HABITAT- AND DENSITY-MEDIATED INFLUENCES ON SNOOK...
Transcript of HABITAT- AND DENSITY-MEDIATED INFLUENCES ON SNOOK...
HABITAT- AND DENSITY-MEDIATED INFLUENCES ON SNOOK ECOLOGY: LESSONS LEARNED FROM MANIPULATIVE RELEASE EXPERIMENTS WITH HATCHERY-
REARED JUVENILE SNOOK
By
NATHAN PAUL BRENNAN
A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
2008
1
© 2008 Nathan Paul Brennan
2
To my father who encouraged and believed in me, who nurtured my intellectual curiosity,
provided perspective, and strengthened me, making this milestone possible. . .
3
ACKNOWLEDGMENTS
I will always be thankful and grateful for my wife, children, my mother and father, and
family members who showed their continued patience, support, and encouragement throughout
this venture. I gratefully acknowledge the contributions of my graduate committee Ken Leber
and Bill Lindberg (co-chairs), Carl Walters, Craig Osenberg, Mike Allen, and Tom Frazer for
their overall direction in experimental design, scientific approach, and editorial assistance.
Throughout, Ken has also provided constant support, direction, and encouragement which
allowed me to pursue independent avenues of research in a larger stock enhancement program.
I thank Carl for his additional assistance with field work, data analysis, and overall direction.
Bill Lindberg also provided extra direction, time, and encouragement during this endeavor.
Mote Marine Laboratory financially supported my graduate work and I commend Mote
administration for supporting the educational pursuits of their employees. Much gratitude is due
to R. DeBruler, D. Wilson, V. Mooney, A. Warner, B. Blackburn, G. Wahl, C. Ondercin, A.
Read, S. Andersen, and many interns and volunteers in the Stock Enhancement Program at Mote
Marine Laboratory. I thank those at Mote’s Center for Aquaculture Research and Development
for rearing snook used in this study. I thank L. Blankenship (Northwest Marine Technology,
Olympia, WA) and staff at the Stock Enhancement Research Facility, Florida Fish and Wildlife
Conservation Commission (FWC) for assistance during tagging and release activities. I thank P.
Hull, D. Doherty, and M. McLeod for assistance during snook spawning trips and snook
releases.
Funding for this work was provided by William Mote for initiating the snook hatchery, the
Florida Fish and Wildlife Conservation Commission, and the National Oceanic and Atmospheric
funded research consortium–the Science Consortium for Ocean Replenishment (SCORE).
4
TABLE OF CONTENTS page
ACKNOWLEDGMENTS ...............................................................................................................4
LIST OF TABLES...........................................................................................................................7
LIST OF FIGURES .........................................................................................................................9
ABSTRACT...................................................................................................................................15
CHAPTER
1 INTRODUCTION ..................................................................................................................17
2 INFLUENCE OF REARING HABITAT ON GROWTH, SURVIVAL, DISPERSAL, AND RELATIVE STOCK CONTRIBUTION OF COMMON SNOOK Centropomus undecimalis........................................................................................26
Introduction.............................................................................................................................26 Methods ..................................................................................................................................29
Experimental Design .......................................................................................................29 Study Habitat and Timeline.............................................................................................30 Hatchery Fish Production, Tagging and Release ............................................................31 Collection of Environmental Variables at Release Sites .................................................32 Short-Term Collections ...................................................................................................34 Long-Term Collections ...................................................................................................34 Sample Processing and Data Analysis ............................................................................36
Results.....................................................................................................................................39 Short-Term collections at Release Sites ..........................................................................40 Long-Term Assessment...................................................................................................43
Discussion...............................................................................................................................45
3 MANIPULATIONS OF STOCKING MAGNITUDE: ADDRESSING DENSITY-DEPENDENCE IN A JUVENILE COHORT OF COMMON SNOOK Centropomus undecimalis........................................................................................85
Introduction.............................................................................................................................85 Methods ..................................................................................................................................88
Study Area .......................................................................................................................88 Experimental Design and Sampling Methods .................................................................89 Gear Efficiency and Population Estimation ....................................................................90 Pre-Release April 2002 Sampling ...................................................................................94 Tagging and Release........................................................................................................94 Post-Release Evaluation ..................................................................................................95
Results.....................................................................................................................................95 Gear Efficiency................................................................................................................95
5
Tagging and Release of Hatchery-Reared Snook............................................................96 Post-Release Evaluation ..................................................................................................96 Creek Density ..................................................................................................................98
Discussion...............................................................................................................................98
4 CRYPTIC CANNIBALISM IN SIZE-STRUCTURED SNOOK POPULATIONS ...........112
Introduction...........................................................................................................................112 Materials and Methods .........................................................................................................116
Generalized Snook Life History in a Tidal Creek .........................................................116 Approach 1: Phenotypic Characteristics .......................................................................116 Approach 2: Field Enclosure Trials...............................................................................117 Approach 3: Cannibalism Intensity in Wild Snook Populations..................................119
Results...................................................................................................................................123 Phenotypic Characteristics ............................................................................................123 Field Enclosure Trials....................................................................................................124 Cannibalism Meta-Analysis and Model Results ...........................................................124
Discussion.............................................................................................................................126
5 CONCLUSION.....................................................................................................................150
LIST OF REFERENCES.............................................................................................................159
BIOGRAPHICAL SKETCH .......................................................................................................183
6
LIST OF TABLES
Table page 2-1 Numbers of released snook in 1998 and 1999 according to creek, sub habitats within
creeks, and release size category .......................................................................................58
2-2 Nomenclature for classification of release sites, release systems, release sub-systems and release habitats. ...........................................................................................................59
2-3 Numbers of snook released according to release system and year. 1997-1 and 1997-2 represent two release experiments performed in 1997. .........................................60
2-4 Release sites, dates and sampling schedule at the release sites .........................................61
2-5 Snook captures organized by sampling type and locations. These data were used to verify results from short-term sampling and monitor recruitment to adult habitats. .........62
2-6 Variables used for the principle component analysis examining characteristics of the release sites and associated observed short-term post-release responses (first summer and winter after release) of hatchery-reared snook age-1 snook .......................................63
2-7 Variables used for the principle component analysis examining both short- (days-at-large <= 365) and long-term (days-at-large > 365) post-release responses of hatchery-reared snook stocked in 1998 and 1999 according to release site characteristics.....................................................................................................................64
2-8 von Bertalanffy growth parameters for snook recaptures (n) sourced from different release habitats. Length and age data only included recaptures of hatchery snook at-large for 180d or longer. ....................................................................................................65
2-9 Numbers of snook recaptured at various distances (km) away from their respective release sites. Data are only for snook recaptured within the first year after releases. ......66
3-1 Results from standardized sampling in April 2002 and physical attributes of experimental creeks (shoreline distance and percent altered habitat)..............................104
3-2 Mark-recapture results for the four experimental creeks organized according to sample month. “Hatchery 2x” and “Wild 2x” represent snook that were recaptured in two subsequent sampling events..................................................................................105
3-3 Numbers of hatchery-reared snook released at the four experimental tidal creeks. Abundance estimates by creek for the age-1 wild snook before releases (April 2002) are provided with the expected and observed (in June 2002) percentage of hatchery snook in post-release collections. ....................................................................................106
7
4-1 Summary of studies examined regarding common snook diets.......................................135
4-2 Numbers of snook processed for stomach contents and observed instances of cannibalism from winter data in nursery habitats (i.e. from Adams and Wolfe, 2006, and Sarasota data, 2004) ..................................................................................................136
4-3 Linear regression statistics for various morphometric comparisons: vertical gape (VG), body depth (BD, anal spine (AS), throat to anus (TRH-AN), standard length (SL), and fork length (FL). ..............................................................................................137
4-4 Number of replicate experimental treatments (N) of field enclosure trials examining cannibalism according to predator size, the inclusion of additional prey (bait), snook prey density, and two prey size categories (expressed as a percent of predator size) .....138
4-5 Details of actual cannibalism occurrences found in snook stomachs collected in the field. .......................................................................................................................139
8
LIST OF FIGURES
Figure page 2-1 Map of experimental release sites located off the southwest coast of Florida USA.
Release creeks are denoted, and release sub-habitats (U= upstream, M=midstream, L=lower stream, and I=barrier island habitat) are enclosed in transparent shaded ovals ..............................................................................................................................67
2-2 Release numbers (grey bars) and post-release results showing percent hatchery snook found in total snook catch (graph A) and recapture ratios (number recaptured/number released) adjusted for effort (from 21 m seine, graph B). ..................68
2-3 Mean recapture ratios (number recaptured/number released x100) of snook from 1998 and 1999 releases, by sub-subsystems. .....................................................................69
2-4 Recapture ratios (number recaptured/number released x 100) of snook by release habitat. Data are from recaptures collected within the first year after release. Mean recapture ratio is indicated by the horizontal line..............................................................70
2-5 Linear regression models using mean catch-per-unit effort during winter of age-1 snook to predict short- (< 1 year post-release) and long-term (>=1 year post-release) recapture ratios (rratio; number recaptured/number released x 100).................................71
2-6 Mean growth rates from fish released in different habitats and sub-habitats. Data are from recaptures collected 6-10 months after releases in 1998, and 1999..........................72
2-7 Size-at-recapture for snook released in different habitats. Data are from recaptures collected 6-10 months after releases in April 1998, and 1999...........................................73
2-8 Mean size-at-release and growth rates of hatchery-reared snook recaptured during the first winter after release (6-10 mo.). ............................................................................74
2-9 Length frequency distributions of wild and hatchery snook at microhabitat release sites for each season...........................................................................................................75
2-10 Plot of first two principle components for short-term data related to characteristics of the release sites ..................................................................................................................76
2-11 Plot of growth rate versus recapture ratio (number recaptured/number released*100) from standardized seine collections during the first winter after release...........................77
2-12 Catch per unit effort (CPUE) of wild and hatchery snook less than 300 mm fork length according to release habitat and associated mean growth rates (mm/d) through the first winter after release................................................................................................78
2-13 Plot of the canonical variables (triangles) for size-at-release (SAR mm FL) and associated response variables. Release sites are plotted as shaded dots ...........................79
9
2-14 Mean dispersal distances (km) organized by size-at-recapture (RecapFL, graph A), and size-at-release (SAR, mm FL, graph B)......................................................................80
2-15 Plot of length and age data from hatchery snook recaptures and associated vonBertalanffy models.......................................................................................................81
2-16 Mean dispersal distances from rearing habitats (minimum distance from release site to recapture site) of snook recaptured within seasons (spring, summer, fall winter) and by snook age (years)....................................................................................................82
2-17 Plot of the first two principle components comparing long-term responses (recaptures at-large for over 1 year, black dots) of the released fish, short-term responses (from recaptures at-large less than 1 yr., also black dots), and characteristics of the release sites (shaded dots)................................................................83
2-18 Numerical and biomass projections to adult stages of juvenile hatchery-reared snook released in various rearing habitats with characteristic growth rates by first winter.........84
3-1 Map of experimental study sites along the coasts of Sarasota and Manatee Counties, Florida 107
3-2 Schematic of various depletion-removal methods including (a) “two-bank depletion”, (b) “single-bank depletion”, and (c) “two-bank extended” depletion..............................108
3-3 Graphical representation of individual capture probabilities (p) of snook from various depletion-removal methods (circles, triangles and squares, graph a) and the likelihood profile for pcommon (graph b) ...........................................................................109
3-4 Abundance estimates of wild (grey area) and hatchery (white area) age-1 snook over time in treatment creeks receiving high augmentation levels (top graphs) and low augmentation levels (bottom graphs)...............................................................................110
3-5 Results from post-release samples collected in June 2002 showing catch-per-unit effort (graph a) and relative abundance (graph b) of hatchery snook (white bars) to wild juvenile (grey bars) abundance (set at 100%)..........................................................111
4-1 Map of Sarasota and Manatee counties where snook collections and stock enhancement research has been performed. Arrows indicate the tidal creeks where snook were captured and subsequently examined for stomach contents.........................140
4-2 Photograph of enclosures where cannibalism trials occurred..........................................141
4-3 Schematic of model process predicting cannibalism-related mortality by the various age groups of snook .........................................................................................................142
4-4 Annual relative abundance of age-1 and adult snook standardized around age-0 abundance ........................................................................................................................143
10
11
4-5 Graphical comparisons of snook morphometrics (in mm) ..............................................144
4-6 Photograph of a 70 mm (TL) and a 178 mm (TL) common snook. Black spot at tip of dorsal fin is obvious in smaller snook, but spot is nearly absent in larger snook.............. 145
4-7 Graphical results from field enclosure trials describing mean cannibalism rates of small and large snook according to various prey treatments ...........................................146
4-8 Plot of body lengths (fork length FL) of snook cannibals and cannibalized prey ...........147
4-10 Monte-Carlo simulation results (n=10,000 iterations) showing proportions of age-0 mortality due to age-specific cannibalism .......................................................................148
4-11 Monte-Carlo simulation results (n=10,000 iterations) showing the proportion of age-0 abundance (at the beginning of the model period) cannibalized by the various age groups of snook................................................................................................................149
LIST OF ABBREVIATIONS
α area covered by a standardized depletion net
β number of snook captured
δ number of snook remaining in a standardized seine sampled area
γage-i age specific estimate of the number of snook in a given cohort
λi field cannibalism rate, i.e. observed number of age i snook with cannibalized prey in their stomachs / total number of age i snook stomachs examined
AS anal spine length (mm)
b scaling parameter of MW function
BC Bowlees Creek
BD body depth (mm), measured at the widest part of the body
c mortality due to cannibalism (number of individuals)
CPUE catch-per-unit effort,
CWT coded-wire tag
Dc density (number per 30 m of shoreline) of juvenile snook in the creeks
D.O. dissolved oxygen (mg/l)
FL fork length (mm)
HB Hudson Bayou
HC total creek shoreline distance
I Barrier island habitat
k vonBertalanffy growth coefficient
L Lower stream mouth habitat
Linf vonBertalanffy parameter, asymptotic length
M Midstream habitat
Mu annual natural mortality at unit weight
12
MW weight specific instantaneous natural mortality rate
NC North Creek
N̂ Ct total in-creek juvenile snook abundance at time t
PC Phillippi Creek
pcommon the conditional maximum likelihood estimate for binomial sampling for a all depletion removal trials
p it proportion of age i snook in a given tidal creek at time t
pmax the conditional maximum likelihood estimate for binomial sampling for a particular depletion removal trial
s seined area of a creek
SAR Size-at-release (FL, mm)
SC South Creek
SL standard length (mm)
snook common snook (Centropomus undecimalis)
t time in days
t0 vonBertalanffy parameter, theoretical time when fish length = 0
TIDY Tidy Peninsula, also known as Tidy Island
TIDY_I Barrier island counter part to TIDY
TL total length (mm)
T0
U Upstream habitat
W snook body weight (or mass, in g)
WB Whitaker Bayou
WtCi typical weight of individual prey that age group i would cannibalize
Wti the weight of an individual in September from cohort i
VA-AN the distance from the ventral gill arch (hypohyal bone) to the anus
13
VG vertical gape, defined as the vertical distance between the anterior tip of the premaxlilla and the anterior tip of the dentary bone
VIE visible implant elastomer
14
Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy
HABITAT- AND DENSITY-MEDIATED INFLUENCES ON SNOOK ECOLOGY: LESSONS LEARNED FROM MANIPULATIVE RELEASE
EXPERIMENTSWITH HATCHERY-REARED JUVENILE SNOOK
By
Nathan Paul Brennan
August 2008
Chair: William J. Lindberg Cochair: Kenneth M. Leber Major: Fisheries and Aquatic Sciences
Common snook Centropomus undecimalis stocks in Florida, USA are experiencing
increasing pressure from the recreational fishing industry and habitat loss. This dissertation
investigates components of snook stock enhancement research, as a potential supplementary
management tool. As juvenile stages are typically used for stocking, I examined influences of
habitat and ambient snook density on post-release fates, and the potential for cannibalism in
juvenile stages.
Chapter 2 examines post-release growth, survival, and dispersal of juveniles stocked into
rearing habitats, and their numerical and biomass contributions to adult stocks. Growth was
inversely related to recapture ratios and release sites producing low growth had the highest
numerical contributions to adult stocks, but highest biomass contributions were from sites
producing intermediate growth rates.
In Chapter 3, manipulations of release magnitude of hatchery-reared juveniles showed tidal
creeks with high augmentation levels (n=2) resulted in overall post-release abundance increases
(10 months post-release) but imperceptible changes in low augmentation creeks (n=2) of wild
and hatchery conspecifics. Disproportionate post-release loss (85%) of hatchery-reared snook
15
16
occurred in one high augmentation creek but was not detected in the other. No loss of wild
conspecifics was detected in any of the creeks.
Snook are cannibalistic and show evidence for this behavior intensifying during early
stages. In Chapter 4, I examined ontogenetic differences in cannibalism potential and its role in
controlling abundance. Morphological comparisons (anal spines, mouth gapes, and confusion
colorations [eyespots]) showed that predatory defense and compensatory feeding adaptations
decreased with size. Examination of existing research on snook diet coupled with model
simulations showed that inter- and intra-cohort cannibalism intensified in nursery habitats and
during winter when older age groups were present.
Overall this work used released snook to quantify tradeoffs in production to adult stocks
based on characteristics of snook rearing habitats. Within an age-1 cohort, stocking increased
overall abundance but sometimes with high post-release mortality in hatchery snook. Intra- and
inter-cohort cannibalism was focused on age-0 snook in rearing habitats, and could be influential
on inter-annual recruitment dynamics. These findings provide valuable input for stakeholders
evaluating snook stock enhancement potential in Florida.
CHAPTER 1 INTRODUCTION
Global declines in coastal habitat quality, ecological diversity, and fishery stocks continue
to concern stakeholders and resource managers. Anthropogenic activities are thought to be the
primary causative agents (i.e. over harvest and habitat degradation, Lotze et al., 2006), and
efforts to reverse or slow this trend have been largely unsuccessful. Accurate knowledge of
ecosystem-wide function and processes, and species-specific ecological knowledge must be
coupled with responsive and adaptive corrective actions to restore ecosystem health.
Florida has one of the highest rates of human population increase in the United States
(Campbell, 1997). With a high coast to land-area ratio, much associated activity and
development is concentrated along Florida’s coast, which can influence the coastal ecosystems
(Bruger and Haddad, 1986, Lorenz, 1999; Lorenz and Sarafy, 2006). Unfortunately, shallow
coastal systems also are some of the most diverse and productive systems, making them
particularly important for conservation and management that target critical biological habitats
and associated resources. Many valuable fishery stocks in these same areas are also the focus of
increasingly popular recreational fisheries. Florida’s USD 5.4 billion recreational fishery is
easily the largest in the USA (American Sportfishing Association, 2004), and is known
internationally as a hot spot for recreational fishing. Traditional management measures are used
to control harvest and harvest laws are continually adjusted to maintain resource levels in the
face of the recreational fishing popularity.
Alongside traditional management practices, Florida’s managers and stakeholders have
directed efforts at stock enhancement to aid in restoring depleted stocks. Stock enhancement
seeks to augment the natural supply of juveniles to overcome recruitment limitation by releasing
animals produced in aquaculture settings into the wild (Bell et al., 2008). This approach is
17
becoming increasingly popular among stakeholders and managers worldwide (Honma, 1993;
Bell et al., 2006; Bell et al., 2008) probably owing to its seemingly direct application (Travis et
al., 1998). Unfortunately, direct evidence of augmentation effects in localized demes,
populations, and ecological communities is sparse (Hilborn, 1998; Travis et al., 1998; Leber
1999; Bell et al., 2006). While economically important stocks experience high harvest rates and
are declining (FAO, 2004), and many stocks have collapsed from overfishing (Steele, 1992;
Hutchings, 2000), responsive changes in community structure (Carscadden et al., 2001; Steneck,
2006) and ecosystem process (Pauley et al., 1998; Jackson et al., 2001) along with alteration and
loss of critical habitats (Vitousek et al., 1997) has added considerable uncertainty toward the
feasibility of stock enhancement programs. Aside from biological impacts, broader measures of
the overall feasibility of stock enhancement programs that weigh economic, social, and
institutional influences have been few (Garaway et al., 2006; Lorenzen, 2008) and add to the
difficulties and uncertainties of this approach.
Nonetheless, advances in culture technology of marine animals since the 1970’s, coupled
with a more focused scientific approach has rapidly progressed stock enhancement research
(Blankenship and Leber, 1995; Munroe and Bell, 1997; Immamura, 1999; Leber, 1999). Major
progress has occurred through improving fitness of hatchery-reared animals, tagging and
subsequent understanding of release techniques and recruitment of stocked animals to harvested
and spawning populations, genetic implications for stock enhancement, and understanding
biological, economic and social criteria necessary for stock enhancement programs to be
feasible.
Improvements in fitness of hatchery-reared animals has provided substantial progress in
the field through an understanding of behavioral deficiencies of hatchery-reared animals (Olla
18
and Davis, 1998; Hossain et al., 2002; Masuda, 2004; Davis et al., 2004), and hatchery selection
effects (Berejikian et al., 1999). Proactive steps toward managing fish health and disease
transmission from hatchery to wild stocks are also important contributions in this field (Kennedy
et al., 1998; Mushiaki and Muroga, 2004).
Tags (or marks) are crucial elements in providing informative feedback necessary for
evaluating the efficacy of stock enhancement programs (Blankenship and Leber, 1995; Walters
and Martell, 2004). Typically, stock enhancement programs require tags to be applied to many
juvenile fishes with minimal harm yet have high information content, low associated costs, and
have high retention rates throughout the lifetime of the organism (Blankenship and Leber, 1995).
A variety of tag types have been used to monitor stocked individuals including genetic tags
(Bravington et al., 2004; Jorstad, 2004; Sekino et al., 2005), chemical and thermal marks (Volk
et al., 1999; Jenkins 2002), internal tags (Davis et al., 2004; Brennan et al., 2005), electronic
archival tags (e.g. Prentice et al. 1990), and external tags (Stoettrup, 2002). The use of
tags along with the application of controlled experimentation and assessment of enhanced stocks
has provided quantitative feedback to substantially progress the field (Leber et al., 2004).
Specific improvements in stocking strategies include determining optimal size-at-release
(Tsukamoto, 1989; Yamashita et al., 1994; Leber, 1995), release season, and size-season
interactions (Leber et al., 1997; Leber et al., 1998; Sanchez-Lamadrid, 2002; Gwak et al., 2003),
and release habitat (Solazzi et al. 1991; Russell et al., 2004; Davis et al., 2005; Andersen et al.,
2005) and acclimation to the release sites (Brennan et al., 2006; Fairchild, 2008). Eventual
recruitment of stocked animals to adulthood and to fishery landings has also been documented
(Svåsand et al., 1990; Kitada et al., 1992; Leber and Arce, 1996; Kaeriyama, 1999; Zohar et al.,
2008).
19
Recent progress has also occurred through understanding the broader potential for stock
enhancement (Taylor et al., 2005; Garaway et al., 2006; Lorenzen, 2006; Lorenzen, 2008). This
involves an integrated analysis of biological, economic, and social criteria for suitable candidate
species for stock enhancement activities. Much of this work directly incorporates stakeholder’s
priorities and needs that are primarily responsible for driving most stock enhancement activities.
Models are needed to evaluate the candidacy of species targeted for stock enhancement
considering what is known about life history traits and other parameters of the stock (e.g.
fecundity, recruitment dynamics, density-dependent growth and survival), cost effectiveness of
hatchery production, fishery economics and logistics, how it integrates with other forms of
fishery management, and intuitional goals and priorities (Lorenzen, 2008). These types of
proactive evaluation measures are needed prior to commitments of large investments in stock
enhancement programs (Hilborn, 1998; Walters and Martell, 2004; Lorenzen, 2008). This has
been the exception rather than the norm for most stock enhancement programs however (Bell et
al., 2006; Bell et al., 2008). The inclusion of stakeholders in the process (e.g. Garaway et al.,
2006; Uki et al., 2006; Becker et al., 2008; Tomiyama et al., 2008; ), however, is a positive step
towards understanding and directing stock enhancement and other fishery management programs
toward adaptive management approaches (Walters, 1997; Perry et al., 1999; Bell et al., 2008).
Less understood and less demonstrated, yet key areas of uncertainty in this field, include
density-dependent effects, genetic effects, and economic feasibility. Numerical responses to
stocking in wild and hatchery-sourced animals (e.g. intra- and inter-specific competition or
cannibalism) and interactive effects within and across generations are rare (Leber et al., 1995),
and remain a high priority (Leber et al., 2004; Walters and Martell, 2004; Lorenzen, 2005).
Studies examining the effects of stock enhancement programs on inter-annual and spatial
20
variation in recruitment across community and ecosystems that aligns with current ecosystem-
based management programs (e.g. Pauley et al., 2000) are also a high priority. Genetic effects
and consequences of releasing hatchery-reared animals on wild populations have been predicted
(e.g. Ward, 2006; De Innocentiis et al., 2008; Tringali et al., 2008a) and serve as useful
guidelines for stock enhancement programs. Application of these guidelines, however, remains a
challenge and adds to the difficulties in achieving cost-effective programs (Lorenzen, 2008).
Preliminary economic analyses generally predict that high costs of hatchery production and
inflexible social implications of hatchery production often preclude economic success (Hilborn,
1998; Walters and Martell, 2004; Bell et al., 2006; Lorenzen, 2008). Optimizing cost-
effectiveness of stocking programs by improving post-release survival (Leber et al., 2005;
Brennan et al., 2006) are positive steps toward this, and creative methods for production may
also be necessary to make stock enhancement programs feasible (e.g. Japanese scallops
Patinopecten yessoensis are produced in hanging cages in situ, bypassing the need for hatcheries
[Uki, 2006]). As with all forms of fisheries management programs, challenges lie in how stock
enhancement programs can adjust to adaptive management measures based on dynamic feedback
from biological, economic and social sources (Walters, 1997; Walters and Martell, 2004; Bell et
al., 2006; Lorenzen, 2008).
The common snook Centropomus undecimalis, is one of the focal species of stock
enhancement research aimed at supplementing Florida’s existing stock management programs.
Common snook are an estuarine-dependent catadromous fish found in tropical and subtropical
habitats of the eastern coasts of the Americas, including Florida. Along with substantial
alteration of snook habitat (Bruger and Haddad, 1986), the increasing popularity of the
recreational fishery has caused concern for managers and stakeholders. Florida’s snook
21
populations are considered overfished and stocks are below management goals (Muller and
Taylor, 2005). Despite increasingly restrictive fishery regulations on common snook, annual
fishing mortality rates have steadily increased over the last 20 years, and recruitment has
generally declined (Muller and Taylor, 2005). High fishing pressure, coupled with relatively few
spawner-sized females in the adult population, has made management of snook stocks difficult
(Muller and Taylor, 2005). Furthermore, snook associate with shoreline habitat (Marshall, 1958;
Gilmore et al., 1983, McMichael et al., 1989; Peters et al., 1998a) and thus depend on coastal
waters that may be subjected to intense anthropogenic influences. Collectively, overfishing and
habitat loss have caused a general decline in the population although the relative influence of
these is unclear (Muller and Taylor, 2005).
In 1985, the Florida Fish and Wildlife Conservation Commission (FWC) and Mote
Marine Laboratory (MML) initiated a long-term stock enhancement research program with
common snook. This was followed a decade later by advancements in snook husbandry
technology that facilitated the production of sufficient numbers of hatchery-reared snook
juveniles to support experimental releases at a reasonable scale (thousands of individuals).
Based in Sarasota, Florida, snook stock enhancement research has prioritized a
responsible approach (sensu Blankenship and Leber, 1995) coupled with intensive field
evaluation to provide experimental feedback. Since 1996, over 52,000 juvenile hatchery-reared
snook have been released through a variety of experimental manipulations. All released snook
were tagged with coded-wire tags (CWT) to identify experimental release groups from juvenile
through adult stages of the released snook (Brennan et al., 2005). Visible implant elastomer
(VIE) tags were also used to provide external identification of hatchery-reared snook to aid in
fishery returns and provide benign evaluation of basic release groups (Brennan et al., 2005).
22
Researchers found that snook acclimated in predator-free enclosures located at release sites had
about twice the survival rates compared to those released without acclimation (Brennan et al.,
2006).
Ongoing sampling of the localized snook population has documented survival to
adulthood of the released hatchery-reared juveniles. Since 1999, individuals from release
cohorts have been regularly documented in annual wild snook spawning aggregations. Both
male and female hatchery-reared snook have been recaptured with mature gonads, and some
recaptured individuals have outgrown the harvestable size slot (27-33 inches total length [TL, or
686-838 mm TL]) (Leber and Brennan, unpublished data). Releasing juvenile snook into a
variety of habitats has provided valuable understanding of essential fish habitat (as defined in the
1996 amendments to the Magnuson-Stephens Act) for juvenile snook (Chapter 2). Quantitative
sampling in juvenile and adult snook habitats over the years also allowed researchers to measure
relative contribution rates of various juvenile rearing habitats to adult demes, and documenting
associated migratory patterns from juvenile to adulthood (Chapter 2). Mark-recapture data from
hatchery-reared and wild snook using a variety of tag technology has provided important insight
into short and long-term dispersal patterns (Chapter 2 and 3; Pine et al, unpublished data). Using
hatchery-reared juvenile snook to experimentally manipulate ambient juvenile abundance in tidal
creeks researchers were able to examine how density-dependent responses within a cohort varied
between habitats and across seasons (Chapter 3).
Stomach content examination of wild and hatchery-reared snook from field collections
has also provided valuable insight into snook dietary preferences, stomach content evacuation
rates Brennan and Heinlein, unpublished data), and diel feeding periodicity (Rock et al.,
unpublished data). With evidence for density-dependent survival in juvenile snook, the
23
potential for cannibalism was also examined using laboratory and field studies (Chapter 4).
Associated research has also measured natural and fishing mortality rates in slot-sized snook
(Bennett, 2006), and provided an important understanding of seasonal and annual variability in
microhabitat use by adult snook (Marcinkiewicz, 2007).
This dissertation represents an integral part of the FWC/MML project and is focused on
understanding snook life history and ecology, especially in juvenile stages, characteristics of
snook rearing habitats, and productive capacities of these habitats in response to variations in
inter-annual recruitment fluctuations. More specifically, in Chapter 2, using theoretical
perspectives of size-structured growth, survival, and movement, and rigorous monitoring of the
stocked fish I examined evidence for habitat-specific tradeoffs in post-release growth, survival,
and dispersal of the stocked fish. I related these post-release responses to characteristics and
conditions at the rearing habitats. Long-term monitoring of the stocked fish for years after the
releases showed that patterns in growth and survival established within the first year after release
were correlated to characteristics of the rearing habitats, and continued to be reflected in the
stocked fish as they matured. Coupling this with related work on post-release survival (Chapter
3), I was able to predict numerical and biomass contributions of specific rearing habitats to adult
stages, then examine differences and tradeoffs between growth and survival of the stocked snook
based on rearing habitat characteristics.
In Chapter 3, I examined the potential for density-dependent mortality in juvenile snook
cohorts. I used releases of juvenile hatchery-reared snook to artificially manipulate localized
recruitment and abundance across different habitats. Depletion-removal trials were first used to
calculate the efficiency of a standardized seine for sampling, then, coupled with an intensive
standardized sampling program, in-creek abundance of juvenile snook was measured. Two tidal
24
creeks received high augmentation release treatments, and two creeks received low augmentation
treatments. Abundance was measured before and through the first year after the experimental
releases. This study focused on augmentation effects within a cohort of juvenile snook, but
future work should also examine inter-cohort predation and mortality and how inter-annual
natural variation on ambient snook abundance could influence similarly staged release
experiments.
As snook are piscivorous and have shown evidence for density-dependent growth and
mortality, in Chapter 4, I examined the potential influence of cannibalism in controlling juvenile
snook abundance. I examined published and unpublished literature on snook diets, and
supplemented this with my own examinations of Sarasota snook stomach contents. Then I used
these data to provide input into a model I developed to understand the role of cannibalism in
snook abundance dynamics and to provide insight into conditions that promote cannibalism.
Overall, this work provided important insight into understanding the potential and feasibility of
stock enhancement with common snook stocks in Florida, and provides a better understanding of
snook ecology, recruitment patterns, and characteristics of essential fish habitat for snook.
25
CHAPTER 2 INFLUENCE OF REARING HABITAT ON GROWTH, SURVIVAL, DISPERSAL, AND RELATIVE STOCK CONTRIBUTION OF COMMON SNOOK Centropomus undecimalis
Introduction
In fish, individuals often make ontogenetic habitat shifts over their lifetime to maximize
growth and overall fitness (Werner and Gilliam, 1984). For example, catadromous fishes spawn
in marine waters, but rear in neashore and often lower-salinity environments making them ideal
organisms for investigating ecological and physiological consequences of habitat shifts and the
evolutionary traits acquired to deal with those consequences. In juvenile stages, predation threat
is primarily linked to body size (Werner and Gilliam, 1984) and is a primary motivational driver
behind behavioral tradeoffs between spending time in refuge habitat or risky foraging to grow
(Werner and Gilliam, 1984; Walters and Martell, 2004)
Foraging arena theory (Walters and Juanes, 1993; Walters and Martell, 2004) states that
predation refugia are the principal microhabitats occupied by young recruits and that fish
abundance is strongly mediated by food availability within and near refuge, with higher mortality
in foragers that move out of refugia. Variation in food resources, forager density, predation
threat within and outside of refuges, and interactions among these can strongly influence growth
and mortality rates (e.g. Connel and Jones, 1991; Tupper and Boutilier, 1997; Walters and
Kitchell, 2001; Bystrom et al., 2004 ). High forager density can reduce localized food
availability, and increase competition, foraging time, and subsequent predation risk (Walters and
Juanes 1993, Walters and Martell 2004). As a result, increased density leads to both reduced
growth and increased mortality (e.g. Love et al 1991, Persson and Elkov, 1995, Jones 1997,
Madon et al. 2001, Seitz et al. 2005). Foraging arena theory predicts greater survival for post-
larvae and juveniles that recruit to refuge habitats with higher forage availability and reduced
predation threat, than in habitats with lower food availability and greater predation threat.
26
This study, conducted in Sarasota and Manatee counties of southwestern Florida, USA,
explores the influence of settlement habitat (via releasing hatchery-reared juvenile snook into
various habitats) on growth, survival, and dispersal and recruitment to adult demes of common
snook Centropomus undecimalis. Common snook (herein referred to as “snook”) are a tropical
to sub-tropical estuarine-dependent piscivorous fish found along the Atlantic Coast of the
Americas. In Florida, snook stocks are subjected to intense recreational fishing pressure (Muller
and Taylor, 2005) and have suffered dramatic coastal habitat alteration (Bruger and Haddad,
1986). Snook are also sensitive to cold waters (Storey and Guder 1936; Marshall 1958; Shafland
and Foote 1983) and Florida represents the northern limit of their latitudinal range (Marshall,
1958). As a result, Florida’s snook populations undergo annual temperature-dependent range
expansions and contractions. During mild winters, snook larval are dispersed north of the 29th
parallel and successfully survive (e.g. Linton and Richards, 1965), but during harsher winters,
cold water (<14 º C) generally causes snook mortality north of this (Storey and Gruder, 1936;
Shafland and Foote, 1983). In Sarasota and Manatee counties, Florida (latitude range = 27º 7’ to
27º 28’ N) snook sometimes suffer mortality during cold winters, but local populations find
thermal refuge in warmer tidal creeks (Marcinkiewicz, 2007; N. Brennan, unpublished data).
Snook spawn in high salinity (>28 ‰) seawater in inlets and tidal passes of estuaries, mouths of
rivers and canals, and off sandy beaches primarily from May - August (Marshal, 1958; Volpe,
1959; Tucker and Campbell, 1988; Taylor et al., 1998). Hatching occurs within 17-18h after
spawning (Lau and Shafland, 1982; Tucker, 2003), and within two weeks recruitment to
nearshore habitats is thought to occur by larvae taking advantage of a two-layered estuarine
circulation system (Norcross and Shaw, 1984; Tolley et al., 1987) and vertical positioning within
the water column (Peters et al., 1998b). Post-larval snook are found mostly along vegetated
27
shorelines of shallow brackish tidal creeks, canals and lagoons in both low- (riverine) and high-
salinity (mangrove swamp and salt marsh) habitats (Tolley et al., 1987; Peters et al., 1998a).
Juvenile snook rearing habitats are diverse, but include estuarine streams and canals, barrier
island and mainland marshes. Juveniles occur in salinities between 0 and 40 ‰, along
moderately sloping shorelines, in basin depths of 1 m or less, mud or sand substrate, and various
forms of shoreline vegetation and structure (Harrington and Harrington, 1961; Fore and Schmidt,
1973; Gilmore et al., 1983; McMichael et al., 1989; Peters et al., 1998a). In Sarasota and
Manatee counties, snook rearing habitats are located in tidal creeks and estuarine backwater
habitats. During winter, snook of all sizes are abundant in these habitats (especially tidal creeks)
as they probably serve as thermal refugia (N. Brennan, unpublished data). During winter,
McMichael et al. (1989) sampled at fixed stations within age-0 snook rearing habitats in Tampa
Bay, have evidence for small-scale repositioning of the cohort during cold winter months. After
winter, Gilmore et al. (1983) hypothesized that adolescent snook from east-central Florida
populations (150 – 400 mm SL, mean = 240 mm SL) undergo an ontogenetic habitat shift to
seagrass beds. Adult snook are also known to disperse to seagrass beds, estuarine inlets, and
beaches by late spring and early summer (Marshall, 1958; Gilmore et al., 1983; Taylor et al.,
1998).
In this study, I tested the following hypotheses: (1) Snook juveniles remain in their
settlement habitat until maturation. (2) Snook that disperse out of settlement habitats return to
those same habitats, suggesting either homing behavior or short dispersal distances to nearby
habitats. (3) Mangrove habitats along shorelines of barrier islands afford as good a foraging and
refuge habitat for juvenile (late age-0, age-1) snook as do mangrove habitats in tidal creeks. (4)
28
Relative contribution of juvenile habitats to adult abundance is proportional to relative
abundances of young snook cohorts in those habitats.
Using hatchery-reared snook as proxies for wild snook (about 8-11 mo. old at release), I
tested the hypotheses by comparing growth, survival, and dispersal patterns in the rearing
habitats (i.e. release habitats). I also examined how these early-on patterns in the various rearing
habitats were reflected in long-term recruitment to adult stocks. The release of hatchery-reared
fish into the wild can be a useful approach to test ecological hypotheses (Leber et al., 1995;
Leber, 1999; Miller and Walters, 2004; Lorenzen, 2005). Although hatchery fish generally
suffer greater initial post-release mortality (within weeks) than wild fish (Lorenzen, 2000;
Fleming and Petersson, 2001; Jonsson et al., 2003; Brennan et al., 2008), by contrast no
consistent differences are reported for growth (Svåsand et al., 2000; Bohlin et al., 2002; Jonsson
et al., 2003; Lochet et al., 2004; Dahl et al., 2006). In this study, I assumed that the relative
growth, survival, and recruitment patterns, established early on in rearing habitats, was reflective
of wild juvenile snook found in the same locales, and reflective of eventual recruitment patterns
to adult stocks.
Methods
Experimental Design
I used a replicated release-recapture experiment to test the effects of release site, and
size-at-release (SAR), on growth, survival, dispersal, and recruitment to adult populations.
Hatchery-reared juvenile snook (80-270 mm fork length [FL], 8-11 months old) were tagged and
released into various coastal habitats in Sarasota and Manatee counties, in southwest Florida,
USA (Table 2-1). Release sites were categorized into sub-systems along a salinity gradient of
typical tidal creek (system) (Table 2-2): (1) upper reaches of the creek with typically low-salinity
29
water (“U”), (2) midstream reaches in brackish water habitats (“M”), (3) lower reaches in higher
salinity waters near the creek mouth (“L” [lower site]), and (4) along barrier island high salinity
habitats (“I”) nearest the respective tidal creek (Table 2-1, Table 2-2, Figure 2-1). Experimental
releases in sub-systems were repeated across the following release systems (or tidal creeks):
South Creek (SC), North Creek (NC), Phillippi Creek (PC), Hudson Bayou (HB), and Bowlees
Creek (BC), (Table 2-1, Figure 2-1). Releases also occurred at Tidy Peninsula (TIDY) a high
salinity mainland mangrove-lined peninsula and a nearby barrier island.
Study Habitat and Timeline
The experimental area encompassed about 30 km of coastline located in Sarasota and
Manatee Counties, on the west coast of Florida, USA (Figure 2-1). Stocking occurred during
April 1998 and 1999 (Table 2-1). In 1998 we stocked the four sub-habitats of four tidal creeks
SC, NC, PC, and BC, plus two subhabitats of TIDY (Table 2-1). In 1999 we stocked the same
sites as in 1998, but replaced PC with HB, a high salinity developed bayou and its adjacent
barrier island habitat (Figure 2-1), and stocked only one sub-system at TIDY. Releases in 1999
were replicated with a second pulse of releases at the same sites two weeks later (Table 2-1).
Barrier island release habitat (sub-system=I) were characterized by high salinity water
(typically 30-34 ‰), with mangroves vegetating the shorelines (i.e. Rhizophora mangel,
Avicennia germinans, and Laguncularia racemosa). Water depths along the shoreline were
generally 1 - 2.5 m at low tide (tidal variation at all sites was approximately 0.6 m), with
substrates of sand, shell, and mud. At the tidal creek mouths (sub-system=L) salinities were
typically more variable due to seasonal rain and tidal flux, but generally around 25 - 32 ‰.
Here, shoreline vegetation was also primarily mangroves. Water depths at the mouth of NC
(non-dredged habitat) were generally < 0.5 – 1.0 m, but 1.0 – 2.0 m at the other sites (dredged).
Substrate was comprised primarily of mud, sand, and live oyster bars (Crassostera virginica).
30
Midstream creek habitats (sub-system=M), are influenced by tidal flux and creek outflow, and
waters were generally less saline. Mangrove, live oak (Quercus virginiana), Juncus sp., and
Brazilian pepper (Schinus terebinthifolius) predominated shoreline vegetation. Water depths in
NC and SC (primarily non-dredged habitat) were generally <1.0 m, and at PC and BC (dredged
habitat) were about 0.5 - 2.0 m. Substrate at the midstream sites was primarily mud, clay and
sand mixtures. Low salinity waters (generally 0-15 ‰) typified upstream tributary habitats (sub-
system = U), but these sites were still influenced by tidal flux. Shoreline vegetation consisted of
some mangrove, but more live oak, leather fern (Acrostichum aureum), Juncus sp., and Brazilian
pepper. Water depths were generally <1 m (except for BC; about 1 - 2 m deep) and substrate
was primarily mud, sand, and some clay.
Hatchery Fish Production, Tagging and Release
Wild adult brood stock snook were captured annually from May – September with large
seines (140-m to 160-m long, 3-m deep, 3-cm multifilament nylon mesh) along barrier island
passes and beaches where snook aggregated for spawning activities each new and full moon.
Snook were held in floating net pens, then individuals were sexed, and eggs and milt were
collected by gently squeezing the abdominal area. Snook eggs were fertilized with milt in 250
ml beakers, transferred to pre-filtered sea water in oxygenated plastic bags, and transported to
2,200 l recirculating seawater tanks (salinity = 28-32 ‰) at the laboratory. Tank salinity was
gradually lowered to 3-6 ‰ by 30 d after hatching. After this, juvenile snook were sorted by size
on a regular basis to prevent cannibalism and reared until tagging (Main, unpublished data).
Snook were tagged according to release site, release lot (within-year replication over
time), and SAR (small 80-129 mm FL, medium 130-179 mm FL, and large 180-270 mm FL).
Tagging occurred in April 1998 and 1999 (about 8-11 mo. after hatching for each respective
release year). Individual snook were tagged with single-length coded-wire tags (CWT, as in
31
Brennan et al., 2005) to identify information about release batches (e.g. release site, release date,
and size class, but no unique information on individuals however). These snook were also fitted
with visible implant elastomer marks as a simple external identifier of hatchery-reared snook
(Brennan et al., 2005). During tagging, sub-samples of snook from each release group were
taken to obtain mean lengths and weights according to release batch. After tagging, juveniles
were held in tanks with recirculating brackish water for 1-24 d until release.
On each release date, snook were taken from their tanks, checked for tag retention, and
then loaded into transport tanks. Transport to the release sites took between 2–5 h depending
upon access conditions at the release sites. At most upstream and midstream release sites, snook
were transported in aerated 150 l coolers, or 800 l hauling tanks by truck, whereas at the stream
mouth and island habitats snook were transported by boat in 800 – 1200 l tanks. If necessary,
transport salinities were gradually adjusted en route to match release site salinity. Snook were
either released directly from transport tanks, from 75 l containers, or 3.5 m long canoes partially
filled with water into shoreline habitat. At each release site, snook were stocked along 100-300
m of shoreline during daylight hours, between 0900 - 1400. The number of fish stocked varied
among size-classes and sites, but was generally between 80 and 300 (fish per date, site, and size
class: Table 2-1). Stocking also occurred in 1997 and 2000-2004 in similar creeks as part of
other stock enhancement studies (Table 2-3).
Collection of Environmental Variables at Release Sites
Aside from the experimentally imposed habitat classifications, I collected environmental
variables at each release site to quantify differences among the release habitats aid in predicting
growth and survival rates of juvenile snook. I measured the following habitat parameters from
three collection sites at each release site: bank slope, substrate softness, percent organic
composition and percent moisture of the substrate, percent shoreline composed of vegetation,
32
and percent shoreline composed of mangrove vegetation. For standardization purposes, I took
samples at perpendicular distances away from the mean high tide elevation indicated on the
bank. I estimated mean high tide elevations on the bank as the upper limit of barnacle, limpet, or
algae distribution on shoreline structure. Shoreline slopes (rise/run) were calculated using
vertical measurements to the bottom at 1m perpendicular from the bank starting at the mean high
tide elevation. Substrate softness was measured 2 m perpendicular from the bank’s mean high
tide elevation. To do this a 0.75 cm2 x 1.5m cylindrical fiberglass rod was held in a vertical
position, resting on the substrate. Then, a 4.1 kg standard weight was placed on the top-side of
the rod and the subsequent penetration distance (cm) of the rod into the sediment was measured.
Substrate core samples (3.8 cm diameter x 10 cm depth) were also collected 2 m perpendicular
from the banks mean high tide elevation at each site. Substrate cores were analyzed for percent
organic material and moisture (by weight; Lenore et al., 1998). I also visually estimated the
percent coverage of macro-vegetation (e.g. trees, grasses, shrubs) along an approximately 100 m
stretch of the shoreline centered on the collection site. Similarly, for the same stretch of given
shoreline, I also estimated the percent coverage of mangrove vegetation (i.e. Rhizophora mangel,
Avicennia germinans, or Laguncularia racemosa).
From each core sample, I removed an approximately 15 ml subsample after thorough
mixing to obtain a homogeneous mixture. Each subsample was placed in an aluminum pan,
weighed (mg), and then dried at 105oC for 24 H. After drying, subsamples were reweighed (to
obtain a percent moisture content), then ashed at 550oC and weighed again to obtain the
percentage of organic material in the sample.
I also collected water quality measurements (temperature [0C], salinity (‰), dissolved
oxygen [mg/l], and pH) with a YSI 556 handheld multi-parameter model (YSI incorporated,
33
Yellow Springs, Ohio, USA) at a minimum of three seine haul locations within each sub-habitat.
These were collected during standardized sampling events within the first month after each
release. For these sites, water quality measurements were taken both at the surface and bottom
of the water column.
Short-Term Collections
After the 1998 releases, we performed standardized sampling efforts in May, June, July
and November 1998, and in January and March 1999 (Table 2-4). After April 1999 releases, we
sampled in May, June, July, August, November 1999, and January 2000 (Table 2-4). From May
– October 1998, a standardized sampling effort (one unit) consisted of nine hauls with a 21.3 x
1.8 m bag seine (1cm nylon mesh) and 18 casts with a 2.4m radius cast net (1 cm monofilament
mesh) at each release site. During November 1998, we modified the sampling effort and gear at
M, L, and I sub-systems to increase tag returns. At these sub-systems we performed six hauls
with a 45.7m x 3m bag seine (1cm nylon mesh) and three hauls with a 21.3 x 1.8 m (1cm nylon
mesh) bag seine. We did not sample with a cast net. Because of the narrowness of the streams at
the upstream sites (sub-system=U), use of the 45.7m x 3m bag seine was impossible, so we
continued sampling with our earlier sampling regimen of cast nets (18 casts) and a 21 m seine (9
hauls). Because of this change in methods, I only used 21m standard seine data for comparisons
between upstream and other sub-systems, and only compared recapture ratios from data collected
during winters for year-to-year comparisons.
Long-Term Collections
Recaptures over the long-term (defined as recaptures at large < one year) were collected
from 1999 – 2006 to validate patterns observed in the rearing habitats (i.e. growth and recapture
ratios), and to document relative contribution rates of stocked juveniles in the various rearing
habitats to adult stocks. Generally, snook do not recruit to adult spawning habitat until age 3,
34
when they can be found in high concentrations in coastal passes and along beaches during
summer months (Taylor et al., 1998). Two year old snook have also been documented with
mature gonads (Taylor et al., 1998) and snook 150-400 mm SL (typical age two sizes) have been
documented to move to seagrass habitats during spring and early summer (Gilmore et al., 1983;
Brennan, unpublished data). Therefore I used data from annual collections at beaches (barrier
island coastlines along the Gulf of Mexico), barrier island passes, and intercoastal seagrass
habitats. Sampling at beaches, passes and seagrass habitats was conducted with seines (137 m
long x 2.4 m deep, with 5 cm multifilament nylon mesh), cast nets (4 m radius, 5 cm
multifilament mesh) and trammel nets (137 m long x 2.4 m deep, 3 cm inner nylon mesh, and 30
cm outer nylon mesh). We also performed standardized sampling efforts at randomly selected
coastal sites (mainland and barrier islands, about 1.6 kilometers of shoreline).
In September 2001 and July 2005, red tide outbreaks caused mass mortality of snook
along beaches and passes. During these events, we scanned individual carcasses for CWTs. I
also used data from snook captured in an annual angling tournament specifically created and
designed to evaluate contribution rates of hatchery-released snook to localized adult stocks.
These tournaments were conducted from 1998- 2005, and anglers were directed to fish between
Anna Maria Island and Venice Inlet (Figure 2-1, Table 2-5). Any snook, captured by an angler,
was measured (FL mm) and scanned for CWTs. Tagged snook were kept for CWT extraction,
untagged snook were released alive.
For analysis of long-term data, I removed migrational bias from recapture ratio
comparisons by only using recaptures from beaches and passes (adult snook spawning habitat)
and those captured along shoreline habitat at least 1 km away from any given release site
(because sampling at locations near release sites could result in higher recapture ratios of snook
35
released nearby). Hatchery snook captured within the same creek as their release creek were not
included in the analysis. I considered recaptured snook that had met these criteria as “dispersed”
and were used in the recapture ratio analysis.
Sample Processing and Data Analysis
All captured snook were counted, measured, and checked for the presence of CWTs with
magnetic CWT detectors (Northwest Marine Technology Inc., Seattle WA). Untagged snook
were released, while tagged hatchery-reared snook were preserved on ice and returned to the
laboratory where lengths and weights were taken. Coded-wire tags were extracted from tissue
and read using a binocular microscope (at 60x). Each tag was read twice by independent readers
to verify tag codes. Recaptured fish’s tag codes were linked to release data through a release-
recapture database system using Microsoft Excel, Microsoft Access, and SAS (Brennan, 1997).
Recapture ratios for each release group (i) were calculated as:
Recapture ratioi = (number recapturedi /tag retentioni) / number releasedi (2-1)
where tag retentioni is the proportional tag retention of fish from group (i). I assumed CWT
retention estimates at the time of release to be valid after release thereafter because other studies
have shown that loss of CWTs implanted in muscle tissue was insignificant after 30 d
(Blankenship, 1990; Dorsey 2004; Meeuwig et al., 2007; Brennan et al., 2007), and in our
studies with snook (Brennan et al., 2005), CWT retention estimates one year after tagging, were
not significantly different from initial CWT retention estimates taken 1-30d after tagging.
Data on short term growth rates (mm/d) were collected from snook recaptured during the
first winter after release (6-10 months after release). To calculate growth rates, I subtracted
mean SAR (FL mm, based on tag codes) from individual recapture lengths (FL mm) divided by
days at liberty. Because of the short-time period between release and recapture dates, I assumed
that a linear relationship (slope = mm/d) was adequate to describe growth during this period.
36
To examine growth over the long-term, I fit von Bertalanffy growth models to length and
age data of snook released in different release habitats (e.g. BC, NC, or islands, Table 2-2) with
least squares regression techniques using Microsoft Excel and the solver add-in. Age of
recaptured snook was estimated by adding 300 d to the days since release, because snook were
about 10 months old at the time of release. For this analysis I only used data for recaptures more
than 180 d since release (to ensure that potential habitat influences on growth had occurred) from
release experiments in 1997-2004 (Table 2-3 for release information).
Dispersal distances were calculated as the minimum in-water distance (km) between
release and recapture sites within the confines of the water body, measured to the nearest 0.1 km.
Because snook spawn on an annual basis (generally during summers), I was able to use length
frequency distributions to differentiate between age-0 (or young-of-year snook), age-1 (those
spawned the summer prior to the age-0 group, same age class as the released snook), and older
snook. I then calculated catch-per-unit-effort (CPUE) of the age-1 cohort captured during
standardized sampling at the release sites during the first summer and winter after each release.
Age-1 CPUE was calculated separately for each standard seine (21 m and 45 m).
Multivariate analyses: To understand how various habitat characteristics were related
to the selected release sites and how these related to short-term responses in the released snook, I
performed a principle component analysis also using SAS procedures (SAS, 1999). The PCA
was based on the correlation matrix of environmental variables collected at the release sites (see
above section), estimates of mean CPUE for age-1 snook collected the time of release (May-July,
summer) and during the first winter after the release (November), and first winter mean growth
rates and recapture ratios of the released fish (summarized in Table 2-6). To quantify the
relationships between release variables (SAR and release site) and response variables (size at
37
recapture, mean dispersal distances from release sites [summer and winter separately], first-
winter growth, and first-winter recapture ratios, I performed a canonical correspondence analysis
using SAS (SAS, 1999). For this, recapture data were organized according to mean SAR (FL
mm) and release site. I also conducted a second principle component analysis to examine how
components of the rearing habitats and short-term patterns in growth and recapture ratios were
related to long-term response patterns in the released snook (long-recapture ratios and growth
variables, Table 2-7).
Predicting contributions of rearing habitats to adult stocks: Because first winter
growth and recapture ratios were strongly correlated with long-term growth and recapture ratios,
I used these data as the basis for predicting contributions of rearing habitats to adults stocks. I
used a linear function of loge (ln) transformed first winter growth rates (growth, mm/d) to predict
loge transformed first winter recapture ratios using least squares procedures and the solver add-in
feature in Microsoft Excel. While recapture ratios were only reflective of relative survival rates
for any particular release group, I used data from a subsequent study (Chapter 3; Brennan et al.,
2008) conducted in 2002 in many of the same habitats to provide estimates of real survival rates
of the released snook. In Brennan et al., (2008, Chapter 3), abundance of age-1 snook was
estimated prior to, and subsequently through the first year after release. While post-release
survival is a function of ambient density and stocking magnitude (among other things), I used
estimates of survival to first winter from creeks that received similar stocking magnitudes (about
1000-2000 stocked) as the releases during 1998 and 1999. Therefore, the “high augmentation”
creeks from the 2002 releases (BC and Whitaker Bayou [WB]) were similar in magnitude to the
1998 and 1999 release years. Of these two creeks WB had very little loss of hatchery snook by
the first winter (about 0.85 of the snook released in May survived to first winter) and I used this
38
rate as a best case survival scenario for released fish. An important difference in 2002, however,
was that we used an in situ acclimation release technique which has been shown to
approximately double the survival rates of stocked snook (Brennan et al., 2006). Therefore, to
standardize with recapture ratios from 1998 and 1999 releases (non acclimated releases) I used
half this rate (i.e. 0.42 for survival to first winter) as a best case scenario. Highest recapture
ratios from 1998 and 1999 were set at 0.42 and others were down-scaled accordingly. After
obtaining abundance estimates at first winter, I used 0.9 as the proportion survived from first
winter to second summer, and 0.7 for subsequent annual survival rates.
To project biomass contributions of juvenile snook released in specific rearing habitats to
adult stocks in subsequent summers, I first plotted growth rates at first winter versus the
associated Linf parameter (FL mm) from von Bertalanffy growth functions specific to release
habitats (Table 2-8). Subsequent parameters (k and t0) for predicted von Bertalanffy growth
functions were calculated as linear relationships to Linf. After this, I calculated length (FL mm)
at age for 2 – 5 summers after the release. Lengths (FL mm) were converted to body weight
(Wt, g) using the function,
Wt = 0.000009*(FL)2.9892 (2-2)
with parameters fit from length and weight data from Sarasota snook recaptures (Brennan,
unpublished data, FL =176-816 mm, n=738) by least squares procedures using Microsoft Excel
and the solver add-in. Individual weight predictions for the respective summers were multiplied
by numbers-at-age estimates to generate biomass estimates. For this analysis I ignored
differential fishing mortality normally subjected to snook of these sizes recruiting to the fishery.
Results
In April 1998 and again in April 1999 a total of 21,392 juvenile snook were tagged and
released (Table 2-1). Mean tag retention in the laboratory by release year and lot varied between
39
83.3% and 97.2%. From April 27, 1998 to August 5, 2005, we sampled 21,304 snook along
shorelines and estuaries from Venice Inlet to Anna Maria Island, in Sarasota and Manatee
counties, representing about 40 km of coast. Of these, 480 were hatchery snook recaptures from
snook released in 1998-1999: 437 contained CWTs that were decoded, 18 CWTs were lost
during processing (3.76%), 7 recaptures contained CWTs but were released alive after recapture,
and 18 snook had only VIE tags. Of the 437 tags recaptured and decoded from this study, 202
were from the April 1998 releases and 235 from 1999. Recapture data described below are only
from deciphered tag codes.
Short-Term collections at Release Sites
General: Sampling at release sites from April 27, 1998 – April 30, 2000 yielded 3,085
snook. Of these 262 (8.49% of the snook catch) were hatchery reared tagged snook: 93 were
from releases in April 1998 (0.78%), and 169 from April 1999 (1.78%). Overall during the first
month after release, hatchery snook from each respective release represented 25.4 - 27.4% of the
snook catch, but gradually declined to 0.0 - 9.0% 6-8 months later. Recapture ratios also
declined over time but were between 0.01% – 0.05% per 21m seine haul within the first month
after release (Figure 2-2).
Highest mean recapture ratios were from fish released in upstream and midstream sub-
systems) (Figure 2-3). Fish released in North Creek had recapture ratios about twice as high as
any other tributary in both 1998 and 1999 even though sampling effort was similar among all
sites (Figure 2-4). Also mean recapture ratios of snook released in creeks were higher than those
of snook released along islands. Recapture ratios were generally correlated with age-1 snook
abundance (CPUE 21 m seine) (Figure 2-5).
Overall, released snook showed high fidelity to their particular release site. Generally,
60% of the recaptured snook were recaptured at their release site and 89% were no farther than 5
40
km from their release sites (Table 2-9). For snook released in BC, PC, NC, and SC, 30.5%
(n=262) were recaptured at other release sites within the first year after release, but no clear
patterns of upstream or downstream movement were apparent. For the same period, only nine
(of 262 recaptured, 3.4%) were captured outside of their respective creek within the first year
after release. For the barrier island release sites, only one recapture (of 16 total, 6.3%) was
found away from its release site. The farthest migration documented within the first year after
release was from a fish released at TIDY that emigrated to the upstream section of NC, 24 km
away.
Generally, first winter mean growth rates showed strong patterns of differentiation
according to release site (see below for statistical analysis). Highest mean growth rates were
found in snook released at barrier island sites (0.67-0.78 mm/d) followed by BC (0.67 mm/d),
TIDY (0.63 mm/d), HB (0.52 mm/d), NC (0.44 mm/d), SC (0.44 mm/d), and PC (0.41 mm/d)
(Figure 2-6). Sizes-at-recapture during the first winter showed similar patterns with largest sizes
from snook released at islands, TIDY, and BC, and smallest sizes at recapture from NC and SC
(Figure 2-7). There was no apparent relationship between SAR and first winter growth rates,
however (Figure 2-8, and see below).
Notable differences were found in patterns of snook size structures among the release
sites and sub-sites throughout the seasons (Figure 2-9). Juvenile snook were common
throughout the year in upstream, mid-stream, and stream mouth sub-habitats of tidal creeks, but
rare at barrier islands except during summer and fall months (170-300 mm FL, Figure 2-9).
Large snook (>350 mm FL) predominated frequency distributions at barrier islands throughout
the year, but were most abundant here during summer and especially fall months. During winter
and spring, relative abundance of large snook increased in the creeks particularly at stream
41
mouth habitats compared to in-creek abundance during summer and fall. Among creeks, length
frequencies of snook were very similar with respect to sub-habitats.
Multivariate analyses with short-term responses: The principle component analysis of
release site variables and short-term post-release responses showed that the first two principle
components explained 27% and 21% of the standardized variance and 67% was explained by the
first three principle components. A plot of the first two principle components (Figure 2-10)
shows that recapture ratios and age-1 densities were inversely related to growth rates at first
winter (Figure 2-10, Figure 2-11, Figure 2-12). High CPUE and recapture ratios were positively
related to sediment softness and moisture content but negatively related to growth, salinity, pH,
dissolved oxygen, and mangrove coverage (Figure 2-10, Figure 2-11). Release sites with high
CPUE and recapture ratios, but low growth included NCU, NCM, and NCL. Release sites with
low CPUE, low recapture ratios, but high growth, salinity, dissolved oxygen and mangrove
coverage included SCA, NCA, TIDY_I, and TIDY. Among our release sites, bank slopes
(steepness) were inversely related to vegetative coverage, sediment organics, and temperature.
Considering bank steepness and vegetative coverage, BCU had steep banks and low vegetative
coverage, but PCL, PCA, and NCL had more gradual bank slopes and more vegetative shoreline
coverage.
The canonical correspondence analysis showed that mean SAR according to release site,
was most correlated to summer (0.77) and winter (0.73) recapture lengths, but not well correlated
with mean dispersal rates during summer (0.33) or winter (0.28), or first winter growth (0.08), or
recapture ratios (-0.33). A plot of the canonical variables for SAR and the response variables
(Figure 2-13) reflects the strongest correlations between SAR and recapture lengths but much
less correlation between the other responses. While canonical variables were not calculated for
42
all release sites (because of missing response data), the canonical variables for TIDY, SCL, and
HBL showed strongest correlation to SAR, but snook released in North Creek showed no
associations with SAR (Figure 2-13). Just examining SAR and dispersal distances however,
snook released in the largest size class (and larger size-at-recapture) showed patterns of greater
dispersal distances compared to the smaller size classes (Figure 2-14).
Long-Term Assessment
General: From August 1, 1998 – August 31, 2005 sampling at beaches, passes, and
barrier island habitats and sampling at release sites from April 1, 2000 to August 31, 2005
resulted in an additional 18,344 snook captured (Table 2-5). Of these, we collected 2,883 snook
from beaches and passes, 1,038 from random sampling at barrier island sites, 1,453 snook from
red tide collections, 1,063 from local snook tournaments, and 11,907 snook from sampling at
release sites. Of the snook released in 1998 and 1999 (at upstream, midstream, stream mouth,
and along island shorelines), 174 were recaptured after over 1 yr (0.95% of total snook captured).
Most long-term recaptures were captured at their release sites (n=132), but 42 had “dispersed”
(i.e., moved >1 km). Overall, the largest recaptured hatchery snook was 816 mm FL (34 inch
total length) and weighed 5,492 g (12.1 pounds). The longest time at liberty was 2,399 days
(6.57 years).
Long-term Growth: I obtained age estimates of 514 hatchery-reared snook at large for 6
mos. or longer (maximum age 7.39 years). Of these, four release habitats had over 40 recaptured
snook (5-7 age cohorts, n=59 recaptures from BC, and n=249 from NC) and I used these to
generate age-at-length von Bertalanffy growth models (Table 2-8). Fastest growth was
predicted for snook released at barrier island habitats (Linf = 824), followed by snook released in
Bowlees Creek (Linf = 704), South Creek ((Linf = 657), and North Creek (Linf = 552) (Figure 2-
15).
43
Dispersal: Generally mean dispersal distances increased over seasons and years (Figure
2-16). Snook recaptured during summer and fall had the highest mean dispersal distances, and
were rarely recaptured in tidal creeks during these seasons, although sampling occurred regularly
there. Mean dispersal distances during winter did not increase across years however, and
showed high rearing site fidelity each winter (Figure 2-16). Snook tended to recruit to adult
spawning habitat nearest their rearing habitats.
Mulitvariate analysis with long-term responses: The second principle component
analysis comparing key features of release sites and short-term post-release responses with long-
term responses (Table 2-7) showed the first three components explained 80% of the standardized
variance (36%, 28%, 16% respectively). Examining a plot of the first two components (Figure
2-17), age-1 summer CPUE, age-1 winter recapture ratio, and long-term recapture ratios (from
dispersers and those captured at beaches and passes) were all within the upper left quadrant.
Short- and long-term growth patterns were closely related and generally inversely related to
recapture ratios.
Contribution to adult stocks: Least squares procedures predicted recapture ratios as a
linear function of loge (ln) transformed first winter growth rates (growth, mm/d) as:
Recapture ratio = -1.9596* Ln (growth) - 0.2578 (2-3)
The correlation coefficient for this function was 0.59. Predictably, plots of numbers of hatchery-
released snook recruiting to adult stocks according to age and rearing habitat (expressed as short-
term growth rates), shows highest contributions from rearing habitats with lowest growth rates
and highest post-release survival (Figure 2-18a). Recruitment of stocked juveniles from rearing
habitats to age-2 adult stocks (Nage-2, to second summer) were expressed as:
Nage-2 = -286 * (ln(growth)) + 81 (2-4)
44
Based 1000 hatchery-reared juveniles stocked in rearing habitats during early summer and the
above assumptions, Creeks such as NC and SC (with first winter growth rates of 0.44 mm/d),
were predicted to produce about 316 age-2 snook while a barrier island release site stocked at the
same level would produce about 171 age-2 snook (Figure 2-18a). Under these conditions, and
not accounting for fishing mortality, the numbers of age-4 snook produced by NC or SC would
be about the same as the number of age-2 snook produced by an island habitat stocked at the
same level. Total biomass estimates of hatchery-released snook recruiting to adult stocks based
on rearing habitat growth rates showed a hump-shaped relationship (Figure 2-18b). Rearing
habitats with intermediate first winter growth rates of stocked snook produced the highest
biomass of adult snook. Biomass production of age-2 hatchery snook from rearing habitats as
NC or SC with slow growth rates produced about the same biomass of age-2 snook as did island
sites where growth rates were much higher (Figure 2-18b). By age-4, however, about 60% more
biomass was predicted for NC or SC sources compared to island sources.
Discussion
In this study I used stock enhancement to understand important life history characteristics
and habitat interactions of juvenile snook stocked into various rearing habitats and examined
contribution rates of the stocked juveniles to adult stages. Due to the vulnerability of Florida’s
snook stocks to overfishing (Muller and Taylor, 2005) and widespread modification of coastal
rearing habitats (Bruger and Haddad, 1986; Lorenz, 1999; Davis et al., 2005b; Lorenz and
Sarafy, 2006) investigations of stock enhancement potential is warranted. Notwithstanding, the
controversial nature of this measure coupled with the status of snook stocks in Florida (Muller
and Taylor, 2005) emphasizes the need for a cautionary approach for stock enhancement
research (Blankenship and Leber, 1995; Walters and Martell, 2004, Lorenzen, 2008). This
45
study, operated on an experimental level, and coupled with extensive post-release evaluation,
provided valuable information on juvenile snook life history characteristics tied to various
characteristics of rearing habitats and eventual contribution of these fish to adults stocks.
Releasing juvenile snook into varied habitats showed evidence of tradeoffs between
growth and survival. Sites where snook growth rates were highest also had comparatively low
recapture ratios, suggesting low survival. Conversely, suppressed growth was exhibited at sites
with high recapture ratios (high survival) and high overall densities (CPUE) of juvenile snook.
Generally, habitats with higher ambient densities of wild juvenile snook correlated with higher
survival of hatchery-reared snook, but at the cost of density-dependent growth. This suggests
superior refuge habitat and overall lower predation threat, but also high competition for prey
items among the juveniles. Beverton-Holt models predict density-dependent mortality with
increasing population size. In fish, adjustments in growth responses usually occur under
predation threat as a probable tradeoff against mortality risk (Werner et al., 1983; Belk, 1995;
Belk, 1998; Krause et al., 1998). In this study, the differential post-release growth rates and
recapture ratios probably reflect variations in refuge quality, predation threat, and competitive
interactions.
The inverse relationship between growth rates and recapture ratios of stocked juvenile
snook was among the strongest quantitative correlations I found. Relationships among release
habitat features and snook production were more complicated, however, and reflects the diversity
of juvenile snook habitat (Gilmore et al., 1983; McMichael et al., 1989; Peters et al., 1998a) and
Florida estuaries (Faunce and Sarafy, 2006). Because mangrove or vegetative shoreline
coverage was not well correlated with recapture ratios or growth rates, other features such as
sediment softness, and organic content are more predictive of snook production in Sarasota
46
rearing habitats. The inverse relationship between bank slope and vegetative coverage probably
reflects common anthropogenic alterations in the form of dredging and vegetative removal.
Rearing habitat production and stock enhancement considerations: Rearing
habitats with characteristic differences in attributes of early-on post-release responses of the
stocked juveniles showed remarkable differences in numerical and biomass contributions to adult
stages. Beck et al. (2001) defined a nursery habitat as those that produce a greater number of
individuals per unit area that recruit to adult populations than the average production of other
habitats where juveniles occur. Others have presented theoretical implications for numerical
contributions of nursery habitats to adult stocks (e.g. Ross et al, 2003; Dahlgren et al., 2006), but
field studies demonstrating this are rare. In fish that attain large body sizes, are highly fecund,
and have variable growth and survival rates (such as snook, Taylor et al., 1998), differences in
numerical and biomass contributions could be substantial. In Sarasota, the model showed that
equal levels of hatchery-reared juveniles stocked into NC and SC (sites with low growth rates)
produced more age-2 individuals per unit area than other habitats, but contributions to age-2
biomass production (i.e. potential egg production) were among the lowest and similar to barrier
island release sites that produced snook with high growth rates, but had low recapture ratios.
Biomass production per unit area was hump shaped (Figure 2-18b) and maximized at release
sites with intermediate growth rates. As age of the recruiting hatchery snook increased, the
maximum biomass began to shift towards rearing habitats with lower growth rates again. Using
this work, resource managers can assess the relative production value (in terms of snook) of
various rearing habitat. Quantification of total habitat area is an important next step to evaluate
overall production according to habitat type.
47
While growth rates and survival are functions of production, I ignored other important
considerations that should not be overlooked. As snook attain larger sizes, fishing mortality
increases and places a disproportionate pressure on larger-sized snook (Muller and Taylor,
2005). This pressure must be accounted for. While adult biomass is an index of potential
recruitment, snook are protandric hermaphrodites (Taylor et al., 1998) and further complicate
stock recruitment models (Muller and Taylor, 2005), it remains an important consideration.
Snook rearing habitats must also be quantified if total production by habitat type is to be
considered (as outlined in Dahlgren et al., 2006). Quantification of rearing habitats might be
accomplished with rough habitat measures based on GIS mapping systems, but some
measurements (e.g. measuring organic content in sediments) would clearly require more effort.
Coupling post-release survival estimates from snook released in similar creeks in 2002
(Brennan et al., 2008; Chapter 3) to this study are not without complication. Post-release
survival and growth are, in part, functions of ambient density of snook juveniles in the rearing
habitats, refuge quality, stocking magnitude, predator and prey abundance, encounter rates, and
feeding rates. Within release sites, I assumed that ambient densities of juvenile snook were
generally the same during the two studies. This might be reasonable because I observed similar
post-release growth responses across the years (e.g. North Creek has had relatively high
recapture rates and relatively low growth rates across the years 1997, 1998, 1999, 2000, 2001,
Brennan, unpublished data) when the same habitats were stocked. Within a given year,
however, snook recruitment can vary across rearing habitats so relative measures across the years
at a particular rearing habitat may not be representative of recruitment in other rearing habitats.
In 2005 red tide blooms (Karenia brevis) in Sarasota Bay may have caused disproportionately
lower recruitment strength in nearby rearing habitats (Brennan, unpublished data). While its
48
difficult account for, physical changes in shoreline habitats (e.g. installation of docks, seawalls,
associated dredging activities, growth or removal of vegetation, or pulses of organics from
upstream sources) may also influence year-to-year results. Increases in experience with this
work (e.g. increases in the number of experimental years, additional stocking levels, and across
habitats) can increase the confidence in our expectations of such activities, but should be coupled
with experimental stock enhancement modeling scenarios (Lorenzen, 2005; Lorenzen, 2008).
Strong localized differences in post-release growth and recapture ratios were found at
release sites as close as 1-km apart. These habitat-specific patterns, established early-on,
continued to be found in snook recaptures for six years after the release. Similar growth and
abundance patterns are seen on a broader scale between snook stocks found along the east and
west coasts of Florida. East coast stocks are generally faster growing, attain larger sizes, but are
less numerous than stocks from the Gulf coast (Muller and Taylor, 2005). This emphasizes how
various features of rearing habitat can influence juvenile snook production to adult stocks.
Although it may be coincidental, in our study area, tidal creeks with highest growth rates but
lowest survival were also highly anthropogenically impacted (shoreline habitats were highly
altered through seawalls, channelization, and creeks were routinely dredged). Habitats along
Florida’s east coast are also considered more developed (Muller and Taylor, 2005). Research is
warranted to further understand the potential effects of differing anthropogenic development on
valuable coastal habitats and their influence on stock-recruitment dynamics for snook stocks and
other important coastal stocks.
Because wild age-1 snook density was strongly correlated to hatchery-reared snook
recapture ratios and growth rates, these relationships probably hold true for wild snook, although
survival estimates of wild conspecifics are probably higher. Hatchery-released animals are
49
generally more vulnerable to mortality compared to wild conspecifics of the same sizes (Mason
et al., 1967; Lorenzen 2006; Jokikokko et al., 2006; Shimizu et al., 2007; Fritts et al., 2007).
This has been attributed to a variety of behavioral and other deficiencies due to rearing in
psychosensory-deprived environments such as hatcheries (Olla et al., 1998, Berejikian et al.,
1999). Experimental releases of hatchery-reared snook in 2002 (Brennan et al.; 2008, Chapter 3)
showed disproportionately high mortality within the first month after release compared to wild
conspecifics. In an earlier study (Brennan et al., 2006), snook acclimated in predator-free
enclosures for 3d had survival rates 1.92x higher than naive snook stocking directly into the
wild. Therefore, highest mortality of stocked age-1 snook apparently occurs soon after release.
Nonetheless, even if subsequent survival rates of wild and hatchery snook are similar, first winter
recapture ratios would continue to reflect the initial post-release mortality. Study is warranted to
distinguish between real survival rates of wild and hatchery snook over time.
Other: Water quality at release sites containing sediments with high percent moisture
and percent organic material was generally hypoxic with relatively low salinity during summer.
Generally these habitat types were found in sections of the tidal creeks (especially NC and SC).
Aquatic habitats containing sediment with high loads of organic material typically experience
hypoxic conditions due to oxygen consumption by microbial activity often most pronounced
during summer (e.g. Kemp et al., 1992; MacPherson et al., 2007). While survival rates and
overall juvenile snook density were generally higher at these sites, snook growth rates were
suppressed. Growth responses were likely influenced by density-dependent competition for
limited food resources (as above), but growth would likely be further suppressed by higher
physiological costs associated with tolerance to hypoxic conditions. This supports physiological
work (Peterson and Gilmore, 1991) demonstrating that small snook (<150 mm SL) are better
50
able to tolerate hypoxic conditions than are larger snook and hypoxic sites may serve as an
important physiological refuge from aquatic predators. Other attributes of the habitats (i.e.
gentle bank slopes, soft sediments, low salinity, and high temperatures) were also typically found
at these sites and other studies characterizing juvenile habitats corroborate this finding (Gilmore
et al., 1983; McMichael et al., 1989; Peters et al. 1998a). Summer time water quality at island
habitats correlated with higher dissolved oxygen levels, higher salinity, pH, and loosely
correlated with steeper banks. Arguably, the low recapture ratios at these sites might reflect
higher predation risk and mortality because more potential predators can tolerate these
conditions. High predation risk could cause size-dependent survival where only the largest and
fastest growing snook survive beyond the vulnerable size range. If this were true, however lower
recapture rates of small SAR snook should have been observed at these sites. I found evidence
that would support this in plots of recapture lengths according to release habitats (Figure 2-7) but
not in plots of growth rates and SAR, or the canonical correlation analysis (Figure 2-8, 2-13).
Within a 42 d window after release, larger SAR individuals showed patterns of higher
mean dispersal distances compared with smaller individuals, and SAR was loosely correlated
with first summer and winter dispersal rates. Other studies have shown that relative predation
risk and dispersal associated with body size (e.g. Werner and Gilliam, 1984; Walters and
Juannes, 1993; Armstrong et al., 1997). Smaller individuals, in general, are at higher risk from
predation than large individuals, especially outside of refugia, and movement is comparatively
reduced. Lorenzen (2000) found a negative, linear relationship between mortality of released
fish and length. In contrast, large fish benefit from improved foraging opportunities therefore
increased growth rates. In this study, patterns of increasing dispersal distances with body size is
only suggestive of lower predation risk in large fish.
51
Size structure analysis from snook collected at the release sub-habitats year-round
demonstrates the importance of tidal creeks as rearing habitats (or nursery habitats), and
important winter refuge for snook of all ages. Generally, tidal creeks served as nursery habitats
for snook throughout the year. During winter, tidal creeks experienced increases in adult snook
abundance, with corresponding declines in abundance at barrier island habitats (highest
abundance during summer and fall). Comparatively high winter abundance of snook in the
creeks and associated temperature profiles suggests that creek habitat serves as a thermal refuge
for snook (Marcinkiewicz, 2007). Furthermore, dissolved oxygen levels in tidal creeks during
summer aregenerally low (e.g. Kemp et al., 1992; MacPherson et al., 2007) compared to other
times of the year, but dissolved oxygen levels along barrier islands are generally high year round.
These patterns imply that juvenile and adult snook’s spatial distribution might also be controlled
by abiotic conditions. In the winter, large-scale aggregation occurs in the creeks where water is
warmer. During summer and fall months, tidal creek waters become hypoxic, and large snook
are more common along beaches, passes (spawning habitats), and grass flats. The year-round
presence of juveniles in tidal creeks (even during summer hypoxic conditions) coupled with
recapture ratios suggests their potential importance as refugia from predation.
Important considerations for release experiments can also be gained from natural size
structures of snook in various habitats and across seasons. Snook are piscivorous (McMichael et
al., 1989; Aliume et al., 1997) and cannibalistic (Tucker, 2003; Adams and Wolfe, 2006; Chapter
4). Resultantly, abundance of larger snook could serve as a proxy for juvenile snook predation
threat. With large snook present, juvenile snook are at high cannibalism risk. Based on size
structures, juvenile snook at barrier island habitat would be subjected to high predation threat
year round. Low abundance of adults in creeks during summer suggests lower associated
52
predation threat for juveniles. Furthermore, relative abundance of wild juvenile snook could also
be used as an indicator of suitable habitat for released fish, although density-dependent effects
should be considered. During winter, snook released in the creeks would be exposed to higher
cannibalism risk due to high snook densities in the creeks at this time (see Adams and Wolfe,
2006; Brennan, unpublished data). These fish would also be only about six months old and
smaller than spring-released snook. Over-winter mortality (e.g. Hurst and Conover 1998) is also
a major concern with snook because of their susceptibility to cold temperatures. For stock
enhancement, these concerns must also be weighed against rearing costs to determine economic
feasibility.
Overall, hatchery-reared snook showed high site fidelity among creeks (or release
habitats) and even for specific locales within creeks (or release sub-habitats). Within the first
year after release, juvenile snook had emigration rates of only 3-6% outside of the release
habitats. This agrees with an earlier study examining recaptures of juvenile snook released in
NC (1% emigration from the creek, Brennan et al., 2006). During subsequent summers and fall
(i.e. spawning months, Taylor et al., 1998), however, mean distances between release and
recapture sites (i.e. dispersal distances) continued to increase with. Young mature snook tended
mostly to disperse to spawning habitat and grass flats nearest their rearing habitats. Older and
larger snook showed increasing dispersal distances during summer and fall. The increasing
dispersal tendencies with increasing age have important implications for stock management.
While young mature snook disperse to local spawning sites, larger snook dispersed more widely.
Because the larger snook are more fecund (Chambers and Leggett, 1996; Berkeley et al., 2004;
Longhurst, 2006; Francis et al., 2007), these fish may have a greater influence on snook
recruitment dynamics and genetic patterns. Widespread and intense harvest of these larger
53
individuals may have negative consequences for snook stocks on both local and regional levels.
During winter, however, mature snook showed consistently high fidelity to rearing habitats
(especially in tidal creeks) demonstrated through consistent use of these sites each winter as they
sought thermal refugia (Figure 2-16). Understanding these snook behaviors and life history
patterns are important for understanding how production and quality of rearing habitats can
influence localized demes and fisheries.
Recapture ratios showed some difficulties when used as indices of survival because of
uncertainties in post-release distribution of the juvenile snook. Our short-term sampling showed
significantly higher recapture ratios from fish released at upstream and midstream habitats than
those released at barrier islands. These results may have been biased in several ways: (1)
Sampling took place at the release sites. Thus, for each release creek system proportionally more
sampling effort occurred within the creeks compared to barrier islands (about three times as
much effort in creek habitat). (2) Barrier island habitats represent a larger and more “open”
habitat than within-stream habitats. Even if snook dispersed randomly, our sampling regimen
would recapture fewer snook released at barrier islands, than within the creeks, because the
creeks were more restrictive, with less overall dispersal area available. (3) Within the creeks, a
general downstream movement of snook may have resulted in disproportionately more
recaptures of snook from “upstream” release locations compared to stream mouth release sites.
Very few juvenile hatchery snook were recaptured outside of the creeks, but this could have been
due to less sampling effort in these habitats. A generalized downstream movement pattern was
not found from our recaptured snook, however. Often, snook released downstream were
captured upstream and midstream, and snook released at the midstream habitats were found in
54
the upstream, midstream and stream mouth habitats. Most often, snook released in a particular
habitat were also recaptured there.
Our long-term sampling program was designed to validate short-term samples at the
release sites. Because snook disperse to adult spawning habitat during summer months, samples
from these habitats should have provided the least biased data for recapture ratio analysis
independent of juvenile release habitat. In this study, of 6,437 snook sampled along beaches and
barrier island passes (where snook were found spawning, Table 3), only 21 snook (0.33%) were
recaptured from these habitats. The low percentage of hatchery snook at large made it difficult
to obtain a large enough sample size. Long-term samples (snook recaptures from any location
that were at sea for over one year) resulted in similar data as the short-term data set with
significantly higher recaptures of snook from midstream and stream mouth habitats compared to
barrier island habitats. Because most of these snook were recaptured in the creeks, geographical
biases may have again influenced these data. These data correlate well with other findings of
high site fidelity in snook. Volpe (1959) and Bruger (1980) conducted snook dispersal studies
using tag returns from anglers and found that 70-79% of snook recaptures were within 8-9 km of
the tag and release site. In these studies angler catches were not particularly linked to the tag and
release sites. This correlates well with our findings of 89% of our recaptures within 5 km of the
release sites. The higher percentage in this study could be because our sampling stations were
centered around release sites and less random than angler’s capture sites.
Conclusion: Florida’s snook stocks are subjected to increasing fishing pressure (Muller
and Taylor, 2005) and coastal rearing habitats have been subjected to widespread ongoing habitat
modification (Bruger and Haddad, 1986; Lorenz, 1999; Davis et al., 2005; Lorenz and Sarafy,
2006). Considering these resource management problems, coupled with considerable uncertainty
55
in understanding stock enhancement potential, empirical research such as this study, can aid in
resolving these issues. In this study I found that rearing habitats where snook were stocked
influenced post-release growth and recapture ratios at first winter and these traits were
perpetuated for years after releases. Generally, post-release growth was inversely related to
recapture ratios and tied to characteristics of the rearing habitats. Sites with highest recapture
ratios and low growth rates (often tidal creeks), generally contained soft sediment with high
organic content and low dissolved oxygen levels in the water. These sites were also correlated
with high ambient densities of juvenile snook. Sites with low recapture ratios and high growth
rates generally had high salinity and pH, firmer sediment, and low ambient densities of juvenile
snook (often barrier island habitat). I tied first winter recapture ratios from this study to real
survival rates of stocked snook at first winter from another study (Brennan et al., 2008, Chapter
3) and used long-term growth data to project contributions of the various rearing habitats to adult
stocks. Sites with highest recapture ratios and low growth made the highest numerical
contributions to adult stocks, but biomass contributions were among the lowest. Hatchery-reared
juveniles stocked at sites that produced intermediate growth rates and recapture ratios, made the
highest biomass contributions to adults stocks. Follow up work is needed to understand
interactive implications between stocking magnitude (Brennan et al., 2008, Chapter 3) and
responses in hatchery and wild snook (across age classes). This work provides important
information for resource managers considering the productivity of various rearing habitats and
has both fishery and habitat management implications. Follow up models examining stock
recruitment and fisheries dynamics under various enhancement scenarios (e.g. Lorenzen, 2006)
can be coupled with results from this study. Together, these can be powerful tools that provide
essential information for responsible and adaptive management of these valuable stocks.
56
57
Table 2-1. Numbers of released snook in 1998 and 1999 according to creek, sub habitats within creeks, and release size category. "Large" refers to snook 180-270 mm FL, “medium” 131-179 mm FL, “small” 180-129 mm FL
Year Season Estuarine Lot SAR Release sub-habitat TotalTributary U M L I
1998 Spring North Creek 1 Large 100 0 100 150 3501 Medium 0 100 0 150 2501 Small 308 310 310 300 1228
Subtotal 408 410 410 600 1828South Creek 98-1 Large 100 100 100 150 450
98-1 Medium 100 0 99 146 34598-1 Small 210 300 210 300 1020
Subtotal 410 400 409 596 1815Phillippi Creek 98-1 Large 300 150 299 0 749
98-1 Medium 300 450 300 400 145098-1 Small 610 610 610 204 2034
Subtotal 1210 1210 1209 604 4233Bowlees Creek 98-1 Large 300 300 300 601 1501Tidy Island 98-1 Large . . 0 500 500
Medium . . 2000 0 2000Lot 98-1 Total 2328 2320 2328 2901 11877
1999 Spring North Creek 99-1 Large 87 88 86 87 34899-1 Medium 87 86 87 85 34599-1 Small 84 85 86 86
Subtotal 258 259 259 258 1034South Creek 99-1 Large 87 87 87 87 348
99-1 Medium 86 105 86 86 36399-1 Small 85 87 87 87
Subtotal 258 279 260 260 1057Bowlees Creek 99-1 Large 87 87 87 85 346
99-1 Medium 87 87 88 86 34899-1 Small 87 87 89 87
Subtotal 261 261 264 258 1044Hudson Bayou 99-1 Large 87 97 87 87 358
99-1 Medium 90 87 87 86 35099-1 Small 87 0 0 0 87
Subtotal 264 184 174 173 795Tidy Island 99-1 Large . . 345 0 345
99-1 Medium . . 85 0 8599-1 Small . . 260 0 260
Subtotals . . 690 0 690Lot 99-1 Total 1041 983 1647 949 4620
North Creek 99-2 Large 85 87 87 90 34999-2 Medium 85 87 87 89 34899-2 Small 87 91 89 94
Subtotal 257 265 263 273 1058South Creek 99-2 Large 78 87 101 86 352
99-2 Medium 78 86 86 86 33699-2 Small 86 102 86 87 361
Subtotal 242 275 273 259 1049Bowlees Creek 99-2 Large 86 0 173 87 346
99-2 Medium 86 87 87 87 34799-2 Small 87 87 87 88
Subtotal 259 174 347 262 1042Hudson Bayou 99-2 Large 86 85 87 97 355
99-2 Medium 86 87 86 87 34699-2 Small 78 0 0 0 78
Subtotal 250 172 173 184 779Tidy Island 99-2 Large . . 408 0 408
99-2 Medium . . 470 0 470Subtotal . . 878 0 878
Lot 99-2 Total 1008 886 1934 978 48061999 Total 2049 1869 3581 1927 9426
Grand Total 4377 4189 5909 4828 21303
341
346
350
361
349
58
Table 2-2. Nomenclature for classification of release sites, release systems, release sub-systems and release habitats
Release Release Release Release HabitatSite System Sub-system Habitat TypeSC_U South U South Creek Tidal CreekSC_M South M South Creek Tidal CreekSC_L South L South Creek Tidal CreekSC_I South I Island Barrier IslandNC_U North U North Creek Tidal CreekNC_M North M North Creek Tidal CreekNC_L North L North Creek Tidal CreekNC_I North I Island Barrier IslandPC_U Phillippi U Phillippi Creek Tidal CreekPC_M Phillippi M Phillippi Creek Tidal CreekPC_L Phillippi L Phillippi Creek Tidal CreekPC_I Phillippi I Island Barrier IslandHB_U Hudson U Hudson Bayou Tidal CreekHB_M Hudson M Hudson Bayou Tidal CreekHB_L Hudson L Hudson Bayou Tidal CreekHB_I Hudson I Island Barrier IslandBC_U Bowlees U Bowlees Creek Tidal CreekBC_M Bowlees M Bowlees Creek Tidal CreekBC_L Bowlees L Bowlees Creek Tidal CreekBC_I Bowlees I Island Barrier IslandTIDY Tidy TIDY Tidy Peninsula Mainland PeninsulaTIDY_I Tidy I Island Barrier Island
59
Table 2-3. Numbers of snook released according to release system and year. 1997-1 and 1997-2 represent two release experiments performed in 1997
Release System Release Years1997-1 1997-2 1998 1999 2000 2001 2002 2004 Totals
Intercoastal Islands 1,667 1,795 2,901 1,927 0 0 0 0 8,Tidy Peninsula 0 0 2,000 1,568 0 0 0 0 3,Bowlees Creek 489 0 900 1,566 0 0 889 3,841 7,685Hudson Bayou 0 0 0 1,217 0 0 0 0 1,Phillippi Creek 1,342 1,768 3,629 0 0 0 0 0 6,North Creek 1,009 891 1,228 1,561 4,183 1,634 436 433 11,375South Creek 0 897 1,219 1,587 0 824 128 369 5,024Totals 4,507 5,351 11,877 9,426 4,183 2,458 1,453 4,643 43,898
290568
217739
60
Table 2-4. Release sites, dates and sampling schedule at the release sites. Dates highlighted in grey represent the original sampling regimen; all other dates represent the modified sampling regimen
Location Release dates and sample months:
Months after 1998 release: Months after 1999 release:
Release 1 2 3 7 9 11 Release 1 2 3 4 7 9South Creek Apr-98 May-98 Jun-98 Jul-98 Nov-98 Jan-99 Mar-99 Apr-99 May-99 Jun-99 Jul-99 Aug-99 Nov-99 Jan-00
North Creek Apr-98 May-98 Jun-98 Jul-98 Nov-98 Jan-99 Mar-99 Apr-99 May-99 Jun-99 Jul-99 Aug-99 Nov-99 Jan-00
Phillippi Creek Apr-98 May-98 Jun-98 Jul-98 Nov-98 Jan-99 Mar-99 Apr-99 May-99 Jun-99 Jul-99 Aug-99 Nov-99 Jan-00
Bowlees Creek Apr-98 May-98 Jun-98 Jul-98 Nov-98 Jan-99 Mar-99
Hudson Bayou Apr-99 May-99 Jun-99 Jul-99 Aug-99 Nov-99 Jan-00
Tidy Island Apr-98 May-98 Jun-98 Jul-98 Nov-98 Jan-99 Mar-99 Apr-99 May-99 Jun-99 Jul-99 Aug-99 Nov-99 Jan-00
61
Table 2-5. Snook captures organized by sampling type and locations. These data were used to verify results from short-term sampling and monitor recruitment to adult habitats
Year Beaches and Passes Intercoastal Red Tide Tournaments Release sites1998 . 112 . 37 5181999 349 603 . 48 1,9042000 393 102 . 191 1,0712001 867 . 1,103 128 2,8382002 431 . . 195 1,8862003 489 13 . 155 2,5872004 354 208 . 309 3,2932005 . . 350 195 888Totals 2,883 1,038 1,453 1,258 14,985Grand Total 21,617
62
Table 2-6. Variables used for the principle component analysis examining characteristics of the release sites and associated observed short-term post-release responses (first summer and winter after release) of hatchery-reared snook age-1 snook stocked in 1998 and 1999
Release siteSCU SCM SCL SCA NCU NCM NCL NCA PCU PCM PCL PCA HBM HBL HBA BCU BCM BCL BCA TIDY TIDYA
Age-1 summer CPUE 0.040 0.077 0.086 0.077 0.299 0.269 0.118 0.028 0.000 0.151 0.148 0.019 0.083 0.000 0.071 0.294 0.154 0.231 0.026 0.063 0.000Age-1 winter CPUE 0.000 0.083 0.000 0.250 0.141 0.256 0.458 0.125 0.000 0.278 0.000 0.000 0.000 0.333 0.381 0.028 0.067 0.042 0.185 0.000Growth (mm/d) 0.34 0.47 0.74 0.44 0.41 0.52 0.71 . . 0.29 0.79 . 0.52 0.69 0.67 0.67 0.63 . 0.62 0.68R_ratio 0.000 0.210 0.743 0.359 1.625 1.713 0.966 0.265 0.000 0.000 0.083 0.166 0.000 0.000 1.120 0.976 0.272 0.110 0.000 0.123 0.400Salinity (ppt) 3.3 26.2 33.7 34.9 6.5 13.7 30.3 32.6 0.3 8.2 30.0 32.1 34.1 33.8 35.8 32.1 33.3 32.9 36.0 33.4 35.13Dissolved oxygen (mg/l) 3.6 4.4 5.1 7.4 6.2 5.8 5.9 9.7 7.1 6.4 5.6 5.2 2.7 3.4 7.1 4.1 6.3 5.8 7.1 5.6 8.54Temperature (celsius) 28.8 31.2 29.6 30.8 30.8 29.2 32.6 31.9 31.8 31.5 32.5 33.4 28.8 28.5 27.7 30.8 30.7 30.1 30.2 28.8 30.84pH 7.9 7.9 8.1 8.3 7.9 7.1 8.0 8.4 7.8 7.9 8.0 8.0 8.1 8.1 8.4 8.2 8.4 8.4 8.3 8.0 8.30Substrate softness (cm) 27 22 15 9 10 28 34 10 26 41 13 23 32 10 7 22 63 19 17 12 1Bank slope (rise/run) 0.25 0.20 0.15 0.30 0.18 0.14 0.09 0.17 0.26 0.21 0.11 0.14 0.34 0.24 0.24 0.31 0.15 0.20 0.38 0.16 0.15Sediment organics (%) 4.78 8.04 6.32 7.35 4.50 8.96 11.28 4.13 7.38 4.96 9.15 23.72 8.25 8.12 5.14 4.59 16.08 4.32 5.22 2.05 1.23Semidmen moisture (%) 43.48 50.23 43.32 48.74 49.05 60.47 66.30 37.73 49.84 43.84 48.02 74.65 49.90 51.30 43.21 39.23 55.26 38.34 46.81 35.69 28.58Shoreline vegetation (%) 100 97 100 100 100 100 100 100 100 100 100 83 50 37 100 27 32 73 48 90 86Shoreline mangroves (%) 13 100 100 100 24 43 97 100 0 37 93 83 42 28 100 8 24 73 48 83 8
8
6 Note: Catch per unit effort (CPUE) of age-1 snook was measured according to 21 long standard seine hauls. Growth and recapture ratios (number recaptured/number released * 100, standardized for effort) were from snook recaptured during first winter after release.
63
Table 2-7. Variables used for the principle component analysis examining both short-term (days-at-large <= 365) and long-term (days-at-large > 365) post-release responses of hatchery-reared snook stocked in 1998 and 1999 according to release site characteristics
Variable SCU SCM SCL SCA NCU NCM NCL NCA PCU PCM PCL PCA HBM HBL HBA BCU BCM BCL BCA tidy TIASalinity (ppt) 3.3 26.2 33.7 34.9 6.5 13.7 30.3 32.6 0.3 8.2 30.0 32.1 34.1 33.8 35.8 32.1 33.3 32.9 36.0 33.4 35.1Dissolved oxygen (mg/l) 3.6 4.4 5.1 7.4 6.2 5.8 5.9 9.7 7.1 6.4 5.6 5.2 2.7 3.4 7.1 4.1 6.3 5.8 7.1 5.6 8.5Temperature (celsius) 28.8 31.2 29.6 30.8 30.8 29.2 32.6 31.9 31.8 31.5 32.5 33.4 28.8 28.5 27.7 30.8 30.7 30.1 30.2 28.8 30.8pH 7.9 7.9 8.1 8.3 7.9 7.1 8.0 8.4 7.8 7.9 8.0 8.0 8.1 8.1 8.4 8.2 8.4 8.4 8.3 8.0 8.3Substrate softness (cm) 26.7 22.3 15.3 8.7 9.6 27.7 34.3 10 26 41 13 23 32 10 7 22 63 19 17 12 18Bank slope (rise/run) 0.25 0.20 0.15 0.30 0.18 0.14 0.09 0.17 0.26 0.21 0.11 0.14 0.34 0.24 0.24 0.31 0.15 0.20 0.38 0.16 0.15Sediment organics (%) 4.78 8.04 6.32 7.35 4.5 8.96 11.28 4.13 7.38 4.96 9.15 23.72 8.25 8.12 5.14 4.59 16.08 4.32 5.22 2.05 1.23Semidmen moisture (%) 43.48 50.23 43.32 48.74 49.05 60.47 66.3 37.73 49.84 43.84 48.02 74.65 49.9 51.3 43.21 39.23 55.26 38.34 46.81 35.69 28.58Shoreline vegetation (%) 100 97 100 100 100 100 100 100 100 100 100 83 50 37 100 27 32 73 48 90 86Shoreline mangroves (%) 13 100 100 100 24 43 97 100 0 37 93 83 42 28 100 8 24 73 48 83 86Age-1 summer CPUE 0.040 0.077 0.086 0.077 0.299 0.269 0.118 0.028 0.000 0.151 0.148 0.019 0.083 0.000 0.071 0.294 0.154 0.231 0.026 0.063 0.000Age-1 winter CPUE . 0.308 0.361 0.241 . 0.454 0.465 0.126 . 0.417 0.167 0.083 0.181 0.472 . 0.296 0.149 0.339 0.111 0.105 0.1671st winter growth rate (mm/d) . 0.34 0.47 0.74 0.44 0.41 0.52 0.71 . . 0.29 0.79 . 0.52 0.69 0.67 0.67 0.63 . 0.62 0.681st winter R ratio 0 0.210 0.743 0.359 1.625 1.713 0.966 0.265 0.000 0.000 0.083 0.166 0.000 0.000 1.120 0.976 0.272 0.110 0.000 0.123 0.400R ratio, long-term dispersers 0 0.105 0.425 0.179 0.217 0.214 0.429 0.000 0.000 0.083 0.165 0.331 0.000 0.288 0.560 0.488 0.408 0.439 0.178 0.084 0.000R ratio, winter habitats (long-term) 0 0.943 0.212 0.000 0.758 0.535 1.502 0.177 0.000 0.000 0.083 0.166 0.000 0.288 0.280 0.366 0.136 0.329 0.000 0.252 0.000R ratio, summer habitats (long-term 0 0.000 0.106 0.090 0.000 0.214 0.322 0.000 0.000 0.000 0.083 0.000 0.000 0.000 0.560 0.488 0.272 0.329 0.178 0.056 0.000Linf (FL mm) 657 657 657 824 595 595 595 824 . . . . . . 824 704 704 704 824 . .3 yr weight (g) . . 665 . 1020 679 856 885 . . 672 . . . 988 1231 951 1290 . 685 . Note: Catch per unit effort (CPUE) of age-1 snook was measured according to 21 long standard seine hauls. Long-term growth was expressed as Linf (von Bertalanffy parameter), and weight of age-3 snook recaptures (days-at-large = 366-720) and recapture ratios (R ratio, number recaptured/number released*100, standardized for effort). Recapture ratios over the long-term were organized according to those recaptured in winter habitats, those recaptured at least 1 km from their release sites (dispersers), and in adult snook summer habitats (beaches, passes, and seagrass habitats).
64
Table 2-8. von Bertalanffy growth parameters for snook recaptures (n) sourced from different release habitats. Length and age data only included recaptures of hatchery snook at-large for 180d or longer.
Linf k t0 nSouth Creek 657 0.250 -0.486 46North Creek 552 0.550 0.470 249Bowlees Creek 704 0.408 -0.502 60Island Habitat 824 0.164 -1.650 40
65
Table 2-9. Numbers of snook recaptured at various distances (km) away from their respective release sites. Data are only for snook recaptured within the first year after releases. Release Site Release Habitat Release sub-system Number recaptured by distance from release site (km)
0 0.1 to 5 5.1 to 10 10.0 to 20 > 20BCU Bowlees Creek Upstream 14 20 1 2 1BCM Bowlees Creek Midstream 4 10 1 3 0BCL Bowlees Creek Stream Mouth 12 4 7 1 1PCU Phillippi Creek Upstream 0 0 0 0 0PCM Phillippi Creek Midstream 0 1 2 0 0PCL Phillippi Creek Stream Mouth 16 2 2 4 0NCU North Creek Upstream 14 36 0 1 0NCM North Creek Midstream 51 19 1 1 0NCL North Creek Stream Mouth 51 27 5 6 0SCU South Creek Upstream 25 1 0 0 0SCM South Creek Midstream 31 4 1 0 0SCL South Creek Stream Mouth 7 9 1 0 0HBU Hudson Bayou Upstream 0 6 0 0 0HBM Hudson Bayou Midstream 4 7 0 0 0HBL Hudson Bayou Stream Mouth 5 5 1 0 0TIDY Tidy Peninsula Mainland Island 35 5 2 2 1TIDYI Tidy Island 1 1 0 0 0BCI Buttonwood Harbor Island 1 1 5 0 1HBI Lido Lagoon Island 5 2 0 0 0PCA Coconut Bayou Island 3 0 3 0 0NCA Bird Keys Island 10 0 1 0 0SCA South Creek Island 6 1 0 1 0Total 295 161 33 21 4Cumulative percent 57% 89% 95% 99% 100%
66
67
Figure 2-1. Map of experimental release sites located off the southwest coast of Florida USA. Release creeks are denoted and release sub-habitats (U= upstream, M=midstream, L=lower stream, and I=barrier island habitat) are enclosed in transparent shaded ovals.
South Creek
INTERCO
ASTAL
WATERW
AYCasey
Key METERS
0 500 1000
27o10’6.75” N
U
LM
I MAINLAND
COAST2500 m
Big
Pass
Gulf ofMexico
LidoKey
Siesta Key
Phill
ippi
Cre
ek
HudsonBayou
I L
U
M
I
UM
L
RobertsBay
Intercoastal
Waterw
ay
1200 M1200 MLongboat
Key
Tidy IslandLongboat Pass
Gulf ofMexico
Anna
Maria Island
Intercoastal Waterway
SarasotaBay
I
2200 MLongboatKey
SarasotaBay
Gulf ofMexico
BowleesCreek
I
LM U
800 M
Barrier Island
North Creek
BirdKeys
Gulf ofMexico
I
L
M
U
Intercoastal Waterw
ay
27o 12’ 58.55” N
82o 29’ 56.21” W
27o 16’ 17.74” N
82o 34’ 2.48” W
82o 39’ 8.98” W
27o 27’ 31.23” N
27o 24’ 47.81” N
82o 34’ 43.95” W
82o 29’ 9.05” W
68
Figure 2-2. Release numbers (grey bars) and post-release results showing percent hatchery
snook found in total snook catch (graph A) and recapture ratios (number recaptured/number released) adjusted for effort (from 21 m seine, graph B). The chronological representation on the horizontal axis is not to scale.
0
2000
4000
6000
8000
10000
12000
r-97
t-97
v-97
c-97
an-9
8eb
-98
r-98
r-98
y-98
un-9
8Ju
l-98
v-98
n-99
r-99
r-99
y-99
un-9
9Ju
l-99
ug-9
9v-
99an
-00
Rel
ease
Num
ber
Ap
Oc
No
De J F Ma
Ap
Ma J No Ja Ma
Ap
Ma J A No J
0
5
10
15
20
25
% h
atch
ery
snoo
k0
2000
4000
6000
8000
10000
12000
Apr
-97
Oct
-97
Nov
-97
Dec
-97
Jan-
98Fe
b-98
Mar
-98
Apr
-98
May
-98
Jun-
98Ju
l-98
Nov
-98
Jan-
99M
ar-9
9A
pr-9
9M
ay-9
9Ju
n-99
Jul-9
9A
ug-9
9N
ov-9
9Ja
n-00
Rel
ease
Num
b e
0
0.025
0.05
0.075
0.1
Rec
aptu
re ra
te
Rel
ease
num
ber
Rel
ease
num
ber
Apr
Jan
Dec
Nov
Oct
Mar
Apr
May
Jun
Jul
Nov
Jan
Mar
Feb
May
Jun
Jul
Apr
Aug
Nov
Jan
1997 1998 1999
Apr
Jan
Dec
Nov
Oct
Mar
Apr
May
Jun
Jul
Nov
Jan
Mar
Feb
May
Jun
Jul
Apr
Aug
Nov
Jan
1997 1998 1999
B
% H
atch
ery
snoo
kR
ecap
ture
rate
A
0
0.5
1
1.5
2
2.5
3
3.5
4
Upstream Barrier IslandsMidstream Stream Mouth
(n=90) (n=84)
(n=52)
(n=15)
Rec
aptu
re ra
te
Figure 2-3. Mean recapture ratios (number recaptured/number released x100) of snook from
1998 and 1999 releases, by sub-subsystems. Means are calculated from replicate release years, stream systems and release lots. Data only include recaptures within one year after release. Numbers in brackets are the numbers of individual snook recaptures for each particular release microhabitat. Error bars are 95% confidence limits of the means.
69
0
0.5
1
1.5
2
2.5
3
3.5
4
Islands Tidy Bowlees Hudson Phillippi North SouthIntercoastalIslands
TidyIsland
BowleesCreek
HudsonBayou
PhillippiCreek
NorthCreek
SouthCreek
Mean recapture rate
Rec
aptu
re ra
te
Figure 2-4. Recapture ratios (number recaptured/number released x 100) of snook by release
habitat. Data are from recaptures collected within the first year after release. Mean recapture ratio is indicated by the horizontal line.
70
Figure 2-5. Linear regression models using mean catch-per-unit effort during winter of age-1
snook to predict short- (< 1 year post-release) and long-term (>=1 year post-release) recapture ratios (rrate; number recaptured/number released x 100).
r2 = 0.7228
r2 = 0.9298
0
0.5
1
1.5
2
2.5
3
3.5
4
0 0.5 1 1.5 2 2.5
MEAN AGE-1 CPUE IN WINTER(in release habitats)
3
SHORT-TERM RRATIOS (< 1 YEAR)
LONG-TERM RRATIOS (1-8 YEARS)
Rec
aptu
re ra
tio
Mean Age-1 CPUE in winter(at release habitats)
71
Figure 2-6. Mean growth rates from fish released in different habitats and sub-habitats. Data are from recaptures collected 6-10 months after releases in 1998, and 1999. Sites from left to right are arranged in a North to South order. Error bars are 95% confidence intervals. Numbers above bars are number of recaptures.
0.35
0.45
0.55
0.65
0.75
0.85
TID Y I TI B C HB LID O KEY PC PC A N C N C A SC SC ATIDYI TIDY BC HB NCAHBA PC PCA NC SCASC
Gro
wth
rate
(mm
/d)
2 14
12
1
4
3
1
44
42
9
72
150
175
200
225
250
275
300
325
350
375
400
0 1 2 3 4 5
Size
at r
ecap
ture
(For
k le
ngth
, mm
)
6ISLANDS TIDY BC NC SC
Figure 2-7. Size-at-recapture for snook released in different habitats. Data are from recaptures collected 6-10 months after releases in April 1998, and 1999. Sites from left to right are arranged in a North to South order.
73
Figure 2-8. Mean size-at-release and growth rates of hatchery-reared snook recaptured during
the first winter after release (6-10 mo.).
0
0.2
0.4
0.6
0.8
1
1.2
0 50 100 150 200 250
Mean size-at-release (FL, mm)
Gro
wth
rate
(mm
/d)
Mean size-at-release (FL, mm)
Gro
wth
rate
(mm
/d)
74
75
Figure 2-9. Length frequency distributions of wild and hatchery snook at microhabitat release sites for each season. Hatchery
snook bars (red) are stacked above wild snook bars (blue).
UPSTREAM
0
20
40
60
50 150 250 350 450 550 650 750 850
MIDSTREAM
0
20
40
60
50 150 250 350 450 550 650 750 850
STREAM MOUTH
0
20
40
60
50 150 250 350 450 550 650 750 850
INTERCOASTAL ISLANDS
0
20
40
60
50 150 250 350 450 550 650 750 850
SPRINGUPSTREAM
0
20
40
60
50 150 250 350 450 550 650 750 850
MIDSTREAM
0
20
40
60
50 150 250 350 450 550 650 750 850
STREAM MOUTH
0
20
40
60
50 150 250 350 450 550 650 750 850
INTERCOASTAL ISLANDS
0
20
40
60
50 150 250 350 450 550 650 750 850
SUMMER
UPSTREAM
0
20
40
60
50 150 250 350 450 550 650 750 850
MIDSTREAM
0
20
40
60
50 150 250 350 450 550 650 750 850
STREAM MOUTH
0
20
40
60
50 150 250 350 450 550 650 750 850
INTERCOASTAL ISLANDS
0
20
40
60
50 150 250 350 450 550 650 750 850
FALLUPSTREAM
0
20
40
60
50 150 250 350 450 550 650 750 850
MIDSTREAM
0
20
40
60
50 150 250 350 450 550 650 750 850
STREAM MOUTH
0
20
40
60
50 150 250 350 450 550 650 750 850
INTERCOASTAL ISLANDS
0
20
40
60
50 150 250 350 450 550 650 750 850
WINTER
Fork Length (mm)
Num
ber o
f Sno
ok
76
Figure 2-10. Plot of first two principle components for short-term data related to characteristics of the release sites. Release sites are represented by shaded dots, release site characteristics are represented as vectors.
ot of first two principle components for short-term data related to characteristics of the release sites. Release sites are represented by shaded dots, release site characteristics are represented as vectors.
NCA
TIDYA
PCA
NCM
SCABCM
BCL
PCL
TIDY
NCU
NCL
0
0.5
0.5)
Pri
om2
()
Sediment organics
Tem
pera
ture
CPUE winter
Mangrove
cover
Dissolved oxygen (mg/l)
Bank slope
Growth
Sediment softness
CPUE summer
Vegetative cover
Sediment moisture
Salinity
Recapture ratio
inc
ple
cpo
nent
21
%
BCU
-0.5-0.5 0
Principle component 1 (27%
y = -1.9596Ln(x) - 0.2578R2 = 0.5923
00.20.40.60.8
11.21.41.61.8
0.4 0.45 0.5 0.55 0.6 0.65 0.7 0.75 0.8
Growth rate (mm/d)
Rec
aptu
re ra
tio
Figure 2-11. Plot of growth rate versus recapture ratio (number recaptured/number
released*100) from standardized seine collections during the first winter after release. The linear model (inset) shows the best fit of loge transformed growth rates to recapture ratios with an associated correlation coefficient.
77
0
0.5
1
1.5
2
2.5
3
SC NC HB BC TI ISLANDS
CP
UE
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
Mea
n gr
owth
rate
(mm
/d)
CPUE
GROWTH
SouthCreek
NorthCreek
HudsonBayou
BowleesCreek
TidyPeninsula
BarrierIslands
CPU
E
Mea
n gr
owth
rate
(mm
/d)
91
12 1415
42
Figure 2-12. Catch per unit effort (CPUE) of wild and hatchery snook less than 300 mm fork
length according to release habitat and associated mean growth rates (mm/d) through the first winter after release. Error bars are 95% confidence intervals. Numbers above bars are the numbers of snook recaptured.
78
BCM
NCM
SCM
SAR
WINDISP
GROWTH
WINFL
RRATIO
SUMDISP
SUMFL
BCUHBL
NCLNCU
TIDY
-1.5
-1
-0.5
0
0.5
1
1.5
-1 -0.5 0 0.5 1
Variable 1
Var
iabl
e 2
Figure 2-13. Plot of the canonical variables (triangles) for size-at-release (SAR mm FL) and associated response variables including first winter recapture ratio (RRATIO), growth to first winter (GROWTH, mm/d), dispersal rates during summer (SUMDISP) and winter (WINDISP), and size-at-recapture (mm FL) during summer (SUMFL) and winter (WINFL). Release sites are plotted as shaded dots.
79
0 0.05 0.1 0.15 0.2 0.25 0.3 0.35
80-129
130-179
180-229
Distance Moved (km)
0 0.05 0.1 0.15 0.2 0.25 0.3
87-129
130-179
180-225
Distance Moved (km)
Rec
apFL
(mm
)SA
R (F
L m
m)
Distance moved (km)
Distance moved (km)
A
B
Figure 2-14. Mean dispersal distances (km) organized by size-at-recapture (RecapFL, graph A), and size-at-release (SAR, mm FL, graph B). Error bars are standard error.
80
0
100
200
300
400
500
600
700
800
1 2 3 4 5 6 7 8
Age (years)
Leng
th (F
L)
BC model NC Model BC obs NC obs
NC length=594(1-e-0.553(age-0.478))
BC length=704(1-e-0.409(age-0.059))
NC length=594(1-e-0.553(age-0.478))
BC length=704(1-e-0.409(age-0.059))
Age (years)
Fork
leng
th (m
m)
Figure 2-15. Plot of length and age data from hatchery snook recaptures. Data include
recaptures from NC (shaded dots) and BC (black triangles). Von Bertalanffy length-at-age predictions are shown for both creeks, with growth equations shown in box.
81
0
2
4
6
8
10
12
0.25 0.5 0.75 1 1.25 1.5 1.75 2 2.25 2.5 2.75 3 3.25 3.5 3.75 4 4.25 4.5 4.75 5 5.25 5.5 5.75 6S S F W S S F W S S F W S S F W S S F W S S F WS S F W S S F W S S F W S S F W S S F W S S F W
AGE-1 AGE-2 AGE-3 AGE-5AGE-4
MEA
N D
ISPE
RSA
L D
ISTA
NC
E (K
M)
AGE-6
125 32 5271
37
17 14
29
10
25 8
24
6
14
9
5
6
8
3 3
1
1
3
2
Figure 2-16. Mean dispersal distances from rearing habitats (minimum distance from release
site to recapture site) of snook recaptured within seasons (spring, summer, fall winter) and by snook age (years). Data are from snook released in the spring of 1998, and 1999. Numbers of individuals recaptured within each season are shown above bars. Error bars are standard error.
82
NCL
BCU
BCL
NCA
Age-1 w inter CPUE
1st w inter recapture ratio
sediment moisture
L_inf
short-term grow th
pHsalinity
Age-3 w eightbank slope
dissolved oxygen
temperature
recapture ratio atbeaches and passes
recapture ratioof long-term dispersers
substrate softness
Age-1 summer density
sediment organics
shoreline vegetation
long-term w inter recapture ratio
NCM
mangrove cover
SCL
BCM
-0.50
0.00
0.50
-0.50 0.00 0.50
Principle component 1 (36%)
Prin
cipl
e co
mpo
nent
2 (2
8%)
Figure 2-17. Plot of the first two principle components comparing long-term responses (recaptures at-large for over 1 year, black dots) of the released fish, short-term responses (from recaptures at-large less than 1 yr., also black dots), and characteristics of the release sites (shaded dots). The percentages in brackets on the axes represent the percent of the overall variation explained by each respective principle component.
83
0
100
200
300
400
500
0.35 0.4 0.45 0.5 0.55 0.6 0.65 0.7 0.75growth rates in rearing habitats
Num
ber o
f sno
ok
age-2age-3age-4age-5age-6
0
25
50
75
100
125
150
175
200
0.35 0.4 0.45 0.5 0.55 0.6 0.65 0.7 0.75
growth rates in rearing habitats
Bio
mas
s (k
g
age-2age-3age-4age-5age-6
Growth rate (mm/d) in rearing habitat
Num
ber o
f sno
okB
iom
ass (
kg)
Growth rate (mm/d) in rearing habitat
A
B
NC
/SC
BC
TID
Y
ISLA
ND
S
HB
NC/SC(N age-2 = 316)
ISLANDS(N age-2 = 171)
Figure 2-18. Numerical (graph A) and biomass (graph B) projections to adult stages of juvenile hatchery-reared snook released in various rearing habitats with characteristic growth rates by first winter. Projections assume equal stocking levels (1000 age-1 snook per rearing habitat). Arrows indicate projections according to rearing habitats observed in Sarasota and Manatee Counties (NC/SC = North Creek or South Creek, HB=Hudson Bayou, TIDY=Tidy Peninsula, BC= Bowlees Creek, ISLANDS=barrier island habitat). Shaded dots in B highlight the differences in biomass production of age-4 snook between NC/SC and ISLANDS.
84
CHAPTER 3 MANIPULATIONS OF STOCKING MAGNITUDE: ADDRESSING DENSITY-
DEPENDENCE IN A JUVENILE COHORT OF COMMON SNOOK Centropomus undecimalis
Introduction
Overfishing and habitat alteration are considered to be the primary anthropogenic
disturbances to coastal ecosystems (Hallegraeff, 1993; Jackson, 2001; Vitousek et al., 1997).
Worldwide demands on ocean fishery resources since the early 1990’s have exceeded 100
million tons of harvest and fishery experts predict global marine catch has approached its upper
limit (Botsford et al., 1997). Furthermore, historical data indicate that long-term overfishing has
been a primary contributor to major structural and functional changes in coastal ecosystems
(Jackson et al., 2001). The primary goal of sustainable fisheries has evidently not been widely
achieved, considering the number of overfished populations and indirect effects of fisheries on
ecosystems (e.g. bycatch) (Steele et al., 1992; Botsford et al., 1997; Pauly et al. 1998; Hamilton
and Haedrich, 1999). Anthropogenic habitat change and pollution, although difficult to quantify,
also pose substantial threats to fishery stocks worldwide. For many marine species, juvenile
nursery habitats are associated with coastlines where development and pollution are
concentrated, resulting in inherently reduced ability of fisheries stocks to recover (Bruger and
Haddad, 1986; Islam and Haque, 2004; Mumby et al., 2004).
In Florida, common snook Centropomus undecimalis (“snook”) are a coastal, warm-
water fish whose populations are a major concern for fishery managers because of their
ecological and economic value. Snook are valued as one of the top marine sport fishes in Florida
and contribute to an annual USD 5.4 billon saltwater recreational fishing industry in Florida
alone (American Sport Fishing Association, 2004). Despite increasingly restrictive fishery
regulations on common snook, these populations are considered overfished and below
management goals (Muller and Taylor, 2005). Annual fishing mortality rates have steadily
85
increased over the last 20 years, and recruitment has generally declined (Muller and Taylor,
2005). High fishing pressure, coupled with relatively few spawner-sized females in the adult
population, has made management of snook stocks difficult (Muller and Taylor, 2005).
Furthermore, snook associate with shoreline habitat (Marshall, 1958; Gilmore et al., 1983,
McMichael et al., 1989; Peters et al., 1998a) and thus depend on coastal waters that may be
subjected to intense anthropogenic influences. Collectively, overfishing and habitat loss have
caused a general decline in the population although the relative influence of these is unclear
(Muller and Taylor, 2005).
In Florida, snook spawn primarily during summer (June – August) in high salinity (>28
ppt) seawater in the inlets and tidal passes of estuaries, mouths of rivers and canals, and along
sandy beaches (Marshal, 1958; Volpe, 1959; Taylor et al., 1998). Post-larval snook recruit
mostly to vegetated shallow brackish tidal creeks, canals and lagoons in both low- (riverine) and
high-salinity (mangrove swamp and salt marsh) habitats (Peters et al., 1998a), with moderately
sloping shorelines and basin depths of 1 m or less, with mud or sand substrate. Age-0 cohorts
are thought to generally remain in their rearing habitats through the winter and early spring
(Gilmore et al., 1983; McMichael et al., 1989). In Sarasota and Manatee counties, these habitats
are found in tidal creeks and estuarine backwaters, and can also serve as thermal refuge for
snook of all sizes during winter (N. Brennan, unpublished data). Gilmore et al. (1983)
hypothesized that after winter, adolescent snook (150 – 400 mm standard length [SL], mean =
240 mm SL, from east-central Florida populations) undergo an ontogenetic habitat shift to
seagrass beds. Adult snook are also known to disperse to seagrass beds, estuarine inlets, and
beaches by late spring and early summer (Marshall, 1958; Gilmore et al., 1983; Taylor et al.,
1998).
86
Faced with rapid human population growth in Florida and the limitation of traditional
management tools, managers have investigated the potential to augment overfished populations
with supply-side approaches such as stocking hatchery-reared juvenile snook. However,
evidence for stock enhancement programs actually accomplishing stock management goals has
been sparse, and such programs can be ineffective and even deleterious to wild stocks
(Nickelson, 2003; Walters and Martell, 2004; Kosteau and Zhou, 2006). Therefore, stock
enhancement programs might best be operated on an experimental basis with rigorous scientific
evaluation prior to full-scale acceptance as a management tool (Blankenship and Leber, 1995;
Leber, 1999; Hilborn, 2004; Walters and Martell, 2004). Experiments with stock enhancement
can provide valuable insight into population dynamics, behavior, growth, and survival responses
to habitat quality, for example (Miller and Walters, 2004).
The efficacy of stock enhancement should depend on the magnitude of density-dependent
processes. With snook, a cannibalistic and piscivorous species, the degree to which habitat
availability and quality, or recruitment limitation, influence stock sizes remains unclear. Many
species suffer elevated and density-dependent mortality during specific ontogenetic stages,
typically during early life stages (Houde, 1987; Byström et al., 2004; Doherty et al., 2004). In
fish, highest mortality rates occur during the larval stages (Houde, 1987; Lorenzen, 1996). In
some cases however, density-dependent mortality is high in later stages and overall recruitment
rates are restricted by these “survival bottlenecks”. For example, Byström et al. (2004) showed
that while age-0 char were not affected by density-dependent processes, age-1 char underwent
highly density-dependent ontogenetic feeding shifts and became exposed to high rates of
predation. With snook stocks, relative contributions of various life stages to overall survival
rates remain unclear. Loss of habitat availability can inherently reduce a stock’s production
87
capacity, yet high density-dependent mortality can still occur in remaining habitats from
competition and predation. An examination of population responses to manipulative stocking
experiments over a variety of habitats would aid in determining the timing and extent of density-
dependent effects. Furthermore, if snook stocks are overfished, then experimental and
manipulative stocking may be necessary to elicit density-dependent responses in augmented
populations.
In this study, I released juvenile (age-1) hatchery-reared snook to manipulate localized
age-1 recruitment and identify potential resultant density-dependent mortality effects. Mortality
responses in the age-0 cohort from age-1 cannibalism or competition were not addressed in this
study because primary recruitment pulses of age-0 snook occurred from June - September, well
after the experimental releases occurred (in May), and related abundance responses would be
confounded with natural variations in creek-specific recruitment. Specifically, I aimed to
determine if snook nursery habitats are naturally filled to capacity, i.e., if strongly density-
dependent survival would preclude increasing the abundance of age-1 snook leading to increased
abundances of older snook. To accomplish this I (1) estimated wild age-1 snook abundance in
four estuarine tributaries, (2) released hatchery-reared age-1 snook to increase the total localized
abundance of age-1 snook by either 100% (high augmentation; n=2 creeks) or 10% (low
augmentation; n=2 creeks); (3) determined if hatchery-released snook were subjected to higher
mortality rates than wild age-1 snook, and (4) determined if overall loss of age-1 snook in creeks
with high augmentation was higher than those in creeks with low augmentation.
Methods
Study Area
Experimental releases were made in four estuarine creeks (Bowlees Creek [BC],
Whitaker Bayou [WB], North Creek [NC], and South Creek [SC]) located in Sarasota and
88
Manatee Counties, on the west coast of Florida, USA (Figure 3-1). These creeks are tidally
influenced and water levels typically fluctuate by about 0.75 m daily. Salinities vary with tidal
influence and seasonal rains (wet season during summer). Water temperatures during summer
(June – September) typically range from ca 28 oC – 34 oC, and winter (December – March)
temperatures range from ca 12oC – 22oC. Dissolved oxygen levels during summer are low
compared to winter (typically <5 mg/l versus >7 mg/l). All creeks are partially influenced by
anthropogenic shoreline alterations, and sections of BC, WB, and SC have been dredged to some
extent for maritime navigation within the last 20 years (Table 3-1). North Creek remains
relatively unaltered, although upstream reaches are surrounded by a golf course development and
shoreline residential properties. Densities of wild age-1 snook (~150-300 mm fork length [FL])
during early summer were about 0.5 – 3.0 fish/30 m of shoreline (Table 3-1).
Experimental Design and Sampling Methods
I defined “age-1 snook” or “juvenile snook” as offspring from the 2001 spawning year.
By April 2002, these snook were approximately 10 (+/-2) months old. Experimental releases of
hatchery-reared juveniles occurred in May, 2002 and our study continued until June 2003 when
the juveniles were about 24 months old. I imposed two levels of manipulation: 1) a large
addition of hatchery-reared juveniles equal to 100% of the wild juvenile abundance (in creeks
BC and WB), and 2) a smaller addition of hatchery-reared fish equal to 10% of the wild stock (in
NC and SC). Quantitative sampling was performed in all creeks: once before releases (to
estimate pre-release age-1 abundance), and six times after releases through the following spring
(approximately in June, August, October, December 2002, and February, June 2003, Table 3-2)
to monitor the effects of the releases on in-creek abundance of juvenile snook.
To estimate in-creek juvenile snook population size, I used a “leap-frog” systematic
sampling design where a standardized bag seine (73 m long x 3 m deep with 1 cm nylon
89
multifilament mesh) was hauled at roughly every third 30 m section of shoreline throughout each
creek (Figure 3-1, Table 3-1). Each haul approximately sampled a 30 m x 21 m rectangle (630
m2). For creek sections less than 21 m wide, we sampled 30 m of creek length but incorporated
both shorelines into the sample. The seine, loaded on a 4 m kayak, was deployed by hauling
about 21 m of the leading end of the seine towards the shoreline. Once there, a second person
held the outside “corner” of the block and a third person pulled the kayak parallel to the bank (or
against the opposite shoreline) for 30 m then turned and deployed the remaining 21 m of net
towards the shoreline. Once against the bank, both persons moved toward each other along the
shoreline and retrieved the remainder of the seine. To avoid overestimating snook abundance, I
divided snook catch by two for samples collected from two-bank sites because of the increased
shoreline habitat. Seines were not hauled in narrow areas (i.e. less than 3 m across) such as
mosquito ditches, but these shorelines were assumed to be valid snook habitat and were used to
calculate total snook abundance.
Gear Efficiency and Population Estimation
Prior to abundance estimates, we first conducted depletion-removal trials in January and
February, 2002 to measure gear efficiency. I used a modified depletion-removal population
estimate (Hillborn and Walters, 1992) within a defined area to predict single-pass seine
efficiency (Figure 3-2). Three types of depletion-removal methods were used depending on
habitat type. In creek areas narrower than 21 m across (i.e., two stream banks, “two-bank
depletion”), two nets (21-24 m long, 0.9 m deep, 1 cm multifilament mesh), about 42 m apart,
blocked off upstream and downstream borders of the sampling area (Figure 3-2a). We used the
standard seine described above, to sample ~ 30 m within this zone, allowing ~ 6 m from the
seine perimeters to the block nets measured along the shoreline to allow for escapement (as
might occur in a single pass seine). After this, we seined the entire area with the above seine two
90
more times, for a total of three seine hauls (most snook were captured within the first seine, so
three seine passes sufficiently described the depletion rate, see below). We performed 10
separate “two bank” depletion experiments. We also performed “single-bank” depletion along
open shorelines (Figure 3-2b), and used the same method if additional shorelines were farther
than 21 m from the sampled shoreline. A 45 m long block net (1 cm multifilament mesh) was
deployed parallel to the shoreline, ~ 21 m away from shore, and two 21 m seines (same sized
mesh, attached at each end of the deployed 45 m long seine) were deployed toward the shore to
enclose a rectangular area of ~ 945 m2. Once block nets were in place, a standardized seine was
hauled along the inside perimeter of the blocked area. The standard net was stretched 30 m
along the inside perimeter of the block net, then each end was pulled toward shore. I allowed ~ 4
m of space between the ends of the block to allow for escapement (Figure 3-2b). Again, second
and third seine hauls sampled the entirety of the blocked area. For each seine haul, captured
snook were counted and measured (FL, mm), then placed in holding containers with aerated
creek water until the depletion trial was complete. We performed eight separate open bank
depletion trials.
We also performed two-bank extended depletion experiments where we blocked off ~
116 m of creek habitat (Figure 3-2c). With a standard seine, we performed three seine hauls
sequentially upstream to the upstream block net. This process continued for a total of nine seine
hauls in the blocked section. We performed this method in two separate areas.
Using data from the three-pass depletion removal trials, I calculated the maximum
likelihood of the number of snook remaining (δ) in the net area α after the last depletion pass i
(which captured βiα juvenile snook) assuming binomially distributed catch for each seine pass,
91
with equal capture probability for each pass. I then estimated the total abundance of snook in the
net area (γα):
γα = ∑ βiα + δ α (3-1)
From this I calculated the number present at the time of each pass (δ 1s, δ 2s, δ 3s …) and
generated an estimate of the capture probability per seine haul (pmax) as the conditional maximum
likelihood estimate for binomial sampling:
p = ∑ βiα / (δ 0s) (3-2)
I also calculated a pcommon value for all depletion trials j:
pcommon = ∑ βiαj / ∑ δ iαj (3-3)
where i and j index seine hauls and depletion trials, respectively. Estimates of age-1 snook
(estimated by length frequency analysis, see below) abundance in a seined area α at time t (δ αt)
were calculated as:
δ αt = βαt / pcommon (3-4)
Where βαt is the catch of juvenile snook in the seine area α at time t. Adjusted mean catch from
all areas s in creek C at time t (∑δ αtC) were then extrapolated to total creek shoreline habitat, Hc
(obtained from aerial photographs [1cm=24 m], expressed in 30 m units) to obtain an in-creek
population ( Ct) of juvenile snook: N̂
N̂ Ct =∑δ αtC * Hc (3-5)
with variance:
V( Ct) = AC2 * V( ∑δ αtC) (3-6) N̂
Therefore, the variance estimate uses the maximum likelihood estimate of catch efficiency
(without associated variability) but incorporates variation around the mean catch in each creek at
time t, and assumes creek area HC is constant.
92
I used repeated measures analysis to compare pre- and post-release abundance for
treatment and control creeks. If the April 2002 abundance estimate of wild snook juveniles in a
particular creek (pre-release) was less than its early summer (post-release June or July 2002
samples) abundance estimate, I used early summer abundance of age-1 wild snook as our pre-
release estimate (I assumed that no new recruitment of wild age-1 snook occurred; age-1 snook
typically begin to decline in abundance in creeks by mid-summer [Brennan and Leber,
unpublished data; Gilmore et al., 1983]). Since large-scale emigration of age-1 snook occurred
from mid-summer through fall, I only used early-summer and winter abundance estimates for
post-release repeated measures analysis. I also used a linear regression to model augmentation
level (expressed as a percentage of standing wild stock) with change in abundance after stocking
(also expressed as a percentage) to test if a significant relationship existed between augmentation
level and change in abundance.
Density (number per 30 m of shoreline) of juvenile snook in the creeks (Dc) at time t was
calculated by dividing total in-creek juvenile snook abundance Ct by total creek shoreline
distance (HC) as follows:
N̂
Dct = Ct / HC (3-7) N̂
I used data from pre- and post-release sampling, to compare observed and expected densities
within the creeks. To generate expected density, I assumed that numbers of stocked fish were
completely additive to wild juvenile numbers, then used a Chi-Square test to compare differences
in observed and expected age-1 densities in the creeks after releases occurred. I also used an
analysis of covariance procedure to compare in-creek changes in density (DCt) of juvenile snook
before and after releases (using pre-release density, peak summer density, and peak winter
density).
93
Pre-Release April 2002 Sampling
To estimate pre-release juvenile snook abundance, we performed creek-wide standard
seining efforts in each creek prior to stocking hatchery-reared snook (Figure 3-1). Captured
snook were also marked with externally visible implant elastomer (VIE) tags implanted in the
caudal fins (see Brennan et al., 2005) with VIE colors specific to creeks to (1) aid in determining
site fidelity and inter-creek migration rates, and (2) identify snook recaptured multiple times.
Tagging and Release
Hatchery-reared snook, F1 progeny from wild parental stocks, were hatched in the
summer of 2001 and reared until spring 2002 when tagging and release activities occurred in
May, 2002. Each released hatchery-reared snook was tagged with (1) a coded-wire tag (CWT)
to identify it’s associated release information (i.e., release creek, release date, size-at-release),
and (2) two red VIE tags implanted in the caudal fin to externally identify it as a hatchery-reared
snook (as in Brennan et al., 2005). Tagged juveniles were then returned to their tanks and held
for one week to recover from the tagging process. Salinities in the rearing tanks ranged from 3 -
6 ppt and water temperature from 26-30 °C.
On the day of release (six days after tagging), fish from each release group were
harvested and checked for tag presence. Snook were transported in tanks with brackish water by
truck and boat, and stocked directly into 22 m3 predator-free acclimation enclosures and held for
3 d (a protocol to improve post-release survival rates, Brennan et al., 2006). Enclosures were
located along shorelines, mostly vegetated with mangrove (Rhizophora mangle), oaks (Quercus
spp.) palms (Sabal palmetto), and Brazilian pepper (Schinus terebinthifolius), and at some
locations near boat docks and riprap. After 3d of acclimation, snook were allowed to swim
freely from the enclosures. Release numbers were proportional (either 100% or 10%) to pre-
release abundance estimates of juvenile snook in each experimental creek.
94
Post-Release Evaluation
Post-release creek-wide standard sampling efforts were performed in each creek six times
after the release (approximately during June, August, October, December 2002, February 2003
and June 2003, Table 2). I identified year classes of wild snook according to length-at-capture
based on (1) length frequency graphs of captured snook, (2) corresponding sizes of cultured
snook, and (3) tag-recapture data on known young-of-year snook. For hatchery-reared snook I
adjusted catch by accounting for expected tag loss over time (e.g., VIE tag loss after 6 months
was 3%, [Brennan et al., 2005]). All captured snook were counted, measured, and checked for
the presence of CWTs with magnetic tag detectors, and visually examined for VIE tags.
Untagged wild snook and first-time recaptured hatchery-reared snook were tagged with VIE tags
in the caudal fins (using a unique color for each creek) and CWTs. For subsequent recaptures,
snook were re-tagged with a VIE of a specific color (BC = orange, WB = green, NC = yellow,
SC = pink) and implant location (caudal or anal fin) to identify the creek in which it was
captured, and the number of times recaptured.
Results
Gear Efficiency
Overall we performed 21 depletion-removal trials. Of these, juvenile snook were
captured in 12 trials and these data were used to generate estimates of seine efficiency. The
maximum likelihood estimate of pcommon value was 0.79 (90% CI = 0.75 – 0.83, 3 for likelihood
profile for the estimate).
In April 2002, our “leap-frog” sampling regime resulted in 117 standardized seine hauls
performed throughout the four experimental creeks. Our samples represented about 4 - 10% of
the total shoreline habitat in these creeks (Table 3-1, 3-1). We captured 505 snook and of these
183 were estimated as age-1 juveniles (2001 cohort, ~ 10 months old). After adjusting for seine
95
efficiency and extrapolating mean CPUE to creek-wide shoreline distance I estimated in-creek
juvenile populations to be 375 juveniles in BC, 476 in WB, 2018 in NC, and 639 in SC (Table
3-1).
Tagging and Release of Hatchery-Reared Snook
Hatchery-reared snook ranged from 84 - 270 mm FL at the time of tagging (mean length
= 177.5 mm FL +/- 2.95 SE) and samples collected in April 2002 of wild juvenile snook of the
same age were 79-219 mm FL (mean= 155 +/- 2.43 SE). We tagged and released a total of
2,372 hatchery-reared juvenile snook (Table 3), and six days after tagging, CWT retention in the
snook averaged 99.5% (from 14 groups with an average of 84 fish / group), and VIE retention
was 100%. Total transport time to the release sites was ca. 2 – 5 hours for all release groups and
releases occurred between 1230 and 1900 hrs.
Post-Release Evaluation
Overall, from 1 June 2002 – 30 March 2003, we performed 556 standard seine hauls,
captured 3261 snook and captured, tagged, and released 1588 juvenile snook (Table 3-2). Of
these, 275 were hatchery-reared snook recaptures, and 155 were wild snook recaptures. I found
little evidence of inter-creek movement; of 155 color-coded wild snook recaptures, only one was
recaptured in a creek other than its release creek (about 0.6% found in other creeks) (Table 3-2).
Because all hatchery-reared snook were originally tagged and released with a common VIE
color, I were not able to obtain estimates of inter-creek migration from first-time recaptures.
Only 11 hatchery-reared snook were recaptured twice, and all of these were captured in the same
creek as the first recapture event.
Samples from June 2002 resulted in a total of 410 captured snook, 294 of which were
juveniles. Relative abundance of wild juvenile snook in June was 0.86x, 1.87x, 0.63x, and 2.45x
the April abundance for BC, WB, NC and SC, respectively (Figure 3-4). However, when
96
hatchery-reared fish were included three creeks showed abundance increases compared to April
(BC =1.20x, WB =2.26x, NC =0.81x, SC = 2.61x that of pre-release estimates). In June,
abundance estimates of hatchery-reared fish were 120% larger than wild juvenile abundance in
WB and 40% of the wild abundance in BC (but 97% in July), 29% of NC wild abundance and
6% of SC wild abundance (control creeks) (Figure 3-5a, Figure 3-5b).
Juvenile snook generally emigrated from all four creeks by fall, then returned in winter
(Figure 3-4). As summer progressed, in-creek abundance of age-1 juveniles steadily declined,
and only young-of-the-year snook (from 2002 summer spawns, < 100 mm FL) numerically
dominated snook catches. For example, in early summer, hatchery-reared snook abundance was
37% of the number released in BC, ~100% in WB, 85% in NC, and 75% in SC, respectively. By
fall, large declines in hatchery-reared fish were evident in all creeks; 6%, 6%, 17%, and 0% of
the original numbers released in BC, WB, NC, and SC, respectively (Figure 3-4). Wild age-1
snook exhibited similar patterns of large, late-summer declines in abundance in all creeks. After
the first significant drop in temperature (below 22oC), both wild and hatchery-reared fish
abundance returned to substantial levels comparable to early-summer abundance; winter
abundance of wild snook was 120% the peak summer abundance in BC, 134% in WB, 96% in
NC, and 89% in SC. Winter abundance of hatchery-reared snook was estimated at 54% of the
summer abundance in BC, 78% in WB, 69% in NC, and 258% in SC (Figure 3-4).
Power analysis showed that, given the variability in the catch data and the small number
of replicates used in the study, power was low (0.3 - 0.8) for detecting a 100% difference in
population means. Nonetheless, repeated measures analysis showed a significant time and
treatment interaction effect for pre- and post-release abundance (p=0.024, F = 10.78, df = 2, 3).
Regression analysis showed no significant relationship between augmentation level and change
97
in abundance from pre- to post-release (using either summer or winter as post-release values;
p=0.42, p=0.45, R2=0.09-0.33).
Creek Density
The treatment creeks used in this study were quite different in their ability to
accommodate augmented snook densities. The analysis of covariance model showed no
significant differences in pre-release snook density (p=0.51, 3 df, Figure 3-6), however, after
releases occurred, WB showed a significant increase in overall density (over twice the pre-
release level) that was maintained through the winter (p=0.034, 3-6). In BC (also a high
augmentation treatment), the expected augmented levels were not sustained (Figure 3-6), and I
estimated that 64-85% of the released cultured fish were lost by the time of the first post-release
sample. Loss within wild snook populations was not detected over this period, however. Creeks
that received low levels of augmentation showed no significant changes in overall juvenile snook
density throughout the study (Figure 3-6). Chi-square analysis of the observed and expected
densities in all four creeks reflected the initial loss of hatchery fish in BC by showing a
significant difference between observed and expected post-release densities (p<0.001, Chi-
square= 306.82, 3 df). Thus, responses of juvenile snook density to various levels of
augmentation were creek specific.
Discussion
Experimental releases of high-densities of age-1 hatchery-reared snook in two estuarine
tributaries elevated total age-1 abundance above pre-release levels throughout the study. At the
same time, I found no detectable evidence of suppressed abundance of wild age-1 snook in these
habitats. Surplus production capacity for age-1 snook existed at these particular experimental
times and locations, although inter-cohort or community-wide responses were unclear. This is
not surprising as others have found population sizes below habitat productive capacities (e.g.,
98
Orth and Maughan, 1982; Conder and Annear, 1987). Regardless of a habitat’s productive
capacity, stocks (such as snook) at low levels, with high fecundity, at the edge of their range, and
exposed to high environmental variation, typically demonstrate highly variable recruitment
(Myers, 2001), and thus can occur at densities below capacity.
This study examines an important assumption about recruitment processes – that density-
dependent survival of wild fish may be reduced by the addition of hatchery-reared fish. In
salmonid populations, evidence exists for density-dependent responses in wild stocks due to
stocking hatchery-reared fish (Nickelson et al., 1986; Nickelson, 2003; Kostow and Zhou, 2006).
Density-dependent growth responses were not measured in this study, but other work in the same
creeks (Brennan and Leber, unpublished data), showed suppressed growth in tributaries with
high densities of age-1 snook. I found no evidence of high augmentation treatment effects on
wild conspecific density, although an effect may have been expressed through initial loss of
hatchery-reared fish in BC where the productive capacity may have been substantially exceeded.
Abundance patterns of wild age-1 snook throughout the study, however, followed similar
patterns in all creeks.
After emigration from the creeks in fall, there was movement back into the creeks of age-
1 wild snook in winter, resulting in abundances slightly higher than those in summer. Given the
low observed rates of inter-creek movement (0.6% mixing between experimental creeks), the
additional abundance in creeks of wild age-1 snook during winter (observed in all creeks) was
probably due to immigration of wild snook from other sources, emphasizing their importance as
thermal refuge habitat. Furthermore, hatchery-reared snook demonstrated the same seasonal
movement pattern and (after initial loss of fish in one treatment creek) early-summer abundances
were very similar to winter abundance in both treatment and control creeks. Declines in
99
abundance of hatchery-reared snook from summer to winter (after the initial high post-release
loss), presumably reflected mortality rates in resident wild age-1 snook.
The use of a fixed-station and systematic sampling program was appropriate in our study
because it incorporated a representative spectrum of population and habitat size/quality gradients
in the creeks (by attempting to sample every third 30 m section of shoreline habitat, see Hillborn
and Walters, 1992). Increases in abundance from April to June (as seen in WB and SC) may
reflect improvements in the efficiency and sampling ability of field workers. Such
improvements in sampling efficiency or catchability with experience are not uncommon (e.g.,
Walters and Maguire, 1996). In our experimental creeks, it is doubtful that abundance of wild
age-1 snook actually increased (via immigration) prior to large-scale emigration of juveniles
from the creeks in the fall. Furthermore, samples from NC and BC showed expected declines in
abundance. Logistical considerations and cost restricted replication and sampling effort for this
study. The low experimental power meant that the probability of a Type-II error was high, yet
repeated measures detected a significant time*treatment interactive effect on population means
and thus treatment-effect size was high. Furthermore, sampling over time showed consistent
patterns in all of the creeks (e.g., abundance declines in fall, and increases in winter), and further
supports our confidence in our estimates of abundance.
Inter-annual variation could have a strong effect on the results of similarly staged release
experiments performed at other times (through variations in carrying capacity, recruitment, and
subsequent competition and predation; see Walters and Martell, 2004; Brennan, unpublished
data). Common snook stocks in Florida have characteristics of high recruitment variability
(Myers, 2001) and, even over a 30 km stretch of coastal habitat, evidence exists for strong
variation in intra-annual recruitment of juveniles (Brennan, unpublished data). As snook are also
100
cannibalistic (Tucker, 2003; Adams and Wolfe, 2006), and age-0 and age-1 snook share nursery
habitat for a prolonged period and increases in abundance of age-1 snook may have negative
consequences on young-of-year snook populations through cannibalism and competition for
refuge and food resources. Cyclic patterns of abundance due to cannibalism are common
(Frankiewicz et al., 1999; Sanderson et al., 1999; Fromentin et al., 2000; Claessen et al., 2000;
Persson and de Roos, 2006). Snook populations in Florida show strong evidence for patterns of
alternating fluctuations of abundance between age-0 and age-1 year classes (Brennan,
unpublished data) as do a cousin of snook, barramundi Lates calcifer in Australian estuaries
(Walters and Martell, 2004; Griffin, unpublished data) due to intra- and inter-cohort cannibalism
and competition. Release programs that elevate age-1 abundance could intensify such effects
and a cautionary approach is warranted. Nonetheless, the potential for elevating stocks that are
below capacity is demonstrated by this study and its potential over the long-term must be
evaluated.
The presence of small (~ 60 mm FL) snook in the creeks during fall reflects continued
use of creek habitat by young snook throughout the year. Dramatic declines in abundance of
age-1 snook in the creeks during late summer – early fall (~ 200-325 mm FL at this time),
probably represents an ontogenetic habitat shift of the age-1 snook. This could possibly be due
to (1) a reduction in predation threat as snook attained larger sizes (Koczaja et al., 2005), (2)
superior foraging habitat and improved prey availability and preferences outside the creeks
(Ruzycki and Wurtsbaugh, 1999), (3) better water quality outside the creeks (Brennan and Leber,
unpublished data), or (4) combinations of the above reflecting tradeoffs with size-dependant
predation threat, and improvements in growth due to better foraging opportunities (e.g., Werner
and Gilliam, 1984; Walters and Juannes, 1993; Olsen et al., 1995).
101
Although our results indicated that stocking was additive in both high augmentation
treatments (BC and WB), depressed survival of the hatchery-reared fish occurred in one creek
(65-85% loss). While density-independent causes cannot be ruled out, it may indicate that the
productive capacity in BC was exceeded, and may reflect variation in the productive potential of
various juvenile snook habitats. Pre-release density of age-1 snook in BC was lower than in the
other creeks, yet density in itself was not predictive of the habitat’s potential to accommodate
more snook. In BC, shoreline habitat alteration in the form of seawalls, and steeply dredged
shorelines has been extensive, and ~ 65% of the total shoreline habitat has been altered (Table 3-
1). Shoreline alteration in the other creeks was much less (about 14% in WB, <2% in NC, and
7% in SC). Age-1 snook densities in NC and SC were similar, however, post-release densities
in WB were ~ 8 times higher than BC densities and four times that of NC and SC. While this
suggests an effect on a habitat’s productive capacity, variation in recruitment, anthropogenic
development, and low sample size confound such a comparison.
High juvenile snook abundance was common in habitat characterized by deep mud,
overhanging vegetation, gently sloping banks, and low current (<0.1 m/s or stagnant) (Chapter
2). In our experience, this type of habitat is scarce in highly developed creeks with dredged
bottoms, and shorelines modified with seawall channelization and boat docks (e.g., BC). These
altered habitats often attract predaceous fish possibly due to improved water exchange (from less
flow restrictions) or dock lights that attract bait fishes. Such conditions may be costly to snook
juveniles through loss of predation refuge and increases in predator abundance. Further research
is needed to quantify the influence of anthropogenic alterations on juvenile snook abundance to
identify potential loss of production.
102
In this study, juvenile snook nurseries accommodated abundance increases that persisted
over time. Clearly, the use of four experimental tributaries during a particular year to conduct
manipulative tests on recruitment, albeit logistically difficult, is insufficient to capture a broad
picture of recruitment dynamics and the long-term productive capacity of various systems.
Future studies should also focus on smaller and younger size classes to identify potential survival
bottlenecks (e.g., Doherty et al., 2004), and effects of multiple year classes sharing important
refugia, where density-dependent growth and mortality may be strongly influential.
Understanding community-wide implications and predator-prey interactive effects (e.g. Walters
and Kitchell, 2001; Persson et al., 2007), and interannual variation in recruitment and habitat
capacity, are important considerations that may best be addressed by coupling ecological models
with empirical experimentation. Stock enhancement programs should consider strategies that
maximize their cost-effectiveness while minimizing threats to wild stocks. Empirical studies
such as this, however, are important first steps toward understanding these dynamics and can
provide baseline information for dynamic age-structured stock recruitment models, often lacking
data for juvenile life stages.
103
Table 3-1. Results from standardized sampling in April 2002 and physical attributes of experimental creeks (shoreline distance and percent altered habitat). Population estimates are of age-1 snook found in each creek.
Experimental Number of Number of Total Sampled Sampled shoreline % Altered Mean CPE Adjusted Variance Calculated Calculatedcreek seine hauls shores sampled shoreline (m) shoreline (m) (% of total) shoreline mean CPE age-1 abundance max abundanceBowlees Creek 28 38 14051 1159 8.25 65 0.64 0.81 2.07 375 1,715
Whitaker Bayou 18 26 7650 793 10.37 14 1.50 1.90 9.30 476 1,913
North Creek 45 66 25189 2013 7.99 2 1.93 2.44 13.82 2,018 8,003
South Creek 26 34 24262 1037 4.27 7 0.63 0.80 2.07 639 2,904 Totals 117 164 71153 5002 7.03 3,508 14,535
104
Table 3-2. Mark-recapture results for the four experimental creeks organized according to sample month
Sample Creek Effort Captured, tagged, and released Recaptures MigrantsMonth (No. seines) Wild Hatchery (recaptures) Hatchery 2x Wild Wild 2x N OriginJune Bowlees Creek 27 16 12 0 1 0 02002 Whitaker Bayou 16 41 65 0 0 0 0
North Creek 45 91 23 0 8 0 0South Creek 26 62 4 0 2 0 0
July-August Bowlees Creek 28 22 22 1 3 0 02002 Whitaker Bayou 16 13 12 0 4 0 0
North Creek 46 79 23 1 13 0 0South Creek 26 33 2 0 7 0 0
Aug-Sept. Bowlees Creek 28 20 4 2 0 0 02002 Whitaker Bayou 16 6 4 0 1 0 0
North Creek 45 124 14 1 5 0 0South Creek 29 19 1 0 2 1 0
Oct. - Dec. Bowlees Creek 26 72 7 0 4 0 02002 Whitaker Bayou 16 37 0 0 2 0 0
North Creek 34 317 5 1 14 0 0South Creek 0 0 0 0 0 0 0
Jan.-March Bowlees Creek 28 4 5 0 7 0 1 Whitaker Bayou2003 Whitaker Bayou 16 36 45 4 11 0 0
North Creek 46 239 18 1 51 1 0South Creek 26 17 8 0 11 0 0
June-July Bowlees Creek 28 14 1 0 1 0 02003 Whitaker Bayou 16 6 0 0 2 0 0
North Creek 45 38 0 0 6 0 0South Creek 27 7 0 0 0 0 0
1313 275 11 155 2 1 Note: “Hatchery 2x” and “Wild 2x” represent snook that were recaptured in two subsequent sampling events. Data are only for the juvenile snook cohort (age-1).
105
Table 3-3. Numbers of hatchery-reared snook released at four experimental release sites Release date Release group Size class Pre-release Expected Jun-02
small medium large jumbo Totals wild N % cultured % culturedMay 20, 2002 Bowlees Creek 106 454 169 160 889 375 70.34 29.03
Whitaker Bayou 147 514 182 182 1025 476 68.28 53.27North Creek 134 178 64 60 436 2018 17.77 23.66South Creek 40 52 18 18 128 639 16.69 6.25
Totals 321 1198 433 420 2372 3508 Note: Snook mean lengths were 138 mm FL for the ‘small’ size class, 160 mm FL for ‘medium’, 188 mm FL for ‘large’, and 218 mm FL for ‘jumbo’. Abundance estimates by creek for the age-1 wild snook before releases (April 2002) are provided with the expected and observed (in June 2002) percentage of hatchery snook in post-release collections.
106
Tampa BayTampa
Bay
SOUTH CREEK
INTERCOASTAL
WATERW
AY
BARRIER ISLAND
METERS
0 500 1000
SARASOTABAY
BOWLEES CREEK
0
METERS
250 500
HWY US 41
SARASOTABAY
HWY US41
WHITAKER BAYOU
METERS
0 250 1000
Mainland C
oast
METERS
NORTH CREEK
0 500INTERCOASTAL
WATERW
AY
82o29’9.05” W
27o 12’ 58.55” N
82o 29’ 56.21” W
82o 33’ 7.00” W
27o 21’ 12.71” N
82o 34’ 43.95” W
27o 24’ 47.81” N
82o 29’ 9.05” W
MAINLAND CO
ASTHW
Y US41
HWY US41
Figure 3-1. Map of experimental study sites along the coasts of Sarasota and Manatee
Counties, Florida. Dots represent locations where collections with the standardized seine occurred.
107
116 m
Blo
ck N
et
Blo
ck N
et
21 m (1,4,7)(2,5,8)(3,6,9)
30 m
30 m
36 mB
lock
Net
Blo
ck N
et
21 m
30 m
36 m
Blo
ck N
et
Blo
ck N
et
21 m
Block Net
(Standard Seine)(Standard Seine)
Shore
Shore
Open Water
Shore
Shore
Shore
A
30 m
B
C
Figure 3-2. Schematic of various depletion-removal methods including (A) “two-bank
depletion”, (B) “single-bank depletion”, and (C) “two-bank extended” depletion. Dark lines indicate block nets, lighter dashed lines indicate locations of seine efforts, and arrows indicate the direction of the seine hauls. The numbers in brackets represent the sequence and location of seine hauls (hauls 1 through 9).
108
0.00
0.20
0.40
0.60
0.80
1.00
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19
"two bank""single bank""two-bank extended"
Depletion ID
Prob
abili
ty o
f cap
ture
(p)
0
1E-12
2E-12
3E-12
4E-12
5E-12
6E-12
7E-12
0.65 0.69 0.73 0.77 0.81 0.85 0.89 0.93
Like
lihoo
d
Probability of capture
A B
Figure 3-3. Graphical representation of individual capture probabilities (p) of snook from
various depletion-removal methods (circles, triangles and squares, graph A) and the likelihood profile for pcommon (graph B).
109
0
500
1000
1500
2000
2500
3000
3500
4000
4500
5000
0
500
1000
1500
2000
2500
3000
3500
4000
4500
5000
0
500
1000
1500
2000
2500
3000
0
200
400
600
800
1000
1200
1400
Bowlees Creek Whitaker Bayou
North Creek
Jun Aug Oct Dec Feb Apr2002 2003
Jul Sep Nov Jan Mar MayMay Jun
South Creek
Jun Aug Oct Dec Feb Apr2002 2003
Jul Sep Nov Jan Mar May Jun
Jun Aug Oct Dec Feb Apr2002
Jul Sep Nov Jan Mar MayMay Jun2003
Jun Aug Oct Dec Feb Apr2002
Jul Sep Nov Jan Mar MayMay Jun2003
Low Augmentation Creeks
High Augmentation CreeksN
umbe
r of s
nook
Num
ber o
f sno
ok Post-release loss
wildhatcheryhatchery
wildhatcheryhatchery
wildhatcheryhatchery
wildhatcheryhatchery
Figure 3-4. Abundance estimates of wild (grey area) and hatchery (white area) age-1 snook
over time in treatment creeks receiving high augmentation levels (top graphs) and low augmentation levels (bottom graphs). Error bars are from abundance estimates obtained from 90% confidence intervals around pcommon values. Arrows indicate the timing of hatchery-reared snook releases. General increase in wild abundance from summer to winter indicates immigration from other sources.
110
0
1
2
3
4
5
6
7
8
9
BowleesCreek
WhitakerBayou
NorthCreek
SouthCreek
0%
50%
100%
150%
200%
BowleesCreek
WhitakerBayou
NorthCreek
SouthCreek
BowleesCreek
WhitakerBayou
NorthCreek
SouthCreek
BowleesCreek
WhitakerBayou
NorthCreek
SouthCreek
Rel
ativ
e A
bund
ance
Cat
ch p
er e
ffor
t
a b
Figure 3-5. Results from post-release samples collected in June 2002 showing catch-per-unit effort (graph a) and relative abundance (graph b) of hatchery snook (white bars) to wild juvenile (grey bars) abundance (set at 100%).
111
CHAPTER 4 CRYPTIC CANNIBALISM IN SIZE-STRUCTURED SNOOK POPULATIONS
Introduction
Many species experience elevated mortality during specific ontogenetic stages, often
highest during early life stages (Smith and Fretwell, 1974; Jakobsson and Eriksson, 2000; Einum
et al., 2006). In fish, this especially occurs during larval stages (Houde 1987; Doherty and
Fowler, 1994; Lorenzen, 1996). While many processes contribute to early-stage mortality,
predation can be a primary contributor (Mol, 1996; Duffy et al., 1997; Taylor and Danila, 2005;
Titelman and Hansson, 2006). Nonetheless, high mortality can also manifest in later stages with
reduced refuge or increased foraging risk (e.g. Mooij et al., 1996; Byström, 2004).
Cannibalism, common among piscivorous fish, can be a significant contributor to
mortality during these early life stages (e.g. in Eurasian perch (Persson et al., 2000), cod
[Nuenfeldt and Koster 2000], cape hake [Pillar and Wilkinson, 1995], pollock [Bailey, 1989;
Wespestad et al., 2000], silver hake [Garrison and Link, 2000], and walleye [Liao et al., 2004]).
In controlled high density settings (i.e. aquaculture) cannibalism is commonly documented
(Baras et al., 2003; Chang and Lio 2003; Tucker, 2003; Correa and Cerqueira, 2007), yet
documentation of cannibalism in the wild is less common. This may be attributable to
inappropriate collection times or sample sizes (Bowen and Allanson, 1982; Johnson and
Dropkin, 1993), difficulties in identifying digested prey items with fast gut evacuation rates
(Kennedy, 1969), sampling logistics,(Rindorf, 2004) or high variation in inter-annual recruitment
patterns if cannibalism is density dependent (Nuenfeldt and Koster, 2000). In cod (Uzars and
Plikshs, 2000), for example, cannibalism rates are density dependent, with elevated cannibalism
during high recruitment, but negligible otherwise.
112
Morphological characteristics (Andraso, 1997; Sillett and Foster, 2000; Magnhagen and
Heibo, 2004; Palma and Stenek, 2001) and behavioral traits (e.g., schooling or ontogenetic
habitat shifts) can affect predation risk (Gallego and Heath, 1994; Wennhage, 2000). Many of
the traits are adaptations that minimize predation risk, such as body depth, spines, or coloration,
and can vary allometrically. Understanding how these traits change with body size can give
insights into life stages most vulnerable to cannibalism and predation.
Cannibalism is common to many species and may provide nutritional and energetic
benefits, reduced competition for food and refuge elimination of potential conspecific predators,
and reduction in foraging risk for juveniles (see Claesson et al., 2004 for review). Cannibalism
risks include choking and suffocating (commonly evidenced in hatcheries), bodily harm from
handling of prey, incorporation of conspecific pathogens, increased predation risk by reducing
school size, and reduction in inclusive fitness from consuming kin. Nonetheless, the prevalence
of cannibalism in many organisms suggests a net evolutionary advantage.
In Florida, aspects of stock abundance, productivity and regulating mechanisms have
been of particular concern for fishery managers of the common snook Centropomus undecimalis
(“snook” herein). A coastal, warm-water species, inhabiting estuaries and river systems
(Marshall, 1958; Fore and Schmidt, 1973), snook are subjected to intense recreational fishing
pressure in Florida; stocks are considered over-fished, and recruitment has declined over the past
20 years (Muller and Taylor, 2005). Widespread habitat loss (Bruger and Haddad, 1986), winter
freezes (Storey and Guder, 1936; Marshall, 1958), and red tide blooms also threaten snook
populations in Florida. These factors have led fishery managers to investigate the feasibility of
using stock enhancement to augment snook populations in Florida (Brennan et al., 2005;
Brennan et al., 2006). Despite the potential benefits, stock enhancement can have deleterious
113
effects on wild stocks (Hilborn, 2004; Walters and Martell, 2004). As a result, experiments are
critical to assess the potential feasibility of stock enhancement prior to implementation of a
large-scale program (Blankenship and Leber 1995; Hillborn, 1998; Walters and Martell, 2004).
Cannibalism is one factor that may lead to unintended effects of stock enhancement.
In snook, cannibalism has been documented in aquaculture (Serfling, 1998; Tucker,
2003) and is most intense during early life stages in hatcheries (Tucker, 2003; N. Brennan,
personal observation). In the wild, snook juveniles actively prey on small fishes and crustaceans
(Fore and Schmidt 1973; Gillmore et al. 1983; McMichael et al. 1989; Sazima 2002), and adults
are highly piscivorous (Blewitt et al., 2006), yet cannibalism is only rarely documented (e.g.,
Adams and Wolfe, 2006; N. Brennan, unpublished data). Therefore, cannibalism could be
perceived as non-significant factor influencing population size. However, even low observed
rates of cannibalism or the threat of cannibalism could strongly influence population dynamics
when considering competition for limited refuge in the presence of predators, and prolonged
seasonal threats of cannibalism. Identifying cannibalism from dietary studies in the field may be
a relatively rare event compared to more common prey choices of a particular predator of various
ages, especially when older age classes are not always occupying the same habitats as juvenile
conspecifics. A single incidence of cannibalism results in mortality of a prey individual and
potentially the predator (if the cannibal also chokes and suffocates). Over time (throughout a
season or year) cannibalism focused on a year class of juveniles could act as a strong force in
structuring abundance of the year class (Claesson et al., 2000; Neuenfeldt and Koster, 2000;
Persson et al., 2000), especially during years of high abundance of particular year classes.
In many species, ontogenetic changes in morphology, and behavior suggests adaptations
in earlier stages for coping with smaller absolute body size in the presence of predation and
114
competitive threats from larger animals sharing the same habitat (Werner and Gilliam, 1984;
Mondor and Roitberg, 2002; Harrel and Gibb, 2006). Such adaptations during early ontogenetic
stages can include proportionally larger gapes, predator avoidance capabilities (Harrel and Gibb,
2006), and a disproportionate tendency to cannibalize conspecifics (Byström, 2006). Evidence
found in the life history of snook suggests strong interactive potential between various
ontogenetic stages. In Florida, snook have a protracted spawning season (Taylor et al., 1998)
and subsequent recruitment period to early-stage rearing habitats (McMichael et al., 1989; Peters
et al., 1998a). Here, early and late season cohorts share rearing habitat and the potential for
competition and cannibalism is high. Evidence for competition and predation within a year
class has been documented in other species (e.g. Olson et al., 1995; Miranda and Hubbard,
1994). With snook, the potential for interspecific cannibalism is also high because during
winter, adult and juvenile share the same habitat (Adams and Wolfe, 2006; also see Chapter 2).
Other work (Chapter 3) has also indicated that juvenile snook abundance is density-dependent
and may be limited by habitat availability.
In this study I evaluate the ontogenetic potential for snook cannibalism in laboratory and
field settings using an allometric approach, and estimate the incidence and impact of cannibalism
in wild snook stocks. My hypotheses were: (1) ontogenetic differences in morphology and
coloration reflect high predation risk and cannibalism potential in small snook, (2) the potential
for intra-cohort cannibalism decreases with increasing predator size, and (3) cannibalism on age-
0 snook cohorts in the wild, although infrequently observed, can have a strong influence on
population size. I used three approaches to address these hypotheses; (i) quantitative assessment
of allometric variation in phenotypic characteristics of snook, (ii) a field experiment to quantify
cannibalism rates of different sized predatory snook, (iii) a review of published and unpublished
115
literature on snook diet, and (iv) the development of an age-structured model parameterized with
data from i-iii that simulates cannibalism on age-0 snook in a tidal creek.
Materials and Methods
Generalized Snook Life History in a Tidal Creek
Snook in Florida are known to primarily spawn from May – September (Taylor et al.,
1998). Spawning grounds include barrier island passes and along beaches (Taylor et al., 1998).
Post-larvae recruit to a variety of nearshore habitat (Peters et al., 1998a) including estuarine tidal
creeks and backwaters through the summer and fall (McMichael et al., 1989; Peters et al.,
1998b). By November age-0 cohorts can weigh approximately 12 g and remain in these rearing
habitats through the winter until the following spring and summer when they undergo an
ontogenetic habitat shift to a variety of habitats including seagrass beds (Gilmore et al., 1983)
and beaches (Taylor et al., 1998). By the following winter, at the onset of the first significant
cold snap (when water temperatures drop below about 22oC), yearling snook (about age-1.5) and
adults move into winter habitats for thermal refuge (often back to their rearing habitats such as
creeks and backwaters where age-0 snook are rearing; creek temperatures are often about 5oC
warmer than non-creek habitat, Chapter 3). After temperatures warm again in the spring,
adolescent and adult snook move back to beaches, passes and grass flats (Brennan, unpublished
data).
Approach 1: Phenotypic Characteristics
To test hypothesis (1), I examined differences in morphological and coloration
characteristics of 256 hatchery-reared snook (39-395 mm Fork Length [FL]) and 170 wild snook
(88-903 mm FL). Hatchery snook (F-1 generation, eggs collected from wild snook) were reared
at Mote Marine Laboratory, Center for Aquaculture Research and Development, located in
Sarasota, Florida, USA. Wild snook were collected in creeks and estuaries of Sarasota and
116
Manatee Counties (Figure 4-1). Measurements included total length (TL), FL, standard length
(SL), vertical gape (VG, defined as the vertical distance between the anterior tip of the
premaxlilla and the anterior tip of the dentary bone, as in Luczkovich et al. 1995), body depth
(BD, measured at the widest part of the body, as in Luczkovich et al. 1995), anal spine (AS)
length, and distance from the ventral gill arch (hypohyal bone) to the anus (VA-AN). I also
examined allometric changes in the first dorsal fin markings, as juvenile snook have a dominant
black “spot” at the tip of the dorsal fin. Morphometrics were compared graphically against body
length (i.e. VG, AS, VA-AN, BD). Because growth is an exponential phenomenon, data were
loge transformed and fit to linear regression models with least squares procedures (with 95%
confidence intervals). Therefore, a slope parameter of 0 meant that changes in compared
variables were independent, slopes =1 showed an isometric scaling and slopes around this
showed negative (the response variable scale proportionally larger with decreasing body length)
or positive allometry (the response variable scale proportionally larger with increasing body
length).
Approach 2: Field Enclosure Trials
The second goal of this study was to test the hypothesis that cannibalism rates are higher
for smaller-sized cannibals compared to larger ones. I examined this in field enclosures using a
factorial experimental design in which I varied: (1) predator size (small vs. large), (2) relative
prey size (30% vs. 50% of the predator size), (3) prey density (1 vs. 10 prey per enclosure), (4)
the presence of other (non-snook) prey species (yes or no). The response variable was binary:
the occurrence of cannibalism. Each experimental combination was replicated three times, and
the entire design was blocked at two field locations, yielding a total of 96 trials (2 predator size
classes x 2 relative prey sizes x 2 prey densities x 2 alternate prey x 2 locations x 3 replicates).
117
Experimental trials were conducted in Sarasota County, Florida, in twenty field
enclosures; ten of these were located in an inland creek estuary (North Creek lagoon), and ten
located along the intra-coastal side of a barrier island (Mote site). Enclosures were 3 m x 3.6 m
rectangular, and 2 m high constructed with 3 mm vinyl mesh in a box shape with four sides and a
bottom sewn together using dental floss (Figure 4-2). Enclosures were attached at the corners to
2.4 m long steel re-bars (20 mm diameter) implanted in the substrate. Tree branches were placed
in each enclosure to provide additional refuge habitat. The tops of the enclosures were covered
with 3 cm square plastic bird mesh.
A single snook predator and selected snook prey were stocked simultaneously into each
enclosure and held for one week. Predator sizes were categorized into “small” (74-163 mm FL),
and “large” (218-332 mm FL) size classes. Snook prey, were approximately 50% (42-56%), and
30% (28-42%) the length of the predator. To test the relative effects of prey density on
cannibalism rates, half of the trials were stocked with one snook prey (about 0.1 fish per m2), and
half with ten snook prey (about one fish per m2). Within each of the above experimental
combinations, half of the enclosures were stocked with additional prey species (captured nearby).
One week after stocking, animals in the enclosures were harvested by removing the cable
ties from the re-bars and lifting the sides and bottoms out of the water. At harvest, all cages were
carefully searched for live and dead snook. A cannibalism incidence was assumed to have
occurred if prey snook were missing at harvest. Trials where the predator was missing or dead at
harvest were not included in the analysis. No stomach contents were collected. After harvest, a
new set of predator/prey combinations were randomly stocked into the enclosures until all trial
combinations were completed. Trials were conducted in October and November, 2003, prior to
onset of cooler winter temperatures. Cannibalism incidence was recorded in a binary format (yes
118
or no) and was fit to linear regression models through maximum likelihood methods using
logistic regression techniques (CATMOD procedure in SAS, [SAS, 1999]). I used Bonferroni
adjustments on treatment comparisons to reduce the likelihood of making a Type I error.
Approach 3: Cannibalism Intensity in Wild Snook Populations
Literature Review: I evaluated cannibalism intensity in wild common snook using
published and unpublished literature, (Table 4-1). Published literature was obtained from the
Web of Science database using the search criteria “Centropomus undecimalis” and “diet or
feeding or cannibalism”. I gained further literature by using citations of other work from the
reference lists provided in these manuscripts. Unpublshed material was obtained from work in
Sarasota from the snook stock enhancement research program. Whenever possible, I extracted
data on the size and numbers of the snook examined, collection year, season, and region of data
collection, habitat descriptions of where samples were taken, the percent of the samples
containing food items in the stomach, and the number of snook reported cannibalized.
Field Collections: I also assessed cannibalism rates in the field, by conducting my own
investigations in Sarasota, Florida with snook collected for Mote Marine Laboratory’s Stock
Enhancement Research Program from 2001- 2005 (Table 4-1). All collections occurred in
estuarine tidal creeks, an important rearing habitat for age-0 snook. Snook were collected with
bag seines during (1) June 2001 (2) February – May 2004, (3) July-August 2004, and (4) May-
August 2005. All snook were measured and stomach contents were extracted using a non-lethal
lavage technique (Adams and Wolfe, 2006), immediately preserved in 10% buffered formalin,
and returned to the laboratory for identification. Stomach contents were visually identified to the
lowest possible taxonomic level. For more digested stomach contents, I used any conspicuous
body parts for identification such as carapaces, rostrums, fish jaws and otoliths. Identification of
stomach contents beyond this were categorized as “unidentified”.
119
Age-Structured Cannibalism Model: Evidence of cannibalism in field collections of
snook (Adams and Wolfe, 2006; N. Brennan, unpublished data) and other fishes (Lang et al.,
2000; Persson et al., 2000; Vik et al., 2001) have shown that predation efforts are typically
focused age-0 individuals. As such, I developed a simplified age-structured forecast model to
estimate predation mortality from age-0, age-1, and adult (“age-2+”) snook age groups on an
age-0 year class in Sarasota, Florida. I extracted cannibalism rates (number of cannibals
observed in stomachs/ total stomachs examined) from studies that were conducted in conditions
where cannibalism was most likely to occur. In Sarasota, juvenile and adult snook over-winter
in tidal creeks (Brennan et al., 2008; Chapter 2 and 3), and emigrate back to summer habitat
through the spring (Gilmore et al., 1983, Brennan, unpublished data). In November, snook from
the age-0 cohort are generally large enough to be captured by our standard seines (1 cm mesh,
snook mean lengths about 100 mm FL). As such, I modeled age-0 cannibalism from November
1 – May 15 of a typical year in a tidal creek.
I used a series of variables in a cohort-based Monte Carlo simulation model to predict
how cannibalism mortality from each of the above age categories decreased the original
abundance of the age-0 year class while over-wintering in a typical rearing habitat. Within the
age groups and across all ages, I also measured the contribution of cannibalism mortality to the
total natural mortality for the period. Model calculations were based on real data collected in
Southwestern Florida in rearing habitats of age-0 snook (tidal creek). The model simulation time
frame was November 1 – May 15 of a hypothetical year. For the model, the number of age-0
snook in the hypothetical creek on November 1 was set to 5000. Length frequency data from
Sarasota snook collections in tidal creeks from 1998- 2005 (Brennan, unpublished data) were
used to obtain typical size estimates of each of the age categories on November 1. Size-specific
120
instantaneous natural mortality rates (MW) in a typical population growth function, were used to
calculate abundance of the age group i (γage-i) over time (t, number of days/365):
γage-it = γage-i, t0 * e (–MW
* t) (4-1)
Weight-specific instantaneous natural mortality rates extrapolated from Lorenzen (1996) were
calculated as a power function of the initial mean weight (W) of individuals in the age-i group on
November 1:
MW = Mu * Wb (4-2)
where, Mu is the annual natural mortality at unit weight and b is a scaling parameter (Lorenzen,
1996). Because age-1 and age-2+ age groups of snook are typically not fully recruited to the
tidal creeks (winter habitat) in November, I used the relative abundance in January (typical peak
abundance for all age groups in the creeks) from samples collected in tidal creeks from 1998 –
2005 (Figure 4-4) to calculate the older age group’s abundance based on the age-0 January
projected abundances. Then using equations 4-1 and 4-2 (and weight-based instantaneous
mortality rates of these age groups [Lorenzen, 1996]) I back calculated the age-1 and age-2+
abundance to November 1 (t=0) to obtain γage-i,t0. During the simulation period, the model
calculated daily abundance estimates for each age group that were weighted according age-
specific residence patterns (p i,t = probability of γage-i,t in the creek at time t) across the time
period (see below, Table 4-2, Figure 4-3).
For the model period, I assumed cannibalism mortality was proportional to natural
mortality of the age-0 cohort. I used field data from snook collections that occurred from 11
November – 14 May (i.e. winter and spring data from Adams and Wolfe, [2006], and Brennan;
see Table 4-1) to obtain an estimate of the proportion of the creek population cannibalizing
snook (λi, i.e. number of age i snook observed with cannibalized prey in their stomachs/total
121
number of age i snook stomachs examined). I then calculated daily cannibalism mortality (c) in
the creek by age group i as:
ci,t = γage-i,t * p i,t * λi (4-3)
Total cannibalism (∑ci,t ) within and across age groups was then calculated to give an
estimate of the total number of age-0 snook cannibalized during the model period.
Sources of variation in the model (Table 4-2) were:
(1) Initial size (FL, mm) of each age-i group in November. These data were based on length
frequency histograms of snook captured in Sarasota during November (1998 – 2005, snook stock
enhancement program standardized sampling [Brennan, unpublished data]). The model sampled
from a normalized distribution, based on the mean length and standard deviation obtained from
the field. After this, I used a power exponential weight-length model to convert lengths (FL,
mm) to weights (W, in g):
Wti = 0.000009 * FL (3.0104) (4-4)
(2) Instantaneous mortality rates. The model selected from normalized distributions of the
parameter estimates Mu and b (equation 4-2, from Lorenzen (1996) with associated standard
deviations.
(3) Creek residence patterns (p, expressed as the probability of being in the creek) were
calculated differently for each age group based on the best understanding of wild snook
abundance in the creeks in Sarasota. This was obtained from (a) length frequency distributions
from seasonal sampling in creeks and nearby habitats (Chapter 2, 3, Brennan, unpublished data),
and (b) recapture data of tagged hatchery-reared and wild snook (Chapter 2, 3). For age-0 snook,
I assumed that they gradually dispersed from the creeks as the winter and spring progressed, and
used an inverse cumulative normal distribution function (mean days in creek = 120, SD = 30). I
122
varied the parameter, “mean days in creek” around a normalized distribution with SD = 2. For
age-1 and age-2+ snook, I assumed they immigrated to the creek from outside habitats to over-
winter, and then gradually dispersed from the creeks as the season warmed. I used a normal
density distribution function, and the model selected from normally distributed parameters
(mean=50, SD = 5 for age-1 snook, and mean= 50, SD = 2 for age2+ snook).
(4) Relative abundance of age-1 snook (compared to the age-0 cohort’s abundance) varied within
a uniform distribution (0.10 – 1.51 for age-1, 0.05 – 0.70 for age-2+).
(5) Estimates of cannibalism rates (λi) were based on a normalized distribution of the mean field
rate (overall number of cannibals found/number stomachs examined) with a variance of 1 SD of
the mean. I generated variance around these rates through bootstrap random sampling of the data
set (individual snook of age group i expressed as 0 [no cannibalism] or 1 [cannibal]). I used
Microsoft Excel for bootstrapping procedures (Teknomo, 2006) where the original data set was
randomly sampled to generate 100 data sets (n=1,000 per set). I used sensitivity analysis to
assess the influence of the above mentioned parameter estimates on the results. The model
performed 10,000 iterations using Microsoft Excel with the Crystal Ball add-in (Werckmand et
al., 1998).
Results
Phenotypic Characteristics
Mouth gape, body depth, and anal spine length showed negative allometric scaling with
body length. Mean slopes from linear relationships of the log transformed variables above were
0.664-0.901 (Table 4-3, and Figure 4-5) with the anal spine (slope=0.664) showing strong
allometric scaling. Isometric scaling was found for log transformed relationships of SL and FL,
and VA-AN and FL, however. The dark spot located at the tip of the first dorsal fin was
123
prominent in snook juveniles (30 – 125 mm FL) but in snook generally larger than 175 mm FL
pigmentation was much reduced and only outlined the tip of the fin (Figure 4-6).
Field Enclosure Trials
One hundred one enclosure trials were attempted, but of these 27 contained combinations
of dead (23) or missing predators (15) or at least one dead prey snook (16). Trials with dead
prey or predator snook occurred mostly (25) during the experimental trial weeks of November
12, 2003 (12), or November 19, 2003 (13). There was no pattern of the abovementioned
mortality between the sites (North Creek lagoon = 11, Mote site = 14). Therefore, 74 trials were
used in the analysis, yielding 3-5 replicates per treatment (Table 4-4). The logistic regression
model showed a goodness of fit of 0.72 indicating a reasonable fit to the mode but after
correcting for Type I errors, no treatments or interactions between factors were significant
(Figure 4-8). The overall frequency of cannibalism was 0.307 +/- 0.057 (SE).
Cannibalism Meta-Analysis and Model Results
I found 17 studies (published and unpublished) describing diets of common snook (Table
4-1). Of these, 11 were conducted in Florida, 1 in Georgia, 2 in Mexico, 1 in Puerto Rico, 1 in
the Caribbean, and 1 in Venezuela. Although there was overlap, 13 studies collected samples
during summer, 12 during fall, 9 during winter, and 7 during spring. Overall 6,768 snook
stomachs were examined for dietary contents and the reported percentage of snook containing
food in their stomachs ranged from 21-89% with a mean of 62% +/- 5.1 SE. Using knowledge
of fish size and collection season, I estimated that 2,604 age-0 snook, 1,603 age-1 snook, and
1,753 age-2+ snook were examined for stomach contents (Table 4-1). In Sarasota, 2,515 snook
were examined for stomach contents, and of these, I estimated 801 age-0 snook, 1,133 age-1, and
518 adult snook. Of these 1,341 snook contained food contents in the stomachs (53.3 %). Two
studies (Adams and Wolfe, 2006, and Brennan, this work) specifically examined stomach
124
contents in juvenile habitats during winter (when juvenile habitat is shared with older age
groups). In these studies, 1,990 snook were examined for stomach contents (Table 4-2), of these
1,405 (70.6%) contained food items
Only two the studies reported cannibalism for a total of 7 instances from 6,768 stomachs
examined (0.10%). Six (reported by Adams and Wolfe, 2006) were found in nursery habitat in
Charlotte Harbor, FL during winter months (Table 4-5). In Sarasota, I found one instance of
cannibalism from creek collections in February 2004 (Table 4-5). Based on body length and
collection time, all cannibalized snook were estimated as age-0 (mean length = 93.6 mm SL,
range 48-141 mm, mean weight = 8.8 g). Field data confirmed that smaller size snook
cannibalized proportionally larger snook prey (Figure 4-9).
Model Results
After 10,000 Monte Carlo simulations, using all relevant data, the model predicted that
cannibalism contributed to 18.9%, 13.3%, and 7% of the accrued age-0 mortality due to
predation from age-0, age-1, and age-2+ snook respectively (Figure 4-10. Considering initial
abundance of the age-0 cohort at the start of the model period (n=5,000, November 1), the model
predicted that by May 15, 0.114 of age-0 cohort had been cannibalized by age-0 snook, 0.174
from age-1 snook, and 0.192 from age-2+ snook.
Analysis of the model’s sensitivity to parameter variation, showed that results were
strongly controlled by variations in the estimates of field cannibalism rates. More specifically,
variations in field cannibalism rates within the age-0 cohort contributed to approximately 95% of
the estimates of cannibalism contribution to the age-0s total natural mortality (Figure 4-10). For
the same measurement, field based cannibalism rates for age-1 snook had an 88% influence on
size-specific results, and age-2+ field estimates of cannibalism rates had a 83% influence on the
125
results. For all age groups, variations in field cannibalism rates had more than a 99% influence
on predictions on the proportion of the initial cohort consumed by cannibalism (Figure 4-11).
Aside from field cannibalism rates, variation in the mean size of the age-0 cohort in November (a
density-dependent characteristic) was the second most influential (contributing to 21% for age-0
results) or third (contributing 13% for age-1, and 18% for age-2+ results) most influential when
predicting contribution of cannibalism to natural mortality (strongly associated “Mu” and “b”
were also highly ranked). Variations in relative abundance had a 40% influence on estimates of
the proportion of age-1 cannibalism to total age-0 mortality, and a 46% influence on the
estimated proportion of age-2+ cannibalism mortality on total age-0 mortality. Variation in
creek residence patterns contributed to 0.63 – 2.84% of the estimates of the proportions
cannibalism contributions to natural mortality or the proportion of the initial age-0 cohort
succumbed to cannibalism.
Discussion
Juvenile snook show many morphological features that are possible adaptations to reduce
predation threat, and evidence for potential cannibalism during juvenile stages. The
proportionally longer anal spine in smaller snook may aid in deterring predators, and even
physically block small fish from being swallowed. In rearing tanks with age-0 snook, I have
commonly observed snook with missing body scales from the anal spine to the tail, but scales
anterior to this were not missing; evidence that the fish had been held in the mouth of a larger
conspecific but not swallowed, most likely due to the anal spine. Snook body depth was
proportionally deeper in smaller snook and might also deter similarly-sized predators (including
cannibals) from swallowing a snook. Small snook also exhibit proportionally larger gapes than
larger snook allowing a proportionally greater prey size range for the smaller fish (see 4-7, 4-9),
and thus potentially allowing higher cannibalism rates. In a study comparing suction feeding
126
performance in catfish Clarias gariepinus of various sizes (111-923 mm TL) (Wassenbergh et
al., 2006), absolute suction feeding performance clearly increased with fish size. Small animals
have competitive disadvantages because of their smaller body size (Werner and Gilliam, 1984).
However, when catfish suction performance was scaled according to predator size and relative
prey size, small fish clearly outperformed the larger fish (Wassenbergh et al., 2006). This work
strengthens arguments by Harrel and Gibb (2006) that young animals of smaller sizes, have
compensatory adaptations to minimize this difference. These compensatory adaptations would
be especially important in fish that cannibalize within a cohort. Cannibalized prey are typically
proportionally larger than other prey items (Juanes, 2003), and cannibalism is often associated
with low alternative prey availability or high competition for prey forcing cannibalism among
similarly sized individuals (Persson et al., 2000; Walters and Martell, 2004).
Snook coloration suggests high predation threat among juveniles. Sazima (2002)
observed that juvenile snook generally resemble mojarra (Eucinostomus gula) in overall body
shape and have a prominent black spot at the tip of the dorsal fin. He theorized that snook use
aggressive mimicry and school with mojarra to take advantage of unsuspecting prey. I suggest
that in addition to aggressive mimicry, this spot (prominent in snook smaller than 125 mm, but
generally much less distinct in snook larger than 175 mm FL) (see 4-6) is a form of self mimicry,
an adaptation to confuse predators, causing them to strike at an apparent “eye spot” in a
peripheral, less-essential, body location (Blest, 1957; Wickler, 1968; Hailman, 1977; Robbins,
1980). The presence of an additional mark or spot on the body (in addition to the eyes) is
common in schooling species and probably adds to the predator “confusion effect” (Welty, 1934;
Landeau and Terborgh, 1986). Some birds have ‘flash marks’ in their plumage possibly as an
adaptation to confuse predators (Brooke, 1998); weasel’s black-tipped tails are thought to
127
confuse hawk predators (Powell, 1982). Fish in schools and shoals engage in anti-predator
behavior, and group size has been inversely related to predator success (Neill and Cullen, 1974;
Milinski and Curio, 1975; Major, 1978; Landeau and Terborgh, 1986). Many of these species
are commonly marked with a prominent body spot (e.g. Muglidae, Clupeidae Mullidae,
Gerreidae, Sciaenidae, Carangidae, Sapridae) that may aid in the “confusion effect” by doubling
the number of “eyes” seen by a predator. Eye spots, common in juveniles, reflect ontogenetic
changes in predation risk as the eye spots are lost in later life stages (e.g. snook, red snapper,
butterfly fish, pork fish, sea bass, pork fish, hog fish, wrasses, and cardinal fish). These
ontogenetic variations in morphology of snook – even within the age-0 year class (30 – 175 mm
FL) may be adaptations to compensate for small body size in the presence of predation threat and
competition (see Werner and Gilliam, 1983; Walters and Martell, 2004; Harrell and Gibb,
2006).
Results of cannibalism behavior in the field enclosure trials reflected morphological
differences in small- and larger-sized snook juveniles. Proportional prey sizes tested in the field
enclosures were below the maximum gape and length consumption capabilities of the cannibals
(82-62% predator size), and are probably more realistic. While means were not significantly
different between experimental groups (except for a significant increase in cannibalism rates
when prey were stocked at high density), mean cannibalism rates of small cannibals were
consistently higher than larger cannibals for all the various experimental treatments (Figure 4-8).
Performing the enclosure studies in the field gave the experiment a more realistic setting for
inducing cannibalistic behavior although actual cannibalism rates here were probably lower than
if treatments had been conducted in laboratory tanks and higher if done at natural prey densities.
Several logistical complications occurred in this study. Prey and predator snook found dead in
128
the enclosures, appeared to have died many days prior to examination (about 5d or more).
Mortality appears to have been handling-induced (93% from trials on November 12 and 19,
2003) and this is supported by evidence that the stocked snook died soon after stocking (highly
decayed remains). Increased replication of these treatments, the use of controls, and examination
of stomach contents to confirm cannibalistic behavior in the enclosures is needed to clarify
relative size-specific cannibalism potential in snook.
Snook diets revealed a surprisingly low rate of cannibalism (7 cases recorded of 6,768
stomachs examined, 0.10%) in populations over a wide variety of locations. Assuming these
rates are reflective of typical snook diet, the average number of snook processed for stomach
contents in a single study (mean=359) was just over 1/3 the number needed to identify a single
cannibalism instance. As snook cannibalism was targeted at and within the age-0 cohort, sample
location and timing could strongly bias cannibalism rates. Samples collected in age-0 habitats,
especially nursery habitats (see Beck et al., 2001), are therefore much more likely to contain
evidence of cannibalism. Within a given year, cannibalism rates are also likely to vary over
seasons due to variations in intra- and inter-cohort abundance and interactions of size-specific
habitat use (e.g. Byström et al., 2003). With snook in northern end of their temperature tolerance
range, older age classes of snook often over-winter in rearing habitats such as tidal creeks and
backwater habitats because of warmer water (Marcinkiewicz, 2007; Chapter 2) . In this
scenario, over-winter habitats coincide with snook nursery habitats and cannibalism may be
intensified.
Inter-annual timing of sample collection could also bias results as cannibalism rates are
strongly linked to snook density. Samples collected during years of low juvenile recruitment
will be less likely to contain evidence for cannibalism, while cannibalism is much more likely
129
during years of high juvenile recruitment (Henderson and Corps, 1997; Uzars and Plikshs, 2000;
Walters and Martell, 2004). Adams and Wolfe (2006) found exceptionally high rates of
cannibalism (6 of 280 snook stomach examined). Their study met two primary requirements for
a greater probability of detecting cannibalism: (1) collections were made in putative snook
nursery habitat and (2) during winter months when snook of all ages are concentrated in these
nursery habitats. In Sarasota, my samples met both of these criteria also, however only one
instance of cannibalism was found out of 1,710 snook stomachs. Both studies used the same
stomach lavage apparatus and techniques. These differences might be explained by differences
in juvenile density, a function of inter-annual and spatial recruitment variation. The samples
were collected during different years and in different regions. Snook habitats in Sarasota have
shown large differences in juvenile snook recruitment in coastal nursery habitats separated by
only 19-24 km (Brennan, unpublished data). The diet samples collected in Sarasota were
primarily collected during the winter of 2003-2004, a relatively weak recruitment year for the
age-0 snook, but a strong year for age-1 snook (Figure 4-4, relative abundance of age-1 and age-
0 snook in Sarasota).
Model results showed that collective cannibalism by the age groups could contribute
substantially to the age-0 natural mortality rates. Because field-based cannibalism rates
overwhelmingly influenced the model output, careful consideration must be given to collection
parameters such as sample number, sizes of the predators examined (and associated prey sizes),
sampling timing, and inter-annual influences in recruitment and ambient density of the various
age groups. In this model, I used a density-dependent relationship with instantaneous mortality
rates linked to body weight (Lonenzen, 1996). Body weight is inversely proportional to density
of competitors (Chapter 2, Keeley, 2001) and thus this function accounted for, in part, density-
130
dependent competition and predation rates. Further expansion of this model should include a
more direct influence of density-dependent functions, however. In the model, the age-specific
cannibalism results are, in some ways, independent of each other. Additionally, cannibalism-
induced mortality could theoretically exceed natural mortality rates because the model does not
relate ambient age-specific densities in the field (where cannibalism rates are a reflection of the
ambient densities) to modeled age-specific densities.
Variations in the simulated relative ratio of older age groups to age-0 abundance in
January also had a strong influence on model results (40% for output predictions of age-1
cannibalism contributions to age-0 natural mortality, and 46% influencing predictions of age-2+
cannibalism contributions to age-0 natural mortality). This strong influence of older cohort
abundance on age-0 snook cannibalism is reflected in the ‘saw tooth’ pattern of relative
abundance of the two age classes (Figure 4-4). Similar patterns have been observed in other
species (Barramundi [Walters and Martell, 2004], Pacific Hake [Smith, 1995], Japanese flounder
[Kellison et al., 2002], brown trout [Vik et al., 2001; Nordwall et al., 2001; Carline, 2006],
walleye [Kocovsky and Carline, 2001], walleye Pollock [Yamamura et al., 2001], Sacramento
pikeminnow [Gard, 2005], European sea bass [Henderson and Corps, 1997], vendace [Helminen
and Sarvala, 1994], freshwater fairy shrimp [Dumont and Jawahar, 2004], blue crab [Lipcius and
Vanengel, 1990]. This suggests that inter-cohort cannibalism could have a mediating influence
on subsequent strong year classes.
Adjustments to other parameters such as residency patterns (and associated variation)
could also be important in determining cannibalism intensity. Modeled variation around the
age-specific creek residence patterns was small (e.g. age-1 residence in the creeks peaked at d50
with only a standard deviation of 5d) and were based on migratory responses to thermal changes.
131
Inter-annual variation in water temperatures could change the timing and breadth of peak
abundance, however.
Bioenergetics approaches (e.g. Kitchell et al., 1977; Rice and Cochran, 1984; Rose et al.,
1999; Burke and Rice, 2002) are becoming increasingly popular in fisheries applications (Chips
and Wahl, 2008) and can be useful because they address physiological and activity-level
responses to environmental variation. Nonetheless, field data has consistently not well correlated
with model predictions (Ney, 1993; Bajer et al., 2004; Selch and Chips, 2007; Chips and Wahl,
2008). Often, prey availability, predator-prey encounter rates, and feeding rates are primary
sources of uncertainty (Ney, 1993; Walters and Martell, 2004; Bajer et al., 2004; Chips and Wah,
2008), and are difficult to measure in the field (Walters and Martell, 2004). With snook, these
and estimates of food conversion efficiencies, activity levels, and physiological responses to
abiotic variation (e.g. temperature, salinity, and oxygen interactive effects on physiological
responses) are not well quantified. Therefore, to minimize unpredictability, in this model I used
a simplified approach that bypassed the need for predicting many of the above bioenergetic
variables. Much of this model was based on data from field examinations of snook stomachs
during winter and spring.
While environmental conditions have a strong influence on gastric evacuation rates (e.g.
temperature and salinity, or interactive effects [Vinagre et al., 2007]), the cannibalism model
only focused on field collections during winter and spring and thus reduced these effects. Prey
size (mass) has also been shown to be directly related to residence times in digestive tracts
(Jobling et al., 1977; dos Santos and Jobling, 1995; Andersen, 1998). Kennedy (1969) suggested
that small-sized prey items were often underrepresented in diet studies, due to faster gastric
evacuation rates. For snook, this could strongly bias observations of age-0 cannibalism rates in
132
the field and actual rates are probably under-represented. In a laboratory study, casual visual
examination of differently sized soft-bodied fish prey (snook and Gobiidae) in the stomachs of
snook held in aquaria over time have shown these prey were generally identifiable to species
until about 70% of original prey mass remained (Heinlein and Brennan, unpublished data). In
the same study, gobies weighing 2g (about 50 mm TL) at consumption were approximately 77%
their consumption weight after 4h in the stomach, but 0.5g (about 30 mm TL) gobies were
unidentifiable after 4h (17% - 24% their original weight remained) (Heinlein and Brennan,
unpublished data).
Diet samples are also known to over-represent large-bodied prey items (Mann and Orr,
1969; Gannon, 1976). In the laboratory, 15g snook prey (mean size=122 mm FL) were more
than 93% their original weights 4h after consumption, and only 50% digested by 18 h after
consumption by larger snook (Heinlein and Brennan, unpublished data). These data align with
predictions that larger-bodied prey are comparatively over-represented in diet studies (Mann and
Orr, 1969; Gannon, 1976) and thus reported cannibalism rates of larger prey are probably closer
to actual rates. Snook also demonstrate peak feeding activity during crepuscular and nocturnal
periods (Rock et al., unpublished data), and diet samples collected during daylight often contain
highly digested prey items that are indistinguishable. All studies reported here sampled snook
stomachs during daylight hours. Crepuscular and nocturnal sampling may have revealed higher
rates of cannibalism. Further study could identify the relative influence of these factors on
evacuation rates and cannibalism rates for common snook.
In conclusion, juvenile snook show many morphological adaptations for predation
defense such as elongated anal spines, proportionally deeper body height, and self mimicry in
dorsal fin spot, and also show greater cannibalism potential with proportionally wider mouth
133
gapes in smaller snook. Field enclosure trials indicated that smaller snook had greater
cannibalism tendencies than larger snook, and increased snook densities increased cannibalism
rates. Snook diet studies to date have shown sparse evidence of cannibalism, however,
cannibalism rates are influenced by density-dependent functions, and an interaction between
foraging arenas (nursery habitats versus adult habitats) and seasonal effects on a population’s
movement. All cases of cannibalism recorded have shown that age-0 snook are the target prey
group for intracohort and intercohort cannibalism and this process could have a mediating
influence on recruitment dynamics. This work underlines the necessity for multifaceted annual
recruitment monitoring programs and could contribute to modeling stock enhancement viability
assessments and overall snook management programs.
134
135
Reference study Sample Region Season Habitat description Juvnelie Size range No. of stomach's Stomach's Percent Observed Estimated sample age Known
dates collected habitat? (mm) Sampled w/contents w/ contents cannibals Age-0 Age-1 Age-2+ sizes (num.)
Marshall, 1958 1954-1955 IRL, West Coast, East Coast FL ? Mangrove shorelines No 230-851 FL 128 61 47.7 0 0 0 128 128
Harrington and Harrington, 1961 Indian River Lagoon, FL summer, fall Marshes Yes <100 FL 172 ? ? 0 172 0 0 172
Chavez, 1963 Veracruz, Mexico all seasons Mangrove shroelines Yes 47-341 TL 60 51 85.0 0 ? ? ? 0
Linton and Richards, 1965 1963-1964 Georgia ? estuarine marsh No 24-75 SL 62 24 38.7 0 62 0 0 62
Austin and Austin, 1971 1967-1968 Western Puerto Rico all seasons Mangrove-lined Yes 23-87 TL 38 8 21.0 0 38 0 0 38
canals and river banks
Rojas, 1975 1972-1973 Laguna de Terminos, Campeche, Mexico fall, winter Mangrove wetlands Yes juveniles - mature 327 207 63.3 0 ? ? ? 0
McMichaels et al., 1989 1984-1985 Tampa Bay, FL all seasons Grass flats, beaches, Yes 11-346 SL 480 427 89.0 0 395 ? ? 395
backwaters, rivers
Gillmore et al., 1983 1974-1980 Indian River Lagoon, FL all seasons Freshwater Yes 11-156 SL 99 60 89.0 0 99 0 0 99
Gillmore et al., 1983 1955-1956 Indian River Lagoon, FL all seasons Marshes Yes ca. 20-275 SL 441 ? ? 0 441 0 0 441
Gillmore et al., 1983 1974-1980 Indian River Lagoon, FL all seasons Seagrass Yes 15-600 SL 321 ? ? 0 21 300 0 321
Luskovitch et al., 1995 1985-1986 St. Lucie County, FL summer, fall, winter Mangrove impoundment Yes 5-119 SL 258 ? ? 0 258 0 0 258
Fore and Schmidt, 1973 1971-1973 Everglades, FL summer, fall Tidal creeks, canals Yes 14-196 TL 183 142 77.5 0 183 0 0 183
Fore and Schmidt, 1973 1971-1973 Everglades, FL summer, fall Mangrove shorelines Yes 224-1020 TL 271 127 46.8 0 0 0 271 271
Aliaume et al., 1997 1992-1994 Puerto Rico spring, summer, fall Rivers, freshwater and Yes <300 SL 439 268 61.0 0 91 ? ? 91
saltwater lagoons,
backwaters
Adams and Wolfe, 2006 Feb-Mar 3003 Charolette Harbor, FL winter Estuarine ponds, creeks Yes 141-680 153 94 61.4 3 43 25 42 110
Adams and Wolfe, 2006 Nov 03-Feb 04 Charolette Harbor, FL winter Estuarine ponds, creeks Yes 190-651 127 3 0 150 19 169
Blewett et al., 2005 2000-2002 Charolette Harbor, FL all seasons Coastal mangrove No 300-882 SL 694 432 62.2 0 0 0 694 694
and seagrass
Brennan, unpublished data 2001 Sarasota, FL summer Tidal creeks Yes 100-250 FL 114 90 79.0 0 0 114 0 114
Brennan, unpublished data 2004 Sarasota, FL summer Tidal creeks Yes 100-735 353 162 45.9 0 12 248 93 353
Brennan, unpublished data 2004 Sarasota, FL winter, spring Tidal creeks, canals Yes 80-965 FL 1710 879 51.4 1 789 513 408 1710
Brennan, unpublished data 2005 Sarasota, FL summer Tidal creeks Yes 103-466 FL 338 210 68.1 0 0 258 80 338
Total 6768 3241 61.7 7 2604 1608 1735 5947
Table 4-1. Summary of studies examined regarding common snook diets. ‘Juvenile habitat’ refers to whether samples were collected in juvenile snook habitat. The numbers of individual snook reported from the various age groups are based on body length information.
136
Age group Stomach examination Nov.-Dec. February March April May Total Percentage-0 Number examined 7 160 502 74 96 838 42.13
Number of cannibals 0 1 0 0 0 1age-1 Number examined 135 110 320 50 62 678 34.07
Number of cannibals 2 0 1 0 0 3age-2+ Number examined 19 87 257 61 50 474 23.80
Number of cannibals 3 0 0 0 0 3Total 161 357 1,079 185 208 1,990 100.00
Table 4-2. Numbers of snook processed for stomach contents and observed instances of cannibalism from 2003 and 2004 winter data in nursery habitats (i.e. from Adams and Wolfe, 2006, and Sarasota data, 2004). Totals and percents are also shown according to age group.
Table 4-3. Linear regression statistics for various morphometric comparisons Comparison N slope L95 U95 R2
Ln(VA-AN) * Ln(FL) 31 1.001 0.957 1.045 0.987Ln(SL) * Ln(FL) 94 1.000 0.997 1.002 0.999Ln(BD) * Ln(SL) 94 0.901 0.892 0.915 0.998Ln(VG) * Ln(SL) 135 0.815 0.792 0.831 0.979Ln(AS) * Ln(SL) 332 0.664 0.646 0.682 0.940 Note: vertical gape (VG), body depth (BD, anal spine (AS), throat to anus (TRH-AN), standard length (SL), and fork length (FL). Prior to regression, all values were loge transformed (Ln) Lower (L95) and upper (U95) 95% confidence limits of the slopes are given.
137
Table 4-4. Number of replicate experimental treatments (N) of field enclosure trials examining cannibalism according to predator size, the inclusion of additional prey (bait), snook prey density, and two prey size categories (expressed as a percent of predator size)
Experimental trials NPredator size Bait Prey density Prey sizeLarge Y 1 30 3Large Y 1 50 5Large Y 10 30 4Large Y 10 50 4Subtotal 16Large N 1 30 5Large N 1 50 5Large N 10 30 4Large N 10 50 5Subtotal 19Small Y 1 30 5Small Y 1 50 5Small Y 10 30 4Small Y 10 50 5Subtotal 19Small N 1 30 5Small N 1 50 5Small N 10 30 5Small N 10 50 5Subtotal 20Total 74
138
Table 4-5. Details of actual cannibalism occurrences found in snook stomachs collected in the field.
Source Sample Cannibalistic snook Cannibalized Prey
date length age Length Wt. (g) % pred. lengthAdams and Wolfe 2006 December-03 278 SL age-1 48 0.7 17.3Adams and Wolfe 2006 November-03 313 SL age-1 77 3.4 24.6Adams and Wolfe 2006 March-03 312 SL age-1 113 11.5 36.2Adams and Wolfe 2006 January-03 591 SL age-2+ 117 12.9 19.8Adams and Wolfe 2006 January-03 610 SL age-2+ 121 14.4 19.8Adams and Wolfe 2006 December-03 651 SL age-2+ 141 23.3 21.7Brennan February-04 145 FL age-0 64 1.5 50.0
139
SarasotaCounty
Manatee County
Figure 4-1. Map of Sarasota and Manatee counties where snook collections and stock enhancement research has been performed. Arrows indicate the tidal creeks where snook were captured and subsequently examined for stomach contents. Map of Florida is offset.
140
Figure 4-2. Photograph of enclosures where cannibalism trials occurred. Steel re-bars
supported the material attached at the corners. Bird mesh prevented birds (seen on first enclosure) from preying on the enclosed snook from above.
141
Number of agei snook stomachs examined
Number of cannibalism occurrences
of agei snook
Number of agei snook stomachs examined
Number of cannibalism occurrences
of agei snook
Simulatedproportion of
cannibalized meals
Simulated relativeabundance in January:
(Age-1, Age-2+)
Simulated individual mean length.(FL, mm)
of age0 snook in November
Simulated weight-based instantaneous natural
mortality rate (zi) (Eqn. 4-2)
Population growth function
(Eqn. 4-1)
Number of snookcannibalized by
agei cohort
Proportion of original
age0 cohort cannibalizedby agei snook
Repeat with 10,000independent simulations
Length frequency data and CPUE from standardized
sampling in nursery habitats in January
Creek residence schedule
of agei snook
Mark-recapture data for agei snook in and around
tidal creeks (wild and hatchery-reared recaptures)
(Chapter 2 and 3, and Brennan, unpublished data)
Length frequency data from standardized
sampling in nursery habitats in November.
Marine species mortality rates
(Lorenzen, 1996)
Mean weight of age0 cohort in November
(Eqn. 4-4)
Relative abundance of age-1 or age-2+
Snook in November (Eqn. 4-4)
Abundance of age0cohort in November
(n = 5000)
Proportion of age0 mortality
due to cannibalism
by agei snook
Field dataN
ovember 1 –
May 15
Figure 4-3. Schematic of model process predicting cannibalism-related mortality by various age groups of snook. Shaded boxes represent model components where simulated variability occurred.
142
0
0.5
1
1.5
2
2.5
3
1998 1999 2000 2001 2002 2003 2004 2005
age 0age-1adults
Rel
ativ
e A
bund
ance
Collection Year
Figure 4-4. Annual relative abundance of age-1 and adult snook standardized around age-0
abundance. Data are from annual samples collected in Sarasota during January – March of each year.
143
0.1
0.3
0.5
0.7
0.9
1.1
1.3
0 50 100 150 200 250 300 350 400
VG/SL BD/SL (Throat-AN)/SL FL/SL
0
0.05
0.1
0.15
0.2
0.25
0.3
0 100 200 300 400 500 600 700 800 900 1000Fork length (mm)
AS/
FL
hatchery snookwild snookhatchery snookwild snook
Standard Length (mm)
Fork Length (mm)
Rat
ioA
S/FL
A
B
Figure 4-5. Graphical comparisons of snook morphometrics (in mm). The ratio of anal spine length (AS) to fork length (graph A) declines as a power exponential model with increasing body length. Graph B shows the ratios VG/SL (i.e. VG=vertical gape, SL=standard length), BD/SL, (i.e. BD=body depth), and (throat-AN)/SL (i.e. distance from throat to anus) plotted against standard length.
144
Figure 4-6. Photograph of a 70 mm (TL) and a 178 mm (TL) common snook. Black spot at
tip of dorsal fin is obvious in smaller snook, but spot is nearly absent in larger snook. Photograph by L. Mitchell.
145
00.10.20.3
0.40.50.60.7
S M ALL ( 7 4 - 16 3 mm FL) LA R GE ( 2 2 0 - 3 3 2 mm FL)
1 PREY 10 P REY
Mea
n ca
nnib
alis
m ra
te
Predator size
0
0 .1
0 .2
0 .3
0 .4
0 .5
0 .6
SM ALL ( 74- 163mm FL) LARGE ( 220- 332mm FL)
BAIT NO BAIT
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
S M ALL ( 7 4 - 16 3 mm FL) LA RGE ( 2 2 0 - 3 3 2 mm FL)
30% 50% A
B
C
Figure 4-7. Graphical results from field enclosure trials describing mean cannibalism rates of small and large snook according to various prey treatments: (A) proportional prey size, (B) snook prey density, and (C) additional “bait” effect. Error bars are standard error.
146
Figure 4-8. Plot of body lengths (fork length FL) of snook cannibals and cannibalized prey.
Data originally expressed as standard length were converted to fork length (FL).
0
100
200
300400
500
600
700
0 100 200 300 400 500 600 700
Predator length (FL mm)
Prey
Siz
e (F
L m
m)
Predator length (FL mm)
Prey
leng
th (F
L m
m)
147
0
500
1000
1500
2000
2500
0.00
0.10
0.20
0.30
0.40
0.50
0.60
0.70
0.80
0.90
1.00
0100200300400500600700
0.00
0.10
0.20
0.30
0.40
0.50
0.60
0.70
0.80
0.90
1.00
0
200
400
600
800
0.00
0.10
0.20
0.30
0.40
0.50
0.60
0.70
0.80
0.90
1.00
Mean = 0.189 +/- 0.003
Mean = 0.133 +/- 0.008
Mean = 0.070 +/- 0.002
Age-0 Cannibalism mortality/ Age-0 natural mortality
Freq
uenc
y
age-0 field cannib. Rate 94.760 %Age-0 FL 20.754 %Mu 14.910 %b 8.527 %Age-0 residence 2.838 %Age-1 FL 1.307 %Age-2+ relative abundance 0.868 %Age-2+ FL 0.392 %Age-2 residence 0.309 %Age-1 residence 0.308 %
Model sensitivity
Age-1 field cannib. Rate 88.462 %Age-1 relative abundance 39.981 %Age-0 FL 12.916 %Mu 9.375 %b 4.582 %Age-2+ FL 1.888 %Age-0 residence 1.770 %Age-2+ residence 1.374 %Age-1 FL 0.277 %Age-1 residence 0.258 %
Model sensitivity
Age-2+ field cannib. rate 82.781 %Age-2+ relative abundance 46.233 %Age-0 FL 18.197 %Mu 9.964 %b 6.636 %Age-0 residence 2.457 %Age-2+ FL 1.164 %Age-0 field cannib. rate 0.926 %Age-1 FL 0.634 %Age-1 relative abundance 0.518 %
Model sensitivity
A
B
C
Figure 4-10. Monte-Carlo simulation results (n=10,000 iterations) showing proportions of age-0 mortality due to age-specific cannibalism: (A) age-0 snook, (B) age-1 snook, and (C) adult snook (>= age-2 snook). Results from each sensitivity analysis are inset in each graph. Respective means and associated 95% confidence limits of the mean are included.
148
0100200300400500600700
0.00
0.10
0.20
0.30
0.40
0.50
0.60
0.70
0.80
0.90
1.00
0
200
400600
800
1000
1200
0.00
0.10
0.20
0.30
0.40
0.50
0.60
0.70
0.80
0.90
1.00
0100200300400500600700800
0.00
0.10
0.20
0.30
0.40
0.50
0.60
0.70
0.80
0.90
1.00
Mean = 0.114 +/- 0.002
Mean = 0.174 +/- 0.008
Mean = 0.192 +/- 0.006
Age-0 field cannib. rate 99.305 %Age-0 FL 8.222 %Mu 4.477 %b 3.172 %Age-0 residence 2.460 %Age-2+ residence 1.352 %Age-1 FL 1.332 %Age-2+ relative abundance 0.874 %Age-2 field cannib. Rate 0.532 %Age-1 relative abundance 0.518 %
Model sensitivity
Number cannibalized/ original Age-0 abundance
Freq
uenc
y
A
Age-1 field cannib. rate 99.964 %Age-1 residence 2.100%Age-0 residence 1.798 %Age-2 residence 1.731 %Age-0 FL 1.629 %Mu 1.199 %Age-2 field cannib. rate 1.189 %Age-2 FL 0.836 %b 0.796 %Age-1 FL 0.694 %
Model sensitivity
Age-2 field cannib. rate 99.949 %Age-2+ residence 2.753 %b 1.818 %Age-2+ FL 1.284 %Mu 1.118 %Age-0 field cannib. rate 0.644 %Age-0 residence 0.629 %Age-0 FL 0.582 %Age-1 relative abundance 0.361 %Age-1 residence 0.162 %
Model sensitivity
B
C
Figure 4-11. Monte-Carlo simulation results (n=10,000 iterations) showing proportions of the
original age-0 abundance (at the beginning of the model period) cannibalized by the various age groups of snook: (A) age-0 snook, (B) age-1 snook, and (C) adult snook (>= age-2 snook). Respective means and associated 95% confidence limits of the mean are included.
149
CHAPTER 5 CONCLUSION
In this dissertation I used stock enhancement to understand, test, and reveal mechanisms
behind juvenile snook life history, stock dynamics, and important ecological and environmental
interactions. The practice of stocking fish for management purposes has occurred for centuries,
yet only within the past 20 years has clear application of a scientific approach occurred (Leber,
1999). With this approach, much progress has occurred including understanding behavioral
differences between hatchery and wild conspecifics (Olla and Davis, 1989; Hossain et al., 2002),
managing fish health and disease transmission from hatchery to wild stocks (Kennedy et al.,
1998; Mushiaki and Muroga, 2004), tagging and marking of released fish (Blankenship, 1990;
Bravington et al., 2004; Davis et al., 2004), release techniques and stocking strategies
(Tsukamoto, 1989; Yamashita et al., 1994; Leber, 1995; Brennan et al., 2006; Fairchild, 2008),
and documenting recruitment to fishery landings (Svåsand et al., 1990; Kitada et al., 1999;
Kaeriyama, 1999; Zohar et al., 2008).
Nonetheless, considerable controversy remains in this field largely due to unresolved
critical uncertainties (Leber et al., 2004; Lorenzen, 2005). Because stakeholders are increasingly
promoting the use of stock enhancement technology as a fishery management tool, (Walters and
Martell, 2004; Hilborn, 2004; Leber et al., 2004; Lorenzen, 2008) the need for understanding
these uncertainties are becoming increasingly important. Evaluating ecosystem wide and intra-
and inter-cohort density-dependent responses to stocking is a top priority (Leber et al., 2004;
Walters and Martel, 2004). Understanding genetic consequences of hatchery-reared individuals
interbreeding with wild stocks is also a high priority (Bell et al., 2008).
150
In this dissertation, I coupled the release of hatchery-reared juvenile snook with rigorous
follow up evaluation, to identify fundamental ecological interactions, and important components
of snook’s life history (Leber, 1999; Miller and Walters, 2004); especially during juvenile stages.
Having thousands of organisms tagged and experimentally situated in situ is difficult to
accomplish with wild animals (e.g. Leber et al., 1995). Post-release responses of hatchery-reared
snook stocked into various habitats allowed me to make quantitative correlations with measured
habitat characteristics. It also allowed me to make comparisons between post-release recapture
ratios and growth rates over time and show how patterns established in rearing habitats can have
long-term consequences on production to adult stocks. Monitoring large numbers of tagged
stocked fish over time also allowed me to identify when and where ontogenetic habitat shifts
were occurring (Chapter 2 and 3), and examine fidelity rates in relation to rearing habitats for
years after the releases (Chapter 2). I was able to predict numerical and biomass contributions of
the released cohorts to adults stages (Chapter 2). In Chapter 3, I addressed the critical issue of
density dependent survival and used variations in release magnitudes to examine this potential in
post-release survival within a juvenile cohort of wild and hatchery-reared snook.
In Chapter 2, I examined how stocking various rearing habitats influenced post-release
responses in the hatchery fish, and coupled with other studies (Chapter 3), I was able to project
numerical and biomass contributions of hatchery snook to adult stocks. I identified key
measures of snook microhabitat, and identified tradeoffs in production to adult stages based on
these features. I found a consistent inverse relationship between growth rates and recapture
ratios. These responses were correlated to release habitats with high juvenile snook densities,
soft sediments with high organic content, and waters of low salinity, and hypoxic conditions.
Release habitats with higher salinity, firmer sediment, and higher dissolved oxygen, and low
151
ambient densities of age-1 snook were correlated with high post-release growth rates, but low
recapture ratios. Over time, sites with low growth rates, but high juvenile densities showed the
highest numerical contributions of hatchery reared snook to adults stocks, but biomass
contributions were among the lowest. Snook stocked into rearing habitats with intermediate
growth rates had the highest contributions to adult biomass however.
Others have presented theoretical implications for numerical contributions of nursery
habitats to adult stocks (e.g. Beck et al., 2001; Dahlgren et al., 2006), but field studies
demonstrating this are rare. Differences in numerical and biomass production could be
substantial for species like snook that attain large body sizes, are highly fecund, and have
variable growth and survival rates. Using this work, resource managers can assess the relative
production value (in terms of snook) according to characteristics of rearing habitats. Habitats
along estuarine shorelines were the focus of this study, but additional work including
representative habitats of a production system is important to evaluating how habitats contribute
to stocks on a broader scale. Quantification of total habitat area is also an important next step to
evaluate overall production according to habitat type.
In Chapter 3 (Brennan et al., 2008), I examined potential density-dependent mortality of
hatchery-reared juvenile snook with manipulations in stocking magnitude. I varied release
magnitudes among four tidal creeks to evaluate the productive capacities of the habitats and
determine potential mortality effects on wild and hatchery-reared conspecifics. Creeks receiving
high augmentation treatments showed overall increases in post-release densities, whereas no
overall changes were detected in age-1 snook densities in low augmentation creeks. Loss of
hatchery-reared snook was remarkably high in one creek (BC) that received a high augmentation
treatment, but loss of wild conspecifics was not detected. This is consistent with earlier studies
152
(Chapter 2) where hatchery-reared juveniles, stocked into BC, had lower recapture ratios
compared to snook released at other sites. In BC, the overall increase in post-release density,
accompanied by high loss rates of hatchery-reared snook suggests the productive capacity may
have been exceeded, but alternatively may be due to density independent losses, or that it was
not conducive to survival of hatchery fish.
Work in Chapter 3 examined an important assumption about recruitment processes – that
density-dependent survival of wild fish may be reduced by the addition of hatchery-reared fish.
In Salmonid populations, evidence exists for density-dependent responses in wild stocks due to
stocking hatchery-reared fish (Nickelson et al., 1986; Nickelson, 2003; Kostow and Zhou, 2006).
Density-dependent growth responses were not measured in this study, but other work in the same
creeks (Chapter 2), showed suppressed growth in tributaries with high densities of age-1 snook.
I found no evidence of high augmentation treatment effects on wild conspecific density, although
an effect may have been expressed through initial loss of hatchery-reared fish in BC where the
productive capacity may have been substantially exceeded. Abundance patterns of wild age-1
snook throughout the study, however, followed similar patterns in all creeks.
Clearly post-release survival and growth are, in part, functions of ambient density of snook
juveniles in the rearing habitats, refuge quality, stocking magnitude, predator and prey
abundance, encounter rates, and feeding rates. Release magnitudes in 1998 and 1999 (Chapter 2)
(hundreds to approximately a thousand, by release habitat) were similar to stocking magnitudes
in the 2002 study (Brennan et al., 2008; Chapter 3), although acclimation techniques were not
used in 1998 and 1999 (which adds other potentially important logistical complications,
(Brennan et al., 2006; Fairchild et al., 2008). Two creeks receiving high augmentation levels in
the 2002 study showed very different responses in hatchery snook survival (in one creek 0.85 of
153
the stocked snook remained by first winter, but in the other creek only 0.15-0.36 remained by
first winter). Negligible losses of wild conspecifics were observed in all stocked creeks
however. These results were also positively related to ambient density. Sites with high ambient
density of wild juveniles also had high post-release survival rates of hatchery juveniles, and sites
with low ambient densities resulted in low survival rates of the stocked fish. This suggests that
while growth responses may have been density-dependent, stocking of age-1 snook (on this order
of magnitude) may not have resulted in density-dependent mortality within the cohort, but might
be controlled by density-independent effects such as rearing habitat quality or refuge availability.
Stocking at higher magnitudes may have very different results and a cautionary approach is
warranted. Inter-annual variation in recruitment of snook, its competitors, prey, or predators
could all influence the dynamics of density-dependent responses (Chapter 4 examines the
potential influence of intercohort cannibalism, see below). Aside from further potential
influence of variability in hatchery-based components (such as variation in health of the stocked
snook, or complications with release logistics [Congleton et al., 2000; Sulikowski et al., 2006]),
stocking the same habitats and release magnitudes could still cause very different results, just
due to year-to-year effects. Nonetheless, increases in experience with this work (e.g. increases in
the number of experimental years, additional stocking levels, and across habitats) can increase
confidence in our expectations of such activities, however. Coupling this with experimental
modeling scenarios of stock enhancement (see below), can further add to our understanding of
these issues.
An important consideration for stocking piscivorous and cannibalistic fishes, such as
snook, is the potential influence of intra- and inter-cohort cannibalism and competition as I
examined in Chapter 4. With snook, this potential is further intensified because, as in northern
154
latitudes such as Sarasota, age-0 and older snook share nursery habitat for a prolonged period,
Increases in abundance of age-1 snook, due to stocking or naturally, may have negative
consequences on young-of-year snook populations through cannibalism and competition for
refuge and food resources. Release programs that elevate age-1 abundance could intensify such
effects and a cautionary approach is warranted. Nonetheless, the potential for elevating stocks
that are below capacity is demonstrated by this study and its potential over the long-term must be
evaluated. In chapter 4, I first examined phenotypic characteristics among snook of different
sizes (39-903 mm FL) to find evidence for variation in cannibalism potential across snook sizes.
Smaller snook had proportionally larger mouth gapes (able to consume proportionally larger
prey), were deeper bodied and had proportionally longer anal spines. These morphological
features make predation more difficult, and are likely a reflection of higher predation threat
among smaller-sized individuals. I also found that snook less than 125 mm FL exhibited a
conspicuous dorsal fin spot, but was absent in snook larger than 175 mm FL. I suggested that
this spot may be a form of self-mimicry, an adaptation to confuse predators, causing them to
strike at an apparent “eye spot” in a peripheral, less-essential, body location. Ontogenetic loss of
this spot may reflect decreases in predation threat with larger body sizes.
I also used field enclosure trials to measure relative cannibalism rates among differently
sized snook. Manipulating predator size, relative prey size, prey density, and the presence of
additional prey options, I found higher (although not statistically significant) mean rates of
cannibalism in small snook than larger snook,. Small snook had higher cannibalism rates on
proportionally larger prey compared to large snook predators, the presence of additional prey
species reduced mean cannibalism rates, but again, means were not significantly different.
155
Increased replication may have reduced variation in the study and results should be verified with
examinations of stomach contents and the use of controls.
As a further consideration of snook cannibalism potential, I also examined published and
unpublished data to measure rates of cannibalism detected in wild snook of different life stages.
All cases of cannibalism found were directed at young-of-year snook. These rates were used in a
model predicting cannibalism potential in a localized snook stock in a tidal creek habitat through
a winter. As prey items in a fish’s stomach generally represents a single day’s or night’s worth
of feeding activity, the collective examination of snook stomach contents, showed the sample
sizes of most studies (hundreds) were grossly insufficient for detecting cannibalism instances in
snook (where thousands needed). Also, because cannibalism is generally directed at very small
snook, it would be extremely difficult to detect this in the field because of the high evacuation
rates of small prey. Field based examination of stomach contents are generally biased towards
detecting large prey items, but negatively biased against detecting small prey items (Kennedy,
1969. Ignoring this, and just examining observed cannibalism rates in the field, my model
showed that a population of juveniles consuming just a few conspecifics throughout a season can
collectively and significantly influence a stock size, however. My model (Chapter 4) predicted
that collective cannibalism by the age groups could contribute substantially to the age-0 natural
mortality rates and cannibalism rates were density-dependent. High cannibalism rates were
predicted when strong year classes of age-0 snook were produced, but low during years of low
recruitment of age-0 snook. This work has strong implications for stock enhancement programs
seeking to numerically supplement particular cohorts of snook.
Rigorous experimental monitoring from standardized sampling efforts over time is integral
to understanding basic progress of the stocked organisms and hatchery-wild fish interactions.
156
Work in this dissertation has highlighted the importance of this. Long-term sampling programs,
especially in juvenile habitats, should be a top priority for management agencies tasked with
performing stock assessments and understanding basic biology and ecology of the target
organism while keeping a broad focus on the ecosystem as a whole (Walters and Martell, 2004).
Our ongoing standardized sampling program has revealed important findings on juvenile snook
feeding ecology (Rock et al., unpublished data; Brennan and Heinlein, unpublished data). Multi-
year sampling programs (10 years and longer) directed at multiple life stages are extremely
powerful for understanding inter-annual recruitment and recruitment across geographical areas.
This dissertation revealed how strong differences in inter-annual recruitment could result in
dramatic differences in density-dependent mortality (i.e. cannibalism) and serves as an example
of this. In other work, I identified strong within-year differences in recruitment strength in
habitats only kilometers away, during years when red tide blooms were especially prevalent in
northern sites of our study area. Multi-year sampling programs allow for adaptive
management, a necessary approach in the face of environmental and social variation. Investing
this towards stocks subjected to high harvest pressures and mortality is certainly warranted.
A clear next step will be modeling the cost-effectiveness and social benefits involving
stakeholders and using various enhancement scenarios (Lorenzen, 2006; Lorenzen, 2008) of a
snook stock enhancement management program in Florida. Work in this dissertation provides
essential input into such models. Whereas snook aquaculture technology has not yet been
developed enough to support large-scale enhancement, modeling the costs, benefits, and
alternative management approaches should precede any large-scale application of stock
enhancement (Hilborn, 1998; Walters and Martell, 2004; Lorenzen, 2008). As snook stocks are
fairly localized and non-migratory (Tringali et al., in press[a]), the results of the stock
157
enhancement work in Sarasota, are useful for providing necessary input into modeling
enhancement scenarios and feasibility in Florida.
158
159
LIST OF REFERENCES
Adams, A.J., Wolfe, R.K., 2006. Cannibalism of juveniles by adult Centropomus undecimalis. Gulf of Mex. Sci. 24 (1-2), 11-13.
Adams. A.J., Wolfe, R.K., Pine, W.E., Thornton, B.L., 2006. Efficacy of PIT tags and an
autonomous antenna system to study juvenile life stage of an estuarine-dependent fish. Estuaries Coasts 29 (2), 311-317.
Aliaume C, Zerbi, A., Miller, J.M., 1997. Nursery habitat and diet of juvenile Centropomus
species in Puerto Rico estuaries. Gulf of Mex. Sci. 15 (2), 77-87. American Sport Fishing Association, 2004. Sport fishing in America: Values of our traditional
pastime. Available: www.asafishing.org/asa/ Andersen, N.G., 1998. The effect of meal size on gastric evacuation in whiting. J. Fish. Biol. 2
(4), 743-755. Andersen, A.K., Schou, J., Sparrebohn, C.R., Nicolajsen, H., Stottrup, J.G., 2005. The quality of
release habitat for reared juvenile flounder, Platichthys flesus, with respect to salinity and depth. Fish. Manage. Ecol. 12 (3), 211-219.
Andraso, G.M., 1997. A comparison of startle response in two morphs of the brook stickleback
(Culaea inconstans): Further evidence for a trade-off between defensive morphology and swimming ability. Evol. Ecol. 11 (1), 83-90.
Armstrong, J.D., Braithwaite, V.A., Huntingford, F.A., 1997. Spatial strategies of wild Atlantic
salmon parr: exploration and settlement in unfamiliar areas. J.Anim. Ecol. 66, 203-211. Austin, H., Austin, S., 1971. The feeding habits of some juvenile marine fishes from the
mangroves in western Puerto Rico. Carib. J. Sci., 11(3-4), 171-178. Bailey, K.M., 1989. Interaction between the vertical distribution of juvenile walleye Pollock,
Theragra chalcogramma, in the eastern Bering Sea and cannibalism. Mar. Ecol. Prog. Ser. 53, 205-213.
Bajer, P.G., Whitledge, G.W., Hayward, R.S., 2004. Widespread consumption-dependent
systematic error in fish bioenergetics models and its implications. Can. J. Fish. Aquat. Sci. 61 (11), 2158-2167.
Baras, E., Kestemont, P., Melard, C., 2003. Effect of stocking density on the dynamics of
cannibalism in sibling larvae of Perca fluviatilis under controlled conditions. Aquacult. 219 (1-4), 241-255.
Beck, M.W., Heck, K.L. Jr., Able, K.W., Childers, D.L., Eggleston, D.B., Gillanders, B.M.,
Halpern, B, Hays, C.G., Hoshino, K., Minello, T.J., Orth, R.J., Sheridan, P.F., Weinstein,
M.P., 2001. The identification, conservation, and management of estuarine and marine nurseries for fish and invertebrates. Bioscience 51 (8), 633-641.
Becker, P., Barringer, C., Marelli, D.C., 2008. Thirty years of sea ranching Manila clams
(Venerupis philippinarum): Successful techniques and lessons learned. Rev. Fish. Sci. 16 (1-3), 44-50.
Belk, M.C., 1995. Variation in growth and age at maturity in bluegill sunfish - genetic or
environmental-effects. J. Fish Biol. 47 (2), 237-247. Belk, M.C., 1998. Predator-induced delayed maturity in bluegill sunfish (Lepomis macrochirus):
Variation among populations. Oecologia 113 (2), 203-209. Bell, J.D., Bartley, D.M., Lorenzen, K., Loneragan, N.R., 2006. Restocking and stock
enhancement of coastal fisheries: potential, problems and progress. Fish Res. 80 (1), 1-8. Bell, J.D., Leber, K.M., Blankenship, H.L., Loneragan, N.R., Masuda, R., 2008. A new era for
restocking, stock enhancement and sea ranching of coastal fisheries resources. Rev. Fish. Sci. 16 (1-3), 1-9.
Bennett, J.P., 2006. Using telemetry to assess natural and fishing mortality of common snook in
Sarasota Bay, Florida. Masters thesis. Fisheries and Aquatic Sciences, University of Florida.
Berejikian, B.A., Smith, R.J.F., Tezak, E.P., Schroder, S.L., Knudsen, C.M., 1999. Chemical
alarm signals and complex hatchery rearing habitats affect antipredator behavior and survival of chinook salmon (Oncorhynchus tshawytscha) juveniles. Can. J. Fish. Aquat. Sci. 56 (5), 830-838.
Berkeley, S.A., Chapman, C., Sogard, S.M., 2004. Maternal age as a determinant of larval
growth and survival in a marine fish, Sebastes melanops. Ecology 85, 1258-1264. Blankenship, H.L., 1990. Effects of time and fish size on coded wire tag loss from Chinook and
Coho salmon. In: Parker, N.C., Giorgi, A.E., Heidinger, R.C., Jester Jr., D.B., Prince, E.D., Winans, G.A. (Eds.), Fish marking techniques. American Fisheries Society Symposium 7. Bethesda, MD, pp. 237-243.
Blankenship, H.L., Leber, K.M., 1995. A Responsible Approach to Marine Stock Enhancement.
In: Schramm Jr., H.L., Piper, R.G. (Eds.), Uses and effects of cultured fishes in aquatic ecosystems. American Fisheries Society Symposium 15. Bethesda, MD, pp.376-387.
Blest, A.D., 1957. The function of eyespot patterns in the Lipidotera. Behaviour 11, 209–256. Blewett, D.A., Hensley, R.A., Stevens, P.W., 2006. Feeding habits of common snook,
Centropomus undecimalis, in Charlotte Harbor, Florida. Gulf Carib. Res. 18, 1-13.
160
Bohlin, T., Sundstrom, L.F., Johnsson, J.I., Hojesjo, J., Pettersson, J., 2002. Density-dependent
growth in brown trout: effects of introducing wild and hatchery fish. J. Anim. Ecol., 71 (4), 683-692.
Botsford, L.W., Castilla, J.C., Peterson, C.H., 1997. The management of fisheries and marine
ecosystems. Science 277 (5325), 509-515. Bowen, S.H., Allanson, B.R., 1982. Behavioral and trophic plasticity of juvenile Tilapia
mossamica in utilization of the unstable littoral habitat. Env. Biol. Fish. 7, 357-362. Bravington, M.V., Ward, R.D., 2004. Microsatellite DNA markers: evaluating their potential for
estimating the proportion of hatchery-reared offspring in a stock enhancement programme. Mol. Ecol. 13 (5), 1287-1297.
Brennan, N.P., 1997. Establishment and management of a multi-state, coded-wire tag,
paddlefish database. Masters thesis. Cooperative Fisheries Research Unit. Tennessee Technological University, Cookeville, Tennessee. 38505.
Brennan, N.P., Leber, K.M., Blankenship, H.L., Ransier, J.M., DeBruler Jr., R., 2005. An
evaluation of coded wire and elastomer tag performance in juvenile common snook under field and laboratory conditions. N. Am. J. Fish. Manage. 25, 437-445.
Brennan, N.P., Darcy, M.C., Leber, K.M., 2006. Predator-free enclosures improve post-release
survival of stocked common snook. J. Exp. Mar. Biol. Ecol. 335 (2), 302-311. Brennan, N.P., Leber, K.M., Blackburn, B.R., 2007. Use of coded-wire and visible implant
elastomer tags for marine stock enhancement with juvenile red snapper Lutjanus campechanus. Fish. Res. 83 (1), 90-97.
Brennan, N.P., Walters, C.J., Leber, K.M., 2008. Manipulations of stocking magnitude:
addressing density-dependence in a juvenile cohort of common snook Centropomus undecimalis. Rev. Fish. Sci. 16 (1-3), 215-227.
Brooke, M.D., 1998. Ecological factors influencing the occurrence of ‘flash marks’ in wading
birds. Funct. Ecol., 12 (3), 339-346. Bruger, G.E., 1980. Preliminary analyses of snook, Centropomus undecimalis, population
dynamics in the Naples-Marco Island region of southwest Florida. Fla. Sci. 43 (suppl.1), 25.
Bruger, G.E., Haddad, K.D., 1986. Management of tarpon, bonefish, and snook in Florida.
Stroud, R.H. (Ed.). Multi-jurisdictional management of marine fishes. National Coalition for Marine Conservation, Inc., Savannah, Georgia, 53-57.
161
Burke, B.J., Rice, J.A., 2002. A linked foraging and bioenergetics model for southern flounder. Trans. Am. Fish. Soc. 131 (1), 120-131.
Byström, P., Persson, L., Wahlstrom, E., Westman, E., 2003. Size- and density-dependent
habitat use in predators: consequences for habitat shifts in young fish. J. Anim. Ecol. 72 (1), 156-168.
Byström, P., Andersson, J., Persson, L., De Roos, A.M., 2004. Size-dependent resource
limitation and foraging-predation risk trade-offs: growth and habitat use in young arctic char. Oikos 104, 109-121.
Byström, P., 2006. Recruitment pulses induce cannibalistic giants in Arctic char. J. Anim. Ecol.
75 (2), 434-444. Campbell, P., 1997. Population projections: States 1995-2025. U.S. Census Bureau.
http://www.census.gov/prod/2/pop/p25/p25-1131.pdf. Carline, R.F., 2006. Regulation of an unexploited brown trout population in Spruce creek,
Pennsylvania. Trans. Am. Fish. Soc., 135 (4), 943-954. Carr, S.D., Hench, J.L., Luettich, R.A., Forward, R.B., Tankersly, R.A., 2005. Spatial patterns
in the ovigerous Callinectes sapidus spawning migration: results from a coupled behavioral-physical model. Mar. Ecol. Prog. Ser. 294, 213-226.
Carscadden, J.E., Frank, K.T., Leggett, W.C., 2001. Ecosystem changes and the effects on
capelin (Mallotus villosus), a major forage species. Can. J. Fish. Aquat. Sci. 58 (1), 73-85.
Caselle, J.E., Warner, R.R., 1996. Variability in recruitment of coral reef fishes: The importance
of habitat at two spatial scales. Ecology 77 (8), 2488-2504. Chambers, R.C., Leggett, W.C., 1996. Maternal influences on variation in egg sizes in temperate
marine fishes. Am. Zool. 36, 180-196. Chang, E. Y. Y., Leo, I.C., 2003. Sibling cannibalism of young red drum, Sciaenops ocellatus,
in relation to size disparity and metabolic rates. Env. Biol. Fish. 68 (4), 407-415. Chavez, H., 1963. Contribucion al conocimiento de la biologia de los robalos, chucumite y
constantino (Centropomus spp.) del estado de Veracruz (pisc., Centrop.). Ciencia, Mex. 22, 141-161.
Chen, Y., Jackson, D.A., Harvey, H.H., 1992. A comparison of von Bertalanffy and polynomial
functions in modeling fish growth data. Can. J. Fish. Aquat. Sci. 49, 1228-1235. Chipps, S.R., Wahl, D.H., 2008. Bioenergetics modeling in the 21st century: Reviewing new
insights and revisiting old constraints. Trans. Am. Fish. Soc. 137 (1), 298-313.
162
Claessen, D., de Roos, A.M., Persson, L., 2000. Dwarfs and giants: Cannibalism and
competition in size-structured populations. Am. Nat. 155 (2), 219-237. Claessen, D., de Roos, A.M., Persson, L., 2004. Population dynamic theory of size-dependent
cannibalism. Proc. Roy. Soc. Lond. Ser. B: Biol. Sci. 271 (1537), 333-340. Clarke, C.W., Levy, D.A., 1988. Diel vertical migrations by juvenile sockeye salmon and the
antipredation window. Am. Nat. 131, 271 – 290. Conder, A.L., Annear, T.C., 1987. Test of weighted usable area estimates derived from a
PHABSIM model for instream flow studies on trout streams. N. Am. J. Fish. Manage 7, 339-350.
Congleton, J.L., LaVoie, W.J., Schreck, C.B., Davis, L.E., 2000. Stress indices in migrating
juvenile Chinook salmon and steelhead of wild and hatchery origin before and after barge transportation. Trans. Am. Fish. Soc. 129 (4), 946-961.
Connell, S.D., Jones, G.P., 1991. The influence of habitat complexity on post-recruitment
processes in a temperate reef fish population. J. Exp. Mar. Biol. Ecol., 151 (2), 271-294. Correa, C.F., Cerqueira, V.R., 2007. Effects of stocking density and size distribution on growth,
survival and cannibalism in juvenile fat snook (Centropomus parallelus Poey). Aquacult. Res. 38 (15), 1627-1634.
Dahl, J., Pettersson, E., Dannewitz, J., Jarvi, T., Lof, A.C., 2006. No difference in survival,
growth and morphology between offspring of wild-born, hatchery and hybrid brown trout (Salmo trutta). Ecol. Freshw. Fish 15 (4), 388-397.
Dahlgren, C.P., Kellison, G.T., Adams, A.J., Gillanders, B.M., Kendall, M.S., Layman, C.A.,
Ley, J.A., Nagelkerken, I., Serafy, J.E., 2006. Marine nurseries and effective juvenile habitats: concepts and applications. Mar. Ecol. Prog. Ser. 312, 291-295.
Davis, J.L.D., Young-Williams, A.C., Aguilar, R., Carswell, B.L., Goodison, M.R., Hines, A.H.,
Kramer, M.A., Zohar, Y., Zmora, O., 2004. Differences between hatchery-raised and wild blue crabs: implications for stock enhancement potential. Trans. Am. Fish. Soc. 133 (1), 1-16.
Davis, J.L.D., Young-Williams, A.C., Hines, A.H., Zohar, Y., 2005. Assessing the potential for
stock enhancement in the case of the Chesapeake Bay blue crab (Callinectes sapidus). Can. J. Fish. Aquat. Sci. 62 (1), 109-122.
Davis, S.M., Childers, D.L., Lorenz, J.J., Wanless, H.R., Hopkins, T.E., 2005. A conceptual
model of ecological interactions in the mangrove estuaries of the Florida Everglades. Wetlands, 25 (4), 832-842.
163
De Innocentiis, S., Longobardi, A., Marino, G., 2008. Molecular tools in a marine restocking program for the endangered dusky grouper, Epinephelus marginatus. Rev. Fish. Sci. 16 (1-3), 269-277.
Doherty, P. J., and A. Fowler. 1994. An empirical test of recruitment limitation in a coral reef
fish. Science 263, 935-939. Doherty, P.J., Dufour, V., Galzin, R., Hixon, M.A., Meekan, M.G., Planes, S., 2004. High
mortality during settlement is a population bottleneck for a tropical surgeonfish. Ecol. 85 (9), 2422-2428.
Dorsey, L.G., 2004. Retention of coded wire tags by age-0 Muskellunge. J. Freshwater Ecol. 19
(2), 333-337. dos Santos, J., Jobling, M., 1991. Factors affecting gastric evacuation in Cod, Gadus morhua-L.,
fed single-meals of natural prey. J. Fish Biol. 38, 697-713. Duffy, J.T., Epifanio, C.E., Fuiman, L.A., 1997. Mortality rates imposed by three scyphozoans
on red drum (Sciaenops ocellatus Linnaeus) larvae in field enclosures. J. Exp. Mar. Biol. Ecol. 212 (1), 123-131.
Dumont, H.J., A. Jawahar, A., 2004. Stage-specific cannibalism and spontaneous cyst hatching
in the freshwater fairy shrimp Streptocephalus proboscideus Frauenfeld. Hydrobiologia, 524, 103-113.
Einum, S., Sundt-Hansen, L., Nislow, K.H., 2006. The partitioning of density-dependent
dispersal, growth and survival throughout ontogeny in a highly fecund organism. Oikos 113 (3), 489-496.
Fairchild, E.A., Rennels, N., Howell, W.H., 2008. Predators are attracted to acclimation cages
used for winter flounder stock enhancement. Rev. Fish. Sci. 16 (1-3), 262-268. FAO, 2004. The state of world fisheries and aquaculture: 2004. FAO, Rome (also available at
ftp://ftp.fao.org/docrep/fao/007/y5600e/y5600r00.pdf) Faunce, C.H., Serafy, J.E., 2006. Mangroves as fish habitat: 50 years of field studies. Mar.
Ecol.-Prog. Ser. 318, 1-18. Fleming, I.A., Petersson, E., 2001. The ability of released, hatchery salmonids to breed and
contribute to the natural productivity of wild populations. Nordic J. Freshw. Res. 75 (5), 71-98.
Fore, P.L. Schmidt, T.W., 1973. Biology of juvenile and adult snook, Centropomus undecimalis,
in the Ten Thousand Islands, Florida. U.S. Environmental Protection Agency, Surveillance and Analysis Division. Publication number EPA 904/9-74-002, Chapter 16. Region IV, Atlanta, GA.
164
Francis, R.C., Hixon, M.A., Clarke, M.E., Murawski, S.A., Ralston, S.A., 2007. Ten
Commandments for ecosystem-based fisheries scientists. Fisheries 32 (5), 217-233. Frankiewicz, P., Dabrowski, K., Martyniak, A., Zalewski, M., 1999. Cannibalism as a regulatory
force of pikeperch, Stizostedion lucioperca (L.), population dynamics in the lowland Sulejow reservoir (Central Poland). Hydrobiologia 408, 47-55.
Fretwell, S.D., 1972. Populations in a Seasonal Environment. Princeton University Press,
Princeton, New Jersey. Fritts, A.L., Scott, J.L., Pearsons, T.N., 2007. The effects of domestication on the relative
vulnerability of hatchery and wild origin spring Chinook salmon (Oncorhynchus tshawytscha) to predation. Can. J. Fish. Aquat. Sci. 64 (5), 813-818.
Fromentin, J.M., Gjosaeter, J., Bjornstad, O.N., Stenseth, N.C., 2000. Biological processes and
environmental factors regulating the dynamics of the Norwegian Skagerrak cod populations since 1919. ICES J. Mar. Sci. 57 (2), 330-338.
Furuta, S., 1996. Predation on juvenile Japanese flounder (Paralichthys olivaceus) by diurnal
piscivorous fish: field observations and laboratory experiments. In Watanabe, Y., Yamashita, Y., Oozeki, Y. (Eds.), Survival Strategies in Early Life Stages of Marine Resources. A. A. Balkema, Rotterdam, pp. 285-294.
Gallego, A., Heath, M.R., 1994. The development of schooling behavior in Atlantic herring
Clupea harengus. J. Fish Biol. 45 (4), 569-588. Gannon, J.E., 1976. The effects of differential digestion rates of zooplankton by alewife, Alosa
pseudoharengus, on determinations of selective feeding. Trans. Am. Fish. Soc., 105, 89-95.
Garaway, C.J., Arthur, R.I., Chamsingh, B., Homekingkeo, P., Lorenzen, K., Saengvilaikham,
B., Sidavong, K., 2006. A social science perspective on stock enhancement outcomes: lessons learned from inland fisheries in southern Lao PDR. Fish. Res. 80, 37- 45.
Gard, M.F., 2005. Ontogenetic microhabitat shifts in Sacramento pikeminnow, Ptychocheilus
grandis: reducing intraspecific predation. Aquat. Ecol., 39 (2), 229-235. Garrison, L.P., Link, J.S., 2000. Diets of five hake species in the northeast United States
continental shelf ecosystem. Mar. Ecol. Prog. Ser. 204, 243-255. Gibson, R.N., Pihl, L., Burrows, M.T., Modin, J., Wennhage, H., Nickell, L.A., 1998. Diel
movements of juvenile plaice Pleuronectes platessa in relation to predators, competitors, food availability and abiotic factors on a microtidal nursery ground. Mar. Ecol. Prog. Ser. 165, 145-159.
165
Gilmore, R.G., Bullock, L.H., F. H. Berry. 1978. Hypothermal mortality in marine fishes on south-central Florida. NE Gulf Sci. 2 (2), 77-97.
Gilmore, R.G., Donohoe, C.J., Cooke, D.W., 1983. Observations on the distribution and biology
of east-central Florida populations of the common snook, Centropomus undecimalis (Bloch). Fl. Sci. 46, 313-336.
Grant, S.M., Brown, J.A., 1998. Nearshore settlement and localized populations of Atlantic cod
(Gadus morhua) in shallow coastal waters of Newfoundland. Can. J. Fish. Aquat. Sci. 55 (6), 1317-1327.
Gwak W.S., Tanaka, Y., Tominaga, O., Tsusaki, T., Tanaka, M., 2003. Field evaluation by
RNA/DNA ratios on post-release nutritional status of released and wild Japanese flounder Paralichthys olivaceus juveniles. J. Exp. Mar. Biol. Ecol. 293 (1), 107-124.
Hailman, T.P., 1977. Optical signals: animal communication and light. Indiana University
Press, Bloomington. Hallegraeff, G.M., 1993. A review of harmful algal blooms and their apparent global increase.
Phycologia 32 (2), 79-99. Harrell, A., Gibb, A.C., 2006. Ontogeny of performance in vertebrates. Phys. Biochem. Zoo.,
79 (1), 1-6. Helminen, H., Sarvala, J., 1994. Population regulation of vendace (Coregonus albula) in lake
Pyhajarvi, southwest Finland. J. Fish. Biol., 45 (3), 387-400. Hamilton, L.C., Haedrich, R.H., 1999. Ecological and population changes in fishing
communities of the North Atlantic. Arc. Polar Res., 18 (2), 383-388. Hampton, S.E., 2004. Habitat overlap of enemies: temporal patterns and the role of spatial
complexity. Oecologia 138 (3), 475-484. Harrington, R.W., Harrington, E.S., 1961. Food selection among fishes invading a high sub-
tropical marsh: from onset of flooding through the progress of a mosquito brood. Ecology 42, 646-666.
Henderson, P.A., Corps, M., 1997. The role of temperature and cannibalism in interannual
recruitment variation of bass in British waters. J. Fish. Biol., 50 (2), 280-295. Hilborn, R., 1998. The economic performance of marine stock enhancement projects. Bull. Mar.
Sci. 62 (2), 661-674. Hilborn, R., 2004. Population management in stock enhancement and sea ranching. In: Leber,
K.M., Kitada, S., Blankenship, H.L., Svsand, T. (Eds.), Stock enhancement and sea ranching. Blackwell Publishing, Oxford, UK, pp. 201–209.
166
Hilborn, R, Walters, C.J., 1992. Quantitative Fisheries Stock Assessment: Choice, Dynamics
and Uncertainty. Chapman and Hall, New York, NY. Honma, A., 1993. Aquaculture in Japan. Japan FAO Association, Tokyo. Hossain MAR, Tanaka M, Masuda R., 2002. Predator-prey interaction between hatchery-reared
Japanese flounder juvenile, Paralichthys olivaceus, and sandy shore crab, Matuta lunaris: daily rhythms, anti-predator conditioning and starvation. J. Exp. Mar. Biol. Ecol 267 (1), 1-14.
Houde, E.D., 1987. Fish early life dynamics and recruitment variability. In: Houde, E.D., (Ed),
10th Annual Larval Fish Conference. American Fisheries Society, Symposium 2, Bethesda, MA, pp. 17-29.
Hurst, T.P., Conover, D.O., 1998. Winter mortality of young-of-the-year Hudson River striped
bass (Morone saxatilis): size-dependent patterns and effects on recruitment. Can. J. Fish. Aquat. Sci. 55, 2709-2709.
Hutchings, J.A., 2000. Collapse and recovery of marine fishes. Nature 406, 882-885. Immamura, K., 1999. The organization and development of sea farming in Japan. In Howell,
B.R., Moksness, E., Svåsand, T., (Eds.), Stock Enhancement and Sea Ranching (Ed. B.R. Howell, E. Moksness, and T.), Blackwell Scientific Publications, Oxford, UK, pp. 91-101.
Islam, M.S., Haque, M., 2004. The mangrove-based coastal and nearshore fisheries of
Bangladesh: ecology, exploitation and management. Rev. Fish Biol. Fish. 14 (2), 153-180.
Jackson, J.B.C., Kirby, M.X., Berger, W.H., Bjorndal, K.A., Botsford, L.W., Bourque, B.J.,
Bradbury, R.H., Cooke, R., Erlandson, J., Estes, J.A., Hughes, T.P., Kidwell, S., Lange, C.B., Lenihan, H.S., Pandolfi, J.M., Peterson, C.H., Steneck, R.S., Tegner, M.J., Warner, R.R., 2001. Historical overfishing and the recent collapse of coastal ecosystems. Science 293 (5530), 629-638.
Jakobsson, A., Eriksson, O., 2000. A comparative study of seed number, seed size, seedling size
and recruitment in grassland plants. Oikos 88 (3), 494-502. Jefferts, K.B., Bergman, P.K., Fiscus, H.F., 1963. A coded wire identification system for macro-
organisms. Nature 198, 460-462. Jenkins, W.E., Denson, M.R., Bridgham, C.B., Collins, M.R., Smith, T.I.J., 2002. Retention of
oxytetracycline-induced marks on Sagittae of red drum. N. Am. J. Fish. Manage. 22 (2), 590-594.
167
Jobling, M., Gwyther, D., Grove, D.J., 1977. Some effects of temperature, meal size and body weight on gastric evacuation time in the dab Limanda limanda (L). J. Fish Biol. 10, 291-298.
Jokikokko, E., Kallio-Nyberg, I., Saloniemi, I., Jutila, E., 2006. The survival of semi-wild, wild
and hatchery-reared Atlantic salmon smolts of the Simojoki River in the Baltic Sea. J. Fish Biol. 68 (2), 430-442.
Johnson, J.H., Dropkin, D.S., 1993. Diel variation in diet composition of a riverine fish
community. Hydrobiologia 271, 149-158. Jonsson, N., Jonsson, B., Hansen, L.P., 2003. The marine survival and growth of wild and
hatchery-reared Atlantic salmon. J. App. Ecol. 40 (5), 900-911. Jones, G.P., 1997. Relationships between recruitment and postrecruitment processes in lagoonal
populations of two coral reef fishes. J. Exp. Mar. Biol. Ecol. 213 (2), 231-246. Jorstad, K., 2004. Genetic studies in marine stock enhancement in Norway. In: Leber, K.M.,
Kitada, S., Blankenship, H.L., Svsand, T. (Eds.), Stock enhancement and sea ranching. Blackwell Publishing, Oxford, UK, pp. 339-352.
Juanes, F., 2003. Allometry of cannibalism in piscivorous fishes. Can. J. Fish. Aquat. Sci. 60,
594-602. Kaeriyama, M., 1999. Hatchery programmes and stock management of Salmonid populations in
Japan. In Howell, B.R., Moksness, E., Svåsand, T., (Eds.), Stock Enhancement and Sea Ranching (Ed. B.R. Howell, E. Moksness, and T.), Blackwell Scientific Publications, Oxford, UK, pp. 153-167.
Keeley, E.R., 2001. Demographic responses to food and space competition by juvenile steelhead
trout. Ecol. 82 (5), 1247-1259. Kellison, G.T., Eggleston, D.B., Tanaka, M., 2002. Density-dependent predation and
implications for stock enhancement with Japanese flounder. J. Fish. Biol., 60 (4), 968-980.
Kellison, G.T., Eggleston, D. B., 2004. Coupling ecology and economy: modeling optimal
release scenarios for summer flounder (Paralichthys dentatus) stock enhancement. Fish. Bull. 102 (1), 78-93.
Kemp, W.M, Sampou, P.A., Garber, J., Tuttle, J., Boynton, W.R., 1992. Seasonal depletion of
oxygen from bottom waters of Chesapeake Bay - roles of benthic and planktonic respiration and physical exchange processes. Mar. Ecol. Prog. Ser. 85 (1-2), 137-152.
Kennedy, C.R. 1969. Tubificid oligochaetes as food of dace. J. Fish Biol. 1, 11-15.
168
Kennedy, T.B., Vinyard, G.L., 1997. Drift ecology of western catostomid larvae with emphasis on Warner suckers, Catostomus warnerensis (Teleostei). Env. Biol. Fish. 49 (2), 187-195.
Kennedy, S.B., Tucker, J.W., Neidig, C.L., Vermeer, G.K., Cooper, V.R., Jarrell, J.L., Sennett,
D.G., 1998. Bacterial management strategies for stock enhancement of warmwater marine fish: a case study with common snook (Centropomus undecimalis). Bull. Mar. Sci. 62 (2), 573-588.
Kitada, S., 1999. Effectiveness of Japan’s stock enhancement programmes: current perspectives.
In Howell, B.R., Moksness, E., Svåsand, T., (Eds.), Stock Enhancement and Sea Ranching (Ed. B.R. Howell, E. Moksness, and T.), Blackwell Scientific Publications, Oxford, UK, pp. 103-131.
Kitchell, J.F., Stewart, D.J., Weininger, D., 1977. Applications of a bioenergetics model to
yellow perch (Perca flavescens) and walleye (Stizostedion vitreum vitreum). J. Fish. Res. Board Can. 34: 1922-1934.
Kocovsky, P.M., Carline, R.F., 2001. Dynamics of the unexploited walleye population of
Pymatuning Sanctuary, Pennsylvania, 1997-1998. N. Am. J. Fish. Manage. 21 (1), 178-187.
Koczaja, C., McCall, L., Fitch, E., Glorioso, B., Hanna, C., Kyzar, J., Niemiller, M., Spiess, J.,
Tolley, A., Wyckoff, R., Mullen, D., 2005. Size-specific habitat segregation and intraspecific interactions in banded sculpin (Cottus carolinae). Southeast Nat. 4 (2), 207-218.
Kostow, K.E., Zhou, S.J., 2006. The effect of an introduced summer steelhead hatchery stock on
the productivity of a wild winter steelhead population. Trans. Am. Fish. Soc. 35 (3), 825-841.
Krause, J., Loader, S.P., McDermott, J., Ruxton, G.D., 1998. Refuge use by fish as a function of
body length-related metabolic expenditure and predation risks. Proc. Roy. Soc. Lond. Ser. B: Biol. Sci. 265 (1413), 2373-2379.
Kristiansen T.S., Svåsand, T., 1990. Enhancement studies of coastal cod in western Norway.
Part III. Interrelationships between reared and indigenous cod in a nearly land-locked fjord. J. Con. Int. Explor. Mer 47, 23-29.
Krivan ,V., Vrkoc, I., 2000. Patch choice under predation hazard. Theor. Popul. Biol. 58 (4),
329-340. Landeau, L., Terborgh, J., 1986. Oddity and the ‘confusion effect’ in predation. Anim. Behav.,
34, 1372-1380.
169
Lang, G.M., Brodeur, R.D., Napp, J.M., Schabetsberger, R., 2000. Variation in groundfish predation on juvenile walleye pollock relative to hydrographic structure near the Pribilof Islands, Alaska. ICES J. Mar. Sci. 57 (2), 265-271.
Lau, S.R., Shafland, P.L., 1982. Larval development of snook, Centropomus undecimalis
(Bloch) (Pices: Centropomidae). Copeia 1982, 618-627. Leber, K.M., 1995. Significance of fish size at release on enhancement of striped mullet fisheries
in Hawaii. J. World Aquacult. Soc. 26, 143 153. Leber, K.M., Brennan, N.P., Arce, S.M., 1998. Recruitment patterns of cultured juvenile Pacific
threadfin, Polydactylus sexfilis (Polynemidae), released along sandy marine shores in Hawaii. Bull. Mar. Sci. 62 (2), 389-408.
Leber, K.M., 1999. Rationale for an experimental approach to stock enhancement. In Howell,
B.R., Moksness, E., Svåsand, T., (Eds.), Stock Enhancement and Sea Ranching (Ed. B.R. Howell, E. Moksness, and T.), Blackwell Scientific Publications, Oxford, UK, pp. 63-75.
Leber, K.M., 2004. Marine Stock Enhancement in the USA: Status, trends and needs. In: Leber,
K.M., Kitada, S., Blankenship, H.L., Svsand, T. (Eds.), Stock enhancement and sea ranching. Blackwell Publishing, Oxford, UK, pp. 11-24.
Leber, K.M., Brennan, N.P., Arce, S.M., 1995. Marine enhancement with striped mullet: are
hatchery releases replenishing or displacing wild stocks? In: Schramm Jr., H.L., Piper, R.G. (Eds.), Uses and effects of cultured fishes in aquatic ecosystems. American Fisheries Society Symposium 15. Bethesda, MD, pp.376-387.
Leber, K.M., Arce, S.M., Sterritt, D.A., Brennan, N.P., 1996. Marine stock enhancement
potential in nursery habitats of striped mullet, Mugil cephalus, in Hawaii. Fish. Bull. 94, 452-471.
Leber, K.M, Blankenship, H.L., Arce, S.M., Brennan, N.P., 1997. The influence of release
season on size-dependent survival of striped mullet, Mugil cephalus, in a Hawaiian estuary. Fish. Bull. 95 (2), 267-279.
Leber, K.M., Kitada, S., Blankenship, L., Svåsand, T., 2004. Stock Enhancement and Sea
Ranching: Developments, Pitfalls and Opportunities. Blackwell Publishing, Oxford, United Kingdom.
Leber, K.M., Cantrell, R.N., Leung, P.S., 2005. Optimizing cost-effectiveness of size at release
in stock enhancement programs. N. Am. J. Fish. Manage. 25, 1596-1608. Lenore, C., Greenberg, A., Eaton, A., 1998. Standard methods for the examination of water and
wastewater. 1998. Section 2540. American Public Health Association, American Waterworks Association, Water Environment Federation. United Book Press, Inc., Baltimore, Maryland.
170
Liao, H.S., Pierce, C.L., Larscheid, J.G., 2004. Consumption dynamics of the adult piscivorous
fish community in Spirit Lake, Iowa. North American Journal of Fisheries Management 24 (3), 890-902.
Linehan, J.E., Gregory, R.S., Schneider, D.C., 2001. Predation risk of age-0 cod (Gadus)
relative to depth and substrate in coastal waters. J. Exp. Mar. Biol. Ecol. 263, 25-44. Linton, T. L., Richards, W. L. 1965. Young common snook on the coast of Gerogia. Quart. J.
Florida Acad. Sci. 28 (2), 185-189. Lipcius, R.N., Vanengel, W.A., 1990. Blue-crab population-dynamics in Chesapeake Bay –
variation in abundance (York River, 1972-1988) and stock-recruitment functions. Bull. Mar. Sci. 46 (1), 180-194.
Lochet, A., Lambert, P., Lepage, M., Rochard, E., 2004. Growth comparison between wild and
hatchery-reared juvenile European sturgeons Acipenser sturio (Acipenseridae) during their stay in the Gironde estuary (France). CYBIUM 28 (1), 91-98 Suppl. S.
Longhurst, A., 2006. The sustainability myth. Fish. Res. 81, 107-112. Lotze, H.K., Lenihan, H.S., Bourque, B.J., Bradbury, R.H., Cooke, R.G., Kay, M.C., Kidwell,
S.M., Kirby, M.X., Peterson, C.H., Jackson, J.B.C., 2006. Depletion, degradation, and recovery potential of estuaries and coastal seas. Science 312 (5781), 1806-1809.
Lorenz, J.J., 1999. The response of fishes to physicochemical changes in the mangroves of
northeast Florida Bay. Estuaries 22 (2B), 500-517. Lorenz, J.J., Sarafy, J.E., 2006. Subtroprical wetland fish assemblages and changing salinity
regimes: Implications for everglades restoration. Hydrobiologia 569, 401-422 Lorenzen, K., 1996. The relationship between body weight and natural mortality in fish: a
comparison of natural ecosystems and aquaculture. J. Fish Biol. 49, 627-647. Lorenzen, K., 2000. Allometry of natural mortality as a basis for assessing optimal release size
in fish-stocking programmes. Can. J. Fish. Aquat. Sci. 57, 2374-2381. Lorenzen, K., 2005. Population dynamics and potential of fisheries stock enhancement: practical
theory for assessment and policy analysis. Philos. T. Roy. Soc. B. 260, 171-189. Lorenzen, K., 2006. Population management in fisheries enhancement: gaining key information
from release experiments through use of a size-dependent mortality model. Fish. Res. 80, 19-27.
Lorenzen, K., 2008. Understanding and managing fisheries enhancement systems. Rev. Fish.
Sci. 16 (1-3), 10-23.
171
Lorenzen, K, Enberg, K., 2002. Density-dependent growth as a key mechanism in the regulation
of fish populations: evidence from among-population comparisons. Proc. Roy. Soc. Lond. Ser. B: Biol. Sci. 269 (1486), 49-54.
Love, M. S., Carr, M.H., Haldorson, L.I., 1991. The ecology of substrate-associated juveniles of
the genus Sebastes. Env. Biol. Fish. 30 (1-2), 225-243. Luczkovich, J. J., Norton, S.F., Gilmore Jr., R.G., 1995. The influence of oral anatomy on prey
selection during the ontogeny of two percoid fishes, Lagodon rhomboids and Centropomus undecimalis. Env. Biol. Fish. 44, 79-95.
MacPherson, T.A., Cahoon, L.B., Mallin, M.A., 2007. Water column oxygen demand and
sediment oxygen flux: patterns of oxygen depletion in tidal creeks. Hydrobiologia 586, 235-248.
Madon, S.P., Williams, G.D., West, J.M., Zedler, J.B., 2001. The importance of marsh access to
growth of the California killifish, Fundulus parvipinnis, evaluated through bioenergetics modeling. Ecol. Model. 136 (2-3), 149-165.
Magnhagen, C., Heibo, E., 2004. Growth in length and in body depth in young-of-the-year perch
with different predation risk. J. Fish Biol. 64 (3), 612-624. Major, P.F., 1978. Predator-prey interactions in two schooling fishes, Caranx ignobilis and
Stolephorus purpureus. Anim. Behav., 27, 855-865. Mann, R.H. K., and D.R. Orr. 1969. A preliminary study of the feeding relationships of fish in a
hard-water and a soft-water stream in southern England. J. Fish Biol., 1, 31-44. Marcinkiewicz, L., 2007. Examining seasonal movement and habitat use patterns of adult
common snook. Masters thesis. Fisheries and Aquatic Sciences, University of Florida. Marshall, A.N., 1958. A survey of the snook fishery of Florida, with studies of the biology of
the principal species, Centropomus undecimalis (Bloch). Florida Board of Conservation, Marine Research Laboratory Technical Series No. 22.
Mason, J.W., Brynilson, O.M., Degurse, P.E., 1967. Comparative survival of wild and domestic
strains of brook trout in streams. Trans. Am. Fish. Soc. 96, 313-319. Masuda, R., 2004. Behavioral approaches to fish stock enhancement: a practical review. In:
Leber, K.M., Kitada, S., Blankenship, H.L., Svsand, T. (Eds.), Stock enhancement and sea ranching. Blackwell Publishing, Oxford, UK, pp. 83-90.
Mather, J.A., Odor, R.K., 1991. Foraging strategies and predation risk shape the natural-history
of juvenile octopus-vulgaris. Bull. Mar. Sci. 49 (1-2), 256-269.
172
Maynard, D.J., Flagg, T.A., Mahnken, C.V.W., Schroder, S.L. , 1996a. Natural rearing technologies for increasing post-release survival of hatchery-reared salmon. Bulletin of National Research Institute, Supp. 2, 71-77.
Maynard, D.J, McDowell, G.C., Tezak, E.P., Flagg, T.A., 1996b. Effects of diets supplemented
with live food on the foraging behavior of cultured fall chinook salmon. Prog. Fish Cult. 58, 187-191.
McMichael Jr., R.H., Peters, K., Parsons, G.R., 1989. Early life history of the snook
Centropomus undecimalis in Tampa Bay, Florida. Northeast Gulf Sci. 10, 113-125. Meeuwig, M.H., Puls, A.L., Bayer, J.M., 2007. Survival and tag retention of Pacific lamprey
larvae and macrophthalmia marked with coded-wire tags. N. Am. J. Fish. Manage. 27 (1), 96-102.
Metcalfe, N.B., Fraser, N.H.C., Burns, M.D., 1998. State dependent shifts between nocturnal
and diurnal activity in salmon. Proc. Roy. Soc. Lond. Ser. B: Biol. Sci. 265 (1405), 1503-1507.
Metcalfe, N.B., Steele, G.I., 2001. Changing nutritional status causes a shift in the balance of
nocturnal to diurnal activity in European Minnows. Funct. Ecol. 15 (3), 304-309 Milinski, M., Curio, E., 1975. Experiments on the eco-ethology of predator-prey relations..
Verh. Dtsch. Zool. Ges. no. 68, pp. 81-89. Miller, J.M., Walters, C.J., 2004. Experimental ecological tests with stocked marine fish. In:
Leber, K.M., Kitada, S., Blankenship, H.L., Svsand, T. (Eds.), Stock enhancement and sea ranching. Blackwell Publishing, Oxford, UK, pp. 142-152.
Miranda, L.E., Hubbard, W.D., 1994. Length-dependent winter survival and lipid-composition
of age-0 largemouth bass in bay springs reservoir, Mississippi. Trans. Am. Fish. Soc. 123 (1), 80-87.
Mittelbach, G. G. 1986. Predator-mediated habitat use: some consequences for species
interactions. Env. Biol. Fish. 16, 159 – 169. Mol, J.H., 1996. Impact of predation on early stages of the armoured catfish Hoplosternum
thoracatum (Siluriformes-Callichthyidae) and implications for the syntopic occurrence with other related catfishes in a neotropical multi-predator swamp. Oecologia 107 (3), 395-410.
Mondor, E.B., Roitberg, B.D., 2002. Pea aphid, Acyrthosiphon pisum, cornicle ontogeny as an
adaptation to differential predation risk. Can. J. Zoo.-Rev. Can. Zoo. 80 (12), 2131-2136.
173
Mooij, W.M., Van Densen, W.L.T., Lammens, E.H.R.R., 1996. Formation of year-class strength in the bream population in the shallow eutrophic Lake Tjeukemeer. J. Fish Biol. 48 (1), 30-39.
Muller, R.G., Taylor, R.G., 2005. The 2005 stock assessment update of common snook,
Centropomus undecimalis. Florida Fish and Wildlife Conservation Commission, Florida Marine Research Institute. Available: www.floridamarine.org
Mumby, P.J., Edwards, A.J., Arias-Gonzalez, J.E., Lindeman, K.C., Blackwell, P.G., Gall, A.,
Gorczynska, M.I., Harborne, A.R., Pescod, C.L., Renken, H., Wabnitz, C.C.C., Llewellyn, G., 2004. Mangroves enhance the biomass of coral reef fish communities in the Caribbean. Nature 427 (6974), 533-536.
Munro, J.L., Bell, J.D., 1997. Enhancement of marine fisheries resources. Rev. Fish. Sci. 5, 185-
222. Mushiake, K., Muroga, K., 2004. Fish health management in seed production. In: Leber, K.M.,
Kitada, S., Blankenship, H.L., Svsand, T. (Eds.), Stock enhancement and sea ranching. Blackwell Publishing, Oxford, UK, pp. 119-129.
Myers, R.A., 2001. Stock and recruitment: generalizations about maximum reproductive rate,
density dependence, and variability using meta-analytic approaches. ICES J. Mar. Sci. 58 (5), 937-951.
Neill, S.R.St. J, Cullen, J.M., 1974. Experiments on whether schooling by their prey affects the
hunting behavior of cephalopods and fish predators. J. Zool. Lond. 172, 549-569. Nemeth, R.S., 1998. The effect of natural variation in substrate architecture on the survival of
juvenile bicolor damselfish. Env. Biol. Fish. 53 (2), 129-141. Neuenfeldt, S., Koster, F.W., 2000. Trophodynamic control on recruitment success in Baltic
cod: the influence of cannibalism. ICES J. Mar. Sci. 57, 300-309. Ney, J.J., 1993. Bioenergetics modeling today - growing pains on the cutting edge. Trans. Am.
Fish. Soc. 122(5), 736-748. Nickelson, T., 2003. The influence of hatchery coho salmon (Oncorhynchus kisutch) on the
productivity of wild coho salmon populations in Oregon coastal basins. Can. J. Fish. Aquat. Sci. 60 (9), 1050-1056.
Nickelson, T.E., Solazzi, M.F., Johnson, S.I., 1986. Use of hatchery coho salmon
(Oncorhynchus kisutch) presmolts to rebuild wild populations in Oregon coastal streams. Can. J. Fish. Aquat. Sci. 43 (12), 2443-2449.
Norcross, B.L., 1984. Oceanic and estuarine transport of fish eggs and larvae - a review. Trans.
Am. Fish. Soc. 113 (2), 153-165.
174
Nordwall, F., Naslund, I., Degerman, E., 2001. Intercohort competition effects on survival,
movement, and growth of brown trout (Salmo trutta) in Sweedish streams. Can. J. Fish. Aquat. Sci., 58 (11), 2298-2308.
Olla, B.L., Davis, M.W., 1989. The role of learning and stress in predator avoidance of
hatchery-reared coho salmon (Onchorhynchus kisutch) smolts in seawater. Aquacult. 76, 209-214.
Olla, B.L., Davis, M.W., Ryer, C.H., 1998. Understanding how the hatchery environment
represses or promotes the development of behavioral survival skills. Bull. Mar. Sci. 62, 531-550.
Olson, M.H., Mittelbach, G.G., Osenberg, C.W., 1995. Competition between predator and prey:
Resource-based mechanisms and implications for stage-structured dynamics. Ecology 76 (6), 1758-1771.
Orth, D.J., Maughan, O.E., 1982. Evaluation of the incremental methodology for recommending
instream flows for fishes. Trans. Am. Fish. Soc. 111 (4), 413-445. Palma, A.T., Steneck, R.S., 2001. Does variable coloration in juvenile marine crabs reduce risk
of visual predation? Ecology 82 (10), 2961-2967. Pauly, D., Christensen, V., Dalsgaard, J., Froese, R., Torres Jr., F., 1998. Fishing down marine
food webs. Science 279 (6), 860-863. Pauly, D., Christensen, V., Walters, C., 2000. Ecopath, Ecosim, and Ecospace as tools for
evaluating ecosystem impact of fisheries. ICES J. Mar. Sci. 57 (3), 697-706. Pérez-Pinzón, M.A, Lutz, P.L., 1991. Activity related cost of osmoregulation in the juvenile
snook (Centropomus undecimalis). Bull. Mar. Sci. 48 (1), 56 – 66. Pergent-Martini C, Pergent G., 2000. Marine phanerogams as a tool in the evaluation of marine
trace-metal contamination: an example from the Mediterranean. Int. J. Env. Poll. 13 (1-6), 126-147.
Perry, R.I., Walters, C.J., Boutillier, J.A., 1999. A framework for providing scientific advice for
the management of new and developing invertebrate fisheries. Rev. Fish. Biol. Fisher. 9 (2), 125-150.
Persson. L., 1993. Predator-mediated competition in prey refuges - the importance of habitat
dependent prey resources. Oikos 68 (1), 12-22. Persson, L, Eklov, P., 1995. Prey refuges affecting interactions between piscivorous perch and
juvenile perch and roach. Ecology 76 (1), 70-81.
175
Persson, L., Bystrom, P., Wahlstrom, E., 2000. Cannibalism and competition in Eurasian perch: population dynamics of an ontogenetic omnivore. Ecol 81 (4), 1058-1071.
Persson, L., de Roos, A.M., 2006. Size-structured interactions and the dynamics of aquatic
systems. Pol. J. Ecol. 54 (4), 621-632. Persson, L., de Roos, A.M., Byström, P., 2007. State-dependent invasion windows for prey in
size-structured predator-prey systems: whole lake experiments. J. Anim. Ecol. 76 (1), 94-104.
Peters, K.M., Matheson, Jr., R.E., Taylor, R.G., 1998a. Reproduction and early life history of
common snook, Centropomus undecimalis (Bloch), in Florida. Bull. Mar. Sci. 62 (2), 509-529.
Peters, K.M., Giordano, S.D., Adams, J.M., Fenwick, R.W., McWilliams, R.W., Berill, N.J.,
1998b. Factors influencing the distribution of larvae of the common snook, Centropomus undecimalis (Bloch), in Southwestern Florida. Fla. Mar. Res. Publ., 53, 33 pp.
Peterson, M.S., Gilmore Jr., R.G., 1991. Eco-physiology of juvenile snook Centropomus
undecimalis (Bloch): life history implications. Bull. Mar. Sci. 48 (1), 46 - 57. Pillar, S.C., Wilkinson, I.S., 1995. The diet of Cape hake Merluccius capensis on the south coast
of South Africa. S. Afr. J. Mar. Sci. 15, 225-239. Powell, R.A., 1982. Evolution of black-tipped tails in weasels: predator confusion. Am. Nat.
119, 126-131. Prentice, E.F., Flagg, T.A., McCutcheon, C.S., 1990. Feasibility of using implantable passive
integrated transponder (PIT) tags in salmonids. In: Parker, N.C., Giorgi, A.E., Heidinger, R.C., Jester Jr., D.B., Prince, E.D., Winans, G.A. (Eds.), Fish marking techniques. American Fisheries Society Symposium 7. Bethesda, MD, pp.317-322.
Railsback, S.F., Stauffer, H.B., Harvey, B.C., 2005. What can habitat preference models tell us?
Tests using a virtual trout population. Ecol. App. 13 (6), 1580-1594. Raimondi, P.T., 1990. Patterns, mechanisms, and consequences of variability in settlement and
recruitment on an intertidal barnacle. Ecol. Monog. 60, 283–309. Rice, J.A., Cochran, P.A., 1984. Independent evaluation of a bioenergetics model for
largemouth bass. Ecol. 65 (3), 732-739. Rose, K.A., Cowan, J.H., Clark, M.E., Houde, E.D., Wang, S.B., 1999. An individual-based
model of bay anchovy population dynamics in the mesohaline region of Chesapeake Bay. Mar. Ecol. Prog. Ser. 185, 113-132.
Rindorf, A., 2004. Estimation of evacuation rates in the field. J. Fish. Biol. 65, 261-284.
176
Robbins, R.K., 1980. The lycaenid “false head” hypothesis: historical review and quantitative
analysis. J. Lepid. Soc. 34, 194-208. Rojas, J.C., 1975. Contribucion al conocimiento de la biologia de los robalos Centropomus
undecimalis y C. Poeyi en la Laguna de Terminos, Campeche, Mexico. Biol. Inst. Oceanogr. Univ. Oriente 14 (1), 51-70.
Ross, S.W., 2003. The relative value of different estuarine nursery areas in North Carolina for
transient juvenile marine fishes. Fish. Bull. 101 (2), 384 - 404. Roussel, J.M., Bardonnet, A., 1999. Ontogeny of diel pattern of stream-margin habitat use by
emerging brown trout, Salmo trutta, in experimental channels: influence of food and predator presence. Env. Biol. Fish. 56 (1-2), 253-262.
Rozas, L.P., Odum, W.E., 1988. Occupation of submerged aquatic vegetation by fishes: testing
the roles of food and refuge. Oecologia 77, 101-106. Russell, D.J., Rimmer, M.A., McDougall, A.J., Kistle, S.E., Johnston, W.L., 2004. Stock
enhancement of barramundi, Lates calcarifer (Bloch), in a coastal river system in Northern Australia: stocking strategies, survival and benefit-cost. In: Leber, K.M., Kitada, S., Blankenship, H.L., Svsand, T. (Eds.), Stock enhancement and sea ranching. Blackwell Publishing, Oxford, UK, pp. 490 – 500.
Ruzycki, J.R., Wurtsbaugh, W.A., 1999. Ontogenetic habitat shifts of juvenile Bear Lake
sculpin. Trans. Am. Fish. Soc. 128 (6), 1201-1212. Sanchez-Lamadrid, A., 2002. Stock enhancement of gilthead sea bream (Sparus aurata, L.):
assessment of season, fish size and place of release in SW Spanish coast. Aquaculture 210 (1-4), 187-202.
Sanderson, B.L., Hrabik, T.R., Magnuson, J.J., Post, D.M., 1999. Cyclic dynamics of a yellow
perch (Perca flavescens) population in an oligotrophic lake: evidence for the role of intraspecific interactions. Can. J. Fish. Aquat. Sci. 56 (9), 1534-1542.
SAS Institute, 1998. SAS/Stat Users Guide. Version 6: volumes 1 and 2. Cary, North Carolina. Savino, J.F., Stein, R.A., 1982. Predator-prey interaction between largemouth bass and bluegills
as influenced by simulated, submersed vegetation. Trans. Am. Fish. Soc. 111 (3), 255-266.
Sazima, I., 2002. Juvenile snooks (Centropomidae) as mimics of mojarras (Gerreidae), with a
review of aggressive mimicry in fishes. Env. Biol. Fish. 65, 37-45.
177
Seitz, R.D., Lipcius, R.N., Seebo, M.S., 2005. Food availability and growth of the blue crab in seagrass and unvegetated nurseries of Chesapeake Bay. J. Exp. Mar. Biol. Ecol. 319 (1-2), 57-68 Sp. Iss. SI
Sekino, M., Saitoh, K., Yamada, T., Yamashita, Y., .2005. Genetic tagging of released Japanese
flounder (Paralichthys olivaceus) based on polymorphic DNA markers. Aquaculture 244(1-4), 49-61.
Selch, T.M., Chipps, S.R., 2007. The cost of capturing prey: measuring largemouth bass
(Micropterus salmoides) foraging activity using glycolytic enzymes (lactate dehydrogenase). Can. J. Fish. Aquat. Sci. 64 (12), 1761-1769.
Serfling, S.A., 1998. Breeding and culture of snook, Centropomus undecimalis, in a closed-
cycle, controlled environmental culture system. World Aquaculture Society. Aquaculture ‘98 Book of Abstracts. P. 482. February 15-19, 1998. Las Vegas, USA.
Shacklee, J.B., Bentzen, P., 1998. Genetic identification of stocks of marine fish and shellfish.
Bull. Mar. Sci. 62, 589 - 621. Shafland, P.L, Foote, K.J.,1983. A lower lethal temperature for fingerling snook, Centropomus
undecimalis. Northeast Gulf Sci. 6 (2),175-177. Shimizu, D., Sakiyama, K., Sakakura, Y., Takatani, T., Takahashi, Y., 2007. Predation
differences between wild and hatchery-reared tiger puffer Takifugu rubripes juveniles in a salt pond mesocosm. Nippon Suisan Gakkaishi 73 (3), 461-469.
Sillett, K.B., Foster, S.A., 2000. Ontogenetic niche shifts in two populations of juvenile
threespine stickleback, Gasterosteus aculeatus, that differ in pelvic spine morphology. Oikos 91 (3), 468-476.
Smith, P.E., 1995. Development of the population biology of the Pacific hake, Merluccius
productus. CA Coop. Ocean. Fish. Invest., 36, 144-152. Smith, C.C., Fretwell, S.D., 1974. Optimal balance between size and number of offspring. Am.
Nat. 108 (962), 499-506. Sogard, S.M. 1992. Variability in growth-rates of juvenile fishes in different estuarine habitats.
Mar. Ecol. Prog. Ser. 85 (1-2), 35 - 53. Solazzi, M.F., Nickelson, T.E., Johnson, S.L., 1991. Survival, contribution, and return of
hatchery coho salmon (Oncorhynchus kisutch) released into freshwater, estuarine, and marine environments. Can. J. Fish. Aquat. Sci. 48, 248-253.
Steele, D. H., Andersen, R., Green, J.M., 1992. The managed commercial annihilation of
northern cod. Newf. Stud., 8 (1), 34-68.
178
Steneck, R.S., 2006. Is the american lobster, Homarus americanus, overfished? A review of overfishing with an ecologically based perspective. Bull. Mar. Sci 78 (3), 607-632.
Stottrup, J.G., Sparrevohn, C.R., Modin, J., Lehmann, K., 2002. The use of releases of reared
fish to enhance natural populations - A case study on turbot Psetta maxima (Linne, 1758). Fish. Res. 59 (1-2), 161-180.
Story, M., Gruder, E.W., 1936. Mortality of fishes due to cold at Sanibel Island, Florida.
Ecology 18(1), 10-26. Sulikowski, J.A., Fairchild, E.A., Rennels, N., Howell, W.H., Tsang, P.C.W., 2006. The effects
of transport density on cortisol levels in juvenile winter flounder, Pseudopleuronectes americanus. J. World Aqua. Soc. 37 (1), 107-112.
Svåsand, T., 1998. Cod Enhancement Studies in Norway - Background and Results with
Emphasis on Releases in the Period 1983-1990. Bull. Mar. Sci. 62,313-324. Svåsand, T., Kristiansen, T.S., 1990. Enhancement studies of coastal cod in western Norway,
part IV. Mortality of reared cod after release. J. Con. Cons. Int. Explor. Mer 47, 30-39. Svåsand, T., Kristiansen T.S., Pedersen, T, Salvanes, A.G., Engelsen, V.R., Nævdal, G.,
Nødtvedt, M., 2000. The enhancement of cod stocks. Fish and Fish. 1, 173 - 205. Taylor, D.L., Danila, D.J., 2005. Predation on winter flounder (Pseudopleuronectes americanus)
eggs by the sand shrimp (Crangon septemspinosa). Can. J. Fish. Aquat. Sci. 62 (7), 1611-1625.
Taylor, R.G, Grier, H.J., Whittington, J.A., 1998. Spawning rhythms of common snook in
Florida. J. Fish Biol. 53 (3), 502-520. Taylor, R.G., Whittington, J.A., Grier, H.J., Crabtree, R.E., 2000. Age, growth, maturation, and
protandric sex reversal in common snook, Centropomus undecimalis, from the east and west coasts of South Florida. Fish. Bull. 98 (3), 612-624.
Taylor, M.D., Palmer, P.J., Fielder, D.S., Suthers, I.M., 2005. Responsible estuarine finfish
stock enhancement: an Australian perspective. J. Fish. Biol. 67 (2), 299-331 Teknomo, K., 2006. Bootstrap Sampling Tutorial. http://people.revoledu.com/kard/
tutorial/bootstrap/ Titelman J., Hansson L.J., 2006. Feeding rates of the jellyfish Aurelia aurita on fish larvae.
Mar. Biol.149 (2), 297-306. Tolley, S.G., Dohner, E.T., Peebles, E.B., 1987. Occurrence of larval snook, Centropomus
undecimalis (Bloch), in Naples Bay, Florida. Florida. Sci. 50 (1), 34-38.
179
Tomiyama, T., Watanabe, M., Fujita, T., 2008. Community-based stock enhancement and fisheries management of the Japanese flounder in Fukushima, Japan. Rev. Fish. Sci. 16 (1-3) 146-153.
Travis, J., Coleman, F., Grimes, C., Conover, D., Bert, T., Tringali, M., 1998. Critically
assessing stock enhancement: an introduction to the Mote symposium. Bull. Mar. Sci. 62 (2), 305-311.
Tringali, M.D., Leber, K.M., Halstead, W.G., McMichael, R., O’Hop, J., Winner, B., Cody, R.,
Young, C., Neidig, C.L., Wolfe, C., Forstchen, A., Barbieri, L., 2008a. Marine stock enhancement in Florida: a multi-disciplinary, stakeholder-supported, accountability-based approach. Rev. Fish. Sci. 16 (1-3), 51-57.
Tringali, M.D., Seyoum, S., Wallace, E.M., Higham, M., Taylor, R.G., Trotter, A.A.,
Whittington, J.A., 2008b. Limits to the use of contemporary genetic analyses in delineating biological populations for restocking and stock enhancement. Rev. Fish. Sci. 16 (1-3), 111-116.
Tsukamoto, K., Kuwada, H., Hirokawa, J., Oya, M., Sekiya, S., Fujimoto, H., Imaizumi, K.,
1989. Size-dependent mortality of red sea bream, Pagrus major, juveniles released with fluorescent otolith tags in News Bay. J. Fish Biol. 35 (Supp. A), 56-69.
Tsukamoto, K., Masuda, S., Endo, M., Otake, T., 1996. Behavioral characteristic of the ayu,
Plecoglossus altivelis, as predictive indices for stocking effectiveness in rivers. Nippon Suisan Gakkaishi, 56, 1177-1186.
Tucker, J., 2003. Snook culture. World Aqua. 43 (4), 42-47. Tucker, J.W., Campbell, S.W., 1988. Spawning season of common snook along the east central
Florida coast. Fla. Sci. 51 (1), 1-6. Tupper, M. Boutilier, R.G., 1997. Effects of habitat on settlement, growth, predation risk and
survival of a temperate reef fish. Mar. Ecol. Prog. Ser. 151 (1-3), 225-236. Uki, N. 2006. Stock enhancement of the Japanese scallop Patinopecten yessoensis in Hokkaido.
Fish. Res. 80, 62-66. Uzars, D., Plikshs, M., 2000. Cod (Gadus morhua L.) cannibalism in the Central Baltic:
interannual variability and influence of recruit abundance and distribution. ICES J. Mar. Sci. 57 (2), 324-329.
Victor, B.C., 1983. Recruitment and population dynamics of a coral reef fish. Science 219,
419–420. Victor, B.C., 1986. Larval settlement and juvenile mortality in recruitment-limited coral reef
fish population. Ecol. Monogr. 56, 145 – 160.
180
Vik, J.O., Borgstrøm, R., Skaala, Ø., 2001. Cannibalism governing mortality of juvneile brown
trout, Salmo trutta, in a regulated stream. Regul. Rivers: Res. Manage., 17 (4-5), 583-594.
Vinagre,C., Maia, A., Cabral, H.N., 2007. Effect of temperature and salinity on the gastric
evacuation of juvenile sole Solea solea and Solea senegalensis. J. App. Ichth. 23 (3), 240-245.
Vitousek, P.M., Aber, J.D., Howarth, R.W., Likens, G.E., Matson, P.A., Schindler, D.W.,
Schlesinger, W.H., Tilman, D.G., 1997. Human alteration of the global nitrogen cycle: Sources and consequences. Ecol. App. 7 (3), 737-750.
Volk, E.C., Schroder, S.L., Grimm, J.J., 1999. Otolith thermal marking. Fish. Res. 43 (1-3),
205-219. Volpe, A.V., 1959. Aspects of the biology of the common snook, Centropomus undecimalis
(Bloch), of southwest Florida. Florida State Board of Conservation, Technical Series Number 31.
Walters, C.J., 1997. Challenges in the management of riparian and coastal ecosystems. Cons.
Ecol. 1(2), http://www.consecol.orgt/vol1/iss2/art1. Walters, C.J., Juanes, F., 1993. Recruitment limitations as a consequence of natural selection for
use of restricted feeding habitats and predation risk taking by juvenile fishes. Can. J. Fish. Aquat. Sci. 50, 2058 – 2070.
Walters, C.J., Maguire, J., 1996. Lessons for stock assessment from the northern cod collapse.
Rev. Fish Biol. Fisher. 6, 125-137. Walters, C.J., Kitchell, J.F., 2001. Cultivation/depensation effects on juvenile survival and
recruitment: implications for the theory of fishing. Can. J. Fish. Aquat. Sci. 58, 39-50. Walters, C. J., Martell, S.J.D., 2004. Fisheries ecology and management. Princeton University
Press, New Jersey, USA. Ward, R.D., 2006. The importance of identifying spatial population structure in restocking and
stock enhancement programmes. Fish. Res. 80, 9-18. Wassenbergh, S.van, Aerts, P., Herrel, A., 2006. Scaling of suction feeding performance in the
catfish Clarias gariepinus. Phys.Biochem. Zoo. 79 (1), 43-56. Weiss, S., Schmutz, S., 1999. Performance of hatchery-reared brown trout and their effects on
wild fish in two small Austrian streams. Trans. Am. Fish. Soc. 128, 302-316. Welty, J.C., 1934. Experiments in group behavior of fishes. Physiol. Zool. 7, 85-128.
181
Wennhage, H., 2000. Vulnerability of settling plaice Pleuronectes platessa to predation: effects of developmental stage and alternative prey. Mar. Ecol. Prog. Ser. 203, 289-299.
Werckmand, C., Hardy, T., Wainwright, E., 1998. Crystal Ball Users Manual. Decisioneering,
Colorado. Werner, E.E., Gilliam, J.F., Hall, D.J., Mittelbach, G.G., 1983. An experimental test of the
effects of predation risk on habitat use in fish. Ecology 64, 1540–1548. Werner, E.E., Gilliam, J.F., 1984. The ontogenetic niche and species interactions in size-
structured populations. An. Rev. Ecol. Systemat. 15, 393-425. Wespestad, V.G., Fritz, L.W., Ingraham, W.J., Megrey, B.A., 2000. On relationships between
cannibalism, climate variability, physical transport, and recruitment success of Bering Sea walleye pollock (Theragra chalcogramma). ICES J. Mar. Sci. 57 (2), 272-278.
Wickler, W., 1968. Mimicry in plants and animals. McGraw-Hill, New York. Yamamura, O., Yabuki, K., Shida, O., Watanabe, K., Honda, S., 2001. Spring cannibalism on 1
year walleye Pollock in the Doto area, northern Japan: is it density-dependent? J. Fish. Biol., 59 (3), 645-656.
Yamashita, Y., Nagahora, S., Yamada, H., Kitagawa, K., 1994. Effects of release size on
survival and growth of Japanese Flounder Paralichthys olivaceus in coastal waters off Iwate Prefecture, northeastern Japan. Mar. Ecol. Prog. Ser. 105, 269-276.
Yoder, J.M., Marschall, E.A., Swanson, D.A., 2004. The cost of dispersal: predation as a
function of movement and site familiarity in ruffed grouse. Behav. Ecol. 15 (3), 469-476. Zohar, Y., Hines, A.H., Zmora, O., Johnson, E.G., Lipcius, R.N., Seitz, R.D., Eggleston, D.B.,
Place, A.R., Schott, E.J., Stubblefield, J.D., Chung, J.S., 2008. The Chesapeake Bay blue crab (Callinectes sapidus): A multidisciplinary approach to responsible stock replenishment. Rev. Fish. Res. 16 (1-3), 24-34.
182
183
BIOGRAPHICAL SKETCH
Nathan Paul Brennan was born in Hartford, Connecticut, USA in 1965. At the age of 2, he
moved with his family to the remote highlands of Papua New Guinea, in the Engan Province
where his father worked in the Lutheran Mission. After attending elementary school at the
Highlands Lutheran School, Nathan moved with his family to the capital city, Port Moresby.
There he attended Port Moresby International High School for three years. In 1981, the
Brennans moved to Kailua, Hawaii, USA, located on the island of Oahu where he graduated
from Kailua High School in 1984. In May, 1989, he received the degree of Bachelors of Arts in
Zoology from the University of Hawaii at Manoa.
From 1990 -1995, Nathan worked at the Oceanic Institute located at Makapu’u studying
fisheries and aquaculture under the Stock Enhancement Program. In August 1995, he was
recruited to Cookeville, Tennessee for graduate work in a master’s program at the Biology
Department of Tennessee Technological University (TTU) studying Paddlefish stocks in the
Mississippi drainage basin and earned a Master’s degree in Biology in December 1997. By
March 1998, he moved to Sarasota gaining employment at Mote Marine Laboratory. While
maintaining his position at Mote, Nathan started his doctoral program at the University of
Florida in September 2002, examining juvenile snook ecology using stock enhancement as his
primary experimental tool. Nathan currently lives in Sarasota with his wife and two children.