Fish–Habitat Relationships in Lakes: Gaining Predictive and Explanatory Insight by Using Artificial Neural Networks

Understanding and predicting the impacts of habitat modification and loss on fish populations are among the main challenges confronting fisheries biologists in the new millennium. Statistical models play an important role in this regard, providing a means to quantify how environmental conditions shape contemporary patterns in fish populations and communities and formulating this knowledge in a framework where future patterns can be predicted. Developing fish–habitat models by traditional statistical approaches is problematic because species often exhibit complex, nonlinear responses to environmental conditions and biotic interactions. We demonstrate the value of a robust statistical technique, artificial neural networks, relative to more traditional regression techniques for modeling such complexities in fish–habitat relationships. Using artificial neural networks, we provide both explanatory and predictive insight into the whole-lake and within-lake habitat factors shaping species occurrence and abundance in lakes from southcentral Ontario, Canada. The results show that species presence or absence is highly predictable based on whole-lake measures of habitat, and that these fish–habitat models show good generality in predicting occurrence in other lakes from an adjacent drainage. Detailed evaluation of these models shows that partitioning the predictive performance of the models into measures such as sensitivity (ability to predict species presence) and specificity (ability to predict species absence) allows assessment of the strengths, weaknesses, and applicability of the models more readily. We show that artificial neural networks are a useful approach for examining the interactive effects of habitat and biotic factors that shape species occurrence, abundance, and spatial occupancy within lakes. Finally, using simulated and empirical examples, we show that artificial neural networks provide greater predictive power than do traditional regression approaches for modeling species occurrence and abundance.