Two decades of widespread drought-induced forest mortality events on every forested continent have raised the specter of future unpredictable, rapid ecosystem changes in 21st century forests. Yet our ability to predict drought stress, much less drought-induced mortality across the landscape remains limited. This uncertainty stems from both incomplete understanding of spatial variation in drought tolerance within widespread species, and an incomplete understanding of the spatial variation in the minimum xylem tensions (variation in exposure) experienced by a species across the landscape. We built on previous research showing limited within-species variation in thresholds to hydraulic damage in blue oak (Quercus douglasii) to quantify genetic and plastic variation in multiple plant hydraulic traits in both a common garden and across the species’ geographic range in California. We then tested whether site climate could predict either spatial variation in plant traits or spatial variation in end-of-growing season water stress (Pmin).
Results/Conclusions
We found limited among-population genetic variation and a range of plastic variation in hydraulic traits. We also found that site climate was a weak predictor of plant traits, and an extremely poor predictor of end-of-season water stress. Instead, it appears that many sites had become largely decoupled from the local climate by the end of the season due to access to subsurface water sources. Moreover, hydraulic traits were not well correlated with end-of-season water stress. Collectively, our results suggestion that spatial variability in vulnerability to drought may be less than anticipated in this widespread oak species, but that spatial variability in exposure to drought is complicated by subsurface hydrology that is generally less well characterized than climate.