DIVERGENCE BETWEEN SUBSPECIES GROUPS OF SWAINSON'S THRUSH (CATHARUS USTULATUS USTULATUS AND C. U. SWAINSONI)

Abstract

—Swainson’s Thrush (Catharus ustulatus) is a long-distance Nearctic–Neotropical migrant that includes two major subspecies groups: the russet-backed group (C. u. ustulatus) of the Pacifi c Coast and the olive-backed inland group (C. u. swainsoni) (American Ornithologists’ Union [AOU] 1998, Evans Mack and Yong 2000). The two groups are most easily distinguished by diff erences in plumage characteristics, breeding and wintering location, and some vocalizations (Evans Mack and Yong 2000). Historical controversy over the taxonomic treatment of the ustulatus and swainsoni groups suggests that they have previously been considered on the border between subspecies and recently diverged sister species (reviewed in Bond 1963). Several authors emphasize the diff erences between the groups in breeding habitat and the lack of intergradation in regions where they co-occur (Bent 1949, Phillips 1991). Other authors emphasize their similarities, citing extensive intergradation along the eastern slope of the Sierra Nevada (Grinnell and Miller 1944). In the past decade, a number of studies have brought further clarity to the extent of genetic, behavioral, ecological, and acoustic divergence between the ustulatus and swainsoni subspecies groups. Here, I review these recent advances in our knowledge, identify future research questions, and discuss potential implications of divergence on the taxonomic treatment of the two groups according to the AOU’s guidelines for naming species (AOU 1998, Johnson et al. 1999). Received, accepted. Resumen.—Catharus ustulatus es un migrador de larga distacia, Neartico–Neotropical. Incluye dos grupos importantes de subespecies: en la costa del Pacífi co, el grupo de “espalda café-rojizo”, C. u. ustulatus, y, en el interior, el grupo de “espalda olivacea”, C. u. swainsoni (American Ornithologists’ Union 1998, Evans Mack and Yong 2000). Ambos grupos se diferencian fácilmente por las caracteristicas del plumaje, sitios de reproducción y área de invernada, además de vocalizaciones (Evans Mack and Yong 2000). La controversia sobre el tratamiento taxonómico de los grupos ustulatus y swainsoni, sugiere que han sido previamente considerados en el límite entre subespecies y especies hermanas que recientemente divergieron (revizado en Bond 1963). Varios autores enfatizan las diferencias entre los grupos en áreas de reproducción y la falta de intergradación en las regiones en que coexisten (Bent 1949, Phillips 1991). Otros autores subrayan sus similitudes, aludiendo además a una extensiva intergradación en la vertiente este de la Sierra Nevada (Grinnell and Miller 1944). En la década pasada, varios estudios clarifi caron los niveles de divergencia genética, de comportamiento, acústicas y ecológica, entre los grupos de subespecie ustulatus y swainsoni. En este estudio, reviso los trabajos anteriores identifi cando futuras preguntas de investigación, y discuto las implicaciones taxonómicas para ambos grupos de acuerdo con las reglas establecidas por la Unión de Ornitólogos Americanos (American Ornithologists’ Union 1998, Johnson et al. 1999). Evolutionary History Recent genetic data have helped shed light on the evolutionary history within the genus Catharus as well as on intraspecifi c variation within the Swainson’s Thrush (C. ustulatus). Two independently derived molecular phylogenies of the genus Catharus using diff erent loci suggest that the Swainson’s Thrush is the oldest lineage in the genus and is sister to all other species of Catharus, including several resident and short-distance migratory species from the Neotropics (Outlaw et al. 2003, Winker and 1E-mail: kruegg@berkeley.edu ORNITHOLOGICAL MONOGRAPHS NO. 63 68 Pruett 2006). The fi ve morphologically similar long-distance migratory species within the genus Catharus do not form a monophyletic clade, which indicates that migratory behavior has evolved multiple times within the genus and that the morphology of the North American long-distance migratory species exhibits strong evolutionary convergence (Winker and Pruett 2006). Outlaw et al. (2003) suggested that the Swainson’s Thrush diverged from other Catharus species ~4 mya, and the combined data sets indicate that species of Catharus most likely diverged somewhere in the Neotropics. An analysis of population structure within the Swainson’s Thrush (Ruegg and Smith 2002) using a rapidly evolving region of the mitochondrial genome, the control region, focused on genetic diff erentiation between the two major subspecies groups (Fig. 1): the russet-backed group of the Pacifi c Coast (C. u. ustulatus) and the olive-backed inland group (C. u. swainsoni) (American Ornithologists’ Union [AOU] 1998, Evans Mack and Yong 2000). These data showed that divergence between the subspecies groups is lower than reported levels of divergence between many sister species, but similar to estimates of divergence between well-diff erentiated subspecies of birds. Individuals of ustulatus and swainsoni from across the breeding range are separated by fi ve diagnostic mutations with a net sequence divergence of 0.7% (Fig. 2; Ruegg and Smith 2002). By comparison, Johnson and Cicero (2004) found that estimates of mitochondrial DNA (mtDNA) sequence divergence in 39 pairs of avian sister species ranged from 0.0% to 8.2%, with an average divergence of 1.9%. Alternatively, control-region divergence between some well-diff erentiated subspecies ranges from 0.07% to 4.8%: 0.07% between Brewer’s Sparrow (Spizella breweri breweri) and Timberline Sparrow (S. b. taverneri) subspecies Fig. 1. Breeding and wintering ranges of coastal ustulatus (black) and inland swainsoni (gray) subspecies groups and potential contact zones (black and white stripes) in Swainson’s Thrush (based on a range map from the Cornell Laboratory of Ornithology, with range data by Nature-Serve). DIVERGENCE IN SWAINSON’S THRUSH 69 (Klicka et al. 1999), 0.1% between Bluethroat (Luscinia svecica) subspecies (Questiau et al. 1998), and 1.7–4.8% between seven subspecies of Common Chaffi nch (Fringilla coelebs; Marshall and Baker 1998). Nuclear genetic distances between ustulatus and swainsoni populations from the Pacifi c Northwest of North America (two ustulatus, two swainsoni, and one mixed population) are lower than microsatellite genetic distances between other designated species but within the range of genetic distances between other subspecies and populations. Microsatellitebased genetic distances, in which an FST of 1 represents no gene fl ow and an FST of 0 represents complete mixing, ranged from 0.018 to 0.043 between pure ustulatus and swainsoni populations (Ruegg et al. 2006b). In comparison, FST-based estimates of genetic distance among some species of birds range from 0.054 (Shy Albatross [Thalassarche cauta] and Whitecapped Albatross [T. steadi]; Abbott and Double 2003) to 0.069 (fi ve species of gulls [Larus spp.]; Crochet et al. 2003), whereas FST-based genetic distances among some well-diff erentiated subspecies and populations range from 0.014 to 0.075 (0.014 for populations of Yellow Warblers [Dendroica petechia] from across North America [Gibbs et al. 2000]; 0.018 among populations of Chestnut-backed Chickadees [Poecile rufenscens; Burg et al. 2006]; and 0.075 among subspecies of North American Steller’s Jays [Cyanoci a stelleri; Burg et al. 2005]). In conclusion, mitochondrial and nuclear genetic divergence between the ustulatus and swainsoni subspecies groups is somewhat lower than estimates of genetic divergence between some other avian sister species, but is within the range of genetic divergence between other well-defi ned subspecies of birds. It is generally accepted that divergence between many well-diff erentiated subspecies and recently diverged sister species of birds can be att ributed to geographic isolation during the midto late Pleistocene (Avise and Walker 1998, Johnson and Cicero 2004, Lovett e 2005). Paleoecological data suggest that suitable songbird habitat was present in the east and west, south of the glacier’s edge, and that the center of the country was occupied by tundra and desert that likely would have been inhospitable to a songbird (Pielou 1991). Genetic data combined with climate models of the distribution of the ustulatus and swainsoni subspecies groups at the last glacial maximum (LGM) are concordant with paleoecological data and suggest that the most likely distribution of populations at the LGM would have been western coastal and southeastern regions (Ruegg et al. 2006a; Fig. 3). In light of the hypothesized distribution at the LGM, regions of parapatry between the ustulatus and swainsoni subspecies groups likely represent regions of secondary contact following postglacial range expansions. Fig. 2. Summary of genetic analysis based on mtDNA control-region haplotypes from Ruegg and Smith (2002). (A) Frequency of coastal ustulatus (white circles) and inland swainsoni (black circles) haplotypes in 17 breeding populations. (B) Haplotype network with bars across branches indicating single nucleotide changes. Four sizes of circle are used to represent the number of individuals sharing the same haplotype (smallest, 1 individual; medium-small, 2–4 individuals; medium-large, 5–11 individuals; largest, 44 individuals). ORNITHOLOGICAL MONOGRAPHS NO. 63 70