Spatial and temporal dynamics of plant populations in salt-desert shrub vegetation grazed by sheep.

摘要

I studied the effect of moderate sheep grazing on a shadscale plant community at the Desert Experimental Range, southwestern Utah, USA, using a 61-yr data set with two grazing treatments (yes vs. no), two seasons (spring vs. winter), and two soil types (loamy-skeletal vs. coarse-loamy). I studied precipitation, total species cover, annuals, shrub survival, seedling recruitment, plant succession, and plant spatial relationships.; Precipitation showed high variability (CV = 31%) masking on short-term cycles, resulting in study intervals with average (1935-58), dry (1958-69), driest (1969-75), and wet (1980-94(5)) regimes. Total cover in both grazed and ungrazed pastures increased between 1935 and 1975 before decreasing to 1994. Treatments diverged with time, however, so cover was higher in ungrazed pastures in 1975 and 1994. Individually, Atriplex confertifolia decreased from 1958-94 and Ceratoides lanata from 1975-94. Artemisia spinescens increased in ungrazed pastures from 1935-94, while remaining very low in grazed pastures. Grasses increased from 1935-94 with little grazing effects. Annuals increased from absence in 1935 to 63% frequency in 1994; precipitation may be related to this increase. Grazing and soil type had few long-term or short effects on shrub survival. Similarly, only C. lanata showed a microhabitat effect, with greater seedling survival in vegetated than open patches. Seedling recruitment was positively correlated with precipitation. Only A. confertifolia recruitment responded to grazing; it was higher in grazed pastures. A fuzzy graph analysis showed a moderate grazing effect on succession. Clumped distributions were common and were unaffected by grazing but increased in wet years. Plant establishment occurred disproportionally in sites occupied or formerly occupied by plants, suggesting facilitation. Negative interference, however, was suggested by new recruitment occurring further from larger existing individuals. Moderate grazing had little effect on spatial relationships.; In conclusion, the multivariate approach yielded broader conclusions than any individual factors. Although some factors showed more grazing effects than others, grazing could not completely explain observed changes; climate and inherent plant attributes must also be considered. Management at moderate grazing levels may only play a limited role in shadscale communities.