Managing Arid and Semi-Arid Watersheds

Three-Bar Wildlife Area

Three-Bar Wildlife Area

The research program on the Three-Bar watersheds represented the first major experimental watershed program in Arizona chaparral shrublands (DeBano et al. 1999a). Four watersheds (A, B, C, and D) were established and instrumented in 1956 on the Three-Bar Wildlife Area west of Lake Roosevelt. This area supported dense chaparral stands and had not been grazed since 1947. All of the watersheds were burned by a wildfire in June of 1959. more...

Treatments and Research

  • Treatments and Practices - how the study was conducted and treatments used, and management practices used and results for each, with photos and maps of the areas within the reserve.
  • Data Collected - this is a new section we're working on that will soon provide access to data on precipitation, streamflow, sediment loads, and other data that were collected.
    • Documentation for the data - what measurements were recorded, how to interpret it, the years covered, etc.
    • Data - select the watersheds, time period, etc. then view and download the results
  • Results, Implications, and Current Status
  • Related Images

Description and History

The research program on the Three-Bar watersheds represented the first major experimental watershed program in Arizona chaparral shrublands (DeBano et al. 1999a). Four watersheds (A, B, C, and D) were established and instrumented in 1956 on the Three-Bar Wildlife Area west of Lake Roosevelt. This area supported dense chaparral stands and had not been grazed since 1947. All of the watersheds were burned by a wildfire in June of 1959. After the burn, watershed A was abandoned, and watershed F, was instrumented in June 1963 to replace it. All four watersheds (B, C, D, and F) are north-facing, at elevations of 3,350 to 4,250 ft, on soils derived from granite, with the upper slopes exceeding 70%.

After a wildfire
After a wildfire

Previous studies on chaparral had been conducted at the Natural Drainages on the Sierra Ancha Experimental Forest (see above). However, the Three-Bar location, provided a better opportunity for evaluating maximum water yields that might be expected from shrub-to-grass conversions, because of its higher yearly precipitation and dense chaparral cover (greater than 60% crown cover). The dense stands were highly productive areas. Experimental watersheds were subsequently established in medium density chaparral (40 to 60%) on the Whitespar watersheds, and low density cover (less than 40%) on the Mingus watersheds. These watersheds are in north-central Arizona (Yavapai County), in the Central Arizona Highlands (DeBano et al. 1999b). The range of densities at Three-Bar, Whitespar, and Mingus were representative of most of Arizona's chaparral shrublands, and allowed researchers and managers to better identify chaparral shrublands which could be economically treated to obtain increased streamflow.

Cooperators

Interest in chaparral management evolved into a research and management program involving several agencies and organizations. The Tonto National Forest was responsible for managing much of the chaparral areas in the Central Arizona Highland. USDA Forest Service research was conducted by the Rocky Mountain Forest and Range Experiment Station and was assigned to the Forestry Science Laboratory at Tempe. Most of the Forest Service research addressed hydrologic and vegetative evaluations. Scientists with the Arizona Game and Fish Department provided wildlife evaluations on the Three-Bar Wildlife Area. Cooperative studies were also carried out with the University of Arizona, Arizona State University, and Colorado State University. Personnel from the Salt River Project and Arizona Water Resources Committee provided support and guidance in many of the watershed evaluations. Streamflow was gaged on some watersheds by the U.S. Geological Survey of the Department of Interior.

The information in this section is excerpted from History of Watershed Research in the Central Arizona Highlands


Treatments and Practices - Three Bar Watersheds

The Three-Bar experimental watersheds were established to determine the effects of chaparral shrub-to-grass conversions on increasing water yields, on dissolved chemical constituents and sediment, and of fire and herbicide applications in controlling shrub re-growth.


Conversion

While these research objectives provided a framework for treatment of Watersheds B, C, and F, other research agendas evolved as the understanding of chaparral response to treatment increased. Foliar sprays, initially used to control chaparral shrubs, inadequately eliminated all of the shrubs and required repeated application (Hibbert et al. 1974). Because of this inadequacy, soil-applied herbicides in subsequent treatments of Watersheds C, B, and F were tested. Above-normal nitrate levels were discovered in stream water as a result of earlier, herbicide treatment on Watershed C. These high nitrate responses led to studies on water quality and nitrogen losses as a result of shrub treatments.

The treatment pattern changed from treating entire watershed areas (Watersheds C and F) to
selectively controlling shrub plants in a mosaic pattern to provide protection from erosion on steep slopes, better habitat diversity for wildlife, and maintenance of increased streamflow (Watershed B). The mosaic treatment pattern of chaparral control was ultimately tested on the Whitespar watersheds (DeBano et al. 1999b).


Watershed A

AREA: ???? ac ( ha)
SLOPE:
ASPECT: Southeast facing
ELEVATION: 5,??? to 5,??? ft (1,??? to 1,??? m)
VEGETATION: Chaparral—Dominant shrubs are shrub live oak, birchleaf mountainmahogany, sugar sumac, and Emory oak
PARENT MATERIAL: Course Granite
GAGE: 120°V-notch weir
PERIOD OF RECORD: 1956 (partial) through 19??

HISTORY: This is one of the Three Bar watersheds located west of Lake Roosevelt in the Three Bar Wildlife Area, which is maintained cattle-free for game management studies. Stream and rain gages were installed in 1956. Streams were intermittent, flowing about one-third of the time during the first 3 years (1956-1059), and yielding less than 1 inch per year average flow (Hibbert et al 1974). The Boulder wildfire swept over the area in June 1959, topkilling all shrubs (Glendening et al. 1961).

OBJECTIVE: To determine how converting chaparral vegetation to grass affects streamflow, erosion and sedimentation, vegetation, and wildlife.

TREATMENT: Watershed A was shut down in ???? and no further records were collected.

SELECTED REFERENCES

Glendening, G. E., C. P. Pase, and P. Ingebo. 1961. Preliminary hydrologic effects of wildfire in chaparral. Arizona Watershed Symposium, Proceedings 5:12-15.

Hibbert, A.R.; Davis, E.A.; Scholl, D.G. 1974. Chaparral conversion. Part I: Water yield response and effects on other resources. USDA Forest Service Research Paper RM-17, 36 p. Rocky Mountain Forest and Range Experiment Station, Fort Collins, CO.

Watershed B

AREA: 46.5 ac (18.8 ha)
SLOPE:
ASPECT: North facing
ELEVATION: 3,300 to 4,200 ft (1,005 to 1,280 m)
VEGETATION: Chaparral—Dominant shrubs are shrub live oak, birchleaf mountainmahogany, sugar sumac, and Emory oak (crown cover averaged 60 to 75%)..
PARENT MATERIAL: Course granite
GAGE: 90° V-notch
PERIOD OF RECORD: 1956 (partial) through 1983

HISTORY: This is one of the Three Bar watersheds located west of Lake Roosevelt in the Three Bar Wildlife Area, which is maintained cattle-free for game management studies. Stream and rain gages were installed in 1956. Streams were intermittent, flowing about one-third of the time during the first 3 years (1956-1959), and yielding less than 1 inch per year average flow (Hibbert et al 1974). The Boulder wildfire swept over the area in June 1959, topkilling all shrubs (Glendening et al. 1961).

OBJECTIVE: To determine how converting chaparral vegetation to grass affects streamflow, erosion and sedimentation, vegetation, and wildlife. Prior to treatment, streamflow form WS B was calibrated against streamflow from the designated control WS D.

TREATMENT: After the Boulder fire, watershed B was seeded with weeping and Lehmann lovegrasses (Eragrostis curvula and E. lehmanniana) and yellow sweetclover (Melilotus officinalis) in July 1959, but the catch was very poor. The watershed was seeded again in May 1960 with weeping, Lehmann, and Boer lovegrasses (E. chloromelas). The resulting fair but spotty catch gradually increased in density where shrubs were controlled.

Shrubs on northeast-facing slopes (40 % of watershed B) were hand treated with chemicals (pelleted fenuron and granular picloram) in 1965 after 6 years of recovery following the wildfire. Surviving shrubs were again treated in 1968. Browse plants of highest wildlife value were not treated; these species made up only about 5 % of the total shrub cover, and included hollyleaf buckthorn, desert ceanothus, and false-mesquite (Calliandra eriophylla).

The remaining 60 % of watershed B (untreated portion) was treated in January 1972 in the same manner as the 1965 treatment, except that the chemical applied was karbutylate formulated as 50 % active ingredient tablets spaced a foot or so apart on the ground surface around each shrub. After 2 growing seasons the treatment appeared successful.

RESPONSE: The chemical treatment in 1965 reduced shrub cover to about 8 %. In spaces between shrubs, damage was light to forbs and grasses, which responded to release from shrub competition. Grass and forb production averaged 690 lb/acre/year on the treated areas (through 1973), but only 300 on the untreated slopes (Hibbert et al 1974).

SELECTED REFERENCES:

Davis, E.A. and P.A. Ingebo. 1973. Picloram movement from a chaparral watershed. Water Resources Research 9:1304-1313.

Davis, Edwin A. and Charles P. Pase. 1969. Selective control of brush on chaparral watersheds with soil-applied fenuron and picloram. USDA Forest Service Research Note RM-140, 4 p.

Glendening, G.E., C.P. Pase, and P. Ingebo. 1961. Preliminary hydrologic effects of wildfire in chaparral. Arizona Watershed Symposium, Proceedings 5:12-15.

Hibbert, A.R. 1971. Increases in streamflow after converting chaparral to grass. Water Resources Research 7:71-80.

Hibbert, A.R.; Davis, E.A.; Scholl, D.G. 1974. Chaparral conversion. Part I: Water yield response and effects on other resources. USDA Forest Service Research Paper RM-17, 36 p. Rocky Mountain Forest and Range Experiment Station, Fort Collins, CO.

Ingebo, P.A. 1969. Effect of heavy late-fall precipitation on runoff from a chaparral watershed. USDA Forest Service Research Note RM-132, 2 p.

McCulloch, Clay Y. 1972. Deer foods and brush control in southern Arizona. Journal of Arizona Academy of Science 7(3):113-119.

Pase, C.P. 1967. Helicopter-applied herbicides control shrub live oak and birchleaf mountainmahogany. U. S. Forest Service Research Note RM-84, 4 p. Rocky Mountain Forest and Range Experiment Station, Fort Collins, CO.

Pase, C.P. 1971. Effect of a February burn on Lehmann lovegrass. Journal of Range Management 24:454-456.

Pase, C.P. and P.A Ingebo. 1965. Burned chaparral to grass: Early effects on water and sediment yields from two granite soil watersheds in Arizona. Arizona Watershed Symposium, Proceedings 9:8-11

Pase, C.P., P.A. Ingebo, E.A. Davis, and C Y. McCulloch. 1967. Improving water yield and game habitat by chemical control of chaparral. XIV International Union for Research Organizational Congress, Munich, September 1967, Proc. V. I., Sect. 01-02-11, p. 463-486.

Smith, Ronald H., T.J. McMichael, and Harley G. Shaw. 1969. Decline of a desert deer population. Arizona Game and Fish Department, Wildlife Digest, Abstract 3, 8 p.

Watershed C

AREA: 95.3 ac (38.6 ha)
SLOPE:
ASPECT: North facing
ELEVATION: 3,450 to 4,250 ft (1,050 to 1,295 m)
VEGETATION: Chaparral—Dominant shrubs are shrub live oak, birchleaf mountainmahogany, sugar sumac, and Emory oak (crown cover averaged 60 to 75%).
PARENT MATERIAL: Course granite
GAGE: 90° V-notch
PERIOD OF RECORD: 1956 (partial) through 1980

HISTORY: This is one of the Three Bar watersheds located west of Lake Roosevelt in the Three Bar Wildlife Area, which is maintained cattle-free for game management studies. Stream and rain gages were installed in 1956. Streams were intermittent, flowing about one-third of the time during the first 3 years (1956-1959), and yielding less than 1 inch per year average flow (Hibbert et al 1974). The Boulder wildfire swept over the area in June 1959, topkilling all shrubs (Glendening et al 1961).

OBJECTIVE: To determine how converting chaparral vegetation to grass affects streamflow, erosion and sedimentation, vegetation, and wildlife. Prior to treatment, streamflow form WS B was calibrated against streamflow from the designated control WS D.

TREATMENT: Although the Boulder fire was not intended to be part of the original study, treatment plans were realined to accommodate the burning part of the treatment on watershed C (Hibbert et al 1974). Therefore, treatment on C is properly described as wildfire followed by conversion to grass using chemicals to suppress the shrubs.

Four annual aerial spray treatments of 2,4,5-T (1960 through 1963) on watershed C suppressed shrub regrowth after the fire, but actually killed only 42 % of the shrub live oak and 72 % of the birchleaf mountainmahogany plants. Hand treatment of surviving shrubs in 1964 (pelleted fenuron) and again in 1968 (pelleted fenuron and granulated karbutylate) resulted in much better control (Hibbert et al 1974).

Watershed C was control burned in March 1971. A similar type of burn was made 2 years earlier on a nearby plot with good results (Pase 1971). Topkill of shrub live oak and birchleaf mountainmahogany, the 2 most numerous shrubs remaining on watershed C, was 71 and 68 %. Grass was stimulated to higher production than without burning. There was no evidence of surface runoff or erosion as a result of the control burn.

RESPONSE: By 1969 shrub crown cover on watershed C was reduced to less than 3 %. Grasses and forbs averaged 1,200 lb/arce/year from 1963 to 1968. Ground cover in 1974 was considered excellent over most of the watershed, and surface infiltration rates remained high. The channel area was particularly well vegetated by weeping lovegrass. Cover was pooredt on the upper, steep north exposures where lovegrasses did not do well. Forbs and half-shrubs provide fair ground cover on these sites, which account for about 10 % of the catchment.

SELECTED REFERENCES

Davis, E. A. and P. A. Ingebo. 1973. Picloram movement from a chaparral watershed. Water Resources Research 9:1304-1313.

Davis, Edwin A. and Charles P. Pase. 1969. Selective control of brush on chaparral watersheds with soil-applied fenuron and picloram. USDA Forest Service Research Note RM-140, 4 p.

Glendening, G. E., C. P. Pase, and P. Ingebo. 1961. Preliminary hydrologic effects of wildfire in chaparral. Arizona Watershed Symposium, Proceedings 5:12-15.

Hibbert, A. R. 1971. Increases in streamflow after converting chaparral to grass. Water Resources Research 7:71-80.

Hibbert, A.R.; Davis, E.A.; Scholl, D.G. 1974. Chaparral conversion. Part I: Water yield response and effects on other resources. USDA Forest Service Research Paper RM-17, 36 p. Rocky Mountain Forest and Range Experiment Station, Fort Collins, CO.

Ingebo, Paul A. 1969. Effect of heavy late-fall precipitation on runoff from a chaparral watershed. USDA Forest Service Research Note RM-132, 2 p.

McCulloch, Clay Y. 1972. Deer foods and brush control in southern Arizona. Journal of Arizona Academy of Science 7(3):113-119.

Pase, C. P. 1967. Helicopter-applied herbicides control shrub live oak and birchleaf mountainmahogany. U. S. Forest Service Research Note RM-84, 4 p. Rocky Mountain Forest and Range Experiment Station, Fort Collins, CO.

Pase, C. P. 1971. Effect of a February burn on Lehmann lovegrass. Journal of Range Management 24:454-456.

Pase, C. P. and P. A Ingebo. 1965. Burned chaparral to grass: Early effects on water and sediment yields from two granite soil watersheds in Arizona. Arizona Watershed Symposium, Proceedings 9:8-11

Pase, C. P., P. A. Ingebo, E. A. Davis, and C. Y. McCulloch. 1967. Improving water yield and game habitat by chemical control of chaparral. XIV International Union for Research Organizational Congress, Munich, September 1967, Proc. V. I., Sect. 01-02-11, p. 463-486.

Smith, Ronald H., T. J. McMichael, and Harley G. Shaw. 1969. Decline of a desert deer population. Arizona Game and Fish Department, Wildlife Digest, Abstract 3, 8 p.

Watershed D

AREA: 80.5 ac (32.6 ha)

SLOPE: Upper slopes are steep, often exceeding 60 %.

ASPECT: Northeast facing

ELEVATION: 3,700 to 5,250 ft (1,130 to 1,600 m)

VEGETATION: Chaparral—Dominant shrubs are shrub live oak, birchleaf mountainmahogany, sugar sumac, and Emory oak (crown cover averaged 60 to 75%).

PARENT MATERIAL: Course granite

GAGE: 90° V-notch

PERIOD OF RECORD: 1956 (partial) through 1983

HISTORY: This is one of the Three Bar watersheds located west of Lake Roosevelt in the Three Bar Wildlife Area, which is maintained cattle-free for game management studies. Stream and rain gages were installed in 1956. Streams were intermittent, flowing about one-third of the time during the first 3 years (1956-1959), and yielding less than 1 inch per year average flow (Hibbert et al 1974). The Boulder wildfire swept over the area in June 1959, topkilling all shrubs (Glendening et al 1961).

TREATMENT: None

OBJECTIVE: Watershed D was used as a hydrologic reference watershed. A paired watershed methodology was used to evaluate treatment response. Two watersheds with similar characteristics (e.g., size, vegetation, precipitation, and soil type) were selected and before any watershed manipulation was done, runoff from each watershed was measured for several years to determine streamflow variations under pretreatment conditions. The number of years required depends on year to year variability normally experienced. In the semi-arid southwestern United States, it usually take about 7 years of pretreatment calibration to define a pretreatment relationship. During this time, the quantity and quality of other natural resources were also inventoried (e.g., soil loss, forage production, animal types and populations).

With pretreatment measurements completed, one of the watersheds was designated to be the untreated or "control" watershed. It was shown to respond to environmental influences in a particular manner to the watershed where the experimental manipulation or treatment was to be applied.

Measurements continued on both the experimental and control watersheds for several years after a treatment was applied. Streamflow, sediment production, and water quality were monitored regularly, and other resources were reinventoried periodically. Changes caused by the management practice applied to the experimental unit were evaluated by comparing posttreatment values to the pretreatment data relationships.

SELECTED REFERENCES

Glendening, G. E., C. P. Pase, and P. Ingebo. 1961. Preliminary hydrologic effects of wildfire in chaparral. Arizona Watershed Symposium, Proceedings 5:12-15.

Hibbert, A.R.; Davis, E.A.; Scholl, D.G. 1974. Chaparral conversion. Part I: Water yield response and effects on other resources. USDA Forest Service Research Paper RM-17, 36 p. Rocky Mountain Forest and Range Experiment Station, Fort Collins, CO.

Watershed F

AREA: 68.4 ac (26.2 ha)
SLOPE: Upper slopes are steep, often exceeding 60 %
ASPECT: Northeast facing
ELEVATION: 3,900 to 5,100 ft (1,190 to 1,555 m)
VEGETATION: Chaparral—Dominant shrubs are shrub live oak, birchleaf mountainmahogany, sugar sumac, and Emory oak (crown cover averaged 60 to 75%).
PARENT MATERIAL: Course granite
GAGE: 90° V-notch
PERIOD OF RECORD: 1963 through 1983

HISTORY: This is one of the Three Bar watersheds located west of Lake Roosevelt in the Three Bar Wildlife Area, which is maintained cattle-free for game management studies. Stream and rain gages were installed in 1956. Streams were intermittent, flowing about one-third of the time during the first 3 years (1956-1959), and yielding less than 1 inch per year average flow (Hibbert et al 1974). The Boulder wildfire swept over the area in June 1959, topkilling all shrubs (Glendening et al 1961).

OBJECTIVE: To determine how converting chaparral vegetation to grass affects streamflow, erosion and sedimentation, vegetation, and wildlife. Prior to treatment, streamflow from WS F was calibrated against streamflow from the designated control WS D.

TREATMENT: After the Boulder fire, watershed F (prior to being instrumented????) was seeded with weeping and Lehmann lovegrasses (Eragrostis curvula and E. lehmanniana) and yellow sweetclover (Melilotus officinalis) in July 1959, but the catch was very poor. The watershed was seeded again in May 1960 with weeping, Lehmann, and Boer lovegrasses (E. chloromelas). The resulting fair but spotty catch gradually increased in density where shrubs were controlled.

Watershed F was chemically treated in February 1969 with a broadcast application by helicopter of granular karbutylate to the soil (Hibbert et al 1974). Shrub crown cover was reduced from 55 to 4 % in the first year. The shrubs began to show visual signs of the chemical application by late April. Injury symptoms progressed rapidly during the spring months, and by early June most of the shrubs were either dead or dying. Total shrub kill increased to more than 95 % after 2 years. All grasses and herbaceous plants were killed by the chemical application.

RESPONSE: Phytotoxic residues remained active in the top 6 inches of soil over most of the watershed for the first 2 years, preventing growth of new plants (Hibbert et al 1974). The only exception was along the lower channel area where flow surfaced and became perennial after the treatment. A variety of forbs and grasses dominated by horseweed (Erigeron canadensis) invaded the channel area, where the plants rooted in the bed of the stream or in the moist channel banks. This narrow band of plants ended abruptly where water was not within a few inches of the soil surface.

During the third growing season (1971) forbs and grasses began to appear on interior ridges in the lower part of the watershed, but many of these showed injury symptoms before fall (Hibbert et al 1974). Invasion by forbs and grasses during the fourth growing season finally produced fair cover over all but the upper 10 to 15 % of the watershed.

Additional information on treatments at Three Bar can be found in: Davis and Ingebo 1973, Davis and Pase 1969, Glendening et al. 1961, Hibbert 1971, Hibbert et al. 1974, Ingebo 1969, McCulloch 1972, Pase 1967, Pase and Ingebo 1965, Pase et al. 1967, and Smith et al. 1969.

SELECTED REFERENCES

Davis, E. A. and P. A. Ingebo. 1973. Picloram movement from a chaparral watershed. Water Resources Research 9:1304-1313.

Davis, Edwin A. and Charles P. Pase. 1969. Selective control of brush on chaparral watersheds with soil-applied fenuron and picloram. USDA Forest Service Research Note RM-140, 4 p.

Glendening, G. E., C. P. Pase, and P. Ingebo. 1961. Preliminary hydrologic effects of wildfire in chaparral. Arizona Watershed Symposium, Proceedings 5:12-15.

Hibbert, A. R. 1971. Increases in streamflow after converting chaparral to grass. Water Resources Research 7:71-80.

Hibbert, A.R.; Davis, E.A.; Scholl, D.G. 1974. Chaparral conversion. Part I: Water yield response and effects on other resources. USDA Forest Service Research Paper RM-17, 36 p. Rocky Mountain Forest and Range Experiment Station, Fort Collins, CO.

Ingebo, Paul A. 1969. Effect of heavy late-fall precipitation on runoff from a chaparral watershed. USDA Forest Service Research Note RM-132, 2 p.

McCulloch, Clay Y. 1972. Deer foods and brush control in southern Arizona. Journal of Arizona Academy of Science 7(3):113-119.

Pase, C. P. 1967. Helicopter-applied herbicides control shrub live oak and birchleaf mountainmahogany. U. S. Forest Service Research Note RM-84, 4 p. Rocky Mountain Forest and Range Experiment Station, Fort Collins, CO.

Pase, C. P. 1971. Effect of a February burn on Lehmann lovegrass. Journal of Range Management 24:454-456.

Pase, C. P. and P. A Ingebo. 1965. Burned chaparral to grass: Early effects on water and sediment yields from two granite soil watersheds in Arizona. Arizona Watershed Symposium, Proceedings 9:8-11

Pase, C. P., P. A. Ingebo, E. A. Davis, and C. Y. McCulloch. 1967. Improving water yield and game habitat by chemical control of chaparral. XIV International Union for Research Organizational Congress, Munich, September 1967, Proc. V. I., Sect. 01-02-11, p. 463-486.

Smith, Ronald H., T. J. McMichael, and Harley G. Shaw. 1969. Decline of a desert deer population. Arizona Game and Fish Department, Wildlife Digest, Abstract 3, 8 p.


Results, Implications, and Current Status

After a wildfire in 1959, Watershed D recovered naturally to be used as a control and Watershed B, Watershed C, and Watershed F received chemical treatments. Results of the treatments were:

  • The effect of the wildfire on streamflow was generally short-lived. A sharp increase in overland flow occurred the first few years after the wildfire, especially during the first summer rains.
  • Stream flows increased on all the Three-Bar watersheds. However, after the third year, the crown cover on the control watershed (Watershed D) was dense enough to prevent any increased water production.

After a wildfire photo 1  After a wildfire photo 2
Wildfire

Following the treatment on Watershed C:

  • Stream flow increased four-fold (5.8 area-inches of water) over a 11 year period when compared with flow from the control watershed D.
      
  • Foliar herbicides killed about 40% of the shrub live oak and 70% of the birchleaf mountainmahogany. Because of the poor rate of shrub control, the remaining shrubs were treated individually with soil-applied herbicides in 1965 and 1968, which reduced the shrub crown cover to less than 3% by 1969.
  • Annual forage production averaged 1,200 lb/ac, and provided a ground cover which maintained high infiltration rates.
  • The increase in streamflow, and particularly yearlong streamflow, allowed riparian stringers to become established below the gaging station (DeBano et al. 1984).
      
  • Bird populations flourished in the newly created riparian stringers, but were reduced in the areas converted to grass (Szaro 1981).

Results from Watershed B were:

  • Nitrates in streamflow rose to relatively high concentrations (about 85 ppm) and was exported from the watershed in amounts up to 125 lb/ac/yr, in comparison to a control watershed value of about 1 lb/ac/yr. High concentrations of nitrates in the streamflow persisted longer from the two?stage treatment on Watershed B than from the one-stage treatment on Watershed F. High nitrate concentrations (44-373 ppm) were found in soil solutions from 5, 10, and 15 ft depths on the converted watershed as compared to low nitrate concentrations (0.2-6.2 ppm) on an adjacent undisturbed area (Davis 1987a, Davis and DeBano 1986).
  • Herbicide (picloram) concentrations in streamflow were higher (360 to 370 ppb) during the initial three months following treatment than thereafter. After 14 months and 40 inches of accumulated rainfall, picloram could not be detected in the streamflow (Davis 1973).
  • Surviving chaparral shrubs were re-treated in 1968 and again in 1978 (Davis 1987a). These additional two treatments reduced the shrub cover to about 8%.
  • Annual grass and forb production averaged 690 lb/ac on the treated areas as compared to 300 lb/ac on nearby untreated slopes.

  

Results from Watershed F were:

  • Nitrate concentrations in streamflow from the control (watershed D) remained less than 1 ppm, while nitrate from the treated watershed increased to a maximum concentration of 56 ppm during the first posttreatment year, with an annual concentration of 16 ppm (Davis 1984, 1987b, 1989).
  • Shrub crown cover was reduced from 55% to less than 5% the first year after treatment. Shrub kill increased to more than 95% after two years.
  • Runoff efficiency (the ratio of streamflow-to-precipitation) was increased to 2.3 times the efficiency of the control watershed (Watershed D), which amounted to an increase of 1.5 area-inches of streamflow.

Other Studies

Results of other studies near the Three-Bar experimental watersheds, and conducted elsewhere in the chaparral shrublands in the Salt-Verde River Basin are summarized below. In addition to testing of the effectiveness of herbicides for shrub control, other studies using prescribed fire and biological control (goat browsing) were tested.

When using fire, more than one-half of the chaparral canopy should be eliminated and prevented from becoming re-established, to obtain relatively high levels of seeded grass production (Pase 1971 and Pond 1961). However, burning can also result in an increase in undesirable plant species (Pase 1965).

Goat browsing reduces total cover in chaparral stands, particularly when done in conjunction with initial brush-crushing (Severson and DeBano 1991). Goat browsing to control chaparral shrubs can generally be expected to result in the consumption of the same plant species that would be preferred by cattle, deer, and elk (Knipe 1983). Successful use of goats to control shrub requires an intensive level of animal management.

The root system of a shrub live oak was excavated to characterize its mass (Davis and Pase 1977). It was determined that:

Shrub live oak root system
Shrub live oak root system

  • The root system included a taproot, many deep-penetrating roots, and profuse lateral roots.
  • The shrub live oak root system is able to effectively deplete both ephemeral surface and deeply stored soil moisture.

A study was conducted on the west side of Lake Roosevelt and north of the Three-Bar Wildlife area, to measure the temporal and spatial sediment delivery to and within a stream network following a wildfire in chaparral shrublands. This study indicated that:

  • Severe erosion following a wildfire deposited large amounts of hillside soil and debris in the channel system (Heede 1988).
  • As vegetation recovery after fire, sediment delivery from the watershed practically ceased.
  • Relatively clear water, upon entering the channel, caused degradation of the sediment deposited in the tributaries, and delivered this sediment into the main channel for many years after active hillslope erosion on the watershed had ceased.
  • The delayed sediment delivery over time made it difficult to interpret the effect of current management activities on erosion responses.

Mule (Odocoileus hemionus) and white-tailed deer (O. viginianus), and black bear (Ursus americanus) were studied on the Three-Bar Wildlife Area by the Arizona Game and Fish Department. These studies indicated that:

  • Mule and white-tailed deer select a variety of plants for food, including forbs, dwarf and half-shrubs, mast and other fruits, and evergreen browse of both chaparral and desert shrub (McCulloch 1973, Urness 1973, Urness and McCulloch 1973).
  • While conversion treatments increased forage production for cattle, they had adverse effects on deer, particularly when conversions of large areas, or entire watersheds, were implemented (McCulloch 1972).
  • Cover and food for black bear are enhanced in habitats composed of shrubs and low trees interspersed with a few forest species in the major drainages. This arrangement provides numerous mast- and fruit-producing species (LeCount 1980).
  • Leaving areas of adequate size as escape cover, and providing a number of seral stages of postburn vegetation should benefit both game and nongame wildlife species (Pase and Granfelt 1977). Less than one-half of the area should be converted (Reynolds 1972).

An inventory was made of 139 chaparral sites totaling almost 335,000 ac was accomplished in the early 1970s (Brown, T.C. et al. 1974). The costs of converting portions of chaparral shrubland areas that met crown cover, slope, and managerial criteria for conversion to grass, and maintaining these conversions over 50 yr, was compared with the benefits to society in terms of increased water yield and forage for livestock, and reduced fire-fighting costs. It was shown that:

  • Using fire as the main conversion tool, 96 of the inventoried sites (69%) had a benefit-cost ratio greater than 1. Using a soil-applied herbicide, 72 sites (52%) met that economic criteria.
  • Proper management would favorably affect soil movement, wildlife habitat, and esthetics. Recreation use would be unaffected in most treated areas.

Implications

Information has been obtained on how chaparral shrubland ecosystems function for land management decision making. Past research has contributed information on shrub control techniques, watershed and soil responses to shrub control, water quality, wildlife habitat changes, and economics. The most important management implication of this research is the ability to determine how to control chaparral shrubs to enhance the production of water and forage and to maintain wildlife habitat diversity.

If chaparral shrub suppression is desired, burning must be combined with other control methods such as applications of soil-applied herbicides or mechanical control methods. A problem associated with mechanical equipment is that it is limited to slopes with less than a 10% grade on rock free soils. One advantage of using prescribed fire is that the environmental changes created are similar to those occurring during the natural evolution of fire-adapted ecosystems. A disadvantage of using prescribed fire is that shrub control is temporary. Therefore, a management objective is often the suppression of shrubs rather than their eradication.

Current Status

Hydrologic evaluations on the Three-Bar watersheds were discontinued in 1983. The Three-Bar Wildlife Area remains ungrazed and provides a study site for current wildlife studies and for monitoring by the Arizona Game and Fish Department.


Related Images

Related Images can be found in the Flagstaff Lab Image Database.