Use of goats for sustainable vegetation management in grazing lands

Final Report for LS01-119

Project Type: Research and Education
Funds awarded in 2001: $172,210.00
Projected End Date: 12/31/2004
Region: Southern
State: Oklahoma
Principal Investigator:
Arthur Goetsch
Langston University
Expand All

Project Information

Abstract:

This investigation was a collaborative study between the E (Kika) de la Garza American Institute for Goat Research and six Native American Nations using goats for vegetation management and providing for technology transfer using goat field days. We visited potential sites at the six Native American Nations and proposed experimental plans for each site and secured approval of the plans from the Native American Nations. The areas were then fenced into pastures to accommodate the experimental plans and means to provide water for goats were developed. Existing vegetation, both woody and herbaceous, was characterized and were animals taken to the field in early summer where they stayed for the growing season. Approximately 300 animals were used the first year and 200 in the second year. Animals were weighed monthly to measure live weight change and fecal samples were taken for microhistological analysis (a technique to determine what plants the animals were eating) and fecal egg counts. At the end of the season woody vegetation less than 6 ft high was measured to determine the degree of defoliation of woody species. Because two years is a short time relative to the time required to shift vegetation, we observed only the beginning of changes in vegetation. Goats did successfully defoliate most of the woody species and forbs. Poison ivy was controlled, as was greenbriar. Goats did kill many honey and black locust trees at several sites. At some sites, they killed eastern red cedar in the short amount of time. Winged sumac was readily consumed and killed at several sites. Plant vigor was noticably reduced for blackberries, dogwood and sand plum. Problems with excess accumulation of litter were solved. One revelation was that the same species were consumed to different degrees at the sites, suggesting effects of the unique soil or rainfall conditions on palatability. At one site there were noticeable differences in consumption of particular plants in different areas. Another revelation was that as most readily consumed vegetation was controlled, there was increased presence of lesser palatable species or ones growing in the cool season. Fecal egg counts demonstrated that goats used for vegetation management did not have worm problems as long as they had vegetation up off the ground. At one location, there was a significant worm problem due to goats patch grazing bermudagrass close to the ground. Field days were held at the end of both grazing periods at each site. The Native American Nations were involved in the field days, providing a meeting place, coffee and donuts and a tribal representative to give the welcome. Notices were also posted in the tribal newspapers as well as farmer meeting places in the community. Tribal council members attended the meetings at each location. Presentations included a general outline of studies and vegetation at other sites, a detailed description of the study at the local site, and general goat management principles. Then there was a tour of the site so that participants could observe the vegetation that was consumed as well as seeing and petting the goats doing the work. There was good attendance at the meetings both years and we had two request for goat field days this year. Also, producers that attended these field days have requested us to speak at local producers groups to address local goat problems. Schools at the Cherokee site and Caddo Nation sites were involved in the studies. At the Caddo site, a Science class used both GPS and a tape measure to measure the field and divide the field into research pastures. GPS data were shared with the students for analysis and seeing the usefulness of GPS for doing research and other GPS applications such as surveying. Also the data revealed the limitations of GPS and which moved beyond theory into the real world application.

Project Objectives:

1) Investigate effects of various goat management methods for vegetation rehabilitation/control in different grazing land settings in the south-central US.

2) Demonstrate and display appropriate means of vegetation management with goats in south-central US grazing lands, as well as to provide education in other related management areas.

3) Develop an information package on optimal use of goats for grazing land vegetation management to ensure long-term, sustainable, and widespread project impact.

Introduction:

Statement of Problem, Rationale, and Significance

Through common past uses of land in the south-central US, many areas now host high levels of invasive plant species with low productive potential and(or) that are not readily consumed by cattle. Responsible factors include periods of abandonment, inappropriate cultivation or grazing practices, erosion, lack or misuse of fire, exclusion of browsing animal species, high costs or concerns about safety or environmental degradation with conventional control. Management tools available to control, manage, and(or) eliminate undesirable plants in grazing lands include mechanical, chemical, and fire techniques, which are often not effective, economical, practical, or environmentally preferred. Hence, livestock producers either do not change stocking rate, accompanied by decreased digestible nutrient intake by livestock, or stocking rate is reduced in accordance with the decreased quantity of consumable forage present. Both practices allow continual increases in levels of undesirable vegetation with decreasing potential food production and farm income.

Goats offer an alternative method for vegetation management in grazing lands. Goats consume numerous invasive plants not extensively utilized by cattle or sheep and have greater capacity to tolerate antinutritional factors often found in such plants. Demand for goat meat in the US is rising rapidly, in large part because of the increasingly diverse ethnicity of the US population. Nonetheless, use of goats for sustainable vegetation control is not widespread. This is because information available in this area is largely anecdotal, with very little research-based management guidelines, and many livestock producers are simply not familiar with goats and their use in vegetation management.

A primary goal of the project was to increase appropriate employment of goats in sustainable vegetation management in grazing lands of the south-central US, with particular emphasis on Native American Nation (NAN) tribal lands or lands of tribal members. The end-result is the development and employment of safe, economical, and sustainable grazing management systems for increased farm profit, elevated food supplies, enhanced long-term land productivity, and preservation or restoration of native plant species and biodiversity. To achieve this goal and end-result, objectives were to determine optimal means of using goats for vegetation management in an array of scenarios at different NAN locations, and to display such practices and provide education in allied topic areas. The NAN collaborated because of considerable problems with invasive plants on their lands that cannot be remedied with conventional techniques.

The issue of the project has greatest applicability to small and mid-sized producers, well represented by collaborating NAN. Technology being developed and demonstrated is well suited for small family farms (limited resource producers), converse to expensive mechanical and chemical controls not proven particularly effective or requiring considerable investment in equipment, training, chemicals, or outside contracting. Vegetation management with goats can have marked impact on small farms with degraded land not presently well suited for cattle, with benefits for farmers owning land as well as to owners wishing to lease land that currently cannot be conventionally utilized because of extensive presence or encroachment of invasive plant species. Development and dissemination of technology to use goats for rehabilitation of such land and control of invasive plants will allow for new economic return from goat production and farm diversification. Furthermore, economic returns from cattle production may be made possible or increased, and land rental rates could be enhanced as well.

Literature Review

Invasive Plant Encroachment: Encroachment entails either invasion of an undesirable plant(s) where not previously present or aggregation of an existing undesirable plant(s) (Trollope et al., 1989). A plant is undesirable if, compared with other plants present, it is less acceptable to the livestock on site, less productive, less nutritious, competes aggressively with other plants and restricts their productivity, and(or) directly causes harm to grazing livestock. A consequence of encroachment by invasive plant species is the decrease in grass or forage production. There are other undesirable changes occurring as well, among which are decreased biodiversity and system robustness and loss of native plant ecosystems.

In general, encroachment results from disruption of interacting biological control mechanisms (Trollope et al., 1989). For example, large areas of grazing lands have been overgrazed, resulting in decreased grass vigor and an increased ability of many invasive plant species to compete. Failure to maintain proper soil fertility can have similar effects. Overgrazing decreases grass use of topsoil water and allows deeper water penetration, again favoring deep-rooted encroaching plants, especially brushy species. The infrequency of fire today compared with the past has promoted invasion by wood species. Without grazing and in particular if browsers are not present, grasslands are invaded first by shrubby plants, with eventual transition to woodland (El Aich and Waterhouse, 1999). Furthermore, ungrazed woodland areas and associated high fuel loads are very susceptible to intense, unnatural fires that kill large trees and sheltering wildlife, as well as posing a threat to human habitation and life (El Aich and Waterhouse, 1999).

There are few options available for vegetation management in grazing and range lands, which include fire, mechanical, chemical, and animal interventions. Fire is an economical and effective tool for controlling many types of vegetation (Powell et al., 1979); however, many woody species sprout from roots or regrow from seed. Mechanical methods of vegetation management include roller chopping, root plowing, bulldozing, chaining, or cabling and mowing. Although dramatic results from these methods are achieved quickly, most are too expensive for practical application and results are often short-lived. Chemical control of weeds has been effective in the past, but the increased expense for registering chemicals has caused the cost to rise considerably. Hence, land must be of high productive potential to recoup chemical treatment costs. In addition, chemicals can kill desirable plants, such as N-fixing forbs (Allan and Holst, 1996). Effectiveness of using animals for vegetation control depends on animal species, time of year, grazing duration, and length of paddock rest. In most cases the mode of action is repeated defoliation and reduction of carbohydrate reserves (Davis et al., 1974). Of available livestock species, goats offer greatest opportunities as biological vehicles of vegetation management because of unique dietary preferences for most invasive plant species (Taylor, 1992).

Diet Selection by Goats: Vegetation selection by goats is considerably different from that of sheep and cattle. Goats selectively harvest herbage from a wide variety of low-quality woody plants (Fraps and Cory, 1940; Davis et al., 1974; Merrill, 1975; Sidahmed et al., 1981; Bryan et al., 1987; Magadlela et al., 1995). Goats consume many of the spiny and prickly weeds that sheep do not graze (Pompay and Field, 1996). Also, goats preferentially consume reproductive stems, reducing the spread of weeds by seed. Even though mature seeds sometimes survive passage through the digestive tract of most animals and germinate, goats usually consume the seeds in an immature stage that do not survive the digestive tract. Allan and Holst (1996) observed that goats reduced the seed bank of thistles when being used for thistle control. Mayo (2000) observed a reduction in seed production when goats were used to control sericea lespedeza. Relatedly, goats are resistant to many plant toxins and anti-nutritive factors, and are thus capable of defoliating many plant species that cattle do not utilize (Hart, 2001). Hence, goats have been used to clear or control trees and brush in areas where this vegetation is competitive with grasses and forbs that are preferably grazed by sheep and cattle (Terrill and Price, 1985; Pompay and Field, 1996; Balogu et al.,1999). In addition to direct actions, goats help restore cycling of plant nutrients sequestered by woody species. For example, when shinnery oak was grazed for 3 years with goats, available N in the topsoil increased from 1.1 to 23.3 kg/ha, available P rose from 5.5 to 25.5 kg/ha, and available K increased from 133 to 348 kg/ha (Escobar et al., 1998). This can be important in that N and P are critical for the establishment of many native grasses (Rice et al., 1960). During this time, oak species cover declined from 95 to 50% and grass and forb cover increased from 5 to 50%.

Although goats are sometimes classified as browsers, the composition of their diets correlates with the botanical composition of pastures and, hence, might more appropriately be thought of as intermediate feeders. Grant et al. (1984) found goats to graze rushes that were in mixture with various grasses, but selectivity for rushes decreased as the proportion of the mixture decreased. Coblentz (1977) found that diets of feral goats on Catalina Island, California contained 90% browse, 4% forbs, and 6% grass during winter months when availability of browse was high relative to forbs and grass, but contained 8% browse, 18% forbs, and 74% grass during the summer when there was rapid growth of grasses and forbs. Other work has also shown goats to select grass during periods when grass is the major component in the mixture (Squires, 1982).

Goats for Sustainable Vegetation Management: There is a variety of invasive plant grazing land scenarios with which research/demonstration activities will take place. In addition, in all but one instance there is not one but several to many encroaching, invasive plants to be managed or controlled by goats. Although it is well known that goats can be effectively used for control of invasive plants in grazing lands, controlled research in this area is somewhat limited. Nonetheless, some available literature pertaining to use by goats of these specific invasive plants is available and will be briefly overviewed.

To address the problem of 43,000 ha of West Virginia farmland seriously infested with weeds or brush, Magadlela et al. (1995) and Dabban et al. (1997) used various controls with pasture infested with 45% brush cover, primarily blackberry. Goat grazing decreased brush cover to 15% in 1 year, with sheep requiring an additional 2 years to achieve the same reduction. Costs at that time (1986 to 1990) were $33, 262, 133, and 593/ha for treatment by goats, sheep, cutting, and herbicide, respectively.

Goats are particularly appropriate to control invasive plants such as multiflora rose and sericea lespedeza compared with other means, since their seeds remain viable in soil for many years. In this regard, Luginbuhl et al. (1999) used an Appalachian pasture not grazed for 15 years and now overgrown primarily with multiflora rose. During the experiment the pasture was grazed with one stocking rate of goats or goats plus cattle, compared with no grazing. Over the 4-year period herbaceous vegetative cover (non-brushy or woody species) was increased by 65 to 86% with goats and 65 to 80% with goats plus cattle, converse to a decrease with the control from 70 to 20%. Grass species cover increased from 16 to 63% and from 13 to 54% with goats and goats plus cattle, respectively, in contrast to an average of 10% over the period for the control treatment. Multiflora rose bushes were almost eliminated, and there were appreciable decreases in prevalence of poison ivy and blackberry.

Eastern red cedar is an invasive plant of considerable importance in a large part of the US, and is present in varying amounts at most of the research/demonstration sites. High invasion levels severely limit forage growth and livestock production, as well as wildlife carrying capacity (Taylor et al., 1997). In the south-central US, the much greater prevalence of cedar now than before intervention by man is in part due to overgrazing and decreased frequency of fire (Scifres, 1980). Even though fire is an effective method of control of cedar encroachment, many factors limit potential for use and efficacy of fire, among which are close proximity to urban areas, inadequate fuel load to eliminate large trees, and insufficient desire, equipment, expertise, etc. to carry out effective prescribed burning programs.

Goats can contribute to rehabilitation of grazing land infested with junipers (Fuhlendorf et al.,1997), and can similarly be used in sustainable, long-term management. Goats ingest eastern red cedar (Powell et al., 1979; Bauni, 1993; Thompson and Swartz, 1991) and, in fact, it is among the most readily consumed junipers (S. Fuhlendorf and M. Moseley, personal communications). In a Cross Timber rangeland site, Bauni (1993) noted influences of summer stocking rate of Angora goats on cedar consumption. Cedar utilization was 17% in lightly stocked pastures and 45% when heavily stocked. Furthermore, in heavily stocked pastures browsing height was greater and there was a more gradual decline in utilization with increasing height compared with the distinct browsing line of 0.75 m for the light stocking rate. However, cedar consumption is typically greater in fall and winter than summer, particularly with low availability of palatable browse species at those times (Buani, 1993). Fuhlendorf et al. (1997) noted that without browsing, prevalence of Ashe juniper increased. Furthermore, browsers stocked at a rate adequate for a browse line limited further encroachment of woody plants and lessened impact on grass growth of large junipers.

Various oaks, such as post and blackjack, have invaded large areas of Oklahoma and Texas, and goats are effective in their control (Davis et al., 1974; Merrill and Taylor, 1976; Wiedemann et al., 1980). Bauni (1993) observed that at least 4 goats per acre were required to defoliate the oak species because of the abundance of more palatable browse. Briggs and Beal (1940) advocated the use of 2 to 5 Angora goats per acre to clear the oak-dominated brush. Darrow and McCully (1959) recommended use of goats to prevent resprouting of blackjack-post oak after it had been mechanically or chemically controlled. Magee (1957) found that on ranches not using goats to control oak sprouts, regrowth had become a serious problem within 5 years of clearing.

In the preceding paragraphs, consumption and control by goats of most invasive plants present at the research/demonstration sites of this project have been noted. In addition, from observations of goat grazing in mid-summer at a study site hosting a wide array of undesirable vegetation, Pinkerton et al. (1992) classed plants into three preference categories: priority choice, preferred, and grazed but to a lesser extent. Bauni (1993) presented a similar classification in another study. Priority choices of Pinkerton et al. (1992) were sumac, winged elm, green briar, and poison ivy; preferred plants included rusty blackhaw, tree sparkleberry, hickory, eastern red cedar, western ragweed, red bud, pepper vine, rough sunflower, hawthorne, multiflora rose, post oak, tall thistle, and butterfly pea; and plants grazed but to a lesser extent were chittamwood, croton, willow oak, swamp chestnut oak, Illinois bundleflower, annual broomweed, rattan vine, beauty berry, flowering dogwood, osage orange, poison oak, witchgrass, slender lespedeza, shumard oak, common sunflower, kobe lespedeza, goldenrod, johnsongrass, hackberry, buckbrush, white snake root, giant ragweed, cherry, and persimmon.

Cooperators

Click linked name(s) to expand
  • Terry Gipson
  • Steven Hart
  • Roger Merkel

Research

Materials and methods:

Visits were made to each Native American Nation to detail collaboration and select a site suitable for study. Six research/demonstration activities were planned in cooperation with Caddo, Cherokee, Choctaw, Greater Seminole, Osage and Sac, and Fox Native American Nations. Characteristics and specific vegetation management objectives for each site are detailed later. Grazing treatment chosen for each site were thought most appropriate based on existing vegetation and vegetation management goals.

Pastures were fenced beginning in mid November 2001 and completed in mid-May 2002. Net wire (1047 Sheep and Goat wire) was used for perimeter fences and was used for the interior if only one interior division fence was needed. However, the study located on the tribal member’s farm (Choctaw Nation) utilized all electric fence so the area could be completely cleared after the study. On some locations, boundaries had to be located prior to putting up fence. The Native American Nations appreciated us fencing their boundaries, a benefit to them that will last beyond the study. Corner braces were constructed and steel posts driven for the net wire fence. At three locations, electric fence was used. Landscape timbers were used for corners and tread-in posts used to hold 4 strands of electric fence. Solar powered shockers were used. Fence lines were usually wead-eated early in the spring and sprayed lightly with Roundup and then weed-eated later in the year to maintain the electric fence. Water was provided in a variety of ways according to resources available. At one location, a pond was divided in two between the pens for water. At three locations water was piped from existing water supply using flexible polyethylene tubing and float valves. At one location, water was hauled weekly by the caretakers using barrels in a truck. At one location, water was pumped out of a creek into a gravity-flow water tank. At least one guard dog was used at each location. At most locations, one guard dog would slide under the fence between the pastures and effectively guard two pastures. Two locations had three or more pastures and more dogs were utilized. We had no predator kills during this study and goats only got out of the pasture once during this study - a record which we are proud of. At the Cherokee site where there were vegetation control treatments in addition to grazing, the strips were marked with steel T posts at 100-ft intervals to facilitate application and sampling of the treatments.

Herbaceous biomass was measured by hand clipping six .25 m2 quadrats to ground level at the beginning of the study and each time animals were weighed. Samples were dried and weight recorded. Areas were randomly selected and were taken in areas not dominated by tree canopy where appropriate. Dried samples were subjected to laboratory analyses. In addition, leaves of the dominant woody species of each location were sampled at the same time as the herbaceous vegetation. Leaves were obtained from several trees in each treatment and composited for drying and subsequent laboratory analyses. Woody and herbaceous samples were analyzed for crude protein by Kjeldahl procedure (AOAC, 1990), and neutral detergent fiber and in vitro dry matter digestibility were determined by filter bag techniques (ANKOM Technology Corp., Fairport, NY).

Plant species composition and woody species canopy were quantified by establishment of permanent line transects. Five permanent line transects were established in each pasture and control area on each study. At one location, which had 45% cover in each pasture by forest, there were three line transects in the forest area and three in the pasture area. At the Cherokee site, two line transects were used in the non-pasture treatments due to time constraints on reading the transects. The transects were blocked by vegetation type in each pasture to increase the statistical power of the observations. Transects were marked by white plastic stakes at the south or east ends and orange plastic stakes on the other end. Permanent digital pictures were taken from the white end of the transect for documentation. A person stood at the 16-foot mark of the transect with a post marked at 1-foot intervals for comparison purposes. The photographer stood 10 feet from the white transect stake and took a picture with the center of the picture aligned on the length of the transect tape. A dry erase board was used to identify the transect, location, and date at the lower center of the picture. Location of the transects were identified using GPS readings and flagging material.

One-hundred-foot line transects were used and a pointer used at each foot and dropped to the ground. What the pointer hit (ground, litter, rock or direct hit on a plant) was recorded. The plant species under the pointer or closest to the pointer on the side of the tape toward the orange stake was identified and recorded. Each plants species was photographed with a Polaroid camera, photographs numbered and placed in an album and plant samples taken and put in a plant press for later identification. In this way, we documented the plant and had a voucher specimen for the species on location. Species were identified with help from several Botanists, and some were keyed out and/or compared to species at the OSU Herbarium or Northeast Oklahoma University herbarium. Taxonomy nomenclature used was that on the USDA site (USDA NRCS Plants database, //plants.usda.gov). In each pasture, plant species were determined at 500 points which were on five 100-ft transects. Transects were read in the late spring just prior to putting animals on pasture. Cool season species had seed heads and warm season species were significantly large enough to enable identification of both types of plants.

Woody species were quantified on the permanent line transects by measuring canopy cover of each woody plant. Canopy cover was divided into a lower strata cover available to goats (6 ft high or less) and an upper strata. Canopy cover was read at the same time species were read in the late spring. In addition, canopy cover was read in the fall following animal removal to quantify removal of lower canopy by goats. Biomass species composition was determined by the dried weigh rank procedure (Gillen and Smith, 1986). It was necessary to establish permanent .25 m2 quadrats so measurements could be made at the same location each year. Fifty quadrats were established in each pasture. Random quadrats were established and marked with 20 d nails in two corners of the quadrat. One nail had a piece of vinyl on which was written the quadrat number. Permanent stakes using either PVC or metal bars were placed 1 m north of the quadrats to facilitate their location. Locations of permanent quadrats were recorded using GPS. A metal detector was used to relocate the nails for correct placement of the quadrat. The major three species contributing to above ground biomass in each quadrat was recorded and data calculated as per the reference. Soil samples were taken at the quadrats when the quadrats were evaluated. Soil samples were obtained from the top 5 cm of soil since this is where most soil nutrients are located and the nutrients would not have to be leached very far to be incorporated into this upper layer. Soil samples were composited from the fifty locations in each pasture and submitted for a basic soil analysis (pH, N, P, K, organic matter and soil buffering) at the Oklahoma State University Soils Laboratory).

Goats were weighed prior to putting on the pastures and at approximately monthly intervals and when removed. Fecal samples were taken for fecal egg counts and microhistology. Fecal egg counts were done by McMaster’s procedure (Whitlock, 1948). Components of an animals diet can be determined by microhistological analysis of the feces. A trained person can determine the species of undigested plant residues by characteristics of shape, crystals, layers or hairs on the cell wall. In addition, the percentage of each species can be determined (Holechek and Gross, 1982). This technique has been validated for cattle diets (Sparks and Maleechek, 1968). Dried fecal samples were submitted to a commercial company ( J. Rentfleish Microcomposition Laboratory) for microhistological analysis.

Site Specific Procedures

The Osage site was located at the headquarters village for one clan of the Osage Nation. The area was 40% upland and 60% bottomland. The bottomland area was quite fertile and had been used as a communal garden up to the 60's when it was abandoned. The area was overgrown with dogwood, sumac, blackjack oak, red cedar and Osage orange. The woody species so dominated the area that less than 20% of the area supported significant grass. Bromus species grew under the trees during the cool season and further choked warm season grasses. The goal for this site was to return the site to be a communal garden again growing the vegetables that they formerly grew, including Indian corn. Mechanical treatment is cost prohibitive and herbicides are not acceptable. There were two goat treatments applied to this location, One was a first year stocking rate of 4 head per acre and the other was 8 head per acre the first year. Both pastures were stocked at the same rate (3.6 hd/acre) the second year. An ungrazed pasture was used as the control. This was to compare intense defoliation as compared to a more gradual control method because animal productivity which provide returns to the enterprise would be expected to be greater with the lower stocking rate. The question to be answered is how long does it take for the lower stocking rate (and expected greater productivity and returns) to achieve the same degree of vegetation control as the higher stocking rate.

The Seminole site was virgin prairie, never plowed and had the greatest biodiversity of any site. It had two areas spoiled by salt water spills. There were two problems to be addressed at this location: First, the area had not been grazed in years and the accumulated litter was smothering the native species grasses, reducing their vigor. The second problem was that woody species, especially winged Sumac and winged elm and Eastern red cedar had severely encroached the area. This is because the area had not been burned in many years like happened before Oklahoma was settled. These natural fires provided control of the woody species. The goal of this study was to reduce the ground cover of woody species to a level consistent with the natural grassland savanna and to alleviate the self smothering by accumulated litter. Goats are ideal for this task since they have a strong preference for woody species and woody species cannot survive repeated defoliation. Also, the goats would trample the previous season’s litter, reducing the smothering effect and would graze enough grass to reduce litter building to a self-smothering level. Two goat treatments were applied to this area. One had a stocking rate of 4 goats per acre and the other had a heavy stocking rate of 8 goats per acre the first year. The stocking rate was the same (3.6) for the second year. This would enable us to see the benefit of aggressive browsing the first year as compared to gradual control.

The site at Cherokee was a native range site that had been unused for several years and had been mowed yearly for several years to control the vegetation. The major vegetation problems were blackberries, buckbrush, winged elm, sumac and eastern red cedar. These species strongly dominated one pasture and were becoming dominant species in the pasture on the other replication. There was a scattering of forbs that needed to be controlled. At this site, the collaborators were interested in comparing the use of goats for vegetation management with herbicides, the existing practice of mowing and doing nothing (control). Therefore, treatments were grazing with goats, spraying with herbicide (one pint of Grazon P+D and one pint of Remedy, sprayed on in the middle of the growing season with 20 gallons of water), mowing each summer and control. The grazed areas were 5 acres and the other treatments and control were 2 acre strips.

The Caddo site had not been grazed in a number of years. It had been an improved pasture of lovegrass, but the non-use had enabled thickets of smooth sumac, sandplum and black locust to become established. Sumac covered 60% of the area and sand plum covered 15% of the area. Under the canopies, most of the lovegrass had been choked out due to competition for sunlight. The lovegrass in open areas as well as that still struggling under the canopy of woody species were being self smothered by the accumulated grass residues from previous years growth. This self smothering had reduced the vigor of the grass and put it at a competitive disadvantage to the sumac. We wanted to use goats to control the sumac and sand plum and break up the accumulated grass residues over the plants. We thought that since there was substantial grass in the open areas, that it may be useful to use sheep in combination with goats to better utilize the grass residues. One pasture had goats alone (6 goats/acre) and the other pasture had equal numbers of sheep and goats (3 sheep and 3 goats/acre). The second year, 5 goats/acre were used for the goats alone and 2.5 goats and 2.5 sheep/acre were used.

The Choctaw site was on a collaborators farm who raised cattle. The area that we were given to work on consisted of 45% mature forest and 55% bermuda/ryegrass improved pasture. We wanted to utilize goats for clearing the forest understory. Therefore, treatments were goats alone, goats and cattle and cattle alone. The forest understory included greenbriar American beautyberry, virginia creeper and winged elm. The overstory consisted of red oak, post oak, black oak, mockernut hickory and winged elm. The treatments consisted of goats alone (4 hd/acre), goats with cattle (550 lb. stocker animals, .5 hd/acre, goats 4 hd/acre) and cattle alone (.87 hd/acre). For the second year, 3 hd of goats per acre was used on both goat plots, and stocker steers were stocked as in the previous year.

The Sac and Fox site was a sandy hillside that had been abandoned for a number of years and was mostly mature forest, except for some open areas created by installation of two pipelines across the area and an open area towards one corner and one naturally open area. The goal was to make the area suitable for recreational use by removal of most of the understory. The understory consisted of buckbrush, red cedar and poison ivy which has dense in localized areas. Since this was a relatively large area, we had opportunity to apply more treatments. We had a control, two set stocking rates (3 or 6 goats/acre) and a rotational grazing area with four paddocks (12 goats per acre. Stocking rate was the same for the second year.

Statistics

Transect data for woody species was analyzed by the general linear models procedures of SAS (SAS, 1999). The difference in ground cover for each woody species in each canopy height between the start and the end of the study was analyzed. In this way, each transect served as its control. The statistical model used the effects of treatment and transects were replicates. Transect data for herbaceous species (hits/100 ft transect) were analyzed by the mixed model of SAS (SAS, 1999). The main effects were treatment and replication was over years with treatment by year interaction. Species biomass was not analyzed statistically since there is one value for each pasture and there were not sufficient observations over time for sufficient degrees of freedom for analysis. Total biomass, in vitro digestion concentrations of crude protein and neutral detergent fiber were analyzed with factors of treatment and time and each clipped sample represented an observation.

Animal performance was analyzed separately for each location and each year. This was because several genotypes were used the first year and only two similar genotypes were used the second year. The mixed model procedure of SAS (SAS, 1999) was used with a statistical model including such effects as sex, genotype and monthly weigh period as a repeated measure. Fecal egg counts were log transformed prior to statistical analysis. The data were analyzed as a split plot in time with treatment and year being in the main plot and month of sampling in the sub-plot.

Research results and discussion:

Vegetation

Vegetation was quantified as described in the procedures. Two years of grazing is inadequate to effect perennial species because root carbohydrate stores will support the plant under severe defoliation for three to five years. Magadlela et al. (1995) observed that goats reduced brush cover significantly in only one year, but 5 years was required to obtain control. An exception would be where goats debarked some species which effectively kills them . Annual plants may be affected since grazing often prevents seed production, but many plants have a seed bank. A confounding factor is climate, since climatic conditions may be more favorable in one year for seed to sprout and become established than another year, regardless of depletion of the seed bank. We have continued to conduct the studies at three of the locations for the third year with cooperators who wanted to continue the study. We are hoping to get funding for three more years in a grant proposal being submitted. We should be able to see some changes in perennial species due to exhaustion of root reserves by defoliation and in annual species due to reduction in seed bank. The trends that have been started by the first two years of the study will be more evident. We have visually observed reductions in plant vigor, but the methods for quantitation plant cover do not quantitate vigor. However, in the long term, reduced plant vigor will cause death of the plant and reduction in its plant cover.

Caddo: Honey locust canopy cover was reduced, but the difference was not statistically significant (Table 1). We observed that all honey locust trees except two were killed after the first year by goats stripping the bark on them, but most of those killed were not on our transects. Smooth sumac cover actually increased during the study even with goats on the pastures. The goats refused to consume smooth sumac at this location even when there was nothing else to eat. However, goats consumed smooth sumac at other locations and debarked some and killed the trees. We have observed differences in palatability of some species across locations on these studies and even within pastures where they leave a clump of highly preferred species untouched while grazing the same species all around it and grazing other herbaceous species out of the clump. Perhaps the smooth sumac could have benefitted from some pretreatment such as mowing or rollerchopping which has been used to improve the palatability of some species (Davis, 1975; Allen, 1999 and Taylor, 1992 ). Du Toit (1972) observed that goats were good at controlling sprout regrowth following chopping of shrubs.

Buckbrush increased dramatically in the goat pasture in the area covered with sand plum. Sand plum only showed up on three transects, but covered only about 20% of the area and therefore the number of observations was inadequate for statistical analysis. Where sand plum was reduced by goat grazing (it is a highly preferred species) buckbrush increased due to decreasing competition. In addition, buckbrush patches expanded which I think may be attributed to soil disturbance by goat hooves causing buckbrush seeds to sprout. We have observed control of buckbrush at other locations at goats and feel that this is a short term shift in vegetation due to competing species having their vigor reduced by preferential browsing of goats. Buckbrush is generally a moderately or less preferred species (Bauni, 1993). Even though there were inadequate statistics on the sandplum, visually, the cover was not reduced, however the vigor was reduced. The second year, the sand plum in the control area produced fruit but that in the grazed area did not set fruit. Also, before grazing the first year the sand plum was very difficult to walk through, but on the latter two measurement periods, it was progressively easier to walk through and the plants were shorter. Sand plum is highly preferred by goats and was usually totally defoliated in the first six weeks of grazing season. Luginbuhl et al. (2000) observed that measuring the cover of rose did not fully reflect the control that goats were providing on rose species. Ground cover was only slightly reduced, but the number of live canes was reduced by 59%. Black locust which was not intersected by transect since it was only present in one pasture. The black locust was over 15 ft tall and had a very dense canopy. Goats killed over 50% of the black locust trees by debarking them and the trees were falling over. The canopy was unchanged due to multiple strata in the canopy, but in two years, all black locust trees would have been killed. Luginbuhl et al. (2000) observed that goats controlled black locust completely over a 4 year period. Although upper strata of honey locust was recorded to increase, there were only two viable trees left after two years of grazing due to debarking the trees. It appears that the upper strata canopy expanded on these trees. Thee was limited upper strata of smooth sumac which was expanding due to an aversion by the goats for whatever reason.

The percent of most forb species was increased, especially on the grazed areas (Table 2). This is because of the prior accumulation of lovegrass biomass which smothered other species and reduced its own vigor. Goats were effective at breaking down the litter due to disturbance by their hooves and by grazing on the lovegrass, preventing accumulation of litter. There was a dramatic increase in especially cheatgrass, both bromus japonicus and bromus tectorium due to this reduction in litter and disturbance by hooves. However, Severson and Debano, (1991) attributed the reduction in litter on their study to consumption of vegetation, preventing it from accumulating as litter. If this study were continued, it would be desirable to put animals out in mid-march to consume the cheatgrass and prevent it from setting seed and replenishing the seed bank. The seedbank of this species deteriorates rapidly and control should be able to be achieved with two years grazing as shown with sheep (Daubenmire, 1940; Finnerty and Klingman, 1962)

The biomass composition could not be statistically analyzed since there is only one data point per field which was a summary of 50 observations at permanent quadrats (Table 13). Biomass composition was not dramatically affected, but followed some trends similar to species composition. Lovegrass increased slightly in the control treatment, as did horse nettle and brome at the expense of common yarrow and other forbs . Common yarrow biomass increased as did horse nettle on the goat treatment. Yarrow was not consumed and carolina horse nettle is toxic. Even though cheatgrass species composition increased, there was a slight reduction in biomass. Brome increased considerably in the cospecies grazed area. This is likely due to soil disturbance by hooves and a much younger stand of sumac in this area which shaded the ground less in the winter and early spring. There was a dramatic decrease in star-thistle biomass in the cospecies plots and goat plots. Canadian horse weed was controlled by grazing either with goats or goats and sheep.

The neutral detergent fiber content of forage samples was significantly lower for the goat treatment for unknown reasons (Table 19). Crude protein content was lower for the cospecies grazed treatment than for the other two treatments. Perhaps the sheep grazed the high protein containing species and left only those low in protein. Yield was reduced for the cospecies treatment as compared to the two other treatments, probably due to the sheep consuming more herbaceous species.

To summarize the impact of goats and cospecies on vegetation at the Caddo site, we were unable to remediate the major problem of smooth sumac spreading because the goats did not readily consume it at this location. The goats did ameliorate the problem with accumulated litter smothering native grasses since the litter was broken up and distributed by hoof disturbance and the grazing prevented additional litter accumulation. Progress was made towards controlling the sand plum thicket. Research could not be continued in this area because of a change in plans by the Caddo Nation for use of this land. However, the most important research that could be conducted is on pretreatment to induce goats to eat the smooth sumac, whether by mowing, roller chopping or even changing soil fertility and/or pH.

Seminole: Browsing by goats changed vegetation little in two years, but there were a few good changes and trends that will likely continue as browsing continues. Surprisingly, woody plants were reduced in the control plot. The plot was on the edge of the study plots and there were significant numbers of deer in the area which may have caused some of the reductions of woody species in the control plot which was unexpected. While there was a reduction of hawthorn in the control plot, it was almost completely eradicated from the grazed plots (Table 2). Canopy cover of persimmon followed a similar pattern, a reduction in the control plot in which there was little initially and a reduction in the grazed plots, more in the high stocking rate than the low stocking rate, but the effect was not significant. Honeysuckle was only present on two of the treatments and increased over time in the control and decreased over time in the browsed plot. Visual observation after the third year of grazing indicated that honeysuckle was very sparse. Mimosa was increasing for unknown reasons. Luginbuhl et al. (1999) observed little control of honeysuckle by goats. There was a dramatic increase in winged sumac on the control plots whereas there was a dramatic decrease on the grazed plots, more for the high stocking density than the low stocking density. Goats destroyed a large thicket of winged sumac the first year, debarking it as well as eating leaves. They consumed all the resprouts from the plants the second year and the stems were falling over. There was little change in blackberry ground cover, but the goats defoliated plants completely both years since it is high on their preference list. Visually, the vigor was reduced although ground cover was not. Luginbuhl et al. (1999) observed a gradual control of blackberries with goats or cattle and goats while Mills and Bryan (1983) observed that goats and sheep controlled blackberries. Buckbrush ground cover was decreased in both control and high stocking density and little change in the low density. Although buckbrush is not highly preferred by goats the lower strata was completely defoliated when goats were removed at the end of the first year (data not shown). Poison ivy was reduced on the high stocking density plots with a slight increase on the moderate stocking density pasture. Goats love poison ivy and defoliate it early in the season. Luginbuhl et al. (2000) also observed significant control of poison ivy by goats. Winged elm which was well defoliated by goats decreased significantly on the control plot with some effect on the moderate plot. In the supper strata, there were increases in ground cover of common persimmon and winged elm. Goats can only affect the upper strata by debarking these mature trees or natural death of the tree. The goats will prevent the juvenile brush size trees from reaching to maturity and in the long term (10-20 years) will control the woody upper story. There was a significant number of cedars in the end of one pasture and only intersected one transect. Cedars were progressively browsed to a greater extent each year, but no tree mortalities were observed as browsing was severe on only a few plants. Fuhlendorf et al. (1997, working with a similar juniper species, observed that it was necessary to have a high stock density to have significant browsing, and control juniper, but that more palatable species would be overutilized at that stocking density. We have observed that goats utilize cedars more extensively during the winter, probably due to the seasonal decrease in volatile oil content (Riddle et al., 1996). In our experience, if goats grazed these areas in the winter, more successful control of cedar would be obtained.

Composition of herbaceous species was changed little by goats over this two year period (Table 8). There were significant changes due to treatment which were usually due to the initial species composition of treatment plots being different. There were some significant changes over time since some species increase in response to environmental conditions, especially rain. Basically, it takes more than two years of grazing to affect perennial plants due to their root carbohydrate reserves. A significant interaction would mean that the species changed over time due to treatments. The only species that did was western ragweed. It was reduced on the high stocking rate where animals were forced to defoliate the plant and was basically unchanged on the moderate grazing treatment. However, it was reduced on the control grazing treatment possibly due to deer browsing. Several species were different from the others in one year, including Carex, jointtail grass, little bluestem and dropseed. Only jointtail grass appeared to be consistently decreasing. Broomsedge bluestem appeared to increase under moderate grazing and be stable on a high grazing density. Broomsedge bluestem has been observed to decrease over years on another goat grazing study. Little bluestem appeared to increase at the high stocking density and not be affected by the other two treatments.

Ashy sunflower, little bluestem and other forbs increased as a proportion of the biomass on the control plot at the expense of other grasses (Table 14). On the high grazing density treatment, bluestems, dropseeds and other forbs increased at the expense of other grasses and sedge. On the moderate grazing treatment, bluestems, increased at the expense of sericea lespedeza and other grasses. The goats did graze the sericea heavily during the summer and have been shown to provide control of sericea lespedeza (Mayo, 2000). Dry matter yield did not differ significantly for plots, although it decreased across the grazing season (Table 20). The dry matter yield for the high grazing density was numerically lower due to the greater grazing pressure. Forage NDF content was similar for all treatments, but clipped material from the high stocking density plots had a higher level of protein, most likely due to most of the mature forage being grazed off.

Osage: Several woody species were affected by grazing at the Osage site (Table 3). Redbud was quite reduced on the high stocking density, and honey locust reduced on the moderate density treatment since the goats highly preferred the honey locust and stripped the bark. Dogwood was significantly decreased by goat grazing and but increased in the control treatment. Honeysuckle increased on the moderate grazing treatment and was basically unchanged on the other two treatments. Luginbuhl et al. (1999) observed little control of honeysuckle by goats in their study over a 4 year period. Winged sumac was significantly decreased by grazing, especially on the high stocking density treatment. Winged sumac increased dramatically on the control treatment. There was a significant increase in upper strata Virginia creeper on the moderate grazing study. It is hard for goats to reach the upper strata since Virginia creeper climbs trees and goats have not been observed to chew on the stems.

Herbaceous species composition was affected little by goats (Table 9). This is because the area was dominated by a tall overstory although there were a few open areas. Most of the herbaceous species would be affected little until the overstory is removed. There was an increase in Bromus spp. due to removal of the winged sumac canopy as discussed previously for. It will take several years before perennial species revegetate these areas. If further grazing work is done, some goats should be put in the pasture in March to graze the Bromus and prevent it from producing seed. Two years of early grazing by goats should bring the brome under control so that it will not compete with natural establishment of perennial grass seedlings. Work with sheep have shown that when cheatgrass is prevented from producing seed, cheatgrass populations drastically declined (Daubenmire, 1940; Finnerty and Klingman, 1962). As goats continue grazing this area, it may be beneficial to work with soil amendments to promote reestablishment of native grass species since Dabaan et al. (1997) observed a benefit to soil amendments.

There was little change in biomass species composition in the control treatment (Table 15). The increased biomass of Carex reflects the increased proportion of species accounted for by this plant as discussed above. The increased biomass of Scribners panic also reflects the numerical increase in the species. White avens also increased in both grazing plots for unknown reasons. It was surprising that there was a reduction in big bluestem biomass under the moderate grazing treatment. Biomass dry matter yield was numerically lower for the grazed treatments than control with a significant effect due to time; the yield decreased across the season as the animals grazed (Table 21). Neutral detergent fiber was lower for the high stocking density as compared to the other treatments likely because of the forage being less mature due to higher grazing pressure and fiber increased as the season progressed. Protein was highest for the control treatment, but there was very little dry matter to clip because of virtually complete closure of the overstory. The high stocking density had greater protein than the moderate stocking density, again, probably because of younger growth, consistent with the lower fiber levels observed on this treatment.

Choctaw: Lower strata at Choctaw consisted of basically three species (Table 4) which were not affected by grazing treatment. The upper strata was a mature forest composed of oak, elm and hickory species. There was no chance of goats affecting these large trees, most of which goats do not debark. There was some natural mortality occurring and goats would prevent the recruitment of seedlings to replace those the mature trees that died and in the long term could affect the upper strata.

There were very few effects of grazing treatment on herbaceous species at the Choctaw site (Table 10). Half the pastures were dense mature forest and the other half that was open was a bermudagrass based pasture which had been encroached by bahiagrass. There were year to year variations in these species, likely due to environmental factors, especially rainfall pattern. Sheep sorrell and flat sedge were the next most common species. Only open flower rosette grass, poor joe, common oxalis, witchgrass and bahiagrass were different for years, again, most likely due to environmental influences. Only bahiagrass was affected differently by treatment over years. Bahiagrass increased on treatments which were grazed with goats and quite dramatically so. Apparently, they must have avoided consuming the grass and it proliferated. They did consume bermudagrass and quite close to the ground in some spots causing severe patch grazing. This may have allowed the bahiagrass to proliferate.

Bermudagrass increased on the goats and cattle pastures and was unchanged for the cattle and goats (Table 16). Bahiagrass increased on pastures grazed by goats, indicating a preference against that species. Also, there was a decrease in other forbs, consistent with a goats dietary preference for forbs. Bromus spp. decreased on the treatments which contained goats and was not present in significant amounts on the cattle grazing study. This is likely a consequence of goats being placed on this study earlier than other studies and the brome was vegetative and consumed by goats, preventing seeding of the brome. Other grasses increased dramatically on the cattle and goats treatment, consistent with a low preference for grass as compared to forbs and woody species. The yield of biomass was similar for all treatments (Table 22) and was not affected by time because this location had good rainfall distribution over the season and the forage outgrew the goats and had to be baled during the grazing season. Fiber levels were quite high in the forage reflecting the dominance of bermudagrass and bahiagrass and their maturity. Protein level was similar for all treatments and decreased as the season progressed.

Cherokee: There was no upper strata of woody plants at Cherokee, a consequence of previous history of annual mowing, therefore, all the woody plants could be accessed by goats (Table 5). Spraying was very effective at taking out woody species, sumac, blackberry, sassafras, and rose without great effect on buckbrush, persimmon, or greenbrier. Goats reduced dogwood, rose, flameleaf sumac and blackberry. Buckbrush dramatically increased on the grazed treatment. This was due to the goats preferentially grazing blackberries opening up the canopy so the buckbrush under the canopy could proliferate. Vere (1979) observed that not only were goats effective at controlling blackberries, they could return a profit with reasonable care. Luginbuhl et al. (1999) and Escobar et al. (1998) observed that goats were effective at controlling blackberries and as blackberries were controlled, grass increased. Luginbuhl et al. (2000) observed that goats were effective at reducing rose, but complete control had not been achieved in four years. We expect goats to control the buckbrush in the future as the cumulative effects of defoliation cause plant mortality. Virtually all woody species were defoliated by the end of the season which will result in their loss of vigor and eventual mortality. Mowing reduced sumac and buckbrush cover, but otherwise provided little control of woody species.

Herbaceous species were not affected much by goat grazing (Table 11). The only significant treatment differences were for Bromus, hogwort, beaked panic grass and Korean clover. These treatment effects were due to preexisting vegetation before the study. There were several year to year differences, reflecting different environments across the years. There was no true treatment effect over years due to any treatment. Biomass of Bromus was reduced between years (Table 17) regardless of treatment and therefore, likely an environmental effect. Scribners panic was reduced on all treatments except for the grazed pastures where it increased. Sericea lespedeza was not reduced over the two years, but on other studies, goats have been very effective at reducing sericea lespedeza after three years (Mayo, 2000). Velvet panic increased dramatically in all treatments but the grazed treatment. Little bluestem was increased by grazing or spraying. Biomass was not significantly different for treatments although visually, the sprayed plots had a great increase in grass production (Table 23). This may be difficult to observe in that we only harvested two quadrats from each sprayed strip. Fiber levels were higher for the grazed and sprayed treatments. The sprayed treatments appeared to have more biomass which was more mature. The mowed treatment, because of mowing was more vegetative. It is unknown why the control had a lower level of fiber. Protein level was highest for the mowed treatment, consistent with its lowest fiber content. The sprayed treatment had the lowest protein level, consistent with the high fiber level. The control may have had a low protein level and low fiber level due to species composition rather than treatment effect per se.

Sac and Fox. Goats did provide significant control of lower strata woody vegetation at Sac and Fox (Table 6). In fact, most woody species were decreased by grazing with goats. Buckbrush, post oak, dogwood, hackberry and sumac were substantially reduced by grazing. It appeared that the goats had a higher preference for buckbrush at this site than at other sites. Poison ivy was basically eradicated by goat grazing, an important factor in that this land was slated for recreational use. Luginbuhl et al. (1999) observed that goats were quite effective at controlling poison ivy. Many cedars less than 8 ft tall were killed by debarking. This was surprising in that goats seldom debark cedar during the summer months. Most of the cedar not killed had an upper strata which the goats could not defoliate and they did not appear to be willing to crawl under the dense canopy to debark the trees at this time, but they may would in the long term. There was a reduction in the upper strata only for hickory, hackberry, hawthorn, blackjack oak, post oak, gum bully and slippery elm. There appears to be less reduction of the upper strata on the rotational grazing system, which may be a consequence of the rest period.

There were very few significant treatment effects on herbaceous species composition (Table 12). There was more slender three seed mercury on the control and moderate treatments which was a location effect. There was more sedge on some plots, an effect of location and sedge was different for years as observed for other locations, probably reflecting a difference in rainfall. Grazing definitely reduced cheat, but cheat was also reduced on the control treatment. Biomass of sedge increased on the control treatment, but remained steady or decreased on the grazed treatments (Table 18). Other forbs were reduced on the grazing treatments, consistent with the goats preference for forbs, but forbs also decreased on the control. There were no other consistent differences in species biomass for grazing treatments. Biomass yield was greatest for the control and was reduced by grazing (Table 24). The rotation grazing had the next highest drmatter accumulation due to the rest period allowing dry matter to accumulate. The continuously grazed plots had similar levels of dry matter accumulation. Neutral detergent fiber was similar for all grazing treatments and increased over time. Protein content was greater for the grazed treatments since they were kept in a more vegetative state by defoliation.

Animal performance

Goats at the Caddo site gained the least weight of any site due to refusal to eat smooth sumac and there being a shortage of grass (Table 26). Since grass was short, digestibility of the diet was likely low since Sidahmed et al. (1981) observed that diet digestibility increased as the proportion of grass in the diet increased. There was a period by pasture treatment interaction (2002) in that goats gained more weight the first period, lost more weight the second period and lost a similar amount to the other treatment in the third period despite removing of sheep at the end of period 2. Animals gained less weight as the study progressed. There was considerable browse biomass available for consumption (smooth sumac), but the goats did not readily consume it in contrast to other sites where it was a minor species and was consumed. Gains dropped off dramatically after the first period because there was very little herbage besides sumac to consume. Escobar et al. (1998) observed similar low rates of gain for goats grazing a mature stand of shinnery oak where there was little grasses or forbs present. There was sex by period interaction in that wethers gained more weight in period 1 and 3 than doelings, but gained less weight in period 2. In general, wethers, because they are larger and have no nutrient requirements for reproduction, generally gain faster than females. Breed was not a significant factor in gains due to a severe nutritional limitation . In 2003, there as little difference in animals grazing either pasture although there was an interaction between pasture and period in that animals on the goat pasture gained more weight the first two periods, but lost more weight the third period. There was a period effect in that goats gained progressively less weight with each period due to increased scarcity of herbage other than smooth sumac. There was no difference due to the sex or interactions of sex. Goats were pressured to consume smooth sumac and refused to, resulting in significant weight losses. This species was readily consumed at other locations, but it dominated this location and we were unable to achieve satisfactory control. There are anecdotal observations of deer having different diet preferences for the same plant species in different regions. This has been attributed to the effect of the environment on the plants.

Body weight gains at Seminole were similar for both stocking rates in 2002, but animals had to be removed from the high stocking rate earlier because of limited forage availability (Table 27). Rates of gain were similar to that observed by Hart (2000) for goats grazing native pastures. Goats gained more (P<.05) in the first two periods than the third period, likely due to the more nutritious herbage being harvested first and the less desirable being what was available in the last period. Doelings gained more than wethers overall, but there was a breed by period interaction in that doelings did not consistently gain more than wethers. The great difference between genders in period 3 is difficult to explain, but females were represented only in the two Boer crosses. The 7/8 Boer goats gained the most weight and the Boer x Spanish and Spanish were similar whereas the Angora goats gained the least weight. Angoras are at a competitive disadvantage as compared to the other breeds in that they are smaller can cannot reach as high as the other breeds for browse. In addition, they have an extra nutrient requirement for mohair production. Only one breed (Boer x Spanish.) was used in the second year at the same stocking rate. Goats on the pasture stocked at high stocking density gained more weight than the moderate stocking rate, a possible consequence of more severe defoliation the first year producing more tender, young resprouts. In addition, goats on the moderate stocking density had not fully penetrated some of the rose, blackberry, and honeysuckle thickets. There were significant differences in animal gains across periods. Goats gained more the first period because of availability of the most palatable species. It is unknown why animals lost weight the third period, but it was consistent across all treatments in 2002. It may have been a consequence of the elevated ambient temperatures that goats were exposed to at this period. Wethers gained more than doelings as would be expected. Gains were not as good as expected, but these were yearling animals put in a new environment where they had to learn new plants to eat. Also at their age they had somewhat limited growth potential. Goats were forced to clean the less palatable herbage up to obtain quicker vegetation management results. A producer should be able to achieve greater performance levels on his animals at a lighter stocking density if he is patient about controlling vegetation. There was a significant pasture by period effect (P<.001) at Osage in 2002 (Table 28). This was due to half of the animals on the high stocking rate being removed during the last period because forage availability was so limited that animals lost substantial weight in period 2. With the lighter stocking density, animals on the high stocking gained significant weight in the third period. There was a breed effect in that Spanish goats performed better than either Boer cross or Alpine goats. This was a very rugged area and Spanish goats are rugged animals and better able to make a living under these conditions. During the last period of the first year at the Seminole site Spanish goats performed better than Boer cross when herbage was very limiting. Goats gained least weight in period 2, a consequence of forage becoming limiting as the season progressed. Gains were greater in period 3 due to the reduction in animals on the high stocking rate pasture. There were no differences in pastures or sex in 2003. Only one breed was used in this year. The only differences were due to periods. There was a reduction in weight gains in period 2 as the higher quality forage was consumed. There was an increase in gains in period 3 due to half of the goats being removed from each pasture due to forage becoming limiting. Data for gains at the Choctaw site are shown in Table 29. There was a pasture*period interaction in that goats on the mixed grazing gained more weight the first and last periods while those on the goats alone plot gained the most the second period. There were no differences due to pasture. Animals gained the least weight in period 3 due to high quality forage becoming limiting. There was a significant sex by period effect due to wethers gaining more weight the first period than doelings, about the same as doelings the second period and gained less weight than doelings the third period. Angoras gained less weight than B.S. at all periods, but the difference decreased as period increased. Angoras are smaller than the B.S. and less able to compete for browse and have an extra nutrient requirement for mohair growth. Only one breed and sex were used on the study in 2003. There was a significant pasture by period interaction (P<.001) in that animals on the mixed pasture gained more the first two periods, but gained less weight the third period. The animals on this pasture had more patch grazing habits-grazing the bermudagrass down very short which resulted in high quality forage. However, with this short grazing habit, parasite problems accumulated, drastically affecting animals as the end of the second period and the third period. Two animals died and there was some morbidity from parasites on this treatment. Goats at the Cherokee site had the most rapid gains of any site (Table 30). Two replicated pastures were used on the study rather than two different goat grazing treatments. In 2002, there was a sex by period interaction in that wethers gained considerably more than doelings the first period, whereas gains were similar between sexes in subsequent periods. Overall, wethers gained 50% more than doelings. There was a period effect in that animals gained the most weight in the first period due to consumption of the most nutritious species, but in later periods their performance was impaired by summer heat and being forced to eat less desirable plant parts. There was a breed by period interaction in that Boer x Spanish gained 50% more weight than Spanish in the first period, although when gains were lower in subsequent periods, they were similar between breeds. This reflects the greater growth potential of the Boer breed when adequate nutrition is available as observed by Waldron et al (1994) observed that the greater growth potential of Boer goats would only be realized if there was sufficient nutrition to support that growth.. In subsequent periods, growth was limited by nutrient availability rather than genetic potential for gain. The lower gains of Angora goats reflect their smaller size and nutrient requirements to produce mohair as discussed above. There was a breed by sex interaction in that B.S. wethers gained much more than doelings, however, wethers of the other two breeds had similar gains to doelings. In 2002, only doelings of two breeds were used. There were no interactions and only a period effect. Gains by period followed the pattern in the previous year, being greater the first period and decreasing in the following periods due to forage availability becoming limiting as grazing progressed and the effects of summer heat. Gains for the Sac and Fox project are shown in Table 31. There were effects due to pasture (P<.001), breed (P<.001) and the interactions of pasture and breed (P<.02) and pasture and period (P<.001). The pasture x period interaction was caused by the removal of some animals after period 1 due to limited forage availability. This was a dry site and they failed to get spring and early summer rains, so we removed eight of eighteen the animals from the moderate treatment and 36 of 48 on the rotationally grazed treatment. This improved gains the last period in these pastures due to less competition for available forage. We started on a third period and supplemented animals for two weeks with hay and then abandoned the study because forage was non-existent and there were no long term predictions for rain. There were significant period differences in 2003 with a period by pasture interaction (P<.03). Again, animals had to be removed during the study because of forage limitations. Since goats lost weight even in period 1, especially on the moderate stocking rate, half animals were removed from this treatment and 4 animals were removed from the rotational pasture. Animal numbers were further reduced after the second period due to forage limitations and 6 animals continued on the rotational grazing pasture and 3 animals continued on the low stocking rate and 3 animals on the moderate stocking rate for the third period. Soil productivity was limiting and the overstory was closed on over 70% of the pastures further limiting animal productivity. The pasture by period interaction was caused by reduction in animal numbers increasing animal productivity, especially on the moderate grazed group. Overall animal productivity was low, but animals were grazed at high stocking densities to maximize impact on the vegetation.
Fecal egg counts

Fecal egg counts (Table 32) were low at most locations both years and remained far below the threshold requiring deworming (1000 egg/gram level) at all but one location (location difference P<.001). This was because animals at the remaining locations were browsing off the ground in late spring and early summer when moisture was conducive to transmission of parasites. Infective larvae typically are located within 8 cm of the ground and therefore goats avoided them. Later in the summer goats were cleaning up their pastures and were grazing close to the ground, it was hot and dry which kills infective larvae. Larvae numbers increased over the grazing season (P<.001) as would be expected but outside of Choctaw, fecal egg counts were insignificant relative to animal health and well-being. Fecal egg counts were much lower the first year at Choctaw than the second year (30 vs 1326). Although higher the first year than at other locations, fecal egg counts did adversely affect animal health although they were high enough levels at the end of the season to cause some concern. The second year, fecal egg counts were health threatening and caused morbidity and two mortalities. The difference between years was due some to a wetter summer the second year, but the grazing habit of the goats was different between years for unknown reasons. The east half of the pastures at the Choctaw location were mature forest while the west half was improved pasture. In the first year, goats browsed underbrush in the forest heavily whereas in the second year the underbrush was poorly utilized. Instead, goats preferred to graze the pasture and preferentially grazed certain patches very close (less than 8 cm high). Since they concentrated on these areas, there was a high concentration of fecal pellets and there was sufficient moisture and warmth for infective larvae to migrate up the grass and be consumed by the closely grazing goats. Goats were dewormed the third month of the second year and needed dewormed again the following month when they were removed due their grazing habits. Basically, fecal egg counts show that when goats are used for vegetation management, parasites will not be a significant problem except in extenuating circumstances. In conclusion, goats did consume many woody and weedy species in this study. However, only partial control was obtained in the two years of this study since perennial species have tremendous root carbohydrate reserves which will take several years to exhaust. A surprising finding was that the same species is not equally palatable at all locations. Environment including soil and rainfall patterns appears to affect palatability and actually these differences in palatability can be observed at adjacent sites. Another surprising finding was that as control of a species is obtained that less palatable species may be released from competition and proliferate until it is controlled by grazing.

Participation Summary

Educational & Outreach Activities

Participation Summary:

Education/outreach description:

Field days were held at each location in Fall of 2002 and 2003. The events were advertised with assistance of the County Extension Agents, through their regular newsletters and through NRCS personnel. In addition, flyers announcing the meeting were posted at farm and convenience stores within a 25-mile radius of the meeting site. The field days were usually held in meeting facilities of the Native American Nations and they provided coffee and donuts for the meeting. We then had talks on what the goats did to the vegetation at the other sites, a short talk on the goat behavior with GPS collars and then a talk with digital pictures about the impact of goats at this location. After lunch, we then took a pasture walk at the sites to actually see the change between the control areas and where goats were. Goats often came up out of curiosity and participants got to watch closely what the goats ate and often pet the goats. We reached over 200 people who attended these field days. Attendance at field days in 2002 were 24 at Osage, 18 at Caddo, 80 at Cherokee, 18 at Sac and Fox and 27 at Choctaw. Attendance at the field days in 2003 were 60 at the Seminole site, 38 at Osage, 25 at Caddo, 68 at Cherokee, 12 at Sac and Fox and 14 at Choctaw. Most participants signed up for our Institute’s Goat Newsletter and picked up handouts on goat management and vegetation management by goats. Some were new to goats and needed further information on goats and were glad to learn of a source for goat information. One Tribal councilman asked us to bring 1200 goats for next year because they had a lot of land in need of vegetation management by goats. Also, when we appeared and spoke at Tribal Council meetings, in every case, there were Native Americans that had been to the sites and spoke up in strong support of the project. We had requests from two locations for field days this year at the conclusion of the studies.

We received phone calls requesting information on goats and what weeds goats controlled from people who passed by the studies and saw the sign about what was going on. The topic of goats for vegetation management has become a popularly requested speaking topic for Langston goat extension specialists. We draw on our experience with this project to answer producer questions relating to the use of goats to control vegetation and give them examples from these studies. Our Goat Institute and this project were highlighted on an agricultural show on TV. We had a number of articles in local and tribal newspapers about the project. There is strong interest among the Native American Nations in this project. Two Nations are acquiring goats to clean their land up for establishing organic gardens. One Nation plans to use the goat project as a cornerstone for their at-risk youth program. At several sites, the farm crew personnel working with the project have gotten their own goats at their own farm and other members of the Native American Nations are watching them and asking questions. We have essentially created more trainers that are closer to those that we are trying to reach. At one site, we employed a unemployed handicapped youth to take care of the goats. We and a tribal elder are trying to help the youth to get a goat enterprise. He is quite gifted in working with animals and has learned considerable about caring for goats, goat behavior, what goats eat and what species the goats are likely to control. This elder sees it as a way to keep the youth out of trouble as well as teach him business principles and facilitate him becoming a self-supporting adult.

Results of the animal behavior part of the study have been presented at the Eighth International Goat Conference this year. Abstracts have been presented on fecal egg counts and animal performance have been presented at national animal science meetings.

Project Outcomes

Project outcomes:

Farmer Adoption

It is difficult to quantify farmer adaption of this technology, especially since the government does not keep goat statistics. However, judging from the number of requests for information and the increased number of goat herds as one travels across Oklahoma, it is obvious that there is significant application of this technology. Also, these are like demonstration sites in that cattlemen passing by do notice the brush that goats consume and are forced to think about it in their situation. In most locations, mechanical and chemical control of brush and weeds is too expensive and the results do not last very long. As a result of years of not using brush control, brush and woody species are a serious problem in grazinglands in Oklahoma.

Recommendations:

Areas needing additional study

It is evident that two years are inadequate for goats to affect vegetation, especially woody and perennial species, but since the goats are completely defoliating nearly all woody species that they can access, these woody species will succumb to successive defoliations. We are applying to SARE for a grant to continue this study for three more years. We have procedures in place to monitor long term changes in not only woody and herbaceous vegetation, but also monitor the effect on soil nutrients and litter. Another area of investigation needs to be the effect of various factors on the palatability of a non-palatable woody species such as red cedar and sumac, the two most common species in the state with palatability problems. Various treatments could be tried such as mowing, rollerchopping, burning, fertilizer or liming as far as inducing animals to consume the species. These treatments are practical to apply. If factors are observed to affect palatability, there could be further investigation of any change in browse quality or chemical content (would probably require GC-Mass spec analysis of plant extracts) coinciding with the change in preference. Also, there is a need to study the impact of goats in riparian areas. It is well known that grazing riparian areas with cattle and sheep contribute to erosion and pollution, but our observations have been that gullies heal with grazing by goats- the sides become more gradual and rounded and vegetation becomes established on the banks. This is likely a consequence of goats having an aversion to walking in water and they will spread across an area when grazing across a riparian area and are less likely to make a trail leading to erosion. In fact, we have observed that as goats cross gullies en mass, they tend to round down the sides and the soil disturbance by their hooves is conducive to seedling establishment. Also, are goats more effective at trampling in seed when being used for rehabilitating grazing areas than are cattle?

Any opinions, findings, conclusions, or recommendations expressed in this publication are those of the author(s) and do not necessarily reflect the view of the U.S. Department of Agriculture or SARE.