HTHS

THESIS

¢\\.
CT.“
(.3
,...

This is to certify that the

thesis entitled

The Effects of Macrophyte Structural Heterogeneity
and Fish Prey Availability on Age-0 Largemouth
Bass Foraging and Growth

presented by

Rahman David Valley

has been accepted towards fulfillment
of the requirements for

M.S. Fish. & Wildl.

degree in

 

 

 

Date September 11, 2000

 

0-7639 MS U is an Affirmative Action/Equal Opportunity Institution '

 

 

LIBRARY

Mlchlgan State
Unlverslty

 

 

 

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

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6/01 chlRC/DateDuepB5—p. 15

THE EFFECTS OF MACROPHYTE STRUCTURAL HETEROGENEITY AND FISH
PREY AVAILABILITY ON AGE-O LARGEMOUTH BASS FORAGING AND
GROWTH
By

Rahman David Valley

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

2000

ABSTRACT
THE EFFECTS OF MACROPHYTE STRUCTURAL HETEROGENEITY AND FISH
PREY AVAILABILITY ON AGE-0 LARGEMOUTH BASS FORAGING AND
GROWTH
By

Rahman David Valley

Age-0 largemouth bass (Micropterus salmoides; LMB), inhabiting vegetated lake areas,
initially consume invertebrates until growing sufficiently large to consume more
energetically-profitable fish (e.g., age-0 bluegill, Lepomis macrochirus; BG). Age-0
LMB success as piscivores may depend both on age-0 BG abundance and macrophyte
assemblages, which can affect BG vulnerability to LMB. The widespread exotic
macrophyte, Eurasian waterrnilfoil (Myriophyllum spicatum; EWM) typically increases
macrophyte abundance and simplifies structure relative to native plant assemblages.
Therefore, I hypothesized that: l) age—0 LMB enjoy higher foraging success where plants
are of moderate density (compared to dense), structurally heterogeneous (compared to
EWM monocultures), and BG prey are abundant, and 2) age-0 LMB growth is higher in
lakes dominated by native plants (compared to lakes dominated by EWM) and where
age-0 BG prey are abundant. In an age-0 LMB foraging behavior experiment, dense
plants and EWM monocultures significantly reduced LMB foraging success. In a multi-
lake field study, EWM coverage did not significantly affect age-0 LMB growth.
Macrophyte assemblages may have been sufficiently heterogeneous such that age-0 LMB
could capture BG prey in all lakes. Age-0 BG prey abundance positively correlated with
age-0 LMB growth. Overall, my data indicate that plant abundance and Structure

function in complex, scale-dependent ways to influence age-0 LMB foraging success.

ACKNOWLEDGMENTS
I owe a great deal of thanks to my major professor, Mary Bremigan for her guidance,
support, and especially her patience. Dr. Bremigan frequently made herself available to
address any of my concerns and to offer constructive criticism on drafts of proposals,
reports, and this thesis. Yet, at the same time, Dr. Bremigan granted me freedom to make
my own decisions and tailor my project according to my research interests. I would also
like to thank my committee, Patricia Soranno, Gary Mittelbach and John Madsen, for
their advice in developing my thesis project and providing constructive criticism on an
earlier draft of this thesis. I am also thankful to numerous graduate and undergraduate
students that provided field and laboratory assistance, especially T. Brunger, K.
Cheruvelil, C. Downing, S. Hanson, A. Mahan, V. Moore, K. Rogers, M. Sanbom, T.
Schuh, R. Serbin, and A. Wilson. The Army Corps of Engineers, especially J. Madsen,
C. Owens, and R. Stewart provided invaluable assistance with the macrophyte surveys. J.
Amberg and S. Hart taught me the difficult task of keeping larval fish alive in the lab.
Funding for this project was provided by grants from the Aquatic Ecosystem Restoration
Foundation, Aquatic Plant Management Society, MSU Agriculture Experiment Station,
MSU Extension, MSU Department of Fisheries and Wildlife, and scholarships from the

Minnesota Conservation Federation, and New Market Sportsman’s club.

iii

TABLE OF CONTENTS

LIST OF TABLES ................................................................................... v
LIST OF FIGURES ................................................................................. vi
INTRODUCTION ................................................................................... 1
EXPERIMENTAL METHODS ................................................................... 4
EXPERIMENTAL RESULTS ..................................................................... 8
FIELD METHODS ................................................................................ 12
FIELD RESULTS ................................................................................. 16
DISCUSSION ...................................................................................... 19
SUMMARY .................................. _ ....................................................... 26
APPENDIX ......................................................................................... 28
LITERATURE CITED ........................................................................... 42

LIST OF TABLES

Table 1. Characteristics of plant assemblages (Species composition and relative
abundance) simulated in each experimental treatment, total water volume displaced by
the plants, estimated total live plant biomass, and plastic plant components. EWM is
Eurasian watermilfoil.

Table 2. Limnological and macrophyte assemblage characteristics of 6 study lakes in
Michigan. Cover refers to the percentage of sampled points at which plants were present.
Total plant cover indicates the percent occurrence of plants sampled over the entire lake,
whereas littoral plant cover indicates the percent occurrence of plants sampled only in the
littoral zone. Littoral zone is defined as the area from shore to the depth at which plants
consistently occurred.

LIST OF FIGURES

Figure 1. Top-down view of the laboratory set-up for the age-O largemouth bass foraging
behavior experiment.

Figure 2. Geometric mean number of seconds (i 95% CI) elapsed before a largemouth
bass attacked (A), and consumed (B) its first bluegill prey in moderate and high plant
density treatments, Eurasian watermilfoil (EWM) monocultures and diverse plant
treatments, and low and high bluegill density treatments. Main effect p-values from a
three-way ANOVA are reported. No interactions were significant. Numbers above bars
denote the number of replicates.

Figure 3. Box plots of attack rate (attacks per minute of search) by largemouth bass on
bluegill in each treatment of the largemouth bass foraging experiment. Comparisons
shown are attack rates in moderate and high plant density treatments, Eurasian
watermilfoil (EWM) monoculture and diverse plant treatments, and low and high bluegill
density treatments. P-values from Wilcoxin rank-sums tests are reported. Boxes
represent the range in which 50% of the data points fall (i.e., interquartile range or IQR).
Wiskers represent the range of data that fall within 1.5 times the IQR. Diamonds
represent data points that are within 3 times the IQR. Circles represent outliers.
Numbers above plots denote the Wilcoxin mean rank score.

Figure 4. Mean number (i 1 SE) of bluegill (BG) consumed by age-0 largemouth bass in
the largemouth bass foraging experiment. Comparisons shown are total consumption in
moderate and high plant density treatments, Eurasian watermilfoil (EWM) monocultures
and diverse plant treatments, and low and high bluegill density treatments. Main effect p-
values from a three-way ANOVA are reported. No interactions were significant.
Numbers above bars denote the number of replicates.

Figure 5. Size distributions of age-0 largemouth bass collected in late September 1999
both from electrofishing (120 volts pulsed-DC) and Shoreline seining.

Figure 6. Relationship between Eurasian watermilfoil coverage and absolute growth
(standardized to growing degree days) of age-0 largemouth bass (LMB). Vertical and
horizontal bars are used for lakes were measured in both years and represent the range of
means in 1998 (solid line) and 1999 (dashed line); points in the middle represent the
mean effect and response for each lake measured in both years.

Figure 7. Catch per effort (CPE; No./ m2) of age-0 bluegill (determined by length
distributions) captured in seine hauls during summer 1998 and 1999. Black bars
represent vulnerable bluegill (individuals 5 40% of the mean largemouth bass TL). Open
bars are bluegill that were too large for age—O largemouth bass to consume. Dashed
vertical lines separate each sampling date and solid vertical lines separate months.
Values in the upper right of each lake’s distribution in 1998 and upper left in 1999
indicate the absolute growth of age-0 largemouth bass. Lakes are ordered by ascending
age-0 largemouth bass absolute growth rates in 1999. Note y-axis scales differ between

vi

1998 and 1999. All lakes with < 42% littoral coverage of EWM were designated as
“low” EWM lakes.

Figure 8. Relationship between mean catch per effort CPE of vulnerable age-O bluegill
[individuals S 40% of the mean length of age-0 largemouth bass (LMB) captured during
the same period] and absolute growth (standardized to growing degree days) of age-0
LMB. Vertical and horizontal bars are used for lakes measured in both years and
represent the range of means in 1998 (solid line) and 1999 (dashed line); points in the
middle represent the mean effect and response for each lake measured in both years. All
lakes with < 42% littoral coverage of EWM were designated as “low” EWM lakes.

Figure 9. Taxonomic diet composition of age-0 largemouth bass in 1998 and 1999.
Values reported in 1998 and 1999 are means (weighted by sample size for each size class
in each lake) of 5 and 6 lakes respectively (i 1 SE).

Figure 10. Absolute growth (standardized to growing degree days) of age—0 largemouth
bass (LMB) as a function of mean percent occurrence of piscivory (A) and mean percent
occurrence of piscivory as a function of bluegill prey availability (CPE of bluegill S 40%
of the mean length of age-0 LMB captured during the same period; B). Frequency of
piscivory was estimated from diet analysis as the proportion of LMB stomachs that
contained fish and reflects seasonal means. Vertical and horizontal bars are used for
lakes measured in both years and represent the range of means in 1998 (solid line) and
1999 (dashed line); points in the middle represent the mean effect and response for each
lake measured in both years. All lakes with < 42% littoral coverage of EWM were
designated as “low” EWM lakes.

Figure 11. The relationship between adult bluegill size and the catch per effort CPE of
bluegill < 25 mm TL during late August. Adult bluegill were captured during spring
1998 and 1999 electrofishing surveys. Vertical and horizontal bars are used for lakes
measured in both years and represent the range of means in 1998 (solid line) and 1999
(dashed line); points in the middle represent the mean effect and response for each lake
measured in both years. All lakes with < 42% littoral coverage of EWM were designated
as “low” EWM lakes.

vii

INTRODUCTION

Largemouth bass (Micropterus salmoides; LMB) is a keystone species of north
temperate lake food webs (Mittelbach et al. 1995, Power et al. 1996) and a popular
sportfish. Therefore, understanding factors underlying LMB recruitment should promote
effective management of both LMB populations and other features of lake ecosystems
influenced by LMB. Mechanisms affecting recruitment of LMB are often size-
dependent, such that rapid first year growth confers a survival advantage (see Garvey et
al. 1998b for review). Specifically, larger age-0 LMB often enjoy higher overwinter
survival than their smaller counterparts, due to greater energy reserves and reduced risk
to predation (Garvey et al. 1998b).

First year growth of LMB and thus size entering winter is influenced by factors
such as prey species composition and availability. Initially, age-0 LMB consume
zooplankton and macroinvertebrates until they attain the Size advantage needed to capture
age-0 fish prey [typically age-0 bluegill (Lepomis macrochirus; henceforth BG) in north
temperate lakes (Olson 1996) or age-0 shad (Dorosoma spp.) in reservoirs (Garvey and
Stein 1998)]. Because fish prey are more energetically profitable than invertebrate prey,
the shift to piscivory by age-0 LMB can be followed by rapid growth (Keast and Eadie
1985, Olson 1996, Garvey et al. 1998a). Together, the timing of the onset of piscivory,
and the extent to which fish prey contribute to age-0 LMB diets once they are
piscivorous, should determine in large part, age-0 LMB growth rates and recruitment.

Fish prey availability is often variable among systems and can affect the timing
and degree of piscivory of age-0 LMB. Potential factors influencing fish prey availability

include relative hatch dates of predator and prey (Miller and Stork 1984, Phillips et al.

1995), relative growth rates of predator and prey (Keast and Eadie 1985, Olson 1996),
and total prey production (Miller and Stork 1984, Phillips et al. 1995, Garvey and Stein
1998).

In addition to prey fish size and abundance, physical habitat features, such as
macrophyte structure, can affect LMB recruitment by mediating interactions between
age-0 LMB and their fish prey. Theory predicts that an intermediate level of macrophyte
abundance should produce maximum LMB growth (Crowder and Cooper 1979).
Numerous studies conducted on lakes and reservoirs (80-27,000 ha) have demonstrated
that macrophyte coverage exceeding 10% of a system’s total area promotes age-0 LMB
abundance, presumably because macrophytes provide cover from predators of age-0
LMB (Durocher et al. 1984, Bettoli et al. 1992, Maceina 1996, Wrenn et al. 1996,
Miranda and Pugh 1997). However, where total areal coverage of macrophytes exceeds
40-60%, low piscivory and/or poor growth by age-0 LMB commonly result (Colle and
Shireman 1980, Bettoli et al. 1992, Maceina 1996, Wrenn et al. 1996, Miranda and Pugh
1997), presumably because visual and swimming barriers created by dense vegetation
reduce the ability of LMB to forage successfully for fish prey (Glass 1971, Savino and
Stein 1982, Anderson 1984). However, even within the 10-60% plant coverage range,
the relationship between macrophyte coverage and LMB recruitment is highly variable
within and among systems, such that a consistent optimal range of macrophyte coverage
for LMB recruitment has not been identified. This relationship may vary, in part,
because coarse measures of total vegetation abundance (such as percent areal coverage)
do not represent the amount of critical edge (or patch) habitat that is consequential to

LMB success (Annett et al. 1996, Dibble et al. 1996, Miranda and Pugh 1997).

Documenting the effects of macrophyte structural heterogeneity on LMB
recruitment is important because invasive macrophyte species are now widespread
throughout North America and fundamentally alter the nature of habitat structure. For
instance, native macrophyte communities are structurally diverse; their patchy spatial
distributions and mosaic of basal, medial, and canopy growth forms create numerous
horizontal and vertical interstitial spaces (i.e., edge or habitat patches). Invasion by
exotic macrophytes such as Hydrilla (Hydrilla verticillata) and Eurasian watermilfoil
(Myriophyllum spicatum; henceforth EWM), reduce edge by forming extensive
homogeneous monocultures (Madsen 1997). In particular, EWM grows rapidly, typically
forming extensive homogeneous surface canopies that displace structurally-diverse native
macrophytes (Madsen et al. 1991, Nichols et al. 1992). Plant biomass in EWM beds is
partitioned towards the surface, thereby reducing sub~canopy light, oxygen, and pH
(Titus and Adams 1979, Carpenter and Lodge 1986, Madsen 1997). As a result of the
inhospitable foraging environment in the sub-canopy, trophic interactions may be
restricted to the canopy where vegetation is extremely dense and interstitial spaces are
few.

Despite the widespread occurrence of exotic macrophyte invasions, their effects
on LMB recruitment are poorly understood. Furthermore, no studies have
simultaneously evaluated the effects of macrophyte density, structure, and prey
availability on age-0 LMB foraging and growth. Accordingly, I evaluated how
macrophyte density, structure, and fish prey availability, affect age-0 LMB foraging
success and growth at both small (aquaria) and large (multiple whole-lakes) scales. The

small-scale experiment evaluated the relative effects of macrophyte density, structure,

and age-0 BG prey abundance on age-0 LMB foraging success. I hypothesized that age-0
LMB Spend less time searching for prey and enjoy higher foraging success (i.e., higher
attack and consumption rates) where plants are of moderate density (compared to high
density), plants are structurally diverse (compared to EWM monocultures), and age-0 BG
prey are abundant (compared to sparse). Furthermore, I expected high BG density to
have a relatively small positive effect on LMB foraging success in high density plant
treatments and EWM monoculture treatments, due to the inability of the LMB to detect
and/or capture BG prey. In contrast, I expected BG density to have large effects on LMB
foraging success in moderate density plant treatments and diverse structure treatments,
due to the greater ability of the LMB to detect and/or capture prey. The multi-lake study
assessed the generality of my experimental findings by evaluating age-0 LMB piscivory
and growth along gradients of EWM coverage and age-0 BG prey abundance. I
hypothesized that age-0 LMB piscivory and growth decline along a gradient of increasing
EWM coverage. I also expected high age-0 BG prey abundance to have a relatively
small positive effect on age—0 LMB piscivory and growth where EWM coverage was
high. In contrast, I expected age-0 BG prey abundance to have large effects where EWM

coverage was low.

EXPERIMENTAL METHODS
Treatments and Replication—I used a 2x2x2 factorial design crossing EWM versus
diverse plants, high versus moderate plant density, and high versus low BG density in 8
114-L rectangular glass aquaria in the laboratory. A preliminary power analysis

indicated that at least 13 replicates would be needed to detect a large effect (i.e., 40%

difference between treatments, [3 = 0.20) in any particular measured response. Therefore,
every other day for 26 days, I conducted a complete set of replicates (i.e., l replicate per
each of 8 treatments per trial day). Across trial days, each aquarium received at least 1
replicate of each treatment in random order. No more than 2 replicates of a given
treatment were conducted in the same aquarium. Each aquarium contained one age-0
LMB foraging for age-0 BG (see “Fish” below).

Experimental units—Black construction paper was placed on three sides of each
aquarium’s walls to confine the LMB’S field of view to within its experimental unit; the
front wall remained open for observation (Figure 1). Small sponge-filters in each of the
aquaria maintained both water quality and dissolved 02 (approximately 7.5 mg/L).
Although light intensity at the water’s surface ranged from 14.1 to 25.1 p mol/s m2
among aquaria, variability in light intensity did not bias results (see “Experimental
Results”). Temperature remained constant at 22 °C.

Plant characteristics—I established plant treatments that differed according to
structure and abundance by using plastic aquatic plants (manufactured by Second Nature
Plantastics Plus, Oakland, NJ). Plastic foxtail plants that were 45.7 cm in length were
modified to resemble EWM. Plants used in diverse native treatments resembled native
plants commonly found in lakes, and differed in number and length (Table 1). I
measured the volume of water displaced by each plant in order to hold plant volume
constant in the EWM and diverse treatments (Table 1). To relate abundances of plants
used in my experimental treatments to the field, I measured water volume displaced and
dry weight of the real plants (dried at 105 °C for 48 h) that the plastic plants resembled.

From these data, I estimated the “live biomass” equivalent of the plastic plants in each

treatment (Table 1). Estimated live biomasses represented by the plastic plants in
experimental treatments were within the range of typical field values (Grace and Wetzel
1978). Eurasian watermilfoil plants were arranged uniformly on a 76 by 32 cm (5 mm
mesh) metal grid. Plants in the diverse treatments were randomly assigned to uniform
positions on a similar grid. The mosaic of vertical interstitial spaces present in native
macrophyte beds in lakes was simulated in the aquaria by using plants of varying lengths.
However, due to the small size of the aquaria, I could not Simulate horizontal interstices
(i.e., vegetation patches) that occur at spatial scales larger than those represented by my
experimental aquaria.

F ish—Age-O LMB and age-0 BG were collected from Big Crooked and Camp
lakes (Kent Co., MI) and maintained in the laboratory. Largemouth bass were fed small
Shiners (Notropis spp.) until they were approximately 75-90 mm total length (TL). Once
LMB reached the appropriate size, each was transferred to separate experimental aquaria.
Each LMB remained in the same aquarium throughout the 26 (1 duration of the
experiment. If a LMB became diseased or chronically inactive (two starved LMB never
responded to the presence of BG prey), it was replaced by a LMB of similar Size from the
laboratory stock. For each trial, four BG were randomly assigned to each high BG
density treatment and two BG were randomly assigned to each low BG density treatment.
Bluegill densities were similar to prey densities used in other LMB foraging experiments
(Savino and Stein 1982, Anderson 1984, Schramm and Zale 1985, McMahon and
Holanov 1995). Because of the fragility of small age-0 BG, I obtained an estimate of BG
size by measuring the TL of a subsample of BG (n = 10 per trial day) chosen randomly

throughout the experimental set up.

Procedure—Forty hours before a trial, the appropriate plant structure was placed
into each aquarium. At the same time, a clear plexiglas divider was placed into the
middle of the aquarium, confining the LMB to one side of the aquarium. On each trial
morning, BG were selected from the stock and were either sacrificed and measured or
placed into one of eight isolation chambers that were numbered according to the
aquarium into which each chamber would be placed. Chambers were placed on the side
of the aquarium without the LMB but directly adjacent to the divider such that both LMB
and BG were visually aware of each other’s presence. After 10 min, the BG in the first
aquarium to be observed were released from the isolation chamber into the aquarium side
opposite from the LMB. After 20 seconds, the divider was lifted and the trial began.
Aquaria were observed sequentially; therefore, time in the isolation chambers ranged
from 10 to 120 minutes. To account for differences in acclimation periods, I alternated
the order in which aquaria were observed each trial day.

Observational methods were similar to those described by Anderson (1984).
Briefly summarized, from behind a blind, I recorded LMB position in the aquarium and
foraging behavior on a voice tape recorder, noting the time spent by the LMB searching
for prey and the success or failure of each capture attempt. When possible, the positions
of the BG were also noted. At the end of each trial day, the tape was replayed and a
stopwatch was used to determine when attacks occurred and the success of each attack.
The amount of time a LMB remained motionless (i.e., at the very beginning of the trial or
when handling a prey; typically < 10 s) was subtracted from total search times. Each trial

was terminated after 10 minutes, or earlier, if all BG were captured.

Statistical Analyses—Data were analyzed using SAS version 6.12 (SAS Institute
Inc. 1990). I quantified LMB foraging success by evaluating the effects of plant density,
plant structure, and BG density on the following LMB response variables: 1) time to first
attack, 2) attack rate (number of attacks per min of search), 3) capture success
(capturezattack ratio), 4) time to first consumption, and 5) total consumption (i.e., total
number of BG consumed). A three-way AN OVA (type IH sums of squares) was used to
evaluate all response variables except attack rate. Due to the extreme skew and
heterogeneous variances among treatments associated with attack rate (O’Brien test for
unequal variances p = 0.05), I evaluated the effect of each main factor (i.e., plant density,
plant structure, and bluegill density) on attack rate with 3 one-way nonparametric
Wilcoxin rank-sum tests. Although I could not statistically conclude the absence of
significant interactions between factors on attack rate, ANOVA results from the other
four response variables suggest interactions were not common in the experiment (see
“Experimental Results”). Capture success values were arcsine square root (x)
transformed and times elapsed prior to the first attack and consumption were loge (x)
transformed to normalize distributions and homogenize variances. If interaction terms

were not at least marginally significant (p < 0.10), they were dropped from each model.

EXPERIMENTAL RESULTS
Because of the unanticipated loss of replicates (i.e., diseased or chronically
inactive LMB), I completed 9-11 replicates per treatment. Largemouth bass grew
approximately 1 mm per week, reaching 83-92 mm TL by experiment’s end. Mean TL

of BG (computed each trial day) ranged from 20 to 27 mm TL (coefficient of variation

ranged from 3 to 17). On average, BG lengths were 29% of IMB TL during the first trial
and 27% during last trial (mean across all trials = 30%). No effects of date or
aquarium/LMB/light intensity were detected for any of the measured variables (two-way
AN OVA p > 0.10); therefore, all trials were considered independent.

Behavioral Observations—Largemouth bass searched for prey by swimming in
short quick movements, frequently changing direction and eye position. In the diverse
plant treatments, LMB often used the entire depth of the aquarium to search for prey. In
contrast, in the EWM treatments, LMB stayed below the canopy (i.e., lower half of the
aquarium) to search for prey. Vegetation in the EWM treatments often obstructed the
mobility of the LMB when it attempted to swim through the canopy and attack a BG. An
attack was characterized by a LMB stopping its search abruptly, slowly stalking the
detected prey, and finally making an accelerated movement towards the prey. Each
accelerated movement was considered one attack. All BG that were detected by a LMB
were attacked. Largemouth bass resumed aggressive search behavior shortly after
ingesting a prey. Bluegill response to predation threat varied within treatments. Most
BG observed (80%; evenly distributed among treatments) remained motionless in
vegetation as the LMB searched. However, some (20%) BG did not attempt to hide from
the LMB (i.e., “naive” BG) and swam into open water (i.e., open spaces in the diverse
treatment or below the canopy in the EWM monoculture treatment). All BG that swam
below the canopy in EWM treatments were quickly detected and attacked by the LMB.
However, naive bluegill in diverse treatments were not detected as quickly as in EWM
treatments, because vegetation patches in the diverse treatments often obstructed the

LMB’s view of a BG in open water.

Time to first attack—Two replicates were omitted from this analysis because no
BG were attacked. Both trials with no attacks were the EWM monoculture, high plant
density, and low BG density treatment. Time elapsed prior to first attack varied
Si gnificantly among treatments (AN OVA, p = 0.04). As expected, time elapsed prior to
first attack was greater in high plant density than moderate plant density treatments
(ANOVA, main effect p = 0.02; Figure 2A). Time elapsed prior to first attack was
marginally higher in low BG than high BG density treatments (ANOVA, main effect p =
0.09; Figure 2A). Contrary to my expectations, time elapsed prior to first attack did not
vary significantly with plant structure (ANOVA, p = 0.49; Figure 2A). Furthermore,
ANOVA did not detect any significant interactions (p > 0.10), thereby failing to support
my hypothesis that BG density would have an effect on search times in moderate plant
density treatments and diverse plant treatments but no effect in high plant density
treatments and EWM monoculture treatments.

Attack Rates—Wilcoxin rank-sum tests indicated significant effects of plant
structure (p = 0.005) and plant density (p = 0.01). As expected, LMB exhibited higher
attack rates in treatments with moderate density plants and diverse plants (Figure 3).
Surprisingly, BG density did not have a significant effect on attack rates by LMB (p =
0.25; Figure 3).

Capture Success-Capture success (capturezattack ratio) averaged 34.6il.6% and
did not vary among treatments (ANOVA; p = 0.98). In other words, each BG that was

attacked had equal probability of capture, regardless of treatment.

10

Time to first consumption-Fifteen replicates were omitted from this analysis
because no BG were consumed. Largemouth bass consumed zero BG in 11 of the 45
EWM monoculture replicates and only 4 of the 38 diverse plant replicates. The time to
first consumption varied significantly among treatments (AN OVA, p = 0.006). As
expected, time elapsed prior to first consumption was greater in high plant density than in
moderate plant density treatments (AN OVA, main effect p = 0.004; Figure ZB) and
greater (although only of marginal statistical significance) in low BG than high BG
density treatments (AN OVA, main effect p = 0.06; Figure 2B). Contrary to my
expectations, time elapsed prior to first consumption did not vary significantly with plant
structure (ANOVA, p = 0.36; Figure 2B). Contrary to my expectations, no interactions
occurred (p > 0.12).

Total Consumption-The total number of BG eaten during the trials differed
among treatments (AN OVA, p < 0.001) and ranged from 0 to 4 (the maximum number of
BC in the high BG density treatments). In the high BG density treatments, LMB
consumed all four BG in 9 of the 42 replicates. In the low BG density treatments, LMB
consumed both BG in 13 of the 41 replicates. Analysis of variance detected significant
main effects of plant density (p = 0.02) and BG density (p < 0.001) and a nearly
significant effect of plant structure (p = 0.06). AS expected, LMB consumed more BG in
moderate plant density, diverse plant, and high BG density treatments (Figure 4). Again,

contrary to expectations, no interactions were Significant (p > 0.19).

11

FIELD METHODS

Study lakes-To assess the generality of my experimental results, I sought to
evaluate the effects of plant density, plant structure, and age-0 BG prey availability on
age-0 LMB diet and growth through a multi-lake field study. Age-0 LMB were sampled
in multiple (five in 1998 and six in 1999) mesotrophic lakes in southern Michigan, each
with a history of EWM infestation. Four lakes (Bass-1999, Big Crooked, Camp, Lobdell)
were treated with herbicides (including 5-7 ppb fluridone) and three lakes (Bass-1998,
Big Seven, and Heron) received no macrophyte management and remained infested with
EWM, thereby establishing a gradient of EWM coverage among lakes (Table 2).

Macrophyte surveys-Percent cover of macrophytes during August 1998 and 1999
was determined using the point-intercept method (Madsen 1999). Briefly summarized,
this entailed overlaying a grid of 150-250 uniformly spaced points (~ 50-100 m apart)
onto each lake map. Using a global positioning system, points were located, depth was
recorded, and Species presence/absence was determined using visual inspection and/or a
double-headed rake thrown from the boat. The outer edge of the littoral zone was defined
as the greatest depth at which macrophytes typically occurred in each lake. Plant
coverage within the littoral zone was used to evaluate LMB responses, even though many
prior studies have used measures of total areal coverage of plants to evaluate LMB
responses. Gradients of plant/EWM coverage within the littoral zone are likely to affect
LMB and BG interactions and measures of total areal coverage may not reflect
differences in littoral zone structure. For example, relatively deep lakes have relatively

small littoral zones; therefore measures of total plant coverage will be inherently lower in

12

deep lakes compared to shallow lakes, even if plants are extremely dense within the
littoral zone of deep lakes.

Largemouth bass and bluegill sampling—During 1998, age-0 LMB and age-O BG
(age estimated from length distributions) were sampled in each of 5 lakes bi-weekly from
late June (except for Lobdell L. in 1998 where LMB were not sampled until 21, July)
through late August using both a 10 by 1.8-m (4-mm mesh) bag seine (3 sites per lake; 20
to 30 m length pulled parallel to shore; 0 to 1.5-m depth) and a 18.3 by 2.4-m (4-mm
mesh) purse seine [3 sites per lake (same general area as bag seine sites); 1 to 2 m depth].
Each site was vegetated with submerged and/or floating leaf macrophytes and Sites were
evenly dispersed around each lake. Because the size of age-0 LMB collected by the 2
gears did not differ (sequential Bonferroni adjusted t-tests p > 0.05; 12 of 15
comparisons), I used largemouth bass from both gears to evaluate growth. However,
estimates of relative abundance were generated for LMB captured in bag seines only
because catch per effort (CPE; No. per m2) of age—0 fish captured with the purse seine did
not correlate with CPE of the bag seine (r = 0.30) and because the purse seine caught
relatively few age-0 LMB. In 1999, I sampled age-0 LMB and age-0 BG solely with the
bag seine (6 sites per lake) on a monthly basis, from late June to late August. All age-0
LMB and age-0 BG captured each summer were preserved in 95 % ethanol. Age-0 LMB
were also sampled in late September 1999 using both the bag seine and electrofishing
(120 volts pulsed-DC). However, size distributions generated from both gears indicated

size-selectivity by both gears (Figure 5).

13

Largemouth bass growth—All LMB preserved (n = 1230) were later measured to
the nearest millimeter TL and counted. A subset of LMB (n = 390; randomly pooled
from all lakes) was measured, dried at 60°C for 72 hours, and weighed to the nearest
milligram to develop a length-dry weight regression. To draw inferences regarding age-0
LMB growth across lakes and years, I computed an absolute growth rate using the mean
dry mass of age-0 LMB captured in each lake during the first and last sampling period
(typically late June through late August) in each year, divided by the number of growing

degree days (gdd) that elapsed between these sampling dates in each year such that:

g(final) — g(1n1t1al) where
gdd

 

absolute growth =

gdd = cumulative gdd(final) — cumulative gdd(initial) and

 

dd _ [ maximum temperature — minimum temperature ] 50
' 2

Growing degree day data were obtained from National Weather Service lSt order climate
stations closest to the study lakes (Grand Rapids for Bass, Big Crooked, and Camp lakes
and Detroit for Big Seven, Lobdell, and Heron lakes).

Largemouth bass diets-Gut contents from LMB were analyzed in a stratified
manner such that all sizes of age-0 LMB captured in each particular lake on each
sampling date were represented equally. A maximum of five LMB stomachs from each
of three size classes (maximum TL — minimum TL / 3; calculated for each lake and each
date) was analyzed per lake and date. Where possible, equal numbers of LMB were
analyzed from each seine Site within each lake. Stomach contents were identified, where

possible, to family for macroinvertebrates and copepod zooplankton, and to genus for

14

cladoceran zooplankton, and measured using a microscope equipped with a drawing tube
and digitizing tablet. Biomass of stomach contents was estimated by recording the
appropriate body dimensions (e. g., head width or body length) and using length-dry mass
regressions that corresponded to the body dimension measured (G.G. Mittelbach,
Michigan State University, Kellogg Biological Station, unpublished data, R.D. Valley
unpublished data). For the fish consumed that could be measured, I regressed the
standard length (SL) of fish prey against LMB TL in order to estimate the SL and
corresponding mass of prey fish ingested by other LMB that were too digested to be
measured. Regressions were developed for both 1998 and 1999 even though 95%
confidence intervals of both regression slopes overlapped (see “Field Results”). I
characterized the degree of piscivory by first calculating the percent of LMB stomachs in
each of the three Size classes that contained at least one fish. Next, using these
percentage values and the abundance of LMB in each size class, I calculated a weighted
mean percent of LMB that had consumed at least one fish across the three Size classes for
each lake on each date. Finally, I averaged the daily mean value across sampling dates
within each lake to characterize piscivory on a seasonal basis.

Prey availability—Because age-0 BG are the predominant fish prey of age-0 LMB
in Michigan lakes (Olson 1996, this study; see “Field Results”), patterns of age-0 LMB
diets were related to availability of vulnerable BG. All age-0 bluegill captured in the
seines were preserved in 95% ethanol and later measured. Mean CPE (bag seine only) of
vulnerable BG (defined as individuals S 40% of the mean LMB TL; Lawrence 1958) was
calculated for each lake on each date and averaged across sampling dates within each

lake to characterize prey availability on a seasonal basis. Because age-0 BG were not

15

inshore by the first sampling period in 1999 in any of the lakes, this date was excluded
from the analysis of relative density of BG prey.

Statistical analyses—Data were analyzed using SAS version 6.12 (SAS Institute
Inc. 1990). Regression analysis was used to evaluate relationships between response
variables and effect variables. Seasonal means for some response variables (e.g., percent
piscivory and prey availability) for each lake were computed for 1998 and 1999 and
values from lakes that were measured in both years were averaged across years. An
exception was Bass L. where 1998 and 1999 were treated as independent in my analyses
because this lake was not managed for EWM in 1998 but was treated with herbicides in
1999 (Table 2). In all, seven data points were used in the regression analyses: one point
for each lake except Bass L., for which two were used. Variable distributions that were

highly skewed were logc (x) transformed. Rejection criterion was set at CI = 0.05.

FIELD RESULTS

Macrophyte surveys-Fluridone among other herbicides greatly reduced EWM
cover in managed lakes while not reducing native plant cover or plant diversity (Madsen
et al. in press). Total lake coverage of macrophytes (i.e., percent of sampled points with
plants) varied from 37 to 84% but was similar between years (Table 2). Although not
statistically significant, total plant coverage tended to decrease with increasing mean
depth (regression, r2 = 0.35 p = 0.16; Table 2). Total littoral coverage of macrophytes
was similar among lakes and years, ranging from 75 to 99% (mean = 88%; Table 2).
Therefore, I did not evaluate whether plant coverage had an effect on LMB piscivory and

growth. Despite the large range in EWM littoral coverage across lakes, native

l6

macrophytes were common in all lakes (ranging from 62-99% littoral coverage; Table 2).
Furthermore, extensive homogeneous beds of EWM did not typify any of the study lakes.
Rather, EWM beds were often patchy with native plants interspersed between EWM
patches (R.D. Valley personal observation).

Effects of Eurasian watermilfoil and bluegill prey availability on age-0
largemouth bass growth—Densities of age-0 LMB captured in bag seines ranged from
0018-0037 LMB per m2 (mean = 0.027) in 1998 and 0007-0028 LMB per m2 (mean =
0.019) in 1999 (Table 2). Because relative abundance of age-0 LMB did not differ
greatly among my lakes (Table 2), I did not evaluate the relationship between age-0 LMB
relative density and EWM coverage or prey density. Age-0 LMB size in late August was
very similar between years [ranging in 1998 from 55 to 73 mm TL (mean = 62 mm TL);
ranging in 1999 from 51 to 76 mm TL (mean 63 mm TL); Table 2]. Size of LMB
collected with the bag seine in late September 1999 strongly correlated with LMB size in
late August (regression, r2 = 0.95, p = 0.001; mean size in late September 1999 = 66 mm
TL). Size of LMB collected by electrofishing in late September 1999 also correlated
with LMB size in late August but displayed a weaker relationship (regression, r2 = 0.79, p
= 0.04; mean size in late September 1999 = 74 mm TL; see Figure 5 for LMB
distribution comparisons for both gears).

Surprisingly, EWM coverage could not explain variation in age-0 LMB absolute
growth (regression, p = 0.62; Figure 6). Both total age-0 BG density and the density of
vulnerable age-0 BG varied considerably across lakes in 1998 and 1999 (Figure 7).
Indeed, mean CPE of vulnerable BG explained a considerable amount of variation in

LMB absolute growth rate (regression, r2 = 0.61, p = 0.04; Figure 8). However, contrary

l7

to my expectations, EWM coverage could not explain variation in the residuals from the
above regression (p = 0.88). In general, several size classes of age-0 BG were vulnerable
in lakes where age-0 LMB grew relatively rapid. In contrast, age-0 LMB generally grew
slowly where age-0 BG were too large or numerically few.

Largemouth bass diets—Of the 627 age-0 LMB diets analyzed, 113 of them
contained fish prey. Of the 65 fish prey that could be identified, 38 (59%) were BG.
Most of the other fish consumed were Shiners. Largemouth bass initially consumed
zooplankton and macroinvertebrates (mostly chironomids, and ephemeropteran and
zygopteran nymphs), and then shifted to a diet of mostly fish at approximately 60-80 mm
TL (Figure 9). Fish constituted the majority of prey biomass in LMB diets at sizes > 60
mm TL and occurrence of piscivory averaged 27% for LMB greater than this size.
Fishless stomachs of LMB > 60 mm TL were typically empty, indicating that these LMB
were specifically targeting fish as their prey. The size of fish consumed by age-0 LMB in
1998 and 1999 could be predicted by the following regression equations, respectively:
fish prey SL = 2.52 + 0.24*LMB TL (r2 = 0.52; p < 0.001); fish SL = -1.27 + 0.25*LMB
TL (r2 = 0.47 p < 0.001 ). Largemouth bass consumed fish prey that were on average
31% of their length.

As expected, piscivory was important for rapid growth of age-0 LMB. Mean
percent occurrence of piscivory positively correlated with LMB absolute growth rate
(regression, r2 = 0.89, p = 0.002; Figure 10A). Largemouth bass exhibited relatively high
growth in lakes where fish were relatively common in their diet. Furthermore, density of

vulnerable BG prey positively correlated with mean percent occurrence of piscivory

l8

(regression, r2 = 0.70, p = 0.02; Figure 10B). Again, EWM coverage could not explain

these residuals (regression, p = 0.46).

DISCUSSION

Integrating results from studies at multiple scales can lead to holistic
understanding of ecosystem structure and function (Frost et al. 1988, Power et al. 1996,
Stein et al. 1996). Accordingly, I combined both experimental and comparative
approaches to evaluate the relative effects of macrophyte abundance, macrophyte
structure, and fish prey availability on age-0 LMB foraging and growth. Despite the
potential importance of these three factors, no studies have explored their relative effects
on LMB recruitment. By mechanistically evaluating how habitat structure and prey
availability affect age-0 LMB foraging success, my experiment provided a predictive
framework for my field comparison. In turn, my field comparison explored the
generality of my experimental results and identified patterns at the scale most relevant to
managers. Below, I compare results both from the experiment and field study to evaluate
the relative effects of macrophyte density, macrophyte structure, and fish prey
availability on age-0 LMB foraging and growth.

Efiect of macrophyte density on age-0 LMB foraging and growth-Dense
macrophytes can reduce the ability of LMB to forage successfully for fish. Similar to
previous experimental studies with juvenile and adult LMB (Glass 1971, Savino and
Stein 1982, Anderson 1984, Gotceitas and Colgan 1987, Hayse and Wissing 1996), in my
experiment, dense macrophytes lengthened LMB search times, and reduced their attack

and consumption rates. Dense vegetation in my experiment reduced interstitial Spaces

19

and created numerous visual and swimming barriers to the LMB. As a result, prey were
not often encountered, and thus not often attacked or consumed in the high plant density
treatment. Furthermore, Similar to other studies (Glass 1971, Savino and Stein 1982,
Anderson 1984), I did not detect effects of plant density on prey capture success by LMB
(but see Gotceitas and Colgan 1987). Rather, foraging success by age-0 LMB depended
primarily on the ability of LMB to locate prey. In my field study, I could not adequately
determine whether age-0 LMB piscivory and growth were negatively related to dense
cover of littoral macrophytes, because littoral macrophyte coverage was consistently high
among my lakes (ranging 74-99% cover). In addition, no prior field studies have
evaluated the effect of littoral plant coverage on LMB recruitment. However, previous
field studies have demonstrated lower piscivory and/or growth by age-0 LMB where
macrophytes (mostly exotics; e.g., EWM and Hydrilla) are relatively dense (i.e., > 40-
60% total areal coverage; Colle and Shireman 1980, Bettoli et al. 1992, Miranda and
Pugh 1997, Maceina 1996). Nevertheless, measures of total areal coverage by previous
investigators may or may not have reflected differences in littoral macrophyte coverage
within/among their system(s). Therefore, large-scale effects of plant density on LMB
recruitment remain unclear.

Efifect of macrophyte structure on age-0 LMB foraging and growth—In addition to
high macrophyte density, homogeneous macrophyte structure also may reduce LMB
foraging success and growth. However, the effects of macrophyte structure on LMB
have been less well studied (Dibble et al. 1996). In my experiment, both attack and
consumption rates were lower in EWM monoculture treatments compared to diverse,

native plant treatments. Age-0 LMB could more easily access BG prey in diverse, native

20

plant treatments compared to EWM monoculture treatments, due to relatively numerous
interstitial Spaces in the native plant treatment. Although search times did not
significantly differ between EWM and native treatments in my experiment, my
experimental set-up did not mimic the harsh environment (i.e., low light, 02, and pH) of
EWM sub-canopies that has been documented in infested lakes (Titus and Adams 1979,
Carpenter and Lodge 1986, Madsen 1997). I hypothesize that in EWM-infested lakes,
age-0 LMB are often forced to forage within EWM canopies (rather than below, as LMB
did in my experiment) where interstitial spaces are few, and thus experience higher prey
search costs than in diverse plant assemblages.

In the field study, given the large gradient of EWM cover among the lakes (8-
92%), I expected LMB to eat fish less frequently and grow more slowly in lakes with
high EWM coverage. However, age-0 LMB piscivory and growth did not significantly
correlate with EWM coverage. The lack of a negative effect of high EWM coverage on
age-0 LMB piscivory and growth may be related to the patchy distribution of EWM in
reference lakes as opposed to the homogeneous monocultures that my experiment
simulated. AS a result of patchy EWM growth and the high frequency of native plants in
the reference lakes, moderate degrees of structural heterogeneity (and thus favorable
foraging environments for age-O LMB) were likely present among all study lakes.
Eurasian watermilfoil has been present in many lower MI lakes since the early 1960’s
(Coffey and McNabb 1974). Apparently, EWM’s aggressive behavior declines through
time, on the order of 10 to 15 years (Carpenter 1980, Smith and Barko 1990, Engel
1995). Perhaps this explains why I did not see extensive homogeneous monocultures of

EWM, and a corresponding negative effect on age-0 LMB.

21

Other studies support my experimental results by demonstrating high foraging
success by multiple age classes of LMB in habitats with numerous interstitial spaces
(Engel 1987, Smith 1993, Hayse and Wissing 1996). Patchy, heterogeneous macrophytes
may maintain a balance between LMB and BG populations by providing refuge for fewer
small BG than is provided by large homogeneous areas of vegetation. When refuge
patches become overly crowded, juvenile (including age-0) BG are forced to search for
new refuge areas, and thus become vulnerable to predation by LMB (Hayse and Wissing
1996). AS a result, both LMB and BG may experience high growth rates in patchy,
heterogeneous macrophyte assemblages. In contrast, poor LMB and BG Size structure is
common in lakes dominated by monocultures of invasive macrophyte species such as
EWM (Engel 1995, Olson et al. 1998, Unmuth et al. 1999).

Despite the potential importance of macrophyte heterogeneity for LMB
production, both conceptual and quantitative models that predict LMB production as a
function of macrophyte structure do not exist and require further exploration at both
whole-system and experimental scales. Coarse measures of total macrophyte coverage
that are commonly used may not accurately measure structural heterogeneity and may
inadequately characterize habitat perceived by LMB (Annett et al. 1996, Dibble et al.
1996, Miranda and Pugh 1997). Alternatively, explicit measures of structure (i.e.,
number, size, surface area, or spatial arrangement of vegetation patches, or macrophyte
species composition and distribution) may provide greater insight into the relationship

between habitat structure and LMB recruitment (Dibble et al. 1996, Weaver et al. 1997).

22

Efiect offish prey availability on age-0 LMB foraging and growth—In addition to
macrophyte effects on LMB population dynamics, factors that influence prey fish size
and abundance may affect age-0 LMB foraging and growth. However, surprisingly, BG
density did not have a large effect on LMB foraging success in my experiment. The
small Size of the aquaria may have limited my ability to create contrasts in BG prey
availability that may be present in lakes. Indeed, CPE of vulnerable age-0 BG was highly
variable in the study lakes and, as expected, BG prey availability and piscivory by age-0
LMB had a positive effect on LMB growth in the field study. Age-0 LMB grew
relatively Slowly if fish were not present in their diets. Olson (1996) also found a high
correlation between age-0 LMB piscivory and growth in Michigan lakes. However,
compared to my lakes, frequency of piscivory by age-0 LMB was generally higher in
Olson’s lakes and LMB switched to piscivory at a smaller size (50 mm TL compared to
60 mm TL in my lakes in 1998 and 70 mm TL in 1999), suggesting length at diet shift to
fish by age-0 LMB varies among lakes and over time.

Catch per effort of vulnerable BG was highly variable among the study lakes.
Factors that could contribute to high variability in CPE of vulnerable BG include those
that influence age-0 BG size: 1) early spawning of LMB (relative to BG), 2) Slow growth
of age-0 BG, 3) fast growth of age-0 LMB, and those that also influence age-0 BG
abundance: 4) long spawning duration of BG (affects both abundance and size), and 5)
high total reproductive output by adult BG. Although I could not rigorously evaluate the
relative contribution of these factors to the observed variability in the CPE of vulnerable

BG, factors 3, 4, and 5 appear especially important.

23

Age-0 LMB growth: Although cause and effect are unclear, my data suggest age-
0 LMB growth positively influenced BG vulnerability. Similar to results reported by
Olson (1996), as fish became increasingly common in their diet, age-0 LMB grew more
rapidly and additional size classes of BG became vulnerable (e.g., Big Crooked and
Camp lakes). In contrast, piscivory and growth by age-0 LMB was relatively low in
Heron L. despite a relatively high CPE of small (although not vulnerable) age-0 bluegill
(i.e., 25-30 mm TL) during both 1998 and 1999. Competition for macroinvertebrate prey
with the relatively abundant juvenile BG (40-120 mm TL) that were present in Heron L.
(R.D. Valley and MT. Bremigan unpublished data) may explain why age-0 LMB were
unable to gain the necessary Size advantage to consume age-0 BG and thus grow rapidly
Heron L. (Olson et al. 1995).

Bluegill spawning duration: Variability in BG spawning duration among my study
lakes may also explain variability in the production of vulnerable BG. The appearance of
small age-0 bluegill late in the season (late August) in some lakes (e.g., Big Crooked,
Camp, and Heron lakes in 1998 and 1999) was indicative of a protracted spawn. Small
age-0 BG were not present in seine hauls during late summer 1998 and 1999 in Bass L.,
where age-0 LMB growth was poor in both years, indicating BG spawned over a
relatively short time period in this lake.

Reproductive output by adult BG: My data also suggest that differences in total
reproductive output by adult bluegill among lakes may explain variability in age-0 BG
prey availability. This variability in age-0 BG CPE did not appear driven by predation
intensity by juvenile and adult LMB because predator density was highest in lakes where

age-0 BG abundance was highest (e.g., Big Crooked, Camp, and Heron lakes; SM.

24

Hanson and MT Bremigan unpublished data). Rather, adult BG size-structure may play
a large role in determining both BG spawning duration and total reproductive output.
Generally speaking, parental male BG provide care for eggs deposited by one or more
Size-selective females (Gross 1980, 1982). Brood size and nest success positively relate
to the amount of farming and guarding care parental males provide (Bain and Helfrich
1983, Coleman et al. 1985, Claussen 1991). In populations dominated by relatively small
BG (e.g., BG < 150 mm TL), relatively few BG fry may be produced over a short period
of time because energetic constraints restrict the amount and duration of parental care put
forth by small males (Gross 1980, Coleman and Fischer 1991, Wiegman and Baylis
1995). Furthermore, small female BG may produce fewer eggs over a relatively short
period of time compared to their larger counterparts (Gross 1980, Coleman et al. 1985).
Small clutches of eggs deposited in male-occupied nests increase the probability that
males will abandon their nests (Claussen 1991, Coleman and Fischer 1991). In fact,
Gross (1980) demonstrated the importance of adult body size for reproduction by
observing a 50% reduction in BG fry production as a result of only a 6% reduction in
adult body size. Therefore, BG populations dominated by relatively small individuals
may produce relatively few BG fry due to their short spawning seasons. In contrast, the
protracted spawning seasons and high BG fry production that appear characteristic of BG
populations with larger individuals may positively influence age-0 BG prey availability
to age-0 LMB.

I explored (post hoc) whether adult BG size could indeed explain variation in the
abundance of late-hatched (i.e., vulnerable) BG in the study lakes. Data from Spring

electrofishing surveys conducted in the study lakes during Spring 1998 and 1999 were

25

used to explore this hypothesis. Adult BG body size was characterized as the 90th
percentile total length. All BG < 25 mm TL in late August were assumed to belong to a
late-hatched cohort because length of the first hatched cohort of BC in late August is
typically 30-40 mm in north temperate lakes (Breck 1993, Cargnelli and Gross 1996).
Indeed, adult BG body Size positively correlated with the CPE of late hatched BG (r2 =
0.61 p = 0.04; Figure 11), supporting the hypothesis that adult BG size structure affects
age-0 BG prey production.

The relative effects of age-0 LMB growth, protracted BG spawning seasons, and
BG reproductive output on CPE of vulnerable BG may vary among lakes. For example,
each factor appeared important for BG prey availability in Big Crooked and Camp lakes.
Rapid LMB growth, protracted BG spawning seasons, and high total reproductive output
by BG appeared to contribute to high availability of vulnerable BG in Big Crooked and
Camp lakes. In Heron L., Slow age-0 LMB growth seemed to negatively affect
vulnerable BG availability. In Bass L., the absence of a protracted spawn by BG may
have negatively affected vulnerable BG availability. Finally, in Big Seven and Lobdell
lakes, BG apparently did not Spawn successfully, and thus vulnerable BG availability was

low.

SUMMARY
Together, my experiment and field study demonstrated that macrophyte density,
macrophyte structure, and BG prey availability can all affect age-0 LMB success and can
potentially influence recruitment. My experiment demonstrated that dense and/or

homogeneous macrophytes can reduce age-0 LMB foraging success. Macrophyte

26

structure in all study lakes appeared heterogeneous, perhaps explaining why I did not
observe effects of macrophyte density and structure on LMB piscivory and growth in the
field study. Future studies should quantify the degree of structural heterogeneity at the
whole-system scale and evaluate whether LMB recruitment is higher in systems with
heterogeneous structure (i.e., patchy macrophyte beds with a diversity of growth forms)
compared to systems with homogeneous structure (i.e., extensive monocultures of
invasive macrophytes).

My field study demonstrated that age-0 BG prey availability is highly variable
among systems. Two previously unidentified factors appear especially important for
explaining variation in age-0 BG prey abundance: 1) spawning duration by adult BG and
2) total reproductive output by adult BG. High BG prey abundance was typical in lakes
where adult BG produced many BG fry over an extended time. I found adult BG size—
structure to positively correlate with age-0 BG prey abundance. Mechanisms underlying
variability in BG fry production need further exploration.

From a lake management perspective, evaluation of the indirect effects of specific
activities, such as herbicide application, on LMB recruitment is difficult given the
complexity of lake ecosystems. However, these analyses are critical to establishing

effective and responsible management policies.

27

APPENDIX

Tables and figures

28

Table 1. Characteristics of plant assemblages (species composition and relative
abundance) simulated in each experimental treatment, total water volume displaced by
the plants, estimated total live plant biomass, and plastic plant components. EWM is

Eurasian watermilfoil.

 

 

Treatment Treatment Volume Estimated Components (qty*) Length
Assemblage Density Displaced Biomass (cm)
(ml) (3 dry/m2)

EWM monoculture Moderate 287 93.01 Modified foxtail plants (14) 45.7
EWM monoculture High 553.5 179.38 Modified foxtail plants (27) 45.7
Diverse Native Moderate 290 76.97 Anarchis = Elodea 30.5
- - - - Anarchis 38.1

. - - - Cabomba 38.1

- - - - Cabomba 45.7

. - - - Hygrophila 38.1

- - - - large-leaft 38.1

- - - - Valisneria 45.7
Diverse Native High 549 161.72 Anarchis (2) 30.5
. . - - Anarchis 38.1

. - - - Cabomba 30.5

. - - - Cabomba 38.1

- - - - Cabomba 45.7

. . - - Hygrophila 30.5

. - - - Hygrophila 38.1

. - - - Hygrophila 45.7

- - - - large-leaf’ 30.5

- - - - large-leafl 38.1

- . - . large-leafl 45.7

. . - - Valisneria (2) 30.5

. . . - Valisneria 38.1

. . - - Valisneria 45.7

- - - - Anarchis 30.5

. - - - Hygrophila 38.1

 

* Quantity = 1 if not noted.

'1' Leaves from Amazon Sword plants were attached to polypropylene rope to resemble a

broad-leaf pondweed (e. g., Potamogeton amplifolius)

29

.33 98 wafl 6:93 E x83 amm— ofi mats—o 380:8 924 Bob 339:8 203 Cotonou mafia um=w=< m2: 532 m
.A..meo£o_2 20$: 33 32 5 Eco 353 map E can wmfl E 328 85a was was
5 Son vegans m3: cog cos—:28 we? uni 5.44 :32 .bco 353 map E c8898 52..— uom coca—:23 203 323 was A
ASE >23 95 3935 DUES—8 DEBS—9:65.83 mama: 3532:
:2: 28 Aan £3.50 was .35.): cocmExo Saracen was: ouncmmonm 2 uoto>=oo 203 38.38% Co 2E8 =< a.

.32 A“: E 3% 8 885: 5; causes a; .wme 5 23m 5:; Bass .93 $8 ...

 

 

QR mad 2 we as 2 Z. 32 .353
new «.86 a 2 8 : 4.2 E R: we: Baocoo .133
”.2. ~86 as 2 as R 3a 82 9:5
SE :3 S 3 2. a n: 2 3. we: .3 9:5
5c «85 3 a B 3 SN 82 3.85 "a
has 5:. 3 w 8 on 4.2 2. S M32 :3 .386 ME
3m 23. 8 E 3 on a: 82 .88
«.5 Sod S 8 8 3w ES 3 E 82 2328 55m as
m. _ n was SW 9. S ms 3.. 32 :2»:
NR Rod E a. E as ad mm mm was. Basso 5.5:
3% was S 3. we as 92 so E :2 :3 team
3955 $8 $8 $3 $3
«team—5‘ ANS \ E 3.50 3.50 .830 .830 as Ema: c5
33 cam 32.— 33 332 23m .55 .5... 2:238...— 595 35
a): :82 who :82 .33.: .22.: .22.: .38. .38. :8: saw a; 55.8 8.3

 

.cotsooo bEBflmcoo flea—m :25: E Emu—o 05 9 0.5% Eob «Ba 05 mm 3.5% mm 2.8 3825 6:8
.895 05 E 3.5 BEES 35E Co 35:38 2.0on DE 82365 uo>oo ESQ .82:— BBEB .83. 0:28 2: ~03 BEES
853 Mo 35:38 83.3 2.: 3:86.: co>oo 22¢ .88. .8805 063 352m 523 3 £50m BEES ho 033083

on. 8 8&8 .260 dawEozz E moo—E 32m x7. 2: no mocmtaofimno owe—95.0.8 oSianE new _au_wo_o=E§ .N 2an

30

 

Blind

 

 

 

321cm
f‘fi
{’x ‘
Observer 1‘ 76 cm

 

\

\

\
\
t
2
l
I

\
\‘ ’r
N C
114 L
—

 

Aquarium _..=

 

Suspended lighting

-
I ’ -‘\‘
’ \
\
\
A
I
I
l
’l
I
an"

 

 

 

 

 

 

 

 

 

 

Plexiglas ,'
divider \ ~\ _,./
E3“iCkiC‘N1$tflJCfiC”1 I-IllI-I-III ‘I-I-I-I-I-I ,l’llllllfl
\ \ ‘l’ V
‘
\ 5" ;
\\ ’1’ Table

 

 

 

 

Figure 1. Top-down view of the laboratory set-up for the age-0 largemouth bass foraging
behavior experiment.

31

 

 

 

 

 

 

   

 

 

 

 

 

 

 

 

 

   

 

 

 

 

 

 

 

 

 

 

40‘ A First Attack
43
A 40
5 30- 41 _-
.\° W
in
m - 45
fl 20
g _-
0 -
2 10
.2 fl
3
0 0
g B First Consumptlon
as

9. 60‘ :2
E 50 ~
g 4o-
l.l.| ’ 38
o 30 ~
.§
I- 20 - "

10 - _ ....

0 7 : 3-1: :

Moderate High EWM Diverse Low High
Plant Density Structure Bluegill Density

Figure 2. Geometric mean number of seconds (i 95% CI) elapsed before a largemouth
bass attacked (A), and consumed (B) its first bluegill prey in moderate and high plant
density treatments, Eurasian watermilfoil (EWM) monocultures and diverse plant
treatments, and low and high bluegill density treatments. Main effect p-values from a
three-way ANOVA are reported. No interactions were significant. Numbers above bars
denote the number of replicates.

32

 

 

 

 

 

 

 

 

 

 

 

18
51 37 41
A 15 - O O 52 0
g 0 O O
.E
E 12 -
h
8
0' 9 - o o o
E 47
g o O o
a o o e
E 6 «
f) O 0 0
g I 37 g °
< 3 " 0 g é ;
O . . dz, :,
Moderate High EWM Diverse Low High
Plant Density Structure Bluegill Density

Figure 3. Box plots of attack rate (attacks per minute of search) by largemouth bass on
bluegill in each treatment of the largemouth bass foraging experiment. Comparisons
shown are attack rates in moderate and high plant density treatments, Eurasian
watermilfoil (EWM) monoculture and diverse plant treatments, and low and high bluegill
density treatments. P-values from Wilcoxin rank-sums tests are reported. Boxes
represent the range in which 50% of the data points fall (i.e., interquartile range or IQR).
Wiskers represent the range of data that fall within 1.5 times the IQR. Diamonds
represent data points that are within 3 times the IQR. Circles represent outliers.
Numbers above plots denote the Wilcoxin mean rank score.

33

 

45
42

45

 

42

.....

Total Consumption
(Mean No. 36 eaten :t1 SE)

 

 

 

 

 

 

 

 

 

 

 

 

Moderate High EWM ”Diverse Low High
Plant Density Structure Bluegill Density

Figure 4. Mean number (i 1 SE) of bluegill (BG) consumed by age-O largemouth bass in
the largemouth bass foraging experiment. Comparisons shown are total consumption in
moderate and high plant density treatments, Eurasian watermilfoil (EWM) monocultures
and diverse plant treatments, and low and high bluegill density treatments. Main effect p-
values from a three-way ANOVA are reported. No interactions were significant.
Numbers above bars denote the number of replicates.

34

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 
   
  

 

 

 

 

 

 

 

 

 

 

 

 

 

10 8 ’
2 Bass 8 ., C5"“P Heron
4] I Seine I Seine 6 I Seine
3‘ 0 Shock 6‘ 0 Shock 4 0 Shock
2
E 1 2
8 0 o

0
1
21
3 ..
4 q
5 " .
6 I I I I I 10 i V V U 8 I I I l

40 50 60 70 8O 90 100 60 70 80 90 100 1 10 40 50 70 BO 90
Largemouth Base Size Class (mm TL) Largemouth Base Size Cl!“ (mm TL) Largemouth Bees Size Class (mm TL)
12 20
Big Crooked Big Seven

9 I Seine 15 d I Seine
6 0 Shock 10 . 0 Shock
3 5 ‘

E o E

0

8 5
3 ‘ 5 ‘
6 ‘ 10 "
9 r 15 ‘

12 I I F r I 20 I I I
50 60 70 80 90 100 1 10 60 70 80 90 100

Largemouth Bees Size Class (mm TL) Largemouth Base Size Class (mm TL)

Figure 5. Size distributions of age-O largemouth bass collected in late September 1999
both from electrofishing (120 volts pulsed-DC) and shoreline seining.

35

 

0.0008

 

 

 

 

 

 

 

 

 

g — 1998
2 ,5. ------- 1999
<5 1: 0.0006 -
'9 3
2 s.
a
2 \ 0.0004 - 0
m in
2 3
—| E e e
2 5." 0.0002 - 1 —§--
In T
<

0 I l l l

0 20 40 60 80 100

Percent Eurasian Watermilfoil Coverage

Figure 6. Relationship between Eurasian watermilfoil coverage and absolute growth
(standardized to growing degree days) of age-0 largemouth bass (LMB). Vertical and
horizontal bars are used for lakes were measured in both years and represent the range of
means in 1998 (solid line) and 1999 (dashed line); points in the middle represent the
mean effect and response for each lake measured in both years.

36

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1998
025 duly Auc ust
0 1411810" ”1911- EWM 000(21
0.154
0.1-1
0.031 n _nn _ne-|
TITTIITII 'UUIIII
A 0.2th LOW- EWM 000013
N 0.154
E 011
h .
d) 0.05<
0" oIIIIITTIIIIIIIIIIfi—rr1rrrrfirI
c 0218888 High- EWM 0.00025
2 0.151
‘_" 0.1-
%, 0.05-
g ofi'I'UIIIIfilIIIUITTITU$TTTU
00 NO DATA
9
a)
a .
< 0'2 Big C. Low- EWM 0.00059
LLI 0.15‘
8 0.1‘
0.05"
.. 41954110.
02 ‘ Camp Low- EWM 0.000461
0.15“
0.1‘
0.05‘ _
0 I I I I ITI r
1018 2026101620253010152025303101520330310152025303540“
Size Class (mm TL)

 

 

 

 

 

 

1999
0 5 July August September
0.4 0.00017 Heron-High EWM I vunerable
°'3 ‘ El Noivulnomud
0.2 '1
0.1 .
O 7 nl'll‘ln
IIIIT iIIlT IIIIIIII
0.4 ~ 0.00121 Lobdell-High EWM
0.3 3
0.2 ~
0.1 .
llTTl IIITTTII iTiTTTII
0.4 ~ 0.00025 Bass-Low EWM
0.3 1
02 -
0'; ' _ _ nnn
Fill! liiliiri IIIIITTI-
0.4 0.00040 Big Seven-
0.3 ' -
1
02 . H gh EWM
0.1 ~
0 III?I ITT‘I'_III1 IITIIITI
0-4 ‘ 00049 Big Crooked-
” ‘ Low EWM
0.2 4
0.1 4
o 1 WI—
0.4 0.00074 Campiow
0.3 EWM
0.2
0.1

 

 

 

 

10152025” 10152025335045 101520§M34045

Size Class (mm TL)

 

Figure 7. Catch per effort (CPE; No./ m2) of age-0 bluegill (determined by length
distributions) captured in seine hauls during summer 1998 and 1999. Black bars

represent vulnerable bluegill (individuals S 40% of the mean largemouth bass TL). Open

bars are bluegill that were too large for age-0 largemouth bass to consume. Dashed
vertical lines separate each sampling date and solid vertical lines separate months.

Values in the upper right of each lake’s distribution in 1998 and upper left in 1999

indicate the absolute growth of age-0 largemouth bass. Lakes are ordered by ascending
age-0 largemouth bass absolute growth rates in 1999. Note y-axis scales differ between
1998 and 1999. All lakes with < 42% littoral coverage of EWM were designated as
“low” EWM lakes.

37

 

 

 

 

 

 

 

 

 

E
2 A
c) F -7.5-
o 1:
§ 8
O L.
3 8’
<1: -8.0-
m \
E 3 e
_| (0
a: E '

<1 -8.54 :
m" g —4,9 ------ — 1998
< ' ------- 1999
E T e Low EWM

El High EWM
-9.0 1 1
-8 -6 -4 -2 0

in Mean CPE Vulnerable Age-O Bluegill (No. I m2)

Figure 8. Relationship between mean CPE of vulnerable age-0 bluegill [individuals S
40% of the mean length of age-O largemouth bass (LMB) captured during the same
period] and absolute growth (standardized to growing degree days) of age-O LMB.
Vertical and horizontal bars are used for lakes measured in both years and represent the
range of means in 1998 (solid line) and 1999 (dashed line); points in the middle represent
the mean effect and response for each lake measured in both years. All lakes with < 42%
littoral coverage of EWM were designated as “low” EWM lakes.

38

 

 

 

 

 

 

 

 

 

1998
100 - g .- ,0
—.e.— Zooplankton I: '\ ,/
~---Ei-~ Insects ,°' \ -"
80 - -0— FiSh j"! 1. .I' Y
60 - if] % 321‘ j.
A ’ i
x" ~..
31 .1 2=
C 1- l.
O 40 4 .° '-.
a H .I
3 £1 .-’
g- g 20 -' ‘ I" . ..... ‘ '1‘".
O m 4 """ I . 1 j \‘1 II
o 5 /'/ \4 .. “I \ ‘~
5.5 E 0 a" . ” .
D g 1999 I!
2 a) " :
E N ,1
O E 80 ‘ T j
8 .\° 1 j
g g .. 1,.» . ”,4
P m 60 -I 5’." xx? ./+
E I". 5...... /o
v 3' ‘~. .°
. / ‘--.. .’ .1
4O _ '3 111‘ " -H-
Ecc’fffl .l. \\ q .-
I. :1
L. . KEN“. Ji-
20 ' -‘ /" {I
/. . .\
./£"\. I
e'/‘
0 ‘12 I I I I I I

20 3040 50607 809100
Largemouth Bass Size Class (mm TL)

Figure 9. Taxonomic diet composition of age-0 largemouth bass in 1998 and 1999.
Values reported in 1998 and 1999 are means (weighted by sample size for each size class
in each lake) of 5 and 6 lakes respectively (i 1 SE).

39

 

 

 

 

 

 

In Ageo LMB Absolute Growth

 

   

 

 

 

 

 

O l l l T
0 10 20 30 40
Mean Occurrence of Piscivory (%)
4
0e- B ' """
o .
8 A 3 -
gs
: 2' '
0 O 2 ‘
3 .2
c g -—— 1998
3 °- 1 4 5 ------- 1999
2 D e Low EWM
S. a High EWM
O 1 1 1
-8 -6 -4 -2 0

in Mean CPE Vulnerable Age-0 Bluegill (No. I m”)

 

 

 

Figure 10. Absolute growth (standardized to growing degree days) of age-0 largemouth
bass (LMB) as a function of mean percent occurrence of piscivory (A) and mean percent
occurrence of piscivory as a function of bluegill prey availability (CPE of bluegill _<. 40%
of the mean length of age-0 LMB captured during the same period; B). Frequency of
piscivory was estimated from diet analysis as the proportion of LMB stomachs that
contained fish and reflects seasonal means. Vertical and horizontal bars are used for
lakes measured in both years and represent the range of means in 1998 (solid line) and
1999 (dashed line); points in the middle represent the mean effect and response for each
lake measured in both years. All lakes with < 42% littoral coverage of EWM were
designated as “low” EWM lakes.

40

 

 

 

 

 

 

 

 

 

0
iii: '2‘
3
'3’ E
7.: 3:
s 4}; '4‘
E a:
5 s
E 3 1998
’6‘ 1999
e Low EWM
a High EWM
'8 I I I I
4.75 4.85 4.95 5.05 5.15

In 90th Percentile Adult Bluegill Length (mm TL)

Figure 11. The relationship between adult bluegill size and the catch per effort (CPE) of
bluegill < 25 mm TL during late August. Adult bluegill were captured during spring
1998 and 1999 electrofishing surveys. Vertical and horizontal bars are used for lakes
measured in both years and represent the range of means in 1998 (solid line) and 1999
(dashed line); points in the middle represent the mean effect and response for each lake
measured in both years. All lakes with < 42% littoral coverage of EWM were designated
as “low” EWM lakes.

41

LITERATURE CITED

42

LITERATURE CITED

Anderson, 0. 1984. Optimal foraging by largemouth bass in structured environments.
Ecology 65: 851-861.

Annett, C., J. Hunt, and ED. Dibble. 1996. The compleat bass: habitat use patterns of all

stages of the life cycle of largemouth bass. American Fisheries Society
Symposium 16: 306-314.

Bain, MB. and LA. Helfrich. 1983. Role of male parental care in survival of larval
bluegills. Transactions of the American Fisheries Society 112: 47-52.

Bettoli, P.W., M.J. Maceina, R.L. Noble, and R.K. Betsill. 1992. Piscivory in largemouth
bass as a function of aquatic vegetation abundance. North American Journal of
Fisheries Management 12: 509-516.

Breck, J. E. 1993. Hurry up and wait: growth of young bluegills in ponds and the
stimulation with an individual-based model. Transactions of the American
Fisheries Society 122: 467—480.

Cargnelli, L.M. and MR. Gross. 1996. The temporal dimension in fish recruitment: birth
date, body size, and size-dependent survival in a sunfish (bluegill: Lepomis
macrochirus). Canadian Journal of Fisheries and Aquatic Sciences 53: 360-367.

Carpenter, SR. 1980. The decline of Myriophyllum spicatum in a eutrophic Wisconsin
USA Lake. Canadian Journal of Botany 58: 527-535.

Carpenter, SR, and D.M. Lodge. 1986. Effect of submersed macrophytes on ecosystem
processes. Aquatic Botany 26: 341-370.

Claussen, J .E. 1991. Annual variation in the reproductive activity of a bluegill
population: effect of clutch size and temperature. MS. Thesis. University of
Toronto, Canada.

Coffey, B.T. and CD. McNabb. 1974. Eurasian watermilfoil in Michigan. The
Michigan Botanist 13: 159-165.

Coleman, R.M., M.R. Gross, and RC. Sargent. 1985. Parental investment decision rules:
a test in bluegill sunfish. Behavioral Ecology and Sociobiology 18: 59-66.

Coleman, RM. and R.U. Fischer. 1991. Brood size, male fanning effort and the
energetics of a nonshareable parental investment in bluegill sunfish, Lepomis
macrochirus (Teleostei: Centrarchidae). Ethology 87: 177-188.

Colle, DE. and J.V. Shireman. 1980. Coefficients of condition for largemouth bass,

bluegill, and redear sunfish in hydrilla-infested lakes. Transactions of the
American Fisheries Society 109: 521-531.

43

Crowder, LB and WE Cooper. 1979. Structural complexity and fish-prey interactions in
ponds: a point of view. pp. 2-10 in D.L. Johnson and RA. Stein (eds.) Response
of fish to habitat structure in standing water. North Central Division American
Fisheries Society Special Publication 6.

Dibble, E.D., K.J. Killgore, and S.L. Harrel. 1996. Assessment of fish-plant interactions.
American Fisheries Society Symposium 16: 357-372.

Durocher, P.P, W.C. Provine, and J .E. Kraai. 1984. Relationship between abundance of
largemouth bass and submerged vegetation in Texas reservoirs. North American
Journal of Fisheries Management 4: 84-88.

Engel, S. 1987. The impact of submerged macrophytes on largemouth bass and bluegills.
Lake and Reservoir Management 3: 227-234.

Engel, S. 1995. Eurasian watermilfoil as a fishery management tool. Fisheries 20: 20-27.

Frost, T.M, D.L. DeAngelis, S.M. Bartel], D]. Hall, and SH. Hurlbert. 1988. Scale in the
design and interpretation of aquatic community research. pp. 229-258 in SR
Carpenter (ed) Complex interactions in lake communities. Springer-Verlag, New
York.

Garvey, J .E., N.A. Dingledine, N.S. Donovan, and RA. Stein. 1998a. Exploring spatial
and temporal variation within reservoir food webs: predictions for fish
assemblages. Ecological Applications 8: 104-120.

Garvey, J .E. and RA. Stein. 1998. Linking bluegill and gizzard shad prey assemblages to
growth of age-0 largemouth bass in reservoirs. Transactions of the American
Fisheries Society 127: 70-83.

Garvey, J .E., R.A. Wright , and RA. Stein. 1998b. Overwinter growth and survival of
age-0 largemouth bass (Micropterus salmoides): revisiting the role of body size.
Canadian Journal of Fisheries and Aquatic Sciences 55: 2414-2424.

Glass, NR. 1971. Computer analysis of predation energetics in the largemouth bass. pp.
325-363. in BC. Patten (ed) Systems analysis and simulation in ecology vol. 1.
Academic Press, New York.

Gotceitas, V., and P. Colgan. 1987. Selection between densities of artificial vegetation by
young bluegills avoiding predation. Transactions of the American Fisheries
Society 116: 40-49.

Grace, J .B. and R.G. Wetzel. 1978. The production biology of Eurasian watermilfoil

(Myriophyllum spicatum L.): a review. Journal of Aquatic Plant Management 16:
1-11.

Gross, MR. 1980. Sexual selection and the evolution or reproductive strategies in
sunfishes (Lepomis: Centrarchidae). Ph.D. dissertation. University of Utah, Salt
Lake City.

Gross, MR. 1982. Sneakers, satellites and parentals: Polymorphic mating strategies in
North American sunfishes. Zeitschrift ffir Tierpsychologie 26: 1-26.

Hayse, J .W. and T.E. Wissing. 1996. Effect of stem density of artificial vegetation on
abundance and growth of age-0 bluegills and predation of largemouth bass.
Transactions of the American Fisheries Society. 125: 422-433.

Keast, A. and J. M. Eadie. 1985. Growth depensation in year-0 largemouth bass: the
influence of diet. Transactions of the American Fisheries Society 114: 204-213.

Lawrence, J .M. 1958. Estimated sizes of various forage fishes largemouth bass can
swallow. Proc. Ann. Conf. SE Assoc. Game Fish Comm. 1]: 220-225.

Maceina, M.J. 1996. Largemouth bass abundance and aquatic vegetation in Florida lakes:
an alternative interpretation. Journal of Aquatic Plant Management 34: 43-47.

Madsen, J .D., J .W. Sutherland, J .A. Bloomfield, L.W. Eichler, and CW. Boylen. 1991.
The decline of native vegetation under dense Eurasian watermilfoil canopies.
Journal of Aquatic Plant Management 29: 94-99.

Madsen, JD. 1997. Methods for management of nonindigenous aquatic plants. pp. 145-
171 in J .O. Luken and J .W. Thieret (eds.). Assessment and management of plant
invasions. Springer, New York.

Madsen, J .D. 1999. Point intercept and line intercept methods for aquatic plant
management. Army Corps of Engineers Waterways Experiment Station Technical
Note MI-02.

Madsen, J .D., K.D. Getsinger, T.J. Koschnick, M.D. Netherland, R.M. Stewart, D.R.
Honnell, A.G. Stadden, and CS. Owens. in press. Whole lake
applications of Sonar for selective control of Eurasian watermilfoil.
Technical Report EL— , US Army Engineer Research and Development Center,
Vicksburg, MS.

McMahon, T.E. and SH. Holanov. 1995. Foraging success of largemouth bass at
different light intensities: implications for time and depth of feeding. Journal of
Fish Biology 46: 759-767.

45

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: 280-282.

Miller, 8.]. and T. Storck. 1984. Temporal spawning distribution of largemouth bass and
young-of-the-year growth, determined from daily otolith rings. Transactions of
the American Fisheries Society 113: 571-578.

Miranda, LE, and LL. Pugh. 1997. Relationship between coverage and abundance, size,
and diet of juvenile largemouth bass in winter. North American Journal of
Fisheries Management 17: 601-609.

Mittelbach, G.G., A.M. Turner, D]. Hall, J .E. Rettig, C.W. Osenberg. 1995. Perturbation
and resilience: a long-term, whole-lake study of predator extinction and
reintroduction. Ecology 76: 2347-2360.

Murphy, J. and L. P. Riley. 1962. A modified single solution method for the
determination of phosphate in natural waters. Analytica Chemica Acta 27:31-36.

Nichols, S.A., R.C. Lathrop, and SR. Carpenter. 1992. Long-term vegetation trends: a
history. pp. 151-171 in J .F. Kitchell (ed.) Food web management: a case study of
Lake Mendota. Springer-Verlag New York.

Olson, M.H., G.G. Mittelbach, and CW. Osenberg 1995. Competition between predator
and prey: resource-based mechanisms and implications for stage-structured
dynamics. Ecology 76: 1758-1771.

Olson, M.H. 1996. Ontogenetic niche shifts in largemouth bass: variability and
consequences for first-year growth. Ecology 77: 179-190.

Olson, M.H., S.R. Carpenter, P. Cunningham, S. Gafney, B.R. Herwig, N.P. Nibbelink,
T. Pellet, C. Storlie, A.S. Trebitz, and K.A. Wilson. 1998. Managing macrophytes
to improve fish growth: a multi-lake experiment. Fisheries 23: 6-12.

Phillips, J .M., J .R. Jackson, R.L. Noble 1995. Hatching date influence on age specific
diets and growth of age-0 largemouth bass. Transactions of the American
Fisheries Society. 124: 370-379.

Power, M.E., D. Tilman, J .A. Estes, B.A. Menge, W.J. Bond, L.S. Mills, G. Daily, J .C.
Castilla, J. Lubchenco, and, RT. Paine. 1996. Challenges in the quest for
keystones. Bioscience 46: 609-619.

SAS Institute Inc. 1990. SAS user’s guide: statistics, version 6.12. SAS institute, Cary,
North Carolina.

46

Savino, J .F. and RA. Stein. 1982. Predator-prey interaction between largemouth bass
and bluegills as influenced by simulated, submersed vegetation. Transactions of
the American Fisheries Society 111: 255-266.

Schramm, H.L. and AV. Zale. 1985. Effects of cover and prey size on preferences of
juvenile largemouth bass for blue tilapias and bluegills in tanks. Transactions of
the American Fisheries Society 114: 725-731.

Smith, CS, and J.W. Barko. 1990. Ecology of Eurasian watermilfoil. Journal of Aquatic
Plant Management 28: 55-64.

Smith, K.D. 1993. Vegetation-open water interface and the predator-prey interaction
between largemouth bass and bluegills. Ph.D. dissertation. University of
Michigan, Ann Arbor.

Stein, R.A., Bremigan, M.T., and J .M. Dettmers. 1996. Understanding reservoir systems
with experimental tests of ecological theory: a prescription for management.
American Fisheries Society Symposium 16: 12-22.

Titus, J .E. and MS. Adams. 1979. Coexistence and the comparative light relations of the
submersed macrophytes Myriophyllum spicatum L. and Vallisneria americana
Michx. Oecologia 40: 273-286.

Unmuth, J.M.L., M.J. Hansen, and TD. Pellett. 1999. Effects of mechanical harvesting of
Eurasian watermilfoil on largemouth bass and bluegill populations in Fish Lake,
Wisconsin. North American Journal of Fisheries Management 19: 1089-1098.

Weaver, M.J., J .J . Magnuson, and MK. Clayton. 1997. Distribution of littoral fishes in

structurally complex macrophytes. Canadian Journal of Fisheries and Aquatic
Science 54:2277-2289.

Wiegmann, DD. and J .R. Baylis. 1995. Male body size and paternal behaviour in
smallmouth bass, Micropterus dolomieui (Pisces: Centrarchidae). Animal
Behaviour 50: 1543-1555.

Wrenn, W.B., D.R. Lowery, M.J. Maceina, and WC. Reeves. 1996. Relationships

between largemouth bass and aquatic plants in Guntersville Reservoir, Alabama.
American Fisheries Society Symposium 16: 382-393.

47