Ecological Correlates of Labile Sex Expression in the Orchid Catasetum Viridiflavum

Abstract

Patterns of sex expression were studied in a natural population of the epiphytic orchid Catasetum viridiflavum on Barro Colorado Island (BCI), Panama. Data on gender and flowering frequency were collected for 2 yr from 186 orchids occurring in both young (70—100 yr old) and mature forests. The distribution of phenotypic gender ratio was strongly bimodal in the population, and sex ratios were significantly affected by forest age. In the young forest was statistically indistinguishable from 1:1. Gender in C. viridiflavum is largely determined by light intensity. Females occurred most frequently in open canopies,and males predominated in closed canopies. Experimental shading of developing inflorescences of well—illuminated plants duplicated this pattern, suggesting sex expression is controlled by cues received by the developing inflorescence. Small, but significant effects of plant size on gender were also detected. Forty—two percent of plants occurred on dead, decaying wood (mostly standing dead trees), while the remainder occurred on live trees. Substrate type and canopy openness are correlated because dead trees present no canopy to shade epiphytes on them. Differences in sex ratio due to forest type resulted from a more frequent occurrence of plants on standing dead trees in the young vs. the mature forest. Theory for the evolution of environmental sex determination (ESD) assumes that environment affects reproductive success (RS), but that the relationship between RS and environment is different for males and females. Light availability and substrate were investigated for their effects on two components of RS, flowering frequency and vegetative growth. C. viridiflavum on dead, decaying wood flowered more often than those on live trees, but no significant effect due to canopy openness was detected. The influence of substrate and light availability on estimated rates of vegetative growth varied from year to year, but increased rates were associated with either dead, decaying wood or high light intensities in at least 1 yr. Thus, explanations for the evolution of ESD in C. viridiflavum must include potential effects of both substrate type and light environment on male and female RS. Plants growing on dead, decaying wood in an open habitat that produced fruit exhibited reduced vegetative growth compared to nonfruiting plants, indicating there is long—term cost to fruit production in C. viridiflavum. In forest habitats, females setting fruit on live trees did not flower in the following year. Most nonfruiting females on this substrate type, as well as fruiting and nonfruiting plants on dead trees, flowered in the following year. Although sample sizes supporting this pattern were small, it is consistent with the hypothesis that sex differences in costs of reproduction in response to changes in substrate (and correlated changes in light availability) may explain the evolution of ESD in C.viridiflavum.