3.6 Redundancy analysis
We used DCA and RDA to analyze the relationship between species characteristics and environmental factors. The plant data were sorted by DCA, and the gradient lengths of the four sorting axes were obtained, of which the maximum value was 0.63 < 3; therefore, we selected the RDA method. The eigenvalues of the sorting axis and the cumulative interpretation showed that daytime temperature (DT), nighttime temperature (NT), and soil moisture condition (SM) explained 35.70% of the plant change variation. The first two axes explained 35.70% of the plant change variation and 100% of the species-environment relationship variation. The results could explain all the variation in the species-environment relationship (Table 5).
Within a certain range, the SM had the greatest influence on the characteristic indexes of species, followed by DT, and NT, which exerted the least influence on the change of the characteristic plant indexes (Figure 4). All vegetation characteristic indexes were positively correlated with SM, which had the greatest impact on Pn and WUE (P <0.01) and the least impact on BGB, followed by Tr, Gs, and LWC, which had a significantly positive correlation with SM (P <0.05). Daytime temperature exerted the greatest positive influence on LWC and H, as it had an extremely significant positive correlation with both (P <0.01), and the least influence on Pn. Daytime temperature exerted the greatest negative influence on AGB and Gs, as it had a very significant negative correlation with these parameters (P <0.01), and the least influence on Ci (Table 6).