertime larval fish assemblage (Rilling and Houde 1993a). Bay 
anchovy is a maj or component of piscivore diets (Hartman and Brandt 1995), while its 
diet consists almost entirely of crustacean mesozooplankton (Klebasco 1991). This 
makes bay anchovy one of th e most ecologica ll y important fi shes in Chesapeake Ba y, 
connecting phytoplankton production to piscivore production with as few as 3 trophic 
steps. The biomass, diet and bioenergetics of bay anchovy larvae and adults in 
Chesapeake Bay has been studied extensively in recent years, providing excellent data 
for describing the ecological ro le of this fi sh in Chesapeake Bay. 
Estimating summer biomass and bioenergetic parameters is complicated by a 
strong change i 11 age structure. Bioenergetic characteri s ti cs of the population are non-
linear functi ons of individual weight, which changes subtly for the age I+ cohort but 
dramaticall y fo r the young of the year (YOY) cohort. Consequently, both average 
biomass and Lhe associated rates were estimated by separate ly modeling growth and 
380 
mortality or the two cohorts. The size of the age 1+  cohort in mid-sum mer was 
es ti mated fr n egg production mode Is, w hi Ic  a II other abundances were projected or 
back-calcuL11 cd from that estimate. 
For the age 1+  cohott, egg production models provide a robust means of 
eS imating th,? biomass (Rilling and Houde J 993a). These models use measurements 
t
of egg abund? nee made using plankton net tows along with estimates of hourly egg 
mortality, batc h fecundity, spawning fraction and sex-ratio to estimate the biomass of 
adult fishes. Using this approach Rilling and Houde (1999a) estimated that adult 
anchovy biomass was 23606 JV[T in 1993, an average density of 3.73 individuals m?' 
Biomass wa, estimated to be approximately the same in June and July, indicating that 
growth bala ,L ed mortality (Table A-15). 
Usi ,,g a similar egg production model Houde (unpublished data) estimated the 
biomass or a; c l+ anchovy in Chesapeake Bay during summer of 1995-1999 (Table 
A-16). Bio, ,ss varied substantially among years, averaging 70% of the 1993 
estimate. . . regional di stribution of biomass was strongly variable among years 
(Jung, unp .,u ,: ;hed data) and even within years (Rilling and Houde 1993a), making 
assigning an average regional di stribution very difficult. A significant component of 
these chang ., in regional di stributions may reflect ontogenetic and other migrations 
(Jung and I , uJ e 2000, Kimura et al. 2000). Although Rilling and Houde (1993a) 
estimated ,, .. ?her biomass density in the lower Bay, data from Jung (unpublished 
data) s ggc, that averaged over se vera I yea rs, densities may be approximate\ y equal 
O 1 
among regi, ,, . Thus, using a Bay-wide estimate of age l+ biomass of 16,300 tons 
wet weight ('A >I), one obtains an average density of 2.6 g WW m?' . Considering the 
38 1 
interannual \"\ riabi li ty , thi s estimate cornpares favorably with a sli ghtly higher 
estimate fo r 1987 bay anchovy biomass in the mid-Bay (3-4 g WW m-2) derived using 
h sh acoustics (Luo and Brandt 1993) . Wet weight was converted to carbon by 
assuming that dry weight is 20% of wet weight, 10% of dry weight is ash (Wang and 
Houde 1994) and carbon content is 42% of ash free dry weight. Thus, the average 
summer biornass of age 1+  bay anchovy is 197 mgC m-2 (Table A- 17) . 
Using mean body size equal to l.18 g (Rilling and Houde 1993a) , the 
abundance of age l+ anchovy in mid summer (assumed lo be 7/1 5) is 2 .20 m-2 , from 
which one can back-calcul ate the average weight (WJ11111) and abundance (N1111 i1) fo r 
June 1 from corresponding values fo r Jul y 15 (W11111 s and N11111 s) and estimates of rates 
of growt h (G) an d morta I1. ty (Z) us.m g N N j 111\ S , d W - W j 11/I S j un\ = - z ,m j un\ - c . Summer 
e e 
values fo r G and Z were computed by proratin g April -June, June-Aug and Aug-Oct 
bimonthly e::--1i mates fo r G and Z (Wang and Houde 1995, see also Rilling and Houde 
t999b). For example, G was computed as (0 .25)(0.13)+(0.18)+(0 .25)(-0 .06)=0.1975 . 
Similarly coir,puted mortali ty rate was estimated to be 0 .30. The back-calculated 
abundance, mean weight and biornass fo r June 1 were 2.56 m-2 , l.07 g, 2.74 g m-2 , and 
respectively. Projecting these values through the summer, one observes that age 1+ 
biomass decreased by 20 mgC m-2 by the end of summer, an - 10% decrease (Table A-
17). 
The biomass of young-of-the-year (YOY) anchovy durin g summer is more 
difficult to measure directl y than that of the age l + cohort. In part thi s ari ses from the 
fact that recru itment to trawl gear is parti al al best, and egg production models cannot 
382 
be used to esti mate the biomass of this sexuall y immature cohort. Thus, biomass was 
determined by modelin g growth and mortality over time. The approach used was 
identical to that of Wang and Houde ( 1995), except that rates of growth (G) and 
mo1tality (Z) were prorated for Jul y 15-September 1 from July-Aug and Aug-Oct 
bimonthly rates as described above. Total summer egg production was computed 
from the age I+ abundance estimated above (2.2 n,-2)_ Followi ng the approach of 
Wang and Houde (1995), it was assumed that all YOY belong to a single cohort 
spawned on .I uly 15 , the approximate date of peak spawning. Estimates of biomass for 
the cohort began with mean egg biomass. Changes in numbers and mean weight (via 
length-weigh t) were used to esti mate cumul ative mortality (Z) and growth (G), which 
were used to estimate mean biomass usin g 
B = !1, (ec-z - 1) 
G -Z 
where B, is the total summer egg production . The population was assumed to be 50% 
females, wh ic h spaw n an average of 50 times with a batch fecundity equal to 
304.79+404. 64 x female wt. (Wang and Houde 1995). Egg production was computed 
2 
to be 43025 111- summe(1, whi ch is equal to 3.3 1 g WW m-2 (15.4 ? g DW egg- ' , 
Tucker (198() ), eggs are 80% water) . The estimates of G and Z for July 15-September 
1 are 7 .01 a1' ! 6.29, which using eq. (1) gives an average YOY biomass in the second 
half of sum 1'l? 'r equal to 4.85 g WW m-2. Since the YOY cohort was absent in the first 
half of sumrner, the YOY contributi on to summer average biomass is half of the late 
summer average, or 184 mgC m-2 (Table A- 17). At the end of the summer, YOY 
biomass is 5 18 mgC n,-2, an increase in biomass offset by a decrease of only 20 mgC 
383 
m-2 in the age l+ cohort. Thus, summer Bay anchovy production contributed a net 
export to the !~all food web of 498 mgC m-2, equal to a mean rate of 5.4 mgC m-2 d-1 
(Table A-1 7). 
Sum mer production by the age 1+  cohort can be computed from the summer 
average bi o1 :1ss ( B ) and the growth rate (G) using P = GB . For June 1-September 
1, G=0.197 '.'i :1s described above . Thus, summer production is (2.6)(0 .1975)=0.51 g 
WW m-2, or '.) S.8 m gC m-2 . Thi s estimate refl ects only somatic production , however. 
Egg product inn amounted to 3.31 g m-2 summe( 1; therefore somatic and reproductive 
production total s 3 .82 g m-2, or 290 mgCm-2 . The equivalent daily rate is 3 .15 mgC m-
2 d-1. 
Sum mer production for the YOY cohort can also be estimated from average 
biomass ( B ) :rnd the growth rate (G) usin g P = GB . For July 15 - September 1, G 
has been esti1nated to be 7.01 , while B =368 mgC m-2 . Thus, summer production is 
(7.01)(367)=- 2573 mgC m-2, in the same range as reported by Wang and Houde (1995). 
The combin ,? I production of the age l+ and YOY cohorts is 2863 mgC m-2 summe(1, 
an average <1  il y production equal to 31 mgC m-2 c1-1. 
The <Lc omposition of bay anchovy consists almost completely of 
zooplankto1 " lebasco 1991). However, the diet changes during summer as the 
zooplankton :1ssemblage and the age structure of the bay anchovy population changes 
(Tablel8). 'I '1 c effect of the zooplankton assemblage is particularly obvious in June, 
when copep r\ abundance in Chesapeake Bay is typically low , but abundance of 
meroplankt ic .g. barnacle cyprids) and small macrobenthic animals (e.g., 
amphipods) i-, greate r. During the remainder o r the summer, however, the diet 
384 
consists aim "? t exclusively of copepods. Small anchovy (<40 mm, probabl y YOY) 
appear to be limited in their ability to eat large prey such as adult copepods by their 
esophageal (1iameter (Klebasco 199 1). Therefore, these individuals prey on abundant 
copepodites . In general , though, bay anchovy exhibit a particle feedin g mechanism 
that prefers h rger prey. Even very small anchovy (i .e . 8 mm) prefer larger particles, 
leaving behi 11Ll copepod nauplii in favor of copepodites in experimental studies 
(Detwyler and Houde 1970, cited in Klebasco 1991). Rotifers constitute a large 
fraction of j uvenile bay anchovy diets in numerical terms, but due to their small size 
do not consti tute a large fraction by biomass. The largest fraction of identified 
copepods W L i"c l k c1r1ia spp., although a substantial fraction was identified only as 
"calano id co )epods" or "unidentified copepods" (Klebasco 1991). 
The daily ration of bay anchovy appears to be weakly food limited, with food 
limitation most li -..c ly in June when copepod abundance is low . In the other summer 
months, food is sufficientl y abundant that daily ration does not depend appreciably on 
food abund ;1 i..:c l Klcbasco 1991). Using a bioenergetics model compared to growth 
estimates, Luo and Brandt (1993) also concluded that consumption was close to 
maximal (60-70c1; ) most of the time. For July and August, Klebasco (199 1) found that 
adult bay anchovy consumed -17 .8% of dry body weight per day (Table A-19), 60% 
of maximum pm,siblc consumption (Luo and Brandt 1993). Juveniles consumed 
30.3% and 3)% u l' the ir dry body weight in July and August, respectively (Klebasco 
1991 , Table A-19 ), also consistent with weight- and temperature-dependent maximum 
consumptio 11 (Lu u and Brandt 1993). Daily consumption was computed as the 
product of cs i 11 , cd daily biomass for each cohort and the appropriate specific 
385 
consumption , then totaled for each month and for the summer. The summer average 
specific cons umption for the entire population was estimated to be 0.26 d-1 (Table A-
19). The amount of consumption for each diet component was computed from the diet 
fraction and tota l consumption , then aggregated into several prey categoiies (Table A-
20). 
Egestion (E) was computed as a temperature-dependent fraction of 
consumptio n (Luo and Brandt 1993). At 27?C, 20.6% of consumption is egested. 
Excretion (U) is 15% of consumption minus egestion, or 11.9% of consumption. 
2 1
Total egestion plus excretion (E&U) is 2957 gC m- summe( ? The resulting 
assimilation efficiency is 68%. 
Respiration depends on weight, temperature and consumption according to 
R-==- ar wb, f (T )A+ s( c - F), with parameters defined in Luo and Brandt (1993). 
Accordingly, respiration was computed at a daily time increment for each cohort, then 
summed for the summer (Table A-21). 
According to the initial estimates developed in the above description, the 
estimated production, excretion, egestion and respiration exceed consumption by 1278 
mgC m-2 or 14% of consumption (Table A-17). To balance the carbon budget, small 
adjustments were made to the initial values, with uncertainty in the likely direction 
and magnitude of initial estimates considered (Table A-22). Because the estimated 
consumpti on was significantly below maximum (-60% of maximum) consumption 
was revised upward. Similarly, respiration was estimated assuming temperature 
condition s leading to a maximal estimate, therefore respiration was revised downward 
slightly. Production was not changed, while egestion and excretion were adjusted to 
386 
maintain constant assimilation efficiency. The balanced carbon budget resulted in net 
growth efficiency equal to 42% and assimi \ation efficiency equal to 68%, consistent 
with average rates for carnivorous fish (Schroeder 1981). 
Menhaden 
Atlantic menhaden (Brevoort ia 1yra11111ts) is a mi gratory clupeid fish found 
along the entire Atl anti c seasboard from F lorida and Maine. Spawning in the mid 
Atlantic region occurs from May to October. Eggs hatch at sea and larvae are moved 
into estuaries where larvae grow and metarnorphose into juveniles . Menhaden occupy 
much of the Chesapeal eBay and its tidal tributari es. In late fall , juveniles migrate 
seaward out of estuarine areas to large Bays or to the open waters of the coastal zone. 
This hi ghly mi gratory nature complicates the task of estimating their trophic role in 
Chesapeake Bay duri ng summer. 
Menhaden support a substanti al purse-seine fishery on the Atlantic coast, with 
landings during the 1990 's between - 170,000 and 400,000 tons (Vaughn et al. 2001). 
The quantity of thi s catch attributable to removals from Chesapeake Bay cannot be 
known directly due to fishing practices and confidenti ality restrictions on reporting of 
landings. However, Vaughn et al. (200 1) reported estimates of this fraction detived 
from Captain 's Dail y F ishin g Reports. Annual removals of menhaden from 
Chesapeake Bay by the purse-seine fl eet between 1985 and 1999 averaged 150,000 
tons , which was obtained entirely from the Virgini a portion of the mainstem Bay . 
These landings amounted to -50% of the entire Atl antic coast landings for the purse-
seine fl eet. Additional land ings occur in Maryland as the result of the pound net 
387 
fi shery . Average Maryland landings during 1985-1999 amounted to 2 ,000 tons, which 
is entire ly due lo the pound net fishery operatin g in both tributary regions and the 
mainstem. Landin gs by the pound net fishery were about 2,000 tons in recent years 
(NOAA Statistics) . Thi s can be neglected as it is well within the variabi li ty for the 
purse-seine landings . 
Vaughn et al. (200 1) used virtual population analysis to estimate the 
abundance of each age class of the Atlantic coast menhaden stock. Biomass was 
estimated as the product of average abundance at age (Vaughn et al. 200 1) and weight 
al age . Average weights at age for age l+, 2+ and 3+ indiv iduals was obtained from 
Vau ghn et al. (200 1), whi le weight at age for age O+ in mid-summer was assumed to 
be 10 g WW (Luo et al. 200 1). Accordingly, the Atlantic coast stock has a biomass 
equal to - 743,000 tons. Assuming that the exploitation rate within Chesapeake Bay 
is the same as outside the Chesapeake, the Chesapeake stock is 
(1/0.39)(150000)=385,000 tons. This may be an upper limit. If the exploitation rate 
within Chesapeake Bay was hi gher than for the Atlantic coast as a whole, then the 
biomass would be lower. However, there is no ev idence to indicate that this is the 
case. Age O+ individual s, which have a different diet fron, that of adults (see below) , 
were estimated on the basis of the VPA to account for 43% of abundance, but onl y 5% 
of summer bi omass, as individual weight reaches a maximum in late fall (Luo et al. 
200 l) . The distribution of n1enhaden within the Chesapeake Bay is broad and appears 
to exc lude very little of the surface area. Therefore, the average biomass of menhaden 
was di stributed over 18,000 km2 , the approximate surface are of the Bay and its 
388 
2 2 
tributa ri es. The resulting biomass esti mate is 21 g m- , or 2136 mgC m- (Table A-
13) . 
Consumption by menh aden was estimated by applying the menh aden fora bo inob  
model described by Luo et al. (2001) . It was assumed that menh aden occupied water 
1 
with an average chl orophyll-a concentration of I O ~Lg i- which was converted to 
carbon assuming C:Chl -a==50. Water temperature was assumed to be 25?C. The 
est i mated age structure deri ved from Vaughn et a I. (200 I ) wu s a pp Ii ed to estimated 
s pee i Fi c con sum pt ion at age to deri ve a popu Ia tio n average s peci Fi c consum ption ra te 
of 0. J d-1 (Tabl e A- 13). 
M enhaden growth rate depends on the habitat in which it grows. Luo et al. 
(200 J) show that menhaden apparentl y seek out and occupy habitat with sui tab le food 
density, achiev ing average growth rates of - O.D25 d? ' during summec Suitab le habitat 
apparentl y consists of substantial phytoplankton blooms, which the schools of fi sh 
may id ent i f y through c hemosensory abi Ii ti es (Fried Ia nd et al. J 98 9) a Menhaden 
assimil ate their diet with very high effi ciency, -87% (Durbin and Durbin 198 L). 
Landings of menhaden from Chesapeake Bay by the purse-seine fl eet, as noted 
above, amount to l 50,000 tons/ . Other menhaden landings amount to - 4,000 tons 
i' and were assumed to come predominantly from outside the mainstem Bay and 
were therefore neglected. Approximately 50% of purse-seine landings occur during 
summer (NOAA Statistics), and all of the landings are from the lower Bay region of 
the Chesapeake Bay mainstem, which has a surface area or 2,66 J km 2. Thus, daily 
1 
_z r1 0 1?  30 6 C -
2 1-
removals during summer amountlO 0 ? 3 g m c ? ? mg m c ( rable A- 13). 
Thi s is equal to 57% of estimated dail y production by the Chesapeake Bay stock in the 
389 
71 .. ,. , ?? ~: ?(?") : \?t .~ , .. ,t .?.. ?, ~'' . ~"l ",) \.1.~ ? ? ? I 
lower Bay . Production in the upper and middle portions of the mainstem Bay is 
largely unexplo ited by the fishery. 
Herring and Shads 
A diverse assemblage of hetTings and shads were captured in Chesapeake Bay 
mid-water traw ls (S. Jung, unpubli shed data). These inc lude include the Ameri can 
shad (A losa sapidiss inw) , blueback herring (A losa aesti va l11 s), alew ife (A losa 
pseudo /i are11 g11 s), hickory shad (A losa 111 ediocris), gizzard shad (Dorosoma 
cepeclicrn11m ), Atl anti c thread herring (Op istho11 e111a captiva i), as we\\ as unclassified 
C lupeidae belonging to vari ous genera. The e species make annual spawning 
mi grati ons to fresh wate rs in summer (EPA l989) and were found in greatest 
abundance in the upper Bay (S . Jung, unpubli shed data). Most individuals that were 
captured were youn g of the year. B ased on trawl data, S . Jung estimated the YOY 
abundance in the upper Bay during summer to have vari ed between near zero in the 
apparently fa il ed recruitments o f 1999 and 2000 up to 1.25 million in l996. Typical 
summer abundances YOY were 200-400 million (S. Jung, unpubli shed data). Based 
on the summer length -frequency, YOY blueback he rrin g were about 6.0 cm while 
YOY alewives were s li ghtly larger, 8- 10 cm. These lengths correspond to weights of 
2 and 9 g, respecti vely. Assuming that the combinati on of these species had an 
average weight pe r indi vidual in between these values, total bi omas can be estimated 
to have been 750- l SOO tons. 
Another approach to estimate the biomass of herrings and shads is based on 
the ir mortality rate and estimated demand by predators. lt was esti rnated that 
390 
c o nsun, ption of a\osids amounts to l.64 mgC m-2 d-1 2 1, or 15 l mgC m- summe{ . 
Bas d on a max. imum \ongevity of 8 years, Hoeni g's ( l983) mode\ suggests a natura\ 
rnorta\ity rate of - 0 .7. However, young-of-the year may be assumed to be sub_\ect to 
hi gher morta\ity . Hoeni g's mode\ predicts M=2 .0 for a near\ y annua.\ fi sh such as Bay 
anchovy , whi\e Wang and Houde (1995) estimated M=6.29 fo r the youn g-of-the year 
Anchovy durin g \ate summer. Assuming that YOY morta\ity for a\osids was -0.5 fo r 
the summer, then 39% of the biomass is consumed by predation each summer and the 
average biomass is equa\ to 382 mgC m-2 , or - \ 800 tons (Tab\e A- 13). 
A third approach to obtain a. minimum estimate of a\osid biomass is to assume 
that summer production by a\osids is at \east equa\ to the predation demand. Based on 
the growth characteri stics for Alosa pseudoharengus (Froese and Pau\y 200 l ) , specific 
growth for an individua\ - 6 cm in \en gth is rough\y 0.0 l d-1 . Therefore, a biomass of 
\ 63 mgC m-2 is required to support a predatory demand of l .63 mgC m-2 d-1. Over the 
area of the upper Bay , thi s is cqua\ to a bioma s of 771 tons . 
Based on the e three approaches, the biomass of young of the year a\osids in 
the upper Bay was estimated to be about 1000 tons, or 2 12 mgC m-2 (Tab\e A- 13). 
Production was assumed to be 0 .0 l d-1 , with assimi\ation and net growth efficiencies 
equal to 0 .68 and 0.4 l , respective\y (Schroeder 198 l ) . The YOY a\osids were 
as umed to be primarily zooplanktivorous (Froese and Pau\y 200 \ ). 
Striped Bass 
The striped bass (Moro11 e saxatilis) is an important piscivore in Chesapeake 
Bay and supports substantia\ rccreationa\ and commercia\ fi sheries (EP A \ 995) . \n 
391 
Maryland, commercial landi ngs from the Bay averaged about 240 tons {' in the early 
990 's (EPA l 994). R ecrea ti ona l Ia ndi n gs accounted for an addit i anal 3 7 4 tons {' . 
1 
Landings from the Potomac River averaged 74 and 69 tons{' for commercial and 
rec reational landings for a total of I 44 tons y' . In Virginia, commercial and 
recreation landings averaged 100 and 235 tons {' for a total of 335 tons {' . Thus, 
overa ll Bay wide landings of stnped bass in the earl y l990's were about 1000 tons y '. 
Maryland landings , which inc lude the coasta l landings , increased 4.5-fo ld in 
0 
0 
the Ia te l 990 's relative to the earl y I 990 's as the s we k continued to recover from Io n 
term lows and a fishing moratorium in the J980's. for the period 1985-1999, average 
' 
Maryland landings were about 2-fo ld greater than the early 1990's levels. Therefore 
Chesapeake Bay Jandings during thi s period may have averaged about 2000 tons{' . 
EPA ( J 995) repmted on the basis of tag-recapture studies that the fi shing mortality 
rate in recent years was f=0.09. Based on that estimate and assuming natural 
mortality rate is M=O. 15, then the exploitation rate ( ? = lFV - e - M + f )V(M + F) ) is 
0.08 y' , giving an estimate of stock biomass equal to 25,000 tons via B = C / ? 
The striped bass stock was ass umed to be distributed over the entire surface 
2
area o f C hesapeake Bay and its tidal tributa1ies (I 8,000 km ), giving an average 
density of ].38 g m?', equivalent to 172 rngC m?' (DW=0.25WW , Ha1tman and Brandt 
1995c, 1 g DW==500 ingC) (Table A-13). 
An estimate of growth for the population was derived by constructing a 
theoretical population structure. A van BertnlanffY  growth equation 
( L, = L~ V_  e - K(,-1, )) with the parameters L,,= !39 cm, K=0.117, and to=O (Mansueti 
392 
\ l)6 \ , Froese and Pau\y 200 \) was used to estimate \ength at age and annua\ growth 
i n c rc menl for each age. Length (en,) and growth increment was converted to wei ght 
(g,) via w = aLb where a=0.006 \ and b=3 .153 (Mansueti 196 1, Froese and Pau\y 
1 
2.00 \ '). Wei ght-specific growth decreased from 2 i for age 0- \ fish to 0 .08 i 1 for a 
\ 5 year o\c\ fish. These functions how that striped bass do not reach \ega\ size unti\ 
aoc 3-4. Therefore, fi shing morta\ity on those age c\asses was reduced to zero. T he 
:::, 
rcsu\ting age structure a urning M=0.15 and F=0.09 for recruited age c\asses (Age 
4+) and the computed growth at age gives a specific growth estimate fo r the 
popu\ation equa\ to 0 .3 / . However, summer production may be more or \css than 
the annua\ mean rate . 
Specific rates of production, consumption, respiration as we\\ as the 
assimi\ation efficiency were obtained by app\ying coefficients reported by Hartman 
and Brandt ( l995b) to the estimated age structure of a Chesapeake Bay striped bass 
popu\ation assuming a water temperature of between 25-28?C . At these temperatures , 
temperature-dependent specific consumption shou\d be near the max.imum, which 
distributed over the age structure gives an average of 0.034 d-1 , simi\ar to that of an 
age-7 individua\. Assimi\ation was found to be 83% of consumption. Active 
respiration rates, d rived as g 0 g-1 2 d-
1 were converted to g WW er' assuming 3240 
ca\ g-1 , 4 .184 J car' (Hartman and Brandt l 995b) and 4000 J g-1 WW (Hartrnan and 
Brandt 1995c). For the estimated age structure, the average specific respiration rate at 
25?C is 0 .016 d-1 . Specific dynamic action amounts to \ 7 .2% of assimi \ation and is 
a\so a component of the respiration rate . A suming max.ima\ con umption the scope 
for growth is 0 .0076 d-1 , -9 times greater than the annua\ rncan rate . Uthe striped 
393 
bass populati on is able to achieve on ly 75% of maximum consumption, then the 
summertime specifi c growth is 0.00 18 d-1, twice the annual mean rate. At that rate of 
consumption, acti ve metabo lic and SD A equal 0.0 19 d-1, giving NGE=0.083 (Table A-
13). At C=Cmax, NGE=0.27. 
Striped bass undergo ontogenetic shifts in the ir diets durin g summer (H artman 
and Brandt l995d). Age-0 fi sh in earl y summer ate primaril y small invertebrates, but 
by late summe r ate a substantial fraction of juvenile naked goby. By age 1+, bay 
anchovy and menhaden accounted fo r 50% of the summer diet, w ith the remainder 
in vertebrates. Age 2+ and o lder fi sh primaril y ate fi sh (80 %), but obtained about 20% 
o f the diet from be nthic and epibenthic invertebrates (Hartman and Brandt 1995d). 
Ho lli s ( 1938) came to similar conc lusions after examining the stomachs o f 1,738 
striped bass in C hesapeake Bay in 1936-37 . In the upper Bay (Rock Hall and 
Conowingo) the die t al so inc luded shad, white perch, eel, and yellow perch (Holli s 
1952). Biomass was estimated to be overwhe lmingly dominated by age 2 and o lder 
fi sh . There fore, the di et for the populati on was assumed to re fl ect their d iet (50 % 
me nhaden , 20 % bay anchovy , 10% other fi sh, 20% invertebrates). 
Blue fi sh 
Abundance o f bluefi sh (Pomatomus saltatrix), a piscivorous predator, in the 
Bay peaked in the late l970's and is known to be hi ghl y vari able. T he blue fi sh stock 
is highly exploited , principall y as a recreational fi shery (>80% o r total explo itation) . 
Based on fi gures presented by EPA ( 1990), recreational landings in the Bay have been 
estimated to average 1360 tons in M aryland, w ith a similar fi gure in Virgini a. T he 
394 
comme rc ial fis hery inc ludes both inshore and offshore components. EPA (1990) 
re ported that from 1970- 1986 th e Bay landings dominated the tota l, but accounted for 
a dec reas in g frac tion . Ana lysis of land in gs by gea r type for 1985-1999 suggests that 
Bay landings accou nted for a small er fraction o f the total in recent years (-30%). 
Acco rdin g ly, commercia l landings in the Bay were estimated to average 64 tons/ in 
Maryland and 154 tons/ in V irginia. T he to ta l land ings of blue fi sh from the Bay 
w he re the refore estimated to average 2940 tons/ over the past 15 years. ln 
Maryland, hook and line fi shing acco unted for two-t hirds of commercia l land ings, 
w ith the remaining fraction coming from pound nets. In Virgi ni a, the vast majority of 
the commercial landings were due to the pound net fishery . All the bluefish fishe ri es 
were concentrated durin g May-October, w hen Bluefi sh are reside nt in the Bay. 
F ishe ry utili zat ion of the bluefish resource was computed by assum ing that 
Maryland C hesapeake Bay landings were from the upper and mid Bay and that 50% of 
these landings were from the mainstem Bay. Half of the landings were ass umed to 
occur durin g the three months of summer, since the season lasts 6 months, from May-
O ctober. Maryland landings were a ll ocated to the uppe r and mid Bay on a re lati ve 
area bas is . Half o f the Virgini a C hesapeake Bay landings were ass umed to come from 
the lower mainstem Bay, with the balance coming fro m Virgini a tributari es. 
Growth and mortality rate es timates for blue fish were derived from Barger 
(1990, c ited by Froese and Paul y, 2001) and EPA ( 1990) . Bluefi sh have an estimated 
lifespan of 10- 12 years. Natural mortality has been estimated to be 0.35 / for 
yo un ger bluefish and 0.6 / for o lder fi sh . Fi shing mortality rates in fo r the Atl anti c 
s tock as a whole ha ve been estimated to be 0 .7 / for youn ger fi sh and 0 .27 / fo r 
395 
p1esence 
older fi sh. Increased fis hing pressure on smaller fis h is due to their increased . 
n approx 1rna e ages rue ure an we1g1t at age was computed using 
Ill in shore c?11?eas . A ? t t t d ? I 
ese fi gures, and assuming the coefficients of the von Bertalanffy growth eq uation are 
th
Ln= 11 5 cm and K=O. l35 and the weight-length relati on has coeffi cients a=0.0 131 
and b=2.934 (Barger 1990, Froese and Paul y 2001). Based on the age-structured 
fi shing mortality rate and the estimated total ex ploitation, the average exploitation rate 
rs ?~0.36 and total biomass is 8 I 30 tons. Young of the year and adult bluefi sh are 
b 
known lo occupy alI  sal i nitres (EPA I 990), while the relatr ve di stribu tion a mono 
b 
regions was unknown. Therefore, the average biomass was assumed equal th rouoh 
? . b0  
out the Chesapeake Bay and tributari es (18,000 km?) and was estimated to be o 45 
WW m?', or 68 mgC m?' assuming dry weight is 30% of wet weight (Hartman and 
Brandt 1995c) and I g DW=500 mgC (Table A-13). 
Bi oene rgeti cs f b Iu efi sh were described bY  Hartman and Brandt ( l 99 5 b). 
O
Temperature dependent consumption was equal to 98% of Cmax when temperature 
was between 23 and 28 "C, therefore consumption was assumed to not be temperature 
limited. Active respiration rates were computed as for striped bass, but using 
coeffi cients for bluefi sh. Moder,te food limitation was assumed by estimating 
consumption lO be onl y 75% of the temperature-dependent maximum. Accordingly, 
summer average scope for growth was -2-foid greater than annual mean growth 
derived from the age- length and weight-length relations. Us ing these assumptions, 
OpLII t? ?,.. . c rates of growth consumption and respiration (includin a 
P a ron average speer r ' o 
spec ? f.. d . . ) wei?e estimated to be 0.0069, 0.094, and 0.052 d? ' , 
? 1 1c ynam1 c acuon 
res pee ti ve ly . Ne l growth effi ciency (NGE) was 0. I 2 (Table A- 13), less than the food-
396 
rcplclc NGE under the same assumptions, which is 0.29. Assimilation efficiency for 
b ludish was assurned lo be the ame a for striped bass, 0.83 (Hartrnan and Brandt 
\ 995b). 
Unlike striped bass, bluefi h ate a diet consisting a\rnost ex.elusively of fish. 
/\<se O+ bluefish consumed almost ex.elusively bay anchovy and si \verside during 
sumrner (Hartman and Brandt l 995a), but this age class was estimated to account for 
onl y 3c11i r) or biomass. Age \+bluefish , which were estimated to account for 2 l % of 
biomass, obtained the large l fractions of their diet from bay anchovy and spot, but 
a\ ?o significant contributions from si\versides, rnenhaden , and other fi sh. Age 2+ fi sh , 
unlike their younger counterparts, obtained a significant fraction of their summer diets 
(60%) from Atlantic croaker. The average diet, accounting for the fraction of biomass 
at each age was approx.irnate\y: Bay anchovy 3 \ % , Menhaden 8%, Spot \ 5%, 
Croaker 46%. ln fa\\, however, the diet of larger bluefish consisted almost ex.elusively 
or menhaden, while that of younger bluefish consisted of bay anchovy and menhaden 
(Hartman and Brandt l995a). H ence, the d pendence on bluefish on spot and croakers 
was limited to summer. 
Weakfish 
Weakfi sh (Cy 11 oscio11 regali s), like striped bass and bluefi sh , is a major 
Chesapeake Bay piscivore that supports both commercial and recreational fi sheries. 
Commercial landings in Maryland and \/i rginia included a substantial coastal 
component. hesapeake Bay commercial landings were estirnated to have been about 
804 tons/ in the l980s (EPA l990b), of which 726 tons/ were frorn Virginia 
397 
1406 
waters. Recreational land ings were aboul 2-fo ld greater, l556 tons /' , of which 
were from Virginia waters. Total landings of weakfish amounted to 2360 tons i'. 
Most (90%) of the commercial weakfish landings in Maryland occun-ed in April , 
0 
October and November, whi le in Virginia, 32% of the landings occurred durino 
summer. It was assumed that about half the the summer landings in both slates were 
f'rom tributary waters, rather than the 111ain ste111 Bay. 
The total biomass of the resident Chesapeake Bay population was estimated 
c,, 
using the tota l landings from the Bay, information about the ages subject to fi shino 
available growth and mortality data, and approximate estimates of young-of-the-year 
abundance based on trawl data (S. Jung, unpubli shed data). There is no estimate of 
10 
the total mortality rate for the weakfish stock. Natural mo1tality rate was estimated 
be - 0. 7 i' based on Hoenig 's (1983) model and a Io nge vit y of a bout 6 years for 
weakfish in the mid Atlantic 1-egion (EPA l990b). The parameters of the von 
Bertalanffy growth equation (V"=67, K=0.35) and the length weight parameters 
(a=0.009 1, b=2.984) used to estimate length and weight at age. These 1-esu lts, 
combined reports of the length-frequency and age-frequency in the catch (Sadzinski 
ct al. J999) were used to estimate that fishin g mo1tality on age 2+ fish was - 0.6 
Accrodin o obtainin o the an nual landings from the Chesapeake Bay requi1-es a youn o-
"' 0 0 
of-the-year (YOY) abundance equal to - 45 million individuals. Projecting thi s 
abundance throught the age structure gives a total weakfish biomass of about 6000 
tons. Based on repo1ted li fe hi story characteri stics (EPA J990b), the TIES trawl data 
(S. Jung, unpublished data) and the relative contribution of Virginia and Marytland 
landings to the total , it was estimated that 75% of the population resided in Virginia 
398 
wate rs. In Maryland YOY biomass was concentrated in the upper Bay (S. Jung, 
unpubli shed data). O lder fi sh were in greater abundance in the mid Bay than the 
upper Bay , reflectin g their sa lin ity pre ferences (EPA 1990b). Lacki ng better data, it 
was ass umed that the bi o mass of the Mary land portion or the popul at ion was equally 
di stributed between the upper Bay and the mid Bay . Half o r the biomass in both 
regions was assumed to inhabit tributary waters rather than the mainstem Bay. T he 
same ass umpti ons were app li ed to Virgini a (lower Bay) waters. Accordin gly, the 
average weakfi sh biomass in the upper and mid Bay during summer was estimated to 
have been 67 mgC m-2, whi le biomass in the lower Bay was estimated to have been 
2 11 mgC m-2 (0 .25 gDW= l g WW , Hartman and Brandt 1995d, l g DW=500 mgC) 
(Tabl e A- 13). 
Detai led age-structured bi oenergetics data for weakfish were obtained from 
Hartman and Brandt (1995b) and were used to compute scope for growth for the age-
structured weakfish population at summertime temperatures. Summer scope for 
growth was found lo be 14 times greater than the annual mean growth rate computed 
rrom the von Bertal anffy growth parameters and the length -weight re lati onship . Since 
scope for growth fo r weakfish is negative at temperatures below l 5?C (Hartman and 
Brandl 1995b) and is nearly maximal at 25?C, it was assumed that summer scope for 
growth was 3-fold greater than annual mean growth , whi ch was achi eved when 
consumpti on was - 60% of C,,,a,. Accord ingly, specifi c consumption was computed lo 
1 
be 0.068 d- and speci fie growth was 0 .0088 d-1? Assimi lation efficiency was reported 
to be 0 .83 (Hartman and Brandt 1995b) . The computed values impli ed that net growth 
e fficiency was 0 .16. Bion1ass accumu lation was computed as summer production 
399 
minus the sum of summer landings and summer mortality (M~0.7/365 ct?') and was 
CS imaled lo be 0.42 mgC m?' ct?' for the upper and mid-Bay and J .03 mgC m?' ct?' fo r 
t
the lower Bay. 
The sum mer diet of age-0 and age- I weaHish consists of about 75% bay 
anchovy with the remaining fraction coming from small invertebrates. At age 1+  the 
diet consists of 70% bay anchovy, 8% menhaden, 10%o ther fi sh, with the balance 
obtai ncd rrom in vertebrates. Hartman and Brandt (1995b) did not report the mid 
summer diet ror age 2+ weakfish in the mrd Bay. However, based on the May-Jun and 
Sep-Oct diets o r age z+ weakfish, it can be infe n?ed that the diet includes about 40% 
C, 
menhaden, 30% spot, about 20% other fi sh, and and JO% invertebrates. The avcraoe 
diet was computed from the proportion at age and the age-specific diets. Jt was 
assumed that a relati vcly larger fraction of the population was YOY in the upper and 
mid Bay and that a larger fraction was age 2+ in the lower Bay? Average diet was 
computed accordingly. 
Atlanti c Croaker anc;LS.)2.Q1 
Atlantic croaker (Micropogonias u11d11/a1us) and spot (Leios1011111s xan//mrns) 
arc important dcmcrsal feeders in Chesapeake Bay. They are an important component 
of the di cl of striped bass and bluefi sh (Hartman and Brandl J9 95a, see above), and 
also support commercial and recreational fi sheri es during summer (EPA J9 9 IA). 
At Ia ntic croaker spawn on the continental she IF, then both ju ven i Jes and adu Its move 
into estuarine areas in late spring through mid fol I. Young of the year are distributed 
throughout the estuary, whi Je adu lts are more abundant when sal in il  y >5  ppl (EPA 
400 
ti 
199 1A ). Croaker have been in relativel y low abundance in recent years, however, 
mid-water trawl surveys (S. Jung, unpubli shed data) suggested that stron g 
rec ruitments occurred in 1996 and 1998, lead ing to increased abundance and landi ngs 
in the late l 990's (S. Jung, unpubli shed data , NOAA Landings Data). 
Commercial landin gs of croakers were dominated in the Bay by the pound net 
l"ishery. Average pound net landin gs durin g 1985-2000 were 884 tons, or whi ch 782 
tons w..is in Virgini a. The rec reati onal fi shery was of similar magnitude, with total 
Chesapeake Bay landin gs of - l.5 milli on pounds (EPA 1991A, equal to 682 tons) . 
Virginia recreati onal landin gs typi call y accounted for about 75% of the rec reati onal 
landings, thereby giving a figure of 511 tons in Virginia and 170 tons in Maryland. 
Thus, average total landin gs of croaker in the past 15 years was estimated to have been 
- 15 65 tons i' . 
EPA (1991) suggested that an appropriate total morta lity rate for Atlantic 
croaker was Z= l . 15 , while Ross (1988) estimated Z=l .3 from catch curve analysis. 
Natural mortality was estimated to be - 0.5 i' by assuming a lon gevity of of 8 years 
(EPA I 991 A) and app lying Hoeni g's (1983) model. This imp lies an ex ploitati on rate 
of 0.38 , and therefore a tota l biomass of 4,118 tons. Given growth parameters for 
croaker (Ross 1988), population production based on thi s estimate was not suffi cient 
to account for estimated bluefi sh predati on, wh ich was -2.2 mgC m-2 d-1. Assuming 
summer growth was 2-fo ld greater than the annual mean rate and assuming a 
population dominated in nun1bers by age O+ and age l+ individua ls, specifi c growth in 
summer may be - 0.0 I d-1. Thus, a bi omass of 222 mgC m-2 d-1 is required to support 
blue fi sh predation. An additional bi omass is r quired to support the fishery . 
401 
Assuming hair the Mary land landin gs were from the mid Bay mainstern and that half 
the landin gs removed summer production , landings amounted to 0.032 mgC m-2 d-1, 
which wo ul d requ ire an additional 3 .2 rngC 21T1- . \n Virginia, landings were greater , 
amounting to 0 .13 m gC m-2 d-1 under the same assumptions, requiring 13 mgC m-2 
addit ional bio mass. T hus, a m inirnum estimate of croaker bion,ass was computed to 
be 225 and 235 m gC m-2 in the mid and lower Bay, respectively (Table A- 13) . 
Croakers were much less abundant in the upper Bay in T \ES mi d-water traw ls. 
Therefore, it was assumed that croakers were not abu ndant the re. Lacking spec ific 
bioenergeti c data fo r croakers, assi mil ati on and net growth effic ienc ies were assumed 
to be 0.68 and 0.4 1, reasonable values fo r carni vorous fish (Schroeder 198 1) . 
Consumpti on , resp iration and excretion plus egestion were computed usin g these 
rati os. 
Spot (Leios to 11111 s xc111tllll r11s) is d istri buted throughout the salinity ran ge of 
C hesapeake B ay (EPA 199 l A) and were caught in the TlES mi d water traw ls (S . 
Jung, unpub li shed data) in interannuall y varyin g abundance in all regions of the Bay. 
Spot supports both a commercial and recreati onal fi she ry, which are most important in 
V irgini a w here spot are generall y larger. T he largest catches in Virgini a occur in late 
summer to earl y fa ll when mi gratin g spot are captured in V irginia . Annual 
commerc ial land in gs in V irgini a averaged 1330 tons, about a third of which was 
landed in summe r. Landin gs in Maryland averaged about 78 tons, about 50% o f 
which was in summer. EPA (l 99 1) reported th at recreati onal landin gs in Maryland 
were about I .3 millio n !'i sh in 1980 , which is equal to about 9 1 tons of age 2 fish . In 
Virgini a , rec reati onal landin gs in the mid- I 980's were 1.6 million pounds, o r 727 tons, 
402 
1986 and about twice that in 1985 (EPA 1991A). Thus, total annual landings of 
111 
spot arc probably 950 tons/, about 330 ton s of which was landed in summer 
Commercial and recreational landings of spot in Maryland were assumed to come 
rrom the mid Bay. Per unit area these landings were equal to 0.003 and 0.021 mgC m? 
, respecti ve ly. In Vi rgini a, recreational and commercial landings per unit area 
2 cl - I 
1
were cq ual to 0.089 and O. 0 I 5 mgC m? 2 d- , respecti vely (Table A- I 3) . 
Since fi shing mortality rates are not known, there is no direct means of 
est imating the abundance of spot in the Bay. Given an est imate or specific growth 
rate, mini mum biomass can be estimated as that required to suppo1t Ia ndi ngs and 
est imated consumption by dominant predators. Growth parameters of spot were 
2
reported by Froese and Paul y (200 1), apparentl y in fe rred from data in Pacheco (196 ). 
150 
Maxi mum Jongevit  y is 4-5 years, in which ti me a maxi mum size of -24 cm and _ 
0 
g (K~0.725, a~0.0092, b~3.072; Froese and Paul y 2001). Most indi viduals are age 
to age 2 (Pacheco l 962), and growth in thi s time is rapid (K~0.725), declining from 
1
- 0.02 d-' to 0.006 d-'. At a speci fi e growth rate equal to 0.01 d- , the biomass 
required to balance summer landings in the mid and lower Bay is 3.9 and 25.2 mgC m? 
, which summed over those regions is equal to a wet weight of 763 tons. Estimated 
2 195
consumption or spot by striped bass, blueri sh and weakfish requires a biomass of , 
2 I 7 and 545 mgC m?' in the upper, mid and lower Bay, respectively, which is equal to 
a total wet weight of 20,506 tons. Total biomass is therefore 20,268 tons. The above 
noted estimate of speci fi e growth (0.01 d-') was assumed, while growth efficiencies 
were estimated to be the same as for croaker (AE~0.68, NGE~0.41) (Table A- I 3). 
403 
While Perch 
Behind bay anchovy, white perch (Morone americano) contributed the second 
largest biomass to T IES mid water trawls. Based on the TlES data, the presence of this 
species in the mainstcm Bay was limited to the upper Bay, however, it is also known 
lo be present in the upper reaches of the mesohaline Bay. This refl ects the salinity 
prefc rcnccs and spawning migrat ions of this species (Mansuclr 196 1) . While perch 
abundance in the mid Bay may be greater than suggested by the TJES data if il was 
abundant in shallower waters than were sampled by TIES. 
Commercial landings in Maryland averaged 466 tons during 1985-2000, with 
landings increasing through the peri od, especially in the l 990's fo llowing strong 
recruitments. It is not known what fraction of these landings were from the mainstem 
Bay, however, a fi gure of SO% seems plausible and was adopted. Recreational 
landings were -200 tons/ in Maryland in 1998 (Sadzinski et al. 1999) and were 
0 
assumed to be - 1Q O tons in the upper Bay. The vast majorit y of commercial Jandinos 
occu1Ted during November through May, with summer land ings <3% of peak spring 
Ia ndi ngs . Vi ,gin ia  Jandin gs averaged l 4 % of Maryland landings and were assumed to 
have been obtained from the tributaries due to high salinity in the Virginia mainstem. 
Biomass in the upper Chesapeake Bay was estimated by modeling growth and 
mortality of a hypothetical stock. Growth was modeled assuming L.=35 cm and 
K=0.2 /, which was increased from the estimate reported by St. Pierre and Davis 
( 1972) to match the length frequencies observed in the upper Bay TIES data in 1997 
Length-weight pa rumeters were estimated lo be a=O.0 07 and b=3. I 7. N alu ra J 
mortality was estimated to be 0.6 based on the catch curve analysis of SL Pierre and 
404 
Davi s ( I 972) and Hoenig's ( l 983) regression 111odel. To obtain landings or 250-350 
tons/ with a size structure consisten t with data reported by Webb and Z lokovitz 
( 1999) and wi th com mercia l size restrictions (>20.3 c111), recruit111ent 111ust be -50 
million/' w ith fishing 111ortality rate equal to - 0.4 on age 3+ fish and negligable on 
younger indi v idual s. This level of recruit111ent is consistent w ith the s111aller, but not 
failed recruit111ents uggested by the TLES data (S . Jung, unpubli shed data. Based on 
the esti111ated recru itment ,co, rowth and 111ortality, tota l bio111ass is about 2000 tons. 
Larger landings (i.e. if recreati onal fi shing is 111ore substantial than suggested) wou ld 
require greater rec ruit111ent since the yield was near the calculated maxi111u111 yield-per-
recruit. One third of the bio111ass was assumed to be in the upper mesohaline Bay, 
w hile the reamaining was assu111ed to be distributed over the surface area or the upper 
2
Bay. According l y, the bi o111ass density is equal to 282 111gC m- , similar in abundance 
to bay anchovy. The average 111id Bay abundance is esti111ated to be 29 mg 2111- , which 
is low due to the large surface area of the 111id Bay region. 
Consistent with the age structure of the populati on, the bio111ass-weighted 
average age in the stock was -3, for which specific growth is - 0 .002 cr1 (Table A-13). 
rr the average age i s somewhat younger, producti on could be 111uch greater, as hi gh as 
0.008 if the population was do111inated by age I fi sh. Total su111mer producti on was 
estimated to be 52 and 5.3 mgC 111-2 in the upper and 111id Bay, respectivel y . Natural 
111ortality in the two regions during sum111er was - 14% o f bio111ass = s(1- e - M/4 ), or 
2
39 and 4 mg m- . Since fi shing 111orta l ity during su111mer was negli gable, bi o111ass 
accumu lation must be -0. 14 and 0.01 111gC 111-2 cr 1 in the two regions. White perch 
have been reported to be primaril y consumers or zoobenthos as adults (Froese and 
405 
Pau IY  200 I); however, significant consumpuon of bay anchovy by white perch has 
also been observed ,n the upper Bay, suggesting that these fish may also be a major 
,et component when abundant. Ju venile bay anchovy are also zooplankti vores 
ct ? . 
Consumption and cgestion was computed assuming AE=0.68 and NGE=0.41 
(Schroeder 198 1). 
Catfi sh 
The channel catfi sh, Jctol"rus P'"'c1a1"s and other catfi shes occupy low 
salinity habitats and were captured in significant abtmdance in TJES mid-water trawls 
in the upper Bay (S. Jung, unpublished data). Commercial landings in Maryland 
average 760 tons i 1, SO% of which was assumed to have come from tributary 
populations. Vi rginia land ings of 477 tons {' were assumed to have been obtained 
entirely from tributaries and are ignored here (NOAA Landings, Sadzinski et al. 
1999). Recreational landings in Maryland were reported to have been -200 tons {' , 
of which SO% were also assumed to have been been obtained from the tributaries. 
Thus, total removals from the mainstem Bay were esti amated to be 480 tons{'. 
Catfi sh land ings occur throughout the year at a fairly constant rate (NOAA Landings). 
Thercfo re, 2S% of annual landings were assumed to occur in summer 
The channel catfi sh has a maximum reported age of 16 yea1s, with growth 
13 2  
para meters K =0.0 6 and L,,= J J9  and length -weight para meters a =0.0 04 2 and b= 3. 
(Pau Iy  and Froese 200 l) . Natural mortality for cal fi sh was estimated to be 0.2 based 
on maxi mum longevity (Boeni g I 983), while Sadzi nski et al. ( J 999) reported that total 
mortality of the Choptank River stock was 0.8 for recruited age classes. Minimum 
406 
I. . ~ .. . "-~~ .., . .. ' .? - .' . . .. 1111 
marketable size is 48 cm (Sadzinski el al. 1999), which corresponds approx imately to 
an age-8 catl'i sh. Thus , fi shin g mortality was assumed to be 0.6 on age 8+ catfi sh. 
Age 6 and 7 fi sh were assumed to be partia ll y rec ruited (F=O. l , 0.2 , respect ive ly), 
consistent wi th size-structued pound net landings reported by Sadzinsk i et al. ( L9 99). 
T hese estimates or ri hing mortali ty al age gave close lo maximum yield-per-rec ruit , 
indi catin g that annual recruitrnent to age I+ was - 6 million . Based on estimated 
growth and mortality, total biomass was estimated lo be 3,200 tons. As with white 
perch, it was assumed that a fraction of the catfi sh population inhabits the upper 
mesoha\ine zone of the Bay. A figure of 30% was assumed. Di stributed over the 
upper Chesapeake Bay, two-thirds of the population has a biomass density of 450 
mgC m-2. In the mesoha\ine Bay , one-third of the population has an average density 
of 45 mgC m-2 (Table A- 13). 
The mean biomass-weighted age of fi sh in the population is 6 y, for which 
speci fi e growth is 0.00 l d-1? Average ass imilati on and growth efficiencies were used 
(AE=0.68 , NGE=0.4 t ) to compute production. The diet of sma\ \er catfi sh includes 
mostly benthi c in vertebrates, however, as catfish grow they apparen tl y consume an 
increasin g fraction or sma\ \ fish (Poe et al. 1991 ). 
Hogchoker 
The hogchoker (Trinecres 111ac11latus) is known to be abundant in the Bay, but 
because it is unex pl o ited and little studi ed, cert ain estimates of it 's biomass throughout 
the Bay will be e lusive . Average baywide abundance, estimated via the TIES trawl 
surveys was 260 ton s (S. Jun g, unpubli shed data) . Most likely, thi is an 
407 
underesLi mate or the biomass o f thi s f ish, which is is closely associated with the 
bouom and may noL be we ll-sampled by mid-waler traw ls. T he traw l-based estimate 
is 20% o f the traw l-based e ti male ror spot, f'or which the mainstem biomass was 
estimated to be - 20,000 tons. Accord ingly, the biomass of hogchokers in the 
mainstem Bay may be 4,300 tons. Dovel et al. ( 1969) report that hogchokers spawn in 
hi gher salinity w ater and that the young juveniles migrate Lo lower saliniLies before 
returnin g as they growth toward maturity. Thi s pattern is reflected in the length-
rrequencies reported for different sa linity zones by D errick and K enndy ( 1997). 
Accordingly, one may suspect that summer average biomass is higher in low and high 
salinity zones than in the mesohaline region. Thi s may also be suspected due to the 
relati vely larger area o f unsuitable habitat in the mid Bay (due to hypox ia) and the 
larger macrobenthic biomass in the upper and lower Bay, which constitute the diet o f 
the hogchoker (D errick and K enndy 1997). Pro-rated over the surface areas o f these 
2 
regions, the estimated bi omass in the respective regions is tOO, 50 and 100 mgC m-
(T able A- 13). 
Peters and Boyd ( t 972) estimated the feeding rate o f hogchokers at summer 
1 
temperatures to be about 0.15 g g-1 d-1 in 0- 15 ppt salinity and 0 .2 1 g g-1 d- in when 
salinity w as 30 ppl. Gross growth effi ciency (P/C) increased with salinity from 0.23 
to 0.27. A ccordingly, growth rate increased from about 0.025 d-1 when salinity was 0-
15 ppt to 0.04 d-1 at hi gher salinity. The diet or hogchokers includes polychaetes and 
amphipods, in accordance with their abundance. They also consume the siphons of 
bi val ves, but not whole bi val ves (Derri ck and Kennedy 1997). 
408 
Ameri can Eel 
' 111 1e upper 
American eel (A nqui!la rostrata) was captured in greatest numbers ? ti 
Bay in T IES mid water trawls (S. Jung, unpublished data). Being catadromous, eels 
spawn in the Sargasso Sea and migrate into the Bay in late spring as elvers. Elvers 
mi grate to lower ?alinity areas, includi ng rreshwatcr, oligohaline, and mesohaline 
areas or the Bay (Wenner and Musick 1975, EPA I 99 lb), reaching these areas by their 
first summer in the Bay. Once into the Bay, eels reside for as long as 24 years before 
mi grating back to the open sea to spawn, during which time they do not eat, and from 
which they do not return (EPA I 99 1b ). 
Eels are subject to two fi sheries in the Chesapeake Bay, both of which use 
pots. One fi shery targets smaller eels, >25 cm, fo r use as crab bait. The other ri shery 
ca pt u res th em Ii vc r r human consumption overseas. The Ia  tter fi shery requires Ia rger 
O
eels, >33 cm (EPA J 99 J b). Landings for the bait Fishery have not been wel I 
documented due minimal reporting requirements; however, EPA (I 99 Jb) describe a 
10 
scheme fo r estimating landings as a ratio or crab harvests and reported use of eels as 
bait by crabbers using trotlines (chicken necks are now widely used as well). Based 
1 11 
on these fi 0a u res > the estimated Maryland landings are about 4 55 tons y- . Landi ba s ,.n  
Virginia are better documented due to reporting requirements by the Potomac Ri ver 
Fisheries Commission. Virgini a landings From the Potomac River have averaged 
- I4 0 tons/ (EPA I 99 Jb ), and on that basis combined landings from other rivers 
were estimated a have been a simil ar amount. Thus, total Virginia landings may 
10 
have been go tons /. EPA ( J99 lb) reported that waterman in the Ii ve fi shery have 
2
complained that the average size or eels is decreaSJ ng and that many of the eels that 
409 
C, 
have been captured are undersized. On this basis, is has been assumed that fishino 
mortali ty is high, sufficient to achieve thi s size structure. The fi shing mortality rate 
C, 
was infe rred by modeling growth using estimated growth parameters dcri ved usino 
length-weight and maximum length data reported by Froese and Pau ly (2001 ) . T he 
growth cocllici cnt K was estimated from EPA (I 99 I b) which noted that the annual 
60 
growth i ncrcmcnt for market size eels, estimated using tagging studies, was about 
2 
mm. Accordingly, K was esti mated to be 0.08 / . Assuming natural mortal ity is o.
ba cd on a max imum age of 25 years, growth overfi shing such that most of the 
harvest is of marginal size for the Ji ve eel fishery occurs when F~0.8. The biomass 
O
necessary to su p porl the catch (73 5 tons / ) is l 692 tons. 0 f the tota I surfa ce area f 
the Chesapeake Bay and tidal tri butaries, about 25% was assumed to be low 
mcsohalinc to freshwater and favored habitat for eels. T hus, the eels were computed 
2 1 1 
to have an average densi ry within that habi tal of 35 mgC m ? d. (0.2 gAFDW= g 
WW , 500 mgC= I g AFDW). Only 25% of the mesohaline mainstem Bay was assmed 
5 
to be favored habitat. Theref 0 1~, biomass in the mid Bay was computed to be 8.7
mgC m?' . T he lower Bay was assumed to not have any habitat in summer for eels and 
biomass was set to zero. 
T he average age of eels, weight by biomass was 4 .8 y, for which annual 
average p/ B is 0.0015 d?1? summer average P/B was assumed to be 2- fo ld greater, or 
0.003 d? ? Average values for assimilation efficiency and net growth efficiency 
1
obtained from values reported by Schroeder (I 98 1) for camivorous fish (AE=0.68, 
NG E=0.4 I , Table A- I 3). A t thi s rate of production, annual eel production in the Bay 
is estimated to be 926 rons, of which the fishery removed 80%. During summer, eels 
4 l0 
producti on was estimated to be 467 wn s. Since maximum demand f'or ee ls occurs 
during summer, it wa ? assumed th at 75% or the landings (55 1 tons) occurred during 
summer, or 120% or summer producti on. Therefore, l'i shcri es removals from the 
1
upper Bay were es timated to be O 126 mgC m?' d. , while the same ror the mid Bay 
was 0.032 mgC m? 2 ct? 1. Thus, summe1 main stem landings were 124 tons, o, 22% or 
the annual baywide landings. To balance the excess landings, it was assumed that 
biomass decreased during summer by 0.021 mgC m?' ct?' in the upper Bay and 0.006 
mgC 111 -2 d-1 in the mid Bay. 
Other Fish 
It is impoitant to note here that the upper Bay has been characteri zed here as 
hav ing more J"i sh species than the mid Bay, which has mo1e than the lower Bay. In 
i:,, 
fact, the lower Bay may have a more diverse fi sh assemblage (e.g. TJES data, s. Jun o 
unpublished data). This dicrepancy ,~fleets an artifact of data analysis: to suitabl y 
represent a species in a network model requires adequate knowledge of its biomass, 
diet and predators. If the species assemblage is more diverse, this becomes a more 
arduous if not impossible. For example, fi sh that are not abundant may be pooled as 
"other fi sh" in studies of predator diets. Although it may be poss ible to quantify the 
abundances of this diverse assemblage, this substantia l task has not been undertaken 
here. Rather, the possible ramifications of such an oversight are di scussed in the 
analysis of the now network s. 
411 
Literature Cited 
Baines, S. B. and M. L. Pace. 1991. The production or dissolved organ ic matter by 
phytoplankton and its irnportance to bacteri a: Patterns across marine and 
freshwater systems. Li,1111 01. Ocea11ogr. 36(6): 1078-1090. 
Baird, D. and R. E. Ulanowicz . 1989. The seasonal dynamics of the Chesapeake Bay 
Ecosystem. Ecol. Mo11 og r. 59(4): 329-364. 
Banse, K. and S. Mosher. 1980. Adult body mass and annual production/biomass 
rel ati onships of fi eld populati ons. Eco l. Mo11ogr . 50(3): 355-379. 
Barger, L.E. 1990. Age and growth of bluefish Po11wto11111s sa lt atrix from the northern 
Gu lf of Mexico and U.S. south Atl an ti c coast. hsli . 13111/ ., U.S. 88:805-809. 
Batiuk , R.R. Orth , K. Moore, W. Denni son, J. Stevenson, L. Stave r, V. Carter, N. 
Rybicki, R. Hickman , S. Kollar, S. Bieber, and P. Heasly. l992. Chesapeake 
Bay Submerged Aquati c Vegetati on Habitat Requirements and Restoration 
Targets: A Technical Synthesis. EPA Chesapeake Bay Program , Annapoli s, 
MD. 
Berggreen, U., B. Ha nsen and T. Ki?rboe. 1988. Food size spectra , ingesti on and 
growth or the copepod Acarti a tonsa during development: implicati ons fo r 
determ ination of copepod producti on. Mar. Biol. 99: 341-352 . 
Bogdan, K. G. and J. J. Gi lbert. 1982. Seasonal patterns of feeding by natural 
populations of Keratella, Polyartl1ra , and B0s1ni11a: Clearance rates , 
selectiviti es and contributi ons to community grazin g. Li111no/. Ocea 11 ogr. 27: 
918-934. 
Bogdan , K. G. and J. J . Gilbert. 1984. Body size and food size in freshw ater 
zooplan l ton. Proceedings of the National Acaderny of Sciences (USA) 81: 
6427-6431. 
Boreman , J. 1997 . Methods for comparin g the impacts of pol luti on and fishing on fish 
populati ons. Trans. A111. Fish. Soc. 126: 506-5 13. 
B?rsheim , K. Y. and G. Bratbak. l987. Cell vo lume to cell carbon conversion !"actors 
ror a bacterivorous Monas sp. enriched fro1T1 seawater. Mar Ecol. Prag. Ser. 
36: l71 - l75. 
Boynton, W. R. and W . M . Kemp. 2000. lnfluence of ri ver fl ow and nutri ent loads on 
se lected ecosystem processes . A synthes is of Chesapeake Bay data . pp. 269-
298 Tn : J .E. Hobbie (ed.). Estuarine Science: A Synthetic ARproach to 
Research and Practi ce. Island Press, Wash ington, DC. 
4 12 
Boyn ton, W.R ., R. M. Stankelis , F. M. Roh land, J. D. Hagy and J.M. Frank. 1999. 
Maryland Chesapeake Bay Water Quality Mon itoring Program. Ecosystem 
Processes Component. Level One Interpretive Report No . 16. Chesapeake 
Biological Laboratory. Uni vers ity of Maryland Center fo r Environmental 
Science. So lomons, MD. 20688-0038. Ref". No. lUMCESlCBL 99-0070. 
Boynton, W. R. , personal communi cati on. UMCES Chesapeake Biological 
Laboratory, Solomons, MD. 
Butowsk i, . H. 1994. Che apeake Bay Oyster Fishery Management Plan . 
Agreemen t Commitment Report Revision. US EPA Chesapeake Bay Program, 
Ann apo li s, MD. 
Cebrian , J. 1999. Pattern s in the fate or producti on in plant com muniti es. /\ 111. Nat. 
154(4): 449-468 . 
Chesson, J. 1983. The estimation and analys is of preference and its relati onship to 
foragin g model s. Ecology 64(5): 1297- 1304. 
Choi, J. W., B. F. Sherr, and E. B. Sherr. 1999. Dead or ali ve? A large rraction of 
ETS-inacti ve marine bacterioplankton cell s, as assessed by reducti on of CTC, 
can become ETS-acti ve with incubati on and substrate additi on. Aq11at. 
Microb. Ecol. 18 (2): 105-11 5. 
Ch ri stensen, V. and D. Pauly . 1992. ECO PATH II - a so ftware for balancing steady-
state ecosystem models and calculatin g network characteri sti cs. Ecol. 
Modelling. 6 1: 169- 185. 
Chri stensen, V. and D. Pauly . 1993 . Flow characteri sti cs of aquati c ecosysten1 s, p. 
338-352. In V. Chri stensen and D. Paul y (eds.) Trophic models of aq uatic 
ecosystems. lCLARM Conference Proceedin gs No. 26. 
Chri stensen, V. , C.J. Walters, and D. Pauly. 2000. Ecopath with Ecos im: A user's 
guide . October 2000 editi on. Fisheri es Centre, Univ . of Briti sh Columbi a, 
Vancouver, Canada and Internati onal Center for Livin g and Aquati c Resources 
Management,, Penan g, Malays ia. 
Co le, J. J. S. Findlay, and M. L. Pace. 1988. Bacterial producti on in fresh and 
saltwater ecosystems: a cross-system overvi ew. Mar. Eco l. Prag. Ser. 43: 1-
10. 
Cowan, J. L. and W. R. Boynton. I 996. Sediment-water oxygen and nutri ent 
exchanges along the longitudinal ax is or Chesapeake Bay: Seasonal patterns, 
controlling fac tors and ecological significance. Estuaries 19(3):562-580. 
4 13 
Cronin , W . B. and D . W . Pritchard . J975. Additi onal stati sti cs on the dimensions or 
Chesapeake Bay and its tri butari es: Cross secti on w idths and segment vo lumes 
per meter depth. Chesapeake Bay InstJtute. Reference 75-3. Spec ial Report 
42. 
del G iorgio, P. A. and J. J. Co le. Bac teri al growth effi ciency in natural aquati c 
systems. /\ nn . Rev. Ecol. Sys!. 29: 503-4 1. 
Derri ck, P. A. and V. S. Kenn edy. J997. Prey se lecti on by the hogchoker, T rinectes 
maculatus (Pi sces : Soleidae) , along summer sa linity gradients in Chesapea ke 
Bay, USA. Mar. Biol. J2 9: 699-7 1l . 
Dclwylcr, R. and E. D. Houde. 1970. food seleclion by laboralory-rearer larvae or 
the scal ed sardine Harengula pensacolae ('.i sces: C lupeidae) and the bay 
anchovy /\ nc/1 00 111itc/Ji!li (Pi sces: Engraul1dae). Mar. Biol. 7: 2 14-222 . 
Diaz, R J., R. J. Neubauer, L. C Schaffn er, L. Pihl , and S. P. Baden. 1992 
Continuous monitoring of di sso lved oxygen in an estuary experi encin o 
periodic hypox ia and Lhe effect or hypoxia on macrobenlhos and fi sh. bSci. Tor. 
Env. (Supplement 1992): J05 5- J067. 
Dolan , J. R. I 991 a. Gui Ids of ciliale mrcrozooplankton in Lhe Chesapeake Bay. Esr 
Coast. S/J e(/ Sci. 33: J 37-J 52. 
Dolan, J. R. 199 I b. Ciliale growth raLes and consumpLi on by cope pods. Mar. Biol 
111 : 303-309 . 
Dolan , J . R. and D. W. Coals. 1990. Seasonal abundances of planklonic ciliaLes and 
micrortagellotes in mesohaline Chesapeake Bay walers. Esr. Coasr. Shelf Sci. 
3 1: 157-175. 
Dolan , J. R. and c L. Gallegos. 1991. Trophic coupling of rolifers, microflagellales, 
and bacleria during fail months ,n the Rhode River estuary. Mar. Ecol. Frog. 
Ser. 77 : 147- 156. 
Dovel, w. L., J. A. Mi hursky  and A. J. McErlean. 1969. Life his Lory aspccls of Lhe 
hogchoker, Trinectes maculatus, in the Patuxent River esluary, Maryland. 
C/1 es. Sci. 10(2): 104- 119. 
Duarte, c M. and c L. Chi scano. 1999. Seagrass biomass and produclion: a 
reassessment. Aq11at. Bot. 65: I 59- 174. 
7 
Duck low, H. w. and f . Shi ah . I9 93. Bacleri al producLi on in esLUaries. pp 26 l -2S
In : fo rd , T. E. (ed.). A ualic Microbiolo 0 : An Eco[oaical A roach. 
Blackwell Sc ientifi c Publicati ons, Cambridge, M A. 5 I 8 pp. 
4 14 
Durbin , E. G. and A.G. Durbin. 198 1. Asss imil ati on effi ciency and nitrogen 
excreti on of' a filter-feedin g plankti vore, the Atl antic menhaden, Brevoortia 
tyra111111s (Pi sces: Clupeidae). Fish. 8 11/1. 79(4): 60 1-6 16. 
Eberso le, E. L. and V. S. Kennedy. 1994. Size se lecti on of Atlanti c Rangia clams 
Ra11gia c1llleata, by blue crabs, Calli11ec1es sapidus. Estuaries 17(3):668-673 
EPA 1989. Chesapeake Bay Alosid Management Pl an. US EPA Chesapeake Bay 
Program , Ann apoli s, MD. 
EPA 1990a. Chesapeake Bay Bluefi sh Fishery Man agement Pl an. Chesapeake Ba y 
Program Agreement Commitment Report. US EPA Chesapeake Bay Program , 
Annapoli s, MD. 
EPA 1990b. Chesapeake Bay Weak fi sh and Spotted Scatrout Fishery Management 
Pl an. Chesapeake Bay Program Agreement Commitment Report. Chesapeake 
Bay Program , Annapoli s, MD. 
EPA 199 La. Chesapeake Bay Atlanti c Croaker and Spot Fishery Management Pl an. 
Chesapeake Bay Program , Annapoli s, MD. 
EPA 1991 b. Chesapeake Bay Ameri can Eel Fishery Management Plan . Chesapeake 
Bay Program, Annapoli s, MD. 
EPA 1995. Chesapeake Bay Striped Bass Fi shery Management Pl an. Annual 
Progress Report. !994. EPA Chesapeake Bay Program, Annapoli s, MD. 
E p A I9 97 . J 997 Chesapeake Bay BI ue Crab Fishery Management Pl an . June I 997. 
US EPA Chesapeake Bay Program , Annapoli s, MD. 
EPA 2000. The 2000 User's Guide to Chesa.peake Bay Program Biological and Living 
Resources Monitorin g Data. US Environmental Protecti on Agency, 
Chesapeake Bay Program , Annapoli s, MD. 
EPA 2000b. Chesapeake 2000, A Watershed Partnership. EPA Chesapeake Bay 
Program , Annapoli s, MD. 
9
Fenchel , T. 978. The ecology of micro- and meiobenthos. A1111. Rev. Ecol. Sys/. : 
1
99- 121. 
Fenchel, T. l 982. Ecology of heterotrophic mi croflagel !ates. !I. B iocnergeti cs and 
growth. Mar Ecol. Prog. Ser. 8: 225-23 1. 
Froese, R. and D. Pauly (eds.) 200 I. Fish Base. World Wide Web electronic 
publication. http://www.fishbase.org. October 4, 200 I. 
4 15 
Flach, E., J. Vanaverbeke and C. H eip. 1999. The meiofauna: macrofauna ratio 
across the continental slope of the Goban Spur (north -east Atlanti c). J. Exp. 
Mar. Biol. Ecol. 79(2): 233-241. 
Fri ed land, K . D ., D. W. Ahrenho lz, and J. F. Guthrie. 1989. Influence of plankton on 
distribution patterns of the filter-feeder Brevoortia tyra111111s (Pisces: 
Clupeidae). Mar Ecol. Prog. Ser. 54: J- 11. 
Gacclke, U. and D. Strail e. 1994. Seasona l changes of the quantitative importance or 
protozoan s in a large lake. An ecosystem approach using mass-balanced 
carbon fl ow diagrams. Mar. Microb. Food Webs. 8(1 -2): L63- I 88. 
Gerl ac h, S. A. 197 1. On the importance or marine meiofauna for bcnthic 
comm uniti es. Oecologia. 6: 176- 190. 
G ifford , D. J. and D. A. Caron. 2000. Sampling preservation, enumerati on and 
biomass or marine protozoop lankton. 1n : R. Harris, P. Wiebe, J. Lenz, H. R. 
Rune Skjo ldal and M. Huntley, (eels .) ICES Zooplankton Methodology 
Manual. Academic Press, New York. 
Harding, L. w., Jr. and E. S. Perry. 1997. Long-term increase of phytoplankton 
biomass in Chesapeake Bay, 1950-94. Mar. Ecol. Prog. Ser. 157: 39-52. 
Harding, L. W., Jr. , M. E. Mallonee, and E. S. Perry. 200 J. Toward a predicli ve 
understanding of primary productivity in a temperate, partially-stratifi ed 
estuary . Est. Coast. Shel/ Sci. (in press). 
Hargis, w. J ., Jr. and D. S. Haven. I 994. The precarious stale of the Chesapeake 
pub I ic oyster resource. 1n: Hi I I, P. and S. Nelson, (eds.). Toward a Sustainable 
Watershed: The Chesapeake Experiment. Proceedings of the 1994 
Chesapeake Research Conference. Chesapeake Research Consortium , 
Publication Number 149. 
Hartman K. J. , and S. B. Brandl. I9 95a. Trophic resource parli tioni ng, diets, and 
growth of sympatric estuarine predators. Trans. J\111. Fish. Soc. J2 4 (4): 520-
537. 
Hartman , K. J. ands. B. Brandl. J995b. Comparative energetics and the development 
or bioenergetics models for sympatnc estuarine predators. Can. J. Fish. /\quar. 
Sci. 52: 1647-1666. 
Hartman , K. J. ands. B. Brandl. 1995c. Estimating energy density of fish . Trans. 
A!II. Fisli. Soc. 124: 347-355 . 
Heinbokel, J. F., o. W. Coats, K. W. Hcnserson , and M. A. Tyler. J988. 
Reprodocllon and secondary producllon of three spec,c s of the roli fer genus 
SynchaeJa in the estuarine Potomac RI ver. ./. Plank. Res. IO( 4): 659-674. 
416 
I lcinle , D.R. 1966. Production or a ca lanoid copepod, Acartia tonsa, in the Patuxen t 
River estuary. C/1 es. Sci. 7(2): 59-74. 
1 lines, A. H., A. M. Haddon, and L.A. Wiechert. 1990. Gu ild structure and forag in g 
impact or blue crabs and ep ibenthi c fi sh in a subestuary of Chesapeake Bay. 
Mar. Ecol. Prog. Ser. 67: 105- 126. 
I loeni g, J. 1983. Empiri ca l use of longev ity data Lo estimate morta lity rates . Fis/1. 
Bull. 82( l ): 899-903. 
Hobbs, C. H., Lll , J . P. J-lalka, R. T. Kerhin, and m. J. Carron. 1992. Chesapeake Bay 
Sed iment Budget. ./. Coasr. Res . 8(2): 292-300. 
Holl and , A. F. , A. T. Shaughnessy, and M. H. Hi ege l. 1987. Long-term va riati on in 
mesohaline Chesapeake Bay macrobenthos: Spatia l and temporal patterns. 
Esruaries I 0(3):227-245. 
Holli s, E. H. 1952. Variations in the feeding habits of the striped bass , Roccus 
saxarilis (Walbaum), in Chesapeake Bay. 8111/ . Bing. Ocean. Coll. 14: l 11 -
13 L. 
Houde, E. D. and E. S. Ruthc1fo rcl. 1993. Recent trend in estuarine fi sheri es: 
Predicti on of fi sh production and yield. Esr11 aries : 16(2): 16 1-176 . 
Houde, E. D., the late J .C. Gamb le, S. E. Dorsey, and J. H. Cowan, Jr. 1994. Drifting 
mesocosms: the influence or ge lat inous zooplankton on mortality or bay 
anchovy, Anc/,oa 111 irc/1illi , eggs and yo lk-sac larvae. ICES .I. Mar. Sci. 5 I: 
383-394. 
1l oude, E. D., S. Jukic-Peladic, S. B. Brandt and S. D. Leach. 1999. Fisheries: Trends 
in catches, abundance and management, pp. 34 1-366. In : T. C. Malone, A. 
Mal ej , L. W. Harding, Jr ., N. Smocl laka, and R. E. Turner (eels.) Ecosystems at 
the Land-Sea Margin , Drainage Basin to Coastal Sea. Ameri can Geophysica l 
Un ion, Washington , DC. 
Iverson, R. L. 1990. Contro l or fish production. Li111nol. Oceanogr. 35(7): 1593-
1604. 
Jackson J.B . C., M . X. Kirb y, W. H. Berger, K. A. Bjorndal , L. W. Bots ford , B. J. 
Bourque, R. H. Bradbury, R. Cooke, J. Erlandson, J. A. Estes, T. P. Hughes, S. 
Kidwe ll , C. B. Lange, H. S. Lenih an, J.M . Pandolfi , C. H. Peterson, R. S. 
Stencel<, M. J. Tegner and R. R. Warner. 200 I. Hi stori cal overfi shing and the 
recent co ll apse or coastal ecosystems. Science. 293 (5530) : 629-638. 
Jassby, A. D. and T. Pl att. I 976 . Mathematical formu lation of the relati onship 
between photosynthes is and li ght fo r phytopl ankton. Li111nol. Oceanogr. 2 1: 
540-547. 
4 17 
Johnson, _c., D. Klum pp, J. Field, and R. Bradbury. Carbon nux on coral reers: 123-
cf lects or large shifts in com muni ty structure. Mar. Ecol. Prog. Ser. 176? 
]43. - . 
Jordan , S. J. , G. E. Kra nt, and C. Judy. 1994. M~1rylan? Chesapeake Bay oyster 
recovery areas: Summary of technical 1n fo rm at1 on. m: Hill , P. ands . Nelson 
(eds.). Tow a rd a Sus tamable Watershed: The C hcsa peake  Ex peri menL  ' 
Proceedings of the 1994 Chesapeake Research Conference. Chesapeake 
Research Consortium , Publi cati on Number 149. 
Ju , S. -J ., 0. H. Secor, and H. R. Harvey. 1999 . Use of ex tractible lipofusc in fo r8 a 
dctcrmi nation of blue crab Calli11ec1es sapid11s. Mar. Ecol. Prog. Ser. J lc 
17 1- 179. 
Jung, S. and E. D. Houde. 2000. Scale and paocrn in recruitment processes or bay 
anchovy 111 Chesapeake Bay. ICES CM2000/M: 13. 
1 
Jung, S ., unpubli shed data . Pre Ii mi nary analys is and presentation of mid-water traw 
data from the Chesapeake Bay LMER (TfES ) project. 
Karl sen, A. W. , T. M. Cronin , S: E. Ishman, D. A. Willard , C. W. Holmes, M. Marol, 
and R. Kerhin. 2~00. H1 stoncal trendsm Chesapeake Bay di ssolved oxy8gen 
based on bent h ic  f oram rm fern from sed rm enl  cores. Es/ ua  ri es. 23 (4 ): 48 _ 
508. 
Ka y, J . J., L. A. Graham, and R. E. Ulanowicz. J989. A detailed guide to network 
anal ys is. 11r Wulff, F. , J. G . Field , and K.H? Mann: (eds.). Netwo rk Anal ys is 
in Marine Ecolo . Methods and A J lrcatr ons. Springer-Verlag, New York . 
3 
Kemp, w.M . , W.R. Boynton, R.R. Twilley, J.C. Stevenson, and J.C. Means. J 93
The decline or submerged vascular plants in Chesapeake Bay: summary ol 
result s concerning possible causes. Mar. Teclr. Soc. I 17(2):78-89. 
4
Kemp, w. M., w. R. Boynton, R.R. Twilley, J.C. Stevenson, and L. G. Ward. l9S . 
Influences or submersed vascular plants on ecological processes in upper 
Chesapeake Bay. Jn : V. S. Kennedy (ed). The Estuary us a Filter Academic 
Press New York , NY . 
' 
Kemp, w. M., E. M. Smith , M. Marvin-DiPasquale, and W.R. Boynton. J997. 
Organic carbon balance and net ecosystem metaboli sm in Chesapeake Bay. 
229 248
Mar. Ecol. Prog. Ser. 150: - ? 
Kemp, w. M., J. faganeli , S. Puskari s, E. M. Smith .and W. R. Boynton. J999. 
Pelao ic-Benthic coupling and Nutrient Cycline. In : Malone, T. c. , L. MalcJ 
L. vJ'. Harding, Jr , N. smodlaka, and R. E. Turner (eds.). Ecosys tems at the ' 
~? Ameri can Geophysical Unron, Washington, DC 
41 8 
Ki0rboc, T. and T. G. Nielsen. 1994. Regulation of zoop lankton biomass and 
production in a temperate, coastal ecosystem. I. Copepods. Lininol. 
Oceanogr. 39(3): 493-507 . 
Ki?rboc, T. , F. M?hlenbcrg and K Hamburger. 1985. B ioenergeti cs of the 
planktonic copepod Acart1 a tonsa: re lation between feeding, egg production 
and respiration, and compos1 uon of spec,lic dynamic action. Mar. Ecol. Pro, 
Ser. 26: 85-97. S 
Ki mmcrcr, W. J. t 987. The theory of secondary production calcul ations fo r 
continuously reproducing populati ons. Li11111 0 /. Oceanogr. 32(1 ) : I - 13. 
Kimura R, D. H. Secor, E. D. Houde ED and P. M . Picco li . 2000. Up-estuary 
di spersal of young-of-the-year bay anchovy Anchoa m, tchi 11 i in the 
Chesapeake Bay: infe rcnces from microprobe analys is of strontium in otol it h s. 
Mar. Ecol. Prag . Ser. 208: 217-227 
Klesbasco, M. J. 199 [ . Feeding ecology and daily ratio of bay anchovy, Anchoa 
mitchilli , in the mid-Chesapeake Bay. M. S. Thesis. University of Maryland 
Co llege Park. Co llege Park, MD. 
Koch, E. W. 2001. Beyond Light: Physical, Geological, and Geochemical Parameters 
as Possible Submersed Aquatic Vegetation Habitat Requirements. Estuaries. 
24(1): 1- 17. 
Kremer, P. and M. R. Reeve. J989. Growth dynamics of a ctenophore (Mnemi opsis) 
in relat ion to vari able food supply. IL Carbon budgets and growth model. J 
Plank. Res. I J (3): 553-574. 
Langland, M.J., R. E. Edwards, L.A. Sprague and. S. E. Yochum. 200 J. Summary of 
T rends and Status Analys is for Flow, Nu tnents, and Sediments at Selected 
Nontidal Sites, Chesapeake Bay Basin , 1985-99: U.S. Geological Survey Open 
File Report 01 -73, 49 P? 
Larson, R. J. J 986. The feeding and growth of the sea nettle, Chrysaora 
quiuquecirrha (DeSor), in the laborato ry. Estuaries. 9: 376-379. 
Laughlin , R. A. 1982. feeding hab its of the blue crab, Calli11ectes sapidus Rathbun, 
in the Apalachicola estuary, Flonda. Bull. Mar. SCI. 32(4): 807-822 
Linley, E. A. s., R. c. Newell , and M.f. Lucas. !983. Quanntative relationshi ps 
between phytoplankton, bactena and heterotroph1c m1 croflagellates in shelf 
12 77 89
waters. Mar. Ecol. Prag. Ser. : - ? 
Loftus, M. E., D. V. Sobba Rao, and H. H. Sc liger. !972. Growth and di ss ipation of 
phytoplankton in Chesapeake Bay J. Response to a large pulse of rainfall. 
Ches. Sci. 13(4):282-299. 
4 19 
Luo , J. and S. B. Brandl. J 993. Bay anchovy Anchoa 111itc/1il/i production and 
consumption in mid-oesapeake Bay based on a bioenergetics model and 
acousltc measures off 1sh abundance. Mar. Ecol. Prag. Ser. 98: 223-236. 
Luo, J ., S. B. Brandt, and M. J. Kl ebaseo. I9 96. Vi 11ual reality of plan kti vores: A2 7 1 
~';~ 's perspective of prey size se lect ion. Mar. Ecol. Prog. Ser. J 40 ( J -3): _ 
Luo, J., K. J. Hartman, S. B. Brandt, C. F. Cereo, and T. H. Rippetoe. 200 1. A 
spatial! y-cx pl ici t approach for cst_i mating carrying capaci ty: An application for 
the Atlantic Menhaden (Brevoortia tyrannus) 111 Chesapeake Bay. Estuaries. 
24( 4 ): 545-556. 
Mac Intyre , H. L., R. J. Gcider, and D. C. Mil ler. 1996. Microphytoben thos: The 
ecological role of the "Secret Garden" of unvegetated, sha llow-water marine 
habitats. [. Distribution, abundance, and primary production. Estuaries 
I 9(2A): J 86-20 l. 
Ma lone, T. c., H. w. Ducklow, E. R. Peele, and S. E. Pike. 1991. Picoplankton 
carbon flux in Chesapeake Bay. Mar. Ecol. Prag. Ser. 78: J l -22. 
Mansueti, R. J. I9 61. Age, growth and movements of the striped bass, Roccus 
saxal ii is, taken in size se lective gear. C!,es. Sci. 2(]-2):9-36. 
Mansueti, R. J. I9 61. J96 l. Movements, reproduction, and mortality of the white 3
perch, Roccus a//lericaaus, in the Patuxent estuary, Maryland. Ches. Sci. 2( -
4 ): l 42-205. 
Marvin-Dipasq uale, M. c. and D. G. Capone. 1998. Benthic sulfa te reduction along 
central chan nel. I. Spaual trends and contro ls Mar 
,e Cl, esapea ke  Bay ? . ti 28
Ecol. Prag. Ser. 168: 2 I 3-2 ? 
Mi ller, D. C., R. J. Geider, and H. L. Maclntte: J996. Microphytobenthos: The 
. f tlie "Secret Garden of unveget<1ted , sht1 ll ow-water marme 
eco l.o  g1ca 1IT r o 1Re   oI  ? sect,?ment stabi.li.t y and shallow-water food webs 
ha  b,tats . . . oet 
11 
12 . 
Estuaries. J9 (2A): 202-2 ? 
? 1, S lk . a,,d R L Lippson . 1975 . Composit ion and seasonal 
1 er, R. E., S. D. ,. u..  11b1 ,l  crab?  C? a ll1. nectes sap1. dus Rathbun , 1. 11 t.h e Chesapeake M ab undance o, t1 , e ? ue dA? d' J.a centWa?t e1.s . C/ ,es..  5c .t ?. . J6 ( J). 27-31. 
and Delaware Ca n.a 
1 
 a_n  200 1. Production of bacterial substrate by marine 
o il er ' E?  ?F ?  and T. G. ..  Nief 1 s en.y to[J lankton b1? on1ass and ce ll si?z e. J. Plank. Res. 
M I1  
cope pods: Ell ect o P 
. 23(5) : 527-536. E UI' ,owicz J 997. Comparative ecosystem troph ic structure 
onaco , M . E. aUn dSR  . . .d?  Atdli1 ntic es?t uan?e s. Mar. ?,c oI.P  rag. Se r. 16 1:  239-254 
M
of three . . 111 1 - ? 
420 
MonLagna, P. A., G. F. Blanchard, and A. Dinel. J9 95 . Effect of producion and 
bro mass of inLerLidal microphytobenLhos on meiofa unal grazing rales. J. Ex, 
Mar. Biol. Ecol. J 85(2): I 49- I 65. I ? 
Namczr.e, D. A. [99 l. A mmonium exereLion by gelatinous zooplankton and ils role 
,n the regeneraLion of niLrogen rn Chesapeake Bay. M. S. Thesis, UnivcrsiLy of 
Maryland College Park, Maryland. 
Newell , R. L. E. J988. Ecological changes in the Chesapeake Bay: Are they Lhe 
resull of ovcrharvesLrng Lhe Amenean oysLer, Crassostrea virgiJJia. pp. 536-
546 In: M . P. Lynch and E. C. Krome (eds.). U nderswnding the EsLuary 
Advances in Chesapeake Bay Research. Proceedings of a Conference 2 9. 3 J 
March t 988. BalLimore, MD. Chesapeake Research Consortium, PublicaLion 
Number 129. 
Nixon, S. W. I 988. Physica l energy inpuLs and the comparalive ecology of lake and 
marine eeos ys,ems LiJJJJJOI. OceaJJ gr. 33( 4, Part 2): 1005-1025. 
North , E.W. 200 1. Transporl and rcLention of fish early-life sLages in Chesapeake 
Bay: mechanisms and implications for recruitment. Ph.D. DisserLation 
University of Maryland, College Park. 306 p. 
OrL h, R. J. and K Moore. J9  84. Di stri bu tr on and ab u ndanee or sub merged aqua Li c 
vegeLaLioo in Chesapeake Bay. EstJJaries 7(4B): 53 L-540. 
Orth , R. J. and K. A. Moore. !986. Seasonal and year-to-year variaLioos in the 
growLh of zoslera marina (L.) (Eelgrass) in the lower Chesapeake Bay. Aquat. 
not. 24: 225-341. 
Orth, R . J ., J. f. Nowak, D. J. Wi [cox, J. R. WhiLi ng and L. S. Nagey. J9 97. 
Distribution of submerged aquatic vegetaLion in the Chesapeake Bay and 
tribuLaries and Lhe coastal bays- J996. US EPA Chesapeake Bay Program, 
Annapolis, MD. 
Pacheco, A. L. 1962. Age and growlh of spot in lower Chesapeake Bay, wilh noLes 
on disLribuLion and abundance of Juveniles in Lhe York River sysLem. Ches. Sci. 
3( 1):18-28. 
PcLers, D . s. and M. T. Boyd. [972. The effect of temperature, sa liniLy, and 
availabi Ii LY of food on tl1e feeding and growlh of Lhe hogchoker, Trinectes 
macJ1iatJ1S (B tock & Schneider). J. Exp. Mar. Bro. Ecol. 7: 20 1-207 
Poe, T. P., ff. Hansel, s. Vigg, D. E. Palmer, and_L A. Predcrgasl. Feeding of 
predaceous fishes on out-migr:11ng1uvenrle sa lmonrds rn John Day Reservoir, 
Columbia River. Tro JJ s. A//1. f,rsh. Soc. 120(4): 405-420 
Po lov i na, J. J. J 984. An overv iew of Lhe ECOP A Tff model. Fish byte 2(2): 5. 7. 
421 
Prager, M.- H. 1996. A simple model of the blue crab, Callinec tes sapidus, spawnino 
mi grat ion in Chesapeake Bay. B. Mar. Sci. 58(2): 42 1-428. " 
Pu ree l I, J. E. l 992. E ffects of predati on by the scyphomeclusan Chrysaora 
quinquecirr/,a on zooplankton popu lations in Chesapeake B,;y, USA. Mar. 
Ecol. Prog . Ser. 87: 65-76. 
Purce II , J E. I 997. Pelagic cnidarians and ctenophores as predators: Selec ti ve 
predation, feeding rates, and el'fects on prey populations. Auna/es de / /nslill// 
Oceanograpliique 73(2): J 25- 137. 
Purcel l, J.E., J. R. White, and M. R. Roman. 1994. Predat ion by gelatinous 11
zooplankton and resource limi tation as potential controls of Acarria ro sa 
copcpod populations in Chesapeake Bay. Li11u10l. Ocea11ogr. 39(2): 263-278. 
Purce ll , J. E. and J. H. Cowan, Jr. 1995. Predation by the scyphomedusan Clirysao9ra 
qninquicirrha on Mn ei11iopsis /eidyi ctenophores. Mar. Ecol. Prog. Ser. 12 : 
63-70. 
Purce ll , JE and Nemazie, DA. 1992. Quantitati ve feeding ecology of the 5
hydromedusan Ne111opsis bacliei in Chesapeake Bay. Mar. Biol. 11 3(2): 30 . 
3 I l. 
Purce ll , JE, White, JR , Nemazie, DA and Wright, DA. 1999. Temperature, sa linity 
and food effects on asexual reproducuon and abundance of the scyphozoan 
Chrysaara quinquecirrha. Mar. Ecol. Prog. Ser. 180: 187- 196 
Purce ll , J.E., T. A. Shiganova, M. B. Decker, and E. D. Houde. 2001. The 
ctenophore M e iopsis in nati ve and exotic hab itats: U. S. estuaries versus 
11 111
the Black Sea bas in . Hydrobiol. , in press. 
Purcell , J.E. personal communication. UMCES Horn Point Laboratory, Cambridge, 
MD. 
Raymond, P.A. and J. E. Bauer. 2000. Bacterial consumption ol' DOC during 
transport through a temperate estuary. Mar Ecol. Prog. Ser. 22: 1-12 
Reeve, M. R. , M. A. s yms, and P. Kremer. 1989. Growth dynamics of a ctenophore 
(Mnerniopsis) in relation to vari.able food suppl y. I. Carbon biomass, fc eding5, 
egg producuon, growth and assin11 latl on etl 1c1enc y. J. Plank. Res . J l (3): 53 . 
552. 
Ricciardi, A. and E. Bourget. J 998. Weight-to-weight conv ersion factors fo r marine 
bcnthic macroinvertibratcs. Mar. Ecol. Prag. Ser. 163: 245-25 1. 
422 
Rilling, G:C . and E. D. Houde. 1999a. Regional and temporal varia1bil ity in 
d,stn bu. tion an .I  abundance ot bay anchovy (A 11choa 111itchill ) eaoaos  , I" a v. a, e and 
a u t biomass 1n the Chesapeake Bay. Estuaries. 22(4): 1096-l !09. 
d I 
Rilling, G. C. and E. D. Houde J999b. Regional and temporal vari abi lity in arowth 
and mortality of bay anchovy, Auchoa mitchiili, larvae in Chesapeak; B?i 
Fish. Bull. 97: 555-569. 'y. 
Rizzo, W. M. and R_. L. Wetzel. I 985. l nterti da l and _shoal benthic community 
metabolism 111 a temperate estuary: Studies of spatial and temporal scales of 
vari ability. Estuaries. 8(4): 342-35 1. 
Roman, M. R., A. L. Gauzens, W. K. Rhinehart, and J. R. White. 1993. Effects of 
low oxygen waters on Chesapeake Bay zooplankton. Li1111wl. Oceanogr 
38(8): 1603- 161 4. 
Roman, M. R., D. V. Holliday and L. P. Sanford. 20_0 1. Temporal and spatial patterns 
ol zooplankton in the Chesapeake Bay turb1d1ty max imum . Mar. Ecol. Prag. 
Ser. 2 L3: 215-227 
Ross, S. W. I 988. Age, growth , and Mortality of Atlantic croaker in north Carolina 
with Comments on Population Dynamics. I 988. Trans. Am. Fish. Soc. Jl / 
46 1-473. 
Rugolo, L. J., K. s. Knotts, A. M. Lange, and V. A. Crecco. !998. Stock assessment 
of Chesapeake Bay blue crab (Cailin ectes sapidus Rathbun). J. Shellfish Res. 
17(2): 493-517 . 
Sadzinski, R., r. Piavis, and A. Jarzynski. !999. Stock Assessment of Aelected Adult 
Resident Migratory Pinfish in Maryland 's Chesapeake Bay. US Fish and 
Wildlife Service, federal Aid Project Annual Report. MD Department of 
Natural Resources, Annapolis , MD-
Sagasti , A., L. c. Schaffner and J. E. Duffy. 2000. Epifounal communities thri ve in 
an estuary with hypoxic episodes. Estuaries. 23(4): 474-487. 
San ford , L. p. I 994. Wave-forced resuspension of upper Chesapeake Bay muds 
Estuaries 17(1B):J48-J65. 
Schroeder, L. A. consumer growth efficiencies: Their8 2li8m i ts and relationships to 93 805
ecological energeti cs. J. Tlieor. Biol. : -
Schubcl, J. R. and H. H. carter. J977. Suspended sediment b_ud gct for Chesapeake 
Bay. n: M. Wiley, ed. Es tuarine Processes. Vol I I. C, rcu I at1on Scdi ment s 
and Tr1a nsfer of Materi al in the Es tuar . Academic Press, New Yo rk . 
423 
Shiah , F. -K. and I-/. W. Ducklow. 1994. Temperature regulntion or heterotrophic 
b:.ic terioplankton abundance, production and speci[ic growth rate in 
Chesapeake Bay. li111110!. Oceanogr. 39(6): 1243- 1258. 
Smith , E. M. l 998. Coherence of microbia l respiration rate :md ce fl -speci ric bacteria l 
acti vity in a coastnl pl:.inktonic community. l1q11a1. Microb. Ecol. 16 (1): 27-35 
Smith , E .M. 2000. FacLors regulnting respiration and its coupling to primwy 
produclion in a coasial planktonic community. Ph.D. Dissertation. Uni versity 
or Mnrylancl nL College Park. College Park, MD. 
Smith , E. M . and W. M . Kemp. 2001. Size structure and the production/respiration 
bahmce ina coastal plankton community. li111nol. Ocea11ogr. 46(3): 473-485. 
Smith , G. F. 2001 . A characterization or oyster habital in mesohaline Chesapeake 
Bay. Ph .D. Dissertation. Uni versity or Maryland al Co flege Park , Co fl ege 
Park, MD. 
Smol, N. , K. A. Williams, J. C. R. Govaere, and A. 1. 1. Sndce. 1994. Composition, 
clistribution nnd biomass or meiobenthos in the Oosterschelde estuary (S W 
Ne therlands). Hydrobiol. 283: l 97-2 17. 
Stem berger, R . S. and J. 1. Gilbert. Body size, food concentration, and poopulation 
growth in pfankLonic rotif'ers. Ecology. 66(4): 1 I 51-11 59. 
Stokes, C. 1996. Jn[/uences of environmental variables on pholosynthetic and 
respiraLory quotients in Chesapeake Bay. M.S. Thesis. Uni versity of 
Maryland at College Park. College Park, MD. 
Sundbiick, K. , P. Nilsson, C. Nilsson, and B. Jonsson. 1996. Balance between 
aulolrophic and heterotrophic components and processes in microbenthic 
communities of sandy sediments: A field study. Est. Coast Sl7e(/Sci. 43: 
689-706. 
S t. Pierre, R. A. and J. Davis. 1972. Age, growth , and mortalit y of the while perch, 
Marone americana, in the James and York Ri vers, Virginia . Cl,es. Sci. 13(4): 
272-281. 
Szyrmer, J. and R. E. Ulanowicz. 1987. Tootal nows in ecosystems. Ecol. Modeling. 
53: 123-136. 
Ta ng, E. P. Y. and R. H. Peters. l 995. The allometry or algal respiration. J P!a11k 
Res. 17(2): 303-3 l 5. 
424 
FE 
1 arnowski, M. L. , M. Homer, and L. Bayli s. I 994. The effects of the I 993 freshet on 
Chesapeake Bay oyster populations. ln : Hill , P. and S. Nelson, (eds.). Toward 
a Sustainable Watershed: The Chesapeake Experiment. Proceedings of the 
1994 Chesapeake Research Conference. Chesapeake Research Consortium, 
Publication Number 149. 
Tucker, J . W ., Jr. 1989. Energy utili zati on in bay anchovy, Anchoa 111itc/1i/li, and 
black sea bass, Centropages stria ta striata, eggs and larvae. 1-' is/1. 8111/ . U.S. 
78: 279-293. 
Ulanow icz, R. E. 1986. Growth and Development : Ecosystems Phenomenology. 
Springer-Verlag. 
Ulanowicz, R. E. I 995. Ecosystem trophic roundations: Lindeman exonerata. pp. 549-
560 in B.C. Patten and S. Jorgensen (eds.) Complex Ecology: The Part-Whole 
Relati on in Ecosystems. Prentice-Hall , Englewood Cliffs, New Jersey. 
Ulanowicz, R. E. 1997. Ecology, The Ascendent Perspective. Columbia University 
Press, New York. 
Ulanowicz, R. E. 1999. NETWRK 4.2a: A Package of Computer Al gorithms to 
Ana lyze Ecological Flow Networks. University of Maryland Center for 
Environmental Science. Chesapeake Biological Laboratory. UMCES Ref No. 
82- 7 CBL. 
Ulanowicz, R.E. and W.M. Kemp. 1979. Toward canonical trophic aggregati ons. 
A111er. Nm. 114:87 1-883. 
Ulanowicz, R. E. and D. Baird. 1986. A network analysis of the Chesapeake Bay 
ecosystem. Maryland Department of Natural Resources, Annapo li s, MD. 
Un iversity of Maryland Center for Environmental Science. REF [UMCEES] 
CBL 86-79. 
Ulanowicz, R. E. and C. J. Puccia. 1990. Mi xed trophic impacts in ecosystems. 
Coenoses 5(1): 7-16. 
Vaughn, D.S ., J. W. Smith, E. H. Wi lli ams, and M. H. Prager. 2001b. Analysis on 
the status of the Atlantic menhaden stock. NOAA Center for Cooastal 
Fisheri e and Habitat Research, Beaufort, NC. Report to the Atlantic States 
Marine Fisheri es Commission Menhaden Technical Committee. June 7, 200 l . 
Verity, P. G. 1985. Grazing, respiration, excretion and growth rates of tintinnids. 
Li11111ol. Oceanogr. 30(6): 1268-1282. 
Verity, P. G. 1991. Measurement and simulation of prey uptake by mari ne planktonic 
ciliates fed plastidic and aplastidic nanoplankton. Li111110/. Oceanogr. 36(4): 
729-750. 
425 
- . . .... . -~ ..  . ..  . . ..  . ., .. ' ' ~ ' ... .... . -
Vi rnste in , R. W . 1977. T he i1T1porlance o r predat ion by crabs and fishes on benthic 
infauna in Chesapeake Bay. Ecology . 58: 1199- 12 17. 
Walters, C. 2000. N atural selecti on fo r predati on avo idance tacti cs: impli cations fo r 
marine popul ation and community dynamics . Ma r. Ecol. Prag. Se r. 208: 309-
3 13. 
Wang, S. -B. and E. D . Houde . 1994. Energy storage and dynamics in bay anchovy 
J\ 11 c/10a 111i1 chilli . Ma r. Biol. 12 1: 2 19-227. 
Wan g, S.-B. and E. D . Houde. 1995. Di stri buti on, relati ve abundance, biomass and 
producti on of bay anchovy A11 c/1oa 111i1 c/1illi in the Chesapeake Bay . Mar. 
Ecol. Prag. Se r. 12 1: 27-38. 
Webb, E. l\l , and E. Z lokov itz . 1999. Assessment o r the 1998 recreational fin fish 
harvests in the Choptank and Chester Ri vers and summer head boat surveys . 
US Fish and Wildlife Service, Federal Aid Project Annual Report. MD 
Department of N atural Resources, Ann apoli s, MD. 
Wenner, C . A. and J . A . Musick. 1975. Food habits and seasonal abundance of the 
Ameri can eel, Anguilla roslrala , from the lower Chesapeake Bay. Cl1 es. Sc i. 
16( 1): 62-66. 
Wetzel, R . L. and P . A. Penh a\e . 1983. Producti on ecology of seagrass communities 
in the lower C hesapeake B ay. Mar. Tec/1. Soc . J. 17 (2): 22-3 I . 
Widbom, B . 1984. Determination of average individual dry weights and ash-free dry 
weights in different s ieve fractions of marine meiofaun a. Mar. Biol. 84: 101 -
108. 
White, ]. R . and M . R . R oman. 1992 . Egg production by the ca\anoid copcpod 
Acartia tonsa in the mesoha\ine Chesapeake Bay: the importance of food 
resources and te rnperature . Mar. Ecol. Prag. Se r. 86: 239-249. 
Zaitsev, Y . P . Recent chan ges in the trophic structure of the Black Sea. Fisli. 
Ocea 11 gr. 1(2): 180-189. 
426 
Table A- I . Storage and physical input and export of detrita/ dissolved organ ic carbon 
(DOC) and detntal parlicu/ate orga 111 c carbon (POC). Tola/ POC and DOC are 
1nc/1 cated in parentheses. Storages ha ve units mgC m-2 while lransport flu xes 1, 
un!? ls mg Cm-?- 1er I . dVe 
DOC POC 
Region Slorage Import Ex port Storage Import Export 
Upper Bay 12180 263 286 2933 90 0 
( 12504) (5249) 
Mid Bay 27000 79 85 5130 0 0 
(28207) ( 10324) 
Lower Bay 26286 93 220 5156 0 0 
(269 15) (8309) 
Table A-2. Es timales of summer average primary production rates derived from 
14
Harding et al. (200 /) and Kemp et al. (1997). (A) Net C-PP was obtai nee/ di reeled 
f'rom Table 2 in Harding et al. (2001), averaging regions as follows: Upper 
Bay=Reoion 6? Mid Bay=Average of Region 3-5.; Lower Bay=Average of Reoion l 
b ' 
and 2. (8) Gross 14
14 b 
C-PP was computed as (1.48/ 1.2)( l.229)(Net C-PP). (C) 
Summer average daily gross Oz-PP was com~uled from data in Kemp el al. (1997) 
assuming PQ=l .2 (Stokes !996). (0) Euphotic zone integrated phytoplankton 
biomass was computed from chi-a oblained from Harding et a/. (2001) assuming 
C:Chl -a=50. 
Region (A) (B) (C) (D) 
Net 14C-PP oross ,-iC-PP gross 0 2-PP b PP Biomass 
2 1
(mg C m-2 er') (mg C m- d- ) (mg C 111 -
2 d-1) mg C 111 -2 
Upper Bay 811 1230 570 1595 
Mid Bay 2176 3298 3030 39 /3 
Lower Bay 1406 2131 3070 2485 
427 
Tab le A-3 . Summer phytoplankton biomass, gross phytop lankton production (GPP), 
algal resp irat ion (R), ex trace llul ar release of DOC (ER), and net POC production by 
phyto pl ankton (PP). 
Biomass GPP R ER pp 
(mg C m-2) (mg C IT1-2 d-
1) (mg C m-2 d- 1) (mg C 2 111 - d-1) (mg C 2 111 - d-1) 
Net Plankton 
UB 1356 984 27 1 71 642 
MB 3326 2638 665 197 1776 
LB 2 11 2 1705 422 128 11 54 
Picoplankton 
UB 239 246 60 L9 168 
MB 587 660 147 51 462 
LB 373 426 93 33 300 
Total 
UB 1595 1230 33 1 90 809 
MB 3913 3298 812 249 2237 
LB 2485 2131 516 162 1454 
428 
Table A-4. Biomass, gross primary producti on, algal respiration, ex tracellular DOC 
release and net production by microphytobenthos in three regions or Chesapeake Bay 
clunng summer. 
Region Biomass GPP R ER PP 
(m~ C m-2 (mg C _? er 1) Ill - ) (mg C 111 -2 d-1) (mg C m-2 d-1) 
UB (mg C m-
2 er') 
293 234 59 16 
MB 265 212 159 53 15 
LB 293 234 144 59 16 159 
Tab le A-5. Coverage of submersed aquatic vegetation (SA V) in Chesapeake Bay by 
region . % Veg. = percent of total area vegeta ted, B-VA=Average biomass in vegetated 
areas , P-y A=average production in vege tated areas, Biomass=regional average 
biomass, Procluction=regional average production. 
Region SAV A rea B-VA P-VA Biomass 
2 Production 
(ha) %Veg. (mgC m- (mgC 2 ) 111 - er 1) (m~C m-2) (m~C m-2 c1 -1) 
UB 2495 3.86 54000 406 2086 16 
MB 3136 1.45 135000 1260 1952 18 
LB 3967 1.47 135000 1260 1986 19 
429 
Table A-6. Biomass, produclion, consumption and rcspiralion of free- li ving and 
panicle allached bucteria in Chesapeake Bay during summer. Bacleri al growth 
effi ciency: UB=0.43, MB=0.52, LB=0.39 (de/ Giorgio and Cole 1998) 
Region Biomass Production Consumption 
C Respiration (m~C m-2) (m~ 111 _-J d-1) (m~C 111 -2 er') (111gC 111 -2 d-1) 
Free Bacteria 
Upper Bay 649 349 8 l 1 462 
Mic/ Bay 241 5 1417 2725 1308 
Lower Bax: 1258 740 1897 11 57 
Panicle-Atli:1ched Bac teria 
Upper Bay 36 28 65 37 
Mid Bay 73 83 159 77 
Lower Bax: 39 41 105 64 
430 
Table A ? 7. Est imates of biomass (B), production (P), res pi ration (C), consumption 
(C) and and egestion plus excret ion (U) fo r the zooplankton in the upper, mid and 2 01 
\owcr Bay. Biomass has units mgC m?'. The rates (P, R, C, U) have uni ts mgC ? d 
p R C u B 
66 285 ] ] 9 
Upper Ba~ 30.2 [00 330 68 
Heterotrophi c M icrofl agell atcs J65 97 65.9 [3.70 5.48 
Cili ates 13.7 4. 11 
4. 1 l 
1.62 5.4 l 2. 16 
Roti f ers 3.2 t. 62 
35 256 11 5 
Merop lankton 282 106 
1. 30 5.80 2.20 
Mesozoop lankton J7 .0 2.30 
Ctenophores 0.00 
Sea Nettl es 
279 J L94 498 
M id Ba~ ----149.3 4 18 733 151 
Heterotrophi c M icro fl agc ll atcs ]47.0 366 
215 
7.02 23.39 9 .36 
C i Ii ates 23.4 7.02 9.0 1 30.02 12.0 1 
Roti f crs J 8.0 9.01 88 637 287 
Meroplankton 526 263 J0.00 42.00 16.00 
Mcsozoopl ankton ]26 16.00 0.30 3.90 
Ctenophorcs 6.29 0. LO 
Sea Nettl es 
Lower Ba ---- 95 407 169 43.-l 142 127 434 89 
Heterotrophi c M icrofl agc ll ates 86.7 2 17 0.00 0.00 0.00 o.o 0.00 Cili ates 
]9.20 l 9.20 64.0 1 
25.60 
Roti fers 38.4 
268 89 
650 293 
M eropl ankton 1073 36.00 14.00 
108 14.00 
8.00 
Mesozoopl ankton 
Ctenophores o.oo 
Sea Nettl es 
43 1 
Table A-8. Biomass, consumption, respiration and egeslion for 
; 0 Pepod nauplii and 111esozooplankton. Bio111ass /ms units mgC m-
-. The rates have unil s mgC 2 111 - d-1. P=production, R=respiralion, 
C==consumpti on, U=egest ion. 
Region Nauplii Mesa- p R C u 
Biomass zooplank ton 
Biomass 
UB 31 25 1 106 35 256 11 5 
MB 83 443 263 88 637 287 
LB 107 966 ?68 89 650 293 
Table A-9. Geometric mean biovo/u111es (111/ 111 -2) of the ctenophore 
M11e111iopsis leidyi fo r summer in three regions of Chesapeake Bay. 
Unpub li shed TIES data courtesy of J.E. Purcell. Bio111ass as carbon (mg m-2>, 
in parentheses, was computed using 1 ml= 1.0 12 111g wet weight= 0.485 
mgC (Nemazie 199 1) . 
Year Upper-Bay Mid-Bay Lower-Bay 
1995 26 ( 13) 38 (20) 202 (104) 
1996 10 (5) 72 1 (372) 585 (302) 
1997 93 (48) 167 (86) 90 (46) 
2000 156 (80) 226 (l 17) 9 (4.6) 
T l ES Average 71 (36) 288 (149) 223 ( I 15) 
CBMP A vernue 0 (0) 240 (124) 
Best Estimate 36 (18) 264 ( /26) 223 (115) 
432 
Ta?/c A- / 0. Geometric mean biovo /umes (ml m-2) of C/11 ysaora 
Cf1_"'1CJl(ec i rrlw for summer in 1hree regions of Chesapeake Bay. 
Biom<.1ss (mg C m-2) is indicaled in parentheses. Unpub lished data 
cou ri csy of J.E. Purce /I. Biovo /u111e to ca rbon conversions were made 
using l ml = 1.75 mgC (Purce// 1992; Purce// , J. E., persona/ 
com munica lion). 
Year vn MB LB 
1995 0.48 (0.84) 8.40 ( 14.7) 0.14 (0.25) 
1996 0. JO (0. 18) 0.89 (1 .56) 0. 33 (0.58) 
1997 0.67 (1 .17) 5.07 (8 .87) 0.00 (0.00) 
2000 0.00 (0. 00) 0.01 (0.02) 0.00 (0.00) 
TIES A vcraoc 0.3 1 (0.55) 3.59 (6 .29) 0. 12 (0.00) 
Table A- 1 l. Active sedimenl carbon pool size, and rates of sediment bacteria/ 
mc1abolism. Sedi111en1 oxygen consumplion (SOC) was es limated by Cowan and 
Boynl on ( l 996) for slali ons in the channel of each region. Integrated benthic su/for e 
reduction was es timated by Marvin-Di Pasquale and Capone (1998) and was converted 
to 0 2 ass uming S:02=2. Total channel metabolism was es timated as SR+0.5SR and 
was conve rted to carbon assum ing respiratory quotient =J.0. Regional average 
benlhic metabo lism was computed from channel rates assuming appropri ate rates 
outside 1he channel were 50% of channel rates (Kemp et a/. 1997). 
Sediment Chan nel 
Carbon soc SR A vcraoe Region Average 2 1 1
Region (mgC m-2) (g 0
2 
2 m- d-
1
) (11111101 S 111 - d- ) (mgC m-~ d- ) (mgC m-2 er') 
UB 346500 0.58 4.64 220 130 
MB 202400 0.18 50.2 1240 839 
LB 77000 0.72 26.4 770 559 
433 
-
Table A- 12. Biomass and bioenergetic rates for benthic bacteria, meiofauna, 
suspension feeders and deposil feeders. UB=Upper Bay, MB=Mid Bay, LB=Lower 
Ba y. B. 7 iomass has unit s mgC m ?-, a 11 rates are mg C m .-) d_1.  
Re~ion Biomass Producti on Respiration Consumet ion Egestion 
fle111/1ic Bacteria 
UB 30 34 80 I 14 
MB 0 298 337 787 11 24 
LB no 0 249 580 829 0 
Meir 4'r11, 11 a 
UB 700 49 49 196 98 
MB 476 33 33 133 67 
LB 700 49 49 196 98 
S{(.\JH:'11.1?io11 Feeders 
UB 27232 218 327 1089 545 
MB 42 1 6 9 29 15 
LB 6962 98 146 488 244 
Deposir Feeders 
UB 2368 33 41 166 92 
MB I030 14 21 71 36 
LB 4089 57 86 287 143 
434 
Table A? 13. Estimates o f biomass (B, mgC m?'). speci f ic growth rate (P/B, d?1), oross 
growth c ffici ency (GG E=.PiC), net growl h cffic ienc.y (NGE), assimilation efficie~cy 
(A ), biomass accumulation (BA), nel m1grauons (1mm1gration-emigration), and 
removal s by recreational fishing (Rf) and commercial fishing (CF). Unless otherwise 
2 1
indicated, rates have units 111gC 111? ct? . Efficiencies are unit less. 
B- -- BA 1-E RF CF GGE NGE AE P/B I. 82 
0.58 o. 13 
Upeer Ba~ 0.004 0.33 
0.57 
6 10 0.87 -48 
Blue crab 0.25 0.29 
2 136 0.025 0.69 4.00 
M enhaden 0.08 I 0.28 
0.40 
287 
Bay anchovy 0.28 0.41 
0.68 
2 12 0.010 0.42 0.68 0.14 Herrin gs and Shads 0.002 0.29 282 0.4 l 0.68 White Perch 0.28 
195 0.010 0.41 0.68 
Spot 50 0.010 
0.28 
0.68 
Croaker 0.025 0.17 
0.25 0.146 0.038 
100 0.68 
Hogchoker 0.37 450 0.001 
0.25 
0.10 0.83 
0.24 
Catl'i sh 0.08 0.197 0.009 172 0.002 o.1 2 0.83 
Striped Bass 68 0.007 
o.10 0.036 0.002 
o. 13 0.16 0.83 
0.42 
Bluefish 0.1 26 67 0.009 0.68 -0.02 
W eakfi sh 0.003 0.28 
0.41 
35 
A meri can ee l 1.25 2.09 
0.57 0.58 
Mid Ba 0.33 390 0.004 0.29 0.87 
48 
Blue crab 0.25 2 136 0.025 0.40 0.69 
5.41 
Menhaden 0.081 0.28 38 1 0.42 0.68 
0.01 
0.01 8 
Bay anchovy 29 0.002 
0.29 0.02 1 
0.4 l 0.68 0.01 2 
White Perch 0.28 0.020 222 0.010 0.41 0.68 
Spot 0.0 10 0.28 226 0.25 0.68 0.004 
Croaker 50 0.025 
o.l  7 0.01 5 
0.37 0.68 
L-logchoker 45 0.001 
0.25 
0.10 0.83 
0.24 
0.002 0.08 0.197 
0.009 
Catfi sh 172 o. J2  0.83 0.036 0.002 
Striped Bass 68 0.007 
0.10 
0.13 o. 16 0.83 
0.42 
0.032 
Bluefish 67 0.009 0.41 0.68 -0.006 0.28 
Weakfi sh 9 0.003 
A meri can eel 
435 
Table A- 13, continued. 
AE BA 1-E RF 
CF 
~E 
13 1.09 
Lower Ba~ 0.575 0.58 
0. 14 
0.004 0.333 30.6 
Blue crab 342 0.87 0.025 0.25 
0.29 
Menhaden 2136 0.46 0.60 
5.41 
0.081 0.28 0.12 0.089 0.163 
Bay anchovy 38 1 0.28 0.41 
0.68 
0.0 10 0.050 0.080 570 0.41 0.68 Spot 0.0 10 0.28 0.68 
Croaker 236 0.040 0. 19 
0.28 
JOO 0. 10 0.83 0.24 Hogchokcr 0.002 0.08 0.210 
0.020 
0.83 
Striped Bass 172 0.10 0.12 0.359 0.059 
68 0.007 0.83 1.03 
Bluefish o. 13 0.16 
Weakri ?h 21 l 
0.009 
436 
Table A-14. (A) The fraction of all blue crab habitat located in the 
Chesapeake Bay and its tributaries that is located in each region of the 
mainstcm Bay. ( B) The total blue crab biomass in each region. (C) The 
Iota/ surface area of each region (includes non blue crab habitat area). 
(D) The blue crab biomass per unit of bottom area. 
Region (A) (B) (C) (D) 
Habital Biomass Tola/ Area Biomass 
Area (tons) (km 2) (mac m-2) 
UB 0.024 3597 472 610 
MB 0.076 l 1389 2338 390 
LB 0.076 11 389 2661 342 
Table A-15. Biomass (MT wet-wt) of adult Bay anchovy in 
Chesapeake Bay during summer of' 1993. Data from Rilling and 
Houde (1999) as recomputed from raw data (G . C. Ri /Jing, 
personal communication). 
Upeer Ba~ Mid Bat Lower Ba,t Total 
June 2653 8556 12808 24017 
July 142 4956 18096 23 194 
Average 1398 6756 15452 23606 
( a 111 -2) ( 1.94) (3.59) (4 .14) (3. 73) 
437 
Table A- I 6. Preliminary cstimalcs of biomass (Ions 
WCI-wt) or adu /1 Bay anchovy in Chesapeake Bay 
during summer of / 995- J 999. Dara obrained from E. 
Houde (pcrsonu / communicarion). 
Year Bay-Wide Biomass % 1993 Esr imate 
(Ions wer-wt) 
1993 1 23606 100 
1995 5597 24 
1996 3058 13 
I 997 25236 107 
1998 2633 1 11 2 
1999 13969 59 
A vcrao c 16300 69 
Dara as reported in Table A-15. 
438 
Table A-17. Age-structured carbon balance for bay anchovy in 
Chesapea ke Bay during summer. Biomass has units mgC m-2_ 
Rares have units mgC m-2 summer-1? Specific rares have units d-1 
and were computed from summer-1 rares assuming 92 days fo r the 
age I+ cohort and 46 days ror the age O+ cohort. 
Age O+ Age l+ Combined 
Average Biomass 368 197 381 l 
Change in Biomass +51 8 -20 +498 
Production 2573 290 2863 
Somatic 2573 39 26 12 
Reproducti ve 0 252 252 
Consumption 6524 2575 9099 
Eges rion+Excretion 2120 837 2957 
Respiration 2799 1758 4557 
Balance -967 -3 11 -1278 
Spec(fic Ra!es 
P/8 7.0 J. 5 
P/R 0.92 0. 17 
Specific Cons. (d-1) 0.39 0. 14 
SpecificResp.(d-1) 0.1 7 0. 10 
1 computed as half of age O+ biomass plus age I+ biomass. 
439 
Table A-18 . Diet composition (%) by weight of adult bay anchovy 
during June in mid-Chesapeake Bay. Data from Klebasco ( 1991) . 
Au ust 
97.5 
Diel Com anent 6.0 96. 1 1.8 
Copepods 56.2 0.3 
Barnac les (cyprids) 3.5 
Pol ychactes 20.6 
Amph ipods J3.0 2.3 0.7 
Unid . Crustaceans 0.7 1.3 
Other 
440 
Tab le A-19. Speci fic consumption and tota l consumption fo r bay anchovy. 
Spec i f ic consumption from Klebasco (199 1). 
- Cohort June July Au~ust Summer 
Spec ifi c consumplion (er 1) 
Age O+ (YOY) n/a 0.303 0.393 n/a 
!',du/rs (Age /+) 0.07 0.184 0.172 nla 
Consumption (mgC m-2 peri od-1) 
Age O+ (YOY) 1468 5056 6524 
Adults (Ao-e l+) 430 11 28 1018 2575 
Whole Populalion 430 2596 6074 9099 
(~p_ec ifi c rate, er 1) (0.26) 
441 
Tab le A -20. Consumpti on for the combined bay anchovy cohorts with diet 
all ocated to diet co mponents accordin g to diet compositi on in Table A- 18. The 
di v ision o r copcpod compositi on into mesozoopl ankton and microzooplankton 
is approx imate and based on the age structure of the bay anchovy populati on. 
Young-o f-the-yea r anchovy were assumed to eat copepodites (Kl ebasco 199 L), 
which arc a component of the microzooplankton. 
Diet Component June Jul y August Summer 
(% of total) 
M csozooplankton 85 l812 4175 6071 
(67%) 
Microzoopl ankton 0 776 1789 2566 
(28%) 
M croplankton 257 8 109 374 
(4%) 
Bcnthi c Susp. 89 0 0 89 
Feeders (L %) 
10TAL 430 2596 6074 9099 
442 
,Table A-21. Monthly and summer torn/ respiration (mgC m-2) by bay 
dnch_ovy. Rares were converted from 0 2 consumption using respiratory 
quo11en1 ==1.0. 
Cohort June July August Summer 
(d -') 
YQy 
0 669 2130 2799 
~ / + 418 694 646 1758 
Tora / 418 1363 2776 4557 
(50) 
Tab le A -22. Adjus tments to initial bioenergetic estimates for bay anchovy 
lo achieve carbon balance fo r the summer. nc == no change. 
Rate Tnitinl Fi nal Value Daily Rate 
- Va lue (% difference) (C se,ecific rate, d-1) 
Consumption 9099 10191 (+1 2%) 111 (0.29) 
Respirnlion 4557 4016 (- 12%) 46(0.12) 
Egeslion+Excretion 2957 33 12 (+1 2%) 36 (0.09) 
Production 2863 2863 (nc) 31 (0.08) 
443 
COMBfNED LITERATURE CITED 
Asper, V. L. 1987. A review of sediment trap technique. Mar. Tech. Soc. J. 21 (2): 
18-25 . 
Baden, S. P. , L. Pihl , and R. Rosenburg. 1990. Effects of eutrophication on benthic 
communities including fish - Swedish west coast. Ambia 19: 11 3-122. 
Baines, S. B. and M. L. Pace. 1991. The production of dissolved organic matter by 
phytoplankton and its importance to bacteria: Patterns across marine and 
freshwater systems. Lim no!. Oceanogr. 36(6): 1078- 1090. 
Baird , D. and R. E. Ulanowicz. 1989. The seasona l dynamics of the Chesapeake Bay 
Ecosystem . Ecol. Monogr. 59(4): 329-3 64 . 
Banse, K. and S. Mosher. 1980. Adult body mass and annual production/biomass 
relati onsh ips of field populations. Ecol. Monogr. 50(3): 355-379. 
Barger, LE. 1990. Age and growth of bluefish Pomatomus saltatrix from the northern 
Gulf of Mexico and U.S. south Atlantic coast. Fish. Bull. , U.S. 88:805-809. 
Batiuk, R. R. Orth, K. Moore, W. Dennison, J. Stevenson, L. Staver, V. Carter, N. 
Rybicki , R. Hickman, S. Kollar, S. Bieber, and P. Heasly. 1992. Chesapeake 
Bay Submerged Aquatic Vegetation Habitat Requirements and Restoration 
Targets: A Technical Synthesis. EPA Chesapeake Bay Program, Annapolis, 
MD. 
Bayne, B. L. and R. C. Newell. 1983. Physiological energetics of marine molluscs. 
pp. 407-515 . In: Saleuddin, A. S. M. and K. L. Wilbur, eds. The Mollusca, 
Volume 4. Physiology, Part I. New York: Academic Press. 
Berggreen, U., B. Hansen and T. Ki0rboe. 1988. Food size spectra, ingestion and 
growth of the copepod Acartia tonsa during development: implications for 
determination of copepod production. Mar. Biol. 99: 341-352. 
Bienfang, P. K. 1981. Sinking rates of heterogeneous, temperate phytoplankton 
populations. J. Plank. Res .. 3(2): 235-253. 
Blomqvist, S. and L. Hakanson. 1981. A review on sediment traps in aquatic 
environments. Arch. Fur Hydrob. 91 ( l ): 101-132. 
Blumenshine, S. and W. M. Kemp. 2000. Food versus habitat limiation for benthic 
macrofauna in mesohaline regions of Chesapeake Bay. In: Kemp, W. M., R. 
Bartleson, S. Blumenshine, J. D. Hagy and W. R. Boynton. Ecosystem models 
of the Chesapeake Bay relating nutrient loadings, environmental conditions, 
and living resources. US EPA Chesapeake Bay Program, Annapolis, MD. 
444 
Boesch D E B . 
?, . urreson, W. Dennison, E. Houde, W. M. Kemp, V. Kennedy R 
Newell , K. Paynter, R. Orth, and R. Ulanowicz. 2001. Factors in the deciine 
coastal ecosystems. Science. 293 (5535) : 1589-1590. of 
Bogdan, K. G. _and J. J. Gilbert. 1982. Seasonal patterns of feeding by natural 
populations of Keratel!a, Po/yarthra, and Bosmina: Clearance rates 
selectivities and contributions to community grazing. Limnof. Ocea~oo- : 
918-934. or. 27
Bogdan, K. G. and J. J. Gilbert. 1984. Body size and food size in freshwater 
zooplankton. Proceedings of the National Academy of Sciences (USA) 8l: 
6427-6431. 
Boicourt, W. C. 1992. Influences of circulation processes on dissolved oxygen in !he 
Chesapeake Bay. In: Smith, D. E. , M. Leffler, and G. Mackieman (eds.). 
Oxygen Dynamics in the Chesapeake Bay. A synthesis ofrecent research. 
Maryland Sea Grant College, College Park, MD. 
Boreman, J. 1997. Methods for comparing the impacts of pollution and fishing on fish 
populations. Trans. Am. Fish. Soc. 126: 506-513. 
B0rsheim, K. Y. and G. Bratbak. 1987. Cell volume to cell carbon conversion factors 
for a bacterivorous Monas sp. enriched from seawater. Mar Ecol. Prog. Ser. 
36: 171-175. 
Boynton, W.R., W.M. Kemp, J.M. Barnes, L.L. Matteson, F.M. Rohland, D.A. 
Jasinski and H.L. Kimble. 1993. Ecosystem Processes Component Level J 
Interpretive Report No. 10. Chesapeake Biological Laboratory (CBL), 
University of Maryland System, Solomons, MD 20688-0038. Ref 
No.[UMCEES]CBL 93-030a. 
Boynton, W. R. and W. M. Kemp. 2000. Influence ofriver flow and nutrient loads on 
selected ecosystem processes. A synthesis of Chesapeake Bay data. pp. 269-
298 In: J.E. Hobbie (ed.). Estuarine Science: A Synthetic Approach to 
Rese;.ch and Practice. Island Press, Washington, DC. 
Boynton, W. R., R. M. Stankelis, F. M. Rohland, 1. D. Hagy and J. M. Frank. 1999. 
Maryland Chesapeake Bay Water Quality ~onitoring Program. Ecosystem 
Processes Component. Level One Interpretive Report No. 16. Chesapeake 
Biological Laboratory. University of Maryland Center for Environmental 
Science. Solomons, MD. 20688-0038. Ref No. [UMCES]CBL 99-0070. 
Boynton, w. R., personal communication. UMCES Chesapeake Biological 
Laboratory, Solomons, MD. 
Breiman, L., J. Friedman, R. Olshen, and C. Stone. 1984. Classification anq 
Regression Trees. Pacific Grove: Wadsworth. 
445 
Breitburg, D. L. 1990. Nearshore hypoxia in the Chesapeake Bay: Patterns and 
relationships among physical factors. Est. Coast. ShelfS ci. . 30(6): 593-609. 
Breitburg, D. L., T. Laher, C. A. Pacey, and A. Gerstein. 1997. Varying effects of 
low dissolved oxygen on trophic interactions in an estuarine food web. Ecol. 
Monogr. 67(4): 489-507. 
Budd, J. D., T. D. Drummer, T. F. Nalepa, and G. L. Fhnenstiel. 2001. Remote 
sensing of biotic effects: Zebra mussels (Dreissena polymo,pha) influence on 
water clarity in Saginaw Bay, Lake Huron. Limnol. Oceanogr. 46(2): 213-
223. 
Bue, C. D. 1968. Monthly surface-water inflow to Chesapeake Bay. U.S. Geological 
Survey Open-File Report. Arlington, VA. October 1968. 
Butman, C. A. J 986. Sediment trap biases in turbulent flow: Results from a 
laboratory flume study. J. Mar. Res. 44: 645-693. 
Butman, C. A., W. D. Grant, and K. D. Stolzenbach. 1986. Predictions of sediment 
trap biases in turbulent flows: A theoretical analysis based on observations 
from the literature. J. Mar. Res. 44: 601-644. 
Butowski, N. H. 1994. Chesapeake Bay Oyster Fishery Management Plan. 
Agreement Commitment Report Revision. US EPA Chesapeake Bay Program, 
Annapolis, MD. 
Carpenter, J. H. 1965. The Chesapeake Bay Institute technique for the Winkler 
titration. Limnol. Oceanogr. 10(1): 141-143. 
Carpenter, J. H. J 965b. The accuracy of the Winkler method for dissolved oxygen 
analysis. Limnol. Oceanogr. 10(1 ): 135-140. 
Cebrian, J. 1999. Patterns in the fate of production in plant communities. Am. Nat. 
154( 4): 449-468. 
Chao, s. Y. and T. Paluszkiewicz. 1991. The hydraulics of density currents over 
estuarine sills. J. Geophys. Res. 96:7065-7076. 
Chesson, J. J 983 . The estimation and analysis of preference and its relationship to 
foraging models. Ecology 64(5): 1297-1304. 
Choi, J. w., B. F. Sherr, and E. B. Sherr. 1999. Dead or alive? A large fraction of 
ETS-inactive marine bacterioplankton cells, as assessed by reduction of CTC 
can become ETS-active with incubation and substrate addition. Aquat. ' 
Microb. Ecol. 18 (2): 105-115. 
446 
Christensen, V. and D. Pauly. 1992. ECO~ATH 11 - a software for balancing steady-
state ecosystem models and calculatmg network characteristics. Ecol. 
Modelling. 61: 169-185. 
Christensen, V. and D. Pa~ly. 1993. Flow characteristics of a9uatic ecosystems, p. 
338-352. Jn V. Chnstensen and D . Pauly (eds.) Trophic models of aquatic 
ecosystems. ICLARM Conference Proceedings No. 26. 
Christensen, V., C.J. Walters, and D. Pauly. 2000. Ecopath with Ecosim: A user's 
g uide. October 2000 edition. Fisheries Centre, Univ. of British Columbia 
Vancouver, Canada and International Center for Living and Aquatic Reso~rces 
Management,, Penang, Malaysia. 
C loern, J. E. 2001. Our evolving conceptual model of the coastal eutrophication 
problem. Mar. Ecol. Prog. Ser. 210: 223-253. 
Cole, J. J. S. Findlay, and M. L. Pace. 1988. Bacterial production in fresh and 
saltwater ecosystems: a cross-system overview. Mar. Ecol. Prag. Ser. 43: 1-
10. 
Conley, D. J. and T. C. Malone. 1992. Annual cycle of dissolved silicate in 
Chesapeake Bay: implications for the production and fate ofp hytoplankton 
biomass. Mar. Ecol. Prog. Ser. 81: 121-128. 
Constable, A. J. 1999. Ecology of benthic macro-invertebrates in soft-sediment 
environments: A review of progress towards quantitative models and 
predictions. Aust. J Ecol. 24(4): 452-476. 
Cowan, J. L. and W. R. Boynton. 1996. Sediment-water oxygen and nutrient 
exchanges along the longitudinal axis of Chesapeake Bay: Seasonal patterns, 
controlling factors and ecological significance. Estuaries 19(3):562-580. 
Cronin, w. B. and D. W. Pritchard. 1975. Additional statistics on the dimensions of 
Chesapeake Bay and its tributaries: Cross section widths and segment volumes 
per meter depth. Chesapeake Bay Institute. Reference 75-3. Special Report 
42. 
de Jonge, v. N., W. Boynton, C. F. D'Elia, R. Elmgren, and B. L. Welsh. 1994. 
Responses to developments in eutrophication in four different North Atlantic 
estuarine systems. Jn: Dyer, K. R. and R. J. Orth. ( eds.). Changes In Fluxes Jn 
Estuaries: Implications From Science To Management. Olsen & Olsen, 
Fredensborg, Denmark. pp. 179-196. 
de! Giorgio, P.A. and J. J. Cole. Bacterial growth efficiency in natural aquatic 
systems. Ann. Rev. Ecol. Syst. 29: 503-41. 
447 
Del/apenna'. T. M.,. S. A. ~ueh/ .and. L. C: Schaffner. 1998. Sea-bed mixing and 
particle residence times m b10/og1cally and physically dominated estuarine 
systems: a comparison of lower Chesapeake Bay and the York River 
subestuary. Est. Coal. ShelfS ci. 46: 777-795. 
D . 
ernck, P. A. and V. S. Kennedy. 1997. Prey selection by the hogchoker, Trinectes 
maculatus (Pisces: Soleidae), along summer salinity gradients in Chesapeake 
Bay, USA. Mar. Biol. 129: 699-711. 
Detwy/er, R. and E. D. Houde. 1970. Food selection by laboratory-rearer larvae of 
the scaled sardine Harengula pensaco/ae (Pisces: Clupeidae) and the bay 
anchovy Anchoa mitchilli (Pisces: Engraulidae). Mar. Biol. 7: 214-222. 
Diaz, R. J. and L. C. Schaffner. 1990. The functional role of the estuarine benthos. 
pp. 25-56 ln: in Haire, M. and E. C. Krome, (eds.). Perspectives on the 
Chesapeake Bay: Advances in Estuarine Science. Chesapeake Research 
Consortium, Solomons, MD. 
Diaz, R. J. , R. J. Neubauer, L. C. Schaffner, L. Pihl, and S. P. Baden. J 992. 
Continuous monitoring of dissolved oxygen in an estuary experiencing 
periodic hypoxia and the effect of hypoxia on macrobenthos and fish. Sci. Tot. 
Env. (Supplement 1992): 1055-1067. 
Dolan, J. R. 1991 a. Guilds of ciliate microzooplankton in the Chesapeake Bay. Est. 
Coast. ShelfS ci. 33: 137-152. 
Dolan, J. R. 1991 b. Ciliate growth rates and consumption by copepods. Mar. Biol. 
111: 303-309. 
Dolan, J. R. and D. w. Coats. 1990. Seasonal abundances of planktonic ciliates and 
microfJageilates in mesohaline Chesapeake Bay waters. Est. Coast. Shelf Sci. 
31: 157-175. 
Dolan, J. R. and c. L. GaIIegos. 1991. Trophic coupling of rotifers, microflage!Jates, 
and bacteria during fall months in the Rhode River estuary. Mar. Ecol. Prog. 
Ser. 77: 147-156. 
Dovel, W. L., J. A. Mihursky and A. 1. McErlean. 1969. Life history aspects of the 
hogchoker, Trinectes maculatus, in the Patuxent River estuary, Maryland. 
Ches. Sci. 10(2): 104-119. 
Duarte, C. M. and C. L. Chiscano. 1999. Seagrass biomass and production: a 
reassessment. Aquat. Bot. 65: 159-174. 
Ducklow, H. w. and F. Shiah. 1993. Bacterial production in estuaries. pp 261-287. 
Jn: Ford, T. E. (ed.). Aquatic Microbiol~gy: An Ecological Approach. 
Blackwell Scientific Publications, Cambndge, MA. 518 pp. 
448 
Durbin, E. G._and A.G. Durbin. 1981. Asssimi/ation efficiency and nitrogen 
excretion of a filter-feeding p/anktivore, the Atlantic menhaden, Brevoortia 
tyrannus (Pisces: Clupeidae). Fish. Bull. 79(4): 601-616. 
Ebersole, E. L. and V. S. Kennedy. 1994. Size selection of Atlantic Rangia clams 
Rangia cuneata, by blue crabs, Caflinectes sapidus. Estuaries 17(3):668-673. 
Edgar, G. I 990. The use of the size structure of benthic macrofauna/ communities to 
estimate fauna/ biomass and secondary production. J Exp. Mar. Bio. Eco!. 
137: 195-214. 
E lliott, A. J. 1978. Observations of the meteorological/y induced circulation in the 
Potomac Estuary. Est. Coast. Sheff Sci. 6: 285-299. 
EPA 1989. Chesapeake Bay A/osid Management Plan. US EPA Chesapeake Bay 
Program, Annapolis, MD. 
EPA 1990a. Chesapeake Bay Bluefish Fishery Management Plan. Chesapeake Bay 
Program Agreement Commitment Report. US EPA Chesapeake Bay Program, 
Annapolis, MD. 
EPA 1990b. Chesapeake Bay Weakfish and Spotted Sea trout Fishery Management 
Plan. Chesapeake Bay Program Agreement Commitment Repo1t. Chesapeake 
Bay Program, Annapolis, MD. 
EPA 1991 a. Chesapeake Bay Atlantic Croaker and Spot Fishery Management Plan. 
Chesapeake Bay Program, Annapolis, MD. 
EPA 1991 b. Chesapeake Bay American Eel Fishery Management Plan. Chesapeake 
Bay Program, Annapolis, MD. 
EPA 1995. Chesapeake Bay Striped Bass Fishery Management Plan. Annual 
Progress Report. 1994. EPA Chesapeake Bay Program, Annapolis, MD. 
EPA 1997. 1997 Chesapeake Bay Blue Crab Fishery Management Plan. June 1997. 
US EPA Chesapeake Bay Program, Annapolis, MD. 
EPA 2000. The 2000 User's Guide to Chesapeake Bay Program Biological and Living 
Resources Monitoring Data. US Environmental Protection Agency, 
Chesapeake Bay Program, Annapolis, MD. 
EPA 2000b. Chesapeake 2000, A Watershed Partnership. EPA Chesapeake Bay 
Program, Annapolis, MD. 
EPA . 2000c. Chesapeake Bay 2000 Agreement. US EPA Chesapeake Bay Program, 
Annapolis, MD. 
449 
Fenchel, T. 1978. 
_1 _ The ecology of micro- and meiobenthos. Ann. Rev. Ecol. Sys!. 9: 99 21 
Fenchel T 19 
' ? 82. Ecology of heterotrophic microfJageJ/ates. II. Bioenergetics and 
growth. Mar Ecol. Prag. Ser. 8: 225-231. 
Fisher y R 
' ? ?, .1.  D. Hagy and Emma Rochelle-Newall. 1998. Dissolved and particulate 
organic carbon in Chesapeake Bay. Estuaries. 21 (2): 215-229. 
Fisher ' T.  ? R_ ? , E . R. Peele, J. W. Ammerman, and L. W. Hardi.n g, Jr. 1992. Nutrient 
~~~itation of phytoplankton in Chesapeake Bay. Mar. Ecol. Prog. Ser. 82: 51-
Flemer, D. A., G. B. Mackiernan, W. Nehlsen, V. K. Tippie, R. B. Biggs, D. Blaylock 
N. H. Burger, L. C. Davidson, D. Haberman, K. S. Price and J. L. Taft. 1983. ' 
Chesapeake Bay: A Profile of environmental change. U.S. EPA Chesapeake 
Bay Program Report. Annapolis, MD. 
Froese, R. and D. Pauly (eds.) 2001. FishBase. World Wide Web electronic 
publication. http://www.fishbase.org. October 4, 2001. 
Flach, E., J. Vanaverbeke and C. Heip. 1999. The meiofauna: macrofauna ratio 
across the continental slope of the Goban Spur (north-east Atlantic). J Exp. 
Mar. Biol. Ecol. 79(2): 233-241. 
Friedland, K. D., D. W. Ahrenholz, and J. F. Guthrie. 1989. Influence of plankton on 
distribution patterns of the filter-feeder Brevoortia tyrannus (Pisces: 
Clupeidae). Mar Ecol. Prog. Ser. 54: 1-11. 
Furlong, E. T. and R. Carpenter. 1988. Pigment preservation and remineralization in 
oxic coastal marine sediments. Geochim. Cosmochim. Acta. 52: 87-99. 
Gaedke, U. and D. Straile. 1994. Seasonal changes of the quantitative importance of 
protozoans in a large Jake. An ecosystem approach using mass-balanced 
carbon flow diagrams. Mar. Microb. Food Webs. 8( 1-2): 163-188. 
Garber, J. H., w. R. Boynton, J. M. Barnes, L. L. Matteson, J. L. Watts, and S. 
Stammerjohn. 1989. Ecosystem Processes Component (EPC) and Benthic 
Exchange and Sediment Transformation (BEST). Maryland Department of the 
Environment, Maryland Chesapeake Bay Water Quality Monitoring Program. 
Chesapeake Biological Laboratory (CBL), Solomons, MD 20688 
Gerlach, S. A. 1971. On the importance of marine meiofauna for benthic 
communities. Oecologia. 6: 176-190. 
450 
----
Gerri tsen J A F H 11 d d D E . 
, ., . . . . o an , an : . lrvme. 1994. . Suspension-feeding biva/v/es and 
the fate of primary production: An estuarine model applied to Chesapeake 
Bay. Estuaries. 17(2): 403-416. 
Gifford,?? 1. and D. A. Caron. 2000. Sampling preservation, enumeration and 
biomass of marine protozooplankton. In: R. Harris, P. Wiebe, J. Lenz, H. R. 
Rune Skjoldal and M. Huntley, (eds.) ICES Zooplankton Methodology 
Manual. Academic Press, New York. 
Gordon, D. C., Jr. , P. R. Boudreau, K. H. Mann, 1.-E. Ong, W. L. SiJvert, S. V. Smith 
G. Wattayakorn, F. Wul ff, and T. Tanagi. 1996. LOICZ Biogeochemica/ ' 
Modelling Guidelines. LOICZIR&S/95-5, vi +96 pp. LOJCZ, Texel, The 
Nether/ands. 
Graf, G., W. Beng tsson, U. Diesner, R. Schultz, and H. Theede. 1982. Benthic 
response to sedimentation of a spring phytoplankton bloom: process and 
budget. Mar. Biol. 67: 201-208. 
Hagy, ./. D. 1996. Residence Times and Net Ecosystem Processes in Patuxent River 
Estuary. M.S. Thesis. University of Maryland at College Park. 
Hagy, 1. D . and W. R. Boynton. 2000. Controls on hypoxia in Chesapeake Bay and 
its major tributaries. pp. 91- 102. In: Kemp, W. M., R. Bartleson, S. 
B/umenshine, J. D. Hagy and W. R. Boynton. Ecosystem Models of the 
Chesapeake Bay Relating Nutrient Loadings, Environmental Conditions, and 
Living Resources. Final Report. VS EPA Chesapeake Bay Program, 
Annapolis, MD. 
Hagy, J. D. 2000. Tree-structured models of water quality and submersed aquatic 
vegeta tion in Chesapeake Bay. In : Kemp, W. M., R. Bartleson, S. 
B/umenshine, J. D. Hagy, and W. R. Boynton. Ecosystem models of the 
Chesapeake Bay relating Nutrient Loadings, Environmental Conditions and 
Living Resources. US EPA Chespeake Bay Program, Annapolis, MD. 
Hagy, J. D. , L. P. Sanford, and W.R. Boynton. 2000. Estimation of net physical 
transport and hydraulic residence times for a coastal plain estuary using box 
models. Estuaries. 23(3): 328-340. 
Hanson, J.M. and w. C. Leggett. 1982. Empirical prediction offish biomass and 
yield. Can. J Fish. Aqua/. Sci. 39: 257-263. 
H ansen, B. w. 1999. Cohort growth of planktotrophic polychaete larvae - are they 
food limited? Mar. Ecol. Prog. Ser. 178: 109- 11 9. 
451 
rd
Ha ing,_ L. W., Jr. and E. S. Perry. 1997. Long-term increase of phytoplankton 
biomass in Chesapeake Bay, I 950-94. Mar. Ecol. Prog. Ser. 157: 39-52. 
I lard ing, L. W., Jr., M. E. Mallonee, and E. S. Perry. 2001. Toward a predictive 
understanding of primary productivity in a temperate, partially-stratified 
estuary. Est. Coast. ShelfS ci. (in press). 
Hargis, W. J. , Jr. and D. S. Haven. 1994. The precarious state of the Chesapeake 
public oyster resource. In : Hill, P. and S. Nelson, (eds.). Toward a Sustainable 
Watershed: The Chesapeake Experiment. Proceedings of the 1994 
Chesap eake Research Conference. Chesapeake Research Consortium 
Publication Number 149. ' 
I lartman K. J., and S. B. Brandt. 1995a. Trophic resource partitioning, diets, and 
growth ofsympatric estuarine predators. Trans. Am. Fish. Soc. 124 (4): 520-
537. 
Ha11111an, K. J. and S. B. Brandt. 1995b. Comparative energetics and the development 
of bioenergetics models for sympatric estuarine predators. Can. J Fish. Aqua/. 
Sci. 52: 1647-1 666. 
Hartman, K. J. and S. B. Brandt. 1995c. Estimating energy density of fish. Trans. 
Am. Fish. Soc. 124: 347-355. 
Heinbokel, J. F. , D. W. Coats, K. W. Henserson, and M . A. Tyler. 1988. 
Reprodoction and secondary production of three species of the rotifer genus 
Synchaeta in the estuarine Potomac River. J Plank. Res. 10(4): 659-674. 
Heinle, D. R. J 966. Production of a calanoid copepod, A cartia lonsa, in the Patuxent 
River estuary. Ches. Sci. 7(2): 59-74. 
Hines, A. H., A. M. Haddon, and L.A. Wiechert. 1990. Guild structure and foraging 
impact of blue crabs and epibenthic fish in a subestuary of Chesapeake Bay. 
Mar. Ecol. Prog. Ser. 67: 105-126. 
Hoenig, J. 1983. Empirical use of longevity da ta to estimate mortality rates. Fish. 
Bull. 82(1): 899-903. 
Hobbs, C. H., III, J. P. Halka, R. T. Kerhin, and m. J. Carron. 1992. Chesapeake Bay 
Sediment Budget. J Coast. Res. 8(2): 292-300. 
Holland, A. F., A . T. Shaughnessy, and M. H. Hiegel. 1~ 87. Long-term varia tion in 
mesoha/ine Chesap eake Bay macrobenthos: Spatial and temporal patterns. 
Estuaries 10(3):227-245. 
452 
Hollis, E. H. 1952. Variations in the feeding habits of the striped bass Roccus 
?;-;~~Ii/is (Wa lbaum), in Chesapeake Bay. Bull. Bing. Ocean. C~ll. 14: 111-
Hood R. R., ~- V. Wang, J.E. Purcell, E. D. Houde, and L. W. Harding, Jr. 1999. 
M ode/mg particles and pe/~gi~ or?anisms in Chesapeake Bay: Convergent 
fea tures control plankton d1stnbut10ns. J Geophys. Res. 104(C2): 3289-3290. 
Hopkins, K. , B. Brown, L. C. Linker, a,~d R. L. Mader, _Jr. 2000. Chesapeake Bay 
watershed model land uses and lmkages to the a,rshed and estuarine models. 
US EPA Chesapeake Bay Program, Annapolis, MD. 
Houde, E. D. and E. S. Rutherford. 1993. Recent trends in estuarine fisheries: 
Prediction of fish production and yield. Estuaries: 16(2): 161-176. 
Houde, E. D., the late J. C. Gamble, S. E. _Dorsey, and I. H. Cowan, Jr. 1994. Drifting 
mesocosms: the influence of gelatmous zooplankton on mortality of bay 
anchovy, Anchoa milchilfi, eggs and yolk-sac larvae. ICES J Mar. Sci. 51: 
383-394. 
Houde, E. D., S. Juki_-Pe/adic, S. B. Brandt and S. D. Leach. 1999. Fisheries: Trends 
in catches, abundance and management, pp. 341-366. ln: T. C. Malone, A. 
Malej, L. W. Harding, Jr. , N. Smodlaka, and R. E. Turner (eds.) Ecosystems at 
the Land-Sea Margin, Drainage Basin to Coastal Sea. American Geophysical 
Union, Washington, DC. 
Iverson, R. L. 1990. Control of fish production. Limn o/. Oceanogr. 35(7): 1593-
1604. 
Jackson J. B . c., M. x. Kirby, W. H. Berger, K. A. Bjorndal, L. W. Botsford, B. J. 
Bourque, R. H. Bradbury, R. Cooke, I. Erlandson, J. A . Estes, T. P. Hughes, S. 
Kidwell, C. B. Lange, H. S. Lenihan, J. M. Pandolfi, C. H. Peterson, R. S. 
Steneck, M. ]. Tegner and R.R. Warner. 2001. Historical overfishing and the 
recent collapse of coastal ecosystems. Science. 293 (5530): 629-638. 
Janzen, C. D. and K.-C. Wong. 1998. On the Low-Frequency Transport Processes in 
a Shallow Coastal Lagoon. Estuaries 21(  4B): 754-766. 
lass by, A. D. and T. Platt. J 976. Mathematical formulation of the relationship 
between photosynthesis and light for phytoplankton. Limno/. Oceanogr. 21: 
540-547. 
Jensen, M. H. E. Lomstein, and J. S0renson. 1990. Benthic NH4 ~ and N03- flux 
following sedimentation of a spring phytoplankton bloom m Aarhus bight, 
Denmark. Mar. Ecol. Prog. Ser. 61: 87-96. 
453 
Johnson B R H h K . 
, ., . ea~ , . Kim and H_? But!e:. 1991. Development and verification of 
a 3-D numerical hydrodynamic sa/irnty and temperature model of Chesapeake 
Bay. !ech. Rept. HL-91-7. Army Corps of Engineers, Waterways 
Experimentation Station, Vicksburg, MS. 192 pp . 
.Johnson, C., D. Klumpp, J. Field, and R. Bradbury. Carbon flux on coral reefs: 
~:~~cts oflarge shifts in community structure. Mar. Ecol. Prog. Ser. J 26: 123-
.Jordan, S. 1., G. E. Krant, and C. .Judy. 1994. Maryland Chesapeake Bay oyster 
recovery areas: Summary of technical information. In : Hill, P. and S. Nelson 
(e ds.). Toward a Sustainable Watershed: The Chesapeake Experiment. ' 
Proceedings of the 1994 Chesapeake Research Conference. Chesapeake 
Research Consortium, Publication Number 149 . 
.Ju, S.-1., D. H. Secor, and H. R. Harvey. 1999. Use of extractib/e lipofuscin for age 
determination of blue crab Callinectes sapidus. Mar. Ecol. Prag. Ser. J 85: 
171-179. 
lung, S. and E. D. Houde. 2000. Scale and pattern in recruitment processes of bay 
anchovy in Chesapeake Bay. ICES CM2000/M: 13. 
lung, S., unpublished data. Preliminary analysis and presentation of mid-water trawl 
data from the Chesapeake Bay LMER (TIES) project. 
Karlsen, A. W., T. M. Cronin, S. E. Ishman, D. A. Willard, C. W. Holmes, M. Marat, 
and R. Kerhin. 2000. Historical trends in Chesapeake Bay dissolved oxygen 
based on benthic foraminifera from sediment cores. Estuaries. 23(4): 488-
508. 
Karakassis, I, E. Hatziyanni, M. Tsapakis, and W. Plaiti. 1999. Benthic recovery 
i'uiiuwiJ11:, L:c:,::,c1tiuJJ ui' ii:si1 i'm 111i11g; c1 :,c;1 ic:s ui' :suu;c;.s.sc.s c111u l;1:1l<1.Stropi1t:.s. 
Mar. Ecol. Prag. Ser. 184: 205-218. 
Kay, J. J., L.A. Graham, and R. E. Ulanowicz. 1989. A detailed guide to network 
analysis. In: Wulff, F., J. G. Field, an~ K._H. Mann: (eds.). Network Analysis 
in Marine Ecology. Methods and Applications. Spnnger-Verlag, New York. 
Ke//er, A. A. and u. Riebesell. 1989. Phytoplankton carbon dynamics during a 
winter-spring diatom bloom in an enclosed :11arine ecosystem: primary 
production, biomass and Joss rates. Mar. Bwl. 103: 131-142. 
Kemp, W.M., W.R. Boynton, R.R. TwiJiey, J.C. ~tevenson, and J.C. Means. 1983. 
The decline of submerged vascular plants m Chesapeake Bay: summary of 
results concerning possible causes. Mar. Tech. Soc. J. 17(2):78-89. 
454 
Kemp, WM WR . 
In? ?, ? _? Boynton, R.R. Twilley, J.C. Stevenson, and L. G. Ward. 1984_ 
fluences of submersed vascular plants on ecological processes in 
Chesapeak e Ba y. 11 1: V . S . K enned  y (ed). The Estuary as a Filter uApp, edr  ? 
Press, New York, NY. . ca em1c 
Kemp, W. M.: E. M. Smith, M. Marvin-DiPasquale, and W.R. Boynton. 1997. 
Organic carbon balance and net ecosystem metabolism in Chesapeake B 
Mar. Ecol. Prog. Ser. 150: 229-248. ay. 
Kemp, W. M_. , J. Fagane/i, S. Puskaris, E. M. Smith and W. R. Boynton. 1999. 
Pelag1c-Be~thic Coupling and Nutrient Cyc/ine. In : Malone, T. C., L. Ma/ej, 
L. W. 1-Jardrng, Jr., N. Smodlaka, and R. E. Turner (eds.). Ecosystems at the 
1.and-Sea Margin. American Geophysical Union, Washington, DC. 
Kerhin, R. T., J. P. Halka, E. L. Hennessee, P. 1. Blakeslee, D. V. Wells, N. Zoltan 
and R. H. Cuthbertson. 1983. Physical characteristics and sediments budg;t 
for bottom sediments in the Mary/and Portion of Chesapeake Bay. United 
States Environmental Protection Agency, Washington, D. C. 
Ki0 rboe, T. and T. G. Nielsen. 1994. Regulation ofzooplankton biomass and 
production in a temperate, coastal ecosystem. I. Copepods. Limno!. 
Oceanogr. 39(3): 493-507. 
Ki0 rboe, T., F. M0 h/enberg and K. Hamburger. 1985. Bioenergetics of the 
planktonic copepod Acartia tonsa: relation between feeding, egg production 
and respiration, and composition of specific dynamic action. Mar. Ecol. Prag. 
Ser. 26: 85-97. 
Kim merer, W. J. 1987. The theory of secondary production calculations for 
continuously reproducing populations. Lim no!. Oceanogr. 32(1 ): 1- 13. 
Kimura R, D. H. Secor, E. D. Houde ED and P. M. Piccoli. 2000. Up-estuary 
dispersal of young-of-the-year bay a_nchovy Anchoa ?1itchilli in_ the. . 
Chesapeake Bay: inferences from m1croprobe analysis of strontium m otoliths. 
Mar. Ecol. Prog. Ser. 208: 217-227 
Kitazato H, Y. Shirayama, T. Nakatsuka, S. Fujiwara, M. Shimanaga, Y. Kato, Y. 
Okada J. Kanda A. Yamaoka, T. Masuzawa, and K. Suzuki. 2000. Seasonal 
phytod~tritus de;osition and responses of bathyal benthic foraminifera/ 
populations in Sagami Bay, Japan: preliminary results from "Project Sagami 
1996-1999." Mar. Micropa/eonto!. 40 (3): 135-149. 
Kles basco, M. J. 1991 . Feeding ecology and daily ratio ~f bay ~nch~vy, Anchoa 
?t c 111 ? 11 ? m?  th e mi? d -Chesapeake Bay?  M. S. Thesis. Urnvers1ty of Maryland m1 1, 
College Park. College Park, MD. 
455 
Knauer, G. A. D . 
se/ecte' . M. Ka~/, J. H. Martm, and c._N . Hunter. 1984. In situ effects of 
s d . d preserva ti ves on total carbon, nitrogen and metals collected in 
e iment traps. J Mar. Res. 42: 445-462. 
Koch, E. W 200 1 B d L. h Pl . 1 G . 
p ? . ? eyon 1g t: 1ys1ca , eolog1cal, and Geochemical Parameters 
2a4s( ossib/e Submersed Aquatic Vegetation Habitat Requirements. Estuaries 1): 1-17. ? 
Kremer p 
~ ? an~ M . R. Reeve. 1989. Growth dynamics of a ctenophore (Mnemiopsis) 
~~ relation to variable food supply. II. Carbon budgets and growth model. J 
ank Res. 11 (3) : 553-574. 
Kuo, A hy  ?, . K_. ~ark, and M. Z. Moustafa. 199 1. Spatrnl and temporal variabilities of 
YPox1a in the Rappahannock River, Virginia. Estuaries. 14(2): 11 3-1 21. 
Laco utur R . 
e, . ? Philadelphia Academy of Na tura l Science, Estuarine Research Center, 
Sl. Leonard, MD. 
Lang /and M 
' .J., R. E. Edwards, L. A. Sprague and S. E. Yochum. 2001 . Summary of 
Trends and Status Analysis for Flow, Nutrients, and Sediments at Selected 
N_ontidal Sites, Chesapeake Bay Basin, 1985-99: U.S. Geological Survey Open 
FJ/e Report O1 -73, 49 p. 
Lardicc1? C , ?, S. Como, S. Corti, and F. Rossi. 2001. Recovery of the 
macrozoobenthic community after severe dystrophic crises in a Mediterranean 
coastal lagoon (Orbetello, Italy). Mar. Poll Bull. 42(3): 202-214. 
Larson ' R ? J . 1986. The feeding and growth of the sea nettle, C' hrysaora 
quinquecirrha (DeSor), in the laboratory. Estuaries. 9: 376-379. 
La ugh/~n, R. A. 1982. Feeding habits of the blue crab, Callinectes sapidus Rathbun, 
111 the A palachicola estuary, Florida. Bull Mar. Sci. 32(4): 807-822. 
Linley, E. A. S., R. c. Newell, and M. J. Lucas. 1983. Quantitative relationships 
between phytoplankton, bacteria and heterotrophic microflage/lates in shelf 
Waters. Mar. Ecol Prog. Ser. 12: 77-89. 
Loftus, M. E., D. V. Subba Rao, and H. H. Seliger. J9 72. Growth and diss~pation of 
phytoplankton in Chesapeake Bay J. Response to a large pulse oframfall. 
Ches. Sci. 13(4):282-299. 
Luo, J. and S. B. Brandt. 1993. Bay anchovy Anchoa mitchilli pro?uction and 
consumption in mid-Chesapeake Bay based on a bioenerget1cs ~ode/ and 
acoustic measures of fish abundance. Mar. Ecol Prog. Ser. 98? 223-236? 
456 
Luo J S 
' ., . B. Brandt d M J Kl b . . 
fish's er ' _a n . . e asco. 1996. Virtual reality ofp lanktivores: A 
283 P spective of prey size selection. Mar. Ecol Prag. Ser. 140 (1-3): 271-
Luo, J., K 1 
? . ? Hartman, S. B. Brandt, C. F. Cereo, and T. H. Rippetoe. 2001. A 
spatially-ex 11?  . h I-". ? ? ? ? th . P c1t approac ,or estimating carrying capacity: An application for 
24e( Atlantic Menhaden (Brevoortia tyrannus) in Chesapeake Bay. Estuaries. 4): 545-556. 
MacIntyre F 
1
' ? L., R. I. Geider, and D. C. Mil/er. 1996. Microphytobenthos: The 
heac o./o gi I ro_1  f ca e ~? th_e "Secret . Garden" of u~vegetated, sh_a JJow-water man.n e 
19 b1tats. I. Distnbut1on, abundance, and primary product10n. Estuaries. (2A): 186-201. 
MaloneN T  ? C.  ? 1992. Effects of Water Column Processes on Dissolved Oxygen, 
utnents, Phytoplankton and Zooplankton. In: Smith, D. E., M. Leffler, and 
G. Mackieman (eds.). Oxygen Dynamics in the Chesapeake Bay. A synthesis 
~ - Maryland Sea Grant Co//ege, College Park, MD. 
Malone T 
j ? C., W. M. Kemp, H. W. Ducklow, W. R. _Boynton, J. H: Tuttle and R. ~ -
onas. 1986. Lateral variation in the product10n and fate ot phytoplankton m a 
Partially stratified estuary. Mar. Ecol Prog. Ser. 32(2-3): 149-160. 
Malone T C . . kt 
' ? ?, H. W. Ducklow, E. R. Peele, and S. E. Pike. 1991. Picoplan on 
carbon flux in Chesapeake Bay. Mar. Ecol Prog. Ser. 78: 11-22. 
Mansueti, R. J. 1961. Age, growth and movements of the striped bass, Roccus 
saxatilis, taken in size selective gear. Ches. Sci. 2(1-2):9-36. 
Mansueti, R. J. 1961. 1961 . Movements, reproduction, and mortality of the whtte 
perch, Roccus americanus, in the Patuxent estuary, Maryland. Ches. Scz. 2(3-
4 ): 142-205. 
Marsh, A. G. and K. R. Tenore. 1990. The role of nutrient in regulating the . 
Population dynamics of opportunistic, surface deposit feeders in a mesohaJme 
community. Limnol Oceanogr. 35(3): 710-724. 
Marsha]], H. G. a d K K N . 1996. Phytoplankton composition in relation to 
. n . . es1us. . 
Pnmary production in Chesapeake Bay. Mar. Bzo! 125 
611-617 
? 
Marvin -ni ?p asqua 1e , M. C. and D. G.  Ca pone. 1998 ?. Benthic suldfa te redlusc tion along tr Mar 
the Chesapeake Bay central channel. I. Spatial trends an con ? ? ? 
Ecol. Prog. Ser. 168: 213-228. 
Mass eJ,  S. R. 1999. Hydrodynamics of Coastal z ones. Elsevier Science, New York. 
457 
Miller, D. c R . . 
e ?, . ? 1. Ge1der, and H. L. MacIntyre. 1996. M1crophytobenthos: The 
~01 !
0 gica l role of the "Secret Garden" of unvegetated, shallow-water marine 
iab, tats. Tl. Role in sediment stabi li ty and shallow-water food webs 
Estuaries. l 9(2A): 202-2 12. ? 
Miller R E S . . .. 
' ? ?, . D. Su/km, and R. L. L1ppson. 1975. Compos1t10n and seasonal 
abundance of the blue crab, Cal/inectes sapidus Rathbun, in the Chesapeake 
and De/aware Canal and Adjacent Waters. Ches. Sci. . 16(1 ): 27-3 1. 
Moller, E. F. and T. G. Nie/sen. 2001 . Production of bacterial substrate by marine 
copepods: Effect of phytoplankton biomass and cell size. J Plank. Res. 
23(5): 527-536. 
Monaco, M. E. and R. E. Ulanowicz. 1997. Comparative ecosystem trophic structure 
of three U. S. mid-A tlantic estuaries. Mar. Ecol Prog. Ser. 161: 239-254. 
Montagna, P.A., G. F. Blanchard, and A. Dinet. 1995. Effect of producion and 
biomass of intertidal microphytobenthos on meiofaunal grazing rates. J Exp. 
Mar. Biol. Ecol. 185(2): 149-1 65. 
Najjar, R. 1999. The wa ter balance of the Susquehanna River Basin and its response to 
climate change. J Hy dro!. 219:7-19. 
Namezie, D. A. 1991. Ammonium excretion by gelatinous zooplankton and its role 
in the regeneration of nitrogen in Chesapeake Bay. M. S. Thesis, University of 
Maryland College Park, Maryland. 
Newell , R. I. E. 1988. Ecological changes in the Chesapeake Bay: Are they the 
result of overharvesting the American oyster, Crassostrea virginia. pp. 536-
546 In: M. P. Lynch and E. C. Krome (eds.). Understanding the Estuary: 
Advances in Chesapeake Bay Research. Proceedings of a Conference 29-31 
March 1988. Baltimore, MD. Chesapeake Research Consortium, Publication 
Number 129. 
Nixon, S. W. 1988. Physical energy inputs and the comparative ecology of lake and 
marine ecosystems. Limnol. Oceangr. 33(4, Part 2): 1005-1025. 
North, E. W. 2001. Transport and retention of fish e~rly-life stages i~ Chesapeake 
Bay: mechanisms and implications for recruitment. Ph.D. D1ssertat10n. 
University of Maryland, College Park. 306 p. 
Officer, C. B. 1980. Box models revisited. In P. Hamilton and R. B. ~acdonald 
( eds.). Estuarine and Wet/and Processes. Marine Sciences Sen es. Vol. 11. 
New York: Plenum Press. 
458 
Officer, c B . 
1984 ?,ch? B. Biggs, J. L. Taft, L. E. Cronin, M. A. Tyler, and W. R. Boynton. 
Sci ? esapeake Bay anoxia: Origin, development and significance 
ence. 233: 22-27. ? 
Orth R 
' ? J and K M 
ve ? e t . ? . oore. 1984. Distribution and abundance of submerged aquatic 
g ation in Chesapeake Bay. Estuaries 7( 4B): 531-540. 
Orth , R. J.  and K A M 
g ? ? oore. 1986. Seasonal and year-to-year variations in the 
Browth o f Zostera marina (L.) (Ee lgrass) in the lower Chesapeake Bay. Aqua/. 
O/. 24: 225-341 . 
Orth, R. J 1 
0 .-, : F. Nowak, D. J. Wilcox, J. R. Whiting and L. S. Nagey. 1997. 
tri~tnb~tion of s ubmerged aquatic vegetation in the Chesapeake Bay and 
A utan es and the coastal bays- 1996. US EPA Chesapeake Bay Program, 
nnapoJis, MD. 
Pacheco A 
' ?. L. 1962. Age and growth of spot in lower Chesapeake Bay, with notes 
3o(n distribution and abundance ofj uveniles in the York River system. Ches. Sci 1): 18-28. 
Parsons T . . . 
/ .R. , Y. Maita and C.M. Lalli. 1984. Determmat10n of chlorophylls and 
otaJ carotenoids: Spectrophotometric method. pp. 101-112 Jn: Parsons, T.R. , 
Y. Maita and C.M. Lalli. A manual of chemical and biological methods for 
~ - Pergamon Press, Oxford. 
Pe ters D 
' ?~-and M. T. Boyd. 1972. The effect of temperature, salinity, and 
availability of food on the feeding and growth of the hogchoker, Trinectes 
tnacufatus (Block & Schneider). J Exp. Mar. Bio. Ecol. 7: 201-207. 
Peinert R 
' ?, A. Saure, P. Stegman, C. Steinen, H. Haardt and V. Smetacek. 1982. 
Dynamics of primary production and sedimentation in a coastal ecosystem. 
N eth. J Sea Res. 16: 276-289. 
Poe T p . 
' ? ?, H. Hansel, s. Vigg, D. E. Palmer, and L.A. Pre~er~ast. Feedmg of . 
Predaceous fishes on out-migrating juvenile salmomds m John Day Reservo1r, 
Columbia River. Trans. Am. Fish. Soc. 120(4): 405-420. 
PoJ ? 
ovina, I. J. 1984. An overview of the ECO PATH model. Fishbyte 2(2): 5-7. 
Pra ? ? 
ger, ~ - H. 1996. A simple model of the blue crab, Callinectes sapidus, spawnmg 
migration in Chesapeake Bay. B. Mar. Sci. 58(2): 421-428? 
Pritchard, D. W. 1952. Salinity distribution and circulation in the Chesapeake 
estuarine system. J Mar. Res. . 11(2): 106-123. 
459 
Purcell, 1. E 19 , 
qui ? ~
2- Effects of predation by the scyphomedusan Chrvsaora 
nquec,rrha on I kt I . . ./" 
Eco/ p zoop an on popu at1ons m Chesapeake Bay USA. Mi 
. rog. Ser. 87: 65-76. ' ar. 
Purcell IE 199 . 
'P r d. . 
7. Pelagic cnidarians and ctenophores as predators: Selective 
e at1on feed . te s, an d ti..., 0 , mg ra e iects on prey populations. Anna/es de / Jnstzlut 
ceanographique 73(2): 125-137. 
Purce/J , J. E. 1 . . 
2 / ? R. White, and M. R. Roman. 1994. Predat10n by gelatinous 
c-oop ankton and resource limitation as potential controls ofA cartia tonsa 
opepod populations in Chesapeake Bay. Limno! Oceanogr. 39(2): 263-278. 
Purce]], J. E . 
:;:.; ? and J. H. Cowan, Jr. 1995. Predation by the scyphomedusan Chrysaora 6_uicirrha on Mnemiopsis leidyi ctenophores. Mar. Eco! Prog. Ser. 129: 
Purce/] IE . 
1 and Nemazie, DA. 1992. Quantitative feeding ecology of the 
3Ydromedusan Nemopsis bachei in Chesapeake Bay. Mar. Biol. 113(2): 305-1. 
Purce/] l E . 
' ' White, JR, Nemazie, DA and Wright, DA. 1999. Temperature, salinity 
and food effects on asexual reproduction and abundance of the scyphozoan 
Chrysaora quinquecirrha. Mar. Ecol Prog. Ser. 180: 187-196 
Purce/] J E 
' ? ?, T. A. Shiganova, M. B. Decker, and E. D. Houde. 2001. The 
ctenophore Mnemiopsis in native and exotic habitats: U. S. estuaries versus 
the Black Sea basin. Hydrobio! , i11 press. 
Purce]] MJ.  E?  personal communication. UMCES Hom Point Laboratory, Cambridge, 
D. 
Rabalais ' N?  N ., R. E. Turner, D. Justi , Q. Dortch and W. J.  Wi?s eman, Jr . 1999 . 
Characterization of Hypoxia. Topic 1 Report for the Integrated Asse~s:71ent on 
Hypoxia in the Gulf of Mexico. NOAA Coastal Ocean Program Dec1s10n 
Analysis Series No. 15. National Oceanic and Atmospheric Administration. 
Silver Spring, MD. 
Raym d . d . 
on , P.A . and J.E. Bauer. 2000. Bacterial consumpt10n of DOC urmg 
transport through a temperate estuary. Mar Ecol Prog. Ser. 22:1 -12? 
Reeve, M. R., M. A. Syms, and P. Kremer. 1989. Growth dynamics of a ctenoph?re 
(Mnemiopsis) in relation to variable food supply. I. Carbon biomass, fe~dmg, 
egg production growth and assimilation efficiency. J Plank Res. 11 (3)- 535-
552. ' 
Ricciardi, A. and E. Bourget. 1998. Weight-to-weight conversion factors for marine 
benthic macroinvertibrates. Mar. Ecol Prag. Ser. 163: 245-251 ? 
460 
. . ? and E. D. Houde. J9 99a. Regional and temporal variability in 
Rilling, aGd . CJ?  . n an. a un ance of bay anc.h ovy (Anchoa rnitchiill/ eggs, larvae and 
disut ntb buitoimo ass din  tbhe  Cdh esapeake Bay. Estuaries. 22(4): 1096-1109. 
'a dC . and E. D. Houde 1999b. Regional and temporal variability in growth 
Rilling ;G  mortality of bay anchovy, Anchoa ,nitchilli, larvae in Chesapeake Bay. 
zsh. Bull. 97: 555-569. 
Rizzo, W ? M. and R. L. Wetzel. 1985. [nteru.d al and shoal benth1.c  communi.t y 
me~abolism in a temperate estuary: Stodies of spatial and temporal scaies of 
var iab ?t?i ty. Estuaries. 8(4): 342-351. 'i ? 1 R., A. L. Gauzens, w. [(. Rhinehart, and J. R. White. 1993. Effects of 
Roman oMw  oxygen waters on Chesapeake Bay zoopJanJcton. L. irnnol. Oceanogr. 
'3 8(8? )R: .,1 D60.3 -16H1V. o4l.l iday and L. P. Sanford. 2001. Temporal and spatial patterns 
Roman oMf z ooplankton in the Chesapeake Bay turbidity maximum. Mar. Ec.o l. frog. 
Ser. 213: 2 15-227 117 
Ross ' s ? W. 1988. Age, growth, and Mortality of Atlanti.c  croaker i.n  north Caro\J?n a,_ 
wi Comments on population Dynamics. 1988. Trans. Arn. fJSh. Soc. ? 
th 
Rugolo,4 6L1. -J4.,7 K3..  S. Knotts, A. M. Lange, and V. A. crecco. I 998. Stock assessment 
f Chesapeake Bay bJ ue crab ( Cailinectes sapiduS Rathbun). J. She/lfis h Re.s ? 
0 
Sadzins1k7i (2R):  4p9 3p-?5 17..  d A J k. J9 99 Stock Assessment of Aelected Adult 
Re,s id.,e nt.  Miaigvr1ast,o arny  fin.f isahrz iynn Ms a1r. Yland's?  Chesapeake Bay. US fi?s h an d 
Wildlife Service federal Aid proiect Annual Report. MD Department of 
N , , 
Sa atural Resources, Annapolis, MD- h. . 
gasti, A., L. C S 1 ff d J E D ffy. 2000. Epifauna4l7 c 4om48m7u nities t rive in 
. c ,a ner an . . u . 2 4 . _
an estuary with hypoxic episodes. Estuaries. 3( )? ? 
Sanfo d Ch ke }3ay muds. 
r 'L. P. 1994. Wave-forced resospension of upper esapea 
SAS l ES. uaries 17( 1B ): 148-165. ? 6 4? Edition volume 2. t
nst1tute, Inc. 1990. SAS/ST AT User's Guide, Version , . 
Sanford 1165. . . 19 90 covariability of dissolved P 
'L. P. , K. . G ? Sellner ' and D. L. B. re1tburg. ? ? e Chesapea ke 13ay ? J?  ,A.,A, ar. 
oxygen with physical processes m the summertim 
Res. 48: 567-590. 
46] 
Schwingh 
P adm er'. P., B. Hargrave, D. Peer and C. M. Hawkins. Partitioning of ro uct1on and ? ? ? 
be t . resp1rat10n among size groups of organisms in an intertidal 
n liic community. Mar. Ecol. Prog. Ser. 31: 131-142. 
Seitz inger S p . . . . . 
E ' ? ? 1988. Denitnficat1on m freshwater and coastal marme ecosystems: 
7oc2ological and geochemical significance. Limnol. Oceanogr. 33(4, p~rt 2): -724. 
Schroeder L A C h r.fi . . Th . 1?  . d 1 
' ? ? onsumer growt e1, 1c1enc1es: elf 1m1ts an re at1.0 ns 11 1. ps to 
ecological energetics. J Theor. Biol. 93: 805-828. 
SchubeJ J 
13 ? R. and H. f!? Carter. 1977. Suspended sediment ~udget (or Che~apeake 
ay. In : M. WIiey, ed. Estuarine Processes, Vol II. Clfculat10n. Sediments. 
illJiLI.mnsfer of Material in the Estuary. Academic Press, New York. 
Se/igerB H  ?H  ? and J. A. Boggs. 1988. Long term patterns of an~xia in the Chesapeake 
ay. In : M. Lynch and E. C. Krome (eds.). Understandmg the Estuary: 
~ces in Chesapeake Bay Research. Publication Number 129. 
hesapeake Research Consortium. Solomons, MD. 
Shiah p . . 
' b.-K. ~nd H. W. Ducklow. 1994. Temperature re?ulat10n ofhete~otroph1c 
actenoplank.ton abundance, production and specific growth rate m 
Chesapeake Bay. Limno/. Oceanogr. 39(6): 1243-1258. 
Srnayda ' T ? The suspension and sinking of phytoplankton ? th e  sear . 1-970 m ? 
Oceanogr. Mar. Biol. Ann. Rev. 8: 353-414. 
Srnetacek, V. 1985. Role of sinking in diatom life-history cycles: ecological, 
evolutionary and geological significance. Mar. Biol 84: 239-251 ? 
Snietacek, V., K. v. Briickel, B. Zeitzscel and W. Zenk. 1978. _Sedimentation :f 
plankton diatoms during a phytoplankton spring bloom m re/at,on to th 
hydrographic regime. Mar. Biol. 47: 211-226. 
Smith E 1AK . ? d cell-specific bacterial 
' ? v1. 1998. Coherence of microbial respiratwn rate a~ b 16 (J)? 27-35 
act1? v1?t y m?  a coastal planktonic commurn? tY ? Aqua!?  Micro 
17 
? nco1 ? ? 
Smith E ,. K1 ? ? d I?t coupling to primary ' -1v ? 2000. Factors regulating resp1ratJOn an s . t? University 
Pr o d uctw. n in a coastal planktom.c  communi? ty ? Ph ?D ?  Disserta 10n. 
of Maryland at College Park. College Park, MD. 
Smith s ? k 1991 Tomales Bay 
' ? V., J. T. Hollibaugh, S. J. Dollar, and S. Vm ? h t ? trophy at the la11d-sea 
?1etabolism: C-N-P stoichiometry and ecosystem e ero 
Interface. Est. Coast. ShelfS ci. 33: 22J-257? 
Smith d he roduction/respiration 
' E. M. and W M Kemp 2001. Size strncture an t p , 46(3)? 473-485. 
c?o ast?a balance ina l plan?k ton . ? f. Oceanog, ? ? commurnty. Limno ? 
462 
ay. Ph D . enza ion o oyster habitat in mesohaline Chesapeake 
Smith ' GB A charact ? t' f Pa? rFk , M20D0.1?   issertation. University of Maryland at College Park, College 
0
Smol, N.i,s Ktr ibution .an. d .' ? ? ? ovaere, and A. !. J. Sndee. J9 94. Composition, dN'e th.e Arla. nWd Ii)l iambsw Jm aCs sR o fG m e1obenthos in the Oosterschelde estuary (SW 
S s ? Hydrobiol. 283: I 97-2 J7  
, okal ' RC oR~  a F. J. Rohlf. 1995. BiometrV- I.) !in[EditlQ!!- freeman and nd 
pany, NewYork ==-- -
Stemb growth i d J. J. _G i?l bert. Body size, food concentrat10n, and poopulation erger, R. S. an . . . 
St k n P Ia nktonic rotifers. Ecology. 66(4): I J5J-!159. 
respirat ? nfluences of environmental variables on photosynthetic and 
o es ' MC ?a r1y9 l9a 6 ;(J q uotients m Chesapeake Bay. M.S. Thesis. University of 
0
S . n at Col lege Park. College Park, MD-
analy . ? ? and T. R. Parsons. I 972, A practical handbook of seawater 
S tnckland, J.D H?  u etm 167 (2nd ed.). J. Fish. Res- Bd. Can. Ottawa, Canada. 
SIS B II . 
au;ot ?, P. Nilsson, C. Nilsson, and B. Jonsson. 1996. Balance between 
undbacckom K r ophic .a nd heterotrophic components and processes in microbenthic 
68 -~;,ities of sandy sediments: A field study. Est. Coast. Shelf Sci. 43: 
9
MR. A. and J. Davis. J9 72. Age, growth, and mortality of the white perch, 
St. Pierr2e ~~~~~-americana, in the James and York Rivers, Virginia. Ches. Sci. 13(4): 
7
A~ al ? a P ? Colla. 1995. CART: Tree-Structured Non-Parametric Data 
Steinb urg o nd 
ysis. San Diego, CA. Salford Systems. 
' . .J, s RI  a~Cd  ? Al ler and C. Lee. I 991. Early diiagenesis of ch JorophY JI -a i.n  L_ ong J. 
Su n MM a Sound sediments: A measure of carbon flUJ( and particle reworking. ? 
S r. cR es. 49: 279-401. un, M.d-eY  ?, ? Lee and R. C. Aller. J9 93a. Laboratory studi?e s of oxi?c  an d an oxic 
c gradation of chlorophyll-a in Long Island sound sediments. Geochim 
SU n, M.-yo mochim Acta. 57: I 47 -157. . . f i?c-
lab., C. Lee and R. C. Al ier. J9 9 3b . Anox ic  and ox1c degradat10n o nd 
s deled chloropigments and a "c-Jabeied dia1to4m51 1n Long JsJand sou 
e iments. Limnol. Oceanogr. 38(7): \438- ? 
463 
Sun, M.-Y., R. C. Aller and C. Lee. 1994. Spatial and temporal distri~utions of 
sed imentary ch loropigments as indicators of benthic processes m Long Island 
Sound. J. Mar. Res. 52: 149-176. 
Szyrrner, J. and R. E. Ulanowicz. 1987. Tootal flows in ecosystems. Ecol. Modeling. 
53: 123-136. 
Taft, .J. L., A. J. Elliott, and W.R. Taylor. 1978. Box model analysis o~?chesapeake 
Bay ammonium and nitrate fluxes. In Wiley, M. L. (ed.), Estuarine 
Interactions. New York: Academic Press. 
Taft, J . L., E. 0. Hartwig, and R. Loftus. 1980. Seasonal oxygen depletion in 
C hesapeake Bay. Estuaries. 3: 242-247. 
Tang, E. P. Y. and R.H. Peters. 1995. The allometry of algal respiration. J. Plank. 
Res . 17(2): 303-315. 
Tarnowski, M. L., M. Homer, and L. Baylis. 1994. The effects of the 1993 freshet on 
C hesapeake Bay oyster populations. In: Hill, P. and S. Nelson, (eds.). Toward 
a Sustainable Watershed: The Chesapeake Experiment. Proceedings of the 
1994 Chesapeake Research Conference. Chesapeake Research Consortium, 
Publication Number 149. 
Thompson, M . L. 2000. Demographics, production and benthic-pelagic coupling by 
the suspension-feeding polychaete Chaetopterus pergamentaceus in the lower 
Chesapeake Bay. Ph. D. Dissertation. Virginia Institute of Marine Science, 
Gloucester Point, VA. 
Townsend, D. W. and L. M. Cammen. 1988. Potential importance of the timing of 
spring plankton blooms to benthic-pelagic coupling and recruitment of juvenile 
demersal fishes. Biol. Oceanogr. 5: 215-229. 
Tucker, J. W. , Jr. 1989. Energy utilization in bay anchovy, Anchoa mitchilli, and 
black sea bass, Centropages striata striata, eggs and larvae. Fish. Bull. U.S. 
78: 279-293. 
Tumbiolo, M. L. and J. A. Downing. 1994. An empirical model for the prediction of 
secondary production in marine benthic invertebrate populations. Mar. Ecol. 
Prag.Ser. 114: 165-174. 
U lanowicz, R. E. 1986. Growth and Development: Ecosystems Phenomenology. 
Springer-Verlag. 
U Janowicz, R . E. 1995. Ecosystem trophic foundations: Lindeman exonerata. pp. 549-
560 in B .C. Patten and S. Jorgensen (eds.) Complex Ecology: The Part-Whole 
Relation in Ecosystems. Prentice-Hall, Englewood Cliffs, New Jersey. 
464 
p ? E. l 997 Ecolo The Ascendent pers ective. Columbia University 
Ulano WI.C rZe,s sR,  Ne. w y o~k=. = ~._!..!~~~:!S!!lD~~~ 
Ulano A. n;ly ? \,1999. _NETWRK 4.2a: A Package of Computer Algorithms to WJCnZv iRr z e cological Flow Networks. University of Maryland Center for 
82- '2;'{ntal Science. Chesapeake Biological Laboratory. uMCES Ref No. 
A,; 7 ? ?a nd W.M. Kemp. 1979. Toward canonical trophic aggregations. 
Ulano WI.C Z RE . 
er. Nat. I 14:871-883. 
e ? ? and D. Baird. I 986. A network analyS1s of the Chesap1e5ake Bay 
Ulano Wu.IC  Z?,  R E_m . aryland.  Department of Natural Reso.u cosyste M rces, Annapo 1? , MD ? 
c;'LverS1ty of Maryland Center for Environmental Science- REF [UMCEES] 
st
86-79. 
C ' ? ? and C. l. Puccia. !990. Mixed trophic impacts in ecosy ems. 
Ulano WJ.C Z R 
V oenoses 5(1): 7-16. 
e:~~son, A. and K.M. M. Lwiza- !995. The effects of channels and shoals on 
alle-LRe  . ,ange between the Chesapeake Bay and the adjacent ocean. J Geophys. 
Va ughn, eDs . SI  OJO (C9): I 8,551-18,563. 200lb Analysis on 
th ? ?, ? W. Smith, E. H Williams, and JV!. H Prager. ? 
p ?e  ta tus f the Atlantic menhaden stock. NOAA Center for cooaSll1l, . 
~sheries aOnd Habitat Research, aeaufort, NC. Report to the Atl?tc ;1a~eio1 
arme Fisheries Commission Menhaden Technical Committee? u11ne 1'  ? 
Verity 'rp??  G. 5, Grazing respiration excretionandgrowthratesoftm? ?? "?d s. 
198
Ve n.t y, P.1 mG n oI?  Oceanogr. 3' 0(6): !268-1' 282. . ke b marine planktoni.c  
c i I? ? 991. Measurement and sim ul at1on of prey upta 1 ~ceanogr. 36(4 ) : 
0
'ate1s  fed plastidic and aplastidic nanopiankt n? L,rnno ? 
v?u nstei7n2 9-750. d fishes on benth!C.  
i ' R. W ? 1977. The importance of predation1 9b9y 1c2ra1b7s an 
\Val Eco/og)I? . rcatio8ns for ters,n Cfa u2na  in Chesapeake Bay. 5. 8: 1 ?d-ance ?t actics:0 unP t . "09-
.. 000. Natural selection for predauon avoi E I p . g. Ser. 20 ?) 
mann . . d 1. s ]vfar. co. r 
3 _ e populauon and commumty ynam c ? 
\Vang, H1 3 . . lication of the seeond 
. V. and B. H. Johnson. 2000. Vahdat1on_ and apP of Chesapeake BaY?. 
~eneration three-dimensional hydrodynarn1c model ]Jodeling. In revieW? 
ubmitted to Journal of Water QualilY and Ecosystem 
Wang, H. V. C. and S.-Y. Chao. 1996. Intensification of subtida l surface currents 
over a deep channel in the upper Chesapeake Bay. Est. Coast. ShelfS ci. 42: 
77 1-785. 
Wang, S.-B. and E. D. Houde. 1994. Energy storage and dynamics in bay anchovy 
Anchoa mitchilli. Mar. Biol. 121: 2 19-227. 
Wang, S .-B. and E. D. Houde. 1995. Distribution, relative abu ndance, biomass and 
production of bay anchovy Anchoa mitchilli in the Chesapeake Bay. Mar. 
Ecol. Prag. Ser. 12 1: 27-38 . 
Webb, E. HI , and E. Zlokovitz. 1999. Assessment of the 1998 recreational fin fish 
harvests in the Choptank and Chester Rivers and summer head boat surveys. 
US Fish and Wildlife Service, Federal Aid Project Annual Report. MD 
Department of Natural Resources, Annapolis, MD. 
Webster, J. T., Par low, J. S. , and S. V. Smith. 2000. Implications of spatial and 
temporal variation for biogeochemical budgets of estuaries. Estuaries. 23(3): 
341-350. 
We isburg, S. B., J. A . Ranasinghe, L. C. Schaffner, R. J. Diaz, D. M . Dauer, and J.B. 
Frithsen. 1997. An estuarine benthic index of biotic integrity (B-IBI) for 
Chesapeake Bay. Estuaries. 20(1): 149-158. 
Wenner, C . A. and J . A . Musick. 1975. Food habits and seasonal abundance of the 
American eel, Anguilla rostrata, from the lower Chesapeake Bay. Ches. Sci. 
16( I): 62-66. 
Wetzel, R. L. and P.A. Penhale. 1983. Production ecology of seagrass communities 
in the lower Chesapeake Bay. Mar. Tech. Soc. J. 17(2): 22-31. 
Widbom, B. 1984. Determination of average individual dry weights and ash-free dry 
weights in different sieve fractions of marine meiofauna. Mar. Biol. 84: JO J-
108. 
White, J. R . and M. R. Roman. l 992. Egg production by the calanoid copepod 
Acartia tonsa in the mesohaline Chesapeake Bay: the importance of food 
resources and temperature. Mar. Ecol. Prag. Ser. 86: 239-249. 
Za itsev, Y. P. Recent changes in the trophic structure of the Black Sea. Fish. 
Oceangr. 1( 2): I 80-189. 
Zajac, R. N . and R. B. Whitlatch. 2001. Response ofmacrobenthic communities to 
restoration efforts in a New England estuary. Estuaries. 24(2) : 167-183. 
466