Introduction:Sagebrush ecosystems in western North America are being replaced by the invasion of annual grasses,particularly Bromus tectorum.In experimental situations and in localized landscapes,prior studies have do...Introduction:Sagebrush ecosystems in western North America are being replaced by the invasion of annual grasses,particularly Bromus tectorum.In experimental situations and in localized landscapes,prior studies have documented that biological soil crusts(biocrusts)can reduce annual grass presence and that biocrusts are highly vulnerable to physical disturbance.Practical conservation would benefit from verification of these patterms at scales that matter to local economies.This study tests if these patterns appear at a regional scale.Methods:A previously collected data set of vegetation provided sampling of biocrust cover across the Great Basin within the state of Nevada,USA Data were analyzed with non-par ametric methods including odds ratios and generalized aditive models(GAM).Results:From a data set of 608 vegetation plots within the Great Basin ecoregion,proportion of plots with high annual grass cover differed between sites with high versus low biocrust cover(p=0.0015).A negative relationship between annual grass cover and biocrust cover was confirmed with GAM(p=0.009).For a model of biocrust cover,cattle disturbance was found to be an explanatory variable(p<0.00001).Conclusions:The patterns do appear at the regional scale,with high levels of cattle activity corresponding to low cover of biocrusts,and low cover of biocrusts corresponding to high cover of annual grasses.展开更多
文摘Introduction:Sagebrush ecosystems in western North America are being replaced by the invasion of annual grasses,particularly Bromus tectorum.In experimental situations and in localized landscapes,prior studies have documented that biological soil crusts(biocrusts)can reduce annual grass presence and that biocrusts are highly vulnerable to physical disturbance.Practical conservation would benefit from verification of these patterms at scales that matter to local economies.This study tests if these patterns appear at a regional scale.Methods:A previously collected data set of vegetation provided sampling of biocrust cover across the Great Basin within the state of Nevada,USA Data were analyzed with non-par ametric methods including odds ratios and generalized aditive models(GAM).Results:From a data set of 608 vegetation plots within the Great Basin ecoregion,proportion of plots with high annual grass cover differed between sites with high versus low biocrust cover(p=0.0015).A negative relationship between annual grass cover and biocrust cover was confirmed with GAM(p=0.009).For a model of biocrust cover,cattle disturbance was found to be an explanatory variable(p<0.00001).Conclusions:The patterns do appear at the regional scale,with high levels of cattle activity corresponding to low cover of biocrusts,and low cover of biocrusts corresponding to high cover of annual grasses.
