Abstract
(1) The brood reduction hypothesis concerning within-clutch hatching patterns of altricial birds was tested on the pied flycatcher (Ficedula hypoleuca). Results from control nests were compared with those from nests manipulated either at the egg stage (102 nests) or at the hatching stage (251 nests) so that normal hatching spread of about 1.3 days (range 0-2.6 days) was changed to 0-4 days. Brood size was also experimentally changed to 4-8 young. (2) Breeding performance was related to initial clutch size. Birds laying few (5-6) eggs had a lower hatching fledging success than those laying many (7-8) eggs. (3) Hatching failure increased with increasing hatching spread. (4) Nestling mortality was also generally higher for ''asynchronous'' broods than for ''synchronous'' ones, but mean nestling weight at fledging of the ''asynchronous'' broods was higher. (5) Synchronous hatching was advantageous when the survival of juveniles did not depend on fledgeling weight, contrary to the prediction of the brood reduction hypothesis. Assuming a moderate degree of dependence, the experimental results corresponded in general with hatching patterns observed in nature. Synchronous hatching may also be selected for when breeding conditions are unfavourable but are not so bad that complete breeding failures occur. (6) There seem to be some proximate constraints on starting incubation during egg production. Birds may hatch the eggs rather synchronously under unfavourable conditions, such as in marginal habitat and cold weather. This possibility of constraints may explain why hatching spread increased the later in the season the clutch was laid. Alternatively, an early initiation of incubation late in the season may reflect the need to hasten breeding to utilize declining food resources.