Natural Selection and Clutch Size in the European Starling

In a field experiment, broods of 1—10 chicks of equal age were artificially established. The natural food supply was supplemented with an ad libitum supply in some of the nests of 6—10 chicks per nest. Among broods that did not receive extra food, weights gradually increased with brood size up to size 3 chicks per nest, gradually decrerased when brood size increased to 6 chicks per nest, and dropped precipitously with additional increase in brood size beyond 6 chicks per nest. Nests of 10 without extra food wer always completely unsuccessful. Broods of sizes 5 and 6 were probably optimal because their chicks were almost as heavy as those of smaller broods. Brood size 6 is probably the upper limit at which parents can increase their foraging efforts with increasing brood size. The observed mean clutch size of 4.68 @+ 0.16 (1.96 SE) is close to, but less than, the optimal clutch size. This evidence is consistent with predictions assuming that the population is reproducing maximally and supports Lack's theory of maximized reproductive rates. Among broods that did have an ad libitum food supplement, chicks in broods of 10 and below achieved weights as large as those of chicks in much smaller broods that did not receive extra food. This indicates that the ability to rear young is a function of both brood size and the current food supply, supporting Lack's theory that the breeding season's food supply is a determinant of the selection for clutch size in nidicolous species. The intrinsic growth rate (ro of the logistic model) seems to be unaffected by either food supply or brood size, but the duration of growth appears to be shortened by an abrupt peak energy demand in excess of the available food. Thermoregulatory stress may cause poor growth in broods of 1 and 2 chicks and the observed larger—than—normal weight recession of the bigger broods. A new method of estimating the value of the parameters ro and K of the logistic equation from empirical data, without a priori knowledge of the asymptote, is briefly presented.