Evaluation of and insights from ALFISH: a spatially explicit, landscape-level simulation of fish populations in the Everglades

We present an evaluation of a spatially explicit, age-structured model created to assess fish density dynamics in the Florida Everglades area. This model, ALFISH, has been used to compare alternative management scenarios for the Florida Everglades region. This area is characterized by periodic dry downs and refloodings. ALFISH uses spatially explicit water depth data to predict patterns of fish density. Here we present a method for calibration of ALFISH, based on information concerning fish movement, pond locations and other field data. With the current information, the greatest coefficient of determination achieved from regressions of ALFISH output to field data is 0.35 for fish density and 0.88 for water depth. The poor predictability of fish density mirrors the empirical findings that hydrology, which is the main driver of the model, only accounts for 20–40% of the variance of fish densities across the Everglades landscape. Sensitivity analyses indicate that fish in this system are very sensitive to frequency, size and location of permanent ponds as well as availability of prey.     

Updated Risk/Benefit Analysis of Fish Consumption Effects on Neurodevelopment: Implications for Setting Advisories

Epidemiological studies indicate that fish consumption during pregnancy leads to improved neurodevelopment. This suggests that the beneficial nutrients in fish may offset the adverse effects of mercury in the case of the average fish meal. However, our previous risk/benefit model predicted a net neurodevelopmental risk for the majority of species analyzed. In this article the previous model is calibrated against fish benefit data and then compared to other fish risk/benefit models, including recent models from the World Health Organization and USFDA. Our calibrated model estimated greater benefit for low mercury species but greater risk for high mercury species than the other models. With respect to a commonly eaten high mercury fish, swordfish, the calibrated model yielded risks that are supportive of current fish advisories but, in contrast, the other models predicted net neurodevelopmental benefits. The calibrated model was used in a proposed 3 step framework for setting fish consumption advisories: (1) Set initial consumption level based upon mercury RfD; (2) Adjust consumption upward if risk/benefit model indicates a net benefit; (3) Cap fish consumption based upon saturation of O-3 benefit. The implications of this approach for 7 varieties of fish are used to illustrate the framework.     

Fish image recognition method based on multi-layer feature fusion convolutional network

Solving the problem of fish image classification is important to conserve fish diversity. This conundrum can be addressed by developing a new fish image classification method based on deep learning by training data with complex backgrounds. To this end, this paper proposes a fusion model, referred to as Tripmix-Net. The backbone network of the proposed model primarily consists of multiscale parallel and improved residual networks that are connected in an alternate manner, and network fusion is used to integrate the information that is extracted from shallow and deep layers. Experiments conducted on the 15-category WildFish fish image dataset verified the efficacy of the proposed Tripmix-Net for classifying same-genus fish images with complex backgrounds. The model achieved an accuracy of 95.31%, which is a significant improvement over traditional methods. The proposed approach serves as a new concept for the fine-grained image classification of fish against complex backgrounds.     

Modelling Production and Biomasses of Prey and Predatory Fish in Lakes

This work presents a new dynamic model to predict two fundamental functional categories of fish in lakes, prey and predatory fish. The model has been developed within the framework of a more comprehensive lake ecosystem model, LakeWeb, which also accounts for phytoplankton, bacterioplankton, two types of zooplankton (herbivorous and predatory), zoobenthos, macrophytes and benthic algae. The new fish model gives seasonal variations (the calculation time is 1 week). It is meant to account for all factors regulating the production of fish for lakes in general. The model has not been calibrated and tested in the traditional way using data from a few well investigated lakes. Instead, it has been tested using empirical data and regressions based on data from many lakes. The basic aim of the model is that it should capture typical functional and structural patterns in many lakes. It accounts in a relatively simple manner for many complicated processes, like fishing (by birds, animals and man), fish migration to and from lakes and how macrophyte cover gives shelter to small fish and reduces predation pressure. Food choices are handled by distribution coefficients regulating how much of the different available food sources a given organism would consume. Beside these distribution coefficients, and the way the food choices are structured (the food choice panel), fundamental concepts in the fish model are: (1) metabolic efficiency ratios, which express how much of the food consumed by the predator that will increase the biomass of the predator and how much that will be lost by respiration and faeces, (2) actual consumption rates, which are defined from the ratio between the actual biomass of the predator and the normal biomass of the predator, and the normal consumption rates, which are related to the turnover time of the predator. We have demonstrated that the new model gives predictions which agree well with the values given by the empirical regressions, and also expected and requested divergences from these regression lines when they do not provide sufficient resolution.     

Species abundance models and patterns in dragonfly communities: effects of fish predators

We investigated if dragonfly larvae community composition and species abundance curves are sensitive to variation in predation intensity, and whether the fit to a particular niche partitioning model could be used to make inferences about mechanisms structuring communities. The approach taken was to compare communities in lakes either having or lacking fish predation. Dragonfly species classified as active, strongly dominated the dragonfly communities in fishless lakes, and low active species dominated fishless lakes. As activity level is known to correlate with susceptibility to fish predation this indicates that these communities are structured by fish predation. Fitting relative abundance data to five niche partitioning models showed that the same model fitted data from both types of habitats (fish/no fish). This means that the observed differences in relative abundances were substitutive, i.e. the relative abundance of a rank stayed constant, even though the identity of the species having this rank changed. The best fit to data from both types of lakes was found for the random assortment model, which is usually interpreted as an indication that the community is not structured by within-guild interactions. This interpretation for fishless lakes did not seem to agree with other community measures (i.e. lowered diversity and evenness and no relationship between species richness and dragonfly biomass), which indicate that the community is structured by within-guild interactions. Moreover, a detail in the fitting procedure, the number of species included in the analysis, affected which model that fitted data best. Thus, we question if fitting niche partitioning models to data can provide mechanistic understanding of how resources are partitioned in natural communities.     

Chaos and Peak-to-Peak Dynamics in a Plankton–Fish Model

A plankton–fish model, comprising phosphorus, algae, zooplankton, and young fish, with light intensity and water temperature varying periodically with the seasons, is analyzed in this paper. For realistic values of the parameters the model behaves chaotically, but its dynamics within the strange attractor can be described by a few one-dimensional maps that allow one to forecast the next yearly peak of plankton or fish from the last peaks. This property is an unambiguous mark of a special form of chaos. Unfortunately, the estimate of such peak-to-peak maps from field data is possible only if plankton or young fish biomass has been sampled accurately and frequently for a paramount number of years. In conclusion, the analysis shows that it might be that plankton dynamics are characterized by an interesting and peculiar form of chaos, but that inferences from recorded data on the existence of these forms of chaos are premature.     

A general production model with dependence between data from multiple surveys

A fishery‐independent survey for stock assessments is made sometimes more than once per year to detect a difference in relative sizes of fish populations (e.g., catch‐per‐unit‐effort [CPUE]) in response to a seasonal change in fish spatial distributions. Many managers tended to treat such data independently instead of systematically synthesizing them. A primary objective of this study was to synthesize all survey data via a simple hierarchical structure. I used the general (Pella‐Tomlinson) surplus production model for the illustration, because the purpose of this study was not a stock assessment, and the model was simpler than an age‐structured model. The surplus production model has about an eight decade history (since Graham's paper in 1935) and has been prominent in fish population dynamics. The logistic (Graham‐Schaefer) version was useful in the sense of simplifying the dynamics of a fish population in relation to its intrinsic growth, natural mortality, recruitment, density‐dependence, and fishery catch, but it was criticized because of its unrealistic limitations. Subsequently, the general version was suggested to accommodate flexibility and be realistic. In this study, I inferred parameters in the general surplus production model, simultaneously synthesizing all available data even from different temporal ranges. I used Georges Bank yellowtail flounder (Limanda ferruginea) data for demonstration.     

Marine reserve spillover: Modelling from multiple data sources

The functional form of spillover, measured as a gradient of abundance of fish, may provide insight about processes that control the spatial distribution of fish inside and outside the MPA. In this study, we aimed to infer on spillover mechanism of Diplodus spp. (family Sparidae) from a Mediterranean MPA (Carry-le-Rouet, France) from visual censuses and artisanal fisheries data. From the existing literature, three potential functional forms of spillover such as a linear gradient, an exponential gradient and a logistic gradient are defined. Each functional form is included in a spatial generalized linear mixed model allowing accounting for spatial autocorrelation of data. We select between the different forms of gradients by using a Bayesian model selection procedure. In a first step, the functional form of the spillover for visual census and artisanal fishing data is assessed separately. For both sets of data, our model selection favoured the negative exponential model, evidencing a decrease of the spatial abundance of fish vanishing around 1000 m from the MPA border. We combined both datasets in a joint model by including an observability parameter. This parameter captures how the different sources of data quantify the underlying spatial distribution of the harvested species. This enabled us to demonstrate that the different sampling methods do not affect the estimation of the underlying spatial distribution of Diplodus spp. inside and outside the MPA. We show that data from different sources can be pooled through spatial generalized linear mixed model. Our findings allow to better understand the underlying mechanisms that control spillover of fish from MPA.     

Modeling possible cooling-water intake system impacts on Ohio River fish populations

To assess the possible impacts caused by cooling-water intake system entrainment and impingement losses, populations of six target fish species near power plants on the Ohio River were modeled. A Leslie matrix model was constructed to allow an evaluation of bluegill, freshwater drum, emerald shiner, gizzard shad, sauger, and white bass populations within five river pools. Site-specific information on fish abundance and length-frequency distribution was obtained from long-term Ohio River Ecological Research Program and Ohio River Sanitation Commission (ORSANCO) electrofishing monitoring programs. Entrainment and impingement data were obtained from 316(b) demonstrations previously completed at eight Ohio River power plants. The model was first run under a scenario representative of current conditions, which included fish losses due to entrainment and impingement. The model was then rerun with these losses added back into the populations, representative of what would happen if all entrainment and impingement losses were eliminated. The model was run to represent a 50-year time period, which is a typical life span for an Ohio River coal-fired power plant. Percent changes between populations modeled with and without entrainment and impingement losses in each pool were compared to the mean interannual coefficient of variation (CV), a measure of normal fish population variability. In 6 of the 22 scenarios of fish species and river pools that were evaluated (6 species x 5 river pools, minus 8 species/river pool combinations that could not be evaluated due to insufficient fish data), the projected fish population change was greater than the expected variability of the existing fish population, indicating a possible adverse environmental impact. Given the number of other variables affecting fish populations and the conservative modeling approach, which assumed 100% mortality for all entrained fish and eggs, it was concluded that the likelihood of impact was by no means assured, even in these six cases. It was concluded that in most cases, current entrainment and impingement losses at six Ohio River power plants have little or no effect at the population level.     

Fish community structure of Oklahoma gulf coastal plains

We collected the common fish species in all available aquatic habitats (streams, oxbow lakes, swamps) in bottomlands of McCurtain County, Oklahoma. Abundance and distribution of fishes, and environmental data were analyzed by a multivariate approach and examined for fit to a hierarchical model. The variables maximum depth, substrate, and presence of flow were the most important variables predicting fish community structure. Our multivariate analyses demonstrate that environmental factors can explain much variation in presence and abundance of the common fish species. Kolasa's hierarchical model relates species to each other by comparing ranges. This habitat-based model explained the relationships of our species ranges.     

Quantifying the interplay between environmental and social effects on aggregated-fish dynamics

Demonstrating and quantifying the respective roles of social interactions and external stimuli governing fish dynamics is key to understanding fish spatial distribution. If seminal studies have contributed to our understanding of fish spatial organization in schools, little experimental information is available on fish in their natural environment, where aggregations often occur in the presence of spatial heterogeneities. Here, we applied novel modeling approaches coupled to accurate acoustic tracking for studying the dynamics of a group of gregarious fish in a heterogeneous environment. To this purpose, we acoustically tracked with submeter resolution the positions of twelve small pelagic fish (Selar crumenophthalmus) in the presence of an anchored floating object, constituting a point of attraction for several fish species. We constructed a field-based model for aggregated-fish dynamics, deriving effective interactions for both social and external stimuli from experiments. We tuned the model parameters that best fit the experimental data and quantified the importance of social interactions in the aggregation, providing an explanation for the spatial structure of fish aggregations found around floating objects. Our results can be generalized to other gregarious species and contexts as long as it is possible to observe the fine-scale movements of a subset of individuals.     

Assessing the Risk of Hydroxybenzene Contamination in Fish Raised on a Chinese Aquaculture Farm

We estimated the risk posed by hydroxybenzene in farm-raised Crucian carps (Carassius carassius) from a pond on a fish farm in Beijing, China, by analyzing a time-series of observed hydroxybenzene concentrations in fish within a culturing season in year 2006. We used the basic linear regression to model the data and forecast the probability of hydroxybenzene concentration in pond fish exceeding selected health effect criteria (the risk) using Monte Carlo simulation. The risk is highly correlated with the time a fish stays in the pond. The results indicate that the risk–days in pond relationship resembles a sigmoid function with an inflection point around 150 days. The resulting model can be used to demonstrate the benefit of improving water quality in terms of increased fish production.     

Trophic levels of fish species of commercial importance in the Colombian Caribbean

Ecological studies on commercial important fish species are of great value to support resource management issues. This study calculated trophic levels of those Colombian Caribbean fish species whose diet has been locally described. Usable diet data of 119 species resulted in 164 trophic level estimates. An ordinary regression model relating trophic level and fish size was formulated. The regression slope was positive and significantly different from zero (p < 0.05) suggesting a scaling of trophic level with fish size. Both the list of trophic levels and the regression model should be of help in the formulation of trophic indicators and models of neotropical ecosystems.     

Variation in the temperature preference and growth rate of individual fish reconciles differences between two growth models

1. A growth model, originally developed for brown trout (Salmo trutta), has now been fitted to data for Atlantic salmon (S. salar) and stone‐loach (Barbatula barbatula) from English populations, and Arctic charr (Salvelinus alpinus) from Sweden. The model relates growth rate to temperature for a fish of standard size and the functional relationship has a triangular shape with a sharp peak at the optimal temperature for growth and zero growth at the base of the triangle. It was unsuitable for growth data for Norwegian salmon, and a curvilinear Ratkowsky model provided a better fit, though the experimental protocol was different in the Norwegian and English experiments. 2. The Norwegian salmon were kept in groups in each tank, had to compete for food, and had to be divided into slow, moderate and fast growers before the Ratkowsky model could be fitted. Each English salmon was kept in its own tank and fed individually. For replicate experiments, fish of similar size were selected. Variation among fish kept under similar conditions was therefore small, and the triangular model was essentially for individual fish, not groups of fish. 3. The present simulation study tests the hypothesis that individual differences in the growth response could account for the curvilinear growth‐temperature relationship for the Norwegian salmon. The triangular model was used to generate the growth response to temperature for a group of salmon, each fish having a slightly different temperature preference and growth rate. The result was a curvilinear response, well approximated by the Ratkowsky model (adjusted R² = 0.96). When the variability in individual temperature preference was increased, the Ratkowsky model was an even better fit (adjusted R² = 0.98). Therefore, the apparent discrepancy between the two models was reconciled by allowing for individual differences in temperature preference and growth rate within groups of fish.     

Local stream habitat variables predicted from catchment scale characteristics are useful for predicting fish distribution

South-east Queensland (Australia) streams were described by 21 local habitat variables that were chosen because of their potential association with fish distribution. An Assessment by a Nearest Neighbour Analysis (ANNA) model used large-scale variables that are robust to human influence to predict what the values of each of the 21 local habitat variables at each site would be without modification from human activity. The ANNA model used elevation, stream order, distance from source and longitude to predict the local habitat variables; other candidate predictor variables (mean rainfall, latitude and catchment area) were not found to be useful. The ANNA model was able to predict five of the 21 local habitat variables (average width, sand (%), cobble (%), rocks (%) and large woody debris) with an R ² of at least 0.2. The observed values of these five local habitat variables were used to model the distributions of individual fish species. The species distribution models were developed using logistic regression based on a subset of the data (some of the data were withheld for model validation) and a forward stepwise model selection procedure. There was no difference in predictive performance of fish distribution models for model predictions based on observed values and model predictions based on ANNA predicted values of local habitat variables in the withheld data (p-value = 0.85). Therefore, it is possible to predict the suitability of sites as habitat for given fish species using estimated (estimates based on large-scale variables) natural values of local habitat variables.     

Analyzing fish movement as a persistent turning walker

The trajectories of Kuhlia mugil fish swimming freely in a tank are analyzed in order to develop a model of spontaneous fish movement. The data show that K. mugil displacement is best described by turning speed and its auto-correlation. The continuous-time process governing this new kind of displacement is modelled by a stochastic differential equation of Ornstein–Uhlenbeck family: the persistent turning walker. The associated diffusive dynamics are compared to the standard persistent random walker model and we show that the resulting diffusion coefficient scales non-linearly with linear swimming speed. In order to illustrate how interactions with other fish or the environment can be added to this spontaneous movement model we quantify the effect of tank walls on the turning speed and adequately reproduce the characteristics of the observed fish trajectories.     

Predictive modelling and spatial mapping of freshwater fish and decapod assemblages using GIS and neural networks

1. We used stream fish and decapod spatial occurrence data extracted from a national database and recent surveys with geospatial landuse data, geomorphologic, climatic, and spatial data in a geographical information system (GIS) to model fish and decapod occurrence in the Wellington Region, New Zealand. 2. To predict the occurrence of each species at a site from a common set of predictor variables we used a multi‐response, artificial neural network (ANN), to produce a single model that predicted the entire fish and decapod assemblage in one procedure. 3. The predictions from the ANN using this landscape scale data proved very accurate based on evaluation metrics that are independent of species abundance or probability thresholds. The important variables contributing to the predictions included the latitudinal and elevational position of the site reach, catchment area, average air temperature, the vegetation type, landuse proportions of the catchment, and catchment geology. 4. Geospatial data available for the entire regional river network were then used to create a habitat‐suitability map for all 14 species over the regional river network using a GIS. This prediction map has many potential uses including: monitoring and predicting temporal changes in fish communities caused by human activities and shifts in climate, identifying areas in need of protection, biodiversity hotspots, and areas suitable for the reintroduction of endangered or rare species.     

Energetic costs, underlying resource allocation patterns, and adaptive value of predator-induced life-history shifts

We studied costs and benefits of life history shifts of water fleas (genus Daphnia ) in response to infochemicals from planktivorous fish. We applied a dynamic energy budget model to investigate the resource allocation patterns underlying the observed life history shifts and their adaptive value under size selective predation in one coherent analysis. Using a published data set of life history shifts in response to fish infochemicals we show that Daphnia invests less energy in somatic growth in the fish treatment. This observation complies with theoretical predictions on optimal resource allocation. However, the observed patterns of phenotypic plasticity cannot be explained by changes in resource allocation patterns alone because our model-based analysis of the empirical data clearly identified additional bioenergetic costs in the fish treatments. Consequently, the response to fish kairomone only becomes adaptive if the intensity of size selective predation surpasses a certain critical level. We believe that this is the first study that puts resource allocation, energetic costs, and adaptive value of predator induced life-history shifts – using empirical data – into one theoretical framework.     

Can we project changes in fish abundance and distribution in response to climate?

Large‐scale and long‐term changes in fish abundance and distribution in response to climate change have been simulated using both statistical and process‐based models. However, national and regional fisheries management requires also shorter term projections on smaller spatial scales, and these need to be validated against fisheries data. A 26‐year time series of fish surveys with high spatial resolution in the North‐East Atlantic provides a unique opportunity to assess the ability of models to correctly simulate the changes in fish distribution and abundance that occurred in response to climate variability and change. We use a dynamic bioclimate envelope model forced by physical–biogeochemical output from eight ocean models to simulate changes in fish abundance and distribution at scales down to a spatial resolution of 0.5°. When comparing with these simulations with annual fish survey data, we found the largest differences at the 0.5° scale. Differences between fishery model runs driven by different biogeochemical models decrease dramatically when results are aggregated to larger scales (e.g. the whole North Sea), to total catches rather than individual species or when the ensemble mean instead of individual simulations are used. Recent improvements in the fidelity of biogeochemical models translate into lower error rates in the fisheries simulations. However, predictions based on different biogeochemical models are often more similar to each other than they are to the survey data, except for some pelagic species. We conclude that model results can be used to guide fisheries management at larger spatial scales, but more caution is needed at smaller scales.     

Order out of Chaos

The current crisis in the world's fisheries indicates the need for a different management method than that now used by Western scientists, which regulates the quantity of fish taken. The authors propose a method called parametric management, which takes into account the complex, chaotic nature offish stocks and emphasizes preserving regular biological processes in the life cycle of fish by controlling how people fish. Supporting data come from 28 folk societies, the Maine lobster industry, and the authors' mathematical model of fish stocks.