Fire Effects Information System (FEIS)

FEIS Home Page

Research Project Summary: Herbaceous responses to seasonal burning in experimental tallgrass prairie plots

• RESEARCH PROJECT SUMMARY CITATION
• INTRODUCTION
• STUDY LOCATION
• SITE DESCRIPTION
• PLANT PHENOLOGY
• FIRE SEASON/SEVERITY CLASSIFICATION
• RESULTS FROM INDIVIDUAL STUDIES
• FIRE MANAGEMENT IMPLICATIONS
• APPENDIX: SPECIES INCLUDED IN THIS SUMMARY
• REFERENCES

Copeland, Tanya E.; Sluis, William; Howe, Henry F. 2002. Fire season and dominance in an Illinois tallgrass prairie restoration. Restoration Ecology. 10(2): 315-323 [1].

Howe, Henry F. 1994. Response of early- and late-flowering plants to fire season in experimental prairies. Ecological Applications. 4(1): 121-133 [3].

Howe, Henry F. 1995. Succession and fire season in experimental prairie plantings. Ecology. 76(6): 1917-1925 [4].

Howe, Henry F. 1999. Response of Zizia aurea to seasonal mowing and fire in a restored prairie. The American Midland Naturalist. 141(2): 373-380 [5].

Howe, Henry F. 2000. Grass response to seasonal burns in experimental plantings. Journal of Range Management. 53(4): 437-441 [6].

INTRODUCTION:
This Research Project Summary covers 4 studies, all investigating the responses of early- and late-flowering herbaceous species to spring and summer fires in sites where tallgrass prairie species had been planted in former agricultural fields. The first study, described in 2 papers [3,4], focuses on grasses and forbs, both early- and late-flowering species. The second [6] focuses exclusively on dominant grasses. The third [5] focuses on a single perennial forb, golden zizia, as an example of an early-flowering subdominant species. The fourth [1] focuses on early- and late-flowering subdominant herbaceous species. Two hypotheses were investigated:

Hypothesis 1, referred to below as the "flowering-season hypothesis": Fire reduces the abundance of species that are experiencing their greatest growth at the time of burning. If this hypothesis is valid, then spring fires should reduce early-flowering species (including C3 grasses), and summer fires should reduce late-flowering species (including C4 grasses).

Hypothesis 2, referred to below as the "subdominant species hypothesis": Summer fires favor subdominant species. The authors describe 3 ways in which summer fire could improve growing conditions for subdominants: first, by reducing shade from late-flowering dominants, which are generally taller than early-flowering species [1]; second, by removing litter and providing openings that persist through most of the subsequent growing season [1,4]; third, by producing more variation in fire behavior and direct effects, thus increasing the number of species that could meet with favorable conditions in the postfire environment [1].

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

STUDY LOCATION:
The first 3 studies [3,4,5,6] were conducted near Viola, southwestern Wisconsin. The fourth study [1] was conducted at Goose Lake Prairie State Natural Area in Grundy County, northeastern Illinois.

SITE DESCRIPTION:
The first 3 studies [3,4,5,6] were conducted in a field on a flat bottomland floodplain classified as "wet-mesic" [6], where shallow standing water accumulates in early spring and drains from April to mid-May. The valley collects cold air, so the frost-free season is limited to mid-May through mid-September. Soils are loams with 3.4% organic matter and pH averaging 7.01 [3].

The 4th study [1] was conducted on sites with level topography, where much of the surrounding area is wet prairie or prairie marsh with poor natural drainage. Two soil types occur in the study area: a loam and a silty loam.

All studies used small treatment plots (maximum size 12 × 15 m [3,4]) in former agricultural sites that had been cleared of all vegetation, then planted with herbs native to tallgrass prairie. (However, some nonnative species may have been planted in the plots used for Study 2.) Before conversion to agriculture, the study sites probably experienced the historical fire regime described in Table 1:

PLANT PHENOLOGY:
These studies focus specifically on relationships between phenology and burn season. Spring fires were conducted as early as March 31 and as late as May 18, when "early-flowering" species (i.e., flower and set seed before mid-July) were likely to be most sensitive. Summer fires were conducted as early as July 15 and as late as September 3, when "late-flowering" species (i.e., flower and set seed after mid-July) were likely to be most sensitive. "Mid-flowering" species (i.e., flower and set seed June through August) were also observed. (Definitions of flowering periods are from [4].)

FIRE SEASON AND SEVERITY CLASSIFICATION:
Spring/summer, mixed severity, especially in summer burns [5]

RESULTS FROM INDIVIDUAL STUDIES:
Plant responses to fire were measured the growing season after the final burn treatment in each study. Preburn conditions and results of the 4 studies are summarized individually below. The Fire Management Implications section integrates results from all studies, discusses ways in which these studies are and are not consistent with the hypotheses above, and summarizes management implications of the research.

Study 1: Responses of early- and late-flowering prairie species to burning in different seasons
Study 2: Responses of early- and late-flowering grasses to burning in different seasons
Study 3: Response of an early-flowering, subdominant forb to burning in different seasons
Study 4: Responses of dominant and subdominant prairie species to burning in different seasons

Study 1: Responses of early- and late-flowering prairie species to burning in different seasons [3,4]

Prefire plant community: After existing vegetation was removed by herbicide application, plowing, and disking, the site was planted in 1986 with seed of more than 70 prairie species. By 1990, an additional 35 species had volunteered (not planted, either native or nonnative) on the site. The 12 most common species are discussed below. Despite the overall variety of species, some study plots were virtual monocultures.

Fire description: One-third of experimental plots were burned twice in the spring (1989 and 1992); one-third were burned twice in summer of the same years; and one-third were left unburned (Table 2).

Fire effects on plant community: Plant responses were measured the growing season after the each burn year (1990 and 1993).

The summer after the first burns (1990), 12 species occurred with ≥10% cover in at least 1 treatment, so their frequency and canopy cover were analyzed for different responses (Table 3). Several early-flowering species increased after 1 summer burn, as predicted; these include blackeyed Susan and quackgrass. However, another early-flowering species, reed canarygrass, did not respond positively to 1 summer burn. Switchgrass, a late-flowering species, decreased as predicted after 1 summer burn, but most other late-flowering species did not decrease.

After the 2nd set of burns, only 8 species were abundant enough to be analyzed quantitatively (Table 4). Consistent with Hypothesis 1 (that vulnerability to fire is greatest at the time of flowering), early-flowering species were more abundant on summer-burned than spring-burned plots. Responses of late-flowering species were less consistent with the flowering-season hypothesis. Switchgrass declined after burn treatments, but the decline was less severe on spring-burned than summer-burned plots. Big bluestem dominated most plots and increased in all treatments over time. Two late-flowering forbs, Canada goldenrod and white panicle aster, increased throughout the study and were not noticeably influenced by fire. The authors note that big bluestem and these 2 forbs are "vigorous cloners", which is likely to reduce their sensitivity to fire.

When species from the same flowering season were grouped for analysis, data for % cover were consistent with the flowering-season hypothesis: Cover of early-flowering herbs was greater after summer burns (32%) than after spring burns (2%) or in unburned treatments (4%) (P≤0.001). Cover of late-flowering herbs in spring burns and unburned treatments averaged 97% and 93%, respectively, and was significantly greater than in summer burns, where they retained dominance but with lower cover (66%) (P≤0.001).

Productivity of individual species was not consistent with the flowering-season hypothesis. In the growing season after the second set of burns, monoculture plots dominated by big bluestem had significantly more biomass than monoculture plots dominated by other species, with no significant differences between treatments (Table 5) [4].

Study 2: Responses of early- and late-flowering grasses to burning in different seasons [6]

Prefire plant community: After existing vegetation was removed by herbicide application, plowing, and disking, the site was planted in 1992 with 3 early-flowering C3 grasses (reed canarygrass, slender wheatgrass, and Virginia rye) and three late-flowering C4 grasses (big bluestem, prairie dropseed, and switchgrass). Spring and summer burns were conducted in 1995 and 1997; postfire responses were measured in summer 1998.

Fire description: One-third of experimental plots were burned twice in the spring (1995 and 1997); one-third were burned twice in summer of the same years; and one-third were left unburned (Table 6). Burns were conducted in the afternoon of nearly windless days when precipitation had been <1 cm in the previous 2 days. Ring-center firing was used for ignition. All burns killed or top-killed nearly all aboveground vegetation. The first 3 burns consumed most aboveground biomass, but the August 1997 fire left approximately 30% of dead vegetation standing, enough to provide shade after burning. The author mentions variation in fire behavior, caused by variation in abundance of green vegetation, as an important characteristic of summer fires in prairie ecosystems.

Fire effects on plant community: In the growing season after the 1st burn year, unburned plots had significantly more litter than burned plots (P<0.025). This difference did not appear to influence productivity.

The summer after the 2nd burns, unburned plots again had more litter than burned plots, 4 times as much as summer-burned plots. Productivity of C3 and C4 grasses was consistent with the flowering-season hypothesis: C3 grasses produced less biomass in spring-burned than summer-burned plots, and C4 grasses produced less biomass in summer-burned than spring-burned plots (P<0.05). Overall productivity was greatest on spring-burned plots, less on summer-burned plots, and least on unburned plots (P<0.05).

Species-specific results were consistent with the flowering-season hypothesis only for spring burns: Cover of two C4 species (big bluestem and switchgrass) increased in spring-burned plots while cover of reed canarygrass (a C3 species) decreased. Summer burning produced more variable results: Cover of reed canarygrass was expected to increase and cover of the C4 grasses to decrease, but instead, reed canarygrass codominated the summer-burned plots with the other C3 grasses and with C4 grasses.

Study 3: Response of an early-flowering, subdominant forb to burning in different seasons [5]

Prefire plant community: After existing vegetation was removed by herbicide, burning, plowing, and disking, a mixture of species native to tallgrass prairie was planted in 1990. The mixture included golden zizia, an early-flowering, nonrhizomatous, perennial forb. This species was used as an example of an early-flowering, subdominant, perennial forb native to tallgrass prairie. Plots were dominated by reed canarygrass from 1992 through the end of the study (1996). Plots were burned by an intense wildfire in April 1994. Experimental burns were conducted in spring and summer 1995; postfire responses were measured in summer 1996.

Fire description: One-third of experimental plots were burned in spring 1995, one-third were burned in summer of the same year, and one-third were left unburned. Each burn covered a 9 × 9 m plot and lasted 3 to 5 minutes (Table 7).

Fire effects on plant community: Results from this study are consistent with both Hypothesis 1, that plants are most vulnerable to fire when they are flowering, and Hypothesis 2, that summer fire favors subdominant species. Golden zizia occurred on more plots in summer-burned than spring-burned treatments, had more stems per plot, and produced more flowering stems. Frequency and density of golden zizia were least on unburned plots (Table 8). Observations on individual plots suggested that higher-intensity fire favored golden zizia: 64% of individuals in summer-burned plots occurred in the 2 plots with the greatest fireline intensity and little or no litter remaining after fire, and all flowering in the summer-burned plots occurred in the 2 high-intensity burn plots.

Study 4: Responses of dominant and subdominant prairie species to burning in different seasons [1]

Prefire plant community: Prairie species were planted in 1976, after existing vegetation had been removed. Between 1987 and 1996, the site was burned 5 times with spring fires. In 1996, C4 grasses (big bluestem, indiangrass, and switchgrass) were the dominant species.

Fire description: Experimental burns were conducted in 1997 (one-third of plots in spring, one-third in summer, and one-third left unburned), and vegetation was measured again in 1998. Fire spread rates were similar in all burns, averaging 1.3 m/min to cross a 10 × 10 m plot (Table 9).

Fire effects on plant community: Contrary to both Hypothesis 1, that plants are most vulnerable to fire when they are flowering, and Hypothesis 2, that summer fire favors subdominant species, summer burning favored big bluestem over all other species, both dominant and subdominant. Increased tillering of big bluestem was observed after both spring and summer burns, with a greater increase in summer-burned than spring-burned plots (Table 10). If this species' responses to summer fire were based only on phenology, it would be expected to decline after summer fire. The authors attribute big bluestem's increase to its ability to reproduce from extensive, sod-forming rhizomes and the fact that the late-summer burn removed litter for much of the subsequent growing season. The other late-flowering dominants (switchgrass and indiangrass) did not respond positively to burning in either spring or summer. Neither of these species produces dense sod like that of big bluestem. The authors indicate that switchgrass reproduces mainly from seed and basal buds; the FEIS review of indiangrass indicates it regenerates from seed and short rhizomes.

The effects of burn season on subdominant species were addressed according to flowering time (early- and late-flowering) rather than individual species. As predicted by Hypothesis 2, that summer fire favors subdominant species, species richness of subdominants increased in summer-burned plots regardless of flowering season, but not in spring-burned plots (Table 11). The similar responses of early- and late-flowering subdominants suggest that flowering season is not an important influence on subdominant species' responses to fire. Instead, their positive response to summer fire may be due to indirect effects, specifically reduction of litter during much of the subsequent growing season and reduction of shade from dominants.

Planted prairie species and volunteer species responded similarly to burn treatments in terms of species richness, indicating that burn treatments did not favor volunteer (likely undesired) species.

FIRE MANAGEMENT IMPLICATIONS:
Data from the first growing season after burning provided some evidence supporting the initial hypotheses but also demonstrated that responses to fire vary within flowering guilds and are often species-specific.

Regarding Hypothesis 1, that fire reduces the abundance of species that are experiencing their greatest growth at the time of burning: Results of spring burns were consistent with this hypothesis: Spring burns tended to reduce early-flowering dominant species and favor late-flowering ones (Studies 1, 2). Results of summer burns were more species-specific and not completely consistent with this hypothesis. Some late-flowering dominants declined after summer burns, but the C4 grasses as a group still tended to dominate, and strongly rhizomatous species (grasses and forbs) often increased (Studies 1, 2, 4).

Regarding Hypothesis 2, that summer fires favor subdominant species and increase species richness: Studies 2, 3, and 4 appeared to validate this hypothesis, although Study 3 addressed only a single species. These studies demonstrated 2 possible mechanisms for positive responses to summer burning. First, summer-burned plots tended to have less litter in the subsequent growing season than spring-burned plots, possibly offering more microsites for establishment and growth of subdominant species regardless of their flowering seasons. Second, summer burns often included patches of very low fire intensity due to green vegetation; the resulting spatial variation in biomass consumption and litter removal could favor a greater variety of species after fire. A third possible mechanism, making more light available to subdominants by reducing cover of late-flowering dominants, was not clearly demonstrated by these studies, since many dominant grasses, both early- and late-flowering species, responded positively to summer fire.

Based on historical patterns, management-imposed fire regimes may need to deliberately vary fire season and fire behavior in tallgrass prairie ecosystems. In the past, a varied array of burn seasons, burn intervals, and grazing disturbances almost certainly produced a greater variety of species assemblages with more biodiversity than any single management regime (such as repeated dormant-season fires) could now produce [3]. Spring and fall fires can suppress woody species, but burning only in these seasons may favor C4 grasses and late-flowering, rhizomatous forbs over all early-flowering species and late-flowering subdominants, eventually reducing biodiversity [5]. Summer burns could enhance the frequency and richness of subdominant species [4] without compromising vigor of dominant late-flowering grasses. Furthermore, it may be possible to thus increase biodiversity without disproportionately encouraging species that are not native to prairie ecosystems [1].

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

REFERENCES:

1. Copeland, Tanya E.; Sluis, William; Howe, Henry F. 2002. Fire season and dominance in an Illinois tallgrass prairie restoration. Restoration Ecology. 10(2): 315-323. [42262]

2. Hann, Wendel; Havlina, Doug; Shlisky, Ayn; [and others]. 2008. Interagency fire regime condition class guidebook. Version 1.3, [Online]. In: Interagency fire regime condition class website. U.S. Department of Agriculture, Forest Service; U.S. Department of the Interior; The Nature Conservancy; Systems for Environmental Management (Producer). 119 p. Available: http://frames.nbii.gov/frcc/documents/FRCC_Guidebook_2008.07.10.pdf [2010, 3 May]. [70966]

3. Howe, Henry F. 1994. Response of early- and late-flowering plants to fire season in experimental prairies. Ecological Applications. 4(1): 121-133. [27810]

4. Howe, Henry F. 1995. Succession and fire season in experimental prairie plantings. Ecology. 76(6): 1917-1925. [48450]

5. Howe, Henry F. 1999. Response of Zizia aurea to seasonal mowing and fire in a restored prairie. The American Midland Naturalist. 141(2): 373-380. [30533]

6. Howe, Henry F. 2000. Grass response to seasonal burns in experimental plantings. Journal of Range Management. 53(4): 437-441. [77370]

7. 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]

8. 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]