TY - JOUR
AU1 - Powell, Alexis F. L, A.
AU2 - Barker, F., Keith
AU3 - Lanyon, Scott, M.
AB - Abstract The grackles (Quiscalus spp.), together with their sister genus Euphagus, compose a clade within the New World blackbirds (Icteridae). We used gene sequences of cytochrome b and NADH dehydrogenase subunit 2 (ND2) to reconstruct relationships within this group. A primary concern was determining the phylogenetic position and genetic distinctiveness of the extinct Slender-billed Grackle (Q. palustris)—a poorly known endemic of the Lerma Basin and the ancient lakes of the Valley of Mexico, last collected and recorded in 1910—and of the Nicaraguan Grackle (Q. nicaraguensis), which is likewise unusual among grackles for its restricted geographic range. Our analysis differs from previous efforts by inclusion of these taxa along with all other recognized grackle species, intraspecific sampling of Greater Antillean (Q. niger), Carib (Q. lugubris), and Great-tailed (Q. mexicanus) Grackles, and inclusion of additional sequence data. The recovered phylogeny reveals Slender-billed Grackle to be most closely related to one of two major haplotype clades of Great-tailed Grackle, the other being sister to Boat-tailed Grackle (Q. major). Nicaraguan Grackle appears sister to Carib Grackle (Q. lugubris). We discuss the implications of these and other relationships in the genus for species limits and biogeography. Resumen Las especies del género Quiscalus, junto con su género hermano Euphagus, componen un clado dentro de los tordos americanos (Icteridae). Usamos secuencias génicas del citocromo b y la subunidad 2 de la NADH deshidrogenasa (ND2) para reconstruir relaciones dentro del grupo. El objetivo principal era determinar la posición filogenética y la distinción genética de Quiscalus palustris—una especie endémica poco conocida de la Cuenca del Lerma y de las antiguas zonas lacustres del Valle de México, colectada y registrada por última vez en 1910—y de Q. nicaraguensis, que también es poco común entre las especies de Quiscalus por su distribución restringida. Nuestro análisis difiere de esfuerzos previos por la inclusión de estos taxones junto con todas las demás especies de Quiscalus, por el muestreo intraespecífico de Q. niger, Q. lugubris y Q. mexicanus; y por la inclusión de secuencias adicionales. La filogenia resultante revela que Q. palustris está más cercanamente relacionado con uno de los dos clados mayores de haplotipos de Q. mexicanus, mientras que el otro es un clado hermano de Q. major. La especie Q. nicaraguensis aparece hermana de Q. lugubris. Discutimos las implicaciones de éstas y otras relaciones en el género para establecer límites entre especies y su biogeografía. Introduction The grackles (Quiscalus) are among the most familiar of blackbirds (Icteridae). Common in anthropogenic landscapes, they are conspicuous by virtue of their gregariousness, their habit of foraging on the ground in open areas, their iridescent black or rich rusty brown (in some females) plumages, and their long wedge-shaped tails, which they flare and keel distinctively in flight and display. The seven species currently recognized (AOU 1998) are all very similar with respect to morphology, behavior, and ecology. Nevertheless, across their collective range, which encompasses much of North America and the Caribbean and extends to the north Pacific and Caribbean coasts of South America (Fig. 1), a great number of forms (currently 30 ssp.; Jaramillo and Burke 1999) have been described. At least 15 of these taxa were recognized as species (Ridgway 1902), but with better understanding of the relative amounts of morphological difference among them, many species, especially island forms, were subsequently reduced to subspecific status (Peters 1921, Hellmayr 1937). Figure 1. Open in new tabDownload slide Breeding distributions (after Ridgely et al. 2007) of the currently recognized grackles (Quiscalus spp.). Slender-billed Grackle (Q. palustris) was known from two small areas, both located approximately at the center of the cross-shaped symbol. The Caribbean island distributions of the Greater Antillean (Q. niger) and Carib (Q. lugubris) grackles occur on opposite sides of the slanted black bar. Figure 1. Open in new tabDownload slide Breeding distributions (after Ridgely et al. 2007) of the currently recognized grackles (Quiscalus spp.). Slender-billed Grackle (Q. palustris) was known from two small areas, both located approximately at the center of the cross-shaped symbol. The Caribbean island distributions of the Greater Antillean (Q. niger) and Carib (Q. lugubris) grackles occur on opposite sides of the slanted black bar. Although several studies (Avise and Zink 1988, Zink et al. 1991, Lovette et al. 1999, Kerr et al. 2007, DaCosta et al. 2008) have examined intraspecific molecular variation in grackles, only one (DaCosta et al. 2008) was a detailed phylogeographic analysis. That study, of Great-tailed (Q. mexicanus) and Boat-tailed (Q. major) Grackles, found that the former species comprises two geographically distinct clades, the eastern of which is more closely related to Boat-tailed Grackle than to the western clade, thus rendering Great-tailed Grackle paraphyletic. Similarly, (Lovette et al. 1999) reported that an unpublished study of the Carib Grackle (Q. lugubris) found that it comprises at least two geographically distinct lineages that are >3% genetically divergent; however, the significance of this finding for grackle phylogenetics has not been investigated. All other molecular studies of the grackles (Johnson and Lanyon 1999, Lanyon and Omland 1999, Eaton 2006) have reconstructed phylogenetic relations among recognized species using a single sample of each. Furthermore, two species, both unusual for their extremely limited distributions (Fig. 1), have not been included in any molecular studies—the Slender-billed (Q. palustris) and Nicaraguan (Q. nicaraguensis) Grackles. The Slender-billed Grackle (hereafter palustris) is the only blackbird (and one of a handful of New World nine-primaried oscines) driven to extinction in historic times. It was endemic to central Mexico, but the lake and marsh systems in which it lived have been extensively drained and diminished over the past five centuries for agricultural and urban development (Peterson and Navarro-Sigüenza 2006). The only records of palustris are from the Valley of Mexico (now the metropolitan area of Mexico City), where the type specimen was collected in 1827 and the species was reportedly still present in ~1890 (Peterson 1998), and from the headwaters of the Río Lerma, where specimens were collected in 1904 and 1910 (Dickerman 1965). The Nicaraguan Grackle (hereafter nicaraguensis) is an endemic of the marshes and lowlands around Lakes Managua and Nicaragua, where populations appear secure (IUCN 2008). Because of their distinctive morphologies, recognition of palustris and nicaraguensis as species has rarely been questioned, but traditional hypotheses of grackle evolutionary relationships, and of these species particularly, have varied. Past taxonomies (Ridgway 1902, Hellmayr 1937) have grouped the species of grackles into three genera—Quiscalus, Holoquiscalus, and Cassidix—which are still sometimes recognized as subgenera. The long, straight, fine bill of palustris was unlike that of any other grackle (Jaramillo and Burke 1999), but since it was otherwise similar to mexicanus and major, all authorities have grouped palustris with them in Cassidix (Table 1). Blake (1968) suggested palustris might have been a local race of mexicanus, whereas (Selander and Giller 1961) speculated on a sister pairing with major and possible connection with nicaraguensis based on the marsh-nesting habits of these species. Generally, nicaraguensis has been included in Cassidix, perhaps because of its long tail and marsh association. (Hellmayr 1937) considered nicaraguensis to be "allied" with palustris. However, some authorities (Bond 1950, Lack 1976) thought nicaraguensis to be closely related to Greater Antillean (Q. niger) and Carib (Q. lugubris) grackles. The latter two species have been considered a superspecies in the subgenus Holoquiscalus (Jaramillo and Burke 1999). Although (Ridgway 1902) placed nicaraguensis in Cassidix (using the synonym Megaquiscalus), he noted that it shares a feather structural character with most forms of Holoquiscalus. He considered these subgenera more closely related to one another than to the Common Grackle (Q. quiscula) on the basis of morphological similarities (regarding nicaraguensis and lugubris as intermediate forms), as did (Jaramillo and Burke 1999), based on similarities in plumage and voice. (Yang and Selander 1968) noted that displays and vocalizations of nicaraguensis are similar to quiscula, niger and lugubris. They also proposed a derivation of quiscula from niger. Table 1. Specimens sequenced for reconstructing phylogenetic relationships among the grackles (Quiscalus spp.) and Euphagus blackbirds using mitochondrial cytochrome b and ND2 gene sequences. Open in new tab Table 1. Specimens sequenced for reconstructing phylogenetic relationships among the grackles (Quiscalus spp.) and Euphagus blackbirds using mitochondrial cytochrome b and ND2 gene sequences. Open in new tab Previous formal analyses of grackle phylogeny have used either morphological or molecular characters. The phylogeny of (Björklund 1991), based on 23 morphological characters, placed nicaraguensis sister to Rusty Blackbird (Euphagus carolinus) and did not recover Quiscalus, Euphagus, or their union as monophyletic. However, a reanalysis of Björklund's data (Johnson and Lanyon 1999) revealed that only one node (pairing major and niger) in his tree had >50% bootstrap support. (Johnson and Lanyon 1999) used mtDNA to analyze relationships among the grackles and related blackbirds using parsimony; (Eaton 2006) reanalyzed the same dataset using maximum likelihood and Bayesian methods. These analyses recovered Quiscalus as monophyletic and sister to Euphagus. Subgenus Holoquiscalus was rendered paraphyletic by the closer relationship of niger to the Cassidix group than to its previously supposed sister taxon, lugubris. Our primary objectives in this study were to determine the phylogenetic positions and genetic distinctiveness of palustris and nicaraguensis in the context of a wider reevaluation of grackle relationships inferred from the sequences of two protein-coding mitochondrial genes, cytochrome b and NADH dehydrogenase subunit 2 (ND2). Our analysis differs from the most comprehensive previous efforts (Johnson and Lanyon 1999, Eaton 2006) insofar as we used ∼20% more sequence data (2292 base pairs total), included for the first time all recognized species of grackles, and included multiple representatives of species known—as in the cases of lugubris and mexicanus—or suspected (as in niger) to harbor deep phylogenetic divergences among populations. Methods Taxon Sampling Ingroup sampling (Table 1) included at least one individual from each species of Quiscalus (Sibley and Monroe 1990, AOU 1998). Further, we included representatives (provided by J. M. DaCosta, Marjorie Barrick Museum) of the principle haplotype clades known from mexicanus (DaCosta et al. 2008), and we sequenced specimens of previously unsampled subspecies (graysoni, obscurus) in the western portion of the species' range. Similarly, we included samples of the divergent lineages within lugubris (Lovette et al. 1999) and specimens of niger from four of the five islands (or island groups) on which it occurs. Previous molecular analyses of blackbird phylogeny (Johnson and Lanyon 1999, Lanyon and Omland 1999, Eaton 2006) recovered Quiscalus as monophyletic and sister to Euphagus with unequivocal support, so outgroup sampling was limited to the two recognized species of the latter genus. Laboratory Procedures We extracted genomic DNA from tissue samples (Table 1) using a DNeasy Tissue Kit (Qiagen, Valencia, California) following manufacturer instructions, except that for specimens sampled from toe pads, 30 μl of 100 mg per ml dithiothreitol (Gold Biotechnology, St. Louis, Missouri) was added to the digestions, and 50 μl Buffer AE used for each DNA elution. Because the palustris sample was taken from a toe pad of a skin prepared in 1904, its intact DNA concentration was low, so we amplified cytochrome b in six fragments using primers pairs ND5emb1 and H15103 (Barker et al. 2008), L15069-Qp (5′CTAGCCATACACTACACAGCAGAC) and H15305-Qp (5′CGGTAGCGCCTCAGAATGATATTT), L15259-Qp (5′GTTGGAGTCATTCTCCTCCTAA) and H15460-Qp (5′GTGAACTAGGGTAAGTCCTACGAT), L15410 (Barker et al. 2008) and H15709 (Barker 2004), L15656-Qp (5′AACCT CCTAGGAGATCCAGA) and H15934 (Barker et al. 2008), and L15848-Qp (5′CAAAACTACGATCAATGACYTTCCG) with H16137 (Sorenson et al. 1999). The mexicanus toe-pad samples were treated similarly but were amplified in five fragments (H15305 and L15259-Qp were not used), and L15656 (Helm-Bychowski and Cracraft 1993) was substituted for L15656-Qp. Reaction preparation and cycling parameters were as described by (Barker et al. 2008). The initial products were reamplified when necessary to obtain sufficient concentrations for sequencing. We amplified cytochrome b from frozen tissue specimens using ND5emb2 (5′GGYCTAAYCAAAGCCTAYCTA) and H16137, reamplifying when necessary with primer pairs ND5emb1 and H15305-Qp, L15069-Qp and H15709, and L15656-Qp with H16137. To amplify ND2 from frozen tissue specimens, we used primers LMET (Hackett 1996) and H1064 (Drovetski et al. 2004). We obtained the complete sequence of ND2, and ~890 bp of cytochrome b sequence of seven individuals in our study from GenBank; for these same individuals, in order to complete the cytochrome b sequences, we amplified the regions at each end of the gene using primers ND5emb1 with H15103, and L15848-Qp with H16137. We purified PCR products through enzymatic digestion using exonuclease 1 and shrimp alkaline phosphatase (USB Corporation, Cleveland, Ohio), following (Werle et al. 1994), and sequenced them following manufacturer recommendations on an ABI 3700 automated sequencer (BigDye v3.1, Applied Biosystems, Foster City, California) at the Biomedical Genomics Center of the University of Minnesota. We used the same primers as for PCR, except that for whole-gene amplifications, we used some additional, internal primers. These were L5758emb (Barker et al. 2008) and H5766emb (Barker et al. 2008) for ND2, and for cytochrome b, they were L15191 (Lanyon and Hall 1994), L15656, H15709, and H15298 (Helm-Bychowski and Cracraft 1993). We used Sequencher 4.7 (Genecodes, Ann Arbor, Michigan) to align and view chromatograms of complimentary reads and overlapping fragments to produce a consensus sequence. Phylogenetic Analyses Two sets of analyses were performed. The primary set utilized all data acquired, which for most individuals comprised the complete ND2 (1040 bp) and cytochrome b (1143 bp) gene sequences, plus a spacer region and tRNA sequence (108 bp) adjoining the latter. The second set utilized only the cytochrome b (with spacer and tRNA) data, and was conducted to determine whether the results of the primary set were distorted by the lack of ND2 data from individuals sequenced from toe pads (Table 1). We used PAUP* 4.0b10 (Swofford 2002) to infer phylogenetic relations among taxa under maximum parsimony and maximum likelihood criteria, and used MrBayes 3.1.2 (Huelsenbeck and Ronquist 2001, Ronquist and Huelsenbeck 2003) to employ Bayesian methods. Parsimony analyses were conducted with branch-and-bound searches. Support for nodes was assessed, after excluding uninformative characters, with 10 000 nonparametric bootstrap replicates. We used DT-ModSel (Minin et al. 2003) to select the most appropriate models for analyses of the data using maximum likelihood and Bayesian methods. Parameter values for maximum likelihood analyses were obtained through estimation on maximum parsimony trees using PAUP*. Heuristic searches for the best maximum likelihood tree for each dataset were conducted using tree bisection and reconnection branch swapping with 1000 random-addition sequences of taxa. Clock- and nonclock-like models of sequence evolution on this tree were not significantly different (α = 0.05) for the complete dataset according to a likelihood ratio test (Δ = 14.1, df = 19, P = 0.78), but were different for cytochrome b alone (Δ = 31.7, df = 19, P = 0.03). We conducted a second set of heuristic searches for the best maximum likelihood tree for each dataset with clock-like evolution enforced. The topologies of the best clock and nonclock trees for cytochrome b alone were slightly different within the eastern clade of mexicanus, but clock and nonclock models were not significantly different when evaluated on these trees (Δ = 14.0, df = 19, P = 0.79). Support for nodes in maximum likelihood trees were assessed with 500 bootstrap replicates (10 random addition sequences each). Sequence divergences were calculated in PAUP* using nonclock maximum likelihood parameter estimates. We used empirical Bayes factors (Kass and Raftery 1995, Nylander et al. 2004) to select partitioning schemes for Bayesian analyses of each dataset. We tested unpartitioned analyses, partitioning by gene, codon position, and codon position by gene, and the effect of assuming clock-like evolution. For both datasets, a maximally partitioned scheme (including a partition for the spacer and tRNA sequence) and enforcing a strict clock was best (2logeB10 = 19 for the complete dataset and 2logeB10 = 25 for cytochrome b in comparison to the next best models). We conducted final analyses for each dataset with and without enforcing clock-like evolution. For each analysis, we ran four coupled chains for three million generations, sampling every 100. Samples prior to reaching stationarity were discarded as "burn-in" and the remaining subsamples used to create 50% majority-rule consensus trees. Results Parsimony, likelihood, and Bayesian methods recovered similar patterns of relationship. Well-supported nodes inferred using cytochrome b sequences alone (Fig. 2, Table 2) were consistent with the better-supported and more finely resolved topology generated using both cytochrome b and ND2 (but lacking ND2 data for four specimens), so we consider the latter (Fig. 3, Table 2) our best estimate of the phylogeny of the group. Table 2. Dataset description, model parameter values, and analysis summaries for the two mtDNA datasets (ND2 with cytochrome b, and cytochrome b alone) used in reconstructing phylogenetic relationships among the grackles (Quiscalus spp.) and Euphagus blackbirds. For maximum likelihood (ML) and some Bayesian parameters, values are given for models with and without clock-like sequence evolution enforced. CI = ensemble consistency index, RI = ensemble retention index, πi = base frequency, Nst = number of substitution types, TI/TV ratio = transition/transversion ratio, piv = proportion of invariant sites,–ln(l) = negative natural log likelihood of best tree. Open in new tab Table 2. Dataset description, model parameter values, and analysis summaries for the two mtDNA datasets (ND2 with cytochrome b, and cytochrome b alone) used in reconstructing phylogenetic relationships among the grackles (Quiscalus spp.) and Euphagus blackbirds. For maximum likelihood (ML) and some Bayesian parameters, values are given for models with and without clock-like sequence evolution enforced. CI = ensemble consistency index, RI = ensemble retention index, πi = base frequency, Nst = number of substitution types, TI/TV ratio = transition/transversion ratio, piv = proportion of invariant sites,–ln(l) = negative natural log likelihood of best tree. Open in new tab Figure 2. Open in new tabDownload slide Phylogeny of the grackles (Quiscalus spp.; rooted using Euphagus), determined from analysis of mitochondrial cytochrome b with adjoining spacer and tRNA sequence. See Table 1 for specimen details. Left: strict consensus of 12 equally parsimonious trees with nonparametric bootstrap support values. Right: one of two best maximum likelihood trees without enforcing molecular clock. The relationships recovered within the eastern clade of Q. mexicanus differed according to the method of analysis; topologies found with clock-enforced maximum likelihood (ML) and with Bayesian analyses are shown at far right. Nonparametric bootstrap support values followed by estimated Bayesian posterior probabilities (× 100) of nodes without molecular clock enforced are shown above values with molecular clock enforced. Arrows connect support values to nodes when they could not be fitted above and below the adjacent branches. Nodal support was <50% when indicated with a dash or not given. Figure 2. Open in new tabDownload slide Phylogeny of the grackles (Quiscalus spp.; rooted using Euphagus), determined from analysis of mitochondrial cytochrome b with adjoining spacer and tRNA sequence. See Table 1 for specimen details. Left: strict consensus of 12 equally parsimonious trees with nonparametric bootstrap support values. Right: one of two best maximum likelihood trees without enforcing molecular clock. The relationships recovered within the eastern clade of Q. mexicanus differed according to the method of analysis; topologies found with clock-enforced maximum likelihood (ML) and with Bayesian analyses are shown at far right. Nonparametric bootstrap support values followed by estimated Bayesian posterior probabilities (× 100) of nodes without molecular clock enforced are shown above values with molecular clock enforced. Arrows connect support values to nodes when they could not be fitted above and below the adjacent branches. Nodal support was <50% when indicated with a dash or not given. Figure 3. Open in new tabDownload slide Phylogeny of the grackles (Quiscalus spp.; rooted using Euphagus) determined from analysis of mitochondrial cytochrome b and ND2 gene sequences. See Table 1 for specimen details. Left: strict consensus of 10 equally parsimonious trees with nonparametric bootstrap support values. Right: single best maximum likelihood tree without enforcing molecular clock (where different, topology found with Bayesian analysis shown with dashed line. Nonparametric bootstrap support followed by Bayesian posterior probabilities (× 100) of nodes without molecular clock enforced are shown above values with molecular clock enforced. Arrows connect support values to nodes when they could not be fitted above and below the adjacent branches. Nodal support was <50% when indicated with a dash or not given. Figure 3. Open in new tabDownload slide Phylogeny of the grackles (Quiscalus spp.; rooted using Euphagus) determined from analysis of mitochondrial cytochrome b and ND2 gene sequences. See Table 1 for specimen details. Left: strict consensus of 10 equally parsimonious trees with nonparametric bootstrap support values. Right: single best maximum likelihood tree without enforcing molecular clock (where different, topology found with Bayesian analysis shown with dashed line. Nonparametric bootstrap support followed by Bayesian posterior probabilities (× 100) of nodes without molecular clock enforced are shown above values with molecular clock enforced. Arrows connect support values to nodes when they could not be fitted above and below the adjacent branches. Nodal support was <50% when indicated with a dash or not given. In accordance with (DaCosta et al. 2008), we found two highly divergent clades of mexicanus. One clade is distributed mostly west of the Sierra Madre Occidental, and the other, to the east of that range. All analyses strongly supported sister relationships between palustris and the western clade of mexicanus and between major and the eastern clade. Amounts of cytochrome b sequence divergence (Table 3) were modest within the western and eastern clades of mexicanus, averaging 0.3% (range: 0.3%–0.4%) and 0.5% (range: 0.2%–0.8%), respectively, in comparison to the 3.1% divergence between those clades and to their divergences from their sister taxa. Q. palustris was 2.0% divergent from western mexicanus, whereas major was 1.4% divergent from the eastern clade. A sister relationship between the palustris-western mexicanus, and major-eastern mexicanus clades was weakly to modestly supported using the full dataset, but the cytochrome b data alone were unable to dichotomously resolve the relationships among these clades and niger. The niger samples composed a strongly supported monophyletic unit; sequence divergences among islands averaged 1.3% (range: 0.5%–1.9%). Table 3. Genetic divergences (average pairwise %) among grackle (Quiscalus spp.) and Euphagus blackbird species based upon analysis of cytochrome b (with adjoining spacer region and tRNA) under a maximum likelihood model of sequence evolution without enforcing molecular clock. Eastern and western clades of Q. mexicanus are listed separately. Average pairwise within-taxon divergences are shown on the diagonal for taxa with two or more samples; a dash indicates only one sample. Open in new tab Table 3. Genetic divergences (average pairwise %) among grackle (Quiscalus spp.) and Euphagus blackbird species based upon analysis of cytochrome b (with adjoining spacer region and tRNA) under a maximum likelihood model of sequence evolution without enforcing molecular clock. Eastern and western clades of Q. mexicanus are listed separately. Average pairwise within-taxon divergences are shown on the diagonal for taxa with two or more samples; a dash indicates only one sample. Open in new tab We found nicaraguensis to be sister to lugubris, but relationships among nicaraguensis and the two highly divergent (3.9% in cytochrome b; Table 3) lugubris subspecies were not well resolved. Using the complete dataset, all analyses recovered these three lineages as a clade with modest to strong support, and Bayesian and maximum likelihood analyses recovered lugubris as monophyletic with weak support (Fig. 3). Using the cytochrome b data alone, maximum likelihood and clock-enforced Bayesian analyses recovered these three lineages as a clade, but support for this relationship, and for a monophyletic lugubris, was lacking. Under parsimony, a sister relationship between nicaraguensis and the Lesser Antillean lineage of lugubris received weak bootstrap support, whereas maximum likelihood and Bayesian analyses grouped nicaraguensis and mainland lugubris with weak support (Fig. 2). Discussion Our best-resolved phylogeny (Fig. 3) of the grackles (Quiscalus spp.) is consistent with earlier molecular analyses (Johnson and Lanyon 1999, Eaton 2006) but reveals a more complex pattern of relationships than previously recognized. We discovered the phylogenetic positions of two species—palustris and nicaraguensis—not included in previous molecular studies, found these taxa genetically distinct from their closest relatives, and found that some named species comprise deeply divergent lineages. Our phylogeny conflicts with some traditional notions about the relationships among grackle species. Sister Lineage Divergences and Intraspecific Variation Our analyses confirmed that Q. mexicanus comprises two deeply divergent haplotype clades (Wehtje 2004, DaCosta et al. 2008) that are also geographically distinct, suggesting that they represent lineages that split ~2 million years ago (assuming 1.6% divergence per million years; Fleischer et al. 1998, but see Weir and Schluter 2008). The western clade corresponds to subspecies nelsoni and graysoni, notable for being the smallest mexicanus forms and having very pale female plumage (Jaramillo and Burke 1999). Prior to the recent expansion of nelsoni northward into California and the desert southwest of the United States (Phillips 1950, Wehtje 2003, 2004), this clade was restricted to Sonora and coastal Sinaloa west of the Sierra Madre Occidental (Fig. 4). The eastern clade, composed of the larger mexicanus forms with dark brown female plumages, was represented in our study by subspecies mexicanus, monsoni, and obscurus, but (DaCosta et al. 2008) found that it also includes prosopidicola and peruvianus, and thus likely encompasses all Q. mexicanus outside the western clade. Prior to its recent expansion to the Gulf Coast, Great Plains, and southwest of the United States (Wehtje 2003, 2004), this clade was distributed from north-central Mexico east to the Gulf Coast and south through Central America to coastal northern South America. Despite its wide distribution, we found that divergences within this clade were shallow, poorly supported, and imperfectly congruent with named subspecies (see also DaCosta et al. 2008). Figure 4. Open in new tabDownload slide Distributions in ~1960 of the subspecies of Great-tailed Grackle (Quiscalus mexicanus) in Mexico and the United States (after Selander and Giller 1961). Unlabeled dark shading indicates areas of range overlap. Dashed lines indicate state boundaries. Figure 4. Open in new tabDownload slide Distributions in ~1960 of the subspecies of Great-tailed Grackle (Quiscalus mexicanus) in Mexico and the United States (after Selander and Giller 1961). Unlabeled dark shading indicates areas of range overlap. Dashed lines indicate state boundaries. In spite of their genetic and morphological divergence, the eastern and western mexicanus clades interbreed freely (Johnson and Peer 2001, Wehtje 2004) in the southwestern United States, where their distributions now overlap due to expansion over the past ~60 years. Ours is the first molecular study to include graysoni and obscurus and thereby discover that these subspecies belong to the western and eastern mexicanus clades respectively, a biogeographically important finding because their point of contact (Fig. 4) along the Pacific slope of the Sierra Madre Occidental just north of its juncture with the Sierra Madre del Sur represents the only known zone of overlap between these clades prior to recent range expansions. The extent to which graysoni and obscurus interbreed is unknown, but interestingly, these forms sing "noticeably different" songs (Johnson and Peer 2001) and represent the extremes of female plumage variation within Q. mexicanus, with graysoni being pale gray-buff and many female obscurus being nearly black (AFLAP, pers. obs.; Jaramillo and Burke 1999). The sister relationship between palustris and the western mexicanus clade that we discovered, and the sister relationship between major and the eastern clade (see also DaCosta et al. 2008) render mexicanus paraphyletic on two counts. Although this result might lead some to question the specific status of both palustris and major, both species are morphologically and ecologically distinct, and their genetic divergences from the mexicanus clades to which they are most closely related are large in comparison to the average amounts of divergence we found within those clades and the 0.2% maximum found within major (DaCosta et al. 2008). Furthermore, interbreeding is rare enough between major and eastern mexicanus where their ranges now overlap along the Gulf Coast (Fig. 1) that they are distinct under the Biological Species Concept (Selander and Giller 1961, Post et al. 1996). Thus, interbreeding of the mexicanus clades likely reflects retention of ancestral compatibilities as seen also in geese (Paxinos et al. 2002). It seems likely that the morphological similarities, including large size and long tail, of the mexicanus clades, major, and palustris are indicative of their collective monophyly, and that the small size and short tail of niger represents retention of ancestral characteristics also seen in lugubris; however, support for this constellation of relationships was weak. It is conceivable that the mexicanus clades are not sister taxa and that the morphological similarity of niger to lugubris is a result of convergence or reversion to ancestral characteristics brought about through adaptation to the Caribbean island context. The 1.3% average genetic divergence among niger samples was much greater than that seen within the mexicanus clades and major, and the 0.3% known from quiscula (Zink et al. 1991). The magnitudes of divergences among islands suggest relatively great evolutionary independence among populations compared to these other species, and their pattern suggests a history of island colonization from Puerto Rico westward. However, this pattern could be coincidental, and additional sampling would be necessary to establish whether our samples characterize interisland haplotype differences (i.e., whether island populations are reciprocally monophyletic). The 3.9% difference between haplotypes of lugubris samples from South America and the Lesser Antilles is a result we anticipated based on comments by (Lovette et al. 1999). They reported that lugubris haplotypes in Barbados were identical to some in Trinidad, but that samples from St. Vincent and the rest of the Lesser Antilles differed from these by >3% in mtATPase sequences. This substantial genetic divide does not correspond to known patterns of morphological similarity among subspecies (e.g., the forms on Barbados and St. Vincent, which have blackish rather than brown female plumages, were once classified together as a species separate from all other lugubris ssp.; Peters 1921). Somewhat surprisingly, we found that nicaraguensis, which is morphologically and behaviorally distinct, is at most only marginally more genetically divergent from these lineages than they are from one another. Given the lack of support for resolving relationships among these three lineages and their apparently very long histories of evolutionary independence, the lugubris lineages are probably best regarded as separate species. The similar morphologies of Antillean lugubris, mainland lugubris, and niger suggest that their appearances are conserved from the common ancestor of all Quiscalus apart from quiscula. Implications for Higher-Level Relationships, Grackle Evolution, and Biogeographic History Our analyses produced a robust phylogeny for Quiscalus that contradicts some past suppositions about grackle relations. First, it does not support the use of the subgenera Holoquiscalus and Cassidix. The former is rendered paraphyletic, even if nicaraguensis were incorporated into it, by the closer relationship of niger to mexicanus, major, and palustris than to lugubris. The latter subgenus is polyphyletic due to the position of nicaraguensis; in addition, support for the monophyly of the remainder of Cassidix is only modest. Second, the species with the strongest ties to marsh habitat—palustris, major, nicaraguensis, and the peruvianus form of mexicanus—do not compose a clade, nor is it clear that this association is symplesiomorphic, as has been asserted (Selander and Giller 1961, Yang and Selander 1968). Third, the richly buff and pale plumage components of females in the aforementioned taxa (Selander and Giller 1961, Jaramillo and Burke 1999), and the strong sexual dimorphism of tail length in these species are likewise not indicative of close relationships. Finally, quiscula appears to be sister to all other Quiscalus, so the hypothesis that it was derived from an ancestral niger isolated in Florida (Yang and Selander 1968) is not supported. The basal divergences of Euphagus and quiscula suggest a continental North American origin for Quiscalus, but our phylogeny does not lend itself to simple inferences about the early biogeography of the genus (the sister to the Quiscalus-Euphagus clade, Dives [FKB and SML, unpubl. data], is represented by one species each in Central America, South America, and the Caribbean). Rather, it perhaps accurately reflects a history of dynamic fluctuations in distributions during the climatically dynamic late Pliocene and Pleistocene, when diversification of the clade appears to have occurred. Such a history is expected given that grackles are water-associated creatures of savanna, woodland edge, and open marsh—ephemeral habitats that vary considerably in their distribution and extent over time—and, as seen from the range expansions of several species in the past century, they are capable of rapid population and distributional responses to habitat availability. The curiously localized palustris appears to have diverged ~1.2 million years ago from its geographically distant sister, western mexicanus. Q. mexicanus is now common throughout the interior of Mexico, but until the last century, its distribution was more limited, and it was most common on the coastal plains (Christensen 2000). Consequently, eastern mexicanus, western mexicanus, and palustris had mostly allopatric distributions. It is the eastern clade, in the form of subspecies mexicanus, that has very recently spread into central Mexico where palustris once occurred. Whether these taxa ever came into contact is an interesting subject for speculation in light of questions of species limits and the possibility that ecological competition had a role in the extinction of palustris. The first modern occurrence of Q. mexicanus in the Valley of Mexico was in ~1960 at Xochimilco (Dickerman 1965), the same area to which palustris was confined in ~1890 when last reported in the valley (Peterson 1998). However, historical accounts indicate that, in ~1500, mexicanus was introduced to the valley from the Gulf Coast by the Aztecs (Haemig 1978). Common in ~1570, it perhaps declined over the next century and disappeared due to changes in land use (Christensen 2000). The species of grackles, judging from pairwise divergences, appear to be quite recent in comparison to other congeneric North American birds (Klicka and Zink 1997), making them well suited to studies of processes related to speciation. The group allows for comparison of closely related species that have diverged substantially in size or in sexual size dimorphism, as well as those that are genetically divergent yet morphologically similar. Furthermore, range expansions in the past century have brought several of these formerly allopatric species and lineages into secondary contact or more extensive sympatry, thus allowing for study of interspecific (or interclade) interactions with respect to ecological competition, behaviors related to mate attraction and selection, and the consequences of interclade hybridization and introgression. In addition, phylogeographic and population genetic studies of mexicanus, niger, and lugubris are needed to better understand their present population structuring and evolutionary histories. We hope that this first complete species-level phylogeny of the grackles will stimulate further work on these and other aspects of the group's evolution, ecology, and behavior. Acknowledgments We thank G. D. Weiblen for use of his lab during preparation of Q. palustris and other toe pad samples, J. M. DaCosta for sharing Q. mexicanus sequence data, H. Vázquez-Miranda for helpful discussion and Spanish translation of our abstract, and two anonymous reviewers for comments on the manuscript. We also thank the following institutions for maintaining and making available the specimens that made this research possible: American Museum of Natural History, James Ford Bell Museum of Natural History, Marjorie Barrick Museum of Natural History, Field Museum of Natural History, University of Kansas Natural History Museum, Natural History Museum of Los Angeles County, University of New Mexico Museum of Southwestern Biology, and the National Museum of Natural History. This study was supported in part by NSF DEB-0316092 to FKB and SML. Literature Cited American Ornithologists' Union. 1998 . Check-list of North American Birds. 7th ed. American Ornithologists' Union Washington, DC . Avise , J. C. , and R. M. Zink . 1988 . Molecular genetic divergence between avian sibling species: King and Clapper Rails, Long-billed and Short-billed Dowitchers, Boat-tailed and Great-tailed Grackles, and Tufted and Black-crested Titmice. Auk 105 : 516 – 528 . WorldCat Barker , F. K. 2004 . Monophyly and relationships of wrens (Aves: Troglodytidae): a congruence analysis of heterogeneous mitochondrial and nuclear DNA sequence data. Molecular Phylogenetics and Evolution 31 : 486 – 504 . Google Scholar Crossref Search ADS PubMed WorldCat Barker , F. K. , A. J. Vandergon , and S. M. Lanyon . 2008 . Assessment of species limits among yellow-breasted meadowlarks (Sturnella spp.) using mitochondrial and sex-linked markers. Auk 125 : 869 – 879 . Google Scholar Crossref Search ADS WorldCat Björklund , M. 1991 . Evolution, phylogeny, sexual dimorphism and mating system in the grackles (Quiscalus spp.: Icterinae). Evolution 45 : 608 – 621 . Google Scholar Crossref Search ADS PubMed WorldCat Blake , E. R. 1968 . Family Icteridae: American orioles and blackbirds. 138 – 202 . Paynter R. A. Jr. , editor. Check-list of birds of the world: a continuation of the work of James L. Peters Museum of Comparative Zoology . Cambridge, MA . Bond , J. 1950 . Check-list of Birds of the West Indies 3rd ed. Academy of Natural Sciences of Philadelphia . Philadelphia, PA . Christensen , A. F. 2000 . The fifteenth- and twentieth-century colonization of the Basin of Mexico by the Great-tailed Grackle (Quiscalus mexicanus). Global Ecology and Biogeography 9 : 415 – 420 . Google Scholar Crossref Search ADS WorldCat Dacosta , J. M. , W. Wehtje , and J. Klicka . 2008 . Historic genetic structuring and paraphyly within the Great-tailed Grackle. Condor 110 : 170 – 177 . Google Scholar Crossref Search ADS WorldCat Dickerman , R. W. 1965 . The juvenal plumage and distribution of Cassidix palustris (Swainson). Auk 82 : 268 – 270 . Google Scholar Crossref Search ADS WorldCat Drovetski , S. V. , R. M. Zink , I. V. Fadeev , E. V. Nesterov , E. A. Koblik , Y. A. Red'Kin , and S. Rohwer . 2004 . Mitochondrial phylogeny of Locustella and related genera. Journal of Avian Biology 35 : 105 – 110 . Google Scholar Crossref Search ADS WorldCat Eaton , M. D. 2006 . A phylogenetic perspective on the evolution of chromatic ultraviolet plumage coloration in grackles and allies (Icteridae). Auk 123 : 211 – 234 . Google Scholar Crossref Search ADS WorldCat Fleischer , R. C. , C. E. Mcintosh , and C. L. Tarr . 1998 . Evolution on a volcanic conveyor belt: using phylogenetic reconstructions and K–Ar based ages of the Hawaiian Islands to estimate molecular evolutionary rates. Molecular Ecology 7 : 533 – 545 . Google Scholar Crossref Search ADS PubMed WorldCat Hackett , S. J. 1996 . Molecular phylogenetics and biogeography of tanagers in the genus Ramphocelus (Aves). Molecular Phylogenetics and Evolution 5 : 368 – 382 . Google Scholar Crossref Search ADS PubMed WorldCat Hasegawa , M. , H. Kishino , and T. Yano . 1985 . Dating of the human-ape splitting by a molecular clock of mitochondrial DNA. Journal of Molecular Evolution 22 : 160 – 174 . Google Scholar Crossref Search ADS PubMed WorldCat Haemig , P. D. 1978 . Aztec emperor Auitzotl and the Great-tailed Grackle. Biotropica 10 : 11 – 17 . Google Scholar Crossref Search ADS WorldCat Hellmayr , C. E. 1937 . Catalogue of birds of the Americas and the adjacent islands. Part X. Icteridae. Field Museum of Natural History, Zoological Series 13 . Helm-Bychowski , K. , and J. Cracraft . 1993 . Recovering phylogenetic signal from DNA sequences: relationships within the corvine assemblage (class Aves) as inferred from the complete sequences of the mitochondrial DNA cytochrome-b gene. Molecular Biology and Evolution 10 : 1196 – 1214 . Google Scholar PubMed WorldCat Huelsenbeck , J. P. , and F. Ronquist . 2001 . MrBayes: Bayesian inference of phylogeny. Bioinformatics 17 : 754 – 755 . Google Scholar Crossref Search ADS PubMed WorldCat IUCN. [Online] . 2008 . Quiscalus nicaraguensis. 2008 IUCN Red List of Threatened Species <http://www.iucnredlist.org/details/150421> (31 October 2008) . Jaramillo , A. , and P. Burke . 1999 . New World blackbirds: the icterids Princeton University Press . Princeton, NJ . Johnson , K. , and B. D. Peer . 2001 . Great-tailed Grackle (Quiscalus mexicanus). Poole , A. , and F. Gill , editors. The birds of North America, No. 576 The Birds of North America, Inc . Philadelphia, PA . Johnson , K. P. , and S. M. Lanyon . 1999 . Molecular systematics of the grackles and allies, and the effect of additional sequence (cytochrome b and ND2). Auk 116 : 759 – 768 . Google Scholar Crossref Search ADS WorldCat Kass , R. E. , and A. E. Raftery . 1995 . Bayes factors. Journal of the American Statistical Association 90 : 773 – 795 . Google Scholar Crossref Search ADS WorldCat Kerr , K. C R. , M. Y. Stoeckle , C. J. Dove , L. A. Weigt , C. M. Francis , and P. D N. Hebert . 2007 . Comprehensive DNA barcode coverage of North American birds. Molecular Ecology Notes 7 : 535 – 543 . Google Scholar Crossref Search ADS PubMed WorldCat Klicka , J. , and R. M. Zink . 1997 . The importance of recent ice ages in speciation: a failed paradigm. Science 277 : 1666 – 1669 . Google Scholar Crossref Search ADS WorldCat Lack , D. 1976 . Island biology illustrated by the land birds of Jamaica University of California Press . Berkeley, CA . Lanyon , S. M. , and J. G. Hall . 1994 . Reexamination of barbet monophyly using mitochondrial-DNA sequence data. Auk 111 : 389 – 397 . Google Scholar Crossref Search ADS WorldCat Lanyon , S. M. , and K. E. Omland . 1999 . A molecular phylogeny of the blackbirds (Icteridae): five lineages revealed by cytochrome-b sequence data. Auk 116 : 629 – 639 . Google Scholar Crossref Search ADS WorldCat Lovette , I. J. , G. Seutin , R. E. Ricklefs , and E. Bermingham . 1999 . The assembly of an island fauna by natural invasion: sources and temporal patterns in the avian colonization of Barbados. Biological Invasions 1 : 33 – 41 . Google Scholar Crossref Search ADS WorldCat Minin , V. , Z. Abdo , P. Joyce , and J. Sullivan . 2003 . Performance-based selection of likelihood models for phylogeny estimation. Systematic Biology 52 : 674 – 683 . Google Scholar Crossref Search ADS PubMed WorldCat Nylander , J. A A. , F. Ronquist , J. P. Huelsenbeck , and J. L. Nieves Aldrey . 2004 . Bayesian phylogenetic analysis of combined data. Systematic Biology 53 : 47 – 67 . Google Scholar Crossref Search ADS PubMed WorldCat Paxinos , E. E. , H. F. James , S. L. Olson , M. D. Sorenson , J. Jackson , and R. C. Fleischer . 2002 . mtDNA from fossils reveals a radiation of Hawaiian geese recently derived from the Canada goose (Branta canadensis). Proceedings of the National Academy of Sciences USA 99 : 1399 – 1404 . Google Scholar Crossref Search ADS WorldCat Peters , J. L. 1921 . A review of the grackles of the genus Holoquiscalus. Auk 38 : 435 – 453 . Google Scholar Crossref Search ADS WorldCat Peterson , A. T. 1998 . The distribution and type locality of the extinct Slender-billed Grackle Quiscalus palustris. Bulletin of the British Ornithologists' Club 118 : 119 – 212 . WorldCat Peterson , A. T. , and A. G. Navarro-Sigüenza . 2006 . Hundred-year changes in the avifauna of the Valley of Mexico, Distrito Federal, Mexico. Huitzil 7 : 4 – 14 . WorldCat Phillips , A. R. 1950 . The Great-tailed Grackles of the Southwest. Condor 52 : 78 – 81 . Google Scholar Crossref Search ADS WorldCat Post , W. , J. P. Poston , and G. T. Bancroft . 1996 . Boat-tailed Grackle (Quiscalus major). Poole , A. , and F. Gill , editors. The birds of North America, No. 207 The Academy of Natural Sciences . Philadelphia, PA The American Ornithologists' Union . Washington, DC . Ridgely , R. S. , T. F. Allnutt , T. Brooks , D. K. Mcnicol , D. W. Mehlman , B. E. Young , and J. R. Zook . 2007 . Digital distribution maps of the birds of the western hemisphere. Version 3.0 NatureServe . Arlington, VA . Ridgway , R. 1902 . The birds of North and Middle America. Part II. Bulletin of the United States National Museum, No. 50 . Ronquist , F. , and J. P. Huelsenbeck . 2003 . MrBayes 3: Bayesian phylogenetic inference under mixed models. Bioinformatics 19 : 1572 – 1574 . Google Scholar Crossref Search ADS PubMed WorldCat Selander , R. K. , and D. R. Giller . 1961 . Analysis of sympatry of Great-tailed and Boat-tailed Grackles. Condor 63 : 29 – 86 . Google Scholar Crossref Search ADS WorldCat Sibley , C. G. , and B. L. Monroe Jr . 1990 . Distribution and taxonomy of birds of the world Yale University Press . New Haven, CT . Sorenson , M. D. , J. C. Ast , D. E. Dimcheff , T. Yuri , and D. P. Mindell . 1999 . Primers for a PCR-based approach to mitochondrial genome sequencing in birds and other vertebrates. Molecular Phylogenetics and Evolution 12 : 105 – 114 . Google Scholar Crossref Search ADS PubMed WorldCat Swofford , D. L. 2002 . PAUP*. Phylogenetic analysis using parsimony (*and other methods). Version 4 Sinauer . Sunderland, MA . Wehtje , W. 2003 . The range expansion of the Great-tailed Grackle (Quiscalus mexicanus Gmelin) in North America since 1880. Journal of Biogeography 30 : 1593 – 1607 . Google Scholar Crossref Search ADS WorldCat Wehtje , W. 2004 . The Great-tailed Grackle (Quiscalus mexicanus Gmelin) in the western USA: range expansion and secondary contact between subspecies Ph.D. dissertation . University of California, Riverside, CA . Weir , J. T. , and D. Schluter . 2008 . Calibrating the avian molecular clock. Molecular Ecology 17 : 2321 – 2328 . Google Scholar Crossref Search ADS PubMed WorldCat Werle , E. , C. Schneider , M. Renner , M. Völker , and W. Fiehn . 1994 . Convenient single-step, one tube purification of PCR products for direct sequencing. Nucleic Acids Research 22 : 4354 – 4355 . Google Scholar Crossref Search ADS PubMed WorldCat Yang , S. Y. , and R. K. Selander . 1968 . Hybridization in the grackle Quiscalus quiscula in Louisiana. Systematic Zoology 17 : 107 – 143 . Google Scholar Crossref Search ADS PubMed WorldCat Zink , R. M. , W. L. Rootes , and D. L. Dittman . 1991 . Mitochondrial DNA variation, population structure, and evolution of the Common Grackle (Quiscalus quiscula). Condor 93 : 318 – 329 . Google Scholar Crossref Search ADS WorldCat Author notes E-mail: alveypowell@yahoo.com © The Cooper Ornithological Society, 2008
TI - A Complete Species-Level Phylogeny of the Grackles (Quiscalus spp.), Including the Extinct Slender-Billed Grackle, Inferred from Mitochondrial DNAFilogenia Completa a Nivel Específico del Género Quiscalus, Incluyendo la Especie Extinta Quiscalus palustris, Inferida a Partir de ADN MitocondrialAlexis F. L. A. Powell Et AlPhylogeny of The Grackles
JF - Condor: Ornithological Applications
DO - 10.1525/cond.2008.8633
DA - 2008-11-01
UR - https://www.deepdyve.com/lp/oxford-university-press/a-complete-species-level-phylogeny-of-the-grackles-quiscalus-spp-XM0ioEZ0OJ
SP - 718
VL - 110
IS - 4
DP - DeepDyve
ER -