Generous-like flowers: nectar production in two epiphytic bromeliads and a meta-analysis of removal effects.

Animal-pollinated angiosperm plants that respond positively to nectar removal by replenishment invest energy that can entail a reproductive cost. We investigated whether or not nectar removal stimulates replenishment in two hummingbird-pollinated bromeliad species. Nectar replenishment rates were also assessed by removing nectar from manually pollinated flowers because pollination events might be used as signals to save energy by preventing allocation to post-pollination nectar production. Then we synthesized the current understanding of nectar removal effects by reviewing existing published studies with a meta-analysis. The magnitude and significance of estimated nectar removal effects and factors associated with variation in size and direction of nectar removal effects were elucidated with the meta-analysis. We found that both Tillandsia species strongly respond to repeated nectar removal by producing >3 times additional nectar. Nectar secretion patterns were not altered by pollination (stigmatic pollen deposition) and we found no evidence of nectar reabsorption. Although the effect size varied widely across systems and/or environmental conditions, the meta-analysis showed that nectar removal had overall a positive effect on nectar replenishment (mainly among species inhabiting wet tropical habitats such as Tillandsia), and a negative effect on the secretion of additional sugar, suggesting that those plants are resource limited and conservative in the secretion of additional sugar.