Abstract
Here we compare the chick-rearing capacity of common terns Sterna hirundo that were given an additional chick at hatching (incurring rearing costs but no production costs), an additional egg at the start of incubation (incurring incubation and rearing costs) and made to lay an additional egg (incurring laying, incubation and rearing costs). The brood enlargements were within the natural range of brood sizes (i.e. two increased to three). Only those pairs that were given a free chick reared a significantly larger brood compared with unmanipulated controls. The results show for the first time a within-clutch trade-off between producing and rearing extra young. Although there was no effect on egg or hatchling size as a result of the increased egg production, parents incurring the full costs of producing extra young subsequently showed reduced chick provisioning, growth and survival. These results cast doubts on the interpretation of previous brood enlargement experiments as providing empirical evidence that observed clutch sizes are often less than the Lack clutch size, as such experiments have failed to include production costs.