SHORT COMMUNICATIONS 457

BAGNARA, J. T., T. D. TAYLOR, AND M. E. HADLEY. LERNER, A. B. 1953. Metabolism of phenylalanine 1968. The dermal chromatophore unit. J. Cell and tyrosinase. Adv. Enzymol. 14:73-128. Biol. 38:67-79. LEWIN, V. 1963. Reproduction and development of ENGELS, W. L. 1959. The influence of different day young in a population of California Quail. Con- lengths on the testes of a transequatorial migrant, dor 65:249-278. the Bobolink (D&&onyx oryzioorus), p. 759- PFEIFFER, C. A., C. W. HOOKER, AND A. KIRKBAUM. 766. In Photoperiodism and related phenomena 1944. Deposition of pigment in the sparrows’ in plants and . Publ. 55, Amer. Assoc. bill in response to direct applications as a Adv. Sci., Washington, D.C. specific and quantitative test for androgen. FISHER, R. A., AND F. YATES. 1949. Statistical tables. Endocrinology 34 : 389-399. Hafner Publ. Co., Inc., New York. 112 p. POTTER, W. P., AND K. S. NORRIS. 1969. Lizard FLETCHER, R. A. 1971. Effects of vitamin A defi- reflectivity change and its effect on light trans- ciency on the pituitary-gonadal axis of California mission through body wall. Science 163:482- Quail, Lopho~rtyx californicus. J. Exp. Zool. 176: 484. 25-34. RAITT, R. J., AND R. D. OHMART. 1966. Annual cycle HADLEY, M. E. 1972. Functional significance of of reproduction and molt in Gambel Quail of vertebrate epidermal pigmentation. Amer. Zool. the Rio Grande Valley, southern New Mexico. 12: 63-76. Condor 68:541-561. HADLEY, M. E., AND W. C. QUEVEDO. 1966. Verte- RALPH, C. L. 1969. The control of color in . brate epidermal melanin unit. Nature 209:1334- Amer. Zool. 9:521-530. 1335. RALPH, C. L., D. L. GRINI~ICH, AND P. F. HALL. HALL, P. F. 1966. Tyrosinase activity in relation to 1967. Studies of the melanogenic response of plumage color in Weaver Birds. (Streganuru regenerating feathers in the Weaver : Com- padisues ) , Comp. Biochem. Physiol. 18:91- parison of two species in response to two gonado- 100. tropins. J. Exp. Zool. 166:283-288. HAMILTON, H. L. 1941. Influence of adrenal and sex hormones on the differentiation of melano- SERVENTY, D. L., AND A. J. MARSHALL. 1956. Factors phores in the chick. J. Exp. Zool. 88:275-305. influencing testis coloration in birds. Emu 56: 219-222. JONES, R. E. 1970. Effect of season and gonadotropin on testicular interstitial cells of California Quail. STEEL, R. G. D., AND J. H. TORRIE. 1960. Principles Auk 87:729-737. and procedures of statistics. McGraw-Hill Book JUHN, M., AND R. G. GUSTAVSON. 1930. The produc- Co., New York, 481 p. tion of female genital subsidiary characters and STURKIE, P. D. 1965. Avian physiology. Comstock plumage sex characteristics by injection of human Publ. Assoc., Cornell Univ. Press, Ithaca. 766 p. placental hormones in fowls. J. Exp. Zool. 56: WILLIA~~S, C. R. 1967. The breeding biology of 31-66. California Quail in New Zealand. New Zealand KRUMREY, W. A., AXD I. 0. Buss. 1969. Observations Ecol. Sot. Proc. 14:88-99. on the adrenal gland of the African elephant. J. Mammal. 50:90-101. Accepted for publication 30 January 1974.

SYSTEMATIC3 OF THE in intensity, in general, to the pigmentation of these WHITE-THROATED areas in adults), whereas all juveniles I saw of P. (PZPZLO ALBZCOLLZS) fuscus and Aberts’ Towhee, P. uberti, (which, with P. uZbico,llis, form the “brown towhee” group) com- KENNETH C. PARKES pletely lacked dorsal streaking. In M. kieneri the Carnegie Museum dorsal markings appeared as bars (actually transverse Pittsburgh, Pennsylvania 15213 expansions at the tips of somewhat suppressed streaks). The one juvenile of P. ulbicollis then avail- The White-throated Towhee, Pipilo ulbicollis, is a able (Moore Collection 32696) showed a “faint Mexican endemic species of limited distribution. barring” on the dorsum. Two additional juveniles Most of the published information about this species now before me (Carnegie Museum 141364; A. R. can be found in two papers, those of Davis ( 1951) Phillips 5195), as well as several in first prebasic and Marshall ( 1964). Davis used the specific name molt that retain some dorsal feathers of the juvenal Pipilo v.dus, but see Stresemann ( 1954). plumage, confirm this. The terminal barbs of the Marshalls’ careful studies of vocalizations indicate juvenal dorsal feathers have a few blackish barbules that PipiZo albicollis is less distinct from the Brown at the very tip, giving the effect of a faintly darker, Towhee, P. fuscus, than I had believed when I wrote narrow transverse barring or scalloping. This is pre- of this genus and its relatives in 1957 ( Parkes 1957). sumably what Marshall (1964:354) meant in stating I called attention to certain striking similarities be- of P. albicollis that its “marks are broader than long.” tween P. ulbicollis and Melozone kieneri and proposed As stated in my 1957 paper and confirmed by addi- that, judging from skins, P. ulbicollis formed a link tional specimens, the ventral markings of juvenile between P. fuscus and M. kieneri. Marshall found ulbicollis resemble those of the juvenile Melozone that some of the Mexican races of P. fuscus exhibited kieneri I examined in being irregularly distributed and some of the characters that I had thought were con- vaguely shaped dense spots, quite unlike those of any fined, in this group, to P. ulbicollis and M. kieneri, but other Pipilo. Marshall deprecated my statement that the latter two species still share characters not found P. fuscus and P. uberti differed from P. erythrophthal- in the other forms. One of the characters I invoked mus and P. chlorurus in completely lacking dorsal was the pattern of the juvenal plumage. I found that streaks. He found that “many juvenal fuscus [he did in true Pipilo (i.e., the erythrophthulmus group plus not mention abetii] are liberally streaked above and chlorurus) the juvenal plumage is distinctly streaked below; and that juvenal kieneri are usually streaked, above and below (the ventral streaking corresponding not spotted below.” My 1957 findings were based 458 SHORT COMMUNICATIONS in large part on borrowed material; at this point I can Davis emphasized the need for additional collecting say o& that I do not doubt Marshall-word,’ but to provide seasonally comparable samples. Paynter that the 53 iuveniles of P. eruthrovhthalmus and 8 ( 197O:lSO) listed parvirostris in the “Peters” Check- of P. chlorurk in the Carnegie”Muskum are all char- list with a query, citing Davis ( 1954) on the un- acterized by distinct longitudinal dorsal streaking. certainty of its validity. There is no trace of such streaking in the Carnegie On hand at Carnegie Museum are specimens of sample (part of a substantially larger series examined Pipilo albicollis recently collected by Allan R. Phillips, in 1957) of six P. fuscus and four I.’ aberti juveniles. Robert W. Dickerman, Otto Epping, and the writer. Marshall speaks several times of the “Canyon For comparison, 28 specimens, including 23 of the Towhee” (Pipilo fuscus part) as the “ancestor” or type series of paruirostris, were borrowed from the “ancestral stock” of the White-throated Towhee. I Moore Laboratory of Zoology. A list of specimens have long objected to the assignment of a living examined will be found at the end of this paper. species as the “ancestor” of another living species; In studying variation in this species, Davis ( 1951) but this objection aside, Marshalls’ interpretation did not wholly segregate his specimens by age class. leaves no room whatsoever for a possible relationship In discussing the molts and nlumaees of the brown between Pipilo albicollis and Melozone kieneri. He towhee group in general, he- wrote (p. 3) : “Birds suggests that Melozone may be related to Atlupetes undergoing the postjuvenal (first fall) [= first pre- and the chlorurus-eruthrovhthalmus eroun of Pivilo basic] molt retain the primaries and secondaries, and (rather than to P. afbicolks), and, f&her, that ;the‘ usually the rectrices.” A few lines later, however, he brown may be connected with some members wrote “. rectrices are often replaced in the post- of the heterogeneous Aimophila.” I do not find these iuvenal molt. . . .” Insofar as P. albicollis is con- suggestions persuasive, based as they are on sub- cerned, the statement about the retention of juvenal jective impressions of a few vocalizations plus, in the remiges is too dogmatic. At least some individuals first case, “a black shiny bill plus addiction to densest do replace remiges. A. R. Phillips 7891, “windows understory brush” and, in the second, “brown colora- cover sk[ull] roof,” 20 November, has several short, tion and occurrence at edges of brush.” Marshalls’ sheathed remiges. Moore 31112, 17 September, is paper, nevertheless, is a major contribution to our molting from the narrow rectrices typical of juveniles knowledge of relationships among the brown towhee to the broad rectrices of the first basic plumage (see segment of Pipilo (if, indeed, they are truly members Davis 1951:3, fig. 1). It is also in heavy wing molt. of that genus. as Marshall and I and several other The old remiges are scarcely worn at all and are authors have wondered). obviously of the juvenal generation, whereas the old Mayr and Short (1970:82) state that “albicollis remiges of adults undergoing wing molt are generally and fuscus meet barely, if at all [italics mine], in worn and frayed. Virtually all of the parvirostris southern Mexico.” They cite Davis ( 1951) to sup- specimens from Moctum, Oaxaca, taken in .September port this statement, although Daviss’ map clearly and October (including Moore 31112) are moltina shows svmuatrv of the two snecies in almost half of remiges; the chances of all of these having been , & the range of albicollis and his text cites three localities adults, at that time of year, are slim, but unfortunately of sympatry. Marshall (1964:353) has shown that no notes on skull pneumatization appear on their the supposed sympatry at Mitla, Oaxaca, was based labels. on mislabeled specimens of P. fuscus, but there is no In table 47 Davis (1951:83) shows that the entire question about the other two localities of sympatry. type series of 31 specimens of parvirostris was in- Dr. Allan R. Phillips (pers. comm.) has informed cluded in the sample for bill length, although some me of an additional area of sympatry, near Nochixtlan, of these are clearly immature, and possibly even a Oaxaca. His specimen of P. fuscus from that locality majority, as might be expected in a fall sample taken is apparently the southernmost known. at random. The bill lengths of nine males of albicollis Davis ( 1951: 84 ) described Pipilo rutilus parvirostris taken in October and November, and thus seasonally as a new subspecies confined to the vicinity of Mount comparable to Daviss’ series of parvirostris, ranged Zempoaltepec, Oaxaca. He based it on two characters: from 9.9 to 16.7 mm, with a mean of 16.32 mm. a “decidedly shorter” bill, and “coloration of pileum This is very close to Daviss’ figure for 22 males of browner, less gray.” However, he pointed out that parvirostris: 9.8-11.0 (mean 10.37). Of the nine the supposed color difference might be attributable albicollis, the five adults had bills measuring 10.4- to the fact that most of the parvirostris series were 10.7 (mean 10.62), whereas the four immatures collected in the fall of 1941, and that the best speci- measured 9.9-10.0 (mean 9.95). If we assume that mens of “rmtilus” available were from February 1948. the age ratio among the specimens of parvirostris was He suggested that “the richer, browner coloration of about the same, then the inevitable conclusion is that parviroatris may be the result of foxing and unworn Davis ( 1954) was correct in his suspicion that sea- plumage.” Recently taken specimens demonstrate sonally comparable samples would not show significant that dorsal browns do become grayer with wear in differences in bill length. Pipilo albicollis. In addition, Daviss’ series of “rutilus” Making allowances for the differences in collecting was composite, and included some specimens of a years, I find that the fresh-plumaged 1941 specimens truly grayer-backed subspecies, to be described be- in the type series of parvirostris are not significantly yond. -In. a later paper, Davis ( 1954:148) cast some different in color from fresh-plumaged albicollis taken doubt about the validitv of the bill lenath character: in central and southern Oaxaca in 1963-65. This fact, in this paper he demonstrated seasonal bill length together with the demonstrated lack of significant changes in several owing to changes in differences in bill length of seasonally comparable feeding habits. Three summer specimens of par- specimens, indicates that Daviss’ skepticism about the virostris had slightlv longer bills. sex for sex. than validity of his “parvirostris” was justified, and that the average of the fall series of I2 males and ’ 7 fe- name becomes a synonym of albicollis. males. The summer series of “rutilus” had mean bill The distribution map published by Davis (1951: measurements longer than a single male and female 79) clearly suggests a disjunct range for the species taken in December, but 9 February females averaged Pipilo albicollis, although this may be in part an very slightly longer-billed than 12 summer females. artifact of collecting. Nevertheless, the specimens I SHORT COMMUNICATIONS 459 have seen from localities additional to those mapped Specimens examined. [Additional locality records by Davis have been from near one or the other of are given by Davis 1951.1 the two clusters of localities he shows: one in southern Puebla and northern Oaxaca, and one in Pipilo a. ulbicollis central to southern Oaxaca. The series of specimens Oaxaca: San Jose de1 Pacifico, 6 (November) assembled for the present study shows that two races 13 km S of Miahuatlan, 2 (November) of Pi&o albicollis are in fact recognizable, corre- El Tule (S. of Oaxaca), 3 (January) sponding to the northern and southern groups of 6 km S of Etla, 4 (January) localities mapped by Davis. The type locality of Mitla, 6 (April, 1; June, 5) Pipilo albicollis Sclater is San Miguel de las Peras, P. a. “puruirostris” ( = albicollis) Oaxaca, which is well within the range of the Oaxaca: Moctum, 20 (September through Decem- southern form, leaving the northern one to be named. ber ) It may be called: Totontepec, 3 (April)

Pipilo albicollis marshalli P. a. marshalli Oaxaca: San Antonio de1 Rio (just S of Puebla new subspecies border, N of Camotlan), 4 (May, 1; Holotype. Carnegie Museum No. 142292 [pre- July, 3) sumably adult] male (testes 2 x 1 mm, cranium fully 4 miles WNW Tamazulapan de1 Progreso, pneumatized, rectrices of adult shape). From San 1 ( May, juvenile ) Vicente Villalegria, 4 miles N of Tehuacin, Puebla, Puebla: San Vicente Villalegria, 4 miles N of Mexico. Collected. by Juan Nava S. for Kenneth C. Tehuacin, 7 (October, 3; February, 4) Parkes (field no. KCP 2445) on 9 Februarv 1965. Diagnosis. Differs from P. a. albicollis in being I am grateful to John William Hardy and Martin colder and grayer dorsally, on face, and on sides of L. Morton for sending me the important specimens neck; flanks duller, less richly rufescent and more of Pipilo albicollis from the Moore Laboratory of olivaceous anteriorly; spots on tips of wing coverts Zoology, Occidental College; to Allan R. Phillips and and upper tail coverts, and edgings of fresh outer Robert W. Dickerman for companionship and guidance secondaries paler (whitish to buff rather than some in the field in Mexico, and for lending me specimens shade of orange-buff); orange-brown area of throat from their personal collections (which are now de- paler and, in series, averaging less extensive. posited in the Delaware Museum of Natural History Range. Southern Puebla and adjacent northern and Cornell University, respectively); to Juan Nava Oaxaca, Mexico. I have not examined the specimens S. and Santos Farfan B. for assistance in the field; from Guerrero mentioned by Davis ( 1951:81), but to the Direction General de la Fauna Silvestre for his description of the dark flank color of the two permits to collect birds in the Republic of Mexico; unworn specimens as well as the locality strongly and to the Edward ONeil’ Fund of Carnegie Museum suggest that these birds are referable to marshalli. for financial support of field work. Etymology. Named for Joe T. Marshall, Jr., who has contributed much to our knowledge of the brown LITERATURE CITED towhees in life, although this represents but a tiny DAVIS, J. 1951. Distribution and variation of the fraction of his total contribution to ornithology. Brown Towhees. Univ. California Publ. Zool. Remarks. Color comparisons above are based on 52: l-120. fall and winter specimens taken in 1963-65, and DAVIS, J. 1954. Seasonal changes in bill length of thus comparable with respect both as to seasonal certain birds. Condor 56: 142-149. wear and museum age. Even in worn birds, however, MARSHALL,J, T., JR. 1964. Voice in communication several of the differences are noticeable. Two ju- and relationships among Brown Towhees. Condor veniles of marshalli are available. Comparing these 66:345-356. MAYH, E., AND L. L. SHORT. 1970. Species taxa of with three specimens of ulbidlis that retain portions North American birds. Publ. Nuttall Ornithol. of the juvenal plumage (A. R. Phillips 7889, 7890, Club 9: 1-127. 7944), the two marshalli have the throat only faintly PARKES,K. C. 1957. The juvenal plumages of the washed with yellow, whereas in juvenile albicollis finch genera Atlupetes and Pipilo. Auk 74:499- the throat is distinctly yellow, this color even per- 502. meating the darker adjacent area of the face to pro- PAYNTER,R. A., JR. 1970. Subfamily Emberizinae. duce a somewhat greenish appearance. As in later In R. A. Paynter, Jr. [ed.] Check-list of birds of plumages, the pale edgings of the inner secondaries the world, p. 3-214. Vol. 13. Museum Compara- are more distinctly rufous in albicollis. The first tive Zoology. STRESEMANN,E. 1954. Ferdinand Deppes’ travels prebasic molt of the three ulbicollis has progressed in Mexico, 182441829. Condor 56:86-92. too far to permit adequate comparisons of flank and dorsum color. Accepted for publication 29 August 1973.