The response of ecosystem carbon cycle to temperature is one of the major topics in the research field of global change ecology.The general pattern of the response of gross primary productivity(GPP) to temperature usually shows that GPP increases with temperature at the lower temperature range to reach a maximum value (GPP_(max)),and then declines as temperature increases further.Thus,GPP_(max) represents the photosynthetic potential of vegetation at the optimum temperature.However,our understanding on the spatial and temporal patterns and main driving factors of GPP_(max) in forest ecosystems are still limited.In this study,we analyzed the temporal and spatial distribution of GPP_(max) and main influencing factors in five typical forest ecosystems based on flux data (temperate coniferous and broad-leaved mixed forest of Changbaishan,subtropical Chinese fir (Cunninghamia lanceolata) plantation of Huitong,subtropical evergreen coniferous forest of Qianyanzhou,subtropical evergreen broad-leaf and coniferous mixed forest of Dinghushan,and tropical monsoon forest of Xishuangbanna) along the North-South Transect of Eastern China (NSTEC),which covered tropical,subtropical,and temperate climate zones.The results showed that the temperature response of GPP showed a unimodal pattern,with GPP_(max) occurring at the optimum temperature in each year for all ecosystems.GPP_(max) at the optimum temperature in forests were ranked following the order:Changbaishan >Qianyanzhou >Xishuangbanna>Huitong >Dinghushan.Temperature played the most important role in driving the spatial variation of GPP_(max) across sites,with GPP_(max) decreasing with the increases of temperatur.Solar radiation,precipitation and VPD affected GPP_(max).For the interannual variation of GPP_(max) in each site,GPP_(max) in Changbaishan was mainly controlled by air temperature and by soil water content in Huitong,Qianyanzhou,and Dinghushan forests.We failed to find the main factors affecting interannual variation of tropical rainforest in Xishuangbanna.Our results benefit the understanding of GPP variation under climate change and provide evidence and parameter for accurate simulation of carbon cycle.