Understanding Fisheries Data with Advanced Distribution Models

Jim Crocker
16th May, 2024

Understanding Fisheries Data with Advanced Distribution Models

The species distribution model reveals a clear habitat partitioning between the two Moreton Bay Bug species, predicting that Reef Bugs (Thenus australiensis) dominate deeper, offshore habitats while Mud Bugs (T. parindicus) are dominant in shallower, inshore waters.

Image adapted from: McMillan et al. / CC BY (Source)

Key Findings

  • The study by the Queensland Department of Agriculture and Fisheries used species distribution models (SDMs) to separate catch data for two species of shovel-nosed lobsters
  • SDMs revealed distinct trends for each lobster species, which were previously hidden in aggregated data
  • The study identified shifts in fishing behavior, showing increased targeting of one species over time
Fisheries data are essential for monitoring the impacts of fishing on wild populations, yet they often come with complications. One major issue is the aggregation of catch data for multiple species, which can obscure individual species' trends and make effective management challenging. This problem becomes more pronounced as fisheries increasingly target species that were previously caught incidentally. A recent study by the Queensland Department of Agriculture and Fisheries[1] addresses this issue by demonstrating the use of species distribution models (SDMs) to untangle multi-species catch data for better stock assessments. The study focuses on two species of shovel-nosed lobsters (Thenus spp.), which were historically recorded together in logbook records. By applying SDMs, the researchers were able to differentiate between the two species in the aggregated data, revealing previously hidden trends. This approach not only allowed for more accurate stock assessments but also highlighted shifts in fishing behavior, such as changes in target species. Species distribution models are statistical tools used to predict the distribution of species across geographic areas based on environmental conditions and species occurrence data. These models have been increasingly used to forecast the impacts of climate change on biodiversity[2]. However, their application in fisheries management, particularly for disaggregating multi-species catch data, remains underexplored. The study by the Queensland Department of Agriculture and Fisheries builds on previous research that has shown the utility of SDMs in understanding species distributions and their responses to environmental changes. For example, a study on the global distribution of tuna found that climate change is causing significant shifts in tuna habitats, with many species moving towards the poles[3]. This kind of information is crucial for developing adaptive management strategies to mitigate the impacts of climate change on fisheries. In the context of shovel-nosed lobsters, the use of SDMs provided a more granular view of catch data, which is often aggregated for non-target species. This is particularly important as fisheries management moves towards ecosystem-based approaches. A study on multispecies fisheries in southeastern Australia demonstrated that integrated management strategies, which consider multiple species and ecological interactions, outperform single-focus measures[4]. By enabling species-specific assessments, SDMs can support more effective ecosystem-based fisheries management (EBFM). The researchers used a long-term dataset from logbook records to develop their SDMs. They incorporated environmental variables such as sea surface temperature and depth, which are known to influence the distribution of marine species. The models were then used to allocate the aggregated catch data to the two shovel-nosed lobster species. This revealed distinct catch trends for each species, which were previously masked by the aggregated data. One key finding was the identification of shifts in fishing behavior. The study showed that as one species became more targeted, its catch rates increased while the other species' catch rates remained stable or declined. This kind of information is vital for fisheries managers, as it can inform decisions on quotas, gear restrictions, and other management measures. While the study demonstrates the potential of SDMs for disaggregating multi-species catch data, it also highlights some limitations. One challenge is the accuracy of the environmental data used in the models, which can vary across different geographic regions. Additionally, the models assume that the relationships between species distributions and environmental variables remain constant over time, which may not always be the case. Despite these limitations, the study provides a valuable template for researchers and managers looking to assess individual species using aggregated data. By revealing species-specific trends and shifts in fishing behavior, SDMs can enhance our understanding of fisheries dynamics and support more effective management strategies. In conclusion, the use of species distribution models to disaggregate multi-species catch data offers a promising approach for improving stock assessments and fisheries management. This study by the Queensland Department of Agriculture and Fisheries illustrates the potential of SDMs to reveal hidden trends and inform adaptive management strategies, building on previous research that underscores the importance of understanding species distributions and ecological interactions[2][3][4].

EnvironmentEcologyMarine Biology

References

Main Study

1) Untangling multi-species fisheries data with species distribution models

Published 15th May, 2024

https://doi.org/10.1007/s11160-024-09863-1

Related Studies

2) Predicting species distribution: offering more than simple habitat models.

https://doi.org/10.1111/j.1461-0248.2005.00792.x

3) Large-scale distribution of tuna species in a warming ocean.

https://doi.org/10.1111/gcb.14630

4) An integrated approach is needed for ecosystem based fisheries management: insights from ecosystem-level management strategy evaluation.

https://doi.org/10.1371/journal.pone.0084242


Related Articles

An unhandled error has occurred. Reload 🗙