Promoting the Development of Astragalus mongholicus Bunge Industry in Guyang County (China) Based on MaxEnt and Remote Sensing.

Abstract

To provide high-quality Astragalus mongholicus Bunge to domestic and foreign markets and maintain sustainable development of the A. mongholicus industry, Firstly, we evaluated the impact of environmental factors and planting areas on the A. mongholicus industry. The maximum entropy method (MaxEnt) was utilized to simulate the suitability distribution of A. mongholicus and establish the relationship between the active component contents of A. mongholicus and ecological factors through linear regression analysis. The random forest algorithm was subsequently used to perform feature selection and classification extraction on Sentinel-2 imagery covering the study area. Furthermore, the planting, processing, and sales of A. mongholicus in Guyang County were investigated, and the roles of stakeholders in the value chains were analyzed. The results demonstrated that precipitation of the warmest quarter, minimum temperature of the coldest month, standard deviation of seasonal temperature changes, range of mean annual temperature, and mean diurnal range [mean of monthly (max temp - min temp)] were the five environmental variables that contributed the most to the growth of A. mongholicus. The most influential factor on the distribution of high-quality A. mongholicus was the mean temperature of the coldest quarter. The classification results of image features showed that the planting areas of A. mongholicus was consistent with the suitable planting areas predicted by MaxEnt, which can provide data support to the relevant departments for the macro development of the A. mongholicus industry. In the production of A. mongholicus, 10 value chains were constructed, and the study demonstrated that the behavior of stakeholders, target markets, and the selected planting area had a significant impact on the quality of A. mongholicus.