Breeding during a Predation Regime Shift: Behavioural and physiological responses of female common eiders facing the recovery of their native predator

Predation is a key selective force driving prey trait evolution. Yet, the extent to which predator-consumptive and predation risk effects shape phenotypes is still poorly understood. Since predation risk increases the cost of current reproduction, long-lived prey are predicted to prioritize survival and future reproduction by reducing reproductive investment under high predation risk (e.g., through changes in life-history, behaviour, physiology or morphology). However, the relative roles played by adaptive plastic responses (i.e., withinindividual phenotypic variation) and selective processes (i.e., among-individual phenotypic variation) in allowing populations and individuals to persist under fluctuating predation threat in the wild have been overlooked.
In my thesis, I shed light on these knowledge gaps by evaluating the responses of a sea duck species, the common eider Somateria mollissima, to fluctuating predation risk from their native predator, the white-tailed eagle Haliaeetus albicilla. I focused on adjustments of behavioural and physiological traits to the recovery of white-tailed eagles in southwestern Finland. As a longlived species characterised by energetically-costly fasting during incubation and frequent intermittent breeding, common eiders are ideal for studying the life-history trade-off between reproduction and survival. I relied on a longterm individual-based monitoring of breeding females (> 2500 breeding attempts from > 1000 individuals in 2008-2022) to explore variations in risktaking behaviour (measured by flight initiation distance) and physiological proxies of energy allocation (baseline corticosterone levels) and parental effort (baseline prolactin levels) in response to temporally and spatially variable predation risk. The longitudinal data collection was not restricted to the focal variables, but also integrated a monitoring of female characteristics (e.g., age, body condition and clutch size) and parameters such as nesting microhabitat and fine- and large-scale predation risk indices, allowing a good understanding of concurrent changes occurring in the population.
First, in agreement with the prediction that long-lived prey should reduce parental investment under elevated predation threat, I found that increasing predation threat was associated with longer flight initiation distances (i.e., reduced risk-taking) over breeding attempts. Importantly, this response was explained by a plastic adjustment of flight initiation distance to the risk of predation on breeding females, but also by the selective disappearance of risktaking phenotypes from islands characterised by high nest predation risk. Supporting the latter notion, I showed that risk-taking strategies affected fitness (including survival, breeding propensity and hatching success), especially under elevated predation risk. My results thus provide among the first evidence that the target of predation (adults or offspring) determines the mechanisms driving adjustment of risk-taking propensity to perceived predation risk. In risky environments, females were also more likely to rely on prior experience with a threat to dampen their flight response when repeatedly confronted with a harmless stimulus (here, a human approach over consecutive days). Such short-term habituation to non-life-threatening stimuli is likely adaptive and could limit the reproductive costs of an overreaction to benign stimuli.
Second, I found that baseline corticosterone levels were positively associated with reproductive investment. In addition, baseline corticosterone levels were downregulated when incubating under high predator activity, again reflecting decreased energetic reproductive investment under threat of predation. Additionally, this physiological adjustment was habitat-dependent, as females downregulated baseline corticosterone levels under higher predation risk on themselves on open islands (associated with lower adult survival) but not on forested islands where nests are visually concealed from predators. Last, while I predicted that incubating females would display lower baseline prolactin levels – mediating reduced parental effort – under high predation threat, this prediction was not directly supported by my results. However, elevated baseline prolactin levels were still found to be indirectly related to predation pressure. Indeed, elevated prolactin levels seemed to promote hatching success in years characterised by elevated predation pressure. Overall, high prolactin levels favoured hatching success under constraining extrinsic (e.g., poor climatic conditions) or intrinsic conditions (e.g., when incubating a large clutch).
Thus, the results drawn from my thesis provide valuable insight into the short- and long-term responses of prey individuals and populations to a changing predation regime. Interestingly, most of the observed responses seemed to be driven by within-individual adjustments of behaviour or physiol...