Pioneer plant communities five years after the 1988 Yellowstone fires
Loading...
Date
1995
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Montana State University - Bozeman, College of Letters & Science
Abstract
The Yellowstone fires of 1988 burned many different types of vegetation. This initiated secondary succession in environments from valley bottoms to alpine tundra. Five years after fire, plant communities were measured. Species presence was recorded in 100 m^2 macroplots and cover was sampled in twenty 1000 cm^2 quadrats. Pioneer community composition after severe fire in late-seral
vegetation was compared across the elevational gradient in nine environmental types with three replications in each. In two of the subalpine fir environments, communities arising from four different
pre-fire serai stages were sampled to test the hypothesis that pioneer community compostion differs when early-seral versus late-seral forests burn in one environmental type. Plant cover tends to decrease with increasing elevation. Along the elevational gradient, the wet grasslands had the strongest recovery from fire (plant cover averaged 97%), while the lowest cover was
in the subalpine zone near treeline (39% average cover). Species richness was between 32 and 42 species per 0.01 hectare in the seven lowest environmental types. Diversity in the two highest elevational environmental types was distinctly low (19 and 20 species/0.01 hectare, respectively). Forty-two of the 262 species identified occurred in nearly all environments. Many of the others were concentrated in various portions of the gradient (i.e. grasslands, montane forests, subalpine fir forests). Each species and its distribution was tabulated. To test the hypothesis that pioneer communties were influenced by previous vegetation, ordinations (principal component analysis and principal coordinate analysis) were conducted on postfire communities representing four pre-fire serai stages. Neither method indicated communities arising from any pre-fire serai stages were distinct from any others. Chi-square goodness-of-fit to random distribution and Monte Carlo randomizations of individual species in these environmental types identified only three species that were significantly non-randomly distributed among postfire communities from pre-fire serai stages. All three were more strongly represented in pioneer communities from early prefire serai stages. Eighteen species in each environmental type possibly had non-random distributions (P=0.06 to 0.15) indicating they may deserve further study.