COS 139-9
Combined influence of soil moisture and vapor pressure deficit on productivity of US forests

Friday, August 14, 2015: 10:50 AM
321, Baltimore Convention Center
Mallory L. Barnes, School of Natural Resources and the Environment, University of Arizona, Tucson, AZ
David J.P. Moore, School of Natural Resources and the Environment, University of Arizona, Tucson, AZ
David D. Breshears, School of Natural Resources and the Environment, University of Arizona, Tucson, AZ
M. Susan Moran, Southwest Watershed Research Center, USDA, ARS, Tucson, AZ
Background/Question/Methods

Forests across the globe are at increasing risk of drought- and heat-induced mortality, and future climate change is likely to exacerbate this phenomenon. Global-change-type drought characterized by reduced soil moisture (θ) and increased atmospheric demand (i.e., vapor pressure deficit; VPD) has already been implicated in forest die-off in the Southwest United States.  Yet, there is limited understanding of the functional response of forest production to these new hydroclimatic conditions. Extreme VPD and θ conditions both cause plants to close stomata and limit transpiration to conserve water; however the mechanism for each response differs. We hypothesized that there is set of VPD and θ conditions that is conducive to high forest productivity, unique to different forest ecosystems with limits related to climate, life history and biogeochemistry.  Further, as θ and VPD vary with predicted climate change, these conditions for high forest productivity will shrink, offering limited opportunity for carbon uptake and forest production during the growing season.  Over time, this lost opportunity could lead to a decline in forest health and ultimately to regional mortality. This hypothesis was tested using decade-long measurements of gross primary productivity (GPP), θ, and VPD at AmeriFlux sites in forests across the United States.

Results/Conclusions

We quantified forest responses for non-zero daily GPP values in VPD/ θ space over a recent 10-year period at each site.  The years with the greatest number of days when GPP was above the mean were considered the most conducive to forest productivity. As the range of climate conditions conducive to GPP>0 narrowed, there were fewer days with GPP above the mean.  The number of days/year with high GPP could not be predicted with just VPD or θ alone, but was a function of the varying combination of the two.  Our results show there is a set of VPD and θ conditions that is conducive to high forest productivity, unique to each site. The co-constraints VPD and θ on forest production are often overlooked in studies using temperature and precipitation to characterize climate conditions. We find that warm droughts (low θ, high VPD) and cool droughts (low θ, low VPD) show distinctly different effects on forest productivity. We conclude that warm droughts in particular lead to severe reductions in productivity, potentially increasing the risk of eventual mortality.