Plant-pollinator interactions are well-known examples of mutualism, but are not free of antagonism. Antagonistic interactions and defenses or counter-defenses are expected particularly in nursery pollination. In these systems, adult insects, while pollinating, lay their eggs in flowers, and juveniles consume the seeds from one or several fruits, thereby substantially reducing plant fitness. The outcome of such interactions will depend, for the plant, on the balance between pollination versus seed predation and for the larvae on the balance between the food and shelter provided versus the costs imposed by plant defenses, e.g., through abortion of infested fruits. Here, we examine the costs and benefits to the larvae in the nurserypollination system Silene latifolia/Hadena bicruris. Using selection lines that varied in flower size (large-vs. smallflowered plants), we investigated the effects of variation in flower and fruit size and of a potential defense, fruit abortion, on larval performance. In this system, infested fruits are significantly more likely to be aborted than noninfested fruits; however, it is unclear whether fruit abortion is effective as a defense. Larger flowers gave rise to larger fruits with more seeds, and larvae that were heavier at emergence. Fruit abortion was frequently observed (ca. 40% of the infested fruits). From aborted fruits, larvae emerged earlier and were substantially lighter than larvae emerging from non-aborted fruits. The lower mass at emergence of larvae from aborted fruits indicates that abortion is a resistance mechanism. Assuming that lower larval mass implies fewer resources invested in the frugivore, these results also suggest that abortion is likely to benefit the plant as a defense mechanism, by limiting both resources invested in attacked fruits, as well as the risk of secondary attack. This suggests that selective fruit abortion may contribute to the stability of mutualism also in this non-obligate system.

In field populations of S. latifolia, fruits are often multiply-sired. Because later-arriving pollen had decreased chances of fertilizing the ovules, this implies that open-pollinated flowers often benefit from pollen carry-over or pollinator visits within short time intervals, which may contribute to increase offspring genetic diversity and fitness.

Interactions, antagonistic or mutualistic, can exert selection on plant traits. We explored the role of Hadena bicruris, a pollinating seed predator, as a selective agent on its host, the dioecious plant Silene latifolia. We exposed females from artificial‐selection lines (many, small flowers (SF) vs. few, large flowers (LF)) to this moth. Infestation did not differ significantly between lines, but the odds of attacked fruits aborting were higher in SF females. We partitioned selection between that caused by moth attack and that resulting from all other factors. In both lines, selection via moth attack for fewer, smaller flowers contrasted with selection via other factors for more flowers. In LF females, selection via the two components was strongest and selection via moth attack also favoured increased fruit abortion. This suggests that the moths act as more of a selective force on flower size and number via their predating than their pollinating role.

Flowering plants rely on vectors for pollen transfer, and cannot choose their mates. Although recipient plants are unable to choose which pollen they receive, post-pollination selection (acting pre-or post-zygotically) may modify the outcome of pollination. Here we show that genetic variation among pollen recipients can predict the outcome of pollen competition (seed paternity) in the dioecious white campion. To investigate whether genetic variation among pollen recipients affects paternity, we applied the same pollen mixture from two males to three females, two of which full sisters and the third one chosen at random (unrelated). To control for maternal environmental effects, the plants used for these crosses were greenhouse-reared F 1 . We replicated this in two populations, for a total of 51 crosses, and genotyped a total of 772 offspring to assign paternity. If genetic variation affects paternity, we expected greater similarity of paternity success of the focal male with the sisters, compared to the unrelated female. Paternity of the focal male was significantly more repeatable over sisters, compared to repeatability over the mean of sisters and the unrelated females. When populations were analyzed separately, this was significant in one of the two populations. Paternity was not significantly correlated with stigma size. This provides evidence that in at least one population, genetic variation among individual plants influences the donors' paternity success, as assessed through genetic analysis of the seedling. Since due to gravitydispersed seeds natural patches may often consist of related plants, the observed effect may contribute to variation in male reproductive success.