ABSTRACT 
 
Title of Dissertation: ALTERNATE STATE THEORY AND 
TIDAL FRESHWATER MUDFLAT 
EXPERIMENTAL ECOLOGY ON 
ANACOSTIA RIVER, WASHINGTON, 
D.C. 
 
  
 
Peter Ian May, Doctor of Philosophy, 2007 
  
Dissertation directed by: Patrick Kangas 
Marine Estuarine Environmental Sciences 
 
The concept that multiple community states may alternately exist for some 
ecosystems has been the subject of controversy for decades.  This theory is tested and 
applied to the mudflats of the low/middle marsh intertidal zone of two restored 
freshwater tidal marshes on the Anacostia River.  It is believed that experimental 
exclosures exposed strong species interactions and provided a window with which to 
view the potential alternate existence of two structurally different systems, intertidal 
mudflat and emergent marsh.  The occurrence, persistence and community 
composition of the two ecosystem states are examined through experimental 
exclosures at the two marsh restoration study areas. 
  
The power of large grazers to deflect the goals of wetland restoration 
practitioners is studied in the context of alternate state theory. Initially unvegetated 
mudflat, native marsh vegetation emerged within exclosure study areas at two 
restoration sites. Resident Canada geese (Branta canadensis maxima) decimated 
planted areas of restored marsh left open to grazing, returning marsh to unvegetated 
mudflats.  Data from exclosures are presented on macrophyte community 
composition, sediment elevation, bird, fish, invertebrate and algae associations from 
two separate sets of Anacostia River experimental exclosure sites, one covering     
588 m2, the other covering 2,700 m2.  Results support the hypothesized alternate 
existence of the two system states in the same space and relative time, each dependant 
upon the access of a critical mass of large grazers.  A description of the mudflat biotic 
community and its interconnectivity is discussed as an important feature of the 
Anacostia River system. 
An emergy analysis of each state and an accounting of fisheries energy flow is 
conducted. Information collected relating to the pre-restoration (tidal mudflat) and 
post-restoration (emergent marsh) physical and biological conditions are detailed and 
analyzed.  A determination of the emergy inputs for a large-scale marsh restoration 
project are calculated and as a final analysis, economic (emdollar) equivalents are 
developed to compare the yield of fisheries production supplied by mudflats vs. a 
restored and mature emergent freshwater tidal marsh.  Through these studies support 
is given to valuing mudflats as important system components of Anacostia River.  
  
 
ALTERNATE STATE THEORY AND TIDAL FRESHWATER 
MUDFLAT EXPERIMENTAL ECOLOGY ON ANACOSTIA 
RIVER, WASHINGTON, D.C. 
 
 
By 
 
Peter Ian May 
 
 
Dissertation submitted to the Faculty of the Graduate School of the  
University of Maryland, College Park, in partial fulfillment 
of the requirements for the degree of 
Doctor of Philosophy 
2007 
 
 
 
 
 
Advisory Committee: 
Patrick Kangas, Chairman / Advisor 
Professor Andy Baldwin 
Dr. Richard Hammerschlag 
Professor J. Court Stevenson 
Professor David Tilley 
 
  
 
 
 
 
 
 
 
 
 
? Copyright by 
Peter Ian May 
2007 
 
 
 
 
 
 
 
 
 
 
ii 
 
Dedication 
 
 
To all my family 
 
To the Anacostia River 
 
 
iii 
 
Acknowledgements 
This dissertation would not have been possible without the lifelong love and 
support of my family. Thank you Mom, Dad, Chris, Klaus, Omi, Myra, Sarah, all the 
cousins, Mays and Spangenbergs.  
My mentor and advisor, Dr. Patrick Kangas, has been a significant influence 
on me academically and philosophically for many years. His encouragement helped 
to keep myself positive during some difficult times and my many travels throughout 
the world with him opened my eyes and will sustain me a lifetime.   
I was encouraged to enter the program by Kristen Hughes, without her love 
and support I would not have tried to reach this far and to her I will always be 
grateful.   
My years working at the District of Columbia Watershed Protection Division 
were some of the most rewarding for the friendships and professional support of my 
academic research on the Anacostia River.  Dr. Hamid Karimi, Sheila Besse, Jeffry 
Zahn, Nicoline Shultebrandt, Massoud Massoumi, Jerusalem Bekele and the crew of 
D.C. Fisheries and Wildlife were all important to me in many ways.  So many others 
helped me with my field work.  Kerry Manire and I survived, I will always give 
thanks. Stephen Syphax of the National Park Service, National Capital Parks East 
influenced my interest in the social and ecological history of the Anacostia River and 
its residents.  Dr. Dick Hammerschlag of the U.S. Geological Survey Patuxent 
Wildlife Research Center helped me significantly with my Kingman Marsh research 
and was a mutual friend of the Anacostia.   
 
 
iv 
 
The final push to complete this dissertation was supported immensely by Erin 
Leone for her help with statistics.  My lifelong friend Tom Juhasz gave me a long 
weekend to remember in helping me organize my final draft.  Chris Streb, a friend 
from the early grad school days of ?GEES? gave me 24 hours of his life to finalize 
the frustrations of formatting. Friends Bryon Salladin and Jeremy Thomas helped me 
out of tight spots. Biohabitats gave me the time and support to complete the work. 
Ginger Chapman cheered me on at all the right times. 19-Q and Old Greenbelt gave 
me solitude as my refugia from the distractions of life. 
 
 
v 
 
Table of Contents 
List of Tables .............................................................................................................. ix 
List of Figures .............................................................................................................. xi 
Chapter 1: Introduction........................................................................................... 1 
1.1 Background and Significance ........................................................... 1 
1.2 History of Tidal Freshwater Emergent Marshes and Mudflats on 
Anacostia River......................................................................................................... 3 
1.3 The Use of Exclosures as a Method of Exposing Alternate States in 
Intertidal Habitat ....................................................................................................... 6 
1.4 Research Questions and Hypotheses ................................................ 8 
1.5 Organization of the Thesis.............................................................. 10 
Chapter 2: Literature Review................................................................................ 11 
2.1 Alternate Stable State Theory ......................................................... 11 
2.2 Tidal Freshwater Mudflats.............................................................. 13 
2.3 Tidal Freshwater Aquatic Community Interconnectedness............ 14 
2.4 Exclosures, Tidal Flats and Emergent Marsh Herbivory................ 16 
Chapter 3: Kenilworth Marsh Experimental Ecology .......................................... 19 
3.1 Introduction..................................................................................... 19 
3.2 Study Area Location Description.................................................... 24 
3.3 Research Methods and Materials.................................................... 25 
3.3.1 Birds............................................................................................ 25 
3.3.2 Fish.............................................................................................. 25 
3.3.3 Exclosures................................................................................... 27 
3.4 Data Analysis Methods................................................................... 32 
3.5 Results............................................................................................. 34 
 
 
vi 
 
3.5.1 Birds............................................................................................ 34 
3.5.2 Fish.............................................................................................. 38 
3.5.3 Fish Survey Site Comparisons.................................................... 44 
3.6 Exclosures....................................................................................... 48 
3.6.1 1997 Invertebrates....................................................................... 48 
3.6.2 1998 Invertebrates....................................................................... 51 
3.6.3 Algae........................................................................................... 53 
3.6.4 Emergent Macrophytes ............................................................... 57 
3.7 Discussion....................................................................................... 62 
3.7.1 Bird and Fish Predation on Mudflat Invertebrates...................... 62 
3.7.2 Algae Cover and Invertebrate Densities ..................................... 65 
3.7.3 Mudflat Invertebrate Population Crash?..................................... 67 
3.7.4 Strong Fish Influence Over Invertebrates................................... 68 
3.7.5 Algal Grazer/Disturbance Activity ............................................. 71 
3.7.6 Response of Emergent Macrophytes to Birds and Fish.............. 72 
3.7.7 River Turtle Influence................................................................. 74 
3.7.8 Mollusks and Mammals on the Mudflats.................................... 75 
3.7.9 The Most Interesting Animal...................................................... 75 
3.8 Conclusions..................................................................................... 76 
Chapter 4: Kingman Lake/Marsh Experimental Ecology .................................... 79 
4.1 Introduction..................................................................................... 79 
4.2 Study Area Description................................................................... 86 
4.3 Research Methods and Materials.................................................... 87 
4.3.1 Birds............................................................................................ 87 
4.3.2 Fish.............................................................................................. 88 
 
 
vii 
 
4.3.3 Exclosures................................................................................... 88 
4.4 Data Analysis Methods................................................................... 91 
4.5 Results............................................................................................. 93 
4.5.1 Birds............................................................................................ 93 
4.5.2 Fish.............................................................................................. 95 
4.5.3 Exclosure Macrophytes............................................................... 96 
4.6 Presence / Height .......................................................................... 109 
4.7 Discussion..................................................................................... 114 
4.8 Conclusion .................................................................................... 124 
Chapter 5: Emergy Analysis On Kingman Fisheries Production ....................... 126 
5.1 Introduction................................................................................... 126 
5.2 Operational Definitions................................................................. 130 
5.2.1 Defining Emergy....................................................................... 130 
5.2.2 Defining Transformity .............................................................. 130 
5.2.3 Defining Emdollars................................................................... 131 
5.2.4 Kingman Mudflats and Marsh Primary and Fisheries Production 
Emergy Analysis............................................................................................... 131 
5.2.5 Energy Systems Diagram and Analysis Table.......................... 133 
5.2.6 Renewable Resources ............................................................... 137 
5.2.7 Inputs to the Kingman Marsh Restoration................................ 138 
5.2.8 Exports from the Kingman System........................................... 139 
5.2.9 Kingman Marsh Area................................................................ 141 
5.2.10 Emergy Indicies Calculation................................................... 142 
5.3 Fisheries Emergy Yield Ratio....................................................... 142 
5.4 Results........................................................................................... 143 
 
 
viii 
 
5.4.1 Emergy Analysis Table............................................................. 143 
5.4.2 Renewable Energies and Production Exports........................... 146 
5.4.3 Fisheries Production and Export............................................... 146 
5.4.4 Catfish and Emergy................................................................... 148 
5.4.5 No Catfish Emergy Scenarios................................................... 150 
5.4.6 Total Annual Emergy Exported (Without Catfish)................... 151 
5.4.7 Emergy Indices ......................................................................... 152 
5.5 Discussion..................................................................................... 154 
5.6 Conclusion .................................................................................... 156 
5.6.1 Fisheries use of Tidal Freshwater Mudflats.............................. 156 
5.6.2 Emergy and Alternate State Theory Informing Restoration..... 159 
Chapter 6: Conclusion ........................................................................................ 174 
6.1 Alternate Ecosystem State Changes on Anacostia River.............. 174 
Appendices................................................................................................................ 181 
Appendix A ? Kenilworth Marsh Ecological Data Tables ....................................... 182 
Appendix B ? Kingman Lake Ecological Data Tables............................................. 219 
Appendix C ? Kingman Fisheries Emergy Analysis Tables .................................... 317 
Appendix D ? Examples of SAS Code..................................................................... 332 
Literature Cited ......................................................................................................... 336 
 
 
 
ix 
 
List of Tables 
Table 1.1 Anacostia River timeline of human involvement in changes to the 
ecosystem.......................................................................................................... 4 
Table 3.1 Kenilworth MF3 mudflat bird surveys 1997-1999 listed in order of 
importance value (relative abundance + frequency / 2).   Functional feeding 
groups are: SB, Shorebirds; WB, Wading birds; DDG, Dabling ducks/geese; 
SC, Scavengers; R, Raptors. ........................................................................... 36 
Table 3.2 Kenilworth MF3 mudflat study area bird functional feeding groups. ........ 36 
Table 3.3 Kenilworth Mass Fill 1 marsh fish community from 8 electrofishing 
surveys; absolute abundance (total # caught); average #/survey; relative 
abundance (% total #); frequency (% survey sp. appearance); ranked by 
importance value (relative abundance + frequency / 2).................................. 40 
Table 3.4  Kenilworth Mass Fill 2 marsh fish community from 8 electrofishing 
surveys; absolute abundance (total # caught); average #/survey; relative 
abundance (% total #); frequency (% survey sp. appearance); ranked by 
importance value (relative abundance + frequency / 2).................................. 41 
Table 3.5  Kenilworth Mass Fill 3 mudflat fish community from 4 electrofishing 
surveys; absolute abundance (total # caught); average #/survey; relative 
abundance (% total #); frequency (% survey sp. appearance); ranked by 
importance value (relative abundance + frequency / 2).................................. 42 
Table 3.6  Dueling Creek marsh fish community from 5 electrofishing surveys; 
absolute abundance (total # caught); average #/survey; relative abundance (% 
total #); frequency (% survey sp. appearance); ranked by importance value 
(relative abundance + frequency / 2). ............................................................. 43 
Table 3.7  Kenilworth Mass Fill 3 mudflat fish community from 4 bag sein surveys; 
absolute abundance (total # caught); average #/survey; relative abundance (% 
total #); frequency (% survey sp. appearance); ranked by importance value 
(relative abundance + frequency / 2). ............................................................. 44 
Table 3.8  Percent (%) similarity matrix comparing common fish species collected 
between sites from all years sampling events. Similarities read as far left 
column % species similarity to next site from left to right............................. 45 
Table 4.1 Kingman bird surveys 1997-2001 listed in order of importance value 
(relative abundance + frequency / 2) .............................................................. 94 
Table 4.2  Kingman Lake mudflat fish community from 6 electrofishing surveys; 
 
 
x 
 
absolute abundance (total # caught); average #/survey; relative abundance (% 
total #); frequency (% survey sp. appearance); ranked by importance value 
(relative abundance + frequency / 2). ............................................................. 95 
Table 4.3  20X exclosure plot mean biomass 2001-20004....................................... 106 
Table 4.4  10X exclosure plot mean biomass 2001-20004....................................... 106 
Table 4.5   Planted Species at Kingman Marsh Restoration compared to those found 
in exclosure area transects. ........................................................................... 109 
Table 4.6  Comparison of Kingman mudflat exclosure plant community attributes 
over four survey years. Averages are arithmetic means combining both plot 
sizes across all elevations.............................................................................. 120 
Table 5.1  Kingman Marsh emergy analysis table.................................................... 136 
Table 5.2 Emergy per mass of river geopotential and river sediments..................... 138 
Table 5.3 Emergy indices for Kingman Marsh restoration....................................... 153 
Table 5.4 Summary calculations of secondary production for middle and upper level 
fish consumers on Anacostia marsh and mudflat sites ................................. 171 
 
 
 
xi 
 
List of Figures 
Figure 1.1  Venn diagram of the two dominant intertidal system states on Anacostia 
River with the alternate potential for either state to occur................................ 2 
Figure 1.2 Vicinity map of exclosure study areas on Anacostia River......................... 7 
Figure 2.1 Schematic of a typical exclosure experimental arrangement. ................... 17 
Figure 3.1 United States Army Corps of Engineers 1891 Map of Anacostia River 
Marshes........................................................................................................... 19 
Figure 3.2 Kenilworth Marsh and Dueling Creek study areas.................................... 21 
Figure 3.3 Anacostia River typical tidal hydrograph.................................................. 22 
Figure 3.4 Schematic of Kenilworth MF3 mudflat exclosure treatment types........... 27 
Figure 3.5. Arrangement of Kenilworth MF3 mudflat exclosure treatment blocks. .. 28 
Figure 3.6 Kenilworth MF3 mudflat bird observations 1997-1999............................ 34 
Figure 3.7 Kenilworth MF1 and MF2 combined electrofishing survey results.  Fish 
averages arithmetic means + SE. .................................................................... 46 
Figure 3.8  Kenilworth MF3 combined electrofishing survey results.  Fish averages 
arithmetic means + SE. ................................................................................... 46 
Figure 3.9  Dueling Creek Marsh electrofishing survey results.  Fish averages are 
arithmetic means + SE. ................................................................................... 47 
Figure 3.10  1997 Invertebrate group percent composition within each exclosure 
treatment type.  (AA) all accessible control treatment; (FA)  fish accessible/ 
bird exclosure treatment; (BA) bird accessible/ fish exclosure treatment; (TX) 
total exclosure treatment................................................................................. 49 
Figure 3.11 Invertebrate densities by year and exclosure type.  Densities are Tukey 
adjusted LS means + SE.  Mean sharing any com on letter are not significantly 
different (p < 0.05). Letter case differentiates years. Means with no letters are 
also significantly different. (AA) all accessible control treatment; (FA) fish 
accessible/bird exclosure treatment; (BA) bird accessible/fish exclosure 
treatment; (TX) total exclosure treatment....................................................... 50 
Figure 3.12  Invertebrate group percent composition within each exclosure treatment 
type. (AA) all accessible control treatment; (FA) fish accessible/bird exclosure 
treatment; (TX) total exclosure treatment....................................................... 52 
 
 
xii 
 
Figure 3.13  1997?1998 combined algae mean percent cover. Means are Tukey 
adjusted LS means + SE. Means with different letters are significantly 
different (p < 0.05). Means sharing the same letter are not significantly 
different. (AA) all accessible control treatment; (FA) fish accessible/bird 
exclosure treatment; (BA) bird accessible/fish exclosure treatment; (TX) total 
exclosure treatment. ........................................................................................ 55 
Figure 3.14  Figure 3.12. 1997?1998 combined algae mean cover by season. Cover 
are Tukey adjusted LS means + SE. Means not sharing common letters are 
significantly different (p < 0.05). Means sharing common letters are not 
significantly different...................................................................................... 56 
Figure 3.15  View of 1997 total exclosure (TX) treatment plot with young Pontederia
......................................................................................................................... 58 
Figure 3.16  View of 1997 mid-summer exclosure plot Pontederia cordata and 
Polygonum ...................................................................................................... 58 
Figure 3.17  1997 Kenilworth Mudflat exclosure Polygonum sp. volunteerism 
grouped by treatment block.  A B and C are replicate treatment blocks. ....... 59 
Figure 3.18  1997 Kenilworth mudflat exclosure Pontederia cordata volunteerism 
grouped by treatment block.  A B and C are replicate treatment blocks. ....... 59 
Figure 3.19  1998 Kenilworth mudflat exclosure treatment Pontederia cordata 
experimental planting survivorship................................................................. 61 
Figure 3.20  1998 Kenilworth mudflat Pontederia cordata experimental planting 
above ground biomasss LS means + SE. Different letters denote significant 
difference in biomass (p < 0.0001). The AA treatment type was significantly 
different than FA or TX with no variance within the exclosure treatment 
between the AA treatment blocks................................................................... 62 
Figure 3.21  Generalized interactions between components of the intertidal mudflat 
exclosure area at the Kenilworth MF3.  Thin layers are likely interactions 
(sometimes with intermediaries not labeled) that were not directly observed or 
revealed through the exclosures.  Bold lines are interactions that were directly 
observed on the mudflat or revealed through exclosures................................ 64 
Figure 3.22  Water levels depicting higher levels in 1998 over 1997. ....................... 67 
Figure 3.23  Exclosure filled with Pontederia cordata. The exclosure to the right is 
fish accessible. ................................................................................................ 77 
Figure 3.24  Goose tracks covering the mudflat study area at Kenilworth Mass Fill 3.
......................................................................................................................... 78 
 
 
xiii 
 
Figure 4.1  Nineteenth-century image of rail hunting on Anacostia River wild rice.. 80 
Figure 4.2 U.S. Army Corps of Engineers 1913 map of existing and planned 
Anacostia River shorelines. ............................................................................ 80 
Figure 4.3  Locations of Anacostia River marsh restoration projects with year of 
completion and Dueling Creek reference area. Kingman Marsh Cell 1 
restoration area at the center bottom. Image from USGS Patuxent Wildlife 
Research Center, 2003. ................................................................................... 82 
Figure 4.4  Upper Kingman Marsh Cell 1 restoration area with study site locations. 84 
Figure 4.5  Kingman Lake/Marsh Canada goose counts showing increasing numbers 
found on the mudflats, open water, and golf course around Upper Kingman 
Cell 1............................................................................................................... 85 
Figure 4.6 Kingman Marsh exclosure study plots with corner surface elevation points.
......................................................................................................................... 89 
Figure 4.7  Kingman Exclosure plot area exaggerated sediment surface elevations.. 92 
Figure 4.8  Kingman mudflat electrofishing survey results........................................ 96 
Figure 4.9  Kingman plot exclosures.  Direction of generally increasing elevation... 97 
Figure 4.10  Kingman exclosure plot total cover arithmetic means and .................... 98 
Figure 4.11  Kingman combined 10x exclosure plot arithmetic mean cover by species 
across years with 1 standard deviation.......................................................... 100 
Figure 4.12  Kingman combined 20x exclosure plot arithmetic mean cover by species 
across years with 1 standard deviation.......................................................... 100 
Figure 4.13  2001 combined exclosure plot arithmetic mean cover by species across 
mudflat surface elevation midpoints with 2 standard deviations. Species are 
represented unless there was no cover.......................................................... 102 
Figure 4.14   2002 combined exclosure plot arithmetic mean cover by species across 
mudflat surface elevation midpoints with 2 standard deviations. Species are 
represented unless there was no cover.......................................................... 102 
Figure 4.15  2003 combined exclosure plot arithmetic mean cover by species across 
mudflat surface elevation midpoints with 2 standard deviations. Species are 
represented unless there was no cover.......................................................... 103 
Figure 4.16  2003 combined exclosure plot arithmetic mean cover by species across 
mudflat surface elevation midpoints with 2 standard deviations. Species are 
 
 
xiv 
 
represented unless there was no cover.......................................................... 103 
Figure 4.17  Combined 10x exclosure plot biomass by species across years. Data are 
arithmetic means with 1 standard deviation.................................................. 105 
Figure 4.18 Combined 20x exclosure plot biomass by species across years. Data are 
arithmetic means with 1 standard deviation.................................................. 105 
Figure 4.19 Combined 10x exclosure plot biomass for all species by year  over 
elevation. Year lines are biomass continuum values of predicted LS means + 
SE from analysis of covariance. Points along the lines are biomass values 
from predicted means within the low to middle marsh zone represented by the 
data................................................................................................................ 108 
Figure 4.20 Combined 10x exclosure plot biomass for all species by year  over 
elevation. Year lines are biomass continuum values of predicted LS means + 
SE from analysis of covariance. Points along the lines are biomass values 
from predicted means within the low to middle marsh zone represented by the 
data................................................................................................................ 108 
Figure 4.21 2001 species presence along exclosure plot transect lines and each 
species height in relation to the exaggerated surface elevation of the plots. 111 
Figure 4.22 2002 species presence along exclosure plot transect lines and each 
species height in relation to the exaggerated surface elevation of the plots. 111 
Figure 4.23 2003 species presence along exclosure plot transect lines and each 
species height in relation to the exaggerated surface elevation of the plots. 113 
Figure 4.24 2004 species presence along exclosure plot transect lines and each 
species height in relation to the exaggerated surface elevation of the plots. 113 
Figure 4.25  View in 2002 southeast across Kingman exclosure plots in the direction 
of generally ................................................................................................... 114 
Figure 4.26 View northeast across the unvegetated mudflat of an unfenced control 
plot. Fence lines of........................................................................................ 115 
Figure 5.1 Kingman Marsh restoration energy circuit diagram detailing renewable 
energy flows, inputs from the economy for restoration, and system exports 
from the mudflat/open water and emergent marsh systems.......................... 134 
Figure 5.2  Simplified energy circuit diagram illustrating the basic flow of energies 
used to calculate the emergy indices............................................................. 142 
Figure 6.1 Human involvement in Anacostia River system state changes over 400 
years. ............................................................................................................. 175 
 
 
xv 
 
Figure 6.2 Representation of the dual influences  of grazing pressure and sediment 
surface elevation in determining the potential extent of mudflats (Kangas et 
al., 2004). ...................................................................................................... 177 
 
 
 
 
1 
 
Chapter 1: Introduction 
Restoration is the ultimate test of ecological theory. 
J.J. Ewel, 1987 
1.1 Background and Significance 
The mudflats of the tidal freshwater Anacostia River in Washington, D.C. are 
the focus of complimentary studies that may provide evidence of two alternate 
community states, intertidal mudflat and emergent marsh, existing within the same 
space and relative time (Figure 1.1).  The theory of alternate or multiple stable 
community states (detailed in the following literature review) may be an appropriate 
framework for attempting to explain, in part, the resilience of mudflats on the 
Anacostia River for over half a century and the tendency of recent emergent marsh 
restoration projects to be altered or confounded. 
The purpose of this study is to define several of the basic ecological elements 
and forcing functions of the mudflat system, and determine the potential for the low-
middle marsh elevation zones on mudflats of Anacostia River to support two alternate 
states, one emergent marsh and the other unvegetated intertidal mudflat. The 
experimental use of exclosures as an artificial switching function between the two 
states is key to determining this potential.  The mudflats of the low-middle marsh 
zone have the potential to support emergent marsh vegetation under certain 
conditions and are here after referred to as biogenic mudflats.  Biogenic mudflats are 
a part of the greater mudflat system on the Anacostia River, in addition to those 
 
 
2 
 
intertidal mudflats existing below the low-marsh elevation gradient which due to 
physical conditions lack the ability to support emergent marsh vegetation, and from 
here after are referred to physical mudflats.  Where both types are considered 
together, mudflats or mudflat system will be used. 
 
Figure 1.1  Venn diagram of the two dominant intertidal system states on Anacostia River with 
the alternate potential for either state to occur.  
 
The results of biogenic mudflat exclosure field experiments implemented 
between 1997-1998 and 2001-2004 at two Anacostia marsh restoration sites are used 
to examine: (1) the basic ecological interactions that characterize the entire mudflat 
system; (2) the applicability of the alternate community state model to the Anacostia 
biogenic mudflat and restored marsh systems; (3) the switching mechanisms of 
change between the two community types; and (4) the patterns of development of 
 
 
3 
 
emergent marsh plant communities under strong herbivory pressure with and without 
exclosures.   
Additionally, an emergy analysis (detailed in Chapter 5) of the resource flows 
and forcing functions of the two community states, mudflat and restored emergent 
marsh, is employed with a conversion of restoration emergy to its economic 
equivalent in emdollars (Odum, 1996).  Based on the significant investment of 
resources from the economy into converting tidal mudflats into emergent marsh, an 
emergy analysis of the two alternate states is applied to one of the large-scale marsh 
restoration efforts completed on Anacostia River mudflats. 
1.2 History of Tidal Freshwater Emergent Marshes and Mudflats on Anacostia 
River 
The Anacostia River has been changed significantly by direct and indirect 
human involvement over the last 400 years (Table 1.1). Historically, the Kenilworth 
and Kingman marshes supported expansive areas of emergent freshwater tidal 
wetlands dominated by Zizania aquatica (Coues and Prentiss, 1883; Biohabitats, 
1990; Syphax and Hammerschlag, 1995).  These wild rice stands were the keystone 
of a thriving sora and carolina rail hunting enterprise at the turn of the 19th Century 
which was renowned as the finest in the region (Coues and Prentiss, 1883).   From the 
1920's through the 1940's, the U.S. Army Corps of Engineers (USACE) was directed 
to dredge and fill these marshes on the Anacostia River as they were considered a 
health threat as a vector for mosquitos and malaria (Biohabitats, 1992; Syphax and 
Hammerschlag, 1995).  The dredging and straightening of the river effectively 
 
 
4 
 
eliminated the emergent marshes, and in their place shallow water boating and 
recreational ?tidal lakes? were envisioned (USACE, 1913). 
  
Table 1.1 Anacostia River timeline of human involvement in changes to the ecosystem. 
 
The Kingman and Kenilworth tidal lakes quickly filled with sediment derived 
from a rapidly developing watershed, in the process creating persistent, expansive 
areas of intertidal mudflats.  These mudflats were a defining characteristic of the 
Anacostia River ecosystem for over half a century until the implementation of large-
scale wetland restoration projects at Kenilworth in 1993 and Kingman Lake in 2000 
converted large areas of mudflats to emergent marsh (USACE, 1993; personal 
observation, 2000).   
The pre-restoration Kenilworth and Kingman mudflats were identified by the 
 
 
5 
 
U.S. National Park Service (NPS) and the U.S. Army Corps of Engineers for 
restoration to emergent marsh utilizing dredged river sediments and extensive 
emergent macrophyte plantings (USACE, 1993). Prior to the Kenilworth restoration 
in 1993, the reasons for a lack of volunteer recruitment of emergent macrophytes to 
the mudflats were first investigated.  Independent field experiments were undertaken 
to determine the conditions that emergent plant species required to successfully 
colonize and vegetate the mudflats (Athanas et al., 1991; Bowers, 1993; Stevenson et 
al., 1995).  Along mainstem Anacostia River mudflats, planting success / survival was 
determined to be influenced by river flow shearing energies, deposition of floatable 
wrack, sediment toxicity and anoxia, and substrate elevation (Stevenson et al., 1995), 
whereas at Kenilworth adjacent to the river channel, sufficient substrate elevation was 
determined to be the primary controlling factor for planting success (Bowers, 1993).   
After what was considered a successful marsh restoration effort at Kenilworth 
in 1993 (Syphax and Hammerschlag, 1995), plans were made by USACE to continue 
the restoration work on the extensive mudflats of Kingman Lake in 2000. While the 
Kingman project was being planted in the summer of 2000, resident Canada goose 
(Branta canadensis maxima) grazer damage to the plantngs required that fences be 
erected around and within the planting areas to allow the plants time to root and take 
hold. Excluding the geese with perimeter and interior fencing at the recommendation 
of the planting contractor, Ecological Restoration and Management, the project areas 
became green with the installation of 700,000 individual plants comprised of six 
species (Hammerschlag et al., 2001).  The following spring of 2001 saw the removal 
of the fencing which had collapsed in many places over the winter.   
 
 
6 
 
1.3 The Use of Exclosures as a Method of Exposing Alternate States in Intertidal 
Habitat 
An experimental exclosure design was first developed at the restored 
Kenilworth Marsh in 1997.  Initially intended to support a study of shorebird and fish 
invertebrate predation and algal grazing and disturbance on mudflats (May, 2000), the 
exclosures revealed the unforseen pressures of larger scale herbivory. The emergent 
macrophyte Pontederia cordata unexpectedly volunteered within the experimental 
biogenic mudflat exclosures which had previously not supported broad-leaved 
emergent plants.  In 1998, the direction of the research moved to address the question 
of herbivory on biogenic mudflats, which appeared to suppress the emergence, 
survivorship and growth of marsh macrophytes, maintaining an unvegetated biogenic 
mudflat system state while suppressing the emergent marsh system state.  
A further investigation of this effect was undertaken at the Kingman Marsh 
restoration site in the spring of 2001 following the previous years marsh construction 
and restoration planting.  An exclosure design and emergent macrophyte assessment 
regime was developed and implemented with the intention of experimentally 
determining the level of goose herbivory pressure the restoration area was 
experiencing as well as determining the effectiveness of two different exclosure sizes.   
Data are presented on marsh plant survivorship, biomass, cover and biogenic 
mudflat community structure from the two Anacostia River study areas, Kenilworth 
Marsh (1997-1998) and Kingman Marsh (2001-2004) (Figure 1.2).  The exclosure 
data appears to support the theory that alternate community states (biogenic mudflat 
 
 
7 
 
and emergent marsh) exist on an elevational gradient of 1.5-2.0 ft above NGVD '29 
(National Geodetic Vertical Datum 1929) within the intertidal zone of Anacostia 
mudflats in the presence of strong herbivory pressure.  
 
Figure 1.2 Vicinity map of exclosure study areas on Anacostia River. 
 
The greater potential of exclosure fencing uses for emergent marsh restoration 
became apparent as early as the Kenilworth exclosure study in 1998, but was not fully 
appreciated until the Kingman Marsh restoration witnessed goose herbivory on a 
massive scale (Syphax personal communication, 2001; Hammerschlag et al., 2001). 
These experiments verify that there is a distinct correlation between the planting 
 
 
8 
 
"success" of tidal marsh restoration projects in this low-middle marsh zone and the 
exclusion of resident Canada geese (Branta canadensis maxima) from some 
restoration areas.  This condition is considered an indication that alternate system 
states do exist with goose herbivory being a primary driver toward the biogenic 
mudflat state while confounding emergent marsh state development being advanced 
by the USACE and NPS. 
1.4 Research Questions and Hypotheses 
Given the power that resident Canada geese have shown in controlling 
biogenic mudflat / emergent macrophyte community dynamics, and the marsh 
restoration process driven by the Army Corps of Engineers, several research 
questions and hypotheses were developed: 
 
Question: Do alternate system states exist on Anacostia River tidal mudflats? 
Hypothesis: The evidence of significant differences in vegetation between  
excluded and unexcluded plots on the biogenic mudflats in the low-
middle (1.5-2.0 ft. NGVD '29) marsh zone suggest alternate system 
states. 
 
Question: Of what significance is the role of resident Canada geese in marsh 
restoration and alternate state theory? 
Hypothesis: Resident geese are an attractor to the biogenic mudflat state and will 
drive marsh restoration efforts toward the biogenic mudflat state on 
 
 
9 
 
Anacostia River. 
 
Question: Are exclosures revealing temporary or persistent alternate states on 
Anacostia mudflats? 
Hypothesis: Exclosures reveal the potential for persistent alternate states on 
Anacostia biogenic mudflats depending upon goose herbivory. 
 
Question: At what elevations are exclosures most successful in facilitating 
emergent vegetation establishment? 
Hypothesis: Exclosures promote successful vegetation establishment within a 
defined  zone of elevation. 
 
Question: What are emergent plant community organizing dynamics within 
exclosures over time? 
Hypothesis: Plant community dynamics within exclosures may not be generally 
predicted. 
 
Question: What are the economic and ecological costs and benefits of emergent 
marsh restoration to the environment and society? 
Hypothesis: The ecological and socioeconomic benefits of emergent marsh 
restoration are worth the economic investment. 
 
 
10 
 
1.5 Organization of the Thesis 
The work in this thesis is organized by research area, each with its own 
introduction, site conditions, methods, results and discussions sections.  A literature 
review chapter of alternate state theory, mudflats, aquatic community 
interconnections, and exclosures is followed by separate chapters on Kenilworth 
Marsh, Kingman Lake/Marsh, and a chapter detailing an emergy analysis of the 
Kingman Marsh restoration.  A discussion and summary chapter concludes the thesis 
with recommendations for the application of the research conducted here to future 
restoration work and thoughts on the restoration of marshes at the expense of mudflat 
habitat. 
 
 
11 
 
Chapter 2: Literature Review 
2.1 Alternate Stable State Theory 
The view that some ecosystems have the potential of existing and persisting as 
different stable community types within the same physical space and time, or along some 
trajectory has been referred to as alternate stable states or multiple stable states (Tansley, 
1935; Sutherland 1974, 1990; May, 1977; Peterson, 1984; Knowlton, 1992; Brinson et 
al., 1995; Lockwood, 1997; Scheffer and Jeppesen, 1998; Petraitis and Latham, 1999; 
van de Koppel et al., 2001; Young et al., 2001).  This theory has been forwarded as a way 
to describe the parallel potential, yet alternate emergence, of different stable communities 
within a single environmental condition.   
There has been debate as to whether the condition actually exists in the natural 
world as it has been shown in mathematical models and theory (Noy-Meir, 1975; May, 
1977; Sutherland, 1974, 1981, 1990; Connell and Sousa, 1983; Sousa and Connell, 1985; 
Rietkerk and van de Koppel, 1997).  Often the controversy focuses on how stability is 
defined and even determined for a natural community (Lewontin, 1969; Noy-Meir, 1975; 
May, 1977; Sutherland, 1974, 1981, 1990; Connell and Sousa, 1983; Sousa and Connell. 
1985; Knowlton, 1992).  Alternate stable state theory assumes that if a system is  
sufficiently disturbed or altered so that it reaches some threshold past which it is difficult 
to return to its original state, then it has essentially flipped into a new stable state of 
existence.   
In a review article 30 years ago, May (1977), acknowledged and agreed with a 
 
 
12 
 
growing body of empirical evidence supporting multiple stable states.  In the same article 
May also stated that "Unfortunately, the complications inherent in multi-species systems 
almost invariably preclude any quantitative confrontation between theory and data, ? for 
multi-species communities, the empirical observations remain largely anecdotal, and the 
theory remains largely metaphorical."  Notwithstanding Connell and Sousa's (1983; 
Sousa and Connell,1985) contrary position on the evidence for multiple stable states, 
others argue that the use of manipulative field experiments can effectively demonstrate 
the existence of multiple states (Sutherland, 1974, 1981, 1990; Peterson, 1984; Hik et al., 
1992; Knowlton, 1992; Petraitis and Latham, 1999; van de Koppel et al., 2001). 
In an effort to avoid the complications inherent in the stability debate with respect 
to alternate states (Knowlton, 1992), the use of the term alternate community states 
(Petraitis and Latham, 1999) or simply alternate states is employed.  The recognition of 
the power of switching forces that trigger changes between alternate states and the point 
at which a community breaks from one state to the other is crucial to the understanding of 
multiple state theory (May, 1977; Knowlton, 1992; Rietkerk and van de Koppel, 1997; 
Petraitis and Latham, 1999; Kangas, 2004a). Under some conditions an alternate state can 
be achieved very quickly, although stability implies that it would be difficult to switch a 
system state back again.  Without the ties of the stability debate, it is believed that in the 
case of Anacostia River biogenic mudflats and restored marshes a rapid change in state 
can be achieved under the right conditions.  The simple trigger of erecting a fence on a 
biogenic mudflat to induce a marsh system state, and then taking it down again to return 
to a biogenic mudflat system state is all that is needed and is investigated in this study. 
 
 
13 
 
2.2 Tidal Freshwater Mudflats 
One system that is under represented in the ecological literature is the freshwater 
tidal mudflat.  Extensive marine tidal flat experimental investigations have been 
undertaken encompassing years of accumulated data and observations (Reise ,1985, 
2001).  The unseen complexity and diversity of the marine intertidal flat system, as well 
as its global ubiquity, has long made these areas a natural choice for ecological study 
(Reise, 1985, 2001; Nordstrom and Roman, 1996).  By contrast, the freshwater tidal 
mudflats are not diverse in benthic assemblages (Brittingham and Hammerschlag, 2006) 
and can easily be overlooked by ecologists who study the diverse flora and fauna 
associated with the emergent marshes (Diaz, 1977, W.E. Odum et al., 1984, 1988; Lopez, 
1988; Findlay et al., 1989; Sacco et al., 1994). 
The ecological role of freshwater tidal mudflats may be more complex than a 
casual observation may suggest (Peterson, 1981).  W.E. Odum and associates (1984), 
some of the first to extensively study tidal fresh marshes from a whole systems 
perspective, believed at the time that "the knowledge of energy flow in tidal freshwater 
wetlands is almost totally speculative."  However, Diaz and Boesch (1977) mentioned 
that freshwater tidal flat benthic fauna appeared to be a significant contributor to the diets 
of benthic feeding fishes.  Wading birds and migratory shorebirds are also a major 
predatory force in the tidal flat feeding on invertebrates in the upper few centimeters of 
the exposed sediment surface (Baker and Baker, 1973; Schneider, 1978; Quammen, 
1984).   
Further investigations into the positive role that freshwater tidal mudflats play in 
 
 
14 
 
migratory bird feeding grounds is warranted.  Tidal marsh restoration efforts in the 
District of Columbia between 1993 and 2003 has converted or attempted to convert 
almost all expansive mudflat areas to emergent marsh utilizing dredge spoil deposited on 
the mudflats to raise sediment elevations to a point suitable for emergent macrophyte 
growth (USACOE, 1993).  While the ecological benefits of emergent marshes are 
generally recognized (Odum et al., 1984; Mitsch and Gosselink, 2000), one potential 
consequence of converting mudflats to emergent marsh is the elimination of valuable 
feeding areas for shorebirds and some waterfowl during their long migrations (Recher, 
1966; Hayman et al., 1986; Evans, 1987; Perry and Deller, 1996).  Reducing this 
significant habitat type from the District could have hidden costs which have not yet been 
accounted for. The mudflat system in the District is not well studied and there may be 
some species, energy flow pathways, nutrient processing or other functions and values 
that have not been uncovered. The question must be asked, are mudflats important and in 
what ways? 
2.3 Tidal Freshwater Aquatic Community Interconnectedness 
The study of food webs and trophic level interactions have been the source of 
significant research since Elton's landmark work Animal Ecology in 1927 and 
Lindeman's work in the trophic dynamic aspect of ecology (1942).  The later work of 
MacArthur (1955), Paine (1966), May (1973), Pimm and Lawton (1977), Reise (1985), 
Power (1992), and others solidified the importance of experimental and model 
investigations into the very basic levels of biotic interrelationships to gain a greater 
understanding of the structure of ecosystems. 
 
 
15 
 
Aquatic community interconnectedness with the biotic and physical environment 
have been experimentally studied extensively in rivers (Power 1984, 1988, 1990), marine 
systems (Paine, 1966, 1969, 1974; Reise, 1985, 2001) and estuarine systems (Teal, 1976; 
Diaz, 1992; Everett and Ruiz, 1993).  These investigations have often revealed interesting 
relationships that were only discovered after extensive observation and field 
experimentation.  The comparatively less extensive tidal freshwater marsh ecosystems 
have much wider gaps in the experimentally derived knowledge of basic functional 
relationships (Diaz and Boesch, 1977; Good et al., 1978; Odum et al., 1984; W.E Odum, 
1988; Keddy, 2000; Mitsch and Gosselink, 2000).  The pivotal compilation of Good, 
Whigham and Simpson (1978) assembled several papers on freshwater wetlands, many of 
which were works on the tidal marshes. The summary combined works of Simpson, 
Good, Leck and Whigham (1983) and Odum et al. (1984) stated the importance of and 
need for further work into the study of these systems.  Numerous studies on these systems 
have been completed (McVay et al., 1980; Doumlele, 1981; Simpson et al., 1983a; 
Bowden, 1984; Rozas and Odum, 1987; Rozas et al., 1988; Hussey and Odum, 1992; 
Leck and Simpson, 1987, 1994; Khan and Brush, 1994; Diaz, 1994; Pasternack and 
Brush, 1998; Perry and Hershner, 1999; Otto et al., 1999; Pasternack et al., 2000; 
Baldwin et al., 2001; Pasternack and Brush, 2001; Seybold et al., 2002; Tanner et al., 
2002; Capers, 2003) adding significantly to the tidal freshwater marsh literature base.  
While many of these works were arguably important, it still appears that Simpson et al., 
(1983) and Odum et al., (1984, W.E.,1988) were alone in looking at tidal freshwater 
marshes from a whole systems perspective. 
 
 
 
16 
 
2.4 Exclosures, Tidal Flats and Emergent Marsh Herbivory 
The roles of exclosures in experimental ecology and restoration have followed a 
parallel evolution for many years.  Exclosures as experiments are widely used to 
quantitatively reveal the effects of predation, herbivory or disturbance on the community 
structure of a studied ecosystem (Paine, 1969, 1974; Power, 1984, 1988; Quammen, 
1984; Reise, 1985, 2001; Bazely and Jeffries, 1989; Belanger and Bedard, 1994; Taylor 
and Grace, 1995; Grace and Ford, 1996; Evers et al., 1998; Gough and Grace, 1998; 
Rachich and Reader 1999; Baldwin and Pendleton 2003).  By establishing some sort of 
barrier to suspected predators, herbivores or bioturbators, an exclosure plot paired with an 
unexclosed control plot releases the excluded study plot from whatever known (or 
unknown) agents that are acting on the site, and with some time the change in biological 
or physical character of the site is revealed when compared to the control site (Figure 
2.1). Restoration work has often made use of some type of exclosure to prevent suspected 
herbivores from disturbing a restored system and even as a method of restoration in and 
of itself (Keller and Burnham, 1982; Platts and Wagstaff, 1984; Opperman and 
Merenlender, 2000; Haramis and Kearns, 2007). 
 
 
17 
 
 
Figure 2.1 Schematic of a typical exclosure experimental arrangement. 
 
In exclosure studies of goose herbivory in salt marshes, Bazely and Jeffries (1986, 
1989) determined that changes in plant community structure due to grazing are often 
abrupt and can be followed by periods of relative stasis.  However, their studies also 
found that the plant community changes within exclosures do not necessarily predict the 
structure of the marsh as a whole under ungrazed conditions. Bazely and Jeffries (1986, 
1989) observed that the plant community within their exclosures was not found elsewhere 
in their study area and that this represented a "cul-de-sac in community development."  
Although exclosures can arrest the effects of grazing within them, unanticipated changes 
in plant community structure can still occur seasonally and yearly as the interactive 
effects of competitive exclusion become evident.  
 Self-organization is typical for newly emerging systems after a significant 
disturbance (Odum, 1989, McKee and Baldwin, 1999) such as a massive restoration 
 
 
18 
 
effort (Kangas, 2004a; Mitsch and Jorgensen, 2004).  Although ironically, what organizes 
in exclosures, while reflecting what the community might be like without a stressor, 
should still not be considered "natural" in the sense that it would undoubtedly change 
significantly once the fencing was removed.   
The large scale fencing effort that was initially employed in the Kingman Marsh 
restoration may be considered to have been maintaining an artificial and unrealistic 
version of a restored marsh given the reality of heavy grazer activity.  Utilizing the 
unintended disturbances caused by the geese as a signal for a reevaluation of marsh 
restoration design can only help to move the discipline of restoration ecology forward 
(Zedler, 2001), instead the traditional engineering tendency toward complete control may 
only set the stage for future disappointments (Mitsch and Jorgensen, 2004). 
 
 
19 
 
Chapter 3: Kenilworth Marsh Experimental Ecology 
3.1 Introduction 
Kenilworth Marsh was once a part of hundreds of hectares of tidal freshwater 
emergent and forested wetlands that fringed the Anacostia River (Figure 3.1).  Since the 
early part of the 20th Century, plans were made to significantly alter the Anacostia 
shoreline (U.S. Army Corps of Engineers, 1913). From the 1920s-40?s the Anacostia was 
dredged by the Army Corps of Engineers to deepen and straighten the river for improved 
navigation (Syphax and Hammerschlag, 1995).  Dredged river material was used to fill 
fringing wetlands to create upland for what was to become Anacostia Park in 1923 
(Hutchinson, 1977).  In the 1940s the Kenilworth Marshes were dredged to create a 
recreational lake, although sedimentation soon filled the shallow tidal area and it became 
dominated by unvegetated intertidal mudflats considered to be of minimal habitat value 
(Syphax and Hammerschlag, 1995).   
 
Figure 3.1 United States Army Corps of Engineers 1891 Map of Anacostia River Marshes. 
 
 
 
20 
 
The National Park Service personnel had a longstanding planning concept to 
restore portions of the area to its pre-existing emergent marsh state.  Previously there was 
a concern that sediment contamination may have been prohibiting vegetative growth 
(Syphax and Hammerschlag, 1995). In 1991, to test the hypothesis that sediment 
elevation was a limiting factor to emergent plant growth, a study was conducted by 
constructing several cells, filling them with adjacent sediment to different elevations, 
then planting 10 species of emergent plants (Bowers, 1993,1995; Syphax and 
Hammerschlag, 1995).  Sediment elevations in the replicated cells ranged from 0.5-3.0 
feet above mean sea level (MSL).  Results of the field testing found that approximately 
2.0-2.1 feet above MSL was the optimal sediment surface elevation for emergent plant 
growth (Bowers, 1993; Syphax and Hammerschlag, 1995).  An unexpected factor that 
influenced plant growth in the study cells was grazing of the plants by Canada geese 
Branta canadensis (Bowers, 1993).  
In 1992 the National Park Service, collaborated with the Baltimore District Army 
Corps of Engineers to perform dredging operations in the tidal Anacostia River and 
utilize the dredged sediment to raise the surface elevations of Kenilworth Marsh to levels 
that were determined to be suitable for marsh vegetation to survive (Syphax and 
Hammerschlag, 1995). A massive planting effort was undertaken in the spring of 1993 
involving over 290,000 plants of 16 species. Channels or ?guts? were cut into the planted 
restoration areas to enhance tidal river water access to the planted areas to mimic prior 
wetland hydrology.  At the end of the first growing season, emergent marsh plant 
coverage of the restoration area was considered successful with over 90% coverage of a 
majority of the target sites (Syphax and Hammerschlag, 1995).  
 
 
21 
 
The restoration areas were defined as Mass Fill 1 (MF1) 4 ha, Mass Fill 2 (MF2) 
6 ha, and smaller dispersed fringe areas, Mass Fill 3 (MF3) of less than 3 ha (Figure 3.2).  
 
Figure 3.2 Kenilworth Marsh and Dueling Creek study areas. 
 
 
22 
 
 
Sediment surface elevations were targeted at 2.5 feet above MSL (0.495 ft.) for 
the high marsh zone of MF1 and 2.1 feet above MSL for the middle marsh zone of MF2.  
Mass Fill 3 was intended to be a middle marsh zone, although after planting it was 
determined that these areas were lower in elevation and likely at less than 2.0 feet above 
MSL making it a low marsh zone inundated to a greater depth and for a greater duration 
than the areas with higher sediment surface elevations.   Average tidal ranges on the 
Anacostia River are approximately 2.9 feet (Figure 3.3). By early September of 1993 the 
marsh vegetation was dominated by Leersia oryzoides in Mass Fill 1 while Sagittaria 
latifolia and Leersia oryzoides were co-dominants in Mass Fill 2 (Hammerschlag, 
unpublished data).  Mass Fill 3 fringe areas were slow to vegetate with mudflats 
dominating.  This was determined to be a result of lower final sediment elevations within 
the MF3 sites (Syphax and Hammerschlag, 1995).  
 
Figure 3.3 Anacostia River typical tidal hydrograph. 
 
 
 
23 
 
One approximately 4,060 m2 (0.4 ha) area of MF3, was situated just 25 m east of 
the Anacostia River inlet to Kenilworth Marsh.  This site was planted twice in 1993 and 
completely denuded each time, presumably by Canada geese. This area, maintained as an 
intertidal mudflat with an average surface elevation of just 1.1 ft (0.4 ft min to 2.0 ft max) 
NGVD above mean sea level (1995 Army Corps of Engineers Survey), was used for 
studying the ecology of the mudflat system state.    
In 1996 a preliminary plan for investigating the ecology of intertidal mudflats was 
developed.  The interrelationship between mudflat benthic macroinvertebrates and 
shorebird predation was the initial focus, with the added effects of fish predation on 
invertebrates and algae coverage also of interest. In January of 1997, initial investigations 
of invertebrate, bird and fish utilization of the Kenilworth MF3 mudflat began.  
Following the exclosure experiments of Paine in the rocky marine intertidal (Paine, 1969; 
1974), Power in riverine systems (Power, 1984; 1988; 1990) and Riese in marine tidal 
flats (Reise, 1985; 2001), in March of 1997 a fenced exclosure experimentation scheme 
was begun to determine the relative influence of shorebird vs. fish predation on mudflat 
benthic macroinvertebrates.  Grazing and disturbance effects on algae coverage of the 
mudflat was also of interest as it is the only form of primary production at the base of the 
mudflat food web.   
In July of 1997, two species of macrophytes volunteered within the biogenic 
mudflat exclosures, Pontederia cordata (17 clonal groups), and Polygonum sp. (38 clonal 
groups). These plants did not emerge anywhere else on the biogenic mudflat and had not 
since initial attempts at planting in 1993 had failed.  The unexpected effects of the 
exclosure fencing in stimulating biogenic mudflats to allow the emergence of 
 
 
24 
 
macrophytes, changed the nature and direction of the study toward the role of grazers as 
strong top-down forces in controlling emergent macrophyte emergence, survivorship and 
growth.   
Eventually, results of the study led to the discussion of the potential for mudflats 
on the Anacostia River to be models of the rationale behind alternate state theory. As 
stated earlier, this theory has been put forward as a way to describe the parallel potential, 
yet alternate emergence, of different stable communities within a single environmental 
condition. This concept, and its application to the results of the Kenilworth mudflat 
research, would have unrealized implications for future emergent marsh restoration on 
Anacostia River. 
3.2 Study Area Location Description 
The Kenilworth study area is located on the Anacostia River in Washington, D.C. 
(38? 54' 37"N, 76? 56' 54"W).  The river is completely freshwater and tidal with an 
estimated tidal amplitude of approximately 1.0m. The marsh and mudflat study areas for 
birds and fish, exclosure invertebrates, algae and plants are situated immediately south 
and east of the mainstem Anacostia River (Figure 3.2) approximately 11 km upriver of its 
confluence with the Potomac River.  Soil variables for Kenilworth surface sediments 
range from 20-57% sand, 31-48% silt and 12-32% clay with 6-11% organic matter and an 
average pH of 6.41 (Neff, 2002). Dueling Creek Marsh, a 1.2 ha site used as a local 
reference condition for fish surveys, is located approximately 0.8 km upriver of the 
Kenilworth Marsh inlet, just past the D.C. / Maryland line entering from the west side of 
the Anacostia River.  
 
 
25 
 
3.3 Research Methods and Materials 
3.3.1 Birds 
In an effort to understand the ecological association of avifauna to mudflats, birds 
were observed twice monthly at the Kenilworth Mass Fill 3 mudflat beginning in 1997, 
continuing through 1999 (Appendix A, Table 3.1).  Observations were made from the 
edge of a wooded area that is adjacent to the mudflat at low tide.  Low tide counts are 
considered representative of average waterbird usage (Burton et al., 2004). A count of all 
birds within the field of view of the mudflat study area was conducted over a period of 30 
minutes, with notes taken differentiating if the birds were physically in the mudflat zone 
or outside the mudflat in the open water, trees or adjacent shoreline.  Only birds with a 
water connection were identified, with all primarily terrestrial bird species not counted in 
the survey aside from the two species of hawk that conspicuously circled the mudflat area 
while scores of killdeer fed at low tide.  After an initial count over 15 minutes, the 
remainder of time was spent observing the behaviors of the birds on the mudflat and 
counts made for new visitors to the area.  All counts of large numbers of birds were 
double-checked during this time period and the lesser of the two numbers used as the 
official count.  
3.3.2 Fish 
To define the system components of fish to mudflat and emergent marsh, surveys 
were conducted within the tidal channels of Kenilworth and Dueling (Figure 3.2) using a 
Smith-Root fish electrofishing boat operated by the D.C. Government Fisheries Division. 
 
 
26 
 
Surveys were conducted on a flooding high to slack spring tide during each shocking run 
to allow for shocker boat access. A total of 25 electrofishing surveys were completed 
covering spring, summer and fall seasons intermittently between 1993 and 1998 
(Appendix A, Tables 3.2 to 3.6).  Shocking levels were set at 240-500 volts at 3-5 amps, 
60 pulses/sec, adjusted for conductivity (D.C. Fisheries standard operating procedures). 
Approximately 1000 seconds of shocking time was conducted within each of the tidal 
channels in Mass Fill areas 1 & 2, as well as the open water adjacent to the Mass Fill 3 
area experimental mudflat. Each survey was completed through two, 100m passes over 
the same area, once in and once out of survey line.  All fish collected were kept in live 
wells on the boat and identified to species, counted, measured for length and released 
after shocking was completed.  Game fish and larger fish species were also weighed to 
the nearest gram.  
A bag seine survey of the fishes leaving the mudflat on an ebb tide was also 
conducted in January and February of 1997, and October of 1998 (Appendix A, Table 
3.6).  A 30m weighted bag seine with surface edge floats was deployed across a tidal gut 
draining the study mudflat on a flood/slack tide and retrieved on the ebb/slack low tide.  
All fish in the bag sein survey were collected, identified to species, counted and released. 
Electrofishing surveys were also conducted at Dueling Creek, a 1.2 ha mature 
emergent tidal marsh just upriver of the Kenilworth Marsh inlet above the District line in 
Maryland.  Dueling Creek had not been physically altered for over 50 years since the 
original straightening of the Anacostia River.  This system was used as a reference 
condition for the Kenilworth Marsh restoration and was surveyed with the same level of 
effort encompassing approximately the same total shocking area as was surveyed in 
 
 
27 
 
Kenilworth, often on the same date and tide. Abundance, numbers of species, frequency 
and relative importance of each species found during the observation period were 
calculated, as well as the percent similarity of the numbers of species to each survey site. 
3.3.3 Exclosures 
In late March of 1997 field exclosure experimentation at Kenilworth Marsh began 
on the mudflat within the Mass Fill 3 restoration area.  Four, 7m x 7m treatment plots in 
the experiment included: mudflat control plots of fish and bird accessible areas (AA, all 
accessible, no exclosure); fish accessible/bird exclosure (FA, partial exclosure); bird 
accessible/fish exclosure (BA); and fish/bird exclosure (TX, total exclosure) (Figure 3.4).   
 
Figure 3.4 Schematic of Kenilworth MF3 mudflat exclosure treatment types. 
 
Three blocks (A-C) with each of the four treatments were established for a total of 
12 treatment plots on the mudflat (Figure 3.5).  All plot areas were randomly assigned 
 
 
28 
 
treatments within each block, with blocks stratified for approximately equal elevation as 
determined by arrangement of the blocks parallel to the incoming tide with flood facing 
fence sides positioned at the edge of low tide. 
 
Figure 3.5. Arrangement of Kenilworth MF3 mudflat exclosure treatment blocks. 
 
The perimeter fencing type used for the exclosures consisted of aquaculture grade 
nylon netting with 1cm mesh diameter and 1.0m fence height wrapped around 5cm by 
5cm, 2m long hardwood stakes sunk approximately 1m into the mudflat at the corners of 
each exclosure.  Total exclosure and bird accessible/fish exclosure plots had fencing flush 
with the mudflat while only total exclosure and fish accessible/bird exclosure treatments 
had lines crossing above the plots to prevent bird landings within the plots.  The bird 
accessible treatment was intended to allow for the landing of shorebirds within the plot to 
 
 
29 
 
permit invertebrate predation while excluding fish access. The 7mx7m exclosure sizing 
for this study was among the largest found to be used for experimental tidal flat research 
at the time (Quammen, 1984; Riese, 1985).  Fish accessible/bird exclosure treatment 
plots had 25cm of fencing trimmed on two sides of each plot from the mud up to exclude 
birds but allow fish entering the mudflat area on the incoming tide.  Non-excluded 
control plots had no fencing on at least two sides of each plot and no overplot bird 
exclusion lines. 
In 1997 the BA, bird accessible/fish exclosure treatment was observationally 
found to not allow for shorebird access with the treatment effectively acting as a total 
exclosure to fish and birds. In 1998 these exclosure treatment plots were dismantled as 
they were functioning as a total exclosure and removed from the experiment while the 
other treatments plots were maintained.  
3.3.3.1 Invertebrates 
Shorebird and fish predation impacts on benthic macroinvertebrate populations 
within the different exclosure treatment types were evaluated.  Samples were collected 
from the mudflat substrate with a manual coring device to evaluate macroinvertebrate 
species composition and densities (Vorberg, 1993). The corer had a diameter of 6.5 cm 
and was sunk into the mudflat to a depth of approximately 10cm.  Five core samples were 
collected from within each treatment (166 cm2), with each core sample site located using 
a random numbers table and values applied to a coordinate grid across each plot. Core 
samples were only taken from locations at least 1m from any fence line to avoid edge 
effects.  The 1m band around the inside perimeter of each treatment plot was used as a 
 
 
30 
 
movement corridor when samples were collected to avoid disturbing the mudflat as much 
as possible. In 1997, sediment core samples were taken from each of the four mudflat 
exclosure treatments on a low tide in March, May, July, September, and November.  In 
1998, invertebrate cores were collected monthly from each of three treatments from April 
through November.  Mudflat sediment samples were immediately rinsed through a U.S. 
Standard #30 sieve, (0.595 mm2 openings), and the remaining material preserved in site 
labeled jars of 70% ethanol and stained with several grams of rose bengal.  In a lab, the 
contents of each jar was rinsed of preservative and spread over a light table in a glass tray 
with all stained invertebrates identified, counted and removed from the tray. 
3.3.3.2 Algae 
Effects of fish and bird grazing and disturbance on algae percent coverage of the 
mudflat were evaluated through the exclosure treatments.  The mudflat exclosure algae 
coverage was determined at low tide once monthly beginning in April throughout 1997 
and once every month in 1998, but for March, which was flooded due to spring storm 
flooding events at several attempted field collection dates.  Algae percent coverage was 
determined through the use of a 1m x 1m square wooden frame with a fishing line grid 
overlay at 10cm intervals. The frame was placed at three randomly generated coordinate 
positions on the mudflat within each treatment plot.  Algal cover measurements were 
only taken from locations at least 1m from any fence line to avoid edge effects.  The 1m 
band around the inside perimeter of each treatment plot was used as a movement corridor 
when samples were collected to avoid disturbing the mudflat as much as possible. Percent 
algae coverage was determined by counting points under grid line intersections where 
 
 
31 
 
algae were visually evident on the mudflat. 
3.3.3.3 Emergent Macrophytes 
During the course of invertebrate and algae data collection, the emergent 
macrophytes, Pontederia cordata and Polygonum sp., were observed to volunteer in the 
in several of the exclosures beginning in July of 1997 and their presence in each 
exclosure treatment noted throughout the growing season.  Exclosure plots were visited 
and maintained throughout the winter. In the second year (1998), active experimentation 
was employed as intentional plantings were used to more fully understand the 
interactions between exclosures and macrophyte survival and growth under different 
levels of fish and bird accessibility as potential grazer or disturbance factors.  Five, 50cm 
average stem height, actively growing bare root Pontederia  plants were grouped in the 
middle of each treatment plot on one meter centers in an ?X? configuration in early June 
of 1998.  Plant presence and evidence of grazing was recorded through the growing 
season for each plot.  At the peak of the growing season, individual plant clonal clusters 
were measured for area coverage. All above-ground plant biomass was collected from 
each of the treatment plots by cutting at the mudflat surface, bagged, plot coded, and 
weighed.  Fresh plant matter was later rinsed of mud and sorted by treatment plot and 
plant number. Biomass was initially air dried for several days and then oven dried at 80? 
C for at least 48 hours and weighed to the nearest 0.01g.  
 
 
 
32 
 
3.4 Data Analysis Methods 
 Absolute and relative abundance, average numbers of individuals 
observed/survey, frequency and importance values of each species recorded during the 
observation periods were calculated for all bird and fish data.   
A Chi-Square analysis was conducted on invertebrate data collected within the 
experimental exclosure treatment plots during years 1997 and 1998.  Repeated sampling 
of each treatment type within each block repetition revealed strong differences in 
invertebrate numbers between similar treatments between blocks, which is often expected 
with populations of invertebrates in tidal flats (Riese, 1985).  These block differences 
were removed by converting invertebrate raw numbers, or counts, into percentages (%) of 
the total invertebrates of each species (group) per sample from each treatment, which 
would transform the effect of the natural variability of invertebrate populations between 
repetitions to one that could be more validly compared between exclosure types with the 
Chi-Square analysis.  
A two-way ANOVA using the SAS mixed model procedure (PROC MIXED) was 
applied to each years invertebrate data converted to densities (#individuals/m2), 
incorporating the blocking structure of the replicated plots as a randomized variable 
(SAS, 2006).  The model was exclosure type as an independent descriptor variable and 
invertebrate density (#/m2) as the dependant response variable.  The variances between 
exclosure types were heterogeneous so the data were log transformed. Tukey test 
adjustments were made to identify significantly different means, with the overall family-
wise experimental error rate not to exceed ? = 0.05 and the standard error of the mean 
 
 
33 
 
presented as the measure of variation. 
The SAS mixed model procedure was also used to analyze the algae data.  The 
algae data were first averaged across both years and by block repetition and exclusion 
type as well as by season.  Seasons were divided into Spring (March-May), Summer 
(June-August), Fall (September-November), and Winter (December-February). As with 
the invertebrate density data, Tukey test adjustments were made to identify significantly 
different means, with the overall family-wise experimental error rate not to exceed ? = 
0.05 and the standard error of the mean presented as the measure of variation. 
A two-way ANOVA using the SAS mixed model procedure (PROC MIXED) was 
applied to the macrophyte biomass data incorporating the blocking structure of the 
replicated plots as a randomized variable (SAS, 2006).  The model was exclosure type as 
an independent descriptor variable and dry weight as the dependant response variable.  
The variances between exclosure types were heterogeneous so the data were log 
transformed.  Because there was no dry weight biomass within each of the three all 
accessible control (AA) treatment plots, there was no variance between each of the 
treatment plots. Tukey test adjustments were made to identify significantly different 
means, with the overall family-wise experimental error rate not to exceed ? = 0.05 and 
the standard error of the mean presented as the measure of variation. 
 
 
 
34 
 
3.5 Results 
3.5.1 Birds 
Bird observations at the Kenilworth MF3 mudflat study area were conducted over 
87 separate site visits at low tide between January 1997 and August 1999 (Appendix A, 
Table 3.1).  Of 33 species of birds representing 8 families, ten species had a frequency of 
occurance of 10% or greater and nine species had a relative abundance of greater than 1% 
(Figure 3.6), with 24 species having a relative abundance of less than 1% (Table 3.1). In 
terms of total numbers of birds counted over the span of the survey the Canada goose, 
Branta canadensis, were most numerous (666 individuals) with the killdeer Charadruis 
vociferous, next with 460 individuals counted, followed by mallards Anas platyrhynchos 
(354) and ring-billed gulls Larus delawarensi (204).  
-1
0
1
2
3
4
5
6
7
8
9
10
Canada goose Killdeer Mallard Ring billed gull Great blue
heron 
Greater
yellowlegs
American crow Pintail Great egret
Bird species representing 1% of the total birds observed 
Av
er
ag
e #
 b
ird
s o
bs
er
ve
d 
 
Figure 3.6 Kenilworth MF3 mudflat bird observations 1997-1999.  
 
 
35 
 
The majority of these birds were identified on or very near the exclosure mudflat.  
The great blue heron, Ardea herodias, was most frequently observed (77% of the time) 
on or around the mudflat area with killdeer following in frequency at 67% and geese at 
46%.    The top five species in terms of relative abundance included: Canada goose 
(22%), killdeer  (15%), mallard (12%), ring-billed gull (7%), and great blue heron (5%).   
The calculation of an importance value (% relative abundance + % frequency / 2) 
gave the relative significance of each species of bird to the mudflat observation area.  
This value integrates the percent relative abundance of numbers of each species to the 
total number of birds counted with the percent frequency, or likelihood of each species to 
be seen at the site during a site visit.  Table 3.1 is ranked in order of importance value.  
The killdeer ranked highest in importance with the great blue heron ranked close behind. 
Canada goose, ring-billed gull and mallard rounded out the top five species of importance 
to the mudflat observation area.  
Of the 8 families of birds that were observed in the mudflat study area, 6 guilds 
were represented (Table 3.2).  The feeding group representing the greatest number of 
species was the shorebirds followed by the dabbling ducks/geese. In terms of the average 
numbers of birds within these groups, the ranking was reversed, due to the numbers of 
geese and mallards representing more than twice the average numbers of shorebirds.  
Essentially these feeding groups were almost entirely represented by the numbers of 
killdeer, geese and mallards.  Lesser numbers of individuals of the other species within 
each of these groups made up the shorebird and dabbling waterfowl feeding groups.   
Scavengers and diving birds were the next feeding groups found in the mudflat 
area that comprised a total of 5 species of birds each. In terms of average numbers of 
 
 
36 
 
 
Table 3.1 Kenilworth MF3 mudflat bird surveys 1997-1999 listed in order of importance value 
(relative abundance + frequency / 2).   Functional feeding groups are: SB, Shorebirds; WB, Wading 
birds; DDG, Dabling ducks/geese; SC, Scavengers; R, Raptors.  
 
  
 
Table 3.2 Kenilworth MF3 mudflat study area bird functional feeding groups. 
 
 
 
 
 
37 
 
birds observed over the duration of the survey, the ring-billed gull, fish crow 
Corvus ossifragus, and herring gull Larus argentatus, contributed the greatest numbers of 
individuals of each species to the average for the scavengers.  Diving birds, while 
representing the same number of species observed over the survey, contributed much 
lesser numbers in terms of the average numbers of birds found in the mudflat area.  The 
belted kingfisher Ceryle alcyon, was frequently represented, likely due to one or more 
individuals whose territorial range included the mudflat study area.  Double crested 
cormorants Phalacrocorax auritus,  hooded merganser Lophodytes cucullatus, and 
common merganser Mergus merganser, all represented divers in similar numbers, 
although the cormorant was twice as likely to frequent the area as the mergansers which 
have shorter migration visits to the area.  Wading birds representing 4 species contributed 
greater average numbers of birds than the diving birds.  This was mostly due to the 
prevalence of the great blue heron, which was the most frequently observed bird in the 
study area and second in importance only to the killdeer, both of which preferentially 
feed in the intertidal zone.   
Four species of raptor were identified in the study area over the survey period.  
The low numbers of individuals observed was indicative of the feeding group at the top 
of the food web and their solitary nature.  The osprey Pandion haliaetus, true to its 
alternate common name ?fish hawk? was infrequently observed in the study area. One 
instance of a bald eagle Haliaeetus leucocephalus, was of note in that it swooped into the 
mudflat exclosure study area to carry away a catfish that had been recently stranded by 
the tide in a shallow pool. 
 
 
38 
 
3.5.2 Fish 
Electrofishing conducted at Kenilworth included 20 surveys in three areas of the 
marsh, Mass Fill 1 (8 shocking events), Mass Fill 2 (8 shocking events), and Mass Fill 3 
(4 shocking events), with an additional 4 bag sein surveys conducted at Mass Fill 3 and 5 
shocking events conducted at the Dueling Creek reference marsh (Appendix A, Tables 
3.2 to 3.6).  A total of 27 species of fish representing 11 families were collected within 
the 3 Mass Fill areas of Kenilworth and 25 species captured of 11 families at Dueling 
Creek.  Families included: Cyprinidae (carps and minnows); Anguillidae (freshwater 
eels); Antherinidae (silversides); Fundulidae (killifishes); Clupeidae (herrings); 
Ictaluridae (bullhead catfishes); Castomomidae (suckers); Percidae (perches); Poeciliidae 
(livebearers); Moronidae (temperate basses); and Centrarchidae (sunfishes). 
 
3.5.2.1 Mass Fill 1 Fish 
Fish captured at the Mass Fill 1 electrofishing site included 22 species in 10 
families and 855 individuals (Appendix A, Table 3.2). The site was dominated by gizzard 
shad Dorosoma cepedianum in terms of total numbers caught (247) and relative 
abundance of the total catch (29%), being captured in 75%, or in 6 of the 8 shocking 
events (Table 3.3).  These large planktivores swim in schools and quickly grow to sizes 
that make them unlikely prey to other fish. The brown bullhead catfish Ameiurus 
nebulosus, was the next largest contributor species to the survey in total numbers (198) 
and relative abundance (23%), also caught with a 75% frequency.  Brown bullhead 
catfish, very common to the Anacostia River, are bottom-feeding omnivores that have the 
 
 
39 
 
unique trait of actually caring for their young.  
While captured in fewer numbers and relative abundance, the mummichog 
Fundulus heteroclitus, and banded killifish Fundulus diaphanus were caught more 
frequently (88%) than the gizzard shad or brown bullhead catfish.  Ten species 
contributed greater than 1% of the total catch although the top two species comprised 
more than 50% of the fish caught with the other 8 each contributing less than 10%.  The 
only piscivore in the top ten for Mass Fill 1 was the white perch Morone americana, with 
a relative abundance of almost 10%, occurring in 75% of the surveys of the site.  A single 
largemouth bass Micropterus salmoides, and one striped bass Morone saxatilis were the 
other piscivores identified in the survey.  
In terms of relative importance of each species to the site, an importance value (% 
relative abundance + % frequency / 2) was determined and Table 3.3 ranked them 
accordingly.  Gizzard shad, brown bullhead and mummichog were the top three, with the 
banded killifish and common carp Cyprinus carpio, rounding out the top five.  Of note at 
MF1 was the mosquitofish Gambusia affinis, not captured at any of the other sites 
including the Dueling reference site.  
3.5.2.2 Mass Fill 2 Fish  
Electrofishing at Mass Fill 2 captured 21 species from 8 families representing 815 
individuals (Appendix A, Table 3.3).  Similar to that of MF1, the gizzard shad and brown 
bullhead catfish were the top two species represented in MF2, with the total numbers and 
relative abundance of brown bullheads (205, 25%) greater than that of the gizzard shad 
(144, 18%) (Table 3.4).  These species, along with the white perch (92, 11%), made up  
 
 
40 
 
 
Table 3.3 Kenilworth Mass Fill 1 marsh fish community from 8 electrofishing surveys; absolute 
abundance (total # caught); average #/survey; relative abundance (% total #); frequency (% survey 
sp. appearance); ranked by importance value (relative abundance + frequency / 2). 
 
 
 
more than 50% of the total catch, with 8 species individually comprising less than 
1% of the total catch.  In terms of frequency of species occurrence, gizzard shad and the 
pumkinseed sunfish Lepomis gibbosus, were caught in 88% of the surveys with brown 
bullheads and white perch caught in 77% of the surveys.  Table 3-4 ranks species of fish 
for MF2 in order of their importance value, with the top five listed as gizzard shad, brown 
bullhead, pumpkinseed sunfish, white perch, and mummichog killifish.  The piscivores 
white perch, striped bass and largemouth bass were all represented. 
 
 
41 
 
 
Table 3.4  Kenilworth Mass Fill 2 marsh fish community from 8 electrofishing surveys; absolute 
abundance (total # caught); average #/survey; relative abundance (% total #); frequency (% survey 
sp. appearance); ranked by importance value (relative abundance + frequency / 2). 
 
 
3.5.2.3 Mass Fill 3 Fish 
This site represents fishes found in the tidal channel adjacent to the mudflat 
exclosure study area.  Electrofishing at Mass Fill 3 captured 19 species from 9 families 
representing 389 individuals (Appendix A, Table 3.4).   Only half the survey events were 
conducted at this site compared to the MF1 and MF2 sites, as the mudflat system was not 
considered for study until 1997.  The top three species in total numbers and relative 
abundance, gizzard shad (155, 40%), brown bullhead (91,23%), and pumpkinseed sunfish 
(66,17%), accounted for 80% of the total fish captured at the site (Table 3.5).  These two 
fish were the only species captured at each of the shocking events with a frequency of 
100%. Ranked by importance value, the top five species were gizzard shad, pumpkinseed 
sunfish, white perch, eastern silvery minnow Hybognathus regius, and brown bullhead. 
 
 
42 
 
Nine species of fish comprised less than 1% of the total catch.  Of the piscivores, the 
white perch comprised 4% of the total catch while a single striped bass was collected. 
 
Table 3.5  Kenilworth Mass Fill 3 mudflat fish community from 4 electrofishing surveys; absolute 
abundance (total # caught); average #/survey; relative abundance (% total #); frequency (% survey 
sp. appearance); ranked by importance value (relative abundance + frequency / 2). 
 
 
3.5.2.4 Dueling Creek Fish 
The Dueling Creek reference marsh electrofishing surveys captured a total of 25 
species of fish representing 9 families (1607 individuals) (Appendix A, Table 3.5).  The 
top four species caught in total numbers and relative abundance were blueback herring 
Alosa aestivalis (451,28%), white perch (308,19%), eastern silvery minnow (250, 16%), 
and pumpkinseed sunfish (206, 41%), accounted for over 75% of the total catch (Table 
3.6).  Six species of fish were captured in every survey with a frequency of 100%.  These 
species, in order of their importance, were white perch, eastern silvery minnow, gizzard 
 
 
43 
 
shad, mummichog killifish, spottail shiner Notropis hudsonius, and banded killifish.  
Thirteen species were found in numbers representing less than 1% of the total catch. A 
species and family captured at Dueling and not at any of the Kenilworth sites was 
American eel Anguilla rostrata, of which only three individuals were found.  
 
Table 3.6  Dueling Creek marsh fish community from 5 electrofishing surveys; absolute abundance 
(total # caught); average #/survey; relative abundance (% total #); frequency (% survey sp. 
appearance); ranked by importance value (relative abundance + frequency / 2).  
 
 
3.5.2.5 Mass Fill 3 Seine 
A minor seining effort was conducted at the Mass Fill 3 mudflat identifying 12 
species of fish representing 7 families and 940 individuals (Appendix A, Table 3.6).  The 
dominant species captured was the mummichog killifish with 670 individuals 
representing 71% of the total catch (Table 3.7).  The banded killifish represented the 
second highest relative abundance of fish in the sein survey with 15% and 143 
 
 
44 
 
individuals. The top three fish in importance value were the mummichog. banded 
killifish, and the brown bullhead catfish. Half of the species of fish were captured less 
than 1% of the time in the survey.  Almost 98% of the fish were captured during one fall 
survey.  
Table 3.7  Kenilworth Mass Fill 3 mudflat fish community from 4 bag sein surveys; absolute 
abundance (total # caught); average #/survey; relative abundance (% total #); frequency (% survey 
sp. appearance); ranked by importance value (relative abundance + frequency / 2). 
 
 
3.5.3 Fish Survey Site Comparisons 
A comparison between each of the sites found that in terms of numbers of species, 
each of the sites was remarkably similar to each other.  A percent similarity matrix 
relating the numbers of common fish species between each site reflected these 
commonalities (Table 3.8).  Similarity was measured from one site to the other as the 
number of common species between two sites, divided by the number of total species 
collected from the first site that is being compared to the other.  Sites in the matrix 
compared to themselves were given a value of 1.  The Dueling Creek reference site had 
72% of its species in common with the Kenilworth Mass Fill 3 mudflat site.  Dueling had 
76% and 84% of its species composition in common with the restored marsh Mass Fill 1 
 
 
45 
 
and Mass Fill 2 sites respectively. The numbers of common species found at Kenilworth 
Mass Fill 1 was 73% similar to the mudflat site of Mass Fill 3.  Nearly all of the species 
collected at MF1 (86%) and MF3 (95%) were also represented at the Dueling reference, 
and 100% of the species captured at MF2 were also caught at Dueling. 
 
Table 3.8  Percent (%) similarity matrix comparing common fish species collected between sites from 
all years sampling events. Similarities read as far left column % species similarity to next site from 
left to right. 
 
Total numbers of fish collected at the restored marsh sites of MF1 and MF2 were 
855 and 815 respectively, with each of the two sites shocked 8 times. When total 
numbers of fish were averaged out for the number of surveys conducted at each site, the 
MF1 and MF2 sites were expectedly close with 107 and 102 fish caught on average per 
shocking event respectively.  Figure 3.7 combines the Mass Fill 1 and Mass Fill 2 
sampling areas and represents the average numbers of each species of fish that were 
captured in each sampling event for these two areas.  Mass Fill 3, shocked on 4 dates, 
collected 97 fish on average per survey (Table 3.5).  Figure 3.8 represents the average 
 
 
46 
 
 
Figure 3.7 Kenilworth MF1 and MF2 combined electrofishing survey results.  Fish averages 
arithmetic means + SE. 
 
Figure 3.8  Kenilworth MF3 combined electrofishing survey results.  Fish averages arithmetic means 
+ SE. 
 
 
 
47 
 
numbers of fish captured for each species on a given survey date at Mass Fill 3.  
Dueling Creek, shocked on 5 dates, captured more than three times as many fish on an 
average survey (321 individuals) than each of the Kenilworth Mass Fill areas (Table 3.6).
 Figure 3.9 represents the average numbers of fish captured for each species at 
Dueling.  Each of the sites surveyed had several species of migratory fish, with a least 
one individual represented from each of the anadromous fish species which included: 
blueback herring, alewife herring Alosa pseudoharengus, American shad Alosa 
sapidissima, and striped bass.  Dueling Creek surveys captured the only catadromous 
species, American eel Anguilla rostrata.    
 
Figure 3.9  Dueling Creek Marsh electrofishing survey results.  Fish averages are arithmetic means + 
SE. 
 
 
 
 
48 
 
Comparisons between the species collected from electrofishing and seining 
surveys of the Mass Fill 3 mudflat revealed that each survey resulted in different total 
numbers of mummichog killifish captured.  Electrofishing collected a total of 4 
mummichogs, while seining collected 670 mummichog killifish from the mudflat area 
with relative abundances of 1% and 71% respectively.  The discrepancies in numbers 
represent a flaw in boat based electrofishing surveys in that they only sample fish that are 
within their electrical field at high tide.  Fish such as Fundulus sp. move in schools in 
shallow water areas with the flooding and ebbing tide to avoid predation so are less likely 
to be captured by boat, and are more likely to be captured by a seine left in a tidal channel 
over a tidal cycle. 
3.6 Exclosures 
3.6.1 1997 Invertebrates 
A total of 300 invertebrate sediment core samples were collected in 1997, 
combined from each of the exclosures treatments in three blocks of treatments. Samples 
were collected every other month between March and November. Repeated sediment 
core sampling within the Kenilworth Mass Fill 3 exclosures produced three primary 
forms of benthic macroinvertebrates.  These included the numerically dominant family of 
aquatic worms, Oligochaeta, the second most numerous mudflat invertebrate from the 
family of midge fly larvae, Chironomidae, and the non-native bivalve mollusk, Asian 
clam Corbicula fluminea of the family Corbiculidae.  
The Chi-Square analysis averaged across the percentages of the three invertebrate 
types according to exclusion type, had six (6) degrees of freedom and a value of 68 with 
 
 
49 
 
p <0.01 indicating that there were significant differences in invertebrate group 
percentages within the each of the exclosure types.  Figure 3.10 illustrates the 
invertebrate group composition within each exclosure type for 1997.  Oligochaetes 
dominated in terms of total numbers of individuals and percentages of the total 
invertebrate population within each exclosure type (Appendix A, Table 3.7).  Across all 
exclosures, oligochates made up 82% of all invertebrates with 3650 individuals counted, 
with chironomids totaling 17% with 739 individuals and corbicula comprising 1% with 
54 individuals counted of the total numbers of invertebrates collected.   
 
Figure 3.10  1997 Invertebrate group percent composition within each exclosure treatment type.  
(AA) all accessible control treatment; (FA)  fish accessible/ bird exclosure treatment; (BA) bird 
accessible/ fish exclosure treatment; (TX) total exclosure treatment.  
 
 
 
50 
 
When converted to mean numbers of invertebrates/m2, analysis of variance for 
each exclosure type revealed significant differences (p<0.05) in the densities of 
oligochaetes collected from both the total exclosure (TX) and bird accessible/fish 
exclosure (BA) plots when compared to the control plot, all accessible (AA) exclosures 
(Figure 3.11).   There were also significant differences in oligochaete densities between 
the total exclosure plots and the fish accessible/bird exclosure plots.  Significant 
differences did not exist for any other exclosure type or invertebrate population in 1997. 
 
Figure 3.11 Invertebrate densities by year and exclosure type.  Densities are Tukey adjusted LS 
means + SE.  Mean sharing any com on letter are not significantly different (p < 0.05). Letter case 
differentiates years. Means with no letters are also significantly different. (AA) all accessible control 
treatment; (FA) fish accessible/bird exclosure treatment; (BA) bird accessible/fish exclosure 
treatment; (TX) total exclosure treatment.    
 
 
 
 
51 
 
The all accessible (AA) treatment type had lower total numbers of individuals and 
absolute percentages of oligochates (242 & 7%) than the fish accessible/bird exclusion 
(FA) (395 & 11%) treatment, with both treatments allowing fish access to the plots.  The 
treatments that prohibited fish access to the plots found higher numbers of individuals 
and percentages of oligochaetes, with values for bird accessible/fish exclosure (BA) 
(1458 & 40%) and total exclosure (TX) (1555 & 42%).  The all accessible control (AA) 
treatment had only slightly lower numbers and percentages of chironomid (120 & 16%) 
and corbicula (8 & 15%) invertebrates within the treatment type than the fish accessible/ 
bird exclosure (FA) (121 & 16% chironomid; 11 & 20% corbicula), while the bird 
accessible /fish exclosure (BA) (363 & 49% chironomid; 25 & 46% corbicula) and total 
exclosure (TX) (135 & 18% chironomid; 10 & 19% corbicula) treatments found higher 
numbers and percentages of invertebrates.   
 
3.6.2 1998 Invertebrates 
The BA, bird accessible/fish exclosure was removed from the study in 1998.  It 
was determined that this treatment effectively functioned as a TX total exclosure, as the 
only treatment differences were crossed lines over the top of the TX exclosures to prevent 
birds from landing with them.   Because there was no evidence of shorebirds or 
waterfowl landing within the fencing of the BA treatment from visual observations or the 
presence or tracks the previous year (an occasional heron was an exception), the 
exclosure fencing for this treatment was taken down and the plot was no longer sampled.  
Although this removed 3 replicate exclosure study plots, 8 months of samples were 
 
 
52 
 
collected monthly between April and November (3 more months than in 1997) so a total 
of 360 invertebrate sediment cores were collected and analyzed.  
The Chi-Square invertebrate analysis for 1998 averaged across the percentages of 
the three invertebrate types according to exclusion type, had (6,4) degrees of freedom and 
a value of 84 with p <0.01 that there were again differences in invertebrate percentages 
within the each of the exclosure types.  Figure 3.12 illustrates the percent invertebrate 
group composition within each exclosure type for 1998. Oligochaetes again dominated in 
terms of numbers of individuals and percentages of the total invertebrate population 
within each exclosure type (Appendix A, Table 3.8).  Across all exclosures, oligochaetes 
made up 75% of all invertebrates, with chironomids totaling 11% and corbicula 
comprising 14% of the total invertebrates collected.   
 
Figure 3.12  Invertebrate group percent composition within each exclosure treatment type. (AA) all 
accessible control treatment; (FA) fish accessible/bird exclosure treatment; (TX) total exclosure 
treatment. 
 
 
53 
 
When converted to mean numbers of invertebrates/m2, analysis of variance for 
each exclosure type revealed significant differences (p<0.05) in the densities of 
oligochaetes collected between the total exclosure plots (TX) and the all accessible 
control exclosure plots (AA) (Figure 3.11).   There were also significant differences in 
oligochaete densities between the total exclosure plots and the fish accessible/bird 
exclosure plots.  Significant differences did not exist for any other exclosure type or 
invertebrate population in 1998. 
The unfenced control (AA) treatment type had lower numbers and percentages of 
oligochates (44 & 11%) than the fish accessible (FA) (75 & 18%), and total exclosure 
(TX) (286 & 71%) treatments respectively.  The (AA) control treatment had similar 
numbers and percentages of chironomid (16 & 28%) and corbicula (26 & 34%) within 
the treatment type to the fish accessible (FA) (17 & 30% chironomid; 21 & 28% 
corbicula) treatment type, while the total (TX) exclosure treatments had higher numbers 
and percentages (24 & 42% chironomid; 28 & 37% corbicula) of those invertebrates.   
The trend in invertebrate response to exclosure treatment type for the 1997 
sampling year was equivalent to the 1998 sampling year when invertebrates were counted 
across all months. In terms of total numbers, in1998 there were less invertebrates 
collected, only 537 individuals of all species, when compared to 1997 counts (4,443 
individuals of all species).  This represents a reduction of 88% even though three 
additional months were sampled in 1998.   
3.6.3 Algae 
The dominant form of algae found on the Kenilworth study mudflat was the 
 
 
54 
 
filamentous blue-green algae Oscillatoria sp. along with numerous diatoms.  Between 
1997 and 1998, 621 separate algae % coverage measurements were made across blocks 
and exclosure treatments (Appendix A, Table 3.9). The model analysis found that algae 
% coverage by exclosure type had an F value of 12.07 with (3, 30) degrees of freedom 
and a p <0.01.  Algae % cover by season had an F value of 81.05 with (3, 30) degrees of 
freedom and a p < 0.01 while exclosure type by season had an F value of 1 with 9 degrees 
of freedom and a p = 0.41. Residuals from the mixed model procedure differences of 
least squares means between exclosure types were found to be within acceptable ranges, 
so there was no need to transform the data. 
The differences in least squares means between exclosure type and differences 
across both years were similar within seasons. Tukey adjusted differences between algal 
percent coverage least square means were found to be highly significant (p < 0.001) 
between the AA control treatments and the TX total exclosure treatments across both 
years (Figure 3.13).  Additionally, there were significant differences (p <0. 01) between 
the AA control treatment and the BA bird accessible/fish exclosure treatment, which was 
tested in 1997 but not 1998.  No significant difference (p = 0.99) was found between the 
BA bird accessible and TX total exclosure types.  The AA control and FA fish accessible 
exclosure types also had no significant differences (p = 0.39) between the treatment types 
across both years.   
The FA fish accessible treatment appears to have reacted similarly to the AA 
control treatment in terms of repressed algal coverage when compared to the TX total 
exclosure and BA bird accessible treatments, both of which effectively blocked fish and 
access.  The accessibility of fish to the AA control and FA fish accessible exclosures was 
 
 
55 
 
the main commonality between the two treatments. 
Significant differences between the FA fish accessible and TX total exclosure 
treatments (p <0.01) and the FA fish accessible and BA bird accessible treatments (p < 
0.01) reflected greater algal percent coverages in the TX total exclosure and BA bird 
accessible exclosures. (Figure 3.13).  As evidenced by the swarms of killifish (Fundulus 
sp.) observed to move across the mudflats on the vanguard of flooding tides (Appendix 
A, Table 3.10), these schools of fish were turning over the top layers of sediment looking 
for food and in the process most likely limiting the potential of algae to develop the dense 
algal mats found in the fish excluding treatment types.  The fact that there was no 
significant difference between the AA all accessible and FA fish accessible/bird 
exclosure treatment types effectively means that a disturbance effect from birds on algae 
could not be differentiated from that of fish alone. 
 
Figure 3.13  1997?1998 combined algae mean percent cover. Means are Tukey adjusted LS means + 
SE. Means with different letters are significantly different (p < 0.05). Means sharing the same letter 
are not significantly different. (AA) all accessible control treatment; (FA) fish accessible/bird 
exclosure treatment; (BA) bird accessible/fish exclosure treatment; (TX) total exclosure treatment. 
 
 
 
56 
 
Differences of least squares means found significant effects of season for all 
treatments between three of four seasons across both years (Figure 3.14).  This result 
suggests the differences in algal percent coverage due to seasonality can be explained by 
differences in photoperiod and temperature, and the physical effects of each of these on 
fish feeding and movement habits that would impact algae grazing and disturbance.  
There was a significant effect of season on algae percent coverage on all treatments: 
between the Fall and Spring (p < 0.01); Fall and Winter (p < 0. 01); Spring and Summer 
(p = 0.02); Spring and Winter (p < 0. 01); and Summer and Winter (p < 0. 01).  The two 
seasons that were not significantly different from each other were with respect to algae 
coverage on all exclosure treatments were Fall and Summer (p = 0.33). 
 
Figure 3.14  Figure 3.12. 1997?1998 combined algae mean cover by season. Cover are Tukey adjusted 
LS means + SE. Means not sharing common letters are significantly different (p < 0.05). Means 
sharing common letters are not significantly different. 
 
 
57 
 
3.6.4 Emergent Macrophytes 
The emergent macrophytes, Pontederia cordata and Polygonum sp., were 
observed to volunteer in several of the exclosures beginning in July of 1997 (Figure 
3.15). The presence of Pontederia or Polygonum clonal groups from each exclosure 
treatment were noted as they developed throughout the growing season (Figure 
3.16)(Appendix A, Table 3.11). By August, Polygonum was found in each exclosure type 
with TX total exclosures and BA bird accessible/fish exclosures having greater numbers 
than the AA all accessible control or the FA fish accessible/bird exclosures (Figure 3.17).  
By the end of October, no Polygonum was found in any of the exclosure treatments.  
Pontederia also volunteered prolifically in both the TX total exclosures and BA bird 
accessible/fish exclosures throughout each of the growing season observations yet not at 
all in the any of the AA control exclosures (Figure 3.18).  In July there were 17 
Pontederia volunteers with the same total number found in August. September and 
October each had 16 and 15 individual plants persisting respectively. Only in July did 
two Pontederia individuals volunteer in one FA fish accessible/bird exclosure but they 
both disappeared from this exclosure type throughout the rest of the season.  
In 1998, intentionally planted Pontederia individuals (5 in each exclosure type 
and block, 45 total) were observed for presence/absence twice within the week after 
planting and 5 times after that throughout the growing season (Appendix A, Table 3.12).  
Within the first week after planting on June 3rd, only three of 15 individuals from the AA 
all accessible control plots remained, and these three were stems only with no leaves.  
While the control plots retained only 20% of their initial plantings, both the TX total 
 
 
58 
 
 
Figure 3.15  View of 1997 total exclosure (TX) treatment plot with young Pontederia 
cordata clonal volunteers.   
 
 
Figure 3.16  View of 1997 mid-summer exclosure plot Pontederia cordata and Polygonum 
sp. clonal volunteers. The 1m x1m wood frame on the left is the algae cover 
measuring quadrant. 
 
 
 
59 
 
 
 
Figure 3.17  1997 Kenilworth Mudflat exclosure Polygonum sp. volunteerism grouped by treatment 
block.  A B and C are replicate treatment blocks. 
 
 
Figure 3.18  1997 Kenilworth mudflat exclosure Pontederia cordata volunteerism grouped by 
treatment block.  A B and C are replicate treatment blocks. 
  
 
 
60 
 
exclosure and FA fish accessible exclosures maintained 100% of their plantings 
throughout the month of June (Figure 3.19).  By July, the control treatment plots that 
provided complete access to fish and birds had 0% Pontederia plants remaining, while 
FA fish accessible/bird excluded plantings dropped to 60% and TX total exclosure 
treatment plots still retained 100% of their planting survival.  At the end of the 
presence/absence survey the TX total exclosure plots retained 100% of their plantings 
that had grown well in all plots throughout the growing season while the FA fish 
accessible/bird exclosure plots retained only 46% of their initial numbers and most were 
visibly stunted in growth and had noticeable grazer damage to leaves and stems.  Canada 
geese Branta canadensis were the likely grazer in the AA control treatment plots, as 
evidenced by numerous easily identifiable goose tracks in the mud within the plot areas 
(Appendix A, Table 3.13).  For the FA fish accessible treatment plots, geese were 
definitely excluded but the grazer was not as easily identified. Likely grazers or 
bioturbators included common carp Cyprinus carpio, and/or one or more of several turtle 
species that inhabit the area and were occasionally found inside of exclosure plots 
(Appendix A, Table 3.14). 
 
 
61 
 
 
Figure 3.19  1998 Kenilworth mudflat exclosure treatment Pontederia cordata experimental planting 
survivorship. 
 
The mudflat exclosure Pontederia cordata biomass data was collected at the end 
of the growing season in 1998 (Appendix A, Table 3.15). There were inherently obvious 
significant differences between each of the non-fenced control AA plots that had no 
plants present and so no biomass, and the TX total exclosure and FA fish accessible/bird 
exclosure treatment plots which had biomass.  There were significant differences 
identified in dry weight biomass (g/m2) lease squares means between the AA control 
plots and the TX total exclosure treatment plots (p < 0.01) as well as between the AA 
control plots and the FA fish accessible/bird exclosure treatment plots (p < 0.01) (Figure 
3.20).  Especially notable is that there was also a significant difference in mean dry 
weight biomass between the FA fish accessible/bird exclosure treatment plots and the TX 
total exclosure treatment plots (p <0.01).  Tukey adjustments made to the data found the 
same significant differences between the treatments with all adjusted least squared means 
 
 
62 
 
with p <0.01. 
 
Figure 3.20  1998 Kenilworth mudflat Pontederia cordata experimental planting above ground 
biomasss LS means + SE. Different letters denote significant difference in biomass (p < 0.0001). The 
AA treatment type was significantly different than FA or TX with no variance within the exclosure 
treatment between the AA treatment blocks. 
3.7 Discussion 
3.7.1 Bird and Fish Predation on Mudflat Invertebrates 
The experimental exclosure research was originally initiated to determine the use 
and impact of bird and fish predation, grazing or bioturbation on mudflat benthic 
macroinvertebrate communities and algal coverage (Quammen, 1984; Reise, 1985). Birds 
in tidal freshwater marshes are known to be significant with respect to species diversity 
and the number of guilds, in large part due to the variety and complexity of habitat types 
(literature review by Odum et al., 1984) that always includes a degree of unvegetated 
 
 
63 
 
intertidal mudflat area.  Some migratory shorebird species and many species of waterfowl 
are attracted to the open spaces of mudflats in which to feed, rest, and avoid predation 
(Recher, 1966; Quammen, 1984; Riese, 1985; Hayman et al., 1986; Evans 1987; Bull and 
Farrand, 1994; Perry and Deller, 1996).   
According to the Weller?s ?hemimarsh? concept (1978), an equal mixture of open 
water and interspersed emergent plant cover in freshwater marshes are most productive 
for waterfowl use. While this concept was applied to non-tidal marshes, the 
understanding that a heterogenous mixture of wetland conditions maximizes bird 
productivity may still be applied in the case of tidal marshes that include mudflats, which 
attract shorebirds and wading birds as well as waterfowl. Maximizing edge effects as a 
means of improving wildlife has long been thought to be important in game management 
(Leopold, 1933).  In the case of shorebirds and some waterfowl this must be tempered by 
the need for the predation protection of wide edge transitions that expansive mudflat 
systems often afford between open water subtidal and emergent marsh and upland 
vegetation (Recher, 1966; Hayman et al., 1986). More than protection for some bird 
species, mudflats provide aquatic invertebrates, algae, fish, mollusks (Quammen, 1984; 
Riese, 1985), and a variety of plant matter from seeds to tubers to whole plants and 
detritus that are consumed by a variety of species from multiple trophic levels illustrated 
in Figure 3.21 (Lodge, 1991).   
 
 
64 
 
 
Figure 3.21  Generalized interactions between components of the intertidal mudflat exclosure area at 
the Kenilworth MF3.  Thin layers are likely interactions (sometimes with intermediaries not labeled) 
that were not directly observed or revealed through the exclosures.  Bold lines are interactions that 
were directly observed on the mudflat or revealed through exclosures. 
 
Excluding birds was found to have significant positive effect on the oligochaete 
aquatic worm population densities (mean #/m2) when compared to the control areas 
(Figure 3.11). The total exclusion effect on these invertebrates was also significantly 
noticeable with regard to the areas that allowed fish access to the study mudflat but 
excluded birds.  The experiments did not detect a significant difference in oligochaete 
population densities between the areas that allowed fish and bird access (unfenced 
control) and those plots that only allowed fish access.  This is interesting because it 
effectively reveals that fish predation is the primary driver of oligochaete population 
control on the mudflats in this study area, more so than the combined effects of bird and 
fish predation which was expected to have a greater effect than fish predation alone but 
did not.  
The numerous species of midge larvae, family Chironomidae, are important 
 
 
65 
 
trophic links in wetlands, consuming algae and detritus and being consumed by other 
invertebrates, fish (Batzer, 1998) and birds (Batzer et al., 1993). The densities of 
chironomids collected at the Kenilworth mudflat were not found to be significantly 
different among the different exclosure plot treatment types (Figure 3.11). Alternatively, 
as a percentage of invertebrate group composition within each exclosure type, 
chironomids actually decreased significantly in both years with greater degrees of 
exclosure (Figures 3.8, 3.10).  This is counter to what was originally hypothesized when 
the study design was created.   
There are numerous species in the family Chrionomidae (Pennak, 1978), with 
many found on the study mudflat of a size and coloration that would seem to make them 
very noticeable to the keen eyes of shorebirds (Evans, 1987) and some fishes (Batzer, 
1998).  It was hypothesized that the combination of shorebird and fish exclusion from the 
mudflat, releasing them from predation pressure, would allow for significant increases of 
chironomids given their ability to regenerate quickly and in high numbers.  This was not 
the case.  
3.7.2 Algae Cover and Invertebrate Densities 
There was an expectation that significantly increased algal coverage found in total 
exclosures over the control plots and fish only accessible plots would stimulate 
chironomid invertebrate populations with an abundant food source.  To the contrary, 
chironomid group percentages within exclosures (Figures 3.10, 3.12) appeared to be 
inversely proportional to algae percent cover within the same exclosures (Figure 3.13), 
With many species of fish and birds known to utilize chironomids as a food source and 
 
 
66 
 
important trophic link between algae and larger predators (Batzer, 1998), the results of 
the exclosure experiments appeared confounding.   
It is possible that the dense mats of algae that formed in total exclosures did have 
significant numbers of chironomids.  These algal mats were observed to become so dense 
that they trapped pockets of gas bubbles under them (methane from the sediments or 
oxygen from the algae), and on subsequent incoming tides the mats would reach a critical 
density peel up, and float to exclosure edges.  It is likely that chironomids went with 
these mats periodically leaving relatively bare mudflat still inhabited by the sub-surface 
dwelling oligochaete worms.  It is also possible that in contrast to the lighter colored, 
higher albedo mud and less algae in control plots, the much darker, low albedo and dense 
blue-green filamentous mats created some kind of physical threshold from heat or oxygen 
depletion beyond which these species of chironomid could not tolerate.  Was this a case 
of an artificially induced bottom-up control of algae on a specific invertebrate, or an 
unforeseen result of the lack of a top-down grazer control exposed by the exclosure 
experiment? (Power, 1990,1992).   
The enclosure/exclosure experiments of adult and larval fish conducted by Batzer 
(1998) found that while fish were consuming chironomids, exclusion of fish actually 
harmed midge populations rather than benefiting them.  Their conclusion was that the 
fish were also consuming other invertebrates that were predators or competitors of the 
chironomids which had an indirect positive feedback to the midges greater than the direct 
effect of fish predation on the midges directly.  This emphasizes the usefulness of 
exclosures in uncovering species interactions, and the difficulty in determining trophic 
level influences, even in the seemingly simple tidal freshwater mudflat system. The 
 
 
67 
 
Kenilworth mudflat interaction between chironomids and mudflat algae should be 
investigated further, as these findings and those of the above studies contradict the 
findings of others that reduced predation and increased algal coverage should enhance 
some invertebrate populations (Campeau et al., 1994; Rader and Richardson, 1994).   
3.7.3 Mudflat Invertebrate Population Crash? 
Another interesting result of the invertebrate exclosure study was the 89% 
decrease in total oligochaete numbers and 92% reduction in total chironomid numbers 
across all treatment plots between 1997 and 1998 yearly total invertebrate counts 
(Appendix A, Tables 3.7 & 3.8). Severe flooding of the Anacostia River in January-
February of 1998 effectively kept the mudflat permanently submerged for several weeks 
with unusually high water levels for several months (Figure 3.22).  
 
Figure 3.22  Water levels depicting higher levels in 1998 over 1997.  
 
 
 
68 
 
While it is not certain why this would have an effect on the oligochaete and 
chironomid invertebrate communities, if it did at all, it was the only outstanding event 
that occurred in 1998 that might explain the drastic reduction in the populations of these 
two groups.  Corbicula was not negatively affected and had a 28% increase in total 
numbers.   
Perhaps the effective temporary conversion of the study mudflat from an intertidal 
to a subtidal zone allowed greater invertebrate predation by fish, although these 
invertebrate communities are known for rapid population increases within short time 
periods and should have recovered to be identified in the monthly sampling throughout 
the year.  Ice shear may have been a factor, or the greater deposition of sediment on the 
study mudflat from the flooding. It is also possible that the 1998 numbers were the 
?normal? populations for these two groups and that some effect in 1997 caused greatly 
increased populations of the two groups to occur.  Without longer term invertebrate 
sampling it is difficult to determine. 
3.7.4 Strong Fish Influence Over Invertebrates 
Direct observations of feeding and evidence of shorebird and goose tracks within 
openly accessible control plots revealed that shorebirds and waterfowl were utilizing the 
areas (Appendix A, Table 3.13). While shorebird predation of invertebrates on mudflats 
is well documented (Quammen, 1984; Riese, 1985, 2001), the evidence from the 
treatment exclosures at the Kenilworth study mudflat indicates that fish exert a stronger 
effect on the oligochaete and chironomid invertebrate populations than shorebirds as 
invertebrate densities were not significantly different between the control AA and fish 
 
 
69 
 
accessible FA treatments.  If the numbers of these invertebrates were significantly lower 
in the control treatment than the fish accessible only treatment, assumptions could be 
made that shorebirds have a strong effect on Kenilworth oligochaete and chironomid 
invertebrate populations.  Although shorebird utilization of the Kenilworth mudflat was 
documented by regular bird surveys, observations of significant numbers of fish (most 
notably Fundulus sp. swarming the mudflats on the edge of flood tides) are most likely 
the driving force affecting the oligochaete and chironomid populations at Kenilworth.  
 Herons were also using the site as successful feeding grounds for fish during tidal 
exchanges.  Predation risk for fish by birds can be great (Power, 1984; Crowder et al., 
1997), but intertidal food sources such as abundant oligochaetes and piscivorous fish 
predator avoidance appear to be an irresistible draw for the Fundulus sp. and other small 
fishes (Rozas and Odum, 1987; McIvor and Odum, 1988; Rozas et al., 1988; Horn et al., 
1999).   
Tidal freshwater marshes are important spawning, nursery and feeding habitat for 
numerous resident freshwater and estuarine fishes, as well as anadromous, semi-
anadromous and catadromous species of fish (Odum et al., 1984; Rozas and Odum, 1987, 
Rozas et al.,1988; Mitsch and Gosselink, 2000). Electrofishing of the Kenilworth restored 
marsh Mass Fill 1 &2 areas, the Kenilworth Mass Fill 3 mudflat area, and the Dueling 
Creek reference marsh all revealed similarities among them in the shared species of fish 
collected (Table 3.8).  There are, however some differences in average numbers of 
particular species and total fish collected as well as the relative abundances of specific 
species of fish (Tables 3.3 to 3.6).   
In terms of average numbers of fish collected, electrofishing at the two restored 
 
 
70 
 
marsh (MF1 & MF2) and one mudflat (MF3) site were dominated by the omnivorous 
bottom feeding brown bullhead catfish Ameiurus nebulosus, and the gizzard shad 
Dorosoma cepedianum, a planktivore (Figures 3.5 to 3.7).  Alternatively, the Dueling 
Creek marsh reference site was dominated by schools of young of the year, migratory 
blueback herring Alosa aestivalis, the semi-migratory piscivore white perch Morone 
americana, and the small prey fish the eastern silvery minnow Hybognathus regius.   
Dueling also supported greater average numbers of the top end tertiary consumer the 
largemouth bass Micropterus salmoides (4.8 fish/survey) when compared to the three 
other Kenilworth sites combined (0.7 fish/survey) (Tables 3.3 to 3.6).  These numbers 
may suggest that the Dueling reference site reflects the fish composition of what might be 
expected of a more mature marsh, with a greater balance of species and trophic level 
interactions.  The Kenilworth sites were surveyed for fish when the restored marsh was 
less than 5 years old.  Kenilworth had existed almost completely as an intertidal mudflat 
prior to restoration.  The restored marsh may still be living with the ecological memory 
(Peterson, 2002) of a mudflat and will take some time for the marsh to develop and the 
fish community to respond. 
Experimental evidence from the Kenilworth fish exclusion treatment plots, when 
compared to the fish accessible plots, suggests that benthic feeding fish found in 
electrofishing surveys from the families Fundulidae, Cyprinidae, and Ictaluridae, were a 
dominant predation, grazing and disturbance factor on the mudflats with regard to 
invertebrates and algae.   
 
 
71 
 
3.7.5 Algal Grazer/Disturbance Activity 
Algal primary production is considered a potentially important energy base for 
tidal marsh food webs (Odum et al., 1984; Stribling and Cornwell, 1997; Mitsch and 
Gosselink, 2000; Currin et al., 2003). It is interesting that grazer pressure on primary 
production at the study mudflat seems to be proportionately related to size, with 
invertebrates and fish affecting a relatively greater influence on the algae than birds, 
while some waterfowl, specifically Canada geese Branta canadansis, appear to have a 
much greater influence on emergent macrophyte growth.  To a different degree the geese 
influence algae as more of a disturbance factor from trampling across exposed mudflats 
and shoveling up mud in search of macrophyte roots, tubers and seeds (Belanger and 
Bedard, 1994).  The initial hypothesis was that there may be an additive effect of bird and 
fish disturbance and grazing factors affecting algae coverage on the mudflat.  Fish 
grazing and disturbance on algae was expected to be from common carp Cyprinis carpio, 
and brown bullhead catfish Ameiurus nebulosus, which both have bottom-feeding 
behaviors (Rohde et al., 1994). The effect of killifish on algae growth may also be 
significant, and for those that have observed them closely possibly analogous to a wave 
of roto-tillers turning over the thin surface layer of the mudflat in their mouths and with 
their tails on the edge of flooding tides while searching for food. 
Although the bird accessible control plots had lower percent cover of algae across 
both years when compared to fish only accessible treatment plots, the effect was not 
statistically significant (Figure 3.13).  Fish only accessible treatments did have a higher 
algae percent cover than the control plots but less than the total exclosures, supporting the 
 
 
72 
 
differential effect of grazing and disturbance by fish and birds.  Because the bird 
accessible/ fish exclosure was not successful in allowing birds into the plots, the reverse 
effect of fish accessible/bird exclosure could not be tested.  What is certain though, is that 
by excluding the grazing and bioperturbation effects of fish and geese from the algae 
growing on the study mudflat, the result would be significantly greater algal coverage.  
3.7.6 Response of Emergent Macrophytes to Birds and Fish 
Bird surveys conducted at the study mudflat were clear in which species were the 
most dominant in total numbers, and likely to be the most important in their effect on 
macrophyte emergence and survivorship.  The Canada goose Branta canadensis, (both 
resident and migratory) was the most numerous bird observed on and around the mudflat 
study area at Kenilworth (Table 3.1). Large waterfowl grazers such as these have been 
known to have a significant impact on emergent macrophyte vegetation when their 
populations become too much for the natural areas they frequent to sustain (Smith and 
Odum, 1981; Lodge, 1991; Kotanen and Jefferies, 1997).  
The unexpected emergence of volunteer plants within exclosure plots and not 
within control plots gave impetus to a controlled intentional planting study which yielded 
results similar to that of the algae. Differences were expected between the non-fenced 
control treatment plots and the total exclosure treatment plots. Almost immediately all of 
the intentional plantings in each of the control plots had completely disappeared within 2 
weeks of their placement (Figure 3.19). There was some question as to whether the fish 
accessible/bird exclosure treatment plots would respond similarly to the control 
treatments, as both were completely open and accessible to carp Cyprinus carpio, which 
 
 
73 
 
were known to be utilizing the area and have been demonstrated to be a disturbance 
factor (King et al., 1997; Lougheed et al., 1998).  The greater numbers of volunteers 
found within block C over the other two block replicates may be due to slightly higher 
elevations within the C block or its proximity to the tidal inlet. 
Interestingly, the separate and combined effects of fish and birds on algae was 
also seen with emergent macrophytes. The fact that all planted plugs persisted within the 
fish accessible/ bird exclosure treatment plots for several weeks suggested that the 
exclosure type might respond more like the total exclosure plots which also had no loss 
of plants.   While this was initially the case, regular visits to the site noted persistent 
grazer pressure on the plants within the fish accessible treatments in the form of missing 
leaves and stalks as well as the complete disappearance of several of the planted plugs 
within each of the those exclosure plots.    
Additionally, there were also notable differences in the growth and vigor of the 
plants from the total exclosure plots, of which every planted plug survived and grew well 
without any evidence of grazing. By comparison, the fish accessible/ bird exclosure 
plants appeared stunted and often had stalks with no leaves from grazing.  There were 
significant differences in the dry weight biomass between the total exclosure and fish 
only accessible treatment types (Figure 3.20).  This could be accounted for by the 
repeated stress of the grazed plants within the fish only accessible plots undergoing the 
stress of having to regenerate biomass while the ungrazed plants were free to grow more 
vigorously without the stress.   Basically, if there was no exclosure barrier to geese, then 
no plants survived at the exclosure mudflat elevations. If fish could access the biogenic 
mudflat but not geese there was some survivorship and moderate biomass while if there 
 
 
74 
 
was no access by fish or geese then there was complete survivorship and significantly 
more biomass then either of the other two treatments that allowed some access by 
grazers. 
3.7.7 River Turtle Influence 
While the most likely grazers to be impacting the fish accessible/bird exclosure 
plots were thought to be carp, turtles such as the painted Chrysemys picta picta, red-eared 
slider Chrysemys scripta elegans, and snapping turtle Chelydra serpintia, were also 
found within exclosures and noted around the mudflat study area (Appendix A, Table 
3.14).  It is believed that they could have also been opportunistically grazing on the 
young shoots, leaves and regrowth of the planted plugs, contributing to the stress and 
reduced dry weight biomass of the macrophytes from the fish accessible treatment plots. 
Evidence of river turtle use of the mudflats was unexpectedly observed on the 
study mudflat and within exclosure plots (Appendix A, Tables 3.14 & 3.16).  Their 
effects on the mudflat are not certain, although when occasionally found trapped in an 
exclosure during spring high tides the disturbance to the biogenic mudflat algae was 
noticeable as they slid about the exclosure searching for an exit at low tide.  Many 
species of  turtles are known to be omnivores, and could be a factor in the grazing or 
bioturbation caused removal of the experimental plantings in the biogenic mudflat 
exclosures (Moll and Moll, 2004).  There is certainly a need for a greater understanding 
of the role that turtles have in the Anacostia River, as it may be significant (Iverson, 
1982; Congdon and Gibbons, 1986; Moll and Moll, 2004).    
 
 
75 
 
3.7.8 Mollusks and Mammals on the Mudflats 
Identified use of the study mudflat by mammals, primarily raccoons Procyon 
lotor, was revealed by obvious signs of digging with tracks found around holes in the 
mud.  The Asiatic clam Corbicula fluminea, are a likely subject of the search as they 
were found in numbers within each of the exclosure treatments (Figure 3.11) and empty 
shells were found scattered around sites where a raccoon had obviously been digging.  
Empty shells of the freshwater mussel Anadonta sp. were also found, although as many 
of the genus with the common name ?floater? suggests, they could have come in on the 
tide as they were not found in subsurface sediments of the study muflat but were found in 
tidal flats on other parts of the Anacostia River.  The apple or ?Asian mystery? snail of 
the family Vivipridae was also regularly found within or near the study exclosures, and 
occasionally Planorbid or Physid snails were also identified (Appendix A, Table 3.17). 
The snails could have also been prey for raccoons as well as some wading birds, fish or 
carnivorous waterfowl. They were not found in numbers on the study mudflat that would 
suggest they are a significant factor in algae grazing and they may be suppressed by 
predation on the exposed flats.  The relative lack of snails within exclosures may be a 
result of their inability to reach the refugia in any numbers without being eaten.  Larger 
snails would also have trouble penetrating the small mesh size. 
3.7.9 The Most Interesting Animal 
Probably the most interesting animal(s) found on the mudflat study area, and 
within several of the exclosure plots, were the colonies of the phylum Ectoprocta, the 
bryzoan Pectinatella magnifica (Appendix A, Table 3.18).  A freshwater species of the 
 
 
76 
 
more common marine forms, they are structured and function similarly to corals but are 
not related.   These animals were likely carried in on the tide and settled on the study site 
adding another interesting, albeit incidental, species to the ecology of the mudflat. 
3.8 Conclusions 
The Kenilworth Marsh experimental biogenic mudflat exclosure study determined 
that fish and birds had differing impacts and unforeseen influences on benthic 
macroinvertebrate densities and algae coverage as well as macrophyte emergence, 
survivability and biomass.  The exclosure experiments followed for two years at 
Kenilworth were proven to be an important tool in revealing the sometimes complex and 
unexpected interrelationships of the intertidal mudflat food web.  Several basic 
interactions are now better understood and the case for a greater appreciation of these 
systems can be made. 
The theory of alternate stable states was first considered with the voluntary 
emergence of plants within the exclosure study plots.  The study mudflat, which had 
existed in an unvegetated state for over 3 years, was released from plant grazing and 
disturbance pressure by fish, possibly river turtles, and most certainly geese.  The 1998 
study of intentional plantings of emergent macrophytes revealed a stark and significant 
contrast between total, partial and non-excluded control treatment plots. What is 
interesting is that there appears to be a gradient of grazer pressure that would have not 
been realized without the partial ?fish only accessible? exclosure treatment.   
The removal of exclosure treatment fencing in 1999 revealed that even a 
completely vegetated 7m x 7m plot of almost 2m high Pontederia cordata (Figure 3.23) 
 
 
77 
 
could be completely removed by geese within a week. Within two weeks there would be 
almost no evidence of roots or tubers as the geese would have dug them out, leaving 
nothing but their numerous tracks in the mud at low tide (Figure 3.24). No regeneration 
of the plants occurred in the following years after the exclosure fencing was removed. 
 
Figure 3.23  Kenilworth exclosure filled with Pontederia cordata. The exclosure to the right is fish 
accessible.   
 
 
 
78 
 
 
Figure 3.24  Goose tracks covering the mudflat study area at Kenilworth Mass Fill 3. 
 
 
This complete change in condition, from biogenic mudflat to emergent marsh and 
back again to biogenic mudflat, exemplifies several of the tenets of alternate state theory.  
Alternate state changes are often sudden, obvious and triggered by some external 
mechanism. In this case the triggers were exclosure fencing, and goose grazer pressure.  
In a review article 30 years ago, May (1977) stated that he felt without empirical 
evidence, multiple or alternate species states would remain a theory, and evidence would 
be largely metaphorical. It is believed that this study of a tidal freshwater biogenic 
mudflat was able to add empirically to the debate over alternate state theory, and in this 
case move the theory from the metaphorical to the verifiable.
 
 
79 
 
Chapter 4: Kingman Lake/Marsh Experimental Ecology 
4.1 Introduction 
The area that was known as Kingman Lake once existed as part of the expansive 
emergent marshes that historically dominated the Anacostia River. Coues and Prentiss 
(1883) describe the marshes ?for three miles above the railroad bridge the channel winds 
tortuously between extensive marshes composed of wild rice (Zizania aquatica)?? 
which were renowned for seasonal bird hunting (Figure 4.1).  The silting in of the river 
and the threat of malaria from mosquitoes spurred plans to dredge the river marshes to 
create a ?lake? for recreational boating, which was undertaken by the U.S. Army Corps 
of Engineers (USACE) between the 1920?s and late 1940?s (Figure 4.2)(USACE, 1913; 
Syphax and Hammerschlag, 1995). After the dredging of the marshes and straightening 
of the river was completed, Kingman Lake began filling in with decades of deposited 
sediment from erosion in the watershed to the river. Intertidal mudflats formed and 
became the dominant feature of the Kingman area at low tide.  
 
 
80 
 
 
Figure 4.1  Nineteenth-century image of rail hunting on Anacostia River wild rice marshes (Coues 
and Prentiss, 1883). 
 
 
 
 
Figure 4.2 U.S. Army Corps of Engineers 1913 map of existing and planned Anacostia River 
shorelines. 
 
 
In 2000, intertidal portions of Kingman Lake were reconstructed by the U.S. 
Army Corps of Engineers to restore the historic emergent marshes.  With management 
oversight by the U.S. National Park Service and the District of Columbia City 
 
 
81 
 
Government, Kingman Lake would become Kingman Marsh.  Dredged river sediments 
from the Anacostia River channel were used to increase mudflat surface elevations to 
levels that would support a restored emergent marsh similar to that of the Kenilworth 
Marsh restoration effort that was completed in 1993. For Kenilworth Marsh, a range of 
target elevations above mean sea level was designed for low (1.75?), middle (2.1?), and 
high (2.5?) marsh NGVD ?29 (National Geodetic Vertical Datum based on mean tide 
levels around the year 1929) (Syphax and Hammerschlag, 1995). Learning from the 
results of the initial Kenilworth Marsh vegetation patterns that revealed a high degree of 
high marsh volunteer native and invasive species (Guerrero and Hille, 1994), the 
Kingman Marsh restoration design elevations were planned to be lower (2.0? NGVD ?29) 
to supress the establishment of invasive non-native species such as purple loosestrife 
(Lythrum salicaria) and common reed (Phragmites australis), that were found to be a 
problem at Kenilworth with higher elevations (Hammerschlag et al., 2006). Optimal 
surface elevations for emergent marsh establishment and macrophyte species diversity 
and cover was determined to be at 2.1? NGVD ?29, which on the Anacostia River is 
approximately mean high tide (Bowers, 1993; Hammerschlag et al., 2006). Tidal 
channels or ?guts? were cut into the planted restoration areas to allow for river waters to 
access the restored marsh areas. 
Approximately 13 hectares of marsh were restored with 700,000 individual 
macrophyte plugs installed comprised of 6 native species spaced evenly throughout the 
area. The restoration areas were defined as Kingman Marsh ?Cell? 1 and ?Cell? 2, with a 
majority of the restoration area (11 ha) located in Cell 1 (Figure 4.3).  While marsh 
restoration planting was taking place, resident Canada geese Branta canadensis maxima, 
 
 
82 
 
(non-migratory descendants of captive birds released in the early 20th Century) began 
grazing newly planted material and triggered adaptive management decisions by 
managers to erect goose exclusion fencing around the perimeter and within the interior of 
the marsh restoration areas (Hammerschlag et al., 2001). With goose exclusion fencing in 
place, by the end of the first growing season in 2000 average emergent marsh plant cover 
of both restoration Cells was considered successful with nearly100% cover attained 
(Hammerschlag et al., 2001).  
 
Figure 4.3  Locations of Anacostia River marsh restoration projects with year of completion and 
Dueling Creek reference area. Kingman Marsh Cell 1 restoration area at the center bottom. Image 
from USGS Patuxent Wildlife Research Center, 2003. 
 
By September of 2000 the marsh vegetation in Cell 1 was colonized by numerous 
species that had not been planted, including small numbers of the non-natives Lythrum 
salicaria and Phragmites australis.  Species planted for the restoration included, 
Pontederia cordata, Sagittaria latifolia, Peltandra virginica, Schoenoplectus 
tabernaemontani, Juncus effusus, and Nuphar luteum. With the removal of goose 
exclosure fencing in early 2001, percent cover at both restoration areas of Kingman 
declined 40-60% (Hammerschlag et al., 2002).  Subsequent years cover also declined 
 
 
83 
 
considerably, a result of combined goose grazer pressure and the erosion and settling of 
the dredged river sediments resulting in lower surface elevations than had been planned 
(Hammerschlag et al., 2006).  It was estimated that roughly 50%, or between 300,000 and 
400,000 plants installed for the restoration were grazed upon or pulled out by the geese, 
preferentially consuming Sagittaria and Pontederia, although the young shoots of other 
typically non-palatable species such as Peltandra have been known to be eaten (Harris, 
2002). 
In 2000, as a result of the unexpectedly intense goose herbivory found during the 
restoration planting, a preliminary plan for the implementation of an experimental goose 
exclosure study was developed.  The study was influenced by the results of the exclosure 
studies conducted at Kenilworth Marsh that found geese to be the limiting factor in 
voluntary macrophyte emergence (Chapter 3).  The Kingman Marsh exclosures 
comprised a much larger area of potentially biogenic mudflat, with the focus of the study 
on the effects of goose grazer pressures suppressing emergent marsh regeneration. The 
exclosures were sited in an area of Cell 1 that had not been previously planted and 
remained a mudflat (Figure 4.4).   
In early June of 2001, the experimental exclosures were constructed.  While the 
entire experimental exclosure study area (including unfenced control plots) at Kenilworth 
comprised 588m2, the Kingman study area covered 2,700m2 .  The exclosures in this area, 
which existed as an intertidal mudflat with surface elevations ranging from 1.3? to 2.0? 
NGVD ?29 (USGS Survey, 2004), were maintained for 4 years to follow the emergent 
macrophyte community that immediately developed under the exclusion of goose grazer 
pressures.    
 
 
84 
 
 
Figure 4.4  Upper Kingman Marsh Cell 1 restoration area with study site locations. 
 
 
 
85 
 
Between in 1997 and 2001 bird surveys were conducted on the mudflats of 
Kingman Cell 1 as well as the golf course on Kingman Island.  These surveys revealed an 
increasing trend in the numbers of geese that resided in and around Kingman Lake that 
was a factor in the restoration of Kingman Marsh (Figure 4.5). Fish surveys of the 
mudflats of Kingman Cell 1 were also conducted beginning in 1996.   
R2 = 0.1268
0
100
200
300
400
500
600
700
800
Survey Years 1997-2001
# 
Ge
es
e
DC Goose Survey Linear (DC Goose Survey)  
Figure 4.5  Kingman Lake/Marsh Canada goose counts showing increasing numbers found on the 
mudflats, open water, and golf course around Upper Kingman Cell 1. 
 
The effects of the exclosure fencing in stimulating the mudflats toward the 
emergence of macrophytes, confirmed the role of herbivore grazing at Kingman Marsh as 
a strong top down force in controlling macrophyte emergence, survivorship and growth 
(Power, 1988,1992).  Just as the results of the Kenilworth exclosure study inferred, the 
Kingman exclosures were also expected to support the potential for the low-middle marsh 
intertidal mudflat areas on the Anacostia River to be models of the rationale behind 
alternate state theory. As already stated in Chapters 2 and 3, this theory has been 
forwarded as a way to describe the parallel potential, yet alternate emergence, of different 
 
 
86 
 
community types within a single environmental condition triggered by some internal or 
external force.  In the case of Anacostia River marsh restoration, the triggers were the 
Army Corps of Engineers as a marsh attractor and resident geese as a mudflat attractor 
(Kangas et al., 2004). The alternate state concept, and its application to the results of the 
Kingman mudflat research, would also apply to future emergent marsh restoration on 
Anacostia River. 
4.2 Study Area Description 
The Kingman Marsh (38? 54' 17"N, 76? 57' 42"W) exists as a 45 ha tidal 
impoundment of which approximately 13 ha of emergent marsh was restored in 2000.  It 
is situated just west of the Anacostia River, separated by Kingman Island and connected 
tidally by northern and southern inlets (Figure 4.3).  Kingman Island is bisected by 
Benning Road bridge, with the northern half of the island managed as a golf course by 
the National Park Service and the southern half of the island forested and administered by 
the District of Columbia City Government. Kingman Marsh is less than 1 km below the 
Kenilworth Marsh on the river. The site is influenced by a semi-diurnal freshwater tide 
with an average tidal amplitude of approximately 1.0 m.  At low tide, large areas of 
unvegetated mudflats still exist around and in between the seasonally dense, restored 
vegetated areas. Soil variables for Kingman surface sediments 48% sand, 33% silt and 
19% clay with 6% organic matter and a pH of 7.06 (Neff, 2002).  Kingman Marsh north 
of Benning Road (Kingman Cell 1) consisted of 11 ha of emergent marsh in 2000, and is 
the location of the work completed for this study (Figure 4-4). The exclosure experiment 
area was located on unplanted tidal flats approximately 300m southwest of the northern 
 
 
87 
 
Kingman inlet and immediately north of the marsh restoration planting area.  
4.3 Research Methods and Materials 
4.3.1 Birds 
Birds were observed at the Kingman mudflats beginning with two surveys at one 
site in the fall/winter of 1997 and then twice monthly beginning in spring 1998 
continuing through January of 2001 (Appendix B, Table 4.1). Observations were initially 
made in 1997 from one site (Site #1) at the northern edge of Kingman Island (Figure 4.4). 
In spring of 1998 another site was added (Site #2) from the edge of a wooded area 
approximately 300m south of Site #1 adjacent to the mudflats.  Counts of geese were also 
made from the golf course on Kingman Island and in the open water area just above 
Benning Road Bridge. All bird surveys were conducted at low tide, as low tide counts are 
considered representative of average waterbird usage (Burton et al, 2004). A count of all 
birds within the field of view of the mudflat study area was conducted over a period of 30 
minutes, with each species and their numbers recorded.  Only birds with a water 
connection were identified, with all primarily terrestrial bird species not counted in the 
survey. After an initial count over 15 minutes, the remainder of the time was spent 
observing the behaviors of the birds on the mudflat with additional counts made for new 
birds entering to the site.  All counts of large numbers of birds were double-checked 
during this time period and the lesser of the two numbers used as the official count.   
Basic abundance, numbers of species, frequency and relative importance of each species 
found during the observation period were calculated.   
 
 
88 
 
4.3.2 Fish 
Fish surveys were conducted over two sites (TG1 and TG2) along tidal ?gut? 
(TG) channels within Kingman Lake (Figure 4.4) using a Smith-Root fish 
electroshocking boat operated by the D.C. Government Fisheries Division. Surveys were 
conducted on a flooding high to slack tide during each shocking run. Shocking levels 
were set at 240-500 volts at 3-5 amps, 60 pulses/sec, adjusted for conductivity (D.C. 
Fisheries standard operating procedures).  Approximately 1000 seconds of shocking time 
was conducted along each electrofishing shock line. Each survey was completed through 
two, 100m passes over the same area, once in and once out of the survey line.  A total of 
6 electrofishing surveys were completed (two on each shocking date) covering a spring, 
summer and fall season in 1996 and 1998 (Appendix B, Table 4.2).  All fish collected 
were kept in live wells on the boat and identified to species, counted, measured for length 
and released after shocking was completed.  Game fish and larger fish species were also 
weighed to the nearest whole gram.  
4.3.3 Exclosures 
In early June of 2001, exclosure fencing was installed in an area of Kingman 
Marsh Cell 1 that had been left unplanted and remained mudflat (Figure 4.4). Two 
different sized treatments of fenced exclosure plots and unexclosed control plots with 
three replicates each were constructed. The two different rectangular exclosure sizes 
(15m x 10m and 15m x 20m) established a total of 6 exclosed and 6 unexclosed plots 
with a total area of 2,700m2 (0.27ha), half of which excluded geese (Figure 4.6). Plots 
were identified by size (10 or 20) and excluded (X) or unexcluded (U), with replicates 
 
 
89 
 
 
Figure 4.6 Kingman Marsh exclosure study plots with corner surface elevation points. 
 
identified by the letters A, B, or C.  The two plot sizes were used for the purpose of 
determining the effect that exclosure size would have on goose accessibility and the 
restoration potential of larger sized fenced plots. Vinal coated wire fencing with 3cm 
diameter mesh and 1.2m in height was utilized as exclosure material and staked with 5cm 
x 5cm, 2m wood stakes that were left in the marsh from the restoration work. The 
mudflat surface elevations at the corners of each plot were surveyed by the USGS and 
tied to NGVD ?29 (National Geodetic Vertical Datum based on mean tide levels in the 
year 1929). Plots were arranged with fenced treatment plots sharing common sides to 
maximize available exclosure area with available fencing.  An attempt was made to 
 
 
90 
 
stratify the location of the entire block of all plots on the mudflat so as to incorporate the 
range of visually observable surface elevations while fitting into the available mudflat 
area.  While treatment plots were not random or independent, the primary goal of the 
exclosure plot design was to test the ability of two sizes of exclosures and unfenced 
control plots to support macrophyte emergence.  Exclosure design was determined 
through collaboration with staff of the USGS.   
4.3.3.1  Emergent Macrophytes 
Transects were run from west to east across each exclosure plot from 3, 6, 9, and 
12m along each plots 15m side (Figure 4.6) using a 100m tape.  Species percent cover 
between each meter mark, species presence/height at each meter mark and 4 randomly 
placed quadrats of above ground biomass were collected along each of the four 10m or 
20m transects crossing each exclosure plot.  The macrophyte data were collected once a 
year in late August or September for four years (2001-2004).  These measurements were 
not necessary in unexcluded (control) plots as no plants volunteered in any year and they 
remained 100% biogenic mudflat.  Access across each plot was consistently made along 
the north side of each transect line to reduce trampling effects on plant measurements 
made on the south side of each transect line (Cahill et al, 2001).  Visual estimations of 
cover for each species present were made between each meter mark to 1m south of the 
transect, with percentages rounded to the nearest 5.0%.  Cover from species with less 
than this were marked as having 1% increments.  The estimation of the percent cover of 
mud was also determined. The presence of the species found at each meter mark on the 
transect line and its height above the biogenic mudflat surface was recorded.   
 
 
91 
 
For biomass collections, a random numbers table generated four locations along 
each of the plots 4 transect lines at which a 0.25 m2 aluminum hoop quadrat was laid on 
the south side of the tape.  All surface vegetation falling within the quadrat was cut at 
mud level, collected and bagged with a plot and transect location code.  This system 
accumulated a total of 4m2 of randomized biomass data for each of the 6 treatment plots 
sampled.  Samples were taken to a lab where each was rinsed of mud and separated into 
different species within each sample.  All samples were initially spread out and air dried 
in the lab, then paper bagged and oven dried at 80? C for not less than 48 hours, after 
which they were weighed to the nearest 0.01g and recorded for species and location code.   
4.4 Data Analysis Methods 
Absolute and relative abundance, average numbers of individuals 
observed/survey, frequency and importance values of each species recorded during the 
observation periods were calculated for all bird and fish data.   
Arithmetic means of all macrophyte cover data for each species were calculated 
through the SAS Proc Means procedure plus/minus one standard deviation (SAS version 
9.1, 2006).  Analysis of variance was not conducted on exclosure macrophytes as there 
was not enough data from each species to support an ANOVA and the lack of 
randomization and independence of the plot design violated the assumptions of the 
ANOVA.  The complete lack of any macrophyte cover in the control plots throughout the 
4-year study created an interesting situation with the data for which there was no 
variance, so no ANOVA could be run to compare exclosure treatments to controls.  The 
arithmetic means plus 1 standard deviation were plotted for cover for each species 
 
 
92 
 
measured in the two exclosure sizes for each year.  Cover data was also applied to 
mudflat surface elevations plus/minus 1 standard deviation. Biomass data was handled 
similarly to cover data.  Arithmetic means were used for each species biomass over the 
four year period with means produced from Proc Means. 
Mudflat surface elevation points taken at the corners of each excluded and 
unexcluded plot (Figure 4.6) analyzed using the SAS procedure Proc G3D (SAS, 2006). 
SAS Proc G3D was set to extrapolate the surface area at 1.0 ft. major intervals to match 
the intervals of the surface elevation survey data to create the three dimensional field 
topography of the experimental plot area (Figure 4.7). Plane coordinates for X and Y for 
each transect were created to fit the field as the entire experimental area was essentially 
set up as a grid. Arithmetic means cover data for transect coordinates were matched 
within extrapolated surface elevation ranges (1.25?, 1.35? NGVD ?29, etc.) and plotted. 
 
Figure 4.7  Kingman Exclosure plot area exaggerated sediment surface elevations. 
 
 
 
 
93 
 
Biomass data were analyzed similarly to cover data.  Arithmetic means were used 
for each species biomass over the four year period with means produced from Proc 
Means.  Mean biomass was plotted against elevation between the 1.5-1.7? NGVD ?29 low 
marsh zone to focus on the elevations points where biomass begins to increase.   
For the species presence data a general linear mixed model procedure was used 
(Proc Glimmix), as the data was set up to be binary (1/0, presence/absence) for each 
species.  The Glimmix allows a binary distribution to be modeled, which is not normally 
distributed as opposed to the assumption of a normal distribution for an ANOVA  (SAS, 
2006).  Species height data was applied to the 3D mudflat surface elevation model and 
plotted at their location along each transect within each exclosure plot. 
4.5 Results 
4.5.1 Birds 
Surveys of birds at Kingman were taken at 103 site visits on 34 dates between 
1997 and 2001 (Appendix B, Table 4.1).  Over this period of time, a total of 7,247 were 
counted representing 7 families of birds. Of that number 5,297 were Canada geese Branta 
canadensis (Appendix B, Table 4.2), representing 73% of the total number of birds 
observed (Table 4.1).  Surveys of goose populations in and around the Kingman Marsh 
over the three years leading up to the restoration found an increasing trend in the 
population of resident geese (Figure 4.5).  The second most abundant species of bird 
observed was the ring-billed gull Larus delawarensis, of which 787 were recorded with a 
relative abundance of 11%.  The killdeer Charadrius vociferous, a common inland 
shorebird followed with 719 individuals recorded comprising 10% of the total number  
 
 
94 
 
Table 4.1 Kingman bird surveys 1997-2001 listed in order of importance value (relative abundance + 
frequency / 2) 
 
 
observed.  While less numerous, 6 other species of shorebirds were observed on the 
mudflats. The mallard Anas platyrhynchos and great blue heron Ardea herodias 
completed the 5 species of birds that comprised greater than 1% relative abundance.   
In terms of frequency of occurance, the great blue heron was most often observed 
in surveys appearing 93% of the time ahead of geese at 85%, and ring-billed gull and 
killdeer at 61% and 54% respectively.  While the geese were top rated in their importance 
value at Kingman, the great blue heron was ranked second due in large part to its high 
frequency of occurance.  Other birds of note were the two crow species American and 
fish crow Corvus sp., which were combined due to their difficulty to distinguish at a 
distance, greater yellowlegs Tringa melanoleuca, and two bald eagles Haliaeetus 
leucocephalus.  
 
 
95 
 
4.5.2 Fish 
Fish collected at Kingman Lake prior to marsh restoration included 16 species of 
fish representing 7 families and 1012 individuals collected from 6 electrofishing surveys 
(Appendix B, Table 4.3).  The greatest numbers of fish captured were the brown bullhead 
catfish Ameiurus nebulosus with 414 individuals, followed by pumpkinseed sunfish 
Lepomis gibbosus, and gizzard shad Dorosoma cepedianum.  Brown bullheads had the 
greatest relative abundance (41%), and along with the spottail shiner Notropis hudsonius 
were observed at a frequency of 83% (Table 4.2).  
 
Table 4.2  Kingman Lake mudflat fish community from 6 electrofishing surveys; absolute abundance 
(total # caught); average #/survey; relative abundance (% total #); frequency (% survey sp. 
appearance); ranked by importance value (relative abundance + frequency / 2). 
 
  
 
Pumpkinseed, gizzard shad, and white perch Morone americana, were each 
collected with 100% frequency.  The white perch and striped bass Morone saxatilis were 
the only piscivorous fish captured making up just over 4% of the total catch. Half of the 
species of fish collected made up at least 1% of the total catch (Figure 4.8).  While the 
 
 
96 
 
brown bullhead was ranked highest in importance on the Kingman mudflats, the other 
benthic feeding fishes, common carp Cyprinis carpio and channel catfish Ictalurus 
punctatus were only ranked 8th and 14th out of 16 species. An average of 169 fish were 
caught per survey at Kingman with brown bullheads representing an average of 69 
captured per survey. 
 
Figure 4.8  Kingman mudflat electrofishing survey results. 
 
4.5.3 Exclosure Macrophytes 
4.5.3.1 Cover  
After erection of the Kingman mudflat exclosures, the voluntary emergence of 
multiple species of macrophtes within each of the 6 exclosure plots was contrasted by the 
 
 
97 
 
complete lack of any plant emergence in each of the 6 control unfenced plots (Figure 
4.9).  Total vegetative cover data collected in the exclosures from each plot size between 
2001-2004 are presented in Figure 4.10 (Appendix B, Tables 4.4 and 4.5). Trends in 
combined species arithmetic mean cover for both plot sizes, 10x and 20x, mirrored each 
other across the four years. Mean cover in 2001 began at 20% for the 10x exclosures and 
60% cover for the 20x exclosures.  Each of these mean covers increased in 2002 with the 
10x cover improving to near 40% while the 20x cover had a more modest increase to 
approximately 65%.  
 
Figure 4.9  Kingman plot exclosures.  Direction of generally increasing elevation. 
 
Year 2003 mean cover for both plot sized dropped considerably, and fell again in 
 
 
98 
 
2004 with 10x cover at less than 5% with 20x cover decreasing to just over 30%.  These 
consecutive yearly decreases are likely attributable to multiple storm events occurring 
within 2003, producing flooding that compromised several of the fenced exclosures with 
large woody debris (logs and trees).  When exclosure fences were repaired, tell tale signs 
of goose access to compromised plots included multiple crater like holes in the mudflats 
where feeding on the tubers of the previous years perennial plants had been.  Once 
emergence of the plants had occurred, another storm and breech of several plot fence 
lines in 2003 allowed the resident geese in again, with evidence of the browsing or 
topping of all the young leaves of plants occurring in several areas.  Repeated grazing and 
lack of growth in several of the plots in 2003 likely inhibited re-emergence in many of 
the exclosure plots in 2004 and drove down mean cover values.   
 
Figure 4.10  Kingman exclosure plot total cover arithmetic means and 
standard deviations for all species of plants across four years for each 
exclosure type. 
 
 
99 
 
The top five macrophyte species that emerged with greater than 1% cover in both 
plot sizes were Sagittaria latifolia, Pontederia cordata, Peltandra virginica, Echinochloa 
walteri, and Ludwigia palustris.  Other species identified that did not comprise a notable 
percentage of the macrophyte cover within the exclosures included Leersia oryzoides, 
Polygonum sp., Bidens sp., and instances of the native invasive Typha sp. and the non-
native invasive Lythrum salicaria.  The last two occurred only at two locations early then 
disappeared from later years data collection. It is notable that the two invasive species 
were dominant in areas of higher elevation not far from the study site and yet failed to 
establish within the exclosure plots. 
The two exclosure sizes behaved similarly with regard to species cover changes 
over the four year study period, although they differed in the mean cover values with the 
smaller 10x exclosures generally reflecting less cover per species than the 20x exclosure 
plots (Figures 4.11, 4.12).  In 2001, Pontederia clearly dominated in cover over all other 
species in both plot sizes with Sagittaria and Echinochloa relatively even although less 
than half the cover of Pontederia.  In 2002 and 2003, Sagittaria mean cover surged in 
both plot sizes, while Pontederia and Echinochloa decreased and disappeared from cover 
transects all together by 2004.  The dramatic increase in Sagittaria was mirrored by a 
decline in Pontederia over the same period. While also eventually decreasing in 2004, 
Sagittaria cover dominated all species cover in both plot sizes but for Peltandra cover in 
the 10x plots, which remained somewhat constant across both plot sizes in most years.  
Of note is the dramatic increase in Ludwigia cover in 2002, becoming the dominant 
species in the 10x plots and then practically disappearing in 2003 the following year.   
 
 
 
100 
 
 
Figure 4.11  Kingman combined 10x exclosure plot arithmetic mean cover by species across years 
with 1 standard deviation. 
 
 
Figure 4.12  Kingman combined 20x exclosure plot arithmetic mean cover by species across years 
with 1 standard deviation. 
 
 
101 
 
When the top 5 species mean cover data for both sized plots were combined and 
plotted against mudflat surface elevations for each year, cover for each species was 
translated into the elevation zone where each species was found (Figures 4.13 to 4.16).  
In 2001 the dominance of Pontederia cover over other species was found across all 
elevations, although mean cover peaked between the 1.5?-1.6? NGVD ?29 elevation range 
(Figure 4.13).  Echinochloa and Sagittaria shared roughly equal cover between 1.5?-1.8? 
NGVD ?29, although above 1.9? NGVD ?29 Sagittaria had greater cover but much less 
still than Pontederia.  By 2002 Sagittaria began to increase its cover across several 
elevations, notably between 1.7?-2.0? NGVD ?29.  Ludwigia also showed gradual 
increases in cover between 1.4?-2.0? NGVD ?29, with peak cover between 1.8?-1.9? 
NGVD ?29.  Pontederia cover between 1.4?-1.7? NGVD ?29 was still greater than other 
species at these elevations (Figure 4-14).  By 2003, Sagittaria had moved into cover 
dominance at elevations from 1.3?-2.0? NGVD ?29with other species representing less 
than 5% mean cover at other elevations and Echinochloa disappearing entirely from the 
cover surveys (Figure 4-15).  The trend in Sagittaria cover dominance increased with 
increasing elevation.  Between 1.4?-1.7? NGVD ?29 mean cover ranged from more than 
20% to 30%, while between 1.7?-1.9? NGVD ?29 cover increased to between 40% and 
50%.  At the highest elevation within the exclosure plots, between 1.9?-2.0? NGVD ?29,  
Sagittaria mean cover reached its peak at over 80%.  In 2004, Ludwigia joined 
Echinochloa in loosing cover representation within the exclosure plots, with Sagittaria 
maintaining cover dominance at all elevations and Peltandra and Pontederia existing at 
elevations between 1.5?-1.8? NGVD ?29 at covers of less than 10% (Figure 4.16). 
 
 
102 
 
 
Figure 4.13  2001 combined exclosure plot arithmetic mean cover by species across mudflat surface 
elevation midpoints with 2 standard deviations. Species are represented unless there was no cover. 
 
 
Figure 4.14   2002 combined exclosure plot arithmetic mean cover by species across mudflat surface 
elevation midpoints with 2 standard deviations. Species are represented unless there was no cover. 
 
 
103 
 
 
Figure 4.15  2003 combined exclosure plot arithmetic mean cover by species across mudflat surface 
elevation midpoints with 2 standard deviations. Species are represented unless there was no cover. 
 
 
Figure 4.16  2003 combined exclosure plot arithmetic mean cover by species across mudflat surface 
elevation midpoints with 2 standard deviations. Species are represented unless there was no cover. 
 
 
 
104 
 
4.5.3.2 Biomass 
Four randomized quadrats of biomass data from each of the four transect lines 
crossing each of the 6 exclosure plots were collected once each year between 2001-2004 
(Appendix B, Tables 4.6 and 4.7). Biomass data collected in the exclosures exhibited 
similar trends to those of the macrophyte cover data.  Pontederia and Echinochloa were 
the largest contributors of mean biomass (g/m2) in both exclosure sizes in 2001, with 
Sagittaria and Ludwigia biomass representing to a lesser degree (Figures 4.17 & 4.18).  
By the next year, Sagittaria biomass began a dramatic increase in the 20x exclosures that 
would continue for the rest of the study period.  In 2002, Echinochloa biomass dropped 
considerably in the 20x exclosures and disappeared altogether in the 10x exclosures.  
Peltandra biomass revealed a sharp increase in the 20x exclosures not seen in the 10x 
plots.  Ludwigia surged to become the largest contributor in mean biomass in the 10x 
plots in 2002, just ahead of Pontederia which had increased in the 10x plots but 
decreased in the 20x plots over the previous year.  Like Echinochloa in 2001, Ludwigia 
had reached its peak biomass in 2002 and declined considerably the following year, 
disappearing from random biomass collections completely by 2004.  The mean biomass 
of Peltandra increased in 2003 in the 10x study plots to dominate all other species in 
these exclosures although its numbers in the 20x plots had decreased considerably.  By 
the last survey year in 2004, Peltandra and Sagittaria were the only two species noted in 
biomass collections, with Sagittaria reaching its highest levels in the 20x exclosure plots 
and Peltandra, while reduced over the previous year still holding a dominant edge over 
Sagittaria in the 10x exclosure plots.  Every species represented in the biomass survey 
 
 
105 
 
 
Figure 4.17  Combined 10x exclosure plot biomass by species across years. Data are arithmetic means 
with 1 standard deviation. 
 
 
Figure 4.18 Combined 20x exclosure plot biomass by species across years. Data are arithmetic means 
with 1 standard deviation. 
 
 
 
106 
 
revealed a net decrease in mean biomass by 2004 when compared to the initial 
year biomass levels, with the exception of Sagittaria which reported marked increases in 
total mean biomass, and Peltandra which had higher mean biomass in the intervening 
years but still ended with greater biomass than when the experiment began (Tables 4.3 
and 4.4).   
Table 4.3  20X exclosure plot mean biomass 2001-20004 
Species 2001 g/m2 2002 g/m2 2003 g/m2 2004 g/m2 01? ? 04? Change 
      
Sagittaira 
latifolia 36.29 84.71 248.94 369.63 Increase 
Pontederia 
cordata 132.2 53.2 4.82 0.0 Decrease 
Ludwigia sp. 0.32 45.67 1.83 0.0 Decrease 
Peltandra 
virginica 5.81 90.82 19.54 13.79 Increase 
Echinachloa  
sp. 220.83 9.59 0.0 0.0 Decrease 
 
Table 4.4  10X exclosure plot mean biomass 2001-20004 
Species 2001 g/m2 2002 g/m2 2003 g/m2 2004 g/m2 01? ? 04? Change 
      
Sagittaira 
latifolia 4.65 26.45 20.08 9.16 Increase 
Pontederia 
cordata 32.98 41.27 2.37 0.0 Decrease 
Ludwigia sp. 12.62 48.4 0.26 0.0 Decrease 
Peltandra 
virginica 0.0 0.0 53.22 11.78 Increase 
Echinachloa  
sp. 13.36 0.0 0.0 0.0 Decrease 
Polyganum  
sp. 0.89 1.32 0.0 0.0 Decrease 
 
Inherently obvious differences in biomass between excluded and control plots 
occurred, with no biomass present within unexcluded plots throughout the four years. 
 
 
107 
 
When biomass from all species were combined, mean biomass from each plot size was 
adjusted to extrapolated mudflat surface elevations and plotted. For the surface elevation 
gradient within exclosure plots between 1.55? and 1.70? NGVD ?29 low marsh 
elevations, a gradually increasing trend in biomass with increasing elevation could be 
detected from 2001 to 2002 in the 10x exclosures (Figure 4.19).  In 2003, beginning with 
reduced mean biomass for the 10x exclosure plots, the previous years increasing trend 
reversed itself with gradually decreasing biomass as elevation increased.  By 2004 the 
mean biomass in the 10x exclosures had declined further with the biomass across 
elevation trend line evening out. The 20x exclosure plot biomass over elevation revealed 
some interesting results (Figure 4.20).  The 2001 mean biomass increased gradually and 
steadily across the low elevations as they had in the 10x plots. In 2002 and 2003 the 
effect of elevation on biomass was sharper and more pronounced, beginning with less 
biomass at the lower elevations in each year and rising to greater biomass each successive 
year approaching the 1.7? NGVD ?29 elevation mark.  There is an apparent break point 
between the 1.63?-1.65? NGVD ?29 elevations where biomass is noted to reverse its trend 
of less biomass over successive years to an increase in biomass over previous years.  The 
2004 mean biomass data also revealed a break at 1.60? NGVD ?29, where small increases 
in elevation resulted in more rapid increases in biomass when compared to 2001.  
 
 
108 
 
 
Figure 4.19 Combined 10x exclosure plot biomass for all species by year  over elevation. Year lines 
are biomass continuum values of predicted LS means + SE from analysis of covariance. Points along 
the lines are biomass values from predicted means within the low to middle marsh zone represented 
by the data. 
 
Figure 4.20 Combined 10x exclosure plot biomass for all species by year  over elevation. Year lines 
are biomass continuum values of predicted LS means + SE from analysis of covariance. Points along 
the lines are biomass values from predicted means within the low to middle marsh zone represented 
by the data. 
 
 
109 
 
4.6 Presence / Height 
Species presence and height data were collected at each meter mark along 12-ten 
meter and 12-twenty meter exclosure transects each year of the four year study period 
(Appendix B, Tables 4.8 and 4.9). The differences in exclosure size was not determined 
to have a significant effect on species presence between years (p>0.8).  Some differences 
were found in terms of the total number of species found along these transects over the 
four year study period when compared to those species intentionally planted by the 
restoration effort in Cell 1 areas outside of the exclosure study area (Table 4.5).  
Table 4.5   Planted Species at Kingman Marsh Restoration compared to those found in exclosure 
area transects. 
 
 
 
110 
 
A total of nine species were found along transect lines in 2001 and 10 species in 
2002 although only four and five species respectively contributed numbers greater than 
1% of all those species observed.  In 2003, total numbers of species found along transect 
lines decreased to six with only Sagittaria latifolia, Pontederia cordata and Peltandra 
virginica contributing to notable presence within the exclosures.  The final sampling year 
(2004) saw the total species presence reported in transect data decline to 4 with only 2 
species, Sagittaria and Peltandra contributing greater than 1% of the total plot 
occurrances.  
The location and height of the species identified along each of the exclosure plot 
transects were plotted over a SAS generated mudflat surface topography to give a 
graphical representation of the location and height of each species found within each of 
the 6 exclosures (Figures 4.21 to 4.24).  The 4 year progression of the species presence 
and height relative to each plot and its position on the mudflat surface elevation gradient 
gives a more realistic visual interpretation of the changes in species composition, position 
relative to elevation, density and height than can be inferred through two dimensional 
figures and data tables.  Figure 4.21 reveals that in 2001, the dominance of Pontederia 
was evident throughout the exclosure plots with height related to the differences in 
mudflat surface elevation.  In 2002, the increased presence of Sagittaria can be seen in 
the higher elevations, with Pontederia consolidation in lower plot elevations and a 
resulting noticeable decrease in species height (Figure 4.22).  Ludwigia also becomes 
pervasive throughout the exclosures in 2002 and contributes to a noticeable increase in 
the density and diversity of species in the lower/middle elevations.   
 
 
111 
 
 
Figure 4.21 2001 species presence along exclosure plot transect lines and each species height in 
relation to the exaggerated surface elevation of the plots. 
 
Figure 4.22 2002 species presence along exclosure plot transect lines and each species height in 
relation to the exaggerated surface elevation of the plots. 
 
 
112 
 
The 2003 survey revealed a noticeable thinning of the density of individual plants 
as well as a marked decrease in the numbers of species represented throughout the plots 
with the effective disappearance of Ludwigia and Echinochloa (Figure 4.23).  The spread 
of Sagittaria becomes obvious while Peltandra and Pontederia presence also becomes 
clear. By 2004, the dominance of Sagittaria in the higher plot elevations is contrasted by 
the sparse presence of Peltandra throughout the lower plot elevations (Figure 4.24). 
Overall plant heights remain high with virtually no lower level  ?understory? found.  This 
is effectively the result of the growth and maturation of Sagittaria as the dominant 
perennial within the exclosure plot that occupied the highest elevations.  The growth in 
height and spread of wide, shade inducing leaves left little else that could compete, 
although Peltandra is still a presence in this area.  Figures 4.21 to 4.24 effectively 
represent the emergence and self-organization of the macrophyte community within the 
exclosure study area throughout the 4 year study period. 
 
 
113 
 
 
Figure 4.23 2003 species presence along exclosure plot transect lines and each species height in 
relation to the exaggerated surface elevation of the plots. 
 
 
Figure 4.24 2004 species presence along exclosure plot transect lines and each species height in 
relation to the exaggerated surface elevation of the plots. 
 
 
114 
 
4.7 Discussion 
The Kingman Marsh exclosure plots revealed the potential for the low-middle 
marsh elevation (1.5-2.0 ft. NGVD '29) biogenic mudflats to support a dense and 
relatively diverse native emergent plant community with the exclusion of resident goose 
grazer pressures (Figure 4.25). It is the geese that appear to be the dominant force 
controlling native plant vegetation dynamics at these elevations absent exclosures. At 
lower sediment surface elevations (< 1.5? NGVD ?29) a relative lack of volunteer plants 
within exclosures is likely due to the stress of increased flooding depth and duration over 
the physical mudflats that effectively suppresses emergent vegetation.  As mudflat 
surface elevation increased across each exclosure, the potential for macrophytes to be 
present and have greater cover, biomass and height increased.   
 
Figure 4.25  View in 2002 southeast across Kingman exclosure plots in the direction of generally 
increasing elevation. Cover, density and diversity is apparent, with native species. Vegetation on 
the far right in the distance past the exclosures is a dense stand of non-native Lythrum salicaria. 
 
 
115 
 
The complete lack of vegetation presence in unfenced control plots (Figure 4.26) at any 
elevation supports the assertion that two alternate states, one biogenic mudflat and 
one emergent marsh, can exist in the same space and time within certain elevation 
zones on Anacostia River. 
 
Figure 4.26 View northeast across the unvegetated mudflat of an unfenced control plot. Fence lines of 
exclosure plots marked a stark contrast between areas that were, or were not accessible to goose 
herbivory. 
 
Herbivory of macrophytes in freshwater wetlands has been well documented 
(literature review by Lodge, 1991), with the effects of goose herbivory on wetlands in 
general also a well studied area (Smith and Odum, 1981; Cargill and Jefferies, 1984; 
Bazely and Jeffries, 1986; Giroux and Bedard, 1987; Kerbes et al., 1990; Hik et al., 1992; 
Belanger and Bedard, 1994; Beaulieu et al., 1996).  Most of these goose herbivory studies 
focused on migratory stopover or breeding grounds, with some of the research finding 
that the migratory goose populations had no effect on plant production (Beaulieu et al., 
1996) or even a positive synergistic effect (Cargill and Jefferies, 1984; Bazely and 
 
 
116 
 
Jeffries, 1986).  While migratory waterfowl are only in an area a portion of the year, 
resident waterfowl such as Canada geese and mallards are present year round and may be 
expected to impact a site to a greater degree (Haramis and Kearns, 2007).  
The effects of resident Canada geese on wetlands have become especially 
apparent in Maryland in the last several decades, but were not a concern in the District of 
Columbia  until 2000 (Harris, 2002). During the Kingman Marsh restoration, the strong 
effects of goose grazing became evident from the beginning of the marsh planting when 
exclosure fencing was installed to prevent rapid losses of planted material by the geese.  
It was thought that a successful growing season with goose exclusion would establish a 
marsh that would be resistant to goose herbivory in the following years by nature of the 
ability of the marsh to ?outgrow? goose grazer pressure. This had been the case with the 
Kenilworth Marsh restoration.  In retrospect, the increasing numbers of geese present 
year round on the adjacent golf course, mudflats and open water areas of Kingman 
(Figure 4.5) made restoration without exclosure fencing not an option if certain species of 
desireable marsh plants (i.e. Sagittaria latifolia and Pontederia cordata) were to exist.   
The context of landscape position (and surrounding land use) is considered to be 
an important element influencing ecosystem development (review by Farina, 2006). The 
behavioral ecology of animals includes decision making for movement, habitat selection, 
and feeding that is integral to the landscape (MacArthur and Pianka, 1966; Gagliardo et 
al., 2001). This is especially true for resident Canada geese (Harris, 2002). At Kingman 
Lake, landscape context was not taken into consideration with regard to its influence on 
the resident goose population prior to restoration activities, as it was not seen as a 
significant problem at Kenilworth Marsh. Although it is clear that in some landscape 
 
 
117 
 
contexts animals can wield considerable control over their environment (Butler, 1995), 
achieving proper surface elevations was understood to be the primary driving force 
behind marsh restoration outcomes in terms of cover and species distribution, as was 
determined by the relative success at Kenilworth Marsh.   
Unlike Kenilworth Marsh, the Kingman Marsh restoration areas were bordered by 
a golf course (Cell 1) and mowed lawn (Cell 2), with numerous access points for geese to 
enter the marsh area. Expanses of open water also allowed for easy flight approaches and 
additional access to the restored marsh.  Kenilworth Marsh did have an expansive 
recreational lawn area nearby, although this area was buffered by wider unbroken 
wooded areas unlikely to be used by the geese for access.  The only large open water 
suitable for goose approach in Kenilworth Marsh was that which surrounded the Mass 
Fill 3 sites and as recounted in Chapter 3, this area was heavily grazed by geese after 
restoration planting, reverting the site back to a mudflat state.  
The presence of overwhelming goose herbivory effectively made the case for two 
alternate marsh states at Kingman, one dominated by low-middle marsh elevation 
mudflats with some high marsh invasives plant species not palatable to geese, and one 
without goose herbivory protected by exclosure fencing.  The experimental exclosures 
were constructed to determine if the low to middle marsh mudflat surface elevations 
(1.5?-2.0? NGVD ?29) would support the type of emergent marsh planned for in the 
Kingman restoration without the pressure of goose herbivory.   
Of the six species that were initially planted for the Kingman Marsh restoration in 
2000, three of these species, Sagittaria latifolia, Pontederia cordata and Peltandra 
virginica, voluntarily emerged in each of the 6 experimental exclosures in 2001. 
 
 
118 
 
Combined plot species cover along transects against mudflat surface elevation reveal that 
Pontederia can emerge and cover at elevations as low as 1.3? NGVD '29 and begin to 
increase in cover above 1.5? NGVD '29 (Figure 4.13).  
Garbish and Coleman (1978) found that Pontederia seed germination at all 
elevations in their tidal fresh vegetation experiment was poor, while transplanted nursery 
stock seedlings survival was satisfactory at all elevations.  However they cautioned that 
their experimental area was also exposed to a large fetch, which subjected the study plots 
to consistent wind/wave energy and debris scour/accumulation that may have negatively 
influenced seed germination. The results of tidal freshwater marsh vegetation emergence 
from the seed bank in the Anacosia restoration areas (Baldwin and DeRico, 1999) and 
with field experiments in other tidal freshwater systems (Baldwin et al., 2001 had more 
positive results, although study plots were not subjected to the energies as Garbish and 
Coleman.  
The differences in exclosure size was not determined to be significant in plant 
species presence or biomass between years, although the prevalence of 2 of the 3 the 
smaller exclosures on the northern side of the study area with exposure to greater fetch 
made them more likely to be breeched by storm driven floating logs and trees. The loss of 
exclosure protection, even temporarily, was an opportunity for goose access and 
herbivory that some of the plot areas with lower elevations could not recover from in 
subsequent years.  The interactive effects of herbivory and elevation are seen as 
important controlling factors in the Kingman Marsh restoration effort (Baldwin and 
Pendleton, 2003; Hammerschlag et al., 2006). 
The species cover dominance of Pontederia in 2001 across all exclosure 
 
 
119 
 
elevations continued in 2002, although only up to the edge of the low marsh elevation of 
1.7? NGVD ?29 where Sagittaria cover began to increase over Pontederia at higher 
elevations approaching 2.0? NGVD '29 (Figure 4.14) which appears to be the classic 
result of competitive exclusion (Grime, 1973; Cronk and Fennessy, 2001). Significantly, 
the 6 control plots revealed that without the human intervention of fenced exclosures, no 
vegetation would voluntarily develop at any low-middle marsh elevations.  
While mudflat surface elevation did have an observed if not significant influence 
on species presence, cover, and biomass of emergent volunteers within the exclosures, 
the complete lack of vegetation in control plots throughout the 4 year study (Table 4.6) 
highlighted the influence of goose herbivory as the inherently significant overriding 
factor in determining plant presence within the range of elevations found in the 
unexcluded plot study areas.  This was further supported by the fence failure of several of 
the exclosure plots in 2003 following successive storm and flood events.  While the 
exclosure fencing was repaired within weeks, the presence of a developing emergent 
macrophyte community prior to the storm events within those plots was severely 
impacted in the intervening period of fence failure.  Direct evidence of goose grazing was 
observed in the form leaf browse, tracks and characteristic crater like ?grub-outs? in the 
mud where the foraging for below ground roots and tubers occurred.  This sub-surface 
grazing effectively reduces or eliminates any chance of vegetation regeneration (Belanger 
and Bedard, 1994).  Even with fencing re-erected and maintained, the 3 exclosure plots 
affected reported reduced cover and biomass for 2003, which continued into 2004 
resulting in almost no cover or biomass in most of those exclosure plot areas. The 
exclosure that was located at the higher elevations kept near complete cover throughout 
 
 
120 
 
the study, although it was also the only one to not have a fence failure. Of the plots that 
did have fence failures, the lower elevation exclosures had a difficult time revegetating 
after an eat-out while those plots at higher elevations were more likely to be recolonized. 
 
Table 4.6  Comparison of Kingman mudflat exclosure plant community attributes over four survey 
years. Averages are arithmetic means combining both plot sizes across all elevations. 
 
 
The effects of different disturbance regimes (Odum et al., 1979) particularly on 
wetlands have been increasingly well studied (review by McKee and Baldwin, 1999), 
with the interactive effects of multiple disturbance agents and stresses on wetland 
vegetation of particular interest to the Kingman exclosure study (Belanger and Bedard, 
1994; Brewer et al., 1997; Gough and Grace, 1998; Rachich and Reader, 1999; Baldwin 
and Pendleton, 2003, Bertness et al., 2004). Several studies of tidal freshwater wetlands 
have found a negative effect of lower sediment surface elevation (and subsequent 
increased depth and duration of tidewater inundation) on wetland plant seedling 
emergence, survival and growth (Baldwin et al., 2001; Peterson and Baldwin, 2004).  
Studies of the interactive effects of herbivory and disturbance with sediment surface 
 
 
121 
 
elevation have also found that these interactions can potentially have an additive negative 
effect on growth and species composition (Brewer et al., 1997; Baldwin and Pendleton, 
2003).  Interestingly, several studies actually found positive synergistic effects between 
plant herbivory and the promotion of alternative species interactions (Noy-Meir, 1975; 
Belanger and Bedard, 1994; Gough and Grace, 1998; Rachich and Reader, 1999) 
although the added effect of elevation was not considered.  
The Kingman restoration target elevation of 2.0? NGVD ?29 was in the middle 
marsh zone, which was expected to keep invasive macrophytes from colonizing at their 
preferred higher elevations (Hammerschlag et al., 2006).  Higher than designed for post 
construction sediment surface elevations in some areas allowed for the establishment of 
the invasive macrophytes Lythrum salicaria, Phragmites australis and Typha sp.  Known 
competitors, these species, particularly Lythrum, appear to have established at the higher 
elevation areas of Kingman Marsh to the exclusion of many native planted species.   It is 
possible that these invasive species might have dominated regardless of goose grazing, 
although the goose herbivory of native planted species created gaps for competitively 
dominant species that were much wider than they would have been without the grazing 
effect (Gough and Grace, 1998; Rachich and Reader, 1999; Cronk and Fennessy, 2001). 
It is notable that the invasive species Lythrum was dominant in areas of higher elevation 
not far from the study site and yet while present in trace numbers, failed to spread within 
the exclosure plots even at the highest elevations approaching 2.0? NGVD ?29.  This 
?elevational resistance? to undesirable plant invaders was part of the design intent of the 
Kingman marsh restoration (Hammerschlag et al., 2006), which was revealed to be a 
successful design strategy as long as target elevations are met as evidenced by the 
 
 
122 
 
experimental exclosure plot species composition and coverage. 
It is believed that goose herbivory acting as the primary control factor, combined 
with the secondary physical stress of lower substrate elevations, accounts for the 
persistence of the mudflat as a dominant feature of the low to middle marsh (1.5-2.0 ft 
NGVD '29) intertidal areas of Kingman Marsh.  Bertness and associates (2004) estimate 
that it takes one adult goose just one hour to remove all above and below ground salt 
marsh vegetation across one square meter leaving bare mudflat.  With the goose counts of 
this study (Appendix B, Table 4.1) and National Park Service estimates of between 500 
and 2000 geese possible in the Kingman area (Harris, 2002), the potential and realized 
effect they have on marsh restoration is obvious.  The capability of the resident geese to 
significantly transform their environment to such a degree qualifies them as ecosystem 
engineers (Jones et al., 1994) able to confound the best efforts of the Army Corps of 
Engineers.  
Once marsh vegetation is removed, the additional combined stress of lower 
mudflat surface elevation becomes too much for the marsh system to rebound from in the 
short term (Baldwin and Pendleton, 2003; Peterson and Baldwin, 2004) even with an 
available seed bank source (Baldwin and DeRico, 1999).  Because of this, the lack of the 
marsh state resiliency at the low-middle marsh elevations allows the switch to the 
alternate biogenic mudflat state, becoming more resilient. Under the continued pressure 
of strong resident goose herbivory, the biogenic mudflat state will remain the dominant 
condition in this zone of elevation without intervention to reduce the goose population 
(Harris, 2002).  Attempts to artificially switch the biogenic mudflat into the emergent 
marsh state again with fenced exclosures is understood to be a temporary solution.   
 
 
123 
 
The concept of ecological memory may be strong with the biogenic mudflat state 
(McKee and Baldwin, 1999; Peterson, 2002), as mudflats had persisted in the landscape 
of the Anacostia River for almost half a century as the dominant feature (Syphax and 
Hammerschlag, 1995) and potentially for more than a century in association with 
emergent marshes given the river?s long depositional history (Coues and Prentiss, 1883). 
The strong mudflat memory reinforces itself with time and tends to inhibit an ecological 
response to change unless the change is rapid and intense (Peterson, 2002).  Pethick 
(1996) has determined that mudflats may persist for centuries without some extraordinary 
change in condition. The restoration work of the Army Corps of Engineers would be 
considered such a change, and in effect the trigger to switch the mudflat state over to the 
emergent marsh alternate state.  This trigger was effective at Kenilworth Marsh and was 
assumed that it could work again at Kingman Lake were it not for the unanticipated 
confounding influence of the resident geese.   
Further efforts at the large scale conversion of mudflats to emergent marsh have 
taken place along the Anacostia River fringe opposite Kingman Island (Figure 
4.3)(Hammerschlag et al., 2006).  This emergent marsh, constructed in 2003 in a similar 
manner to Kenilworth and Kingman Marshes, has been artificially kept in the vegetated 
marsh state with a complex matrix of perimeter and interior exclosure fencing as well as 
aerial exclusion lines crossing the tops of the exclosures to keep birds from entering.  It is 
anticipated that the goose exclusion will remain throughout the Fringe Marsh until a plan 
for managing the geese can be developed.  As of this writing, a goose management plan 
is still in development. 
 
 
124 
 
4.8 Conclusion 
With any large-scale restoration project, uncertainty and natural variation from 
intended goals can be expected (Zedler, 2001, Mitsch and Jorgensen, 2004; Baldwin, 
2004).  The dramatic loss of 40-60% of the total restored marsh cover from the Kingman 
Marsh restoration in 2001 was unexpected and primarily from resident goose herbivory 
(Hammerschlag et al., 2006) despite relatively less goose grazer impacts at Kenilworth 
Marsh.  Some would argue that resident geese (or any non-native species) are not natural 
and are not a proper part of the Anacostia River system (Kangas, 2004a, 2004b; Kangas 
et al., 2004).  While efforts to cull their numbers in rural and even some urban areas have 
proven successful, the decision to reduce their numbers in the District of Columbia will 
be an interesting case in political ecology, the result of which is not yet known but has the 
potential to be controversial (Peterson, 2000; Harris, 2002).  In a quote highlighted in 
Kangas? comprehensive book on ecological engineering (2004a), Marston Bates (1961) 
has this to say about exotic species:  
?The animals and plants that have been accidentally or purposefully introduced 
into various parts of the world in the past offer many opportunities for study that have  
hardly been utilized. They can, in a way, be considered as gigantic, though unplanned, 
experiments in ecology, geography, and evolution, and surely we can learn much from 
them.? 
Ecologically engineered systems are attempts to design and direct the natural 
energies of ecosystem processes to serve some societal purpose that also benefits the 
natural world in some way (Odum, 1989; Kangas, 2004a; Mitsch and Jorgensen, 1989, 
 
 
125 
 
2004). Understanding and accepting the ability of ecosystems to organize themselves is 
essential to ecological engineering and wetland restoration ecology (Mitsch and Wilson, 
1996; Odum, 1989; Mitsch, 2000).  While adaptive management decisions are important 
in directing the restoration of complex systems like tidal marshes toward a desired 
outcome, the recognition that unexpected forces may dramatically alter restoration goals 
beyond human management fixes may need to be accepted (Mitsch and Jorgensen, 2004).  
Restoration projects are excellent opportunities to learn from successes and failures and 
every effort should be made to apply the knowledge gained in future restoration work in 
order to advance the field of ecological engineering and restoration ecology.   
The Army Corps of Engineers has made efforts to learn and adapt from its 
restoration efforts, as the funding for vegetation, sediment elevation, seed bank studies 
and several other research efforts attest (Hammerschlag et al., 2006).  The problem with 
some of its failures in the restoration planning and decision making process may be more 
of a result of restoration ecologists and designers not asking the right questions, posing 
hypotheses and sufficiently testing them (Kangas, 2004a, 2004b).  Taking the time and 
investing comparatively modest sums of money in smaller trial and error research efforts 
with a view of the next restoration project from a whole systems perspective would 
probably prove more cost effective in the long run. Instead it appears that the important 
yet distracting budget cycles and implementation schedules of large restoration efforts 
tend to overshadow and in many ways preclude the ?go slow and learn? approach.  
Regardless, large scale and expensive restoration efforts that do not result in desired 
outcomes are still important opportunities for lines of research that can advance the field 
of ecological engineering and the help to improve the discipline of restoration ecology.    
 
 
126 
 
Chapter 5: Emergy Analysis of Kingman Fisheries Production 
5.1 Introduction 
The discipline of ecological engineering has evolved greatly since the coining of 
the term in the early 1960?s by pioneering ecologist Howard T. Odum (Odum, 1962; 
Odum et al., 1963; Kangas, 2004a; Mitch and Jorgensen, 1989, 2004).  One of the 
primary goals of ecological engineering, ecosystem restoration, has even developed to the 
point of becoming an important tool of the U.S. Army Corps of Engineers.  
The use of dredge material for the restoration of emergent marsh habitat (U.S. 
Army Engineer Waterways Experiment Station, 1978, Priest et al., 1996, Streever, 2000; 
Yozzo et al., 2001) has become the primary ecologically engineered method of wetland 
re-establishment on Anacostia River in Washington, D.C (USACE, 1993).  The most 
suitable sites for this type of restoration approach are shallow water areas and intertidal 
mudflats, which are too low in sediment elevation to support emergent marsh vegetation 
(Syphax and Hammerschlag, 1995).   
As recounted in Chapter 4, the Kingman Lake tidal area was historically 
dominated by intertidal mudflats at low tide following its creation by dredging in the mid 
20th Century.  This condition made it attractive for a Kenilworth Marsh type beneficial 
use of dredge material restoration effort (1993) reported in Chapter 3, and became the 
model for the Kingman restoration. Onsite marsh restoration activities began on Kingman 
Lake in early 2000 by the U.S. Army Corps of Engineers.  Dredging, filling, grading, 
marsh planting and fencing was completed in 2000, transforming the Kingman area from 
 
 
127 
 
its status as a tidal ?lake? to that of Kingman Marsh. 
Tidal marshes have often been regarded as valuable for their potential to increase 
sediment trapping, organic matter and nutrient cycling, fisheries/wildlife production and 
habitat as well as flood control and intrinsic human aesthetic value (Greeson et al., 1979; 
Odum et al., 1984; Keddy, 2000, Mitsch and Gosselink, 2000, Zedler, 2001; Batzer et al., 
2006). These functions and values are often and correctly forwarded as the rationale for 
the restoration of wetland systems.  
One important function of tidal marshes that is often cited is their importance as 
productive nursery and feeding areas for migratory and resident fishes (Van Engel and 
Joseph, 1968; E.P. Odum, 1971; W.E. Odum et al., 1984; Mitsch and Gosselink, 2000). 
E.P. Odum, in describing salt marshes, compared them to ?primary production pumps? 
feeding coastal waters in what he referred to as outwelling (E.P. Odum, 1980, 2000).  The 
flow of energy from salt marshes to adjacent waters was affirmed by Childers et al. 
(2000), emphasizing the association of the marshes ?outwelling? with nearby tidal creeks 
and the inner estuary. Studies have related commercial fishery catch with the area of 
intertidal vegetation (Turner, 1977) and marsh edge length (Teal and Howes, 2000). 
While the outwelling hypothesis has been characterized in terms of salt marshes and 
estuaries, the same relationship can be made with tidal freshwater marshes which are 
known to export considerable percentages of their above ground macrophyte biomass 
(Odum and Heywood, 1978; Odum et al., 1984; Mitsch and Gosselink, 2000).  The 
energy subsidy of the regular tidal pulse to and from intertidal marshes maximizes the 
efficiency of the system in transforming and exporting energy flows to adjacent waters 
(Odum, Odum and Odum, 1995). A major pathway for the transfer of energy from lower 
 
 
128 
 
to higher trophic levels in wetlands is made through fish (Van Engel and Joseph, 1968; 
Smith et al., 2000) which are then as a ?fishery? the link to the human social and 
economic system (Bahr et al., 1982; Martin, 2002).   
The transfer, or export, of restored Kingman Marsh energies through the fishery is 
examined through the use of emergy analysis. The comparison of production exports with 
the resulting changes in energy signatures from an existing, pre-restored condition to a 
restored condition for natural systems are rarely undertaken (H.T. Odum, 1983, 1996; 
W.E. Odum et al., 1984; Martin, 2002; Kangas et al., 2004).  The structure and function 
of a pre-existing or restored system is influenced by the sum of the natural and human 
energies that contribute to it (Nixon, 1988). While economic analyses recognize 
monetary values, they typically do not incorporate or greatly underestimate the value of 
natural environmental inputs and exports due to the lack of a direct monetary valuation 
for natural services (Odum, 2001).  An environmental accounting methodology that 
evaluates the total natural and human economic energy flows of a system on a common 
energy basis is termed an emergy evaluation or emergy analysis, with emergy the 
measure of natural or human work previously required to support an environmental 
system, product or service (Odum, 1996; Martin, 2002).  
The value of large scale environmental restoration projects to the human economy 
and society, such as improved fisheries production, as well as to the natural systems they 
are constructed in, (increased primary production) is worth a detailed systems level 
investigation to advance the understanding of some of the results of restoration (Odum, 
1984, 1996; Martin, 2002; Kangas, 2004a; Kangas et al., 2004; Mitsch and Jorgensen, 
2004).  The work of Martin (2002) and his emergy analysis of U.S. Army Corps of 
 
 
129 
 
Engineers river diversions intended to restore the tidal marshes of the Mississippi River 
delta is utilized as an appropriate template for an emergy analysis of the Kingman Marsh 
primary production and secondary fishery production under pre-restoration mudflat 
conditions, early restored marsh conditions and a mature marsh condition scenario.  
Using data specific to Kingman, the Anacostia River, the region, and several referenced 
assumptions, the emergy evaluation seeks to determine if the new marsh restoration 
would improve the exports of primary production and sport fisheries over the existing 
mudflat condition.   
Cairns (1988) described restoration ecology as ?full or partial placement of 
structural or functional characteristics that have been extinguished or diminished and the 
substitution of alternative qualities or characteristics than the ones originally present with 
the proviso that they have more social, economic, or ecological value than existed in the 
disturbed or displaced state.? If some form of Cairns definition is to be accepted, then an 
accounting and comparison of the energy flows of the alternate states of pre-restoration 
mudflat and restored marsh is warranted to determine the relative contributions of each. 
To understand ecosystems and effectively restore aspects of them, the rates of energy 
circulation within a system as well as the rates of flows of energy and matter into and out 
of the system must be understood (Odum, H.T., 1971). The usefulness of the emergy 
analysis is found in objectively quantifying the total amounts of energy from the 
environment and the economy that are used.   Comparing the restored marsh primary and 
fisheries production to the production of the mudflat alternate state can be useful when 
observed through an emergy analysis.  In other words, in terms of primary production and 
fisheries export, is a restored marsh at Kingman more productive than the mudflats that 
 
 
130 
 
they replaced?  The following emergy analysis is intended to attempt to answer these 
questions. 
5.2 Operational Definitions 
5.2.1 Defining Emergy  
An expression of energy that accounts for all of the available energy flows of one 
type of energy that are used directly or indirectly in creating a resource or service is 
known as emergy, short for energy memory (Odum, 1996). Emergy analysis is a method 
by which all of the energy inputs to and exports from the natural and human constructed 
systems may be valued and accounted for in a common non-monetary way with the 
energy base of solar energy (Odum, 1996).  Solar emergy, termed the solar emjoule (sej) 
is the equivalent solar energy required to produce a natural or human product or service.  
Solar emergy is calculated by multiplying units of energy in Joules (the international 
standard of energy measure) by emergy per energy ratios called transformities (Odum, 
1988, 1996).  
5.2.2 Defining Transformity 
These ratios, called transformities, are obtained by dividing the total emergy that 
was used in a process by the energy yielded by the process (Odum, 1988, 1996).  The 
transformity ratios effectively convert different energy types such as fuel, to the common 
unit of emergy the solar emjoule.  Transformity ratios have been calculated and placed in 
tables for numerous natural and human products and services and are central to 
quantifying the common emergy values needed when performing emergy analyses 
 
 
131 
 
(Odum, 1996).  Several ratios used for calculation are emergy per mass and emergy per 
dollar ratios, also termed emdollars.   
5.2.3 Defining Emdollars 
The emdollar (em$) ratio is important in converting emergy flow to a common 
monetary amount, which is useful when translating natural products or services, which 
are not often given an economic value, to that which can be related to the human 
economy (Odum, 1996; 2001). The emdollar is defined as the total amount of money 
flow generated by a given amount of emergy input (Odum, H.T., 1996; 2000, 2001; 
Martin 2002). Emdollars allow the direct comparison and calculation of non-monetary 
emergy found in environmental inputs, services and functions to that of standard 
economic equivalents in dollars (Odum, 1996; 2001; Martin, 2002). Emdollars are 
obtained by dividing the emergy flows by the emergy/money ratio (which is determined 
as the total emergy of a region divided by the gross economic product of that region in a 
given year) (Odum 1996, Odum et. al. 2000, 2001; Martin, 2002).  In this way the value 
of natural resources and services to the human economy can be equated so that an 
analysis can inform management decisions concerning the optimum use of human 
resources. 
5.2.4 Kingman Mudflats and Marsh, Primary and Fisheries Production Emergy 
Analysis 
An emergy analysis of the production exports that result from the marsh 
restoration of the Kingman area of Anacostia River (profiled in Chapter 4) is detailed for 
 
 
132 
 
the purposes of determining the net value and benefits of this type of restoration to the 
primary and fisheries production.  The natural environmental contributions to an existing 
or newly restored system are often characterized individually, but rarely aggregated and 
quantified in a manner which is directly comparable.  Determining the net benefit of 
environmental restoration in terms of emergy is seen as a way to guide management 
policy and efficiently utilize money for environmental projects (Odum, 1996, 2001; 
Kangas et al., 2004). The results of the emergy analysis will help inform future 
restoration projects of this type.  The analysis conducted here is patterned on the work 
conducted by Martin (2002). 
Information collected relating to the Kingman area pre-restoration (intertidal 
mudflat/open water) and post-restoration (emergent marsh, mudflat/open water) physical 
and biological conditions are compiled and detailed. Fishery production exports for the 
restored marsh condition and mature marsh condition are based upon fishery data 
collected within the newly restored Kenilworth Marsh and Dueling Creek mature marsh 
systems detailed in Chapter 3. A determination of the natural and human emergy inputs 
required for the Kingman Marsh restoration project and its emergy exports are first 
diagrammed and then listed and calculated.  As a final analysis, economic (emdollar) 
equivalents are developed to compare the environmental economic services of primary 
and fisheries production supplied by the pre-existing Kingman condition dominated by 
intertidal mudflats vs. that of the restored Kingman emergent tidal marsh as it existed the 
year it was completed in 2000 and with a mature marsh projection 50 years into the 
future.  
It is fundamental to emergy analysis to assume that the contribution of a resource 
 
 
133 
 
or service is proportional to the amount of energy required to create the resource or 
service (Brown and Herendeen, 1996).  The emergy analysis of the total economic and 
natural emergy (person hours, fuel, information, solar insolation, tides, biomass 
production, sediment processes, etc.) required to build and maintain the restored 
Kingman marsh are calculated and displayed in a table. The transformation of these 
different energy types into the common unit of emergy, the solar emjoule (sej), provides a 
platform for a clear analysis of the relative benefits of the Kingman marsh restoration  
over mudflats, with the expectation that over time the primary and fisheries production 
provided by the marsh restoration will be improved (Odum, 1996). 
5.2.4.1 Methods  
5.2.5 Energy Systems Diagram and Analysis Table 
First an energy systems diagram for the mudflat/open water and newly restored 
emergent marsh conditions at Kingman is developed to help conceptually organize the 
relationships between the system components and their energy flow pathways (Figure 
5.1).  The diagram represents the structure and functions of the system in order to help 
identify the inputs and particular outflows that are to be evaluated (Odum, 1996; Odum 
and Odum, 2000, Odum et.al., 2000; Martin, 2002).  The important system components 
that contribute to the Kingman boundary area include the sun, rain, tides, river 
components, wind and the resources imported to restore emergent marsh.  While many of 
the values are based on referenced data particular to the area and the region, assumptions 
are made that they apply particularly to the Kingman study site and several assumptions 
are made that are applied from the literature. The primary productivity of mudflats, open  
 
 
134 
 
 
Figure 5.1 Kingman Marsh restoration energy circuit diagram detailing renewable energy flows, 
inputs from the economy for restoration, and system exports from the mudflat/open water and 
emergent marsh systems. 
 
 
 
135 
 
water and emergent marsh are calculated from literature values and applied in 
proportion to their assumed area coverage in the pre and post restoration conditions.  
Secondary production in the form of the contribution of the sport fishery utilization of the 
two conditions are also calculated from actual fishery data collected earlier (Chapters 3 
and 4) and applied to this study.   
An emergy analysis table (Table 5.1) is then constructed directly from the systems 
diagram and values of the annual amounts of import or export of each energy flow are 
quantified in raw units (joules, grams, dollars) using inflows and outflows crossing the 
system boundary as row headings (Odum, 1996; Odum and Odum 2000; Odum et.al. 
2000; Martin 2002). Transformities reported by Odum (1996) and other cited literature 
sources are applied to convert the flows of different raw unit energies to emergy of the 
same type in solar emjoules (sej) for comparison and analysis (Odum, 1996; Odum and 
Odum, 2000; Odum et.al., 2000; Martin 2002).  The calculations of each of these raw unit 
flows of energy that contribute directly to the emergy analysis Table 5.1 are detailed in 
the footnotes to the table in Appendix C, Table 5.1. 
 
 
136 
 
Table 5.1  Kingman Marsh emergy analysis table 
 
 
 
137 
 
5.2.6 Renewable Resources 
Because the sun, wind and rain are all supported by and interdependent on the 
same initial solar energies, their values are not cumulatively added to the total value of 
renewable resources.  To avoid double counting the contributions of these inputs, the 
largest input is selected as the surrogate for the sum of the inputs (Odum, 1996; Martin, 
2002).  For the Kingman boundary area, this would be the chemical potential emergy of 
the rain.  The river geopotential emergy was determined by multiplying the percentage of 
flow of Anacostia River water into the Kingman area by the transformity calculated for 
geopotential specific to the Anacostia River (Table 5.2).   
River sediment flow was determined as the annual total suspended solids (TSS) 
flow into Kingman as a percentage of that from the Anacostia River minus sediment 
reduced due to subsidence in the Kingman study area.  Sediment was multiplied by the 
emergy per mass for mineral sediment from Table 5.2 as assumed by Martin (2002).  
Geologic uplift was considered negligible for the Kingman area and was left out of the 
calculations while river geopotential was added to the sum of the total renewable 
resources. 
 
 
138 
 
Table 5.2 Emergy per mass of river geopotential and river sediments 
 
5.2.7 Inputs to the Kingman Marsh Restoration 
The energy inputs to the marsh restoration in the form of dollars and volume of 
dredged river sediment were broken into several groups as detailed by the U.S. Army 
Corps of Engineers (Appendix C, Table 5.2)(personal communication, 2004).  
Construction management and design, physical construction, dredged river sediment, 
marsh plantings, goose exclosure fencing, and monitoring and maintenance costs were all 
added to a total imported emergy for the Kingman Marsh restoration work.  The emergy 
to dollar ratio calculated for the year of the restoration in 2000 (1.0 x 1012 sej/$ ) was 
used as a conservative estimate to convert dollar values to solar emjoules (Odum, 2001). 
It is calculated as the annual U.S. inflow of emergy divided by the U.S. Gross National 
Product (Odum, 1996).  
 
 
 
139 
 
5.2.8 Exports from the Kingman System 
There are several components of production in tidal freshwater systems that are 
exported to different degrees.  The concept of resource export, or outwelling, from tidal 
marshes and shallow water zones in estuaries discussed earlier has been articulated since 
the 1960?s (E.P Odum, 1971, 1980, 2000).  This is considered to be true of the Kingman 
Marsh system which is predominantly intertidal, with most water leaving the system to 
move further down the Anacostia River on an ebbing tide.  The model created for this 
emergy analysis attempts to account for the primary and fishery production exports from 
the Kingman system with a similar methodology and accounting to that of Martin (2002).  
While Odum et al., (1984) reported that energy flow in tidal freshwater marshes was 
largely speculative, a rough energy flow description identified three primary sources of 
energy to support the marsh food webs: emergent macrophytes, terrestrial organic 
material and phytoplankton, with benthic microflora also considered of possible 
importance. The primary production group contributions estimated in this analysis for 
export from Kingman under pre and post-marsh restoration conditions included: open 
water phytoplankton (Footnote 5.1), mudflat algae (Footnote 5.2), and emergent marsh 
(Footnote 5.3) primary production.  These are considered to be the largest contributors to 
export from the marsh (Odum et al., 1984) and are unique features of estuaries in having 
three distinct classes of primary producers (Odum, 1980).   
Fish and bird groups are also represented. Although both fish and bird data were 
collected at Kingman, only the fish data was utilized for the emergy analysis as a primary 
exporter of energy from the system (Footnote 5.4).  This fits with the accepted paradigm 
 
 
140 
 
of the marsh as a nursery for fisheries (Boesch and Turner, 1984; Deegan, 1993; Houde 
and Rutherford, 1993; Mitch and Gosselink, 2000). The role of the higher level fish 
consumers in the emergy analysis is to represent the ultimate extraction and export of the 
saprotrophs, primary and secondary consumers that are converting flows of energy from 
terrestrial carbon and the detrital pool, which are important components of the system 
(Odum, E.P, 1971, 1980; Odum et al., 1984; Mitch and Gosselink, 2000).  
Approximately 100% turnover of aboveground macrophyte primary production 
on an annual basis is possible for many tidal freshwater marshes, with almost all above 
ground plant matter from several dominant species of middle and lower marsh zones 
completely decomposing and leaving bare mud within months after senescence (Odum 
and Heywood, 1978; Odum et al., 1984).  Phytoplankton in Kingman also have 100% 
export potential on an annual basis with the outflow of almost all open water on an ebb 
tide as well as higher trophic level consumption and export. Algae biomass transition 
through higher trophic levels is also assumed a large proportion of export.  The observed 
tendency of mudflat algae to form algal mats that detach and float away once reaching a 
critical density is assumed to increase export combined with trophic transfers at Kingman 
to 100% (personal observation).   
Emergent marsh macrophyte biomass export was set at 80% to account for several 
factors, including: incomplete decomposition of some species; trapping of plant material 
in marsh sediments; and the inclusion of belowground root biomass in the production 
calculations which generally does not leave the system, although a case could be made 
for some export through waterfowl that dig up marsh plant tubers (Odum and Heywood, 
1978; Whigham et al., 1978).  Exports of energy flow through fish populations are also 
 
 
141 
 
assumed to be 100% on an annual basis.  
5.2.9 Kingman Marsh Area 
Marsh areas for Kingman were assumed to be that of the initial year of restoration 
in 2000 at 13 ha (Hammerschlag et al., 2001) which was the full extent of marsh 
restoration without the impacts of latter years herbivory.  Under typical marsh 
development trajectories with ample sediment supplies it might be assumed that the 
emergent marshes would expand their coverage over time or at least retain their current 
coverage in the immediate timeframe (Stevenson et al., 1988). This was found to not be 
the case, due to extreme goose herbivory as discussed in Chapter 4.  For this reason, the 
same area coverage for emergent marsh and fishery useable open water area at high tide 
was assumed for Kingman Marsh (year 2000) when fishery production numbers for the 
Kenilworth restored marsh and Dueling mature marsh hypothetical scenarios were 
applied.  The same marsh area coverage was assumed over the 50 year time frame with 
the Dueling Creek fishery numbers applied to this hypothetical future of Kingman Marsh 
as a mature marsh fishery similar to that of Dueling.  Fishery useable open water area is 
defined for this analysis as all areas not occupied by emergent marsh at high tide or 31.2 
ha.  This is considered because the species comprising the sport fishery do not typically 
penetrate the marsh itself, but utilize the adjacent open water areas at high and low tide to 
hunt for prey fish that are utilizing the marsh for food and refugia (Rozas and Odum, 
1987). 
 
 
 
142 
 
5.2.10 Emergy Indicies Calculation 
Emergy indices are calculated from the fisheries export values found in the 
emergy analysis table and are used to draw inferences from the emergy analyses.  The 
emergy indicies used are as follows with their basic terms represented in Figure 5-2. 
 
Figure 5.2  Simplified energy circuit diagram illustrating the basic flow of energies used to calculate 
the emergy indices. 
5.3 Fisheries Emergy Yield Ratio 
 An interpretation of fisheries benefit, this ratio is the emergy of the fisheries 
production export, or yield from the system (Y) divided by the emergy of the renewable 
resources (Re) plus the primary production that supports the fishery (P). This ratio 
(Y/(Re+P)) indicates which system state (intertidal mudflat or emergent marsh) and 
condition (unchanged, newly restored or mature marsh) is more productive from a 
fisheries perspective.  The process with a greater value produces more fisheries to the 
 
 
143 
 
system with larger values representing larger contributions. 
5.4 Results 
5.4.1 Emergy Analysis Table 
5.4.1.1 Renewable Resources 
Total emergy from marsh restoration in year one (2000) estimated from 
renewable resources increased by 3.47 x 1016 sej over estimated mudflat renewable 
resources due to the Kingman Marsh restoration (Table 5.1).  Fifty year total increases in 
emergy flow due to the marsh restoration was calculated at 1.73 x 1018 sej over the purely 
mudflat state during that time period. As stated in the methods, a fifty-year time horizon 
for Kingman was chosen as a comparison to the marsh development point that the 
Dueling Creek mature marsh reference condition had attained.  The total emergy values 
did not include those that would be considered ?double counting? as described earlier in 
the methods section. River input contributions, including riverine water used by the 
restored marsh and to a lesser extent tidal energy absorbed by the marsh, were 
determined as the difference between the restored marsh and pre-existing mudflat 
conditions.  Annual river geopotential (Table 5.1, row 3) and the annual chemical 
potential energy of the rain (Table 5.1, row 2), acting on both state conditions equally, 
were the largest emergy contributors to the system with annual sediments captured and 
stored (Table 5.1, rows 9-10) and annual river water used (Table 5.1, rows 7-8) following 
with their contributions.   
The differential between water emergy used followed the differences between the 
 
 
144 
 
open water available under the marsh and mudflat conditions, with the restored marsh 
condition reflecting an annual increase of 2.99 x 1016 sej in the year one 2000 analysis 
(Table 5.1, rows 7-8).  Both conditions received the same sediment subsidy from the tidal 
Anacostia River.  From the three surface elevation table (SET) readings recorded nearest 
to the study period of the emergy analysis, it appeared that the lower elevation mudflat 
SET station (1.7? NGVD ?29) showed a greater average net gain of sediment while the 
higher elevation marsh SET station (2.4? NGVD ?29) had a more modest average net gain 
(Hammerschlag, personal communication).  While three readings are difficult to judge a 
long-term trend by, the emergy analysis for the restoration year of 2000 were assumed to 
be indicative of that year.  
Even as a rough approximations, it seems inconsistent with tidal wetland function 
conventional wisdom that a vegetated marsh would capture and store less sediment than 
an unvegetated mudflat or subtidal area (Stevenson et al., 1988), although both these 
systems are known for their potential to have seasonal and yearly changes in 
sedimentation due to localized conditions (Serodes and Troude, 1984).   
Net emergy for sediments captured and stored under the pre-restoration, mudflat 
condition reflected a modest annual increase of 3.88 x 1015 sej over the restored marsh 
condition (Table 1, rows 9-10).  The lack of difference in sediments captured vs. 
sediments stored within each condition was a result of the relatively insignificant effect 
of geologic subsidence assumed for the area for one year (Kearney and Stevenson, 1991), 
when compared to the potential volume of sediments captured in the system under both 
conditions.  The sediments captured and stored values turned out to not be important in 
the analysis, however, and to avoid double counting of emergy flows only the largest of 
 
 
145 
 
the riverine sediment inputs (river geopotential) was used leaving sediment captured and 
sediment stored out of the total renewable resource calculation. 
Tide was the next largest emergy contributor of renewable resources, with the 
marsh area assumed to absorb and more efficiently utilize a greater amount of tidal 
energy (5.53 x1015 sej annually) due to the increased roughness of the marsh surface over 
that of the open water / mudflat condition (7.8 x 1014 sej) (Table 5.1, row 6).  The daily 
pulses of this renewable energy source, in following nature?s pulsing paradigm (W.E., 
E.P. and H.T. Odum, 1995), subsidizes both the mudflat/open water and the emergent 
marsh states, although the marsh is assumed to maximally utilize and convert more of 
these energies into production than the mudflat state. Wind was assumed to have the 
same relative effect on both conditions (2.49 x 1015 sej annually) while wave 
contributions in the largely sheltered tidal embayment were comparatively insignificant, 
although also differed due to the relative areas of open water under both conditions 
(Table 5.1, rows 4-5). 
5.4.1.2 Inputs and Exports 
Emergy imported from the economy in 2000 for the restoration of Kingman was 
calculated to be 3.74 x 1018 sej.  This included year one of the 50 year amortized dollar 
emergy amount from over $5.7 million in direct economic investment as well as the 
emergy of the imported dredged sediments for year 2000 (Table 5.1, rows 11-15).  When 
expressed in emdollars (em$) by dividing the solar emjoules by the emergy to dollar ratio 
for year 2000 (1.00 x 1012 sej/$), this was calculated to be 3.74 x 106 em$ amortized over 
50 years.  No investment from the economy was being directly expended on Kingman to 
 
 
146 
 
keep it in the mudflat condition, as it was supporting itself entirely with natural energies, 
this state was assumed to have zero direct monetary imports from the economy.   
5.4.2 Renewable Energies and Production Exports 
5.4.2.1 Primary Production 
By investing money in marsh restoration, the Kingman system was able to capture 
more renewable energies, and resulted in annual exports of primary production emergy 
increasing from 6.26 x 1015 sej under the pre-restoration mudflat condition, to 2.83 x 1016 
sej with a restored marsh, an annual net increase of 2.20 x 1016 sej in primary 
productivity due to the restoration (Table 5.1, rows 16-17).  Over the 50-year time scale, 
this accounted for a net increase of 1.1 x 1018 sej in exported primary production from 
marsh restoration over the pre-existing mudflat state. 
Under the pre-restoration mudflat state, primary productivity accounted for just 
over 3% of total annual emergy exports, with the remainder owing to the large 
contribution of the fisheries.  Under the restored marsh state, primary production 
increased to over 35% of the total emergy exported from the system annually.  
Renewable resource emergy increased modestly under the restored system state, as much 
of the natural energy contributions to Kingman from the environment were applied 
almost equally under both state conditions.   
5.4.3 Fisheries Production and Export 
Primary production from phytoplankton, mudflat algae, and emergent 
macrophytes were not the only exports accounted for in the emergy analysis.  The 
 
 
147 
 
considerable effort to account for fisheries production under the pre-restoration mudflat 
condition in Kingman to compare to a restored marsh and a mature marsh condition 
revealed some interesting and unexpected results (Chapters 3 & 4).  Under pre-restoration 
mudflat conditions that had existed for decades in Kingman Lake, the total annual sport 
fishery emergy production (with catfish, 1.76 x 1017 sej) was greater than three times that 
of the restored marsh scenario (5.32 x 1016 sej) (Table 5.1, row 18). The greater emergy 
from fisheries production that was supported by the mudflat state over restored marsh 
(1.23 x 1017 sej) was 75% attributable to the biomass of bullhead catfishes coming from 
the mudflat state (1.59 x 1017 sej).   
To take this a step further, the fishery production estimates for the mature Dueling 
Creek marsh system were applied to the Kingman Marsh. This was done with the 
assumption that given enough time (several decades or more) the Kingman Marsh fishery 
might develop in much the same way as Dueling Marsh to reflect somewhat less catfish 
and a more diverse species composition that included top level piscivores.  When these 
calculations were completed, the result was only slightly better, with the mudflat 
condition still producing an emergy surplus of 1.18 x 1017 sej over the mature marsh 
hypothetical scenario (Table 5.4). 
The total emergy exports from the two alternate Kingman systems, including 
primary production and fisheries (with catfish), found the pre-restoration mudflat state to 
be exporting 1.82 x 1017 sej, with the restored Kingman Marsh state only exporting 8.15 x 
1016 sej, a net positive export from the mudflat state of 1.00 x 1017 sej.  The District of 
Columbia fishery is estimated to be a $1.35 million annual contributor to the local 
economy (Appendix C, Table 5.7). This value is also thought to be an underestimate 
 
 
148 
 
considering the basis of the economic numbers were from shoreline angler surveys that 
did not include boat based fishers, or the highly economically intensive bass boat fishing 
public (Cummins and Rockland, 1987; Byers, 2005).  
5.4.4 Catfish and Emergy 
The inclusion of bullhead catfish is the ?normal? state of the fishery, with the 
emergy analysis reflecting more emergy exported for the pre-restoration condition, which 
was greatly influenced by the biomass of catfish caught in the mudflat state as compared 
to the restored marsh.  This is interesting, in that previous emergy analyses that included 
local fisheries estimates and economic data revealed a direct connection between wetland 
area and fishery production (Bahr et al., 1982; Martin, 2002).  Of course these analyses 
were conducted on the highly productive and economically important Louisiana coastal 
fishery, and a direct comparison of results from these studies is not intended.   
The family of bullhead catfish (Ictaluridae) are considered to be commercially 
important, and for the individual angler, an important gamefish (Rohde et al., 1994).  
While there is no commercial fishing allowed in District waters for any species, 
individual anglers are generally free to fish.  Catfish are bottom dwelling omnivorous fish 
and they come into close contact with riverine sediments. In the Anacostia River, bottom 
sediments in several locations were found to be contaminated to different degrees by 
location by PCBs, PAHs, chlordane, DDTs, and several heavy metals (Velinsky et al., 
1997). Due to this acknowledged contamination and the findings of elevated levels of 
many of these contaminants in catfish tissue (Velinsky and Cummins, 1996), the D.C. 
Department of Health, Environmental Health Administration imposed a public health 
 
 
149 
 
advisory on their consumption in the city that has been in effect for over a decade (Byers, 
2005). Potomac River commercial and hook and line catfish harvests for Maryland and 
Virginia were almost 200,000 lbs in 2000, yet dropped to over 1,500 lbs in 2004 and 550 
lbs in 2005 (Potomac River Fisheries Commission statistics, 2007 personal 
communication).  This is likely due to the extension of a fishery advisory for catfish to 
these downriver portions of the Potomac by their resource management agencies. 
A shoreline angler fishery survey conducted in 2004 by the Government of the 
District of Columbia, Department of Health, Environmental Health Administration,  
Fisheries and Wildlife Management Division, found that of the anglers surveyed, about 
19% had caught catfish and that the fish species (family) favored by the anglers was 48% 
catfish (Byers, 2005).  The same survey estimated the annual angler catch of all gamefish 
(Footnote 5.4, Table 1) in the District to be approximately 909kg, of which about 400kg 
(44%) were estimated to be catfish (Appendix C, Table 5.7).    
The catfish public health advisory and its ban on consumption effectively 
removes catfish from the list of important gamefish in the District, just as it has for the 
Potomac River fishery which essentially saw the complete removal of catfish from 
commercial and hook and line harvests in just a few short years despite the fact that some 
numbers are still reported.  For these reasons, and to determination of what kind of 
economically important ?fishery? a restored Kingman Marsh would have, assumptions 
about the fishery production from Kingman also created scenarios that excluded catfish 
from the emergy calculations (Tables 5.1, 5.4, Appendix C, Tables 5.3 to 5.5).  This was 
done much as the 19 other species of fish that were captured during the electrofishing 
surveys (Appendix A, Table 3.5, Appendix B, Table 4.3) were left out of the emergy 
 
 
150 
 
analysis in order to evaluate the effect that the marsh restoration had on the ?fishery? as 
opposed to the fishes as a whole, which was the intention of the analysis. 
5.4.5 No Catfish Emergy Scenarios 
The same emergy evaluations for fishery production that were calculated from the 
full suite of gamefish species typical for freshwater systems (Footnote 5.4, Table 1) were 
recalculated without the bullhead catfishes.  Once this was done, the estimated annual 
fishery production of dry weight fish biomass for Kingman Lake mudflats prior to 
restoration declined from 3.0 x 106 g to 3.12 x 105 g, a difference of 2.67 x 106 g (Table 
5.4).  The restored Kingman Marsh fishery scenario also declined without the catfish 
biomass, from 9.17 x 105 g to 2.13 x 105 g, a difference of 7.04 x 105 g.  This still left the 
restored marsh condition with less annual biomass production than the pre-restoration 
mudflat state, in large part due to larger mudflat biomass values for yellow perch, 
pumpkinseed and bluegill sunfish and values for white perch that were more than half 
that of the restored marsh condition (Appendix C, Table 5.4).    
When translated into emergy values, the pre-restoration mudflat state for annual 
fishery emergy production transformed to 1.82 x 1016 sej (Table 5.1, row 18).  Restored 
marsh fishery emergy production totaled 1.33 x 1016 sej, a decrease in emergy from the 
mudflat condition of 4.9 x 1015 sej, making the marsh restoration a net loss in terms of 
sport fishery emergy production even without catfish contributions.  Mature marsh 
emergy from fisheries without catfish came to 3.73 x 1016 sej.  This represented the first 
fishery scenario in which the marsh emergy came out ahead of the mudflat emergy, 
revealing the value of a mature marsh to fisheries that may only come with development 
 
 
151 
 
over time. 
5.4.6 Total Annual Emergy Exported (Without Catfish) 
The total annual emergy exports from the two alternate Kingman systems, 
including primary production and fisheries without catfish, found the pre-restoration 
mudflat state to be exporting 2.45 x 1016 sej, with the restored Kingman Marsh state 
exporting 4.16 x 1016 sej, a positive export from the restored marsh state of 1.71 x 1016 
sej, or 1.7 times that of the mudflat state (Table 5.1, row 18).  The same was true of the 
mature marsh fishery scenario with 3.28 x 1018 sej in total emergy exported over 50 
years, an increase of 1.41 x 1018 sej, or 2.7 times that of the 50 year old mudflat state.  
Conversely, if catfish were left in the equation, the mudflat state would have had 4.78 x 
1018 sej, or 2.1 times that of the mature marsh.  This scenario, without catfish, revealed 
that a simple adjustment in species biomass allocation to fall in line with social 
management objectives can have much different emergy outcomes that are not 
immediately evident from total biomass or even energy conversions to joules.   
The annual total energy in joules from the combined primary production and 
fisheries exported from the restored and mature marsh scenarios (with or without catfish) 
were greater than that of the mudflat state, although once the joule energy was 
transformed by their respective emergy to joule ratios, only the marsh conditions without 
catfish came out ahead in emergy over the mudflats.  This attests to the power and value 
of having a common unit basis in solar energy, and transformities that adjust the relative 
quality, or true value of the energy.  This methodology is intended to be used in 
developing and analyzing differing types of data in a comparative way to be able to make 
 
 
152 
 
more informed management decisions.  The system that exports the greatest amount of 
emergy to the public economy is usually considered to be the socially more beneficial, 
and it would seem that without catfish being considered a part of the sport fishery in the 
District of Columbia, a newly restored and mature marsh system improves the human 
economy with respect to the fisheries. 
5.4.7 Emergy Indices 
Emergy indices (Table 5.3) are developed from calculations applied to the data 
from the emergy analysis (Table 5.1) to help quantify the production export contributions 
of the mudflat and emergent marsh systems to the Anacostia River ecosystem. 
Additionally, when alternately compared natural systems provide an economic service to 
the local economy, as with a sport fishery, the system that exports the greatest amount of 
fishery emergy to the public economy is considered to be the best system for the 
economy in providing certain services (fisheries) to society (Brown and McClanahan, 
1996; Martin, 2002).   
 
 
153 
 
 
Table 5.3 Emergy indices for Kingman fisheries 
Kingman Kingman Kingman Kingman
Description Calculation Units Without With With Without 
Marsh New Marsh Mature Marsh Marsh
(Mudflats) (Restored) (Restored) (Mudflats)
(Years 1-50) (Years 1-50)
Note
1 Renewable Emergy Flow Re E+13 sej 96288 99756 4987810 4814382
2 Total System Export Y
     (system with catfish) E+13 sej 18191 8145 431930 909567
     (system without catfish) E+13 sej 2447 4161 327768 122361
3 P E+13 sej 626 2827 141372 31280
4 Fisheries Emergy Yield Ratio Y/(Re+P)
     (system with catfish) sej/sej 0.19 0.08 0.08 0.19
     (system without catfish) sej/sej 0.03 0.04 0.06 0.03
5 Fisheries Emergy
     (system with catfish) E+13 sej 17566 5318 290559 878287
     (system without catfish) E+13 sej 1822 1334 186396 91082
6 Fisheries Emdollar Value
     (system with catfish) em$ 175,660 53,180 2,905,590 8,782,870
     (system without catfish) em$ 18,220 13,340 1,863,960 910,820
Notes on values taken from Table 5.1
1 Renewable emergy flow: Rain + River geopotential + Waves + Tide + Riverine water used-mudflat + Riverine water used-marsh
2 Exports (yield) from Kingman: NPP mudflat and open water + NPP marsh + Fisheries scenarios with and without catfish
3 Primary Production: NPP mudflat and open water + NPP marsh
4 Fisheries Emergy Yield Ratio: Total exports under each scenario divided by renewable resources plus primary production
5 Fisheries Emergy: Values for each scenario taken from Table 5.1
6 Fisheries Emdollar Value: Fisheries emergy divided by emergy to dollar ratio for US for year 2000 (1.0E+12 sej/$)(Odum,2001)
Primary Production
 
5.4.7.1 Fisheries Emergy Yield Ratio 
 An interpretation of net fisheries benefit, this ratio is the emergy of the fisheries 
production export, or yield from the system (Y) divided by the emergy of the renewable 
resources (Re) plus the primary production (P) that ultimately supports the fisheries.  This 
ratio (Y/(Re+P)) indicates which system state (intertidal mudflat or emergent marsh) and 
condition (unchanged mudflat, newly restored or mature marsh) is more productive from 
a fisheries perspective.  The process with a greater ratio produces more fisheries to the 
system with larger values representing larger contributions.  Fisheries production emergy 
 
 
154 
 
yield for the mudflat, restored marsh and 50 year mudflat and mature marsh are 
calculated in Table 5.3.  Under the one year mudflat state the fisheries production ratio 
with catfish calculated to 0.19, more than twice that of the restored marsh state of 0.08.  
If catfish were taken out of the fisheries equation, the mudflat state produced the lesser 
value of 0.03 while the restored marsh state fishery production ratio value was greater at 
0.04.  When the 50 year time frame is applied the index calculation reveals that under the 
fishery scenario that includes catfish the mudflat state is again more productive (0.19) 
than the mature marsh (0.08) while removing the catfish from the fishery reverses the 
result with the mature marsh producing a fishery index value of 0.06, twice that of the 
mudflat state. 
5.5 Discussion 
The emergy analysis from Martin (2002), on which much of the Kingman 
analysis is patterned, created a year 1 and 50 year time horizon for the analysis based on 
the life expectancy of the river diversions that were constructed to supply sediment to 
build and expand Louisiana coastal marshes.  Reasonable assumptions were made about 
marsh accretion rates as a result of the diversions, and the emergy and emdollar return on 
investment was significant given the time allowance and the economic importance of 
expanded wetland coverage to the lucrative coastal fishery as well as providing for 
greater flow of renewable energies (Bahr et al., 1982; Martin, 2002).  With the Kingman 
Marsh, the unexpected effects of goose herbivory and surface elevation reduction on the 
marsh restoration significantly decreased the area and extent of emergent marsh coverage 
from its original planted cover in year 2000 of 100% to only 25% coverage four years 
 
 
155 
 
later (Hammerschlag et al., 2006). For these reasons, an estimated expansion of marsh 
area from its existing cover in year 2000 was not considered and for the hypothetical 50 
year time horizon the year 2000 coverage was kept. 
Large amounts of low transformity, renewable river inputs and primary 
productivity support higher quality emergy products such as fish.  The investment in 
restoration of coastal tidal marshes can in some ways be compared to agricultural systems 
where imported energies are used to support and capture the natural energies that yield a 
valuable economic export.  The capture of sun, rain and wind by agriculture systems still 
requires substantial imported energy from the human economy, which results in a lower 
renewable emergy captured. Unlike agriculture, the investment in marsh restoration is 
usually a single up front cost that sets the stage for natural processes to take over and 
sustain the system supporting processes and production (as with fisheries) to enhance the 
ecological and human economy. 
As stated earlier, the D.C. urban recreational fishery was estimated at $1.35 
million annually.  With the significant dependence of many species of migratory and 
resident gamefish on wetland energy transformation and flow, improving the state of 
coastal wetlands is usually seen as a very positive action providing benefits to local and 
regional economies (Bahr et al., 1982; Odum et al., 1984; Houde and Rutherford, 1993; 
Mitsch and Gosselink, 2000; Martin, 2002).  
The conversion of fisheries emergy to an emdollar (em$) equivalent allows the 
marsh fisheries production export to be related to the economy.  This is done by dividing 
the fisheries emergy flow by the emergy to dollar ratio for the year of the investment, 
which in 2000 was 1.0 x 1012 sej/$ (H.T. Odum, 2001). Interestingly, the mudflat 
 
 
156 
 
condition produced a higher fishery emergy and emdollar value than the restored marsh 
condition at Kingman, unless catfish were removed from the calculation under the 50 
year projection (Table 5.3).  
Marsh systems also benefit societies in ways that often transcend easily accounted 
for economic benefits.  These benefits, such as aesthetic enjoyment and recreational use, 
potential educational opportunities for schools and the general public, scientific studies 
related to restored systems, nutrient transformation, sediment trapping, flood attenuation 
and wildlife habitat, are all amplified when marsh restoration occurs in close proximity to 
human development, such as in urban areas like the District of Columbia.  These benefits, 
as well as creating a diversity of habitat types and species found in the Anacostia system, 
are important factors to consider when valuing restoration activities. 
5.6 Conclusion 
Through the application of emergy analysis and environmental accounting 
methodologies, emergy calculations can illustrate the potential of investment in 
ecologically engineered natural systems to harness natural renewable "free" energies for 
the benefit of society and the environment (Odum ,1996; Martin, 2002; Kangas, 2004a; 
Mitsch and Jorgensen, 2004). With emergy analysis, the value of marsh restoration to the 
natural and human economy can be better understood in general and insights drawn with 
the Kingman Marsh restoration in particular.  
5.6.1 Fisheries use of Tidal Freshwater Mudflats 
Kingman?s longtime status as a shallow water tidal impoundment, dominated by 
 
 
157 
 
mudflats at low tide, was considered to be of limited habitat value (Syphax and 
Hammerschlag, 1995). There has been significant research on and recognition of the 
habitat value of intertidal mudflats in marine and some estuarine systems (Peterson, 
1981; Quammen, 1984; Zedler, 2001), especially in other countries such as Germany and 
Japan (Riese, 1985, 2001; Hosokawa, 1997; Imamura, 2000).  The benthic invertebrate / 
fish / bird interactions are especially notable. 
The importance and ecological role of tidal freshwater marshes have been studied 
as mentioned earlier, although their associated intertidal mudflats have not been widely 
investigated (Serodes and Troude, 1984).  The initial focus of the fishery survey of 
Kingman was to characterize the fish utilization of the expansive mudflat areas in the 
total absence of any emergent marsh vegetation.  A similar survey was also ongoing in 
the restored Kenilworth emergent marshes and mudflat areas as well as a nearby mature 
tidal marsh at Dueling Creek, with the goal of making generalized comparisons of the 
fish populations from each of these nearby systems in different stages of development.   
The fisheries analysis provided for an interesting dilemma in that the pre-
restoration mudflat state produced greater annual gamefish biomass and emergy (with 
catfish) than either the restored or mature marsh states when these conditions were 
applied to the Kingman restored marsh scenario.  This result was due entirely to the 
massive contributions of bullhead catfish biomass that were surveyed by electrofishing 
boat from the Kingman mudflats in the years prior to the marsh restoration.  While the 
restored marsh of Kenilworth and the mature marsh of Dueling Creek both supported 
catfish as well, their fish communities were somewhat more balanced in terms of species, 
numbers and biomass. These marsh sites also had numbers of higher trophic level 
 
 
158 
 
piscivores in largemouth bass while Kingman had none.  Because catfish are typically 
included as a part of any sportfishery, initial emergy calculations naturally included them.  
A rationale was created for the removal of catfish from each fishery emergy evaluation 
scenario to see what effect their removal would have on the emergy analysis in terms of 
fishery production export and total production export.  For strictly fishery production 
export, only the mature marsh scenario without catfish produced more fish emergy, while 
under total exports including primary production the restored marsh and mature marsh 
(again without catfish) produced a greater total emergy export scenario. 
These created scenarios, and their emergy analysis, reveal how significant an 
impact that the mudflat condition has on the energy flow of the system when compared to 
the marsh system.  Removing the bullhead catfish from the ?fishery? might seem 
arbitrary and not a true representation of the natural state of the fishery. This has been 
done for other species of fish for health reasons as in the District ban on carp and eel 
consumption, and stock recovery as was done with striped bass. A fishery is really an 
arbitrary human creation dictated by the norms of the society that utilize it.  In many 
cultures carp are a significant aspect of a fishery, as are eels and numerous other species 
that some cultures don?t even consider attempting to catch much less eat.   Catfish have 
been and will probably continue to be a significant aspect of the D.C. fishery despite a 
health advisory, with 65% of anglers surveyed reporting that they eat the fish they catch 
and of those 48% responding that the fish they prefer to catch and eat is catfish, despite 
the fact that 75% of anglers acknowledge that they are aware of the public health 
advisory on the fish (Byers, 2005). 
When comparing the fisheries of the two system states of Kingman (intertidal 
 
 
159 
 
mudflat and emergent marsh), from the human fisheries oriented perspective the one 
considered the most beneficial is that which exports the most fishery emergy to the public 
economy and natural systems (Brown and McClanahan, 1996; Martin, 2002).  With this 
in mind, the mudflat state is the greater of the two given the local anglers desire to keep 
catching catfish. 
5.6.2 Emergy and Alternate State Theory Informing Restoration  
The application of the emergy analysis method of environmental accounting to 
the comparison of mudflats and emergent marshes helps to build a logical framework 
from which alternate state theory may be better understood in relation to the Anacostia 
River's natural history and present change.  As ecological processes are thought to be 
shaped by a system's history (Peterson, 2002; Price and Morin, 2004) an assessment and 
valuation of the mudflat characteristics that defined the Anacostia for almost 60 years 
should be considered invaluable in understanding and refining marsh restoration 
strategies.  The mudflats that evolved over the last half of the 20th century essentially 
created a system model that had adapted to its present state.  Dominated by mudflat algae 
and plankton as the primary production drivers, catfish thrived on the smooth mud plains 
that supported benthic aquatic worms, aquatic insect larvae, mollusks and the smaller 
forage fish that eat them.  Migratory shorebirds also benefited from the mudflat 
condition.   There is existing value in mudflats, which are valued differently in different 
societies such as Germany and Japan, and even in regions within the same society (Reise, 
1985, Hosokawa, 1997; Imamura, 2000) 
The forcing functions involved in a change to an alternate state are often sudden, 
 
 
160 
 
powerful and difficult to reverse (Knowlton, 1992; Peterson, 2002).  An analysis of the 
human (Army Corps of Engineers construction) and natural (goose herbivory) switching 
functions between the marsh and biogenic mudflat alternate states described in Chapters 
3 & 4 revealed linkages and disconnects to outcomes that were not immediately obvious 
to planners prior to the completion of restoration activities at Kingman Marsh.  While the 
evidence of goose herbivory at Kenilworth Mass Fill 3 was noticed by resorationists, it 
was a small area and the emergence of a marsh in a majority of the other sites at 
Kenilworth was seen as a success.  Nature does give us clues, whether we decide to 
follow them is up to us. 
Because the restoration and management of ecosystems involves an inherent 
degree of uncertainty (Zedler, 2001), especially in systems that have alternate state 
potentials (Peterson et al., 2003), the value of an emergy analysis and its potential to 
inform the decision making process and shape environmental policy could save managers 
and society a considerable amount of time and expense (Odum, 1996; Martin, 2002). 
Fisheries improvement was never a primary goal of the restoration, as it was a driver of 
the emergy analysis completed here.  Nutrient uptake, sediment trapping, marsh bird 
habitat, public education, research and aesthetics, as well as a desire to reclaim a piece of 
marshland that was an important part of the Anacostia River were some of the additional 
goals of the restoration. 
The nature of political ecology (Peterson, 2000) and the ecological mores of our 
society were the driving forces behind the restoration of Kingman Lake to Kingman 
Marsh.  The US Army Corps of Engineers had dredged away almost all of the Anacostia 
River?s expansive marshes over the first half of the 20th century.  The restoration of these 
 
 
161 
 
marshes is generally seen as a positive endeavor, tidal marshes are now valued in most of 
U.S. society.  The value of tidal flats to other societies, such as Japan, reveals how 
different cultures can look at ecosystems differently, and subject their societal mores 
upon them.  And so it becomes a social value judgment in the end as to what perceived 
benefits are desired through the restoration of any natural system, and emergy analysis 
can attempt to address and inform these judgements.  
 As resource managers commit to becoming increasingly involved in the often 
costly restoration of ecosystems, there is the potential for societies and their economies to 
expect that these expenditures be utilized in more efficient ways under the likely scenario 
of shrinking future regional and world economies (Odum and Odum, 2000).  It will also 
become significantly more important to have information which will help to make 
informed decisions about the direction that restoration will take, as with fisheries used in 
this analysis, that may be expected with large capital expenditures (Odum, 1996, 
Peterson, 2000; Martin, 2002, Kangas et al. 2004).   
In the case of the Kingman Marsh restoration, the fishery emergy yields under the 
mudflat state and that of the restored marsh state were entirely dependent upon the 
inclusion or exclusion of catfish from the analysis. Only once catfish were removed from 
the evaluation under the future mature marsh scenario was the fisheries yield from a 
restored marsh greater than that of the pre-existing mudflat condition. Mudflats are 
understood to be ecologically valuable in other societies, their value to the Anacostia 
River and the local economy was poorly understood until the completion of the fishery 
emergy analysis comparing them to restored and mature emergent marshes.  The 
complete services that the Anacostia River tidal freshwater marshes provide are still 
 
 
162 
 
poorly understood, and should be investigated further in future work. The restoration of 
ecosystems is still an evolving discipline, and with emergy analysis informing 
environmental decision making projects that are in early planning stages may be balanced 
against natural existing conditions to potentially provide improved restoration outcomes 
depending upon the restoration goals.   
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
163 
 
FOOTNOTES 
Footnote 5-1. Phytoplankton 
Phytoplankton primary productivity in tidal freshwater marshes is considered to 
be largely unknown although critical to any energy flow analysis (Odum et al., 1984; 
Mitsch and Gosselink, 2000). Primary production in the form of phytoplankton in 
Kingman may be considered to be a function of the percentage of time that the waters of 
the Anacostia River are within the Kingman area that is subtidal. Under both pre and post 
restoration scenerios this is considered to be the same, as the only areas that were ?filled? 
with dredged river sediments were not subtidal, but were those that included intertidal 
mudflats which were exposed at low tide.  The annual phytoplankton production for 
Kingman was extrapolated from Maryland Department of Natural Resources data for the 
tidal freshwater Potomac River (DNR, 2007).  While the Maryland data was a measure of 
the rate of annual phytoplankton production (315.7 g C/m2/yr), it was also a measure for 
the entire water column of the mainstem Potomac River, often upwards of 30 ft. in depth 
in this area. Although under high turbidity the euphotic zone is limited to the upper 1 to 
2.5 m with highest photosynthetic rates in the upper 0.5 m of the water column, tides and 
winds do mix the water column and lower rates of productivity extend to some greater 
depths (Ragotzkie, 1959).  
To account for the reported production rate of the whole Potomac water column, 
the rate was divided by a factor of five to adjust the rate to the depth of the Kingman 
subtidal areas, which at high tide are on average approximately 6 ft (personal depth 
observations).  The estimated value used for this analysis (63.1 g C/m2/yr), when 
 
 
164 
 
compared to reported values for southeastern estuaries and coastal waters (67-375 g C 
/m2/yr) are on the low end with even the higher values considered conservative (Pomeroy 
and Wiegert, 1981).  Additionally, the value that is used for Kingman which is applied 
only to the subtidal area, was divided in half to account for the turnover of half the tidal 
prism in the marsh each day.  Annually, 100% of the phytoplankton is assumed to be 
exported from Kingman. 
Footnote 5-2.  Mudflat Algae 
Primary production values for Kingman included estimates for mudflat algae.  
Algae on mudflats and in emergent marshes make important contributions to numerous 
trophic levels and taxa, including nematodes, chironomids, gastropods and numerous fish 
(Reise, 1985; Campeau et al., 1994).  Investigations into intertidal epibenthic algae 
productivity have been primarily found in saltmarsh systems (Gallagher and Daiber, 
1974; Van Raalte et al., 1976a; Pomeroy and Wiegert, 1981). Some measurements of 
algal productivity within the emergent saltmarsh plant community were found to be as 
high as 30% of net annual production (Gallagher and Daiber, 1974).  A study of algae 
specifically separating out a bare intertidal creekbank from measurements made within 
Spartina alterniflora and S. patens at high and low tide found the bare creekbank surface 
algae production to be 12-19 times more productive respectively when exposed than that 
of algae measurements made within the Spartina (Whitney and Darley, unpubl. data, 
from Pomeroy and Wiegert, 1981).   An annual estimate of net productivity of 
approximately 190 g C/m2 of mudflat algae was determined from the above study with 
more than 75% of the production occurring during ebb tide. 
Algal production in tidal freshwater wetlands has been asserted to be less than one 
 
 
165 
 
percent of net annual production (Whigham and Simpson, 1975; Whigham and Simpson, 
1976), far less than the 30% reported by Gallagher and Daiber for saltmarsh algae (1974).  
This is not incongruous considering that the much taller and broader leaves of tidal 
freshwater macrophytes provide for considerable shading during the growing season 
(Mitch and Gosselink, 2000) than that of the short and slender Spartina sp.  While 
shading would account for reduced algae growth within stands of freshwater marsh 
vegetation as compared to that of saltmarshes, the lack of research into tidal freshwater 
mudflat algae production reasonably allows for the assumption that algal productivity 
there may be roughly equivalent to that of saltmarsh exposed creekbanks.   
Observational evidence of dense algal mats forming on Kenilworth mudflat 
exclosures (detailed in Chapter 3) supports the assumption that algae production on 
Anacostia tidal flats can be a significant contributor to overall net primary productivity.  
This is especially important considering that the turnover rate of algae is much faster than 
that of emergent macrophytes (Mitsch and Gosselink, 2000) and is also growing year 
round.  An assumption that 100% of algae NPP is exported from the Kingman system is 
made for the emergy analysis. 
Footnote 5-3.  Emergent Macrophytes 
Typically the largest component of primary production in the tidal freshwater 
marsh comes from the emergent marsh plants (Whigham et al., 1978; Odum et al., 1984; 
Mitsch and Gosselink, 2000).  Estimates of annual marsh plant net primary production 
that are obtained by a single, aboveground harvest method are not considered accurate, as 
they do not include data on belowground biomass, growing season mortality of whole 
plants and plant parts, herbivore consumption, or the production of species that may 
 
 
166 
 
dominate before or after the single harvest (Whigham et al., 1978).   
The value used as the best estimate surrogate for the Kingman emergent marsh 
annual NPP (2,346.5 g/m2) was obtained from an estimate that included multiple harvests 
throughout the growing season of 14 species of annuals and perennials, as well as 
estimations of leaf mortality and belowground production (Whigham et al., 1978).  While 
this estimate is more complete than the many estimates of single harvest aboveground 
annual NPP often quoted in the literature for tidal freshwater wetlands (Whigham et al., 
1978; Simpson et al., 1983; Odum et al., 1984; Mitsch and Gosselink, 2000), it is still 
considered an underestimate as it did not include belowground production for all species 
sampled or account for herbivory (Whigham et al., 1978).   
As stated earlier, annual marsh NPP export was set at 80% to account for 
reductions in total production export from incomplete decomposition, some trapping of 
plant material in marsh sediments and the inclusion of belowground root biomass in the 
calculations which is not exported (Odum and Heywood, 1978; Whigham et al., 1978).  
Footnote 5-4.  Fisheries  
Tidal freshwater marshes are important spawning, nursery and feeding habitat for 
numerous resident freshwater, estuarine, anadromous, semi-anadromous and catadromous 
species of fish (Odum et al., 1984; Rozas and Odum, 1987; Mitsch and Gosselink, 2000), 
with estuarine and coastal marshes in general considered critical to the support of fishery 
stocks (Boesch and Turner, 1984; Deegan, 1993; Houde and Rutherford, 1993; Mitsch 
and Gosselink, 2000; Martin, 2002).    Improvement of the fishery is often an argument 
for the restoration of emergent marsh habitat (Able et al., 2000, Mitsch and Gosselink, 
2000; Zedler, 2001).  The emergy analysis conducted by Martin (2002) relied on a linear 
 
 
167 
 
relationship between primary production and fisheries yield developed by Nixon (1988) 
and supported with commercial fishery / energy information from Bahr et al. (1982).  For 
the Kingman emergy analysis, site specific fisheries data was collected from Kingman 
itself, from the adjacent restored Kenilworth Marsh (1993), as well as from Dueling 
Creek marsh, a nearby reference system that has been unaltered for over 50 years.  The 
methodology and complete fisheries dataset, collected over multiple years and seasons, is 
reported in Chapters 3 and 4.   
For the purposes of this analysis, the fishery data used for Kingman included only 
those species considered to be a part of the D.C. urban fishery as sport / game fish (Byers, 
2005) (Footnote 5.4, Table 1).  This followed the methods of Bahr et al. (1982) and 
Martin (2002), as well as representing the potential of Kingman to support the top trophic 
energy levels that could be available to the D.C. fishery. 
 
 
 
168 
 
Footnote 5.4, Species of fish used in the Kingman emergy analysis 
 
As detailed in the fish survey methodology of Chapter 3, approximately the same 
level of electrofishing effort was expended at each site and the surveys covered 
approximately the same total area.  The only differences were if the shocking occurred 
over mudflats with no emergent marsh vegetation, or in flooded tidal ?gut? channels 
surrounded by emergent marsh vegetation.  Kingman was only sampled in the pre-
restoration, or mudflat condition, while Kenilworth had two marsh survey locations and 
one mudflat location, and Dueling had one marsh survey location representing a mature 
marsh.  The Kenilworth marsh fish survey data was used as a surrogate for the Kingman 
restored marsh condition and the Dueling fish survey data was used as a hypothetical, 
future Kingman mature marsh condition.   
To determine the total annual production of fish biomass from each system for the 
emergy analysis, several calculations were made from the fish survey data.  First, the 
total numbers of each fish species captured from each survey area were aggregated by 
 
 
169 
 
each location and each system (mudflat or marsh), then divided by the number of surveys 
conducted at each ecotype to obtain an average fish abundance for each species at every 
site (Appendix C, Tables 5.3-5.5).  Because each site was sampled at least once and often 
more than once in each of the Spring, Summer and Fall seasons, the fish survey numbers 
were assumed to be indicative of annual average species abundance and composition.  
The average numbers of individuals of each fish species caught at each survey site 
were multiplied by an average individual fish wet weight as determined from the actual 
fish survey values (Chapters 3 & 4) or as found in the literature (Appendix C, Table 5.6) 
to get an average wet weight biomass for each species of fish for each location. An 
interesting aspect of the fish data was realized when this was done.  It became apparent 
that the Kingman and Kenilworth mudflat survey areas were heavily dominated by 
catfish biomass and absent of the top piscivore the largemouth bass.  At the same time, 
both the Kenilworth and Dueling emergent marsh sites had largemouth bass and greater 
biomass of white perch while the Dueling site had also greater average fish weights of 
yellow perch, striped bass and pumpkinseed sunfish.  The difference in species 
composition was likely due to feeding strategies, with catfish being tactile omnivorous 
bottom feeders comfortable with the mudflat environment and piscivores such as the 
largemouth bass more visual and comfortable in areas with some shade and structure 
where they can hide and wait for prey to swim by (Rohde et al., 1994).  Because of the 
potential impact that the different species composition and abundances was expected to 
have on total fish productivity for the mudflat vs. restored marsh vs. mature marsh 
evaluation scenarios being developed for the emergy analysis, it was decided to create 
two sportfish biomass production scenarios, one with catfish and one without. 
 
 
170 
 
The average wet weights of sportfish species were segregated out from the other 
species caught and tabulated under each sampling location/system type and a separate 
table created for the no catfish scenario (Tables 5.4). These values were then divided by 
the total area sampled, which was standardized across all sites (approximately 500 m2) to 
give the average biomass (g/m2) collected for each species at each sampling location.  
Annual fish production was determined by multiplying the average biomass for 
each species by a production to biomass (P/B) ratio.  Production is generally defined as 
the total net generation of new fish tissue over a period of time by a species-population, 
usually reported in g/m2/yr (Chapman, 1978). Because only a few reasonable estimates of 
production for some freshwater species have been made, and P/B ratios developed for 
them to aid in research on fish populations, an average of P/B ratios was taken for the 
various species where reported values were given (Chapman, 1978; Wetzel, 1983), and 
that average ratio (1.5) applied to each of the gamefish species to obtain annual fish 
production numbers for the emergy analysis. 
Production was then converted to an energy value per unit area.  This was 
calculated by first converting fish wet weight to dry weight with the assumption that dry 
weight is approximately 20% of wet weight (Waters, 1977).  The production of fish dry 
weight biomass (g/m2/yr) for the exact area sampled was totaled across all species under 
the catfish/no catfish scenarios and then multiplied by an estimation of the total area (m2)   
that could support the upper level consumer fishery at each location to give an estimated 
total gram dry weight production for each site.  These values were then converted to 
energy equivalents. The approximate dry weight of 1g of fish is assumed to have the 
caloric equivalent of 4.5 Kcal (Bahr et al., 1982).   
 
 
171 
 
Table 5.4 Summary calculations of secondary production for middle and upper level fish 
consumers  
 
 
 
 
 
 
172 
 
 
Multiplying the values gave an energy equivalent production in Kcal/m2/yr.  
 
 
Transformities for fish energy conversions were calculated from Bahr et al., 
(1982).  As stated earlier, transformites are obtained by dividing the total emergy that was 
used in a process by the energy yielded by the process (Odum, 1996).  The energy flux 
rates from Bahr et al. detailed solar inputs and outflows through the Louisiana coastal 
ecosystem and are used as best approximations for this analysis.  Solar inputs (1.0 J solar 
energy = 1.0 solar emjoule) to wetland producers (369 x 1012 Kcal/yr) and solar inputs to 
phytoplankton and benthic algae (134 x 1012 Kcal/yr) gave a total of 503 x 1012 Kcal/yr.  
This value was divided by the outflows of energy from middle consumers (0.38 x 1012 
Kcal/yr) and upper consumers (0.29 x 1012 Kcal/yr) and converted to Joules to give 
transformities of 3.02 x 106 sej/J for middle consumers and 3.48 x 106 sej/J for upper 
consumers.   
These transformities were applied to each of the gamefish species to give an 
annual emergy estimate for the fish found in each area, with separate tables created for 
gamefish communities with catfish and without catfish (Table 5.4; Appendix C, Tables 
5.3-5.4).  All gamefish were considered to be middle level consumers due to their 
reported feeding habits (Rohde et al., 1994), except for the largemouth bass, which was 
considered an upper level consumer.  The striped bass would have been given upper level 
consumer status but for the fact that the majority of these fish captured in the surveys 
were juveniles so they were considered middle level consumers. 
In aggregating the individual species production rates for each area, a new table 
was formed (Table 5.3), which broke down each sampling location and again created 
 
 
173 
 
different scenarios for catfish.  Again, these production numbers were applied to the total 
amount of fishery useable area for each location.  As reported in Chapter 4, electrofishing 
surveys were not conducted in Kingman after marsh restoration was completed. To 
determine what the Kingman Marsh fishery production would be under the restored 
marsh scenario, the Kenilworth Marsh mass fill areas (MF1 & MF2) fish abundances 
were aggregated and averaged from data reported in Chapter 3 to use as a hypothetical 
surrogate for Kingman Marsh post-restoration.  These values were entered into Table 5.3.  
Additionally, the Dueling Creek Marsh fishery production values were also used as a 
hypothetical surrogate for Kingman Marsh fishery production after many decades of 
marsh maturation and development. An average fishery transformity (3.07 x 106 sej/J) 
was used incorporating the middle and upper level fish consumer transformities and 
applied to the energy values calculated for each location and hypothetical scenario (Table 
5.3) and applied to the general fishery in the analysis Table 5.1. 
 
 
174 
 
Chapter 6: Conclusion 
6.1 Alternate Ecosystem State Changes on Anacostia River 
The Anacostia River has a long history of human involvement which has many 
times significantly altered the structure of the system at general and discrete points in 
time (Figure 6.1).  Impacts such as erosion and sedimentation from colonial tobacco 
farming are believed to be indirectly responsible for the expansion and maintenance of 
renowned 19th Century wild rice (Zizania aquatica) marshes.  Subsequent dredging and 
filling of these marshes from the 1920's-1940's by the Army Corps of Engineers 
essentially switched the tidal river system from a shallow emergent marsh dominated 
state to a short lived deeper water state lacking emergent marshes.   
The high sediment loads associated with post WWII development in the 
watershed switched the system, including Kenilworth and Kingman which had been 
dredged of its marshes, to one dominated by intertidal mudflats.  Mudflats are known to 
potentially maintain themselves for several hundred years and more if not forced into 
another state by a significant disturbance (Pethick, 1996), and for 50 -60 years the 
Anacostia River had large expanses of intertidal mudflats.  The implementation of the 
Kenilworth Marsh (1993) Kingman Marsh (2000) and River Fringe (2003) restoration 
plans developed in the early '90's by the Army Corps of Engineers (USACE, 1992) was 
an effort to effectively switch almost 40 ha of mudflat back to emergent marsh again.   
These defining events in the Anacostia's natural history may be interpreted as 
massive disturbance pulses that have been the triggers for significant alternative state  
 
 
175 
 
 
Figure 6.1 Human involvement in Anacostia River system state changes over 400 years. 
 
 
 
176 
 
changes (May 1977; Odum, 1995; Knowlton, 1992; Rietkerk and van de Koppel, 
1997).  In essence, the Army Corps of Engineers dredging project has been the catalyst 
for changes to two alternate states within the last 60 years, marsh to mudflat and mudflat 
to marsh.  While these were human induced impacts and effectively caused multiple state 
changes to occur within the Anacostia River, the unplanned response of the system to 
these disturbances is of equal if not greater interest.  In Zedler's book on restoring tidal 
wetlands (2001), she notes that ?the effects of unplanned disturbances on natural or 
restored ecosystems offer an opportunity to learn a great deal about restoration site 
management and the resiliency or frailty of restored marshes?.   
Learning from unplanned or planned disturbances of wetland restoration sites can 
prompt lines of questioning and hypothesis testing that will invariably further the field of 
ecological engineering and restoration ecology (Zedler, 2001, Kangas et al., 2003; Mitsch 
and Jorgensen, 2004).  To this end field experimentation is essential.  As the only 
manipulative research experiment employed to monitor and measure the direct effects of 
the goose herbivory impacts to the Kingman Marsh restoration, these exclosure studies 
represent an important contribution to the understanding of wetland restoration ecology.   
Figure 6.2 is an elegant model of the interactive effects that sediment surface 
elevation and grazing pressure have on wetland vegetation communities from the subtidal 
through the intertidal to the terrestrial zones of a hypothetical marsh system on Anacostia 
River.  As detailed in the exclosure study of Kingman Marsh mudflats in Chapter 4, 
resident goose grazing pressure was determined to be the dominant force controlling the 
presence of emergent marsh vegetation within the low to middle marsh sediment surface 
elevation gradient.  It was also recognized that in the absence of grazing pressure through 
 
 
177 
 
 
Figure 6.2 Representation of the dual influences  of grazing pressure and sediment surface elevation 
in determining the potential extent of mudflats (Kangas et al., 2004). 
 
the use of exclosures, emergent vegetation would develop within this middle level 
elevation gradient into an emergent marsh condition.  This is represented by the marsh 
quarter of the elevation axis in Figure 6.2, which extends to some hypothetical break 
point halfway between the low and high grazer pressure axis after which grazing pressure 
becomes so high that the mudflat state will eventually take over from its alternate marsh 
state.  
The results of the Kingman Marsh exclosure study also revealed the strong effect 
of elevation on the emergent plant community below a break point on the elevation 
gradient.  This is represented as the upper quarter of Figure 6.2 labeled mudflat.  Even 
with little or no grazer pressure, the physical effects of elevation influencing depth and 
 
 
178 
 
duration of tidal inundation prevents most species of emergent plants from establishing 
themselves.  Areas between tidal mean low water and the lower end of the low marsh 
elevation gradient may be expected to exist as physical mudflat at low tide.  These 
physical mudflats are typically associated with tidal marshes to some extent due to 
physical stresses and do not have the potential to switch to an alternate marsh state easily.  
It is these areas that the Army Corps of Engineers believed they were targeting for their 
restoration work.  The fact that the intertidal mudflats of the Anacostia River supported 
no emergent marsh vegetation was thought to be purely elevation influenced and that by 
filling and grading the mudflats to increase the sediment elevations, restoration plantings 
and volunteer vegetation would re-establish a marsh system.   
The Kenilworth and Kingman exclosure studies revealed that some of these 
mudflats that were thought to be too low in elevation to support emergent marshes were 
actually biogenic, influenced by herbivory and bioturbation as well as elevation.  It is 
possible that many of the original Anacostia River mudflats within the low to middle 
marsh elevation gradient could have supported some depth tolerant species of emergent 
marsh vegetation without the Army Corps of Engineers dredging and filling operations.  
The lack of voluntary emergence of species such as Pontederia cordata, Sagittaria 
latifolia and Zizania aquatica on the original mudflats may have been hindered by 
several factors. While much of the original mudflat area may have existed at elevations 
below which even depth tolerant species could survive, in other areas it is likely that the 
combined interactive effects of lower elevation stresses and predation or disturbance by 
birds and fish could have suppressed germination and reduced the stock of viable seeds 
(Baldwin and Pendleton, 2003). Herbivory of those seedlings that did emerge would be 
 
 
179 
 
likely on open mudflats as they would be obvious targets even under low levels of 
grazing pressure.   
The planning for the Kingman Marsh restoration in 2000 relied solely on the 
elevation model of plant zonation that was used for the 1993 Kenilworth project, and had 
not factored into their designs the influence of grazing pressure on marsh vegetation 
establishment. The unexpected effect of voluntary emergent plant growth in Kenilworth 
mudflat exclosures in 1997 revealed that goose herbivory was strongly influencing plant 
emergence and growth on a small mudflat study area.  In 1998 an experimental study of 
this effect in the same exclosures verified the top-down control of emergent vegetation in 
large part through goose herbivory. Because this effect was only seen in a small area of 
Kenilworth with a majority of the restoration plantings not impacted, the grazer influence 
was not expected to be an issue with the Kingman restoration work.   
In hindsight, a careful consideration of the different landscape contexts that 
Kenilworth and Kingman represented and their effects on influencing herbivory would 
have been useful.  The fact that the majority of the Kingman Marsh restoration area was 
surrounded by a golf course with a known resident goose population was an incredible 
oversight in restoration planning.  The surveys of birds at Kingman between 1997 and 
2001 revealed an increasing trend in the numbers of resident geese using the mudflats, 
open water and golf course areas.   
The Kingman and Kenilworth Marsh exclosures revealed that in the absence or 
exclusion of goose grazer pressure a relatively diverse and vigorous vegetative 
community can develop on the biogenic mudflats.  A recognition that predetermined 
marsh restoration goals may not be naturally attainable in the presence of these grazer 
 
 
180 
 
pressures must be realized.  The complete lack of any vegetative growth in the non-
excluded plots speaks volumes of the ability of large grazers to foil the goals of 
expensive marsh restoration projects if their effects are not taken into account.  The 
understanding that the Anacostia River system has alternate system states in biogenic 
mudflats and emergent marshes and that both are of value may move the restoration 
debate away from a rigid imposition of wetland design, toward an understanding that 
unintended change in restoration may be useful in understanding and creating new 
ecosystems. 
Intertidal habitats are widely understood to be of great importance to migrating 
shorebirds and waterfowl (Recher, 1966; Quammen, 1984; Lane and Jensen, 1999; 
Burton et al., 2004).  The dramatic decline of these habitat types over the last century has 
coincided with the increase in development and filling of shoreline areas and the 
hardening of the land/water interface to prevent erosion (Perry and Deller, 1996).  While 
wetlands and other shallow water habitats have been lost at increasing rates, the 
emergence of the intertidal mudflat in low energy urban/urbanizing coastal areas may 
have provided a "new" acceptable alternative for some species of birds.  These altered 
mudflat habitats have the potential to provide for alternative feeding and resting areas for 
migratory shorebirds as well as waterfowl (Lane and Jensen, 1999).  A recognition of the 
value of these habitat types as an integral system component with emergent marshes is 
essential to an understanding of the new ecological dynamic on Anacostia River. 
 
 
 
181 
 
Appendices 
 
 
182 
 
Appendix A ? Kenilworth Marsh Ecological Data Tables  
 
 
 
 
183 
 
 
 
 
184 
 
 
 
 
185 
 
 
 
 
186 
 
 
 
 
187 
 
 
 
 
188 
 
 
 
 
189 
 
 
 
 
190 
 
 
 
 
191 
 
 
 
 
192 
 
 
 
 
193 
 
 
 
 
194 
 
 
 
 
195 
 
 
 
 
196 
 
 
 
 
197 
 
 
 
 
198 
 
 
 
 
199 
 
 
 
 
200 
 
 
 
 
201 
 
 
 
 
202 
 
 
 
 
203 
 
 
 
 
204 
 
 
 
 
205 
 
 
 
 
206 
 
 
 
 
207 
 
 
 
 
208 
 
 
 
 
209 
 
 
 
 
210 
 
 
 
 
211 
 
 
 
 
212 
 
 
 
 
213 
 
 
 
 
214 
 
 
 
 
215 
 
 
 
 
216 
 
 
 
 
217 
 
 
 
 
218 
 
 
 
 
219 
 
Appendix B ? Kingman Lake Ecological Data Tables
 
 
220 
 
 
 
 
221 
 
 
 
 
222 
 
 
 
 
223 
 
 
 
 
224 
 
 
 
 
225 
 
 
 
 
226 
 
 
 
 
227 
 
 
 
 
228 
 
 
 
 
229 
 
 
 
 
230 
 
 
 
 
231 
 
 
 
 
232 
 
 
 
 
233 
 
 
 
 
234 
 
 
 
 
235 
 
 
 
 
236 
 
 
 
 
237 
 
 
 
 
238 
 
 
 
 
239 
 
 
 
 
240 
 
 
 
 
241 
 
 
 
 
242 
 
 
 
 
243 
 
 
 
 
244 
 
 
 
 
245 
 
 
 
 
246 
 
 
 
 
247 
 
 
 
 
248 
 
 
 
 
249 
 
 
 
 
250 
 
 
 
 
251 
 
 
 
 
252 
 
 
 
 
253 
 
 
 
 
254 
 
 
 
 
255 
 
 
 
 
256 
 
 
 
 
257 
 
 
 
 
258 
 
 
 
 
259 
 
 
 
 
260 
 
 
 
 
261 
 
 
 
 
262 
 
 
 
 
263 
 
 
 
 
264 
 
 
 
 
265 
 
 
 
 
266 
 
 
 
 
267 
 
 
 
 
268 
 
 
 
 
269 
 
 
 
 
270 
 
 
 
 
271 
 
 
 
 
272 
 
 
 
 
273 
 
 
 
 
274 
 
 
 
 
275 
 
 
 
 
276 
 
 
 
 
277 
 
 
 
 
278 
 
 
 
 
279 
 
 
 
 
280 
 
 
 
 
281 
 
 
 
 
282 
 
 
 
 
283 
 
 
 
 
284 
 
 
 
 
285 
 
 
 
 
286 
 
 
 
 
287 
 
 
 
 
288 
 
 
 
 
289 
 
 
 
 
290 
 
 
 
 
291 
 
 
 
 
292 
 
 
 
 
293 
 
 
 
 
294 
 
 
 
 
295 
 
 
 
 
296 
 
 
 
 
297 
 
 
 
 
298 
 
 
 
 
299 
 
 
 
 
300 
 
 
 
 
301 
 
 
 
 
302 
 
 
 
 
303 
 
 
 
 
304 
 
 
 
 
305 
 
 
 
 
306 
 
 
 
 
307 
 
 
 
 
308 
 
 
 
 
309 
 
 
 
 
310 
 
 
 
 
311 
 
 
 
 
312 
 
 
 
 
313 
 
 
 
 
314 
 
 
 
 
315 
 
 
 
 
316 
 
 
 
 
317 
 
Appendix C ? Kingman Fisheries Emergy Analysis Tables
 
 
318 
 
 
 
 
319 
 
 
 
 
320 
 
 
 
 
321 
 
 
 
 
322 
 
 
 
 
323 
 
 
 
 
324 
 
 
 
 
325 
 
 
 
 
326 
 
 
 
 
327 
 
 
 
 
328 
 
 
 
 
329 
 
 
 
 
330 
 
 
 
 
331 
 
 
 
 
332 
 
Appendix D ? Examples of SAS Code 
 
 
333 
 
 
 
 
334 
 
 
 
 
335 
 
 
 
336 
 
Literature Cited  
Able, K.W., D.M. Nemerson, R.R. Light, R.O. Bush. 2000. Initial response of fishes to 
marsh restoration at a former salt hay farm bordering Delaware Bay.  In: Concepts 
and Controversies in Tidal Marsh Ecology. Pgs. 749-773. Eds. M.P. Weinstein 
and D.A. Kreeger. 2000. Kluwer Academic Publishers, Dordrecht, The 
Netherlands. 875pp.  
Abtew, W., J. Obeysekera. 1995. Lysimeter study of evapotranspiration of cattails and 
comparison of three estimation methods. Transactions of the American Society of 
Agricultural Engineers 38: 121-129. 
Athanas, C., J. Cornwell, and C. Stevenson. 1991. Emergent Wetland Establishment 
Under Differing Habitat Conditions In The Anacostia And Potomac River Basins.  
Horn Point Environmental Lab, Center for Environmental And Estuarine Studies, 
UMD, Cambridge, MD and Metropolitan Washington Council of Governments. 
Submitted to District of Columbia, Office of Policy and Planning, Department of 
Public Works. 51pp. 
Bahr, L.M., J.W. Day, J.H. Stone. 1982. Energy cost-accounting of Louisiana fishery 
production.  Estuaries 5(3): 209-215. 
Baker, M.C., and E.M. Baker. 1973. Niche relationships among six species of shorebirds 
on their wintering and breeding ranges. Ecological Monographs 43:193-212. 
Baldwin, A.H., and E.F. DeRico. 1999. The seed bank of a restored tidal freshwater 
marsh in Washington, D.C. Urban Ecosystems 3:5-20. 
Baldwin, A.H. and F.N. Pendleton. 2003. Interactive effects of animal disturbance and 
elevation on vegetation of a tidal freshwater marsh. Estuaries 26(4A):905-915. 
Baldwin, A.H., M.S. Egnotovich, and E. Clarke. 2001. Hydrologic change and vegetation 
of tidal freshwater marshes: field, greenhouse, and seed-bank experiments. 
Wetlands 21(4):519-531. 
Baldwin, A.H. 2004. Restoring complex vegetation in urban settings: The case of tidal 
freshwater marshes. Urban Ecosystems 7:125-137. 
Bates, M. 1961. Man in nature. Prentice Hall, Englewood Cliffs, N.J. 
Batzer, D.P. 1998. Trophic interactions among detritus, benthic midges, and predatory 
fish in a freshwater marsh. Ecology 79: 1688-1689. 
Batzer, D.P. and R. Sharitz. (eds.). 2006. Ecology of Freshwater and Estuarine Wetlands. 
University of California Press, Berkeley, CA. 568pp. 
 
 
337 
 
Batzer, D.P., M. McGee, V.H. Resh, and R.R. Smith. 1993. Characteristics of 
invertebrates consumed by mallards and prey response to wetland flooding 
schedules. Wetlands 13: 41-49. 
Bazely, D.R. and R.L. Jefferies. 1986. Changes in the composition and standing crop of 
 salt-marsh communities in response to the removal of a grazer.  Journal of 
 Ecology 74:693-706. 
Bazely, D.R. and R.L. Jefferies. 1989. Lesser snow geese and the nitrogen economy of a 
grazed salt marsh.  Journal of Ecology 77:24-34. 
Beaulieu, J., G. Gauthier, and L. Rochefort. 1996. The growth response of graminoid 
plants to goose grazing in the high Arctic environment. Journal of Ecology 
84:905-914. 
Belanger, L. and J. Bedard. 1994. Role of ice scouring and goose grubbing in marsh plant 
dynamics. Journal of Ecology 82:437-445. 
Bertness, M.D. and G.H. Leonard. 1997. The role of positive interactions in 
communities: lessons from intertidal habitats. Ecology 78(7):1976-1989. 
Bertness, M.D., B.R. Silliman and R. Jefferies. 2004. Salt marshes under siege. American 
Scientist 92:54-61. 
Biohabitats. 1990. Kenilworth Marsh Restoration. Vol. 1, Historical and Existing 
Conditions. Submitted to Metropolitan Washington Council of Governments, 
Washington, D.C. 
Boesch, D.F. and R.E. Turner. 1984. Dependence of fishery species on salt marshes: the 
role of food and refuge. Estuaries 7: 460-468. 
Bowden, W.B. 1984. Nitrogen and phosphorus in the sediments of a tidal freshwater 
marsh in Massachusetts. Estuaries 7(2):108-118. 
Bowers, K. 1993. Bioengineering and the art of landscaping. Land and Water, 
January/February, 14-17. 
Bowers, K. 1995. Innovations in tidal marsh restoration: the Kenilworth Marsh account. 
Restoration and Management Notes 13:155-161. 
Brewer, J.S., J.M. Levine and M.D. Bertness. 1997. Effects of biomass removal and 
elevation on species richness in a New England salt marsh. Oikos 80:333-341. 
Brinson, M.M., R.R. Christian and L.K. Blum. 1995. Multiple states in the sea-level 
induced transition from terrestrial forest to estuary. Estuaries 18(4):648-659. 
 
 
 
338 
 
Brittingham, K.D., R.H. Hammerschlag. 2006. Final Report (2002-2004): Benthic 
macroinvertebrate communities of reconstructed freshwater tidal wetlands in the 
Anacostia River, Washington, D.C. U.S.G.S. Patuxent Wildlife Research Center. 
Brown, L.D. 1978. Recent vertical crustal movement along the east coast of the United 
States. Tectonophysics, 44:205-231. 
Brown, M.T., and R.A. Herendeen. 1996. Embodied energy analysis and emergy 
analysis: a comparative view. Ecological Economics 19: 219-235. 
Brown, M.T., and T.R. McClanahan. 1996. Emergy analysis perspectives of Thailand and 
Mekong River dam proposals. Ecological Modeling 91: 105-130. 
Bull, J. and J. Farrand. 1994. National Audubon Society Field Guide to North American 
Birds: Eastern Region. Alfred A. Knopf, New York. 796pp. 
Burton, N.H.K., A.J. Musgrove, and M.M. Rehfisch. 2004. Tidal variation in numbers of 
waterbirds: how frequently should birds be counted to detect change and do low 
tide counts provide a realistic average? Bird Study 51:48-57. 
Butler, D.R. 1995. Zoogeomorphology: Animals as geomorphic agents. Cambridge 
University Press, New York. 
Byers, T.F. 2005. 2004 District of Columbia Shoreline Angler Survey.  Report of the 
District of Columbia, Department of Health, Environmental Health 
Administration, Fisheries and Wildlife Management Division. Washington, D.C.  
Cairns, J. Jr. 1988. Restoration ecology: The new frontier. Pgs. 1-11. In: J. Cairns, Jr., 
ed., Rehabilitating Damaged Ecosystems, Vol. I. CRC Press, Boca Raton, FL. 
Cahill Jr., J.F., J.P. Castelli, and B.B. Casper. 2001. The herbivory uncertainty principle: 
visiting plants can alter herbivory. Ecology 82(2):307-312.  
Campeau, S., H.R. Murkin, and R.D. Titman. 1994. Relative importance of algae and 
emergent polant litter to freshwater marsh invertebrates. Canadian Journal of 
Fisheries and Aquatic Sciences 51:681-692.    
Capers, R.S. 2003. Macrophyte colonization in a freshwater tidal wetland. (Lyme, CT, 
USA). Aquatic Botany 77:325-338. 
Cargill, S.M. and R.L. Jefferies. 1984. The effects of grazing by lesser Snow  Geese on 
the Vegetation of a sub-artic salt marsh. Journal of Applied Ecology 21:669-686. 
Center for Watershed Protection. 1997. District of Columbia Wetland Conservation Plan. 
Prepared for District of Columbia Environmental Regulation Administration. 
Grant# 95G-97-WRMD03. Washington, D.C. 
 
 
339 
 
Chapman, D.W. 1978. Production in fish populations. In: Ecology of Freshwater Fish 
Production. Pgs. 5-25. S.D. Gerking, Ed. 1978. John Wiley & Sons. New York. 
520p. 
Childers, D.L., H.N. McKellar, Jr., R. Dame, F. Sklar, and E. Blood. 2000. Twenty more 
years of marsh and estuarine flux studies: Revisiting Nixon (1980). In M.P. 
Weinstein and D. A. Kreeger, eds. International Symposium: Concepts and 
Controversies in Tidal Marsh Ecology. Kluwer Academic Publishers, Dordrecht, 
The Netherlands, pp. 389-241. 
Cogdon, J.D. and J.W. Gibbons. 1989. Biomass of freshwater turtles: a geographic 
comparison. Amer. Midland Naturalist 115:165-173. 
Connell, J. H. and R.O. Slayter. 1977. Mechanisms of succession in natural communities 
and their role in community stability and organization. The  American Naturalist 
111:1114-1119. 
Connell, J.H. and W.P. Sousa. 1983. On the evidence needed to judge ecological stability 
or persistence. The American Naturalist 121(6):789-824. 
Coues, E., and D. W. Prentiss. 1883. Avifauna Columbiana: Being a list of birds 
ascertained to inhabit the District of Columbia, with the times of arrival and 
departure of such as are non-residents, and brief notices of habits, etc. 
Washington: Government Printing Office. 
Cronk, J.K. and M.S. Fennessy. 2001. Wetland Plants, Biology and Ecology. Lewis 
Publishers, Boca Raton, FL. 
Crowder, L.B., D.D. Squires, and J.A. Rice. 1997. Nonadditive effects of terrestrial and 
aquatic predators on juvenile estuarine fish. Ecology 78:11796-1804. 
Cummins, J.D., and D.B. Rockland. 1987. Economic value and demographics of an urban 
fishery: Washington, D.C. Presented at the 117th Annual Meeting of the 
American fisheries Society, Winston Salem, North Carolina, September 13-17, 
1987. 
Currin, C.A., S.C. Wainright, K.W. Able, M.P. Weinstein, and C.M. Fuller. 2003. 
Determination of food web support and trophic position of the mummichog, 
Fundulus heteroclitus, in New Jersey smooth cordgrass (Spartina alterniflora), 
common reed (Phragmites australis) and restored salt marshes. Estuaries 26: 485-
510. 
Deegan, L.A. 1993. Nutrient and energy transport between estuaries and coastal marine 
ecosystems by fish migration.  Canadian Journal of Fish and Aquatic Sciences 50: 
74-79. 
 
 
 
340 
 
Diaz, R.J. 1977. The effects of pollution on benthic communities of the tidal James River, 
Virginia. PhD Dissertation. University of Virginia, Charlottesville. 149pp. 
Diaz, R.J. 1992. Ecosystem assessment using estuarine and marine benthic community 
structure. Pgs. 67-85. In: G.A. Burton (ed.), Sediment Toxicity Assessment. 
Lewis Publishers, Boca Raton, FL. 
Diaz, R.J. 1994. Response of tidal freshwater macrobenthos to sediment disturbance. 
Hydrobiologica 278:201-212. 
Diaz, R.J., and J.L. Boesch. 1977. Habitat development field investigations, Windmill 
Point Marsh development site, James River, Virginia, Appendix C:Environmental 
impacts of marsh development with dredged material: acute impacts on the 
macrobenthic community. Vicksburg, MS: US Army Waterways Experiment 
Station. USAWES Technical Report D-77-23. 
Doumlele, D.G. 1981. Primary production and seasonal aspects of emergent plants in a 
tidal freshwater marsh. Estuaries 4(2):139-142. 
DNR(www.dnr.state.md.us/Bay/monitoring/mon_mngmt_actions/chapter5.html.map6. 
Maryland Department of Natural Resources phytoplankton annual production 
rates for the tidal freshwater Potomac River. 
Elton, C. 1927. Animal Ecology. Sidgwick and Jackson, London, England. 
Evans, A. 1987. Relative availablility of the prey of wading birds by day and night. 
Marine Ecology Progress Series 37:103-107. 
Everett, R.A. and G.M Ruiz. 1993. Coarse woody debris as a refuge from predation in 
aquatic communities: and experimental test. Oecologia 93:475-486. 
Evers, D.E., C.E.Sasser, J.G. Gosselink, D.E. Fuller, J.M. Visser. 1998. The impact of 
vertebrate herbivores on wetland vegetation in Atchafalaya Bay, Louisiana. 
Estuaries 21(1):1-13. 
Ewel, J.J. 1987. Restoration is the ultimate test of ecological theory. Pp. 31-34. 
In:Restoration ecology. W.R. Jordan III, M.E. Gilpin, J.D. Aber. (Eds.). 
Cambridge University Press, Cambridge, UK. 
Farina, A. 2006. Principles and Methods in Landscape Ecology: Toward a Science of 
Landscape. Landscape Series Volume 3. Springer, Dordrecht, The Netherlands. 
412pp. 
Findlay, S., K. Schoeberl and B. Wagner. 1989. Abundance, composition, and dynamics 
of the invertebrate fauna of a tidal freshwater wetland. Journal of the North 
American Benthological Society 8:140-148. 
 
 
341 
 
Gagliardo, A., F. Odeti, and P. Ioale. 2001. Relavance of visual cues for orientation at 
familiar sites by homing pigeown: an experiment in a circular arena. Proc. R. Soc. 
Lond. 268:2065-2070. 
Gallagher, J.L. and F.C. Daiber. 1974. Primary production of edaphic communities in a 
Delaware salt marsh. Limnol. Oceanogr. 19:390-395. 
Garbish, E.W., Jr., and L.B. Coleman. 1978. Tidal freshwater marsh establishment in 
upper Chesapeake Bay: Pontederia cordata and Peltandra virginica. Pgs. 63-78 In: 
Good, R.E., D.F. Whigham, and R.L. Simpson (eds.) Freshwater Wetlands: 
Ecological Processes and Management Potential. Academic Press, New York, 
NY.  
Giroux, J.F., and J. Bedard. 1987. The effects of grazing by greater snow geese on the 
vegetation of tidal marshes in the St. Lawrence estuary. Journal of Applied 
Ecology 24:773-788. 
Good, R.E., D.F. Whigham, and R.L. Simpson, Eds. 1978. Freshwater Wetlands: 
Ecological Processes and Management Potential. Academic Press, New York. Pp. 
378. 
Gough, L. and J.B. Grace. 1998. Herbivore effects on plant species density at varying 
productivity levels. Ecology 79(5):1586-1594. 
Grace, J.B. and M.A. Ford. 1996. The potential impact of herbivores on the susceptibility 
of the marsh plant Sagittaria lancifolia to saltwater intrusion in coastal wetlands.  
Estuaries 19(1):13-20. 
Greeson, P.E., J.R. Clark, and J.E. Clark (eds.). 1979. Wetland Function and Values: The 
State of Our Understanding. American Water Resources Association, Minn., MN. 
Grime, J.P. 1973. Competitive exclusion in herbaceous vegetation. Nature 242:344-347. 
Guerrero, V.C. and M. Hille. 1994. The ecology of Kenilworth Marsh: inventory of 
vegetation and restoration efforts. Department of Biological And Environmental 
Science, University of the District of Columbia. Washington, DC. Submitted to 
the Kenilworth Marsh Monitoring Committee, National Park Service, National 
Capitol Parks-East.  
Hammerschlag, D. 2002. Personal communication on data from Kingman Marsh surface 
elevation tables. 
Hammerschalg, D., C.C. Krafft, K. Phyllaier, and M.M. Paul. 2001. First Year Annual 
Report (2000) for the Kingman Marsh Vegetation Monitoring Project. USGS 
PWRC Report. Laurel, Maryland. 
 
 
 
342 
 
Hammerschalg, D., K.P. Neff, C.C. Krafft, M.M. Paul, and K.D. Brittingham. 2002. 
Second Year Annual Report (2001) for the Kingman Marsh Vegetation 
Monitoring Project. USGS PWRC Report. Laurel, Maryland. 
Hammerschlag, D., A.H. Baldwin, C.C. Krafft, K.P. Neff, M.M. Paul, K.D. Brittingham, 
K. Rusello, J.S. Hatfield. 2006. Final Report: Five Years of Monitoring 
Resconstructed Freshwater Tidal Wetlands in the Urban Anacostia River (2000-
2004). U.S.G.S. Patuxent Wildlife Research Center and University of Maryland 
Department of Biological Resources Engineering. Laurel, Maryland. 
Haramis, G.M. and G.D. Kearns. 2007. Herbivory by resident geese: the loss and 
recovery of wild rice along the tidal Patuxent River. Journal of Wildlife 
Management 71(3):788-794. 
Harris, H.B. 2002. The impact of resident Canada geese on wetland restoration, and an 
evaluation of available mitigation techniques. Prepared for the Anacostia 
Watershed Society. Sustainable Development and Conservation Biology Graduate 
Program. University of Maryland, College Park, Maryland. 
Hayman, P., J. Marchant, and T. Prater. 1986. Shorebirds. Boston: Houghton Mifflin 
Company. 360pp. 
Havens, K.J., L.M. Varnell and J.G. Bradshaw. 1995. An assessment of ecological 
conditions in a constructed tidal marsh and two natural reference tidal marshes in 
coastal Virginia. Ecological Engineering 4:117-141. 
Hik, D.S., R.L.Jefferies, and A.R.E. Sinclair. 1992. Foraging by geese, isostatic uplift and 
asymmetry in the development of salt-marsh plant communities. Journal of 
Ecology 80:395-406. 
Holdahl, S.R., and N.L. Morrison. 1974. Regional investigations of vertical crustal in the 
U.S., using precise relevelings and mareograph data. Tectonophysics, 23:373-390. 
Horn, M.H., K.L.M. Martin, and M.A. Chotkowski. 1999. Intertidal Fishes: Life In Two 
Worlds. Academic Press. San Diego. 399pp. 
Hosokawa, Y. 1997. Restoration of coastal tidal flat in Japan. Port and Harbour Research 
Institute, Ministry of Transport, Nagase, Yokosuka, Japan. 
Houde, E.D. and E.S. Rutherford. 1993. Recent trends in estuarine fisheries: predictions 
of fish production and yield. Estuaries 16: 161-176. 
Hussey, B.H. and W.E. Odum. 1992. Evapotranspiration in tidal marshes. Estuaries 
15(1):59-67. 
Hutchinson, L.D. 1977. The Anacostia Story: 1608-1930. Smithsonian Institution Press, 
Washington, D.C. 158pp. 
 
 
343 
 
Imamura, H. 2000. Current status of artifical tidal flats in Japan. From the Wetland 
Restoration in Coastal Ennvioronments Workshop: construction techniques, 
ecology of intertidal habitat, and function assessment procedures.  December 5-8, 
2000, Patuxent Wildlife Research Center, Laurel, Maryland. 
Iverson, J.B, 1982. Biomass in turtle populations: a neglected subject. Oecologica 55:69-
75. 
Jones, C.G., J.H. Lawton and M. Shachak. 1994. Organisms as ecosystem engineers. 
Oikos 69:373-386. 
Kangas, P.C. 1990. An energy theory of landscape for classifying wetlands. In:. 
 A.E.Lugo, M.Brinson and S.Brown (Eds), Ecosystems of the World 15: 
 Forested Wetlands. Elsevier, Amsterdam, The Netherlands pp. 15-24.  
Kangas, P.C. 2004a. Ecological Engineering, Principles and Practice. Lewis Publishers, 
Boca Raton, FL.  
Kangas, P.C. 2004b. The role of invasive species in a complex ecosystem: mute swans in 
the Chesapeake Bay. Mute Swan Habitat Symposium Proceedings. 
Kangas, P., P. May and S. Kassner. 2004. A case study in Anacostia River Restoration.  
Proceedings of the 30th Annual Conference on Ecosystems Restoration and 
Creation, Hillsborough Community College, Tampa, FL. 
Kearney, M.S., and J.C. Stevenson. 1991. Island land loss and marsh vertical accretion 
rate evidence for historical sea-level changes in Chesapeake Bay. Journal of 
Coastal Research 7(2): 403-415. 
Keddy, P.A.2000. Wetland Ecology: Principles and Conservation. Cambridge University 
Press. 
Keller, C.R. and K.P. Burnham. 1982. Riparian fencing, grazing and trout habitat 
preference on Summit Creek, Idaho. North American Journal of Fisheries 
Management 2:53-59. 
Kerbes, R.H., P.M. Kotanen, and R.L.  Jefferies. 1990. destruction of wetland habitat by 
lesser snow geese: a keystone species on the west coast of Hudson Bay. Journal of 
Applied Ecology 27: 242-258. 
Khan, H. and G.S.Brush. 1994. Nutrient and metal accumulation in a freshwater tidal 
marsh. Estuaries 17(2):345-360. 
King, A.J., A.I. Robertson, and M.R. Healey. 1997. Experimental manipulations of the 
biomass of introduced carp (Cyprinus carpio) in billabongs. Impacts on water-
column properties. Marine and Freshwater Research 48:435-443. 
 
 
344 
 
Knowlton, N. 1992. Thresholds and multiple stabe states in coral reef community 
dynamics. American Zoologist 32:674-682. 
Kotanen, P.M., and R.L. Jefferies. 1997. Long-term destruction of sub-arctic wetland 
vegetation by lesser snow geese. Ecoscience 4:179-182. 
Lane, J.J. and K.C. Jensen. 1999. Moist soil impoundments for wetland wildlife. U.S. 
Army Corps of Engineers, Ecosystem Management and Restoration Research 
Program. Technical Report EMRRP-99-March 1999. 
Leck, M.A. and R.L.Simpson. 1987. Seed bank of a freshwater tidal wetland: turnover 
and relationship to vegetation change. American Journal of Botany 74(3):360-
370. 
Leck, M.A. and R.L. Simpson. 1994. Tidal freshwater wetland zonation: seed and 
seedling dynamics. Aquatic Botany 47:61-75. 
Leopold, A. 1933. Game management. New York, Scribners. 
Lindeman, R.L. 1942. The trophic dynamic aspect of ecology. Ecology 23:399-418. 
Lodge, D.M. 1991. Herbivory on freshwater macrophytes. Aquatic Botany 41:195-224. 
Lockwood, J.L. 1997. An alternative to succession ? assembly rules offer guide to 
restoration efforts. Restoration and Management Notes 15(1):45-50. 
Lopez, G.R. 1988. Comparative ecology of the macrofauna of freshwater and marine 
muds. Limnological Oceanography 33:946-962. 
Lotka, A.J. 1922. A contribution to the energetics of evolution. Proceedings of the 
National Academy of Sciences, U.S., 8:147-155. 
Louda, S.M. 1989. Differential predation pressure: a general mechanism for structureing 
plant communities along complex environmental gradients?  Trends in Ecology 
and Evolution 4(6):158-159. 
Lougheed, V.L., B. Crosbie, and P. Chow-Frazer. 1998. Predictions on the effect of 
common carp (Cyprinus carpio) exclusion on water quality, zooplankton, and 
submergent macrophytes in a Great Lakes wetland. Canadian Journal of Fisheries 
and Aquatic Sciences 55:1189-1197. 
MacArthur, R.H. 1955. Fluctuations of animal populations, and a measure of community 
stability. Ecology 36:533-536. 
MacArthur, R.H. and E.R. Pianka. 1966. On optimal use of patchy environment. 
American Naturalist 100:603-610. 
 
 
345 
 
 
Mandel,R., S. Kim, A. Nagel, and C.L. Schultz. 2007. Anacostia sediment models: Phase 
3 Anacostia HSPF watershed model and Version 3 TAM/WASP water clarity 
model. Interstate Commission on the Potomac River Basin, Rockville, MD. 
Martin, J.F. 2002. Emergy valuation of diversions of river water to marshes in the 
Mississippi River Delta. Ecological Engineering 18: 265-286. 
May, P.I. 2000. How (and why) to build a tidal freshwater mudflat mesocosm. Pp. 45-58. 
In: Proceedings of the Annual Ecosystems Restoration and Creation Conference, 
P.J. Cannizzaro (ed.). Hillsborough Community College, Plant City, FL. 
May, R.M. 1973. Stability and complexity in model ecosystems. Princeton University 
Press, Princeton. 
May, R.M. 1977. Thresholds and breakpoints in ecosystems with a multiplicity of stable 
states. Nature 269:471-477. 
McIvor, C., and W.E. Odum. 1988. Food, predation risk, and microhabitat selection in a 
marsh fish assemblage. Ecology 69(5):1341-1351. 
McKee, K.L. and A.H. Baldwin. 1999. Distrubance regimes in North American wetlands. 
Pgs. 331-363 In: Ecosystems of Disturbed Ground, L.W. Walker (ed.). From: 
Ecosystems of the World 16. Elseviver, Amsterdam, The Netherlands. 
McVay, M.E., P.E. Heilman, D.M. Greer, S.E. Brauen, and A.S. Baker. 1980. Tidal 
freshwater marsh establishment on dredge spoils in the Columbia River Estuary. 
Journal of Environmental Quality 9(3):488-493. 
Mitsch, W.J. 2000. Self-design applied to coastal restoration. Pgs. 554-564 In: Concepts 
and Controversies in Tidal Marsh Ecology. M.P. Weinstein and D.A. Greeger 
(eds.). Kluwer Academic, Dordrecht, The Netherlands. 
Mitsch, W.J., and J.G. Gosselink. 2000. Wetlands. 3rd Edition. John Wiley & Sons, Inc. 
New York. 
Mitsch,W.J., and S.E. Jorgensen. 1989. Ecological Engineering: An Introduction to 
Ecotechnology. John Wiley & Sons, New York. 
Mitsch,W.J., and S.E. Jorgensen. 2004. Ecological Engineering and Ecosystem 
Restoration. John Wiley & Sons, New York. 
Mitsch, W.J. and R.F. Wilson. 1996. Improving the success of wetland creation and 
restoration with know-how, time, and self-design. Ecological Applications 6:77-
83. 
 
 
346 
 
Moll, D. and E.O. Moll. 2004. The Ecology, Exploitation, and Conservation of River 
Turtles. Oxford University Press, New York. 383pp. 
Neff, K.P. 2002. Plant Colonization and Vegetation Change in a Restored Tidal 
Freshwater Wetland in Washington, D.C. M.S. Thesis. University of Maryland, 
College Park, MD. 
Nixon, S.W. 1988. Physical energy inputs and the comparative ecology of lake and 
marine ecosystems. Limnology and Oceanography 33: 1005-1025. 
Nordstrom, K.F., and C.T. Roman (eds.). 1996. Estuarine Shores: Evolution, 
Environments and Human Alterations.  John Wiley & Sons, New York. 
Noy-Meir, I. 1975. Stability of grazing systems: an application of predator-prey graphs. 
Journal of Ecology 63:459-481. 
Odum, E.P. 1971. Fundamentals of Ecology. 3rd Edition. W.B. Saunders, Philadelphia, 
PA. 574pp. 
Odum, E.P. 1980. The status of three ecosystem-level hypothesis regarding salt marsh 
estuaries: tidal subsidy, outwelling, and detritus-based food chains. In: V.S. 
Kennedy ed., Estuarine Perspectives. Academic Press, New York, pp. 485-495. 
Odum, E.P. 2000. Tidal marshes as outwelling/pulsing systems. Pgs. 3-8. In: M.P. 
Weinstein and D.A. Kreeger, eds., Concepts and Controversies in Tidal Marsh 
Ecology. Kluwer Academic, Dordrecht, The Netherlands. 
Odum, E.P., J.T. Finn, and E.H. Franz. 1979. Perturbation theory and the subsidy-stress 
gradient.  Bioscience 29(6):349-352. 
Odum, H.T. 1962. Man in the ecosystem. Pgs. 57-75. In: Proceedings of the Lockwood 
Conference on the Suburban Forest and Ecology. Bulletin 652 Connecticut 
Agricultural Station, Storrs, CT. 
Odum, H.T. 1971. Environment, Power, and Society. John Wiley, NY, 286pp. 
Odum, H.T. 1983. Systems Ecology. John Wiley and Sons, New York, N.Y., 644pp. 
Odum, H.T. 1984. Energy analysis evaluation of coastal alternatives. Water Science and 
Technology 16:717-734. 
Odum, H.T. 1988. Self organization, transformity, and information. Science 242:1132-
1139. 
Odum, H.T. 1989. Ecological engineering and self-organization. In: Ecological 
Engineering. W.J. Mitsch and S.E. Jorgensen (eds.). John Wiley & Sons, New 
York. 
 
 
347 
 
Odum, H.T. 1996. Environmental Accounting: Emergy and Environmental Decision 
Making. John Wiley & Sons, New York, New York. 370pp. 
Odum, H.T. 2000. Emergy ? emdollar evaluation and the Everglades. North American 
Lake Management. Miami. 28p. 
Odum, H.T. 2001. Emergy Accounting. Chapter in Unveiling Wealth ? On Money, 
Quality of Life and Sustainability, ed. By P. Bartelmus. Kluwer Academic Publ. 
23pp. 
Odum, H.T., and E.C. Odum. 1999. The Prosporus Way Down, Principles and Policies.  
Draft given by H.T. to P.I.M. University Press of Colorado. 375pp. 
Odum, H.T. and E.C. Odum. 2000. Modeling for all Scales: An Introduction to System 
Simulation. Academic Press, New York. 
Odum, H.T., S.J. Doherty, F.N. Scatena, and P.A. Kharecha. 2000. Emergy evaluation of 
reforestation alternatives in Puerto Rico. Forest Science 46 (4):521-530. 
Odum, H.T., W.L. Siler, R.J. Beyers, and N. Armstrong.  1963.  Experiments with 
engineering of marine ecosystems. Publ. Inst. Marine Sci., University of Texas 
9:374-403. 
Odum, W.E. 1988. Comparative ecology of tidal freshwater and salt marshes. American 
Review of Ecological Systems 19:147-176. 
Odum, W.E. and M.A. Heywood. 1978. Decomposition of intertidal freshwater marsh 
plants. Pgs. 89-97. In: Freshwater Wetlands: Ecological Processes and 
Management Potential. Eds. R.E. Good, D.F. Whigham, and R.L. Simpson. 1978. 
Academic Press, New York. 378 pp. 
Odum, W.E., E.P. Odum, and H.T. Odum. 1995. Nature?s pulsing paradigm. Estuaries. 
18:547-555. 
Odum, W.E., T.J. Smith III, J.K. Hoover, and C.C. McIvor. 1984. The Ecology of Tidal 
Freshwater Marshes of the United States East Coast: A Community Profile. 
University of Virginia, Charlottesville, VA. For: U.S. Fish and Wildlife Service 
FWS/OBS-83/17. 
Officer, C.B., D.R. Lynch, G.H. Setlock, and G.R. Helz. 1984. Recent sedimentation 
rates in Chesapeake Bay. In: V.S. Kennedy ed., The Estuary as a Filter. 
Academic Press, New York, pp. 131-157. 
Opperman, J.J. and A.M. Merenlender. 2000. Deer herbivory as an ecological constraint 
to restoration of degraded riparian corridors. Restoration Ecology 8(1):41-47. 
 
 
 
348 
 
Otto, S., P.M. Groffman, S.E.G. Findlay, and A.E. Arreola. 1999. Invasive plant species 
and microbial processes in a tidal freshwater marsh. Journal of Environmental 
Quality 28:1252-1257. 
Paine, R.T. 1966. Food web complexity and species diversity. American Naturalist 
100:65-75. 
Paine, R.T. 1969. The pisaster-tegula interaction: prey patches, predator food preference 
and intertidal community structure. Ecology 50:950-961. 
Paine, R.T. 1974. Intertidal community structure: experimental studies on the relationship 
between a dominant competitor and its principal predator. Oecologia 15:93-120. 
Pasternack, G.B. and G.S. Brush. 1998. Sedimentation cycles in a river mouth tidal 
freshwater marsh. Estuaries 21(3):407-415. 
Pasternack, G.B., W.G. Hilgartner, and G.S. Brush. 2000. Biogeomorphology of an upper 
Chesapeake Bay river mouth tidal freshwater marsh. Wetlands 20(3):520-537. 
Pasternack, G.B. and G.S. Brush. 2001. Seasonal variations in sedimentation and organic 
content in five plant associations on a Chesapeake Bay tidal freshwater delta. 
Estuarine, Coastal and Shelf Science 53:93-106. 
Pennak, R.W. 1978. Freshwater Invertebrates of the United States. New York: Wiley-
Interscience Publications. 2nd ed. 
Perry, M.C. and A.S. Deller. 1996. Review of factors affecting the distribution and 
abundance of waterfowl in shallow-water habitats of Chesapeake Bay. Estuaries 
19 (2A):272-278. 
Perry, J.E. and C.H. Hershner. 1999. Temporal changes in the vegetation patterns in a 
tidal freshwater marsh. Wetlands 19(1):90-99. 
Peterson, C.H. 1981. The ecological role of mud flats in estuarine systems. Pgs. 184-192. 
In: R.C. Carey, P.S. Markovits, and J.B. Kirkwood, eds. Proceedings of the U.S. 
Fish and Wildlife Service Workshop on Coastal Ecosystems of the Southeastern 
United States. U.S. Fish and Wildlife Service, Office of Biological Services, 
Washington, D.C. 
Peterson, C.H. 1984. Does a rigorous criterion for environmental identity preclude the 
existence of multiple stable points?  The American Naturalist 124(1):127-133. 
Peterson, G.D. 2000. Political ecology and ecological resilience: an integration of human 
and ecological dynamics. Ecological Economics 35:323-336. 
Peterson, G.D. 2002. Contagious disturbance, ecological memory, and the emergence of 
landscape pattern. Ecosystems 5:329-338. 
 
 
349 
 
Peterson, G.D., S.R. Carpenter, and W.A. Brock. 2003. Uncertainty and the management 
of multistate ecosystems: an apparently rational route to collapse. Ecology 
84(6):1403-1411. 
Peterson, J.E. and A.H. Baldwin. 2004. Seedling emergence from seed banks of tidal 
freshwater wetlands: response to inundation and sedimentation. Aquatic Botany 
78:243-254. 
Pethick, J.S. 1996. The geomorphology of mudflats. In: Estuarine Shores: Evolution, 
Environments and Human Alterations. K.F.Nordstrom and C.T. Roman eds. John 
Wiley and Sons, New York. 
Petraitis, P.S., and R.E. Latham. 1999. The importance of scale in testing the origins of 
alternative community states. Ecology 80(2):429-442. 
Pimm, S.L., and J.H. Lawton. 1977. Number of trophic levels in ecological communities. 
Nature 268:329-331. 
Platts, W.S. and F.J. Wagstaff. 1984. Fencing to control livestock grazing on riparian 
habitats along streams: is it a viable alternative? North American Journal of 
Fisheries Management 4:266-272. 
Pomeroy, R.L., and R.G. Wiegert. 1981. The Ecology of a Salt Marsh. Ecological Studies 
38. Springer-Verlag, New York. 271p. 
Potomac River Fisheries Commission. 2007. Personal phone and email communication to 
obtain unpublished statistics on Potomac River fish harvests.  
Power, M.E. 1984. Depth distributions of armored catfish: predator induced resource 
avoidance? Ecology 65:523-528. 
Power, M.E. 1988. Grazer control of algae in an ozark mountain stream: effects of short 
term exclusion. Ecology 69(6):1894-1898. 
Power, M.E. 1990. Resource enhancement by indirect effects of grazers: armored catfish, 
algae, and sediment. Ecology 71(3):897-904. 
Power, M.E. 1992. Top-down and bottom-up forces in food webs: do plants have 
primacy? Ecology 73(3):733-746. 
Price, J.E. and P.J. Morin, 2004. Colonization history determines alternate community 
states in a food web of intraguild predators. Ecology 85(4):1017-1028. 
Priest III, W.I., C.W. Frye, J. Nestlerode, and R.J. Byrne. 1996. Beneficial Uses of 
Dredges Material from the Waterway on the Coast of Virginia (WCV). Final 
Report of the Virginia Marine Resources Commission, Virginia Institute of 
Marine Science to The Department of Environmental Quality, Virginia Coastal 
 
 
350 
 
Resource Management Program and NOAA.60pp. 
Quammen, M.L. 1984. Predation by shorebirds, fish and crabs on invertebrates in 
intertidal mudflats: an experimental test. Ecology 65(2):529-537. 
Rachich, J. and R. Reader. 1999. Interactive effects of Herbivory and competition on blue 
vervain (Verbena hastata L.: Verbenaceae). Wetlands 19(1):1156-161. 
Rader, R.B. and C.J. Richardson. 1994. Response of macroinvertebrates and small fish to 
nutrient enrichment in the northern Everglades. Wetlands 14:134-146. 
Ragotzkie, R.A. 1959. Plankton productivity in estuarine waters of Georgia. Publ. Inst. 
Mar. Sci. Univ. Texas. 6:146-158. 
Recher, H.F. 1966. Some aspects of the ecology of migrant shorebirds. Ecology 
47(3):393-407. 
Reise, K. ed. 1985. Tidal Flat Ecology. Springer-Verlag, Berlin. 191pp. 
Reise, K. ed. 2001. Ecological Comparisons of Sedimentary Shores. Springer-Verlag, 
Berlin. 384pp. 
Rietkerk, M., and J. van de Koppel. 1997. Alternate stable states and threshold effects in 
semi-arid grazing systems. Oikos 79:69-76. 
Rohde, F.C., R.G. Arndt, D.G. Lindquist, and J.P. Parnell. 1994. Freshwater Fishes of the 
Carolinas, Virginia, Maryland, & Delaware. The University of North Carolina 
Press, Chapel Hill, NC. 222p. 
Rozas, L.P. and W.E.Odum. 1987. Fish and macrocrustacean use of submerged plant 
beds in tidal freshwater marsh creeks. Marine Ecology Progress Series 38:101-
108. 
Rozas, L.P., C.C. McIvor, and W.E.Odum. 1988. Intertidal rivulets and creekbanks: 
corridors between tidal creeks and marshes. Marine Ecology Progress Series 
47:303-307. 
SAS. 2006. Version 9.2 for windows. SAS Institute, Cary, North Carolina. 
Sacco, J.N., E.D. Seneca, and T.R. Wentworth. 1994. Infaunal community development 
of artifically established salt marshes in North Carolina. Estuaries 17(2):489-500. 
Scheffer, M. and E. Jeppesen. 1998. Alternative stable states. P.p. 397-406. In: The 
Structuring Role of Submerged Macrophytes in Lakes. E.Jeppesen, M.   
Sondergaard, and K. Christoffersen (eds.). Springer, New York, NY. 
 
 
 
351 
 
Schneider, D. 1978. Equilization of prey numbers by migratiory shorebirds. Nature 
271:353-354. 
Serodes, J.B. and J.P. Troude. 1984. Sedimentation cycle of a freshwater tidal flat in the 
St. Lawrence Estuary. Estuaries 7(2): 119-127. 
Seybold, C.A., W. Mersie, J. Huang, and C. McNamee. 2002. Soil redox, pH, 
temperature, and water-table patterns of a freshwater tidal wetland. Wetlands 
22(1):149-158. 
Simpson, R.L., R.E. Good, M.A. Leck, and D.E. Whigham. 1983. The ecology of 
freshwater tidal wetlands. Bioscience 33(4):255-259. 
Simpson, R.L., R.E. Good, R.A.Walker, and B.R.Frasco. 1983a. The role of Delaware 
River tidal freshwater wetlands in the retention of nutrients and heavy metals. 
Journal of Environmental Quality 12(1):41-48.  
Smith, K.J., G.L. Taghon, and K.W. Able. 2000. Trophic linkages in marshes: 
Ontogenetic changes in diet for young-of-the-year mummichog, Fundulus 
heteroclitus. In M.P. Weinstein and D. A. Kreeger, eds. International Symposium: 
Concepts and Controversies in Tidal Marsh Ecology. Kluwer Academic 
Publishers, Dordrecht, The Netherlands, pp. 221-236. 
Smith, T.J. and W.E. Odum. 1981. The effects of grazing by snow geese on coastal 
marshes. Ecology 62:98-106. 
Stevenson, J.C., L.G. Ward, and M.S. Kearney. 1988. Sediment transport and trapping in 
marsh systems: implications of tidal flux studies. Marine Geology 80: 37-59. 
Stevenson, C., J. Cornwell, L. Staver. 1995. Anacostia Wetlands Demonstration Project 
Phase II. Horn Point Environmental Laboratories, Center for Environmental and  
Estuarine Studies, University of Maryland. Draft Final Report to: District of 
Columbia department of Public Works, Office of Policy and Planning. 57pp. 
Streever, B. 2000. Dredged material marshes: summary of three projects. Wetlands 
Research Bulletin 2(1):1-4. 
Stribling, J.M., and J.C. Cornwell. 1997.  Identification of important primary producers in 
a Chesapeake Bay tidal creek system using stable isotopes of carbon and sulfur. 
Estuaries 20:77-85. 
Sutherland, J.P. 1974. Multiple stable points in natural communities. The American 
Naturalist 108(964):859-873. 
Sutherland, J.P. 1990. Perturbations, resistance, and alternative ciews of the existence of 
multiple stable points in nature. The American Naturalist 136(2):270-275. 
 
 
352 
 
Syphax, S.W. 2001. Personal Communication on observations of goose herbivory at 
Kingman Marsh. 
Syphax, S.W. and R.S.Hammerschlag. 1995. The reconstruction of Kenilworth Marsh, 
The last tidal marsh in Washington, D.C. Park Science. Publication of the 
National Park Service-U.S. Department of the Interior 15 (1) 15-19. 
Tanner, C.D., J.R. Cordell, J. Rubey, and L.M. Tear. 2002. Restoration of freshwater 
intertidal habitat functions at Spencer Island, Everett, Washington. Restoration 
Ecology 10(3):564-576. 
Tansley, A.G. 1935. The use and abuse of vegetative concepts and terms. Ecology 
16(3):284-307. 
Taylor K.L. and J.B.Grace. 1995. The effects of vertebrate herbivory on plant community 
structure in the coastal marshes of the Pearl River, Louisiana, USA. Wetlands 
15(1):68-73. 
Teal, J.M. 1976. Predation by the salt marsh killifish Fundulus heteroclitus in relation to 
prey size and habitat structure: consequences for prey distribution and abundance. 
Journal of Experimental Marine Biology and Ecology 23:255-266. 
Teal, J.M., and B.L. Howes. 2000. Salt marsh values: Retrospection from the end of the 
century. In M.P. Weinstein and D. A. Kreeger, eds. International Symposium: 
Concepts and Controversies in Tidal Marsh Ecology. Kluwer Academic 
Publishers, Dordrecht, The Netherlands, pp. 9-19. 
Turner, R.E. 1977. Intertidal vegetation and commercial yields of penaeid shrimp. 
American Fisheries Society Transactions 106:411-416. 
U.S. Army Corps of Engineers. 1891. Map of Anacostia River in the District of Columbia 
and Maryland. U.S. Engineer Office, Washington, D.C. National Archives. 
U.S. Army Corps of Engineers. 1913. Map of Anacostia River, D.C. Reclamation and 
Development of River and Flats, Anacostia Bridge to District Line. U.S. Engineer 
Office, Washington, D.C. National Archives. 
U.S. Army Corp of Engineers, Baltimore District. 1992. Environmental Assessment, 
Maintenance Dredging Federal Navigation Project Anacostia River Basin,  District 
of Columbia and Maryland. Washington, D.C. 
U.S. Army Corps of Engineers, Baltimore District. 1993. Innovative Dredged Material 
Disposal Strategies for the Anacostia River. Prepared for the District of Columbia 
under the authority of Section 22 of Public Law 93-251. 
U.S. Army Corps of Engineers. 1995. Unpublished surface elevation data for study 
mudflat at Kenilworth Mass Fill 3. 
 
 
353 
 
U.S. Army Corps of Engineers. 2004. (personal communication, email attachment of 
Kingman Restoration expenses) 
U.S. Army Engineer Waterways Experiment Station. 1978. Wetland habitat development 
with dredged material: engineering and plant propagation. U.S. Army Eng. 
Waterways Exp. Station. Tech. Rep. DD-78-16. 158pp. 
U.S. Geological Survey. 2004. Corners of Kingman exclosure plot mudflat surface 
elevations surveyed by Dr. Dick Hammerschlag and Kevin Brittingham. 
Van de Koppel, J., J. Huisman, R.van de Wal, and H. Olff. 1996. Patterns of herbivory 
along a productivity gradient: an empirical and theoretical investigation. Ecology 
77(3):736-745. 
Van de Koppel, J., P.M.J. Herman, P. Thoolen, and C.H.R. Heip. 2001. Do alternate 
stable states occur in natural excosystems? Evidence from a tidal flat. Ecology 
82(12):3449-3461. 
Van Engel, W.A., and E.B. Joseph. 1968. Characterization of Coastal and Estuarine Fish 
Nursery Grounds and Natural Communities. Final Report, Bureau of Commercial 
Fisheries, Virginia Institute of Marine Science, Gloucester Point, VA. 43pp. 
Van Raalte, C.D., I. Valiela and J.M. Teal. 1976. Production of epibenthic salt marsh 
algae: light and nutrient limitation. Limn. Oceanogr. 21:862-872. 
Velinsky, D.J., and J.D. Cummins. 1996. The District of Columbia Fish Tissue Analysis: 
Distribution of Chemical Contaminants in 1993-1995 Wild Fish Species in the 
District of Columbia.  Prepared by the Interstate Commission on the Potomac 
River Basin for the D.C. Government Environmental Health Administration, 
Water Resources Management Division. Washington, D.C. 
Velinsky, D.J., B. Gruessner, C. Haywood, J. Cornwell, R. Gammisch, and T.L. Wade. 
1997. The District of Columbia Sediment Core Analysis: Determination of the 
Volume of Contaminated Sediments in the Anacostia River, District of 
Columbia.Prepared and Submitted by the Interstate Commission on the Potomac 
River Basin to the D.C. Environmental Regulation Administration. 
Vorberg, R. 1993. Effects of different corer sizes on sampling strategy with regard to 
patchiness in a freshwater tidal flat area. Arch. Hydrobiol. Suppl. 75(3-4):397-
405. 
Warner, A., D. Shepp, K. Corish,  and J. Galli. 1997. An Existing Assessment of 
Contaminants in the Anacostia Watershed.  Metropolitan Washington Council of 
Governments. Washington, D.C. 
Waters, T.F. 1977. Secondary production in inland waters. Advances in Ecological 
Research 10:91-164.  
 
 
354 
 
Weller, M.W. 1978. Management of freshwater marshes for wildlife.  In Freshwater 
wetlands: Ecological processes and management potential, eds. R.E. Good, D.F. 
Whigham, and R.L. Simpson, 267-284. New York: Academic Press. 
Wetzel, R.G. 1983. Limnology. CBS College Publishing. New York. 760p. 
Whigham, D.R., and R. L. Simpson. 1975. Ecological studies of the Hamilton marshes. 
Progress report for the period June 1974 ? January 1975. Rider College, Biology 
Dep., Lawrenceville, N.J. 
Whigham, D.R., and R. L. Simpson. 1976. The potential use of freshwater tidal marshes 
in the management of water quality in the Delaware River. Pages 173-186 in J. 
Tourbier and R. W. Pierson, eds. Biological Control of Water Pollution. 
University of Pennsylvania Press, Philadelphia, P.A. 
Whigham, D.R., J. McCormick, R.E. Good, and R. L. Simpson. 1978. Biomass and 
Primary Production in Freshwater Tidal Wetlands of the Middle Atlantic Coast. 
Pgs. 3-20, In: Good, R.E., D.F. Whigham, and R. L. Simpson eds. 1978. 
Freshwater Wetlands: Ecological Processes and Management Potential. Academic 
Press, New York.  
White, D.A. 1993. Vascular plant community development on mudflats in the Mississippi 
River delta, Louisiana, USA. Aquatic Botany 45:171-194. 
Whitney, D.E. and W.M. Darley. Unpublished data reported in Pomeroy L.R. and R.G. 
Wiegert, eds. 1981. The Ecology of a Salt Marsh. Ecological Studies; Vol. 38. 
Springer-Verlag: New York. 271p. 
Young, T.P., J.M. Chase, and R.T. Huddleston. 2001. Community succession and 
assembly-comparing, contrasting and combining paradigms in the context of 
ecological restoration. Ecological Restoration 19(1):5-18. 
Yozzo, D.J., Rhoads, J.M., Wilber, P., Nuckols, W., Hollen, L., and Will, R. 2001. 
Beneficial uses of dredged material for habitat creation, enhancement, and 
restoration in NY/NJ Harbor. U.S. Army Corps of Engineers, New York District, 
New York, N.Y. 
Zedler, J.B. Ed. 2001. Handbook for Restoring Tidal Wetlands. CRC Press. Boca Raton, 
Florida. 439pp.