Brood size and the economy of brood defence: examining Lack's hypothesis in a biparental cichlid fish

Synopsis We tested the explanatory value of two hypotheses reviewed by Lack (1954) in the maintenance of brood size in free-ranging convict cichlidsCichlasoma nigrofasciatum: (1) physiological constraints on egg production, and (2) behavioural constraints imposed by brood defence. Number of free-swimming young in 13 experimental (E) broods was augmented to the upper limit of the size distribution of natural broods (150 young); 18 control (C) broods were handled in the same way but brood size was not changed (mean ± SE = 69.5 ± 11.0). E and C brood sizes were measured at 5 day intervals. At day 20 (just before independence from parental care), 50.3 ± 9.4 (n = 9) young remained in E broods and 30.8 ± 7.8 (n = 8) young remained in C broods (p> 0.05). Offspring number did not differ significantly (p> 0.05) between C and E broods after day 10. Mean growth rate of offspring was significantly lower in E broods than in C broods, perhaps in response to increased density of young in the former. Both the convergence of offspring number in E and C broods and suppression of growth in E broods support a behavioural constraint; that during the first 10 days in which the young are free swimming, two parents are unable to defend large broods as successfully as small broods. A trade-off exists in parental investment between current and future reproduction. Extra-parental investment in current reproduction (eggs) does not result in an increased number of young at independence, therefore a behavioural constraint during brood defence should stabilize the evolution of clutch size.