Context-dependent anti-predator responses in wild ungulates: inter-individual variation in relation to the prevailing predation regime
Predation exerts one of the strongest selective pressures in natural systems. Prey commonly adjust their behaviour in response to predation risk, and these anti-predator responses can shape individual fitness, population dynamics, and even community structure. For large herbivores, humans have become the dominant predator across much of the world. However, their impact on prey behaviour may differ fundamentally from that of natural predators, due to differences in hunting tactics, spatio-temporal predictability, and selectivity. Individuals also consistently differ in behavioural traits (i.e. inter-individual variation or animal personality), such as boldness, or risk-taking. These differences can strongly influence vulnerability to predation. Yet, little is known about how predator selectivity and the co-occurrence of multiple risks (e.g. natural predators and humans) affect both the mean level and the variance of behavioural traits within wild prey populations (Leclerc et al. 2017; Bonnot et al., in prep). The maintenance of inter-individual variation in behaviour has important eco-evolutionary consequences (Réale et al. 2007; Wolf & Weissing 2012). For example, the non-random distribution of behavioural phenotypes across landscapes (Khazar et al. 2025) can affect spatial population structure, gene flow, and ultimately genetic diversity. If predator selectivity disproportionately targets certain personality types, this may lead to behavioural homogeneisation and reduced resilience to environmental change.
As part of the ANR-funded AnthropoFear project (“Landscapes of FEAR in Europe’s Anthropocene”), this postdoctoral project aims to investigate the role of predation context in shaping anti-predator behaviour (i.e. mean and variance) of large herbivores, with a focus on the roe deer (Capreolus capreolus) as a model species. It combines experimental manipulations, analyses of long-term datasets, and collaborative research networks. The project will mainly address the following core questions:
1/ How do average anti-predator responses of large herbivore prey differ in relation to the prevailing predation context?
Spatio-temporal variation in anti-predator behaviours, eg vigilance levels and escape, should all depend on the predation context (e.g. predator hunting method; single vs multiple predators). For instance, while vigilance is an efficient anti-predator behaviour for escaping a cursorial predator (i.e. wolf) that typically chases their prey over long distances across the landscape, it is likely less efficient against an ambush predator, like the lynx, that hunts predominantly in forested habitats. Using a playback experiment protocol (see Suraci et al. 2017, Bhardwaj et al. 2023), we will assess the anti-predator responses of wild prey ungulates to various predator sound stimuli (human, wolf, lynx) that differ in terms of hunting method and intensity. Using camera traps, the behavioural responses of ungulates to prey sounds will be video-recorded (eg vigilance, flight, latency to resume activity, habituation). This protocol has been successfully displayed during winter 2024–25 at the Grimsö Wildlife Research Station (Sweden) thanks to our long-term collaboration with SLU University. Those data have been preliminary analysed as part of an internship and could be turned into a standalone publication. The next step is to replicate the experiment across multiple sites with contrasting predation regimes, including in study sites with GPS- monitored, individuals. In these cases, we will be able to investigate the influence of other individual characteristics (eg personality-type) and habituation on the observed anti-predator responses. The postdoctoral researcher will be expected to take the lead in implementing the experimental protocol. To do so, he/she will be able to rely on the support of the CEFS technician staff and students.
2/ How does the predation context modulate among-individual variation in anti-predator behaviour?
When facing a stress or a threat, individuals within populations do not react in the same way depending on their personalities or behavioural types. If these behavioural types experience consistent differences in context-dependent or habitat-dependent performance, this could result in varying diversity of behavioural types (i.e. inter-individual variation) among (sub)populations. Such changes in diversity of behavioural types can have cascading impacts on population composition, dynamics, resilience, and ecosystem services and sustainability and thus, eco-evolutionary dynamics (Sanderson et al 2023). For example, if predators disproportionately select against certain personality types, this should result in homogeneisation of behavioural variation (i.e. lower diversity of behavioural types) and reduce the ability to respond to future threats. On the other hand, the coexistence of different predator species, with different hunting modes and prey selectivity, should result in a wider diversity of behavioural tactics within populations.
To test the hypothesis that populations exposed to multiple predators will maintain a greater diversity of behavioural tactics, we will exploit the long-term GPS monitoring dataset managed by the Eurodeer network (22 roe deer populations across Europe). Based on preliminary work (Khazar et al. 2025), we will use repeatable behavioural metrics (movement, space use, activity) to identify individual risk-management tactics and assign individuals along a shy-bold continuum. The within-population variance and the structure of the behavioural syndromes will then be compared across populations faced with different predation contexts (eg intense hunting, vs wolf and hunting).
3/ Depending on his/her research interests, other datasets and questions are also available for the postdoctoral researcher to tackle. For instance, the impact of individual personality phenotype on life-history performance could be investigated thanks to an intensive dataset available on roe deer demographic parameters (neonate growth rate, survival, phenology of parturition, female reproductive success). Preliminary analyses on hunting-induced behavioural adjustments derived from accelerometery data have also been performed recently and need to be investigated more deeply.
Required qualifications
– PhD in behavioural ecology
– Demonstrated research experience on predator-prey interactions, prey antipredator behaviour or animal personality
– Knowledge or interest on population demography / life-history parameters
– Demonstrated skills in scientific writing
– Demonstrated skills in data analyses using R
– Excellent collaborative, communication and interpersonal skills
– Speaking French is not required but would be an asset
– A valid driving license is necessary
Job environment
The person will be based in the Comportement et Ecologie de la Faune Sauvage (CEFS) unit at INRAE-Toulouse (France). He/she will interact closely with supervisors and peers, at the CEFS (notably with Mark Hewison and Laura Gervais), and benefit from our long-standing collaboration with scientific partners throughout Europe (Francesca Cagnacci, Petter Kjellander, and many more), and have privileged access to the Eurodeer network datasets. The postdoctoral project duration will be a minimum of 18 months, with the possibility of extension depending on the funding level. The salary is dependent on the candidate’s experience.
For more information, potential candidates should contact Nadège Bonnot (nadege.bonnot@inrae.fr).
To apply, please send your CV, a cover letter describing your motivation and relevant experience, and contact details for references. Interviews with candidates will take place on a rolling basis, starting at the end of August 2026. The recruitment process will continue until the post is filled. Ideally the successful applicant will start in late 2026 or early 2027, but the starting date can be adjusted depending in the candidate’s availability.
Commentaires récents