COS 232-2 - Impacts of magnitude, timing, and depth of water availability on above and belowground savanna dynamics

Abstract: Tree cover in savannas tends to increase with mean annual precipitation, but we lack an understanding of whether this increase is driven primarily by rainfall intensity, deep water infiltration, or longer growing seasons associated with greater annual precipitation. While increases in deep water availability are known to benefit trees more than grasses, the mechanisms associated with changes in tree growth with increased availability of deeper soil water below the grass rooting zone have not been directly tested. Understanding these mechanisms for tree saplings is particularly insightful because the sapling stage limits tree recruitment in savannas.





The goals of this study are (1) assess how the intensity, frequency, and depth of water availability impacts growth of common savanna tree saplings and (2) understand how variation in water availability affects belowground processes such as root productivity and water partitioning among trees and grasses. To address these goals, we present a novel field experiment manipulating depth, intensity, and timing of irrigation in a semi-arid savanna at Wits Rural Facility, South Africa. Grasses and tree seedlings of six common savanna species were planted in 2018 and water manipulation began in October 2019. Irrigation was used to increase shallow or deep soil water availability, simulate heavy rainfall events, and extend the wet season in different treatments. We measured tree and grass growth and physiology (stomatal conductance) and belowground net primary productivity in response to each treatment. We predicted that treatments that increased deep water availability would favor tree growth, and that differences among trees would reflect distinct growth strategies to mitigate competition with grasses.




Here we show that common tree species show differential growth responses to magnitude, timing, and depth of water availability in the sapling stage, which is the limiting demographic stage for savanna trees. We found that tree species with higher stomatal conductance (Senegalia nigrescens) showed greater growth responses to irrigation treatments than species with low conductance values (Colophospermum mopane). Annual belowground primary productivity was highest in plots that were irrigated during the dry season. Depth of water uptake differed among C. mopane, S. nigrescens, and C₄ grasses, illustrating spatial niche partitioning between sapling trees and grasses. Together, these data provide a comprehensive look at how manipulation of rainfall dynamics impact tree sapling growth, belowground dynamics, and tree-grass competition in lowveld savannas. These data will be used to parameterize models predicting savanna ecosystem responses to future altered rainfall regimes.