Capture success and efficiency of dragonflies pursuing different types of prey

The dynamics of predator-prey interactions vary enormously, due both to the heterogeneity of natural environments and to wide variability in the sensorimotor systems of predator and prey. In addition, most predators pursue a range of different types of prey, and most organisms are preyed upon by a variety of predators. We do not yet know whether predators employ a general kinematic and behavioral strategy, or whether they tailor their pursuits to each type of prey; nor do we know how widely prey differ in their survival strategies and sensorimotor capabilities. To gain insight into these questions, we compared aerial predation in 4 species of libelluid dragonflies pursuing 4 types of dipteran prey, spanning a range of sizes. We quantified the proportion of predation attempts that were successful (capture success), as well as the total time spent and the distance flown in pursuit of prey (capture efficiency). Our results show that dragonfly prey-capture success and efficiency both decrease with increasing size of prey, and that average prey velocity generally increases with size. However, it is not clear that the greater distances and times required for capturing larger prey are due solely to the flight performance (e.g., speed or evasiveness) of the prey, as predicted. Dragonflies initiated pursuits of large prey when they were located farther away, on average, as compared to small prey, and the total distance flown in pursuit was correlated with initial distance to the prey. The greater initial distances observed during pursuits of larger prey may arise from constraints on dragonflies' visual perception; dragonflies typically pursued prey subtending a visual angle of 1°, and rarely pursued prey at visual angles greater than 3°. Thus, dragonflies may be unable to perceive large prey flying very close to their perch (subtending a visual angle greater than 3-4°) as a distinct target. In comparing the performance of different dragonfly species that co-occur in the same habitat, we found significant differences that are not explained by body size, suggesting that some dragonflies may be specialized for pursuing particular types of prey. Our results underscore the importance of performing comparative studies of predator-prey interactions with freely behaving subjects in natural settings, to provide insight into how the behavior of both participants influences the dynamics of the interaction. In addition, it is clear that gaining a full understanding of predator-prey interactions requires detailed knowledge not only of locomotory mechanics and behavior, but also of the sensory capabilities and constraints of both predator and prey.