Patterns of species richness and turnover rate along the elevation gradients.
Our results suggest that PTEs, with the interaction of topographic heterogeneity and climate change, have profoundly affected the biodiversity trends around metal mines along the elevation gradients of the Qilian Mountains. In the habitats classified as “natural” (i.e., not contaminated or lightly contaminated), the plant richness showed a unimodal distribution with increasing elevation (Fig. 1a, explained deviance = 38%, P < 0.005). Compared to the natural habitats, the presence of PTEs in soil caused an increase in plant richness (Fig. 1a, explained deviance = 58%, P < 0.005) and a decrease in plant evenness (Fig. 1b, explained deviance = 76%, P < 0.005). For the variations in plant richness, evenness, and cover, the models involving Climate × PTEs or Topography × PTEs described the data better than the models of climate, PTEs, or topography alone or the models involving additive effects (Supplementary materials Tables 2, 3, 4). Nevertheless, the deposition of PTEs in the soil altered the plant diversity distribution unevenly along the elevation gradient. Compared to the natural habitats, the plant species richness in PTE-polluted habitats located in the low-elevation desert (1578-2183 m a.s.l) and lowland grassland (2365-3079 m a.s.l) showed significant changes, with maximum increases of 25% and 18%, respectively. However, the evenness in the PTE-polluted habitats in the same elevation zone decreased by 44% and 45%, respectively (Fig. 1b and Supplementary Materials Fig. 7). Unlike the obvious changes in plant richness between the natural habitats and PTE-polluted habitats, the change in plant coverage between these two areas was relatively small (Fig. 1c, explained deviance = 95%, P < 0.005).
———————-
Place Fig 1 here
———————-
Alongside changes in species richness, the species composition of plant communities changed with the presence of PTEs and the interaction of topographic heterogeneity and climate change. Nonmetric multidimensional scaling (NMDS) demonstrated the spatial changes in plant communities shaped by PTE deposition in different study sites (n = 64) of the mountain range (average turnover rate = 0.51, range = 0.16 - 1.00) (Fig. 1d). The turnover rate reflected the high similarity in the plant composition at the mine sites on the same elevation gradient (PERMANOVA, R2 = 47%, P = 0.01). The results showed that PTE deposition changed the plant species composition of PTE-polluted habitats in most classified groups compared with that in the natural habitats (F = 9.06, df = 2696, P = 0.027, Fig. 1d). Throughout the four studied climate zones along the elevation gradient (mountain desert steppe, mountain grasslands, desert grassland, and mountain shrub-steppe), we found that the turnover rates in the PTE-polluted habitats showed greater changes in plant composition in the low-elevation mountain desert steppe (ANOVA, P = 0.052) and mountain grassland (ANOVA, P < 0.01) (Fig. 2).
———————-
Place Fig 2 here
———————-