More info for the terms: competition, cool-season, cover, density, dough stage, fire frequency, fire management, fire regime, forb, forbs, formation, frequency, fuel, invasive species, litter, natural, nonnative species, prescribed fire, presence, resistance, restoration, severity, shrub, shrubs, succession, wildfire, xeric
Impacts: The impacts of cheatgrass invasion vary with plant community and degree of cheatgrass dominance. At low densities, cheatgrass may simply be a part of the understory. At high densities, cheatgrass dominance can lead to complete community type conversions from perennial bunchgrass, sagebrush, salt-desert shrub, or pinyon-juniper communities to cheatgrass monocultures. The presence and dominance of cheatgrass affects many aspects of community structure, process, and function including diversity of plant and animal species, disturbance regimes, succession to other undesirable nonnative plants, nutrient cycling, and soil attributes. These changes may require substantial human intervention to convert to more desirable ecosystems.
Impacts on native plants: Cheatgrass is very persistent across a range of habitats [93,112,115], and can displace both rare and common plant species, thus reducing the number and genetic diversity of native plants in invaded communities. These changes can result from direct competition, from increased fire frequency, and possibly from indirect effects of changes in plant litter dynamics, nutrient cycling and soil ecology [377]. Soil water depletion is one of the principal mechanisms that allows cheatgrass to successfully compete with perennial grasses [138,281] and may negatively impact root growth of native species [280], especially during the establishment of perennial grass seedlings. Cheatgrass has been shown to deplete soil moisture and reduce growth of natives such as Idaho fescue [316], bluebunch wheatgrass [3], green rabbitbrush, and needle-and-thread grass [281]. It has also been observed to impede the establishment of native seedlings such as big sagebrush, green rabbitbrush, antelope bitterbrush [132,144,476], Stansbury cliffrose (Purshia mexicana var. stansburiana) [352], and several perennial herbaceous species [373]. Wallace and Nelson [447] suggest that cheatgrass infestations may contribute to shrub die-off in western rangelands. Introduction of cheatgrass and heavy grazing of native perennials have irrevocably altered the understories of Oregon white oak woodlands in the eastern Cascades [2]. A review by Mosley and others [307] suggests that plant biomass production on cheatgrass-infested rangeland is substantially less, and much more variable from year to year, than on rangeland dominated by native vegetation.
Cheatgrass can supply reasonably good forage and ground cover when desirable perennials are removed, although production is unpredictable and unreliable [59]. For details on cheatgrass' impacts on forage availability see Importance to Livestock and Wildlife. It has been suggested that in the absence of burning or heavy grazing, cheatgrass can help prevent erosion, promote water infiltration, and increase soil organic matter as a result of litter accumulation on areas where perennial cover has been reduced [1,210,406,454]. On the other hand, cheatgrass litter provides the fine fuel that can lead to wildfires, which may ultimately lead to plant community type conversions, subsequent impacts on wildlife, and the proliferation of other nonnative plants.
On the Snake River Plain, as in many areas throughout the Columbia River Basin and the Great Basin, the introduction of cheatgrass and other annual grasses, such as medusahead, has altered fuel loads and fuel distribution, which in turn has changed fire frequency, intensity, severity, timing, and extent [78,239,340,406,461,476,484,487,492]. For more information on fire regime changes in cheatgrass dominated ecosystems, see Fire Ecology. These fire regime changes, coupled with the impacts of domestic livestock grazing, have greatly modified plant recruitment, species composition and distribution, and the physiognomy and functioning of many vegetation types within the Intermountain West [71,340,365,461]. Several plant communities that are not adapted to the frequent fires that cheatgrass infestations engender have become increasingly susceptible to loss by wildfire. Examples include old growth pinyon-juniper woodlands [295], vast areas of sagebrush steppe (e.g. [239,340,461,484]), extensive stands of antelope bitterbrush in south central Idaho, central Utah, Oregon, and southern California [300], and salt-desert shrublands that often have high numbers of endemic plant and animal species [59,71].
Cheatgrass alters successional trajectories of postfire plant communities by interfering with native seedling establishment [3,132,144,281,316,352,373,406,476], by competing with established perennials for resources [138,280,281], and by shortening the interval between fires [48,261,406,455,461,487]. The ecological consequences of repeated burning include reduced species diversity, as the proportion of annuals in the community increases at the expense of other life forms [340,461]. In the desert shrublands of North America, individual fires may not have serious lasting effects, but recurrent fires may lead to resource homogenization that can complicate restoration efforts. More research is needed to determine how widespread this process is and to develop management tools to mitigate its effects [71].
The vitality and integrity of several native, fire-sensitive communities has been so reduced by repeated wildfires that extensive areas of shrub and perennial grass communities have been converted to annual grasslands [287,461,483,484,487,492]. Knick [239] reports rates of loss of big sagebrush and native salt-desert shrub communities in the Snake River Birds of Prey National Conservation Area in southwestern Idaho. In 1979, 250,000 acres (100,000 ha), or 51% of the total area, was comprised of big sagebrush and native salt-desert shrub communities. Because of fires and other disturbance, only 145,000 acres (58,000 ha) of shrubland remained in 1994, and another 30,000 acres (12,000 ha) of shrublands burned by 1998. Shrubland losses of this magnitude are typical of many sagebrush ecosystems in lower elevations and in xeric climates throughout the Columbia River Basin and Great Basin [71,239,340,461,492]. Monsen and Shaw [304] report "cheatgrass invasion has completely transformed ecological communities on millions of acres in the sagebrush grasslands of North America."
Despite the trend of improving range conditions on Intermountain rangelands attributed to improved grazing management [487], the dominance and distribution of cheatgrass has continued to increase in various rangeland ecosystems of the Intermountain area. This continuing increase in cheatgrass is attributed to grazing management systems that favor cheatgrass establishment [477] and to the increasing role of wildfire [48,71,239,340,461,475,492]. Some of the native plant and animal species in ecosystems that are now prone to widespread wildfires due to the presence of cheatgrass are at considerable risk of going extinct at the population level: locally or even regionally. This implies a threat to the existence of large, integrated ecosystems that have existed since the Pleistocene in the relatively arid lands between the Sierra Nevada and Rocky Mountains. The results could be the conversion of these native ecosystems to annual grasslands lacking not only the native vertebrates but also those invertebrates and biologic soil crusts that are involved in the energy flow, water cycling, and nutrient balance of the ecosystem [48].
Impacts on wildlife: Vegetation type conversion can affect wildlife ranging from herbivores to carnivores and reduce overall biodiversity [71]. While fires at low to moderate severities on sites with good productivity, deep soils, and good native plant components can benefit some wildlife habitats, high-severity wildfire in sagebrush-annual grass rangelands is considered detrimental to most wildlife species, as it promotes dominance of nonnative annual grasses and high fire frequency [418]. Large-scale change in structure of plant communities of this type can cause reductions in suitable habitat and concomitant shifts in the abundance of some species of birds [383] including Brewer's and sage sparrows and sage-grouse [94,239]. Areas in which sagebrush has been removed may not be used by nesting or brood-rearing sage-grouse until sagebrush re-establishes [97,234] (see sage-grouse).
Cheatgrass fires can adversely affect wildlife habitat by eliminating important forage species such as antelope bitterbrush on elk and mule deer winter range in Lassen County, northeastern California, and in northwestern Nevada [92,439]. The habitable winter range in this area is relatively small, so when fire promotes cheatgrass and reduces scattered foothill clumps of antelope bitterbrush and sagebrush, browse, thermal cover, and hiding cover are reduced [252]. Additionally, the length of time that cheatgrass is green and actively growing is shorter than that of native perennial vegetation. Thus compared to native species, cheatgrass dominance reduces the green-feed period for foraging animals [307].
Conversion of sagebrush habitat to cheatgrass is associated with a decline in densities of black-tailed jackrabbits [241], Townsend's ground squirrels [472], and other small mammals [164,176]. Two factors may limit small mammal populations after fire: 1) the loss of shrub cover may result in increased predation, and 2) thick stands of cheatgrass may impede small mammal movements which may affect breeding success and population size [176]. Similarly, dense cheatgrass can blockade newly hatched ducklings from making the vital trek from upland nest to lowland water [252]. Because cheatgrass-dominated communities support fewer small mammals than shrub-dominated communities, predator species such as the gopher snake, coyote, badger, and raptorial birds are also affected by large-scale losses of shrub habitat [161,241]. Loss of small mammals results in an increasingly unstable prey base for raptors in the Snake River Birds of Prey Area [161,472] and elsewhere. Kochert and others [241] have also documented that golden eagles avoid previously burned areas, and that golden eagle fledging success declines as the extent of burned area increase in their territory. Slow-moving fauna such as desert tortoises are sometimes killed in rapidly moving fires such as those that burn in cheatgrass [492]. The effective management of many wildlife species can depend on the control of invasive plants like cheatgrass and in the maintenance of appropriate FIRE REGIMES [71].
Other nonnative invasive species: Rangelands dominated by cheatgrass may also be susceptible to establishment of other nonnative, invasive annual grasses such as interrupted windgrass, corn brome, little lovegrass, poverty grass, and ventenata that are already present in the Pacific Northwest and are capable of invading and naturalizing in cheatgrass-infested areas [318]. The concept of a cheatgrass stand being a community closed to the establishment of perennial species [373] is challenged when one considers nonnative perennial and annual species that have established and persisted in cheatgrass stands such as forage kochia [90,271,305,331,476], squarrose knapweed [374], spotted knapweed, diffuse knapweed, rush skeletonweed, leafy spurge, common St. Johnswort, Dalmatian toadflax [187], yellow starthistle [187,306], and medusahead [187,189,196,295]. Medusahead has invaded and replaced dense stands of cheatgrass [189,295] over large areas in California, Idaho, Oregon, and Washington during the past 40 years [196]. Medusahead herbage is less palatable to ungulates, thus reducing grazing capacity where it replaces cheatgrass [187], and its seeds are not digestible by upland game birds, which are large consumers of cheatgrass seeds [381]. Evans and Young [141] suggest that cheatgrass enhances seed germination of several nonnative species in desert shrublands because of improved water availability associated with cheatgrass litter.
Impacts on soil resource: On ranges in which sagebrush and other native species have been eliminated by recurrent fires, cultivation, or grazing and now support nearly pure stands of annual grasses or weeds, soil losses can be severe [60]. A review by Upadhyaya and others [438] suggests that cheatgrass fires may leave land vulnerable to soil erosion because cheatgrass burns so completely. Cheatgrass litter may affect the rate and quality of nutrient cycling differently than native species, although more research is needed to understand these dynamics [62]. Nutrients may cycle faster in systems dominated by cheatgrass because its fine above- and belowground plant material can decompose quickly and because infested areas burn frequently, releasing minerals rapidly [326]. MacDonald [260] suggests that because cheatgrass does not catch and hold snow like a diverse perennial stand of vegetation, the site becomes drier. Cheatgrass dominance and associated fires also reduce populations, diversity, and recovery of biological soil crusts, which affect nutrient cycling, water infiltration, and potential soil erosion. When annual nonnatives dominate the plant community, the makeup of biological soil crusts changes, and their reestablishment is impaired [41].
The impacts of cheatgrass on ecosystems that it now dominates are such that the goal of managing vegetation to reflect conditions thought to have been present before European settlement may be impossible. Even the concept that preservation of native plant communities will prevent, eliminate, or control cheatgrass is not supported by evidence from sites in Utah [167]; in other areas where cheatgrass has displaced native bunchgrasses and shrubs in Great Basin; and where it continues to dominate sites in southeastern Washington, almost to the total exclusion of native grasses [326]. Dominance of cheatgrass and medusahead will not be changed simply by removing cattle or reducing their numbers [325]. The decision to reclassify vegetation types and to manage for cheatgrass instead of native perennial grasses may be considered for areas with sparse stands of desirable perennials [487] that are dominated instead by cheatgrass, halogeton, and/or medusahead. In such situations neither complete protection nor conservative grazing can restore a desirable cover within a reasonable period because a seed source of desirable species is lacking and competition from the undesirable plants is severe [60].
Control: Because cheatgrass is very persistent once it becomes established, eradication of large infestations is not usually a reasonable goal. The extent to which cheatgrass dominates a plant community greatly determines the appropriate suppression strategy [274,307], and cheatgrass response to management options is very site specific [189,478]. Some current research on control and management considerations for cheatgrass was presented at the Cheatgrass Awareness Conference in February 2003. Summaries and presentations are available online (Cheatgrass Awareness Conference).
Care must be taken with methods employed to control cheatgrass so that any void left by cheatgrass removal is not filled with another nonnative invasive species that may be even less desirable. A thorough site reconnaissance and evaluation is recommended before initiating any form of plant control.
Monitoring is an important part of weed management [218], as are early detection and local eradication of new infestations [497]. Estimates of cheatgrass coverage should be designed to determine how dominant cheatgrass is in an area compared to other vegetation. Areas should be monitored every spring, and relative cover and boundaries of any infestation should be recorded. Special attention should be paid to roadsides and other disturbed areas where cheatgrass is commonly found. If an infestation is found, the location should be recorded and monitored to measure the rate in which the infestation is spreading. Determining the relative cover of cheatgrass can be difficult due to the fluctuations in cover between years in accordance with weather variation [84]. Tueller [433] discusses the use of remote sensing to extract information about burns and annual vegetation patterns in the sagebrush grass ecosystems of the Great Basin relative to cheatgrass dominance.
Information regarding the extent of cheatgrass dominance in comparison to native species can be used to decide appropriate management strategies. Strategies may range from protection of areas not yet invaded, to reclassification of areas dominated by cheatgrass and management of these areas as cheatgrass rangelands. Intermediate to these states are areas with varying densities of cheatgrass, many having the potential to be converted to annual grasslands after fire. A review by McIver and Starr [274] suggests that "restoration" requires not only the reestablishment of historical plant community structures and species compositions, but also of the processes needed to sustain these into the future. It may be difficult to predict the outcome of control efforts or changes in disturbance regimes at any given site because native plant species respond differently to disturbance events [270,274].
Once cheatgrass is established, complete protection from grazing or other disturbances will not usually reduce cheatgrass abundance. Protection from all grazing can in fact increase the chance for fire and cause subsequent increases in cheatgrass dominance [461,478]. A mountain big sagebrush/bluebunch wheatgrass site in southwestern Idaho that was invaded by cheatgrass and subsequently protected from further disturbance continued to be dominated by cheatgrass for decades. Perennial grasses recovered slowly in protected areas, requiring more than 45 years to increase in cover from about 1.4 to 6.7 % [298].
Effective control of cheatgrass requires 1) eliminating live plants, 2) preventing seed formation, and 3) controlling seed germination and emerging seedlings [300]. In plant communities where cheatgrass is present but herbaceous perennials remain abundant, cheatgrass control measures should include the needs of the perennial plants. Control without replacement by desirable perennials will likely result in the reestablishment of cheatgrass or some other undesirable species [307]. In order to maintain dominance on a site, cheatgrass must produce a viable seed crop each year. If not, perennial plants can pre-empt the site. On the other hand, only a few perennial grass seedlings need to establish each year to maintain a perennial stand if mortality in established plants is not excessive [189]. Grazing management to favor native perennial grasses is doomed to failure unless there are sufficient perennial grasses in the ecosystem to outcompete cheatgrass. It is inappropriate to manage cheatgrass ranges as if they were perennial grass-dominated ranges [476].
Large areas that are mostly devoid of perennials and have fire-free intervals of 5 or fewer years have probably crossed a threshold, and the cheatgrass community probably represents a relatively stable "steady state" [251,307], such as exists on many depleted sites within Wyoming big sagebrush habitat types of the Snake River Plain and other portions of the Columbia River Basin [307]. For such sites only aggressive methods may have a chance of restoring more desirable sagebrush steppe communities [251,379,440]. Some authors suggest that it may be best to reclassify these communities as annual grasslands, and manage them accordingly [440,487].
Prevention: The most important aspect of invasive species management is prevention [191,388]. Whether or not cheatgrass establishment can be prevented in arid and semiarid rangelands is uncertain. It has been suggested, however, that maintaining an adequate cover of native plant species [14,59,112] and biological soil crusts [41] can render some communities more resistant to cheatgrass invasion. Managers must be aware that cheatgrass is especially prevalent in the early stages of fire succession or following other disturbances when shrubs, trees, perennial grasses, and other invasive plants are removed [34,34,122,139,145,210,273,291,295,382,388,396], and that cheatgrass density can increase dramatically 2 to 3 years after disturbance.
Proper grazing management may be an important tool in preventing or delaying further encroachments of cheatgrass into perennial vegetation [59,60,175,440].
Management practices that maximize the amount of organic debris left onsite and minimize the period of time between vegetation removal and reestablishment can limit the establishment of invasive plant species [400].
Integrated management: Once established, cheatgrass can rarely be controlled or eradicated with a single method, and most researchers and managers recommend combining physical, biological, chemical, and cultural control methods in some fashion [492]. Individual control methods (e.g. disking, herbicides, and prescribed fire) have limited application for control of cheatgrass on rangeland, and are more effective when used as site preparation for revegetation with desirable species [307].
Controlling live cheatgrass plants and the existing seed bank requires a combination of treatments conducted over a 1- to 2-year period. For example, mature plants can be killed before seed dispersal in spring by tilling or burning, with fall tillage or herbicide application as a follow-up treatment to eliminate any new seedlings. Artificial seeding of desired species is then conducted in the late fall or early winter [300]. Another strategy is to use prescribed fire in autumn to prepare sites for seeding the following spring [78,307,362]. Disking, herbicides, or prescribed livestock grazing can then be used in late spring to reduce vigor and seed production of the cheatgrass plants that establish after the fire. Seeding the site can be delayed until after the disking, herbicide, or prescribed grazing treatment, or the site can be broadcast-seeded immediately before grazing treatment, in order to use livestock to trample in the seed [307]. Another effective approach combines summer burning, when the majority of cheatgrass seeds were still held in the inflorescences, followed by fall herbicide application after cheatgrass seedlings have emerged [180]. It is unclear how long reductions of cheatgrass may last. Harris and Goebel [189] suggest that burning alone is usually not enough to destroy seeds, but can be combined with mechanical or chemical treatment to improve effectiveness.
Sheley and others [389] provide examples of successional weed management systems that integrate various control techniques to direct successional processes, resulting in different successional patterns and usefulness to range managers. The resultant plant community is dependent on the specific weed management system and the plant community prior to weed management [388,389].
On the Lawrence Memorial Grassland Preserve in Oregon, combinations of prescribed burning, herbicide applications (glyphosate), and mowing treatments were used to prevent cheatgrass and medusahead seed maturation. Preliminary results indicate that glyphosate treatment, and summer prescribed burning followed by mowing the next year, were equally effective at reducing invasive annual grass cover [349]. A multi-state, interdisciplinary research project has been proposed to examine integrated restoration strategies on western rangelands [323].
Physical/mechanical: Mixed results are reported for controlling cheatgrass with physical methods such as hand pulling, cutting or mowing, and disking or tilling. Tillage is often cited as an effective control method when combined with other methods; however, such intensive treatments are not usually appropriate for natural areas or wildlands and often lead to establishment of other undesirable plants. Regardless of which method is used, the cheatgrass seed bank must be depleted to get effective control [84].
In small areas, hand pulling can effectively control cheatgrass [365].
Cutting or mowing is not a recommended control method for cheatgrass unless it can be repeated several times per year, for several years. Plants that are cut before seed ripening can generate new culms and produce seeds at the cut height. Plants that are cut after seed ripening will still leave viable seeds [154,349]. Because seeds may be viable as early as the dough stage, before any purple coloration appears, there is only a short period after inflorescences appear that the plants can be mowed without danger of dispersing viable seeds. Therefore there is no single phenological stage that can be mowed which will assure complete kill or absence of seeds [207]. Cutting cheatgrass plants along trails before seeds are ripe and throughout the season can reduce spread of the invasion. Cutting cheatgrass plants for 3 consecutive years reduced densely populated areas at the Northeast Preserves in Oregon [365]. Repeated mowing (every 3 weeks) during the spring and summer was as effective at controlling cheatgrass seed production as was an application of glyphosate, though it was very labor intensive [349].
A review by Mosley and others [307] suggests that tillage and other mechanical control methods applied alone are often ineffective for controlling cheatgrass on rangeland. To be effective, tillage must be 4 to 6 inches (10-15 cm) deep to bury cheatgrass seeds in the soil and prevent them from germinating. More than 1 tillage treatment is usually needed, because the tillage equipment does not consistently cut deeply enough to bury the seed. To effectively suppress cheatgrass, tillage must be followed by sowing of desirable plants [307]. Tillage is not usually appropriate in wildlands and rangelands since it can damage important desirable species, increase erosion, alter soil structure, and expose soil for rapid invasion of cheatgrass and other invasive plants.
Fire: See Fire as a control agent for cheatgrass.
Biological: No insects or fungi have been approved by the USDA for use as biological control agents for cheatgrass. Cheatgrass is often infected with a head smut fungus (Ustilago bulleta) that may reduce seed yield when severe. Meyer and others [288] are exploring whether using this fungus may be feasible for biocontrol of cheatgrass. Some research has been conducted on the potential of pink snow mold (Fusarium nivale) as a control agent. In addition to these fungi, over 20 diseases of cheatgrass have been reported [492]. Kennedy [230] summarizes developmental research on the use of soil bacteria from the genus Pseudomonas as biocontrol agents for cheatgrass. Reviews by Carpenter and Murray [84] and Mosely and others [307] suggest that these microbes colonize the seeds and roots of cheatgrass and produce a toxin that appears to inhibit cell elongation in the growing plant. The toxin may also inhibit cheatgrass germination and seedling vigor. One particular strain, P. fluorescens D7, inhibits several biotypes of cheatgrass from throughout the world. This strain of bacteria also attacks other species in the Bromus genus.
Livestock grazing can reduce cheatgrass cover [45,116,343] and can be purposely manipulated to control cheatgrass [112,308,440], although some authors recommend against it (e.g. [137]). This tool is probably best suited to localized areas, either for protecting existing stands of perennial plants from fire, or for aiding the artificial seeding of severely depleted sites [308].
To prevent seed production, Mosely [308] recommends that cheatgrass plants be grazed before they turn purple in color. At least 2 defoliations are needed in the spring of each year, for a minimum of 2 consecutive years, to control cheatgrass [308]. Clipping was used experimentally to simulate grazing for cheatgrass control. Late spring clipping had the largest reduction in both density and biomass of cheatgrass compared to fall and early spring clipping or clipping in early spring only [414]. On salt-desert shrubland in western Utah, cheatgrass cover remained the same over 50 years on ungrazed range. It increased only very slightly on range grazed by domestic sheep in the fall, while spring grazing tended to increase cover of cheatgrass in general [460].
Prescribed domestic sheep grazing can be used to suppress cheatgrass density, growth, seed production, and mulch accumulations [111,112,116,308]. This is more easily accomplished in areas devoid of desirable perennials than in areas where remnant desirable perennials are present and need to be protected [308]. Where desirable perennials are present, grazing intensity should be light enough so that desirable grasses have a residual stubble height of 3 inches (8 cm) or more [308]. Grazing dried cheatgrass during winter dormancy will help reduce mulch accumulations and enhance seedling establishment of perennials. Grazing intensities in winter can be moderately heavy without damaging desirable plants, if soils are dry and firm [307]. Grazing in late autumn, when herbaceous plants are dormant, can reduce sagebrush density and enhance the ability of perennial herbs to compete with cheatgrass and sagebrush [308]. Encouraging reestablishment of native plants improves the effectiveness of grazing as a control method [492]. Where cheatgrass is a problem in bluebunch wheatgrass communities, properly timed grazing (before the boot stage of bluebunch wheatgrass, which coincides with shoot and flower development of cheatgrass) for a short duration is a helpful control technique [293]. Mixtures of crested wheatgrass and cheatgrass soon become dominated by crested wheatgrass if grazing on the area is well managed [406]. Prescribed domestic sheep grazing can also be combined with prescribed fires, herbicides, disking, and broadcast seeding [308].
In mixed stands with desirable perennials, livestock grazing can be directed to 1) reduce cheatgrass competition by concentrating grazing of cheatgrass during the dough seed stage, providing perennials have the opportunity to complete their life cycle, or 2) focus grazing on the needs of perennials while mostly ignoring cheatgrass [440]. Daubenmire [112] noted that cheatgrass did not dominate under heavy grazing pressure during spring, when livestock feed on the plants so heavily that no seeds matured. In this case, scattered plants may escape grazing, and when animals are removed, cheatgrass quickly dominates the area. Excessive grazing in the early spring, year after year, weakens the native cool-season perennial grasses, thereby providing habitat for cheatgrass to spread, but it also reduces abundance of cheatgrass. "The fact that excessive spring grazing both enhances the presence and biologically suppresses the abundance of cheatgrass is one of the most misunderstood aspects of the biology of this grass" [478].
Sanders [379] suggests that annual rangelands receiving 14 or more inches (356 mm) of annual precipitation can be converted to perennial rangeland through grazing management, providing there are sufficient perennial plants present as a seed source. Even in this precipitation zone, however, conversion seldom occurs without some means of reducing the annual grass competition (herbicides, fire, heavy spring grazing). Sanders adds that there is little evidence that annual rangelands receiving less than 12 inches (300 mm) of annual precipitation can be converted through grazing management alone [379].
Fire-free intervals can be lengthened by using prescribed domestic sheep grazing to disrupt fine-fuel continuity and reduce fine-fuel loads [308]. Grazed firelines should be at least 250 feet (75 m) wide [307,469].
Chemical: Several herbicides have been used alone or in combination to provide effective control of cheatgrass, including quizalofop, fluazifop-p-butyl, sethoxydim, paraquat, glyphosate, imazapic, sulfometuron methyl, and atrazine [84]. Carpenter and Murray [84] give more details and contacts in various states with specific chemical control information. The Weed Control Methods Handbook has detailed information on some of these chemicals as well [432].
Chemical control can be used effectively when desired perennial plants are still abundant. One year of chemical application will only temporarily thin the cheatgrass population and may actually increase cheatgrass seed production. Treatment must be repeated from 2 to 5 years consecutively. Several newer herbicides, especially imazapic, are being tested for selective control of cheatgrass in perennial broadleaf seedling stands [492]. Summaries of some current research on the use of imazapic for cheatgrass control are available (Cheatgrass Awareness Conference).
Sulfometuron methyl has been used to control cheatgrass on rangelands for 1 to 2 years and thus improve the success of rehabilitation projects. Projects in Nevada and Idaho comparing sulfometuron methyl treatment with burning and disking are summarized by Pellant and others [338]. There are various restrictions associated with sulfometuron methyl use. Ongoing and available studies include rates, season [56], effects of sulfometuron methyl on survival and reproduction of native plants and biological soil crusts, and effects on seeded perennial plants [56,338]. Spring treatments were considerably less effective; and damage to native bunchgrasses (Sandberg bluegrass and bottlebrush squirreltail) occurred with high application rates in spring [387]. Other studies explore the effects of combining chemical control with prescribed burning and seeding of desirable plants [131,132,143,444,463,480].
Some cheatgrass biotypes have evolved resistance to herbicides including atrazine, simazine, chlorotoluron, primisulfuron-methyl, and sulfosulfuron; and may be cross-resistant to other photosystem II inhibitors, urea, amides, and/or acetolactase inhibitors. More information on herbicide resistance is available at WeedScience.org.
Chemical control of other invasive species can result in increases in cheatgrass. For example, the use of 2,4-D alone to control spotted knapweed resulted in an increase in biomass and cover of cheatgrass, while the use of 2,4-D and grazing by domestic sheep resulted in lower biomass and cover of cheatgrass compared to treatment with 2,4-D alone [249]. Similarly, application of clopyralid to control yellow starthistle can increase cheatgrass cover [122].
Cultural: Cheatgrass is not competitive with established perennials, particularly grasses; therefore, biological suppression with desirable perennials may be an effective method of controlling cheatgrass [492]. Because cheatgrass seedlings can outcompete seedlings of most perennial species and prevent their establishment, successful establishment of desirable perennials 1st requires the depletion or removal of cheatgrass plants and seed bank [138,307,373]. Robertson and Pearse [373] suggest that openings in cheatgrass communities caused by severe drought, fire, or grazing can provide conditions favorable for artificial reseeding with desirable species, provided sufficient soil remains. Evans [138], however, found that as few as 4 cheatgrass seedlings per square foot (43/m2) can inhibit survival of crested wheatgrass seedlings, implying that is necessary to achieve near-perfect cheatgrass control for successful perennial seedings [140,147,373]. Techniques for accomplishing this are presented in the above sections. Haferkamp and others [180] examine several different seedbed preparations and seeding methods on sites in Washington and Oregon. Effectiveness of these techniques varies widely, depending on terrain, local weather conditions, treatment timing relative to cheatgrass development, and recovery of annuals from residual seed reserves left near the soil surface where they are capable of germinating [387]. For information regarding revegetation after prescribed burning or wildfire see Fire Management Considerations.
Young [477] summarizes the history of revegetation programs in the West, comparing species, sites, and ecosystems. Monsen and McArthur [302] provide a discussion of the history of revegetation projects in the Intermountain West, including constraints and benefits of using native and nonnative species. Most attempts to artificially seed big sagebrush ranges with native grasses prior to 1945 were largely unsuccessful. The introduction of crested wheatgrass, a nonnative, drought-tolerant perennial grass, made large-scale seeding of sagebrush ranges possible. From 1945 to 1965, several million acres of sagebrush rangeland were seeded to crested wheatgrass in the Intermountain area. Cheatgrass distribution was limited by extreme grazing pressure at that time, so these seedings took place in seedbeds with limited cheatgrass competition. The universal grazing management systems that followed ultimately favored cheatgrass establishment (see [477] for details). Subsequent increases in the distribution and density of cheatgrass now precludes the successful establishment of crested wheatgrass and other perennials without prior cheatgrass control on most of these sites [477].
Crested wheatgrass is a reliable and persistent grass for low-elevation, drought-prone areas of the Intermountain West. It establishes rapidly even under relatively dry conditions and tends to persist for many years [225,300,328]. It is best adapted to the sagebrush grassland portions of the Intermountain range, but not to salt-desert ranges [373]. While studies have indicated that cheatgrass is a superior competitor to crested wheatgrass [3,23,26,158,190,208], they have also indicated that crested wheatgrass competes more effectively with cheatgrass than does bluebunch wheatgrass [3,136,190].
In addition to crested wheatgrass, several other nonnative, herbaceous forage species have been used for revegetation and cultural control efforts, with limited to mixed success [67,71,117,328,487]. A long-term study to identify the species best suited to seeding semiarid rangeland sites in northeastern Washington indicated that hard fescue (Festuca trachyphylla) was the most aggressive competitor, and cultivars of crested wheatgrass provided the highest yields. Several other grasses were also tested and compared, and results are presented by Harris and Dobrowolski [188]. A review by Mosley and others [307] provides the following summary: 'Hycrest' crested wheatgrass, 'Sodar' streambank wheatgrass (Elymus lanceolatus), and 'Luna' intermediate wheatgrass are most suitable on unirrigated sites where annual precipitation is 12 to 16 inches (300-400 mm). 'Regar' meadow bromegrass (Bromus biebersteinii) and orchardgrass (Dactylis glomerata) are suitable for reseeding most cheatgrass-infested rangeland where annual precipitation exceeds 16 inches (400 mm), although orchardgrass can also be invasive. Drought-resistant species such as 'Hycrest' crested wheatgrass and Siberian wheatgrass (Agropyron fragile) are most suitable in areas receiving less than 12 inches (300 mm) of annual precipitation, although successful establishment on such dry sites is especially difficult [307]. Native grasses are more difficult to establish, less persistent, and less productive than nonnative grasses on sites with less than 10 to 12 inches (254-300 mm) of annual precipitation [24,300]. Mountain rye (Secale montanum) was seeded on reclaimed coal mined lands in southeastern Montana to suppress cheatgrass and Japanese brome (Bromus japonicus). It was effective for short-term suppression (2 years), but not for longer-term control [10,11].
Some managers are experimenting with another nonnative plant, forage kochia, for use on severely altered semiarid rangelands that support monocultures of cheatgrass and/or tumblemustard and are subject to repeated fires in Nevada [90], Utah [271,331], and Idaho [305]. It is argued that forage kochia provides mule deer with highly digestible protein that crested wheatgrass lacks; is adapted to very dry sites (5 to 27 inches (127-956 mm) precipitation); is more easily established and has higher growth rates than antelope bitterbrush; is persistent; and is one of the few perennial species that can compete with cheatgrass. It is also suggested that forage kochia can create an open window for the return of native plants by decreasing fire frequency (see Fuel management). In a seeding operation on a degraded range in Nevada, managers planted a mixture of forage kochia, Wyoming big sagebrush, fourwing saltbush, and a variety of bunchgrasses. By the 4th growing season, Wyoming big sagebrush, thickspike wheatgrass (Elymus lanceolatus), and other native bunchgrasses and forbs were becoming more visible [90]. On study sites in Utah, forage kochia established and persisted in cheatgrass stands along roads; and it established, persisted, and spread over 437 yards (400 m) throughout a cheatgrass-infested, abandoned farm site. It also spread about 50 feet (15 m) into the edge of a native Wyoming big sagebrush-pinyon-juniper community adjacent to the farm site [271]. Clements and others [90] suggest that forage kochia does not appear to be invasive in the sagebrush/bunchgrass region of the Great Basin, and may provide an opportunity for native plants to reestablish by decreasing fire frequency in cheatgrass-dominated stands. "We have observed big sagebrush seedlings in stands of forage kochia that had first suppressed the cheatgrass” [90]. Nonetheless, caution is warranted and long-term monitoring recommended when replacing nonnative invasive plants with other highly competitive nonnative species.
The practice of seeding nonnative perennial grasses such as crested wheatgrass to contain nonnative annuals, stabilize disturbances, and provide forage for livestock was initially successful, but resulting stands do not provide the structure, functions, resilience, or values of native communities any more than monocultures of cheatgrass [239,253,304]. These plant conversions have negatively impacted wildlife habitat in big sagebrush communities by reducing the diversity of shrubs and forbs that was present in native communities. Big sagebrush communities throughout the western U.S. have seen rapid and continuing declines in populations of small mammals, raptors, sage-grouse, songbirds, other vertebrates, and invertebrates, particularly in the last 20 to 40 years [304]. Monsen and Shaw [304] review the current status of rehabilitation efforts in big sagebrush communities and suggest that failure to establish native species in revegetation plantings is due in part to inexperience and inappropriate planting techniques, and may also be related to species, subspecies, and ecotypes of plants used.
It is commonly thought that restoration of rangeland plant communities using native grasses is difficult [477]. However, natural recovery of some cheatgrass sites by native species has been observed (e.g. [300,456]). Sandberg bluegrass, bottlebrush squirreltail, Thurber needlegrass, western wheatgrass, and thickspike wheatgrass have recolonized and gained dominance on sites once infested by cheatgrass [300]. The natural recovery process certainly suggests that these same species can be effectively used in artificial seedings if seeds are readily available and planting requirements are better understood. Some progress has been made using selected native species to restore cheatgrass-infested sites, and carefully planned seedings in sagebrush benchlands and pinyon-juniper woodlands can often restore the native understory and prevent further spread of cheatgrass. Additional information on the use of native plant materials for sagebrush steppe restoration is given by Jones [221].
As is true for nonnative perennials, establishing native perennials is most effective when cheatgrass is removed or its density substantially reduced [189,303,359,492]. Replacement species for cheatgrass must be able to germinate and produce a vigorous seedling with 1 to 2 days of wet soil, and the seedling must be morphologically adapted to survive drought periods of 10 or more days following a 2-day wet period in order to be competitive with cheatgrass [160]. Hardegree [182] suggests seed priming to increase germination rates of native perennial grasses relative to cheatgrass. Research by Chambers [85] indicates the need to create soil surface features that trap and retain seeds and provide favorable conditions for seedling establishment. Monsen and Shaw [299,301,304] provide detailed discussions of restoration of big sagebrush communities in the Intermountain West using native plants such as big sagebrush, bottlebrush squirreltail, thickspike wheatgrass, and Sandberg bluegrass.
A review by Ott and others [328,329] suggests that native perennial grasses such as bluebunch wheatgrass and bottlebrush squirreltail can have a competitive impact on cheatgrass growth and reproduction, particularly as mature plants. Suppression of cheatgrass invasion following fire has been observed in areas where these grasses or other fire-tolerant native plants were abundant in the prefire community and readily recovered following fire [328,363,456]. Cheatgrass has high relative shoot and root growth rates compared with bottlebrush squirreltail and bluebunch wheatgrass [23]. Detailed treatment of competition between cheatgrass and bluebunch wheatgrass is given by Harris [186]. Bluebunch wheatgrass seedlings may have a better chance of establishing among bottlebrush squirreltail plants than among cheatgrass or medusahead [196].
Bottlebrush squirreltail is a promising candidate for assisting ecological restoration of rangelands dominated by medusahead and cheatgrass [196,220]. It has successfully colonized stands of annual grasses in southern Idaho [23,198], and can establish naturally in rangelands infested by cheatgrass [38,300]. Bottlebrush squirreltail appears particularly well adapted to postfire establishment and competition with cheatgrass [62]. Early fall seeding of bottlebrush squirreltail in cheatgrass-infested rangelands could be effective in future restoration projects [38].
Idaho fescue is a native perennial bunchgrass that can persist in areas with high densities of cheatgrass. Idaho fescue plants from cheatgrass-infested areas tend to be more effective competitors than plants from pristine areas. This suggests that continual competition with cheatgrass selects for a more competitive group of Idaho fescue plants. Hence, seeds from Idaho fescue plants in cheatgrass-infested areas may prove to be more successful at suppressing cheatgrass when planted elsewhere [316,317].
Desert needlegrass (Achnatherum speciosum) is a potentially valuable native species for use in restoration seedings in the more arid portions of the Great Basin [359]. Kitchen [233] summarizes information on perennial forb life-history strategies on semiarid rangelands and their implications for revegetation. Meyer [284] discusses establishment of big sagebrush in community restoration projects in the Intermountain West. She suggests 1st establishing early seral species such as bottlebrush squirreltail, since they appear better able to compete with cheatgrass and other annuals, and then seeding big sagebrush along with late-seral understory species. Shaw and Haferkamp [386] discuss using spiny hopsage for revegetation in a Wyoming big sagebrush community in southern Idaho. Belnap and Sharpe [42] present a study using native grasses to revegetate a bunchgrass community in Utah. Winterfat was successfully established after fire in Idaho [303]. Stevens [404] cites evidence and provides guidelines for establishing plants such as big sagebrush, rubber rabbitbrush, fourwing saltbush, antelope bitterbrush, Palmer's penstemon (Penstemon palmeri), western yarrow (Achillea millefolium), globemallow (Sphaeralcea spp.), and other native and some nonnative species by interseeding into existing stands of cheatgrass.
Jones and others [219] conducted a study to determine if it is possible to restore the moss component of biological soil crusts to a nonnative annual grassland devoid of perennial mosses. Moss recovery or initiation was low over the entire experiment, regardless of the site preparation methods used (sulfometuron methyl, burning, and tilling), all of which were effective for controlling the cheatgrass [219].
In the absence of vesicular-arbuscular mycorrhizae (VAM), mycorrhizal bunchgrasses (e.g. bluebunch wheatgrass and Idaho fescue) are at a competitive disadvantage compared to nonmycorrhizal species (e.g. Russian-thistle) or facultatively mycorrhizal species (e.g. cheatgrass) in capturing limited soil resources. When a plant community is predominantly nonmycorrhizal, VAM fungi decline in abundance and remain depressed until host plants reestablish. Therefore, reestablishment of VAM-dependent plant species may not occur until both plant propagules and the spores of VAM fungi occupy the same site at the same time [405,464]. Rehabilitation of disturbed sites may need to include techniques designed to stimulate the reestablishment of VAM symbiosis [464].
