Abstract:
In this study, we examined the utility of pollen morphology for resolving questions about the evolutionary history of Billia, which is a poorly known genus of Neotropical trees. Billia has been traditionally circumscribed with two species and treated as sister to Aesculus L. However, the number of species in Billia is uncertain, because the genus exhibits abundant morphological diversity but little discontinuous variation. Therefore, Billia may be monotypic and highly polymorphic, or it may have two species with blurred boundaries due to incipient speciation and/or hybridization. Moreover, one recent molecular phylogenetic study shows Billia nested within Aesculus. Our work sought to address the following questions: (i) Are there discontinuities in the pollen of Billia that may suggest species boundaries? (ii) Does the pollen of Billia show evidence for inter-specific hybridization? (iii) Do the exine morphology and size of pollen in Billia differ from those in Aesculus? Our results from scanning electron microscopy showed that pollen exine morphology is not taxonomically informative in Billia but that there are significant differences in pollen size between red- and white-flowered individuals. Thus, our pollen data support the utility of flower color in Billia for species delimitation. Our assessments of pollen viability do not support hybridization in the genus, but cannot be used to rule it out. Finally, pollen exine morphology may lend some support to an evolutionary origin of Billia within eastern North American Aesculus. In contrast, data on pollen size suggest that Billia may belong in a topological position outside of Aesculus.
