Plants serve as vital components of ecosystems, with their contamination status acting as sensitive indicators of environmental pollution. Therefore, the precise assessment of plant heavy metal contamination and source identification are crucial for regional ecological conservation and sustainable development. This study investigated heavy metal pollution in four characteristic plant species (Anabasis aphylla L., Alhagi camelorum Fisch., Reaumuria songonica (PalL)Maxim., and Haloxylon ammodendron (C. A. Mey.) Bunge.) within the Kalamaili Mountain Nature Reserve, employing comprehensive methodologies including pollution indices, bioconcentration factors (BCFs), absolute principal component score–multiple linear regression (APCS-MLR), and the random forest model (RF). The key findings revealed the following: The soil exhibited severe Cd and Hg contamination. The plant Cr concentrations exceeded standard limits by 31.89 to 147 fold. The Pb, Hg, and As content in plants showed significant differences. The plants displayed differential metal enrichment capacities, ranked as Cr (BCF = 3.28) > Hg (1.22) > Cd (0.92) > Cu (0.25) > Zn (0.15) > Pb (0.125) > As (0.125), highlighting Cr, Hg, and Cd as priority ecological hazards. Complex interactions were observed, with Reaumuria songonica (PalL)Maxim. showing strong Cd soil–plant correlation (r = 0.78), whereas Alhagi camelorum Fisch. demonstrated negative associations (Cd: r = −0.21). APCS-MLR identified mining/smelting as primary contributors to Cd (63.49%), Zn (55.66%), and Cr (45.51%), while transportation dominated Pb emissions (72.92%). Mercury pollution originated from mixed sources (56.18%), likely involving atmospheric deposition, and RF modeling corroborated these patterns, confirming industrial and transportation synergies for Cd, Zn, Cr, Cu, Hg, and As, with Pb predominantly linked to vehicular emissions. This multidisciplinary approach provides scientific evidence for establishing heavy metal monitoring systems and formulating targeted remediation strategies in arid ecologically fragile regions.

To investigate the contamination status and analyze the sources of soil toxic metal contamination on the northern slopes of the East Tianshan mountain industrial belt in Xinjiang, northwest China, this study measured the contents of six common toxic metals such as Zn, Cu, Cr, Pb, Hg and As in 82 surface soil (0–20 cm), and using the ground accumulation index, pollution load index, and improved weighted index assessed the contamination characteristics of the soil and using a semi-variance function and APCS-MLR model the identified the potential sources of contamination. The results indicate that the average concentrations of Pb, Hg, and As are significantly higher than the background values in Xinjiang. The average ranking of toxic metal content is as follows: Zn > Cr > Pb > Cu > Hg > as. A single-factor pollution index analysis shows that As and Pb pollution are severe, while Hg and Cu pollution are moderate. The improved weighted index shows that moderate lead pollution accounts for 6.1% and severe lead pollution accounts for 54.88%; 98.88% of arsenic is severely contaminated. The APCS-MLR model identified three main sources of heavy metals: Cu and Cr as industrial production sources, Pb and Zn as transportation and agricultural activity sources, and As, Cr, Cu, Hg, and Zn as natural and mixed pollution sources. This study provides a solid scientific basis for the prevention and control of toxic metal pollution in agricultural soils, thus ensuring food security and sustainable development in the region.