OOS 14-6
Community and ecosystem scale responses to precipitation variability and extremes in semiarid grasslands, shrublands and woodlands

Tuesday, August 6, 2013: 3:00 PM
101F, Minneapolis Convention Center
William T. Pockman, Department of Biology, University of New Mexico, Albuquerque, NM
Amanda L. Boutz, Department of Biology, University of New Mexico, Albuquerque, NM
Patrick J. Hudson, Biology, University of New Mexico- Albuquerque, Albuquerque, NM
Jean-Marc Limousin, Biology, University of New Mexico, Albuquerque, NM
Robert E. Pangle, Department of Biology, University of New Mexico, Albuquerque, NM
Jennifer A. Plaut, Biology, University of New Mexico, Albuquerque, NM
Michell L. Thomey, Department of Biology, University of New Mexico, Albuquerque, NM
Scott L. Collins, Department of Biology, University of New Mexico, Albuquerque, NM
Marcy E. Litvak, Department of Biology, University of New Mexico, Albuquerque, NM
Nate G. McDowell, Earth and Environmental Sciences, Los Alamos National Laboratory, Los Alamos, NM
Eric E. Small, Department of Geological Sciences, University of Colorado, Boulder, CO
Background/Question/Methods

Soil water availability is the principal control on plant function in semiarid regions. In the southwestern USA, climate change is expected to generally increase aridity as well as the frequency of severe drought.  To investigate the causes and consequences of plant, community and ecosystem responses to extreme events, we conducted a series of manipulative experiments in grama grassland, creosote bush shrubland and pinon-juniper woodland ecosystems at the Sevilleta LTER.  In each ecosystem, rainout shelters were used to sustain severe drought for five years.  In grassland, precipitation variability was also manipulated by supplementing monsoon rainfall or excluding monsoon rainfall entirely and replacing it with many small events or a small number of large events.  Measurements of cover, productivity, leaf level gas exchange and/or sap flow were used to assess drought impacts and acclimation to experimentally-altered water availability.  The ecosystem consequences of observed pinon mortality were assessed using eddy flux measurements at a nearby ecosystem-scale girdling experiment that simulated pinon mortality during extreme drought.

Results/Conclusions

In all ecosystems, severe drought limited plant function. Grass cover decreased with sustained drought in both Black grama grassland and in a mixed grass-shrub community in the ecotone between grassland and shrubland.  In shrubland, there was little change in shrub cover after five years of drought, although leaf area index decreased.  In pinon-juniper woodland, severe drought led to rapid pinon mortality on hillslope plots, and was associated with with bark beetle activity.  Hillslope juniper exhibited progressive branch dieback leading to slow mortality during years 2-5 of the experiment.  On deeper soils with less slope both pinon and juniper survived for five years before pinon mortality began.  During this extended drought, both pinon and juniper exhibited growth limitations, decreased leaf area and acclimation of gas exchange and hydraulic characteristics.  Ecosystem scale measurements of carbon fluxes following girdling revealed that carbon sequestration decreased by 50%.  The expected increases in respiration associated with decomposition of the dead trees and increases in understory production lagged the girdling treatment by nearly two years while surface temperatures increased due to altered energy balance.  Our results suggest that future changes in precipitation regime and drought frequency will alter plant performance and survival with consequences for community composition and ecosystem structure and function.