Zygomorphic flowers last longer: the evolution of floral symmetry and floral longevity.

Floral longevity, the length of time a flower remains open and functional, is a phylogenetically conserved trait that balances floral costs against the rate at which flowers are pollinated. Floral symmetry has long been considered a key trait in floral evolution. Although zygomorphic (bilaterally symmetric) flowers typically receive fewer floral visitors than actinomorphic (radially symmetric) flowers, it is yet to be determined whether this could be associated with longer floral longevity. Using newly collected field data combined with data from the literature on 1452 species in 168 families, we assess whether floral longevity covaries with floral symmetry in a phylogenetic framework. We find that zygomorphic flowers last on average 1.1 days longer than actinomorphic flowers, a 26.5% increase in longevity, with considerable variation across both groups. Our results provide a basis to discuss the ecological and evolutionary costs of zygomorphy for plants. Despite these costs, zygomorphy has evolved numerous times throughout angiosperm history, and we discuss which rewards may outweigh the costs of slower pollination in zygomorphic flowers.

Phylogeny of the bee genus Halictus (Hymenoptera: halictidae) based on parsimony and likelihood analyses of nuclear EF-1alpha sequence data.

We investigated higher-level phylogenetic relationships within the genus Halictus based on parsimony and maximum likelihood (ML) analysis of elongation factor-1alpha DNA sequence data. Our data set includes 41 OTUs representing 35 species of halictine bees from a diverse sample of outgroup genera and from the three widely recognized subgenera of Halictus (Halictus s.s., Seladonia, and Vestitohalictus). We analyzed 1513 total aligned nucleotide sites spanning three exons and two introns. Equal-weights parsimony analysis of the overall data set yielded 144 equally parsimonious trees. Major conclusions supported in this analysis (and in all subsequent analyses) included the following: (1) Thrincohalictus is the sister group to Halictus s.l., (2) Halictus s.l. is monophyletic, (3) Vestitohalictus renders Seladonia paraphyletic but together Seladonia + Vestitohalictus is monophyletic, (4) Michener's Groups 1 and 3 are monophyletic, and (5) Michener's Group 1 renders Group 2 paraphyletic. In order to resolve basal relationships within Halictus we applied various weighting schemes under parsimony (successive approximations character weighting and implied weights) and employed ML under 17 models of sequence evolution. Weighted parsimony yielded conflicting results but, in general, supported the hypothesis that Seladonia + Vestitohalictus is sister to Michener's Group 3 and renders Halictus s.s. paraphyletic. ML analyses using the GTR model with site-specific rates supported an alternative hypothesis: Seladonia + Vestitohalictus is sister to Halictus s.s. We mapped social behavior onto trees obtained under ML and parsimony in order to reconstruct the likely historical pattern of social evolution. Our results are unambiguous: the ancestral state for the genus Halictus is eusociality. Reversal to solitary behavior has occurred at least four times among the species included in our analysis.