Weighted stream temperature tolerance index is insensitive to changes in stream fish composition

Abstract

Freshwater fish species require and compete for specific water temperatures and thermal regimes to maximize physiological and behavioral processes; however, the thermal habitats on which freshwater fish rely are being modified by local human activities and global climate change. Water temperature measurements that are often used to determine thermal regimes allow predictions of species occurrence in the absence of assemblage surveys. Methods that connect assemblage composition with temperature provide additional information that otherwise is lost when temperature is measured alone. Using stream fish assemblage data collected over 2 decades from various biomonitoring programs in southern Ontario, Canada, we tracked site-level changes in stream fish assemblage composition from numerous watersheds. We also assessed the utility of using a weighted species association tolerance index with respect to water temperature (WSATI-WT) that weights the abundance of each species in an assemblage by their thermal preference to generate an assemblage temperature preference. We tracked temporal changes in assemblage temperature preferences and used bootstrapping to assess the effects of sampling variability on this index. Fish assemblage composition changed over time in all watersheds, but these changes had differential effects on WSATI-WT values. Temporal changes in WSATI-WT values ranged from −9.43 to 5.04°C, of which 68% were negative and 32% positive. These changes imply that, over time, the thermal regime of some sites changed, but bootstrapping showed that WSATI-WT estimates could differ by ≤10°C (mean = 1.57°C) due to sampling variability alone. Examination of the variation in bootstrapped values relative to species composition indicates that uncommon or underrepresented species that have atypical thermal preferences relative to the other species within individual samples may have a large influence on WSATI-WT values. We must develop robust and resilient methods of tracking assemblage responses to warming temperatures if we hope to manage watersheds and species that are sensitive to warming temperatures.