Sleep in the landscape of fear

Like us, nearly all animals sleep. Sleep provides many key benefits: it consolidates memories, cleanses the brain from detrimental by-products of metabolism, boosts the immune system and supports the immune response to infections, for example. Sleep loss is thus accompanied by many detrimental effects on health and, likely to limit these effects, sleep loss is often followed by longer sleep. But are there costs associated with sleeping? If so, what are the selective pressures that constrain sleep time? Finally, how do animals cope when they cannot sleep as much as needed? We currently have no answers to these questions because, due to technological limitations, sleep is mostly studied in the laboratory where individuals do no face any of the constraints on sleep time, nor the costs and benefits of sleep loss. To begin to address this major knowledge gap, we have successfully pioneered the use of state-of-the-art multi-sensor biologgers Daily Diaries (DDs), designed to study complex behaviour in the wild, to identify sleep postures accurately in wild fallow deer fawns. In this project we will use DDs to answer the long-standing question of how predation risk affects sleep in an ideal model: wild red deer under predation by wolves and hunting by people.

Sleep is potentially costly for wild animals since, compared to when awake, sleeping individuals are less aware of what happens around them and less quick to respond to changing conditions, such as approaching predators. The predation risk hypothesis proposes that higher predation risk favours shorter sleep time, an idea supported by comparative studies in mammals. Whether and how predation risk affects sleep at the individual level and the strength of selection it imposes at the population level are however unknown. The ecological theory of the landscape of fear posits that prey change their behaviour (e.g. more vigilance, less foraging) and physiology (e.g. higher stress) when predators are present and when using areas perceived to be risky. Likewise, fear of humans alters behaviour across entire wildlife communities and increases stress levels in hunted deer. Given the reduced awareness and increased vulnerability during sleep, individuals should adjust sleep when facing higher predation risk or hunting. We will test the predictions derived from the predation risk hypothesis combined with the landscape of fear that, under increased predation risk or hunting pressure, daily sleep time is reduced and more fragmented; sleep onset is delayed; sleep loss is followed by longer sleep.

We will compare sleep in two wild red deer populations in the Czech Republic that have been studied since 2019 with DDs and GPS. At Doupov, wolves established a pack and red deer constitute 30% of their diet. Year-long data show that deer altered their spatial behaviour after the arrival of wolves, as predicted by the theory of the landscape of fear. The Kladská population is predator free. At both sites, ~80% of deer, including collared individuals, use winter enclosures for supplementary feeding. Wolves can enter the Doupov enclosure and a third of their 30 predation events is accurately dated allowing us to link predation events to changes in sleep in deer. Both deer populations are culled with known dates and locations of hunts so that we can assess how hunting affects sleep. We will collar captive red deer with DDs and record their behaviour during standard sleep experiments to refine the identification rule for sleep using Daily Diaries and to quantify changes in sleep and stress levels after exposure to predator cues.

This work will be the first to demonstrate how predation risk affects sleep and will be the template for much needed studies on sleep under natural conditions.