Effects of Late Summer Fire on Tallgrass Prairie Microclimate and Community Composition

Two tallgrass prairie sites with different grazing histories and fuel accumulation were burned to investigate vegetation response to wild fires of differing intensity. Headfires were ignited 5 September 1985 under wildfire weather conditions (37 C air temperature, 36%7o relative humidity and wind gusts to 40 km/h). One of the burned plots had not been grazed for the past 3 years and had 1000 g/m2 fine fuel load; the other had been grazed moderately and had 450 g/m2 fine fuel. Fire intensity at the soil surface on the high fuel-load plot was four times that on the the low fuel burn plot. Microhabitat and vegetation responses were monitored the year following burning. Measured vegetative parameters, including tiller density, basal area and production, indicated that rhizomatous tallgrasses revegetated both burned plots quickly. Bunchgrasses decreased in both basal area and production on the high fuel plot. Community composition and productivity differences persisted 1 yr after the fires and were mainly attributable to bunchgrass mortality and microhabitat improvement for ruderal species. The data suggest that intensity is an important factor to consider when developing disturbance theory for tallgrass prairie. INTRODUCTION The tallgrass prairie ecosystem is often cited as an example of a fire-tolerant system which evolved with as a part of its natural disturbance regime (Adams et al., 1982). Tallgrass prairie is not regarded as a strict fire type but it has been shown that protection results in loss of productivity by the dominant grasses (Weaver and Rowland, 1952; Aikman, 1955; Ehrenreich, 1959; Hulbert, 1969; Anderson et al., 1970) and eventual replacement by woody species (Weaver, 1968; Bragg and Hulbert, 1976). Aboriginal and lightning-set fires are believed to have maintained the northern and eastern borders of this grassland well into climatic regions where forest would be considered climax vegetation (Vogl, 1974; Vankat, 1979; Axelrod, 1985). Most controlled experimental fires in tallgrass prairie have been conducted in the spring for a variety of reasons, including: (1) Researchers have long been interested in the increased productivity and renewed vigor of dominant grasses such as Andropogon gerardii Vit. following a spring burn (Curtis and Partch, 1950; Dix and Butler, 1954; Ehrenreich, 1959; Kucera and Ehrenreich, 1962; Hadley and Kieckhefer, 1963; McMurphy and Anderson, 1965; Old, 1969; Kucera, 1970; Launchbaugh and Owensby, 1978; Knapp, 1984a); (2) spring burns can be conducted in relative safety (Anderson, 1982; Wright and Bailey, 1982); (3) it had been incorrectly perceived that in other seasons, particularly summer, was not possible because of green or insufficient fuel (Bragg, 1982) and low probability of lighting-caused ignition (Hulbert, 1973), and (4) summer is not desirable because it removes forage valued for livestock grazing (Bragg, 1982). Prescribed spring burning does little damage to dominant warm-season grasses (C4) while reducing the vigor of cool-season plants (C3), thus giving the vegetation a boost toward theoretical tallgrass climax. This is a much studied and welldocumented phenomenon (Anderson et al., 1970). Additionally, there is a stimulatory effect on the rhizomatous tallgrasses due to the fire, by the removal of litter and timely microclimate improvement (Old, 1969; Knapp, 1984a). However, the intensity of spring fires may differ from that of fires occurring in summer and autumn. Historical accounts (Higgins, 1986) show that 85% of the 'This is journal article no. 5251 of the Oklahoma Agricultural Experiment Station.