Biomass Dynamic Model Research Articles

The use of the swept area method for assessing the seabob shrimp Xiphopenaeus kroyeri (Heller, 1862) biomass and removal rates based on artisanal fishery-derived data in southern Brazil: using depletion models to reduce uncertainty

The seabob shrimp (Xiphopenaeus kroyeri) represents an important fishing resource for ar- tisanal fishermen in coastal areas of southern Brazil. Stock assessments of this species have generally relied on biomass dynamics models as applied to CPUE time-series, which (a) are only available for a small offshore fraction of the exploited population and (b) does not comprise patterns of the shallowest artisanal fishing grounds. This work explores the use of extensive catch and effort data derived from a small-scale trawl fishery to obtain swept-area estimates of abundance and removal rates in a limited coastal area of southern Brazil (Ti- jucas Bay, Santa Catarina State, 27 o 15'S-48 o 33'W). Data were obtained from 7,198 fishing trips monitored at the fishing communities between June 2004 and August 2005. Because three parameters of the swept-area equation (i.e. trawl velocity, catch efficiency and wing spread) were unknown, they were defined through a stochastic procedure and calibrated by estimates produced by a Leslie depletion model applied to concurrent catches obtained in one fishing ground. A 21.7% removal rate was estimated for the period June 2004-January 2005; this increased to nearly 34% between February and July 2005. This removal scenario predicted that a five-month fishery would suffice to remove 90% of the biomass available in the Tijucas Bay, nearing the 87% CPUE reduction observed in the same period. Whereas abundance and harvest rate estimates were likely af- fected by inadequate knowledge of the swept-area equation parameters, the similarity of these estimates with relative abundance indexes supports the convenience of the proposed method and justifies future efforts to im- prove its accuracy.

The relationship between stream periphyton dynamics and near-bed turbulence in rough open-channel flow

The main goal of the present work is to characterize, at the laboratory scale, the physical relationships between periphyton biomass and water velocity. Experiments, conducted in a rough open-channel, and numerical simulations, have been performed for a large range of values of the main hydrodynamic parameters. The major role of the shear stress exerted by the flow on the bed, in the periphyton dynamics, is clearly proved. We have developed a model of periphyton biomass dynamics taking the friction velocity, as external physical variable. We fitted the model to our laboratory data and, despite their poorness, the model has shown a good fit, clearly showing that local physical processes are the main factors controlling the dynamics in running waters.

Management strategies for short lived species: The case of Australia's Northern Prawn Fishery: 2. Choosing appropriate management strategies using input controls

A Management Strategy Evaluation framework is used to evaluate management strategies based on input controls for the fishery for two tiger prawn species ( Penaeus esculentus and Penaeus semisulcatus) in Australia's Northern Prawn Fishery. Three “assessment procedures” are considered and two forms of decision rule. The performance of the management strategies is evaluated in terms of whether stocks are left at (or above) the spawning stock size at which Maximum Sustainable Yield is achieved ( S MSY), the long-term discounted total catch and the extent of inter-annual variation in catches. The focus of the analysis is on management strategies based on the current method of stock assessment because an alternative method of assessment based on a biomass dynamics model is found to be highly variable. None of the management strategies tested is able to leave the spawning stock size of P. esculentus near S MSY if the target effort level used in the management strategy is set to E MSY. Accounting for stock structure through the application of a spatially- (stock-) structured assessment approach fails to resolve this problem. Since the assessment method is generally close to unbiased, the failure to leave the stocks close to S MSY is because the measure of control is total effort and the two species are found (and caught) together. Reducing the target effort level to below E MSY increases the final stock size, but the reduced risk comes at a cost of reduced catches. The best management strategy in terms of leaving both species close to S MSY is found to be one that changes the timing of the fishing season so that effort is shifted from P. esculentus to P. semisulcatus and sets more precautionary effort targets for P. esculentus.

Some refinements of the assessment of the South African squid resource, Loligo vulgaris reynaudii

A biomass dynamic model making use of an observation error estimator has been used over recent years to assess the status of the squid resource, Loligo vulgaris reynaudii, off the coast of South Africa. This assessment has indicated the resource to be at a low level and the risk of collapse under the current level of fishing effort to be high. Accordingly, past scientific management advice has been that effort should be reduced. This assessment is refined here to take account of both observation and process error, and a number of sensitivity tests related to assumptions made in the original model are conducted. Bayesian analyses to take full account of estimation uncertainty are also undertaken. The results from the respective analyses are reported, and it is shown that changing the “hockey-stick” stock–recruitment function of the original model to a Beverton–Holt form opens the possibility that the resource is under less risk from the current level of effort than has been supposed.

A framework to model shrimp (Pandalus borealis) stock dynamics and to quantify the risk associated with alternative management options, using Bayesian methods

Abstract A new integrated Bayesian framework for making quantitative assessments, predictions, and risk analyses of shrimp (Pandalus borealis) stock development is constructed. A biomass dynamic model, based on the logistic function but including an explicit term for cod predation, suggests that the quantity of shrimp consumed by cod could equal that taken by the fishery. The model proved superior to an alternative model in its ability to estimate parameters central to the assessment; the alternative model subsumed cod predation as part of an overall population growth effect without a time trend. Two series of shrimp biomass indices, catch, cod biomass estimates, cod consumption estimates, and prior distributions of model parameters provided information to the models. Process and observation errors were incorporated simultaneously using a state-space modelling framework. A Bayesian approach was used to construct posterior probability distributions of model parameters and derived variables relevant for management advice, including quantification of future risk of transgressing reference points in relation to alternative management options.

Applying game theory to analyze the influence of biological characteristics on fishers’ cooperation in fisheries co-management

Abstract The concept of mixed strategies from game theory is combined with basic economic concepts and a biomass dynamics model to capture important aspects of a fishery cooperative system. The concept of discounting, which is assumed to be a function of the intrinsic growth rate of the fish stock, is applied to mimic the fishers’ tendency to cooperate. Profits from fishing, discounted by the tendency to cooperate, are specified for each fisher under various circumstances and viewed as a two by two matrix game, with two players and two strategies (cooperative and non-cooperative). The payoffs in the matrix determine the equilibrium outcome for the game. The results indicate that the biological parameters, intrinsic growth rate and stock size, can have important influence on the fishers’ cooperation.

Is pelagic top-down control in lakes augmented by benthic energy pathways?

Modern food web studies are typically conducted from a trophic dynamic perspective that focuses on combined roles of top-down and bottom-up forces in regulating food web structure. Recognition of spatial food web subsidies in diverse ecosystems highlights the importance of energy flow as a foundation for understanding trophic dynamics. Here, we consider how different energy flow configurations might affect trophic dynamics in north-temperate lakes. A literature review revealed that littoral piscivores exert top-down control on prey fishes. In contrast, analysis of littoral predator diets indicated extensive omnivory and heavy reliance on zoobenthic prey. We explored this uncoupling between trophic dynamics (piscivores regulate prey fish) and energy flow (zoobenthos in piscivore diets) using a biomass dynamic model. This model compared top-down impacts of a piscivore on prey fishes under two scenarios: consumption of prey fish only and consumption of prey fish plus zoobenthos. The model predicted that elimination of zoobenthivory leads to a 50% reduction in piscivore standing stock and concomitant 2.5-fold increase in prey fish abundance (i.e., zoobenthivory plays a key role in mediating pelagic top-down control). These results highlight the role of benthic–pelagic linkages in regulating trophic dynamics and underscore the value of whole-ecosystem approaches to the study of food webs.

Interactions of productivity, predation risk, and fishing effort in the efficacy of marine protected areas for the central Pacific

Ecospace biomass-dynamics models for the central North Pacific predict strong space-time variation in abundances of various trophic groups in relation to nutrient-driven patterns in primary production and circulation-driven concentration of production in convergence areas. The model predicts simple patterns in ecosystem organization and abundances along productivity gradients. Predicted patterns are robust to alternative assumptions about how mobile organisms may alter dispersal behavior in relation to local fitness (per capita gain from net food intake minus predation mortality). Large marine protected areas (MPAs) would be needed to rebuild endangered populations and counter impacts of growing fishing effort. We expected that increases in dispersal rate in response to lower food availability and higher predation risk would reduce efficacy of MPAs as a management tool. Instead, simulations indicated that this negative effect may generally be eliminated or even reversed by positive effects of fitness-maximizing behaviors. Interannual variability in ocean circulation can further reduce the efficacy of MPAs; therefore, the dynamic nature of pelagic environments should also be considered in MPA design. Anomalies in ocean circulation may shift productive areas relative to MPA location, resulting in increased fishing mortality and (or) misinterpretation of catch statistics.

Use of state–space modeling with a Bayesian approach to estimate target reference points for green sea urchin (Strongylocentrotus droebachiensis) stocks in the Queen Charlotte Strait region, British Columbia, Canada

Abstract This study develops a state–space modeling and Bayesian approach to a biomass dynamic model of green sea urchins ( Strongylocentrotus droebachiensis ) in British Columbia, Canada. The objective is to determine limit and target reference points for harvest quota options for this fishery, and to assess the degree of risk associated with these target reference points (in which risk is defined as the probability that the target reference point will include the limit reference point when all uncertainties are considered). These uncertainties include random variability both in catch per unit effort and in annual biomass dynamics, and uncertainties in catch and effort data in the developing years of this fishery. In addition, the approach uses fishery-independent survey data as auxiliary information to index green urchin population abundance. Compared with the traditional methods of fitting the biomass dynamic model, this new approach incorporates more realistic error structures. The resulting probability distribution for the limit reference point (maximum sustainable yield, MSY) is likely to reflect the true uncertainties about MSY more closely, and can be used to measure the risk of exceeding MSY for setting fishing quotas for this green urchin stock. It is concluded that the current practice of selecting target reference points for this population at 25–50% of MSY has a probability of including the limit reference point (MSY) of

A hierarchical approach to determining reference points for Pandalid shrimp

Reference points for harvesting Pandalid shrimp are categorized into five general approaches: historical proxies, biomass dynamics models, dynamic pool models, stock–recruit models, and demographic production models. Each of these approaches has different data requirements and underlying assumptions. Estimation of biological reference points from these methods can be viewed as a hierarchy, using data-poor proxies in the lowest tier to applying more informative demographic production models in the highest. Based on a review of Pandalid life histories, precautionary-approach reference points, and methodologies for estimating reference points and their applications to Pandalid shrimp stocks, we advocate a progression from proxies to more informative models and the requisite advancement of research programs to develop reliable reference points for Pandalid shrimp stocks.

Methods for stock assessment of crustacean fisheries

Crustacean fisheries account for nearly 30% of all fish and shellfish landings by value world-wide. This paper reviews and evaluates the application of various stock assessment methods to crustacean fisheries. Methods covered include biomass dynamics models, delay-difference models, depletion methods, equilibrium yield and egg per recruit models and dynamic size-structured models. The development, underlying assumptions, data requirements, benefits and disadvantages of each model are briefly considered. The implications and treatment of uncertainty in data, parameter estimates and model structure are also considered. Finally, we speculate how crustacean stock assessment methodology may evolve in the future.

A biomass dynamics assessment of the southeastern Caribbean snapper–grouper fishery

Southern red snapper, Lutjanus purpureus, and yellowedge grouper, Epinephelus flavolimbatus, are the main species exploited by handliners and longliners of the Venezuelan medium range snapper–grouper fleet in the southeastern Caribbean. Catch and effort data for the period 1981–2000 are used to assess both species by means of a Schaefer biomass dynamic model. Due to lack of contrast in both time series empirical estimates of the intrinsic rate of population increase ( r) were used as auxiliary information for fitting the model to the data. Model parameters and derived variables uncertainty was estimated using a conditioned non-parametric bootstrap with fixed r values and normally distributed r values with estimated coefficients of variation of 55% for yellowedge grouper and 42% for red snapper. Also, 10-year biomass projections under different catch scenarios for both species were used to assess the impact of fishing on future population levels. Results indicate that L. purpureus is more intensely exploited than E. flavolimbatus and biomass levels in the year 2000, depending on the initial biomass B i hypothesis, were on average between 22% ( B i=0.5 K) and 39% ( B i=K) and 13% ( B i=0.5 K) and 25% ( B i= K) of virgin biomass ( K) for E. flavolimbatus and L. purpureus, respectively. Significant population recoveries are possible if catches are maintained at relatively low levels for both species, but biomass recovery to MSY levels depending on species and scenarios may require a longer time frame than the 10-year period used for projections.

Applying biomass dynamic models to the southern horse mackerel stock (Atlantic waters of Iberian Peninsula). A comparison with VPA-based methods

The horse mackerel (Trachurus trachurus) fishery of the Northeast Atlantic, is currently subject to assessment and management in the ICES area. Assessment of the Southern horse mackerel (ICES Divisions VIIIc and IXa) is made with Virtual Population Analysis (VPA), using time series of catch-at-age data and catch-per-unit-of-effort (CPUE) from 1985 to present. The application of biomass-dynamic models to the assessment and catch prediction of this stock has never been attempted before. In this paper, a production model is applied to the Southern horse mackerel stock. To quantify uncertainty in parameter estimates, bootstrap confidence intervals were computed, which showed that estimates could be considered as reliable. The bootstrap standard deviations of Ft, r, MSY and FMSY were relatively small, despite the lack of trends in the available effort series. On the contrary, the catchability, q, showed a coefficient of variation of 74%. The current level of fishing mortality for 1998 was considered as unsustainable for the resource, being well above FMSY according to the biomass-dynamic models, and above Fpa according to the age-structured model. Both models were in close agreement on the time series of fishing mortality and in the perception of the state of the stock. Differences existed in the changes in stock biomass especially over the last years. Over this time period the age-structured model showed an increasing trend in biomass. The estimates of MSY and FMSY were in accordance with the precautionary approach philosophy. The biomass-dynamic model used here was usefully applied to the Southern horse mackerel stock, giving complementary information to the age-structured model, both in the perception of the state of the stock and in the definition of management targets.

A new approach of fitting biomass dynamics models to data

A non-traditional approach of fitting dynamic resource biomass models to data is developed in this paper. A variational adjoint technique is used for dynamic parameter estimation. In the variational formulation, a cost function measuring the distance between the model solution and the observations is minimized. The data assimilation method provides a novel and computationally efficient procedure for combining all available information, i.e., the data and the model in the analysis of a resource system. This technique will be used to analyze data for the North-east Arctic cod stock. Two alternative population growth models: the logistic and the Gompertz model are used for estimating parameters of simple bioeconomic models by the method of constrained least squares. Estimates of the parameters of the models dynamics are reasonable and can be accepted. The main inference from the work is that the average fishing mortality is found to be significantly above the maximum sustainable yield value.

A MODEL FOR THE BIOECONOMIC EVALUATION OF MARINE PROTECTED AREA SIZE AND PLACEMENT IN THE NORTH SEA

ABSTRACT. The use of marine protected areas (MPAs) as a basic management tool to limit exploitation rates in marine fisheries has been widely suggested. Models are important in predicting the consequences of management decisions and the design of monitoring programs in terms of policy goals. However, few tools are available that consider both multiple fleets and ecosystem scale dynamics. We use a new applied game theory tool, Ecoseed, that operates within a temporally and spatially explicit biomass dynamics model, Ecopath with Ecosim, to evaluate the efficacy of marine protected areas in the North Sea in both ecological and economic terms. The Ecoseed model builds MPAs based on the change in values of predicted economic rents of fisheries and the existence value of biomass pools in the ecosystem. We consider the market values of four fisheries operating in the North Sea: a trawl fishery, a gill net fishery, a seine fishery, and an industrial (reduction) fishery. We apply existence values, scaled such that their aggregate is similar to the total fishery value, to six biomass pools of concern: juvenile cod, haddock, whiting, saithe, seals, and the collective pool ‘Other predators’ that include marine mammals. Four policy options were considered: to maximize the rent only; to maximize the existence values only; to maximize the sum of the rent and existence values; and, finally, to maximize the sum of the rent and the existence values, but excluding only the trawl fleet from the MPA. The Ecoseed model suggests that policy goals that do not include ecological considerations can negatively impact the rents obtained by the different fishing sectors. The existence values will also be negatively impacted unless the MPA is very large. The Ecoseed model also suggests that policy goals based solely on existence values will negatively impact most fisheries. Under policy options that included ecological considerations, maximum benefits were derived from an MPA that covered 25–40% of the North Sea, placed along the southern and eastern coasts. Finally, the Ecoseed model suggests that an exclusion of the trawl fishery only from the MPA can provide small‐to‐substantial positive impacts to most species and fleets; this relative impact depends on level of interaction between the trawl fleet and the other fleets target species (e.g., through bycatch).

Least median of squares: a suitable objective function for stock assessment models?

Robust fitting methods, intended for data sets possibly contaminated with invalid observations, are gaining increased use in analysis of fishery data. In particular, the method of least median of squares (LMS) has attracted attention. Its hallmark is high statistical resistance, which makes it immune to up to 50% contamination in the data. However, the same property makes it inefficient and can cause faulty fitting of typical fishery data. The LMS fit can be in conflict with important sections of a time series, a problem we illustrate by fitting a biomass dynamic (surplus production) model to simulated and actual fishery data. Additionally, we illustrate that LMS parameter estimates can be highly sensitive to small perturbations in the data. Other robust methods, like the method of least absolute values (LAV), appear less prone to such problems. A key reference on LMS recommends using the method as part of an exploratory procedure to identify outliers, rather than as an objective function for final model fitting. We concur with that recommendation.

USING A FISHERY-INDEPENDENT SURVEY TO ASSESS THE STATUS OF A SPANNER CRAB RANINA RANINA FISHERY: UNIVARIATE ANALYSES AND BIOMASS MODELLING

A fishery-independent survey of the spanner crab, Ranina ranina fishing grounds off the coast of New South Wales, Australia was completed in 1988-89 and repeated in 1997-98. This survey pro vided an empirical assessment of this stock and allowed the development of a population model. The results showed the marked sexual dimorphism of this species with males generally attaining larger sizes than females and a sex ratio that was slightly biased towards females. Seventy-one to 73% of the population was found to be under the legal minimum size, male crabs comprised most of the portion of the population that was targeted by fishers and 96% of females were protected by the size limit. Comparing catch rates from the two surveys revealed large variabilities in abundances across sev eral of the spatial and temporal scales investigated. Catches of crabs (especially females) decreased during the spawning period each year (October to December). In one location (Tallows Beach) these decreases did not occur, suggesting that this site was the focus of a spawning migration. A significant 6% decrease in the relative abundances of all crabs over the 11 year period of the study and a 55% decline in the numbers of large male crabs (> 108 mm CL) provided some cause for concern. The stock was represented with a biomass dynamic model conditioned on catch and fitted to the results of the fishery-independent surveys and/or commercial catch per unit effort using non-linear optimization. The history of catches showed an increasing rate of exploitation and a decreasing index of abundance (i.e., a one-way trip) and therefore we were unable to estimate the exploitable biomass of this stock. However, by assuming a series of growth rate scenarios of the exploitable stock, we estimated patterns of relative exploitation and stock productivity. Confidence intervals for the parameter estimates were determined using bootstrap methods. Recommending a Total Allowable Commercial Catch (TACC) for this fishery from this study is difficult because estimates of the extent to which the stock has been fished down depend upon the dataset used for the analysis and the assumed growth rate of the stock. Quantifying a TACC also requires assumptions about the objectives for the stock and the certainty of achieving that objective. The NSW spanner crab fishery must continue to operate with uncertainty about its sustainable harvest rate. Fishery-independent surveys and flexibility in the TACC will be the most effective strategies to ensure that no adverse biological, economic or social outcomes occur in this fishery.

Dynamic catch-effort model for brown shrimp Farfantepenaeus californiensis (holmes) from the Gulf of california, Mexico

We analyzed the catch and effort data for the brown shrimp (Farfantepenaeus californiensis) fishery over 22 years. We used a biomass dynamic model in a stochastic version to analyze the catch-per-unit effort of the trawl fishery in the Gulf of California, Mexico. Two hypotheses were proposed: the first (observation error) assumes that the catch-per-unit effort (CPUE) is measured with error, while in the second (process error), the variability of the population is produced by factors not included in the model. The results show that the hypothesis of the process error was accepted. In this case, the hypothesis of the process error recognizes several sources of perturbation in the model. For example, changes in the size structure of the shrimp population, changes in the natural mortality, variations in the fecundity or recruitment, and changes in the environment. Evidence of the relation between the environment and the shrimp population is discussed, as is the possibility of proposing a biological hypothesis for the trends of the population and the CPUE.

Representing Density Dependent Consequences of Life History Strategies in Aquatic Ecosystems: EcoSim II

EcoSim II uses results from the Ecopath procedure for trophic mass-balance analysis to define biomass dynamics models for predicting temporal change in exploited ecosystems. Key populations can be represented in further detail by using delay-difference models to account for both biomass and numbers dynamics. A major problem revealed by linking the population and biomass dynamics models is in representation of population responses to changes in food supply; simple proportional growth and reproductive responses lead to unrealistic predictions of changes in mean body size with changes in fishing mortality. EcoSim II allows users to specify life history mechanisms to avoid such unrealistic predictions: animals may translate changes in feeding rate into changes in reproductive rather than growth rates, or they may translate changes in food availability into changes in foraging time that in turn affects predation risk. These options, along with model relationships for limits on prey availability caused by predation avoidance tactics, tend to cause strong compensatory responses in modeled populations. It is likely that such compensatory responses are responsible for our inability to find obvious correlations between interacting trophic components in fisheries time-series data. But Ecosim II does not just predict strong compensatory responses: it also suggests that large piscivores may be vulnerable to delayed recruitment collapses caused by increases in prey species that are in turn competitors/predators of juvenile piscivores.

Giant panda ( Ailuropoda melanoleuca) population dynamics and bamboo (subfamily Bambusoideae) life history: a structured population approach to examining carrying capacity when the prey are semelparous

The giant panda, Ailuropoda melanoleuca, is a highly specialized Ursid whose diet consists almost entirely of various species of bamboo. Bamboo (Bambusoideae) is a grass subfamily whose species often exhibit synchronous semelparity. Synchronous semelparity can create local drops in carrying capacity for the panda. We modeled the interaction of pandas and their bamboo food resources with an age structured panda population model linked to a natural history model of bamboo biomass dynamics based on literature values of bamboo biomass, and giant panda life history dynamics. This paper reports the results of our examination of the interaction between pandas and their bamboo food resource and its implications for panda conservation. In the model all panda populations were well below the carrying capacity of the habitat. The giant panda populations growth was most sensitive to changes in birth rates and removal of reproductive aged individuals. Periodic starvation that has been documented in conjunction with bamboo die-offs is probably related to the inability to move to other areas within the region where bamboo is still available. Based on the results of this model, giant panda conservation should concentrate on keeping breeding individuals in the wild, keep corridors to different bamboo species open to pandas, and to concentrate research on bamboo life history.