Metabolomics has become a popular and accessible tool for scientists seeking to understand phenotypic changes at the metabolic level, providing a snapshot of cellular activity distinct from what can be inferred by proteomics or transcriptome analysis. But can this approach be used to characterize chemical compounds acting in natural environments, affecting organismal behavior and physiology? We have developed a metabolomics-based strategy to leverage the existing variation in chemical cues across different environmental conditions towards identifying ecologically important waterborne molecules. High concentration patches of signaling molecules are sensed by local animals, alerting them to the risk of attack from predators, resulting in modified behavior with cascading effects on coastal ecosystems. As expected, among the many pressures that organisms face, predation has contributed to the evolution of sensory systems for detecting chemical cues found in natural systems.