| Abstract: | Density counts were taken for 37 plant species on 12 plots which were assigned to three treatments: spring burn, autumn burn, and control. Post-fire sampling was carried out at intervals of up to 6 years when a second fire treatment was applied, followed by further sampling at 1 year. Most pre-fire species returned by 12 months after fire. Initial recovery was slower on the autumn burn plots and these were surpassed after 12 months by the spring burn plots which remained significantly more diverse until 6 years after fire when the treatments converged. The second fire led to further loss of species, especially on the autumn burn plots. Overall, numbers of individuals were lower in the post-fire community and did not fully recover during the 6 years before the second fire. Of the 37 species followed over the study period, eight had population numbers consistently below pre-fire values, seven showed better recovery on the autumn burn plots, eight showed better recovery on the spring burn plots, and fourteen had population numbers equal to or greater than pre-fire values. Most of the species not fully recovering had relatively poor survival rates. The better recovery rate of species on the autumn burn plots was attributable to better recruitment while the species recovering on the spring burn plots showed better survival and recruitment. The species increasing their numbers after the fire did so through good recruitment or survival, or both. For some species post-fire rainfall or temperature was significantly correlated with greater post-fire recruitment; others were more strongly affected by the treatment itself or more directly by the somewhat different fire intensities in spring and autumn, the differences in soil and litter moisture content, or seasonal variations in soil seed populations. |
