Taxonomy of Crimson-winged Finch Rhodopechys sanguineus: a test case for defining species limits between disjunct taxa

2006·6.3 MB·

by  Guy M Kirwan

—
Taxonomy of the Crimson-winged Finch
Rhodopechys sanguineus: a test case for defining
species limits between disjunct taxa?
GuyM. Kirwarf, PhilW.Atkinsonb ArnoudB. van den BergandHadoram Shirihaid
,
Taxonomie du Roselin a ailes rousses Rhodopechys sanguineus: un cas d’etude pour definir les
limites des especes entre des taxons geographiquement eloignes? Nous avons examine la
taxonomie du Roselin a ailes rousses Rhodopechys sanguineus en utilisant des donnees de la
morphologie, des mensurations et des vocalisations. Cette espece a traditionellement ete traitee
comme polytypique, comprenant deux taxons, la sous-espece nominale sanguineus d’Asie
occidentale etcentrale, etalienusd’Afriquedunord-ouest. Peu d’auteurs ontreleveles differences
&
morphologiques manifestes entreles deux, quoique Fry Keith (2004) aientrecemmentsuggere
qu’ils puissent constituer des especes phylogenetiques. Nos analyses indiquent que jusqu’a neuf
caracteristiques deplumage separentles males des deuxtaxons (dont quatre sontdiagnostiques et
plusieurs autres presque), et trois caracteristiques peuvent etre utilisees de fa$on fiable pour
distinguer les femelles (dont toutes sont entierement ou quasiment diagnostiques). Nous
decrivons egalement des variations du plumage saisonnieres et liees a Page, chez les deux taxons.
De plus, des donnees morphometriques soumises a l’Analyse en Composantes Principales
indiquent que les deux taxons, et surtout les femelles, sont plutot mieux separes au niveau de la
taille et des proportions qu’on ne le pensait jusqua present. II a ete impossible de faire des
comparaisons adequates entre les vocalisations des deux taxons, peut-etre parce que les
enregistrements disponibles proviennent de differentes saisons, et sans doute a cause des
variations individuelles considerables des cris. Bien que nos resultats exigent un examen
moleculaire, utilisantparticulierementplusieursautresmembersdes roselinsdeszonesdesertiques
comme outgroup’, ils suggerent assez bien qu’il s’agit de deux allo-especes, peut-etre meme deux
especes apart entiere, si on se base surla definition du rang d’espece de Helbig etal. (2002). On
peut trouver egalement une repartition biogeographique identique ou similaire (a celle des deux
taxons de Rhodopechys) chezplusieurs autres formes qu’il semble preferable de considerer comme
specifiquement distinctes (telles que les Fauvettes na—ines africaine Sylvia deserti et asiatique S.
nana). Nos resultats renforcent en outre l’importance apparemment de plus en plus negligee
des specimens conserves dans les musees pour la taxonomie aviaire, a une epoque ou les etudes
moleculaires semblent avoir acquis une importance supreme (Collar 2004).
Summary. Usingmorphology, morphometries andvocalisationswe investigated the taxonomyof
theCrimson-wingedFinchRhodopechyssanguineus, whichhas traditionallybeenviewedasapoly-
typic species, comprising two taxa, nominatesanguineusinwestern and CentralAsia, andalienus
in north-west Africa. Few previous commentators have remarked on the obvious morphological
&
differences between the two, although Fry Keith (2004) recentlysuggested that theymight be
phylogenetic species. Our analyses suggest that as many as nine plumage features separate males
ofthe two taxa (four beingdiagnosticandseveral others nearlyso), and three features can be reli-
ably used to distinguish females (ofwhich all are diagnostic orvirtuallyso). We also describe sea-
sonal and age-related plumage variation in both taxa. Furthermore, morphometric data subject-
ed to a Principal ComponentsAnalysis suggestthat the two are ratherbetterseparated in sizeand
shape than previously thought, especiallyin females. It proved impossible to drawadequate com-
parisons between the vocalisations ofthe two taxa, perhaps due to the available recordings being
from different seasons, and certainly because ofconsiderable individual variation in calls. Our
results demand molecular testing, using especially various other members ofthe ‘desert finches’
as an outgroup, but provide strong indication that two allospecies, perhaps even full species, are
involved, based on the guidelines for assigning species rank ofHelbig etal. (2002). The same or
136-BullABCVol13No2(2006) Taxonomyofthe Crimson-wingedFinch:Kirwanetal.
a similar biogeographical pattern as found in the two Rhodopechys taxa is also evident in a num-
ber ofother forms which seem best considered as being specifically distinct (e.g. African Desert
Warbler SylviadesertiandAsian DesertWarbler S. nand). Our results further reinforce the seem-
ingly increasingly neglected importance of the museum skin in avian taxonomy, in an epoch
where molecular studies appear to have acquired paramount importance (Collar 2004).
C rimson-winged Finch Rhodopechys san- and flanks (though brown in sanguineusand sooty
guineus (Gould, 1838) was described from black in brunneonucha), and pale spots on chest
the environs ofErzurum, in north-easternTurkey. (the latter especially in alienus). Also, they share
Thespeciesisconsideredpolytypic,withthenom- possession ofabifurcated gularpouch, though the
inate form breeding in montane areas from west- occurrence ofthis in fringillids otherthanPinicola
central Turkey and, very patchily, in the Levant, and Pyrrhula is poorly documented (see
somewhat more continuously east to CentralAsia Niethammer 1966). It might also be remarked
and extreme north-west China (in Xinjiang), that some of the North American forms of rosy
whilst the taxon R. s. alienus Whitaker, 1897, finches share almost as many morphological char-
inhabits similar high-altitude areas in north-west acters with Crimson-winged Finch. The impor-
Africa, principally in the Moroccan High Atlas, tantworkofGroth (1998 andinprogress) thusfar
but also extremely locally in the Aures massif of suggests, amongst many other results, that Desert
north-east Algeria (Cramp & Perrins 1994). As Finch occupies a clade containing the many
recently noted by Kirwan & Gregory (2003), the canaries, some Carduelis and Golden-winged
taxonomy ofthe species has rarely been discussed Grosbeak Rhynchostruthus socotranus\ Callacanthis
in the literature, with the only detailed data con- clusters with the mountain finches Leucosticte,
cerning geographical variation being those pre- Carpodacus nipalensis and Pyrrhoplectes epauletta’,
sented by Vaurie (1949) and C. S. Roselaar in and that Crimson-winged Finch and Mongolian
Cramp & Perrins (1994). Even moderately Finch form a separate clade.
detailed specialist works dealing with the cardue- Roselaar (op. cit.) consideredgeographicalvari-
lines have provided only relatively limited and ation in Crimson-winged Finch to be ‘fairly
incomplete discussions of variation within the strong,’ albeit involving colour alone, and, whilst
&
species (e.g. Clement et al. 1993). Indeed, in this study was in progress, Fry Keith (2004)
recent years arguably more attention has focused noted that Rhodopechys sanguineus alienus might
on generic limits and phylogenetic relationships representaphylogeneticspecies, althoughstrange-
amongst the so-called ‘desert finches,’ i.e. ly the illustrations in the same work appear to
Crimson-winged Finch, Desert Finch Rhodospiza show characteristics of nominate sanguineus\ The
obsoleta, Trumpeter Finch Bucanetes githagineus purpose of the present contribution is to draw
and Mongolian Finch B. mongolicus (Groth 1998, attention to the larger differences that exist
Kirwan & Gregory 2005), although it had been betweenR. s. sanguineusandR. s. alienusthanhave
suggested that, alone of these, R. sanguineus is heretoforebeensuspected. (Throughoutwefollow
&
more closely related to Red-browed Finch David Gosselin2002forspellingsofthevarious
Callacanthis burtoni from the Himalayas (see taxa in the ‘desert finches.’)
Desfayes 1969). Some further evidence ofthe lat-
ter relationship was acquired during the course of Methods
GMK
the present study, but this is one of the many acquired mensural data from specimens of
problems confronting workers with Asiatic finch- both relevant taxa (see Table 1), and all others
es (see, e.g., Voous 1977) that demands further within the desert finches grouping, held at the
testing using molecular methods. C. S. Roselaar Natural History Museum (NHM, Tring), as fol-
(in litt. 2006) has pointed out the morphological lows: Callacanthis burtoni (Punjab and north-west
similarities between Crimson-wingedFinchandat India: n-Y], including nine males); Rhodopechys
leastonetaxonofAsianRosyFinchLeucostictearc- sanguineus (Armenia, Syria, Iran, Turkey,
toa brunneonucha. They share a black cap, pale Lebanon, Kazakhstan and Samarkand: «=30,
nape, pink rump, pinkwing-fringing, dark throat including22 males); R. s. alienus (MoroccanAtlas:
TaxonomyoftheCrimson-wingedFinch:Kirwan etal. BullABCVol13No2(2006)-137
Table 1. Means± SD and samplesizesforwing, tail and bill measurementsofthe‘desertfinches,’basedon specimens held
inThe Natural HistoryMuseum (Tring).Anyspecimenforwhichonedatasetormorecould notbe measuredwasexcluded
fromtheanalysisandtable.
Tableau 1. Moyenne±SDetnombred’echantillons pourles mensurationsde I’aile, laqueueetle beedes roselinsdes
zonesdesertiques, basessurdesspecimensdu Natural HistoryMuseum (Tring).Toutspecimen pourlequel unou plusieurs
ensemblesdedonnees ne pouvaientpasetre mesuresaeteexclu de I’analyseetdutableau.
BGG=Bucanetesgithagineusgithagineus, BGA=B.g.amantum, BGC=B.g.crassirostris, BGZ=B.g.zedlitzi, BM=B.mongolicus,CB=
Callacanthisburtoni, RO=Rhodospizaobsoleta, RS=Rhodopechyssanguineussanguineus, RSA=R.s.alienus.
Females Males
Wing(mm) Tail(mm) Bill(mm) N Wing(mm) Tail(mm) Bill(mm) N
BGG 79.33±1.15 50.33±0.58 10.23±0.81 3 83.33±2.67 51.58±2.39 10.31 ±0.43 12
BGA 80.14±2.34 51.14±2.04 10.87±0.31 7 82.43±1.51 53.14±2.27 10.79±0.52 7
BGC 83.6±2.88 54.4±2.63 10.41 ±0.47 10 87.18±2.32 57.27±4.13 10.80±0.28 11
BGZ 83.67±1.53 54.00±1.00 10.50±0.46 3 85.75±1.59 54.65±3.23 10.47±0.31 20
BM 86.94±1.69 55.75±3.71 10.17±0.59 16 88.77±1.88 56.65±3.14 10.07±0.52 26
CB 94.44±1.24 64.50±3.89 15.25±0.84 8 98.94±1.84 67.11 ±3.18 15.38±0.46 9
RO 84.13±1.64 64.25±3.81 11.66±0.46 8 86.36±1.91 64.21 ±3.62 11.79±0.58 14
RS 99.38±1.85 56.63±2.56 13.64±0.92 8 104.05±2.59 60.26±3.26 13.76±0.65 19
RSA 102±1.87 61.40±1.14 13.60±0.43 5 105.57±2.76 64.00±1.15 13.30±0.4 7
n=13, including seven males); Bucanetes mongoli- the latter category being further subdivided into
cus (China, Central Asia and Afghanistan: n=44, good- and good+ (these subdivisions can be con-
including 27 males); B. g. githagineus (all Egypt: sideredasbeingVirtuallydiagnostic’ and ‘diagnos-
«=15, including 12 males); B. g. zedlitzi tic’). Abroad range ofmaterial, pertainingto both
(Morocco, Algeria and Tunisia: 72=24, including forms, was photographed, using a Nikon Coolpix
20 males); B.g. amantum (Canaries: 72=16, includ- 885 digital camera, in indirect natural light (see
ing eight males); B. g. crassirostris (Punjab and Figs. 3-10).
Sind: 72=21, including 11 males); and Rhodospiza
obsoleta (CentralAsia: 72=22, including 14 males). Results —
Specimens were generally sexed according to label Mensural data. A Principal Components
data, but these were checked closely against rele- Analysis (PCA) was performed on the net mensu-
vant literature (Svensson 1992, Cramp & Perrins ral data and the results mapped on both two-
&
1994) and work in progress (Shirihai Svensson dimensional (Fig. 1) and three-dimensional plots
in prep.) in the case ofsuspectidentifications.The (Fig. 2).Thetwo-dimensionalplotconfirmedthat
followingdatawere obtained from each specimen: (as iswell known for these taxa) males aregeneral-
wing (flattened), tail-length and culmen-length ly larger than females, and further revealed that
(to base offeathers), using a standard metal wing- Rhodospiza obsoleta, Callacanthis burtoni,
rule with a perpendicular stop at zero (accurate to Rhodopechys sanguineus and R. s. alienus are all
0.5 mm), and dial callipers (accurate to 0.1 mm). rather well-differentiated taxa, with the two-
Specimens forwhich an incomplete series ofmen- dimensional plot also revealing the close relation-
sural data was available were excluded from the ship that has been suggested between Callacanthis
statistical analysis. andRhodopechys). Plottedthree-dimensionally, the
Notes on plumage variation in both sexes of degree ofseparation between R. s. sanguineus and
the two forms of Crimson-winged Finch were R. s. alienusis particularlyclear, and is rather larg-
taken and, following comparison with those fea- er than that between most other subspecies stud-
tures listed as separating nominate sanguineus and iedwithinthisgroupoffinches.As isevidentfrom
alienus by Roselaar {op. cit.), were ranked accord- Table 1, both sexes of alienus tend to be longer
ing to their usefulness in distinguishing the two. winged and longer tailed than the same sexes of
None of the features discussed by Roselaar was sanguineus, with particularly little overlap in
found to be invalid, buttherewasaclearhierarchy females (perhaps influenced by the smaller sample
in their relative usefulness. Thus, they were arbi- sizes). Culmen-length appears very similar in the
trarilygraded as beingeitheraverageorgood, with two taxa.
138-BullABCVol13No2(2006) Taxonomyofthe Crimson-wingedFinch:Kirwan etal.
andtail -1-5
-3 -2 -1 2 3
smallerwing,billandtail largerwing,billandtail
Principalcomponentaxis1
Figure 1. Results ofaPrincipal ComponentsAnalysisfor Figure2. ResultsofaPrincipal ComponentsAnalysisfor
wing-, tail-andbill- (culmen-) lengthsofalltaxaof wing-, tail- andbill- (culmen) lengths ofalltaxaof‘desert
‘desertfinches’ plottedtwo-dimensionally. finches’ plottedthree-dimensionally.Axis 1 and2 arecor-
Resultats d’uneAnalyseenComposantes Principalespour relatedasin Fig. 1, whilstaxis 3 ispositivelycorrelated
lalongueurdel’aile, laqueueetlebee (culmen) de withwingandtailandnegativelycorrelatedwithbill.
l’ensembledes taxonsdes roselins deszonesdesertiques Resultats d’uneAnalyseen Composantes Principalespour
indiquesdefa9onbi-dimensionnelle. lalongueurde1’aile, laqueueetlebee (culmen) de
BGG =Bucanetesgithagineusgithagineus(12 & 3 ??), l’ensembledes taxonsdes roselinsdeszones desertiques
BGA= B.g. amantum (7 ^^,7 ??), BGC = B.g. cras- indiquesdefat^ontri-dimensionnelle. Les axes 1 et2sont
sirostris(11 10 ??), BGZ= B.g. zedlitzi(20 ^^,3 correles commedanslaFig. 1, tandis quel’axe3 estcor-
??), BM =B. mongolicus(26 16 ??), CB = relepositivementavecl’aileetlaqueueetnegativement
Callacanthisburtoni(9 8 ??), RO =Rhodospiza aveclebee.
obsoleta(14^^, 8 ??), RS =Rhodopechyssanguineussan- BGG = Bucanetesgithagineusgithagineus(12 3 ?'?),
guineus(19 8 ??), andRSA= R. s. alienus(7 ^^,5 BGA=B.g. amantum (7 ^^,7 ??), BGC= B.g. cras-
??) sirostris(11 10 ??), BGZ= B.g. zedlitzi(20 a*0*, 3
??), BM = B. mongolicus(26 aV 16 ??), CB =
Callacanthisburtoni(9 °*°’', 8 ??), RO = Rhodospiza
—A obsoleta(14 <?<?, 8 ??), RS = Rhodopechyssanguineussan-
Pusleufmualgef.or sespadreattaiinlgedmianlTeasbloef2,thneintewfoeattuarxeas,aoref g¥u?i)n.eus(19 0*<A 8 ??), andRSA=R. s. alienus(7 5
which the majority are better than average and
fourarerankedas extremelygood (i.e. diagnostic).
Fewer features, just three, separate females of is at least one summer (July) record of birds in
alienus from sanguineus, but all are diagnostic or suitable breeding habitat.
virtually so (Table 3). As noted byVaurie (1949), —
RoselaarinCramp &Perrins (1994), andRoselaar Vocalisations. The vocal analysis was performed
(1993), there is no evidence ofgeographicalvaria- by AvdB and Magnus Robb. Approximately 30
tion in either plumage or size within Asian popu- recordings belonging to the two taxa, from
ltahtoiuognshtofthRa.tsA.zsearnbgauiijnaenuis,biarldtshomuigghhtVatuernidet(o19h4a9v)e TMohreocScoou,ndTuArpkperyoaacnhd, GUeoKrgi/aN,etahnedrlaanrdcshi,veddatai-n
a slightly longer tail. We have not examined any base, were included in the comparison. Individual
specimens from Algeria and are thus unable to variation proved to be considerable, to the point
comment as to the presence (or not) ofanyvaria- where any real differences between the two are
tion within alienus although Clement et al. obscured. Like several other carduelines, it seems
,
(1993) erroneously suggested that the species that individual pairs may produce their own vari-
might only be a winter visitor to this region. In ants of certain calls, at least during the breeding
&
fact, as noted byIsenmann Moali (2000), there season. Furthermore, the available recordings of
Taxonomyofthe Crimson-wingedFinch:Kirwanetal. BullABCVol13No2(2006)-139
Figures 3-4. Dorsalandventralviewsofthesamespecimensofadultmales ofRhodopechyssanguineussanguineus
(upper) andR. s. alienus(lower) inworn (spring) plumage (GuyM. Kirwan, ©TheNatural HistoryMuseum,Tring)
Vuesdorsales etventralesdesmemesspecimens demalesadultesdeRhodopechyssanguineussanguineus(enhaut) andR.
s. alienus(enhas) enplumage use (printemps) (GuyM. Kirwan, ©TheNatural HistoryMuseum,Tring)
Figures 3-6. Dorsal andventralviewsofthesamespecimens ofadultfemalesofRhodopechyssanguineussanguineus
(upper) andR. s. alienus(lower) inworn (spring) plumage (GuyM. Kirwan, ©TheNatural HistoryMuseum,Tring)
Vues dorsales etventrales des memesspecimens defemelles adultes deRhodopechyssanguineussanguineus(enhaut) and
R. s. alienus(enhas) enplumage use (printemps) (GuyM. Kirwan, ©TheNatural HistoryMuseum,Tring)
Figures 7-8. Dorsal andventral viewsofthesamespecimensofadultmales ofRhodopechyssanguineussanguineus
(upper) and R. s. alienus (lower) in fresh (autumn) plumage (GuyM. Kirwan, ©TheNatural HistoryMuseum,Tring)
Vuesdorsales etventralesdes memesspecimensdemales adultes deRhodopechyssanguineussanguineus(enhaut) andR.
s. alienus(en has) en plumage frais (automne) (GuyM. Kirwan, ©TheNatural HistoryMuseum,Tring)
140-BullABCVol13No2(2006) Taxonomyofthe Crimson-wingedFinch:Kirwanetal.
Figures 9-10. Dorsalandventralviews ofthesamespecimensoffirst-summermalesofRhodopechyssanguineussan-
guineus(upperthree) andR. s. alienus(lowerbird) showinginter- andintra-taxonvariation (GuyM. Kirwan, ©The
NaturalFlistoryMuseum,Tring)
Vuesdorsales etventrales desmemesspecimensdemalesde leretedeRhodopechyssanguineussanguineus(les trois d’en
haut) andR. s. alienus(l’oiseauduhas) illustrantlesvariations inter-etintra-taxons (GuyM. Kirwan, ©TheNatural
HistoryMuseum,Tring)
Figures 11-12. MaleRhodopechyssanguineusalienus, Oukaimeden, Morocco, 5April2005 (Arnoud B. vandenBerg)
Rhodopechyssanguineusalienus, male, Oukaimeden, Maroc, 5 avril2005 (Arnoud B.vandenBerg)
Figure 13. FemaleRhodopechyssanguineusalienus, Figure 14. MaleRhodopechyssanguineussanguineus,
Oukaimeden, Morocco, 5April2005 (Arnoud B. van Kazbegi, Georgia, 22June2005 (Rene Pop)
denBerg) Rhodopechyssanguineusalienus, male, Kazbegi, Georgie,
Rhodopechyssanguineusalienus, femelle, Oukaimeden, 22juin2005 (RenePop)
Maroc, 5 avril2005 (ArnoudB. vanden Berg)
Taxonomyofthe Crimson-wingedFinch:Kirwan etal. BullABCVol13No2(2006)-141
Table2. Relative hierachyinthosefeatures listed by Roselaar(inCramp& Perrins 1994) as usefulforseparating malesof
Rhodopechyssanguineussanguineusand R.s.alienus, onthe basisofspecimen examination ofmaterial held in
The Natural HistoryMuseum (Tring).
Tableau 2. Hierarchie relativedescaracteristiquesmentionneesparRoselaar(en Cramp& Perrins 1994)commeetant
utiles pourseparerles malesde Rhodopechyssanguineussanguineusetde R.s.alienus, surlabasede I'examendespeci-
mensconservesau Natural HistoryMuseum (Tring).
Feature Ranking Comments
R.s.alienushasonlyveryindistinctblack good+
onmantle
R.s.alienuslackspinkinuppertail-coverts good+ asnotedbyRoselaar,theremaybeavinouswashinfreshplumage
R.s.alienuslacksblackspottingonear- good+ seealsocommentbelow*
covertsandbreast
R.s.alienushascentralchinandthroat good+ nominatehasthroatmoreorlessconcolorouswithbreast,andthebrown
grey-whitetingedrosy,withnarrowbrown markingsonlowerbreast/bellyandflanksaremoreextensivethaninalienus
upper-breast-band
R.s.alienushastailmoreextensively good-
dark(lessblack)onouterfeathers
R.s.alienushaslessredontheflight- good- entirewingseemstoshowlesspinkandwhiteelements,butthereissome
feathers overlapwiththenominate
R.s.alienushaslessredonfaceand average/ somehaveforesuperciliummarkedwithred(e.g.NHM 1937.12.28.27)
noneonforesupercilium good
R.s.alienushasblackofcrownmore average
restrictedtoforepart
R.s.alienuslacksrufousonback, mantle average thisfeatureisnotalwaysapparentinR.s.sanguineus
andhead-sides
*Veryworn R.s.sanguineuscanshowquitesomepinkontheunderparts,butR.s.alienusnevershowssuchcoloration(e.g.NHM
1949.Whistler.8802).
alienusweremade atadifferentstageofthe breed- variation, both age-related and seasonal. For fur-
&
ing season to those of nominate sanguineus. therdetails see Shirihai Svensson (inprep.), and
Furtherworkon this issuewould be interesting to also Figs. 3-10.
perform, nonetheless, especially to record and
compare the vocalisations of the two in winter, R. s. sanguineus
Sexes moderately differentiated, mainly in
when their repertoires are presumably less exten-
spring/summer, otherwise seasonal plumagevaria-
sive and flock/species cohesion is more important
tion rather limited and mostly due to wear. Post-
than pair cohesion or individual advertisement.
breedingmoult (complete) andpost-juvenile (par-
Even so, even ifclear-cut differences did become
tial) moult chiefly in August-September, but pre-
evident as the result of such fresh analysis, this
would not necessarilyserve as anythingother than breeding moult (adult and first-year) is apparently
absent. SPRING Worn. Adult cf Wear increases
interesting additional support for regarding the
two taxaasspecies. Furthermore,wehavenotcon- contrast ofuniform black cap, face pattern (espe-
cially pale supercilium and collar), rose tone to
ducted playback experiments to ascertain any
lower back/rump and red at bill base, on lores and
measureofresponsivenessofonetaxon to thesong
ofthe other. In any case, we note that such tests around eye; also breast and breast-sides more
rufous-cinnamon and black-spotted central breast
do not provide conclusive proof one way or the
and flanks, and red basal areas on wings and tail
other’ (Helbig etal. 2002).
are more conspicuous, but (in both sexes) white
tips to remigesvirtuallywearoff. Afewalso devel-
Age-related and seasonal plumage variation op a slightly pale pinkish cream-brown throat but
Based on examination of specimen material at not to the same extent as some first-summer
NHM
we present the following notes on plumage Bill changes from greyish to dull warm yellow
142-BullABCVol13No2(2006) Taxonomyofthe Crimson-wingedFinch:Kirwanetal.
Table3. Relative hierachyinthosefeatures listed by Roselaar(inCramp & Perrins 1994) as usefulforseparatingfemalesof
Rhodopechyssanguineussanguineusand R.s.alienus, onthe basisofspecimen examinationofmaterial held inThe
Natural History Museum (Tring).
Tableau 3. Hierarchie relativedescaracteristiques mentionneesparRoselaar(en Cramp & Perrins 1994) commeetant
utiles pourseparerlesfemellesde Rhodopechyssanguineussanguineusetde R.s.alienus surlabasede I'examen de
,
specimensconservesau Natural HistoryMuseum (Tring).
Feature Ranked Comments
Throatandbreastpatternisasmaleand good+ femaleR.s.alienushassomepalespotsinbreast-bandandear-coverts,
differsfromnominate whereasfemaleR.s.sanguineushasmoretawny-brownonflanksandlower
breast-sides,andhassolidtawny-brownear-coverts, breastandthroat,whilst
someevenshowaslightlypinkfore-face(nevershownbyalienus)
R.s.alienushaslessblackincrown good-
R.s.alienushasupperpartsgreyerandrump good-
lesspink/white
when breeding. In the hand, r5 white, except Black crown patch is narrowly fringed cinnamon-
broad but incomplete dark subterminal field on buff whilst carmine-red of face is duller, upper-
innerweb (concentratedonouterpart) andalmost parts more buff/rufous-brown, less heavily
entirely dark outer web, whilst r6 is white, except streaked but broadlyfringed, and rump/uppertail-
usuallyfor the black shaft and, occasionally, a dif- coverts tinged pale rosy-pink, although often con-
fuse andnarrowdarkareaon the edgeofthe inner cealed. Whitish band between breast and flanks
web. Adult ? Duller than cT, with paler crown and upper bellywashed pale pink, with yellowish-
patch, reduced capped appearance, more white buff breast and flanks tipped whitish and
and duller pink in wings and virtually no pink in unstreaked or virtually so. Adult ? Much as
tail, and has whiter, less rufous and less spotted spring, but broad greyish-buff fringes to crown,
underparts (some have chin and throat cream upperparts and upperwings, and plumage even
white). The central remiges and primary-coverts, duller. In both sexes pale fringes towing- and tail-
especially, are more narrowly fringed paler pink- feathers are broader, with carmine-pink and red
red (with browner and less obvious centres); r5 basal area partially concealed (thus overall wing
white as in c? but has broader and more complete patternless contrastingthan in spring). Bothsexes
black subterminal field, and r6 has broader and differ from first-winter in being evenlyfresh, with
darkerareaalongedge ofinnerweb andattip, not broader and whiter primary tips. First-winter
uniformwhite. Mantle andscapulars brownerand (both sexes following post-juvenile renewal of
much less heavily streaked, and lower back to head, body, lesser and median coverts, perhaps
uppertail-coverts paler grey-brown, tinged some inner greater coverts and tertials, and a few
isabelline, whilst lores and eye surround greyish apparently replace even some inner primaries,
cinnamon-buff(usuallyalmost no red), with paler primary-coverts, secondaries andrl.) cThas overall
ear-coverts and less contrasting supercilium. Bill plumage like fresh adult but stronglyapproach-
greyish-horn (much less yellow). First-summer es ?, thus sexual dimorphism obscured. Adult
Very like respective adults and best aged by head and underparts patterns strongly reduced
retained juvenile wing- and tail-feathers, with and have more extensive pale flecks (chin/throat
moult limits as first-winter, but even more con- variable), upperparts essentially warm brown and
trastingly worn, and extent of subterminal black paler rump than adult. Pink and white areas in
areas in r5-6 as latter. Much individual variation, remiges and rectrices also generally duller or
especially in with some approaching adult ? reduced. Retained juvenile greater coverts have
,
inoverallcoloration (butusuallysafelysexedusing blackish-brown inner web and brown-buff on
same criteria as for adults). Some have, to a vary- most ofouterweb, except narrowpinkfringe; the
ing degree, a white throat and upper breast with retained primary-coverts are mostly dull brown
almost unstreaked breast and flanks, and thus with narrow fringes. Tail has more extensive dark
approach R. s. alienus. AUTUMN Fresh. Adult areas than in adult and pattern approaches
TaxonomyoftheCrimson-wingedFinch:Kirwanetal. BullABCVol13No2(2006)-143
.
adult ? First-winter ? Much like first-winter can approach adult ? inoverall plumage.AUTUMN
.
but plumage paler with generally much-reduced Fresh. Adult Black crown patch narrowly
pink in wing and fringes to greater coverts mainly fringed cinnamon-buff and duller, pinkish facial
buffwith limited or no pink; terminal halfofr5 areas reduced or lacking, upperparts slightly more
blackexceptdiffusewhitetip on innerweb, andr6 huffish, and brown-buff breast and flanks more
also has slightlymore extensive darkthan adult ? obviouslytippedwhitish.Adult ? Muchasspring.
Both sexes differfrom ads in having retainedjuve- Both sexes differ from first-winters in being even-
nile primaries and tertials browner and less fresh ly fresh with broader, more solid and whiter pri-
with narrower, less sharply defined and less pure marytips. However, incomparisonwithadultsan-
white tips; tail-feathers obviously pointed. JUVE- guineustheprimarytips arenarrowerandlesspure
NILE Soft fluffy bodyfeathering is generally rather white (whitish-cream with a pale buff-brown
&
sandy or sandy-brown, with few dark feather cen- tinge). First-winter Like fresh adult but over-
tres,verylittlepinkvisibleonclosedwing, andbill all plumage strongly approaches ? (especially
is dark horn-yellow becoming brownish at tip. adult), thus sexual dimorphism obscured.
However, unlike most first-winter ?? crown
,
R. s. alienus darker and more clearly defined, and some have a
Sexes generally less strongly differentiated than in hint of grey on nape (lacking in first-winter ?),
R. s. sanguineus but otherwise they seem to have whilst pink edges of wing substantially broader
very similar moult and ageing characteristics. and brighter. Retained juv greater coverts have
SPRING Worn. Adult Wear increases contrast of brown inner webs and brown-buff on most of
black cap, whitish tips to remiges virtually wear outerweb; the retained primary-coverts are most-
off, and pale areas of face become slightly more lyblackish brown exceptforathinpinkish edgeto
pronounced (often with slight pinkish hue to outer web. Tail (retained juvenile feathers) has
cheeks), but much more limited seasonalvariation more extensive dark areas than adult and pat-
than in R. s. sanguineus. Like latter, pink-red basal tern approaches adult ? with even more diffuse
areas to wings and tail more exposed and bill and huffier tips, and reduced (or virtually lacks)
changes from greyish to dull warm yellow when pale wedge on inner web of r6. First-winter ?
breeding. In thehand, unlikeR. s. sanguineus, r5 is Much like first-winter & but paler with strongly
mostly dark/black except forwhitish tip, and thus reduced dark in cap, no grey on nape, reduced
lacksdarksubterminalfieldto innerweb,whilstr6 pink in wing and has fringes to greater coverts
isalsoalmostcompletelydark/black, includingthe mainly buff; tail pattern variable, as in first-winter
outer web, except for an usually sharp wedge on orwitheven moreobscurepaleareas. Bothsexes
the inner web. Adult ? Very similar to (many differ from adults in having retained juvenile pri-
probably impossible to sex in the field), but over- maries and tertials, which are browner basally, less
all duller with paler and less solid crown and fresh and have considerably more diffuse and
much-reduced greyish nape; also white throat buffer tips; tail-feathers distinctly pointed. JUVE-
patch slightly less sharply defined and cheeks NILE Not examined but probably close to R. s.
mostly lack pinkish; underparts virtually identical sanguineus.
to o", but remiges (especially central part) and
primary-coverts more narrowly fringed palerpink. Discussion
&
Tail pattern recalls but pale tips rather conspic- Allopatric taxa, as noted by Helbig et al. (2002),
uously reduced and these and wedge on innerweb always present particularlyproblematic caseswhen
ofr6 diffuse and sullied pale buff-brown. Bill gen- endeavouringto ascertainwhethertheysuchforms
erally duller with less yellow tinge. First-summer should be regarded specifically, foras these authors
Very like respective adults and best differentiated succinctlystate: Assignmentofspeciesrankinsuch
by retained juvenile wing- and tail-feathers, and caseswill necessarilybebasedon hypothesis, rather
moult limits as in first-winter. Extent ofsubtermi- than on proven facts.’ In the present case, as dis-
nal black areas in r5—6 also as first-winter. Due to cussed elsewhere (e.g. Kirwan & Gregory 2005),
reduced dark cap and pink in wing, both sexes are the two taxa concerned may well prove to be the
lessstronglypatterned then adults and are less eas- soleconstituentsofthegenusRhodopechysandthey
ily sexed, whilst especially some first-summer areclearlyratherclosein general morphology, ecol-
144-BullABCVol13No2(2006) Taxonomyofthe Crimson-wingedFinch:Kirwanetal.
5
ogy and habits. Nonetheless, they are also easily Nonetheless, we believe, as this paper demon-
diagnosable in virtually all plumages, with only strates, that the museum skin continues to hold a
first-summer males liable to anyconfusion at their high value in avian taxonomic studies and, like
only point of contact, the museum cabinet! In Collar (2004), we bemoan the current trend to
addition, females of the two forms, especially, seemingly regard molecular tools as the only
clearly separate using a multivariate statistical means to adequatelyprogress such research.
analysis of mensural data (see Table 1, and Figs.
1-2), andthedegreeofsegregationisquitemarked Acknowledgements
in comparison to that exhibited bywhat have tra- Wearegratefultothefollowingforvariouskindness-
ditionally been viewed as closely related taxa, esassociatedwiththepreparationofthismanuscript:
although Groth (1998) found Rhodopechys sensu Vasil Ananian, Goran Ekstrom, Andrew Grieve,
lato to be polyphyletic. In sum, again bearing in Giuseppe Micali, Rene Pop, Sandy Rae, and Mark
mindtheguidelinesofHelbigetal. (2002) itseems Adams,Alison Harding, Robert Pfys-Jones, Michael
that the two Rhodopechys demand recognition Walters and Effie Warr (of The Natural History
under any ofthe pattern-defined species concepts Museum, Tring). Magnus Robb assisted with the
currently operating (see Sluys & Hazevoet 1999) vocal analysis. Ron Demey and Fran^oise Dowsett-
Lemaire prepared the French summary. Kees
and have certainly achieved allospecies status, but
Roselaar refereed the manuscript. GMK’s field stud-
whether theyhave achievedfull species rankunder
the modern definition of the Biological Species iesinTurkey(andelsewhere)wereabettedbyanum-
ber ofadditional people and organisations that are
Conceptmustprobablyawait the results ofmolec- &
fully credited in Kirwan Konrad (1995) and
ular analysis. In contrast, the two taxa discussed Kirwan et al. (2000). GMK is grateful to Roy &
here would surely be recognised as species under Moira Hargreaves, and John & Effie Warr for
the framework of the Metapopulation Lineage accommodation duringhis visits toTring.
Concept ofspecies (or General Species Concept),
application ofwhich it was argued recently by de References
fQyuienigrozh(o2w005m)ondoetronn-ldyapyrovbiidoelsogaismtesansdioafgnuonsie- ClemSepnatr,rowP.s,:HaArnrisI,deAn.tif&icaDtaivoins,GuJ.id1e9.93L.onFdinocnh,estaJnK:d
‘species,’ but also returns more closely to Mayr’s ChristopherHelm.
original conceptualisation of what constitutes a Collar, N. J. 2004. Species limits in some Indonesian
species, rather than merely focusing on the attrib- thrushes. Forktail20: 71-87.
ute ofreproductive isolation. Cramp, S. & Perrins, C. M. (eds.) 1994. The Birds of
Like several other taxa recently assigned the Western Palearctic. Vol. 8. Oxford: Oxford
allospecies or full species status, the distributions UniversityPress.
of the two finches discussed here accord rather David, N. &Gosselin, M. 2002.Thegrammaticalgen-
wellwith an increasingly recognised biogeograph- der ofavian genera. Bull. Br. Ornithol. Club 122:
ical phenomenon under which North Africa is 257-282.
viewed as something of a refugium for endemic Desfayes, M. 1969. Remarques sur les affinites des
taxa. In several cases, e.g. the two DesertWarblers Fringilli&des des genres Rhodopechys et Callacanthis.
cSlyolsveisatdeesxetratnitanredlaSt.ivneanisar(eSshtirriichtaeidettoalC.e2nt0r0a1l),Astihae Fry,OCi.seHau. &R.KFe.iOt.h,39S:.2K1.-2(7e.ds.) 2004. The Birds of
and the Middle East. As already noted for the Afica. Vol. 7. London, UK:AcademicPress.
genus Sylvia but probably for many other addi- Groth,J. G. 1998. Posterabstracts 118:Molecularphy-
, logenyofthecarduelinefinchesandHawaiianhon-
tional groups, quite plausibly even the genus
eycreepers. Ostrich69: 401.
Rhodopechys the long drought known as the
MMeYssAinainadn wCh,riiscihs,lewdhtiochthepeenatkierde Isroanmoe-Tu5.r5a-n8i.an Helb&ig,CAo.llJi.n,sKonn,o1xM,.A.2G0.0,2.PaGrkuiind,elDi.neTs.,foSrangasstseiDrg,niOGn.g
species rank. Ibis 114: 518-525.
region becoming extremely dry and to the Isenmann,P. &Moali,A.2000. OiseauxdAlgerie. Paris:
Mediterranean shrinking in size (Sue 1984, Societe d’Etudes Ornithologiques de France.
Tchernov 1988), may have played an important Kirwan, G. M., Bechtolsheim, M. & Willig, S. 2000.
role in this process. Again, to some extent the Distribution ofMongolian Finch inTurkey. Dutch
answers to such enigmas lie in genetics. Birding22: 149-150.
TaxonomyoftheCrimson-wingedFinch:Kirwanetal. BullABCVol13No2(2006)-145