Two-way splits are all the rage these days: Sagebrush and Bell’s sparrows; Winter and Pacific wrens; Common Moorhen and Common Gallinule; Snowy and Kentish plovers; Yellow-throated and Bahama warblers; and a bunch more. Three-way splits are rarer, but they do happen: Think of the three-way splits in the 1990s of the scrub-jays and of the “Solitary” vireos.
How about a five-way split of two species? That’s been proposed by James Maley and Robb Brumfield for the King Rail–Clapper Rail complex. In his column in the print version of the September/October 2013 Birding (“Rails Relations Revisited,” pp. 25–26) Paul Hess, writing with his trademark clarity and succinctness, sums up the Maley–Brumfield proposal. The five species in Maley and Brumfield’s formulation would be: Ridgway’s Rail (Rallus obsoletus) of western North America; Clapper Rail (R. crepitans) of eastern coastal North America; Mangrove Rail (R. longirostris), currently classified as six subspecies of the Clapper Rail, of South America; Aztec Rail (R. tenuirostris), currently classified as a King Rail subspecies but apparently more closely related to Ridgway’s Rail, endemic to highland freshwater marshes in central Mexico; and King Rail (R. elegans) of the eastern U.S., southern Ontario, and Cuba.
Focusing for a moment just on North America, we can say with certainty that the diverse taxa in the King Rail–Clapper Rail complex exhibit great variation in plumage. That’s the takeaway from the photo salon appearing on pp. 28–39 of the September/October Birding. It’s hard to imagine that the ghostly birds photographed by Corey Finger (Brooklyn, New York) and Diana Doyle (Hampton, Virginia) are the same species as the ruddy birds photographed by Naomi Blinick (Sonora) and Marco Antonio González Bernal (Sinaloa). Which, of course, is the whole point of the Maley–Brumfield proposal: They’re not the same species.
So we have Paul Hess’s summary and ABA members’ photos, but what are the actual differences, taxon by taxon, and perhaps species by species? This is a good news–bad news scenario. The bad news is that the situation is fantastically complex, involving far too many taxa and literature citations for a proper overview on the print pages of Birding. The good news—and it far outweighs the bad news—is that all of this complexity has been brilliantly and authoritatively distilled in the Clapper Rail account in Birds of North America Online (BNA Online).
BNA Online Series Editor Alan Poole and Taxonomy Editor Michael Patten have graciously given us permission to reproduce the entire systematics account from William R. Eddleman and Courtney J. Conway’s monograph on the Clapper Rail. We’ll get to that in a moment, but, first a few words about BNA Online.
I, personally, can’t imagine birding without BNA Online. I visit the site at least weekly. I’m sure I average out to more than once a day. And I’ve been doing so for the better part of the past decade. The literature citations alone would justify the entire series. The sections on “Systematics,” “Sounds,” and “Molts” are favorites, but I frequently find myself wandering into the realm of “Locomotion” and “Diet,” even “Agonistic Behavior” and “Body Parasites.” It’s all in there.
The print version of Birds of North America takes up two long bookshelves—and it doesn’t begin to compare with the immense amounts of information, including audio and video, in BNA Online. For the exceedingly modest rate of $42 per year (less, prorated, if you subscribe for several years at a time), you can have full access to this incomparable reference to the birds of North America. Get started here:
Back to the Clapper Rail–King Rail complex. Here is Michael Patten’s overview, annotated with photos from the photo salon in the current issue of Birding:
by Michael Patten
[with very minor editing for style and format. —Ted Floyd]
Plumage coloration and body size and proportions vary considerably across this species’ vast geographic range, a range that includes much of North America and n. South America. Overall coloration is darker and more richly colored from north to south along the coasts of the e. United States, w. Mexico, and Baja California peninsula (Ripley 1977). Plumage on the East Coast of the United States and in the Caribbean varies from gray to dull gray brown or buff, with grayer plumage to the north (Oberholser 1937, Ripley 1977), but plumage in w. North America, including in freshwater habitats, is buff to rich cinnamon. In general, populations from e. Mexico, Central America, and the n. Caribbean are paler than those farther north (Ripley 1977), although it may be that insular populations in the w. Gulf of Mexico and s. Caribbean tend to be darker than those on the nearest mainland. Some populations are dimorphic, having both a morph less saturated with color and a morph more heavily pigmented. Individuals from northern and western populations average larger in all measurements except culmen length, which is shortest on the West Coast and in s. North America, heaviest South America, and most slender in e. North America and the Caribbean (Olson 1997). Wing chord declines clinally in sw. North America, from se. California and sw. Arizona south to Nayarit (Ripley 1977).
Twenty-one subspecies, following Ripley (1977) but with due consideration to Banks and Tomlinson (1974) and skepticism toward two insular subspecies off the Yucatan peninsula known in each case solely from the single type specimen. With regard to the insular taxon named from Belize (formerly British Honduras), Russell (1964:57) wisely noted that “Clapper Rails exhibit much individual variation, and a series will be needed to establish the taxonomic status of coastal and insular birds in the colony.” Retained subspecies differ in dorsal color and pattern, ventral color, body size, and bill length. Marked individual variation and the existence of color morphs help to explain why authorities varied for years with respect to the number of subspecies; moreover, there may be considerable overlap in plumage color and measurements where subspecies distributions overlap (Heard 1983). In the end, it may be impossible to determine the true number of subspecies until two issues are settled: (1) intrepid taxonomists need to analyze specimens amassed across the entirety of the species’ range—piecemeal attempts will go only so far—and (2) systematists need to determine if R. longirostris andR. elegans (the King Rail) are indeed different biological species . . . and if they are to specify exactly which subspecies belong to which parent species.
Oberholser (1937) was the last taxonomist to attempt a range-wide analysis of geographic variation and subspecies limits on the basis of primary data (i.e., museum specimens) rather than on secondary data (i.e., literature). For all of his powers of observation, Oberholser had a marked and notorious tendency to over-split. It is thus likely that, as has been the case in plenty of other cases (see Browning 1978), many of his taxa will prove to be invalid. The most recent literature-based revision was Ripley’s (1977), and he ignored some analyses, such as that of Banks and Tomlinson (1974), that would have changed his conclusions. The lack of consensus is exemplified by the wide range of subspecies of R. longirostris listed in standards references, with a mere 11 in Peters (1934) to 21 in Ripley (1977)—whose taxonomy Taylor (1996) followed exactly—to a staggering 25 in Ridgway and Friedmann (1941).
A more vexing issue, one that may need to be tackled first, is the question of species limits between R. longirostris and R. elegans (see Related Species, below). Taxonomic opinion has been divided for decades. Peters (1934), in his monumental Checklist of the Birds of the World, recognized two species but considered various taxa to be subspecies of R. elegans that are now almost universally classified as subspecies of R. longirostris. Ripley (1977) solved all such problems by lumping the two species completely under the name R. longirostris. More recently, Olson (1997) argued for a recognition of species limits nearer to Peters (1934) treatment, with all subspecies of R. longirostris in w. North America reallocated to R. elegans, and a comprehensive survey of mitochondrial DNA argues for breaking the complex into five separate species (Maley and Brumfield ms.). Even so, mtDNA diverges little among subspecies in the obsoletus and creptians groups, with coastal and inland subspecies in the former group having a common ancestor <50,000 years ago (Fleischer et al. 1995).
Obsoletus Group (Sw. United States And W. Mexico):
R. l. obsoletus Ridgway, 1874. Resident in coastal salt marshes of California from Humboldt Bay south to Monterey Bay [type locality = San Francisco], although now extirpated from much of its former range. Cheeks brownish; flanks blackish with bold white barring; breast ochre-buff; dorsal feathers edged olive-gray; large (wing chord > 140 mm).
R. l. levipes Bangs, 1899. Resident in coastal salt marshes from Santa Barbara, California, south to at least Estero de Punta Banda near Ensenada, Baja California [type locality = Newport Bay, Orange Co., California]. Similar to R. l. obsoletus but breast pinkish cinnamon, dorsum darker, and body size averages smaller.
R. l. beldingi Ridgway, 1882. Includes R. l. magdalenae van Rossem, 1947. Coastal salt marshes and mangroves on both coasts of the Baja California peninsula, from San Quintín southward on the Pacific Coast and I. Espíritu Santo on the Gulf Coast [type locality = I. Espíritu Santo, Baja California Sur]. Similar to R. l. levipes but breast rich hazel and dorsum paler.
R. l. yumanensis Dickey, 1923. Resident in freshwater, brackish, and salt marshes, both inland and coastal, from se. California and sw. Arizona (Salton Sea and lower Colorado R. valley) south through ne. Baja California and nw. Sonora to the Colorado R. delta [type locality = Bard, Imperial Co., California]. Similar to R. l. levipes but cheeks ashy gray and dorsal feathers paler brown and edged grayish; wing pointed (Banks and Tomlinson 1974).
R. l. rhizophorae Dickey, 1930. Resident in coastal salt marshes in nw. Mexico from central Sonora south to central Sinaloa [type locality = Tóbari Bay, Sonora]. Like R. l. yumanensis but pale gray overall and wing rounded (Banks and Tomlinson 1974).
R. l. nayaritensis McLellan, 1927. Resident in coastal salt marshes and mangroves from central Sinaloa to s. Nayarit [type locality = San Blas, Nayarit]. Like R. l. rhizophorae but dark gray overall (Banks and Tomlinson 1974).
Crepitans Group (E. United States, Caribbean, and Yucatan Peninsula):
R. l. crepitans Gmelin, 1789. Breeds in coastal salt marshes from Massachusetts to s. North Carolina [type locality = Long Island, New York]; winters from breeding range, although typically from New Jersey, south to ne. Florida. Cheeks gray; breast pale cinnamon; dorsum pale brown with feathers edged olive-gray; white bars on flanks broad; bill long (> 55 mm) and relatively slender.
R. l. saturatus Ridgway 1880. Resident along the Gulf Coast from Florida Panhandle west to Texas and south to Tamaulipas, Mexico [type locality = Rigolets, Lake Borgne, Louisiana]. Similar to gray morph of R. l. scottii but browner overall and breast orange-cinnamon.
R. l. insularum Brooks, 1920. Includes R. l. helius Oberholser, 1920. Resident in salt marshes and mangroves of the Florida Keys from Key Largo to Boca Grande [type locality = Big Pine Key] and, presumably, the Marquesas Keys (Owre 1978). Similar to R. l. waynei but with the bill short (< 60 mm), paler overall, and dorsal feather edges broader and purer gray.
R. l. coryi Maynard, 1887. Includes R. l. corrius Oberholser, 1937. Resident in mangrove swamps of The Bahamas [type locality = island off s. shore of Andros]. Like R. l. insularum but breast pinkish buff and dorsum pale buff-brown.
R. l. caribaeus Ridgway, 1880. Includes (see Peters 1934, Ripley, 1977), R. l. cubanusChapman, 1892; R. l. leucophaeus Todd, 1913; R. l. vafer Wetmore, 1928, R. l. manglecola Danforth, 1934; R. l. limnetis Oberholser, 1937. Resident in mangroves of the Greater Antilles and the Virgin Is., with scattered records in the n. Lesser Antilles [type locality = vic. Spanish Town, Jamaica]. Dimorphic, with a pale morph and a dark morph; pale morph like R. l. coryi but dorsal feathers fuscous.
R. l. pallidus Nelson, 1905. Includes R. l. grossi Paynter, 1950, is “known only from the type specimen” (Paynter 1955), and R. l. belizensis Oberholser, 1937, of which “the one specimen from British Honduras is the type” (Russell 1964). Resident in coastal marshes, mangroves, and cayes of the Yucatan Peninsula south to central Belize [type locality = Río Lagartos, Yucatán]. Like R. l. saturatus but breast less orange, dorsum paler brown, and bill short (< 55 mm).
Longirostris Group (N. South America):
R. l. cypereti Taczanowski, 1877. Resident coastally in Ecuador and nw. Peru [type locality = Santa Luzia, Tumbes delta, Perú]. Cheeks brown; breast ochre-buff; dorsal feathers edged olive; white bars on flanks narrow; wing short (< 130 mm); bill short (< 55 mm) and stout.
R. l. phelpsi Wetmore, 1941. Resident along n. coast of Colombia and nw. Venezuela [type locality = Laguna de Tucacas, Puerto López, La Guajira, Colombia]. Size and shape of R. l. cypereti but cheeks gray, breast pinkish cinnamon, and dorsal feathers edged brown.
R. l. margaritae Zimmer and Phelps, 1944. Resident on I. Margarita [type locality = Boca del Río, I. Margarit, Venezuela]. Like R. l. cypereti but breast warm hazel and dorsum darker.
R. l. pelodramus Oberholser, 1937. Resident in coastal marshes of Trindad. Similar to R. l. phelpsi but breast purer cinnamon, white bars on flanks wide, and dorsal feathers blackish with rufous edges.
R. l. longirostris Boddaert, 1783. Resident coastally in the Guianas [type locality = Cayenne, French Guiana]. Like R. l. pelodramus but dorsal feathers paler and brown.
R. l. crassirostris Lawrence, 1871. Resident in coastal salt marshes of e. Brazil, from Maranhão south to s. São Paulo [type locality = Bahia, Brazil]. Like R. l. pelodramus but dorsal feathers edged olive.
The avian order Gruiformes, which includes the rails, coots, and gallinules (Rallidae), long included a host of oddball taxa, many in their own taxonomic families but all united by a distinctive morphology of the leg bones. Yet the traditional Gruiformes apparently are not monophyletic, although the Rallidae are clearly part of a well-supported core clade of gruiform birds (Fain et al. 2007, Hackett et al. 2008). Within the order, the Rallidae is sister to the Heliornithidae, the tiny family of finfoots that includes a mere three species (Fain et al. 2007). Without an extensive phylogeny of species within the Rallidae it is difficult to say exactly where the genus Rallus falls relative to other genera, yet on the basis of morphology the nearest relatives are likely to be Gallirallus and Pardirallus, each of which has been merged into Rallus at various times (Ripley 1977, Taylor 1996).
It is clear that R. longirostris is sister to R. elegans, although even this statement assumes the two are good biological species (e.g., Dickerman 1971, Ripley 1977). Early studies of mitochrondrial DNA variation are inconclusive with regard to species limits (Avise and Zink 1988). If separate species are involved, these two taxa, with R. wetmorei, the Plain-flanked Rail of coastal n. Venezuela, form a superspecies (Sibley and Monroe 1990). But it is perhaps debatable whether R. wetmorei is itself a species, with some arguing it be treated as a distinctive subspecies of R. longirostris (e.g., Meyer de Schauensee 1966). Blake (1977) stated that R. wetmorei occurred sympatrically with R. longirostris in Carabobo, Venezuela, which would seem to settle the matter. . . yet Olson (1997) intriguingly posited that R. wetmorei may not be a taxon at all but is rather a well-marked color morph, one that simply lacks black-and-white barring on the flanks (it is structurally identical to adjacent subspecies of R. longirostris).
Putting R. wetmorei aside, and under the assumption that R. longirostris and R. elegansare biological species, there remains the question of which subspecies to allocate to which species. For example, although subspecies in the obsoletus group are nowadays typically classified with R. longirostris, they have been classified with R. elegans (e.g., Peters 1934, Hellmayr and Conover 1942). By contrast, R. elegans tenuirostris Ridgway, 1874, of central Mexico has been grouped with R. longirostris (e.g., Oberholser 1937, Williams 1989). Species limits may even exceed the traditional two taxa of R. longirostrisand R. elegans. For example, Bent (1926) treated the obsoletus group as a separate species. And a recent comprehensive survey of mitochondrial DNA argues for breaking the complex into five separate species (Maley and Brumfield ms.): R. elegans of the e. United States and Cuba; R. crepitans of the Atlantic seaboard, Gulf Coast, Greater Antilles, and Yucatan peninsula; R. obsoletus of w. North America; R. tenuirostris of central Mexico; and R. longirostris of South America. Only time will tell if this taxonomy is found acceptable by the American Ornithologists’ Union’s checklist committees for North America and for South America.
Regardless, R. longirostris, as currently constituted, hybridizes with R. elegans in brackish marshes where geographic ranges overlap (Meanley and Wetherbee 1962, Meanley 1969, Bledsoe 1988, Olson 1997). Hybrids exhibit a variety of plumage intermediacy (Meanley 1969), and gene flow may partly explain the color morphs of someR. longirostris subspecies.