Impact of climate changes on an Antarctic seabird: Role of individual quality

Background/Question/Methods

Individuals within a population are different (e.g. age or breeding experience). Taking into account individual differences in responses to climatic fluctuations can change our understanding of how populations may cope with a warming climate.

In this study, we examined the effect of climate fluctuations over a 47-year period on a seabird breeding in Antarctica, the Southern fulmar (Fulmarus glacialoides). First, we built a life cycle based on reproductive status including individuals coming back to the colony (observable birds) or not (unobservable non breeders). Among birds that returned to the colony, we distinguished birds breeding or not. Among breeders, we distinguished birds that succeeded in having a fledging chick or not (egg broken or chick dead). We thus considered four stages: successful breeders, failed breeders, observable and unobservable non breeders.

We then estimate demographic parameters using multistate mark-recapture model. Those parameters described transitions between stages from time t and t+1; thus may depend on stages at time t (i.e. reproductive stage at the previous breeding season) and will exhibit or not temporal variation. To study the effect of climate variations on vital rates, we included climate variables as covariates in the mark-recapture model.

Results/Conclusions

Demographic rates showed strong differences among individuals related to their reproductive status the previous year.

The return probabilities showed temporal variations. Petrels are known to skip breeding as a result of a trade-off between survival and reproduction, with individuals skipping breeding when their body conditions are insufficient to breed. Previous breeders return probabilities were higher than those of previous non breeders.

Previous successful breeders showed higher breeding and success probabilities than other stages. Individual quality may explain this pattern. Previous successful breeders could be considered as “good quality” individuals compared to “lower quality” individuals (failed breeders, non breeders). They had a higher probability to come back and breed because they may have better body conditions.  They were also more likely to raise successfully a chick.

Observables non breeders seemed to have the highest survival probability followed by successful breeders, failed breeders and unobservable non breeders, although model selection did not support strong differences among stages. Mechanisms underlying this pattern remained unclear.

We are examining the effect of sea ice extent, sea surface temperature and a-Chlorophyll rate over different seasons of the life cycle and expecting contrasted responses to climate among breeding stages.