干旱区生态环境知识资源中心

Factors with variation at broad (e.g., climate) and fine scales (e.g., soil texture) that influence local processes at the plant scale (e.g., competition) have often been used to infer controls on spatial patterns and temporal trends in vegetation. However, these factors can be insufficient to explain spatial and temporal variation in grass cover for and and semiarid grasslands during an extreme drought that promotes woody plant encroachment. Transport of materials among patches may also be important to this variation. We used long-term cover data (19152001) combined with recently collected field data and spatial databases from a site in the northern Chihuahuan Desert to assess temporal trends in cover and the relative importance of factors at three scales (plant. patch, landscape unit) in explaining spatial variation in grass cover. We examined cover of five important grass species from two topographic positions before, during, and after the extreme drought of the 1950s. Our results show that dynamics before, during, and after the drought varied by species rather than by topographic position. Different factors were related to cover of each species in each time period. Factors at the landscape unit scale (rainfall, stocking rate) were related to grass cover in the pre- and post-drought periods whereas only the plant-scale factor of soil texture was significantly related to cover of two upland species during the drought. Patch-scale factors associated with the redistribution of water (microtopography) were important for different species in the pre- and post-drought period. Another patch-scale factor, distance from historic shrub populations, was important to the persistence of the dominant grass in uplands (Bouteloua eriopoda) through time. Our results suggest the importance of local processes during the drought, and transport processes before and after the drought with different relationships for different species. Disentangling the relative importance of factors at different spatial scales to spatial patterns and long-term trends in grass cover can provide new insights into the key processes driving these historic patterns, and can be used to improve forecasts of vegetation change in and and semiarid areas.