Pierre-Marc Brousseau, Dominique Gravel and I. Tanya Handa
Ground beetles are remarkable hunters with the capacity to kill and consume a huge diversity of potential prey ranging from snails and poison-secreting millipedes to fast jumping springtails. However, in natural ecosystems, predators will not eat all prey they encounter. Predators are selective, in part based on their ability to catch, handle and eat potential prey. While the characteristics that can impede predator/ prey interactions may seem quite intuitive, developing mathematical models that allow us to predict which predator species will interact with which prey species is complex, due to the high diversity of morphologies, physiologies and offensive and defensive behaviours. Our objective was to develop a model to predict predator-prey interactions by identifying potential constraints, or trait matches, in predator-prey interactions. For example, the success of a predator might depend on having sufficient strength to deal with the cuticular toughness of a potential prey or its ability to detect fast moving prey. We conducted 475 feeding experiments using 20 species of ground beetles and 115 prey species. Each beetle was offered prey with different morphological characteristics to determine if it could kill and eat it. We observed that the main constraining characteristic determining a successful outcome for the predator was the biting force of the predator in relation to the cuticular toughness of the prey. Another important constraint was the predator/ prey body size ratio. However, these morphological characteristics were insufficient to correctly predict failures of the predators to consume prey. To do so, we also had to consider the phylogeny (evolutionary relatedness) of predator and prey species. This information served as a proxy for characteristics such as chemical defences of prey that are difficult to measure, but are expected to be similar in closely related species. Combining both constraining trait matches and phylogeny, we were able to correctly predict the outcome of over 82 % of ground beetle interactions with their prey. Our mathematical model can be easily extended to other predators and will be useful to better understand how food webs are structured, and to predict the impact of adding or removing species on interactions in ecosystems undergoing global change.
Image caption: Synuchus impunctatus. Image provided by authors.