Environmental vs. organismal controls on big sagebrush plant communities

Background/Question/Methods

The impacts of wildlife and livestock grazing on plant communities have always produced mixed effects. Some ecosystems benefit from moderate levels of grazing, while in other systems low levels of grazing can lead to biodiversity and productivity loss. To explain this response widely accepted theoretical models suggest that climate and environmental factors play a large role in determining whether grazing will have positive consequences for biodiversity and productivity; however, meta-analyses reveal a non-significant relationship between grazing impacts and abiotic factors. Many of the field studies that find these contradictory effects were conducted in semi-arid to arid systems where grazing has wide ranging consequences. Our study explores how environmental filtering may disguise grazing impacts particularly in dryland ecosystems where changes in mean annual precipitation influence potential vegetation. To test this hypothesis, we surveyed 13 established grazing gradients around livestock watering points surrounded by big sagebrush (Artemisia tridentata Nutt.) plant communities in southwestern Wyoming. Our watering points span a precipitation gradient and have variable soil texture. At each, we selected two radial transects extending 1.5 kilometers and sampled functional type cover and plant community composition at 12 distances as a proxy for grazing intensity.

Results/Conclusions

Our linear models show the effect of grazing was minimal or non-existent across all sites for functional types particularly when compared to climate and soil texture. For perennial bunchgrasses, cover decreased with grazing intensity (p = 0.039) but only accounted for 4% of the variability. However, models including percent sand (p < 0.001) and precipitation (p < 0.001) increase the variance explained to 39%. Forb cover was not significantly related to grazing intensity and was best explained by a log-transformed model using mean annual precipitation as a predictor (p < 0.001; R2 = 0.416). Big sagebrush, had a weakly positive relationship with grazing intensity for both density (p = 0.020) and cover (p = 0.001). At the species level, some bunchgrass species had a nonsignificant relationship with grazing intensity (i.e. Poa secunda) and were instead highly related to soil texture (p < 0.001). Our results suggest that climate and soil variability exist in a hierarchical relationship with grazing as controls on plant communities and averaging across sites may dampen the effect of grazing in some studies. Future experiments should take care to account for differences in soil texture and climate when determining the directional effect of grazing on plant communities.