Geochemical modelling and ecological risk of heavy metals in industrial agricultural soils: a Python-driven approach from southwestern Nigeria
Abstract
Heavy metal contamination of agricultural soils near the Ota Industrial Estate, southwestern Nigeria, was assessed using a six-module Python pipeline that integrates geochemical indexing, ecological-risk modelling, and human-health risk assessment. Cu, Pb, Zn, and Cd were analysed at 12 sites. Cu, Pb, and Zn were within WHO/FAO permissible limits. Cd concentrations exceeded the WHO/FAO guideline of 0.80 mg/kg at half of the locations (IN1--IN6, up to 3.09 mg/kg), and two locations exceeded the upper limit of 3.0 mg/kg. Contamination-factor analysis classified Pb at the vehicle-workshop zones (IN7--IN9) as considerable to very high, with CF up to 12.61, and classified Cu/Zn as moderate to very high at fabrication and galvanisation zones. The mean ecological risk index (RI) for the estate was 64.5 ± 24.1, indicating low overall risk, although Cd and Pb fell within moderate Er bands in some sub-zones. Independent recomputation showed that none of five originally proposed metal-pair correlations was significant; however, a strong Pb--Cd correlation was Bonferroni-significant (r = 0.946, p = 0.0012), indicating a shared anthropogenic source. US EPA RAGS non-carcinogenic hazard indices were below unity for adults (HI = 0.014) and children (HI = 0.126). Carcinogenic risk was not quantified because validated oral slope factors for Cd or Pb are unavailable in IRIS. The principal regulatory basis for clean-up is the Cd exceedance. The open-source pipeline provides a reproducible framework for soil-contamination assessment in West African regulatory contexts.