LAUSR

Abstract Agriculture is greatly influenced by climate change, especially in developing countries. Farmers are both the basic executors of agricultural adaptation and among the most vulnerable groups to climate change. However, the understanding on climate change perception and adaptive behavior of local farmers is still very limited. This study develops a binary logistic regression (BLR) model to explore the underlying principles of local farmer's perception and adaptive behavior toward climate change, with a focus on influential farmers. Through a field survey with 117 head farmers from 89 farmers’ cooperatives in Chongming Island of China - the largest alluvial island in the world, we found that: 1) 92% of the respondents thought climate change is happening; 59% of the respondents thought climate change has an obvious impact on agriculture production; and 45% of the respondents had a plan for future adaptation. 2) Based on BLR analysis, it was found that 3 out of 16 factors have significant impacts on head farmer’s adaptive behavior toward climate change, including agricultural training, perceived temperature change, and education level. 3) The percentage of consistency(POC) is proposed to describe the performance of our BLR model. The overall POC of BLR model is 68.4%, with a higher POC (85.7%) for farmers who have no adaptive behavior. 4) Key measures to enhance local climate adaptation include an integrated and coordinated plan of farmer-level adaptation, tailored and specialized training programs, improvement of scientific research, and publicity of best practices.