Behavior-Based Thresholds Support Screening-Level Prioritization of Caffeine as a Neuroactive Wastewater Micropollutant in Coastal Waters

Abstract Sensitive functional end points reveal early responses to neuroactive wastewater contaminants but are rarely translated into screening thresholds. Using caffeine as a sentinel, we evaluated a behavior-based framework for coastal waters. Systematic searches of Web of Science Core Collection, PubMed, and Embase through 24 August 2026 identified 66 eligible studies and 2181 effect sizes for meta-analysis. The near-zero grand pooled signed ln RR may reflect cancellation among end points with opposing response directions. Responses varied by life stage, exposure duration, concentration, and end point category. Standardized light–dark assays across eight aquatic species revealed taxon- and context-dependent behavioral disruptions, including responses at low concentration and non-monotonic patterns. Species sensitivity distributions based on behavioral benchmark concentrations yielded a conservative hazardous concentration affecting 5% of species (HC5; 0.0585 μg L–1) and a screening-level behavior-based predicted no-effect concentration (PNEC; 0.0195 μg L–1). A separate systematic literature search, supplemented with NORMAN EMPODAT records, compiled 1421 quantifiable caffeine measurements in coastal and estuarine waters. Concentrations were right-skewed (median, 44.9 ng L–1; IQR, 14.60–190.00 ng L–1), and 69.88% of measurements fell into the high-priority risk quotient category (RQ ≥ 1). These findings support behavior-based thresholds for prioritizing neuroactive wastewater pollution in coastal waters.