\ I
l

135
292

 

THFSIS

,9

Date

0-7 639

This is to certify that the
thesis entitled

COMPLEX INTERACTIONS BETWEEN ZEBRA MUSSELS
AND THEIR PLANKTONIC PREY

presented by
Alan Elliott Wilson

has been accepted towards fulfillment
of the requirements for

M.S. degree in Fisheries and Wildlife

ﬂ/M/wa

Major professor

5 May 2001

MS U is an Afﬁrmative Action/Equal Opportunity Institution

 

.__. _.__ ’ _

 

 

LIBRARY

Michigan State
University

 

 

 

PLACE IN RETURN Box to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

 

 

 

DATE DUE [ DATE DUE DATE DUE
l . {OCT 2 6 2004
* z ,‘ a 2
f4? “ :1"
£29404 ’

 

 

 

 

 

 

 

 

 

 

 

 

6/01 cJCIHC/DatoDue.p65-p.15

COMPLEX INTERACTIONS BETWEEN ZEBRA MUSSELS
AND THEIR PLANKTONIC PREY
By

Alan Elliott Wilson

A THESIS
Submitted to
Michigan State University
In partial fulﬁllment of the requirements
for the degree of
MASTER OF SCIENCE
Department of Fisheries and Wildlife

2001

ABSTRACT

COMPLEX INTERACTIONS BETWEEN ZEBRA MUSSELS
AND THEIR PLANKTONIC PREY

By

Alan Elliott Wilson

Many studies in North American lakes have documented decreases in
phytoplankton and ciliate abundance after the invasion of the zebra mussel (Dreissena
polymorpha). However, fewer studies have examined the effect of zebra mussels on
phytoplankton species composition or investigated the effects at a scale appropriate to
phytoplankton. In Chapter 1, I derived an equation that predicts zebra mussel dry tissue
biomass from total phosphorus concentration to estimate a reasonable stocking density of
mussels for my experiment. Chapter 2 describes a ﬁve-week replicated in situ mesocosm
experiment I conducted to evaluate the impact of zebra mussels on phytoplankton and
ciliate communities in a pond that lacked mussels. Within one week, zebra mussels non-
selectively reduced phytoplankton biomass by 53%. The effect of zebra mussels on total
phytoplankton biomass gradually declined over the remaining four weeks of the
experiment. By the end of the experiment, no algal groups were statistically different
between treatments. This waning effect of zebra mussels on algal abundance could not
be explained by a shift towards less edible algal species and may have been due to the
deteriorating condition of the zebra mussels. The algal data suggested that the zebra
mussels suffered from food limitation after the ﬁrst week. In contrast, the mussels

reduced ciliate abundance by 70% or greater throughout the entire study.

ACKNOWLEDGEMENTS

I sincerely thank O. Samelle, 8. Hamilton, G. Mittelbach, and P. Soranno for their
helpful comments and criticisms throughout the entire project; E. Cebelak, L. Dillon, G.
Gerrish, P. Hamilton, and N. Samelle for ﬁeld assistance; J. Chiotti and T. Toda for zebra
mussel total wet to dry tissue conversion; D. Donham and K. Hoffman for technical
assistance; S. Hamilton and D. Raikow for Gull Lake TP data; N. Dorn and R. Valley for
their suggestions on the experimental design; E. Cebelak and D. Raikow for their help
building the enclosures; M. Machavaram for sharing his knowledge about chemistry; W.
Wilson for helping process samples; K. Rogers for helping identify zooplankton; J. Bence
and M. Rutter for their statistical expertise, N. Consolatti and the rest of the KBS staff for
their kindness and support; T. Coon and the rest of the F&W staff for their enthusiastic
support; R. Valley for his thorough criticisms of earlier drafts of this manuscript; C.
Steiner for use of his stapler; and my family and friends for their endless encouragement
for all of my endeavors. This project was funded by a George H. Lauff Scholarship, the
Department of Fisheries and Wildlife at Michigan State University, National Science
Foundation Research Training Grant DIR-9602252, and by the Michigan Sea Grant
College Program, project number R/NIS-4, under grant number NA76RG0133 from the

Ofﬁce of Sea Grant, NCAA, and funds from the State of Michigan.

iii

TABLE OF CONTENTS

LIST OF TABLES .................................................................................... v
LIST OF FIGURES ................................................................................. vi
INTRODUCTION .................................................................................... 1
CHAPTER 1
RELATIONSHIP BETWEEN ZEBRA MU SSEL BIOMASS AND TOTAL
PHOSPHORUS IN EUROPEAN AND NORTH AMERICAN LAKES ..................... 2
Introduction ................................................................................... 2
Materials and Methods ...................................................................... 3
Literature Data ...................................................................... 3
Gull Lake Sampling ................................................................ 5
Data Analysis ........................................................................ 6
Results ......................................................................................... 7
Discussion .................................................................................... 8
CHAPTER 2
EFFECTS OF ZEBRA MUSSELS ON PHYTOPLANKTON AND CILIATES: A
MESOCOSM EXPERIMENT .................................................................... 14
Introduction ................................................................................. 14
Materials and Methods ..................................................................... 15
Experimental Setup ............................................................... 15
Sampling and Laboratory Analyses ............................................. 19
Data Analyses ..................................................................... 21
Results ....................................................................................... 22
Physical and Chemical Parameters ............................................. 22
Biological Parameters ............................................................ 23
Pond Conditions ................................................................... 26
Discussion ................................................................................... 26
APPENDICES
Tables ........................................................................................ 32
Figures ....................................................................................... 40
BIBLIOGRAPHY ................................................................................... 49

Table 1:

Table 2:

Table 3:

Table 4:

Table 5:

Table 6:

LIST OF TABLES

Lake characteristics and dreissenid abundance for Polish lakes. Mussel
biomass expressed as dry tissue mass. Year of sampling (when known) in
parentheses.

Total phosphorus, mean depth and dreissenid abundance for North
American lakes. Mussel biomass expressed as dry tissue mass.

Multiple regression statistics for the inﬂuence of mean depth (m), calcium
(mg 0 L") and log total phosphorus (mg 0 m3) on zebra mussel biomass in
Polish lakes. N = 16.

Treatment and pond averages (Days 8-36), ranges (Days 1- 36), p-values,
and treatment effects (100 * [(zebra mussel — control)/control]) for
chemical, physical, and biological parameters. On July 19, pH was not
recorded due to a malfunction with the electrode. NA = not available

Absolute observed treatment differences (zm+ vs. control) in algal groups
throughout the study. All groups are measured as total algal dry biomass
(pg 0 LI) and “Others” denotes desmids, diatoms, and ﬁlamentous

algae.

Relative observed treatment differences (zm+ vs. control) in al gal groups
throughout the study. All groups are measured as relative dry biomass (%
of total biomass) and “Others” denotes desmids, diatoms, and
ﬁlamentous algae.

Figure 1:

Figure 2:

Figure 3.

Figure 4.

Figure 5.

Figure 6.

Figure 7.

Figure 8.

Figure 9.

LIST OF FIGURES

Relationship between total phosphorus and dreissenid biomass for Polish
and North American lakes. Data from Tables 1 & 2. Regression line and
95% conﬁdence bands for equation estimates were calculated for Polish
lakes after excluding the outlier (Lake Olow).

Chlorophyll concentration in the control and zebra mussel enclosures
(daily treatment averages :t 1 standard error).

Measured effect size (100 * [(zebra mussel — control)/ control]) of zebra
mussels on phytoplankton abundance (measured as chlorophyll a) over the
37 days of the experiment.

Observed chlorophyll a concentrations for Day 8 plotted in relation to the
Dillon-Rigler TP-chlorophyll relationship (logChl a = 1.449 * logTP -
1.136). 95% prediction intervals for Dillon-Rigler are provided. Error
bars represent i 1 standard error.

Relative algal group dry biomass (%) for zebra mussel enclosures, control
enclosures, and the pond for Day 8, Day 22, and Day 36.

Mean ciliate dry biomass (pg 0 L", i 1 standard error) for zebra mussel
and control enclosures and pond for Days 8, 22, and 36.

Dry tissue biomass (mg 0 mm!) for average mussel length (12mm) for
pre- and post-experimental zebra mussels. A correction factor for the
effect of freezing as a preservative technique was used to determine the
post-experimental mussel weights. Error bars represent 1 standard error.

Time-averaged (Days 8, 22, and 36) data on small green alga] group
dry biomass (pg 0 L") vs. ciliate dry biomass (pg 0 LI) for the four
replicates of both treatments.

Average available carbon (rim3 0 day) in the zebra mussel enclosures.
Horizontal dotted line indicates minimum amount of carbon necessary for

routine maintenance without growth based on mussels fed Nitzschia (from
Walz 1978).

vi

Introduction

The following two chapters represent part of the work I completed as part of my
Master’s degree requirements in Fisheries and Wildlife. I wanted to perform an
ecologically sound mesocosm experiment aimed at examining the effect of zebra mussels
on phytoplankton and ciliates. To do this correctly, I needed to determine a suitable
amount of zebra mussels to stock into my enclosures. Since the pond where the
experiment was conducted was mussel-free, I was unable to determine a zebra mussel
biomass estimate for the pond. Chapter 1 describes an analysis of published data on
zebra mussel biomass and limnological variables. From this data, a predictive equation
was developed for calculating zebra mussel biomass from total phosphorus concentration.
At the time when this thesis was submitted, a manuscript based on Chapter 1 was
submitted to Archiv fiir Hydrobiologie. I used this equation from Chapter 1 to stock my
experimental enclosures with a realistic, naturally occurring amount of zebra mussels. In
Chapter 2, I describe an experiment I conducted to determine how zebra mussels affect
phytoplankton and ciliates in a newly invaded system. A manuscript based on Chapter 2
is currently being prepared for submission to Canadian Journal of Fisheries and Aquatic

Sciences.

Chapter 1: Relationship between zebra mussel biomass and total phosphorus in

European and North American lakes

Introduction

Since its invasion from Europe in the mid-1980’s, the zebra mussel (Dreissena
polymorpha) has spread rapidly into freshwater systems throughout eastern North
America (Hebert et al. 1991; Ludyanskiy et al. 1993), with often dramatic effects on
community structure and ecosystem processes (Gillis and Mackie 1994; Johengen et al.
1995; Lavrentyev et al. 1995; MacIsaac et al. 1995; Nalepa et al. 1996; Bastviken et al.
1998; Pace et al. 1998). Forecasting the effects of zebra mussels on ecosystems yet to be
invaded is currently limited by, among other things, the ability to predict the eventual
abundance of Dreissena. Ramcharan et al. (1992a, b) were successful in building
empirical models of steady-state (i. e., long-term average) abundance and population
fluctuations based on lake characteristics, but these models only predict the density of

mussels (number 0 m'z). Many ecosystem impacts of invading species should be more
closely linked to population biomass (g - m'z) than to density (Mellina et al. 1995; Amott

and Vanni 1996; Young et al. 1996), so it would also be useful to develop empirical
models that can predict zebra mussel biomass from easily-measured lake characteristics.
For example, the relationship between zebra mussel body mass and the rate at which
particles are ﬁltered is only weakly nonlinear (logzlog slope ~O.9, Kryger and Riisgard
1988), so biomass predictions can be used to roughly predict potential ﬁltration rates by

future mussel populations. A predictive model for zebra mussel biomass would also be

helpful in the selection of mussel stocking levels for manipulative experiments, especially
in habitats where no estimates of natural abundance are available.

A number of lake characteristics may potentially inﬂuence the biomass of zebra
mussels in freshwater lakes, including: lake depth, bottom slope, substrate type, degree of
mixing, turbidity, nutrient concentrations, and phytoplankton biomass (Hanson and Peters
1984; Rasmussen and Kalff 1987; Ramcharan et al. 1992b; Mellina and Rasmussen
1994). We expected mussel biomass to be positively related to TP, as seen for
zoobenthic biomass in general (Hanson and Peters 1984; Rasmussen and Kalff 1987),
because of the strong inﬂuence of phosphorus in limiting lake, and in particular,
phytoplankton productivity (Schindler 1977; 1978). However, Stanczykowska (1984)
reported a tendency for mussel density to be reduced in lakes with very high TP (> 300

mg 0 m3) and Ramcharan et al. (1992b) found a negative relationship between mussel

density and orthophosphate concentration.

Materials and Methods

Literature data
We found published data on zebra mussel biomass (reported here as dry tissue
biomass) and one or more predictor variables, including; total phosphorus concentration
(TP; summer and spring), calcium concentration (Can), lake area and depth (mean and
maximum), Secchi depth, and chlorophyll concentration (summer and spring). We were
only able to find sufﬁcient literature data for three of these potential predictors; depth,
Ca”, and TP. Of these three predictors, we did not expect there to be a strong inﬂuence

of depth or calcium in the data set. Shallow lakes might be expected to have higher areal

biomass than deep lakes, since a greater fraction of phytoplankton production should be
available to benthic ﬁlter feeders, and there should be less oxygen depletion near the
bottom in shallow well-mixed systems. However, most literature data on zebra mussel
abundance refer to biomass in the depth zone of mussel occurrence, so deep areas with no
mussels would presumably not affect the average biomass reported for deeper lakes. This
makes it less likely that lake depth, independent of lake productivity, will inﬂuence zebra
mussel biomass as typically reported in the literature. In addition, we limited our
analyses to lakes that contain zebra mussels, so presumably all lakes in the data set
should have sufﬁcient calcium for mussel growth. Consequently, we did not expect Ca+2
to be an important factor inﬂuencing zebra mussel biomass (Sprung 1987; Ramcharan et
al. 1992b). The data set comprised 32 lakes in Poland and six lakes in North America

(T ables 1, 2). Note that the set of studies that report mussel biomass is only a small
fraction of all studies that have estimated zebra mussel density.

Several of the Polish lakes were part of three lake systems: Beldany-Mikolajskie-
Sniardwy; Bozcne-Niegocin; and Dargin-Dobskie-Kisajno-Mamry. We followed
Ramcharan et al. (1992a) and considered each of these lakes as an independent
observation, since there was considerable variation in mussel biomass and both stable and
unstable mussel populations (Ramcharan et al. 1992a) among lakes within the same
system. Treating these lakes as independent observations did not inﬂuence our
conclusions with respect to the statistical signiﬁcance (at P < 0.05) of the two
relationships between mussel biomass and predictor variables that we report.

We excluded data from studies in which we judged that the lake bottom was

sampled in a biased manner; as in studies that collected samples only from hard

substrates or reefs (e. g., Hamilton et al. 1994; Kornobis 1977). We limited our data set
to include only those studies that presented zebra mussel biomass as dry tissue weight,
and in most cases, we relied on mean biomass values reported by the authors (all Polish
lakes, Saginaw Bay, Lake St. Clair, Lake Erie). For the remaining lakes we made our
own determinations of lake-wide mean biomass, as described below.

For Oneida Lake, we extracted size distribution data from Figure 7 of Mellina et
al. (1995) with a digitizer, and applied their tissue mass to shell length relationship (dry
mass = 0.006221ength2'61) to determine biomass. For Lake Ontario, we relied on depth-
speciﬁc biomass estimates (reported as kg 0 10 min trawl'l) from Figure 2 of Mills et al.
(1999) and their estimate of 0.73 ha swept per 10 min trawl. Mills et al. (1999) reported
dreissenid biomass for 8 depth strata ranging from 15 m to 85 m. We converted their
depth-speciﬁc biomass estimates to an area-weighted lake average by determining the
proportion of lake bottom within each of the sampled depth strata from a digitized

bathymetric map (NOAA, National Geophysical Data Center).

Gull Lake sampling
We estimated zebra mussel biomass in Gull Lake (42°24’N, 85°24’W) on 8 - 9
July, 1999, 5 years after Dreissena was ﬁrst sighted in this lake (see Moss 1972 for a
description of the lake). Four sampling sites were selected randomly from each of four
depths: 2.5, 5, 7.5, and 10 m. SCUBA divers collected all mussels and macrophytes by
hand within a 1 m2 quadrat from each site. Samples were frozen for a few days before
sorting. We counted and measured all mussels larger than 15 mm, and subsampled

smaller mussels. Some sites had large numbers of very small mussels attached to

macrophytes. In these cases, macrophytes were subsampled by measuring the total wet
weight of the macrophytes, then weighing out subsamples from which mussels were
counted and measured. We developed a dry tissue mass (g) versus shell length (mm)
relationship for fresh Gull Lake mussels: (log dry tissue mass = 2.5429 * log length -
4.9396, R2 = 0.93, N = 50) to convert size distributions to biomass.

The epilimnion of Gull Lake was sampled four times from June to August, 1998,
with a depth integrating tube sampler. Water samples were filtered through Whatman
GF/F glass ﬁber ﬁlters on the day of collection. Total phosphorus (the sum of dissolved
and particulate fractions) was measured via persulfate digestion (Valderrama 1981)

followed by molybdate blue colorimetry (Murphy and Riley 1962).

Data analysis

Much of the mussel biomass data represented single-year estimates, and most of
the measurements of Ca+2 and TP (summer averages for the epilimnion) in the Polish data
set were made many years after zebra mussel biomass was estimated (Table 1). Although
Ca+2 should not change drastically from year to year, large temporal changes in TP are
possible due to human inﬂuence. These factors should increase the unexplained error of
a predictive relationship, especially given that zebra mussel abundance can vary greatly
from year to year (Ramcharan et al. 1992a; Stanczykowska 1984; Stanczykowska and
Lewandowski 1993). Restricting the data set to lakes in which mussel biomass and TP
were measured within three years of each other did not improve ﬁt (N = 6). We log-

transforrned the TP data which greatly reduced skewness in this variable.

Initial statistical analyses indicated the presence of two very large outliers (Lake
Olow and Lake Stregiel, Table l) in the Polish data set. For example, when we regressed
mussel biomass against TP, Ca+2 and mean depth, the standardized residual for Lake
Olow was 6.1, indicating an extreme outlier. There were no TP data for Lake Stregiel, so
inclusion/exclusion of this lake was inconsequential to analyses involving TP. These two
lakes had unusually high biomass (> 40 g o m'z) and were responsible for strong
skewness in the mussel data that could not be alleviated via transformation.
Consequently, we excluded these two lakes from all subsequent statistical analyses. We

strongly suspect that values of mussel biomass greater than 40 g 0 m"; represent transient,

nonsustainable excursions from long-term average biomass (see Discussion). For this
reason, we also restricted the data from Saginaw Bay, Lake Huron to data collected in

1993. Mussel biomass in Saginaw Bay increased from 10 g o m‘2 in 1991 (ﬁrst year after
invasion) to 62 g 0 In2 in 1992, then declined to 4.5 g o m'2 in 1993 (Nalepa et al. 1995),

suggesting that zebra mussel biomass was far above long-term sustainable levels in 1992.

Results

The Polish data consisted of mesotrophic to hyper-eutrophic lakes (mean and
range of TP: 70, 19 - 233 mg 0 m3) with mean depths from 1 - 14 m, and Ca+2
concentrations from 32 - 75 mg 0 L’1 (Table 1). Not surprisingly, mean depth and
maximum depth were highly intercorrelated (r = 0.81, P < 0.0001, N = 25), so we only
used mean depth in regression analyses. There was only a marginally signiﬁcant

correlation between Ca+2 and logTP (r = 0.43, P = 0.09, N = 16) and no signiﬁcant

correlation between mean depth and logTP (r = -0.32, P = 0.13, N = 24) in the Polish
data. Stepwise multiple regression indicated that Ca+2 and mean depth had no
statistically signiﬁcant inﬂuence on mussel biomass in these data (P > 0.3). The latter
result was robust to the order in which variables were entered and whether variables were
forced into the model. Multiple regression indicated that logTP was the only potentially
inﬂuential variable (Table 3). Based on these results, we calculated a predictive equation
for the Polish data (Figure 1): mussel biomass = -10.8 (18.8) + 11.0 (14.9) logroTP, R2 =
0.19, P = 0.04, N = 24 (standard errors in parentheses).

The TP range of the six North American lakes (Table 2, mean and range of TP:
15.9, 9.3 - 22.5 mg 0 m3) was much smaller and extended lower than the range of TP in
the Polish data. With the exception of Oneida Lake, biomass estimates in these recently-
invaded North American lakes ﬁt within the 95% conﬁdence limits of predictions from
the equation above (Figure 1). The data for Oneida Lake were collected from within two
years of initial invasion, and so probably represent a transient, nonsustainable biomass, as
seen in Saginaw Bay (see Discussion).

Combining the Polish and North American data and excluding Lake Olow and
Oneida Lake (not near steady-state), we calculated the following predictive equation:
mussel biomass = -6.5 (15.4) + 8.7 (13.2) lOgroTP, R2 = 0.22, P = 0.01, N = 29 (standard

errors in parentheses).

Discussion

Total phosphorus was the only variable that signiﬁcantly predicted zebra mussel

biomass in the Polish data set. Lack of Ca+2 inﬂuence was expected, since Ca+2

concentrations were above 30 mg 0 L'1 in every lake (Table 1), a level that is above
minimum thresholds for successful Dreissena growth and reproduction (Ramcharan et al.
1992b). Likewise, mean and maximum depth should have little direct inﬂuence given
that mussel biomass is generally reported from Polish lakes for the restricted depth zone
of mussel occurrence.

The amount of variation in mussel biomass explained by TP (R2 = 0.19, 0.22) was
low but comparable to that reported for total zoobenthic biomass by Rasmussen and Kalff
(1987) (R2 = 0.20, 0.26). In contrast, Hanson and Peters (1984) found a stronger
relationship between zoobenthic biomass and TP (1984; R2 = 0.48), which may be related

to the lower TP range in their study (3 - 117 mg 0 m3) relative to the range in Rasmussen
and Kalff (4 - 390 mg 0 m3) and our study (9 - 233 mg 0 m3). The response of lake

productivity (as indexed by phytoplankton biomass) to increases in TP tends to be weaker

for lakes with TP greater than ~200 mg - rn'3 (Sarnelle et al. 1998). At high levels of TP

other factors begin to limit lake productivity (Smith 1982, McCauley et al. 1989), so the
inﬂuence of increased phosphorus on benthic biomass should weaken.

Given that the response of a single species to enrichment is likely to be much
more variable than the response of total zoobenthic biomass, the statistical signiﬁcance of
the TPzzebra mussel relationship is encouraging. However, we were only able to

establish a TP inﬂuence after excluding lakes with mussel biomass > 40 g o m’z. The

status of these excluded lakes is thus critical to our analysis.

Based on both Polish and North American data, we propose that a dreissenid

2

biomass in excess of ~40 g 0 m' is not sustainable (i. e., much higher than steady-state

biomass) in lakes. This hypothesis is supported by evidence suggesting that mussel

populations in outlier lakes were unstable at the time that biomass was measured. One of
the three high-biomass lakes (Lake Stregiel, Table 1) was classiﬁed as having an unstable
mussel population by Ramcharan et al. (1992a), based on the magnitude of interannual
density ﬂuctuations. Lake Olow, a second high-biomass lake, was not explicitly
classiﬁed as unstable by Ramcharan et al. (1992a), but we note that mussel density was

reported as 1830 0 In2 in 1978 (the year in which biomass was estimated, Lewandowski
1991) and as 514 o m'2 (year unspeciﬁed) by Stanczykowska and Lewandowski (1993).

Thus, the mussel population in Lake Olow may have been at a transient high biomass
level in 1978. Biomass data for Oneida Lake, the third high-biomass lake, were limited
to the ﬁrst two years after invasion (Mellina et al. 1995), and thus may represent an initial
overshoot of steady-state biomass, as documented in Saginaw Bay (N alepa et al. 1995).
Based on our empirical relationship, we would expect that zebra mussel biomass in
Oneida Lake should be about ﬁve times lower than that observed in 1992 and 1993. This
prediction can be readily tested as a way of evaluating our contention that a zebra mussel

biomass in excess of ~40 g 0 m'2 is far above steady-state levels.

To more rigorously examine population instability as a factor producing residual
variation in the Tszussel biomass relationship, we restricted the Polish data to lakes
classiﬁed as stable by Ramcharan et al. (1992a), and recalculated the logTsziomass
regression. This restriction greatly improved the fit: mussel biomass = -33.5 (116.3) +
23.2 (18.4) longP, R2 = 0.60, P = 0.04, N = 7 (standard errors in parentheses). We
caution against using this equation for prediction because it is based on very few lakes,
but the improvement in ﬁt suggests that population instability may be a major source of

residual variation in the TP-mussel biomass relationship. Additional factors that may

10

account for residual variation are: substrate quality (Mellina and Rasmussen 1994), the
temporal mismatch between measurements of TP and mussel biomass for most of the
Polish lakes, and within-lake spatial variation in mussel biomass estimates. Standard
deviations of single-year biomass estimates for North American lakes vary from 65%
(Gull Lake) to >>100% (Lake Erie, Lake St. Clair, Saginaw Bay) of the mean (Nalepa et
a1. 1995; N alepa et a1. 1996; Derrnott and Kerec 1997), so this source of residual
variation may be important.

With the exception of the first two years after invasion, we found no evidence to
suggest that time since colonization inﬂuences steady-state dreissenid biomass. North
American lakes three years after invasion do not seem to support detectably higher
biomass of zebra mussels than Polish lakes in which mussels have existed for >50 years.
The limited data available on population dynamics immediately after invasion (Saginaw
Bay, Nalepa et a]. 1995) support the suggestion that dreissenid populations in North
America require only about three years to approach steady-state biomass. Ironically,
mussel populations in some recently invaded North American lakes may be closer to
steady-state biomass than populations in some EurOpean lakes. Long-term presence is no
guarantee that a mussel population will be near steady-state biomass in any given year,
especially when one considers the precipitous declines and rapid recoveries that
characterize some long-established populations (Stanczykowska et al. 1975). Clearly,
more study of Dreissena population dynamics is needed.

To compare the response of dreissenid biomass to TP enrichment with the
response of total zoobenthic biomass reported in previous studies requires conversion of

the data to common scales. To this end, we calculated logloTP vs. loglobiomass and lnTP

ll

vs. (biomass)°'1 regressions for the combined Polish and North American data to enable
comparison with the relationships reported by Hanson and Peters (1984) for profundal
benthos, and Rasmussen and Kalff (1987) for profundal and sublittoral benthos,
respectively. For both of these comparisons, the slope of the zebra mussel response to TP
enrichment (logzlog slope:0.55, ln:tenth root slope: 0.06) was roughly similar to the slope
for total zoobenthic biomass (logzlog slope: 0.71, ln:tenth root slope: 0.08-0.09). These
comparisons suggest that dreissenid biomass increases at a roughly similar rate with
enrichment as total zoobenthic biomass. The elevation of the dreissenid regression was
similar to those reported by Rasmussen and Kalff (1987), but somewhat lower than that
reported by Hanson and Peters (1984) (-0.22 versus -0.09). A lower elevation is expected
for the response of a single taxon relative to total zoobenthic biomass.

The positive relationship that we found between dreissenid biomass and TP is not
surprising given that zoobenthic biomass responds positively to nutrient enrichment, but
contrasts with the negative correlation found between mussel density and orthophosphate
concentration by Ramcharan et al. (1992b). It is difficult to compare these contrasting
results because of the difference in independent variables employed, but we can suggest
that the data set analyzed by Ramcharan et al. (1992b) included lakes with much higher
levels of phosphorus loading than the lakes that we analyzed. Their data set included 15

lakes with orthophosphate concentrations in excess of 100 mg - m'3. Many of these lakes
probably had TP in excess of 233 mg 0 m3, the maximum in our data set. In extremely

eutrophic waters, anoxia and toxic cyanobacteria may lead to reduced zebra mussel
biomass, as suggested by Stanczykowska (1984). In any case, orthophosphate

concentration is a poorer surrogate variable for lake productivity than TP, because the

12

former is subject to much greater seasonal variation and is under much greater control by
the biota (via uptake and excretion) than is TP.
In conclusion, we found that dreissenid biomass can be predicted from TP for

lakes with TP less than 233 mg 0 m3, and that steady-state biomass in recently invaded

North American lakes generally ﬁts the positive Tsziomass relationship for Polish lakes.
The latter ﬁt, however, is at least in part a function of the low R2 of the Polish
relationship. The Tsziomass relationship for Polish and North American lakes
combined can be used to predict future zebra mussel biomass in uninfested lakes, and to
suggest reasonable biomass stocking levels for experiments in habitats for which biomass
estimates are lacking. These predictions, however, carry a large degree of uncertainty,
much of which may stem from large interannual and spatial variation in mussel
abundance for individual lakes. To reﬁne these predictions, more unbiased estimates of
dreissenid biomass over multiple years are needed. We strongly recommend that dry
tissue mass (rather than total or shell-free wet mass) be determined in future studies given
that most existing studies report dry tissue mass and that dry mass estimates are more

generally reliable and reproducible.

13

Chapter 2: Effects of zebra mussels on phytoplankton and ciliates: a mesocosm

experiment

Introduction

Since their introduction into North America in the mid-1980’s, zebra mussels
have been implicated in drastic ecosystem-level changes, including a shift in energy ﬂow
from the pelagic to the benthic zone (Fahnenstiel et al. 1995; Johengen et al. 1995; Amott
and Vanni 1996; Lavrentyev et a]. 2000), declines in microzooplankton and
phytoplankton biomass (Holland 1993; Heath et al. 1995; Lavrentvey et al. 1995;
MacIsaac et al. 1995; Bastviken et a1. 1998; Pace et al. 1998; Jack and Thorp 2000;
James et a]. 2000; Yu and Culver 2000), and shifts in phytoplankton species composition
from communities dominated by palatable species, like diatoms, to inedible blue-greens
(Lowe and Pillsbury 1995) and vice versa (Reeders and bij de Vaate 1990; Caraco et al.
1997; Smith et al. 1998; Yu and Culver 2000).

Several manipulative studies have been performed to assess the effects of zebra
mussels on phytoplankton and ciliates (Reeders and bij de Vaate 1990; Heath et al. 1995;
Lavrentyev et al. 1995; Mellina et al. 1995; Klerks et al. 1996; Roditi et al. 1996; James
et al. 1997; Bastviken et al. 1998; Jack and Thorp 2000; James et al. 2000). Of these,
laboratory experiments have documented zebra mussel feeding preferences for particular
groups of protozoans and algae (Lavrentyev et al. 1995; Bastviken et al. 1998), as well as
decoupling of a well-established total phosphorus-chlorophyll relationship (Mellina et al.
1995). Although laboratory studies can be useful for examining small-scale effects on

individual species and populations, they are inappropriate for studying community

14

processes such as shifts in species composition. For questions of this type, larger-scaled
experiments are needed. Of those experiments aimed at examining the effects of zebra
mussels on phytoplankton communities in the ﬁeld, only three were large—scale ﬁeld
experiments (Reeders and bij de Vaate 1990; Heath et al. 1995; Jack and Thorp 2000).
The ﬁrst of these was unreplicated and the latter two studies lasted less than a week.
Because phytoplankton and ciliates can take several weeks to establish a new community
equilibrium after a disturbance, studies aimed at examining the effect of a newly
introduced grazer on these guilds should allow enough time for populations to reach a
new equilibrium. In this paper, I examine the effect of zebra mussels on the
phytoplankton and ciliate community with a five-week replicated ﬁeld experiment.
Speciﬁcally, I address the following questions:
I. Do zebra mussels negatively affect phytoplankton, and if so, over what time
scale?
2. Do zebra mussels cause a shift in phytoplankton species composition from
palatable to unpalatable species?
3. How does the effect of zebra mussels on ciliates compare to their effects on

phytoplankton?

Materials and Methods

Experimental Setup

This experiment was performed in an experimental pond at the Kellogg

Biological Station (Michigan State University, Hickory Comers, Michigan). The pond

15

was 30 m in diameter and 1.8 m in depth and almost completely surrounded by cattails.
The pond contained abundant macrophytes, sunﬁsh (Lepomis spp.), and
macroinvertebrates, but no zebra mussels.

Two treatments with four replicates each were used in the experiment: zebra
mussels present and zebra mussels absent (control). Eight experimental enclosures were
constructed out of clear polyethylene (1.13 m in diameter by 1.5 min depth and heat
sealed at the bottom) that was stapled to 1.0 m x 1.0 m Styrofoam-supported wooden
frames. Enclosures were covered by plastic window screen to prevent pond organisms
from entering the enclosures.

The enclosures were ﬁlled on 15 June 1999 with pond water pumped through a

149 um mesh net to remove all macrozooplankton. Zooplankton were then collected
with a 102 um mesh net from the pond and stocked into each enclosure on two occasions

(22 June and 9 July) to achieve a natural zooplankton density. The second stocking was
performed because zooplankton densities were much lower in both sets of enclosures
relative to the pond during the ﬁrst two weeks. Despite the additional stocking,
zooplankton densities remained low in the enclosures relative to the pond throughout the
entire experiment.

In order to prevent large zooplankton from dominating in the enclosures, one
juvenile bluegill (Lepomis macrochirus; 402tl mm total length) was added to each
enclosure. Stocking density was below the natural density of bluegill for local lakes
(Mittelbach 1988) because the pond was less productive than local lakes. The ﬁsh were
seined from a local lake on 27 May and held in aquaria until they were transferred to the

enclosures on 18 June. The enclosures were checked daily for fish mortality. Dead ﬁsh

16

were removed and immediately replaced with a similar-sized fish. A total of ﬁve fish
were replaced during the course of the experiment, but ﬁsh mortality did not have a
signiﬁcant impact on any measured responses (ANOVA; p > 0.35)

I used an equation that predicts zebra mussel dry tissue biomass (g o m'z) from
total phosphorus concentration (pg 0 L") to estimate a reasonable stocking density (2 1 g
0 m'z) for the experiment (dry tissue biomass = -10.8 + 11.0 * logTP, R2 = 0.19, P < 0.04,
N = 24; Wilson and Samelle submitted). On 20 May, zebra mussels were collected from
Gull Lake and quickly transported directly to the lab where druses were separated with a
razor and all detritus was gently scraped from each mussel with a coarse scouring pad.
Next, all mussels between 10 and 20 mm were placed into a ﬂow-through tank (2.5 m x
0.3 m x 0.3 m) where they were allowed to attach to one of eight substrates made from
PVC pipe (0.1 m in diameter, 0.5 min length, and 0.05 m thick) cut lengthwise to create
symmetrical pipe halves. Fresh water was pumped from Gull Lake (19 L 0 min") and
ﬁltered for large debris with a l-mm-mesh net before entering the tank. Settled detritus
and feces were siphoned from the tank twice daily.

To determine the relative condition of the mussels before and after the
experiment, length-weight relationships were derived from randomly chosen mussels.
On the same day that the mussels were collected and added to the ﬂow-through tank, 50
fresh mussels (mean length = 13.9 mm, length ranging from 6.3 — 24.4 mm) were
measured with calipers to the nearest 0.1 mm, and the soft body tissue was removed with

a scalpel. The tissue was then placed into a drying oven at 55 °C for 22 hours until a
constant weight was observed. Dried mussel tissues were weighed to the nearest i 0.01

mg. Post-experimental mussels were removed from the experimental substrates on 7

l7

August and frozen for later analysis. Twenty-ﬁve post-experimental mussels from each
enclosure (for all post-experimental mussels: mean length =13.7 mm, range 9.0 - 21.0
mm) were later thawed and a length-weight relationship for post-experimental mussels
was derived following the same protocol used for pre-experimental mussels. Because dry
weight can be lost through freezing (J. Chiotti, A. E. Wilson, and T. Toda, unpublished
data), a correction factor was used to adjust the dry weight of frozen samples. To
determine the correction factor, additional mussels were collected from Gull Lake and the
length-weight protocol was repeated using a set of fresh mussels versus a matched set
that had been frozen. The correction factor derived from a linear regression of frozen dry
tissue mass on fresh dry tissue mass was: Frozen (g) = 1.3406 * Fresh (g) - 0.0003 (R2 =
0.997, P < 0.001).

On Day 1 (21 June), one substrate was hung into each of the eight enclosures
from a 0.5 m nylon rope attached to a PVC pipe (30 mm in diameter and 1.2 min length)
secured across the tops of each enclosure. This allowed each substrate to be hung
directly in the middle of each enclosure. All zebra mussels were removed from four of
the eight mussel substrates and these substrates were placed into the control enclosures.
All mussels on the remaining four treatment substrates were counted and measured
before being deployed. Mussels used in the experiment averaged 12.2i2.7 mm in length
(mean i standard error), and the average initial dry tissue biomass used in each enclosure
was 1.2:I:0.1 g (dry tissue, mean i- standard error). This biomass is within the bounds of
the 95 % conﬁdence intervals predicted for the total phosphorus concentration of the pond
(13 pg 0 L4). Mussel density averaged 162:10 mussels (mean i standard error) per

treatment enclosure. Zebra mussels were monitored at each sampling date for mortality.

18

Sampling and laboratory analyses

The enclosures and the pond were sampled on 20 June (Day 0), 21 June (Day 1),
and at seven day intervals for the next ﬁve weeks (N = 37 days). All sampling was
conducted from a small boat.

A YSI multisensor (model 600XL) was used to measure temperature (°C), pH,
and dissolved oxygen (mg 0 L") at three depths (surface, 0.5 m, and 1.0 m) in each
enclosure and the pond. Readings were averaged over all depths for all analyses.

A clear plastic tube (5 cm in diameter and 1 min length) was used to take
integrated water samples (:2 L o tube’l) from the enclosures and the pond. The contents
of two tubes were placed into 10 L plastic cubitainers and stored in the dark on ice. At
the lab, each cubitainer was poured into a bucket and the sample was thoroughly mixed.

A 250 ml aliquot of water was ﬁltered through a Gelman A/E ﬁlter for particulate
phosphorus analysis (PP). After the ﬁlters were dried for 24 hours at 30 °C, they were
stored in a closed container with desiccant until further analysis. Filtrate was collected in
60 ml Nalgene bottles for analyses of total dissolved phosphorus (TDP) and ammonium
(NHX). PP and TDP were determined spectrophotometrically (Lambda 20, Perkin
Elmer) after potassium persulfate digestion (Menzel and Corwin 1965). N114+ was
measured by indophenol blue colorimetry (Wetzel and Likens 1991). Total phosphorus
(TP) was calculated as the sum of PP and TDP.

A 500 ml aliquot of water was ﬁltered through a Gelman A/E filter for
chlorophyll analysis. After ﬁltration, the ﬁlter was placed into a ti ghtly-sealed ﬁlm
canister and frozen. Chlorophyll a (jig 0 LI) was determined ﬂuorometrically (Turner

Designs 10A) after dark extraction in 95% ethanol for 30 hours.

19

A 100 ml sample was preserved with 1% Lugol’s solution for phytoplankton
counting. Phytoplankton samples from each enclosure and the pond were counted for
Days 8, 22, and 36. Depending on the chlorophyll concentration, 10 - 100 ml aliquots

from each enclosure were settled in Uterrnohl settling chambers, and sufﬁcient time (2 10

hours per cm of chamber height) was allowed for complete settling. Each chamber was
divided into circular inner and outer halves and an equal number of visual ﬁelds were
counted in each half (Sandgren and Robinson 1984). At least 15 ﬁelds per chamber half
were counted for the most abundant species and at most 100 ﬁelds from each half were
counted for most species. Phytoplankton were identiﬁed and enumerated to genus or
species with an inverted microscope at 400x and 1000x.

Phytoplankton were grouped into six categories according to morphological and
functional characteristics, as well as abundance. These groups consisted of small greens

< 10 um (Elakothrix spp., Nannochloris spp., Oocystis spp.), dinoﬂagellates (Ceratium

spp., Peridinium spp.), cryptomonads (Cryptomonas erosa, Cryptomonas pusilla,
Rhodomonas spp.), miscellaneous ﬂagellates < 10 um (Dinobryon spp., Trachlemonas
spp.), colonials (Uroglenopsis spp.), and others (desmids {Closterium spp., Cosman'um
spp., Staurastrum spp. }, diatoms {Nitzschia spp., Synedra spp. }, and ﬁlamentous greens).
For each sample, 10 randomly selected individuals of each common species were
measured with a micrometer. Formulas for simple geometric volumes that most
resembled particular species were used to calculate phytoplankton biovolume

(rim3 0 ml'l). Average phytoplankton biovolume per cell for each algal species did not
vary among treatments or dates so a single average (across treatment and dates) cell

volume was calculated for each species.

20

Ciliates were counted and measured in the same manner as the phytoplankton,
although ciliates were not identiﬁed into speciﬁc categories. Unlike phytoplankton
biovolume, average ciliate biovolume differed signiﬁcantly (p < 0.05) between
treatments, so I used separate average cell volumes for ciliates from zebra mussel
enclosures, control enclosures, and the pond for each sampling date. Ciliate and
phytoplankton biovolumes were converted to dry biomass (pg 0 LI) assuming a speciﬁc
gravity of 1 and a dry mass to wet mass ratio of 0.10.

Macrozooplankton were sampled by pouring the contents of seven integrated

tubes (z 14 L) through a 102 um mesh net, and preserved in 95% ethanol. For most

dates, the entire sample was counted and measured, otherwise 2 ml subsamples were
taken with a Henson-Stempel pipette and counted until 50 individuals of each species
were measured. Zooplankton subsamples were counted with a Ward zooplankton
counting wheel at magnifications between 20x and 80x. Cladoceran taxa included:
Bosmina spp., Ceriodaphnia spp., Chydorus spp. Daphnia retrocurva, and
Diaphanosoma spp. The copepod taxa measured were calanoid juveniles, cyclopoid
juveniles, Diacyclops spp., Diaptomus, Mesocyclops spp., Tropocyclops spp., and nauplii.
Dry biomass (pg 0 L") of each species was estimated from measured lengths using

length-weight regressions derived by Culver et al. (1985).

Data analyses
I used one-way analysis of variance (ANOVA) to assess treatment effects for
most parameters on Days 0, 1, 8, and 36 and on time-averaged data (Days 8 — 36). If no

statistically signiﬁcant treatment effects were detected for time-averaged data (P > 0.05),

21

I performed repeated-measures AN OVA on all data from Day 8 through Day 36 to
determine if any time x treatment interactions were present. None were found, so results
from the repeated-measures ANOVA are not shown. Effect sizes were calculated from
treatment means using the following formula: (100 * [(zebra mussel — control)/control]).
Linear regression was used to determine zebra mussel dry body weight (g) from length
(mm), and analysis of covariance was used to compare length-weight regressions for
mussels analyzed at the beginning of the experiment and for zebra mussels in each of the
four treatment enclosures at the end of the study. To determine if Tchhlorophyll
decoupling had occurred on Day 8, a two-sample t-test assuming unknown variance

(W elch test) was used. The observed chlorophyll concentrations for the two treatments
and the pond were compared to the predicted chlorophyll concentrations from the
Tchhlorophyll regression provided in Dillon and Rigler (1974). Log transformations
were applied to data if they were skewed or their variances were heterogeneous. Arcsine
transformations were applied to relative algal biomass estimates. All statistical analyses

were performed with Systat 8.0 (SPSS 1998). Rejection criterion was set at or < 0.05.

Results

Physical and chemical parameters
Although no measured parameters differed significantly between treatments on
Day 0 (the day before mussels were added), three variables (temperature, TDP, and TP)
were statistically different four hours after the addition of the zebra mussels on Day 1

(Table 4).

22

Temperature varied from 22 to 27 °C for all enclosures for all days, and the
average temperature throughout the experiment was approximately 25.7 °C (Table 4).
The pH in the enclosures ranged from 7.4 to 7.8 and was similar between treatments
(Table 4). Similarly, dissolved oxygen concentrations in the enclosures were comparable
and averaged approximately 6 mg 0 L'1 in the enclosures (Table 4).

Total dissolved phosphorus tended to decline over the course of the experiment
and averaged 3.5 pg 0 L'1 for all enclosures (Table 4). As expected, mean PP and NH;
differed signiﬁcantly between treatments (p = 0.01 and p = 0.02, respectively, Table 4).
Particulate phosphorus averaged < 3 pg o L'1 in the enclosures and was consistently
higher in the control enclosures. A similar trend was observed for TP (treatments
averaged < 6.5 ug - L'l). As early as Day 8, N114+ was higher in the zebra mussel
enclosures, and this effect was maintained throughout the remainder of the experiment

(except for Day 36). Ammonium ranged from 7.0 to 14.5 pg 0 L'1 in zebra mussel

enclosures and from 7.5 to 11.4 pg 0 L'1 in the control enclosures.

Biological parameters
Zebra mussels had a dramatic but ephemeral effect on phytoplankton abundance.
Chlorophyll concentrations differed signiﬁcantly between treatments (5-week average, p
= 0.001; Table 4) and averaged 1.1 ug 0 L'1 in the control enclosures and 0.6 pg o L'1 in
the zebra mussel enclosures. Although a large negative effect on chlorophyll
concentration was observed early in the experiment in the zebra mussel enclosures
(71.4% less chlorophyll in the zebra mussel enclosures when compared to the control

enclosures; Day 8), chlorophyll in the zebra mussel enclosures consistently increased

23

throughout the remainder of the study to eventually reach a concentration no different
than the control enclosures on Day 0 (p = 0.57; Figures 2, 3). The response of total
phytoplankton biomass (from microscope counts) paralleled that observed for chlorophyll
concentration (Table 4).

The Dillon-Rigler Tchhlorophyll equation (from Dillon and Rigler 1974)
predicted the chlorophyll concentrations observed in the control enclosures and the pond,
however the observed chlorophyll concentration for the zebra mussel enclosures on Day
8 was signiﬁcantly lower than predicted (p < 0.0001; Figure 4). Thus, the negative
effects of zebra mussels on phytoplankton biomass was not driven by a negative effect on
total phosphorus.

The effect of zebra mussels on phytoplankton species composition was modest.
Two algal groups were signiﬁcantly greater in the control enclosures (cryptomonads p =
0.003 and small greens p = 0.01, Table 4). When examining phytoplankton species
composition on individual days, all groups except colonials were signiﬁcantly lower in
the zebra mussel enclosures when compared to the control enclosures on Day 8 (Table 5).
However, the effect of zebra mussels on the phytoplankton community composition was
not maintained after Day 8. By Day 22, only one group signiﬁcantly differed between
treatments (small greens), and by the end of the study, no groups differed (Table 5). The
relative phytoplankton biomass data suggest that the mussels were predominantly non-
selective in their filtering of the phytoplankton throughout the entire study, with no
groups being different by Day 8 (Table 6, Figure 5) and only one algal group’s relative
biomass being signiﬁcantly different between treatments (5-week average, misc.

ﬂagellates p = 0.03; Table 6).

24

Ciliates signiﬁcantly declined throughout the entire study in the zebra mussel
enclosures. Ciliate biomass in the zebra mussel enclosures averaged 77.5% lower than
the control enclosures throughout the entire study (p < 0.0001, Table 4). Ciliate biomass
in the' zebra mussel enclosures was reduced by 70.6% by Day 8 and was maintained at
these reduced levels (or greater) until the end of the study (Figure 6).

Bosmina spp. and Diaphanosoma spp. accounted for > 99% of all cladocerans,
and calanoid and cyclopoid juveniles accounted for > 73% of all copepods for both
treatments. Copepods were almost 3 times more abundant than cladocerans in the zebra
mussel enclosures, and both groups were equally abundant in the control enclosures.
Zebra mussels had a signiﬁcant effect on total zooplankton biomass averaged over all
dates (p = 0.048, Table 4). Zebra mussels signiﬁcantly reduced total cladocerans
throughout the study (time-averaged, p = 0.035), but had no effect on copepods.

The ﬁrst sign of zebra mussel mortality was observed on 12 July (Day 22). I
continued to observe dead mussels throughout the experiment, however the average total
mortality observed in the zebra mussel enclosures over the entire experiment accounted
for only 7.6:3.6% (mean .+.. standard error) of the total density. The mussels that survived
were shown to have lost 56% of their initial weight by the completion of the study
(Figure 7). No differences were observed for post-experimental weights between
replicates of the zebra mussel treatment (AN COVA; p = 0.71), but there was a highly
signiﬁcant difference between length-weight relationships for pre-experimental mussels
[log dry tissue mass (g) = 2.5429 * log length (mm) — 4.9396, R2 = 0.93, N = 50] and

post-experimental mussels [log dry tissue mass (g) = 2.1455 * log length (mm) — 4.8135,

25

R2 = 0.76, N = 100 ) (Tukey-Kramer test of slopes, p < 0.0001), after correcting for the

weight lost due to freezing.

Pond Conditions

The pond was similar to the control enclosures for several parameters (i.e.,
temperature, pH, and NHI), but chlorophyll, PP, TDP, and TP were almost twice as high
in the pond as the control enclosures throughout the entire experiment (ﬁve-week
treatment averages, Table 4). Only on Day 8 were algal assemblages in the control
enclosures similar to that observed in the pond (Figure 5). Ciliate abundance in the pond
was similar to the control enclosures until Day 22, but by Day 36 the control enclosures
had 4 times greater ciliate biomass than the zebra mussel enclosures and the pond (Table
4). In addition, after two zooplankton inoculations, total zooplankton biomass measured
in the enclosures averaged 6.2 times lower than that observed in the pond (Table 4).
Finally, the zooplankton assemblage in the pond was very similar to that observed in the

control enclosures.

Discussion

As expected, zebra mussels had a dramatic negative impact on algal abundance
early in the experiment. By the end of the first week, the mussels reduced algal biomass
and chlorophyll concentrations by 53% and 71%, respectively. Similar but less steep
declines in chlorophyll occurred in the control enclosures shortly after zooplankton were
inoculated in all enclosures on Days 8 and 22. Although zebra mussels removed a

majority of the algae from the enclosures early in the study, their effect on algal

26

abundance was not maintained throughout the experiment. After Day 8, chlorophyll
concentrations continued to increase until the completion of the experiment, at which
time concentrations were similar Io the control enclosures. The inability of the mussels
to maintain the phytoplankton at a low level could be a result of an algal species shift to
inedible species, altered algal size structure, nutrient enrichment via zebra mussel
mortality, and/or to the declining health of the mussels.

Although the zebra mussels signiﬁcantly reduced most algal groups (except
colonials) by Day 8, a shift in phytoplankton species composition was not observed
because abundances of all algal groups were similar between treatments by Day 36.
Additionally, I did not encounter “inedible” algal species, such as colonial blue-greens,
thus it is unreasonable to conclude that a shift to less palatable species occurred. It also is
not likely that the size-structure of the algal assemblages affected mussel grazing because
all algal species were well within the size range mussels have been shown to consume
(range 1 — 150 pm; Ten Winkel and Davids 1982; Sprung and Rose 1988; Horgan and
Mills 1997) and all algal groups have been shown to be grazed by mussels in other
studies (Heath et al. 1995; Lavrentyev et al. 1995; Bastviken et al 1998; Smith et a1.
1998).

Zebra mussels immediately reduced and maintained low levels of ciliates
throughout the experiment (Figure 6), while phytoplankton abundance continued to
increase after Day 8 (Figure 2). This result is not surprising given that phytoplankton
were capable of acquiring available nutrients created via mussel excretion and mortality,
while ciliates competed with and were preyed upon by zebra mussels. Others have

shown similar impacts of zebra mussels on protozoan abundances. For example,

27

Lavrentyev et al. (1995) conducted laboratory experiments where they showed that zebra
mussels reduced protozoan abundance by > 70%, while the mussels had less of an impact

on phytoplankton abundance (z 45% decline). Additonally, competition between zebra

mussels and ciliates for small edible algae, like small greens, could help explain the
observed effects on ciliate and algal abundances. A positive correlation between ciliates
and small green algae for the enclosures (Figure 8) suggests that although the zebra
mussels reduced all algal groups equally, the large absolute reduction of small green
algae (68%, Table 5) could have aided in keeping ciliate numbers low. Thus, zebra
mussels are extremely effective at controlling ciliates, however the importance of
competition and predation in zebra mussels’ ability to control ciliates is currently
unresolved. Thus, further research directed at understanding the competitive and
predative interactions between protozoans and zebra mussels will aid in developing more
complete and accurate food-web models.

The deteriorating health of the mussels could also help explain the lack of effect
on algal concentration observed later in this experiment. Several studies have examined
the role of physical and chemical parameters in regulating zebra mussels in lakes
(Ramcharan et al. 1992; Ludyanskiy et al. 1993; Mackie and Schloesser 1996; Karatayev
et al. 1998). Speciﬁcally, calcium provides the material for mussel shell construction and

mussels require at least 20 mg 0 L'1 Ca”2 to establish populations (Ludyanskiy et al.
1993). The pond used in this study had calcium concentrations above 50 mg 0 L'1 Ca+2

(S. Hamilton and D. Raikow, personal communication), thus calcium limitation likely did
not affect mussel growth and maintenance. Ramcharan et al. (1992) demonstrated that

mussels are sensitive to pH and found mussels to be absent from lakes with pH < 7.3.

28

The pH in the enclosures ranged from 7.4 to 7.8; therefore, it is not likely that pH
signiﬁcantly affected zebra mussel health.

Most temperate species are adapted to seasonally changing thermal environments,
however extremely high (or low) temperatures can be lethal. Karatayev et al. (1998)
suggest that zebra mussels thrive in temperatures below 27°C and have difﬁculty
surviving in temperatures greater than 32 °C. Although the enclosures reached 27 °C,

the average measured temperature experienced by the mussels throughout the entire study
was 25 °C. Thus it seems unlikely that temperature by itself affected mussel health and
grazing, however, higher temperatures could have made routine physiological
maintenance more difﬁcult due to higher food requirements at higher temperatures (W alz
1978; Aldridge et al. 1995; Fanslow et al. 1995; Horgan and Mills 1997).

An individual zebra mussel’s growth rate is dependent on body size and
temperature, among other factors (Walz 1978). Walz (1978) indicates that higher
temperatures severely restrict zebra mussel growth rates due to a greater demand for food
and that larger mussels require less food per gram of mussel than smaller mussels (W alz
197 8; James et al. 2000). Comparisons of length-weight relationships performed on the
mussels before and after the study show that the mussels lost 56% of their weight during
the experiment (Figure 7). Average available daily rations were calculated by converting
measured chlorophyll concentrations into carbon (u g) (carbonzchlorophyll = 67: 1;
Riemann et al. 1989) and then calculating ﬁltration rates based on the size distribution
and density of mussels used in the study (Kryger and Riisgard 1988). The average daily

ration on Day 8 of the experiment was 31 u g carbon - clay'1 in the mussel enclosures.

Walz (197 8) used Nitzschia, a high-quality laboratory cultured diatom, to show that a 5

29

mg (dry tissue mass) mussel requires 42.5 p g carbon 0 day’1 for routine maintenance at
20 °C. Although this size is slightly larger than the average mussel used in this study (3.7
mg dry tissue mass), smaller mussels typically require more food per unit body weight
than larger mussels (W alz 1978). Additionally, given the lower quality pond water,
which contained a considerable amount of detritus, the zebra mussels would have needed
to ﬁlter more enclosure water when compared to a similar amount of cultured Nitzschia
medium to acquire the carbon needed for basic maintenance. My calculations indicated
that food abundance in the zebra mussel enclosures was below that required for basic
maintenance on Day 8 (Figure 9). Thus, it is reasonable to conclude that the mussels lost
weight due to insufficient food availability soon after they were added to the enclosures.
Although I attempted to predict a reasonable stocking density of zebra mussels for my
enclosures based on the TP concentration of the pond (13 pg 0 LI), the enclosures were
half as productive as the pond (Table 4), and, consequently, I over-estimated the amount
of mussels required. With this in mind, careful consideration of the mussels’ food
availability must be accounted for when designing ﬁeld experiments with zebra mussels
in enclosed, previously mussel-free systems. Given that many earlier experimental
studies involving zebra mussels used excessively high amounts of zebra mussels,
presumably to see a grazing effect, my study clearly shows how critical determining a
natural density of zebra mussels can be to outcome of the study. Also, this is ﬁrst study
to compare length-weight relationships to monitor the health of zebra mussels during an
experiment. Without this type of data, conclusions based on the health of the mussels
would only be conjecture. Thus, future experiments using zebra mussels should; 1) take

precautions to not overstock their zebra mussel treatments by either taking benthic

30

samples to determine a biomass estimate for water bodies where mussels have already
invaded or by measuring total phosphorus concentration of the experimental units for
mussel-less systems and then calculating dry tissue biomass, 2) monitor total phosphorus
levels in enclosures related to the natural system, and 3) monitor the health of the mussels
throughout the experiment.

In conclusion, zebra mussels have been shown to quickly reduce algae and ciliates
soon after entering a new water body. However, their impact on phytoplankton biomass
was shown to diminish within two weeks of their introduction. The deteriorating health
of the mussels has been proposed to help explain this effect. Future work incorporating a
long-term, large-scale, controlled ﬁeld experiment aimed at examining the gradual
development of a zebra mussel founder population and its effects on the community will
elucidate the important complex interactions between zebra mussels and other food web

components.

31

Table 1. Lake characteristics and dreissenid abundance for Polish lakes. Mussel

biomass expressed as dry tissue mass. Year of sampling (when known) in parentheses.

 

 

Mean Maximum TP Calcium Mussel
Lake depth depth (mg . m3) (mg - L") Biomass

(m) (m) (g°m‘2)
Beldany 10.0 31.0 55.0" (’76) 33.0' 0.2I (’62)
Boczne 8.7 15.0 157.0" (’76) 54.0“ 19.6' (’62)
Dargin 10.6 37.0 63.0k (’76) 63.1“ 7.71l (’62)
Dobskie 7.8 21.0 60.0k (’76) 53.2“" 5.41 (’62)
Glebokie 11.8 34.3 62.5"f 5.0j (’76)
Goldopiwo 24.5 465’ 9.71 (’62)
Inulec 4.6 10.1 147.0"f 6.7j (’76)
Jagodno 8.7 34.0 92.0k (’76) 64.0a 18.4' (’62)
Jorzek 5.5 11.6 111.0"f 7.0j (’76)
Kierzlinski 11.7 44.0 38.0k (’77) 45.0g 2.5d (’77)
Kisajno 8.4 24.0 54.4“" (’76) 54.0” 7.4' (’62)
Kolowin 4.0 7.2 53.0“ (’78) 15.1(1 (’78)
Kotek 1.0 2.5 103.01‘ (’76) 50.02’ 4.7I (’62)
Kuc 8.0 28.0 40.01‘ (’77) 1.1d (’78)
Majcz Wielki 6.0 16.4 18.5f 12.0"J(’76)
Mamry 11.7 40.0 43.4""‘(’76) 34.58l 13.41 (’62)
Mikolajskie 11.1 27.8 60.0k (’76) 36.01 0.6I (’62)

32

Table l. (cont’d)

 

 

Niegocin 10.0 40.0 233.0" (’76) 66.1“1 17.7I (’62)
Olow 12.9 40.1 29.5““ (’77) 46.0g 43.2d (’78)
Pilakno 13.0 56.6 20.0" (’77) 38.08 0.8(1 (’77)
Probarskie 9.2 31.0 38.0" (’77) 8.8d (’77)
Ros 29.0 13.91 (’62)
Sniardwy 5.9 25.0 38.01“ (’76) 32.01 5.2I (’62)
Stregiel 12.5 49.0“ 51.3' (’62)
Szymon 1.1 2.9 87.0“ (’76) 21.0I (’62)
Tajty 7.6 34.0 55.0k (’76) 54.0“ 17.1' (’62)
Taltowisko 14.0 38.4 54.0k (’76) 75.0“ 12.6I (’62)
Talty 13.6 37.5 36.0k (’76) 64.0“ 0.6I (’62)
Wilkus 5.5 44.0“ 25.8I (’62)
Zabinska 42.5 32.2I (’62)
Zelwazek 3.7 7.4 5.5j (’76)
3 Mean 8.5 26.2 69.9 50.1 12.7 I

 

Data sources: aGieysztor and Odechowska 1958; bHillbricht-Ilkowska et al. 1984; cKajak
and Zdanowski 1983; dLewandowski 1991; °Patalas 1960; fPlanter and Wisniewski 1985;
gPrusik et al. 1989; hSpodniewska 1978; ’Stanczykowska 1977; jStanczykowska et al.

1983; "Zdanowski 1982.

33

Table 2. Total phosphorus, mean depth and dreissenid abundance for North American

lakes. Mussel biomass expressed as dry tissue mass.

 

Mean Mussel

 

TP Year Year Depth biomass
Lake (mg - m'3) invaded sampled (m) (g - m'z)
Erie (eastern basin) 11.7“ 1988 ‘92-‘93 25.0 7.7b
Gull 14.0i 1994 ’99 12.5 6.11
Oneida 190" 1991 ‘92-’93 6.8 440"
Ontario 9.3d 1990 ’95 86.0 0.9f
Saginaw Bay, Lake Huron (inner bay) l9.0° 1990 ‘93 7.2 4.53
St. Clair 22.5c 1986 ‘90-‘94 3.8 3.8h

Data sources: alBertram 1993; bDermott and Kerec 1997 cFahnenstiel et al. 1995;

dJohengen et al. 1994; eMellina et al. 1995; fMills et al. 1999; gNalepa et al. 1995;

hNalepa et al. 1996; ithis study.

34

' Tl

“mam-AH.

Table 3. Multiple regression statistics for the inﬂuence of mean
depth (m), calcium (mg - LI) and log total phosphorus (mg 0 m3)

on zebra mussel biomass in Polish lakes. N = 16.

 

 

Variable Slope Standard error P
Log mean depth 5.5 5.8 0.36
Log calcium 11.7 13.9 0.42
Log total phosphorus 16.9 6.8 0.03
Full model 0.04

 

35

 

 

 

 

$_ 3.8 $3 2 8 $3.28 $8.88 $8.28 84.3-8.8 a: 73.8 2 82-3.8 3.2 8:
8.0 moo Ed who mod 2.2 end a: 828.52
:2 was
$32.8 $2.8 $8.23 $8.2 8 $8.2 Z 83-8.8 25.2-8 .8 $82.38 8229.28
33 who 88 3o 88 can 43 8.2 38.
:2 8:
$838 $388 $8.21 $8.88 $8.88 821 2.8 A 8.38 8 63-5.8 88688.22
3o :.o 88 N3 :3 8% ohm mom 8:685 38
:2 m8
$28-8 $388 $83-8 $8.3 $2.28 84.28.: 5.3.828 20.3-88.8 3688.3
3o 88 «so 26 So 2 .N 32 Rs 823252
:2 was
$8.8 $8.3 $88 $2. 2 -8 $8.3 2 2.88.8 63- 2 38 28.4-8.8 5&5
88 Ed 83 :8 :3 New Rd SA 83685
$2.8 $8.8 $8.3 $88 $2 .3 2883.8 5.23.8 243.28
8.0 mg 8 .o 3o 38 on: 8: n: ma
$8.8 $88 $8 .8 $2.88 $8.88 88.8-2.8 seem-8&8 2 ~33.va 68
mod woo 5o 88 «no 28mm 92 8.2 sausage
om warm w ED 2 man 2 3D swab; +2N 38:00 econ 8335/
goes-m

Quote “2253.3
mos—Si

Eobcoo .m> 232:: men-.5 ﬂue-cm 22585. 28 358$:me

I 6828 8205. ... 83 acute “2253.: can Jon—«>8 .3m 4 Page memes-2 .Gm-w Page mews-82” 252 was Eugen-Q.

Gm - 2 825 Hows—Ev
moments “2253;.

22282“ 8: u < Z 688820 05 5:» 2285.888
a 8 26 3982 so: me? In .3 32m :0 92088.82 Rama—OE 28 286.38 288223 28 5822850528

.v 2an

36

:2 8:

 

 

 

 

$38-8 $3.2 3 $283 m£2522 22
moo owe <2 <2 86 Ed 8: 3.0 825328
2.2 m8
$38-8 $3.03 $82.3 $255 22
88 :3 <2 <2 cod 88 m3 8% 225
9-2 8:
$33-8 $_ 33 $ 382 $285 22
as 88 <2 <2 88 83 8 .2 8.3 535823
$82-8 $3 2 E $89-8 $8.28 $3.83 28.2-8.8 8272.8 83-2.8 2.2 88
So cod 85 ed So one :2 SN =an820
cm awn w an _ 2mm o 39 owﬁog +EN 32:00 ucom 833:3
283%

288:0 258235

32?-2
305:8 .m> 2838 2208 ﬂootm 2:253; mo 8:32.:ch

Gm - 2 825 Human:
mamas; 25823;

8.288 .v 222-

37

 

 

 

ood okvvm- 86 $9”- cod $3- 86 $3. .808
3.0 «3 N00 9&3”- cod 9.3.0- 36 oxymm- 550
«m6 §w ﬂ - 2 .o 05;- vmd 95m? vmd 955m- 3:200
A81 om vv
mod $5- cod 05 3- mod 95 :1 36 95%. £880 =aEm
o3 gin- 86 $2- 85 03w- cod $8- ”2859,50
E: 2 vv
omd 83m Ed §~- cod @091 mod 95: 33:0qu 8:2
3.0 $3- wad $3“ mod Q. a m- 3.6 $3- m8m__owaao:5
mos—gd 855:6 mos—9d 855:6 mos—gd oocobtﬁ mos—.35 858:6 98m Em?
Euﬁﬁoﬁ. 2253; 8258; E2535.
6m Ea .S .w 38
cm ‘80 an E5 w 3Q mowﬁo>< 22:305.
9%.“ 308885 can .mEoEB .mEEmou 38:06 :550: 25 9.4 a we @3805 iv Rm? :38 mm 8.532:

8a 335. =< $95. 05 Sonwaobt anew Rm.“ 5 205:8 .m> +83 mooceotﬁ E2535 82890 82o3< .m 053.

38

 

 

 

who $3 3.0 $m- Ed $vm- mod $w 350
wmd $mm hmd $Nm- 3 .o $mom wwd $w- 3:200
ea 2 vv
2 .0 $2- 5.0 $31 and $m~ a: .0 $9“- 535 :55
3.0 $5M- wwd $w- 00.0 $3- 3 .o $5“- 5505930
ea 2 vv
nod $oo_ mmd $>m wed $o mod $00 33:0me 0&2
omd $3- 2.0 $03 mud $5 3.0 $m 33:03:85
3295 33:56 3295 833:6 323m 833:5 3235 333:6 9.on Em?
5253; 5253; 5353; 535535.
6m Ba .8 .w 38
©m 3Q mm 3Q w 39 3333‘ 5253;

dam—m 55335 Ba £5535 .5535 38:3 :350: was $3505 :35 95 $v $5505 5 3522 3 35305

03 3:on :< 265.0, 05 52.385 529% Rm.“ 5 205:8 .m> +55 moo—85.5% 505535 3E3? gnu—om no 2an

39

 

 

 

 

 

:5"
E. 40 - OneidaO . Olow . Poland
on O N. America
5 30 -
a
g 20 - .0 .1
¢ . O
:5 10 - T
Q) 0 O . . .
a 0 _ o O .
.59.
a
t} l I l
g 0.8 1.2 1.6 2.0 2.4
Log total phosphorus (mg - m'3)
Figure 1. Relationship between total phosphorus and dreissenid biomass for Polish

and North American lakes. Data from Tables 1 & 2. Regression line and
95% conﬁdence bands for equation estimates were calculated for Polish
lakes after excluding the outlier (Lake Olow).

4O

   

 

—0— Control

Chlorophyll (pg . L")

 

v —V— ZM +
0
0 10 20 30 40
Time (Days)
Figure 2: Chlorophyll concentration in the control and zebra mussel enclosures

(daily treatment averages i 1 standard error).

41

Figure 3.

Effect size

 

-10%

-20% '

-30% '

-40% '

-50% '
-60% '
-70% .

 

 

-80%

 

O 10 20 30 40

Time (Days)

Measured effect size (100 * [(zebra mussel — control) / control]) of zebra
mussels on phytoplankton abundance (measured as chlorophyll a) over the
37 days of the experiment.

42

 

      
  
    

 

 

Regression and ...... '.
prediction intervals from ..... V.
.11" Dillon and Rigler 1974 ........

.4 ..............
. .............
w ...............

1 .............

v ............
cu ................

= 0 ..........
>3 °°°°°

.ﬂ .......

a. ......
¢ ,,,,,
3.. .....

.2 .......

u: .....

D
g” D Control

A V ZM+

-1
O 5 1 1 5
Log Total Phosphorus (pg . L'l)
Figure 4. Observed chlorophyll a concentrations for Day 8 plotted in relation to the

Dillon-Rigler TP — chlorophyll relationship (logChl a = 1.449 * logTP —
1.136). 95% prediction intervals for Dillon-Rigler are provided. Error
bars represent i 1 standard error.

43

 

 

 

 

 

 

 

 

 

 

 

 

co
ca
2
.2 1.0
CD
30.8
65 0.6
To
200.4
<1
2 0.2
'ﬁ
3 00V
Di Z C P Z C p Z C P
M 0 o M 0 o M o 0
t d t d t d
r r 1'
o o 0
l l 1
I Others
EH Colonials
a Small Greens
3 Cryptomonads
E] Flagellates (<10um)
El Dinoﬂagellates
Figure 5. Relative al gal group dry biomass (%) for zebra mussel enclosures, control

enclosures, and the pond for Day 8, Day 22, and Day 36.

 

100 '

 

 

ZM +
80 ' m Control

 

 

 

 

 

Ciliate Dry Biomass (pg . L'l)

 

Figure 6. Mean ciliate dry biomass (pg 0 L'l, :t 1 standard error) for zebra mussel
and control enclosures and pond for Days 8, 22, and 36.

45

Mussel dry tissue weight (mg . mm'l)

Figure 7.

 

0.8

 

0.7 ‘ - Pre-experimental

Post-experimental

 

 

 

0.6 ~
0.5 ~
0.4 l
0.3 -

0.2 ‘

 

0.1 "

 

 

 

 

0.0 ‘

Dry tissue biomass (mg 0 mm“) for average mussel length (12mm) for
pre- and post-experimental zebra mussels. A correction factor for the
effect of freezing as a preservative technique was used to determine the
post-experimental mussel weights. Error bars represent 1 standard error.

46

 

 

 

 

 

 

 

.3" 0.6
A El
’ 0.5 . El
DD
:3.
V
a 0.4 .
a a
c 0.3 - '3
0-
no ,,
>. El Control
3.. 0.2 -
a v ZM+
3

0.1 ~ W
.3 a
0‘

o , C ﬁ ﬁ , , ,

O 0.1 0.2 0.3 0.4 0,5 0.6 0.7 0.8

Small Green Dry Biomass (ug - L'l)

Figure 8. Time-averaged (Days 8, 22, and 36) data on small green algal group
Dry biomass (ug 0 LI) vs. ciliate dry biomass (pg 0 LI) for the four
replicates of both treatments.

47

Available Carbon (ug . day'l)

Figure 9.

180

160 ‘

14o -

120 -

100

60 ‘

40

20

 

 

F."

IIIIJI rIIIIIIIIIIIIIIIIIiIII-IIIIIIIII-Illdi

 

 

 

 

 

 

 

 

 

 

UV ‘1
T I

lull-Ildi

 

 

 

 

0 10 20 30

Time (Days)

Average available carbon (um’3 0 day) in the zebra mussel enclosures.
Horizontal dotted line indicates minimum amount of carbon necessary for
routine maintenance without growth based on mussels fed Nitzschia (from

Walz 1978).

48

BIBLIOGRAPHY

Aldridge, D. W., B. S. Payne, and A. C. Miller. 1995. Oxygen consumption,
nitrogenous excretion, and ﬁltration rates of Dreissena polymorpha at acclimation
temperatures between 20 and 32 °C. Canadian Journal of Fisheries and Aquatic
Sciences. 52: 1761-1767.

Amott, D. L., and M. J. Vanni. 1996. Nitrogen and phosphorus recycling by the zebra

mussel (Dreissena polymorpha) in the western basin of Lake Erie. Canadian Journal
of Fisheries and Aquatic Sciences 53:646-659.

Bastviken, D. T. E., N. F. Caraco, and J. J. Cole. 1998. Experimental measurements of
zebra mussel (Dreissena polymorpha) impacts on phytoplankton community
composition. Freshwater Biology 39:375-386.

 

Bertram, P. E. 1993. Total phosphorus and dissolved oxygen trends in the central basin
of Lake Erie, 1970-1991. Journal of Great Lakes Research 19:224-236.

Caraco, N. F., J. J. Cole, P. A. Raymond, D. L. Strayer, M. L. Pace, 8. E. G. Findlay,
and D. T. Fischer. 1997. Zebra mussel invasion in a large, turbid river: phytoplankton
response to increased grazing. Ecology 78:588-602.

Culver, D. A., M. M. Boucherle, D. J. Bean, and J. W. Fletcher. 1985. Biomass of
freshwater crustacean zooplankton from length-weight regressions. Canadian Journal
of Fisheries and Aquatic Sciences 42: 1380-1390.

Dermott, R., and D. Kerec. 1997. Changes to the deepwater benthos of eastern Lake
Erie since the invasion of Dreissena: 1979-1993. Canadian Journal of Fisheries and
Aquatic Sciences 54:922-930.

Dillon, P. J. and F. H. Rigler. 1974. The phosphorus-chlorophyll relationship in lakes.
Limnology and Oceanography 19:767-773.

Fahnenstiel, G. L., T. B. Bridgeman, G. A. Lang, M. J. McCormick, and T. F. Nalepa.
1995. Phytoplankton productivity in Saginaw Bay, Lake Huron: effects of zebra

mussel (Dreissena polymorpha) colonization. Journal of Great Lakes Research
21:465-475.

Fahnenstiel, G. L., G. A. Lang, T. F. Nalepa, and T. H. Johengen. 1995. Effects of zebra
mussel (Dreissena polymorpha) colonization on water quality parameters in Saginaw
Bay, Lake Huron. Journal of Great Lakes Research 21:435-448.

Fanslow, D. L., T. F. Nalepa, and G. A. Lang. 1995. Filtration rates of the zebra mussel

(Dreissena polymorpha) on natural seston from Saginaw Bay, Lake Huron. Journal of
Great Lakes Research 21:489-500.

49

Gieysztor, M., and Z. Odechowska. 1958. Observations on the thermal and chemical
properties of Mazurian lakes in the Gizycko region. Polskie Archiwum Hydrobiologii
4: 123-152.

Gillis, P. L., and G. L. Mackie. 1994. The impact of Dreissena polymorpha on
populations of Unionidae in Lake St. Clair. Canadian Journal of Zoology 72: 1260-
1271.

Hamilton, D. J ., C. D. Ankney, and R. C. Bailey. 1994. Predation of zebra mussels by
diving ducks: an enclosure study. Ecology 75:521-531.

Hanson, J. M., and R. H. Peters. 1984. Empirical prediction of crustacean zooplankton
biomass and profundal macrobenthos biomass in lakes. Canadian Journal of Fisheries
and Aquatic Sciences 41:439—445.

Heath, R. T., G. L. Fahnenstiel, W. S. Gardner, J. F. Cavaletto, and S.-J. Hwang. 1995.
Ecosystem-level effects of zebra mussels (Dreissena polymorpha): an enclosure
experiment in Saginaw Bay, Lake Huron. Journal of Great Lakes Research 21:501-516.

Hebert, P. D. N ., C. C. Wilson, M. H. Murdoch, and R. Lazar. 1991. Demography and
ecological impacts of the invading mollusc Dreissena polymorpha. Canadian Journal
of Zoology 69:405-409.

Hillbricht-Ilkowska, A., W. Lawacz, and R. Wisniewski. 1984. External and internal
loading and retention of phosphorus in the R. Jorka lakes (Masurian Lakeland,
Poland) vs. their trophic status. Internationale Vereinigung fiir Theoretische und
Angewandte Limnologie, Verhandlungen 22:973-977.

Holland, R. E. 1993. Changes in planktonic diatoms and water transparency in Hatchery
Bay, Bass Island area, western Lake Erie since the establishment of the zebra mussel.
Journal of Great Lakes Research 19:617-624.

Horgan, M. J ., and E. L. Mills. 1997. Clearance rates and ﬁltering activity of zebra
mussel (Dreissena polymorpha): implications for freshwater lakes. Canadian Journal
of Fisheries and Aquatic Sciences 54:249-255.

Jack, J. D., and J. H. Thorp. 2000. Effects of the benthic suspension feeder Dreissena
polymorpha on zooplankton in a large river. Freshwater Biology 44:569-479.

James, W. F., J. W. Barko, M. Davis, H. L. Eakin, J. T. Rogala, and A. C. Miller. 2000.
Filtration and excretion by zebra mussels: implications for water quality in Lake Pepin,
Upper Mississippi River. Journal of Freshwater Ecology 15:429-437.

James, W. F., J. W. Barko, and H. L. Eakin. 1997. Nutrient regeneration by the zebra
mussel (Dreissena polymorpha). Journal of Freshwater Ecology 12:209-216

50

 

 

Johengen, T. H., O. E. Johannsson, G. L. Pemie, and E. S. Millard. 1994. Temporal and
seasonal trends in nutrient dynamics and biomass measures in Lake Michigan and

Ontario in response to phosphorus control. Canadian Journal of Fisheries and Aquatic
Sciences 51:2570-2578.

Johengen, T. H., T. F. Nalepa, G. L. Fahnenstiel, and G. Goudy. 1995. Nutrient changes
in Saginaw Bay, Lake Huron, after the establishment of the zebra mussel (Dreissena
polymorpha). Journal of Great Lakes Research 21:449-464.

Kajak, Z., and B. Zdanowski. 1983. Ecological characteristics of lakes in north-eastern
Poland versus their trophic gradient. 1. General characteristics of 42 lakes and their
phosphorus load, study objectives and scope. Ekologia Polska 31:239-256.

Karatayev, A. Y., L. E. Burkalova, and D. K. Padilla. 1998. Physical factors that limit
the distribution and abundance of Dreissena polymorpha (Pallas). Journal of Shellﬁsh
Research 17:1219-1235.

Klerks, P. L., P. C. Fraleigh, and J. E. Lawniczak. 1996. Effects of zebra mussels
(Dreissena polymorpha) on seston levels and sediment deposition in western Lake
Erie. Canadian Journal of Fisheries and Aquatic Sciences 53:2284-2291.

Kornobis, S. 1977. Ecology of Dreissena polymorpha (Pallas) (Dreissenidae, Bivalvia)
in lakes receiving heated water discharges. Polskie Archiwum Hydrobiologii 24:531-
545.

Kryger, J ., and H. U. Riisgard. 1988. Filtration rate capacities in 6 species of European
freshwater bivalves. Oecologia 77:34-38.

Lavrentyev, P. J ., W. S. Gardner, J. F. Cavaletto, and J. R. Beaver. 1995. Effects of
zebra mussel (Dreissena polymorpha Pallas) on protozoa and phytoplankton from
Saginaw Bay, Lake Huron. Journal of Great Lakes Research 21:545-557.

Lavrentyev, P. J ., W. S. Garner, and L. Y. Yang. 2000. Effects of the zebra mussel on
nitrogen dynamics and the microbial community at the sediment-water interface.
Aquatic Microbial Ecology 21:187-194.

Lewandowski, K. 1991. The occurrence of Dreissena polymorpha (Pall) in some
mesotrophic lakes of the Masurian Lakeland (Poland). Ekologia Polska 39:273-286.

Lowe, R. L., and R. W. Pillsbury. 1995. Shifts in benthic algal community structure and
function following the appearance of zebra mussels (Dreissena polymorpha) in
Saginaw Bay, Lake Huron. Journal of Great Lakes Research 21:558-566.

Ludyanskiy, M. L., D. McDonald, and D. MacNeill. 1993. Impact of the zebra mussel, a
bivalve invader. Bioscience 43:533-544.

51

McCauley, E., J. A. Downing, and S. Watson. 1989. Sigmoid relationships between
nutrients and chlorophyll among lakes. Canadian Journal of Fisheries and Aquatic
Sciences 46: 1 171-1 175.

MacIsaac, H. J ., C. J. Lonnee, and J. H. Leach. 1995. Suppression of microzooplankton
by zebra mussels: importance of mussel size. Freshwater Biology 34:379-387.

Mackie, G. L., and D. W. Schloesser. 1996. Comparative biology of zebra mussels in
Europe and North America: an overview. American Zoologist 36:244-258.

Mellina, E., and J. B. Rasmussen. 1994. Patterns in the distribution and abundance of
zebra mussel (Dreissena polymorpha) in rivers and lakes in relation to substrate and

other physicochemical factors. Canadian Journal of Fisheries and Aquatic Sciences
51:1024-1036.

Mellina, E., J. B. Rasmussen, and E. L. Mills. 1995. Impact of zebra mussel (Dreissena
polymorpha) on phosphorus cycling and chlorophyll in lakes. Canadian Journal of
Fisheries and Aquatic Sciences 52:2553-2573.

Menzel, D. W. and N. Corwin. 1965. The measurement of total phosphorus in seawater
based on the liberation of organically bound fractions by persulfate oxidation.
Limnology and Oceanography 10:280-281.

Mills, E. L., J. R. Chrisman, B. Baldwin, R. W. Owens, R. O’Gorman, T. Howell, E. F.
Roseman, and M. K. Raths. 1999. Changes in the dreissenid community in the lower

great lakes with emphasis on southern Lake Ontario. Journal of Great Lakes
Research. 25:187-197.

Mittelbach, G. G. 1988. Competition among refuging sunﬁshes and effects of ﬁsh
density on littoral zone invertebrates. Ecology 69:614-623.

Moss, B. 1972. Studies on Gull Lake, Michigan. 1. Seasonal and depth distribution of
phytoplankton. Freshwater Biology 22289-307.

Murphy, J ., and L. P. Riley. 1962. A modiﬁed single solution method for the
determination of phosphate in natural waters. Analytica Chirnica Acta 27 :3 1-36.

Nalepa, T. F., D. J. Hartson, G. W. Gostenik, D. L. Fanslow, and G. A. Lang. 1996.
Changes in the freshwater mussel community of Lake St. Clair: from Unionidae to
Dreissena polymorpha in 8 years. Journal of Great Lakes Research 22:354-369.

Nalepa, T. F., J. A. Wojcik, D. L. Fanslow, and G. A. Lang. 1995. Initial colonization of

the zebra mussel (Dreissena polymorpha) in Saginaw Bay, Lake Huron: population
recruitment, density and size structure. Journal of Great Lakes Research 21:417-434.

52

Pace, M. L., S. E. G. Findlay, and D. Fischer. 1998. Effects of an invasive bivalve on
the zooplankton community of the Hudson River. Freshwater Biology 39:103-116.

Patalas, K. 1960. Charakterystyka skladu chernicznego wody 48 jezior okolic
wegorzewa. Roczniki Nauk Rolniczch, Serial B 77:243-297.

Planter, M., and R. J. Wisniewski. 1985. Factors affecting nutrient budget in lakes of the
R. J orka watershed (Masurian Lakeland, Poland). IX. The exchange of phosphorus
between sediments and water. Ekologia Polska 33:329-344.

Prusik, S., B. Zdanowski, and A. Hutorowicz. 1989. Sezonome zmiany warunkow
srodowiska w jeziorach dymiktycznych o roznym stopniu eutroﬁzacji. Roczniki Nauk
Rolniczych, Seria H 102:43-75.

Ramcharan, C. W., D. K. Padilla, and S. I. Dodson. 1992a. A multivariate model for
predicting population ﬂuctuations of Dreissena polymorpha in North American lakes.
Canadian Journal of Fisheries and Aquatic Sciences 49: 150-158.

Ramcharan, C., D. K. Padilla, and S. I. Dodson. 1992b. Models to predict the potential
occurrence and density of the zebra mussel, Dreissena polymorpha. Canadian Journal
of Fisheries and Aquatic Sciences 49:2611-2620.

Rasmussen, J. B., and J. Kalff. 1987. Empirical models for zoobenthic biomass in lakes.
Canadian Journal of Fisheries and Aquatic Sciences 44:990-1001.

Reeders, H. H., and A. Bij de Vaate. 1990. Zebra mussels (Dreissena polymorpha): a
new perspective for water quality management. Hydrobiologia 200]201:437-450.

Riemann, B., Simonsen, P., and Stensgaard, L. 1989. The carbon and chlorophyll content
of phytoplankton from various nutrient regimes. Journal of Plankton Research 11:1037-
1045.

Roditi, H. A., N. F. Caraco, J. J. Cole, and D. L. Strayer. 1996. Filtration of Hudson
River water by the zebra mussel (Dreissena polymorpha). Estuaries 19:824-832.

Sandgren, C. D., and J. V. Robinson. 1984. A stratiﬁed sampling approach to
compensating for non-random sedimentation of phytoplankton cells in inverted
microscope settling chambers. British Phycological Journal 19:67-72.

Samelle, O., S. D. Cooper, S. Wiseman, and K. M. Mavuti. 1998. The relationship
between nutrients and trophic-level biomass in turbid tropical ponds. Freshwater
Biology 40:65-75.

Schindler, D. W. 1977. The evolution of phosphorus limitation in lakes. Science
195:260-262.

53

Schindler, D. W. 1978. Factors regulating phytoplankton production and standing crop
in the world’s freshwaters. Limnology and Oceanography 23:478486.

Smith, T. E., R. J. Stevenson, N. F. Caraco, and J. J. Cole. 1998. Changes in
phytoplankton community structure during the zebra mussel (Dreissena polymorpha)
invasion of the Hudson River (New York). Journal of Plankton Research 20: 1567-1579.

Smith, V. H. 1982. The nitrogen and phosphorus dependence of algal biomass in lakes:
an empirical and theoretical analysis. Limnology and Oceanography 27:1101-1112.

Spodniewska, I. 1978. Phytoplankton as the indicator of lake eutrophication. 1. Summer
situation in 34 Masurian lakes in 1973. Ekologia Polska 26:53-70.

Sprung, M. 1987. Ecological requirements of developing Dreissena polymorpha eggs.
Archiv fiir Hydrobiologie/ Supplementband 79:69-86.

Sprung, M., and U. Rose. 1988. Inﬂuence of food size and food quantity on the feeding
of the mussel Dreissena polymorpha. Oecologia 77:526-532.

Stanczykowska, A. 1977. Ecology of Dreissena polymorpha (Pall.) (Bivalvia) in lakes.
Polskie Archiwum Hydrobiologii 24:461-530.

Stanczykowska, A. 1984. The effect of various phosphorus loadings on the occurrence
of Dreissena polymorpha (Pallas). Limnologica 15:535-539.

Stanczykowska, A., H. J. Schenker, and Z. Farara. 1975. Comparative characteristics of
populations of Dreissena polymorpha (Pall.) in 1962 and 1972 in 13 Mazurian lakes.
Bulletin de l'Académie polonaise des sciences. Série des sciences biologiques.
23:383-390.

Stanczykowska, A., K. Lewandowski, and J. Ejsmont-Karabim. 1983. Biotic structure
and processes in the lake system of R. Jorka watershed (Masurian Lakeland, Poland).
IX. Occurrence and distribution of molluscs with special consideration to Dreissena
polymorpha (Pall.). Ekologia Polska 31:761-780.

Ten Winkel, E. H., and C. Davids. 1982. Food selection by Dreissena polymorpha
Pallas (Mollusca: Bivalvia). Freshwater Biology 12:553-558.

Valderrama, J. C. 1981. The simultaneous analysis of total nitrogen and total
phosphorus in natural waters. Marine Chemistry 10: 109-122.

Walz, N. 1978. The energy balance of the freshwater mussel Dreissena polymorpha
Pallas in laboratory experiments and in Lake Constance III. Growth under standard
conditions. Archiv fiir Hydrobiologie Supplement 55:121-141.

Wetzel, R. G. and G. E. Likens. 1991. Limnological analyses, 2nd edition. Springer.

54

F “M iii

Young, B. L., D. K. Padilla, D. W. Schneider, and S. W. Hewett. 1996. The importance
of size-frequency relationships for predicting ecological impact of zebra mussel
populations. Hydrobiologia 332: 151-158.

Yu, N. and D. A. Culver. 2000. Can zebra mussels change stratification patterns in a
small reservoir? Hydrobiologia 431: 175-184.

Zdanowski, B. 1982. Variability of nitrogen and phosphorus contents and lake
eutrophication. Polskie Archiwum Hydrobiologii 29 (3-4):541-597.

55

IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII

lHIIILHIllllﬂlll(lillllllllllllllll