Siberian (A. r. japonicus) versus American Pipits (A. r. rubescens, pacificus, alticola) in basic plumage

Cin-Ty Lee

Plates by: Andy Birch & Brian J. Small

Photographs by: Toru Akiba, Don Desjardin, Jim Gain, Kim Hyun-Tae, Takashi Koike, Scott Robinson, Tetsu Sato, Jim Stasz.

Contents:

Page 1 of Article

Page 2 of Article click here

Photo Page 1 (Siberian Pipits) click here

Photo Page 2 (American Pipits) click here

Plate of Rock Pipits by Brian J. Small click here

Plate of Water and Buff-bellied Pipits by Brian J. Small click here

Plate by Andy Birch click here

1. Introduction
Four subspecies of Buff-bellied Pipits are currently recognized by the American Ornithologist’s Union: japonicus, pacificus, alticola, and rubescens. The first subspecies, japonicus, may be a candidate for species status according to the American Ornithologist’s Union (AOU; 1989) because it breeds in eastern Asia and is considered to be significantly different, in terms of plumage characters, from the latter three, which breed entirely within North America and western Greenland. Japonicus is herein referred to as the "Siberian" Pipit and the latter subspecies, which breed and winter in North America, are known as the "American" pipits. The Siberian Pipit is a very rare vagrant to Europe, the Middle East, and along the Pacific Coast of North America. The American Pipits are vagrants to Western Europe (rubescens) and possibly even rare winterers to eastern Asia (pacificus). In detail, however, the vagrancy patterns of Siberian and American Pipits are poorly known.

There is also somewhat of a gap in the literature regarding the subspecific identification of Buff-bellied Pipits. To our knowledge, the only article devoted to the Buff-bellied Pipit complex is Parkes (1982), in which subspecies identification in alternate (breeding) plumage was discussed, but neither the identification in basic (winter) plumage nor the identification of Siberian pipits was discussed. Alström and Mild (1996), Lewington et al. (1991), and Beaman and Madge (1998) discussed the identification of Siberian Pipit in relation to the rubescens subspecies of the American Pipits, but did not address the high degree of subspecific plumage variation within the American Pipit group.

While we originally set out to write a definitive identification article on all four subspecies, we soon discovered from our field and museum studies that plumage variation within the American Pipit group was too complex for subspecific identification at a reasonable level of confidence. As a first step, towards our original goal, this manuscript focuses on separating Siberian from American Pipits in basic (winter) plumage. We also present a preliminary guideline for subspecific separation of the American Pipit complex. Our research on alternate (breeding) plumage identification is still in progress and will be presented in the future.

2. Systematics and nomenclature
Knox (1988) and Alström and Mild (1996) have already reviewed the history of the taxonomic split of Water Pipit (A. spinoletta) into Buff-bellied Pipit, Rock, and Water Pipit, but a brief review is in order here. Vaurie (1959), Hall (1961) and Williamson (1965) split the species into two ecological groups, the "water pipits", which breed in mountainous regions in Europe, Asia, and North America, and the "rock pipits", which breed along rocky coastlines in northwest Europe. These two groups were considered to be separate species by Bannerman (1953) and Oreel (1980). Nazarenko (1978) showed that two subspecies of the "water" pipit, blakistoni and japonicus, overlapped in terms of their breeding range in central Asia, but that they preferred different habitats. As a result of this study, Glutz (1985) divided the "water" pipit into two species, the Water Pipit and the Buff-bellied Pipit. Devillers (1980), British Ornithologist’s Union Records Committee (1986), Alström and Mild (1987) and Knox (1988) suggested that the entire complex be divided into three species, Rock (A. petrosus), Water (A. spinoletta), and Buff-bellied Pipit (A. rubescens). Alström and Mild (1996) showed that Rock, Water, and Buff-bellied Pipits can be distinguished from each other based on plumage characters, regardless of one’s opinions regarding their species status. The American Ornithologist’s Union (1989) currently recognizes the latter taxonomy, so this is the taxonomy followed here.

Within each species, there exists considerable geographic and subspecific variation. Rock Pipit consists of nominate petrosus, which breeds in Ireland, Britain, and northwest France, and littoralis, which breeds in Fennoscandia and northwest Russia (Cramp, 1988). Williamson (1965) also recognized two other distinct Rock Pipit races worthy of subspecific status: "kleinschmidti" (breeding on the Faroe Islands and possibly on the Scottish outlying islands) and "meinertzhageni" (breeding on the Outer Hebrides). However, the British Ornithologist’s Union Records Committee (1986) does not formally recognize their subspecific status. The Water Pipit consists of three subspecies: spinoletta, which breeds in southern and central Europe; coutellii, which breeds in Asia Minor; and blakistoni, which breeds in Central Asia. We will discuss the subspecific distribution and breakdown of the Buff-bellied Pipit complex below .

Our choice of "common" names for these subspecies is arbitrary. We choose to call japonicus the Siberian pipit, rather than "Japanese" Pipit, because it’s breeding range (see below) is restricted to Siberia. In addition, the name Siberian Pipit has been applied to this subspecies before. We refer to the remaining subspecies of the Buff-bellied Pipit complex as American Pipits because their breeding and wintering range is confined primarily to North America. We use "Buff-bellied Pipit" to describe all four subspecies as a whole.

3. Distribution of Buff-bellied Pipit subspecies
3.1 Breeding range
Below, we discuss the current understanding of the breeding and wintering ranges of the four subspecies of the Buff-bellied Pipit. We proceed with the cautionary note that the exact boundaries between breeding ranges within the American group are probably very uncertain given the difficulty in subspecific identification.
The Siberian Pipit A.r. japonicus, breeds in central and western Siberia from Tunguska to Kamchatka and south to northern Sakhalin and the Kurile Islands (American Ornithologist’s Union, 1989). Breeding has not yet been documented in Japan (Brazil, 1991). It is not thought to breed in Alaska (American Ornithologist’s Union, 1991).

Pacificus breeds in the Pacific Cordillera from the mountains of Oregon north to the alpine and arctic tundra of Alaska (Miller and Green, 1987; Campbell et al., 1997), including the Aleutian (Gabrielson and Lincoln, 1959) and Pribolof Islands (Thompson and Delong, 1969). Overall, the breeding range is primarily believed to be west of the Rocky Mountains.

Alticola, sometimes known as the "Rocky Mountain" race", breeds throughout the Rocky Mountains and outlying ranges from southern British Columbia and Montana, south to New Mexico and Arizona, and west to the eastern Great Basin Ranges (Miller and Green, 1987). In the southern part of its range, it has bred in the White Mountains, AZ and the Sangre de Cristo Range, NM (Hubbard, 1978). More recently, it has been discovered as a breeder in California, occurring in the Sierra Nevada as far north as Mono County and as far south as Tulare County (Miller and Green, 1987). Isolated breeding pairs have also been reported from the summit of Mount San Gorgonio in the San Bernardino Mountains, California (McCaskie, 1978; Miller and Green, 1987; Miller, 1988). Breeding is also suspected above timberline in the White Mountains and on nearby Telescope Peak in the Panamint Mountains (Small, 1994). Howell and Webb (1992) reported several birds, presumably of this subspecies, at 2,450 m in the Sierra San Pedro Martir, Baja California, including a male who made repeated song flights during June, which is well past the time of spring migration.

Rubescens breeds from northern Yukon east to western Greenland, and south to southwestern Yukon, northern British Columbia, Northwest Territory, northern Manitoba, northern Ontario, northern Quebec, southern Labrador and Newfoundland (American Ornithologist’s Union, 1983). Isolated populations breed at higher elevations on Gaspe Peninsula and on Mount Katahdin, Maine (American Ornithologist’s Union, 1983) and Mount Washington, New Hampshire (Petersen, 1991; Veit and Petersen, 1993). A sub-population, along the Hudson Bay, was considered by Oberholser (1974) to be a separate subspecies (A. r. ludovicianus). However, this population is essentially identical to rubescens and has never been accepted by the AOU.

A fifth subspecies, A. r. geophilus, was proposed by Oberholser (1974) based on slightly different plumage characteristics. However, this subspecies has also never been recognized by the AOU because it is only marginally distinct from pacificus. Oberholser described this subspecies as breeding primarily in coastal southern Alaska. We have no field experience with geophilus, and it is poorly represented in specimen collections. We combine our discussion of geophilus with pacificus, its closest counterpart in terms of plumage.

3.2 Wintering Range
The Siberian Pipit winters in eastern China and Japan (Honshu) (Gabrielson and Lincoln, 1959; AOU, 1983; Brazil, 1991), Nepal (Inskipp and Inskipp, 1985), Pakistan (Grimmett et al., 1999), northern India, northern Burma, northern Vietnam and southern China (AOU, 1983), Hong Kong (King and Dickenson, 1975), Taiwan (Chang, 1980), and South Korea (Gore and Won, 1971). It has been recorded as a vagrant as far south as Myanmar (Burma), NE Thailand and western Tonkin, and may be regular enough in Israel to be considered as winterers (cf 3.3.2).

The American Pipits winter in largest numbers along coastlines, grasslands, and farmlands in California, Arizona, New Mexico, and Texas, south through Baja California and Mexico to Guatamela (AOU, 1983; Howell and Webb, 1995), and east through the Gulf states to the Atlantic coastal plain (Root, 1988). Small numbers of wintering birds are found as far north as southern British Columbia, northern Great Basin, southern New England and rarely in the Great Plains (Root, 1988; Veit and Petersen, 1993; Campbell et al. 1997).

Unfortunately, the breakdown of the wintering range by subspecies is poorly demarcated at present, primarily because of the lack of knowledge in identifying these subspecies. As such, the foregoing should be considered tentative. In general, it is believed that rubescens accounts for most of the eastern wintering populations (Oberholser, 1974). The AOU (1957) describes rubescens as wintering from "Texas, Arkansas, Tennessee, West Virginia (Upshur County), and the lower Delware Valley south through eastern Mexcio to Guatemela, the Gulf coast, and Florida." Oberholser (1972) stated that this subspecies was a common winterer in north-central Texas and as far west as El Paso. Pulich (1988) confirmed that at least some of the specimens taken in Texas were indeed of this subspecies.

It is believed that pacificus accounts for most of the Pacific Coast populations (Grinnell and Miller, 1944), with the AOU (1983) stating that wintering occurs from "southern coastal British Columbia, Oregon, west-central Nevada (Lahontan Valley), and southern Utah to Baja California and western Mexico (south to Oaxaca)." Monson and Phillips (1981) were of the opinion that all wintering pipits in Arizona were of this subspecies. Oberholser (1972) claimed that pacificus is a fairly common winterer in Texas, although Pulich (1988) recommended that its status in Texas be reinvestigated. Exactly how far east pacificus ranges during winter is unknown, but considering the general lack of reports east of Texas, it is probable that Texas is the easternmost limit of regularly wintering pacificus.

According to the AOU (1983), the wintering range of alticola is largely unknown, but it has been recorded in December from Arizona. Monson and Phillips (1981), however, stated that all winterers in Arizona are pacificus. Grinnell and Miller (1944) described alticola as a rare winter visitant to California, but did not elaborate. Their account includes only specimens collected from early to mid-April. Oberholser (1974) indicated that this subspecies occurred in Texas on 22 March and 10 April. However, these dates coincide with the spring migration (see below), and thus do not necessarily imply that wintering occurs in Texas. Regardless of the origin, Pulich (1988) was unable to relocate these specimens to verify the report.

3.3 Timing of migration and vagrancy
The migratory statuses of the Siberian and American Pipits are described here. The latter is described as whole because so little is known about the distribution and migratory status of the American subspecies. Roughly speaking, however, pacificus probably pertains to birds in western North America while rubescens pertains to those in eastern North America.

3.3.1 American Pipits in western North America
In western North America, American Pipits are on the move from their breeding grounds by late August in Alaska and have mostly departed by early September. Peak movements occur in the northern part of British Columbia during early September and from late September to early October in the southern part of the province (Campbell et al., 1997). Interestingly, fall transients are about eight times more numerous along the coast than in the interior in British Columbia. Migration is largely over in British Columbia by the end of October, with stragglers continuing into early November (Campbell et al., 1997). Wintering birds arrive in California and Arizona by middle to late September, with numbers peaking in middle October (Grinnell and Miller, 1944; Garrett and Dunn, 1981; Roberson, 1984; Small, 1994). Arrivals as early as the first week of September are considered exceptional (Rosenberg et al., 1991; Small, 1994), with arrivals in late August even more so (Grinnell and Miller, 1944). A single record on July 1 from Lassen Peak in northern California was regarded by Grinnell and Miller (1944) as an early fall transient, but may have been an individual trying to summer, considering that breeding colonies exist in the Oregon Cascades not far to the north.

Winterers in the southern part of their range (e.g., southern California) typically stay until mid-April, with a few lingering until early May (Rosenberg et al., 1991). Small (1994) states that spring transients in California occur from mid-April into May. However, a sharp increase in the number of birds between late March and April in British Columbia (Campbell et al., 1997) suggests that wintering birds are on the move well-before mid-April, and in all likelihood, the spring migration is probably quite protracted. In British Columbia, Campbell et al. (1997) noted that the total number of records of spring transients is lower than the number of fall transients by a factor of roughly three. Breeding individuals arrive by the last week of April in southeastern Alaska and by the first week of May in northern Alaska (Gabrielson and Lincoln, 1959).

The local populations (most likely alticola; Miller and Green, 1987) that breed in the Sierra Nevada and San Bernardino mountains in California arrive on breeding grounds between mid-April and early May (Miller, 1988). These stay until late September, but occasionally into November depending upon the weather conditions (see references in Small, 1994).

American Pipits in eastern and central North America
Individuals are on the move from their breeding grounds in northern Canada by late August as evidenced by arrivals of fall transients in Ontario between the first week of September through October (Speirs, 1985). Sadler and Myres (1976) noted movements during late August in Alberta. In the southern part of Ontario, the earliest arrivals occur in early September and peak during mid- to late October (Speirs, 1985). Interestingly, in New York, earliest arrivals are 2 August and 13 August in inland and coastal counties, respectively (Bull, 1974); numbers of transients peak in October. In Alberta, transients pass through Edmonton during the last week of September and have completely passed through by early October (Sadler and Myres, 1976). In Minnesota, earliest arrivals appear by mid-September, and in Missouri by mid-September, peaking in early to middle October (Robbins and Easterla, 1992). Fall transients pass through Massachusetts primarily during October (Veit and Petersen, 1993), and through Cape May, New Jersey from October through November (Stone, 1965; Sibley, 1997). Winterers arrive in Tennessee by early October, with exceptionally early arrivals occuring in September and even late August (Robinson, 1990); in Arkansas by late September, peaking in October and November (James, 1986); in Louisiana during October (Lowery, 1955); in Alabama by early October (Imhof, 1962); in Florida by October (Kale and Maeher, 1990); and in Texas by late September (Pulich, 1988). Fall transients apparently pass through by the end of November, with stragglers continuing into December through January. Small wintering flocks are occasionally encountered in December or January in coastal New York (e.g., Jamaica Bay, Bull, 1974), along the shores of Lake Erie (Speirs, 1985), and at Cape May, New Jersey (Stone, 1965; Sibley, 1997). Wintering individuals have been recorded as late as February in coastal Massachusetts (Veit and Petersen, 1993) and on Christmas Bird Counts in the northern Great Plains (e.g., Missouri; Robbins and Easterla, 1992).

Northward migration during spring is probably protracted. In eastern and central North America, pipits have largely departed their wintering grounds by early May (e.g., Alabama, Imhof, 1962; Texas, Pulich, 1988; Tennessee, Robinson, 1990). However, spring transients are clearly on the move by March throughout much of the interior United States as exemplified by peak movements during March and early April in Arkansas (James, 1986), late March to early April in New York (Bull, 1974), and mid-March in Missouri (Robbins and Easterla, 1992). At Point Pelee, Ontario spring transients pass through between late March and late May (Speirs, 1985). In Massachusetts, spring transients pass through from late March to mid-May, with peak counts occurring in April (Veit and Petersen, 1993). A curious observation is that spring transients are extremely rare at Cape May, New Jersey (Stone, 1965) and considerably less numerous than fall transients in Massachusetts (Veit and Petersen, 1993). In Alberta, spring transients pass through during the first week of April (Sadler and Myres, 1976). We were unable to find information regarding the arrival of breeders in northernmost Canada, but based on arrival dates in northern Ontario around early May (Speirs, 1985), it is likely that breeders arrive by mid-May. In the southern part of its breeding range, for instance at Guanella Pass in Colorado (alticola), they arrive during late April to early May (Conry, 1978).

American Pipits: vagrancy
Published extralimital occurrences of the American subspecies are few. Grinnell and Miller (1944) documented four specimens of alticola taken separately during the first two weeks of April in California. Bull (1974) described a specimen believed to be alticola, which was collected in Suffolk County in New York on May 10, 1882 (American Museum of Natural History 25964). Grinnell and Miller (1944) reported no occurrences of rubescens in California, but Rosenberg et al. (1991) suggest that one specimen collected on 22 December, 1902 from Yuma, Arizona (near California border) might pertain to this subspecies. Extralimital occurrences of Buff-bellied Pipits in Bermuda probably pertain to this subspecies (AOU, 1973). Vagrant rubescens have been reported in Germany, Italy, Iceland, Ireland, and Britain primarily between the dates of mid-September and late October (see L.G.R Evans, 1994 for review). At least one, if not both, of the Italy sightings may in fact be Siberian Pipits (Shirihai and Colston, 1987). Two records of Buff-bellied Pipit exist for Scandinavia (one in Norway and one in Sweden), both of which occurred between October and January. The Scandinavian pipits were identified as rubescens, however some features may suggest Siberian (Brian Small, pers. comm.). The western North American subspecies pacificus, has been recorded in Okinawa, Japan during the winters of 1982/83 and 1984/85 and from January to February, 1987 (see Brazil, 1991 for references).

There is no doubt that rubescens is a vagrant to Europe. However, the above compilation suggests that pacificus and alticola may also exhibit vagrancy patterns, insofar as the above records represent accurate identifications. In addition, one should not assume that rubescens does not appear on the Pacific coast during migration or even during winter.