GIS-fuzzy neural network-based evaluation of tobacco ecologicalsuitability in southwest mountains of China

Evaluation of tobacco ecological suitability is important for the scientific tobacco planting and mountain agro-resources utilization in the southwest mountains of China. In this study, Qianjiang County tobacco plantation in southeast Chongqing was used as a case to evaluate tobacco ecological suitability. After analysis of tobacco biological characteristics, thirteen ecological factors with strong influence on local tobacco growth were selected to construct a tobacco planting ecological suitability evaluation indexes system. The spatial distributions of the factors of temperature, precipitation, sunshine and soil were modeled by gridded interpolation, multi-factor sunshine simulation and co-Kriging on GIS platform. This was followed by evaluation of Qianjiang tobacco planting suitability using the fuzzy neural network approach. The results showed that the average acreage of the most suitable regions was 648.63 km². This accounted for 27.03% of the total study area and mainly covered the moderate hillslope and middle mountain areas within altitude 800~1 100 m. The combined acreage of suitable and sub-suitable regions was 964.13 km², which covered 40.18% of the total area and mainly distributed in the hilly, low mountains within altitude 600~800 m. Unsuitable region with an acreage coverage of 775.16 km² mainly covered the mountain areas above 1 600 m of altitude or 25 degree of slope. The study not only provided guidance for adjustments of local tobacco planting but also offered a new evaluation mode of crop planting ecological suitability in mountain terrains.