Scale and Hierarchical Relationships When Incorporating Observed Data into Fish Models

Identifying correlations between environmental variables and fish presence or density is usually the main focus of efforts to model fish-habitat relationships. These relationships, however, can be confounded by scale and hierarchical effects.  In particular the strength of fish – habitat correlations can change across both spatial and temporal scales as well as longitudinal gradients.  Hierarchical relationships among habitat variables can also complicate the modeling process because weighting all statistically significant relationships equally may not accurately reflect fish response.  We provide examples of scale and hierarchical effects on fish – habitat relationships both from the literature and from previous work. We also outline our approach to modeling fish – habitat relationships in our current efforts.