Predation can affect prey indirectly by altering prey phenotype. Prey phenotype can change in response to cues related to predator-prey interactions [e.g. visual, mechanical or chemical cues from predators (i.e. kairomones), or chemical cues excreted by injured conspecifics (i.e. alarm cues)]. In Caribbean streams, the rostrum of the amphidromous shrimp Xiphocaris elongata is significantly longer below waterfall, where predatory fishes are present, than above waterfalls, where fishes are absent. Previous experiments showed that rostrum growth in X. elongata was higher when long-rostrum and short-rostrum shrimp were exposed to either kairomones and alarm cues or kairomones cues alone, than when exposed to no cues. These experiments demonstrated that predator and prey cues in Caribbean streams affect shrimp morphology. Our objective here is to examine if predator and prey cues in Caribbean streams also affect shrimp behavior. We performed laboratory experiments in which we exposed long-rostrum and short-rostrum X. elongata for 24 hours to either visual, mechanical, and chemical cues from predatory fish (kairomones); visual, kairomones and alarm cues; visual and kairomones cues; kairomones and alarm cues; kairomones cues alone; or no cues, to examine the proportion of shrimp that were hiding in crevices and the proportion of shrimp that were inactive.
Results/Conclusions
Shrimp of both phenotypes hid at a higher proportion when receiving kairomone, visual, and mechanical cues from fish and alarm cues from shrimp (average = 0.65), than when receiving a combination of kairomone, visual, and alarm cues (average = 0.35); kairomone and alarm cues (average = 0.33); kairomone and visual cues (average = 0.35); kairomone cues alone (average = 0.41) or when receiving no cues (average = 0.34) (P <0.0001). Shrimp of both phenotypes were stiller (i.e. less active) when receiving either a combination of visual, chemical, and mechanical cues from fish (average = 0.81); kairomone, visual, and alarm cues (average= 0.79); or kairomone and alarm cues (average = 0.69); than when receiving kairomone and visual cues (average = 0.62); kairomone cues alone (average = 0.61) or no cues (average = 0.45) (P <0.0001). Our work demonstrates that fish predation in Caribbean streams affects X. elongata shrimp indirectly through morphological and behavioral inducible defenses. Knowledge of how predation-related cues generate a response in their prey will contribute to our general understanding of the ecological and evolutionary mechanisms responsible for predator-prey communication and specific understanding of their consequences in aquatic habitats.