Drylands are socioeconomically critical systems that house the majority of the world's livestock and a rapidly growing fraction of the human population. Climate models suggest decreases in mean annual precipitation (MAP) and an increase in extreme events (e.g., droughts and floods) in many drylands for the future. Despite the importance of drylands to Earth system processes and human livelihoods, understanding of the controls over ecosystem processes in these systems lags far behind our understanding of wetter systems. Similarly, relatively little is known about the consequences of livestock management to ecosystem pools and processes (e.g., soil carbon pools and stability) in the context of changes in climate, particularly in Africa. Rainfall gradients across areas with similar soil offer the possibility to explore ecological responses to climate. Previous studies across rainfall gradients often focus on 'pristine' locations with minimal human impact; these studies may not reflect patterns typical of present-day heavily human-impacted areas or future patterns in a more densely populated world.
We set up a study across a rainfall gradient in the Namibian portion of the Kalahari Desert. We selected four rainfall locations across the gradient. All rainfall locations were on Kalahari sands, with MAP ranging from 200 mm in the south to 550 mm in the north. Within each rainfall location, we selected four villages in areas with communal grazing. Villages were centered around wells, leading to decreasing grazing intensity with increasing distance from village centers. We assessed woody vegetation cover, species composition, and soil carbon.
Woody cover was lowest at the driest rainfall location (mean = 14%). There was little difference in percent cover among the four wetter sites (27-33% cover), although the wetter sites were typically characterized by large trees in comparison to small shrubs at the drier sites. There was a log-linear relationship between distance from village and woody cover at the two sites in the middle of the gradient, where cover stabilized by 1.2 km from the village centers. There was little response to distance at the wet and dry ends of the gradient. Soil carbon was influenced by woody vegetation, with low carbon content in canopy interspaces and low rainfall sites.
An understanding of how ecosystems respond to climate change in the context of human land use is critical for developing a predictive understanding of future trajectories of these systems, implementing responsive and sustainable land management strategies, and developing informed mitigation or adaptation plans.