Climatic water deficit, tree species ranges, and climate change in Yosemite National Park

Abstract Aim (1) To calculate annual potential evapotranspiration (PET), actual evapotranspiration (AET) and climatic water deficit (Deficit) with high spatial resolution; (2) to describe distributions for 17 tree species over a 2300‐m elevation gradient in a 3000‐km 2 landscape relative to AET and Deficit; (3) to examine changes in AET and Deficit between past ( c. 1700), present (1971–2000) and future (2020–49) climatological means derived from proxies, observations and projections; and (4) to infer how the magnitude of changing Deficit may contribute to changes in forest structure and composition. Location Yosemite National Park, California, USA. Methods We calculated the water balance within Yosemite National Park using a modified Thornthwaite‐type method and correlated AET and Deficit with tree species distribution. We used input data sets with different spatial resolutions parameterized for variation in latitude, precipitation, temperature, soil water‐holding capacity, slope and aspect. We used climate proxies and climate projections to model AET and Deficit for past and future climate. We compared the modelled future water balance in Yosemite with current species water‐balance ranges in North America. Results We calculated species climatic envelopes over broad ranges of environmental gradients – a range of 310 mm for soil water‐holding capacity, 48.3°C for mean monthly temperature (January minima to July maxima), and 918 mm yr −1 for annual precipitation. Tree species means were differentiated by AET and Deficit, and at higher levels of Deficit, species means were increasingly differentiated. Modelled Deficit for all species increased by a mean of 5% between past ( c. 1700) and present (1971–2000). Projected increases in Deficit between present and future (2020–49) were 23% across all plots. Main conclusions Modelled changes in Deficit between past, present and future climate scenarios suggest that recent past changes in forest structure and composition may accelerate in the future, with species responding individualistically to further declines in water availability. Declining water availability may disproportionately affect Pinus monticola and Tsuga mertensiana . Fine‐scale heterogeneity in soil water‐holding capacity, aspect and slope implies that plant water balance may vary considerably within the grid cells of kilometre‐scale climate models. Sub‐grid‐cell soil and topographical data can partially compensate for the lack of spatial heterogeneity in gridded climate data, potentially improving vegetation‐change projections in mountainous landscapes with heterogeneous topography.