Transferability of predictive models of coral reef fish species richness
Ana Sequeira, Jessica Meeuwig | Dec 17, 2015
Diver-based survey of coral reef fish communities along a transect line in the western Indian Ocean.
Photo: Dr. Emily Darling (Wildlife Conservation Society).
Sequeira AMM, Mellin C, Lozano-Montes H, Vanderklift MA, Babcock RC, Haywood MDE, Meeuwig JJ, Caley J. 2015. Transferability of predictive models of coral reef fish species richness. Journal of Applied Ecology, 53(1): 64-72.
- Coral reefs are the most diverse of all marine systems, but are under increasing threat from global change.
- Predictive models are useful for assessing the likely impacts of global change on tropical species.
- Data have been collected for the Great Barrier Reef (GBR) for the last 20 years, but research and monitoring work at Ningaloo Reef (NR) is only comparatively recent.
- Despite huge differences in habitat and climate at the two sites, the models developed for the GBR (where data are plentiful) are transferable in space – meaning they can be used to learn about, and make predictions for, NR (where data are sparse).
Understanding biodiversity patterns depends on data collection, which in marine environments can be prohibitively expensive. Transferable predictive models could therefore provide time- and cost-effective tools for understanding biodiversity–environment relationships. We used fish species counts and spatial and environmental predictors to develop predictive models of fish species richness (S) for two major coral reefs located in separate ocean basins: Australia’s Great Barrier Reef (GBR; Queensland) and Ningaloo Reef (NR; Western Australia). We tested the ability of the GBR model to predict S at NR (its transferability) under various scenarios using different sampling durations, years sampled and transect sizes. Based on Rsq, the GBR model poorly predicted S at NR (Rsq < 16%) with few predicted values strongly correlated with observations. However, comparable spatial patterns in S across NR were predicted by both the NR and the GBR models when calibrated at similar spatio-temporal scales.
This result suggests that poor validation of the transferred models may indicate low deviance explained by the predictors in the new system (where other predictors not included might have a more direct effect on the response) and that in some situations, model transferability may be considerably improved by using data sets of similar spatio-temporal scales. Therefore, data filtering by time and space may be required prior to transferring models. Transferable models can provide initial estimates of fish species richness patterns in poorly sampled systems, and thereby guide the design of better and more efficient sampling programs. Further improvements in model transferability will increase their predictive power and utility in conservation planning and management.
NINGALOO FROM THE AIR
The Ningaloo Coast is a pristine World Heritage site boasting exceptional marine biodiversity. Predictive models can help assess threats to animal populations in the region and give insights into how they may respond to future climate and human impacts (Photo: Dr. Ana Sequeira).
Prediction maps of fish species richness (S) at Ningaloo Reef (NR) obtained from Great Barrier Reef (GBR) scenarios A–G and the NR models (Sequeira et al. 2015).
FUNDING & ACKNOWLEDGEMENTS
A.M.M.S. was supported by an IOMRC (UWA/AIMS/CSIRO) collaborative Postdoctoral Fellowship. C.M. was funded by the Marine Biodiversity Hub and by ARC Grant (DE140100701). Thanks to R. Pitcher and M. Case for access to environmental data (collected under CERF and NERP), and to all collectors of the fish data.
Challenges of transferring models of fish abundance between coral reefs
Reliable abundance estimates for species are fundamental in ecology, fisheries, and conservation. Consequently, predictive models ...
Transferring biodiversity models for conservation: Opportunities and challenges
The ability of a model to predict biodiversity metrics in novel environments is termed ‘transferability’, ...