Communities are shaped by interspecific interactions and specific adaptations, depending on the nature of the interaction (direct or indirect), allowing coexistence . Effects of competition might be masked by behavioural adaptations as a result of a shared evolutionary history. Studies of interspecific competition have mainly focused on how indirect exploitation competition affect behaviour, e.g., via territoriality and foraging strategy , and have neglected the implications of direct interference for community structure  and evolution .
Direct antagonistic interactions in competition for food, space, nest sites, or mating partners can be very costly, especially if the competition reduces individual fitness (direct interference) [3, 5–7]. Aggressive behavioural confrontations between competing species are likely if adults prey on each others offspring . Interference may result from resource-related aggression between heterospecific or conspecific individuals with potential effects on niche use and community structure (e.g., dominance or exclusion) . Nest predation has been considered a special case of direct interference via interspecific killing [3, 9]. It is a significant source of offspring mortality and can strongly affect parental life history strategies [10, 11].
Although numerous population studies have shown that nest predation is widespread in animal communities [12–14], its importance as an evolutionary force has been neglected [15, 16]. Recent studies on avian breeding biology demonstrate that the direct fitness effects of nest predation can drive evolution [10, 12, 17, 18]. Safety of nest sites and optimised breeding conditions are key intrinsic components of breeding habitat quality [15, 19]. Adults should adapt their behaviour to predation risk [18, 20] by, e.g., choosing concealed nest sites , adjusting clutch size or nestling period , and temporally partitioning foraging activity [21, 22].
Comprehensive knowledge about adaptive strategies against nest predation in mammals is lacking. Yet defence of a nest and its vulnerable young is a powerful explanation for the ultimate function of female aggression and territoriality, which is most intense during lactation and near the nest site . In addition, nests (in birds) or burrows (in mammals) serve as micro-refuges from predators and therefore have anti-predatory benefits for the adult as well .
Parental anti-nest predatory behaviour
According to parental investment theory, parents must trade off the benefits of investment into current offspring with possible negative effects on their future inclusive fitness [13, 25–27]. In addition, parents might suffer extra costs of defending offspring against predators, including time, energy, and missed opportunity costs [26, 28, 29]. Short-term adjustments are often trade-offs between behaviours, e.g., predator avoidance and foraging , and escalate in defending offspring against predators or infanticidal conspecifics [8, 23, 26, 31, 32]. The degree of escalation is influenced by offspring vulnerability and reproductive value to the parent [13, 25]. Female rodents become more aggressive during their pregnancy . Female European rabbits (Oryctolagus cuniculus) increase vigilance during late pregnancy to minimise predation . In mice, mothers successfully defend their nestling pups against infanticidal conspecifics . Attacks on pups decrease rapidly after they reach a less vulnerable stage . Maternal defence behaviour against rattlesnakes declines in California ground squirrels as snake activity declines when pups grow older .
According to the “harm to offspring hypothesis” of Dale et al. , the level of parental predation risk-taking is adjusted according to the harm the offspring will suffer without parental care. Most vulnerable are altricial nestlings incapable of escaping and defending themselves [13, 36, 37]. Predation risk is highest when the parents are away (e.g., during foraging trips) [32, 33, 37, 38]. In response, parents might alter their foraging time (e.g., reducing activity) or seek temporal or spatial foraging refuges to decrease risk [39, 40]. Cresswell  showed that nesting blackbirds altered their nest defence behaviour to compensate for predation risk demonstrating that parental behavioural flexibility in response to predation can yield fitness benefits  compared with fixed strategies . For semi-fossorial mammals, the morphology and complexity of burrows have high defence value for nestlings and thereby fulfil an anti-predator function, especially for animals with helpless altricial young [24, 43]. Hunting intruders might be confused by very complex burrow systems [44, 45]. Some species dig special parturition chambers in addition to the main burrow (e.g., European rabbits, ) and plug them to minimise predation (e.g., Columbian ground squirrel, Urocitellus columbianus, ). Increased vigilance (e.g., high scanning rates in European rabbits [47, 48]) at burrow entrances and adjusting burrow attributes (diameters, depths, or lengths of the burrow, tunnels, and nest chambers) may further lower the risk of nest predation [17, 45, 49].
Physiological stress responses to nest predation risk
Physiological responses in stressful situations, e.g., encounters of prey with predators, are evolutionarily conserved and represent a widespread and fundamental mechanism for ecosystem functioning in animal systems [40, 50]. Predator-induced changes in stress hormone metabolites associated with acute or chronic risks (reviewed in ) aim to increase survival . Animals without adequate alternative defence responses (e.g., alteration of life histories, defensive morphologies) to mitigate predation risk must engage in costly physiological responses . Elevated plasma glucocorticoids, or their faecal metabolites, are often measured as indicators of such physiological stress responses [50, 52–55].
Study system and hypotheses
Here we compare behavioural strategies among different interaction types: interspecific resource competition, intraspecific interference competition, and interspecific nest predation. We used a small mammal study system including the semi-fossorial common voles (Microtus arvalis) as focus animals, the greater white-toothed shrews (Crocidura russula) as potential nest predators, and field voles (Microtus agrestis) as interspecific competitors. All three species coexist in many habitats of the northern hemisphere, overlapping in their common habitat and, to a smaller extent, in their diets [56–58]. Voles live in large burrow systems with underground tunnels and corresponding runway systems above ground. Shrews explore vole tunnel systems to search for invertebrate food . Because of the limited space underground, shrews may react aggressively when encountering tunnel inhabitants. Additionally, they can act as nest predators on altricial vole nestlings by plundering their easily accessible nest chambers . This shrew behaviour might intensify during environmentally-adverse seasons when invertebrate food is scarce, e.g., in autumn .
To protect vulnerable altricial nestlings from nest predators, voles should have evolved anti-nest predator strategies to secure their reproductive success. Getz et al. reported nest defences in two American vole species, Microtus pennsylvanicus and M. ochrogaster, even against larger shrew species, e.g., Blarina brevicauda. Shrews only successfully preyed on nests when vole mothers were absent on foraging trips .
What kinds of parental behavioural strategies can secure nestling survival and allow the coexistence of competing species in a nest predator–prey system? How do these strategies differ from responses to non-nest predator antagonists?
We hypothesised that vole mothers adjust their burrow systems in response to antagonists . If she encounters a nest predator during pregnancy, we expected to find different nest architectures at parturition than in the absence of predators.
We further hypothesised that vole mothers react to nest predation risk by altering their time budgets and habitat use. To guard vulnerable nestlings, vole mothers should increase vigilance and nest guarding behaviour (e.g., at burrow entrances), especially during the first sensitive nestling phase. We expected guarding behaviour to be more intense in the presence of a potential nest predator than in the presence of an inter- or intraspecific resource competitor. In addition, if guarding behaviour incurs extra costs for vole mothers, e.g., owing to shortened foraging periods, we expected that to be reflected in the mother’s or nestlings’ body conditions.
If vole mothers exhibit physiological stress reactions concomitant to these behavioural reactions (A and B), we expected levels of faecal corticosterone metabolites (FCM) to correlate to the presence or absence of different antagonists.
Thus, we hypothesised that the levels of stress hormone metabolites in expectant vole mothers increased when they encountered shrews as nest predators.