Influence of grazing and vegetation on water yields and erosion

Back to contents - Previous file - Next file

Erosion is a natural process and few lands are without it. Natural erosion on most lands is not faster than soil formation. Accelerated erosion is faster than soil formation which, if continued, can result in considerable soil loss. Erosion is the movement of soil as a result of the forces of water and wind. These forces dislodge and transport the soil particles. Other factors determine the degree of erosion. Erosion is a result of weather (mainly rainfall and wind); soil (texture, structure, detachability, infiltration, transportability); topography (slope degree, length); and vegetation (cover, roots, height).

Surface runoff of water is a normal part of the hydrological cycle and few lands are without it, as well. It contributes to streamflow, although seepage of underground (springs, etc.) water is the most desirable source of streamflow. Excessive runoff contributes to erosion, deprives the ecosystem of needed water and reduces underground water supplies. Many of the factors related to erosion are also related to runoff. Thus, the two are interrelated.

Water erosion and runoff begin with the raindrop which incidentally is not tear-shaped as popularly believed. It is kidney-shaped. The impact of falling raindrops is quite large. Fifty millimeters of rain per hectare deliver enough force to lift 18 cm of soil to a height of one meter. On bare ground or heavily trampled rangelands, large raindrops are delivered in blows that dislodge soil particles. The raindrops themselves further compact the soil. Infiltration is reduced and the excess water runs off, carrying the soil particles with it. An objective of range management is to reduce trampling and to create a kind of vegetation that will intercept the raindrops, reducing their force. Water that trickles down the vegetation to the soil does not splash and dislodge soil particles. Infiltration is enhanced and runoff is reduced.

Hydrologic outputs from woodland, shrubland, and grassland ecosystems have been researched by a variety of methods over a period of years in the United States. There are difficulties in comparing the information, however, because studies have not tested the outputs with any degree of uniformity of methods. Nevertheless, ample evidence exists to point out the negative hydrologic impacts of heavy or abusive grazing in terms of reduced infiltration and increased compaction rates associated with that level of use. What is less well understood is the role that properly balanced grazing strategies would have on the long range hydrologic condition of the rangeland watersheds (Gaither and Buckhouse, 1981). It is not the author's intention to review the literature. Some selected and relevant publications are cited to illustrate certain points. Those wanting further information on rangeland watershed management in arid and semiarid regions are referred to the Society for Range Management's publication, "Rangeland Hydrology" (Benson, et al., 1972).

Some personal experiences are also worth mentioning. The beneficial effects of improved range was witnessed during a visit to the Khanasiri Range Station in Jordan. The station, located in a 250 mm per year rainfall zone, is characterized by a rough topography and shallow soils, a situation highly subject to runoff and erosion. The vegetation cover was in good to excellent condition owing to years of controlled grazing and proper management. The erodibility of the adjacent lands was considerably less because of an undulating topography and the dominance of the deep and well developed Mediterranean terra rosa soil type. Some of this area was cultivated and the remainder was overgrazed rangeland in poor condition.

A rainstorm occurred prior to and during the visit. Runoff from the ploughed lands was tremendous and the waterways were soon overflowing with red water which was ample testimony as to the amount of erosion that had taken place. There was also considerable runoff from the overgrazed rangelands, but erosion was less as was evidenced by the color of the water. Runoff from the Station's rangelands was practically nil and the small amounts of water that did run off was crystal clear.

A spring was reborn and a dry creek started flowing water on a ranch in Texas following brush control. The owner stated that he had never seen the creek flow water except during floods during forty years of ownership. However, an elderly man over 90 years of age stated he caught fish in the creek as a child, until it dried up. Increased spring and streamflow has been observed on various occasions following range improvement. On one occasion, a spring near extinction was saved by the elimination of phreatophytes.

Runoff-Erosion as related to Range Condition

Changes in vegetal type, modifications of cover and grazing intensities also result in corresponding changes in the hydrological regime. These changes may be beneficial or disastrous. The beneficial changes are more conducive to sustained maximum livestock grazing than disastrous changes. Generally, infiltration increases and runoff and erosion decrease with improvement in range condition. Leithead (1959) found that runoff increased in the Davis Mountain-Big Bend area of Texas as range condition deteriorated and moisture absorption by soil became slower. It was concluded that a range site in good condition could absorb moisture five to six times faster than the same range site in poor condition. Allred (1950) showed that rainfed infiltration rate was drastically reduced with reduction in vegetation cover and organic matter (Table 12). Infiltration rate on bare ground was only 0.5 inches per hour compared with a rate of 1.0 inches per hour on rangelands protected by 750 pounds of forage and organic material per acre and 9.4 inches per hour on rangelands with 5,800 pounds of vegetative material per acre. Organic matter in this case refers to ungrazed plant material on the soil surface in various stages of decomposition. This is also called "mulch" or litter.

Table 12. Relationship between forage and organic matter content and rainfall infiltration rate per hour. (From Allred, 1950)

Two high elevation watersheds in central Utah were studied to determine differences in surface runoff as related to soil and plant cover. Approximately 3 to 5 times as much surface runoff was produced on the watershed with poorer plant cover. Runoff during the three year period study was 10.33, 8.74 and 5.49% of the total annual rainfall on the area with the poorer plant cover as compared to 4.10, 2.88 and 1.05% for the continuously well-vegetated watershed (Stewart and Forsling, 1931). Aldon and García (1973) stated that average annual rate of sediment production declined 71% in the period 1967-71 compared with the period 1956-66 on a 471-acre watershed on the Rio Puerco drainage in New Mexico. This decline was a result of an increase in plant size and litter production.

Juniper (Juniperus spp. ) and sagebrush (Artemisia spp. ) are undesirable species in the Rocky Mountain range. Gaither (1981) found that sediment values ranged from 1,572 kg/ha in juniper ecosystems to 1,284 kg/ha in sagebrush ecosystems to 431 kg/ha in grassland ecosystems following investigation with a Rocky Mountain infiltrometer set to simulate a 28-minute convectional storm with an intensity of about 10 cm/hr. Rauzi (1960) conducted infiltration studies on rangelands in three locations in the Northern Plains. High condition range sites absorbed water almost three times as much as did low condition range sites. Results indicated that on two of the three areas the amounts of standing vegetation contributed more to infiltration than mulch material. Meeuwig (1970) investigated the influences of vegetation, soil properties and slope gradient on infiltration capacity and soil stability of high elevation herbland on the Wasatch Front in northern Utah under simulated rainfall conditions. Results emphasized the importance of vegetation and litter cover in maintaining infiltration capacity and soil stability. Infiltration is also affected significantly by soil properties, notably bulk density, aggregation, and moisture content.

Finally, Osborne (1956) classified various rangeland soils in the U.S. as to their inherent credibility based on certain soil characteristics, and then conducted studies to determine their stabilization requirements. He found that the least erodible soils required about 1,500 pounds of vegetal material per acre for stabilization and the most erodible required 5,000 pounds. Except for the highly erodible soils, these levels of productivity were equivalent to those for properly utilized ranges in low good or high fair condition. The results of the study also suggested that some land should not be utilized because of watershed value as illustrated in Table 11.

Runoff-Erosion as related to Grazing Intensity

It is generally considered that livestock compact the soil by trampling, although some disagree, claiming that animals break up surface crusts, enhancing water absorption. Research indicates that runoff and possibly erosion increase with grazing intensity. The influence of trampling cannot be ruled out because there is likely an interaction between trampling and defoliation which reduces the protective cover regardless of range condition.

Blackburn, et al. (1982) conducted a literature review attempting to tie the effects of grazing to hydrologic and watershed response. Rangeland areas were divided into several eco-geographic zones and studied separately. These zones were: sagebrush/grass, salt-desert shrub, southwest semidesert shrub/grass, California grasslands, Northern Great Plains, South Great Plain, pinyon-juniper woodland, ponderosa pine/bunchgrass, high elevation rangeland, eastern hardwood or pine forest. They pointed out that livestock grazing influences water hydrologic properties by removing protective plant cover and by trampling. They found that the literature is filled with examples of the adverse impacts of heavy or abusive grazing on watersheds.

They concluded, however, that runoff was greater with moderate stocking rates compared to no use, but that the differences in erosion were nil as illustrated in Figure 15. Excessive runoff and erosion occurs only with excessive grazing (Figure 15). These and other data indicate that animal production can be sustained without damage to the environment regarding erosion. Several studies have supported this.

Smeins (1975) concluded from his studies that moderate grazing may not increase erosion, but may significantly increase runoff as compared to lightly grazed or ungrazed areas. Moderate grazing has the potential to maintain a favorable forage, may not increase the hazard of erosion, and could possibly produce good quality runoff water for use outside the watershed. In Alberta, Canada, ungrazed areas and areas that had been grazed by cattle at four rates for 10 years were studied. Results indicated that soil erosion by water was not a critical factor in management even after 10 years of heavy grazing. Increasing amounts of standing vegetation and natural mulch were associated with increased water-intake rates (Johnston, 1962). Dunford (1949) measured runoff and erosion over a period of 12 years at the Manitou Experimental Forest near Colorado Springs. Two grazing intensities were studied for their effect on watershed values. Results indicated that moderate grazing would be allowable on relatively gentle slopes if the resulting loss of runoff water did not cause critical shortages of moisture for vegetative productivity.

Figure 15. Impact of grazing intensities on runoff and erosion (Adapted from Blackburn, et al., 1982).

Water Quality

Water full of sediment is not quality water. The silt fills waterways and reservoirs. It has been shown that quality water from this point of view can be produced with proper grazing and managerial practices. The picture is not so clear from a bacterial point of view. An understanding of relationships between livestock grazing and water contamination is needed in order to properly manage ranges near streams.

Two studies in arid and semiarid areas indicate that water contamination by livestock is minimal. A study in Utah showed that bacterial analyses (fecal and total coliform) of runoff water from infiltrometer plots indicated that the potential public health hazard of livestock grazing on semiarid open range on gentle slopes is probably minimal (Gifford, et al., 1976). In another study, no adverse effects from fecal contamination were detected after cattle grazing was introduced in a semiarid watershed near Coyote Flat in southeastern Utah. The area was seeded to crested wheatgrass in 1967 after pinyon-juniper chaining and windrowing of debris, and it was protected from grazing until 1974 when it was cattle-stocked at 2 ha/AUM. There were no significant changes in fecal and total coliform production, indicating that potential health hazards from fecal pollution during such grazing are minimal. Most dry rangelands such as those covering the southwestern United States have few, if any, permanent streams. Thus, there is little or no effective streambank area from which bacteria can be flushed into a water course. On most chainings, especially those with debris in place, runoff water cannot flow any distance overland and very little rainfall runs off (Buckhouse, 1976).

Contamination does appear to be a problem in the mountainous areas in the United States, but it is not all due to livestock. Studies on two adjacent pastures along Trout Creek in central Colorado indicated only minor effect of cattle grazing on water quality. Bacterial contamination of the water did, however, increase. Bacterial counts dropped to levels similar to those in the ungrazed pasture after removal of the cattle (Johnson, et al., 1978).

Sources and variations in bacterial indicators were reported by Stephenson et al. (1978) from stream sites over a three-year period on a 233-km² rangeland in southwest Idaho. The occurrence of fecal coliforms was directly related to the presence of cattle on summer range and winter pastures. Fecal coliform counts in adjacent streams were found to increase soon after cattle were turned in and remained high for several months after cattle were removed. Runoff from rainstorms increased both total and fecal coliform concentrations in streams on summer range with limited management and adjacent to winter pastures, but runoff from snowmelt had little effect. Total coliform counts varied more with change in streamflow than did fecal coliform counts. In fenced summer range allotments, under deferred grazing management, the effects were the same, except that bacterial counts were not as high or persistent. The decrease in bacterial concentrations at several downstream sampling sites indicated that certain stream segments were selfpurifying. The presence or absence of livestock along the streams overshaded any effect that variations in chemical concentration of the water might have had on bacterial concentrations.

White (1976) collected surface water samples from September, 1972 through August, 1973 on the northern slopes of Mount Taylor in west-central New Mexico. The effects of grazing, irrigation, water impoundment, vegetation clearing, and road construction upon the natural water quality was determined. Road construction had the greatest effect on water quality of these land-use practices, but all of the practices had a significant effect on at least one or more of the dissolved constituents. Shinner, et al. (1974) measured the water quality of a watershed for three summers. The standard plate count of 20 c and the aerobic bacteria enumerated on trypticase soy agar at 20 c numbered about 5 x 103 cells/ml in July, and remained at this concentration for the rest of the summer. The standard plate count at 35 c was generally constant at near 50 cells/ml each summer at all sample sites. The total coliforms in 9 of 17 cases exhibited a tenfold increase in numbers in late July or August over the values of 10 or less per 100 ml found in June. Fecal coliforms in 11 of 17 cases increased 10- to 100-fold in concentration in July or August over values obtained in June or September. Maximum of fecal streptococci were usually seen at all sample sites in July or August except for 1972, when sheep were being grazed in September. A stream in a natural area of the watershed appeared to be less polluted than other streams in the watershed which were influenced by human use.

Synopsis

Upset water regimes created by reduced infiltrations and excess surface runoff and soil erosion are desertification symptoms (Figure 5). These in turn promote other symptoms: microclimate changes, increasing microaridity, decreasing productivity and diminution of usefulness. Studies and experiences have shown that infiltration rates increase and that erosion and runoff decrease with increases in range condition. One study showed, except for highly erodible soils, that stabilization can be attained with low good or high fair condition. Studies have also shown that moderate grazing intensity (proper degree of utilization) can maintain a favourable forage resource without increasing the hazard of erosion. It could produce good quality runoff water for use outside the watershed. Both improved range condition and proper degree of utilization can be achieved with manipulations of livestock. Superior livestock production and desertification control will also be achieved.

The influence of livestock on bacterial contamination of water appears to be minimal in arid and semiarid areas. There is evidence that livestock grazing near mountain streams can contribute to some bacterial contamination, although recreationists contribute as well and probably more.