Strangers in strange lands: Evidence for generalized risk assessment strategies in non-native marine snails

Background/Question/Methods

Though it was originally supposed that prey in invaded predator-prey interactions would be unlikely to respond defensively to an evolutionarily novel predator, observations of this phenomenon are increasingly common. Several mechanisms for recognition have been proposed, including adaptation, learning, and generalized risk assessment. The prevalence of any one of these mechanisms has implications for invasion ecology, potentially enabling managers to identify when invasive prey might elude biotic resistance, or when the impact of invasive predators is likely to be mitigated by appropriate prey responses. Moreover, invasive species offer an opportunity to test theory on the role of co-evolution in selecting for information strategies in risk assessment. 

In mesocosm experiments, I investigated the relative importance of single and combined chemical cues indicating predation risk (fed or starved native crab, injured conspecifics or bland prey) in influencing risk assessment in 2 native and 3 non-native marine snails. I compared avoidance behavior of prey, defined as crawl-out, burying/hiding, and reduced feeding rates, among treatments, and to a control treatment lacking predation cues. Recognition strategies were defined as generalized if snails demonstrated the greatest magnitude of avoidance response when exposed to cues of injured conspecifics. 

Results/Conclusions

All non-native snails assayed demonstrated generalized risk assessment, based on: 1) avoidance behavior was greater in response to the general (injured conspecifics) vs. specific cue (crab); and 2) the presence of the full predation cue (conspecific snails consumed by crab) did not further increase avoidance behavior. Natives, on the other hand, avoided maximally only when presented with the full predation cue, thus requiring maximum information specificity to demonstrate the strongest avoidance behavior. The responses of natives either did not differ between the component cues (Nucella lamellosa), or were stronger to the specific predator cues (Littorina sitkana).

Despite evidence for generalized risk assessment, two of the three non-native species (Ocenebra inornata and Urosalpinx cinerea) also demonstrated the ability to recognize a predator with which they share only a limited evolutionary co-history, while there was no evidence for recognition in Batillaria attramentaria. Only one of the two species of native snails demonstrated a significant increase in avoidance in response to the crab predator alone (Nucella lamellosa), though predator cues are hypothesized to be more reliable than cues generated from injured conspecific prey.