2.4 Division of suitable areas and climatic characteristics
When converting a continuous prediction result into the Boolean form of
”suitable habitat” and ”unsuitable habitat”, it is important to select
the appropriate threshold (Chen et al., 2022). We used the maximum
training sensitivity plus specificity threshold, which maximizes the TSS
value to create binary maps (de Andrade et al., 2020; Jiménez-Valverde
& Lobo, 2007). The threshold value (P) of the suitable areas ofT. chinense was 0.275. In addition, the suitable areas ofT. chinense also included three levels of suitable areas, which
were low suitable areas (0.275 ≤ P < 0.5), medium suitable
areas (0.5 ≤ P < 0.7), and high suitable areas (P ≥ 0.7). We
used SDMtools to compare the spatial changes of the suitable areas ofT. chinense under different climate scenarios.
To analyze the change of environmental characteristics in the suitable
areas of T. chinense , we randomly selected 5000 points in the
current suitable areas of T. chinense . We extracted the
corresponding values from the 5000 points in the layers of different
climatic conditions corresponding to the dominant environmental factors.
R 4.2.0 software was used to calculate the 95% quantile and average
value of the extracted values, and what kind of environmental pressure
will be faced by T. chinense in the future was intuitively
verified (Tagliari et al., 2021).