Floral scents attract pollinators to plant rewards; in nectarless flowers, pollen grains are the only reward. Thus, pollen not only fertilizes ovules, but also feeds pollinators. This dilemma is resolved by specialization of anthers (i.e., heteranthery): feeding anthers that feed pollinators and pollinating anthers for fertilization. We hypothesized that the chemical composition of floral volatiles differs between the two types of anther and influences pollination preference for feeding anthers. We used Solanum rostratum as a study model because its heterantherous flowers produce a floral scent that suggests a close association with their pollinators. The main aim of this study was to determine the chemical composition of the two types of anther and to investigate how they influence foraging behaviour of pollinators. To characterize this composition, we used solid phase microextraction and hexane extraction followed by gas chromatography-mass spectrometry. We registered 12 volatile compounds in S. rostratum floral extracts, mainly aromatic and sesquiterpene compounds. The proportion of these compounds differed between feeding and pollinating anthers. Some of these compounds were probably emitted by osmophores located in both anther types. Also, we used electroantennography to investigate Melipona solani antennal response to floral volatiles. The M. solani antennae are receptive to the highest floral extract dose tested. Finally, we conducted two behavioural bioassays to test bee attraction for each type of floral extract: a) multiple-choice in a feeding arena using M. solani and b) Y-olfactometer bioassay using Bombus impatiens. Both bee species preferred feeding anthers in bioassays. In conclusion, heteranthery involves chemical differentiation (i.e., proportion of volatiles compounds) in anther specialization that influences bee preference for feeding anthers over pollinating anthers.

No abstract

No abstract

Floral scents shape plant-pollinator interactions. Although, populations of the same species can vary in their floral scent, little is known about how this variation affects pollinator visitation. In this study, we compare the scents emitted by buzz-pollinated Solanum rostratum (Solanaceae) in two areas of its distribution (Mexico and USA), and investigate how these differences in scent affect pollinator preferences and attraction. We determined the variation of floral volatile compounds using hexane extraction followed by gas chromatography coupled with spectrometry. We also performed a field cage multiple-choice bioassay and a Y-tube behavioural bioassay using Bombus impatiens. We recorded 13 volatile compounds in floral extracts for plants from both ranges that varied qualitative and semi-quantitatively among populations. We found that in the field cage experiment, bumble bees visited plants from the USA populations more frequently than plants from Mexican populations. However, bees showed no difference in preference between extracts from Mexican or USA flowers. We conclude that although bees show differential visitation to whole plants of the different regions, variation in floral extract alone does not translate into differences in preference by B. impatiens. The potential effects of variation in floral scent on the other native bee pollinators remain to be assessed.

Domestication is an evolutionary process with an impact on plant reproduction. Many domesticated plants are self-compatible (i.e., they lack mechanisms to reject their own pollen), but few domesticated plants are fully or partially self-incompatible. We used the husk tomato, Physalis philadelphica, as a study model to investigate changes in the reproductive strategy of an annual partially self-incompatible plant during the process of domestication. Wild and cultivated populations of this species coexist in close proximity. These different populations present a high level of morphological and genetic variation associated with different degrees of domestication. We hypothesized that artificial selection favors self-compatibility in cultivated plants through changes in their reproductive strategy and some reproductive parameters associated with domestication. To test this hypothesis, we characterized the floral morphology and some reproductive parameters of weedy plants (wild plants), landraces (semi-domesticated plant), and commercial plants (domesticated plants). We conducted an artificial crossing experiment, germinated the seeds, and recorded seedling growth. Commercial plants had the largest flowers and the highest number of ovules. Yet, they did not differ in other reproductive parameters (e.g., herkogamy, size of pollen grains, stigmatic area, and pollen:ovule ratio) from landraces and weedy plants. Physalis philadelphica produced fruits by autonomous autogamy in the artificial crossing experiment. These fruits were the smallest and lightest fruits at all degrees of domestication; however, fruit set of autonomous autogamy was higher in weedy plants. In addition, fruit production was higher when weedy plants donated pollen to commercial plants. Although seeds produced by autonomous autogamy of weedy plants had a low germination percentage, their cotyledons and the embryonic foliage leaves appeared earlier than in landraces and commercial plants. In conclusion, the domestication syndrome in this plant was manifested as increments in flower size and ovule production. Contrary to expectations, there was higher fruit production by autonomous autogamy in weedy plants than in cultivated plants. It seems that artificial selection in P. philadelphica favors self-incompatibility in cultivated plants. Nonetheless, spontaneous self-pollination seems to be advantageous in weedy populations because they produced viable seeds from which cotyledons and the embryonic foliage leaves emerged earlier than in cultivated plants.