Footprints of predatory lady beetles stimulate increased dispersal of aphid prey, but do not alter feeding behavior or spread of a non-persistently transmitted plant virus

Abstract

Predators impact prey directly, through consumption, and indirectly, through non-consumptive effects that modify prey physiology and behavior in ways that affect survival and reproduction. Non-consumptive effects (NCEs) are well documented across prey taxa in response to different predator cues. However, for most prey species, both the mechanisms underlying NCEs and the impacts of NCEs on broader ecological processes are poorly understood. We addressed these knowledge gaps for an aphid prey species (Myzus persicae) by studying dispersal, in-leaf feeding behavior, and plant virus transmission in response to chemical cues deposited by a walking predator (Hippodamia convergens), a non-predator, non-competitor (Drosophila melanogaster), and an artificial feeding deterrent (2% mineral oil). We used this approach to better understand the specificity of M. persicae responses to chemical footprints of threatening and non-threatening organisms, as well as the magnitude of behavioral responses relative to a known deterrent. We found that chemical footprints deposited by H. convergens stimulated M. persicae to disperse from a suitable host (Brassica napus) at a rate equivalent to the 2% mineral oil positive control, while footprints of D. melanogaster did not modify M. persicae dispersal. Through electrical penetration graphing (EPG) recordings, we found that mineral oil stimulated aphids to make more probes, but footprint treatments did not significantly modify feeding behavior. In mesocosm-based virus transmission assays, H. convergens footprints and 2% mineral oil also stimulated increased dispersal, but this did not translate into increased transmission of turnip mosaic virus. We identified components of H. convergens and D. melanogaster footprints using gas chromatography and mass spectrometry and found that H. convergens footprints are chemically similar to footprints of related Coccinellidae, but distinct from D. melanogaster footprints. Our results provide evidence of specificity in M. persicae responses to predator footprints, which may help this prey species avoid costly dispersal behavior in response to cues from non-predators. We also provide ecological context by demonstrating that increased dispersal in response to predator cues does not always lead to increased virus transmission, as previously assumed, particularly when in-leaf feeding behaviors underlying virus acquisition do not change in response to predator cues.