The linkage between runoff and shrub cover on rangelands is a complex one. Depending on local conditions, increases in shrub cover may cause runoff to increase, stay the same, or decrease. For semiarid landscapes characterized by Horton overland flow, we can expect that shrub encroachment will result in a "net" decrease in landscape infiltration capacity of the soil and increased interconnectedness of intercanopy patches, which in turn will lead to increased runoff-at least at the hillslope scale. The exceptions are semiarid landscapes in which shrubs occur in bands parallel to the contour of the hillslope, allowing them to efficiently capture runoff. As scale increases, runoff will likely be captured at different points in the landscape. In other, wetter landscapes, the response of runoff to changes in shrub cover will differ depending upon how runoff is generated. For example, on mesquite rangelands where runoff occurs as Horton overland flow and where soils are deep and herbaceous growth is potentially high, changes in shrub cover have a minor effect on runoff. This is because a very high percentage of soil water will be evapotranspired regardless of what the vegetation cover is. Where subsurface flow occurs, increasing shrub cover can reduce runoff because of the dual effect of interception and transpiration loss by shrubs. These conditions exist for some chaparral rangelands in the southwestern United States and some juniper rangelands in Texas. The literature on the topic linking shrub cover and runoff provides few examples from rangelands where shrub control will likely result in increased streamflow. Two vegetation communities with the greatest potential for increasing runoff through shrub control are the chaparral woodland in the southwestern United States and portions of juniper woodlands in Central Texas. Both have high density evergreen shrubs that intercept and transpire large amounts of water and are in regions where runoff occurs as subsurface flow. These conclusions are on the basis of relatively small-scale studies. Success has been demonstrated on small watersheds (up to 225 acres) for chaparral rangelands. As of yet, increased streamflow has not been documented from juniper rangelands, but smaller scale studies have been encouraging. Runoff processes are scale-dependent. As the scale increases from the hillslope and small catchment to the watershed level, the linkage between shrub cover and runoff diminishes, but little work has been done to explicitly document watershed runoff dynamics at larger scales. I believe that the greatest potential for large-scale augmentation of streamflow from brush control lies in control of undesirable and invasive riparian and phreatophytic vegetation, especially salt cedar. In addition to increasing streamflow, control of salt cedar will result in substantial and economically significant improvement in flood and sediment control. 2004 by Texas A&M University Press. All Rights Reserved.
