|UFDC Home||myUFDC Home | Help|
This item has the following downloads:
1 ENERGETIC CONSEQUENCES OF HABIT AT LOSS: TRADE-OFFS IN ENERGY ACQUISITION AND ENERGY EXPENDITURE BY Micropterus salmoides By JAKOB C. TETZLAFF A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUREMENTS FOR THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2008
2 2008 Jakob C. Tetzlaff
3 ACKNOWLEDGMENTS I would like to thank my graduate advisors, Dr. William E. Pine, III and Dr. Thomas K. Frazer, for their advice, support, and guidance. I woul d also like to thank Dr. Carl J. Walters for his support and constructive crit icism. Additionally, I would like to thank the members of the Florida Rivers Research Lab for their assistance : Matt Lauretta, Drew Dutterer, Oliver Burgess, Jared Flowers, Lauren Marcinkiewicz, Elissa Buttermore, Elise Bergman, and especially Ed Camp for helping me objectively observe the coasta l rivers. I thank Sherry Giardini and Michelle Quire for helping me with all the forms a nd deadlines I would have otherwise missed. Finally, I would like to thank my family a nd friends. I thank my pa rents for their support throughout this endeavor. Most importantly I would like to thank Casey Knutson for her understanding and commitment to me throughout this process. This study was supported in part, by a Stat e Wildlife Grant from the Florida Fish and Wildlife Conservation Commission.
4 TABLE OF CONTENTS page ACKNOWLEDGMENTS...............................................................................................................3 LIST OF TABLES................................................................................................................. ..........6 LIST OF FIGURES................................................................................................................ .........7 ABSTRACT....................................................................................................................... ..............9 CHAPTER 1 GENERAL INTRODUCTION..............................................................................................10 Foraging Efficiency............................................................................................................ ....10 Physical Structure and Foraging Efficiency...........................................................................13 Study Systems.................................................................................................................. .......14 Hypotheses..................................................................................................................... .........17 2 GROWTH, DIET AND ENERGETICS OF Micropterus salmoides IN TWO SYSTEMS WITH CONSTRASTING VE GETATIVE CHARACTERISTICS....................22 Introduction................................................................................................................... ..........22 Methods........................................................................................................................ ..........23 Study Systems.................................................................................................................23 Fish Sampling..................................................................................................................24 Growth......................................................................................................................... ....25 Diets.......................................................................................................................... .......26 Bioenergetics Model........................................................................................................27 Results........................................................................................................................ .............28 Growth......................................................................................................................... ....28 Reproductive Investment.................................................................................................29 Diet........................................................................................................................... .......29 Bioenergetics Model........................................................................................................31 Discussion..................................................................................................................... ..........31 3 MOVEMENT AND HOME RANGE OF Micropterus salmoides IN TWO SYSTEMS WITH CONSTRASTING VEGETA TIVE CHARACTERISTICS.......................................48 Introduction................................................................................................................... ..........48 Methods........................................................................................................................ ..........50 Study Systems.................................................................................................................50 Telemetry...................................................................................................................... ...50 Movement Patterns..........................................................................................................52 Time to Independence.....................................................................................................52 Home Range Estimation..................................................................................................53
5 Results........................................................................................................................ .............55 Discussion..................................................................................................................... ..........57 Ecological Findings.........................................................................................................57 Methodological Findings.................................................................................................59 Conclusion..................................................................................................................... ..63 4 SUMMARY AND CONCLUSIONS.....................................................................................78 LIST OF REFERENCES............................................................................................................. ..81 BIOGRAPHICAL SKETCH.........................................................................................................89
6 LIST OF TABLES Table page 2-1 Largemouth bass growth parameters for the Chassahowitzka and Homosassa Rivers.....38 2-2 Diets of largemouth bass from the Ch assahowitzka and Homosassa Rivers by size class.......................................................................................................................... ..........39 2-3 Percent composition by weight of fish prey for largemouth bass in the Chassahowitzka and Homosassa Rivers............................................................................40 3-1 Summary of largemouth bass tagged w ithin the Chassahowitzka and Homosassa Rivers......................................................................................................................... ........64 3-2 Summary of fish movement for la rgemouth bass in the Chassahowitzka and Homosassa Rivers; total movement recorded (m), mean daily movement (m/d) and standard deviation............................................................................................................. .65 3-3 Summary of observations used for determ ining kernel density home range estimates; the number of random observations pl aced within each river zone for each individual..................................................................................................................... ......66 3-4 Summary of independent observations used for determining kernel density home range estimates; the number of random obser vations placed within each river zone for each individual............................................................................................................ .68
7 LIST OF FIGURES Figure page 1-1 Locations of the Chassahowitzka a nd Homosassa Rivers within Florida.........................19 1-2 Site map of the Chassahowitzka River, Florida......... 1-3 Site map of the Homosassa River, Florida.............................21 2-1 Size structure of largemouth bass during summer and winter sampling periods in the Chassahowitzka and Homosassa Rivers............................................................................41 2-2 Observed total length (mm) vs. age (year s) and predicted total length vs. age for largemouth bass in the Chassahowitzka and Homosassa Rivers ......................................42 2-3 ln(body mass) vs. ln(gonad mass) for fe male largemouth bass in the Chassahowitzka and Homosassa Rivers.......................................................................................................43 2-4 Total length of predator vs. total length of fish prey for largemouth bass in the Chassahowitzka and Homosassa Rivers............................................................................44 2-5 Example of a diagnostic plot of parameter estimates for the four main variables of the bioenergetics model from the Chassahowitzka River..................................................45 2-6 Scenario 1 of bioenergetics model; estimates of energy acquisition and energy expenditure of largemouth bass from th e Chassahowitzka River and Homosassa River.......................................................................................................................... .........46 2-7 Scenario 2 of bioenergetics model; estimates of energy acquisition and energy expenditure of largemouth bass from th e Chassahowitzka River and Homosassa River.......................................................................................................................... .........47 3-1 Map of the Homosassa River displaying the location of acoustic receivers, their detection radius and the partitioning of observable and unobservable zones....................70 3-2 Map of the Chassahowitzka River displayi ng the location of acoustic receivers, their detection radius and the partitioning of observable and unobservable zones....................71 3-3 Average proportion of movement for largemouth bass over a 24 hour period for the Chassahowitzka and Homosassa Rivers............................................................................72 3-4 Kernel density home range estimates of largemouth bass within the Chassahowitzka and Homosassa Rivers.......................................................................................................73 3-5 Relationship between fish total length and home range size for largemouth bass in the Chassahowitzka and Homosassa Rivers......................................................................74
8 3-6 Schoeners ratio index for various time lags of autonomous receiver data collected in the Chassahowitzka River..................................................................................................75 3-7 Schoeners ratio index for various time lags of autonomous receiver data collected in the Homosassa River..........................................................................................................76 3-8 Relationship between sample size and ho me range size for largemouth bass in the Chassahowitzka and Homosassa Rivers............................................................................77
9 Abstract of Thesis Presen ted to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Masters of Science ENERGETIC CONSEQUENCES OF HABIT AT LOSS: TRADE-OFFS IN ENERGY ACQUISITION AND ENERGY EXPENDITURE BY Micropterus salmoides By Jakob C. Tetzlaff August 2008 Chair: William E. Pine, III Cochair: Thomas K. Frazer Major: Fisheries a nd Aquatic Science Increased nutrient loading and altered flow regimes in Floridas spring-fed, coastal rivers are possible drivers of documented changes in the structure of submer sed aquatic vegetation communities (i.e., abundance, distribution, and species composition). The broader ecological consequences of these changes in vegetative habitat, however, are not well understood. The objective of this study is to investigate the ro le of structural habitat (submersed aquatic vegetation) as it relates to the foraging behavior and foraging e fficiency of a freshwater apex predator, Micropterus salmoides Specifically, I investigate the trade-offs in energy acquisition and energy expenditure of wild fish in rivers that vary markedly in their vegetative character, but are otherwise similar with resp ect to their water chemistry (e .g., temperature, salinity and nutrient concentrations) and physic al attributes (e.g., depth and flow). My approach is to combine multiple independent methods of research including growth analysis, diet analysis, bioenergetics modeling, and telemetry to determ ine the potential energetic consequences of foraging in rivers that afford different degrees of structural habitat complexity. This study attempts to link broad-scale changes in habitat to both individual and population growth rates using an energetics framework.
10 CHAPTER 1 GENERAL INTRODUCTION The act of foraging (i.e., searching for and acq uiring food) is an esse ntial aspect of an animals ecology that has both costs and bene fits (Stephens and Krebs 1986). How animals balance these costs and benefits in variable en vironments is a central topic in the ecological literature. The theoretical basis of foraging behavior binds multiple ecological theories, including: optimal foraging theory (Schoe ner 1971, Charnov 1976, Stephens and Krebs 1986), life history theory (Stearns 1976, Abrams 1991), and predator-prey inte ractions (Holling 1966, Connell 1972). Predicting how animals balance the costs and bene fits of foraging in response to changes in food supply requires linking physiology with ecological interactions, since costs include both physiological and time costs of acq uiring food and also pred ation risks that are linked to foraging activities. Understanding the natu re of animal responses to variation in the availability of resources has implications for food web dynamics and many areas of community ecology. Foraging Efficiency Foraging efficiency can be defined as th e difference between the amounts of energy expended acquiring food and the amount of energy ingested from food (Schoener 1971, Pyke et al. 1977). If risk of predation wh ile foraging is fixed, an animal should attempt to maximize net energy gain (intake expe nditure). Net energy gain determines the amount of energy an animal can allocate to maintenance, growth, and repr oduction (Charnov 1976). By balancing the costs and benefits of foraging, animals can acquire e qual amounts of energy using alternative foraging strategies. The optimal foraging strategy, including foraging effort and the tactics used to acquire food resources, is a function of a number of va riables including: prey resource abundance, prey size, prey vulnerability, predati on risk, and the physical environm ent (e.g., structural complexity)
11 of the foraging universe. Changes to these variables can directly affect the foraging behaviors of predators, their resulting energy budgets, and ultimately thei r fitness potential. Predators should employ a foraging mode that maximizes net energy gain while feeding, unless that mode implies substa ntially higher predation risk th an other modes. Two general categories of foraging modes can be applied for most freshwater fish predators: sit-and-wait (ambush) or pursuing (active) (Pianka 1966, Schoener 1971). Most predators fall somewhere along this continuum of foraging be haviors, with individuals of so me species able to adopt both tactics, switching from a sit-and-wait to an activ ely foraging tactic when prey resources decline (Huey and Pianka 1981, Helfman 1990). Active foragi ng predators may encounter prey at greater rates than ambush predators; however, this incr eased encounter rate often comes at increased metabolic cost and increased predation risk (Huey and Pianka 1981). The foraging mode a predator employs represents a trade-off in the av erage long-term benefits of a switch against the costs prey must be captured for the predator to persist, but the energe tic costs of acquisition must not exceed the value of the prey captured (Krebs a nd McCleery 1984, Schoener 1987). Predators can also switch their foraging behavi or in response to changes in their habitat (Savino and Stein 1989, James 1994). To optimize fitn ess, predators should vary their behavior in habitats offering different leve ls of structural complexity to maintain levels of net energy intake that satisfy metabolic and reproduc tive demands, while simultaneously minimizing predation risk (Stehpens and Krebs 1986). Whethe r predators are able to recognize such tradeoffs and make optimal choices is an area of continuing research (Eli assen et al. 2007, Noonburg et al. 2007). In addition to predator foraging mode, prey avai lability and prey qual ity (e.g., type or size) have long been hypothesized to affect the net ra te of a predators en ergy intake via their
12 influence on the quantity of energy expended by a predator during foraging (Paloheimo and Dickie 1966, Kerr 1971, Werner 1979). In general, when food density decreases, an animal has to spend progressively more time searching for f ood. Foraging costs, in te rms of the time spent actively searching and the number of feeding attempts that would lead to satiation, should also increase as prey become small in relation to th e size of the predator (K err 1971). As the predator grows, foraging efficiency will be enhanced if the size of available prey also increases (Paloheimo & Dickie, 1966, Sherw ood et al. 2002). Maximum foraging efficiency occurs when the maximum energy is obtained from each prey with minimum effort. When larger prey are available, predators are likely to be less active and often exhibit greater growth potential (Rennie et al. 2005, Kaufman et al. 2006). Pr esumably, variation in growth rate, body size, and fecundity are effected through variation in these two important components of foraging, i.e., amount of time spent foraging and energy intake. Recent advances in the ability to es timate activity costs of wild fish in situ suggest that activity metabolism may represent a potentially large and variable fraction of energy budgets (Boisclair and Legget 1989, Boisclair and Sirois 1993, Rennie et al. 2005). Assuming that the budgeting of metabolic costs is additive, the energe tic demand of activity is in direct competition with other physiological functions for allocation within an animals energy budget. This would suggest a trade-off between energy for activity, an d energy for other physiological functions such as growth and reproduction. For example, across a number of fish species, individuals displaying elevated levels of activity have been found to have decreased growth ra tes (Henderson et al. 2004, Rennie et al. 2005, Kaufman et al. 2006). Thus, active metabolism can represent a significant foraging cost and may influence the foraging mode a predator employs.
13 Physical Structure and Foraging Efficiency The role of physical structur e, particularly submerged aquatic vegetation (SAV), in predator-prey interactions has been well documented in freshwater fish systems (Crowder and Cooper 1982, Anderson 1984, Heck and Crowder 1991) In general, predator-feeding rates should be greatest at intermediate levels of structural complexity because of the contrasting effects of decreasing foraging efficiency a nd increasing prey abundance with increasing structural complexity (Crowder and Cooper 1982, Crowder et al. 1998). While the overall influence of structural complexity on foraging be havior is well underst ood, most studies have focused on the effects of high structural density on predator efficiency (Savino and Stein 1982, Anderson 1984, Bettoli et al. 1992). The relatio nship between energy acquisition rate and physical structure at low struct ural complexity is less clear and likely varies by species and between aquatic systems (Heck and Orth 1980, Crowder and Cooper 1982). The ability of a predator to feed efficiently at low structural complexity de pends largely on the behavioral response of the predator to changes such as prey abundance, prey composition, and prey vulnerability associated with habitat change. Variation in levels of structural comple xity of aquatic ecosystems has important consequences for the foraging ecology of predator s and may affect growth and survival through its influence on the availability and vulnerability of prey species. Habita t structure can influence the abundance, diversity, and composition of pr ey communities (Bettoli et al. 1993, Crowder et al. 1998, Wyda et al. 2002). Changes in the prey community associated with habitat changes may force predators to alter their foraging behavior or forage on suboptimal prey (Sass et al. 2006). The structural characteristics of a particular habitat type can al so influence a predators foraging efficiency. Nelson (1979) and Sa vino and Stein (1989) reported that some predator species forage most successfully in habita ts having structure present at an optimal level, above or below
14 which success decreases. For ambush predator s, physical structure aids predation by concentrating prey (that typically use structure for concealment, but must move in order to forage themselves), concealing a predators appro ach, reducing its visibility to potential prey and increasing its capture success (Heck and Or th 1980, Coen et al. 1981, James 1994, Flynn and Ritz 1999). Thus, habitat complex ity has the potential to influe nce a variety of components related to the energy acquis ition rates of predators. Loss of structural complexity can increase th e metabolic costs of foraging for predatory fish through three primary mechanisms: reduced prey abundance, reduced capture success, and altered foraging mode. For a given size prey, a decrease in abundance will force predators to spend more time actively searching (Kerr 1971) In addition, capture success for ambush predators is often reduced when structural habitat is absent. Pr edators may also switch foraging modes from an ambush mode to an actively foraging mode as structural habitat decreases (Savino and Stein 1989). In order to balance increased energetic expenditures at low structural complexity, predators must increase their ener gy intake or else will have lower net energy available for maintenance, growth and reproduc tion. The combination of direct and indirect foraging costs associated with low structural ha bitat has the potential to depress the long-term rate of energy intake for predatory fishes. Decreases in the daily rate of production for individuals can have direct c onsequences at the individual (growth, condition, maturity, and fecundity), population (age stru cture, density, biomass, and production) and community (species interactions, biomass, productivity, co mposition, and distribution) levels. Study Systems In lotic systems, nutrient enrichment often l eads to increases in the biomass of suspended and/or benthic algae and peri phyton associated with submer sed aquatic vegetation (SAV) (Duarte 1995, Smith et al. 1999). These effects can cause substantial change s to the structural
15 characteristics of rivers and stre ams, leading to nearly complete elimination of SAV (Wetzel and Likens 1991, Frazer et al. 2006). Declines in S AV associated with increasing nutrient loading may reduce the quantity and complexity of habitat and therefore directly influence predator-prey interactions. Along the interior west coast of peninsular Florida, increases in nitrate concentrations have been documented in several first magnitude spri ng systems (Jones et al. 1997, Frazer et al. 2006). These springs serve as the origin of flow for ma ny coastal Florida streams that, because of their shallow depths, favorable substrate characteristic s and clear water, have historically supported dense assemblages of SAV and associated fa unal communities. Recent research indicates a precipitous decline in macrophyt e abundance in a number of thes e systems coincident with marked increases in nutrient load ing rates and periphyton associated with the plants (Frazer et al. 2006). Of particular concern is the decline of native species such as Vallisneria americana and Sagittaria kurziana The ecological consequences of SAV lo ss in these systems have yet to be fully investigated. My study sites are the Chassahowitzka and Homosassa Rivers, which are both small, spring-fed rivers along Floridas southwest Gulf (Figure 1-1). These syst ems provide ideal study sites for investigating potential ecological effects of declines in SAV because while they similar with respect to their physical (size, discharge, temperate, depth, substrate) and chemical (nutrients, salinity) characteristics, SAV in the Homosassa River is markedly reduced compared to the Chassahowitzka Rive r (Frazer et al. 2006). The Chassahowitzka River runs west approximat ely 4 km from the main spring boil to the beginning of the associated coastal marsh complex and then another 4 km to the Gulf of Mexico (Figure 1-2). Flow throughout the river is tidally influen ced (Yobbi and Knochenmus 1989).
16 Above the marsh complex, mid-stream channel de pth averages 1.0 m and the average width of the river is 92 m. The river is bordered by th e Chassahowitzka National Wildlife Refuge and is characterized by limited riparian development (Fi gure 1-2). Submersed aquatic vegetation occurs throughout the majority of the fr eshwater portion of the river a nd is characterized by a patchy heterogeneous distribution. Common macrophytes include Vallisneria americana Sagittaria kurziana, Potamogeton pectinatus Najas guadalupensis Myriophyllum spicatum and Hydrilla verticillata Filamentous macroalgae, including Lyngbya sp. and Chaetomorpha sp., are also abundant. The Homosassa River runs west approximately 5 km from the main spring complex to the beginning of the associated coastal marsh complex and then another 7 km to the Gulf of Mexico (Figure 1-3). The majority of stream discharge emanates from a main spring complex; however, smaller spring runs in the upper river contribut e additional flow. Tidal cycles influence both spring discharge and flow within the river (Yobbi and Knochenmus 1989). The average width of the river is 130 m. The average mid-stream channe l depth of the freshwater portion of the river is 1.8 m. Riparian zones are extensively develope d by homes and businesses. A majority of the shoreline in the upper reaches of th e river consists of sea walls a nd lacks natural riparian habitat (Figure 1-3). Trends in submersed aquatic vege tation show drastic declines over the past 11 years with current composition of SAV limited pr imarily to filamentous algae (Frazer et al. 2006). The river historically supported a dens e macrophyte community, comprised of a similar composition to that found currently in the Chassahowitzka River, including Vallisneria americana Sagittaria kurzian, Potamogeton pectinatus and Najas guadalupensis (Frazer et al. 2006).
17 Hypotheses The research purpose of this study was to inves tigate how structural habitat influences the foraging efficiency, both food intake and activity level, of a freshwater fish predator, largemouth bass Micropterus salmoides I utilized a case-study approach to examine potential trade-offs in foraging behaviors by quantifying the diets and movement pattern s of largemouth bass in two river systems with contrasting levels of structural complex ity. My first objective was to characterize the size structure, ag e structure, and growth patter ns of largemouth bass in each system. Assuming similar patterns of allocat ion towards reproduction, this allowed me to examine if there were differences in the am ount of energy each populat ion had available to allocate towards growth. My second objective was to quantify diet charac teristics of largemouth bass in each system. This allowed me to determin e if differences in energy intake alone could explain possible differences in growth of largem outh bass between the rivers. My third objective was to quantify activity, movement, and home range size of largemouth bass in each system. My expectation was that largemouth bass will employ a lternative foraging strategies as a result of differences in their foraging environment. I then combined results from each of these aspects of the study in an attempt to discern how trade-o ffs in prey intake and activity patterns could explain differences in growth patte rns and/or size/age structure. I hypothesized that the average energy acqui sition rates of largemouth bass in the Homosassa River (low structural complexity) w ould be lower than for largemouth bass in the Chassahowitzka River (intermediat e structural complexity) as a result of increased foraging costs. I hypothesized that differences in foragi ng efficiency would be recognizable at both individual and population levels Specifically, I tested: H1: Growth by largemouth bass is great er in the Chassahowitzka River, H2: Consumption by largemouth bass is greater in the Homosassa River,
18 H3: Activity and space use by largemouth bass is greater in the Homosassa River. I used information on the consumption and activity levels of largemouth bass in each river to explore alternative hypothesis re garding possible consequences of an altered energy budget, driven by changes in habitat structural complexity, on populations of largemouth bass.
19 Figure 1-1. Locations of the Chassahowitz ka and Homosassa Rivers in Florida.
20 Figure 1-2. Site map of the Chassahowitzka River, Florida.
21 Figure 1-3. Site map of the Homosassa River, Florida.
22 CHAPTER 2 GROWTH, DIET AND ENERGETICS OF Micropterus salmoides IN TWO SYSTEMS WITH CONSTRASTING VEGETATIVE CHARACTERISTICS Introduction The role of physical structur e, particularly submerged aquatic vegetation (SAV), in predator-prey interactions has been well documented for freshwater fish communities (Crowder and Cooper 1982, Anderson 1984, Heck and Crowder 1991). In general, predator-feeding rates should be greatest at intermediate levels of structural complexity because of the contrasting effects of decreasing foraging efficiency and increasing prey abundance and availability with increasing structural complexity (Crowder and Cooper 1982, Crowder et al. 1998). While the overall influence of structural complexity on fo raging behavior is well understood, most studies have focused on the effects of high structure de nsity on predator efficiency (Savino and Stein 1982, Anderson 1984, Bettoli et al. 1992). The relati onship between energy acquisition rate and physical structure at low struct ural complexity is less clear (Heck and Orth 1980, Crowder and Cooper 1982). The ability of a predator to feed e fficiently at low structur al complexity depends largely on the response of the predator to changes such as prey abundance, prey composition, and prey vulnerability relative to habitats with greater complexity. Variation in structural complexity am ong aquatic ecosystems may have important consequences for the foraging ecology of predators, and may affect growth and survival through its influence on the availability and vulnerability of prey species. Habita t structure can influence the abundance, diversity, and composition of pr ey communities (Bettoli et al. 1993, Crowder et al. 1998, Wyda et al. 2002). Changes in the prey community associated with habitat changes may force predators to alter their foraging behavior or forage on suboptimal prey (Sass et al. 2006). Structural habitat can also influence predator foraging efficiency. Ne lson (1979) and Savino and Stein (1989) reported that some predator specie s forage most successfully in habitats having
23 structure present at an optimal complexity, abov e or below which success decreases. For ambush predators, physical structure aids predation by concealing a predators approach, reducing its visibility to potential prey, a nd increasing its capture success (Heck and Orth 1980, Coen et al. 1981, James 1994, Flynn and Ritz 1999). Thus, habitat complexity has the potential to influence a variety of components rela ted to the energy acquisit ion rates of predators. In Chapter 2, I compare growth and food habits of largemouth bass in two river systems which vary markedly with respect to their structural complexity due to differences in abundance, and composition of SAV. I hypothesize that net energy intake rates by largemouth bass will be depressed in the system with lowe r structural complexity as a resu lt of increased foraging costs. These differences in net energy intake will be manifest as differences in energy allocated to growth and reproduction. I examine growth patter ns of largemouth bass in each system as a proxy for differences in net energy intake rates. I compare diet composition and relative prey sizes between the two po pulations as indices of energy acqui sition. I then use a bioenergetics model to examine scenarios that demonstrate how largemouth bass may trade-off energy acquisition with energy expenditures to maximize fo raging efficiency. I combine results from the bioenergetics model with observed diet data to assess how differing combinations of energy acquisition and energy expe nditure could explain observed growth rates. Methods Study Systems Two spring-fed rivers systems along Floridas southwest Gulf coast, the Chassahowitzka and Homosassa Rivers, were selected for this pr oject. These two systems are similar with respect to their physical (temperate, dept h, substrate) and chemical (nutri ents, salinity) characteristics, but the submerged aquatic vegetation (SAV) communities are markedly different (Frazer et al. 2006). The Chassahowitzka River is composed of a patchily distributed heterogeneous SAV
24 community, while the Homosassa River is nearly devoid of rooted macrophytes and macroalgal abundance is comparatively sparse. A detailed desc ription of these rivers is provided in Chapter 1. Fish Sampling Largemouth bass populations were sample d from January 2007 through March of 2008. All fish were collected using boat electrofishi ng (Smith-Root Inc.; Mark IX GPP unit pulsed-DC; 20-30 A). Sampling was conducted one day each mont h on each river to obtain diet and growth information. Intensive three day mark-recaptur e studies were carried out in July 2007 and January 2008 to obtain largemouth bass population estimates (Frazer et al. 2008). In October 2007 and February 2008 a majority of the largemout h bass collected were sacrificed (N = 97 and 98 in the Chassahowitzka and Homosassa Rivers, respectively) to obtain information on growth and reproduction. Most largemouth bass collected from Janua ry 2007 to March 2008 were weighed (g), measured (total length [TL]; mm), marked with individually numbered T-bar anchor tags (Hallprint Pty Ltd), and released. Tags were insert ed into the muscle tissue near the base of the soft dorsal fin at a 45 degree an gle so that the anchor locked behind the pterygiophores (Guy et al. 1996). Diet samples were taken from at least th irty fish with total le ngth greater than 250-mm during each monthly sample. Diet contents were collected with a gastric lavage technique (Seaburg 1957). Diets were stored in 118-ml Wh irl-Pak sample bags and placed on wet ice. Diets were frozen upon return to the laboratory. During the October and February sampling pe riods a majority of the largemouth bass collected were sacrificed and pl aced on ice in the field. Total le ngth (mm) and weight (g) were measured and each fish was sexed if possible. Gonads, muscle tissue, stomachs, and saggital
25 otoliths were collected from all fish sacrifice d. Gonad wet weight to dr y weight ratio and gonadal somatic index (GSI) were calcu lated for sacrificed fish. Saggital otoliths were remove d, cleaned, and allowed to dr y before mounting. Otoliths were mounted on microscope slides using Crys tal Bond cement. Otoliths were sectioned using a low speed diamond wheel saw (Model 650 South Bay Technologies), cutting along a dorsoventral plane passing through the otolith nucleus. Otolith sect ions were viewed independently by two readers th rough dissecting microscopes at 6.3 40X magnification with transmitted light. All otolith sections were read without knowledge of fish length or weight. Any discrepancies in age determination were solved by re-reading the otolith in concert with a second experienced reader. Growth Length-frequency distributions of largem outh bass collected during the summer and winter extended sampling periods were compared between each river. Shapiro-Wilk tests were first used to test for normality in length-fre quency distributions. Since data were non-normal, Kolmogorov-Smirnov tests were used to test for differences in the cumulative length frequency distribution among populations. In addition, largemouth bass were divided into four length groups: < 200 mm TL, 200 to < 300 mm TL, 300 to < 380 mm TL, and 380 mm TL, and chisquare tests were used to assess if the propor tion of fish in each length group varied between rivers for the two sampling periods. A von Bertalanffy growth equation with an addi tive error structure (R icker 1975) was fit to age data collected on each population. The su m-of-squares differences between observed and predicted length-at-age was minimized using th e Solver tool in Microsoft Excel (Frontline systems, http://www.solver.com) to fit the equation
26 ] e [1 ) *( -ot t k tL L (2-1) where Lt is the predicted length at a given age t, L is the asymptotic mean length, k is the metabolic coefficient, t is age (years), and to is the hypothetical age when the mean fish total length is 0. Growth curves were compared using a likelihood ratio test (Kimura 1980). Chisquare statistics were used to test for growth differences between the populations based on the von Bertalanffy growth parameters. This procedure involves te sting the hypothesis that a common growth curve is adequate to fit data from both populati ons. If the common growth curve hypothesis is rejected, then a series of more restrictive models and hypotheses that omit parameters and populations are compared to eluc idate differences in growth parameters. In addition, age data for largemouth bass collected from the Homosassa River in 1985 (Porak et al. 1987) were compared with data collected in 2007-2008 to test if growth parameters for largemouth bass in the Homosassa River had cha nged over time. There were no historical data available for the Chassahowitzka River. Diets All prey items were i ndentified to the lowest possible taxonomic order. Items that were not identifiable to family (e.g., bones, scales, or pieces of tissue) were grouped as either unknown fish or unknown invertebrates. The total length (mm) and dry weight of each individual prey item was measured. Items were dried at 60C for 48 hours, and weighed to 0.0001 g. Diet items were grouped into broad categories for comparison. To characterize largemouth diets for each populat ion, largemouth bass were divided into the same four length groups used to compare size structure. The per cent composition of prey items by weight, total weight of stomach contents and standardized consumption (weight of all stomach contents [g] / fish body weight [g]) were used to characterize diets (Hyslop 1980). In
27 addition, mean number of identifiable fish prey per stomach, mean length of whole fish prey, and the percentage of empty stomachs were compar ed between largemouth bass populations in the two rivers. Bioenergetics Model Size-at-age data and growth increment data from tagging were used to fit a general growth model following methods outlined by C. J. Walters and T. E. Essington (University of British Columbia and University of Washington, personal communication ). The model allows for estimation of seasonal, temperature driven feeding and metabolism parameters. The model uses a basic bioenergetics equati on to describe the rate of chan ge in weight as (Paloheimo and Dickie 1965): ) ( ) ( T fm mW T fc HW dt dWn d (2-2) In this equation, the first term, HWd, describes the anabolic pro cesses associated with food acquisition; where, H represents the rate at wh ich an animal acquires mass, W represents the body mass of the animal, and d represents a parameter for scali ng anabolic processes with mass. The second term, mWn, represents the catabolic processes where m is the rate at which an animal looses mass and n is the scaling factor of catabolic processes with mass. The function fc(T) represents variation in food intake rate with environmental temperature T and fm(T) represents variation in metabolic rate with temperat ure. These functions were modeled using Q10 equations of the form f(T)= Q10 (T-10)/10, with different Q10 values for consumption and metabolism. The bioenergetics model was used to invest igate a range of scen arios concerning how feeding and metabolism parameters for each pop ulation could explain the observed growth patterns in size-at-age and growth increment data. A large number of model runs were conducted in which some parameters were fixed between populations while others were allowed to vary.
28 Fixing some parameters while allowing others to vary allowed for ex amining the range of parameter combinations that could credibly e xplain (fit) the data, using a complex likelihood criterion devised by C. J. Walters and T. E. Essington (University of British Columbia and University of Washington, personal communication ). All possible combinations of fixed and varying parameters were run to evaluate common patterns in parameter estimates. Results Growth Cumulative length-frequency di stributions were significantly different between rivers during each sampling period (KS test, D = 1, P < 0.001, for summer and winter). During both the winter and summer expanded samp ling periods the proportion of largemouth bass sampled from each size class was significantly different between rivers: summer size structure ( 2 = 118.53, df = 3, P < 0.001), winter size structure ( 2 = 14.17, df = 3, P = 0.002) (Figure 2-1). The lifetime growth patterns of largemouth bass also diffe red between rivers (Figure 2-2). Growth of individuals in the Homosassa River exceeded grow th in the Chassahowitzka River for the first 2 years of life. Beyond age 4, largemouth bass in the Chassahowitzka River were larger than largemouth bass in the Homosassa River. Largem outh bass in the Chassahowitzka River reached a greater maximum body length than bass in the Homosassa Rive r (Table 2-1). Von Bertalanffy growth models fit the age data from both populations well (Table 2-1, Figure 2-2). Growth curves were sign ificantly different between rivers ( 2 = 11.14, df = 3, N = 194, P = 0.01). Differences in growth curves between rivers were best ex plained by differences in the asymptotic length ( 2 = 2.02, df = 1, P = 0.15). Growth curves of largemouth bass from the Homosassa River collected in 1985 were significantly different than those collected in 20072008 ( 2 = 78.96, df = 3, N = 223, P < 0.001). Differences between time periods were best explained by differences in the asymptotic length ( 2 = 1.65, df = 1, P = 0.19); the metabolic
29 coefficient ( k ) and hypothetical age at whic h fish length equals 0 ( to) did not differ ( 2 = 0.14, df = 1, P > 0.90). Reproductive Investment Reproductive investment by mature fema les was compared by plotting gonad mass against body mass (Figure 2-3). Fe males with undeveloped ovaries were excluded from analysis. GSI of mature females was not signif icantly different between populations (F[1,36] = 9.458, P = 0.21; for Homosassa River, mean GSI = 0. 044 +0.00026 SD, N = 21; for Chassahowitzka River, mean GSI = 0.036 +0.00005 SD, N = 17). For males, GSI was also not significantly different between populations (F[1,36] = 1.486, P = 0.231; for Homosassa River, mean GSI = 0.016 +0.000089 SD, N = 22; for Chassahowitzk a River, mean GSI = 0.012 +0.000033 SD, N = 15). In the Chassahowitzka River four of th e six females sampled less than 300 mm TL had immature ovaries. In the Homosassa River, onl y one of ten females sampled less than 300 mm TL had immature ovaries (Figure 2-3). Diet A total of 664 largemouth bass were examined for food habits, 353 in the Chassahowitzka River and 311 in the Homosassa Rive r. Diet consisted of mostly fish in both systems, comprising 62% and 81% of the total co mposition by weight in the Chassahowitzka and Homosassa Rivers, respectively. Percent compositi on of diet items by size class for largemouth bass in each river is given in Table 2-2. No di fferences in prey composition were observed in the smallest size class. For the 200 to < 300 mm si ze class, largemouth bass in the Chassahowitzka River consumed a greater proporti on of crayfish and a lower propor tion of fish than largemouth bass in the Homosassa River. For the 300 to < 380 mm size class, largemouth bass in the Chassahowitzka River consumed a greater proporti on of crayfish and lower proportion of marine crustaceans than largemouth bass in the Homosassa River. For the largest size class, largemouth
30 bass in the Chassahowitzka Rive r consumed a greater proportion of amphibians and a lower proportion of fish than largemouth bass in the Homosassa River. Across all size classes, the percentage of empty stomachs was greater in the Chassahowitzka River than the Homosassa River (T able 2-2). Average mass of stomach contents was greater in the Homosassa River across all size classes (Table 2-2), however, the difference was only statistically significant for the smallest size class (F[1,101] = 4.96, P = 0.03). Although standardized consumption rate tended to be hi gher for largemouth bass in the Homosassa River than in the Chassahowitzka River, these differences were not sta tistically significant (Table 2-2). Only for the 200 to < 300 mm size class was standardized consumption rate greater for largemouth bass in the Chassahowitzka River. The number of identifiable fish prey per stom ach was greater in the Homosassa River for all but the smallest size classe s of largemouth bass (Table 22). For the 300 to < 380 mm size class, bass in the Chassahowitzka River consumed on average significantly fewer number of prey fish (F[1,77] =5.99, P = 0.02). The size of prey fish consumed varied inversely w ith the number of prey fish consumed. The size of prey fish cons umed was greater in the Homosassa River for the smallest size class (Table 2-2). Across all othe r size classes, the size of fish consumed was greater in the Chassahowitzka Ri ver (Figure 4); this difference was only statistically significant for the 300 to < 380 mm size class (F[1,68] =4.59, P = 0.03). The composition of prey fish consumed varied across all size classes (Table 2-3). For all size classes except the largest, largemouth bass in the Homosassa River had a greater diet breadth of prey species. In a ddition, diets of largemouth bass in the Chassahowitzka River tended to be dominated by the largest available prey species at each size cla ss (Frazer et al. 2008) whereas largemouth bass in the Homosassa Rive r consumed a range of prey species.
31 Bioenergetics Model The bioenergetics model produced two contras ting, but equally plausible scenarios, to explain the patterns of observed growth in each system. The bioenergetics parameter estimates obtained were relatively unsta ble and highly confounded as indicated by Markov chain Monte Carlo sampling of the model likelihood function (Figure 2-5). The diagnostic plot illustrating pairs of parameter combinations with relatively high likelihood (relatively good fit to the data) showed that the food consumption parameter (H) and metabolic parameter (m) as well as their associated scaling parameters (d) and (n) were highly confounded (Figur e 2-5). That is, the model can explain (fit) th e patterns of growth equally well using a wide range of parameter combinations for energy intake and energy e xpenditure. Despite the uncertainty in exact parameter estimates, the model produced two consis tent scenarios to explain the growth data. The first scenario had both energy intake and energy expenditure rates of largemouth bass higher in the Chassahowitzka River than in the Homosassa River (Figure 2-6). In this case, the amount by which energy intake exceeded energy expenditu re in the Chassahowitzka River was greater than the difference in the Homosassa River. The second scenario had energy intake by largemouth bass slightly higher in the Homosassa River but en ergy expenditures much greater than in the Chassahowitzka River (Figure 2-7). In this case, the difference between energy intake and expenditure was also greater in the Chassaho witzka River. Both scenarios predicted greater foraging efficiency by largemouth bass in the Ch assahowitzka River, allowing the model to fit the observed pattern of growth rates in both rivers. Discussion Observed differences in growth curves of largemouth bass between the two rivers investigated could not be explained with cons umption data, suggesting a possible difference in foraging efficiency between the two systems. Si ze structure and growth curves of largemouth
32 bass were significantly different between the Ch assahowitzka and Homosassa Rivers with the catch from the Chassahowitzka River having a hi gher abundance of older a nd larger individuals. Largemouth bass in the Homosassa River had fast er initial growth rate s through the first two years of life, but lower asymptotic lengths. Few individuals greater than 380 mm and no individuals greater than age-5 were sampled in the Homosassa River. Comparisons of age-length data collected in 1985 with those collected in 2008 from the Homosassa River revealed a lower asymptotic length at present compared to 1985. Largemouth bass in the Homosassa River had greater standardized consumption rates across all sizes compared to largemouth bass in the Chassahowitzka Rive r. Greater consumption by the smallest size class of largemouth bass in the Homosassa River likely contributed to the higher growth rates observed at younger ages in the Homosassa River. However, despite greater consumption rates, higher stomach content mass, and a lower proportion of empty stomachs, the growth rates of largemouth bass in the Homosa ssa River slowed after age-2 and reached an asymptote at a total length that was 100 mm less than that of largemouth bass in the Chassahowitzka River. The fact that food consump tion failed to explain differences in observed growth suggests that growth differences may resu lt from greater foraging costs in the Homosassa River. Differences in the size of prey consumed by largemouth bass in each system lend additional support to the hypothesi s that foraging costs are grea ter in the Homosassa River. Increased energetic costs associat ed with foraging on sub-optimal prey sizes has been proposed as an explanation for slow growth in a vari ety of species (Sherwood et al. 2002). Largemouth bass in the Homosassa River tended to consume gr eater numbers, but smaller sizes of prey fish, as they grew. Within the smallest size class, largemouth bass in the Homosassa River consumed
33 larger average fish prey and a lower mean nu mber than largemouth ba ss in the Chassahowitzka River. However, as fish grew in the Homosassa River they tended to rely on relatively smaller prey. Average size of fish prey consumed increased by a factor of 2.5 in the Chassahowitzka River between the juvenile and adult size classes while average size only increased by 1.4 in the Homosassa River between the same size classes. As predators grow, an increase in ration is necessary to maintain a positive growth rate (K err 1971). If large prey are not available, the number of small prey needed to maintain positiv e growth will increase with predator body size, and the amount of energy required to obtain incr easing numbers of small prey will also rise (Sherwood et al. 2002). Increased foraging costs a ssociated with a contin uing reliance on smaller prey ( Sherwood et al. 2002, Rennie et al. 2005, Kaufman et al. 2006 ) may be one factor contributing to the patterns of growth rates observed in this study. However, the differences observed in prey size are likely not large enough to completely explain the patterns of growth in each population (Bajer et al. 2004). Bioenergetics modeling revealed two plausibl e scenarios of energy acquisition and energy expenditure which could explain the observed differences in grow th between rivers. The first scenario is that largemouth bass in the Chassahowitzka Rive r maximize foraging efficiency by having greater acquisition rates al ong with greater metabolic rates than largemouth bass in the Homosassa River. The second scenario is that largemouth bass maximize foraging efficiency in the Chassahowitzka River by having similar or slightly lower, ener gy acquisition rates but substantially lower metabolic rates than largemou th bass in the Homosassa River. Each of these foraging strategies has theoretical support: the first scenario is pr emised on the strategy of energy maximization, that the assumed goal of foraging is to maximize the long-term rate of energy intake (Stephens and Krebs 1986); the second scen ario is premised on the strategy of time
34 minimization, that an animal forages only long enough to obtain energy requirements, thus decreasing predation risks (Schoener 1971). While th ese strategies are often considered mutually exclusive or at least two e nd points along a continuum of fo raging strategies (Hixon 1982), ambush predators can maximize energy intake while minimizing energy expenditures when selectively foraging in productive environments Selection is likely to favor any animal maximizing its net energy input in a minimum amount of time while actually foraging (Pyke et al. 1977). The consumption data from this study supports the bioenergetics model scenario 2 where largemouth bass in the Chassahowitzka River maximize foraging efficiency by minimizing foraging costs. There are at least four altern ative explanations which could explain the differences in the growth patterns observed: (1) differences in thermal regime, (2) differences in the energy content of prey items, (3) differences in the allocation of energy, and (4) differenc es in adult mortality. Temperature is unlikely to explai n the differences in growth as both systems are spring-fed and exhibit similar water temperatures (Frazer et al 2006). Differences in the composition or energy content of prey were also unlikely to explai n the growth patterns observed. Largemouth bass consumed primarily fish prey across all sizes in both rivers. The major difference in prey composition between populations was that largemouth bass consum ed a greater proportion of crayfish in the Chassahowitzka River, and crayfi sh have a lower caloric density than fish (Pope et al. 2001). Fish prey contai ned the highest energy content of available food items in both systems and the higher proportion of fish in the Homosassa River diets should have provided largemouth bass in the Homosassa River with gr eater energy acquisition rates than in the Chassahowitzka River resulting in higher growth which was not observed.
35 Life history theory could be used to explai n the observed growth patterns of largemouth bass in the Chassahowitzka and Homosassa Rivers. One of the fundamental tenets of life history theory is that age at maturity and level of repr oductive effort evolve in response to age-specific potential for growth and survival and that trade-offs among thes e parameters are optimized to maximize fitness (Hutchings 1993). An increase in reproductive inve stment by itself will lead to a lower somatic growth rate for adults (Leste r et al. 2004), and high re productive investment could be a viable strategy for dealing with a foraging environment that only provides low densities of smaller prey. However, the data co llected on GSI of female s did not support this hypothesis as GSI of mature females did not di ffer significantly between populations. Qualitative differences in the proportion of females with im mature or undeveloped ovaries were observed; however, these results lacked sta tistical power to test. Small sample size and lack of samples over the entire spawning period pr evented strong inferences on po ssible differences in allocation of energy towards reproduction. Differences in size specific mortality can l ead to differences in size structure and bias estimates of growth rate parameters (Parma a nd Deriso 1990, Taylor et al. 2005). Relatively high natural or fishing mortality can bias the growth parameters when fitting a von Bertalanffy growth curve, particularly by selectiv ely removing faster gr owing individuals from the population so that the sample of older fish consists mainly of slower-growing indivi duals (Biro et al. 2006). Populations experiencing high fish ing mortality are likely to have resultant growth curves which have a downward bias in the asymptotic length (Linf), upward bias in the growth coefficient ( k ), and an upward bias in the hypothetical age when fish total length is 0 (to) (Taylor et al. 2005). This trend in bias in growth parameters is the same as the trend that was observed between populations. In addition, high natura l mortality could have biased growth parameters and explain
36 the differences observed in size structure. Few individuals greater than 380 mm TL and no individuals greater than age 5 were sampled in th e Homosassa River. Lack of older individuals in the age-length sample could have biased the asymptotic length downw ard especially if the individuals that were sampled te nded to be intrinsically slowe r-growing individuals. Direct estimates of fishing mortality rates were not ava ilable and therefore this hypothesis could not be rejected; however, observations of very low fishing effort reduce the likelihood of this effect. Estimates of fishing and natural mortality rate s would help explain the patterns of growth observed. Comparisons of age-length data collected in 1985 with those colle cted in 2008 from the Homosassa River show that the asymptotic leng th of largemouth bass was lower than in 1985. This comparison provides circumstantial evidence that the current growth potential of largemouth bass in the Homosassa River is redu ced relative to past conditions. However, no growth data were available from the Chassahowit zka River in 1985, so tre nds within each river could not be compared over time. In addition, there is no accompanying plant data from the Homosassa River in 1985; however, data that do exist suggest that the SAV community in the Homosassa River in 1985 was much closer to th at currently found in the Chassahowtizka River. Frazer et al. (2006) documented the plant charac teristics of the Homosassa River since 1998 and show that a variety of SAV types were pres ent in 1998 and that the abundance of SAV has steadily declined since. While the differences in growth curves observed in the Homosassa River between 1985 and 2007 can not be directly linked to changes in the SAV community, the trend observed is consistent with the hypotheses outlined in Chapter 1. Significant differences in the size structure and growth rates between two populations of largemouth bass in this study suggest that fora ging efficiency may vary between populations.
37 Study systems were similar in respect to most phys ical and chemical characteristics as well as their overall community composition (Frazer et al. 2008). However, they varied drastically in respect to the abundance, and composition of S AV. Food intake measured by diet consumption and prey composition could not explain the di fferences in observed growth. Growth by largemouth bass in the Homosassa River was lowe r than in the Chassahowitzka River despite greater consumption. Differences in the physical environment between rivers likely influences foraging efficiency as both prey availability a nd prey vulnerability are higher in intermediate levels of structural complexity. In patchily di stributed vegetation, pred ators are provided areas from which to ambush prey as well as areas c ontaining high densities of prey. Thus, in such environments predators should be able to maxi mize net energy intake by selectively foraging on high utility prey, minimizing fora ging costs associated activity, and minimizing the time spent foraging. Bioenergetics models revealed that a strategy of minimizing foraging costs while forgoing some opportunities to acquire energy coul d lead to a greater fo raging efficiency. Diet data collected from the two rivers supports this hypothesis. While largemouth bass in the Chassahowitzka River were more likely to have empty stomach, they had lower diet breadths and consumed larger prey. Alte rnatively, largemouth bass in the Homosassa River had greater average mass of stomach contents but had a grea ter diet breadth and were consuming smaller prey. Higher attack rates and smaller sized prey li kely contributed to increased foraging costs in the Homosassa River. Additional knowledge of behavior patterns of largemouth bass in each river could reveal how foraging m ode or foraging effort contri butes to the efficiencies of largemouth bass in each river. These results sugg est that foraging in low structure environments may have greater costs relative to foraging in in termediate structure environments for some fish predators.
38 Table 2-1. Largemouth bass growth parameters for the Chassahowitzka and Homosassa Rivers (L = asymptotic length; k = growth coeffi cient; to = hypothetical age at which fish length = 0). River Chassahowitzka (2007-2008) Homosassa (2007-2008) Homosassa (1985) L (mm) 545.96 439.58 519.31 k 0.24 0.35 0.33 to (years) -0.63 -0.72 -0.73 Sample size 97 98 125
39Table 2-2. Diets of largemouth bass from the Chassahowitzka (Cha ) and Homosassa (Hom) Rivers, s howing the mean composition by weight of diet items, sample size, percent of empty stomach s, mean stomach content mass, mean consumption rate (g/g), mean number of fish prey, and mean length of fish prey for different body lengths. < 200 mm 200 to < 300 mm 300 to < 380 mm > 380 mm Diet Item Cha Hom Cha Hom Cha Hom Cha Hom Amphibian 0.00 0.00 1.50 0.00 0.00 0.00 24.45 0.00 Crayfish 4.75 4.60 34.56 5.57 25.09 10.35 6.00 0.00 Crustacean 0.00 1.71 8.87 5.96 3.58 16.35 2.65 8.80 Fish 84.01 79.69 44.86 81.88 63.53 64.95 59.11 88.77 Invertebrates 1.56 0.75 0.58 0.38 0.07 1.39 0.03 0.11 Shrimp 3.22 3.98 0.15 0.99 0.06 1.69 0.00 0.00 Plant/Algae/Detritus 6.15 8.93 8.60 4.68 6.88 5.25 7.75 2.32 Unknown 0.31 0.33 0.57 0.17 0.79 0.01 0.00 0.00 Sample Size 47 56 100 153 125 63 54 21 Percent Empty 0.26 0.18 0.23 0.11 0.24 0.14 0.17 0.21 Mean g/diet 0.054*0.202* 0.284 0.387 0.496 0.666 1.836 2.590 Mean g/g 0.00090.0014 0.00160.0014 0.00110.0019 0.00180.0021 Mean Number of Fish Prey 1.90 1.51 1.83 2.03 1.33* 2.28* 1.00 1.15 Mean Fish Prey Length 24.64 30.47 36.41 33.57 62.23* 44.03* 98.59 80.64 Significant difference at P < 0.05.
40Table 2-3. Percent composition by weight of fish prey for largemouth bass in the Chassahowitzka (Cha) and Homosassa (Hom) Rivers. < 200 mm 200 to < 300 mm 300 to < 380 mm > 380 mm Fish Prey Cha Hom Cha Hom Cha Hom Cha Hom Atlantic needlefish 0 0 0 0 0.04 0.01 0 0.03 Bluefin killifish 0.05 0.03 0.01 0.02 0 0.02 0 0 Coastal shiner 0 0.21 0 0.02 0 0.02 0 0 Golden shiner 0 0 0 0 0 0 0.22 0 Goby 0.01 0.15 0.04 0.05 0.02 0.02 0 0 Grey snapper 0 0 0 0 0.31 0 0.28 0.72 Inland silverside 0.05 0.08 0 0.02 0 0 0 0 Lake chubsucker 0 0 0 0 0 0 0.06 0 Spotted sunfish 0.63 0.11 0.14 0.08 0.26 0.20 0.36 0.21 Largemouth bass 0 0 0 0.12 0 0 0.04 0.03 Mojarra 0 0 0.10 0.27 0.11 0.57 0 0 Pinfish 0 0 0.40 0 0.14 0 0.01 0 Rainwater killifish 0.21 0.27 0.11 0.12 0.01 0.06 0 0 Sailfin molly 0 0 0.12 0 0 0.03 0 0 Seminal killifish 0 0 0 0.04 0.07 0 0 0 Toadfish 0 0 0 0.12 0 0.02 0 0 Unknown fish 0.05 0.15 0.07 0.13 0.02 0.05 0.03 0
41 Figure 2-1. Size structure of largemouth bass during summer (sum) and winter (win) sampling periods in the Chassahowitz ka (Cha) and Homosassa (Hom) Rivers. Boxes depict the lower quartile, median, and upper quartile of the sample; dotted lines represent the smallest and largest non-outlier obs ervations; dots represent outliers.
42 Figure 2-2. Observed total length (mm) vs. age (years) (circles) a nd predicted tota l length vs. age (lines) for largemouth bass in the Chassahow itzka (closed circles and solid line) and Homosassa Rivers (open circles and dashed line).
43 Figure 2-3. ln(body mass) vs. ln(gonad mass) fo r female largemouth bass in the Chassahowitzka (closed circles and solid line; y = 0.8 31x 2.267) and Homosassa Rivers (open circles and dashed line; y = 0.845x 2. 235). Squares represent females with undeveloped ovaries (closed = Chassahowitz ka River, open= Homosassa River).
44 Figure 2-4. Total length of predat or vs. total length of fish prey for largemouth bass in the Chassahowitzka (closed circ les and solid line; y = 0. 2879x 28.175) and Homosassa Rivers (open circles and dashed line; y = 0.1439x 0.8649).
45 Figure 2-5. Example of a diagnostic plot of parameter estimates for the four main variables of the bioenergetics model from the Chassahowitzka River.
46 Figure 2-6. Scenario 1 of bioenergetics mode l; estimates of energy acquisition (circles) and energy expenditure (squares) of largemout h bass from the Chassahowitzka (closed) and Homosassa (open) Rivers. Net energy in take is equal to the difference between energy acquisition (circles) and energy expenditure (squares).
47 Figure 2-7. Scenario 2 of bioenergetics mode l; estimates of energy acquisition (circles) and energy expenditure (squares) of largemout h bass from the Chassahowitzka (closed) and Homosassa (open) Rivers. Net energy in take is equal to the difference between energy acquisition (circles) and energy expenditure (squares).
48 CHAPTER 3 MOVEMENT AND HOME RANGE OF Micropterus salmoides IN TWO SYSTEMS WITH CONSTRASTING VEGETATIVE CHARACTERISTICS Introduction How often and far an animal m oves and how much space it uses are principal aspects of its ecology, and can be important indicators of re source requirements and av ailability (Plummer and Congdon 1994, Charland and Gregory 1995, Johnson 2002). An animals home range must be large enough to provide the key resources for its survival (primarily food and refuge) and reproduction. Many factors influence the area requ ired by an animal to obtain these resources, and these factors can vary among individuals within a species in a single population, and also among populations across a wider spa tial area. In any given environmen t, there is also a range of behaviors that allow predator s to obtain food (Grant and Noakes 1987, Tyler and Rose 1994). Intraspecific variation in movement behavior reflects di fferent tactics used by individuals within a species or population to meet energetic demands and cope with predation risks. Comparison of animal movements, foraging behavior, and home range size among populations may allow for delineation of the mechanisms which force animals to behave differently in contrasting environments. In aquatic ecosystems, the structure of the phys ical environment has been shown to impact the dynamics of predator-prey interactions (Crowder and C ooper 1979). Both prey abundance and richness are often positively correlated with structural complexity of habitats (Heck and Crowder 1991). In addition, capture success and prey vulnerability vary as a function of habitat complexity (Savino and Stein 1982, Gotceitas and Colgan 1989). Thus, predators should adjust their foraging behavior and fora ging effort depending on the comple xity of the local habitat in order to maximize foraging efficiency (Stephens and Krebs 1986). Observed foraging behavior represents the perceived trade-o ff in costs and benefits of obt aining resources. Variation in
49 foraging behavior and foraging effort have impor tant consequences for energy acquisition rates and space use requirements of predators and may vary between systems due to direct and indirect consequences associated with the structure of physi cal habitat. Loss of physical habitat assumed to be importa nt to fish is a common management issue for aquatic ecosystems worldwide (Dibble et al. 1996). In lotic systems, two key drivers of loss of physical habitat include nutrient enrichment a nd development of riparian shoreline. Nutrient enrichment often leads to increases in the bi omass of suspended and/or benthic algae and periphyton associated with submersed aquatic vegetation (SAV) (Duarte 1995, Smith et al. 1999). These increases can cause su bstantial changes to the structural characteristics of rivers and streams, leading to nearly complete elimination of SAV (W etzel and Likens 1979, Frazer et al. 2006). Also, human development of riparian s horeline is directly a ssociated with loss of structural complexity in the form of large woody debris (LWD) (Chris tiansen et al. 1996). Declines in both LWD and SAV have the potential to alter the behaviors and space use patterns of predatory fishes. In this study, movement patterns and home range sizes of largemouth bass ( Micropterus salmoides) were examined in two rivers that vary ma rkedly in respect to their physical habitat characteristics, particularly distribution and abundance of SAV. The Chassahowitzka River has relatively high SAV abundance, while the Homo sassa River is nearly devoid of rooted macrophytes (Chapter 1). Acoustic telemetry ta gs and an autonomous receiver array were utilized to track the movements of largemouth bass in each system. Movement rates, movement distances, and activity time were compared between populations in the two river systems to draw inferences about foraging beha vior and foraging effort. Kernel density home range estimates were used to compare differences in overal l space use requirement s between populations. I
50 hypothesized that largemouth bass would displa y higher activity and movement patterns and have greater space use requirements in the system where structural habitat, as measured by abundance of SAV, has been markedly reduced. Methods Study Systems Two spring-fed rivers systems along Floridas southwest Gulf coast, the Chassahowitzka and Homosassa Rivers, were selected for this pr oject. These two systems are similar with respect to their physical (temperate, depth, substrate) and chemical (nutri ents, salinity) characteristics, but the submerged aquatic vegetation (SAV) communities are markedly different (Frazer et al. 2006). The Chassahowitzka River is composed of a patchily distributed heterogeneous SAV community, while the Homosassa River is nearly devoid of rooted macrophytes and macroalgal abundance is comparatively sparse. A detailed desc ription of these rivers is provided in Chapter 1. Telemetry A total of 20 adult largemouth bass from each river were surgically implanted with individually coded V-7 and V-13 Vemco acoustic telemetry tags (V-7: 7-mm diameter, 21-mm length, 1.0-grams, 140-day battery life; V-13: 13-mm diameter, 40-mm length, 6 grams, 240-day battery life; Vemco Ltd., Shad Bay, Nova Sco tia, Canada). All largemouth bass were collected using boat electrofishing (Sm ith-Root Inc.; Mark IX GPP unit pulsed-DC; 20-30 A). All largemouth bass collected were measured (total le ngth, mm), weighed (g), and each fish selected for surgery was anesthetized with sodium bicarbona te. All fish selected for tagging were greater than 250-mm TL. Fish of this size were considered to be mature adults (Cha pter 2). No fish were implanted with a tag greater than 1.25% of its body weight to minimize the influence on behavior and movement (Winter 1996). Tags were implanted into the peri toneal cavity of all
51 fish. Abdominal incisions were cl osed with 3-4 absorbable suture s and ethyl cyanoacrylate. Each tagged fish was observed until the effects of anesthes ia ceased, fish were able to keep themselves upright and regained normal fin mo vement, and then released into the same area from which they were collected. A staggered-entry tagging design was used with ta gs implanted in small batches beginning in January 2007 through November 2007 in order to maximize the temporal resolution of the data (Pollock et al. 1989). Each acoustic tag outputs a uniquely identif iable, pseudo-random pulse every 40 to 80 seconds (mean = 60 seconds) at a 69.0 kHz fre quency. Signals were detected by Vemco VR2 acoustic receivers (Vemco Ltd.) which record the unique tag number al ong with a date-time stamp. VR2 receivers were pla ced strategically throughout the ri vers for continuous monitoring of largemouth bass. Each VR2 receiver was anchor ed in a fixed location within each river. Remote monitoring using autonomous receiv ers such as VR2s creates a reception environment conducive for a high proportion of fish activity to be recorded within each river. When tagged fish swim within the detection ra nge of the receiver (approximately 150-m radius; Tetzlaff, unpublished data), the re ceiver records the unique tag id, date, and time. Receivers were able to provide presence-absence da ta within set detection zones al ong the river, at time scales of minutes, for time periods of months. The two rivers were divided into observable and unobservable zones based on receiver location and detection range (Figure 2 and 3). Detection range varied across receivers due to depth of water, physical stru cture within the water column, and substrate type. The detection range of the receivers allowed fo r coverage of the entire width of the rivers in most locations (Figures 2 and 3) Having detection ranges as large as the width of the river ensured that a high proportion of upstream-downstream movement was detected because the receivers acted as ga tes to monitor fish movement into and out of larger river
52 reaches. On average, there were 8 receivers ac tive in each river most months. Data from all receivers were manually retrie ved each month during which ti me the receivers and anchor materials were checked and cleaned to assure proper performance. Receiver battery changes occurred on yearly intervals to preven t loss of acoustic coverage and data. Movement Patterns Movement patterns were analyzed using data collected from VR2 receivers. Only individuals that were at large for at least 30 days and with at least 3000 observations (VR2 identifications) were used for analysis. Data from the first week following implantation was not used in analyses, so as to allow a recovery pe riod and avoid potential effects of capture and surgery. Activity time, daily minimum displacement, and home ranges were calculated for each tagged fish in the two rivers. Activity rate was calculate d as the proportion of hours in which a fish was observed moving between multiple receivers relative to the total number of hours that the fish was observed within study site. Movement (meter s/day) was estimated as the daily minimum displacement. Displacement distan ces were calculated as the di stances between two receivers. A non-zero movement distance was assigned each time an individual fish was recorded as moving from one receiver to another. C onsecutive observations recorded within 40 seconds when a fish was within the detection zone of two receiver s simultaneously were removed when determining movement distances. Time to Independence If a fish was located within a detection zone (F igures 3-2 and 3-3), this fish was potentially detected every 40-80 seconds leading to a high degr ee of autocorrelation in the data. The Time to Independence (TTI) (Swihart and Slade 1985) wa s calculated for tagged fish to determine appropriate sampling intervals to subset the tota l records of fish observations for determining
53 kernel density home ranges. The TTI examines the way that the distance between location records changes with sampling interv als using Schoeners ratio statistic 2 2 r tV (3-1) where t2 is the mean squared distance between consecutive location records and r2 is the mean squared distance from each location to the range center (the arithmetic mean of all coordinates). TTI between locations is indicated when the firs t of three successive time intervals exceed V = 2. This is roughly equivalent to the time required for an animal to traverse its entire home range (Swihart and Slade 1985). Home Range Estimation Kernel density home ranges were calculated for individuals with sufficient relocation data (N=36, 18 in each river). Data collected from autonomous receivers were used to assign relocation points. At present, there are no gene rally accepted criteria for using data from autonomous receivers to determine kernel density home range estimates. There are two main analytical challenges when using data collect ed from autonomous receivers for determining home range. First, tags output a uniquely identifiable signal every 40 80 seconds, typically resulting in extreme autocorrelati on of data due to consecutive re locations of individual fish on individual receivers. Second, each record of fi sh presence does not represent an exact GPS location. Instead, each observation could have oc curred anywhere within a receivers detection radius. Thus, a set of a priori rules was developed to aid in filtering data and interpreting relocations of telemetered fish within the array. These rules aided in reducing autocorrelation and provided the basis for data compilation us ed in the analyses described below. Two sets of data were used to calculate kern el density estimates in order to evaluate the effects of autocorrelation: (1) a subset of all data (Table 3-3) and (2) a subset of observations at
54 constant time intervals according to the TTI (Table 3-4). In addition to th e data collected from the VR2 receivers, unobserved estimated positions were incorporated into home range estimates in order to generate more rea listic utilization distributions. Unobserved position estimates were added as follows: (1) all data were sorted in chronological order for individual fish, (2) river reaches were divided into obs ervable and unobservable zones (F igure 3-2 and 3-3), (3) when consecutive observations were recorded on differe nt receivers, unobserved positions equal to the proportion of time spent between detection zones we re assigned to the re gion in between the two detection zones. Detection probability was eval uated by sorting all dete ctions in chronological order and determining the proportion of hits whic h occurred in non-consecutive detection zones. For each of the data sets, the number of relo cation points exceeded the number of data points necessary to determine accurate kernel home ranges (Seaman et al. 1999). Thus, for each data set, the total data was scaled down to facilitate the determination of kernel home range estimates. The number of observations for each fish was weighted by summing the total number of observations from all fish, and then determin ing the proportion of tota l observations for which each fish was responsible. Between 30 and 300 observations were assigned to each fish, weighted by the individuals re lative number of observations (T able 3-3). Thirty observations were used as a minimum based on literature recommendations for kernel-based home range methods (Seaman et al. 1999). Three hundred obs ervations were used as a maximum as observations beyond 300 would not have substantia lly improved the fit of the kernel estimate (Seaman et al. 1999). The proportion of observati ons which occurred within each observable and unobservable zone was then multiplied by the numb er of observations assigned to each fish. Observation points for each zone were assigne d X and Y coordinates by randomly distributing
55 relocation points within each zone using Hawt hs Analysis Tools in ArcGIS (Beyer 2004, http://www.spatialecology.com/htools). Kernel density estimates were calculate d following Worton (1989) where the throughwater distances between all obser vation points and all re ferences points were estimated using a raster-based GIS approach desc ribed by Jensen et al. (2006). With this approach, a grid resolution of 10 m was used for reference locations and then these locations were overlain with the observation points. Calculations of through-water distances allo wed for incorporation of land boundaries between points and ther efore obtain unbiased estimate s of home ranges within the rivers. The kernel density smoothing parameter (h) was calculated for each individual fish using the least squares cross validation approach (S eaman et al. 1999). Smoothing parameters were calculated using the animal movement extens ion in ArcView 3.1 (Hooge and Eichenlaub 2000). Results A total of 36 largemouth bass were used for m ovement and home range analysis. Fish were monitored for 38 to 240 days with a total of 2,237 and 2,303 cumulative days recorded on the Chassahowitzka and Homosassa Rivers, respective ly (Table 3-1). Observ ations ranged from 501 to 267,296 per individual (Table 3-1). Mean relo cations of individual fish were 46,871 in the Chassahowitzka and 66,373 in the Homosassa. Th e proportion of hours in which a fish was observed relative to total number of hours at large ranged from 14.6 to 92.9% in the Chassahowitzka River and 18.1 to 77.5% in the Homosassa River. On average, fish were observed during 52.3 and 55.9% of the total hour s in which they were available in the Chassahowitzka and Homosassa Rivers, respectively. In general, largemouth bass displayed crepuscu lar activity patterns. In both rivers, fish activity peaked at 07:00 with a smaller pe ak between 19:00 and 20:00 and limited movement during mid-day hours (Figure 3-3). The proportion of hours in which a fish was observed moving
56 among multiple receivers was significantly greater in the Homosassa River ( t = -2.84, df = 17, P < 0.01). On average, largemouth bass in the Homo sassa River were active 24.2% of all hours observed, while largemouth bass in the Chassaho witzka River were active 9.5% of all hours observed. Most fish displayed similar pattern in their movements with respect to time, however, there was high individual variation in the fr equency of movements. Mean minimum daily displacement rates were highly va riable in both populations, but on average were significantly higher in the Homosassa River (mean = 24953 meters +/7213 SD) compared to the Chassahowitzka River (mean = 9583 meters +/2722 SD) ( t = -2.98, df = 17, p < 0.01) (Table 32). Home range sizes of largemouth bass were significantly larger in the Homosassa River than in the Chassahowitzka Rive r for all comparisons. When using all data (Table 3-3), 90% kernel home range estimates ranged from 2.83 to 13.58 ha in the Homosassa River and 1.49 to 5.08 ha in the Chassahowitzka River. For the same data, 50% kernel home range estimates ranged from 1.62 to 4.95 in the Homosassa River and 0.78 to 2.23 in the Chassahowitzka River. Both 90% and 50% kernel home ranges were signif icantly larger in the Homosassa River (mean = 7.48 ha +/3.25 SD and 2.90 ha +/1.19 SD, re spectively) compared to the Chassahowitzka River (mean = 3.05 ha +/1.16 SD and 1.28 ha +/0.34 SD, respectively) ( t = -4.08, df = 17, p < 0.001, t = -5.87, df = 17, p < 0.001) (Figure 3-4). Home ra nge size was positively correlated with body length in both rivers (Figure 3-5). The autonomous receiver data were highly autocorrelated. The average TTI was highly variable among individual largemouth bass in both rivers (Figure 36 and Figure 3-7). The average Schoeners index peaked when the sa mpling interval was between 12 and 16 hours in both rivers, with a maximum at 15 hours. However, for a number of individuals the observations
57 were not independent regardless of the sampli ng interval. For the Chassahowtizka River, 90% and 50% kernel home range estimates were sma ller when using only observations separated by 15 hours compared to when using all data (mean = 2.97 ha +/1.33 SD and 1.24 ha +/0.43 SD, respectively) (90%: t = 1.17, df = 17, p = 0.13, 50%: t = 1.50, df = 17, p = 0.08). For the Homosassa River, 90% and 50% kernel home range estimates were slightly larger, but not significantly different when using observations separated by 15 hours compared to when using all data (mean = 7.51 ha +/3.47 SD and 2.92 ha +/1.31 SD, respectively) (90%: t = 0.29, df = 17, p = 0.39, 50%: t = 0.41, df = 17, p = 0.34). The relationship between sample size and kernel home range estimate differed between rivers (Fig ure 3-8). In the Chassahowitzka River, home range size was negatively correlated with sample size. In the Homosassa River, home range size was positively correlated with sample size. Discussion Ecological Findings Activity, movement, and home range size of largemouth bass varied greatly between the two river systems. Observations of higher m ovement rates and larger space use by largemouth bass in the Homosassa River follow predictions in Ch apter 1 that structural complexity of habitat strongly influences the foraging beha vior of fish predators. This analysis intended to characterize the movement and space use patterns of largemouth in contrasting systems rather than correlate movement behaviors with specific abiotic or biotic factors. However, given the similarities in the physical and chemical characteristics as well as in the overall community composition of these systems (Frazer et al. 2008), it is likely that th e differences in physical habitat played a major role influencing the pa tterns of movement and space use observed. The physical habitat template of aquatic syst ems has been shown to have strong effects on the foraging behaviors (Savino a nd Stein 1989), movement rates (Harvey et al. 1999), and spatial
58 distribution (Schuerell and Schindl er 2004) of fishes. In general, fish tend to exhibit lower movement rates and higher spatial aggregati on in structurally complex habitats. My study supports this relationship, with largemouth bass in the Homosassa River displaying a greater degree of movement and larger home range sizes than largemouth bass in the Chassahowitzka River. A number of plausible mechanisms might explain the observed differences in behavior such as differences in prey abundance, compositi on, and vulnerability, or differences in foraging strategy. Most of these mechanisms revolve around the role of structural complexity in predatorprey interactions (Ch2). Movement and home range are important aspect s of an animals feed ing strategy and thus movement rate and home range size should reflect the relative profitabil ity of the habitat in which a predator resides. Lower relative foraging profitability resulting from differences in prey availability or prey size could force largemouth bass in the Homosa ssa River to traverse greater distances in order to obtain sufficient food to meet energetic demands. The greater movement rates observed may be the result of having to comp ensate for reduced availa bility of resources by more intensively searching for food (Huey a nd Pianka 1981). Helfman (1990) predicted that ectotherms such as fishes should switch from am bush to active foraging modes as prey density decreases in order to maintain a minimum enco unter rate. Fausch et al. (1997) provided support for this hypothesis; by experiment ally reducing prey availability they were able to induce a foraging mode switch in char from an am bush to an actively foraging mode. Differences in movement and space use ma y also reflect two di fferent, but equally profitable, foraging strategies given differences in the physical environment between rivers. Largemouth bass in experimental systems have been shown to change predation tactics in response to decreases in habita t complexity, switching from an ambush foraging mode to an
59 actively foraging mode (Savino and Stein 1982). Active foraging increases the encounter rate between predator and prey, but is associated w ith increased metabolic cost and predation risk (Huey and Pianka 1981). Ambush predation often resu lts in lower encounter rates with prey, but also lower metabolic cost and predation risk. Th e behavior and space use patterns observed and reported herein may represent the best strate gy for the given suite of ecological conditions present in each river. However, increased m ovement and space use by largemouth bass in the Homosassa River likely results in greater foraging costs; both physiological costs and increased predation risk. Observed differences in the gr owth rates of largemouth bass between rivers (Chapter 2) are likely influen ced by differences in the movement and activity of largemouth bass between rivers. Methodological Findings Continuous monitoring of activity through time via autonomous receivers can provide powerful insight into animal behavior and ener getic requirements withou t the large investments in time or resources usually associated with ma nual tracking of fishes. However, the scale of inferences that can be made from remote mon itoring depends on the type of data collected. The omni-directional receiver array utilized in this study could only provide presence-absence data. Simpfendorfer et al. (2002) used a mean-positi on algorithm to determine fish position using a grided array of VR2 receivers. The linear design of the receiv ers in this study precluded the determination of X and Y coordinates for each ob servation; thus, the X and Y coordinates of each VR2 receiver were assigned to each observation when determining daily displacement. Lack of exact location information reduced the le vel of inference to broad scale space use and introduced bias into the determination of movement rates. Displacement rates calculated in this study we re prone to two sources of bias. First, the majority of a fishs activity occurs on a spatial scale that is less than th e detection radius of a
60 single receiver. Movement distan ces were only assigned when cons ecutive observations occurred on different receivers. All movements that occu r on a fine scale can no t be quantified and are therefore underestimated. Activit y and movement that occurs at a spatial scale too small for detection by VR2 receivers likely represents a significant portion of daily activity and movement (Demers et al. 1996). Second, consecutive observa tions on different receivers were assigned a movement distance equal to the distance between th e centers of each receiver. As a consequence, movements occurring around the periphery of two receivers were overestimated. Animals which occupied space in areas where de tection zones of receivers were close were more likely to have overestimated movement rates. Largemouth bass in the Chassahowitzka River were more likely to have movement overes timated because receivers were closer together. Despite this bias, increased movement rates were observed in the Homosassa River compared to the Chassahowitzka River. Movement rates report ed in this study are difficult to compare to other movement studies on largemouth bass, or ri verine fish in general as they are likely overestimated; however, the comparisons betwee n systems in this study should hold true. Fixed kernel density estimates were used to estimate the space use patterns of largemouth bass in the Chassahowitzka and Homosassa Rivers. For lotic systems, the home range of fishes has traditionally been expressed as the range span of the fish. Ho wever, this method often places undue emphasis on outlying relocat ions and does not allow for i nvestigation of the internal structure of the home range (Kernohan et al. 2001). Kernel methods have become the method of choice for describing the space use of animals (Marzluff et al. 2004) and allowed for more accurate comparisons of space use patterns by la rgemouth bass between the two study systems. Using core areas in which a fish was estimated to spend 50% of its time rather than the absolute distance between the most upstream and downstr eam relocation offers a more biologically
61 significant comparison of space use and more accur ate estimate of an animals home range (Seaman et al. 1999). Kernel home range estimates were determ ined based on the proportion of time spent in different reaches throughout the ri ver. Utilization distributions of kernel density estimates are usually interpreted as the proportion of time spen t in a given area over a specified time period (Seaman and Powell 1996). The general approach us ed in this study of assigning locations in different reaches based on the proportion of to tal observations follows this definition. The addition of points in unobservable reaches was meant to generate realistic distributions of observations throughout the river. When relocations were only assigned to detection zones, the utilization distributions generated often consisted of multiple modes of use centered on the VR2 receivers even when high bandwidths were used. Given that the detecti on ranges of receivers covered the entire width of th e river in all but one location, and the detection probability exceeded 99% for both systems, assigning locations in reaches where fish were unobserved seemed appropriate and facil itated interpretation of home range size and shape. The sample size used when determining ho me ranges varied between individuals. For most home range studies sample size varies between individuals, however, observations of individuals in this study varied by a factor of 80. In additi on, the number of observations recorded for all individuals exceeded a realistic sample size which could be used for home range analysis. Data collected were scaled down by assigning between 30 and 300 observations to each animal relative to the total number of relocatio ns observed for each animal. These values were chosen following the recommendations of Seaman et al. (1999), who used computer simulations to determine the effect of sample size on the accu racy of home range estimates. They determined that a minimum of 30 observations was require d for accurate home ranges and that smaller
62 sample sizes overestimated home range size. In a ddition, they concluded that bias and variance of home range size approached an asymptote at about 100 observations. Th e receiver array used in this study allowed for a high proportion of fish activity to be observed, with a number of individuals on record for greater than 80% of the time they were at large. For animals which were recorded for a majority of time at larg e, between 100 and 300 locations were assigned to ensure accurate estimates. While the range in the data collected exceeded the range in observations used for determining home range, in cluding more than 300 observations would not have improved home range estimates (Seaman et al. 1999). GPS coordinates were randomly assigned to observations used for determining home range estimates. The exact location of each observation was unknown; instead of assigning each observation the same X and Y coordinate, observ ations were randomly distributed within each assigned zones. Randomly assigning X and Y coor dinates to each observation given a detection zone, allowed for creation of realistic utilization distributions. However, th e resulting utilization distribution was not used to make inferences be yond the scale of the data collected. Utilization distributions estimate the intensity or probabi lity of use by animal throughout its home range (Kernohan 2001). Using the utilizat ion distribution to determine resource selection for animals has become a widely used approach (. Howeve r, since observations we re randomly distributed within each zone, only inferences on space use at the scale of river reaches was appropriate for this study. This method allowed for objective delin eation of heavily used reaches or sections of the river, but not for specific habitat utilization. Use of autocorrelated observations did not sign ificantly bias kernel home range estimates. There was no pattern of bias when calculating kernel home range estimates using autocorrelated data compared to kernel home range estimates calculated using observations separated by 15
63 hours. Use of independent observations increase d the kernel home range estimate for some individuals and reduced estimates for other individuals relative to estimates calculated using autocorrelated data. While use of independent ob servations for kernel home range estimates has traditionally been emphasized (Swihart and Slad e 1985), more recent investigation into the use of autocorrelated data sets suggests that use of independent observations is not required when estimating home range (de Solla et al. 1999). In this study, data collected were highly autocorrelated, however, the general location of each animal was known for a majority of the time in which they were available. Thus, use of autocorrelated data in this case allowed for an accurate estimate of home range size. Conclusion The comparison of activity, movement and hom e ranges of largemouth bass in two coastal river systems revealed definitive differences in behavior and space use between two populations which reside in environments offering contras ting levels of physical habitat. Largemouth bass spent more time actively moving, traversed greater distances, and occupied larger home ranges in the river with lower structur al habitat complexity. Differences in activity and space use likely stem from differences in the availability and vu lnerability of prey resources between systems. Largemouth bass in the Homosassa River appe ar to compensate for depressed resource availability by increasing foragi ng effort and occupying larger home ranges. While fish predators should alter their foraging tactics in order to meet energetic requirements in varying environments, alternative foraging modes likely vary in the fitness potential they offer. Increased activity and space use has both energetic costs as well as higher predation risks. A greater understanding of the energetic trade-offs asso ciated with foraging is required to provide inferences on the ultimate outcome of the behavi oral differences demons trated in this study.
64 Table 3-1. Summary of largemouth bass tagge d within the Chassahowitzka and Homosassa Rivers; tag identification, total length (mm), date ta gged, date of last record days at large (DAL), days on record (DOR), number of obse rvations, and fate. River TagID Length First obs. Last obs. DAL DOR Total obs. Fate Chassahowitzka 2094 539 1/19/2007 9/15/2007 239 193 49188 Tag expired Chassahowitzka 2095 468 1/19/2007 3/20/2007 60 48 3519 Unknown Chassahowitzka 2096 454 1/19/2007 4/19/2007 90 76 7190 Mortality Chassahowitzka 2097 350 1/19/2007 9/16/2007 240 236 106136 Tag expired Chassahowitzka 2099 441 1/19/2007 3/07/2007 47 30 13664 Unknown Chassahowitzka 2098 370 1/19/2007 9/16/2007 240 149 75183 Tag expired Chassahowitzka 2100 385 1/19/2007 8/15/2007 208 136 109472 Tag expired Chassahowitzka 2101 402 1/19/2007 9/16/2007 240 177 52719 Tag expired Chassahowitzka 3008 315 8/01/2007 9/24/2007 54 36 3375 Unknown Chassahowitzka 3055 410 8/01/2007 3/04/2008 216 209 103435 Tag expired Chassahowitzka 5812 349 8/01/2007 12/08/2007 129 124 15394 Tag expired Chassahowitzka 5813 336 8/01/2007 11/20/2007 111 112 68112 Tag expired Chassahowitzka 5818 252 8/01/2007 12/07/2007 128 28 501 Left study site Chassahowitzka 3056 400 9/10/2007 1/21/2008 133 132 98958 Unknown Chassahowitzka 5822 405 9/10/2007 10/22/2007 42 42 32423 Unknown Chassahowitzka 5827 420 9/10/2007 2/11/2008 154 155 110647 Unknown Chassahowitzka 5830 378 9/10/2007 3/11/2008 183 118 53354 Tag expired Chassahowitzka 8781 397 11/05/2007 3/23/2008 139 121 13294 Tag expired Chassahowitzka 8782 374 11/05/2007 11/18/2007 13 13 2453 Unknown Chassahowitzka 8785 360 11/05/2007 3/23/2008 139 138 21419 Tag expired Homosassa 3058 403 3/06/2007 10/12/2007 221 192 115847 Tag expired Homosassa 3059 375 3/06/2007 10/12/2007 220 219 42221 Tag expired Homosassa 3060 427 3/06/2007 4/13/2007 38 32 11256 Unknown Homosassa 3061 321 3/06/2007 3/16/2007 11 10 2981 Unknown Homosassa 3062 325 3/06/2007 5/04/2007 60 52 29644 Unknown Homosassa 2092 357 6/15/2007 12/27/2007 196 146 10101 Tag expired Homosassa 3009 320 6/15/2007 6/17/2007 2 3 1375 Mortality Homosassa 2093 340 6/15/2007 11/01/2007 139 140 77889 Tag expired Homosassa 3047 380 6/15/2007 10/15/2007 122 116 10758 Fish removed Homosassa 3049 327 6/15/2007 9/17/2007 94 94 15550 Unknown Homosassa 3057 534 6/15/2007 1/21/2008 220 222 267296 Tag expired Homosassa 5823 374 9/17/2007 3/28/2008 193 184 155016 Tag expired Homosassa 5824 338 9/17/2007 3/30/2008 195 196 245893 Tag expired Homosassa 5825 435 9/17/2007 1/29/2008 134 131 68094 Unknown Homosassa 5826 426 9/17/2007 3/29/2008 194 180 74665 Tag expired Homosassa 5828 387 9/17/2007 2/12/2008 149 120 42331 Unknown Homosassa 5829 437 9/17/2007 1/30/2008 136 132 108680 Unknown Homosassa 5831 402 9/17/2007 11/04/2007 48 37 19329 Tag malfunction Homosassa 8779 312 11/12/2007 3/12/2008 121 74 16472 Unknown Homosassa 8788 362 11/12/2007 3/21/2008 130 64 12069 Tag expired Denotes fish removed from data analysis
65 Table 3-2. Summary of fish movement fo r largemouth bass in the Chassahowitzka and Homosassa Rivers; total movement recorded (m), mean daily movement (m/d) and standard deviation. River Tag ID Total movement (m) Mean da ily displacement (m/d) Standard deviation Chassahowtizka 2094 399145 4338 7616 Chassahowtizka 2095 59155 1739 1663 Chassahowtizka 2096 19575 797 887 Chassahowtizka 2097 1359480 12587 15630 Chassahowtizka 2098 292680 3658 6217 Chassahowtizka 2099 29580 1368 1560 Chassahowtizka 2100 217200 6033 7476 Chassahowtizka 2101 261835 3154 5522 Chassahowtizka 3008 16801 1133 518 Chassahowtizka 3055 3882930 20764 26148 Chassahowtizka 3056 1513125 27511 31165 Chassahowtizka 5812 98100 1114 3884 Chassahowtizka 5813 31860 838 691 Chassahowtizka 5822 656320 16007 14471 Chassahowtizka 5827 816480 21486 31919 Chassahowtizka 5830 4656901 42723 50072 Chassahowtizka 8781 94694 2705 6211 Chassahowtizka 8785 140573 4534 4920 Homosassa 2092 17275 808 342 Homosassa 2093 94985 2499 2781 Homosassa 3047 24145 561 335 Homosassa 3049 121005 2333 1790 Homosassa 3057 10761875 48917 50471 Homosassa 3058 3169280 22800 36753 Homosassa 3059 1168915 15796 24248 Homosassa 3060 50620 3540 2546 Homosassa 3062 29895 2299 2231 Homosassa 5823 11410015 66725 57519 Homosassa 5824 15962255 8227 49787 Homosassa 5825 2831985 25285 22446 Homosassa 5826 3151025 20070 23057 Homosassa 5828 1721665 17934 20340 Homosassa 5829 4452650 34786 34059 Homosassa 5831 1812570 100698 53454 Homosassa 8779 38690 1137 1250 Homosassa 8788 15520 690 227
66Table 3-3. Summary of observations used fo r determining kernel density home range estimates; the number of random observations placed within each river zone for each individual. Columns with multiple numb ers represent transition zones between receivers which were unobserved by VR2 receivers. Observable and unobservable zones River Tag ID 0 1 1 2 2 2 3 3 2 4 3 4 4 4 5 5 5 6 6 6 7 7 5 8 6 8 7 8 8 8 9 9 Total Chassahowitzka 2094 0 0 0 1 0 0 0 0 0 0 12 3 53 3 19 0 2 0 1 0 0 94 Chassahowitzka 2095 0 0 0 9 0 0 0 0 0 0 21 2 1 0 0 0 0 0 0 0 0 33 Chassahowitzka 2096 0 0 0 25 0 0 0 0 0 0 6 1 1 0 0 0 0 0 0 0 0 33 Chassahowitzka 2097 0 0 0 0 0 0 0 0 0 0 12 2 136 14 29 0 2 0 1 0 0 196 Chassahowitzka 2098 0 0 0 8 0 0 2 0 115 2 2 0 0 0 0 0 0 0 0 0 0 129 Chassahowitzka 2099 0 0 0 1 0 0 0 0 0 0 1 2 29 0 0 0 2 0 1 0 0 36 Chassahowitzka 2100 0 0 0 0 0 0 0 0 0 0 1 2 175 3 10 0 2 0 1 0 0 194 Chassahowitzka 2101 0 1 2 3 1 1 2 1 70 2 14 2 1 2 1 0 2 0 1 0 0 106 Chassahowitzka 3008 0 2 2 12 0 0 2 0 1 2 1 2 3 2 11 0 0 0 0 0 0 40 Chassahowitzka 3055 1 0 2 2 0 0 1 0 1 1 2 2 49 33 106 0 2 2 48 2 1 255 Chassahowitzka 3056 0 0 0 12 0 0 12 0 155 0 0 0 0 0 0 0 0 0 0 0 0 179 Chassahowitzka 5812 0 0 0 0 0 0 0 0 0 0 25 2 4 2 2 0 0 0 0 0 0 35 Chassahowitzka 5813 0 0 0 0 0 0 0 0 0 0 107 2 2 0 0 0 0 0 0 0 1 112 Chassahowitzka 5822 0 0 0 1 0 0 2 0 19 5 28 2 1 0 1 0 0 0 1 0 0 60 Chassahowitzka 5827 0 0 0 0 0 0 0 0 0 0 7 9 286 2 1 0 0 0 0 0 0 305 Chassahowitzka 5830 0 1 2 2 0 0 2 0 2 2 27 21 66 21 62 0 2 2 2 2 1 217 Chassahowitzka 8781 0 1 2 44 0 0 2 0 2 2 1 2 1 2 1 0 1 0 1 0 1 63 Chassahowitzka 8785 1 10 4 89 0 0 2 0 1 2 3 1 1 2 1 0 0 0 0 0 0 117 Homosassa 2092 0 0 0 31 2 1 2 0 1 0 0 0 0 0 0 0 0 0 0 0 0 37 Homosassa 2093 0 0 0 1 0 0 1 0 16 2 219 0 0 0 0 0 0 0 0 0 0 239 Homosassa 3047 0 0 0 0 0 0 0 0 5 2 29 0 0 0 0 0 0 0 1 0 0 37 Homosassa 3049 0 1 1 33 7 14 2 0 1 0 0 0 0 0 0 0 0 0 0 0 0 59 Homosassa 3057 0 6 4 118 20 57 15 1 44 2 13 2 2 2 11 2 0 2 7 2 2 312 Homosassa 3058 0 0 0 0 0 0 0 0 5 2 2 2 20 5 17 6 0 6 236 3 3 307 Homosassa 3059 0 95 10 16 0 0 2 0 16 2 1 0 0 0 0 0 0 0 0 0 0 142 Homosassa 3060 0 1 0 0 0 0 0 0 1 0 0 0 33 0 0 1 0 0 1 0 0 37 Homosassa 3062 0 2 0 1 0 2 2 1 85 2 2 0 0 0 1 0 0 1 1 1 1 102 Homosassa 5823 1 98 37 117 0 0 8 0 15 2 9 2 3 2 2 2 0 2 7 2 6 315 Homosassa 5824 0 1 1 1 0 0 2 0 1 2 5 2 8 2 135 5 0 55 93 2 1 316 Homosassa 5825 0 99 24 103 0 0 2 0 1 2 1 0 0 0 1 0 0 2 1 2 3 241 Homosassa 5826 0 0 0 0 0 0 0 0 1 2 16 2 5 2 49 2 0 20 155 2 1 257
67Table 3-3. Continued Observable and unobservable zones River Tag ID 0 1 1 2 2 2 3 3 2 4 3 4 4 4 5 5 5 6 6 6 7 7 5 8 6 8 7 8 8 8 9 9 Total Homosassa 5828 0 1 1 3 0 0 2 0 2 2 12 1 1 2 28 2 0 10 84 2 1 154 Homosassa 5829 0 0 0 1 0 0 2 0 137 26 82 2 3 2 25 2 0 6 19 1 1 309 Homosassa 5831 0 19 16 40 0 0 2 0 1 0 0 0 0 0 0 0 0 0 0 0 0 78 Homosassa 8779 0 0 0 1 0 0 2 0 4 2 45 0 0 0 2 0 0 2 1 0 0 59 Homosassa 8788 0 1 1 1 0 0 2 0 36 2 1 0 0 0 0 0 0 0 0 0 0 44
68Table 3-4. Summary of independe nt observations used for determining kernel de nsity home range estimat es; the number of random observations placed within each river zone for each individua l. Columns with multiple numbers represent transition zones between receivers which were unobserved by VR2 receivers. Observable and unobservable zones River Tag ID 0 1 1 2 2 2 3 3 2 4 3 4 4 4 5 5 5 6 6 6 7 7 5 8 6 8 7 8 8 8 9 9 Total Chassahowitzka 2094 0 0 0 0 0 0 0 0 0 0 11 2 146 10 44 0 0 1 1 0 0 216 Chassahowitzka 2095 0 0 0 30 0 0 0 0 0 0 12 0 0 0 0 0 0 0 0 0 0 42 Chassahowitzka 2096 0 0 0 60 0 0 0 0 0 0 10 0 0 0 0 0 0 0 0 0 0 70 Chassahowitzka 2097 0 0 0 0 0 0 0 0 0 0 15 2 210 13 35 0 0 0 0 0 0 276 Chassahowitzka 2098 0 0 0 45 0 0 5 0 123 2 5 0 0 0 0 0 0 0 0 0 0 180 Chassahowitzka 2099 0 0 0 1 0 0 0 0 0 0 1 1 26 0 0 0 1 0 1 0 0 30 Chassahowitzka 2100 0 0 0 0 0 0 0 0 0 0 1 1 140 11 12 0 0 0 0 0 0 166 Chassahowitzka 2101 0 0 0 50 0 0 23 0 123 7 21 0 0 0 0 0 0 0 0 0 0 223 Chassahowitzka 3008 0 14 4 7 0 0 0 0 0 0 0 0 1 1 4 0 0 0 0 0 0 30 Chassahowitzka 3055 1 0 0 2 0 0 0 0 0 0 1 1 25 18 102 0 4 30 116 1 1 300 Chassahowitzka 3056 0 0 0 5 0 0 2 0 157 0 0 0 0 0 0 0 0 0 0 0 0 165 Chassahowitzka 5812 0 0 0 0 0 0 0 0 0 0 102 9 24 4 7 0 0 0 0 0 0 144 Chassahowitzka 5813 0 0 0 0 0 0 0 0 0 0 139 2 6 0 0 0 0 0 0 0 1 148 Chassahowitzka 5822 0 0 0 0 0 0 0 0 25 14 30 0 0 0 0 0 0 0 0 0 0 70 Chassahowitzka 5827 0 0 0 0 0 0 0 0 0 0 2 1 205 0 0 0 0 0 0 0 0 207 Chassahowitzka 5830 0 1 1 2 0 0 0 0 1 1 12 6 64 20 43 0 0 1 2 1 6 160 Chassahowitzka 8781 0 1 1 114 0 0 1 0 1 1 1 1 3 1 2 0 0 0 0 0 1 125 Chassahowitzka 8785 2 7 2 142 0 0 0 0 0 0 3 0 0 0 0 0 0 0 0 0 0 156 Homosassa 2092 0 0 0 106 1 1 1 1 3 0 0 0 0 0 0 0 0 0 0 0 0 111 Homosassa 2093 0 0 0 0 0 0 0 0 13 8 135 0 0 0 0 0 0 0 0 0 0 156 Homosassa 3047 0 0 0 0 0 0 0 0 9 7 86 0 0 0 0 0 0 0 0 0 0 103 Homosassa 3049 0 0 0 49 18 29 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 96 Homosassa 3057 0 8 4 124 22 38 11 11 34 4 9 1 1 1 9 2 0 4 11 3 6 300 Homosassa 3058 0 0 0 0 0 0 0 0 3 1 4 1 8 3 21 1 0 12 146 1 1 199 Homosassa 3059 0 188 2 6 0 0 2 0 5 0 0 0 0 0 0 0 0 0 0 0 0 203 Homosassa 3060 0 1 0 0 0 0 0 0 4 0 0 0 24 0 0 0 0 0 1 0 0 30 Homosassa 3062 0 2 0 0 0 6 0 4 37 1 1 0 0 0 0 0 0 0 0 0 0 51 Homosassa 5823 0 74 34 79 0 0 2 0 8 2 4 0 1 1 1 1 0 0 6 5 14 230 Homosassa 5824 0 1 1 1 0 0 1 0 1 1 4 3 23 16 111 0 0 45 72 0 0 277 Homosassa 5825 0 72 30 49 0 0 1 0 1 0 0 0 0 0 0 0 0 0 0 0 3 155
69Table 3-4. Continued Observable and unobservable zones River Tag ID 0 1 1 2 2 2 3 3 2 4 3 4 4 4 5 5 5 6 6 6 7 7 5 8 6 8 7 8 8 8 9 9 Total Homosassa 5826 0 0 0 0 0 0 0 0 0 0 11 1 4 1 46 3 0 20 94 5 8 192 Homosassa 5828 0 1 1 1 0 0 0 0 0 0 9 0 0 0 31 2 0 16 57 4 7 129 Homosassa 5829 0 0 0 0 0 0 0 0 61 29 38 1 1 1 23 5 0 4 12 0 0 174 Homosassa 5831 0 8 4 17 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 30 Homosassa 8779 0 0 0 0 0 0 0 0 5 3 54 0 0 0 3 0 0 1 1 0 0 66 Homosassa 8788 0 0 0 2 0 0 2 0 35 1 1 0 0 0 0 0 0 0 0 0 0 40
70 Figure 3-1. Map of the Homosassa River show ing the location of ac oustic receivers, their detection radius and the partitioning of observable (f illed red) and unobservable (dashed lines) area.
71 Figure 3-2. Map of the Chassahowitzka River show ing the location of acoustic receivers, their detection radius and the partitioning of observable (f illed red) and unobservable (dashed lines) area.
72 Figure 3-3. Average proportion of movement for largemouth bass over a 24 hour period for the Chassahowitzka (solid line) and Homosassa Rivers (dashed line).
73 Figure 3-4. Kernel density home range estimates of largemouth bass within the Chassahowitzka (filled bars) and Homosassa (open bars) Rivers using a subset of all data from autonomous receivers.
74 Figure 3-5. Relationship between fish total leng th and home range size for largemouth bass in the Chassahowitzka (filled circles) and Homosassa (open circles) Rivers.
75 Figure 3-6. Schoeners ratio index for various time lags of autono mous receiver data collected in the Chassahowitzka River. The expected value of V when successive observations are independent is two. Observations with va lues less than two are not independent.
76 Figure 3-7. Schoeners ratio index for various time lags of autono mous receiver data collected in the Homosassa River. The expected valu e of V when successive observations are independent is two. Observations with va lues less than two are not independent.
77 Figure 3-8. Relationship between sample size an d home range size for largemouth bass in the Chassahowitzka (filled circles) and Homosassa (open circles) Rivers.
78 CHAPTER 4 SUMMARY AND CONCLUSIONS This research project highlighted how changes in the structural characteristics of aquatic habitats may alter the foraging efficiency of a predatory fish. I utilized a case-study approach, comparing differences between energy intake an d energy expenditure by largemouth bass in two adjacent riverine systems, to test hypotheses re lated to habitat mediated influences on foraging behavior. The two rivers, which served as trea tments in a natural experiment (Diamond 1986), were similar with respect to their physical, ch emical and geomorphic char acteristics, but differed markedly in the amount of structural habita t (SAV) they afforded. Hypotheses about how largemouth bass foraging behavior might respond to different levels of st ructural habitat were formulated from prior knowledge through a litera ture review in Chapter 1. I attempted to combine multiple lines of inference to test hypotheses and strengthen conclusions regarding potential foraging strategies by largemouth bass in the two contrasting systems. Decreased foraging efficiency by largemouth bass in the Homosassa River apparently resulted from direct and indirect effects associ ated with reduced structural habitat complexity. Largemouth bass in the Homosassa River had lowe r asymptotic lengths, and older individuals were absent from the age structure samples. F ood intake alone could not explain the differences in observed growth rates, as largemouth bass in the Homosassa River consumed more prey than largemouth bass in the Chassahowitzka River. In addition, largemouth bass in the Homosassa River displayed significantly greater activity, movement rates and space use. In combination, these findings support the hypothesi s that largemouth bass in the Homosassa River have lower foraging efficiencies than largemout h bass in the Chassahowitzka River. Differences in foraging activity and space us e by largemouth bass between rivers likely stem from a combination of differences in ava ilable and vulnerable prey resources. Both prey
79 availability and prey vulnerabili ty are generally higher at intermediate levels of structural complexity. Largemouth bass in the Homosassa River appear to compensate for depressed resource availability by increasing foraging effo rt and occupying larger home ranges. Rather than searching open water and pursuing prey, la rgemouth bass in the Chassahowitzka River can reduce energetic costs by sitting in vegetation and waiting until prey pass nearby. In patchily distributed vegetation, predators are provided areas from which to ambush prey as well as areas containing high densities of prey. Largemouth bass in the Homosassa River likely incur greater metabolic costs, and possibly pred ation risks in order to acquire sufficient food resources to meet their energetic needs. Alternatively, largemout h bass in the Chassahowitzka River appear to conserve energy and reduce preda tion risks while still securing necessary resources by foraging more efficiently. Schoener (1971) originally pr oposed that animals should fo llow one of two theoretical foraging strategies: energy maxi mizing or time minimizing. While these strategies represent two end points along a continuum of fo raging strategies (Hixon 1982); selection should favor animals which can maximize energy intake while minimizing the amount of time spent foraging (Pyke et al. 1977). Bioenergetics models, fit using incremental growth data and size-at-age data, revealed two possible foraging parameter scenarios to e xplain the growth patterns observed in the Homosassa and Chassahowitzka Rivers. In each scenario, the foraging efficiency of largemouth bass was greater in the Chassahowitzka River. However, the foraging scenarios involved opposite patterns in energy acquisition and energy expenditure. In the firs t scenario, largemouth bass in the Chassahowitzka River had greater food consumption and greater associated metabolic costs than in the Homosassa River. In the second scenario, largemouth bass in the Chassahowitzka River had lower food consumpti on and lower metabolic costs than in the
80 Homosassa River. Both food consumption data and telemetry data supported the second scenario, in which largemouth bass in the Cha ssahowitzka River maximize foraging efficiency by minimizing foraging costs. This study highlighted a number of mechan isms through which habitat alterations can impact the energy acquisition rates of a pred ator. Largemouth bass in the Homosassa River incurred greater risk in order to secure pr oportionally lower energetic gains compared to largemouth bass in the Chassahowit zka River likely as a result of increased metabolic costs of foraging. While this study was unable to id entify all consequences of altered foraging efficiencies at the population and community leve ls, by combining multiple independent lines of inquiry this study provide d strong support for consequences of reduced foraging efficiency at the individual level. Although foraging, energy al location and life history characteristics can be studied independently, they must be integrated in order to fully understand an organisms ecology and the dynamics of populations in cont rasting environments (Boggs 1992). Elucidating the ultimate consequences of habitat loss for pr edators is difficult becau se of the myriad of factors that influence predator-p rey interactions. How a predator responds to changes in prey abundance, prey composition and prey vulnerability with changes to an ecosystems habitat template will invariably be system-specific. Incorporating knowledge of altered foraging efficiencies or species interactions will prove critical towards understanding and predicting the effects of such changes on ecosystem dynamics.
81 LIST OF REFERENCES Abrams, P. A. 1991. Life history and the rela tionship between food availability and foraging effort. Ecology 72:1242-1252. Anderson, O. 1984. Optimal foraging by Largemout h Bass in structured environments. Ecology 65:851-861. Bajer, P. G., G. W. Whitledge, and R. S. Ha yward. 2004. Widespread consumption-dependent systematic error in fish bioenergetics models and its implications. Canadian Journal of Fisheries and Aquatic Sciences 61:2158-2167. Bettoli, P. W., M. J. Maceina, R. L Noble, and R. K. Betsill. 1992. Piscivory in largemouth bass as a function of aquatic vegetation abundan ce. North American Journal of Fisheries Management 12:509-516. Bettoli, P. W., M. J. Maceina, R. L Noble, and R. K. Betsill. 1993. Response of a reservoir fish community to aquatic vegetation removal. North American Journal of Fisheries Management 13:110-124. Beyer, H. L. 2004. Hawth's analysis tools for ArcGIS. http://www.spatialecology.com/htools. Accessed Feb 2008. Biro, P. A., M. V. Abrahams, J. R. Post, and E. A. Parkinson. 2006. Behavioral trade-offs between growth and mortality explain evoluti on of submaximal growth rates. Journal of Animal Ecology 75:1165-1171. Boisclair, D., and W. C. Leggett. 1989. The im portance of activity in bioenergetics models applied to actively foraging fishes. Canadian Journal of Fisheries and Aquatic Sciences 46:1859-1867. Biosclair, D., and P. Sirios. 1993. Testing assump tions of fish bioenergetics models by direct estimation of growth, consumption, and activ ity rates. Transactions of the American Fisheries Society 122:784-796. Boggs, C. L. 1992. Resource allocation: exploring connections between foraging and life history. Functional Ecology 6:508-518. Brown, S. J. and M. J. Maceina. 2002. The influe nce of disparate levels of submersed aquatic vegetation on Largemouth Bass population char acteristics in a Georgi a reservoir. Journal of Aquatic Plant Management 40:28-35. Charland, M. B. and P. T. Gregory. 1995. Move ments and habitat use in gravid and nongravid female garter snakes (Colubridae: Th amnophis). Journal of Zoology 236:543-561. Charnov, E.L. 1976. Optimal foraging; attack st rategy of a mantid. American Naturalist 110:141151.
82 Childress, J. J., and G. N. Somero. 1990. Metaboli c scaling: a new perspective based on scaling of glycolytic enzyme activities American Zoologist 30:161-173. Christensen, D. L., B. R. Herwig, D. E. Schi ndler, and S. R. Carpenter. 1996. Impact of lakeshore development on coarse woody debris in north temperate lakes. Ecological Applications 6:1143-1149. Coen, L. D., K. L. J. Heck, and L. G. Abele. 1981. Experiments on competition and predation among shrimps of seagrass meadows. Ecology 62:1484 Connell, J. H. 1972. Community interactions on marine intertidal shores. Annual Review of Ecology and Systematics 3:169-192. Crowder, L. B., and W. E. Cooper. 1979. Structur al complexity and fish -prey interactions in ponds: a point of view. Pages 2-10 in D. L. Johnson and R. A. Stein, editors. Response of fish to habitat structure in standing water. American Fisheries Society, North Central Division, Special Publication 6, Bethesda, Maryland, USA. Crowder, L. B. and W. E. Cooper. 1982. Habita t structural complexity and the interaction between bluegills and their prey. Ecology 63:802-1813. Crowder, L. B., E. W. McCollum, and T. H. Martin. 1998. Changing perspectives on food web interactions in lake li ttoral zones. Pages 240-249 in E. Jeppesen, M. Sondergaard, M. Sondergaard, K. Christoffersen, editors. Th e structuring roles of submerged macrophytes in lakes. Springer Berlin Heidel berg, New York, New York, USA. Demers, E., R. S. McKinley, A. H. Weatherley and D. I. McQueen. 1996. Activity patterns of largemouth bass and smallmouth bass determined with electromyogram biotelemetry. Transactions of the American Fisheries Society 125:434-439. Diamond, J. 1986. Overview: laboratory expe riments, field experiments, and natural experiments. Pages 3-22 in J. Diamond and T. J. Case, editors. Community ecology. Harper and Row, New York, New York, USA. Dibble, E. D., K. J. Killgore, and S. H. Harrel. 1996. Assessment of fish-plant interactions. Multidimensional approaches to reservoir fisheries management. American Fisheries Society Symposium 16:357-372. Duarte, C. M. 1995. Submerged aquatic vegetati on in relation to different nutrient regimes. Ophelia. 41:87-112. Eliassen, S., C. Jorgensen, M. Mangel, and J. Giske. 2007. Explorati on or exploitation: life expectancy changes the valu e of learning in foraging strategies. Oikos 116:513-523. Fausch, K. D., S. Nakano, and S. Kitano. 1997. Experimentally induced foraging mode shift by sympatric charrs in a Japanese mounta in stream. Behavioral Ecology 8:414-420.
83 Flynn, A. J., and D. A. Ritz. 1999. Effect of habitat complexity and predatory style on the capture success of fish feeding on aggreg ated prey. Journal of Marine Biological Association:87. Frazer, T. K., S. K. Notestein, and W. E. Pi ne. 2006. Changes in the Physical, Chemical and Vegetative Characteristics of the Homosassa, Chassahowitzka and Weeki Wachee Rivers. Final Report: Southwest Florida Wa ter Management District, Brooksville, Florida. 163 pp. Frazer, T. K., W. E. Pine III, M. V. Lauretta, G. Warren, E. Nagis, W. Strong, and T. Tuten. 2008. Increased nutrient loading of spring-fed coastal rivers: effect s on habitat and faunal communities. Annual Report: Florida Fish and Wildlife Conservation Commission, Tallahassee, Florida. 74 pp. Gotceitas, V., and P. Colgan. 1989. Predat or foraging success and habitat complexity: quantitative test of the threshold hypothesis. Oecologia 80:158-166. Grant, J.W.A, and D.L.G. Noakes. 1987. Movers and stayers: foraging tactics of young-og-theyear brook charr, Salvelinus fontinalis Journal of Animal Ecology 56: 1001-1013. Guy, C. S., H. L. Blankeship, and L. A. Nielsen. 1996. Tagging and marking. Pages 353-379 in B. R. Murphy and D. W. Willis, editors. Fi sheries techniques, 2nd edition. American Fisheries Society, Benthesda, Maryland, USA. Harvey, B. C., R. J. Nakamoto, and J. L. White. 1999. Influence of large woody debris and a bankfull flood on movement of adult reside nt coastal cutthroat trout (Oncorhynchus clarki) during fall and winter. Canadian J ournal of Fisheries and Aquatic Sciences 56:2161. Heck K. L. J., and R. J. Orth. 1980. Seagra ss habitats: the roles of habitat complexity, competition and predation in structuring associated fish and motile macroinvertebrate assemblages. Pages 449-464 in V.S. Kennedy, editor. Estuarine perspectives. Academic Press, London, UK. Heck, K. L. J., and L. B. Crowder. 1991. Habita t structure and predato r-prey interactions in vegetated aquatic systems. Pages 281-299 in S. S. Ball, E. D. McCoy, and H. R. Mushinsky, editors. Habitat structure: th e physical arrangement of objects in space. Chapman and Hall, London, UK. Henderson, B. A., G. E. Morgan, and A. Va illancourt. 2004. Growth, ingestion rates and metabolic activity of walleye in lakes with and without lake herri ng. Journal of Fish Biology 65:1270. Helfman, G. S. 1990. Mode selection and m ode switching in foraging animals. Pages 249-297 in P. J. B. Slater, J. S. Rosenblatt, and C. Beer editors. Advances in the Study of Behavior, volume 19. Academic Press, San Diego, California, USA.
84 Hixon, M. A. 1982. Energy maximizers and time mi nimizers: theory and reality. The American Naturalist 119:596-599. Hodgson, J. R., and J. F. Kitchell. 1987. Oppor tunistic foraging by largemouth bass (Micropterus salmoides). The American Mi dland Naturalist 118:323-336. Holling, C. S. 1966. The functional response of inve rtebrate predators to prey density. Memoirs of the Entomological Society of Canada 48:3-86. Hooge, P. N., and B. Eichenlaub. 1997. Animal movement extension to Arcview, version 1.1. US Geological Survey, Anchorage. Huey, R., and E. C. Pianka. 1981. Ecological consequences of foraging mode. Ecology 62:991999. Hutchings, J. A., and R. A. Myers. 1993. The eff ect of age on the seasonality of maturation and spawning of Atlantic cod, Gadus morhua Canadian Journal of Fisheries and Aquatic Sciences 50:2468. Hyslop, E. J. 1980. Stomach contents analysis: a review of methods a nd their application. Journal of Fish Biology 17:411.429. James, P. L. and K. L. Heck. 1994. The effect s of habitat complexity and light intensity on ambush predation within a simulated seagrass habitat. Journal of Experimental Marine Biology and Ecology 176:187-200. Jensen O. P., M. C. Christman, and T. J. Miller. 2006. Landscape-base d geostatistics: a case study of the distribution of blue crab in Chesapeake Bay. Environmetrics 17:605. Johnson, C. J., K. L. Parker, D. C. Heard, a nd M. P. Gillingham. 2002. A multiscale behavioral approach to understanding the movements of woodland caribou. Ecological Applications 12:1840-1860. Jones, G. W., S. B. Upchurch, K. M. Champion and D. J. Dewitt. 1997. Water-quality and Hydrology of The Homosassa, Chassahowitz ka, Weeki Wachee, and Aripeka Spring Complexes, Citrus and Hernando Counties, Florida (Origin of Increasing Nitrate Concentrations). Ambient Ground-Water Qu ality Monitoring Program. Southwest Florida Water Management District. 167 pp. Kaufman, S. D., J. M. Gunn, G. E. Morgan, and P. Couture. 2006. Muscle enzymes reveal walleye (Sander vitreus) are less active when larger prey (cisco, Coregonus artedi) are present. Canadian Journal of Fisher ies and Aquatic Sciences 63:970-979. Kernohan, B. J., R. A. Gitzen, and J. J. Mi llspaugh. 2001. Analysis of animal space use and movements. Pages 126 in J. J. Millspaugh and J. M. Marzluff, editors. Radiotracking and animal populations. Academic Press, San Diego, California, USA.
85 Kerr, S. R. 1971. Prediction of fish growth effici ency in nature. Journal of Fisheries Research Board of Canada 28:815. Kimura, M. 1980. A simple method for estimating e volutionary rate of ba se substitutions through comparative studies of nucleotide sequen ces. Journal of Molecular Evolution 16:111 120. Krebs, J. R., and R. H McCleery. 1984. Op timization in behavioral ecology. Pages 91-121 in J. R. Krebs and N. B. Davies, editors. Beha vioral ecology, an evolutionary approach. Blackwell Scientific Publications, Oxford, UK. Lester, N. P., B. J. Shuter, and P. A. Abrams 2004. Interpreting the von Bertalanffy model of somatic growth in fish: the cost of reproduc tion. Proceedings of the Royal Society of London Series B 271:1625. Marzluff, J. M., J. J. Millspaugh, P. Hurvitz, and M. S. Handcock. 2004. Relating resources to a probabilistic measure of sp ace use: forests fragments and Stellers Jays. Ecology 86:1411-1427. McNamara, J. M. and A. I. Houston. 1994. The eff ect of a change in foraging options on intake rate and predation rate. Am erican Naturalist 144:978-1000. Nelson, W. G. 1979. Experiment al studies of selective pred ation on amphipods: consequences for amphipod distribution and abundance. Jour nal of Experimental Marine Biology and Ecology 38:225. Noonburga, E. G., L. A. Newmanb, M. Lewis, R. L. Crabtree, and A. B. Potapov. 2007. Sequential decision-making in a variable environment: modeling elk movement in Yellowstone National Park as a dynamic ga me. Theoretical Population Biology 71:182195. Paloheimo, J. E., and L. M. Dickie. 1965. Food a nd growth of fishes. I. A growth curve derived from experimental data. Journal of Fish eries Research Board of Canada 22:521. Parma, A. M., and R. B. Deriso. 1990. Dynami cs of age and size composition in a population subject to size-selective mortal ity: Effects of phenotypic vari ability in growth. Canadian Journal of Fisheries and A quatic Sciences 47:274-289. Pianka, E. R. 1966. Convexity, desert lizards and spatial heteroge neity. Ecology 47:1055-1059. Plummer, M. V., and J. D. Congdon. 1994. Radiot elemetric Study of Activity and Movements of Racers (Coluber constrictor) Associated with a Carolina Bay in South Carolina. Copeia 1994:20-26. Pollock, K. H., R. Winterstein, C. M. Bunck, and P. D. Curtis. 1989. Survival analysis in telemetry studies: the staggered entry design. Journal of Wildlife Management 53:7-15.
86 Pope, K. L., M. L. Brown, W. G. Duffy, and P. H. Michaletz. 2001. A caloric-based evaluation of diet indices for largemouth bass. E nvironmental Biology of Fishes 61:329-339. Porak, W. F., S. Crawford, and J. R. Este s. 1987. Population dynamics of largemouth bass resources. Final Report: St ate of Florida Game and Fresh Water Fish Commission, Tallahassee, Florida. 98 pp. Pyke, G. H. 1979. Optimal foraging in fish. Pages 199-202 in H. Clepper, editor. Predator-prey systems in fisheries management. Sport Fishing Institute, Washington, D.C., USA. Rice, J. A., and P. A. Cochran. 1984. Indepe ndent evaluation of a bioenergetics model for largemouth bass. Ecology 65: 732-739. Ricker, W. E. 1975. Computation a nd interpretation of biological statistics of fish populations. Fisheries Reseach Board of Canada Bulletin 191. Rennie M. D, N. C. Collins, B. J. Shuter, J. W. Rajotte, and P. Couture. 2005. A comparison of methods for estimating activity costs of wild fish populations: more active fish observed to grow slower. Canadian Journal of Fisheries and Aquatic Sciences 62:767. Sammons, S. M., M. J. Maceina, and D. G. Pa rtridge. 2003. Changes in behavior, movement, and home ranges of largemouth bass following large-scale Hydrilla removal in Lake Seminole, Georgia. Journal of Aquatic Plant Management 41:31. Sass, G. G., J. F. Kitchell, S. R. Carpenter, T. R. Hrabik, A. E. Marburg, and M. G. Turner. 2005. Fish communitiy and food web responses to a whole-lake removal of coarse woody habitat. Fisheries 31:321-220. Savino, J. F., and R. A. Stein. 1982. Predator -prey interactions between largemouth bass and bluegills as influenced by simulated, s ubmersed, vegetation. Transactions of the American Fisheries Society 111:255-266. Savino, J. F., and R. A. Stein. 1989. Behavior of fish predator s and their prey: habitat choice between open water and dense vegetation. E nvironmental Biology of Fishes 4:287-293. Schoener, T. W. 1971. Theory of feeding strate gies. Annual Review of Ecology and Systematics 2:369-404. Schoener, T. W. 1987. Time budgets and territor y size: some simultaneous optimization models for energy maximizers. American Zoologists 27:259. Seaburg, K. G. 1957. A stomach sampler for li ve fish. The Progressive Fish-Culturist 19:137139. Seaman, D. E., and R. A. Powell. 1996. An ev aluation of the accuracy of kernel density estimators for home range analysis. Ecology 77:2075-2085.
87 Seaman, D. E., J. J. Millspaugh, B. J. Kernohan, G. C. Brundige, K. J. Raedeke, and R. A. Gitnzen. 1999. Effects of sample size on ke rnel home range estimates. The Journal of Wildlife Management 63:739-747. Scheuerell, M. D., and D. E. Schindler. 2004. Cha nges in the spatial dist ribution of fishes in lakes along a residential developmen t gradient. Ecosystems 7:98-106. Sherwood, G. D., I. Pazzia, A. Moeser, A. Hont ela, and J. B. Rasmussen. 2002. Shifting gears: enzymatic evidence for the energetic advantag e of switching diet in wild-living fish. Canadian Journal of Fisheries a nd Aquatic Sciences 59:229241. Simpfendorfer, C. A., M. R. Heupel, and R. E. Heuter. 2002. Estimation of short term centers of activity from an array of omnidirectional hydrophones, and its use in studying animal movements. Canadian Journal of Fish eries and Aquatic Sciences 59:23-32. Smith, V. H., G. D. Tilman, and J. C. Nekola. 1999. Eutrophication: impacts of excess nutrients inputs on freshwater, marine and terrestrial ecosystems. Environmental Pollution 100:179-196. Stearns, S. C. 1976. Life-history tactics: a review of the ideas. The Quarterly Review of Biology 51:3-47. Swihart, R. K., and N. A. Slade. 1985. Influe nce of sampling interval on estimates of homerange size. The Journal of W ildlife Management 49:1019-1025. Stephens, D. W., and J. R. Krebs. 1986. Foraging Theory. Princeton University Press, Princeton, New Jersey, USA. Taylor, N. G., C. J. Walters, and S. J. D. Martell. 2006. A new likelihood for simultaneously estimating von Bertalanffy growth parameters gear selectivity, and natural and fishing mortality. Canadian Journal of Fish eries and Aquatic Sciences 62:215-223. Tyler, J. A., and K. A. Rose. 1994. Individual variability and spatial heterogeneity in fish population models. Reviews in Fish Biology and Fisheries 4:91-123. Werner, E. E. 1979. Niche partitioning by food size in fish communities. Pages 311-322 in H. Clepper, editor. Predator-prey systems in fisheries manageme nt. Sport Fishing Institute, Washington, D. C., USA. Werner, E. E., and G. G. Mittelbach. 1981. Optim al foraging: field tests of diet choice and habitat switching. Ameri can Zoologist 21:813-829. Werner, E. E., G. G. Mittelbach, D. J. Hall, and J. F. Gilliam. 1983. Experimental tests of optimal habitat use in fish: the role of relative habitat prof itability. Ecology 64:15251539. Wetzel, R. G., and G. E. Likens 1991. Limnological Methods, 2nd edition. Springer-Verlag, New York, New York, USA.
88 Whitledge, G. W., and R. S. Hayward. 1997. Labor atory evaluation of a bioenergetics model for largemouth bass at two temperature and feedi ng levels. Transactions of the American Fisheries Society 126:1030-1035. Winter, J. D. 1996. Advances in Un derwater biotelemetry. Pages 555-590 in B.R. Murphy and D. W. Willis, editors. Fisheries Techniques, 2nd edition. American Fisheries Society, Bethesda, Maryland, USA. Worton, B. J. 1989. Kernel methods for estima ting the utilization distribution in home-range studies. Ecology 70:164-168. Worton, B. J. 1995. Using Monte Carlo simula tion to evaluate kernel-based home range estimators. Journal of Wildlife Management 59:794-800. Wyda, J. C., L. A. Deegan, J. E. Hughes, and M. J. Weaver. 2002. The response of fishes to submerged aquatic vegetation complexity in two ecoregions of the mid-Atlantic bight: Buzzards Bay and Chesapeake Bay. Estuaries and Coasts 25:86-100. Yobbia, D. K., and L. A. Knochenmus. 1989. Salin ity and flow relations and effets of reduced flow in the Chassahowitzka River and Homosassa River Estuaries, Southwest Florida. USGS Water-Resources Investig ations Report 88-4044:1-48.
BIOGRAPHICAL SKETCH Jakob Tetzlaff was born to Dennis and Deanna Tetzlaff in summer, 1983. He grew up on Pewaukee Lake in the great state of Wisconsin. Ja ke earned his B.S. in wildlife ecology from the University of Wisconsin-Madison in 2005. He comple ted his masters research at the University of Florida in 2008. He will continue his graduate career at the University of Florida in a Ph.D. program under Dr. William Pine III.