Fire Effects Information System (FEIS)

FEIS Home Page

Research Project Summary: Prescribed and wildfire in clearcut mixed-conifer forests of Miller Creek and Newman Ridge, Montana

Table of Contents:

• SUMMARY
• PROJECT AND SITE DESCRIPTION
• PREFIRE PLANT COMMUNITY AND FUELS
• PLANT PHENOLOGY AT BURN TIME
• FIRE SEASON AND SEVERITY CLASSIFICATION
• FIRE DESCRIPTION
• FIRE EFFECTS ON FUELS, PLANT SPECIES, AND SMALL MAMMAL SPECIES
• FIRE MANAGEMENT IMPLICATIONS
• APPENDIX: SPECIES INCLUDED IN THE SUMMARY
• REFERENCES

Common names are used throughout this summary. See the Appendix for a complete list of the common and scientific names of plant and animal species discussed in this summary and for links to FEIS Species Reviews.

The researchers conducting this set of studies set the significance level at P = 0.05 for all variables tested [3].

Citation:
Fryer, Janet L.; Scher, Janette S.; Crane, Marti, comps. 2017. Research Project Summary: Prescribed fire and wildfire in clearcut mixed-conifer forests on Miller Creek and Newman Ridge, Montana. In: Fire Effects Information System, [Online]. U.S. Department of Agriculture, Forest Service, Rocky Mountain Research Station, Missoula Fire Sciences Laboratory (Producer). Available: https://www.fs.usda.gov/database/feis/management_project_summaries/Shearer98/all.html [].

Sources:
This summary compiles research that was originally published in the following sources: [2,3,4,5,6,9,10,12,13,14,15,16,17,18,19,20,21].

Project Locations:
There were two study sites, both in northwestern Montana: Miller Creek and Newman Ridge. The Miller Creek site was 641 acres (259 ha) and located in the Miller and Martin creek drainages of the Miller Creek Demonstration Forest on the Flathead National Forest. The Newman Ridge site was 526 acres (213 ha) and located between Two Mile and Ward creeks on the Lolo National Forest [2,3,14,18]. Winters at the study sites are long, cool, and wet; summers are short and warm. Approximately two-thirds of annual precipitation falls as snow. Snowmelt and most rainfall occurs in April, May, and June. Climate is cooler and drier on Newman Ridge than at Miller Creek. Surface loess deposits containing ash from the Mt. Mazama and Glacier Peak volcanic eruptions are 0.5 to 2.5 inches (1-6 cm) thick at Miller Creek and 2 to 3 inches (5-8 cm) thick on Newman Ridge [3,4,14,18].

Figure 1—Locations of Miller Creek and Newman Ridge study sites in northwestern Montana [3].

Site Descriptions:
Miller Creek: Elevation at Miller Creek ranges from 4,200 to 5,000 feet (1,280-1,524 m), with slopes averaging 24% and ranging from 9% to 35% [3,4,14,18]. Soils are Andic Cryoboralfs that developed in glacial till from argillites and quartzites of the Wallace (Belt) formation. Mean annual precipitation is about 25 inches (630 mm) [3,4,14,18].

Newman Ridge: Elevation on Newman Ridge ranges from 4,400 to 5,400 feet (1,341-1,646 m), with slopes averaging 55% and ranging from 44% to 76%. Soils are Andic Cryochrepts that either developed in residuum or in colluvium from argillites and quartzites of the Belt formation. Mean annual precipitation is nearly 40 inches (1,000 mm) [3,18].

Treatment Descriptions:
Four treatments were originally planned: clearcut only; clearcut and spring burn; clearcut and summer/early fall burn; and an uncut, unburned control. Sixty treatment units were established at Miller Creek and 16 on Newman Ridge, with an equal number of units facing each of the four cardinal directions. Units were 10 acres (4 ha) in size at Miller Creek and ranged from 21 to 58 acres (8.5-23 ha) on Newman Ridge. Prior to prescribed fire treatments, all study units except the 8 control units were clearcut between 1966 and 1969. Tree limbs were slashed in place and dispersed by hand crews for fuel continuity. Uncut control units were adjacent to the clearcuts. Sample plots were 2.5 acres (1 ha) in size and centered in each unit. At Miller Creek, a total of 73 clearcut plots in 55 different units were broadcast burned: 11 in spring, 51 in summer, and 11 in fall. Wildfire "treatments" were added after wildfires burned portions of clearcut-and-burned treatments and unburned controls [2,9,15,18,20].

Miller Creek: Mixed-conifer stands at Miller Creek ranged from 200 to 250 years old [9,20]. In order of dominance, conifers were western larch, Douglas-fir, and Engelmann spruce, with some lodgepole pine, grand fir, and subalpine fir. Timber volume was about even between Douglas-fir, Engelmann spruce, and western larch. Most of Miller Creek was in one of three phases (major variations) of the subalpine fir/bride's bonnet habitat type. Four habitat types were identified (table 1) [2,3,14,18,20].

Newman Ridge: Mixed-conifer stands on Newman Ridge ranged from 180 to 200 years old. They were more diverse than stands at Miller Creek. In order of dominance, conifers were Douglas-fir, western larch, and lodgepole pine, with some ponderosa pine, grand fir, subalpine fir, western white pine, Engelmann spruce, and western redcedar. Seven habitat types were identified (table 2) [2,3,14,18].

Study sites are classified in the following Biophysical Settings and probably historically experienced the fire regimes described in table 3.

Prefire Fuels:
Clearcutting added considerable slash to the forest floor. Duff depth averaged 2.2 inches (5.5 cm) at Miller Creek and 2.1 inches (5.2 cm) on Newman Ridge, with deeper duff on north- and east-facing aspects than on south- and west-facing aspects. Excluding duff, fuel loads after clearcutting and before prescribed fire ranged from 60 to 165 tons/acre (135-370 T/ha). Fuel loads at Miller Creek were slightly more than on Newman Ridge. Mass of prefire fuel loads by size class is shown in table 4 [2].

Mean moisture content of fuels at Miller Creek is shown in table 5 [2].

Mean moisture content of fuels on Newman Ridge is shown in table 6 [2].

Only times of conifer seed dispersal and germination were provided. On Miller Creek and lower slopes of Newman Ridge, conifers dispersed seeds from mid-September to early October; on upper slopes of Newman Ridge, conifers dispersed seeds in mid-October [3]. Depending on species, conifer seeds germinated around the time of snowmelt [5] (see Germination, establishment, and growth).

Prescribed fires were set in spring, summer, and fall (May-October). Fire severity ranged from low (spring prescribed fires) [2,3,13,15,16] to moderate (mostly midsummer) and severe (mid- to late summer and fall; includes the summer wildfires). Fire severity was determined by ground char depth [20], described in Ryan and Noste [11].

Miller Creek:
Spring prescribed fires: 18 May to 18 June of 1967 and 1968
Summer prescribed fires: 3 July to 9 September of 1967 and 1968
Fall prescribed fires: 30 September 1967 and 2 October 1968
Summer wildfire: 23 August 1967

Newman Ridge:
Spring prescribed fires: 1 June to 18 June 1969 and 1970
Summer prescribed fires: 9 July to 23 July 1969 and 1970
Fall prescribed fires: 1 October 1969 and 28 September 1970
Summer wildfire: 25 July 1969

Figure 2—Buildup index during burning days at Miller Creek and Newman Ridge [2].

Prescribed fires: Slash fuels were allowed to cure for an average of 9 months before burning (range: 2-18 months) [2,18]. Broadcast burning was conducted in the spring, summer, and fall, in 1967 and 1968 at Miller Creek and in 1969 and 1970 on Newman Ridge [2,3,17,18]. Fires were ignited by center-firing, flanking, or a combination of these. Most fires were ignited in late afternoon or evening after winds had decreased [3,10]. Clearcuts with heavy fuel loads were burned under wetter conditions than clearcuts with lighter fuel loads for fire control [3].   Wildfires: The 1967 wildfire at Miller Creek burned portions of 5 units that had been clearcut and burned under prescription and 4 uncut, control units [3,17]. The 1969 wildfire on Newman Ridge burned portions of 1 clearcut unit [16].

Figure 3—A tour group viewing study units and fall color of western larch at Miller Creek. U.S. Forest Service photo.

Prescribed fires removed most fine and woody fuels. During prescribed fires at Miller Creek, 75% of the fuels less than 3.9 inches (10 cm) burned, and 60% of larger fuels burned. On Newman Ridge, 89% of the fuels less than 3.9 inches (10 cm) burned, and 55% of larger fuels burned [2].

Plants
Stickney and Shearer [20,21] monitored posttreatment successional development for 7 years on Newman Ridge and for 20 years at Miller Creek. On both sites in postfire year 1, sprouting species and species that established from seed dominated. Common beargrass and white spirea had highest cover among sprouting species. For species that established from seed, fireweed, snowbrush ceanothus, lodgepole pine, and western larch had highest cover. Herbaceous species dominated from about postfire year 2 to 7, with fireweed having highest cover. Shrubs dominated from about postfire year 6 to 15, with snowbrush ceanothus providing highest cover. At Miller Creek, conifer seedlings began overtopping shrubs around postfire year 7 and became dominant around postfire year 20 [20,21].

• Conifers
• Other plants

• Natural regeneration
• Artificial regeneration

Prescribed fires at Miller Creek killed roots (all species) within only 2 inches (5 cm) of the soil surface, on average, so the target goal of root kill depth to 4 inches (10 cm) was not met. Compared to the number of living roots before burning, the number of living roots remaining after fire generally decreased as duff and soil dried with date of burning (figure 4). The summer wildfire killed more roots than summer or fall prescribed fires, possibly because it burned when duff and soil moisture content were lowest and fuels driest. Seventy percent of roots in the 3.3- to 4-inch (8.4- to 10-cm) soil layer died from the wildfire, and >90% in the upper 2.4 inches (6.1 cm) died [13]. Root kill was determined using a chemical test [3].

Figure 4—Root mortality within 4 inches of the soil surface by date of burning on clearcut plots [13].

Seed supply: Nearby uncut trees [13,17,20] provided seed sources. Postfire dispersal of seed into the clearcuts from western larches near clearcuts was plentiful. Overall, the best seed production year for conifers was 1971; however, heavy frost in May killed some of western larch's potential seed crop [12,13,14]. Another good seed crop year for conifers occurred in 1980 [12]. Seed fall was highest on south-facing aspects, followed in order by west-, east-, and north-facing aspects [15]. On Newman Ridge, seed dispersal from off-site parents decreased sharply from the burn edge to about 300 feet (90 m) within burn perimeters. Seed dispersal continued at a low but uniform rate to about 800 feet (220 m) within burn perimeters [13]. Among conifers, western larch was most successful at establishing in burn interiors [3].

On-site seed sources were lacking on clearcuts [20], but on-site seed sources were available from parent trees on adjacent control plots. On uncut, wildfire-burned plots, seeds also dispersed from scorched cones of fire-killed western larch and ponderosa pine [13,17,20].

In addition to natural seeding from off- and on-site seed sources, seeds were sown on test plots in 1967, and bareroot seedlings were planted on four clearcuts at Miller Creek from 1970 through 1973 and on eight clearcuts on Newman Ridge from 1970 through 1975 (see Artificial regeneration).

Conifers with heavy seeds and rapid seedling root growth had the highest ratios of seedlings established to seeds dispersed [15]. Seedling:seed ratios were developed using the number of seeds caught in seed traps from 1969 to 1974, and the number of seedlings on plots closest to each seed trap (table 9) [16]. From postfire years 1 to 5, density of emergent and small, established seedlings was measured on 0.001-acre (0.0004 ha) plots. Plots were remeasured in postfire year 10, along with density of large seedlings. Density of large seedlings (defined in table 11) was measured on 0.003-acre (0.0013-ha) plots (n = 110-177 plots) [16].

Germination, establishment, and growth: Douglas-fir, lodgepole pine, subalpine fir, and western larch were the primary conifer pioneers on cut-and-burned plots [13,20,21]. Most conifers established from the 1971 cone crop, which was rated "good" for all species [20]. Most conifer germination occurred on mineral soil or soil with <0.5 inch (1.3 cm) of duff remaining [13]. Germination of western larch was greater on mineral soil than on unburned duff >0.5 inch thick. Very few western larch seedlings established on unburned duff despite adequate seed sources [13,15,17]. For all conifers, more germination occurred on north- and east-facing slopes than south- and west-facing slopes. Timing of germination varied by species, with some western larch seeds beginning germination several days before snowmelt and Engelmann spruce beginning germination about 3 weeks after snowmelt. Douglas-fir, grand fir, lodgepole pine, and ponderosa pine began germination soon after snowmelt [13]. DeByle and Packer [5] reported that Engelmann spruce benefitted the most, and Douglas-fir the least, from exposure of mineral soil on these study sites.

Overall, conifer establishment was higher at Miller Creek than on Newman Ridge. Highest conifer establishment was on north-facing slopes (table 10). Douglas-fir had high establishment rates on both study sites, while establishment of Engelmann spruce varied greatly between sites. On Newman Ridge, Douglas-fir established well on all but south-facing slopes, while grand fir established well only on north- and east-facing slopes. At Miller Creek, Douglas-fir, Engelmann spruce, and western larch showed good establishment on all aspects. Lodgepole pine established on nearly all clearcuts, while western redcedar established only on moist sites, mostly stream bottoms [18].

At Miller Creek, cover of conifer seedlings increased faster on uncut, wildfire-burned plots than on clearcut plots burned under prescription. On south-facing clearcut plots burned in May, conifer cover increased from about 1% in postfire year 10 to 15% in postfire year 20. On uncut plots that burned in the August wildfire, cover of conifer seedlings increased from about 1% in postfire year 3 to nearly 100% in postfire year 20. On wildfire-burned plots, nearly all regeneration was western larch and lodgepole pine that established from seeds dispersed from cones of on-site parent trees killed by the wildfire [20].

Natural regeneration of western larch at Miller Creek was higher on burned than on unburned sites (table 10). In 1978, stocking of large western larch seedlings at Miller Creek averaged 54% on burned clearcuts and 5% on unburned clearcuts. By 1984, stocking averaged 71% on burned clearcuts and 1% on unburned clearcuts [17,18]. Western larch seedlings outnumbered other conifers on 10 of 18 units burned under prescription. Western larch was well established on all aspects, with highest stocking on west-facing slopes [18]. On seven south- and east-facing blocks, however, western larch regeneration decreased from 1974 through 1984, while regeneration of other conifers more than doubled [19].

Natural regeneration of western larch was higher at Miller Creek than on Newman Ridge (table 10, table 11). Lower natural regeneration of western larch on Newman Ridge was likely due to harsher site conditions, larger clearcuts, and lower seed production [18]. In 1979 on Newman Ridge, stocking of large western larch seedlings was highest in the western redcedar-bride's bonnet habitat type; moderate in the grand fir-bride's bonnet, grand fir-common beargrass, and Douglas-fir-dwarf huckleberry habitat types; and lowest in the Douglas-fir-ninebark habitat type. Survivorship of western larch seedlings to postfire year 10 was less than expected; across habitat types, it averaged only 38% [12,17,18]. Survivorship was highest on north-facing slopes [18]. By postfire year 7, mean height of western larch and lodgepole pine seedlings was about that of shrubs (mostly snowbrush ceanothus) [20].

Height growth of western larch at Miller Creek was related to habitat type and phase, with the tallest seedlings on the warm, dry common beargrass phase of the subalpine fir/bride's bonnet habitat type and shortest seedlings on the cooler and moister rusty menziesia phase of the subalpine fir/bride's bonnet habitat type. Western larch seedlings in warm habitat types began growth earlier in the spring and might have benefited from extra nitrogen provided by snowbrush ceanothus, a nitrogen-fixing species that prefers warm sites. On Newman Ridge, western larch seedlings were significantly taller on east- and south-facing aspects than on west- or north-facing aspects. In 1978, the tallest western larch seedlings were on uncut stands burned by the wildfire (table 14) [17].

On burned clearcuts on Newman Ridge, lodgepole pine showed highest overall height growth among conifers; grand fir showed least height growth. Douglas-fir established in highest numbers (table 13).

Mortality: Seed and seedling losses were caused by drought and/or high temperatures at the soil surface, browsing animals, fungi, and frost heaving [14,15,17]. Drought was the leading cause of death on south-facing slopes and the second highest cause on other aspects [13]. Several animals killed conifer seedlings. Rodents browsed seedlings, and in 1968, migrating dark-eyed juncos killed over 90% of conifer germinants by removing seedcoats and clipping the cotyledons [13]. Death from fungal infection peaked in spring and ceased when soil dried, with most mortality occurring from mid- to late June. Fungi were the primary cause of conifer seedling death on Newman Ridge, with losses greater on plots with scorched duff than on plots burned down to mineral soil. Frost heaving killed some conifer seedlings in their first fall and early spring [13]. Ten years after treatments, mean survival of western larch seedlings on Newman Ridge was 38% [16].

The fires killed Pacific yew, which had not been logged before the fires. Although Pacific yew had 1% to 63% cover and 80% mean frequency at Miller Creek before fire, it was absent from all burned plots as of postfire year 8 [13].

Artificial regeneration
On Newman Ridge from postfire years 1 to 6, bareroot Douglas-fir, Engelmann spruce, grand fir, ponderosa pine, lodgepole pine, and western larch were auger planted on burned clearcuts in May. For each unit planted, species selection was based on what had been harvested. Planted seedlings were measured in postfire year 10 to determine survival rates [13]. Survivorship of ponderosa pine [3] and western larch was less than expected. For western larch, this was attributed to frost damage after seedling buds opened in May 1971. Other conifers were minimally affected by that frost [13].

For most species planted on Newman Ridge, establishment rates were highest on north-facing slopes (table 14). Survival of Douglas-fir, lodgepole pine, and western larch was similar on west- and east-facing slopes, but grand fir survival was higher on west- than on east-facing slopes [16]. In postfire year 10, mean survivorship of artificial regeneration across aspects was highest for Douglas-fir and least for western larch:

Douglas-fir: 59%
lodgepole pine: 57%
ponderosa pine: 53%
grand fir: 48%
Engelmann spruce: 44%
western larch 38% [13].

The researchers did not provide height growth data for artificial regeneration.

Other plants
Herbs dominated treated sites on Newman Ridge through postfire year 15, although shrubs also established soon after fire. Fireweed, tall annual willowherb, snowbrush ceanothus, and Scouler's willow established within 3 postfire years [20,21]. Common beargrass and fireweed constituted most herbaceous cover. Partially burned common beargrass sprouted within a few weeks of prescribed burning; its sprouting was "occasional" on wildfire-burned plots [13]. Fireweed seedlings established in postfire year 1 [20]. On plots that were clearcut and burned under prescription on Newman Ridge, herbs reached maximum cover (61%) in postfire year 3 and still had high cover (50%) in postfire years 9 through 15 [20,21]. Shrub cover exceeded that of herbs by postfire year 16 [20]. Sprouting shrubs or hardwoods on Newman Ridge included Sitka alder, Rocky Mountain maple, Scouler's willow, and white spirea [13].

Although not present before treatments, fireweed, tall annual willowherb, and bull thistle established on burn units from off-site, wind-blown seed. Fireweed was the most frequent herb from about postfire years 1 to 4, occurring on all but one burn plot [20]. It consistently had highest postfire cover of all herbaceous species on both sites. Fireweed attained highest cover on moist sites at Miller Creek that burned at high severity; however, its cover declined rapidly in early postfire years. At Miller Creek on block East-6, which had a severe prescribed fire, fireweed cover climbed from 0% before fire to about 57% in postfire year 2, then dropped to 35% in postfire year 9. On block South-1, which had a low-severity prescribed fire, fireweed cover climbed from 0% in postfire year 1 to 35% in postfire year 8. Tall annual willowherb was present on only 1 unit after fire. On that unit, it was the most common herb in postfire year 2. After that, its cover dropped to <1% [21].

Other herbs benefitted from openings created by clearcutting and fire [20]. Herbs sprouting after fire included common beargrass, broadleaf arnica, pinegrass, and northwestern sedge. At Miller Creek, common beargrass exceeded pretreatment cover on burned plots where duff remained intact. Since common beargrass has rhizomes on or near the soil surface, its cover was greatly reduced on plots where fire was severe. It had not gained pretreatment cover by postfire years 7 or 9 on plots where duff was either burned off or greatly reduced by fire [21]. Broadleaf arnica increased in cover in early postfire succession at Miller Creek but then lost cover rapidly. On block West-15, for example, its pretreatment cover averaged about 37%. In postfire year 2, it averaged about 55% cover and produced an "abundant" flower crop. By postfire year 6, its cover dropped to 0%. On the same plot, pinegrass showed a steady increase in cover. Its cover averaged about 12% before treatments and about 17% in postfire years 4 and 6 [21]. Pinegrass and northwestern sedge showed highest increases in cover at Miller Creek, exceeding their pretreatment cover by postfire years 5 to 9, depending on unit [21].

Most shrubs began sprouting within 2 weeks after prescribed fires, regardless of time of burning. On plots burned under prescription in May, shrubs reached highest cover in postfire year 16 (~60%) and declined by postfire year 20 (~40%) [20]. White spirea and Rocky Mountain maple were the most common sprouting shrubs [13,20]. However, Rocky Mountain maple, rusty menziesia, and Idaho goldthread (a perennial forb) were slow to recover on plots where duff consumption was high [13]. White spirea and twinflower showed highest increases in cover at Miller Creek, exceeding their pretreatment cover by postfire years 5 to 9, depending on unit. On both sites, thimbleberry averaged 60% frequency (about twice its pretreatment cover) by postfire year 9 [21]. Rusty menziesia, thinleaf huckleberry, and Sitka alder had higher cover on plots that were clearcut but not burned than those that were burned [20].

Snowbrush ceanothus and Scouler's willow were not present before fire. Snowbrush ceanothus established after fire from the soil seedbank [20,21], and Scouler's willow established from off-site, wind-blown seed [13]. Best establishment of snowbrush ceanothus was on severely burned plots [20]. For example, part of block South-3 on Newman Ridge burned in the severe wildfire. There, snowbrush ceanothus seedlings established in postfire year 1, averaging about 8% cover. Snowbrush ceanothus reached maximum cover in postfire year 6 at 40%. Block South-13 at Miller Creek had a high-severity prescribed fire. Snowbrush ceanothus did not establish on South-13 until postfire year 3, when it averaged 1% cover. It reached maximum cover of 37% in postfire year 8, and its cover dropped to about 30% in postfire year 9 [21].

Burning generally favored sprouting species, but two fires in quick succession did not. At Miller Creek, sprouting was minimal on plots that were clearcut, burned under prescription on 8 August 1967, then reburned in a wildfire just weeks later, on 23 August 1967. Fireweed dominated from postfire years 1 to 4 on these clearcut, twice-burned plots. Snowbrush ceanothus established from the soil seedbank and dominated twice-burned plots by postfire year 7. Conifer seedlings overtopped shrubs by postfire year 17. Most conifer seedlings (65%) were Douglas-fir and western larch [20].

Fireweed and white spirea dominated plots that were uncut-unburned prior to the wildfire (originally, control plots). By postfire year 20, lodgepole pine and snowbrush ceanothus dominated these wildfire-only plots [20].

Small mammals
Trapping was used to compare habitat use by small mammals. On both study sites, grids were laid out in four habitats: mature, uncut timber (control), clearcut only (logging slash), clearcut with low-severity prescribed fire, and clearcut with high-severity prescribed fire. Grids were 4.6 acres (1.9 ha) and contained 81 traps. Traps were monitored for 8 years (1969-1976), for 5 days in the fall of each year [15]. A total of 1,800 individuals were trapped at Miller Creek and on Newman Ridge, comprising 11 small mammal species. Of these, deer mice and red-tailed chipmunks were most frequent on all grids. Deer mice, red-tailed chipmunks, and northern red-backed voles constituted >90% of all small mammal species trapped. Deer mice and red-tailed chipmunks eat seeds; northern red-backed voles primarily eat fruits. These food choices probably attracted them to newly burned sites [6].

Northern red-backed voles were the most common small mammal in mature timber, particularly on north-facing slopes. It was nearly absent on burned plots. Red-tailed chipmunks were the second most common in mature timber and was about twice as frequent on south- than on north-facing slopes. Ermine, vagrant shrews, and northern flying squirrels occurred on all plot types in about equal, but low, numbers. Northern flying squirrels require durable snags for nesting, preferring snags of ponderosa pine and western larch. On mature timber plots on Newman Ridge, they were trapped more often on south-facing slopes, which had both conifer species, than on north-facing slopes, which had only western larch (table 16) [6].

Habitat use by small mammals was similar at both sites. Deer mice were most common on disturbed sites, while red-tailed chipmunks and red-backed voles were more common on less disturbed sites. The North American deermouse comprised <20% of total small mammal species composition in mature timber sites compared to >70% on clearcut-and-burned sites. Red-tailed chipmunk numbers were about 30% less on clearcut-and-burned than on mature timber sites, and the northern red-backed vole was absent from burned sites. However, other vole species, particularly the long-tailed vole, were trapped on burned sites [6].

North American deermouse numbers peaked in 1972, following the heavy conifer seed crop year of 1971. Increased availability of conifer seeds likely explains the peak in North American deermouse numbers [6].

Figure 5—North American deermouse. Image by Tom Brakefield © California Academy of Sciences, used with permission.		Figure 6—Red-tailed chipmunk. Creative Commons photo by Gena Bentall.

Small mammal use differed following clearcut-high severity fire versus clearcut-low severity fire. On Newman Ridge, all of block South-3 burned severely in mid-September, while 2 weeks later, only 35% of block North-3 burned in a patchy, low-severity fire. Deer mice occurred in large numbers on the severely burned block; 95% of small mammals trapped were deer mice. Deer mice were less common on the low-severity block, although they were still the most frequent small mammal species. Red-tailed chipmunks favored the low-severity burn; they were twice as frequent in the low-severity block as in the high-severity block (figure 5). Long-tailed voles were mostly trapped on the low-severity block [6].

Figure 7—Numbers of small mammal species trapped on Newman Ridge, by species and year, on South-3 and block North-3. Block South-3 was clearcut in June 1969 and burned by high-severity prescribed fire in September 1970. Block North-3 was clearcut in June 1969 and burned by low-severity prescribed fire in September 1970. "Chipmunk" refers to red-tailed chipmunk [6].

These studies found that after clearcutting, broadcast burning in spring or early summer removed little duff; therefore, early-season fires did not create the mineral soil seedbed needed for establishment of early-seral conifers. Late summer and early fall fires on south-, east-, and west-facing slopes consumed more duff and exposed mineral soil most effectively. On north-facing slopes, only late-season fires are likely to remove enough duff to allow for establishment of early-seral conifers [3,15]. When a significant rainfall event occurs late in the fire season, DeByle [3] recommended waiting several days before burning, to let duff and fine fuels dry. In dry years, water content of fine fuels should be within 10% to 17% for safe and effective burning [3].

Fire of at least moderate severity was needed to prepare a mineral soil seedbed. On Newman Ridge, moderate-severity fires removed most of the duff and prepared an adequate seedbed. At Miller Creek, moderate-severity fire exposed less mineral soil because the duff was thicker and wetter, so fewer conifer seedlings established [15]. Norum [10] presents a table to predict duff consumption from prescribed fire based on duff moisture content in the lower half of the duff layer.

Timing of burning can be coordinated with seed dispersal to enhance conifer establishment. Time of burning is not critical in a poor seed crop year, but the researchers recommended against burning after mid-September in a good seed crop year. Fires set later than mid-September may kill newly dispersed seed [3,15].

Prescribed fires had little effect on roots and rhizomes buried more than 1 inch (2.5 cm) below the mineral soil surface. These studies suggests that severe surface or ground fires are needed to kill roots lying >2 inches (5 cm) below the mineral soil surface [3,15].

These studies confirm that shade-tolerant species such as grand fir can establish in early postfire succession, and they may interfere with establishment and growth of fire-dependent, seral species. On dry sites, however, these species may show greater mortality and slower growth rates than fire-tolerant seral species over time. Shearer [15] noted that artificial regeneration of ponderosa pine, lodgepole pine, and western larch was most successful when seedlings were planted 1 or 2 years after fire. With an early start, seedlings of these seral species became dominant over shade-tolerant seedlings. Sites planted later were more likely to be dominated by Douglas-fir, firs, and Engelmann spruce. Faster growth of seral species was particularly evident on sites with snowbrush ceanothus seedlings. Nitrogen-fixing ability of snowbrush ceanothus might have increased soil fertility, increasing growth rates of seral conifers [15].

In general, sites that had a mineral soil seedbed and were relatively free of vegetation favored establishment of western larch [9]. The light, windborne seeds of western larch blew into in the center of large (~50 acres (20 ha)) clearcut-and-burned sites, so western larch comprised a relatively high proportion of the seedlings in burn interiors. However, planting may be necessary on steep, harsh sites [3]. Although western larch showed poor initial establishment in this study, it was the fastest-growing species overall [15]. On favorable sites, western larch seedlings established quickly and grew taller than other vegetation. Where establishment was delayed, shrubs suppressed growth of western larch seedlings [9].

REFERENCES:

1. Barrett, S.; Havlina, D.; Jones, J.; Hann, W.; Frame, C.; Hamilton, D.; Schon, K.; Demeo, T.; Hutter, L.; Menakis, J. 2010. Interagency fire regime condition class guidebook (FRCC), [Online], (Version 3.0). In: Interagency fire regime condition class website. U.S. Department of Agriculture, Forest Service; U.S. Department of the Interior; The Nature Conservancy (Producers). Available: https://www.frames.gov/files/7313/8388/1679/FRCC_Guidebook_2010_final.pdf [2017, March 1]. [85876]

2. Beaufait, William R.; Hardy, Charles E.; Fischer, William C. 1977 [Revised]. Broadcast burning in larch-fir clearcuts: The Miller Creek-Newman Ridge study. Res. Pap. INT-175. Ogden, UT: U.S. Department of Agriculture, Forest Service, Intermountain Forest and Range Experiment Station. 53 p. [11114]

3. DeByle, Norbert V., ed. 1981. Clearcutting and fire in the larch/Douglas-fir forests of western Montana--a multifaceted research summary. Gen. Tech. Rep. INT-99. Ogden, UT: U.S. Department of Agriculture, Forest Service, Intermountain Forest and Range Experiment Station. 73 p. [7270]

4. DeByle, Norbert V.; Packer, Paul E. 1972. Plant nutrient and soil losses in overland flow from burned forest clearcuts. National Symposium on Watersheds in Transition. 1972: 296-307. [8627]

5. DeByle, Norbert V.; Packer, Paul E. 1981. Soils and watershed. In: DeByle, Norbert V., ed. Clearcutting and fire in the larch/Douglas-fir forests of western Montana--A multifaceted research summary. Gen. Tech. Rep. INT-99. Ogden, UT: U.S. Department of Agriculture, Forest Service, Intermountain Forest and Range Experiment Station: 47-52. [18632]

6. Halvorson, Curtis H. 1981. Small mammal populations. In: DeByle, Norbert V., ed. Clearcutting and fire in the larch/Douglas-fir forests of western Montana--A multifaceted research summary. Gen. Tech. Rep. INT-99. Ogden, UT: U.S. Department of Agriculture, Forest Service, Intermountain Forest and Range Experiment Station: 41-46. [18631]

7. LANDFIRE Rapid Assessment. 2005. Reference condition modeling manual (Version 2.1). Cooperative Agreement 04-CA-11132543-189. Boulder, CO: The Nature Conservancy; U.S. Department of Agriculture, Forest Service; U.S. Department of the Interior. 72 p. On file at: U.S. Department of Agriculture, Forest Service, Rocky Mountain Research Station, Fire Sciences Laboratory, Missoula, MT. [66741]

8. LANDFIRE. 2008. CONUS refresh (LANDFIRE 1.1.0). Biophysical settings layer, LANDFIRE data distribution site, [Online]. In: LANDFIRE. U.S. Department of the Interior, Geological Survey (Producer). Available: https://landfire.cr.usgs.gov/viewer/ [2017, January 10]. [89416]

9. Latham, Penelope A.; Shearer, Raymond C.; O'Hara, Kevin L. 1998. Miller Creek Demonstration Forest--A forest born of fire: A field guide. Gen. Tech. Rep. RMRS-GTR-7. Ogden, UT: U.S. Department of Agriculture, Forest Service, Rocky Mountain Research Station. 68 p. [+ supplementary information]. [28988]

10. Norum, Rodney A. 1981. Fire behavior and effects. In: DeByle, Norbert V., ed. Clearcutting and fire in the larch/Douglas-fir forests of western Montana--A multifaceted research summary. Gen. Tech. Rep. INT-99. Ogden, UT: U.S. Department of Agriculture, Forest Service, Intermountain Forest and Range Experiment Station: 17-18. [18629]

11. Ryan, Kevin C.; Noste, Nonan V. 1985. Evaluating prescribed fires. In: Lotan, James E.; Kilgore, Bruce M.; Fischer, William C.; Mutch, Robert W., technical coordinators. Proceedings--symposium and workshop on wilderness fire; 1983 November 15-18; Missoula, MT. Gen. Tech. Rep. INT-182. Ogden, UT: U.S. Department of Agriculture, Forest Service, Intermountain Forest and Range Experiment Station: 230-238. [12456]

12. Shearer, R. C. 1989. Seed and pollen cone production in Larix occidentalis. In: Turnbull, J. W., ed. Tropical Tree Seed Research: ACIAR Proceedings no. 28, Proceedings of an International Workshop. 1989 August 21-24; Queensland, Australia. ACIAR Proceedings No. 28. Canberra, Australia: Australian Centre for International Agricultural Research: 14-17. [15534]

13. Shearer, Raymond C. 1975. Seedbed characteristics in western larch forests after prescribed burning. Res. Pap. INT-167. Ogden, UT: U.S. Department of Agriculture, Forest Service, Intermountain Forest and Range Experiment Station. 26 p. [12342]

14. Shearer, Raymond C. 1976. Early establishment of conifers following prescribed broadcast burning in western larch/Douglas-fir forests. In: Proceedings, Tall Timbers fire ecology conference and fire and land management symposium; 1974 October 8-10; Missoula, MT. No. 14. Tallahassee, FL: Tall Timbers Research Station: 481-500. [12499]

15. Shearer, Raymond C. 1981. Silviculture. In: DeByle, Norbert V., ed. Clearcutting and fire in the larch/Douglas-fir forests of western Montana--A multifaceted research summary. Gen. Tech. Rep. INT-99. Ogden, UT: U.S. Department of Agriculture, Forest Service, Intermountain Forest and Range Experiment Station: 27-31. [12819]

16. Shearer, Raymond C. 1982. Establishment and growth of natural and planted conifers 10 years after clearcutting and burning in a Montana larch forest. In: Baumgartner, David M., ed. Site preparation and fuels management of steep terrain: Proceedings of a symposium; 1982 February 15-17; Spokane, WA. Pullman, WA: Washington State University, Cooperative Extension: 149-157. [12818]

17. Shearer, Raymond C. 1984. Effects of prescribed burning and wildfire on regeneration in a larch forest in northwest Montana. In: New forests for a changing world: Proceedings, Society of American Foresters convention; 1983 October 16-20; Portland, OR. Washington, DC: Society of American Foresters: 266-270. [6730]

18. Shearer, Raymond C. 1989. Fire effects on natural conifer regeneration in western Montana. In: Baumgartner, David M.; Breuer, David W.; Zamora, Benjamin A.; Neuenschwander, Leon F.; Wakimoto, Ronald H., comps. Prescribed fire in the Intermountain region: Forest site preparation and range improvement: Symposium proceedings; 1986 March 3-5; Spokane, WA. Pullman, WA: Washington State University, Department of Natural Resources, Cooperative Extension: 19-33. [11242]

19. Shearer, Raymond C.; Halvorson, Curtis H. 1967. Establishment of western larch by spring spot seeding. Journal of Forestry. 65: 188-193. [12501]

20. Shearer, Raymond C.; Stickney, Peter F. 1991. Natural revegetation of burned and unburned clearcuts in western larch forests of northwest Montana. In: Nodvin, Stephen C.; Waldrop, Thomas A., eds. Fire and the environment: Ecological and cultural perspectives: Proceedings of an international symposium; 1990 March 20-24; Knoxville, TN. Gen. Tech. Rep. SE-69. Asheville, NC: U.S. Department of Agriculture, Forest Service, Southeastern Forest Experiment Station: 66-74. [16635]

21. Stickney, Peter F. 1985. Initial stages of a natural forest succession following wildfire in the Northern Rocky Mountains, a case study. In: Lotan, James E.; Kilgore, Bruce M.; Fischer, William C.; Mutch, Robert W., technical coordinators. Proceedings--symposium and workshop on wilderness fire; 1983 November 15-18; Missoula, MT. Gen. Tech. Rep. INT-181. Ogden, UT: U.S. Department of Agriculture, Forest Service, Intermountain Research Station: 383-384. [7367]