Fruit set in a deceptive orchid: The effect of flowering phenology,display size,and local floral abundance

We investigated the effect of flowering time, display size, and local floral density on fruit set in Tolumnia variegata, a pollination-limited orchid that offers no reward to its pollinator(s). During 1990, natural variation in flowering time, display size, and fruit set were monitored in 508 plants at one locality in Puerto Rico. The following season, orchid floral abundance per host tree (Randia aculeata) was manipulated to investigate its effect on fruit set. Four floral abundance treatments were established (700, 500, 300, and 100), each replicated four times. Flowering time was the most important trait affecting fruit set. The proportion of plants setting at least one fruit was significantly high early and late in the season, but low during the flowering peak. Thus, strong disruptive selection differential on flowering phenology was found. Display size had little effect on fruit set. A weak, but significant disruptive selection differential on display size was found. Orchid floral abundance per host tree had a significant effect on fruit set. Early in the season, T. variegata flowers with intermediate number of conspecific flowers exhibited a greater probability of setting fruit than those in host trees with fewer or more flowers. Our results show that flowering phenology may be evolutionarily unstable, possibly a consequence of the deception pollination system. Furthermore, a deception strategy would be relatively unsuccessful in populations where plants are found in either very dense or sparse patches.