Abstract: Tree mortality is an important ecological process in forests. It is crucial to understand how tree mortality affects spatial patterns and interspecific associations for revealing the mechanisms of tree mortality and community dynamics. We employed the correlation method of spatial point pattern analysis to analyse the variations in spatial patterns and interspecific relations before and after mortality using data obtained from two surveys of a 25 hm² plot in the Mulun National Natural Reserve, China. The results showed that most species had an aggregated distribution both pre- and post-mortality. The proportion of species with aggregated distribution reduced slightly post-mortality compared with that for pre-mortality. Increases in the number of species with random distribution at small scale indicated that tree death was not random. At the species level, there were significant positive associations between dead and live trees of the 13 common species at different levels of 0-30 m range, suggesting weak intraspecific and interspecific competition among dominant species. Pre- and post-mortality interspecific associations were mostly positive, which remained stable during the period of two surveys for most species, indicating that the community had reached a rather stable stage. Following tree mortality, the number of species with positive associations increased at 1-30 m scales, whereas the number of species with negative and no associations decreased at most scales. These results indicated that the pressure of interspecific competition was relieved to some extent after individual death.

Key words: tree mortality, spatial point pattern analysis, spatial pattern, interspecific association, karst ecosystem