LIBRARY OF THE UNIVERSITY OF ILLINOIS AT URBANA-CHAMPAIGN 590.5 FI V.5I cop. 3 SURVEY P r JAN 3 i3o9 v.. 51 FIELDIANA . ZOOLOGY ^j.^, ,s , '^ Published by / FIELD MUSEUM OF NATURAL HISTORY Volume 51 November 13, 1968 No. 8 Geographical Variation in the Canary Serinus sulphuratus Austin L. Rand Chief Curator, Zoology, Field Museum of Natural History This canary ranges widely east and south of the African equatorial forest with a favorite habitat of grasslands with many scattered trees and shrubs. It shuns the forest and also thick brush. Although widespread, it is often local in occurrence, and is absent frorp some of the more arid areas, such as much of western South Africa and eastern Kenya. In terms of political units the species range is from western Kenya, Uganda, and adjacent Congo south to Cape Prov- ince, and west across the continent through Northern Rhodesia (and parts of adjacent Congo) and Angola. The main geographical variation is of the "more or less" type: in general size (wing length) and size of bill (not only in length, but even more so in depth and consequent bulk), and in coloration (yellowish- ness or greenishness of plumage). The extremes, the large, very heavy-billed, green bird from Cape Province, and the smaller, mod- erate-billed, and yellower bird from' Uganda, etc., appear sharply different. But there are intermediates of various sorts. Size and color do not always march together, and neither follow uniformly smooth clines, although some could be considered stepped reversing clines. To sketch this briefly: The larger-billed, green-yellow Cape Province bird is replaced by a moderately yellow, moderately large-billed bird in northeastern South Africa; a small-billed, yellow bird occurs from Southern Rho- desia to Uganda, a small-billed, moderately green bird in Angola, and a moderately large-billed, moderately green-yellow bird in Kenya. The difficulty of codifying this variation in terms of species and subspecies is shown by various taxonomic treatment. Roberts (1948 ■ Library of Congress Catalog Card Number: 68-5902U No. 1059 119 120 FIELDIANA: ZOOLOGY, VOLUME 51 (6th impression), Birds of South Africa, pp. 367-368) recognized two species: S. sulphurata with three subspecies, and S. shelleyi with none. The revised 1957 edition recognizes only one species with three sub- species, as does White (1963, Revised Check-list Afr. — finches — , p. 771) (but not all the same three that Sclater, 1930, Syst. Av. Aethiop., p. 816, recognized), while Clancey (1966, Durban Mus. Novit., 7, pt. 13, p. 608) recognizes five races. Thanks to the series in Field Museum of Natural History, and the loan of material from the Museum of Comparative Zoology (courtesy of Dr. R. A. Paynter), the American Museum of Natural History (courtesy of Dr. Dean Amadon), and the United States National Museum (courtesy of Dr. G. E. Watson), it is possible to re-evaluate the situation with results differing somewhat from all the above. Perhaps the most notable of the changes is the recognition of the remarkable similarity between the populations from Kenya and those of the far distant southern Portuguese East Africa-Natal area. Angola birds are similar to these in color but not in bill size. The problem can be considered from one point of view as lumping or splitting and the identification of specimens. But, from another point of view, the subspecies concept is more meaningful when con- sidered as a way of codifying geographical variation with names so we can more easily talk about it. How many names are useful for this? Two treatments are possible: one, using two, or the other, using five subspecies names. I. From a broad, over-all point of view, only two names could be used: 1. S. s. sulphuratus Linnaeus, 1776 Range. — Cape Province. Diagnosis.- — The largest, most heavy-billed, and green- est birds, with a dark olive breast band contrasting with the vivid yellow throat. Remarks. — This is the only population that is clearly separate from all the others. 2. S. s. sharpei Neumann, 1900 Range. — That of the species north of Cape Province. Diagnosis. — This includes variable populations, but in color and size all are separate from the above in being less greenish and more yellowish; most populations are also smaller in size. RAND: GEOGRAPHICAL VARIATION IN SERINUS 121 II. If this treatment be not accepted, it becomes necessary to use at least the five subspecies outlined below: 1. S. s. sulphuratus Linnaeus, 1776 — Cape of Good Hope Range.— Cape Province. Diagnosis. — The darkest, greenest subspecies, with dis- tinct dark olive breast band, also larger and more heavy- billed. 2. S. s. wilsoni Roberts, 1936^ — Kloof, Natal Synonym: S. s. languens Clancey, 1962, Durban Mus. Novit., 6, pt. 15, p. 193 — Sul do Save, southern Portuguese East Africa. Range. — From Natal and Orange Free State to southern Portuguese East Africa. Diagnosis. — Like sulphuratus, but yellower on breast, flanks, and upper parts. Remarks.- — S. s. languens was described as like wilsoni in color but smaller. Male wing 74-78 (av. 76) vs. 80.5- 85.5 (av. 83.1), but see measurements in table. At most, this can be considered a slight tendency toward the smaller, brighter yellow S. s. shelleyi to the north. 3. S. s. shelleyi Neumann, 1903 — Kafuro, northeastern Tan- ganyika. Synonym: S. s. loveridgei van Someren, 1921. — Lumbo, Portuguese East Africa. Range. — Southern Rhodesia, extreme southern Nyasa- land, and northern Portuguese East Africa, presumably across central Tanganyika to Uganda and adjacent Congo. Diagnosis. — Differs from both the above in being yel- lower, the olive breast band being faint or obsolete, in the shorter wing, and in the much smaller bill, more evident in its lesser bulk than its somewhat shorter length. Remarks. — A female from Lumbo, northern Portuguese East Africa (the type locality of loveridgei) is small, wing 72, culmen 10, and very yellow, agreeing with females of shelleyi from Uganda in this). Vincent (1936, Ibis, p. 116) says that shelleyi from the littoral of Portuguese East Africa are smaller than those from inland. Four specimens from eastern southern Nyasaland area, and two from Southern Rhodesia are slightly greener than Uganda shelleyi but are best placed here. 122 FIELDIANA: ZOOLOGY, VOLUME 51 4. S. s. frommi Kothe, 1911 — Namanjera southwestern Tan- ganyika. Range. — Southwestern Tanganyika, Northern Rhodesia (and probably adjacent Congo and northern Nyasaland) and Angola. Diagnosis. — A small subspecies like shelleyi with a small bill but greener, with a more pronounced olive breast band, thus approaching in color both wilsoni and sharpei. Remarks. — This name has usually been placed in the synonymy of shelleyi, although the characters of this race, greenishness and small bill, have been pointed out by Traylor (1962, Publ. Cult. Co. Diam. Ang., Lisboa, no. 58, p. 130). I have not seen the type of frommi which is from Na- manjera, in Ufipa, latitude 7° 34' S., longitude 31° 07' E., but the description says "... like sharpei but smaller — and in regard to color^under parts sulphur yellow with breast and flanks greenish yellow. ..." I have a single specimen from Tukuyu, latitude 9° 16' S., longitude 33° 38' E., that agrees well with this, and with two Northern Rhodesia birds (Ndola and Lundazi) and most Angola birds. From northern Nyasaland (Vipya Plateau, 6,000 ft.) I have but a single female (= cT?) wing 76, culmen 12 mm. It is much greener below than Uganda shelleyi and agrees with male frommi. Certain specimens in Field Museum from Mt. Moco and Mt. Soque in Angola are much darker and greener than a series in the American Museum from various parts of An- gola and one from Mt. Moco. The meaning of this dis- crepancy is not clear. 5. S. s. sharpei Neumann, 1903^ — Marangu, Kilimanjaro Range. — Western Kenya highlands and Kilimanjaro. Diagnosis. — Differs from its geographical closest rela- tive shelleyi in being larger, larger-billed, and more greenish with a light olive breast band; very similar to the far dis- tant wilsoni. Remarks. — The main character of this subspecies, on the northwestern edge of the range of the species and sep- arated from wilsoni by shelleyi, is one of geography. As RAND: GEOGRAPHICAL VARIATION IN SERINUS 123 TABLE OF MEASUREMENTS I found it difficult to measure the bill satisfactorily. The measurements used here are the result of remeasuring the whole series in one day. Larger bills tend to be very much more massive than shorter ones, so that the observable differences are greater than measurements indicate. Wing Culmen (mm.) (mm.) S. s. sulphurata Cape Province sex ? (9) 81-87 (av. 83.6) (9) 12-14 (av. 13) S. s. wilsoni Southern Portuguese East Africa cf (4) 77-79 (av. 78.1) (4) 12-13 (av. 12.7) Natal sex ? (6) 81-87 (av. 83.6) (6) 12-13.5 (av. 12.7) S. s. shelleyi Uganda cf (10) 76-78 (av. 76.8) (10) 11-12 (av. 11.4) Lake Kivu to Lake Tanganyika area c? (7) 76-80 (av. 77.2) (7) 11-12 (av. 11.5) Extreme southern Nyasaland cf ? (4) 72-73 (av. 72.3) (4) 10.5-11 (av. 10.7) Southern Rhodesia c? (2) 76,76 (2) 12, 12 S. s. frommi Angola cf (7) 77-80 (av. 78.5) (6) 10.5-12 (av. 11.1) Northern Rhodesia