Monday, August 6, 2007

PS 3-45: Effect of experimental drought and water addition on community dynamics in three aridland plant communities

Jennifer Johnson1, Scott L. Collins1, Eric E. Small2, and William T. Pockman1. (1) University of New Mexico, (2) University of Colorado

Aridland ecosystems are subjected to wet and dry cycles that vary seasonally and annually. Evidence suggests that many aridland ecosystems are highly responsive to wet years and resistant to short term drought periods, but severe chronic drought may greatly alter community structure and ecosystem function. To test community response to different precipitation regimes, we experimentally altered precipitation inputs in desert grassland, creosotebush shrubland and mixed grass-shrub communities at the Sevilleta LTER site in central New Mexico. Manipulations (N=3) included a 50% reduction in ambient annual rainfall starting in 2001 and a 50% rainfall addition treatment based on the long-term average of 250mm per year starting in 2005. Plant community composition was measured seasonally in nine permanently located 1m2 quadrats in each replicate of each treatment. We report here on changes in plant species composition from 2002-2006. Seasonal and annual rainfall differed among years: non-monsoon precipitation was 50.8, 55.5, 176.0, 214.0, and 29.8 mm (long-term mean is 98.0 mm) and monsoon precipitation was 181.6, 70.0, 114.7, 163.5 and 257.9 mm (long-term means is 140.0) in 2002-2006, respectively. Rainfall reduction treatment had little impact on total plant cover in creosotebush shrubland, but richness averaged 1-2 species lower in drought plots compared to control and water addition plots. Total plant cover was nearly half that of control plots and water addition treatments under experimental drought in the mixed grass-shrub community but both richness and cover increased rapidly in 2006, a year of above average monsoon precipitation. Chronic drought also reduced cover in the grassland community, but drought treatments had higher plant species richness, on average than controls or rain addition plots. Results suggest that desert plant communities differentially respond to chronic drought and interannual precipitation variation, with greatest changes in cover and richness occurring in areas dominated by herbaceous vegetation.