2. LGT Environmental Impacts

2.1 Range Utilization

2.1.1 Direct Indicators

2.1.1.1 Soil Structure

See Section III-2.1.1.1 and Section V-2.3.1.2.

2.1.1.2 Soil Erosion

Livestock grazing has profound influences on the amount and continuity of vegetative cover on rangelands. There are different thresholds of cover for different combinations of soils and topography related to soil loss from wind- and/or water-based erosion. Vegetative cover provides a physical barrier to the direct contact between free-falling precipitation and the soil surface, interrupting the release of the energy of raindrops on the soil aggregates. Direct “splash effect” of raindrops breaks soil particles away from aggregates and disperses them into moving surface water. The disaggregated particles may also form solid masses or crusts on the surface that harden and suppress seedling emergence as well as resist water infiltration, thus promoting runoff. Standing live and dead crop and a mulch component on the soil surface provide retardance to overland surface flow and increase the opportunity time for water to infiltrate into the soil profile. In the case of wind erosion, vegetative matter on the soil surface acts as a physical retardance against wind velocity and captures moving particles. Through the action of plant defoliation and removal of biomass, grazing can in large part determine the development and maintenance of sufficient plant material to hold erosion to tolerable levels. Documentation of the utilization of vegetation by grazing animals past the level of amounts adequate for soil protection are manifold and worldwide.

Lim (1993) showed that excessive grazing (primarily by cattle) with a high density of animals under limited species (2-3) contributed to soil degradation in the Mongolian steppes. As the content of soil humus and nitrogen decreased, the soil structure was destroyed, the top horizons became dried and condensed (increased bulk density), and the area of bared spots increased. Zhaohesitu (1993) reported that dry and windy climate and loose sandy surface soils are the potential factors causing desertification in the grassland regions of Inner Mongolia and that any improper land utilization, including overgrazing, can destroy the fragile ecosystem and promote the desertification process.

McCalla et al. (1984) studied the influence of short duration grazing (SDG), moderate continuous grazing (MCG), heavy continuous grazing (HCG), and grazing exclusion on sediment production of midgrass- and shortgrass-dominated communities over a 20-month period in the Texas Edwards Plateau land resource area. A combination of cattle, sheep, and goats was used in each grazing treatment. Sediment production was consistently less from the midgrass (bunchgrass) than from the shortgrass (sodgrass) community. The HCG pasture was severely overgrazed and resulted in excessive soil loss. The midgrasses in the pasture were destroyed after 26 months of overgrazing. Sediment production from the SDG pasture stocked at double the recommended rate increased during the study period. The SDG pasture, by the end of the study, had lost more sediment from both the midgrass- and shortgrass-dominated communities than the MCG pasture. Sediment loss from the midgrass community in the MCG pasture was consistently low during the study; however, sediment production from the shortgrass community decreased in the MCG pasture. Sediment production from the midgrass community in the nongrazed pasture remained consistently low throughout the study, but the shortgrass community showed a strong decrease in sediment loss during the study. The results of this study documented that sediment production was consistently less from the midgrass than from the shortgrass community. A decline in midgrass, regardless of the cause, results in greater soil loss. Short-duration grazing at double the recommended stocking rate resulted in greater soil loss from both the midgrass- and shortgrass-dominated communities than the moderately stocked continuously grazed pasture.

A definitive study of the hydrologic characteristics of vegetation types as affected by livestock grazing systems was done by Thurow et al. (1986) in the Edwards Plateau major land resource area of Texas. Infiltration rate and sediment production were assessed in oak, bunchgrass, and sodgrass vegetation types in moderate continuous (MCG), heavy continuous (HCG), and intensive rotation (short-duration, SDG) grazing systems and in a livestock exclosure (LEX). Infiltration rate was related to the total organic cover and bulk density characteristics of the site. The amount of cover was more important than type, indicating that protection of soil structure from direct raindrop impact was the primary function of cover on infiltration. The SDG and the HCG pastures had lower total organic cover with correspondingly lower infiltration rates compared to the MCG and LEX pastures. Sediment production was related to the total aboveground biomass and the bunchgrass cover of the site. Obstruction to overland sediment transport and protection from the disaggregating effect of direct raindrop impact were the primary functions of the total aboveground biomass and bunchgrass cover. Midgrass cover and total aboveground biomass in the MCG and LEX pastures was significantly greater than in the SDG and HCG pastures; thus, sediment production from the MCG and LEX pastures was significantly lower than from the SDG and HCG pastures.

Soil hydrologic characteristics influenced by livestock trampling under intensive rotation grazing were also studied by Warren et al. (1986a). Four rates of trampling were used to simulate stocking rates of moderate (8.1 ha/AU/yr), double (4.1 ha/AU/yr), triple (2.7 ha/AU/yr), and no livestock trampling. Infiltration rate decreased significantly and sediment production increased significantly on a site with a silty clay surface soil devoid of vegetation following periodic trampling typical of intensive rotation grazing systems (double and triple the recommended moderate stocking rate). The deleterious impact of livestock trampling generally increased as stocking rate increased. Damage was augmented when the soil was moist at the time of trampling. Thirty days of rest were insufficient to allow hydrologic recovery. Soil bulk density, aggregate stability, aggregate size distribution, and surface microrelief were related to the soil hydrologic response of the trampling treatments.

Warren et al. (1986b) in a similar study at the same location studied the soil hydrologic response to number of pastures and stocking density under intensive rotation grazing. Regardless of whether the pasture grazed at the highest stocking density was in similar or poorer hydrologic condition in terms of treatment response, the authors concluded that the data do not support the hypothesized advantages of increased stocking density via manipulation of pasture size and numbers. Rest is more important than livestock activity to soil hydrologic stability. Therefore, very little benefit in terms of soil hydrologic condition should be expected from large increases in the number of pastures within rotation grazing systems.

The same authors (Warren et al. 1986c) also studied the effects of season and stage of rotation cycle on hydrologic condition of rangeland under intensive rotation grazing. They measured infiltration and sediment production for two years prior to the movement of livestock onto the pasture, soon after their removal, and approximately midway through the subsequent rest period of each rotation through the system. As found in other experiments, midgrass-dominated interspaces were characterized by significantly higher infiltration rates and lower sediment production than shortgrass-dominated interspaces. Infiltration rate declined and sediment production increased following the short-term intense grazing periods inherent in the rotational system. The detrimental effect was significant during periods of drought or winter dormancy but not during periods of active growth. Soil characteristics relating to higher hydrologic condition were significantly more stable during the growing season, providing greater resistance to and resilience from the damaging impact of livestock activity. The authors suggested that where growing seasons are short and often unpredictable, as in most semiarid and arid regions, repeated intense trampling may lead to long-term degradation of the soil resource.

Gamougoun et al. (1984) evaluated vegetation, soils, infiltration rates, and sediment production as they relate to livestock exclusion, continuous heavy grazing, continuous moderate grazing, and rotation grazing on a homogeneous plant-soil complex. The exclusion of livestock resulted in infiltration rates significantly higher than when the pastures were grazed in any system. No differences were found between heavily and moderately stocked pastures. This lack of change was attributed to organic matter additions from forbs that replaced grasses when the area was heavily grazed. The rotation treatment had infiltration rates that were lower than the exclosures or continuous grazing treatments. Sediment production from interrill erosion was similar in all treatments except when the livestock were concentrated into a fourth of the rotations system's area, which resulted in higher sediment levels.

Soil bulk density is inversely related to infiltration rates. Lower infiltration means more runoff and, in many cases, greater potential for soil erosion. Several researchers have reported the relationship between soil bulk density and livestock grazing. Rauzi and Hanson (1966) reported the probability that moderately grazed pastures have more pore spaces than heavily grazed pastures to be about 70 percent. Pore space in a lightly grazed treatment was significantly larger than in moderately and heavily grazed pastures. Unrestricted grazing of stream bottoms in the Black Hills caused increases in bulk density in the surface to 5 cm on 3 out of 4 sites (Orr 1960). Unrestricted grazing of the Southern High Plains caused increased bulk density of the surface soil in a study done by Brown and Schuster (1969). Laycock and Conrad (1967), in a study in northern Utah, showed that seasonal changes in bulk density were greater than those that could be attributed to grazing effects. Rhoades et al. (1964) and Linnartz et al. (1966) reported that grazing increased soil bulk density to 40 cm or more. However, these findings are contrary to other studies that found grazing impacts were minimal or restricted to the surface 5 cm. Wood and Blackburn (1984) found that a high-intensity, low-frequency grazing system (HILF) encouraged grazing and trampling underneath shrubs as well as in the interspaces, thus promoting increased bulk densities in shrub zones where bulk densities are normally lower than shortgrass communities. The same study found that the percentage of water-stable soil aggregates in the HILF system were lower than in the grazed and rested deferred-rotation and moderately stocked, continuously grazed treatments. Soil aggregates from heavily stocked, continuously grazed treatment were significantly less stable than from moderately stocked, continuously grazed, rested or grazed deferred-rotation and rested HILF treatments but were not different from the grazed HILF.

Data on the impacts of livestock grazing on infiltration rates were summarized by Gifford and Hawkins (1978). These authors found that a) the results are often confounded by range improvement activities, past grazing, and/or climatic fluctuations; b) the results may be very site specific; and that c) differences between light and moderate grazing are usually very small. Additionally, d) heavy grazing almost always causes a reduction in infiltration rate. Again, the predominance of evidence suggests that heavy grazing, through increases in bulk density and reduction of infiltration rates, can be expected to increase runoff and the potential for soil erosion.

Several studies have documented the direct effect of grazing on sediment production and erosion. Lusby (1970) reported that the sediment produced from an ungrazed watershed at Badger Wash, Colorado, was 45 percent less than from a heavily grazed watershed, based on an uncalibrated paired watershed comparison. Studies conducted in a Louisiana forest (Duvall and Linnartz 1967) and a Canadian grassland (Johnston 1962) did not show an increase in sediment production from heavy grazing. In the Canadian study, heavy grazing was defined as 6 Ac/Au for a 6-month summer season on good condition range. A study on the Manitou Experimental Forest in Colorado found no difference in sediment production between moderate continuously grazed and nongrazed areas (Dunford 1954). Similar results were reported by Rich and Reynolds (1963) in a study conducted on the chaparral lands of Arizona. However, moderate grazing of a subalpine range in Utah increased sediment production over that of nongrazed areas (Meeuwig 1965). Meeuwig emphasized the need for proper management to reduce grazing impacts. Buckhouse and Gifford (1976) reported that grazing at 2 ha/AUM for two weeks on a chained pinyon-juniper site in southern Utah showed no increase in sediment production over that recorded from an ungrazed area.

Heavy continuously grazed pastures on the Edwards Plateau of Texas produced more sediment and less than one-half the infiltration rate of a pasture grazed under a 4-pasture deferred rotation system and an exclosure (McGinty et al. 1978). Wood et al. (1978) measured sediment loss from 3 vegetation areas in 8 different grazing/nongrazing treatments on the Edwards Plateau and Rolling Plains of Texas. Heavy continuously grazed midgrass interspace areas produced more sediment than a 4-pasture deferred-rotation grazing system and a 20-year-old livestock exclosure. Pastures grazed in a short duration system were intermediate.

Knight et al. (1984) found that infiltration rates of undisturbed oak mottes in the Texas Edwards Plateau were greater than those on adjacent grass-dominated areas. Infiltration rates and sediment production of oak mottes were most influenced by grass standing crop, mulch and organic matter, and water-stable aggregates. As the amount of vegetation increases, infiltration rates should increase and soil loss decrease. However, some plants have a greater influence on infiltration rates and sediment production than others. The soils under oak mottes are strongly aggregated and covered with mulch, organic layers, and bunchgrasses, which results in greater infiltration rates and lower soil loss than from adjacent grassland areas. Likewise, midgrasses (bunchgrasses) are more effective in modifying adjacent grassland areas than shortgrass (sodgrasses). Thus, greater infiltration rates and lower soil loss occurred from midgrass-dominated areas than shortgrass-dominated areas. Livestock grazing could have profound effects on the relative percentage of midgrasses to shortgrasses, and consequent infiltration rates and soil erosion.

Blackburn et al. (1982) summarized the literature on impact of grazing on watershed parameters, including sediment production or erosion. This study has great application, since it covers vegetation types of sagebrush/grass (native and seeded), salt-desert shrub, southwest semidesert shrub/grass, California grasslands, Northern Great Plains, Southern Great Plains, pinyon-juniper woodland, ponderosa pine/bunchgrass, higher elevation rangelands, and Eastern hardwood or pine forests. These vegetation types would be analogous to many areas of the world's rangelands. Blackburn concludes that grazing, whether by arthropods or ungulates, has impact on watershed parameters and that it has been recognized for seventy years (now over eighty years) that heavy grazing accelerates erosion and runoff. Blackburn cites Love (1958): “There is a large body of information leading to the conclusion that heavy grazing has had bad hydrologic consequences. It is doubtful that more investigations are needed to emphasize this conclusion.” Conversely, Blackburn reports that there was relatively little information on the hydrologic impacts of light or moderate grazing intensities. Most of the grazing studies up to 1982 compared the impacts of heavy grazing with no grazing, thus giving the impression that heavy grazing is a viable management objective or that livestock grazing is universally equivalent to heavy grazing; however, no such oversimplification is justified (Blackburn et al. 1982). The available data on light or moderate grazing intensity strongly suggest that hydrologic differences between pastures continuously grazed lightly or moderately are not significant. There appears to be no hydrologic advantage to grazing a watershed lightly rather than moderately. Some studies have failed to show a difference in soil loss, infiltration capacity, or soil bulk density among light, moderate, and ungrazed pastures.

In the same literature review, Blackburn et al. (1982) referred to studies from the Rolling Plains and Edwards Plateau of Texas for impacts of specialized grazing systems on watershed characteristics. The results of studies by McGinty et al. (1978), Wood and Blackburn (1981a, 1981b), and Blackburn et al. (1980) indicate that pastures grazed under a 4-pasture deferred-rotation system are hydrologically similar to those of livestock exclosures. Conversely, the HILF grazing systems were either similar hydrologically to heavy continuous grazing in the Rolling Plains or to moderate continuous grazing in the Edwards Plateau. The summary by Blackburn et al. (1982) concludes that the major pollutant from rangeland watersheds is sediment. Moderate continuous grazing or specialized grazing systems should reduce sediment losses to a minimum from most watersheds. However, it should be recognized that if watersheds have been severely overgrazed, instituting a moderate continuous or specialized grazing system may not reduce sediment losses. Bacteria as potential pollutants from livestock grazing do not appear to be a problem on areas not included in riparian zones. Ungrazed areas may contribute in excess of recommended levels of bacterial counts for water quality based on use by wildlife (Doran and Linn 1979).

The impact of brush control on watershed parameters has been of interest primarily for two reasons. Increased water can be made available on-site to forage plants for livestock by controlling less palatable shrubs, and/or off-site by replacing deep-rooted shrubs with shallow-rooted grasses or forbs that consume less water. Relatively little information, however, is available to evaluate the impact of brush control on runoff, erosion, or water quality (Blackburn 1983).

Influence of sagebrush control on infiltration rates, sediment production, runoff, or erosion has varied with location and treatment. Work in the sagebrush (Artemisia spp.) areas of the western U.S. has generally shown that little improvement in infiltration rate or sediment production can be expected from mechanical control of sagebrush. Controlling pinyon-juniper on gentle slopes had little effect on infiltration rates or sediment production. Research in Utah, however, reported larger sediment loss from chained, debris-windrowed sites than from chained, debris-left-in-place or undisturbed woodland. Controlling juniper on moderate or steep slopes in Texas had a significant influence on soil loss for the first two years after treatment (Blackburn 1983).

Burning chaparral watersheds reduces soil protective cover exposing the sites to erosion forces. The resultant impacts may be increased overland flow, increased peak flow, increased sediment loss by water, slumping, and dry gravel. The use of fire in the chaparral-type rangeland must be prescribed with caution and a sound fire plan. Controlling honey mesquite by several methods increased infiltration rates and either had no effect or decreased sediment production on the Rolling Plains of Texas. The application of prescribed burning following herbicides had little influence on infiltration rates of whitebrush or running mesquite-dominated sites on the South Texas plains, but decreased sediment loss when compared to herbicide treatment or no treatment. Loss of soil cover following herbicide treatment of brush is apparently compensated by the positive influence of prescribed burning on herbaceous cover (Knight et al. 1983). Improvement in hydrologic variables, including erosion, following brush control is not automatic and depends on site conditions before and after treatment, method used, weather, and management (Blackburn 1983).

Thurow et al. (1988a) sampled infiltration rate and interrill erosion bimonthly from 1978 to 1984 on pastures grazed continuously, moderately stocked; continuously and heavily stocked; high-intensity, low-frequency, and moderately stocked; and short duration and heavily stocked. The moderately stocked, continuously grazed and high-intensity, low-frequency pastures were able to recover from droughts and maintain initial infiltration rates and interrill erosion. In contrast, infiltration rates decreased and interrill erosion increased on heavily stocked continuously grazed and heavily stocked short-duration pastures. The trend of infiltration rate and interrill erosion deterioration in the heavily stocked, short-duration grazing and heavily stocked, continuously grazed pastures was not gradual; rather, it followed a stair-step pattern typified by decreasing condition during drought and an inability to recover to predrought level during periods of above-normal precipitation. During drought the grazing pressure on these pastures reduced grass basal cover (Thurow et al. 1988). The high crowns of bunchgrass made them susceptible to damage from heavy grazing pressure. The litter base of the heavily grazed pastures also declined during drought due to more complete consumption of forage. The continued heavy stocking rate after drought was sufficient to prevent a sustained recovery of total basal cover, litter, and midgrass cover to predrought levels. Thus, increases in erosion occurred in a stair-step fashion rather than as a gradual increase. The heavy stocking rate and climate rather than grazing strategy were the primary factors influencing the hydrologic responses.

Thurow et al. (1986) assessed infiltration rate and sediment production of oak motte, bunchgrass, and sodgrass vegetation types in four different treatment pastures in the Edwards Plateau of Texas. Treatments were moderate continuous grazing (MCG), heavy continuous (HCG), and intensive rotation short-duration, SDG grazing systems and a livestock exclosure (LEX). The SDG and HCG pastures had lower total organic cover with correspondingly lower infiltration rates compared to the MCG and LEX pastures. Sediment production was related to the total aboveground biomass and the bunchgrass cover of the site. Obstruction to overland sediment transport and protection from the disaggregating effect of direct raindrop impact were the primary functions of the total aboveground biomass and bunchgrass cover. Total aboveground biomass was greatest in the oak motte and least in the sodgrass interspace. Consequently, the sodgrass interspace had the greatest amount of sediment production and the oak mottes had the least sediment production. Midgrass cover and total aboveground biomass in the MCG and LEX pastures was significantly greater than in the SDG and HCG pastures; thus, sediment production from the MCG and LEX pastures was significantly lower than from the SDG and HCG pastures.

2.1.1.3 Water Quality

Riparian areas and streams are critical resource areas for livestock production characteristic of Livestock Grazing Temperate (LGT) and Livestock Grazing Humid (LGH) systems. Livestock are attracted to riparian areas because of succulent forage, accessibility, shade, a reliable water supply, and, at times, a favorable microenvironment compared to associated upland communities (Skovlin 1984). In semiarid rangelands, one hectare of meadow may have a potential grazing capacity equal to 10 to 15 hectares of surrounding forested rangeland. In some areas, riparian meadows may constitute only 1 to 2 percent of livestock summer range but may provide 20 percent of summer forage (Clary and Webster 1990). Eighty percent of forage consumed by livestock may be obtained from riparian meadows (Kauffman and Krueger 1984).

However, the presence of livestock in these critical riparian/stream zones is often viewed as a negative association. In LGT and LGH production systems, the availability of drinking water for livestock is often a major constraint to livestock production. Consequently, available water sources and associated stream riparian zones often are a focal point for livestock and other herbivore use. Animals converge on these areas seeking both water and shade, and, as a result, areas around water often become a major zone of degradation. When this occurs, maintaining the biological and ecological structure of the stream riparian zone can become an issue.

2.1.1.3.1 Stream Riparian Issues

Livestock grazing and trampling affects stream habitats by reducing or eliminating riparian vegetation, changing stream bank and channel morphology, and increasing sediment load in streams (Clary and Webster 1990). Streamside areas typically receive 20 to 30 percent greater use than adjacent upland areas (Platts, 1991). Although it is widely acknowledged that excessive livestock grazing and trampling can negatively affect condition of stream riparian zones and the stream itself, many factors actually determine the impact livestock will have on the riparian zone, including a) confinement of livestock to stream riparian areas, b) total availability and type of water sources, c) factors associated with terrain that predispose the grazing animal to increase the amount of time in areas near water sources, and d) season of the year which influences the animal's need for water or influences the animal's behavior. A sufficient number of disparate variables are involved in livestock grazing in general and for stream riparian zones specifically. The variables which each stream or animal water source needs must be evaluated in terms of problems associated with livestock grazing rather than in terms of generic assumptions that any and all livestock grazing is detrimental to stream and riparian zone ecosystems.

Because of their close association with watersheds, riparian areas must be viewed in the context of the entire watershed, not just as individual stream reaches. As noted in the Salmon Recovery Plan, designed to enable recovery of Chinook salmon, an anadomous fish listed as endangered in the Pacific Northwest what happens in the uplands associated with riparian zones affects what happens to the stream and the riparian zone. In the watershed, the primary effects of livestock grazing are alteration of plant cover and/or plant composition and soil compaction from the physical action of animal hooves.

2.1.1.3.2 Direct Effects of Livestock Grazing

Livestock have the potential of detrimentally affecting water quality by a) causing changes in the chemical, physical, and bacteriological characteristics of water; b) modifying habitat by changing stream channel and associated vegetation; and c) changing stream flow patterns.

According to Platts (1989), the potential effects of grazing on aquatic and riparian resources are numerous and are listed below.

Effect on Water Column:

a. withdrawal of stream flow to irrigate grazing lands;

b. drainage of wet meadows or lowering of the groundwater table to facilitate grazing access;

c. pollutants (e.g., sediments) in return water from grazed pasture lands;

d. change in magnitude and timing of organic and inorganic energy inputs to the stream (i.e., solar radiation, debris, nutrients);

e. an increase in fecal contamination;

f. change in water column channel morphology, such as an increase in stream width and a decrease in stream depth, including reduction of streamshore water depth;

g. change in timing and magnitude of stream flow events from change in watershed vegetative cover;

h. increase in stream temperature.

a. shearing or sloughing of stream bank soils by hoof or head action;

b. water, ice, and wind erosion of exposed stream bank and channel soils because of loss of vegetation cover;

c. elimination or loss of stream bank vegetation;

d. reduction of the quality or quantity of stream bank undercuts;

e. increasing stream bank angle which increases water width and decreases water depth.

a. change in channel morphology,

b. altered stream sediment transport processes

a. change in plant species composition (e.g., brush to grass to forbs);

b. reduction of floodplain and stream bank vegetation, including vegetation hanging over or entering the water column;

c. decrease in plant vigor;

d. changes in timing and amount of organic energy leaving the riparian zone;

e. elimination of riparian plant communities (i.e., lowering of the water table allowing xeric plants to replace riparian plants).

Schepers and Francis (1982) found higher nutrient levels in a cow-calf pasture in Nebraska; nitrates increased 45 percent and total phosphorus increased 37 percent. Nutrient levels were correlated primarily with grazing density (Schepers et al., 1982). However, the risk of nutrient enrichment is low in arid rangelands where animal wastes are distributed and runoff is comparatively light. Studies by the Agricultural Research Service and Bureau of Land Management found little evidence of nutrient enrichment from unconfined livestock grazing in Reynolds Creek, an arid watershed in southern Idaho (ARS, 1983).

Nutrient loss from pasture to stream is minimal where streamside pastures remain in good condition. Vegetation buffers the stream from direct waste input and assimilates the nutrients into plant tissue. Gary et al. (1983) evaluated how spring cattle grazing on pastures affected a small stream in central Colorado. Manure recovered within 3 meter strips on each side of the stream accounted for only 4 to 6 percent of the total expected manure production. Nitrate nitrogen did not increase significantly and ammonia increased significantly only once. Although the authors document direct stream deposition, nutrient increase was limited because the pastures were in good condition and grazed only moderately in the spring. Nutrient concentrations were low in a continuously grazed unimproved pasture in a humid site in east-central Ohio (Owens et al., 1989). An earlier study on the same site, using only summer grazing, showed that nutrients did not increase significantly (Owens et al., 1983). Dixon et al. (1983) examined chemical and bacteriological loss from a cow-calf wintering area in southern Idaho, which was irrigated and where return flows entered the stream. The concentration of nitrogen and phosphorus in runoff was a fraction of that observed from cattle feedlots and the authors concluded that the loss of nutrients was small.

Bacteria from the intestinal tract of warm-blooded animals are indicators of fecal contamination and the presence of microbial pathogens. Most state water quality standards use fecal coliform bacteria (FC) as the indicator for determining suitability of the water for recreational and domestic water supply use.

Studies have shown that livestock grazing increases fecal coliform counts over background (ie., coliform counts expected without livestock grazing) (Doran and Linn 1979, Gary et al. 1983, Tiedeman 1987). Bacterial counts increase after cattle are turned in and may remain high after cattle are removed (Stephenson and Street 1978, Jawson et al. 1982). The primary mechanism for bacterial contamination appears to be direct deposition of fecal material in the stream or transport of fecal material to the stream via overland flow (Miner 1992).

In arid rangelands, bacterial contamination may be minimal. Coliform bacteria stayed within a few feet of the manure on a dry Utah rangeland (Buckhouse and Gifford 1976). On rangeland sites in Reynolds Creek in southwestern Idaho, geometric mean coliform count values did not exceed 50/100 ml (ARS 1983).

Once bacteria reaches a stream, bottom sediment may act as a reservoir. When manure was deliberately added to the stream, most of the organisms (90 percent or more) settled to the stream bottom (Sherer et al. 1988). Resuspension of sediments may increase bacterial numbers (Stephenson and Rychert 1982). Sherer et al. (1988) reported that deliberate stream disturbance, such as raking the stream bottom, resuspended sediment and increased bacterial counts. Resuspension also occurs when stream flow increases or when animals walk through streams. However, some additional considerations are provided by Bohn and Buckhouse (1985). For example, coliform populations exhibit daily and seasonal cycles which may influence results. Therefore, individual samples represent the status only at the time of sampling. Coliforms may survive in feces for long periods, up to a year, and coliform concentrations increase with storm and runoff events. Wildlife also contribute to bacterial numbers which may influence results. Thus, coliforms may not be satisfactory indicators since they may die off while pathogens remain viable.

Baxter-Potter and Gilliland (1988) made the following conclusions from a literature review of bacterial pollution from agricultural lands. Like other researchers, they, too, found that the proximity of fecal contamination to the watercourse is significant. If the bacteria cannot be transported by overland flow, its contribution to pollution will be minor. They also found that increased discharge during storms may increase bacterial densities. Other factors, including temperature, wildlife activity, fecal deposit age, and channel and bank storage, affect bacterial densities in runoff.

2.1.1.4 Soil Fertility

See Section IV-2.3.1.1, Section IV-2.8.1.1, and Section IV-2.5.1.1

2.1.1.5 Water Quality/Pesticide Residues

See Section III-2.1.1.3 and Section IV-2.1.1.3

2.1.1.6 Air Quality

See Section III-2.1.1.5 and Section IV-2.4.1.1

2.1.1.7 Botanical Composition, Upper and Lower Layers

In drylands around the world, including the North American southwest, desertification and/or the replacement of productive grasslands and savannas with shrublands and woodlands dominated by unpalatable species appears to have occurred since European settlement. A report by the World Resources Institute (1994) stated that nearly all the world's rangelands except those in the Arctic have been substantially degraded by livestock grazing, the introduction of exotic species, fuelwood harvesting, suppression of the natural fire cycle, wildlife depredation, and conversion to cropland or housing. Rangelands in North America are not subject to the same uncontrolled access and population pressures as in many developing countries. Thus, the outlook is more favorable for reversing negative trends and restoring degraded sites. Even so, it is unlikely that many of the continent's rangelands will ever again regain their former ecological complexity (World Resources Institute 1994).

Warm-temperate grasslands and savannas, such as those which comprised many landscapes in southwestern North America at the time of European settlement, have been replaced by shrublands and woodlands (Archer 1994a). Numerous studies in North America indicate the replacement of grasslands and savannas by shrublands and woodlands since settlement (Chew and Chew 1965, Blackburn and Tueller 1970, Young and Evans 1981, McPherson et al. 1988, Archer 1989). In many instances these changes in vegetation structure are regarded as undesirable because they have reduced the carrying capacity of the land for livestock or have contributed to retrogression. Similar changes have been reported for Africa, India, Australia, and South America. Explanations for the proliferation of woody plants and the associated decline of graminoid species are usually associated with alterations in climatic, grazing, and fire regimes. More recently, studies to elucidate the influence of increasing levels of atmospheric CO₂ offer evidence of another agent of vegetation change. None of these factors individually, or in combinations of less than all, appear to have been sufficient to cause the dramatic changes manifested in rangeland vegetation. For example, during the period 1903-1925, CO₂ concentration increased 11 percent over that of the late 1700s; yet, significant woody plant encroachment had occurred by this time on many range sites. Would an 11 percent increase in atmospheric CO₂ have been sufficient to cause or contribute to woody plant encroachment into grassland (Archer 1994a)? Acting together, herbivory, lack of fire, atmospheric enrichment, and climate have interacted to produce the recent changes, with selective grazing by large numbers and high concentrations of livestock having been the primary force in altering plant life-form interactions to favor unpalatable woody species over graminoids (Archer 1994b). Disturbance and climate can interact to offset or reinforce ecosystem propensity to change. Anthropogenic activities may cause changes independent of climate, or they may reinforce, magnify, or accelerate changes instigated by natural processes. For example, grassland retrogression associated with livestock grazing may be mitigated in years of normal or above-normal rainfall and magnified during years of below-normal precipitation (Archer 1993). As a result, retrogression and desertification can be natural, human-induced, or a combination of the two.

2.1.1.7.1 Vegetation Changes

Shifts in grass and woody plant abundance have broad implications for biodiversity, primary and secondary productivity, soil development and stability, livestock and wildlife composition and carrying capacity, recreational opportunities, water quality, and water distribution. On many rangelands, soil compaction and reduced plant cover from overgrazing decreased the soil's ability to absorb infrequent rain and, thus, started a general drying trend. Lower water tables and a decreased flow are common symptoms of the phenomenon in many western (U.S.) waterways. In many areas, changes in plant cover, documented by photos and historic accounts, have been dramatic. Overgrazing virtually wiped out native perennial grasses in some rangelands, allowing shrubby vegetation, such as sagebrush, mesquite, and juniper, plants whose deep roots can lower water tables and alter the local hydrological balance, to move in. The disruption of the natural fire cycle, which acted as a check on shrub invasion, exacerbated this trend (World Resources Institute 1994). Experts also attribute some loss of range productivity to continued brush invasion, which is especially severe on western rangeland now that the incidence of wildfire has been reduced. Most artificial brush removal efforts employing herbicides or mechanical equipment have been constrained since the 1970s because of environmental and financial costs. Even prescribed burning, the least costly and most environmentally acceptable control technique, faces new restrictions owing to the effects of smoke on air quality. Researchers estimate that approximately 40.5 million hectares (100 million acres) of western rangeland are threatened by sagebrush or juniper encroachment (World Resources Institute 1994). Historical increases in the density of unpalatable shrubs and trees have reduced the carrying capacity and now threaten the sustainability of livestock production in arid and semiarid grasslands and savannas around the world. Biodiversity, wildlife habitat, and nutrient cycling (rates, magnitude, seasonality, and spatial patterns) are also affected by these changes in vegetation (Archer 1995a). Potential explanations for increased abundance of woody plants in dryland ecosystems and how activities of domestic and native herbivores might influence the balance between grasses and woody plants were reviewed by Archer (1995a). Archer argues that in many cases thresholds of herbaceous utilization required to enable woody plants to successfully establish appear to be easily exceeded, even at light levels of grazing. He concludes that grazing management schemes must aggressively incorporate the use of fire as well as emphasize grass utilization and maintenance of species composition if the abundance or dominance of aggressive woody invaders is to be successfully regulated. Archer suggests that cases studies with Prosopis glandulosa, an invasive North American arborescent, show that grazing management alone will not suffice to maintain a favorable grass-woody plant balance. While such management may slow the rate at which stands of woody plants develop, it will not prevent their occurrence if seeds are being actively dispersed into an area. Therefore, grazing management strategies must seek ways to curtail production and dispersal of invasive woody plants (high levels of grass biomass, favoring woody plant seed predators and deferring areas where seed-producing plants occur) and enable the use of periodic fire to limit the size and competitiveness of the Prosopis plants (Archer 1995a).

Changes from herbaceous to woody plant domination also constitute a potentially important global climate feedback by affecting carbon sequestering, nonmethane hydrocarbon emissions, and biophysical land surface-atmosphere interactions (albedo, evapotranspiration, surface roughness, boundary-layer dynamics) (Archer et al. 1994 and Archer 1994b). After considering both the climate change and elevated atmospheric CO₂ hypotheses, in a summary statement, Archer et al. (1994) stated that widespread invasion of woody plants into grasslands and thicketization of savannas is often coincident with intensification of grazing. Direct and indirect effects of livestock grazing (preferential utilization of grasses, alteration of soil structure and chemistry, woody legume seed dispersal, reductions in fire frequency, and/or intensity) significantly influence woody plant population dynamics, generally enhancing woody plant seedling establishment, seed production, plant longevity, and stand development.

Within a climatic regime, the climax vegetation, or natural potential plant community, is a product of all the biotic and abiotic forces that contributed to its formation (Dyksterhuis 1958). This vegetation developed to a “dynamic equilibrium” characterized by an association of species that persisted over time (Figure 1). It is generally agreed that fire is one of the most important of the complex of factors which created and maintained open grasslands (Daubenmire 1968, Sauer 1950) and savannas (Humphrey 1962, Stewart 1956); therefore, it is logical that its removal or reduction in both frequency and scope would contribute to the displacement of grasses by other life forms. This profound influence of fire on grasslands was from naturally occurring sources, such as lightning, as well as from human activities. While activities of pre-European inhabitants may have put little additional grazing pressure on U.S. rangelands, it seems incontestable that these humans affected vegetation in numerous ways. It is most likely that fire used for such purposes as an aid in hunting, in creating areas that were attractive to game, and for additional uses, such as to communicate through burning, to increase visibility and mobility, to reduce insect pests, and to minimize attacks by enemies, helped to shape rangeland vegetation (Sauer 1950, Stewart 1956, Komarek 1962).

Europeans and their activities within many of the world's rangeland areas have had a dramatic effect on rangeland vegetation composition. Although there are several generally accepted, human-associated causes for rangeland vegetation change, overgrazing by domestic livestock is a major factor. As humans became more numerous and sedentary, and their livestock herds increased, they formed proprietary boundaries and began to influence grazing resources more seriously, creating a new disturbance regime and major associated shifts in plant communities. In ancient developed societies, such as middle eastern and Asian countries, increased grazing pressure became a function of population and subsequent increased animal numbers. In the Autonomous Region of Inner Mongolia of the People's Republic of China, for example, increased pressure on grazinglands have occurred within the last four to five decades by the movement of Han Chinese from the south into the vast and formerly sparsely populated steppes.

Schuster (1995) stated that the major cause of overexploitation and deterioration of rangeland in arid and semiarid regions of the world is the human population increase and resulting increase in livestock numbers. This increase results in an imbalance between the resource and its uses. While the infusion of large numbers of domestic livestock and continued overutilization of the range is an obvious cause for the shift, suppression of wildfires and removal of adequate fuel beds to carry fires is also cited as a cause of dramatic vegetation changes in some regions (Scifres et al. 1985, Scifres and Hamilton 1993). Reduction of soil cover and retardance to surface flow following rainfall reduced the effectiveness of precipitation and contributed to soil loss by both wind and water erosion. As early as 1895, Jared Smith concluded:

There has been much written during the past 10 years about the deterioration of the ranges. Cattlemen say that grasses are not what they used to be, that the valuable perennial species are disappearing, and that their place is being taken by less nutritious annuals. This is true to a very marked degree in many sectors of the grazing country (Smith 1895).

Degradation of rangeland is most often accompanied by a loss of productivity and a move from a more mesic, stable environment to a more xeric, more erodible environment (Figure 3). As erosion occurs, soil, which is the base resource available to support future production of vegetation as a moisture-holding reservoir and nutrient pool, becomes less capable of sustaining rangeland yields.

By the early 1900s, uncontrolled grazing by horses, sheep, cattle, and burros had so degraded vegetation and soils in the U.S. that federal legislation (Taylor Grazing Act of 1934) was enacted in an attempt to curtail further deterioration (Stoddart et al. 1975). Scenarios paralleling those of the North American Southwest occurred in Australia and South America. The widespread invasion of woody plants into grasslands, “thicketization,” and increases in the stature and density in savannas typically coincide with development of the livestock industry and intensification of grazing (Archer 1994a). Unfortunately, this has led many to associate the act of grazing with rangeland degradation, when it is excessive grazing that is the cause. Historically, producers have had a natural tendency to stock at a rate that gradually leads to some degree of rangeland deterioration (Pieper and Heitschmidt 1988). Therefore, it is not surprising that a common perception is that livestock grazing is not compatible with many other uses of western rangeland.

The great grassland areas of the world, including the Great Plains and southwestern U.S. evolved with grazing animals as an integral part of the ecological forces which formed them. Yet, in contrast to native herbivores whose numbers or patterns of grazing may vary widely, concentration of domestic animals can be artificially maintained at consistently high levels by supplemental feeding and watering and by protection from natural predation and disease. Fences prevent migration to new areas when the abundance of preferred forages decreases, resulting in higher frequencies and intensities of defoliation and maintenance of grazing pressure over a greater portion of the year and over a higher frequency of years than might otherwise occur (Archer 1994a). The results can be radical changes in species composition and significant soil erosion that are not reversible over time frames relevant to management. Conversely, Lauenroth et al. (1994) found that while settlement of the American Great Plains by European agriculturalists has had major impacts in the past 150 years, there is no evidence that the change from wild ungulates to properly managed domestic livestock has had detrimental ecological impacts. There is an obvious and profound difference between proper grazing use and abusive grazing use on rangeland vegetation.

Pieper (1994) from Strange (1980) used a “retrogressive spiral” to show the process of vegetation changes resulting from excessive herbivory on rangeland (Figure 4). Production losses on rangeland are generally first associated with the livestock products they support. Livestock losses are twofold. As stocking rate of the range is increased to a level where forage availability and quality are diminished below the requirements for individual animals to meet their genetic potential for gains and reproduction, production per animal is reduced. Simultaneously, as overutilization of range resources continues to occur, the production per unit of land area must also diminish as a function of both reduced individual animal performance and degradation of the resource (Figure 5). Native grazing lands are a renewable resource, but they have a finite capacity to fully recover from long-term exploitive use. The process of recovery through secondary succession is often very slow (or even impractical without artificial revegetation), particularly in the more arid regions. There are areas of rangelands within the U.S., North China, and Inner Mongolia and other regions of the world where topsoil loss has rendered landscapes incapable of regaining the pristine vegetation composition and productivity within the lifetime of living humans.

Statistics are available on the condition of approximately 88 percent of the 770 million acres of U.S. rangelands for four reporting periods. These lands include those administered by the Bureau of Land Management (42 percent) and nonfederal lands (46 percent) evaluated by the Natural Resource Conservation Service (Figure 6). There has been a significant shift on nonfederal rangelands from poor to fair and from fair to good range condition during the period from 1963 to 1987. In 1987 there was less than one-half as much poor condition range and twice as much good condition as in 1963 (USDA 1980, 1987, 1988) (Figure 7). Bureau of Land Management-administered rangelands generally followed the same trend as nonfederal lands with only about one-half as much poor condition in 1984 as in 1936 and over twice as much good range condition in 1984 as in 1936 (Figure 8). Caution must be used in attempting to interpret these data on both nonfederal and federal lands. It is probable that different evaluation methods were used for some reports and that all data presented do not represent true range condition because neither PNC (potential natural plant community) nor ecological status were adequately described. Available data in some cases may have been overextended to comply with administrative requests for reports (SRM 1989). Even with such caveats, however, it is generally accepted that U.S. range condition has improved over the past fifty years as indicated by the data in figure 8.

F. E. Clements's concept of a climax or natural potential plant community on North American prairie is often credited with being the first significant contribution to functional ecological principles for use in rangeland management. In 1935 L. A. Stoddart developed a range condition scheme based on climax vegetation from the USDA Natural Resource Conservation Service (SRM 1989). It was not until 1949, however, when E. J. Dyksterhuis published his classic paper on condition and management of rangeland based on quantitative ecology that a deviation-from-climax basis for determining range condition (Figure 9) and for relating range condition to management decisions became widely accepted (Dyksterhuis 1949). The Natural Resource Conservation Service began to use the Dyksterhuis method in the early 1950s. While other agencies were slower to accept the theory and its application to range management, today virtually all U.S. land management agencies use some form of quantitative ecology for determining range condition and relating it to appropriate range stocking rates in order to induce range improvement through secondary plant succession (SRM 1989, World Resources Institute 1994).

It should be noted that the Clementsian paradigm has been challenged both inside and outside the U.S. by range professionals who do not necessarily subscribe to the relationships of climax vegetation and range condition. In their discussion of range ecology and implication for rangeland management in Africa, Behnke and Scoones (1993) state that mainstream range management techniques (described as “equilibrium” systems and referenced primarily to the Clements, Sampson, Stoddart, Dyksterhuis models) do not work in the African situation. Behnke and Scoones state that where perturbances in the system are very frequent, these events dominate and actually drive the system in a continuum of disequilibrium. Range management by African pastoralists in such systems is opportunistic, responding flexibly to stress rather than preventing it, and movement provides a means of circumventing stress under certain ecological conditions. However, Behnke and Scoones also recognize that “mainstream range management techniques are ideally suited to addressing the needs of settled forms of animal husbandry operating under equilibrium conditions (exemplified by fenced ranches in temperate climates).”

This conflict of concepts has received focus recently in the U.S., resulting in an effort to redefine a basis for determining rangeland “health” not corresponding entirely on the percentage of climax vegetation remaining on the site at a point in time. However, two facts are clear: that there is no other more generally accepted, tried, and tested method yet being used, and that the quantitative ecology method has contributed to and facilitated range condition improvement for over four decades in widespread use within the U.S. Although not all people agree that U.S. rangelands have been saved from depletion by quantitative ecology, many do agree. Box and Sisson (1975) stated:

We firmly believe that public lands are in much better condition today than they have been in this century and that this improved condition has been made through management programs of range professionals.

2.1.1.7.2 Grazing and Vegetation Change-Cause and Effect Relationships

Willms et al. (1985) examined the effects of 4 stocking rates on the vegetation in a rough fescue grassland vegetation in southwestern Alberta. Stocking at a light rate (1.2 AUM/ha) for 32 years did not affect range condition. However, a modest increase in stocking rate (1.6 AUM/ha) led to a marked decline in range condition. This result was associated with a change in the composition of rough fescue from 38 to 21 percent of basal area. Rough fescue (Festuca scabrella) was nearly eliminated with a stocking rate of 2.4 AUM/ha. Rough fescue was replaced by Parry oat grass (Danthonia parryi) which increased from 24 percent at 1.2 AUM/ha to 48 percent at 2.4 AUM/ha. However, stocking at 4.8 AUM/ha resulted in severe deterioration of the grassland. This result required the annual adjustment of the stocking rate to avoid animal losses. The recommended stocking rate for good condition range in the area is 1.6 AUM/ha. Recovery of the vegetation within exclosures to a stable condition, took from 14 years in the lightly grazed field to more than the length of the study in the very heavily grazed field. The duration required for recovery was related to the original range condition of the exclosures.

Yonghong and Jargalsaihan (1993) found that as grazing pressure increased, the community abundance decreased in the northeast Mongolian Plateau. The successional series in the gradient were Stipa grandis + Leymus chinensis in the lightly grazed site, Stipa keylovii + Artemisia frigioa + low grasses in the moderately grazed sites, and Carex duriuscula + Artemisia scorparia + annuals in the heavily grazed site. Both species number and richness decreased with grazing pressure increase. Aneurolopidium chinense grassland, a meadow steppe grassland, which distributes from northeastern China to Inner Mongolia, produces a large amount of high quality forage. However, the grassland vegetation has severely degraded, and production has seriously decreased during the past twenty to thirty years, especially in the pastures near villages that receive heavy grazing. During this time period, the number of animals in the villages has increased as many as two to three times. In a study of the ecology of degradation of Aneurolopidium in Heilongjiang Province, Kawanabe et al. (1993) determined the percent coverage of bare ground in lightly degraded pastures was 27 to 33 percent, whereas that of heavily degraded pastures was 47 to 78 percent. Annual plants such as Chloris virgata and alkali-tolerant plants such as Iris loctea invaded degraded pastures, indicating disturbance by heavy grazing on the vegetation. Cutting pastures (pastures used for hay) were little degraded in the study area and had pH levels of 6.17 to 6.46, as opposed to degraded grazed pastures with ph levels of 8.47 to 8.90, thus indicating physical and chemical properties of the soil are negatively modified in heavily grazed pastures. In still another study on Aneurolopidium chinense grasslands, Yuxum (1993) found that aboveground yield was reduced as the increase in grazing intensity increased. Aboveground yield was significantly lower on sites that had been grazed more than moderately compared to no grazing. The effects of root weight to grazing intensity had the same trend as the aboveground biomass but were less dramatic.

McPherson and Wright (1990) studied the effects of redberry juniper (juniperus pinchotii) on species composition and biomass production of the herbaceous layer for three years on an ungrazed site and a site with past cattle grazing in western Texas. Grass production decreased as juniper cover increased. Species composition differed between ungrazed and formerly grazed sites, and the formerly grazed site produced less grass biomass than the ungrazed site during each year. The inverse relationship between juniper cover and grass production was linear on the ungrazed site and logarithmic on the formerly grazed site, suggesting that grazing resulted in a competitive advantage to the overstory species and that succession following grazing will proceed slowly or will be unpredictable. This relationship persisted for a least five years following release from grazing.

In the McPherson and Wright study, basal cover, density, biomass, and species richness of the understory were measured in concentric zones from the stem bases of large redberry juniper (juniperus pinchotii) trees to 6 meters beyond their canopy edges on a shallow, rocky soil and 2 deep soils in the northern Edwards Plateau of Texas. The juniper-driven successional processes of tree dominance, debilitation of understory dominant, influx of subsidiary species, and the general reduction in diversity, density, and biomass of the herbaceous species were evident on all 3 sites. Juniper interference intensified with increasing proximity to the stem bases. Herbaceous biomass in the presence of interference by large junipers on the 3 soils was 1,300, 1,780 and 1,290 Kg/ha, compared to 2,140, 2,140 and 1,560 kg/ha 2 years after the junipers were killed on the 3 sites. Projected herbaceous biomass when juniper populations on the sites develop into closed-canopy woodlands was 320, 880, and 270 kg/ha for the 3 soils. Work by Birdwell in central Oklahoma has produced similar relationships between eastern red cedar density increase and reduction in herbaceous production (Birdwell 1993) (Figure 10).

Homan and Cangyang (1993) reported the degradation of grasslands in the Xinjiang Province of China as a result of overgrazing, excessive reclamation, gathering medicinal herbs and fuel, as well as backward management and decision making. Approximately one-third of Chinese grasslands are degraded, while those of some pastoral banners (counties) are up to four-fifths degraded. In Xilinguole League, degraded grasslands make up about 50 percent of the total area with degradation increasing on an east to west gradient (Hao and Yaozhong 1993). Degeneration caused by the combination of uncertain responsibility, overgrazing and, less money input to the grassland resources have also been reported in the Songmen grassland region (Zisheng and Zhongxin 1993). In a study of plant species diversity of typical steppe grasslands of Inner Mongolia, Xiaoxia et al. (1993) found that light and moderate grazing increased plant species diversity and the stability of the plant communities were maintained. Overgrazing, however, resulted in a decrease of species diversity and simplicity of structure.

Zhijun (1993) studied the static indicators and benefits of investments under different grazing intensities on the Stipa breviflora desert steppe in midwest Inner Mongolia. Results showed that the coverage of the plant community and herbage yield decreased with heavy grazing; the investment in grasslands rose as grazing intensity increased; the annual total output value of grassland was the highest in the moderately grazed rotation, lower in the heavy and light grazing; the annual net output value of grassland was the highest in the moderately grazed rotation, the second highest in the moderately grazed seasonal continuous grazing, and lowest in the heavy grazing; the payoff periods of investment in grasslands were 3.9 and 3.7 years in the moderately grazed rotation and season continuous, respectively, 4.4 years in the light grazing and 40 years in the heavy grazing.

Zhong (1993) reported the total area of Inner Mongolia natural grasslands as 78 million hectares with a utilizable area of 63 million hectares. The grasslands cover approximately 67.5 percent of the total land area of Inner Mongolia. Long-term overgrazing has caused degeneration of 25 million hectares of the steppe and accounts for 39.3% of the available, utilizable grasslands. Even Hulunbier League, an area with pristine and near pristine grasslands in many areas has been the subject of concern because of degradation (Baolin 1993). Zhonghou (1993) reported a GIS study of grassland resources utilization in the Xilingol League of Inner Mongolia. Of twelve counties within the League, 10 counties were overstocked. This overstocking has led to gradual deterioration and sandification in many places of the region, and there are imbalances between segments of grassland not fully utilized because of lack of water and some segments seriously overused and subjected to complete destruction of the ecosystem.

Aixing et al. (1993) found that production varied significantly on two pasture types in the Stipa steppe in Yanchi County, Ningxia between 1964 and 1990. The average sward height of low flat pastures and low gradient hilly pastures decreased from 10.27 and 14.27 cm, respectively, in 1964 to 3.06 and 7.30 cm in 1990. The aboveground biomass decreased from 53.7 g/m² to 15.76 g/m² in low flat pastures between 1964 and 1990 and increased from 45.9 g/m² to 47.42 g/m² in low gradient hilly pastures during the same period. The stocking rate in both pastures decreased from 1.15 sheep unit (SU)/ha and 1.47 SU/ha, respectively, in 1964 to 0.36 SU/ha and 0.54 SU/ha in 1990.

Excessive livestock grazing also has been implicated in global desertification problems. Overgrazing is listed as the principal mechanism in many of the steps leading to desertification of different areas in Chile in South America (Mabbutt and Floret 1980). Many other examples were presented by these authors concerning the detrimental role of grazing in the desertification process in their UNESCO publication.

Elmore and Kauffman (1994), Popolizio et al. (1994) and Armour et al. (1994) indicate that human activities, particularly the introduction of livestock, are an influential perturbation of western riparian ecosystems in the U.S. By the end of the nineteenth century, yearlong grazing resulted in extensive degradation of stream riparian systems throughout the West. Armour et al. (1994) estimated that more than 50 percent of the habitat in western U.S. riparian and stream ecosystems is damaged. Overgrazing in stream corridors has been so damaging to habitat that range managers have referred to these sites as “sacrifice zones” (Stoddart and Smith 1955). In arid and semiarid ecosystems, riparian zones are extremely attractive to grazing animals. They are mesic zones within arid systems and thus furnish an inordinate amount of forage of higher nutritional quality, shade, and protective cover. It is logical that without protection riparian zones will be subjected to extraordinary intensity and frequency of defoliation by grazing animals. This stress results in the reduction of species diversity, often including the loss of trees and shrubs such as Salicaceae.

Grazing systems have been studied in order to ascertain their role in causing change in rangeland vegetation composition. For example, Pitts and Bryant (1987) made comparisons over a 4-year period between 1-herd, 16-pasture, short-duration grazing (SDG) and continuous grazing (CG) on the Texas High Plains. They evaluated animal performance, vegetation response, and diet quality. At equal stocking rates, animal performance on SDG and CG were the same. As stocking rate was increased on the SDG system to double the CG rate, steer gains were lower than those of CG. At 1.5 times the CG stocking rate, SDG gains and CG gains were similar. Other studies have indicated that individual animal performance may decline with increased stocking under SDG (Reese 1986, Heitschmidt 1986), but similar decreases in steer gains have been observed when stocking under CG has been increased (Hart et al. 1986). Diet quality under SDG and CG were not different in the fourth year of grazing, indicating that SDG did not increase or decrease forage quality. The SDG system did not increase standing crop biomass, animal performance, range condition, or improve diet quality over CG in the Texas High Plains.

Willms et al. (1990) tested the hypothesis that time-controlled grazing with high animal densities and high stocking rates will improve grassland condition. Three sites were fenced into cell grazing system with from 12 to 17 paddocks and grazed according to principles of HRM (Savory 1983). Stocking rates were between double and triple the recommended rates. Grazing periods varied from 1 to 3 days during the growing season to 4 to 7 days during the senescent period. The study was conducted over a 6-year period and concluded that timed grazing with high stocking rates, as applied in the study, resulted in no improvement in the condition of grassland. Range condition on the grazed pastures was lower than in exclosures after 6 years of protection, while vegetation densities were greater in protected than in grazed native prairie areas.

Hart et al. (1988) tested claims that rotation grazing systems will increase stocking capacity of range while maintaining or improving animal gain, range condition, and forage production. They compared continuous, 4-pasture rotationally deferred and 8-paddock short-duration rotation grazing at light, moderate, and heavy stocking rates on mixed-grass range near Cheyenne, Wyoming. There were no significant differences in total herbage production, as measured by peak standing crop, among the grazing systems or stocking rates in any year. No differences in botanical composition of peak standing crop were detected among grazing systems, stocking rates, or years. This study concluded that ranchers could stock at rates considerably heavier (60 to 80 percent) than those recommended by the Natural Resource Conservation Service for greatest return/ha. However, the authors suggest that the increase in returns at the high stocking rates is small and perhaps does not compensate for the risk of range deterioration.

Ralphs et al. (1990) also tested the hypothesis that standing crop could be maintained as stocking rate increased using short duration grazing, and thus sustain higher stocking rates. Four stocking rate treatments ranging from the recommended rate for moderate continuous grazing to 2.5 times (1.0x, 1.5x, 2.0x, and 2.5x) the recommended rate were applied in a simulated 8-pasture SDG system. There was little change in frequency and composition of shortgrasses over the study, but midgrass frequency and composition both declined. Standing crop of all major forage classes declined as stocking rates increased. However, the rate of decline was less than proportional to the increase in stocking rate during the growing season. By fall, standing crop was inversely proportional to stocking rate, leading to the conclusion that standing crop could not be maintained at the higher stocking rates. Low standing crop in the fall indicated a potential shortage of forage at the high stocking rates during the subsequent winter. The authors also concluded that forage responses to increasing stocking rates observed under the SDG systems were similar to those expected from continuous grazing at the same stocking rates. In a concurrent watershed study, Thurow (1988b) reported that midgrass cover was eliminated in pastures that were heavily continuously grazed and declined under heavily stocked SDG over a 6-year period compared to moderately stocked continuous grazing. Midgrass cover was maintained in moderately stocked, high-intensity, low-frequency (HILF) grazing, and midgrasses increased under moderately stocked, continuous grazing and where livestock were excluded. During one year of the study, sideoats grama (Bouteloua curtipendula) basal diameter in the moderately continuously grazed pasture and the livestock exclosure was significantly greater than in the SDG pasture. Heavy grazing intensity used in the study, regardless of the grazing strategy, does not appear suited for maintenance of midgrass species.

Martin and Severson (1988) compared the effects of the Santa Rita grazing system, a 1-herd, 3-pasture, 3-year rotation schedule with those of yearlong grazing on semidesert in southern Arizona. Differences in densities of plant species between the grazing treatments were small, and there were no significant differences in herbage production and utilization or shrub density and cover between the grazing treatments. The authors concluded that favorable vegetation responses to rotation grazing were most apparent on ranges initially in poor condition. The authors stated that their experience and the results of their study suggest that the Santa Rita grazing system, like other systems, can accelerate range improvement if the initial condition is poor to fair but may show little benefit on range initially in good condition.

Heavy grazing of sagebrush (Artemisia spp.) grass ranges caused total plant density to decrease and the proportion of the stand as sagebrush to increase (Pickford 1932). These changes caused the grazing capacity of these areas to be reduced by 40 to 75 percent. At the Sheep Experimental Station near Dubois, Idaho, a pasture grazed heavily in late fall from 1924 to 1949 remained in good range condition with an open stand of sagebrush and a good understory of perennial grasses and forbs (Laycock 1967). An adjacent pasture, grazed heavily in both spring and fall, deteriorated to poor condition as grasses and forbs decreased and sagebrush increased. Laycock (1967) concluded that sagebrush/grass ranges can be improved by heavy grazing in the fall by sheep.

Overgrazing of Southwest semidesert shrub/grass ranges in the western U.S. during the early 1900s allowed brush to increase at the expense of perennial grasses and resulted in widespread erosion (Blackburn et al. 1982). Cable (1975) stated that by 1926 many areas of the chaparral type which had been covered with stirrup-high grass stands fifty years earlier were dense stands of brush. Overgrazing has altered the character of California grasslands as much as any other in the U.S. Native perennial bunchgrasses were replaced in the late 1800s by annuals introduced at the Spanish missions (Blackburn et al. 1982).

The Rio Grande Plains of southern Texas and northeastern Mexico offer some distinct examples of processes involved in the conversion of grasslands to woodlands (Archer 1989). Grazing animals are an important component of these processes of change. Brown and Archer (1987) evaluated the relationship between domestic cattle and vegetation change in a savanna woodland with respect to dung deposition and the dispersal and establishment of mesquite (Prosopis glandulosa). On the site with cattle, seedlings were found in 75 percent of dung pats surveyed in September. Although seedling survival in dung (79 percent) was only 16 percent greater than that of mesquite emerging from seeds experimentally sown away from dung, no seedlings were found on areas without cattle. The high density of seedlings with cattle, in contrast to absence of seedlings on the area without cattle, suggests rates of invasion of grasslands by mesquite would have increased substantially in North America following the settlement and introduction of domestic ungulates. Prior to the introduction of livestock, poor seed dissemination and germination may have limited its Holocene spread (Brown and Archer 1987).

In a central Texas experiment, Brown and Archer (1989) determined that grasses are ineffective in excluding honey mesquite except perhaps when there are large accumulations of litter, and that density of mesquite plants could have slowly increased in grassland throughout the Holocene. These researchers addressed the question, why then has Prosopis invasion apparently occurred only recently on many sites? Mesquite seed were apparently available from drainages and riparian zones adjacent to upland prairies; however, limited dispersal prior to the introduction of domestic livestock may have been a primary constraint to grassland invasion. The loss of North American megafauna potentially utilizing Prosopis and its fruit (pods) may have subsequently restricted its spread into grasslands until the introduction of livestock. Relative to Holocene fauna known to consume Prosopis pods and seed, domestic livestock are much more effective agents of dispersal in that they transplant large numbers of germinable seeds away from parent plants harboring host-specific seed predators and deposit them in grazed habits where emergence and survival could be high. Alternatively, Prosopis may have always been present in grasslands, but recurring fires kept plants from developing in structure and producing seed. With the introduction of large numbers and high concentrations of domestic livestock into North America, dispersal of germinable Prosopis seed into grassland would have increased greatly (Brown and Archer 1987), and even moderate grazing would have caused a many-fold increase in emergence. As herbaceous biomass was reduced by continuing grazing, probabilities of Prosopis seedling survival would not have necessarily increased. However, frequency and intensity of fire would have decreased, thus allowing more plants to grow to reproductive maturity and increase the availability of seed for additional dispersal (Figure 11).

Following the “pioneering” of upland prairie sites by Prosopis, these plants served as a nucleus of shrub cluster organization (Archer et al. 1988). Mesquite served as recruitment foci for bird-disseminated seed of other woody species previously restricted to other habitats. The result was a landscape composed of discrete chronosequences of woody plant assemblages organized about a mesquite nucleus (Figure 12). The landscape moves toward a monophasic woodland as new clusters are initiated and existing clusters expand and coalesce. Because the conversion of grasslands and savannas to woodlands in the Rio Grande Plains of South Texas is initiated by mesquite, factors, including livestock grazing, regulating its dispersal, establishment and role as a facilitator of woody community development are emphasized.

Stable carbon isotope ratios of soil organic carbon in southern Texas indicate C₃ woody plants currently occupy sites once dominated by C₄ grasses. Historical aerial photographs (1941-90), tree ring analysis, and plant growth models all indicate this displacement has occurred over the past 100 to 200 years. Succession from grass- to woody plant-domination occurs when the N₂-fixing arborescent, honey mesquite (Prosopis glandulosa) invades and establishes in herbaceous patches. Over time, this plant modifies soils and microclimate to facilitate the ingress and establishment of additional woody species. The result is a landscape comprised of shrub clusters of varying ages organized around a Prosopis nucleus. As new clusters form and existing clusters enlarge, coalescence occurs. Shrubs initially subordinate to the Prosopis eventually contribute to its demise and prevent reestablishment. Changes in livestock grazing and fire regimes appear to have been the driving force behind these historic changes in vegetation; however, the possibility of climatic change and the effects of increases in atmospheric CO₂ must also be considered (Archer 1995b).

Bogush (1952) stated that although cattle are widely accused of being responsible for the change of the vegetation of the Rio Grande Plains of South Texas from grass to brush, they are only part of a more complex series of interrelated factors. As long as the grassland was grazed only by deer, antelope, moving herds of buffalo, and the like, there was little evidence of invasion by the woody members of the arroyo floras (including mesquite) except marginally. Buffalo alone could not be much of a factor, even in the physical distribution of the seeds of the shrubs. The migratory habits of the buffalo carried these animals beyond the range of the mesquite months before the first beans matured. Fencing did much to promote a rapid invasion of the brush and was apparently a strong factor. It was restriction of the range of the cattle rather than the cattle drives which favored the encroachment of the brush. Opposed to these facts are such statements by laymen about exceptional winters between 1860 and 1875, during which time the cattle were said to have drifted into Mexico ahead of northers. In Mexico, sources say, these cattle ate the mesquite and planted seeds on the northward trek back to the home range. Actually, the mesquite and other woody vegetation was already established in the arroyos and elsewhere. Bogush states that during their early ecesic development, seedlings of the woody plants require favorable moisture conditions and are no less vulnerable to drought than seedlings of grasses. Once established, however, woody plant roots, in several species, possess phenomenal penetrating powers and can reach sources of water which are often unavailable to other plants.

Changes in plant composition among herbaceous species is influenced by defoliation. Plant response to defoliation varies depending upon species and growth form differences. Plants which decrease under grazing pressure do so either because they are intolerant to defoliation or because they are highly preferred by herbivores and are grazed more heavily than other plants. However, plants which increase under grazing pressure may do so because they are either relatively tolerant of defoliation or they are less frequently or intensively grazed than others in the community (Archer and Tieszen 1986). Milchunas and Lauenroth (1993) reported that consistent response to grazing appears to be selection for low-growing, prostrate growth forms, an avoidance mechanism.

In their paper on the quantitative effects of grazing on vegetation and soils over a global range or environments, Milchunas and Lauenroth (1993) reported the use of multiple regression analyses on a worldwide 236-site data set compiled from studies that compared species composition, aboveground net primary production (ANPP), root biomass, and soil nutrients of grazed versus protected, ungrazed sites. The authors found that changes in species composition with grazing were primarily a function of ANPP and the evolutionary history of grazing of the site, with level of consumption third in importance. Changes in species composition increased with increasing productivity and with longer, more intense evolutionary histories of grazing. These three variables explained more than 50 percent of the variance in the species' response of grasslands or of grasslands plus shrublands to grazing, even though methods of measurements and grazing systems varied among studies. Years of protection from grazing was a significant variable only in the model for shrublands.

Swank and Oechel (1991) conducted an experiment in the California chaparral to test the interactions among the effects of herbivory, competition, and resource limitation on chaparral herbs and to explain the absence of herbs in gaps in the canopy of chaparral. Four main factors in the experiment included trenching to reduce root competition, caging to reduce herbivory, nutrient addition, and water addition in chaparral dominated by chamise (Adenostoma fasciculatum). No single factor was regarded as causing the absence of herbs. Herb establishment was limited by herbivory and root competition with shrubs for limited nutrients and water. Survival of herbs was enhanced by trenching, caging, and water. Growth was enhanced by nutrient addition, but the magnitude of this nutrient effect became appreciable only within trenched, caged plots. The vegetative cover of annual herbs in gaps was limited by the combined effects of herbivory, competition with the roots of shrubs, inadequate nutrients, and inadequate water. When protected from these effects, herbs flourished in chaparral soil.

Grazing-induced modifications in species composition have been documented in numerous grasslands throughout the world (Branson and Weaver 1953, Ellison 1960, Williams 1969, Noy-Meir et al. 1989). Compositional changes frequently involve the replacement of higher successional species by lower successional species (Canfield 1957). The lower successional species are frequently mid- or shortgrass species held in a subordinate position by competitive interactions with species possessing greater stature (Arnold 1955). Grazing reduces the competitive ability of the mid- and tallgrasses, thereby increasing the relative abundance of lower successional grasses and forbs and establishing the potential for shrub invasion. This scenario of species replacement in response to grazing has frequently been inferred from field observation but has not been experimentally verified. Briske (1991) presents a scenario incorporating elements of developmental morphology, grazing resistance, competitive interactions, and population structure as a mechanistic explanation for species replacement in grasslands.

Plant species whose adaptations to the prevailing climate and soils would make them the competitive dominant of the community under conditions of light grazing may assume subordinate roles, or even face local extinction, as grazing pressure increases. Grazing animals affect plants directly and indirectly. Direct effects of grazing are those associated with alterations in plant physiology and morphology resulting from defoliation and trampling. Grazing also influences plant performance indirectly by altering microclimate, soil properties, and plant-competitive interactions. These indirect effects accentuate plant response to defoliation in ways not readily simulated by clipping experiments. Over time, the combined direct and indirect effects of grazing on plant growth and reproduction are manifested in plant population dynamics. Herbivores affect the productivity, composition, and stability of plant assemblages through mediation of plant natality, recruitment, and mortality and may cause directional changes in community structure and function.

The impact of livestock grazing on ecosystems varies in relation to the evolutionary history of the site and the level of grazing pressure (Stebbins 1981, Milchunas et al. 1988). Intermountain grasslands of North America evolved with light grazing and have changed markedly since the introduction of livestock (Mack and Thompson 1982). In contrast, tall-, mixed, and shortgrass prairies of North America, which evolved with bison, pronghorn, and prairie dogs, have been relatively resistant to stresses associated with livestock grazing. Although plant species in ecosystems that evolved with grazing are well adapted to defoliation, domestic livestock can substantially impact their growth and persistence in numerous ways (Pieper and Heitschmidt 1988). In grassland or savanna systems that occur in areas climatically and edaphically capable of supporting trees and shrubs, prolonged grazing may decrease the capacity of grasses to competitively exclude woody plants, while at the same time reducing fire frequency and intensity by preventing the accumulation of fine fuels.

Where grazing intensity is high, the grazing avoidance-type plants will inevitably dominate the site because of the overriding influence of utilization. Evidence for the premise is most striking from southwestern hot deserts, the Great Basin, and the Southern Plains, where extensive areas of grasslands characterized by grazing tolerant-type plants (perennial grasses) have given way to shrublands dominated by unpalatable grazing avoidance-type plants. In California, grazing-tolerant perennial grasses appear to have given way to annual grasses which could be viewed as employing an avoidance or escape-type strategy (Archer and Smeins 1991).

Grasslands are able to tolerate a moderate degree of grazing intensity before changing in composition, diversity, or productivity. However, as grazing intensity is increased or becomes continuous, tall- and midgrasses eventually give way to short-statured perennial grasses, which, in turn, give way to annuals and unpalatable perennials with a concomitant loss of primary and secondary productivity, diversity, cover, and soil (Archer and Smeins 1991).

Also, see Section III-2.1.1.6 and III-2.1.1.8.

2.2 Forest Utilization

2.2.1 Direct Indicators

2.2.1.1 General Background

Forests of the LGT region of the northern United States are different from those which existed before European settlement. Conditions and events, both natural and man-caused, have combined to shape the forests as they exist now. Major changes in species composition of forest grazinglands have occurred since European settlement (Adams, 1995). These species shifts are attributable to a combination of logging, grazing, fire suppression, accidental introduction of exotic pests, changing climate, and related events over the past century. For example, in Idaho for the period 1952 to 1987, western white pine declined 60 percent, ponderosa pine declined 40 percent, and true firs, lodgepole pine, and Douglas-fir increased 60, 39, and 15 percent, respectively.

In some ecosystems species composition of trees has changed to species which are more susceptible to insect-and disease-caused mortality and growth loss. These species are generally more shade tolerant, resulting in retention of lower branches and foliage and creating “ladder” fuels which carry ground fires into the crowns. Tree densities are also generally greater than in the past. Increased densities create moisture stress in trees during draught years and cause trees to become more susceptible to damage from insects and diseases. Large-scale mortality of trees can create severe fire control problems.

A second change is in forest density. This is particularly true of ponderosa pine forests which were developed under a regime of frequent, low-intensity, ground fires. Fire suppression has contributed to a dense, overcrowded forest. In a study area on Arizona's Coconino National Forest, pre-settlement density averaged 23 trees per acre while current stem counts are 851 trees per acre. On the Boise National Forest in Idaho, density on a study plot was estimated to be 29 trees per acre for the 300+ years before 1906. Density increased to 533 trees per acre and the species composition also shifted from predominately ponderosa pine to predominantly Douglas-fir.

2.2.1.2 Effect of Livestock Grazing

A common contention is that grass and shrub communities of the Intermountain Region evolved largely in the absence of large-hooved herbivores (Covington 1984). Although this contention is not fully substantiated, it can be stated that with the introduction of domestic livestock, unprecedented herbivory began on forest and rangeland ecosystems of the entire region. Livestock grazing differed significantly from that of native herbivores in intensity, season, and dispersal of use across the landscape. Among the results of the introduction of domestic livestock was a reduction in forage yields and litter, which can alter fire intensity and extent. High-frequency, low-severity fire regimes characteristic of forest and rangeland ecosystems in the Intermountain Region prior to livestock grazing have changed to ones of lower frequency but much greater intensity.

2.2.1.4 Botanical Composition

Payne (1985) identified three major cattle/tree crop (sylvo-pastoral) systems in the tropics: a) grazing and/or browsing in natural forest; b) grazing or harvesting forage grown under planted trees, including trees grown for timber and/or firewood and trees grown for nut, fruit, and industrial crop production as well as for timber and/or firewood; and c) browsing or harvesting of tree forage. A major advantage of cattle/tree crop systems in the wetter tropics is that, to a limited extent, they simulate the rainforest ecosystem that they replace, which has the following biological advantages: a) available solar energy is used efficiently by the plant biomass because of the vertical stratification of vegetation components of the system and because the soil is protected from severe erosion by two or more plant storys; b) stratification of the root systems of the different plant species, whose roots occupy different soil horizons, ensures that as wide a range as possible of essential nutrients are removed from the soil and c) if the trees used in the system are legume species or other species that can fix nitrogen, such as the alder (Alnus acuminata), which is planted in the Costa Rican highlands and fixes nitrogen through the agency of a fungus (Actinomyces alni), these assist in improving soil fertility, as does the use of forage legumes in the pasture mixture grown under the trees. Long-term improvement in soil fertility is possible due to the effect of cattle continuously grazing grass legume pastures. Where cattle are grazed under fruit or nut trees, soil fertility may be further enhanced if tree crop by-product feeds, such as coconut meal, are returned to the tree crop area and fed to cattle (Payne 1985).

In addition to the biological advantages, there are four major economic advantages in cattle/tree crop systems. First, annual costs of weeding under trees is reduced, since grazing reduces weed competition. In fact, the first use of cattle in many coconut plantations was as “weeders” or “brushers,” and foresters now realize that the use of cattle in forest plantations may appreciably decrease annual weeding costs. Second, product output and labor input becomes more diversified, along with the more effective utilization of labor on an annual basis. Third, a possible increase in total product output per unit area of land results, and, where cattle/tree crop systems replace monocultural tree crop systems, an increase in the total value of output per unit area occurs, as animal products normally possess a higher unit value than plant products. Finally, the possibility emerges of managing higher grade and, therefore, more productive cattle under tree crops than on open grazing land in the same climate. This last possibility exists because ambient temperatures and heat stress on the cattle is generally lower on grazing under tree shade than on open pastures.

In drier regions of the tropics, trees are valuable not only as a source of nutrients to enhance soil fertility and shade for livestock but also as protection for the soil from excessive erosion. Trees are also a major source of nutritious forage during the long dry season. Continuous overstocking depletes the ground vegetation, induces erosion, and, combined with continuous firing undertaken to induce new grass growth, reduces the number of tree species to those that are fire resistant. If fire resistant trees are then continually lopped to obtain forage in the dry season, these dry-land ecosystems rapidly degrade. The major problem, therefore, in the drier regions of the tropics is not to learn how to introduce new cattle/tree crop systems but to reverse the degradation of traditional cattle/tree crop systems (Payne 1985).

Knowles (1991) reviewed 20 years of experience in New Zealand with silvo-pastoral systems. Trials were established in the early 1970s to measure the effect of the tree crop on understory. Sheep were the main experimental animals in the trial, but beef cattle, deer, and goats have also been used. In the trials, sheep were confined to a particular treatment for the year, with stocking rates set according to the measured pasture yields. The increasing effect of a developing canopy caused a decline in feed quality, as indicated in reduced wool weights and livestock growth rates, particularly for young animals at any given feed allowance. Livestock continuously grazed under trees results in a significant negative effect on sheep performance, but if livestock spend only a part of their time under trees, the effect may not be noticeable. Although abortions in cattle grazing on radiata pine have been reported, it does not appear to be a consistent problem with large herds with a history of continuous grazing among trees. Current guidelines recommend that during the last three months of pregnancy in-calf cows should not be grazed among stand of young radiata pine receiving silvicultural treatment (Knowles 1991).

Tsiouvaras et al. (1989) studied the effects of goats on understory vegetation and fire hazard reduction in a coastal forest of Monterrey pine (Pinus radiata) and red gum (Eucalyptus camaldulensis) in California. Stocking rates of 600 Spanish goats per hectare for 1 day reduced cover of the brush understory in the forest by 41 percent and 48 percent, respectively, at heights of <0.5m and 0.5-1.5m. In the fuelbreak stocked at 280 goats per hectare for 3 days, the reduction was 46 percent and 82 percent at heights of <0.5m and 0.5-1.5m, respectively. Forage biomass utilization reached 75 percent in the brush understory and 84 percent in the fuelbreak. In the brush understory, goats grazed more at the upper height (84 percent) than at the lower (64 percent). Grazing reduced 1- and 10-hour dead fuels 33 percent and 58 percent, respectively, while the litter depth was reduced as much as 27 percent.

Much of the Douglas fir-zone in the northern Rocky Mountains of the U.S. is characterized by understorys dominated by tall and medium shrubs. The value for browse of the shrubs has been reported as a variable, depending upon species. Shrub species that are preferred by domestic livestock and big game mostly occur in open sites following logging or fire, while shrub species, most prevalent as a stand approaches climax, are unpalatable and seldom utilized as forage (Pengelly 1963). A 2-year food habit and distribution study by Mitchell and Rodgers (1985) of a cow-calf herd on summer range in northern Idaho showed that by early July up to one-half of the cattle diet came from forest species, primarily browse.

Lewis et al. (1988) studied plant responses to pine management and deferred-rotation grazing in north Florida. Frequency of occurrence of herbaceous species and foliar cover of woody species were determined in natural stands of 50-year-old slash and longleaf pine (Pinus elliottii and Pinus palustris). Occurrences were compared to similar forest sites that were harvested and site prepared by double-roller chopping and not replanted with slash pine, or replanted to 1,112 trees/ha in single- and double-row configurations. In addition, these sites were ungrazed or grazed using 3 deferred-rotation systems. Both burning and mechanical disturbances initially reduced foliar ground cover of most woody species; however, few species were eliminated from the community. Most woody species were recovering within 6 years from treatment, but succession was somewhat slower on mechanically treated areas. Survival and growth of planted pines were not affected by grazing; and planting configuration did not affect pine growth. Grazing per se and deferred-rotation treatments had no significant influence on any of the parameters measured on planted pines through age 5. Tree growth was not affected in any year by any deferred-rotation grazing treatment. Also, crown development was not affected by grazing treatments. Although cattle sometimes break lower limbs by rubbing on stems, this does not change the height to the lowest living branch.

Clearing and thinning of caatinga vegetation in northeastern Brazil are viewed as methods of optimizing forage and wood production. Schacht and Malechek (1990) compared the botanical composition of goat diets relative to forage availability in undisturbed, cleared and 2 levels of thinned (25% and 55% canopy cover) stands of tropical woodland. Clearing and thinning of caatinga vegetation resulted in higher amounts of available forage through the wet season and up to the time of leaf fall. At the end of the growing season, available herbaceous biomass was generally 7 to 8 times higher on the treated pastures than on the control; biomass of available browse was about 4 times greater. After leaf fall, total available forage was similar for all 4 treatments but about 90% of the available forage on the control was leaf litter. Even though browse availability was high throughout the wet season on the treated pastures, herbaceous vegetation was the primary dietary constituent. Only during the mid to late dry season, when herbaceous vegetation was dead and leaf:stem ratios were low, was browse consistently selected at high levels. The authors concluded that clearing and thinning increases the amount and diversity of available forage; thereby, improving foraging conditions. Moreover, production of herbaceous vegetation declines towards control levels only at some canopy cover higher than 55%.

Sharrow et al. (1989) conducted a study to evaluate the potential for using herded sheep to control competing vegetation in Douglas-fir plantations in Pacific Northwest coniferous forests. Three 4- to 6-year old plantations were grazed once each year during the May to September grazing season. Estimates of current year's growth present in October, both inside and outside a livestock exclosure on each study plantation, were used to evaluate the effects of grazing. In general, utilization of brush by sheep was moderate to heavy, except in the spring of one year during the 5-year study, when brush was lightly utilized. Sheep grazing effectively reduced both total understory plant growth and brush net current year's growth on all plantations. Reduced brush biomass on grazed areas was associated with greater Douglas fir diameter growth in 2 study years. By the fourth year of the study, trees in grazed areas were 5 percent taller and 7 percent greater in diameter compared to ungrazed controls. These data suggest that sheep may be effectively used as a biological control agent for brush control in coastal Douglas fir forests.

Putman et al. (1989) reported an experiment that established two 5.6 ha enclosures within an area of heavily grazed deciduous woodland in the New Forest, Hampshire, England. A constant grazing pressure was maintained (c. 1 fallow deer ha^-1) in one enclosure, while the other was kept free of all large herbivores. The vegetation of both was surveyed 6 years, 14 years, and 22 years after enclosure. Clear differences were apparent between the two enclosures and with the ungrazed site over time. While in the grazed plot no regeneration was apparent, rapid regeneration of birch, beech, oak, Scots pine, Douglas fir, and holly had occurred in the ungrazed plot by the 6-year evaluation; by the 22- year evaluation, with closure of the canopy, establishment had virtually ceased. Clear differences were also recorded in species composition of both trees and ground flora. Not only did species resistant to grazing become more abundant in the grazed plot but also many graze-sensitive or palatable species that were absent in that plot became reestablished in the ungrazed area. See also Sections IV-2.2.1.3 and V-2.2.1.3

2.8 Habitat Composition Changes and Biodiversity

2.8.1 Direct Indicators

2.8.1.1 Range Utilization

2.8.1.1.1 Spatial

Most rangelands in LGT systems are grazed by more than one grazing animal. Most management strategies use the animal unit equivalent (AUE) method to allocate forage to the different grazers using the same area. The AUE method places the different animal grazers on an equivalent base of forage consumption. For example, if a cow eats 25 pounds (dry weight) of forage per day and a deer consumes 5 pounds, then 5 deer equal 1 cow. Arguments have ensued over the equivalency of various herbivores, often with the argument being based on the special interest of the involved parties. Hunters want 3 elk to equal 1 cow, while ranchers want 2 elk to equal 1 cow. In only some cases are spatial and forage preference differences recognized. This simplistic approach to forage allocation may work as a basis for formulating management plans, but it all too often ends up as the actual mechanism for allocating forage to animal grazers.

Research projects have often recognized both spatial and dietary differences of various herbivores. Models predicting competition (Nelson 1984) have been developed. However, Bowyer and Bleich (1984) found fewer deer grazing meadows in montane southern California that had been previously grazed by cattle. Klein (1984) considered competition for habitat from livestock to be one of the critical problems facing large mammals in North America. The authors also listed accelerated energy development and forest-cutting practices as contributors to habitat loss. In some cases facilitation has been suggested (Anderson and Scherzinger 1975, Urness 1982). Anderson and Scherzinger (1975) reported that elk numbers on winter range increased after controlled cattle grazing was initiated to precondition forage for increased nutritive quality. Cattle, horses, and sheep were used successfully to improve deer winter range in Utah (Urness 1982). Briefly, then, on rangeland grazed by multiple animal grazers, anything is possible. Management must be as dynamic as the system. Simplistic paradigms and broad brush management plans do not work.

2.8.1.1.2 Richness/Selectivity

Conversion of native communities to those more suitable to livestock production is often associated with livestock grazing (Vavra 1994). Brush removal with or without reseeding to exotic species is often detrimental to some species of wildlife. In sagebrush communities where the brush is removed, declines in sage grouse, Brewer's sparrow, sage thrasher, sage sparrow, and raptors can be expected. Horned larks and vesper sparrows may increase. Studies of population changes due to habitat alteration were not easy and needed long time frames. Baker et al. (1976) recognized grazing by livestock as a desirable use of sagebrush rangelands if grazing was controlled to prevent range deterioration and guidelines were established for wildlife.

In the LGT region of the north-central U.S., riparian areas provide important habitat for terrestrial animals and critically influence the stream systems they border. Water quality and quantity are increasingly being referred to as the most important issues in Western North America. Riparian vegetation is of critical importance for fish because it provides escape cover, lowers summer water temperatures through shading and reduces stream bank erosion that can silt in spawning and rearing areas. Livestock make important use of riparian areas, particularly in extensive management systems with little control (Gillen et al. 1985).

2.8.1.1.3 Temporal

Cattle grazing can degrade or improve rangeland vegetation through herbivory and trampling. Evidence is scant that herbivory directly benefits plants (Anderson and Scherzinger 1975, Ellison 1960). Biomass production (Anderson and Scherzinger 1975, Nelsen 1984) seed size and yield (Jameson 1963), longevity (West 1979, Wright and Van Dyne 1976), tiller production (Branson 1956), and water-use efficiency (Nowak and Caldwell 1984) are not increased by herbivory. However, cattle herbivory can indirectly benefit plants by reducing fires and rodent herbivory resulting from accumulations of vegetation. Grazing can also prevent excessive mulch from inhibiting seed germination and seedling establishment, at least on sites receiving high amounts of precipitation (Klein 1984). Cattle herbivory can affect the type of plant tissue produced. For example, cattle herbivory can remove apical dominance and cause plants to initiate new tillers from basal buds (Dahl and Hyder 1977), which can improve the vegetation's nutritional quality and accessibility to herbivores (Urness 1982).

Most rangeland plants evolved under large-ungulate grazing (Platou and Tueller 1985) and developed morphological and physiologicl mechanisms to cope with herbivory. In addition to these mechanisms, plant tolerance to cattle herbivory depends greatly on the competitive pressures exerted by surrounding plants (Caldwell 1984). Competitive pressures among plants in a community also largely determine plant community composition. Herbivory can alter these competitive relationships. Shifts in plant community composition are termed improvement or degradation depending upon the value system used to evaluate the changes (Heitschmidt and Taylor 1991).

Cattle herbivory can also affect plant species richness. Light to moderate cattle grazing usually results in greater plant species richness than does heavy grazing or no grazing at all. Grazing especially improves plant diversity wherever plant growth is abundant. Light to modrate grazing increases diversity by opening the vegetation canopy and enabling more species to compete successfully. Diversity is reduced on ungrazed sites because the thick vegetation canopy intercepts light and moisture; heavy grazing reduces diversity by gradually eliminating the most palatable plant species.

Cattle may aid seed dispersal of rangeland plants by physically transporting seeds between sites. Judicious trampling can help reduce excessive accumulations of mulch, but excessive trampling can injure plants, especially when soils are moist. On dry soils cattle trampling may increase seedling emergence of some plant species but decrease emergence of others (Eckert et al. 1986, Weigel et al. 1990).

2.8.1.2 Forest Utilization

2.8.1.2.1 Spatial

Throughout the late Pliocene and the Quarternary ice ages, species of ponderosa pine bunchgrass ecosystems have migrated up and down in elevation and latitude, tracking favorable bioclimatic conditions (Covington 1994). At various points in time, ponderosa pine bunchgrass communities were much more prevalent than they are today. These communities were most common during the Pliocene (two to five million years ago), when these ecosystems provided extensive habitat for evolving prehistorical megafauna. Ponderosa pine-bunchgrass ecological systems have coevolved with frequent surface fires and open parklike conditions for at least the past two to five million years. As a consequence, many constituent species have evolved to require frequent fire and open stand conditions for their survival.

Landscape patterns and disturbance events occur at a variety of spatial and temporal scales, so their evaluation and creation must be made in a hierarchical framework. To achieve positive cumulative effects and increase flexibility in future management options, the context and justification for action at one scale must be defined at the next higher scale. Vegetation patterns and disturbances at a lower scale have site-specific values, but they must first contribute to the desired values at the larger scale if sustainability of the larger system is to be achieved.

Past forest management practices did not conserve disturbance or recovery processes. Fire suppression has altered fire regimes and ecosystem structures that developed under a high-frequency and low-severity fire regime. Extending the time frame between disturbance events caused significant increases in fuels and increased potential for intense and large fires. In other areas, harvesting and grazing exceeded historical levels such that sites were disturbed prior to their full recovery from the previous disturbance events. In both instances, successional development created stands and landscape patterns that differ from historical conditions (Everett 1994).

In the forest ecosystem of the Blue Mountains of northeastern Oregon and southeastern Washington, large-scale systematic disturbances, primarily logging and fire supression, have profoundly influenced vegetation patterns and vegetation-herbivore interactions. The most important forest ecosystems for large herbivores in the Blue Mountains are those at mid-elevations that include open plant communities in ponderosa pine and Douglas fir zones and mixed conifer communities at higher elevation. Large herbivores often graze early seral stages heavily after timber harvesting in the mixed conifer zone (Skovlin et al.1976). Steppe and shrub-steppe zones at lower elevations and high mountain and subalpine zones at higher elevations contain plant communities of great importance to the grazing herbivore. The low elevation zones include grasslands, sagebrush, and juniper woodlands in association with foothills and canyons that are particularly important as spring, fall, and winter range for all classes of ungulates (Sheehy and Vavra 1996).

2.8.1.2.2 Temporal

The greatest impact of cattle on wildlife is through the alteration of the quality of the habitat (Vavra 1994). This impact is both short-term and a long-term and may involve quite obvious changes in habitat structure or more subtle changes in compositon of the plant or animal community. Examples of short-term effects of cattle grazing on habitat quality include reduction in understory cover. Understory cover is critical for many species of ground nesting birds during the nesting season. It is also important as hiding cover for fawns of such big game species as pronghorn antelope and mule deer. Nesting/fawning success with such species is often the most important factor affecting population size from year to year.

Long-term effects of cattle grazing include drastic alteration of habitat structural diversity. This alteration is most noticeable and most deleterious on riparian zones where areas that originally supported a complex mixture of tree, shrubs, and herbaceous species have now been reduced to limited herbaceous ground cover or bare gound. These changes are accompanied by drastic reductions in wildlife species richness. Typically, this change in habitat structure has been accompanied by significant hydrological degradation of watersheds and streams including depressed water tables, increased runoff, lower dry season flows, and increased water temperatures. These latter changes cause reductions in fish biomass and species richness.

The assertion that cattle grazing under certain circumstances can cause ecosystem damage does not imply that it always does so. There are areas of the western United States and Mongolia that have sustained many years of cattle grazing with little evidence of loss of ecosystem structure or function or any known loss of biodiversity that can be attributed to grazing practices.

The seriousness of a dysfunctional ecosystem from grazing makes it imperative that we understand better the role of cattle grazing in causing, exacerbating, and/or alleviating such breakdowns. The Great Basin is now experiencing widespread invasions of exotic plants - a symptom of ecosystem dysfunction. Managers need to understand the role of cattle grazing in contributing to this phenomena.

2.8.1.2.3 Richness/Selectivity

The impacts of Euro-American settlement on ponderosa pine ecosystems has been devastating to native biotic communities (Covington 1994). For example, fire scars on ponderosa pine trees near Flagstaff, Arizona, stopped suddenly in 1876. The reason for the lack of fire scars after 1876 is linked to the thousands of head of livestock which were introduced into the area in 1877. The ensuing overgrazing effectively eliminated the herbaceous fuels which had previously carried fire across the landscape and disrupted the frequent fires which had been so important in maintaining open parklike forests.

Measurement of current tree stand and understory conditions near Flagstaff, Arizona, indicate that in 1876 over 80 percent of the area was dominated by herbaceous vegetation while only 17.3 percent was under tree canopies (Covington 1994). By 1990 virtually all of the bunchgrass area had been converted to pole stands and dog-haired thickets of sapling-sized trees. The relict herbaceous openings now constitute less than 4 percent of the area studied. Early photographs taken before the heavy livestock grazing and from early inventories indicate that bunchgrasses, wildflowers, and shrubs originally grew close to the base of old-growth trees. Today, a careful excavation of the thick forest floors reveals the buried silicate skeletons of these plants at the soil surface. The conclusion reached was that a tremendous simplification of net primary productivity throughout ponderosa pine ecosystems has occurred to the point that in many areas virtually all of the net primary production is concentrated in trees

2.8.1.2.4 Suitability of Animal to the Environment

Livestock grazing can change habitat composition, especially where grazing use is concentrated and distribution constraints are major limiting factors to maintaining an equilibrium carrying capacity. Studies of the effect of large herbivores on plants and soils in the Blue Mountains of Northeastern Oregon suggest that herbivory by large ungulates can cause moderate to severe reduction of shrubs in a variety of logged and unlogged forest communities and that deer and elk can have important effects even in the absence of livestock. Forest communities in the Blue Mountains respond relatively rapidly to reduced grazing pressure, requiring a decade or less for measurable effects to occur after ungulates are excluded. However, long-term loss of seed viability of preferred shrubs may diminish the potential for natural regeneration, causing an important loss of plant biodiversity.

On summer forested rangeland subject to herbivory, changes in forest management that maximize herbivory by both domestic and wild ungulates are possible. Previously, dispersed timber-harvest units were of limited size to optimize arrangement of cover for wild ungulate security and open areas used for foraging. Creating larger timber harvest units would increase the probability that seral shrubs in treated areas would escape intense browsing to allow for greater expression of understory vegetation

2.8.1.2.5 Livestock-Wildlife Interactions

2.8.1.2.5.1 Effects of Livestock Grazing on Salmonids

Incubation of embryos and emergence of fry are the most sensitive stages in salmonid reproduction cycles. Recommended levels for successful incubation are at or near oxygen saturation with temporary reductions in dissolved oxygen no lower than 5 mg/l (Bjornn and Reiser 1991). Low-dissolved oxygen concentrations affect growth, food conversion efficiency, swimming performance, and survival. High turbidity reduces sight feeding and growth and interferes with migration. Salmonid sight feeding is impaired at turbidities in the range of 25-70 NTU. Salmonids will migrate in water of higher turbidity; however, they avoid such waters for rearing and feeding (Lloyd et al. 1987). Recommended levels to protect salmonids is 50 NTU, measured instantaneously, or 25 NTU, measured over a ten-day period (Harvey 1989).

Clean substrates are important habitat components because salmonids build nests (redds) in gravel and cobble substrate. Clean substrates are required to provide dissolved oxygen to the embryo, remove metabolic wastes, and allow alevins (fry) to emerge from the redd. Sediment from erosion reduces the survival of embryos. During spawning, salmonids also need adequate cover for escape and hiding. This cover is provided by overhanging vegetation, undercut banks, submerged vegetation, submerged objects such as logs and rocks, attached floating debris, deep water, turbulence, and turbidity. Stream flow is also important because it determines the amount of spawning area available by regulating the area covered by water and the velocity and depth of water over the gravel beds. Preferred water depth and velocity have been established for many species. Grazing management practices can alter the hydrologic regime by increasing peak flows and decreasing base flows. These changes decrease the amount of habitat available for salmonids at critical times in their life cycle.

2.8.1.2.5.2 Effect of Livestock Grazing on Wild Herbivores

Multiple herbivore systems are common in Livestock Grazing Temperate (LGT) systems of the United States, Mongolia, and Inner Mongolia. In these regions, large wild herbivore and livestock use of habitat often overlaps, at least during a portion of the year. In the western United States, it is common to have as many as three large wild herbivores potentially competing with livestock for habitat (Vavra 1994). Most grazingland management systems developed for rangelands in the western United States need to account for the active use of one or more wild herbivore in developing grazing management strategies.

2.8.1.2.5.2.1 Competition between Cattle and Wildlife

Potential competition between cattle and wildlife on public lands in LGT systems, including destruction and/or degradation of forage and cover, migration routes, water, and disease interactions is not well understood (Vavra 1994). Often the public perception of this interaction is incorrect. However, the public perception of these interactions influences management and actions of land administrators. Cattle and wildlife have the highest potential competition for feed, cover, and space but somewhat less for water (except for wild horses and burros). Indirectly, the potential for competition exists in the alteration of wildlife habitat by cattle with both short-term (change of cover) and long-term (change of plant community and habitat quality) effects. Fencing of traditional migration routes, when it occurs, has been shown to be easily remedied by several sources in the literature and has been overplayed as a major concern in wildlife management. Eradication of native wildlife species that pose as potential disease reservoirs to beef cattle in particular and livestock in general should not be practiced.

Cattle can compete with wildlife for food or water. Of the two, potential competition for food is the most important. In the LGT of the north-central United States, roughly 90 percent of the forage utilized by big game species prior to 1850 is now being utilized by cattle (Cooperrider 1994). These estimates are based upon a number of necessary but simplistic assumptions, but a vast amount of evidence supports the general statement that a major proportion (over 50 percent) of the forage once available and utilized by big game animals such as bison, (Bison bison), bighorn sheep (Ovis canadensis), elk (Cervus elaphus), antelope (Antelocapra americana), and mule deer (Odocoileus hemionus hemionus) is now being utilized by cattle. This level of usurpation of available forage is partly responsible for the lower numbers of most big-game species now present in the Western U.S. However, numerous other activities have occurred during this period which have also contributed to declines in big game species. Therefore, at the national level few generalizations can be made about the impact on big game species.

Cattle can impact big-game species by utilizing key forages that big-game species need during critical times of year (Cooperrider 1994). Typically this occurs on winter ranges in the north, high altitude or other summer ranges in the south, and in places such as California it may be on transitional ranges. Cattle are not normally affected by big-game species because a) big-game species are found in such low numbers and/or biomass compared to cattle, b) most species of big game have dissimilar forage preferences compared to cattle, c) big-game species often utilize or prefer different habitat from free-ranging cattle, and d) cattle are managed to avoid having to forage on ranges during the critical times of year, i.e., they are usually moved to other ranges or fed hay during the critical months.

Beyond direct competition, cattle can interfere with use of resources by wildlife, and they alter wildlife habitat quality, both short-term and long-term (Cooperrider 1994). “Interference” competition occurs when one species prevents another from utilizing a resource. A few examples of this type of competition have been suggested. Both elk and bighorn sheep have been observed to avoid range areas used by cattle. However, detrimental effects on populations of either species have not been documented. Predation by wildlife on cattle, which may be viewed as an additional form of “competition,” has not recently been an important impact on livestock production, but with reintroduction of certain predators and increased protection, predation may again become a major factor influencing livestock production in some areas.

Cattle can transmit diseases to wildlife species and vice versa, with minimal effects to both, with detrimental effects to only one, or with detrimental effects to both (Cooperrider 1994). Although local situations, such as with bison and brucellosis in and around Yellowstone National Park, can be contentious, they are neither widespread nor unmanageable. Thus, disease transmission to or from cattle is neither a major problem with wildlife nor a major drain on the cattle industry from a national perspective. This statement is made from the perspective that a) cattle grazing on wildlands anywhere will be subject to a normal array of native diseases carried by native wildlife populations, b) dealing with such diseases is a normal cost of doing business, and © extirpation of native wildlife to eradicate cattle diseases is neither acceptable nor economical and certainly is not sustainable.

Cooperrider (1994) has summed up many of the potential conflicts that exist between livestock and wildlife. These include

a) hay farming and other agricultural practices accompanying cattle operations, which typically requires diversion or development of water which may severely reduce the quality of both the aquatic habitat and also the adjacent riparian vegetation, causing significant declines in both aquatic and terrestrial population;

b) removal of large predators, particularly the red and gray wolf (Canis rufus and Canis lupus) and the grizzly bear (Urus arctos), and current efforts to reestablish these species into portions of their former range;

c) fencing associated with cattle ranching, which is not a major problem with wildlife in most places but need not be a problem anywhere; and

d) plant community-type conversion use of herbicides, chaining, or other artificial means followed by seeding with exotic grasses for the purpose of increasing livestock forage.

The use of cattle as a tool in wildlife management, including localized management, especially on riparian areas, wetlands, lowland hardwood forests, and other areas critical to wildlife, may not always have beneficial effects for wildlife unless wildlife dietary and habitat requirements are understood and utilized (Vavra 1994). Cattle can be used as a tool for wildlife management, but some claims about the benefits of cattle grazing to support this objective have been extravagant. Although much data supports the use of cattle as a beneficial tool for wildlife management, many areas of interaction, especially with nongame species, are unknown.

2.3 Indirect Indicators

2.3.1 Cropping Systems

See Section V-2.3.2.3.

2.3.2 Animal-Mechanical Power

See Section V-2.3.2.1.

2.3.3 Population

See Section V-2.1.2.5.

2.3.4 Land Use-Rights

See Section V-2.1.2.4.

2.3.5 Feeds and Feeding

See Section IV-2.1.2.2.

2.3.6 Microeconomics

See Section III-2.1.2.6.