Integrating Empirical Ecotoxicology and Predictive Modelling to Assess Ecological and Human Health Risks from Heavy Metals in Malaysia's Chini Lake UNESCO Biosphere Reserve.

Chini Lake, Malaysia's first UNESCO-designated biosphere reserve, exemplifies the growing challenge of safeguarding freshwater ecosystems exposed to legacy mining activities and ongoing land-use disturbance. Mining and other anthropogenic activities have contributed to persistent heavy metal contamination in freshwater ecosystems, underscoring the need for robust and reliable risk assessment frameworks to inform conservation and policy. This study applies an integrated approach combining empirical (ECOTOX) and predictive (ECOSAR) toxicity data to evaluate ecological and human health risks across 18 surface water sampling stations within Chini Lake tropical freshwater ecosystem. The study determines the abundance and distribution of heavy metals, heavy metal pollution index (HPI), risk quotient (RQ) and hazard quotient (HQ). The RQ derived from the empirical ECOTOX database revealed high ecological risks (RQ ≥ 1) from mean concentrations of aluminium, zinc, copper, nickel, and lead. This level of threat was underestimated by several orders of magnitude when using the ECOSAR model. This finding highlights a significant potential for the underestimation of ecological threats in freshwater bodies if predictive models are used in isolation for metals. For human health, all HQ values for direct water ingestion were below the risk threshold (HQ < 1), though arsenic and cadmium were the primary risk drivers, with children being 2-4 times more vulnerable than adults. This research provides a critical methodological framework for reconciling different risk assessment paradigms. It underscores the paramount importance of leveraging field-relevant, empirical ecotoxicological data over predictive models for metals to accurately characterise ecological threats and guide sustainable water governance in mining-impacted and climate-sensitive tropical environments.