­­Effect of multimodal cues from a predatory fish on refuge use and foraging on an amphidromous shrimp

BACKGROUND: Prey can alter their behavior when detecting predator cues. Little is known about which sensory channel, number of channels, or the interaction among channels that shrimp species use to evaluate the threat from predators. The amphidromous shrimp Xiphocaris elongata has an induced defense, an elongated rostrum, where predatory fishes are present. We sought to test if kairomones or visual cues when presented singly from fish either eating flakes or shrimp, had more effect on altering the temporal feeding and refuge use patterns of long-rostrum (LR) X. elongata. We were also interested in elucidating potential interactions among cues when presented simultaneously in different combinations (kairomones + visual + mechanosensory, kairomones + alarm + visual, kairomones + alarm, kairomones + visual) on the same response variables. We expected that when presented alone kairomones will significantly increase refuge use and decrease foraging, particularly late at night, in comparison to visual cues alone, and that multiple cues when presented simultaneously will further increase refuge use and decrease foraging at night. METHODS: We exposed shrimp to individual or multiple cues from the predatory fish mountain mullet, Augonostomus monticola. We examined shrimp behavior with respect to refuge use and foraging activity during four time periods (after sunset, nighttime, sunrise, and sunset) in a 24-hour period. RESULTS: Shrimp presented fish visual and chemical cues singly did not differ from one another but differed from control shrimp (no cues) with respect to refuge use or foraging. The number of shrimp using refuge in the treatment with most cues (KVM: kairomones+ visual + mechanosensory) was higher than in all the treatments with less cues. A significant decline in foraging was observed when multiple cues were presented simultaneously. The highest number of shrimp foraged one hour after sunset and at nighttime. A significant interaction was observed between cue treatments and time periods, with shrimp in the KVM treatment foraging less and using more refuge late at night and at sunrise than shrimp in other treatments or time periods. CONCLUSIONS: The observation that fish chemical and visual cues when presented singly produced similar refuge use and foraging patterns was contrary to expectation and suggests that visual and chemical cues, when presented alone, provide redundant information to X. elongata with regards to predation threat. The significant increase in refuge use and reduction in foraging observed in the KVM treatment suggest multimodal signal enhancement in the perception of threat. This makes evolutionary sense in “noisy” environments, such as streams, where detection, localization, and intention of predators is much improved when cues are received through multiple sensory channels.