Introduction
In recent decades, human activities have caused a significant loss of global biodiversity due to their impact on climate and vegetation1,2. In some “hotspots”, vegetation diversity variations have led to major changes in ecosystem function, stability, and serviceability3,4. The establishment of global conservation priorities and an understanding of the alteration processes in different environments have become necessary to focus the limited economic resources on those areas with the highest protection value and that are most threatened by environmental changes5,6. As demonstrated by several studies, long-term metal mining and processing have disturbed mountain ecosystems and altered the succession of plant communities in many regions. For example, in the mountains of Eastern Europe (Alps)7, North America (Appalachian Mountains)8, and South Africa9, as well as in other mineral-rich mountain ranges, extensive mining has triggered the spread of potentially toxic elements (PTEs, such as Cd, Zn, Cu, Pb, Ni, Hg, and Cr), which profoundly alter the local environment for the growth of plant species10,11,12. However, some researchers argued that the effect of PTEs on vegetation succession was shown to affect plant species richness within limited small areas12,13. Other studies on habitats and abiotic factors have demonstrated that regional level variations (e.g., climate patterns and topographic heterogeneity) may lead to greater changes in community succession processes14-16. For example, the effects of PTE transport can overlap with rapidly changing climatic conditions along elevation gradients, causing variations in plant diversity and ecosystem function17,18and, eventually, concealing the effects of PTEs on succession19,20,21. Understanding the potential additive and interactive effects of climate, topography, and PTEs on biodiversity and ecosystem functions (ESFs) at multiple spatial scales is therefore crucial to comprehending the ecological consequences of abandoned mines worldwide10,20,22.
Although some authors have been highlighted the influence of PTEs transport in small-scale experiments or in study regions with narrow climatic gradients8,9,23, most mining ecosystems still lack long-term biophysical and ecological data sets for assessments of possible ecosystem responses. Ideally, the prediction of climate-related ecological changes requires the simultaneous consideration of multiple climate drivers and the assessment of the mid- to long-term temporal changes to capture the complex, interacting responses24. “Space-for-time” substitutions, also known as ergodic gradient studies, allows for quantitative predictions to be realized in the absence of long-term data sets or a detailed mechanistic knowledge of possible responses to climatic changes25,26. This approach uses multiple sites across environmental gradients (i.e., the elevation gradient) to predict the time trajectory in ecological changes, which is considered to be causally correlated with the changes across the gradient24. Studies employing this approach have been applied in paleoenvironmental investigations of pollen27, altitude changes in bird distribution28 and biodiversity changes in land use29.
The objective of this study was to quantify the combined impact of climate, mountain topography, and PTEs on plant biodiversity and ecosystem functions through a broad survey of the multiregional mining sites of the Qilian Mountain metallogenic belt in the northeastern Tibetan Plateau (China) (Supplementary materials Fig. 1, Table 1). Within an elevation gradient rising from 700 to 5,808 m above sea level (a.s.l.) and an extension of approximately 800 km, the Qilian Mountains are a metal-rich mountain range that covers several major natural ecosystems of temperate regions, from low-elevation desert to mountain, mountain shrub-steppe, and alpine grassland, including high-elevation desert ecosystems. Based on the “space-for-time” method, a multimodel framework was proposed to assess the statistical distribution of ecological responses arising from exposure to one or more metal stressors along different elevation gradients (1500-3600 m a.s.l) and determine which elevation zone (or climate condition) has habitats that are more vulnerable to PTE threats. The importance of a series of environmental covariates correlated with ecological risks, PTE transport, and species richness patterns was quantified. For instance, elevation and ecological risk caused by PTE transport were found to be significantly correlated in all the elevation gradients (r = 0.68). Climate variables such as the mean annual precipitation (MAP, r = -0.92), mean annual temperature (MAT, r = 0.79), transpiration rate (ET0, r = 0.45), and wind speed (win10, r = -0.83) were found to be correlated with elevation. Among the topographic factors, both slope (r = 0.41) and aspect (r = 0.40) were correlated with the distribution of the ecological risk related to mines. Afterward, we selected four variables to characterize the intensity of PTEs (ecological risk index, mining years, closure period, and the relative proportion of mining area/plant/factory in the surrounding landscape) and calculated their normalized indicators (PTE intensity). Finally, we compared the statistical support of generalized additive models (GAMs), including the individual and interactive effects of PTEs, climate, and topography, on biodiversity and ecosystem functions.