Discordance between genomic divergence and phenotypic variation in a rapidly evolving avian genus (Motacilla)
Graphical abstract
Introduction
In birds, male plumage differences among closely related taxa are often believed to be the result of sexual selection, and to play an important role in reproductive isolation (Price, 2008). Plumage differences can evolve rapidly (Olsson et al., 2010, Omland and Lanyon, 2000, Milá et al., 2007), and when populations are geographically structured, may result from spatial variation in selection regimes (Price, 2008). Recent studies have demonstrated that a small number of genes can cause dramatic plumage differences despite limited genetic differentiation throughout the remainder of the genome (Poelstra et al., 2015, Toews et al., 2016, Vijay et al., 2016, Mason and Taylor, 2015). Lack of overall genetic differentiation in taxa with distinct phenotypic differences is likely due to either (1) recent divergence, with strong selection on phenotype, or (2) large-scale introgression, except on pre-existing adaptive genetic differences. In such cases, it is unlikely that phylogenetic relationships gleaned from few loci accurately reflect true species trajectories.
Genera that contain widespread species complexes are useful systems for investigating geographic variation in phenotypes because they offer comparisons between populations and species at different stages of the speciation continuum. Species complexes are often characterized by high frequencies of hybridization and poorly developed isolation barriers, despite being structured geographically (Price, 2008). Traditionally, it was thought that sympatric species would respond similarly to environmental factors influencing divergence, and therefore the study of sympatric complexes might reveal important biogeographic barriers. However, a growing body of literature suggests that species specific differences have a direct effect on demography, and spatially concordant genetic breaks should not be the expectation (Zamudio et al., 2016). For example, sexually selected traits, such as plumage, can directly affect genetic diversity via assortative mating or species recognition (Price, 1998). The complex interactions of gene flow, drift, and selection play a large role in determining the outcome of speciation and the rate at which species move along the speciation continuum. While the placement of recently diverged and introgressed lineages may be difficult, these can be contrasted with older, reproductively isolated groups within the same genus.
One bird system that is particularly well suited for such studies is the passerine genus Motacilla in the family Motacillidae. Motacilla consists of 12 species distributed throughout the Old World (Alström and Mild, 2003, del Hoyo et al., 2004) that have earned the common name wagtails due to their propensity to pump their long tails up and down. Within wagtails, there are multiple examples of taxa at different stages in the speciation process: from barely differentiated parapatric populations, to subspecies/species with distinct plumages that meet in hybrid zones, to fully reproductively isolated species (Alström and Mild, 2003). Previous phylogenetic and phylogeographic studies of Motacilla report mitochondrial relationships incongruent with both taxonomy (Pavlova et al., 2005, Voelker, 2002, Pavlova et al., 2003, Li et al., 2016, Alström and Ödeen, 2002, Ödeen and Björklund, 2003, Ödeen and Alström, 2001) and nuclear relationships (Alström and Ödeen, 2002, Ödeen and Björklund, 2003, Ödeen and Alström, 2001), with suggestions that mitochondrial DNA (mtDNA) poorly reflects the true phylogeny (Alström and Mild, 2003, Alström and Ödeen, 2002, Ödeen and Björklund, 2003, Ödeen and Alström, 2001). Several of these studies focused on aspects of wagtails’ plumage diversity, some proposing cases of remarkable parallel plumage evolution (Alström and Mild, 2003, Alström and Ödeen, 2002, Ödeen and Alström, 2001) and others implicating the role of selection in rapid plumage evolution (Pavlova et al., 2005).
Of particular interest have been the four sympatric, migratory wagtail species White Wagtail M. alba, Grey Wagtail M. cinerea, Citrine Wagtail M. citreola, and Yellow Wagtail M. flava which are widely distributed across the Palearctic during the breeding season. These species represent a striking contrast in spatial variation in male breeding plumage (cf. Fig. 1). Currently, subspecies are defined by differences in both color and pattern of head plumage in the M. flava complex (13 subspecies) and by head, back, and wing-covert plumage in M. alba (9 subspecies) (Alström and Mild, 2003). On the basis of both genetic and plumage data, many of these subspecies have been treated as separate species (reviewed in Alström and Mild, 2003). Plumage differences are thought to have evolved rapidly and in conflict with phylogeographic structure (Pavlova et al., 2005, Pavlova et al., 2003, Li et al., 2016, Ödeen and Björklund, 2003, Ödeen and Alström, 2001). In contrast, the other two Palearctic breeding species, M. cinerea and M. citreola, lack this extreme plumage variation (3 subtly different and 2 distinct subspecies, respectively (Alström and Mild, 2003).
Wagtails can be broadly categorized by breeding distribution (i.e. Palearctic, Afrotropical). Whereas only some of the Palearctic species are migratory, all of the Afrotropical species are resident (Alström and Mild, 2003, del Hoyo et al., 2004). Species with Palearctic breeding distributions can be further categorized by plumage color (i.e. “black-and-white” and yellow). Past phylogenetic reconstructions have not fully supported these groupings. M. cinerea, M. citreola, and the M. flava complex all have yellow plumage, but the latter two species have repeatedly been found to be polyphyletic (Voelker, 2002, Pavlova et al., 2003, Alström and Ödeen, 2002, Ödeen and Alström, 2001). Genetic data places the polytypic M. alba within the “black-and-white” plumage group, along with three monotypic species with rather restricted allopatric distributions in the Indian subcontinent (White-browed Wagtail M. maderaspatensis), Cambodia (Mekong Wagtail M. samveasnae), and Japan (Japanese Wagtail M. grandis) (Alström and Mild, 2003, Alström and Ödeen, 2002). The black-and-white Afrotropical African Pied Wagtail M. aguimp has also been placed within this group (Alström and Ödeen, 2002, Ödeen and Alström, 2001). The other two Afrotropical (Cape Wagtail M. capensis and Mountain Wagtail M. clara) and the Malagasy (Madagascar Wagtail M. flaviventris) species are closely related (Voelker, 2002, Alström and Ödeen, 2002, Alström et al., 2015) and have slight or no geographical variation in plumage (del Hoyo et al., 2004). A recent phylogenetic exploration of the family found the São Tomé endemic São Tomé Shorttail Amaurocichla bocagii nested within Motacilla, and proposed its inclusion within this genus (Alström et al., 2015). Overall, relationships among species are unclear and are in need of reexamination.
In this study, we utilize genome-wide SNPs, nuclear introns, and mtDNA to analyze phylogenetic relationships and divergence patterns in Motacilla, with complete species-level sampling and comprehensive coverage of the three most diverse Palearctic species. We (1) estimate the first complete time-calibrated species tree for this group; (2) use genome-wide SNPs to reconstruct the phylogeny and investigate the agreement between genotype and phenotype in the three most variable wagtail species; and (3) demonstrate conclusively that mtDNA alone is inappropriate for phylogenetic studies of Motacilla.
Section snippets
Sanger data
Throughout the manuscript, we follow the taxonomy of Alström and Mild, 2003, Alström et al., 2015.
To resolve species-level relationships and infer divergence times, we utilize previously published and unpublished sequences from (1) three nuclear introns (CHD1Z, ODC, Mb) for 42 individuals across all 12 Motacilla species (Alström and Ödeen, 2002, Ödeen and Björklund, 2003), and (2) two mitochondrial regions (ND2, CR) for 103 individuals across all species, including all subspecies of M. alba, M.
Results
We successfully constructed a 956 Mbp reference genome for M. alba which was used to assemble ddRADseq loci. A total of 219 million quality filtered reads were aligned with 5x average coverage to 73% of the zebra finch genome. Reflecting this low coverage, the N50 of the alignment was only ∼3 kbp. Reference mapping of 442.5 million ddRADseq reads resulted in 8.2 × 104 unique loci across 246 individuals with an average coverage of 29.7 (BioProject PRJNA356768). See Table S1 for individual-level
Discussion
Modern phylogeographic studies hope that the population divergence history of widespread and variable species can be disentangled with sufficient numbers of variable, neutrally evolving SNPs (Brumfield et al., 2003, McCormack et al., 2013). In the present study, genome-wide SNPs provide a robust estimate of species-level relationships, most of which are corroborated by the combined analysis of mtDNA and nuclear introns, but disagree stronger with the mtDNA gene tree. This is a major step
Data accessibility
- -
Raw, demultiplexed ddRAD reads: NBCI SRA under PRJNA356768.
- -
SNP datasets, input files for analyses, and tree files: Dryad repository doi:https://doi.org//10.5061/dryad.008bq.
Acknowledgements
We thank the museums and those who sent us samples: Ulf Johansson, Swedish Museum of Natural History; Brian Schmidt, USNM; Isao Nishiumi, NMSM; Ben Marks, FMNH; Mark Robbins, KU; Jack Withrow, UAM; Paul Sweet, AMNH; Kristof Zyskowski, YPM; Michael Westberg, Bell Museum; Gary Voelker, TCWC; Sharon Birks, UWBM. We thank Martim Melo for sharing tissues and Dai Shizuka facilitating contacts in Japan.
For helpful discussion and comments, we thank Sievert Rohwer, John Klicka, Nick Sly, Shawn
References (92)
- et al.
The utility of single nucleotide polymorphisms in inferences of population history
Trends Ecol. Evol.
(2003) - et al.
Applications of next-generation sequencing to phylogeography and phylogenetics
Mol. Phylogenet. Evol.
(2013) - et al.
The causes of mitochondrial paraphyly
Mol. Phylogenet. Evol.
(2010) - et al.
Missing inaction: the dangers of ignoring missing data
Trends Ecol. Evol.
(2008) - et al.
The Lanius excubitor (Aves, Passeriformes) conundrum—taxonomic dilemma when molecular and non-molecular data tell different stories
Mol. Phylogenet. Evol.
(2010) - et al.
Inferring genetic ancestry: opportunities, challenges, and implications
Am. J. Hum. Genet.
(2010) - et al.
Plumage genes and little else distinguish the genomes of hybridizing warblers
Curr. Biol.
(2016) - et al.
Past hybridization between two East Asian long-tailed tits (Aegithalos bonvaloti and A. fuliginosus)
Front. Zool.
(2014) - et al.
Fast model-based estimation of ancestry in unrelated individuals
Gen. Res.
(2009) - et al.
Pipits and Wagtails of Europe, Asia and North America. A&C Black
(2003)