Montane coniferous forests in the Intermountain West, USA are bound by upper and lower treelines. Lower treelines in the Intermountain West are generally assumed to be limited by water, but ecophysiological evidence supporting this claim is limited. A widely used metric of drought tolerance in plants is the vulnerability to hydraulic dysfunction due to embolism in the hydraulic pathway. Thus, we measured intraspecific variation in hydraulic parameters and functional traits in Rocky Mountain Douglas-fir (Pseudotsuga menziesii var. glauca) at five elevations along an elevation gradient in southeastern Idaho. We measured leaf- and branch-level vulnerability to drought (focusing on (1) P50leaf and P50branch, i.e., the water potential inducing 50% loss in hydraulic functioning and (2) the slope of the steepest portion of the branch vulnerability curve). We also measured leaf-area-to-sapwood-area ratio and leaf mass per area at each elevation. We hypothesized that Douglas-fir at lower elevations would show less hydraulic vulnerability ( more negative P50 in both branches and leaves and a more gradual slope of the steepest portion of the branch vulnerability curves) to drought than individuals growing upslope and that leaves would be more vulnerable than branches at all sites. Our research provides information about the hydraulic strategies of Douglas-fir—a common lower treeline species in the Intermountain West—from its lower to upper elevation limits.
Results/Conclusions
Branch-level vulnerability to drought increased with elevation. P50branch increased with elevation (R2=0.93, p=0.03, excluding the highest elevation site) and the slope of the steepest portion of the branch vulnerability curves became steeper with elevation (R2=0.93, p=0.01). In trees at lower elevations, declines in hydraulic conductivity are less sensitive to changes in xylem tension, demonstrating that branches from the trees at lower elevation experience less drought-induced hydraulic dysfunction. Leaf-level vulnerability to drought (interpreted from P50leaf) showed no significant relationship with elevation (R2=0.28, p=0.35), but was approximately three times less negative than P50branch on average (-1.56 MPa vs. -4.89 MPa). A less negative P50leaf suggests that the leaves are more vulnerable to drought than branches, which has many ecophysiological implications with respect to whole-tree water transport. There was no significant relationship between leaf-area-to-sapwood-area ratio and elevation (R2=0.024, p=0.40), but leaf mass per area increased with decreasing elevation (R2=0.26, p=0.01, excluding the highest elevation site). This may indicate that trees at lower elevations are growing thicker leaves in response to less water availability.