Abstract
The perennial herb genus Hemerocallis (Asphodelaceae) shows four flowering types: diurnal half-day, diurnal one-day, nocturnal half-day, and nocturnal one-day flowering. These flowering types are corresponding to their main pollinators, and probably act as a primary mechanism of reproductive isolation. To examine how the four flowering types diverged, we reconstructed the phylogeny of the Japanese species of Hemerocallis using 1615 loci of nuclear genome-wide SNPs and 2078 bp sequences of four cpDNA regions. We also examined interspecific gene flows among taxa by an Isolation-with-Migration model and a population structure analysis. Our study revealed an inconsistency between chloroplast and nuclear genome phylogenies, which may have resulted from chloroplast capture. Each of the following five clusters is monophyletic and clearly separated on the nuclear genome-wide phylogenetic tree: (I) two nocturnal flowering species with lemon-yellow flowers, H. citrina (half-day flowering) and H. lilioasphodelus (one-day flowering); (II) a diurnal one-day flowering species with yellow-orange flowers, H. middendorffii; (III) a variety of a diurnal half-day flowering species with reddish orange flowers, H. fulva var. disticha; (IV) another variety of a diurnal half-day flowering species with reddish orange flowers, H. fulva var. aurantiaca, and a diurnal one-day flowering species with yellow-orange flowers, H. major; (V) a diurnal half-day flowering species with yellow-orange flowers, H. hakuunensis. The five clusters are consistent with traditional phenotype-based taxonomy (cluster I, cluster II, and clusters III-V correspond to Hemerocallis sect. Hemerocallis, Capitatae, and Fulvae, respectively). These findings could indicate that three flowering types (nocturnal flowering, diurnal one-day flowering, and diurnal half-day flowering) diverged in early evolutionary stages of Hemerocallis and subsequently a change from diurnal half-day flowering to diurnal one-day flowering occurred in a lineage of H. major. While genetic differentiation among the five clusters was well maintained, significant gene flow was detected between most pairs of taxa, suggesting that repeated hybridization played a role in the evolution of those taxa.
Highlights
Since the study of Darwin (1862), a wealth of evidence has been accumulated that adaptation to different pollinator groups promotes the divergence of floral traits and leads to speciation
Nucleotide diversity within a taxon was the highest in H. middendorffii var. esculenta (0.00214) and the lowest in H. major (0.00134) (Table S3, but the latter was sampled from only one population)
The nuclear genome phylogeny is more reliable than the chloroplast phylogeny because the population structure based on the neutral loci showed that genetic differentiation was well maintained among the five major clusters significant interspecific gene flow was detected between most pairs of taxa
Summary
Since the study of Darwin (1862), a wealth of evidence has been accumulated that adaptation to different pollinator groups promotes the divergence of floral traits and leads to speciation (reviewed in Schiestl and Johnson, 2013). In the process of adaptive divergence and speciation mediated by pollinators, floral displays attracting animal pollinators co-vary with the preferences and perceptual abilities of pollinators (Dyer et al, 2012) Under this co-variation, a change in a floral display trait can cause pollinator shifts and promote premating reproductive isolation between nascent species (Bradshaw and Schemske, 2003). Interspecific hybridization often leads to adaptive introgression (SuarezGonzalez et al, 2018) and contributes to new combinations of floral traits that can promote hybrid speciation (Rieseberg, 1997; Rieseberg et al, 1995). Such introgression between sister species adapted to distinct pollinators is a complicated process. A better understanding of this process requires studies that combine, highly-resolved phylogenies, detailed observations on pollinator visits, and reliable estimation of gene flow between the species (van der Niet and Johnson, 2012)
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
