1 EVALUATING RELATIONSHIPS BETWEEN ANGLER EFFORT, CATCH RATES, AND MANAGEMENT OPTIONS IN FLORIDA RECREATIONAL FISHERIES By NICHOLAS WAYNE COLE A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2014
2 Â© 2014 Nicholas Wayne Cole
3 To Dad, Mom, and all the rest of my family
4 ACKNOWL E DG MENTS I would like to thank my advisor, Dr. Mike Allen for providing me the opportunities and responsibility I needed to grow as a research professional and transition my style of thinking from that of a technician to a scientist. Without his blunt and thoughtfu l tutelage, I would not be the scientist I am today. I would also like to thank the other members of my committee, Dr. Kai Lorenzen and Travis Tuten for their consistent support both with my work in this thesis but also in all aspects of my time at the Uni versity of Florida. I would like to thank Dr. Rob Ahrens for always being willing to take time out of your day to help me understand an analysis. I would like to thank Drew Dutterer, Bill Pouder, Bob Eisenhauer, Brandon Thompson, Jay Holder, Kevin McDaniel s and Jason Dotson with the Florida Fish and Wildlife Commission for providing me data and support in my research. I would also like to thank University of Florida L AKEWATCH director Mark Hoyer for providing me data and encouragement in my research. I woul d like thank Zak Slagle and Stephanie Shaw and all the other members of the Dequine building for being great lab mates and their reliability in assisting me whenever I needed it. I would especially like to Merrill Rudd for her absolute unwavering support t hroughout my time at UF. She provided me the encouragement I needed to succeed and was working with me every weekend, even if it was via video chat. Thank you to Mendy Willis and Cynthia Hight for helping me achieve all the additional projects I set out to do outside of my research. Lastly, I would like to thank my parents and family for their hard work and sacrifice in providing me these opportunities. I would not have been able to reach my goals without it.
5 TABLE OF CONTENTS page ACKNOWL E DG MENTS ................................ ................................ ................................ ............... 4 LIST OF TABLES ................................ ................................ ................................ ........................... 6 LIST OF FIGURES ................................ ................................ ................................ ......................... 7 ABSTRACT ................................ ................................ ................................ ................................ ..... 8 CHAPTER 1 INTRODUCTION ................................ ................................ ................................ .................. 10 2 FACTORS INFLUENCING BLACK CRAPPIE Pomoxis nigromaculatus AND FLORIDA BASS Micropterus floridanus ANGLER SITE SELECTION IN THE STATE OF FLORIDA ................................ ................................ ................................ ........... 14 Methods ................................ ................................ ................................ ................................ .. 15 Data Collection ................................ ................................ ................................ ................ 15 Analysis ................................ ................................ ................................ ........................... 18 Results ................................ ................................ ................................ ................................ ..... 18 Discussion ................................ ................................ ................................ ............................... 19 3 DECREASING CAPTURE VULNERABILITY OF FLORIDA BASS Micropterus floridanus WITH EXPOSURE TO CATCH AND RELEASE ANGLING. ......................... 32 Methods ................................ ................................ ................................ ................................ .. 34 Capture Recapture Population Estimates ................................ ................................ ........ 34 Experimental Angling ................................ ................................ ................................ ..... 35 Analysis ................................ ................................ ................................ ........................... 36 Results ................................ ................................ ................................ ................................ ..... 38 Discussion ................................ ................................ ................................ ............................... 40 4 CONCLUSIONS ................................ ................................ ................................ .................... 51 LIST OF REFERENCES ................................ ................................ ................................ ............... 54 BIOGRAPHIC AL SKETCH ................................ ................................ ................................ ......... 59
6 LIST OF TABLES Table page 2 1 List of hypotheses tested through GLM model fits and the literature used to select the appropriate components included. ................................ ................................ ..................... 24 2 2 Second order model selection (AICc) for Black Crappie hypotheses. .............................. 25 2 3 Second order model selection (AIC c ) for Florida Bass hypotheses. ................................ .. 26 2 4 AIC c selected model (Expected Catch) summary values for Black Crappie fishery. ........ 27 2 5 AIC c selected model (Aquatic Structure) summary values for Florida Bass fishery. ........ 27 3 1 Average weekly effort per hectare of Florida lakes during peak season. .......................... 45 3 2 Florida Bass angled during the study. ................................ ................................ ................ 46 3 3 Model selection scores (AICc) bet ween lure type and CPUE ................................ .......... 46
7 LIST OF FIGURES Figure page 2 1 The ten predictive variables used to create representative models in this study, plotted against Florida Black Crappie angling hou rs per hectare ................................ ...... 28 2 2 The ten predictive variables used to create representative models in this study, plotted against Florida Bass angling hours per he ctare ................................ ..................... 29 2 3 Diagnostic assessments of expected catch model for Black Crappie. ............................... 30 2 4 Diagnostic assessments of aquatic structure model for Florida Bass. ............................... 31 3 1 Length f requency distribution of the Florida Bass angled in Devils Hole Lak e ............... 47 3 2 Catch per unit effort (CPUE) of Florida Bass at cumulative angler catches usin g active and finesse lures ................................ ................................ ................................ ...... 48 3 3 Catch per angler hour plotte d over cumulative catches ................................ ..................... 49 3 4 Simulated catch per unit effort plotted over time in angling days with alternate m anagement strategies ................................ ................................ ................................ ....... 50
8 Abstract of Thesis Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Science EVALUATING RELATIONSHIPS BETWEEN ANGLER EFFORT, CATCH RATES, AND MANAGEMENT OPTIONS IN FLORIDA RECREATIONAL FISHERIES By Nicholas W. Cole August 2014 Chair: Michael Allen Major: Fisheries and Aquatic Sciences Recreational fishery managers have seldom considered fishing effort restrictions an appropriate management tool . Because angler catch rates a re often assumed to be linearly related to fish abundance, management actions that increase fish abundanc e are assumed to improve angler catch rates , despite unlimited entry . T he objectives of this thesis were to : 1) identify factors that are related to fishing effort patterns in open access recreational fisheries , 2) identify whether exposure to catch and release angling makes fish less vulnerable to capture and thus influences angler CPUE, and 3) simulate regional effort based management strategies to quantify potential improvement in catch rates. I developed normal generalized linear models to a ssess the relationships between angler effort and lake and year specific influencing variables for Florida Bass Micropterus floridanus and Black Crappie Pomoxis nigromaculatus . For Black Crappie, both proportion of littoral area (P<0.001) and catch per effort (P=0.07) were positively related to effort . For Florida Bass , proportion of littoral area was positively related to angler effort (P < 0 .05), suggesting that aquatic structure and expected catches were important factors for attracting anglers . I sought to identify whether an gler CPUE would decline after exposing Florida Bas s to catch and release angling . I used fou r weeks of experimental angling to test active and finesse
9 lur es . The finesse lure exhibited a linear decline in CPUE from 1.77 fish per angler hour to 0.83 and the active lure showed an exponential decl ine in CPUE from 2.33 fish per angler h our to 0.25. Using the vulnerable pool framework, I simulated the exchange rates between vulnerable and invulnerabl e pools . I showed that 36 angling hours per week in Devils Hole Lake would reduce available catch rate s to <1 fish p er angling hour after 30 angling days but a 30 day seasonal closure of the fishery increase d catch rates back to level s similar to an unfished state. Because angler catch per effort may be negatively impacted by behavior modification of fish , this makes active management important to sustainable economic benefits in recreational fisheries .
10 CHAPTER 1 INTRODUCTION Recreational fisheries often occur in a landscape of systems that provide varying degrees of satisfaction to anglers. Cox et al. (2002) noted that in open access fisheries, anglers will be attracted to high quality fishing opportunities until their satisfa ction declines, at which point they may select alternate fishing sites or choose not to fish (Johnston et al. 2010; Beardmore et al. 2011; Hunt et al. 2011) . Expected catch rates are usually considered an important driver in this relationship between fishi ng and satisfaction (Holland and Ditton 1992; Carpenter and Brock 2004) . Non catch related factors like enjoying nature and relaxation can also be important drivers for satisfaction, but are often over looked and more difficult to manage than catch related metrics (Hansen et al. 2005; Hunt et al. 2007; Post et al. 2008; Hunt et al. 2011) . The satisfaction that anglers receive from a fishing site will vary substantially depending on the different levels of importance placed on the outcomes of angling (Fed ler and Ditton 1994). A consumptive fisher will seek the highest expected catch of harvestable fish and may be less interested in rare trophy events (Bryan 1977; Ditton et al. 1992). A trophy oriented fisher is more likely to release fish and prefers a few large fish over many average sized fish (Bryan 1977; Ditton et al.1992) . The diversity of these angling types can vary greatly between fisheries mode or target species. Thus, all management strategies should consider the typologies of anglers taking part in the fishery (Johnston et al. 2010). Recent work has proven that recreational fisheries can cause both growth and recruitment overfishing (Post et al. 2002; Parkinson et al. 2003; Sullivan 2003; Cooke and Cowx 2004; Allen et al. 2013) . Thus, understandi ng regional angling effort dynamics is important for maintaining high angler satisfaction, and can also be important for preventing overfishing (Post et al. 2002; Ward et al. 2014). For example, if angling effort is strongly tied to angler catch per effort , then
11 declining catch rates would cause effort to shift to other fisheries . Alternately, if angler effort remains high despite stock depletions , then overfishing is likely (Allen et al. 2013) . Thus, understanding how fishing effort responds to changes in catch per effort is a key need in recreational fisheries management. Most fisheries managers typically assume that angler catch rates will increase proportionately to fish abundance . However, it has been shown that in some Largemouth Bass M. salmoides and trout fisheries, catch and release fishing can decrease angler catch rates over time without changing abundance (Anderson and Heman 1969 ; Beukema 1970 ; Askey et al. 2006). This probably occurs via some form of behavior modification like spatial avoidance, decreased risk taking, or learned avoidance of lures (Cox and Walters 2002). Managers typically try to maintain high angler satisfaction and prevent overfishing by stabilizing or improving the density of desirable fish species and sizes , which is assumed to improve angler catch rates . However, if the relationship between population abundance and angler catch rates is not proportional, there is still the potential for decreased angler satisfaction. In most cases, angler access to fishing sites is only limi ted by a nominal license fee and the opportunity costs associated with fishing a particular site. There is rarely a limitation on the number of anglers allowed to fish in a system. Managers instead use bag limits and varying styles of length based harvest limitations to protect against overfishing and improve fish abundance. Exploitation in a lake system is influenced by three main factors; discard mortality, harvest restrictions (e.g., bag/size limits), and the amount of fishing effort (Cox et al. 2002) . B ecause managers usually only have direct control over regulations, this can lead to great difficulty in managing open access recreational fisheries (Cox et al. 2002; Post et al. 2002; Allen et al. 2013) .
12 The importance of recreational fisheries and the difficulties managing them suggest that in some cases, novel management strategies like regional effort based management will improve long term sustainability of a fishery and its economic and social benefits. Cox et al. (2003) showed that considering the fisheries independent of each other within a single region can have deleterious effects on the overall value within that region. Angler mobility is important , as anglers seeking high satisfaction can distribute throughout a region in patterns unforeseeable to managers, limiting the efficacy of bag and minimum length limiting regulations (Cox et al. 2003). Objectives and chapter structure: In this thesis, I sought to quantify the components of a recreational fishery that can make classic output control stra tegies (i.e. minimum length regulations and bag limits) ineffective and evaluate the potential for regional effort based management as an alternative management strategy. My objectives were:  Identify the important factors that drive angler site selecti on within the state of Florida,  Identify whether catch and release angling can influence angler catch rates and if it is specific to the tactics of the individual angler,  Simulate regional effort based management strategies based on objective 2 and quantify the potential improvement available in angler catch rates. Chapter I introduces the concepts of behavioral modification in targeted fish species, effort based management as a potential strategy, and the objectives necessary to assess these in t he state of Florida. Chapter II highlights the first steps in this process by evaluating the factors that influence angler site selection within two Florida fisheries; a harvest oriented fishery in Black Crappie and a catch & release oriented fishery in Fl orida Bass. Chapter III defines an empirical study that identifies how angler catch rates may be impacted fish behavioral modification due to being caught and release and how this may differ with different lure types.
13 This is used to present a simulation o f the potential that effort based management may have to maintain high catch rates in catch and release driven fisheries. Chapter IV provides conclusions about the potential management implications of this research in the state of Florida and recommends fu ture research objectives that will expand on this work.
14 CHAPTER 2 FACTORS INFLUENCING BLACK CRAPPIE Pomoxis nigromaculatus AND FLORIDA BASS Micropterus floridanus ANGLER SITE SELECTION IN THE STATE OF FLORIDA Inland freshwater recreational fisheries are generally open access, and fishing effort is driven by a variety of socioeconomic factors that lead anglers to decide when and where to allocate fishing effort (Hunt et al. 2007; Johnston et al. 2010) . Recreatio nal anglers are mobile and can move throughout the landscape in order to maximize their own satisfaction. Some have hypothesized that angler movement dynamics are based solely on the expected catch available in each lake system (Parkinson et al. 2004) . Thi s assumes that the only factors influencing angler decision to allocate effort is the expected catch or harvest rate of fish. This assumption does not account for the many non catch related factors that have been shown to be important in an tion of effort such as cost of travel, natural beauty of the site, and quality of access facilities (Post et al. 2002; Hunt 2005; Hunt et al. 2007; Post et al. 2008; Hunt et al. 2011; Post and Parkinson 2012) . Angler satisfaction can be positively and n egatively affected by both catch and non catch related factors within a fishery (Hunt 2005; Post et al. 2002) . This means that ignoring non catch related influences on angling effort can preclude a full understand ing of the spatial and temporal fishing eff ort dynamics and potential for overfishing. Thus, understanding factors driving regional effort dynamics is important and should play a role in the management of recreational fisheries. Because many anglers seek different outcomes from fishing (e.g., trophy catch vs. high harvest), the satisfaction that individual anglers derive from a fishery can differ greatly among angler types (Bryan et al. 1977). This differing perception of satisfaction also leads to very different effects on fish population qual ity and structure (Johnston et al. 2010) . For example, harvest driven anglers will consistently fish lakes that provide high expected catch or have
15 minimum size limits that allow more opportunity to harvest fish compared to other available lakes (Johnston et al. 2010) . Trophy anglers base their site selection on factors that do not directly relate to catch rates or harvest possibilities (Johnston et al. 2010) . Because these angler es adequately given variability in angler expectations and variation in fish population metrics across landscapes (Post et al. 2002 ; Hunt et al. 2011 ; Johnston et al. 2012 ). If angler site selection is driven by factors independent of catch rates, then managing fishing effort in order to increase catch rates would have very little impact on angler satisfaction . Alternately, if angler effort is allocated based on catch related utility metrics, then efforts to improve catch rates should result in higher an gler effort in individual lakes and regions of lakes. However, relatively few studies have evaluated how catch and non catch related factors influence the spatial and temporal patterns in fishing effort (Parkinson et al. 2004 ; Hunt et al. 2007 ; Post et al. 2008 ; Post and Parkinson 2012). The objectives of this study were to (1) evaluate specific hypotheses of catch and non catch related metrics that influence the observed fishing effort for Black Crappie Pomoxis nigromaculatus and (Florida Bass Micropterus floridanus fisheries . The Black Crappie fishery is harvest oriented, whereas the Florida Bass fishery is primarily catch and release . Thus, these two species offer different case studies that could vary in fishing effort responses to both catch and non cat ch related attributes of lakes . Methods Data Collection Creel surveys were conducted by Florida Fish and Wildlife Commission on 26 total lakes between 1985 and 2012. Catch, harvest, and effort were assessed in seasonal periods for each lake and for each study species. Data were collected during 16 week long peak season (January May) creel surveys . Four week periods were used with ten sampling days per period; six being
16 on weekdays and four being on weekends. Each sampling day consisted of a four hour samp ling period. Each stratified weekend or weekday sampling day was chosen at random using uniform probability distribution . Both boat and bank anglers were sampled , and for most lake years a roving creel survey technique was used. For a minority of lake year s, access point creel survey methods were used. Some sampling occurred outside of peak season sampling periods , but for this analysis only peak season data was used . The lakes assessed were of varying size and geographically representative of fishing oppor tunities in Central Florida. Ten variables were chosen to form five models to predict fishing effort , with each model representing specific catch and non catch related components for potential angler site selection . There is overlap in how each variable can be interpreted . For example, habitat characteristics of a lake influence both the abundance of fish (via recruitment) and the natural beauty of a fishing site . Therefore, predictors such as proportion of littoral area could influence anglers expected c atch via fish abundance , or also influence their site choice due to natural aquatic structure . I developed models that represented hypotheses about expected catch and non catch factors that could influence angler site selection for both Black Crappie and F lorida Bass fisheries , based on variables from the peer reviewed literature (Table 2 1). Only one catch related model was developed, which included angler catch per effort (CPUE) and proportion littoral area (Table 2 1) . This model hypothesized that anglers would expect higher catch rates from sites with higher realized catch rates and those with high quality littoral habitat. The remaining four models represented non catch related components (Table 2 1) . The model represe nting aquatic structure hypothesized that anglers would choose sites with more aquatic and riparian habitat and less shoreline residential development . The seclusion model postulated that fishing effort would decrease for sites closer to urban centers and for lakes
17 with highly developed shorelines. The travel cost model considered proximity to urban areas and parking costs . Finally, the facilities model hypothesized that anglers would choose sites with more boat ramp and parking space, better restrooms, and lower cost to park (Table 2 1). Catch per effort was calculated using a mean of ratios estimator where the individual ratios of catch over the hours angled that occurred with each angler trip is used . These individual ratios are then summed and divided over the number of anglers interviewed to calculate a mean estimate of CPUE (Zale, Parrish, and Sutton 2013). These daily estimates then receive an equal weighting within the overall estimate of CPUE for the season. Thus, this is estimate of CPUE is indepe ndent of the effor t values used in the study, allowing me to use CPUE as a predictor of effort within expected catch model. Proportion of littoral area was calculated using the total surface area of each lake and the total littoral area . This was e stimate d using mean littoral widths found for each lake year in the AKEWATCH long term monitoring database. Each littoral width was calculated by averaging >10 transects in each lake year. Any littoral widths in unsampled lake years were Proportion of developed shoreline was calculated by using Google Earth P ro to quantify the total perimeter length of scaled and digitized satellite images of each a lake and the leng th of the perimeter with any type of developmental structure, including dwellings, ramps, or docks. Florida Fish and Wildlife Commission records provided the type of bathroom, the number of trailer parking spots available, the cost to park, and the number of boat ramps on the lake. The type of bathroom was indexed where 0 represent ed no bathroom present, a 1 represents a porto potty style bathroom, a 2 represents an outhouse style bathroom, and a 3 represents a standard flush bathroom. Lastly, the number o f people within a 50 kilometer radius, 100 kilometer radius,
18 and 150 kilometer radius was quantified for each lake during each lake year . I created a spatially oriented distance matrix that defined town centers that fell within defined radii from a boat ra mp on each study lake using the fields package in Program R. This matrix uses the Euclidian distance between each location. Population estimates for each town center were defined in each lake year using US Census estimated populations numbers. Analysis To identify catch and non catch related components that may influence fishing effort, I formed additive generalized linear models (GLM) representative of the specific hypotheses (Table 2 1). I fitted each GLM assuming normal error distribution, and the dat a represented each lake year. I used second order Akaike Information Criterion (AIC c ) model selection through the Program R package AICcmodavg (V. 1.35 by Marc J. Mazerolle), to build model selection tables to identify the most parsimonious model for each fishery . After models were selected using AICc , I assessed the goodness of fit for the final selected models using simple linear regression to evaluate the significance of each parameter. I plotted the selected models residuals vs. the fitted data and the scale location to identify heteroscedasticity and a normal Q Q plot and a Results Plotting each of the 10 predictive variables against angler effort per hectare showed the relationship between the magnitude of fishing effort and the predictive variables that accounts for lake size (Figure 2 1, 2 2). For both Florida Bass and Black Crappie, the relationships exhibited few distinct trends. The proportion of littoral area appeared to have an inverse relationship with effort per hectare and catch per effort; showing a slightly dome shaped relationship in the Florida Bass fishery with high variability (Figure 2 2).
19 c score for the Black Crappie fishery (Table 2 2). This consisted of three parameters, the model inter cept, catch per effort, and the proportion littoral area . The model had an AIC c weight of 0.8 , suggesting it was the only model with substantial support . The parameter for proportion of littoral area was significant (P < 0.0 1), and the coefficient suggeste d a positive relationship with angling effort (Table 2 4). The c atch per effort parameter was positive and marginally significant (P = 0.07) suggesting that it was still important . Plotting the residuals against the predicted values, there were some issues with outliers in the data specifically associated with very high period of catch in a single lake (Figure 2 3). Plotting the residuals versus the leverage shows that much of the data was concentrated around a leverage of 0.05 but the variability in the Pe 2 3). Thus, the model assumptions were determined to be adequately met. In the Florida Bass fishery, the mo del representing aquatic structure in the lake was chosen by second order model selection with an AIC c wei ght of 0.75 (Table 2 3). Using this model, the proportion of littoral area positively related to fishing effort (P < 0 .01, Table 2 5), and the proportion of developed shoreline was not significant (P = 0.15) . Plotting the residuals against the predicted va lues, there was negligible evidence for bias at very low fitted values due to a few outliers within the lakes (Figure 2 4). Plotting the residuals versus the leverage show ed uals were highly variable (Figure 2 4) . Thus, for Florida Bass the only support was that aquatic habitat characteristics were related to fishing effort. Discussion I found that the angling effort dynamics of the study lakes may differ greatly from those that have been assessed in other regions. My results did not support hypotheses that many common non catch related predictors (i.e. Facilities, Seclusion, and Travel Cost) would influence
20 angling effort per hectare in Florida Bass or Black Crappie fisherie s. I found evidence in support of one non catch related hypothesis suggesting that aquatic structure could be an important predictor of angling effort within Florida Bass fisheries. Previous studies have found non catch related variables drive angler site selection; often independent of expected catches (Ben Akiva and Morikama 1990 ; Morey and Waldman 199 8 ; Jakus and Shaw 1999 ; Banzhaf et al. 2001). Because anglers have been known to seek social benefits from recreational angling unrelated to actually cat ching fish (Ditton et al. 1992), the weak relationship found in this study is an important conclusion f or recreational angling managers . Alternatively, I found little evidence for non catch related factors influencing angling effort within the Black Crappie fisheries, and instead found support for the expected catch model. Some past studies have suggested that expected catches can be the most important driver of angling effort dynamics (Parsons et al. 2000 ; Schuhmann and Schwabe 2004 ; Beardmore et al. 2011). This is especially true when the life history characteristics of the fish in the fishery are consis tent with a harvest oriented target species (Arlinghaus 2006) . For these anglers, expected catches can drive site choice so strongly that even the potential limitation of expected catch through reductions in bag limits will cause angler effort shifts to ot her fisheries (Beard et al. 2003). The methods used to evaluate fishing site selection vary widely, and in this study I used a revealed preference approach. This approach focuses on associating actual measures of angling effort with measures of potential predictor variables relative to that particular fishery (Hunt 2005). Alternatively, the stated preference methods focus on using surveyed angler opinions to identify hypothetical fishing behaviors relative to potential predictor variables important to the m (Hunt 2005). I chose to use the realized preference approach over the stated preference because
21 it takes advantage of previously collected monitoring studies commonly conducted by state and federal agencies and does not have the biases of angler intentio ns relating to actual actions (Fenichel et al. 2012) . However, the disadvantage of this approach is that I could not identify specific mechanisms causing angler site choice, only variables that were correlated with fishing effort per hectare. Th is inabili ty to differentiate between mechanisms is a limitation because some of my predictor variables for fishing effort could have resulted from different mechanisms . For example, the proportion of littoral area was a significant parameter and included in the two c scored models for both species. Within my hypothesized models, we included proportion of littoral area in catch and non catch related models since its perception can differ among anglers (Hunt 2005). Jones and Lupi (1999) found shoreline aest hetics to be positively related to angler site selection, but Schuhmann and Schwabe (2004) also suggested that anglers can associate increased structure in a lake with increased catches. While it is difficult to discern exactly how littoral habitat is bein g perceived by anglers taking part in these fisheries within the revealed preference framework, it does strongly suggest that available emergent and floating vegetation is very important to recreational anglers when choosing where to fish in Florida . Alter nately, the proportion of littoral areas also influences fish recruitment, and thus, could also be influencing angler expected catches . Thus, more work using stated preference methods should be conducted in the future to identify the mechanisms of angler s ite choice. One of the more interesting findings of this study was the weak relationships between predictive variables like travel costs or available facilities and angling effort in my study lakes. This may be attributed to the lake rich, population rich nature of central Florida, where my study lakes were located. There were over 18 million people in the state of Florida at the time of the
22 2010 United States Census and this number is expected to grow to 20 million by 2015 (U.S. Census Bureau 2013). This large number of people resides in a relatively small geographic area in comparison to many previous studies, some of which were conducted in areas with a clear gradient from urban centers to remote lake access (Hunt et al. 2011 ; Post and Parkinson et al. 2 012) . This would likely dilute the signal that many of the socio economic factors could exert through a realized preferences approach. For instance, travel cost may have a large impact on anglers realized preferences in fishing, but many of the major popul ation centers in Florida are within a reasonable daily driving distance from each other. This means that as travel costs begin to increase away from one urban center, the water bodies then begin moving closer to another . Thus, travel costs may not be an ef fective predictor of fishing effort in lake rich landscapes with multiple large urban centers . The unique regional dynamics become an important consideration with increasing numbers of angling opportunities that may decrease the impact and signal that ca tch and non catch related factors have on realized effort within a particular fishery. Thus, realized effort patterns may not adequately express the importance of these predictive factors compared to a data rich joint assessments of preference like those d one by Ben Akiva and Morikawa (1990). Future studies attempting to predict variables of importance in the site selection of anglers in lake rich systems should invest in survey techniques like long term angler diaries to associate the stated preference app roaches with fishery specific angler typologies. By using these two forms in a joint preference approach, the assessment benefits from the validity of correlating actual behaviors with the stated intentions of anglers (Hunt 2005). The results of thi s study did not suggest that many non catch related factors are not important in lake rich systems. Rather, it suggested that the relationship between angler site
23 selection and predictive variables may be more complicated in some geographical locations tha n those used in previous studies. This was an important result of this study and shows that to fully understand these complicated relationships requires a concerted effort with managing bodies conducting long term creel monitoring programs and researchers attempting to understand what drive angling effort dynamics in a social benefit driven activ ity like recreational fishing.
24 Table 2 1. List of hypotheses tested through GLM model fits and the literature used to select the appropriate components included . Hypothesis Components Literature Citation Expected Catch CPUE+Proportion Littoral Area Morey and Waldman 1998 Schuhmann and Schwabe 2004 Aquatic Structure Proportion Littoral Area +Prop. Developed Shoreline Peters et al. 1995 Jones and Lupi 1999 Seclusion Prop. Developed Shoreline+People per 50km+People per 100km+People per 150km Banzhaf, Johnson, and Mathews 2001 Travel Cost Cost to Park+People per 50km+People per 100km+People per 150km Ben Akiva and Morikama 1990 Hunt 2007 Facilities Quality of Restroom+Parking Spots+Boat Ramps+Cost to Park Morey and Waldman 1998 Jakus and Shaw 1999
25 Table 2 2. Second or der model selection (AICc) for Black Cr appie hypotheses. Representation Model k AICc AICc Wt LL Expected Catch CPUE+Proportion Littoral Area 3 423.6 0.0 0.8 207.6 Aquatic Structure Proportion Littoral Area+Prop. Developed Shoreline 3 425.9 2.2 0.2 208.7 Seclusion Prop. Developed Shoreline+People per 50km+People per 100km+People per 150km 5 544.4 120.7 0.0 265.8 Travel Cost Cost to Park+People per 50km+People per 100km+People per 150km 5 550.2 126.5 0.0 268.7 Facilities Quality of Restroom+Parking Spots+Boat Ramps+Cost to Park 5 638.8 215.2 0.0 313.1 *All models were te sted assuming normal error distributions and included the model intercept as a parameter.
26 Table 2 3. Second order model selection (AIC c ) for Florida Bass hypotheses . Representation Models k AICc AICcWt LL Aquatic Structure Prop. Developed Shoreline+Proportion Littoral Area 3 420.9 0.0 1.0 206.2 Expected Catch CPUE+Proportion Littoral Area 3 435.6 14.7 0.0 213.6 Seclusion Prop. Developed Shoreline+People per 50km+People per 100km+People per 150km 5 498.3 77.4 0.0 242.8 Travel Cost Cost to Park+People per 50km+People per 100km+People per 150km 5 542.5 121.6 0.0 264.9 Facilities Quality of Restroom+Parking Spots+Boat Ramps+Cost to Park 5 634.5 213.6 0.0 310.9 *All mode ls were tested assuming normal error distributions and included the model intercept as a parameter.
27 Table 2 4. AIC c selected model (Expected C atch) summary values for Black Crappie fishery. Parameter Coefficient Standard Error P Value Intercept 0.64 0.52 0.23 Catch Per Effort 0.59 0.32 0.0 7 Proportion Littoral Area 3.84 1.32 <0.01 * *Parameters with p values less than or equal to 0.05 are marked with * Table 2 5. AIC c selected model (Aquatic Structure ) summary values for Florida Bass fishery. Parameter Coefficient Standard Error P Value Intercept 1.03 0.41 0.01 Proportion Littoral Area 1.83 0.67 <0.01 * Proportion Developed Shoreline 3.00 1.25 0.15 *Parameters with p values less that 0.05 are marked with *
28 Figure 2 1. The ten predictive variables used to create representative models in this study, plotted against Florida Black Crappie angling hours per hectare. (Note: Each data point is a single lake year and is assessed independent from each other. People per 50,100,150 kilometers are plotted in hundreds of people. )
29 Figure 2 2. The ten predictive variables used to create representative models in this st udy, plotted against Florida Bass angling hours per hectare. (Note: Each data point is a single lake year and is assessed independent from each other. People per 50,100,150 kilometers ar e plotted in hundreds of peopl e .)
30 Figure 2 3. Diagnostic assessments of expected catch model for Black Crappie .
31 Figure 2 4 . Diagnostic assessments of aquatic structure model for Florida Bass.
32 CHAPTER 3 DECREASING CAPTURE VULNERABILITY OF FLORIDA BASS Micropterus floridanus WITH EXPOSURE TO CATCH AND RELEASE ANGLING. Globally, recreational fisheries constitute a major human use of many freshwater and marine systems (Arlinghaus et al. 2002; Idhe et al. 2011; Arlinghaus et al. 2013) and are important from a socioeconomic pers pective (Finn and Loomis 2001; Welcome et al. 2010) . Socioeconomic value of recreational fisheries includes both economic impacts (e.g. jobs/revenue from fishing related purchases) and socioeconomic benefits in the form of satisfaction experienced by angle rs (Propst and Gavrilis 1987) . Angler satisfaction is resultant from both catch related and non catch related factors (Finn and Loomis 2001; Hunt 2005; Arlinghaus 2006) . Catch related components (e.g., catch per unit effort, CPUE) are usually more variable between angling events than non catch factors, and thus, often drive overall changes in angler utility (Arlinghaus 2006) . primary metrics targeted by fisheries managers for improvement t hrough stock enhancement, habitat manipulations, or harvest regulations (Schramm et al. 1998; van Poorten & Post 2005; Askey et al. 2006) . Managers seeking to maximize CPUE often implement strategies designed to sustain high fish densities, such as stringe nt size limits or mandatory catch and release (Arlinghaus et al. 2007) . Catch and release regulations can improve fish abundance as long as discard mortality is not too high (Coggins et al. 2007), but direct increases in angler CPUE requires that catchabil ity (i.e., the fraction of the fish stock caught per unit effort) remains constant . Potential benefits of catch and release or similarly restrictive regulations may not be realized if angler CPUE changes after exposure to angling (i.e. if fish become les s catchable with increasing fishing effort) . Modifications of behavior that result in better avoidance of capture, or
33 other strategies that do not directly l imit angler effort (Cox and Walters 2002; Askey et al. 2006) . This has been empirically shown in several fisheries (Askey et al. 2006; Ward et al. 2013) . It is not known how ubiquitous (across species or regions) such effects are, though some studies sugge st they are certainly possible in other fisheries (Clark 1983; Spaet et al. 2010) . Further, it is not known if potential fishing induced changes to catchability may be mediated by altered angler behavior (i.e. anglers adopting novel fishing strategies to c . The potential for fish to become invulnerable after capture is well described by the vulnerable pool dynamic framework (Cox and Walters 2002). This framework suggests that within a population fish can reside in a group that is invul nerable to angling and a group that is vulnerable. There is defined natural rate of exchange between these two states that is constantly changing depending on seasonal effects, behavioral change, and community interactions. The introduction of angling infl uences these exchange rates by forcing behavioral or spatial changes in fish, thereby temporarily removing fish from the vulnerable population . Thus, understanding the exchange rate between vulnerable and invulnerable pools of fish provides a measure of ho w fishing effort is likely to influence angler catch rates. Given the implications for management and economies, it is surprising that so many gaps exist in our knowledge of how sport fish behaviorally respond to exposure to fishing . Thus, there is a need to evaluate responses of fish catchability when exposed to angling . This could influence angler CPUE and ultimately the efficacy of many recreational fisheries management strategies on improving angler satisfaction through catch related metrics. I evaluat ed how angler CPUE and fish catchability changed after exposure to angling. I used a fishing experiment on a naÃ¯ve Florida Bass Micropterus floridanus population to assess
34 whether angler CPUE changes through time after fish are exposed to fishing . The obje ctives of this chapter were to:  evaluate whether angler CPUE and thus fish catchability change when a Florida Bass population is exposed to catch and release angling,  evaluate whether changes in CPUE and catchability were similar for two different lures types that exhibit markedly different sensory cues to the fish, and  assess the potential for fishing effort management to improve angler catch rates . Results of this study have strong implications for the ability of fishery managers to improve an gler CPUE with regulations that protect fish from harvest and/or limit angler access. Methods Capture Recapture Population E stimates The fish population at Devils Hole Lake is closed to immigration or emigration between neighboring water bodies . Florida Ba ss abundance was estimated using capture recapture methods, using fish captured by shoreline electrofishing and through hook and line fishing during our fishing experiment (below) . Capture events were conducted via electrofishing surveys of the littoral zo ne of the lake using a Smith Root Â® 9.0 generator powered pulsator (GPP) electrofishing unit with a boom mounted electrode . All bass collected were measured for total length (TL; mm), and marked by a pelvic fin clip . Bass were implanted with a uniquely numbered passive integrated transponder (PIT) tag (Biomark Â®) inserted into the body cavity between the pelvic fins (Harvey and Campbell 1989) . Recapture events for population estimation included 600 s transects covering the entire perimeter of the lake. U pon recapture bass were scanned for the presence of a PIT tag. The Jolly Seber population estimator for multiple recapture events was used to estimate the population size of Florida Bass at Devils Hole Lake (Jolly 1965; Seber 1982) .
35 Experimental Angling owned, mesotrophic lake and has received very little fishing effort over the past decade prior to our experiment where we introduced experimentally controlled hook and line fishing . Two anglers fished six hours a day for thre e days each week over four total weeks, with no more than four days between angling events . Two lure types were fished on 9.1 kg test braided line, with 1.5 m fluorocarbon leaders of 9.1 kg test. High test line was used to minimize the capture time for eac h fish in attempt to minimize the stress of being caught, as Cooke et al. (2002) suggested its importance in recovery time to pre angling heart rates . The active lure was a chrome and black lipless crank bait (Rat L TrapÂ© Bill Lewis Lures), which is shiny and exhibits a loud rattle when retrieved . The finesse style bait was a 125 mm soft plastic stick bait color plum with emerald flake (Senko, Gary Yamamoto Custom BaitsÂ©) fished weightless with a 3/0 offset worm hook . Anglers alternated fishing each lure ev ery hour so that each lure type was fished by each angler equally throughout the day. Anglers were encouraged to fish in multiple styles and in all areas of the lake in an attempt to maintain high catch rates over the full length of the study . Angled bass were measured for total length (TL; mm) and received a pelvic fin clip. If they did not already have a unique PIT tag, a tag was inserted following the same protocols as the mark recapture portion of the study. Bass were immediately released and monitored until they swam of their own volition. If an angling induced injury occurred that may have influenced chose not to hold caught fish in pens to monitor for potential hooking mortality to minimize additional stress on the fish, and because hooking mortality rates for black basses are generally low (Mouneke and Childress 1990).
36 To evaluate whether the fishing effort we used was realistic for open access fisher ies, I effort in other Florida lakes . Roving creel surveys were conducted by the Florida Fish and Wildlife Conservation Commission, which were typically conducted d uring peak fishing season (January May) each year . We estimated weekly fishing effort for Florida Bass for a wide range access recreational fisheries. Analysis I evaluated how angler catch per unit effort (CPUE) changed when a naÃ¯ve Florida Bass population was exposed to recreational angling. I estimated CPUE as, where effort ( E ) was angler hours extended with each lure each sample day. I used the standard equation for catch ( C ), where ( q ) is the catchability coefficient and re ordered it to solve for catchability using CPUE calculated as, where was the previously estimated adult Florida Bass abundance of Devils Hole Lake . The CPUE in the fished condition ( CPUEf ) was estimated as a function of cumulative catch (Leslie Method) for both lure types as,
37 where is CPUE in the unfished condition and is a reduction factor that described the shape of the relationship. For each lure type, the expected CPUE f was fitted to the observed CPUE assuming a log normal probability distribution. I calculated log likelihoods for each model and maximized them by c hanging the parameters , , and normal distribution, using the optimization program (SOLVER in Excel). I used AICc model selection to evaluate whether the temporal changes in CPUE were best expl ained by separate models for each lure type, or one combined model using the total log likelihood . To identify the potential for regional effort based management to influence angler catch rates, I predicted the vulnerable pool exchange rates within the l ake. For this analysis, I chose to combine the separate lure catches into a single CPUE time series in order to better represent a realistic fishery . I calculated the vulnerable pool ( V ) over the cumulative fishing effort of the study. I estimated CPUE usi ng the vuln erable pool framework and initialized the differential equation given by Cox and Walters (2002) as, where equals the number of fish in the vulnerable pool, k 1 refers to the exchange rate from the invulnerab le pool to the vulnerable and k 2 refers to the exchange from the vulnerable to the invulnerable pool. I estimated the daily rate of change in the vulnerable pool using where is the estimated total adult population of bass, the k 3 parameter is composed of the catchability coefficient multiplied by E t , the number of catch and release angling hours each day. Change from the initialized V 1 was calculated using the Euler differential approximation as,
38 and subsequent change after that was calculated using the Adams Bashforth differential approximation as I then fit observed CPUE from this study to predicted CPUE estimated as, assuming lognormal probability distribution and by minimizing the negative log likelihood using the optimization program Solver in excel by changing the variable k 1 , k 2 , and q . The final model parameters were used to simulate fishing effort clo sures representative of possible management actions . The first management action was to maintain an open access fishery that received an average of 36 angling hours per week, the second was to maintain a constant fishery but limit entry to only 18 angling hours per week, and lastly, conducted a limited access fishery that had a closure period of 30 days before returning to open access (36 angling hours per week). Results The angling study took place from June 3 27 th 2013 with a total of 12 sampling days and 144 cumulative angler hours, amounting to an average of 3.18 angling hours per hectare per week. This was in the upper third of average peak season bass fishing effort for Florida lakes (Table 3 1). Thus, the stud y exerted a high level of fishing effort, but the levels were not dissimilar from values found in real open access fisheries . A total of 260 (75%) Florida Bass were captured from an estimated population size of 347 (344 388 95% CIs) adult Florida Bass ( . Thus,
39 approximately 87 (25%) fish were never angled . Of the 260 individuals, individual fish were recaptured 27 times (Table 3 2). A total of ten fish were angled but did not receive a PIT tag due to sampling error . This comb ined to 297 total catches by the anglers . The length distribution of angled bass was dominated by individuals between 26.0 and 36.0 cm (85%; Figure 3 2 ) . Less than 4% of all bass caught experienced any observable injury as a result of angling. The two lures used had different rates of success . The finesse lure had the highest number of catches with a total of 160 fish, whereas the active lure res ulted in 90 fish angled (Table 3 2 ) . The active lure also had the lowest incidence of recaptures with only tw o bass being angled twice . The finesse lure had a total of 25 recaptures with the majority of those (n=20) being caught twice, and two bass being caught multiple times (i.e., three times and four times respectively; Table 3 2 ). After the naÃ¯ve Florida Ba ss population was exposed to recreational angling, CPUE declined with cumulative angler effort (hr) for both lure types. The finesse lure expressed a linear decline in CPUE dropping from initial 1.77 fish per angler hour to 1.17 fish per angler hour a t 144 cumulative hours (Figure 3 2B) . Changes in c atchability also declined with an initial q of 0.005 proportion caught per angler hour and a final observed q of 0.003 proportion caught per angler hour . The active lure showed an exponential decline in CPUE fro m initial values of 2.33 to a final observed value at 144 cumulative angler hours o f CPUE 0.17 (Figure 3 2A, C ) . Catchability declined from 0.0064 to 0.0005 for the active lure . Thus, angler CPUE declined by an order of magnitude for the active lure, but o nly dropped by approximately half for the finesse lure . Model selection suggested that lure types had a different effect on CPUE over cumulative angler hours (Table 3 3). The active lure showed a more drastic rate of decline in CPUE and catchability over c umulative angler hours and also had the lowest number of recaptures overall
40 compared to the finesse lure (Table 3 1; Figure 3 2) . Thus, catchability dropped more dramatically for the active lure than for the finesse lure, but q changed after consistent exp osure to angling for both bait types. I was able to fit the observed total catch per effort from the study to the predicted CPUE with initial predicted CPUE beginning at 3.36 and falling to 1.36 fi sh per angling hour (Figure 3 3 ). The forward simulated p redicted catch per effort that maintained effort level of the experiment (36 hours per week) predicted CPUE to stagnate at 1.3 fi sh per angling hour (Figure 3 3A ). Decreasing the effort level by half, (18 hours per week) predicted that CPUE available to an glers would increase to ~1.9 f ish per angler hour (Figure 3 3B ). The third simulation using the cyclical strategy predicts available CPUE to increase to ~3.2 fish per angler hours after the 30 day restricted period, a level similar to the naÃ¯ve state of th e lake (Figure 3 3C ). Discussion My results showed that catchability (and thus angler CPUE) falls quickly for naive fish after exposure to fishing. This suggests that the ability to maintain high angler CPUE by maintaining high population abundance coul d be compromised, and thus, management actions that increase abundance may not proportionately influence angler CPUE . Some past studies found incidental evidence for decreased vulnerability to angling after an initial angling event (Anderson and Heman 1969 ; Beukema 1970) . Askey et al. (2006) showed empirically similar overall declines in CPUE with Rainbow Trout Oncorhyncus mykiss using only finesse style fly pattern lures . In my study, angler CPUE quickly fell below average levels for the state of Florida w ith the active bait (0.58 fish/angler hours , Florida Fish and Wildlife Conservation Commission, unpublished data). This is important because most fishery managers operate under the assumption that angler catch rates can be improved through management actio ns that increase
41 fish abundance . If fish behaviorally learn to avoid capture, then angler CPUE may not respond to changes in fish abundance. This infers that maintaining high catch rates may require effort limitation regulations (Cox and Walters 2002 ; Cox and Walters 2003) . The results of my simulation supported the results of Cox and Walters (2002 ; 2003). My study showed that in Devils Hole Lake long term constant effort is likely to result in a stagnation of available catch rates at effort levels even lower than the 36 angling hours per week used in the study. Thus, even with stringent effort restri ctions, anglers will only experience moderate catch rate increases. Considering how recreational fisheries in North America have experienced moderately high angling effort for many years, the baseline of what is considered quality catch rates may have shif ted to a lower level compared to truly pristine systems (Ward et al. 2013). This suggests that an active management strategy that provides a few cyclically restricted lakes amongst a region of lakes could provide higher net regional angler satisfaction . Th is becomes especially important with the recognition that substitution of anglers in and out of a fishery is an active component of recreational fishing dynamics (Hunt et al. 2007). Increasing the overall available regional catch rates may help maintain hi gher interest in the fishery over time. Mechanisms for the decline in catchability could be behavioral and/or changes in fish spatial distribution. Thus, decline in vulnerability to angling that I showed could be due to 1) fish moving to areas where they were not exposed to angling after the initial fishing days, 2) less aggression in overall predatory behavior within the system, or 3) a learned response to avoid the lures . I believe that it is unlikely fish moved to areas where they were not exposed to a ngling, because the lake had a maximum depth of 4.5 m. and anglers were able to cover the entire lake in each angling day. It is possible that the declining angler CPUE was due to a shift in overall level of aggression. If the act of being caught drives th e individual fish to be more risk averse and feed
42 less within the system, we would likely see the overall decrease in angling catchability depicted in the study but the rate at which catchability decreases should be similar between the two lures . Thus, I f eel the most credible hypothesis was that fish learned to avoid the lures, and the degree of fish learning was influenced by lure type. Few fish were recaptured overall but recaptures occurred at much lower rates with the active lure compared to the fines se lure . Because the behavior modification occurred at a much higher rate in the active lure with more sensory cues available to the fish, than the cryptic finesse lure, it is likely that the degree of learning was influenced by lure type . This process is not dissimilar to cases where aposematic prey illicit an aversion to being predated upon by inducing an undesirable experience, often through toxins (Matthews 1977). Predators recognize these species through specific color or patterns and it has been shown that naÃ¯ve predators can develop this recognition much more quickly with conspicuous prey over cryptic prey (Guilford 1986). These similarities introduce strong evidence that the learned behavior modification is developed specifically with experiences ass ociated with individual lures, and thus, in this case individual fish learning is the most likely driver of decreasing angler catchability independent of abundance. Because I found evidence that declines in CPUE, catchability, and a lack of recaptures we re lure specific, anglers may be able to moderate for their own loss in CPUE by changing lures often or using subtle baits that may be less likely to illicit a learned response. Wilde et al. (2003) showed that lure size could significantly influence catch rates and the size of fish caught behavioral modification fro m the fish (Wilde et al.2003; Rapp et al. 2008). It is unknown how
43 this learned response would vary when many different lures are used by anglers with a wide range of skill, which is typical of recreational fisheries in practice . However, my study showed c lear evidence that angler CPUE declines after exposure to lures , and the vulnerability of fish to specific lures can be markedly different. Lakes that incur especially high angling effort would theoretically also increase exposure of fish to a larger varie ty of lures. This could have implications to an economically important tackle industry. Considering the differing influences that my study lures played on the vulnerability of fish, understanding empirically how these influences occur and the duration they may influence a fishers behavior must be important to manufacturers, anglers, and managers alike. Further, the decline in CPUE happened very rapidly for the active bait, indicating that minimal angler effort is required for lower catch rates . Managing a ngling effort in individual lakes or on a broad regional scale could be required to prevent overfishing in many recreational fisheries (Cox et al. 2003 ; Coggins et al. 2007 ; Allen et al. 2012) . My results suggested that even a low amount of fishing effort (e.g. 60 cumulative angling hours) caused angler CPUE to decline by half and vulnerability to decrease by almost seven times for the active lure . Thus, maintaining high angler catch rates may be extremely difficult if CPUE declines after relatively low eff ort. Ward et al. (2013) found no relationship between angler catchability and angler effort density or the proportion of fish released in open access rainbow trout fisheries. While counter to our findings, they suggested that this may be due to the establi shed nature of their lake fisheries. If much of the population has already experienced being caught and released, angler catchability may already be depressed to baseline low levels . This issue of a shifting baseline in catchability is supported in our stu dy by the quick drop in catchability with the active lure at a naÃ¯ve state and stagnation at mean statewide Florida Bass CPUE with low cumulative angling effort (E=60). It is
44 possible that angling effort restrictions as proposed in Cox et al. (2003) may me diate for a shifted baseline and thus, increase angler satisfaction for some anglers. I inferred behavior modification through fish learning as a likely mechanism for changes in catchability, but this study did not measure individual fish responses when e xposed to fishing. Studies to measure behavior and evaluate whether individual fish are in fact less likely to be responsive to fishing lures are needed to empirically draw this conclusion. Reactivity within individual fisheries to angling plays an importa nt role in vulnerable pool dynamics as pointed out by Cox and Walters (2002) and thus in resiliency of recreational fisheries. My study restricted anglers to only two lure choices, but future work should evaluate how CPUE and catchability changes with a va riety of lure types and anglers being free to change lures in attempt to maintain high catch rates. As my study showed, different lures may influence behavior modification at markedly different rates and assessing the sensory cues associated with these lur es types may be very important. An understanding of angler social norms within individual fisheries, in conjunction with an understanding of influence they have on fish behavior could provide a novel and important tool to managers. Additionally, future wor k should look empirically at the possibility of decreasing CPUE occurring due to decreased aggression in the individually caught fish as opposed to a learned behavior to avoid specific lure types.
45 Table 3 1 . A verage weekly e ffort per hectare of Fl orida lakes during peak season. Lake Mean Weekly Peak Season Angling Hours per Lake Hectare Lochloosa 0.64 Panasoffkee 0.80 Tohopekaliga 0.86 Sampson 0.91 Yale 1.00 Weohyakapka 1.12 Juniper 1.22 Dexter 1.28 Jackson 1.70 Starke 2.07 Medard 2.33 Ivanhoe 2.37 Merritts Mill 2.51 Rowell 2.60 Tarpon 2.71 3.18* George 3.59 Thonotosassa 3.61 Saddle Creek 4.54 Bear 4.63 Karick 5.19 Hurricane 5.48
46 Table 3 2 . Florida Bass angled during the study. Individual Fish Caught Lure Type One Two Three Four Active 88 (34%) 2 (<1%) 0 0 Finesse 148 (57%) 20 (8.0%) 1 (<1%) 1 (<1%) * Percentages in parentheses are based off 260 total individual fish caught. Table 3 3 . Model selection scores (AICc) between lure type and CPUE Model LL Parameters AICc 0.189 6 10.49 0 22.24 3 51.68 41.18
47 Figure 3 1. Length f requency distribution of the Florida Bass angled in Devils Hole Lake ( Note: Lure catches combined ).
48 Figure 3 2. Catch per unit effort (CPUE) of Florida Bass at cumulative angler catches using active and finesse lures. A) CPUE with active lure , B) CPUE with finesse lure . (Note: Obser ved values are represented by black diamonds . Dotted grey lines represent predicted relationship of exponential decline in catch per unit effort . )
49 Figure 3 3 . Catch per angler hour plotted over cumulative catches. (Note: Observed values are represented by black diamonds. Dotted grey line represents predicted relationship of exponential decline.)
50 Figure 3 4 . Simulated catch per unit effort plotted over time in angling days with alternate management strategies. A) constan t 36 angling hours per week, B) constan t 18 angling hours per week. C) 36 angling hours per week from day 1 27, 0 angling hours from day 28 58, and a return to 36 angling hours per week thereafter. (Note: Observed values are represented by black diamonds . D otted grey lines represent predicted relationship of vulnerable pool exchange .)
51 CHAPTER 4 CONCLUSIONS My goal was to better understand the implications and potential for regional effort based management to positively impact freshwater angler satisfaction in the state of Florida. I chose to approach these using three separate objectives within Florida lake fisheries. I first attempted to quantify factors that may influence fishing effort within lake fisheries in Florida. I then identified how high effort in a lake fishery might negatively impact angler satisfaction through changes in catchability, even independent of changes in target species population abundance. Lastly, I showed how the implications of decreasing capture vulnerability in a system, independ ent of population abundance, may lead to a shifted baseline in angler catches and this may be mediated for through regional effort based management. To quantify the catch and non catch related influences on angler site selection, I assumed that fishing e ffort in lake fisheries was representative of the realized preferences of anglers choosing where to fish. The data were collected by the Florida Fish and Wildlife commission through roving and point access creel surveys. I found that expected catch may pla y a role in angler site selection in Florida for the harvested oriented Black Crappie fishery and lake aesthetics may play a role in the catch and release oriented Florida Bass fisheries. I did not find some common non catch related influences to be import ant in Florida, and I believe that this is representative of the lake rich nature of the landscape in Florida combined with the geography of human population centers. Future work to identify site choice mechanisms will require long term consistent monitori ng of effort in Florida fisheries and an additional focus on the stated preference of anglers taking part in the fisheries. The complicated nature of Florida fisheries suggests that managers would benefit from continuing this work to develop a joint prefer ence representation of angler site selection.
52 My angling experiment tested the hypothesis that catchability was constant after fish exposure to catch and release angling. My results showed that caught and released Florida Bass will modify their behavior t o avoid being caught and this modification occurs at different rates with lures that exhibit different sensory cues. This avoidance has a direct implication on the available catch rates of Florida Bass anglers and in turn angler catchability. This implicat ion is very important when considering that most freshwater fisheries management is focused on increasing fish abundance, which is assumed to increase angler satisfaction . More work is needed to evaluate how fishing effort and fish abundance combine to inf luence catchability in recreational fisheries. To identify how regional effort based management may improve catches and this angler satisfaction, I conducted a simulation of the vulnerable pool dynamics within our experimental lake system . I showed that based on the exchange rates in our fished system and the level of effort exerted, angler catch rates would be depressed to a level indicative of most Florida Bass fisheries in the state. Decreasing this effort was predicted to minimally increase CPUE, but short term cyclical closures of the fishery was able to bring catch rates back to similar levels to a naÃ¯ve fishery. My thesis indicates that spatial management strategies within freshwater fisheries in Florida could shift management focus from objectives of improving population abundance and size structure to understanding how anglers ultimately impact their own satisfaction. This does not suggest that these classic management strategies are not still important to stock management. The implications of thes e studies point more towards the potential to improve regional satisfaction within the state of Florida by sacrificing minimal lake specific angling opportunities in the short term for long term region wide benefits in catch rates. This is not a new sugges tion
53 as Cox and Walters (2002) suggested that closed access concepts could improve overall angling quality in a region and decrease the risks of over fishing. This is also being utilized successfully in marine systems through marine protected areas, season al closures, or by requiring additional cost to entering the fishery (i.e. tarpon tags) . However, for lake rich landscapes like Florida, maintaining a few sites with effort restrictions could substantially improve angler CPUE at those sites, which could be used as a tool to recruit new anglers, or provide unique angling opportunities to a few anglers. Future work in testing these concepts should look to improve on the replication of these concepts in the Southeastern US. The work presented here is based on an overly simplified design (e.g., only two fishing lures), and testing these concepts in real world situations where there are many more confounding factors within the system will show to what degree regional effort based management strategies can improv e angler satisfaction. It is important to understand that fully identifying the human dimensions within these fisheries are necessary to quantify the role that improving catches will have on angler satisfaction. I was able to show that catch rates may be i mportant to Florida anglers but more work must be done to understand how the impacts of alternate management strategies, like regional effort based management, will manifest in angler satisfactions. Managers should focus on improving overall angler satisfa ction, which requires more than increasing fish abundance. Stock abundance and structure are very important not only to recreational fisheries but overall environmental health. In terms of recreational angling management, these are only underlying symptoms of the true product of recreational fisheries; the satisfaction of those taking part in the activity.
54 LIST OF REFERENCES Akaike, H. 1973. Maximum likelihood identification of Gaussian autoregressive moving average models. Biomet rika 60(2): 255 265 All en, M. S., N. M. Ahrens , M. J. Hansen, and R. Arlinghaus. 2013. Dynamic angling effort influences the value of minimum length limits to prevent recruitment overfishing. Fishe ries Management and Ecology 20: 247 257 Anderson, R., and M. L. Heman. 1969. Angling as a factor influencing catch ability of largemouth bass. Transactions of American Fisheries Society 98: 317 320 Arlinghaus, R., T. Mehner, and I. G. Cowx. 2000. Reconciling traditional inland fisheries management and sustainability in industrializ ed countries, with emphasis o n Europe. Fish and Fisheries 3: 261 316 Arlinghaus, R. 2006 . On the apparently striking disconnect between motivation and satisfaction in recreational fishing: the case of catch orientation of German anglers . North American Jour nal of Fisheries Management 26: 592 605 Arlinghaus R., S. J. Cooke , J. Lyman , D. Policansky , A. Schwab , C. D. Suski et al. 2007. Understanding the complexity of catch and release in recreational fishing: an integrative synthesis of global knowledge fr om historical, ethical, social, and biological perspectives. R eviews in Fisheries Science 15: 75 167 Arlinghaus R., S. J. Cooke , and W. Potts. 2013. Towards resilient recreational fisheries on a global scale through improved understanding of fish and fishe r behavior . Fishe ries Management and Ecology 20: 91 98 Askey , P. J., S. A. Richards , J. R. Post, and E. A. Parkinson. 2006. Linking angling catch rates and fish learning under catch and release regulations. North American Journal of F isheries Management 26: 1020 1029 Banzhaf, M. R. , F. R. Johnson , and K. E. Mathews. 2001. Opt stated preferences. Pages 157 177 i n J. Bennett and R. Blamey, e d itors. Th e choice modelling approach to environmental valuation . MIT Press, Northampton , Massachusetts Beard, D. T., S. P. Cox , S. R. Carpenter. 2003. Impacts of daily bag limit reductions on angling effort in Wisconsin Walleye lakes. North American Journ al of Fisheries Management 23(4): 1283 1293 Beardmore, B., W. Haider, L. M. Hunt , an d R. Arlinghaus. 2011. The importance of trip context for determining primary angler motivations: Are more s pecialized anglers more catch oriented than previously believed? North American Jo urnal of Fisheries Management 31(5): 861 879
55 B en Akiva, M. E. and T. Morikawa. 1990. Estimation of switching models from revealed preferences and stated intentions. Transportation Research. 24(6): 485 495 Beukema, J. J. (1970) Acquired hook avoidance in the pike Esox lucius L. fished with artificial and natural ba its. Journal of Fish Biology 2: 155 160 Bryan, H. 1977. Leisure value systems and recreational specialization case of trout fisherman. Jo urnal of Leisure Research 9(3): 174 187 Ca rpenter, S. R., and W. A. Brock. 2004. Spatial complexity, resilience, and poli cy diversity: Fishing on lake rich landscapes. Ecology and Society 9(1) Clark, Jr. R. D. 1983. Potential effects of voluntary catch and release of fish on recreational fisheries. North American Journal of Fisheries Management 3:306 314 Coggins, L. G., M. J. Catalano , M. S. Allen, W. E. Pine , and C. J. Walters. 2007. Effects of cryptic mortality and the hidden costs of using length limits in fishery manage ment. Fish and Fisheries, 8(3): 196 210 Cooke, S. J., and I. G. Cowx. 2004. The role of recreational fishing in global fish crises. Bioscience 54(9) : 857 859 Cooke, S. J., J. F. Schreer , D. H. Wahl , and D. P. Philipp. 2002. Physiological impacts of catch and release angling practices on largemo uth bass and smallmouth bass. Black Bass 2000: ecology, conservation, and management. American Fisheries Society Symposium 31:489 512 Cox S. P. and C. Walters. 2002. Modeling exploitation in recreational fisheries and implications for effort management on British Columbia rainbow trout lakes. North American Journal of Fi sheries Management 22: 21 34 Cox, S. P., C. Walters, and J. R. Post. 2003. A model based evaluation of active management of recreational fishing effort. North American Journal of Fisheries Management 23 : 1294 1302 Ditton, R. B., D. K. Loomis , Choi, S. 1992. Recreation specialization : re conceptualization from a social worlds perspective. Journal of Leisure Research 24(1) : 33 51 Fedler, A. J. 1994. Understanding angler motivations in fisheries management. Fisheries 19(4) : 6 13 Fenichel, E. P., J. K. Abbott , and B. Huang. 2013. Modelling angler behaviour as a part of the management system: synthesizing a multi disciplinary literature. Fish and Fisheries 14 : 137 157
56 Finn, K. L., and D. K. Loomi s. 2001. The importance of catch motives to recreational anglers: The effects of catch satiation and deprivation. Human Dimensions of Wildlife 6(3) : 173 187 Guilford T. 1986. How do warning colors work? Conspicousness may reduce recognition errors in exper ienced predators. Animal Behavior. 34 : 286 288 Hansen, M. J., T. D. Beard , and S. W. Hewett. 2005. Effect of measurement error on tests of density dependence of catchability for walleyes in northern Wisconsin angling and spearing fisheries. North American Journal of Fisheries Management 25(3) : 1010 1015 Harvey, W. D., and D. L. Campbell. 1989. Technical Notes: Retention of passive integrated transponder tags in largemouth bass brood fish. The Progressive Fish Culturist 164 166 Holland, S. M., and R. B. Ditton. 1992. Fishing trip satisfaction: a typology of anglers. North American Journal of Fisheries Management 12(1) : 28 33 Hunt L. M. 2005 . Recreational fishing site choice models: insights and future opportunities. Human Dimensions of Wildlife 10 : 153 172 Hunt, L. M., B. N. Boots , and P. C. Boxall. 2007. Predicting fishing participation and site choice while accounting for spatial substitution, trip timing, and trip context. North American Journal of Fisheries Management 27(3) : 832 847 Hunt L. M. , R. Arlinghaus , N. Lester and R. Kushneriuk. (2011) . The effects of regional angling effort, angler behavior, and harvesting efficiency on landscape patterns of overfishing. Ecological Applications 21 : 2555 2575 Ihde, J. F., M. J. Wilberg , D. A. Loewe nsteiner , D. H. Secor, and T. J. Miller. 2011. The increasing importance of marine recreational fishing in the US: challenges for management. Fisheries Research 108 : 268 276 Johnston, F. D., R. Arlinghaus, and U. Dieckmann. 2010. Diversity and complexity of angler behaviour drive socially optimal input and output regulations in a bioeconomic recreational fisheries model. Canadian Journal of Fisheries and Aquatic Sciences 67(9) : 1507 1531 Jolly, G. M. 1965. Explicit estimates from capture recapture data with both death and immigration stochastic model. Biometrika 52 :225 247 Jones, C. A. and F. Lupi. 1999. The effect of modeling sustitute activities on recreational benefit estimates. The Jo urnal of Risk and Uncertainty 26(1) : 77 92 Matthews, E.G. 1977. Signal based frequency dependent defense s trategies and the evolution of mimicry. American Naturalist 11 : 213 222
57 Morey, E. R. and D. M. Waldman. 1998. Measurement error in recreation demand m odels: The joint estimation of participation, site choice, and site characteristics. Journal of Environmental Economics and Management 35 : 262 276 Mouneke, M.I. and M.W. Childress. 1994. Hooking mortality: a review for recreational fisheries. Reviews in Fisheries Science 2 : 123 156 Murphy, B. R., and D. W. Willis, editors. 1996. Fisheries techniques, 2nd edition. American Fisheries Society, Bethesda, Maryland. Odling Smee, L. and V.A. Braithwaite. 2003. The role of learning in fish orientation. Fish and Fisheries 4:235 236 Parkinson, E. A., J. Post , and S. P. Cox . 2003. Population level consequences of not limiting effort in a sport fishery. Fishery Report Department of Business, Industry & Resource Development, Northern Territory Government (67) Parkinson, E. A., J. R. Post , and S. P. Cox. 2004. Linking the dynamics of harvest effort to recruitment dynamics in a multistock, spatially structured fishery. Canadian Journal of Fisheries and Aqu atic Sciences 61(9) : 1658 1670 Parsons, G. R., A. J. Plantinga , and K. J. Boyle. 2000. Narrow choice sets in a random utility model of recreation demand. Land Economic, 76(1) : 86 99 Post, J. R., and E. A. Parkinson . 2012. Temporal and spatial patterns of a ngler effort across lake districts and policy options to sustain recreational fisheries. Canadian Journal of Fisheries and Aquatic Sciences 69(2) Post, J. R., L. Persson , E. A. Parkinson , and T. Van Kooten. 2008. Angler numerical response across landscape s and the collapse of freshwater fisheries. Ecological Applications 18(4) Post, J. R., M. Sullivan , S. Cox , N. P. Lester , C. J. Walters et al. 2002. Canada's recreational fisheries: The invisible collapse? Fisheries 27(1) : 6 17 Propst, D. B., and D. G. G avrilis. 1987. Role of economic impact assessment procedures in recreational fisheries management. Transactions of American Fisheries Society 116 : 450 460 Rapp, T., S. J. Cooke , R. Arlinghaus. 2008. Exploitation of Specialised Fisheries Resources: The Importance of Hook Size in Recreational Angling for Large Common Carp (Cyprinus carpio L.). Fisheries Research Schuhmann, P.W., and K. A. Schwabe. 2004. An analysis of congestion measures and he terogeneous angler preferences in a random utility mode of recreational fishing. Environemental and Resource Economic 13 : 21 32
58 Seber, G. A. F. 1982. The estimation of animal abundance and related parameters. Edward Arnold, London . Spaet, J. L. Y., S. T. Kessel , and S. H. Gruber. 2010. Learned hook avoidance of lemon sharks (Negaprion brevirostris) based on electroreception and shock treatment. Marine Biology Research 6 : 399 407 Sullivan, M. G. 2003. Active management of walleye fisheries in Alberta: Dil emmas of managing recovering fisheries. North American Journal of Fisheries Management 23(4) U.S. Census Bureau. (2013, August). State & C ounty Q uickfacts: Gainesville, Florida. Retrieved August, 2013, from http://quickfacts.census.gov Van Poorten, B.T., and J.R. Post. 2005. Seasonal fishery dynam ics of a previously unexploited rainbow trout population with contrasts to established fis heries . North American Journal of Fisheries Management 25 : 329 345 Walters , C. J and S. Martell . 2004 . Fisheries Ecology and Management . Princeton University Press, Princeton, New Jersey. Ward, H.G.M., Askey, P.J., and Post, J. R. 2013. A mechanistic u nderstanding of hyperstability in catch per unit effort and density dependent catchabilit y in a multistock recreational fishery. Canadian Journal of Fisheries and Aquatic Science 70(10) : 1542 1550 Welcomme , R. L., I. G. Cowx , C. Bene , S. Funge Smith , A. Halls and K. Lorenzen. (2010) Inland capture fisheries. Philosophical Transacti ons of the Royal Society B 365 : 2881 2896. Wilde G. R., K. L. Pope , B. W. Durham. 2003. Lure size Restrictions in Recreational Fisheries. Fisheries 28(6) : 18 26
59 BIOGRAPHICAL SKETCH Nicholas Cole was born in St. Joseph, Missouri. He was raised on a farm in Northwest Missouri and graduated from Savannah High School in 2005. He had developed an interest in natural resource conservation early on and conducted his undergraduate studies at the University of Missouri, earning a B.S. in fisheries and wildlife s ciences. He worked on numerous fisheries research projects as an undergrad with Missouri Department of Conservation and the U.S. Fish and Wildlife Service. He assisted in freshwater eco logy res earch with North Carolina State University and University of Nebraska before join ing the University of Florida.