Research Project Summary: Changes in grassland vegetation following fire in northern Idaho



Common names are used throughout this summary. For a complete list of the common and scientific names of species discussed in this summary and for links to FEIS species reviews, see the Appendix.

RESEARCH PROJECT SUMMARY AUTHORSHIP AND CITATION:
Gucker, Corey. 2011. Research Project Summary: Changes in grassland vegetation following fire in northern Idaho. In: Fire Effects Information System, [Online]. U.S. Department of Agriculture, Forest Service, Rocky Mountain Research Station, Fire Sciences Laboratory (Producer). Available: http://www.fs.fed.us/database/feis/ [].

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

Gucker, Corey. 2004. Canyon grassland vegetation changes following the Maloney Creek wildfire. Moscow, ID: University of Idaho. 80 p. Thesis. [2].

Gucker, Corey L.; Bunting, Stephen C. 2011. Canyon grassland vegetation changes following fire in northern Idaho. Western North American Naturalist. 71(1): 97-105. [1].

STUDY LOCATION:
Research was conducted in the China and Corral Creek watersheds of the Garden Creek Nature Preserve in the Craig Mountain Wildlife Area. The Craig Mountain Wildlife Area is roughly 20 miles (40 km) south of Lewiston, Idaho, in southern Nez Perce and western Lewis counties [1,2].

SITE DESCRIPTION:
Study plots occurred in early, mid-, and late-seral communities within the bluebunch wheatgrass-Sandberg bluegrass-arrowleaf balsamroot habitat type [7]. Plots occupied southeastern to southwestern aspects with slopes of 16% to 52% at elevations of 2,612 to 3,432 feet (796-1,046 m).

PREFIRE PLANT COMMUNITY:
Within the study area, plots are dominated by bluebunch wheatgrass, Sandberg bluegrass, arrowleaf balsamroot, yellow starthistle, and brome grasses. Increased abundance of nonnative species including yellow starthistle and cheatgrass indicated early-seral conditions. Increased abundance of bluebunch wheatgrass, Sandberg bluegrass, and arrowleaf balsamroot was characteristic of late-seral plots. Typically, late-seral plots occurred on steep slopes where past grazing and farming impacts on the vegetation were less evident.

Study sites are classified in the following plant community and probably historically experienced the fire regime described below:

Table 1. Fire regime information on the vegetation community studied in this Research Project Summary. Fire regime characteristics are taken from the LANDFIRE Rapid Assessment Vegetation Model [5], which was developed by using expert opinion as documented in the PDF file linked from the Potential Natural Vegetation Group listed below. Cells are blank where information was not available in the Rapid Assessment Vegetation Model.
Northern and Central Rockies
Vegetation Community (Potential Natural Vegetation Group) Fire severity* Fire regime characteristics
Percent of fires Mean interval
(years)
Minimum interval
(years)
Maximum interval
(years)
Mountain grassland Replacement 60% 20 10  
Mixed 40% 30    
*Fire Severities:
Replacement=Any fire that causes greater than 75% top removal of a vegetation-fuel type, resulting in general replacement of existing vegetation; may or may not cause a lethal effect on the plants.
Mixed=Any fire burning more than 5% of an area that does not qualify as a replacement, surface, or low-severity fire; includes mosaic and other fires that are intermediate in effects.
Surface or low=Any fire that causes less than 25% upper layer replacement and/or removal in a vegetation-fuel class but burns 5% or more of the area [3,4].

PLANT PHENOLOGY:
Most species were dormant at the time of the mid-August fire.

FIRE SEASON AND SEVERITY CLASSIFICATION:
Summer, low severity

FIRE DESCRIPTION:
This study was conducted on burned and unburned sites that were studied for 2 years before [6] and 2 and 3 years following the Maloney Creek Wildfire. All prefire data were obtained from Robins [6]. Study objectives were to evaluate the effects of fire on native and nonnative species.

The lighting-ignited Maloney Creek Wildfire burned approximately 74,515 acres (30,155 ha) and a large portion of the Garden Creek Nature Preserve. Only plots in the China Creek Watershed burned. Plots burned in a low-severity backing fire on 20 August 2000. Humidity ranged from 27% to 33%, air temperature averaged 74.5 °F (23.6 C), and winds were 0 to 10.4 miles (16.7 km)/h with gusts of up to 21.9 miles (35.2 km)/h when plots in the China Creek Watershed burned.

FIRE EFFECTS ON PLANT COMMUNITY:
Species cover was averaged from a total of 14 unburned plots. There were 14 unburned plots: 4 early-seral, 5 midseral, and 5 late-seral. There were 11 burned plots: 2 early-seral, 6 midseral, and 3 late-seral. Seral stage of the communities was determined by species composition and relative abundance of native and nonnative species. Early-seral communities were dominated by yellow starthistle; late-seral communities were dominated by bluebunch wheatgrass. Midseral communities had moderate levels of yellow starthistle and bluebunch wheatgrass.

Native grasses: Bluebunch wheatgrass decreased briefly following fire. Bluebunch wheatgrass cover was significantly (P=0.035) lower on burned than unburned sites in the 2nd postfire year, but average cover on burned and unburned plots was not significantly different by the 3rd postfire year. In the 3rd postfire year, bluebunch wheatgrass coverage on burned plots exceeded prefire coverage. Sandberg bluegrass coverage was significantly greater on burned than unburned plots in the 2nd (P=0.045) and 3rd postfire years (P=<0.0001).

Nonnative grasses: Cheatgrass cover was significantly (P=0.002) greater on unburned than burned plots just before the fire (2000), but by the 3rd postfire year these differences were lost (P=0.343) due to large increases in cheatgrass cover on burned plots. Cover of Japanese brome increased on both burned and unburned plots following the fire, but cover was significantly (P=0.017) greater on unburned than burned plots only in the 2nd postfire year. Coverage on burned and unburned plots was not significantly (P>0.05) different in the 3rd postfire year.

Native perennial forbs: Individual fire responses were analyzed for arrowleaf balsamroot, silky lupine, and western yarrow. There were no significant differences in arrowleaf balsamroot coverage on burned and unburned plots in any year. However, average coverage did decrease from nearly 13% just before the fire to 6.8% in the 2nd postfire year. Silky lupine cover decreased slightly on unburned plots and increased on burned plots. By the 3rd postfire year, burned plots had significantly (P<0.0001) greater silky lupine coverage than unburned plots. Western yarrow decreased dramatically on burned plots. Western yarrow cover was significantly greater on burned than unburned plots (P=0.002 in 1999, P=0.014 in 2000) prior to burning. Due to decreases on burned plots, differences in western yarrow coverage on burned and unburned plots were lost after the fire.

Nonnative forbs: Yellow starthistle coverage was not significantly different on burned and unburned plots in any time before or after the fire. However, yellow starthistle coverage generally decreased over time on unburned plots and was relatively unchanged on burned plots.

Life form groups: Many species in the study area contributed low overall plot coverage and were analyzed as groups. Nonnative annual forb cover was significantly (P=0.018) lower on burned plots than on unburned plots prior to the fire, but by the 3rd postfire year coverage was significantly (P=0.003) greater on burned plots. Nonnative annual forb cover increased on both burned and unburned plots following the fire. Native annual forbs also increased on both burned and unburned plots after the fire, but native annual forb cover was significantly (P=0.025) greater on burned than unburned plots in the 3rd postfire year. Data were pooled for several perennial forbs, not analyzed individually (see Native perennial forbs). These forbs increased on both burned and unburned plots, and differences in perennial forb cover on burned and unburned plots were not significantly different. Increases were likely due in part to favorable postfire moisture conditions.

Table 2. Pre- and postfire cover on burned and unburned plots (for species with >0.5% average cover in at least 1 pre- or postfire year)
Year 1999 2000 2002 2003
Time since fire (years) pre pre 2 3
Common name

Average cover (%)

Native perennial forbs
western yarrow UB* 2.26a** 1.35a 1.77a 1.19a
B 4.03b 2.81b 1.32a 0.78a
waha milkvetch UB 0.03 0.24 0.31 0.03
B 0.44 0.76 1.25 0
Cusick's milkvetch UB 0.50 0.34 0.13 1.24
B 1.05 0.19 0.48 2.18
arrowleaf balsamroot UB 8.57a 6.81a 6.85a 6.54a
B 9.57a 12.95a 6.82a 7.40a
fernleaf biscuitroot UB 0.14 0.14 0.54 1.35
B 0.04 0.02 0.08 0.84
nineleaf biscuitroot UB 0.16 0.42 1.58 1.04
B 0.40 0.39 2.20 4.02
silky lupine UB 2.34a 2.39a 2.26a 1.31a
B 3.89a 4.45a 3.37a 6.93b
Snake River phlox UB 0.20a 0.27a 0.71a 0.39a
B 0.24a 0.17a 0.41a 0.09b
narrowleaf scullcap UB 0.81 1.62 2.66 1.87
B 0.19 0.33 0.47 0.36
shiny chickweed UB 0 0.07 0.25 0.76
B 0 0.02 0.97 2.18
meadow deathcamas UB 0.09 0.24 0.43 2.06
B 0.07 0.22 0.31 1.76
Nonnative annual forbs
thymeleaf sandwort UB 0.32 0.53 2.01 0.79
B 0.17 0.11 3.73 0.99
cutleaf filaree UB 0 0.04 0.28 1.33
B 0 0.03 0.4 1.46
strict forget-me-not UB 0 0.01 0.01 0.33
B 0.02 0.05 0.17 4.08
Lewiston cornsalad UB 0 0 0.23 0.05
B 0.04 0.04 0.99 0.01
Nonnative perennial forbs
yellow starthistle UB 13.91a 7.11a 9.52a 4.43a
B 7.13a 7.08a 6.32a 6.62a
Native perennial grasses
Sandberg bluegrass UB 0.27a 0.29a 0.39a 0.28a
B 0.17a 0.15a 1.67b 3.28b
bluebunch wheatgrass UB 20.10a 16.38a 25.28a 26.34a
B 17.74a 18.66a 15.56b 21.53a
Nonnative perennial grasses
bulbous bluegrass UB 0 0 0.08 0.07
B 0.05 0.09 1.11 1.04
Nonnative annual grasses
Japanese brome UB 6.99a 6.93a 13.02a 10.25a
B 9.95a 5.22a 6.75b 8.90a
cheatgrass UB 1.36a 1.46a 5.96a 6.26a
B 0.76a 0.22b 1.74b 7.73a
rattail sixweeks grass UB 0.08 0.42 1.70 0
B 0 0 0.05 0
* UB=unburned, B=burned.
**Burned and unburned coverage values with different subscripts within the same species or year are significantly (P<0.05) different.

Diversity and evenness: Species richness, Pielou's evenness, Simpson's diversity, and Shannon-Weiner's diversity were evaluated for burned and unburned plots. Species richness on burned and unburned plots was not significantly different in any study year before or after the fire. Species composition was relatively unchanged by fire. Pielou's evenness was not different before the fire but was significantly (P=0.001) greater on burned than unburned plots in the 3rd postfire year. Simpson's and Shannon-Weiner's diversity index were both significantly (P<0.03) greater on burned than unburned plots in the 2nd and 3rd postfire years. Increased evenness and relatively unchanged species richness on burned plots suggests that changes in Simpson's and Shannon-Weiner's diversity reflect increased homogeneity for species coverage and not an addition of species on burned plots.

Seral stage: For those species typically used to determine seral stage within the bluebunch wheatgrass-Sandberg bluegrass-arrowleaf balsamroot habitat type, differences in a given year between burned and unburned sites were evaluated within early, mid-, and late-seral plots. Sample sizes within a given seral stage were small.

Native species: There were no significant differences in bluebunch wheatgrass cover on burned and unburned plots in any year or seral stage. However, average bluebunch wheatgrass cover nearly doubled on early-seral burned plots. For all seral stages, the cover of Sandberg bluegrass was significantly (P<0.02) greater on burned than unburned plots in the 3rd postfire year. Arrowleaf balsamroot cover was variable, and high variances affected significance results. Average arrowleaf balsamroot cover decreased slightly but not significantly following the fire in all but the late-seral burned plots.

Nonnative species: Overall, yellow starthistle cover decreased on both burned and unburned plots (see Table 2), but slight increases occurred on mid- and late-seral burned plots when data were analyzed by seral stage (see Table 3). Cheatgrass increased on early-, mid-, and late-seral burned and unburned plots. In the late-seral plots, cheatgrass cover on unburned sites was significantly (P<0.01) greater than burned sites until the 3rd postfire year.

Table 3. Average pre- and postfire cover on early, mid-, and late-seral burned and unburned plots for species used to determine seral stage
Year 1999 2000 2002 2003
Time since fire (years) pre pre 2 3
 

Average cover (%)

bluebunch wheatgrass
Early UB* 7.95a** 10.85a 14.80a 15.54a
B 7.17a 7.31a 10.01a 17.32a
Mid UB 20.20a 14.39a 27.13a 27.95a
B 17.4a 19.09a 16.22a 23.36a
Late UB 29.7a 22.8a 31.82a 34.17a
B 25.5a 25.37a 17.92a 20.66a
Sandberg bluegrass
Early UB 0.24a 0.07a 0.22a 0.02a
B 0a 0.07a 1.32a 3.23b
Mid UB 0.16a 0.32a 0.37a 0.50a
B 0.15a 0.18a 1.43a 2.26b
Late UB 0.4a 0.44a 0.54a 0.26a
B 0.32a 0.16a 2.37a 5.35b
arrowleaf balsamroot
Early UB 6.77a 3.97a 2.35a 3.44a
B 14.9a 17.55a 5.65a 5.14a
Mid UB 7.54a 5.85a 9.42a 5.12a
B 10.8a 14.4a 8.30a 8.08a
Late UB 11a 10.03a 7.88a 10.44a
B 3.62a 7.0a 4.63a 7.55a
yellow starthistle
Early UB 40.6a 16.72a 22.68a 10.22a
B 32.0a 33.04a 21.1a 16.95a
Mid UB 6.33a 6.29a 7.57a 4.03a
B 2.42a 1.94b 4.46a 4.91a
Late UB 0.13a 0.23a 0.93a 0.19a
B 0.02a 0.06a 0.18a 3.17a
cheatgrass
Early UB 0.89a 0.58a 8.01a 10.70a
B 0.09a 0.04a 1.48a 10.59a
Mid UB 2.51a 2.52a 3.70a 4.65a
B 1.37a 0.34a 2.38a 9.07a
Late UB 0.59a 1.09a 6.57a 4.33a
B 0b 0.11b 0.63b 3.14a
* UB=unburned, B=burned.
**Burned and unburned coverage values with different letters within the same species, year, or seral stage are significantly (P<0.05) different.

FIRE MANAGEMENT IMPLICATIONS:
The summer fire in the Garden Creek Preserve canyon grasslands increased the cover of most native and nonnative species. In general, species composition and dominance were unchanged. Findings from this study will be useful in anticipating desirable and undesirable changes in native and nonnative species after fire. Based on the findings from this study, fire exclusion from this area for the sake of weed management does not appear warranted.

APPENDIX: SPECIES INCLUDED IN THIS SUMMARY
This Research Project Summary contains fire effects and/or fire response information on the following species. For more information, follow the highlighted links to the FEIS reviews of those species.

Common name Scientific name
forbs
western yarrow Achillea millefolium
thymeleaf sandwort Arenaria serpyllifolia
waha milkvetch Astragalus arthurii
Cusick's milkvetch Astragalus cusickii
arrowleaf balsamroot Balsamorhiza sagittata
yellow starthistle Centaurea solstitialis
cutleaf filaree Erodium cicutarium
fernleaf biscuitroot Lomatium dissectum
nineleaf biscuitroot Lomatium triternatum
silky lupine Lupinus sericeus
strict forget-me-not Myosotis stricta
Snake River phlox Phlox colubrina
narrowleaf scullcap Scutellaria angustifolia
shiny chickweed Stellaria nitens
Lewiston cornsalad Valerianella locusta
meadow deathcamas Zigadenus venenosus
graminoids
Japanese brome Bromus japonicus
cheatgrass Bromus tectorum
bulbous bluegrass Poa bulbosa
Sandberg bluegrass Poa secunda
bluebunch wheatgrass Pseudoroegneria spicata
rattail sixweeks grass Vulpia myuros

REFERENCES:


1. Gucker, Corey L.; Bunting, Stephen C. 2011. Canyon grassland vegetation changes following fire in northern Idaho. Western North American Naturalist. 71(1): 97-105. [82531]
2. Gucker, Corey. 2004. Canyon grassland vegetation changes following the Maloney Creek wildfire. Moscow, ID: University of Idaho. 80 p. Thesis. [51512]
3. Hann, Wendel; Havlina, Doug; Shlisky, Ayn; [and others]. 2010. Interagency fire regime condition class (FRCC) guidebook, [Online]. Version 3.0. In: FRAMES (Fire Research and Management Exchange System). National Interagency Fuels, Fire & Vegetation Technology Transfer (NIFTT) (Producer). Available: http://www.fire.org/niftt/released/FRCC_Guidebook_2010_final.pdf. [81749]
4. LANDFIRE Rapid Assessment. 2005. Reference condition modeling manual (Version 2.1), [Online]. In: LANDFIRE. Cooperative Agreement 04-CA-11132543-189. Boulder, CO: The Nature Conservancy; U.S. Department of Agriculture, Forest Service; U.S. Department of the Interior (Producers). 72 p. Available: http://www.landfire.gov/downloadfile.php?file=RA_Modeling_Manual_v2_1.pdf [2007, May 24]. [66741]
5. LANDFIRE Rapid Assessment. 2007. Rapid assessment reference condition models, [Online]. In: LANDFIRE. U.S. Department of Agriculture, Forest Service, Rocky Mountain Research Station, Fire Sciences Lab; U.S. Geological Survey; The Nature Conservancy (Producers). Available: http://www.landfire.gov/models_EW.php [2008, April 18] [66533]
6. Robins, Sandra S. 2001. Effects of perennial plant competition on the invasibility of canyon grassland communities by Centaurea solstitialis. Moscow, ID: University of Idaho. 54 p. Thesis. [66570]
7. Tisdale, E. W. 1986. Canyon grasslands and associated shrublands of west-central Idaho and adjacent areas. Bulletin No. 40. Moscow, ID: University of Idaho, College of Forestry, Wildlife and Range Sciences. 42 p. [Contribution No. 284; Forest, Wildlife and Range Experiment Station]. [2338]

FEIS Home Page