Assessing climate-induced range shifts of stream fishes using a consensus framework

Documenting climate-induced range shifts of freshwater fish is difficult because there are few spatially representative, long-term data sets with re-surveys of the same location over time. While such standardised data sets would allow for the identification of range shifts, most data sets are confounded by changing sampling priorities and methodologies through time. Finding a robust framework for detecting climate-induced range shifts thus remains a research priority. Using a database from stream surveys conducted during the period 1957-2019, we employed a multi-method framework for determining whether 10 stream fish species (nine warm-water, one cold-water) have exhibited elevational range shifts within the North Platte River watershed, Wyoming, USA. By employing a combination of different approaches (predictive reference models, quantile regressions, spatial aggregations), we produced a consensus framework for determining whether observed range expansions and range contractions were consistent across species. The weight of evidence from our consensus framework (>50% of all modelling approaches) indicated strong support for upstream range expansions of three (33%) warm-water species, including a long-established nonnative species (common carp [Cyprinus carpio]). The consensus framework revealed no evidence for range contractions of one cold-water trout species (brown trout [Salmo trutta]). As global climate change continues, the value of historical datasets for evaluating pre-warming, baseline conditions will increase. Given the methodological limitations of different analytical approaches, our findings illustrate that consensus frameworks based on the weight of evidence can improve confidence in the detection of climate-induced range shifts when robust historical data sets are lacking.