Long-Term Changes in Population Statistics of Freshwater Drum (Aplodinotus Grunniens) in Lake Winnebago, Wisconsin, Using Otolit

Home , Freshwater drum, Percopsis omiscomaycus, Rough fish, Turbo sarmaticus

LONGTERM CHANGES IN POPULATION STATISTICS OF FRESHWATER DRUM ( APLODINOTUS GRUNNIENS ) IN LAKE WINNEBAGO, WISCONSIN, USING OTOLITH GROWTH CHRONOLOGIES AND BOMB RADIOCARBON AGE VALIDATION

Shannon L. DavisFoust

A Dissertation Submitted in

Partial Fulfillment of the

Requirements for the Degree of

Doctor of Philosophy

in Biological Sciences

The University of WisconsinMilwaukee

August 2012

LONGTERM CHANGES IN POPULATION STATISTICS OF FRESHWATER DRUM ( APLODINOTUS GRUNNIENS ) IN LAKE WINNEBAGO, WISCONSIN, USING OTOLITH GROWTH CHRONOLOGIES AND BOMB RADIOCARBON AGE VALIDATION

Shannon L. DavisFoust

A Dissertation Submitted in

Partial Fulfillment of the

Requirements for the Degree of

Doctor of Philosophy

in Biological Sciences

The University of WisconsinMilwaukee

August 2012

Major Professor Dr. Rebecca Klaper Date

Graduate School Approval Date

ABSTRACT

LONGTERM CHANGES IN POPULATION STATISTICS OF FRESHWATER DRUM ( APLODINOTUS GRUNNIENS ) IN LAKE WINNEBAGO, WISCONSIN, USING OTOLITH GROWTH CHRONOLOGIES AND BOMB RADIOCARBON AGE VALIDATION

Shannon DavisFoust

The University of WisconsinMilwaukee, 2012 Under the Supervision of Dr. Rebecca Klaper

Estimating fish population statistics such as mortality and survival requires the

use of fish age data, so it is important that age determinations are accurate. Scales have

traditionally, and erroneously, been used to determine age in freshwater drum

(Aplodinotus grunniens ) in the majority of past studies. I used bomb radiocarbon dating

to validate that sagittal otoliths of drum are the only accurate structure (versus spines or

scales) for age determinations. Drum grow slower and live longer than previously

recognized by scale reading.

Drum otoliths can be used not only to determine age, but also to estimate body length. I examined archaeological drum otoliths dated to the 1830's and older from around the Winnebago system. Maximum body length attained by modern drum had not significantly changed over time; however, longevity is reduced and growth rates have increased in modern drum. The changes in these demographic statistics could be due to several environmental factors that have changed in the Lake Winnebago system since

European settlement, and in particular the effects from the removal trawling program that

iii

removed large quantities of drum annually from 1936 to 1990.

Disentangling growth responses to environmental factors such as climate and trophic interactions is a major goal of ecology. Annual growth rates measured from otoliths have frequently been reported to correspond to temperature, but not to food abundance. I constructed growth chronologies of drum using a linear mixed model that separates endogenous and exogenous growth effects. Diet analysis of drum revealed an adult lengthrelated diet shift. The growth rates of large drum from 2001 to 2008 were significantly greater than small drum, which coincided with the arrival of zebra mussels

(Dreissena polymorpha ) in the Lake Winnebago system. The increase in growth rates of large drum were positively correlated to, but departed from, regional temperatures compared to small drum. The increased growth rates of large drum post zebra mussel establishment were corroborated by improved in body condition of large drum. Zebra mussels provide a significant source of nutrition to drum, which is unique because most effects of dreissenids are reported at lower trophic levels.

Major Professor Dr. Rebecca Klaper Date

TABLE OF CONTENTS

LIST OF FIGURES ...... viii LIST OF TABLES...... xi PREFACE...... xii DEDICATION...... xiii ACKNOWLEDGEMENTS...... xiv CHAPTER ONE: INTRODUCTION TO USING A BIOLOGICAL CHRONOMETER TO GAUGE ENVIRONMENTAL CHANGE...... 1 Background of the Lake Winnebago System ...... 4

Background of Freshwater Drum ...... 11

Ecological Roles of Freshwater Drum...... 14

Historical Management of Lake Winnebago Freshwater Drum...... 18

CHAPTER TWO: AGE VALIDATION OF FRESHWATER DRUM USING BOMB RADIOCARBON ...... 24 Introduction...... 24

Methods ...... 27

Results...... 32

Discussion...... 40

CHAPTER THREE: LONGTERM CHANGES IN THE DEMOGRAPHY OF A FRESHWATER FISH CORRESPONDING TO ANTHROPOGENIC DISTURBANCES46 Introduction...... 46

Methods ...... 48

Results...... 54

Discussion...... 63

CHAPTER FOUR: FRESHWATER DRUM RESPONSE TO THE INTRODUCTION OF ZEBRA MUSSELS IS REFLECTED IN THEIR OTOLITH GROWTH CHRONOLOGY...... 69

Introduction...... 69

Methods ...... 72

Results...... 77

Discussion...... 84

CHAPTER FIVE – CONCLUSIONS: LINKING IMPACTS OF ROUGH FISH REMOVAL AND OTHER CHANGING ENVIRONMENTAL FACTORS TO THE FRESHWATER DRUM POPULATION IN LAKE WINNEBAGO ...... 87 APPENDICES ...... 93 Appendix A: Area Swept Population Estimate of Lake Winnebago Freshwater Drum94

Appendix B: The Linear Mixed Effects Model...... 97

Appendix C: The Lake Winnebago Freshwater Drum Growth Chronology Index.101

Appendix D: Length Distribution Trends of Lake Winnebago Freshwater Drum from 19862011 ...... 103

Appendix E: Year Class Strength of Lake Winnebago Freshwater Drum ...... 104

Appendix F: Stock Recruitment Analysis of Lake Winnebago Freshwater Drum..105

LITERATURE CITED ...... 107 CURRICULUM VITAE...... 121

vii

LIST OF FIGURES

Figure 1.1. Lake Winnebago is part of the FoxWolf Watershed, which covers approximately 17% of the land area of Wisconsin (map modified from Wisconsin Department of Natural Resources 1989). The map on the left depicts the Fox River (dotted area) and the Wolf River basins (striped area)...... 5

Figure 1.2. Approximate timings for major anthropogenic disturbances to Lake Winnebago since the impoundment of the lower Fox River in 1850...... 7

Figure 1.3. “Molars” in the freshwater drum pharynx, which are adapted for crushing mussel shells...... 13

Figure 2.1. Otolith core 14 C chronologies for freshwater drum (triangles), Arctic char and lake trout (small squares; Campana et al. 2008), black drum from Chesapeake Bay (plus signs; Campana and Jones 1998), gray snapper from the Gulf of Mexico (large squares; adapted from Fischer et al. 2003), together with the atmospheric values from the Western Hemisphere (times signs; adapted from Nydal 1993). The 14 C values are fitted with locally weighted leastsquare regressions...... 34

Figure 2.2. Catch per unit effort (CPUE) of age0 freshwater drum from experimental trawl assessments, Lake Winnebago, 1962–1984...... 35

Figure 2.3. Age frequency histograms for freshwater drum sampled in assessment trawl surveys in 1986 and from 2003 to 2007 at Lake Winnebago showing the progression of the strong 1983 yearclass (asterisks). Fish older than age 40 were omitted because of the small sample size...... 36

Figure 2.4. Growth increments of four structures from a 564 mm, 2,268 g freshwater drum from Lake Winnebago in 2003 (sex not determined) yielding different age estimates: (A) scale (14 years), (B) anal spine (10 years), (C) dorsal spine (12 years), and (D) transversely sectioned otolith (20 years). The bars represent 1 mm; the black circles indicate the approximate locations of the interpreted annuli. Edges were not counted as complete annuli. If unequal numbers of annuli were counted on the two sides of a spine, the side with the higher number was used...... 38

Figure 2.5. Otolithestimated ages of freshwater drum from Lake Winnebago in 2003 versus the mean age estimated from scales, anal spines, and dorsal spines. All values are

viii

years; the error bars represent 95% confidence intervals. Values that fall on the 1:1 diagonal lines represent full agreement between the respective structures...... 39

Figure 3.1. Locations for the recovery of drum otoliths at archaeological sites surrounding the Lake Winnebago system. At each site, the number of otoliths recovered and the approximate date of site occupancy is shown...... 51

Figure 3.2. Freshwater drum otolith length distributions for three archaeological sites (Bell, Kargus, and Doty) and three modern capture methods (Angling, Blast, and Trawl). Note that “Bell/Kargus” and “Blast/Trawl” represent distributions from pooled samples...... 56

Figure 3.3. Freshwater drum age distributions for the four subset groups. “Bell/Kargus” and “Blast/Trawl” represent distributions from pooled samples. Archaeological subsets were given the prefix “arch” and modern subsets the prefix “mdrn.”...... 59

Figure 3.4. von Bertalanffy growth models were fit to the four subsets using otolith length at age. Archaeological subsets were given the prefix “arch,” and modern subsets the prefix “mdrn.” ...... 61

Figure 3.5. von Bertalanffy growth models were fit to the four subsets using otolith length at age. Archaeological subsets were given the prefix “arch,” and modern subsets the prefix “mdrn.” Confidence intervals were fit with the profile likelihood method...... 61

Figure 3.6. Otolith length versus total (body) length of 481 modern drum...... 62

Figure 4.1. Orientation and location of the sulcus in a sagittal otolith section from a 58 year old drum. The black line indicates the axis where increments were measured...... 74

Figure 4.2. The proportion of freshwater drum within 20 mm total length classes that had consumed items within each food category. Food items making up <2% of the diet are not shown...... 78

Figure 4.3. The length class (384 mm TL) at which half of the drum were found to consume zebra mussels was used to create small and large size classes for the growth chronology and body condition analyses...... 78

Figure 4.4. The master growth chronology of Lake Winnebago freshwater drum was plotted by year (A). Positive year effects indicate years when drum had relatively good

growth, and negative year effects indicate years when drum had relatively poor growth. Year effects and growing degree days from a 10°C base (GDD) were standardized to a mean of 0 and a standard deviation of one (B). The vertical line marker at 1998 indicates when zebra mussels were first reported in Lake Winnebago. Confidence intervals represent +/ two standard errors...... 80

Figure 4.5. Growth chronologies for sex (A) and size (B) subsets of freshwater drum. The vertical line marker at 1998 indicates when zebra mussels were first reported in Lake Winnebago. Confidence intervals represent +/ two standard errors. The asterisks denote years of significant difference between the year effect values...... 82

Figure 4.6. Relative weights (Wr) of each size subset of drum were examined for changes pre and postzebra mussel arrival. The vertical line marker at 1998 indicates when zebra mussels were first reported in Lake Winnebago. Confidence intervals represent ±2 standard errors...... 83

Figure A.1. The mean catch per unit effort (CPUE) of adult Lake Winnebago freshwater drum from assessment trawling conducted in August, September, and October each year increased from 19862010...... 96

Figure D.1. Length frequencies of freshwater drum measured using random selection during October Lake Winnebago assessment trawling...... 103

Figure E.1. Year class strengths of freshwater drum spanning approximately 60 years...... 104

Figure F.1. Lake Winnebago freshwater drum adult catch per unit effort (CPUE) and age0 CPUE’s were collected from fall assessment trawling from 19862010...... 106

LIST OF TABLES

Table 2.1. Collection year, core weight, age (based on otolith growth increments), year class (based on otolith age), and 14 C and δ 13 C assay values for freshwater drum sagittal otolith cores sampled from Lake Winnebago, Wisconsin, in 1986, 2003, and 2006...... 33

Table 3.1. Summary statistics for freshwater drum otolith lengths at three archaeological sites (Bell, Doty Island, and Kargus) and the modern gears (Blast, Angling, and Trawl). “Bell/Kargus” and “Blast/Trawl” represent pooled samples...... 57

Table 3.2. Pairwise comparisons of mean otolith lengths among freshwater drum within each of three archaeological sites (Bell, Doty, Kargus), three modern gears (Angling, Blast, Trawl), and among four subset groups using Tukey’s method and a resampling procedure. The results for the resampling procedure are the percentage of 1000 resamplings where the comparison was significantly different...... 57

Table 3.3. Summary statistics for freshwater drum ages for the four subsets. “Bell/Kargus” and “Blast/Trawl” represent pooled samples...... 59

Table 3.4. Pairwise comparisons of mean ages among freshwater drum within each of the four subset groups using Tukey’s method and a resampling procedure. The results for the resampling procedure are the percentage of 1000 resamplings where the comparison was significantly different. Archaeological subsets were given the prefix “arch,” and modern subsets the prefix “mdrn.” ...... 60

Table 3.5. The analysis of the archaeological otoliths was built on the following assumptions and simplifications for unknown factors regarding the archaeological drum otoliths...... 65

Table A.1. Catch per unit effort (CPUE) was calculated for Lake Winnebago freshwater drum from assessment trawling conducted from 19862010 and used to estimate abundance within the lake...... 95

PREFACE

This dissertation is the culmination of a research project that I began in 2003. The project gained momentum with the recognition that freshwater drum live far longer and grow much slower than earlier studies had indicated. Further, there were a lot of freshwater drum structures and data at hand. Hundreds of archaeological freshwater drum otoliths had been recovered from around the Lake Winnebago system, and data on the Lake Winnebago fish community has been collected by the Wisconsin Department of

Natural Resources for several decades.

Chapter 2 completes the “forgotten requirement of age validation” (Beamish and

McFarlane 1983) for freshwater drum. This work was done in cooperation with Dr.

Steven Campana from the Bedford Institute of Oceanography in Nova Scotia, who is a prominent scientist in the fields of fish age and growth and bomb radiocarbon dating.

This chapter has been published in Transactions of the American Fisheries Society .

Chapter 3 is an analysis of archaeological freshwater drum population statistics.

Archaeologists have recovered hundreds of otoliths from in the Lake Winnebago vicinity that date back to the 1830’s and earlier. These structures were loaned for this research by

Dr. Jeffrey Behm and Richard Mason of the University of Wisconsin Oshkosh.

Chapter 4 is on the growth chronology developed from modern Lake Winnebago freshwater drum otoliths. The growth chronology models were developed with the assistance of Dr. George Spangler and Dr. Sanford Weisberg, University of Minnesota, and with Dr. Derek Ogle, Northland College, who also consulted me with programming in R.

xii

DEDICATION

My research is dedicated to all fish biologists and aquatic ecologists, with reverence for those who came before me and philanthropy to those who will come after me.

xiii

ACKNOWLEDGEMENTS

First and foremost, I thank my children, my husband, my parents, and friends for being supportive and patient while I have worked on this degree for eight years. The

time that I have taken from my loved ones will be made up for – especially by means of

enjoying the outdoors.

This journey would never have begun without Dr. Ronald Bruch, my supervisor

in the WDNR Fisheries Management Program. I am greatly indebted to him for support,

encouragement, and invaluable advising all along the way. I am grateful to have and

distinguished committee members Dr. Rebecca Klaper, Dr. John Berges, Dr. Jerry

Kaster, Dr. Derek Ogle, and Dr. Filipe Alberto. Other important role models to me have been Dr. Steven Campana, Bedford Institute of Oceanography, Fisheries and Oceans

Canada, who was a vital part of validating freshwater drum age; Dr. George Spangler,

University of Minnesota, St. Paul, MN, who helped me understand the value of growth

chronologies; and his former students, Dr. Donald Pereira and Lyn Annette Bergquist,

whose doctoral research provided the foundation that paved the way for my research.

Special acknowledgement goes to Dr. Jeffery Behm and Richard Mason, University of

Wisconsin Oshkosh, for entrusting me with archaeological artifacts. I also want to thank

my WDNR coworkers: Kendall Kamke, Robert Olynyk, Arthur Techlow III, Douglas

Rinzel, Jack O’Brien, Scott Koehnke, Ryan Zernzach, Cory Wienandt, Ryan Koenigs,

Danielle Sippel, Hilary Meyer, Seth Herbst, Rebecca Pawlek, Brian Finch, and numerous

dedicated volunteers for helping collect data and for their consideration in giving me time

and space to do my research. Finally, I want to thank my friends and colleagues at the

Great Lakes WATER Institute, Milwaukee, WI, and the University of Wisconsin

Oshkosh for their amazing support. I am fortunate to have such great comrades.

xiv 1

CHAPTER ONE: INTRODUCTION TO USING A BIOLOGICAL

CHRONOMETER TO GAUGE ENVIRONMENTAL CHANGE

The science of ecology concerns the study of organisms and their interrelationships with the environment. On some temporal and spatial scale, every ecosystem on earth has variations in abiotic factors such as temperature, precipitation, and elemental abundances and in biotic factors, particularly the relative abundances of the species that make up the community. Looking at ecology from two perspectives, the interrelationships include how an organism influences their environment, including both biotic and abiotic factors, and in turn, how these factors influence the organism. To fully understand the great complexity of an ecosystem then, an ecologist must obtain data on both the organism and many factors in their environment.

The growth of an organism has the potential to reflect some of the great multitude of factors that may change within an environment (e.g., Black et al. 2008; Coffin et al.

2003; Drake et al. 2002; Jones et al. 1989; LeBreton and Beamish 2000; Marchitto et al.

2000; Ostazeski and Spangler 2001; Rhoads and Lutz 1980; Rutherford et al. 1995; Rypel

2009; Schöne et al. 2004; Weatherley et al. 1987), and therefore can provide clues regarding population and community statistics. For example, body condition is inherently linked to reproductive fitness (e.g., Foster et al. 1999) and population growth

(e.g., Caswell 1978), and therefore these traits have strong potential to correspond to the same environmental factors, unless demographic statistics such as survival rate are compromised by other factors. Therefore, if population statistics of a species has been, or

2 will be, impacted in some way by factors such as invasive species, climate change, exploitation, or water quality parameters related to industrial, residential or agricultural inputs; it is likely these changes will be reflected in individual growth patterns.

Understanding the linkages between growth rates and environmental variables allow inferences to be made about the present state of an ecosystem, and relative abundances of future community assemblages can be predicted in response to impending environmental changes.

Lake Winnebago, Wisconsin, is part of a large ecosystem that has continually faced common anthropogenic pressures particularly since European settlement in the

1800’s. Some of the major events that have occurred in the recent history of this ecosystem include the manipulation of natural water levels, nutrient and chemical inputs from agriculture and urban sprawl, the removal of millions of pounds of rough fish species for several decades, and the arrival of aquatic invasive species (AIS) including the common carp (Cyprinus carpio ) and the zebra mussel ( Dreissena polymorpha ) (e.g.,

Kahl 1993; Otis 1988; Wisconsin Department of Natural Resources 2004). We have little

information about how the fish community was impacted by past influences, while future

changes are anticipated. Current global and regional trends suggest that urban sprawl

will continue, more invasive species will become established, the climate will warm, and

nutrient loading will fluctuate. These forthcoming changes are reasons for urgent

assessment of baseline fish community statistics and continuous monitoring programs.

Continuously monitoring the effects of environmental change on an ecosystem provides essential information regarding its health and managerial needs, but logistic

limitations make it difficult to impossible to run the necessary data collection programs,

3 especially on a large or widespread system. A longlived species with indeterminant growth and a chronological growth record can serve as an important tool for detecting longterm ecosystem changes (e.g., Drake et al. 2002; Helama et al. 2006; Pereira et al.

1995a; Rypel et al. 2009; Schöne et al. 2004; Strom et al. 2004). Freshwater drum

(Aplodinotus grunniens Rafinesque) fits these criteria. Further, drum have persisted as the fish with the greatest biomass in Lake Winnebago since the earliest management records (Becker 1964; Wirth 1958; Wisconsin Department of Natural Resources 2004), and their high abundance inherently makes them an important component in the energy flow of this ecosystem.

The specific goal of my research is to understand past relationships between freshwater drum growth and environmental factors in the Lake Winnebago system.

Assessing fish population statistics such as growth and mortality rates requires the use of age data, so it is critical that age determinations are accurate. Therefore, it is necessary to work with a validated age structure, which is consequently the first objective of this research (DavisFoust et al. 2009; Chapter 2). The second objective is to examine long term changes in population statistics of freshwater drum from pre to postEuropean settlement (Chapter 3). The age structure, growth patterns, and mortality rates of archaeological freshwater drum are compared to modern drum, resulting in a comprehensive analysis of changes in their growth patterns over time. The third objective is to examine how growth rates of freshwater drum have corresponded to environmental factors since 1950 (Chapter 4). In this chapter, a growth chronology is developed from the otolith increment measurements to make key associations between annual growth rates and environmental factors. The final objective of my research is to

4 bring all this information together to make recommendations for future monitoring programs and management decisions (Chapter 5). Understanding the pressures that influence the largest component of the fish community in Lake Winnebago may resolve unanswered questions about past anthropogenic impacts and predict future changes to this ecosystem.

In order to gauge the effects of environmental change, one has to fully understand the organism and its environment. The remainder of this chapter provides a generous review of the Lake Winnebago system and freshwater drum for the purpose of considering the many possible factors that may have impacted, or will impact, drum growth in the past, present, and future.

Background of the Lake Winnebago System

Much data has been accrued on the Lake Winnebago system, which has been studied and managed since at least the mid 1800’s. Lake Winnebago covers 55,728 hectares, making it the largest inland lake in Wisconsin (Wisconsin Department of

Natural Resources 2004). It comprises the largest part of the Winnebago system and is part of the FoxWolf Watershed, which includes all the tributaries to the Wolf and upper

Fox Rivers ( Figure 1.1 ). It is surrounded primarily by the agricultural land of central

Wisconsin. From the Wolf and Fox Rivers, water flows through three upper river pool lakes of Lake Butte des Morts, Lake Poygan, and Lake Winneconne before reaching

Lake Winnebago. The total drainage area of the FoxWolf watershed is approximately

16,576 km 2, which is almost 17% of the area of Wisconsin. For its size, Lake Winnebago

is a relatively shallow lake with a maximum depth of 6.4 meters and an average depth of

4.7 meters, having a flat bottom predominantly composed of organic silt or sand. A few rock reefs exist on the northnortheast end of the lake and along the west and east shores.

Lake Winnebago is ultimately connected to the Lake Michigan via a series of 17 locks and 12 dams on the lower Fox River.

Figure 1.1. Lake Winnebago is part of the FoxWolf Watershed, which covers

approximately 17% of the land area of Wisconsin (map modified from Wisconsin

Department of Natural Resources 1989). The map on the left depicts the Fox River

(dotted area) and the Wolf River basins (striped area).

Lake Winnebago, along with the upper and lower Fox Rivers, the Wolf River, and

the upper river pool lakes, are considered a vitally important water resource for the region

(Wisconsin Department of Natural Resources 2004). The Winnebago Pool lies within the

Fox Valley where over two million people live within 120 km of the lakes. Lake

Winnebago supplies municipal drinking water to the cities of Appleton, Menasha,

Neenah, and Oshkosh. Many cities and sanitary districts directly discharge wastewater effluent into Winnebago waters. Despite regulated monitoring, there is always the threat of wastewater effluent contamination (J. Kaiser, Chemist at Fond du Lac Water Pollution

Control Plant, pers. comm.). In contrast to the good health of Lake Winnebago, the lower

Fox River downstream is a U.S. Environmental Protection Agency areaofconcern

(AOC) that is lined with pulp and paper mills, which use the water for cooling, processing, and power. The sport fishery of Lake Winnebago was estimated to provide

$234 million to the counties surrounding the Lake Winnebago system in 2005

(University of Wisconsin Extension 2007). The diversity of sport fisheries most notably includes walleye ( Sander vitreus ), white bass ( Morone chrysops ), and the largest harvestable lake sturgeon (Acipenser fulvescens ) population in North America (Bruch et al. 2009). Further, it has fewer consumption advisories than downstream in the lower

Fox River and Green Bay region (Wisconsin Department of Natural Resources 2006), and it has substantially fewer exotic species than the Great Lakes.

Even with its confirmed economic importance, Lake Winnebago still must endure many continuing anthropogenic pressures ( Figure 1.2) . Prior to European settlement, this lake was relatively clear and mostly consisted of large riverine marshes dominated by floating and emergent macrophytes (Sloey 1970; Thompson 1959; Wisconsin

Department of Natural Resources 1993), but in modern times Lake Winnebago has been categorized as hypereutrophic and turbid (Lillie and Mason 1983). The open marsh situation was lost due to unnaturally high and constant water levels maintained by the impoundment of the lower Fox River in the 1850’s and also from inadequate stormwater management and erosion control practices (Wisconsin Department of Natural Resources

1993). It took several decades of sedimentation and constant high water levels to realize the total amount of wetland loss; large bogs of vegetation that had floated down from the upper river pool lakes were reported in Oshkosh in the early 1900’s. The vegetation loss propagated itself; the perimeters of the lakes expanded and wind fetch increased, which produced greater wave action. Wave action stirred up sediments and caused turbidity that further limited light penetration. This scenario has perpetually limited the growth of submergent vegetation that may restore water clarity, provide quality habitat, and ultimately restore a more productive ecosystem. In addition, rock shorelines used to control erosion are associated with decreased biodiversity in this system (Gabriel and

Bodensteiner 2012).

1850 1875 1900 1925 1950 1975 2000

1850present: Water levels maintained at unnaturally high levels (WDNR 1993) 18501937: Large scale formation and disintegration of floating bog (WDNR 1993) 19371960: Replacement of floating bog by other macrophytes (WDNR 1993) 19602000: Decline of macrophytes (WDNR, unpublished anecdotal reports) 19602000: High nutrient loading (WDNR, unpublished data) 19602000: Reduced water clarity (WDNR, unpublished anecdotal reports) 1960: Severe and prolonged flooding event (WDNR 1993) 1969: Severe and prolonged flooding event (WDNR 1993) 1973: Severe and prolonged flooding event (WDNR 1993) 1984present: Best Management Practices reduce nutrient loading (WDNR 2004) 2000present: Improved water clarity (Bruch 2008) 1900present: Carp present (WDNR memo, unpublished) 1970present: Rusty crayfish present (Hobbs III and Jass 1988) 1998present: Zebra mussels present (WDNR memo, unpublished) 19361990: State rough fish removal (emphasizing drum) (Staggs and Otis 1996) 19551966: State intensive rough fish removal period (Staggs and Otis 1996) 1966present: Lake sturgeon abundance recovering from low levels (Bruch 2008)

Figure 1.2. Approximate timings for major anthropogenic disturbances to Lake

Winnebago since the impoundment of the lower Fox River in 1850.

Water clarity has anecdotally improved in recent years (Wisconsin Department of

Natural Resources 2004). The Wisconsin Citizens Lake Monitoring program reports greater mean Secchi disk depths in Lake Winnebago (Wisconsin Department of Natural

Resources 2010) than Thompson (1959) did in the 1950’s. This improvement is also evidenced by winter lake sturgeon spearing harvest being at record highs and increased growth of vegetation in the littoral zone of the lake (R.M. Bruch, Wisconsin Department of Natural Resources, pers. comm.). The increased water clarity is mostly attributed to the implementation of nonpoint source water pollution control programs, including more stringent agricultural regulations and the implementation of Best Management Practices for Water Quality that were established in the Winnebago system in the early 1990’s

(R.M. Bruch & R.P. Olynyk, Wisconsin Department of Natural Resources, pers. comm.).

Zebra mussels, which were first observed in Lake Winnebago in 1998, may also have helped improve water clarity (e.g., Fahnenstiel et al. 1995; Holland 1993; Idrisi et al.

2001).

During the 20 th century, common carp, rusty crayfish (Orconectes rusticus ), curlyleaf pondweed ( Potamogeton crispus ), Eurasian watermilfoil ( Myriophyllum spicatum ), and zebra mussels are aquatic invasive species (AIS) that have been introduced to the Lake Winnebago ecosystem. These invasive species generally reduce biodiversity and habitat quality by replacing native species (e.g., Griffiths 1993; Wilson

2002), growing to nuisance quantities and blocking sunlight (e.g., Bolduan et al. 1994), and by uprooting vegetation and keeping sediment suspended (e.g., Miller and Crowl

2006).

The recent restoration of the lower Fox River locks for historical and commercial

9 purposes may accelerate the introduction of new AIS that presently reside in the Great

Lakes, but not in Lake Winnebago. This lock system had been closed for several decades, providing a comprehensive barrier to AIS that already plague most of the Great

Lakes aquatic community. All but one of the 17 locks of the lower Fox River, which have provided 63 kilometers of barriers to introduced species and diseases that presently inhabit the Great Lakes, are in the process of being restored after being shut down since the 1980’s (Friends of the Fox 2011; United States Army Corps of Engineers 1997). The

AIS of primary concern of being introduced into the Winnebago system include:

• Round gobies ( Neogobius melanostomus ) are small, bottomdwelling fish that

consume zebra mussels and the eggs and fry of any species within their habitat

(Chotkowski and Marsden 1999; Corkum et al. 2004).

• Sea lamprey is a predator/parasite that primarily targets fish species without

scales (Farmer and Beamish 1973; Scott and Crossman 1973) and larger over

smaller individuals (Swink 2003). The most wellknown hosts to sea lamprey do

not inhabit the Lake Winnebago system, but landlocked sea lampreys are known

to attack all but the smallest of fish species (Christie 1974). There is evidence

that sea lamprey can affect the survival, growth and condition of lake sturgeon

(Patrick et al. 2009), burbot ( Lota lota) (Stapanian and Madenjian 2007; Swink

2003), and white suckers ( Catostomus commersonii ) (Henderson 1986). In

addition, lampricide treatments are known to affect nontarget species (e.g.,

Boogaard et al. 2003; Marsden et al. 2003; Waller et al. 2003), including native

lampreys (Moser et al. 2007).

• The spiny ( Bythotrephes longimanus ) and fish hook ( Cercopagis pengoi ) water

fleas replace large quantities of zooplankton that are part of the diet of small fish.

B. longimanus has been found to alter zooplankton communities (Yan and

Pawson 1997) and C. pengoi may compete with small fish for zooplankton

(Vanderploeg et al. 2002).

• Quagga mussels ( Dreissena bugensis ) could potentially dominate the dreissenid

population as they have in other lacustrine, albeit deeper, locations such as the

Canadian shoreline of Lake Ontario (Wilson et al. 2006). Quagga mussels can

survive in a wider variety of habitats than zebra mussels (Berkman et al. 2000;

Dermott and Munawar 1993), and therefore will likely result in an increase the

total biomass of mussels. Dreissenids reduce phytoplankton (e.g., Fahnenstiel et

al. 2010), thereby are thought to indirectly affect food abundance for fish

populations (e.g., Charlton 1994; Hondorp et al. 2005).

• White perch ( Morone americana ) are an invasive fish that would likely become

an abundant predator and competitor in the Lake Winnebago system as it has in

Green Bay (Cochran and Hesse 1994). White perch are opportunistic feeders

(Stanley and Danie 1983), are ovivores, have diet overlaps at various life stages

with largemouth bass ( Micropterus salmoides ) and black crappie ( Pomoxis

nigromaculatus ), and utilization of white perch by adult piscivores is reportedly

low (Harris 2006).

There are several species of fish that are suspected to be nonnative, but are naturalized in the Winnebago system because they became established prior to fisheries management records and do not appear to have had any invasive tendencies or

11 deleterious impacts on the ecosystem. Most of these fish are native to the Mississippi

River and probably migrated to Lake Winnebago via a manmade canal known as the Fox

Wisconsin connection at Portage, which was opened in 1837 and closed in the 1960's

(Becker 1983). Flathead catfish (Pylodictis olivaris ), shortnose gar ( Lepisosteus platostomus ), bullhead minnow ( Pimephales vigilax ), pugnose minnow ( Opsopoeodus emiliae ), blackstripe topminnow ( Fundulus notatus ), western sand darter ( Ammocrypta clara ), the river darter ( Percina shumardi ), and sauger (Sander canadensis ) (Greens

1935) are all species suspected of migrating during the FoxWisconsin connection.

Archaeological studies can provide evidence of these migrations; for example, no flathead catfish skeletal remains have been reported at any archaeological site in the Lake

Michigan drainage area (R. Koziarski, UW Milwaukee, pers. comm.).

Background of Freshwater Drum

It is the wide distribution and frequent high abundance of the freshwater drum that makes it an extraordinary species. It is especially abundant in Lake Winnebago where it has the highest biomass of any fish species (Becker 1964). Freshwater drum probably inhabited Lake Winnebago since the receding of the last glaciers 10,000 years ago (Barney 1926). Barney (1926) speculated that this species originated from the Gulf of Mexico and became isolated some time prior to the last glaciations of the northern

U.S. during the Pleistocene and its range is limited by geography rather than by climate.

Today, it occurs throughout the eastern twothirds of North America, from the Hudson

Bay drainage to the Gulf States and from Montana and eastern Mexico to Pennsylvania

(NatureServe Explorer 2010), which is probably the largest latitudinal range of any

12 freshwater fish species in North America (Barney 1926; Fremling 1978). Longitudinally, freshwater drum are common to abundant in the Mississippi drainage and the drainage of all the Great Lakes except for Lake Superior and range as far south as Guatemala. Its range has also been artificially extended; freshwater drum were first introduced in the

1950’s as a game fish in Colorado (Everhart and Seaman 1971). It was also introduced accidentally in the Fox River drainage in Illinois (Burr and Page 1986), and in lakes in

Wyoming (NatureServe Explorer 2010; Stone 1995) and Colorado (NatureServe Explorer

2010). Although it is not known to be indigenous to the upper Wisconsin River, there were two reports in 1947, which were likely a result of a fish rescue and transfer operation from the Mississippi River in the 1930’s (Becker 1983).

There are thirtythree species and 18 genera of the Sciaenidae family that range around North America, and freshwater drum are the sole inhabitant of freshwater (Becker

1983). Members of this family are recognized by a high arched back, a covering of ctenoid scales, and a distinct lateral line that extends to the end of the caudal fin. They also have a thick skull with large cavities, which hold the mucous glands of their lateral line system. Their jaws contain many small, sharp teeth, and their pharynx contains flat, round teeth used for crushing and grinding ( Figure 1.3 ). They have spiny (anterior) and soft dorsal (posterior) fins that are slightly connected, and their anal fin has two spines.

Members of this family have a large sagittal otolith located in the sacculus of the inner ear. The name Aplodinotus is Greek for ‘singleback’, and grunniens refers to the sounds that this fish can produce by contracting muscles that cause a tendon to rub across on the swim bladder (Priegel 1967b). Ramcharitar et al. (2006) summarize that there are two main purposes of sciaenid sounds, which are reproductive calls and disturbance calls

(e.g. pain, alarm, annoyance).

Figure 1.3. “Molars” in the freshwater drum pharynx, which are adapted for crushing

mussel shells.

Exact biomass and population estimates of freshwater drum in Lake Winnebago

are extremely rough. Drum constitute most of the biomass of the lake (Rice 1987), but

numerically they are probably second to trout perch ( Percopsis omiscomaycus ) (Otis and

Staggs 1988). Becker (1964) notes that freshwater drum are “numerically the most successful species of large fish in Lake Winnebago,” and Becker (1983) notes that they are “abundant in Lake Winnebago and common in Lakes Poygan, Winneconne, and Lake

Butte des Morts.” A tagging program in the late 1980’s for a markrecapture study resulted in too few recaps to make a population estimate with any accuracy (Kamke and

Bruch 1991). Kamke and Bruch (1991) tried to estimate the population size of freshwater drum using Bailey’s method (Ricker 1975), which is a mark and recapture technique. Population estimates ranged from 88.6 to 154.0 million fish, but it was felt that there were too few recaps to calculate accurate numbers. There are many known drawbacks to the area swept method (Alverson and Pereyra 1969), which expands the mean calculated density from the area sampled to the total area of the lake, but it is

14 probably more accurate than the tagging study for estimating the freshwater drum population. The freshwater drum adult population estimate from the area swept method was a mean of 27.7 million (SD = 13.6) for the period 19621990 in the trawlable area of the lake (Kamke and Bruch 1991). Estimates ranged from 2.2 million in 1979 to 51.0 million in 1987. Current population estimates of drum within Lake Winnebago range from 19862010 range from a mean of 46.5 million as an October to 76.4 million in

August (Appendix A).

The history of fluctuating water clarity in Lake Winnebago could be a factor in affecting freshwater drum growth and abundance over time. Many members of the

Sciaenidae family are typically found in association with turbid environments (Rodríguez and Lewis Jr 1997). While freshwater drum occur in a wide variety of habitats throughout the middle of North America, their largest populations are reported to occur in large silty lakes and medium to large rivers (Lee et al. 1991; Page and Burr 1991).

Turbid and sluggish waters are the habitats that are most frequently associated with freshwater drum (Priegel 1967a), but they are occasionally found in clear water (Becker

1983; Trautman 1982). Given their prominent lateral line, which provides mechanosensory information (Montgomery et al. 1995), turbidity may be a factor that allows freshwater drum an advantage in certain conditions.

Ecological Roles of Freshwater Drum

From an ecological standpoint, the high biomass of freshwater drum must certainly play a key role in the nutrient cycles and energy flow of Lake Winnebago, but their specific community interactions are ambiguous. Age0 freshwater drum often make

15 up a high proportion of forage (prey) species during the summer (R.M. Bruch, WDNR, pers. comm.) and their temporal and spatial behaviors make them vulnerable to a wide spectrum of predators walleye, sauger, white bass, yellow perch ( Perca flavescens ),

largemouth ( Micropterus salmoides ) and smallmouth bass ( Micropterus dolomieu ), and

northern pike ( Esox lucius ). However, there are no formal studies of predation rates on

freshwater drum in Lake Winnebago. Becker (1983) states that freshwater drum eggs

and larvae are vulnerable, but also that freshwater drum have few natural predators, “their

eggs, larvae, and fry are seldom taken by other species.” Butler (1965) also stated that,

“drum, even when abundant, are not used to any great extent as food by other species.”

Butler (1965) and Priegel (1967b) reported that young freshwater drum are eaten

infrequently by walleye, burbot ( Lota lota ), sauger, and white bass. Staggs and Otis

(1996) found that the growth of walleye and sauger, which are two of the most abundant predatory fish in Lake Winnebago, corresponded to years of high age0 freshwater drum

or troutperch ( Percopsis omiscomaycus ) abundance. After their first year, freshwater

drum probably no longer serve as a forage base for predator fishes because of protection provided by the development of their body size.

Freshwater drum are benthic generalist feeders. Diet analyses vary substantially

among studies, and correspond to their size (e.g., Swedberg 1965), geographical location

(e.g., Priegel 1967a), and season (e.g., Griswold and Tubb 1977). In many waters,

freshwater drum feed chiefly on mollusks (Priegel 1967a), which would be expected

considering their large pharyngeal molars. Wirth (1959) found that mollusks, crayfish,

and fish were a part of the Winnebago freshwater drum diet to a “lesser proportion.”

Preigel (1967a) reported only 0.1% of the total food volume of freshwater drum in Lake

Winnebago was the fingernail clam ( Pisidium sp.), and he reported no other mollusks.

Small freshwater drum in Lake Winnebago primarily consumed minute crustaceans and gradually shifted to midge larvae, specifically Chironomus sp, by the time they reached

40 mm in length. Midge larvae accounted for 88.399.3% of the total food volume for

672 adult male and female freshwater drum. Preigel (1967a) also reported only 0.2% of

the total diet of male freshwater drum was comprised of fish (of unidentified species),

and Daiber (1952) reported similar findings of freshwater drum in western Lake Erie. In

contrast, fish (yellow perch and bluegills) were the second most important item in the

diets of freshwater drum in the lower Missouri River (Berner 1951) and in four lakes

sampled in Iowa (Moen 1955). Many of these analyses were shortterm and used

samples from either one type of sampling gear or one habitat type; therefore, the full diet

range of drum was not captured in any single study. For example, Priegel’s (1967a)

stomach analyses were conducted entirely on freshwater drum captured by trawl in 1953.

The diet of freshwater drum overlaps with a variety of gamefish species. As youngoftheyear, Winnebago freshwater drum chiefly consume copepods and cladocerans, which is similar to the diet of youngoftheyear game fish including walleye, sauger, white bass, and yellow perch (Priegel 1967a). Butler (1965) stated that freshwater drum and white bass share similar habitats in Lake Erie and had previously been found to compete for food by Daiber (1952). Freshwater drum consume leeches,

Helobdella sp., which many predator fish also consume. Chironomid larvae and gizzard shad comprise a significant portion of both freshwater drum (Priegel 1967a, pers. obs.) and lake sturgeon (Stelzer et al. 2008) diets in Lake Winnebago.

For competition to occur, there must be a limiting resource (Gause 1934), and

17 many of the abovementioned prey items have never been shown to be limiting. Staggs and Otis (1996) reported evidence of competition between freshwater drum and troutperch, but no evidence that freshwater drum competed with any game fish species.

However, recent evidence suggests that the chironomid population in Lake Winnebago does fluctuate and may be limiting at times (Anderson 2010). Stomach analysis of lake sturgeon speared in February of 2005 found relatively few chironomid larvae compared to other years of analysis (G.H. Drecktrah, UW Oshkosh, pers. comm.). Additional evidence is more anecdotal. Chironomid hatches were observed by local residents to be smaller than most could remember, and bottom samples from Lake Winnebago that spring turned up fewer densities of chironomid larvae than ever before. The 2005 lake sturgeon spearing season was nicknamed “year of the skinny fish” (Wisconsin

Department of Natural Resources 2009).

In considering all potential controllers of growth, disease and parasite loads should not be overlooked. Freshwater drum captured through DNR assessment programs occasionally appear to have internal or external parasites, and appear to be more common some years (pers. obs.). Freshwater drum are known serve as intermediate hosts for parasites including mussels (Becker 1983), flukes and roundworms (Priegel 1967b). The glochidia larvae of mussels pass through metamorphosis on freshwater drum gills, and they can carry an relatively high number of them (Surber 1912). Fuller (1974) lists several clams that use the freshwater drum as a host: Megalonaias gigantea, Anodonta grandis, Arcidens confragosa, Lampsilis orbiculata, Leptodea fragilis, Proptera alata,

Proptera laevissima, Proptera purpurata, Truncilla donaciformis, and Truncilla truncata .

Historical Management of Lake Winnebago Freshwater Drum

Early rough fish (deemed undesirable for sport fishing) removal programs were based on the premise that by removing undesirable fish species, the abundance of

gamefish would increase through reduced competition and increased living space (Harris

1964; Priegel 1965; Priegel 1971). In the late 1970’s, more support for rough fish

removal efforts came from a new theory that bottom feeding fish (mostly rough fish)

limit bound phosphorus for algae growth by consuming detritus and bottom organisms

(Sloey et al. 1978). Thus, by removing the rough fish, excess nutrients would also be

removed from the lake.

Rough fish removal in Wisconsin dates back to at least 1899 (Harris 1964) and

ended in Lake Winnebago in 1990 (Kamke and Bruch 1991). The first removal efforts

were in response to sportsmen’s complaints of too many carp and freshwater drum. In

Lake Winnebago, prior to 1936, there had been some systematic rough fish removal by private commercial fishermen under contract with the Wisconsin Conservation

Commission (Priegel 1971). State crews began working with the contracted commercial

fishermen in 1936. Early crews used hoop nets and trap nets (at all times of the year), but

later they switched to trawling along the bottom because trawling was more efficient for

capturing rough fish. An intensive freshwater drum removal program was launched in

1955, shifting removal emphasis from common carp to freshwater drum. The intensive

removal period ended in 1966 when commercial crews quit fishing due to the poor

market for freshwater drum. From 1967 to 1976, two trawlers (the Sheepshead and

Winnebago ) ran year round. A third trawler, the Calumet , was added to the fleet in 1976.

This trawler was so efficient that it increased freshwater drum harvest by 45%. Due to budget constraints, the Sheepshead and Winnebago were retired in 1980. State

freshwater drum removal operations were discontinued in 1990 (Kamke and Bruch

1991).

The costbenefit of the removal program on Lake Winnebago was repeatedly

challenged throughout the years (Kamke and Bruch 1991). Most biologists came to the

conclusion that calculating changes in freshwater drum abundance could not be made

with any reliability due to inaccuracies in the data collected (e.g., poor estimates of sizes

and number of fish caught). Inaccuracies were further exacerbated by using data from

the removal program itself, which had changing removal effort, changing capture

methods, and increased removal efficiency over time (Michael Staggs, WDNR, internal

memo 1984). Proxies were used as potential indicators of change in freshwater drum

abundance. Increasing abundances of game fish species (for which there were presumably more accurate records) would suggest decreasing freshwater drum density

(Priegel 1971). A decrease in the average age and length of freshwater drum would be an indication of an increase in density, assuming that growth rates remained the same

(Rounsefell and Everhart 1953). Condition factors were calculated because an improvement in condition would also be expected when a population is at a lower density.

Early assessments of the Lake Winnebago removal program suggested changes in the fish community, but later studies determined that most findings were either inaccurate or not consistent. The first major assessment was completed by Priegel (1971), who was a big proponent of the program. His study covered the 12year “intensive removal

20 period” from 19551966. He found that while growth rates of freshwater drum remained relatively stable, there was a decrease in the age composition. The problem with this conclusion is that age determinations were made from scale samples, which are now invalidated for drum over age four years (DavisFoust et al. 2009). Priegel (1971) also reported that the proportion of freshwater drum captured each year after 1959 had decreased, while the proportion of gamefish netted distinctively increased. However, fishing effort constantly fluctuated, the primary gear used changed from trap nets to trawl nets in 1962, and part of the decline of netted game fish in 1964 was attributed to improvements to eliminate the handling of game fish. Priegel (1971) had first detected an improvement in condition factors at the beginning of the study, but later this improvement was lost. One notable finding in the Priegel (1971) report is that mean lengths of drum captured from trap nets had dramatically declined after 1962. Chapter 3 of this dissertation examines a potential shift in drum length frequencies from pre

European settlement to present day.

During the intensive removal period, approximately 1.3 million kg (2.8 million lbs) of freshwater drum had been removed per year (Priegel 1971). Priegel estimated that to keep the Winnebago freshwater drum population at an optimum abundance, approximately 1.1 to 1.4 million kilograms (2.5 to 3.0 million lbs) needed to be removed annually. Despite Priegel’s removal recommendation, the intensive removal period ended in 1966 due to reduced funding. During the next 14 years (19671981), it was estimated that an average of only 0.5 million kg (1.2 million lbs) per year were removed

(Otis 1988).

Otis (1988) evaluated the effects of the removal program from 196781. His

21 report was based on mean catch per haul, lengths and weights of individual freshwater drum, and scalederived age data from 19671981 (there was no age data in 1970, 1972, and 1974). Data analysis over this time period was difficult because two smaller towing vessels were replaced with a larger stern trawler in 1976, the Calumet . Thus, catch per unit effort was lower from 19671975 than from 19761981 because longer towing duration and larger trawl sizes resulted in higher mean catch per haul. A correction factor was applied, and Otis (1988) determined that overall freshwater drum growth and condition during 19671981 were significantly better than during the years of intensive removal. At first this finding seems contrary to the expected effects of removal trawling, but because fish from age three to five had better condition factors than older fish, and because removal gear selected larger (and presumably older) fish, the change in overall condition factors could be a reflection of a decline in the age structure. Otis (1988) concluded that the annual removal of 0.5 million kg of freshwater drum from 19671981 was not enough to significantly affect freshwater drum abundance. He also concluded that a separate randomized sampling design would be necessary to adequately assess the impacts of removal trawling. Removal trawling was geared toward maximizing the number of freshwater drum removed and minimizing the numbers of game fish removed, and therefore provided biased data. He also recommended the use finer mesh nets to capture younger freshwater drum and game fishes. One serious shortcoming of the Otis

(1988) study is that his conclusions were primarily based on changes in growth and mortality rates. These statistics rely on accurate age determinations, but age determinations for fish over age 4 in this study were inaccurate because they were read from scales (DavisFoust et al. 2009). Otis (1988) did not examine the data for shifts in

22 body length frequencies over time.

The third and last major assessment of the removal program was conducted by

Kamke and Bruch (1991) This study was based on removal trawling conducted from

19821990 during which a mean of 0.7 million kg (1.5 million lbs) of freshwater drum were removed each year and data from standardized trawling (assessment trawling). The objectives were to determine if freshwater drum removal at Priegel’s recommended annual rate of 1.1 to 1.4 million kilograms for 58 years would 1) measurably enhance sport fish populations, 2) substantially decrease freshwater drum biomass, 3) alter the freshwater drum population structure, and 4) affect the abundance of forage and non game fish species.

The critical problem with conducting this assessment was that the goal of 1.1 to

1.4 million kilograms of freshwater drum removed per year was never reached. The highest removal rate reached was 1.0 million kg (2.1 million lbs) in 1990. They analyzed the data anyway. Correction factors had to be used to link the data sets between the removal trawler, the Tarred Twine , and the assessment trawler, the Winnebago . They

concluded that the removal program was not impacting the freshwater drum population to

any appreciable amount due to the size of the lake and the resiliency of the species, and:

1. freshwater drum do not appear to have a negative impact on any life stage of

walleye (also may be change in relation to gizzard shad (Dorosoma

cepedianum ) abundance, which is a primary food resource of walleye);

2. adult walleye abundance fluctuates independently of both adult freshwater

drum biomass and freshwater drum year class strength [later contradicted by

the findings of Staggs and Otis (1996)];

3. freshwater drum in Lake Winnebago have a 68.5% annual mortality rate, of

which less than 10% is attributable to removal operations [based on inaccurate

age structure data (DavisFoust et al. 2009)], and

4. freshwater drum mortality rates and growth rates have not changed

significantly from 1955 to 1990 [also based on inaccurate age data].

There were many reviews (e.g. David Hildreth, Lee Kernen, James March, Lee Meyers,

Keith Otis, Gordon Priegel, Fish Biologist of the WDNR) of the Kamke and Bruch draft

report. In summary, the report was criticized for reasons such as inaccurate correction

factors used to make direct comparisons among the data sets, being statistically unsound,

and that Priegel’s recommendation for the necessary removal quantity was not attained.

The impacts of removal trawling were never fully agreed upon, and it appeared

that any impacts would remain undetermined based on existing data. Despite public

support to continue rough fish removal efforts, evidenced by editorials in the Oshkosh

Northwestern and other local newspapers throughout the 1980’s, the program was

terminated in the fall of 1990. Fall assessment trawling was initiated in 1986 and remains

in operation today. It was designed to assess the fish community of Lake Winnebago and

does not remove any appreciable numbers of freshwater drum.

CHAPTER TWO: AGE VALIDATION OF FRESHWATER DRUM USING

BOMB RADIOCARBON

Introduction

Estimates of fish age are the foundation for understanding and forecasting

fisheries population dynamics. Accurate age estimation is critical for correctly

calculating age structure, growth rates, survival, mortality rates, and age at maturity

(Campana 2001; Hoxmeier et al. 2001; Ricker 1975) The use of aging structures needs to be validated for all ages of a species since the frequency of increment formation may

change during a fish’s life history (Beamish and McFarlane 1983; Campana 2001).

Freshwater drum Aplodinotus grunniens , the only freshwater member of the

family Sciaenidae, has the broadest latitudinal distribution of any freshwater fish species

in North America (Rypel et al. 2006; Stewart and Watkinson 2004). Although not

considered a sport fish, its common abundance makes it a significant member of many

fish communities as a forage fish and multilevel predator. Stomach content analysis has

revealed that small freshwater drum are eaten by walleye Sander vitreus , burbot Lota

lota , sauger S. canadensis and white bass Morone chrysops in the Mississippi River and

Lake Winnebago, Wisconsin (Butler 1965; Priegel 1963a; Priegel 1967b). Walleye and

sauger, two of the most abundant piscivorous fishes in Lake Winnebago, had greater

growth rates during years of high age0 freshwater drum abundance, strongly suggesting

that drum are an important prey item (Staggs and Otis 1996). Freshwater drum are benthic generalist feeders, and larger individuals will consume small fish (Becker 1983;

Daiber 1952). In Lake Winnebago, midge larvae (Chironomidae) have historically been

a primary component of freshwater drum diets (Priegel 1967a), making this species a

25 potential competitor with lake sturgeon Acipenser fulvescens , which are known to also depend heavily on chironomids (Choudhury et al. 1996; Stelzer et al. 2008).

Before 1994, most of the published demographic parameters for freshwater drum, such as growth rates, mortality rates, and age at maturity, were based on age estimates from scales (e.g., Becker 1983; Bur 1984; Daiber 1953; Edsall 1967; Houser 1960;

Klaassen and Cook 1974; Priegel 1969; Schoffman 1940; Van Oosten 1938; Wrenn

1968). After 1994, sagittal otoliths became the most frequently used structure (e.g.,

Palmer et al. 1995; Pereira et al. 1995a; Pereira et al. 1995b; Rypel 2007; Rypel et al.

2006), although some studies continued to use scales (e.g., Braaten and Guy 2004; Phelps et al. 2000). Ages estimated from otoliths have been used to determine maturation rates palm (Palmer et al. 1995), compare growth rates of freshwater drum from different habitats (Rypel et al. 2006), and detect sexual dimorphism (Rypel 2007). In addition, widths of otolith growth increments have been used in biochronological studies to investigate the influence of environmental conditions (Pereira et al. 1995a; Pereira et al.

1995b) and community interactions (Ostazeski and Spangler 2001) on growth.

Despite their widespread use in estimating age, scales and otoliths of freshwater drum have not been validated as aging structures, thus weakening the credibility of studies that have used age estimates from these structures. Timing of annulus formation on freshwater drum scales has been evaluated, but without validation of age estimates

(e.g., Edsall 1967; Swedberg 1965; Wrenn 1968). Goeman et al. (1984) reported age validation of freshwater drum using sagittal otolith age estimates to follow the progression of individuals in a strong yearclass for three consecutive years, but this methodology is considered age corroboration, which can support but not replace age

26 validation (Campana 2001). Also, anal and dorsal spines of freshwater drum have not been evaluated as valid aging structures.

The most unambiguous method for validating the periodicity of growth increments is using fish of known age (Beamish and McFarlane 1983; Campana 2001;

Casselman 1987). This method is more difficult, though, for large populations of fish, fish located in larger water bodies, and longlived fishes. Further, rough fish species like freshwater drum are generally considered undesirable and typically do not receive the attention nor the funding needed for intensive studies.

For longlived fishes, the best alternative method for validating age estimates is assaying the cores of their otoliths for atomic bomb radiocarbon (Campana 2001). The thermonuclear bomb testing era in the 1950s and 1960s resulted in a spike in the quantity of radiocarbon ( 14 C) in the earth’s hydrosphere, leaving a detectible temporal signature in

otoliths and other calcified structures of organisms living in that era. Bomb radiocarbon

dating does not use radioactive decay, as does the traditional method of radiocarbon

dating; rather, it is a measure of the change in atmospheric radiocarbon that was released

from atmospheric bomb testing and incorporated into carbonbased structures of growing

organisms. Subsequently, bomb radiocarbon dating is best used on structures that (1)

were growing during the bombtesting era, (2) have visible growth increments from

which to estimate age, (3) are metabolically inert after carbon deposition, and (4) provide

enough material for 14 C assay (minimum of 3 mg). When large enough, the cores of fish

otoliths that formed during the bombtesting period meet these criteria.

Bomb radiocarbon dating has been completed on the otoliths of numerous marine

and semimarine species but relatively few freshwater species. Marine species include

27 haddock Melanogrammus aeglefinus (Campana 1997), red snapper Lutjanus

campechanus (Baker Jr and Wilson 2001), gray snapper L. griseus (Fischer et al. 2005),

and canary rockfish Sebastes pinniger (Andrews et al. 2007; Piner et al. 2005).

Semimarine species include black drum Pogonias cromis of the Chesapeake Bay region

(Campana and Jones 1998), which reside in estuarine waters during their first year, and

Arctic char Salvelinus alpinus , which is an anadromous species (Campana et al. 2008).

The only entirely freshwater species that have been assayed using this approach are lake trout S. namaycush (Campana et al. 2008) and lake sturgeon (Bruch et al. 2009).

The objectives of this study were to (1) validate the age of freshwater drum with bomb radiocarbon analysis of sagittal otolith cores and evaluate the accuracy of the otolith age estimates; (2) support the age validation with corroboratory evidence on drum yearclass strength; and (3) evaluate the accuracy of age estimates derived from freshwater drum scales, anal spines, and dorsal spines.

Methods

Study Site .—Lake Winnebago, at 55,728 ha, is the largest inland lake in

Wisconsin (WDNR 2004), with a maximum depth of 6.4 m and an average depth of 4.7

m. It is part of the eutrophic Winnebago–Upper Fox–Wolf watershed and is connected to

the Great Lakes at Green Bay by the lower Fox River. Of the at least 76 species of fish

found in the Winnebago system, freshwater drum have historically been estimated to

consistently have the highest biomass (Priegel 1967; Staggs and Otis 1996).

Sampling .—We sampled freshwater drum captured during Lake Winnebago

assessment trawling in October of 1986 and 2003–2007, during Winnebago system

28 fishing tournaments in July 2003 and 2006, and following underwater blasting events as part of a bridge construction project on the Fox River between Lakes Winnebago and

Butte des Morts in April 2007. Trawling was conducted during daylight hours in August,

September, and October with a 7.9mhead rope bottom trawl with a 3.8cm stretch mesh body and a 1.3cm stretchmesh cod end liner. The trawl was towed at 6.4–7.2 km/h for 5 min in each of four to seven randomly selected 1´latitude X 1´longitude sampling grids within each of five Lake Winnebago sampling areas.

Otoliths were collected from trawlcaptured freshwater drum in stratifiedrandom subsamples in 1986 and 2003–2006 (15 per 25.4mm length interval), from random subsamples in 2007 (standard volume per cast), from drum greater than 457 mm during fishing tournaments in 2003 and 2006, and from all drum collected following underwater blasting in 2007. Scales were collected in 1986 and from trawl sampling in 2003. Dorsal and anal spines were collected from trawl sampling in 2003. Scales were removed from midway between the lateral line and the midbase of the spiny dorsal fin, and the second spines of the dorsal and anal fins were cut at their bases using surgical nail nippers. All drum sampled were measured to the nearest 2.5 mm total length (TL) and weighed to the nearest gram. Sex and maturity were discerned for all trawlsampled drum in 2007. Fish with oocyte or general testes development were considered mature.

Otoliths were washed to remove all adhering tissues and prepared for sectioning by being embedded in EpoKwick fastcure epoxy resin to prevent fracturing while being

cut. Two to four transverse sections, 0.25– 0.48 mm thick, were cut through the core of

each otolith with an Isomet lowspeed diamond blade saw and mounted on glass

microscopy slides with cyanoacrylate glue for viewing and storage. Spines were cleaned

29 to remove excess soft tissue, and a 0.25–0.48mm section was cut with the saw through the basal portion of each. Otolith and spine sections were examined through an Olympus

SZX7 dissection microscope equipped with an Olympus DP71 camera using a combination of transmitted and reflected light after either mineral oil or ethanol was applied for clarification. Opaque zones were considered the boundaries of annual growth increments (Casselman 1987). Scales were soaked in water, cleaned with a brush, and viewed on a microfiche projector at 40X magnification. Annual growth increments on scales were defined as continuous opaque zones.

Age Validation Using Bomb Radiocarbon Dating.— Eighteen otoliths were selected from specimens taken during 1986 and from 2003 to 2006 with an estimated year of core formation between 1948 and 1980 to measure radiocarbon values from before, during, and after the bombtesting era. One otolith of the original pair was processed for age estimation as described previously and the second otolith was sectioned, aged, and micromilled after being embedded in a hard epoxy (Araldite epoxy

GY502 and hardener HY956 in a 5:1 weight ratio). Three adjacent 1mmthick transverse sections through the core of the second otolith were cut using multiple blades on an Isomet lowspeed diamond blade saw and lightly polished to improve clarity. The growth increment sequence was examined and digitally photographed at 16–40X magnification with reflected light, at a resolution of 2,048 3 2,048 pixels, and then digitally enhanced with Adobe Photoshop CS2 to improve contrast. Age estimates were based on the enhanced images, and aging precision was quantified with coefficient of variation (CV = 100 X SD/mean; Campana 2001.)

Otolith cores representing what was assumed to be the first year of life were

30 isolated from the central section of each otolith as a solid piece with a Merchantek computercontrolled micromilling machine with 300µmdiameter steel cutting bits and burrs. The assumed date of core sample formation was calculated as the year of fish collection minus the number of growth increments between the otolith edge and onehalf way along the growth axis of the extracted core. After sonification in Super Q water

(Millipore Corporation) and drying, the sample was weighed to the nearest 0.1 mg in preparation for 14 C assay with accelerator mass spectrometry (AMS). The AMS assays

also provided δ 13 C (‰) values, which were used to correct for isotopic fractionation

effects. Radiocarbon values were subsequently reported as 14 C, which is the per mille

(‰) deviation of the sample from the radiocarbon concentration of 19thcentury wood, corrected for sample decay before 1950 according to methods outlined by Stuiver and

Polach (1977).

The feature of a bomb radiocarbon chronology that is most stable across locations and environments (and thus most useful as a dated marker) is the year of initial increase above prebomb levels in response to the period of atmospheric testing of nuclear weapons. Campana et al. (2008) demonstrated that a 14 C value 10% above the prebomb background is a robust and accurate indicator of the year of initial appearance of bomb

14 C, and one that is consistent with atmospheric sources. Therefore, we estimated the

14 value corresponding to the 10% threshold contribution of C ( CT) by calculating 90%

14 of the range in C between its lowest ( CL) and peak ( CP) values and subtracting it from

the peak value, that is,

CT = CP – 0.9( CP – CL)

where CL occurs on or after 1952, the year of initial release of bomb radiocarbon into the

14 atmosphere. The year of initial appearance of bomb C ( YT) is then defined as the year

14 in which the C chronology first exceeds CT. To further substantiate the calculated

year of initial rise, a second method by Kerr et al. (2004) was used, whereby the year of

initial rise is the year that is significantly greater (±2 SD) than the mean prebomb level.

Age Corroboration .—Catchperuniteffort (CPUE) indices of the age0 year class abundance of freshwater drum were collected from 1962 to 1984 with experimental trawling conducted during daylight hours from June to November with a 3.7mheadrope bottom trawl with a 3.8cm stretchmesh body and a 0.65cm stretchmesh cod end liner towed at 6.4–7.2 km/h for 7 min in various locations, primarily along the west shore of

Lake Winnebago. The average CPUE (number of age0 freshwater drum per trawl cast) was calculated for each year by averaging the CPUE’s over all casts during the months from August to October, and this was used to document strong hatches of drum that might show up in trawlcaught age frequencies during subsequent years.

Age frequencies based on otolith age estimates of freshwater drum caught in assessment trawls on Lake Winnebago were calculated from pooled data in 1986 and

2003–2007 and examined for progressions of strong yearclasses over the 21year period.

Aging accuracy of alternative structures .—Two experienced readers independently estimated the age of each fish by counting the number of visible growth increments on the scales and otolith sections; one experienced reader examined spine sections. Coefficients of variation were calculated for scales and otoliths between the two readers to examine the degree of agreement.

Results

Age Estimation

We collected pairs of sagittal otoliths from 1,361 freshwater drum—1,170 (287 in

1986, 154 in 2003, 127 in 2004, 107 in 2005, 110 in 2006, and 385 in 2007) captured

during trawling on Lake Winnebago, 121 obtained from fishing tournaments in 2003 and

2006, and 70 collected following underwater blasting events in 2007. From the 1,361 pairs of otoliths collected, we estimated ages for 1,351 freshwater drum ranging from age

0 (61 mm) to age 58 (599 mm). Male drum began to mature at age 2 (226 mm), and

female drum began to mature at age 5 (272 mm). Ten otoliths were not readable due to

faulty sectioning, and one otolith was rejected due to a structural deformity making the

growth increments indistinct. Growth increments on otoliths were clear and easily

interpretable. The betweenreader CV was 0.7%.

Age Validation

Micromilling removed the first 1 to 3 years growth of the freshwater drum otolith

cores, which provided adequate sample masses for AMS assays ranging from 5.1 to 33.4

mg. Bomb radiocarbon values of otolith cores (as 14 C) ranged from 124.0 to 234.2 producing a classic 14 C curve that correlated well with known 14 C reference chronologies ( Table 2.1; Figure 2.1). Bomb radiocarbon 14 C values showed a sharp

increase beginning in 1957, a peak value of 234.2 in 1965, and a steady decline to the

most recent sample of 1983. Based on the equation by Campana (2008), 1957 was the

initial year of increase in 14 C above prebomb levels in the Lake Winnebago freshwater drum chronology. Using the methods of Kerr et al. (2004), the initial year of increase above the mean prebomb level was 1956. Since consistent under or overaging of the

33 freshwater drum otolith growth sequences would have phaseshifted the entire freshwater drum bomb chronology to the right or left, the close correspondence of the freshwater drum and reference chronologies, and the similarities in their calculated initial years of increase, indicate that the freshwater drum otolith growth increments provide accurate estimates of age. Estimated otolith ages of the drum sampled for 14 C ranged from 17 to

52 years. The betweenreader CV of the age estimates for the 18 otoliths assayed was

1.72%. The mean SD of the individual radiocarbon assays was approximately 5‰.

Table 2.1. Collection year, core weight, age (based on otolith growth increments), yearclass (based on otolith age), and 14 C and δ 13 C assay values for freshwater drum sagittal otolith cores sampled from Lake Winnebago, Wisconsin, in 1986, 2003, and 2006.

Collection Core weight year (mg) Age Yearclass 14 C (‰) δ13 C (‰) a 2006 11.64 52 1954 117.1 9.69 2003 27.42 49 1954 116.8 10.2 1986 8.81 31 1955 124 11.7 1986 5.87 29 1957 82.7 10.05 1986 5.11 29 1957 42.7 8.19 1986 7.82 27 1959 38.9 9.72 1986 6.82 26 1960 113.2 8.71 1986 7.24 24 1962 121 9.31 1986 7.52 23 1963 112.4 9.6 2006 8.37 43 1963 113.5 8.84 1986 7.18 22 1964 215.9 8.98 2006 7.39 41 1965 234.2 10.41 1986 5.76 20 1966 217.1 12.63 2003 33.39 37 1966 219.8 12.4 1986 7.04 18 1968 181.4 10.14 1986 9.3 17 1969 170.6 10.42 2003 5.07 29 1974 102.3 b 2006 9.01 23 1983 77.6 12.21 a δ 13 C was used to correct for isotopic fractionation to derive 14 C, which is he per mille (‰) deviation from the radiocarbon concentration of 19 th century wood (see text). b Data not available.

1000

∆1 800 4C

600 C 14

400

200

200 1950 1960 1970 1980 1990

Year of formation

Age Corroboration

Examination of freshwater drum yearclass strength from experimental trawling samples of age0 fish from 1962 to 1984 showed a very strong yearclass in 1983 ( Figure

2.2 ). This strong yearclass was consistently and clearly the most abundant for each

sampling year in 1986 and 2003–2007 based on otolith age estimates ( Figure 2.3 ). The

interpretation of Figures 1.2 and 1.3 thus corroborates otolith age estimates up to age 24,

35 since this ageclass was still abundant in the 2007 survey year.

300

200

100 C.P.E. (number/cast) C.P.E. CPUE (number/cast)

0 1962 1963 1964 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 Year Year

Figure 2.2. Catch per unit effort (CPUE) of age0 freshwater drum from experimental trawl assessments, Lake Winnebago, 1962–1984.

1986 (n=283) 2005 (n=106)

50 * 40 40 30 30 20 20 * 10 10 Frequency (%) 0 Frequency (%) 0 0 5 10 15 20 25 30 35 40 0 5 10 15 20 25 30 35 40 Age class (years) Age class (years)

2003 (n=197) 2006 (n=178)

40 * 40 * 30 30 20 20 10 10

Frequency (%) 0 Frequency (%) 0 0 5 10 15 20 25 30 35 40 0 5 10 15 20 25 30 35 40 Age class (years) Age class (years)

2004 (n=125) 2007 (n=452)

30 20 * 25 * 15 20 15 10 10 5 5 Frequency (%) 0 Frequency (%) 0 0 5 10 15 20 25 30 35 40 0 5 10 15 20 25 30 35 40 Age class (years) Age class (years)

Aging Accuracy of Alternative Structures

Growth increments could be seen on scales and the basal sections of anal and

dorsal spines, although they were not as clearly distinguishable as those on otoliths

(Figure 2.4 ). In addition, the lumen of both anal and dorsal spines was often

37 deteriorated, particularly on specimens taken from older individuals. Age bias plots revealed similar relationships between otoliths and scales, anal spines, and dorsal spines

(Figure 2.5 ). The age estimates from otoliths begin to exceed estimates from the three other structures starting at a length of about 280 mm (age 3). After this point, the otolith age estimates continue to diverge from the scale age estimates by as much as 36 years.

The average CV between otoliths and scales was 43.3%, that between otoliths and anal spines 46.5%, and that between otoliths and dorsal spines 49.0%. Log10transformed otolith ages and log10transformed scale ages were significantly correlated ( r2 = 0.71, P

< 0.00001, n = 475) in a linear regression, although when only data from fish 10 years and older were used in a regression, much less of the variance was explained (r 2 = 0.23, P

< 0.00001, n = 267). The two correlation coefficients were significantly different ( Z =

9.45, P < 0.0001) using the statistical test recommended by Zar (1996).

A B

C D

80 75 70 65 60 55 50 45 40 35 30 25 Scale age determination age Scale 20 15 10 5

Scale age estimation 0 0 5 10 15 20 25 30 35 40 45 60

55 Otolith age determination 50

15 Anal spine age determination age spine Anal 10 Anal spine age estimation 5 0 0 5 10 15 20 25 30 35 40 45 90 85 Otolith age determination 80 75 70 65 60 55 50 45 40 35 30 25 Dorsal spine age estimation 20 Dorsal spine age determination age spine Dorsal 15 10 5 0 0 5 10 15 20 25 30 35 40 45

Otolith age determination Otolith age estimation Figure 2.5. Otolithestimated ages of freshwater drum from Lake Winnebago in 2003 versus the mean age estimated from scales, anal spines, and dorsal spines. All values are years; the error bars represent 95% confidence intervals. Values that fall on the 1:1 diagonal lines represent full agreement between the respective structures.

Discussion

The onset and peak of the freshwater drum 14 C chronology for Lake Winnebago

closely reflects other published 14 C values for freshwater and marine fish species, thus validating otoliths as an accurate aging structure to 52 years with an error of no more than

±2 years. Compared with the knownage chronology from Canadian Arctic char and lake trout (Campana et al. 2008), the freshwater drum chronology begins to increase the same year, and peak values occur within 2 years of each other. The black drum chronology from Chesapeake Bay (Campana and Jones 1998) is the most similar to that of the freshwater drum, with an identical initial year of increase of the 14 C value 10% above the prebomb background and a slightly lagged peak. The peak freshwater drum 14 C

values lag slightly behind the atmospheric chronology (Nydal 1993), which would be

expected. During the years after the peak, the freshwater drum 14 C values are at levels

quite similar to those of the gray snapper from the Gulf of Mexico (Fischer et al. 2005).

There are several factors that can affect the timing of the peak value and shape of bomb radiocarbon curves. Peak 14 C values often lag slightly behind the atmospheric values due to the time lag between geographic distribution and incorporation of carbon into living tissue (Nydal 1993). For this same reason, 14 C values often vary slightly with geographical locations (e.g., Kerr et al. 2004). The watermixing time tends to be lower in marine systems (e.g., Campana and Jones 1998), and the trophic position or origin of diet items of the organism (e.g., Campana et al. 2002) may cause a lag in the onset of the curve and differences in 14 C values. Large differences in the magnitude of

the peak and the post bomb radiocarbon values can reflect regional differences in water

mixing rates, which can dilute the bomb signal. Some imprecision in the otolithcoring

41 technique also can affect the shape of the radiocarbon curve.

Freshwater drum otoliths were noted to contain clear periodic increments in the early 1980s (Becker 1983), but at the time these were not validated as annual increments.

Goeman et al. (1984) reported age validation of Mississippi River freshwater drum otoliths by following the progression of strong year classes for three consecutive years.

While their study provided strong evidence that freshwater drum otoliths produced accurate ages, the method they used was actually age corroboration, not age validation

(Campana 2001). Additionally, in the Goeman et al. (1984) study the progression of strong yearclasses did not have consistent representation each year, there were no fish over age 10 in the age frequency histograms, and the overall rigor of the study is unknown because the number of fish plotted on the age frequency histograms is not reported. We consistently identified the strong 1983 yearclass of drum in Lake

Winnebago in age frequencies using otolith age estimates of drum captured in assessment trawling in 1986 and 2003–2007. The 1983 yearclass was first detected in experimental trawl samples as age0 fish, and persisted in otolith age frequencies in assessment trawl sampling in 1986 and 2003–2007. This yearclass progression corroborates the age validation of the older drum assayed for 14 C and also supports the accuracy of the age

estimates from otoliths of freshwater drum of younger ages.

Published studies on life history characteristics of freshwater drum before 1994

were based on scale age estimates (e.g., Butler and Smith Jr 1950; Daiber 1953; Edsall

1967). After 1994, published studies were primarily based on otolith age estimates

(Pereira et al. 1995a; Pereira et al. 1995b; Rypel 2007; Rypel et al. 2006), although some

(e.g., Braaten and Guy 2004; French III and Bur 1996) still relied on scale age estimates,

42 perhaps because freshwater drum otoliths had never been truly validated and because virtually all reference books (e.g., Becker 1983; Schultz 2004; Werner 2004) cited age estimates from scales. Von Bertalanffy parameters and sexual dimorphism in growth rates (Palmer et al. 1995), and ageatmaturation parameters (Rypel 2007) based on drum otolith ages have been reported, but without validation of otolith growth increments.

Since our results show that scale ages are inaccurate, all demographic parameters based on scale ages must be incorrect. One effect of this inaccuracy is that changes in parameters, such as mortality and growth rates, cannot be detected over time. We explored the possibility of using the relationship between otolith and scale age estimates from archived scale age data to reconstruct a usable age structure for historic freshwater drum populations. Although we found a significant relationship between scale and otolith age estimates up to age 3, the diminished relationship for structures from older fish reduced the likelihood of accurately discerning true age structure from archived scale age data.

There may be regional or geographical differences in agreement between otolith and scale ages. For Mississippi River freshwater drum, scales were found to overestimate age through age 9 (Goeman et al. 1984). In our study, the TL at which otolith age estimates begin to diverge from scale age estimates corresponds to the TL of the onset maturity of Winnebago freshwater drum. Otolith growth is less likely to be disrupted by maturation than scale growth because otoliths are a vital component of a sensory organ of the nervous system and, unlike scales, otoliths grow throughout the lifetime of a fish and are not subject to resorption (Campana and Neilson 1985). The agreement of spines and scales in our study demonstrates that the growth of otoliths is

43 controlled separately from scale and skeletal tissues. Our validation confirms that the freshwater drum is one of the longestlived fishes of the Lake Winnebago system, surpassed only by lake sturgeon, which are estimated to attain ages up to 96 years (R. M.

Bruch, unpublished data). Other longlived fishes within the Great Lakes drainage are known to rarely exceed 40 years. Based on otolith age estimates, lake trout in Lake

Superior are estimated to live up to 42 years (Schram and Fabrizio 1998), while lake trout in the Arctic were recently validated to live at least 50 years (Campana et al. 2008).

Flathead catfish Pylodictis olivaris have been reported to attain a maximum age of 17 years in the Lake Michigan drainage based on pectoral spine age estimates (Daugherty and Sutton 2005) and 28 years in the Tallapoosa River, Alabama, based on otolith age estimates (Nash and Irwin 2000). Otolith age estimates show that flathead catfish in the

Lake Winnebago system may reach 30 years of age (Allen Niebur, Wisconsin

Department of Natural Resources, personal communication). The Great Lakes cisco

Coregonus artedi was recently reported to reach at least age 18 based on otolith age estimates, much longer than previously thought based on scale age estimates (Yule et al.

2008).

Slowgrowing, latematuring species are more vulnerable to human exploitation

(Musick 1999). Early freshwater drum management decisions on the Winnebago system were based on the premise that freshwater drum were a fastgrowing, shortlived species with a high mortality rate (Priegel 1967a). Our age validation study indicates, however, that freshwater drum live much longer than the majority of other species in the Lake

Winnebago fish community and, unlike most other longlived species (e.g., lake sturgeon), mature at a relatively young age and spawn annually. This strategy optimizes

44 an individual’s reproductive value. For example, a female lake sturgeon living to age 80 will spawn an average of 15 times within her lifespan, while a female freshwater drum living to an age of 50 will spawn approximately 45 times in her life span. This unique trait allows freshwater drum to be more prolific than fish species with a latematuring life history strategy, which may partially explain why freshwater drum are geographically widespread and frequently abundant where present. This life history trait undoubtedly contributed to the poor success of 55 years of rough fish removal programs on Lake

Winnebago designed to reduce drum abundance (Kamke and Bruch 1991; Priegel 1967a).

Our results support revision of reference books that base the life history characteristics of freshwater drum on scale age estimates. For example, the maximum lifespan reported in Becker (1983), a commonly cited fisheries reference book for fishes from Wisconsin, is 17 years. The maximum age of Lake Winnebago freshwater drum in this study was 58 years based on the validated otolith ages. A freshwater drum sampled from Lake Winnebago in the late 1980s was estimated from otoliths to be age 70

(R.M.B., personal observation). The greatest maximum published age based on otoliths of freshwater drum from the Red Lakes, Minnesota, is 71 years (Pereira et al. 1995a).

The Red Lakes and Lake Winnebago provide similar habitats, occur within similar latitudes, and are both large, shallow systems with no seasonal thermal stratification. The maximum lifespan of freshwater drum in Alabama based on otolith age estimates was reported to be about 30 years (Rypel et al. 2006). With the validation of ages derived from freshwater drum sagittal otoliths, it is important that all freshwater drum age estimation is based on otoliths, the only structure in drum that provides precise and accurate estimates of age for calculation of meaningful demographic population

45 parameters.

CHAPTER THREE: LONGTERM CHANGES IN THE DEMOGRAPHY OF A FRESHWATER FISH CORRESPONDING TO ANTHROPOGENIC DISTURBANCES

Introduction

Subfossilized otoliths from freshwater drum (Aplodinotus grunniens ) have been

recovered from Native American encampments dating to preEuropean settlement in the

North American Midwest. Drum otoliths from archaeological excavations have been

used to construct drum growth indices (Bergquist 1996) and make longterm

comparisons of body length (Mason 1996; Priegel 1963b; Witt Jr 1960). It was recently

determined that age of drum can also be accurately estimated from otoliths (DavisFoust

et al. 2009), but age compositions and growth rates have not been compared between the

two time periods. In this study, ages and growth rates of archaeological and modern

drum collected from the Lake Winnebago region are examined and the results are

compared to identify longterm changes in drum demographics.

Freshwater drum presently dominate the fish community (Becker 1964; Wirth

1958; Wisconsin Department of Natural Resources 2004) of Lake Winnebago, the largest

inland lake in Wisconsin, and drum were important to the Native Americans in the Lake

Winnebago region as hundreds of drum otoliths recovered from cooking pits and middens

in excavated Native American encampments attest (e.g., Dirst 1985; Mason 1996;

Overstreet 1997; Seurer 1978). Mason (1996), Priegel (1963), and Witt Jr (1960)

concluded that freshwater drum that lived prior to European settlement attained greater body lengths than modern drum. Using the otolith length to body length relationship for

modern drum, Witt Jr (1960) calculated body lengths from ancient drum otoliths

47 recovered from five archaeological sites in the United States that dated from 7000 B.C.

1600 A.D. He concluded that archaeological drum had a longer maximum length than modern drum captured in nearby sites. Similar methods were used by Priegel (1963b) and Mason (1996) for archaeological sites on the Lake Winnebago system to also conclude that the maximum length of archaeological drum was greater than modern drum. Archaeological otoliths have not been used to make longterm comparisons of age composition and growth rates in freshwater drum.

Lake Winnebago has faced intense anthropogenic pressures since the early to

middle 1800’s (Wisconsin Department of Natural Resources 1989; 1993; 2004). A series

of 17 locks and 12 dams were installed on the lower Fox River in the 1850’s and

maintain unnaturally high and constant water levels that have caused the gradual loss of

thousands of acres of emergent wetlands on Lake Winnebago. The impoundment has

isolated the Lake Winnebago fish community from the Great Lakes; however, tag return

data show that some species do occasionally migrate from Lake Winnebago to the Great

Lakes (Wisconsin Department of Natural Resources, unpublished data). Motor boats have kept sediments suspended, thereby maintaining turbidity and preventing the establishment and growth of both submerged and emergent aquatic plants. Nutrient inputs from agriculture and urban sprawl have intensified the presumed natural eutrophic condition of this lake (Sloey and Spangler 1977; Wisconsin Department of Natural

Resources 1989; 1993; 2004). Nonnative aquatic invasive species have been introduced since at least the early 1900’s, including rusty crayfish (Orconectes rusticus ; Hobbs III

and Jass 1988) and zebra mussels (Dreissena polymorpha ; Wisconsin Department of

Natural Resources 2004) that presumably have trophic linkages with freshwater drum

(e.g., French III and Bur 1996).

In an attempt to decrease the abundance of drum, which are locally considered undesirable, and increase game fish abundance, millions of kilograms of drum were removed annually from 1936 to 1990 (Otis 1988; Staggs and Otis 1996). The removal program was terminated in 1990 in part because the desired effect was never confirmed

(Kamke and Bruch 1991; Staggs and Otis 1996). While drum are not commercially or recreationally important, their great biomass inherently makes them a significant component in the energy flow of this ecosystem.

Thus, in this study, length distribution, age composition, and growth rates of archaeological and modern drum collected from the Lake Winnebago region are examined and the results are compared to identify longterm changes in drum demographics. Otolith lengths will be used exclusively for this comparison as the body lengths are unavailable for archaeological drum. The observed longterm changes in the demography of Lake Winnebago drum will be discussed as evidence for the potential impacts of the anthropogenic disturbances of the 19th and 20 th centuries on aquatic ecosystems.

Methods Study site

Lake Winnebago spans 55,728 hectares with a maximum approximate length of

48 km and width of 16 km (Wisconsin Department of Natural Resources 2004). It has a maximum depth of 6.4 meters, an average depth of 4.7 meters, and is part of the eutrophic WinnebagoUpper FoxWolf Watershed, connected to Lake Michigan via the lower Fox River.

Data collection for modern drum

Drum were obtained at weighins of local fishing tournaments in July 2003, 2006,

2008, and 2009 (the “angling” subset); by netting casualties immediately following underwater blasts for a construction project in the Fox River between Lake Butte des

Morts and Lake Winnebago in April 2007 (the “blast” subset); and from assessment trawling in 2007 (the “trawl” subset). Assessment trawling was conducted at random locations within historical lake management areas to broaden the geographic scope of lake wide sampling. A bottom trawl with a 7.9 meter head rope, 3.8 cm stretch mesh body and a 1.3 cm stretch mesh codend was towed 6.47.2 km/hr for five minutes in each of two randomly selected one minute latitude by one minute longitude sampling grids within each of six Lake Winnebago sampling areas. Trawling took place during daylight hours and usually covered areas from 4.3 to 5.8 meters in depth. Total lengths were measured to the nearest 2.5 mm and sagittal otoliths were extracted for age determination.

Otoliths were washed to remove all adhering tissues and a vernier caliper was used to measure the length of an otolith’s longest axis to the nearest 0.01 millimeters.

Each otolith was prepared for sectioning by embedding in epoxy (Epoxycure TM resin and hardener in a 5:1 weight ratio) to prevent fracturing while being cut. Two to four transverse sections of 0.2540.475 mm thickness cut from the core of each otolith with a lowspeed diamond blade saw were mounted with cyanoacrylate on glass slides for viewing and storage. An Olympus SZX7 dissection microscope with an Olympus DP71 camera was used for microscopic analysis and capturing digital images. The clearest sections that were cut closest to the core of the otolith were used for age determination.

Seventy percent ethanol was applied to the sections for clarification and transmitted light was used to identify opaque zones, which were counted as annual increments (Casselman

1987). Otolith sections were assessed for age by at least two experienced readers.

Samples were omitted from further analysis when ages from the two readers did not agree, which seldom occurred and was usually due to poorly cut sections or malformations of the otoliths.

Data collection for archaeological drum

Drum sagittal otoliths recovered from seven archaeological sites surrounding

Lake Winnebago dating from approximately 8500 BC to 1730 AD ( Figure 3.1 ) were

acquired from the University of Wisconsin Oshkosh Anthropology Department. The

estimated dates of the otoliths are based on radiometric dating, ethnohistoric data, and

other culturally diagnostic artifacts found in general association with the sites (Behm

2008; Dirst 1985; Overstreet 1997; Seurer 1978).1

1 The estimate of the Sauer Resort time period was calibrated using quickcal 2007 version 1.5.

Metzig Garden 47WN283 1 otolith Doty Island 47WN670 ca.85006500 BC 448 otoliths Sauer Resort 47WN207 ca.16801712 9 otoliths ca.13901530 (calibrated ca.13171419) Lake Winneconne Lake Winnebago

Kargus 47WN216 50 otoliths Lake Butte des Morts ca.13301390

Bell 47WN9 2 otoliths Menominee Park II 47WN544 ca.12221275 12 otoliths 58 otoliths ca.13501640 ca.16801730

Fahrney Point 47WN754 38 otoliths ca.17281730

Archaeological otoliths that were structurally complete were prepared for analysis

in a manner similar to the modern otoliths, except that they were not washed in order to prevent potential or additional exfoliation of the outer annuli. Otoliths were embedded in

a relatively hard epoxy (Araldite epoxy GY502 resin and hardener HY956 in a 5:1 weight

ratio), were sectioned following the same methods as for the modern otoliths, and were

assessed with reflected light as they were mineralized to the point that light could not be

transmitted through them. Age determinations were discarded from the data if outer

annuli were particularly unclear or exfoliated. The archaeological otoliths were more

difficult to read than the modern otoliths, so not all ages were agreed upon between

52 readers. Age precision was evaluated for the archaeological otoliths by calculating the coefficient of variation (CV) among assigned ages (Campana 2001). The final CV was

0.37%. When ages were not agreed upon, the age determination of the more experienced reader was used.

Data analysis

Differences in mean otolith length among fish captured at the three archaeological sites with n>50 (Doty, Bell, and Kargus) and among the modern fish captured in the three gears (angling, blast, and trawl) were assessed with a oneway ANOVA. Otoliths <9.2 mm in length were not used because this was the length of the smallest recovered archaeological otolith. One otolith from each pair was randomly selected (either left or right) from each modern fish. Prior to conducting the oneway ANOVA, homogeneity of variances was tested with Levene’s test and normality was assessed with an Anderson

Darling test and visual inspection of a histogram of residuals. In situations where a transformation could not be found to adjust for heterogeneous variances, a weighted one way ANOVA was used with the inverse of the group residuals variance as weights.

Tukey’s posthoc multiple comparison method was used to determine pairwise differences following a significant ANOVA result. As sample sizes varied greatly among sites and gears, the robustness of the ANOVA results to sample size differences was examined by extracting 1000 random samples of n otolith lengths from each site or gear

group with replacement, where n is the minimum sample size among the three sites or

three gears. The overall ANOVA p value and each Tukey’s pairwise p was recorded for each of the 1000 resamples. If p<0.05 on fewer than 5% of the resamples, then a

53 significant difference was declared. Sites or gears that did not have significantly different mean otolith lengths were pooled. All subsequent analyses examined differences among the archaeological “sites” and the modern “gears” (some sites and gears may have been pooled) which, hereafter, will be called subset groups. Differences in mean otolith length and differences in mean otolith age among subset groups were assessed similarly with a weighted oneway ANOVA and resampling procedure.

Von Bertalanffy growth models (VBGM) of otolith length at age were used to compare growth patterns among subset groups. Both the traditional (Ricker 1975) and

Francis parameterizations (Francis 1988) of the VBGM were used. The Francis parameterization was used once with parameters set at mean length at ages 5, 15, and 25 and a second time at ages 10, 35, and 50. The different ages were used in the Francis parameterization in order to better describe the growth of drum throughout the observed range of ages. Models were fit assuming a multiplicative error structure (i.e., residual errors were log transformed). A sequence of extra sumofsquares tests (Ritz and Streibig

2008) were used to identify whether or not parameters were equal among all subset groups. If parameters were not equal among the subset groups, then confidence intervals for each parameter from the profile likelihood method were examined to identify differences in parameters among subset groups, where nonoverlapping confidence intervals suggested a significant difference.

Size and growth rates were assessed by analyzing otolith lengths as body lengths were not available for archaeological drum. Previous studies (Priegel 1963b; Witt Jr

1960) estimated the body lengths of archaeological drum using the relationship between body length and otolith length for modern drum. Body lengths were not estimated from

54 otolith lengths for archaeological drum because the accuracy of these estimates cannot be assessed. However, the relationship between otolith length and body length of modern drum was modeled to ascertain the strength of the relationship between these two variables. Piecewise linear regression (Toms and Lesperance 2003) was used because a break was readily apparent (see Figure 3.7) when these data were plotted.

All statistical tests were performed using the R environment version 2.14.1 (R

Development Core Team 2012). The Tukey procedure was implemented with glht() from the multcomp package (Hothorn et al. 2008), the von Bertalanffy models were fit with nls() and the profile confidence intervals were constructed with confint() from the nlstools package (Baty and DelignetteMuller 2011), and the piecewise linear regression was fit with piecewise.linear() from the SiZer package (Sonderegger 2011). A significance level of α=0.05 was used for all statistical tests.

Results

Length Distributions

Mean otolith length from archaeological samples differed among the three sites

(F(2,553)=39.277, p<0.0001; Figure 3.2 ), with the mean otolith length greater at Doty

Island than at Bell and Kargus, which did not differ significantly (Tables 3.1 and 3.2).

All further analyses of archaeological samples were with Doty Island separated from Bell and Kargus, which were pooled.

Mean otolith length from modern samples differed among gears

(F(2,478)=87.750, p<0.0001; Figure 3.2 ). Tukey’s method suggested that all three gears had different mean otolith lengths ( Table 3.2). However, the resampling procedure

55 indicated that blastcaptured fish did not differ from trawlcaptured fish. Thus, angling captured fish had significantly larger mean otolith lengths then blast or trawlcaptured fish. All further analyses of modern samples were with anglingcaptured fish separated from blast and trawlcaptured fish, which were pooled.

Mean otolith length differed among the four subset groups ( F(3,478)=237.7, p<0.0001; Figure 3.2 ). Tukey’s pairwise comparisons suggested that the mean otolith length differed among all subset groups, but the resampling procedure indicated that the modern anglingcaptured fish did not differ significantly from either subset of archaeological fish ( Table 3.2). Thus, mean otolith length was greater at Doty Island

than at the combined Bell/Kargus subset, which were both larger than the modern blast

and trawlcaptured fish ( Tables 3.1 and 3.2). Modern anglingcaptured fish were larger than modern blast and trawlcaptured fish.

Bell Angling

0 5 15 0 4 8 10 15 20 25 10 15 20 25

Doty Blast

0 5 10 0 40 80

Frequency Frequency 10 15 20 25 10 15 20 25

Kargus Trawl

0 4 8 12 0 20 40 60 10 15 20 25 10 15 20 25

Bell/Kargus Blast/Trawl mdrn-BlTr arch-BeKa

0 40 0 10 10 15 20 25 10 15 20 25

Otolith length (mm)

Subset n mean st. dev. min. max. Bell 58 16.22 2.79 9.20 23.40 Doty 448 18.50 2.38 10.11 24.88 Kargus 50 16.23 1.99 12.11 21.00 Angling 106 17.36 3.03 9.69 26.50 Blast 51 12.84 1.30 10.00 15.34 Trawl 324 13.61 3.22 9.32 24.15 Bell/Kargus 108 16.22 2.44 9.2 23.40 Blast/Trawl 375 13.51 3.04 9.32 24.15

Tukey contrasts comparisons between t statistic p resampling Doty – Bell 6.819 <0.0001 98.1% Kargus – Bell 0.017 0.9998 1.6% Kargus – Doty 6.360 <0.0001 99.5% Blast – Angling 4.523 <0.0001 100.0% Trawl – Angling 3.756 <0.0001 100.0% Trawl – Blast 0.767 0.0073 15.8% Bell/Kargus – Doty 7.697 <0.0001 100.0% Angling – Doty 3.893 0.0007 65.5% Blast/Trawl – Doty 26.386 <0.0001 100.0% Angling – Bell/Kargus 3.044 0.0119 65.0% Blast/Trawl – Bell/Kargus 9.023 <0.0001 100.0% Blast/Trawl –Angling 12.966 <0.0001 100.0%

Age Distributions

Mean otolith age differed significantly among the four subset groups (F

(3,1031)=305.04, p<0.0001; Table 3.3). Tukey comparisons indicated that the mean ages of all four subset groups were significantly different, but the resampling procedure indicated that the modern anglingcaptured fish were not different from the archaeological Bell/Kargus fish ( Table 3.4, Figure 3.3 ). The oldest mean otolith age was found for archaeological fish at Doty Island, followed by the archaeological fish at

Bell/Kargus and the modern anglingcaptured fish. Three archaeological drum had lived over age 70, and the single oldest modern drum had reached age 58. Less than 0.1% of the modern drum and >16% of archaeological drum were over age 50.

arch-Doty arch-Bell/Kargus

30 50 20

20 10

0 0 0 20 40 60 80 0 20 40 60 80

mdrn -Blast/Trawling mdr n-Angling Frequency 30 80 20 40 10

0 0 0 20 40 60 80 0 20 40 60 80

Age

Table 3.3. Summary statistics for freshwater drum ages for the four subsets. “Bell/Kargus” and “Blast/Trawl” represent pooled samples.

subset n mean age st. dev. min max Doty 446 38.5 15.3 2 73 Bell/Kargus 108 24.2 10.5 5 57 Angling 106 19.0 9.1 0 47 Blast/Trawl 375 12.5 8.6 0 38

Tukey contrasts comparisons between t statistic P resampling archBell/Kargus – archDoty 13.918 <0.0001 100.0% mdrnAngling – archDoty 14.584 <0.0001 100.0% mdrnBlast/Trawl – archDoty 29.596 <0.0001 100.0% mdrnAngling – archBell/Kargus 3.836 <0.0001 89.4% mdrnBlast/Trawl – archBell/Kargus 11.344 <0.0001 100.0% mdrnBlast/Trawl – Angling 4.525 <0.0001 97.5%

Growth Models

The L∞ parameter from the VBGM did not differ significantly (F(3,1023)=1.853, p=0.1359), but K ( F(3,1023)=5.299, p=0.0012) and t0 ( F(3,1023)=7.095, p<0.0001) did differ significantly among the four subset groups (Figure 3.4 ). The L∞ parameter is poorly estimated for the archaeological Bell/Kargus fish and the modern blast and trawl captured fish due to the lack of older fish that would have approached asymptotic growth.

In addition, the VBGM parameters estimated from the Francis parameterization for age

50 were not significantly different among subset groups ( F(3,1023)=1.889, p=0.1297), but suffers similarly from the lack of older fish recovered from the Bell/Kargus archaeological sites and the blast and trawlcaptured modern samples ( Figure 3.5).

Modern anglingcaptured fish frequently achieved greater lengths at age than the blast or trawlcaptured fish, and the drum from the archaeological Doty Island site frequently achieved greater lengths at age than fish from the Bell/Kargus sites.

archDoty archBell/Kargus mdrnAngling Otolithlength (mm) mdrnBlast/Trawl 10 12 14 16 18 20

0 10 20 30

Age (years)

Figure 3.4. von Bertalanffy growth models were fit to the four subsets using otolith length at age. Archaeological subsets were given the prefix “arch,” and modern subsets the prefix “mdrn.”

age50

age35 age25

age15

age10 12 14 16 18 20 22 24

Predicted mean otolith length (mm) lengthotolith mean Predicted age5

archarchDotyDoty arch archBeKaBell/Kargus mdrn mdrnAnglingAngling mdrn mdrnBlTrBlast/Trawl

Group

Estimation of archaeological drum body lengths

A broken line relationship was modeled between the otolith lengths and body lengths of 481 modern drum ( Figure 3.6 ). There was a strong linear relationship between otolith and body length for otoliths both under and over 14.20 mm ( r2=0.97, r2=0.96, respectively). Total length (mm) Total length 200 400 600 10 15 20 Otolith length (mm)

Figure 3.6. Otolith length versus total (body) length of 481 modern drum.

Discussion

Demographic statistics of freshwater drum from the Lake Winnebago system have changed since European settlement. Growth rates have increased and longevity has declined, but potential maximum lengths have remained similar. Considering the high proportion of archaeological drum that survived over age 50 relative to modern drum, the evidence for decreased longevity is strong. The potential maximum length of modern freshwater drum from the Lake Winnebago system could not be directly examined due to the absence of older ages; however, growth models predicted that drum from both time periods have the potential to reach similar mean maximum body lengths.

An increased growth rate, or compensatory growth response, in fish can be triggered by a decline in population density (Bystrom and GarciaBerthou 1999; Ricker

1975). An increase in growth rates would explain why changes in length frequencies or population densities of drum were never detected during sixty years of removal operations in Lake Winnebago. Changes in the age distribution (i.e., more young fish) would not have been detected because scales were used for age determinations at that time, which were later invalidated (DavisFoust et al. 2009).

Faster growth rates of modern drum could be a response to factors that govern metabolic activity, such as warmer temperatures or increased food availability (Beverton and Holt 1959). The climate is warmer today than during the time period that most of the archaeological drum in this study lived. The Little Ice Age cooled temperatures in the

Northern Hemisphere from the 15 th to 19 th centuries (Mann et al. 2009), and seasonal

temperatures are positively correlated to drum growth and recruitment rates (DavisFoust,

unpublished data), as in many other fish species (e.g., Lankford Jr and Targett 2001;

Pepin 1991; Pörtner et al. 2001). Further, food abundance for small drum may be greater today than in the past, which could affect recruitment rates and population densities.

Small drum rely heavily on chironomids larvae as a food resource (Priegel 1967a; see

Chapter 4 of this dissertation), and the environment of Lake Winnebago has purportedly changed to support greater densities of chironomids larvae. Apparently nuisance levels of chironomids from Lake Winnebago were unknown until the early 1900’s (Burrill

1913), which corresponds to the enhancement of naturally eutrophic condition of Lake

Winnebago beginning in the 1800’s. Hilsenhoff (1967) found that the abundance of adult chironomids that emerged in May was positively correlated to the amount of spring runoff. The feeding stimulus for the larvae to pupate and emerge as adult flies was diatoms, and diatom blooms were dependent on the amount spring runoff. Sawdust from sawmills contributed an enormous amount of seston to Lake Winnebago starting around

1850 (Mason 1993). The inputs from sawmills declined during the early 1900’s, but sewerage pollution from Fond du Lac and Oshkosh and nutrient runoff from agricultural practices were increasing. Damming of the lower Fox River in the 1850’s caused slower water flow due to higher water levels, so organic muck built up at the bottom of the lake and provided optimal habitat for chironomids larvae. In addition, drum and lake sturgeon

(Acipenser fulvescens ), which also rely on chironomid larvae for food (Stelzer et al.

2008), were both intensively harvested by the early 1900’s (Smith 1968), perhaps reducing predation intensity.

There are a number of assumptions and simplifications that have to be made when studying archaeological fauna ( Table 3.5 ), which affects this study as well as the original conclusions drawn by Witt Jr (1960), Priegel (1963b), and Mason (1996). The most

65 severe limitation of these former studies was the lack of age data, so it may have appeared that archaeological drum attained greater lengths only because they lived longer, but there are other potential problems. Firstly, it is unknown whether indigenous people selected fish based on size, and any selectivity they did practice may have varied over time and location depending on fish abundance and general food availability.

Different fishing gear may have biased size distributions. The practice of angling by

Native Americans is evidenced by fishing lures that have been recovered from many archaeological sites, and it is also likely that nets and spears were used (Mason 1993;

Stevenson 1985). Further, there is no certainty of location of capture. Canoes and other floatation devices may have been used for fishing and traveling to fishing areas.

Table 3.5. The analysis of the archaeological otoliths was built on the following assumptions and simplifications for unknown factors regarding the archaeological drum otoliths. Problem Assumption/Simplification Small otoliths are less likely to be Reduce the modern data set to exclude otoliths smaller than the recovered smallest archaeological otolith

Location of capture is unknown Assume that drum were captured nearby encampment and subset archaeological otoliths by site to detect potential differences among the sites

Capture methods and size selectivity Subset archaeological otoliths by site to detect temporal and spatial are unknown and may have changed differences in demographic statistics; compare archaeological over time samples to modern samples obtained from angling competition for largest drum

Demographic statistics may have Assume demographic statistics were relatively consistent by site; changed over time (preEuropean simplify models by pooling sexes and subset archaeological settlement) in relation to exploitation otoliths by site to detect temporal and spatial differences in and other environmental factors demographic statistics

Growth rates are sexually dimorphic Assume sex ratios are constant; simplify modern models by (Rypel 2007), but sexes of pooling sexes; subset archaeological otoliths by site to detect archaeological drum cannot be temporal and spatial differences in demographic statistics due to separated differing sex ratios

Relationship between body length Assume relationship between body length and otolith length has and otolith length may not be remained constant over time consistent

The results of this study do not reaffirm earlier studies that freshwater drum

attained greater lengths prior to European settlement (Mason 1996; Priegel 1963b; Witt Jr

1960). The random body length distributions they used for modern drum could have been skewed towards smaller individuals by the early to mid1900’s due to commercial

and removal trawling cropping off the largest individuals. The different results also may

have been due to the otolith length to body length conversion model employed in the

former studies. The breakpoint in the relationship between otolith length and body length

was missed, perhaps due to relatively small sample sizes of large drum, and the resulting

models had steeper slopes that overestimated body lengths.

The breakpoint in the otolith length to body length relationship may be explained by the “metabolically inert” property of otoliths, whereby, unlike all other bones in the

skeletal system, deposited minerals are not remobilized (Campana and Neilson 1985). As body growth slows and mineral deposits in bone tissue are resorbed, the growth of the

otolith would continue as the growth of skull cavity slows, perhaps altering the axis of

growth in the otoliths. The archaeological otoliths, which were from slower growing

individuals, generally appeared more robust ( i.e., thicker through the transverse section)

than modern otoliths of similar ages (S. DavisFoust, pers. obs.). Perhaps minerals are

deposited on the distal or medial surfaces rather than the posterior and anterior edges

when longitudinal growth of the otolith is limited by cavity space.

The evenness of incremental spacing in otolith sections may reflect the stability of the environment and should be compared between archaeological and modern drum.

Pereira et al. (1995a) were unable to detect regular periodicity in modern drum growth patterns, but they suggested that there could have been regular periodicity that had been

67 extinguished by human influences. An increase in variability in incremental spacing over time would suggest that environmental variability has increased. Further, environmental variability could be used to gauge changes in recruitment variability.

The changes in the life history traits detected in this study are probably a response to anthropogenic disturbances, and it is unlikely that freshwater drum are the only species undergoing these changes. Unfortunately, there is little longterm data on Lake

Winnebago to demonstrate how much it has been affected by anthropogenic disturbances.

For example, much of the baseline data in Wisconsin has only been collected over the past 30 years. In this short time frame, ongoing declines in nongame native fish species

have been observed, suggesting that environmental conditions continue to change (Lyons

et al. 2000). Management has recently brought about improvements to water quality in

Lake Winnebago; however, the establishment of zebra mussels and other aquatic invasive

species continue to alter the ecosystem causing many unknown and poorly understood

effects on the aquatic ecosystem.

Freshwater drum have important traits that make them a key species in continuing

to monitor ecosystem changes in Lake Winnebago. For example, if either drum

longevity increases or growth rates decline, and these changes correspond to either water

clarity or a decline in chironomid densities, it would point to the factor that caused the

change. If water quality does not improve, but drum life history traits rebound back to

what was observed prior to European settlement, it could indicate that there were effects

of rough fish removal in Lake Winnebago. If water clarity improves, but drum

demography does not change, it would be indicative of long term effects due to

continuing artificially high water levels, the establishment of nonnative species, or

68 climate change. Freshwater drum do not presently have social or economic value, but they have great value as an indicator of environmental change.

CHAPTER FOUR: FRESHWATER DRUM RESPONSE TO THE INTRODUCTION

OF ZEBRA MUSSELS IS REFLECTED IN THEIR OTOLITH GROWTH CHRONOLOGY

Introduction

Zebra mussels ( Dreissena polymorpha ) have negatively affected the lower trophic

levels of many ecosystems (e.g., Fahnenstiel et al. 2010; Hecky et al. 2004; Idrisi et al.

2001; Malkin et al. 2008; Mayer et al. 2000; Ozersky et al. 2009; Richardson and Bartsch

1997; Strayer and Malcom 2006). The effects of dreissenid introductions on fish growth

have largely been indirect through the alteration of availability of favored food resources

(e.g., Bartsch et al. 2003; Giuliano 2011; Johannsson et al. 2011; Karataev and

Burlakova 1995; Karatayev et al. 1997; Pothoven and Madenjian 2008; Strayer et al.

2004). A direct effect of the consumption of dreissenids on the growth of molluscivorous

fish has not been identified (e.g., French III and Bur 1996; Smith et al. 2008), presumably because of the relatively low energetic content of shelled prey (Pothoven and Madenjian

2008; Pothoven and Nalepa 2006).

Freshwater drum ( Aplodinotus grunniens ) is a molluscivorous fish and is the most

abundant fish species in Lake Winnebago, the largest inland lake in Wisconsin (WDNR

2004). This watershed has relatively few of the aquatic invasive species that now inhabit

much of the Laurentian Great Lakes. Zebra mussels were initially found in Lake

Winnebago in 1998 (Wisconsin Department of Natural Resources, unpublished data);

however, there are no estimates of their abundance. Preliminary observations suggested

that drum in Lake Winnebago did not initially consume zebra mussels, consistent with previous observations of the relative lack of molluscivory by freshwater drum in Lake

Winnebago (Priegel 1967a). However, drum are one of the main fish species to feed on dreissenids throughout North America (Eggleton et al. 2004), with dreissenids forming as much as 60% of diet of drum in some situations (Watzin et al. 2008).

The identification of environmental parameters that control growth rates is vital to understanding ecological interactions within a community. Growth rates are controlled by the interplay of endogenous and exogenous variables; endogenous variables include

age, maturation, and individual variability (Rhoads and Lutz 1980; Weisberg 1993),

whereas exogenous variables include effects from the biotic (e.g., resource abundance

and community interactions) and abiotic environment (e.g., temperature, photoperiod, precipitation, dissolved oxygen; Evans and Claiborne 2006). The summation of these

variables constitutes the growth of an individual within a given time frame, and the

relative contribution of variables influencing total annual growth can be statistically partitioned to provide information on regulation of growth of an organism or to serve as

an archive of environmental change experienced by the organism (Weisberg 1993;

Weisberg et al. 2010).

Many field studies have established a link between otolith growth chronologies

(time series of annual growth indices) and temperature for adult fish (Black et al. 2011;

Cyterski and Spangler 1996; Gauldie 1991; Hoff and Fuiman 1993; Matta et al. 2010;

Morrongiello et al. 2011; Ostazeski and Spangler 2001; Pereira et al. 1995a).

Relationships between food resources and otolith growth have been observed at the

microstructure level for larval fish in laboratory experiments (e.g., Barber and Jenkins

2001; Campana 1983; Neilson and Geen 1982; Radtke and Fey 1996; Tonkin et al.

2008a; Tonkin et al. 2008b) and less commonly in natural settings (e.g., Meekan et al.

2003). However, the effect of resource availability on otolith growth has not been reported in the otolith macrostructure of mature fish. Identifying a relationship between otolith growth and resource abundance is more difficult, in part, because creating precise indices of food abundance is intensive, whereas climatic indices are often readily available. As a consequence, fish growth chronologies are often hypothesized, but not shown, to be indirectly related to resource availability, such as from eutrophication (e.g.,

Rutherford et al. 1995; Smith et al. 2008) or changes in population density (Morrongiello et al. 2011; Ostazeski and Spangler 2001).

It is likely that consumption of dreissenids by drum has increased since zebra mussels were first discovered in Lake Winnebago and that consumption of mussels is related to drum body size. However, it is uncertain if consumption of zebra mussels would result in changes in body growth that would ultimately be reflected in the growth chronology developed from drum otoliths. Thus, the objectives of this research were to determine (1) if freshwater drum were consuming zebra mussels, (2) if zebra mussel consumption varied according to drum body length, (3) if there was a relationship between the drum otolith growth chronology and the establishment of zebra mussels, and

(4) if there was a relationship between body condition of drum and the establishment of zebra mussels. The relationship between the drum otolith growth chronology and regional climate indices was also examined to ensure that any observed changes in the otolith growth chronology were not a response to regional climatic trends.

Methods

Data Collection

Drum were collected by trawling, angling, and boom electroshocking, and from underwater blasts used for a bridge construction project. Trawling was conducted with a

7.9 meter head rope, 3.8 cm stretch mesh body and 1.3 cm stretch mesh codend bottom trawl towed at 6.47.2 km/hr for five minutes at depths of 4.3 to 5.8 meters in October from 19842009 at two randomly selected one minute latitude by one minute longitude sampling grids within each of six sampling areas. Angling samples were collected from fishing tournaments held in the cities of Omro and Oshkosh in July 2003, 2006, and

2009. Electroshocking was conducted along the western shore of Lake Winnebago from

April through September 2009. Fish that were casualties from the underwater blasts were collected immediately after the blast in the Fox River between Lake Butte des Morts and

Lake Winnebago in April 2007.

Total length (TL; +0.1 inch, except +1 mm in 2009) and weight (+0.1 pounds)

were recorded for all drum. In October 19842006 and 20082009, 15 drum per one inch

TL interval were subsampled from trawlcaptured fish. In 2007, drum were randomly

sampled using a standard volume (36.7 liters) removed from the trawl catch. Sub

sampling was not used for drum collected with the other methods of capture. From 2006

2009, sex and maturity were determined by visual inspection of the gonads of all sampled

fish. Sagittal otoliths were collected from all sampled drum in 1986 and 20032009.

Stomach contents were removed from drum captured in the trawls and by electrofishing

in 20062009.

Diet Analysis

The diet of freshwater drum was broadly characterized by identifying the occurrence (presence/absence) of Amphipoda, Isopoda, larval Chironomidae, adult

Chironomidae, Oligochaeta/Hirudinea, Dreissena, nonDreissenid Mollusca, crayfish, fish, Trichoptera, Ephemeroptera, and microcrustaceans in the stomach contents. The proportion of drum within 20mm TL classes that consumed each food item was

calculated, and a lengthrelated diet shift was visually apparent. To identify the length at

which more than 50% of drum had consumed zebra mussels, a four parameter model

(Richards 1959) was fit to the proportion ( p) of fish in 20mm TL intervals l that had

consumed the food item,

b2 − b1 pl = b1 + −b  l  4 1+    b3  where b1 is the minimum proportion that had consumed the item, b2 is the maximum proportion that had consumed the item, b3 is the TL at which one half of the model

maximum proportion consumed the food item (the inflection point), and b4 is the slope of the curve at b3. The b3 value for zebra mussels was used as the breakpoint to create small and large subsets of drum.

medial

sulcus posterior anterior

distal Figure 4.1. Orientation and location of the sulcus in a sagittal otolith section from a 58 year old drum. The black line indicates the axis where increments were measured.

Growth Analysis

Otoliths were washed to remove all adhering tissues and prepared for sectioning by embedding each structure in EpoKwick Fast Cure Epoxy to prevent fracturing while being cut. Two to four transverse sections of 0.250.48 mm thickness were cut through the core of each otolith with an Isomet saw and were mounted on glass slides for viewing and storage using cyanoacrylate glue. The section with the most visually clear annual increments was used for growth analysis. Microscopic analysis of the sections was done using an Olympus SZX7 dissection microscope equipped with an Olympus DP71 camera.

Growth increments were examined using transmitted and reflected light after 70% ethanol was applied for clarification. Opaque zones were considered the end of an annual growth increment (Casselman 1987). Photographs were taken at 20X magnification and increment widths were consistently measured along a straightline from the focus to the margin within 1 mm of the sulcus using ImagePro Express ( Figure 4.1 ). The last

75 increment was not used in further analyses because it represents incomplete growth from the year of capture.

A growth chronology index was developed using a linear mixedeffects model

(Weisberg et al. 2010) (see Appendix B for a more complete description of the model).

This model decomposes an increment width ( y) into five terms: the intrinsic fixed age effect ( ι), extrinsic random environmental (or year) effect ( η ), intrinsic random fish effect ( f), an age by year interaction effect, and an additive error term ( e),

y =ι + η + f + (ιη) + e cka a c+a−1 ck a,c+a−1 cka where a represents the ath year of life, c represents the cth yearclass, and k represents the kth fish within the cth year class. The term f allows each fish to have its own overall level of growth that applies to all increments on the fish. A master growth chronology was developed for all drum combined, and additional chronologies were also developed for subsets of male and female and small and large drum.

Significant differences between individual years on growth chronologies fit to

separate subsets of individuals were determined using the formula for a twosample ttest

assuming unequal variances,

η −η t = A,c+a−1 B,c+a−1 c+a−1 2 2 SEη + SEη A,c+a−1 B,c+a−1

where SE is the standard error for a year effect and A and B represent the two subsets of

individuals (S. Weisberg, pers. comm.). A sequential Bonferroni adjustment (Holm

1979) was used to correct for multiple comparisons from testing all years on the growth

chronologies.

The relationship between growth chronologies and regional temperature data was

76 examined using Pearson’s correlation coefficient ( r). Mean daily and monthly air temperatures measured near the west side of Lake Winnebago at Oshkosh Wittman

Airport (43°59'N / 88°33'W) from 1948 to 2008 were obtained from the National

Oceanographic and Atmospheric Administration (NOAA) Satellite and Information

Service archives. These air temperatures were used as a proxy for water temperatures because a strong correlation between mean monthly Lake Winnebago water temperatures obtained from City of Appleton Water Treatment Plant (44°13'N / 88°24'W) and mean monthly air temperature data between 1992 and 2008 was observed ( r=0.95, p<0.0001).

Growing degree days (GDD) were calculated annually using a base temperature of 10°C because this temperature is in the midrange of temperatures typically experienced by freshwater drum (Neuheimer and Taggart 2007).

Body Condition

Relative weight (Wr ) values, calculated using the 75 th percentile standard weight equation for freshwater drum proposed by Blackwell et al. (1995), were calculated to determine if the body condition of freshwater drum varied over time and among length classes. To ensure seasonal consistency, only October trawl captured drum were used for this analysis. Twoway ANOVA was used to determine if mean Wr ’s differed between size classes ( small and large ) and pre (19841997) and post (19982009) zebra mussel

establishment time periods. Individual relative weight values were natural log

transformed prior to analysis to meet the assumptions of the twoway ANOVA.

A significance level of α=0.05 was used for all statistical tests, and sequential

Bonferroni adjustments were made for multiple comparisons. The linear mixed model,

77 sequential Bonferroni, and ANOVA tests were performed using the R environment version 2.14.1 (R Development Core Team 2012). The linear mixed model procedure was implemented with lmer() from the lme4 package (Bates et al. 2010). All other analyses were conducted in Excel.

Results

Diet Analysis

Of the total number of drum stomachs sampled, 202 (13.7%) were empty and 21

(1.4%) had no identifiable contents; the diets of the remaining 1254 drum 40760 mm TL were characterized. Excluding unknown items, the frequencies of diet items encountered were larval Chironomidae (23.8%), Dreissena (23.8%), microcrustaceans (19.3%), fish

(7.0%), Amphipoda (6.9%), crayfish (5.6%), Oligochaeta/Hirudinea (4.7%), Isopoda

(3.9%), nonDreissena Mollusca (2.8%), Trichoptera (1.8%), adult Chironomidae (0.5%) and Ephemeroptera (0.1%).

Most drum less than approximately 360 mm TL had consumed larval chironomidae and microcrustaceans, whereas most drum larger than approximately 360 mm TL had consumed zebra mussels ( Figure 4.2). Fish were also common in drum larger than approximately 360 mm TL, and crayfish were common in drum larger than approximately 540 mm TL. Other food items were only common in drum between approximately 200 and 400 mm TL. Zebra mussel shells observed within small freshwater drum were relatively small and few in number; larger drum were often found with their digestive tract completely engorged with crushed zebra mussel shells. The TL at which 50% or more of drum had consumed zebra mussels was 384 mm (r2=0.97)

(Figure 4.3). Thus, drum <384 mm TL were classified as small and ≥384 mm TL as

large .

larval Chironomidae Oligochaeta/Hirodinea zebra mussels 1 1 1 0.8 0.8 0.8 0.6 0.6 0.6 0.4 0.4 0.4 0.2 0.2 0.2 0 0 0 60 60 60 120 180 240 300 360 420 480 540 600 120 180 240 300 360 420 480 540 600 120 180 240 300 360 420 480 540 600

microcrustaceans Isopoda fish 1 1 1 0.8 0.8 0.8 0.6 0.6 0.6 0.4 0.4 0.4 0.2 0.2 0.2 Proportion 0 0 0 60 60 60 120 180 240 300 360 420 480 540 600 120 180 240 300 360 420 480 540 600 120 180 240 300 360 420 480 540 600

Amphipoda nonDreissena Mollusca crayfish 1 1 1 0.8 0.8 0.8 0.6 0.6 0.6 0.4 0.4 0.4 0.2 0.2 0.2 0 0 0 60 60 60 120 180 240 300 360 420 480 540 600 120 180 240 300 360 420 480 540 600 120 180 240 300 360 420 480 540 600

Total length (20 mm class)

Figure 4.2. The proportion of freshwater drum within 20 mm total length classes that had consumed items within each food category. Food items making up <2% of the diet are not shown.

small large 1

0.8

0.6

0.4 Proportion

0.2

0 60 100 140 180 220 260 300 340 380 420 460 500 540 580 620 Total length (20 mm class) Figure 4.3. The length class (384 mm TL) at which half of the drum were found to consume zebra mussels was used to create small and large size classes for the growth chronology and body condition analyses.

Growth Analysis

There were a total of 22,373 increment measurements from 1632 fish; ages

ranged from 258 and calendar years represented by growth increments ranged from

19482008. The most notable trends in the master growth chronology were a peak in

19871988, a sharp decline to 1992, and an increasing trend from 2001 to 2008 (Figure

4.4A, see Appendix C for the year effect values for the growth chronology).

The annual fluctuations of the master growth chronology were positively correlated to air temperatures. GDD base 10°C was positively correlated to the master chronology ( r=0.69, p<0.001) (Figure 4.4B ). Mean monthly air temperatures for summer months were also positively correlated with the master growth chronology (April r=0.40, p=0.0163; June r=0.41, p<0.0001; July r=0.57, p=0.0103; August r=0.48, p=0.0015; September r=0.47, p=0.0018), but not with any other months ( r<0.21, p>0.7821).

2007 2007

2003 2003

1999 1999

1995 1995

1991 1991

1987 1987

1983 1983

1979 1979 Year Year 1975 1975

1971 1971

1967 1967

1963 GDD, 10°C base 1963

1959 1959

1955 1955 Master

1951 1951

1947 1947 8 4 0 2 0 4 4 8 2 4 A. Year effect B. Standardized values

Figure 4.4. The master growth chronology of Lake Winnebago freshwater drum was plotted by year (A). Positive year effects indicate years when drum had relatively good growth, and negative year effects indicate years when drum had relatively poor growth. Year effects and growing degree days from a 10°C base (GDD) were standardized to a mean of 0 and a standard deviation of one (B). The vertical line marker at 1998 indicates when zebra mussels were first reported in Lake Winnebago. Confidence intervals represent +/ two standard errors.

The year effects of males (n=295) and females (n=852) were significantly different only in 1992 and 2005 ( Figure 4.5A ). In 1992, female growth was less and in 2005 female growth was greater than males. The year effects of small (<384 mm TL, n=1140) and large (≥384 mm TL, n=497) drum were not significantly different prior to 2001, but

growth of large drum increased while growth of small drum remained constant after 2000

such that growth differed significantly between the two groups in every year after 2000

(Figure 4.5B ).

Body Condition

A strong interaction effect between the size class (small and large) of freshwater

drum and time period (pre and postzebra mussel discovery) was evident on mean

relative weight ( F(2,6208)=757.2, p<0.0001). The mean relative weight of small drum

decreased between 2.5 and 4.5% (t=8.57, p<0.0001) from the pre to postzebra mussel

discovery periods. In contrast, the mean relative weight of large drum increased between

32.4 and 40.6% (t=26.15, p<0.0001) from pre to postzebra mussel discovery ( Figure

4.6).

2007 * 2007 * * * *

2003 * 2003 * * 1999 1999

1995 1995 * 1991 1991

1987 1987

1983 1983

1979 1979 Year Year 1975 1975

1971 1971

1967 1967 ≥ 384≥ mm TL

1963 1963 male

1959 1959

1955 1955 female <384 mm TL 1951 1951

1947 1947 8 4 0 8 4 0 4 8 4 8 16 12 16 12 12 16 12 16 A. Year effect B. Year effect

130 ≥384 mm TL <384 mm TL

120

110

100

Wr 90

60 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 Year

Discussion

This study has linked the consumption of zebra mussels by freshwater drum to increased growth rates and body condition. There is an adult lengthrelated diet shift; only large drum (≥384 mm TL) consume significant quantities of zebra mussels, while small drum (<384 mm TL) seldom consume them. Correspondingly, the growth chronology of large drum, but not small drum, signals the establishment of zebra mussels in Lake Winnebago, and the mean body condition of large drum, but not small drum, has increased from pre (19841997) to post (19982009) zebra mussel discovery, which signifies that zebra mussels are providing a significant source of nutrition. There was also a strong positive correlation between drum growth chronologies and regional temperatures, but the growth chronology of large drum increased and diverged from small drum after 2000, and therefore ruled out temperature being the dominant factor controlling the recent increase in growth rates. Wrenn (1968) reported a similar diet shift in freshwater drum in Alabama where Corbicula (Asiatic clams) had invaded. At 10 inches (254 mm TL), Alabama drum shifted from predominantly consuming dipteran larvae and copepods to gizzard shad and Corbicula .

Drum near the transitional stage from small to large size displayed the greatest diet diversity. There may be a physiological limitation that causes drum to stop consuming smaller prey items when they attain a certain size, or a morphological change that allows drum to utilize different resources. Food resource utilization in fish tends to relate to body size (Werner and Gilliam 1984); however, the diet shift in Lake

Winnebago drum is not likely explained by gape limitation. There may be a limit imposed by the development of their pharyngeal teeth. French III (1997) determined that

85 over 85% of drum have their villiform teeth replaced by molariform teeth by 265 mm TL.

In contrast, lake sturgeon ( Acipenser fulvescens ) rely on chironomids as a primary food resource at all body lengths. Perhaps either the mouth parts of lake sturgeon are more specialized to strain Chironomidae larvae from sediment or the gill rakers are better able to retain Chironomidae larvae. Unlike drum, sturgeon have a unique jaw protrusion mechanism that is highly specialized for benthic feeding (Carroll and Wainwright 2003).

If drum become less able to obtain Chironomidae larvae as they grow larger, then there likely is a learning curve for drum to find and consume alternative food items.

This study has demonstrated that a fish growth chronology can record a distinct signal such as from the establishment of an invasive species, which was reflected by body condition. Impacts of dreissenid establishment have been predominantly reported at lower trophic levels (e.g., Fahnenstiel et al. 2010; Hecky et al. 2004; Idrisi et al. 2001;

Malkin et al. 2008; Mayer et al. 2000; Ozersky et al. 2009; Richardson and Bartsch 1997;

Strayer and Malcom 2006), and dreissenids have only been linked indirectly to fish growth (e.g., Bartsch et al. 2003; Giuliano 2011; Johannsson et al. 2011; Karataev and

Burlakova 1995; Karatayev et al. 1997; Pothoven and Madenjian 2008; Strayer et al.

2004). If the improved body condition of large drum is permanent, then the establishment of dreissenids could affect many additional life history traits of Lake

Winnebago drum. There is no evidence as of 2011 that the increased growth rates are beginning to decline. Zebra mussels are either providing a more energyrich diet than the previously available food items, or zebra mussels have not yet become a limiting resource in Lake Winnebago.

Growth chronologies developed from stable anatomical structures should be

86 considered a valuable tool for fisheries management programs for several reasons.

Traditional fisheries growth models often include age effects, thereby obscuring annual variations in growth patterns and making it difficult to detect environmental factors that govern growth. Removal of age effects make it easier to observe annual changes in growth rates. If age effects had not been removed in this study, the increased growth rates of large drum may not have been detected by examining cross sectional samples of body lengths at age. In addition, the effects of zebra mussel establishment may not have been detected if the growth chronology had not been examined by size classes defined by a diet shift. Further, growth rates over many years can be determined from a single sampling event to obtain age determination structures, and therefore in some situations can address the problem of afterthefact data acquisition and avoid the costs associated with a rigorous sampling protocol to obtain annual measurements from marked and recaptured fish. In the absence of empirical body measurements, a growth chronology can ultimately help improve our understanding of ecosystem relationships.

CHAPTER FIVE: CONCLUSIONS

LINKING IMPACTS OF ROUGH FISH REMOVAL AND OTHER CHANGING

ENVIRONMENTAL FACTORS TO THE FRESHWATER DRUM POPULATION

IN LAKE WINNEBAGO

The chapters of this dissertation have described results using new tools to understand the freshwater drum population in Lake Winnebago. These tools were:

• calculation of demographic statistics using validated ages (Chapter 2 ),

• the examination of the longterm changes in longevity and growth rates by

comparing modern to archaeological drum ( Chapter 3 ), and

• the development of a growth chronology of drum dating back to 1948, and

calculating growth chronologies by sex and size subsets ( Chapter 4 ).

These approaches have revealed changing statistics in this population. In Chapter

3, I reported that Lake Winnebago drum from pre and postEuropean settlement had similar body length potential, but the historical populations had significantly slower growth rates and greater longevity. In Chapter 4, I reported that the growth rates of large drum (≥384 mm total length) have increased corresponding to the discovery of zebra mussels in Lake Winnebago.

These changes are indicative of major ecosystem disturbances, and the cause of the changes reported in Chapter 4 is more straightforward than in Chapter 3. The increase in growth rates of large drum, but not small drum, every year since 2001 corresponding to the probable time of zebra mussel establishment in Lake Winnebago, along with increased body condition, provides strong evidence of a cause and effect relationship. However, cause(s) of the changes in longevity and growth rates observed

88 throughout the entire drum population post European settlement are not nearly as clear cut because there have been many changes on Lake Winnebago since the arrival of

Europeans: higher water levels and subsequent wetland loss, the advent of largescale agricultural practices and urban growth, and the addition of municipal and industrial wastewater influents, etc. (Wisconsin Department of Natural Resources 1989; 1993;

2004; see Chapter 1). These factors have undoubtedly affected the water quality of Lake

Winnebago, but the specific connections between drum demography and water quality can only be speculated. For example, drum are often reported to tolerate turbid water better than other fish species (Fremling 1980; Matthews 1984).

As discussed in Chapter 3, the rough fish removal program could have had effects on the population densities and therefore also on the age composition and length distribution of drum. It is possible that six decades of intensive exploitation cropped off the largest and oldest individuals, and compensatory growth responses from reduced population densities would explain the increase in growth rates of modern drum (Ricker

1975). In support of the effectiveness of the rough fish removal program, catch per unit effort (CPUE) of adult drum has increased over the assessment trawling period from

19862010 ( r=0.43, p=0.032, see Appendix A ). Further, the reduced longevity of

modern drum compared to archaeological drum further supports the effectiveness of the

removal program because reduced longevity is indicative of increased annual mortality

rates.

Another major change in drum population statistics that was not reported in the previous chapters is that there has been shift in the length distributions during assessment trawling from 19862010 (see Appendix D ). The proportion of large freshwater drum

89 captured in Lake Winnebago during fall assessment trawling increased appreciably from

19862010. Drum larger than 480 mm total length were never captured during assessment trawling from 19862002. In addition, the age composition of the drum population appears to be shifting towards older individuals, but this assessment is biased by the influence of the strong 1983 year class (see Appendix E ). The appearance of older and larger drum could be evidence of recovering from the impacts of removal trawling two decades after the program ended.

The recent appearance of large drum in Lake Winnebago provides a clue, but does not provide a definitive answer for assessing potential impacts of removal trawling.

Firstly, it is unknown whether or not early rough fish removal efforts on the lake encountered large drum. There are no records of drum lengths at the time of the earliest removal efforts beginning in the early 1930’s, so there is no baseline data to compare the current population statistics to. Second, the timing of the arrival and establishment of zebra mussels coincides with the appearance of large drum in the lentic areas of Lake

Winnebago. Zebra mussels evidently provide enough nutrition to enhance drum growth rates, but it is unlikely that they cause drum to grow quickly enough to achieve the large sizes that were abruptly observed during lake assessment trawling. A more probable explanation is that movements of large drum, which are locally known to reside in riverine areas, were constrained due to diet limitations, but have dispersed in pursuit of more abundant or easily acquired resources as zebra mussels became available. This behavior would be consistent with ideal free distribution theory (Fretwell and Lucas

1970). Under this scenario, large drum were always present in the Lake Winnebago system, but they did not inhabit the lentic portion of the ecosystem. This movement

90 pattern of large modern drum appears to be consistent with the behavior of large archaeological drum (by examining length distributions of drum at each archaeological site); at least until the establishment of zebra mussels (see the Doty Island subset in

Chapter 3 ).

One last, and important, analysis is considered. Catch per unit effort (CPUE) data

have been collected by standardized trawling methods since 1986 (see Appendix A ).

There have been no clear changes in CPUE trends over time. A stock recruitment analysis (see Appendix F ) was attempted in order to examine correlations between the

adult drum population size and the number of age0 drum. Adult drum CPUE’s and age

0 CPUE’s were used from 19862010. According to this analysis, drum abundance never

declined enough to affect recruitment rates, perhaps explaining why reduced densities

were never detected. This analysis should be considered extremely crude because the

data did not conform to assumptions for this analysis and the confidence intervals were

wide.

When compared to growth and survival of drum that lived prior to European

settlement, some of the strongest evidence supports that modern drum have undergone

compensatory growth in response to reduced population densities brought on by the

rough fish removal program. However, correlation does not imply causation. Other

contemporary changes, particularly the arrival and establishment of a new food resource

(zebra mussels) one decade after the termination of the rough fish removal program, preclude a foregone conclusion. The major confounding factors are that the exact timing

of the demographic changes is unknown, and water quality has not been monitored over a

long enough time span. Definitively determining whether or not removal trawling had an

91 impact on the freshwater drum population would be useful because other environmental factors would become more significant if we rule out effects of the removal program. It should be noted that my research was not focused on determining potential impacts of removal trawling, and thus has not used every tool available to assess changes in the freshwater drum population statistics. There is still work that could be done.

In conclusion, the freshwater drum population of Lake Winnebago has undergone significant changes in population statistics between pre and post European settlement, and the reasons for these changes are undetermined. The changes are most likely the effects of multiple factors, but too much time has spanned without the availability of monitoring data to examine correlations with many of the potential contributing factors.

Drum otoliths collected between 1850 and 1950 would potentially provide key information.

In summary, freshwater drum have important traits conducive to monitoring ecosystem changes. My research has demonstrated that a drum growth chronology developed from otoliths can record important environmental changes, such as annual temperature fluctuations and the establishment of invasive species reported here.

Combined with this fact, drum are an optimal species for growth chronology studies.

They have great longevity; their lifespan is greater than many other members of their fish community. They have unusually large sagittal otoliths, which makes them easy to obtain. The large size of their otoliths also contributes to the clarity of their increments, aiding accuracy in age determinations and resolution for measurements. Their large otoliths commonly persist years after an individual is deceased, and therefore are often recovered from archaeological sites. The archaeological otoliths have potential to

92 provide a longterm historical growth chronology, which opens up the avenues for a multitude of studies. Lastly, although they were heavily exploited until 1990, drum in

Lake Winnebago are now under relatively little fishing pressure, meaning that future populations changes will more directly reflect environmental impacts other than exploitation.

The one thing that is certain in an uncertain world is that change will continue.

The freshwater drum growth chronology will continue to reflect changes in its environment, and detecting these connections will help us to better understand the ecosystem. Freshwater drum should be regularly monitored in the Lake Winnebago system because they have major ecosystem interactions and make up the largest component of the fish community. And, considering the ease of collection and the great amount of information that can potentially be acquired, freshwater drum sagittal otoliths should continue to be collected from all body length ranges on an annual basis.

APPENDICES

Appendix A: Area Swept Population Estimate of Lake Winnebago Freshwater Drum

Population abundance of freshwater drum in Lake Winnebago was calculated via the swept area method (Alverson and Pereyra 1969; King 1995) based on catch per unit effort (CPUE) from assessment trawling for one week in August, September, and October each year from 1986 to present (see Chapter 1 ). The swept area method assumes that vulnerability to the trawl is 100% (i.e. no drum evaded the trawling net) (Beamish 1969), and that the species sampled is within 0.9 meters of the bottom 100% of the time.

However, drum may not be near the bottom 100% of the time, so population estimates should be considered conservative. The area swept per cast of the net was one acre (2.47 hectares).

Mean CPUE for each year was calculated by dividing the total number of drum captured by the total number of casts. The mean CPUE was then multiplied by the trawlable area (depths > 3.7 meters) of the lake and the entire area of the lake, 110,200 acres (44,596 hectares) and 137,708 acres (55,729 hectares), respectively. The population estimate for the entire area of the lake assumes that drum are equally distributed between the trawlable area and entire area of the lake.

The total pounds of adult freshwater drum in Lake Winnebago ranges between 45 and 127 million lbs (20.4 and 57.6 million kg) with a mean of 73.4 million pounds (33.3 million kg) ( Table A.1 ). The confidence intervals were based on low and high CPUE’s

from assessment trawling from 19862010. August and September CPUE’s were used because they are statistically in agreement, t(23)=1.87, p=0.075; October CPUE’s were

significantly different from August CPUE’s, t(23)=6.98, p<0.001, and September

CPUE’s, t(23)=5.76, p<0.001.

Table A.1. Catch per unit effort (CPUE) was calculated for Lake Winnebago freshwater drum from assessment trawling conducted from 19862010 and used to estimate abundance within the lake. 469.30 CPUE 370.30 354.96 413.42 640.01 481.45 496.64 508.38 471.42 137,708 110,200 7 64,626,621.89 41,452,103.77 51,717,066.06 112,073,203.47 138 138 138 138 138 136 138 138 of casts of 0 91 343.63 ater number drum 51,102 48,984 57,052 88,321 66,440 67,543 70,157 65,056 17,443 89 43,283 50,462 138 138 313.64 365.67 78 53,955 41,540 138 138 390.98 301.01 5 46,719 112 417.13 86 65,812 126 522.32 9192 51,671 78,340 13794 138 377.16 62,49496 567.68 135 74,101 462.92 13899 63,640 536.96 01 13802 59,285 111,497 461.16 138 137 429.60 813.85 06 92,74808 13809 72,659 672.09 62,682 136 129 534.26 485.91 SD year 1987 1993 1995 2000 2003 2004 2007 2010 338.05 80.87 CPUE 133.63 269.43 434.28 382.28 228.95 532.98 381.00 110,200 137,708 37,253,559.67 46,552,751.32 11,136,386.09 98,628,864.52 46 46 46 46 46 44 46 46 of casts of 1 44.21 338.05 mean 63,072.52 133.88 469.30 3720 drum 6,147 7,693 12,394 19,977 17,585 10,074 24,517 17,526 sd year 1987 1993 1995 2000 2003 2004 2007 2010 510.87 CPUE 432.63 435.41 439.98 733.17 469.57 693.54 541.11 528.28 110,200 137,708 56,297,607.06 70,350,552.38 45,003,573.13 46 46 46 46 46 46 46 46 of casts of drum 6,496 19,901 20,029 20,239 33,726 21,600 31,903 24,891 24,301 sd year 1987 1993 1995 2000 2003 2004 2007 2010 September October Overall 555.12 110,200 137,708 CPUE 597.41 495.83 530.85 752.57 592.50 555.78 451.07 504.98 45,150,262.09 61,174,273.35 76,444,526.63 110,866,914.61 127,049,400.80 46 46 46 46 46 46 46 46 of casts total acres total low estimate low high estimate high trawlable acres trawlable drum 6,157 27481 22808 24419 34618 27255 25566 20749 23229 freshwater number freshwater number freshwater number freshw (using area swept per cast = 1 acre) 1 cast per = swept (using area mean no. adults per acre per adults mean no. n 25 25 n 25 25 n 24 24 n 25 25 sd min 15,082 328 min 15,033 327 min 3,720 81 min 31,270 301 max 37,034 805 max 41,517 923 max 32,946 716 max 111,497 814 year sum 608,114 1,096 sum 584,289 1,144 sum 361,181 180 sum 1,576,813 3,34 19861987 264301988 42 16024 46 629.291993 348.351994 19861995 24212 20,2271996 1988 43 32503 42 19845 481.60 46 563.07 462000 2001 1986 431.41 706.59 24273 1994 191552003 1988 27,0032004 46 1996 422005 7414 46 29,346 456.07 23398 527.672007 587.02 46 462008 19 46 161.17 637.96 1994 30570 20012010 11,279 198 508.65 15,033 46 1996 46 46 12,252 2005 664.57 245.20 326.80 46 18,875 19 266.35 2001 2008 46 19,979 19,807 19 410.33 46 46 2005 434.33 430.59 4,446 20 2008 20 22,282 222.30 44 200 506.41 20 19891990 235781991 150821992 46 23968 46 34815 45 512.57 46 327.87 532.621997 1989 756.851998 20066 1990 18,9241999 16355 1991 16,188 46 15634 1992 46 16,209 46 45 24,2162002 411.39 46 46 436.22 359.73 37034 46 355.54 352.37 1989 339.87 526.43 46 1990 1997 7,9602006 1991 1998 24,791 1992 na 33321 1999 11,494 805.09 16,765 46 46 19,309 22,0892009 46 173.04 46 46 na 538.93 27955 46 46 2002 249.87 364.46 198 724.37 41,517 419.76 na 480.20 46 1997 19 1998 45 9,098 19 1999 2006 1990 607.72 8,420 922.60 25,917 34,289 46 31,27 46 2002 46 197.78 46 2009 183.04 32,946 22,575 563.41 745.41 199 199 46 46 19 2006 716.22 490.76 25,138 20 2009 46 12,152 546.48 37 20 328.43 20 mean 25,338.08 45.67 555.12 mean 23,371.56 45.76 510.87 mean 15,049.2 August No. of adultNo. total of drum in acres No. of adult drum in trawlable area trawlable drumNo. in adult of CPUE’s of adult drum combined from all months have significantly increased

from 19862010 ( r=0.43, p=0.032) ( Figure A.1 ). However, there is not a significant

96 increase when any single of month of assessment trawling adult CPUE’s are examined.

This significant positive relationship suggests that the freshwater drum population has increased over this 16 year period and that it may have increased because the removal trawling is no longer controlling the population size.

800

700

600

500

400 Adult drum CPUE Adult

300

200 1986 1989 1992 1995 1998 2001 2004 2007 2010 Year

Figure A.1. The mean catch per unit effort (CPUE) of adult Lake Winnebago freshwater drum from assessment trawling conducted in August, September, and

October each year increased from 19862010.

Appendix B: The Linear Mixed Effects Model

Growth chronologies were developed from otolith increment measurements using

a mixedeffects model (Weisberg et al. 2010). This model was selected over the former

fixedeffects linear model (Weisberg 1993) to overcome three shortcomings. First, the

earlier model assumes that the increments are independently distributed, which is

unnatural. In reality, biological growth is likely to have withinfish autocorrelations in

increment growth and autocorrelations among fish in a population for any given year.

Second, the year effect and age effect parameters were not unique because they were

aliased to an arbitrary additive constant. Comparisons within a single study were

feasible, but comparisons to other chronologies would not be meaningful. Third, an

interaction term is not interpretable in the fixed effects model.

In the basic mixedeffects linear model (Equation 1), an increment width ( y) is

decomposed into four terms: the extrinsic random environmental (or year) effect ( η ),

intrinsic random fish effect ( f), intrinsic fixed age effect ( ι), and an additive error term ( e).

y = ι + η + f + e (1) cka a c+a−1 ck cka

In this equation, a represents the ath year of life, c represents the yearclass, and k represents the kth fish within the cth yearclass. The term f allows each fish to have its own overall level of growth that applies to all increments on the fish; i.e., a fish with slower growth would be assigned a lower individual growth value than a fish with faster growth.

Equation 1 can be adjusted to account for randomeffect interactions (Equation 2).

y =ι + η + f + (ιη) + e (2) cka a c+a−1 ck a,c+a−1 cka

Equation 2 is only used if the interaction between age and year are significant. More specifically, it will only be included for instances where it is represents 10% or more of the variance (S. Weisberg, pers. comm.). ANOVA and Akaike’s information criterion

(AIC) values are used to select the most appropriate model.

The Weisberg et al. (2010) model is solved using matrix algebra (Equation 3): Y X Z = β + + εˆ (3) u where the terms in bold represent matrices or vectors. The Y vector contains the

increment measurements for each fish arranged in a column. The number of entries is

equal to the sum of the ages of all the fish, as each fish will have an increment

measurement for each year of its life. Let the sum of the ages of all the fish equal N.

The X matrix contains a number of columns that is equal to the oldest fish and number of rows equal to N. This matrix contains mostly zeros; there is a single number one in each row. The one is located in the column representing the age at which the corresponding increment measurement (located in the Y vector) was made. For example,

the matrix values for a three year old fish in a data set in which the maximum age is four

would look like:

1 0 0 0 0 1 0 0 0 0 1 0 0 0 0 0 The β vector is the estimated intrinsic, or age effect, coefficients. The age effects are considered fixed effects because the ages are not a representative sample; all of the ages are present. Ages also have a systematic pattern of responses that decrease with age associated with the entire population. The restricted maximum likelihood algorithm is used to make estimates of fixed effects.

The number of columns in the Z matrix is equal to the total number of years in the data set (from the earliest year class to the most recent year a fish was caught), m, plus the total number of fish, t. The Z matrix is also mostly zeros. There are two ones in each row; one corresponds to the year at which the corresponding increment measurement

(located in the Y vector) was made, and the second one corresponds to the individual fish that the measurement was taken from. For example, consider a three year old fish of the yearclass that is the second year of a data set that spans five years. It is also the second fish in a data set of three fish. The matrix values would be as follows:

0 1 0 0 0 0 1 0

0 0 1 0 0 0 1 0 0 0 0 1 0 0 1 0

The vector u contains the extrinsic or predicted coefficients for year effects (one is assigned to each year) and fish effects (one is assigned to each fish). The length

(number of rows) of u is equal to the number of columns in the Z matrix ( m + t ). The first m entries contain the year effects and the next t entries contain the fish effects. Year effects are classified as random effects because the years are considered sampled from all possible years. Years are modeled as random draws from the same normal distribution with a mean of zero. Fish effects are also considered random because the fish are a sample from the population. Fish are modeled as random draws from the same normal distribution with mean zero. A best linear unbiased prediction (BLUP) algorithm is used to make the predictions for fish effects.

The predicted residual errors are represented by εˆ . The three components of the model (age effects, year effects, and fish effects) add up to the predicted growth increment for each fish. The residual errors are calculated as the difference between the

100 observed values and the predicted values.

101

Appendix C: The Lake Winnebago Freshwater Drum Growth Chronology Index

The growth chronology index is presented in Table C.1 . The values are year effects that were calculated using the Weisberg et al. (2010) linear mixed model (see

Chapter 3 ). The master chronology was calculated with all data combined, and the chronologies calculated from small (<384 mm TL) and large drum (≥384 mm TL ) subsets are also presented.

Table C.1. The growth chronology indices from Lake Winnebago freshwater drum were calculated by using the Weisberg linear mixed model. The year effect values and their standard errors ( s.e. ) are reported. All combined Small (<384 Large (≥384 (master) Females Males mm TL) mm TL) Year Year Year Year Year Year effect s.e. effect s.e. effect s.e. effect s.e. effect s.e. 1948 0.799 2.001 3.789 2.744 0.225 1.592 0.913 1.859 2.796 3.234 1949 1.097 2.001 3.287 2.744 0.089 1.592 1.365 1.859 3.080 3.234 1950 2.752 2.001 1.461 2.744 1.344 1.592 1.665 1.859 6.992 3.234 1951 0.504 2.001 0.688 2.744 0.268 1.592 0.072 1.859 5.000 3.234 1952 1.403 1.802 4.223 2.227 0.236 1.557 1.079 1.743 1.024 2.692 1953 1.744 1.718 2.076 2.138 0.803 1.525 0.680 1.743 4.763 2.356 1954 0.822 1.718 1.870 2.138 0.002 1.525 0.333 1.743 1.895 2.356 1955 3.437 1.575 5.194 1.925 1.301 1.473 1.595 1.670 3.580 1.945 1956 3.337 1.575 3.750 1.925 1.598 1.473 1.248 1.670 8.559 1.945 1957 0.177 1.542 1.298 1.867 0.171 1.473 0.422 1.670 3.214 1.872 1958 2.082 1.489 2.605 1.813 0.375 1.446 0.875 1.633 5.060 1.806 1959 1.788 1.345 2.775 1.640 1.427 1.338 1.951 1.501 0.969 1.368 1960 1.998 1.309 1.075 1.640 0.132 1.338 0.746 1.501 1.108 1.318 1961 1.889 1.309 0.882 1.640 0.454 1.338 0.952 1.501 2.257 1.318 1962 0.931 1.309 0.244 1.640 0.886 1.338 1.565 1.501 2.080 1.318 1963 1.354 1.211 0.809 1.386 0.545 1.269 1.030 1.383 4.041 1.197 1964 2.368 1.159 1.948 1.364 1.495 1.231 1.569 1.347 0.447 1.132 1965 0.475 1.159 0.422 1.364 0.393 1.231 0.347 1.347 1.646 1.132 1966 0.384 1.091 0.270 1.168 0.826 1.196 0.063 1.257 0.512 1.090 1967 0.126 1.072 1.305 1.155 1.151 1.196 1.844 1.257 1.333 1.064 1968 0.963 1.061 2.034 1.142 0.542 1.196 1.228 1.257 0.382 1.052 1969 0.497 1.023 0.244 1.086 0.771 1.172 0.639 1.230 2.373 0.997 1970 1.703 0.964 2.868 0.920 1.499 1.108 0.783 1.142 3.060 0.894 1971 0.860 0.932 1.971 0.895 1.461 1.108 0.525 1.121 0.930 0.853 1972 1.296 0.932 2.618 0.895 0.826 1.108 1.086 1.121 2.432 0.853 1973 0.013 0.896 0.409 0.845 0.699 1.075 0.395 1.070 0.343 0.829 1974 0.335 0.875 0.639 0.831 0.087 1.075 0.269 1.060 1.096 0.802 1975 0.465 0.844 1.078 0.788 0.284 1.052 0.520 1.014 0.242 0.787 1976 0.623 0.828 0.682 0.781 0.466 1.036 0.268 0.992 0.635 0.782 1977 0.112 0.828 0.099 0.781 0.379 1.036 0.449 0.992 0.624 0.782 1978 1.087 0.820 1.744 0.777 0.542 1.036 0.572 0.992 1.929 0.772 1979 2.003 0.812 2.649 0.777 1.012 1.036 1.388 0.984 2.664 0.768

102

All combined Small (<384 Large (≥384 (master) Females Males mm TL) mm TL) Year Year Year Year Year Year effect s.e. effect s.e. effect s.e. effect s.e. effect s.e. 1980 1.529 0.791 1.811 0.757 0.628 1.010 0.879 0.954 1.466 0.759 1981 1.049 0.778 0.982 0.743 0.081 1.010 0.995 0.947 0.736 0.742 1982 0.875 0.769 1.650 0.743 0.890 1.010 0.584 0.941 0.860 0.734 1983 0.516 0.739 1.090 0.460 0.509 0.965 1.527 0.892 2.773 0.446 1984 2.957 0.713 4.692 0.458 1.912 0.957 3.442 0.857 4.589 0.385 1985 2.461 0.708 3.044 0.457 1.164 0.957 1.466 0.857 3.604 0.384 1986 0.811 0.854 2.238 0.510 0.060 1.451 0.191 1.348 1.286 0.396 1987 2.311 0.822 6.997 0.494 1.938 1.393 2.502 1.246 4.042 0.392 1988 2.845 0.788 5.091 0.451 0.658 1.277 3.635 1.146 4.959 0.378 1989 0.643 0.761 3.694 0.445 0.357 1.210 0.581 1.083 0.724 0.371 1990 2.416 0.742 3.758 0.438 0.644 1.195 1.085 1.038 2.938 0.368 1991 1.344 0.719 1.406 0.421 1.934 1.171 0.746 0.986 1.849 0.365 1992 5.007 0.709 8.963 0.420 2.196 1.157 5.428 0.978 4.768 0.363 1993 2.978 0.709 1.994 0.420 1.244 1.157 1.026 0.978 3.711 0.363 1994 1.313 0.698 2.095 0.417 0.717 1.144 0.213 0.955 1.489 0.363 1995 0.620 0.677 2.205 0.392 2.090 1.097 2.591 0.908 0.985 0.356 1996 1.425 0.661 3.146 0.391 0.287 1.065 0.611 0.883 0.591 0.350 1997 3.193 0.659 4.048 0.390 1.271 1.065 1.972 0.887 2.191 0.350 1998 0.810 0.640 2.964 0.362 1.952 1.018 2.646 0.847 0.307 0.347 1999 0.943 0.626 1.565 0.349 1.738 0.978 2.635 0.814 1.375 0.341 2000 1.122 0.611 2.000 0.346 0.327 0.953 0.113 0.791 0.158 0.338 2001 0.932 0.600 1.299 0.342 1.222 0.941 1.158 0.770 1.685 0.337 2002 1.224 0.587 0.589 0.335 1.217 0.915 0.680 0.746 4.464 0.337 2003 1.634 0.575 0.409 0.327 0.119 0.884 1.093 0.722 6.239 0.337 2004 0.546 0.579 1.547 0.326 1.029 0.868 1.418 0.711 5.082 0.350 2005 4.415 0.573 6.599 0.316 2.728 0.847 3.098 0.699 8.354 0.359 2006 4.161 0.582 0.531 0.339 2.230 0.848 1.045 0.692 12.149 0.391 2007 4.307 0.600 4.148 0.471 1.415 0.833 1.145 0.723 12.193 0.417 2008 3.831 0.597 4.208 0.486 1.312 0.838 2.128 0.716 10.199 0.440

103

Appendix D: Length Distribution Trends of Lake Winnebago Freshwater Drum from 19862011

Total lengths of Lake Winnebago freshwater drum (n=21,891) were measured

during October assessment trawling from 1986 to 2011. These samples were randomly

selected from trawl captures. Detailed assessment trawling methods are provided in

Chapters 2, 3, and 4). Young of the year (approximately less than 177 mm TL) were not measured every year. Drum larger than 480 mm TL were never captured during assessment trawling from 19862002. The peak body length of drum under 400 mm TL gradually decreased from 19862010. The peaks at 250 mm in 1986 and at 280 mm in

1987 are the strong 1983 year class.

Figure D.1. Length frequencies of freshwater drum measured using random selection during October Lake Winnebago assessment trawling.

104

Appendix E: Year Class Strength of Lake Winnebago Freshwater Drum

Ages of freshwater drum from 1986 and from 2003 to 2010 (n=2004) were used to examine year class strength. Many of the ages came from drum captured by stratified subsampling, but these graphs depict strong year classes because of the poor relationship between body length and age in freshwater drum. An age error key was not used because of this poor relationship. The strong 1983 year class has remained predominant for more than 27 years.

1986 2003 2004

40 20 Frequency Frequency 0 10 20 30 0 40 80 0

1940 1960 1980 2000 1940 1960 1980 2000 1940 1960 1980 2000 Year Class Year Class Year Class

2005 2006 2007 30 20 10 Frequency Frequency 0 20 60 0 10 20 30 0

1940 1960 1980 2000 1940 1960 1980 2000 1940 1960 1980 2000 Year Class Year Class Year Class

2008 2009 2010 60 40 20 Frequency Frequency 0 20 40 0 20 60 0

1940 1960 1980 2000 1940 1960 1980 2000 1940 1960 1980 2000 Year Class Year Class Year Class

Figure E.1. Year class strengths of freshwater drum spanning approximately 60 years.

105

Appendix F: Stock Recruitment Analysis of Lake Winnebago Freshwater Drum

Stock recruitment relationships were examined using Lake Winnebago assessment trawling catch per unit effort (CPUE) from 19862010 (see Chapter 4 for assessment trawling methods). The assessment trawling methods separated age0 and adult drum; age0 drum were roughly defined as individuals < 162 mm total length and adult drum were individuals ≥ 162 mm total length. For this assessment, age0 drum

CPUE’s were defined as recruits, and adult drum CPUE’s were defined as stock. Mean age0 CPUE between September and October was used as recruits because paired ttests between CPUE’s by year indicated no significant difference between these two months

(t(23)=0.72, p=0.474 between Sept and Oct; t(23)=4.74, p<0.001 between Aug and

Sept; and t(23)=3.36, p=0.003 between Aug and Oct). The mean adult CPUE between

August and September was considered stock because there was no significant difference between those two months (see Appendix A ). Parameters of the Beverton Holt (1957) and Ricker (1975) models were calculated. The stock recruitment models indicated that the adult population had never declined enough to cause a decline in recruitment. With log transformation, the data also did not appear to fully meet normality requirements as residual values were skewed. In the BevertonHolt model, the parameters a and b resulted in nonsensical negative values, a=0.1032 (3.1511 0.0827) and 0.0042 (

0.080– 0.0033) (95% confidence limits were found with bootstrapping, 200 iterations).

The Ricker model did not provide as good of a fit as the BevertonHolt model as indicated by AIC values, but a and b were positive values a=0.4835 (0.19661.6940) and b=0.0022 (0.00040.0050). The highest recruitment levels occurred at 412.2 (196.5

1873.5) stock (age 1+) per acre.

106 recruits Recruits(age0 CPUE) 0 50 100 150 200 250 300

400 500 600 700 800

Parental (spawner)stock stock (adult CPUE)

Figure F.1. Lake Winnebago freshwater drum adult catch per unit effort (CPUE) and age0 CPUE’s were collected from fall assessment trawling from 19862010.

107

LITERATURE CITED

Alverson, D., and W. Pereyra. 1969. Demersal fish explorations in the northeastern Pacific Ocean—an evaluation of exploratory fishing methods and analytical approaches to stock size and yield forecasts. Journal of the Fisheries Board of Canada 26(8):19852001. Anderson, T. J. 2010. Secondary production of Chironomidae in a large eutrophic lake. Master of Science Thesis. University of Wisconsin Oshkosh. Andrews, A. H., L. A. Kerr, G. M. Cailliet, T. A. Brown, C. C. Lundstrom, and R. D. Stanley. 2007. Age validation of canary rockfish (Sebastes pinniger) using two independent otolith techniques: leadradium and bomb radiocarbon dating. Marine and Freshwater Research 58(6):531541. Baker Jr, M. S., and C. A. Wilson. 2001. Use of bomb radiocarbon to validate otolith section ages of red snapper Lutjanus campechanus from the northern Gulf of Mexico. Limnology and Oceanography 46(7):18191824. Barber, M., and G. Jenkins. 2001. Differential effects of food and temperature lead to decoupling of shortterm otolith and somatic growth rates in juvenile King George whiting. Journal of Fish Biology 58(5):13201330. Barney, R. L. 1926. The distribution of the freshwater sheepshead, Aplodinotus grunniens Rafinesque, in respect to the glacial history of North America. Ecology 7:351364. Bartsch, L., W. Richardson, and M. Sandheinrich. 2003. Zebra mussels ( Dreissena polymorpha ) limit food for larval fish ( Pimephales promelas ) in turbulent systems: a bioenergetics analysis. Hydrobiologia 495(1):5972. Bates, D., M. Maechler, and B. Dai. 2010. lme4: linear mixedeffects models using S4 classes. CRAN Rproject. R Foundation for Statistical Computing., Vienna, Austria. Baty, F., and M. L. DelignetteMuller. 2011. nlstools: tools for nonlinear regression diagnostics. http://www.rproject.org . Beamish, F. 1969. Photographic observations on reactions of fish ahead of otter trawls. Food and Agriculture Organization Fishery Report 62:511522. Beamish, R. J., and G. A. McFarlane. 1983. The forgotten requirement for age validation in fisheries biology. Transactions of the American Fisheries Society 112(6):735 743. Becker, G. 1964. The fishes of lakes Poygan and Winnebago. Transactions of the Wisconsin Academy of Sciences, Arts and Letters 53:2952. Becker, G. C. 1983. Fishes of Wisconsin. University of Wisconsin Press Madison. Behm, J. A. 2008. The Meskwaki in Eastern Wisconsin: Ethnohistory and Archaeology. The Wisconsin Archeologist 89(1&2):785. Berkman, P., D. Garton, M. Haltuch, G. Kennedy, and L. Febo. 2000. Habitat shift in invading species: zebra and quagga mussel population characteristics on shallow soft substrates. Biological Invasions 2(1):16. Berner, L. 1951. Limnology of the lower Missouri River. Ecology 32(1):112. Beverton, R., and S. Holt. 1957. On the dynamics of exploited fish populations. Fishery Investigations Series II Volume XIX, Ministry of Agriculture, Fisheries and Food Fishery Investment, Series 2, No. 19, 1957 19.

108

Beverton, R., and S. Holt. 1959. A review of the lifespans and mortality rates of fish in nature, and their relation to growth and other physiological characteristics. CIBA Foundation Colloquia on Ageing 5:142180. Black, B., G. Boehlert, and M. Yoklavich. 2008. Establishing climategrowth relationships for yelloweye rockfish ( Sebastes ruberrimus ) in the northeast Pacific using a dendrochronological approach. Fisheries Oceanography 17(5):368379. Black, B. A., R. J. Allman, I. D. Schroeder, and M. J. Schirripa. 2011. Multidecadal otolith growth histories for red and gray snapper (Lutjanus spp.) in the northern Gulf of Mexico, USA. Fisheries Oceanography . Blackwell, B. G., M. L. Brown, and B. R. Murphy. 1995. Development and evaluation of a standard weight ( Ws ) equation for freshwater drum. Prairie Naturalist 27(1):51 61. Bolduan, B. R., G. C. Van Eeckhout, H. W. Quade, and J. E. Gannon. 1994. Potamogeton crispus –the other invader. Lake and Reservoir Management 10(2):113125. Boogaard, M. A., T. D. Bills, and D. A. Johnson. 2003. Acute toxicity of TFM and a TFM/niclosamide mixture to selected species of fish, including lake sturgeon (Acipenser fulvescens ) and mudpuppies ( Necturus maculosu s), in laboratory and field exposures. Journal of Great Lakes Research 29:529541. Braaten, P. J., and C. S. Guy. 2004. Firstyear growth, condition, and sizeselective winter mortality of freshwater drum in the lower Missouri River. Transactions of the American Fisheries Society 133(2):385398. Bruch, R. M., S. E. Campana, S. L. DavisFoust, M. J. Hansen, and J. Janssen. 2009. Lake sturgeon age validation using bomb radiocarbon and knownage fish. Transactions of the American Fisheries Society 138(2):361372. Bur, M. T. 1984. Growth, reproduction, mortality, distribution, and biomass of freshwater drum in Lake Erie. Journal of Great Lakes Research 10:4858. Burr, B., and L. Page. 1986. Zoogeography of fishes of the lower Ohioupper Mississippi basin. Pages 287–324 in The Zoogeography of North American Freshwater Fishes . John Wiley & Sons. Burrill, A. C. 1913. Economic and biologic notes on the giant midge, Chironomus (Tendipes ) plumosus Meigen. Bulletin of the Natural History Society 10:124163. Butler, R. L. 1965. Freshwater drum, Aplodinotus grunniens , in the navigational impoundments of the Upper Mississippi River. Transactions of the American Fisheries Society 94(4):339349. Butler, R. L., and L. L. Smith Jr. 1950. The age and rate of growth of the sheepshead, Aplodinotus grunniens Rafinesque, in the upper Mississippi River navigation pools. Transactions of the American Fisheries Society 79(1):4354. Bystrom, P., and E. GarciaBerthou. 1999. Density dependent growth and size specific competitive interactions in young fish. Oikos 86(2):217232. Campana, S. E. 1983. Feeding periodicity and the production of daily growth increments in otoliths of steelhead trout ( Salmo gairdneri ) and starry flounder ( Platichthys stellatus ). Canadian Journal of Zoology 61(7):15911597. Campana, S. E. 1997. Use of radiocarbon from nuclear fallout as a dated marker in the otoliths of haddock Melanogrammus aeglefinus . Marine Ecology Progress Series 150(1):4956.

109

Campana, S. E. 2001. Accuracy, precision and quality control in age determination, including a review of the use and abuse of age validation methods. Journal of Fish Biology 59:197242. Campana, S. E., J. M. Casselman, and C. M. Jones. 2008. Bomb radiocarbon chronologies in the Arctic, with implications for the age validation of lake trout (Salvelinus namaycush ) and other Arctic species. Canadian Journal of Fisheries and Aquatic Sciences 65(4):733743. Campana, S. E., and C. M. Jones. 1998. Radiocarbon from nuclear testing applied to age validation of black drum, Pogonias cromis . Fishery Bulletin 96(2):185192. Campana, S. E., L. J. Natanson, and S. Myklevoll. 2002. Bomb dating and age determination of large pelagic sharks. Canadian Journal of Fisheries and Aquatic Sciences 59(3):450455. Campana, S. E., and J. D. Neilson. 1985. Microstructure of fish otoliths. Canadian Journal of Fisheries and Aquatic Sciences 42(5):10141032. Carroll, A., and P. Wainwright. 2003. Functional morphology of prey capture in the sturgeon, Scaphirhynchus albus . Journal of Morphology 256(3):270284. Casselman, J. M. 1987. Determination of age and growth. Pages 209–242 in S. Gill, editor. The Biology of Fish Growth. Academic Press, London. Caswell, H. 1978. A general formula for the sensitivity of population growth rate to changes in life history parameters. Theoretical Population Biology 14(2):215230. Charlton, M. N. 1994. The case for research on the effects of zebra mussels in Lake Erie: Visualization of information from August and September 1993. Journal of Biological Systems 2(4):467480. Chotkowski, M., and J. Marsden. 1999. Round goby and mottled sculpin predation on lake trout eggs and fry: field predictions from laboratory experiments. Journal of Great Lakes Research 25(1):2635. Choudhury, A., R. Bruch, and T. A. Dick. 1996. Helminths and food habits of lake sturgeon Acipenser fulvescens from the Lake Winnebago system, Wisconsin. American Midland Naturalist 135:274282. Christie, W. 1974. Changes in the fish species composition of the Great Lakes. Journal of the Fisheries Board of Canada 31(5):827854. Cochran, P., and P. Hesse. 1994. Observations on the white perch ( Morone americana ) early in its invasion of Wisconsin. Transactions of the Wisconsin Academy of Sciences, Arts and Letters 82:2531. Coffin, A., D. Pereira, and G. Spangler. 2003. Stockspecific growth rates of lake herring, Coregonus artedi , in western Lake Superior. Environmental Biology of Fishes 68(1):3948. Corkum, L., M. Sapota, and K. Skora. 2004. The round goby, Neogobius melanostomu s, a fish invader on both sides of the Atlantic Ocean. Biological Invasions 6(2):173 181. Cyterski, M., and G. Spangler. 1996. Development and utilization of a population growth history of Red Lake walleye, Stizostedion vitreum . Environmental Biology of Fishes 46(1):4559. Daiber, F. C. 1952. The food and feeding relationships of the freshwater drum, Aplodinotus grunniens Rafinesque in western Lake Erie. Ohio Journal of Science (Ohio Academy of Science) 52(1).

110

Daiber, F. C. 1953. Notes on the spawning population of the freshwater drum (Aplodinotus grunniens Rafinesque) in western Lake Erie. American Midland Naturalist :159171. Daugherty, D. J., and T. M. Sutton. 2005. Population abundance and stock characteristics of flathead catfish in the lower St. Joseph River, Michigan. North American Journal of Fisheries Management 25(3):11911201. DavisFoust, S. L., R. M. Bruch, S. E. Campana, R. P. Olynyk, and J. Janssen. 2009. Age validation of freshwater drum using bomb radiocarbon. Transactions of the American Fisheries Society 138(2):385396. Dermott, R., and M. Munawar. 1993. Invasion of Lake Erie offshore sediments by Dreissena , and its ecological implications. Canadian Journal of Fisheries and Aquatic Sciences 50(11):22982304. Dirst, V. 1985. Three class Oneota sites in East Central Wisconsin. University of Wisconsin Oshkosh, Department of Sociology and Anthropology. Drake, D., R. Naiman, and J. Helfield. 2002. Reconstructing salmon abundance in rivers: an initial dendrochronological evaluation. Ecology 83(11):29712977. Edsall, T. A. 1967. Biology of the freshwater drum in western Lake Erie. The Ohio Journal of Science 67(6):321. Eggleton, M., L. Miranda, and J. Kirk. 2004. Assessing the potential for fish predation to impact zebra mussels ( Dreissena polymorpha ): insight from bioenergetics models. Ecology of Freshwater Fish 13(2):8595. Evans, D. H., and J. B. Claiborne, editors. 2006. The Physiology of Fishes, Third edition. CRC Press, Boca Raton, FL. Everhart, W., and W. Seaman. 1971. Fishes of Colorado. Colorado Game, Fish, and Parks Division, Denver . Fahnenstiel, G., G. Lang, T. Nalepa, and T. Johengen. 1995. Effects of zebra mussel (Dreissena polymorpha ) colonization on water quality parameters in Saginaw Bay, Lake Huron. Journal of Great Lakes Research 21(4):435448. Fahnenstiel, G., T. Nalepa, S. Pothoven, H. Carrick, and D. Scavia. 2010. Lake Michigan lower food web: Longterm observations and Dreissena impact. Journal of Great Lakes Research 36(sp3):14. Farmer, G., and F. Beamish. 1973. Sea lamprey ( Petromyzon marinus ) predation on freshwater teleosts. Journal of the Fisheries Research Board of Canada 30:601 605. Fischer, A. J., M. S. Baker Jr, C. A. Wilson, and D. L. Nieland. 2005. Age, growth, mortality, and radiometric age validation of gray snapper ( Lutjanus griseus ) from Louisiana. US National Marine Fisheries Service Fishery Bulletin 103:307319. Foster, G., A. Hodgson, and M. Balarin. 1999. Effect of diet on growth rate and reproductive fitness of Turbo sarmaticus (Mollusca: Vetigastropoda: Turbinidae). Marine Biology 134(2):307315. Francis, R. 1988. Maximum likelihood estimation of growth and growth variability from tagging data. New Zealand Journal of Marine and Freshwater Research 22(1):43 51. Fremling, C. R. 1978. Biology and functional anatomy of the freshwater drum Aplodinotus Grunniens Rafinesque: Laboratory Manual. Nasco, Fort Atkinson, Wisconsin.

111

Fremling, C. R. 1980. Aplodinotus grunniens (Rafinesque), Freshwater Drum. Pages 756 in D. S. Lee, C. Gilbert, C. Hocutt, R. Jenkins, and McAllister, editors. Atlas of North American Freshwater Fishes. North Carolina State Museum of Natural History, Raliegh, NC. French III, J., and M. T. Bur. 1996. The effect of zebra mussel consumption on growth of freshwater drum in Lake Erie. Journal of Freshwater Ecology 11(3):283289. French III, J. R. P. 1997. Pharyngeal teeth of the freshwater drum ( Aplodinotus grunniens ) a predator of the zebra mussel ( Dreissena polymorpha ). Journal of Freshwater Ecology 12(3):495498. Fretwell, S., and H. Lucas. 1970. On territorial behavior and other factors influencing habitat distribution in birds. Acta Biotheoretica 19(1):1636. Friends of the Fox. 2011. FoxWisconsin Heritage Area Feasibility Study. http://heritageparkway.org/planning/feasibilitystudy/ . Last accessed 6/13/2012. Fuller, S. 1974. Clams and mussels (Mollusca: Bivalvia). C. J. Hart, and S. Fuller, editors. Pollution Ecology of Freshwater Invertebrates. Academic Press, Inc., New York. Gabriel, A. O., and L. R. Bodensteiner. 2012. Impacts of riprap on wetland shorelines, Upper Winnebago Pool Lakes, Wisconsin. Wetlands :113. Gauldie, R. 1991. The morphology and periodic structures of the otolith of the chinook salmon ( Oncorhynchus tshawytscha ), and temperaturedependent variation in otolith microscopic growth increment width. Acta Zoologica 72(3):159179. Gause, G. 1934. The Struggle for Existence. Williams and Wilkins, Baltimore, MD. Giuliano, A. 2011. Effects of Zebra Mussel Invasion on Juvenile Steelhead Distribution, Diet, Growth, and Condition in the Muskegon River, Michigan. Master of Science Thesis. University of Michigan. Goeman, T. J., D. R. Helms, and R. C. Heidinger. 1984. Comparison of otolith and scale age determinations for freshwater drum from the Mississippi River. Proceedings of the Iowa Academy of Science. 91(2):4951. Greens, C. W. 1935. Distribution of Wisconsin Fishes. Wisconsin Conservation Commission, editor, Madison, WI. Griffiths, R. W. 1993. Effects of zebra mussels ( Dreissena polymorpha ) on the benthic fauna of Lake St. Clair. Pages 415437 in T. F. Nalepa, and D. W. Schloesser, editors. Zebra Mussels: Biology, Impacts, and Control. Lewis Publishers, Boca Raton, FL. Griswold, B. L., and R. A. Tubb. 1977. Food of yellow perch, white bass, freshwater drum, and channel catfish in Sandusky Bay, Lake Erie. Ohio Journal of Science (Ohio Academy of Science) 77(1). Harris, J. L. 2006. Impacts of the Invasive White Perch of Kerr Reservoir, Virginia. Virginia Polytechnic Institute and State University. Harris, R. F. 1964. Lake Winnebago and Its Management. Wisconsin Conservation Department. Hecky, R., R. Smith, D. Barton, S. Guildford, W. Taylor, M. Charlton, and T. Howell. 2004. The nearshore phosphorus shunt: a consequence of ecosystem engineering by dreissenids in the Laurentian Great Lakes. Canadian Journal of Fisheries and Aquatic Sciences 61(7):12851293. Helama, S., B. R. Schoene, B. A. Black, and E. Dunca. 2006. Constructing longterm

112

proxy series for aquatic environments with absolute dating control using a sclerochronological approach: introduction and advanced applications. Marine and Freshwater Research 57(6):591599. Henderson, B. 1986. Effect of sea lamprey ( Petromyzon marinus ) parasitism on the abundance of white suckers ( Catostomus commersoni ) in South Bay, Lake Huron. Journal of Applied Ecology 23:381389. Hilsenhoff, W. L. 1967. Ecology and population dynamics of Chironomus plumosus (Diptera: Chironomidae) in Lake Winnebago, Wisconsin. Annals of the Entomological Society of America 60:11831194. Hobbs III, H., and J. P. Jass. 1988. The crayfishes and shrimp of Wisconsin (Cambaridae, Palaemonidae). Milwaukee, WI (USA), Milwaukee Public Museum. Hoff, G., and L. Fuiman. 1993. Morphometry and composition of red drum otoliths: changes associated with temperature, somatic growth rate, and age. Comparative Biochemistry and Physiology 106(2):209219. Holland, R. 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(3):617624. Holm, S. 1979. A simple sequentially rejective multiple test procedure. Scandinavian Journal of Statistics :6570. Hondorp, D. W., S. A. Pothoven, and S. B. Brandt. 2005. Influence of Diporeia density on diet composition, relative abundance, and energy density of planktivorous fishes in southeast Lake Michigan. Transactions of the American Fisheries Society 134(3):588601. Hothorn, T., F. Bretz, and P. Westfall. 2008. Simultaneous inference in general parametric models. Biometrical Journal 50(3):346363. Houser, A. 1960. Growth of freshwater drum in Oklahoma. Oklahoma Fishery Research Laboratory. Hoxmeier, R. J. H., D. D. Aday, and D. H. Wahl. 2001. Factors influencing precision of age estimation from scales and otoliths of bluegills in Illinois reservoirs. North American Journal of Fisheries Management 21(2):374380. Idrisi, N., E. Mills, L. Rudstam, and D. Stewart. 2001. Impact of zebra mussels (Dreissena polymorpha ) on the pelagic lower trophic levels of Oneida Lake, New York. Canadian Journal of Fisheries and Aquatic Sciences 58(7):14301441. Johannsson, O. E., K. L. Bowen, K. T. Holeck, and M. G. Walsh. 2011. Mysis diluviana population and cohort dynamics in Lake Ontario before and after the establishment of Dreissena spp., Cercopagis pengoi , and Bythotrephes longimanus . Canadian Journal of Fisheries and Aquatic Sciences 68(5):795811. Jones, D. S., M. Arthur, and D. Allard. 1989. Sclerochronological records of temperature and growth from shells of Mercenaria mercenaria from Narragansett Bay, Rhode Island. Marine Biology 102(2):225234. Kahl, R. 1993. Aquatic Macrophyte Ecology in the Upper Winnebago Pool Lakes, Wisconsin Technical Bulletin. W. D. o. N. Resources, editor, Madison, WI, . Kamke, K. K., and R. M. Bruch. 1991. Assessment of the effects of freshwater drum removal on the fish community dynamics of Lake Winnebago. Wisconsin Department of Natural Resources, Internal Fish Management Report, Oshkosh. Karataev, A. Y., and L. Burlakova. 1995. The role of Dreissena polymorpha in lake

113

ecosystems Rol'drejsseny v ozernykh ehkosistemakh. Ehkologiya Moscow (3):232236. Karatayev, A. Y., L. E. Burlakova, and D. K. Padilla. 1997. The effects of Dreissena polymorpha (Pallas) invasion on aquatic communities in eastern Europe. Journal of Shellfish Research 16(I):187203. Kerr, L. A., A. H. Andrews, B. R. Frantz, K. H. Coale, T. A. Brown, and G. M. Cailliet. 2004. Radiocarbon in otoliths of yelloweye rockfish ( Sebastes ruberrimus ): a reference time series for the coastal waters of southeast Alaska. Canadian Journal of Fisheries and Aquatic Sciences 61(3):443451. King, M. 1995. Fisheries Biology, Assessment and Management. Fishing News Books Victoria, Australia. Klaassen, H. E., and F. W. Cook. 1974. Age and growth of the freshwater drum in Tuttle Creek Reservoir, Kansas. Transactions of the Kansas Academy of Science (1903) 76:244247. Lankford Jr, T. E., and T. E. Targett. 2001. Lowtemperature tolerance of age0 Atlantic croakers: recruitment implications for US midAtlantic estuaries. Transactions of the American Fisheries Society 130(2):236249. LeBreton, G., and F. Beamish. 2000. Interannual growth variation in fish and tree rings. Canadian Journal of Fisheries and Aquatic Sciences 57(11):23452356. Lee, D., G. Burgess, and S. Platania. 1991. Atlas of North American Freshwater Fishes. North Carolina Biological Survey, Raleigh, N.C. Lillie, R., and J. Mason. 1983. Limnological characteristics of Wisconsin Lakes, Technical Bulletin No. 138. Department of Natural Resources, Madison, WI. Lyons, J., P. Cochran, and D. Fago. 2000. Wisconsin Fishes 2000: Status and Distribution. U. S. G. Institute, editor, Madison, WI. Malkin, S., S. Guildford, and R. Hecky. 2008. Modeling the growth response of Cladophora in a Laurentian Great Lake to the exotic invader Dreissena and to lake warming. Limnology and Oceanography 53(3):11111124. Mann, M. E., Z. Zhang, S. Rutherford, R. S. Bradley, M. K. Hughes, D. Shindell, C. Ammann, G. Faluvegi, and F. Ni. 2009. Global signatures and dynamical origins of the Little Ice Age and Medieval Climate Anomaly. Science 326(5957):1256 1260. Marchitto, T. M., Jr., G. A. Jones, G. A. Goodfriend, and C. R. Weidman. 2000. Precise temporal correlation of holocene mollusk shells using sclerochronology. Quaternary Research (Orlando) 53(2):236246. Marsden, J. E., B. D. Chipman, L. J. Nashett, J. K. Anderson, W. Bouffard, L. Durfey, J. E. Gersmehl, W. F. Schoch, N. R. Staats, and A. Zerrenner. 2003. Sea lamprey control in Lake Champlain. Journal of Great Lakes Research 29:655676. Mason, R. P. 1993. Archaeology, cavalry, and lakeflies. Fox Valley Archeology 21:14. Mason, R. P. 1996. A study of freshwater drum otoliths from the Doty Island village site in Winnebago County, Wisconsin. Fox Valley Archeology 25:90110. Matta, M. E., B. A. Black, and T. K. Wilderbuer. 2010. Climatedriven synchrony in otolith growthincrement chronologies for three Bering Sea flatfish species. Marine EcologyProgress Series 413:137145. Matthews, W. J. 1984. Influence of turbid inflows on vertical distribution of larval shad and freshwater drum. Enriched title Influence of turbid inflows on vertical

114

distribution of larval shad and freshwater drum [Dorosoma spp., Aplodinotus grunniens] 113(2):192198. Mayer, C., A. VanDeValk, J. Forney, L. Rudstam, and E. Mills. 2000. Response of yellow perch ( Perca flavescens ) in Oneida Lake, New York, to the establishment of zebra mussels ( Dreissena polymorpha ). Canadian Journal of Fisheries and Aquatic Sciences 57(4):742754. Meekan, M., J. Carleton, A. McKinnon, K. Flynn, and M. Furnas. 2003. What determines the growth of tropical reef fish larvae in the plankton: food or temperature? Marine Ecology Progress Series 256:193204. Miller, S., and T. Crowl. 2006. Effects of common carp ( Cyprinus carpio ) on macrophytes and invertebrate communities in a shallow lake. Freshwater Biology 51(1):8594. Moen, T. 1955. Food of the Freshwater Drum, Aplodinotus grunniens Rafinesque, in Four Dickinson County, Iowa, Lakes. The Proceedings of the Iowa Academy of Science Food of the Freshwater Drum, Aplodinotus grunniens Rafinesque, in Four Dickinson County, Iowa, Lakes:584598. Montgomery, J., S. Coombs, and M. Halstead. 1995. Biology of the mechanosensory lateral line in fishes. Reviews in Fish Biology and Fisheries 5(4):399416. Morrongiello, J. R., D. A. Crook, A. J. King, D. S. L. Ramsey, and P. Brown. 2011. Impacts of drought and predicted effects of climate change on fish growth in temperate Australian lakes. Global Change Biology 17(2):745755. Moser, M. L., J. M. Butzerin, and D. B. Dey. 2007. Capture and collection of lampreys: the state of the science. Reviews in Fish Biology and Fisheries 17(1):4556. Musick, J. A. 1999. Ecology and Conservation of LongLived Marine Animals. Pages 1 7 in J. A. Musick, editor. Life in the Slow Lane: Ecology and Conservation of LongLived Marine Animals. American Fisheries Society Symposium 23, Bethesda, Maryland, USA. Nash, M., and E. Irwin. 2000. Use of otoliths versus pectoral spines for aging adult flathead catfish. Pages 309–316 in E. R. Irwin, W. A. Hubert, C. F. Rabeni, H. L. J. Schramm, and T. Coon, editors. Catfish 2000: Proceedings of the International Ictalurid Symposium. American Fisheries Society Symposium 24, Bethesda, Maryland. NatureServe Explorer. 2010. NatureServe Explorer: An Online Encyclopedia of Life. http://www.natureserve.org/explorer/ . Last accessed 6/13/2012. Neilson, J. D., and G. H. Geen. 1982. Otoliths of chinook salmon ( Oncorhynchus tshawytscha ): daily growth increments and factors influencing their production. Canadian Journal of Fisheries and Aquatic Sciences 39(10):13401347. Neuheimer, A. B., and C. T. Taggart. 2007. The growing degreeday and fish sizeatage: the overlooked metric. Canadian Journal of Fisheries and Aquatic Sciences 64(2):375385. Nydal, R. 1993. Application of bomb 14 C as a tracer in the global carbon cycle. Trends in Geophysical Research 2:355364. Ostazeski, J. J., and G. R. Spangler. 2001. Use of biochronology to examine interactions of freshwater drum, walleye and yellow perch in the Red Lakes of Minnesota. Environmental Biology of Fishes 61(4):381393. Otis, K. 1988. Effects of Freshwater Drum Removal in Lake Winnebago, 19671981,

115

Research Report 145. Wisconsin Department of Natural Resources, Madison, WI. Otis, K., and M. Staggs. 1988. Analysis of Fish Trawling Data from Lake Winnebago, 19621981, Research Report 146. Department of Natural Resources, Madison, WI. Overstreet, D. F. 1997. Oneota prehistory and history. The Wisconsin Archeologist 78(1 2):250296. Ozersky, T., S. Malkin, D. Barton, and R. Hecky. 2009. Dreissenid phosphorus excretion can sustain C. glomerata growth along a portion of Lake Ontario shoreline. Journal of Great Lakes Research 35(3):321328. Page, L., and B. Burr. 1991. Peterson Field Guides: Freshwater Fishes. Houghton Mifflin Company, New York. Palmer, E. E., P. W. Sorensen, and I. R. Adelman. 1995. A histological study of seasonal ovarian development in freshwater drum in the Red Lakes, Minnesota. Journal of Fish Biology 47(2):199210. Patrick, H. K., T. M. Sutton, and W. D. Swink. 2009. Lethality of sea lamprey parasitism on lake sturgeon. Transactions of the American Fisheries Society 138(5):1065 1075. Pepin, P. 1991. Effect of temperature and size on development, mortality, and survival rates of the pelagic early life history stages of marine fish. Canadian Journal of Fisheries and Aquatic Sciences 48(3):503518. Pereira, D. L., C. Bingham, G. R. Spangler, Y. Cohen, D. J. Conner, and P. K. Cunningham. 1995a. Growth and recruitment of freshwater drum ( Aplodinotus grunniens ) as related to longterm temperature patterns. Canadian Special Publication of Fisheries and Aquatic Sciences 121:617629. Pereira, D. L., C. Bingham, G. R. Spangler, D. J. Conner, and P. K. Cunningham. 1995b. Construction of a 110year biochronology from sagittae of freshwater drum (Aplodinotus grunniens ). Pages 177196 in D. H. D. Secor, J. M. Campana, S. E., editor. Recent Developments in Fish Otolith Research, volume 19. Belle W. Baruch Library of Marine Science. Phelps, A., C. B. Renaud, and F. Chapleau. 2000. First record of a Freshwater Drum, Aplodinotus grunniens , in the Rideau River, Ottawa, Ontario. Canadian Field Naturalist 114(1):121125. Piner, K. R., O. S. Hamel, J. L. Menkel, J. R. Wallace, and C. E. Hutchinson. 2005. Age validation of canary rockfish (Sebastes pinniger ) from off the Oregon coast (USA) using the bomb radiocarbon method. Canadian Journal of Fisheries and Aquatic Sciences 62(5):10601066. Pörtner, H. O., B. Berdal, R. Blust, O. Brix, A. Colosimo, B. De Wachter, A. Giuliani, T. Johansen, T. Fischer, and R. Knust. 2001. Climate induced temperature effects on growth performance, fecundity and recruitment in marine fish: developing a hypothesis for cause and effect relationships in Atlantic cod ( Gadus morhua ) and common eelpout ( Zoarces viviparus ). Continental Shelf Research 21(18 19):19751997. Pothoven, S., and C. Madenjian. 2008. Changes in consumption by alewives and lake whitefish after dreissenid mussel invasions in Lakes Michigan and Huron. North American Journal of Fisheries Management 28(1):308320. Pothoven, S. A., and T. F. Nalepa. 2006. Feeding ecology of lake whitefish in Lake

116

Huron. Journal of Great Lakes Research 32(3):489501. Priegel, G. R. 1963a. Food of walleye and sauger in Lake Winnebago, Wisconsin. Transactions of the American Fisheries Society 92(3):312313. Priegel, G. R. 1963b. Use of otoliths to determine length and weight of ancient freshwater drum in the Lake Winnebago area. Transactions of the Wisconsin Academy of Sciences, Arts and Letters 52:2735. Priegel, G. R. 1965. Drum out the drum. Wisconsin Conservation Bulletin 30(2):1011. Priegel, G. R. 1967a. Food of the freshwater drum, Aplodinotus grunniens, Lake Winnebago, Wisconsin. Transactions of the American Fisheries Society 96(2):218220. Priegel, G. R. 1967b. The freshwater drum: its life history, ecology and management. Department of Natural Resources, Madison, WI. Priegel, G. R. 1969. Age and rate of growth of the freshwater drum in Lake Winnebago, Wisconsin. Transactions of the American Fisheries Society 98(1):116118. Priegel, G. R. 1971. Evaluation of intensive freshwater drum removal in Lake Winnebago, Wisconsin, 19551966. Wisconsin Department of Natural Resources, Madison, WI. R Development Core Team. 2012. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. Radtke, R., and D. Fey. 1996. Environmental effects on primary increment formation in the otoliths of newlyhatched Arctic charr. Journal of Fish Biology 48(6):1238 1255. Ramcharitar, J., D. Gannon, and A. Popper. 2006. Bioacoustics of fishes of the family Sciaenidae (croakers and drums). Transactions of the American Fisheries Society 135:14091431. Rhoads, D. C., and R. A. Lutz. 1980. Skeletal Growth of Aquatic Organisms: Biological Records of Environmental Change, volume 1. Plenum Press, New York. Rice, J. 1987. Intensive removal of freshwater drum from Lake Winnebago: a preliminary study of the feasibility of computer modeling and simulation. University of Wisconsin Oshkosh, unpublished report. Richards, F. 1959. A flexible growth function for empirical use. Journal of Experimental Botany 10(2):290301. Richardson, W., and L. Bartsch. 1997. Effects of zebra mussels on food webs: interactions with juvenile bluegill and water residence time. Hydrobiologia 354(1):141150. Ricker, W. 1975. Computation and interpretation of biological statistics of fish populations. Fisheries Research Board of Canada Bulletin 191:1382. Ritz, C., and J. C. Streibig. 2008. Nonlinear regression with R. Springer Verlag. Rodríguez, M. A., and W. M. Lewis Jr. 1997. Structure of fish assemblages along environmental gradients in floodplain lakes of the Orinoco River. Ecological Monographs 67(1):109128. Rounsefell, G. A., and H. W. Everhart. 1953. Fishery Science: Its Methods and Applications. Wiley, New York. Rutherford, D. A., W. E. Kelso, C. F. Bryan, and G. C. Constant. 1995. Influence of physicochemical characteristics on annual growth increments of four fishes from the lower Mississippi River. Transactions of the American Fisheries Society

117

124(5):687697. Rypel, A. 2009. Climate–growth relationships for largemouth bass (Mi cropterus salmoides) across three southeastern US states. Ecology of Freshwater Fish 18:620628. Rypel, A., W. Haag, and R. Findlay. 2009. Pervasive hydrologic effects on freshwater mussels and riparian trees in southeastern floodplain ecosystems. Wetlands 29(2):497504. Rypel, A. L. 2007. Sexual dimorphism in growth of freshwater drum. Southeastern Naturalist 6(2):333342. Rypel, A. L., D. R. Bayne, and J. B. Mitchell. 2006. Growth of freshwater drum from lotic and lentic habitats in Alabama. Transactions of the American Fisheries Society 135(4):987997. Schoffman, R. J. 1940. Age and growth of the drum in Reelfoot Lake. Journal of the Tennessee Academy of Science 16:100110. Schöne, B., E. Dunca, H. Mutvei, and U. Norlund. 2004. A 217year record of summer air temperature reconstructed from freshwater pearl mussels ( M. margarifitera , Sweden). Quaternary Science Reviews 23(1617):18031816. Schram, S. T., and M. C. Fabrizio. 1998. Longevity of Lake Superior lake trout. North American Journal of Fisheries Management 18(3):700703. Schultz, K. 2004. Ken Schultz's field guide to freshwater fish. J. Wiley & Sons, Hoboken, N.J. Scott, W., and E. Crossman. 1973. Freshwater Fishes of Canada. Fisheries Research Board of Canada, Bulletin 184, Ottawa. Seurer, D. M. 1978. Excavations at the Sauer Resort site, a Lake Winnebago phase site in Winnebago County. Fox Valley Chapter Wisconsin Archeological Society 15(Special Paper Number 1):6567. Sloey, W., F. Spangler, and C. Fetter. 1978. Management of freshwater wetlands for nutrient assimilation. Pages 321340 in R. E. Good, D. F. Whigham, and R. L. Simpson, editors. Freshwater Wetlands: Ecological Processes and Management Potential. Academic Press, New York. Sloey, W. E. 1970. The limnology of hypereutrophic Lake Butte des Morts, Wisconsin. Proceedings of the 13th Conference Great Lakes Research :951968. Sloey, W. E., and F. L. Spangler. 1977. Trophic Status of the Winnebago Pool Lakes: Report to the Fox Valley Water Quality Planning Agency. University of Wisconsin Oshkosh. Smith, N., C. Krueger, and J. Casselman. 2008. Growth chronologies of white sucker, Catostomus commerson i, and lake trout, Salvelinus namaycush : a comparison among lakes and between trophic levels. Environmental Biology of Fishes 81(4):375386. Smith, S. 1968. Species succession and fishery exploitation in the Great Lakes. Journal of the Fisheries Research Board of Canada 25(4):667693. Sonderegger, D. 2011. SiZer: Significant Zero Crossings. Staggs, M. D., and K. J. Otis. 1996. Factors affecting firstyear growth of fishes in Lake Winnebago, Wisconsin. North American Journal of Fisheries Management 16(3):608618. Stanley, J. G., and D. S. Danie. 1983. Species profiles: life histories and environmental

118

requirements of coastal fishes and invertebrates (North Atlantic) white perch. Maine Cooperative Fishery Research Unit, U.S. Fish and Wildlife Service. Stapanian, M. A., and C. P. Madenjian. 2007. Evidence that lake trout served as a buffer against sea lamprey predation on burbot in Lake Erie. North American Journal of Fisheries Management 27(1):238245. Stelzer, R. S., H. G. Drecktrah, M. P. Shupryt, and R. M. Bruch. 2008. Carbon sources for lake sturgeon in Lake Winnebago, Wisconsin. Transactions of the American Fisheries Society 137(4):10181028. Stevenson, K. P. 1985. Oneota subsistencerelated behavior in the Driftless Area: a study of the Valley View site near La Crosse, Wisconsin. University of Wisconsin Madison. Stewart, K. W., and D. A. Watkinson. 2004. The freshwater fishes of Manitoba. University of Manitoba Press. Stone, M. 1995. Fish stocking programs in Wyoming: a balanced perspective. Uses and Effects of Cultured Fishes in Aquatic Ecosystems 15:4751. Strayer, D. L., K. A. Hattala, and A. W. Kahnle. 2004. Effects of an invasive bivalve (Dreissena polymorpha ) on fish in the Hudson River estuary. Canadian Journal of Fisheries and Aquatic Sciences 61(6):924941. Strayer, D. L., and H. M. Malcom. 2006. Longterm demography of a zebra mussel (Dreissena polymorpha ) population. Freshwater Biology 51(1):117130. Strom, A., R. Francis, N. Mantua, E. Miles, and D. Peterson. 2004. North Pacific climate recorded in growth rings of geoduck clams: A new tool for paleoenvironmental reconstruction. Geophysical Research Letters 31(6):L06206. Stuiver, M., and H. A. Polach. 1977. Discussion: reporting of 14 C data. Radiocarbon 19(3):355363. Surber, T. 1912. Identification of the glochidia of freshwater mussels. U. B. o. Fisheries, editor. Swedberg, D. V. 1965. Age and rate of growth of freshwater drum, Lewis and Clark Lake, Missouri River. Proceedings of the South Dakota Academy of Science 44:160168. Swink, W. D. 2003. Host selection and lethality of attacks by sea lampreys ( Petromyzon marinus ) in laboratory studies. Journal of Great Lakes Research 29:307319. Thompson, D. Q. 1959. Biological Investigation of the Upper Fox River. Wisconsin Conservation Department. Toms, J. D., and M. L. Lesperance. 2003. Piecewise regression: a tool for identifying ecological thresholds. Ecology 84(8):20342041. Tonkin, Z., A. King, and D. Ramsey. 2008a. Otolith increment width responses of juvenile Australian smelt Retropinna semoni to sudden changes in food levels: the importance of feeding history. Journal of Fish Biology 73(4):853860. Tonkin, Z., A. King, and A. Robertson. 2008b. Validation of daily increment formation and the effects of different temperatures and feeding regimes on shortterm otolith growth in Australian smelt Retropinna semoni . Ecology of Freshwater Fish 17(2):312317. Trautman, M. B. 1982. The Fishes of Ohio: with Illustrated Keys. Ohio State University Press Columbus, OH, USA. United States Army Corps of Engineers, D. D. 1997. Fox River Project, Wisconsin

119

Navigation System: Draft Final Disposition Report, Draft Environmental Impact Statement. US Army Corps of Engineers, Detroit District. University of Wisconsin Extension. 2007. The Economic Impact of Angling on the Lake Winnebago System. http://winnebago.uwex.edu/files/2010/09/Summary_4_page_version5.pdf . Last accessed 6/13/2012. Van Oosten, J. 1938. The age and growth of the Lake Erie sheepshead, Aplodinotus grunniens Rafinesque. Michigan Acadamey of Science, Arts, and Letters 23(1937):651668. Vanderploeg, H., T. Nalepa, D. Jude, E. Mills, K. Holeck, J. Liebig, I. Grigorovich, and H. Ojaveer. 2002. Dispersal and emerging ecological impacts of PontoCaspian species in the Laurentian Great Lakes. Canadian Journal of Fisheries and Aquatic Sciences 59(7):12091228. Waller, D. L., T. D. Bills, M. A. Boogaard, D. A. Johnson, and T. C. J. Doolittle. 2003. Effects of lampricide exposure on the survival, growth, and behavior of the unionid mussels Elliptio complanata and Pyganadon cataracta . Journal of Great Lakes Research 29:542551. Watzin, M. C., K. JoppeMercure, J. Rowder, B. Lancaster, and L. Bronson. 2008. Significant fish predation on zebra mussels Dreissena polymorpha in Lake Champlain, U.S.A. Journal of Fish Biology 73(7):15851599. Weatherley, A. H., H. Gill, and J. M. Casselman. 1987. The Biology of Fish Growth. Academic Press, London. Weisberg, S. 1993. Using hardpart increment data to estimate age and environmental effects. Canadian Journal of Fisheries and Aquatic Sciences 50(6):12291237. Weisberg, S., G. Spangler, and L. Richmond. 2010. Mixed effects models for fish growth. Canadian Journal of Fisheries and Aquatic Sciences 67(2):269277. Werner, E. E., and J. F. Gilliam. 1984. The ontogenetic niche and species interactions in sizestructured populations. Annual Review of Ecology and Systematics 15:393 425. Werner, R. G. 2004. Freshwater Fishes of the Northeastern United States : A Field Guide, 1st edition. Syracuse University Press, Syracuse, N.Y. Wilson, K., E. Howell, and D. Jackson. 2006. Replacement of zebra mussels by quagga mussels in the Canadian nearshore of Lake Ontario: the importance of substrate, round goby abundance, and upwelling frequency. Journal of Great Lakes Research 32(1):1128. Wilson, K. A. 2002. Impacts of the invasive rusty crayfish ( Orconectes rusticus ) in northern Wisconsin lakes. University of Wisconsin Madison. Wirth, T. 1958. Lake Winnebago freshwater drum. Wisconsin Conservation Bulletin 23(51):3032. Wirth, T. L. 1959. Winnebago: The big lake. Wisconsin Conservation Bulletin 24:1519. Wisconsin Department of Natural Resources. 1989. Winnebago Comprehensive Management Plan. Wisconsin Department of Natural Resources, Oshkosh, WI. Wisconsin Department of Natural Resources. 1993. Aquatic Macrophyte Ecology in the Upper Winnebago Pool Lakes, Wisconsin. Wisconsin Department of Natural Resources, Oshkosh, WI. Wisconsin Department of Natural Resources. 2004. Water quality in the Lake Winnebago

120

pool. Wisconsin Department of Natural Resources, Publication FH22904, Madison. Wisconsin Department of Natural Resources. 2006. Choose wisely: a health guide for eating fish in Wisconsin. http://dnr.wi.gov/fish/consumption/FishAdvweb09lo.pdf . Last accessed 6/13/2012. Wisconsin Department of Natural Resources. 2009. Winnebago Sturgeon Spearing Regulations and Information System. http://dnr.wi.gov/fish/sturgeon/lakewinnebago/2009WinnebagoSturgeonSpearing RegulationPamphlet.pdf . Last accessed 6/13/2012. Wisconsin Department of Natural Resources. 2010. Citizen Lake Monitoring Network. http://dnr.wi.gov/lakes/CLMN/ . Last accessed 06/13/2012. Witt Jr, A. 1960. Length and weight of ancient freshwater drum, Aplodinotus grunniens , calculated from otoliths found in indian middens. Copeia 3:181185. Wrenn, W. B. 1968. Life history aspects of smallmouth buffalo and freshwater drum in Wheeler Reservoir, Alabama. Proceedings of the Southeastern Associated Game and Fisheries Commission 22:479495. Yan, N., and T. Pawson. 1997. Changes in the crustacean zooplankton community of Harp Lake, Canada, following invasion by Bythotrephes cederstroemi . Freshwater Biology 37(2):409425. Yule, D. L., J. D. Stockwell, J. A. Black, K. I. Cullis, G. A. Cholwek, and J. T. Myers. 2008. How systematic age underestimation can impede understanding of fish population dynamics: lessons learned from a Lake Superior cisco stock. Transactions of the American Fisheries Society 137(2):481495. Zar, J. H. 1996. Biostatistical Analysis. 3rd edn Prentice Hall. Upper Saddle River, New Jersey .

121

CURRICULUM VITAE SHANNON L. DAVISFOUST 627 PINE ST • OMRO, WI 54963 Email: [email protected] EDUCATION University of Wisconsin Milwaukee Ph.D., Biological Sciences August 2012 “Longterm Population Statistics of Freshwater Drum ( Aplodinotus grunniens ) in Lake Winnebago, Wisconsin, using otolith growth chronologies and bomb radiocarbon age validation” Secondary area of concentration: Natural Resource Management University of Wisconsin Oshkosh M.S., ZoologyEcology December 2001 “Environmental Variation and the Demography of a Tropical Forest Rodent” University of Wisconsin Oshkosh B.A., BiologyZoology major, Spanish & Chemistry minors, magna cum laude December 1997 APPOINTMENTS Wisconsin Department of Natural Resources Fisheries Management: Fisheries Technician 05/2000 to present Remediation and Redevelopment: Sample Custodian 02/2002 to 06/2004 Water Management: Water Management Specialist 05/2001 to 08/2002 Septage and Wastewater: Program Assistant 06/2001 to 01/2002 University of Wisconsin Oshkosh Clean Boats Clean Waters Coordinator 09/2010 to present Reviewer for student faculty collaborative grant proposals 2008 to present Department of Biology and Microbiology, Lab Instructor 08/2006 to present BIO 104 “Ecosphere in Crisis” BIO 105 and Honors BIO 108 “Concepts of Biological Unity” BIO 106 “Concepts of Biological Diversity” Environmental Studies Program, Lecturer and Student Advisor 05/2010 to present ES 101 “Seminar in Environmental Issues” ES 396 “Field Studies in Environmental Sciences: Waterways of Wisconsin” University of Wisconsin Fond du Lac 01/2008 to 05/2008 Department of Biological Sciences, Lecturer BIO 103 “Human Environmental Biology”

122

SHANNON L. DAVISFOUST (PAGE 2 OF 5)

APPOINTMENTS (CONTINUED) Moraine Park Technical College, Fond du Lac, WI 08/2006 to 12/2006 Veterinary Technician Program, Lecturer 806105 (hybrid course) “Principles of Animal Biology” Great Lakes WATER Institute, Milwaukee, WI 05/2004 to 05/2006 Research Assistant and Field Technician SCUBA surveys (advanced PADI certified) University of Wisconsin Milwaukee 08/2004 to 05/2006 Department of Biological Sciences, Teaching Assistant BioSci 152 “Foundations of Biological Sciences II” Menominee Park Zoo, Oshkosh, WI 05/1998 to 05/2001 Seasonal Zookeeper

PUBLICATIONS Adler, G.H., A. Carvajal, S.L. DavisFoust, and S.W. Brewer. 2012. Habitat associations of opossums and rodents in a lowland forest in French Guiana. Mammalian Biology 2012:8489.

DavisFoust, S.L. 2011. Lab Manual for Introductory Level Environmental Course, Bio 104: Ecosphere in Crisis. University of Wisconsin Oshkosh .

DavisFoust, S.L., R.M. Bruch, S.E. Campana, R.P. Olynyk, and J. Janssen. 2009. Age validation of freshwater drum using bomb radiocarbon. Transactions of the American Fisheries Society 138:385–396.

Bruch, R.M., S.E., Campana, S. DavisFoust, M.J. Hansen, and J. Janssen. 2009. Lake sturgeon age validation using bomb radiocarbon and knownage fish. Transactions of the American Fisheries Society 138:361–372.

Adler, G.H., A. Carvajal, S.W. Brewer, and S.L. Davis. 2006. First record of Didelphis albiventris (Didelphimorphia: Didelphidae) from Paracou, French Guiana. Mammalia 2006:319320.

Adler, G.H., S.L. Davis, and A. Carvajal. 2003. Bots (Diptera: Oestridae) infesting a neotropical forest rodent, Proechimys semispinosus , (Rodentia: Echimyidae), in Panama. Journal of Parasitology. 89(4): 693697.

123

SHANNON L. DAVISFOUST (PAGE 3 OF 5)

MANUSCRIPTS IN PROGRESS DavisFoust, S.L., R.M. Bruch, and D.H. Ogle. n.d . Longterm changes in the demography of a freshwater fish corresponding to anthropogenic disturbances. To be submitted to The American Naturalist .

DavisFoust, S.L., R.M. Bruch, D.H. Ogle, and G.R. Spangler. n.d . Using an otolith growth chronology to examine the effects of changing resource abundance due to the establishment of an invasive species. To be submitted to Environmental Biology of Fishes .

PRESENTATIONS

Invasive Species Education Summit, Horicon , WI “UW Oshkosh Clean Boats Clean Waters Internship Program” 14 June 2012

57 th Annual Midwest Archaeological Conference, LaCrosse, WI “Using Sclerochronology to Date the Occupancy of an Archaeological Site” Coauthor: Richard Mason 14 Oct 2011

40 th Annual Meeting of the Wisconsin American Fisheries Society, Stevens Point, WI “Backcalculations vs. Biochronology: Is there a Better Method for Interpreting Fish Growth Responses?” 1 Feb 2011

39 th Annual Meeting WI Chapter American Fisheries Society, Green Bay, WI “Using a Biochronology to Detect the Effects of Zebra Mussel Establishment and Food Resource Abundance” Coauthors: Ronald Bruch, John Janssen, George Spangler, Derek Ogle, and Donald Pereira 2 Feb 2010

139 th Annual Meeting of the American Fisheries Society, Nashville, TN “Drumming out the Truth in Lake Winnebago Part II: Using Otolith Biochronologies to Assess Environmental Impacts to Freshwater Drum Populations” Coauthors: Ronald Bruch, John Janssen, George Spangler, Derek Ogle, and Donald Pereira 31 Aug 2009

4th Annual Otolith Symposium, Monterey, CA “Linking Otolith Biochronology to the Establishment of an Invasive Species and Other Environmental Perturbations” Coauthors: Ronald Bruch, John Janssen, George Spangler, Derek Ogle, and Donald Pereira 25 Aug 2009

124

SHANNON L. DAVISFOUST (PAGE 4 OF 5)

PRESENTATIONS (CONTINUED)

TriChapter American Fisheries Society Meeting, Duluth, MN “Drumming Out the Truth in Lake Winnebago [Part I]” Coauthors: Ronald Bruch, Steven Campana, John Janssen, Kendall Kamke, and Robert Olynyk 3 Feb 2009

69 th Annual Midwest Fisheries and Wildlife Conference, Columbus, OH “Age Validation of Freshwater Drum using Nuclear Fallout and Population Level Responses” Coauthors: Ronald Bruch, Steven Campana, John Janssen, Kendall Kamke, and Robert Olynyk 16 Dec 2008

2nd Annual Natural Shoreline Expo, Oshkosh, WI 31 May 2008 “Beliefs and Attitudes about the Waterfront: What is a Beautiful Shoreline?”

Great Lakes WATER Institute Anchorwatch Seminar, Milwaukee, WI 30 Mar 2007 "Drumming out the truth in Lake Winnebago: Are 'Lucky Stones' more than just Lucky?"

Wisconsin DNR Age Estimation Workshop, Oshkosh, WI, “Estimating Age of Freshwater Drum” Coauthor: Robert Olynyk 28 Jun 2006

Joint WIIA American Fisheries Society Meeting, Dubuque, IA, “Lake Sturgeon Reproductive Success” 18 Jan 2006

CURRENT MEMBERSHIPS Wild Ones Fox Valley Area Chapter Mar 2010 to present

American Institute of Fishery Research Biologists Sept 2009 to present

World Sturgeon Conservation Society (WSCS) c.V. Sept 2009 to present

American Fisheries Society, Wisconsin Chapter Aug 2004 to present

VOLUNTEER ACTIVITIES American Fisheries Society Wisconsin Chapter Jan 2012 Judging presentations

Wild Ones Fox Valley Area Chapter Board of Directors Web Chair Mar 2012 to present

125

SHANNON L. DAVISFOUST (PAGE 5 OF 5)

VOLUNTEER ACTIVITIES (CONTINUED)

Guest speaking at local schools “Cracking the Chronological Code: Interpreting Ecological Changes from Fish Ear Bones” Ripon College, Ripon, WI Oct 2009 “What Kills Fish?” St. Mary’s Springs High School, Fond du Lac, WI Nov 2009

Assistant for Ichthyology course at Pigeon Lake Field Station, WI June 2009 I taught a fish age and growth session and assisted with field activities for an ichthyology course, University of Wisconsin – Stout

Wisconsin DNR Hunter Safety Course 2001 to present Each fall and spring I teach portions of the course that are related to ecology, animal identification, natural resource conservation, archery, gun, and tree stand safety

Omro Rushford Volunteer Fire Department 2001 to present Emergency response, public outreach to promote fire safety, various fundraising events.

The Great Lakes Marsh Monitoring Program 2001 to 2005 Duties included surveying amphibian populations by identifying their calls at eight stations and completing an annual wetland vegetation survey.

Winnebago County Master Gardeners 2003 to 2005 Certified Master Gardener. Volunteer activities included horticultural advising and landscaping.

AWARDS • Best Student Poster Award, Wisconsin Chapter American Fisheries Society, Feb 2012

• North American Dendroecological Fieldweek Graduate Research Scholarship, August 2011

• Best Student Paper Award, Wisconsin Chapter American Fisheries Society, Feb 2010

• University of WisconsinMilwaukee Graduate School Student Travel Award, 2009 th • Invited to compete in the Best Student Paper Symposia at the 139 Annual Meeting of the American Fisheries Society, Nashville, TN, 2009

• University of WisconsinMilwaukee Ruth Walker Student Travel Award, 2008

• University of WisconsinMilwaukee Graduate School Student Travel Award, 2008

• Ivy Balsam MilwaukeeAudubon Society Scholarship, 2007

• First recipient of the Statewide Wisconsin Chapter American Fisheries Society

126

Carroll Norden Memorial Scholarship, 2005

• University of WisconsinMilwaukee Graduate Student Chancellor’s Award 2005

Major Professor Dr. Rebecca Klaper Date