关键词: 香禾糯, 气候变化, MaxEnt模型, 潜在适生区, 模型优化

Abstract: Based on the actual distribution data of Guizhou kam sweet rice in field records and literature, the Maximum Entropy Model (MaxEnt) optimised by the Kuenm packet was used to study the suitable regions of Guizhou kam sweet rice and the dominant climatic factors affecting the distribution of kam sweet rice under the historical and future climate scenarios, which provides a reference basis for rational planning of distribution of kam sweet rice production in the context of climate change. The results showed that: (1) during the historical period of 1970-2000, the potential suitable region of kam sweet rice were mainly concentrated in the southeastern part of Qiandongnan prefecture, Guizhou province, and the highly suitable regions were distributed in Congjiang, Rongjiang, and Liping counties. (2) Compared with the historical period, the highly suitable region of kam sweet rice under the SSP126 climate scenario increases in region and moves to higher altitude in 2041−2060, and decreases in region and moves to higher latitude in 2081−2100, and increases in region and moves to higher latitude in 2041−2060, and decreases in region and moves to higher latitude in 2081−2100 under the SSP245 climate scenario. In the SSP585 climate scenario, the highly suitable region of kam sweet rice increases in region and moves to higher altitudes in 2081−2100. (3) Among the 19 climate impact factors, the min. temperature of the coldest month and the mean diurnal range difference are the main climate factors affecting the distribution of kam sweet rice, with contribution rates of 34.8% and 13.8%, respectively. Generally speaking, climate change will lead to the expansion of the highly suitable region of kam sweet rice in 2041−2060 and the reduction of the region in 2081−2100, so it is necessary to pay attention to the climate change, and rationally plan the production layout of kam sweet rice to ensure the food security.

Key words: Kam sweet rice, Climate change, MaxEnt, Potential suitable region, Model optimization