COS 173-5 - Species cover explains only a small amount of spatial variability in plant water sensitivity

Abstract: Spatio-temporal patterns of plant water uptake, loss, and storage exert a first-order control on photosynthesis, evapotranspiration dynamics, and thus, land-atmosphere interactions. Many studies of plant water uptake response to drought stress have focused on differences in drought response shown by different species, as a consequence of their different stomatal closure, xylem conductance, and root traits. However, several other factors can affect spatial variability in plant water stress response, including but not limited to topographic and soil properties that influence rainfall infiltration and water distribution in the unsaturated zone, adaptation to local climatic variations, and changes in vegetation density. Understanding the degree to which species variability or geographic factors control the overall dynamics of plant water uptake is fundamental to understanding the response of vegetation to hydrologic variations, but has traditionally been hindered by data limitations. In situ measurements are generally too sparse to sufficiently span the range of possible variations across different geographic factors. Here, we study the effects of different geographic factors on plant-water interactions by analyzing the dynamics of remotely sensed live fuel moisture content (LFMC, defined as the vegetation water content divided by dry biomass) determined from Sentinel-1 synthetic aperture radar and Landsat multispectral observations. LFMC directly reflects vegetation water content and therefore patterns of plant water uptake and evapotranspiration. We quantify the "plant-water sensitivity" by using an auto-regressive model comparing LFMC to water-related climate variables and analyze the spatial patterns of plant water sensitivity at 4 km resolution across 12,888 locations across the Western United States. We used data from the Forest Inventory Analysis (FIA) program to ensure that these locations were each dominated by a single species. Species cover information was able to explain only 12% of the observed variability in PWS across these sites. By contrast, elevation and topographic features, soil properties, climate, and vegetation density explained 43% of variability in PWS, based on a random forest model. Thus, only a small fraction (less than 30%) of spatial variability in plant water stress is explained by species cover. This suggests that studies that analyze drought response across species are able to address only a limited component of overall plant drought response dynamics. Among non-species factors, the local climate is a key control on plant water stress variability.