Shahaama Abdul Sattar, Christian Jørgensen, and Øyvind Fiksen. University of Bergen.

Fishing is identified as a potential cause for driving evolution towards earlier maturation in many fish stocks and has been suggested to lead to earlier sex change in sex-changing species such as groupers. Many studies have focused on the ecological effects of fisheries on sex-changing fish, but little attention has been directed towards understanding the evolutionary responses to high and size-dependent fishing mortality. We have developed an individual-based model of emergent size at maturity and energy allocation under varying levels of fishing mortality. In the model, individuals differ in their age at maturation and energy allocated to reproduction in male and female phase. Our results predict that these traits are quite sensitive to even low fishing mortalities. Age and length at both maturation and sex change decrease in the population with increasing fishing mortalities. The model predicts shifts (at the population level) between a hermaphroditic and dioecious strategy with increasing fishing mortalities, as well as decreasing population size and increasing female to male sex ratios with fishing effort. Yield peaks at low to intermediate levels of fishing mortality (about 0.08 year,^-1). The simplest form of management of the fishery requires implementing a low fishing mortality and choosing proper size limits for the fishery.

Suzanne H. Alonzo,¹ Teresa Ish,² Meisha Key,³ Marc Mangel,² and Alec D. MacCall.^{2 1}Yale University; ²University of California, Santa Cruz; ³California Department of Fish and Game.

Understanding population dynamics and predicting the response of species to exploitation depends on our knowledge of vital rates such survival, growth and reproduction. However, in any species there may be spatial, temporal and individual variation in these parameters. One of the most striking patterns of individual variation occurs in species that exhibit protogynous sex change where individuals reproduce first as females and later as males. Although many sex-changing species are exploited, we still have little information on how this life history pattern influences the response of a species to fishing and the performance of existing assessment methods. We use the California sheephead (Semicossyphus pulcher, a protogynous sex-changing species) as an illustrative example to determine whether existing methods can be used to assess the status of a sex-changing species and determine how uncertainty in life history parameters affect the estimated status of the stock and its predicted response to fishing. Using data from multiple sources, we developed individual based models to assess the status of the stock, determine the performance of spawning per recruit measures and examine the effect of uncertainty in key life history parameters on our ability to assess the stock dynamics of this sex-changing species.

David S. Boukal,¹ Andr M. de Roos,² and Lennart Persson.^{3 1}University of Bergen; ²University of Amsterdam; ³Ume University

Fish in exploited stocks mature earlier and usually at smaller sizes because of genetic and plastic responses. The latter occur e.g. when individual fish grow faster at lower population sizes due to reduced competition for food. Using a size-structured consumer-resource model based on a planktivorous fish life history, we show that exploitation can easily induce irreversible evolutionary changes in individual life histories and stock properties. As a result of annual spawning, early maturation at small sizes and late maturation at large sizes can become alternative, evolutionary and ecologically stable states in the same environment. Exploitation of late-maturing populations can induce evolution to smaller maturation sizes associated with stepwise decreases in age at first reproduction. We show that complete and early fishing moratoria slowly reverse this process, but belated or partial moratoria can accelerate or even instigate further evolution to smaller sizes at maturation.

Elizabeth N. Brooks,¹ Joseph E. Powers,² and Enric Cortes.^{1 1}National Marine Fisheries Service; ²Louisiana State University.

We derive analytic results for MSY benchmarks and show that they can be calculated directly from maximum lifetime reproductive rate. This rate can be calculated directly from biological parameters of maturity, fecundity, and natural mortality, or a distribution can be derived from appropriate metadata. Given an index of relative abundance, current stock status can be evaluated relative to the benchmarks to determine overfished and overfishing status. Our derivations lead to a reparameterization of the common stock recruit relationships, Beverton-Holt and Ricker, in terms of SPR_MSY. Often, the parameters in a stock recruit relationship are set based on a hypothesized or presumed level of stock resiliency; parameterizing directly in terms of SPR_MSY makes those hypotheses more transparent and more intuitive. The ability to directly calculate benchmarks from biological data makes the method an attractive option for datapoor fisheries such as those for many sharks, artisanal fisheries (e.g., the Caribbean), or bycatch species. The benchmark derivation in terms of life history information provides a link between biological characteristics and appropriate management.

Mark Butler,¹ Alison MacDiarmid,² Thomas Dolan,¹ and Michael Goodrich.^{1 1}Old Dominion University; ²NIWA, New Zealand.

Fishing-induced changes in the size structure of exploited populations may alter reproductive success not only through its effects on female size and egg production, but also via effects on male size. Fertilization success may decline in fished populations if mating depends on male-female size relationships and frequency- or size-dependent behavioral interactions. We are using spatially-explicit, individual-based modeling to integrate laboratory and field data, and to disentangle the roles that male and female abundance and size structure play in determining reproductive success. We compare spiny lobster populations in MPAs and fished regions, as well as species with contrasting mating systems and reproductive strategies, namely: the New Zealand Rock Lobster (Jasus edwardsii) and the Caribbean spiny lobster (Panulirus argus). Our results indicate that brood size may be sperm limited in heavily exploited New Zealand populations where operational sex ratios are heavily skewed toward females by fishing. In Florida, where fishing removes large individuals of both sexes from the population, we find no evidence of sperm limitation. Model results suggest that the large number of small males in the Florida population, each mating with only a few small females, may counter-balance the loss of large males that would otherwise dominate matings.

John Caddy. International Fisheries Consultant, Rome Italy.

The literature on habitat requirements of early life history stages of motile demersal macrofauna are reviewed from the perspective of the essential role of cover in many life history functions. Post-planktonic stages often settle in complex habitats with fractal characteristics that protect small organisms, and post-larval and juvenile stages may depend on topographical features which are limited in extent. A majority of benthic habitats are sedimentary and low in structural complexity and structural elements may have been further reduced by anthropogenic activities. Many marine demersal organisms move between specific habitats with size, stage and age, but the stage-specific risk of death from natural causes remains approximately the same for all stages, since as stage duration increases, predation risk per unit time declines. An evolutionary motive for this generalisation is suggested. In addition to maintaining spawning potential as suggested by a stock-recruit focus, stock replenishment is equally threatened by degradation of the habitats of successive life history stages. Migration often occurs across seascapes where cover connectivity is limited and predation risk is high. Bottlenecks in recruitment supply may occur naturally and nullify spawning success, and can be further accentuated by human activities which fragment cover resources.

For motile marine organisms, two rates of natural mortality apply: one while ‘under cover’, and a higher rate ‘in the open’ of the foraging arena, or when migrating between stage-specific habitats. Hence, loss of cover increases risk of death. It also reduces foraging success by restricting the area of adjacent foraging arenas, and affects other life history activities tied to cover. The carrying capacity of a marine habitat may thus be determined by the holding capacity of that stage-specific habitat with the lowest carrying capacity, and not by gamete production or the trophic supplies available to adults. Where a habitat bottleneck exists or is created, trophic resources and suitable habitat for adults are underused and do not necessarily limit fishery productivity. Mature individuals may also require spawning refugia where habitat protection from harvesting is essential to stock replenishment. Remediation of limited areas of critical habitat offers the potential for production enhancement or counteracting the effects of overfishing, but cannot co-exist with trawl gear designed to fish on rough bottom. A summary of representative literature on the role of marine cover is provided as an annex, and the relevance of this perspective to standard assessment and stock rebuilding theory is briefly discussed.

Lora M. Clarke,¹ Benjamin D. Walther,² Stephan B. Munch,¹ Simon R. Thorrold,² and David O. Conover.^{1 1}Stony Brook University; ²Woods Hole Oceanographic Institution.

Patterns of connectivity are important in understanding the geographic scale of local adaptation. High connectivity, or the exchange of individuals among subpopulations, is assumed in most marine species because of life histories that include widely dispersive stages. However, evidence of local adaptation in coastal species raises questions concerning the degree of connectivity. We examined geochemical signatures in the otoliths of juvenile Atlantic silversides, Menidia menidia, collected in 16 locations along the northeastern coast of the United States from New Jersey to Maine in 2003 and 9 locations in 2004 using laser ablation-inductively coupled plasma mass spectrometry (LA-ICPMS) and isotope ratio mass spectrometry. Spatial and temporal trends in otolith geochemistry were examined. Juvenile fish showed significant site-specific differences and were assigned to natal sites with 70% classification accuracy using quadratic discriminant analysis. Use of this algorithm allowed us to assign natal origin to spawning adults captured the following year in the same locations. Results show that Menidia menidia exhibit highly dispersive behavior with over 50% migrating greater than 200 km from natal sites. These findings suggest high connectivity and demonstrate marine species with largely open populations are capable of local adaptation despite apparently high gene flow.

David O. Conover and Stephan B. Munch. Stony Brook University.

Knowledge of geographic and temporal scales of adaptive genetic variation is crucial to species conservation yet our understanding of these phenomena, particularly in marine systems, is scant. Until recently, the belief has been that because most marine species have highly dispersive or mobile life stages, local adaptation could occur only on broad geographic scales. Similarly, the time scale of adaptive divergence has also been assumed to be very long, requiring thousands of generations. Recent studies of a variety of species have challenged these beliefs. There is strong evidence of geographically structured local adaptation in physiological and morphological traits and the proportion of quantitative trait variation at the among-population level (Q_ST) is much higher than it is for neutral markers (F_ST). Moreover, evidence is accumulating that selection can cause rapid adaptive divergence on contemporary time scales. The differing spatial and temporal scales of adaptive vs. neutral genetic divergence call for a new paradigm in our thinking about the relationship between phenogeography (the distribution of phenotypic variation) and phylogeography (the distribution of lineages) in marine species. The idea that contemporary selective processes can cause fine scale spatial and temporal divergence underscores the need for a new emphasis on Darwinian fishery science.

James H. Cowan. Louisiana State University.

The success of single species approaches to fisheries management has been disappointing, and has resulted in a national dialogue about concepts of ecosystem management, as emphasized in the President’s Commission Report on the State of the Ocean, the Pew Ocean’s Report, language in the Sustainable Fisheries Act (SFA), and in a recent NRC (2006) study. As a first step towards the ecosystem approach, the SFA mandated delineation of “essential fisheries habitat”, but provided only a vague description of how this was to be defined. Subsequent guidelines and research in response to this mandate have attempted to clarify the definition of EFH, and to describe important fish/habitat relationships for a variety of species, but neither have convincingly excluded or included many habitats from consideration. I contend that problems have arisen because delineation of EFH is an extension of the single species approach, albeit with more consideration of geography and life history. Moreover, in several recent studies it has been shown that man-induced changes in ecosystem function can result from top-down effects such as fishing, habitat modifications, pollution, eutrophication, etc., resulting in a shift in the ecological baseline. In such cases, the altered ecosystems are often much less responsive to simple management actions that attempt to recover ecosystem functionality. This occurs for a variety of reasons ranging from reductions or changes in habitat, to reorganizations of foodwebs because of the removal of strong interactors in the ecosystem. Regardless of the mechanisms, however, alternate steady states that have been caused by forcing from the top-down may be less likely to return to a state that resembles “pristine”, and thus less likely to provide ecological goods and services and fisheries productivity that are similar to pre-disturbed conditions. In my opinion, this further compromises the need to define EFH from a single species perspective, and reiterates the need to move as quickly as possible towards ecosystems-based fisheries management.

Sean P. Cox¹ and Allen Robert Kronlund.^{2 1}Simon Fraser University; ²Fisheries and Oceans Canada.

Maternal age effects, in which survival rates of progeny from older mature fish are greater than those for younger mature fish, are likely to cause non-stationarity in stock production relationships. Such non-stationarity could cause rapid stock declines, or even collapse, under intense size-selective fishing. Although maternal age effects have been demonstrated in laboratory studies, it is unlikely that this phenomenon could be detected and taken into account in quantitative stock assessments for large-scale commercial fisheries. However, this does not imply that such effects should be ignored in developing robust fishery management procedures. We developed a closed-loop simulation procedure to test the performance of adaptive and non-adaptive harvest decision rules under severe, moderate, and no maternal age effects on juvenile survival. Neither the assessment models nor the decision rules took maternal age effects into account. We applied the approach to British Columbia sablefish (Anoplopoma fimbria), which is a long-lived, deep water species that grows relatively rapidly in early life and matures at 5-7 years of age despite longevity of approximately 70-80 years.

Kate Cresswell. University of California, Santa Cruz.

Antarctic krill are central to Southern Ocean ecosystems, forming in many areas a direct link between the primary production and higher predators. The chief economic interest in krill stems from a krill fishery that began in the late 1970’s, and although the catch has remained at a low steady rate, recent advancements in krill processing threaten to increase drastically the impact of this fishery. Krill are under additional pressure due to a particularly high predation rate from land-based predators during summer in the Southern Ocean. In addition, some authors have reported a long-term decline in krill by as much as 80% over the last 30 years, due most probably to a life-history constraint that links juvenile survival to a sea-ice extent that is declining rapidly with global warming. The abundance of krill manifests itself in the breeding success of land-based predators during sensitive life-history stages; predators in some areas suffer breeding failure in years of low krill abundance. We used an individual-based modeling technique to investigate the interaction between krill and their predators in specific areas of the Southern Ocean, to predict the effects of an increasing krill fishery.

Katy K. Doctor. University of Washington.

Fisheries have the ability to create strong selection pressures on life history traits of target species. The effects of fishery selection on the upriver migration timing of sockeye salmon (Oncorhynchus nerka) populations in the commercially productive Bristol Bay region of Alaska is one such example. This fishery is highly productive, as indicated by record sockeye catches occurring over the past 20 years. One explanation for this sustained success is the inherent biocomplexity of these stocks. Local adaptation of life history traits has enabled stocks to sustain productivity and has provided the system as a whole with resilience during different climatic regimes and events. Migration timing is a highly heritable trait that exists within many aquatic and marine species. The predictable and consistent upriver migration timing of Pacific salmon provide an excellent model to study fishery selection. Consistent upriver migrations of locally adapted populations are driven by a combination of environmental and genetic controls. Evidence of segregation in upriver migration timing among spawning populations has been tested and could prove important when examining the effects of commercial fishing. The Bristol Bay fishery is managed by escapement and, due to the nature of this strategy, is temporally biased towards heavier fishing at the end of the run. This bias in fishing pressure allows the examination of the effects that selective fishing may have on segregated migratory stocks.

Tara A. Duffy, David O. Conover, Anne E. McElroy. Stony Brook University.

Reproductive capacity, sex ratio skew, and population success of aquatic organisms are affected by xenobiotic chemicals in coastal waters, demonstrated in numerous species of fish. Organisms that possess unique sex determining systems, may be particularly sensitive indicators of these anthropogenic impacts. The Atlantic silverside, Menidia menidia, is used to address reproductive impacts as it exhibits both temperature-dependent sex determination (TSD) and genetic sex determination (GSD), depending upon latitude. Populations across the species range were reared at two temperatures to determine the geographic distribution of the sex determining mechanism. Fish tissues were examined to determine patterns of reproductive development. Endocrine disruption of populations was addressed by exposing silversides to varying concentrations of an exogenous estrogen. Exposure during the period of sex differentiation indicated differential sensitivity to xenoestrogens between populations, with greater impact in fish exhibiting TSD. Microgeographic variation ofsex ratio in relation to urbanization was assessed using fish collected from New York estuaries. Histological sections were examined to determine reproductive impairment. Disruption was noted both in fish tissues and in skewed sex ratios. Species with ESD appear to be more sensitive to xenoestrogens than those with strict genetic sex determination, and may therefore be better indicators of anthropogenic influence.

Erin S. Dunlop,¹ Marissa L. Baskett,² Mikko Heino,³ and Ulf Dieckmann.^{1 1}International Institute for Applied Systems Analysis, Laxenburg, Austria; ²Princeton University; ³University of Bergen.

Several recent theoretical and empirical studies have provided evidence that fishing is capable of inducing evolutionary changes in key life history traits. These evolutionary changes can have unwanted consequences, such as reduced body sizes in the catch, which might lead to a deterioration of the quality of the fishery. Therefore, managers need viable options for slowing, stopping, or reversing the evolutionary consequences of fishing. In this study, we explore one potential management strategy by developing and analyzing an eco-genetic model aimed at studying the effects of marine reserves on fishing-induced evolution. Our model advances previous theoretical approaches by including features such as phenotypic plasticity, density-dependent growth, and evolution of multiple life history traits. We parameterize our model for a population of cod that undergoes an annual migration from feeding grounds to spawning grounds. Using our model, we explore the consequences of marine reserve location (either in the feeding grounds or in the spawning grounds) and proportion of area protected on the speed, direction, and magnitude of evolutionary responses. The results of our model underscore the importance of having an evolutionary perspective when implementing management strategies aimed at protecting commercially important fish stocks.

William Eldridge,¹ Jeff Hard,² and Kerry Naish.^{1 1}University of Washington; ²NOAA, Seattle.

There are several empirical examples of the role of harvest in causing evolution of fish populations. Few methods, however, have explicitly used quantitative genetic approaches to understand the relationships between life history variation, harvest strategies and fishery induced evolution. We developed a genetic-based model of a Chinook salmon fishery, parameterized from empirical data and incorporating heritability values for length at age, to assess a range of harvest regimes on long-term abundance. Lengths at age for each age group were treated as different, but correlated, traits. The model showed that a constant exploitation rate above a minimum size will reduce abundance below levels predicted by a model that does not consider genetic diversity. All age groups, including those not under selection, respond to selection by becoming smaller and less fecund. When harvest occurs between a minimum and maximum size limit only, the population will evolve rapid growth to avoid the fishery, and the population may have higher abundance after 100 years than would be predicted by a nongenetic model. Under both selection regimes a faster growing population could sustain higher harvest rates and therefore had greater responses to selection.

Katja Enberg, Erin S. Dunlop and Ulf Dieckmann. International Institute for Applied Systems Analysis, Austria.

As a result of the ongoing declining trends in the abundance of many exploited fish stocks, fisheries management is compelled to deal also with stocks that have collapsed and are in the phase of recovery. Evolutionary changes caused by fisheries are known to affect the genetic and phenotypic structure of the exploited fish stocks. These changes have been most visible on the life history characteristics influencing age and size at maturation traits that have a major influence on the reproductive potential of an individual. Is the recovery potential of fished stocks also influenced by such changes? Moreover, are such adaptive changes in, for example, maturity schedule, restorable? In our paper we study, by means of an eco-genetic model with multiple evolving traits, how fisheries-induces evolution affects and is affected by the collapse and recovery of fish stocks. The population dynamical component of our model is influenced by environmental variability, and we have parameterized the model with life history traits resembling those of Atlantic cod. We investigate the interplay between ecological (population size and biomass) and evolutionary (genetic composition and adaptability) recovery and study whether it is possible to facilitate the recovery process.

Timothy E. Essington¹ and Phillip Levin.^{2 1}University of Washington; ²NOAA Fisheries, Seattle.

One justification for developing fisheries on lower-trophic level species is that fishing "replaces" predation that would otherwise be carried out by the depleted high trophic level species. Yet, fish predators and fisheries generally have markedly different size selectivities, so they remove individuals in different life-history stages. We evaluated the equivalancy of predation and fishing losses in generic population models that were based on empirically derived life-history invariants. These invariants, which are markedly similar across taxonomically related species, characterize the gross nature of life-history strategies adopted by taxonomic groups. We parameterized the model for the orders pleuronectiformes, clupeiformes, gadiformes, and the genus Sebastes and evaluated whether the magnitude of population regulation imposed by predation is equivalent to that imposed by fishing. We find that in most cases, fishing has the largest effect on population growth rate because it tends to remove individuals in life stages that have the highest reproductive value. This pattern was most pronounced for Sebastes and pleuronectiformes and was least pronounced for clupeiformes. These contrasts result from the markedly different life-history strategies adopted by these species. We suggest that fishing be viewed as distinct from predation when the cumulative impacts of fisheries development on food webs are evaluated.

Mikko Heino¹ and Ulf Dieckmann.^{2 1}University of Bergen; ²Leiden University.

Today, fishing is the dominant source of mortality in most commercially exploited fish stocks. Life-history theory predicts that changes in mortality regimes cause selection on life-history traits. In particular, increased mortality can strongly favor earlier maturation. Indeed, commercially exploited fish stocks often show trends towards earlier maturation. However, earlier maturation may also simply reflect phenotypic plasticity—triggered, for example, by improved individual growth in overexploited stocks. Until recently, the difficulties involved in disentangling plastic and evolutionary components of life-history changes have hindered understanding the nature of these phenotypic changes. To help overcome this problem, we introduced probabilistic reaction norms for age and size at maturation: by estimating such reaction norms, one can control for growth-related phenotypic plasticity and changes in mortality. A suite of methods for estimating these reaction norms is now available. Addressing different types of data, these methods have been applied to at least 17 stocks, representing eight different species of marine and freshwater fish. All but two of these case studies suggest that a significant evolutionary component has contributed to the observed trends in age and size at maturation. Remarkably, this component is often detectable at time scales as short as a couple of decades.

Selina Heppell and Scott Heppell. Oregon State University.

Simple compensation models for stock assessment assume a constant and instantaneous relationship between population density and growth. While the models can be modified for slow species such as sharks, or fast species, such as herring, it has long been recognized that this fast-slow continuum is insufficient for teleosts because many species follow a periodic or bet-hedging spawning strategy linked to extreme variability in environmental conditions that affect larval survival. The importance of strong year classes is well recognized, but selection pressures that have shaped the bet-hedging strategy have generally not been considered in management. For long-lived species such as Pacific rockfishes, Atlantic cod, orange roughy, sturgeon and freshwater suckers, populations may typically exhibit a declining annual population growth rate marked by occasional pulses of strong recruitment. Occasional may be as rare as once per generation episodic, rather than periodic and density-dependence may be manifested differently from Beverton-Holt-type stock-recruit relationships. Age-structure may be critical in these populations to assure adequate spawning during the good years, either through protracted spawning seasons or larval condition factors determined by maternal effects. These species must be managed in ways that promote, rather than suppress, their natural resilience to environmental perturbations, including climate change.

William F. Herrnkind,¹ Mark J. Butler,² and John H. Hunt.^{3 1}Florida State University; ²Old Dominion University; ³Florida Marine Research Institute.

On the basis of extensive empirical research on the early life history of Caribbean spiny lobster (Panulirus argus) in the Florida Keys nursery, we hypothesized that recruitment is limited in a density-dependent manner by shelter-imposed thresholds, below which the population fluctuates in response to local changes in postlarval supply. Thus, we envisioned the nursery to be a mosaic where, at the extremes, juvenile recruitment at some sites is limited by nursery-habitat suitability and at others by postlarval supply. Local juvenile populations are linked by the subsequent movement of older nomadic juveniles among sites, any of which can be influenced by temporal variation in postlarval supply or environmental degradation. Our spatially explicit individual-based model, constructed from measured biological and ecological parameters of early life stages, suggested that recruitment operates in this way. We recently concluded a four-year study designed to test the hypothesis and to specify how the supply of postlarvae varies along the Florida Keys archipelago and whether these conditions vary according to regional oceanographic features. We also modeled and experimentally tested how postlarval supply and nursery habitat structure link to the local recruitment of juveniles at sites throughout the nursery. Results support our hypothesis, suggesting that integrating life-stage features, spatial and temporal heterogeneity in habitat structure, and postlarval supply is essential to understanding and predicting population-level consequences to recruitment.

Lyndie A. Hice and David O. Conover. Stony Brook University.

The Atlantic silverside, Menidia menidia, displays strong latitudinal variation in vertebral number along the east coast of North America. To determine how gene flow and local adaptation might influence this pattern, we are investigating fine-scaled variation in vertebral number. Field data showed an increase in the mean number of vertebrae with increasing latitude, in accordance with Jordan’s rule. Progeny of field populations reared in a common environment show a similar trend, indicating a genetic basis to the trait. Such strong correlation with an environmental gradient implies that variation in vertebral number is adaptive. Moreover, we present evidence from a long-term selection experiment that vertebral number evolves rapidly in concert with selection on body size. Our field results also provided evidence of a flattening of correlation with latitude at the northern and southern ends of the distribution, implying that gene flow constraints local adaptation at the edges of the range. There were also abrupt shifts in the relationship between vertebral number and latitude near regions of distinct habitat change such as Cape Hatteras and Cape Cod. Taken together, these results reveal the interplay of gene flow and local adaptation in determining geography of vertebral number.

Ray Hilborn and Carolina V. Minte-Vera. University of Washington.

In recent years there has been considerable interest in the evolutionary impacts of fishing, and many have argued that selective pressure on large fish can cause growth rates to decline with major implications to sustainability of fishing. In this paper we review the data on heavily exploited fish populations and show that contrary to these expectations, most marine fish stocks tend to grow faster rather than slower when subjected to heavy fishing. Further we show that the actual selective regimes in marine fish stocks are vastly different from the experimental fishing regimes which are the basis for most concerns, and the selective pressure in most fisheries is not particularly strong. While certainly there is some selective pressure in many fisheries against fast growing individuals, the trophic dynamics of ecosystem appear to trump any evolutionary impacts.

K. T. Honey¹ and R. M. Fujita.^{2 1}Stanford University; ²Environmental Defense.

Many California fisheries have declined in recent decades, both ecologically and economically. Historically, fishery managers focused on maximizing access to public trust ocean resources. This management approach fostered an emphasis on short-term cash flow, rather than on longer-term asset building in many fisheries. Alternative management models exist to create stewardship incentives with potential to simultaneously benefit the environment, fishermen, and coastal communities. The aim of our project is to ascertain the key biological, economic, and cultural determinants of success in meeting specific objectives for a variety of fishery management regimes (e.g., community-based management, fishery cooperatives, area-based management, community development quotas, and individual transferable quota programs). Results from our primary literature review and preliminary field interviews suggest that collaborative research, localized management, and regulatory reform will be necessary. We are developing several pilot projects in collaboration with California fishermen incorporating these elements. We are also developing an integrated approach to reforming the harvest, processing, and distribution/marketing sectors in order to achieve synergies. These integrated research and management models will include detailed proposals of preferred management options—explicitly incorporating fisherman "life-history" and coastal community preferences—to enhance prospects for protecting and enhancing ocean ecosystems, fishery revenues, coastal communities, and working waterfronts.

Sergey Ignatyev. Institute of Biology of the Southern Seas NASU, Sevastopol, Ukraine.

My research is based on a year-round study of the ichthyofauna in coastal waters near the Ukrainian Vernadsky Antarctic Station in 2002-03. Of the eight species collected (seven Notothenidae, one Harpagiferidae), bullhead notothen (Notothenia coriiceps) occurred most frequently (71%), followed by emerald notothen (Trematomus bernacchii) (17%) and dusky notothen (T. newnesi) (7%). All species were benthic or near-benthic, and most were predators-benthophages. Amphipods and isopods dominated their diet, which also include algae and mollusks. Krill made up less than 3%. Bullhead notothen mean length was 22.8 cm (ranged 12.5–43.5 cm); mean weight was 387.4 g (range 60–1820 g); 35% of individuals were 12–25 cm in length and 65% over 25 cm. Male:female ratio was 1:1.6. Males averaged 25.7 cm in length (range 13.0–37.0 cm); females 23.5 cm (range 12.5–43.5 cm). Gonadsomatical index revealed one pulse in gametogenesis and autumn-winter spawning. It peaked in April–May, but began increasing as early as September. Emerald notothen mean length was 14.3 cm (range 5.4–21.0 cm); mean weight was 72.8 g. Male:female ratio was 1:1.6. Males averaged 11.95 cm (range 5.4–15.5 cm), females 15.7 cm (range 11.5–21.0 cm. Spawning took place in October–November.

Christian Jørgensen,¹ Øyvind Fiksen,¹ and Bruno Ernande.^{2 1}University of Bergen; ²IFREMER, Port-en-Bessin, France.

Field and lab experiments, analyses of fisheries data, theory, and models all corroborate that harvest may induce rapid and substantial life-history changes in many exploited fish stocks. Which life history trait will change and to what degree depend on the selectivity of the fishery and the ecology and life history of the species in question, making it hard to reach generalizable conclusions about potentially successful management options. Here, we ask to what degree different management regimes may decrease or reverse the negative effects of harvest-induced life-history evolution in the Northeast Arctic cod. We use a state-dependent energy-allocation life-history model for cod. A set of life-history strategies are optimized by varying the mortality, and we then use quantitative genetics to describe how these strategies change in frequency in a population as a result of harvest-induced selection. We can thus quantify the evolutionary effect of fishing on life-history change and assess evolutionary rates. We present the effects of three different management regimes: marine protected area, maximum size limit, and minimum size limit. Marine protected areas and minimum size limits result in reduced evolutionary rates compared to current fishing regimes, while maximum size limits produce little change.

Neala Kendall and Tom Quinn. University of Washington.

Life history traits of wild animals can be strongly influenced by human activities. Fishing gear, specifically gillnets, selectively remove certain individuals within a population and can affect life-history traits such as size and age at maturity. A sockeye salmon gillnet fishery has been located at the Wood River system of Bristol Bay, Alaska, for over 100 years. Past research suggests that this fishery can be selective on size and age based on mesh size and fishery timing. However, fishing pressure and fishery management have varied greatly among years, and long-term selection has not been examined. I am investigating the magnitude and nature of gear selectivity by this fishery on length and age at maturity of the sockeye salmon over time. These processes will be examined for the lake system as a whole and individual spawning populations within the system using over 50 years of data. A historical reconstruction of size- and age-selective fishing will be performed. Selection metrics to estimate vulnerability of sockeye of different lengths and ages and selection variability will be calculated. I will also assess population-specific fishery exploitation, and will model the effects of fishing on size and age at maturity on these populations.

L. A. Kerr and D. H. Secor. Chesapeake Biological Laboratory, University of Maryland.

During their first year of life, estuarine-dependent white perch in the Patuxent River will either exhibit retentive behaviors causing juveniles to persist in freshwater natal habitats (retentive contingent) or disperse into down-estuary brackish habitats (dispersive contingent). Otolith microchemical analysis showed that these two behaviors are discrete and have consequences to population resiliency. Here, we test bioenergetic consequences of either behavior. A randomized factorial experiment with two contingent types (fresh and brackish water) and two salinity treatments (1 and 8) was conducted over a 30 day period. Based upon osmoregulatory costs, we hypothesized that fish reared in mesohaline conditions would devote a greater apportionment of energy towards growth, and feeding metabolism than those reared in freshwater, regardless of contingent membership. Alternatively, if contingent membership is associated with varying energetic tactics, a contingent effect should be observed. Experiments supported higher allocation of energy to growth in mesohaline conditions and evidence a small amplitude contingent effect. We conclude that habitat use of mesohaline environments by juvenile white perch results in higher growth rates regardless of dispersal history. We speculate that the retention of juveniles in freshwater nurseries is related to survival benefits, perhaps due to reduced exposure to predation that occurs during dispersal.

Holly K. Kindsvater,¹ Marc Mangel,² and Michael B. Bonsall.^{3 1}University of Florida; ²University of California, Santa Cruz; ³University of Oxford, U.K.

Life history has recently been recognized to be a critical component of species management, especially for the long-lived members of the genus Sebastes, a diverse group of fishes in the Northeast Pacific. We examine correlations in ecological and physical habitat traits with life-history. Some members of this genus exhibit extraordinary longevity (>150 years). Life-history theory provides the expectation that ecological traits (asymptotic body size, age, and size at maturity) are correlated with lifespan. Alternatively, habitat traits (depth, temperature, and oxygen concentration) may underlie physiological mechanisms explaining lifespan. We used phylogenetically independent contrasts in order to control for the possibility that the correlation in traits was due to evolutionary history. We used the most recent phylogeny available with estimates of ecological (maximum size, age at maturity and size at maturity) and physical traits (depth, temperature, and dissolved oxygen concentration) compiled from the literature. We found that, after phylogeny is corrected for, age and size at maturity predicted lifespan, suggesting the primary importance of natural mortality in shaping the evolution of Sebastes species and supporting a more comprehensive treatment of this parameter in fishery stock-assessment models.

Christopher C. Koenig, Felicia C. Coleman, and Maurizio Tomaiuolo. Florida State University.

During the 1990s researchers discovered that the percent males in the gag, Mycteroperca microlepis, population had declined from a historical level of 17–20% to 2.4% in both the Gulf of Mexico and the South Atlantic Bight. The decline was correlated with, and thus assumed related to, increased fishing pressure. Our work on gag in the Madison Swanson Fishery Reserve (NE Gulf) and other work suggest a mechanism for fishing-induced changes in sex ratio. We found that males remain on spawning sites year-round, sex change is initiated during the spawning period, and most transitionals and new males occur after this period. Males increase in the catch after the spawning season, as fishers target aggregation sites year round. It is therefore likely that the species is constantly compensating for the paucity of males in the spawning aggregations, but never realizing the unfished spawning sex ratio because males are caught up between spawning (and sex-change induction) periods. Based on this model only closed areas, not closed seasons, protect the normal sex ratio in this species.

A. T. Laugen,¹ P. Boudry,² and B. Ernande.^{1 1}IFREMER, Port-en-Bessin, France; ²IFREMER, La Tremblade, France.

In addition to demographic consequences for the target species, commercial exploitation of living resources may induce adaptive changes in life history traits because of selective harvesting. Such changes may be expressed as either immediate plastic responses to environmental variation or as microevolution potentially occurring within decades. Using a suitable model system, the Pacific oyster (Crassostrea gigas), we aim to disentangle plastic and evolutionary components of life-history changes in this commercially important species. Following the extinction of the Portuguese oyster (C. angulata) 35 years ago, Pacific oysters originating from Japan were introduced to the France to sustain production. Since then, there has been a gradual decrease in growth rates and delay in timing of spawning. By using time series of environmental parameters to remove noise due to plastic responses in time series of phenotypic traits we may observe temporal trends in residuals and thereby separate the effects of plasticity and evolution on the observed changes in growth and reproduction. Whereas plastic responses can be reversed within a generation, backtracking undesirable evolutionary changes generally requires several generations. Investigating the causes of life-history changes is therefore crucial for the selection of correct management strategies to obtain long-term sustainable yields in this species.

Robert Lessard, Ray Hilborn, and Brandon Chasco. University of Washington.

Managing salmon stocks generally involves establishing escapement goals by fitting mathematical curves to spawner-recruit data. The objective is usually to optimize for optimal sustained yield, a biological conservation target, or some combination of the two. Analyses have generally not considered more detailed life-history either because of a lack of data on abundances at intermediate stages of life history against which to test models, or because maturation and migration rates do not vary enough for consideration. Sockeye salmon populations in Bristol Bay, Alaska, show a great deal of variation in smolt migration timing and duration of ocean residency, distinguishing them from other salmonid species as well as sockeye populations from many other systems. We compare the results of establishing escapement goals without life-history details to those of an analysis where age-structured migration and survival are considered in greater detail. We show explicitly that these life-history details can explain variation otherwise indiscernible in simple spawner-recruit models, lending different perspectives to harvest strategies.

Kai Lorenzen. Imperial College London.

Most fish population models and stock assessment methods assume that population regulation occurs exclusively in the pre-recruit phase of the life cycle, through density-dependent mortality. However, there is increasing evidence that processes in the recruited phase of the life cycle can play an important role in population regulation and dynamics. Quantitatively the most important processes is density-dependent growth which affects both current population biomass and, through interaction with size-dependent reproductive development, reproductive output and future recruitment. I review the literature and use population modelling and comparative analyses to synthesize current understanding of the role of density-dependent growth in fish population dynamics. The effect of density-dependent growth on reproductive output depends on population structure and thus, the processes driving variation in abundance. In general, the strength of density-dependence in the recruited phase is somewhat lower than that in the pre-recruit phase of the life cycle, but not insignificant. Density-dependent effects in the pre-recruit and recruited phases are not independent, so that population carrying capacity can be predicted from the density-dependent parameters of either the stock-recruitment or body growth models. This suggests a need to review the concept of recruitment limitation in fish populations and its implications. Analysis of density-dependence in growth, which reflects resource limitation, also helps to disentangle the action of bottom-up and top-down effects on population regulation. I close by outlining implications of density-dependent growth for fisheries management, including the design of marine reserves and stock enhancement initiatives.

Susan K. Lowerre-Barbieri, Joel Bickford, and Sarah Walters. Florida Fish and Wildlife Research Institute.

Estimates of reproductive output, or fecundity, play an important roll in stock assessment. In multiply spawning fish, fecundity is the product of batch fecundity (the number of eggs released in one event) and spawning frequency (the number of times a female spawns within a spawning season). Population spawning frequency estimates are calculated from the percentage of females that are active spawners and on the assumption that there is not immigration or emigration from the spawning population. This method was developed based on northern anchovy, a schooling species, and has since been applied to numerous species with varying levels of aggregation. The degree to which spotted seatrout aggregate to spawn varies with spawning habitat/location. Because spawning fish are contagious, estimates of the percentage of active spawners will vary due to the proximity of sampling to a spawning site. Spawning frequencies based on fish collected the following morning would be expected to be more representative of the population. However, with the advent of telemetry we begin to be able to evaluate individual spawning frequencies. For a nonschooling fish, such as seatrout, individual spawning frequencies appear to be both sex-specific and much lower for females than expected from population estimates.

Yasmin Lucero. University of California, Santa Cruz.

Laboratory research has measured an age-dependent maternal effect in black rockfish (Sebastes melanops); larvae from old mothers grow faster and survive longer under starvation conditions than do larvae from young mothers. This observation raises the question: Has fishing, and its concurrent change in population age-structure, affected rockfish productivity more than we thought? The answer to this question depends on how the observed maternal effect interacts with the highly variable recruitment and complex early-life history of rockfish. I use an explicit population model to address the question. I have adapted a Beverton-Holt stock recruitment model to depend on maternal age-distribution, and modeled mortality parameters to be functions of maternal age. This yields a very general and tunable model of a population with an age-dependent maternal effect. I will show how population productivity and fisheries metrics change with the addition of a maternal effect. I will show how time to recovery is affected by a marine reserve in the presence of age-dependent maternal effects.

Marc Mangel. University of California, Santa Cruz.

One of the great challenges of biology is to understand, simultaneously, pattern (across time or space) and variation (differences in pattern). In the salmonids, this challenge arises in the context of the major life-history events of migration from fresh water to the sea and returning from seawater to fresh water. I will present life-history models that combine proximate (physiological mechanism) and ultimate (natural selection) considerations and that allow us to understand both the pattern and variation in Atlantic, Chinook, and coho salmon, Arctic charr, and steelhead. The models suggest generalizations about top-down and bottom-up control of life histories, about the evolution of diadromy, and about implications for the management of fisheries, the recovery of salmonid populations, and effective aquaculture.

Steven J. D. Martell,¹ Meaghan Darcy,¹ Gerard DiNardo,² and Carl J. Walters.^{1 1}University of British Columbia; ²NOAA Fisheries, Hawaii.

Benthic fishes and invertebrates in archipelago regions are isolated at the post-recruitment stage due to vast distances and extreme depths between adjacent islands. Connectivity between islands primarily occurs through larval dispersal, and settlement rates at each island are largely dictated by physical oceanography. Consequences of ignoring connectivity on estimates of F_MSY and optimal harvest strategies are unknown. We develop a hierarchical assessment framework that assumes density-independent dispersal and density-dependent settlement rates.. The model is fit to commercial catch-effort data from 13 fishing banks for two lobster species in the northwester Hawaiian Islands. Ignoring meta-population structure biases estimates of maximum fishing mortality thresholds (MFMT) upwards and minimum stock size thresholds at each bank (MSST) downwards. Including meta-population structure protracts recovery times and requires alternative strategies for the allocation of fishing effort.

Ivan Mateo. University of Rhode Island.

Several authors state that foraging conditions and food web dynamics may be contributing to declines in Atlantic cod stocks. Therefore, it is essential to take a food web perspective to understand the complicated array of potential interactions affecting marine communities. The widely used Wisconsin bioenergetics model uses an energy-balance approach calculated on a daily time step, which allows for a fine-grained analysis of trophic interactions over various time scales. Bioenergetics modeling syntheses have been made for many important fishes within the Great Lakes. However, few have been developed for US Northeastern Continental Shelf.

Growth performance of Georges Bank and Gulf of Maine Atlantic cod during year 2004 was examined using bioenergetics modeling to synthesize information on their trophic dynamics. Growth efficiency, which incorporates daily growth and consumption rates, was used as a measure of growth performance. Overall growth performance for Atlantic cod was significantly lower at Georges Bank than in Gulf of Maine. Monthly individual consumption demand and specific growth rates for Atlantic cod were significantly higher on Georges Bank than in the Gulf of Maine. Increasing water temperatures, which approached the upper limits of thermal tolerances for cod in Georges Bank, led to decreasing growth efficiencies for cod. Temperatures and energetic content of diets were less variable in Gulf of Maine, which generated more consistent growth efficiencies.

Thomas R. Matthews,¹ Kerry E. Maxwell,^1,2 Rodney D. Bertelsen,¹ and Charles D. Derby.^{2 1}Florida Fish and Wildlife Conservation Commission, ²Georgia State University.

Accurate age estimates in the commercially important Caribbean spiny lobster, P. argus, are an essential element of life history and population analyses. We histologically determined the lipofuscin content in the central nervous system of known-age spiny lobsters reared in the laboratory to verify lipofuscin accumulation patterns. The quantification of neurolipofuscin is a relatively new but established technique for aging crustaceans but has not been previously attempted on P. argus. We then applied the technique to determine the age of wild lobsters in the Florida Keys and explored the potential effect of lobster age on egg production. Modal progression suggested that the vast majority of lobsters in the population were in the year 1 or year 2 age class and those older lobsters may reproduce earlier in the year and produce more clutches of eggs.

E. J. Niklitschek¹ and David H. Secor.^{2 1}Universidad Austral de Chile; ²University of Maryland.

We tested and modeled bioenergetics, growth, survival and behavior responses of juvenile Atlantic sturgeon to temperature, salinity and dissolved oxygen levels currently observed in the Chesapeake Bay. A semi-mechanistic multivariable bioenergetics model for predicting growth was built based upon Fry’s paradigm. The model explained a higher fraction of the variance (Akaike’s index) than one simply fit by multinomial quadratic functions. In laboratory behavioral studies, sturgeon selected those water quality conditions according to preferred habitats predicted by the bioenergetics model. Explicit maps of potential production were generated for a year coinciding with an experimental hatchery release of juvenile Atlantic sturgeon. Recapture locations of released and wild juvenile Atlantic sturgeon occurred in regions of the Chesapeake Bay predicted to support positive production. On the other hand, higher than expected growth rates suggested active habitat selection by released fish during the time from release to recapture.

Michael R. O’Farrell¹ and Louis W. Botsford.^{2 1}NMFS, Santa Cruz; ²University of California, Davis.

A common goal of fisheries management is to maintain a stock’s reproductive potential at a level that will allow for population sustainability, given inherent uncertainties in the relationship between stock and recruitment. However, measures of reproductive potential commonly used in fisheries generally do not include functions accounting for variation in larval survival dependent on maternal age. Here, we incorporate maternal age dependent larval mortality into calculations of lifetime reproduction to evaluate the performance of conventional fisheries management when the manager is ignorant to the existence of maternal age effects. Results suggest that managing without knowledge of maternal effects can be effective when high larval mortality rates occur only for the progeny of the youngest reproductive females. However, if larval mortality is a continuously decreasing function of maternal age over a large portion of the maternal lifespan, ignorance of maternal effects may be costly for sustainability. We conclude by considering the role of marine reserves in managing a population in which maternal age effects exist. This paper investigates the converse of a central theme for this Mote symposium: the fishery consequences of life histories.

Guido Plaza¹ and Minoru Ishida.^{2 1}Universidad Católica de Valparaíso; ²National Research Insitute of Fisheries Science, Kochi, Japan.

The bigger is better mechanism (i.e., mortality is a decreasing function of size) was tested for larvae of the Japanese sardine Sardinops melanostictus using (i) general linear models (GLM) and (ii) the traditional approach. In GLMs the radii at age were used as the dependent variables, and the standardized residuals of both regressions OR on age and OR on total length (TL), age class (AC), and day of the year (DOY) as independent variables. In the traditional approach the increment width and otolith radius of original cohorts were compared with those of survivors sampled in later stages. GLMs showed a significant increase in radii at age from younger to older age classes supporting the bigger is better mechanism. Likewise, the traditional approach showed that survivors from 10-days-hatched cohorts had significantly wider daily increment widths and otolith radii at age, throughout their first month of life, than the original populations. A further outcome was a distinctive linear growth pattern throughout larval and early juvenile stages in the relationships TL on age (r² = 0.87), OR on age (r² = 0.90) and OR on TL (r² = 0.95), contrasting with the allometric growth patterns previously documented for the larval stage of this species and other clupeoids.

Peter J. Rubec,¹ Jesse Lewis,¹ David Reed,¹ Charles F. Ashbaugh,² Curt Lashley,² Salvatore Versaggi,³ Robert H. Weisberg,⁴ Lianyuan Zheng,⁴ Ruoying He,⁴ and Chris Jenkins.^{5 1}Florida Fish and Wildlife Conservation Commission, ²SASCO Inc., ³Versaggi Shrimp Corp., ⁴University of South Florida, ⁵University of Colorado at Boulder.

A study was conducted to model and map the spatial distributions and abundances of pink shrimp (Farfantepenaeus duorarum) on the West Florida Shelf (WFS) using habitat suitability modeling (HSM). Data loggers and an electronic logbook system installed on three shrimp boats were used to gather data concerning catch (lbs) and effort (hours fished) for pink shrimp along with associated bottom temperature, salinity, and depth data at known coordinates during fishing operations. Data provided by the fishing company collected using a vessel monitoring system (VMS) allowed the creation of a map depicting areas with high fishing effort. Significantly higher mean catch rates (CPUEs) of pink shrimp occurred on the WFS during June to September, and October to December 2004 in comparison to January to March, and April to June 2005. Oceanographic modeling predicted monthly averaged bottom currents (speed and direction) and temperatures for a 16-month period from March 2004 to June 2005. Current speed and direction data indicated marked upwelling onto the WFS during 2004, and downwelling from the shelf during 2005. Sediment data from the WFS were interpolated to produce a sediment distribution map. Suitability functions were created across environmental gradients to predict CPUEs in relation to depth, aspect, bottom type, bottom temperature, current speed, current direction, and VMS zones. The HSM linked to geographic information systems were used to predict spatial distributions and abundances of pink shrimp monthly from March 2004 to June 2005. The areas with the most pronounced upwelling were also the areas that the HSM analyses predicted should have the highest catch rates. This was verified by overlaying observed CPUEs from the fishing vessels onto the suitability zones predicted by the HSM. Nutrients carried onto the shelf promoted higher shrimp abundances. The analyses estimated mean CPUEs by HSM zones. Linking fisheries to oceanography can explain how the ecosystem functions to the benefit of both the fishing industry and fisheries management.

Bernard Sainte-Marie,¹ Thierry Gosselin,^1,2 Jean-Marie Sévigny,¹ and Nicola Urbani.^{3 1}Fisheries and Oceans Canada, Québec; ²Université du Québec à Rimouski; ³GeneChem Management Inc., Montréal.

The impact of fishing as a force of sexual selection is not well understood in crustaceans. Fishing must be viewed as acting in conjunction with, or in opposition to, natural factors which also modify the context for sexual competition and mate choice. We review knowledge of the polygynandrous mating system of the snow crab and evaluate the likely interplay between natural and fishing forces in the process of sexual selection. Snow crab has determinate growth and two female reproductive stages (primiparous, multiparous) with discrete mating seasons. Temperature shifts the spectrum of size-at-maturity in both sexes and determines female reproductive tempo, thereby altering sperm supply, egg production and operational sex ratio. “Boom-and-bust” population dynamics modulate the phenotype of receptive individuals and the direction and intensity of sexual competition and mate choice. Fishing directed only at large males may attenuate or exacerbate some aspects of sexual conflict at primiparous mating, depending on the natural context, but otherwise it consistently promotes mating of small mature males, reduces opportunity for female mate choice and increases the likelihood of sperm limitation. These changes have mixed but still incompletely appreciated effects on female fitness. The long-term potential for selection against large size-at-maturity remains uncertain.

C. M. St. Mary and H. K. Kindsvater. University of Florida.

Life history theory predicts that allocation of energy to growth, reproduction, and to offspring size and number depends on maximizing lifetime expected fitness; that is, number of grandchildren. The optimal strategy will vary under different types of selective pressure. However, existing theory is static: it does not account for the role of variation in maternal quality and offspring quality, although such variation has recently been documented in a variety of fishes in several families. We generate a simulated population of individual fish according to the predicted constraints, and explore the optimal life history strategy in a dynamic environment, using a genetic algorithm approach. We are able to evaluate the optimal strategy of allocation to growth, offspring number, and offspring size under a range of selective environments, focusing on the role of predation and/or fishing pressure. In some cases, our simulations converge to the expected result: for instance, uncertain adult survival results in an earlier age at first reproduction. However, there is more variation within individuals and within populations in the optimal phenotypes than is seen in wild fish populations, suggesting that a more comprehensive approach is needed to address all the forces shaping life history strategies.

William H. Satterthwaite. University of California, Santa Cruz.

Steelhead (Oncorhynchus mykiss) display diverse life histories; fish are potentially anadromous or nonanadromous and vary in their time to smolting and/or maturity. Steelhead in central California face very different environmental conditions in cooler, relatively undammed coastal streams than in warmer, heavily regulated streams in the Central Valley. In an attempt to predict smolting and maturity patterns in steelhead and how they differ in coastal and valley streams, I have adopted the framework developed by Thorpe et al. (1998) predicting smolting and maturity in salmonids as a function of projected growth over time windows prior to the actual smolting or maturation process. Based on literature searches, I have identified the approximate timing of these life history windows for steelhead and obtained estimates of growth, mortality, and fecundity in both stream systems. I used stochastic dynamic programming to predict optimal thresholds for smolting and maturity for fish in the different stream systems, and generated additional fitness estimates for fish using nonoptimal thresholds. I will compare my model results with empirical results to assess the degree to which the smolting and maturity patterns in these populations are optimal in the current, human-affected state of these systems and the likely impacts of changes in water flow.

David H. Secor¹ and Richard Kraus.^{2 1}University of Maryland; ²George Mason University.

A modern theory that has application to the recruitment problem and life cycles is the storage effect, whereby spawning stock biomass accumulates each year in long lived species so that when early survival conditions are favorable, stored egg production can result in explosive population growth. Contingent structure—dispersive and retentive modalities in life cycles within populations—is an attribute that contributes to the storage effect. Here, a nursery habitat associated with one contingent behavior may make a small contribution in a given year, but over a decade contribute significantly to spawning stock biomass. Based upon retrospective microchemical analysis of otoliths, we apportioned white perch contingent structure associated with fresh water and brackish nursery habitats in the Patuxent River estuary, Maryland over an 11 year period. The dispersive contingent was associated with brackish nursery habitats, high flow years, and dominant year-classes. The retentive freshwater contingent predominated in drought conditions and was associated with weak year-classes. Thus, under extended drought cycles, the minority contingent behavior is predicted to contribute to resiliency. The dominant role of climate on recruitment in many exploited marine fishes highlights the importance in conserving contingent structure as a population attribute.

Kate I. Siegfried. University of California, Santa Cruz.

Calculating natural mortality, /M/, for long-lived fishes is often difficult. We rarely have a data set that is long enough to derive the parameter directly, in which case we depend on established models for estimating /M/ that require life history or length data. A prime example of this dilemma is the recent stock assessment of California sheephead, Semicossyphus pulcher.

California sheephead are a protogynous (female to male) sequential hermaphroditic species found in shallow, temperate waters from Monterey Bay to Cabo San Lucas, Mexico. There are active commercial (live-fish) and recreational fisheries for sheephead in Californian waters. The first stock assessment was undertaken by the California Department of Fish and Game in 2005, aided by researchers on the UC Santa Cruz campus. The researchers used Hoenig's method to estimate /M/ proportional to the maximum age (53 years).

Since there are other models available to calculate /M/, I will compare those model results using data from the sheephead commercial fishery. I determine estimates of /M/, using weight and life history-based methods. I find that the estimate of /M/ converges to a constant if the fish recruit to the model after age two. Therefore it may be reasonable, under certain assumptions, to use a constant natural mortality for California sheephead.

J. D. Simons,¹ T. J. Shirley,² J. Wood,² J. Lester,³ S. Glenn,³ L. Gonzalez,³ J. Ditty,⁴ J. E. Smith,⁵ K. Withers,⁶ and M. E. Vega.^{7 1}Texas Parks and Wildlife Department; ²Harte Research Institute, Corpus Christi, Texas; ³Houston Advanced Research Center, The Woodlands, Texas; ⁴National Marine Fisheries Service, Galveston; ⁵Texas State Aquarium, Corpus Christi, Texas; ⁶Texas A&M University-Corpus Christi, ⁷CINVESTAV-IPN, Merida, Yucatan, Mexico.

With the ever increasing computing power being developed and the increasing volumes of data being collected, a new era of gigantic databases has been ushered in. Projects such as the Census of Marine Life (CoML), National Biological Information Infrastructure (NBII), WhyWhere?, BioGeomancer, FishBase, and others are capturing taxonomic and biological data into, in most cases, spatially explicit databases. We report on a proposed project to build a temporally and spatially explicit web-based marine and estuarine trophic database for the Gulf of Mexico. It will include food-habit data for marine mammals, sea turtles, sea and shore birds, fishes, crustaceans, ctenophores, and jellyfish extracted from published and unpublished literature resources. This database will provide data for the analysis of spatial and temporal trends in the trophodynamics of the Gulf of Mexico. It will provide a rich database for theoretical investigations into food-web structure by groups such as Webs on the Web. These data will be made available to fisheries and food web modelers using Ecopath and other modeling projects. Techniques will be developed to make these data available to the general public through organizations such as the Texas State Aquarium and other education and outreach groups.

Martin D. Smith. Duke University.

Life history models of exploited fish populations typically treat fishing mortality as an exogenous parameter. Implicitly, this approach assumes that the supply of fishing effort is perfectly inelastic. That is, the supply curve of effort is vertical. Fishery modelers often run simulations for different values of fishing mortality, but his exercise also assumes vertical supply and simply explores a series of these curves as different scenarios. The seemingly innocuous assumption of vertical supply conflicts with a large body of empirical work on behavior of fishermen and fishing fleets. Economists and fisheries scientists consistently find that fishing behavior is responsive to economic opportunities over time and space as well as across target species. Accounting for this phenomenon requires that fishing mortality be made endogenous. This paper explores approaches to endogenizing fishing mortality in life history models by allowing the fish stock in the previous period and other behavioral drivers to enters into the equation that predicts fishing effort in the next period. The paper examines conditions under which the standard approach is approximately accurate and when endogenous fishing mortality dramatically alters model predictions. Accounting for fishing behavior ultimately will help to improve predictions from management models and avoid fisheries management failures.

Melissa Snover, Pacific Islands Fisheries Science Center, Hawaii.

Many marine animals must increase several orders of magnitude in size as they grow from egg or larvae to adults and ecological scaling properties limit the size range over which certain life style are exploitable. Hence, many organisms undergo one or several ontogenetic habitat shifts as they grow to maximize growth rates while minimizing predation risk. An understanding of the mechanisms that drive these shifts is critical in managing both target and bycatch populations impacted by fisheries. Fundamental differences between habitats impacting survival and growth rates include food availability, both total density and individual availability (e.g. gape limits or jaw crushing strength), temperature, and predation intensity. I developed a model building on previous theoretical studies of ontogenetic habitat shifts that considers each of these factors with the goal of teasing apart how these top-down and bottom-up influences may interact to affect the timing, and the variability in timing, of ontogenetic habitat shifts. While the model is very general in applicability, as a case study, I look at differences in the timing of the shift from pelagic to neritic habitats in populations of the loggerhead sea turtle (Caretta caretta), which are vulnerable to bycatch in different fisheries in the different habitats.

David R. Swank¹ and Edward S. Rutherford.^{2 1}University of California, Santa Cruz; ²University of Michigan.

To determine how life-history traits of fish populations may be expected to vary over time in response to environmental changes, I examined temporal variation in life-history traits of five wild Great Lakes steelhead populations from historic (1950s, 1960s, and 1980s) and recent (1997–2000) samples taken from past scale collections, new population sampling, and the literature, all collected from steelhead spawning runs in northern Michigan streams. I also analyzed a long-term (22-year) data set from one of these wild steelhead populations for significant temporal changes in life-history traits. Significant temporal variation in stream age structure, female age at maturity, and adult mortality was found within most steelhead populations. Female age at maturity increased in four populations, while adult mortality decreased in all five populations. There were no changes in growth rates for four populations. Female age at maturity was negatively correlated with adult mortality, as measured by percent repeat spawners. The Black River steelhead population had significant changes in female age at first maturity, mean length at age, and adult survival that were related to a large decrease in sea-lamprey predation since the 1950s. Variation in fall run size was considerable and was positively related to mean September stream flow.

Jill H. Swasey,¹ Jennie M. Harrington,¹ and Andrew A. Rosenberg.^{2 1}MRAG Americas, Inc. Essex, MA; ²University of New Hampshire.

In this paper we will explore the interaction between fishing pressure and life history characteristics in the application of rebuilding strategies for U.S. fisheries. In most situations, an overfished stock has been exploited beyond an explicit limit that would ensure a level of safe reproduction. The Magnuson-Stevens Act mandates that overfished fishery resources must be rebuilt within a specified time period to the biomass level that will support maximum sustainable yield, and be sustainable over the long term. This time period for rebuilding should be as short as possible, not to exceed ten years except for biologically compelling reasons, often dictated by life history characteristics. In those cases, timelines have often been set for much longer than 10 years, up to 90 years. Typically, fisheries comprised of slow-growing species will have longer rebuilding timelines. However, the Councils and NMFS often apply the longest timelines allowed which slows the process of recovery. Of particular interest are the situations in which overfishing continues into rebuilding and the possibility that excessive fishing mortality rates are more affected by what is happening in the fishery than the actual life history (i.e. slow growth) of the stock.

Carl Walters,¹ Steven J. D. Martell,¹ and Behzad Mahmoudi.^{2 1}University of British Columbia, Vancouver; ²Florida Marine Research Institute.

An Ecopath-Ecosim ecosystem model under development for coastal areas of the Gulf of Mexico simulates responses of 63 biomass pools to changes in fisheries and primary productivity. 10 key species are represented by detailed, multistanza population dynamics models (31 of the biomass pools) that attempt to explicitly account for possible changes in recruitment rates due to changes in bycatch rates and trophic interactions. Over a 1950-2004 historical reference period, the model shows good simulated agreement with time series patterns estimated from stock assessment and relative abundance index data for many of the species, and in particular offers an explanation for apparent nonstationarity in natural mortality rates of menhaden (declining apparent M over time). It makes one highly counterintuitive policy prediction about impacts of management efforts aimed at reducing bycatch in the shrimp trawl fishery, namely that bycatch reduction may cause negative impacts on productivity of several valued species (menhaden Brevoortia patronus, red drum Sciaenops ocellatus, red snapper Lutjanus campechanus) by allowing recovery of some benthic predators such as catfishes that have been impacted by trawling but are also potentially important predators on juveniles of the valued species. Recognition of this policy implication would have been impossible without explicit, multistanza representation of juvenile life histories and trophic interactions, since the predicted changes in predation regimes represent only very small overall biomass fluxes.

Sarah Walters,¹ Susan K. Lowerre-Barbieri,¹ Joel Bickford,¹ and David Mann.^{2 1}Florida Fish and Wildlife Research Institute; ²University of South Florida.

Evaluating essential spawning habitat and its usage are important aspects of fisheries management. Classifying these habitats can be challenging as the definition of habitat itself is multifaceted, involving physical location, environmental parameters, and/or the interaction between the two. Spatial and temporal variations must also be accounted for as spawning habitat is not necessarily static. A comprehensive, multiple year research study was conducted in Tampa Bay, Florida examining spawning habitat of spotted seatrout (Cynoscion nebulosus). Spawning habitat of these estuarine, multiple-batch spawning fish was evaluated using passive acoustics, an effective and noninvasive methodology involving permanent as well as mobile hydrophones. As spotted seatrout males produce sounds associated with reproduction, spawning aggregations were easily located with hydrophones. Areas where aggregations were detected were compared to areas from which spawning was absent to determine factors potentially driving spawning-site selection such as bottom type, water quality, and current. To evaluate temporal variation in spawning-habitat use, one spawning site was continually monitored for five years to assess diel, seasonal, lunar, and tidal periodicities. Both of these long-term spawning habitat studies contribute to our understanding of what spawning habitat is and how it can shift over time and space.

Robert Warner. University of California, Santa Barbara.

Evidence from a variety of sources suggests two important features of marine larval dispersal that can affect management and also may be an important source of selection on life-history characteristics. First, larval dispersal distances appear to be much shorter than previously thought, and include substantial amounts of self-recruitment. Second, larval recruitment into coastal habitats appears to be intermittent and heterogeneous on annual time scales, driven by advection in turbulent coastal circulation. The stochastic nature of larval transport in particular will create unavoidable uncertainty that complicates the management of nearshore ecosystems.

Short dispersal paths and self-recruitment suggest that management may be more effective if carried out on a local scale, and that the effects of spatial management approaches may be limited in their extent. The pulsed aspect of recruitment, even at long distances from a source, may alleviate the Allee effects that limit the success of long-distance colonization. High variation in recruitment, especially occasional large pulses of recruitment, may enhance the contribution of the storage effect on species persistence and coexistence. On the other hand, local rates of larval settlement may be largely decoupled from local stock abundances, even if self-recruitment is substantial. This provides an unexplored source of uncertainty in stock-recruitment relationships. Finally, the stochastic nature of connectivity may make it difficult to assess the effects of spatial fisheries management policies, because it is likely to take long periods of sampling in order to detect recruitment responses to a management change.

Models suggest that the stochastic nature of successful settlement could exert strong selection on life histories, leading to extreme iteroparity and fine partitioning of reproductive effort. Thus the long lives and high fecundity of many marine organisms may be an evolutionary response to coastal ocean circulation.

John Wiedenmann. University of California, Santa Cruz.

An overfished stock is typically considered rebuilt when the population biomass or spawning biomass cross some threshold, often (but not always) the population size that produces maximum sustainable yield (B_MSY or SSB_MSY). In cases where rebuilding is the result of a single strong year class, and if that year class does not produce a large number of recruits, then the population will likely drop below the rebuilding threshold as the cohort ages and moves out of the population. I model the effects of a skewed age-distribution (relative to the stable age-distribution) on the probability of long-term recovery for rebuilt populations with different life histories. For the model populations, having a skewed age-distribution does affect the probability of long-term recovery, but how it affects them depends not only on magnitude and direction of skew, but also on the life history of the population. Therefore, effective management of rebuilt stocks might benefit from considerations of the population age structure.

J. Woodland and D.H. Secor. University of Maryland Center for Environmental Science.

In comparison to coastal species, anadromous fishes such as sturgeons are highly sensitive to alterations in nursery habitat volumes due to small nursery size and its proximity to sources of anthropogenic degradation. Here, we present a case study where geometric growth of a population of endangered sturgeon coincided with a two-fold increase in habitat nursery volume. Shortnose sturgeon (Acipenser brevirostrum), a U.S. Endangered Species, experienced a four-fold increase in abundance in the Hudson River during the period 1985-2000. Life table elasticity analysis indicated that such population growth can only occur through large changes to first year survivorship. We evaluated the hypothesis that improvements in water quality during the 1970’s stimulated population recovery by increasing nursery habitat volume. Hindcast year-class strengths, estimated by age structure of the extant population (corrected for gear selectivity and cumulative mortality), indicated high recruitments (31,000-52,000 yearlings) during 1986-1992. This period was preceded and succeeded by c. 5 year-periods of lower recruitment (6,000-17,500 yearlings); trends corroborated by shortnose sturgeon bycatch from a beam trawl survey. This evidence supports the view that improved nursery volumes can promote species recovery in endangered coastal species.