Pourquoi le Goéland argenté, Larus argentatus pond-il un troisième œuf plus petit que les deux précédents ?

Abstract

Breeding performances of Herring Gulls (Larus argentatus) nesting on Trébéron island (Brittany) were compared during two successive periods. During the first period (1983-88) feeding conditions were favourable : breeders regularly visited the very large Brest’s refuse tip, situated at 12 km from the colony, where food resources were abundant and predictable. During the second period (1989-90), following the building of an incineration plant, the amount of garbage dumped at the tip decreased, thus reducing the amount of food available to the gulls. Between 1983 and 1988, mean clutch sizes were very high (2.8-2.9, Table I) and comparable with the highest estimates made at other European colonies. The estimates for 1989-90 were significantly lower (2.7) than during the previous period. In 1989-90, modal clutch size was still of three eggs, but the proportion of clutches with one or two eggs was three times larger (Table I). When garbage was incinerated, egg volume decreased by 2.1 % (second egg) and 3.8 % (third egg, Table II). Whatever the year, there was no significant difference in volume between the first and second eggs (except in 1988). In contrast, the third egg was about 10 % smaller each year than the two previous eggs (Table II). The production of young decreased from an average of 1.30 per breeding pair in 1983-88 to 0.65 in 1989-90. It was noticed that even when the feeding conditions were very favourable, 13 % of the breeding pairs only produced three fledglings (Table IV). The proportion of breeding pairs successfully raising two or three fledglings markedly dropped in 1989-90, but not the proportion of pairs raising a single young (Fig. 1). This suggests that the brood-rearing capacity of the parents is food-limited and that their reproductive effort greatly increases with brood size. The results obtained in 1989-90 show that a low food availability limits egg production in the Herring Gull. However, those obtained between 1983 and 1988 suggest that food limitation does not explain either the truncated distribution of clutch size or the small size of the third egg. It is therefore proposed that laying three eggs represents the best possible trade-off between the cost of reproduction and the lifetime reproductive success, and that the third egg is a «spare-egg» producing a fledgling only if the (a) egg and/or the (b) egg do not hatch.