Birding In Taiwan

SPOTLIGHT ON TAIWAN

Endemic subspecies of Taiwan birds— first impressions

N. J. COLLAR

Taiwan possesses 15 endemic bird species that every birder wants to get during the course of a visit. These are (in order of their scientific description, with date provided) Taiwan Whistling Thrush Myophonus insularis (1862), Swinhoe’s Pheasant Lophura swinhoii (1863), Taiwan Blue Magpie Urocissa caerulea (1863), Taiwan Partridge Arborophila crudigularis (1864), White-eared Sibia Heterophasia auricularis (1864), Steere’s or Taiwan Liocichla Liocichla steerii (1877), Styan’s or Taiwan Bulbul Pycnonotus taivanus (1893), Yellow or Taiwan Tit Parus holsti (1894), Mikado Pheasant Syrmaticus mikado (1906), Collared Bush Robin Tarsiger johnstoniae (1906), Flamecrest Regulus goodfellowi (1906), Whitewhiskered Laughingthrush Garrulax morrisonianus (1906), Taiwan Barwing Actinodura morrisoniana (1906), Taiwan Yuhina Yuhina brunneiceps (1906) and Taiwan Bush Warbler Bradypterus alishanensis (2000). You can fairly well gauge the elevations at which these things occur—and hence the order in which they are likely to be ticked off— by the dates of their description, starting with Robert Swinhoe’s multiple gleanings from the lowlands in the 1860s (his pheasant must have been a captive bird brought to him) and ending (almost) with the climactic six-species achievement of Walter Goodfellow forty years later as he reached the summit of Yu Shan, then Mt Morrison (see Ogilvie-Grant 1906). Most of these species are extremely distinctive, although none is in its own genus (the Yellow Tit has its own subgenus Machlolophus); only the whistling thrush, Flamecrest, bulbul and bush warbler appear to be very close to mainland Chinese forms. But does the taxonomic distinctiveness of Taiwan end with these species? Taiwan possesses a good many endemic subspecies as well, but how distinctive are they? Which among them might be eligible for upgrading to species level in the event of a new taxonomic revision? While it is always difficult to fix a complete inventory, owing to different judgements about the validity of many taxa and the extent to which they figure in modern treatments, a very preliminary review of the island’s endemic subspecies (78 in this compilation —all, I hope, of those currently or recently claimed) is attempted here, based on a list originally compiled by Brian Sykes and a three-day review of material at the Natural History Museum (NHM) in Tring, U.K. In this short exploratory survey, I treat each of these taxa in turn. I have sought for and, where found, provided diagnostic characterisation of each subspecies from the standard or original sources, and deal very briefly with various types of uncertainty associated with particular taxa. Where a subspecies is said to be “undescribed” by a source, this still means that the source recognises it as valid. The first paragraph of the entry accounts for the foregoing information. The second mainly provides a commentary based on my examination of specimens at NHM. Each taxon is marked with a symbol to indicate my tentative reckoning: ** = possible full species; * = strong (clearly marked) subspecies; • = weak (poorly marked) subspecies; † = dubious subspecies; ‡ = subspecies but not endemic; – = no decision possible. I use “eyeballing” to mean just a brief visual comparison, including matching wing and tail lengths against each other. To save space, Handbook of the birds of the world is indicated, with appropriate volume, as HBW, and I should add that as I wrote the quoted HBW (in press) texts for the thrushes and chats I offer no commentary on their accuracy.

Chinese Bamboo Partridge Bambusicola thoracicus sonorivox** Generally darker than nominate; much smaller rufous throat-patch; chestnut (not black) blotches on flanks (HBW 2). The differences between the two taxa are striking. In the nominate the “rufous throat-patch” is the entire area from below the eyes to the necksides, with a grey breast-patch below (but with the rufous encompassing it by forming a narrow line below). In sonorivox the grey of the breast extends around the rufous chin and throat to the necksides and over the eyes. The underparts of the nominate, below the breast, are buff with small sparse blackish spotting, almost entirely on the flanks; in sonorivox the underparts are rufous-buff with extensive broad chestnut spotting and scaling (a few nominate show more extensive and broader spotting, but not the same degree) (see Plate 1). There are further, lesser, differences above (sonorivox more suffused grey and olive, with more rufous-tinged crown markings, fewer white spots on wing-coverts and mantle, reduced buff edges to wing-coverts).

Common Pheasant Phasianus colchicus formosanus* Undescribed in HBW 2. Paler than E Chinese race torquatus, with flanks almost whitish-buff and mantle straw-coloured (Madge and McGowan 2002). The differences between formosanus and torquatus are as stated, the former also having a green-based rather than blue-based rump, but the two are extremely similar.

Barred Buttonquail Turnix suscitator rostratus • Undescribed in HBW 3 and Johnsgard (1991), although the latter indicates “a duller non-breeding plumage also exists in some races (rostrata [sic], blakistoni)”. The NHM material, which is diversely labelled, suggests that Taiwan females are slightly more rufous-tinged below and slightly buffier in the wing-coverts and the males have reduced barring on the throat and flanks, but the effects of age and season may be in play. At any rate, rostratus appears to be a poorly marked subspecies.

Grey-capped Pygmy Woodpecker Dendrocopos canicapillus kaleensis • Treated by Short (1982)—and hence by Winkler et al. (1995) and HBW 7—as encompassing several mainland taxa, but Dickinson (2003) followed Cheng (1987) in retaining kaleensis as referring exclusively to Taiwanese birds. Short (1982) admitted that Taiwanese and Vietnamese birds are smaller than the others in this group, implying that the former has a longer tail. Otherwise there is no published comparative description. La Touche (1931 [=1931-1934]) separated the adjacent mainland form (nagamichii, although he first used kurodae) from more northerly scintilliceps for having less white on back and wings, less white on upper back (and barred with black), wing-spots smaller, underparts richer and buffier and more broadly streaked, with underwing-coverts always marked with black and sometimes broadly barred; and from kaleensis as “very similar… but larger”. La Touche’s diagnosis of nagamichii from scintilliceps holds. However, eyeballing five males from Taiwan and from the adjacent mainland I cannot see a convincing size difference, although some Taiwan birds may be shorter-tailed. Clearly the two populations are so similar that taxonomic separation is questionable. However, mainland birds (including females) consistently show larger white markings on the tertials, and on this basis I tentatively accept kaleensis as endemic to Taiwan.

White-backed Woodpecker Dendrocopos leucotos insularis • Smallest race, similar to D. l. tangi but more white on back, more extensive pink below (HBW 7). NHM specimens are decidedly shorter-winged and shorter-tailed than tangi, but they have neither more white on the back nor more pink on the underparts.

Grey-headed Woodpecker Picus canus tancolo • Treated by Short (1982)—hence by Winkler et al. 1995), HBW 7 and Dickinson (2003)—as encompassing the population on Hainan, but Cheng 1987) continued to recognise hainanus from the latter. Given how improbable it is for the same subspecies to occur on two widely separated islands, especially when there are adjacent, essentially intervening mainland populations, Cheng’s treatment seems appropriate, although Short’s only diagnosis is that “Hainan birds average slightly smaller than those from Formosa, but there is great overlap”. Taiwan birds are clearly smaller than those on the adjacent mainland. Hainan birds are less clearly smaller than Taiwan birds. The separation of the latter goes back to Stuart Baker (1919), who commented: “There are only two specimens of Hainan birds in the British Museum, but these are smaller than Formosan birds, with smaller bills, and are possibly also rather darker above and less brown below. It is with some doubt that I keep them separate, but Dr Hartert, who formerly considered the two subspecies identical (Novitates Zool. xvii p.222), informs me that a series of 12 birds in the Tring Museum [these will now be in New York, where Short will have seen them] bears out the above characters…”. Both Hainan birds (a male and female) have blacker-streaked crowns than any of the six from Taiwan and than most of those from the adjacent mainland. They are infinitesimally darker green above than Taiwan birds but similar below. The slight size difference appears to be real (a larger sample of both taxa is needed to confirm or refute Short’s “great overlap”). However, on these tenuous distinctions I accept hainanus, making tancolo endemic to Taiwan.

Black-browed Barbet Megalaima oorti nuchalis** Intermediate (with M. o. annamensis) between nominate race and black-crowned races sini and faber, nuchalis with red on mantle (HBW 7). Although nuchalis can be thought of as intermediate in the way HBW suggests, it is in fact less so morphologically than at first glance, and not at all so in terms of geography. Races oorti and annamensis from SE Asia clearly belong together rather than annamensis with nuchalis), differing in distinct but small ways; the same is true of faber and sini, which are approximately adjacent in S China. The Taiwan taxon shares the full red breast-patch with the latter pair, and the crown colour with the former pair, but both former and latter are more similar in having red on the hindcrown, whereas nuchalis has a red stain on the mantle (not the nape, despite its scientific name). My inclination would be to treat oorti and annamensis as one species, faber and sini as a second, and nuchalis as a third. See Plates 2 & 3.

House Swift Apus nipalensis kuntzi or A. affinis kuntzi – Intermediate between nominate (i.e. nipalensis) and subfurcatus, with most heavily streaked rump of any subspecies (HBW 5; also Chantler 2000). This is all based on Brooke (1971): “intermediate in general colour between nipalensis and subfurcatus, but has the rump ashy white and usually with dark shaft-streaks, not clear white as in all other races”. The original description (Deignan 1958) also mentions “general coloration… almost matt black” with a hint of greenish not bluish gloss, and the greyish-white of throat making the definition of light and dark areas less clear. There appear to be no specimens in NHM, but one bird (1914.4.8.13), simply marked “Takao” and taken by Styan, must be from Taiwan (Styan collected there, Swinhoe lived and took birds in "Takow" (see, e.g., Swinhoe 1865), and a 1922 Times atlas marks Takao on the site of modernday Kaohsiung and indicates no place of the same or similar spelling for anywhere in China). This does not possess a rump any more heavily streaked, ashier or less clear-cut than specimens taken at random from NHM trays for nipalensis and subfurcatus (latter from the adjacent mainland). Why in any case would a geographical outlier be intermediate between two subspecies that lie to the west and far south-west of it? Of course a sample size of one with a question-mark over its provenance is not the basis of a taxonomic judgement, and I think the distinguished authorities above should be allowed to prevail; but a new look at the evidence would be worthwhile.

Grass Owl Tyto capensis pithecops (Eastern Grass Owl T. longimembris pithecops) • Treated as synonymous with chinensis by König et al. (1999). Larger (than [nominate] race longimembris or else race chinensis) with some buff suffusion (HBW 5). Four specimens of pithecops in NHM are not obviously larger than the five of chinensis; however, all are whiter-faced, and three are distinctly paler buff on the breast, whiter on the lower abdomen, and paler on the uppertail (the fourth is matched in these features by one relatively pale chinensis). On this slim basis pithecops seems valid.

Mountain Scops Owl Otus spilocephalus hambroecki* Has no rufous morph; dark brown above with prominent pale collar, buffish and finely streaked below, with very distinct pale facial disc and long bicoloured ear-tufts (HBW 5). König et al. (1999) suggest its voice is typical of the species. NHM specimens are as indicated, and marginally less rufous-tinged below and above. Facial pattern, long ear-tufts and whitish nuchal collar make this a very distinctive form (albeit vanderwateri from Sumatra possesses the same collar pattern).

Elegant Scops Owl Otus elegans botelensis – Darker and even more finely marked than Japanese birds, often occurring in dark rufescent morph (HBW 5). NHM has no botelensis.

Collared Scops Owl Otus bakkamoena glabripes or O. lettia glabripes • Paler than nominate (lettia) (HBW 5); toes bare (König et al. 1999). Compared to adjacent mainland race erythrocampe (name not used in NHM), glabripes is not discernibly paler. However, erythrocampe has a ruff of feathers bordering the lower facial disc which are largely buff and distinct from the stonewhite in the rest of the underparts and face, whereas in glabripes this area is stone-white; moreover, there is more buff in the upperparts on erythrocampe than in glabripes, and the latter’s toes are indeed less feathered basally. This therefore seems a moderately distinct subspecies.

Brown Wood Owl Strix leptogrammica caligata or S.newarensis caligata † This taxon is universally treated as present on both Taiwan and Hainan (Cheng 1987, HBW 5, König et al. 1999, Dickinson 2003) but, given the presence of populations (allocated to ticehursti) all along the intervening Chinese mainland, the logic and likelihood of this arrangement are questionable. König et al. say “slightly larger than nominate… possibly a synonym of nominate”. NHM has two specimens from Taiwan (one female, one unsexed), none from Hainan. Wing and tail lengths, eyeballed, do not differ from three (all male) mainland Chinese specimens (race ticehursti). The underpart barring on the two Taiwan specimens is slightly vaguer and browner than that on mainland birds, but this may be a sex, age or individual difference; a female ticehursti from Vietnam appears identical to the Taiwan birds. On this inadequate sample and review, I am inclined to regard caligata as invalid.

Tawny Owl Strix aluco yamadae • Undescribed in both HBW 5 and König et al. (1999). Yamashina (1936) diagnosed yamadae as similar to nivicola (the race from adjacent China) but “decidedly smaller”, with minimal overlap in wing length of eight specimens (256-282 vs 282- 312 mm). Eyeballing wing length of the only yamadae in NHM against the single race ma and a large series of nivicola (split as Himalayan Wood Owl in Rasmussen and Anderson in press) suggests that Yamashina was correct, and I find that Mees (1970) also had a specimen that confirmed the diagnosis.

Collared Owlet Glaucidium brodiei pardalotum* Darkest race (HBW 5). König et al. (1999) acknowledged no other race except pardalotum, but their description involves no comparison. Specimens in NHM are not noticeably darker than mainland Chinese birds, but the streaks on the lower underparts are darker and more distinct, and the head is all spotted, not (mainly) barred and (slightly) spotted. The pale markings on the upperparts of most pardalotum are buffier than in mainland birds. It would be worth checking how different the voice of Taiwan birds is, but on this evidence pardalotum is a good subspecies.

Savanna Nightjar Caprimulgus affinis stictomus • This taxon breeds in Taiwan and may or may not winter in Indochina (Cleere 1998, Holyoak 2001), but does not breed in the latter as implied in Dickinson (2003). Cheng (1987) and HBW 5 list it for Taiwan only. Placed in monticolus group by Cleere, and considered in the past to be synonymous with monticolus, but the five specimens in NHM are greyer overall (Holyoak 2001). Cleere noted that the type, a female, is buffish-brown above and cinnamon-tinged below, and that two male specimens are also very buffish. Swinhoe (1866) thought Taiwan birds extremely close to monticolus and only gave them a name because Blyth, having seen a female specimen, encouraged him to do so. The five NHM specimens of stictomus are indeed less rufous both above and below than specimens of race amoyensis from adjacent mainland China.

Oriental Turtle Dove Streptopelia orientalis orii † Duller than nominate (HBW 4). Very similar to nominate but underparts average slightly paler; probably not valid, but sample size too small (Gibbs et al. 2001). Yamashina (1932) based his description on its smaller size, and “many specimens… somewhat paler in colour than S. o. orientalis… breasts more purplish-red and… rump… more greyish-blue”. NHM holds five specimens from Taiwan. Three are relatively paler-bellied than apparently comparable material from mainland areas and Japan, but I cannot see any reliable way of distinguishing these birds—their wing lengths match those of Japanese birds—and orii seems best placed in synonymy with orientalis.

Spotted Dove Streptopelia chinensis formosa † This race was not mentioned by HBW 4 and was rejected by Gibbs et al. (2001), followed by Dickinson (2003), but Cheng (1987) admitted it, and Gibbs et al. acknowledged its minor differences (“may average a little larger and have more white on the outerwebs of the outer tail feathers”). In describing it, Kuroda (1927) regarded it as identical in plumage and merely shorter-winged on average. Eyeballing three Taiwan specimens in NHM (which does not recognise the taxon) against birds from mainland China, many of the latter have comparable wing lengths and none shows more white on the outertail; so formosa seems a likely synonym of chinensis.

Emerald Dove Chalcophaps indica formosanus † This taxon was rejected by Gibbs et al. (2001) as one of a number of subspecies which “show no significant differences and are best treated as synonyms of the nominate”. Even Cheng (1987) took this line; HBW 4 and Dickinson (2003) make no reference to the taxon. Swinhoe (1865) diagnosed it on the basis of black, not greyish-brown undertail- coverts. NHM has seven specimens from Taiwan, all in essence identical to mainland birds. The undertailcoverts of Taiwanese birds are blackish-grey, but some (collected after Swinhoe) are ash-grey, and this feature appears variable in continental forms also.

White-bellied Green Pigeon Treron sieboldii sororius • This subspecies was accepted by Cheng (1987)— who listed it as wintering in adjacent mainland China—and hence by Dickinson (2003), but was ignored as one of several unspecified but invalid races in HBW 4, and explicitly rejected by Gibbs et al. (2001): “examination of a good series… could confirm none of the diagnostic features nor establish any other significant differences from the nominate”. NHM contains at least eight females and six males from Taiwan, and a larger series from Japan. In the Taiwan tray D. Goodwin, author of Pigeons and doves of the world, has left a note: “I can see no distinction between the Japanese & Formosan specimens. The original statement (based on comparison with a written description) of the colour differences of back & rectrices are [sic] incorrect”. Indeed, back and tail feathers are uniform. However, males from Taiwan all show yellower crowns, napes and neck-sides than any from Japan; and the females do likewise, though less obviously. On this basis sororius emerges as valid, but the difference could scarcely be slighter.

Whistling Green Pigeon Treron formosae formosae – (**) From two Japanese races by golden-bronze crown, smaller size, brighter plumage; Gibbs et al. (2001) listed several characters of the fourth race filipina that do not conform with their own description of the nominate—abdomen white, maroon extending onto mantle, iris blue not red, bill without yellowish or greyish tip—but comment that “none of these features seems to distinguish this form convincingly from the nominate”. They do, however, retain filipina, as does HBW 4; otherwise nominate formosae would not be endemic to Taiwan. Nominate formosae is very and fairly distinct from the northern and southern Ryukyu races respectively, and its separation at species level may be appropriate; this needs urgent investigation, as the conservation status of the species, as currently constituted, is Near Threatened (BirdLife 2001), and a split into two species would probably upgrade both to a higher category of threat (but it is uncertain whether the southern Ryukyu race would belong with its northern or southern neighbour). NHM contains no filipina.

Black-chinned Fruit Dove Ptilinopus leclancheri taiwanus – Larger than nominate with a longer tail, much more massive bill, darker green upperparts with blue blotching from crown to back, larger, paler, more maroon breast-band, posterior of chin-patch more maroon, underparts much darker. This description is taken from Ripley (1962) by Gibbs et al. (2001), who remarked that taiwanus is “known from very few specimens, most of which are now lost”. HBW 4 did not recognise it but may have missed Ripley (1962) altogether, as they did not admit Taiwan itself in the range of the species, merely “Lan Hsü” (presumably Lanyu). NHM holds no specimens. From the description by Ripley, this ought to be a highly distinctive form, and the fact that it is hardly known on Taiwan is completely mystifying.

Slaty-breasted Rail Gallirallus striatus taiwanus † Palest race, along with race gularis (HBW 3); as pale as race jouyi, but smaller; paler than gularis, upperparts more greyish-olive, white on belly more extensive (Taylor 1998). Yamashina (1932) distinguished it from nominate striatus and race gularis by its being “strikingly pale” above and below, with grey edges to back feathers, grey ground colour of wings and flanks, and broader (twice as broad) white stripes on flanks; and from jouyi by its smaller size. NHM only holds a small sample of the species in general, and only two Taiwan birds. However, one of the Taiwan birds matches most of those from the adjacent mainland (jouyi) in wing length, and the colour shading is indeed the same; I could find no specimens in NHM from the range of gularis (S China), but as Cheng (1987) did not recognise the latter the differences are presumably vanishingly small. On NHM evidence, Yamashina’s diagnosis only works with respect to nominate striatus, not jouyi, so the validity of taiwanus appears tenuous.

Slaty-legged Crake Rallina eurizonoides formosana – Colour of throat purer white (HBW 3); upperparts darker than in race amauroptera (Taylor 1998). This latter seems to be based on Seebohm (1895), who remarked that “compared with immature examples of Rallina eurizonoides… the upper parts… are much darker”. NHM only possesses the type, which is in immature plumage; in any case the throat feathering is matted. Its lower underparts have the moreblack, less-white barring of Philippine populations, and its upperparts are precisely similar dark brownish-olive. Everything hinges on the whiter throat of formosana than nominate eurizonoides.

Black Kite Milvus migrans formosanus † This taxon was not recognised by Cheng (1987) nor by Ferguson-Lees and Christie (2001), but was accepted by HBW 4, which offered no diagnosis, and by Dickinson (2003), although both latter sources indicated that it also occurs on Hainan (in respect of which the same considerations as for Grey-headed Woodpecker and Brown Wood Owl apply). Ferguson-Lees and Christie (2001) described formosanus as sedentary in Taiwan (not mentioning Hainan) and as “very like lineatus but averages smaller” (hence in their view invalid). NHM does not possess a copy of the obscurely published original diagnosis by Kuroda in 1920, and Kuroda (1927) merely gave the dimensions of the type. NHM possesses six Taiwan birds and five from “Amoy” (Xiamen) in adjacent Fujian province. In terms of plumage and eyeballed measurements I see nothing to distinguish the Taiwan birds; I doubt they even average smaller.

Crested Serpent Eagle Spilornis cheela hoya* Undescribed in HBW 2. Size of adjacent race ricketti; much darker, cheeks and throat blackish, breast nearly plain, rest of underparts clearly spotted (Ferguson-Lees and Christie 2001). Upheld in NHM.

Crested Goshawk Accipiter trivirgatus formosae • Undescribed in HBW 2. Although relatively distant from other populations, the only distinction given is “probably averaging larger even than indicus” (Ferguson-Lees and Christie 2001). The original description (Mayr 1949), however, stressed this taxon’s darkness (but disclosing that the sample size was two!): adult male “very dark, particularly below”, immature female “very dark and heavily marked underneath”, the tendency towards darker coloration being “noticeable also in certain specimens from northern Burma and northern Indo-China, but not reaching quite the extreme shown by the two Formosan birds”. NHM possesses four Taiwan birds, two adults (male and female) and two immatures (male and female). The adult female and immatures are no darker above or below than various birds from elsewhere in the range. The adult male is matched for darkness above by two males from Thailand; however, the pale rufous-buff breast-streaks are much browner in the Taiwan bird, and the barring on the belly is mid-brown with paler centres. No size differences suggest themselves. On this meagre basis formosae must provisionally stand. Besra Accipiter virgatus fuscipectus • Undescribed in HBW 2. Placed with mainland “affinis” group and judged the largest race, darker and browner above (Ferguson-Lees and Christie 2001). The original description (Mees 1970), comparing with mainland affinis, indicated that adult males have “breast and barring on the underparts browner, not so rufous”, and “upperparts… brownish-grey rather than pure grey”, while adult females are likewise slightly browner below but greyer (less brown) above (so that viewed from above males and females of fuscipectus show hardly any difference); moreover, male wings are usually >170 mm in fuscipectus, usually <170 mm in affinis. Mees (1970) borrowed four NHM Taiwan specimens for his extensive comparisons, but in fact NHM holds no fewer than 13 specimens from the island. His diagnosis of differences between males holds, and on this basis fuscipectus is validated; but I am stumped by his assertion that females are greyer above than female affinis (some specimens show a greyish gloss, including one borrowed by Mees, but so do several affinis), nor can I see any difference between NHM’s two Hainan specimens (both female) which Mees said are brown above and therefore belong with affinis. In the absence of a male from Hainan it is not possible to determine whether fuscipectus is present there or not; for the moment it is appropriate to assume not (particularly given my doubts about subspeciessharing between the two islands), leaving fuscipectus a poorly marked Taiwan endemic.

Long-tailed Shrike Lanius schach formosae † This taxon was accepted by Cheng (1987) but not mentioned by Lefranc (1997), Harris (2000) or Dickinson (2003). Specimens from Taiwan are identical to certain birds from adjacent mainland China.

Eurasian Jay Garrulus glandarius taivanus* Placed in the “bispecularis” group, distinguished by “blackish mottling on forehead and dusky nasal tuft” (Madge and Burn 1993). The nasal tufts are black, not dusky, and the mottling is really just a few streaks immediately above the tufts. Other diagnostic features are: smaller and smaller-billed than adjacent mainland sinensis, whiter on abdomen and greyer on scapulars.

Grey Treepie Dendrocitta formosae formosae* Placed in the eastern group which is characterised by darker, less patterned plumage with a very white rump and shorter, wholly black tail; nominate formosae fairly brown above and light below, with basal third of central tail feathers grey (Madge and Burn 1993). Taiwan birds differ from those on the adjacent mainland not only by the grey bases to the tail (black on mainland) but also by the whitish-grey rump (whitish on mainland).

Spotted Nutcracker Nucifraga caryocatactes owstoni* Placed in the southern group which is distinguished from the northern group by a mid-brown body, smaller, less intense spotting and relatively longer tail with more white in the outermost feathers; in owstoni ground colour of body sooty-brown (Madge and Burn 1993). The deeper brown body coloration in owstoni is confirmed, and its blacker cap; moreover, the white mantle-streaks in the nearest Chinese population (race macella) are reduced in owstoni to a few nape-flecks, and the size of breast-spots is also greatly reduced.

Maroon Oriole Oriolus traillii ardens ‡ Having eight years earlier named ardens as a species, Swinhoe (1870) distinguished as a variant the population he found on Hainan under the name nigellicauda, asserting it to be shorterwinged and longer-tailed, with black of neck extending less far down the breast, and black “shafts” to tail feathers. Dickinson (2003) extended the range of nigellicauda to N Vietnam and SE Yunnan, China. NHM has a good series from Taiwan, with six full-plumaged males including the type of ardens, but only a single full-plumaged male from Hainan, the type of nigellicauda. It also has a good series from Indochina. I cannot see any difference between adult males from Taiwan and from Indochina; and I think the sooty-shaded vanes (not shafts) of the type of nigellicauda are an individual variation (birds of all populations show this in varying degree). Females and immatures from all three areas also appear similar. Measurement of all specimens might disclose some consistent differences of the kind to which Swinhoe alluded, but eyeballing wings and tails suggests no real discontinuities. I suspect the name ardens should apply to all these populations.

Black Drongo Dicrurus macrocercus harterti • Stuart Baker (1918) described this taxon as the biggest of all in the species, with the exception of Himalayan birds (race albirictus), but “separable at a glance by its curiously short tail”. Stuart Baker’s diagnosis seems to hold in NHM birds, although some specimens have tails that are comparable in length to those of other taxa. The geographically closest race, cathoecus of E China, is certainly often as long-tailed and, from Stuart Baker’s own data, entirely overlaps harterti in wing length; but the bill of harterti appears consistently larger (deeper), and it is more on this basis than any other that harterti seems valid.

Bronzed Drongo Dicrurus aeneus braunianus* Dickinson (2003) admitted two other subspecies, of which the closest to approach Taiwan is the nominate, ranging from India through Indochina to S China. Swinhoe (1863) compared his specimens of braunianus with specimens from India, and noted that the bill is always shorter and broader-based, with higher feathers on the culmen, while the head and back feathers are shorter and rounder, reflecting purple and blue instead of copper-green; breastsheen similar, but feathers there much larger and round instead of lanceolate; wings and tail black below (not brown), and wings longer; feathers of tail much broader, sheen again blue not green. Although NHM lacks material from adjacent mainland China, many specimens of nominate aeneus from Indochina are available. Comparison of these with braunianus bears out Swinhoe’s diagnosis well, with the exception of bill dimensions and shade of undersides of wings and tail.

Black-naped Monarch Hypothymis azurea oberholseri • Stresemann (1913) diagnosed this taxon as very like nominate azurea in colour but on average larger with slightly less violet on the back and generally larger nape-mark. He added that the adjacent mainland race styani is on average smaller, and the white of the undersides reaches higher on the belly. At a glance I cannot see any obvious difference in back colour between oberholseri and azurea, or in the extent of white on the belly between oberholseri and styani; however, the black nape-mark is indeed generally larger in oberholseri than azurea, and both azurea and styani appear smaller. The bill of oberholseri looks stouter than in azurea, shorter than in styani.

Island Thrush Turdus poliocephalus niveiceps** Possibly best candidate for species status among at least 51 subspecies of Island Thrush, being highly distinctive in plumage, the only subspecies with pronounced sexual dimorphism, and much the most northerly outlier; male white-hooded with blackish upperparts and breast-band shading to orange belly with broad black-and-whitestriped vent, female similar but browner above, head brown with buff postocular supercilium, buff-speckled cheeks, buff submoustachial and throat with dark-stippled malar and throat-streaks; only in this race does female show complex facial pattern and dark throat-streaks (HBW in press). I am not in favour of a major splitting exercise on poliocephalus, but re-checking NHM material certainly reinvigorates my sense that this taxon may be a special case within the poliocephalus complex. See Plate 4.