Abstract
BACKGROUND AND AIMS: We examined the relationship between reproductive allocation and vegetative growth in three monoicous sexual systems of bryophytes. The sexual systems show a gradient of increasing distance between the sexes, from gonioautoicous to cladautoicous to rhizautoicous. Here, we investigated the following two hypotheses: (1) reproductive allocation differs between sexes and sexual systems, and male reproductive allocation increases with increasing distance between male and female gametangia; and (2) reproductive allocation is negatively related to vegetative growth. METHODS: We sampled the three sexual systems, represented by three moss species of the genus Fissidens in the Atlantic Forest of Southeastern Brazil. Ramets were washed in the laboratory; the reproductive structures were detached from the vegetative ramets and sorted regarding sex and individual, dried at 70 °C for 72 h, and weighed in an ultramicrobalance. We calculated the mean reproductive and vegetative mass and reproductive allocation and used generalized linear models to test our predictions. KEY RESULTS: Reproductive allocation differed between species and sexes. It was higher in the rhizautoicous than in the cladautoicous and gonioautoicous species. Mean reproductive allocation was greater in males than in females of the rhizautoicous species, greater in females than males of the cladautoicous species, and did not differ between the sexes in the gonioautoicous species. Estimates of reproductive and vegetative mass were positively related in females of the rhizautoicous species. Vegetative mass was not related to reproductive allocation in the gonioautoicous species, but negatively related to reproductive allocation in the male and female branchlets of the cladautoicous species and in the female ramets of the rhizautoicous species. CONCLUSIONS: The reproductive allocation patterns differ between the rhizautoicous species and the 'truly' monoicous species, with shorter intersexual distances, which implies that our hypotheses were supported only in part. We suggest that the hypotheses should be reformulated and tested further by comparing 'truly' monoicous species with dioicous species and by including other genera.