Intermediate tree cover can maximize groundwater recharge in the seasonally dry tropics

Two-thirds of the world’s population may live in water-limited regions by 20251. In Africa about 340 million people already lack access to adequate and hygienic sources of water, such as groundwater2. Limited water constrains food production, nutrition, and health as well as impacting opportunities for education, work, and improved livelihoods. Reliable access to clean water is essential for achieving the UN Sustainable Development Goals. Forests have often been described as ‘sponges’ storing rain water and slowly releasing it to maintain groundwater and streams during dry periods3,4,5. Formerly, this sponge theory and related ideas motivated policies aimed at conserving and restoring forests5,6,7. In recent decades, however, these ideas have lost credibility as studies show that forest clearance generally leads to increased and afforestation to reduced water yields3,8,9. Therefore a contrasting trade-off theory–in which more trees means less water–has become the dominant paradigm. This trade-off theory predicts that as tree densities increase, water losses from transpiration and interception dominate their hydrological effects6,8,10. In the tropics the trade-off theory rests on limited evidence. The few available studies compare extremes: open land versus closed forest, or grasslands versus dense plantations11. Despite the recognition that trees can improve soil hydraulic conductivity and reduce overland water flow4,12,13, and other findings that question the generality of the trade-off theory14,15,16, we find no available data on the relationship between tree cover and water yields at intermediate tree densities, and few data concerning the specific mechanisms that determine groundwater reserves and dry season stream flows11,12,17. The neglect of intermediate tree cover is a striking omission given the importance of such open vegetation in terms of extent and biodiversity18, and the fact that it supports many of the world’s poorest people. In Africa, there are 350 Mha of open and fragmented forests and 514 Mha of other wooded lands (including savannah, agroforests etc.)19–more than the area under closed forest and plantations (277 and 8 Mha, respectively). Such open vegetation also plays a major role in the global carbon balance; regions with 10 to 30% tree canopy cover are estimated to store 23% of the total forest biomass carbon stock (above- and belowground) in sub Saharan Africa and 15% of the total forest carbon stocks for the global tropics20. Tree planting is, or would be, a major element in many climate mitigation projects, in efforts to combat desertification, and in livelihood focused development proposals seeking to improve access to firewood and other products. But the trade-off theory has reduced the application of such planting projects due to concerns that these efforts would jeopardize scarce water resources6,8,9,10,21,22. Here we present and test an optimum tree cover theory for groundwater recharge that can reconcile the available scientific evidence and contrasting perceptions about forests and groundwater in the seasonally dry tropics (Fig. 1). We hypothesized that under conditions typical of the seasonally dry tropics an intermediate tree cover maximizes groundwater recharge. Below this optimum cover, the hydrological benefits gained from more trees outweigh their extra water use, while at higher values of tree cover the water use from additional trees exceeds any positive effect they might have on groundwater recharge (Fig. 1). We recognize that the tree cover value where this optimum occurs depends on various factors including tree species, local soil and climatic conditions. But before considering these influences we needed to evaluate our theory. We chose a common African semi-arid landscape known as “agroforestry parkland” where water shortage is a recognised livelihood constraint23. Parklands, in which annual crops are grown under scattered trees, constitute the most extensive farming system in semi-arid West Africa23. Indeed, about 1.9 million km2 (47%) of the total agricultural land in Sub-Saharan Africa has a tree cover above 10%24. We want to know how such intermediate tree cover influences groundwater recharge. Figure 1 Conceptual water budget of the optimum tree cover theory.