<%BANNER%>

Using acoustic telemetry to estimate aatural and fishing mortality of common snook in Sarasota Bay, Florida

HIDE
 Title Page
 Dedication
 Acknowledgement
 Table of Contents
 List of Tables
 List of Figures
 Abstract
 Introduction
 Methods
 Results
 Discussion
 Appendix: Hightower model and Cormack-jolly-sebers...
 References
 Biographical sketch
University of Florida Institutional Repository

PAGE 1

USING ACOUSTIC TELEMETRY TO ES TIMATE NATURAL AND FISHING MORTALITY OF COMMON SNOOK IN SARASOTA BAY, FLORIDA By JASON P. BENNETT A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2006

PAGE 2

Copyright 2006 by Jason P. Bennett

PAGE 3

This document is dedicated to my supporti ng family and the memory of my father.

PAGE 4

iv ACKNOWLEDGMENTS I would like to thank my graduate adviso r, Dr. Bill Pine, and the rest of my graduate committee; Dr. Mike Allen, Dr. To m Frazer, and Dr. Robert Muller. I would also like to thank everyone from the following organizations: University of Florida, Department of Fisheries and Aquatic Sciences : Lauren Marcinkiewicz, Drew Dutterer, Mark R ogers, Greg Binion, Travis Tuten, Galen Kaufman, Vince Palitano, Nick Trippel Florida Fish and Wildlife Conservation Commission : Ron Taylor, Luiz Barbieri, Behzad Mahoudi, Alexis Trotter Mote Marine Laboratory : Nate Brennen, Dr. Kenneth Leber, Meagan Darcy, Jason Rock, Michelle Heupel, Colin Simpfendorfer, Dave Wilson, Brett Blackburn, Sarasota Bay Explorers I also thank Captain John Dixon, without whom we could not have caught any snook.

PAGE 5

v TABLE OF CONTENTS page ACKNOWLEDGMENTS.................................................................................................iv LIST OF TABLES............................................................................................................vii LIST OF FIGURES...........................................................................................................ix CHAPTER 1 INTRODUCTION........................................................................................................1 Study Species................................................................................................................2 Life History...........................................................................................................3 Stock Status in Florida...........................................................................................4 Project Justification......................................................................................................6 Fishery Dependent Methods to Estimate F and M................................................6 Fishery Independent Methods to Estimate F and M..............................................9 Parameter Estimation...........................................................................................10 2 METHODS.................................................................................................................15 Field Methods.............................................................................................................15 Study Site.............................................................................................................15 Tagging Procedure...............................................................................................15 Tag and Receiver Description.............................................................................16 Acoustic Array Network......................................................................................16 Array deployment and testing......................................................................16 Manual tracking............................................................................................18 Data Analysis..............................................................................................................18 Assigning Fish Fates............................................................................................18 Total Mortality Estimates....................................................................................22 3 RESULTS...................................................................................................................26 Field Results...............................................................................................................26 Tag Reception......................................................................................................26 Fish Detection......................................................................................................26 Data Analysis Results.................................................................................................28 Assigning Fates...................................................................................................28

PAGE 6

vi Secondary Analysis.............................................................................................31 4 DISCUSSION.............................................................................................................46 Acoustic Array, Tagging and Detection.....................................................................46 Model Assumptions....................................................................................................48 Fishing and Natural Mortality Analysis.....................................................................49 Total Mortality Estimation.........................................................................................50 Comparison to Assessment.........................................................................................51 Method Comparison...................................................................................................52 Utility in Florida.........................................................................................................53 Conclusions.................................................................................................................54 APPENDIX HIGHTOWER MODEL AND CORMACKJOLLY-SEBER MODEL RESULTS........56 LIST OF REFERENCES...................................................................................................61 BIOGRAPHICAL SKETCH.............................................................................................67

PAGE 7

vii LIST OF TABLES Table page 1-1. Review of snook regulations in the state of Florida.................................................14 1-2. Alternative methods for estimating na tural mortality based on life history attributes...................................................................................................................14 2-1. Catch of study snook per gear type..........................................................................25 2-2. Descriptions of the five models used to estimate capture probabilities, F and M in all three different scenarios. Each occasion for which an estimate was generated was a month.............................................................................................25 2-3. Descriptions of the five models us ed to estimate Apparent Survival ( ) and capture probability (Error! Objects cannot be created from editing field codes.). Each occasion for which an estimate was generated was a month...................................25 3-1. Location of study site exits and cr eeks and the determined number of VR2 receivers needed.......................................................................................................44 3-2. Fates as of January 2006 of 66 tagged, harvestable size fish used to estimate mortality...................................................................................................................44 3-3. 2005 yearly estimates of fishing mo rtality, natural mortality, and their corresponding AIC values attained using the methods outlined in Hightower et. al (2001) for all three approaches of fa te assignments of questionable fish not heard after the red tide events of su mmer 2005. The number of fish used is noted for each approach...........................................................................................44 3-4. 2005 yearly total mortality estimates by modeling total mortality directly using methods from Hightower et. al (2001 ) and by simply adding together the Hightower fishing and natu ral mortality estimates..................................................44 3-5. Estimates of total mortality (Z) were used along with natural mortality (M) estimates from the Hoenig equation and estimates of emigration rate (E) to figure fishing mortality (F) from all Co rmack-Jolly-Seber models performed. AIC values and degrees of freedom (D F) are shown for each model used ( t indicates a time dependent variable, a pe riod indicates a fixe d variable, and an s indicates a variable fixed by ope n and closed harvest seasons)...............................45

PAGE 8

viii 3-6. Review of the methods and results for the mortality estimations. The ranges for the Hightower method are based on all five models used in the three different modeling approaches................................................................................................45 A-1. Results for models assuming that all fish last heard during the summer of 2005 were emigrants and fishing mortalities....................................................................57 A-2. Results for all models assuming that a ll fish last heard during the summer of 2005 were natural mortalities...................................................................................58 A-3. Results for all models assuming that th e fish last heard during the summer of 2005 were 50/50 emigrants & fishing mortalities/natural mortalities......................59 A-4. Results from all Cormack-Jolly_Seber models........................................................60

PAGE 9

ix LIST OF FIGURES Figure page 1-1. Total snook catch and release for the Gu lf of Mexico and the Atlantic from 1981-2004.................................................................................................................12 1-2. Harvest, post-release fishing deaths, and to tal snook deaths due to fishing related mortality for the Gulf of Mexico and the Atlantic from 1981-2004........................13 2-1. Map of Sarasota Bay with the study site borders of Cortez to the Northwest and Venice to the Southeast (large arrows). Passes leaving the study site are indicated by small arrows while creeks are indicated by stars.................................24 3-1. A detection history for ev ery fish in the study. Each row is an individual fish and each column is a month of the study.................................................................33 3-2. Percent detections and mean water temperatures by months of the study...............34 3-3. 2005 Red tide ( K. brevis ) cell counts for Sarasota Bay. The dashed line indicates the cell count level that begins to cau se mortality in some fish species..................34 3-4. Model p ˆt Ft Mt capture probabilities for all thr ee modeling approaches. This model had time dependent capture proba bilities, F and M estimates. These approaches differ only in the fates assi gned to 17 fish last heard during the summer of 2005........................................................................................................35 3-5. Model p ˆt Ft Mt F and M results for all three m odeling approaches. This model had time dependent capture probabilities, F and M estimates. These approaches differ only in the fates assigned to 17 fi sh last heard during the summer of 2005..36 3-6. Model p ˆt Fs Mt capture probabilities for all th ree modeling approaches. This model had time dependent capture probabi lities, and M, with F estimates fixed by open and closed harvest seasons. Thes e approaches differ only in the fates assigned to 17 fish last hear d during the summer of 2005.......................................37 3-7. Model p ˆt Fs Mt F and M for all three modeling approaches. This model had time dependent capture probabilities, and M, with F estimates fixed by open and closed harvest seasons. These approach es differ only in the fates assigned to 17 fish last heard duri ng the summer of 2005...............................................................38

PAGE 10

x 3-8. Model p ˆt Fs M. capture probabilities for all three modeling approaches. This model had time dependent capture proba bilities, fixed M estimates, with F estimates fixed by open and closed harves t seasons. These approaches differ only in the fates assigned to 17 fish la st heard during the summer of 2005............39 3-9. Model p ˆt Fs M. capture probabilities for a ll three modeling approaches. This model had time dependent capture proba bilities, fixed M estimates, with F estimates fixed by open and closed harves t seasons. These approaches differ only in the fates assigned to 17 fish la st heard during the summer of 2005............40 3-10. Comparison of fishing mortality and natural mortality estimates using the p ˆt Ft Mt model (time dependent capture probabi lities, F and M) for all three fate assignment approaches. These approaches differ only in the fates assigned to 17 fish last heard duri ng the summer of 2005...............................................................41 3-11. Catch curve analysis of 286 dead a dult snook collected during the red tide bloom of 2005..........................................................................................................42 3-12. Results for the Cormack-Jolly-Seber models tp ˆ. and tp ˆt These models each had time dependent apparent survival ( ) but differ in time dependent and fixed capture probabilities (p ˆ)................................................................................43

PAGE 11

xi Abstract of Thesis Presen ted to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Master of Science USING ACOUSTIC TELEMETRY TO ES TIMATE NATURAL AND FISHING MORTALITY OF COMMON SNOOK IN SARASOTA BAY, FLORIDA By Jason P. Bennett December 2006 Chair: William E. Pine, III Major Department: Fisher ies and Aquatic Sciences The common snook Centropomus undecimalis is a popular, saltwater gamefish found in southern Florida that has been act ively managed by Florida’s natural resource management agencies to prevent overexploita tion since the 1950’s. Despite increasingly restrictive management regulations, the status of the population and the effectiveness of these regulations remain uncertain. Most fi sheries management activities are focused on regulating fishing mortality (F). Because of this, an important aspect of population assessments is an accurate estimate of F to provide insight into the magnitude of fishing mortality relative to natural mortality (M). Here, telemetry methods for relocating radiotagged fish were used to estimate total mort ality (Z), F, and M for adult common snook in Sarasota Bay, Florida. After tagging, fish were relocated using a series of remote, autonomous receivers in conjuncti on with active tracking effort s. These relocations were evaluated using a suite of a priori assumptions to determine whether an animal was live or dead. These fates were then converte d to mortality rates using known-fate type

PAGE 12

xii models in program SURVIV. Three different modeling approaches were used to assess uncertainty in the mortality estimates to assign ing fates to 17 fish that were not relocated after a large harmful algal bloom in Saraso ta Bay during summer 2005. For the period from October 2004 through December 2005 Z values (0.68 1.08), F values (0.24 0.66), and M values (0.12 0.65) were estimated us ing different approaches which depended on how fish fates were assigned. However, estima ted parameter values were similar to more traditional stock assessment mortality estimation methods and provided insight into the performance of the mortality estimation methods currently used in the assessment. Using telemetry methods to relocate fish and es timate mortality rates has advantages over traditional assessment methods, but each me thod has assumptions and potential biases that must be acknowledged. This study se rved as a good example of estimating fishing and natural mortality for a Fl orida fishery as a complement to current stock assessment methods.

PAGE 13

1 CHAPTER 1 INTRODUCTION The common snook ( Centropomus undecimalis “snook”) is a popular gamefish species found throughout south Florida. Since the 1950’s, snook have been actively managed by the state to enhance recreational fishing and to pr otect the species from being over-fished. The management plan for this sp ecies has evolved drastically over the last 50 years and currently consists of bag, size, and seasonal limits all of which are designed to reduce fishing mortality (F) to a sustainable level (Muller and Taylor 2006). However, the effectiveness of these regulations for attaining this objective remains unclear. Stock assessments of snook and other co mmercially and recreationally important fish species in Florida are conducted at regu lar intervals to monito r population trends and fishing mortality (see for example Murphy 2003; Murphy 200 5; Mahmoudi 2005; Munyandorero et al. 2006; Muller and Taylor 2006). An important aspect of these assessments is an accurate estimation of total mortality (Z), which for management purposes, is often then broken into its compone nt parts: fishing mort ality (F) and natural mortality (M). Most assessment approaches generally use virtua l-population-analysis (VPA) type methods or tradi tional angler-dependent taggi ng studies to estimate F (e.g., Mahmoudi 2005; Munyandorero 2006; Muller and Taylor 2006). However, each of these methods relies on a diverse set of data n eeds and often complicated assumptions to indirectly estimate fish ing mortality rates. Additionall y, neither of these methods can directly estimate M (both methods often use assumed M values) as M is rarely observed or estimable from fishery-dependent data.

PAGE 14

2 Recent advances in telemetry technologies have provided fisheries researchers with new tools to investigate the ecology of aquatic animals. While using telemetered animals to evaluate the movements and habitat use of fish species is not new, until recently technological restrictions have made it difficult to use this technique to estimate vital rates in fish populations. Wild life biologists have a history of using telemetry tags to document harvest or natural mortalities of birds and mammals (White and Garrott 1990). In wildlife applications the ta gged animal can often be relocated and its fate (live or dead) visually observed yet these methods have not been widely used in fisheries applications because mortalities of aquatic animals are rarely observed. The purpose of this study was to use tele metry methods to estimate fishing and natural mortality for adult common snook in Sarasota Bay, Florida. The “fates” of individual fish were hypot hesized through a combination of active tracking and observations from an array of remote, aut onomous receivers. Each time a tagged snook was relocated, the live/dead status of th at fish was evaluated using a suite of a priori rules to help determine whether an animal was a live, emigrated from Sarasota Bay, harvested, or died from natural causes. These fates we re then converted to vital rates using knownfate type models developed by Hightower et al. (2001). These mortality estimates are critical parameters for assessing the effectiven ess of current harvest regulations used to manage common snook populations in Fl orida (Muller and Taylor 2006). Study Species The common snook is a large, predatory fish native to rivers, estuaries, and inshore tropical and subtropical waters of the wester n Atlantic Ocean, the Gulf of Mexico, and Caribbean Sea from the United States to Brazil (Gilmore et al. 1983; Rivas 1986,

PAGE 15

3 McMicheal et. al. 1989; Muller and Taylor 2006 ). Snook are cold sensitive with low lethal temperature limits from 6-13 C (43-55 F) (Marshall 1958; Howe lls et al. 1990). In Florida, snook are abundant along the Gulf of Mexico coast from Tampa Bay south to the Florida Keys and north to Cape Canaveral along the Atlantic coast (Muller and Taylor 2006). The spawning season ranges from spring to summer along both Florida coasts, generally beginning when water temperatures reach 25 C. Spawning takes place during afternoons and evenings in la rge aggregations in or near passes that provide water exchange with the open ocean (Taylor et al. 1998). Snook use a variety of habitats throughout their life history. Two to three weeks following hatching, snook larvae can be found along mangrove roots and grass edges (Peters et al. 1998). Juvenile snook tend to inhabit shallow, warm channels and creeks with few macrophytes and slow moving wate r (Fore and Schmidt 1973). Juvenile and adult snook are most commonly found in ar eas associated with mangroves (Marshall 1958; Gilmore et al. 1983) but can also be found in rivers, lake s, salt marshes, reefs, and along beaches (Muller and Taylor 2006). Duri ng cool winter months, snook seek thermal refuges, usually in creeks and canals, to avoid lethal low temperatures. Life History Snook in Florida demonstrate unique life hi story attributes by location (Atlantic Ocean or Gulf of Mexico) and also by se x. Snook from each Florida coast have been shown to be genetically isolated; snook from Florida Bay and the Flor ida Keys are closer to Atlantic phenotypes, whereas the Gu lf coast population demonstrates a unique phenotype (Tringali and Bert 1996). Snook gr owth rates differ between the coasts,

PAGE 16

4 leading to different management regulations by the Fish and Wildlife Research Institute (FWRI) for each coast (Taylor et al 1993; Tringali and Bert 1996). Snook can live to be about 20 years of age with a maximum observed age of 21 years (Taylor et al. 2000). Maximum size is about 1,100 mm TL. The Florida state record weighed 20.04 kilograms and the worl d record (from Costa Rica) weighed 24.32 kilograms (International Game Fish Association). On average, snook on the Gulf coast grow at a slower rate and r each a smaller maximum size than those on the Atlantic coast (see below, Taylor et al. 2000). Snook growth patterns are described by a von Bertalanffy growth model, with each coast ha ving its own unique growth curve. Growth equations combined for each sex are (Taylor et al. 2000): Atlantic fork length (mm) =) 1 ( 3 989) 0976 0 ( 235 0 agee (1) Gulf fork length (mm) =) 1 ( 3 947) 352 1 ( 175 0 agee (2) Snook are protandrous hermaphrodites, bei ng born and maturing as males and later changing into females. This change results in different growth rates for the sexes, with females usually growing larger than males. Males can be sexually mature by age-1, and sexual transition occurs from ages 1-7 (240 – 824mm) resulting in an uneven sex-at-age ratio. The majority of small snook (under 300-mm) are males, and most large snook (greater than 700-mm) are females (Thue et al. 1982; Taylor et al. 2000). Due to this sexually dimorphic growth, fish that are gene rally targeted and harvested are most likely females. Stock Status in Florida Commercial harvest of snook was prohi bited in 1957 when the species was declared a gamefish by the state of Florida (Muller and Taylor 2006), and this species has

PAGE 17

5 been actively managed for more than 50 years by the State (Table 1-1) In 2004 (last year data are available), snook were the fourth most recreationally sought after fish on the Gulf coast and fifth on the Atlantic coast ba sed on angler preferences from the National Marine Fisheries Service’s Ma rine Recreational Fisheries Statistics Survey interviews (Muller and Taylor 2006). Total catch (harve st and release) has increased on the Gulf coast to 2.1 (1.5-2.7) million caught in 2004, but peaked on the Atlantic coast in 1995 with an estimated 737,000 (483,000-1,002,000) snook caught (Figure 1-1). Over 90% of snook caught are released aliv e and 2.13% of snook caught and released are estimated to die post release (Taylor et al 2001). In 2004, total harvest of snook, including estimated non-harvest fishing deaths, was estima ted to be 112,000 (80,000-146,000) on the Gulf coast and 40,000 (16,000-68,000) fish on the Atlan tic coast (Figure 1-2), with 60% of Gulf harvest occurring from March to May. Recreational effort on the Atlantic coast increased until 1995 and has remained relatively constant since then. Gulf coast snook angler effort increased six-fold from 1982 to 2001 and has continued to increase since. In 2004-2005, thirty-three percent (217,000) of all Fl orida resident fishing license holders also purchased an optional endorsement “snook st amp” that is required only if the angler intends to harvest a snook (M uller and Taylor 2006). Dire cted effort for common snook is large and, at least on the Gulf coast, grow ing; as is evident from increases in angler effort (Muller and Taylor 2006). The snook fishery is one of the most h eavily managed recreational fisheries in Florida. Bag limits, size limits, allowable gears, and seasonal closures for snook were first enacted by the Florida legislature. Th e Marine Fisheries Commission was created in 1983 and addressed snook in 1985 with a ha rvest slot of 609-mm to 863-mm TL and one

PAGE 18

6 fish in possession allowed over the slot. In 1994, the winter closure was shortened by 10 days by starting on December 15 but opening the month of February, and the Florida Fish and Wildlife Conservation Commission (FWC) implemented a snook management plan to maintain snook stock size at or above 40% spawning potential ratio (SPR). In 1999, a slot length limit of 660-863mm was implemen ted and, in 2002 on the Gulf coast, the daily bag limit was reduced from two fish to one and May was added to the closed season. Although these regulations are mo re restrictive than for mo st other recreational fish species in Florida, the 40% SPR goal established in 1994 has yet to be met. The goals of these regulations are to reduce F and to prot ect the most fecund females from fishing mortality (Muller and Taylor 2006). Closed ha rvest seasons are used to eliminate harvest when snook are highly concentrated and could be easily exploited; i.e. when animals are aggregated to spawn or are in thermal refuge s. Closed seasons regulate harvest only, not effort, thus snook are targeted year-round by anglers, but can only be kept during relatively short harvest seas ons. Directed snook fishing continues during the closed seasons, and the magnitude of mortality relate d to angling during closed seasons is still being investigated. Project Justification Fishery Dependent Methods to Estimate F and M Estimates of total mortality and its compone nt parts (F and M) are key parameters used in stock assessments to manage comme rcial and recreational fish stocks. Many tools used by fishery managers (i.e.; harvest restrictions) are based on regulating fishing mortality, yet this parameter is difficult to accurately estimate. While harvest can often be observed, natural mortality is rarely obs erved, making the estimation of this parameter

PAGE 19

7 difficult. Total instantaneous mortality is commonly estimated using a catch-curve by regressing the natural log of the numbers of fi sh at age (that are fully recruited to the fishery) vs. age class. This approach es timates total mortality, but does not directly provide information on fishing or natura l mortality. A similar method is a nonregression-based total mortality estimati on using the Chapman-Robson method. This method provides an estimate of su rvival (S) using the equation: N i i N i ix xN S1 11 (3) Where xi= the number of years the i th fish is older than the age at full recruitment; and N = the total number of fully recr uited fish (Chapman and Robson 1960). These methods have a suite of assumptions that can be difficult to meet, and if violated can decrease the accuracy of Z (Ricker 1975; Van Den Avyle1993; Murphy 1997; Haddon 2001). These assumptions include: 1. Constant recruitment for all years. 2. All ages have been exposed to the same history of fishing mortality (constant fishing mortality across all represented years). 3. Random sampling of the population. Little is known about the recruitment vari ation of snook in Sarasota Bay, so any mortality values obtained from this method warrant concern. Additionally, the current slot limit functionally leads to violations of the second assumpti on because length is poorly correlated to age for snook and fish from ages 3 to 18 can be in the slot limit. However, this assumption violation is not near ly as bad as violation of the assumption of constant inter-year natural mortality and recru itment. For these reasons a catch curve was

PAGE 20

8 included in the 2005 snook stock assessment to serve simply as an approximation for comparison to other results (Muller and Taylor 2006). In most age-structured stock assessmen ts (spotted seatrout, Murphy 2003; red drum, Murphy 2005; mullet, Mahmoudi 2005; sheepshead, Munyandorero 2006; snook, Muller and Taylor 2006), F is primarily estimated by using fishery-dependent methods based on landings data and size-to-age estima tes (i.e., Virtual Popul ation Analysis [VPA] or Statistical Catch At Age [CAA] met hods), passive tagging methods (e.g., FLOY tag programs), or more recently through fisheryindependent methods. VPA methods require catch and age data and incorporate changes in vulnerability and F with age. VPA methods require assumptions of M and a termin al age, both of which can be difficult to estimate. The estimates of F for the populat ion in both VPA and CAA are most precise for year classes that have already moved through the fishery (reached terminal age). When using an untuned VPA, estimates of F fo r age classes currently in the fishery are generally not as accura te as for ages that have move d through the fishery (Walters and Martell 2004). This limits the utility of cat ch-based methods used to assess trends in fishing mortality rates over the relatively short time intervals in which regulatory decisions are often made by many management agencies, particularly for long-lived species. Another method to estimate F is to cr eate a known population of fish by collecting, marking, and releasing tagged fish. After co rrecting for tagging mo rtality, tag loss, and angler reporting rate, an estimate of F can be obtained from the ratio of the number of tags returned to the number of fish tagged. Tagging mortality and ta g loss estimates can be obtained through simple experiments, how ever, angler reporting rate can be quite

PAGE 21

9 difficult to estimate and can cause large biases in the estimate of F (Pollock et al. 2001; Pollock et. al. 2002). On small or closed systems this method is often accompanied by creel surveys, which aid in determining angler reporti ng (Van Den Avyle 1993; Ney 1993). Other approaches to estimating re porting rates include high-reward tagging programs or the use of surreptit iously planted tags (see revi ew by Pollock et al. 2001). Estimating M is difficult because natural mortality events are rarely observed and difficult to measure in aquatic systems (Quinn and Deriso 1999). If an estimate of F is available, M can be indirectly obtained fr om change in abundance at age (i.e., catch curve) by subtracting F from Z (Van Den Avyle 1993). Natural mortality is most commonly estimated by relating M to some other (often estimated) parameter related to a fish’s life history. For example, k from a Von Bertalanffy growth equation is thought to approximate M, where M = 1.5* k (Jensen 1996). However, estimators that use life history parameters can vary greatly thr oughout an animal’s life (Beverton and Holt 1959). Other methods to estimate M from life hi story information are reviewed in Table 1-2. The instantaneous natural mortality rates used by FWRI in the 2005 assessment were assumed to be the same as those us ed in previous snook assessments: M = 0.20 on the Atlantic and 0.25 on the Gulf coast. Na tural mortality estimates differ between the coasts because Gulf-coast snook are thought to be more likely to die due to cold and lethal red-tide blooms than Atlantic snook (Muller and Taylor 2006). The natural mortality rates are assumed to be consta nt across years in the FWRI assessment. Fishery Independent Methods to Estimate F and M Acoustic telemetry is a powerful tool th at can provide direct observations and estimates of F and M without many of the problems and assumptions of the methods

PAGE 22

10 mentioned above (Pine et al. 2003; Walters a nd Martell 2004). In this method, a group of animals are tagged with long-life telemetry ta gs and monitored for an extended period of time (months or years depending on battery life and tag size). The fates of relocated tagged animals can then be us ed to directly estimate F a nd M for the current study year without relying on anglers returning tags (Hightower et al. 2001). The assumptions from Hightower et al. (2001) of using a telemetry study to es timate mortality rates are: 1. All tagged snook present in the study site are alive or dead due to natural or post hooking release causes (these cause s cannot be distinguished). 2. Each snook alive in the study site has the sa me probability of surviving to the next tracking event. 3. All tags have been retained and are working properly. 4. Each snook is behaving independently with regard to emigration and harvest. 5. Snook not present in the study site have been harvested or have emigrated from the study site. 6. Snook repeatedly located in the same loca tion died due to non harvest mortality (i.e., hooking or natural mortality). Because of these assumptions, this method works best for spatially closed systems where the probability of tracki ng the fate of an animal is high. Open systems must be made “closed” by monitoring each exit point, which facilitates the monitoring of fish emigration from the study location. Parameter Estimation Estimates of F and M can be obtained by using the methods outlined by Hightower et al. (2001). This method is similar to a known-fate Kaplan-Meier model modified by Pollock et al. (1995) that allows for relocations of both live and dead animals. Most known-fate models assume a relocation proba bility of 1.0 (Williams et al. 2001). The combined model from Pollock et al. (1995) al lows for relocations of both live and dead

PAGE 23

11 animals and also allows for relocation probabi lities less than 1.0. The Hightower et al. (2001) model allows for distinct estimates of F and M by modeling only relocated fish for each individual occasion (i.e., month), and not for all fish potentially in the system. This is necessary because not all snook present in the study site will be located on each tracking sweep or by stationary receivers. Because only relocated fish are modeled, each relocation acts as a “release” of that animal back into the system. For example, a fish located in June on multiple receivers is known to be alive and within the system until it is relocated again at a future date either in the study array or until the animal is known to have emigrated. If the fish is not located again after June it is assumed to have been harvested if it was not detected leaving the system.

PAGE 24

12 Gulf of Mexico 0 500,000 1,000,000 1,500,000 2,000,000 2,500,0001981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003Number of Snook Total Snook Catch ReleasedA Atlantic 0 200,000 400,000 600,000 800,000 1,000,000 1,200,000 1,400,000 1,600,0001981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003Number of Snook Total Snook Catch ReleasedB Figure 1-1. Total snook catch and release for th e Gulf of Mexico a nd the Atlantic from 1981-2004 (Muller and Taylor 2006).

PAGE 25

13 Gulf of Mexico 0 20,000 40,000 60,000 80,000 100,000 120,000 140,0001981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003Number of Snook Snook Harvested Post-release Deaths Total Fishing DeathsA Atlantic 0 20,000 40,000 60,000 80,000 100,000 120,0001981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003Number of Snook Snook Harvested Post-release Deaths Total Fishing DeathsB Figure 1-2. Harvest, post-release fishing deaths, a nd total snook deaths due to fishing related mortality for the Gulf of Mexico and the Atlantic from 1981-2004 (Muller and Taylor 2006).

PAGE 26

14 Table 1-1. Review of snook regulatio ns in the state of Florida. Year Regulation 1953 Minimum size set at 457mm (18 inches) FL 1957 Snook made illegal to buy or sell; Bag limit set at four snook > 457mm (18 inches) FL 1981 Bag limit reduced to two snook per day. No snook < 660mm (26 inches) FL maybe taken in June or July during 1982 – 1986 1983 January, February, June, and July 1983 – 1986 closed to snook possession 1985 January, Feb, June, and July closed permanently August 1985-1986 closed Minimum size increased to 609mm (24 inches) TL Only one snook may be > 863mm (34 inches) TL 1987 All species of Centropomus covered by the regulations August is closed permanently Use of treble hooks prohib ited with natural baits 1994 Winter closed during 16 December – January 31 SPR goal set at 40% 1997 Population separated into Atlantic and Gulf Stocks 1999 Harvest slot set at 660mm ( 26 inches) to 863mm (34 inches) TL 2002 Gulf stock: closed during May and daily bag reduced to one snook from two (Muller and Taylor 2002) Table 1-2. Alternative methods for estimati ng natural mortality based on life history attributes. Equations for estimating M Citation )) ( (log 4643 0 )) ( (log 643 0 )) ( (log 279 0 0066 0 ) ( log10 10 10 10e temperatur water annual average k L M Pauly (1980) ) ln( 982 0 44 1 ) ln(max ^t M :where tmax equals the maximum observed age in the un-fished stock Hoenig (1983) 25 092 1aW M :where Wa equals the weight at age a Peterson and Wroblewski (1984)

PAGE 27

15 CHAPTER 2 METHODS Field Methods Study Site This study occurred in Sarasota Bay, Flor ida. Sarasota Bay is oriented along a north-south axis and is traver sed along this axis by the Intercoastal Waterway (ICW), which enters to the north at the Cortez Bridge near Cortez, Florida and exits to the south at the Albee Point bridge near the Venice inle t in Venice, Florida. The bay covers an area of 13,467.9 hectares and a linear dist ance of 43.45 kilometers, and includes multiple sub-bays each with unique flow and habi tat characteristics: H udson Bayou, Blackburn Bay, and Little Sarasota Bay. The site also contains three passes between the bay and the Gulf of Mexico (Longboat Pass, New Pass, and Big Pass) as well as five major tributary creeks (Bowlees Creek, Whitaker Creek, Phil lipi Creek, North Creek and South Creek) (Figure 2-1). Tagging Procedure Beginning in October of 2004, adult common snook were collected in Sarasota Bay using hook and line, seines, and trammel nets Trammel netting by actively searching for legal snook and then setting th e net on these fish was the most successful method (Table 2-1). All snook collected were measured ( TL), and each fish selected for surgery was anesthetized with sodium bicarbonate, and a Vemco acoustic telemetry tag implanted into its abdominal cavity. Two different t ypes of tags were used; 65 V-16 tags (16-mm diameter, 70-mm length, 700-da y battery life) and 10 V-8 tags (8-mm diameter, 40-mm

PAGE 28

16 length, 300-day battery life; Vemco Ltd., Shad Bay, Nova Scotia, Canada). Sixty-three of the 75 tags were implanted in snook wi thin the legal size range; five tags were implanted in snook over legal si ze and seven in snook just unde r legal size. Three angler returned tags were later reus ed and implanted in newly ca ught snook, bringing the sample size of legally harvestable tagged fish to 66. Abdominal incisions were closed with 3-4 absorbable sutures and ethyl cyanoacrylate. Each snook was observed until fully recovered and then released into the same area from which it was collected. No ex ternal tag or mark was added or visible (except for the surgery scar and temporary sutu res) to avoid influenc ing angler behavior. The goal of this project was to generate fishery independent estimates of F and M, therefore any tag or mark that could change the probability of harvest was unwanted. Tag and Receiver Description Each acoustic tag outputs a uniquely iden tifiable signal, pseudo-randomly between 30 to 90 seconds (mean = 60 seconds). The Vemco VR2 acoustic receivers (Vemco Ltd.) recorded the unique tag number along with the date and time of reception. Data from all receivers were manually retrieved on one to three month intervals during which time the receivers and anchor materials we re checked and cleaned to assure proper performance. Receiver battery changes occurr ed on yearly intervals to prevent loss of acoustic coverage and data. Acoustic Array Network Array deployment and testing The use of remote receivers requires that snook move within the relatively small detection range of the receiver in order to be detected. To increase th e probability of tags being detected, receivers were positioned in ar eas likely to be frequented by snook, or in

PAGE 29

17 areas that offered physical boundaries, such as a narrow area of the bay (Klimley et al. 1998; Lacroix and Voegeli 2000; Clements et al 2005, Heupel et al. 2006). Each of the five exits from the study site were acousti cally “closed” with 1-3 fully automated stationary VR2 receivers depending on the width, depth, and ambient noise of each site. Each of the three bay passes and the two IC W exits from the study site (Cortez in the north and Venice in the south, Figure 2-1) we re range tested with temporarily placed VR2 receivers and active tags to ensure that number and placement of receivers provided complete acoustical spatial coverage of the pa ss or exit. Because these sections of the bay differ in shape, depth, and area; each pass/exit required different numbers of receivers and receiver placements. Receivers were placed in potential locations while tags were drifted through the pass at severa l locations in an attempt to mimic possible routes a snook could take through the pass. Data from each VR2 receiver was then downloaded and compared to the actual times the tags were drifte d through the pass to evaluate receiver performance. This dia gnostic procedure was repeated until receivers were placed such that a test tag drifting through the pass was detected at least twice during the time the tag was passing through th e detection field for complete acoustic coverage of each pass or exit (Clements et al 2005). Each receiver was then anchored on the bottom with either a large concrete anch or or attached to an existing piling or navigational structure. Receivers were deployed in the creeks to monitor tagged snook movements in and out of the creek systems. Because the creek depths were strongly dependent on tides, the receivers were generally placed in deeper holes near the creek mouths such that the receivers would remain submerged even at lo w tide. Additional receivers were located

PAGE 30

18 upstream throughout the individual creeks as pa rt of a separate te lemetry study evaluating juvenile snook movement patterns. Manual tracking Manual tracking using a Vemco VR60 or VR100 acoustic receiver by boat was used to search for fish in areas not cove red by the VR2 receivers. Manual tracking was key for trying to locate specific fish that may not have been recently detected on the receiver array (by searching n ear their last known location) as well as locating tags from animals that died. Manual tracking also aide d in identifying whether a fish had emigrated from the study site by allowing for checking along the beaches outside of Sarasota Bay and within the passes in areas that exte nded beyond the detection area of the VR2 receivers. Manual tracking was conducted appr oximately monthly during the summer and every 6-8 weeks at other times of the year. From field testing, it was determined that a conservative detection distance using the ma nual receiver was approximately 300 m. The manual tracking protocol then consisted of tracking the shoreline of Sarasota Bay and stopping every 300 m for 3 min to listen for ta gged snook. Cross-bay transects were also conducted when possible. Data Analysis Assigning Fish Fates For the purposes of this thesis only the legally harvestable fish were used to estimate F and M (Hightower et. al 2001). All fish used in the analys is were detected at least once after tagging before being included in the study to reduce the chance of tagging and surgery effects influenc ing the study results. Relocations from the VR2 and VR100 receiv ers were used to assign fates to each fish in the study based on the previously me ntioned assumptions fr om Hightower et. al

PAGE 31

19 (2001). Fish fates were defined as: within site and heard aliv e, within site but not heard (i.e. up creek beyond detection range of receivers ), within site and heard dead (i.e. tag lying on bottom), or emigrated from site. These fate assignments were based on when and where each fish was last located via ac tive tracking and/or on the array. Fate assignments were made by two people reviewin g the relocations and jointly determining the fate assigned to that animal for each month. The individual fates of the fish were then used to construct a monthly binary “capture” hi story of each fish; with months of live, in site relocation receiving a “1” and months with other fates assigned a “0”. Fish determined to be emigrants were censored from the models. These fish could be added back into the models later if the fish r eentered the site. Program MARK (White and Burnham 1999) was used to create the data summary matrix (“Yij” matrix, Burnham et al. 1987) to use in the data analysis. Fishing and natural mortality rates were estimated using program SURVIV (White 1983) based on general models and procedures described by Hightow er et al (2001). With this model, for R snook released at time t the number of fish still alive at time t + 1 ( St +1) is expressed as: 1 ) ( 1 t M F t tp e R St t (4) Here Ft and Mt are instantaneous mortality rates during period t, and pt+1 is the relocation probability at time t+1. Survivors to time t+1 were considered new releases for the next period. The number of fi sh still alive at time t+2 (St+2), but not seen at time t+1, is expressed as: ) ( ) 1 (2 ) ( 1 ) ( 21 1 t M F t M F t tp e p e R St t t t (5)

PAGE 32

20 Capture probabilities (p ˆ) and values of F and M are made on the time scale of the data input; in this case, mont hly estimation. The precision of the parameter estimates are driven by the magnitude of the event (data “c ontrast”, Hilborn and Walters 1992). Sparse data, such as infrequently obs erved mortality events, result in limited contrast; thus, some models were constructed at l onger time scales (“fishing seas on” or annually) to allow for monthly observations to be pooled (Hightower et al. 2001, Heupel and Simpfendorfer 2002). A suite of biologically reasonable models were used to estimatep ˆ, F, and M. The models used different time intervals for each parameterfixed models (denoted by a “ X.” subscript), time (monthly) dependent models (“ Xt”), and seasonal models ( “Xs” ) that corresponded to the open and cl osed harvest season for snook ( X equaled p ˆ for capture probability, F for fishing mortality, or M fo r natural mortality) (see Table 2-2 for all model combinations). Akaike Information Criteria (AIC) values (Akaiki 1974; Anderson et. al 1998) were used to pr ovide guidance on model fit. Since fate assignments were based on a ssumptions and subjective inference, it was possible that fates could be mis-assigned (i.e a natural mortality mis-classified as an emigrant or vice versa). The data was firs t analyzed with the fate assignments based strictly on the rules described in Hightower et al (2001). Tw o alternative da ta sets were then developed which reassigned fates for a gro up of 17 fish that were last detected alive during the summer of 2005 and had subjectiv e fate assignments. These 17 fish were originally designated as either fishing mort alities or emigrants depending on there last known location in the receiver ar ray (8 last heard within th e array and 9 last heard on a pass/exit). A fish that was last detected at a pass/exit receiver was classified as an

PAGE 33

21 emigrant, while a fish last detected on one of the bay receivers was classified as a fishing mortality. The classification of these fish was different than others because, of the 18 fish no longer heard beyond the summer of 2005 seventeen tags were not located and, thus, could not confirm that thes e fish were indeed dead. Thes e fish were suspected to be natural mortalities because they were lost fr om the receiver array during a relatively short time interval, the harvest season was closed, a nd there was an extremely large and intense red tide event occurring at this time. Rec overy of these tags was not possible as local governmental agencies hired aquatic plant removal companies to use floating plant harvesters and prison work crews to remove dead fish floating on the surface and transport the fish to the landfill. To eval uate how possibly mis-assigning these fish would change F and M estimates, these 17 fish were assigned alternative fates by classifying them as all emigrants and fishing mo rtalities (notated as lo st fish as emigrants and fishing mortalities), all as natural mort alities (notated as lo st fish as natural mortalities) or an equal combination of the two (notated as 50/50). In the “lost fish as emigrants and fishing mortalities approach” fish that were last detected on a pass/exit receiver were considered emigrants and were censored from the analysis; fish last heard on receivers inside the site were considered fi shing mortalities (46 fish were used in this approach). Conversely, in the “lost fish as natural mortality appr oach” the 17 lost fish were considered natural mortalities no matter what the location of th eir last detection (55 fish were used in this approach). A “50/50 approach” was used to examine the sensitivity of the models to fate assignments by showi ng how the models reacted to more subtle changes in fate assignments compared to the more extreme differences between the lost fish as emigrants and fishing mortalities approa ch and the lost fish as natural mortalities

PAGE 34

22 approach (50 fish were used in this approach). The differences in the number of fish used for each approach were due to the assignmen t of emigration as emigrants were censored from the analyses. Total Mortality Estimates The Hightower et al. (2001) approach was also used to estimate Z for the tagged fish. In this method, no dis tinction was made between fish no longer heard within the site and known natural mortalities. In this framework, the mo rtality components (F and M) were pooled to generate a total mortality es timate. This approach eliminated some subjectivity in assigning fates of fish and provi ded an estimate of total mortality that was useful for comparing mortality trends across the study period. Annual apparent survival, survival based on loss of detection, (apparent survival, = 1 – mortality – emigration) was also es timated using Cormack-Jolly-Seber (CJS) methods (Cormack 1964; Jolly 1965; Seber 196 5, Pollock et al. 1990) to compare with the total mortality trends estimated with the other models. This method used only capture histories for each fish and was not dependent on fates assigned to the tagged fish. The model estimated two parameters, and p ˆ, which could be either fixed, time dependent, or fishing season dependent. Table 2-3 c ontains a description of all models. The assumptions of the CJS method are: 7. All previously tagged snook alive at the time of one census are equally likely to be captured in that census. 8. Tagging does not increase the likelihood of mortality. 9. The probability of a snook surviving to the next census is independent of its age. Ages of the fish were unknown, but only fi sh within the legal harvest slot were used in this analysis. Survival estimate s from CJS models are robust to permanent

PAGE 35

23 emigration, but can be biased by temporary emigration when capture probability is low (Kendall and Nichols 1995, Zehf uss et. al 1999). The apparent survival estimates were adjusted to correct for five known permanent emigrants, but it was not necessary to correct for temporary emigrants as the time st ep of the emigration was often less than the time step of the estimate as small, sometime s days long, emigrations occurred with some fish. Trends in apparent survival from th is model should have mirrored mortality events from the Hightower models (Pine et. al 2003). Total mortality and emigration rates (E) were estimated for the CJS models with the following equations: ) ( log S Ze (6) ) 1 ( log proportion emigration Ee (7) Emigration proportion was the proportion of pe rmanent emigrants to total study fish. M was estimated using the Hoenig equation outlin ed in Table 1-2. Using results from equations 5 and 6, F was estimated where: E M Z F (8) These estimates of F were then compared w ith F estimates from the Hightower models. A potentially large source of natural mortality in marine fish populations can be from harmful algal blooms (“red tides”) commonly caused by large blooms of the phytoplankton Karenia brevis. These dinoflagellates release a nonproteinaceous endotoxin called brevitoxin (Steidinger and Ha ddad 1981). Concentrations of red tide cells greater than 100,000-200, 000 cells/liter can cause fish mortality (Sea Stats 2000). In the summer of 2005, a spatially and tempor ally large red tide bloom occurred off of southwestern Florida with highest concentratio ns occurring in Saraso ta Bay. Concurrent

PAGE 36

24 with this red tide, widespread mortalities of a variety of fish and marine mammal species were recorded in southwest Florida. Duri ng July, large numbers of snook were observed dead or dying around Sarasota Bay, particular ly in the New Pass area and bay and ocean shorelines near Sarasota. Two hundred eight y-six adult snook of various sizes were collected during this morta lity event for size and age structure sampling. All snook collected were measured, their otoliths remove d, and the fish were checked for acoustic tags. The otoliths were sent to FWRI for age analysis and the resulting age estimates were used in the catch-curve analysis. Th is estimate of Z was independent of the mortality estimates from the tagging data. Figure 2-1. Map of Sarasota Bay with the st udy site borders of Cortez to the Northwest and Venice to the Southeast (large arro ws). Passes leaving the study site are indicated by small arrows while creeks are indicated by stars.

PAGE 37

25 Table 2-1. Catch of study snook per gear type. Gear type Number of study snook caught Hook and Line 6 Seine net 10 Trammel net 62 Table 2-2. Descriptions of the five models used to estimate capture probabilities, F and M in all three different scenarios. Each occasion for which an estimate was generated was a month. Model Number of Parameters Description p ˆt Ft Mt 45 Time dependent capture probabilities, F and M p ˆ. F. M. 3 Fixed capture probabilities, F and M p ˆ. Fs M. 4 Fixed capture probabilities and M, and F fixed by open and closed harvest seasons p ˆt Fs M. 18 Time dependent capture probabi lities, fixed M, and F fixed by open and closed harvest seasons p ˆt Fs Mt 32 Time dependent capture proba bilities and M, and F fixed by open and closed harvest seasons Table 2-3. Descriptions of the five mode ls used to estimate Apparent Survival ( ) and capture probability (p ˆ). Each occasion for which an estimate was generated was a month. Model Number of Parameters Description tp ˆ. 16 Time dependent apparent survival and fixed capture probabilities tp ˆt 30 Time dependent apparent surv ival and capture probabilities .p ˆt 16 Fixed apparent survival and tim e dependent capture probabilities sp ˆt 17 Apparent survival fixed by open and closed harvest seasons and time dependent capture probabilities p ˆ. 2 Fixed apparent survival and capture probabilities

PAGE 38

26 CHAPTER 3 RESULTS Field Results Tag Reception Table 3-1 lists the number of VR2 receive rs used in all passes and creeks. Receivers were placed in the passes/exits su ch that there was no possible route a snook could take through the pass/exit without being detected. During range tests of receiver placement performance, tags were heard at least twice on every test drift through the pass/exit, implying total acousti c coverage. Reception rate was calculated as the actual number of detections divided by the number of times, on average, a tag could have been detected while the tag was in the receivers estimated detection range. On average total tag detections were about 50%, such that if, in theory, a tag should have been detected 10 times during a drift through the pass, th e tag was detected at least 5 times. Fish Detection A detailed detection history for all fish us ed in the analysis is provided in Figure 31. Four fish were never detected post rel ease and were excluded from the analysis. These fish may have succumbed to tagging mortality, immediate harvest, or tag failure. Because the fates of these four fish could not be determined, they were censured from the analysis. Tag failure was unlikely as it wa s evaluated by examining 10 recovered tags (nine from anglers and one found in a dead s nook); all were worki ng when returned and later implanted again into snook. The VR2 receivers could occasionally interpret an ambient noise (i.e., from a depth sounder) as a single pulse from an acoustic tag.

PAGE 39

27 Following the guidelines in Clements et. al (2005), multiple receptions (more than one reception pulse over several minutes) were re quired before a receiver reception was considered a true detection. Nearly all other study fish (72 of 78) were heard on at least one occasion on a stationary receiver. Detection rates approached 100% (p ˆ= 1.0) early in the study as water temperatures declined towards leth al limit for snook (Figure 3-2). As water temperatures declined, snook entered their thermal refuges within th e tidal creeks, passing the creek receivers on their way to these refuges, thus increasing detection rates. When the water temperatures were at their lowest levels, snook movement rates greatly declined and snook likely remained stationary within their thermal ref uges, thereby causing det ection rates to drop. As water temperatures increased in the spring, detection rates also increased as the snook left the creeks (detected again on the creek r eceivers) and entered the bay. During this time, receptions on the intra-bay and pass receivers increased as the snook resumed their spring and summer movement patterns. A sharp decline in detections was observed during the summer of 2005 (Figure 32), during which a major red tide bloom was occurring (Figure 3-3). Eighteen study fish were not detected beyond this point of time. One of these fish was an observed natural mortality (dead fish was recovered), but the fates of the other 17 were unknown. The last known locations for 8 of these fish were from bay and exit receivers, but 9 were last heard in locations inside the bay. Duri ng the fall and winter of 2005/2006, as water temperatures again declined, detection ra tes continued to decrease and never again reached the levels from the pr evious year (Figure 3-2).

PAGE 40

28 Two fish were detected only through ma nual tracking. Manual tracking also located five tags that were classified as na tural mortalities based on repeated relocations in a single location. Recovery of these tags was attempted, but was unsuccessful. However, it was believed that if these fish ha d been alive and simply stationary, attempts to recover the tags would likely have cause d the fish to swim away from the area. Data Analysis Results Assigning Fates Fates of all 66 legally harvestable fish used in this analysis are presented in Table 3-2. Fish whose tags were returned were know n to be fishing mortal ities (they were, by default, counted toward F by the model due to their disappearance from within the system). Fates of fish that died of natura l mortality were more difficult to classify. During the red tide bloom only one tag was reco vered from a dead snook that was part of the age sample collection. More fish were assumed to have died, but it was thought that their carcasses were removed by local cleanup crews before they could be checked for tags. Concurrent with this large fish kill was an extreme drop in tag detections (July 2005, Figure 3-2); several fish that had commonl y been detected were not detected again on the array or via manual tracking. Because of the events which occurred at this time it was necessary to consider these fish using alternative fate assignment approaches. These fates were used to examine how the model resu lts would differ if these fish were treated as if they had been illegally harvested, died of natural mo rtality (even though tags were never relocated), or emigrated. The result s from modeling approaches with the “lost” fish with different fates assigned during the red tide bloom are presented below. For each of the three approaches (the lo st 17 fish as both emigrants and fishing mortalities approach, lost fish as natural mortalities approach, and lost fish as 50/50

PAGE 41

29 approach) the three best performing models, based on AIC values, were always p ˆt Ft Mt, p ˆ t Fs Mt, and p ˆt Fs M.; AIC and yearly F and M values (summed monthly values) of all models for all three approaches are provided in Table 3-3. Results from all models performed in all three appr oaches are provided in Tabl es A-1, A-2, and A-3. Model p ˆt Ft Mt This model was fully time dependant, with variable monthly capture probability, F and M estimates. Capture probabilities were high, ranging from 0.7 to 1.0 for all three approaches (Figure 3-4). As was expected, F and M estimates varied between the three approaches. In the lost fish as emigrants and fishing mortalities approach, F peaked in August 2005 (0.18 per month, SE = 0.09) a nd estimated a yearly F (monthly values summed) of 0.66 per year. The natural mortality estimate was 0.16 per year. In the lost fish as natural mortalities, F was lower at 0.41 per year. However, natural mortality was much higher, peaking in August 2005 (0.35 per month, SE = 0.11) leading to a yearly M of 0.65 per year. The lost fish as 50/50 a pproach had estimates in-between the other approaches with F equal to 0.48 per year a nd M equal to 0.42 per year (Figure 3-5). Model p ˆt Fs Mt This model had time dependant monthly capture probability and M estimates, but the F estimates were fixed by open and closed harvest seasons. Capture probabilities were again high, ranging from 0.7 to 1.0 for all three approaches (Figure 3-6). Estimates of F and M again varied with the major diffe rence in F coming from the values for the closed harvest seasons. In the lost fish as emigrants and fishing mo rtalities approach the F for the open harvest season months equa led 0.05 per month (SE = 0.02), while the estimates for the months of the closed harv est seasons dropped to 0.03 per month (SE =

PAGE 42

30 0.01); leading to a yearly F value of 0.50. In the lost fish as natural mortalities approach, F for the open harvest season months ag ain equaled 0.05 per month (SE = 0.02), however, the estimates for the months of the closed harvest seasons dropped to 0.00 per month (SE = 0.00); leading to a yearly F value equal to 0.30. The F for the lost fish as 50/50 approach was again somewhere between th e other approaches w ith a yearly F of 0.34 per year. Monthly natural mortality estimates were almost identical to the previously describedp ˆt Ft Mt model for all approaches (Figure 3-7). Model p ˆt Fs M. This model had time dependant monthly cap ture probability, but a fixed monthly M and F fixed by open and closed harvest seasons Capture probabilities were again high, ranging from 0.7 to 1.0 for all three approaches (Figure 3-8). Estimates of F and M again varied, but monthly F estimates for all ap proaches were almost identical to the p ˆt Fs Mt previously described. The major difference in results for the different approaches for this model came from the M estimates. The lost fish as emigrants a nd fishing mortalities approach had monthly M values equal to 0.01 per month (SE = 0.01) for all months in the study. This gave a yearly M equal to 0.12 per year. Conversely, the lost fish as natural mortalities approach had monthly M values equal to 0.05 per month (SE = 0.01) for all months in the study leading to a yearly M of 0.60 per year. The lost fish as 5/50 approach estimate was again centered within the approach estimates with a yearly M equal to 0.36 (Figure 3-9). Clearly there were differences between a ll the models for the three approaches used. Figure 3-10 shows the differences betw een the monthly F and M estimates for all three approaches for the p ˆt Ft Mt (fully time dependent) model. The estimates for F

PAGE 43

31 varied between the approaches, but the ma jor source of variati on between the three approaches appears to be in the M estimation. Secondary Analysis Total mortality was modeled in two ways. The first method modeled Z in the same way that F and M were modeled using methods from Hightower et. al (2001); with each of the five models and the three different a pproaches of assigned fates. This method resulted in a range of yearly Z estimates for 2005 of 0.68 – 1.08 per year. The second method simply added together F and M estimates from the previously described Hightower models as per the e quation Z = F + M and resulted in a range of yearly 2005 Z estimates of 0.63 – 1.06 per year (Table 3-4). A catch curve was included in the 20 05 Snook Stock Assessment to provide a “rough estimate of the magnitude of total mortality” (Muller and Taylor 2006). The 286 dead adult snook collected during the red tide event were used to construct a comparable catch-curve using a similar range of fish ages The catch-curve was estimated using fish ages 5 through 8 and 6 through 8 because of the 286 fish only four were age 9 or greater. This generated a Z estimate of 0.37 per ye ar when using ages 5 through 8 and 0.83 per year when using ages 6 through 8 (Figure 3-11) Age-5 was used as it was included in the catch curve for the 2005 snook stock assessment (Muller and Taylor 2006). However, it is unlikely that age-5 fish are fully vulnerable. For the Cormack-Jolly-Seber analysis, the tp ˆ. and tp ˆt models were virtually identical in model fit AIC values of 698.58 a nd 699.22. AIC scores from all models fit are provided in Table 3-5 and results in Table A-4. Capture probabilities were as high as those from the Hightower models, ranging from 0.7 to 1.0. Model tp ˆ. showed a strong

PAGE 44

32 drop in survival in April (0.83, SE = 0.06) and August (0.69, SE = 0.08) of 2005. Those are the same months that s how high natural mortality in the Hightower models. Model tp ˆt showed a dip in capture probability duri ng the cold months following the beginning of the study similar to that shown by previous ly mentioned models. It also showed a decline in survival during Ap ril (0.85, SE = 0.05) and August (0.70, SE = 0.08) of 2005 (Figure 3-12). Survival estimates from the CJS models were used to estimate total mortality. These rates were used along with estimate s of natural mortality and emigration to estimate fishing mortality (Eq. 8). Z estimates ranged from 0.43 to 0.56 per year giving a range of F estimates from 0.18 to 0.31 per year (Table 3-5). These several methods gave variable es timates of mortality. Components of mortality (Z, F, and M) were estimated using two different methods each. A review of each component, method used, and the result are presented in Table 3-6.

PAGE 45

33 Not heard Heard alive Known Fishing mortality Natural mortality Month before deployment Possible tagging mortality Oct 2004 Nov 2004 Dec 2004 Jan 2005 Feb 2005 Mar 2005 Apr 2005 May 2005 Jun 2005 Jul 2005 Aug 2005 Sep 2005 Oct 2005 Nov 2005 Dec 2005 Jan 2006 Figure 3-1. A detection hi story for every fish in the study. Each row is an individual fish and each column is a month of the study.

PAGE 46

34 0 10 20 30 40 50 60 70 80 90 100Oct-04 Nov-04 Dec-04 Jan-05 Feb-05 Mar-05 Apr-05 May-05 Jun-05 Jul-05 Aug-05 Sep-05 Oct-05 Nov-05 Dec-05 Jan-06Percent0 5 10 15 20 25 30 35Mean Water Temp (C) % Detections per month Water Temp (C) Lethal Temp Limit Figure 3-2. Percent detections and mean water temperatur es by months of the study. 0 2, 0 00 000 4, 0 00 000 6 ,0 00 0 0 0Jan-05 Feb-05 Mar-05 Apr-05 May-05 Jun-05 Jul-05 Aug-05 Sep-05 Oct-05 Nov-05 Dec-05Cell coun t Red tide cells/Liter Observed Fish Mortality Figure 3-3. 2005 Red tide (K. brevis) cell counts for Sarasota Bay. The dashed line indicates the cell count leve l that begins to cause mortality in some fish species.

PAGE 47

35 Emigrants and Fishing Mortalities Approach0.50 0.60 0.70 0.80 0.90 1.00Nov-04 Dec-04 Jan-05 Feb-05 Mar-05 Apr-05 May-05 Jun-05 Jul-05 Aug-05 Sep-05 Oct-05 Nov-05 Dec-05A Natural Mortalities Approach0.50 0.60 0.70 0.80 0.90 1.00Nov-04 Dec-04 Jan-05 Feb-05 Mar-05 Apr-05 May-05 Jun-05 Jul-05 Aug-05 Sep-05 Oct-05 Nov-05 Dec-05B 50/50 Approach 0.50 0.60 0.70 0.80 0.90 1.00Nov-04 Dec-04 Jan-05 Feb-05 Mar-05 Apr-05 May-05 Jun-05 Jul-05 Aug-05 Sep-05 Oct-05 Nov-05 Dec-05C Figure 3-4. Model p ˆt Ft Mt capture probabilities for all thr ee modeling approaches. This model had time dependent capture proba bilities, F and M estimates. These approaches differ only in the fates assi gned to 17 fish last heard during the summer of 2005.

PAGE 48

36 Emigrant and Fishing Mortalities Approach0.00 0.10 0.20 0.30 0.40 0.50Nov-04 Dec-04 Jan-05 Feb-05 Mar-05 Apr-05 May-05 Jun-05 Jul-05 Aug-05 Sep-05 Oct-05 Nov-05 Dec-05Rate F MA Natural Mortalities Approach0.00 0.10 0.20 0.30 0.40 0.50Nov-04 Dec-04 Jan-05 Feb-05 Mar-05 Apr-05 May-05 Jun-05 Jul-05 Aug-05 Sep-05 Oct-05 Nov-05 Dec-05Rate F MB 50/50 Approach0.00 0.10 0.20 0.30 0.40 0.50Nov-04 Dec-04 Jan-05 Feb-05 Mar-05 Apr-05 May-05 Jun-05 Jul-05 Aug-05 Sep-05 Oct-05 Nov-05 Dec-05Rate F MC Figure 3-5. Model p ˆt Ft Mt F and M results for all three modeling approaches. This model had time dependent capture proba bilities, F and M estimates. These approaches differ only in the fates assi gned to 17 fish last heard during the summer of 2005.

PAGE 49

37 Emigrants and Fishing Mortalities Approach 0.60 0.70 0.80 0.90 1.00Nov-04 Dec-04 Jan-05 Feb-05 Mar-05 Apr-05 May-05 Jun-05 Jul-05 Aug-05 Sep-05 Oct-05 Nov-05 Dec-05A Natural Mortalities Approach 0.60 0.70 0.80 0.90 1.00Nov-04 Dec-04 Jan-05 Feb-05 Mar-05 Apr-05 May-05 Jun-05 Jul-05 Aug-05 Sep-05 Oct-05 Nov-05 Dec-05B 50/50 Approach 0.60 0.70 0.80 0.90 1.00Nov-04 Dec-04 Jan-05 Feb-05 Mar-05 Apr-05 May-05 Jun-05 Jul-05 Aug-05 Sep-05 Oct-05 Nov-05 Dec-05C Figure 3-6. Model p ˆt Fs Mt capture probabilities for all th ree modeling approaches. This model had time dependent capture probabi lities, and M, with F estimates fixed by open and closed harvest seasons. Thes e approaches differ only in the fates assigned to 17 fish last hear d during the summer of 2005.

PAGE 50

38 Emigrants and Fishing Mortalities Approach0.00 0.10 0.20 0.30 0.40 0.50Nov-04 Dec-04 Jan-05 Feb-05 Mar-05 Apr-05 May-05 Jun-05 Jul-05 Aug-05 Sep-05 Oct-05 Nov-05 Dec-05Rate F MA Natural Mortalities Approach 0.00 0.10 0.20 0.30 0.40 0.50Nov-04 Dec-04 Jan-05 Feb-05 Mar-05 Apr-05 May-05 Jun-05 Jul-05 Aug-05 Sep-05 Oct-05 Nov-05 Dec-05Rate F MB 50/50 Approach 0.00 0.10 0.20 0.30 0.40 0.50Nov-04 Dec-04 Jan-05 Feb-05 Mar-05 Apr-05 May-05 Jun-05 Jul-05 Aug-05 Sep-05 Oct-05 Nov-05 Dec-05Rate F MC Figure 3-7. Model p ˆt Fs Mt F and M for all three modeling approaches. This model had time dependent capture probabilities, a nd M, with F estimates fixed by open and closed harvest seasons. These approaches differ only in the fates assigned to 17 fish last heard during the summer of 2005.

PAGE 51

39 Emigrants and Fishing Mortalities Approach 0.60 0.70 0.80 0.90 1.00Nov-04 Dec-04 Jan-05 Feb-05 Mar-05 Apr-05 May-05 Jun-05 Jul-05 Aug-05 Sep-05 Oct-05 Nov-05 Dec-05A Natural Mortalities Approach 0.60 0.70 0.80 0.90 1.00Nov-04 Dec-04 Jan-05 Feb-05 Mar-05 Apr-05 May-05 Jun-05 Jul-05 Aug-05 Sep-05 Oct-05 Nov-05 Dec-05B 50/50 Approach 0.60 0.70 0.80 0.90 1.00Nov-04 Dec-04 Jan-05 Feb-05 Mar-05 Apr-05 May-05 Jun-05 Jul-05 Aug-05 Sep-05 Oct-05 Nov-05 Dec-05C Figure 3-8. Model p ˆt Fs M. capture probabilities for all three modeling approaches. This model had time dependent capture proba bilities, fixed M estimates, with F estimates fixed by open and closed harvest seasons. These approaches differ only in the fates assigned to 17 fish last heard during the summer of 2005.

PAGE 52

40 Emigrants and Fishing Mortalities Approach0.00 0.10 0.20 0.30 0.40 0.50Nov-04 Dec-04 Jan-05 Feb-05 Mar-05 Apr-05 May-05 Jun-05 Jul-05 Aug-05 Sep-05 Oct-05 Nov-05 Dec-05Rate F MA Natural Mortalities Approach0.00 0.10 0.20 0.30 0.40 0.50Nov-04 Dec-04 Jan-05 Feb-05 Mar-05 Apr-05 May-05 Jun-05 Jul-05 Aug-05 Sep-05 Oct-05 Nov-05 Dec-05Rate F MB 50/50 Approach0.00 0.10 0.20 0.30 0.40 0.50Nov-04 Dec-04 Jan-05 Feb-05 Mar-05 Apr-05 May-05 Jun-05 Jul-05 Aug-05 Sep-05 Oct-05 Nov-05 Dec-05Rate F MC Figure 3-9. Model p ˆt Fs M. capture probabilities for all three modeling approaches. This model had time dependent capture probabilities, fixed M estimates, with F estimates fixed by open and closed harvest seasons. These approaches differ only in the fates assigned to 17 fish last heard during the summer of 2005.

PAGE 53

41 Fishing Mortality 0.00 0.10 0.20 0.30 0.40 0.50Nov-04 Dec-04 Jan-05 Feb-05 Mar-05 Apr-05 May-05 Jun-05 Jul-05 Aug-05 Sep-05 Oct-05 Nov-05 Dec-05Rate Emigrants and Fishing Mortalities Approach Natural Mortalities Approach 50/50 ApproachA Natural Mortality0.00 0.10 0.20 0.30 0.40 0.50Nov-04 Dec-04 Jan-05 Feb-05 Mar-05 Apr-05 May-05 Jun-05 Jul-05 Aug-05 Sep-05 Oct-05 Nov-05 Dec-05Rate Emigrants and Fishing Mortalities Approach Natural Mortalities Approach 50/50 ApproachB Figure 3-10. Comparison of fishing mortality and natural mortality estimates using the p ˆt Ft Mt model (time dependent capture proba bilities, F and M) for all three fate assignment approaches. These approaches differ only in the fates assigned to 17 fish last heard during the summer of 2005.

PAGE 54

42 Catch Curve ages 5 through 8y = -0.3709x + 5.8734 R2 = 0.4875 0.00 1.00 2.00 3.00 4.00 5.00 0246810121416agenatural log of # of fis h A Catch Curve ages 6 through 8y = -0.8346x + 9.2739 R2 = 0.9955 0.00 1.00 2.00 3.00 4.00 5.00 0246810121416agenatural log of # of fis h B Figure 3-11. Catch curve analysis of 286 d ead adult snook collected during the red tide bloom of 2005.

PAGE 55

43 tp. 0.50 0.60 0.70 0.80 0.90 1.00Nov-04 Dec-04 Jan-05 Feb-05 Mar-05 Apr-05 May-05 Jun-05 Jul-05 Aug-05 Sep-05 Oct-05 Nov-05 Dec-05Rate Apparent Survival p^A tpt0.50 0.60 0.70 0.80 0.90 1.00Nov-04 Dec-04 Jan-05 Feb-05 Mar-05 Apr-05 May-05 Jun-05 Jul-05 Aug-05 Sep-05 Oct-05 Nov-05 Dec-05Rate Apparent Survival p^B Figure 3-12. Results for the Cormack-Jolly-Seber models tp ˆ. and tp ˆt These models each had time dependent apparent survival ( ) but differ in time dependent and fixed captur e probabilities (p ˆ).

PAGE 56

44 Table 3-1. Location of study site exits and creeks and the determined number of VR2 receivers needed. Exit Location VR2 Receivers Creek Location VR2 Receivers New Pass 2 Bowless Creek 3 Big Pass 3 Whitaker Bayou 2 Longboat Pass 3 Phillippi Creek 1 Cortez (ICW) 3 North Creek 3 Venice (ICW) 1 South Creek 2 Table 3-2. Fates as of January 2006 of 66 tagge d, harvestable size fish used to estimate mortality. Censored Emigrants Fishing Mortalities Natural Mortalities Unknown Still Alive 4 8 16 6 17 15 Table 3-3. 2005 yearly estimates of fishi ng mortality, natural mortality, and their corresponding AIC values attained fo r all three approaches of fate assignments of questionable fish not heard after the red tide events of summer 2005. The number of fish used is noted for each approach. Model Emigrants and Fishing Mortalities Approach (46 fish) Natural Mortalities Approach (55 fish) 50/50 Approach (50 fish) F M AIC F M AIC F M AIC p ˆt Ft Mt 0.66 0.16 228.97 0.41 0.65 232.60 0.48 0.42 233.32 p ˆ. F. M. 0.60 0.12 234.71 0.24 0.60 285.19 0.48 0.36 253.56 p ˆ. Fs M. 0.55 0.12 236.05 0.30 0.60 282.43 0.42 0.36 252.59 p ˆt Fs M. 0.51 0.12 219.32 0.30 0.60 261.37 0.39 0.36 232.15 p ˆt Fs Mt 0.50 0.16 225.76 0.30 0.65 223.81 0.34 0.50 218.78 Table 3-4. 2005 yearly total mortality estimat es by modeling total mort ality directly using methods from Hightower et. al (2001 ) and by simply adding together the Hightower fishing and natural mortality estimates. Model Emigrants and Fishing Mortalities Approach Natural Mortalities Approach 50/50 Approach Direct Addition Direct Addition Direct Addition p ˆt Ft Mt 0.80 0.82 1.08 1.06 0.93 0.90 p ˆ. F. M. 0.72 0.72 0.93 0.84 0.81 0.84 p ˆ. Fs M. 0.72 0.67 0.92 0.90 0.81 0.78 p ˆt Fs M. 0.70 0.63 0.90 0.90 0.77 0.75 p ˆt Fs Mt 0.68 0.66 0.87 0.95 0.77 0.84

PAGE 57

45 Table 3-5. Estimates of total mortality (Z) were used along with natural mortality (M) estimates from the Hoenig equation and estimates of emigration rate (E) to figure fishing mortality (F) from all Co rmack-Jolly-Seber models performed. AIC values and degrees of freedom (D F) are shown for each model used (t indicates a time dependent variable, a pe riod indicates a fixed variable, and an s indicates a variable fixed by open and closed harvest seasons). Model AIC DF Z M – Hoenig E F tp ˆ. 698.58 16 0.56 0.21 0.04 0.31 tp ˆt 699.22 30 0.54 0.21 0.04 0.29 .p ˆt 704.57 16 0.43 0.21 0.04 0.18 sp ˆt 706.51 17 0.43 0.21 0.04 0.18 p ˆ. 713.23 2 0.47 0.21 0.04 0.22 Table 3-6. Review of the methods and result s for the mortality estimations. The ranges for the Hightower method are based on a ll five models used in the three different modeling approaches. Mortality Component Hightower Method Results Alternative Method Results Z 0.68 1.08 Catch Curve: 0.37 0.83 F 0.24 0.66 CJS: 0.18 0.31 M 0.12 0.65 Hoenig equation: 0.21

PAGE 58

46 CHAPTER 4 DISCUSSION Acoustic Array, Tagging and Detection A key factor in this stud y was the ability to determine whether a tagged snook had emigrated from Sarasota Bay with a high degr ee of certainty. The tag reception rate, based on the initial range tests described earlie r, was approximately 50%; however, this is a bit misleading. During the tag testing, all tags were heard by the receivers during every trial (100% detection rate), but not every po ssible tag signal was received. Thus, the 50% reception rate represented the chance of receiving all of the signals from a tag as it moves through the receiver de tection field in the pass. All tags were detected on every test conducted, but only 50% of the total potenti al soundings were heard. While this reception rate was much lower than originally anticipated, the actua l performance of the receivers with tags implanted in snook was mo st likely better than the rates measured in the trials. The trials were conducted by dr ifting the tags through multiple pass or exit locations on a moving tide, which caused th e tag to constantly move through each location. Clements et al. (2005) and Heupel et al. (2006) found, and additional trials with the manual receiver showed, that fast flowing wa ter, such as during tide changes, greatly reduced reception distance. The initial range tests were performed during the falling tide change so the tags would drift past the rece ivers without the assist ance or interference from a boat motor to mimic snook movements as closely as possible and minimize boat motor noise interference. Boat noise, although not the greatest source of interference, can negatively influence range tests and acousti c array performance (Klimley et al. 1998;

PAGE 59

47 Lacroix and Voegeli 2000). The data from the receivers and ma nual tracking suggest, however, that snook made erratic movements within the passes, and thus were not moving constantly through th e receiver reception field. Also, snook primarily were found to move throughout the bay and only a ggregated in the passes during the spawning season. This erratic movement behavior le d to higher detection probabilities of animals on each of our time intervals, and overall p ˆranged from about 0.70 to 1.00 per month. Given that snook were observed to move erratically through th e passes and not in a fairly straight line as our test tags moved, it is likely that the average detection frequency was greater than 50%. On trips with the ma nual receiver during periods of low water movement in the passes or on trips within the bay, detection of 300-1000 meters was regularly observed. Heupel et al. (2006) observed reception ranges of around 800-meters with similar tags and receiv ers in similar locations, but did find that at these long distances the reception rate does decrease. Tag detection rate was highest during peri ods of seasonal temperature change when snook were actively moving between summer sp awning passes (Taylo r et al. 1998) and winter thermal refuges (Marsha ll 1958; Howells et al. 1990). During this time, fish were moving long distances, increasi ng the likelihood of swimming within the detection range of a receiver. Detection rates dropped drastically during the summer months of 2005. This was expected to some degree as snook would likely have left the bay to spawn in the passes and then, possibly, move to the beaches to f eed (Taylor et al. 1998). The winter of 20052006 was expected to serve as a “check” period for the tagged fish that had survived the spring and fall harvest seasons and the summer red tide events, as the remaining snook

PAGE 60

48 would have to return to the creeks as a ther mal refuge. However, detection rates did not increase during the cold water period. If the tagged fish were alive and viable, they should have been detected as they reente red the study site and the creeks the following winter. The extreme drop in detection rate seen starting in July 2005 was most likely from the effects of the intense red tide bloom th at occurred in the study area at that time. Model Assumptions The Hightower et al. (2001) method has seve ral assumptions; some of which can be problematic. If a fish emigrated from the st udy site, but was not detected on a pass/exit receiver as it left, it would be considered a fi shing mortality as the last data point for the fish would have most likely come from a r eceiver within the site. This would obviously affect the model results by inflating the F estimates. Attempts were made early in the study to reduce the chance of this happening. Much effort went into the pass/exit receiver placement and range testing to assu re that there was no r oute a snook could take to circumvent detection. Tags were detected on every test performed after the optimal receiver location was determined. Alt hough undetected escape possibly could have happened and should be considered when exam ining these results, th is possibility is believed to be unlikely. Equal vulnerability of all fish used in the F and M estimation was assumed, however, vulnerability likely changed with ag e or size. The assessment compensates for this by using age-7 fish as a reference point for F estimates (Muller and Taylor 2006). This is difficult to account for in this type of study where age of the tagged fish is unknown. Using only harvestabl e size fish was the best opt ion possible to reduce the effects of violation of this a ssumption (Hightower et. al 2001).

PAGE 61

49 Tag failure could have also been a problem Tag failure could not be distinguished from fishing mortality as both situations would have looked id entical in the data either would result in a fish seeming to be no longe r present in the study site. Tag failure was always a possibility, but all tags were in working order prior to deployment and all 10 returned tags (9 from angler and 1 from redtide killed fish) conti nued to work throughout the duration of their e xpected battery life. Fishing and Natural Mortality Analysis It was thought that the best approach to estimating fishing and natural mortality was to stray from the guidelines of Hightower et al. (2001); the lost fish as natural mortalities approach. In this approach, the 17 fish that were last heard during the red tide event were considered natural mortalities no matter the location of their last detection (inside array or on pass/exit). These fates were different than those that would have been assigned if the guidelines from Hightower et al. (2001) were followed. These fates would have been fishing mortalities and emigrants, but this did not seem to be supported by what was being observed in the field. The results from this modeling approach s howed that fishing mortality for snook is high and similar to the values estimated in the stock assessmen t during the harvest seasons and that natural mortal ity from red tide events can also be an unexpectedly large source of mortality. These results are the first direct estimates of natural mortality for a sportfish population in Florida and are also the first direct estimates of natural mortality from a red tide event anywhere. Although the fate assignments were subj ective, uncertainty was evaluated by modeling with three fate assignment approach es. While the magnitude of the parameter estimates from these approaches was differen t, the overall patterns in fishing mortality

PAGE 62

50 were similar between the different approaches. Incorrect fate assignments for even a low number of fish would influence the estimates A key assumption in the lost fish as emigrants and fishing mortalities approach wa s that the eight fish that were unaccounted for in the array were assumed to have died due to fishing mortality. If these fish did disappear from the system during this time peri od due to fishing, this would indicate a high level of illegal harv est during the summer closed period – a scenario that is likely of interest to resource ag encies. However, given the sudde n drop in the frequency of tag relocations, the closed harvest season, the exceptionally large red tide event, and the observed large numbers of dead snook during this red tide bloom, it is more likely that these eight fish died due to natural mortality. Total Mortality Estimation Total mortality estimates from this study were higher than those estimated by the snook stock assessment (Muller and Taylor 2006). This is not unexpected, as the assessment estimates were for the entire coas t, and the red-tide event which caused much of the mortality for this study was more locali zed. There appeared to be little difference between the Z estimates atta ined through the Hightower modeling approach versus simply adding the F and M estimates together. This implied that total mortality was higher than expected, even if there was uncertainty about how that total was partitioned. Estimates of Z attained through the catch curv e analyses were lower than those from the Hightower models likely because not all assumpti ons of the catch-curve were met. As in Muller and Taylor (2006), the use of a cat ch-curve here is just to provide an approximation of the total mortality values. In this case, the catch-curve was not very informative because few ages were represented in the sample and the high variation in

PAGE 63

51 snook growth patterns results in no age likel y ever being fully recruited (Muller and Taylor 2006). The CJS models showed drops in apparent survival in April and August of 2005, at the same time the other models indicated mark ed increases in mortality. This method used only capture history and was not reliant on the assignm ent of natural or fishing mortality as in the Hightower models (Cor mack 1964; Jolly 1965; Seber 1965, Pollock et al. 1990). However, this method did require that fate assignments with regard to emigration be made. Given the extensive testing of the receiver array prior to the study, it was believed that this design could reliably detect an emigrant, thus these survival trends were likely not biased by mis-assigning a fish as an emigrant. The F estimates from the CJS models were also similar to the F estimates from the lost fish as natural mortalities approach Hightower type models. Comparison to Assessment The FWC 2005 Snook Stock Assessment (Muller and Taylor 2006) reported a yearly F range for 2004 of 0.36 to 0.48 per ye ar. These estimates were obtained using fisherydependent CAA models with an assumed natural mortality estimate of 0.25 per year. This project estimated F values th at ranged between 0.30 and 0.66 per year depending on which approach was used to as sign fates to questionable fish. These estimates were obtained without fisher depe ndence and were independent of the natural mortality estimates which ranged from 0. 12 to 0.65 per year, depending on fate assignment. Although these were estimates made in successive years, some general comparisons are possible. Snook harvest is indeed high and the methods used by FWC are most likely providing accurate estimates of F. However, a large discrepancy exists between the estimates of natural mortality. This is due to the high, regional natural

PAGE 64

52 mortality observed in Sarasota Bay in 2005. This high rate is most likely not constant, but rather event driven. It may, however, become more important to consider these possibly large regional sources of natural mortality if red ti de becomes more prevalent in the future. Method Comparison This project was funded for a number of r easons: 1) to respond to a need mentioned in the assessment for regional data, 2) to di rectly estimate M, 3) to provide a fisherindependent estimate of F, and 4) to provide an example of the benefits, problems, and utility of using this method with other species in Florida. In formation regarding this last reason may be the most useful in the long r un. Walters and Marte ll (2004) extensively discuss tradeoffs in fisheries management, bot h in the methods that are used to make decisions and the tradeoffs that must o ccur when making social and management decisions related to limited fish resources This study is an example of evaluating tradeoffs in estimating model parameters wh ich can provide different information for making a management decision. Using teleme try methods to directly estimate fishing and natural mortality has several key attribut es including: 1) dire ct, fisher-independent measures of F and M, 2) continuous data collection for multi ple fish/organisms simultaneously, and 3) movement and habitat da ta (Hightower et al. 2001; Pine et al. 2003; Walters and Martell 2004). One way in which telemetry can be used to compensate for the problems with the traditio nal assessment methods mentioned earlier, is by combining it with the other methods; such as using telemetry to estimate reporting rate and natural mortality in a traditional ta gging study (Pollock et al. 2004). This would solve two of the problems associated with the traditional tagging method. The tradeoff for this telemetry method is that it also has its disadvantages. Thes e include: 1) expense

PAGE 65

53 (high equipment cost), 2) high maintenance (high man hours with cer tain skills required depending on how receivers are deployed), 3) potential problems with fate assignment and potential need to assume fates of questionable fish, and 4) potentially strong influence of fate mis-classification on parameter estimates due to small sample size. Utility in Florida This project was done using snook as th e study species, however, this method can be used on many marine species and the st udy could be designed for multiple species simultaneously. Telemetry methods to directly estimate F and/or M have been used for a variety of fish species incl uding striped bass (Hightower et al. 2001), blacktip sharks (Heupel and Simpfendorfer 2002), largemouth bass (Waters et al. 2005), salmon (Bendock and Alexandersdottir 1993), and lingc od (Starr et al. 2005) for example. Telemetry has been used to assess movement and habitat use of fish species in many studies (Standora and Nelson 1977; Klimley et al. 1988; Lagardre et al. 1990; Lacroix and McCurdy 1996; Arendt et al. 20 01b; and Heupel et al. 2003). Stock assessments for other fish specie s in Florida including red drum (Murphy 2005), spotted seatrout (Murphy 2003), mullet (Mahmoudi 2005), and sheepshead (Munyandorero 2006); all use CAA methods simila r to the snook stock assessment. In each of these assessments, the authors have id entified the need for comparative mortality estimates to provide a “check” of the mortal ity estimates currently estimated to help better manage these stocks. Telemetry studies such as this could fill that need. Future applications of this method may want to work in smaller, more geographically closed systems than what wa s used for this study (e.g., Klimley et al. 1998; Lacroix and Voegeli 2000; Clements et al 2005; Heupel et. al 200 6). Large, more open systems require more recei vers to close the exits, more field effort to identify

PAGE 66

54 natural mortalities through manual tracking, a nd a greater chance that a fish will emigrate from the system or not be identified as mortality. Heupel and Simpfendorfer (2002) and Heupel et al. (2003) worked in an ideal situ ation where they were able to have total acoustic coverage for their study site. This el iminated the need for assuming fates of lost fish as the data resolution was high enough to identify an animal’s location at all times allowing swim-speed, dispersion rate, predati on, and detailed moveme nt patterns to be calculated for juvenile blacktip sharks. In creasing the number of receivers, or an improvement in the available receiver technology, could reduce the need for manual tracking and would likely increase the detail with which mortality and movement patterns could be examined (Klimley et al. 1998; Lacr oix and Voegeli 2000; Clements et. al 2005; Heupel et al. 2006). Conclusions A major finding of this study is that natural mortality can be very high related to red tide blooms at discrete spatial and tempor al levels. Much of the data used in FWRI assessments is collected at regional field offi ces around the state. However, much of the assessment estimations are conducted at la rger, coast wide scales. This study demonstrates that natural mortality events may be large locally and if red tide events increase with frequency, intensity, or spatial area, then these larger natural mortality rates may need to be incorporated into the stock assessment programs. This study provided the first direct meas urements of F and M for any game-fish species in Florida and will serv e as an excellent comparison to the derived estimates for snook currently used in the stock assessment. These direct measurements identified temporal changes in natural mortality due to red tide events. This study demonstrated the problems, but also the utility of using a large acoustic ta gging program in an estuarine

PAGE 67

55 setting which could be used to address iden tified management needs for other species; such as providing rapid estimation of mortal ity rates in evaluating the effectiveness of newly implemented harvest regulatio ns or other management actions.

PAGE 68

APPENDIX HIGHTOWER MODELS AND CORMAC K-JOLLY-SEBERS MODEL RESULTS

PAGE 69

57Table A-1. Results for models assuming that all fish last heard during the summer of 2005 were emigrants & fishing mortalities p ˆt Ft Mt p ˆ. F. M. p ˆ. Fs M. p ˆt Fs M. p ˆt Fs Mt Variable Value SE Value SE Value SE Value SE Value SE F-Nov 2004 0.00 0.01 0.05 0.01 0.06 0.02 0.05 0.02 0.05 0.02 F-Dec 2004 0.00 0.01 0.05 0.01 0.04 0.01 0.03 0.01 0.03 0.01 F-Jan 2005 0.00 0.02 0.05 0.01 0.04 0.01 0.03 0.01 0.03 0.01 F-Feb 2005 0.05 0.04 0.05 0.01 0.06 0.02 0.05 0.02 0.05 0.02 F-Mar 2005 0.02 0.02 0.05 0.01 0.06 0.02 0.05 0.02 0.05 0.02 F-Apr 2005 0.07 0.04 0.05 0.01 0.06 0.02 0.05 0.02 0.05 0.02 F-May 2005 0.00 0.01 0.05 0.01 0.04 0.01 0.03 0.01 0.03 0.01 F-Jun 2005 0.02 0.03 0.05 0.01 0.04 0.01 0.03 0.01 0.03 0.01 F-Jul 2005 0.03 0.04 0.05 0.01 0.04 0.01 0.03 0.01 0.03 0.01 F-Aug 2005 0.18 0.09 0.05 0.01 0.04 0.01 0.03 0.01 0.03 0.01 F-Sep 2005 0.12 0.08 0.05 0.01 0.06 0.02 0.05 0.02 0.05 0.02 F-Oct 2005 0.08 0.07 0.05 0.01 0.06 0.02 0.05 0.02 0.05 0.02 F-Nov 2005 0.03 0.07 0.05 0.01 0.06 0.02 0.05 0.02 0.05 0.02 F-Dec 2005 0.04 0.12 0.05 0.01 0.04 0.01 0.03 0.01 0.03 0.01 F-Jan 2006 0.47 0.26 0.05 0.01 0.04 0.01 0.03 0.01 0.03 0.01 M-Nov 2004 0.00 0.00 0.01 0.01 0.01 0.01 0.01 0.01 0.00 0.01 M-Dec 2004 0.00 0.00 0.01 0.01 0.01 0.01 0.01 0.01 0.00 0.00 M-Jan 2005 0.00 0.00 0.01 0.01 0.01 0.01 0.01 0.01 0.00 0.00 M-Feb 2005 0.00 0.00 0.01 0.01 0.01 0.01 0.01 0.01 0.00 0.00 M-Mar 2005 0.00 0.00 0.01 0.01 0.01 0.01 0.01 0.01 0.00 0.00 M-Apr 2005 0.08 0.04 0.01 0.01 0.01 0.01 0.01 0.01 0.08 0.04 M-May 2005 0.00 0.00 0.01 0.01 0.01 0.01 0.01 0.01 0.00 0.00 M-Jun 2005 0.00 0.00 0.01 0.01 0.01 0.01 0.01 0.01 0.00 0.00 M-Jul 2005 0.00 0.00 0.01 0.01 0.01 0.01 0.01 0.01 0.00 0.00 M-Aug 2005 0.08 0.05 0.01 0.01 0.01 0.01 0.01 0.01 0.08 0.05 M-Sep 2005 0.00 0.00 0.01 0.01 0.01 0.01 0.01 0.01 0.00 0.00 M-Oct 2005 0.00 0.00 0.01 0.01 0.01 0.01 0.01 0.01 0.00 0.00 M-Nov 2005 0.00 0.00 0.01 0.01 0.01 0.01 0.01 0.01 0.00 0.00 M-Dec 2005 0.00 0.00 0.01 0.01 0.01 0.01 0.01 0.01 0.00 0.00 M-Jan 2006 0.00 0.00 0.01 0.01 0.01 0.01 0.01 0.01 0.00 0.01

PAGE 70

58Table A-2. Results for all models assuming that all fish la st heard during the summer of 2005 were natural mortalities p ˆt Ft Mt p ˆ. F. M. p ˆ. Fs M. p ˆt Fs M. p ˆt Fs Mt Variable Value SE Value SE Value Variable Value SE Value SE F-Nov 2004 0.00 0.01 0.02 0.01 0.05 0.02 0.05 0.02 0.05 0.02 F-Dec 2004 0.00 0.01 0.02 0.01 0.00 0.01 0.00 0.01 0.00 0.00 F-Jan 2005 0.01 0.02 0.02 0.01 0.00 0.01 0.00 0.01 0.00 0.00 F-Feb 2005 0.04 0.04 0.02 0.01 0.05 0.02 0.05 0.02 0.05 0.02 F-Mar 2005 0.02 0.03 0.02 0.01 0.05 0.02 0.05 0.02 0.05 0.02 F-Apr 2005 0.08 0.04 0.02 0.01 0.05 0.02 0.05 0.02 0.05 0.02 F-May 2005 0.00 0.00 0.02 0.01 0.00 0.01 0.00 0.01 0.00 0.00 F-Jun 2005 0.00 0.00 0.02 0.01 0.00 0.01 0.00 0.01 0.00 0.00 F-Jul 2005 0.00 0.02 0.02 0.01 0.00 0.01 0.00 0.01 0.00 0.00 F-Aug 2005 0.00 0.04 0.02 0.01 0.00 0.01 0.00 0.01 0.00 0.00 F-Sep 2005 0.12 0.07 0.02 0.01 0.05 0.02 0.05 0.02 0.05 0.02 F-Oct 2005 0.08 0.07 0.02 0.01 0.05 0.02 0.05 0.02 0.05 0.02 F-Nov 2005 0.03 0.07 0.02 0.01 0.05 0.02 0.05 0.02 0.05 0.02 F-Dec 2005 0.04 0.12 0.02 0.01 0.00 0.01 0.00 0.01 0.00 0.00 F-Jan 2006 0.46 0.26 0.02 0.01 0.00 0.01 0.00 0.01 0.00 0.00 M-Nov 2004 0.00 0.00 0.05 0.01 0.05 0.01 0.05 0.01 0.00 0.00 M-Dec 2004 0.00 0.00 0.05 0.01 0.05 0.01 0.05 0.01 0.00 0.00 M-Jan 2005 0.00 0.00 0.05 0.01 0.05 0.01 0.05 0.01 0.00 0.00 M-Feb 2005 0.00 0.00 0.05 0.01 0.05 0.01 0.05 0.01 0.00 0.00 M-Mar 2005 0.00 0.00 0.05 0.01 0.05 0.01 0.05 0.01 0.00 0.00 M-Apr 2005 0.07 0.04 0.05 0.01 0.05 0.01 0.05 0.01 0.08 0.04 M-May 2005 0.00 0.00 0.05 0.01 0.05 0.01 0.05 0.01 0.00 0.00 M-Jun 2005 0.05 0.03 0.05 0.01 0.05 0.01 0.05 0.01 0.04 0.03 M-Jul 2005 0.18 0.07 0.05 0.01 0.05 0.01 0.05 0.01 0.17 0.07 M-Aug 2005 0.35 0.11 0.05 0.01 0.05 0.01 0.05 0.01 0.35 0.11 M-Sep 2005 0.00 0.00 0.05 0.01 0.05 0.01 0.05 0.01 0.00 0.00 M-Oct 2005 0.00 0.00 0.05 0.01 0.05 0.01 0.05 0.01 0.00 0.00 M-Nov 2005 0.00 0.00 0.05 0.01 0.05 0.01 0.05 0.01 0.00 0.00 M-Dec 2005 0.00 0.00 0.05 0.01 0.05 0.01 0.05 0.01 0.00 0.00 M-Jan 2006 0.00 0.00 0.05 0.01 0.05 0.01 0.05 0.01 0.00 0.00

PAGE 71

59Table A-3. Results for all models assuming that the fish last heard during the summer of 2005 were 50/50 emigrants and fishing mortalities/natural mortalities p ˆt Ft Mt p ˆ. F. M. p ˆ. Fs M. p ˆt Fs M. p ˆt Fs Mt Variable Value SE Value SE Value Variable Value SE Value SE F-Nov 2004 0.00 0.01 0.04 0.01 0.05 0.02 0.05 0.02 0.06 0.02 F-Dec 2004 0.00 0.01 0.04 0.01 0.02 0.01 0.01 0.01 0.00 0.00 F-Jan 2005 0.00 0.02 0.04 0.01 0.02 0.01 0.01 0.01 0.00 0.00 F-Feb 2005 0.05 0.04 0.04 0.01 0.05 0.02 0.05 0.02 0.06 0.02 F-Mar 2005 0.02 0.02 0.04 0.01 0.05 0.02 0.05 0.02 0.06 0.02 F-Apr 2005 0.07 0.04 0.04 0.01 0.05 0.02 0.05 0.02 0.06 0.02 F-May 2005 0.00 0.00 0.04 0.01 0.02 0.01 0.01 0.01 0.00 0.00 F-Jun 2005 0.00 0.00 0.04 0.01 0.02 0.01 0.01 0.01 0.00 0.00 F-Jul 2005 0.05 0.04 0.04 0.01 0.02 0.01 0.01 0.01 0.00 0.00 F-Aug 2005 0.04 0.05 0.04 0.01 0.02 0.01 0.01 0.01 0.00 0.00 F-Sep 2005 0.12 0.08 0.04 0.01 0.05 0.02 0.05 0.02 0.06 0.02 F-Oct 2005 0.09 0.07 0.04 0.01 0.05 0.02 0.05 0.02 0.06 0.02 F-Nov 2005 0.02 0.06 0.04 0.01 0.05 0.02 0.05 0.02 0.06 0.02 F-Dec 2005 0.01 0.12 0.04 0.01 0.02 0.01 0.01 0.01 0.00 0.00 F-Jan 2006 0.51 0.28 0.04 0.01 0.02 0.01 0.01 0.01 0.00 0.00 M-Nov 2004 0.00 0.00 0.03 0.01 0.03 0.01 0.03 0.01 0.00 0.00 M-Dec 2004 0.00 0.00 0.03 0.01 0.03 0.01 0.03 0.01 0.00 0.00 M-Jan 2005 0.00 0.00 0.03 0.01 0.03 0.01 0.03 0.01 0.00 0.00 M-Feb 2005 0.00 0.00 0.03 0.01 0.03 0.01 0.03 0.01 0.00 0.00 M-Mar 2005 0.00 0.00 0.03 0.01 0.03 0.01 0.03 0.01 0.00 0.00 M-Apr 2005 0.07 0.04 0.03 0.01 0.03 0.01 0.03 0.01 0.08 0.04 M-May 2005 0.00 0.00 0.03 0.01 0.03 0.01 0.03 0.01 0.00 0.00 M-Jun 2005 0.02 0.02 0.03 0.01 0.03 0.01 0.03 0.01 0.02 0.02 M-Jul 2005 0.14 0.06 0.03 0.01 0.03 0.01 0.03 0.01 0.17 0.06 M-Aug 2005 0.18 0.08 0.03 0.01 0.03 0.01 0.03 0.01 0.23 0.09 M-Sep 2005 0.00 0.00 0.03 0.01 0.03 0.01 0.03 0.01 0.00 0.00 M-Oct 2005 0.00 0.00 0.03 0.01 0.03 0.01 0.03 0.01 0.00 0.00 M-Nov 2005 0.00 0.00 0.03 0.01 0.03 0.01 0.03 0.01 0.00 0.00 M-Dec 2005 0.00 0.00 0.03 0.01 0.03 0.01 0.03 0.01 0.00 0.00 M-Jan 2006 0.00 0.00 0.03 0.01 0.03 0.01 0.03 0.01 0.00 0.01

PAGE 72

60Table A-4. Results from all Cormack-Jolly_Seber models. tp ˆ. tp ˆt .p ˆt sp ˆt .p ˆ. Variable Value SE Value SE Value Variable Value SE Value SE –Nov 2004 1 0.00 10.000.920.01 0.910.020.910.01 -Dec 2004 0.98 0.02 0.980.020.82 0.12 0.930.020.910.01 -Jan 2005 0.98 0.02 0.990.020.92 0.04 0.930.020.910.01 -Feb 2005 0.96 0.03 0.960.040.73 0.06 0.910.020.910.01 -Mar 2005 0.98 0.03 0.960.030.77 0.06 0.910.020.910.01 -Apr 2005 0.83 0.06 0.850.050.98 0.02 0.910.020.910.01 -May 2005 0.98 0.03 0.980.030.80 0.06 0.930.020.910.01 -Jun 2005 0.97 0.04 0.950.030.86 0.05 0.930.020.910.01 -Jul 2005 0.86 0.06 0.870.060.98 0.02 0.930.020.910.01 -Aug 2005 0.69 0.08 0.700.080.91 0.05 0.930.020.910.01 -Sep 2005 0.88 0.07 0.880.070.84 0.08 0.910.020.910.01 -Oct 2005 0.92 0.07 0.930.070.90 0.07 0.910.020.910.01 Nov 2005 0.93 0.07 0.950.090.80 0.09 0.910.020.910.01 -Dec 2005 0.83 0.11 0.840.150.76 0.10 0.930.020.910.01 -Jan 2006 0.61 0.15 0.7166.560.73 0.12 0.930.020.910.01 p-Nov 2004 0.85 0.02 0.820.120.51 0.13 0.820.120.850.02 p-Dec 2004 0.85 0.02 0.920.040.92 0.01 0.920.040.850.02 p-Jan 2005 0.85 0.02 0.720.060.92 0.01 0.730.060.850.02 p-Feb 2005 0.85 0.02 0.760.060.92 0.01 0.770.060.850.02 p-Mar 2005 0.85 0.02 0.980.020.92 0.01 0.980.020.850.02 p-Apr 2005 0.85 0.02 0.820.060.92 0.01 0.790.060.850.02 p-May 2005 0.85 0.02 0.860.050.92 0.01 0.860.050.850.02 p-Jun 2005 0.85 0.02 0.970.030.92 0.01 0.980.020.850.02 p-Jul 2005 0.85 0.02 0.890.060.92 0.01 0.910.050.850.02 p-Aug 2005 0.85 0.02 0.880.070.92 0.01 0.850.080.850.02 p-Sep 2005 0.85 0.02 0.910.060.92 0.01 0.900.070.850.02 p-Oct 2005 0.85 0.02 0.800.090.92 0.01 0.800.090.850.02 pNov 2005 0.85 0.02 0.750.110.92 0.01 0.760.100.850.02 p-Dec 2005 0.85 0.02 0.780.140.92 0.01 0.740.120.850.02 p-Jan 2006 0.85 0.02 0.7166.560.92 0.01 0.520.130.850.02

PAGE 73

61 LIST OF REFERENCES Akaike, H. 1974. A New Look at the St atistical Model Identification. IEEE Transactions on Automatic Control 19(6):716-723 Anderson, D.R., K.P. Burnham, and G.C. White. 1998. Comparison of Akaike Information Criterion for Model Selection and Statistical Inference from Capturerecapture Studies. Journal of A pplied Statistics 25(2):263-282 Arendt, M.D., A.J. Lucy, and D.A. Evans. 2001b. Diel and Seasonal Activity Patterns of Adult Tautog, Tautoga onitis, in Lower Chesapeake Bay, Inferred form Ultrasonic Telemetry. Environmental Biology of Fishes 62:379-391. Bendock, T., and M. Alexandersdottir. 1993. Hooking Mortality of Chinook Salmon Released in the Kenai River, Alaska. North American Journal of Fisheries Management 13:540-549. Beverton, R.J., and S.J. Holt. 1959. A Review of the Life Spans and Mortality rates of Fish in Nature, and Their Relation to Growth and Other Physiological Characteristics. In CIBA Foundation Co lloquia on Aging, the Lifespans of Animals. (Wolshenholme, G.E.W. and M. O’Connor, Eds.) London, Churchill. Volume 5:142-180 Burnham, K.P., D.R. Anderson, G.C. White C. Brownie, and K.H. Pollock. 1987. Design and Analysis Methods for Fish Survival Experiments Based on ReleaseRecapture. American Fish eries Society, Maryland. Chapman, D. G., and D.S. Robson. 1960. The Analysis of a Catch Curve. Biometrics 16:354-368. Clements, S., D. Jepsen, M. Karnowski, and C.B. Schreck. 2005. Optimization on an Acoustic Telemetry Array for Detecting Transmitter-Implanted Fish. North American Journal of Fish eries Management 25:429-436 Cormack, R.M. 1964. Estimates of Survival from the Sighting of Marked Animals. Biometrika 51(3/4):429-438 Fore, P. L., and T. W. Schmidt. 1973. Biology of Juvenile and Adult Snook, Centropomus undecimalis, in the Ten Thousand Islands. Pp. XVI 1-18. in Ecosystems analysis of the Big Cypress Sw amp and estuaries. U. S. Environmental Protection AGENCY Region IV, Atlanta, Georgia Publication Number EPA 904/974-002.

PAGE 74

62 Gilmore, R. G., C. J. Donahoe and D.W. Cooke. 1983. Observations on the Distribution and Biology of the Common Snook, Centropomus undecimalis (Bloch). Florida Scientist 46: 313-336. Haddon, M. 2001. Modeling and Quantitative Methods in Fisheries. Chapman and Hall/CRC, Boca Raton, London, New York, and Washington, D.C. Heupel, M. R., and C. A. Simpfendorfer. 2002. Estimation of Mortality of Juvenile Blacktip Sharks, Carcharhinus limbatus, within a Nursery Area Using Telemetry Data. Canadian Journal of Fisher ies and Aquatic Sciences 59:624-632. Heupel, M.R., C.A. Simpfendorfer & R.E. Hueter. 2003. Running Before the Storm: Blacktip Sharks Respond to Falling Barome tric Pressure Associated with Tropical Storm Gabrielle. Journal of Fish Biology 63:1357-1363. Heupel, M.R., J.M. Semmens, and A.J. Hobday. 2006. Automated Acoustic Tracking of Aquatic Animals: Scales, Design and Depl oyment of Listening Station Arrays. Marine and Freshwater Research 57:1-13. Hightower, J.E., J. R., Jackson, and K. H. Pollock. 2001. Use of Telemetry Methods to Estimate Natural and Fishing Mortality of Striped Bass in Lake Gaston, North Carolina. Transactions of the North American Fisheries Society 130:557-567. Hilborn R., and C.J. Walters. 1992. Quan titative Fisheries Stock Assessment: Choice, Dynamics, and Uncertainty. Chapman & Ha ll, International Thomson Publishing, New York, New York. Hoenig, J.M. 1983. Empirical Use of Longevity Data to Estimate Mortality Rates. Fishery Bulletin 82:898-903 Howells, R. G., A. J. Sonski, P. L. Shafla nd, and B. D. Hilton 1990. Lower Temperature Tolerance of snook, Centropomus undecimalis. Short Papers and Notes. Northeast Gulf Sciences 11(2):155-158 Jensen, A. L. 1996. Beverton and Holt Life History in Variants Result from Optimal Trade-off of Reproduction and Survival. Canadian Journal of Fisheries and Aquatic Sciences 53:820-822. Jolly, G.M. 1965. Explicit Estimates from Ca pture-recapture Data With Both Death and Immigration-stochastic Model. Biometrika 52(1/2):225-247 Kendall, W.L. and J.D. Nichols. On the Us e of Secondary Capture-recapture Samples to Estimate temporary Emigra tion and Breeding Proportions Journal of Applied Statistics 22(5/6):751-762 Klimley, A.P., S.B. Butler, D.R. Nelson, A. Stull. 1988. Die Movements of Scalloped Hammerhead Sharks, Sphurna lewini, Griffith and Smith, to and from a Seamount in the Gulf of California. J ournal of Fish Biology 33:751-761

PAGE 75

63 Klimley, A.P., F. Voegeli, S.C. Beavers, and B.J. Le Boeuf. 1998. Automated Listening Stations for Tagged Marine Fishes. Ma rine Technology Society Journal 32:94-101. Lacroix, G.L., and P. McCurdy. 1996. Mi gratory Behavior of Post-smolt Atlantic Salmon during Initial Stages of Seaward Migration. Journal of Fish Biology 49:1086-1101. Lacroix, G.L., and F.A. Voegeli. 2000. Development of Automated Monitoring Systems for Ultrasonic Transmitters. In Advances in Fish Telemetry. Eds A. Moore and I. Russell, editors. CEFAS, Lowestoft. Lagardre, J.P., J.J. Ducamp, L. Favre, J. Mosneron Dupin, and M. Sprandio. 1990. A Method for the Quantitative Evaluation of Fish Movements in Salt Ponds by Acoustic Telemetry. Journal of Expe rimental Marine Bi ology and Ecology 141:221-236. Mahmoudi, B. 2005. The 2005 Update of th e Stock Assessment for Striped Mullet, Mugil cephalus in Florida. Fish and Wildlif e Conservation Commission, Florida Marine Research Institute, St. Petersburg, FL. Marshall, A. R. 1958. A Survey of the Snook Fishery of Florida, with Studies of the Biology of the Principal Species, Centropomus undecimalis (Bloch). Florida Board of Conservation Marine Research Labor atory Technical Series Number 22. McMicheal, R. H., Jr., K. M. Peters and G. R. Parsons. 1989. Early Life History of the Snook, Centropomus undecimalis in Tampa Bay, Florida. Northeast Gulf Science 10(2):113-126. Muller, R. G., and R. G. Taylor. 2006. The 2005 Stock Assessment Update of Common Snook, Centropomus undecimalis. Fish and Wildlife Conservation Commission, Florida Marine Research Inst itute, St. Petersburg, FL. Munyandorero, J., M.D. Murphy, and T.C. MacDonald. 2006. An Assessment of the Status of Sheepshead in Florida Wa ters through 2004. Fish and Wildlife Conservation Commission, Fl orida Marine Research Inst itute, St. Petersburg, FL. Murphy, M. D. 1997. Bias in ChapmanRobson and Least-squares Estimators of Mortality Rates for Steady-state Popula tions. Fishery Bulletin 95: 863-868. Murphy, M.D. 2003. A Stock Asse ssment of Spotted Seatrout Cynoscion nebulosus in Florida: Status of the Stocks through 2001. Fish and Wildlife Conservation Commission, Florida Mari ne Research Institute, St. Petersburg, FL. Murphy, M.D. 2005. A Stock Assessment of Red Drum, Sciaenops ocellatus, in Florida: Status of the Stocks through 2003. Fi sh and Wildlife Conservation Commission, Florida Marine Research Inst itute, St. Petersburg, FL.

PAGE 76

64 Ney, J. J. 1993. Dynamics of exploited fi sh populations. Inland Fisheries Management in North America. Edited by C. C. Kohler and W. A. Hubert. American Fisheries Society, Bethesda, Maryland. Pauly, D. 1980. On the Interrelations hips between Natural Mortality, Growth Parameters, and Mean Environmental Temperature in 175 fish stocks. Conseil International pour l’Exploration de la Mer Journal 39(2):175-192 Peters, K. M., R. E. Matheson, Jr. and R. G. Taylor. 1998b. Repr oduction and Early Life History of Common Snook, Centropomus undecimalis (Bloch), in Florida. Bulletin of Marine Sciences 62:509-529. Peterson, I., and J. S Wroblewski. 1984. Mortality Rates of Fishes in the Pelagic Ecosystem. Canadian Journal of Fi sheries and Aquatic Sciences 41:1117-1120. Pine, W.E. III, K.H. Pollock, J.E. Hightower T.J. Kwak, and J.A. Rice. 2003. A Review of Tagging Methods for Estimating Fish Population Size and Components of Mortality. Fisheries 28(10):10-23 Pollock, K. H., C. M. Bunck, S. R. Winterstein, and C. L. Chen. 1995. A Capturerecapture Survival Analysis Model for Radi o-tagged Animals. Journal of Applied Statistics 22:661-672. Pollock, K. H., J. M. Hoenig, W. S. Hearn, and B. Calingaert. 2001. Tag Reporting Rate Estimation: 1. An evaluation of the hi gh-reward tagging method. North American Journal of Fisheries Mana gement 21 (3): 521-532 Pollock, K. H., J. M. Hoenig, W. S. Hearn, and B. Calingaert. 2002. Tag Reporting Rate Estimation: 2. Use of High Reward Tagging and Observers in Multi-component Fisheries. North American Journa l of Fisheries Management 22:727-736 Pollock, K.H., H. Jiang, and J.E. Hightower 2004. Combining Telemetry and Fisheries Tagging Models to Estimate Fishing and Natural Mortality Rates. Transactions of the American Fisheries Society 133:639-648. Pollock, K.H. J.D. Nichols, C. Brownie, and J.E. Hines. 1990. St atistical Inference for Capture Recapture Experiment s. Wildlife Monograph 107. Quinn, T. J. II, and R. B. Deriso. 1999. Quantitative Fish Dynamics. Oxford University Press, Oxford. Ricker, W.E. 1975. Computation and Interpre tation of Biological Statistics of Fish Populations. Fisheries Research Board of Canada Bulletin 191. Rivas, L. R. 1986. Systematic Review of the Perciform Fishes of the Genus Centropomus. Copeia 1986(3):579–611.

PAGE 77

65 Sea Stats. 2000. Florida Fish & Wildlif e Conservation Commission, Florida Marine Research Institute. St. Petersburg, FL. http://www.fmri.usf.edu Seber, G.A. 1965. A Note on the Multiple-recapture Census. Biometrika 52(1/2):249259 Standora, E.A., and D.R. Nelson. 1977. A Telemetric Study of the Behavior of Freeswimming Pacific Angel Sharks, Squatina califonica. Southern California Academy of Sciences 76:193-201. Starr R.M., O'Connell V., Ralston S., Breaker L. 2005. Use of Acoustic Tags to Estimate Natural Mortality, Spillove r, and Movements of Lingc od (Ophiodon elongatzu) in a Marine Reserve. Marine Technology Society Journal 39 (1): 19-30. Steidinger K.A., and K. Haddad. 1981. Biol ogic and Hydrographic Aspects of Red Tide. Bioscience 31(11):814-819 Taylor R. G., H. J. Grier, and J. A. Wh ittington. 1998. Spawning Rhythms of Common Snook in Florida. Journal of Fisheries Biology 53:502-520. Taylor, R. G., J. A. Whittington, and H. J. Grier. 1993. Biology of Common Snook from the East and West Coasts of Florida. Study 3, Sect. 1. In investigations into nearshore and estuarine gamefish dist ributions and abundance, ecology, life history, and population genetics in Florida. Edited by R. E. Crabtree, T. M. Bert, and R. G. Taylor. FDNR/FMRI Rep. No F0165-F0296-88-93-C. U.S. Department of the Interior, Fish and Wild life Service, Washington, D.C. Taylor, R. G., J. A. Whittington, H. J. Grie r, R. E. Crabtree. 2000. Age, Growth, Maturation, and Protantric Sex Reversal in the Common Snook, Centropomus undecimalis, from South Florida Waters. Fisheries Bulletin 98(3): 612-624. Taylor, R. G., J. A. Whittington, D. E. Ha ymans. 2001. Catch-and-Release Mortality Rates of Common Snook in Florida. North American Journal of Fisheries Management. 21:70–75. Thue, E. B., E. S. Rutherford and D. G. B uker. 1982. Age, Growth and Mortality of the Common Snook, Centropomus undecimalis (Bloc), in the Everglades National Park, Florida. U.S. National Park Servi ce South Florida Research Center Report T683. Tringali, M. D. and T. M. Bert. 1996. Th e Genetic Stock Structure of Common Snook (Centropomus undecimalis). Canadian Journal of Fish eries and Aquatic Sciences. 53:974-984. Van Den Avyle, M. J. 1993. Dynamics of Exploited Fish Populations. Inland Fisheries Management in North America. Edited by C. C. Kohler and W. A. Hubert. American Fisheries Society, Bethesda, Maryland.

PAGE 78

66 Waters, D.S., R.L. Noble, and J.E. Hightower 2005. Fishing and Natural Mortality of Adult Largemouth Bass in a Tropical Reser voir. Transactions of the American Fisheries Society 134:563-571. Walters, C.J., and S. J. D. Martell 2004. Fisheries Ecology and Management. Princeton University Press, Princeton, New Jersey. White, G. C. 1983. Numerical Estimation of Survival Rates from Band Recovery and Biotelemetry Data. Journal of Wildlife Management 47:716-728. White, G.C., and K. P. Burnham. 1999. Program MARK: Survival Estimation from Populations of Marked Animals. Bird Study 46 Supplement, 120-138. White, G.C., and R.A. Garrott. 1990. Anal ysis of Wildlife Ra dio-tracking Data. Acadimic Press, San Diego, California. Williams, B.K., J.D.Nichols, and M.J. C onroy. 2002. Analysis and Management of Animal Populations. Academic Press, San Diego, California. Zehfuss, K. P., J. E. Hightower, and K. H. Pollock. 1999. Abundance of Gulf sturgeon in the Apalachicola River, Florida. Tran sactions of the Ameri can Fisheries Society 128: 130-143.

PAGE 79

67 BIOGRAPHICAL SKETCH Jason P. Bennett is from Kansas City, MO. He earned his B.S. in biology from the University of Missouri – Kansas City in 2003. He came to the University of Florida in 2003 and began his master’s research in 2004. He completed his master’s research in 2006.


Permanent Link: http://ufdc.ufl.edu/UFE0016067/00001

Material Information

Title: Using acoustic telemetry to estimate aatural and fishing mortality of common snook in Sarasota Bay, Florida
Physical Description: xii, 67 p. ; ill.
Language: English
Creator: Bennett, Jason Perry ( Dissertant )
Pine, William. ( Thesis advisor )
Allen, Mike ( Reviewer )
Frazer, Tom ( Reviewer )
Muller, Robert ( Reviewer )
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2006
Copyright Date: 2006

Subjects

Subjects / Keywords: Fisheries and Aquatic Sciences thesis, M.S
Dissertations, Academic -- UF -- Fisheries and Aquatic Sciences

Notes

Abstract: The common snook Centropomus undecimalis is a popular, saltwater gamefish found in southern Florida that has been actively managed by Florida's natural resource management agencies to prevent overexploitation since the 1950's. Despite increasingly restrictive management regulations, the status of the population and the effectiveness of these regulations remain uncertain. Most fisheries management activities are focused on regulating fishing mortality (F). Because of this, an important aspect of population assessments is an accurate estimate of F to provide insight into the magnitude of fishing mortality relative to natural mortality (M). Here, telemetry methods for relocating radio-tagged fish were used to estimate total mortality (Z), F, and M for adult common snook in Sarasota Bay, Florida. After tagging, fish were relocated using a series of remote, autonomous receivers in conjunction with active tracking efforts. These relocations were evaluated using a suite of a priori assumptions to determine whether an animal was live or dead. These fates were then converted to mortality rates using known-fate type models in program SURVIV. Three different modeling approaches were used to assess uncertainty in the mortality estimates to assigning fates to 17 fish that were not relocated after a large harmful algal bloom in Sarasota Bay during summer 2005. For the period from October 2004 through December 2005 Z values (0.68 - 1.08), F values (0.24 - 0.66), and M values (0.12 - 0.65) were estimated using different approaches which depended on how fish fates were assigned. However, estimated parameter values were similar to more traditional stock assessment mortality estimation methods and provided insight into the performance of the mortality estimation methods currently used in the assessment. Using telemetry methods to relocate fish and estimate mortality rates has advantages over traditional assessment methods, but each method has assumptions and potential biases that must be acknowledged. This study served as a good example of estimating fishing and natural mortality for a Florida fishery as a complement to current stock assessment methods.
Subject: mortality, snook, telemetry
General Note: Title from title page of source document.
General Note: Document formatted into pages; contains 79 pages.
General Note: Includes vita.
Thesis: Thesis (M.S.)--University of Florida, 2006.
Bibliography: Includes bibliographical references.
Original Version: Text (Electronic thesis) in PDF format.

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0016067:00001

Permanent Link: http://ufdc.ufl.edu/UFE0016067/00001

Material Information

Title: Using acoustic telemetry to estimate aatural and fishing mortality of common snook in Sarasota Bay, Florida
Physical Description: xii, 67 p. ; ill.
Language: English
Creator: Bennett, Jason Perry ( Dissertant )
Pine, William. ( Thesis advisor )
Allen, Mike ( Reviewer )
Frazer, Tom ( Reviewer )
Muller, Robert ( Reviewer )
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2006
Copyright Date: 2006

Subjects

Subjects / Keywords: Fisheries and Aquatic Sciences thesis, M.S
Dissertations, Academic -- UF -- Fisheries and Aquatic Sciences

Notes

Abstract: The common snook Centropomus undecimalis is a popular, saltwater gamefish found in southern Florida that has been actively managed by Florida's natural resource management agencies to prevent overexploitation since the 1950's. Despite increasingly restrictive management regulations, the status of the population and the effectiveness of these regulations remain uncertain. Most fisheries management activities are focused on regulating fishing mortality (F). Because of this, an important aspect of population assessments is an accurate estimate of F to provide insight into the magnitude of fishing mortality relative to natural mortality (M). Here, telemetry methods for relocating radio-tagged fish were used to estimate total mortality (Z), F, and M for adult common snook in Sarasota Bay, Florida. After tagging, fish were relocated using a series of remote, autonomous receivers in conjunction with active tracking efforts. These relocations were evaluated using a suite of a priori assumptions to determine whether an animal was live or dead. These fates were then converted to mortality rates using known-fate type models in program SURVIV. Three different modeling approaches were used to assess uncertainty in the mortality estimates to assigning fates to 17 fish that were not relocated after a large harmful algal bloom in Sarasota Bay during summer 2005. For the period from October 2004 through December 2005 Z values (0.68 - 1.08), F values (0.24 - 0.66), and M values (0.12 - 0.65) were estimated using different approaches which depended on how fish fates were assigned. However, estimated parameter values were similar to more traditional stock assessment mortality estimation methods and provided insight into the performance of the mortality estimation methods currently used in the assessment. Using telemetry methods to relocate fish and estimate mortality rates has advantages over traditional assessment methods, but each method has assumptions and potential biases that must be acknowledged. This study served as a good example of estimating fishing and natural mortality for a Florida fishery as a complement to current stock assessment methods.
Subject: mortality, snook, telemetry
General Note: Title from title page of source document.
General Note: Document formatted into pages; contains 79 pages.
General Note: Includes vita.
Thesis: Thesis (M.S.)--University of Florida, 2006.
Bibliography: Includes bibliographical references.
Original Version: Text (Electronic thesis) in PDF format.

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0016067:00001


This item has the following downloads:


Table of Contents
    Title Page
        Page i
        Page ii
    Dedication
        Page iii
    Acknowledgement
        Page iv
    Table of Contents
        Page v
        Page vi
    List of Tables
        Page vii
        Page viii
    List of Figures
        Page ix
        Page x
    Abstract
        Page xi
        Page xii
    Introduction
        Page 1
        Page 2
        Page 3
        Page 4
        Page 5
        Page 6
        Page 7
        Page 8
        Page 9
        Page 10
        Page 11
        Page 12
        Page 13
        Page 14
    Methods
        Page 15
        Page 16
        Page 17
        Page 18
        Page 19
        Page 20
        Page 21
        Page 22
        Page 23
        Page 24
        Page 25
    Results
        Page 26
        Page 27
        Page 28
        Page 29
        Page 30
        Page 31
        Page 32
        Page 33
        Page 34
        Page 35
        Page 36
        Page 37
        Page 38
        Page 39
        Page 40
        Page 41
        Page 42
        Page 43
        Page 44
        Page 45
    Discussion
        Page 46
        Page 47
        Page 48
        Page 49
        Page 50
        Page 51
        Page 52
        Page 53
        Page 54
        Page 55
    Appendix: Hightower model and Cormack-jolly-sebers model results
        Page 56
        Page 57
        Page 58
        Page 59
        Page 60
    References
        Page 61
        Page 62
        Page 63
        Page 64
        Page 65
        Page 66
    Biographical sketch
        Page 67
Full Text












USING ACOUSTIC TELEMETRY TO ESTIMATE NATURAL AND FISHING
MORTALITY OF COMMON SNOOK IN SARASOTA BAY, FLORIDA














By

JASON P. BENNETT


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


2006

































Copyright 2006

by

Jason P. Bennett
































This document is dedicated to my supporting family and the memory of my father.















ACKNOWLEDGMENTS

I would like to thank my graduate advisor, Dr. Bill Pine, and the rest of my

graduate committee; Dr. Mike Allen, Dr. Tom Frazer, and Dr. Robert Muller. I would

also like to thank everyone from the following organizations:

University of Florida, Department of Fisheries and Aquatic Sciences: Lauren

Marcinkiewicz, Drew Dutterer, Mark Rogers, Greg Binion, Travis Tuten, Galen

Kaufman, Vince Palitano, Nick Trippel

Florida Fish and Wildlife Conservation Commission: Ron Taylor, Luiz Barbieri,

Behzad Mahoudi, Alexis Trotter

Mote Marine Laboratory: Nate Brennen, Dr. Kenneth Leber, Meagan Darcy, Jason

Rock, Michelle Heupel, Colin Simpfendorfer, Dave Wilson, Brett Blackburn, Sarasota

Bay Explorers

I also thank Captain John Dixon, without whom we could not have caught any

snook.
















TABLE OF CONTENTS



A C K N O W L E D G M E N T S ................................................................................................. iv

LIST OF TABLES ................................................... vii

LIST OF FIGURES ....................................... .. ......... ............................ ix

CHAPTER

1 IN TR O D U C T IO N ........ .. ......................................... ..........................................1.

Study Species ................................................................................. . .. ...............2
L ife H history ............................................................................................ . 3
Stock Status in Florida...................................................................................... .4
P project Ju stification .. .... ... .................. ...... ................................6.. .. ... 6
Fishery Dependent Methods to Estimate F and M..........................................6...
Fishery Independent Methods to Estimate F and M ................... ..................... 9
P aram eter E stim ation ........................................................................................... 10

2 M E T H O D S .............. ................................................ ................................... . 15

F field M eth o d s ............................................................................................................. 15
S tu d y S ite ............................................................................................................. 1 5
Tagging Procedure ................................................................................... 15
T ag and R receiver D description ........................................................ ................ 16
A acoustic A rray N etw ork....................................... ....................... ............... 16
Array deployment and testing ................................................... 16
M anu al track in g ....................................................................... ............... 18
D ata A n aly sis .............................................................................................................. 1 8
A signing F ish F ates....................................................................... ............... 18
T otal M ortality E stim ates ...................................... ...................... ................ 22

3 R E S U L T S ................................................................................................................. .. 2 6

F field R e su lts ............................................................................................................... 2 6
T ag R ecep tio n ...................................................................................................... 2 6
F ish D etectio n ...................................................................................................... 2 6
D ata A naly sis R results .... ... ......................................... ....................... . .......... 28
A signing F ates .......................................................................................... 28


v









Secondary A analysis ................ .............. .............................................. 31

4 D IS C U S SIO N ................................................................................................... 4 6

Acoustic Array, Tagging and Detection ................................................................46
M odel A ssu m ption s ................................................... .. .......................................... 4 8
Fishing and Natural Mortality Analysis ................................................................49
T otal M ortality E stim ation ......................................... ........................ ................ 50
C om prison to A ssessm ent......................................... ........................ ............... 51
M ethod C om prison ................................................. ............................................ 52
U utility in F lo rid a ......................................................................................................... 5 3
C o n c lu sio n s............................................................................................................... .. 5 4

APPENDIX

HIGHTOWER MODEL AND CORMACK-JOLLY-SEBER MODEL RESULTS ........56

L IST O F R E F E R E N C E S ...................................................................................................6 1

BIOGRAPHICAL SKETCH ..................................................................................... 67















LIST OF TABLES


Table page

1-1. Review of snook regulations in the state of Florida...........................................14

1-2. Alternative methods for estimating natural mortality based on life history
a ttrib u te s ............................................................................................................... ... 1 4

2-1. Catch of study snook per gear type ..................................................... ................ 25

2-2. Descriptions of the five models used to estimate capture probabilities, F and M
in all three different scenarios. Each occasion for which an estimate was
generated w as a m onth .................................................................. ................ 25

2-3. Descriptions of the five models used to estimate Apparent Survival (0) and
capture probability (Error! Objects cannot be created from editing field codes.). Each
occasion for which an estimate was generated was a month. ..............................25

3-1. Location of study site exits and creeks and the determined number of VR2
receive ers n eed ed ...................................................................................................... 4 4

3-2. Fates as of January 2006 of 66 tagged, harvestable size fish used to estimate
m o rta lity ............................................................................................................. .. 4 4

3-3. 2005 yearly estimates of fishing mortality, natural mortality, and their
corresponding AIC values attained using the methods outlined in Hightower et.
al (2001) for all three approaches of fate assignments of questionable fish not
heard after the red tide events of summer 2005. The number of fish used is
noted for each approach.. .. ............................................................... ................ 44

3-4. 2005 yearly total mortality estimates by modeling total mortality directly using
methods from Hightower et. al (2001) and by simply adding together the
Hightower fishing and natural mortality estimates. ............................................44

3-5. Estimates of total mortality (Z) were used along with natural mortality (M)
estimates from the Hoenig equation and estimates of emigration rate (E) to
figure fishing mortality (F) from all Cormack-Jolly-Seber models performed.
AIC values and degrees of freedom (DF) are shown for each model used (t
indicates a time dependent variable, a period indicates a fixed variable, and an s
indicates a variable fixed by open and closed harvest seasons). ..............................45









3-6. Review of the methods and results for the mortality estimations. The ranges for
the Hightower method are based on all five models used in the three different
m odeling approaches .. ................................................................... ............... 45

A-1. Results for models assuming that all fish last heard during the summer of 2005
w ere em grants and fishing m ortalities ............................................... ................ 57

A-2. Results for all models assuming that all fish last heard during the summer of
2005 w ere natural m ortalities.............................................................. ................ 58

A-3. Results for all models assuming that the fish last heard during the summer of
2005 were 50/50 emigrants & fishing mortalities/natural mortalities................... 59

A-4. Results from all Cormack-Jolly_Seber models...................................................60















LIST OF FIGURES


Figure page

1-1. Total snook catch and release for the Gulf of Mexico and the Atlantic from
1 9 8 1 -2 0 0 4 ............................................................................................................ .. 12

1-2. Harvest, post-release fishing deaths, and total snook deaths due to fishing related
mortality for the Gulf of Mexico and the Atlantic from 1981-2004 .....................13

2-1. Map of Sarasota Bay with the study site borders of Cortez to the Northwest and
Venice to the Southeast (large arrows). Passes leaving the study site are
indicated by small arrows while creeks are indicated by stars...............................24

3-1. A detection history for every fish in the study. Each row is an individual fish
and each colum n is a m onth of the study ............................................ ................ 33

3-2. Percent detections and mean water temperatures by months of the study...............34

3-3. 2005 Red tide (K. brevis) cell counts for Sarasota Bay. The dashed line indicates
the cell count level that begins to cause mortality in some fish species ...............34

3-4. Model p t Ft Mt capture probabilities for all three modeling approaches. This
model had time dependent capture probabilities, F and M estimates. These
approaches differ only in the fates assigned to 17 fish last heard during the
su m m er of 2 0 0 5 ........................................................................................................ 3 5

3-5. Model p t Ft Mt F and M results for all three modeling approaches. This model
had time dependent capture probabilities, F and M estimates. These approaches
differ only in the fates assigned to 17 fish last heard during the summer of 2005. .36

3-6. Model p t Fs Mt capture probabilities for all three modeling approaches. This
model had time dependent capture probabilities, and M, with F estimates fixed
by open and closed harvest seasons. These approaches differ only in the fates
assigned to 17 fish last heard during the summer of 2005............... ...............37

3-7. Model P t Fs Mt F and M for all three modeling approaches. This model had
time dependent capture probabilities, and M, with F estimates fixed by open and
closed harvest seasons. These approaches differ only in the fates assigned to 17
fish last heard during the sum m er of 2005 ......................................... ................ 38









3-8. Model P t Fs M. capture probabilities for all three modeling approaches. This
model had time dependent capture probabilities, fixed M estimates, with F
estimates fixed by open and closed harvest seasons. These approaches differ
only in the fates assigned to 17 fish last heard during the summer of 2005. ...........39

3-9. Model P t Fs M. capture probabilities for all three modeling approaches. This
model had time dependent capture probabilities, fixed M estimates, with F
estimates fixed by open and closed harvest seasons. These approaches differ
only in the fates assigned to 17 fish last heard during the summer of 2005. ...........40

3-10. Comparison of fishing mortality and natural mortality estimates using the t Ft
Mt model (time dependent capture probabilities, F and M) for all three fate
assignment approaches. These approaches differ only in the fates assigned to 17
fish last heard during the sum m er of 2005 ......................................... ................ 41

3-11. Catch curve analysis of 286 dead adult snook collected during the red tide
b lo o m o f 2 0 0 5 ......................................................................................................... 4 2

3-12. Results for the Cormack-Jolly-Seber models Ptp. and Otp t. These models
each had time dependent apparent survival (0) but differ in time dependent and
fixed capture probabilities ( p ) ............................................... .......................... 43















Abstract of Thesis Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Master of Science

USING ACOUSTIC TELEMETRY TO ESTIMATE NATURAL AND FISHING
MORTALITY OF COMMON SNOOK IN SARASOTA BAY, FLORIDA
By

Jason P. Bennett

December 2006

Chair: William E. Pine, III
Major Department: Fisheries and Aquatic Sciences

The common snook Centropomus undecimalis is a popular, saltwater gamefish

found in southern Florida that has been actively managed by Florida's natural resource

management agencies to prevent overexploitation since the 1950's. Despite increasingly

restrictive management regulations, the status of the population and the effectiveness of

these regulations remain uncertain. Most fisheries management activities are focused on

regulating fishing mortality (F). Because of this, an important aspect of population

assessments is an accurate estimate of F to provide insight into the magnitude of fishing

mortality relative to natural mortality (M). Here, telemetry methods for relocating radio-

tagged fish were used to estimate total mortality (Z), F, and M for adult common snook in

Sarasota Bay, Florida. After tagging, fish were relocated using a series of remote,

autonomous receivers in conjunction with active tracking efforts. These relocations were

evaluated using a suite of apriori assumptions to determine whether an animal was live

or dead. These fates were then converted to mortality rates using known-fate type









models in program SURVIV. Three different modeling approaches were used to assess

uncertainty in the mortality estimates to assigning fates to 17 fish that were not relocated

after a large harmful algal bloom in Sarasota Bay during summer 2005. For the period

from October 2004 through December 2005 Z values (0.68 1.08), F values (0.24 0.66),

and M values (0.12 0.65) were estimated using different approaches which depended on

how fish fates were assigned. However, estimated parameter values were similar to more

traditional stock assessment mortality estimation methods and provided insight into the

performance of the mortality estimation methods currently used in the assessment. Using

telemetry methods to relocate fish and estimate mortality rates has advantages over

traditional assessment methods, but each method has assumptions and potential biases

that must be acknowledged. This study served as a good example of estimating fishing

and natural mortality for a Florida fishery as a complement to current stock assessment

methods.














CHAPTER 1
INTRODUCTION

The common snook (Centropomus undecimalis, "snook") is a popular gamefish

species found throughout south Florida. Since the 1950's, snook have been actively

managed by the state to enhance recreational fishing and to protect the species from being

over-fished. The management plan for this species has evolved drastically over the last

50 years and currently consists of bag, size, and seasonal limits all of which are designed

to reduce fishing mortality (F) to a sustainable level (Muller and Taylor 2006). However,

the effectiveness of these regulations for attaining this objective remains unclear.

Stock assessments of snook and other commercially and recreationally important

fish species in Florida are conducted at regular intervals to monitor population trends and

fishing mortality (see for example Murphy 2003; Murphy 2005; Mahmoudi 2005;

Munyandorero et al. 2006; Muller and Taylor 2006). An important aspect of these

assessments is an accurate estimation of total mortality (Z), which for management

purposes, is often then broken into its component parts: fishing mortality (F) and natural

mortality (M). Most assessment approaches generally use virtual-population-analysis

(VPA) type methods or traditional angler-dependent tagging studies to estimate F (e.g.,

Mahmoudi 2005; Munyandorero 2006; Muller and Taylor 2006). However, each of these

methods relies on a diverse set of data needs and often complicated assumptions to

indirectly estimate fishing mortality rates. Additionally, neither of these methods can

directly estimate M (both methods often use assumed M values) as M is rarely observed

or estimable from fishery-dependent data.









Recent advances in telemetry technologies have provided fisheries researchers with

new tools to investigate the ecology of aquatic animals. While using telemetered animals

to evaluate the movements and habitat use of fish species is not new, until recently

technological restrictions have made it difficult to use this technique to estimate vital

rates in fish populations. Wildlife biologists have a history of using telemetry tags to

document harvest or natural mortalities of birds and mammals (White and Garrott 1990).

In wildlife applications the tagged animal can often be relocated and its fate (live or dead)

visually observed yet these methods have not been widely used in fisheries applications

because mortalities of aquatic animals are rarely observed.

The purpose of this study was to use telemetry methods to estimate fishing and

natural mortality for adult common snook in Sarasota Bay, Florida. The "fates" of

individual fish were hypothesized through a combination of active tracking and

observations from an array of remote, autonomous receivers. Each time a tagged snook

was relocated, the live/dead status of that fish was evaluated using a suite of a priori rules

to help determine whether an animal was alive, emigrated from Sarasota Bay, harvested,

or died from natural causes. These fates were then converted to vital rates using known-

fate type models developed by Hightower et al. (2001). These mortality estimates are

critical parameters for assessing the effectiveness of current harvest regulations used to

manage common snook populations in Florida (Muller and Taylor 2006).

Study Species

The common snook is a large, predatory fish native to rivers, estuaries, and inshore

tropical and subtropical waters of the western Atlantic Ocean, the Gulf of Mexico, and

Caribbean Sea from the United States to Brazil (Gilmore et al. 1983; Rivas 1986,









McMicheal et. al. 1989; Muller and Taylor 2006). Snook are cold sensitive with low

lethal temperature limits from 6-130C (43-550F) (Marshall 1958; Howells et al. 1990).

In Florida, snook are abundant along the Gulf of Mexico coast from Tampa Bay

south to the Florida Keys and north to Cape Canaveral along the Atlantic coast (Muller

and Taylor 2006). The spawning season ranges from spring to summer along both

Florida coasts, generally beginning when water temperatures reach 250C. Spawning

takes place during afternoons and evenings in large aggregations in or near passes that

provide water exchange with the open ocean (Taylor et al. 1998).

Snook use a variety of habitats throughout their life history. Two to three weeks

following hatching, snook larvae can be found along mangrove roots and grass edges

(Peters et al. 1998). Juvenile snook tend to inhabit shallow, warm channels and creeks

with few macrophytes and slow moving water (Fore and Schmidt 1973). Juvenile and

adult snook are most commonly found in areas associated with mangroves (Marshall

1958; Gilmore et al. 1983) but can also be found in rivers, lakes, salt marshes, reefs, and

along beaches (Muller and Taylor 2006). During cool winter months, snook seek thermal

refuges, usually in creeks and canals, to avoid lethal low temperatures.

Life History

Snook in Florida demonstrate unique life history attributes by location (Atlantic

Ocean or Gulf of Mexico) and also by sex. Snook from each Florida coast have been

shown to be genetically isolated; snook from Florida Bay and the Florida Keys are closer

to Atlantic phenotypes, whereas the Gulf coast population demonstrates a unique

phenotype (Tringali and Bert 1996). Snook growth rates differ between the coasts,









leading to different management regulations by the Fish and Wildlife Research Institute

(FWRI) for each coast (Taylor et al. 1993; Tringali and Bert 1996).

Snook can live to be about 20 years of age with a maximum observed age of 21

years (Taylor et al. 2000). Maximum size is about 1,100 mm TL. The Florida state

record weighed 20.04 kilograms and the world record (from Costa Rica) weighed 24.32

kilograms (International Game Fish Association). On average, snook on the Gulf coast

grow at a slower rate and reach a smaller maximum size than those on the Atlantic coast

(see below, Taylor et al. 2000). Snook growth patterns are described by a von

Bertalanffy growth model, with each coast having its own unique growth curve. Growth

equations combined for each sex are (Taylor et al. 2000):

Atlantic fork length (mm) = 989.3(1 e 0235(age- 0976)) (1)

Gulf fork length (mm) = 947.3(1 e0 175(age+1 352)) (2)

Snook are protandrous hermaphrodites, being born and maturing as males and later

changing into females. This change results in different growth rates for the sexes, with

females usually growing larger than males. Males can be sexually mature by age-1, and

sexual transition occurs from ages 1-7 (240 824mm) resulting in an uneven sex-at-age

ratio. The majority of small snook (under 300-mm) are males, and most large snook

(greater than 700-mm) are females (Thue et al. 1982; Taylor et al. 2000). Due to this

sexually dimorphic growth, fish that are generally targeted and harvested are most likely

females.

Stock Status in Florida

Commercial harvest of snook was prohibited in 1957 when the species was

declared a gamefish by the state of Florida (Muller and Taylor 2006), and this species has









been actively managed for more than 50 years by the State (Table 1-1). In 2004 (last year

data are available), snook were the fourth most recreationally sought after fish on the

Gulf coast and fifth on the Atlantic coast based on angler preferences from the National

Marine Fisheries Service's Marine Recreational Fisheries Statistics Survey interviews

(Muller and Taylor 2006). Total catch (harvest and release) has increased on the Gulf

coast to 2.1 (1.5-2.7) million caught in 2004, but peaked on the Atlantic coast in 1995

with an estimated 737,000 (483,000-1,002,000) snook caught (Figure 1-1). Over 90% of

snook caught are released alive and 2.13% of snook caught and released are estimated to

die post release (Taylor et al. 2001). In 2004, total harvest of snook, including estimated

non-harvest fishing deaths, was estimated to be 112,000 (80,000-146,000) on the Gulf

coast and 40,000 (16,000-68,000) fish on the Atlantic coast (Figure 1-2), with 60% of

Gulf harvest occurring from March to May. Recreational effort on the Atlantic coast

increased until 1995 and has remained relatively constant since then. Gulf coast snook

angler effort increased six-fold from 1982 to 2001 and has continued to increase since. In

2004-2005, thirty-three percent (217,000) of all Florida resident fishing license holders

also purchased an optional endorsement "snook stamp" that is required only if the angler

intends to harvest a snook (Muller and Taylor 2006). Directed effort for common snook

is large and, at least on the Gulf coast, growing; as is evident from increases in angler

effort (Muller and Taylor 2006).

The snook fishery is one of the most heavily managed recreational fisheries in

Florida. Bag limits, size limits, allowable gears, and seasonal closures for snook were

first enacted by the Florida legislature. The Marine Fisheries Commission was created in

1983 and addressed snook in 1985 with a harvest slot of 609-mm to 863-mm TL and one









fish in possession allowed over the slot. In 1994, the winter closure was shortened by 10

days by starting on December 15 but opening the month of February, and the Florida Fish

and Wildlife Conservation Commission (FWC) implemented a snook management plan

to maintain snook stock size at or above 40% spawning potential ratio (SPR). In 1999, a

slot length limit of 660-863mm was implemented and, in 2002 on the Gulf coast, the

daily bag limit was reduced from two fish to one and May was added to the closed

season.

Although these regulations are more restrictive than for most other recreational fish

species in Florida, the 40% SPR goal established in 1994 has yet to be met. The goals of

these regulations are to reduce F and to protect the most fecund females from fishing

mortality (Muller and Taylor 2006). Closed harvest seasons are used to eliminate harvest

when snook are highly concentrated and could be easily exploited; i.e. when animals are

aggregated to spawn or are in thermal refuges. Closed seasons regulate harvest only, not

effort, thus snook are targeted year-round by anglers, but can only be kept during

relatively short harvest seasons. Directed snook fishing continues during the closed

seasons, and the magnitude of mortality related to angling during closed seasons is still

being investigated.

Project Justification

Fishery Dependent Methods to Estimate F and M

Estimates of total mortality and its component parts (F and M) are key parameters

used in stock assessments to manage commercial and recreational fish stocks. Many

tools used by fishery managers (i.e.; harvest restrictions) are based on regulating fishing

mortality, yet this parameter is difficult to accurately estimate. While harvest can often

be observed, natural mortality is rarely observed, making the estimation of this parameter









difficult. Total instantaneous mortality is commonly estimated using a catch-curve by

regressing the natural log of the numbers of fish at age (that are fully recruited to the

fishery) vs. age class. This approach estimates total mortality, but does not directly

provide information on fishing or natural mortality. A similar method is a non-

regression-based total mortality estimation using the Chapman-Robson method. This

method provides an estimate of survival (S) using the equation:

N
Zx
S = -I (3)
NZx
N + -X, -1
1=1

Where X, = the number of years the ith fish is older than the age at full recruitment; and

N = the total number of fully recruited fish (Chapman and Robson 1960).

These methods have a suite of assumptions that can be difficult to meet, and if

violated can decrease the accuracy of Z (Ricker 1975; Van Den Avylel993; Murphy

1997; Haddon 2001). These assumptions include:

1. Constant recruitment for all years.

2. All ages have been exposed to the same history of fishing mortality (constant
fishing mortality across all represented years).

3. Random sampling of the population.

Little is known about the recruitment variation of snook in Sarasota Bay, so any

mortality values obtained from this method warrant concern. Additionally, the current

slot limit functionally leads to violations of the second assumption because length is

poorly correlated to age for snook and fish from ages 3 to 18 can be in the slot limit.

However, this assumption violation is not nearly as bad as violation of the assumption of

constant inter-year natural mortality and recruitment. For these reasons a catch curve was









included in the 2005 snook stock assessment to serve simply as an approximation for

comparison to other results (Muller and Taylor 2006).

In most age-structured stock assessments (spotted seatrout, Murphy 2003; red

drum, Murphy 2005; mullet, Mahmoudi 2005; sheepshead, Munyandorero 2006; snook,

Muller and Taylor 2006), F is primarily estimated by using fishery-dependent methods

based on landings data and size-to-age estimates (i.e., Virtual Population Analysis [VPA]

or Statistical Catch At Age [CAA] methods), passive tagging methods (e.g., FLOY tag

programs), or more recently through fishery-independent methods. VPA methods require

catch and age data and incorporate changes in vulnerability and F with age. VPA

methods require assumptions of M and a terminal age, both of which can be difficult to

estimate. The estimates ofF for the population in both VPA and CAA are most precise

for year classes that have already moved through the fishery (reached terminal age).

When using an untuned VPA, estimates ofF for age classes currently in the fishery are

generally not as accurate as for ages that have moved through the fishery (Walters and

Martell 2004). This limits the utility of catch-based methods used to assess trends in

fishing mortality rates over the relatively short time intervals in which regulatory

decisions are often made by many management agencies, particularly for long-lived

species.

Another method to estimate F is to create a known population of fish by collecting,

marking, and releasing tagged fish. After correcting for tagging mortality, tag loss, and

angler reporting rate, an estimate ofF can be obtained from the ratio of the number of

tags returned to the number of fish tagged. Tagging mortality and tag loss estimates can

be obtained through simple experiments, however, angler reporting rate can be quite









difficult to estimate and can cause large biases in the estimate of F (Pollock et al. 2001;

Pollock et. al. 2002). On small or closed systems this method is often accompanied by

creel surveys, which aid in determining angler reporting (Van Den Avyle 1993; Ney

1993). Other approaches to estimating reporting rates include high-reward tagging

programs or the use of surreptitiously planted tags (see review by Pollock et al. 2001).

Estimating M is difficult because natural mortality events are rarely observed and

difficult to measure in aquatic systems (Quinn and Deriso 1999). If an estimate ofF is

available, M can be indirectly obtained from change in abundance at age (i.e., catch

curve) by subtracting F from Z (Van Den Avyle 1993). Natural mortality is most

commonly estimated by relating M to some other (often estimated) parameter related to a

fish's life history. For example, k from a Von Bertalanffy growth equation is thought to

approximate M, where M = 1.5*k (Jensen 1996). However, estimators that use life

history parameters can vary greatly throughout an animal's life (Beverton and Holt

1959). Other methods to estimate M from life history information are reviewed in Table

1-2.

The instantaneous natural mortality rates used by FWRI in the 2005 assessment

were assumed to be the same as those used in previous snook assessments: M = 0.20 on

the Atlantic and 0.25 on the Gulf coast. Natural mortality estimates differ between the

coasts because Gulf-coast snook are thought to be more likely to die due to cold and

lethal red-tide blooms than Atlantic snook (Muller and Taylor 2006). The natural

mortality rates are assumed to be constant across years in the FWRI assessment.

Fishery Independent Methods to Estimate F and M

Acoustic telemetry is a powerful tool that can provide direct observations and

estimates of F and M without many of the problems and assumptions of the methods









mentioned above (Pine et al. 2003; Walters and Martell 2004). In this method, a group of

animals are tagged with long-life telemetry tags and monitored for an extended period of

time (months or years depending on battery life and tag size). The fates of relocated

tagged animals can then be used to directly estimate F and M for the current study year

without relying on anglers returning tags (Hightower et al. 2001). The assumptions from

Hightower et al. (2001) of using a telemetry study to estimate mortality rates are:

1. All tagged snook present in the study site are alive or dead due to natural or post -
hooking release causes (these causes cannot be distinguished).

2. Each snook alive in the study site has the same probability of surviving to the next
tracking event.

3. All tags have been retained and are working properly.

4. Each snook is behaving independently with regard to emigration and harvest.

5. Snook not present in the study site have been harvested or have emigrated from the
study site.

6. Snook repeatedly located in the same location died due to non harvest mortality
(i.e., hooking or natural mortality).

Because of these assumptions, this method works best for spatially closed systems

where the probability of tracking the fate of an animal is high. Open systems must be

made "closed" by monitoring each exit point, which facilitates the monitoring of fish

emigration from the study location.

Parameter Estimation

Estimates ofF and M can be obtained by using the methods outlined by Hightower

et al. (2001). This method is similar to a known-fate Kaplan-Meier model modified by

Pollock et al. (1995) that allows for relocations of both live and dead animals. Most

known-fate models assume a relocation probability of 1.0 (Williams et al. 2001). The

combined model from Pollock et al. (1995) allows for relocations of both live and dead









animals and also allows for relocation probabilities less than 1.0. The Hightower et al.

(2001) model allows for distinct estimates ofF and M by modeling only relocated fish for

each individual occasion (i.e., month), and not for all fish potentially in the system. This

is necessary because not all snook present in the study site will be located on each

tracking sweep or by stationary receivers. Because only relocated fish are modeled, each

relocation acts as a "release" of that animal back into the system. For example, a fish

located in June on multiple receivers is known to be alive and within the system until it is

relocated again at a future date either in the study array or until the animal is known to

have emigrated. If the fish is not located again after June it is assumed to have been

harvested if it was not detected leaving the system.







12





Gulf of Mexico
2,500,000 -
---Total Snook Catch

2,000,000 -
0
0
C
c0 1,500,000


_0 1,000,000 -
E
z
500,000 -


0
--- -IL -I -I -i --I N)J N)
(D (D (D (D (D (D D (D O O
CO cO i t .O .- CO cO i tD O CO


Atlantic
1,600,000
Total Snook Catch
1,400,000 .... Released

1,200,000
O
O
c 1,000,000
)'
800,000

E 600,000
z 400,000

200,000

-0 -1 7 1 1 -1 -1 -1 -1 -1 1 1, 1 1
(D (D (D (D (D (D (D (D (D (D O O
OC O0 0 0 (0 C. 0 (.0 (. (D O O
w 03 O"n "-4 (O --A 03 w n "-4 O --- 03 w B


Figure 1-1. Total snook catch and release for the Gulf of Mexico and the Atlantic from
1981-2004 (Muller and Taylor 2006).







13


Gulf of Mexico
140,000 -
---Snook Harvested
120,000 Post-release Deaths
Total Fishing Deaths ,
0 100,000 IF ,


0
D 60,000 -
E
z 40,000 -





tO tO tO tO tO tO tO tO t.O O O
O0 O0 O0 O0 O0 ,. tD t.O t.O O O


Atlantic
120,000 -
-- Snook Harvested
- Post-release Deaths
100,000 -- Total Fishing Deaths

8 80,000 '


S60,000 \ \

S40,000 /0 00 ( 0

20,000 \ /


-- -i -Ps -re -Da -h -s -i -i )
(, (0 (, (0 (0 (0 (0 (0 (0 O O
O0 O0 O0 C0 C0 (0,(. (0 (0 O O
-- CO U "-4 (0 -- CO 0 "-4 (0 -- CO B


Figure 1-2. Harvest, post-release fishing deaths, and total snook deaths due to fishing
related mortality for the Gulf of Mexico and the Atlantic from 1981-2004
(Muller and Taylor 2006).









Table 1-1. Review of snook regulations in the state of Florida.
Year Regulation
1953 Minimum size set at 457mm (18 inches) FL
1957 Snook made illegal to buy or sell; Bag limit set at four snook > 457mm (18
inches) FL
1981 Bag limit reduced to two snook per day.
No snook < 660mm (26 inches) FL maybe taken in June or July during 1982 -
1986
1983 January, February, June, and July 1983 1986 closed to snook possession
1985 January, Feb, June, and July closed permanently
August 1985-1986 closed
Minimum size increased to 609mm (24 inches) TL
Only one snook may be > 863mm (34 inches) TL
1987 All species of Centropomus covered by the regulations
August is closed permanently
Use of treble hooks prohibited with natural baits
1994 Winter closed during 16 December January 31
SPR goal set at 40%
1997 Population separated into Atlantic and Gulf Stocks
1999 Harvest slot set at 660mm (26 inches) to 863mm (34 inches) TL
2002 Gulf stock: closed during May and daily bag reduced to one snook from two
(Muller and Taylor 2002)


Table 1-2. Alternative methods for estimating natural mortality based on life history
attributes.
Equations for estimating M Citation
logo0 (M) = -0.0066 0.279 (logo0 (L.)) Pauly
+ 0.643 (log0l (k)) + 0.4643 (log0l (average annual water temperature)) (1980)

Hoenig
In(M) = 1.44 0.982 x ln(tmax) :where tmax equals the maximum (1983)
observed age in the un-fished stock


M = 1.92Wa-025 :where Wa equals the weight at age a Peterson
and
Wroblewski
(1984)














CHAPTER 2
METHODS

Field Methods

Study Site

This study occurred in Sarasota Bay, Florida. Sarasota Bay is oriented along a

north-south axis and is traversed along this axis by the Intercoastal Waterway (ICW),

which enters to the north at the Cortez Bridge near Cortez, Florida and exits to the south

at the Albee Point bridge near the Venice inlet in Venice, Florida. The bay covers an

area of 13,467.9 hectares and a linear distance of 43.45 kilometers, and includes multiple

sub-bays each with unique flow and habitat characteristics: Hudson Bayou, Blackburn

Bay, and Little Sarasota Bay. The site also contains three passes between the bay and the

Gulf of Mexico (Longboat Pass, New Pass, and Big Pass) as well as five major tributary

creeks (Bowlees Creek, Whitaker Creek, Phillipi Creek, North Creek, and South Creek)

(Figure 2-1).

Tagging Procedure

Beginning in October of 2004, adult common snook were collected in Sarasota Bay

using hook and line, seines, and trammel nets. Trammel netting by actively searching for

legal snook and then setting the net on these fish was the most successful method (Table

2-1). All snook collected were measured (TL), and each fish selected for surgery was

anesthetized with sodium bicarbonate, and a Vemco acoustic telemetry tag implanted

into its abdominal cavity. Two different types of tags were used; 65 V-16 tags (16-mm

diameter, 70-mm length, 700-day battery life) and 10 V-8 tags (8-mm diameter, 40-mm









length, 300-day battery life; Vemco Ltd., Shad Bay, Nova Scotia, Canada). Sixty-three

of the 75 tags were implanted in snook within the legal size range; five tags were

implanted in snook over legal size and seven in snook just under legal size. Three angler

returned tags were later reused and implanted in newly caught snook, bringing the sample

size of legally harvestable tagged fish to 66.

Abdominal incisions were closed with 3-4 absorbable sutures and ethyl

cyanoacrylate. Each snook was observed until fully recovered and then released into the

same area from which it was collected. No external tag or mark was added or visible

(except for the surgery scar and temporary sutures) to avoid influencing angler behavior.

The goal of this project was to generate fishery independent estimates ofF and M,

therefore any tag or mark that could change the probability of harvest was unwanted.

Tag and Receiver Description

Each acoustic tag outputs a uniquely identifiable signal, pseudo-randomly between

30 to 90 seconds (mean = 60 seconds). The Vemco VR2 acoustic receivers (Vemco

Ltd.) recorded the unique tag number along with the date and time of reception. Data

from all receivers were manually retrieved on one to three month intervals during which

time the receivers and anchor materials were checked and cleaned to assure proper

performance. Receiver battery changes occurred on yearly intervals to prevent loss of

acoustic coverage and data.

Acoustic Array Network

Array deployment and testing

The use of remote receivers requires that snook move within the relatively small

detection range of the receiver in order to be detected. To increase the probability of tags

being detected, receivers were positioned in areas likely to be frequented by snook, or in









areas that offered physical boundaries, such as a narrow area of the bay (Klimley et al.

1998; Lacroix and Voegeli 2000; Clements et al. 2005, Heupel et al. 2006). Each of the

five exits from the study site were acoustically "closed" with 1-3 fully automated

stationary VR2 receivers depending on the width, depth, and ambient noise of each site.

Each of the three bay passes and the two ICW exits from the study site (Cortez in the

north and Venice in the south, Figure 2-1) were range tested with temporarily placed

VR2 receivers and active tags to ensure that number and placement of receivers provided

complete acoustical spatial coverage of the pass or exit. Because these sections of the

bay differ in shape, depth, and area; each pass/exit required different numbers of

receivers and receiver placements. Receivers were placed in potential locations while

tags were drifted through the pass at several locations in an attempt to mimic possible

routes a snook could take through the pass. Data from each VR2 receiver was then

downloaded and compared to the actual times the tags were drifted through the pass to

evaluate receiver performance. This diagnostic procedure was repeated until receivers

were placed such that a test tag drifting through the pass was detected at least twice

during the time the tag was passing through the detection field for complete acoustic

coverage of each pass or exit (Clements et al. 2005). Each receiver was then anchored on

the bottom with either a large concrete anchor or attached to an existing piling or

navigational structure.

Receivers were deployed in the creeks to monitor tagged snook movements in and

out of the creek systems. Because the creek depths were strongly dependent on tides, the

receivers were generally placed in deeper holes near the creek mouths such that the

receivers would remain submerged even at low tide. Additional receivers were located









upstream throughout the individual creeks as part of a separate telemetry study evaluating

juvenile snook movement patterns.

Manual tracking

Manual tracking using a Vemco VR60 or VR100 acoustic receiver by boat was

used to search for fish in areas not covered by the VR2 receivers. Manual tracking was

key for trying to locate specific fish that may not have been recently detected on the

receiver array (by searching near their last known location) as well as locating tags from

animals that died. Manual tracking also aided in identifying whether a fish had emigrated

from the study site by allowing for checking along the beaches outside of Sarasota Bay

and within the passes in areas that extended beyond the detection area of the VR2

receivers. Manual tracking was conducted approximately monthly during the summer and

every 6-8 weeks at other times of the year. From field testing, it was determined that a

conservative detection distance using the manual receiver was approximately 300 m. The

manual tracking protocol then consisted of tracking the shoreline of Sarasota Bay and

stopping every 300 m for 3 min to listen for tagged snook. Cross-bay transects were also

conducted when possible.

Data Analysis

Assigning Fish Fates

For the purposes of this thesis only the legally harvestable fish were used to

estimate F and M (Hightower et. al 2001). All fish used in the analysis were detected at

least once after tagging before being included in the study to reduce the chance of tagging

and surgery effects influencing the study results.

Relocations from the VR2 and VR100 receivers were used to assign fates to each

fish in the study based on the previously mentioned assumptions from Hightower et. al









(2001). Fish fates were defined as: within site and heard alive, within site but not heard

(i.e. up creek beyond detection range of receivers), within site and heard dead (i.e. tag

lying on bottom), or emigrated from site. These fate assignments were based on when

and where each fish was last located via active tracking and/or on the array. Fate

assignments were made by two people reviewing the relocations and jointly determining

the fate assigned to that animal for each month. The individual fates of the fish were then

used to construct a monthly binary "capture" history of each fish; with months of live, in

site relocation receiving a "1" and months with other fates assigned a "0". Fish

determined to be emigrants were censored from the models. These fish could be added

back into the models later if the fish reentered the site. Program MARK (White and

Burnham 1999) was used to create the data summary matrix ("Yij" matrix, Burnham et al.

1987) to use in the data analysis.

Fishing and natural mortality rates were estimated using program SURVIV (White

1983) based on general models and procedures described by Hightower et al (2001).

With this model, for R snook released at time t, the number of fish still alive at time t + 1

(St+ ) is expressed as:

S, = Re(-F Mtp t+ (4)

Here Ft and Mt are instantaneous mortality rates during period t, andpt+l is the relocation

probability at time t+l. Survivors to time t+1 were considered new releases for the next

period. The number of fish still alive at time t+2 (St+2), but not seen at time t+ 1, is

expressed as:


st2 =R~e(F mt (I)(- p, )e-F~ 1 ~lv )(P+ 2)









Capture probabilities (p ) and values ofF and M are made on the time scale of the

data input; in this case, monthly estimation. The precision of the parameter estimates are

driven by the magnitude of the event (data "contrast", Hilborn and Walters 1992). Sparse

data, such as infrequently observed mortality events, result in limited contrast; thus, some

models were constructed at longer time scales ("fishing season" or annually) to allow for

monthly observations to be pooled (Hightower et al. 2001, Heupel and Simpfendorfer

2002).

A suite of biologically reasonable models were used to estimate P, F, and M. The

models used different time intervals for each parameter- fixed models (denoted by a "X."

subscript), time (monthly) dependent models ("X,"), and seasonal models ("X, ") that

corresponded to the open and closed harvest season for snook (X equaled p for capture

probability, F for fishing mortality, or M for natural mortality) (see Table 2-2 for all

model combinations). Akaike Information Criteria (AIC) values (Akaiki 1974; Anderson

et. al 1998) were used to provide guidance on model fit.

Since fate assignments were based on assumptions and subjective inference, it was

possible that fates could be mis-assigned (i.e. a natural mortality mis-classified as an

emigrant or vice versa). The data was first analyzed with the fate assignments based

strictly on the rules described in Hightower et al (2001). Two alternative data sets were

then developed which reassigned fates for a group of 17 fish that were last detected alive

during the summer of 2005 and had subjective fate assignments. These 17 fish were

originally designated as either fishing mortalities or emigrants depending on there last

known location in the receiver array (8 last heard within the array and 9 last heard on a

pass/exit). A fish that was last detected at a pass/exit receiver was classified as an









emigrant, while a fish last detected on one of the bay receivers was classified as a fishing

mortality. The classification of these fish was different than others because, of the 18

fish no longer heard beyond the summer of 2005, seventeen tags were not located and,

thus, could not confirm that these fish were indeed dead. These fish were suspected to be

natural mortalities because they were lost from the receiver array during a relatively short

time interval, the harvest season was closed, and there was an extremely large and intense

red tide event occurring at this time. Recovery of these tags was not possible as local

governmental agencies hired aquatic plant removal companies to use floating plant

harvesters and prison work crews to remove dead fish floating on the surface and

transport the fish to the landfill. To evaluate how possibly mis-assigning these fish

would change F and M estimates, these 17 fish were assigned alternative fates by

classifying them as all emigrants and fishing mortalities notatedd as lost fish as emigrants

and fishing mortalities), all as natural mortalities notatedd as lost fish as natural

mortalities) or an equal combination of the two notatedd as 50/50). In the "lost fish as

emigrants and fishing mortalities approach" fish that were last detected on a pass/exit

receiver were considered emigrants and were censored from the analysis; fish last heard

on receivers inside the site were considered fishing mortalities (46 fish were used in this

approach). Conversely, in the "lost fish as natural mortality approach" the 17 lost fish

were considered natural mortalities no matter what the location of their last detection (55

fish were used in this approach). A "50/50 approach" was used to examine the sensitivity

of the models to fate assignments by showing how the models reacted to more subtle

changes in fate assignments compared to the more extreme differences between the lost

fish as emigrants and fishing mortalities approach and the lost fish as natural mortalities









approach (50 fish were used in this approach). The differences in the number of fish used

for each approach were due to the assignment of emigration as emigrants were censored

from the analyses.

Total Mortality Estimates

The Hightower et al. (2001) approach was also used to estimate Z for the tagged

fish. In this method, no distinction was made between fish no longer heard within the site

and known natural mortalities. In this framework, the mortality components (F and M)

were pooled to generate a total mortality estimate. This approach eliminated some

subjectivity in assigning fates of fish and provided an estimate of total mortality that was

useful for comparing mortality trends across the study period.

Annual apparent survival, survival based on loss of detection, (apparent survival, P

= 1 mortality emigration) was also estimated using Cormack-Jolly-Seber (CJS)

methods (Cormack 1964; Jolly 1965; Seber 1965, Pollock et al. 1990) to compare with

the total mortality trends estimated with the other models. This method used only capture

histories for each fish and was not dependent on fates assigned to the tagged fish. The

model estimated two parameters, PD and p, which could be either fixed, time dependent,

or fishing season dependent. Table 2-3 contains a description of all models. The

assumptions of the CJS method are:

7. All previously tagged snook alive at the time of one census are equally likely to be
captured in that census.

8. Tagging does not increase the likelihood of mortality.

9. The probability of a snook surviving to the next census is independent of its age.

Ages of the fish were unknown, but only fish within the legal harvest slot were

used in this analysis. Survival estimates from CJS models are robust to permanent









emigration, but can be biased by temporary emigration when capture probability is low

(Kendall and Nichols 1995, Zehfuss et. al 1999). The apparent survival estimates were

adjusted to correct for five known permanent emigrants, but it was not necessary to

correct for temporary emigrants as the time step of the emigration was often less than the

time step of the estimate as small, sometimes days long, emigrations occurred with some

fish. Trends in apparent survival from this model should have mirrored mortality events

from the Hightower models (Pine et. al 2003).

Total mortality and emigration rates (E) were estimated for the CJS models with

the following equations:

Z= -log,(S) (6)

E = loge (1 emigration proportion) (7)

Emigration proportion was the proportion of permanent emigrants to total study fish. M

was estimated using the Hoenig equation outlined in Table 1-2. Using results from

equations 5 and 6, F was estimated where:

F=Z-M-E (8)

These estimates of F were then compared with F estimates from the Hightower models.

A potentially large source of natural mortality in marine fish populations can be

from harmful algal blooms ("red tides") commonly caused by large blooms of the

phytoplankton Karenia brevis. These dinoflagellates release a nonproteinaceous

endotoxin called brevitoxin (Steidinger and Haddad 1981). Concentrations of red tide

cells greater than 100,000-200,000 cells/liter can cause fish mortality (Sea Stats 2000).

In the summer of 2005, a spatially and temporally large red tide bloom occurred off of

southwestern Florida with highest concentrations occurring in Sarasota Bay. Concurrent









with this red tide, widespread mortalities of a variety of fish and marine mammal species

were recorded in southwest Florida. During July, large numbers of snook were observed

dead or dying around Sarasota Bay, particularly in the New Pass area and bay and ocean

shorelines near Sarasota. Two hundred eighty-six adult snook of various sizes were

collected during this mortality event for size and age structure sampling. All snook

collected were measured, their otoliths removed, and the fish were checked for acoustic

tags. The otoliths were sent to FWRI for age analysis and the resulting age estimates

were used in the catch-curve analysis. This estimate of Z was independent of the

mortality estimates from the tagging data.


Figure 2-1. Map of Sarasota Bay with the study site borders of Cortez to the Northwest
and Venice to the Southeast (large arrows). Passes leaving the study site are
indicated by small arrows while creeks are indicated by stars.









Table 2-1. Catch of study snook per gear type.
Gear type Number of study snook caught
Hook and Line 6
Seine net 10
Trammel net 62


Table 2-2. Descriptions of the five models used to estimate capture probabilities, F and M
in all three different scenarios. Each occasion for which an estimate was
generated was a month.
Model Number of Description
Parameters
pt Ft Mt 45 Time dependent capture probabilities, F and M


kF. M.

kF, M.

Pt FM.

Pt FMt


3 Fixed capture probabilities, F and M

4 Fixed capture probabilities and M, and F fixed by open and
closed harvest seasons
18 Time dependent capture probabilities, fixed M, and F fixed by
open and closed harvest seasons
32 Time dependent capture probabilities and M, and F fixed by
open and closed harvest seasons


Table 2-3. Descriptions of the five models used to estimate Apparent Survival (P) and
capture probability (p). Each occasion for which an estimate was generated
was a month.
Model Number of Description
Parameters
Q(t 16 Time dependent apparent survival and fixed capture probabilities
(t Jt 30 Time dependent apparent survival and capture probabilities

(D. t 16 Fixed apparent survival and time dependent capture probabilities
(Ds t 17 Apparent survival fixed by open and closed harvest seasons and
time dependent capture probabilities
(D. 2 2 Fixed apparent survival and capture probabilities














CHAPTER 3
RESULTS

Field Results

Tag Reception

Table 3-1 lists the number of VR2 receivers used in all passes and creeks.

Receivers were placed in the passes/exits such that there was no possible route a snook

could take through the pass/exit without being detected. During range tests of receiver

placement performance, tags were heard at least twice on every test drift through the

pass/exit, implying total acoustic coverage. Reception rate was calculated as the actual

number of detections divided by the number of times, on average, a tag could have been

detected while the tag was in the receivers estimated detection range. On average total

tag detections were about 50%, such that if, in theory, a tag should have been detected 10

times during a drift through the pass, the tag was detected at least 5 times.

Fish Detection

A detailed detection history for all fish used in the analysis is provided in Figure 3-

1. Four fish were never detected post release and were excluded from the analysis.

These fish may have succumbed to tagging mortality, immediate harvest, or tag failure.

Because the fates of these four fish could not be determined, they were censured from the

analysis. Tag failure was unlikely as it was evaluated by examining 10 recovered tags

(nine from anglers and one found in a dead snook); all were working when returned and

later implanted again into snook. The VR2 receivers could occasionally interpret an

ambient noise (i.e., from a depth sounder) as a single pulse from an acoustic tag.









Following the guidelines in Clements et. al (2005), multiple receptions (more than one

reception pulse over several minutes) were required before a receiver reception was

considered a true detection. Nearly all other study fish (72 of 78) were heard on at least

one occasion on a stationary receiver.

Detection rates approached 100% (p = 1.0) early in the study as water temperatures

declined towards lethal limit for snook (Figure 3-2). As water temperatures declined,

snook entered their thermal refuges within the tidal creeks, passing the creek receivers on

their way to these refuges, thus increasing detection rates. When the water temperatures

were at their lowest levels, snook movement rates greatly declined and snook likely

remained stationary within their thermal refuges, thereby causing detection rates to drop.

As water temperatures increased in the spring, detection rates also increased as the snook

left the creeks (detected again on the creek receivers) and entered the bay. During this

time, receptions on the intra-bay and pass receivers increased as the snook resumed their

spring and summer movement patterns.

A sharp decline in detections was observed during the summer of 2005 (Figure 3-

2), during which a major red tide bloom was occurring (Figure 3-3). Eighteen study fish

were not detected beyond this point of time. One of these fish was an observed natural

mortality (dead fish was recovered), but the fates of the other 17 were unknown. The last

known locations for 8 of these fish were from bay and exit receivers, but 9 were last

heard in locations inside the bay. During the fall and winter of 2005/2006, as water

temperatures again declined, detection rates continued to decrease and never again

reached the levels from the previous year (Figure 3-2).









Two fish were detected only through manual tracking. Manual tracking also

located five tags that were classified as natural mortalities based on repeated relocations

in a single location. Recovery of these tags was attempted, but was unsuccessful.

However, it was believed that if these fish had been alive and simply stationary, attempts

to recover the tags would likely have caused the fish to swim away from the area.

Data Analysis Results

Assigning Fates

Fates of all 66 legally harvestable fish used in this analysis are presented in Table

3-2. Fish whose tags were returned were known to be fishing mortalities (they were, by

default, counted toward F by the model due to their disappearance from within the

system). Fates of fish that died of natural mortality were more difficult to classify.

During the red tide bloom only one tag was recovered from a dead snook that was part of

the age sample collection. More fish were assumed to have died, but it was thought that

their carcasses were removed by local cleanup crews before they could be checked for

tags. Concurrent with this large fish kill was an extreme drop in tag detections (July

2005, Figure 3-2); several fish that had commonly been detected were not detected again

on the array or via manual tracking. Because of the events which occurred at this time it

was necessary to consider these fish using alternative fate assignment approaches. These

fates were used to examine how the model results would differ if these fish were treated

as if they had been illegally harvested, died of natural mortality (even though tags were

never relocated), or emigrated. The results from modeling approaches with the "lost"

fish with different fates assigned during the red tide bloom are presented below.

For each of the three approaches (the lost 17 fish as both emigrants and fishing

mortalities approach, lost fish as natural mortalities approach, and lost fish as 50/50









approach) the three best performing models, based on AIC values, were always p t Ft Mt,

p t Fs Mt, and p t Fs M.; AIC and yearly F and M values (summed monthly values) of all

models for all three approaches are provided in Table 3-3. Results from all models

performed in all three approaches are provided in Tables A-1, A-2, and A-3.

Model p t Ft Mt

This model was fully time dependant, with variable monthly capture probability, F

and M estimates. Capture probabilities were high, ranging from 0.7 to 1.0 for all three

approaches (Figure 3-4). As was expected, F and M estimates varied between the three

approaches. In the lost fish as emigrants and fishing mortalities approach, F peaked in

August 2005 (0.18 per month, SE = 0.09) and estimated a yearly F (monthly values

summed) of 0.66 per year. The natural mortality estimate was 0.16 per year. In the lost

fish as natural mortalities, F was lower at 0.41 per year. However, natural mortality was

much higher, peaking in August 2005 (0.35 per month, SE = 0.11) leading to a yearly M

of 0.65 per year. The lost fish as 50/50 approach had estimates in-between the other

approaches with F equal to 0.48 per year and M equal to 0.42 per year (Figure 3-5).

Model p t Fs Mt

This model had time dependant monthly capture probability and M estimates, but

the F estimates were fixed by open and closed harvest seasons. Capture probabilities

were again high, ranging from 0.7 to 1.0 for all three approaches (Figure 3-6). Estimates

ofF and M again varied with the major difference in F coming from the values for the

closed harvest seasons. In the lost fish as emigrants and fishing mortalities approach the

F for the open harvest season months equaled 0.05 per month (SE = 0.02), while the

estimates for the months of the closed harvest seasons dropped to 0.03 per month (SE =









0.01); leading to a yearly F value of 0.50. In the lost fish as natural mortalities approach,

F for the open harvest season months again equaled 0.05 per month (SE = 0.02),

however, the estimates for the months of the closed harvest seasons dropped to 0.00 per

month (SE = 0.00); leading to a yearly F value equal to 0.30. The F for the lost fish as

50/50 approach was again somewhere between the other approaches with a yearly F of

0.34 per year. Monthly natural mortality estimates were almost identical to the

previously described p t Ft Mt model for all approaches (Figure 3-7).

Model p t Fs M.

This model had time dependant monthly capture probability, but a fixed monthly M

and F fixed by open and closed harvest seasons. Capture probabilities were again high,

ranging from 0.7 to 1.0 for all three approaches (Figure 3-8). Estimates ofF and M again

varied, but monthly F estimates for all approaches were almost identical to the p t Fs Mt

previously described. The major difference in results for the different approaches for this

model came from the M estimates. The lost fish as emigrants and fishing mortalities

approach had monthly M values equal to 0.01 per month (SE = 0.01) for all months in the

study. This gave a yearly M equal to 0.12 per year. Conversely, the lost fish as natural

mortalities approach had monthly M values equal to 0.05 per month (SE = 0.01) for all

months in the study leading to a yearly M of 0.60 per year. The lost fish as 5/50

approach estimate was again centered within the approach estimates with a yearly M

equal to 0.36 (Figure 3-9).

Clearly there were differences between all the models for the three approaches

used. Figure 3-10 shows the differences between the monthly F and M estimates for all

three approaches for the p t Ft Mt (fully time dependent) model. The estimates for F









varied between the approaches, but the major source of variation between the three

approaches appears to be in the M estimation.

Secondary Analysis

Total mortality was modeled in two ways. The first method modeled Z in the same

way that F and M were modeled using methods from Hightower et. al (2001); with each

of the five models and the three different approaches of assigned fates. This method

resulted in a range of yearly Z estimates for 2005 of 0.68 1.08 per year. The second

method simply added together F and M estimates from the previously described

Hightower models as per the equation Z = F + M and resulted in a range of yearly 2005 Z

estimates of 0.63 1.06 per year (Table 3-4).

A catch curve was included in the 2005 Snook Stock Assessment to provide a

"rough estimate of the magnitude of total mortality" (Muller and Taylor 2006). The 286

dead adult snook collected during the red tide event were used to construct a comparable

catch-curve using a similar range of fish ages. The catch-curve was estimated using fish

ages 5 through 8 and 6 through 8 because of the 286 fish only four were age 9 or greater.

This generated a Z estimate of 0.37 per year when using ages 5 through 8 and 0.83 per

year when using ages 6 through 8 (Figure 3-11). Age-5 was used as it was included in

the catch curve for the 2005 snook stock assessment (Muller and Taylor 2006).

However, it is unlikely that age-5 fish are fully vulnerable.

For the Cormack-Jolly-Seber analysis, the t/p and tp pt models were virtually

identical in model fit AIC values of 698.58 and 699.22. AIC scores from all models fit

are provided in Table 3-5 and results in Table A-4. Capture probabilities were as high as

those from the Hightower models, ranging from 0.7 to 1.0. Model rt/p showed a strong









drop in survival in April (0.83, SE = 0.06) and August (0.69, SE = 0.08) of 2005. Those

are the same months that show high natural mortality in the Hightower models. Model

(It~P t showed a dip in capture probability during the cold months following the beginning

of the study similar to that shown by previously mentioned models. It also showed a

decline in survival during April (0.85, SE = 0.05) and August (0.70, SE = 0.08) of 2005

(Figure 3-12).

Survival estimates from the CJS models were used to estimate total mortality.

These rates were used along with estimates of natural mortality and emigration to

estimate fishing mortality (Eq. 8). Z estimates ranged from 0.43 to 0.56 per year giving a

range ofF estimates from 0.18 to 0.31 per year (Table 3-5).

These several methods gave variable estimates of mortality. Components of

mortality (Z, F, and M) were estimated using two different methods each. A review of

each component, method used, and the result are presented in Table 3-6.













Not heard
Heard alive
Known Fishing mortality
Natural mortality
Month before deployment
Possible tagging mortality


Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Jan


Figure 3-1. A detection history for every fish in the study. Each row is an individual fish
and each column is a month of the study.












100 35
90 -30
80 CD
o ,, 25 S
70 .
-60 \ 20." 20
2 50 '- .' *--
840 D15

30 % Detections per month -10-
20 ..- -Water Temp (C) 5
10 ----- Lethal Temp Limit
0 0
o z cc c- -n c> c c > O z o c-
F < 0 D ? -F < 0


Figure 3-2. Percent detections and mean water temperatures by months of the study.



-- Red tide cells/Liter

Observed Fish
Mortality
0 0








SZC- -n > C -- C- > Un O z 0
D ) 0 C- CD CD
0 c -- 1 ,- (C) -0 IF- < o
6 o o 0 o o 5 6 6 o o o

Figure 3-3. 2005 Red tide (K. brevis) cell counts for Sarasota Bay. The dashed line
indicates the cell count level that begins to cause mortality in some fish
species.













1.00

0.90

0.80

0.70

0.60

0.50


Emigrants and Fishing Mortalities Approach


z o c- -n S c-c- > c 0 z Co
o CD CD -0C CD 0) 0 CD
C C) C C 6 C Co
C 7 H C) C C C C)
0 0 0 CT5 C55 5 g C a


Natural Mortalities Approach


1.00

0.90

0.80

0.70

0.60

0.50





1.00

0.90

0.80

0.70


z C c- -n > c- c > 0 z> 0
0 (D (Di -9 0) ED o C CD
?< 7 1 S ( a -0 < o
oo o o o o < -" o o o o o
- I I I I 0"I C "I C5 C C
-P C" 0" B


50/50


0.60

0.50
z ) c- -m K > K c- c- > 0 O Z -




Figure 3-4. Model p t Ft Mt capture probabilities for all three modeling approaches. This
model had time dependent capture probabilities, F and M estimates. These
approaches differ only in the fates assigned to 17 fish last heard during the
summer of 2005.












Emigrant and Fishing Mortalities Approach
-F


0.50

0.40

0.30

0.20

0.10

0.00






0.50

0.40

a 0.30

0.20

0.10

0.00


Natural Mortalities Approach


--F
- M


zo 0 M K > K L- L- >co0 zo0
o0 CD M w -a w MD a 0 MD
0 Cr 7 7 1 -
CDiA ii 6 6 D DC
00 00 q Ln D CD n 6
Ln n n n n L L L L B


50/50 Approach


-F


0.40 iv

0.30

0.20 ,

0.10 -

0.00 -

,- p7 < = ( < ,0


Figure 3-5. Model P t Ft Mt F and M results for all three modeling approaches. This
model had time dependent capture probabilities, F and M estimates. These
approaches differ only in the fates assigned to 17 fish last heard during the
summer of 2005.


z o C- Mn E > E C-C-> wn 0 z o
O CD M ( D m O V C CD C) 0 D
? :P 7 -. :P -0 -'~ < 0
0)C I I
C) 0) 0 0 6 ~ 61 C 0 C
CP C-I 01I 0I C1 0I C) C1 01 01 01 0


0.50












1.00 -

0.90

0.80

0.70 -

0.60






1.00

0.90

0.80

0.70

0.60






1.00

0.90

0.80

0.70


Emigrants and Fishing Mortalities Approach


z C- M > K C > 0 z 0
o0 CD' WCW C: CD C' 0 CD
S0 0" 1 7 < (a -" t0 7"I < 0"
CD D CD D CD CD 6 CD CP 6 6 CD 6 6
.41> 4L> CrI cOl COl C71 CI C CP CP CP CP CP


Natural Mortalities Approach


z 0 C-- > C C- C > 0 z 0
o M W ,- W C C: C: o 0 M
< 9 ," 7, ,< ? -, (a _0 < 0
000 ? 6 6 o 0 n 6 o
C O'l C' 'l Enl C' l U'l *B


50/50 Approach


0 .6 0 ...... .
z O -a l > > 0 z 0
o CD=3aC C C CD C) 0 CD
< C 7 P ( -a < 0


Figure 3-6. Model p t Fs Mt capture probabilities for all three modeling approaches. This
model had time dependent capture probabilities, and M, with F estimates fixed
by open and closed harvest seasons. These approaches differ only in the fates
assigned to 17 fish last heard during the summer of 2005.











Emigrants and Fishing Mortalities Approach
0.50 -
F
0.40 M

0.30 -
Co
0.20

0.10 -

0.00
z oc -n K > K > o z o
0 M 7 7D 0 0 CD



Natural Mortalities Approach
0.50 -
-F
0.40 M

a, 0.30 *

0.20 *

0.10 -

0.00 ,--- > ."----,O,
z 0 C- M K > K Co C- > 0 0 z 0
z< o 7" 7D- '< 7 O
o CD ^CD Cro CD C7 0 C CD


50/50 Approach F
0.50
-M
0.40

S0.30 -

0.20 -I

0.10 -

0.00 -
z o -n s > Ecc o z o
0 CD CD C ( 0 (D
< 7 7< M -0 '" < 0o
= 4 On O On (n 5 n O1 6n D n 6

Figure 3-7. Model p t Fs Mt F and M for all three modeling approaches. This model had
time dependent capture probabilities, and M, with F estimates fixed by open
and closed harvest seasons. These approaches differ only in the fates assigned
to 17 fish last heard during the summer of 2005.












1.00

0.90

0.80

0.70

0.60






1.00

0.90

0.80

0.70

0.60


1.00


0.90

0.80

0.70


Emigrants and Fishing Mortalities Approach


o o 1- M K > > 5 0 z 0
0~ CD 01 M1 -0 C1 D 0 01 CD


Natural Mortalities Approach


z 0 MK>K>W0 0
0 CD CD -0 C CD 0 M
5 5 5 6 : cD, ) C
0nL, L, Li LI U 1 U P C


50/50 Approach


0 .6 0 ...... .
Z C- > C- C- > C) 0 Z 0
0 CD a CD a C C C C D 0 0 CD
o o o 6 o o o o o o o
-I Ol Ol (l Ol 0l l l C

Figure 3-8. Model P t Fs M. capture probabilities for all three modeling approaches. This
model had time dependent capture probabilities, fixed M estimates, with F
estimates fixed by open and closed harvest seasons. These approaches differ
only in the fates assigned to 17 fish last heard during the summer of 2005.












0.50

0.40


Emigrants and Fishing Mortalities Approach

-F


a 0.30

0.20

0.10

0.00
z o o- K t > K t > o 0 z 0
< C) -o o <
oo 0 o 0 o0 o o o A


Natural Mortalities Approach


0.50


-F


0.40 -

0.30

0.20

0.10

0.00 -
z 0 C- -n > C- c -c > W 0 z 0
0 CD CD C ( C D 0 0 CD
< ? 7 7 < 7 (a -0 '-" < o
6 o o o 6 oD 6 o o 6 6 o6
P 4 n 0 n 4C <.1 01 n 0Cn Cn Cn Cn


50/50 Approach


-F


S p i-- .3* -_j- -_ _^ -;- -^,
z o 0 E > E o > c o z o
O CD M CD M o m CC (D O O (D
< 0 7 : C) 7 7y CD I C)7 0 CD
2P I Cg CP CP CP C S g g C


Figure 3-9. Model P t Fs M. capture probabilities for all three modeling approaches.
This model had time dependent capture probabilities, fixed M estimates, with
F estimates fixed by open and closed harvest seasons. These approaches
differ only in the fates assigned to 17 fish last heard during the summer of
2005.


0.50


0.40

0.30

0.20


0.10

0.00






41




Fishing Mortality
0.50 Emigrants and Fishing Mortalities Approach
- Natural Mortalities Approach
0.40 ....... 50/50 Approach

a 0.30
(U
0.20

0.10 .

0.00
o CD (D a) -0 a) CD 0 0 CD
< o ,o- < z, o < < o



Natural Mortality
0.50
Emigrants and Fishing Mortalities Approach
- Natural Mortalities Approach
0.40 ....... 50/50 Approach

0.30 *

0.20 -

0.10 *

0.00 /o, /2
z 0 c- -n K > K c- c- > (l) O z 0
0C C C (1) C 0 0 C
oD 6 5 o p o I
B
Figure 3-10. Comparison of fishing mortality and natural mortality estimates using the
P t Ft Mt model (time dependent capture probabilities, F and M) for all three
fate assignment approaches. These approaches differ only in the fates assigned
to 17 fish last heard during the summer of 2005.












Catch Curve ages 5 through 8


0

S


y = -0.3709x + 5.8734
R2 = 0.4875


0 2 4 6 8 10 12 14 16
age


Catch Curve ages 6 through 8


y = -0.8346x + 9.2739
R2 = 0.9955


0 2 4 6 8 10 12 14 16
age


Figure 3-11. Catch curve analysis of 286 dead adult snook collected during the red tide
bloom of 2005.


5.00

4 4.00
O
- 3.00
0

- 2.00

- 1.00
C0.00
0.00


5.00
_-
()
4 4.00

4-- 3.00
0

- 2.00

- 1.00
C0.00
0.00











OtP.
1.00 -

0.90 -

D 0.80 -

0.70 -
-Apparent Survival
0.60 -
pA

0.50
z u c- -n > a -- <- > w O z 0
0 CD wD -0 w CD 0 0 CD
< 0 o- < I 0 (Q o 0 0 < )0



OtPt
1.00

0.90 -







0.50
0 CD ( a) -C a ) CD o 0 CD
,< 0 1- o- 0 < ,01 01 0< 0

B
Figure 3-12. Results for the Cormack-Jolly-Seber models (tP. and (tPJ These models
each had time dependent apparent survival (() but differ in time dependent
and fixed capture probabilities (p ).









Table 3-1. Location of study site exits and creeks and the determined number of VR2
receivers needed.
Exit Location VR2 Receivers Creek Location VR2 Receivers
New Pass 2 Bowless Creek 3
Big Pass 3 Whitaker Bayou 2
Longboat Pass 3 Phillippi Creek 1
Cortez (ICW) 3 North Creek 3
Venice (ICW) 1 South Creek 2


Table 3-2. Fates as of January 2006 of 66 tagged, harvestable size fish used to estimate
mortality.
Censored Emigrants Fishing Natural Unknown Still Alive
Mortalities Mortalities
4 8 16 6 17 15


Table 3-3. 2005 yearly estimates of fishing mortality, natural mortality, and their
corresponding AIC values attained for all three approaches of fate
assignments of questionable fish not heard after the red tide events of summer
2005. The number of fish used is noted for each approach.
Model Emigrants and Fishing Natural Mortalities 50/50 Approach
Mortalities Approach Approach (50 fish)
(46 fish) (55 fish)
F M AIC F M AIC F M AIC
ptFt Mt 0.66 0.16 228.97 0.41 0.65 232.60 0.48 0.42 233.32
F. M. 0.60 0.12 234.71 0.24 0.60 285.19 0.48 0.36 253.56
Fs M. 0.55 0.12 236.05 0.30 0.60 282.43 0.42 0.36 252.59
t Fs M. 0.51 0.12 219.32 0.30 0.60 261.37 0.39 0.36 232.15
t Fs Mt 0.50 0.16 225.76 0.30 0.65 223.81 0.34 0.50 218.78


Table 3-4. 2005 yearly total mortality estimates by modeling total mortality directly using
methods from Hightower et. al (2001) and by simply adding together the
Hightower fishing and natural mortality estimates.
Model Emigrants and Fishing Natural Mortalities 50/50 Approach
Mortalities Approach Approach
Direct Addition Direct Addition Direct Addition
ptFt Mt 0.80 0.82 1.08 1.06 0.93 0.90
F. M. 0.72 0.72 0.93 0.84 0.81 0.84
Fs M. 0.72 0.67 0.92 0.90 0.81 0.78
t Fs M. 0.70 0.63 0.90 0.90 0.77 0.75
t Fs Mt 0.68 0.66 0.87 0.95 0.77 0.84









Table 3-5. Estimates of total mortality (Z) were used along with natural mortality (M)
estimates from the Hoenig equation and estimates of emigration rate (E) to
figure fishing mortality (F) from all Cormack-Jolly-Seber models performed.
AIC values and degrees of freedom (DF) are shown for each model used (t
indicates a time dependent variable, a period indicates a fixed variable, and an
s indicates a variable fixed by open and closed harvest seasons).
Model AIC DF Z M Hoenig E F
P7tP. 698.58 16 0.56 0.21 0.04 0.31
(It t 699.22 30 0.54 0.21 0.04 0.29
(D./t 704.57 16 0.43 0.21 0.04 0.18
s/ t 706.51 17 0.43 0.21 0.04 0.18
pQ. /5. 713.23 2 0.47 0.21 0.04 0.22


Table 3-6. Review of the methods and results for the mortality estimations. The ranges
for the Hightower method are based on all five models used in the three
different modeling approaches.
Mortality Component Hightower Method Alternative Method Results
Results
Z 0.68 1.08 Catch Curve: 0.37 0.83
F 0.24-0.66 CJS: 0.18 0.31
M 0.12-0.65 Hoenig equation: 0.21














CHAPTER 4
DISCUSSION

Acoustic Array, Tagging and Detection

A key factor in this study was the ability to determine whether a tagged snook had

emigrated from Sarasota Bay with a high degree of certainty. The tag reception rate,

based on the initial range tests described earlier, was approximately 50%; however, this is

a bit misleading. During the tag testing, all tags were heard by the receivers during every

trial (100% detection rate), but not every possible tag signal was received. Thus, the 50%

reception rate represented the chance of receiving all of the signals from a tag as it moves

through the receiver detection field in the pass. All tags were detected on every test

conducted, but only 50% of the total potential soundings were heard. While this

reception rate was much lower than originally anticipated, the actual performance of the

receivers with tags implanted in snook was most likely better than the rates measured in

the trials. The trials were conducted by drifting the tags through multiple pass or exit

locations on a moving tide, which caused the tag to constantly move through each

location. Clements et al. (2005) and Heupel et al. (2006) found, and additional trials with

the manual receiver showed, that fast flowing water, such as during tide changes, greatly

reduced reception distance. The initial range tests were performed during the falling tide

change so the tags would drift past the receivers without the assistance or interference

from a boat motor to mimic snook movements as closely as possible and minimize boat

motor noise interference. Boat noise, although not the greatest source of interference, can

negatively influence range tests and acoustic array performance (Klimley et al. 1998;









Lacroix and Voegeli 2000). The data from the receivers and manual tracking suggest,

however, that snook made erratic movements within the passes, and thus were not

moving constantly through the receiver reception field. Also, snook primarily were

found to move throughout the bay and only aggregated in the passes during the spawning

season. This erratic movement behavior led to higher detection probabilities of animals

on each of our time intervals, and overall p ranged from about 0.70 to 1.00 per month.

Given that snook were observed to move erratically through the passes and not in a fairly

straight line as our test tags moved, it is likely that the average detection frequency was

greater than 50%. On trips with the manual receiver during periods of low water

movement in the passes or on trips within the bay, detection of 300-1000 meters was

regularly observed. Heupel et al. (2006) observed reception ranges of around 800-meters

with similar tags and receivers in similar locations, but did find that at these long

distances the reception rate does decrease.

Tag detection rate was highest during periods of seasonal temperature change when

snook were actively moving between summer spawning passes (Taylor et al. 1998) and

winter thermal refuges (Marshall 1958; Howells et al. 1990). During this time, fish were

moving long distances, increasing the likelihood of swimming within the detection range

of a receiver.

Detection rates dropped drastically during the summer months of 2005. This was

expected to some degree as snook would likely have left the bay to spawn in the passes

and then, possibly, move to the beaches to feed (Taylor et al. 1998). The winter of 2005-

2006 was expected to serve as a "check" period for the tagged fish that had survived the

spring and fall harvest seasons and the summer red tide events, as the remaining snook









would have to return to the creeks as a thermal refuge. However, detection rates did not

increase during the cold water period. If the tagged fish were alive and viable, they

should have been detected as they reentered the study site and the creeks the following

winter. The extreme drop in detection rate seen starting in July 2005 was most likely

from the effects of the intense red tide bloom that occurred in the study area at that time.

Model Assumptions

The Hightower et al. (2001) method has several assumptions; some of which can be

problematic. If a fish emigrated from the study site, but was not detected on a pass/exit

receiver as it left, it would be considered a fishing mortality as the last data point for the

fish would have most likely come from a receiver within the site. This would obviously

affect the model results by inflating the F estimates. Attempts were made early in the

study to reduce the chance of this happening. Much effort went into the pass/exit

receiver placement and range testing to assure that there was no route a snook could take

to circumvent detection. Tags were detected on every test performed after the optimal

receiver location was determined. Although undetected escape possibly could have

happened and should be considered when examining these results, this possibility is

believed to be unlikely.

Equal vulnerability of all fish used in the F and M estimation was assumed,

however, vulnerability likely changed with age or size. The assessment compensates for

this by using age-7 fish as a reference point for F estimates (Muller and Taylor 2006).

This is difficult to account for in this type of study where age of the tagged fish is

unknown. Using only harvestable size fish was the best option possible to reduce the

effects of violation of this assumption (Hightower et. al 2001).









Tag failure could have also been a problem. Tag failure could not be distinguished

from fishing mortality as both situations would have looked identical in the data either

would result in a fish seeming to be no longer present in the study site. Tag failure was

always a possibility, but all tags were in working order prior to deployment and all 10

returned tags (9 from angler and 1 from red-tide killed fish) continued to work throughout

the duration of their expected battery life.

Fishing and Natural Mortality Analysis

It was thought that the best approach to estimating fishing and natural mortality

was to stray from the guidelines of Hightower et al. (2001); the lost fish as natural

mortalities approach. In this approach, the 17 fish that were last heard during the red tide

event were considered natural mortalities no matter the location of their last detection

(inside array or on pass/exit). These fates were different than those that would have been

assigned if the guidelines from Hightower et al. (2001) were followed. These fates would

have been fishing mortalities and emigrants, but this did not seem to be supported by

what was being observed in the field.

The results from this modeling approach showed that fishing mortality for snook is

high and similar to the values estimated in the stock assessment during the harvest

seasons and that natural mortality from red tide events can also be an unexpectedly large

source of mortality. These results are the first direct estimates of natural mortality for a

sportfish population in Florida and are also the first direct estimates of natural mortality

from a red tide event anywhere.

Although the fate assignments were subjective, uncertainty was evaluated by

modeling with three fate assignment approaches. While the magnitude of the parameter

estimates from these approaches was different, the overall patterns in fishing mortality









were similar between the different approaches. Incorrect fate assignments for even a low

number of fish would influence the estimates. A key assumption in the lost fish as

emigrants and fishing mortalities approach was that the eight fish that were unaccounted

for in the array were assumed to have died due to fishing mortality. If these fish did

disappear from the system during this time period due to fishing, this would indicate a

high level of illegal harvest during the summer closed period a scenario that is likely of

interest to resource agencies. However, given the sudden drop in the frequency of tag

relocations, the closed harvest season, the exceptionally large red tide event, and the

observed large numbers of dead snook during this red tide bloom, it is more likely that

these eight fish died due to natural mortality.

Total Mortality Estimation

Total mortality estimates from this study were higher than those estimated by the

snook stock assessment (Muller and Taylor 2006). This is not unexpected, as the

assessment estimates were for the entire coast, and the red-tide event which caused much

of the mortality for this study was more localized. There appeared to be little difference

between the Z estimates attained through the Hightower modeling approach versus

simply adding the F and M estimates together. This implied that total mortality was

higher than expected, even if there was uncertainty about how that total was partitioned.

Estimates of Z attained through the catch curve analyses were lower than those from the

Hightower models likely because not all assumptions of the catch-curve were met. As in

Muller and Taylor (2006), the use of a catch-curve here is just to provide an

approximation of the total mortality values. In this case, the catch-curve was not very

informative because few ages were represented in the sample and the high variation in









snook growth patterns results in no age likely ever being fully recruited (Muller and

Taylor 2006).

The CJS models showed drops in apparent survival in April and August of 2005, at

the same time the other models indicated marked increases in mortality. This method

used only capture history and was not reliant on the assignment of natural or fishing

mortality as in the Hightower models (Cormack 1964; Jolly 1965; Seber 1965, Pollock et

al. 1990). However, this method did require that fate assignments with regard to

emigration be made. Given the extensive testing of the receiver array prior to the study,

it was believed that this design could reliably detect an emigrant, thus these survival

trends were likely not biased by mis-assigning a fish as an emigrant. The F estimates

from the CJS models were also similar to the F estimates from the lost fish as natural

mortalities approach Hightower type models.

Comparison to Assessment

The FWC 2005 Snook Stock Assessment (Muller and Taylor 2006) reported a

yearly F range for 2004 of 0.36 to 0.48 per year. These estimates were obtained using

fishery- dependent CAA models with an assumed natural mortality estimate of 0.25 per

year. This project estimated F values that ranged between 0.30 and 0.66 per year

depending on which approach was used to assign fates to questionable fish. These

estimates were obtained without fisher dependence and were independent of the natural

mortality estimates which ranged from 0.12 to 0.65 per year, depending on fate

assignment. Although these were estimates made in successive years, some general

comparisons are possible. Snook harvest is indeed high and the methods used by FWC

are most likely providing accurate estimates of F. However, a large discrepancy exists

between the estimates of natural mortality. This is due to the high, regional natural









mortality observed in Sarasota Bay in 2005. This high rate is most likely not constant,

but rather event driven. It may, however, become more important to consider these

possibly large regional sources of natural mortality if red tide becomes more prevalent in

the future.

Method Comparison

This project was funded for a number of reasons: 1) to respond to a need mentioned

in the assessment for regional data, 2) to directly estimate M, 3) to provide a fisher-

independent estimate ofF, and 4) to provide an example of the benefits, problems, and

utility of using this method with other species in Florida. Information regarding this last

reason may be the most useful in the long run. Walters and Martell (2004) extensively

discuss tradeoffs in fisheries management, both in the methods that are used to make

decisions and the tradeoffs that must occur when making social and management

decisions related to limited fish resources. This study is an example of evaluating

tradeoffs in estimating model parameters which can provide different information for

making a management decision. Using telemetry methods to directly estimate fishing

and natural mortality has several key attributes including: 1) direct, fisher-independent

measures ofF and M, 2) continuous data collection for multiple fish/organisms

simultaneously, and 3) movement and habitat data (Hightower et al. 2001; Pine et al.

2003; Walters and Martell 2004). One way in which telemetry can be used to

compensate for the problems with the traditional assessment methods mentioned earlier,

is by combining it with the other methods; such as using telemetry to estimate reporting

rate and natural mortality in a traditional tagging study (Pollock et al. 2004). This would

solve two of the problems associated with the traditional tagging method. The tradeoff

for this telemetry method is that it also has its disadvantages. These include: 1) expense









(high equipment cost), 2) high maintenance (high man hours with certain skills required

depending on how receivers are deployed), 3) potential problems with fate assignment

and potential need to assume fates of questionable fish, and 4) potentially strong

influence of fate mis-classification on parameter estimates due to small sample size.

Utility in Florida

This project was done using snook as the study species, however, this method can

be used on many marine species and the study could be designed for multiple species

simultaneously. Telemetry methods to directly estimate F and/or M have been used for a

variety of fish species including striped bass (Hightower et al. 2001), blacktip sharks

(Heupel and Simpfendorfer 2002), largemouth bass (Waters et al. 2005), salmon

(Bendock and Alexandersdottir 1993), and lingcod (Starr et al. 2005) for example.

Telemetry has been used to assess movement and habitat use of fish species in many

studies (Standora and Nelson 1977; Klimley et al. 1988; Lagardere et al. 1990; Lacroix

and McCurdy 1996; Arendt et al. 2001b; and Heupel et al. 2003).

Stock assessments for other fish species in Florida including red drum (Murphy

2005), spotted seatrout (Murphy 2003), mullet (Mahmoudi 2005), and sheepshead

(Munyandorero 2006); all use CAA methods similar to the snook stock assessment. In

each of these assessments, the authors have identified the need for comparative mortality

estimates to provide a "check" of the mortality estimates currently estimated to help

better manage these stocks. Telemetry studies such as this could fill that need.

Future applications of this method may want to work in smaller, more

geographically closed systems than what was used for this study (e.g., Klimley et al.

1998; Lacroix and Voegeli 2000; Clements et al. 2005; Heupel et. al 2006). Large, more

open systems require more receivers to close the exits, more field effort to identify









natural mortalities through manual tracking, and a greater chance that a fish will emigrate

from the system or not be identified as mortality. Heupel and Simpfendorfer (2002) and

Heupel et al. (2003) worked in an ideal situation where they were able to have total

acoustic coverage for their study site. This eliminated the need for assuming fates of lost

fish as the data resolution was high enough to identify an animal's location at all times

allowing swim-speed, dispersion rate, predation, and detailed movement patterns to be

calculated for juvenile blacktip sharks. Increasing the number of receivers, or an

improvement in the available receiver technology, could reduce the need for manual

tracking and would likely increase the detail with which mortality and movement patterns

could be examined (Klimley et al. 1998; Lacroix and Voegeli 2000; Clements et. al 2005;

Heupel et al. 2006).

Conclusions

A major finding of this study is that natural mortality can be very high related to

red tide blooms at discrete spatial and temporal levels. Much of the data used in FWRI

assessments is collected at regional field offices around the state. However, much of the

assessment estimations are conducted at larger, coast wide scales. This study

demonstrates that natural mortality events may be large locally and if red tide events

increase with frequency, intensity, or spatial area, then these larger natural mortality rates

may need to be incorporated into the stock assessment programs.

This study provided the first direct measurements ofF and M for any game-fish

species in Florida and will serve as an excellent comparison to the derived estimates for

snook currently used in the stock assessment. These direct measurements identified

temporal changes in natural mortality due to red tide events. This study demonstrated the

problems, but also the utility, of using a large acoustic tagging program in an estuarine






55


setting which could be used to address identified management needs for other species;

such as providing rapid estimation of mortality rates in evaluating the effectiveness of

newly implemented harvest regulations or other management actions.















APPENDIX
HIGHTOWER MODELS AND CORMACK-JOLLY-SEBERS MODEL RESULTS














Table A-1. Results for models assuming that all fish last heard during the summer of 2005 were emigrants & fishing mortalities
I pt Ft Mt p F.M. I Fs M. It Fs M. It Fs Mt
Variable Value SE Value SE Value SE Value SE Value SE


F-Nov 2004
F-Dec 2004
F-Jan 2005
F-Feb 2005
F-Mar 2005
F-Apr 2005
F-May 2005
F-Jun 2005
F-Jul 2005
F-Aug 2005
F-Sep 2005
F-Oct 2005
F-Nov 2005
F-Dec 2005
F-Jan 2006
M-Nov 2004
M-Dec 2004
M-Jan 2005
M-Feb 2005
M-Mar 2005
M-Apr 2005
M-May 2005
M-Jun 2005
M-Jul 2005
M-Aug 2005
M-Sep 2005
M-Oct 2005
M-Nov 2005
M-Dec 2005
M-Jan 2006


0.00
0.00
0.00
0.05
0.02
0.07
0.00
0.02
0.03
0.18
0.12
0.08
0.03
0.04
0.47
0.00
0.00
0.00
0.00
0.00
0.08
0.00
0.00
0.00
0.08
0.00
0.00
0.00
0.00
0.00


0.01
0.01
0.02
0.04
0.02
0.04
0.01
0.03
0.04
0.09
0.08
0.07
0.07
0.12
0.26
0.00
0.00
0.00
0.00
0.00
0.04
0.00
0.00
0.00
0.05
0.00
0.00
0.00
0.00
0.00


0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01


0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01


0.06
0.04
0.04
0.06
0.06
0.06
0.04
0.04
0.04
0.04
0.06
0.06
0.06
0.04
0.04
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01


0.02
0.01
0.01
0.02
0.02
0.02
0.01
0.01
0.01
0.01
0.02
0.02
0.02
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01


0.05
0.03
0.03
0.05
0.05
0.05
0.03
0.03
0.03
0.03
0.05
0.05
0.05
0.03
0.03
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01


0.02
0.01
0.01
0.02
0.02
0.02
0.01
0.01
0.01
0.01
0.02
0.02
0.02
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01


0.05
0.03
0.03
0.05
0.05
0.05
0.03
0.03
0.03
0.03
0.05
0.05
0.05
0.03
0.03
0.00
0.00
0.00
0.00
0.00
0.08
0.00
0.00
0.00
0.08
0.00
0.00
0.00
0.00
0.00


0.02
0.01
0.01
0.02
0.02
0.02
0.01
0.01
0.01
0.01
0.02
0.02
0.02
0.01
0.01
0.01
0.00
0.00
0.00
0.00
0.04
0.00
0.00
0.00
0.05
0.00
0.00
0.00
0.00
0.01














Table A-2. Results for all models assuming that all fish last heard during the summer of 2005 were natural mortalities
Spt Ft Mt p F.M. I Fs M. Pt Fs M. I t Fs Mt
Variable Value SE Value SE Value Variable Value SE Value SE


F-Nov 2004
F-Dec 2004
F-Jan 2005
F-Feb 2005
F-Mar 2005
F-Apr 2005
F-May 2005
F-Jun 2005
F-Jul 2005
F-Aug 2005
F-Sep 2005
F-Oct 2005
F-Nov 2005
F-Dec 2005
F-Jan 2006
M-Nov 2004
M-Dec 2004
M-Jan 2005
M-Feb 2005
M-Mar 2005
M-Apr 2005
M-May 2005
M-Jun 2005
M-Jul 2005
M-Aug 2005
M-Sep 2005
M-Oct 2005
M-Nov 2005
M-Dec 2005
M-Jan 2006


0.00
0.00
0.01
0.04
0.02
0.08
0.00
0.00
0.00
0.00
0.12
0.08
0.03
0.04
0.46
0.00
0.00
0.00
0.00
0.00
0.07
0.00
0.05
0.18
0.35
0.00
0.00
0.00
0.00
0.00


0.01
0.01
0.02
0.04
0.03
0.04
0.00
0.00
0.02
0.04
0.07
0.07
0.07
0.12
0.26
0.00
0.00
0.00
0.00
0.00
0.04
0.00
0.03
0.07
0.11
0.00
0.00
0.00
0.00
0.00


0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.05


0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01


0.05
0.00
0.00
0.05
0.05
0.05
0.00
0.00
0.00
0.00
0.05
0.05
0.05
0.00
0.00
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.05


0.02
0.01
0.01
0.02
0.02
0.02
0.01
0.01
0.01
0.01
0.02
0.02
0.02
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01


0.05
0.00
0.00
0.05
0.05
0.05
0.00
0.00
0.00
0.00
0.05
0.05
0.05
0.00
0.00
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.05


0.02
0.01
0.01
0.02
0.02
0.02
0.01
0.01
0.01
0.01
0.02
0.02
0.02
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01


0.05
0.00
0.00
0.05
0.05
0.05
0.00
0.00
0.00
0.00
0.05
0.05
0.05
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.08
0.00
0.04
0.17
0.35
0.00
0.00
0.00
0.00
0.00


0.02
0.00
0.00
0.02
0.02
0.02
0.00
0.00
0.00
0.00
0.02
0.02
0.02
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.04
0.00
0.03
0.07
0.11
0.00
0.00
0.00
0.00
0.00














Table


A-3. Results for all models assuming that the fish last heard during the summer
mortalities/natural mortalities


of 2005 were 50/50 emigrants and fishing


p t Ft Mt p F. M. p Fs M. pt Fs M. pt Fs Mt
Variable Value SE Value SE Value Variable Value SE Value SE


F-Nov 2004
F-Dec 2004
F-Jan 2005
F-Feb 2005
F-Mar 2005
F-Apr 2005
F-May 2005
F-Jun 2005
F-Jul 2005
F-Aug 2005
F-Sep 2005
F-Oct 2005
F-Nov 2005
F-Dec 2005
F-Jan 2006
M-Nov 2004
M-Dec 2004
M-Jan 2005
M-Feb 2005
M-Mar 2005
M-Apr 2005
M-May 2005
M-Jun 2005
M-Jul 2005
M-Aug 2005
M-Sep 2005
M-Oct 2005
M-Nov 2005
M-Dec 2005
M-Jan 2006


0.00
0.00
0.00
0.05
0.02
0.07
0.00
0.00
0.05
0.04
0.12
0.09
0.02
0.01
0.51
0.00
0.00
0.00
0.00
0.00
0.07
0.00
0.02
0.14
0.18
0.00
0.00
0.00
0.00
0.00


0.01
0.01
0.02
0.04
0.02
0.04
0.00
0.00
0.04
0.05
0.08
0.07
0.06
0.12
0.28
0.00
0.00
0.00
0.00
0.00
0.04
0.00
0.02
0.06
0.08
0.00
0.00
0.00
0.00
0.00


0.04
0.04
0.04
0.04
0.04
0.04
0.04
0.04
0.04
0.04
0.04
0.04
0.04
0.04
0.04
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03


0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01


0.05
0.02
0.02
0.05
0.05
0.05
0.02
0.02
0.02
0.02
0.05
0.05
0.05
0.02
0.02
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03


0.02
0.01
0.01
0.02
0.02
0.02
0.01
0.01
0.01
0.01
0.02
0.02
0.02
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01


0.05
0.01
0.01
0.05
0.05
0.05
0.01
0.01
0.01
0.01
0.05
0.05
0.05
0.01
0.01
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03


0.02
0.01
0.01
0.02
0.02
0.02
0.01
0.01
0.01
0.01
0.02
0.02
0.02
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01


0.06
0.00
0.00
0.06
0.06
0.06
0.00
0.00
0.00
0.00
0.06
0.06
0.06
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.08
0.00
0.02
0.17
0.23
0.00
0.00
0.00
0.00
0.00


0.02
0.00
0.00
0.02
0.02
0.02
0.00
0.00
0.00
0.00
0.02
0.02
0.02
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.04
0.00
0.02
0.06
0.09
0.00
0.00
0.00
0.00
0.01














Table A-4. Results from all Cormack-Jolly Seber models.
trp Vatpt 1E.pt i spt S.u p
Variable Value SE Value SE Value Variable Value SE Value SE


O-Nov 2004
O-Dec 2004
O-Jan 2005
O-Feb 2005
O-Mar 2005
O-Apr 2005
O-May 2005
O-Jun 2005
O-Jul 2005
O-Aug 2005
O-Sep 2005
O-Oct 2005
0- Nov 2005
O-Dec 2005
O-Jan 2006
p-Nov 2004
p-Dec 2004
p-Jan 2005
p-Feb 2005
p-Mar 2005
p-Apr 2005
p-May 2005
p-Jun 2005
p-Jul 2005
p-Aug 2005
p-Sep 2005
p-Oct 2005
p- Nov 2005
p-Dec 2005
p-Jan 2006


1
0.98
0.98
0.96
0.98
0.83
0.98
0.97
0.86
0.69
0.88
0.92
0.93
0.83
0.61
0.85
0.85
0.85
0.85
0.85
0.85
0.85
0.85
0.85
0.85
0.85
0.85
0.85
0.85
0.85


0.00
0.02
0.02
0.03
0.03
0.06
0.03
0.04
0.06
0.08
0.07
0.07
0.07
0.11
0.15
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02


1
0.98
0.99
0.96
0.96
0.85
0.98
0.95
0.87
0.70
0.88
0.93
0.95
0.84
0.71
0.82
0.92
0.72
0.76
0.98
0.82
0.86
0.97
0.89
0.88
0.91
0.80
0.75
0.78
0.71


0.00
0.02
0.02
0.04
0.03
0.05
0.03
0.03
0.06
0.08
0.07
0.07
0.09
0.15
66.56
0.12
0.04
0.06
0.06
0.02
0.06
0.05
0.03
0.06
0.07
0.06
0.09
0.11
0.14
66.56


0.92
0.82
0.92
0.73
0.77
0.98
0.80
0.86
0.98
0.91
0.84
0.90
0.80
0.76
0.73
0.51
0.92
0.92
0.92
0.92
0.92
0.92
0.92
0.92
0.92
0.92
0.92
0.92
0.92
0.92


0.01
0.12
0.04
0.06
0.06
0.02
0.06
0.05
0.02
0.05
0.08
0.07
0.09
0.10
0.12
0.13
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01


0.91
0.93
0.93
0.91
0.91
0.91
0.93
0.93
0.93
0.93
0.91
0.91
0.91
0.93
0.93
0.82
0.92
0.73
0.77
0.98
0.79
0.86
0.98
0.91
0.85
0.90
0.80
0.76
0.74
0.52


0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.12
0.04
0.06
0.06
0.02
0.06
0.05
0.02
0.05
0.08
0.07
0.09
0.10
0.12
0.13


0.91
0.91
0.91
0.91
0.91
0.91
0.91
0.91
0.91
0.91
0.91
0.91
0.91
0.91
0.91
0.85
0.85
0.85
0.85
0.85
0.85
0.85
0.85
0.85
0.85
0.85
0.85
0.85
0.85
0.85


0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02















LIST OF REFERENCES


Akaike, H. 1974. A New Look at the Statistical Model Identification. IEEE
Transactions on Automatic Control 19(6):716-723

Anderson, D.R., K.P. Burnham, and G.C. White. 1998. Comparison of Akaike
Information Criterion for Model Selection and Statistical Inference from Capture-
recapture Studies. Journal of Applied Statistics 25(2):263-282

Arendt, M.D., A.J. Lucy, and D.A. Evans. 2001b. Diel and Seasonal Activity Patterns of
Adult Tautog, Tautoga onitis, in Lower Chesapeake Bay, Inferred form Ultrasonic
Telemetry. Environmental Biology of Fishes 62:379-391.

Bendock, T., and M. Alexandersdottir. 1993. Hooking Mortality of Chinook Salmon
Released in the Kenai River, Alaska. North American Journal of Fisheries
Management 13:540-549.

Beverton, R.J., and S.J. Holt. 1959. A Review of the Life Spans and Mortality rates of
Fish in Nature, and Their Relation to Growth and Other Physiological
Characteristics. In CIBA Foundation Colloquia on Aging, the Lifespans of
Animals. (Wolshenholme, G.E.W. and M. O'Connor, Eds.) London, Churchill.
Volume 5:142-180

Burnham, K.P., D.R. Anderson, G.C. White, C. Brownie, and K.H. Pollock. 1987.
Design and Analysis Methods for Fish Survival Experiments Based on Release-
Recapture. American Fisheries Society, Maryland.

Chapman, D. G., and D.S. Robson. 1960. The Analysis of a Catch Curve. Biometrics
16:354-368.

Clements, S., D. Jepsen, M. Karnowski, and C.B. Schreck. 2005. Optimization on an
Acoustic Telemetry Array for Detecting Transmitter-Implanted Fish. North
American Journal of Fisheries Management 25:429-436

Cormack, R.M. 1964. Estimates of Survival from the Sighting of Marked Animals.
Biometrika 51(3/4):429-438

Fore, P. L., and T. W. Schmidt. 1973. Biology of Juvenile and Adult Snook,
Centropomus undecimalis, in the Ten Thousand Islands. Pp. XVI 1-18. in
Ecosystems analysis of the Big Cypress Swamp and estuaries. U. S. Environmental
Protection AGENCY Region IV, Atlanta, Georgia Publication Number EPA 904/9-
74-002.









Gilmore, R. G., C. J. Donahoe and D.W. Cooke. 1983. Observations on the Distribution
and Biology of the Common Snook, Centropomus undecimalis (Bloch). Florida
Scientist 46: 313-336.

Haddon, M. 2001. Modeling and Quantitative Methods in Fisheries. Chapman and
Hall/CRC, Boca Raton, London, New York, and Washington, D.C.

Heupel, M. R., and C. A. Simpfendorfer. 2002. Estimation of Mortality of Juvenile
Blacktip Sharks, Carcharhinus limbatus, within a Nursery Area Using Telemetry
Data. Canadian Journal of Fisheries and Aquatic Sciences 59:624-632.

Heupel, M.R., C.A. Simpfendorfer & R.E. Hueter. 2003. Running Before the Storm:
Blacktip Sharks Respond to Falling Barometric Pressure Associated with Tropical
Storm Gabrielle. Journal of Fish Biology 63:1357-1363.

Heupel, M.R., J.M. Semmens, and A.J. Hobday. 2006. Automated Acoustic Tracking of
Aquatic Animals: Scales, Design and Deployment of Listening Station Arrays.
Marine and Freshwater Research 57:1-13.

Hightower, J.E., J. R., Jackson, and K. H. Pollock. 2001. Use of Telemetry Methods to
Estimate Natural and Fishing Mortality of Striped Bass in Lake Gaston, North
Carolina. Transactions of the North American Fisheries Society 130:557-567.

Hilborn R., and C.J. Walters. 1992. Quantitative Fisheries Stock Assessment: Choice,
Dynamics, and Uncertainty. Chapman & Hall, International Thomson Publishing,
New York, New York.

Hoenig, J.M. 1983. Empirical Use of Longevity Data to Estimate Mortality Rates.
Fishery Bulletin 82:898-903

Howells, R. G., A. J. Sonski, P. L. Shafland, and B. D. Hilton 1990. Lower Temperature
Tolerance of snook, Centropomus undecimalis. Short Papers and Notes. Northeast
Gulf Sciences 11(2):155-158

Jensen, A. L. 1996. Beverton and Holt Life History in Variants Result from Optimal
Trade-off of Reproduction and Survival. Canadian Journal of Fisheries and
Aquatic Sciences 53:820-822.

Jolly, G.M. 1965. Explicit Estimates from Capture-recapture Data With Both Death and
Immigration-stochastic Model. Biometrika 52(1/2):225-247

Kendall, W.L. and J.D. Nichols. On the Use of Secondary Capture-recapture Samples to
Estimate temporary Emigration and Breeding Proportions. Journal of Applied
Statistics 22(5/6):751-762

Klimley, A.P., S.B. Butler, D.R. Nelson, A. Stull. 1988. Die Movements of Scalloped
Hammerhead Sharks, Sphurna lewini, Griffith and Smith, to and from a Seamount
in the Gulf of California. Journal of Fish Biology 33:751-761









Klimley, A.P., F. Voegeli, S.C. Beavers, and B.J. Le Boeuf. 1998. Automated Listening
Stations for Tagged Marine Fishes. Marine Technology Society Journal 32:94-101.

Lacroix, G.L., and P. McCurdy. 1996. Migratory Behavior of Post-smolt Atlantic
Salmon during Initial Stages of Seaward Migration. Journal of Fish Biology
49:1086-1101.

Lacroix, G.L., and F.A. Voegeli. 2000. Development of Automated Monitoring Systems
for Ultrasonic Transmitters. In Advances in Fish Telemetry. Eds A. Moore and I.
Russell, editors. CEFAS, Lowestoft.

Lagardere, J.P., J.J. Ducamp, L. Favre, J. Mosneron Dupin, and M. Sperandio. 1990. A
Method for the Quantitative Evaluation of Fish Movements in Salt Ponds by
Acoustic Telemetry. Journal of Experimental Marine Biology and Ecology
141:221-236.

Mahmoudi, B. 2005. The 2005 Update of the Stock Assessment for Striped Mullet,
Mugil cephalus in Florida. Fish and Wildlife Conservation Commission, Florida
Marine Research Institute, St. Petersburg, FL.

Marshall, A. R. 1958. A Survey of the Snook Fishery of Florida, with Studies of the
Biology of the Principal Species, Centropomus undecimalis (Bloch). Florida Board
of Conservation Marine Research Laboratory Technical Series Number 22.

McMicheal, R. H., Jr., K. M. Peters and G. R. Parsons. 1989. Early Life History of the
Snook, Centropomus undecimalis in Tampa Bay, Florida. Northeast Gulf Science
10(2):113-126.

Muller, R. G., and R. G. Taylor. 2006. The 2005 Stock Assessment Update of Common
Snook, Centropomus undecimalis. Fish and Wildlife Conservation Commission,
Florida Marine Research Institute, St. Petersburg, FL.

Munyandorero, J., M.D. Murphy, and T.C. MacDonald. 2006. An Assessment of the
Status of Sheepshead in Florida Waters through 2004. Fish and Wildlife
Conservation Commission, Florida Marine Research Institute, St. Petersburg, FL.

Murphy, M. D. 1997. Bias in Chapman-Robson and Least-squares Estimators of
Mortality Rates for Steady-state Populations. Fishery Bulletin 95: 863-868.

Murphy, M.D. 2003. A Stock Assessment of Spotted Seatrout Cynoscion nebulosus in
Florida: Status of the Stocks through 2001. Fish and Wildlife Conservation
Commission, Florida Marine Research Institute, St. Petersburg, FL.

Murphy, M.D. 2005. A Stock Assessment of Red Drum, Sciaenops ocellatus, in Florida:
Status of the Stocks through 2003. Fish and Wildlife Conservation Commission,
Florida Marine Research Institute, St. Petersburg, FL.









Ney, J. J. 1993. Dynamics of exploited fish populations. Inland Fisheries Management
in North America. Edited by C. C. Kohler and W. A. Hubert. American Fisheries
Society, Bethesda, Maryland.

Pauly, D. 1980. On the Interrelationships between Natural Mortality, Growth
Parameters, and Mean Environmental Temperature in 175 fish stocks. Conseil
International pour l'Exploration de la Mer Journal 39(2): 175-192

Peters, K. M., R. E. Matheson, Jr. and R. G. Taylor. 1998b. Reproduction and Early Life
History of Common Snook, Centropomus undecimalis (Bloch), in Florida. Bulletin
of Marine Sciences 62:509-529.

Peterson, I., and J. S Wroblewski. 1984. Mortality Rates of Fishes in the Pelagic
Ecosystem. Canadian Journal of Fisheries and Aquatic Sciences 41:1117-1120.

Pine, W.E. III, K.H. Pollock, J.E. Hightower, T.J. Kwak, and J.A. Rice. 2003. A Review
of Tagging Methods for Estimating Fish Population Size and Components of
Mortality. Fisheries 28(10): 10-23

Pollock, K. H., C. M. Bunck, S. R. Winterstein, and C. L. Chen. 1995. A Capture-
recapture Survival Analysis Model for Radio-tagged Animals. Journal of Applied
Statistics 22:661-672.

Pollock, K. H., J. M. Hoenig, W. S. Hearn, and B. Calingaert. 2001. Tag Reporting Rate
Estimation: 1. An evaluation of the high-reward tagging method. North American
Journal of Fisheries Management 21 (3): 521-532

Pollock, K. H., J. M. Hoenig, W. S. Hearn, and B. Calingaert. 2002. Tag Reporting Rate
Estimation: 2. Use of High Reward Tagging and Observers in Multi-component
Fisheries. North American Journal of Fisheries Management 22:727-736

Pollock, K.H., H. Jiang, and J.E. Hightower. 2004. Combining Telemetry and Fisheries
Tagging Models to Estimate Fishing and Natural Mortality Rates. Transactions of
the American Fisheries Society 133:639-648.

Pollock, K.H. J.D. Nichols, C. Brownie, and J.E. Hines. 1990. Statistical Inference for
Capture Recapture Experiments. Wildlife Monograph 107.

Quinn, T. J. II, and R. B. Deriso. 1999. Quantitative Fish Dynamics. Oxford University
Press, Oxford.

Ricker, W.E. 1975. Computation and Interpretation of Biological Statistics of Fish
Populations. Fisheries Research Board of Canada Bulletin 191.

Rivas, L. R. 1986. Systematic Review of the Perciform Fishes of the Genus
Centropomus. Copeia 1986(3):579-611.









Sea Stats. 2000. Florida Fish & Wildlife Conservation Commission, Florida Marine
Research Institute. St. Petersburg, FL. http://www.fmri.usf.edu

Seber, G.A. 1965. A Note on the Multiple-recapture Census. Biometrika 52(1/2):249-
259

Standora, E.A., and D.R. Nelson. 1977. A Telemetric Study of the Behavior of Free-
swimming Pacific Angel Sharks, Squatina califonica. Southern California
Academy of Sciences 76:193-201.

Starr R.M., O'Connell V., Ralston S., Breaker L. 2005. Use of Acoustic Tags to Estimate
Natural Mortality, Spillover, and Movements of Lingcod (Ophiodon elongatzu) in a
Marine Reserve. Marine Technology Society Journal 39 (1): 19-30.

Steidinger K.A., and K. Haddad. 1981. Biologic and Hydrographic Aspects of Red Tide.
Bioscience 31(11):814-819

Taylor R. G., H. J. Grier, and J. A. Whittington. 1998. Spawning Rhythms of Common
Snook in Florida. Journal of Fisheries Biology 53:502-520.

Taylor, R. G., J. A. Whittington, and H. J. Grier. 1993. Biology of Common Snook from
the East and West Coasts of Florida. Study 3, Sect. 1. In investigations into
nearshore and estuarine gamefish distributions and abundance, ecology, life
history, and population genetics in Florida. Edited by R. E. Crabtree, T. M. Bert,
and R. G. Taylor. FDNR/FMRI Rep. No. F0165-F0296-88-93-C. U.S. Department
of the Interior, Fish and Wildlife Service, Washington, D.C.

Taylor, R. G., J. A. Whittington, H. J. Grier, R. E. Crabtree. 2000. Age, Growth,
Maturation, and Protantric Sex Reversal in the Common Snook, Centropomus
undecimalis, from South Florida Waters. Fisheries Bulletin 98(3): 612-624.

Taylor, R. G., J. A. Whittington, D. E. Haymans. 2001. Catch-and-Release Mortality
Rates of Common Snook in Florida. North American Journal of Fisheries
Management. 21:70-75.

Thue, E. B., E. S. Rutherford and D. G. Buker. 1982. Age, Growth and Mortality of the
Common Snook, Centropomus undecimalis (Bloc), in the Everglades National
Park, Florida. U.S. National Park Service South Florida Research Center Report T-
683.

Tringali, M. D. and T. M. Bert. 1996. The Genetic Stock Structure of Common Snook
(Centropomus undecimalis). Canadian Journal of Fisheries and Aquatic Sciences.
53:974-984.

Van Den Avyle, M. J. 1993. Dynamics of Exploited Fish Populations. Inland Fisheries
Management in North America. Edited by C. C. Kohler and W. A. Hubert.
American Fisheries Society, Bethesda, Maryland.









Waters, D.S., R.L. Noble, and J.E. Hightower. 2005. Fishing and Natural Mortality of
Adult Largemouth Bass in a Tropical Reservoir. Transactions of the American
Fisheries Society 134:563-571.

Walters, C.J., and S. J. D. Martell 2004. Fisheries Ecology and Management. Princeton
University Press, Princeton, New Jersey.

White, G. C. 1983. Numerical Estimation of Survival Rates from Band Recovery and
Biotelemetry Data. Journal of Wildlife Management 47:716-728.

White, G.C., and K. P. Burnham. 1999. Program MARK: Survival Estimation from
Populations of Marked Animals. Bird Study 46 Supplement, 120-138.

White, G.C., and R.A. Garrott. 1990. Analysis of Wildlife Radio-tracking Data.
Acadimic Press, San Diego, California.

Williams, B.K., J.D.Nichols, and M.J. Conroy. 2002. Analysis and Management of
Animal Populations. Academic Press, San Diego, California.

Zehfuss, K. P., J. E. Hightower, and K. H. Pollock. 1999. Abundance of Gulf sturgeon
in the Apalachicola River, Florida. Transactions of the American Fisheries Society
128: 130-143.
















BIOGRAPHICAL SKETCH

Jason P. Bennett is from Kansas City, MO. He earned his B.S. in biology from the

University of Missouri Kansas City in 2003. He came to the University of Florida in

2003 and began his master's research in 2004. He completed his master's research in

2006.