Shrubs, streamflow, and the paradox of scale

In this paper, we examine the linkage between woody plants and the water budget for three important woody plant communities in Texas, USA: saltcedar (Tamarix chinensis, Tamarix ramosissima), Ashe juniper (Juniperus ashei Buchholz), and mesquite (Prosopis glandulosa Torr. var. glandulosa). In most cases, these species are found in distinct physiographic and soil settings. Saltcedar is restricted to stream channels and floodplains; Ashe juniper is found mostly on karst limestone outcrops with shallow soils; and mesquite is found on deep soils. Because of these differences, changes in woody plant cover in each community will have a different effect on the water budget. For each type, we review the available literature and explicitly report the scale of observation (tree, stand, catchment, or landscape). A simple framework called the shrub-streamflow framework, which recognizes differences in response due to differences in physiographic setting, climate, and potential for deep drainage or subsurface flow, enables us to generalize the results. The fundamental premise of the framework is simple: for shrublands to be hydrologically sensitive to changes in woody plant cover, soil water or groundwater must be accessible to deep-rooting plants but too deep for shallow-rooting ones. Such a situation exists if groundwater is close to the surface (within 3-5 m and/or if deep drainage occurs (because of either high precipitation input or bypass flow in the soil). We argue that on an area basis, conversion of saltcedar stands to herbaceous plants in riparian regions has a much greater potential for increasing water yield than does conversion of woodlands to grasslands in upland regions where deep drainage does not occur. On upland sites where deep drainage does occur, conversion from woody to herbaceous vegetation may result in a savings of 40-80 mm year(-1) of water. But such savings have been observed only up to the small-catchment scale, and until further work is done it is uncertain whether they can be achieved at larger scales. Copyright (C) 2006 John Wiley & Sons, Ltd.