Characterizing Lentic Fish Assemblages and Community

Iowa State University Capstones, Theses and Graduate Theses and Dissertations Dissertations

2012 Characterizing lentic fish assemblages and community-environment relationships: An evaluation of natural lakes and reservoirs in Iowa, USA Jesse Robert Fischer Iowa State University

Follow this and additional works at: https://lib.dr.iastate.edu/etd Part of the Agriculture Commons, and the Aquaculture and Fisheries Commons

Recommended Citation Fischer, Jesse Robert, "Characterizing lentic fish assemblages and community-environment relationships: An evaluation of natural lakes and reservoirs in Iowa, USA" (2012). Graduate Theses and Dissertations. 12854. https://lib.dr.iastate.edu/etd/12854

This Dissertation is brought to you for free and open access by the Iowa State University Capstones, Theses and Dissertations at Iowa State University Digital Repository. It has been accepted for inclusion in Graduate Theses and Dissertations by an authorized administrator of Iowa State University Digital Repository. For more information, please contact [email protected].

Characterizing lentic fish assemblages and community-environment relationships: An evaluation of natural lakes and reservoirs in Iowa, USA

Jesse Robert Fischer

A dissertation submitted to the graduate faculty

in partial fulfillment of the requirements for the degree of

DOCTOR OF PHILOSOPHY

Major: Fisheries Biology

Program of Study Committee: Michael C. Quist, Major Professor John A. Downing Philip M. Dixon Clay L. Pierce Kevin J. Roe

Iowa State University

Ames, Iowa

2012

TABLE OF CONTENTS

LIST OF TABLES iv LIST OF FIGURES vi ACKNOWLEDGEMENTS viii

CHAPTER 1. GENERAL INTRODUCTION 1 Dissertation Organization 4 References 5

CHAPTER 2. CHARACTERIZING LENTIC FRESHWATER FISH ASSEMBLAGES USING MULTIPLE SAMPLING METHODS 10 Abstract 10 Introduction 11 Methods 15 Results 18 Discussion 22 Acknowledgments 26 References 26 Tables 33 Figures 36

CHAPTER 3. COMPARISON OF MULTIPLE SAMPLING METHODS FOR ASSESSING LENTIC FRESHWATER FISH POPULATIONS 39 Abstract 39 Introduction 40 Methods 43 Results 48 Discussion 51 Acknowledgements 57 References 57 Tables 66 Figures 70

CHAPTER 4. UNDERSTANDING FISH COMMUNITY STRUCTURE DYNAMICS IN LENTIC ECOSYSTEMS: A COMPARISON OF NATURAL LAKES AND RESERVOIRS IN IOWA, USA 76 Abstract 76 Introduction 77 Methods 80 Results 85 iii

Discussion 88 Acknowledgements 93 References 94 Tables 100 Figures 106

CHAPTER 5. GENERAL CONCLUSIONS 111

LIST OF TABLES

CHAPTER 2

TABLE 2.1.—Water body type, area (ha), mean depth (Z; m), sampling effort 33 (number of net nights for passive gears, number of seine hauls, number of 3-minute trawls, and number of 5-minute electrofishing runs conducted within a season), mean secchi depth (m), mean chlorophyll a (Chl-a; µg/L) concentration, mean total phosphorus (TP; µg/L) concentrations. Water body information obtained from the Iowa Department of Natural Resources Lakes Information Report.

TABLE 2.2.—Probability of detection and the number of water bodies present (N) 34 for seven sampling methods used in the spring (Sp), summer (Su), and fall (Fa) to characterize fish assemblages of six lakes and impoundments in Iowa, 2008.

CHAPTER 3

TABLE 3.1.—Water body, type, area, mean depth (Z), sampling effort (number of 66 net nights for passive gears, number of seine hauls, number of 3-minute trawls, and number of 5-minute electrofishing runs conducted within a season), mean Secchi depth, mean chlorophyll a (Chl-a) concentration, mean total phosphorus (TP) concentrations.

TABLE 3.2.—Total number of individuals sampled in spring (Sp), summer (Su), 67 and fall (Fa) for seven sampling methods from six lakes and impoundments in Iowa, 2008.

TABLE 3.3.—Proportion of lakes and impoundments in Iowa where 50 and 125 68 (parentheses) stock-length individuals were sampled with seine (S), trawl (T), day electrofishing (D), night electrofishing (N), fyke net (F), mini-fyke net (M), and gill net (G) in spring, summer, and fall 2008.

TABLE 3.4.—Mean proportional size distribution (PSD) and standard error 69 (parentheses) of 12 fish species sampled with seine (S), trawl (T), day electrofishing (D), night electrofishing (N), fyke net (F), mini-fyke net (M), and gill net (G) in spring, summer, and fall from six lake and impoundments in Iowa, 2008. Mean PSD was only calculated for seasons, sampling methods, and lakes with greater 50 stock-length individuals sampled.

CHAPTER 4

TABLE 4.1.—Physiochemical characteristics of 45 lentic waterbodies located in 100 Iowa, USA. v

TABLE 4.2.—Spearman rank correlation coefficient of surface area (ha), watershed- 101 to-lake-area ratio (W-L ratio), mean depth (m), maximum depth (m), Secchi depth (m), chlorophyll a (µg/L; Chl a), total phosphorus (µg/L, TP), total nitrogen (µg/L, TN), and total suspended solids (µg/L, TSS) of 45 lentic water bodies located in Iowa, USA. Correlation coefficients with P ≤ 0.05 indicated in bold.

TABLE 4.3.—Physicochemical correlations for nonmetric multidimensional scaling 102 ordinations of fish species and trophic composition for natural lakes, reservoirs, and combined fish assemblage data collected using summer benthic trawling, fall modified-fyke nets, and fall night electrofishing from 2008-2011 in Iowa, USA. Correlation coefficients with P ≤ 0.05 indicated in bold.

TABLE 4.4.—Spearman’s rank correlation coefficients (rs) between selected 103 physiochemical characteristics and species richness from natural lakes, reservoirs, and combined fish assemblage data collected using summer benthic trawling, fall modified-fyke nets, and fall night electrofishing from 2008-2011 in Iowa, USA. Correlation coefficients with P ≤ 0.05 indicated in bold.

TABLE 4.5.—Fishes sampled from 45 water bodies in Iowa, 2008-2011. Species 104 without abbreviated species codes occurred at fewer than 5% of the water bodies sampled and removed from taxonomic analyses. Trophic categories assigned to species were based on those proposed by Goldstein and Simon (1999) and combined to account for species that belonged to more than one trophic category during different life history stages following Miranda et al. (2008).

LIST OF FIGURES

CHAPTER 2

FIGURE 1.—Mean and maximum number of species and individuals sampled 36 seasonally with seven methods for six Iowa lakes and impoundments, 2008. Error bars represent one SE.

FIGURE 2.—Species accumulation curves for six lakes and impoundments in Iowa 37 sampled in 2008 with seven methods.

FIGURE 1.—Number of species and individuals sampled with combinations of 38 seven methods (e.g., one gear, two gears) for six Iowa lakes and impoundments, 2008.

CHAPTER 3

FIGURE 1. Mean catch-per-unit effort (CPUE) of substock-length individuals from 70 seasonal sampling with seven methods in six lakes and impoundments in Iowa, 2008. Values for CPUE were calculated for independently for seine (number per haul); trawl (number per 3-minute trawl); day and night electrofishing (EF; number per hour); and fyke net, mini-fyke net, and gill net (number per net night). Error bars represent one SE.

FIGURE 2.—Mean catch-per-unit effort (CPUE) of stock-length individuals from 72 seasonal sampling with seven methods in six lakes and impoundments in Iowa, 2008. Values for CPUE were calculated for independently for seine (number per haul); trawl (number per 3-minute trawl); day and night electrofishing (EF; number per hour); and fyke net, mini-fyke net, and gill net (number per net night). Error bars represent one SE.

FIGURE 3.—Mean minimum numbers of samples (i.e., five minute electrofishing 74 runs, deployments, sets) necessary to sample 125 stock-length fish from seasonal sampling with seven methods in six lakes and impoundments in Iowa, 2008. Error bars represent one SE.

CHAPTER 4

FIGURE 1.—Nonmetric multidimensional scaling (NMS) ordinations of species 106 composition (top) and trophic composition (bottom) for natural lakes (open circles) and reservoirs (filled circles) with 90% confidence ellipses for summer benthic trawling fish assemblage data collected from 45 water bodies in Iowa, USA from 2008 to 2011. Environmental vectors indicate the direction and strength of significant (P ≤ 0.05) correlations within the NMDS ordination for: L_Size = lake surface area (ha), TN = total nitrogen concentration (µg/L). vii

Species abbreviation codes are found in Table 5. FIGURE 2.—Nonmetric multidimensional scaling (NMS) ordinations of species 107 composition (top) and trophic composition (bottom) for natural lakes (open circles) and reservoirs (filled circles) with 90% confidence ellipses for fall modified-fyke netting fish assemblage data collected from 45 water bodies in Iowa, USA from 2008 to 2011. Environmental vectors indicate the direction and strength of significant (P ≤ 0.05) correlations within the NMDS ordination for: Chla = chlorophyll a concentration (µg/L), L_Size = lake surface area (ha), Mn_Depth = mean waterbody depth (m), Mx_Depth = maximum waterbody depth (m), TP = total phosphorus concentration (µg/L), Secchi = Secchi disc depth (m), TSS = total suspended solids concentration (mg/L), WS_Ratio = watershed-to-lake-area ratio, and TN = total nitrogen concentration (µg/L). Species abbreviation codes are found in Table 5.

FIGURE 3.—Nonmetric multidimensional scaling (NMS) ordinations of species 108 composition (top) and trophic composition (bottom) for natural lakes (open circles) and reservoirs (filled circles) with 90% confidence ellipses for fall night electrofishing fish assemblage data collected from 45 waterbodies in Iowa, USA from 2008 to 2011. Environmental vectors indicate the direction and strength of significant (P ≤ 0.05) correlations within the NMDS ordination for: L_Size = lake surface area (ha), Secchi = Secchi disc depth (m), and WS_Ratio = lake surface area to watershed area ratio. Species abbreviation codes are found in Table 5.

FIGURE 4.—Physical characteristics of natural lakes and reservoirs in relation to 109 species richness for 45 waterbodies in Iowa, USA from 2008 to 2011.

FIGURE 5.—Water quality characteristics of natural lakes and reservoirs in relation 110 to species richness for 45 waterbodies in Iowa, USA from 2008 to 2011.

viii

ACKNOWLEDGEMENTS

I am indebted to the guidance of my advisor Dr. Michael Quist and thank him for his support and direction throughout the development of this project. Mike’s commitment to his students is without equal and his mentorship has been fundamental to my professional development while at Iowa State University (ISU). I also deeply appreciate the assistance provided from my committee members Drs. John Downing, Philip Dixon, Clay Pierce, and

Kevin Roe. Their input and direction contributed greatly to the development of this project and my education.

For their numerous discussions with me and assistance with the completion of this project, I am very appreciative of fellow ISU graduate students: Mike Colvin, Jeff Koch, Travis

Neebling, Tony Sindt, Maria Dzul, Bryan Bakevich, Tim Parks, Jim Church, Adam Heathcote,

Paul Skrade, and Drake Larsen. I am also extremely grateful for all of the fantastic technicians that have supported me in the field and laboratory. The help of Matt Mork, Scott Bisping,

Andrew Schaefer, Wade Massure, Nick Johnson, Rebecca Krogman, Josh Bruegge, Colin Hinz,

Chris Smith, Jared Brashears, Brett Meyer, Alex Dierker, Tony Grimm, Casey Bergthold, Grant

Scholten, Michael Sundberg, and Cole Harty have made this project a much more meaningful experience than it would have without them.

All of this would not have been possible without financial and logistical support from the

Iowa Department of Natural Resource (IDNR). I am particularly grateful for the assistance, direction, and conversations with IDNR biologists, including George Antoniou, Mike McGee,

Joe Larscheid, Randy Schultz, George Scholten, Mike Hawkins, Ben Dodd, Bryan Hayes, Mark

Flammang, Chad Dolan, Ben Wallace, Don Herrig, and Andy Fowler. Support for this project ix was also provided by Department of Natural Resource Ecology and Management at ISU and the

Idaho Cooperative Fish and Wildlife Research Unit at the University of Idaho.

I dedicate this dissertation to my parents, Robert and Sandra, whose instilled work ethic and unwavering encouragement are responsible for who I am today. Finally, I thank my amazing wife, Heather. Her patience, support, and love have been instrumental to me and the completion of this project. 1

CHAPTER 1. GENERAL INTRODUCTION

Despite constituting less than 4% of the earth’s continental surface (Downing et al.

2006), lentic freshwater ecosystems provide a disproportionate array of economic and ecological services (e.g., municipal and agricultural water supply, flood control, biodiversity) whose values are dependent on the quality of water resources and ecological integrity. However, freshwater ecosystems and their biota are among the most endangered in the world as a result of habitat degradation and loss, introduction of invasive species, water pollution, and overexploitation

(Rahel 2000; Sala et al. 2000; Dudgeon et al. 2006). Lentic freshwater ecosystems in the United

States demonstrate this vulnerability as water quality continues to decline in response to anthropogenic alterations. For example, the proportion of lakes and reservoirs surveyed in the

United States considered impaired for one or more designated uses has steadily increased from

35% in 1992 to 58% in 2006 (U.S. EPA 2012). One of the largest factors contributing to impairment, non-point source pollution, is commonly associated with land-use modification

(e.g., agriculture, urbanization) and is a worldwide problem leading to the eutrophication of surface waters (Carpenter et al. 1998; Dodds et al. 2009).

Water quality degradation has been particularly drastic in the midwestern United States where the native terrestrial landscape has been extensively altered. Less than 0.1% of the land area support native terrestrial vegetation in the highly agricultural regions of Illinois, Indiana, and Iowa (Sampson and Knopf 1994). The replacement of perennial tallgrass prairies with row crops and pastures has decreased water quality through increases in sediment and nutrient inputs, and thereby increased eutrophication and decreased biological integrity in aquatic systems.

Although increased eutrophication can increase fish production (Hanson and Leggett 1982; 2

Downing et al. 1990), fish assemblage composition in lentic systems often shifts from sight- feeding predators to benthivorous species as water clarity decreases and nutrient concentrations increases (Persson et al. 1991; Jeppesen et al. 2000; Egertson and Downing 2004). Furthermore, the introduction of non-native species can severely alter lentic environments with similar impacts to those attributed to eutrohpication. Common carp Cyprinus carpio has long been associated with disrupted ecological processes in aquatic environments (e.g., Breukelaar et al. 1994;

Lougheed et al. 1998), particularly in agriculturally-influenced Midwestern lentic ecosystems

(Schrage and Downing 2004; Jackson et al. 2010). In Iowa, severe alterations to the ecological integrity of lentic ecosystems have resulted in over one third (i.e., 46 of 132) of the principal recreation lakes considered impaired. The majority of impaired waters (i.e., 30) are impaired due to anthropogenic eutrophication (Iowa Department of Natural Resources 2007). However, as our understanding of threats to aquatic ecosystem integrity increases, little is still known about factors that shape fish communities in complex lentic habitats.

Assessing the occurrence and abundance freshwater fish is particularly challenging due to numerous sampling bias (e.g., gear, season, location) that affect accurate characterization of populations and assemblages (Hayes et al. 1996; Huber et al. 1996; Pope and Willis 1996).

Unlike rivers and streams where relatively few sampling methods (e.g., electrofishing, seines) are commonly used (Guy et al. 2009; Rabeni et al. 2009), numerous methods (e.g., electrofishing, seines, fyke nets, gill nets, trawling) have been used to sample fish assemblages in lakes and impoundments (Jackson and Harvey 1997; Miranda and Boxrucker 2009; Pope et al.

2009). The need to use multiple gears is generally the result of habitat heterogeneity and associated species use in lentic ecosystems. Differences in habitat characteristics (e.g., depth, water clarity, vegetation) and selectivity for certain species and sizes of fish affect the 3 efficiencies of sampling methods in different habitats. For example, multiple gears are often necessary to sample both juvenile and adult fish of the same species due to shifts in habitat use and size biases associated with different gears (Boxrucker et al. 1995). However, few studies have evaluated multiple sampling gears simultaneously, and those that exist have focused on a limited number of species (e.g., gizzard shad Dorosoma cepedianum, black crappie Poxomis nigromaculatus) at small spatial scales (e.g, Boxrucker et al. 1995; Guy et al. 1996; Allen et. al

1999). Therefore, additional quantitative studies of lake fish assemblage assessment are needed to understand the influence of sampling methods on characterizing whole fish assemblages in lakes and impoundments.

Reservoirs are unique ecosystems that can exhibit characteristics intermediate to lentic and lotic habitats (Wetzel 2001). Reservoirs are often referred to as artificial lakes and the two ecosystems (i.e., natural lakes and reservoirs) are generally managed similarly in regions where they coexist (Irz et al. 2006). However, environmental relationships of fish assemblages in reservoirs has received considerably less attention than natural systems (Irz et al. 2002; Miranda et al. 2008), despite the potential to support similar management strategies, monitoring programs, and fish faunas (e.g., Guy and Willis et al. 1996; Launois et al. 2011, Menezes et al. in press).

Comparative ecological studies of lake and reservoir fish assemblage structure and environmental influences are even less common (e.g., Irz et al. 2006; Irz et al 2007). Therefore, understanding the dynamics of fish assemblage structure across ecosystems may be beneficial to natural resource agencies charged with simultaneously managing fisheries and conserving biodiversity. Furthermore, identifying environmental factors important for structuring assemblages in lentic systems will be critical for monitoring further anthropogenic changes or restoration efforts. 4

The goal of this project was to provide greater knowledge of factors that influence lentic

fish fauna characterization. Specifically, the influence of sampling methodologies and

assemblage-environmental relationships were evaluated. This dissertation is comprised of four

subsequent chapters. The next chapter (i.e., chapter 2) is a manuscript that will be submitted for

publication in Fisheries. The chapter describes the evaluation of multiple freshwater fish

sampling protocols used to characterize lentic fish assemblages in artificial and natural

ecosystems. The third chapter identifies temporal patterns in population characteristics of 12

ecologically-important species from several lentic fish sampling methods that will also be

submitted to Fisheries for publication. The fourth chapter contrasts fish assemblage structure from 45 natural lakes and reservoirs throughout Iowa, USA. This chapter will be submitted as a manuscript to Transactions of the American Fisheries Society. Chapters two, three, four, are

structured similarly and include an abstract, introduction, methods, results, discussion,

acknowledgements, references, tables, and figures. The final chapter provides a synthesis and

general conclusions from the three previous chapters.

REFERENCES

Allen, M. S., M. M. Hale, and W. E. Pine III. 1999. Comparison of trap nets and otter trawls for

sampling black crappie in two Florida lakes. North American Journal of Fisheries

Management 19:977–983.

Boxrucker, J., P. Michaletz, M. J. Van Den Avyle, and B. Vondracek. 1995. Overview of gear

evaluation study for sampling gizzard shad and threadfin shad populations in reservoirs.

North American Journal of Fisheries Management 15:885–890.

Breukelaar, A. W., E. Lammens, J. Breteler, and I. Tatrai. 1994. Effects of benthivorous bream

(Abramis brama) and carp (Cyprinus carpio) on sediment resuspension and

concentrations of nutrients and chlorophyll-a. Freshwater Biology 32:113-121.

Carpenter, S. R., N. F. Caraco, D. L. Correll, R. W. Howarth, A. N. Sharpley, and V. H. Smith.

1998. Non-point pollution of surface waters with phosphorus and nitrogen. Ecological

Applications 8:559–568.

Dodds, W. K., W. Bouska, J. L. Eitzmann, T. J. Pilger, K. L. Pitts, A. J. Riley, J. T. Schloesser,

and D. J. Thronbrugh. 2009. Eutrophication of U.S. freshwaters: analysis of potential

economic damages. Environmental Science and Technology 43:12–19.

Downing, J. A., C. Plante, and S. Lalonde. 1990. Fish production correlated with primary

productivity, not the morphoedaphic index. Canadian Journal of Fisheries and Aquatic

Sciences 47:1929–1936.

Downing, J.A., Y.T. Prairie, J.J. Cole, C.M. Duarte, L.J. Tranvik, R.G. Striegl, W.H. McDowell,

P. Kortelainen, N.F. Caraco, J.M. Melack, and J. Middelburg. 2006. The global

abundance and size distribution of lakes, ponds, and impoundments. Limnology and

Oceanography 51:2388–2397 6

Dudgeon, D., A. H. Arthington, M. O. Gessner, Z. Kawabata, D. J. Knowler, C. Lévêque, R. J.

Naiman, A. Prieur-Richard, D. Soto, M. L. J. Stiassny, and C. A. Sullivan. 2006.

Freshwater biodiversity: importance, threats, status and conservation challenges.

Biological Reviews 81:163–182.

Egertson, C. J., and J. A. Downing. 2004. Relationship of fish catch and composition to water

quality in a suite of agriculturally eutrophic lakes. Canadian Journal of Fisheries and

Aquatic Sciences 61:1784–1796.

Guy, C. S., and D. W. Willis. 1995. Population characteristics of black crappies in South

Dakota waters: a case for ecosystem-specific management. North American Journal of

Fisheries Management 15:754-765.

Guy, C. S., D. W. Willis, and R. D. Schultz. 1996. Comparison of catch per unit effort and size

structure of white crappies collected with trap nets and gill nets. North American Journal

of Fisheries Management 16:947–951.

Guy, C. S., P. J. Braaten, D. P. Herzog, J. Pitlo, and R. S. Rogers. 2009. Warmwater fish in

rivers. Pages 59–84 in S. Bonar, W. Hubert, and D. Willis, editors. Standard sampling

methods for North American freshwater fishes. American Fisheries Society, Bethesda,

Maryland.

Hanson, J. M., and W. C. Leggett. 1982. Empirical predictions of fish biomass and yield.

Canadian Journal of Fisheries and Aquatic Sciences 39:257–263.

Hubert, W. A. 1996. Passive capture techniques. Pages 157–192 in B. R. Murphy and D. W.

Willis, editors. Fisheries techniques, second edition. American Fisheries Society,

Bethesda, Maryland. 7

Iowa Department of Natural Resources. 2007. Water quality in Iowa 2004 through 2005:

assessment database for the 2006 Section 305(b)/303(d) assessment/listing cycle. Iowa

Department of Natural Resources, Des Moines, Iowa.

Irz, P., A. Laurent, S. Messad, O. Pronier, and C. Argillier. 2002. Influence of site

characteristics on fish community patterns in French reservoirs. Ecology of Freshwater

Fish 11:123–136.

Irz, P., M. Odion, C. Argillier, and D. Pont. 2006. Comparison between the fish communities

of lakes, reservoirs and rivers: can natural systems help define the ecological potential of

reservoirs? Aquatic Sciences 68:109–116.

Irz, P., M. François, T. Oberdorff, T. Whittier, N. Lamouroux, D. Mouillot, and C. Argillier.

2007. Fish community comparisons along environmental gradients in lakes of France

and north-east USA. Global Ecology and Biogeography 16:350–366.

Jackson, D. A., and H. H. Harvey. 1997. Qualitative and quantitative sampling of lake fish

communities. Canadian Journal of Fisheries and Aquatic Sciences 54:2807–2813.

Jackson, Z. J., M. C. Quist, J. A. Downing, and J. G. Larscheid. 2010. Common carp (Cyprinus

carpio), sport fishes, and water quality: ecological thresholds in agriculturally eutrophic

lakes. Lake and Reservoir Management 26:14–22.

Jeppesen, E., J. P. Jensen, M. Sondergaard, T. Lauridsen, and F. Landkildehus. 2000. Trophic

structure, species richness and biodiversity in Danish lakes: changes along a phosphorus

gradient. Freshwater Biology 45:201–218.

Launois, L., J. Veslot, P. Irz, and C. Argillier. 2011. Selecting fish-based metrics responding to

human pressures in French natural lakes and reservoirs: towards the development of a

fish-based index (FBI) for French lakes. Ecology of Freshwater Fish 20:120–132. 8

Lougheed, V. L., B. Crosbie, and P. Chow-Fraser. 1998. Predictions on the effect of common

carp (Cyprinus carpio) exclusion on water quality, zooplankton, and submergent

macrophytes in a Great Lakes wetland. Canadian Journal of Fisheries and Aquatic

Sciences 55:1189–1197.

Menezes, R. F., F. Borchsenius, J-C. Svenning, M. Søndergaard, T. L. Lauridsen, F.

Landkildehus, and E. Jeppesen. In Press. Variation in fish community structure,

richness, and diversity in 56 Danish lakes with contrasting depth, size, and trophic state:

does the method matter? Hydrobiologia.

Miranda, L. E., M. D. Habrat, and S. Miyazono. 2008. Longitudinal gradients along a reservoir

cascade. Transactions of the American Fisheries Society 137:1851–1865.

Miranda, L. E., and J. Boxrucker. 2009. Warmwater fish in large standing waters. Pages 29–42

in S. Bonar, W. Hubert, and D. Willis, editors. Standard sampling methods for North

American freshwater fishes. American Fisheries Society, Bethesda, Maryland.

Persson, L., S. Diehl, L. Johansson, and G. Andersson. 1991. Shifts in fish communities along

the productivity gradient of temperate lakes: patterns and the importance of size-

structured interactions. Journal of Fish Biology 38:281–294.

Pope, K. L., and D. W. Willis. 1996. Seasonal influences of freshwater fisheries sampling data.

Reviews in Fisheries Science 4:57–73.

Pope, K. L., R. M. Neumann, and S. D. Bryan. 2009. Warmwater fish in small standing waters.

Pages 13–28 in S. Bonar, W. Hubert, and D. Willis, editors. Standard sampling methods

for North American freshwater fishes. American Fisheries Society, Bethesda, Maryland.

Rabeni, C., J. Lyons, N. Mercado-Silva, and J. Peterson. 2009. Warmwater fish in wadeable

streams. Pages 43–58 in S. Bonar, W. Hubert, and D Willis, editors. Standard sampling 9

methods for North American freshwater fishes. American Fisheries Society, Bethesda,

Maryland.

Rahel, F. J. 2000. Homogenization of fish faunas across the United States. Science 288:854–

856.

Sala, O. E., F. S. Chapin, J. J. Armesto, R. Berlow, J. Bloomfield, R. Dirzo, E. Huber-Sanwald,

L. F. Huenneke, R. B. Jackson, A. Kinzig, R. Leemans, D. Lodge, H. A. Mooney, M.

Oesterheld, N. L. Poff, M. T. Sykes, B. H. Walker, M. Walker, and D. H. Wall. 2000.

Global biodiversity scenarios for the year 2100. Science 287:1770–1774.

Samson, F., and F. Knopf. 1994. Prairie conservation in North America. Bioscience 44:418–

421.

Schrage, L. J., and J. A. Downing. 2004. Pathways of increased water clarity after fish removal

from Ventura Marsh; a shallow, eutrophic wetland. Hydrobiologia 511:215–231.

U. S. Environmental Protection Agency. 2012. Watershed Assessment, Tracking, and

Environmental Results: National Summary of Impaired Waters and Total Maximum

Daily Loads Summary; http://ofmpub.epa.gov/waters10/attains_nation_cy.control.

Wetzel, R. G. 2001. Limnology: lake and river ecosystems, third edition. Academic Press, San

Diego, California.

CHAPTER 2. CHARACTERIZING LENTIC FRESHWATER FISH ASSEMBLAGES

USING MULTIPLE SAMPLING METHODS

A manuscript to be submitted for publication in North American Journal of Fisheries

Management

Jesse R. Fischer1 and Michael C. Quist2

Abstract

Characterizing fish assemblages in lentic ecosystems is difficult due to high habitat

heterogeneity and the diversity of methods available to sample fish. Therefore, multiple

sampling methods are almost always necessary for gaining reliable estimates of species richness

and for sampling different sizes of a particular species. However, most research on the use of

multiple sampling methods has targeted recreationally-important species. As such, little

information is available regarding the influence of multiple methods and timing (i.e., temporal

variation) on characterization of lentic fish assemblages. We sampled six lakes and

impoundments seasonally (i.e., spring, summer, fall) with beach seines, benthic trawls, boat

electrofishing (i.e., day and night), modified fyke nets, mini-fyke nets, and gill nets (fall only) to

evaluate the combined influence of sampling methods and timing on the number of species and

individuals sampled. Probabilities of detection for species indicated strong selectivities and seasonal trends. Additionally, patterns in seasonal probabilities of detection and species

1 Iowa State University, Department of Natural Resource Ecology and Management, 339 Science II, Ames, IA 50010, USA. 2 U.S. Geological Survey, Idaho Cooperative Fish and Wildlife Research Unit, Department of Fish and Wildlife Sciences, University of Idaho, Moscow, Idaho 83844, USA.

accumulation curves provided guidance on optimal seasons to use gears when targeting multiple

species. The evaluation of species richness and number of individuals sampled using multiple

gear combinations demonstrated that appreciable benefits over relatively few gears (e.g., three to

four) used in optimal seasons were not present. Our results indicated that the characterization of

lentic fish assemblages was highly influenced by the selection of sampling gears and seasons, but

did not appear to be influenced by waterbody type (i.e., natural lake, impoundment). The

standardization of data collected with multiple methods and seasons to account for bias will

allow fisheries scientists to have greater reliability in their interpretations and decisions made

using information on lentic fish assemblages.

INTRODUCTION

Knowledge of fish assemblage composition is critical for the effective management of

fisheries and conservation of biodiversity in aquatic ecosystems. The importance of

understanding fish assemblage structure is supported by the inclusion of biological components

in the 1972 U.S. Clean Water Act (CWA), which focuses on maintaining and restoring

ecological integrity of surface waters. A consequence of the CWA’s legal mandate to focus on

aquatic organisms has been extensive research on the ecology of fishes in streams and rivers.

Furthermore, the formative research on fish-based indices of biological integrity (e.g., Karr

1981) have greatly improved the methods and techniques used to monitor fish (e.g., Lyons 1992;

Angermeier and Smogor 1995) and furthered our understanding of factors structuring the occurrence, abundance, and composition for lotic fish assemblages (e.g., Angermeier and

Schlosser 1989; Poff and Allan 1995). Unfortunately, a similar focus on lacustrine environments and their aquatic biota has progressed at a much slower rate and lags far behind lotic systems. 12

The reasons for the disparity between our understanding of lentic and lotic fish assemblage structure are due, in part, to the physical diversity of lacustrine ecosystems and associated problems with sampling lakes compared to rivers and streams. For example, few methods are generally needed to efficiently sample lotic fishes in stream reaches (e.g., Neebling and Quist 2011), whereas a wide variety of lentic fish sampling methods and techniques have been evaluated for sampling entire lake fish assemblages (e.g., Jackson and Harvey 1997;

McInenry and Cross 2004; Clark et al. 2007). Multiple fish sampling methods are used in lentic habitats because lakes can have distinct physiochemical zones (i.e., littoral, pelagic). Differences in fish habitat use between littoral and pelagic zones can affect the efficiency of sampling methods and necessitate the use multiple of gears to characterize fish assemblages. As such, zonation in lentic ecosystems has led to research focused on describing fish assemblages of individual zones (e.g., offshore; McQueen et al. 1986; Gido and Matthews 2000; littoral; Weaver et al. 1993; Ruetz et al. 2007) as opposed to attempts to describe whole-lake assemblages. In contrast, studies of lotic systems have long attempted to describe the structure and function of multiple species that are often considered representative of the fish assemblage. The lack of information regarding sampling of multiple species to characterize whole-lake fish assemblages is also the result of confusion as to what is considered lentic (e.g., lakes, wetlands, ponds, reservoirs). Perhaps the greatest discrepancy arises from natural lakes and impounded riverine systems (hereafter referred to as impoundments). Impoundments are socially and economically important, and provide numerous ecological goods and services that are similar to natural lakes.

However, impoundments often have chemical, physical, and biological characteristics that are similar to both lakes and rivers that can also have substantial effects on the efficiency of fish sampling methods. 13

An additional factor that has contributed to slower progress of developing standard

sampling methods in lentic ecosystems is the well-known effects of temporal variation in fish

abundance and size structure (Guy and Willis 1991; Pope and Willis 1996). Whereas rivers and

streams are commonly sampled at baseflow when sampling efficiency is maximized, lentic

environments are commonly sampled with different gears at specific times of the year due to

temporal patterns of fish use (e.g., summer offshore movement of fish, spawning) and

recruitment of fish to sampling gears. For instance, young cohorts of small-bodied species that hatch in spring may not be susceptible to standard methods until the following year, whereas age-0 large-bodied species hatched in spring may be collected during their first fall. The diversity of factors limiting the development of widely accepted lentic fish sampling methods has resulted in the practice of targeted sampling for species or groups of species. Examples of targeted sampling, primarily for sport fishes, in lentic systems are ubiquitous in the literature

(e.g., Paragamian 1989; Guy and Willis 1991; Allen et al. 1999; Sammons et al. 2002;

Bonvechio et al. 2008). However, describing the influence of multiple sampling techniques on fish assemblage characterization is crucial to understanding complex ecosystems such as lakes and impoundments.

All fish sampling methods are selective (e.g., species, size) due to physical attributes of the gear and(or) the sampling protocol. For example, differences in construction materials and dimensions of fish sampling gears have been repeatedly demonstrated to result in samples of different species and sizes of fish (Henderson and Nepszy 1992; Schultz and Haines 2005;

Fischer et al. 2010). The gear-specific sampling protocol (e.g., night versus day electrofishing) can also affect fish assemblage characterization and has received considerable attention

(McInerny and Cross 1996; Thurow and Shill 1996; Pierce et al. 2001; Riha et al. 2011). 14

Therefore, the use of multiple methods to reduce bias can account for underrepresentation of

species that are hard to detect and is often necessary for a variety of aquatic habitats (e.g.,

Neebling and Quist 2011) including lakes (Jackson and Harvey 1997). Accurately characterizing

assemblage structure is particularly important for biological assessment, where the presence and

relative abundance of species and functional groups (e.g., trophic guilds) in a system are crucial to the interpretation and evaluation of ecological impairment. Since all fish sampling methodologies have biases, evaluating the use of multiple sampling methods is important for developing and implementing monitoring designs that represent the biological assemblage of an ecosystem. Additionally, the recent development of standard freshwater fish sampling methods

(Bonar et al. 2009) includes different techniques (i.e., gears and seasons) for sampling small (i.e.,

≤ 200 ha; Pope et al. 2009) and large standing waters (i.e., > 200 ha; Miranda and Boxrucker

2009). Disparity in techniques would greatly limit comparisons of data collected from different water bodies. However, we are unaware of previous studies that have evaluated the influence of season on multiple sampling methods for characterizing fish assemblage structure in lentic ecosystems. The goal of our study was provide researchers and natural resource managers with information on the influence of multiple sampling gears and protocols (e.g., sample timing) used to characterize lentic fish assemblages. Specifically, our objectives were to 1) compare several passive and active gears to detect the presence of species sampled seasonally, 2) evaluate the influence of sampling intensity and timing on estimates of species richness using multiple sampling methods, and 3) estimate the potential trade-offs of using multiple sampling gears to characterize lentic fish assemblages while maximizing species richness and number of individuals encountered. Our results provide a relative comparison of both commonly used and novel sampling methods for freshwater fish in lentic habitats. Because sampling was consistent 15

across multiple seasons, the results provide insights on optimizing sampling strategies to

characterize fish assemblages in lakes and impoundments.

METHODS

Study sites— Lakes and impoundments were selected to represent the wide range of trophic

conditions present in Iowa (Table 1). Three natural lakes and three impoundments of low (e.g.,

high total phosphorus, low water clarity), intermediate, and high water quality were selected.

Water quality designations were based on research by Iowa Department of Natural Resources

and Iowa State University (Downing et al. 2005). The three natural lakes included West Okoboji

Lake, Lake Minnewashta, and Silver Lake. All natural lakes were located in Dickinson County

in northwest Iowa. Impoundments selected for sampling included Pleasant Creek Lake (Linn

County), Don Williams Lake (Boone County), and Prairie Rose Lake (Shelby County).

Fish sampling— Sampling gears included a variety of passive and active gears and were selected based on the recommended standard sampling methods for small and large standing waters (i.e., boat electrofishing, modified fyke nets, gill nets, seine; Bonar et al. 2009).

Supplementary gears (e.g., benthic trawl, mini-fyke nets) were also included to maximize the number of species and individuals sampled in each water body. Standard modified-fyke nets (1 m × 2 m frame, 12.7-mm bar-measure mesh, with a 15.2 m lead; Miranda and Boxrucker 2009) were used to target active species located in littoral habitats (e.g., centrarchids; Hubert 1996).

Mini-fyke nets (0.6 m × 1.2 m frame, 6.4-mm ace mesh, with a 7.6 m lead) were also used to sample small-bodied species common in shallow littoral areas (Fago 1998; Barko et al. 2004).

Experimental gill nets (i.e., sinking) were used to sample pelagic species that are not commonly sampled with fyke nets. Gill nets were 30.5 m long × 2 m deep with ten 3.1-m long panels in a 16

quasi-random order (127, 38, 57, 25, 44, 19, 64, 32, 51, 102-mm bar-measure mesh) based on recommendations of Miranda and Boxrucker (2009) and Pope et al. (2009). Fyke nets and gill nets were set at dusk and retrieved the following morning. A beach seine (9.1 m long × 2 m deep, 2-m × 2-m × 2-m bag, 6.4-mm ace mesh) was used to sample littoral fishes in areas conducive for seining (i.e., little aquatic vegetation or woody debris, few large boulders).

Quarter-arc seine hauls were conducted during the day. A small-mesh benthic trawl (i.e., mini-

Missouri trawl; Herzog 2005) was used to sample small-bodied species and juveniles of larger species from littoral and profundal benthic habitats. The trawl had a headrope length of 2.4 m, footrope length of 3.7 m, and upright height of 0.6 m. The trawl body consisted of a small (6.3- mm delta mesh) outer mesh and a large (34.9-mm bar mesh of 1.0-mm multifilament nylon) inner mesh. Trawl towlines were 38.1 m long to allow for a maximum effective depth of 5.4 m with a 7:1 drop ratio. Additional information on the design, development, and specifications of this trawl is provided by Herzog et al. (2005), Guy et al. (2009), Neebling and Quist (2011).

Trawls were towed perpendicular to the shore for 3 minutes at approximately 3.2 km/h during the

day. Pulsed DC electrofishing was used to target littoral fishes not collected with the other

sampling gears. Electrofishing efficiency is often affected by diel period (Sanders 1992;

Reynolds 1996; McInerny and Cross 2004); therefore, boat electrofishing was conducted during the day and night to evaluate differences in sample timing. Electrofishing efforts were standardized to have a 3,000W power transfer to fish (Burkhardt and Gutreuter 1995; Miranda

2009). Boat electrofishing was conducted in 5 minute runs (i.e., pedal down time) that were conducted parallel to the shoreline with two netters (6.3-mm delta mesh dipnets). All sampled fish were identified to species and measured to the nearest millimeter (total length). Fish greater 17

than 100 mm were weighed to the nearest gram. Unidentified specimens were preserved in 10%

formalin and identified in the laboratory.

Because changes in physical and chemical characteristics (e.g., water temperature, water clarity) and fish behavior throughout year (e.g., spawning, emigration) can influence estimates of population characteristics (e.g., births, recruitment to sampling methods) and assemblage composition (Pope and Willis 1996; Gido and Mathews 2000; Jordan and Willis 2001), fish were

sampled seasonally to evaluate the influence of sample timing. Sampling was conducted in the

spring (i.e., April 11 – May 31), summer (i.e., late June 22 – July 13), and fall (i.e., September 14

– October 31) of 2008. Gill nets were only used in the fall to minimize mortality. Samples were

allocated for each waterbody using a systematic random sampling design to ensure that sampling

included a diversity of habitats and that all gears were represented throughout lakes and

impoundments. Specifically, the shoreline was divided into segments that included at least one

sample from each gear (i.e., seven gears total). The number of shoreline segments, delineated for

each lake or impoundment, was based on the effort required for all sampling gears (Table 1).

For example, a 75 ha lake included 10 samples of each gear. Shoreline segments were further

divided into eight reaches. A total of eight reaches was selected to include an individual reach

for each of the seven sampling gears (i.e., mini-fyke, standard fyke, gill net, seine, trawl, day

electrofishing, night electrofishing) in addition to an alternative reach that was used to allocate

gears that were unable to be deployed in a preselected reach (e.g., enclosed swimming beach).

Individual gears were randomly assigned to reaches in each segment. Once a specific gear was

assigned to a reach, the gear was used to sample fish in that reach across all seasons.

Data analysis.—The mean number of species and individuals sampled with each gear and

season was estimated across lakes and impoundments. Additionally, the probability of detecting 18 a species when present in a waterbody was the number of samples where the species was captured divided by the total number of samples conducted. Probability of detection estimates were calculated for gears and seasons individually and across water bodies where a species was present. The presence of species across seasons in a waterbody was assumed to be constant (i.e., no emigration or immigration). Therefore, the focus of our evaluation was estimating detection probabilities when a species was known to be present in a lake or impoundment and not on estimates of site occupancy with imperfect detection rates (e.g., MacKenzie et al. 2005). For example, if at least one individual of a species was found in a lake, probability of detection was calculated for all gears and seasons regardless of encounters for the species in the lake. Species accumulation curves were constructed for water bodies, gears, and seasons. Species accumulation curves were used to evaluate the influence of season and increasing number of samples on estimates of species richness. Additionally, species accumulation results were used assess the correspondence in patterns of assemblage representation among lakes and impoundments. The total number of species and individuals encountered with combinations of gears used in a single season (i.e., with replacement) was plotted to evaluate the gain in species or individuals sampled using multiple methods. Specifically, totals of species and individuals sampled with one gear, two gears, three gears, and up to seven gears used in a single season were combined while allowing seasons to vary by gear (e.g., summer seining and fall modified-fyke nets).

RESULTS

Totals of 43 species and 61,293 fish were sampled from all six water bodies and three seasons (Table 1; Figure 1). Thirty-five species and 10,800 individuals were sampled in spring, 19

40 species and 18,189 individuals in summer, and 36 species and 32,304 individuals in fall.

Regardless of season, mini-fyke nets sampled 81% of the species observed across water bodies,

followed by night electrofishing (79%), day electrofishing (74%), fyke nets and trawling (67%),

seining (65%), and gill nets (54%). Modified-fyke nets sampled the greatest number of individuals (24,806) representing 29 species. Mini-fyke nets sampled the second largest number of individuals (10,853) representing 35 species, followed by night electrofishing with 7,234 individuals representing 34 species. Trawling sampled 8,275 individuals across 29 species, seining sampled 4,763 individuals and 28 species, and day electrofishing sampled 2,903 individuals and 32 species. Gill nets sampled the fewest number of species (23) and individuals

(2,459), but were only used in the fall. The highest number of species sampled in a season was

29 with night electrofishing (fall) and day electrofishing (fall), while the most individuals

sampled in a season was 10,085 with modified-fyke nets (fall). Spring sampling did not result in

higher a number of species relative to summer or fall for any of the gears evaluated. However,

trawling (24), modified-fyke nets (26), and mini-fyke nets (28) had the highest number of species

sampled in summer across all six lakes and impoundments.

Total and unique species by gear

The majority of species were sampled with multiple gears (median = 6 gears; Table 2).

However, four gears sampled species that were not sampled with other methods. Trout-perch

(scientific names provided in Table 2) and sauger were only sampled with the benthic trawl.

Night electrofishing was only gear that sampled emerald shiner. Shorthead redhorse and orangespotted sunfish were only sampled with modified-fyke nets, while the mini-fyke net was the only gear to sample common shiner and tadpole madtom. Common shiner, emerald shiner, 20

tadepole madtom, orangespotted sunfish, and sauger were represented by a single individual, yet

accounted for approximately 12% of the total number of species sampled across all water bodies and seasons. In contrast, several species were ubiquitous. Common carp, channel catfish, green sunfish, bluegill, largemouth bass, black crappie, and walleye were sampled in all of the water

bodies included in our study. Of these species, only green sunfish was not sampled with all of

the gears (i.e., gill nets). Several other species (i.e., gizzard shad, golden shiner, black bullhead,

yellow bullhead, white bass, smallmouth bass, white crappie, yellow perch, freshwater drum)

were sampled with every gear in the lakes and impoundments they were encountered (Table 2).

Probability of detection was consistent among seasons using a single gear for several of

the species evaluated. For example, probabilities of detection for common carp, white sucker,

and walleye sampled with fyke nets were consistent across seasons (Table 2). In contrast,

probabilities of detection of several species decreased in summer and peaked in spring and fall.

Probabilities of detection for white bass (night electrofishing), largemouth bass (day and night

electrofishing), black crappie (modified fyke nets), and walleye (night electrofishing) were

lowest during summer. Peaks in the fall likely reflect recruitment of a species to particular gear.

For example, the probability of detecting bluegill sampled with the trawl increased from 0.40 in

spring and 0.50 in summer to 0.89 in fall. Similar increases were observed for bluegill sampled

with other gears that targeted small-bodied individuals (i.e., beach seine, mini-fyke).

Differences in probabilities of detection between gears were generally large for an

individual species, indicating strong gear selectivity for many of the species sampled (Table 2).

Nearly half of the species encountered had maximum probabilities of detection with gill nets (10

species) or modified fyke nets (11 species). In contrast, substantially fewer species had

maximum probabilities of detection for the other gears evaluated. Specifically, the number of 21

species sampled with maximum probabilities of detection for night electrofishing (7 species),

mini-fyke nets (6 species), trawling (6 species), seining (4 species), and day electrofishing (3

species) were all much lower for modified fyke nets and gill nets.

Species accumulation curves (Figure 2) indicated that increased sampling effort

consistently increased the number of species encountered in individual lakes and impoundments.

Nonetheless, consistent patterns in seasonal species accumulations were observed for individual

gears. Fall maximized the number of species encountered with seining (67% of water bodies),

day electrofishing (50%), and night electrofishing (67%), while summer maximized species

encountered with trawling (83%), modified-fyke nets (83%), and mini-fyke nets (50%). Species richness from gill nets tended to asymptote with fewer samples relative to the other gears for each waterbody sampled. In contrast, species accumulation curves for seining and mini-fykes demonstrated the slowest rates of asymptotic species richness.

Plots of total species richness and number of individuals sampled with separate gear

combinations illustrated that the use of a single sampling method substantially underrepresented

the fish assemblages of the study lakes and impoundments (Figure 3). In fact, appreciable increases in the total number of species and individuals were not observed until at least three sampling methods were combined. Improvement in the number of species and individuals sampled were substantially less for gear combinations above four methods. Additionally, gear combination results were consistent between natural lakes and impoundments despite differences in the number of species and individuals sampled.

DISCUSSION

Designing protocols and choosing fish sampling methods is ultimately a compromise between logistics (e.g., time, cost), and the precision and accuracy needed to answer research and management questions. Our results indicated that the characterization of lentic fish assemblages was influenced by the selection of sampling gears and seasons. Seasonal patterns in detection

probabilities and species accumulations suggested that there were optimal seasons to use each

sampling gear when attempting to maximize the number of species sampled. However, the

optimal season to use a single gear may change depending on the sampling objectives (e.g., total number of individuals), but did not appear to be affected by lake type (i.e., natural, impoundment). Our study also demonstrated that certain gears used in a single season sampled consistently more species and individuals than others (e.g., day versus night electrofishing).

Finally, the use of multiple techniques demonstrated diminishing returns of more than four sampling gears due to strong selectivities and high catch rates of relatively few sampling methods. The consistency of our results for natural lakes and impoundments indicated that similar methods may be adequate for sampling these different ecosystems. Therefore, careful selection of multiple gears and seasons would improve fish assemblage characterization over a single gear.

The dominance of biological communities by relatively few species with the majority of taxa considered rare or uncommon has long been of scientific interest (e.g., Williams 1944).

Rare species are often of disproportionally greater management and conservation interest due to losses of biodiversity and risk of extinction (e.g., Gaston 1994; Fagan et al. 2002). However, it is unrealistic to assume that all species present in a water body can be consistently detected, regardless of the number sampling methods used and the sampling effort exerted (Krebs 1998). 23

Reliance on a single sampling method has been consistently demonstrated to underestimate

species richness in lentic habitats (e.g., Weaver et al. 1993; Jackson and Harvey 1997), further

corroborated by the current study. Therefore, sampling protocols designed to target biological

assemblages are a compromise between effort and the diminishing returns of additional species from additional sampling. Although our sampling methods were not exhaustive, they were considerably more diverse than other published studies focused on lentic fish assemblages.

Furthermore, the repeated use of several methods across multiple seasons provided a relative comparison for several commonly-used and novel freshwater fish sampling techniques. Our results suggested that relatively few gears can be used by maximize assemblage characterization of lentic habitats. For example, the combination of four methods (fall night electrofishing, summer trawl, summer fyke net, summer mini-fyke) detected 91% of the species and 28% of all individuals sampled with seven gears across all water bodies and three seasons. The representation of over 90% of the species encountered with a subset of methods is particularly promising considering 12% of the species sampled were singletons. Accounting for the majority of species with relatively few sampling methods was likely due to strong species selectivities for the sampling gears and seasons observed. However, the use of more than one gear with strong selectivities is desirable for reducing bias and developing standard fish sampling protocols that characterize assemblages.

Characterizing lacustrine fish assemblages (e.g., species richness) is difficult due to distinct physicochemical zones that vary in location throughout the year. Substantial effects of gears and seasons were observed for the lakes and impoundments included in our research.

Therefore, a combination of multiple gears used in more than one season might be necessary to adequately sample lentic fish assemblages. Several direct comparisons between sampling 24 methods provided useful information for the selection of optimal sampling methods. This was most clearly demonstrated by the comparison of day and night electrofishing. Night electrofishing consistently sampled more species and individuals than day electrofishing.

Consequently, probabilities of detection were often higher and species accumulation curves tended to asymptote with fewer samples for night relative to day electrofishing. Several other studies that have compared the influence of diel period on electrofishing sampling data (e.g.,

Paragamian 1989; Pierce et al. 2001; McInerny and Cross 2004) have found similar results.

However, the difference between the number of fish and species sampled with day and night electrofishing is likely due to high water clarity of the study lakes.

Another finding worth noting is with regard to gears that sampled small-bodied species or age-0 individuals (i.e., beach seine, benthic trawl, mini-fyke nets). Beach seines are a commonly used sampling gear for targeting small-bodied species in standing waters (Hayes et al. 1996;

Pope et al. 2009). However, capture efficiencies of seines are generally low for benthic species relative to those that inhabit the water column (Lyons 1986; Pierce et al. 1990). Furthermore, seining catch rates can often be inconsistent as a result of obstructions (e.g., woody debris, boulders). The mini-Missouri trawl has been an effective sampling method for small-bodied species in lotic systems (Herzog et al. 2009; Neebling and Quist 2011). For example, Herzog et al. (2009) demonstrated numerous detections of rare species (e.g., shoal chub Macrohybopsis hyostoma, sturgeon chub Macrhybopsis gelida, crystal darter Crystallaria asprella) at previously undocumented locations or re-discovery of species thought to extirpated throughout lotic habitats in the Mississippi River basin. Neebling and Quist (2008) documented the first collection of western sand darter Ammocrypta clara in Iowa’s interior rivers since 1958 using the mini-

Missouri trawl. Our results suggest that a mini-Missouri trawl may be an alternative to seining 25

as trout-perch, Iowa darter, and Johnny darter had the highest probabilities of detection with the

benthic trawl. We found the mini-Missouri trawl to be an effective sampling gear for littoral

areas when towed perpendicular from the shore relative to seining and mini-fyke nets.

Specifically, the ability to sample water bodies in a single visit (i.e., seining, trawling) is often

beneficial over passive gears (e.g., mini-fykes) when lengthy travel is necessary. Sampling with

trawls can also be less physically demanding than seining, because the watercraft is used to pull

the net through the sampled habitat. Therefore, additional research on the use (e.g., cost, labor)

of the mini-Missouri trawl in lentic habitats is warranted given our results.

Management implications

The choice of freshwater fish sampling methods for lentic ecosystems can substantially

influence the interpretation of data. For example, the common practice of targeting

recreationally-important species with a single method will infrequently be representative of the

fish assemblage present in a waterbody. Furthermore, data are increasingly collected from aquatic ecosystems to achieve multiple research, management, and conservation objectives (e.g., sport fish, biomonitoring). Although increasing the number of methods to sample more species may not always be justified, a multiple-gear approach provides a more complete characterization

of the fish assemblage. Furthermore, using multiple sampling techniques may afford additional

understanding of population characteristics (e.g., recruitment variability) by providing estimates of abundance for different life history stages of particular species. Additional information from multiple gears can be crucial to determining when an insufficient number of fish may limit inferences for targeted populations (e.g., monitoring rare species). Obviously more samples and gears will always provide more information, but researchers will remain constrained by 26

logistical, social, and economic limitations. Therefore, developing consistent sampling methods

that can be used across a wide variety of lentic systems is desirable to maximize the information gained and provide comparable data across temporal and spatial scales. Our comparison of

several sampling techniques provides guidance on the development of fish sampling protocols

designed to characterize lentic fish assemblages.

ACKNOWLEDGEMENTS

We thank M. Mork, S. Bisping, T. Neebling, J. Koch, W. Massure, A. Schaefer, R.

Schultz, A. Fowler, and J. Christiansen for assistance with field collection. We also thank P.

Dixon, J. Downing, C. Paukert, C. Pierce, and K. Roe for helpful comments on a previous

version of this manuscript. This project was supported, in part, by the Department of Natural

Resource Ecology and Management-Iowa State University, Iowa Department of Natural

Resources, and Idaho Cooperative Fish and Wildlife Research Unit. The Unit is jointly

sponsored by the University of Idaho, U.S. Geological Survey, Idaho Department of Fish and

Game, and Wildlife Management Institute. The use of trade, firm, or product names is for

descriptive purposes only and does not imply endorsement by the U.S. Government. #1-08-

6493-I.

REFERENCES

Allen, M. S., M. M. Hale, and W. E. Pine III. 1999. Comparison of trap nets and otter trawls for

sampling black crappie in two Florida lakes. North American Journal of Fisheries

Management 19:977–983. 27

Angermeier, P. L., and R. A. Smogor. 1995. Estimating number of species and relative

abundances in stream-fish communities: effects of sampling effort and discontinuous

spatial distributions. Canadian Journal of Fisheries and Aquatic Sciences 52:936–949.

Angermeier, P. L., and I. J. Schlosser. 1989. Species–area relationships for stream fishes.

Ecology 70:1450–1462.

Barko, V. A., D. P. Herzog, R. A. Hrabik, and J. S. Scheibe. 2004. Relationship among fish

assemblages and main-channel-border physical habitats in the unimpounded upper

Mississippi River. Transactions of the American Fisheries Society 133:371–384.

Bonar S. A., W. A. Hubert, and D. W. Willis. 2009. Standard sampling methods for North

American freshwater fishes. American Fisheries Society, Bethesda, Maryland.

Bonvechio, T. F., W. F. Pouder, and M. M. Hale. 2008. Variation between electrofishing and

otter trawling for sampling black crappies in two Florida lakes. North American Journal

of Fisheries Management 28:188–192.

Burkhardt, R. W., and S. Gutreuter. 1995. Improving electrofishing consistency by

standardizing power. North American Journal of Fisheries Management 15:375–381.

Clark, S. J., J. R. Jackson, and S.E. Lochmann. 2007. A comparison of shoreline seines with

fyke nets for sampling littoral fish communities in floodplain lakes. North American

Journal of Fisheries Management 27:676–680.

Downing, J. A., J. Li, G. Antoniou, D. Kendall, C. Kling, J. Herriges, R. Castro, P. Van Meter,

D. Woolnough, K. Egan, Y. Jeon, R. Andrews, S. Conrad, and L. Boatwright. 2005.

Iowa Lakes Classification for Restoration. Iowa Department of Natural Resources, Des

Moines. 28

Fagan, W. F., P. J. Unmack, C. Burgess, and W. L. Minckley. 2002. Rarity, fragmentation, and

extinction risk in desert fishes. Ecology 83:3250–3256.

Fago, D. 1998. Comparison of littoral fish assemblages sampled with a mini-fyke net or a

combination of electrofishing and small-mesh seine in Wisconsin lakes. North American

Journal of Fisheries Management 18:731–738.

Fischer, J. R., N. P. Johnson, R. D. Schultz, and M. C. Quist. 2010. A comparison of modified

fyke nets for evaluating fish assemblages and population structure. Journal of Freshwater

Ecology 25:555–563.

Gaston, K. J. 1994. Rarity. Chapman and Hall, London, UK.

Gido, K. B., and W. J. Matthews. 2000. Dynamics of the offshore fish assemblage in a

southwestern reservoir (Lake Texoma, Oklahoma-Texas). Copeia 2000:917–930.

Guy, C. S., and D. W. Willis. 1991. Seasonal variation in catch rate and body condition for four

fish species in a South Dakota natural lake. Journal of Freshwater Ecology 6:281–292.

Guy, C. S., P. J. Braaten, D. P. Herzog, J. Pitlo, and R. S. Rogers. 2009. Warmwater fish in

rivers. Pages 59–84 in S. Bonar, W. Hubert, and D. Willis, editors. Standard sampling

methods for North American freshwater fishes. American Fisheries Society, Bethesda,

Maryland.

Hayes D. B., C. P. Ferreri, and W. W. Taylor. 1996. Active fish capture methods. Pages 193–

220 in B. R. Murphy and D. W. Willis, editors. Fisheries techniques, 2nd edition.

American Fisheries Society, Bethesda, Maryland.

Henderson, B. A., and S. J. Nepszy. 1992. Comparison of catches in mono- and multifilament

gill nets in Lake Erie. North American Journal of Fisheries Management 12:618–624 29

Herzog, D. P., V. A. Barko, J. S. Scheibe, R. A. Hrabik, and D. E. Ostendorf. 2005. Efficacy of

a benthic trawl for sampling small-bodied fishes in large river systems. North American

Journal of Fisheries Management 25:594–603.

Herzog, D. P., D. E. Ostendorf, R. A. Hrabik, and V. A. Barko. 2009. The mini-Missouri trawl:

a useful methodology for sampling small-bodied fishes in small and large river systems.

Journal of Freshwater Ecology 24:103–108.

Hubert, W. A. 1996. Passive capture techniques. Pages 157–192 in B. R. Murphy and D. W.

Willis, editors. Fisheries techniques, second edition. American Fisheries Society,

Bethesda, Maryland.

Jackson, D. A., and H. H. Harvey. 1997. Qualitative and quantitative sampling of lake fish

communities. Canadian Journal of Fisheries and Aquatic Sciences 54:2807–2813.

Jordon, G. R., and D. W. Willis. 2001. Seasonal variation in sampling indices for shovelnose

sturgeon, river carpsucker, and shorthead redhorse collected from the Missouri River

below Fort Randall Dam, South Dakota. Journal of Freshwater Ecology 16:331–340.

Karr, J. R. 1981. Assessment of biotic integrity using fish communities. Fisheries 6(6):21–27.

Krebs, C. J. 1998. Ecological methodology, second edition. Benjamin Cummings, Menlo Park,

California.

Lyons, J. 1986. Capture efficiency of a beach seine for seven freshwater fishes in a North-

temperate lake. North American Journal of Fisheries Management 6:288–289.

Lyons, J. 1992. The length of stream to sample with a towed electrofishing unit when species

richness is estimated. North American Journal of Fisheries Management 12:198–203. 30

MacKenzie, D. I., J. D. Nichols, J. A. Royle, K. H. Pollock, L. L. Bailey, and J. E. Hines. 2005.

Occupancy estimation and modeling: Inferring patterns and dynamics of species

occurrence. Elsevier, San Diego, California.

McInerny, M. C., and T. K. Cross. 1996. Seasonal and diel variation in electrofishing size-

selectivity and catch-per-hour of largemouth bass in Minnesota lakes. Investigational

Report 451. Minnesota Department of Natural Resources, Hutchinson.

McInerny, M. C., and T. K. Cross. 2004. Comparison of day electrofishing, night

electrofishing, and trap netting for sampling inshore fish in Minnesota Lakes. Special

publication 161. Minnesota Department of Natural Resources, Hutchinson.

McQueen, D. J., J. R. Post, and E. L. Mills. 1986. Trophic relationships in freshwater pelagic

ecosystems. Canadian Journal of Fisheries and Aquatic Sciences 43:1571–1581.

Miranda, L. E. 2009. Standardizing electrofishing power for boat electrofishing. Pages 223–

230 in S. Bonar, W. Hubert, and D. Willis, editors. Standard sampling methods for North

American freshwater fishes. American Fisheries Society, Bethesda, Maryland.

Miranda, L. E., and J. Boxrucker. 2009. Warmwater fish in large standing waters. Pages 29–42

in S. Bonar, W. Hubert, and D. Willis, editors. Standard sampling methods for North

American freshwater fishes. American Fisheries Society, Bethesda, Maryland.

Neebling, T. E., and M. C. Quist. 2008. Observations on the distribution and status of western

sand darter, spotted gar, and skipjack herring in Iowa rivers. Journal of the Iowa

Academy of Science 115:24–27.

Neebling, T. E., and M. C. Quist. 2011. Comparison of boat Electrofishing, trawling, and

seining for sampling fish assemblages in Iowa's nonwadeable rivers. North American

Journal of Fisheries Management 31:390–402. 31

Paragamian, V. L. 1989. A comparison of day and night electrofishing: size structure and catch

per unit effort for smallmouth bass. North American Journal of Fisheries Management

9:500–503.

Pierce C. L., J. B. Rasmussen, W. C. Leggett. 1990. Sampling littoral fish with a seine:

corrections for variable capture efficiency. Canadian Journal of Fisheries and Aquatic

Sciences 47:1004–1010.

Pierce, C. L., A. M. Corcoran, A. N. Gronbach, S. Hsia, B. J. Mularkey, and A. J. Schwartzhoff.

2001. Influence of diel period on electrofishing and beach seining assessments of littoral

fish assemblages. North American Journal of Fisheries Management 21:918–926.

Poff N. L., and J. D. Allan. 1995. Functional organization of stream fish assemblages in relation

to hydrological variability. Ecology 76:606–627.

Pope, K. L., and D. W. Willis. 1996. Seasonal influences of freshwater fisheries sampling data.

Reviews in Fisheries Science 4:57–73.

Pope, K. L., R. M. Neumann, and S. D. Bryan. 2009. Warmwater fish in small standing waters.

Pages 13–28 in S. Bonar, W. Hubert, and D. Willis, editors. Standard sampling methods

for North American freshwater fishes. American Fisheries Society, Bethesda, Maryland.

Reynolds, J. B. 1996. Electrofishing. Pages 221–253 in B. R. Murphy and D. W. Willis,

editors. Fisheries techniques, second edition. American Fisheries Society, Bethesda,

Maryland.

Riha, M., J. Kubecka, M. Prchalova, T. Mrkvicka, M. Cech, V. Drastik, J. Frouzova, E.

Hohausova, T. Juza, M. Kratochvil, J. Peterka, M. Tuser, and M. Vasek. 2011. The

influence of diel period on fish assemblage in the unstructured littoral of reservoirs.

Fisheries Management and Ecology 18:339–347. 32

Ruetz C. R., D. G. Uzarski, D. M. Krueger, and E. S. Rutherford. 2007. Sampling a littoral fish

assemblage: comparing small-mesh fyke netting and boat electrofishing. North

American Journal of Fisheries Management 27:825–831.

Sammons, S. M., D. A. Isermann, and P. W. Bettoli. 2002. Variation in population

characteristics and gear selection between black and white crappies in Tennessee

reservoirs: Potential effects on management decisions. North American Journal of

Fisheries Management 22:863–869.

Sanders, R. E. 1992. Day versus night electrofishing catches from near-shore waters of the Ohio

and Muskingum Rivers. Ohio Journal of Science 92(3):51–59.

Schultz, R. D., and D. E. Haines. 2005. Comparison of seasonal bluegill catch rates and size

distributions obtained with trap nets and electrofishing in a large, heated impoundment.

North American Journal of Fisheries Management 25:220–224.

Thurow, R. F., and D. J. Schill. 1996. Comparison of day snorkeling, night snorkeling, and

electrofishing to estimate bull trout abundance and size structure in a second order Idaho

stream. North American Journal of Fisheries Management 16:314–323.

Weaver, M. J., J. J. Magnuson, and M. K. Clayton. 1993. Analyses for differentiating littoral

fish assemblages with catch data from multiple sampling gears. Transactions of the

American Fisheries Society 122:1111–1119.

Williams, C. B. 1944. Some applications of the logarithmic series, and the index of diversity to

ecological problems. Journal of Ecology 32:1–44.

TABLE 1.—Water body type, area (ha), mean depth (Z; m), sampling effort (number of net nights for passive gears, number of seine hauls, number of 3-minute trawls, and number of 5- minute electrofishing runs conducted within a season), mean secchi depth (m), mean chlorophyll a (Chl-a; µg/L) concentration, mean total phosphorus (TP; µg/L) concentrations. Water body information obtained from the Iowa Department of Natural Resources Lakes Information Report. Sampling Secchi Water body Type Area Z effort depth Chl-a TP Prairie Rose Lake Impoundment 70 2.7 10 0.7 48 91 Don Williams Lake Impoundment 60 5.5 10 1.7 24 84 Pleasant Creek Lake Impoundment 162 4.8 12 2.1 15 40 Silver Lake Natural lake 432 2.3 20 0.7 37 167 Lake Minnewashta Natural lake 48 3.1 10 1.7 22 116 West Okoboji Lake Natural lake 1,557 11.6 20 5.4 4 27

TABLE 2.—Probability of detection and the number of water bodies present (N) for seven sampling methods used in the spring (Sp), summer (Su), and fall (Fa) to characterize fish assemblages of six lakes and impoundments in Iowa, 2008 Gill Seine Trawl Day EF Night EF Fyke net Mini-fyke net net Family and species N Sp Su Fa Sp Su Fa Sp Su Fa Sp Su Fa Sp Su Fa Sp Su Fa Fa Lepisosteidae Shortnose gar Lepisosteus platostomus 3 0.13 0.10 0.08 0.18 Clupeidae Gizzard shad Dorosoma cepedianum 2 0.05 0.05 0.23 0.05 0.36 0.64 0.27 0.91 0.59 0.09 0.05 0.14 0.05 0.96 Cyprinidae Spotfin shiner Cyprinella spiloptera 2 0.07 0.07 Common carp Cyprinus carpio 6 0.01 0.01 0.05 0.04 0.01 0.26 0.24 0.21 0.27 0.20 0.24 0.57 0.60 0.56 0.01 0.01 0.02 0.45 Common shiner Luxilus cornutus 1 0.05 Golden shiner Notemigonus crysoleucas 4 0.08 0.02 0.02 0.19 0.10 0.14 0.08 0.17 0.06 0.04 0.02 0.06 0.06 0.10 Emerald shiner Notropis atherinoides 1 0.05 Spottail shiner Notropis hudsonius 4 0.08 0.03 0.03 0.07 0.12 0.03 0.05 0.18 0.05 0.13 0.02 0.05 0.02 0.05 Bluntnose minnow Pimephales notatus 4 0.05 0.17 0.12 0.15 0.13 0.12 0.03 0.15 0.02 0.08 0.03 0.08 0.02 Fathead minnow Pimephales promelas 4 0.03 0.02 0.03 0.03 0.02 0.02 0.02 0.02 Bullhead minnow Pimephales vigilax 3 0.04 0.18 0.04 0.18 0.12 0.10 0.06 0.02 0.08 0.14 0.02 0.04 0.08 Creek chub Semotilus atromaculatus 1 0.30 0.40 Catostomidae River carpsucker Carpiodes carpio 4 0.19 0.06 0.23 0.02 0.15 0.06 0.10 Quillback Carpiodes cyprinus 2 0.05 0.05 0.20 0.20 0.10 0.60 White sucker Catostomus commersonii 4 0.02 0.02 0.03 0.08 0.08 0.07 0.12 0.13 0.07 0.42 0.43 0.45 0.07 0.08 0.05 0.48 Bigmouth buffalo Ictiobus cyprinellus 4 0.05 0.02 0.13 0.08 0.20 0.12 0.07 0.03 0.03 0.18 0.03 0.15 Shorthead redhorse Moxostoma macrolepidotum 2 0.03 0.03 Ictaluridae Black bullhead Ameiurus melas 4 0.02 0.02 0.02 0.12 0.08 0.10 0.02 0.18 0.05 0.35 0.30 0.17 0.77 0.83 0.42 0.57 0.47 0.33 0.45 Yellow bullhead Ameiurus natalis 5 0.01 0.01 0.01 0.01 0.01 0.01 0.13 0.10 0.13 0.26 0.39 0.22 0.13 0.01 0.08 0.15 Tadpole madtom Noturus gyrinus 1 0.05 Channel catfish Ictalurus punctatus 6 0.01 0.01 0.05 0.01 0.06 0.07 0.07 0.11 0.16 0.16 0.05 0.45 0.06 0.02 0.01 0.63 Flathead catfish Pylodictis olivaris 3 0.06 0.03 0.06 Esocidae Northern pike Esox lucius 3 0.02 0.02 0.12 0.08 0.04 0.06 0.04 0.12 0.16 0.32 0.28 0.08 0.02 0.30 Muskellunge Esox masquinongy 3 0.02 0.02 0.05 0.07 0.19 0.02 0.05 0.02 0.02 0.02 Percopsidae Trout-perch Percopsis omiscomaycus 2 0.10 0.03 Moronidae White bass Morone chrysops 5 0.03 0.06 0.03 0.07 0.26 0.11 0.28 0.51 0.22 0.61 0.54 0.72 0.75 0.11 0.10 0.25 0.94

TABLE 2.—Continued Yellow bass Morone mississippiensis 2 0.20 0.10 0.33 0.30 0.33 0.43 0.03 0.37 0.20 0.03 0.87 Centrarchidae Green sunfish Lepomis cyanellus 6 0.04 0.01 0.01 0.16 0.12 0.21 0.22 0.11 0.17 0.11 0.28 0.13 0.09 0.32 0.27 Pumpkinseed Lepomis gibbosus 3 0.05 0.05 0.07 0.12 0.05 0.12 0.19 0.29 0.26 0.07 0.05 0.17 Orangespotted sunfish Lepomis humilis 1 0.05 Bluegill Lepomis macrochirus 6 0.16 0.34 0.56 0.40 0.50 0.89 0.49 0.52 0.57 0.76 0.73 0.70 0.88 0.92 0.79 0.43 0.43 0.66 0.18 Redear sunfish Lepomis microlophus 2 0.03 0.03 0.03 0.13 Smallmouth bass Micropterus dolomieu 2 0.03 0.03 0.23 0.20 0.27 0.27 0.43 0.20 0.37 0.03 0.03 0.23 Largemouth bass Micropterus salmoides 6 0.05 0.27 0.26 0.02 0.11 0.11 0.55 0.46 0.65 0.63 0.45 0.57 0.09 0.07 0.20 0.05 0.04 0.05 0.06 White crappie Pomoxis annularis 4 0.10 0.06 0.14 0.02 0.06 0.04 0.08 0.10 0.21 0.04 0.14 Black crappie Pomoxis nigromaculatus 6 0.17 0.16 0.13 0.52 0.40 0.11 0.10 0.17 0.34 0.15 0.33 0.72 0.59 0.87 0.11 0.18 0.52 0.39 Percidae Iowa darter Etheostoma exile 3 0.10 0.06 0.32 0.28 0.30 0.08 0.02 0.02 0.06 0.02 Johnny darter Etheostoma nigrum 5 0.06 0.08 0.10 0.22 0.42 0.17 0.03 0.03 0.01 0.07 0.01 Logperch Percina caprodes 2 0.13 0.07 0.17 0.03 0.17 0.03 0.07 0.20 0.07 0.37 0.03 Yellow perch Perca flavescens 5 0.14 0.03 0.15 0.39 0.17 0.08 0.19 0.22 0.24 0.32 0.38 0.39 0.53 0.54 0.01 0.11 0.11 0.76 Sauger Sander canadensis 1 0.10 Walleye Sander vitreus 6 0.04 0.02 0.04 0.22 0.09 0.04 0.09 0.20 0.42 0.35 0.55 0.43 0.48 0.55 0.04 0.12 0.15 0.71 Sciaenidae Freshwater drum Aplodinotus grunniens 4 0.02 0.02 0.38 0.20 0.20 0.25 0.30 0.30 0.43 0.63 0.15 0.38 0.65 0.07 0.30 0.83

FIGURE 1.—Mean and maximum number of species and individuals sampled seasonally with seven methods for six Iowa lakes and impoundments, 2008. Error bars represent one SE.

West Okoboji Minnewashta Silver P 20 Seine Spring 15 Summer Fall 10

20 Traw l 15

20 Day EF 15

20 Night EF 15

Number of species of Number 10

20 Fyke net 15

20 Mini-fyke net 15

20 Gill net 15

0 5 10 15 20 0 2 4 6 8 10 0 5 10 15 20 0 2 4 6 8 10 12 0 2 4 6 8 10 0 2 4 6 8 10 Number of samples

FIGURE 2.—Species accumulation curves for six lakes and impoundments in Iowa sampled in 2008 with seven methods.

Natural lakes Impoundments 40

Species 15

0 105

104

103

102 Numberof individuals

101 1 2 3 4 5 6 7 1 2 3 4 5 6 7 Gears Gears

FIGURE 3.—Number of species and individuals sampled with combinations of seven methods (e.g., one gear, two gears) for six Iowa lakes and impoundments, 2008.

CHAPTER 3. COMPARISON OF MULTIPLE SAMPLING METHODS FOR

ASSESSING LENTIC FRESHWATER FISH POPULATIONS

A manuscript to be submitted for publication in North American Journal of Fisheries

Management

Jesse R. Fischer3 and Michael C. Quist4

Abstract

All freshwater fish sampling methods are biased towards particular species, sizes, and

sexes and are further influenced by season, habitat, and fish behavior. Thus, the development of

standard sampling methods is important for repeatability of assessments and comparisons of data

collected across various temporal and spatial scales. We sampled six lakes and impoundments

representing a diversity of trophic conditions in Iowa, USA, using multiple gears (i.e., modified

fyke-nets, experimental gill nets, beach seine, benthic trawl, boat electrofishing) to determine the

influence of sampling methodology and season on freshwater fisheries assessments.

Specifically, we were interested in describing the influence of season on catch-per-unit-effort

(CPUE), size structure (proportional size distribution [PSD]), and the number of samples required to obtain 125 stock-length individuals for 12 species of recreational and ecological importance (e.g., invasive). Gear selectivity varied from those species sampled with a variety of methods (e.g., bluegill Lepomis macrochirus, black crappie Pomoxis nigromaculatus, freshwater

drum Aplodinotus grunniens) to species predominately sampled with a single gear (e.g.,

largemouth bass Micropterus salmoides, electrofishing). Mean CPUE generally peaked in the spring and fall as a result of sampling method effectiveness (e.g., shallow littoral oriented) and habitat use of species (i.e., movement offshore during summer). Using modified-fyke nets, mean

PSD was consistent among seasons for black bullhead Ameiurus melas and yellow bullhead

Ameiurus natalis, but decreased from spring to fall for white bass Morone chrysops, bluegill, and

black crappie. The mean number of samples required to sample 125 stock-length individuals

was minimized with fall gill nets for 5 of the 12 species evaluated (walleye Sander vitreus,

yellow perch Perca flavescens, channel catfish Ictalurus punctatus, freshwater drum, and white

bass). Our results provide fisheries scientists with relative comparisons between several

recommended standard sampling methods and illustrate the effects of seasonal variation on

estimates of population indices.

INTRODUCTION

Assessment of fish populations with standard sampling methodologies is crucial to the

appropriate management of fisheries, accurate environmental monitoring, and comparison of

data across large spatial and temporal scales (Willis and Murphy 1996; Yoder and Smith 1998).

However, assessing fish populations is challenging due to numerous biases (e.g., gear, season,

location) inherent to any sampling design or technique (Ricker 1969; Hayes et al. 1996; Pope and

Willis 1996), and standard sampling methods (i.e., gears, timing) are necessary to maintain

consistency and repeatability in assessments of freshwater fishes (Bonar et al. 2009).

Quantitative assessment of fish sampling methods and constraining factors (e.g., season, time of

day, gear construction specifications and material), however, is often limited by the high cost and

41 labor associated with sampling across large temporal and spatial scales. Therefore, sampling protocols often rely on previous sampling methods; or else specific techniques must be used to convert data when sampling methods are altered from previous protocols (Peterson and Paukert

2009).

Unlike rivers and streams where relatively few standard sampling gears (e.g., electrofishing, seines, trammel nets) are typically used to characterize fisheries (Guy et al. 2009;

Rabeni et al. 2009), numerous gears (e.g., electrofishing, seines, fyke nets, gill nets, trawling, hoop nets) are regularly used to sample fish in lentic systems (Murphy and Willis 1996; Miranda and Boxrucker 2009). Multiple methods are typically required because lakes and impoundments have distinct zones (i.e., pelagic, littoral) that differ in physicochemical characteristics and fish use. Furthermore, differences in physical characteristics (e.g., depth, water clarity, vegetation) can affect the efficiency of sampling methods in differing zones. For example, more than one sampling method is often necessary to sample both juvenile and adult fish of the same species due to differing habitat use and size biases associated with sampling gears (e.g., Boxrucker et al.

1995; Bonvechio et al. 2008).

Studies evaluating the influence of multiple sampling methods are important because they inform fisheries managers and researchers about the relative bias of different sampling techniques. Gear comparison studies often focus on descriptive measures of fish assemblages

(e.g., species richness, species unique to a gear) or estimates of population characteristics (e.g., relative abundance, size distribution). Managers are often interested in determining the most appropriate methods for sampling recreationally-important fish populations (e.g., sport fish, forage species) in lentic systems (e.g., Boxrucker et al. 1995; Sammons et al. 2002). As such, estimates of catch-per-unit-effort (CPUE) and size structure (e.g., proportional size distribution

[PSD]) provide comparable measures when evaluating one or more sampling methods.

Additionally, determining the number of sampled fish required to accurately characterize fish

populations should also be considered when developing sampling strategies (Vokoun et al. 2001;

Miranda 2007; Quist et al. 2009). Sample timing (e.g., diel period, season) can also have strong influences on the interpretation of fish population assessments (e.g., Neumann and Willis 1993;

Thurow and Schill 1996; Quist et al. 2001; also see review by Pope and Willis 1996). However, previous studies that have evaluated multiple sampling gears in lentic ecosystems have generally focused on single species assessments at small spatial scales (e.g., Boxrucker et al. 1995; Cross et al. 1995; Guy et al. 1996; Allen et. al 1999) or on few gears used in a single season (e.g., Fago

1998; Pierce et al. 2001; Sammons et al. 2002; Bonvechio et al. 2008). The lack of information on the influence of sampling methods and timing to assess fish populations of species frequently targeted in lentic ecosystems necessitates additional research. Furthermore, the recent formation of standard sampling methods for sampling freshwater fishes in North America (see Bonar et al.

2009) emerged from a need to have comparable information among natural resource agencies.

The widespread adoption of standard sampling methods, like all things new, will likely meet some degree of resistance. Information on the effects of sampling methodology and timing will likely be needed for natural resource managers and agencies to fully adopt standard methods, particularly if large financial requirements are needed (e.g., purchasing new equipment for an entire state or province).

The goal of this study was to evaluate methods for sampling fish populations of 12 recreationally- and ecologically-important species (e.g., invasive) in Iowa lakes and

impoundments. The specific objectives of this study were to evaluate seasonal patterns of CPUE

and size structure estimates (i.e., PSD), and determine the number of samples (e.g., individual

hauls, net deployments) required to sample 125 stock-length individuals using a variety of sampling methods. This study provides fisheries managers and researchers with information on how available and recommended methods (e.g., Bonar et al. 2009) influence the characterization of fish populations in lentic water bodies.

METHODS

Study sites— Six lakes and impoundments were selected to represent the range of trophic

conditions present in Iowa (Table 1). Three natural lakes and three impoundments of low (i.e.,

high total phosphorus, low water clarity), intermediate, and high water quality were selected.

Impoundments included Pleasant Creek Lake (Linn County), Don Williams Lake (Boone

County), and Prairie Rose Lake (Shelby County). Natural lakes included West Okoboji Lake,

Lake Minnewashta, and Silver Lake located in Dickinson County in northwest Iowa.

Fish sampling— Fish sampling techniques included both active and passive methods.

Several of the gears were selected based on the recommended standard sampling methods for

small and large standing waters (i.e., boat electrofishing, modified fyke nets, gill nets, seine)

provided in Bonar et al. (2009). However, additional sampling methods (e.g., benthic trawl,

mini-modified-fyke nets) were also used. Standard modified-fyke nets (1 m × 2 m frame, 12.7-

mm bar-measure mesh, 15.2 m lead; Miranda and Boxrucker 2009) were used to sample

structure-oriented species, such as centrarchids located in littoral habitats (Hubert 1996). In

addition to using standard fyke nets, mini-modified-fyke nets (0.6 m × 1.2 m frame, 6.4-mm ace

mesh, 7.6 m lead) were used to sample small-bodied species and age-0 fish common in littoral

habitats (Fago 1998; Barko et al. 2004). Both types of fyke nets were set at dusk and retrieved

the following morning. Sinking experimental gill nets were used to sample fishes not typically

sampled with fyke nets (e.g., pelagic species). Gill nets were 30.5 m long × 2 m deep with ten

3.1-m long panels in a quasi-random order (127, 38, 57, 25, 44, 19, 64, 32, 51, 102-mm bar- measure mesh; Miranda and Boxrucker 2009; Pope et al. 2009). Similar to fyke nets, gill nets were deployed overnight and retrieved the following morning. A beach seine (9.1 m long × 2 m deep, 2-m × 2-m × 2-m bag, 6.4-mm ace mesh) was used to sample littoral habitats without extensive woody debris or large boulders. Seining was conducted during the day using quarter- arc hauls. Littoral and benthic habitats were also sampled with a small benthic trawl. The trawl was used to sample small-bodied species and juveniles of larger species and had a headrope length of 2.4 m, footrope length of 3.7 m, and upright length of 0.6 m. The trawl body consisted of a small (6.3-mm delta mesh) outer mesh and a large (34.9-mm bar mesh of 1.0-mm multifilament nylon) inner mesh. Trawl towlines were 38.1 m long to allow for a maximum effective depth of 5.4 m with a 7:1 drop ratio. Additional information on the design, development, and specification of the Mini-Missouri trawl as used in lotic habitats is described by Herzog et al. (2005), Guy et al. (2009), Neebling in Quist (2011). Trawls were towed perpendicular from the shore for 3 minutes at approximately 3.2 km/h during the day. Lastly, pulsed DC electrofishing was used to sample fishes not collected with other sampling gears.

Because electrofishing catch rates are influenced by diel period (Sanders 1992; Reynolds 1996;

McInerny and Cross 2004), electrofishing was conducted during the day and night. Boat electrofishing output was standardized at 2,750-3,250 W (Burkhardt and Gutreuter 1995;

Miranda and Boxrucker 2009; Miranda 2009). Electrofishing runs were conducted for 5 minutes, parallel to the shoreline, and with two netters using 6.3-mm delta mesh dipnets. All sampled fish were identified to species and measured to the nearest millimeter of total length

(TL). Fish greater than 100 mm TL were weighed to the nearest gram. Unidentified specimens were preserved in 10% formalin and identified in the laboratory.

Substantial temporal variation in samples can be caused by changes in physicochemical characteristics (e.g., water temperature, water clarity) and differences in fish behavior (e.g., spawning, emigration). Seasonal variation can influence estimates of population characteristics

(e.g., births, recruitment to sampling methods), as well as assemblage composition (Pope and

Willis 1996; Gido and Mathews 2000; Jordan and Willis 2001). Thus, fish were sampled seasonally to evaluate the optimal time of year to characterize fish populations for each sampling gear. Samples were collected in the spring (i.e., April 11 – May 31), summer (i.e., late June 22 –

July 13), and fall (i.e., September 14 – October 31). Experimental gill nets were only used in the fall to minimize mortality associated with overnight sets.

A systematic random sampling design was used to allocate samples for each waterbody and ensure that a diversity of habitats was sampled throughout lakes and impoundments.

Specifically, the shoreline was divided into segments that included at least one sample from each gear. The number of shoreline segments, delineated for each lake or impoundment, was based on the effort required for all sampling gears (Table 1). For example, a 75 ha lake included 10 samples of each gear. Therefore, at least 10 shoreline segments were identified to ensure a sample from each sampling method. Shoreline segments were further divided into eight reaches.

A total of eight reaches was selected to include an individual reach for each of the seven sampling gears (i.e., mini-fyke, standard fyke, gill net, seine, trawl, day electrofishing, night electrofishing) in addition to an alternate reach that was used to allocate gears that were unable to be used in a preselected reach (e.g., enclosed swimming beach, spillway). Individual gears were

randomly assigned to reaches in each segment. Gears assigned to specific reach were used to

sample fish in that reach throughout the study (i.e., spring, summer, fall).

Data analysis.— Although species were not directly targeted during sampling, comparisons between sampling gears and timing focused on 12 species that commonly occur in

lentic ecosystems throughout North America (Lee et al. 1980). Species of interest included

common carp Cyprinus carpio, black bullhead Ameiurus melas, yellow bullhead A. natalis,

channel catfish Ictalurus punctatus, white bass Morone chrysops, bluegill Lepomis macrochirus,

largemouth bass Micropterus salmoides, white crappie Pomoxis annularis, black crappie P.

nigromaculatus, yellow perch Perca flavescens, walleye Sander vitreus, and freshwater drum

Aplodinotus grunniens. Catch-per-unit-effort for species sampled with an individual gear would

be expected to change throughout the year with differences in behavior (e.g., spawning) and

growth (e.g., size-selectivity). Therefore, mean CPUE was estimated as the number of fish sampled for each species, gear, and season across water bodies. Estimates of CPUE for mini- fyke nets, standard fyke nets, and gill nets were the number of fish per net night. Estimates of

CPUE for the seine and trawl were the number of fish per haul. Estimates of CPUE for day and night electrofishing were calculated as the number of fish per hour of electrofishing. Mean

CPUE was estimated separately for stock- and substock-length individuals to evaluate the selectivity of each gear.

Proportional size distribution is a common index used to numerically describe length- frequency distributions, thereby providing useful information on recruitment, growth, and mortality of fishes (Willis et al. 1993; Anderson and Neumann 1996; Guy et al. 2007).

Proportional size distributions can be influenced by biased sampling associated with selectivity for different sizes of fish throughout the year and should be considered when determining

sampling methods (see Willis et al. 1993; Pope and Willis 1996). Size structure of populations

was estimated using mean PSD for all species using standard length categories for all gear and

seasons where at least 50 stock-length individuals were sampled (Gabelhouse 1984; Anderson

and Neumann 1996; Bister et al. 2000; Guy et al. 2007).

Determining the number of individual fish required (i.e., sampled and measured) to

accurately estimate characteristics (e.g., mean length, size structure, PSD) of a fish population is

essential to choosing standard methods for fisheries assessments. Furthermore, the required

number of individuals depends on characteristics of the species (e.g., body size) and the

population (e.g., size and age structure; Vokoun et al. 2001; Miranda 2007). For example,

Vokoun et al. (2001) recommended measuring 300-400 individuals to estimate length frequency

using information on channel catfish and bluegill populations. Miranda (2007) recommended

measuring 375-1,200 and 150-425 individuals to estimate 1-cm and 2.5-cm length-frequency histograms, respectively, using data from black crappie, bluegill, and largemouth bass populations. Additionally, the characteristic being estimated can also have an influence on the

required number of individuals. Miranda (2007) reported that 75-160 and 75-140 fish were required to accurately estimate mean length and PSD, respectively, for black crappie, bluegill, and largemouth bass. Following these guidelines, Quist et al. (2009) recommend a minimum sample size of 125 for calculating PSD. Therefore, the mean number of samples required to collect 125 stock-length individuals were calculated for each season and gear. Additionally, the proportion of water bodies where 50 and 125 stock-length individuals were sampled was calculated for each gear and season.

RESULTS

Across all water bodies, seasons, and gears, 43 species and 61,293 fish were sampled.

The most abundant species sampled was bluegill (27,940 individuals), followed by black crappie

(9,308), freshwater drum (6,021), black bullhead (3,306), white bass (2,383) and largemouth bass (2,085). The 12 focal species accounted for 91% of all individuals sampled among all seasons and sampling methods. Bluegill, black bullhead, and largemouth bass were sampled with every sampling method in every season, while several other species (i.e., common carp, black crappie, yellow perch, walleye) were sampled with all methods in nearly every season

(Table 2).

Gears that specifically targeted small-bodied and age-0 fish (i.e., seine, trawl, mini-fyke nets) were effective at sampling substock-length individuals for several of the focal species

(Figure 1). Specifically, mean CPUE was highest for substock bluegill and yellow perch with mini-fyke nets. Substock-length individuals of some species were almost exclusively sampled with a single method in an individual season. For example, mean CPUE was highest for substock-length largemouth bass (seine) and black crappie (trawl) during the summer. In contrast, substock-length white bass were sampled with several gears with higher mean CPUE in the fall. Seasonal patterns in mean CPUE of substock-length individuals were also observed for several of the gears. Mean CPUE of substock-length walleye and yellow perch was lowest in the summer with peaks in spring and fall. Increases in mean CPUE throughout the year were observed for substock-length bluegill (trawl) and yellow perch (day electrofishing), while decreases from spring to fall were observed for common carp (fyke nets). Similar patterns between mean CPUE of substock-length individuals and CPUE of stock-length individuals were found for common carp (fyke nets) and walleye (night electrofishing). Gill nets were generally

49 not effective at sampling substock-length individuals, but resulted in the highest mean CPUE for common carp and channel catfish.

Seasonal patterns in CPUE of stock-length individuals were observed for numerous species (Figure 2). For example, mean CPUE of stock-length individuals increased throughout the year (i.e., from spring to fall) for walleye (fyke nets), yellow perch (night electrofishing and fyke nets) and freshwater drum (day electrofishing). Similarly, mean CPUE of stock-length individuals decreased throughout the year for common carp with fyke nets. Catch-per-unit-effort of stock-length individuals was generally lower in the summer for several of the focal species.

Specifically, mean CPUE of stock-length individuals was lowest in the summer for black crappie

(fyke nets), largemouth bass (day and night electrofishing), walleye (night electrofishing), channel catfish (day and night electrofishing), black bullhead (fyke nets), yellow bullhead (night electrofishing and fyke nets), and white bass (day and night electrofishing, fyke nets). Although individuals were sampled with nearly every method in each season, some species were noticeably more susceptible to one or a small subset of sampling methods. For instance, stock- length largemouth bass were predominately sampled with day and night electrofishing. Black bullhead (79.6% of total individuals sampled), freshwater drum (65.8%), common carp (54.8%), white bass (49.9%), white crappie (49.5%), black crappie (47.7), and bluegill (35.0%) were primarily sampled with fyke nets regardless of season.

At least 50 stock-length bluegills were sampled in all lakes with summer fyke nets (Table

3). However, the proportion of lakes where at least 125 stock-length bluegill were sampled was greatest with fyke nets in the fall (83.3%). The proportion of lakes where 50 stock-length black crappie were sampled was also greatest with fall fyke nets (66.7%); however, spring fyke nets maximized the proportion of lakes where 125 stock-length black crappie were sampled (33.3%).

Fifty stock-length individuals were not sampled in a single lake for any of the season and gear combinations for channel catfish and white crappie (Table 3). Mean bluegill PSD from fyke nets was lowest in fall (36 ± 11, [overall mean ± SE]), but similar for bluegills sampled in the spring

(56 ± 9) and summer (56 ± 13; Table 4). Mean PSD for bluegill sampled with night electrofishing was also lowest in the fall (39 ± 12). Black crappie mean PSD was highest in the spring for fyke net samples (75 ± 21). Mean largemouth bass PSD decreased from spring through fall with night electrofishing. Black bullhead and yellow bullhead mean PSDs were consistent among all seasons using fyke nets, while mean PSD of white bass was similar between spring and summer, but lower in the fall for fyke net samples. Mean PSD for yellow perch and walleye sampled with fyke nets were lower in fall than summer. Mean PSD for both species was greater for fall gill nets than fall fyke nets.

Gill nets minimized the mean number of samples needed to sample 125 stock-length individuals for walleye (147 ± 58 net nights), yellow perch (34 ± 3), channel catfish (150 ± 96), freshwater drum (87 ± 55), and white bass (29 ± 5; Figure 3). The mean number of samples needed to sample 125 stock-length individuals was minimized with fyke nets in the spring for common carp (83 ± 37 net nights), in the summer for bluegill (10 ± 5) and black bullhead (14 ±

5), and in the fall for black crappie (45 ± 22). Spring night electrofishing minimized the mean number of samples required to sample 125 stock-length largemouth bass (39 ± 5 five minute electrofishing runs) followed closely by spring day electrofishing (47 ± 9) and fall night electrofishing (54 ± 12).

DISCUSSION

Several well-known patterns of sampling bias were observed in our study. For instance, species and size selectivities for an individual gear were detected for many of the species.

Largemouth bass were almost exclusively sampled with electrofishing (i.e., stock-length) and seining (i.e., substock-length). Selectivity for adult and juvenile largemouth bass with electrofishing and seining, respectively, has long been known and these two methods are regularly recommended (e.g., Swingle 1956; Jackson and Noble 1995). In contrast, the use of multiple gears are commonly recommended for targeting bluegill (Bettross and Willis 1988), black crappie (Sammons et al. 2002), and yellow perch (Robillard et al. 1995). Our results showed that specific gears and seasons consistently resulted in higher mean CPUE for stock- or substock-length individuals for the 12 focal species evaluated. Specifically, modified-fyke nets are frequently used to target structure-oriented species and were particularly effective at sampling black crappie, white crappie, bluegill, walleye, common carp, black bullhead, yellow bullhead, and white bass. Gill nets are a commonly used passive sampling technique that target pelagic and highly mobile species (Hubert 1996). In our study, gill nets generally resulted in higher mean CPUE of stock-length walleye, yellow perch, common carp, channel catfish, freshwater drum, and white bass. Electrofishing is used to target a variety of freshwater fish species and was effective at sampling the majority of species evaluated in our study. As such, our results provide fisheries scientists with relative comparisons of sampling data for several recreationally- and ecologically-important species collected using a variety of techniques across seasons.

The use of multiple sampling methods to quantify abundance of adults and juveniles of the same species is common (e.g., Boxrucker et al. 1995; Bonvechio et al. 2008). Seining is

generally considered an effective method for sampling littoral areas of lakes, as it can be

conducted with minimal personnel and equipment needs, and it provides quantitative estimates

of fish abundance. However, capture efficiencies of beach seining are often highly variable due

to snags (e.g., woody debris, boulders) and high macrophyte densities (Pierce et al. 1990).

Miranda and Boxrucker (2009) did not recommend gears that directly targeted small-bodied or

age-0 fish in large standing water bodies, but Pope et al. (2009) recommended summer seining for small-bodied species (e.g., minnows) and age-0 fish in small-standing water bodies. We attempted to evaluate seasonal influences on estimates of seining, while providing similar comparisons for alternative methods (i.e., mini-modified-fyke nets and a benthic trawl). Mini-

fyke nets constructed with small mesh (e.g., ≤ 6 mm) have been commonly used to sample age-0

sport fish and small-bodied species (Weaver et al. 1993; Fago 1998; Ruetz et al. 2007). Small-

mesh fyke nets often sample more individuals and species than seining (Gritters 1994; Clark et

al. 2007). In our evaluation, mean CPUE of substock-individuals of bluegill, yellow perch, and

freshwater drum was highest with mini-fyke nets. An additional sampling method that has proved productive in sampling previously undocumented small-bodied species is the mini-

Missouri trawl (Herzog et al. 2005; Herzog et al. 2009). Although trawling is most commonly

employed in lotic ecosystems, most trawls require a powerful vessel with multiple personnel to

operate safely (Hayes et al. 1996) and was not recommended by Pope et al. (2009) or Miranda

and Boxrucker (2009) to sample standing water bodies. We used a small benthic trawl that could

sample littoral areas and provide comparable information to other littoral fish sampling methods.

The trawl used in our study was particularly effective at sampling substock-length black crappie

and yellow perch during summer. While seining was effective at sampling largemouth bass

(summer) and bluegill (fall), it was substantially more variable (i.e., over 50% of hauls yielded

53 zero fish) compared to the trawl and mini-fyke net. Furthermore, the trawl could be used to sample lentic ecosystems without multiple trips (i.e., deployment and retrieval) that are required for passive gears set overnight.

Understanding seasonal variation of sampling data is important to the interpretation of fish population and assemblage estimates by fisheries managers. The disproportionate contribution of older individuals due to sampling bias can make interpretation of dynamic rate functions (e.g., recruitment and mortality) difficult or impossible (Ricker 1969). For example,

Cross et al. (1995) observed increased CPUE of age-2 bluegill and decreased CPUE of age-4 and older bluegill as the year progressed due to differences in reproductive behavior. Seasonal differences in sampling bias generally result in peaks in the spring and fall, and are often attributed to a variety of factors, including changes in physical conditions (e.g., water temperature, macrophyte growth), fish behavior (e.g., spawning, changes in prey resources), and recruitment of smaller individuals to the sampling gear as a result of somatic growth (Pope and

Willis 1996). Bimodal patterns in abundance and size structure of fish have long been observed for freshwater species sampled seasonally with a single gear (Kelly 1953; Congdon 1968). We observed peaks in mean CPUE in the spring and fall for several of the species evaluated. For instance, largemouth bass (stock-length), walleye (stock- and substock-length), yellow perch

(stock- and substock-length), black bullhead (stock- and substock-lenght), and white bass (stock- length) exhibited lower catch rates in the summer with one or more gears. Like CPUE, size structure commonly peaks in the spring and fall (Pope and Willis 1996) and bimodal patterns in

PSD have long been observed in fish population assessments for largemouth bass (Carline et al.

1984; Gilliland 1985; Bettross and Willis 1988), bluegill (Bettross and Willis 1988), yellow perch (Lott and Willis 1991), and walleyes (Mero and Willis 1991). Peaks in PSD during the

spring and fall were only observed for black crappie sampled with fyke nets. Mean PSD for

bluegill was similar for spring and summer, but was lowest in the fall for both electrofishing and

fyke net samples. Mean PSD of largemouth bass also decreased from spring through fall with

night electrofishing. Therefore, sampling in the spring and fall can often increase catch rates, but

potential bias (e.g., size and age structure) associated with sample timing should be considered to ensure representativeness of the targeted populations. Sampling methods that reduce biases caused by differing capture efficiencies for various ages and sizes are desirable when selecting methods to monitor fish populations.

In addition to differences associated with sampling at various times of the year, diel period can influence fish population and assemblage assessments because of changes in habitat use and potential gear avoidance during the day. Numerous studies have evaluated the influence of diel period on littoral fish assemblage assessments with electrofishing (e.g., Paragamian 1989;

Pierce et al. 2001) and seining (Pierce et al. 2001; Riha et al. 2011). Mean CPUE in our study was consistently higher for night electrofishing compared to day electrofishing for stock- and substock-length individuals. However, this may have been due to the majority of the lakes sampled having high water clarity (i.e., > 1.5 m seechi depth; Table 1). McInerny and Cross

(1996) observed increased CPUE of largemouth bass greater than 200 mm TL with night

electrofishing compared to day electrofishing when Secchi depths were greater than 2 m.

Therefore, estimates of size structure from samples conducted at different times may be biased

and not accurately represent the population. Although we were limited in our ability to compare

day and night electrofishing estimates of mean PSD, our results suggested that mean PSD of

bluegill collected with day and night electrofishing were similar for spring and summer samples.

In addition to size selectivity, diel period can also influence species selectivity and capture

55 efficiency. Pierce et al. (2001) observed greater CPUE and species richness with night electrofishing and seining compared to similar methods during the day. While seining at night may have been more productive than during the day, we only seined during the day due to logistical constraints and safety concerns. Generally, increases in abundance and species richness observed from night sampling are attributed to the increased use of littoral habitats that make fish more susceptible to sampling (Pierce et al. 2001; Riha et al. 2011). Not only can diel period have implications for sampling with a single gear, but it may also influence comparisons of different sampling methods. For example, Gritters (1994) attributed the increased number of species and individuals sampled with mini-fyke nets relative to seining conducted during the day to differences in behavior. Therefore, passive gears that are set overnight (e.g., fyke nets, gill nets) may be more effective than active methods at sampling nocturnal species.

Although comparisons of multiple sampling methods are often difficult due to differences in the unit (e.g., number per net-night, number per hour of effort), our estimation of the sample size required to sample 125 stock-length individuals provides a useful comparison of sampling methods across seasons. Fisheries managers and researchers are often interested in describing populations with estimates of mean length or indices of size structure. However, the number of samples needed to adequately estimate length distributions of fish populations is rarely quantified (e.g., Miranda 1993; Vokoun et al. 2001; Miranda et al. 2007). While the estimated number of samples needed to collect 125 stock-length individuals in our study are useful, their interpretation should be applied with caution. For example, the effort to sample 125 stock-length bluegill or largemouth bass would be minimized in the spring with fyke nets and night electrofishing, respectively. However, estimates of size structure for bluegill and largemouth bass sampled during the spring or early summer would likely overestimate true size structure due

to the disproportionate sampling of spawning individuals (Novinger and Legler 1978).

McInerny and Cross (1996) observed selectivity for largemouth bass greater than 200 mm TL

with spring night electrofishing and increased CPUE of bluegill occurred when gonad

development was greatest (Cross et al. 1995). Despite seasonal biases associated with spawning behavior, the mean number of samples required to sample 125 stock-length individuals was minimized for 5 of the 12 species evaluated with fall gill nets. Although gill nets were constructed of multiple sizes of mesh, gill nets generally sampled larger individuals than fyke nets used in the same season. As such, mean PSD was higher with gill nets relative to fyke nets for white bass, yellow perch, and walleye. Differences in mean PSD between gears may have also been due to differences in habitat use by smaller individuals, because gill nets were set in deeper habitats (i.e., > 2 m of water) to ensure that the net was fully deployed. Therefore, sampling regimes that attempt to assess the relative abundance and size structure of multiple species in lake and reservoir ecosystems will likely require a combination of gears and seasons.

Management implications

Given the guidelines recommended by Bonar et al. (2009) for sampling lentic freshwater ecosystems, it is important for fisheries managers to understand how interpretations of assessments may be influenced by different sampling methods used at various times of the year.

For example, Miranda and Boxrucker (2009) recommended fall fyke netting for large standing waters (> 200 ha), while Pope et al. (2009) recommended spring fyke netting for small standing waters (< 200 ha). Although there may be justifiable and logical reasons to sample different sizes of water bodies in separate seasons, data collected with similar gears in different seasons may not be comparable. Therefore, our study provides a relative framework for comparing data

57 collected with various sampling methods across seasons for multiple species that have recreational and ecological importance. Ultimately, sampling methods used to assess freshwater fisheries will be the compromise between objectives (e.g., maximize species richness, minimize variation), logistical constraints (e.g., cost, labor), and tradition.

ACKNOWLEDGEMENTS

We thank M. Mork, S. Bisping, T. Neebling, J. Koch, W. Massure, A. Schaefer, R.

Schultz, A. Fowler, and J. Christiansen for assistance with field collection. We also thank P.

Dixon, J. Downing, C. Paukert, C. Pierce, and K. Roe for helpful comments on a previous version of this manuscript. This project was supported, in part, by the Department of Natural

Resource Ecology and Management Iowa State University, Iowa Department of Natural

Resources, and Idaho Cooperative Fish and Wildlife Research Unit. The Unit is jointly sponsored by the University of Idaho, US Geological Survey, Idaho Department of Fish and

Game, and Wildlife Management Institute. The use of trade, firm, or product names is for descriptive purposes only and does not imply endorsement by the US Government. This study was performed under the auspices of Iowa State University protocol #1-08-6493-I.

REFERENCES

Allen, M. S., M. M. Hale, and W. E. Pine III. 1999. Comparison of trap nets and otter trawls for

sampling black crappie in two Florida lakes. North American Journal of Fisheries

Management 19:977–983.

Anderson, R. O., and R. M. Neumann. 1996. Length, weight, and associated structural indices.

Pages 447–482 in B. R. Murphy and D. W. Willis, editors. Fisheries techniques, 2nd

edition. American Fisheries Society, Bethesda, Maryland.

Barko, V. A., D. P. Herzog, R. A. Hrabik, and J. S. Scheibe. 2004. Relationship among fish

assemblages and main-channel-border physical habitats in the unimpounded upper

Mississippi River. Transactions of the American Fisheries Society 133:371–384.

Bettross, E. A., and D. W. Willis. 1988. Seasonal patterns in sampling data for largemouth bass

and bluegills in a northern Great Plains impoundment. Prairie Naturalist 20:193–202.

Bister, T. J., D. W. Willis, M. L. Brown, S. M. Jordan, R. M. Neumann, M. C. Quist, and C. S.

Guy. 2000. Proposed standard weight (Ws) equations and standard length categories for

18 warmwater nongame and riverine fish species. North American Journal of Fisheries

Management 20:570–574.

Bonar S. A., W. A. Hubert, and D. W. Willis. 2009. Standard sampling methods for North

American freshwater fishes. American Fisheries Society, Bethesda, Maryland.

Bonvechio, T. F., W. F. Pouder, and M. M. Hale. 2008. Variation between electrofishing and

otter trawling for sampling black crappies in two Florida lakes. North American Journal

of Fisheries Management 28:188–192.

Boxrucker, J., P. Michaletz, M. J. Van Den Avyle, and B. Vondracek. 1995. Overview of gear

evaluation study for sampling gizzard shad and threadfin shad populations in reservoirs.

North American Journal of Fisheries Management 15:885–890.

Burkhardt, R. W., and S. Gutreuter. 1995. Improving electrofishing consistency by

standardizing power. North American Journal of Fisheries Management 15:375–381.

Carline, R. F., B. L. Johnson, and T. J. Hall. 1984. Estimation and interpretation of proportional

stock density for fish populations in Ohio impoundments. North American Journal of

Fisheries Management 4:139–154.

Clark, S. J., J. R. Jackson, and S.E. Lochmann. 2007. A comparison of shoreline seines with

fyke nets for sampling littoral fish communities in floodplain lakes. North American

Journal of Fisheries Management 27:676–680.

Congdon, J. C. 1968. The fish population of Lake Poinsett, South Dakota, as indicated by four

types of gear. M.S. thesis, South Dakota State University, Brookings.

Cross, T. K., M.C. McInerny, and D.H. Schupp. 1995. Seasonal variation in trap-net catches of

bluegill in Minnesota lakes. North American Journal of Fisheries Management 15:382–

389.

Fago, D. 1998. Comparison of littoral fish assemblages sampled with a mini-fyke net or a

combination of electrofishing and small-mesh seine in Wisconsin lakes. North American

Journal of Fisheries Management 18:731–738.

Gabelhouse, D. W. 1984. A length-categorization system to assess fish stocks. North American

Journal of Fisheries Management 4:273–285.

Gido, K. B., and W. J. Matthews. 2000. Dynamics of the offshore fish assemblage in a

southwestern reservoir (Lake Texoma, Oklahoma-Texas). Copeia 2000:917–930.

Gilliland, E. 1985. Evaluation of Oklahoma’s standardized electrofishing in calculating

population structure indices. Proceedings of the Annual Conference of Southeastern

Associations of Fish and Wildlife Agencies 39:277–287.

Gritters, S. A. 1994. Comparison of fish catch between a mini fyke net and a 10.7-meter bag

seine in the upper Mississippi River. National Biological Survey, Environmental

Management Technical Center, Special Report 94-S008, Onalaska, Wisconsin.

Guy, C. S., D. W. Willis, and R. D. Schultz. 1996. Comparison of catch per unit effort and size

structure of white crappies collected with trap nets and gill nets. North American Journal

of Fisheries Management 16:947–951.

Guy, C. S., R. M. Neumann, D. W. Willis, and R. O. Anderson. 2007. Proportional size

distribution (PSD): A further refinement of population size structure index terminology.

Fisheries 32:348.

Guy, C. S., P. J. Braaten, D. P. Herzog, J. Pitlo, and R. S. Rogers. 2009. Warmwater fish in

rivers. Pages 59–84 in S. Bonar, W. Hubert, and D. Willis, editors. Standard sampling

methods for North American freshwater fishes. American Fisheries Society, Bethesda,

Maryland.

Hayes D. B., C. P. Ferreri, and W. W. Taylor. 1996. Active fish capture methods. Pages 193–

220 in B. R. Murphy and D. W. Willis, editors. Fisheries techniques, 2nd edition.

American Fisheries Society, Bethesda, Maryland.

Herzog, D. P., V. A. Barko, J. S. Scheibe, R. A. Hrabik, and D. E. Ostendorf. 2005. Efficacy of

a benthic trawl for sampling small-bodied fishes in large river systems. North American

Journal of Fisheries Management 25:594–603.

Herzog, D. P., D. E. Ostendorf, R. A. Hrabik, and V. A. Barko. 2009. The mini-Missouri trawl:

a useful methodology for sampling small-bodied fishes in small and large river systems.

Journal of Freshwater Ecology 24:103–108.

Hubert, W. A. 1996. Passive capture techniques. Pages 157–192 in B. R. Murphy and D. W.

Willis, editors. Fisheries techniques, second edition. American Fisheries Society,

Bethesda, Maryland.

Jackson, J. R., and R. L. Noble. 1995. Selectivity of sampling methods for juvenile largemouth

bass in assessments of recruitment processes. North American Journal of Fisheries

Management 15:408–418.

Jordon, G. R., and D. W. Willis. 2001. Seasonal variation in sampling indices for shovelnose

sturgeon, river carpsucker, and shorthead redhorse collected from the Missouri River

below Fort Randall Dam, South Dakota. Journal of Freshwater Ecology 16:331–340.

Kelly, D. W. 1953. Fluctuation of trap net catches in the upper Mississippi River. U.S. Fish and

Wildlife Service, Special Scientific Report, Fisheries Number 101.

Lee, D. S., C. R. Gilbert, C. H. Hocutt, R. E. Jenkins, D. E. McAllister, and J. R. Stauffer Jr.

1980. Atlas of North American Fishes. North Carolina Biological Survey, Raleigh.

Lott, J. P., and D. W. Willis. 1991. Gill net mesh size efficiency for yellow perch. Prairie

Naturalist 23:139–144.

Mero, S. W., and D. W. Willis. 1992. Seasonal variation in sampling data for walleye and

sauger collected with gill nets form Lake Sakakawea, North Dakota. Prairie Naturalist

24:231–240.

McInerny, M. C., and T. K. Cross. 1996. Seasonal and diel variation in electrofishing size-

selectivity and catch-per-hour of largemouth bass in Minnesota lakes. Investigational

report 451. Minnesota Department of Natural Resources, Hutchinson.

McInerny, M. C., and T. K. Cross. 2004. Comparison of day electrofishing, night

electrofishing, and trap netting for sampling inshore fish in Minnesota Lakes. Special

publication 161. Minnesota Department of Natural Resources, Hutchinson.

Miranda, L. E. 1993. Sample sizes for estimating and comparing proportion-based indices.

North American Journal of Fisheries Management 13:383–386.

Miranda, L. E. 2007. Approximate sample sizes required to estimate length distributions.

Transactions of the American Fisheries Society 136:409–415.

Miranda, L. E. 2009. Standardizing electrofishing power for boat electrofishing. Pages 223–

230 in S. Bonar, W. Hubert, and D. Willis, editors. Standard sampling methods for North

American freshwater fishes. American Fisheries Society, Bethesda, Maryland.

Miranda, L. E., and J. Boxrucker. 2009. Warmwater fish in large standing waters. Pages 29–42

in S. Bonar, W. Hubert, and D. Willis, editors. Standard sampling methods for North

American freshwater fishes. American Fisheries Society, Bethesda, Maryland.

Novinger, G. D., and R. E. Legler. 1978. Bluegill population structure and dynamics. Pages 37-

49 in G. D. Novinger and J. G. Dillard, editors. New approaches to the management of

small impoundments. American Fisheries Society, North Central Division, Special

Publication 5, Bethesda, Maryland.

Paragamian, V. L. 1989. A comparison of day and night electrofishing: size structure and catch

per unit effort for smallmouth bass. North American Journal of Fisheries Management

9:500–503.

Peterson, J. T., C. P. Paukert. 2009. Converting nonstandard fish sampling data to standardized

data. Pages 195–216 in S. Bonar, W. Hubert, and D. Willis, editors. Standard sampling

methods for North American freshwater fishes. American Fisheries Society, Bethesda,

Maryland.

Pierce C. L., J. B. Rasmussen, W. C. Leggett. 1990. Sampling littoral fish with a seine:

corrections for variable capture efficiency. Canadian Journal of Fisheries and Aquatic

Sciences 47:1004–1010.

Pierce, C. L., A. M. Corcoran, A. N. Gronbach, S. Hsia, B. J. Mularkey, and A. J. Schwartzhoff.

2001. Influence of diel period on electrofishing and beach seining assessments of littoral

fish assemblages. North American Journal of Fisheries Management 21:918–926.

Pope, K. L., and D. W. Willis. 1996. Seasonal influences of freshwater fisheries sampling data.

Reviews in Fisheries Science 4:57–73.

Pope, K. L., R. M. Neumann, and S. D. Bryan. 2009. Warmwater fish in small standing waters.

Pages 13–28 in S. Bonar, W. Hubert, and D. Willis, editors. Standard sampling methods

for North American freshwater fishes. American Fisheries Society, Bethesda, Maryland.

Quist M. C., R. J. Bernot, C. S. Guy, and J. L. Stephen. 2001. Seasonal variation in population

characteristics of gizzard shad. Journal of Freshwater Ecology 16:641–646.

Quist, M. C., K. I. Bonvechio, and M. S. Allen. 2009. Statistical analysis and data management.

Pages 171–194 in S. Bonar, W. Hubert, and D Willis, editors. Standard sampling

methods for North American freshwater fishes. American Fisheries Society, Bethesda,

Maryland.

Rabeni, C., J. Lyons, N. Mercado-Silva, and J. Peterson. 2009. Warmwater fish in wadeable

streams. Pages 43–58 in S. Bonar, W. Hubert, and D Willis, editors. Standard sampling

methods for North American freshwater fishes. American Fisheries Society, Bethesda,

Maryland.

Reynolds, J. B. 1996. Electrofishing. Pages 221–253 in B. R. Murphy and D. W. Willis,

editors. Fisheries techniques, second edition. American Fisheries Society, Bethesda,

Maryland.

Ricker, W. E. 1969. Effects of size-selective mortality and sampling bias on estimates of

growth, mortality, production, and yield. Journal of the Fisheries Research Board of

Canada 26:479–541.

Riha, M., J. Kubecka, M. Prchalova, T. Mrkvicka, M. Cech, V. Drastik, J. Frouzova, E.

Hohausova, T. Juza, M. Kratochvil, J. Peterka, M. Tuser, and M. Vasek. 2011. The

influence of diel period on fish assemblage in the unstructured littoral of reservoirs.

Fisheries Management and Ecology 18:339–347.

Robillard, S. R., T. Kassler, and J. E. Marsden. 1995. Yellow perch population assessment in

southwestern Lake Michigan, including evaluation of sampling techniques. Aquatic

Ecology Technical Report 95/9. Illinois Natural History Survey, Champaign.

Ruetz C. R., D. G. Uzarski, D. M. Krueger, and E. S. Rutherford. 2007. Sampling a littoral fish

assemblage: comparing small-mesh fyke netting and boat electrofishing. North

American Journal of Fisheries Management 27:825–831.

Sammons, S. M., D. A. Isermann, and P. W. Bettoli. 2002. Variation in population

characteristics and gear selection between black and white crappies in Tennessee

reservoirs: Potential effects on management decisions. North American Journal of

Fisheries Management 22:863–869.

Sanders, R. E. 1992. Day versus night electrofishing catches from near-shore waters of the Ohio

and Muskingum Rivers. Ohio Journal of Science 92(3):51–59.

Swingle, H. S. 1956. Appraisal of methods of fish population study, part 4: determination of

balance in farm fish ponds. Transactions of the North American Wildlife Conference

21:298–322.

Thurow, R. F., and D. J. Schill. 1996. Comparison of day snorkeling, night snorkeling, and

electrofishing to estimate bull trout abundance and size structure in a second order Idaho

stream. North American Journal of Fisheries Management 16:314–323.

Vokoun, J. C., C. F. Rabeni, and J. S. Stanovick. 2001. Sample-size requirements for evaluating

population size-structure. North American Journal of Fisheries Management 21:660–

665.

Weaver, M. J., J. J. Magnuson, and M. K. Clayton. 1993. Analyses for differentiating littoral

fish assemblages with catch data from multiple sampling gears. Transactions of the

North American Fisheries Society 122:1111–1119.

Willis, D. W., and B. R. Murphy. 1996. Planning for sampling. Pages 1–15 in B. R. Murphy

and D. W. Willis, editors. Fisheries techniques, 2nd edition. American Fisheries Society,

Bethesda, Maryland.

Willis, D. W., B. R. Murphy, and C. S. Guy. 1993. Stock density indices: development, use,

and limitations. Reviews in Fisheries Science 1:203–222.

Yoder, C. O., and M. A. Smith. 1998. Using fish assemblages in a state biological assessment

and criteria program: essential concepts and considerations. Pages 17–56 in T. P. Simon,

editor. Assessing the sustainability and biological integrity of water resources using fish

communities. CRC Press, Boca Raton, Florida.

TABLE 1.—Water body, type, area, mean depth (Z), sampling effort (number of net nights for passive gears, number of seine hauls, number of 3-minute trawls, and number of 5-minute electrofishing runs conducted within a season), mean Secchi depth, mean chlorophyll a (Chl-a) concentration, mean total phosphorus (TP) concentrations. Area Sampling Secchi Chl-a TP Water body Type (ha) Z (m) effort depth (m) (µg/L) (µg/L) Prairie Rose Impoundment 70 2.7 10 0.7 48 91 Don Williams Impoundment 60 5.5 10 1.7 24 84 Pleasant Creek Impoundment 162 4.8 12 2.1 15 40 Silver Natural lake 432 2.3 20 0.7 37 167 Minnewashta Natural lake 48 3.1 10 1.7 22 116 West Okoboji Natural lake 1,557 11.6 20 5.4 4 27

TABLE 2.—Total number of individuals sampled in spring (Sp), summer (Su), and fall (Fa) for seven sampling methods from six lakes and impoundments in Iowa, 2008. Gill Seine Trawl Day EF Night EF Fyke net Mini-fyke net net

Family and species Sp Su Fa Sp Su Fa Sp Su Fa Sp Su Fa Sp Su Fa Sp Su Fa Fa Total Cyprinidae Common carp Cyprinus carpio 1 3 4 4 1 50 44 30 34 29 42 166 157 118 1 1 2 118 805 Ictaluridae

Black bullhead Ameiurus melas 1 1 3 8 5 8 1 16 8 85 40 17 1,423 655 553 182 91 36 173 3,306

Yellow bullhead Ameiurus natalis 70 1 1 1 1 1 18 10 15 282 139 169 21 1 9 22 761

Channel catfish Ictalurus punctatus 1 2 13 1 9 12 9 12 14 21 6 78 12 2 1 188 381 Moronidae

White bass Morone chrysops 2 7 2 7 47 17 68 99 23 406 405 231 552 13 8 50 446 2,383 Centrarchidae

Bluegill Lepomis macrochirus 50 142 2,744 210 667 1,815 199 162 414 1,643 716 869 1,947 5,502 2,338 146 418 7,929 29 27,940

Smallmouth bass Micropterus dolomieu 2 1 20 17 12 22 33 11 24 1 1 11 155

Largemouth bass Micropterus salmoides 6 687 57 2 48 10 192 99 249 277 172 226 9 7 27 4 3 4 6 2,085

White crappie Pomoxis annularis 6 4 15 2 5 3 21 9 18 5 9 97

Black crappie Pomoxis nigromaculatus 256 36 139 2,991 616 36 16 49 172 21 49 769 1,799 1,873 15 34 201 236 9,308 Percidae

Yellow perch Perca flavescens 196 5 19 159 18 11 49 73 163 63 134 100 176 165 1 10 46 225 1,613 Walleye Sander vitreus 4 4 4 39 8 3 12 27 119 54 132 65 112 145 3 11 12 152 906 Sciaenidae

Freshwater drum Aplodinotus grunniens 1 58 228 22 25 25 35 37 65 154 10 84 3,868 6 1,134 269 6,021

Total 61 1,289 2,994 394 4,175 2,522 593 465 990 2,692 1,218 2,092 5,203 8,950 9,838 386 586 9,429 1,884 55,761

TABLE 3.—Proportion of lakes and impoundments in Iowa where 50 and 125 (parentheses) stock-length individuals were sampled with seine (S), trawl (T), day electrofishing (D), night electrofishing (N), fyke net (F), mini-fyke net (M), and gill net (G) in spring, summer, and fall 2008. Spring Summer Fall Family and species S T D N F M S T D N F M S T D N F M G Cyprinidae 16.7 16.7 16.7 Common carp x x x x x x x x x x x x x x x x Cyprinus carpio (0) (0) (0)

Ictaluridae 25.0 50.0 25.0 75.0 25.0 50.0 25.0 Black bullhead x x x x x x x x x x x x Ameiurus melas (0) (25.0) (25.0) (25.0) (0) (25.0) (25.0)

Yellow bullhead 20.0 20.0 20.0 x x x x x x x x x x x x x x x x Ameiurus natalis (20.0) (0) (20.0) Channel catfish x x x x x x x x x x x x x x x x x x x Ictalurus punctatus

Moronidae 40.0 40.0 20.0 40.0 60.0 White bass x x x x x x x x x x x x x x Morone chrysops (20.0) (0) (0) (40.0) (20.0)

Centrarchidae 33.0 83.3 83.3 83.3 100 16.7 16.7 33.3 66.7 83.3 Bluegill x x x x x x x x x Lepomis macrochirus (0) (50.0) (66.7) (16.7) (66.7) (0) (0) (0) (16.7) (83.3) 16.7 16.7 33.3 16.7 Largemouth bass x x x x x x x x x x x x x x x Micropterus salmoides (0) (0) (0) (0)

White crappie x x x x x x x x x x x x x x x x x x x Pomoxis annularis 50.0 16.7 66.7 16.7 Black crappie x x x x x x x x x x x x x x x Pomoxis nigromaculatus (33.3) (16.7) (16.7) (16.7)

Percidae 20.0 20.0 20.0 20.0 Yellow perch x x x x x x x x x x x x x x x Perca flavescens (0) (0) (0) (0) 16.7 16.7 16.7 Walleye x x x x x x x x x x x x x x x x Sander vitreus (0) (0) (0)

Sciaenidae 25.0 Freshwater drum x x x x x x x x x x x x x x x x x x Aplodinotus grunniens (25.0)

TABLE 4.—Mean proportional size distribution (PSD) and standard error (parentheses) of 12 fish species sampled with seine (S), trawl (T), day electrofishing (D), night electrofishing (N), fyke net (F), mini-fyke net (M), and gill net (G) in spring, summer, and fall from six lake and impoundments in Iowa, 2008. Mean PSD was only calculated for seasons, sampling methods, and lakes with greater 50 stock-length individuals sampled. Spring Summer Fall Family and species S T D N F M S T D N F M S T D N F M G Cyprinidae 100 98.2 41.4 Common carp x x x x x x x x x x x x x x x x Cyprinus carpio (0) (0) (0)

Ictaluridae 100 93.0 100 95.7 100 97.2 99.3 Black bullhead x x x x x x x x x x x x Ameiurus melas (0) (6.9) (0) (2.6) (0) (2.6) (0)

Yellow bullhead 99.2 98.0 98.1 x x x x x x x x x x x x x x x x Ameiurus natalis (0) (0) (0) Channel catfish x x x x x x x x x x x x x x x x x x x Ictalurus punctatus

Moronidae 99.4 98.3 29.3 71.2 81.8 White bass x x x x x x x x x x x x x x Morone chrysops (0.6) (0.8) (0) (28.8) (18.2)

Centrarchidae 45.6 45.2 55.9 52.7 55.9 50.7 17.7 45.6 38.9 35.9 Bluegill x x x x x x x x x Lepomis macrochirus (28.9) (12.4) (9.4) (13.2) (12.5) (0) (0) (23.2) (12.0) (11.4) 84.6 66.7 49.5 54.9 Largemouth bass x x x x x x x x x x x x x x x Micropterus salmoides (0) (0) (10.2) (0)

White crappie x x x x x x x x x x x x x x x x x x x Pomoxis annularis 75.4 12.6 50.4 16.5 Black crappie x x x x x x x x x x x x x x x Pomoxis nigromaculatus (21.3) (0) (13.4) (0)

Percidae 74.1 28.8 37.9 73.6 Yellow perch x x x x x x x x x x x x x x x Perca flavescens (0) (0) (0) (26.4) 35.2 24.6 43.2 Walleye x x x x x x x x x x x x x x x x Sander vitreus (0) (0) (0)

Sciaenidae 99.3 Freshwater drum x x x x x x x x x x x x x x x x x x Aplodinotus grunniens (25.0)

80 Black crappie 60 Spring Summer 40 Fall

0 2.0 White crappie 1.5

1.0

0.5

0.0 250 Bluegill 200 150 100 50 25 0 25 Largemouth bass 20 15

Substock CPUE 10 5 0 2.0 Walleye 1.5

1.0

0.5

0.0 8 Yellow perch 6

Seine Trawl Fyke net Gill net Daytime EF Nighttime EF Mini-fyke net FIGURE 1.—Mean catch-per-unit effort (CPUE) of substock-length individuals from seasonal sampling with seven methods in six lakes and impoundments in Iowa, 2008. Values for CPUE were calculated for independently for seine (number per haul); trawl (number per 3-minute trawl); day and night electrofishing (EF; number per hour); and fyke net, mini-fyke net, and gill net (number per net night). Error bars represent one SE.

0.5 Common carp 0.4 Spring Summer 0.3 Fall 0.2 0.1 0.0 3.0 Channel catfish 2.5 2.0 1.5 1.0 0.5 0.0 1.0 Black bullhead 0.8 0.6 0.4 0.2 0.0 4 Yellow bullhead 3

Substock CPUE 2

0 140 Freshwater drum 120 100 80 60 40 20 0 8 White bass 6

Seine Trawl Fyke net Gill net Daytime EF Nighttime EF Mini-fyke net FIGURE 1.—Continued.

60 Black crappie 50 Spring 40 Summer 30 Fall 20 10 0 3.0 White crappie 2.5 2.0 1.5 1.0 0.5 0.0 120 Bluegill 100 80 60 40 20 0 60 Largemouth bass 50

Stock CPUE 40 30 20 10 0 3.0 Walleye 2.5 2.0 1.5 1.0 0.5 0.0 8 Yellow perch 6

Seine Trawl Fyke net Gill net Daytime EF Nighttime EF Mini-fyke net FIGURE 2.—Mean catch-per-unit effort (CPUE) of stock-length individuals from seasonal sampling with seven methods in six lakes and impoundments in Iowa, 2008. Values for CPUE were calculated for independently for seine (number per haul); trawl (number per 3-minute trawl); day and night electrofishing (EF; number per hour); and fyke net, mini-fyke net, and gill net (number per net night). Error bars represent one SE.

3.0 Common carp 2.5 Spring 2.0 Summer 1.5 Fall 1.0 0.5 0.0 3.0 Channel catfish 2.5 2.0 1.5 1.0 0.5 0.0 40 Black bullhead 30

0 20 Yellow bullhead 15 Stock CPUE

0 6 Freshwater drum 5 4 3 2 1 0 10 White bass 8 6 4 2 0

Seine Trawl Fyke net Gill net Daytime EF Nighttime EF Mini-fyke net

FIGURE 2.—Continued.

3000 Black crappie 2500 Spring 2000 Summer 1500 Fall 1000 500 0 1400 White crappie 1200 1000 800 600 400 200 0 1400 Bluegill 1200 1000 800 600 400 200 0 3000 Largemouth bass 2500 2000 1500

Number of samples 1000 500 0 3000 Walleye 2500 2000 1500 1000 500 0 4000 Yellow perch 3000

2000

1000

Seine Trawl Fyke net Gill net Daytime EF Nighttime EF Mini-fyke net FIGURE 3.—Mean minimum numbers of samples (i.e., five minute electrofishing runs, deployments, sets) necessary to sample 125 stock-length fish from seasonal sampling with seven methods in six lakes and impoundments in Iowa, 2008. Error bars represent one SE.

3000 Common carp 2500 Spring 2000 Summer 1500 Fall 1000 500 0 2000 Channel catfish 1500

1000

500

0 3000 Black bullhead 2500 2000 1500 1000 500 0 3000 Yellow bullhead 2500 2000 1500

Number of samples 1000 500 0 2000 Freshwater drum 1500

1000

500

0 3000 White bass 2500 2000 1500 1000 500 0

Seine Trawl Fyke net Gill net Daytime EF Nighttime EF Mini-fyke net FIGURE 3.—Continued.

CHAPTER 4. UNDERSTANDING FISH ASSEMBLAGE STRUCTURE IN LENTIC

ECOSYSTEMS: A COMPARISON OF NATURAL LAKES AND RESERVOIRS IN

IOWA, USA

A manuscript to be submitted for publication in Transactions of the American Fisheries Society

Jesse R. Fischer1 and Michael C. Quist2

Abstract

Reservoirs are often managed similarly to natural lakes because they are assumed to be

functionally comparable. However, direct comparisons of fish assemblage-environment

relationships between these two ecosystems are rare and this assumption has not been adequately

evaluated. We investigated associations of fish assemblage structure from 45 natural lakes and

reservoirs in Iowa, USA. Fish sampling was conducted with benthic trawls, modified-fyke nets,

and night electrofishing. Increased species diversity in reservoirs was most strongly related to

morphometric characteristics (i.e., larger surface area, increased depth); whereas, fewer species

were observed in natural lakes with low water clarity and high suspended solids. Fish

assemblage structure between natural lakes and reservoirs was consistently dissimilar for all

sampling methods. Structuring of species composition in reservoirs was correlated with a variety

of limnological and physical characteristics, but was largely dependent on the sampling method.

In contrast, trophic structure of fishes in reservoirs was weakly associated with the

environmental factors evaluated and was similar to fish species structure of natural lakes. Fish

trophic composition of natural lakes was related to waterbody size, but was consistent among

sampling methods. Overall, distinct differences in fish assemblage structure were observed

between natural and artificial lentic ecosystems. Our results emphasize the need to consider

waterbody origin (i.e., natural or artificial) on fish assemblage characterization and subsequent

inferences made from environmental correlations.

INTRODUCTION

Lentic fish assemblages have long been of interest to community ecologists, largely because they have ecosystem characteristics that are analogous to islands (e.g., discrete, isolated, repeated; Barbour and Brown 1974; Brown 1981). Consequently, numerous studies have evaluated the influence of physical factors (e.g., morphology, water chemistry, size, location) on fish assemblage structure in natural lakes (e.g., Magnuson et al. 1998; Tonn and Magnuson 1982;

Eadie et al. 1986; Jackson and Harvey 1993). However, the structure of fish assemblages in artificial lentic habitats (i.e., reservoirs) has received considerably less attention (Irz et al. 2002;

Irz et al. 2006; Miranda et al. 2008), despite their potential to support similar management strategies, monitoring programs, and fish faunas (e.g., Guy and Willis et al. 1995; Launois et al.

2011, Menezes et al. in press).

A poor understanding of fish assemblage structure in reservoirs is likely the result of their artificial and highly-managed nature and difficulties inherent in quantifying whole-lake fish assemblages. Sampling lentic fish assemblages often requires multiple methods. Consequently, most research in reservoirs has focused on sport fish population structure and dynamics rather than quantifying non-sport fish occurrences and abundances. Furthermore, sampling biases are

78 regularly associated with distinct physicochemical zones (i.e., pelagic, littoral) and habitat characteristics that can effect interpretation of environmental structuring (Menezes et al. in press). However, natural lakes and reservoirs are frequently managed similarly and monitored with many of the same fish sampling techniques (Miranda and Boxrucker 2009; Launois et al.

2011). Therefore, understanding the dynamics of fish assemblage structure across ecosystems may be beneficial to natural resource agencies charged with simultaneously managing fisheries and conserving biodiversity.

Distinct differences between natural lakes and reservoirs present inherent challenges, but also provide an opportunity to understand how lentic fish assemblages are structured by a wide range of both physical and chemical characteristics. For instance, structurally-complex habitats are often assumed to support a greater diversity of species and functional traits than simple habitats (Klopfer and MacArthur 1960). Species richness and habitat complexity have been observed to be positively associated for a variety of taxa and ecosystems (e.g., Karr 1968;

Murdoch et al. 1972; Kohn and Leviten 1976; Menge et al. 1985), including lotic environments

(e.g., Gorman and Karr 1978) and natural lakes (e.g., Eadie and Keast 1984; Tonn and

Magnuson). Reservoirs are unique because they are engineered, but also because they can exhibit characteristics of both natural lakes and riverine ecosystems. Therefore, increased biodiversity associated with increased habitat complexity might allow reservoirs to support a greater number of fish species than similar natural lakes. Miranda et al. (2008) observed that large reservoirs in the Tennessee River valley harbored up to 67 species and that species richness generally increased as riverine species were added to the assemblage downstream. However, physical characteristics (e.g., size, proportion of littoral habitat) were also strongly associated with position in the river (Miranda et al. 2008). Therefore, it is uncertain whether within-lake

79 habitat characteristics alone were responsible for the observed patterns in fish occurrence and abundance (Miranda et al. 2008). Fish assemblages of natural lentic systems are also structured by lake morphology and chemical characteristics (Tonn and Magnusson 1982; Eadie and Keast

1984; Jackson and Harvey 1993). Despite the potential for similar environmental regulation of biotic communities, comparative ecological studies of lake and reservoir fish assemblage structure and environmental influences are uncommon (e.g., Irz et al. 2006; Irz et al 2007).

Comparisons of reservoirs and natural lakes are also warranted because both ecosystems have the propensity to share similar fish faunas. This is, in part, the result of similar physical characteristics (e.g. lacustrine habitat) and due to the widespread introduction of lentic species in both systems. Sport fish stocking is common in both ecosystems to enhance fisheries, but unintentional introductions (i.e., invasive species) are more likely in lakes that are stocked more frequently (Radomski and Goerman 1995). Sport fish introductions in reservoirs are particularly concerning because they can result in losses of biodiversity and biotic homogenization of lotic fish assemblages (e.g., Clavero and Hermoso 2011). Therefore, understanding environmental relationships of fish assemblage structure in natural and artificial lentic habits is essential to the conservation and management of these similar systems (Launois et al. 2011).

Because previous research on lentic fish assemblage-environment relationships has primarily focused on either reservoirs (e.g. Miranda et al. 2008) or natural lakes (e.g., Tonn and

Magnuson 1982), the goal of our study was to examine the influence of physicochemical characteristics on fish assemblages in both lentic ecosystems. We chose to reservoirs and natural lakes within a spatially-limited extent (i.e., Iowa, USA) to increase understanding of factors regulating fish assemblage structure. The specific objectives of our study were to determine if broad physical and chemical features regulate fish assemblages in natural lakes and reservoirs

similarly and identify possible factors responsible for dissimilarities of fish assemblages between

artificial and natural ecosystems. We anticipate that greater habitat complexity and connectivity

to lotic environments in reservoirs will distinguish fish assemblages and their environmental

relationships from natural lakes. However, ubiquitous and recreationally-important species (i.e.,

stocked sport fish) will likely decrease dissimilarity of fish faunas between natural lakes and

reservoirs despite differences in habitat characteristics.

METHODS

Study area.—Iowa is located in the agriculturally-dominated Western Corn Belt Plains ecoregion of the Mississippi River basin. As such, standing water bodies in Iowa are among the most eutrophic environments in the world due to the predominance of row-crop agriculture across the landscape (Arbuckle and Downing 2001). The presence of both natural (i.e., glacially-formed lakes) and artificial (i.e., reservoirs) lentic ecosystems in Iowa provided an opportunity to understand fish assemblage structure dynamics from a limited geographic extent and similar source populations of native species. A total of 45 lakes and reservoirs across Iowa was selected to represent the range of physical and water quality conditions present in Iowa’s lentic systems

(Table 1).

Physicochemical characteristics.—Lakes and reservoirs were sampled three times yearly between May and August (2000-2010) for measures of Secchi depth (m), chlorophyll a (µg/L), total phosphorus (µg/L), total nitrogen (µg/L), and total suspended solids (mg/L). Water samples were taken from the epilimnion in the presence of a thermocline. When a thermocline was absent, the entire water column was sampled. Processing and analysis of water samples followed

standard methods detailed in Egertson and Downing (2004). Values from years prior to fish

sampling were averaged to broadly characterize the lentic environment for each waterbody.

Additionally, morphometric characteristics of the water bodies were estimated and included

surface area (ha), watershed-to-lake-area ratio, mean depth (m), and maximum depth (m). Water

quality parameters and physical characteristics were those identified as important in previous

studies of lentic fish assemblages (Jackson and Harvey 1993; Irz et al. 2007; Miranda et al.

2008). We also attempted to select variables that reduced the number of correlated factors.

Because we were interested in describing the influence of environmental characteristics for

different ecosystems that may be dissimilarly influenced, we included factors that were deemed

important despite high correlation (e.g., Spearman’s rank correlation; rs ≥ 0.70) with other factors (Table 2). For example, Secchi depth was strongly correlated with total phosphorus concentration (rs = 0.80; P < 0.0001), but both were included in our analysis due to the potential

for physicochemical factors to regulate fish assemblages in natural lakes and reservoirs

differently.

Fish assemblages.—Fish assemblages were sampled from 2008 to 2011 with both active and

passive methods that targeted a diversity of habitats. Sampling protocols were based on previous

evaluations of the gears and timing (e.g., season, diel period) that concurrently maximized the

number of species and total number of individuals sampled in Iowa lakes (Fischer, unpublished

information). Sampling of each waterbody was conducted within a calendar year to minimize

temporal variability, yet exploit dissimilarities in species represented in different seasons. Small-

bodied adults and juveniles of large-bodied species that commonly use littoral and benthic

habitats were targeted with a benthic trawl. The trawl had a headrope length of 2.4 m, footrope

82 length of 3.7 m, and upright height of 0.6 m. The trawl body consisted of a small outer mesh

(6.3-mm delta mesh) and a large inner mesh (34.9-mm bar mesh of 1.0-mm multifilament nylon) to protect smaller individuals from injury and damage due to larger fish or debris (e.g., rocks).

Trawl towlines were 38.1 m long to allow for a maximum effective depth of 5.4 m with a 7:1 drop ratio. Trawling was conducted perpendicular to shore for 3 minutes at approximately 3.2 km/h during the summer (June 24-July 13). Further details regarding the design, development, and specification of the benthic trawl (also known as the mini-Missouri trawl) used in our study are provided by Herzog et al. (2005), Guy et al. (2009), and Neebling and Quist (2011).

Modified-fyke nets (1 m × 2 m frame, 12.7-mm bar-measure mesh, 15.2 m lead; Miranda and

Boxrucker 2009) were used to sample mobile and structure-oriented species, such as sunfishes

(Lepomis spp.) and crappies (Poxomis spp.) located in littoral habitats (Hubert 1996). Fyke nets were set at dusk and retrieved the following day. Pulsed-DC electrofishing was used to sample fishes not collected with other sampling gears. Electrofishing catch rates are influenced by diel period (Sanders 1992; Reynolds 1996; McInerny and Cross 2004); therefore, electrofishing was only conducted at night (i.e., 30 minutes after sunset to 30 minutes before sunrise) to increase the number of species and individuals encountered (Fischer unpublished data). Electrofishing output was standardized at 2,750-3,250 W (Burkhardt and Gutreuter 1995; Miranda and Boxrucker

2009; Miranda 2009). Boat electrofishing was conducted with 5-minute runs, parallel to the shoreline, using two netters (6.3-mm delta mesh dipnets). Sampling with modified-fyke nets and boat electrofishing were conducted during the fall (September 14 to November 3). All sampled fish were identified to species in the field, if possible. Unidentified specimens were preserved in

10% formalin and identified in the laboratory. Catch-per-unit-effort (CPUE) was calculated for all species as the mean number of individuals per haul for trawls, mean number of individuals

per net-night for modified-fyke nets, and the mean number of individuals per hour of

electrofishing for each waterbody.

A systematic random sampling design was used to allocate samples for each waterbody

to ensure that a diversity of habitats was sampled. Specifically, the shoreline was divided into

segments that included at least one sample from each gear. The number of shoreline segments

(i.e., samples) in each waterbody was based on surface area. Water bodies less than 101 ha

received 10 samples, 102-202 ha received 12 samples, 203-405 received 15 samples, and those

greater than 405 ha received 20 samples of each gear. Shoreline segments were further divided

into reaches that were randomly assigned one of the three sampling gears.

Differences between habitat characteristics of reservoirs and natural lakes were tested

with a permutational multivariate analysis of variance (PERMANOVA). The PERMANOVA is

a non-parametric test that can directly partition variation from dissimilarity-based matrices

(Anderson 2001). The PERMANONVA does not rely on an assumption that the data are multivariate normal, or require continuous values and is, therefore, well suited for ecological assemblage data (Anderson 2001). A multivariate analogue to Fisher’s F-ratio and permutational-derived P-value are estimated with PERMANOVA (Anderson 2001). If differences in habitat characteristics between natural lakes and reservoirs were observed with the

PERMANOVA (i.e., P ≤ 0.05), a Wilcoxon test was used to evaluate differences between waterbody type for individual environmental characteristics.

Nonmetric multidimensional scaling (NMDS) was used to identify patterns of fish assemblage structure in the study systems. Fish assemblage composition was evaluated separately for each of the sampling methods (i.e., trawl, fyke net, electrofishing) to reduce the effect of sampling bias. Fish composition data used in the NMDS analyses consisted of species

CPUE and trophic guild CPUE. Trophic categories were assigned to individual species and based on those proposed by Goldstein and Simon (1999). Specific trophic categories used in our comparison included herbivores, detritivores, planktivores, invertivores, and carnivores.

Following Miranda et al. (2008), trophic guilds were combined to account for species that belonged to more than one trophic category during different life history stages. The combined classifications included invertivore-detritivore, invertivore-carnivore, invertivore-herbivore, planktivore-invertivore, and planktivore-detritivore for a total of nine trophic guilds. Rare species or guilds were defined as those occurring at less than 5% of the water bodies and excluded from analyses. The NMDS ordinations were conducted using Bray-Curtis dissimilarity matrices of fish assemblage data. Differences in fish assemblage structure between natural lakes and reservoirs were tested using PERMANOVA. Similar to the NMDS analysis, the

PERMANOVA analyses were conducted separately for each sampling method and assemblage composition measure (i.e., species, trophic). If differences in fish assemblage structure between natural lakes and reservoirs were observed (i.e., P ≤ 0.05), correlations between physicochemical and morphometric characteristics of lentic ecosystems and fish assemblage structure were evaluated using rotational vector fitting (Faith and Norris 1989). Vector fitting with NMDS ordinations was used to identify significantly correlated environmental vectors and to evaluate the direction of the maximum correlation. Environmental vector significance (P ≤ 0.05) was estimated using 999 random permutations of the data (Faith and Norris 1989). Rotational vector fitting analysis was conducted for combined assemblage data and ecosystem-specific (i.e., natural lake or reservoir) NMDS ordinations.

RESULTS

Physicochemical characteristics among water bodies were highly variable, but generally overlapped between natural lakes and reservoirs (Table 1). For example, reservoir size varied from 10 to 352 ha, while natural lake size varied from 48 to 2,174 ha. Overall, physicochemical characteristics differed between natural lakes and reservoirs (PERMANOVA; F1,43 = 12.2; P =

0.001). Natural lakes tended to be larger (W = 490.0; P < 0.0001), have smaller surface-area-to- watershed-area ratios (W = 158.0; P < 0.0001), shallower maximum depths (W = 212.0; P =

0.007), lower water clarity (W = 226; P = 0.0019), and higher suspended solids (W = 418.5; P =

0.019) than reservoirs. Mean depth, total nitrogen, total phosphorus, and chlorophyll a

concentrations were similar between waterbody types (P > 0.05).

Totals of 50 species, one hybrid, and 149,108 individuals were sampled from all water

bodies with all sampling methods. Trawling sampled 34 species and was the only method to

sample trout-perch Percopsis omiscomaycus and brook stickleback Culaea inconstans.

Electrofishing sampled 43 species and was the only method to detect spotfin shiner Cyprinella

spiloptera, sand shiner Notropis stramineus, and goldfish Carassius auratus. Modified-fyke nets

sampled 38 species and were the only method that sampled shortnose gar Lepisosteus

platostomus, river carpsucker Carpiodes carpio, and shorthead redhorse Moxostoma

macrolepidotum. Several of the species that were only sampled in reservoirs included sand

shiner, goldfish, rainbow trout Oncorhynchus mykiss, warmouth L. gulosus, and spotted bass

Micropterus punctulatus. Shortnose gar, spottail shiner N. hudsonius, river carpsucker,

shorthead redhorse, tadpole madtom Noturus gyrinus, trout-perch, brook stickleback, Iowa darter

Etheostoma exile and logperch Percina caprodes were only sampled in natural lakes. Channel

catfish Ictalurus punctatus (37 water bodies), green sunfish L. cyanellus (40 water bodies),

largemouth bass M. salmoides (44 water bodies), and black crappie P. nigromaculatus (44 water

bodies) were sampled in over 80% of the natural lakes and reservoirs. Bluegill L. macrochirus

was the only species sampled in every waterbody. Species richness varied from 10 to 30 (20.6 ±

1.5 species [overall mean ± SE]) for natural lakes and from 5 to 22 (10.7 ± 0.7 species) for

reservoirs. Overall, species richness differed between waterbody type (W = 498.5; P < 0.0001).

The NMDS ordinations of species composition from reservoirs and natural lakes produced stable solutions for samples from the benthic trawl (2 axes; stress = 0.16; Figure 1), modified-fyke nets (2 axes; stress value of 0.14; Figure 2), and night electrofishing (2 axes; stress value of 0.14; Figure 3). Results from the PERMANOVA indicated that fish species assemblage structure differed between natural lakes and reservoirs for sampling with the benthic trawl (F1,43 = 2.38; P = 0.005), modified-fyke nets (F1,43 = 6.80; P = 0.001), and night

electrofishing (F1,43 = 12.5; P < 0.001). Similar to species structure, NMDS ordinations of

trophic composition produced stable solutions for the benthic trawl (2 axes; stress = 0.07; Figure

1), modified-fyke nets (2 axes; stress = 0.09; Figure 2), and night electrofishing (2 axes; stress =

0.11; Figure 3). Trophic guild assemblage structure between reservoirs and natural lakes was

similar using data from the benthic trawl (F1,43 = 0.99; P = 0.388), but differed for modified-fyke

nets (F1,43 = 9.39; P < 0.001) and night electrofishing (F1,43 = 7.26; P < 0.001). Separate NMDS ordinations of fish species composition from the benthic trawl resulted in stable ordinations for natural lakes (2 axes; stress = 0.13) and reservoirs (4 axes; stress value of 0.07). Modified-fyke net species assemblage NMDS resulted in stable ordinations for natural lakes (2 axes; stress =

0.15) and reservoirs (2 axes; stress = 0.11). Stable NMDS ordinations for species composition night electrofishing were obtained for natural lakes (3 axes; stress = 0.08) and reservoirs (2 axes; stress = 0.10). Similar to species composition, stable NMDS ordinations were obtained for

87 trophic composition data from natural lakes for sampling with the benthic trawl (2 axes; stress =

0.12), modified-fyke nets (2 axes; stress = 0.05), and night electrofishing (2 axes; stress = 0.08).

Reservoir NMDS ordinations for trophic assemblage data from benthic trawl sampling (2 axes; stress = 0.07), modified-fyke nets (2 axes; stress = 0.07), and night electrofishing (2 axes; stress

= 0.07) resulted in stable ordinations.

Environmental vector fitting for combined NMDS ordinations of species composition data revealed strong differentiation of reservoir and natural lake assemblage structure along a gradient of waterbody size for all three sampling gears (Figures 1-3). Other significantly correlated physicochemical variables were primarily associated with NMDS axis 2 (modified- fyke net, Figure 2; night electrofishing Figure 3), suggesting limited influence on the separation of species composition between reservoirs and natural lakes. Environmental vector fitting for individual NMDS ordinations illustrated inconsistent patterns of habitat relationships between waterbody types. For instance, combined NMDS ordinations of fish species sampled with modified-fyke nets showed significant correlations for all of the physicochemical variables included in our evaluation, except watershed-to-lake-area ratio and TN concentration (Figure 2).

Reservoir fish species composition sampled with modified-fyke nets demonstrated similar structure that was significantly correlated with all of the variables evaluated; whereas, TN was the only factor related to natural lake species composition from modified-fyke net data (Table 3).

Similar to species composition, the environmental vector fitting for combined NMDS ordinations of trophic assemblage structure indicated that waterbody size differentiated reservoirs and natural lakes from modified-fyke net and electrofishing data (Table 3, Figure 2,

Figure 3). However, consistent patterns of correlated environmental variables were observed between individual NMDS ordinations for separate waterbody types (i.e., natural lakes and

reservoirs; Table 3). Although all of the environmental factors evaluated were significant (P <

0.05) for combined NMDS ordinations of trophic composition from fyke net and night

electrofishing sampling, no structuring for reservoir trophic composition was observed when

evaluated individually for both gears. In contrast, waterbody size was consistently correlated

with NMDS ordination axes for natural lake trophic composition for all three sampling methods

(Table 3).

Relationships between species richness and environmental factors also demonstrated

inconsistent patterns between natural lakes and reservoirs. Overall, more species were observed

in larger water bodies (rs = 0.77; P < 0.0001; Figure 4) and for reservoirs (rs = 0.69; P < 0.0001).

However, natural lake species richness was not significantly correlated with surface area (rs =

0.17; P = 0.55). Reservoir species richness also increased with other morphometric characteristics (Figure 4). Specifically, greater richness was observed as mean depth (rs = 0.45;

P = 0.01) and maximum depth (rs = 0.48; P = 0.006) increased in reservoirs. In contrast, species

richness of natural lakes was related to water quality variables (Figure 5). Fewer species were

observed in natural lakes with low Secchi depth (rs = 0.61; P = 0.022) and high TSS (rs = -0.65;

P = 0.013). Overall, diversity of natural lakes was associated with limnological factors, whereas

reservoirs appeared to be more strongly influenced by morphometric characteristics.

DISCUSSION

Lentic fish assemblages are commonly shaped by a multitude of factors simultaneously

acting at different scales (e.g., Eadie et al. 1986; Jackson et al. 1992 Jackson et al. 2001; Irz et al.

2007). In our study, fish composition of natural lakes was strongly associated with lake size and

few other physicochemical factors. In contrast, reservoir assemblage-environmental linkages

89 were observed for various habitat variables, but were generally weak and differed among sampling techniques. Species richness, however, was correlated with increasing lake size for reservoirs and increasing water clarity in natural lakes. Therefore, observed differences between the structuring of fish assemblages from natural and artificial lentic ecosystems may have resulted from a variety of interrelated constraints (e.g., abiotic and biotic factors, management legacies).

Physical differences between natural and artificial lacustrine ecosystems may have been responsible for the observed dissimilarity in faunal composition. For example, natural lakes tended to have larger surface areas with smaller watersheds. Fish diversity of lentic environments is often positively associated with waterbody surface area for natural and artificial ecosystems (e.g., Eadie et al. 1986; Irz et al. 2007). Miranda et al. (2008) demonstrated that fish assemblages of large reservoirs on the Tennessee River were structured by longitudinal gradients within the river system, where downstream reservoirs increased in surface area (i.e., littoral and riverine zones), habitat complexity, nutrient inputs, and species diversity (Miranda et al. 2008).

Species richness in reservoirs from our study was also strongly correlated with habitat size (i.e., surface area, mean depth, maximum depth). However, unlike Eadie et al. (1986) and Irz et al.

(2007), we did not observe a similar relationship between surface area and species richness in natural lakes. Species richness in natural lakes was positively associated with water clarity, but reservoirs frequently had greater water clarity (i.e., Secchi depth, TSS) and fewer species than natural lakes. Although decreased water clarity in natural lakes relative to reservoirs may seem counterintuitive, our study area included some of most eutrophic lentic environments in the world (Arbuckle and Downing 2001). Shallow, natural lakes are highly susceptible to resuspension of sediments and associated nutrients by abiotic (e.g., wind) and biotic (e.g.,

90 bioturbation) processes (Scheffer et al. 1993). Turbid natural lakes had fewer species than lakes with increased water clarity, whereas water clarity did not appear to be strongly associated with reservoir species richness. Therefore, factors like waterbody size and water clarity appeared to differentially influence fish assemblage composition of natural and artificial ecosystems.

Lack of environmental associations other than waterbody size for natural lakes and generally weak correlations for reservoir fish assemblage structure suggests that other factors not included in our analysis (e.g., basin characteristics, shoreline development) may have been more important. Decreased assemblage variability (i.e., similar patterns of species occurrence and abundance among water bodies) within ecosystems could have limited our ability to find habitat associations, despite heterogeneity in environmental conditions. Like physical habitat, fundamental differences in fish community dynamics between natural lakes and reservoirs likely resulted in the consistently dissimilar assemblage structure observed. Specifically, reservoirs are engineered ecosystems that exhibit chemical and physical characteristics intermediate to riverine and lentic habitats (Wetzel 2001). Despite the potential to support a diversity of lentic and riverine species (e.g., increased connectivity to lotic systems and greater habitat complexity), reservoirs often support a subset of stream fauna that possess ecological traits favorable for lacustrine habitats (Gido et al. 2009). Fewer lotic species present in reservoirs than available riverine species pool is likely the result of unsuitable habitat characteristics and negative biotic interactions (e.g., competition, predation) with the established fish fauna. For example, lotic and natural lake fish communities and their biotic interactions have developed over much longer temporal periods (i.e., thousands of years) than reservoirs constructed in the last century.

Although natural lakes have undoubtedly been altered by humans (e.g., pollution, stocking), natural systems may maintain more complex ecological interactions (e.g., predator-prey) than

91 those present in reservoirs that results in increased coexistence (Noble 1986). We observed several ubiquitous and abundant species (e.g., bluegill, black crappie) among waterbody types.

However, species richness was consistently lower in reservoirs and several species (e.g. spottail shiner, Iowa darter, and logperch) occurred in almost all natural lakes, but not in artificial lentic habitat. Therefore, it is not surprising that fewer habitat associations were observed for natural lakes compared to less diverse and sport-fish-dominated reservoirs.

Dissimilar management legacies could have also influenced findings from our contrast of natural and artificial lentic ecosystems. For example, differentiation in assemblage composition between natural lakes and reservoirs was likely the result of systematic differences in patterns of sport fish stocking between waterbody types. Redear sunfish L. microlophus is not native to

Iowa (Lee et al. 1980), but has been propagated and stocked in Iowa’s reservoirs since the 1960s

(Harlan et al. 1987). Redear sunfish was sampled in 38% of reservoirs compared to one natural lake. In contrast, walleye Sander vitreus was generally encountered more frequently in natural lakes (Harlan et al. 1987). Although walleye is also stocked in large reservoirs throughout Iowa, abundance was generally lower in reservoirs than that of natural lakes. Largemouth bass and channel catfish were more abundant in reservoirs despite being stocked into most lentic bodies throughout Iowa (Harlan et al. 1987). In addition to differences in stocking between natural and artificial lentic ecosystems, fish diversity in reservoirs may also be influenced by management activities associated with sport fish enhancement. For instance, rotenone has been widely used to eliminate invasive and undesirable native species with the goal of enhancing recreational fisheries across the U.S. (Flick and Webster 1992; Bettoli and Maceina 1996; Wydoski and

Wiley 1999) and Iowa (Lennon 1970). Therefore, the use of rotenone in the watershed during reservoir construction could have eliminated source populations of native species commonly

found in natural lentic water bodies. Elimination of source populations coupled with predator

heavy sport fish stocking may also have resulted in less diverse reservoir habitats.

Observed patterns of reservoir species composition structure were inconsistent among

fish sampling methods. Similarly, Menezes et al. (in press) found differences in environmental

relationships of Danish lake fish assemblages sampled with electrofishing, littoral gill nets, and

pelagic gill nets. However, it is not surprising that sampling selectivities and biases for different

methods may result in dissimilar characterizations of fish assemblages and subsequent habitat

relationships. Because multiple sampling methods are generally required to characterize fish

assemblages in lentic ecosystems (e.g., Jackson and Harvey 1997), the disparity between

observed fish-assemblage-habitat-associations for differing ecosystems and sampling methods

emphasizes the need to consider ecosystem type and sampling methodology when evaluating

lentic ecosystems. For instance, combined fish environmental correlations with fish assemblage

structure would not have reflected natural lake structuring with modified-fyke nets and night

electrofishing in our study.

Overall, the direct comparisons of fish assemblage-environment relationships indicated distinct differences between natural lakes and reservoirs that may provide guidance on the

management of lentic ecosystems. For instance, stronger relationships between environmental

characteristics and species composition in reservoirs implies that management activities (e.g.,

habitat restoration, nutrient load reduction) could increase benefits over similar actions in natural

lentic ecosystems. However, inconsistencies of assemblage-environment relationships between

sampling methods suggests that reservoir fish assemblage structure is complex and additional

research is needed to identify the relative importance of habitat characteristics to individual

species or taxonomic groups. Lack of strong physicochemical structuring of natural lake fish

93 assemblages suggests that further evaluation of biotic regulation and additional abiotic factors may be necessary to conserve and manage natural lake fish assemblages. Furthermore, identification of fish assemblage (e.g., Launois et al. 2011) and population characteristic (e.g.,

Guy and Willis 1995) relationships to environmental factors in lake and reservoir ecosystems may be critical development of ecological monitoring programs for lentic ecosystems and are particularly relevant to the management and conservation of biodiversity in lentic environments as anthropogenic factors increasingly threaten ecological integrity.

ACKNOWLEDGEMENTS

We thank M. Mork, S. Bisping, T. Neebling, J. Koch, M. Colvin, W. Massure, A.

Schaefer, R. Schultz, A. Fowler, J. Christiansen, T. Grimm, C. Bergthold, J. Bruegge, G.

Scholten, M. Sundberg, C. Smith, R. Krogman, C. Hinz, and N. Johnson for assistance with data collection. We also thank P. Dixon, J. Downing, C. Pierce, and K. Roe for helpful comments on a previous version of this manuscript. This project was supported, in part, by the Department of

Natural Resource Ecology and Management Iowa State University, Iowa Department of Natural

Resources, and Idaho Cooperative Fish and Wildlife Research Unit. The Unit is jointly sponsored by the University of Idaho, US Geological Survey, Idaho Department of Fish and

Game, and Wildlife Management Institute. The use of trade, firm, or product names is for descriptive purposes only and does not imply endorsement by the US Government.

REFERENCES

Anderson, M. J. 2001. A new method for non-parametric multivariate analysis of variance.

Austral Ecology 26:32-46.

Arbuckle, K. E., and J. A. Downing. 2001. The influence of watershed land use on lake N:P in a

predominantly agricultural landscape. Limnology and Oceanography 46:970–975.

Bettoli, P. W., and M. J. Maceina. 1996. Sampling with toxicants. Pages 303–333 in B. R.

Murphy and D. W. Willis, editors. Fisheries techniques, second edition. American

Fisheries Society, Bethesda, Maryland.

Barbour, C. D., J. H. Brown. 1974. Fish species diversity in lakes. American Naturalist

108:473–489.

Brown, J. H. 1981. Lakes as islands: biogeographic distribution, turnover rates, and species

composition in the lakes of central New York. Journal of Biogeography 8:75–83.

Burkhardt, R. W., and S. Gutreuter. 1995. Improving electrofishing consistency by

standardizing power. North American Journal of Fisheries Management 15:375–381.

Clavero, M., and V. Hermoso. 2011. Reservoirs promote the taxonomic homogenization of fish

communities within river basins. Biodiversity and Conservation 20:41–57.

Eadie, J. M. and A. Keast. 1984. Resource heterogeneity and fish species diversity in lakes.

Canadian Journal of Zoology 62:1689–1695.

Eadie, J. M. , T. A. Hurly, R. D. Montgomerie, K. L. Teather. 1986. Lakes and rivers as islands:

species-area relationships in the fish faunas of Ontario. Environmental Biology of Fishes

15:81–89.

Egertson, C. J., and J. A. Downing. 2004. Relationship of fish catch and composition to water

quality in a suite of agriculturally eutrophic lakes. Canadian Journal of Fisheries and

Aquatic Sciences 61:1784–1796.

Faith, D. P., and R. H. Norris. 1989. Correlation of environmental variables with patterns of

distribution and abundance of common and rare freshwater macroinvertebrates.

Biological Conservation 50:77–98.

Flick, W. A., and D. W. Webster. 1992. Standing crops of brook trout in Adirondack waters

before and after removal of nontrout species. North American Journal of Fisheries

Management 12:783–796.

Goldstein, R. M., and T. P. Simon. 1999. Toward a united definition of guild structure for

feeding ecology of North American freshwater fishes. Pages 123-202 in T. P. Simon,

editor. Assessing the sustainability and biological integrity of water resources using fish

communities. CRC Press, Boca Raton, Florida.

Gorman, O. T., and J. R. Karr. 1978. Habitat structure and stream fish communities. Ecology

59:507–515.

Guy, C. S., and D. W. Willis. 1995. Population characteristics of black crappies in South

Dakota waters: a case for ecosystem-specific management. North American Journal of

Fisheries Management 15:754–765.

Guy, C. S., P. J. Braaten, D. P. Herzog, J. Pitlo, and R. S. Rogers. 2009. Warmwater fish in

rivers. Pages 59–84 in S. Bonar, W. Hubert, and D. Willis, editors. Standard sampling

methods for North American freshwater fishes. American Fisheries Society, Bethesda,

Maryland.

Irz, P., A. Laurent, S. Messad, O. Pronier, and C. Argillier. 2002. Influence of site

characteristics on fish community patterns in French reservoirs. Ecology of Freshwater

Fish 11:123–136.

Irz, P., M. Odion, C. Argillier, and D. Pont. 2006. Comparison between the fish communities

of lakes, reservoirs and rivers: can natural systems help define the ecological potential of

reservoirs? Aquatic Sciences 68:109–116.

Irz, P., M. François, T. Oberdorff, T. Whittier, N. Lamouroux, D. Mouillot, and C. Argillier.

2007. Fish community comparisons along environmental gradients in lakes of France

and north-east USA. Global Ecology and Biogeography 16:350–366.

Harlan, J. R., E. B. Speaker, and J. Mayhew. 1987. Iowa fish and fishing. Iowa Department of

Natural Resources, Des Moines, Iowa.

Herzog, D. P., V. A. Barko, J. S. Scheibe, R. A. Hrabik, and D. E. Ostendorf. 2005. Efficacy of

a benthic trawl for sampling small-bodied fishes in large river systems. North American

Journal of Fisheries Management 25:594–603.

Hubert, W. A. 1996. Passive capture techniques. Pages 157–192 in B. R. Murphy and D. W.

Willis, editors. Fisheries techniques, second edition. American Fisheries Society,

Bethesda, Maryland.

Jackson, D. A., K. M. Somers, and H. H. Harvey. 1992. Null models and fish communities:

evidence of nonrandom patterns. American Naturalist 139:930–943.

Jackson, D. A., and H. H. Harvey. 1993. Fish and benthic invertebrates: community

concordance and community–environment relationships. Canadian Journal of Fisheries

and Aquatic Sciences 54:2807–2813.

Jackson, D. A., and H. H. Harvey. 1997. Qualitative and quantitative sampling of lake fish

communities. Canadian Journal of Fisheries and Aquatic Sciences 50:2641–2651.

Jackson, D. A., P. R. Peres-Neto, and J. D. Olden. 2001. What controls who is where in

freshwater fish communities: the roles of biotic, abiotic, and spatial factors. Canadian

Journal of Fisheries and Aquatic Sciences 58:157–170.

Karr, J. R. 1968. Habitat and avian diversity on strip mined land in east-central Illinois. Condor

70:348–357.

Klopfer, P. H., and R. H. MacArthur. 1960. Niche size and faunal diversity. American

Naturalist, 94:293–300.

Kohn, A. J., and P. J. Leviten. 1976. Effect of habitat complexity on population density and

species richness in tropical intertidal gastropod assemblages. Oecologia 25:199–210.

Launois, L., J. Veslot, P. Irz, and C. Argillier. 2011. Selecting fish-based metrics responding to

human pressures in French natural lakes and reservoirs: towards the development of a

fish-based index (FBI) for French lakes. Ecology of Freshwater Fish 20:120–132.

Lee, D. S., C. R. Gilbert, C. H. Hocutt, R. E. Jenkins, D. E. McAllister, and J. R. Stauffer Jr.

1980. Atlas of North American Fishes. North Carolina Biological Survey, Raleigh.

Lennon, R. E., J. B. Hunn, R. A. Schnick, and R. M. Burress. 1970. Reclamation of ponds,

lakes, and streams with fish toxicants: a review. Food and Agriculture Organization of

the United Nations, Fisheries Technical Paper 100.

Magnuson, J. J., W. M. Tonn, A. Banerjee, J. Toivonen, O. Sanchez, and M. Rask. 1998.

Isolation vs. extinction in the assembly of fishes in small northern lakes. Ecology

79:2941–2956.

McInerny, M. C., and T. K. Cross. 2004. Comparison of day electrofishing, night

electrofishing, and trap netting for sampling inshore fish in Minnesota Lakes. Special

publication 161. Minnesota Department of Natural Resources, Hutchinson.

Menezes, R. F., F. Borchsenius, J-C. Svenning, M. Søndergaard, T. L. Lauridsen, F.

Landkildehus, and E. Jeppesen. In Press. Variation in fish community structure,

richness, and diversity in 56 Danish lakes with contrasting depth, size, and trophic state:

does the method matter? Hydrobiologia.

Menge, B. A., J. Lubchenco, and L. R. Ashkenas. 1985. Diversity, heterogeneity and consumer

pressure in a tropical rocky intertidal community. Oecologia 65:394–405.