Implementing Noxious Weed Control Through Multi-Species Grazing

Final Report for SW03-006

Project Type: Research and Education
Funds awarded in 2003: $187,935.00
Projected End Date: 12/31/2007
Matching Federal Funds: $187,935.00
Region: Western
State: Washington
Principal Investigator:
Dr. Donald D. Nelson
Washington State University
Expand All

Project Information


A partnership including private landowners, a contract sheep and goat grazier, Washington State University Extension, USDA/ARS, Washington Department of Fish and Wildlife and the Big Bend Resource Conservation and Development Council joined forces to investigate the use of multi-species grazing as a tool, in combination with herbicides, mechanical clipping and/or tillage, in the control of noxious weed infestations. Grazing combinations of cattle, sheep and goats were evaluated as a tool in an integrated pest management strategy in the control of invasive plants. Two field sites were selected in Eastern Washington (i.e., Sprague Lake and the Barker Ranch) to monitor treatment impact over three grazing seasons (2004-2006).

Project Objectives:
Objectives/Peerformance Targets
  1. Management Support Group (MSG): One MSG group made up of 6-12 participants and livestock owners involved in the SARE/PDP, Noxious Weed Control Through Multi-Species Grazing and other cooperators will be formed by January 1, 2003.

    Holistic system: Develop a holistic systems approach on two ranch units, involving 2,600 acres, to address the control of invasive plants and noxious weeds, including Russian olive, scotch thistle, perennial pepper-weed and knapweed species. This system will evaluate weed management effectiveness; long-term sustainability of control and it will outline a process that is economically, environmentally and socially sound.

    Overall effectiveness: Compare overall effectiveness of the holistic system to an herbicide only system.

    Information dispersal: Provide information to producers and the community to improve their ability to formulate land management decisions based on sustainability of the weed control methods.

    Producer involvement in studies: Involve 10 additional producers in designing and implementing studies to insure the information meets the needs of producers and is transferred to a larger audience.

    Hands-on teaching: Provide participants with a deeper understanding of the concepts of multi-species grazing by teaching what they have learned to the public and their peers at 3 tours held one per year beginning in 2005.

    Enterprise evaluation: Evaluate the feasibility of adding additional enterprises such as hair sheep, cashmere fiber producing goats and meat goats.

    Forage quality: Determine if forage quality is adequate to meet animal nutritional needs at various stages of plant growth and biological stages of the livestock species used.

    Livestock management: Evaluate the effectiveness of each livestock species in utilizing each noxious weed species. Determine the most appropriate livestock species to be used at different stages of growth for each noxious weed. Gain knowledge of when, how much and how often to graze the target plant species for maximum impact on them while having minimum impact on the desirable species.

    Secondary compounds: Determine if there are any secondary compounds that may inhibit the utilization of the noxious weeds, as well as the impact on soils and plant germination.

    Professional publications: Publish information and results of the study in two professional journals and two regional media publications.


Sheep, goats and cattle are effective weed management tools, as well as important sources of food and fiber when properly managed. In the western United States, cattle have been the primary grazers that have favored a shift in the plant community to forbs and shrubs. Herbicides have been the main tool used to combat this shift and to reduce invading species. Sheep and/or goats as a tool to harvest vegetation and balance the plant community have seen limited use in the United States. As a result, the knowledge needed to effectively apply the tool of multi-species grazing is limited.

Weeds degrade soil and water resources, reduce biodiversity, reduce wildlife habitat and cattle forage production, and alter the integrity and function of the ecosystem. "Biology and Management of Noxious Rangeland Weeds” (Sheley 1999) focuses on how noxious weeds affect the ecological and economic systems. In 1988 it was estimated that knapweed contributed to the loss of forage on range and pasture land in eastern Washington of 130 million pounds annually. In 1988 dollars that loss translated into $951,000 if replaced as pasture or rangeland animal unit months (AUMs), or $2.9 million if lost forage was replaced as hay, based on $44/ton (Roche' 1988). Those figures were based on an infestation of approximately 427,000 acres. In Colorado, diffuse knapweed expanded from 30,000 acres in 1989 (Lacey 1989) to an estimated 100,000 acres in 1997 that is over a 3-fold increase in 8 years. If we assume the rapid increase of diffuse knapweed in Colorado represents what is happening throughout the West with noxious weeds, and factoring in today's forage costs, the economic impact on the cattle industry could be significantly higher. These estimations of economic loss of forage production as a result of noxious weeds are based on cattle as the primary forage harvester, which does not take in account the value of the forage (often destroyed by herbicides) for other livestock, such as sheep and goats.

Sheep, goats, and cattle used in conjunction as weed control tools might reduce herbicide costs by utilizing forages (weeds) that contribute to high livestock forage and pasture losses under cattle-only management. In this way, lands infested with noxious weeds can generate revenue by producing significant food and fiber instead of providing little usable forage for cattle coupled with high herbicide bills. In this project, field trials tested and compared the effectiveness of cattle, sheep, goats, biological controls, and herbicide for managing weeds on private land.

An integrated approach using managed multi-species grazing, and selective herbicides makes the most efficient use of nonrenewable and on-farm resources, and integrates natural biological cycles and controls. This project looked at how to integrate sheep and/or goats into an existing cattle operation with minimal additions to infrastructure. The use of timing of grazing, and dietary preferences to shift the plant community to the desired state should reduce the reliance on herbicides and fossil fuels. This process is powered principally by sun energy in place of fossil fuels. Managed grazing should increase the litter cover on the soil and reduce the potential for erosion as plant material is returned to the soil surface via manure and incorporated by trampling.

This project set out to identify the most sustainable, effective, and economically viable alternative for weed control on private land. Input and involvement of landowners and community representatives in the design, implementation, and review of the project were employed in an attempt to ensure enhanced quality of life for agricultural landowners, improve wildlife habitat, and bring dollars to their community.

Riparian ecosystems

Riparian areas, including lakeshores and stream banks, are dynamic, directly influenced by water for at least part of the growing season, and depend on periodic flooding for new seedling establishment. They are comprised of an herbaceous understory below shrubs, young and mature trees such as willows (Salix spp.) and cottonwoods (Populus spp.), as well as standing, dead trees called snags. Some important functional attributes of riparian areas include: (1) a diverse plant community supported by a unique hydrology that combines groundwater and surface water interactions and deposited fine sediments with greater water-holding capacity, (2) carbon transfer through litter deposition supporting both in-stream and downstream aquatic organisms, (3) possible reduction of downstream flooding by stabilizing stream banks, providing surface roughness that attenuates flow energies promoting infiltration and uptake by plants, and (4) habitat for many different aquatic and terrestrial species (NRCS 1996b; Parker and Williamson 2003).

Figure 1: Example of dense Russian olive trees in a Riparian area

Over time, many riparian areas have been heavily disturbed by fire suppression, dam construction, unmanaged recreation, and overgrazing, allowing for encroachment of nonnative invaders such as Russian olive (Elaeagnus angustifolia L.) (Parker and Williamson 2003). Russian olive is widespread in riparian areas in all of the western United States (Figure 1) (Shafroth et al. 1995; Stoleson and Finch 2001; Collins 2002; Stannard et al. 2002) and continues to degrade these riparian ecosystems (Parker and Williamson 2003). Russian olives can maintain a competitive edge by fixing nitrogen and survive bare, mineral substrates to fill in riparian vegetation gaps after native over-story trees have died, replacing trees, shrubs, and displacing associated wildlife (Muzika and Swearingen 1998; Stoleson and Finch 2001; Collins 2002). Russian olive modifies riparian ecosystems by blocking stream channels, lessening the abundance and diversity of wildlife, and possibly altering stream hydrology and nutrient cycling (Parker and Williamson 2003; Tu 2003).

The majority of documented effective Russian olive control has been achieved using herbicide alone or physical removal followed by an herbicide application (Katz and Shafroth 2003). These approaches can be time-consuming and costly, and in the case of some herbicides, inappropriate because of proximity to water (Katz and Shafroth 2003).

Justification for Russian olive control research on the Barker Ranch

On January 22, 2004, the United States District Court for the Western District of Washington issued an order establishing herbicide buffer zones adjacent to certain waterways in Washington, Oregon, and California (Figure 2) (US EPA 2004). These buffer zones extend 18 m from the defined water’s edge for ground herbicide application and 91 m for aerial application to protect threatened and endangered species such as steelhead and salmon (US EPA 2004). Therefore, the need to manage invasive riparian area vegetation while protecting aquatic species necessitates an alternative, environmentally sound method of weed control within these herbicide-prohibited buffer zones.

Figure 2: Counties in Washington, Oregon, and California (shaded green) in which pesticide-use buffers may have been ordered by the US District Court and support threatened and endangered salmon or steelhead habitat
(US EPA 2004).

Russian olive dominates much of the riparian area along parts of the Yakima River in West Richland, WA that supports anadromous salmon and steelhead. However, the United States Department of Agriculture’s Natural Resources Conservation Service (NRCS) recommends a 91-m herbicide- and livestock-free riparian buffer along the particular stretch of the Yakima River that runs through the Barker Ranch (Barker) study site to maintain wildlife habitat (NRCS 2002). In pastures adjacent to this buffer zone, the effects of a US EPA-approved herbicide and two grazing combinations of cattle and goats on Russian olive were compared. Livestock, such as goats and cattle, have been successfully used to control many types of undesirable vegetation (Frost and Launchbaugh 2003) such as blackberries (Rubus discolor) (Cox 2003), larkspur (Delphinium barbeyi L.) (Lamming 2001), cheatgrass (Bromus tectorum) (Mosely 1996), and leafy spurge (Euphorbia esula L.) (Walker 1994).

Objectives and Hypotheses

The overall objective of Amy Hummer’s M.S. thesis study was to determine if multi-species grazing provides an effective alternative to herbicide for Russian olive control. Specific objectives were to compare the effects of herbicide and multi-species grazing on Russian olive and to compare the effects of grazing order (goats followed by cattle and cattle followed by goats) on Russian olive. The hypotheses were: (1) Russian olive density is not affected by treatment (grazing, herbicide application, and no treatment), (2) Russian olive aboveground biomass is not affected by treatment, (3) Russian olive density is not affected by grazing order, and (4) Russian olive above ground biomass is not affected by grazing order.

Literature Review

Russian olive

Each year in the United States, more than$ 2B is spent in an attempt to eradicate invasive weeds that have managed to successfully elude 50 years of herbicide technology (Dekker 1997; DiTomaso 2000). Russian olive, a large deciduous shrub or small tree growing up to 10 m in height, has been classified as a noxious weed in both New Mexico and Colorado (Collins 2002; Stannard et al. 2002; Miller 2003; NRCS 2007), and it is characterized by silvery, alternate, lanceolate leaves (Muzika and Swearingen 1998; Stannard et al. 2002; Katz and Shafroth 2003) approximately 4- to 10-cm long on thorny branches (Collins 2002; Miller 2003). At age three flowers and small fruits (Collins 2002; Stannard et al. 2002) begin to appear on the trees (Muzika and Swearingen 1998). The fruits, present from August through October (Miller 2003), contain seeds able to grow in a wide range of soil types (Kindschy 1998).

Introduction to United States, Historical Uses, and Demographics

Russian olive is native to southern Europe and western Asia (Muzika and Swearingen 1998; Collins 2002; Miller 2003). It was first introduced to the United States in the late 1800s as an ornamental shrub (Muzika and Swearingen 1998; Collins 2002; Stannard et al. 2002; Parker and Williamson 2003). By the 1940s, Russian olive was widely used in the Great Plains for windbreaks (Kindschy 1998; Stannard et al. 2002). Its use has been advocated as recently as the 1990s for bee nectar and erosion management by the NRCS (Collins 2002; Katz and Shafroth 2003). Today, Russian olive can be found in all but 13 states in the southern US and has been reported as an invasive species in greater than two-thirds of all states (Katz and Shafroth 2003). The largest populations are found in the central, western, and Great Plains states (Collins 2002). In central Washington, where shrub steppe has been converted to mitigated wetland, Russian olive occurs along riparian corridors often growing in combination with common cattail (Typha ssp.), bulrush (Scirpus spp.), common reed (Phragmites ssp.), and purple loosestrife (Lythrum salicaria L.) (Zouhar 2005).

Environmental Tolerances and Dispersal Mechanisms

Russian olive is heavily invasive in wet saline and riparian areas (Stannard et al. 2002). A successful opportunist, Russian olive easily colonizes an area following a disturbance (Collins 2002). It spreads rapidly, and with relative ease; due to its wide range of tolerance for varying soil moisture, soil type (most commonly Mollisols or Entisols), soil texture (from clay loam to sand), salinity, temperature range (-45°C to +46°C), and shade. It is also aided by its nitrogen fixing capability and swift seed dispersal by a number of different birds (Shafroth et al. 1995; Muzika and Swearingen 1998; Collins 2002; Katz and Shafroth 2003; and Zouhar 2005). The seeds are dispersed in the fall and winter (Katz and Shafroth 2003). Russian olive seeds are well protected against digestive enzymes and effectively survive in a range of soil types for up to three years upon excretion by animals (Collins 2002).

Ecological Value

Russian olive expansion along riparian corridors presents an interesting paradox. It fixes nitrogen, which is beneficial to surrounding vegetation, provides habitat and food for birds, and provides an effective barrier against soil erosion with its extensive root system (Shafroth et al. 1995; Collins 2002; Katz and Shafroth 2003). However, it displaces native trees, reduces avian species richness, and often invades irrigation canals and moist or irrigated pastures (Collins 2002; Katz and Shafroth 2003; Zouhar 2005).

Physical and Chemical Defenses

Physical and chemical defense mechanisms of mature Russian olive against grazing include the presence of thorny branches and leaves, which contain defense compounds (Katz and Shafroth 2003). The main function of thorns and spines is not to eliminate grazing altogether, but to restrict the bite size of browsing ungulates thereby increasing handling time (Owen-Smith and Novellie 1982; Cooper and Owen-Smith 1986). Unless coupled with small leaves to significantly increase handling time, spines themselves are not effective in deterring goat grazing (Cooper and Owen-Smith 1986).

Secondary metabolites are a plant’s chemical defense against grazing (Provenza et al. 1994). Cipollini and Levey (1997) define secondary metabolites as compounds found in plants serving no primary purpose, but which may be toxic, or at least deter predators such as herbivores. Secondary metabolites are not always detectable by animals, as they make lack an odor, taste, or may even imitate nutrients. Herbivores are capable of processing secondary metabolites, but with limitations (Freeland and Janzen 1974).

Control and Elimination

Complete elimination of Russian olive is often impractical for extensive stands, but practical for small, isolated stands (Stannard et al. 2002). Traditional approaches include physical removal, herbicide application, or a combination of the two, and are met with variable success. These methods can be time-consuming and costly, or in the case of herbicides, may not be suitable in environmentally sensitive riparian areas (Katz and Shafroth 2003).

Physical removal includes mowing, burning, cutting, bulldozing, and grazing, but is variably successful when used alone, as new shoots and roots sprout readily from the stumps and crowns (Collins 2002; Stannard et al. 2002). It is thought by some that the spread of young Russian olive trees can be reduced using introduced livestock (Katz and Shafroth 2003; Zouhar 2005), and Lamming (2001) states that mature, male goats actually prefer to forage on Russian olive. The NRCS (2002) has recommended goats as a management tool in Russian olive-infested pastures at Barker. For cattle, Zouhar (2005) ranked the palatability of Russian olive as poor, and it does not appear to be significantly impacted by wildlife grazing (Collins 2002; Katz and Shafroth 2003).

Chemical herbicides such as Garlon-4® (active ingredient triclopyr) and Arsenal® (active ingredient imazapyr) are commonly used for selectively or non-selectively controlling woody plants such as Russian olive (Parker and Williamson 2003). Concentration and application method differ for young and mature trees. For saplings, Miller (2003) recommends an even leaf coverage of a 2% solution of Garlon® in water combined with a surfactant between July and October, or a 20% solution mixed with oil or fuel as a penetrant applied as a basal spray. Katz and Shafroth (2003) found a 4% solution of imazapyr provided significant initial damage to mature Russian olive; however, long-term damage was not evaluated. Although Russian olive is responsive to triclopyr and imazapyr, a follow-up application within one or two years will almost always be required (Stannard et al. 2002).

A combination of treatments is likely the best approach to Russian olive control (Muzika and Swearingen 1998; Collins 2002; Katz and Shafroth 2003). Burning followed by an herbicide application is one control option, although physical removal followed by an herbicide application to the remaining stumps has proven to have an even greater effect (Muzika and Swearingen 1998; Collins 2002). For example, a study by Washington State University’s Plant Materials Center showed that after removal of Russian olive trees, mowing the 0.6-m-high re-growth followed by an application of Garlon® provided long-term control (NRCS 2005). For best results, smaller trees should be removed from moist soil and larger trees cut at ground level followed by an immediate herbicide application to remaining stumps (Katz and Shafroth 2003).

In general, any method of Russian olive removal will require follow-up attention to suppress new root crowns from sprouting (Katz and Shafroth 2003). Additionally, preventing initial seedling establishment in Russian olive-prone areas through temporary inundation or target-specific grazing can reduce costly and time-consuming control efforts later (Katz and Shafroth 2003).

Grazing Animals/Prescriptive Grazing

The concept of using livestock for prescriptive grazing purposes first appeared in the scientific literature in the 1930’s (Mosley 1996). In any prescriptive grazing regime, livestock dietary niches, preferences, and body size, as well as timing, stocking intensities, and durations should be considered (Freeland and Janzen 1974; Sidahmed et al. 1981; Provenza and Balph 1988; Distel and Provenza 1991; Gross et al. 1993; Walker 1994; Perryman et al. 1995; Bailey et al. 1996; NRCS 1996a; Madsen 2004).

Livestock fall into one of two dietary niches defined by digestive system fermentation type – either ruminant (foregut) or cecal (hindgut) (Walker 1994). Ruminants can be further classified as concentrate selectors, intermediate feeders, or grass/roughage feeders based on physiological and morphological needs (Walker 1994). Goats are intermediate feeders adaptable to variable forage. Cattle are grass/roughage feeders based on morphological features such as mouthparts that refine grazing selections (Provenza and Balph 1988; Walker 1994). Cattle, having no incisors in their upper jaw, use their tongue to sweep vegetation into their mouths, and “bite” off pieces using their dental pad and teeth on the lower jaw. Goats have a flexible upper lip and prehensile tongue allowing them to graze vegetation close to the ground (Walker 1994).

Cattle are known to prefer grass over shrubs, plant leaves to stalks, and green matter to dry matter (Walker 1994; Madsen 2004). Goats prefer broadleaf and woody plants (NRCS 1996a; Madsen 2004), and are useful for modifying shrub growth to increase their nutritional value and productivity (Distel and Provenza 1991). Goats are a feasible alternative to chemical or physical control of invasive vegetation (Distel and Provenza 1991).

Body size is also associated with an animal’s ability to regulate the degree of forage ingested (Provenza and Balph 1988; Walker 1994; Bailey et al. 1996). Smaller herbivores have the advantage of allocating more time to searching for higher quality forage while larger herbivores, requiring more energy, cannot afford to be as selective in the same timeframe (Walker 1994; Bailey et al. 1996).

The most advantageous time to turn animals onto target forage is when the target species is most susceptible to defoliation and/or when desirable vegetation will incur minimal impact (DiTomaso 2000). Grazing applications can cause a shift in plant communities when pressure is timely and correctly applied, but patience is required, as results may not be evident for three or four years (Madsen 2004). Currently, no published information exists on timing of grazing applications to control Russian olive, nor does information exist on specifically using prescriptive livestock grazing to control Russian olive.

Stocking intensity is also important to consider in prescriptive grazing applications. Sidahmed et al. (1981) found the chemical and botanical make-up in the diet of goats could be correlated to range condition and stocking intensities. Perryman et al. (1995) believe that the essence of plant damage does not lie as much in the timing of grazing applications as in the duration and intensity.

Multi-Species Grazing

Prescriptive grazing management strategies can be used to control the spread of invasive weeds by: (1) moderately grazing a single species to shift plant composition through dietary preference, (2) intensely grazing a single species using high stocking rates to force animals to utilize all vegetation, or (3) grazing with multispecies (DiTomaso 2000). Multi-species grazing can occur simultaneously (common use) or sequentially. In this study, cattle and goats were grazed sequentially; however, as discussed below it refers to simultaneous grazing, as a lack of published information exists on grazing multiple species sequentially.

Walker (1994) promotes multi-species grazing for prescriptive weed control. He discusses the benefits of different forage preferences among species leading to different dietary niche occupation and increased overall forage utilization. He states that for a single plant species, intra-specific grazing (between individuals of the same species) will always prove more competitive than inter-specific grazing (between different species).

Multi-species grazing is beneficial to ecosystems, can maximize animal production, and is economically advantageous, yet still is not widely practiced (Sidahmed et al. 1981; Squires 1982; Walker 1994). Multi-species grazing, using a combination of animals with differing morphological and physiological attributes, leads to differing pressures on foliage and is almost always be more beneficial to an ecosystem than single-species grazing (Walker 1994). Sidahmed et al. (1981) found that when feral goats grazed alone, their diet was comprised of 50% browse. When they grazed after sheep, this number increased to 70%-90%. Multi-species grazing has also been used to maximize animal production (Squires 1982). For example, when averaged over a number of studies, Walker (1994) found multi-species grazing yielded a 24% increase in meat production compared to cattle grazing alone, and a 9% increase compared to sheep grazing alone. However, despite the benefits of multi-species grazing, it is still not commonly practiced due to: (1) goat and sheep losses to predation, (2) difficulty in obtaining multi-species grazing permits, (3) lack of producer knowledge, and 4) prejudices about certain livestock species (Walker 1994).

Cattle and sheep share the greatest degree of forage preference overlap (Squires 1982; Walker 1994), while cattle and goats share the least as summarized from greater than 200 studies (Walker 1994). The combined results of these studies have shown that sheep consume 50% grass, 30% forbs, and 20% browse; while cattle consume 70% grass, 15% forbs, and 15% browse; and goats consume 30% grass, 10% forbs, and 60% browse (Walker 1994). Differences in dietary preference and documented benefits of multi-species grazing, combined with a lack of published information on sequential grazing itself and for control of Russian olive, make up the experimental framework for this study.

Literature Cited

Bailey, D. W., J. E. Gross, E. A. Laca, L. R. Rittenhouse, M. B. Coughenour, D. M. Swift, and P. L. Sims. 1996. Mechanisms that result in large herbivore grazing distribution patterns. Journal of Range Management 49:386-400.

Cipollini, M. L., and D. J. Levey. 1997. Secondary metabolites of fleshy vertebrate-dispersed fruits: adaptive hypotheses and implications for seed dispersal. The American Naturalist 150:346-372.

Collins, E. 2002. Introduced species summary project: Russian olive (Elaeagnus angustifolia L.). Invasion Biology Introduced Species Summary Project – Columbia University.

Cooper, S. M., and N. Owen-Smith. 1986. Effect of plant spinescence on large mammalian herbivores. Oecologia 68:446-455.

Cox, C. 2003. Nonchemical methods for removing unwanted blackberry plants. Journal of Pesticide Reform. 23:10-11.

Dekker, J. 1997. Weed diversity and weed management. Weed Science 45:357-363.

Distel, R. A., and F. D. Provenza. 1991. Experience early in life affects voluntary intake of blackbrush by goats. Journal of Chemical Ecology 17:431-450.

DiTomaso, J. M. 2000. Invasive weeds in rangelands: species, impacts, and management. Weed Science 48:255-265.

Freeland, W. J., and D. H. Janzen. 1974. Strategies in herbivory by mammals: the role of plant secondary compounds. The American Naturalist 108:269-289.

Frost, R. A., and K. L. Launchbaugh. 2003. Prescriptive grazing for wildland weed management. A new look at an old tool for controlling weeds on rangelands. Rangeland 25:43-47.

Gross, J. E., L. A. Shipley, N. T. Hobbs, D. E. Spalinger, and B. A. Wunder. 1993. Functional response of herbivores in food-concentrated patches: tests of a mechanistic model. Ecology 74:778-791.

Katz, G. L., and P.B. Shafroth. 2003. Biology, ecology, and management of Elaeagnus angustifolia L. (Russian olive) in western North America. Wetlands 23:763-777.

Kindschy, R. R. 1998. European starlings disseminate viable Russian-olive seeds. Northwestern Naturalist 79:119-120.

Lacey, J.R., C.B. Marlow, and J.R. Lane. 1989 Influence of spotted knapweed (Centaurea maculosa) on surface water runoff and sediment yield. Weed Technol. 3:627-31.

Lamming, L. 2001. Successfully controlling noxious weeds with goats. Alternative Weed Strategies 21:19-23.

Madsen, C. 2004. Vegetation management the natural way with goats and sheep. Sustaining the Pacific Northwest: Food, Farm, & Natural Resource Systems 2:1-3.

Miller, J. H. 2003. Russian olive nonnative invasive plants of southern forests: a field guide for identification and control. Gen. Tech. Rep. SRS-62. Asheville, NC: U. S. Department of Agriculture, Forest Service, Southern Research Station. 93 p.

Mosely, J. C. 1996. Prescribed sheep grazing to suppress cheatgrass: a review. Sheep & Goat Research Journal 12:74-81.

Muzika, R., and J. M. Swearingen. 1998. Plant Conservation Alliance, Alien Plant Working Group. Available at: Accessed 18 November 2004.

NRCS (Natural Resources Conservation Service). 1996a. Non-federal grazing lands in the United States. Program Aid Number 1552w.

NRCS (Natural Resources Conservation Service). 1996b. Riparian Areas Environmental Uniqueness, Functions, and Values. RCA Issue Brief #11.

NRCS (Natural Resources Conservation Service). 2002. The Barker Ranch LTD. Wetland reserve program plan, West Richland, WA.

NRCS (Natural Resources Conservation Service). 2005. Weeds – description of selected species, control, herbicide technology, and Washington State noxious weed lists. Available at: Accessed 3 May 2006.

NRCS (Natural Resources Conservation Service). 2007. Plants Database. Available at: Accessed 24 February 2007.

Owen-Smith, N., and P. Novellie. 1982. What should a clever ungulate eat? The American Naturalist 119:151-178.

Parker, D., and M. Williamson. 2003. Low-impact, selective herbicide application for control of exotic trees in riparian areas: saltcedar, Russian-olive and Siberian elm. USDA Forest Service. Southwestern Region, Albuquerque, NM. 4 p.

Perryman, B. L., W. A. Laycock, L. B. Bruce, K. K. Crane, and J. W. Burkhardt. 1995. Range readiness is an obsolete management tool. Rangelands 27:36-41.

Provenza, F. D., et al. 1983. Biological manipulation of blackbrush by goat browsing Coleogyne ramosissima in the southwestern United States. Journal of Range Management 36:513-518.

Provenza, F. D., and D. F. Balph. 1988. Development of dietary choice in livestock on rangelands and its implications for management. Journal of Animal Science 66:2356-2368.

Provenza, F. D., J. J. Lynch, E. A. Burritt, and C. B. Scott. 1994. How goats learn to distinguish between novel foods that differ in postingestive consequences. Journal of Chemical Ecology 20:609-625.

Roche', C.T., and B.F. Roche', Jr. 1988. Distribution and amount of four knapweed (Centaurea L.) species in eastern Washington. Northwest Sci. 62:242-51.

Schreuder, H.T., R. Ernst, and H. Ramirez-Maldonado. 2004. Statistical techniques for sampling and monitoring natural resources. USDA Forest Service Gen. Tech. Rep. RMRS-GTR-126. Fort Collins, CO. 111 p.

Shafroth, P. B., G. T. Auble, and M. L. Scott. 1995. Germination and establishment of the native plains cottonwood (Populus deltoides Marshall subsp. Monilifera) and the exotic Russian-olive (Elaeagnus angustifolia L.). Conservation Biology 9:1169-1175.

Sheley, R.L., and J.K. Petroff. 1999. Biology and management of noxious rangeland weeds. Oregon State University Press, Corvallis,OR.

Sidahmed, A. E., J. G. Morris, and S. R. Radosevich. 1981. Summer diet of Spanish goats grazing chaparral. Journal of Range Management 34:33-35.

Squires, V. R. 1982. Dietary overlap between sheep, cattle, and goats when grazing in common. Journal of Range Management 35:116-119.

Stannard, M., D. Ogle, L. Holzworth, J. Scianna, and E. Sunleaf. 2002. History, biology, ecology, suppression and revegetation of Russian-olive sites (Elaeagnus angustifolia L.). USDA – Natural Resources Conservation Service, Boise, ID, USA. Plant Materials No. 47, Technical Notes. 14 p.

Stoleson, S. H., and D. M. Finch. 2001. Breeding bird used of and nesting success in exotic Russian olive in New Mexico. The Wilson Bulletin 113:452-455.

Tu, Mandy. 2003. Element stewardship abstract: Elaeagnus angustifolia L., Invasives on the web: The Nature Conservancy wildland invasive species program. Davis, CA: The Nature Conservancy (Producer). Available at: Accessed 22 February 2007.

US EPA (U.S. Environmental Protection Agency). 2004. Court ordered buffers around pacific salmon-supporting waters. U.S. Environmental Protection Agency. Available at: Accessed 5 April 2006.

Walker, J. W. 1994. Multispecies grazing: the ecological advantage. In:Sheep Research Journal Special issue. p 52-64.

WRCC (Western Regional Climate Center). Available at: Accessed 4 April 2006.

WSU (Washington State University). Soils of Washington. Available at: Accessed 11 April 2006.

Zouhar, K. 2005. Elaeagnus angustifolia. In: fire effects information system. U.S. Department of Agriculture, Forest Service, Rocky Mountain Research Station, Fire Science Laboratory. Available at: Accessed 4 April 2006.


Click linked name(s) to expand/collapse or show everyone's info
  • Michael Carpinelli
  • Michael Crowder
  • James Dobrowolski
  • Amy Hummer
  • Craig Madsen
  • Andrea Mann
  • Skip Stonesifer


Materials and methods:

Project Experimental Design: Barker Ranch

Twelve pastures, each 30 X 60 m, were randomly established and received one of four experimental treatments: C (grazed first by cattle followed immediately by goats), G (grazed first by goats followed immediately by cattle, H (treated with Arsenal [impazpyr] or Garlon-4 [triclopyr] herbicide, or U (untreated control). Each of these treatments was replicated three times. All of the pastures were either adjacent to and/or diagonal from another pasture to maximize the number of Russian olive within each pasture.

In each pasture, two 60 m transect lines ran the pasture length, as to intercept Russian olive clusters and contained five permanent plots of 20 X 50 cm. The plot size and number of plots per pasture were incorporated into the design based on successful sampling in past similar experiments (M.F. Carpinelli, personal communication, April 2006). Systematic sampling with a random start was used to determine the location of the first plot on each transect (Schreuder 2004). The first plot began at 3 m, 6m or 9 m corresponding to the pasture’s replication (i.e., the first plot in pastures C1, G1, H1 and U1 began at the 3 m mark, while the first plot in C2, G2, H2 and U2 began at the 6 m mark and so on)) (Schreuder 2004). Subsequent plots were 12 m apart. Sampling times were chosen based on plant phenology and forage availability in response to spring time flood irrigation and fall re-growth after the first series of grazing treatments (M.F. Carpinelli, personal communication, June 2004).

Figure 3: Plot layout for conventional herbicide and grazing treatments at Barker Ranch near Richland, Washington.

Project Design: Sprague Lake

The project-planning group decided that the Sprague Lake site was going to be a learning site (i.e., an observational study), rather than an experimental layout like the one at the Barker Ranch. Sprague Lake is located in Adams and Lincoln counties in WA, two miles from the town of Sprague. The lake is six miles long and one mile wide. The learning site is located on the southwest end of the lake on land acquired by the Washington Fish and Wildlife Commission in 2003 from the Hercules Ranch owned and operated by Rex Harder. It has been designated as a “watchable wildlife’ site.

The main target plants at this learning site are diffuse knapweed, Dalmatian toadflax, rush skeleton weed and annual forbs. The plots have a mixture of native perennial grasses, native forbs, native shrubs, annuals grasses, annual forbs and scattered patches of noxious weeds. The site was grazed with goats and sheep twice each year in 2004, 2005 and 2006 in order to prevent seed production on the target plants. For a plot map refer to Figure 4 and for the plot sizes refer to Table 1.

Figure 4: Sprague Lake plot map

Table 1: Sprague Lake plot sizes

Study Site: Barker Ranch

This study was conducted at the Barker Ranch (Barker) during spring and fall grazing seasons in 2004, 2005 and 2006. Barker, just west of Richland, Washington, is located in Benton County and encompasses Sections 13, 23, 24, and 25 of Township 10 N, Range 27 E WM, as well as sections 18, 19, 29, 30, and 31 of Township 10 N, Range 28 E WM. The study site consisted of about 6 ha located within 809 ha of privately managed wetland and adjacent upland ecosystems along the Yakima River. The study site elevation was 128 m and the average low and high temperatures during the months in which data were collected in 2004 were 13.5°C and 27.8°C, respectively. In 2005, they were 12.7°C and 27.8°C, respectively (WRCC 2006). The average amount of rainfall during the data collection period during 2004 and 2005 were 1.83 cm and 0.99 cm, respectively. The average total annual precipitation in West Richland was 18 cm (WRCC 2006).

Predominant study site soil types include Xeric Torripsamments, Xeric Torriorthents, and Xerollic Torriorthents (WSU 2006). These soil types are characterized by dry, sandy soils formed beneath sparse dune or shrub-steppe vegetation having low water-holding capability.

Nineteen plant communities characterized the study site area. The most abundant communities included annual grasses/saltgrass (Distichlis spicata L.) (25%), dense Russian olive (Elaeagnus angustifolia L.)/locust (Robinia pseudoacacia L. and Gleditsia triacanthos L.) (16%), tall wheatgrass (Agropyron elongatum) (15%), tall fescue (Lolium arundinaceum)/Kentucky bluegrass (Poa pratensis L.)/clover/sedge (15%), annual grass/tall wheatgrass (10%), tall wheatgrass/tall fescue/Kentucky bluegrass/clover/sedge (10%), mowable Russian olive with tall wheatgrass (5%), and sedge/rush/Nebraska sedge (Carex nebrascensis) (4%) (Table 2) (NRCS 2002).

Table 2: Most abundant plant communities identified at the Barker Ranch
(from NRCS 2002).

The Barker Ranch is managed as a series of wetland and adjacent upland ecosystems with the intent of providing habitat for migratory waterfowl and upland species such as ring-necked pheasants (Phasianus colchicus), mule deer (Odocoileus hemionus hemionus), and California quail (Callipepla californica). Vegetation management, including Russian olive control, is in accordance with a set of planned goals designed by the NRCS and includes flood irrigation and livestock grazing to best meet the habitat and forage needs of these species (NRCS 2002). The ranch is flood irrigated, generally from March through mid-October, using water from the Yakima River. The diverted water is passed through a series of canals and water control structures for irrigation purposes to maintain desirable plant communities to sustain wildlife habitat . Livestock grazing has been used on the Barker Ranch to control unwanted vegetation in an effort to create suitable waterfowl and other wildlife habitat (NRCS 2002). Cattle grazing is currently permitted in designated areas for the purpose of maintaining desired vegetation heights to maximize its usefulness to nesting birds and other wildlife species. Historically at Barker, several tactics have been tried in an effort to control Russian olive and maintain open waterfowl habitat including cattle grazing, mowing, and/or herbicide application (NRCS 2002). Cattle are used to graze older perennial grasses and to reduce overall vegetation, but have been ineffective in reducing Russian olive densities. Due to root and shoot sprouting, mowing Russian olive alone has increased the number of trees. A Garlon® application followed by mowing dead Russian olive “skeletons”, in subsequent years, has been a common method of control. This method has not reduced the number of Russian olive trees because the stands are so dense, but has kept them from increasing and has helped maintain open grassy areas for migratory birds .

Treatments: Barker Ranch

Vegetation measurements were recorded each season before (pre-graze) the first species grazed, between (mid-graze) species grazing application, and after (post-graze) the second species grazed in the cow and goat pastures. The cows grazed pastures C1, C2, and C3 and then moved into G1, G2, and G3, while the goats did the opposite at the same time. The goal was to allow each animal species to utilize approximately 80% of the forage before moving to the next pasture, leaving 20% of the forage for the following species. Eighty-percent-forage utilization allowed for a measurable impact from the grazing applications. The goal of grazing complimentary livestock (i.e., species with different forage preferences) was to allow the first grazing species to selectively forage, thereby facilitating utilization of the remaining forage by the second species. This also allowed for a utilization analysis by species.

Adaptive management strategies were used as needed to adjust goat and cow herd size to shorten grazing duration in pastures. In spring 2004, a Boer cross and cashmere/Spanish goatherd of 44 adult does and 20 kids was used. After a 10-day adjustment period, an additional 41 adults and 55 kids were added to the herd for a total of 160 head. In the spring, the goats grazed the goat and cow-pastures in 18.5 days. Seven Black Angus cow-calf pairs (estimated body weights 545 kg and 135-160 kg, respectively) comprised the cattle herd and grazed the cow- and goat-pastures in 25 days. During the fall grazing season in 2004, 130 goats (58 adult and 72 kids) grazed all six pastures in 15 days. The 7 cow-calf pairs (estimated body weights 545 kg and 205-250 kg, respectively) grazed all six pastures in 17 days.

During the spring grazing season in 2005, a herd of 61 adult and 74 kid Boer cross and cashmere/Spanish and 7 Black Angus cow-calf pairs (estimated body weights 545 kg and 70-115 kg, respectively) grazed the plots. Four days after the start of the season, an additional 14 kids were added to the goatherd. Twenty-three days into the grazing season, the goatherd contained 131 adults and 150 kids. Ten days after the commencement of the spring grazing season, six additional cows and calves were added to the cattle herd. That spring, the goat-herd grazed all 6 pastures in 16 days, while the cow-calf pairs grazed them in 34 days. For the fall 2005 grazing season, 143 goats (65 adults and 78 kids) and eight cows (estimated body weight 545 kg) without calves composed the grazing herds. The animals were not given any supplemental feed. During the fall, the goats grazed thesix pastures in 10 days, while eight cows grazed the pastures in 13 days. Animals were allowed free access to water, and goats had free access to loose trace mineral salt with selenium.

Following the cessation of sampling each year, Russian olive in the herbicide-treated pastures was treated with 1% Arsenal® in 2004 and 2% Garlon-4® in 2005. Both applications included methylated seed oil (MSO) as a surfactant. Herbicide was applied with sprayers attached to all-terrain vehicles. Standard timing for Russian olive herbicide application at Barker is late July, August, or early September before leaves senesce. In the first year of the study, vegetation sampling took place before Arsenal® was applied to Russian olive in the herbicide-treated pastures due to scheduling conflicts. These three pastures were essentially treated as control pastures for the first year as the effects of the herbicide was not apparent until the vegetation was sampled the second year. Arsenal® was chosen in 2004 because it had been an effective method of chemical weed control at Barker in the past. Garlon-4® was substituted in 2005 because Arsenal® killed non-target vegetation surrounding Russian olive after the 2004 application. During each study year, herbicide-treated pastures were sampled at the beginning of the spring and fall grazing seasons only. These pastures were sampled once each season because vegetation was not utilized as it was in the grazed pastures. These end-of-season data were considered equivalent to the post-graze data and were used to compare the four pasture treatments with respect to density and biomass.

Three pastures were left untreated as controls. In 2003 these pastures had been treated with 1% Arsenal® and MSO and were grazed in early 2004. During each year of the study, the untreated pastures were sampled at the commencement of the spring and fall grazing seasons only with the reasoning the same as stated above for the herbicide pastures.

Density: Barker Ranch

At each plot, understory and over story density measurements were made for forbs and Russian olive. The density of each forb species (number of stems per 0.5 X 1 m plot) was recorded. Russian olive understory density measurements were made by recording the number of branches intersecting the 0.5 m2 plot. Russian olive over story density measurements were made by counting the number of living trees that intersected a 24.6 m2 imaginary cylinder (2.8-m dia. infinite height) created using a Hastings TEL-O-POLE® Measuring Stick, centered over the plot corner at the lower end (closer to start) of the transect line.

Aboveground Biomass: Barker Ranch

For each plot, a representative sample of the aboveground biomass for grasses, forbs, and Russian olive understory (defined as all branches and leaves from ground-level to approximately 6 m in height that intersected the three-dimensional area created by the 0.5 X 1 m plot) was hand clipped to the soil surface, air dried at 60ºC for 3 to five days, and weighed. Net weight, pasture number, replication, plot number, species name, and collection date were recorded on data sheets. Gross weight, as well as net leaf and woody material weight, were recorded for Russian olive understory clippings.

Vegetation Heights and Cover: Barker Ranch

For each plot, average height (cm) and cover of grasses, forbs, and Russian olive were determined and recorded. Changes in Russian olive understory cover and height and over story cover and minimum and maximum canopy heights were measured as a way to estimate the animals’ impact on the over story trees. Cover of litter, rock, and bare ground was visually estimated and recorded for each plot.

Research results and discussion:

Holistic Systems Approach

A holistic approach needs to look at the economic, social and environmental impacts of decisions. How can we design a weed management system that is economically, socially and environmentally sound? Is there a way to incorporate targeted livestock grazing as part of an integrated approach to manage weeds? Is there a need to change how we look at weed management? Is it possible for us to change our mindset from eradication of weeds to finding a way to utilizing them as a source of feed for other types of livestock and/or training cattle to eat certain weeds? In most publications on weed management there is a discussion on the economic impact of the weeds on the livestock industry due to loss of forage and the cost of control. The economic evaluation is typically based on cattle. There is a need to look at incorporating other types of livestock, such as goats and sheep, as a way to manage weeds and shift weed management from an annual cost to a source of income.

One critically important item learned was that a three year study is the minimum time needed to see changes in perennial plants that have an extensive root system, re-sprout rapidly or have seeds that stay viable in the soil for many years. In this study, we worked at two sites: (1) Barker Ranch where goats, cattle and herbicides were incorporated in the management of Russian olive in replicated plots and (2) Sprague Lake where goats and sheep were evaluated as a tool to manage noxious weeds on a publicly accessible learning site.

Figure 5: A portion of Barker Ranch showing the extensive stands of Russian olive

Barker Ranch

The management of Russian olive is an annual challenge at Barker Ranch. It is managed primarily for waterfowl habitat. The ranch is flood irrigated to provide food and cover for waterfowl. Cattle grazing is utilized as a tool to help manage the vegetation for waterfowl. The irrigation provides the necessary water to produce the habitat, but at the same time it is a constant source of Russian olive seeds. The large stands of Russian olive through out the ranch and surrounding area also provide a source of seed spread by birds and other animals. In order to keep the ranch from becoming a solid stand of Russian olive the open areas are spot sprayed at least twice annually.

To use livestock for vegetation management, it is important to understand animal preferences as well as the importance of the animal’s experience with eating different plants. A good source of information is the work Fred Provenza has done at Utah State University on animal behavior ( ). During this project we learned that the goats did not initially eat Russian olive. The goats had not been exposed to Russian olive prior to the study and there was two to three day period before the goats began to eat Russian olive. After that initial learning period the goats readily ate Russian olive for the rest of the three year study. The concept of a learning curve for animals to learn to eat new plants needs to be understood by people who want to use livestock as tools to help manage weeds.

The experiment at the Barker ranch was designed to look at the ways to incorporate the use of herbicides and livestock as a means to control Russian olive. Arsenal and Garlon 4 were the two herbicides used for control of Russian olive. Both were effective in controlling Russian olive. The application of Arsenal has to be done very carefully to avoid killing non-target vegetation. Barker Ranch has switched to Garlon 4 since it is a more selective herbicide.

The goats and cattle were grazed on the plots twice per year - once in June/July and a second time in August/September. After the initial 2-3 day learning curve for the goats to start utilizing Russian olive, the goats selectively browsed the Russian olive. The goats also selected for other weeds such as perennial pepperweed, and yellow starthistle. Including the cattle in the rotation made it possible to graze off the tall wheatgrass plants that removed all the old dead stalks. The grass production on the grazed plots increased significantly over the three years of the study.

With two treatments per year the goats were able to suppress the Russian olive. Goats were able to keep the stands open which allowed the cattle easier access and the reduced canopy cover allowed more sunlight to reach the grasses. If additional control was desired, the goats could be rotated back onto the Russian olive as soon as there was enough re-growth. Since Russian olive is a sprouter and due to the constant source of seed, in order for the goats to be effective the goats would need to be part of the long-term management plan for the ranch. The goats were most effective on Russian olive that was 6 feet or less in height. The nutritional quality of the diet for the goats was quite good. The leaves of the Russian olive were over 20% crude protein (CP) and the perennial pepperweed was over 15% CP.

Based on the results of this study, goats could be used as a tool to manage Russian olive, reducing the amount of herbicide used and providing another source of income for the ranch by selling the goats at the end of the year.

Challenges of using goats include:

The level of management required is higher for a herd of goats than it is for two temporary workers on ATVs spraying weeds

It is another enterprise for the ranch manager to coordinate with his limited time

Predator issues

Cost for new infrastructure – current fences will not hold goats

Large pastures – goats will need to herded or fenced on specific areas to obtain desired results

Benefits of using goats include:

Reduce the amount of herbicide used and reduce labor spent spraying weeds

The initial cost for new infrastructure will be offset by income from selling goats (i.e., converting weeds into a source of income)

Provide weed management services for other landowners in the area

Partner with goat producer to reduce expenses and time needed to manage goat enterprise.

Data were analyzed using the General Linear Model (GLM) procedure and Ryan-Einot-Gabriel-Welsch (REGWQ) multiple range test at P = 0.05 in SAS 9.0 (SAS 2006). The model included treatment, year, season, replication and their interactions.

Pasture Treatment

Russian olive Density

Density (trees/24.6 m2) was not equal (F = 3.08, 3 df, P = 0.0272, N = 480) (Table 3) among the 4 pasture treatment types at the end of Year 2. I rejected my hypothesis that Russian olive density was not affected by treatment. The REGWQ test showed goat-grazed-first (goat) pastures ( = 0.42) were different from both the untreated ( = 0.26) and herbicide ( = 0.25) pastures, but not from cow-grazed-first (cow) pastures ( = 0.30). Cow, untreated, and herbicide pastures were equal to one another.

Table 3: Summary comparing 4 pasture treatment types on Russian olive density at the end of Year 2 as determined by SAS General Linear Model (GLM) procedure and Ryan-Einot-Gabriel-Welsch (REGWQ) multiple range tests.

At the end of Year 2, using the post-graze means and herbicide and untreated “post-graze equivalent” means, densities increased by an average of 0.11, 0.05, and 0.03 trees m-2 in the cow, goat, and untreated pastures, respectively, while densities decreased by an average of 0.14 trees m-2 in the herbicide pastures (Table 4; Fig. 6). No difference among treatments occurred between years or seasons (Table 3).

Table 4: Average Russian olive densities (trees m-2) and aboveground biomass (g m-2) measured at the end of each grazing season in spring and fall of 2004 and 2005 at Barker Ranch near West Richland, WA.

Russian olive density increased overall in the goat and cow pastures by the end of the two-year study (Figure 6). This was unexpected as goats were observed stripping all leaves and much of the Russian olive tree bark in the grazed pastures. Cows were observed sampling and nibbling on leaves occasionally, although the leaves did not appear to be a significant component of their diet. In the cow pastures, densities remained approximately the same in spring and fall of 2004 and increased slightly in spring of 2005 (Figure 6). In the cow pastures, cows reduced the height of the tall bunchgrasses and forbs (data not included here) allowing the goats easier access to tree branches and exposing small saplings that the goats also ate. This system, along with the death of mature trees and the sprouting of saplings, feasibly kept density levels in the cow pastures roughly equal at the end of each season until fall 2005 when they increased. The most probable explanation for this increase is that not as many mature trees died that season due to natural causes or from being stripped of leaves and bark. It is not likely that the goats ignored saplings (the only supply of new tree density) as a food source, as vegetation was even more scarce at the end of the fall grazing season than the spring coupled with much drier and less palatable vegetation in fall 2005 overall.

Figure 6: Mean number of living Russian olive trees m-2 at the end of each season over a 2-year period.

Density increased in the goat pastures until the end of the spring 2005 grazing season. This trend is likely a result of goats grazing annual grasses, forbs, and easily accessible Russian olive branches first, and cows eating down the tall bunchgrasses and exposing saplings afterward. This could have been coupled with the possibility of new trees sprouting during the time the cows were in these pastures as well. The decrease in density following the spring 2005 grazing period is likely due to goats killing mature trees by stripping the leaves and bark with no subsequent re-sprouting as vegetation was much drier and appeared to come back less abundantly than in 2004. Interestingly, although densities differed between the cow and goat pastures at the end of spring 2004, they were similar at the end of the study. In addition, the same trend was true for biomass (see following section), which leads to the possible conclusion that grazing order had no effect on mean Russian olive density or biomass (see “Grazing Order” below).

The increase in density in the herbicide pastures from spring 2004 to fall 2004 was directly related to the fact that herbicide was applied that year following the fall 2004 sampling. In 2004, the herbicide pastures were essentially considered untreated pasture (as described in the methods section), and thus exhibited a similar rate of increase in density as the untreated pastures (Figure 6). A decrease in density in spring 2005 and further still in fall 2005 coincides with visual observations in spring 2005 of severe leaf necrosis on all trees treated with Arsenal® in fall 2004.

As expected, at the end of two-years into the study, density increased overall in the untreated pastures as they were not grazed nor chemically treated (Figure 6). However, a decrease in density occurred between fall 2004 and spring 2005, dropping from 0.33 to 0.16 trees m-2. It took a full season to return to 0.33 trees m-2 in these pastures. This temporary decrease in density may have been attributed to winter weather conditions.

Russian olive Aboveground Biomass

Biomass was different (F = 15.40, 3 df, P = <0.0001, N = 480) among the 4 pasture treatment types at the end of Year 2 (Table 5). This rejected the hypothesis that Russian olive aboveground biomass was not affected by treatment. The REGWQ test showed biomass was greater in the untreated pastures ( = 24.16) than goat ( = 0.03), cow ( = 0.98), and herbicide pastures ( = 6.04). Goat, cow, and herbicide pastures were not different from one another. At the end of Year 2, biomass increased by an average of 1.5 and 6.64 g m-2 in the cow and untreated pastures, respectively, while biomass decreased by an average of 0.48 and 5.62 g m-2 in the goat and herbicide pastures, respectively (Table 4). No differences occurred between years or seasons (Table 5).

Table 5: Summary comparing four pasture treatment types on Russian olive aboveground biomass at Barker Ranch, West Richland, WA as determined by SAS General Linear Model (GLM) procedure and Ryan-Einot-Gabriel-Welsch (REGWQ) multiple range tests.

Biomass in the cow pastures increased steadily, but very slightly (1.5 g m-2) from spring 2004 to fall 2005 (Figure 7). Post-graze measurements in these pastures, (i.e., following goats) showed that between the two species, almost all above ground leaf biomass was utilized. Visual observations and measurements by grazing order (see “Grazing Order” below) concluded that the cows sampled leaves occasionally, but the goats had the most significant impact on the Russian olive.

Figure 7: Mean Russian olive aboveground biomass (g m-2) at the end of each season over a 2-year period

Biomass decreased overall (0.48 g m-2) in the goat pastures (Figure 7), and varied only slightly over the two-year study period. Notably, biomass increased from fall 2004 to spring 2005 and then decreased at a similar rate from spring 2005 to fall 2005 to almost the same level as in fall 2004. This was also true for density in the goat pastures and likely for the same reason. Interestingly, biomass in both pastures was almost equal when measured at the end of every season; that is, grazing order did not affect biomass (see “Grazing Order” below).

Herbicide pasture biomass decreased overall by the end of the study as expected (Figure 7). Biomass increased from spring to fall 2004 due to timing of the Arsenal® application. Biomass decreased from fall 2004 to spring 2005, which paralleled the observed extreme leaf necrosis/leaf loss on treated trees. Between spring and fall 2005, biomass increased slightly, which coincided with visual observations, and was likely a result of one year passing since the 2004 Arsenal® application.

Biomass in the untreated pastures increased overall by the end of the study as expected. Biomass increased from spring to fall 2004, but then decreased from fall 2004 to spring 2005 before increasing again. Again, this temporary decrease in biomass may have been attributed to winter weather conditions.

Grazing Order

Russian olive Density

Density was similar between the cow and goat pastures regardless of grazing order as shown by the post-graze means (F = 3.05, 1 df, P = 0.0823, N = 240) (Table 5). The hypothesis that Russian olive density was not affected by grazing order proved to be true. Density between the two pastures was also similar when comparing the pre-graze means (F = 0.32, 1 df, P = 0.5746, N = 240), but was different at mid-graze (F = 6.15, 1 df, P = 0.0138, N = 240). Mean density increased after the cows grazed and decreased after the goats grazed in the cow pastures, while density decreased after the goats grazed and increased after the cows grazed in the goat pastures (Figure 8).

Figure 8: Mean number of living Russian olive trees m-2 before the grazing application of the first animal species, in the middle of the first and second species grazing application, and after the grazing application of the second species over a 2-year period.

Post-graze density did not change, regardless of grazing order between 2004 ( = 0.35) and 2005 ( = 0.36) or spring ( = 0.33) and fall ( = 0.37) (Table 5). Mid-graze density was different between 2004 ( = 0.43) and 2005 ( = 0.27) and spring ( = 0.29) and fall ( = 0.41). Pre-graze density was also different between 2004 ( = 0.35) and 2005 ( = 0.34), but was not different between spring ( = 0.26) and fall ( = 0.42).

Grazing order did not affect density; however, some trends by species may provide insight on the usefulness of goats and cows for Russian olive control. Density always decreased after goat grazing which means that the goats not only ate/killed very small trees, but were able to strip trees of leaves and bark to the point that they died (Figure 8). The thought behind grazing cows before goats was that cows might utilize forbs and grasses desirable to the goats and encourage them to eat the new Russian olive trees. However, after a two to three day learning curve, the goats acquired a taste for Russian olive and were able to cause a density decrease whether grazing before or after the cows. Density always increased following cow grazing (Figure 8). The cows ate small quantities of leaves, but not enough to kill trees. Additionally, small trees had time to sprout while cows were grazing, and cows reduced the height of many large groupings of bunchgrasses (up to 1.5-m high), exposing smaller trees. At first glance, it would appear that the cows stimulated tree growth, but more likely, new trees grew and/or were exposed during and after cow grazing, thus explaining the density increase.

Grazing order affected mid-graze density within both years as expected because, as noted above, goats killed small trees while cows removed competing bunchgrasses. Grazing order also affected pre- and mid-graze density within season in 2004 and 2005. Pre-graze differences were likely due to trees not being grazed by goats before spring 2004 (and more than a year had passed since trees had been treated with Arsenal® and subjected to cattle grazing at this study site) and being stripped of leaves and bark by goats before fall 2004. Additionally, trees had time to recover and re-grow leaves between fall 2004 and spring 2005 and were heavily grazed again before the pre-graze period in fall 2005. As expected, mid-graze effects by season paralleled mid-graze effects by treatment and year.

Russian olive Aboveground Biomass

Biomass was similar in the cow and goat pastures regardless of grazing order as shown by the post-graze means (F = 0.99, 1 df, P = 0.3204, N = 240) (Table 6). The investigator’s working hypothesis was that Russian olive aboveground biomass was not affected by grazing order. Biomass was also similar between the two pastures pre-graze (F = 0.67, 1 df, P = 0.4132, N = 240) and mid-graze (F = 1.90, 1 df, P= 0.1689, N = 240) (Table 6). In the cow pastures, biomass decreased at almost the same rate following cow and goat grazing (Figure 8). In the goat pastures, biomass decreased at a greater rate than in the cow pastures following goat grazing, and increased slightly following cow grazing, but decreased overall after both species grazed (Figure 9). Biomass was similar in the cow and goat pastures pre, mid-, and post-graze between years and seasons (Table 6).

Table 6: Summary comparing the individual and combined grazing effects of cattle and goats on Russian olive aboveground biomass as determined by SAS General Linear Model (GLM) procedure and Ryan-Einot-Gabriel-Welsch (REGWQ) multiple range tests.

Figure 9: Mean Russian olive aboveground biomass (g m-2) before the grazing

Although grazing order did not affect biomass, some trends emerged which may provide insight on effects of sustained multi-species grazing on Russian olive biomass. Biomass always decreased following goat grazing. When goats grazed first, biomass decreased sharply (Figure 9) because after a brief learning curve, they ate the Russian olive as readily, if not more readily, than the grasses and forbs. Biomass increased slightly when cows grazed following goats (Figure 9). Again, this is not to say cows stimulated leaf growth, but apparently new leaves appeared at a slightly faster rate than cow grazing diminished them. When cows grazed first, biomass also decreased, suggesting cows had some effect on biomass, but not nearly the impact of goats (Figure 9). When goats grazed after cows, biomass decreased, but not at as much as when goats grazed first (Figure 9). This is due to the fact that when goats grazed following cows, less Russian olive biomass was available.

Notation regarding loss of collaborating scientists and impact on Barker Ranch data collection and analysis

In Year 1 and Year 2 of the Barker Ranch study, the data were collected and analyzed by Amy Hummer who used this project as her Master’s thesis study. She completed her M.S. degree at the end of Year 2 (2005). Also during 2005, Michael Carpinelli, the collaborating scientist who designed the study, resigned his position with USDA/ARS in Burns, OR and accepted a position with the NRCS in Grants, NM. As a result, he was no longer active in this project. We hired another Natural Resource Sciences graduate student (Len Zeoli) to collect the Year 3 (2006 data). In mid-2006, the other collaborating scientist, Jim Dobrowolski, resigned his position with Washington State University (WSU) to become CSREES National Program Leader, Rangeland and Grasslands Resources in Washington DC. As a result, he was no longer active in this project. In August 2006, Skip Stonesifer resigned from the U.S. Department of Fish and Wildlife and accepted a position with the Bureau of Land Management in Wyoming. He has not been involved in the project since his departure. Then the question became, who is going to statistically analyze the Year 3 data and the overall effects of all three years of the project at the Barker Ranch? The project coordinator, Donald D. Nelson, recruited a fellow faculty member and statistician in the WSU Department Animal Sciences to attempt to do this. His efforts produced Table 7 that summarizes the grazing treatment effects on Russian olive.

Table 7: Averages for each treatment before grazing, when the animals were switched and after grazing.
All of this data was collected in the fall 2006 (August and September)

The amount of data here to base conclusions on is limited. No data was available for Controls or Herbicide Treated plots. Plots with Cows First had much taller Russian olive, 159 cm vs. 65 cm with greater cover, 52% vs.65% and greater density, 51 vs. 21 for Cows First and Goats First, respectively. In the Cows First plots there was little change in percent cover or density of Russian olive, but in the Goats First plots there was some decrease in percent cover and density, but an increase in height. Based on this data, Dr. Gaskins was unable to determine the overall effect of multi-species grazing on Russian olive for the three years of the project.

Sprague Lake

At Sprague Lake observations were made on how well the goats utilized the noxious weeds and the impact on the native plant community. Diffuse Knapweed, Dalmatian Toadflax, rush skeletonweed and Canada thistle were the main weeds on the site. The noxious weeds, except for the Canada thistle, were intermixed with the native grasses (bluebunch wheatgrass, Idaho fescue, sandberg bluegrass, basin wildrye), native forbs (arrowleaf balsamroot, yarrow, lupine) and shrubs (golden current, wild rose, green rabbitbrush, stiff sagebrush, Wyoming sagebrush, and hawthorn). The Canada thistle was on a site that was predominately reed canary grass.

The site is a mixture of very shallow, shallow and loamy ecological sites in the 9-12 inch rainfall zone. The noxious weeds were scattered through out the sites mainly in small patches. The goats readily ate diffuse knapweed, Dalmatian toadflax, rush skeletonweed, and Canada thistle. Over the three years of the study the following observations were noted:

Dalmatian toadflax disappeared from the shallow sites and were spindly on the loamy sites

Diffuse knapweed plants were reduced in size and two treatments prevented seed production

Rush skeletonweed was grazed by the goats but a change in plant vigor or density was not observed.

Canada thistle did not produce seed in the grazed plots after the first year of treatments. During a site visit in October of 2007 (treatments ended July 2006), it was noted that there were no Canada thistle plants producing seed within the grazed plot. The three years of treatment had reduced the vigor of the Canada thistle enough to have a multiple year affect on the plants.

The goat grazing reduced the vigor of the native shrubs

The goats were moved to the next plot when the noxious weeds had been heavily grazed, and as a result the utilization on the native grasses was kept low (generally less than 10%)

The native forbs such as yarrow and arrowleaf balsamroot were heavily utilized.

Goats and sheep with their preference for broadleaf plants can be used as a tool for managing noxious weeds on rangelands. The benefits of using goats and sheep as a tool for converting noxious weeds to income and reducing herbicide use has to be balanced with the challenges such as impact on native shrubs and forbs and cost associated with improving fencing, predator control, and increased labor for management. There are also the time considerations related to starting a new enterprise. There may be creative ways of dealing with some of these challenges, such as a cattle producer partnering with a goat or sheep producer for weed control.

Research conclusions:
Impact of Results/Outcomes

Barker Ranch: Treatment effects
After four grazing seasons over a three-year study period, grazing and herbicide treatments had an effect on Russian olive density and aboveground biomass. The results also showed that while grazing order had no overall effect on Russian olive density and biomass, identifiable trends by grazing species exist and may be important to future use of these animal species for prescriptive grazing.

The results show that biomass and density always decreased after goat grazing. Using repeated grazing applications over a number of years, goat grazing alone, or following cattle grazing, may be a feasible alternative method of weed control for small, isolated stands of Russian olive (Stannard et al. 2002; Zouhar 2005). Goats have been successfully used in past prescriptive grazing regimes on woody vegetation and on vegetation undesirable to other livestock (Provenza et al. 1983; Cox 2003), as was true in this study. Results also showed that biomass and density always increased, with one exception after cattle grazing. This is not to say that cattle stimulate leaf and tree growth, but that cattle do not impact Russian olive significantly enough to counter new growth. Essentially, net leaf and tree growth was, on average, greater following cattle grazing. Therefore, cattle grazing alone would not be recommended for control of Russian olive. However, cattle would be useful for grazing grasses and forbs in a Russian olive-infested pasture, encouraging goats to focus on the remaining vegetation, thus lessening the time it would take goats to impact Russian olive.

Barker Ranch: Management implications
Although the overall impact (averaged at the end of each season) of grazing did not significantly lessen Russian olive density or aboveground biomass, the immediate impact (measured immediately following grazing) may prove fruitful for future weed control. Using goats alone, or following cattle, may be a feasible alternative to chemical weed control in herbicide-restricted riparian areas for small, isolated stands. This may be possible if prescriptive grazing applications are repeated over several seasons, as four seasons of grazing applications in this study did not lessen density or biomass. For this type of grazing regime to be successful, land and/or animal managers must consider size, location, and age of the Russian olive stand and use appropriate herd size, adaptive management, and proper timing. For Russian olive, this would be before thee years of age, as this is the time when they begin to bear fruit and flowers (Collins 2002; Stannard et al. 2002; Miller 2003). When properly managed, goats can effectively defoliate target vegetation, thereby lessening seed production, reducing viability, and depleting nutrient reserves, thus deteriorating root systems (Madsen 2004). This is important because Russian olive trees have extensive root systems with “well developed laterals” (Zouhar 2005), however, over-grazing of desirable vegetation must always be avoided.

Undoubtedly the most effective control is prevention through awareness and education, as well as a team effort with adjacent landowners to stop any additional planting and to remove existing Russian olive. In addition, the prescriptive grazing regime must be properly managed, natural disturbances must remain uninhibited, and human disturbances minimized (Zouhar 2005).

Barker Ranch

Photographic comparison of treatments on Russian olive
The following series of photos will provide visual documentation on the effectiveness of using goats for Russian olive management. The photos will also be used to show how the plant community changed over the three years of the study that resulted from the use of goats and cattle.

The following series of photos are all taken at the same location over the three years of the study. This series of photos show how the goats reduced the canopy cover and the height of some of the Russian olive plants. Two treatments per year with the goats suppressed the Russian olive and allowed more sunlight to reach the grass in the understory.

Figure 9: Plot G2 June 2004 - Before any study treatments

Figure 10: Plot G2 June 2006 – After two years of treatment by goats and cows and before 2006 treatments

Figure 11: Plot G2 June 2006 – After the first goat treatment during 2006

Figure 12: Plot G2 April 2007 – Early spring after three years of treatments as the grasses were beginning to green up and the Russian olive was starting to leaf out.

Re-growth of Russian olive

Russian olive will re-sprout and re-grow quickly after it has be mowed, cut or browsed by goats. The two treatments per year completed in this study should be considered the minimum treatment level needed to suppress Russian olive.

The following series of photos illustrates the impact of the goats on the amount of re-growth on the Russian olive over the three years of the study. The number of days between the first and second treatment in 2004 was 61 days, 2005 was 57 days, and 2006 was 69 days.

In 2004 the amount of re-growth before the second treatment in September was similar to the growth prior to the start of treatments in June of 2004. The pictures of the re-growth in 2005 and 2006 show the impact of the goat browsing on the Russian olive. The picture taken in June 2006 also shows a significant increase in the grass production over that shown in June 2004.

Figure 13: Plot G2 June 2004 – Before any study

Figure 14: Plot G2 Sept 2004 – Re-growth Before second goat treatment

Figure 15: G2 June 2005 - Before 1st goat treatment

Figure 16: G2 August 15 2005 – Re-growth Before 2nd goat treatment

Figure 17: G2 June 29, 2006 – Before first goat

Figure 18: G2 Sept 12, 2006 – Re-growth before second goat treatment

The impact of goat browsing on the Russian olive is seen in two main ways: 1) eating the leaves and small branches of the plants that they can reach, as goats are bipedal and can reach vegetation that is 5-6 feet tall, and 2) eating the bark off of the stems of the plant. Goats will eat the bark of stems that range in diameter from ½ inch to 4 inches.

Figure 19: Goats at work on the Russian olive

Figure 20: Lower branches have been defoliated and are trying to reach the higher branches

Figure 21: G3 – Prior to first goat treatment in July 2006

Figure 22: G3 – After goat treatment in July 2006.Plants shorter than 5 feet are completely defoliated, taller plants have a browse line. Goats preferred the Russian olive to the tall wheatgrass

Figure 23: Barking of small and larger diameter
Russian olive branches

Figure 24: Another example of barking small and larger diameter Russian olive branches

Figure 25: Russian olive tree barked in
September 2006

Figure 26: April 2007 Dead stems as a result of barking by goats. Russian olive will re-sprout, but new growth is within easy reach of goats and quite palatable.

Figure 27: Re-growth from a Russian olive that had been previously grazed

Figure 28: New re-growth was completely removed by goats

There were other changes in the plant community as the result of goat and cattle grazing. In the photos below you can see that there was a significant change in the production as well as some changes in the amount of certain forbs such as dog fennel. The goats also utilized other weeds such as perennial pepperweed and yellow starthistle.

Figure 29: G3- June 2004 - Before any treatments

Figure 30: G3 June 2006 - Before 2006 treatments

Figure 31: G3 – July 4, 2006 Before treatments in 2006. The plant that is flowering is perennial pepperweed

Figure 32: G3 July 7, 2006 After first goat treatment in 2006. Goats selected for the pepperweed and Russian olive.

Summary of the overall effectiveness of the goats, as well as the impact of the goats and cattle, on the plant community:
The following changes were observed over the three years of the study:
· Increase in the number of stems of Russian olive in goat treated plots
· Reduced height and canopy cover of Russian olive in the goat treated plots
· Some change in the amount of forbs in goat treated plots (i.e., a reduction in dog fennel)
· Increase in grass production in grazed plots

The increase in the number of stems was due to the fact that Russian olive is a strong sprouter. When the goats barked the stems, the plant sent out new shoots increasing the number of stems. The new stems were within easy reach of the goats and quite palatable. If goats continued to be browsed on the site the number of stems would decline as the root system of the Russian olive was weakened by the continued defoliation by the goats. To gain better control of the Russian olive, the goats would need to be rotated on a continuous basis through the treated areas whenever the Russian olive had re-grown. The continuous rotation would increase the amount of stress placed on the plant versus two treatments per year as was done in this study. Three years is the very minimum length of study necessary to show a change in a perennial plant and the plant community due to an applied stress.

The barking of Russian olive stems and defoliation by the goats impacted the height and canopy cover of the Russian olive during the three-year study. The goats’ preference for forbs resulted in some changes to the plant community. As shown in the photos above, there appears to have been a reduction in the amount of dog fennel and the goats readily ate the perennial pepperweed in the plots. The grass production increased over the study period and was probably due to a number of factors such as, improved nutrient cycling due to removal of old grass stubble, manure provided by the cows and goats; increased sunlight due to reduction in Russian olive canopy.

If goats are to be used as the main tool for managing Russian olive, they need to part of a long-term management plan in order keep the sprouts and the new seedlings in check.

Figure 33: April 2007 - After 3 years of treatments with goats and cows

Figure 34: April 2007 - Untreated control plot immediately to the right of treated photo

Barker Ranch – Bulrush

Michael Crowder, the manager of the Barker Ranch, wanted to see if we could use the goats to help manage bulrush in some of the shallow wetlands. The bulrush was thick and over eight feet tall in some areas and was not providing good quality waterfowl habitat. The objective on the bulrush site was to create more open water for waterfowl for the fall and winter.

Figure 35: June 2004 Bulrush Before goats

Figure 36: June 2004 Bulrush after goats

Figure 37: June 2004 Bulrush before goats

Figure 38: June 2005 – Bulrush site maintained by cattle grazing only

Cattle have grazed the bulrush site for many years. The pasture is several hundred acres in size. The cattle had made a few trails through the tall bulrush but were not concentrated enough to graze the tall stands. The goats along with a few hair sheep were placed on the bulrush site in early June 2004. Since the water level could be controlled, a majority of the area did not have standing water so the goats and sheep would graze into the stands of tall bulrush. The animals were fenced in on areas between ½ to 1 acre in size and grazed on bulrush for about a month. During that time, three goats were observed limping and were later treated for foot rot. When grazing goats on wet sites, well-trimmed hooves will help to reduce the likelihood of animals getting foot rot.

The bulrush re-growth was grazed a second time in late August/early September with the goats and sheep. The bulrush that had not been previously grazed was turning brown. The palatability of the brown bulrush was low which made using the animals as a means of clearing the bulrush ineffective. After the second treatment the area was mowed with a tractor. The grazing by the goats and sheep made it possible to mow the area by reducing the volume of standing material and by opening up the stand, which allowed more sunlight to reach the ground and dry it out. In 2000, cattle readily grazed the bulrush re-growth. Since the combination of treatments in 2000 cattle grazing alone has been used to keep the bulrush short. The increased open water significantly increased the use of the area by waterfowl.

In 2005, the goats were put on another area with bulrush, but the treatment was not as effective primarily because of the difficulty in controlling the water level. The goats did not like going into the water deeper than 6-8 inches and in some areas the water was over 12 inches deep. The deep water also made it difficult to fence the bulrush into smaller pens that force the goats to focus on the bulrush. The control of water levels is critical to the successful use of goats and cattle on the bulrush.

Figure 39: Goats at work on the bulrush June 2004

Sprague Lake

At Sprague Lake, goats and a few hair sheep were grazed on 10 plots. The vegetation was bluebunch wheatgrass, Idaho fescue, basin wildrye, native forbs and shrubs. The noxious weeds on the site were diffuse knapweed, Dalmatian toadflax, Rush skeletonweed and Canada thistle.

Dalmatian toadflax

Figure 40: June 2, 2004 – Dalmatian toadflax
before treatments

Figure 41: June 3, 2004-Dalmatian toadflax after
grazing by goats and sheep

Figure 42: Late May 2006 – Dalmatian toadflax after 2 years treatment and before first treatment for 2006

Figure 43: Late May 2006 - Close-up of Dalmatian toadflax before third year of treatments

Figure 44: October 2007 – Same area as above photos. Dalmatian toadflax present, but plants spindly

Figure 45: Late May 2006 - After 1st treatment in 2006. Heavy utilization on toadflax; light utilization on bluebunch wheatgrass and Idaho fescue.

Figure 46: Late May 2005 - Idaho fescue and
Dalmatian toadflax - before treatment

Figure 47: Late May 2005 - After treatment

Figure 48: Late May 2006 - Dalmatian toadflax – After two years of treatment and before third year of treatment

Figure 49: October 2007 – No treatments

After two treatments per year for three years, the Dalmatian toadflax was removed from the shallow ecological sites and significantly reduced on the loamy sites. The Dalmatian toadflax on the loamy ecological sites with the deeper soils did re-grow after the second treatment in mid-July. The fall re-growth was adequate to sustain the plant. A third treatment in the fall would probably be needed to provide control of Dalmatian toadflax on sites with deeper soils.

Diffuse Knapweed

In this series of photos the post is next to the same knapweed plant for all the photos, which is also adjacent to a basin wildrye plant.

Figure 50: Diffuse knapweed – Before 2nd treatment July 2004.

Figure 51: Diffuse knapweed – After 2nd treatment July 2004, beginning to flower

Figure 52: May 2006-Before third year of treatment, Knapweed bolting

Figure 53: May 2006-After first treatment

The goats readily ate diffuse knapweed. Two treatments per year were necessary to prevent seed production on these drier sites. In areas with higher rainfall or soils with greater moisture holding capacity more than two treatments may be needed to prevent seed production. The grazing by the goats reduced the size of the knapweed plants and seed production. The seed-bank in the soil provided the source for new diffuse knapweed plants. If the knapweed was grazed during the bolting stage, the goats ate most of the plant (see second set of photos). If grazed when flowering (first set of photos), the knapweed is coarser and the goats will strip the leaves, seeds and branches but do not prefer the coarser stems. As the palatability of the knapweed changes, the goats may shift to utilizing more of the native grasses and the utilization on non-target plants would need to be monitored. Based on our observations goats and sheep can be an effective tool to manage diffuse knapweed on rangeland. The challenge with the knapweed is the soil seed bank where the seed can remain viable for over eight years.

Canada Thistle

Figure 54: July 2004 - Before second treatment

Figure 55: July 2006 - Before second treatment

Canada thistle
The plants on this site were Canada thistle, Reed canarygrass, large stem bulrush and stinging nettle. The goats ate the Canada thistle, Reed canarygrass and bulrush, but did not eat the stinging nettle. During the first year of treatments some of the Canada thistle did go to seed in the grazed plots. For the next two years, no Canada thistle plants flowered within the grazed area. The following set of photos was taken in October 2007. No grazing treatments were done in 2007. The Canada thistle was not flowering in the grazed plot so the grazing treatments had suppressed the Canada thistle. In order to prevent the Canada thistle from reoccupying the site to levels prior to goat grazing, the goat treatments would have to continue on a maintenance basis. The goats selected for the Canada thistle on this site, but the preference would change depending on the other plants present.

Figure 56: October 2007 – Untreated area

Figure 57: October 2007 – Treated for three years (2004-2006)

Rush skeletonweed

Figure 58: May 2005-Before goats late

Figure 59: May 2005-After goats late

Figure 60: Late May 2006-Rush skeletonweed and
sagebrush seedling before treatment

Figure 61: Late May 2006-Rush skeletonweed and sagebrush seedling after goats grazing

Rush skeletonweed

The goats ate the Rush skeletonweed and young sagebrush seedlings. Bluebunch wheatgrass received light utilization and the native forbs such as yarrow and arrowleaf balsamroot were heavy utilized.

The goats’ preference for shrubs that worked well for the Russian olive study was a challenge when using them on the Sprague Lake site. The native shrubs on the site were golden currant, wild rose, green rabbitbrush, hawthorn and greasewood. The goats selected for the shrubs and reduced their vigor, especially the currant, over the three years of the study. Using goats for control of weeds in areas with desirable shrubs needs to be carefully planned to reduce the negative impact on the shrubs. Timing (season of grazing, duration, and length of rest need to be considered); fencing (fencing the goats out of large areas of shrubs); and the use of other livestock, such as sheep, should be considered during the planning process.

Forage quality

Forage samples were taken of perennial pepperweed, prickly lettuce, prostrate knotweed, Bermuda grass, black medic, tall wheatgrass seedheads, Russian olive leaves and new growth from Russian olive. The samples were air dried prior to being sent to lab for nutritional analysis and that is the reason for the high dry matter percentage (Table 8).

Table 8: Nutritional analysis for summer 2004 forage samples from the Barker Ranch. All results are on a dry matter basis. DM = dry matter; NDF = neutral detergent fiber; ADF = acid detergent fiber; CP = crude protein.

The goats maintained or improved in body condition when they were on the Russian olive plots.

Bulrush observations

A second nutritional analysis was done on bulrush in June 2005. The sample was taken for comparison with 2004 data as well to obtain a true dry matter reading (Table 9).

Table 9: 2005 bulrush nutritional analysis. Sample was taken June 28, 2005.

The crude protein values for bulrush were very similar between the samples taken in 2004 and 2005, 11.75 and 11.8 respectively. The acid detergent fiber was higher in 2005 than in 2004, 36.6% and 30.6% respectively.

On the bulrush sites when the size of the paddocks were kept small and thus more frequent moves, the animals were able to maintain body condition. In areas where the water was deeper and the size of the paddock had to be increased, the length of time in the paddock also had to increase to have the same impact on the bulrush. The longer the animals stayed in the paddock the nutritional quality of the forage declined. In that situation some of the does with twins showed a slight decline in body condition. The following comments on the nutrient quality of bulrush were provided by Tom Platt, WSU Livestock Extension Agent in Davenport, Washington: “The nutritional analysis for bulrush looks adequate in protein and mineral content. The high level of fiber (ADF and NDF) will limit digestibility and intake, so you should monitor lactating does for loss of body condition. If they can't eat (and digest) enough calories to maintain their body condition, then they also will be short on other nutrients. If they can maintain their body condition, they should be ok. If the does lose too much condition and you need to keep them on the bulrush, then you will probably have to wean the kids, which will greatly reduce the does' nutrient requirements.”

Livestock management:Barker Ranch

Two treatments were completed on the Russian olive plots during 2004

Table 10: 2004 first grazing treatment on Russian olive plots at Barker Ranch.

The goats first ate the forbs in the plot (china lettuce and perennial pepperweed). They were also utilizing Bermuda grass, Kentucky bluegrass and some tall wheatgrass. It took 10 days for the goats to have a significant impact on the Russian olive in the first ¼ acre plot. As a result we decided to increase the number of goats from 64 head (44 does and 20 kids) to 160 head (85 mature does and 75 kids). When the goats were moved into the second plot, the first plant that most of the animals started eating was Russian olive. Goats were moved to the next plot when they had cleared all of the leaves off of the Russian olive plants within reach (about 6 feet high).

Table 11: 2004 second grazing treatment on Russian olive plots at Barker Ranch.

In the two months after the first treatment (refer to Table 2) on the Russian olive plots, the Russian olive had re-grown to levels similar to before first treatment, Bermuda grass had re-grown significantly, and tall wheatgrass had limited re-growth. 130 head of goats were turned into plot G1 on August 27. When the goats were initially turned into G1, they were observed eating a variety of plants, Bermuda grass, perennial pepperweed, seed heads of tall wheatgrass and new re-growth on Russian olive trees. On the second treatment they were held in the plot until a significant number of the stems had been barked. In order to get that level of impact on the Russian olive, the goats also utilized the re-growth of the Bermuda grass, tall wheatgrass and forbs in the plot. Since the goats were getting low on feed they were harder to keep in the plots. The goats did escape briefly from a couple of the plots. They were going under the fence so plastic fence posts were used to stake down the fence. One of the reasons that the goats were getting out is the charger was putting out around 3,000 volts, which is below the 5000 volts normally considered adequate to hold goats. The lower voltage seems not to be enough when the goats are being pushed to bark the Russian olive.

Refer to Table 12 for body condition scores (BCS) of goats for second treatment on Russian olive plots. The first date is three days after start of second treatment and the second date is five5 days after the end of the second treatment.

Table 12: Body condition (BCS) scores on selected goats grazing Russian olive plots at
the Barker Ranch site.

Cows were moved when utilization on the grasses was around 80%. The cattle were observed eating leaves off the Russian olive trees. It appeared to be a minor component of their diet. The cows seemed to be bothered by flies more than the goats. Goats were moved to the next plot when the Russian olive had been completely defoliated up to five to six feet.

Due to the location of the Russian olive plots water had to be hauled to the goats and cows. On average the 7 cows and calves utilized 92 gallons of water per day and the 130 head of goats utilized 54 gallons per day.

Table 13: 2005 First Treatment Russian Olive Plots at Barker Ranch

Table 14: 2005 second treatment of Russian olive plots at Barker Ranch

Table 15: 2006 First treatment of Russian olive lots at Barker Ranch

Table 16: 2006 2nd Treatment Russian Olive Plots at Barker Ranch

Goat Treatments

Table 17: Comparison of changes in carrying capacity during the three years of the study

Figure 62: Graphic depiction of changes in goat carrying capacity during the three years of the study

The changes in carrying capacity (Table 17 and Figure 62) were calculated by multiplying the pounds of goats in each plot by the number of days the goats where in each plot. The number provides a way to compare how the carrying capacity of the goats changed from 2004 to 2006. By looking at the change in carrying capacity of the goats during the first and second treatments over the three years of the project, some generalizations can be made about the treatments. Comparing the carrying capacity during the first treatment across the 3 years, the carrying capacity increased by about 20% from 2004 to 2005 and there was less than a one percent increase from 2005 to 2006. The increase in carrying capacity from 2004 to 2005 is probably due mainly to the Russian olive re-sprouting after BEING stressed by the browsing and barking done by the goats. The fact that there was no increase in carrying capacity for the goats in year three for the first treatment may indicate that the goat browsing was having enough impact on the Russian olive to reduce the growth the following year.

The significant decrease in carrying capacity during the second treatment (25% from 2004 to 2005, 42.5% from 2004 to 2006) indicates that the goat browsing was having a significant impact on the ability of the Russian olive to re-grow between treatments. If this trend continued, the goat carrying capacity during the first treatment would also begin to decline. It appears that three years is the minimum length of a study necessary to begin to determine the changes that are occurring in a plant community. A four to five year study would provide better insights on the changes in a plant community caused by a stress such as targeted grazing.

In three years, the goats were able to reduce the height of the Russian olive plants and to keep the stands more open that allowed for increased forage production for the cattle. The goats also developed a preference for the Russian olive. Russian olive is good forage for the goats as shown by the nutritional analysis and the does with kids maintained body condition during the study.

Cow Treatments

Comparing the changes in carrying capacity of the cows over the three years of the project. The numbers are calculated by multiplying the number of cows by the number of days the cows were in the plots (Table 18 and Figure 62).

Table 18: Changes in cow carrying capacity during the three years of the study

Figure 62: Graphic depiction of cow carrying capacity during the three years of the study

The cow carrying capacity during the first treatment from 2004 to 2005 increased by over 230% and from 2004 to 2006 the increase was over 470%. There are a number of factors contributing to this increase: increased spring rainfall, removal of old plant material from the tall wheatgrass improving plant vigor and palatability, improved nutrient cycling – increase in nutrients from manure and breakdown of the increased litter on the ground, and the reduction in over story canopy of the Russian olive due to goat browsing. The cow carrying capacity during the second treatment declined in 2005 and 2006. The decline could be due to a number of reasons: 1) the first treatment in 2004 and 2005 grazing ended on July 14 and 15 respectively and in 2006 grazing did not end until August 1. With the later end date of grazing in 2006 there is a good chance there was less moisture available for re-growth, and 2) the percent utilization in 2005 and 2006 was higher during the first treatments than in 2004. The old plant material in the tall wheatgrass plants was mostly removed in 2004 and that increased the palatability of the tall wheatgrass in the subsequent years, resulting in increased utilization.

Sprague Lake
The project-planning group decided that the Sprague Lake site was going to be a learning site (i.e., an observational study) rather than an experimental layout like the one at the Barker Ranch. The land located on the shores of Sprague Lake is owned by the Washington State Department of Fish and Wildlife. The main target plants at this site are diffuse knapweed, Dalmatian toadflax, Rush skeletonweed and annual forbs. The plots have a mixture of native perennial grasses, native forbs, native shrubs, annual grasses, annual forbs and scattered patches of noxious weeds. The site was grazed with goats and sheep twice in order to prevent seed production on the target plants. For a plot map refer to Figure 4 and for the plot sizes refer to Table 1.

Goats and sheep grazed the site together. The first treatment started on May 28 (refer to Table 6). The annual grass weeds, Bromus tectorum and medusa head were headed out and the animals did not utilize them. The animals did utilize the annual mustards as well as diffuse knapweed and Dalmatian toadflax. The goats did browse two of the native shrubs, currant and wild rose, with limited use of sagebrush (rigid and Wyoming). The utilization on the annual grasses was low, less than 20%.

Table 19: 2004 first grazing treatments at the Sprague Lake learning site. A total of 250
head of animals on site composed of 207 goats and 43 hair sheep

The second treatment started on July 26. By that time some of toadflax plants had flowered and formed mature seed (refer to Table 19). The diffuse knapweed had re-grown, but had not flowered. The Canadian thistle in plot 6 was flowering and some had set seed. In plot 4 the Rush skeletonweed was flowering and some plants had set seed. There was only limited re-growth on the annual forbs. It appears that the second treatment was one to two weeks late. The animals ate the leaves and flowers off the Dalmatian toadflax, and diffuse knapweed. The Rush skeletonweed received moderate use.

Table 20: 2004 second grazing treatments at the Sprague Lake learning site.

Table 21: First goat grazing of Sprague Lake plots

Table 22: 2005 second goat grazing of Sprague Lake plots

Table 23: 2006 first goat grazing treatment of Sprague Lake plots.

Table 24: 2006 second goat grazing treatment of Sprague Lake plots.

Participation Summary

Research Outcomes

No research outcomes

Education and Outreach

Participation Summary:

Education and outreach methods and analyses:


5/11/04: News release written and widely distributed by Andrea Mann (Coordinator, Big Bend RC & D) describing the SARE Implementing Noxious Weed Control Through Multi-Species Grazing project.

9/04: A moist soil management tour occurred in September 2004 at the Barker Ranch (a private 2000 acre duck club) near West Richland, Washington. Barker Ranch cooperator and manager, Michael Crowder, hosted the tour. The tour theme revolved around the remarkable progress in restoring uplands and promoting the growth of moist soil plants such as smartweed and millet, using livestock, cultivation, herbicides and water manipulation as management tools.


6/5/05: A press release entitled, Goats and Cattle Used for Invasive Vegetation Control, was widely distributed to promote the field tour at the Barker Ranch scheduled for 8/16/05.

8/16/05: Field Day at Barker Ranch--Results from the second year of the three-year study on multi-species grazing using cattle and goats to control Russian olive, bulrush, perennial pepper weed and other invasive plants were highlighted during a field tour at the Barker Ranch that was attended by 25 people.

8/25/05: A press release entitled, Big Bend RC&D Hosts Successful SARE Tour, was widely distributed

Articles were also submitted to Washington NRCS Current Developments publication and the SARE section of the National NRCS News website.


1/15/06: Press Release describing the 2005 end of year status of the project sent to the regional media.

2/15/06: Craig Madsen presented two posters at the 59th Annual Meeting of the Society for Range Management in Vancouver, BC.

4/12/06: Press Release sent to the regional media and the Washington NRCS Current Developments publication containing a notice of the Sustainable Rural Enterprise Conference and Tour scheduled for May 23-24, 2006 in Ritzville, WA.

5/1/06: WSU media article sent to the regional media regarding status of project.

5/23/06: Project Conference, “Sustainable Rural Enterprises”, Ritzville, WA co-sponsored by the Adams County Economic Development, Mid-Columbia Chapter of the Society of Range Management, Big Bend Resource Conservation and Development Council, Washington State University Extension and the Adams County Cattlemen’s Association. The keynote speaker was An Peischel, Tennessee State University Extension Goat and Small Ruminant Specialist, who discussed goat management and grazing behavior in relation to invasive vegetation management, creating niche markets and enterprise development. Forty people attended the conference.

5/24/06: In conjunction with the ‘Sustainable Rural Enterprises” conference, a field day was conducted to look at the vegetation management learning site located adjacent to the Washington State Fish and Wildlife boat launch area on the south shore of Sprague Lake. The tour included a review of invasive vegetation management using goats in wetland areas. Approximately 30 people who attended the previous day’s conference attended this field day.

6/5/06: Press Release sent to the regional media and the Washington NRCS Current Developments publication regarding the outcome of the conference and tour.

11/2/06: Craig Madsen made a presentation on this project at the Washington State Weed Association 56th Annual Weed Conference in Yakima, WA. Estimate that 250 people were at the presentation.

Additional project outreach occurred at the Barker Ranch as reported by Michael Crowder, Resident Manager:

Wetland Restoration class from WSU Tri-Cities Campus, Summer 2006

Cattlemen who rent grazing on Barker Ranch and ranch workers visited and inquired about project several times during the project.

Barker Ranch neighbors visited and inquired about project at various times during the project.

Refuge, wetland managers and duck club owners visited the Barker Ranch several times during the project.

Twenty-nine co-owners of the Barker Ranch visited and inquired about project many times during the project.


5/8/07: Multi-Species Grazing Conference, WSU Tri-Cities Campus, Richland, WA. Topics covered by nationally recognized goat and cattle grazers included: (1) management and marketing of meat goats, (2) what you can accomplish with goats, (3) grazing and browsing principles, (4) fencing, water and mineral needs, (5) use of guard dogs, (6) contract vegetation management; fee grazing, (7) range monitoring and (8) creating wildlife habitat. Fifteen people attended this conference.

5/9/07: Field Day at Barker Ranch, W. Richland, WA. Topics included on-the-ground demonstration of: (1) how to build and troubleshoot electric fences, (2) plant preferences of goats, (3) body condition scoring of goats, (4) famacha training (i.e., a visual internal parasite diagnosis technique) and (5) a range monitoring technique and browsing planning. Fifteen people attended the field day.

July 2007: Front page article in the Western Farmer Stockman entitled “Munching Machines” and a second article on page 5 entitled “Pro Shares Her Fencing Tips” by T.J. Burnham who attended the May 9, 2007 Field Day at the Barker Ranch.

March/April 2008: Article published in issue of In Practice publication of Holistic Management International entitled “Creating Open Water Wildlife Habitat—Using Multi-Species Grazing for Bulrush Management” written by Craig Madsen and Michael Crowder.

Education and Outreach Outcomes

Recommendations for education and outreach:

Areas needing additional study

Because no published information currently exists on using prescriptive grazing for controlling Russian olive, many possibilities exist for future studies in this area. For example, future research using single-species-grazing, common-use grazing (multiple species grazing simultaneously), or a study similar to this using sheep in place of cattle would increase our knowledge base. In this study, forb and grass utilization by species was recorded, but was not part of the analyses as the focus was impact on Russian olive biomass and density. In the future, it would be useful to plot forb and grass species utilization vs. Russian olive utilization by animal species to determine if important trends exist there, too. Additionally, research on combinations of grazing, herbicide, and mechanical removal is needed. Finally, supplementary research using prescriptive grazing on Russian olive stands of various ages would reveal the most effective treatment time.

If this study were to be repeated in the future, two changes to the design may be advantageous. First, it may be beneficial to include at least two, possibly three, data collection periods in the untreated and herbicide-treated pastures. This way these pastures would have measurements corresponding to the pre- and post, if not, pre-, mid-, and post-graze values in the grazed pastures. This would have allowed the measurement of any changes in density and biomass in un-grazed pastures instead of being limited to an end-of-season difference. Next, it would have been beneficial to have applied Arsenal® to the trees the first year before sampling so that the herbicide-treated pastures would be considered as such the first year. Although, as stated above, this was not the case, and herbicide was applied after the Russian olive was sampled due to a conflict in sampling and application timing. Additionally, Arsenal® killed not only the Russian olive trees, but all of the vegetation surrounding the trees; therefore the recommendation would be to use Garlon-4® to avoid killing non-target vegetation.

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.