Assessing the potential of species loss caused by deforestation in a mature subtropical broadleaf forest in central China

Deforestation is a major type of land use change to accommodate growing population, especially in developing countries. The risk of diversity loss due to habitat loss can be estimated using the species-area relationship based on abundance of each species. However, deforestation often occurs before there is any understanding of the impact of deforestation on tree diversity. Here, we assessed the potential effect of forest habitat destruction on the loss of species richness in a mature subtropical broadleaf forest in central China. We surveyed and constructed the species-area relationship for 54 400 m(2) plots, and simulated habitat loss scenarios by randomly and aggregately sampling plots. Rank-abundance of the 21 tree species was best fitted by the Zipf-Mandelbrot model, and our sample size was sufficient by the criterion of Hill numbers at orders q = 0, 1, and 2. We found that the number of species lost due to habitat loss was well predicted by the random placement species loss-area loss curve, and was lower than that due to aggregated habitat destruction by less than one species. The probability of losing one species reached 40% when losing 16 plots by aggregated sampling, 10 plots fewer than that by random sampling. Moreover, the probability of losing two species was 10-22 % higher by aggregately sampling than that by randomly sampling when losing 17 - 34 plots (0.68 - 1.36 ha). Considering that aggregated deforestation is common in reality, the results imply that the number of tree species lost due to deforestation could be higher than the theoretical estimation. Our study suggests the importance of assessing the impact of deforestation on tree diversity before selective logging in subtropical forests.