Climate change impacts ecosystems in many ways, from effects on species to phenology to wildfire dynamics. Assessing the potential vulnerability of ecosystems to future changes in climate is an important first step in prioritizing and planning for conservation. Although assessments of climate change vulnerability commonly are done for species, fewer have been done for ecosystems. To aid regional conservation planning efforts, we assessed climate change vulnerability for ecosystems in the Southeastern United States and Caribbean.

First, we solicited input from experts to create a list of candidate ecosystems for assessment. From that list, 12 ecosystems were selected for a vulnerability assessment that was based on a synthesis of available geographic information system (GIS) data and literature related to 3 components of vulnerability—sensitivity, exposure, and adaptive capacity. This literature and data synthesis comprised “Phase I” of the assessment. Sensitivity is the degree to which the species or processes in the ecosystem are affected by climate. Exposure is the likely future change in important climate and sea level variables. Adaptive capacity is the degree to which ecosystems can adjust to changing conditions. Where available, GIS data relevant to each of these components were used. For example, we summarized observed and projected climate, protected areas existing in 2011, projected sea-level rise, and projected urbanization across each ecosystem’s distribution. These summaries were supplemented with information in the literature, and a short narrative assessment was compiled for each ecosystem. We also summarized all information into a qualitative vulnerability rating for each ecosystem.

Next, for 2 of the 12 ecosystems (East Gulf Coastal Plain Near-Coast Pine Flatwoods and Nashville Basin Limestone Glade and Woodland), the NatureServe Habitat Climate Change Vulnerability Index (HCCVI) framework was used as an alternative approach for assessing vulnerability. Use of the HCCVI approach comprised “Phase II” of the assessment. This approach uses summaries of GIS data and models to develop a series of numeric indices for components of vulnerability. We incorporated many of the data sources used in Phase I, but added the results of several other data sources, including climate envelope modeling and vegetation dynamics modeling. The results of Phase II were high and low numeric vulnerability ratings for mid-century and the end of century for each ecosystem. The high and low ratings represented the potential range of vulnerability scores owing to uncertainties in future climate conditions and ecosystem effects.

Of the 12 ecosystems assessed in the first approach, five were rated as having high vulnerability (Caribbean Coastal Mangrove, Caribbean Montane Wet Elfin Forest, East Gulf Coastal Plain Southern Loess Bluff Forest, Edwards Plateau Limestone Shrubland, and Nashville Basin Limestone Glade and Woodland). Six ecosystems had medium vulnerability, and one ecosystem had low vulnerability. For the two ecosystems assessed with both approaches, vulnerability ratings generally agreed. The assessment concluded by comparing the two approaches, identifying critical research needs, and making suggestions for future ecosystem vulnerability assessments in the Southeast and beyond. Research needs include reducing uncertainty in the degree of climate exposure likely in the future, as well as acquiring more information on how climate might affect biotic interactions and hydrologic processes. Ideally, a comprehensive vulnerability assessment would include both the narrative summaries that resulted from the synthesis in Phase I, as well as a numeric index that incorporates uncertainty as in Phase II.