Sierra Ancha Experimental Forest

Treatments

Short-Term Studies

Short-term studies were used to investigate hydrological and ecological relationships of different plant communities found on the Sierra Ancha Experimental Forest. The results of these studies provided the basis for establishing long-term watershed studies to more fully evaluate water yield responses to brush control in chaparral (Natural Drainages), and to timber harvesting (Workman Creek) in mixed conifer forests. Studies were designed to test erosion control and revegetation techniques (Hendricks 1936, 1942, Hendricks and Grabe 1939). Cooperrider and Hendricks (1940) and Hendricks and Johnson (1944) studied the effects of grazing or wildfire on soil erosion. Sykes (1938) and Cooperrider et al. (1945) evaluated winter hydrographs for both Parker Creek and the Salt River.

Watershed treatment in mixed conifer vegetation

Low density chaparral vegetation

The effect of consumptive use of range vegetation on soil and water resources was determined by using small lysimeters. A lysimeter is an instrument for measuring the amount of water percolating through soils and for determining materials dissolved by the water. Evaporation from bare soil was compared to evapotranspiration from perennial grasses and shrubs (Rich 1951).

Fletcher and Rich (1955) developed a method of classifying southwestern watersheds on the basis of precipitation, potential evapotranspiration, and potential water yields. Most water yields result from water stored in the soil during the winter when evapotranspiration is low. High water yielding areas should be managed for water augmentation while intermediate and low-water yielding areas should be managed to reduce erosion.

Early ecological studies were concerned with characteristics of the Arizona white and Emory oak trees (Bliss 1937, Pase 1969, Pond 1971). Little (1938, 1939) conducted several botanical studies. The forest floor was recognized as playing an important role providing soil protection, increasing water holding capacity and availability to vegetation, and enhancing plant germination (Hendricks 1941, Pase 1972, Garcia and Pase 1967, Pase and Glendening 1965).

Watershed assessments suggested that streamflow was related to the interaction of precipitation and the different native vegetation types occupying the upland watersheds of the Salt River Basin (Cooperrider and Sykes 1938, Cooperrider and Hendricks 1940, and Cooperrider et al. 1945). This finding encouraged further research on techniques of vegetation manipulations and the hydrologic responses to these manipulations. Of particular interest was development of a method for controlling deep-rooted brush species and replacing them with shallower-rooted grass plants.

Early studies tested herbicides for shrub control (Cable 1957, Lillie et al. 1964), while other studies evaluated the combined use of herbicide and prescribed burning for controlling vegetation (Lindenmuth and Glendening 1962, Lillie et al. 1964, Pase and Glendening 1965, Pase and Lindenmuth 1971).

After a wildfire on Workman Creek

Short-term studies were also conducted at Workman Creek to supplement information gained from watershed-level studies. Investigations were conducted on the control of New Mexico locust and Gambel oak (Gottfried 1980, Davis and Gottfried 1983). The biomass of locust, both above-and below-ground, was also related to water yield (Gottfried and DeBano 1984). Another concern of managers was the potential impact of pocket gophers on ponderosa pine seedling survival (Gottfried and Patton 1984).

Long-term Studies

Important study areas dedicated to long-term research were located at the Base Rock lysimeters, Natural Drainages (watersheds), and the Workman Creek watersheds.

Base Rock Lysimeters

Hydrologic research in the 1930s used lysimetry to obtain quantitative information on the water balance for different vegetation types, because it offered a degree of experimental control that was not attainable using experimental runoff plots or small and large watersheds (Martin and Rich 1948). However, limitations associated with the use of lysimeters were recognized. Typical limitations include difficulty on making the lysimeters of sufficient size to reduce border effects and the disruption or destruction of the natural soil profile. The lysimeters constructed on the experimental forest mitigated these limitations by using large, undisturbed soil blocks overlying bedrock and, therefore, were named the Base Rock Lysimeters. Three lysimeters (18 ft wide and 50 ft long) with undisturbed soil profiles were established on an area supporting a deteriorated stand of perennial grasses, snakeweed (Gutlerrezia sarothrae) and yerba-santa (Eriodictyon angustifloium) (less than 2% total plant cover). Revegetation treatments including seeding, fertilizing and watering the lysimeters were carried out during 1934 and 1935. By 1942, the cover had increased to about 8% and consisted of mainly sprangletop (Leptochioa dubia), and sideoats (Bouteloua curtipendula) and hairy grama (Bouteloua hirsuta). Three grazing treatments applied between 1942 and 1948 were:

• Overgrazing with sheep at the rate of 2 mature ewes for 4 days annually, usually in October.
• More moderate grazing at half the above rate.
• Ungrazed.

Conclusions from the Base Rock Lysimeters show that (Martin and Rich 1948):

• The major portion of annual water yield occurred during winter as sub-surface flow from long-duration, low-intensity storms.
• Most surface run-off and soil erosion occurred during summer storm events from short-duration, high-intensity thunderstorms.
• Overgrazing caused increases in summer surface runoff and erosion and decreases in areal infiltration capacities, while the amount of winter water percolation was independent of the grazing treatment. Soil losses during summer storms increased from 60 to 307 tons/mi2 on heavily grazed plots compared to the ungrazed controls. In contrast, winter soil losses only increased from 15 to 68 tons/mi2 on moderately grazed plots.
• Results from late summer frontal-type storms were intermediate between those from summer and winter storm events.

Natural Drainages

Four chaparral-covered watersheds, called the Natural Drainages, were established on the Sierra Ancha Experimental Forest in 1934. These watersheds ranged in size from 9 to 19 ac. Precipitation and runoff were measured. The upper slopes of the watersheds are covered with diabase soils, which cover between 28 and 54% of the watersheds. Soils on the lower slopes were derived from quartzite. The original vegetation on these watersheds was sparse, with low density chaparral stands on southerly exposures.

Before treatment, crown cover of the chaparral shrubs was 20 to 25% compared with covers two-to-three times this density on the Whitespar and Three Bar watersheds (DeBano et al.1999a and DeBano et al.1999b). Shrub live oak was the most abundant shrub. Livestock grazing first started about 1880 and continued until 1934, when experimental watersheds were established on the Sierra Ancha Experimental Forest.

The Natural Drainages were one of the first watershed-level studies used in the Central Arizona Highlands to evaluate the effect of grazing on vegetative change and on production of streamflow and sediment. Two of the four watersheds were grazed by cattle and horses for short periods during the fall and spring beginning in 1939, and 2 watersheds were the controls (Rich and Reynolds 1963). Herbaceous vegetation was quantified using meter-square quadrants. Streamflow and sediment was measured using 90° V-notch weirs and sediment was measured in weir basins at the bottom of each of the 4 watersheds. These studies were terminated in 1952 when it was determined that the intensities of grazing used in the studies had no effect on total water yield or sediment trapped in the weir ponds (Rich and Reynolds 1963).

A growing interest in augmenting streamflow in the Central Arizona Highlands by manipulating native vegetation developed among water users in central Arizona during the 1950s. As a result, many of the research efforts during this period were initiated to evaluate the feasibility of increasing streamflow. A second experiment designed to determine the effects of chaparral cover manipulations on streamflow was started on the Natural Drainages in 1954 (Ingebo and Hibbert 1974). Chaparral cover was suppressed on 2 watersheds by treating the shrubs with herbicides, while the other 2 watersheds were maintained as control areas.

Overall, major amounts of sediment-free water come from areas with good grass cover and soil erosion is greatest where vegetation densities have been decreased, as by overgrazing.

Results from herbicide studies on the Natural Drainages watersheds indicated:

• There was 3-times more grass, forb, and half-shrub production on the treated areas having quartzite soils than on similar soils on the control areas (Pond 1964). No differences in plant production were observed on the diabase soils.
• An increase of 22% in streamflow occurred on treated areas (Ingebo and Hibbert 1974). Pretreatment average annual streamflow was 1.65 inches.
• The treated areas showed a 30% increase in quickflows, a 32% increase in delayed flows (the rising and falling stages of a streamflow hydrograph), and a 26% increase in peak flows (Alberhasky 1983).
• A decline of 72% in annual sedimentation was attributed to the increase in grass cover on the treated areas.

Streamflow increases from vegetation manipulations were attributed to lower evapotranspiration demands by the replacement grass cover. The streamflow increases from the Natural Drainages watersheds were low compared to other chaparral areas (Hibbert et al. 1974); this was related to the initial low density of shrubs and to the southeastern exposure of the area that results in relatively high energy inputs for evapotranspiration.

Workman Creek Watersheds

A major project was conducted on the Workman Creek watersheds to evaluate the hydrology of higher elevation mixed conifer forests, and to determine the changes in streamflow and sedimentation resulting from manipulating the forest vegetation.