Forests and Water: A State of the Art Review for Colorado

Home Browse Search

At Wagon Wheel Gap, for example, the average annual precipitation is only 21 inches, the mean pre - treatment water yield was 6.1 inches, and clearing one of the two catchments increased annual water yields by an average of only one inch. At Fool Creek in the Fraser Experimental Forest the average annual precipitation is approximately 26 inches, the average annual runoff prior to harvest was 8.7 inches, and the average increase in flow from cutting 50% of the vegetation or 40% of the watershed was 3.1 in, or approximately three times the value observed at Wagon Wheel Gap. The water yield increases obtained from other watershed experiments in the FEF and Coon Creek in south - central Wyoming are quite consistent, as they are in similar forest types and exhibit similar hydrologic behavior. There is greater variation in the increases in runoff measured from the paired- watershed studies at Beaver Creek, as there is more variation among these watersheds and they ex- hibit greater variability in the processes controlling the amount and timing of runoff (Baker, 1986). In summary, a certain amount of evaporation will occur independent of the vegetative cover. Until annual pre- cipitation exceeds this threshold, a change in vegetation generally will have no effect on annual water yields as long as the basic runoff processes are not changed (e.g., the infiltration rate and soil moisture storage capacity are not altered by compaction, paving, or soil erosion). As annual precipitation exceeds this threshold, the veg- etation plays an increasingly important role in the water balance equation by intercepting rain and snow and transpiring water. An increase or decrease in the density of the vegetation cover — as indexed by basal area or leaf area — will have a corresponding effect on runoff. Both equation 2 and paired- watershed studies in the Rocky Mountains show that the threshold for forest manage- ment to affect water yields is approximately 18 -19 inches, and this threshold corresponds with the threshold of 18 -20 inches suggested by Bosch and Hewlett (1982) (Figure 2.2). The only way to reliably increase water yields in areas that receive less than 18 -19 inches of annual precipita- tion is to reduce infiltration by paving or other land sur- face treatments and collecting the resulting runoff. 2.2.2. Timing of an increase in runoff. The timing of an increase in water yield depends on the timing of soil moisture recharge. In environments with dry summers and wet rainy winters, most of the water yield increase after forest clearing shows up in the fall and early winter because the reduction in summer tran- spiration causes the soils to be wetter at the beginning of the rainy season (e.g., Harr et al., 1975; Ziemer, 1981). Hence less rain is needed to recharge the soil and ground- water at the beginning of the wet season, and more of the initial rainfall in the winter wet season is converted into runoff. Later in the winter the proportional increases in runoff are much smaller, as once the soil is recharged the presence or absence of trees will primarily affect the amount of rain lost to interception. While annual inter- ception losses in coniferous forests are generally in the range of 20 -30% (Dunne and Leopold, 1978), the per- centage of rainfall lost to interception during large storm events is much less (Zinke, 1967). Hence an increase or decrease in the density of the forest vegetation has a pro- portionally smaller effect on runoff once the soils have been fully recharged. In snow - dominated areas the basic water balance equa- tion still applies, but there are some important differ- ences with respect to the timing of the observed water yield increases. In Colorado nearly all of the water yield increase occurs in early spring, and the difference in the timing of the water yield increase is due primarily to the difference in the timing of soil moisture recharge. As in rain - dominated areas, removal of the forest canopy results in less soil water depletion during the summer, so less water is needed for soil moisture recharge and more of the early snowmelt is converted into runoff. The removal of the forest canopy also increases the rate of spring snowmelt, so the increased snowpack typically does not extend the duration of spring snowmelt. The net result is that in snowmelt- dominated areas the increase in runoff due to forest harvest occurs primarily in spring, on the rising limb of the snowmelt hydrograph, rather than in the fall and early winter. This pattern is clearly shown by comparing the average annual hydrographs prior to and after timber harvest for the snowmelt -domi- nated Fool Creek watershed on the Fraser Experimental Forest (Figure 2.3). Month -by -month comparisons of the pre- and post- treatment flows for Fool Creek confirm that May is the only month with a consistent, statisti- cally- significant increase in runoff (Troendle and King, 1985). A significant increase in June runoff can occur in places where, or years when, peak snowmelt occurs slightly later in the year (Troendle et al., 2001). As dis- cussed in section 2.2.5, an increase in soil moisture after timber harvest has no detectable effect on low flows in Colorado. 2.2.3. Causes and variability of water yield increases after forest harvest. In most forested areas the primary effect of forest harvest or afforestation is to alter the amount of growing season