Nanoplastics are increasingly recognized as biologically active contaminants in aquatic ecosystems, yet whether their toxicity follows common physiological
principles across habitats remains unclear. Here, we combine hierarchical three-level meta-analysis with conservative machine-learning analyses to test
whether habitat acts as a higher-order organizer of nanoplastic toxicity in microalgae. We synthesize 1,287 effect sizes from 40 independent studies
spanning freshwater and marine environments. Oxidative stress and growth inhibition emerge as universal responses to nanoplastic exposure, whereas
downstream physiological responses differ consistently between habitats. Freshwater microalgae exhibit greater disruption of photosynthesis, energy
metabolism and growth, while marine taxa show stronger investment in oxidative stress regulation, antioxidant defence and nutrient allocation,
revealing distinct habitat-dependent physiological response architectures. These architectures are further modified by exposure concentration, exposure
duration, particle size and taxonomic identity, with exposure concentration representing the strongest overall predictor of toxicity. Grouped cross-validation
demonstrates that habitat-dependent response architectures are reproducible across independent studies. Our findings show that habitat does not simply
alter the magnitude of nanoplastic toxicity but systematically reorganizes how physiological stress is expressed across aquatic ecosystems with contrasting
environmental conditions. This architecture-based perspective provides a mechanistic framework for habitat-informed ecological risk assessment and
predictive modelling of nanoplastic pollution.

暂无