Incorporating vertical movement of fishes in habitat use models

Fish telemetry studies now routinely collect positional and depth data, yet analytical approaches that integrate three-dimensional data are limited. Here we apply the potential path volume (PPV) model, a method previously developed to estimate habitat volume based on rates of avian movement, to free-swimming fish. Using a telemetry dataset of white sucker (Catastomus commersonii) from Turkey Lake (Ontario, Canada), we evaluated the effects of the number of spatial positions and different methods of selecting swim speed (v_swim), a key parameter for PPV models, on habitat volume estimates. We subsequently compared habitat volume estimates and habitat overlap among white sucker pairs from the PPV models to those calculated using kernel utilization distribution-based approaches. The number of spatial positions in the PPV model had a significant effect on habitat volume estimates, whereas the magnitude of the v_swim parameter or its specificity (constant value vs. fish-season specific parameter values) did not affect habitat volume estimates. The PPV method resulted in larger habitat volume estimates and greater habitat overlap estimates among fish pairs relative to those obtained from a three-dimensional kernel utilization distribution method. The PPV model is a useful analytical tool that, by incorporating potential animal movement into habitat use evaluations, can help answer key ecological questions and provide insight into fish space use in a wide range of conservation and management applications.