Abstract
Background: Pollen stainability is often used in hybridisation studies as a proxy for true pollen viability. Its value for assessing fertility and ploidy level was assessed in British Sorbus L.
Aims: To assess variation in pollen stainability within and between Sorbus taxa, and its relationship to ploidy level.
Methods: Stainability of pollen was examined in herbarium specimens using Alexander's Stain.
Results: The diploid species (S. aria, S. aucuparia, S. torminalis) have the highest pollen stainability, followed by tetraploids (S. croceocarpa, S. devoniensis, S. eminens, S. intermedia, S. lancastriensis, S. latifolia, S. leptophylla, S. pseudofennica, S. rupicola). The diploid hybrids (S. × thuringiaca, S. × vagensis) have lower stainability than their parents. Triploids (S. arranensis, S. bristoliensis, S. decipiens, S. leyana, S. minima, S. pseudomeinichii, S. whiteana, S. wilmottiana) usually have the lowest stainability.
Conclusions: There is significant variation between taxa within each group, so stainability cannot be used as a proxy for ploidy level. The stainability also varies according to the parental combinations involved; hybrids with S. aucuparia have significantly lower stainability than other combinations. Assuming pollen stainability reflects pollen viability, there is potential for pollen grains of nearly all taxa to pass on genetic information to future generations through pollination of sexual taxa.
Acknowledgements
I would like to thank Andy Karran, Alex Patto and Joseph Rogers for their voluntary work counting pollen grains.