Predicting patterns of vulnerability to climate change in near coastal species using an algorithm-based risk assessment framework
Lee, H., C. Folger, D. Reusser, Pat Clinton, AND R. Graham. Predicting patterns of vulnerability to climate change in near coastal species using an algorithm-based risk assessment framework. U.S. Environmental Protection Agency, Washington, DC, EPA/600/R-17/052, 2017.
With projected increases in air and ocean temperatures, reductions in ocean pH (ocean acidification), and increasing sea level rise, climate change is arguably the greatest environmental threat facing near-coastal (0-200 m depth) species. Over the next 20 to 100 years, climate change is likely to affect both the biodiversity of near-coastal ecosystems and the ecosystem services provided by these critical habitats. However, species vary in their vulnerability to specific climatic changes and climate impacts will vary geographically. For management to focus on the species and locations at the greatest risk, it is critical to have a basic knowledge of the extent and pattern of climate risk. To address these needs, we developed a rule-based framework that predicts the relative risk of near-coastal species to ocean temperatures, ocean acidification, and sea level rise at regional scales. For ocean acidification, we developed a risk approach by treating reductions in pH as a “pollutant”. A unique feature of the current effort is that the risk analysis is automated using an online tool, the Coastal Biodiversity Risk Analysis Tool (CBRAT; http://www.cbrat.org/), that removes the limitations imposed by expert opinion. As a proof-of-concept, we conducted preliminary climate risk analyses on all the crabs (417 species) from the Beaufort Sea to Southern California, and partial analyses on all the rockfish and bivalves. The preliminary analyses indicated strong geographical patterns of risk; with minimal impacts of sea level rise in the north (Alaska) and greater risk along the southern coast of North America. Temperature impacts are largely limited to the southernmost range of a species. The current report documents the rules, assumptions, and limitations of the risk analysis approach, while the CBRAT site allows managers and researchers to evaluate different climate scenarios. This work contributes to ACE Project 2 - Climate Impacts on Watersheds, Water Quality, and Ecosystems.
Near-coastal (0-200 depth) ecosystems and species are under threat from increasing temperatures, ocean acidification, and sea level rise. However, species vary in their vulnerability to specific climatic changes and climate impacts will vary geographically. For management to respond in a scientifically-sound fashion, it is critical to know the extent and geographic patterns of risk. To address these needs, we developed a rule-based framework to predict the relative risk of near-coastal species due to climate change at regional scales. The framework synthesizes risks from biotic traits and population trends with risks associated with increasing ocean temperatures, ocean acidification, and sea level rise. Key objectives in crafting the framework were to: a) predict climate risks for rare species as well as for better studied species; b) identify major climate stressor(s) affecting each species within each region; c) predict geographic patterns for the importance of different climate stressors; and d) assess how risks change under different climate scenarios. A unique feature of the current effort is that we developed an ecoinformatics website called the Coastal Biodiversity Risk Analysis Tool (CBRAT; http://www.cbrat.org/) to conduct climate risk analyses and to serve as a tool for managers and researchers to address specific climate and species inquiries. CBRAT uses an algorithm approach where the risk is automatically generated from a centralized knowledgebase and a set of explicit rules, thereby avoiding the limitations and potential biases of risks based on expert opinion. The first method to identify species at risk was a set of rules using “baseline/status”; biotic traits, such as endemicity, population trends and relative abundance, that are associated with increased climate vulnerability or resilience. These baseline/status rules help identify species that might not be identified as at risk through the analysis of specific climate stressors. The core method to predict temperature risks was the Ecoregional Thermal Windows approach that compares the projected sea surface temperature (SST) in each ecoregion to the historic range of SST values in the “warmest occupied ecoregion” or WOE. Temperatures in the WOE are assumed to represent the upper ecological thermal limit for a species to maintain a viable population. Though ocean acidification is the least well understood of the climate stressors, it is possible to conduct a first-order risk assessment by treating it as a contaminant. Specifically, we derive “maximum allowable toxicant concentrations” (MATCs) for pH and aragonite saturation state from a synthesis of exposure experiments. Because species within a taxon vary greatly in their sensitivity, we developed an approach to generate high, moderate, and low sensitivity effects thresholds to pH, though the analysis is limited by the paucity of experiments based on endpoints directly related to population viability. To predict the regional effects of sea level rise on intertidal species, we integrated predictions for regional net sea level rise, depth and habitat preferences of the species, and habitat thresholds, which are estimates of the proportion of each major intertidal habitat type lost to different levels of sea level rise. A preliminary uncertainty analysis suggests that the present framework is sufficient to elucidate geographical patterns of risk and identify high risk species, though it does not have sufficient resolution for fisheries management. Two geographic patterns that are emerging from the preliminary risk assessment are that sea level rise appears to be minor threat to intertidal species in much of Alaska and that temperature impacts are primarily limited to the southern range of species.