Discussion
This experimental study investigated the effect of food availability on the intensity of the anti-predator behaviour of breeding red kites. Under experimentally enhanced food conditions, the intensity of anti-predator behaviour was higher in red kite parents with old broods than with young broods. This age-dependence disappeared in the un-supplemented control group, where anti-predator behaviour was constant over the entire nestling period. These results support the previous theoretical prediction of Dale et al. (1996) that the “reproductive value of offspring” hypothesis has greater relevance under favourable breeding conditions, while the “harm-to-offspring” hypothesis becomes more relevant under poor breeding conditions. These results suggest that food availability affects parental anti-predator behaviour by changing the nestlings’ body condition which represents a new pathway of how food conditions drive parental investment.
Under a wide range of food conditions, food supplementation to parents might mainly affect parental condition rather than vulnerability of the brood (Boutin 1990; Michel et al. 2022; Ruffino et al. 2014). However, in recent studies we showed that enhanced food conditions due to our experimental food supplementation increased nest and nestling survival, as well as body condition of nestlings compared to control broods (Catitti et al. 2022; Nägeli et al. 2022). This confirms that food supplementation reduces the harm that offspring suffer due to a period of parental absence by increasing their baseline body condition – and probably also increases the reproductive value of the brood due to increased post-fledging survival probability of nestlings (Naef-Daenzer and Grüebler 2016).
The greater relevance of the “harm-to-offspring” hypothesis under normal than enhanced food conditions suggests that also large species, such as the red kite, can be affected by predator presence in the nest area. Field observations during predator exposure trials support that, even if mobbing intensity is low, red kite parents invest time in supervising their brood and the predator; time that could otherwise be spent in food provisioning (Ghalambor et al. 2013; Ibáñez-Álamo et al. 2015; Martin and Briskie 2009; Mutzel et al. 2013). In this respect, parents of nestlings that are susceptible to predation are faced with a trade-off between time invested in nest guarding and time invested in foraging (e.g. Komdeur 1999; Rothenbach and Kelly 2012). The outcome of this trade-off likely represents the underlying mechanism of adjustments in mobbing intensity. During food shortages, when nestlings have more urgent food requirements and are more susceptible to harm from starvation or developmental stress (see Catitti et al. 2023), chasing away predators should expedite the return to foraging. On the other hand, during favourable food conditions, breeding pairs can spend more time passively guarding the nest and, thus, can save energy and avoid risking themselves during active mobbing behaviour. This might be true for bird species where nest guarding and attendance can be efficient anti-predator strategies (Catry et al. 2006; Dewey and Kennedy 2001; Hu et al. 2017; Rothenbach and Kelly 2012; Samelius and Alisauskas 2001). We suggest that nest guarding and mobbing represent two different nest defence strategies that both reduce nest predation (Caro 2005; Montgomerie and Weatherhead 1988), but their cost-benefit ratio changes with food availability and vulnerability of the brood.
As the two hypotheses show contrasting predictions for the effect of nestling age on mobbing behaviour, we used nestling age to investigate the relevance of the two hypotheses under different food conditions. However, also brood size showed a clear effect on mobbing behaviour. While the “harm-to-offspring” hypothesis does not give a clear prediction regarding brood size, the “reproductive value of offspring” hypothesis does (Montgomerie and Weatherhead 1988) and is supported by increased mobbing intensity in parents with large versus small broods. We recently showed that large brood sizes were associated with reduced body mass and increased corticosterone levels in red kite nestlings (Catitti et al. 2022; Nägeli et al. 2022), illustrating the general life-history trade-off between offspring number and offspring quality (Stearns 1992). Since it is the body condition of the nestlings that is expected to affect parental mobbing behaviour under the “harm-to-offspring” hypothesis (Dale et al. 1996), these results suggest both, increased harm to offspring, as well as increased reproductive value in large broods. Thus, increased anti-predator investment into large broods suggests that the reproductive value of an additional nestling is larger than the increased costs of reduced body condition arising in a period of absence no parental care. This might be a general pattern as increased parental care investment into large broods, including anti-predator investment, has been shown in many studies (Clutton-Brock 1991; Lazarus and Inglis 1986; Royle et al. 2012).
Quantifying time-to-detection was important in this study because differential detection could bias the investigation of capture probability, which was used as proxy for mobbing intensity. Predator detection time is not only a methodological issue, but is also likely associated with nest predation risk, should low detection probability be due to parental absence from the nest area (Behrens et al. 2019; Duncan Rastogi et al. 2006; Samelius and Alisauskas 2001; Schmidt and Whelan 2005). Under poor natural food conditions, parents detected the decoy predator later than under favourable food conditions probably due to longer foraging trips. Supplementary feeding resulted in earlier detection under poor natural food conditions, but later detection under favourable food conditions. In addition, parents with large broods detected the decoy predator earlier than parents with small broods indicating a higher nest visitation rate. These results are in line with recent studies showing that, in years with low food availability, the home-range size of red kite pairs is considerably larger than in years with high food availability (Pfeiffer and Meyburg 2015), and that prey delivery rate is elevated in large versus small broods (Andereggen 2020), but can be reduced under ad libitum food conditions. Together, these results indicate that time-to-detection depends on factors affecting movement behaviour within home-ranges and corroborates that nest predation risk by avian predators can be increased in pairs with large home-ranges (Lameris et al. 2018), in situations of low food availability (Duncan Rastogi et al. 2006), and in pairs with small brood-sizes (Schmidt 1999), all being consequences of low habitat quality.
Finally, weather conditions during the predator exposure trial affected time-to-detection probability, as well as time-to-capture. This was expected for time-to-detection, because weather conditions are shown to affect ranging behaviour (Baucks 2018) and food delivery rates (Andereggen 2020). However, while we added weather variables mainly to account for potential biases in the analysis of time-to-capture, our study is one of the very few showing that weather conditions affect mobbing intensity (but see Fisher et al. 2004). Mobbing intensity strongly decreased i.e. time-to-capture increased with high ambient temperatures and windy conditions. When ambient temperatures are outside their thermal neutral zone, birds face additional energy costs during activities, which could affect their decisions regarding nest defence (Fisher et al. 2004). Also, windy conditions can affect control of swoops and, thus, increase injury and predation risk of parent birds. This is also supported by the fact that time-to-capture was increased when the predator was placed closer to trees, impeding manoeuvrability. Both underlying mechanisms, energetic trade-offs and increased threat of injury, might be particularly relevant in large bird species exhibiting predominantly soaring flight, such as red kites, where flapping flight is energetically costly, and manoeuvrability limited (Sapir et al. 2010; Shepard et al. 2013; Shepard et al. 2016; Shepard et al. 2019).
In conclusion, we show that nest defence depends not only on predation risk due to predator type and behaviour, but also on environmental conditions. While, in general, parental anti-predator investment seems to be adjusted to the reproductive value of the brood, this investment is modulated by a multitude of factors associated with the brood, the environment, and the approaching predator. In particular, this study provides evidence that food availability affects anti-predator behaviour by altering the body condition of nestlings. The vulnerability of the offspring is therefore important for the choice of the nest defence strategy. It represents a driver of mobbing intensity and is important for the outcome of trade-offs between different forms of parental care. Thus, low food availability might have mobbing-mediated consequences for reproduction and reproductive costs beyond the consequences mediated through changes in foraging behaviour, even if predation rate remains unchanged. In addition to potential survival costs for the parents, the additional parental effort due to frequent mobbing may even be a reason for brood desertion under poor environmental conditions (see Nägeli et al. 2022). Ultimately, large-scale environmental factors affecting investment into nest defence may have significant demographic consequences.