Monday, August 4, 2008 - 4:40 PM

COS 5-10: Local biotic interactions drive community dynamics in a Sonoran Desert perennial plant community

Brad J. Butterfield1, John M. Briggs1, Julio L. Betancourt2, Raymond M. Turner1, and Janice M. Bowers1. (1) Arizona State University, (2) U.S. Geological Survey

Background/Question/Methods

Temporal dynamics of desert perennial plant communities are highly erratic, and weakly reflect variation in climatic variables, suggesting the importance of local biotic interactions in driving community dynamics.  To assess the direct and indirect effects of climate, competition, and facilitation on recruitment and mortality, we analyzed 75 years of permanent plot data from Tumamoc Hill on the grounds of the University of Arizona Desert Laboratory near Tucson, AZ.  The perennial vegetation in the plots is of the Arizona Upland floristic group of the Sonoran Desert, consisting of small trees, shrubs, and cacti .  Recruitment, mortality, and changes in cover were extracted from digitized, spatially-explicit maps using GIS, allowing for determination of species’ recruitment niches.  We used structural equation modeling and other multivariate methods to determine the relative importance of climatic versus local biotic variables on mortality, recruitment of colonizer species, and recruitment of facilitation-dependents.

Results/Conclusions

Recruitment of colonizer species was positively correlated with precipitation, particularly fall cyclones, however facilitation-dependent species exhibited no direct effect of precipitation.  Instead, the latter responded strongly to increases in nurse plant cover.  Increases in overall density and cover also led to higher rates of mortality, indicating a strong competitive effect, whereas drought had a less-substantial effect.  Our results support the relatively strong influence of local biotic interactions on community dynamics, and a less-pronounced direct effect of climatic variation.  Due to the multitude of feedbacks and time lags produced by these complex interaction networks, we suggest a model of community dynamics based on the relative abundance of different functional types within a community, allowing us to extend the results of our study to other desert assemblages.