科研之友
微信
新浪微博
Facebook
分享链接
摘要
Litterfall is an important process that links vegetation and soil pools and plays an important role in the maintenance of soil fertility. Although studies indicated that climate will significantly affect forest litterfall, the role of biotic factors such as the spatial heterogeneity of forest age, remains unclear. In this study, we built an updated dataset of litterfall in China and explored the key drivers affecting forest litterfall by establishing optimal linear mixed models (OLMMs). The potential bias of models and their spatial patterns were then evaluated based on the OLMMs and remotely sensed and China's forest inventory data. The results showed the mean annual temperature (MAT) and forest age were the key drivers affecting forest litterfall. Abiotic factors and forest age and height together accounted for 77.5% of the variation in observed litterfall. Although forest age and height did not apparently enhance the coefficient of determination (R-2), these factors significantly decreased spatial errors. Therefore, if the model contains only climate factors and the spatial patterns of biotic factors are ignored, it will produce high spatial errors (-52% to 92%). In addition, when forest age and height were not considered, variation of litterfall explained by forest age was inappropriately attributed to MAT, which significantly overestimated the importance of climate factors on forest litterfall. Specifically, litterfall was overestimated for young forests and underestimated for old forests if the model did not contain forest age in China. Models that ignored forest age significantly overestimated the contribution of climatic factors on forest litterfall and produced high spatially specific errors. The comparison of the litterfall modeled by OLMMs and the remote sensing-based net primary production (NPP) indicated that litterfall and NPP are strongly dependent, and the ratio of litterfall to NPP linearly increased with forest age.
- 出版日期2019-4-15
- 单位北京师范大学
