Contents lists available at ScienceDirect Behavioural Processes journal homepage: www.elsevier.com/locate/behavproc Ninja owl; Gerbils over-anticipate an unexpected flying predator Sonny Shlomo Bleicher a,b,c, *, Burt Philip Kotler c , Keren Embar c a Department of Biological Sciences, Washington and Lee University, Lexington, VA, USA b Department of Environmental Science and Policy, George Mason University, Fairfax, VA, USA c Mitrani Department for Dryland Ecology, Ben-Gurion University of the Negev, Sde Boker, Israel ARTICLE INFO Keywords: Apprehension Behavioural ecology Evolutionary game theory Harvest-rate curves Predator-prey dynamics Risk-estimation ABSTRACT Foragers make decisions based on cues, information collected from their environment, processed into strategic behaviours. This information, processed in multiple regions of the brain, ultimately result in the production of stress hormones and visible changes in behaviour of animals – both reflexively to avoid depredation and stra- tegically to avoid an encounter with the predator. In a common-garden experiment we tested how imperfect information from visual cues of a predator impacts foraging and apprehension of a desert rodent, the Egyptian gerbil (Gerbillus pyramidum). The gerbils were exposed to predation by barn owls (Tyto alba), one camouflaged on dark nights using black dye. Gerbils’ response to the owls was measured using patch-use measured in giving- up densities (GUDs) and time spent in vigilance activity. Owl lethality was extrapolated from mean times spent in attacks and number of attempted strikes. Dyed owls attack-rate was lower and attack duration greater than those of the white owls. During the full moon, when dyed owls were visible, gerbils responded with extreme vigilance and minimal foraging (high GUDs). During the new moon when the owls were most stealthy, the gerbils showed low vigilance coupled with a similar high GUD. The inconsistency between gerbils’ foraging and vigilance behaviours, suggest a likely mismatch between perceived risk and actual measurement of predator lethality gathered by the gerbils’ observations in real time. 1. Introduction As the survival of the fittest shapes the living communities on earth (Darwin, 1859), no pressure is stronger than that of predation risk. Predators directly impact the populations of their prey in the environ- ment through consumption, but additionally indirectly through non- consumptive forces, i.e. “fear” (Orrock et al., 2008). Fear has been defined as an increase in stress-hormone production by an individual due to its interaction with environmental cues that suggest increased impending mortality (Gross and Canteras, 2012). However, ecologically it is perhaps more usefully defined as the behavioural, or strategic, response to predation costs of foraging arising from tradeoffs of food and safety (Brown et al., 2001; Bleicher, 2017a). Predation risk impacts many of the basic ecological functions of populations and individuals. For example, the evidence is strong for predator presence impacting habitat suitability in a “landscape of fear” (e.g. Bleicher, 2017b; Laundré et al., 2001; Rosenzweig, 1991). Also, predation risk cues drive down reproductive success in small mammals and hatch-rates in pas- serines (Ylönen and Ronkainen, 1994; Zanette et al., 2011). Once born, risk-related stress impacts development and survivorship of offspring (Cheng and Martin, 2012). And even on the community scale, fear from predators can facilitate coexistence of competing species (Bleicher et al., 2019; Fox, 2011; Kotler, 1984; Rosenzweig and MacArthur, 1963). The cognitive analysis of interpretation of cues, leading to beha- vioural decisions, is imperative to the study of how animals use re- sources, move through space, and interact with each other. Optimal patch-use theory (Brown 1988), offers a quantitative model for non- invasive analysis of risks animals sense in their environment. The model states that foragers exploiting resource patches are aware of their har- vest rate (H) and balance it with the energetic costs of harvesting (C), the predation costs (P), and the missed opportunity costs (MOC). The model, rigorously used to measure predation costs (cf. Bedoya-Perez et al., 2013), assumes the foragers recognize, and instantaneously cal- culate the risk in the environment based on the rate of substitution of survivorship with energy acquisition (f/e), probability of falling prey while foraging μ, and the perceived probability of surviving p (Brown, 1999). As we assume foragers are aware of their harvest rates, we are able to use these rates as functions of actual resource density and thus use the point when foraging ceases as an estimate of risk, i.e. a giving- up density (GUD, Brown et al., 1992). When the environment is per- ceived safer the foragers will forage more, to a lower GUD, than when https://doi.org/10.1016/j.beproc.2020.104161 Received 22 January 2020; Received in revised form 30 May 2020; Accepted 1 June 2020 ⁎ Corresponding author at: Department of Biological Sciences, Washington and Lee University, Lexington, VA, USA. E-mail address: sbleicher@wlu.edu (S.S. Bleicher). Behavioural Processes 178 (2020) 104161 Available online 04 June 2020 0376-6357/ © 2020 Elsevier B.V. All rights reserved. T