Forests and Water - Friends or Foes? : Hydrological implications of deforestation and land degradation in semi-arid Tanzania

Abstract: This is a study of landscape hydrology; a type of hydrology which focuses on large sca1e implications of hydrologica1 processes as a function of hydroclimate, surface and sub-surface land properties, and the effect of human activities. The focus of the thesis is groundwater recharge and dry season flow following deforestation and land degradation. This is a highly controversial issue. In many countries, inc1uding those in humid-temperate regions, an a1most intuitiona1 feeling causes many to many believe that springs and streams will flow more steadily and at higher rates in a forested as compared to a deforested catchment. The trees are not seen as giant pumps transferring stored soilwater to the atmosphere, but as creators of an environment of "sponges" that can "buffer" tbe runoff and support dry season flow. Both of these views are simplistic. The aim of this thesis is to try to separate myth from science in regard to forests and water. In tbe study area in Babati District in Tanzania a multi-component research approach was attempted. Two catchments, one forested and one deforesteddegraded, were studied regarding soj] properties, runoff and groundwater recharge. This was done botb in tbe field and with tbe use of two computer models: one simulating groundwater recharge as a function of rainfa1l variability, and one simulating hydrological implications of massive land cover conversion on tbe flooding ofnearby Lake Babati. Three major findings came out of the study. The first is that most forested catchments (in various hydroclimates and landscapes) will increase the runoff following deforestation (due to less evapotranspiration). This is well-established knowledge, but it also depends on the actual conditions at hand. These conditions are defined as hydroclimate, soil texture and slope. In humid-temperate climates with coarse soils on flat land, the conditions strongly favor increased runoff following deforestation. However, in dry tropical regions with fine textured soils on billy ground, and where deforestation also implies land degradation, less dry season flow is likely to develop after a considerable adjustment period has been allowed. Secondly, the prevalence of preferential flow in a forest soil, as compared to a compacted and eroded soil, must be a key component in an explanation of why more dry season flow can emerge from a forested as compared to a deforested catchment in the dry tropics. Thirdly, there are several aspects of semi-arid and arid tropical hydrology which make comparisions with humid-temperate regions difficult and require special attention in the management ofwater resources in the dry tropics.