Experimental impacts of grazing on grassland biodiversity and function are explained by aridity

Grasslands, covering over 40% of the terrestrial surface, provide crucial biodiversity and ecosystem services. Livestock grazing, the most widespread land use globally, significantly impacts these ecosystems. While grazing is essential for food production and economies, concerns exist about its sustainability, especially given the high biomass of domestic livestock in many systems. Understanding the long-term effects of grazing on biodiversity and ecosystem function is crucial for sustainable management. Four major uncertainties hinder this understanding: 1) lack of experimental evidence on the interaction between aridity and grazing impacts on biodiversity and multifunctionality across different sites; 2) reliance on meta-analyses and unstandardized field investigations instead of regional paired comparison experiments; 3) limited long-term experimental data to capture the time-lag responses of ecosystems; and 4) unclear understanding of the relative contributions of aboveground (plants) and belowground (soil microbes) biodiversity to ecosystem multifunctionality under long-term grazing. This study addresses these gaps using a large-scale, multi-site experiment.

Previous research on grazing impacts has limitations. Most studies are conducted at single sites, failing to capture the crucial role of aridity, a key determinant of dryland ecosystems. Existing work often relies on meta-analyses or unstandardized field data, lacking the rigor of replicated regional experiments needed to establish causal relationships. Short-term grazing experiments are common, ignoring potential time-lags in ecosystem responses. The relative importance of aboveground (plant) and belowground (soil) biodiversity in supporting multiple ecosystem functions under long-term grazing remains unclear. This lack of robust experimental evidence hampers effective management strategies for grazing, the most extensive land use globally.

This study utilized a network of 10 experimental sites across an aridity gradient in Northern China, encompassing three major grassland types (meadow steppe, typical steppe, and desert steppe). Paired grazed and ungrazed plots, established with >10 years of livestock exclusion, were sampled to assess the long-term effects of grazing. Eleven ecosystem functions were measured: above- and below-ground biomass, plant community nitrogen and phosphorus, soil organic carbon, soil available nitrogen, microbial biomass carbon and nitrogen, decomposers, pathogen control, and mycorrhizal mutualism. Aboveground plant diversity and belowground soil microbial diversity (bacteria, fungi, protists) were assessed using high-throughput sequencing. Multifunctionality was calculated using three approaches: averaging multifunctionality, weighted multifunctionality (accounting for correlated functions), and multi-threshold multifunctionality. Linear mixed effects models were used to analyze the interactive effects of grazing and grassland type (related to aridity) on biodiversity and multifunctionality. Regression models explored the relationship between aridity and the effects of grazing. Additional statistical analyses examined the relative contributions of plant and soil biodiversity to multifunctionality under grazed and ungrazed conditions, controlling for climate variables through structural equation modeling.

The study found a significant interaction between grazing and aridity in influencing biodiversity and multifunctionality. Long-term grazing had no significant effects in meadow steppes (lower aridity) but significantly reduced biodiversity and multifunctionality in desert steppes (higher aridity). The negative effects of grazing increased with increasing aridity. Long-term grazing did not affect plant diversity in higher aridity steppes, but soil biodiversity was significantly reduced in desert steppes. Importantly, the study demonstrated a shift in the drivers of multifunctionality from plant diversity in ungrazed grasslands to soil biodiversity in long-term grazed grasslands. This shift was consistent across different measures of multifunctionality and individual functions. The findings were robust even after controlling for environmental factors like mean annual temperature and precipitation using Structural Equation Modeling.

The results provide strong experimental evidence supporting the hypothesis that the long-term impacts of domestic herbivore grazing on grassland biodiversity and multifunctionality are significantly influenced by aridity. The study advances current observational-level knowledge by providing long-term experimental data from a network of sites. The observed shift from plant diversity to soil biodiversity as a driver of multifunctionality under long-term grazing highlights the crucial role of soil biodiversity conservation in grazed grasslands, especially in arid ecosystems. The findings emphasize the need for careful management of livestock grazing in arid grasslands, where multifunctionality is vulnerable and highly dependent on soil biodiversity.

This study demonstrates that the effects of long-term grazing on grassland biodiversity and multifunctionality are strongly dependent on aridity. Higher aridity intensifies the negative effects of grazing, particularly impacting soil biodiversity. The shift in the importance of soil biodiversity for ecosystem multifunctionality in grazed systems underscores the need for conservation strategies focused on belowground biodiversity. Future research should investigate these effects over even longer timescales and broader spatial scales, incorporating various grazing intensities and management practices. Adaptive grazing management strategies that consider aridity levels are essential for the sustainable management of grasslands worldwide.

While the study represents a significant advance in understanding grazing impacts, some limitations exist. The study was conducted in a specific region of Northern China, and the findings might not be universally applicable to all grassland ecosystems. The specific livestock types and grazing intensities might influence the results. The study focused on a specific time point, and future research should investigate temporal dynamics in biodiversity-function relationships. Finally, the relatively small number of sites may limit the generalizability of the results.