Abstract
The introduction and expansion of the alien annual cheatgrass (Bromus tectorum L. [Poaceae]) has led to significant changes in the North American sagebrush biome fire regime. Fires have become larger and more frequent due to the creation of continuous fine fuel load of cheatgrass biomass that fills large expanses of the Intermountain West. In an effort to compartmentalize and slow fire progress, land management agencies continue to install hundreds of kilometers of vegetative greenstrips, largely comprising introduced forage grasses and sub-shrubs. While effective at suppressing fire progress, these greenstrips provide little ecological functionality because of their limited species composition. Curlycup gumweed (Grindelia squarrosa (Pursh) Dunal [Asteraceae]) is a short-lived native perennial forb, which may have potential for inclusion in Intermountain greenstrips to provide a pollen and nectar source without compromising the fire-suppressing capabilities of the greenstrip. We compared flammability traits of curlycup gumweed against 4 commonly utilized greenstrip species and cheatgrass during the summer months of June through September using collections made in southern Idaho. Curlycup gumweed maintained moisture content levels similar to those of forage kochia (Bassia prostrata (L.) A.J. Scott [Chenopodiaceae]). Curlycup gumweed out-performed crested wheatgrass (Agropyron cristatum (L.) Gaertn. [Poaceae]), Siberian wheatgrass (A. fragile (Roth) P. Candargy [Poaceae]), and Russian wildrye (Psathyrostachys juncea (Fisch.) Nevski [Poaceae]) for time to ignition and duration of combustion. Based on these results, curlycup gumweed should be considered for use in Intermountain greenstrip seedings.
Curlycup gumweed (Grindelia squarrosa) occurs naturally in disturbed habitats throughout western North America. It is currently being evaluated for selected class germplasm release for use in pollinator and wildlife habitat and may have potential for use in Intermountain greenstrip seedings.
In recent decades, the sagebrush (Artemisia tridentata Nutt. [Asteraceae]) steppe biome of the Intermountain West of North America has been significantly affected by wildfires of increasing size and frequency (Pellant 1990; Shinneman and others 2018). Although wildfires were an important factor in the development and maintenance of sagebrush ecosystems, recent changes, largely influenced by human activities, have dramatically altered the dynamics and impacts of wildfire beyond natural limits, creating huge expanses affected by fire (Shinneman and others 2018). For example, Brooks and others (2015) reported that 8.4 million ha (20.8 million ac) of Greater Sage-Grouse (Centrocercus urophasianus Bonaparte [Phasianidae]) range had burned between 1984 and 2013. In another report, a total of 566,000 burned ha (1.4 million ac) was documented between 2000 and 2018 in Idaho public lands alone (EIIFC 2019).
Much of the change in fire impact in the region can be attributed to the invasion of introduced annual grasses such as cheatgrass (Bromus tectorum L. [Poaceae]). High-density cheatgrass monocultures have resulted in accelerated fire frequencies (more frequent fires) (Balch and others 2013; Brooks and others 2015) and a significant increase in the number of large “megafires” caused by longer, hotter, and drier fire seasons (Westerling and others 2006). Cheatgrass has caused fire frequency in the Great Basin to decrease from once every 75 to 100 y to as little as every 3 to 5 y in certain areas (Whisenant 1990; Monsen and Memmott 1999). Invasive annual grasses fill available open spaces and provide an abundance of continuous fine fuel (Pilliod and others 2017) (Figure 1). Even in areas where the perennial vegetation is largely intact, the invasion of annuals creates conditions conducive to wildfire. Likewise, cheatgrass responds quickly following fires with a new flush of seedlings perpetuating the cheatgrass fire cycle (D’Antonio and Vitousek 1992).
Intermountain shrublands in the sagebrush biome typically have a strong component of bare ground (foreground). Cheatgrass (background), however, invades these interspaces and creates a continuous fuel load allowing fires to travel much farther than they would naturally.
Most fires in the region occur in the summer months of May through September (Shinneman and others 2018) with peak fire activity beginning in July (EIIFC 2019). During this time in the Intermountain basins, several weeks of continuous drought are met with frequent lightning storms. In Idaho, lightning strikes account for 20 to 200 fires each year (EIIFC 2019). Added to that are the human-caused fires resulting from cigarettes, automobiles, mowing, and gunfire.
Fuel breaks (any fuel treatment involving the removal or modification of vegetation in strips to disrupt fuel continuity and reduce fuel loads) are commonly implemented by US Bureau of Land Management (BLM) and other land management agencies to slow and compartmentalize wildfires, allowing fire crews time to respond when a fire occurs (Maestas and others 2016). For example, the goal of the BLM’s recent Paradigm Project in southwestern Idaho was to install approximately 480 km (298 mi) of fuel breaks in an area highly frequented by wildfire (USDI BLM 2011). Greenstrips are a form of fuel break developed to strategically install perennial plants with beneficial fire-suppressing attributes in locations where they might suppress or slow the advancement of a wildfire by changing the dynamics of the fire behavior triangle (fuels, weather, topography). Properly developed greenstrips increase the proportion of plants with high moisture content and exclude those species that produce fine fuels (Pellant 1994; Maestas and others 2016).
Plant species used in Great Basin greenstrips should possess many of the following attributes: 1) stay green and retain moisture during the wildfire season; 2) be adapted to the site and have the ability to persist through periodic extended drought conditions; 3) be grazing tolerant; 4) reduce fuel continuity by having widely separated individual plants or by producing relatively low amounts of fuel; 5) be capable of establishing and persisting among competitive annual species; and 6) be fire tolerant themselves (Monsen 1994; Pellant 1994; Davison and Smith 1997; Maestas and others 2016). Few species have been shown to meet these requirements. The most commonly recommended species for Intermountain greenstrip plantings include forage kochia (Bassia prostrata (L.) A.J. Scott [Chenopodiaceae]), crested wheatgrass (Agropyron cristatum (L.) Gaertn. [Poaceae]), Siberian wheatgrass (A. fragile (Roth) P. Candargy [Poaceae]), and Russian wildrye (Psathyrostachys juncea (Fisch.) Nevski [Poaceae]) (St John and Ogle 2009; Maestas and others 2016; Shinneman and others 2018).
Crested wheatgrass, Siberian wheatgrass, and Russian wildrye are all long-lived and highly drought-tolerant perennial grasses introduced from Eurasia. These species were initially planted in the region to provide forage for livestock, but their drought tolerance and ability to stay (at least somewhat) green during the dry season have led to their inclusion in greenstrip seedings. The introduced sub-shrub, forage kochia, has been shown to be effective at suppressing wildfires by disrupting fuel continuity and maintaining high moisture levels (Harrison and others 2002; Waldron 2011). Forage kochia was the primary species to be seeded in the BLM Paradigm Project (USDI BLM 2011). In field trials conducted by Monsen and Memmott (1999), flame lengths going from dry grass stubble into plots of forage kochia were reduced from 3 to 4.5 m (10–15 ft) down to 0.6 to 1.5 m (2–5 ft). Fires in 26 km/h (16 mi/h) winds burned into the forage kochia plot an average of only 0.6 m (2 ft) before visibly diminishing (Monsen and Memmott 1999).
One notable drawback of greenstrips in the Great Basin is the lack of plant species diversity and thus ecological function. Greenstrips are often monocultures of non-native perennials such as crested wheatgrass or forage kochia, which offer little to native wildlife. Merriam and others (2006) found that non-native plant abundance was greater in fuel breaks than in adjacent areas. In monotypic greenstrips of forage kochia one often sees only forage kochia and bur buttercup (Ceratocephala testiculata (Crantz) Roth [Ranunculaceae]) (personal observation). The lack of diversity, in particular native forbs, creates miles of non-functional habitat effectively void of valuable resources such as native pollen and nectar. Adapted forbs that meet the fire management requirements could be beneficial in suppressing weed encroachment and could provide food and habitat for native insects, upland birds, and other wildlife. Monsen and Memmott (1999) concluded that greenstrips using a combination of species is practical and advisable; however, only a limited number of forbs have been evaluated for use in greenstrips in the Great Basin region.
Monsen and Memmott (1999) tested common yarrow (Achillea millefolium L. [Asteraceae]) and small burnet (Sanguisorba minor Scop. [Rosaceae]) and found significant reduction in flame lengths comparable to those observed with forage kochia. Maestas and others (2016) and St John and Ogle (2009) recommended common yarrow, blue flax (Linum perenne L. [Linaceae]), small burnet, and alfalfa (Medicago sativa L. [Fabaceae]) for inclusion in greenstrips based on their tendency to remain green into late summer. Davison and Smith (1997) further recommended Palmer’s penstemon (Penstemon palmeri A. Gray [Scrophulariaceae]) for use in Nevada greenstrips.
One species worthy of consideration in Intermountain seedings is curlycup gumweed (Grindelia squarrosa (Pursh) Dunal) (Figure 2). Curlycup gumweed is a native, short-lived perennial forb of the sunflower family (Asteraceae) found throughout North America with the exception of the southeastern states. It is well-adapted to conditions of disturbed, early-seral areas in the Intermountain West where it is commonly found along road margins and in degraded range (Welsh and others 2003). Curlycup gumweed seems to meet many of the desirable attributes for greenstrip species; namely, green during the fire season, able to withstand drought (Welsh and others 2003; Tilley and Pickett 2016), and grazing tolerant (via low palatability) (Ogle and Brazee 2009). A close relative, Great Valley gumweed (G. camporum Greene) native to the San Joaquin Valley of California, has been credited in helping reduce the spread of wildfires in California grasslands (Dremann 1994) and is listed as a “preferred” plant species for fire-wise landscaping in Marin County (FIRE Safe MARIN 2020).
Curlycup gumweed is often found naturally occurring along roadsides in the disturbed ground between the road and established perennials. Here shown in mid-July, still green.
Curlycup gumweed has potential to increase ecosystem function while meeting the requirements of desired greenstrip species. Highly attractive to native bees, curlycup gumweed is recommended for inclusion in pollinator-friendly habitat plantings (Lee-Mäder and others 2016). The species’ drought tolerance and late-season flowering make it especially valuable for the Conservation Reserve Program (CRP) and other range plantings in the arid West where late-blooming forbs are limited. It has also been documented as a food source for Sage-Grouse (Peterson 1970), a species being considered for listing under the US Endangered Species Act (ESA) by the US Fish and Wildlife Service (USFWS 2005).
We initiated a study to describe curlycup gumweed’s greenstrip attributes and compare its performance to that of commonly utilized species as well as cheatgrass to determine its potential for use in greenstrip seed mixes.
Burning characteristics are highly variable and dependent on numerous external factors including ambient temperature, wind speed, and relative humidity, so it is difficult to formalize flammability scores for various plant species (Kauf and others 2014; Essaghi and others 2017). Dried cheatgrass has an ignition temperature of approximately 270 °C (518 °F) (Kaminsky 1974). Launchbaugh and others (2008) further defined 225 kg/ha (200 lb/ac) or greater of cheatgrass as a fuel load with the potential to generate extreme fire behavior, equivalent to a fire line intensity of 100 BTU (British Thermal Units) per foot square per second (BTU/ft2/s). Beckstead and others (2011), however, measured flame temperatures at 5 cm (2 in) above the soil surface in prescribed burn sites of near-monoculture cheatgrass sites in Utah and Washington using temperature-indicating lacquer paints and found peak temperatures reaching just over 150 °C (302 °F). Though they reported cover values of 85%, they did not report on plant density or fuel load. Brooks (2002) likewise measured fire behavior in Mojave Desert species and found peak temperatures averaging 140° C (284 °F) in the understory, which was largely made up of annual Bromus spp. with biomass averaging 1000 kg/ha (890 lb/ac). Their results seem at odds with Kaminsky’s ignition temperatures (1974).
We found considerable variability in laboratory testing methods used to determine burning characteristics. Kauf and others (2014) conducted flammability tests with an epiradiator and 1 g samples at 250 °C (482 °F) and 420 °C (788 °F). Batista and others (2012) similarly used 250 °C for their epiradiator tests, while Essaghi and others (2017) used 2 g samples at 600 °C (1112 °F). We decided to use larger samples and heat sources similar to those found in a cheatgrass/sagebrush fire.
MATERIALS AND METHODS
Collection Locations
Percent moisture and fire behavior of 3 commonly used introduced perennial grasses, crested wheatgrass, Siberian wheatgrass, and Russian wildrye, and one introduced perennial shrub, forage kochia, were compared to the native forb curlycup gumweed. We also included samples of cheatgrass in our trial to observe undesirable characteristics. Biomass samples of approximately 1 kg (2.2 lb) were collected monthly throughout the fire season (June, July, August, and September) from populations in the Snake River Plain of southern Idaho for use in the individual experiments described below (Figure 3). Cheatgrass, curlycup gumweed, crested wheatgrass, and forage kochia were collected 8 km (5 mi) N of Minidoka, Idaho (42°49.709′, −113°26.067′) at an elevation of 1450 m (4785 ft) (site 1). This site is a degraded, basin big sagebrush (Artemisia tridentata Nutt. ssp. tridentata [Asteraceae]) and bluebunch wheatgrass (Pseudoroegneria spicata (Pursh) Á. Löve [Poaceae]) plant community receiving an average of 20 to 30 cm (8–12 in) precipitation. Collections of Russian wildrye and Siberian wheatgrass (site 2) were made 57 km (35.4 mi) NW of Aberdeen, Idaho (43°9.585′, −112°58.55767′) at 1395 m (4604 ft) in a site that historically supported a Wyoming big sagebrush (A. tridentata Nutt. ssp. wyomingensis Beetle & Young [Asteraceae]) and bluebunch wheatgrass community with similar precipitation (Web Soil Survey 2019). Samples were hand harvested and put in 1-gal freezer storage bags and stored in an ice-filled cooler for transport.
We made biomass collections from 2 sites in the Snake River Plain of southern Idaho. Also shown is the location of the Burley, Idaho, weather station and Aberdeen Plant Materials Center. Image courtesy of Google Earth.
Moisture Content
To determine moisture content (MC), 4 subsamples of 10 to 20 g (0.35–0.70 oz) were weighed within 4 h of collection, then dried in a forage oven at 60 °C (140 °F) for 5 d. We computed MC as follows:
where MC is the moisture content of the combustible material (in %), WW is weight (wet) of material at the time of collection, and DW is weight (dry) of material.
Burning Characteristics
To estimate the temperatures of a cheatgrass fire, we dug a 1.45 m2 (4.8 ft2) circle of high-density dry cheatgrass monoculture, including the top 5 cm (2 in) of soil. We placed it in a galvanized washtub of the same diameter to maintain fuel orientation and ignited it on one end with a propane torch (Figure 4). Average aboveground biomass of the site was determined by taking 5 samples, 0.5 m2 in size, and oven-drying them at 60 °C (140 °F) for 5 d. Average biomass at the site ranged from 1200 to 4560 kg/ha (1068–4058 lb/ac) with a mean of 2460 kg/ha (2189 lb/ac), a rate well above Launchbaugh and others’ (2008) definition of high fuel load and Brooks’ (2002) Mojave Desert biomass. We measured peak temperatures with a Lasergrip1022 infrared thermometer (Etekcity Corporation, Anaheim, California) held 0.5 m (1.65 ft) from the flames. Site conditions during the burn were 20 °C (68 °F) and 42% relative humidity. In 2 replications we recorded peak temperatures of 522 °C (971.6 °F) and 524 °C (975.2 °F).
We determined combustion temperatures of a high-density cheatgrass fire by igniting a 1.45 m2 area of freshly dug material in a galvanized washtub. In 2 replications, we recorded peak temperatures of slightly more than 500 °C (932 °F).
To quantify burning characteristics, a 12 g (0.4 oz) sample of loosely arranged fresh material was centered in a 10 × 15 cm (4 × 6 in) cage made of 13 mm hardware cloth and placed directly into the flames of a liquid propane burner (Kenmore RB2518TS) gas grill with individual burner output of 170 BTU/ft2/s. We measured flame temperatures of 300 °C to 400 °C (572–752 °F) using a 7.6 cm (3 in) stainless steel temperature gauge from GasSaf (Zengcheng Guangzhou, Guangdong) mounted with the sensor end held directly above the flame at the same height as the burn samples.
Time to ignition (TI), defined as visible flame sustained for more than 1 s, and duration of combustion (DC), defined as ignition to extinction from consumption of available fuel, were measured using a hand-held digital timer. Each test was replicated 4 times. Data were not analyzed for significance, but means are presented here to show trends.
Weather data were obtained from Burley, Idaho, an area of similar site characteristics and historic plant communities that lies 27 km (16.8 mi) SE of site 1 and 79 km (49 mi) SW of site 2 (Web Soil Survey 2019). Weather graphs for the period of 15 June to 15 September 2019 (Figure 5) were prepared using NOWData (NOAA NWS 2019). Precipitation was normal for the testing period, totaling 300 mm (12 in). There were 4 rainfall events recorded in the study area during the period of evaluation: on 3 July, the region received 23 mm (0.92 in); on 25 July, 127 mm (5.08 in); on 8 August, 23 mm (0.92 in); and between 7 September and 10 September, the area received 127 mm (5.08 in). Observed temperatures ranged from a low of 0.6 °C (33.1 °F) on 9 June to a high of 37 °C (98.6 °F) on 22 July.
Accumulated precipitation (in) and daily temperature data from nearby Burley, Idaho. Courtesy of NOWData.
Moisture content (MC) of wildland-harvested cheatgrass, 3 introduced forage grasses, forage kochia, and curlycup gumweed from 15 June to 15 September 2019 in the Intermountain West. BRTE = cheatgrass, GRSQ = curlycup gumweed, PSJU = Russian wildrye, AGCR = crested wheatgrass, BAPR = forage kochia, AGFR = Siberian wheatgrass. Error bars represent 1 standard error.
RESULTS
Moisture Content
In mid-June all 6 species had MC of more than 50% (Figure 6). Cheatgrass moisture levels dropped sharply between June and July as flowering had concluded, plants senesced, and seed matured. In July, August, and September, cheatgrass MC was at essentially 0%. The 2 perennial wheatgrass species showed similar trends in MC with percent moisture starting at approximately 60% in June and decreasing to 28% in August, which it maintained for the remainder of the trial. Russian wildrye began with moisture levels similar to the other perennial grasses but maintained somewhat higher levels in August (40%). Russian wildrye also responded to late summer rain in early September with an increase of average MC to 49%. Forage kochia maintained consistent MC throughout the trial duration averaging just above 60% from June through August. In September we observed a slight decline in MC of forage kochia to 56%. Curlycup gumweed had the highest MC in June (79%), after which MC declined through July and August from 68% to 55% and ending at 50% in September.
Burning Characteristics
Cheatgrass had the fastest TI in every month of the trial (Figure 7). In June it took an average of 7 s for cheatgrass to ignite. At this point the cheatgrass was a mixture of dried and green material with some plants at anthesis and others undergoing seed maturation. Cheatgrass ignition in July, August, and September was essentially instantaneous. Crested wheatgrass, Siberian wheatgrass, and Russian wildrye had similar trends of ignition with TIs averaging from 17 s (crested wheatgrass) to 25 s (Russian wildrye) in June and decreasing slightly during the latter months when plants were drier. August and September TI for the perennial grasses ranged from 5 to 13 s. Forage kochia TIs were consistently greater than those of the grasses. Curlycup gumweed was very difficult to ignite in June with an average TI of 90 s. TIs dropped in the subsequent months, corresponding to decreasing moisture levels; however, curlycup gumweed (Figure 8) had the longest ignition times of the evaluated species each month with the exception of September when forage kochia had slightly slower ignition (38 s) compared to curlycup gumweed (28 s).
Time to ignition (TI) at 440 °C (824 °F) of 12 g samples of wildland-harvested materials. BRTE = cheatgrass, GRSQ = curlycup gumweed, PSJU = Russian wildrye, AGCR = crested wheatgrass, BAPR = forage kochia, AGFR = Siberian wheatgrass. Error bars represent 1 standard error.
Curlycup gumweed had the longest time to ignition (TI) for each month except September. Photo shows material collected in June.
As MC decreased over the summer months, fires burned much more quickly in all species (Figure 9). Cheatgrass samples burned more quickly than any other species examined (Figure 10). The longest cheatgrass DC (80 s) was in June when much of the sample was still green. The dry, post-shatter cheatgrass tested from July, August, and September burned almost explosively with fires averaging approximately 10 to 20 s. We observed intermediate DCs from the perennial grasses. In the dry summer months of August and September, all 3 perennial grass species exhibited DCs of approximately 60 s. DCs of forage kochia and curlycup gumweed were notably longer than those of the other species, averaging approximately 100 s in August and September.
Duration of combustion (DC) of 12 g biomass at 440 °C (824 °F) of wildland-harvested materials. BRTE = cheatgrass, GRSQ = curlycup gumweed, PSJU = Russian wildrye, AGCR = crested wheatgrass, BAPR = forage kochia, AGFR = Siberian wheatgrass. Error bars represent 1 standard error.
A 12 g sample of cheatgrass harvested in mid-September ignites quickly and burns explosively.
CONCLUSIONS
Burning characteristics are dependent on numerous external factors and are difficult to determine precisely (Kauf and others 2014; Essaghi and others 2017); however, between-species comparisons can be made. Our results from using low-tech test methods show trends and comparative values for 6 species important in the Intermountain West sagebrush ecosystem, as they occurred throughout the summer.
MC is perhaps the most important factor when determining a species’ flammability. Cheatgrass TI and DC changed significantly as MC levels dropped. Russian wildrye responded to late-summer rain events with green-up and increased MC, although that did not appear to translate to longer TI or DC. Curlycup gumweed and forage kochia both maintained high MC throughout the summer. Curlycup gumweed response to fire was similar to forage kochia, though decreased MC of curlycup gumweed in September corresponded with slightly decreased TI. Both curlycup gumweed and forage kochia were quite slow to ignite and to burn, mostly just drying up and withering away rather than creating a significant flame. Curlycup gumweed was clearly less prone to ignition and the creation of a hot flash during the fire season than the commonly planted perennial grasses.
To our knowledge, curlycup gumweed has rarely been recommended for seeding in the Intermountain West restoration and reclamation projects. This omission is likely attributable to being considered a weedy species by some because of its low palatability and its tendency to increase under grazing (Whitson and others 1996; Ogle and Brazee 2009). And yet, curlycup gumweed possesses numerous attributes desirable for native wildlife habitat restoration practices; namely, drought tolerance, easy establishment, adaptation to disturbed sites, attractive to pollinators, and fall flowering (Tilley and Pickett 2016). As such, multiple accessions are currently being investigated by the authors for selected class germplasm release at the USDA NRCS Plant Materials Center in Aberdeen, Idaho.
Despite its performance in our evaluations, curlycup gumweed may possess some attributes that limit its feasibility in greenstrips. Curlycup gumweed is a short-lived perennial or biennial that naturally serves as an early-seral colonizer of disturbed areas. Typically, curlycup gumweed decreases in abundance after a few years, giving way to longer-lived, late-successional climax species such as sagebrush and perennial bunch grasses. It is possible that the niches created as curlycup gumweed decreases can be exploited by invasive annuals (Maestas and others 2016). However, greenstrip areas do not typically phase toward a climax community; rather, they remain in a state of semi-disturbance thanks to periodic mowing. Under such management, curlycup gumweed may continue to reestablish and persist if allowed.
Of note, curlycup gumweed has been investigated as the feedstock for producing sustainable biodiesel and aviation fuels (Neupane and others 2016). Researchers found that 12 to 25% of the dry weight of the plant (depending on agronomic factors) is made of terpenoid compounds, primarily grindelic acid, which is a diterpene acid that could be converted to hydrocarbons that can be used directly as an aviation fuel. Despite containing these flammable compounds, we saw no evidence that field-harvested curlycup gumweed material was abnormally combustible.
Our results indicate curlycup gumweed possesses many traits desirable for a greenstrip species and should be considered for inclusion in Intermountain seed mixes. Its presence in greenstrips could increase species diversity and provide a pollen and nectar source for native insects.
Footnotes
Photos by Derek Tilley
