Abstract

Climate change models for many ecosystems predict more extreme climatic events in the future, including exacerbated drought conditions. Here we assess the effects of drought by quantifying temporal variation in community composition of a complex montane meadow landscape characterized by a hydrological gradient. The meadows occur in two regions of the Greater Yellowstone Ecosystem (Gallatin and Teton) and were classified into six categories (M1–M6, designating hydric to xeric) based upon Satellite pour l'Observation de la Terre (SPOT) satellite imagery. Both regions have similar plant communities, but patch sizes of meadows are much smaller in the Gallatin region. We measured changes in the percent cover of bare ground and plants by species and functional groups during five years between 1997 and 2007. We hypothesized that drought effects would not be manifested evenly across the hydrological gradient, but rather would be observed as hotspots of change in some areas and minimally evident in others. We also expected varying responses by plant functional groups (forbs vs. woody plants). Forbs, which typically use water from relatively shallow soils compared to woody plants, were expected to decrease in cover in mesic meadows, but increase in hydric meadows. Woody plants, such as Artemisia, were expected to increase, especially in mesic meadows. We identified several important trends in our meadow plant communities during this period of drought: (1) bare ground increased significantly in xeric meadows of both regions (Gallatin M6 and Teton M5) and in mesic (M3) meadows of the Teton, (2) forbs decreased significantly in the mesic and xeric meadows in both regions, (3) forbs increased in hydric (M1) meadows of the Gallatin region, and (4) woody species showed increases in M2 and M5 meadows of the Teton region and in M3 meadows of the Gallatin region. The woody response was dominated by changes in Artemisia spp. and Chrysothamnus viscidiflorus. Thus, our results supported our expectations that community change was not uniform across the landscape, but instead could be predicted based upon functional group responses to the spatial and temporal patterns of water availability, which are largely a function of plant water use and the hydrological gradient.