Characterizing Dryland Surface Hydrological Dynamics Using Ecohydrological Modeling and Geophysical Observations

The two greatest challenges facing dryland ecosystems are: 1) recent and rapid changes in land use and 2) the predicted effects due to climate change, which have potentially devastating consequences on the 2.5 billion people living in drylands around the world. The situation is particularly sensitive in sub-Saharan Africa, where 800 million people are reliant on local crop and animal agricultural food production. The following dissertation will address three main objectives: 1) providing research that focuses on agricultural systems where a direct coupling exists between the natural distribution of vegetation and livestock instead of crop based agricultural systems, 2) advancing the theoretical understanding of the organization of dryland vegetation patterns by developing novel ecohydrologic models, 3) advancing geophysical methods towards providing hydrological datasets at novel spatial and temporal scales in drylands. In Chapters 2 and 5, two novel ecohydrologic models are developed that utilize the resource tradeoff hypothesis on the organization of dryland vegetation patterns. Chapter 2 presents changes in the regional distribution of woody species in central Kenya varying rainfall patterns. Model results of future rainfall scenarios indicate expansion of the most drought tolerant species and reduction of the species with the highest water use. Chapter 5 presents results from a spatially explicit ecohydrological model used to understand the spatial patterns of vegetation on hillslopes. The model results suggest a switch in optimal vegetation patterns, from highly organized vegetation clusters to a random distribution of vegetation, by varying two model parameters that control the length scales of plant water facilitation and competition. Chapters 3 and 4 each present novel soil moisture datasets collected with different geophysical methods. In Chapter 3, direct-current resistivity is used to collect high-temporal resolution soil moisture measurements in a tank laboratory setting. The dataset supports the potential use of the method to field settings and the use of a semi-analytical solution to model subsurface flow following a rapid infiltration event into a large macropore. In Chapter 4, patch water balance measurements and electromagnetic induction maps are presented to characterize the spatiotemporal redistribution of soil moisture in central Kenya. The datasets were essential in identifying scale-dependent positive feedback that may have led to the recent proliferation of the undesirable succulent S. volkensii. The continued proliferation of S. volkensii in degraded landscapes threatens the sustainability of this pastoralist ecosystem. By integrating the positive feedback into ecohydrological models, future work has the potential to provide land manages with diagnostic tools to predict system behavior and thus guide decisions with the long-term goal of rehabilitating this ecosystem back to its tree-grass state.