Probabilistic measures of climate change vulnerability, adaptation action benefits, and related uncertainty from maximum temperature metric selection

Abstract

Predictions of the projected changes in species distributions and potential adaptation action benefits can help guide conservation actions. There is substantial uncertainty in projecting species distributions into an unknown future, however, which can undermine confidence in predictions or misdirect conservation actions if not properly considered. Recent studies have shown that the selection of alternative climate metrics describing very different climatic aspects (e.g., mean air temperature vs. mean precipitation) can be a substantial source of projection uncertainty. It is unclear, however, how much projection uncertainty might stem from selecting among highly correlated, ecologically similar climate metrics (e.g., maximum temperature in July, maximum 30-day temperature) describing the same climatic aspect (e.g., maximum temperatures) known to limit a species' distribution. It is also unclear how projection uncertainty might propagate into predictions of the potential benefits of adaptation actions that might lessen climate change effects. We provide probabilistic measures of climate change vulnerability, adaptation action benefits, and related uncertainty stemming from the selection of four maximum temperature metrics for brook trout (Salvelinus fontinalis), a cold-water salmonid of conservation concern in the eastern United States. Projected losses in suitable stream length varied by as much as 20% among alternative maximum temperature metrics for mid-century climate projections, which was similar to variation among three climate models. Similarly, the regional average predicted increase in brook trout occurrence probability under an adaptation action scenario of full riparian forest restoration varied by as much as .2 among metrics. Our use of Bayesian inference provides probabilistic measures of vulnerability and adaptation action benefits for individual stream reaches that properly address statistical uncertainty and can help guide conservation actions. Our study demonstrates that even relatively small differences in the definitions of climate metrics can result in very different projections and reveal high uncertainty in predicted climate change effects.

An assessment of marine, estuarine, and riverine habitat vulnerability to climate change in the Northeast U.S.

“…Stream temperatures are projected to increase significantly by the end of the century, with the largest increases in the southern Mid-Atlantic and northern New England [ 107 ]. These changes make cold- and cool-water rocky-bottom river systems in the Northeast U.S., and the species they support, particularly vulnerable, with implications to food web structure and energy flow in riverine communities [ 165 – 169 ]. In addition, riverine habitats have been historically altered by a host of non-climate perturbations, including damming, water diversion, mineral mining, and storm water pollution [ 170 – 173 ], which can exacerbate climate-related changes in temperature and streamflow.…”

Evaluating Forest Visitors’ Place Attachment, Recreational Activities, and Travel Intentions under Different Climate Scenarios

“…Moreover, the record-breaking high temperature of 40.2 °C was recorded in Taitung, Taiwan, in July of 2020. The degree of preference of visitors toward climate change adaptation policies was assessed by listing nine potentially suitable strategies for Wuling National Forest Recreation Area [79][80][81]. These were listed alongside existing measures.…”

Individual behaviour and resource use of thermally stressed brook trout Salvelinus fontinalis portend the conservation potential of thermal refugia

“…Often, this is because data used to construct predictive models are collected at spatial scales that are not commensurate with the scale at which fish use habitat (Fullerton et al, 2018;Snyder et al, 2015). Consider maximum summer temperature or maximum 30day average temperatures, which are often used to estimate reach-level occupancy probability for a fish species of interest (DeWeber & Wagner, 2018). These metrics summarise ambient conditions and can provide insights into occupancy probabilities for an entire stream reach, but they fail to account for fine-scale spatial heterogeneity in stream temperature that can increase local survival (Ebersole et al, 2003;Kanno et al, 2014;Snyder et al, 2015;Torgersen et al, 1999).…”