BROOD SIZE, HATCHING ORDER AND HATCHING DATE - EFFECTS ON 4 LIFE-HISTORY STAGES FROM HATCHING TO RECRUITMENT IN WESTERN GULLS

1. We examined survival and recruitment (into the breeding population) in three cohorts of western gulls Larus occidentalis (Audubon) in rel ation to brood size at hatching, hatching order and hatching date. We divided the pre-breeding period into four stages: (i) the natal stage- 'pre-fledging' from hatching to independence; (ii) the juvenile stage- the period from independence to age 1 year; (iii) the subadult stage- the second and third years of life; and (iv) the recruitment stage-the period when adults breed for the first time. 2. Hatching date and hat ching order (in order of importance) each had a strong, negative effec t on survival in the natal stage. Brood size had little effect on nata l survival; however, after controlling for hatching date there was a s ignificant negative effect. Juvenile survival covaried positively with brood size, negatively with hatching date, and negatively (but weakly ) with hatching order. No variate correlated with survival in the suba dult stage; however, recruitment probability of adults covaried negati vely with hatching date. Effects of each variate did not reflect confo unding by parental age. 3. Of the three variables, only the effect of hatching date varied between cohorts, and only in the natal stage. 4. The post-fledging survival advantage of hatching early in the season ( expressed in the juvenile and recruitment stages) may reflect a direct effect; chicks fledging early in the season may establish lasting dom inance over other chicks, and/or an indirect effect (better parental q uality). We suggest that the post-fledging advantage of hatching into a large brood probably reflects a direct effect; competitive abilities may covary positively with brood size due to interaction between sibl ings, and/or an indirect effect (parental quality). 5. Relative recuit ment rate of different cohorts was accurately predicted by natal survi val. However, natal survival was a poor indicator of the probability t hat a given hatchling recruited because of selection occurring post-fl edging on differences related to brood size and to a lesser extent on hatching order and hatching date. As a result, each parent's reproduct ive success was poorly measured by natal survival of their young. The data suggest that survival of young to age 1 year provides a more reli able estimate of recruitment probability and a parent's reproductive s uccess.