中文摘要：

以人类活动为主导的城市扩张和土地覆被变化对城市生态环境产生了重要影响,并与气候变化共同影响植被净初级生产力（NPP）,但目前从时空尺度上脱离气候干扰仅以人类活动为主导因素来定量分析其对植被NPP影响的研究尚不充分.本研究以广州市为研究区,利用CASA模型估算2001—2013年实际净初级生产力（NPPact）,结合CHIKUGO模型估算得到的潜在净初级生产力（NPPp）计算因土地覆被变化导致的NPP损失（NPPlulc）,并建立相对贡献指数（RCI）定量分析和评价在城市扩张过程中人类活动对NPP的影响.结果表明：2001—2013年间,广州总体及其5片区NPPact和NPPlulc分别呈减少和增加趋势,并存在明显的空间差异性;RCI呈明显增加趋势,东北片区RCI值最低,为0.31,表明气候变化是其NPP变化的主要原因,其他4个片区的RCI值均高于0.5,说明4个片区人为干扰严重,人类活动是其NPP减少的主导因素;广州市及其5片区的RCI变化斜率均大于0,人类活动对植被的干扰逐年增强,北部片区RCI变化斜率值最大（0.693）,人为干扰增加趋势最明显.

英文摘要：

Urban expansion and land cover changes driven primarily by human activities have sig- nificant influences on urban eco-environment, and together with climatic change jointly affect net primary productivity （NPP）. However, quantitative analysis about the impacts of human activities on NPP change isolated from climatic change in the spatiotemporal scale is poorly understood. We took Guangzhou City as study area to estimate the actual NPP （NPPact ） and the potential NPP （NPPp） from 2001 to 2013 based on Carnegie Ames Stanford Approach （CASA） model and CHIK- UGO model, and calculated the loss of NPP due to land use and land cover change （ NPPlulc ） from NPPact and NPPp. The impact of human activities on NPP in the process of urban sprawl was quanti- tatively analyzed and assessed by examining a relative contribution index （RCI） based on NPPp and NPPlulc. Guangzhou City and its five regions showed a declined trend of NPPact and an increased trend of NPPlulc from 2001 to 2013, and significant spatial differences of NPPact and NPPlulc were found in all regions. RCI had an increasing trend over 13 years, the smallest value of average RCI occurred in northeastern region （0.31） , indicating climatic change was the main cause of NPP change, while the average RCI was higher than 0.5 in the other four regions, indicating that these regions were subjected to severe anthropogenie disturbances and human activities were the dominant factors of NPP reduction. The slopes of RCI change were positive in Guangzhou and its five regions,revealing an increasing human disturbance trend. Northern region had the largest RCI slope of 0.693, suggesting the trend was most obviously in this region.