ABSTRACT 
 
 
 
 
Title of Document: TRANSPORT, SOURCES, AND QUALITY OF 
SESTON IN A PIEDMONT HEADWATER 
STREAM 
  
 David Charles Richardson, Doctor of Philosophy 
2008 
Directed By: Dr. Margaret Palmer, Marine, Estuarine and 
Environmental Science, Chesapeake Biological 
Laboratory 
Dr. Louis A. Kaplan, Stroud Water Research 
Center 
 
 
Streams transport and process particulate organic carbon (POC) within the 
suspended load (seston) after terrestrially-fixed (allochthonous) carbon enters lotic 
ecosystems or as instream (autochthonous) production is suspended from the 
streambed.  POC provides a basal resource for upstream food webs, but can also 
support heterotrophic metabolism in downstream rivers.  Yet, the controls on 
transport, sources, and biological availability (quality) of POC from headwaters are 
poorly understood.  I examined seston and POC dynamics in a 3
rd
-order headwater 
stream in SE Pennsylvania.  I studied the temporal controls on seston transport, 
composition, and sources, and the biological quality of POC.  
  I present evidence that seasonal patterns in seston and POC transport and 
composition reflect stream organism activity and cycles of autumnal leaf litter inputs 
  
and vernal algal production.  This work also provides the first evidence of recurrent 
nighttime peaks of seston transport; I attribute this pattern to bioturbation of 
streambed sediments that suspends particles during stream-organism nest digging, 
foraging, and movement.  I present the first effort to demonstrate that stream 
salamanders contribute to ecosystem level processes such as modification of seston 
and organic carbon flow.   
Mixing model analyses of seston composition indicate that seston in small 
streams is predominantly mineral-core particles; however, POC fluxes are primarily 
organic-core particles.  Furthermore, the traditional view of headwater POC as leaf 
detritus should be expanded to include algal-derived particles, even in forested 
headwater streams.  Finally, I report a new method for measuring the quality of 
suspended POC from aquatic environments using heterotrophic respiration as a metric 
for lability.  Heterotrophic respiration rates for suspended POC measured using this 
method were an order of magnitude higher than respiration rates previously reported 
for benthic POC. 
The results from this research highlight the importance of seston for both 
headwater streams and as a longitudinal linkage of allochthonous and autochthonous 
organic matter to downstream ecosystems.  The magnitude of this connection is 
controlled by the quality, quantity, and timing of POC delivery. 
 
  
 
 
 
 
 
 
 
 
 
 
 
TRANSPORT, SOURCES, AND QUALITY OF SESTON IN A PIEDMONT 
HEADWATER STREAM   
 
 
 
By 
 
 
David Charles Richardson 
 
 
 
 
 
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 
2008 
 
 
 
 
 
 
 
 
Advisory Committee: 
Professor Margaret A. Palmer, co-Chair 
Dr. Louis A. Kaplan, co-Chair 
Professor William F. Fagan 
Professor Roberta L. Marinelli 
Dr. Michael J. Paul 
 
  
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
? Copyright by 
David C. Richardson 
2008 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ii 
 
Preface 
This dissertation contains a single introduction section, and four research chapters.  
Chapter 1, 2, 3, and 4 are presented in manuscript form with introduction, methods, 
results, and discussion followed by tables, figure captions and figures.  A single 
literature cited section occurs at the end for citations throughout the dissertation.   
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 iii 
 
 Dedication 
For my grandfather and Maryland alumnus, Pete Sante. 
 
 iv 
 
Acknowledgements 
I would like to thank my two dissertation advisors, Lou Kaplan and Margaret Palmer, 
who have supported, guided and pushed me as I pursued this work.  I cannot thank 
Lou and Margaret enough for exposing me to the breadth of ecology, while 
continuing to challenge me to do my best.  Thank you to my committee, Bill Fagan, 
Roberta Marinelli, and Mike Paul for providing feedback and valuable conversations 
that have greatly improved this research. 
 
I owe thanks to the members of the Palmer Lab, both past and present, who have 
always been willing to lend assistance and support throughout this dissertation, 
including idea development, sleepless nights of sestosalamandology, last minute 
editing, and overall encouragement: Holly Menninger, Evan Grant, Brian Laub, 
Emily Bernhardt, Lauren Culler, Solange Filoso, Brooke Hassett, Owen McDonough, 
Aaron Moore, Karen Nelson, Chris Patrick, Bob Smith, Chris Swan, and Lie?Ann 
Van-Tull.  I especially want to thank Laura Craig for her friendship and assistance 
from the beginning. 
 
The staff at the Stroud Water Research Center and the Stroud Family has been 
instrumental in supporting my research and providing me with a home away from 
home.  A number of scientists and technicians assisted me in the field and laboratory 
including Melanie Arnold, Heather Brooks, Stephanie Dix, Dave Funk, Sara 
Geleskie, Mike Gentile, Sue Herbert, Mark Monk, Dave Montgomery, Jan Surma, 
Sherman Roberts, Tracy Wiegner, and summer interns and teachers including Jessie 
Dow, Kyle Cordes, Ira Glass, Tim Ligget, Jackie McIlwain, Bill Richardson, and 
Maddie Winters.  Karen Hogan, my fellow grad student at Stroud, provided me with 
invaluable advice, field help, and encouragement throughout this dissertation.  I 
would like to thank the senior science staff at Stroud for their ever-open doors: Bern 
Sweeney, Tom Bott, Charlie Dow, and John Jackson.  Denis Newbold and Anthony 
Aufdenkampe have been my secondary committee, providing insight, advice, 
exposure to new methods and literature, and were vital to my graduate education. 
 
My research was financially supported by NSF grants DEB-0404801 (Ecological 
Circuitry Collaboratory) and DEB-0543526.  The Ecological Circuitry Collaboratory, 
including Holly Ewing, Kathie Weathers, and Peter Groffman, provided me with my 
first real exposure to ecosystem ecology and introduced me to a number of top 
ecology graduate students and scientists and for that, I am grateful.   
 
I thank my family and friends for support and encouragement, and for instilling me 
with the value of the pursuit of knowledge and hard work during my lifelong foray 
into school and academia.  I owe Kat Cappillino a huge thank you, for being my best 
friend and love, for supporting me, and most of all, for teaching me that a scientist 
makes his own luck.  
 
 v 
 
Table of Contents 
 
 
Preface........................................................................................................................... ii 
Dedication....................................................................................................................iii 
Acknowledgements...................................................................................................... iv 
Table of Contents.......................................................................................................... v 
List of Tables .............................................................................................................. vii 
List of Figures............................................................................................................viii 
Introduction................................................................................................................... 1 
Chapter 1: Temporal dynamics of seston: A recurring nighttime peak and seasonal 
shifts in composition in a stream ecosystem................................................................. 8 
Introduction............................................................................................................... 8 
Methods................................................................................................................... 10 
Results..................................................................................................................... 16 
Discussion............................................................................................................... 21 
Tables...................................................................................................................... 31 
Figure Captions....................................................................................................... 32 
Figures..................................................................................................................... 35 
Chapter 2: Contributions of salamanders and macroinvertebrates to seston export in 
artificial stream ecosystems ........................................................................................ 44 
Introduction............................................................................................................. 44 
Methods................................................................................................................... 46 
Results..................................................................................................................... 51 
Discussion............................................................................................................... 54 
Tables...................................................................................................................... 59 
Figure Captions....................................................................................................... 60 
Figures..................................................................................................................... 61 
Chapter 3: Modeling sources and seasonal dynamics of seston and suspended 
particulate organic carbon within a stream ecosystem ............................................... 66 
Introduction............................................................................................................. 66 
Methods................................................................................................................... 69 
Results..................................................................................................................... 77 
Discussion............................................................................................................... 80 
Tables...................................................................................................................... 90 
Figure Captions....................................................................................................... 92 
Figures..................................................................................................................... 94 
Chapter 4: A method for measuring bacterial mineralization rates of suspended 
particulate organic carbon in stream ecosystems...................................................... 102 
Introduction........................................................................................................... 102 
Methods................................................................................................................. 106 
Results................................................................................................................... 112 
Discussion............................................................................................................. 114 
Tables.................................................................................................................... 122 
Figure Captions..................................................................................................... 127 
 
 vi 
 
Figures................................................................................................................... 129 
Appendices................................................................................................................ 134 
Appendix 1............................................................................................................ 134 
Appendix 1 Figure ................................................................................................ 136 
Literature Cited ......................................................................................................... 137 
 
 
 
 
 
 
 
 
 
 
 
 vii 
 
List of Tables 
 
Chapter 1 - Temporal dynamics of seston: A recurring nighttime peak and 
seasonal shifts in composition in a stream ecosystem 
 
Table 1.1. Mixing model results presented as the percent of seston and POC concentration by 
mass (mean ? SE) (pg. 31).    
 
Chapter 2 - Contributions of salamanders and macroinvertebrates to seston 
export in artificial stream ecosystems  
 
Table 2.1. Experiment design for the four day experiment.  ?Macroinverts? represents the 
macroinvertebrate community excluding crayfish (pg. 59). 
 
Chapter 3 - Modeling sources and seasonal dynamics of seston and suspended 
particulate organic carbon within a stream ecosystem 
 
Table 3.1. End members from literature values and this study used for 3-compartment mixing 
model to separate seston carbon (C
x
) and nitrogen content (N
x
) in leaf detritus (d), algal cells 
(a), and mineral-core (m) particles (pg. 90). 
 
Table 3.2. Seston concentration (conc.) and composition (means ? 1 SE) from watershed 
habitat survey in June 2007 (pg. 91). 
 
Chapter 4 - A method for measuring bacterial mineralization rates of suspended 
particulate organic carbon in stream ecosystems 
 
Table 4.1. Experimental Design for January 2006 and April 2007 experiments (pg. 122). 
 
Table 4.2. Initial particulate and dissolved organic carbon concentrations for all experiments 
(pg. 123).  
 
Table 4.3. Oxygen consumption curve parameter fits for each treatment and experiment (pg. 
124).   
 
Table 4.4. Temporal loss rate coefficients (k) and biological turnover times (T
b-observed
) for all 
treatments with either POC corrected for DOC respiration or DOC alone (pg. 125).   
 
Table 4.5. Ranges of respiration rates for different carbon sources (pg. 126).    
 
 
 viii 
 
List of Figures 
Chapter 1 - Temporal dynamics of seston: A recurring nighttime peak and 
seasonal shifts in composition in a stream ecosystem 
 
Figure 1.1. Hourly discharge of White Clay Creek over the sampling period.  Arrows and 
circles indicate sampling days (11 Aug 05, 21 Sep 05, 01 Nov 05, 07 Dec 05, 12 Jan 06, 24 
Feb 06, 31 Mar 06, 27 Apr 06, 24 May 06, 20 Jun 06, 26 Jul 06) (pg. 35).   
 
Figure 1.2. Seasonal changes in (a) seston concentration, (b) particulate organic carbon 
concentration, (c) seston carbon content, and (d) particulate organic carbon flux for day and 
night samples (pg. 36).  
 
Figure 1.3. Example diel curve from June sampling of seston concentration and particulate 
organic carbon (POC) concentration (pg. 37). 
 
Figure 1.4. Least squares linear regressions between temperature and seston concentration 
for both day and night samples (pg. 38).   
 
Figure 1.5. Least squares linear regressions between the maximum discharge of the 
antecedent storm and the increase of total seston concentration from day to night for each 
sampling month (pg. 39). 
 
Figure 1.6. Seasonal changes in (a) Chl a and (b) Pheo a concentrations for day and night 
samples (pg. 40).   
 
Figure 1.7. Seasonal changes in ?
13
C of POC averaged across day and night samples. Boxes 
(mean ? 1 SD) represent the range of ?
13
C of (a) epilithic algae as autochthonous and (b) four 
species of senesced leaf species as allochthonous end-members (pg. 41). 
 
Figure 1.8. Least squares linear regressions between total pigment concentrations (Chl a + 
Pheo a) and ?
13
C of POC (pg. 42). 
 
Figure 1.9. Environmental scanning electron microscope images of seston. (a) Most particles 
were small angular clays and silts with adsorbed bacteria or organic matter, while (b) other 
particles were recognizable as diatoms cells (e.g. Pinnularia) (pg. 43).    
 
 
 ix 
 
Chapter 2 - Contributions of salamanders and macroinvertebrates to seston 
export in artificial stream ecosystems  
 
Figure 2.1. Artificial streams and salamander additions. (a) Evan Grant adding salamanders 
into the artificial streams. (b) Salamander following addition to artificial stream (pg. 61). 
 
Figure 2.2. Seston inflow and outflow for Experiments A and B.  (a) Inflow seston 
concentration from header tank and (b) outflow seston concentration, (c) % seston exported 
and (d) transport distance (S
w
) in flume 1 and flume 2 (pg. 62).  
 
Figure 2.3. Discharge over the course of the experiment for flumes 1 and 2.  Discharge was 
measured prior to each treatment and once following treatment 4 (pg. 63).   
 
Figure 2.4. Interaction effects for Experiment B: (a) treatment by flume interaction for Sw, 
(b) treatment by flume interaction for % seston exported, (c) treatment by day or night 
interaction for Sw, and (d) treatment by day or night interaction for % seston exported (pg. 
64). 
 
Figure 2.5. Chl a for experiments A and B. (a) Chl a outflow concentrations and (b) % Chl a 
exported (pg. 65). 
 
 
 x 
 
Chapter 3 - Modeling sources and seasonal dynamics of seston and suspended 
particulate organic carbon within a stream ecosystem 
 
Figure 3.1.  Map of White Clay Creek, SE Pennsylvania where P1-P4 are ponds, S1-S4 are 
seeps, R1-R3 are riffles and N1-N3 are runs sampled for the watershed habitat sampling (pg. 
94).     
 
Figure 3.2.  Hypothetical partitioning of source contributions from three selected end 
members (leaf detritus, algal cells and mineral associated particles for mixtures collected as 
seston (pg. 95). 
 
Figure 3.3.  Seasonal changes in (a) seston carbon:nitrogen (C:N) ratios and modeling results 
for (b) seston concentration, (c) suspended POC concentration, and (d) suspended particulate 
nitrogen concentration by particle type for WCC 2005-2006 samples (pg. 96).   
 
Figure 3.4.  Least squares linear regression between pigment concentration (?g L
-1
) and algal 
POC (mg L
-1
) from WCC 2005-2006 baseflow modeling (pg. 97). 
 
Figure 3.5.  WCC historic baseflow seston characteristics averaged over 14 years (1993-
2006).  Panels show seasonal changes in (a) seston concentration, (b) suspended POC 
concentration, and (c) carbon content.  Modeling results of seston composition for seston 
concentration and suspended POC concentration are shown in panels (d) and (e) (pg. 98).  
 
Figure 3.6.  Comparisons between pond outflow and seep (a) seston concentration, (b) 
carbon content, (c) nitrogen content, and (d) carbon:nitrogen (C:N) ratios for watershed 
habitat samples collected in 2004 (pg. 99).   
 
Figure 3.7.  Elemental composition by habitat (pond, riffle, run, seep) and sample type 
(seston, herbs, CPOM, soil sediments, bank sediments) for watershed habitat samples 
collected in 2007 (pg. 100).   
 
Figure 3.8. Stable isotope ratios by habitat (pond, riffle, run, seep) and sample type (seston, 
herbs, CPOM, soil sediments, bank sediments) for watershed habitat samples collected in 
2007 (pg. 101).   
 
 
 xi 
 
Chapter 4 - A method for measuring bacterial mineralization rates of suspended 
particulate organic carbon in stream ecosystems 
 
Figure 4.1.  The left side of the incubation axle in the temperature controlled water bath (pg. 
129). 
 
Figure 4.2. (a) Oxygen consumption and (b) respiration rates for Jan 2006 experiment (pg. 
130).   
 
Figure 4.3. (a) Oxygen consumption and (b) respiration rates for Apr 2007 experiment (pg. 
131). 
 
Figure 4.4. (a) Oxygen consumption and (b) respiration rates for DOC lability experiment 
(pg. 132).  
 
Figure 4.5. Respiration rates for the dispersed, suspended particle treatments from Jan 2006 
and Apr 2007 experiments (pg. 133).  
 
 1 
 
Introduction 
Each year rivers export approximately 0.2 Pg of particulate organic carbon 
(POC) to the oceans (Spitzy and Ittekot 1991; Battin et al. 2008).  POC within the 
suspended load (seston) enters lotic ecosystems as terrestrially-fixed (allochthonous) 
carbon or from instream (autochthonous) production.  Streams and rivers not only 
generate and transport POC, but also play an active role in the processing of POC 
(Cole and Caraco 2001; Wipfli and Gregovich 2002; Mayorga et al. 2005).  
Knowledge of the sources of the carbon (allochthonous vs. autochthonous), timing 
and distance of transport, and biological interaction has implications for the energetic 
state of aquatic ecosystems (Vannote et al. 1980), the importance of headwaters to 
downstream ecosystems and management of watersheds (Meyer and Wallace 2001), 
and understanding the role of lotic ecosystems in the global carbon budget (Cole et al. 
2007). 
 
  Potential sources of seston include organic-core particles such as algal or 
bacterial cells, leaf or wood fragments, dissolved organic carbon flocculates, and low-
organic content silt or clay particles with adsorbed organic carbon.  Seston in streams 
has historically been considered to be primarily organic-core particles of terrestrial 
origin, such as fragments of leaves and wood created by macroinvertebrate feeding 
and egestion upstream (Vannote et al. 1980).  However, elemental, radioisotope, and 
microscopic analyses showed that most seston in streams and rivers is composed of 
small (<20 ?m), old, high mineral content particles with carbon to nitrogen ratios 
lower than leaves or wood (Meybeck 1982; Cole and Caraco 2001; Dodds and Whiles 
 
 2 
 
2004).  These seemingly contradictory results could be reconciled if seston were a 
mixture of inorganic particles with adsorbed terrestrially-derived organic matter 
(mineral-core, Aufdenkampe et al. 2001), algal cells, and fragmented leaves or woody 
debris. 
 
POC dynamics include transport downstream, deposition to and resuspension 
from the streambed (Thomas et al. 2001; Newbold et al. 2005).  Hydrological and 
biological dynamics control the timing of POC transport from headwaters.  Storms 
are the most important factor regulating variability in the magnitude of annual or 
monthly seston and POC fluxes (Golladay et al. 2000).  However, because small 
streams are under storm conditions for only a small fraction of time, temporal POC 
dynamics under baseflow conditions may disproportionately regulate or reflect stream 
ecosystem processes.  At baseflow, POC travels short distances downstream in a 
series of hops (Cushing et al. 1993; Webster et al. 1999; Thomas et al. 2001) 
interspersed with periods of retention on the streambed (Newbold et al. 2005).  
Regulation of suspension from the streambed has been infrequently examined, 
especially on a diel time-scale.  However, seasonal peaks of POC concentration have 
been attributed to increased macroinvertebrate fragmentation of leaf litter as stream 
temperature increases (Webster et al. 1990; Molla et al. 2006).       
 
The location of POC mineralization by aquatic organisms depends on the 
timing and distance of transport, but also on the quality of the organic matter 
(Webster et al. 1999; Rosi-Marshall 2004).  Investigation of CO
2
 dynamics in large 
 
 3 
 
rivers indicate that old, terrestrially-derived POC is an important component of 
heterotrophic river respiration (Cole and Caraco 2001; Richey et al. 2002).  However, 
POC transported under baseflow has been considered metabolically inert (Vannote et 
al. 1980; Webster et al. 1999), and neither nutritious to macroinvertebrates (Benke 
and Wallace 1980) nor very important to downstream food webs (Finlay et al. 2002; 
McCutchan and Lewis 2002).  A continuous array of labilities likely exists within the 
suspended load (seston), with POC ranging from labile (e.g. fresh algal cells) to 
refractory (e.g. humified soil organic matter) (e.g. Trumbore 2000), but more 
measurements of suspended POC quality are needed.   
 
My goals for this dissertation were to (1) determine the sources of seston and 
POC to headwater stream ecosystems, (2) understand controls on seston composition 
and transport, and (3) determine if suspended POC is biologically available to 
heterotrophic microorganisms.  I used a combination of field surveys, field and lab 
experiments, and modeling to address these goals and attempted to understand seston 
dynamics in a 3
rd
-order Pennsylvania piedmont stream.  I provide evidence that 
seston transport at baseflow is regulated by seasonal and diel patterns in litterfall, 
instream primary production, and stream organism activity.  This work provides the 
first evidence of recurrent nighttime peaks of seston transport; I attribute this pattern 
to bioturbation of streambed sediments.  I also show that algal carbon is an important 
component of seston, even in forested headwater streams.  Finally, I present a new 
method for measurement of POC quality, and report some of the few measurements 
 
 4 
 
of suspended POC lability in headwater streams.  A brief description of each chapter 
follows.     
 
Chapter 1 ? I measured the baseflow concentration and composition of seston over a 
one-year period in White Clay Creek, a third-order stream in the southeastern 
Pennsylvania Piedmont, to assess the temporal variability in seston concentration and 
quality at seasonal and diel time scales. Each month, I sampled baseflow seston every 
1.5 hours over a 24-hour period and measured concentration, carbon composition, 
pigment content, and 
13
C isotopic ratios.  Seston and particulate organic carbon 
(POC) concentrations exhibited a strong diel pattern; nighttime concentrations 
exceeded daytime concentrations by 80% and 43%, respectively.  Chlorophyll a 
concentration in seston (1.3 ? 0.8 ?g L
-1
: mean ? SE) did not exhibit a diel pattern.  I 
attribute the diel pattern of seston concentration to bioturbation by the nocturnal 
stream community, including crayfish, amphibians, eels, and macroinvertebrates.  
Seasonally, carbon content of seston increased from 9% throughout most of the year 
to 15% during November, December, and March, while seston ?
13
C was depleted in 
the late fall and enriched in early spring months relative to the rest of the year.  
Chlorophyll a and pheophytin a concentrations in seston also peaked during the early 
spring.  Seasonal patterns in seston and POC composition reflect cycles of autumnal 
leaf litter inputs and vernal algal production.  Bioturbation and shifts in organic 
carbon inputs mediate changes in POC quality and fluxes, which affect the 
bioavailability of POC and ultimately influence rates of heterotrophic respiration. 
 
 
 5 
 
Chapter 2 ? I tested the hypothesis that bioturbation by salamanders and stream 
macroinvertebrates is a mechanism for nighttime increases in seston transport.  I 
measured the concentration of seston over four consecutive days and nights in both 
the presence and absence of macroinvertebrates and salamanders in stream-side 
flumes.  I describe results for seston transport including the outflow concentration of 
seston, percentage of seston exported from the flumes, average seston transport 
distance (S
w
), and chlorophyll a content of seston.  During the daytime of the 
salamander addition, percent seston exported (29.3% ? 2.2: mean ? SE) and S
w
 (24.5 
m ? 1.5: mean ? SE) increased by 44% and 30%, respectively.  Macroinvertebrate 
additions caused no effects on seston transport, and there were no patterns in 
suspended chlorophyll a concentrations.  Salamanders caused daytime increases in 
seston, an opposite trend to the hypothesized nighttime increases.  However, this was 
the first effort to demonstrate stream salamander contribution to ecosystem level 
processes such as modification of seston and organic carbon flow.  Salamander 
activity, density, and size make them ideal candidates for influencing transport of 
seston and should be considered further along with other stream organisms such as 
eels and crayfish. 
 
Chapter 3 ? I used a new approach that defines seston end-members by particle type 
? mineral-dominated, terrestrial plant detritus and algal ? with distinct organic carbon 
and nitrogen contents and used a mixing model analysis to quantify fractions of 
seston, in a 3rd-order southeastern Pennsylvania stream.  I observed that seston mass 
concentrations were dominated by mineral particles (65% ? 1%: mean ? SE).  
 
 6 
 
However, algal and terrestrial plant particles, because of their high organic content, 
still represented a large fraction of the particulate organic carbon (89% ? 0.3%: mean 
? SE) and nitrogen (87% ? 0.4%: mean ? SE) in transport.  Furthermore, seasonal 
patterns in seston sources reflected cycles of autumnal leaf litter inputs and vernal 
algal production.  The results suggest that the biological view of seston (organic 
matter) in small streams as fragmented and egested leaf-litter is compatible with the 
biogeochemical perspective that seston (mass) is mineral-derived.  However, the 
biological view should be expanded to include algal-derived particles.  The source of 
organic matter can indicate lability for biological use and should ultimately influence 
rates of heterotrophic respiration in downstream rivers. 
 
Chapter 4 ? I developed a method to measure heterotrophic respiration rates (quality) 
of POC.  The method involves keeping particles in suspension, correcting for 
respiration of dissolved organic carbon, and extending rate measurements to 30-40 
days.  BOD bottles were filled with POC at elevated (5X) concentration using 
particles collected and concentrated by tangential flow filtration.  Glass beads added 
to the BOD bottles prevented aggregation while the bottles were rotated around an 
incubation axel to ensure that the particles were suspended. Replicate samples were 
periodically sacrificed for DO measurements.  I observed that, at the beginning of the 
experiment, suspended particles exhibited POC mineralization rates that were 5-fold 
higher than particles that were allowed to settle.  Mineralization rates were highest at 
the onset of the experiment at 180 to 300 mg O
2
 g initial POC
-1 
day
-1
 and then 
declined in the following weeks to 10 mg O
2
 g initial POC
-1 
day
-1
.  Further, I 
 
 7 
 
compared respiration rates from this study to previous studies of stream POC and 
found that POC is more labile than previously assumed.  Labile POC may be 
mineralized by bacteria within upstream reaches; more resistant POC may subsidize 
downstream ecosystems with carbon from the headwaters.     
 
 8 
 
Chapter 1: Temporal dynamics of seston: A recurring nighttime 
peak and seasonal shifts in composition in a stream ecosystem 
 
Introduction 
Particulate organic carbon (POC) provides energy for microorganisms and 
macroinvertebrates in headwater streams (Hedin 1990), and when transported in the 
suspended load (seston), supports heterotrophic metabolism in downstream reaches 
and rivers (Cole and Caraco 2001; Wipfli and Gregovich 2002).  POC transport 
represents a longitudinal linkage between upstream autochthonous and allochthonous 
production and downstream ecosystems with the magnitude of connection controlled 
by the quality, quantity, velocity, and timing of POC delivery (Mayorga et al. 2005).   
 
Hydrological and biological dynamics control the timing of POC transport 
from headwaters.  Storms are the most important factor regulating variability in the 
magnitude of annual or monthly seston and POC fluxes (Golladay et al. 2000).  
However, because small streams are under storm conditions for only a small fraction 
of time, temporal POC dynamics under baseflow conditions may ultimately regulate 
or reflect stream ecosystem processes.   At baseflow, seasonal peaks of POC 
concentration have been attributed to increased macroinvertebrate fragmentation of 
leaf litter as stream temperature increases (Webster et al. 1990; Molla et al. 2006).  
However, diel patterns of seston transport have not been extensively examined.  
Peaks in seston concentration, found in a deciduous European forest reach at mid-
morning and evening, were attributed to diurnal rhythms in algal drift (Pozo et al. 
 
 9 
 
1994), but no diel patterns were found in either boreal (Naiman 1982) or glacier-fed 
alpine streams (Hieber et al. 2003).  Knowledge of temporal patterns can help 
evaluate the sources of POC (Kendall et al. 2001), quality of the energetic subsidy to 
downstream biota (Rosi-Marshall 2004), controls on transport (Pozo et al. 1994), and 
lead to better estimates of annual POC fluxes, especially in streams with pulsed inputs 
of organic matter (Fisher et al. 2004).   
 
  Potential sources of seston include organic-core particles such as algal or 
bacterial cells, leaf or wood fragments, dissolved organic carbon flocculates, and low-
organic content silt or clay particles with adsorbed organic carbon.  Seston in streams 
has historically been considered to be primarily organic-core particles of terrestrial 
origin, such as fragments of leaves and wood created by macroinvertebrate feeding 
and egestion upstream (Vannote et al. 1980).  Stable isotope and chlorophyll a (Chl a) 
analyses supported the idea that stream POC is terrestrial (Finlay et al. 2002) with 
minimal contributions from autochthonous production (Angradi 1993).  However, 
elemental, radioisotope, and microscopic analyses showed that most seston in streams 
and rivers is composed of small (<20 ?m), old, high mineral content particles with 
carbon to nitrogen ratios lower than leaves or wood (Meybeck 1982; Cole and Caraco 
2001; Dodds and Whiles 2004).  These seemingly contradictory results could be 
reconciled if seston were a mixture of inorganic particles with adsorbed terrestrially-
derived organic matter (Aufdenkampe et al. 2001), algal cells, and fragmented leaves 
or woody debris.   
 
 
 10 
 
I report results from intensive field surveys of seston quantity and quality in a 
3
rd
-order piedmont stream in southeastern Pennsylvania, U.S.A.  I measured seston 
concentration and composition using a comprehensive array of measurements through 
monthly diel measurements over a one-year period to assess the temporal variability 
in seston transport and quality at both seasonal and diel time scales.  Here, I report (1) 
a consistent and repeatable diel pattern of nighttime peaks in seston and POC 
concentrations that I attribute to bioturbation and (2) seasonal shifts in composition 
associated with annual cycles of autumnal leaf-fall and vernal algal production prior 
to canopy closure. 
Methods 
Site description ? The upper east branch of White Clay Creek (WCC: N39?51?33?, 
W75?46?59?) is a 3
rd
 order watershed that drains 7.3 km
2
 of eastern piedmont land.  
The watershed has 23% of total area as temperate deciduous forest, 52% as pastures 
or hay fields for horses or cattle, and 22% as row crop agriculture, but has a largely 
intact riparian forest and livestock have minimal access to the stream (Newbold et al. 
1997).  The mean daily baseflow discharge averaged 122 L s
-1
 over the past five years 
with a high annual mean discharge of 174 L s
-1
 and low of 59 L s
-1
.  At baseflow, 
WCC has high nutrient concentrations (3-5 mg NO
3
-N L
-1
; 4-33 ?g PO
4
-P L
-1
) from 
agricultural land use and dissolved organic carbon (DOC) concentrations of 1.5 mg 
DOC L
-1
 (Newbold et al. 1997).  Most of the particulate organic matter input 
(predominantly leaves) occurs from late October to mid-November (J. D. Newbold 
unpubl.).   
 
 
 11 
 
Temperature and discharge ? Throughout the sampling period, stream water 
temperature was recorded using automated temperature loggers (Onset Computer 
Corporation) and stream stage was recorded using an automated pressure transducer 
(Telog Instruments), both at fifteen minute intervals.  Stream stage was converted to 
stream discharge using a 10-year rating curve, according to the method of Dunne and 
Leopold (1978), and averaged to hourly discharge.  I used the discharge record to 
identify the antecedent storm discharge and defined storms as two times baseflow 
discharge because such flows typically result in bed mobilization in WCC (Newbold 
et al. 1997).   
 
Sampling protocol ? Under baseflow conditions, stream water was collected every 1.5 
hours during one 24-hour period each month, from August 2005 to July 2006 (Fig. 
1.1) using an automated water sampler (Teledyne Isco), with the collection tube (9.5 
mm inner diameter) at 2/3 depth in the middle of the stream.  Sampling days (11 Aug 
05, 21 Sep 05, 01 Nov 05, 07 Dec 05, 12 Jan 06, 24 Feb 06, 31 Mar 06, 27 Apr 06, 24 
May 06, 20 Jun 06, 26 Jul 06) were not evenly spaced; however, I delineated names 
based on the sampling month (Aug, Sep, Nov, Dec, Jan, Feb, Mar, Apr, May, Jun, 
Jul).  On 22 August 2006, sampling (see spatial sampling section) included three 
additional sites, all 2
nd
-order streams within the watershed upstream of the WCC site.  
Simultaneously at each of the four locations, three samples were taken at night and 
three were taken during the day to confirm that diel patterns were not unique to the 
main sampling location.    
 
 
 12 
 
The automated sampler was set to collect four 1-L samples at each time 
period.  The first sample was considered a rinse of the inflow tubing and discarded, 
the second was used for pigment analysis, and the final two samples were used for 
replicate seston and POC analysis.  Only the first replicate of seston and POC 
analysis is reported; the second replicate was used to ensure the precision of the 
automated sampling or as a substitute if the first replicate was lost.  Once or twice 
each sampling month, three replicate grab samples were taken by submerging 1-L 
bottles at 2/3 depth to ensure that automated sampling was unbiased in selection of 
particles and accurate in representing stream seston.       
 
Seston ? Seston concentration was determined gravimetrically using the dry weight of 
total suspended solids.  Entire 1-L stream water samples were filtered, after recording 
precise volume, through stacked pairs of pre-weighed and pre-ashed (480?C for 4.5 
hours) 47 mm glass fiber filters (Millipore AP40 with nominal pore-size of 0.7 ?m).  
The sample was filtered through two filters to check analytical accuracy, assess the 
integrity of the top filter and ensure even distribution of seston across the filter.  The 
sample was discarded if the bottom filter gained or lost more than 0.2 mg.  The filters 
were dried for 12-18 hours at 60?C, re-weighed to calculate seston mass on the filter, 
and divided by filtered volume to calculate seston concentration (mg seston L
-1
).   
 
Pigments ? Samples for pigment analysis were collected on a glass fiber filter 
(Millipore AP40), frozen and processed within 30 days using a modified hot ethanol 
extraction method (Biggs and Kilroy 2000).  The samples were heated to 78?C for 10 
 
 13 
 
minutes in 90% ethanol and then stored frozen for 24 hours.  Chl a and pheophytin a 
(Pheo a) concentrations (?g L
-1
) were calculated using measurements of absorbance 
on a spectrophotometer (Beckman Coulter DU-640) before and after acidification 
(Biggs and Kilroy 2000).  Pheo a, a degradation product of Chl a, suggests dead or 
senescing algal cells (Peterson and Stevenson 1992). 
 
Elemental and stable isotope analysis ? Subsamples of each seston filter were wet 
with nanopure water and fumed with hydrochloric acid vapors for 18 hours, 
effectively removing carbonates (A. K. Aufdenkampe unpubl.).  Subsamples were 
analyzed in batches on an Elemental Analyzer (Costech ECS 4010) interfaced with an 
Isotope Ratio Mass Spectrometer (Thermo DeltaPlus XP).  Standard curves for the 
elemental and isotopic analysis were created for each batch using 8 or 9 isotopically 
enriched (+37.63?) and depleted (-26.39?) L-glutamic acid standards over a range 
of carbon masses that encompassed the subsample masses.  Stable carbon isotope 
ratios, reported using ?
13
C notation, were normalized relative to the Vienna Peedee 
Belemnite standard (Coplen et al. 2006).  POC concentration (mg C L
-1
) was 
calculated from the mass of C analyzed, the fraction of total sample analyzed, and the 
volume of stream water filtered.  Weight percent organic carbon (%OC) of seston was 
calculated as the ratio of POC (mg POC L
-1
) to seston (mg seston L
-1
) concentrations 
and POC flux (mg s
-1
) was calculated by multiplying the POC concentration (mg L
-1
) 
by the average hourly discharge (L s
-1
). 
 
 
 14 
 
End member analysis ? Four species of dried, senesced leaves (American beech: 
Fagus grandifolia Ehrh., northern red oak: Quercus rubra L., red maple: Acer rubrum 
L., and tulip poplar: Liriodendron tulipifera L.) were collected from the forest floor, 
dried, and analyzed for stable carbon isotopes as potential allochthonous sources of 
POC.  In May 2007, epilithon (predominantly Melosira varians), collected from six 
rocks in riffles in WCC and rinsed carefully to remove sediments and trapped leaf 
pieces, was analyzed for stable carbon isotopes as potential autochthonous sources of 
POC.    
 
Mixing model ? I modeled seston composition as a mixture of two distinct particle 
types: organic-core particles composed of algal cells, leaf fragments, and bacterial 
cells (carbon content for mixture = 45%); and, mineral-core particles with adsorbed 
organic carbon and bacterial cells (carbon content = 1.5%) (Meybeck 1982; Hedges et 
al. 1986).  For each sample, I simultaneously solved equations (1.1) and (1.2)  
 
Equation 1.1  S = P
o
 + P
m
                                                                                                             
             
Equation 1.2  POC = 0.45 x P
o
 + 0.015 x P
m
                                                                                     
 
using measured seston (S) and POC concentrations to determine the concentration of 
organic-core (P
o
) and mineral-core (P
m
) particles in seston and the concentration of 
organic-core (0.45 x P
o
) and mineral-core (0.015 x P
m
) particles in POC.  Mixing 
model results are reported as the fraction of total concentration by particle type (e.g. 
 
 15 
 
P
o
/S as the organic-core particle fraction of seston by weight).  Results were split into 
two groups: (1) months with increased organic inputs from allochthonous (Nov, Dec) 
and autochthonous sources (Mar) and (2) lower organic inputs (all other months). 
 
Environmental scanning electron microscopy (ESEM) ? Seston was collected on 16 
January 2006 from WCC and images of particles in stream water were captured at 
1300X to14500X magnification using an environmental scanning electron 
microscope (Electro Scan) with an accelerating voltage of 28 kV.   
    
Statistical analyses ? All statistical analyses were conducted using SAS v9.1 Level 
1M3 (2007, SAS Institute Inc., Cary, North Carolina, United States of America).  I 
performed two-way ANOVAs (PROC MIXED) for seston, POC, Chl a and Pheo a 
concentrations, POC flux, percent carbon of seston, and ?
13
C data with month and 
day or night as treatments.  Samples collected at 1.5 hour intervals throughout each 
day or night were considered replicates.  For the spatial sampling, I performed two-
way ANOVAs for seston, POC and Chl a with station and day or night as treatments.  
To ensure homogeneity of variances, it was necessary to log-transform several 
variables (seston, POC); these are reported as back-transformed geometric means.  
Following each ANOVA, I made multiple mean comparisons using the Tukey?s test.  
Tests of auto-correlation (Durbin-Watson) revealed a violation of ANOVA 
assumptions of independence of residuals for seston, POC, carbon content, POC flux, 
and stable isotopes within each month.  Therefore, I used the interaction mean 
squares (MS) as the denominator in the F-statistic, rather than the residual MS, for the 
 
 16 
 
test of diel main effects.  I did, however, use the within-month mean squared error in 
conjunction with the Tukey?s test to identify specific months in which the diel effect 
was observed, reasoning that most or all of the serial auto-correlation reflected real 
(mechanistic) trends over the course of the day or night, rather than over-sampling of 
the random error.    
 
Linear correlation (PROC REG) was used to identify relationships between 
dependent and explanatory variables of interest including temperature, discharge, and 
antecedent storm timing and magnitude.  Outliers were identified (PROC 
ROBUSTREG) using a prespecified residual value (3.0).  For quality control, I 
analyzed the correlation between the automated sampler and grab samples and 
compared the slope and intercept to the 1:1 line using t-tests (PROC TTEST).  For the 
mixing model results, fractions of seston and POC as mineral-core and organic-core 
particles were compared between the grouped months using paired t-tests (PROC 
TTEST).  
Results 
Over the year, one hundred eighty-three samples were taken during baseflow 
conditions on 11 monthly sampling days over 24 hours to 25.5 hours (Fig. 1.1).  
Mean water temperatures ranged from 2?C in December to 20?C in August and July, 
while mean daytime temperature exceeded nighttime temperatures between 0.15?C 
(November) and 1.5?C (April).  WCC discharges, during sampling, ranged from 23 L 
s
-1
 in June to 120 L s
-1
 in February (Fig. 1.1), with minimum discharge in late 
summer, maximum discharge in winter.  Daytime discharge was not different than 
 
 17 
 
nighttime discharge (F
1,10
=0.94, p=0.35).  A series of storms occurred during the 
winter; one large summer storm, on 28 June 2006, reached twice the maximum 
discharge of any other storm (Fig. 1.1).   
   
Precision and accuracy of automated sampler ? There was a positive correlation in 
seston concentrations between grab and automated samples that explained almost 
90% of the variance (r
2
=0.88, F
1,13
=106.5, p<0.001, n=14).  One observation was 
excluded from the regression because it was identified as an outlier with a 
standardized robust residual (6.3) greater than the prespecified value (3.0).  The line 
was statistically identical to the 1:1 line, (intercept = 0, p =0.68; slope = 1, p=0.68) 
and all automated samples were within ?10% of grab samples.  Furthermore, I found 
low coefficients of variation for duplicate automated seston (6.5% ? 6.5: mean ? SD, 
n=183 pairs) and POC samples (8.8% ? 9.8: mean ? SD, n=104 pairs).         
 
Seston and POC ? Over the year, nighttime seston concentrations (3.0 mg L
-1
 ? 0.07: 
mean ? SE) exceeded daytime concentrations (1.7 mg L
-1
 ? 0.07: mean ? SE) by 80% 
(Fig. 1.2a, F
1,10
=76, p<0.001).  While mean nighttime concentrations were always 
greater than daytime concentrations, the difference was not significant during winter 
months (interaction effect: F
10,163
=2.4, p=0.01).  Seasonally, seston concentrations 
were lowest during the late fall and winter months and highest in the summer 
(F
10,163
=67, p<0.001).  I observed the largest diel variation of seston in June, when the 
nighttime concentration of seston (6.4 mg L
-1
 ? 0.43: mean ? SE) exceeded the 
daytime concentration (2.5 mg L
-1
 ? 0.21: mean ? SE; p<0.001) by 155% (Fig. 1.3).       
 
 18 
 
 
POC followed the same trends as seston, both on the diel and seasonal scales.  
Nighttime POC concentrations (0.24 mg C L
-1
 ? 0.005: mean ? SE) exceeded 
daytime concentrations (0.17 mg C L
-1
 ? 0.005: mean ? SE) by 43% (Fig. 1.2b, 
F
1,10
=15, p=0.003).  These differences, while observed each month, were significant 
only during August, April, May, and June (interaction effect: F
10,156
=8.6, p<0.001).  
POC concentrations varied by month and were lower in the winter months 
(F
10,156
=42, p<0.001).  The largest diel change for POC also occurred in June with a 
105% increase in POC concentration (Fig. 1.3) from day (0.21 mg C L
-1
 ? 0.013: 
mean ? SE; p<0.001) to night (0.43 mg C L
-1
 ? 0.017: mean ? SE).                
 
Both seston (r
2
 = 0.32, p=0.07) and POC (r
2
 = 0.17, p=0.21) concentrations, 
averaged by month, were not correlated with stream discharge. However, seston 
concentration correlated with other physical changes in the stream.  Water 
temperature was positively correlated with both day (Fig. 1.4: n=11, p=0.02, r
2
=0.45) 
and night (Fig. 1.4: n=11, p=0.02, r
2
=0.45) seston concentrations.  The slopes for day 
(0.07 ? 0.03: mean ? SE) and night (0.18 ? 0.07: mean ? SE) regression fits were not 
statistically different from each other (Fig. 1.4, t=1.55, df=8, p=0.13).   Mean 
nighttime POC concentrations were also positively correlated with water temperature 
(n=11, p=0.05, r
2
=0.37) but daytime concentrations were not (n=11, p>0.05, r
2
=0.26).  
The maximum discharge from the antecedent storm was negatively correlated with 
the percentage of increase of seston from day to night (Fig. 1.5: n=11, p=0.005, 
 
 19 
 
r
2
=0.61).  However, the antecedent storm discharge did not explain diel differences in 
POC concentrations (n=11, p=0.09, r
2
=0.28).      
 
Carbon content ? Overall, the daytime carbon content (11.4% ? 0.2: mean ? SE) was 
greater than the nighttime carbon content (9.3% ? 0.2: mean ? SE) by 1.2% (Fig. 
1.2c, F
1,10
=15, p=0.003).  Although the mean carbon content was higher in the 
daytime than in the nighttime in each month, the main effect was driven by 
significant differences in November and March.  The percent carbon varied 
seasonally, with the highest values in November, December, and March (F
10,156
=18, 
p<0.001).   
 
POC flux ? The downstream flux of POC had both diel and seasonal trends.  The 
nighttime POC fluxes (Fig. 1.2d: 17.3 mg s
-1
 ? 0.9: mean ? SE) were 45% greater 
than daytime fluxes (11.9 mg s
-1
 ? 0.6: mean ? SE; F
1,10
=20, p=0.001).  The 
magnitude of this effect was buffered in the winter, when both the day-night 
differences in POC concentration were lower and the stream discharge was higher 
(interaction effect: F
10,155
=3.7, p<0.001).  POC flux varied seasonally, peaking during 
the late winter and early spring (F
10,155
=45, p<0.001).  
 
Pigments ? Chl a concentration (1.26 ?g L
-1
 ? 0.8: mean ? SE) was less variable than 
seston or POC concentrations.  Chl a varied by month (Fig. 1.6a, F
10,165
=5.32, 
p<0.001), but there were neither diel (F
1,10
=0.04, p=0.77) nor interaction effects 
(F
10,165
=1.33, p=0.22).  The mean Chl a concentrations for March were significantly 
 
 20 
 
greater than all other months (p<0.05).  Pheo a, unlike Chl a, showed a diel difference 
(Fig. 1.6b, F
1,165
=28, p<0.001), with nighttime Pheo a concentrations (2.10 ?g L
-1
 ? 
0.07: mean ? SE) greater than daytime concentrations (1.56 ?g L
-1
 ? 0.07: mean ? 
SE).  Pheo a concentrations also varied by month (F
10,165
=122, p<0.001), and could 
be divided into three groups: the greatest concentrations during February and March, 
intermediate during late spring and summer and lowest during the late fall and early 
winter.  Pheo a concentrations were more variable than Chl a, and included monthly 
means both lower than and double the average Chl a concentration (Fig. 1.6a, b).         
 
Stable carbon isotopes ? The highest ?
13
C value of POC occurred during March, 
while the lowest values occurred during December and January (Fig. 1.7: F
10,142
=28, 
p<0.001).  For potential allochthonous end-members, all four species of leaves had 
similar ?
13
C values (Fig. 1.7, box b: -28.2? ? 0.6: overall mean ? SD), but the 
epilithic algae were more enriched (Fig. 1.7, box a: p=0.03, -26.5? ? 1.6: overall 
mean ? SD).  ?
13
C was positively correlated with mean pigment (Chl a + Pheo a) 
concentrations (Fig. 1.8: n=11, p=0.016, r
2
=0.49); this trend was driven by the 
February and March peaks in pigment concentrations.  The ?
13
C of March samples (-
25.7? 0.24: mean ? SE) was greater than the end-member for epilithic algae, but was 
within 1 SD of the end-member mean.   ?
13
C value of POC had neither diel 
(F
1,142
=0.43, p=0.52) nor interaction effects (F
10,142
=1.48, p=0.15). 
 
Mixing model ? Throughout the year, mineral-dominated or mineral-core particles 
contributed to >70% of seston concentration, with organic-core particles contributing 
 
 21 
 
<30% of seston concentration by weight.  However, organic-core particles 
consistently contributed the majority of POC concentration by weight (Table 1.1).  
The contribution of organic-core particles to both seston and POC increased during 
periods of increased carbon influx (Nov, Dec, and Mar) from leaf fall and vernal algal 
blooms compared to other months (Table 1.1).  
 
Spatial sampling ? In all four streams sampled within the WCC watershed in August, 
nighttime concentrations of seston (9.0 mg L
-1
 ? 5.5: mean ? SE) and POC (0.64 mg 
L
-1
 ? 0.37: mean ? SE) were greater than daytime concentrations (2.2 mg L
-1
 ? 0.4: 
mean ? SE) and (0.19 mg L
-1
 ? 0.02: mean ? SE), respectively (p<0.05), but there 
were no differences across the streams.  Suspended Chl a concentrations were low in 
all four streams and there was neither diel nor spatial differences among streams. 
Discussion 
Diel patterns ? Possible mechanisms for the consistent nighttime increases in seston 
and POC concentrations (Fig. 1.2a, b) include diel cycles in stream flow, primary 
production, feeding-related particle processing (e.g. shredding, suspension of non-
ingested food particles, egestion), and  bioturbation.  Stream flow can be higher at 
night due to decreases in evapotranspiration in deciduous forests (Czikowsky and 
Fitzjarrald 2004).  In WCC, nighttime baseflow is <5% greater than daytime 
baseflow, but only during the growing season (J. D. Newbold unpubl.); I observed 
diel trends for seston and POC concentration in all seasons.  Primary production in 
WCC causes diurnal increases in dissolved organic matter concentrations (Kaplan and 
Bott 1982) and could be expected to cause similar diurnal increases in the Chl a 
 
 22 
 
portion of seston due to algal drift (Pozo et al. 1994).  However, I observed the 
opposite pattern with nighttime increases in seston concentration, and I saw no diel 
variations in Chl a in any of the sampling months (Fig. 1.6a) or across the watershed.  
Feeding-related particle processing should increase at night due to increased 
nocturnal macroinvertebrate activity (Wagner 1991), and could contribute to seston 
through particle shredding (Wallace et al. 1982), suspension of non-ingested food 
particles (?sloppy feeding?), or suspension of egested particles.  I would expect, 
however, that all of these processes would preferentially contribute carbon-rich 
particles to seston, whereas the carbon content of seston actually declined at night 
(Fig 1.2c).  Thus, none of these explanations prove satisfactory.  The remaining 
explanation is bioturbation, activity that suspends streambed particles during nest 
digging, foraging, and movement (Moore 2006).  
 
Several lines of evidence suggest that the diel pattern is caused by 
bioturbation.  The diel pattern appears to be biologically driven because rising water 
temperatures, and subsequent increases in organism activity (Wallace et al. 1991), 
may result in greater nighttime suspension (Fig. 1.4).  Furthermore, in this study, 
seston and POC concentrations increased after twilight without a lag (Fig. 1.3), which 
matches the activity patterns of nocturnal aquatic organisms (Schloss and Haney 
2006).  For example, mayfly and stonefly larvae are most active from sundown to 
sunrise to minimize risk of fish predation (Cowan and Peckarsky 1994; Elliott 2000).  
While I cannot exclude the activities of aquatic or terrestrial mammals, such as deer 
or raccoons, as a mechanism for the diel patterns, I suspect that those organisms 
 
 23 
 
would generate intermittent pulses of seston resulting from isolated movements, such 
as fishing for food, drinking, or crossing the stream.  Rather than intermittent pulses, 
nighttime increases of seston and POC concentrations appear to be maintained 
throughout the night (e.g. Fig. 1.3).   
 
Furthermore, Pheo a concentrations, which indicate dead or senescing algae, 
increased during the nighttime (Fig. 1.6b); it is feasible that these algae are more 
susceptible to be suspended into the water column through bioturbation or nighttime 
sloughing.  I did not see concomitant increases in suspended Chl a which would have 
been evidence for excess production or grazing.   
 
Bioturbation is known to affect ecosystem function in marine, freshwater, and 
terrestrial ecosystems (Palmer et al. 1997; Meysman et al. 2006; Moore 2006).  
Studies, within freshwater aquatic ecosystems of various biomes, have identified 
nocturnal species of macroinvertebrates and amphibians as drivers of bioturbation 
(Zanetell and Peckarsky 1996; Statzner et al. 2003).  Predaceous stoneflies increased 
sediment removal from the streambed while searching the rocks and sediments for 
prey (Zanetell and Peckarsky 1996).  Crayfish and amphibians, such as tadpoles, 
suspend particles through predator avoidance, conspecific aggression, and feeding at 
night (Creed and Reed 2004; Ranvestel et al. 2004; Usio and Townsend 2004), but 
spend much of the day in burrows or under rocks and other refugia (Petranka 1984; 
Statzner et al. 2000).  Key fish species, in both tropical and Alaskan streams, cause 
appreciable sediment and POC suspension and downstream export through 
 
 24 
 
movement, migration, spawning, feeding, and egestion (Moore et al. 2004; Taylor et 
al. 2006).   
 
In WCC, the nocturnal community that could contribute to bioturbation of the 
stream sediments includes 104 species of mayflies (Ephemeroptera), 45 species of 
stoneflies (Plecoptera) (J. K. Jackson unpubl.), two species of crayfish (D. A. Lieb 
pers. comm.), and three species of salamanders (E. H. Grant unpubl.).  Fish probably 
do not drive the diel pattern because, of all 20 fish species in WCC, only eels are 
nocturnal (R. J. Horwitz pers. comm.).  Furthermore, the diel patterns occurred 
throughout the watershed, including several 2
nd
 order streams (<10 cm deep) that 
would exclude large fish such as trout or sunfish.   
 
Seasonal patterns ? Warm season peaks of POC and seston concentrations have been 
attributed to increased macroinvertebrate fragmentation of leaf litter in other 
deciduous forest streams in North Carolina (Webster et al. 1990; Wallace et al. 1991) 
and Spain (Molla et al. 2006).  Similarly, I observed peak concentrations during the 
warm seasons in addition to the consistently observed diel patterns (Fig. 1.2a, b).  The 
peaks, in spring and summer, can be explained by stream organism activity, such as 
feeding and bioturbation, because maximum macroinvertebrate activity and biomass 
occurs during the warm seasons (Wallace et al. 1991).  Bioturbation likely helps 
control seasonal patterns because the organic content of seston decreased during the 
warm months (Fig. 1.2c), indicative of bioturbation rather than feeding behaviors 
(sloppy feeding, egestion).  Filter feeding macroinvertebrates could decrease the 
 
 25 
 
carbon content of seston by selectively feeding on high organic content particles 
(Parkes et al. 2004), but are likely not responsible for ecosystem-level decreases in 
POC content.  Monaghan et al. (2001) showed that filter feeders are only responsible 
for 11% of particulate organic matter removal from the water column in an Idaho 
stream.  
 
Seasonal changes within the deciduous riparian vegetation appeared to drive 
seasonal patterns of seston quality.  In late October and early November, WCC 
receives 40-50% of annual organic matter fall-in (313 g m
-2
 yr
-1
, Newbold et al. 
1997), while in late winter and early spring, the open canopy, increasing hours of 
daylight and warming temperatures are conducive to the growth of long algal 
filaments on rocks and sediments (Kaplan and Bott 1982).  The two seasonal peaks in 
carbon content coincided with autumnal leaf litter inputs and vernal algal blooms 
(Fig. 1.2c).  Despite the seasonal changes in quality, the overall carbon content (9% to 
15%) was still low compared to organic-core material like leaves, wood or algae 
(~45%), but higher than seston in large rivers with high sediment loads (~1-3%; 
Hedges et al. 1986; Onstad et al. 2000).  Seston is likely a mixture of these two types 
of particles (organic-core and mineral-core, Meybeck 1982), with peaks in organic-
core particles occurring in autumn and spring (Table 1.1).  Mineral-core particles 
comprised most of total seston mass (Table 1.1); microscopic analysis showed that 
most seston particles are clays or empty diatom frustules (Fig. 1.9).  However, 
organic-core particles, despite being a relatively small fraction of total seston mass 
 
 26 
 
and rarely seen in microscopic analysis (Fig. 1.9), still represent a large fraction of 
carbon flux because of their high carbon content (Table 1.1).      
 
Autochthonous sources provide a fraction of organic-core particles in seston; 
it is necessary to convert pigment concentration to algal carbon to determine the 
proportion of POC from autochthonous carbon.  If a fixed C:Chl a ratio of 30 (Smith 
1980) is assumed, algal carbon would vary between 70% of POC during early spring 
and 30% during the summer.  However, C:Chl a ratios vary with growth rate, 
irradiance, temperature, and nutrient concentration (Cloern et al. 1995) and 
pheopigments, degradation products of Chl a including Pheo a, are not included in 
these calculations.  This is problematic because pheopigments represent algal-derived 
carbon from old, dying cells sloughed from epilithon (Peterson and Stevenson 1992), 
or are otherwise suspended by stream organism activity including bioturbation, 
herbivory, and filter-feeding egestion (Downs and Lorenzen 1985).  Microscopic 
examination of WCC seston revealed that, of the few particles that were algal cells, 
most were individual diatoms and relatively few of these had retained their 
chloroplasts (Fig. 1.9).  Thus, two types of algal cells appear to exist in the suspended 
load: fresh algal cells or older, senesced and grazed cells which likely have different 
algal carbon:pigment ratios (Downs and Lorenzen 1985).  Therefore, algal 
contributions to POC may exceed the preliminary estimate of 30-70%.  More work is 
needed to accurately estimate the contribution of algal carbon to downstream POC 
transport by determining the temporal variation of C:pigment ratios in seston and 
epilithon on the streambed.   
 
 27 
 
  
Studies from large rivers (Kendall et al. 2001) and streams (Walters et al. 
2007) are consistent with observations that POC is isotopically enriched in spring and 
depleted in the late fall months (Fig. 1.7).  Autumnal POC-?
13
C was even depleted 
relative to whole deciduous leaves (-28.2?, Fig. 1.7) suggesting that there was a 
selective preservation of lignin (~-30?) relative to proteins and cellulose (~-27?) 
(Benner et al. 1987).  As much as 25% of leaf carbon leaches within the initial 24 
hours of wetting (Lush and Hynes 1973) with lignin remaining as a considerable 
portion of leaf carbon available for physical and biological fragmentation (Benner et 
al. 1987).  The spring samples reflect a greater ?
13
C epilithon signal (Fig. 1.7).  
However, algal ?
13
C is known to vary spatially as a function of the concentration of 
dissolved free CO
2
, the ?
13
C of dissolved inorganic carbon, water-velocity effects on 
cell-wall CO
2
 gradients, and primary production rates, which in turn depend on 
position within the watershed relative to carbonate lithologies and on reach 
morphology (Finlay et al. 1999; Walters et al. 2007).  In this study, I sampled an 
autochthonous source from only one riffle at one time.  Temporally and spatially 
explicit samples of epilithon are necessary to determine the algal ?
13
C contributions 
throughout the WCC watershed.  
  
Hydrologic patterns ? Most POC and seston export from WCC occurs during storms 
(Newbold et al. 1997), as is typical of small streams (Golladay et al. 2000).  However, 
during storms, seston is transported long distances downstream (Mosisch and Bunn 
1997) and carbon content decreases, lowering POC quantity and quality available for 
 
 28 
 
local biological use.  Baseflow transport, though comprising a smaller proportion of 
the total load, delivers upstream POC to relatively close reaches downstream 
(Thomas et al. 2001).  Furthermore, seston and POC concentrations were not 
correlated with baseflow discharge.  From measurements of particle deposition 
velocities (0.1 to 0.4 mm s
-1
; Thomas et al. 2001), WCC mean baseflow water 
velocity (0.12 m s
-1
), and depth (0.10 m), I can calculate that, at baseflow, particles 
travel between 31 to 124 meters on average before they are deposited on the 
streambed.  Furthermore, WCC is at baseflow 90% of the time (Newbold et al. 1997), 
when most heterotrophic respiration occurs.  Therefore, I focused on baseflow 
transport of seston and POC.      
 
Hydrology can still influence baseflow seston concentrations.  A large storm 
or series of successive storms can deplete the seston supply by removing readily 
suspended particles from the streambed, and clearing rocks and sand of epilithon 
(Carling 1983; Van Sickle and Beschta 1983).  As a result, POC concentrations are 
higher during baseflow periods following infrequent storms rather than frequent 
scouring floods (Pozo et al. 1994).  In this study, the magnitude, duration and time 
since the antecedent storm could not adequately explain the baseflow concentrations 
or characteristics of seston.  However, the time since the antecedent storm did 
influence the diel variation in seston suggesting that large antecedent storms reduced 
the amount of seston available for nighttime increases in concentration in the days 
following the storm (Fig. 1.5). 
 
 
 29 
 
Implications for energy flow ? In a tropical stream, bioturbation increased 
downstream POC flux by 60% (Taylor et al. 2006), similar to the observation of 45% 
increases in nighttime POC fluxes in WCC (Fig. 1.2d).  For extrapolation to annual 
POC fluxes, many studies typically use flow-weighted concentrations or rating curves 
derived from baseflow samples (Golladay 1997).  However, I demonstrate here that 
this approach is inappropriate for estimating temporal dynamics or magnitudes of 
POC fluxes at baseflow.  For example, using only daytime POC concentrations in 
WCC, the annual baseflow POC transport fluxes for the sampling period would be 
337 kg POC yr
-1
; however, integrating over day and night concentrations, the annual 
baseflow POC transport flux is 414 kg POC yr
-1
. 
 Without consideration of nighttime 
differences in POC concentration, annual baseflow POC flux would be 
underestimated by 23%.  
 
My quantitative and qualitative geochemical analyses of particles in transport 
suggest that temporal patterns in seston and POC concentration and composition were 
generated by nocturnal bioturbation and seasonal cycles of leaf fall and algal 
production.  These results argue that the traditional ecological perspective of POC as 
fragmented and egested leaf-litter (Vannote et al. 1980) requires modification to 
include mineral-core and algal-derived particles as components of the suspended 
load.  Furthermore, the results detail variability in POC fluxes and quality (i.e. carbon 
content of seston), which controls carbon bioavailability (Baisden et al. 2002) and 
should ultimately influence rates of heterotrophic respiration in downstream rivers 
(Cole and Caraco 2001; Mayorga et al. 2005).  Future work is needed to advance the 
 
 30 
 
understanding of the role of river networks in controlling the transformations of 
allochthonous and autochthonous POC transported from inland waters to the oceans. 
 
 
 31 
 
Tables 
Table 1.1. Mixing model results presented as the percent of seston and POC 
concentration by mass (mean ? SE), where m represents mineral-core and o 
represents organic-core particles, and p<0.05 indicates significant difference between 
high organic matter input months (Nov, Dec, Mar) and the rest of the year from 
paired t-tests (df=75 for each test).    
      
% of 
mass 
Type of 
particle 
Nov, Dec, 
Mar  
Other 
months 
t-test  
p-value 
seston m 70% ? 1.4% 81% ? 0.4% <0.001 
seston o 29% ? 1.4% 18% ? 0.4% <0.001 
POC m 5% ? 0.4% 9% ? 0.2% <0.001 
POC o 94% ? 0.4% 90% ? 0.2% <0.001 
 
  
 
 
 32 
 
Figure Captions 
 
Figure 1.1. Hourly discharge of White Clay Creek over the sampling period.  Arrows 
and circles indicate sampling days (11 Aug 05, 21 Sep 05, 01 Nov 05, 07 Dec 05, 12 
Jan 06, 24 Feb 06, 31 Mar 06, 27 Apr 06, 24 May 06, 20 Jun 06, 26 Jul 06).  Note the 
break in the y-axis scale to account for one large storm in June 2007.  
 
Figure 1.2. Seasonal changes in (a) seston concentration, (b) particulate organic 
carbon concentration, (c) seston carbon content, and (d) particulate organic carbon 
flux for day and night samples. The means were plotted against julian day because of 
uneven intervals between sampling periods, but were identified as the month in which 
they were sampled.  Error bars are standard errors; some cannot be seen when smaller 
than the symbol. 
 
Figure 1.3. Example diel curve from June sampling of seston concentration (left 
vertical axis) and particulate organic carbon (POC) concentration (right vertical axis). 
 
Figure 1.4. Least squares linear regressions between temperature and seston 
concentration for both day (seston = 0.07 Temperature + 0.78, r
2
=0.45, p=0.02) and 
night (seston = 0.18 Temperature + 0.83, r
2
=0.45, p=0.02) samples.   
 
 
 33 
 
Figure 1.5. Least squares linear regressions between the maximum discharge of the 
antecedent storm and the increase of total seston concentration from day to night for 
each sampling month (% Increase = ? 0.09 Max Discharge + 119, r
2
=0.61, p=0.005). 
 
Figure 1.6. Seasonal changes in (a) Chl a and (b) Pheo a concentrations for day and 
night samples.  The means were plotted against julian day because of uneven intervals 
between sampling periods, but were identified as the month in which they were 
sampled.  Error bars are standard errors; some cannot be seen when smaller than the 
symbol. 
 
Figure 1.7. Seasonal changes in ?
13
C of POC averaged across day and night samples. 
The means were plotted against julian day because of uneven intervals between 
sampling periods, but were identified as the month in which they were sampled.  
Error bars are standard errors; some cannot be seen when smaller than the symbol.  
Boxes (mean ? 1 SD) represent the range of ?
13
C of (a) epilithic algae as 
autochthonous and (b) four species of senesced leaf species as allochthonous end-
members. 
 
Figure 1.8. Least squares linear regressions between total pigment concentrations 
(Chl a + Pheo a) and ?
13
C of POC (?
13
C = 0.35 pigments ? 29.2, r
2
=0.49, p=0.015). 
 
 
 34 
 
Figure 1.9. Environmental scanning electron microscope images of seston. (a) Most 
particles were small angular clays and silts with adsorbed bacteria or organic matter, 
while (b) other particles were recognizable as diatoms cells (e.g. Pinnularia).    
4000
1000
500
250
0
16 Jul 05 19 Oct 05 21 Jan 06 26 Apr 06 29 Jul 06
Figures
Figure 1.1
Hourly dischar
ge (L
 s
-1
) 
0.5
7
6
5
4
3
2
1
0
01 Aug 05 02 Oct 05 03 Dec 05 03 Feb 06 06 Apr 06 07 Jun 06 08 Aug 06
Seston (mg L
-1
) 
POC (mg L
-1
) 
Seston carbon 
content (%) 
POC flux  (mg s
-1
) 
0.4
0.3
0.2
0.1
0
20
15
10
5
20
15
10
5
0
25
30
Night 
Day
a
b
c
d
Aug
Sep
Nov Dec
Jan
Feb
Mar
Apr
May
Jun
Jul
Figure 1.2
20:00 h 2:00 h 8:00 h 14:00 h 20:00 h
0
3
Seston (mg L
-1
) POC (mg L
-1
)
0.5
0.3
0.1
0.7
0.9Night
Seston
POC
6
9
21 Jun 06 22 Jun 06
Figure 1.3
7
6
5
4
3
2
1
0
Seston (mg L
-1
) 
630 9 12 15 18 21
Temperature (?C)
Night
Day
Figure 1.4
160
120
Diel increase of seston (%)
80
40
0
0 200 400 600 800 1000
Antecedent storm max discharge (L s
-1
)
Figure 1.5
02 Oct 05 03 Dec 05 03 Feb 06 06 Apr 06 07 Jun 06 08 Aug 06
3
Night 
Day
a
b
Chl 
a
 (
?
g L
-1
) 
01 Aug 05
Pheo 
a
 (
?
g L
-1
) 
2
1
0
7
6
5
4
3
2
1
0
Aug
Sep
Nov
Dec
Jan
Feb Mar
Apr
May Jun
Jul
Figure 1.6
a
b
algae
leaves
-25
01 Aug 05 02 Oct 05 03 Dec 05 03 Feb 06 06 Apr 06 07 Jun 06 08 Aug 06
?
13
C
-26
-27
-28
-29
-30
Aug Sep Nov Dec Jan Feb Mar Apr May Jun Jul
Figure 1.7
-25
?
13
C
-26
-27
-28
-29
-30
0 2 4 6 8
Pigments (?g L
-1
)
Figure 1.8
1 ?m 10 ?m
Figure 1.9
 
 44 
 
Chapter 2: Contributions of salamanders and macroinvertebrates 
to seston export in artificial stream ecosystems 
 
Introduction 
Particulate organic carbon (POC) provides energy for microorganisms and 
macroinvertebrates in headwater streams (Hedin 1990; Mulholland et al. 2000), and 
when transported in the suspended load (as seston), supports heterotrophic 
metabolism in downstream reaches and rivers (Wipfli and Gregovich 2002; Cole and 
Caraco 2001).  Stream organisms can control the movement of seston through 
retention by filtration (Wotton et al. 1996), generation of particles by sloppy feeding 
and egestion (Wallace et al. 1991), or bioturbation that suspends particles during nest 
digging, foraging, and movement (Moore 2006).  While the ecosystem-level effects 
of bioturbation have been studied extensively in soft-sediment environments of 
oceans, lakes, and rivers (Mermillod-Blondin and Rosenberg 2006; Meysman et al. 
2006; Nogaro et al. 2006), relatively few studies have addressed this issue in streams, 
despite the fact that stream organisms are known to impact bottom sediments (Jones 
et al. 1997; Palmer et al. 1997).  For example, predaceous stonefly larvae can disturb 
streambed sediment while searching for prey (Zanetell and Peckarsky 1996).  
Crayfish and amphibians, particularly tadpoles, may suspend particles through 
predator avoidance, conspecific aggression, and feeding (Creed and Reed 2004; 
Ranvestel et al. 2004; Usio and Townsend 2004).  Through their movement, 
migration, feeding, and spawning activities, key fish species suspend sediment and 
 
 45 
 
POC which is then exported downstream in both tundra and tropical streams (Moore 
et al. 2004, Taylor et al. 2006).   
 
In a southeastern Pennsylvania piedmont stream, daily fluctuations in seston 
transport included 80% increases in nighttime seston concentrations and 43% 
increases in nighttime POC concentrations (Chapter 1).  The macroinvertebrate and 
salamander communities may contribute to the observed nighttime increase of seston 
through bioturbation of bed sediments because of their nocturnal behavior and high 
streambed population densities (Moore 2006).  Macroinvertebrates and salamanders 
both exhibit diel patterns in their activities on the streambed.  Many 
macroinvertebrates are nocturnal, moving to the surface of streambeds to feed on 
detritus, periphyton, and other organisms at night (e.g. Schloss and Haney 2006) and 
often exhibiting peak downstream drift during nighttime hours (Brittain and Eikeland 
1988).  Similarly, many salamanders are most active at night as they search for 
macroinvertebrate prey on the sediment surface (Petranka 1984).  Macroinvertebrates, 
excluding crayfish, are small in body size relative to other stream organisms such as 
fish and amphibians, but their densities exceed those of larger organisms by several 
orders of magnitude, suggesting they could have a larger cumulative effect on seston 
transport.  Previous work has linked amphibian activity with bioturbation of 
streambed sediments (Ranvestel et al. 2004), and salamander larvae can reach 
considerable densities in small streams, where they often replace fish as dominant 
predators (Davic and Welsh 2004).  However, I have not found any work that has 
focused on the effect of salamander feeding and foraging activity on ecosystem level 
 
 46 
 
processes, despite their high diversity, density, and predator position in the food chain 
of small order streams.   
 
Here, I test the hypothesis that bioturbation by salamanders and stream 
macroinvertebrates is a mechanism for nighttime increases in seston transport.  Using 
stream-side flumes, I measured the concentration of seston over four consecutive days 
and nights in both the presence and absence of macroinvertebrates and salamanders.  I 
also measured suspended chlorophyll a to determine if the seston was a byproduct of 
grazing or sloughing from within the autochthonous biofilm (Peterson and Stevenson 
1992).  Throughout the experiment, I describe the percentage of seston export and 
average seston transport distance as metrics for bioturbation and particle suspension 
within the artificial streams.  I show significant daytime increases in seston transport 
in experimental treatments with salamanders.  However, macroinvertebrates did not 
affect seston transport and there were no patterns in suspended chlorophyll a (Chl a) 
concentrations.  
Methods 
Experimental streams ? I conducted a four day experiment in two stream-side flumes 
at the Stroud Water Research Center, in southeastern Pennsylvania in September 
2006.  The two flumes (30m long; 0.3 m wide; 0.04 m water depth; approx. 3 cm s
-1
 
water velocity; surface area = 9 m
2
) were filled with stones (approximately 5.7 cm 
median diameter) that overlaid a sand bed (8 mm median diameter, ?hyporheic zone?) 
(Fig. 2.1a; Battin et al. 2003).  Prior to the experiment, all the stones were power-
washed and the streambed sand was stirred while the flumes were running to remove 
 
 47 
 
particles and macroinvertebrates and to ensure minimal differences in sediment 
accumulation between flumes at the start of the experiment.  The flumes were 
gravity-fed stream water from White Clay Creek (WCC) from one shared header tank 
and run in continuous through-flow mode at ~0.5 L s
-1
 for 21 days before the 
experiment began to allow periphyton to colonize the rocks and seston, from storms 
and baseflow within WCC, to fill interstitial spaces (Battin et al. 2003).   
   
Experimental design ? The experiment was run for four consecutive days with 
different treatments in each flume and during each day (Table 2.1).  The first 
treatment was a control, with no added macroinvertebrates or salamanders to either 
flume to compare the behavior of the two flumes.  For the second treatment, I added 
macroinvertebrates freshly collected from WCC to flume 1 (see macroinvertebrate 
and salamander collection) to examine the effects of the diel behavior of 
macroinvertebrates on particle transport.  For the third treatment, I added a 
macroinvertebrate community to flume 2, without removing any macroinvertebrates 
from flume 1 to ensure both flumes had macroinvertebrates when I added the 
salamanders.  For the fourth and final treatment, I added salamanders (see 
macroinvertebrate and salamander collection) to flume 1 to examine the effect of 
diel behavior of salamanders on particle transport.  After organism additions, 16:00 to 
18:00 on the afternoon preceding nighttime sampling, salamanders and 
macroinvertebrates could be observed moving on and between the rocks.  Between 4-
6 hours passed before I began sampling for each treatment.  I collected water samples 
at three time points per night (22:00, 1:00, 4:00) and day (10:00, 13:00 and 16:00) to 
 
 48 
 
measure seston and suspended Chl a (see below for sampling procedure).  The 
sampling design allows for two experimental comparisons of day/night and treatment 
differences (Table 2.1); thus, hereafter Experiment A includes the initial 48 hours 
(treatments 1 and 2) and Experiment B the final 48 hours of manipulations 
(treatments 3 and 4). 
     
Macroinvertebrate and salamander collection - WCC drains a temperate deciduous 
basin (7.3 km
2
) of agricultural and forested piedmont land and harbors an abundant 
biotic community (Newbold et al. 1997).  For each of the two macroinvertebrate 
additions (treatments 2 and 3), I collected macroinvertebrates from Surber samples 
(50 x 50 cm opening, 85 cm net length, 250 ?m mesh size) in 3 WCC riffles with a 
total sample area approximately equal to half of the bed area of a flume (4.5 m
2
).  I 
supplemented the Surber samples with 18+ D-net jab samples collected along the 
stream banks, debris, and pools.  The Surber and D-net samplers are two standard 
collecting devices for stream macroinvertebrates involving disturbing the bottom 
sediments and catching organisms in a net held downstream.  Surber samplers are 
used for quantitative analysis of macroinvertebrate density using a square sampling 
frame to border the sampling area.  For each of the two macroinvertebrate additions 
(treatments 2 and 3), I added ~17,000 individuals for a final density of 1800 
individuals m
-2
.  Families that appeared the most frequently included water pennies 
(Coleoptera: Psephenidae), riffle beetles (Coleoptera: Elmidae), chironomids 
(Diptera: Chironomidae), crawling mayfly larvae (Ephemeroptera: Ephemerellidae), 
clinging mayfly larvae (Ephemeroptera: Heptageniidae), portable-case caddisfly 
 
 49 
 
larvae (Trichoptera: Brachycentridae) and net-spinning caddisfly larvae (Trichoptera: 
Philopotamidae; Trichoptera: Hydropsychidae).  
 
Macroinvertebrates from all samples were held for a maximum of 8 hours at 
stream water temperature before they were added to the flumes.  I used a swirl and 
decant technique (Palmer et al. 2006) to collect organisms from a 5 gallon bucket and 
then distribute them at 3 meter intervals along the length of each flume.  Crayfish 
were excluded from the macroinvertebrate additions because of their size and to limit 
the experiment to testing differences between salamander and small 
macroinvertebrate additions.  Flume outflow passed continuously through a 250 ?m 
sieve to collect drifting macroinvertebrates and salamanders.   
 
For the salamander addition treatments, two-lined salamander larvae (Fig. 
2.1b: Eurycea bislineata) were collected by turning over rocks and wood within 
several 2
nd
 order tributaries to WCC and in the 3
rd
 order stream itself.  Salamanders 
were placed individually into bags filled with stream water and held in coolers 
streamside for <8 hours.  At sunset following treatment 3, I released 89 salamanders 
at intervals of 0.1 m from the first meter to the 29
th
 meter of flume 1 for a final 
salamander density of 9.9 individuals m
-2
.  I added realistic densities of two-lined 
salamander larvae to the artificial flumes; similar salamander species such as the Blue 
Ridge Two-lined salamander (Eurycea wilderae) have been shown to have densities 
between 8.6 and 10.1 individuals m
-2
 (Davic and Welsh 2004).     
 
 
 50 
 
Seston ? At each sampling time, water samples were taken from the inflow to the 
flumes (header tank outflow) and the outflows of flume 1 and flume 2.  In the 
laboratory, seston concentration was determined gravimetrically using the dry weight 
of total suspended solids.  Aliquots of stream water were filtered through stacked 
pairs of pre-weighed and pre-ashed (480?C for 4.5 hours) glass fiber filters (Millipore 
AP40 with nominal pore-size of 0.7 ?m).  The sample was filtered through two filters 
to check analytical accuracy, assess the integrity of the top filter and ensure even 
distribution of seston across the filter.  The sample was discarded if the bottom filter 
gained or lost more than 0.2 mg.  The filters were dried for 12-18 hours at 60?C and 
re-weighed to calculate seston concentration (mg seston L
-1
).   
 
Seston outflow concentration (S
out
) was divided by inflow concentration (S
in
) 
to determine the percentage of seston (P
out
) exported from each flume (eq. 2.1).  The 
net seston transport distance (S
w
), the average length of travel before deposition, was 
calculated using S
in
, S
out
 and the length (L) of the flume (eq. 2.2; Battin et al. 2003).   
 
Equation 2.1   P
out
 = 100 x (S
out
 / S
in
) 
 
Equation 2.2   S
w
 = L / ln(S
in
 / S
out
 ) 
 
Pigments ? Measured volumes of water for suspended Chl a concentrations were 
filtered through a glass fiber filter (Millipore AP40).  The filters were frozen and 
processed within 30 days using a modified hot ethanol extraction method (Biggs and 
 
 51 
 
Kilroy 2000).  The samples were heated to 78?C for 10 minutes in 90% ethanol and 
then stored frozen for 24 hours.  Chl a concentrations (?g L
-1
) were calculated using 
measurements of absorbance on a spectrophotometer (Beckman Coulter DU-640) 
before and after acidification (Lorenzen 1967; Biggs and Kilroy 2000).   
 
Statistical analysis ? All statistical analyses were conducted using SAS v9.1 (Level 
1M3, 2007, SAS Institute Inc., Cary, NC, USA).  I analyzed the data as two separate 
experiments (Table 2.1) testing the effects of macroinvertebrate additions in 
experiment A and the effect of salamander additions in experiment B.  For each 
experiment, I performed BACI three-way ANOVAs (proc GLM) for seston outflow 
concentration, percent seston exported, transport distance, Chl a outflow 
concentration and percent Chl a exported with day or night, treatment and flume as 
fixed effects (Skalski and McKenzie 1982; Smith 2002).  Three- and two-way 
interaction terms were removed from the model if the effects were non-significant.  I 
acknowledge limitations in the statistical analysis due to potential serial auto-
correlation and lack of independence between samples, treatments, and experiments 
(see Smith et al. 1993).  However, the BACI model is most effective if the variance in 
the difference between flumes does not change over sampling time (Smith et al. 1993) 
and I only saw random fluctuations in the variance. 
Results 
Macroinvertebrate and salamander additions ? Within 3 hours of each addition, I 
captured a small number of macroinvertebrates (<50) in the outflow, but added those 
immediately back to the top of the flume.  In the flumes without macroinvertebrates 
 
 52 
 
additions, I captured no individuals in the outflow.  After the initial drift, fewer than 4 
individuals hour
-1
 were captured in the flume outflow.  Throughout the course of the 
experiment, no salamanders were recovered in the flume effluent and following the 
experiment, I recovered >95% of the salamanders during a thorough search of the 
flume.  
 
Flume conditions ? The inflow seston concentration was greater at night than day 
during both experiments (Exp 1: F
1,20
=8.1, p=0.0099; Exp 2: F
1,20
=23.6, p<0.0001) 
and there was a trend of decreasing inflow concentrations throughout experiments A 
and B (Fig. 2.2a).  WCC experienced storm flows  36 hours before the start of the 
treatment 1 and discharge within the creek slowly decreased from 80 L s
-1
 to 65 L s
-1
 
during the four day experiment.  Inflow seston concentrations were lowest in the day 
during treatment 4 and highest in the night during treatment 2 (Fig. 2.2a).  Flume 1 
discharge was constant (0.4 L s
-1
); however, flume 2 discharge declined by 40% 
during the last two treatments (0.5 L s
-1
 to 0.3 L s
-1
) (Fig. 2.3).   
 
Seston, experiment A (treatments 1 and 2) ? Nighttime seston outflow concentration 
(0.68 mg L
-1
 ? 0.02: mean ? SE) exceeded daytime seston outflow concentration 
(0.46 mg L
-1
 ? 0.02: mean ? SE) by 48% (Fig. 2.2b; F
1,20
=38, p<0.0001) following 
the diel patterns of the flume inflow concentrations.  However, there were no other 
significant effects for seston outflow concentration, % seston exported or transport 
distances when considering diel macroinvertebrate activity (Fig. 2.2b, c, d).   
 
 
 53 
 
Seston, experiment B (treatments 3 and 4) ? All three-way interactions were non-
significant and removed from the ANOVA model; I examined 2-way interactions and 
main effects (flume, day/night, treatment 3 vs. 4) for seston outflow concentration, 
percent seston exported, and transport distances.  Seston outflow concentration (Fig. 
2.2b) treatment effects are superimposed on inflow concentration effects resulting in 
significant treatment by flume (F
1,17
=8.1, p=0.01), day/night by flume (F
1,17
=4.9, 
p=0.04), and treatment by day/night (F
1,17
=10.0, p=0.0006) interaction effects.  
However, because the inflow concentration had a diel pattern with nighttime peaks, 
examining metrics that correct for inflow concentration (seston transport distance and 
percent seston exported) was a more effective approach when examining the diel 
effects of salamander additions on seston transport.  Both seston transport distances 
and percent seston exported still had significant treatment by flume (Fig. 2.4a, b) and 
treatment by day/night (Fig. 2.4c, d) interactions.  These interaction effects were 
driven by a 44% and 30% increase in percent seston exported (29.3% ? 2.2: mean ? 
SE) and S
w
 (24.5 m ? 1.5: mean ? SE), respectively, during the daytime results of the 
salamander addition (Fig. 2.2c, d).     
 
Chlorophyll a ? Chl a outflow concentration (Fig. 2.5a; 0.79 ?g L
-1
 ? 0.05: mean ? 
SE) was more variable than seston data, but with few clear patterns throughout 
experiments A and B.  Chl a only had a significant main effect in experiment B when 
daytime outflow concentration was greater than nighttime concentration (F
1,20
=4.9, 
p=0.04), but there were neither treatment nor flume effects.  On average, the flumes 
 
 54 
 
exported 4% more suspended Chl a than they received (Fig. 2.5b) with percentage of 
Chl a export averaging 104 % ? 8 (mean ? SE).   
Discussion 
Salamander effect ? Salamanders did not increase the nighttime export of seston as I 
predicted, but did increase daytime export of seston from the artificial flumes (Fig 2c, 
2d).  Salamander activity is likely to result in the suspension of particles because they 
are large stream organisms, cover considerable areas in their movement and actively 
search for macroinvertebrate prey (Petranka 1984).  However, the 9% increase in 
seston export was only a fraction of the nighttime increase in seston concentration 
(80%) that was seen in WCC (Chapter 1) or in the inflow concentration during the 
day and night of treatment 4 of this experiment (Fig. 2.2a: 39% nighttime increases of 
seston concentration).   
 
The increase in seston transport due to salamander additions during the day in 
treatment 4 was surprising given nocturnal salamander activity patterns measured in 
other studies (Petranka 1984).  Salamanders might have increased daytime activity 
because they were released from visual stimulus of predatory fish.  Stream and pond 
macroinvertebrates and amphibians will be nocturnal in the presence of predatory fish 
and diurnal when released from this visual pressure (Flecker 1992; Hampton and 
Duggan 2003).  Alternatively, it is possible I did not capture normal nocturnal 
salamander activity because the salamanders were added to the flume only four to 
five hours before the first nighttime measurement.  After experiencing handling stress 
for up to 8 hours and removal from the stream (translocation), the salamanders may 
 
 55 
 
not have resumed typical behavior (Matthews 2003; Artigas et al. 2005).  
Salamanders could have been recovering from handling during the daytime and been 
digging under rocks and in the sediment to establish refugia or search for prey.  
Allowing more time for salamanders to recover might have been needed for a realistic 
reproduction of normal diel behavior patterns.   
 
Macroinvertebrate effect ? I added ~1800 individuals m
-2
 to the flumes, densities that 
are comparable to lower ranges of densities of macroinvertebrates found in 
northeastern deciduous streams and WCC (Kratzer et al. 2006; J.K. Jackson pers. 
comm.).  I added a diverse array of functional feeding groups representative of the 
stream community including predators, scrapers, shredders, collectors-filterers, and 
collector-gatherers (Merritt and Cummins 1995).  Despite the addition of a diverse 
macroinvertebrate community and particles available for suspension to the flumes, I 
found no significant effect of macroinvertebrates on seston concentration.  The only 
small macroinvertebrates (i.e. excluding shrimp and crayfish) that have been 
documented to cause particle suspension in stream ecosystems are large predatory 
stonefly larvae that actively search for their prey on stream sediments (Zanetell and 
Peckarsky 1996).  The additions included very few of these macroinvertebrates (0.1% 
of all individuals).  Furthermore, in the study conducted by Zanetell and Peckarsky 
(1996), the stonefly larvae were examined in small in-stream experiments designed to 
maximize bioturbation effects.  Therefore, given the low density of predatory 
macroinvertebrates that may have contributed to bioturbation, macroinvertebrates 
 
 56 
 
additions did not increase seston concentration, export or transport distances (Fig. 
2.2b, c, d: Experiment A).        
 
Non-treatment effects ? Other than the treatment effects, inflow concentrations of 
seston, flow velocity and discharge can also affect seston transport within the 
artificial streams.  I saw patterns in the inflow concentrations that reflect the diel 
patterns of seston in WCC (Fig. 2.2a; Chapter 1).  The inflow concentrations, in turn, 
caused diel variation in the seston outflow concentration (Fig. 2.2b).  The inflow 
concentrations also declined throughout the experiment (Fig. 2.2a) due to decreasing 
discharge in WCC following a storm ~36 hours prior to the beginning of treatment 1.  
The outflow seston concentrations mirror inflow concentrations resulting in nighttime 
peaks and overall decline in outflow concentration in both experiments A and B (Fig. 
2.2b).  However, I was able to correct for inflow effects by using the percent seston 
exported and transport distance metrics.   
 
The flow within each artificial stream could also affect the deposition and 
outflow concentrations.  For example, seston transport distances were closely 
correlated with discharge throughout a range of different stream sizes (Thomas et al. 
2001; Paul and Hall 2002).  Unfortunately, the flume discharge decreased throughout 
experiment B in flume 2 (Fig. 2.3).  If the discharge in flume 2 had remained 
constant, the percent of seston exported and S
w
 for flume 2 might have been higher, 
especially in treatments 3 and 4.  However, there seemed to be little effect of flow 
because the outflow concentration of seston of flume 2 continued to match flume 1 
 
 57 
 
(Fig. 2.2b).  There was only real divergence between measurements in the two flumes 
in the daytime of treatment 4 (Fig. 2.2b).  Regardless, the BACI statistical design 
cannot account for external effects other than treatment effects (Smith et al. 1993), so 
the statistical results should be viewed with caution.        
 
Chlorophyll a ? I used Chl a as an indicator of algae in the suspended load (Elliott et 
al. 2004).  In previous studies in these flumes, the biofilm grew for ~25 days before 
the algal biomass was limited by biofilm detachment or grazing (Battin et al. 2003).  
In both experiment A and B, Chl a outflow often exceeded Chl a inflow (Fig. 2.5b) 
indicating production and senescence of algal particulate matter within the flumes.  
Chl a export did not increase even during the daytime of treatment 4 (Fig. 2.5b) 
indicating that the additional particles from salamander bioturbation were more likely 
derived from stream sediments than grazed or sloughed periphyton.    
 
Directions for future research ? Although previous work has demonstrated amphibian 
bioturbation (Ranvestel et al. 2004), this is the first effort to demonstrate stream 
salamander contribution to ecosystem level processes such as modification of seston 
and organic carbon flow.  Stream-associated species account for approximately 35% 
of total amphibian diversity in the United States (Davic and Welsh 2004), and up to 
45% of all salamander species rely on streams for part or all of their life cycle 
(Lannoo 2005).  Despite the significance of stream-associated amphibians (Davic and 
Welsh 2004; Lannoo 2005), knowledge of their ecosystem-level importance is 
 
 58 
 
lacking (Corn et al. 2003).  Their activity, population density, and size make them 
ideal candidates for influencing transport of seston and should be considered further. 
 
I could modify future experiments to test whether results from salamander and 
macroinvertebrate additions were artifacts of experimental design or true effects.  I 
could improve the statistical power by using replicated flumes for each treatment, 
allow the flumes to equilibrate for 6 months or more (inflow concentration = outflow 
concentration; J. D. Newbold pers. comm.), and maintain constant discharge 
throughout the experiment.  Furthermore, I could add more macroinvertebrates, to 
better mimic the higher densities measured in WCC (up to 25,000 individuals m
-2
; J. 
K. Jackson pers. comm.).  The 10-fold increase in individuals would include more 
relevant individuals (i.e. mobile predators; Zanetell and Peckarsky 1996), although at 
similar proportions to the community in this study.  Also, it is probably necessary to 
allow the salamanders more time to acclimate to the artificial streams and recover 
from handling stress and translocation.  Finally, other nocturnal stream organisms 
might contribute to the bioturbation of the sediments.  For example, crayfish suspend 
particles from the streambed through movement related activity, including predation, 
conspecific aggression, and feeding activities (Creed and Reed 2004, Usio and 
Townsend 2004).  Including crayfish additions in the experiment might allow me to 
further understand mechanisms for nighttime increases in seston transport occurring 
in WCC.          
 
 59 
 
Tables 
Table 2.1. Experiment design for the four day experiment.  ?Macroinverts? represents 
the macroinvertebrate community excluding crayfish. 
 
 
Experiment A Experiment B 
  
Treatment 1 Treatment 2 Treatment 3 Treatment 4 
 
Night Day Night Day Night Day Night Day 
Flume 1 
No Macroinverts    
No Salamanders 
Macroinverts      
No Salamanders 
Macroinverts       
No Salamanders 
Macroinverts       
Salamanders 
Flume 2 
No Macroinverts    
No Salamanders 
No Macroinverts   
No Salamanders 
Macroinverts       
No Salamanders 
Macroinverts       
No Salamanders 
 
 
 60 
 
Figure Captions 
 
Figure 2.1. Artificial streams and salamander additions. (a) Evan Grant adding 
salamanders into the artificial streams. (b) Salamander (Eurycea bislineata) following 
addition to artificial stream. 
 
Figure 2.2. Seston inflow and outflow for Experiments A and B.  (a) Inflow seston 
concentration from header tank and (b) outflow seston concentration, (c) % seston 
exported and (d) transport distance (S
w
) in flume 1 and flume 2.  
 
Figure 2.3. Discharge over the course of the experiment for flumes 1 and 2.  
Discharge was measured prior to each treatment and once following treatment 4.   
 
Figure 2.4. Interaction effects for Experiment B: (a) treatment by flume interaction 
for Sw (F
1,17
=4.9, p=0.04), (b) treatment by flume interaction for % seston exported 
(F
1,17
=4.9, p=0.04), (c) treatment by day or night interaction for Sw (F
1,17
=4.4, 
p=0.05), and (d) treatment by day or night interaction for % seston exported 
(F
1,17
=4.4, p=0.05). 
 
Figure 2.5. Chl a for experiments A and B. (a) Chl a outflow concentrations and (b) 
% Chl a exported. 
 
a
b
Figures
Figure 2.1
a
b
c
d
A B
Figure 2.2
Figure 2.3
a
b
c
d
Figure 2.4
?
a
a
a
b
A B
Figure 2.5
 
 66 
 
Chapter 3: Modeling sources and seasonal dynamics of seston 
and suspended particulate organic carbon within a stream 
ecosystem 
 
Introduction 
Particulate organic carbon (POC) provides energy for microorganisms and 
macroinvertebrates in headwater streams (Hedin 1990), and when transported in the 
suspended load (seston), supports heterotrophic metabolism in downstream reaches 
and rivers (Cole and Caraco 2001; Wipfli and Gregovich 2002).  POC transport 
represents a longitudinal linkage between upstream autochthonous and allochthonous 
production and downstream ecosystems with the magnitude of connection controlled 
by the quality, quantity and timing of POC delivery (Mayorga et al. 2005).  
Knowledge of the sources of POC and temporal patterns of POC transport can help 
evaluate the quality of the energetic subsidy to downstream biota (Mayorga et al. 
2005), controls on transport (Pozo et al. 1994), and lead to better understanding of the 
metabolic state of freshwater ecosystems (Cole et al. 2007).   
 
Potential sources of seston include organic-core particles such as algal or 
bacterial cells, leaf or wood fragments, and dissolved organic carbon flocculates or 
mineral-core particles such as silts or clays with adsorbed organic carbon.  Seston in 
small streams has historically been considered to be primarily fragments of leaves and 
wood created by macroinvertebrate feeding and egestion upstream (Vannote et al. 
1980).  Stable isotope and chlorophyll a analyses support the idea that stream POC is 
 
 67 
 
terrestrial with minimal contributions from autochthonous production (Finlay et al. 
2002; Kaiser et al. 2004).  However, elemental, radioisotope and microscopic 
analyses showed that most seston in streams and rivers is composed of small (<20 
?m) particles with high mineral content and elemental compositions closer to algae 
(C:N<10) than leaves or wood (C:N>30) (Meybeck 1982; Cole and Caraco 2001; 
Dodds and Whiles 2004).  These seemingly contradictory results could be reconciled 
if seston were a mixture of algal cells, fragmented leaves or woody debris, and 
mineral-core particles (Aufdenkampe et al. 2001); especially given that POC 
associated with mineral-core particles has a C:N ratio lower than its terrestrial source 
because of preferential adsorption of high nitrogen content amino acids (C:N<10; 
Hedges et al. 1986; Aufdenkampe et al. 2001).  The composition and source of seston 
could influence the array of POC quality (lability) that exists within seston.  For 
example, POC quality of different particle types could range from labile (e.g. fresh 
algal cells) to semi-labile (e.g. leaf detritus) to refractory (e.g. humified soil organic 
matter).  Understanding the diversity of POC quality by identifying its source and 
composition could help reveal the magnitude of the link between headwater streams 
and downstream aquatic ecosystems (Thorp et al. 2006).   
 
In addition to POC quality, hydrological and biological dynamics regulate the 
longitudinal linkage of POC by controlling the timing and distance of seston 
transport.  Storms are the most important factor regulating variability in the 
magnitude of annual or monthly seston and POC fluxes (Golladay et al. 2000), and 
transport particles long distances downstream during high flow events (Verhoff et al. 
 
 68 
 
1979).  However, because small streams are under storm conditions for only a small 
fraction of time (4% for Appalachian mountain streams: Buffam et al. 2001), 
temporal POC dynamics under baseflow conditions may play a disproportionately 
large role in regulating stream ecosystem processes.  At baseflow, riparian vegetation 
and aquatic organisms may regulate seasonal patterns in seston transport more than 
hydrological patterns.  Seasonally, the composition of seston at baseflow can vary 
with autumnal peaks in leaf litter inputs from deciduous riparian vegetation or vernal 
increases of algal production (Chapter 1).  Webster et al. (1990) and Molla et al. 
(2006) attributed baseflow peaks of POC during warmer months to increased 
macroinvertebrate fragmentation of leaf litter as community biomass and activity 
increased.  However, these studies did not verify that seasonal increases in seston 
concentration resulted from increased concentrations of leaf fragments.  In some large 
temperate streams and rivers, autochthonous production dominates seston during 
seasons with high primary production (Pozo et al. 1994; Walters et al. 2007). 
  
 My goals for this study were to (1) identify the sources of seston, (2) 
determine if increased seston concentrations during warmer months result from 
increased processing of leaf detritus, and (3) determine if algal carbon is an important 
component of seston in headwater streams.  I used a mixing model to examine 
baseflow seasonal fluctuations in seston composition and sources from an intensive 
field survey (Chapter 1) and from 14 years of sampling in a 3
rd
-order piedmont stream 
in southeastern Pennsylvania, U.S.A.  The mixing model identified contributions to 
seston concentration from three distinct particle types (algal cells, detrital leaf 
 
 69 
 
fragments and mineral-core particles) using carbon and nitrogen contents of seston.  
When only carbon data were available, seston was identified as one of two distinct 
particle types (organic-core and mineral-core particles).  I also present data from a 
watershed habitat survey of seston, soils, and vegetation and identify potential 
patterns in seston sources from ponds, seeps, riffles and runs.       
 
Methods 
Site description ? The upper east branch of White Clay Creek (WCC: N39?51?33.9?, 
W75?46?59.6?) is a 3
rd
 order watershed draining 7.3 km
2
 of the southeastern 
Pennsylvania Piedmont.  The watershed has 23% of total area as temperate deciduous 
forest, 52% as pastures or hay fields for horses or cattle, and 22% as row crop 
agriculture, but has a largely intact riparian forest and livestock have minimal access 
to the stream (Fig. 3.1; Newbold et al. 1997).  The dominant tree species are 
American beech (Fagus grandifolia Ehrh.), red maple (Acer rubrum L.), northern red 
oak (Quercus rubra L.), black oak (Quercus velutina Lam.), and tulip poplar 
(Liriodendron tulipifera L.) (Wiegner et al. 2005).  Herbaceous vegetation in the 
riparian zone of streams includes Japanese honeysuckle (Lonicera japonica Thunb.), 
multiflora rose (Rosa multiflora Thunb.), knotweed (Polygonum aviculare L.), skunk 
cabbage (Symplocarpus foetidus L.), and jewelweed (Impatiens capensis Meerb.).  
The soils in the WCC watershed are typic hypludults, however, in the riparian zone, 
aquic fragiudults dominate (Wiegner et al. 2005).  The mean daily baseflow discharge 
averaged 122 L s
-1
 over the past five years with a high annual mean discharge of 174 
L s
-1
 and low of 59 L s
-1
.  At baseflow, WCC has high nutrient concentrations (3-5 
 
 70 
 
mg NO
3
-N L
-1
; 4-33 ?g PO
4
-P L
-1
) from agricultural land use and moderate dissolved 
organic carbon (1.5 mg DOC L
-1
) concentrations (Newbold et al. 1997).  Streambed 
sediments are composed of clay-, silt-, and sand-sized particles in runs with gneiss- 
and schist-derived gravel and cobble in riffles (Kaplan et al. 1980).  Most of the 
particulate organic matter input, predominantly leaves, occurs from late October to 
mid-November (J. D. Newbold unpubl.). 
 
WCC 2005-2006 baseflow sampling ? Under baseflow conditions, stream water was 
collected every 1.5 hours during one 24-hour period each month, from August 2005 
to July 2006 using an automated water sampler (Teledyne Isco), with the collection 
tube (9.5 mm inner diameter) at 2/3 depth in the middle of the thalweg (see Chapter 1 
for complete sampling design).  The automated sampler was set to collect four 1-L 
samples at each time period.  The first sample was considered a rinse of the inflow 
tubing and discarded, the second was used for pigment analysis, and the final two 
samples were used for replicate seston and POC analysis.  Only the first replicate of 
seston carbon and nitrogen analysis is reported.  Samples were processed within 9 
hours of collection.   I use carbon (Fig. 1.2b) and pigment (Fig. 1.6) concentrations as 
well as carbon content of seston (%, Fig. 1.2c) from Chapter 1 for modeling and 
analysis of modeling results.       
 
WCC historical baseflow sampling ? From Jan 1993 to Dec 2006, WCC was sampled 
weekly, monthly, or bimonthly at one or two sites by submerging a 5 L sample bottle 
at 2/3 depth in the thalweg of the stream.  Samples were carried to the lab and stored 
 
 71 
 
for <1 week at 4?C before they were processed for seston (see seston analysis 
section).   
 
Watershed habitat site descriptions (2004 and 2007) ? In Aug 2004, three ponds 
outflows (Fig. 3.1: P1, P3, P4) and four seeps were sampled for seston concentration 
and carbon and nitrogen content.  P1 (area = 9409 m
2
) and P4 (pond area = 340 m
2
) 
are ponds in agricultural and residential grass fields with connections to tributaries of 
WCC by pipe outflows.  P3 (area = 631 m
2
) flows to a 2
nd
-order tributary of WCC 
through a small stream.  Seston samples were also taken from four zero-order streams 
(seeps) flowing from groundwater springs that emerge within forests (Fig. 3.1: S1, 
S2, S3, S4).  In Jun 2007, I sampled 2 ponds (P1, P2), 3 seeps (S1, S2, S3), 3 riffles 
(R1, R2, R3) and 3 runs (N1, N2, N3) within the WCC watershed for seston, riparian 
soils from the O-horizon, bank and stream sediments, coarse particulate organic 
matter (CPOM) and herbaceous vegetation (Fig. 3.1).  The two ponds, located in 
agricultural and residential grass fields, P1 and P2 (area = 4808 m
2
), had direct 
outflows to WCC that were approximately 3 L s
-1
 or <10% of the stream discharge.  
Samples from three forested groundwater seeps were collected just downstream of 
where groundwater emerged and formed a small channel (<1 L s
-1
).  Three riffles and 
three runs were randomly selected along a 1 km stretch of the 3
rd
-order branch of 
WCC with discharges ranging from 28 to 39 L s
-1
. 
 
Watershed habitat sample collection and processing ? For both 2004 and 2007 
sampling, triplicate water samples for seston analysis were collected at 2/3 depth in 
 
 72 
 
the water column at each riffle and run site, and directly from each pond outflow.  
Sample water was collected from each seep using a small sampling container, taking 
care to avoid disturbing the seep bed sediments because of the shallow depths (<5 
cm).  For the 2007 watershed habitat sampling, three bank sediment samples were 
collected above water level in the stream channel, and three soil samples were 
collected in the riparian zone within 2-10 meters of the channel.  Both bank sediments 
and soils were collected at each site using a trowel, stored in a plastic bag, dried at 
60-100?C, sifted by hand to remove vegetation, visible roots, and animals, and then 
ground into a fine powder using a mortar and pestle.  At each site, three composites of 
CPOM were collected from in-stream leaf packs and rinsed in the water column to 
remove sediment and stream organisms.  Also, three composites of herbaceous 
vegetation and grasses (herbs) were cut from live plants growing next to the water, 
within the channel and often hanging into the stream.  The vegetation samples 
(CPOM and herbs) were immediately dried at 60-100?C and then ground in a Wiley 
Mill through a 250-?m sieve.  All samples were processed for carbon and nitrogen 
content (%), 
13
C and 
15
N stable isotopic ratios and C:N elemental ratios (see below).  
The three replicates for each measure at each site were averaged together and the 
mean was used for statistical analyses.      
 
Seston analysis ? Seston concentration was determined gravimetrically using the dry 
weight of total suspended solids.  Stream water samples were filtered, after recording 
volume, through stacked pairs of pre-weighed and pre-ashed (480?C for 4.5 hours) 47 
mm glass fiber filters (Millipore AP40 or Whatman GF/F with nominal pore-size of 
 
 73 
 
0.7 ?m) and dried for 12-18 hours at 60?C.  Seston concentration (mg seston L
-1
) was 
calculated by dividing the dry mass of seston collected on the filter by the filtered 
volume.
 
 WCC 2005-2006 baseflow and watershed habitat seston samples for 2004 
and 2007 sampling were processed for carbon and nitrogen content and isotopes (see 
below), while seston samples from WCC historical baseflow were ashed (480?C for 
4.5 hours) and reweighed to determine ash-free dry mass which was converted to 
POC concentration (mg POC L
-1
) by assuming AFDM was 45% carbon and dividing 
by the filtered volume (Webster et al. 1999).   
 
Carbon and nitrogen content and stable isotope analysis ? Subsamples of each seston 
filter, soil, bank, and stream sediment sample were wet with nanopure water and 
fumed with hydrochloric acid vapors for 18 hours, effectively removing carbonates 
(A.K. Aufdenkampe unpubl.).  Subsamples were analyzed in batches on an Elemental 
Analyzer (Costech ECS 4010) interfaced with an Isotope Ratio Mass Spectrometer 
(Thermo DeltaPlus XP).  Standard curves for the elemental and isotopic analysis were 
created for each batch using 8 or 9 isotopically enriched (+37.63?) or depleted (-
26.39?) L-glutamic acid standards over a range of carbon masses that encompassed 
the subsample masses.  Stable carbon isotope ratios, reported using ?
13
C notation, 
were normalized relative to the Vienna Peedee Belemnite standard, while stable 
nitrogen isotope ratios, reported using ?
15
N notation, were normalized relative to air 
N
2
 as the standard.  Carbon and nitrogen content (%) were calculated as the ratio of 
measured carbon or nitrogen mass to dry mass of subsample for all seston, soil, 
sediment and vegetation samples. 
 
 74 
 
 
Two-compartment mixing model ? For the historical WCC baseflow data, because I 
only had carbon content for seston, I modeled seston composition as a mixture of two 
distinct particle types: organic-core and mineral-core particles.  Organic-core particles 
are composed of algal cells, leaf fragments and bacterial cells (carbon content for 
mixture = 34%), while, mineral-core particles are clays and silts with adsorbed 
organic carbon and bacterial cells (carbon content = 1.5%) (Meybeck 1982; Hedges et 
al. 1986).  For each sample, I simultaneously solved equations (3.1) and (3.2): 
 
Equation 3.1  S = P
o
 + P
m
 
             
Equation 3.2  POC = 34% * P
o
 + 1.5% * P
m
 
 
 using measured seston (S) and POC (POC) concentrations to determine the 
concentration of organic-core (P
o
) and mineral-core (P
m
) particles in seston and the 
concentration of organic-core (34% x P
o
) and mineral-core (1.5% x P
m
) particles in 
POC (Chapter 1).  Modeling results were averaged by month. 
 
Three-compartment mixing model ? For the WCC 2005-2006 baseflow, I used carbon 
and nitrogen content (%) to model seston composition as a mixture of three distinct 
particle types: mineral-core (m), terrestrial plant detritus (d) and algal (a) particles.  I 
define mineral-core particles as small clays and silts with adsorbed organic matter and 
microorganisms (mineral-associated POC) (Aufdenkampe et al. 2001) and terrestrial 
 
 75 
 
plant detritus particles as small fragments of deciduous leaves.  I define algal particles 
as individual diatoms because using microscopy, I can clearly identify individual 
diatom cells, but few other products of instream production.  Each end member has a 
distinct combination of carbon and nitrogen contents that were calculated by 
averaging a range of literature values (Table 3.1).  Mineral-core particles and algal 
cells have low C:N ratios, but, mineral-core particles have low carbon and nitrogen 
content while algal particles have higher carbon and nitrogen content.  Leaf detritus 
particles have the highest carbon content, but also the highest C:N ratio resulting in 
low nitrogen content organic matter.       
 For each sample, I simultaneously solved equations (3.3), (3.4), and (3.5) 
 
Equation 3.3  S = P
d
 + P
a
 + P
m
 
         
Equation 3.4  POC = C
d
 * P
d
 + C
a
 * P
a
 + C
m
 * P
m
 
 
Equation 3.5  PN = N
d
 * P
d
 + N
a
 * P
a
 + N
m
 * P
m 
 
using measured seston (S), POC (POC), and particulate nitrogen (PN) concentrations 
carbon contents and nitrogen contents for each end member derived from literature 
values (Table 3.1).  P
x
 represents seston concentration from detrital (d), algal (a) and 
mineral-core (m) particles while C
x
 and N
x
 are the carbon and nitrogen contents (%) 
of each particle type (Table 3.1).  Because of the standard deviation associated with 
the end member means (Table 3.1), I used a Monte Carlo simulation where I solved 
 
 76 
 
equations (3.3), (3.4), and (3.5) for 1000 randomly selected combinations of the end 
members (C
x
 and N
x
) for all three particle types for each sample (Manly 2006).  I 
report the mean contribution (P
x
, mg L
-1
) of detrital, algal, and mineral-core particles 
to seston concentration from this simulation.  Furthermore, I can use these solutions 
and the end members to get the contribution of each particle type to POC (C
d
 * P
d
, C
a
 
* P
a
, C
m
 * P
m
; mg L
-1
) and PN (N
d
 * P
d
, N
a
 * P
a
, N
m
 * P
m
; mg L
-1
) concentration. 
 
Statistical analysis ? All statistical analyses were conducted using SAS v9.1 Level 
1M3 (2007, SAS Institute Inc., Cary, NC, USA).  For WCC 2005-2006 baseflow, I 
performed one-way ANOVAs (proc MIXED) for C:N ratios and modeling results, P
d
, 
P
a
, and P
m
, concentrations with month as the treatment.  Since the contributions of 
detrital, algal, and mineral-core particles to POC and PN are linear scalars of P
d
, P
a
, 
and P
m
, results from seston ANOVAs for each type of particle can be applied to 
seasonal trends in both POC and PN contributions.  For WCC historical baseflow, I 
performed one-way ANOVAs for carbon content (%), seston, and POC 
concentrations with month as the treatment.  I performed one-way ANOVAs on 
modeling results for seston contributions from mineral-core and organic-core 
particles.  Since the contributions of mineral-core and organic-core fractions to POC 
are linear scalars of their respective contributions to seston, I applied ANOVA results 
from the contributions of seston to the POC contributions.  Seasonal differences were 
examined by making multiple mean comparisons using the Tukey?s test for all pair-
wise differences between months and describe the non-significant differences 
between months by using seasons (winter, spring, summer, autumn).  For seston 
 
 77 
 
samples from the Jun 2007 watershed habitat sampling, I performed one-way 
ANOVAs for seston concentration, carbon and nitrogen content, ?
13
C, ?
15
N, and C:N 
ratios with habitat (riffle, run, pond, seep) as the treatment.  I performed two-way 
ANOVAs for carbon and nitrogen content, ?
13
C, ?
15
N, and C:N with habitat (riffle, 
run, pond, seep) and type (soil, bank sediments, CPOM, herbs) as the treatments.  To 
ensure homogeneity of variances, it was necessary to log-transform several variables 
(seston, POC, modeling results); these are reported as back-transformed geometric 
means.  Following each ANOVA, I made multiple mean comparisons using the 
Tukey?s test.  For seston samples from the watershed habitat 2004 sampling, I 
performed paired t-tests between ponds and seeps for seston concentration, carbon 
and nitrogen contents, and C:N ratios. 
Results 
WCC 2005-2006 baseflow sampling ? Over the year, one hundred eighty-three 
samples were taken during baseflow conditions on 11 monthly sampling days.  The 
carbon and nitrogen contents of these samples were contained by the mixing model 
space made by the end members (Fig. 3.2).  Low seston C:N ratios (Fig. 3.3a : 9.2 ? 
0.1: mean ? SE) caused the particles to fall along the axis created by algal and 
mineral-core end members (Fig. 3.2).  Seasonally, seston C:N ratios peaked in Aug 
and Dec, and were lowest in Mar (F
10,169
=9.3, p<0.001).  Throughout the year, 
mineral-core particles dominated seston concentration, followed by algal and detrital 
particles, respectively (Fig. 3.3b).  However, algal particles consistently contributed 
the most to POC concentration, followed by detrital and mineral-associated POC, 
respectively (Fig. 3.3c).  Algal particles also consistently contributed the majority of 
 
 78 
 
PN concentration, followed by mineral-associated nitrogen and detrital particles, 
respectively (Fig. 3.3d). Mineral-core particles showed seasonal trends for seston, 
POC, and PN with high concentrations in the summer and low concentrations in the 
winter (Fig. 3.3b, c, d: F
10,153
=20.2, p<0.001).  Algal particles also showed seasonal 
trends for seston, POC, and PN with the highest concentrations in spring and summer 
and lowest concentrations in the late autumn and winter (Fig. 3.3b, c, d: F
10,153
=34.8, 
p<0.001).  Detrital particles showed seasonal trends for seston, POC, and PN with the 
lowest concentrations in March and July (Fig. 3.3b, c, d: F
10,153
=8.5, p<0.001).  Algal 
POC was positively correlated with pigment (chlorophyll a + pheophytin a) 
concentration (Fig. 3.4; r
2
=0.36; F
1,160
=92, p<0.001; algal POC = 0.09 + 0.01 x 
pigments).   
 
WCC historical baseflow sampling ? Over the 14 years of sampling, from 1993 to 
2006, 502 total seston samples were taken on weekly, bimonthly, or monthly 
sampling intervals with 26 to 50 samples taken for each month.  Seasonally, seston 
concentration was highest in late spring and summer months (April-July), and lowest 
in late autumn and winter months (Oct-Dec) (Fig. 3.5a: F
11,490
=24.0, p<0.001).  
Similarly, POC concentration was highest in spring and summer months (April-
August) and lowest in late autumn and winter (October-Feb) (Fig. 3.5b: F
11,490
=11.4, 
p<0.001).  Stream discharge was not correlated with seston (r
2
=0.05, p = 0.75) nor 
POC concentration (r
2
=0.03, p = 0.77).  Carbon content of seston increases through 
the summer until a peak in November at 17% ? 0.8 (mean ? SE) (Fig. 3.5c: 
F
11,490
=13.1, p<0.001).  Mineral-core particles showed seasonal trends for seston and 
 
 79 
 
POC with high concentrations in the summer and low concentrations in the winter 
(Fig. 3.5d, e: F
11,489
=27.8, p<0.001).  Organic-core particles also showed seasonal 
trends for seston and POC with high concentrations in the summer and low 
concentrations in the winter (Fig. 3.5d, e: F
11,489
=9.5, p<0.001).  Mineral-core 
particles dominated seston concentration for most of the year, except in late autumn 
when mineral-core and organic-core particles made up equal proportions of seston 
(Fig. 3.5d).  Organic-core particles comprised most of POC concentration for the 
entire year (Fig. 3.5e).          
 
Watershed habitat sampling ? For the Aug 2004 sampling, seep seston C:N ratio was 
higher than pond outflow seston C:N ratio, but there was no difference between seep 
and pond outflow seston concentration, carbon content, or nitrogen content (Fig. 3.6).  
For the Jun 2007 sampling, seston concentration and carbon content were highest in 
the pond outflows and seeps and lowest in riffles and runs (Table 3.2).  Seston 
nitrogen content was highest in the pond outflows; while ?
13
C was lowest in pond 
outflows (Table 3.2).  Seston C:N ratios were highest in seeps, and there was no 
difference in ?
15
N among the habitats (Table 3.2).      
 
CPOM and herbs had higher carbon content (F
3,27
=115.1, p<0.001) and 
nitrogen content (F
3,27
=53.4, p<0.001) than soils or bank sediments (Fig. 3.7).  
However, there were no effects of habitat on carbon or nitrogen content.  Herbs had 
lower ?
13
C than CPOM, soils or bank sediments (F
3,27
=17.9, p<0.001), but the was no 
effect of habitat on ?
13
C (Fig. 3.8).  Riffle and run ?
15
N was lower than seep and pond 
 
 80 
 
?
15
N (F
3,27
=7.5, p<0.001), but there was no significant effect of sample type (Fig. 
3.8).  CPOM had higher C:N ratios (26.7 ? 0.9: mean ? SE) than herbs, soils or bank 
sediments (F
3,27
=4.0, p=0.02). 
Discussion 
Seston source ? Seston in streams has historically been considered to be primarily 
terrestrial leaf fragments (Vannote et al. 1980), however, modeling results indicate 
mineral-core particles dominate total seston concentration (Fig. 3.3b, 3.5d).  
Microscopic analysis confirms that most particles are amorphous clay or silt particles 
that are unidentifiable as leaf fragments or algal cells (Meybeck 1982; D.C. 
Richardson pers. obs.).  However, algal cells and leaf detritus, with high carbon 
content, dominate POC concentrations (Fig. 3.3c, 3.5e; Chapter 1).  A small 
percentage of organic-core particles can dominate POC concentration (by mass) 
because of their high carbon content (Table 3.1), which is congruent with microscopic 
analysis of seston where few particles (by number) can be identified as having an 
organic-core.  Algal POC was always greater than detrital POC indicating the 
importance of primary production to POC flux downstream, even in forested 
headwater streams (e.g. Elliott et al. 2004).  Using similar mixing models, 
Aufdenkampe and Buckaveckas (unpubl.) found that Amazon and Mississippi river 
seston is also comprised mostly of mineral-core particles, while POC concentration is 
predominantly algal and detrital leaf particles.  These results explain the importance 
of both mineral-core particles to the mass and number of stream particles (e.g. 
Meybeck 1982; Cole and Caraco 2001), but also the importance of organic-core 
 
 81 
 
(algal and detrital leaf) particles to the flux of carbon transported downstream (e.g. 
Vannote et al. 1980; Elliot et al. 2004). 
 
Seasonal trends in seston composition ? Most POC and seston export from WCC 
occurs during storms (Newbold et al. 1997), as is typical of small streams where 
seston and POC concentrations are positively correlated with stream discharge 
(Golladay et al. 2000).  However, in this study and others in WCC at baseflow, seston 
and POC concentrations are not correlated with stream discharge (Newbold et al. 
1997; Chapter 1).  Thus, alternative hypotheses are necessary to explain seasonal 
patterns of seston concentration and composition (Fig. 3.3, 3.5).  Warm-season peaks 
of seston and POC concentrations are attributed to increasing stream organism 
biomass and activity including feeding and bioturbation (Wallace et al. 1991; Molla 
et al. 2006; Chapter 1).  I attribute this pattern to an increase in all types of activity, 
not simply shredding, because while leaf detritus and organic based particles peak in 
the summer, so do algal particles and mineral-core particles (Fig. 3.3, 3.5).  I 
hypothesize that increases in seston and POC concentrations during the warm months 
(Fig. 3.5a, b) result from shredding and egestion of leaf detritus, increases in 
periphyton growth and grazing on the streambed (Mulholland et al. 1991; Hill et al. 
2001) and increases in stream organism activity (bioturbation) that suspends bed 
sediments (Moore 2006).  In autumn, the seston concentration decreased dramatically 
at the same time as litterfall (Fig. 3.3, 3.6), which could have resulted from the 
filtration of particles by leaf packs instream (Wallace et al. 1993).  This filtration 
could occur simultaneously with increased fragmentation of leaf litter both by 
 
 82 
 
physical and biological degradation (Swan and Palmer 2004).  These two 
simultaneously occurring mechanisms are supported by trends in October through 
December where seston (Fig. 3.3b, 3.5d) and POC (Fig. 3.3c, 3.5e) concentration are 
comprised of higher proportions of organic-core particles, especially leaf detritus, 
particles compared to the rest of the year.  While leaf detritus filtered out all particles 
contributing to an overall decrease in seston concentration, leaf detritus particles are 
generated at a higher rate during the autumn than the rest of the year. 
 
In late winter and early spring, the open canopy, increasing hours of daylight 
and warming temperatures are conducive to the growth of long algal filaments on 
rocks and sediments (Kaplan and Bott 1982; Hill et al. 2001) resulting in peaks of 
algal POC (Fig. 3.3c).  However, the early spring peak did not occur in the organic-
core particles in the historical data (Fig. 3.5d, e).  In the historical data, carbon 
content of seston averaged approx. 11% for March and April averaged over the 14 
year period, but carbon content was 16 and 14% for March 2006 and April 2006 from 
the historical data, respectively, which matches spring peaks in carbon content from 
WCC 2005-2006 data (Chapter 1).  The high algal content of seston during the spring 
of 2006 may have resulted from exceptionally high primary production during the 
spring because of warmer than average temperatures and fewer storms (e.g. Fisher et 
al. 1982; Power and Stewart 1987).     
 
Validation of the model ? The mixing model is a valid tool for characterization of 
seston composition and sources; however there are some limitations and inherent 
 
 83 
 
assumptions made when using the model.  The three end members seemed to be an 
appropriate description of the seston data because all sample points were 
encompassed by the end member space (Fig. 3.2) indicating that they could all be 
separated by these end members.  Microscopic analyses match with the results 
derived from the model.  Most particles are small (<20 ?m) clays and silts with high 
mineral content (Meybeck 1982; Chapter 1).  Less frequently, single diatom cells or 
leaf fragments can be recognized but few particles are clearly identifiable as insect 
parts, macroinvertebrate frass, or large flocs under the microscope.        
 
The algal composition was derived using elemental compositions of cultures 
of marine or lake diatoms in phytoplankton (Table 3.1) because I could primarily 
identify individual algal cells (diatoms) as the only algal derived particles.  I had to 
make this assumption, because, in this study, I was unable to get an adequate end 
member sample from rock scrapings during the watershed habitat sampling due to the 
thin coating of periphyton on rocks, possibly because of macroinvertebrate grazing or 
low primary production with the closed summer canopy.  Furthermore, I was unable 
to easily separate algal from sediment or allochthonous matter in periphyton by hand 
for previous sampling (Chapter 1).  Future work will take advantage of centrifugation 
techniques to isolate algal cells from epilithon (e.g. McCutchan and Lewis 2002).  
From this study, the best example of algal particles was pond seston, which was 
similar to literature derived algal end members with high carbon and nitrogen content 
(Fig. 3.8), low C:N ratios (<9) and ?
13
C that was closest to periphyton stable carbon 
isotopes (Fig. 3.8; Chapter 1; Walters et al. 2007).  Furthermore, algal carbon was 
 
 84 
 
positively correlated with pigment concentration (Fig. 3.4) and C:pigment ratios were 
within realistic values calculated in other studies (Angradi 1993), indicating the 
validity of the algal end member. However, while individual diatom cells can be 
clearly identified using a microscope, some suspended POC in the stream could come 
from extracellular polymeric substances produced by primary production on rocks 
and sediments (Decho and Lopez 1993; Sutherland et al. 1998) and was not 
accounted for using this model.    
  
For most temperate deciduous forests with developed riparian zones, 
including WCC, litterfall is an important source of particulate organic matter to 
stream ecosystems (Fisher and Likens 1973; Benfield 1997).  In WCC, most litterfall 
enters the stream as deciduous leaf detritus during mid to late October (J. D. Newbold 
unpubl.).  Leaf detritus has low variation in carbon content (44%) and consistent 
nitrogen composition despite small variations in C:N ratio among different tree 
species (Ostrofsky 1997; Swan and Palmer 2004).  The leaf detritus end member was 
assumed to have elemental composition similar to the senesced litterfall entering the 
stream.  However, as leaf material is colonized by microorganisms in the stream the 
nitrogen content increases as the leaves decay because attached microorganisms will 
use inorganic nitrogen from the water column for cell growth (Suberkropp et al. 
1976).  This effect might lower the C:N ratio of the fragmented leaf material causing 
the leaf detritus end member to shift towards the algal end member (Fig. 3.2).  This 
idea is supported by the CPOM sampled for the watershed habitat sampling.  The 
 
 85 
 
CPOM was collected from leaf packs within the stream and has higher nitrogen 
contents (Fig. 3.7, 1-2%) than the end member from the model (Table 3.1, 1%).    
 
In this study, seston is divided into three distinct particle types.  However, 
seston composition can include a number of other particle types including small 
particles (0.7 to 1 ?m) such as bacterial cells, DOC aggregates, and clay particles 
with adsorbed inorganic nitrogen.  Bacterial cells and DOC aggregates contain lipids, 
polysaccharides and proteins resulting in C:N ratios (5) lower than any of the end 
members for this study (Vallino et al. 1996; Kerner et al. 2003; Table 3.1).  However, 
both bacterial cell and DOC aggregate carbon only contribute a small fraction of POC 
concentration in streams (Kaplan and Newbold 1995; Kerner et al. 2003).  Mineral-
core particles could include ammonium adsorbed to clays or silts through ionic 
bonding (Mayer et al. 1998).  Theoretically, this could add a fourth particle end 
member with high nitrogen content with low carbon content.  Streambed sediments 
have very low nitrogen content (0.03%; D. C. Richardson, unpubl.) indicating this 
fourth end member does not comprise a large proportion of seston in WCC.    
 
Stable isotope analyses offer an interesting tool to elucidate seasonal trends of 
seston transport (Kendall et al. 2001) and to differentiate terrestrial organic matter 
sources from instream production (Finlay et al. 2002) since these sources have 
differing ?
13
C values (Fry 2007).  However, separation of POC sources in streams 
using end members such as periphyton, terrestrial vegetation, and soil OM is 
challenging because the ranges of ?
13
C are large, variable and overlapping and seston 
 
 86 
 
?
13
C was occasionally outside the range of measured isotope end members (Kendall 
et al. 2001; Chapter 1).  Mixing models are ineffective without clear separation 
between end members and end members that encompass the samples (Fry 2007).  
Furthermore, algal ?
13
C is known to vary spatially as a function of the concentration 
of dissolved free CO
2
, the ?
13
C of dissolved inorganic carbon, water-velocity effects 
on cell-wall CO
2
 gradients, and primary production rates, creating a range of ?
13
C 
that depends on position within the watershed and local reach morphology (Finlay et 
al. 1999; Walters et al. 2007).  The proportion of seston existing as organic-core and 
mineral-core particles may be more easily obtained using carbon content because of 
the distinct separation between the two end members (Meybeck 1982).  Because algal 
carbon and detrital organic matter from terrestrial leaves have differing C:N ratios 
(Table 3.1), further separation of seston into three particle types (mineral-core, algal 
cells, and leaf detritus fragments) was possible using both carbon and nitrogen 
content.  Also, in the watershed sampling, the separation by elemental composition 
was significantly clearer than stable isotopes (Fig. 3.7 vs. Fig. 3.8).   
 
Watershed habitat sampling ? The watershed habitat sampling served to identify 
potential sources of seston to the 3
rd
-order WCC.  Seston could come from litter of 
riparian vegetation entering by direct litterfall or lateral movements (Benfield 1997), 
slumping riparian herbaceous vegetation (Menninger and Palmer 2007), bank, soil or 
streambed sediments (Golladay et al. 2000), autochthonous production of periphyton 
(Angradi 1993; Sutherland et al. 1998), or from upstream ecosystems.  Upstream 
ecosystems can include habitats such as zero order streams (seeps) or pond outlets 
 
 87 
 
that are connected directly to stream networks.  The transport and downstream 
influence of seston from seeps and ponds to a stream network depends on the 
magnitude of discharge coming from the habitat and the transport distance of the 
particle (Vadeboncoeur 1994; Thomas et al. 2001).   
 
Seeps are fed by groundwater resulting in low disturbance environments with 
accumulations of decomposing CPOM (Burns et al. 1998; Duval and Hill 2006).  
Therefore, seeps contribute high concentrations of fragmented and egested leaf litter 
to the stream network; this idea is supported by the similarity of seep seston 
composition to CPOM (Fig. 3.7).  Each individual seep has little flow to add fluxes of 
these particles to the stream network, but the cumulative total flux from all seeps 
across a forested stream network could contribute significantly to leaf detritus in the 
stream network. 
 
Fluvial discontinuities such as lakes (natural and manmade) and ponds (e.g. 
natural, beaver, mill, farm, recreational) are common features of most landscapes 
(Winter 2001).  Ponds are open, slowly moving or stagnant water that result in high 
primary productivity, and can export high quality autochthonous material to streams 
(Fig. 3.6, 3.7; Richardson and Mackay 1991).  In this study, seston content at pond 
outflows could account for up to 50% of total instream seston fluxes (3 L s
-1
 x 
concentration from Table 3.2), although the pond seston likely only travel short 
distances downstream at low flow (Vadeboncoeur 1994; Thomas et al. 2001).     
  
 
 88 
 
Riparian herbs and grasses can contribute significant quantities of detritus to 
streams through live or senesced overhanging vegetation or material entering the 
water column during floods (Gurnell 2007; Menninger and Palmer 2007).  This can 
be especially important in deforested riparian zones where streamside herbs and 
grasses are the dominant terrestrial riparian vegetation.  Within the WCC watershed, a 
portion of streams run through agricultural fields and residential property where 
grasses and herbs are the dominant riparian vegetation.  Also, along the forested 3
rd
-
order reach and in locations such as seeps and ponds, herbs and grasses grow 
abundantly alongside and within the stream channel.  Live herbs and grasses, because 
of high nitrogen content, are a better quality resource (Fig. 3.7; MacKay et al. 1992; 
Menninger and Palmer 2007) and decompose more rapidly than leaf litter (Menninger 
and Palmer 2007).  Decomposition of herbs and grasses could provide a source of 
high quality particles with a unique stable isotope signature (Fig. 3.8) to the stream 
ecosystem, but would require quantification of the mass of herbs and grasses entering 
WCC and whether the plants enter the stream channel live or senesced.   
 
Importance of headwater stream seston ? Within a 3
rd
-order piedmont stream, seston 
transport is dominated by mineral-core particles, while POC transport is dominated 
by organic-core particles including algal cells and leaf detritus.  Seasonal patterns of 
seston transport and composition reflect seasonal cycles of litterfall, stream organism 
activity such as feeding and egestion of leaf litter, bioturbation and grazing, and, 
during some years, increased algal carbon from vernal algal production.  Even in 
headwater streams, where historically, algal carbon was not considered an important 
 
 89 
 
component of downstream carbon flux (Fisher and Likens 1973; Vannote et al. 1980), 
algal cells are a key component of POC transport throughout the year and their high 
lability could link carbon from headwater streams to immediate downstream reaches.  
More refractory leaf detritus and terrestrially-derived mineral-associated organic 
material may link headwaters to mid-order streams, rivers and estuaries (Cole and 
Caraco 2001).  Furthermore, connections of ponds and seeps to streams within the 
watershed provide a local link of high carbon content particles to the stream network.  
Future work is needed to link particle composition to organic matter quality and 
transport to better understand the role of stream ecosystems in controlling the 
transformations of allochthonous and autochthonous POC transported from inland 
waters to the oceans (Thorp et al. 2006; Cole et al. 2007). 
 
 90 
 
Tables 
 
Table 3.1. End members (means ? 1 SE) from literature values and this study used 
for 3-compartment mixing model to separate seston carbon (C
x
) and nitrogen content 
(N
x
) in leaf detritus (d), algal cells (a), and mineral-core (m) particles. 
 
  End members 
 
Detrital 
(C
d
 or N
d
) 
Algal 
(C
a
 or N
a
) 
Mineral-core  
(C
m
 or N
m
) 
Carbon 
content (%) 
42.1% ? 3.1
 
26% ? 4.9 1.5% ? 0.2 
Nitrogen 
content (%) 
0.98% ? 0.36 4.33% ? 1.41 0.22% ? 0.04 
C:N 50 ? 17.9 7 ? 1.9 9 ? 1 
Citations 
3,4,7,10
 
3,4,5,6,8,9,11,12
 
1,2,7
 
1
Aufdenkampe et al. 2001; 
2
Aufdenkampe et al. 2007; 
3
Cross et al. 2005; 
4
Elser et al. 2000; 
5
Geider and La Roche 2002; 
6
Hedges 2002; 
7
Hedges et al. 1986; 
8
Hillerbrand and Sommer 
1999; 
9
Leonardos and Geider 2004; 
10
Chapter 1; 
11
This study (pond seston); 
12
Ho et al. 2003 
 
 
 91 
 
Table 3.2. Seston concentration (conc.) and composition (means ? 1 SE) from 
watershed habitat survey in June 2007.  One-way ANOVAs for each seston variable 
test differences across habitats (riffles, runs, pond outflows, and seeps).  Within each 
variable, habitats with different alphabetic superscripts are significantly different 
(Tukey?s test). 
  
  Seston variables 
Habitat 
Conc. 
(mg L
-1
) 
Carbon 
content (%) 
Nitrogen 
content (%) ?
13
C ?
15
N C:N 
Riffle 4.2 ? 1.1
 a
 7.9 ? 0.9
 a
 0.9 ? 0.1
 a
 -27.7 ? 0.2
 a
 5.1 ? 0.3 10.0 ? 0.2
 a
 
Run 3.0 ? 0.4
 a
 8.6 ? 0.7
 a
 1.0 ? 0.1
 a
 -29.3 ? 0.1
 a
 5.7 ? 0.5 9.6 ? 0.1
 a
 
Pond 21.7 ? 4.2
 b
 26.9 ? 4.3
 b
 3.4 ? 0.5
 b
 -24.8 ? 1.1
 b
 5.7 ? 0.5 9.2 ? 0.1
 a
 
Seep 20.2 ? 3.2
 b
 23.2 ? 4.4
 b
 1.6 ? 0.3
 a
 -27.8 ? 0.6
 a
 4.6 ? 0.6 17.0 ? 0.4
 b
 
F-stat F
3,7
=17.7 F
3,7
=10.8 F
3,7
=15.9 F
3,7
=11.2 F
3,7
=1.1 F
3,7
=2.3 
p-value 0.001 0.005 0.002 0.005 0.42 <0.001 
 
 
 92 
 
Figure Captions 
 
Figure 3.1.  Map of White Clay Creek, SE Pennsylvania where P1-P4 are ponds, S1-
S4 are seeps, R1-R3 are riffles and N1-N3 are runs sampled for the watershed habitat 
sampling.  The watershed boundary is indicated by the red line.  Streams (lines) and 
ponds (circles) within the watershed indicated in white.   
 
Figure 3.2.  Hypothetical partitioning of source contributions from three selected end 
members (leaf detritus, algal cells and mineral associated particles for mixtures 
collected as seston.  Open circles represent individual data points from WCC 05-06 
sampling, while the boxes represent ?1 SD for end members collected from literature 
values. 
 
Figure 3.3.  Seasonal changes in (a) seston carbon:nitrogen (C:N) ratios and 
modeling results for (b) seston concentration, (c) suspended POC concentration, and 
(d) suspended particulate nitrogen concentration by particle type for WCC 2005-2006 
samples.  Data is plotted by julian day with sampling months are identified in panel a; 
error bars represent ? 1 SE.     
 
Figure 3.4.  Least squares linear regression between pigment concentration (?g L
-1
) 
and algal POC (mg L
-1
) from WCC 2005-2006 baseflow modeling (algal POC = 0.09 
+ 0.01 pigments, r
2
=0.36, p<0.001).  
 
 
 93 
 
Figure 3.5.  WCC historic baseflow seston characteristics averaged over 14 years 
(1993-2006).  Panels show seasonal changes in (a) seston concentration, (b) 
suspended POC concentration, and (c) carbon content.  Modeling results of seston 
composition for seston concentration and suspended POC concentration are shown in 
panels (d) and (e). Error bars represent ? 1 SE.   
 
Figure 3.6.  Comparisons between pond outflow and seep (a) seston concentration, 
(b) carbon content, (c) nitrogen content, and (d) carbon:nitrogen ratios for watershed 
habitat samples collected in 2004.  Error bars represent ? 1 SE and * indicates 
significantly different means (p<0.05). 
 
Figure 3.7.  Elemental composition by habitat (pond, riffle, run, seep) and sample 
type (seston, herbs, CPOM, soil sediments, bank sediments) for watershed habitat 
samples collected in 2007.   
 
Figure 3.8. Stable isotope ratios by habitat (pond, riffle, run, seep) and sample type 
(seston, herbs, CPOM, soil sediments, bank sediments) for watershed habitat samples 
collected in 2007.   
P1
P2
S1
S2
R1
R2
N1
N2
P3
S3
R3
N3
P4
0 500 1000
meters
S4
Figures
Figure 3.1
Leaf 
detritus
Algal 
cells
Mineral
core
Nitrogen content (%)
Carbon content (%)
48
40
32
24
16
8
0
0 21 3 4 5 6
Figure 3.2
POC (mg L
-1
) 
01 Aug 05 02 Oct 05 03 Dec 05 03 Feb 06 06 Apr 06 07 Jun 06 08 Aug 06
Aug Sep Nov Dec Jan Feb Mar Apr May Jun
Jul
PN (mg L
-1
) 
Seston (mg L
-1
) 
C:N
12
11
10
9
8
7
3.5
2.5
2.0
1.5
1.0
0.5
0
3.0
0.20
0.15
0.10
0.05
0
0.04
0.03
0.02
0.01
a
b
c
d
0
mineral-core algal cells
leaf detritus
Figure 3.3
Pigments (?g L
-1
)
Algal POC (mg L
-1
)
0 1 2 3 4 5 6 7 8 9 10
0
0.1
0.2
0.3
Figure 3.4
Seston
(mg L
-1
)
POC
(mg L
-1
)
Carbon 
content (%)
a
b
c
4
3
2
1
0.5
0.4
0.3
0.2
0.1
18.0
16.5
15.0
13.5
12.0
10.5
9.0
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Seston
(mg L
-1
)
POC
(mg L
-1
)
d
e
3
2
1
0
0
0.1
0.2
0.3
0.4
0.5 Mineral-coreOrganic-core
Mineral-coreOrganic-core
Figure 3.5
a b c d
*
Pond     Seep Pond     Seep Pond     Seep Pond     Seep
Seston (mg L
-1
)
C content (%) N content (%)
C:N
10
8
6
4
2
0
40
30
20
10
4
3
2
1
25
10
5
20
15
Figure 3.6
CPOM
Herbs
Sediments
& soils
Seston
Legend
Nitrogen content (%)
Carbon content (%)
30
25
15
10
5
0
43210
Pond
Riffle
Run
Seep
35
40
45
20
Figure 3.7
Legend
CPOM
Herbs
Seston
Soils and 
sediments
?
13
C
?
15
N
10
8
6
4
2
0
-34 -32 -30 -28 -26 -24
Pond
Riffle
Run
Seep
Figure 3.8
 
 102 
 
 
Chapter 4: A method for measuring bacterial mineralization 
rates of suspended particulate organic carbon in stream 
ecosystems 
 
Introduction 
Each year rivers export approximately 0.2 Pg of particulate organic carbon 
(POC) to the oceans (Spitzy and Ittekot 1991; Battin et al. 2008). Streams and rivers 
not only transport POC, but also play an active role in the processing of POC through 
mineralization of allochthonous and autochthonous organic matter.  For example, 
CO
2
 out-gassing from large rivers indicates that POC is metabolically important and 
at least a fraction of heterotrophic metabolism in the water column is supported by 
POC exported from upstream reaches  (Cole and Caraco 2001; Richey et al. 2002; 
Mayorga et al. 2005; Battin et al. 2008).  Yet, despite the ecological importance of 
POC, little is known about respiration rates of suspended POC in streams and, 
therefore, the quality of small (<1 mm) particles in transport (but see Naiman and 
Sedell 1979).   
 
Most measurements of POC quality use benthic POC, which supports local 
ecosystems and is relatively easy to collect from the streambed.  Webster et al. (1999) 
summarized results of respiration rates on benthic POC from several streams and 
determined that POC quality in small streams is low.  In those studies, the biological 
turnover time, defined as the average time before organic carbon is respired as 
inorganic CO
2
, was 2.6 years, implying that POC would travel tens to hundreds of 
 
 103 
 
kilometers into downstream rivers, estuaries, and oceans before it is used by the 
heterotrophic community (Webster et al. 1999).   
 
POC transport would therefore appear to be a weak longitudinal link between 
upstream production, both terrestrial and aquatic, and downstream rivers.  However, 
the strength of the link is controlled by the quality (lability) of the transported POC 
(Mayorga et al. 2005).  Labile POC should be metabolized in upstream reaches within 
short transport distances, while refractory (non-labile) POC should travel greater 
distances downstream with minimal biological processing (e.g. Kaplan and Newbold 
2003).   A continuous array of labilities exists within the suspended load (seston), 
with POC ranging from labile (e.g. fresh algal cells) to refractory (e.g. humified soil 
organic matter).  Understanding the diversity of POC quality is critical in determining 
the fraction of ecosystem respiration supported by POC transported from upstream 
and the resulting degree of longitudinal energy transfer.   
      
Microbial respiration rates are used to estimate POC lability and link energy 
use with organic matter quality (Benner et al. 1995; Opsahl 2005).   The rate of 
dissolved oxygen (O
2
) consumption under dark conditions is a simple and direct 
assessment of organic carbon oxidation resulting from heterotrophic metabolic 
activity and therefore an appropriate measure of carbon quality.  
 
However, there are several methodological concerns regarding the ability to 
measure POC quality: (1) oxygen measurements need to be precise, (2) dissolved 
 
 104 
 
organic carbon (DOC) respiration may mask POC respiration in small streams, and 
(3) present methods do not replicate instream conditions or measure the range of 
suspended POC lability.  In most small forested streams, concentrations of suspended 
POC are quite low, and heterotrophic respiration yields small changes in oxygen.  
Without precise measurements of oxygen concentrations, the quality of suspended 
POC is difficult to assess.  For example, an average suspended POC concentration in 
a piedmont 3
rd
-order stream is 0.15 mg POC L
-1
 (Chapter 1).  Heterotrophic 
respiration of 0.03 mg POC L
-1
, i.e. 20% of the total concentration, requires oxygen 
measurements with <1% coefficient of variation to detect the associated consumption 
of oxygen, assuming a respiratory quotient of 1.  Furthermore, DOC concentrations 
often exceed POC concentrations in northeastern Piedmont streams (DOC:POC>1; 
Kaplan et al. 2006) and respiration of the larger DOC pool could mask oxygen 
consumption from POC respiration. The incubation of POC is also critical to 
measurements of heterotrophic oxygen consumption, and therefore measurements of 
POC quality.  Even though seston is typically in a turbulent environment and 
constantly bouncing off the stream bed (McNair and Newbold 2001), most methods 
for measuring respiration require the samples to be shaken lightly, if at all (Benner et 
al. 1995; Webster et al. 1999; Bonin et al. 2000).  Replicating suspension under 
laboratory conditions could be important for accurately measuring respiration rates, 
because suspension may maximize oxygen diffusion to the microorganisms attached 
to individual particles (Ploug and Grossart 1999).  Also, most respiration rates are 
measured after short-term incubations (hours: Webster et al. 1999; Benner et al. 
1995), yet, POC in streams and rivers could be comprised of a small, rapidly-used, 
 
 105 
 
labile fraction mixed with a larger, gradually-consumed, refractory fraction.  The idea 
that the carbon pool is comprised of a range of labilities has been similarly proposed 
for stream DOC (Kaplan and Newbold 2003), soil organic matter (Trumbore 2000) 
and marine sedimentary organic matter (Berner 1995).  Short-term measurements do 
not sufficiently represent the continuum of POC lability, but long-term measurements 
(weeks to months) begin to distinguish labile from refractory or recalcitrant fractions.     
 
In this chapter, I (1) report a method for measuring the quality of suspended 
POC from aquatic environments using heterotrophic respiration as a metric for 
lability, (2) test different components of the method, and (3) compare respiration rates 
of suspended POC from this study to reported respiration rates of POC.  The method 
includes concentrating POC relative to DOC, keeping particles in suspension, making 
precise oxygen measurements over a period of 35-40 days, and correcting for 
respiration of DOC.  I measured suspended POC quality from a 3
rd
-order SE 
Pennsylvania stream on two different dates and compared treatments where POC was 
allowed to settle versus those where POC was kept continually in suspension.  I also 
tested the effect of adding beads, which disperse POC and break up flocs.  Finally, I 
tested the effect that concentration of POC had on DOC lability.  Using this method, 
initial respiration rates for suspended POC were an order of magnitude higher than 
those reported in the literature for benthic POC (Webster et al. 1999). 
 
 
 106 
 
Methods 
Site description ? Three experiments were conducted to assess a new method for 
measuring heterotrophic respiration of suspended particulate and dissolved organic 
carbon.  The three experiments started on 17 January 2006 (Jan 2006), 04 April 2007 
(Apr 2007) and 15 August 2007 (DOC lability).  Stream water was collected for all 
three experiments from the upper east branch of White Clay Creek (WCC: 
N39?51?33.9?, W75?46?59.6?), a 3
rd
-order watershed draining 7.3 km
2
 of eastern 
piedmont land.  The watershed has 23% of total area as temperate deciduous forest 
and 74% as agricultural land, but has a largely intact riparian forest and livestock 
have minimal access to the stream (Newbold et al. 1997).  At baseflow, WCC 
averages 122 L s
-1
 discharge, has high nutrient concentrations (3-5 mg NO
3
-N L
-1
; 4-
33 ?g PO
4
-P L
-1
) from agricultural land use, moderate dissolved organic carbon 
concentrations (1.5 mg DOC L
-1
; Newbold et al. 1997) and between 0.1 and 0.4 mg 
suspended POC L
-1
 (Chapter 1).   
 
Sample collection and concentration ? For each experiment, stream water was 
collected from WCC at baseflow using a submersible battery operated pump (Whale 
pump) suspended at 2/3 depth in the water column (approximately 15 cm deep).  
Water was pumped through a 53 ?m sieve to remove any large particles, while still 
collecting ~90% of baseflow seston particles (Kaplan et al. 2002; Kaplan et al. 2006).  
I concentrated the particles using tangential flow filtration (TFF; Millipore Pellicon) 
to increase POC concentrations.  TFF applies flow across the surface of a membrane 
(0.65 ?m pore-size Millipore Pellicon 2 Cassette Filter), driven by a peristaltic pump.  
 
 107 
 
Particles that are larger than the membrane pore size are kept in suspension to 
minimize accumulation and clogging on the membrane while pressure forces a 
portion of the water through the membrane as filtrate.  Approximately 200 L was 
concentrated to 20 L of sample for Jan 2006 and Apr 2007 experiments and 100 L of 
stream water to 10 L of sample for the DOC lability experiment (Aug 2007).  
However, the TFF process may involve loss of particle mass because of adsorption to 
the membrane (Minor and Nallathamby 2004).  
 
Experimental design ? For the four treatments tested in each experiment (Jan 2006 
and Apr 2007), the final volume of concentrated water (20 L) was divided into four 5-
L aliquots (Table 4.1).  I measured heterotrophic respiration on DOC alone as a 
control and to correct for respiration of DOC in bottles containing both DOC and 
POC.  I also tested the effects of (1) keeping particles suspended using a rotating axle 
(see incubations section), (2) dispersing particles and aggregates using glass beads, 
and (3) allowing the particles to settle in static treatments.  For the DOC lability 
experiment, there were only two treatments and both were particle-free (preTFF and 
postTFF) to test the effect of TFF on DOC quantity and quality.   
 
To remove particles for the DOC control treatments, I filtered 5 L of 
concentrated particles through glass fiber filters (Millipore AP40 with nominal pore-
size of 0.7 ?m) using vacuum filtration.  All 5 L subsamples, with and without POC, 
were left in a water bath (<8 hrs) to equilibrate with the incubation temperature 
(18?C) and then were shaken vigorously by hand for 5 minutes to saturate the water 
 
 108 
 
with oxygen.  Sample water from each 5L bottle was divided into 21 muffled 
borosilicate glass 60-mL BOD bottles (Wheaton) using a 5 L churn splitter (USGS 
2002) to ensure homogenous distribution of POC into each BOD bottle.  Each bottle 
was filled at a slow but steady rate to overflow to avoid adding air bubbles to the 
bottle.  The bottle was capped with a glass stopper and another plastic cap was placed 
over the stopper while the bottle was submerged in a water bath to eliminate any 
potential for atmospheric exchange with gases in the sample water.  To test the effect 
of dispersing particle and bacterial flocs in both POC (Jan 2006) and DOC alone (Apr 
2007) treatments (Table 4.1), I added 20 acid washed (1N HCl) and muffled (6 hours, 
480?C) glass beads (4 mm diameter) to the 60-mL BOD bottles prior to adding the 
sample water.  The beads provided additional surface area (50 mm
2
 bead
-1
) for 
bacterial growth.  However, adding 20 beads only increased the internal surface area 
of the BOD bottle (11,000 mm
2
) by 9%.   
 
Carbon measurements ? Replicate aliquots of sample water were taken from the 
churn splitter at the beginning, middle and end of each sample distribution to analyze 
for initial POC and DOC concentrations.  Sample water for all treatments was filtered 
through stacked pairs of identical pre-weighed and pre-ashed (480?C for 4.5 hours) 
glass fiber filters (Millipore AP40 with nominal pore-size of 0.7 ?m).  The filters 
were dried for 12-18 hours at 60?C and re-weighed to calculate seston concentration 
(mg L
-1
).  Subsamples of each seston filter were analyzed for carbon content using an 
Elemental Analyzer (Carlo Erba) (Chapter 1).  DOC concentration was measured 
from the filtrate water (<0.7 ?m) using UV catalyzed persulfate oxidation detected by 
conductivity (Sievers 800 or 900 TOC analyzers; Kaplan 1992). 
 
 109 
 
 
Incubations ? Each 60-mL BOD bottle was incubated in the dark in a large 
temperature controlled water bath.  For the settled particle treatment, BOD bottles 
were placed on the water bath floor and particles settled to the bottom of the bottle 
within 1 hour.  The remaining bottles (suspended treatments) were strapped to the 
incubation axle to keep particles in suspension (Fig. 4.1).  The incubation axle was a 
2 m PVC axle with maximum capacity of 108 sixty-mL BOD bottles.  The axle, 
connected to a motor by a plastic gear and chain, rotated at 5 revolutions per minute 
to keep particles in constant suspension.  Temperature in the water bath was 
maintained at 18?C for all experiments by circulating water between the water bath 
and a thermostatically regulated reservoir.         
 
Dissolved oxygen measurements ? Three randomly selected replicate bottles were 
removed at 7 time periods for Jan 2006 and Apr 2007 experiments (approximate 
sampling times after initialization of the experiment: 0 hours, 5-7 hours, 24-36 hours, 
7, 14, 28, and 35 days).  Only the first 5 time periods were sampled for the DOC 
lability experiment.  I quantified bacterial respiration by measuring changes in 
oxygen concentration over time using the Winkler technique with auto-titration 
(Bryan et al. 1976; Graneli and Graneli 1991).  Each bottle was treated with 0.4 mL 
of manganous sulfate and 0.4 mL of alkaline iodide solutions followed by 0.8 mL of 
hydrosulfuric acid (18 N) (Bryan et al. 1976).  A 25-mL aliquot was drawn from the 
acidified solution with a volumetric pipette and was titrated using sodium thiosulfate 
(0.0125 N).  Titration equivalence points were determined potentiometrically using an 
 
 110 
 
auto-titrator (Mettler DL-21) equipped with a platinum combination electrode 
(Mettler DM 140-SC) (Graneli and Graneli 1991; Furuya and Harada 1995).  Oxygen 
concentrations can be routinely measured with 0.25% coefficient of variation using 
careful handling procedures and automated titrations (A.K. Aufdenkampe unpubl.).  
As a further check, each time I ran sequences of samples on the auto-titrator, one 
duplicate sample was measured immediately after the first.  Despite known issues 
with volatilization of iodine, analytical uncertainty was 0.6%, measured as percent 
coefficient of variation.   
 
Model fitting and treatment comparisons ? For each treatment in each experiment, I 
fit the loss of oxygen concentration over time (oxygen consumption curve) with a 
first order loss rate curve (eq. 4.1) using nonlinear least squares analysis (proc NLIN, 
SAS 8.1; SAS Institute Inc., Cary, NC, U.S.A.)   
 
Equation 4.1  O
2
 = c
1
 e
-k * t
 + c
0
 
 
where t is time (hours).  I fit three constant parameters to eq. 4.1: the first parameter, 
k, is the temporal loss rate (hours
-1
); the reciprocal of the loss rate is the observed 
biological turnover time (T
b-observed
) of POC.  The second parameter, c
1
 (mg O
2
 L
-1
), is 
the total possible oxygen consumed from heterotrophic respiration.  The third 
parameter, c
0
 (mg O
2
 L
-1
), is the oxygen concentration that is unconsumed by 
respiration; c
1
 + c
0
 is the total initial oxygen concentration in each bottle at the start of 
the experiment.   
 
 111 
 
 
To derive oxygen consumption for POC alone, oxygen consumption for the 
POC + DOC treatments was corrected for heterotrophic respiration of DOC (eq. 4.2, 
see Appendix 1 for derivation).   
 
Equation 4.2  O
2 corrected
 = A
exp
 ? (c
1doc
 e
-kdoc * t
 + c
0doc
)
 
+ (c
1
 e
-k * t
 + c
0
) 
 
For each experiment, the oxygen concentration consumed by DOC respiration was 
calculated by subtracting the DOC consumption curve, using k
doc
, c
1doc
, and c
0doc
 
parameters for the DOC control, from A
exp
, the initial oxygen concentration for each 
experiment (8.3 mg L
-1
 for Jan 2006 and 8.05 mg L
-1
 for Apr 2007).  Oxygen 
consumption from DOC respiration was added to the oxygen consumption curve from 
total carbon (POC + DOC) using k, c
1
, and c
0
, parameters from the POC + DOC 
consumption curve of interest.  The resulting oxygen consumption curve, for POC 
alone, was fit with a new curve with parameters k, c
1
, and c
0
 (eq. 4.1).   
 
Model fits (r
2
, p-values from F-statistics) were calculated for each curve using 
the extra sums of squares principal (Draper and Smith 1998).  I report approximate 
95% confidence limits for all model parameters (c
0
, c
1
, k) as calculated by the least 
squares analysis (proc NLIN).  Comparisons between treatments (t-tests) were made 
using estimates of the rate coefficients and standard errors from nonlinear least 
squares analysis.  The absolute value of the first derivative (instantaneous slope) of 
eq. 4.1 was derived (eq. 4.3) to calculate change in oxygen per unit time (eq. 4.3).   
 
 112 
 
 
Equation 4.3  dO
2
 dt
-1
 = c
1 
k e
-k * t
 
 
To calculate respiration rates, eq. 4.3 was scaled to the change in oxygen mass per 
initial mass of carbon (mg O
2
 initial g C
-1
 day
-1
) using the BOD bottle volume (60 
mL) and the initial mass of DOC or POC in each bottle.  The instantaneous biological 
turnover time (T
b-instant
) was calculated by taking the reciprocal of the decomposition 
rates calculated from respiration rates assuming a respiratory quotient of 1 (Webster 
et al. 1999).       
 
Results 
Jan 2006 and Apr 2007 experiments ? POC concentrations were elevated 5-fold 
above stream POC concentrations after tangential flow filtration for both 
experiments.  POC, as the % of total organic carbon (POC + DOC), was increased 
from 11% to 44% for Jan 2006 and 9% to 33% for Apr 2007 (Table 4.2); however 
TFF concentration of particles resulted in ~50% mass recovery in both experiments.  
For the Jan 2006 experiment, heterotrophic oxygen consumption was highest in the 
POC + DOC treatments (dispersed, suspended particles>suspended particles>settled 
particles) and lowest in the DOC control treatment (Fig. 4.2a).  For the Apr 2007 
experiment, heterotrophic oxygen consumption was highest in the POC + DOC 
treatments (dispersed, suspended particle and settled particle) and lowest in the DOC 
control treatments (DOC control and dispersed DOC control) (Fig. 4.3a).  The first 
order decay model (eq. 4.1) was an excellent fit for all curves in both experiments 
 
 113 
 
(Table 4.3, r
2
>0.97).  After correcting for DOC driven oxygen consumption, at the 
onset of the experiment, respiration rates for POC were 180 mg O
2
 initial g C
-1
 day
-1
 
for the Jan 2006 experiment and 300 mg O
2
 initial g C
-1
 day
-1
 for the Apr 2007 
experiment.  After 30-35 days, respiration rates slowed to approximately 10 mg O
2
 
initial g C
-1
 day
-1
, or 5% and 3% of the initial rates, respectively (Fig. 4.2b, 4.3b).  For 
the Jan 2006 experiment, the largest difference among POC treatments occurred 
within the first week, with respiration rates reduced by 33% from the dispersed, 
suspended particle to the suspended particle treatments and 50% from suspended 
particle to settled particle treatments (Fig. 4.2b).  For the Apr 2007 experiment, the 
largest difference among POC treatments also occurred within the first week, with 
respiration rates reduced by 26% from dispersed, suspended particle to settled particle 
treatments (Fig. 4.3b).  The respiration rates for dispersed, suspended particles were, 
at first, 4- to 6-times greater than the respiration on DOC alone, although after 3+ 
weeks, the respiration rates approached the same asymptote at approximately 10 mg 
O
2
 initial g C
-1
 day
-1
 (Fig. 4.2b, 4.3b).   
Using the temporal loss rate coefficient (k) for the Jan 2006 experiment, I 
calculated that the biological turnover time for carbon (T
b-observed
) was shortest in the 
dispersed, suspended particle treatment, followed by the suspended particle treatment, 
settled particle treatment, and lastly in the DOC control (Table 4.4).  For the Apr 
2007 experiment, T
b-observed
 was shortest in the dispersed, suspended particle and 
settled particle treatments, and those turnover times were not significantly different 
from each other (Table 4.4; t=-2.17, df=8, p=0.06).  The two DOC control treatments 
had slower turnover times, and those rates were not significantly different from each 
 
 114 
 
other (Table 4.4; t=-0.88, df=36, p=0.38).  The dispersed, particle treatment turnover 
time was 3x faster than the DOC control (Table 4.4; t=9.12, df=22, p<0.001).  The 
turnover time for DOC control was not statistically different between the Jan 2006 
and Apr 2007 experiments (Table 4.4; t=0.14, df=36, p=0.88).  However, when 
comparing dispersed, suspended particle treatments from both experiments, the 
turnover time for Apr 2007 was 40% shorter than for the Jan 2006 experiment (Table 
4.4; t=3.2, df=8; p=0.01).       
   
DOC lability experiment ? Tangential flow filtration elevated DOC concentrations by 
60% from 1.3 mg DOC L
-1
 to 2.1 mg DOC L
-1
.  Heterotrophic consumption of 
oxygen was higher in the postTFF than preTFF treatment (Fig. 4.4a) and the first 
order decay model (eq. 4.1) was an adequate fit for both treatments (r
2
>0.92; Table 
4.3).  T
b-observed
 for preTFF DOC (4.6 days) and postTFF (12.8 days) were not 
significantly different (Fig. 4.4b, Table 4.4; t=1.27, df=24, p=0.21). 
 
Discussion 
I found that TFF elevated POC concentration enough to measure respiration 
over short incubation times with small changes in oxygen, especially when using the 
Winkler method with auto-titration for oxygen measurements.  Since respiration of 
DOC is a significant fraction of oxygen consumption, correction for respiration of 
DOC was necessary in treatments with both POC and DOC to estimate the 
heterotrophic respiration rate of POC alone.  Keeping particles suspended and 
dispersed maximized oxygen delivery to the particles and provided the most realistic 
 
 115 
 
conditions for incubations.  Finally, long-term incubations provided insight to the 
continuum of POC lability.  I discuss each of these components of the method below.    
 
Small changes in oxygen concentration due to heterotrophic respiration of 
suspended POC are difficult to measure due to high DOC:POC ratios and low 
concentrations of POC.  I was unable to find any other studies of POC lability that 
corrected for respiration of DOC during incubations.  This correction is necessary 
because DOC consumption was a significant portion of total oxygen consumption in 
POC + DOC treatments (Fig. 4.2a, 4.3a).  Without correction for DOC respiration, 
oxygen consumption due to POC would have been overestimated by at least 25% at 
the beginning of the study and by 50% the end of the incubation.   
 
Although benthic POC with a mean particle size >125 ?m has been used for 
most studies of POC quality in streams (Fuss and Smock 1996; Webster et al. 1999; 
Bonin et al. 2000), most suspended POC in headwater streams is between 0.7 and 20 
?m in diameter (Kaplan et al. 2006).  Methods to collect POC with nets or sieves only 
concentrates particles ?15-20 ?m and recovery of gravity-filtered particles off glass 
fiber or membrane filters has proved to be challenging and ineffective (D. C. 
Richardson pers. obs.).  Using TFF, I was able to concentrate small particles well 
above stream concentrations and relative to DOC, causing greater changes in oxygen 
concentration due to heterotrophic consumption of POC than would be typically 
measured.  TFF might not be necessary to measure POC quality when POC 
concentrations are high such as in samples from large rivers like the Amazon or 
 
 116 
 
Mississippi, anthropogenically impacted areas, pond and reservoir outflows, or those 
collected during storms (Whiles and Dodds 2002; Dodds and Whiles 2004).  
However, in forested headwater streams with high DOC:POC ratios and low POC 
concentrations, concentration of suspended POC is necessary to adequately measure 
POC respiration rates.      
 
Although TFF was advantageous in elevating POC concentrations, recovery 
of POC from TFF (50-60%) was less than optimal.  This inefficiency led to more 
processing time and extended the length of time before the incubation could start.  
Poor recovery may be due to adsorption of particles to the membrane (Minor et al. 
1999), and I am currently investigating its causes and modifying the TFF procedure to 
maximize POC return.  Additionally, electrostatic gradients repel DOC molecules 
(Benner et al. 1997) and cause an unintended concentration of DOC by 60% (Aug 
2007; Table 4.2).  I am uncertain if TFF modified DOC quality because I could not 
detect significant differences in turnover times between pre- and post-TFF DOC in 
the DOC lability experiment (Table 4.4).  If TFF concentrates DOC without 
discrimination between DOC molecules of differing lability, then the respiration 
rates, corrected for initial DOC concentration, and turnover times should be equal 
between pre- and post-TFF DOC.  In this study, I accounted for this potential change 
in DOC quality by using post-TFF DOC as the DOC correction for all POC 
treatments.    
 
 
 117 
 
Rotation and mixing of BOD bottles using the incubation axle increased 
oxygen consumption and respiration rates over static bottles for both experiments, 
especially during the initial week (Fig. 4.2, 4.3).  Diffusion of oxygen to particle-
attached microorganisms and POC is faster for suspended rather than settled particles 
since layering in settled treatments will limit solute exchange to lower layers of 
particles (Ploug and Grossart 1999).  For example, Ploug and Grossart (1999, 2000) 
found that bacterial production on particles increased 5- to 10-fold when seston was 
kept in suspension during incubations compared to those under static conditions.  In 
addition to rotating samples, I added beads to some bottles to generate turbulence and 
break up flocs.  In initial experiments using the incubation axle, flocs, visible to the 
naked eye, formed in the BOD bottles while rotating on the incubation axle (D. C. 
Richardson pers. obs.).  These flocs likely form because of binding bacterial 
extracellular polymeric substances and charge attractions in quiescent water (Droppo 
2004), but I did not see them in any microscopy of WCC seston.  The beads increase 
internal mixing and break up flocs, but do not artificially increase substratum for 
bacterial growth and use of DOC (Fig. 4.3b).  Conditions of mass transfer and 
chemical microenvironments for particles are likely more similar to natural stream 
conditions using the incubation axle and beads due to suspension and mixing of 
particles.      
 
    When measuring the oxygen concentrations in each bottle, I used the 
Winkler method with auto-titration.  This is an efficient, accurate and precise method 
to measure oxygen concentration, especially when attempting to detect small amounts 
 
 118 
 
of oxygen consumption (Furuya and Harada 1995).  Auto-titration was accurate 
within ?0.06 mg O
2
 L
-1
 for Winkler measurements (Ploug et al. 2002), which is 
advantageous over previously used oxygen concentration measurements such as the 
Gilson respirometer (Webster et al.1999) or using polargraphic measurements of 
oxygen using probes, which can lack the precision and accuracy necessary to measure 
the small changes in oxygen concentration (CWT 2004).  This precision is especially 
important during the initial 24 hours of oxygen measurements when concentrations 
changed by <0.5 mg L
-1
 (Fig. 4.2a, 4.3a).   
 
 Incubations and measurements over 30+ days enable separation of high 
quality, labile organic matter from refractory organic matter (Fig. 4.2b, 4.3b).  Short-
term measurements are adequate for measuring initial rates but do not allow for 
separation of a continuous range of labilities including labile, resistant, and refractory 
classes (Berner 1995; Trumbore 2000; Kaplan and Newbold 2003).  Combining the 
range of lability classes with differential transport of suspended POC produces 
upstream-downstream linkages of varying scales.  Labile POC, with turnover times of 
less than 10 days (Table 4.4), transported under base-flow conditions could support 
metabolism in local downstream reaches.  Labile POC could also be transported long 
distances by storms and would support heterotrophic respiration in mid-order streams.  
Less labile material could be transported by storms and baseflow until the POC is 
used for metabolism in rivers far downstream (e.g. Cole and Caraco 2001; Mayorga 
et al. 2005) or is exported from the river network.     
   
 
 119 
 
 The observed difference in oxygen consumption between experiments (Jan 
2006 and Apr 2007) was most evident in the initial respiration rates (Fig. 4.5).  POC 
lability is likely linked to seasonal differences in the composition of seston, 
specifically increases in algal derived carbon.  In early April, the stream canopy is 
open, light levels are higher, and periphyton growth is at a maximum on the 
streambed leading to high chlorophyll a and carbon content of seston (Chapter 1; 
Pozo et al. 1994).  In this study, I saw a 2-fold increase in initial respiration rates from 
the January to the April experiment (Fig. 4.5).  This lability could be linked to an 
increase in high nutrient content organic matter (algal cells) (Chapter 3; Pozo et al. 
1994).  Similarly, in a eutrophic lake, suspended POC quality dramatically increased 
during warm season algal blooms (Parparov et al. 1998) and increasing nitrogen 
content of benthic POC was correlated with increasing heterotrophic respiration rates 
(Fuss and Smock 1996).       
 
Respiration rate comparisons ? I compared the respiration rates from both dispersed, 
suspended particle treatments to a range of POC respiration rates measured in a 
variety of small streams and organic matter from different aquatic environments 
(Table 4.5).  The respiration rates for POC measured using the method reported here 
were initially 15-fold greater than those reported by Webster et al. (1999) for benthic 
POC from a range of locations including Michigan, Oregon, Idaho, Virginia, and 
North Carolina streams (Fig. 4.5).  However, the initial rates from the current 
experiment were comparable to respiration rates for marine DOM from shallow water 
(Biddanda et al. 1994), eutrophic lake seston (Parparov et al. 1998), and diatom 
 
 120 
 
formed aggregates (Ploug and Grossart 2000).  After 25-27 days, the respiration rates 
had decreased to the range of literature reported rates for stream POC (Fig. 4.5; Fuss 
and Smock 1996), leaves and wood veneers incubated in the stream (Stelzer et al. 
2003), and wetland DOM (Opsahl 2005).  These comparisons support the hypothesis 
that high initial rates of respiration in this study were driven by algal sources with 
rapid biological turnover times (T
b-instant
<1 week, Fig. 4.5).  Respiration rates after one 
month were possibly driven by remaining semi-labile and refractory POC including 
mineral-organic complexes (i.e. clay particles with adsorbed organic carbon) and leaf 
fragments (T
b-instant
 >4 months, Fig. 4.5).   
     
The differences between measurements of suspended POC from this study and 
benthic POC from other studies (Table 4.5) appear to be related to both the ability of 
the method reported here to measure rapid initial respiration rates and to the greater 
lability of suspended POC relative to benthic POC.  This difference in lability may be 
related to particle size, i.e. the majority of benthic POC is >50 ?m, while the majority 
of suspended POC is <20 ?m (Minshall et al. 1983; Kaplan et al. 2002).  At baseflow, 
suspended POC likely represents a subset of small and labile or semi-labile benthic 
particles that is easily suspended and transported downstream including individual 
diatom cells, bacterial cells, leaf fragments and small clays with adsorbed organic 
matter.  Anthropogenic point sources, such as farm ponds, may also contribute labile 
POC during baseflow (Chapter 3).   
 
 
 121 
 
Directions for future research ? Over a 35 day period, I measured a range of 
respiration rates for suspended POC, from 10 to 300 mg O
2
 initial g C
-1
 day
-1
 with 
observed biological turnover times of approximately 10 days (Table 4.4).  The results 
from this study showed the continuous range of lability of POC and indicated that 
suspended POC has the potential to influence metabolism in reaches immediately 
downstream as well as rivers (Cole and Caraco 2001; Mayorga et al. 2005).  At low 
flows, small particles move downstream at 150 m day
-1
 (Newbold et al. 2005) and 
with a turnover time of 10 days (this study), the associated POC would be 
metabolized after traveling only 1.5 km downstream.  This travel distance is far 
shorter than the 42 km of travel distance estimated by Webster et al. (1999).  Using 
this method to measure suspended POC quality from a range of stream ecosystems, 
including streams with high primary production (e.g. meadow streams) or streams 
dominated by fluxes of terrestrially-derived organic matter (e.g. small forested 
headwaters), will help determine ranges of POC lability in stream ecosystems and 
quality of organic matter exported from headwaters.  Application of the method 
presented here to other stream ecosystems will help advance the understanding of the 
role of river networks in controlling the transformations of allochthonous and 
autochthonous POC transported from inland waters to the oceans (Cole et al. 2007). 
 
 122 
 
Tables 
Table 4.1. Experimental Design for January 2006 and April 2007 experiments.  ? 
indicates the treatment combination was used in that experiment; X indicates that 
there were particles in the bottle (POC), particles were kept in suspension (Incubation 
axle) and the bottle was mixed and particles were dispersed (Beads).   
 
Treatment Jan 2006 Apr 2007 POC 
Incubation 
axle Beads 
Dispersed, suspended 
particles 
? ? X X X 
Suspended particles ?  X X  
Settled particles ? ? X   
DOC control ? ?  X  
Dispersed DOC 
control 
 ?  X X 
 
 123 
 
Table 4.2. Initial particulate and dissolved organic carbon concentrations (mean ? 
SD) for all experiments. N.A. indicates that POC concentration was not measured.  
 
      
Experiment Treatment POC (mg L
-1
) DOC (mg L
-1
) 
Jan 2006 DOC control <0.03 1.5 ? 0.05 
 Settled particles 0.98 ? 0.04 1.5 ? 0.02 
 Suspended particles 0.92 ? 0.10 1.5 ? 0.04 
 
Dispersed, suspended 
particles 
0.99 ? 0.06 1.5 ? 0.02 
Apr 2007 DOC control <0.03 2.1 ? 0.08 
 
Dispersed DOC 
control 
<0.03 2.2 ? 0.09 
 Settled particles 1.12 ? 0.08 2.1 ? 0.08 
 
Dispersed, suspended 
particles 
0.97 ? 0.07 2.1 ? 0.05 
DOC lability preTFF N.A. 1.3 ? 0.02 
 postTFF N.A. 2.1 ? 0.08 
 
 
 124 
 
Table 4.3. Oxygen consumption curve (O
2
 = c
1
 e
-k X t
 + c
0
) parameter fits for each 
treatment and experiment.  Estimates (95% confidence intervals) for parameters k, c
0
, 
and c
1
 included with model fits (r
2
).  All models were significant fits (p<0.001). 
 
Exp. Treatment 
k  
(hour
-1
) 
c
1
  
(mg O
2
 L
-1
) 
c
0
  
(mg O
2
 L
-1
) r
2
 
Jan 
2006 
DOC control 
0.0013  
(0.0003-0.0023)
2.0  
(1.1-2.8)
6.3  
(5.4-7.2) 
0.96 
 Settled particles 
0.0016  
(0.0013-0.0019)
3.1  
(2.8-3.5)
5.3  
(4.9-5.7) 
0.99 
 Suspended particles 
0.0023  
(0.0013-0.0033)
3.3  
(2.7-3.8)
5.0  
(4.4-5.6) 
0.97 
 
Dispersed, suspended 
particles 
0.0028  
(0.0020-0.0036)
3.6 
(3.3-4.0)
4.8  
(4.4-5.2) 
0.98 
Apr 
2007 
DOC control 
0.0014  
(0.0008-0.0021)
3.6  
(2.8-4.5)
4.5  
(3.6-5.5) 
0.98 
 
Dispersed DOC 
control 
0.0018 
(0.0012-0.0023)
4.0 
(3.4-4.6)
4.2  
(3.6-4.9) 
0.98 
 Settled particles 
0.0030  
(0.0027-0.0034)
5.1  
(4.9-5.3)
3.4  
(3.2-3.6) 
0.99 
 
Dispersed, suspended 
particles 
0.0030  
(0.0024-0.0036)
6.1  
(5.7-6.5)
2.5  
(2.1-2.9) 
0.99 
DOC 
lability 
preTFF 
0.0090 
(0.0011-0.0168)
0.6 
(0.5-0.8)
7.8  
(7.6-7.9) 
0.96 
 postTFF 
0.0033  
(-0.0026-0.0091)
1.5  
(-0.1-3.1)
6.9 
(5.2-8.6) 
0.92 
 
 125 
 
Table 4.4. Temporal loss rate coefficients (k) and biological turnover times (T
b-
observed
) for all treatments with either POC corrected for DOC respiration or DOC 
alone.  Loss rate coefficients of treatments with the same letter were not significantly 
different when analyzed using all pairwise t-tests within the experiment where 
p<0.05.  T-tests were completed using estimates of k and standard errors generated 
from nonlinear least squares analysis.   
    
Exp. Treatment 
Carbon 
Source k (hour
-1
) 
Biological 
Turnover Time 
(T
b-observed
 = 1/k, 
days) 
Within 
Experiment 
Grouping 
Jan 2006 DOC control DOC 0.0013 31.6  a 
 Settled particles POC 0.0020 21.3  a 
 Suspended particles POC 0.0032 12.9  b 
 
Dispersed, suspended 
particles 
POC 0.0040 10.5  c 
Apr 2007 DOC control DOC 0.0014 29.8  a 
 
Dispersed DOC 
control 
DOC 0.0018 23.7  a 
 Settled particles POC 0.0056 7.5  b 
 
Dispersed, suspended 
particles 
POC 0.0046 9.0  b 
DOC 
lability 
preTFF DOC 0.0090 12.8  a 
 postTFF DOC 0.0033 4.6  a 
 
 
 126 
 
Table 4.5. Ranges of respiration rates for different carbon sources.  Biological 
turnover time (T
b-instant
) was calculated from respiration rates assuming a respiratory 
quotient of 1.    
 
Carbon Source 
Respiration rate  
(mg O
2
 g C
-1
 day
-1
)  
Biological 
turnover time 
(T
b-instant
, days)  Citation 
Stream suspended POC 10 - 300 9-270 This study 
Benthic POC 1 - 21 130-2660 Webster et al. 1999
a
 
Benthic POC 5 - 20 130-530 Bonin et al. 2000 
Benthic POC, Leaves, Wood 3 - 7 380-890 Fuss and Smock 1996 
Leaves and Wood 2 - 22 120-1330 Stelzer et al. 2003 
Wetland DOC 1 - 19 140-2660 Opsahl 2005 
Marine DOC 101 - 143 19-26 Biddanda et al. 1994 
Lake suspended POC 69 - 187 14-39  Parparov et al. 1998
a
 
a
See citation for literature summary of respiration rates 
 
 
 
 
 
 
 127 
 
Figure Captions 
 
Figure 4.1.  The left side of the incubation axle in the temperature controlled water 
bath. 
 
Figure 4.2. (a) Oxygen consumption, where lines indicates model fit (O
2
 = c
1
 e
-k * t
 + 
c
0
), and  (b) respiration rates for Jan 2006 experiment.  POC respiration rates are 
corrected for DOC respiration.  Treatments are DOC control (DOC), settled particles 
(Pset), suspended particles (Psus), and dispersed, suspended particles (Psus,dis).  
Error bars are standard errors. 
 
Figure 4.3. (a) Oxygen consumption, where lines indicates model fit (O
2
 = c
1
 e
-k * t
 + 
c
0
), and (b) respiration rates for Apr 2007 experiment.  POC respiration rates are 
corrected for DOC respiration.  Treatments are DOC control (DOC), dispersed DOC 
control (DOCdis), settled particles (Psus), and dispersed, suspended particles 
(Psus,dis).  Error bars are standard errors.  
 
Figure 4.4. (a) Oxygen consumption, where line indicates model fit (O
2
 = c
1
 e
-k * t
 + 
c
0
), and (b) respiration rates for DOC lability experiment. Treatments are pre-
tangential flow filtration (preTFF) and post-tangential flow filtration (postTFF).  
Error bars are standard errors. 
 
 
 128 
 
Figure 4.5. Respiration rates for the dispersed, suspended particle treatments from 
Jan 2006 and Apr 2007 experiments.  The grey box represents the range of measured 
respiration rates from Webster et al. (1999).  Right y-axis values indicate biological 
turnover time (T
b-instant
) assuming respiratory quotient of 1 (Webster et al. 1999).
Figures
Figure 4.1
a
Respiration Rate
(mg O
2
 initial g C
-1
 day
-1
)
b
Time (day)
O
2
 concentration
(mg L
-1
)
8.5
8.0
7.5
7.0
6.5
6.0
5.5
5.0
200
160
120
80
40
0
0 10 20 30 40
DOC
Pset
Psus
Psus,dis
Figure 4.2
a
b
Respiration Rate
(mg O
2
 initial g C
-1
 day
-1
)
O
2
 concentration
(mg L
-1
)
Time (day)
8.5
7.5
6.5
5.5
4.5
350
300
250
200
0
0 10 20 30 40
3.5
2.5
150
100
50
DOC
Psus
Psus,dis
DOCdis
Figure 4.3
a
b
Respiration Rate
(mg O
2
 initial g C
-1
 day
-1
)
O
2
 concentration
(mg L
-1
)
Time (day)
8.5
7.5
8.0
90
0 2 104 6 8
80
70
60
50
40
30
20
10
0
PreTFF
PostTFF
Figure 4.4
3
27
266
2664
Respiration Rate
(mg O
2
 initial g C
-1
 day
-1
)
Time (day)
0 10
10
Apr 2007
Jan 2006
T
urnover time
(day)
1
100
1000
20 30 40
Figure 4.5
 
 134 
 
Appendices 
 
Appendix 1 
Derivation of Equation 4.2: O
2 corrected
 = A
exp
 ? (c
1doc
 e
-kdoc *  t
 + c
0doc
)
 
+ (c
1
 e
-k *  t
 + c
0
) 
 
P = c
1
 e
-k *  t
 + c
0
  
P is the equation for oxygen consumption curve from POC + DOC with parameters 
estimated from least squares analysis (Figure A1). 
 
D = c
1doc
 e
-kdoc *  t
 + c
0doc 
D is the equation for oxygen consumption curve from DOC alone treatments for each 
experiment with parameters estimated from least squares analysis (Fig. A1). 
 
X(t) = A
exp
 ? P 
X(t) is the total oxygen consumed by respiration of both POC and DOC at time (t) 
and A
exp
 is the initial oxygen concentration for each experiment (Fig. A1). 
 
Y(t) = A
exp
 ? D 
Y(t) is the total oxygen consumed by respiration of only DOC at time (t) (Fig. A1). 
 
X(t) ? Y(t) = the oxygen consumed by POC alone at time (t). 
 
N(t) is the oxygen consumption curve for POC alone (Fig. A1). 
 
 135 
 
N(t) = A
exp
 ? {X(t) ? Y(t)}  
 
Simplify and rearrange N(t): 
N(t) = A
exp
 ? { A
exp
 ? P ? (A
exp
 ? D)}  
N(t) = A
exp
 ? { A
exp
 ? P ?  A
exp
 + D }  
N(t) = A
exp
 ?    A
exp
 + P +  A
exp
 ? D  
N(t) = A
exp
 ? D + P  
 
Hence eq. 4.2:  O
2 corrected
 = A
exp
 ? (c
1doc
 e
-kdoc *  t
 + c
0doc
)
 
+ (c
1
 e
-k *  t
 + c
0
) 
D
P
Y(t)
X(t)
Aexp
N(t)
Time (day)
O
2
 concentration 
(mg L
-1
)
Figure A1.  Theoretical oxygen consumption curves for DOC 
alone (D), POC + DOC (P) and estimation of the DO consump-
tion curve for POC alone (N(t)).  A
exp
 is the initial oxygen 
concentration and Y(t) and X(t) are the total oxygen consumed by 
respiration for each curve.
Appendix 1 Figure
 
 137 
 
Literature Cited 
 
Angradi, T. R. 1993. Chlorophyll content of seston in a regulated Rocky Mountain river, 
Idaho, USA. Hydrobiologia 259:39-46. 
 
Artigas, M. L., J. E. Skjaeraasen, A. C. Utne-Palm, and T. Nilsen. 2005. Recovery from 
handling stress in Gadus morhua. Journal of Fish Biology 67:384-391. 
 
Aufdenkampe, A. K., J. I. Hedges, J. E. Richey, A. V. Krusche, and C. A. Llerena. 2001. 
Sorptive fractionation of dissolved organic nitrogen and amino acids onto fine 
sediments within the Amazon Basin. Limnology and Oceanography 46:1921-1935. 
 
Aufdenkampe, A. K., E. Mayorga, J. I. Hedges, C. Llerena, P. D. Quay, J. Gudeman, A. V. 
Krusche, and J. E. Richey. 2007. Organic matter in the Peruvian headwaters of the 
Amazon: Compositional evolution from the Andes to the lowland Amazon mainstem. 
Organic Geochemistry 38:337-364. 
 
Baisden, W. T., R. Amundson, D. L. Brenner, A. C. Cook, C. Kendall, and J. W. Harden. 
2002. A multi-isotope C and N modeling analysis of soil organic matter turnover and 
transport as a function of soil depth in a California annual grassland soil 
chronosequence. Global Biogeochemical Cycles 16:1135, 
doi:10.1029/2001GB001823. 
 
Battin, T. J., L. A. Kaplan, S. Findlay, C. S. Hopkinson, E. Marti, A.I. Packman, J. D. 
Newbold, and F. Sabater.  2008. Biophysical controls on organic carbon fluxes in 
fluvial networks.  Nature Geosciences 1:95-100. 
 
Battin, T. J., L. A. Kaplan, J. D. Newbold, and C. M. E. Hansen. 2003. Contributions of 
microbial biofilms to ecosystem processes in stream mesocosms. Nature 426:439-
442. 
 
Benfield, E. F. 1997. Comparison of litterfall input to streams. Journal of the North American 
Benthological Society 16:104-108. 
 
Benke, A. C., and J. B. Wallace. 1980. Trophic basis of production among net-spinning 
caddisflies in a southern Appalachian stream. Ecology 61:108-118. 
 
Benner, R., B. Biddanda, B. Black, and M. McCarthy. 1997. Abundance, size distribution, 
stable carbon and nitrogen isotopic compositions of marine organic matter isolated by 
tangential flow ultrafiltration. Marine Chemistry 57:243-263. 
 
Benner, R., M. L. Fogel, E. K. Sprague, and R. E. Hodson. 1987. Depletion of 
13
C in lignin 
and its implications for stable carbon isotope studies. Nature 329:708-710. 
 
Benner, R., S. Opsahl, G. Chin-Leo, J. E. Richey, and B. R. Forsberg. 1995. Bacterial carbon 
metabolism in the Amazon River system. Limnology and Oceanography 40:1262-
1270. 
 
 
 138 
 
Berner, R. A. 1995. Sedimentary organic matter preservation - An assessment and speculative 
synthesis - a comment. Marine Chemistry 49:121-122. 
 
Biddanda, B., S. Opsahl, and R. Benner. 1994. Plankton respiration and carbon flux through 
bacterioplankton on the Louisiana Shelf. Limnology and Oceanography 39:1259-
1275. 
 
Biggs, B. J. F., and C. Kilroy. 2000. Stream periphyton monitoring manual. New Zealand 
Ministry of the Environment, Christchurch, New Zealand. Available online: 
http://www.niwascience.co.nz/ncwr/tools/periphyton. 
 
Bonin, H. L., R. P. Griffiths, and B. A. Caldwell. 2000. Nutrient and microbiological 
characteristics of fine benthic organic matter in mountain streams. Journal of the 
North American Benthological Society 19:235-249. 
 
Brittain, J. E., and T. J. Eikeland. 1988. Invertebrate drift - a review. Hydrobiologia 166:77-
93. 
 
Bryan, J. R., J. P. Riley, and P. J. L. Williams. 1976. Winkler procedure for making precise 
measurements of oxygen concentration for productivity and related studies. Journal 
of Experimental Marine Biology and Ecology 21:191-197. 
 
Buffam, I., J. N. Galloway, L. K. Blum, and K. J. McGlathery. 2001. A stormflow/baseflow 
comparison of dissolved organic matter concentrations and bioavailability in an 
Applachian stream. Biogeochemistry 53:269-306.  
 
Burns, D. A., P. S. Murdoch, G. B. Lawrence, and R. L. Michel. 1998. Effect of groundwater 
springs on NO
3
 concentrations during summer in Catskill Mountain streams. Water 
Resources Research 34: 1987-1996. 
 
Carling, P. A. 1983. Particulate dynamics, dissolved and total load, in two small basins, 
northern Pennines, UK. Hydrological Sciences Journal 28:355-375. 
 
Clean Water Team (CWT). 2004. Dissolved oxygen measurements, principles, and methods, 
DQM IP-3.1.1.  In: The Clean Water Team guidance compendium for watershed 
monitoring and assessment, version 2.0.  Division of Water Quality, California State 
Water Resources Control Board (SWRCB), Sacramento, California.    
 
Cloern, J. E., C. Grenz, and L. Vidergar-Lucas. 1995. An empirical model of the 
phytoplankton chlorophyll:carbon ratio - the conversion factor between productivity 
and growth rate. Limnology and Oceanography 40:1313-1321. 
 
Cole, J. J., and N. F. Caraco. 2001. Carbon in catchments: Connecting terrestrial carbon 
losses with aquatic metabolism. Marine and Freshwater Research 52:101-110. 
 
Cole, J. J., Y. T. Prairie, N. F. Caraco, W. H. McDowell, L. J. Tranvik, R. G. Striegl, C. M. 
Duarte, P. Kortelainen, J. A. Downing, J. J. Middelburg, and J. Melack. 2007. 
Plumbing the global carbon cycle: Integrating inland waters into the terrestrial carbon 
budget. Ecosystems 10:171-184. 
 
 
 139 
 
Coplen, T. B., W. A. Brand, M. Gehre, M. Gr?ning, H. A. J. Meijer, B. Toman, and R. M. 
Verkouteren. 2006. New guidelines for ?
13
C measurements. Analytical Chemistry 
78:2439-2441. 
 
Corn, P. S., R. B. Bury, and E. J. Hyde. 2003.  Conservation of stream amphibians in North 
America. In, Semlitsch, R. (ed.), Amphibian Conservation. Smithsonian Institution. 
Washington, District of Columbia.   
 
Cowan, C. A., and B. L. Peckarsky. 1994. Diel feeding and positioning periodicity of a 
grazing mayfly in a trout stream and a fishless stream. Canadian Journal of Fisheries 
and Aquatic Sciences 51:450-459. 
 
Creed, R. P., and J. M. Reed. 2004. Ecosystem engineering by crayfish in a headwater stream 
community. Journal of the North American Benthological Society 23:224-236. 
 
Cross, W. F., J. P. Benstead, P. C. Frost, and S. A. Thomas. 2005. Ecological stoichiometry 
in freshwater benthic systems: Recent progress and perspectives. Freshwater Biology 
50:1895-1912. 
 
Cushing, C. E., G. W. Minshall, and J. D. Newbold. 1993. Transport Dynamics of Fine 
Particulate Organic-Matter in 2 Idaho Streams. Limnology and Oceanography 
38:1101-1115. 
 
Czikowsky M. J., and D. R. Fitzjarrald. 2004. Evidence of seasonal changes in 
evapotranspiration in eastern U.S. hydrological records. Journal of 
Hydrometeorology 5:974-988. 
 
Davic, R. D., and H. H. Welsh. 2004. On the ecological roles of salamanders. Annual Review 
of Ecology Evolution and Systematics 35:405-434. 
 
Decho, A. W., and G. R. Lopez. 1993. Exopolymer microenvironments of microbial flora: 
Multiple and interactive effects on trophic relationships. Limnology and 
Oceanography 38:1633-1645. 
 
Dodds, W. K., and M. R. Whiles. 2004. Quality and quantity of suspended particles in rivers: 
Continent-scale patterns in the United States. Environmental Management 33: 355-
367. 
 
Downs, J. N., and C. J. Lorenzen. 1985. Carbon:pheopigment ratios of zooplankton fecal 
pellets as an index of herbivorous feeding. Limnology and Oceanography 30:1024-
1036. 
 
Draper, N. R., and H. Smith. 1998. Applied regression analysis. J. W. Wiley, Hoboken, New 
Jersey. 
 
Droppo, I. G. 2004. Structural controls on floc strength and transport. Canadian Journal of 
Civil Engineering 31:569-578. 
 
Dunne, T., and L. B. Leopold. 1978. Water in environmental planning. W.H. Freeman and 
Company, San Francisco, California. 
 
 
 140 
 
Duval, T. P., and A. R. Hill. 2006. Influence of stream bank seepage during low-flow 
conditions on riparian zone hydrology. Water Resources Research 42:W10425, 
doi:10.1029/2006WR004861. 
 
Elliott, J. M. 2000. Contrasting diel activity and feeding patterns of four species of 
carnivorous stoneflies. Ecological Entomology 25:26-34. 
 
Elliott, S. R., R. J. Naiman, and P. A. Bisson. 2004. Riparian influences on the biophysical 
characteristics of seston in headwater streams. Northwest Science 78:150-157. 
 
Elser, J. J., W. F. Fagan, R. F. Denno, D. R. Dobberfuhl, A. Folarin, A. Huberty, S. 
Interlandi, S. S. Kilham, E. McCauley, K. L. Schulz, E. H. Siemann, and R. W. 
Sterner. 2000. Nutritional constraints in terrestrial and freshwater food webs. Nature 
408:578-580. 
 
Finlay, J.  C., S. Khandwala, and M. E. Power. 2002. Spatial scales of carbon flow in a river 
food web. Ecology 83:1845-1859. 
 
Finlay, J. C., M. E. Power, and G. Cabana. 1999. Effects of water velocity on algal carbon 
isotope ratios: Implications for river food web studies. Limnology and Oceanography 
44:1198-1203. 
 
Fisher, S. G., L. J. Gray, N. B. Grimm, and D. E. Busch. 1982. Temporal succession in a 
desert stream ecosystem following flash flooding. Ecological Monographs 52:93-
110. 
 
Fisher, S. G., and G. E. Likens. 1973. Energy flow in Bear Brook, New Hampshire: 
Integrative approach to stream ecosystem metabolism. Ecological Monographs 
43:421-439. 
 
Fisher, S. G., R. A. Sponseller, and J. B. Heffernan. 2004. Horizons in stream 
biogeochemistry: Flowpaths to progress. Ecology 85:2369-2379. 
 
Flecker, A. S. 1992. Fish predation and the evolution of invertebrate drift periodicity: 
evidence from neotropical streams. Ecology 73:438-448. 
 
Fry, B. 2007. Stable isotope ecology. Springer, New York. 
 
Furuya, K., and K. Harada. 1995. An automated precise Winkler titration for determining 
dissolved oxygen on board ship. Journal of Oceanography 51:375-383. 
 
Fuss, C. L., and L. A. Smock. 1996. Spatial and temporal variation of microbial respiration 
rates in a blackwater stream. Freshwater Biology 36:339-349. 
 
Geider, R. J., and J. La Roche. 2002. Redfield revisited: Variability of C:N:P in marine 
microalgae and its biochemical basis. European Journal of Phycology 37:1-17. 
 
Golladay, S. W. 1997. Suspended particulate organic matter concentration and export in 
streams. Journal of the North American Benthological Society 16:122-131. 
 
 
 141 
 
Golladay, S. W., K. Watt, S. Entrekin, and J. Battle. 2000. Hydrologic and geomorphic 
controls on suspended particulate organic matter concentration and transport in 
Ichawaynochaway Creek, Georgia, USA. Archiv Fur Hydrobiologie 149:655-678. 
 
Graneli, W., and E. Graneli. 1991. Automatic Potentiometric Determination of Dissolved-
Oxygen. Marine Biology 108:341-348. 
 
Gurnell, A. M. 2007. Analogies between mineral sediment and vegetative particle dynamics 
in fluvial systems. Geomorphology 89:9-22. 
 
Hampton, S. E. and I. C. Duggan. 2003. Diel habitat shifts of macrofauna in a fishless pond.  
Marine and Freshwater Research 54:797-805. 
 
Hedges, J. I., J. A. Baldock, Y. Gelinas, C. Lee, M. L. Peterson, and S. G. Wakeham. 2002. 
The biochemcial and elemental compositions of marine plankton: A NMR 
perspective. Marine Chemistry 78:47-63. 
 
Hedges, J. I., W. A. Clark, P. D. Quay, J. E. Richey, A. H. Devol, and U. D. Santos. 1986. 
Compositions and fluxes of particulate organic material in the Amazon River. 
Limnology and Oceanography 31:717-738. 
 
Hedin, L. O. 1990. Factors controlling sediment community respiration in woodland stream 
ecosystems. Oikos 57:94-105. 
 
Hieber, M., C. T. Robinson, and U. Uehlinger. 2003. Seasonal and diel patterns of 
invertebrate drift in different alpine stream types. Freshwater Biology 48:1078-1092. 
 
Hill, W. R., P. J. Mulholland, and E. R. Marzolf. 2001. Stream ecosystem responses to forest 
leaf emergence in spring. Ecology 82:2306-2319. 
 
Hillerbrand, H., and U. Sommer. 1999. The nutrient stoichiometry of benthic microalgal 
growth: Redfield proportions are optimal. Limnology and Oceanography 44:440-446. 
 
Ho, T. Y., A. Quigg, Z. V. Finkel, A. J. Milligan, K. Wyman, P. G. Falkowski, and F. M. M. 
Morel. 2003. The elemental composition of some marine phytoplankton. Journal of 
Phycology 39:1145-1159. 
 
Jones, C. G., J. H. Lawton, and M. Shachak. 1997. Positive and negative effects of organisms 
as physical ecosystem engineers. Ecology 78:1946-1957. 
 
Kaiser, E., D. B. Arscott, K. Tockner, and B. Sulzberger. 2004. Sources and distribution of 
organic carbon and nitrogen in the Tagliamento River, Italy. Aquatic Sciences 
66:103-116. 
 
Kaplan, L. A. 1992. Comparison of High-Temperature and Persulfate Oxidation Methods for 
Determination of Dissolved Organic-Carbon in Fresh-Waters. Limnology and 
Oceanography 37:1119-1125. 
 
Kaplan, L. A., and T. L. Bott. 1982. Diel fluctuations of DOC generated by algae in a 
piedmont stream. Limnology and Oceanography 27:1091-1100. 
 
 
 142 
 
Kaplan, L. A., R. A. Larson, and T. L. Bott. 1980. Patterns of dissolved organic carbon in 
transport. Limnology and Oceanography 25:1034-1043. 
 
Kaplan, L. A., and J. D. Newbold. 1995. Measurement of streamwater biodegradable 
dissolved organic carbon with a plug-flow bioreactor. Water Research 29:2696-2706. 
 
Kaplan, L. A., and J. D. Newbold. 2003. Monomers in Stream Ecosystem Metabolism. Pages 
97-119 in S. Findlay and R. Sinsabaugh, editors. Aquatic Ecosystems: Interactivity of 
Dissolved Organic Matter. Academic Press, New York. 
 
Kaplan, L. A., J. D. Newbold, D. J. Van Horn, C. L. Dow, A. K. Aufdenkampe, and J. K. 
Jackson. 2006. Organic matter transport in New York City drinking-water-supply 
watersheds. Journal of the North American Benthological Society 25:912-927. 
 
Kaplan, L. A., Standley, L. J., J. D. Newbold, J. H. Standridge, A. L. Mager, S. M. Kluender, 
L. L. Peterson, D. B. Smith, W. C. Hession, and P. Luitweiler. 2002. Evaluation of 
sources of pathogens and NOM in watersheds.  Awwa Research Foundation and 
American Water Works Association, Denver, Colorado.  
 
Kendall, C., S. R. Silva, and V. J. Kelly. 2001. Carbon and nitrogen isotopic compositions of 
particulate organic matter in four large river systems across the United States. 
Hydrological Processes 15:1301-1346. 
 
Kerner, M., H. Hohenberg, S. Ertl, M. Reckermann, and A. Spitzy. 2003. Self-organization of 
dissolved organic matter to micelle-like microparticles in river water. Nature 
422:150-154. 
 
Kratzer, E. B., J. K. Jackson, D. B. Arscott, A. K. Aufdenkampe, C. L. Dow, L. A. Kaplan, J. 
D. Newbold, and B. W. Sweeney. 2006. Macroinvertebrate distribution in relation to 
land use and water chemistry in New York City drinking-water-supply watersheds. 
Journal of the North American Benthological Society 25:954-976. 
 
Lannoo, M.J. (ed.). 2005. Amphibian declines: the conservation status of North American 
species. University of California Press. Berkeley, California. 
 
Leonardos, N., and R. J. Geider. 2004. Responses of elemental and biochemical composition 
of Chaetoceros muelleri to growth under varying light and nitrate: phosphate supply 
ratios and their influence on critical N : P. Limnology and Oceanography 49:2105-
2114. 
 
Lorenzen, C. J. 1967. Determination of chlorophyll and pheopigments: Spectrophotometric 
equations. Limnology and Oceanography 12:343-345. 
 
Lush, D. L., and H. B. N. Hynes. 1973. Formation of particles in freshwater leachates of dead 
leaves. Limnology and Oceanography 18:968-977. 
 
Mackay, W. P., J. Zak, and W. G. Whitford. 1992. Litter decomposition in a Chihuahuan 
desert playa. American Midland Naturalist 128:89-94. 
 
Manly, B. F. J. 2006. Randomization, bootstrap and Monte Carlo methods in biology.  
Chapman and Hall, London.   
 
 143 
 
 
Matthews, K. R. 2003. Response of mountain yellow-legged frogs, Rana muscosa, to short 
distance translocation. Journal of Herpetology 37:621-626. 
 
Mayer, L. M., R. G. Keil, S. A. Macko, S. B. Joye, K. C. Ruttenberg, and R. C. Aller. 1998. 
Importance of suspended particulates in riverine delivery of bioavailable nitrogen to 
coastal zones. Global Biogeochemical Cycles 12:573-579. 
 
Mayorga, E., A. K. Aufdenkampe, C. A. Masiello, A. V. Krusche, J. I. Hedges, P. D. Quay, J. 
E. Richey, and T. A. Brown. 2005. Young organic matter as a source of carbon 
dioxide outgassing from Amazonian rivers. Nature 436:538-541. 
 
McCutchan, J. H., and W. M. Lewis. 2002. Relative importance of carbon sources for 
macroinvertebrates in a Rocky Mountain stream. Limnology and Oceanography 
47:742-752. 
 
McNair, J. N., and J. D. Newbold. 2001. Turbulent transport of suspended particles and 
dispersing benthic organisms: the hitting-distance problem for the local exchange 
model. Journal of Theoretical Biology 209:351-369. 
 
Menninger, H. L., and M. A. Palmer. 2007. Herbs and grasses as an allochthonous resource in 
open-canopy headwater streams. Freshwater Biology 52:1689-1699. 
 
Mermillod-Blondin, F. and Rosenberg, R. 2006. Ecosystem engineering: The impact of 
bioturbation on biogeochemical processes in marine and freshwater benthic habitats.  
Aquatic Sciences 68:434-442 
 
Merritt, R. and Cummins, K. 1995. Aquatic insects of North America.  Kendall/Hunt 
Publishing Company, Dubuque, Idaho.  
 
Meybeck, M. 1982. Carbon, nitrogen, and phosphorus transport by world rivers. American 
Journal of Science 282:401-450. 
 
Meyer, J. L., and J. B. Wallace. 2001. Lost linkages and lotic ecology: rediscovering small 
streams. Pages 295-317 in M. C. Press, N. J. Huntly, and S. Levin, editors. Ecology: 
Achievement and Challenge. Blackwell Scientific, Oxford, United Kingdom. 
 
Meysman, F.J.R., J.J. Middelburg, and Heip, C.H.R. 2006. Bioturbation: a fresh look at 
Darwin's last idea. Trends in Ecology & Evolution 21:688-695.  
 
Minor, E. C., T. I. Eglinton, J. J. Boon, and R. Olson. 1999. Protocol for the characterization 
of oceanic particles via flow cytometric sorting and direct temperature-resolved mass 
spectrometry. Analytical Chemistry 71:2003-2013. 
 
Minor, E. C., and P. S. Nallathamby. 2004. "Cellular" vs. "detrital" POM: a preliminary study 
using fluorescent stains, flow cytometry, and mass spectrometry. Marine Chemistry 
92:9-21. 
 
Minshall, G. W., R. C. Petersen, K. W. Cummins, T. L. Bott, J. R. Sedell, C. E. Cushing, and 
R. L. Vannote. 1983. Interbiome Comparison of Stream Ecosystem Dynamics. 
Ecological Monographs 53:1-25. 
 
 144 
 
 
Molla, S., S. Robles, and C. Casado. 2006. Seasonal variability of particulate organic matter 
in a mountain stream in central Spain. International Review of Hydrobiology 91:406-
422. 
 
Monaghan, M. T., S. A. Thomas, G. W. Minshall, J. D. Newbold, and C. E. Cushing. 2001. 
The influence of filter-feeding benthic macroinvertebrates on the transport and 
deposition of particulate organic matter and diatoms in two streams. Limnology and 
Oceanography 46:1091-1099. 
 
Moore, J. W. 2006. Animal ecosystem engineers in streams. Bioscience 56:237-246. 
 
Moore, J. W., D. E. Schindler, and M. D. Scheuerell. 2004. Disturbance of freshwater 
habitats by anadromous salmon in Alaska. Oecologia 139:298-308. 
 
Mosisch, T. D., and S. E. Bunn. 1997. Temporal patterns of rainforest stream epilithic algae 
in relation to flow-related disturbance. Aquatic Botany 58:181-193. 
 
Mulholland, P. J., A. D. Steinman, A. V. Palumbo, J. W. Elwood, and D. B. Kirschtel. 1991. 
Role of nutrient cycling and herbivory in regulating periphyton communities in 
laboratory streams. Ecology 72:966-982. 
 
Mulholland, P. J., J. L. Tank, D. M. Sanzone, W. M. Wolheim, B. J. Peterson, J. R. Webster 
and J. L. Meyer. 2000. Food resources of stream macroinvertebrates determined by 
natural-abundance stable C and N isotopes and a 
15
N tracer addition.  Journal of 
North American Benthological Society 19:145-157. 
 
Naiman, R. J. 1982. Characteristics of sediment and organic-carbon export from pristine 
boreal forest watersheds. Canadian Journal of Fisheries and Aquatic Sciences 
39:1699-1718. 
 
Naiman, R. J., and J. R. Sedell. 1979. Characterization of Particulate Organic-Matter 
Transported by Some Cascade Mountain Streams. Journal of the Fisheries Research 
Board of Canada 36:17-31. 
 
Newbold, J. D., T. L. Bott, L. A. Kaplan, B. W. Sweeney, and R. L. Vannote. 1997. Organic 
matter dynamics in White Clay Creek, Pennsylvania, USA. Journal of the North 
American Benthological Society 16:46-50. 
 
Newbold, J. D., S. A. Thomas, G. W. Minshall, C. E. Cushing, and T. Georgian. 2005. 
Deposition, benthic residence, and resuspension of fine organic particles in a 
mountain stream. Limnology and Oceanography 50:1571-1580. 
 
Nogaro, G., Mermillod-Blondin, F., Carcaillet, F. F., Gaudet, J. P., Lafont, M., Gibert, J. 
2006. Invertebrate bioturbation can reduce the clogging of sediment: an experimental 
study using infiltration sediment columns. Freshwater Biology 51:1458?1473.  
 
Onstad, G. D., D. E. Canfield, P. D. Quay, and J. I. Hedges. 2000. Sources of particulate 
organic matter in rivers from the continental USA: Lignin phenol and stable carbon 
isotope compositions. Geochimica et  Cosmochimica Acta 64:3539-3546. 
 
 
 145 
 
Opsahl, S. P. 2005. Organic carbon composition and oxygen metabolism across a gradient of 
seasonally inundated limesink and riparian wetlands in the southeast Coastal Plain, 
USA. Biogeochemistry 76:47-68. 
 
Ostrofsky, M. L. 1997. Relationship between chemical characteristics of autumn-shed leaves 
and aquatic processing rates. Journal of the North American Benthological Society 
16:750-759. 
 
Palmer, M. A., A. P. Covich, B. J. Finlay, J. Gibert, K. D. Hyde, R. K. Johnson, T. Kairesalo, 
S. Lake, C. R. Lovell, R. J. Naiman, C. Ricci, F. Sabater, and D. Strayer. 1997. 
Biodiversity and ecosystem processes in freshwater sediments. Ambio 26:571-577. 
 
Palmer, M. A., D. L. Strayer, and S. D. Rundle. 2006. Meiofauna, pgs. 415-434.  In F. R. 
Hauer and G. A. Lamberti (eds.), Methods in stream ecology, 2
nd
 ed. Elsevier, San 
Diego, California.  
 
Parkes, A. H., J. Kalff, J. Boisvert, and G. Cabana. 2004. Feeding by black fly (Diptera : 
Simuliidae) larvae causes downstream losses in phytoplankton, but not bacteria. 
Journal of the North American Benthological Society 23:780-792. 
 
Parparov, A. S., T. Berman, H. P. Grossart, and M. Simon. 1998. Metabolic activity 
associated with lacustrine seston. Aquatic Microbial Ecology 15:77-87. 
 
Paul, M. J., and R. O. Hall. 2002. Particle transport and transient storage along a stream-size 
gradient in the Hubbard Brook Experimental Forest. Journal of the North American 
Benthological Society 21:195-205. 
 
Peckarsky, B. L., S. D. Cooper, and A. R. McIntosh. 1997. Extrapolating from individual 
behavior to populations and communities in streams. Journal of the North American 
Benthological Society 16:375-390. 
 
Peterson, C. G., and R. J. Stevenson. 1992. Resistance and resilience of lotic algal 
communities: Importance of disturbance timing and current. Ecology 73:1445-1461. 
 
Petranka, J. W. 1984. Ontogeny of the diet and feeding behavior of Eurycea bislineata larvae. 
Journal of Herpetology 18:48-55. 
 
Ploug, H., and H. P. Grossart. 1999. Bacterial production and respiration in suspended 
aggregates: A matter of the incubation method. Aquatic Microbial Ecology 20:21-29. 
 
Ploug, H., and H. P. Grossart. 2000. Bacterial growth and grazing on diatom aggregates: 
Respiratory carbon turnover as a function of aggregate size and sinking velocity. 
Limnology and Oceanography 45:1467-1475. 
 
Ploug, H., S. Hietanen, and J. Kuparinen. 2002. Diffusion and advection within and around 
sinking, porous diatom aggregates. Limnology and Oceanography 47:1129-1136. 
 
Power, M. E., and A. J. Stewart. 1987. Disturbance and recovery of an algal assemblage 
following flooding in an Oklahoma stream. American Midland Naturalist 117:333-
345. 
  
 
 146 
 
Pozo, J., A. Elosegui, and A. Basaguren. 1994. Seston transport variability at different spatial 
and temporal scales in the Aguera watershed (North Spain). Water Research 28:125-
136. 
 
Ranvestel, A. W., K. R. Lips, C. M. Pringle, M. R. Whiles, and R. J. Bixby. 2004. 
Neotropical tadpoles influence stream benthos: Evidence for the ecological 
consequences of decline in amphibian populations. Freshwater Biology 49: 274-285. 
 
Richardson, J. S., and R. J. Mackay. 1991. Lake outlets and the distribution of filter feeders: 
An assessment of hypotheses. Oikos 62:370-380. 
 
Richey, J., J. Melack, A. Aufdenkampe, V. Ballester, and L. Hess. 2002. Outgassing from 
Amazonian rivers and wetlands as a large tropical source of atmospheric CO
2
. Nature 
416:617-620. 
 
Rosi-Marshall, E. J. 2004. Decline in the quality of suspended fine particulate matter as a 
food resource for chironomids downstream of an urban area. Freshwater Biology 
49:515-525. 
 
Schloss, A. L., and J. F. Haney. 2006. Clouds, shadows or twilight? Mayfly nymphs 
recognise the difference. Freshwater Biology 51:1079-1089. 
 
Skalski, J. R., and D. H. McKenzie. 1982. A design for aquatic monitoring programs. Journal 
of Environmental Management 14:237-251. 
 
Smith, E. P. 2002. BACI design, pgs. 141-148.  In A.H. El-Shaarawi and W.W. Piegorsch 
(eds.), Encyclopedia of Environmetrics. John Wiley & Sons, Hoboken, New Jersey. 
 
Smith, E. P., D. R. Orvos, and J. Cairns. 1993. Impact assessment using the before-after-
control-impact (BACI) model - concerns and comments. Canadian Journal of 
Fisheries and Aquatic Sciences 50:627-637. 
 
Smith, R. A. 1980. The theoretical basis for estimating phytoplankton production and specific 
growth rate from chlorophyll, light, and temperature data. Ecological Modelling 
10:243-264. 
 
Spitzy, A. and Ittekot, V. 1991. Dissolved and particulate organic matter in rivers. In: 
Mantoura, R. F. C., Martin, J. M. and Wollast, R. (eds). Ocean margin in global 
change. John Wiley and Sons, Ltd, New York, pgs. 5-17. 
 
Statzner, B., E. Fievet, J. Y. Champagne, R. Morel, and E. Herouin. 2000. Crayfish as 
geomorphic agents and ecosystem engineers: Biological behavior affects sand and 
gravel erosion in experimental streams. Limnology and Oceanography 45:1030-1040. 
 
Statzner, B., O. Peltret, and S. Tomanova. 2003. Crayfish as geomorphic agents and 
ecosystem engineers: Effect of a biomass gradient on baseflow and flood-induced 
transport of gravel and sand in experimental streams. Freshwater Biology 48:147-
163. 
 
 
 147 
 
Stelzer, R. S., J. Heffernan, and G. E. Likens. 2003. The influence of dissolved nutrients and 
particulate organic matter quality on microbial respiration and biomass in a forest 
stream. Freshwater Biology 48:1925-1937. 
 
Suberkropp, K., G. L. Godshalk, and M. J. Klug. 1976. Changes in chemical composition of 
leaves during processing in a woodland stream. Ecology 57:720-727. 
 
Sutherland, T. F., J. Grant, and C. L. Amos. 1998. The effect of carbohydrate production by 
the diatom Nitzschia curvilineata on the erodibility of sediment. Limnology and 
Oceanography 43:65-72. 
 
Swan, C. M., and M. A. Palmer. 2004. Leaf diversity alters litter breakdown in a Piedmont 
stream. Journal of the North American Benthological Society 23:15-28. 
 
Taylor, B. W., A. S. Flecker, and R. O. Hall. 2006. Loss of a harvested fish species disrupts 
carbon flow in a diverse tropical river. Science 313:833-836. 
 
Thomas, S. A., J. D. Newbold, M. T. Monaghan, G. W. Minshall, T. Georgian, and C. E. 
Cushing. 2001. The influence of particle size on seston deposition in streams. 
Limnology and Oceanography 46:1415-1424. 
 
Thorp, J. H., M. C. Thoms, and M. D. Delong. 2006. The riverine ecosystem synthesis: 
Biocomplexity in river networks across space and time. River Research and 
Applications 22:123-147. 
 
Trumbore, S. 2000. Age of soil organic matter and soil respiration: Radiocarbon constraints 
on belowground C dynamics. Ecological Applications 10:399-411. 
 
United States Geological Survey (USGS). 2002. Selection of equipment for water sampling, 
chap. A2 in U.S. Geological Survey, National field manual for the collection of 
water-quality data: p. 3-1 to 3-100. 
 
Usio, N., and C. R. Townsend. 2004. Roles of crayfish: Consequences of predation and 
bioturbation for stream invertebrates. Ecology 85:807-822. 
 
Vadeboncoeur, Y. 1994. Longitudinal dynamics of seston concentration and composition in a 
lake outlet stream. Journal of the North American Benthological Society 13:181-189. 
 
Vallino, J. J., C. S. Hopkinson, and J. E. Hobbie. 1996. Modeling bacterial utilization of 
dissolved organic matter: Optimization replaces Monod growth kinetics. Limnology 
and Oceanography 41:1591-1609. 
 
Van Sickle, J. and R. L. Beschta. 1983. Supply-based models of suspended sediment 
transport in streams.  Water Resources Research 19:768-778. 
 
Vannote, R. L., G. W. Minshall, K. W. Cummins, J. R. Sedell, and C. E. Cushing. 1980. The 
river continuum concept. Canadian Journal of Fisheries and Aquatic Sciences 
37:130-137. 
 
 
 148 
 
Verhoff, F. H., D. A. Melfi, and S. M. Yaksich. 1979. Storm travel distance calculations for 
total phosphorus and suspended materials in rivers. Water Resources Research 
15:1354-1360. 
 
Wagner, R. 1991. The influence of the diel activity pattern of the larvae of Sericostoma 
personatum (Kirby and Spence) (Trichoptera) on organic matter distribution in 
stream-bed sediments - A laboratory study. Hydrobiologia 224:65-70. 
 
Wallace, J. B., T. F. Cuffney, J. R. Webster, G. J. Lugthart, K. Chung, and B. S. Goldowitz. 
1991. Export of fine organic particles from headwater streams: Effects of season, 
extreme discharges, and invertebrate manipulation. Limnology and Oceanography 
36:670-682. 
 
Wallace, J. B., J. R. Webster, and T. F. Cuffney. 1982. Stream detritus dynamics: Regulation 
by invertebrate consumers. Oecologia 53:197-200. 
 
Wallace, J. B., M. R. Whiles, J. R. Webster, T. F. Cuffney, G. J. Lugthart, and K. Chung. 
1993. Dynamics of inorganic particles in headwater streams: Linkages with 
invertebrates. Journal of the North American Benthological Society 12:112-125. 
 
Walters, D. M., K. M. Fritz, and D. L. Phillips. 2007. Reach-scale geomorphology affects 
organic matter and consumer ?
13
C in a forested Piedmont stream. Freshwater Biology 
52:1105-1119. 
 
Webster, J. R., E. F. Benfield, T. P. Ehrman, M. A. Schaeffer, J. L. Tank, J. J. Hutchens, and 
D. J. D'Angelo. 1999. What happens to allochthonous material that falls into streams? 
A synthesis of new and published information from Coweeta. Freshwater Biology 
41:687-705. 
 
Webster, J. R., S. W. Golladay, E. F. Benfield, D. J. Dangelo, and G. T. Peters. 1990. Effects 
of forest disturbance on particulate organic matter budgets of small streams. Journal 
of the North American Benthological Society 9:120-140. 
 
Whiles, M. R., and W. K. Dodds. 2002. Relationships between stream size, suspended 
particles, and filter-feeding macroinvertebrates in a great plains drainage network. 
Journal of Environmental Quality 31:1589-1600. 
 
Wiegner, T. N., L. A. Kaplan, J. D. Newbold, and P. H. Ostrom. 2005. Contribution of 
dissolved organic C to stream metabolism: a mesocosm study using 
13
C-enriched 
tree-tissue leachate. Journal of the North American Benthological Society 24:48-67. 
 
Winter, T. C. 2001. The concept of hydrologic landscapes. Journal of the American Water 
Resources Association 37:335-349. 
 
Wipfli, M. S., and D. P. Gregovich. 2002. Export of invertebrates and detritus from fishless 
headwater streams in southeastern Alaska: Implications for downstream salmonid 
production. Freshwater Biology 47:957-969. 
 
Wotton, R. S., C. P. Joicey, and B. Malmqvist. 1996. Spiralling of particles by suspension 
feeders in a small lake-outlet stream. Canadian Journal of Zoology 74:758-761. 
 
 
 149 
 
Zanetell, B. A., and B. L. Peckarsky. 1996. Stoneflies as ecological engineers - Hungry 
predators reduce fine sediments in stream beds. Freshwater Biology 36:569-577.