Restoring pre-development forest cover to provide watershed services has become a major driver of conservation, with significant uptake in Latin America and recent expansion into Africa and China. The services desired from such conservation include an improved water yield and moderation of peak and dry-season river flows. However, it is unclear whether these programs will provide the desired benefits.
We investigate the hydrologic response to land-cover change of 10% of watershed area—an area larger than that typically selected for watershed service investments—in high-resolution hydrologic models of 29 watersheds across six continents and diverse climate zones. The watershed site models are used to explore streamflow changes after forest restoration, as well as agricultural development or urban cover expansion. The responses are generally small, with the average water yield and low flow reduced 1-3% under forest restoration and increased by a similar magnitude for agricultural or urban development; peak flow responds slightly more strongly.
We perform a cluster analysis to investigate the differences in the hydrologic response across sites under forest restoration. Three clusters of watersheds show consistent streamflow responses to forest restoration despite the limited response magnitude, but only one cluster is consistent with the desired flow-regulating behavior of decreasing peak flow and increasing dry-season flow. The clusters are identified using a graph-connectedness approach developed for this project. Non-parametric hypothesis testing finds that soil depth and hydraulic conductivity are the only statistically distinct watershed characteristics across these clusters, and so likely have significant effects on the streamflow response to forest restoration. We present ongoing work to assess the importance of soils to address unexplained variance in historical forest change experiments.
About the Speaker: James Dennedy-Frank (Postdoctoral Scholar, Natural Capital Project, Stanford)
P. James Dennedy-Frank is a postdoctoral scholar at the Natural Capital Project at Stanford. James works at the intersection of ecohydrology and water resources, where he seeks to incorporate hydrologic modeling, remote sensing, and statistics to guide sustainable land and water management. James received his Ph.D. and M.S. from Stanford, an M.S. from MIT, and a B.S. from Caltech.
Hosted by: Erica Woodburn