Interactions between species richness, herbivory and precipitation affect standing biomass in Mongolian rangelands

QUESTIONS: Livestock management in rangelands depends on the production of plant biomass. Biomass production is driven by the temporal and spatial variability in precipitation, but our understanding of how precipitation variability mediates grazing effects on biomass production is still fragmented. Along a 600-km precipitation gradient we extracted biomass data to ask the questions: (a) what are the effects of grazing intensity on biomass production; (b) does grazing intensity interact with plant species richness to affect biomass production; and (c) how do plant functional groups respond to grazing and precipitation? LOCATION: Mongolia. METHODS: Biomass was sampled along 15 grazing intensity transects within the precipitation gradient over two consecutive years. We modeled spatial variability in above-ground plant biomass using mixed-effects models. Normalized difference vegetation index (NDVI) data were combined with field-sampled biomass data to correct for inter-annual precipitation variation. The effects of species richness were modeled with respect to possible interactions with grazing intensity, and the composition of plant functional groups was modeled with respect to possible interactions between grazing intensity and precipitation. RESULTS: Biomass was negatively correlated with grazing intensity and this effect increased as precipitation increased. Biomass was positively correlated with species richness in both years, but the strength of this effect and the interaction between species richness and grazing intensity differed between 2014 and 2015 in line with highly variable precipitation between both years. The plant functional groups grasses, sedges, legumes, wormwood and forbs had contrasting responses to grazing and precipitation. CONCLUSION: Biomass production in drylands is more vulnerable to changes in precipitation variability and grazing intensities in relatively moist and productive rangelands than in dry and unproductive ones. Future rangeland management needs to address potentially increasing precipitation variability in order to promote desired forage plants, and to preserve the positive effects of biodiversity for biomass production