Avoidance behaviour of bats and moths: when is it predator defence?

Abstract

Most animals live in environments where they risk becoming the prey of other animals. Because predation is such a catastrophic event for the prey, we could reasonably expect that most animals will exhibit behaviours that have an anti-predatory function. Typical examples include spatial and temporal avoidance of areas where, and periods when, predators are active. This kind of avoidance behaviour is illustrated by the vertical migration of zooplankton in lakes with fish (e.g. Gliwicz 1986), reduced feeding activity by rodents on moonlit nights (Clarke 1983), and avoidance of areas with high vegetation, that may harbour predators, by grazing antelopes (Underwood 1982). We have both worked on avoidance behaviours that we believe may serve an anti-predator function, and that presumably also have evolved in that context. In particular, we have argued that predation risk imposed by raptorial birds during the day is the principal reason why bat foraging activity is usually restricted to the night (Speakman 1990, 1991, 1995, Jones and Rydell 1994, Speakman et al. 2000) and also that nocturnality in bats first evolved for this reason (Rydell and Speakman 1995). Likewise, we have argued that the flight activity and behaviour of certain species of crepuscular moths are also shaped by predation risk, this time imposed by insectivorous birds during the day and bats during the night (Andersson et al. 1998, Rydell 1998, Svensson et al. 1999). Clearly, why bats are nocturnal and not diurnal is a central question for those of us who want to understand the biology of these animals, and the same can be said about many moths. However, when trying to publish our work, we have both repeatedly encountered the same problems when we argue that the temporal activity patterns in bats and moths indeed seem to be maintained by predation. Hence, this note serves to highlight some of the methodological difficulties that attend demonstrating that an avoidance behaviour has an anti-predatory function. It is to some extent borne out of frustration, because we think that the arguments most frequently delivered by reviewers have the consequence that it becomes impossible to test the hypothesis. The problem is accentuated in situations where most or all individuals of an animal behave according to a certain norm, from which they seldom or never deviate. Insectivorous bats and many insects are normally strictly nocturnal, and only fly in daylight under exceptional circumstances (Speakman 1991, Svensson et al. 1999). An analogous example is the ghost swift moth Hepialus humuli, among which the lekking males display intensively together for no more than 30 min each evening, triggered by the falling light intensity, and otherwise remain inactive (Rydell 1998). The methodological problems may not apply so much to the (much more studied) situation where the behaviour in question changes on a short-term basis or varies between individuals, so that the variation in behaviour can be related to variation in predation risk. Examples of this situation include the male Tungara frog, which stops calling and dives when a predatory bat comes within striking distance (Ryan 1985), and guppies, where some males are bolder than others and therefore are selected for mating by females (Godin and Dugatkin 1996). In the case of nocturnality in bats it is usually argued (from the reviewers' side) that predation is unlikely to explain the avoidance behaviour because there are few or no direct observations indicating that predation actually occurs (or at least not enough to indicate that it is important). The same applies to the flight activity of some moths, although for some species the problem is the reverse; predation pressure seems so high that it is hard to imagine that the moths perform any anti-preda-