Predator avoidance behavior of nocturnal and diurnal rodents

Highlights

• Diurnal rodents responded differently to predator presence.

• Camera traps were used to survey feeders.

• Trade-off between feeding and predator avoidance dependent on visitation order.

Abstract

Animals trade-off predation risk against feeding opportunities and prey species may use signals or cues of predators to assess predation risk. We analyzed the mesopredators pine and stone marten (Martes martes, M. foina) and nocturnal and diurnal rodents (Glis glis, Apodemus spp., Sciurus vulgaris). The non-experimental approach used camera traps at feeders which were visited by both, predator and prey. As prey species can eavesdrop on predator signals/cues, there should show some avoidance behavior. The study was conducted on a small mountain in Germany, largely covered by wood, between 29.6.2018 and 5.10.2018. Camera traps were placed 0.6 m near a feeder. Food was replenished regularly to provide a continuous food supply. 34 camera traps provided data for an analysis; total trap nights were 513 (12,312 h). Martens detected the food sources first in 10 instances, and prey species Apodemus/G. glis in 24 instances. G. glis seemed to generally avoid places where martens were feeding while Apodemus and Sciurus did not. The visitations of G. glis depended on whether martens were the first visitors and it significantly avoided such places. Similarly, Apodemus appeared less often at a feeder when martens have been present as a first visitor. The time interval to resume feeding to a monitored feeder after a marten visit was significantly longer compared to a control in G. glis, but not in Apodemus and S. vulgaris. The study shows different responses, with the weakest in the diurnal rodent, and the highest in G. glis. Thus, if a food resource was known by prey species before a predator occurred, the trade-off was shifted towards feeding, but when the predators detect the food source first, the trade-off was shifted to predator avoidance.

Introduction

Effect of predators on prey species can be direct and lethal impacting on an individual’s survival (Caro, 2005) as well as indirect and non-lethal (Cresswell, 2008), e.g., by imposing stress and altering physiological condition or influencing reproduction (Caro, 2005, Cresswell, 2008). Also, predation may influence other activities, such as feeding or maintenance behaviour (Randler, 2006a, b). Therefore, avoidance behavior may have an important influence on populations and community ecology (Cresswell, 2008). Animals take risk of predation into account when trading-off this risk against opportunities for feeding (Kats and Dill, 1998). The threat-sensitive predator avoidance hypothesis (Helfman, 1989) suggests that prey will trade-off predator avoidance against other activities, such as feeding, and that the avoidance responses are related to the magnitude of the predatory threat (Helfman, 1989). On the other side, nutrient-rich food may shift the trade-off to more risky behavior. To assess this situation, animals need information about predation risk (Kats and Dill, 1998) but also about the quality and amount of food. In prey species, specific adaptations allow predator recognition and avoidance behaviour (Apfelbach et al., 2005). Different direct cues or signals can be used for predator recognition, such as visual, acoustic, or olfactory ones; but also, indirect cues can be used (Orrock et al., 2004). For example, tammar wallabies Macropus eugenii were able to use visual, but not acoustic cues from predators, to assess the predation risk (Blumstein et al., 2000). On the other hand, acoustic predator discrimination seems to be frequent across many taxa (Barrera et al., 2011; Hettena et al., 2014). Olfactory eavesdropping has also been reported in many mammals, fish, and in some birds (Apfelbach et al., 2005). In birds, for example, great tits (Parus major) avoid places where a predator has operated previously (experimentally induced by presenting fur and mangled feathers within nest boxes; Ekner and Tryjanowski, 2008): However, olfaction is often linked with visual signals (e.g., feces).

In this study, we analyzed the mesopredators pine and stone marten (Martes martes, M. foina) and nocturnal and diurnal rodents (Glis glis, Apodemus spec., Sciurus vulgaris) as our focal system. Many small mammals are reported to respond to olfactory predator cues, such as the whole animal, feces, urine, anal gland secretions or fur, by altering their space use, feeding patterns, habitat shifts or changes in circadian activity (see Apfelbach et al., 2005 for details). In a comparable mammalian system (least weasel Mustela nivalis nivalis– field vole Microtus agrestis interaction), voles avoided the scents of a predator (Hughes et al., 2010). Also, house mice Mus musculus receiving predator scents in an experiment reduced their relative levels of visitation to feeding places (Hughes et al., 2009). Sánchez-González et al. (2018) placed live traps for wood mice Apodemus sylvaticus in different experimental plots with differing concentrations of red fox scents V. vulpes to confirm this hypothesis (see also Navarro-Castilla and Barja, 2014). In Sciurus, and dormice, predator recognition (or eavesdropping on predator calls) has been also well documented (Hendrie et al., 1998; Randler, 2006a, b). Concerning visual cues, wild caught voles removed fewer seeds from a tray when a stoat Mustela erminea was presented in a cage about 0.15 m and 1.5 m away, but no difference to a control group was found at 3.5 m (Koivisto and Pusenius, 2006).

Here, we used a slightly different and non-experimental approach by using field data from camera trap surveys to investigate whether prey species fat dormouse G. glis, yellow-necked/wood mouse Apodemus spec. and Eurasian red squirrel Scirurus vulgaris respond to predator presence of a real predator and whether they avoid places where a predator feeds in a natural environment or whether they change their behaviour.

Camera traps can be a powerful tool for many ecological questions, like occupancy, species distribution and many other aspects (Rovero and Zimmermann, 2016), but they have been rarely used in behavioural biology, e.g. to measure behavioral responses. Saxon-Mills et al. (2018) and Steindler et al. (2018) used similar camera traps in their behavioral analysis of two different burrowing, nocturnal and largely solitary marsupials. In such cases, the presence of an observer may disturb and influence the behavior of the animals. Camera traps, instead of lab experiments, work on natural conditions where the behaviour of animals is not altered by other confounding factors (Campos et al., 2017). Therefore, natural observations provide fruitful information in addition to laboratory experiments. Moreover, the combination of both can be beneficial to foster our understanding of ecological processes (see, example of seed dispersal in rodents: Campos et al., 2017). Camera traps can capture small mammals of the size of a mouse at distances of about 0.6 m with a high probability (own lab observations; data not published). Further, Randler and Kalb (2018) reported that these cameras are also able to capture even bird species of similar size to a high proportion, despite birds were significantly lighter and smaller than mammals.

As prey species are able to recognize or eavesdrop on predator signals or cues, we hypothesize, that prey should show some avoidance behavior, because martens may leave chemical signals or chemical cues (via their paws) or probably make some noise or sound during their feeder visitations.