Modeling relationships between landscape-level attributes and snorkel counts of chinook salmon and steelhead parr in Idaho

Knowledge of environmental factors impacting anadromous salmonids in their freshwater habitats, particularly at large spatial scales, may be important for restoring them to previously recorded levels in the northwestern Unite d States. Consequently, we used existing data sets and an information-theor etic approach to model landscape-level attributes and snorkel count categor ies of spring-summer chinook salmon (Oncorhynchus tshawytscha) and steelhea d (Oncorhynchus mykiss) parr within index areas in Idaho. Count categories of chinook salmon parr were negatively related to geometric mean road densi ty and positively related to mean annual precipitation, whereas those for s teelhead parr were negatively related to percent unconsolidated lithology. Our models predicted that chinook salmon parr would be in low count categor ies within subwatersheds with > 1 km.km(-2) geometric mean road densities a nd (or) < 700 mm mean annual precipitation. Similarly, steelhead parr were predicted to be in low count categories in subwatersheds with > 30% unconso lidated lithology. These results provide a starting point for fish biologis ts and managers attempting to map approximate status and quality of rearing habitats for chinook salmon and steelhead at large spatial scales.