GuyM. Kirwan 83 Bull. B.O.C. 2008 128(2) Studies of Socotran birds III. Morphological and mensural evidence for a 'new'' species in the Rufous Sparrow Passer motitensis complex endemic to the island of Abd 'Al-Kuri, with the validation of Passer insularis & Sclater Hartlaub, 1881 hy Guy M. Kirwan Received 15February2007 Thepresent studyrepresents the third ina series ofpapers that seeks to re-examine the taxonomicstatusofavianformsdescribedfromtheSocotranarchipelago. Itfollowspublica- tions concerning the population of Nubian Nightjar Caprimiilgiis nuhicus, which was previously regarded as an endemic subspecies,jonesi (Kirwan 2004), and a re-evaluation of species limits in Golden-winged Grosbeak Rhynchostruthiis socotranus (Kirwan & Grieve 2007). Thesenotesseekto stimulaterenewedinterestintaxonomic studiesofSocotranbirds, specificallytomeetthechallengesetbyMartins (1996),whostated: Thereisaclearneedfor a review ofthe avifauna ofSocotra which reflects contemporary systematic thinking.' Alloriginallydescribedasspecies,thesix allopatricgroups (andnineconstituenttaxa) within the almostexclusivelyAfrotropical Rufous Sparrow Passer motitensis complex have suffered a rather checkered taxonomic history of late. Following their demotion to sub- species, this arrangementpersisted throughMoreau & Greenway (1962), White (1963) and Hall & Moreau (1970). Thus, despite the contrary opinions of van Someren (1922), Lynes (1926), Grant & Mackworth-Praed (1944), Bannerman (1948) and Macdonald (1957) con- cerningsome orallofthesetaxa, itwasnotuntilWolters (1982),who splitP. insularis,from the island ofSocotra, offnorth-eastAfrica, and Summers-Smith (1984, 1988), who separat- ed P. iagoensis, fromthe CapeVerde archipelago, atthe level ofspecies thatmore than one species tended to be recognised by major works. This notwithstanding the riposte of Bourne (1986) toSummers-Smith (1984)whereinBournecouldfindlittletorecommendthe advancement of iagoensis to specific status beyond the need for 'a tiresome change of name'. Thus,Sibley&Monroe (1990) recognised thefollowingspecifically: P. iagoensis,P. insu- laris, P. rufocinctiis (including cordofanicus and shelleyi), from East Africa, and P. motitensis, from South Africa north to southern Angola (which arrangement was followed by Gill & Wright2006). Incontrast,Dowsett&Dowsett-Lemaire (1993),followedbyDickinson (2003) in his important world checklist, preferred to recognise just motitensis and iagoensis as species, and pointed to errors and inconsistencies in the work of Wolters and Sibley & Monroe. However, therelevantvolume oftheinfluential ThebirdsofAfrica elected to recog- nise P. shelleyi (spottily in Ethiopia and Somalia south to Kenya) and P. cordofanicus (from west-central Sudan to eastern Chad), in addition to the four taxa separated by Sibley & Monroe, specifically (Urban in Fry & Keith 2004), albeit not without criticism (Leonard & Demey 2006). Thus, in recent years, only motitensis and iagoensis have received reasonably widespread recognition as being meritorious of specific status (Cramp & Perrins 1994 and . Hazevoet 1995 also afforded iagoensis such treatment). Passer[motitensis] insularisSclater&Hartlaub, 1881,wasunsurprisingly(givenitsabun- dance, which is currently placed at c.230,000 individuals: R. F. Porter in litt. 2007) amongst the initial wave of taxa endemic to the ancient island ofSocotra to be described, following 1 GuyM. Kirwan 84 Bull.B.O.C.2008 128(2) the first scientific visit to the archipelago, by Sir Isaac Bayley Balfour, who spent almost seven weeks there in 1880. It was not until the considerably more extensive survey by Ogilvie-Grant & Forbes, in 1898-99, that Passer [motitensis] hemiUuciis Ogilvie-Grant & Forbes, 1899 (hereafter referred to as P. hemileucus), was discovered. The latter is endemic to Abd 'Al Kuri, a rather inhospitable island with no permanent running water, c.36.5 km eastto westand a maximumofc.5 kmnorthto south, whichlies c.145 kmwestofthe main island and rises to amaximum 743 m (Cheung & DeVantier2006). Abd 'AlKuri covers 133 km^whereasSocotrais3,625km^inarea.Ourknowledgeofinsularishasincreasedsubstan- tially (see, e.g., Kirwan etal. 1996) since even the work ofSummers-Smith (1988), who was able to make scarcely even the most basic comments about the bird's natural history, but hemileucus has remained a mysterious taxon known almost solely from specimens, namely the type series and sevenbirds collected by Alec Forbes-Watson in spring 1964. In report- ing on the latter collection, Ripley & Bond (1966), in an immense understatement, referred onlytohemileucusbeingpalerthaninsularis. Inconsequencethetaxon'sobviousdistinctive- ness has gone unappreciated, though its describers (Ogilvie-Grant & Forbes 1899) were clearlyawareofthisastheirmanuscriptmakesplain(evenallowingforthefactthatallsuch novelties were then afforded species status). Summers-Smith (1988) opined that 'although paler and slightly smaller the differences are not great enough to warrant their separation from the birds on Socotra even as a different race.' Clement et al. (1993), presumably impressed by Summers-Smith's statement as to the weak distinction, simply ignored hemileucus. Urban (2004), in contrast, noted it as being 'Much paler than insularis, under- partsnearlypurewhite,blackpatchonchinof smaller, ? withoutduskypatchonthroat', and also remarked on the overall smaller size of hemileucus (something which had not escaped, but apparently failed to impress, Summers-Smith). None of these commentators, withtheexceptionofDillonRipleyandBond,appearstohaveexaminedtheForbes-Watson specimens, and itmightbewonderedwhether the firsttwo authors looked carefullyatthe, admittedlylimited,material tohandinTheNaturalHistoryMuseum (Tring). Infact.Passer hemileucus appears as easily diagnosable as any other member ofthe Rufous Sparrow com- plex admitted to species statusbyUrban (2004). Methods I acquired mensural data from specimens of both Socotran taxa held at The Natural History Museum (NHM, Tring), as follows: Passer insularis (Socotra: n=15, including nine males, one of them juvenile which was not included in the mensural analysis), and Passer hemileucus (Abd 'Al Kuri: n=2, including one male), and the National Museum of Natural History (SmithsonianInstitution),WashingtonDC: P. insularis (Socotra: n=20,includingten males), and P. hemileucus (Abd 'Al Kuri: n=7, including four males). I did not examine the singlemalespecimenofhemileucus ortheseven (bothsexes) ofinsularisheldintheNational Museums and Galleries onMerseyside, Liverpool (C. W. Fisher & T. Parker in litt. 2007) or the two specimens of insularis held atthe University Museum ofZoology, Cambridge, UK. Note that Forbes-Watson's expedition report (which was never published) records that he took 28 males and 23 females of insularis, but the whereabouts of the additional specimens are unknown; notes in the Smithsonian collection report that specimens of some other species were sent as exchanges with museums in Africa, e.g. the Natural History Museum in Bulawayo (M. P. S. Irwin in litt. 2007). Mass data for these additional specimens are pre- sented in the unpublished report and are repeated in Table 1. The types of both were examined: P. insularis (NHM 1881.3.21.20) and P. hemileucus (NHM 1899.8.11.131). The fol- lowing data were obtained from each specimen: wing-chord (flattened) and tail-length. GuyM. Kirwan 85 Bull. B.O.C. 2008 128(2) using a standard metal wing-rule with a perpendicular stop at zero (accurate to 0.5 mm), and culmen-length (to skull) and culmen-depth (at the feathers), using digital callipers (accurate to 0.01 mm). I also conducted a morphological examination of33 specimens ofP. NHM NHM cordofanicus and 29 ofP. shelleyi held in (including the holotype, 1887.9.28.314, ofshelleyi), thesebeingthegeographicallymostproximateformswithinthecomplexincon- tinental Africa, especially P. shelleyi which at least formerly maintained a toehold in north-westernmost Somalia (Ash & Miskell 1998). Notes on plumage variation in both sexes of the two forms were taken and ranked according to their usefulness in distinguishing them. I attempted to conservatively score character differences for males using the system elucidated by Collar (2006), which willbe fullytabledbyCollaretal. (inprep.). Numbersinbracketsrefertothisscoringsystem,rang- ing from 3 for a dramatic difference to 1 for a more minor difference. A broad range of material, pertaining to both forms, was photographed, using a Nikon Coolpix 885 digital camera (see Figs. 1-6). Results — Plumage analysis. Comparison of plumage characters in Passer insularis and P. hemileucus revealed strikingdifferences, coincidentwiththe recognitionoftwo species. Inmales, com- pared to insularis, hemileucus has muchpalerunderparts lacking any ofthe dirty grey tones with which the entire underparts of insularis are washed and which contrast much more noticeablywith the white cheeks in the latter form; the dark throatpatch is confined to the bib in hemileucus but is far more expansive and more solidly black in insularis (Fig. 1), though R. F. Porter {in litt. 2007) cautions that it may appear smaller and less striking in February/March; the median coverts show prominent white tips in hemileucus, but much smaller, less obvious and grey-coloured ones in insularis, whilst the chestnut tones in the wing ofhemileucus are muchpurer and less saturated than on insularis, recalling the differ- ence between Eurasian Tree P. montanus and Spanish Sparrows P. hispaniolensis; the black ear-coverts are more restricted in hemileucus, and insularis also has more extensive and deeper black lores (Fig. 2); the crown-feathers have much paler centres in hemileucus; and finally the mantle and back are much browner and less heavily and darkly streaked in hemileucus,lackingmanyofthegreybackgroundhuesofinsularis overlainwithbroaderand blacker streaks (Fig. 3). I scored two points each for the differences in head pattern, upper- parts pattern and underparts coloration, with additional single points for the difference in wing pattern and size (see below), thus giving a total of eight points. The minor range of variation in these features in four males ofeach taxon is depicted in Figs. 1-3. Females of the two taxa are almost equally distinctive. As evidenced by Figs. 4-6, female hemileucus has overall much whiter underparts than insularis, though like males hemileucus femalesexhibitsomebuffytonesespeciallyonthebreastandflanks,whilst insu- laris has overall much greyer underparts, though some variation is evident. No specimens ofhemileucus show any evidence ofa darkbib, a feature evident in all insularis, though one (NMNH 518324) almost lacks any trace, dark feathering being confined to the bases of the centralthroat(Fig. 4). P. hemileucushasamoreobviouspalesuperciliumbehindtheeyeand lacks any trace of dark feathering on the ear-coverts (female insularis are more male-like); like males, female hemileucus shows reasonably prominentpale tips to the median coverts, which are not apparent in insularis; and, again as in males, hemileucus has generally much brownerupperpartswithslightlybrownerandlessbroadstreakingonthemantleandback, compared to the generally much greyer and darker upperparts of insularis, marked with generally broader and blacker streaking. Females of these two taxa would seem to be far GuyM. Kirwan 86 Bull. B.O.C.2008 128(2) GuyM.Kirwan 87 Bull. B.O.C. 2008 128(2) Figures4-6. Ventral, lateralanddorsalviewsoffemalesofPasserhemileucus (left) andP. insularis, collected in the Socotran archipelago, in spring 1964, by A. D. Forbes-Watson (Guy M. Kirwan / Smithsonian Institution,Washington) GuyM.Kirwan 88 Bull. B.O.C.2008 128(2) more readily identifiable in a field context than, for example, those of House P. domesticus and Spanish Sparrows. Male hemileucus differs from shelleyi in having greyer upperparts with much paler and narrower dark streaking, paler brown tertial centres and rectrices, a greyer crown, lack of anyrufousintherump, lessbold and solidblackear-covertsandbib, andtherufoussuper- ciliary does notwrap around the rear edge of the ear-coverts. Female hemileucus also lacks the rufous rump of female shelleyi, any rufous in the mantle, has much fainter upperparts streaking,amuchpalercrown,tertialsandflight-feathers,andlacksanytraceofabib (quite noticeable in shelleyi). Compared tomalehemileucus, the same sexofcordofanicushas arufous rump, theman- tle and back are overlain with sparse butvery deep black streaking (darkbrown streaking inhemileucus),theflight-feathers aredarker,thewhitetipstothewing-covertsareseeming- ly narrower, the rufous superciliary continues to wrap around the ear-coverts, and the bib andupperbordertotheear-covertsarebothdeepblack. Unlikehemileucus,femalecordofan- icus is rather similar to the male of that taxon, except that it has a much paler (greyer) but still rather extensive bib; furthermore the crown is much darker grey than in female hemileucus. Finally, compared to these twomainland Africanforms, insularis is clearlydifferentiat- ed by a broad range of characters, e.g. rump coloration, underparts coloration etc. The closest of the motitensis group in plumage is rufocinctus (Kenya and northern Tanzania; 53 specimens examined in NHM). Male insularis differs from rufocinctus in the following: lack of rufous rump; darker grey upperparts with narrower, slightly less black streaking (espe- ciallynoticeableonthecrown / nape);morediffuse, slightlylargerand less deepblackbib; onaveragegreyerunderparts;ear-covertslargelywhite (greyin rufocinctus)withblackeye- stripe (absent in rufocinctus); greater coverts with much more rufous (generally none in rufocinctus);alargerbill;whilsttheiridesarepaleinrufocinctusbutdarkininsularis. Female rufocinctushasarufousrump andlessercoverts,heavierblackupperpartsstreaking,amore noticeable and more extensive grey ('shadow') bib, greyer ear-coverts (white in insularis), generallymore grey-saturated crownandnape (darkfeatherbasesmorenoticeable in insu- laris), and a shorterbill. — Morphometries. Mensural data appertaining to the two Socotran taxa are presented in Table 1. These reveal the rather obvious differences in their relative sizes. Compared to insularis, hemileucus is shorter winged, shorter tailed, smallerbilled (particularly obviously in bill-depth, much less so in culmen-length), and lighter in weight. These differences are readily apparent inboth sexes and, particularly in wing- and tail-length, therebeing no or verylittle overlap between the two taxa. Discussion Vocal data are non-existent for P. hemileucus and rather weak (few sound-recordings) for P. insularis. In any case differentiation in Passer is generally rather weakly expressed in termsofvocalisations, songsbeingratherunspecialisedalthoughtheydopossessanadver- tising function (Cramp & Perrins 1994). The available data concerning plumage and morphometriesunambiguouslysupportsrecognitionoftwo speciesunderanyspeciescon- cept currently operating, including all of the pattern-defined concepts (Sluys & Hazevoet 1999), the Metapopulation Lineage Concept or General Species Concept (de Queiroz 2005) or any ofthe more modern interpretations ofthe Biological Species Concept (e.g. Helbiget al.2002,Collar2006,Collaretal. inprep.). Thedifferencesbetweeneithersexofinsularisand — GuyM. Kirwan 89 Bull. B.O.C. 2008 128(2) hemileucus are clearly as great asbetween any ofthe other members ofthe Rufous Sparrow P. motitensiscomplexcurrentlyrecognisedatspecieslevel,aswellasbetweenanyofthefive membersofthe (alsoAfrotropical) Grey-headed SparrowsuperspeciesP.griseus sometimes accorded specific recognition (see Urban 2004). Indeed, it is interesting to note the quite close morphological resemblance between both sexes (but especially males) of hemileucus and those of Sind Jungle Sparrow P. pyrrhonotus of south-east Iran to north-west India, which fact was noted by the authors of the new taxon but which has apparently escaped more recent commentators (Fig. 7). The substantial difference in size between hemileucus and insularis would presumably act as a significant barrier to the chances of their inter- breeding should they come into contact. Almostnothinghasbeenpublished concerning thehabits and behaviour ofP. hemileu- cus, although it has been found in most parts of the island of Abd 'Al Kuri (R. F. Porter in litt. 2007), the onlydatabeing those recordedbyOgilvie-Grant & Forbes (1903), who found it tobe unassociated with Man, in obvious contrast to insularis. They stated: Tt was never seen in the neighbourhood of the native village, but appeared to be con- finedtothebush-cladslopesofoneofthehighestpoints,whereenormouslimestoneblocks which have fallen away from the summit lie scattered over the hillside. Here it makes its home, and we found itbynomeans an easy task to secure specimens for they are very shy and not very numerous [contra the statement in Kirwan et al. 1996] ... A small flock, how- ever, kept flitting about near me on the stems of the bizarre Milk-bushes {Euphorbia Abdelkuri) growingaboutinthemiddleofthemountain,whileIwasengagedindiggingup the fine specimen ofthis new plant, which eventually reached home alive H.O.F.' In 1964, Alec Forbes-Watson's experience (as recorded in his unpublished expedition report) was slightly different. He managed to collect seven of the dozen birds seen, but intiallyfound themverywild andunapproachableinthehills. However,twodayslater (on 22 March), AF-W found hemileucus at a settlement on the north coast 'behaving like P. domesticus or their relatives on Socotra'. Despite their shyness and AF-W's persistence in collecting them, the birds consistently returned to a single Salvadoria bush, which was 'the onlygreeneryinsight'. Ajuvenilehasbeen trapped inlateMarch (Fig. 8; R. F. Porter in litt. 2006). Future ornithological studies on Abd 'Al-Kuri will prioritise documenting this spar- row's population, range and habitat preferences, and any apparent threats (R. F. Porter in UTM litt. 2007). P. hemileucus has been confirmed to breed in four of the island's six grid squares and presence established in one other (R. F. Porter in litt. 2007), suggesting thatits populationisperhapssecureatpresent,butitsconservationstatus,ifspecies statuswereto begenerallyadmitted,clearlyrequiresevaluation,ashemileucusisconfinedtoasmallisland and probablyhas an overall small population. Freshmaterialpertainingto hemileucus and insularis willbe ofvalue to molecular stud- ies that should seek to date the divergence between the two forms, although the Forbes-Watson material is sufficiently recent to permit such an undertaking with relative ease(Payne&Sorensen2003). SocotraoriginallyformedpartoftheAfrican-Arabiantecton- icplate (itrepresentsacontinuationoftheSomalipeninsula) andprobablybecameisolated bythesameseriesofdislocationsduringthebreak-upofGondwanathatproducedtheGulf ofAdeninthe lateTertiary, atleast lOMYA (Laughtonetal. 1970). The Socotra Platformon which all of the islands lie, a small continental granite block over which the seas are rela- tively shallow, has experienced relative tectonic stability since C.6-8MYA and particularly since the onset of the Pliocene (Fournier et al. 2001). Periods of uplift in the general region of the southern Gulf of Aden terminated C.2MYA. During the Last Glacial Maximum m (20,000-25,000 years ago), when sea levels were up to 120 lower thanin the present day, thecoastalplainofthemainisland ofSocotrawasmuchlargerthannow and encompassed GuyM.Kirwan 90 Bull. B.O.C.2008 128(2) TABLE 1 MensuralandweightdataforPasserinsularis andP. hemileucus. Allmeasurementsinmm;weighting. Personalmeasurements (fromspecimensinTheNaturalHistoryMuseum,Tring,andNationalMuseumof NaturalHistory,WashingtonDC)weretakenusingastandardmetalwing-rulewithaperpendicularstop atzero (accurateto0.5mm),anddigitalcallipers (accurateto0.01 mm). Dymond (1993) didnottake comparablemeasurementsforculmen-lengthanddidnotmeasurebill-depth. Taxon Sex Sample Wing- Tail- Culmen Bill- Mass size length length (toskull) depth (reference) P. insularis male 18 72.5-78.5 58-65 15.12-17.04 8.58-10.13 20-33 (n=16) (Forbes-Watsonunpubl.) means 76.19 60.55 16.16 9.37 26.32 (n=28) female 16 71.0-76.5 52-63 15.41-17.51 8.80-10.06 22-35 (Forbes-Watsonunpubl.) means 73.75 58.12 16.49 9.44 - 27.52(n=23) Publisheddata male 5 74-79 53.5-61.0 25.0-29.2 fromDymond 74 58.6 27.5 (1996),range female 4 72-76 58.0-60.5 26.0-29.1 plusmeans 74.7 59.2 27.1 P. hemileucus male 5 70.5-73.0 53-57 15.05-16.74 7.90-8.83 20-25 (Forbes-Watsonunpubl.) means 72.3 55 15.75 8.3 23.0 (n=4) female 4 67.0-69.5 50-54 14.66-15.57 7.39-8.94 20-24 (Forbes-Watsonunpubl.) means 68.6 52.5 15.18 8.22 21.33 (n=3) the islands of Darsa and Samha (The Brothers) which lie almost equally equidistant between the main island and Abd 'Al Kuri. Indeed, because of the generally shallow seas betweenSocotra andTheBrothers (30-50m),landbridgesbetweentheseislandsexistedon severaloccasionsduringtheUpperPleistocene,occurringin100,000-yearcycles (Cheung& DeVantier 2006). It is unsurprising in the present context therefore that Samha and Darsa shouldhosttypicalP. insularis (R. F. Porter in litt. 2006), thoughP. G. Ryan (in litt. 2007) has remarked that populations on these two islands are small, rather like hemileucus, and lack the grey extending onto the breast. Populations on all four islands thus require some fur- ther research into their taxonomyand relationships. Abd 'Al Kuri, unlike the other satellite islands of Socotra, has been long separated from the main island with resultant levels of endemism,e.g. amongstreptiles(Rosier&Wranik2004),plants (Miller&Morris2004),mol- luscs (Neubert 2002) and marine fishes (Klaus & Turner 2004), whilst one other endemic bird taxon, Onychognathus blythii creaghi, has been described from Abd 'Al Kuri, though its validity requires further investigation (Kirwan 2007). The degree of differentiationbetween the Passer taxa in the Socotra archipelago might suggest separate colonisation events, perhaps even involving different metapopulations from Africa and Asia. The phylogeographic study of western Indian Ocean sunbirds con- ducted by Warren et al. (2003) offers an interesting case study with some parallels to the Socotran situation. Most ofthe latter islands' breedingbirds are clearly ofAfrotropical ori- gin, as long ago noted by Chapin (1932), but there has been, as yet, no attempt to date the different colonisation events thatpresumably occurred, nor efforts to testwhether some of the endemic taxa might be of Asian ancestry, for example the grosbeak Rhynchostruthus socotranus (see Kirwan & Grieve 2007). GuyM.Kirwan 91 Bull. B.O.C.2008 128(2) Figure 7. Comparisonbetweenbothsexes ofPasserhemileucus andP. pyrrhonotiis, fromleftto right: maleP. pyrrhonotiis, male P. hemileucus, femaleP. pyrrhonotus and femaleP. hemileucus (GuyM. Kirwan, ©Natural HistoryMuseum,Tring) Figure 8. Juvenile Passer hemileucus, Abd 'Al Kuri, Socotra, late March (Ahmed Saeed Suleiman & Nadim Taleb / SocotraConservation&DevelopmentProgramme) GuyM.Kirwan 92 Bull. B.O.C.2008 128(2) This study reconfirms the importance of the Forbes-Watson collection from the Socotranarchipelagoandtheunpublishedexpeditionreport(seeKirwan1997),notonlyfor taxonomic studies of the islands' birds but also many aspects of their natural history. The 523 specimens that originally constituted the collection are a remarkable testament to Forbes-Watson and his two Kenyan taxidermists, and continue to represent the best (and mostbeautifully prepared) series ofbirds from Socotra; the previous lack of attention that hasbeenpaid to this resourceborders on the extraordinary. Acknowledgements I am grateful to staff members at the Natural History Museum, Tring (Robert Prys-Jones, Katrina Cook, Alison Harding and Mark Adams) and National Museum of Natural History, Washington DC (James P. Dean,Carla Dove, Christina A. Gebhard, GaryGraves and BrianK. Schmitt), for their assistance atthese institutions.RJ.ichardPortersentfieldphotographsofSocotrasparrowsformyuseandClemencyFisherand TonyParkerassistedwitha copyoftheLiverpoolMuseumdatabaserelatingto theirSocotraholdings and some reference material. With the permission of the authors, Stuart Butchart (BirdLife International) for- warded a copy of the Collar et al. paper in preparation. Michael Jennings' persistence was essential in tracking down the unpublished expedition report of Alec Forbes-Watson, and I am grateful to Jennings, Porter and Peter Ryan for their constructive commentaries on this manuscript. Field work in Socotra was undertaken within the framework of the 1993 Ornithological Society of the Middle East (OSME) South Yemen expedition. OSME also provided financial support that permitted my museum studies in North America. References: Ash,J. &Miskell,J. 1998. BirdsofSomalia. PicaPress,Robertsbridge. Bannerman,D.A. 1948. Birdsoftropical ]NestAfrica.CrownAgents,London. Bourne,W.R.P. 1986.RecentworkontheoriginandsuppressionofbirdspeciesintheCapeVerdeislands, especiallytheshearwaters,thekitesandthesparrows. Bull. Brit. Orn. CI. 106: 163-170. Chapin,J. P. 1932. ThebirdsoftheBelgianCongo,PartI. Bull.Amer.Mus. Nat. Hist. 65: 1-756. Cheung,C. &DeVantier,L.2006. Socotra:anaturalhistoryoftheislandsandtheirpeople.OdysseyBooks,Hong Kong. Clement, P., Harris, A. & Davis, J. 1993. Finches and sparrows: an identification guide. Christopher Helm, London. Collar,N.J.2006. ApartialrevisionoftheAsianbabblers(Timaliidae).Forktail22: 85-112. Collar,N.J.,Fishpool,L.D.C,Pilgrim,J.D.,Seddon,N.&Spottiswoode,C.N. (inprep.)Theassessmentof avianspeciesrank,2: quantificationofcharactersignificance. Cramp,S. &Perrins,C.M. (eds.) 1994. Thebirdsofthe WesternPalearctic,vol. 8.OxfordUniv.Press. Dickinson, E. C. (ed.) 2003. The Howard & Moore complete checklist of the birds of the world. Third edn. ChristopherHelm,London. Dowsett, R. J. & Dowsett-Lemaire, F. 1993. A contribution to the distribution and taxonomy ofAfrotropical and Malagasybirds.TauracoRes. Rep.5. TauracoPress,Liege. Dymond, J. N. 1996. Biometric data ofbirds in southern Yemen and Socotra, spring 1993. Sandgrouse 17: 158-164. Fournier,M.,Patriat,P.&Leroy,S.2001.ReappraisaloftheArabia-India-Somaliatriplejunctionkinematics. Earth&PlanetarySci. Letters 189: 103-114. Gill,F. &Wright,M. 2006. Birdsoftheworld:recommendedEnglishnames.ChristopherHelm,London. Grant,C.H.B.&Mackworth-Praed,C.W. 1944.NotesoneasternAfricanbirds.Bull. Brit. Orn. CI.64:35-36. Hall,B.P.&Moreau,R.E. 1970.Anatlasofspeciation inAfricanpasserinebirds.Brit.Mus. (Nat.Hist.),London. Hazevoet, C. J. 1995. The birds ofthe Cape VerdeIslands:an annotated check-list. British Ornithologists' Union Check-list13.BritishOrnithologists'Union,Tring. Helbig,A.J.,Knox, A. G.,Parkin, D. T.,Sangster,G. &Collinson,M. 2002. Guidelinesforassigningspecies rank.Ibis 114: 518-525. Kirwan,G.M. 1997. Socotra: Forbes-Watsonmanuscript.Phoenix 14: 6-7. Kirwan,G. M. 2004. Someremarks onthetaxonomyofNubianNightjarCaprimulgus nubicus,withparticu- larreferencetoC. n.jonesiOgilvie-Grant&Forbes, 1899. Bull.Afr. BirdCI. 11: 117-125. Kirwan,G. M. 2007. StudiesofSocotranbirdsIV. Synonymizationofsixendemicbirdtaxa,withcomments onthenameOnychognathusblythiicreaghi. Sandgrouse29: 135-148. Kirwan,G. M. &Grieve, A. 2007. Studies ofSocotranbirdsII. One, twoorthreespecies: towardsarational taxonomyfortheGolden-wingedGrosbeakRhynchostruthussocotranus. Bull.Afr. BirdCI. 14: 159-169. Kirwan,G. M.,Martins,R. P.,Morton,K. M. &Showier,D. A. 1996.ThestatusofbirdsinSocotraand'Abd Al-KuriandtherecordsoftheOSMEsurveyinspring 1993. Sandgrouse 17: 83-101. Klaus,R. &Turner,J. R.2004.ThemarineisotopesoftheSocotraArchipelago.FaunaofArabia20:45-115.