Mycerobas genera and all other species represented in
figure 2 and 3. It showed the same placement in all trees
(data not shown). The genetic distances (maximum like-
lihood) from hawfinch to Loxia, Pyrrhula, Carpodacus,
Mycerobas, and Eophona were relatively high and very
similar (between 0.09–0.12). Mean distances between
Carduelis, Pyrrhula, Carpodacus,
Mycerobas
and
Eophona were placed between 0.001 and 0.03. Therefore,
hawfinch molecular phylogeny supports the classifica-
tion as a single and separate species from other Cardu-
elini finches. However, the entire mt cyt b DNA gene
should be tested for further assessment of hawfinch phy-
logeny, particularly with possible relatives, such as the
New World Hesperiphona finches [30].
The genera Eophona and Mycerobas are closely related
in all trees with high bootstrap and CP values (figs 2, 3).
Whether they should be considered a single or two differ-
ent genera is unclear, and the two missing Mycerobas
species (M. icteroides and M. melanozanthos) should be
tested in the future. The phenotypical relationship be-
tween Mycerobas and Eophona genera had been estab-
lished previously [30].
Grosbeaks seem to have appeared on Earth earlier than
the genera Carduelis, Loxia, Carpodacus and Pyrrhula in
accordance with the hypothesized timing in figure 2 (14
MYA).
4) Rosefinches (genus Carpodacus, 21 species)
Most rosefinches are closer to the genus Carduelis than
the others birds studied here. Carpodacus nipalensis
(dark rosefinch) is an outlier from the group. The head,
beak, and body characteristics are more like those of a
chaffinch (Montifringilla sp.) than a rosefinch (unpub-
lished results). It clusters in figure 2 (but not in fig. 3)
with the house finch (Carpodacus mexicanus) because
they are both outliers and long branches may attract each
other in this type of analysis [32]. Furthermore, it is re-
markable that the only American species studied in this
work (house finch) is an outlier; it is separate from Asian
rosefinches and it seems more closely related to genus
Carduelis. When more rosefinch species are added to the
dendrograms (figs 2, 3), the house finch may possibly
become more integrated in the Carpodacus group. It
shows no relationship with Old or New World sparrows
[33]. This is more doubtful for C. nipalensis, which has
such different phenotypic characters, and its relation-
ships with geographically closer Montifringilla species
should be studied in the future. Uragus sibiricus, which
was classified as the only species belonging to the genus
Uragus because it has a rosefinch body and a Pyrrhula
beak, is definitely a sister species of Carpodacus rubi-
cilloides. It may have changed its beak from that of C.
rubicilloides (or its ancestor) as an adaptation to eat
mainly buds, like Pyrrhula. This is another example of
beak change in a relatively short time (figs 2, 3) [29].
The relationship of U. sibiricus with Urocynchramus
pylzowi (pink-tailed rosefinch) should be studied in the
future because of their phenotypic similarities and close
geographical range (more limited for the latter, in central
China).
Conclusions
1) The time of appearance of the species of crossbills,
bullfinches, grosbeaks, and rosefinches studied is sug-
gested in figure 2; the species may have arisen between
the Miocene and Pliocene. However, although concor-
dant times have been found for Serinus and Carduelis [4,
5], these early timings for passerine emergence are ac-
cepted by some [8, 10, 27] but not by all [28] authors.
2) Crossbills (genus Loxia) are integrated within the
genus Carduelis, together with redpolls. The common
crossbill shows subspeciation with L. japonica in the
Pleistocene epoch.
3) Pine grosbeak (P. enucleator) groups with bullfinches
and is probably one of the most ancient finches of the
genus Pyrrhula, together with Pyrrhula nipalensis (figs
2, 3).
4) Hawfinch (C. coccothraustes) is not related to any of
the bird groups studied in the present work (in particular
not with Eophona or Mycerobas, which were considered
by some investigators to be grosbeaks together with the
hawfinch [30]). The phenotypic separation of the genus
Coccothraustes is supported and its probably close rela-
tionship with the New World Hesperiphona species
should be studied. Mycerobas and Eophona birds are
closely related, the latter being more ancient and possible
the common ancestor to both genera.
5) The long-tailed rosefinch (U. sibiricus) is a sister
species of Carpodacus and is not related to Pyrrhula
species. Its probable relationship with U. pylzowi should
be studied in the future.
6) The large Haematospiza sipahi is included within the
genus Carpodacus.
7) C. nipalensis, a Fringilla- or Montifringilla-like finch,
is not included within genus Carpodacus.
Acknowledgements. We are indebted to the following Spanish or-
nithologists: Bernardino Yebes, Gloria Gardó, Francisco Mira Chin-
chilla, Arturo Fernández Cagiao, and Alvaro Guillén. Alberto Gar-
cía helped in designing the figures. This work was supported in part
by grants from the Spanish Ministry of Education (PM95-57,
PM96-21, and PM99-23) and the Madrid Regional Government
(06/70/97 and 8.3/14/98).
1 Sibley C. G. and Ahlquist J. E. (1990) Phylogeny and classifi-
cation of birds. Yale University Press, New Haven, Conn.
2 Tordoff H. B. (1954) A systematic study of the avian family
Fringillidae based on the structure of the skull. Misc. Publ.
Mus. Zool. Univ. Mich. 81: 7–41
CMLS, Cell. Mol. Life Sci.
Vol. 58, 2001
Research Article
7
