资源环境科技发展态势分析平台

Understanding the influences of global climate change on soil microbial communities is essential in evaluating the terrestrial biosphere's feedback to this alarming anthropogenic disturbance. However, little is known about how intra-site historical climate variability can mediate the influences of current climate differences on community dissimilarity and assembly. To fill this gap, we examined and disentangled the interactive effects of historical climate variability and current climate differences on the soil bacterial community dissimilarity and stochasticity of community assembly among 143 sites from 28 forests across eastern China. We hypothesize that the relative importance of stochasticity and community dissimilarity are related to historical climate variability and that an increasing sum of intra-site historical variability enhances stochasticity while reduces dissimilarity between two communities. To test our hypothesis, we statistically controlled for covariates between sites including differences in soil chemistry, plant diversity, spatial distance, and seasonal climate variations at annual timescales. We observed that an increase in inter-site current climate differences led to a reduced impact of stochasticity in community assembly and a pronounced divergence between communities. In stark contrast, when communities were subjected to a high level of intra-site historical climate fluctuation, the observed impact incurred from current climate differences was substantially weakened. Moreover, the influence of increased historical variability was consistent along the gradient of current temperature differences between sites. However, effects induced by historical fluctuation in precipitation were disproportional and only evident when small inter-site differences were observed. Consequently, if the prior climate variability is ignored, especially regarding environmental factors like temperature, we assert that the influence current climate differentiation has on regulating community dissimilarity and assembly stochasticity will be underestimated. Together, our findings highlight the importance and need of explicitly controlling the mean and the historical variability of climate factors for the next “generation” of climate change experiments to come.