Predation is a key biological process that influences interactions among individuals, shapes species’ distributions and alters the structure of ecological communities 1,2. For example, predators can have dramatic effects on the ecosystems in which they reside, with their reintroduction or removal triggering trophic cascades 3–5. In cases where predation has an influence on prey fitness, it can drive the evolution of prey behaviours that minimise detection by predators, such as foraging in dense habitat or at times when visibility is low 6.
Optimal foraging theory predicts that prey species trade-off the need to acquire food with the risk of predation associated with foraging 7,8. In order to behave in this way, prey species must evaluate predation risk, which they may do both directly and indirectly 9. For example, prey can detect predators directly through visual, olfactory or echolocation cues 10. However, in the absence of direct cues, many prey species respond to indirect cues of predation risk, including abiotic factors such as moonlight and habitat structure 9,11.
Approximately sixty-nine percent of mammals are nocturnal 12, thus changes in moonlight over the lunar cycle is expected to influence the behaviour of many species. Increased moonlight can improve the hunting efficiency of both mammalian and non-mammalian predators 13,14, resulting in increased predation risk and reduced activity for prey species 15–18. For example, the detectability of new holland mice (Pseudomys navaehollandiae) decreased with increasing moonlight and six times more survey effort was required to reach 95% confidence that the species was absent during full moon compared to new moon 15. These observed behavioural patterns form the basis of the predation risk hypothesis, which predicts that increased moonlight will result in lowered activity in prey species and increased activity in predators 19–21.
Dense vegetation provides mammals with a variety of resources including shelter from predators 14,22,23. For example, Allenby’s gerbils (Gerbillus andersoni allenbyi) exposed to owls and vipers harvested substantially more seed (a measure of perceived predation risk) from sheltered compared to open patches 22. Potentially, reduced predation risk in sheltered environments can interact with increased predation risk associated with illumination, resulting in the suppressive influence of moonlight on prey activity diminishing with increasing shelter. This interaction is described by the habitat mediated predation risk hypothesis21, although tests of this hypothesis have rendered mixed results. For example, Meriam’s kangaroo rat (Dipodomys merriami) removed fewer seeds from open habitat compared to sheltered habitat during a full moon 24. In contrast moonlight had no effect on the use of dense or open microhabitats by the agile antechinus (Antechinus agilis) 25.
In addition to improving the hunting effectiveness of predators, increased moonlight may also increase the visual acuity of some prey species, resulting in greater foraging efficiency, a heightened ability to detect predators, or both 26–28. If the benefits of moonlight to prey outweigh the potential costs, the visual acuity hypothesis predicts a positive association between activity and moonlight 21. For example, capture rates of the giant kangaroo rat (Dipodomys ingens) were positively associated with moonlight, a result explained by their use of open habitat where early detection of predators may reduce predation risk 28. The visual acuity hypothesis is more likely to predict the response of prey to moonlight for species who rely primarily on vision as their primary means of locating food and predators 21.
In this study, we investigated the response of seven small–medium sized mammals ranging from 20–2500 grams (prey species) and two introduced predators (feral cats, Felis catus and red foxes, Vulpes vulpes) to moonlight and vegetation cover in the woodlands of south-western Victoria, Australia. Wildfire is a natural process in this ecosystem, and prescribed fire is used as a management tool to reduce fuel loads and manage biodiversity. Infrequent wildfire and annual application of prescribed fire in our study area has resulted in a large range of vegetation states from open recently burnt sites to dense regenerating areas, making it an ideal location to investigate the responses of prey and predator species to moonlight and vegetation cover. Furthermore, there is increasing evidence that fire-induced vegetation change can increase predation on small mammals by introduced predators 1,29,30 thus determining the role of vegetation cover in mediating perceived predation risk is important. For example, increased predation in recently burnt areas with reduced vegetation cover could potentially be moderated by implementing patchy burns resulting in unburnt vegetation within the fire perimeter for prey species to shelter in.
We tested the following alternative hypotheses (Fig. 1):
-
Predation risk hypothesis. Moonlight increases predation risk. We predict a negative relationship between moonlight and prey species activity and a positive relationship between moonlight and predator activity.
-
Habitat mediated predation risk hypothesis. Moonlight increases predation risk to a greater extent in open compared to sheltered habitats. Both the negative relationship between moonlight and prey activity and the positive relationship between moonlight and predator activity will be stronger in open compared to sheltered habitats.
-
Visual acuity hypothesis. Moonlight reduces predation risk. There will be a positive relationship between moonlight and prey species activity and a negative relationship between moonlight and predator activity.