Determining vegetation responses to global-change-type drought extremes: will warming amplify tree die-off?

Principle Investigator: David D. Breshears, University of Arizona, daveb@email.arizona.edu

Abstract:: This project proposes to develop an approach useful in the prediction of potential effects of climate change on semiarid piñon-juniper woodlands.  The study will assess tree mortality in concert with an ongoing, leveraged study being conducted in Biosphere 2.   The results of the study will be a comprehensive assessment of tree mortality response to extreme drought events spanning drought conditions of dry, warm-dry and hot-dry and integrated with field observations of mortality.

Location: Our factorial, manipulative field experiment will be conducted along the C. Hart Merriam Elevation Gradient, located in northern Arizona near Flagstaff and operated by Northern Arizona University. The approximate location of the area where trees will be removed from the Babbitt Ranch property is N 35.57 W111.96 at an elevation of 2002 m. The lower elevation site is located on Babbitt Ranch property as well and is located at N35.69 W111.43 at an elevation of 1542 m. The lower elevation site for this experiment is also the lowest elevation site on the C. Hart Merriam Elevation Gradient, operated by Northern Arizona University, and has a full met station at that site (http://www.mpcer.nau.edu/gradient/gbd_mesonet.html) and several other projects in place that are collecting various sets of ecological data (see attached letter of support from Merriam-Powell Director Neil Cobb). We will integrate the experiment results with other field observations from New Mexico sites and with basic modeling evaluations using FOREST-BGC.

Hypotheses: Our hypotheses stem from the two important questions raised above: 1) can the mortality threshold of dominant tree species be predicted based on a period of time-integrated water stress prior to mortality, and if so, at what value does such a threshold occur?; and 2) do warmer temperatures exacerbate time-integrated water stress sufficiently to hasten tree mortality, such that global-change-type drought can trigger more extensive vegetation die-off? Specifically, we hypothesize that 1) timeintegrated plant water stress, reflected in soil water content and plant water potential, triggers piñon pine die-off after an extended period, and 2) a temperature increase (~2-4 oC) is sufficient to exacerbate mortality of adult trees. Notably, the existing dynamic global vegetation models explicitly or implicitly assume that some level of time-integrated water stress triggers die-off, even though this threshold has not been quantified or tested. Even though other factors such as bark beetle infestations and stand densities could influence patterns of tree mortality (Shaw et al. 2005, Breshears et al. 2005), most researchers point to plant water stress as the underlying cause of die-off patterns (Waring and Cobb 1992, Allen and Breshears 1998, Breshears et al 2005).

Methods: As introduced within the context of the hypotheses, we will use a manipulative, factorial field experiment with mature piñon trees with treatments of precipitation (supplemental, ambient, extreme drought), temperature (ambient, +2oC), and mechanical disturbance (in situ, transplanted). Below-canopy rainout shelters will be applied to individual trees both in existing and transplanted locations, where the transplants serve as a surrogate for warmer climate conditions (an approach similar to that used by others at NAU). Our experimental design for the field experiment will utilize the largest available tree transplanting spade in the Southwest to transplant mature trees (> 50 years old). Trees will be transplanted both at the ambient site from where they were originally located and to the warmer site, located at a lower elevation utilizing the Merriam Powell Environmental Gradient. The climate at the Warmer Site corresponds to a small but relevant temperature increase (~2oC). At both sites, rainout shelters will be applied to half of the transplanted trees. Watering treatments and comparison of transplanted vs. in situ trees will allow us to correct for transplant and background water stress effects on mortality.



last updated: 28 February 2008 PLH