The evolutionary divergence of island populations, and in particular the tempo and relative importance of neutral and selective factors, is of central interest to the study of speciation. The rate of phenotypic evolution upon island colonization can vary greatly among taxa, and cases of convergent evolution can further confound the inference of correct evolutionary histories. Given the potential lability of phenotypic characters, molecular dating of insular lineages analyzed in a phylogenetic framework provides a critical tool to test hypotheses of phenotypic divergence since colonization. The Guadalupe junco is the only insular form of the polymorphic dark-eyed junco (Junco hyemalis), and shares eye and plumage color with continental morphs, yet presents an enlarged bill and reduced body size. Here we use variation in mtDNA sequence, morphological traits and song variables to test whether the Guadalupe junco evolved rapidly following a recent colonization by a mainland form of the dark-eyed junco, or instead represents a well-differentiated “cryptic” lineage adapted to the insular environment through long-term isolation, with plumage coloration a result of evolutionary convergence. We found high mtDNA divergence of the island lineage with respect to both continental J. hyemalis and J. phaeonotus, representing a history of isolation of about 600,000 years. The island lineage was also significantly differentiated in morphological and male song variables. Moreover, and contrary to predictions regarding diversity loss on small oceanic islands, we document relatively high levels of both haplotypic and song-unit diversity on Guadalupe Island despite long-term isolation in a very small geographic area. In contrast to prevailing taxonomy, the Guadalupe junco is an old, well-differentiated evolutionary lineage, whose similarity to mainland juncos in plumage and eye color is due to evolutionary convergence. Our findings confirm the role of remote islands in driving divergence and speciation, but also their potential role as repositories of ancestral diversity.
Citation: Aleixandre P, Hernández Montoya J, Milá B (2013) Speciation on Oceanic Islands: Rapid Adaptive Divergence vs. Cryptic Speciation in a Guadalupe Island Songbird (Aves: Junco). PLoS ONE 8(5): e63242. doi:10.1371/journal.pone.0063242
Editor: Bengt Hansson, Lund University, Sweden
Received: October 22, 2012; Accepted: March 30, 2013; Published: May 10, 2013
Copyright: © 2013 Aleixandre et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: This research was funded by a grant from the Spanish Ministry of Science and Innovation (CGL2011-25866) to BM. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
As naturally isolated systems, oceanic islands have long served as unique experimental settings for evolutionary biologists –. Insular radiations such as, Anolis lizards in the Antilles, Drosophila flies or drepanidid honeycreepers in Hawaii, and Darwin’s finches in the Galapagos Islands, have contributed considerably to our understanding of speciation processes –. However, the rates of evolutionary divergence upon colonization of oceanic islands, as well as the relative roles of drift and selection in driving this divergence are still poorly understood , . Natural selection can quickly modify morphological traits following colonization of a new island or isolated habitat type as has been shown in mammals  and birds , , yet the rate and magnitude of change can also be slow and depend strongly on the taxonomic group , . In turn, some phenotypic traits can rapidly change in some cases , , yet can also show limited differentiation over millions of years . Moreover, due to the availability of unoccupied ecological niches in the early stages of island colonization, natural selection can give rise to convergent evolution during diversification, as demonstrated by various dramatic examples in groups as disparate as insects and birds , . Due to this variability in patterns and rates of phenotypic divergence, inferring the evolutionary history of insular taxa requires a phylogenetic framework that establishes affinities with mainland relatives and an estimate of time since island colonization, as can be obtained by applying a molecular clock to divergence values from neutral molecular markers , , , .
The insular form of the North American dark-eyed junco (Junco hyemalis) on Guadalupe Island provides a suitable system in which to investigate the relative roles of isolation and selection in the process of evolutionary divergence. The dark-eyed junco is a common and widespread songbird species in North America, and is composed of several geographically structured color morphs ,  (Fig. 1A). One of these is the island junco Junco h. insularis, restricted to the island of Guadalupe (Mexico), 257 km offshore from Baja California (Fig. 1B). This insular population shows a pattern of plumage coloration that is similar to that of the pink-sided junco found in the Rocky Mountains (J. h. mearnsi) (Fig. 2), yet it differs from it morphologically, showing a larger bill and smaller body size , and a very different song to that of mainland juncos .
A. Distribution map of the genus Junco showing schematics of main plumage morphs. Lineage abbreviations, clockwise from south, are the following: VOJU (volcano junco J. vulcani), YEJU (yellow-eyed junco J. phaeonotus), and the following dark-eyed junco (J. hyemalis) morphs: GUJU (Guadalupe junco J. h. insularis), GHJU (gray-headed junco J. h. caniceps), ORJU (Oregon junco J. h. oreganus), SCJU (slate-colored junco J. h. hyemalis) and PSJU (pink-sided junco J. h. mearnsi). Sampling localities for each morph are marked with colored dots. B. Relative position of Guadalupe island with respect to the California shoreline. Juncos on the island are restricted to patches of cypress forest on the northern central plateau (dark green) and the few pine areas along the northern ridge (brown). Main junco distribution ranges are delimited by light green areas.
A. Oregon junco (J. h. oreganus), B. gray-headed junco (J. h. caniceps), C. pink-sided junco (J. h. mearnsi) D. yellow-eyed junco (J. phaeonotus) and E. Guadalupe junco (J. h. insularis). Note the plumage similarity between C and E.
Continental forms of the dark-eyed junco in the United States and Canada diversified recently as the yellow-eyed junco (Junco phaeonotus) of mainland Mexico expanded northward following the last glacial maximum . The Guadalupe junco has a dark iris and plumage coloration similar to some other dark-eyed forms, and thus it is widely thought to be part of this recent radiation , , . If the Guadalupe junco represents a recent colonization of the island by a continental dark-eyed junco, then morphological changes in bill shape and body size would have taken place in just a few thousand years. The rapid differentiation in morphological traits shown by the recently established dark-eyed junco population in San Diego  suggests a capability in the group for rapid adaptation, and an increase in bill size is expected on island populations as it increases potential access to a broader range of food resources .
Here we use divergence time estimates obtained from mitochondrial DNA (mtDNA) sequence variation to test whether the Guadalupe junco evolved recently following a colonization by mainland dark-eyed juncos (as iris and plumage color suggests), or instead represents a “cryptic”, well-differentiated lineage adapted to the insular environment through long-term isolation, with plumage coloration being a result of evolutionary convergence. To do so, we use mtDNA sequence, morphological measurements and song variables to compare the Guadalupe population (insularis) to several continental populations of the dark-eyed junco (including the Oregon junco J. h. oreganus, pink-sided junco J. h. mearnsi and gray-headed junco J. h. caniceps), as well as the yellow-eyed junco (Junco phaeonotus) from mainland Mexico (Fig. 2).
Genetic Divergence and Diversity
Sequencing of two mtDNA markers in 87 individuals (Table 1) produced 9 haplotypes for the control region (CR), and 7 haplotypes for the cytochrome c oxidase I (COI) gene, resulting in 14 haplotypes for the concatenated sequence (989 bp). Phylogenetic analysis of the concatenated mtDNA haplotypes revealed the presence of two highly divergent genetic lineages: one corresponding to insularis individuals from Guadalupe Island, and one that included all continental dark-eyed and yellow-eyed juncos (Fig. 3A). One individual from Guadalupe Island (showing a typical Guadalupe phenotype) was found to carry mainland COI haplotype “A” (Fig. 3B), suggesting it descends from a female that reached the island relatively recently, yet long enough ago to show some mtDNA divergence, as its control region haplotype (“U”, Fig. 3C) does cluster with the remaining Guadalupe CR haplotypes. The mainland lineage showed a “star-like” pattern consisting of a high-frequency haplotype (“AA”) surrounded by various lower frequency haplotypes, a pattern characteristic of a recently expanded population as previously reported for these populations , and one that is not apparent in Guadalupe Island.
Shown are median-joining networks for the concatenated CR and COI sequences (A), COI (B) and CR (C). Haplotypes are represented by circles, their size proportional to their frequency in the population. Color patterns are as follows: orange – Junco phaeonotus, blue tones – continental Junco hyemalis (dark: ORJU, light: PSJU, purple: GHJU), green – Junco h. insularis. Each branch represents a single nucleotide change, with additional mutations indicated by bars along branches. Numbers in squares indicate nucleotide changes separating the volcano junco.
Genetic divergence based on the concatenated dataset was greater between insularis and either continental lineage (1.20% between insularis and continental hyemalis, and 1.22% between insularis and phaeonotus) than between both continental lineages (0.07%) (Table 2).
Mean number of substitutions for CR (349 bp) are shown below the diagonal and for COI (640 bp) above the diagonal. Lineage abbreviations are as follows: continental dark-eyed junco (DEJU), yellow-eyed junco (YEJU) and Guadalupe junco (GUJU).The marked differentiation of insularis with respect to continental lineages is visible in both mtDNA markers. However, the number of nucleotide changes in the CR sequences is lower compared to those in COI, reflecting a lower mutation rate or, more likely, high rates of homoplasy in the control region, as documented in previous studies , .
Applying a standard clock calibration of 2.1% divergence per million years (Ma) for coding mtDNA genes in passerines  to our COI sequences, we estimated the time to the last common ancestor between insularis and the continental hyemalis forms at 0.58 Ma ago, within the middle Pleistocene. Rate correction using the curve for avian mitochondrial protein coding genes in Ho et al.  places the divergence time around 0.43 Ma. Isolation-with-migration results from the MCMC runs on IMa2, with a 0.01 mutation rate per site per lineage per Ma for COI sequence data, estimated a divergence time between the continental and island lineages of 0.60 Ma, and considered a migration rate (m) from the continent to the island of 1.6×10−6 individuals per year, with no gene flow detected from the island to the continent. Analysis with BEAST 1.7.2 revealed homogeneous substitution rates across branches, and clock rate estimation with a fixed clock model set the clock rate to a mean of 0.023 divergence per site per Ma (SE = 0.0014). Since the distribution of the parameter was not normal, a better estimator of divergence time for the most recent common ancestor between Guadalupe and continental haplotypes is the median time, which was 0.57 Ma (with a mean time of 0.77 Ma, SE = 0.0072).
Genetic diversity was higher in Guadalupe Island than in the mainland populations, even when the recent immigrant to the island was excluded from analysis (Table 3). Nucleotide diversity and haplotype diversity indices on the island were an order of magnitude larger than in the continental populations for the COI sequences (Table 3). The values for the control region show less contrast between island and continent, partly because of the high frequency of control region haplotype T on the island, likely due to homoplasy as higher mutation rates in this region are likely to give rise to the same changes independently in different lineages. Fs values are negative and significant for continental dark-eyed juncos for the COI region, consistent with a population expansion as previously reported by Milá et al. . The insularis lineage, however, does not show a significant pattern of population growth, and Fs values are consistent with long-term demographic stability.
The discriminant function analysis of morphological traits shows a clear separation between all three forms (insularis, hyemalis and phaeonotus), and reflects the overall small size and large bill of insularis (Fig. 4). The first discriminant function represents overall larger birds with smaller bills, and the second function corresponds to larger birds with smaller wings (Table 4).
J. h. insularis (green squares), J. h. oreganus (blue squares), J. h. mearnsi (blue circles), J. h. caniceps (purple triangles) and J. phaeonotus (yellow circles). Empty circles represent group centroids. Correlation values between discriminant functions and transformed morphological variables are shown in table 4.
Morphological analysis showed marked divergence in trait mean values between insularis and continental forms (Fig. 5). Insularis individuals showed significantly smaller mean values for weight, wing length and tail length (weight: F2,104 = 85.3, P<0.001; wing: F2,106 = 242.7, P<0.001; tail: F2,102 = 155.3, P<0.001), yet had larger bills than both mainland forms (bill culmen: F2,108 = 146.6, P<0.001; exposed bill culmen: F2,108 = 82.5, P<0.001; bill width: F2,108 = 18.0, P<0.001 and bill depth F2,108 = 56.9, P<0.001), and an intermediate tarsus length (F1,74 = 16.9, P<0.001 insularis vs. continental hyemalis). Members of the phaeonotus lineage also tend to be larger and heavier than their northern relatives.
Bars correspond to 95% confidence intervals. Abbreviations correspond to Guadalupe junco (GUJU), continental dark-eyed juncos (DEJU) and yellow-eyed junco (YEJU). Units of measurement for weight are grams, and wing, tail, tarsus and bill variables are in mm.
Differences in song structure and note diversity are apparent between the three lineages, with the song of continental hyemalis (represented here by the Oregon junco J. h. oreganus) consisting of a simple trill, whereas the songs of phaeonotus and insularis contain a greater diversity of syllables (Fig. 6). A discriminant function analysis of song variables clearly separates the three lineages along the axis of the first discriminant function (Fig. 7), which correlates with increasing frequency range and number of different syllables per song (Table 5). Diversity in phaeonotus songs is intermediate between insularis and oreganus song types, showing less variability and more repetition of syllables than the insularis song (Fig. 6). There are significant differences between insularis and mainland forms (hyemalis+phaeonotus) in number of different syllables (F2,29 = 130.5, P<0.001), maximum frequency (F2,29 = 16.3, P<0.001), minimum frequency (F2,29 = 32.4, P<0.001) and number of repeated elements (F2,29 = 25.7, P<0.001).
A. J. h. oreganus, B. J. phaeonotus and C. J. h. insularis. Amplitude (y-axis) is given in KHz, and time (x-axis) is in seconds.
J. h. insularis (green squares), J. h. oreganus (blue triangles) and J. phaeonotus (orange circles). Empty circles represent group centroids. See Table 5 for correlation values between discriminant functions and song variables
Genetic Divergence and Phenotypic Convergence on Oceanic Islands
Our results reveal marked mtDNA divergence between insularis and continental junco lineages, supporting the hypothesis of a long history of isolation for the Guadalupe junco. Despite plumage and eye color similarities with mainland dark-eyed juncos, Guadalupe juncos represent a well-differentiated lineage that split from continental forms about 0.5 million years ago, well before the start of the postglacial expansion and radiation that gave rise to the continental dark-eyed juncos in the Holocene . Thus the prevailing hypothesis postulating that the Guadalupe junco represents a recent colonization of a continental hyemalis dark-eyed form ,  can be rejected, in favor of the alternative hypothesis for a case of convergence in plumage color.
The study of insular organisms with molecular tools has revealed numerous cases of evolutionary convergence in many different taxonomic groups, from insects  to birds , and not just in oceanic islands but also in similarly isolated continental environments organisms, such as cichlid fishes in isolated freshwater African lakes . Convergence often reveals the action of similar selective factors acting in a similar way to modify phenotypic traits in unrelated lineages, providing insight into the specific factors that are important in the early stages of divergence in insular organisms. The convergent plumage color found in the Guadalupe junco and the continental pink-sided junco, consisting of similarly patterned patches of gray, brown and rufous areas, is unlikely to be the result of environment-driven natural selection, given for instance the marked differences in the habitats where each taxon lives (arid cypress forest and scrub in Guadalupe vs. high-elevation conifer forest and alpine meadows in the Rocky Mountains, respectively). However, it could be due to sexual selection acting on plumage color traits through female choice. Plumage pattern may have appeared in the population from standing variation or by a stochastic mutation-order process  in which a de-novo mutation could give rise to a similar plumage by chance, and be subsequently driven to fixation through sexual selection. Analysis of plumage patterns across all Junco taxa in a phylogenetic context will be necessary to better understand plumage evolution in the group.
Genetic Diversity on Guadalupe Island
The relatively high amount of haplotypic and nucleotide diversity within the insularis lineage, together with the non-significant Fs values, reveal a surprising absence of population bottlenecks in the history of the lineage, despite the necessarily low population sizes juncos there must have experienced. At least during the last 200 years, a feral goat infestation severely reduced the area of suitable junco habitat from at least 780 ha  to currently less than 210 ha. Although this anthropogenic reduction in population size could be too recent for diversity to have been lost by drift, our result is inconsistent with the expected diversity loss often detected in small oceanic islands , –, and indicates that historical effective population sizes (Ne) have been sufficiently large to maintain variability, at least until recently. Other studies reporting relatively high genetic diversity in insular bird populations correspond to natural colonizations by large flocks or management-driven reintroductions in historical times , .
Islands can represent refugia for a species’ genetic diversity in cases where habitat destruction or genetic introgression by invasive competitors has affected original diversity in mainland parts of its range , . In the case of the Guadalupe junco, this insular population constitutes a repository of ancestral genetic diversity in a species complex that has suffered a widespread loss of genetic diversity across the continent, not because of recent anthropogenic factors, but due to the impact of natural glacial cycles on the historical demography of continental populations .
The presence of haplotype “UA” in the insularis lineage indicates an event of secondary colonization by at least one mainland female, as supported by the migration rates towards the island population in the coalescence simulations. The phenotype of the individual in question is however indistinguishable from other Guadalupe juncos, suggesting that the nuclear genome has become homogenized through recombination since colonization. Arrival of migratory continental hyemalis individuals to the island is probably not uncommon, which combined with the low level of introgression detected here, suggests that premating reproductive isolating barriers might have evolved, as suggested by an interesting sighting by Bryant  in the late nineteen-hundreds, who witnessed an Oregon junco (J. h. oreganus) being attacked by Guadalupe juncos during a windy day in the pine forest on the northern tip of the island.
Phenotypic Divergence of the insularis Lineage
Our morphological analysis confirmed that insularis birds are smaller than mainland individuals, yet have significantly larger beaks. Increased beak size has been previously observed in insular passerines, and is commonly associated with a wider range of food resources resulting in part from a lesser number of competing species . In the case of Guadalupe Island, a potential additional factor driving changes in bill size and shape could be related to feeding on the large seeds within the hard cones of the endemic cypress Cupressus guadalupensis. Juncos could have relied heavily on these seeds over at least the last 200 years, a period during which the feral goat infestation severely restricted the juncos’ diet, until the extirpation of the goats just five years ago . Experimental approaches will be necessary to properly test the role of cypress seeds in driving adaptive morphological divergence in Guadalupe juncos.
The song of insularis males is also markedly different from that of mainland males. Although divergence from continental song was expected due to the population's long-term isolation, complexity of the long-range song performed by insularis males goes against the expected loss of diversity in small islands –. Similarly high levels of song diversity and complexity have been recently reported by Pieplow and Francis , for Baird's junco (Junco phaeonotus bairdi) in the Sierra de la Laguna Mountains at the southern end of the Baja California peninsula (Mexico), a small “sky island” located more than 300 km away from the nearest yellow-eyed junco population. It remains to be tested whether song types in these isolated lineages represent ancestral levels of diversity since lost in mainland populations, or alternatively, whether diversity could be the product of increased competition for resources and thus increased sexual selection in small areas with restricted resources.
In contrast to the observed divergence in morphology and song, iris and plumage color in the Guadalupe junco are very similar to that of some continental populations, which led most previous experts to consider the insular form as one of the “dark-eyed juncos” . Our genetic results reveal that plumage color in Guadalupe juncos represents a case of evolutionary convergence and that dark eye color seems to have evolved more than once in the complex. Thanks to molecular dating, the lability and often fast evolution of plumage color has been documented in several groups, including orioles , wagtails , parakeets  and warblers . Within the latter group, a recent study of insular forms of the yellow warbler (Setophaga petechia) revealed that the color pattern found on the Galapagos Archipelago has evolved at least once more independently on the mainland, rendering plumage color uninformative for reconstructing the evolutionary history of the group .
Overall, both drift and selection appear to have played a role in driving the divergence of insularis. Divergence in neutral genetic loci suggests the role of drift, yet marked divergence in morphological structures related to feeding (bill shape) and locomotion (wing and tarsi), as well as male song characteristics, strongly suggest the role of natural and sexual selection in driving divergence, as has been demonstrated in other better studied natural systems –. Further research on the adaptive value of these phenotypic traits and their effect on the fitness of individuals will be necessary to fully understand divergence mechanisms in this island system.
Taxonomic and Conservation Implications
Junco insularis was considered a separate species by most authorities for most of the twentieth century, yet it was lumped with mainland dark-eyed juncos by the American Ornithologists’ Union in 1973  and has remained a subspecific entity (Junco hyemalis insularis) until the present . Previous taxonomic assessments emphasized plumage and iris color as taxonomically informative traits, which led to erroneous interpretations: Ridgway  found affinities between the Guadalupe junco and the pink-sided junco (his annectens), Dwight  classified it as a subspecies of the pink-sided junco (J. mearnsi insularis), and Miller  placed it within the “oreganus artenkreiss”. Our genetic results reveal that neither plumage nor iris color are phylogenetically informative in this case. The Guadalupe junco is strongly divergent from mainland juncos in mtDNA sequence, morphology and song, and phylogenetic analysis reveals that it renders “dark-eyed juncos” paraphyletic. We therefore recommend that the Guadalupe junco be reinstated as a full species (Junco insularis).
Juncos on Guadalupe Island are critically endangered. Restricted to small patches of cypress (Cupressus guadalupensis) and pine (Pinus contorta) in the north end of the island, we estimate the total area of suitable junco habitat at just 170 to 210 ha. The cypress groove area was estimated at about 780 ha in expedition reports from the early to mid twentieth century ,  and described already as an “old forest” due to the lack of regeneration caused by the browsing of the introduced goat population. By the time Jehl and Everett  visited the island in the early1980’s, the forest had been reduced to a narrow grove of cypresses, increasingly decimated by high mortality due to the old age of most trees. The timely eradication of feral goats in 2007, has since allowed the growth of seedlings and a promising though incipient forest regeneration. However, a forest fire in 2008 reduced prime cypress habitat to less than half, and feral cats and mice still plague the island, so that long-term survival of juncos is anything but guaranteed. Targeted conservation efforts are urgently needed to preserve the unique biodiversity of Guadalupe Island, including the unique Guadalupe junco and its habitat.
In contrast to prevailing taxonomy, genetic data reveal the Guadalupe junco as an old, well-differentiated lineage, whose similarity to mainland juncos in plumage and eye color is due to evolutionary convergence. Congruent patterns of marked genetic, morphological, and bioacoustic divergence support the recognition of the Guadalupe junco as a well-differentiated evolutionary lineage, demonstrating the limited value of plumage and eye color for inferring evolutionary history and species limits in this genus. Our findings confirm the role of remote islands in driving divergence and speciation, but also their potential role as repositories of ancestral diversity as shown by the richness of the genetic pool and song structure of the Guadalupe junco. Our results also underscore the urgent need to preserve this critically endangered insular bird.
Junco populations were sampled using mist nets and song recording playbacks to attract individuals to the net when needed. Each individual captured was aged, sexed, and marked with a numbered aluminum band. Morphological measurements were taken (see below), and a blood sample was collected by venipuncture of the sub-brachial vein and stored in lysis buffer at −20°C in the laboratory. After processing, birds were released unharmed at the site of capture. All sampling activities were conducted in compliance with UCLA’s Animal Care and Use Program regulations, and with state and federal scientific collecting permits in the USA, and a special research permit to sample on Guadalupe Island issued by the Secretaria de Medio Ambiente y Recursos Naturales of Mexico (Permit No. SGPA/DGVS/01217/11). Sampling localities for genetic and morphological data (with sample sizes for morphology/genetics specified in parentheses) were as follows: J. h. insularis (34/32): Isla Guadalupe, Mexico (N 29.110°, W 118.3267°); J. h. oreganus (18/22): Payette National Forest, Idaho (N 45.093°, W 116.635°), Sawtooth National Forest, Idaho (43.898°N, 114.901°W), Malibu, California (N 34.064°, W 118.612°), USA; J. h. mearnsi (13/4): Shoshone National Forest, Wyoming (N 43.982°, W 109.515°), USA; J. h. caniceps (10/5): Uinta range, Arizona (N 40.805°, W 110.877°), USA; J. phaeonotus (34/24): La Cima, D. F. (N 19.253°, W 99.336°), Mexico.
We extracted genomic DNA from blood samples using a Qiagen DNeasy kit (Qiagen™, Valencia, CA) following the manufacturer's protocol. We amplified two regions of the mtDNA: 349 base pairs (bp) of the hypervariable region I of the control region (CR) using primers LGL2 (L2263, 5′-GGCCACATCAGACAGTCCAT-3′) and H417 (H2607, 5′-AGTAGCTCGGTTCTCGTGAG-3′) ; and 640 bp of the cytochrome c oxidase I (COI) gene using primers “BirdF1” (5′-TTCTCCAACCACAAAGACATT-3′) and “BirdR1” (5′-CGTGGGAGATAATTCCAAATCCTG-3′) . Primers L16150 (5'-CCTCYAYCWCCARCTCCCAAAGC-3')  and “BirdR2” (5'-ACTACATGTGAGATGATTCCGAATCCAG-3')  were used instead of LGL2 and “BirdR1” to amplify samples from Guadalupe island. Details on amplification conditions are available from the corresponding author upon request. PCR products were purified with an ethanol precipitation and sequenced in an ABI 3730X automated sequencer. The amplification of nuclear copies of the control region is unlikely since we amplified it together with a coding fragment that was unambiguously translated into its aminoacid sequence. All sequences used in the study have been deposited in GenBank (accessions KC862058–KC862069).
Sequences were aligned and trimmed to equal lengths using Sequencher 4.1.4 (GeneCodes) and the accuracy of variable sites was checked visually on the individual chromatograms. We calculated genetic distances using the concatenation of both mtDNA sequences. Percent divergence was calculated using the number of nucleotide changes between haplotypes (p distances) and correcting for within population variation  in Arlequin 126.96.36.199 . We obtained an estimate for divergence time using a standard molecular-clock calibration for coding mtDNA in songbirds of 2.1% per Ma. . In addition, we obtained estimates of divergence time and gene flow between the island and mainland lineages using coalescence analysis of the COI sequences under an isolation-with-migration model as implemented in the program IMa2 . To account for the possibility of multiple nucleotide substitutions per site we used a Hasegawa-Kishino-Yano (HKY) model  for the calculations, since an infinite sites model does not apply to our data. Based on preliminary runs, we set prior maxima for divergence time, population size and mutation rate to 25, 20 and 3, respectively. We run ten chains of 10,000,000 steps each, sampling every 100 steps after a 50,000-step burn-in, with heating in a geometric increment model with a term 1 of 0.96 and a term 2 of 0.9. We tracked chain convergence using Tracer v1.5 . Further estimation of divergence times was conducted using BEAST 1.7.2 , in MCMC runs of 10,000,000 steps each with sampling every 1,000 steps. We used a strict molecular clock (after ruling out rate variation across branches), and modeled rate variation with a lognormal distribution with a mean of 0.021 and a standard deviation of 0.5.
Nei's  unbiased haplotype diversity (h) and nucleotide diversity (π) values within groups for all three lineages were calculated in Arlequin 3.5. To test for past sudden changes in population size we used Fu's  test of neutrality, which detects departures from neutrality in scenarios with biased frequencies of rare alleles or young mutations in non-recombinant sequences. We used Arlequin 3.5 to obtain Fu's Fs values, and large and significant negative values of the Fs statistic were interpreted as an excess of recent mutations caused by a population expansion.
To establish the relationship between the different haplotypes present in the populations we constructed haplotype networks for both regions and for the concatenated sequence using a median-joining (MJ) algorithm  in the program Network 4.6 (fluxus-engineering.com). Haplotypes of the volcano junco (Junco vulcani) from Costa Rica, the most divergent junco species (2.82% divergence in COI sequence from the continental radiation lineages and 2.50% from insularis), were included for comparison.
The following morphological measurements were obtained from each bird: unflattened wing length using a ruler to the nearest 0.5 mm; and tail length, tarsus length, bill culmen, exposed bill culmen, bill width and bill depth using a 0.1-mm precision caliper. For details on each measurement see Milá et al. . We measured mass with a digital scale of 0.1-g precision. All measurements were taken by a single observer (BM). To test the significance of the morphological differences between the island and the continental lineages we compared population means for all morphological traits using one-way ANOVA. We also conducted a discriminant function analysis to assess if the three lineages (Guadalupe juncos, mainland dark-eyed juncos, and yellow-eyed juncos) clustered into morphologically differentiated groups. To calculate discriminant functions, variables were transformed using natural logarithms, and mass was linearized using a cubic-root transformation. All analysis were conducted in SPSS 11.5 (SPSS Inc 2002).
Song recordings were obtained using a Roland Edirol R-09 stereo digital recorder and a Sennheiser ME66 shotgun microphone and stored as uncompressed wave files. Localities for the recordings were the following: J. h. insularis (n = 12): Isla Guadalupe (29.110° N, 118.3267° W), Baja California, Mexico; J. h. oreganus (n = 11): Mt. Laguna (32.864° N, 116.442° W), and University of California San Diego (32.876° N, 117.237° W), California, USA; and J. phaeonotus (n = 7): San Cristóbal de las Casas (16.733° N, 92.604° W), Chiapas, and Ajusco (19.211°N, 99.283°W), D.F., Mexico.
We analyzed sonograms from song recordings using Raven Pro 1.3 . To measure song variables, we produced a spectrogram for each individual song using a 256 Hann window with a 50% time grid overlap and a discrete Fourier transform (DFT) size of 512 samples. We used Marler and Isaac's  definition of “syllable” for junco song, a coherent unit formed by a group of two or more notes, each note being a continuous vocal utterance. Variables measured included maximum and minimum frequencies, peak frequency, song duration, number of different syllables and total number of syllables, from which we obtained a measure of repeated syllables. Mean values for all these measures were calculated for each lineage, and differences were analyzed with SPSS 11.5 by means of one-way ANOVA and discriminant function analysis.
John McCormack provided valuable comments on a previous version of the manuscript. Ellen Ketterson, Jonathan Atwell, Steve Burns, Christie Burns and Jatziri Calderón contributed to field work on Guadalupe Island and to fruitful discussions on junco evolution. We are grateful to Jonathan Atwell for donating recordings of J. h. oreganus.
Conceived and designed the experiments: BM PA. Performed the experiments: BM PA JHM. Analyzed the data: PA BM. Wrote the paper: PA BM.
- 1. Whittaker RJ, Fernández-Palacios JM (2007) Island biogeography: ecology, evolution and conservation. New York: Oxford University Press; 304p.
- 2. Grant PR (1998) Evolution on islands. Oxford, U.K.: Oxford University Press; 352p.
- 3. Coyne JA, Orr HA (2004) Speciation. Sunderland, Massachusetts: Sinauer Associates, Inc.; 545p.
- 4. Price T (2008) Speciation in birds. Greenwood Village, Colorado: Roberts and Company; 470p.
- 5. Losos JB, Ricklefs RE (2009) Adaptation and diversification on islands. Nature. 457(7231): 830–836. doi: 10.1038/nature07893
- 6. Grant PR, Grant BR (2007) How and why species multiply: The radiation of Darwin's finches. Princeton & Oxford: Princeton University Press; 256 p.
- 7. Grant PR (2001) Reconstructing the evolution of birds on islands: 100 years of research. Oikos. 92(3): 385–403. doi: 10.1034/j.1600-0706.2001.920301.x
- 8. Carson HL, Kaneshiro KY (1976) Drosophila of Hawaii - systematics and ecological genetics. Annu Rev Ecol Syst. 7: 311–345. doi: 10.1146/annurev.es.07.110176.001523
- 9. Emerson BC, Paradis E, Thebaud C (2001) Revealing the demographic histories of species using DNA sequences. Trends Ecol Evol. 16(12): 707–716. doi: 10.1016/s0169-5347(01)02305-9
- 10. Emerson BC, Gillespie RG (2008) Phylogenetic analysis of community assembly and structure over space and time. Trends Ecol Evol. 23(11): 619–630. doi: 10.1016/j.tree.2008.07.005
- 11. Millien V (2006) Morphological evolution is accelerated among island mammals. PLoS Biol. 4(10): 1863–1868. doi: 10.1371/journal.pbio.0040321
- 12. Clegg SM, Phillimore AB (2010) The influence of gene flow and drift on genetic and phenotypic divergence in two species of Zosterops in Vanuatu. Philos Trans R Soc Lond B Biol Sci. 365(1543): 1077–1092. doi: 10.1098/rstb.2009.0281
- 13. Mathys BA, Lockwood JL (2011) Contemporary morphological diversification of passerine birds introduced to the Hawaiian archipelago. Philos Trans R Soc Lond B Biol Sci. 278(1716): 2392–2400. doi: 10.1098/rspb.2010.2302
- 14. Phillimore AB, Owens IPF, Black RA, Chittock J, Burke T, et al. (2008) Complex patterns of genetic and phenotypic divergence in an island bird and the consequences for delimiting conservation units. Mol Ecol. 17(12): 2839–2853. doi: 10.1111/j.1365-294x.2008.03794.x
- 15. Lovette IJ, Bermingham E, Ricklefs RE (2002) Clade-specific morphological diversification and adaptive radiation in Hawaiian songbirds. Philos Trans R Soc Lond B Biol Sci. 269(1486): 37–42. doi: 10.1098/rspb.2001.1789
- 16. Tinghitella RM, Zuk M, Beveridge M, Simmons LW (2011) Island hopping introduces Polynesian field crickets to novel environments, genetic bottlenecks and rapid evolution. J Evol Biol. 24(6): 1199–1211. doi: 10.1111/j.1420-9101.2011.02255.x
- 17. Keogh JS, Scott IAW, Hayes C (2005) Rapid and repeated origin of insular gigantism and dwarfism in Australian tiger snakes. Evolution. 59(1): 226–233. doi: 10.1554/04-310
- 18. Buckley TR, Attanayake D, Bradler S (2009) Extreme convergence in stick insect evolution: Phylogenetic placement of the Lord Howe island tree lobster. Philos Trans R Soc Lond B Biol Sci. 276(1659): 1055–1062. doi: 10.1098/rspb.2008.1552
- 19. Fleischer RC, James HF, Olson SL (2008) Convergent evolution of Hawaiian and Australo-Pacific honeyeaters from distant songbird ancestors. Current Biology. 18(24): 1927–1931. doi: 10.1016/j.cub.2008.10.051
- 20. Illera JC, Emerson BC, Richardson DS (2007) Population history of Berthelot's pipit: Colonization, gene flow and morphological divergence in Macaronesia. Mol Ecol. 16(21): 4599–4612. doi: 10.1111/j.1365-294x.2007.03543.x
- 21. Milá B, Warren BH, Heeb P, Thebaud C (2010) The geographic scale of diversification on islands: Genetic and morphological divergence at a very small spatial scale in the Mascarene grey white-eye (Aves: Zosterops borbonicus). BMC Evol Biol. 10.
- 22. Miller AH (1941) Speciation in the avian genus Junco. University of California publications in zoölogy. Vol. 44. Berkeley: University of California press; 434p.
- 23. Nolan V Jr, Ketterson ED, Cristol DA, Rogers CM, Clotfelter ED, et al.. (2002) Dark-eyed junco: Junco hyemalis. Philadelphia: Birds of North America Inc. 44p.
- 24. Mirsky EN (1976) Song divergence in hummingbird and junco populations on Guadalupe island. Condor. 78(2): 230–235. doi: 10.2307/1366858
- 25. Milá B, McCormack JE, Castaneda G, Wayne RK, Smith TB (2007) Recent postglacial range expansion drives the rapid diversification of a songbird lineage in the genus Junco. Proc R Soc Lond B Biol Sci. 274(1626): 2653–2660. doi: 10.1098/rspb.2007.0852
- 26. Dwight J (1918) The geographic distribution of color and of other variable characters in the genus Junco: A new aspect of specific and subspecific values. Bull. Am. Mus. Nat. Hist. 38: 269–309. doi: 10.2307/4073241
- 27. Rasner CA, Yeh P, Eggert LS, Hunt KE, Woodruff DS, et al. (2004) Genetic and morphological evolution following a founder event in the dark-eyed junco, Junco hyemalis thurberi. Mol Ecol. 13(3): 671–681. doi: 10.1046/j.1365-294x.2004.02104.x
- 28. Scott SN, Clegg SM, Blomberg SP, Kikkawa J, Owens IPF (2003) Morphological shifts in island-dwelling birds: The roles of generalist foraging and niche expansion. Evolution. 57(9): 2147–2156. doi: 10.1554/03-021
- 29. McCormack JE, Smith TB (2008) Niche expansion leads to small-scale adaptive divergence along an elevation gradient in a medium-sized passerine bird. Proc R Soc Lond B Biol Sci. 275(1647): 2155–2164. doi: 10.1098/rspb.2008.0470
- 30. Milá B, Smith TB, Wayne RK (2007) Speciation and rapid phenotypic differentiation in the yellow-rumped warbler Dendroica coronata complex. Mol Ecol. 16(1): 159–173. doi: 10.1111/j.1365-294x.2006.03119.x
- 31. Weir JT, Schluter D (2008) Calibrating the avian molecular clock. Mol Ecol. 17(10): 2321–2328. doi: 10.1111/j.1365-294x.2008.03742.x
- 32. Ho SYW, Phillips MJ, Cooper A, Drummond AJ (2005) Time dependency of molecular rate estimates and systematic overestimation of recent divergence times. Mol Biol Evol. 22(7): 1561–1568. doi: 10.1093/molbev/msi145
- 33. Ruber L, Adams DC (2001) Evolutionary convergence of body shape and trophic morphology in cichlids from lake Tanganyika. J Evol Biol. 14(2): 325–332. doi: 10.1046/j.1420-9101.2001.00269.x
- 34. Mani GS, Clarke BC (1990) Mutational order - a major stochastic-process in evolution. Philos Trans R Soc Lond B Biol Sci. 240(1297): 29–37. doi: 10.1098/rspb.1990.0025
- 35. Tayer OB, Bangs O (1908) The present state of the ornis of Guadalupe island. Condor. 10(3): 101–106. doi: 10.2307/1360977
- 36. Degnan SM (1993) Genetic-variability and population differentiation inferred from DNA-fingerprinting in silvereyes (Aves, Zosteropidae). Evolution. 47(4): 1105–1117. doi: 10.2307/2409978
- 37. Rocamora GJ, Richardson DS (2003) Genetic and morphological differentiation between remnant populations of an endangered species: The case of the Seychelles white-eye. Ibis. 145(1): E34–E44. doi: 10.1046/j.1474-919x.2003.00140.x
- 38. Boessenkool S, Taylor SS, Tepolt CK, Komdeur J, Jamieson IG (2007) Large mainland populations of South Island robins retain greater genetic diversity than offshore island refuges. Conserv Genet. 8(3): 705–714. doi: 10.1007/s10592-006-9219-5
- 39. Tzika AC, Rosa SFP, Fabiani A, Snell HL, Snell HM, et al. (2008) Population genetics of Galapagos land iguana (genus Conolophus) remnant populations. Mol Ecol. 17(23): 4943–4952. doi: 10.1111/j.1365-294x.2008.03967.x
- 40. Johnson JA, Tingay RE, Culver M, Hailer F, Clarke ML, et al. (2009) Long-term survival despite low genetic diversity in the critically endangered Madagascar fish-eagle. Mol Ecol. 18(1): 54–63. doi: 10.1111/j.1365-294x.2008.04012.x
- 41. Mason RAB, Browning TL, Eldridge MDB (2011) Reduced MHC class II diversity in island compared to mainland populations of the black-footed rock-wallaby (Petrogale lateralis lateralis). Conserv Genet. 12(1): 91–103. doi: 10.1007/s10592-009-9993-y
- 42. Frankham R (1997) Do island populations have less genetic variation than mainland populations? Heredity. 78: 311–327. doi: 10.1038/sj.hdy.6880980
- 43. Clegg SM, Degnan SM, Moritz C, Estoup A, Kikkawa J, et al. (2002) Microevolution in island forms: The roles of drift and directional selection in morphological divergence of a passerine bird. Evolution. 56(10): 2090–2099. doi: 10.1554/0014-3820(2002)056[2090:miiftr]2.0.co;2
- 44. Brekke P, Bennett PM, Santure AW, Ewen JG (2011) High genetic diversity in the remnant island population of hihi and the genetic consequences of re-introduction. Mol Ecol. 20(1): 29–45. doi: 10.1111/j.1365-294x.2010.04923.x
- 45. Sanaa A, Ben Fadhel N (2010) Genetic diversity in mainland and island populations of the endangered Pancratium maritimum L. (Amaryllidaceae) in Tunisia. Sci Hortic. 125(4): 740–747. doi: 10.1016/j.scienta.2010.05.014
- 46. Hargreaves S, Maxted N, Hirano R, Abberton M, Skøt L, et al. (2010) Islands as refugia of Trifolium repens genetic diversity. Conserv Genet. 11(4): 1317–1326. doi: 10.1007/s10592-009-9960-7
- 47. Bryant WE (1887) Additions to the ornithology of Guadalupe island, in Bulletin of the California academy of sciences. San Francisco: California Academy of Sciences. 269–318.
- 48. Randall JM, Granillo I, Shaw R, Keitt B, Junak S (2008) El monitoreo de los impactos de la erradicación de cabras y otras acciones de manejo sobre las plantas y animales de isla Guadalupe. In KS del Prado and E Peters, editors. Isla Guadalupe restauración y conservación. Mexico D. F. 231–238.
- 49. Baker MC (1996) Depauperate meme pool of vocal signals in an island population of singing honeyeaters. Anim Behav. 51: 853–858. doi: 10.1006/anbe.1996.0089
- 50. Hamao S, Ueda K (2000) Simplified song in an island population of the bush warbler Cettia diphone. J Ethol. 18(1): 53–57. doi: 10.1007/s101640070025
- 51. Baker MC, Baker MSA, Tilghman LM (2006) Differing effects of isolation on evolution of bird songs: Examples from an island-mainland comparison of three species. Biol J Linn Soc Lond. 89(2): 331–342. doi: 10.1111/j.1095-8312.2006.00677.x
- 52. Pieplow ND, Francis CD (2011) Song differences among subspecies of yellow-eyed juncos (Junco phaeonotus). Wilson J Ornithol. 123(3): 464–471. doi: 10.1676/10-126.1
- 53. Omland KE, Lanyon SM, Fritz SJ (1999) A molecular phylogeny of the new world orioles (Icterus): The importance of dense taxon sampling. Mol Phylogenet Evol. 12(2): 224–239. doi: 10.1006/mpev.1999.0611
- 54. Ödeen A, Björklund M (2003) Dynamics in the evolution of sexual traits: Losses and gains, radiation and convergence in yellow wagtails (Motacilla flava). Mol Ecol. 12(8): 2113–2130. doi: 10.1046/j.1365-294x.2003.01883.x
- 55. Ribas CC, Joseph L, Miyaki CY (2006) Molecular systematics and patterns of diversification in Pyrrhura (Psittacidae), with special reference to the picta-leucotis complex. Auk. 123(3): 660–680. doi: 10.1642/0004-8038(2006)123[660:msapod]2.0.co;2
- 56. Lovette IJ, Perez-Eman JL, Sullivan JP, Banks RC, Fiorentino I, et al. (2010) A comprehensive multilocus phylogeny for the wood-warblers and a revised classification of the Parulidae (Aves). Mol Phylogenet Evol. 57(2): 753–770. doi: 10.1016/j.ympev.2010.07.018
- 57. Chaves JA, Parker PG, Smith TB (2012) Origin and population history of a recent colonizer, the yellow warbler in Galápagos and Cocos islands. J Evol Biol. 25(3): 509–521. doi: 10.1111/j.1420-9101.2011.02447.x
- 58. Scott SN, Clegg SM, Blomberg SP, Kikkawa J, Owens IPF (2003) Morphological shifts in island-dwelling birds: The roles of generalist foraging and niche expansion. Evolution. 57(9): 2147–2156. doi: 10.1554/03-021
- 59. Grant PR (1999) Ecology and evolution of Darwin's finches. New Jersey: Princeton University Press; 519p.
- 60. Benkman CW (2003) Divergent selection drives the adaptive radiation of crossbills. Evolution. 57(5): 1176–1181. doi: 10.1111/j.0014-3820.2003.tb00326.x
- 61. Hendry AP, Grant PR, Grant BR, Ford HA, Brewer MJ, et al. (2006) Possible human impacts on adaptive radiation: Beak size bimodality in Darwin's finches. Proc R Soc Lond B Biol Sci. 273(1596): 1887–1894. doi: 10.1098/rspb.2006.3534
- 62. AOU (1973) Thirty-second supplement to the American ornithologists' union check-list of North American birds. Auk. 90(2): 411–419. doi: 10.2307/4088140
- 63. Ridgway R (1877) The birds of Guadalupe island: Discussed with reference to the present genesis of species.
- 64. Dwight J (1918) The geographic distribution of color and of other variable characters in the genus Junco: A new aspect of specific and subspecific values. Bull. Am. Mus. Nat. Hist. 38: 269–309. doi: 10.2307/4073241
- 65. Howell TR, Cade TJ (1954) The birds of Guadalupe island in 1953. Condor. 56(5): 283–294. doi: 10.2307/1364852
- 66. Jehl JR, Everett WT (1985) History and status of the avifauna of Isla Guadalupe, Mexico. Trans S Diego Soc Nat Hist. 20(17): 313–336.
- 67. Tarr CL (1995) Primers for amplification and determination of mitochondrial control-region sequences in oscine passerines. Mol Ecol. 4(4): 527–529. doi: 10.1111/j.1365-294x.1995.tb00251.x
- 68. Hebert PDN, Stoeckle MY, Zemlak TS, Francis CM (2004) Identification of birds through DNA barcodes. PLoS Biol. 2(10): 1657–1663. doi: 10.1371/journal.pbio.0020312
- 69. Sorenson MD, Ast JC, Dimcheff DE, Yuri T, Mindell DP (1999) Primers for a PCR-based approach to mitochondrial genome sequencing in birds and other vertebrates. Mol Phylogenet Evol. 12(2): 105–114. doi: 10.1006/mpev.1998.0602
- 70. Wilson AC, Cann RL, Carr SM, George M, Gyllensten UB, et al. (1985) Mitochondrial-DNA and two perspectives on evolutionary genetics. Biol J Linn Soc Lond. 26(4): 375–400. doi: 10.1111/j.1095-8312.1985.tb02048.x
- 71. Excoffier L, Laval G, Schneider S (2005) Arlequin (version 3.0): An integrated software package for population genetics data analysis. Evol Bioinformatics. 1: 47–50.
- 72. Hey J (2011) Documentation for IMa2. http://genfaculty.rutgers.edu/hey/software.
- 73. Hasegawa M, Kishino H, Yano TA (1985) Dating of the human ape splitting by a molecular clock of mitochondrial-DNA. J Mol Evol. 22(2): 160–174. doi: 10.1007/bf02101694
- 74. Rambaut A, Drummond AJ (2007) Tracer v1.4, available from http://beast.Bio.Ed.Ac.Uk/tracer.
- 75. Drummond AJ, Suchard MA, Xie D, Rambaut A (2012) Bayesian phylogenetics with BEAUti and the BEAST 1.7. Mol Biol Evol. 29(8): 1969–73. doi: 10.1093/molbev/mss075
- 76. Nei M.(1987) Molecular evolutionary genetics. New York & Guildford, England: Columbia University Press. 512p.
- 77. Fu YX (1997) Statistical tests of neutrality of mutations against population growth, hitchhiking and background selection. Genetics. 147(2): 915–925.
- 78. Bandelt HJ, Forster P, Rohl A (1999) Median-joining networks for inferring intraspecific phylogenies. Mol Biol Evol. 16(1): 37–48. doi: 10.1093/oxfordjournals.molbev.a026036
- 79. Milá B, Wayne RK, Smith TB (2008) Ecomorphology of migratory and sedentary populations of the yellow-rumped warbler (Dendroica coronata). Condor. 110(2): 335–344. doi: 10.1525/cond.2008.8396
- 80. Charif RA, Waack AM, Strickman LM (2008) Raven pro 1.3 user's manual, ed. Cornell Laboratory of Ornithology. New York: Ithaca.
- 81. Marler P, Isaac D (1961) Song variation in a population of Mexican juncos. Wilson Bull. 73: 193–206.