Simulation of suitable habitats and geographic tracing based on medicinal and edible plants with Gastrodia elata Bl. as an example

Abstract

Gastrodia elata Bl. (G. elata) is a herb with important medicinal and edible values. Wild G. elata has been subjected to overharvesting and habitat fragmentation due to people’s pursuit of wild quality, resulting in resource depletion. Furthermore, yield cannot keep up with demand, and the contradiction between supply and demand has become increasingly prominent, giving rise to unscrupulous merchants selling counterfeit and shoddy herbs and confusing origins. Therefore, it is necessary to find methods to predict suitable habitats for G. elata to standardize cultivation and authenticate the origins. The research utilized the MaxEnt model and ATR-FTIR spectra to explore the distribution of G. elata’s suitable habitats and to authenticate G. elata from various origins, followed by the Mantel test to analyze the association between environmental variables and infrared spectra. Under the current climate scenario, G. elata’s suitable habitats were primarily found in provinces such as Yunnan, Sichuan, Guizhou, Tibet, and Shaanxi, among which Yunnan had the largest suitable habitat (21.75 %). The ResNet model based on simultaneous 3DOCS spectral data had the best authentication performance, with 100 % accuracy in both the train and test sets. The infrared spectra of G. elata samples from different origins significantly correlate with Prec09, demonstrating that this environmental variable may affect the accumulation of gastrodin components in G. elata. This research gives a scientific basis for the artificial cultivation of G. elata, avoiding resource waste (both human and material) and losses caused by blind introduction. It is conducive to exploring and protecting wild G. elata resources.

Keywords

• Gastrodia elata Bl.
• Suitable habitats
• Chemometrics
• Origin authentication
• Correlation