Lok A LR A te Pd -
eft ic sander tem toe p ' ~ vw)
LS DON ELIT ery “7 4 - - ri ~ae
FRIST G Se ist fe PZ Pees, daar By " - 2 es l- ya se a a ae
7 y PO OT RP 4 ee: pot A ai Wy o. } ‘. er o he =" .~
5 : ‘ . EE ie Ee eT een ae Cee
: it neh Recta an bhi tae te Dg 5 Preteen eer eke eng ere see ""
Ae cae 8 hall £0% ha atycs Ne Oe es Care Prey Pe eS 5h aay The even el Pieclntad mea Sab cite Ba. 9-8. ty
: dn 0 Sl aberecwecsvere: West yw aw mt patrares es.
ree eee eo Pees TT: teas ree ental y, Yariot nes SPAM DSS ILe a MAR BATE MNO E A TS ATR
A ms Lape ea ee te peecun pre nation Were STS) Serer ee eee Sn CE EES Re ele eH Ie
Apts Ps re ea ehh ek re om sk EN AY KW Se m a soenctamae eA Rar erg eee Seek eee
7 ee ees ere Te ee eee aa Rie Ra ttneek akee lo toes rintiasioeeeinciaeh
pers Y pcr toe yee se BRIA SRD Wet, Rot Ae he a gmarsn a HTN Is Ba
: Ses e Nee Ae
ete ek dental

ne
arent tet Ts)
> La tT

pee
Leta ashe! Peek eS
Dactty ye e

of gambit rennet tn Mine Sn on de
OO OE ca hii han Mnin tins SABE
. Aik tone ty et F OR POM
Ae 2 hel Arn rhe A 2 BOT,
was high ira #
epee *

ee

.
ed Ob nace’ wm ty Saw A Made

pe a
pee mae

Perr a reo

eae hn oh eek reef at
en pace Rate permet ome

yee See Se

é ae

La ee

dP Mt

Lad Sond © Be RAP
tt te

<n

a Te, So a a
“5 Tn Deeatat, Boa ASE ears

SO ee en eee ete

LIBRARY OF THE
UNIVERSITY OF ILLINOIS
AT URBANA-CHAMPAIGN

590.5
a

NATURAL HIS I ORY
SURVEY

Sa

Sr ee a ee ee ai

names — int tsetse ail et

zo ; aS ae
et hee ha ates Dio LOR ee i aap ME? = a a 7
ree i rein: = e av : A.

hens

A aevetrwst kh 88. CFH Eg CSTRESPAIMA — Gof tii Filial

Teen hk Bb UF b, UNDANA-LHAMPAIH SY

FO-S Nast Hist Surv

_— hi {

N 2
S| FIELDIANA - ZOOLOGY?
ec S] é PIPRADY
- “ Published by mit
i: FIELD MUSEUM OF NATURAL HISTORY
Volume 51 NOVEMBER 14, 1968 oe a ‘No. 9

Intra-Relations of African Canaries,
Genus Serinus

AUSTIN L. RAND
CHISF CURATOR, ZOOLOGY, FIELD MUSEUM OF NATURAL HISTORY

In preparing the section of African canaries for a forthcoming
volume of Peter’s Check-List of Birds of the World, 1 found a welcome
guide in White’s 1963 Revised Check-List of African Flycatchers. . .
Finches ... and Wazxbills. However, I have made some changes in
the arrangement of the species, and a few in species and subspecies
limits.

Before going on to detail, the list of 25 species is presented with,
where appropriate, the more aberrant subspecies included. Some of
the latter have been considered, and perhaps will prove to be sep-
arate allopatric species. The 25 species recognized are arranged into
six groups to indicate relationships and to aid discussion.

AFRICAN SERINUS
Group I
1. canicollis (including flavivertex)
2. nigriceps
Grovp II

3. citrinelloides
4. frontalis

5. capistratus
6. koliensis

7. scotops

Group III

8. leucopygius
9. atrogularis (includes reichenowi, xanthopygius, and rothschildi;
flavigula is considered a mutant)

Library of Congress Catalog Card Number: 68-59025
No. 1060 125

.' ey
etil Uli a |

1269

126 FIELDIANA: ZOOLOGY, VOLUME 51

10. citrinipectus (of hybrid atrogularis X mozambicus origin)
11. mozambicus

12. donaldsoni

13. flaviventris (includes maculicollis and dorsostriatus)

14. sulphuratus

15. albogularis

Group IV
16. gularis (includes reichard2)
17. mennelli
18. tristriatus
19. menachensis

Group V
20. striolatus (includes whytiz)
21. burtoni (includes melanochrous)
22. rufobrunneus
23. leucopterus

GrouP VI

24. totta (includes symonsz)
25. alario (includes lewcolaema)

This list is not the only arrangement possible, but is a compromise.
Notably, Group II might have been placed first, Group V might have
preceded Group IV. But, this is an attempt to put first the forms
most like the Eurasian Chloris—Serinus—Carduelis—Spinus group, from
which the African forms presumably originated, and place last, the
most different African species without obvious close relatives else-

where.

In making this arrangement, like White, I found it impractical to
use the four genera in which Sclater (1930, Systema Aviwm Aethiopi-

carum) grouped them:

1. Spinus for the birds with wedge-shaped, fine-pointed bills

(vs. stubby or heavy bills with curved culmen):
nigriceps, frontalis, citrinelloides and totta.

b. Serinus for the predominantly yellow-green species:
mozambicus, sulphuratus, ete.

c. Poliospiza for the predominantly grey-brown species:
tristriatus, atrogularis, gularis, burtoni, albogularis, ete.

d. Alario for the one black, chestnut and white species:
alario.

RAND: AFRICAN CANARIES, GENUS SERINUS 127

The other proposed genera, including the eight genera or sub-
genera proposed since 1920, are not recognized.

Certain lines of evident species relationship cut across that indi-
cated by the unit characters used by Sclater. General similarities in
plumage and pattern seemed the most useful in arranging the species.
In regard to distribution, it must be remembered that almost all are
non-forest birds in Africa, and thus range patterns exclude for the
most part the tropical rain forests of the Congo and West Africa.

The six groups and their species are discussed below. The gen-
eralized range is given for each species.

GrRoupP I

1. canicollis—EKastern Africa: Eritrea to Cape Province; west to
Angola.

2. nigriceps—Northeast Africa: (Ethiopia) above 8,000 feet.

Both are predominantly yellow-green species with rump only
slightly yellower than back; canicollis (including flavivertex) is a rather
plain species with a stubby bill; head pattern restricted to a dark area
through eye and a yellowish forecrown; the female is duller, plain
below; but the immature is more brownish above, whitish below and
quite heavily streaked, both above and below. It appears to be dis-
tantly related to S. canarius and S. serinus. The conspicuous geo-
graphical variation includes: grey on back and sides of neck in the
South African S. c. canicollis, lacking in other races; and tail being
yellow or black: six subspecies.

S. nigriceps is also a green-yellow bird, but with a black “‘hood”’
(i.e., head and neck) in the male. It has a rather more slender bill
than caniceps and slightly larger tarsus and toes and has been put in
Spinus. Nosubspecies. While those two species are not particularly
closely related, they seem as close to the Eurasian members of Seri-
nus-Spinus group as to any of the African members, and thus are
placed first.

Group II

3. citrinelloides—Eastern Africa: Eritrea to Kenya, eastern North-
ern Rhodesia and northern Portuguese East Africa (3 sub-
species).

4. frontalis—Eastern Africa: Uganda and adjacent Congo, to
northeastern Northern Rhodesia (? and subspecies Angola?)

5. capistratus—Gabun to Angola and south and east of the for-
ests to north end of Lake Tanganyika.

128 FIELDIANA: ZOOLOGY, VOLUME 51

6. koliensis—East Africa: eastern Congo to western Kenya.
7. scotops—South Africa.

These five species of small, yellow-green birds with rump a little
yellower than the back, form a closely intra-related group despite the
differences in bill shape and in sexual dimorphism. The bill is thin-
nest and most pointed in frontalis, and becomes progressively more
stubby and heavier through citrinelloides, capistratus, koliensis, and
scotops. Sexual dimorphism in color is pronounced in frontalis, capi-
stratus, and some populations of citrinelloides, in all of which the
males have a black mask (lacking in all females) and, correlated with
this, are unstreaked below. The females of citrinelloides and capi-
stratus are heavily streaked below, while that of frontalis is plain
yellow. Sexual dimorphism in color is much less in koliensis, scotops,
and some populations of cztrinellocdes, in which the black mask is
reduced to dusky or greyish, or is absent, and both males and females
are heavily streaked below.

Needless to say, the arrangement of these species has been vari-
ous, citrinelloides and frontalis have been put in Spinus at times in the
past, the others in Serznus. In recent treatments of the central Afri-
can forms, citrinelloides and frontalis have been considered conspecific
as have capistratus and koliensis, while Chapin (1954, Bull. Amer.
Mus. Nat. Hist., 75B, pp. 606, 608) says that in eastern Congo where
both frontalis and capistratus occur, but not together, their behavior
is similar and if it were not for the differences in bill shape, he would
consider them conspecific.

It is perhaps significant that the two most different central Afri-
can forms, frontalis and koliensis, are the only ones with wide over-
laps in ranges. What is found out about the manner in which citri-
nelloides meets frontalis and koliensis in western Kenya and eastern
Uganda will probably influence our species concepts here. In the
meantime, it seems advisable to keep them as species.

As in Group I, the relationship of this group to the Serinus—Spinus
group of Eurasia seems evident, but cannot be pin-pointed.

Group III
8. leucopygius—northern Ethiopian region: Senegal to Eritrea.

9. atrogularis—Arabia, eastern and southern Africa: Eritrea to
Cape Province and Angola.

10. citrinipectus—southern Portuguese East Africa area.
11. mozambicus—widespread outside forests in Ethiopian regions.

RAND: AFRICAN CANARIES, GENUS SERINUS 129

12. donaldsoni—northeastern and eastern Africa (Somalia to Tan-
ganyika).

13. flaviventris—northeastern (Somalia), eastern and southern
Africa and Angola.

14. sulphuratus—eastern Africa (Kenya) to South Africa (where
chiefly eastern) and Angola.

15. albogularis—southern Africa, chiefly western.

In this group, the bill is of the stubby ‘Serinus’ type, becoming
very heavy in some species. Four species are predominantly yellow-
green, while the remaining four others are predominantly grey-brown.
The rump is differently colored, contrasting with the rest of the back,
white in one species, yellow in the others; the ‘‘typical Serinus’’ head
pattern (yellow or white forehead, and line over eye, a yellow or white
fleck in cheek, a dark band through or back of the eye, and dark malar
stripe) appears for the first time in this series and is usual, though is
not invariably present even within a species (e.g., atrogularis).

These eight species form a fairly satisfactory group with stepped
similarities, beginning with brown and grey, streaked species to vari-
ous intermediate forms, to a species which is green-yellow in both
sexes, and ending with a predominantly grey species.

S. leucopygius is a small grey-brown, streaked species with a white
rump, completely lacking yellow or green. In part at least this is a
geographical representative of the next species. S. atrogularis, also
a small, grey-brown streaked species, has a yellow rump; the geo-
graphical variation in this species is great, and although the black-
throated, eastern African form somereni is linked by intermediates to
the very different white-throated, streaked deserti of South West Af-
rica, the three northeastern forms, reichenowi, xanthopygius, and
rothschildi, could, on present knowledge, be considered allopatric spe-
cies. The close relationship of this species to the following yellow-
green forms is also indicated by occasional occurrence of yellow-
throated mutants in the northeastern forms of Africa, which White
considered a separate species, flavigula (Rand, 1968, Bull. Brit. Or-
nith. Club, 88, p. 116), and by the fact that citrinipectus of south-
ern Portuguese East Africa is a species that seems to be of hybrid
atrogularis-mozambicus origin (Irwin, 1961, Durban Mus. Novit., 6,
pt. 11, pp. 138-39).

S. mozambicus is small, greenish above, yellow below, with a
“Serinus head pattern” of black and yellow, and with little sexual

130 FIELDIANA: ZOOLOGY, VOLUME 51

dimorphism in color, though the female is duller. Even the young
are similar but duller and with only a little spotting on breast.

The next species, S. donaldsoni, though much larger, and with a
heavier bill, has a male that is somewhat of the yellow-green S. mo-
zambicus type of coloration, with the head pattern more subdued,
while the female is somewhat of the white-throated, streaked breast,
brown-grey and whitish S. atrogularis reichenowi type!

S. flaviventris has a moderate bill and the male with the type of
plumage that is similar to that of donaldsoni but more like that of
sulphuratus. However, in the northern part of the range, the male
has lower breast and abdomen white. The female has the upper
parts brownish to olive green, streaked, and the under parts yellow
and white, with breast more or less streaked. S. flaviventris as used
here includes dorsostriatus.

S. flaviventris of South Africa, with six races, and S. dorsostriatus
of East Africa, with two races, have usually been kept as separate spe-
cies, with a wide geographical gap between their ranges. However,
the differences between the two races that most closely approach each
other is less than those characterizing some geographically isolated
subspecies in such species as atrogularis, striolatus, burtont, and totta.

The characters of the relevant races are:

S. f. damarensis; southern Angola and South West Africa to ex-
treme northwestern Southern Rhodesia. (For revision of South Afri-
can forms see White, 1967, Bull. Brit. Ornith. Club, 87, p. 111 and
references therein.) Male, wing (8) 70-75 (av. 71.1); tail (3) 58-55
(av. 54); culmen 9-10 mm.

S. f. dorsostriatus; northwest Tanganyika, southwestern Kenya,
and central eastern Uganda: male like damarensis but upper parts
greener with black streaking heavier, yellow of forehead wider; ab-
domen with a small patch of white, often obscured by surrounding
yellow feathers; bill averages shorter. Male, wing (7) 72-77 (av.
73.4); tail (7) 47-52 (av. 50.7); culmen 8-10 mm. Female like that
of damarensis in having conspicuous white abdomen and yellow
breast more or less streaked, and differs chiefly in duller more olive
upper parts with broader blackish streakings.

S. f. maculicollis; Somaliland to eastern Kenya and west to south-
eastern Sudan and northeastern Uganda. Male like that of dorso-
striatus but yellow band on forehead much narrower, and whole
abdomen conspicuously white; wing (9) 67-74 (av. 69.7); tail (9)
47-51 (av. 48.6); culmen 8-10 mm. Female: much like that of dorso-

RAND: AFRICAN CANARIES, GENUS SERINUS 131

striatus but white of abdomen more extensive, extending onto lower
breast; streaking on breast more restricted sometimes forming a neck-
lace across the yellow lower throat-upper breast.

The most extreme of the yellow-green series is S. sulphuratus, an
intensely colored species with both male and female green and lightly
streaked above, green-yellow below, unstreaked, and with the “‘typ-
ical Serinus’’ head pattern present but rather obscure, and even the
immature are yellow-green, unstreaked below. The subspecies to be
recognized have been discussed separately (Rand, 1968, Fieldiana:
Zool., 51, no. 8).

The change to the predominantly grey and white S. albogularis
seems a sudden one, but in reality is not so great. S. albogularis looks
in size, bill, and color much like S. sulphuratus that has lost all the
yellow-green pigmentation except that on the rump and upper tail
coverts.

GrouP IV

16. gularis—north Ethiopian region, northeastern and eastern
Africa south to South Africa and Angola.

17. mennelli—Angola to Portuguese East Africa.
18. tristriatus—northeastern Africa: Eritrea to Ethiopia.
19. menachensis—southwestern Arabia.

In this group of brown-grey birds (without yellow or green), the
rump is like the back in color; the side of the head is uniformly brown
or blackish, except for a white eyebrow stripe in three species, and a
black malar stripe in one, tristriatus. The crown tends to be more
heavily streaked than the back due to the paler edges of the feathers
which in several species are white and so extensive in the forecrown
as to suggest an incipient white forehead. The throat tends to be
white, and streaking of the breast and flanks none, to obscure or
moderate. The bill is more or less stubby Serinus type but with a
tendency for slight elongation and slight slenderness.

This group seems a fairly close-knit one. Perhaps the closest
relative of the first species, gularis, is with albogularis to which it
has a vague similarity; at the other extreme, menachensis seems to
have no other relatives outside this group.

S. gularis has a white throat, brownish face, white eyebrow stripe,
much white in forecrown, and in some subspecies there is moderate
streaking on breast, flanks and back. Geographical variation is con-
siderable, and the question of the specific status of rezchardi, the

132 FIELDIANA: ZOOLOGY, VOLUME 51

browner, more streaked bird of parts of eastern Africa, has been
raised. Very close to gularis is mennelli, which co-exists with gularis
and differs only in detail of coloration (black vs. brown face, etc.).
S. tristriatus is a very plain bird with little streaking even on crown.
The small area of white on chin and upper throat is bordered by
black malar stripe, and the white eye stripe is narrow, but sharply
defined. The Arabian representative, menachensis, is still plainer,
but more streaked above and below.

GROUP V
20. striolatus—eastern Africa (Eritrea to Nyasaland).

21. burtoni—mountains of central Africa, Mt. Cameroon, Mt.
Kenya, etc.

22. rufobrunneus—islands in Gulf of Guinea.
23. leucopterus—mountains of southwestern Cape Province.

This is a group of medium size to large, predominantly brown and
grey canaries with moderate to heavy bills. Three species (in part at
least) of the four have some tinge of greenish-yellow. The rump is
colored like the back. The typical canary head pattern is present in
one species (in brown and whitish), and suggested in another. Al-
though not as closely knit a group as the three preceding ones, the
limitations of a lineal arrangement justify grouping them here.

The only species with a widespread range is striolatus, which is
heavily streaked above and on breast and flanks, and a pronounced
“canary head pattern.” It recalls a giant S. atrogularis reichenowi
of eastern Africa, but it has greenish edgings to the wing quills, and
one subspecies of restricted range, S. s. whytii, has yellow in throat
and head, indicating the relationships with the yellow-green series.
This last could be considered a separate species of restricted range in
the highlands of southern Tanganyika and northern Nyasaland.

The relationship of the above species to the next, burtonz, of re-
stricted, fragmented range, is indicated by the race S. b. melanochrous
of the highlands of southern Tanganyika, which on the upper parts
and head is much like burtonz but on the under parts has the streak-
ing of striolatus and could be kept as a separate species. Another
unusual, though minor character, of melanochrous is that the white
in the forehead is due to white feather edging, as in gularis, not more
solid areas as in burtont. S. burtoni is a large canary, heavy billed,
upper parts nearly uniform brown, with whitish tips to wing coverts
(some races); a variable amount of white markings in forehead and

RAND: AFRICAN CANARIES, GENUS SERINUS 133

sides of head, and black and white in chin, suggesting a canary head
pattern that is nearly lost, as the green of the plumage also has nearly
disappeared, being present only as edgings to wing quills. Except
for the one distinctly streaked subspecies mentioned above, the breast,
belly, and flanks are ochraceous brown, with a little obscure streaking.

The Gulf of Guinea islands endemic S. rufobrunneus is a very
plain reddish brown, above and below, bird with obscure streakings
or mottling, throat paler with some dark markings on chin, and
greater and medium wing coverts with pale rufous tips. Presumably,
this is the result of a later invasion by burtoni-stock, after an earlier
one gave rise to the genus Neospiza of Saint Thomas Island.

The South African species of limited distribution, S. leucopterus,
a large, heavy-billed canary, is a rather plain, grey-brown, obscurely-
mottled bird with white throat and abdomen and with the greenish
influence appearing as a greenish tinge over back and wings. It could
also be considered a relative of albogularis, or, more probably per-
haps, of gularis. However, I put it here, agreeing with Clancey
(1963, Durban Mus. Novit., 6, pt. 19, p. 262).

Group VI
24. totta—South Africa.
25. alario—South Africa.

These two small species are not closely related, but are grouped
as being the most extreme of the genus, without obvious specific re-
lationships elsewhere in the genus. S. totta has been put in the genus
Spinus because of the wedge-shaped bill. Even Roberts (1922, Ann.
Transvaal Museum, 8, p. 261) in coining so many new subgeneric
names did not change this one. The pattern of obscurely mottled
rufous-brown mantle, yellow-green striped head (without canary type
pattern), and plain yellow, or yellow and rufous under parts, and the
white line on the inner web of the outer tail feathers, is without par-
allel in the genus, though the contrasting yellow-green rump, of one
race but not the others, is Serinus in character. In one race the fe-
male differs from the male only in degree; in the other, it is mostly
rufous brown above, paler brown below. Although two rather dif-
ferent subspecies are recognized, this is an arbitrary decision, and
some would consider them as two species.

S. alario (including lewcolaema as a subspecies) with a stubby bill

has had a genus proposed for it, because of its coloration, but Roberts
(1922, Ann. Transvaal Museum, 8, p. 261) wrote that he thought it

134 FIELDIANA: ZOOLOGY, VOLUME 51

not far removed from S. atrogularis. Nicolai (1959, Zool. Jahr.
biicher., Syst., pp. 317-861) on the basis of behavior and hybridiza-
tion relates it to S. serinus and canicollis. However, the coloration
of the male with black hood, rufous mantle, and white lower breast
and abdomen, and the female being drab-grey-brown and pale chest-
nut, and the similar but duller young with streaked upper parts pro-
vide little in the way of clues to relationships.

SUMMARY |

This paper does not include the non-African species of the Cardu-
elis—Serinus—Spinus group from which the African forms presumably
evolved. The arrangement, new here, is based on similarities in pat-
terns and is presumably phyletic. One genus, instead of several
which have sometimes been used in the past, has been used for the
above 25 species. A comprehensive species concept has also been
used. Of the 25 species concerned, eight contain units which could be,
and often have been, considered as separate species. These are indi-
cated on the list on page 125. Only one of these is newly arranged
here, the merging of S. dorsostriatus in the species S. flaviventris. In
only one case is a traditional species treated as representing two
species, i.e., Serznus frontalis and S. citrinelloides. These represent a
group meriting further study.

The question of subspecies to be recognized have been extensively
reviewed in the voluminous literature of African birds, of which the
major items will be cited in the forthcoming volume of ‘‘Peters.’”’ The
subjective element in the number of subspecies to be recognized
makes a definitive treatment impossible.

At times it seems that the South African species have a dispro-
portionate number of subspecies. But it must be remembered that
the landscape there is very complex, with diversified vegetation and
climate. Numbers of local workers, with large series and familiar
with the birds in life, have reviewed the races, not always, it must be
admitted, with the same results. I have had to make decisions. In
only one case, S. sulphuratus, have I resurrected a name for a long
overlooked race.

ee ee
4

en
ee

oe

= oe ee ee

: ; 7
i i
Ae *
;
ae, :
i
’
4
.
a ;
oe
1
/
3
ee
ae t
ohi-.4
,
;
: fi
‘
Oo:
S i
'
Lae
F
ri ‘
‘ Sie,
Tike
is
=e
(re .
eal)
j 2”
tet 7
a 3
‘
i age
; nis
ot
Au at

oe ~ ae. ae!

=

aa a Play.