Vegetation changes in infertile grasslands occur only after decades of fire frequency differences

Background/Question/Methods Prairie ecology research indicates that fire is one of the key factors determining vegetation composition and that fire can change ecosystem nitrogen cycling and species composition. In addition to eliminating woody species, fire can directly impact herbaceous vegetation by eliminating fire-intolerant species, altering competitive interactions among species, and eliminating aboveground litter. All of these factors impact competition among species and ultimately the structure of the plant community. In the long term, fire has the potential to decrease ecosystem nitrogen pools through volatilization of the nitrogen present in aboveground tissues. We used an experiment with four different fire frequencies (Annual, two-year, four-year and control fire), replicated 6 times. This experiment was located in infertile grassland with annual aboveground productivity ranging from 60 to 140 grams per m2 at Cedar Creek LTER. This experiment was established in 1983 and burning started in 1984 We sampled the vegetation composition, diversity and carbon and nitrogen pools and fluxes in 2000 (after 17 years) and 2010 (after 27 years).

Results/Conclusions We found that 27 years of different burning frequencies had only a minor impact on the vegetation composition. We found no change in the most abundant species, Schizachyrium scoparium, nor diversity. Poa pratensis decreased from 10% of the plant biomass to less than 2%, but only in the annual and two-year burns. The main impact was found in the increasing abundance of the legume Lespedeza capitata, which was absent initially. The first individual Lespedeza capitata was found after 8 years. It reached 3% of the plot total biomass after 17 years and 30% after 27 years in the annual burned treatment. We found no change in any pool or flux of carbon or nitrogen, even after 27 years. This result contrasts with previous studies in savannas and more productive grasslands, where the balance between trees, grasses, and the elimination of the litter layer can result in large vegetation changes over a relatively short time period. In this grassland, primary productivity was low, litter did not accumulate, and the increase in one legume species only occurred after decades. Thus, our results indicate that examining fire impacts on vegetation and ecosystems in infertile grasslands, requires decade-long experiments and that the rate of change is likely related to ecosystem productivity.