Interspecies Synchrony in Salmonid Densities Associated with Large-Scale Bioclimatic Conditions in Central Idaho

The abundance of lotic salmonids varies substantially through time, but the extent to which several sympatric species respond coherently to large-scale bioclimatic conditions has rarely been investigated for freshwater fishes. We compared correlations in salmonid density in central Idaho and examined the relationships between changes in salmonid density and the variation in large-scale bioclimatic conditions as indexed by stream flow, air temperature, drought, coastal upwelling, and the number of Chinook salmon Oncorhynchus tshawytscha redds (a surrogate for nutrient influx and therefore increased stream productivity). The average densities of six stream-dwelling salmonid fishes were highly synchronous, with declines from the mid-1980s to the mid-1990s followed by a rebound through 2003. All pairwise correlations were positive and 8 of 15 were statistically significant. Stream flow and Chinook salmon redds were correlated with fish densities as a group, but the importance of bioclimatic indices differed by species and varied by the lag times applied. Stream flow 3 and 4 years previous was most important for brook trout Salvelinus fontinalis and bull trout S. confluentus. The presence of Chinook salmon redds from the previous year was most important for Chinook salmon. The coastal upwelling index 4 years previous was most important for mountain whitefish Prosopium williamsoni and steelhead O. mykiss. No variable captured much of the variation in the density of westslope cutthroat trout O. clarkii lewisi. Models based on data from 1985 to 2003 correctly predicted the synchronous declines in fish density observed for all species from 2004 to 2009, suggesting that the bioclimatic indices we chose were useful surrogates for large-scale factors influencing the temporal changes in salmonid densities in central Idaho. For fishery managers, interspecies synchrony may allow a more powerful multispecies approach to monitoring and reveal the tractability of large systems to management control. However, causal mechanisms are hard to infer and require further research (e.g., the effects of ocean conditions). Our study provides a basis for such work.