부산대학교 도서관

Water temperatures on river networks fluctuate in time and over space. This variation creates complex thermal regimes through which fishes move over the course of their lives. Using water temperature data collected every 30 min for 6 y at 45 locations throughout the Snoqualmie River Basin, Washington, USA, we developed a set of metrics representing facets of the thermal regime important to native Chinook Salmon (Oncorhynchus tshawytscha) and Bull Trout (Salvelinus confluentus) and to non-native Largemouth Bass (Micropterus salmoides) during multiple life-history stages. We modeled these facets using spatial stream network models (SSNMs) built from covariates describing landform, land use, and climate to better understand the potential ecological drivers of thermal habitats for these fishes. The explanatory power (predictive r 2) of the final SSNMs ranged from 23 to 97%, with at least ½ the models >69%. The best models tended to include climate and landform covariates, with elevation and mean annual air temperature most often included. The relative contribution of each SSNM component differed by species: thermal metric models for Chinook Salmon tended to have a covariate- and site-based structure, those for Bull Trout tended to have a yearly structure, and those for Largemouth Bass tended to have a spatial structure. We applied our models to map the thermal landscape in ways that should be of particular interest to fish managers. Given the challenging lack of empirical data for rare fish species or fish species that are currently not present, we illustrate how the proliferation of fine-scale temperature data can be leveraged to provide ecological insight. For example, we identified several tributaries and small low-elevation reaches as areas potentially worthy of protection efforts for Chinook Salmon, and we identified the mainstem Snoqualmie River as an area of potential interaction between native salmon and non-native bass. [ABSTRACT FROM AUTHOR]