Pollination in a patchily distributed lousewort is facilitated by presence of a co-flowering plant due to enhancement of quantity and quality of pollinator visits

Background and Aims

Plants surrounded by individuals of other co-flowering species may suffer a reproductive cost from interspecific pollen transfer (IPT). However, differences in floral architecture may reduce or eliminate IPT.

Methods

A study was made of Pedicularis densispica (lousewort) and its common co-flowering species, Astragalus pastorius, to compare reproductive and pollination success of lousewort plants from pure and mixed patches. Floral architecture and pollinator behaviour on flowers of the two plants were compared along with the composition of stigmatic pollen load of the louseworts. The extent of pollen limitation of plants from pure and mixed patches was also explored through supplemental pollination with self- and outcross pollen (PLs and PLx).

Key Results

Mixed patches attracted many more nectar-searching individuals of Bombus richardsi. These bumble-bees moved frequently between flowers of the two species. However, they pollinated P. densispica with their dorsum and A. pastorius with their abdomen. This difference in handling almost completely eliminated IPT. Lousewort plants from mixed patches yielded more seeds, and seeds of higher mass and germinability, than those from pure patches. Moreover, louseworts from mixed patches had lower PLs and PLx compared with those from pure patches.

Conclusions

Differences in floral architecture induced differences in pollinator behaviour that minimized IPT, such that co-flowering plants significantly enhanced quantity and quality of pollinator visits for the lousewort plants in patchy habitat. These findings add to our understanding of the mechanisms of pollination facilitation.     

Nectarless flowers with deep corolla tubes in Pedicularis: does long pistil length provide an arena for male competition?

Pollinator‐mediated selection does not seem to have a direct influence on the evolution of a long corolla tube in a nectarless flower. We hypothesized that the long pistil length of the nectarless flower with a deep corolla tube provided an opportunity for male competition. Pedicularis siphonantha, a nectarless and partially self‐incompatible lousewort with substantial variation in corolla tube length, was used to test the hypothesis. We tested whether and how corolla tube length affected seed production per capsule and seed germination rate with different pollination treatments. Flowers were hand‐pollinated with pollen from one self donor and one outcross donor and mixed pollen grains consisting of equal amounts from the two donor types, respectively. Additionally, seeds from open‐pollinated flowers with different corolla tube lengths were collected separately for measurement of germination rate. Pollination treatment and corolla tube length significantly affected number of seeds per capsule. Moreover, a significant positive relationship between seeds per capsule and corolla tube length was found when mixed hand pollination was conducted. Seeds of self hand‐pollinated flowers had a lower germination rate than those from outcross‐pollinated flowers. Under open pollination, seeds from flowers with longer corolla tubes tended to have higher germination rate. In P. siphonantha, outcross pollen may have a higher probability of contributing to the next generation when transferred to flowers with longer corolla tubes. The pistil length, therefore, should be seen as a female choice mechanism, which provides an arena for male‐to‐male competition. © 2015 The Linnean Society of London, Botanical Journal of the Linnean Society, 2015, 179 , 526–532.     

Genetic variation in populations of the arctic perennial Pedicularis dasyantha (Scrophulariaceae), on Svalbard,Norway

Isozyme variability was examined in 13 geographically isolated populations of the endemic arctic hairy lousewort (Pedicularis dasyantha) in the Svalbard Archipelago, 80° N latitude, Norway. Of the 23 enzyme systems screened on five buffer systems 18 were interpretable. Of the 31 reliable loci, only 6-phosphogluconate dehydrogenase (6-Pgd), was polymorphic. However, no heterozygotes were detected. Frequencies for allele 1 among the populations varied from 1.00 in the north to 0.00 in the south and 0.53 in the central “overlap” region. At the species level the mean number of alleles per locus (A) was 1.03. Percentage of polymorphic loci (P) was 3%. Expected heterozygosity (H_e) was 0.016. At the population level the mean number of alleles per locus was 1.01, and 1.1% of the loci were polymorphic. H_e was 0.004. These values are low compared to endemic, widespread, selling, and outcrossed species. Flower color morphs were distinct. They varied within and among the 13 populations. The frequency of color morphs coincided with allele frequencies of 6-PGD: allele 1 was found in dark purple morphs, and allele 2 was found in light morphs. This species shows more isozyme genetic variability than the five other species reported in the genus but generally less variation than other species with limited regional distributions. Low-level genetic variation in this diploid species may be a result of colonization events coupled with genetic drift, founder effects, and strong natural selection. Additional factors include the self-compatible reproductive system and the long-lived perennial habit.