A clustering approach based on high-resolution ecological vulnerability index reveals spatial patterns of per- and polyfluoroalkyl substances pollution in lakes on the Tibetan Plateau

Per- and polyfluoroalkyl substances (PFAS) are persistent organic pollutants (POPs) with toxicity, chemical stability, and long-range transport potential. The transport and accumulation mechanisms of PFAS in specific or typical lakes have been reported. In the wake of global PFAS pollution, it is more important to unravel the distribution patterns of PFAS across larger-scale, multi-lake systems. However, traditional lake classification methods are often overly simplistic and inflexible to adapt to large lake systems with complex ecological characteristics. Here, an improved ecological vulnerability index (EVI) was introduced and applied for the first time to classify lakes in a regional, multi-lake study of PFAS pollution. We evaluated the effectiveness of EVI that integrated multi-dimensional environmental factors in revealing PFAS distribution in 12 lakes on the Tibetan Plateau. The results showed that the composition, concentration, and diversity of PFAS in water and sediment samples significantly differed between high-vulnerability lakes (HVL) and low-vulnerability lakes (LVL) clustered by EVI. The linear regression of PFAS concentration and diversity on EVI was most pronounced at the 1-km buffer zone scale compared to larger scales. EVI was strongly associated with PFAS concentration and diversity in HVL dominated by natural factors, and these associations were weakened in LVL with prevalent human interference. Our findings indicate the greater potential of EVI to predict the spatial patterns of PFAS in lakes at smaller scales and across regions with comparable dominance of natural factors. The proposed clustering approach is adaptable, as the indicators and weights in the EVI system can be adjusted based on regional ecological characteristics. This study provides a tool for unveiling the distribution patterns of PFAS and their driving mechanisms in complex lake environments.