<%BANNER%>

Age-Structured Population Model for Evaluating Gulf Sturgeon Recovery on the Apalachicola River, Florida

Permanent Link: http://ufdc.ufl.edu/UFE0024115/00001

Material Information

Title: Age-Structured Population Model for Evaluating Gulf Sturgeon Recovery on the Apalachicola River, Florida
Physical Description: 1 online resource (74 p.)
Language: english
Creator: Flowers, Henry
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2008

Subjects

Subjects / Keywords: Fisheries and Aquatic Sciences -- Dissertations, Academic -- UF
Genre: Fisheries and Aquatic Sciences thesis, M.S.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Gulf sturgeon Acipenser oxyrinchus desotoi is a U.S. federally threatened species found throughout the northern Gulf of Mexico. The viability of the Apalachicola River, Florida Gulf sturgeon population is a concern to fishery managers because of the small population size, slow rate of population recovery, and numerous threats to recovery, including habitat loss and interstate water allocation battles. Although Gulf sturgeon harvest has been illegal since 1985, managers and researchers are concerned that the population is not recovering at the desired rate to meet the recovery criteria objective of delisting by 2023. I developed an age-structured population model for the Apalachicola River Gulf sturgeon to describe population recovery and assess the efficacy of proposed management options. These options represent predicted population responses to competing hypotheses of why the Apalachicola River Gulf sturgeon population has not recovered to the expected level after twenty years of fishery closure. Hypothesis 1 proposes Gulf sturgeon populations are not recovering as a result of very low population size at the end of fishing and given life history characteristics of Gulf sturgeon. Hypothesis 2 proposes Gulf sturgeon recruitment was low due to lack of access to historic spawning areas limiting population recovery. I evaluated these hypothesis and corollary management options using 5 modeling scenarios. Model evaluation of these scenarios indicated that the Apalachicola River Gulf sturgeon population is greatly reduced from historic, pre-exploitation abundance and full population recovery will take in excess of 100 years from time of fishery closure. The observed abundance at the end of exploitation in 1984 was significantly lower than the level I estimated would allow population recovery by 2023 (target year from USFWS recovery plan), and any increases in total mortality would further slow recovery. The Apalachicola River Gulf sturgeon population recovery is not likely spawning habitat limited at this time (hypothesis 2). A stock enhancement program may provide short-term population increases, but the short-term gain in population size may not be worth the risks to the wild stock of initiating a stocking program. A key finding is that Gulf sturgeon recovery will be protracted because of reduced spawning potential due to erosion of the adult age-structure likely related to historic fishing pracitices. This loss of large, mature, highly fecund individuals within the population greatly reduces annual reproductive output slowing population recovery. To promote population recovery, I recommend no short term changes to current Gulf sturgeon management strategies. However, efforts should be made to minimize any adult mortality from direct (targeted sampling or fishing) or indirect (by-catch, incidental take) sources. Recovery criteria should be revised to incorporate quantitative and/or management metrics. Improvements in monitoring methods should provide greater resolution on the current status and trends in the Apalachicola River Gulf sturgeon.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Henry Flowers.
Thesis: Thesis (M.S.)--University of Florida, 2008.
Local: Adviser: Pine, William.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2008
System ID: UFE0024115:00001

Permanent Link: http://ufdc.ufl.edu/UFE0024115/00001

Material Information

Title: Age-Structured Population Model for Evaluating Gulf Sturgeon Recovery on the Apalachicola River, Florida
Physical Description: 1 online resource (74 p.)
Language: english
Creator: Flowers, Henry
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2008

Subjects

Subjects / Keywords: Fisheries and Aquatic Sciences -- Dissertations, Academic -- UF
Genre: Fisheries and Aquatic Sciences thesis, M.S.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Gulf sturgeon Acipenser oxyrinchus desotoi is a U.S. federally threatened species found throughout the northern Gulf of Mexico. The viability of the Apalachicola River, Florida Gulf sturgeon population is a concern to fishery managers because of the small population size, slow rate of population recovery, and numerous threats to recovery, including habitat loss and interstate water allocation battles. Although Gulf sturgeon harvest has been illegal since 1985, managers and researchers are concerned that the population is not recovering at the desired rate to meet the recovery criteria objective of delisting by 2023. I developed an age-structured population model for the Apalachicola River Gulf sturgeon to describe population recovery and assess the efficacy of proposed management options. These options represent predicted population responses to competing hypotheses of why the Apalachicola River Gulf sturgeon population has not recovered to the expected level after twenty years of fishery closure. Hypothesis 1 proposes Gulf sturgeon populations are not recovering as a result of very low population size at the end of fishing and given life history characteristics of Gulf sturgeon. Hypothesis 2 proposes Gulf sturgeon recruitment was low due to lack of access to historic spawning areas limiting population recovery. I evaluated these hypothesis and corollary management options using 5 modeling scenarios. Model evaluation of these scenarios indicated that the Apalachicola River Gulf sturgeon population is greatly reduced from historic, pre-exploitation abundance and full population recovery will take in excess of 100 years from time of fishery closure. The observed abundance at the end of exploitation in 1984 was significantly lower than the level I estimated would allow population recovery by 2023 (target year from USFWS recovery plan), and any increases in total mortality would further slow recovery. The Apalachicola River Gulf sturgeon population recovery is not likely spawning habitat limited at this time (hypothesis 2). A stock enhancement program may provide short-term population increases, but the short-term gain in population size may not be worth the risks to the wild stock of initiating a stocking program. A key finding is that Gulf sturgeon recovery will be protracted because of reduced spawning potential due to erosion of the adult age-structure likely related to historic fishing pracitices. This loss of large, mature, highly fecund individuals within the population greatly reduces annual reproductive output slowing population recovery. To promote population recovery, I recommend no short term changes to current Gulf sturgeon management strategies. However, efforts should be made to minimize any adult mortality from direct (targeted sampling or fishing) or indirect (by-catch, incidental take) sources. Recovery criteria should be revised to incorporate quantitative and/or management metrics. Improvements in monitoring methods should provide greater resolution on the current status and trends in the Apalachicola River Gulf sturgeon.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Henry Flowers.
Thesis: Thesis (M.S.)--University of Florida, 2008.
Local: Adviser: Pine, William.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2008
System ID: UFE0024115:00001


This item has the following downloads:


Full Text

PAGE 1

AGE-STRUCTURED POPULATION MODEL FOR EVALUATING GULF STURGEON RECOVERY ON THE APALACHICOLA RIVER, FLORIDA By H. JARED FLOWERS 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 2008 1

PAGE 2

2008 H. Jared Flowers 2

PAGE 3

For Opa and Oma 3

PAGE 4

ACKNOWLEDGMENTS This research was funded by the Florida Fish and Wildlife Commission and National Oceanic and Atmospheric Administration. I would like to thank my advisor, Dr. Bill Pine, for his support, guidance, and patien ce throughout my graduate program. I would also like to thank the other members of my committee, Dr. Mike Allen and Jim Estes for their service and assistance with my research. I also would like to thank Dr. Carl Walters, Dr. Steve Martell, and Dr. Lew Coggins for teaching me much of what I now know about population dynamics and modeling, allowing me to accomplish this project. I thank Frank Parauka and Dr. Ken Sulak for their input and insight coming from years of field observations a nd experiences. I also thank Dan Gwinn and all the members of the Pine and Alle n labs at the University of Florida who have helped me with my project and who have help ed make my experience here a memorable one. Finally Id like to thank my pare nts Hank and Debra, sister Ami, and the rest of my family who have loved and supported me throughout my ro undabout educational career. 4

PAGE 5

TABLE OF CONTENTS page ACKNOWLEDGMENTS...............................................................................................................4 LIST OF TABLES................................................................................................................. ..........7 LIST OF FIGURES.........................................................................................................................8 ABSTRACT.....................................................................................................................................9 CHAPTER 1 INTRODUCTION..................................................................................................................11 Study Site..................................................................................................................... ...........12 Exploitation History and Current Threats...............................................................................13 2 HYPOTHESES AND THE MODELING APPROACH........................................................17 Using a Modeling Approach to Evaluate Management Options............................................18 Population Models in Speci es Conservation-A Review.........................................................19 Study Model............................................................................................................................21 Age-Structured Population Model Case Studies....................................................................23 Model Scenarios................................................................................................................ .....24 3 METHODS.............................................................................................................................28 Model Construction................................................................................................................28 Source Data and Inputs......................................................................................................... ..28 Biological Parameters.......................................................................................................... ...29 Lengthand Weight-At-Age............................................................................................29 Mortality and Vulnerability.............................................................................................30 Fecundity...................................................................................................................... ...32 Skip Spawning.................................................................................................................33 Recruitment.................................................................................................................... .34 Leading Parameters: Initial Population Size and Recruitment Compensation.......................34 Model Initialization........................................................................................................... .....35 4 RESULTS...............................................................................................................................41 Hypothesis 1: Low Post-Harvest Po pulation Size Limiting Recovery...................................41 Scenario 1..................................................................................................................... ...41 Scenario 2..................................................................................................................... ...41 Scenario 3..................................................................................................................... ...42 Scenario 4..................................................................................................................... ...42 Hypothesis 2: Spawning Habitat Limiting Population Recovery...........................................43 5

PAGE 6

Evaluating Model Uncertainty and Parameter Sensitivity......................................................43 5 DISCUSSION.........................................................................................................................55 Hypothesis Evaluation.......................................................................................................... ..55 Management Implications......................................................................................................56 Mortality..........................................................................................................................56 Supplemental Stocking....................................................................................................56 Jim Woodruff Lock-and-Dam Passage...........................................................................58 Recovery Criteria Options......................................................................................................59 Monitoring program and future research................................................................................61 LIST OF REFERENCES...............................................................................................................66 BIOGRAPHICAL SKETCH.........................................................................................................74 6

PAGE 7

LIST OF TABLES Table page 1-1 Harvest history data sources.............................................................................................. 15 2-1 Model scenarios.......................................................................................................... .......26 3-1 Parameter definitions and values.......................................................................................37 7

PAGE 8

LIST OF FIGURES Figure page 1-1 Historic (Year, x-axis) Gulf sturgeon landings (Total Landi ngs (kg), y-axis) in Florida from 1897-1984 (m ultiple sources).......................................................................16 2-1 Study design............................................................................................................... ........27 3-1 Spreadsheet model structure..............................................................................................38 3-2 Skip spawning effects (Sk) on fecundity...................................................................39 3-3 Probability distribution of recK parameter estimat es using method proposed by Martell et al. (2008).......................................................................................................... .40 4-1 Model scenario 1, recovery rates based on the range of N1985 estimates...................45 4-2 Model scenario 2, N1985 population sizes required for recovery by 2023..........................46 4-3 Model scenario 3, the effect of a dditional mortality on population recovery....................47 4-4 Model Scenario 4, effects of enhanc ing Gulf sturgeon stock with hatchery individuals.................................................................................................................... ......48 4-5 Model Scenario 5, population recovery trends based on reduced access to spawning habitat.................................................................................................................................49 4-6 Model sensitivity to a range of recrui tment compensation values (recK, x-axis) as measured by the percent recovery of the population at the 2023 level..............................50 4-7 Model sensitivity to a range of natural mortality values (M, x-axis) as measured by the percent recovery of the population to the 2023 level...............................................51 4-8 Model sensitivity to a range of initial age at maturity Mai values (Mai, x-axis) as measured by the percent recovery of the population to the 2023 level.............52 4-9 Model sensitivity to a range of von Bertalanffy k values ( k x-axis) as measured by the percent recovery of the population to the 2023 level...................................................53 4-10 Model sensitivity to a range skip spawn in terval Sk values (Sk, x-axis) as measured by the percent recovery of th e population to the 2023 level..........................................54 5-1 Surface plot representing theoretical Gulf sturgeon popul ation subjected to harvest.......63 8

PAGE 9

Abstract of Thesis Presen ted to the Graduate School of the University of Florida In Partial Fulfillment of the Requirements for the Degree of Master of Science AGE-STRUCTURED POPULATION MODEL FOR EVALUATING GULF STURGEON RECOVERY ON THE APALACHICOLA RIVER, FLORIDA By H. Jared Flowers December 2008 Chair: William E. Pine, III Major: Fisheries a nd Aquatic Sciences Gulf sturgeon Acipenser oxyrinchus desotoi is a U.S. federally threatened species found throughout the northern Gulf of Me xico. The viability of the Ap alachicola River, Florida Gulf sturgeon population is a concern to fishery managers because of the small population size, slow rate of population recovery, and numerous threats to recovery, including habitat loss and interstate water allocation battles. Although Gulf sturgeon harvest has been illegal since 1985, managers and researchers are concerned that the population is not recoveri ng at the desired rate to meet the recovery criteria objective of delisting by 2023. I developed an age-structured population mode l for the Apalachicola River Gulf sturgeon to describe population recovery and assess the efficacy of proposed management options. These options represent predicted population res ponses to competing hypotheses of why the Apalachicola River Gulf sturgeon po pulation has not recovered to the expected level after twenty years of fishery closure. Hypot hesis 1 proposes Gulf sturgeon populations are not recovering as a result of very low population si ze at the end of fish ing and given life hist ory characteristics of Gulf sturgeon. Hypothesis 2 proposes Gulf stur geon recruitment was low due to lack of access to historic spawning areas limiting population recovery. I evaluated these hypothesis and corollary management options using 5 modeling scenarios. 9

PAGE 10

10 Model evaluation of these scenarios indicated that the Apalachicola River Gulf sturgeon population is greatly reduced from historic, pre-exploitation abunda nce and full population recovery will take in excess of 100 years from time of fishery closure. The observed abundance at the end of exploitation in 1984 was significantly lower than th e level I estimated would allow population recovery by 2023 (target year from USFWS recovery plan), and any increases in total mortality would further slow recovery. The Apalachicola River Gulf sturgeon population recovery is not likely sp awning habitat limited at this time (hypothesis 2). A stock enhancement program may provide short-term population increase s, but the short-term gain in population size may not be worth the risks to the wild stock of initiating a stocking program. A key finding is that Gulf sturgeon recovery will be protracted because of reduced spawning potential due to erosion of the adult age-structure likel y related to historic fishing pracitices. This loss of large, mature, highly fecund individuals within the population greatly reduces annual reproductive output slowing population recovery. To promote population recovery, I recommend no short term changes to current Gulf sturgeon management strategies. However, efforts should be made to minimize any adult mortality from di rect (targeted sampling or fishing) or indirect (by-catc h, incidental take) sources. Recove ry criteria should be revised to incorporate quantitative and/or management me trics. Improvements in monitoring methods should provide greater resolution on the current st atus and trends in the Apalachicola River Gulf sturgeon.

PAGE 11

CHAPTER 1 INTRODUCTION The Gulf of Mexico sturgeon Acipenser oxyrinchus desotoi (Gulf sturgeon) is a subspecies of the Atlantic sturgeon A. oxyrinchus oxyrinchus, historically found throughout much of the northern Gulf of Mexico. Gulf sturgeon were listed as Threatened under the Endangered Species Act (ESA) in 1991 and the current Gulf Sturgeon Re covery Plan (GSRP) outlines a variety of criteria that must be met before Gulf sturgeon populations can be considered recovered and delisting of this species proposed (U.S. Fi sh and Wildlife Service [USFWS] 1995). Management agencies are particularly concerned with determining what factors are currently limiting the recovery of Gulf sturgeon populations. The Apalachicola-ChattahoocheeFlint River basin (ACF) is the largest of the historic Gulf sturgeon drainages and may have contained the largest population of Gulf sturgeon (Wooley and Cr ateau 1985). The Apalachicola River Gulf sturgeon population is of special concern because of ongoing water allocation disputes in the ACF and becau se Jim Woodruff Lock and Dam (JWLD) may reduce access to historical spawning grounds and potentially affect Gulf sturgeon recovery. A short-term goal of the GSRP is to ensure wild stocks are not cu rrently declining as measured by catch-per-unit-effort indices of ab undance (CPUE) associated with standardized gill-net sampling by USFWS and U. S. Geological Survey (USGS) personnel as part of riverine monitoring plans(USFWS 1995). The primary long-term goal is to esta blish self-sustaining population levels that could allow delisting of the species by 2023. A second long-term goal is the recovery of populations to a point at which they could su stain directed fishing (USFWS 1995). Under the Critical Habitat Designation, the Gulf sturgeon population in each of seven river systems represents an individual manage ment unit defined as the (1) Suwannee, (2) Apalachicola, (3) Choctawhatchee, (4) Yellow, and (5) Escambia River populations in Florida 11

PAGE 12

(parts of Alabama); (6) Pascagoula and (7) Pear l River populations in Mi ssissippi and Louisiana (USFWS 2003). Management agencies have expressed concern that Gulf sturgeon in the Apalachicola River are not recovering at a rate that will lead to the species being delisted by the target recovery date of 2023. There is uncertainty whether or not this slow rate of recovery is real or perceived; however, management agencies are proposing management actions to enhance recovery. These actions are prim arily designed to increase the num ber of Gulf sturgeon recruits with the expectation of ultimately increasing the number of adult fish. Proposed management actions include modifying JWLD operations to create passage opportuni ties for adult Gulf sturgeon to potential hi storic spawning grounds and developi ng stocking programs to release large numbers of juvenile Gulf sturgeon in the Apalachicola River (see Chapter 2). Study Site The Apalachicola River is the largest river, by discharge, in Florida. Major tributaries include the Chattahoochee and Flint rivers in Georgia, and the Ch ipola and Brothers Rivers in Florida. The ACF watershed drains an area of 31,375 km in Georgia, Florida, and Alabama (Wooley and Crateau 1985). The Apalachicola River is unique among rivers known to support Gulf sturgeon because the JWLD complex blocks approximately 78% of historic riverine habitat (Wooley and Crateau 1985) within the ACF. Add itionally, the impacts of modified flow regimes on Gulf sturgeon population viability in the Apal achicola River related to dam operations are unknown. Three known and several potential Gulf st urgeon spawning sites have been identified in the upper portion of the Apalach icola River, all of which are within about 40 km downstream of the JWLD (U.S. Army Corps of Engineers [USACE] 2004; Pine et al. 2006; F. Parauka, USFWS, personal communication ). 12

PAGE 13

Exploitation History and Current Threats Gulf sturgeon historically ranged throughout the northern Gulf of Mexico from Tampa Bay, Florida to the Mississippi Ri ver drainage (Wooley and Crat eau 1985). Gulf sturgeon were exploited in Florida from the late 1800s until 1985 (Figure 1-1). The first fishery for Gulf sturgeon began in 1886 in Tampa Bay and ended shortly after in 1890 when only seven were caught (Huff 1975). The fishery then spread we st through the Florida panhandle, reaching the Suwannee River in 1895, Ochlockonee River in 1898, Apalachicola River in 1900, and limited fishing occurred in other smaller rivers, mostly in the Florida panhandl e that supported Gulf sturgeon by 1901. Only a few directed fisherie s for Gulf sturgeon existed in Alabama, Mississippi, and Louisiana (USF WS 1995). Sturgeon meat was the primary product in early Gulf sturgeon fisheries with caviar production becomi ng more important in later years (Huff 1975). On average, the Apalachicola River provided ap proximately 40 percent of the total annual commercial Gulf sturgeon harvest in Fl orida during the first half of the 20th century (various sources [Table 1-1]). In addition, a recreational fishery briefly emerged following the completion of the JWLD in 1957 (Wooley and Cr ateau 1985). By the 1970s and 1980s Gulf sturgeon catch rates had declined to the point we re most remaining fisheries were untenable and take was prohibited in Alabama in 1972, Mississi ppi in 1974, Florida in 1984, and Louisiana in 1990 (USFWS 1995). A similar pattern of high fish ery harvests, followed by a rapid decline in landings, was also observed in ot her sturgeon fisheries, such as Atlantic sturgeon during this same time period (Spear 2007). Although directed Gulf sturgeon fishing has ended, threats to population viability still exist for this species throughout its range. Human activities such as dredging and beach rebuilding can harm Gulf sturgeon or important habitats, such as riverine spawning shoals. Fisheries by-catch, especially in marine habitats, may still remain a significant source of 13

PAGE 14

mortality. Wooley and Crateau (1 985) estimated a 10 percent annual by-catch mortality rate for Apalachicola River Gulf sturgeon, which may be sk ewed towards larger, mature fish. Because of Gulf sturgeon behavior (such as use of benthic habitats) and gear types used in most nearshore fisheries (trawls), Gulf sturgeon, especially mature individuals, are most vulnerable to by-catch harvest during winter months when they are ou t of the rivers and f eeding in estuaries. Occasional incidental mortality incurred through sampling programs may be a factor as well (F. Parauka, USFWS, personal communication ). Emerging water allocation issues within the ACF potentially reducing the amount of upstream water reaching the Apalachicola River also pose a threat to the Apalachicola River Gulf sturge on population. Reduced st reamflow from JWLD could negatively affect Gulf sturgeon popul ation recruitment through the dewatering of spawning sites (Flowers et al. In review ) or decreasing habitat suitability for young-of year individuals (Randall and Sulak 2007; Flowers and Pine 2008). 14

PAGE 15

Table 1-1. Harvest history data sources Source Years of fishery covered U.S. Bureau of Fisheries. 1898, 1900, 1902, 1905, 1917, 1922, 1927, 1931, 1932, 1934, 1936, 1939, 1940. Report of the United States Commissioner of Fisheries 1896, 1897, 1900, 1902, 1908, 1910-1915 (estimates of harvest), 1918, 1923, 1927-1932, 1934, 1936 Florida State Board of Conservation. 1939, 1941, 1943, 1945, 1947, 1949, 1951, 1953, 1955, 1957, 1959, 1961, 1963, 1965, 1967, 1969, 1971, 1973, 1975, 1977, 1979, 1981, 1983, 1985. Biennial Report. 1938-1985 15

PAGE 16

0 30000 60000 90000 120000 15000018 97 19 03 1909 1915 1921 1927 19 3 3 19 3 9 1945 1951 19 57 19 63 19 69 1975 1981YearTotal Landings (kgs.) Figure 1-1. Historic (Year, xaxis) Gulf sturgeon landings (T otal Landings (kg), y-axis) in Florida from 1897-1984 (multiple sources) 16

PAGE 17

CHAPTER 2 HYPOTHESES AND THE MODELING APPROACH Despite the reduction in mortality with the cessation of the Gulf sturgeon fishery, Gulf sturgeon populations in the Apal achicola River have not recovere d. A key challenge for Gulf sturgeon managers is to determine if population r ecovery is limited by human factors (i.e., loss of spawning habitat due to dam construction) that could be mitigated by directing management actions at these recovery impedi ments. If recovery is being hampered by human activities, then given our knowledge of the species life hi story and population ecol ogy, what are realistic expectations of the time until recovery? Curren tly two competing hypotheses as to why Gulf sturgeon populations in the Apalach icola River have not increased at the expected rate following species protection exist: (1) Gulf sturgeon population recovery is hinde red by low population size present at the end of the exploitation period and has not recovered but will given sufficient time; (2) Gulf sturgeon population recovery is lim ited by reduced spawning production caused by restricted access to historic spawning habitat resulting from the construction of JWLD. Regarding hypothesis 1, the recovery rate of a severely depleted fish population is often greatly prolonged when compared to light to mode rately exploited populations due to the erosion in population age-structure often in older, more fecund individuals When these individuals are removed by the fishery more time is required for these highly exploited populations to rebuild their age-structure and related repr oductive capacity (Walters et al. In press ). The effects of degraded age-structure and reduced reproductive potential on the time required for population recovery may be even more pronounced in slow growing and late maturing species such as sturgeons (Paragamian et al. 2005, Walters et al. In press ). 17

PAGE 18

The impact of lost spawning habitat on fi sh population viability (hypothesis 2) is a widespread fisheries management concer n and area of research (Gunn and Sein 2000, Paragamian et al. 2005). Many populations of fish and other fauna, especia lly riverine migratory fish species, have been impacted by dams restri cting access to historic spawning and feeding areas (Freeman et al. 2003). Gu lf sturgeon in the Apalachicola River are currently limited to approximately 22 percent of the historic ACF basi n (Wooley and Crateau 1985). However, it is unknown how much of the basin was actual historic spawning habitat for Gulf sturgeon. Three spawning locations have been id entified for Gulf sturgeon below the JWLD, and water flow at these sites has likely been altered by JWLD. Using a Modeling Approach to Evaluate Management Options A model is a scientific tool that allows the synthesis of data and identification of important parameters and features of a system (Hilborn and Mangel 1997). Modeling exercises are valuable not for their ability to make precise predictions but for providing broad scenarios against which hypotheses and ideas can be test ed (Walters 1986). Models can evaluate hypotheses through their ability to simulate existi ng data and provide predictions of additional characteristics of a system (Hilborn and Mange l 1997). Previous Gulf sturgeon studies have focused on growth, reproduction, movement, a nd feeding throughout their range (Wooley and Crateau 1985, Sulak and Clugston 1998, Fox et al. 2000) Life history traits such as these should be used in population models to gain an understanding of how populations may respond to management actions or disturbances (Music k 1999; Dulvy et al. 2003). In this study I synthesized the life history information from th ese previous studies and applied principles of ecology and population dynamics in a modeling fr amework to evaluate the hypotheses presented above. 18

PAGE 19

The GSRP suggested the use of population mode ls to assess restoration and management options for Gulf sturgeon, identify future re search needs, and forecast time to population recovery (USFWS 1995). Model parameter inputs were readily derived from available literature and data on the Apalachicola River Gulf stur geon population. Populati on models have been created for Gulf sturgeon in the Apalachicola (Zehfuss et al. 1999), Pearl (Morrow et al. 1999), Suwannee (Pine et al. 2001), and Yellow Rivers (B erg et al. 2007), but most of these previous models focused on assessing population size, popul ation growth, or monitoring program design. Unlike these previous Gulf sturgeon models, the model in this study was specifically designed to examine what factors may be limiting population recovery by screening policy options proposed by managers. The basic outline of this study is diagrammed in Figure 2-1. Population Models in Speci es Conservation-A Review Models are useful tools for assessing anim al populations and policy options, and have greatly improved with (1) advances in model design, including statisti cal and mathematical approaches (2) better incorpor ation of biological data into these approaches, (3) improved understanding of the most informative usage of models within a management framework, and (4) advances in software and computing power ove r time (Walters and Martell 2004). Computer models have been used in fisheries and w ildlife management since the 1960s to make predictions about population responses to mana gement actions (i.e., Walters 1969) such as evaluating fisheries yield and sustainability to different harvest policie s (Walters and Martell 2004). Because of improvements in computing th ere has been a tendency for models to become more sophisticated and complex over time. Howe ver, more complex models do not necessarily provide better predictions and car e should be taken to use a mode l with a balance of simplicity and accuracy (Walters and Martell 2004). Today a variety of modeling programs are widely available, many tailored to specific population model structures or analysis of specific data types. 19

PAGE 20

Population models used in fisheries a nd wildlife conservation are commonly demographically based (i.e., age or stage based) meaning they are based on rates of change occurring in life events (i.e., mortality, immigration/emigration, reproduction, etc.) between and during different stages. These stag es can be based on the age, length, or weight of an animal, or the life history progression of an individual (i.e., juvenile, immatu re, adult stages). Models can be based on the individual animals, population/ groups of animals, or a combination of both, depending on the situation or requirements of a st udy. The structure of a model is important and may have implications in the predictions it makes (Pascual et al. 1997). Depending on the individual study, models can be designed to project forward to simulated populations into the future, or they can be designed to project back ward either from some point into the past. Two main classes of population models are commonly used in fisheries and wildlife conservation stock assessment models (origina lly designed to aid in harvest planning to maximize yields) and population viability analysis (PVA) models (originally designed for conservation evaluation). These models can use similar structures, parameters, and have similar modeling utility data synthesi s and identifying life stages that should be targeted for management activities (Morris et al. 2002, Walte rs and Martell 2004), however the type of predictions these model classes make are different The PVA models provide estimates of future extinction risks over varying peri ods of time and because of this are often used to simulate populations of conservation concer n (Morris et al. 2002). Stock a ssessment models also provide simulations of future population abundance and bi omass, which could incl ude extinction as an outcome, but do not expressly prov ide probabilities of extinction. The decision by a researcher as to which class of model to use, a PVA or stock assessment, should depend largely on the questions being addressed by the study. 20

PAGE 21

One application of stock assessment models for non harvested species is the assessment of permissible take under the ESA. The ESA can allow for limited take of listed species for specific purposes given that th is take will not decrease the li kelihood of population recovery. Permissible take guidelines ar e often developed using models however, PVA models may not include management indices that can be used to establish these guide lines. Even though numerous PVA models exist for marine mammals and sea turtles, separate dedicated take models have been developed to evaluate allo wable harvest levels fo r these species (Heppell 2005). Both PVA and stock assessment models can be custom programmed, but several existing programs are available and widely used by ma nagers. Two example PVA model programs are VORTEX (Chicago Zoological Society, Brookfiel d, Illinois, 2008), an age-structured PVA model that takes into acc ount genetic population dynamics factors and RAMAS (Applied Biomathematics, Setauket, NY, 2008), an age/stage-structured PVA m odel with different versions capable of including spatial and mu ltispecies data. The VORTEX model was recently used in an assessment of endangered mussel popul ations in the ACF basi n (USFWS 2008) while RAMAS is being used by Gulf sturgeon managers in systems in the western Gulf of Mexico (J.P. Kirk, USACE, Vicksburg, MS, personal communication). An example of an age-structured stock assessment population modeling program is MOCPOP (Beamesderfer 1991), which has been used to assess a variety of sturgeon populations (Rieman and Beamesderfer 1990; Morrow et al. 1999; Pine et al. 2001). Study Model Hypotheses in this study were related to population recove ry rates and not extinction risks so I decided that a stock assessment model would be more appropriate than a PVA model. The model I used in this study is designed to be a simple, yet powerful tool for resource 21

PAGE 22

managers to use to evaluate recovery times in populations of concern. The design of the model is similar to MOCPOP; however, it overcomes some of the limita tions found in MOCPOP and is flexible enough to be adapted to individual situations (e.g., specific policy scenarios for individual sturgeon stocks). The MOCPOP model is limited in its ability to model separate groups within a population and describe uncerta inty around parameter and simulation estimates (Beamesderfer 1991, Paragamian et al. 2005). It is also nearly obsolescent, having last been updated in 1991 and its MS-DOS based interface is not user friendly. I constructed the model for this study in th e readily available program Microsoft Excel (Microsoft Corporation, Re dmond, WA 2003), which has advant ages over existing canned models. By design, spreadsheet models are open and transparent which allows users to easily see the relationships among different model components so that they can better understand the behavior of different aspects with in the model. Excel also makes it very easy for the model to be modified based on specific situations allowing the model to be readily updateable with new parameter data. This model is also inexpensiv e, an advantage over some of the other models which must be purchased (e.g., RAMAS Me tapop 5.0 licenses range from $595-1495). There are a few disadvantages of buildi ng a model like this in Excel. Because spreadsheets rely on functions bui lt into individual cells, increasing model capability may greatly increase model complexity to a point wher e the model may be to unwieldy within the spreadsheet format. In this case models may be better constructed throug h a coded language in a statistical program such as SA S (SAS Institute Inc., Cary, NC 2008), R (R-Project 2008), or ADMB (Otter Research Ltd. 2008), which can conde nse large numbers of repeating functions into simpler code. Excel has been criticized for flaws in its statistic al operations (McCullough and Wilson 1999, 2002), but these statis tical operations are not used in this model and therefore 22

PAGE 23

not a concern. If so desired, this basic model c ould be constructed in any available spreadsheet or statistical software pack age, like those mentioned above. Because it is stock assessment based, the m odel in this study was already well adapted to estimate allowable harvest levels and its popul ation effects. Commonly used management metrics, such as fishing mortality at maximum sustain yield ( Fmsy) and spawning potential ratio ( SPR ; [annual egg production under ha rvest]/[annual egg production of an unexploited stock]) (Goodyear 1993), can be readily calculated us ing the model without significant parameter additions. These guidelines are commonly as part of fishery management plans for harvested species but are not widely used for species of co ncern. In conservation these metrics can be used to establish targets, such as maximum allo wable total mortality based on estimates of Fmsy plus natural mortality. Gulf sturgeon populations coul d be considered recovered when they reach an SPR value of >0.30, a level considered healt hy for exploited stocks (Goodyear 1993). Age-Structured Population Model Case Studies Age-structured population models like the one used in this study have been used for a wide variety of species in a range of situati ons to evaluate management scenarios. Agestructured population models were used to exam ine recruitment and recovery characteristics of white sturgeon A. transmontanus in Kootenai Lake (Paragamian et al. 2005). This study used a spreadsheet-based, assessment type model simila r to MOCPOP to assess factors affecting the recovery of this population. Included in these factors were the effects of fishing mortality, population augmentation using hatchery fish, and a reduction in the success of natural recruitment as a result of altere d hydrologic conditions resulting from impoundments. Kareiva et al. (2000) and Wilson (2003) used similar agestructured population models to evaluate management activities on threat ened Snake River chinook salmon Oncorhynchus tshawytsch populations. These activities include d habitat restoration efforts to increase juvenile survival and 23

PAGE 24

dam passage activities to allow both adults access to upstream spawning sites and smolts downstream passage to rearing areas. In this situation, a mult i-agency group responsible for management of these salmon stocks discussed and suggested the scenar ios modeled in these studies (Wilson 2003). Age-structure assessment models have been us ed to evaluate management consequences on a variety of non-fish species. Milner-Gulland (1994) used agestructured modeling techniques to evaluate the effects of different harvest scenarios on Saiga antelope Saiga tatarica in Central Asia. These included different sexually, temporally, and climatically ba sed harvest strategies. Gerber et al. (2004) used age-structured mode ling techniques to assess mortality sources in southern sea otter Enhydra lutris populations and eval uated the effectiveness of strategies to reduce mortality. Age-structured m odels were also used to evalua te harvest effects on green sea turtles Chelonia mydas in Australia (Chaloupka 2002). A st age-structured model was used to evaluate conservation scenarios for a Canadi an endangered species, the spotted turtle Clemmys guttata (Enneson and Litzgus 2008). This study eval uated the potential effects of several activities designed to increase juvenile survival in order to increase population growth rates. Model Scenarios To test hypotheses using the model, five te st scenarios were created to evaluate the effects on the simulated population in which certain parameters were adjusted for each situation (Table 2-1). The condition of low post-exploitation population sl owing recovery was tested by evaluating post harvest population recovery. I evaluated the recovery effect of varying population abundance at the end of the fishery in 1985 ( N1985) (Scenario 1) and estimated the value of N1985 that could produce a fully recovered population in 2023 (Scenario 2). I also examined the effects of added mortality from sampling, management practices, and/or fisheries by-catch (Scenario 3), and supplemental stoc king effects on the population (Scenario 4). 24

PAGE 25

Scenario 5 assessed effects of decreased spaw ning production, representing spawning habitat loss, by capping maximum egg production at various points below the historic maximum. The goal was to test if reduced egg production is impacting Gulf stur geon recovery. Descriptions of each of the five scenarios can be found in Table 2-1. 25

PAGE 26

Table 2-1. Model scenarios Scenario Hypothesis Tested Description Base N1985= 282, F = 0, Stocking= 0 individuals, Egg production capacity= 100% 1 1 Variable N1985= 181,282,645 2 1 Test for N1985 that provides 90, 95, and 99% of No at 2023 3 1 Test for F effects on recovery, F = 0.01, 0.05, 0.1 4 1 Stocking effects on recovery: 500, 2500, 5000 individuals for 5 years (2010-2014), 2500 for 5 years close to fishery end (1990-1994), 2500 20 years (2010-2029) 5 2 Cap population egg production at different levels: 50, 25, 12.5% 26

PAGE 27

27 Inputs L No k N1985 recK Wmat M Sk F Vuln Fec Amax Hypotheses Gulf sturgeon recovery is spawning habitat limited. Gulf sturgeon recovery is limited by low population size at the end of ex p loitation. AgeStructured Model Outputs Projected recovery time and trends under the different hypotheses. Test Scenarios Variation of N1985 Variation of F Stocking regimes Variation of spawning limitation Additional Data Needs Spawner-recruit relationship Environmental effects Introduced species effects Figure 2-1. Study design.

PAGE 28

CHAPTER 3 METHODS Model Construction My population model for assessing Apalachicol a River Gulf sturgeon consists of three tables (Figure 3-1): the first features attributeat-age information on a pe r-recruit basis for wild born individuals, while the sec ond contains attributeat-age information for hatchery stocked individuals. The third is derive d from the first two tables and simulates the actual population by adding numbers of individuals-at-age per year. Each table is contained as a page within an Excel spreadsheet. Population numbers-at-age in a ny given year was calcula ted using the function: ))((1)t1,( ata, aSN N (3-1) where a is age, t is time and Sa is age-specific survival. Other parameters included in the model and used in assessing population responses to management actions include natural mortality ( M ), anthropogenic mortality ( F ), fecundity ( f ), and vulnerability-at-age ( v), initial population size (No), and skip spawning effects ( Sk ). A list of all parameters and their values used in this study is provided in Table 3-1. Source Data and Inputs Model inputs were extracted from historic data and from othe r studies throughout the Gulf of Mexico (Table 3-1). As part of ESA mandated GSRP for this species (USFWS 1995), information on Gulf sturgeon life history, particul arly the location, timing and characteristics of spawning habitats, have been collected for se veral Gulf sturgeon populations (Marchant and Shutters 1996; Sulak and Clugston 1998; Fox et al. 2000). The age-attribute table for wild fish was developed utilizing data from an ongoing tagging and recaptur e effort for Gulf sturgeon in the Apalachicola River that has been conducte d regularly since 1977 by the U.S. Fish and Wildlife Service. This tagging study was not co ntinuous, but has occurred with regularity over 28

PAGE 29

this time period with similar methods and proj ect personnel throughout the entire time period. Zehfuss et al. (1999) used this same dataset to assess population status and evaluate sampling programs for Gulf sturgeon in the Apalachicola, and I updated this dataset to include over 1,800 captures of Gulf sturgeon up to the year 2007. The record for each individual Gulf sturgeon included date, location of capture, gear type, tota l length (TL), fork length (FL), weight, tag type and number, and whether or not the specimen wa s a recapture. Also, observed mortalities resulting from sampling and sex information, if available, were recorded. Some length-at-age data based on pectoral fin rays was available from L. Jenkins (USFWS, unpublished). Biological Parameters Lengthand Weight-At-Age The length-at-age relationship for Gulf stur geon is a very important component of my population model because this relati onship is used to directly or indirectly define other model parameters such as weight-at-age, sampling vulne rability, mortality, and fecundity. For managed fish species, length-at-age curves are commonly es timated by collecting a wi de size range of fish and then estimating ages of a sub-set of these fish using co mmon aging techniques (such as scales, otoliths, or fin-rays [Devries and Frie 1996]). Because Gulf sturgeon are a threatened species, lethally sampling fish for age analyses is not permitted. Common non-destructive aging methods for sturgeon using pectoral fin rays or scutes have been shown to be unreliable at older ages compared to otoliths (Rien and Beames derfer 1994; Rossiter 1995). Huff (1975), Jenkins (USFWS, unpublished data), and Su lak et al. (2002) all contain information on Gulf sturgeon age and growth. Alternatively, the use of incrementa l growth data has been suggested as a way to estimate growth model parameters (Fabens 1965). Using incremental growth data from the existing Gulf sturgeon tagging study to estimate growth has the advantag es of not requiring direct age estimates of individuals important in a case such as this for Gulf sturgeon where lethal 29

PAGE 30

aging methods (otoliths) are impracticable and alternatives (fin rays) are inaccurate (Coggins 2007). I used length-at-age information from th e Apalachicola River in Jenkins (USFWS, unpublished data) combined with Gulf st urgeon tagging data from 1978-2007 in the Apalachicola River to construct a growth curve using a single likelihood framework (C. Walters, University of British Columbia; T. E ssington, University of Washington; personal communication ). The advantages of combining direct aging and tagging data are that the sample size of individuals used for aging is increased, potentially including a better representation of the population, and that the effect of age-assignment errors is diluted because of this additional data. The output from this method was then reparamete rized into a simplified von Bertalanffy growth curve (Ricker 1975), recommended by Johnson et al. (2005) for st urgeon species, )e(1ka aLL (3-2) where L is the asymptotic length parameter, k is the Brody growth parameter, and La is lengthat-age. This simplified formulation of th e von Bertalanffy assumes that the variable to (age at length zero) is zero and eliminates problems arisi ng from limited size-structure representation in the data set (few very young or ol d individuals) that could lead to biologically unrealistic estimates of to (Johnson et al. 2005). A single-growth model was used for both sexes combined because of a lack of published support for sexually dimorphic growth (Johnson et al. 2005). Mortality and Vulnerability Natural mortality is the aver age annual rate at which in dividuals exit the population by death. Estimates of M are commonly acquired from tagging st udies (such as Sulak and Clugston [1999] for Gulf sturgeon) or ca lculated from population dynamics aspects such as longevity (Hewitt and Hoenig 2005) or individual growth rate. Jensen (1996) proposed using the von 30

PAGE 31

Bertalanffy k parameter to estimate overall M where 1.5 k=M. However, I used alternative method recently proposed where: M = k (3-3) (C. Walters, University of British Columbia, personal communication). Evidence for this M =k relationship was found in a review of fish species on Fishbase (www.fishbase.org, 2008) where estimates of both M and k were available. Correlation analysis between these two parameters found a 1:1 relationship such that M and k were proxies. Annual M was variable at age and dependent on overall length, simulated here by Lorenzens (2000) method to predict age specific mortality: a amaxL L *MMa. (3-4) where Ma is mortalityat-age and Mamax is mortality-at-maximum age. Mamax will be solved for as the value which yields an overall M=k averaged across all ages. Because Gulf sturgeon are protected from harv est, F was not a significant factor to the population in its current state and natural mortality represented most of total mortality (Z). However, F was used to simulate harvest that occurred during the Gulf sturgeon fishery and reduce pre-exploitation abundance to that observed at the end of fishing in 1985. This F parameter was also used in scenario 2 to simula te the effects of genera l anthropogenic mortality resulting from human sources in cluding sampling programs, habita t modification operations (i.e. dredging, beach rebuilding), and by-catch in commercial fishing operations. The F was not evenly distributed across all ages and was applied to the population with a vulnerability schedule that determined what proportion of the popul ation was removed through harvest. Field observations of Gulf sturgeon netting activities suggested a dome -shaped vulnerability schedule, where Gulf sturgeon first recruit to the gear at around 450 mm (Huff 1975) after which the oldest 31

PAGE 32

individuals were less vulner able than younger mature indi viduals (F. Parauka, USFWS, personal communication). A double logistic vulnerability function was used (adapted from Martell et al. 2008): )) (-( (-( 50h 50%lL ov50a L yv50A a L-L L-L v)/ )/exp+1 1 ) exp+1 1 (3-5) where va is vulnerability-at-age, Lyv50 is length at 50% vul nerability, young age, Lyv50 is standard deviation length at 50% vulnerability, young age, Lov50 is length at 50% vulnerability, old age and Lov50 is standard deviation length at 50% vulnerability, old age. Fecundity Fecundity is the mean egg production for individual Gulf sturgeon, which is zero until an individual matures, after which it increases appr oximately linearly to fish weight (Walters and Martell 2004). Gulf sturgeon maturation occurs between ages 8-12 for females and 7-10 for males (Huff 1975). Fecundity is important because it determines the poten tial number of recruits that an individual can produce. Gulf sturgeon, like other Acipenseri ds, are highly fecund and produce large numbers of eggs at spawning. Chapman et al. (1993) estimated there were 20,652 Gulf sturgeon eggs/kg eggs, whereas Huff ( 1975) found gonad weight represented 12.7% of a ripe female Gulf sturgeons overall body wei ght. This means a single 75 kg female could produce close to 200,000 eggs. However, Gulf sturgeon mortality from age-0 and age-1 is extremely high, estimated between 99.9-100.0% (Pine et al. 2001), meaning large numbers of eggs do not produce large numbers of age-1 recruits. In this model, fecundity at a given age was approximated by: mat aafWW (3-6) where fa is fecundity-at-age, Wa is weight-at-age, and Wmat is the weight at the age of maturity. 32

PAGE 33

Maturation follows a knife-edge schedule in this fecundity function, meaning all individuals mature at the same age. Skip Spawning An important aspect of Gulf sturgeon life history is skip spawning, where individuals may not spawn every year. In any given year the spawning population is less than the total population (Sulak and Clugston 1998), but how much lower depends on the periodicity of the skip spawning events (Jorgensen et al. 2005). Fe male Gulf sturgeon likely spawn at intervals ranging from every 3-5 years, and males every 15 years (Smith 1985; Fox et al. 2000). Because of skip spawning and the late age-at-maturity, female Gulf sturgeon may only spawn a few times during their life (Sulak and Ra ndall 2002). This relationship between skip spawning and fecundity level is not unique to sturgeon and has been documented in other species as a life history adaptation based in energy availability and allocation (Jor gensen et al. 2005, Rideout et al. 2005). In incorporated skip spawning in the recruitment and fecundity aspects of my model using a modified Ricker curve (Figure 3-2): where aL L i ih ih r iSk Maa MaMa MaMaMa MaaSpP^1 1 exp* 11 *e* (3-7) with P(Sp) as the probability of an indi vidual spawning in a given year, Mai is the initial maturation age, Mar is the number of years needed for entire population to mature, Mah is the age of 50% of population maturation, and Sk is the average skip spawn interval (years) of fully mature individuals. I used this equation to e xplore the population level e ffects of skip spawning by assessing whether skip spawning significantly altered the number of yearly recruits and ultimately recovery trajectory. 33

PAGE 34

Recruitment Population recruitment in the model was si mulated using yearly estimates of population egg production controlled by a density dependent recruitment relationship. A Beverton-Holt (Beverton and Holt 1957) recruitment relationship wa s used in this study and in Pine et al. (2001), although there is little data available on the actual spawner-recruit relationship exhibited by Gulf sturgeon populations. The stock recruit relationship followed the form: b a R 1 (3-8) where R is annual recruits to age-1, a and b are stock recruitment parameters, and = annual population egg production. Leading Parameters: Initial Population Size and Recruitment Compensation Two parameters in my model, the Goodyear compensation ratio ( recK Goodyear 1977, 1980) and the initial population size prior to fishing ( No), were leading parameters selected input parameters estimated by the other input pa rameters through optimizing model fits (Hilborn and Walters 1992). The Goodyear compensation ratio is defined as the ratio of juvenile survival rate at low stock sizes relative to juvenile su rvival in the unexploited condition, representing the recruitment compensation potential of the population. The recK parameter is used to describe population recruitment response to depletion. Higher recK values imply populations that have higher compensatory juvenile survival at lo w population sizes relative to unexploited population sizes. Conversely, populations with lower recK values have a lower compensatory increase in juvenile survival at low populati on sizes. This implies that high recK populations are more resilient to exploitation than populations with low recK values because they have a stronger compensatory response when depleted (W alters and Martell 2004). I estimated recK following the approach in Martell et al (2008) which used the manage ment parameters of maximum 34

PAGE 35

sustained yield ( MSY ) and the exploitation rate needed to achieve this yield ( FMSY) (Pine et al. 2008). This approach provides probabil ity density estimates for a range of recK parameter values (Figure 3-3). The recK value used in the model was 4, which is relatively low across species, and was similar to the value used by Walters et al. (2006) for white sturgeon. The No parameter was estimated using result s from a Stock Reduction Analysis (SRA) performed on the Florida Gulf sturgeon populatio n in a separate ongoing project (Pine et al. 2008). This analysis simulate d historical pre-exploitati on Gulf sturgeon abundance and population structure in Florida waters and examin ed the effects of harv est through time on Gulf sturgeon populations. Unexplo ited Apalachicola River Gulf sturgeon abundance was based on proportion of total historic catch during peak y ears of harvest (1900-1925). The estimate of total catch proportion was similar to th e proportion of ACF basin area to total basin area for all Gulf sturgeon river systems. These unexploited abundance estimates were useful guidelines for initializing the model; however, precise estimat es were not required for model operation or predictions. Within the model, recK and No could be confounded where population characteristics can be expl ained equally well by a high recK and low No or a low recK and high N0. Model Initialization After development, the model was initialized with parameter estimates developed in the SRA model (Table 3-1) to repres ent the initial, preexploitation population of Gulf sturgeon in the Apalachicola River. The population then underwent simulated harvest for 25 years until abundance was at the level observed at the end of harvest in 1985 (Wooley and Crateau 1985). From this point the model was run for 100 years (until 2084), tuning the model to observed data on population size or size structure when possible, through 2023 (tar get year for recovery) plus 35

PAGE 36

the additional years for recovery assessment. Pa rameters and the model scenarios (Chapter 4) were evaluated using this general framework. 36

PAGE 37

Table 3-1. Parameter definitions and values Parameter Description Value Source F Anthropogenic mortality (fishing, etc.) variable Fmsy Fishing mortality at maximum sustained yield 0.05-0.08 Pine et al. 2008 K Brody growth parameter 0.13 Tagging data 1978-2006 L Von Bertalanffy asymptotic length parameter 220 cm Tagging data 1978-2006 M Annual natural mortality, overall wild = 0.13 stocked = 0.14 Tagging data 1978-2006 N Population size in a given year estimated No Initial pre-exploitation population size ~24,000 Pine et al. 2008 N1985 Population size at end of harvest 282 (181-645) Wooley and Crateau 1985 recK Goodyear recruitment compensation parameter 4 Tagging data 1978-2006, Martell et al. 2008 Sk Factor to adjust skip spawning effects 1-5 year intervals Sulak and Clugston 1999, Pine et al. 2001 Wmat Weight at maturity 10.8 kg Huff 1975, Tagging data 1978-2006 Mai 1st age at maturity 6 Huff 1975 v Vulnerability-at-age variable-at-age Tagging data 1978-2006 F. Parauka, personal communication Z Total mortality variable 37

PAGE 38

Parameters -Growth -Mortality -Fecundity -Population -Perturbations Attributes at Age Population 1 -Length-at-age -Weight-at-age -Fecundity-at-age -Survival-at-age Population -N-at-age per year -Total N per year -Total B per year -Average individual weight -Total eggs produced Outputs -Predicted future N -Predicted future B -Predicted future average individual weight -SPR -Fmsy Attributes at Age Population 2 -Length-at-age -Weight-at-age -Fecundity-at-age -Survival-at-age Figure 3-1. Spreadsheet model structure 38

PAGE 39

0.0 10.0 20.0 30.0 40.0 50.0 60.0 1471013161922252831343740434649AgeFecundity0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0Probability of Spawnin g Fec AdjFec Sk Figure 3-2. Skip spawning effects ( Sk ) on fecundity. Fecundity va lue on primary y-axis, annual probability of spawning on secondary y-axis. AdjFec represents skip-spawning adjusted fecundity-at-age, the product of fecundity -at-age and annual spawning probability. Annual population fecundity is lower because only a proportion of individuals are spawning at a given time. 39

PAGE 40

Figure 3-3. Probabil ity distribution of recK parameter estimates using method proposed by Martell et al. (2008). 40

PAGE 41

CHAPTER 4 RESULTS Hypothesis 1: Low Post-Harvest Po pulation Size Limiting Recovery Scenarios 1-4 were designed to examine the time until recovery of Apalachicola River Gulf sturgeon population by estimating recovery time for this populat ion using cumulative knowledge of Gulf sturgeon population ecology Hypothesis 1 states that the Apalachicola River Gulf sturgeon population has not recove red despite over twenty years of protection due to extremely low population biomass at the conclusion of the commercial fishery. Scenario 1 In this scenario I evalua ted the effects of varying N1985 on population recovery times. Gulf sturgeon population abundance at th e end of harvest had a significant effect on the recovery rate of the population, with higher N1985 values resulting in shorter recovery time than low N1985 values. Using the range of N1985 from Wooley and Crateau (1985) abundance at the recovery goal date of 2023 ranges by a factor of nearly two depending on whether the N1985 lower bound ( = 191) or upper bound ( = 645) is used for the initial popul ation abundance. Extending the time interval of recovery until 2084 reduces the effect of differe nt initial population starting values as populations are expected to reach 94 -97% of the pre-exploitation levels by this time (Figure 4-1). The effects on the recovery rate of the population at shorter time intervals are driven by life history attributes such as late maturity combined with the impact on the reproductive potential of the populat ion at low population sizes and the absence of large, older fish in the population (discussed further in Chapter 5). N N Scenario 2 In this scenario I examine a range of values of N1985 required to allow the Apalachicola River Gulf sturgeon population recovery to >90% of its unexploited abundance by 2023 41

PAGE 42

(USFWS 1985). Given currently available best biological input parameters, to meet this goal an N1985 population size must have been 49% of the unexploited population (F igure 4-2). For 95 and 99% recovery levels in 2023, N1985 must have been 69% and 89%. Values of N1985 in this range are >24 times actual estimates (Wooley and Carteau 1985). Scenario 3 This scenario examined how increasing Z after the closure of the directed fishery altered time till recovery. The increase in Z was applied using the an thropogenic mortality term F simulating the effects of non-specific human-induced mortality. Increased total mortality had a strong negative effect on populati on recovery. With an annual F of 1% (total mortality of 14%), population size at the recovery goal years of 2023 and in 2084 wa s estimated at 27 and 84% of the historic population size, resp ectively (Figure 4-3). With F values of 5 and 10% (total mortality 18 and 23%), these recovery values become 15 and 7% in 2023 and 45 and 14% in 2084, respectively. Scenario 4 Several different enhancement regimes were used to simulate potential effects of stocking age-1 individuals on population r ecovery in this scenario. These regimes were designed to simulate short-term conservation stocking progr ams where stocking is intended to stabilize and give the targeted population a boos t at low abundance levels to aid recovery, as opposed to a more traditional, long term stocking program used for some sport and commercial fisheries. The effect of different stocking rates, time of stocking, and duration of stocking schemes were varied. My results show the effects of enhancing Gu lf sturgeon populations th rough stocking programs may have mixed success in decreasing population time to full recovery. Results show that a supplemental stocking program would have the greatest benefits on th e recovery rate of Apalachicola River Gulf sturgeon if stocking programs had been initiated when the population 42

PAGE 43

biomass (and population fecundity) was at its lowest point. For example, a stocking program of 2,500 age-1 fish released for 10 y ears after the end of the commer cial fishery in 1984 would have led to population recovery of approximately 46% by 2023 vs. 38% recovery if the same stocking program were initiated in 2010 (F igure 4-4). If stocking rates for this program starting in 2010 were doubled (5,000 age-1 individuals) the popula tion would have recovered to 44% by 2023. Doubling the number of years, from 10 to 20, stocking 2,500 age-1 individuals starting in 2010 does not significantly affect r ecovery at 2084 (96 vs. 97%). Hypothesis 2: Spawning Habitat Limiting Population Recovery This scenario evaluated the effects of lim ited spawning area availability on population growth. My results show that effects on popul ation recovery from reducing spawning habitat would not be significant for around 20 years after th e closure of the fishery, but thereafter could have large impacts. Reducing spawning habi tat availability by be tween 13 and 50% would reduce recovery by similar fracti ons throughout time. Simulating 50% of historic spawning area availability, the population would recover to 22 and 48% at 2023 and 2084, while 12.5% availability of historic spawning area results in recovery of 10% in 2023 and 13% in 2084 (Figure 4-5). Reducing availabl e spawning habitat area has the poten tial to severely affect Gulf sturgeon population recovery. Evaluating Model Uncertainty and Parameter Sensitivity I evaluated uncertainty and model se nsitivity around the leading parameter recK when constructing this model. The range of recK values (Figure 3-3) described by the Martell et al. (2008) method were input into the model to te st the simulated populat ion recovery response (Figure 4-6). This response was evaluated by examining estimates of population abundance at the year 2023 relative to pre-ex ploitation population abundance. Pre-exploitation abundance was estimated using the parameter values listed in Table 3-1. Higher recK values required high 43

PAGE 44

exploitation rates to fish the population down and allowed the population to rebound much faster. Field data suggest recK is low because of the relativel y low sustainable catch observed during the later years of the fisher y and the slow recovery rate after harvest ended (Figure 1-1). Model sensitivity analyses were also performed for the model parameters M Mai, k and Sk Similar to the analysis described for the recK parameter, sensitivity was evaluated by examining the effect ranges of parameter values had on estimates of popul ation abundance at the year 2023 relative to pre-expl oitation population abundance estimat es derived from parameter values in Table 3-1. Increasing M has the effect of reducing popul ation recovery linearly, such that for the value of M =1.5k time to recovery is almost halved (Figure 4-7). Recovery is faster because the population age-structure contains fewe r old fish that have to be replaced during recovery. Increasing Mai linearly increased population recove ry time slightly (Figure 4-8), longer time to maturity slightly reducing overa ll reproductive output by removing fecundity contributions of younger ages. Increasing k increased recovery time (Figure 4-9), by increasing time for individuals to reach terminal length, indi rectly decreasing weightand fecundity-at-age. Because individuals were smaller longer, more time was spent at smaller, less fecund ages and total reproductive potential of the population was lower. Adjusting Sk had no appreciable affect on recovery time (Figure 4-10). Within the model it appeared that skip-s pawning reductions in annual egg production were offset by increased survival of eggs to age-1 due to density dependence factors within the re cruitment relationship. 44

PAGE 45

0 20 40 60 80 100 Unex p l oi t e d | | End of Ha r vest 1992 2000 2008 2016 20 2 4 2032 2040 2048 2056 20 6 4 2072 2080YearPercent of Original N0 2000 4000 6000 8000 10000 12000 14000 N1985=181 N1985=282 N1985=645 Figure 4-1. Model scenario 1, rec overy rates based on the range of N1985 estimates. Percent of original pre-exploitation populati on on the y-axis, year of simulation on the x-axis. 45

PAGE 46

0 20 40 60 80 100 Unex p l oi t e d | | End of Ha r vest 1992 2000 2008 2016 20 2 4 2032 2040 2048 2056 20 6 4 2072 2080YearPercent of Original N0 2000 4000 6000 8000 10000 12000 14000 Baseline N1985=99% Recovery N1985=95% Recovery N1985=90% Recovery Figure 4-2. Model scenario 2, N1985 population sizes required for recovery by 2023. Percent of original pre-exploitation populati on on the y-axis, year of simulation on the x-axis. 46

PAGE 47

0 20 40 60 80 100 Unex p l oit ed | | End of Harvest 1992 2000 2008 2016 2024 2032 2 0 40 2048 2056 2064 2072 2 0 80YearPercent of Original N0 2000 4000 6000 8000 10000 12000 14000 Baseline, F=0 F=0.01 F=0.05 F=0.1 Figure 4-3. Model scenario 3, th e effect of additional mortality on population recovery. Percent of original pre-exploitation population on the y-ax is, year of simulation on the x-axis. 47

PAGE 48

0 20 40 60 80 100 Unexploi t ed | | End of Harv e s t 1992 2000 2008 2016 2024 203 2 204 0 2 0 4 8 2 056 2064 2072 2080YearPercent of Original N0 2000 4000 6000 8000 10000 12000 14000 Baseline, No stocking Stock 5000, 2010-2014 Stock 2500, 2010-2014 Stock 500, 2010-2014 Stock 2500, 2010-2029 Stock 2500, 1990-1994 Figure 4-4. Model Scenario 4, effects of enhancing Gulf sturgeon stock with hatchery individuals. Percent of or iginal pre-exploitation populat ion on the y-axis, year of simulation on the x-axis. 48

PAGE 49

0 20 40 60 80 100 Unexploi t ed | | End of Harves t 1992 2000 2008 2016 2024 203 2 204 0 2 0 4 8 2 056 2064 2072 2080YearPercent of Original N500 1500 3500 5500 7500 9500 11500 13500 15500 Baseline Spawning Habitat 1/2 Spawning Habitat 1/4 Spawning Habitat 1/8 Figure 4-5. Model Scenario 5, population recove ry trends based on reduced access to spawning habitat. Percent of original pre-expl oitation population on the y-axis, year of simulation on the x-axis. 49

PAGE 50

0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 0 5 10 15 20 25 30recK% Pre-exploitation abundance at 2023 Figure 4-6. Model sensitivity to a range of recruitment compensation values ( recK x-axis) as measured by the percent recovery of the population at the 2023 le vel. Percent preexploitation population (y-axi s) was calculated by firs t calculating 2023 population size using optimum recK value from Figure 3-3. Alternative recK values were then used to estimate 2023 population size, a nd then these population estimates were compared to the population size from the optimum recK value (Figure 3-3). 50

PAGE 51

0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 0 0.05 0.1 0.15 0.2 0.25Natural Mortality (M)% Pre-exploitation abundance at 2023 Figure 4-7. Model sensitivity to a ra nge of natural mortality values ( M x-axis) as measured by the percent recovery of th e population to the 2023 level. Percent pre-exploitation population (y-axis) was calcu lated by first calculating 2023 population size using M = k value calculated using the von Bertla nffy growth function. Alternative M values were then used to estimate 2023 population size, and then these population estimates were compared to the population size from the optimum M value (Table 3-1). 51

PAGE 52

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 024681 01 21Age at Maturity% Pre-exploitation abundance at 2023 4 Figure 4-8. Model sensitivity to a range of initial age at maturity Mai values ( Mai, x-axis) as measured by the percent recovery of the population to the 2023 le vel. Percent preexploitation population (y-axi s) was calculated by firs t calculating 2023 population size using Mai estimates from Huff 1975. Alternative Mai values were then used to estimate 2023 population size, and then these population estimates were compared to the population size from the optimum Mai value (Table 3-1). 52

PAGE 53

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0 0.05 0.1 0.15 0.2 0.25von Bertlanffy K% Pre-exploitation abundance at 2023 Figure 4-9. Model sensitivity to a range of von Bertalanffy k values ( k x-axis) as measured by the percent recovery of th e population to the 2023 level. Percent pre-exploitation population (y-axis) was calcu lated by first calculating 2023 population size using k estimates from growth curves developed with tagging data. Alternative k values were then used to estimate 2023 population size, and then these population estimates were compared to the population size from the optimum k value (Table 3-1). 53

PAGE 54

54 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0123456Skip spawn year interval% Pre-exploitation abundance at 2023 7 Figure 4-10. Model sensitivity to a range skip spawn interval Sk values ( Sk x-axis) as measured by the percent recovery of the population to the 2023 level. Percent pre-exploitation population (y-axis) was calcu lated by first calculating 2023 population size using Sk estimates from Sulak and Clugston (1999). Alternative Sk values were then used to estimate 2023 population size, and then these population estimates were compared to the population size from the optimum Sk value (Table 3-1).

PAGE 55

CHAPTER 5 DISCUSSION Hypothesis Evaluation Based on the results from this simulation model, it is likely that slow recovery of the Apalachicola River Gulf sturgeon population is a ttributable to hypothesis 1, low population size at the end of directed Gulf sturgeon harvest. Gulf sturgeon are adapted to a life history strategy common for other large-bodied, slow-growing fish species producing small numbers of recruits per year over a long life span (Goodwin et al. 2006). Becau se of their long life span and high age-at-maturity long periods of time ar e needed to recover a populations spawning capacity, even with ideal recruitm ent conditions and no increase in total mortality. Individuals spawned at the end of harvest in 1985 are only no w reaching their full fecundity potential. Figure 5-1 shows the effects of a simulated collapse an d recovery on the numbers-at-age of a Gulf sturgeon population. This figur e illustrates the slow recovery time caused by the need for the population to rebuild its age-structure and spawning capacity (Walters et al. In press). At the present time hypothesis 2, recovery lim ited by spawning habitat, does not seem to be a primary factor slowing populat ion recovery at the present tim e. However, attributes of hypothesis 1 may affect hypothesis 2, as low population size likely puts the population under threshold of spawning habitat restrictions. Spawning habitat area limitation may become a limiting factor in the near future as abundance increases with recovery. Gulf sturgeon spawning behavior may mitigate spawning habitat limitation effects with characteristics such as skip spawning, relatively short inc ubation periods (2-3 days), and annual spawning periods lasting several having the effect of reduc ing the density of spawners at the spawning sites at a given time, to percentage of the adult population (Sulak and Clugs ton 1998). Factors other than spawning habitat availability may limit Gulf st urgeon recruitment, with possible bottlenecks 55

PAGE 56

occurring later within the life history of the sp ecies. For example, Bradford et al. (1997) suggested that in-river rearing area might be a limiting factor for salmon smolt production. Management Implications Mortality Mortality is a critical aspect in any populati on if the mortality rate exceeds the birth rate the population will not grow and eventually will go extinct. Concerns that total mortality exceeded population growth rate was the motivati on to close the Gulf sturgeon fishery in 1985. This same concern motivated the end of white sturgeon harvest in the Kootenai River system (Paragamian et al. 2005) and management actions to reduce anthropogenic mortality for other species of concern, such as turtle-exclusion devi ces in commercial trawls to reduce sea turtle bycatch. Pine et al. (2008) found that sustainable e xploitation rates for Gulf sturgeon are relatively low (<25 percent annually) and populations are especially sensitive to harvest with small increases in mortality affecting populations (Morrow et al. 1998, Pine et al. 2001). Care should be taken to eliminate mortality from anthropoge nic sources including indirect mortality from sampling programs, fishery by-catch, or morta lities from beach restoration or maintenance dredging operations. Supplemental Stocking Gulf sturgeon stocking in Florida has been a contentious issue among managers in the past (Hoover 2002). As predicted by the mode l stocking, would not decrease the absolute recovery time, but would primarily increase short-term population size. This may be beneficial for severely depressed populations at a high risk of extinction, but is not required in populations that are stable or slowly growing such as the Apalachicola River Gulf sturgeon population. The stocking rates used in this model we re much lower than those recommended by the Atlantic States Marine Fisheries Commi ssion ([ASMFC] 1996), who recommended stocking 56

PAGE 57

approximately 2.5 million individuals over a ten year period, per river, for both Atlantic and Gulf sturgeon subspecies. For comparison, model esti mates of annual age-1 recruitment ranged from the hundreds to several thousa nds depending on yearly population abundance. High stocking rates over large periods of time such as these could genetically swamp these depleted populations with hatchery reared animals raising concerns over the genetic integrity of the population (Tringali and Bert 1998). This is the case with Kootenai River white sturgeon where natural recruitment was extremely low and the populatio n was maintained through a stocking program (Paragamian et al. 2005). These authors estima ted the Kootenai River white sturgeon population would reach a genetic bottleneck in 2005, afte r which the population would be dominated by hatchery-origin white sturgeon. Stocked individuals are less ecologically fit than native individuals, often having higher natural mortality rates (Lorenzen 2000, Flemi ng and Petersson 2001) and lower reproductive success (Fleming and Petersson 2001). Less fit hatchery individuals wi ll reduce the success of enhancement programs, however, over time less fit individuals will be removed from the population through natural selection and the remaining hatchery in dividuals and their offspring will more closely resemble wild individuals (Lorenzen 2005). The stocking component of my model included a function to control the rate at which stocked individuals will produce phenotypically wild offspring. The rate at which stocked Gulf sturgeon would produce wild offspring is unknown and in this study, the rate was set at 50 percent, meaning stocked fish would produce an equal number of stockand wild-type recruits. Greater rates of wild type production would increase the benefit of a Gu lf sturgeon stocking pr ogram by increasing the abundance and reproductive capacity of the wild population. 57

PAGE 58

A second key consideration for a hatchery prog ram is the long maturation time of Gulf sturgeon means that any stocked individuals mu st survive 6-10 years after stocking before contributing to the spawning biomass of the population. K. Sulak (USGS, personal communication ) has observed that survival rates of stocked Gulf st urgeon in the Suwannee River are significantly lower than those of native indivi duals. If an unidentified bottleneck exists in the system at an older age, stock individuals will still be subjected to th at bottleneck, decreasing their impact. There is a well established tradeo ff between size at stocking, survival, and costs that would be critical to evaluate in developing any stocking program (Leber et al. 2005). Jim Woodruff Lock-and-Dam Passage My model predicted that the Apalachicola Gu lf sturgeon population will not be limited by spawning habitat availability until reaching higher population de nsities. Because of this, establishing upstream passage thr ough JWLD may not result in th e expected population benefits and could be deleterious (see below). For upstr eam passage to be successful as a means of increasing access to upstream spawning sites, it would have to be accompanied by successful downstream passage as well (Jag er 2006). Freshwater life-his tory characteristics of Gulf sturgeon, such as lack of feeding and riverine sp ecialization, likely inhibit the success of this management activity. Unlike other sturgeon sp ecies, there are no landl ocked or freshwater resident Gulf sturgeon populations Uncertainty exists over how Gulf sturgeon would utilize the existing lock at JWLD for passage, as observed in shortnose sturgeon A. brevirostrum in South Carolina (Cooke et al. 2002). If Gulf sturgeon di d not use existing dam locks for passage, then dam modifications such as low-gradient fish ladders (Kynard et al. 2003) or fish elevators (Kynard 1998) may be necessary. If Gulf st urgeon did successfully pass JWLD, it is unknown whether or not Gulf sturgeon passed above JWLD would be able to navigate Lake Seminole to reach potential spawning sites in the Flint and Chattahoochee Rivers. From a conservative 58

PAGE 59

population management standpoint, until further in formation is known, passage through JWLD should be considered a one-way trip mortality event to the spawning biomass. If further evidence is discovered showing spawning habitat is a limiting factor to recovery, such as declines in recruitment and population size from the monitoring program, then other management options might prove to be more effective than passage at JWLD. Construction of artificial spawning areas has proven to be effective for increasing recruitment success of other sturgeon species (Khoroskho an d Vlasenko 1970, LaHaye et al. 1992, Johnson et al. 2006) and was recommended specifically in the Apalachi cola (Wakeford 2001). Another technique would be to optimize river flows duri ng spawning season to maximize the availability of spawning habitat area. Flows between 420 and 570 m3/s at JWLD have been suggested as suitable to accomplish this (USFWS 2006). Pursuing these methods may be more beneficial and less costly (because of the increased mortality risk) to the Gulf stur geon population. Recovery Criteria Options A review of available USFWS recovery plans reveals that common themes in recovery criteria include stable or nondeclining populations, minimizing ta ke (incidental or directed), protection of important or crit ical habitat, establishment of subpopulations, maintain genetic diversity, and recovery to minimum population sizes. Some of th ese criteria are often poorly defined or qualitative in nature Without clear recovery goals, it is difficult for management agencies to assess to the effectiveness of recovery actions (Gerber and Hatch 2002). Quantitative metrics are more informative than qualitative metrics regarding recovery success and should be emphasized in all reco very plans (Gerber and Hatch 2002). The long-term goal of the GSRP is curre ntly based on the qualitative goal that populations must be stable and self-sustaining through natural recruitmen t (USFWS 1995). My model showed, under steady recruitment and mortal ity conditions, that Gulf sturgeon populations 59

PAGE 60

are stable at relatively low popula tion sizes, such as those present at the end of harvest. This agrees with the findings of Mo rrow et al. (1998) who determined that the Pearl River Gulf sturgeon population would be stable with as few as 100 adults. However, even if stable, smaller populations are more susceptible to deleterious e ffects of anomalous events such as large scale mortality events due to hurricanes. Consideratio ns of future Gulf stur geon recovery criteria should include a minimum abundance level, relati ve to historic stock size and incorporating genetic considerations to satisfy recovery goals. A lternatively, instead of relying on abundance as an indicator of population health, common metr ics from managed fish species such as SPR could also be used. As mentioned in Chapter 2, SPR measures of >0.30 are considered healthy in exploited fish populations (Goodyear 1993). Genetic effective population size ( Ne) may be a useful recovery metric for Gulf sturgeon populations and recommended by Tringali and Bert (1995) for this species a nd used as part of recovery goals for other species (Siberian tiger Panthera tigris altaica [Seal and Foose 1984], bull trout Salvelinus confluentus [Rieman and Allendorf 2001]). The /500 rule is also commonly used for species conservation where long-term average Ne should be no less than 500 and short-term declines to no less than 50 in or der to avoid bottlenecks, inbreeding, and other negative genetic population effects (Franklin 1980). Bowen and Avise (1990), estimated current Gulf sturgeon populations have an Ne of around 50, but it is unknown whether or not this is an artifact of the historic harvest of the species or a natural result of the founder effect occurring when the Gulf sturgeon subpopulation was isolat ed from the greater Atlantic sturgeon population. If Gulf sturgeon Ne is naturally low a variation of the /500 rule would have to be used. 60

PAGE 61

The timeframe set in the GSRP (USFWS 1995) may not be realistic for the Apalachicola Gulf sturgeon stock. Population recovery approaching pre-exploitation le vels is probably not reasonable given the life-history characteristics of th is species. The recovery interval from end of harvest to present is only 38 years, which is within the lifespan of an individual Gulf sturgeon and likely incorporates just over three generations of new spawners. Because of these population characteristics, there may be little that managers can actively do to hasten the recovery of the Apalachicola Gulf sturgeon population. If the popul ation is currently incr easing, the best option for managers may be to do nothing and wait for the population to recover on its own. If the population is currently stable or decreasing then steps should be ta ken to specifically identify the limiting factor(s) and life-history stag e(s) that are primarily being aff ected. It is th erefore critical to know the population trends of Gulf sturgeon st ocks from a well define d monitoring program. Monitoring program and future research The most important component of any mana gement program is regular monitoring to assess whether the population is resp onding to the management action. It is critical that the monitoring program is designed to provide accu rate information at a level of precisions necessary to assess progress of the population to recovery (Ger ber and Hatch 2002, Campbell et al. 2002). The primary goal of this program must be to detect trends in population abundance. Without reliable information on these trends it is difficult to make proper management recommendations. Zehfuss (2000) r ecommended that to detect a 10% annual change in population size seven consecutive years of mon itoring would likely be necessary. Current Apalachicola River Gulf sturgeon monitoring performed on a 3-5 year interval is likely not adequate to detect population trends unless significant deviations (~50%) in population size occur. 61

PAGE 62

In conjunction with monitoring programs, re search should be conducted to fill in Gulf sturgeon ecology knowledge gaps. In this study pa rameter uncertainty and sensitivity analysis has provided guidance for future research needs. Primarily this involves the spawner-recruit relationship and recK parameter, growth rates, and natural mortality rate of Gulf sturgeon. Researching spawner-recruit relationships would involve surveying spawning biomass on spawning sites and later sampling for larvae downstream of these sites or for young-of-year individuals later in other parts of the river. Growth and mortality rates could be investigated using the ongoing monitoring tagging studies as well as by gatheri ng new direct ag eing data such as fin rays or otoliths sourced from Gulf stur geon mortalities occurring naturally or through sampling. Future research in these areas could help managers better understand factors affecting Gulf sturgeon populations and provide better in formation for use by population models to make new, improved predictions about possible trends in population status and recovery. In turn, new knowledge may identify other areas in need of data to answer new management questions. 62

PAGE 63

63 Figure 5-1. Surface plot represen ting theoretical Gulf sturgeon population subjected to harvest. Year of simulation on x-axis, age-class on y-ax is. Each cell represents an age class in a given year with the color representing numbers of individua ls in that age class. Large white area at top represents age classe s with 0 individuals while light blue area at bottom represents >150 individuals. Gr adient between these two areas is described by contour lines in 5-individual intervals. Population was harvested for first 25 years, ending at dark horizontal line, and then allowed to rec over. Erosion of population age-structure can be seen prior to 25 y ears, after which popula tion is recovering and age-structure is filling out.

PAGE 64

CHAPTER 6 CONCLUSION Population models are a valuable tool in natural resource conservation. Because of this, they are often required by the rec overy plans of species of concer n. In this study, a population model was used to evaluate several manageme nt scenarios for a threatened Gulf sturgeon population in the Apalachicola River. This was accomplished using existing data and more importantly, for any species of concern, withou t affecting or harming actual populations. The model in this study produced demonstrations of what may actually be occurring in wild Gulf sturgeon populations so that res ource managers will not be forced to make assumptions about the behavior of these populations. This model has al so identified existing problems with current recovery policy, such as vague recovery criteria and weak monitoring program. Specifically, the model has provided insight into the possible levels of change that may occur, which monitor programs must be designed to detect. My results show that the Apalachicola Gu lf sturgeon population is likely limited by population size at the end of harvest and will no t be fully recovered by the current long-term recovery date of 2023. Population recovery at this time may only be around 25-30% of the historic population size while the time to full recovery ma y be in excess of 100 years. Uncertainty in estimates of N1985 will not greatly affect recove ry time, but recovery will be slowed by any additional mortality added to the population. Limited conservation stocking would increase population size ove r the short-term, but would not reduce the time to full population recovery. Stock enhancement would ha ve been more effective if it was performed earlier in the recovery process. Populati on recovery could be limited by reduced spawning habitat area in the future, especi ally as population growth occurs. 64

PAGE 65

65 This Apalachicola River Gulf sturgeon case also serves as an example of the utility of modeling tools in a practical management framew ork. This same model can be modified and applied to a variety of other systems. When making policy d ecisions involving species of concern, managers must take extra precaution to ensure they make the best decisions possible using all means at hand. As new information a nd analysis methods become available, managers should use these tools to aid them in making informed policy recommendations. This model will serve as a flexible framework to help with current and future policy decisions. It is hoped that the model used in this study will be easily understood and of use to resource managers to answer questions of their own in the futu re. Effective management programs are thos e that successfully integrate modeling approaches with field research. Models can ne ver be right and are not replacements for well planned field experiments. If management actio ns are taken with the intent to improve Gulf sturgeon populations, then this model could be us ed to make predictions of population response, and then these predictions compared to data collected following the management action to assess model performance. This iterative process, termed adaptive management, of identifying management objectives, identifying management actions, using models to predict population response to management actions, implementing th e management action, and then testing the model with data from the management action, ha s been a successful approach in the past for maximizing learning and management effectiv eness (Walters 1986). A true adaptive management framework, using models in conjunction with field studies to evaluate and inform management policies, may be the best and most powerful method available for managers to use in managing animal populations.

PAGE 66

LIST OF REFERENCES ASMFC. 1996. Breeding and stocking protocol for cultured Atlantic sturgeon. Fishery Management Report 68. Atlantic States Marine Fisheries Commission. Beamesderfer, R. C. 1991. MOCPOP: A flexible simulator for analysis of age-structured populations and stock-related functions. Oregon Department of Fish and Wildlife, Information Report 91-4, Portland. Berg, J.J., M. S. Allen, and K.J. Sulak. 2007. Population assessment of the Gulf of Mexico sturgeon in the Yellow River, Florida. Pages 365-379 in J. Munro, D. Hatin, J. E. Hightower, K. McKown, K. J. Sulak, A. W. Kahnle, and F. Caron, editors. Anadromous Sturgeons: Habitats, Threats, and Manageme nt American Fisheries Society Symposium 56, Bethesda, Maryland. Beverton, R. J. H. and S.J. Holt. 1957. On the Dynamics of Explo ited Fish Populations. Chapman and Hall, London. Bowen, B.W., and J.C. Avise. 1990. Genetic st ructure of Atlantic and Gulf of Mexico populations of sea bass, menhaden, and sturge on: Influence of zooge ographic factors and life-history patterns. Marine Biology 107:371-381. Campbell, S.P., J.A. Clark, L.H. Crampton, A.D. Guerry, L.T. Hatch, P.R. Hosseini, J.J. Lawler, and, R.C. OConnor. 2002. An assessment of monitoring efforts in endangered species recovery plans. Ecological Applications 12:674-681. Chaloupka, M. 2002. Stochastic simulation modeling of southern Great Barrier Reef green turtle population dynami cs. Ecological Modeling 148:79-109. Chapman, F.A., S.F. OKeefe, and D.E. Campton. 1993. Establishment of parameters critical for the culture and commercialization of Gulf of Mexico sturgeon, Acipenser oxyrhynchus desotoi. Fisheries and Aquatic Sciences Dept., Food Science and Human Nutrition Dept., University of Florida, Gainesville, FL Project Final Report. NOAA No. NA27FD006601. National Marine Fisheries Service. St. Petersburg, FL. Clugston, J. P., A.M. Foster, and S.H. Carr, 1995. Gulf sturgeon, Acipenser oxyrinchus desotoi, in the Suwannee River, Florida, USA. Pages 215 in A. D. Gershanovich and T. I. J. Smith, editors, Proceedings of the Intern ational Symposium of Sturgeons. VNIRO Publications.,Moscow, Coggins, L.G. 2007. Active adaptive management for native fish conservation in the Grand Canyon: implementation and evaluation. Univers ity of Florida. Doctoral Dissertation. 66

PAGE 67

Cooke, D.W., S.D. Leach, and J.J. Isely. 2002 Behavior and lack of upstream passage of shortnose sturgeon at a hydroe lectric facility/navigation lock complex. Pages 101-110 in W. Van Winkle, P. Anders, D. Dixon, and D. Secor, editors. Biology, Management and Protection of North American Sturgeons. Am erican Fisheries Society Symposium 28, Bethesda, Maryland. Devries, D.R. and R. V. Frie. 1996. Dete rmination of age and growth. Pages 483-508 in B.R. Murphy and D.W. Willis, editors. Fisheries Techniques, 2nd Edition. American Fisheries Society, Bethesda, Maryland. Dulvy, N.K., Y. Sadovy, and J.D. Reynolds. 2003. Extinction vulnerability in marine populations. Fish and Fisheries 4:25-64. Enneson, J.J. and J.D. Litzgus. 2008. Using long-term data and a stage-classified matrix to assess conservation strategies for an endangered turtle ( Clemmys guttata ). Biological Conservation 141:1560-1568. Fabens, A. J. 1965. Properties and fitting of the von Bertalanffy growth curve. Growth 29:265289. Fleming, I. A. and Petersson, E. 2001 The ability of released, hatchery salmonids to breed and contribute to the natural produc tivity of wild populations. No rdic Journal of Freshwater Resources. 75:71. Flowers, H.J. and W.E. Pine. 2008. Observation of a juvenile Gulf sturgeon in the Santa Fe River, Florida. Southeastern Naturalist. 7:559-561. Flowers, H.J., W.E. Pine, A.C. Dutterer, K.G. J ohnson, J.W. Ziewitz, M.S. Allen. Implications of Modified Flow Regimes on Gulf Sturgeon Spaw ning in the Apalachicola River, Florida. In review Fox, D. A., J. E. Hightower, and F. M. Pa rauka. 2000. Gulf sturgeon spawning migration and habitat in the Choctawhatchee River system Alabama-Florida. Transactions of the American Fisheries Society 129:811-826. Franklin, I. A. 1980. Evolutionary changes in small populations. Pages 135 in M. Soule and B. A. Wilcox, editors. Conservation Biology: An Evolutionary Ecological Perspective. Sinauer Associates, Sunderland, Massachusetts. Freeman, M.C., C.M. Pringle, E.A. Greathous e, and B.J. Freeman. 2003. Ecosystem-level consequences of migratory faunal de pletion caused by dams. Pages 255-266 in K.E. Limburg and J.R. Waldman, editors. Biodive rsity, Status, and Conservation of the Worlds Shads. American Fisheries Society, Symposium 35, Bethesda, Maryland. Gerber, L. R. and L. T. Hatch. 2002. Are we re covering? An evaluation of recovery criteria under the U.S. Endangered Species Act. Ecological Applications 12: 668. 67

PAGE 68

Gerber, L.R., M.T. Tinker, D.F. Doak, J.A. Este s, and D.A. Jessup. 2004. Mortality sensitivity in life-stage simulation analysis: a case study of southern sea otters. Ecological Applications 15:1554-1565. Goodwin, N.B., A. Grant, A.L. Perry, N.K. Dulvy, and J.D. Reynolds. 2006. Life history correlates of density-dependent recruitment in marine fi shes. Canadian Journal of Fisheries and Aquatic Sciences 63:494-509. Goodyear, C.P. 1977. Assessing the impact of power plant mortality on the compensatory reserve of fish popul ations. Pages 186-195 in W. Van Winkle., editor, 1977. Proceedings of the Conference on Assessing the Effects of Power-Plant -Induced Mortality on Fish Populations. Pergamon Press, New York. Goodyear, C. P. 1980. Compensation in fish populations. Pages 253 in C. H.Hocutt and J. R. Stauffer, Editors, 1980. Biological Monito ring of Fish. Lexington Books, Lexington, Massachusetts. Goodyear, C.P. 1993. Spawning stock biomass per recruit in fisheries management: foundation and current use. Pages 67-81 in S.J. Smith, J.J. Hunt, and D. Rivard, editors, 1993. Risk Evaluation and Biological Reference Points for Fisheries Management. Canadian Special Publication in Fisheries and Aquatic Sciences. Gunn, J.M. and R. Sein. 2000. Effects of forestry roads on reproductive habitat and exploitation of lake trout ( Salvelinus namaycush ) in three experimental lakes. Canadian Journal of Fisheries and Aquatic Sciences 57:97-104. Heppell, S. S. 2005. Development of alternativ e quantitative tools to assist in jeopardy evaluation for sea turtles. Oregon State Univ ersity. Final Report. Southeast Fisheries Science CenterNational Oceanic and Atmospheric Administration. Hewitt, D. A. and J.M. Hoenig. 2005. Comparis on of two approaches for estimating natural mortality based on longevity. Fisheries Bulletin 103:433-437. Hilborn, R. and M. Mangel. 1997. The Ecologica l Detective. Princet on University Press. Princeton, New Jersey. Hoover, A. 2002. A century of sturgeon: The history, biology, and future of the Gulf of Mexico sturgeon in Florida. Techni cal paper. The University of Florida, Gainesville, FL. Huff, J. A. 1975. Life history of Gulf of Mexico sturgeon, Acipenser oxyrinchus desotoi in Suwannee River, Florida. Florida Marine Research Publications Number 16. Jager, H.I. 2006. Chutes and ladders and other game s we play with rivers. II. Simulated effects of translocation on white sturgeon. Canadian Journal of Fisheries and Aquatic Sciences 63:165-175. 68

PAGE 69

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 53:820822. Johnson, J.H., J.E. McKenna, Jr., D.S. Dropkin, and W.D. Andrews. 2005. A novel approach to fitting the von Bertalanffy relationship to a mixed stock of Atlantic Sturgeon harvested off the New Jersey coast. Northeaste rn Naturalist 12:195-202. Johnson, J.H., S.R. LaPan, R.M. Klindt, a nd A. Schiavone. 2006. Lake sturgeon spawning on artificial habitat in the St. Lawrence Rive r. Journal of Applied Ichthyology 22:465-470. Jorgensen, C., B. Ernande, O. Fiksen, and U. Dieckmann. 2005. The logic of skipped spawning in fish. Canadian Journal of Fish eries and Aquatic Sciences 63: 200. Kareiva, P., M. Marvier, and M. McClure. 2000. Recovery and management options for spring/summer chinook salmon in the Columb ia River Basin. Science 290:977-979. Kynard, B. 1998. Twenty-two years of passing shor tnose sturgeon in fish lifts on the Connecticut River: What has been learned? Pages 252 in M. Jungwirth, S.Schmutz, and S. Weiss, editors. Fish Migration and Bypasses. Blackwell Science Inc., Malden, Massachusetts. Kynard, B., D. Pugh, and T. Parker. 2003. Developm ent of fish passage for lake sturgeon. Final report to the Great Lakes Fishery Trust. S.O. Conte Anadromous Fish Research Center, Turner Falls, Massachusetts. Khoroshko, P. A. and A. D. Vlasenko. 1970. Ar tificial spawning grounds of sturgeon. Journal Ichthyology 10:286. LaHaye, M., A. Branchaud, M. Gendron, R. Ve rdon, and R. Fortin. 1992. Reproduction, early life history, and characteristics of the spaw ning grounds of the lake sturgeon (Acipenser fulvescens) in Des Prairies and L'Assompti on rivers, near Montreal, Quebec. Canadian Journal of Zoology 70:1681. Leber, K.M., R.N. Cantrell, and P. Leung. 2005. Optimizing cost-effectivene ss of size at release in stock enhancement programs. North Amer ican Journal of Fisheries Management 25:1596-1608. Lorenzen, K. 2000. Allometry of natural mortality as a basis for assessing optimal release size in fish-stocking programmes. Canadian Journal of Fisheries and Aquatic Sciences 57:23742381. Lorenzen, K. 2006. Population dynamics and potentia l of fisheries stock enhancement: practical theory of assessment and policy analysis. Philo sophical Transactions of the Royal Society B 360:171-189. 69

PAGE 70

Marchant, R. S., and M. K. Shutters. 1996. Artific ial substrates collect Gulf sturgeon eggs. North American Journal of Fish eries Management 16:445-447. Martell, S.J.D, W.E. Pine III, and C.J. Walters. 2008. Parameterizing age-structured models from a fisheries management perspective. Canadian Journal of Fisheries and Aquatic Sciences 65:1586-1600. McCullough, B.D. and B. Wilson. 1999. On the accur acy of statistical pro cedures in Microsoft Excel 97. Computational Statistics and Data Analysis 31:27-37. McCullough, B.D. and B. Wilson. 2002. On the accur acy of statistical pro cedures in Microsoft Excel 2000 and Excel XP. Computational St atistics and Data Analysis 40:713-721. Milner-Gulland, E.J. 1994. A population model for the model of the saiga antelope. The Journal of Applied Ecology 31:25-39. Morris, W.F., P.L. Bloch, B.R. Hudgens, and L.C. Moyle, J.R. Stinchcombe. 2002. Population viability analysis in endangered species recovery plans: past use and future improvements. Ecological Applications 12:708-712. Morrow, J. V., Jr., J. P. Kirk, K. J. Killgore, H. E. Rogillio, and C. Knight. 1998. Status and recovery potential of Gulf st urgeon in the Pearl River syst em, Louisiana-Mississippi. North American Journal of Fisheries Management 18:798-808. Morrow, J. V., Jr., J. P. Kirk, K. J. Killgore, and H. E. Rogillio. 1999. Recommended enhancements to the Gulf sturgeon recovery and management plan based on Pearl River studies. North American Journal of Fisheries Management 19:1117-1121. Musick, J.A. 1999. Criteria to defi ne extinction risk in marine fisheries. Fisheries 24:6-14. Paragamian, V.L, R.C.P. Beamesderfer, and S.C. Ireland. 2005. Status, population dynamics, and future prospects of the e ndangered Kootenai River white sturgeon population with and without hatchery intervention. Transactions of the Ameri can Fisheries Society 134:518532. Pascual, M.A., P. Kareiva, and R. Hilborn. 1997. The influence of model structure on conclusions about the viability and harvesting of Serengeti wildebeest. Conservation Biology 11:966-976 Pine, W. E. III., M. S. Allen, and V. J. Dreitz. 2001. Population viability of the Gulf of Mexico Sturgeon: Inferences from capture-recapture and age-structured models. Transactions of the American Fisheries Society 130:1164-1174. 70

PAGE 71

Pine, W.E. III, H.J. Flowers, K.G. Johnson, and M.L. Jones. 2006. An assessment of Gulf sturgeon movement, spawning site selecti on, and post-spawn holding areas in the Apalachicola River, Florida. University of Florida. Final Report. Florida Fish and Wildlife Conservation Commission Project PR640342. Pine, W.E. III, S.J.D. Martell, and H.J. Flowers. 2008. Status of Gulf stur geon in Florida waters: A reconstruction of historical population trends to provi de guidance on conservation targets. 4th Interim report. University of Flor ida USGS Cooperativ e Unit Project NG06004. Randall, M. and K.J. Sulak. 2007. Relationship be tween recruitment of Gu lf sturgeon and water flow in the Suwannee River, Florida. Am erican Fisheries Society, Symposium 56, Bethesda, Maryland. Ricker, W.E. 1975. Computation and interpretation of biological statistic s of fish populations. Fisheries Research Board of Canada Bulletin No. 191. Rideout, R.M., Rose, G.A., and Burton, M.P.M. 2005. Skipped spawning in female iteroparous fishes. Fish and Fisheries 6:50. Rieman, B.E. and R.C. Beamesderfer. 1990. White st urgeon in the lower Columbia River: Is the stock overexploited? North American Jour nal of Fisheries Management 10:388-396. Rieman, B.E. and F.W. Allendorf. 2001. Eff ective population size and genetic conservation criteria for bull trout. North American J ournal of Fisheries Management 21:756-764. Rien, T.A. and R.C. Beamesderfer. 1994. Accu racy and precision of white sturgeon age estimates from pectoral fin rays. Transact ions of the American Fisheries Society 123:255-265. Rossiter, A., D.L.G. Noakes, and F.W.H. Beam ish. 1995. Validation of age estimation for the lake sturgeon. Transactions of the American Fisheries Society 124:777-781. Seal, U.S, and T. Foose. 1984. Siberian tiger species survival plan: a strategy for survival. Journal of the Minnesota Acad emy of Science 49:3-9. Smith, T.I.J. 1985. The fishery, biology, and management of Atlantic sturgeon, Acipenser oxyrhynchus in North America. Environmen tal Biology of Fishes 14:61-72. Spear, B.J. 2007. U.S. management of Atlantic sturgeon. Pages 339-346 in J. Munro, D. Hatin, J. E. Hightower, K. McKown, K. J. Sulak, A. W. Kahnle, and F. Caron, editors. Anadromous Sturgeons: Habitats, Threats, a nd Management American Fisheries Society Symposium 56, Bethesda, Maryland. 71

PAGE 72

Sulak, K. J. and J. P. Clugston. 1998. Early life history stages of Gulf Sturgeon in the Suwannee River, Florida. Transactions of the American Fisheries Society 127:758-771. Sulak, K. J. and J. P. Clugston. 1999. Recent ad vances in life history of the Gulf of Mexico Sturgeon Acipenser oxyrinchus desotoi in the Suwannee River, Florida, USA: A synopsis. Journal of App lied Ichthyology 15:116-128. Sulak, K.J. and M. Randall. 2002. Understanding sturgeon life history: Enigmas, myths, and insights from scientific studies. Journa l of Applied Ichthyology 18:519-528. Tringali, M. D. and T. M. Bert. 1998. Risk to genetic effective populat ion size should be an important consideration in fish stock-enhan cement programs. Bulletin of Marine Science 62:641-659. USACE. 2004. Gulf sturgeon spawning habitat on the Apalachicola River: continuation of habitat survey and mapping, 06-07/20-21 November 2003, 21-22 January 2004. U.S. Army Corps of Engineers, Mobile District. USFWS. 1995. Gulf sturgeon recovery plan. U.S. Fish and Wildlife Service and Gulf States Marine Fisheries Commission. USFWS. 2003. Designation of critical habitat for the Gulf sturgeon. US Fish and Wildlife Service and National Marine Fisheries Service. USFWS. 2008. Biological opinion on the U.S. Ar my Corps of Engineers, Mobile District, revised interim operating plan for Jim Woodruf f Dam and the associat ed releases to the Apalachicola River. U.S. Fish and Wildli fe Service, Panama City, Florida. Wakeford, A. 2001. State of Florida conservation plan for Gulf sturgeon Acipenser oxyrinchus desotoi. FMRI Technical Report TR-8. Florid a Fish and Wildlife Conservation Commission. Walters, C.J. 1969. A generalized computer si mulation model for fish population studies. Transactions of the Ameri can Fisheries Society 3:505. Walters, C.J. 1986. Adaptive Management of Renewable Resources. The Blackburn Press. Caldwell, NJ. Walters, C.J., R. Hilborn, and V. Christensen. In press. Surplus production dynamics in declining and recovering fish populations. Canadian J ournal of Fisheries and Aquatic Sciences. Walters, C.J. and S.J.D. Martell. 2004. Fi sheries Ecology and Management. Princeton University Press. Princeton, N.J. Walters, C.J., S.J.D. Martell, and J. Korman. 2006. A stochastic appro ach to stock reduction analysis. Canadian Journal of Fisher ies and Aquatic Sciences 63:212-223. 72

PAGE 73

73 Wilson, P.H. 2003. Using population projection matr ices to evaluate recovery strategies for Snake River spring and summer chinook sa lmon. Conservation Biology 17:782-794. Wooley, C. M. and E. J. Crateau. 1985. Moveme nt, microhabitat, exploitation, and management of Gulf of Mexico sturgeon, Apalachicola River, Florida. North American Journal of Fisheries Management 5:590-605. Zehfuss, K.P., J.E. Hightower, and K.H. Pollock. 1999. Abundance of Gulf sturgeon in the Apalachicola River, Florida. Transactions of the American Fish eries Society 128:130143. Zehfuss, K. P. 2000. The status, movement, hab itat preference, and monitoring of Gulf sturgeon in several Florida rivers. Doct oral Dissertation. North Carolin a State University. Raleigh, North Carolina.

PAGE 74

BIOGRAPHICAL SKETCH I was born in Augusta, GA, in 1981, to Hank and Debra Flowers. I was raised and schooled in the Augusta, GA, area, graduating from Greenbrier High School in 1998. I also spent a significant amount of time in southwest Georgia, fishing, hunting, and exploring the outdoors with my grandparents, Opa and Oma; a nd uncle, Mick. These experiences, along with fishing trips with my dad, exploring the bac kyard with my sister, Ami, and high school FFA program with teacher Larry Moore that spurred my interest in natural reso urces and fisheries. I started college at the Georgia Institute of Technology in 1998, and then transferred to Abraham Baldwin Agriculture College in 2001, where I rece ived my A.S. degree in Forest Resources. I then transferred to the Univers ity of Georgia where I completed my B.S.F.R. in both Fisheries Management and Forest Envir onmental Resources. After gr aduation I worked as a field technician with the University of Georgia in Michigan with lake sturgeon on the Muskegon River and then worked as a field technician with the University of Arkansas, Fayetteville USGS Coop Unit on trout species in the White River system. I then came to the University of Florida in 2006 to work on my M.S. After completion of this degree I plan to attend North Carolina State University to pursue my Ph .D under Dr. Joseph Hightower.