The role of biotic interactions in shaping plant flowering phenology has long been controversial; plastic responses to the abiotic environment, limited precision of biological clocks and inconsistency of selection pressures have generally been emphasized to explain phenological variation. However, part of this variation is heritable and selection analyses show that biotic interactions can modulate selection on flowering phenology. Our review of the literature indicates that pollinators tend to favour peak or earlier flowering, whereas pre-dispersal seed predators tend to favour off-peak or later flowering. However, effects strongly vary among study systems. To understand such variation, future studies should address the impact of mutualist and antagonist dispersal ability, ecological specialization, and habitat and plant population characteristics. Here, we outline future directions to study how such interactions shape flowering phenology. IntroductionFor plant reproduction, timing is everything. An individual plant that flowers too early, before it has had time to accumulate sufficient material resources, will have a limited capacity for seed production. One that delays flowering might gain higher capacity, but might also run out of time to use it before the end of the season. Flowering phenology is affected by many environmental factors, among which temperature and photoperiod, which are reliable signals of seasons, are probably the best studied. Accurate detection of such environmental cues and the resulting plastic response of plants enable flowering to occur when climatic conditions are most suitable for reproduction. Thus, resources and conditions impose bottom-up selective forces on phenology.By contrast, top-down forces act on reproductive timing, particularly those imposed by mutualists (pollinators and seed dispersers) and antagonists (floral pathogens and predispersal seed predators). Here, we review recent progress in understanding some of the top-down selective forces that act on reproductive timing. We highlight what is known,

The orchid Dactylorhiza sambucina shows a stable and dramatic flower-color polymorphism, with both yellow-and purple-flowered individuals present in natural populations throughout the range of the species in Europe. The evolutionary significance of flower-color polymorphisms found in many rewardless orchid species has been discussed at length, but the mechanisms responsible for their maintenance remain unclear. Laboratory experiments have suggested that behavioral responses by pollinators to lack of reward availability might result in a reproductive advantage for rare-color morphs. Consequently, we performed an experiment varying the relative frequency of the two color morphs of D. sambucina to test whether rare morph advantage acted in the natural habitat of the species. We show here clear evidence from this manipulative experiment that rare-color morphs have reproductive advantage through male and female components. This is the first demonstration, to our knowledge, that negative frequency-dependent selection through pollinator preference for rare morphs can cause the maintenance of a flower-color polymorphism.

More than one-third of orchid species do not provide their pollinators with either pollen or nectar rewards. Floral mimicry could explain the maintenance of these rewardless orchid species, but most rewardless orchids do not appear to have a rewarding plant that they mimic specifically. We tested the hypothesis that floral mimicry can occur through similarity based on corolla colour alone, using naive bumble-bees foraging on arrays of plants with one rewarding model species, and one rewardless putative mimic species (Dactylorhiza sambucina) which had two colour morphs. We found that when bees were inexperienced, they visited both rewardless morphs randomly. However, after bees had gained experience with the rewarding model, and it was removed from the experiment, bees resampled preferentially the rewardless morph most similar to it in corolla colour. This is the first clear evidence, to our knowledge, that pollinators could select for floral mimicry. We suggest that floral mimicry can be a selective force acting on rewardless orchids, but only under some ecological conditions. In particular, we argue that selection on early-flowering rewardless orchids that receive visits from a large pool of naive pollinators will be weakly influenced by mimicry.

The Orchidaceae characteristically contain a very large number of species that attract pollinators but do not offer them any form of reward in return for visitation. Such a strategy is highly unusual in the plant kingdom. We conducted experiments in order to manipulate the reward strategy of the rewardless bumble-bee-pollinated orchid Barlia robertiana by adding sucrose solution to inflorescences. We found that supplementation decreased the probability of a pollinator removing pollinia by approximately ten times. Despite pollinators visiting many more flowers per inflorescence on supplemented plants, eight times fewer pollinia were removed from supplemented inflorescences during each visit. Pollinia deposition patterns were not significantly affected by supplementation and no geitonogamous deposition was recorded. In populations where inflorescences were supplemented for 20 days, pollinia removal was reduced by over half for supplemented inflorescences, whereas fruit set was unmodified by supplementation. We conclude that rewardlessness would increase total seed paternity, but not change either total seed maternity or the probability that offspring were outcrossed in this species. To the authors' knowledge this is the first time that there has been an unequivocal experimental demonstration of an evolutionary advantage for rewardlessness in the Orchidaceae.

The evolution of plants that provide no form of reward for their pollinators is puzzling because they receive low numbers of pollinator visits and so have low reproductive success. To predict the evolutionary dynamics of empty morphs within a plant population, we modeled different foraging strategies that pollinators could use to avoid them. We predicted that the optimal strategy was to visit empty inflorescences randomly when these were infrequent but to use strategies such as visiting fewer flowers per inflorescence to avoid wasting time on them. As the frequencies of empty inflorescences increased, discriminating directly against empty morphs was more likely to be an optimal strategy than was avoiding the species altogether and switching to an alternative one. An experimental test of this model using artificial inflorescences showed that bumblebees used a variety of strategies to minimize time wasted on empty inflorescences. They showed weak discrimination against empty inflorescences but switched to an alternative type of inflorescence as the frequency of empty inflorescences increased. We predicted that empty morphs would be at a visitation rate disadvantage even when at low frequencies in a plant population. Differences in outcrossing rates, or male function, may explain how rewardlessness spreads in a plant population.