Site-specific biogeochemical response to livestock grazing and climate change differs across four continents

Drylands make up 40% of terrestrial ecosystems and hold up to 20% of the global soil organic carbon pool. Most semi-arid drylands are, to some extent, grazed by livestock. However, the impact of livestock grazing on carbon cycle dynamics over large spatial and temporal scales remains uncertain, especially as the effects of climate change become more pronounced. Thus far, there has been little work, which has explored how site-specific land management may interact with localized shifts in climate to affect biogeochemical processes in dryland ecosystems globally, particularly in the tropics. We explored, using simulation modeling, how grazing intensity and projected climate change may impact biogeochemical dynamics in dryland regions in North America, South America, Asia, and Africa. Our simulation results showed a site-specific biogeochemical response to livestock grazing and climate change, even across ecologically similar dryland systems. In sites that had smaller projected shifts in climate (i.e., North and South America), heavy grazing decreased soil carbon inputs, outputs, and storage. In the other two sites, particularly in the African site, shifts in climate had the largest impact on simulated biogeochemical processes, with a projected 20% decrease in the soil organic carbon pool in the African site by the end of the century. Our study highlights the importance of considering how localized shifts in climate may affect dryland ecosystem function as this may overwhelm land management effects over longer time scales. Our work also suggests that more research is needed to better understand how small-scale, site-specific sensitivity to climate change and land use may influence dryland carbon cycle dynamics at the global scale, particularly in tropical regions.

Methods
Simulation model files for DAYCENT for four dryland sites located in North America, South America, Asia, and Africa. Files in ‘Initial Model Runs’ used observed data for the model to reach equilibrium and are the base input files used for further simulations. Files in ‘Simulation Runs’ were the simulations through the end of the century, altering grazing intensity and SSP scenario. Climate data (precipitation and temperature datasheets) were compiled using observed historical data from nearby weather stations and downscaled CMIP6 climate data for three SSP scenarios provided by Predictia Solutions. Simulated soil organic carbon and nitrogen, net primary productivity, and greenhouse gas fluxes are model output from DAYCENT.