Mixed grazing maintains the spatial invariability of alpine grassland biomass by fostering a more diversified community structure and spatial heterogeneity

Using a 3-year field experiment in an alpine grassland implementing moderate grazing (medium grazing intensity) across various grazing types (including yaks grazing, sheep grazing, and mixed grazing), we quantified the impact of different combinations of grazing types on the community spatial invariability of biomass and underlying processes including spatial covariation and invariability at species level.

To conduct the grazing experiment, twelve equal sized plots measuring 20 m × 30 m were established, with flat topography and similar environmental conditions. These plots were enclosed in 2020 to prevent livestock grazing before the beginning of the grazing experiment. We employed a randomized block design and divided into three blocks. Within each block, four grazing types were randomly arranged: no grazing (NG), yaks grazing (YG), sheep grazing (SG), and mixed grazing of yaks and sheep (MG). The grazing experiment started in 2021 and lasted three years (2021, 2022, and 2023). A consistent moderate grazing strategy was applied in three grazing types, with grazing intensity regulated to a medium level, resulting in herbivores consuming approximately 50% of herbage production (Liu et al., 2015). Using the NY/T 635–2015 calculation of the rangeland carrying capacity in China's agricultural industry standard (Ministry of Agriculture of the People’s Republic of China, 2015), the daily intake of an adult sheep was determined to be 1.25 kg of hay, equivalent to 1/4 of that of an adult yak. We have uniformly stipulated that each grazing period lasts 2 days. The number of yaks and sheep in each grazing treatment was determined by combining pregrazing aboveground biomass and daily herbivore intake. In the case of mixed grazing, the ratio of yaks to sheep was 1:4. Grazing was carried out once a year during the cold season and twice a year during the warm season.

Three quadrat samples were randomly selected in each plot for investigation. These quadrat samples were situated approximately 5 m from the fence boundary and at least 5 m apart from each other. All measurements were made during the growing season, with quadrat samples measuring 50 cm × 50 cm. Data recorded for each quadrat sample included the name of species and total number. Then, residual living aboveground tissue of all vascular plants (i.e. green plant biomass) was harvested at ground level with scissors to represent the aboveground biomass (AGB). Plant samples were then oven-dried at 65 °C for 48 hours and weighed to determine biomass. Overall, the data were collected over a 3-year period, from July to August between 2021 and 2023, resulting in a total of 216 quadrat samples (3 years × 2 months × 4 treatments × 3 replicates × 3 quadrat samples).

Methods
See README for data.pdf for more details on Methods.