Long-term effects of rangeland improvements on cattle distribution: A ranch-scale case study

Resource matching can be a common assumption that underlies anthropogenic rangeland improvements for wildlife or livestock. Countless resources are spent on targeted vegetation manipulations for both livestock and wildlife, but selection for improved areas has not been well quantified in livestock. This case study investigates whether cattle exhibit preferential use of historically improved rangeland sites within sagebrush steppe ecosystems, and whether these effects are relevant long-term compared with inherent landscape heterogeneity. We used a generalized linear mixed-effects resource selection function on Global Positioning System points collected at a fix rate of 2 h from a subset of animals grazing throughout 14 pastures (totaling 13 079 ha) over 3 yr to investigate cattle distribution in relation to historical rangeland improvements. These consisted of various methods for reducing woody plant cover and increasing forage production, most followed by grass and legume reseeding. Overall selection for treated areas was positive (0.16, P < 0.0 0 01), but highly variable (pasture estimates range from-1.81 to 0.77). Differences in enhanced vegetation index (EVI) accounted for 25% of the variation in treatment selection, whereas herbaceous production did not have an appreciable effect on treatment selection. Within individual pastures, the effects of landscape variables (distance to water, slope, herbaceous production, EVI, tree cover, and shrub cover) were more variable than treatment effects and on average 2.4-fold more influential to selection than anthropogenic treatments. Still, the combined effects of all landscape variables accounted for only 8% of the variation in site selection. The observed inconsistency among selection patterns underscores the complexity of large herbivore resource selection behavior, and the statistical challenges to modeling that behavior. These results provide limited and inconsistent support for the hypothesis that large herbivores select long-term (5–20 yr post establishment) for range improvements in relation to their forage quality/quantity in a resource-matching manner.