Breeding phenology in a marine top-predator: individual and population response to environmental variation

Abstract

[eng] An emerging consequence of global climate change is a shift of seasonal biological events. In animals, climatic divers have always been found to guide population trajectories and individual life-history tactics, but it is not clear whether individuals will be able to adapt to the fast paste at which temperature and rainfall are changing. The analysis of long-term data sets reveals different degrees of plasticity in the phenological responses of individuals, within and between species, but the pattern and the payoff of this plasticity are not always known. Understanding individual responses is essential to reveal the mechanisms and implications of climate-related changes on population dynamics and individual life-history tactics. We used longitudinal data collected over eighteen breeding seasons (2001-2018) on the laying date of Cory’s shearwaters Calonectris diomedea, at Pantaleu islet at Dragonera Natural Park, Mallorca (Balearic Archipelago, Spain). The Cory’s shearwater is a trans-equatorial migratory seabird that spends the winter (October-March) into the Atlantic Ocean before returning into the Mediterranean Sea to breed (April-September). Individuals breed in colonies and egg laying occurs more synchronous than in other seabirds, however, the laying date at population as well as individual level change every year, possibly as a response to winter ocean state (e.g. through carrying over effects) and/or conditions at the breeding grounds. We first considered the winter North Atlantic Oscillation index (wNAO) as a proxy of environmental variability and assessed whether its variation correlated with the variation in the average laying date at population level. Subsequently we used longitudinal data to investigate whether individuals were consistent in their laying date relative to the population average, i.e. repeatability of laying date. Finally, using mother-daughter data we assessed whether the laying date has a heritable component. We found that reproductive onset at population level correlated positively with the climatic index. This was fully explained by individual plasticity because, despite a change in the average laying date, individuals were consistent in their trait expression. Indeed, the repeatable of this trait was high (R = 0.517), with some birds breeding consistently earlier than others. Interestingly, trait repeatability was higher for those females born at the colony (R = 0.653) compared with all females. The heritability of laying date assessed using 29 half parent- half offspring pairs was high, but the value suffered of a large uncertainty due to a relatively small sample size (h 2 = 1.14, R2 = 0.12, p=0.07). We thus used another approach and sorted females into two groups, namely ‘early’ and ‘late’ breeders, according to whether laying date occurred before or after the median of the population respectively. We compared the proportion of each combination through a 2x2 contingency table. Results showed a statistically significant difference (X2 1=7.79, p<0.01) confirming a high heritability of laying date. Results from general linear mixed models indicated that early breeding birds had 27% more chance to breed successfully than late breeders (breeding success: 0.83 and 0.60, for the earliest and the latest laying date, respectively; z=-2.68, p<0.01). In conclusion, our work showed that changes in the laying date correlate to environmental conditions and that despite a high plasticity, individual repeatability is high as it is the heritability of this trait. Despite this, the advantage of an early breeding seems small suggesting that inter-individual variability is probably the result of differences in phenotypic quality intrinsic to the population.