COS 133-3 - CANCELLED - Contingent responses of exotic annual grass invasion to grazing across Northern Great Basin rangelands

Background/Question/Methods

Exotic annual grasses such as cheatgrass (Bromus tectorum) are an increasingly widespread management challenge throughout Great Basin rangelands: displacing native perennial vegetation, threatening forage availability for cattle and wildlife, and fueling more frequent and catastrophic wildfires. Although many environmental correlates of annual grass prevalence have been identified, there is still substantial uncertainty about what makes some sites more resistant to invasion than others. A key question for land managers is to what extent cattle grazing influences annual grass invasion. Evaluations of grazing impacts are often confounded by treating grazing as a binary rather than considering gradients of grazing pressure, or by not considering how effects of grazing may vary with fire history and environmental context. Here, we conducted a large-scale field survey of 267 vegetation plots, sampling plant species cover and density, site characteristics, and herbivore dung counts in burned and unburned pastures on public grazing lands throughout Eastern Oregon and Southern Idaho. To examine existing gradients of grazing on the landscape we used a piosphere design with plots stratified by distance to livestock water sources. We used nested multivariate models to examine drivers of annual grass invasion across scales, and asked whether effects of grazing depended on environmental context.

Results/Conclusions

Although annual grass invasion is commonly associated with fire, invasive annual grasses were present in most plots regardless of fire history. Burned vs. unburned plots were more strongly distinguished by perennial grass vs. shrub abundance, respectively, with annual grasses slightly more abundant in burned plots but present to varying degrees in both. Dung counts of cattle and other herbivores were not as tightly correlated with distance from water source as expected, suggesting an attenuated piosphere effect in the context of other landscape variation. Using log-transformed dung counts as an indicator of variation in grazing pressure among plots, we found that the correlation with grazing varied with abiotic context. For example, annual grass cover increased in plots with more grazing pressure within lower-elevation pastures, but decreased with grazing pressure in pastures at higher elevations. In these frequently grazed ecosystems, patterns such as elevational gradients in annual grass invasion may relate to not only the direct response of plant communities to abiotic variation, but also context-dependence in grazing impacts that exacerbate positive or negative interactions. Understanding how grazing affects vegetation differently across the landscape can help inform grazing management to mitigate annual grass invasion in the Great Basin.