Marine ecosystems, particularly in high‐latitude regions such as the Arctic, have been significantly affected by human activities and contributions to climate change. Evaluating how fish populations responded to past changes in their environment is helpful for evaluating their future patterns, but is often hindered by the lack of long‐term biological data available. Using otolith increments of Northeast Arctic cod (Gadus morhua) as a proxy for individual growth, we developed a century‐scale biochronology (1924–2014) based on the measurements of 3,894 fish, which revealed significant variations in cod growth over the last 91 years. We combined mixed‐effect modeling and path analysis to relate these growth variations to selected climate, population and fishing‐related factors. Cod growth was negatively related to cod population size and positively related to capelin population size, one of the most important prey items. This suggests that density‐dependent effects are the main source of growth variability due to competition for resources and cannibalism. Growth was also positively correlated with warming sea temperatures but negatively correlated with the Atlantic Multidecadal Oscillation, suggesting contrasting effects of climate warming at different spatial scales. Fishing pressure had a significant but weak negative direct impact on growth. Additionally, path analysis revealed that the selected growth factors were interrelated. Capelin biomass was positively related to sea temperature and negatively influenced by herring biomass, while cod biomass was mainly driven by fishing mortality. Together, these results give a better understanding of how multiple interacting factors have shaped cod growth throughout a century, both directly and indirectly.
Marine spatial planning (MSP) is considered a valuable tool in the ecosystem-based management of marine areas. Predictive modelling may be applied in the MSP framework to obtain spatially explicit information about biodiversity patterns. The growing number of statistical approaches used for this purpose implies the urgent need for comparisons between different predictive techniques. In this study, we evaluated the performance of selected machine learning and regression-based methods that were applied for modelling fish community indices. We hypothesized that habitat features can influence fish assemblage and investigated the effect of environmental gradients on demersal fish diversity (species richness and Shannon–Weaver Index). We used fish data from the Baltic International Trawl Surveys (2001–2014) and maps of six potential predictors: bottom salinity, depth, seabed slope, growth season bottom temperature, seabed sediments and annual mean bottom current velocity. We compared the performance of six alternative modelling approaches: generalized linear models, generalized additive models, multivariate adaptive regression splines, support vector machines, boosted regression trees and random forests. We applied repeated 10-fold cross-validation, using accuracy as the measure of model quality. Finally, we selected random forest as the best performing algorithm and implemented it for the spatial prediction of fish diversity from the Baltic Proper to the Kattegat. To obtain information on the data reliability and confidence of the developed models, which are essential for MSP, we estimated the uncertainty of predictions with standard deviation of predictions obtained from all the trees in the ensemble random forest method. We showed how state-of-the-art predictive techniques, based on easily available data and simple Geographic Information System tools, can be used to obtain reliable spatial information about fish diversity. Our comparative work highlighted the potential of machine learning method to reduce prediction error in modelling of demersal fish diversity in the framework of MSP.
Management strategy evaluation (MSE) is the state‐of‐the‐art approach for testing and comparing management strategies in a way that accounts for multiple sources of uncertainty (e.g. monitoring, estimation, and implementation). Management strategy evaluation can help identify management strategies that are robust to uncertainty about the life history of the target species and its relationship to other species in the food web. Small pelagic fish (e.g. anchovy, herring and sardine) fulfil an important ecological role in marine food webs and present challenges to the use of MSE and other simulation‐based evaluation approaches. This is due to considerable stochastic variation in their ecology and life history, which leads to substantial observation and process uncertainty. Here, we summarize the current state of MSE for small pelagic fishes worldwide. We leverage expert input from ecologists and modellers to draw attention to sources of process and observation uncertainty for small pelagic species, providing examples from geographical regions where these species are ecologically, economically and culturally important. Temporal variation in recruitment and other life‐history rates, spatial structure and movement, and species interactions are key considerations for small pelagic fishes. We discuss tools for building these into the MSE process, with examples from existing fisheries. We argue that model complexity should be informed by management priorities and whether ecosystem information will be used to generate dynamics or to inform reference points. We recommend that our list of considerations be used in the initial phases of the MSE process for small pelagic fishes or to build complexity on existing single‐species models.
1 April 2020 in Canadian Journal of Fisheries and Aquatic Sciences
The assignment of individual fish to its stock of origin is important for reliable stock assessment and fisheries management. Otolith shape is commonly used as the marker of distinct stocks in discrimination studies. Our literature review showed that the application and comparison of alternative statistical classifiers to discriminate fish stocks based on otolith shape is limited. Therefore, we compared the performance of two traditional and four machine learning classifiers based on Fourier analysis of otolith shape using selected stocks of Atlantic cod (Gadus morhua) in the southern Baltic Sea and Atlantic herring (Clupea harengus) in the western Norwegian Sea, Skagerrak, and the southern Baltic Sea. Our results showed that the stocks can be successfully discriminated based on their otolith shapes. We observed significant differences in the accuracy obtained by the tested classifiers. For both species, support vector machines (SVM) resulted in the highest classification accuracy. These findings suggest that modern machine learning algorithms, like SVM, can help to improve the accuracy of fish stock discrimination systems based on the otolith shape.
The isotopic composition of inorganic carbon in otoliths (δ13Coto) can be a useful tracer of metabolic rates and a method to study ecophysiology in wild fish. We evaluated environmental and physiological sources of δ13Coto variation in Icelandic and Northeast Arctic (NEA) cod (Gadus morhua) over the years 1914–2013. Individual annual growth increments of otoliths formed at age 3 and 8 were micromilled and measured by isotope-ratio mass spectrometry. Simultaneously, all annual increment widths of the otoliths were measured providing a proxy of fish somatic growth. We hypothesized that changes in the physiological state of the organism, reflected by the isotopic composition of otoliths, can affect the growth rate. Using univariate and multivariate mixed-effects models we estimated conditional correlations between carbon isotopic composition and growth of fish at different levels (within individuals, between individuals, and between years), controlling for intrinsic and extrinsic effects on both otolith measurements. δ13Coto was correlated with growth within individuals and between years, which was attributed to the intrinsic effects (fish age or total length). There was no significant correlation between δ13Coto and growth between individuals, which suggests that caution is needed when interpreting δ13Coto signals. We found a significant decrease in δ13Coto through the century which was explained by the oceanic Suess effect-admixture of isotopically light carbon from fossil fuel. We calculated the proportion of the respired carbon in otolith carbonate (Cresp) using carbon isotopic composition in diet and dissolved inorganic carbon of the seawater. This approach allowed us to correct the values for each stock in relation to these two environmental baselines. Cresp was on average 0.275 and 0.295 in Icelandic and NEA stock, respectively. Our results provide an insight into the physiological basis for differences in growth characteristics between these two cod stocks, and how that may vary over time.