Fire Effects Information System (FEIS)
FEIS Home Page

Research Project Summary: Nonnative annual grass fuels and fire in California's Mojave Desert

Howard, Janet L., comp. 2006. Research Project Summary: Nonnative annual grass fuels and fire in California's Mojave Desert. In: Fire Effects Information System, [Online]. U.S. Department of Agriculture, Forest Service, Rocky Mountain Research Station, Fire Sciences Laboratory (Producer). Available: [ ].

Unless otherwise indicated, the information in this Research Project Summary comes from the following papers:

Brooks, Matthew Lamar. 1998. Ecology of a biological invasion: Alien annual plants in the Mojave Desert. Riverside, CA: University of California. 186 p. Dissertation.

Brooks, Matthew L. 1999. Alien annual grasses and fire in the Mojave Desert. Madrono. 46(1): 13-19.

See the Appendix.

There were 2 phases to this Research Project Summary: a fuels study and a prescribed fire study. For the fuels study, 34 study sites were located in western (n=10), central (n=16), and southern (n=8) regions of the Mojave Desert in southeastern California. Sites were >50 m from dirt roads and >2 km from paved roads [4].

For the prescribed fire study, three 2.25-ha study sites were located in the western (35o14'30"N, 117o51'15"W), central (35o07'30"N, 117o07'45"W), and southern (34o41'30"N, 117o57'30"W) desert regions. For mapped locations of the 34 fuel and 3 fire study sites, see [4].

Soils were gravelly to cobbly, well-drained, sandy loams over granitic parent materials. Average annual precipitation across fuel and fire sites was approximately 150 mm, with 87% occurring as rainfall from October through April. Winter rains before the spring and summer studies were 200% of normal. Mean minimum and maximum temperatures from the nearest weather station are 19 °C and 34 °C in August and 0 °C and 7 °C in December, respectively [4]. All sites were located on alluvial bajadas, none were subject to livestock grazing, and none had burned for at least 25 years [6].

Study sites were within a creosotebush-burrobush community. Winterfat, burrobush, and water jacket were common shrub associates [6]. Historical stand structure of the community was widely spaced shrubs with little annual plant biomass in shrub interspaces [15]. Stand structure and species composition of the community differed from historical conditions, with annual species occupying many of the interspaces between shrubs. Nonnative red brome was the most common dominant annual (see Fuel load results). The nonnative grasses Arabian schismus and common Mediterranean grass dominated some study sites and codominated with red brome on others. Native bristly fiddleneck, western tansymustard, and lacey phacelia were dominant forbs. Annuals with minor coverage in the community included nonnative Chilean chess, cheatgrass, small sixweeks grass, redstem stork's bill, and native sixweeks grass [4].

Study sites were on the following vegetation classifications:

FRES30 Desert shrub [10]
K042 Creosotebush-bursage [13]
SRM 211 Creosote bush scrub [17]

Prescribed burning was conducted in August. Annual plants had senesced months before. With the exception of dry nonnative annual grasses, most dead annuals had lost their form and crumbled. Creosotebush, burrobush, and other shrubs were probably dispersing their seeds at the time of the fires [1,10]. Prefire annual plant inventories, done to assess prefire fine fuel loads, were 1st done from 12 April 1995 through 11 May 1995, when most annual species had reached peak biomass, were presumably setting fruit, and had not senesced. Second prefire annual plant inventories were done from 7 July 1995 through 14 July 1995, when annual plants were already dead [4,10].

western Mojave site: Summer/mixed severity
central Mojave site: Summer/low severity
southern Mojave site: Summer/mixed severity

The fire management objective of this study was to compare the role of nonnative annual grasses and other annual plants in facilitating fire spread in the Mojave Desert. Methods employed were measuring cover and frequency of fine fuels produced by different annual plant species and describing how flames spread through these fuels during prescribed summer fires. Frequency was used to evaluate fine fuel continuity; cover evaluated fine fuel amounts [4].

Fuels study
On each of the 34 fuel study locations, 25 frequency and cover measurements per location were taken beneath large (>50-cm diameter) shrub canopies and on interspaces between large shrubs (>1 m between shrubs) [4,6].

Fuel load results: Spring and summer frequencies were highest for nonnative annual brome and schismus species, intermediate for redstem stork's bill, and lowest for most native annuals. Spring and summer cover was highest for the bromes, with schismus providing 2nd-highest coverage. Native forbs provided intermediate cover, and cover was relatively low for redstem stork's bill compared to native forbs. Native annuals other than bristly fiddleneck, western tansymustard, and lacey phacelia provided least cover. Prefire, growing-season coverage and frequency are given below. Data are means; Bromus (red brome, Chilean chess, and cheatgrass), Schismus (Arabian schismus and common Mediterranean grass), and Vulpia (sixweeks grass and small sixweeks grass) spp. data are pooled [4]:

  Cover (%) Frequency (%)
Nonnative annuals
Bromus* spp. 70a  15a
Schismus** spp. 45b  37b
redstem stork's bill 23c  17c
Native-nonnative annual mix
Vulpia*** spp. 23c  1d
Native annuals
bristly fiddleneck 44b  2d
western tansymustard 30bc  1d
lacey phacelia 44b  1d
other native annuals 10d  4cd
different letters indicate significant differences (p<0.05)
*90% red brome and 10% mixed Chilean grass-cheatgrass
**Schismus spp. could not be distinguished in the field
***60% sixweeks grass and 40% small sixweeks grass

Prescribed fire study: Fires were conducted on 16, 22, and 24 August 1995 on the western, central, and southern sites, respectively. Plot size was 2.25 ha; plots were located where bromes, schismus or mixed brome-schismus dominated and contributed the highest load of dead fine fuels. Fine fuels were ignited with a drip torch in a continuous flame line on the upwind border of each site. Fuel loads were [4]:


Fine fuels (kg/ha)

Dominant annuals (>25% relative cover)

Interspace fire spread (m/min) (x)

Area burned (% of 2.25 ha)

Beneath canopy Interspace Beneath canopy Interspace
western 800 100 brome-schismus schismus 1 (1) 50, patchy
central 300 25 brome-schismus schismus 0 (0) 0
southern 700 200 brome brome-schismus 12 (8) 50, continuous

There was no cloud cover during the fires. Other weather data at the beginning and end of the fires were [4]:

Site Time Air temp
Wind direction Wind speed (km/hr)
western begin 11:30 37 19 NNE 16 (18-32)
end 12:15 38 10 NNE 16 (18-32)
central begin 10:30 35 32 SSW 0 (8)
end 11:15 41 25 SSW 0 (8)
southern begin 10:20 33 17 SSE 5 (5)
end 11:30 35 15 SSE 8 (13)

Fire behavior
Site differences: The western Mojave site had large amounts of fuels beneath canopies and moderate amounts in interspaces. Relative humidity was low and wind speed high. Fifty percent of the site burned in many small patches. The central Mojave site had the lowest amount of fine fuels and wind speed and highest relative humidity. Fire did not spread beyond ignition points. The southern Mojave site had the greatest amount of fine fuels both beneath shrub canopies and in interspaces. Relative humidity was low and wind speeds moderate, so fire spread rapidly across interspaces. Fifty percent of the area burned in large, continuous areas [4,6].

Species effect on fire behavior: Among grass species, fire spread was most even, extensive, and rapid where bromes and schismus codominated. Slow, patchy fires resulted when schismus dominated interspaces; rate of spread was 1 m/minute and flame lengths were 5 to 10 cm in interspaces with schismus. In contrast, where bromes dominated interspaces fire spread approximated 12 m/minute, with flame lengths up to 30 cm. Only brome fuels produced flame lengths sufficient to consume large amounts of plant biomass. Redstem stork's bill and native annuals including bristly fiddleneck, western tansymustard, and lacey phacelia did not contribute substantially to fire spread due to low frequency and cover [4,6].

Direct fire effects
Since this was a fuels study, characterizing fire effects was not an objective, so fire-induced shrub mortality or postfire recovery was not measured. Flames fueled by bromes consumed small shrubs including burrobush, winterfat, burrobrush, and Anderson wolfberry. Schismus fires rarely ignited small shrubs. Creosotebush, a large shrub, did not ignite with schismus fires, and usually not with brome fires. However, creosotebush with large accumulations of brome and dead shrub stems in the subcanopy were highly flammable, and large creosotebush shrubs with brome beneath the canopy were usually consumed [4,6]. Of the burned shrubs, only burrobrush is a strong sprouter; the other shrubs are either weak sprouters that rarely survive fire (white bursage and creosotebush) or moderate sprouters following low-severity fire (winterfat and water jacket). For further information on postfire responses of these shrubs, use the links at the beginning of this Research Project Summary to access FEIS reviews on these shrub species.

Fire had no direct effect on the annuals, which were already dead at time of burning. Heat generated by bromes ignited and consumed dead stems and litter of native annuals, while most native stems and litter did not burn with schismus-fueled fire [4,6].

Fire management objective (fine fuels) Nonnative bromes and schismus were necessary for fire spread on these Mojave Desert sites. They were the only annual species that produced enough biomass and continuous cover of fine fuels that persisted into the fire season. Fine fuels from large forbs contributed to fire continuity and spread, although large annual forbs did not produce enough biomass to sustain fire (Brooks, personal observation cited in [4]). Large forbs that contributed to fire spread included nonnative Asian mustard, shortpod mustard, tall tumblemustard, London rocket, and flixweed tansymustard. Native bristly fiddleneck also contributed to fire spread [4]. Large nonnative forbs are especially common along roads, where fires often start, and are expanding their range into roadless areas [3,11]. Large nonnative forbs may present a greater fire hazard in the future [4].

Annual litter also contributes to fine fuel loads, and thick litter layers often develop where nonnative bromes and schismus are present [3]. Thick litter leads to high fire temperatures, long flame residence times, and continuous burn patterns (Brooks, unpublished data in [4]). Because plant litter decomposes slowly in deserts, and grass species can be among the slowest to decompose (review by [8]), brome and schismus litter accumulation may facilitate fire spread in the Mojave Desert [4].

With the exception of bristly fiddleneck, native annuals did not greatly affect fuel loads. Small nonnative forbs such as redstem stork's bill, and annual grasses besides bromes and schismus, did not contribute substantiality to fuel loads either. Prefire coverage and frequency of native sixweeks grass and nonnative small sixweeks grass were low, so the Vulpia species were not important fuels [4,6]. By summer, dead plant matter from these annual grass species was mostly crumbled and scattered, while biomass of dead brome and schismus was not [6].

Other fire management implications: This study showed that nonnative bromes fuel hot, fast-moving, continuous fires, whereas schismus fuels cooler, slower-moving, patchy fires. However, schismus can facilitate fire spread between brome patches [6]. Schismus can, and has, fueled fires in low-elevation Mojave sites. Several wildfires in the 1990s that exceeded 40 ha were fueled mostly by schismus species [4].

The nonnative grass/fire cycle is documented for cheatgrass in the Great Basin (e.g., [2,14,17]), but to date (2006), is less well documented for the nonnative grasses that dominate some southwestern desert ecosystems. This study provides documentation for a nonnative grass/fire cycle in the Mojave Desert. As in the Great Basin, previously burned areas in warm (Mojave) and hot (Sonoran and Chihuahuan) desert ecosystems that are infested with nonnative annual grasses may be more susceptible to fire than unburned areas. The grass/fire cycle is a serious ecological threat in southwestern deserts because native shrubs such as creosotebush and burrobush are poorly adapted to frequent fire [2,7,13]. The grass/fire cycle is most likely to occur on high-elevation sites in the Mojave Desert. High-elevation desert sites provide more mesic conditions for red brome establishment, and high-elevation red brome populations survive drought better than bromes on low-elevation desert sites (Minnich 1998, personal communication cited in [4]).

A review by Rundel and Gibson [15] that preceded this Research Project Summary noted that nonnative bromes, especially red brome, may be exceedingly dense in years of above-average rainfall in the Mojave Desert. For example, Rock Valley, Nevada, is a relatively undisturbed site due to its proximity to the Nevada Test Site, which has restricted access. Inventories showed that before 1960, red brome density in Rock Valley was 14 plants/m² or less. In 1976, density averaged 91 plants/m². In 1988, a wet year, red brome density averaged 2,034 plants/m²: 25 times the 1988 density of all native annuals combined [15].

In this Research Project Summary, Brooks [4,6] emphasizes that minimizing outbreaks and spread of nonnative annual grasses and large forbs is critical for reducing fire occurrence and spread in the Mojave Desert. Human ignition sources should be minimized, especially where nonnative annuals are abundant and topography is conducive to fire spread. Limiting or restricting livestock grazing and off-road vehicle use may reduce rate of nonnative seed spread. Monitoring roadways, washes, and other corridors for nonnative invasion, and controlling new outbreaks, may slow rate of nonnative grass invasion. On infested sites, controlling nonnative grasses in interspaces can reduce fuel continuity and reduce fire spread [4,6].

Sensitive species: The Mojave Desert contains several rare plant and animal species. Some of these species are endemic to the region [10,15], and fire regimes altered from very patchy, infrequent fires to frequent, mixed- and high-severity fires are endangering Mojave Desert habitats [4]. For example, a genetically distinct population of the federally threatened desert tortoise occurs in the Mojave Desert. Frequent fires fueled by red brome are degrading desert tortoise habitat in some areas [3,5].

This Research Project Summary provides direct fire effects information on the following shrubs and fine fuels information on the following herbaceous species. For further information, follow the highlighted links to the FEIS reviews for those species.


Common name Scientific name
Native shrubs
burrobush Ambrosia dumosa
burrobrush Hymenoclea salsola
winterfat Krascheninnikovia lanata
creosotebush Larrea tridentata
water jacket Lycium andersonii
Native forbs
bristly fiddleneck Amsinckia tessellata
western tansymustard Descurainia pinnata
lacey phacelia Phacelia tanacetifolia
Nonnative forbs
Asian mustard Brassica tournefortii
herb sophia Descurainia sophia
redstem stork's bill Erodium cicutarium
shortpod mustard Hirschfeldia incana
tall tumblemustard Sisymbrium altissimum
London rocket Sisymbrium irio
Native grasses
sixweeks grass Vulpia octoflora
Nonnative grasses
red brome Bromus rubens
cheatgrass Bromus tectorum
Chilean chess Bromus trinii
Arabian schismus Schismus arabicus
common Mediterranean grass Schismus barbatus
small sixweeks grass Vulpia microstachys


1. Barbour, Michael G. 1968. Germination requirements of the desert shrub Larrea divaricata. Ecology. 49: 915-923. [4212]
2. Billings, W. D. 1994. Ecological impacts of cheatgrass and resultant fire on ecosystems in the western Great Basin. In: Monsen, Stephen B.; Kitchen, Stanley G., comps. Proceedings--ecology and management of annual rangelands; 1992 May 18-22; Boise, ID. Gen. Tech. Rep. INT-GTR-313. Ogden, UT: U.S. Department of Agriculture, Forest Service, Intermountain Research Station: 22-30. [24248]
3. Brooks, Matt; Berry, Kristin. 1999. Ecology and management of alien annual plants in the California deserts. CalEPPC News (California Exotic Pest Plant Council Newsletter). 7(3/4): 4-6. [61753]
4. Brooks, Matthew L. 1999. Alien annual grasses and fire in the Mojave Desert. Madrono. 46(1): 13-19. [34386]
5. Brooks, Matthew L.; Esque, Todd C.; Schwalbe, Cecil R. 1999. Effects of exotic grasses via wildfire on desert tortoises and their habitat. In: 24th annual symposium of the Desert Tortoise Council: proceedings of the 1999 symposium; 1999 March 5-8; St. George, UT. Wrightwood, CA: Desert Tortoise Council: 40-41. [46285]
6. Brooks, Matthew Lamar. 1998. Ecology of a biological invasion: alien annual plants in the Mojave Desert. Riverside, CA: University of California. 186 p. Dissertation. [37220]
7. D'Antonio, Carla M.; Vitousek, Peter M. 1992. Biological invasions by exotic grasses, the grass/fire cycle, and global change. Annual Review of Ecology and Systematics. 23: 63-87. [20148]
8. Facelli, Jose M.; Pickett, Steward T. A. 1991. Plant litter: its dynamics and effects on plant community structure. The Botanical Review. 57(1): 1-32. [61756]
9. Garrison, George A.; Bjugstad, Ardell J.; Duncan, Don A.; Lewis, Mont E.; Smith, Dixie R. 1977. Vegetation and environmental features of forest and range ecosystems. Agric. Handb. 475. Washington, DC: U.S. Department of Agriculture, Forest Service. 68 p. [998]
10. Hickman, James C., ed. 1993. The Jepson manual: Higher plants of California. Berkeley, CA: University of California Press. 1400 p. [21992]
11. Kemp, Paul R.; Brooks, Matthew L. 1998. Exotic species of California deserts. Fremontia. 26(4): 30-34. [50447]
12. Kuchler, A. W. 1964. Manual to accompany the map of potential vegetation of the conterminous United States. Special Publication No. 36. New York: American Geographical Society. 77 p. [1384]
13. O'Leary, John F.; Minnich, Richard A. 1981. Postfire recovery of creosote bush scrub vegetation in the western Colorado Desert. Madrono. 28(2): 61-66. [3973]
14. Peters, Erin F.; Bunting, Stephen C. 1994. Fire conditions pre- and postoccurrence of annual grasses on the Snake River Plain. In: Monsen, Stephen B.; Kitchen, Stanley G., comps. Proceedings--ecology and management of annual rangelands; 1992 May 18-22; Boise, ID. Gen. Tech. Rep. INT-GTR-313. Ogden, UT: U.S. Department of Agriculture, Forest Service, Intermountain Research Station: 31-36. [24249]
15. Rundel, Philip W.; Gibson, Arthur C. 1996. Ecological communities and processes in a Mojave Desert ecosystem: Rock Valley, Nevada. Cambridge; New York: Cambridge University Press. 369 p. [61799]
16. Shiflet, Thomas N., ed. 1994. Rangeland cover types of the United States. Denver, CO: Society for Range Management. 152 p. [23362]
17. Whisenant, Steven G. 1990. Changing fire frequencies on Idaho's Snake River Plains: ecological and management implications. In: McArthur, E. Durant; Romney, Evan M.; Smith, Stanley D.; Tueller, Paul T., comps. Proceedings--symposium on cheatgrass invasion, shrub die-off, and other aspects of shrub biology and management; 1989 April 5-7; Las Vegas, NV. Gen. Tech. Rep. INT-276. Ogden, UT: U.S. Department of Agriculture, Forest Service, Intermountain Research Station: 4-10. [12729]

FEIS Home Page