RELATIONSHIPS OF THE AVIFAUNA OF SAN ESTiBAN ISLAND,

RICHARD C. BANKS

Bureau of Sport Fisheries and Wildlife U.S. National Museum Washington, D.C. 20560

San Esteban Island, Sonora, , lies in of the avifaunal areas of Sonora, where he the approximately midway states (van Rossem 1945b:24): “The San between the mainland of Mexico and #the Lucan, a term coined by Dice for the southern peninsula of , at a latitude of three-fourths of Baja California, is used with 28” 40 ’ N. It forms the center link of a chain extreme reservation. If employed here it does of islands across the Gulf, being slightly nearer not imply acceptance of the proposal to rec- Tiburon Island on the Sonoran shore than to ognize Baja California as a separate Biotic‘ South San Lorenzo Island on the Baja Cali- Province ’ but to emphasize the character of fornia side. Tiburon is a large island, very the avifauna of San Esteban Island. This is- near shore, with a large avifauna and several land which lies almost exactly in mid-Gulf is endemic forms (van1 Rossem 1932), whereas a political accident so far as the present re- South San Lorenzo is small with a meager avi- port is concerned for only one bird, a thrasher, fauna. The birds of Tiburon are all closely belongs to the Sonoran mainland avifauna. So related to those of Sonora, and the species on far as is known at this time all the other in- San Lorenzo are all identical with those of dicative resident races (except for an Am- Baja California. Although the avifauna of San phispiza) are those of Baja California or more Esteban Island is small, it includes seven spe- particularly of the San Ignacio District (of cies which vary geographically and of which, Grinnell, 1928).” The San Ignacio District as therefore, one can attempt to determine rela- outlined by Grinnell (1928:8) covered about tionships. the central two-thirds of the peninsula and was The first impression that one obtains, from characterized as “poorly developed as regards looking at maps or from visiting the area, is endemic differentiates.” of the proximity of San Estdban to Tiburbn, The subspecific assignments of the birds of which suggests that the affinities of the biota San Esteban made by van Rossem (194513) of the former should be with Sonoran forms. were accepted by the committees that pre- A. J. van Rossem, who is responsible for vir- pared the most recent A.O.U. and Mexican tually all the published information about the check-lists. My own work in Baja California birds of San Esteban Island, originally held and on the islands in the Gulf of California that viewpoint and assigned specimens from (Banks 1963; see also Lindsay 1962,1986) has that island to subspecies found on Tiburon and led me to question some of the reported sub- the Sonoran coast. For example, van Rossem specific determinations and thus to reopen the (1930b) stated that his “Tiburon Island Dis- question of the relationships and affinities of trict” of Sonora “includes also San Esteban the San Esteban avifauna. In the course of Island.” the reassessment reported here I have exam- In later writings, however, van Rossem re- ined all available material from that island, investigated the subspecific status of the geo- including most of that seen by van Rossem graphically variable San Esteban birds and and several recently taken specimens. In each came to quite different conclusions. In 1942 of the seven species accounts which follows, I he assigned specimens of three species to races present briefly the history of the nomenclature found in Baja California rather than to Sonoran of the San Esteban population prior to setting forms, and he emphasized his changed view- forth my own concept of its relationship. point on a number of subsequent occasions, stating, for example, “San Esteban Island, al- DENDROCOPOS SCALARIS though politically a part of Sonora, is Baja When he first reported the Ladder-backed Californian in its avifaunal relationships . . .” Woodpecker on San Es&ban Island, van Ros- (van Rossem 1945a:243). The clearest expres- sem ( 1931) listed his single specimen under sion of this concept is set forth in a discussion the name D. s. cactophilus. Later (van Rossem The Condor, 71:88-93, 1969 1881 AVIFAUNA OF SAN ESTBBAN ISLAND, SONORA 89

1942) he assigned that specimen and one ad- TABLE 1. Mean measurements (mm) of Myiurchus ditional one to the race D. s. lucasanus of cinerascens. southern Baja California, noting that the race All Mexican on the peninsula opposite the island was not cinerascenscinerascens San Est6ban pmtinax lucasanus but eremicus. Birds of this species Wing length, 8 99.9 98.9 98.5 96.4 from nearby Tiburon Island were referred to Winn length. Q 94.2 93.3 92.2 90.9 cactophilus while those from the mainland Tailieng&, 2 91.4 90.9 91.6 89.8 Taillength, 0 86.0 85.6 85.2 84.2 opposite San Esteban (Kino Bay) were placed Bill length, $ 14.57 14.49 14.55 14.83 first in cactophilus but later in sinaloen& Bill length, 0 14.05 14.10 14.30 14.37 (van Rossem 1932, 1945). Recent check-lists Data, except for San EstBban, from Lanyon (1961 and pen. ( A.O.U. 1957; Miller et al. 1957) followed van comm.). Wing length is of flattened wing. Rossems’ treatment of the insular populations but placed the northern limits of lucasanus at inquietus is now considered a part, from the about 29” on the peninsula, following Grinnell species cinerascens ad,ds uncertainty to any ( 1928), to the north of San Esteban. More interpretation of his early concept of the re- recently, Banks (1963) reported a specimen lationships of the San Esteban population. of D. s. sinabensis from Tiburon Island. Thus, Later, when commenting on the Baja Cali- by current treatment the San Esteban popula- fornia affinities of the San Esteban avifauna, tion is flanked by lucusanus on the peninsula van Rossem (1942) referred the same speci- and sinaloensis on the mainland, with the mens to M. c. pertinax of the southern part of southern limits of both cactophilus and eremi- the peninsula. He maintained that position cus not far away. (van Rossem 1945b) when he recognized the There are but three specimens from San distinction of nuttingi and cinerascens. This Esteban available for comparison, including final assignment of the San Esteban birds to the two mentioned’ by van Rossem and one pertinax was followed by the A.O.U. (1957) taken in 1966. All are males; two were taken and by Miller et al. (1957). Lanyon (1961) in April and one in January. The two April did not specifically mention the San Esteban birds, both of which were breeding, are quite birds; although his range map shows a symbol worn. Only the most recent specimen has a for pert&x on San Esteban Island, at the lati- complete upper mandible. tude of intergradation of pertinax and cineras- None of these specimens is like the small tens, he discussed pertinax (p. 443) only with sinaloensis, but I cannot with certainty assign reference to the peninsula of Baja California them to any subspecies. Each individual seems south of latitude 29”. to be variously intermediate, resembling one I have been able to examine all material race in some characters and another race in previously reported from the island (Town- other ways. The relatively fresh plumaged send 1923; van Rossem 194513)as well as speci- January bird is most like catophilus dorsally, mens taken in 1966, a total of eight males and but is more similar to lucasanus in the reduc- five females. All specimens were taken 13-27 tion of markings on the posterior underparts. April. Although migrants of cinerascens are Both April birds are similar tot lucasanus dor- not unexpected at this latitude at this time, I sally, but one is close to eremicus and one to consider the material examined to be members cactophiks ventrally. Thus there seems to be of the resident population. Males taken on influence from both the mainland and the 27 April 1966 had testes 13 x 5, 10 X 5, and peninsula. Certainly this species cannot be 11 x 8 mm; others taken on 17 and 18 April said to demonstrate particularly close affinities 1930 are labeled as being in “breeding condi- between the San Esteban and southern Baja tion.” California avifaunas. M. c. pertinux is a weakly defined race that averages smaller in wing and tail length but MYIARCHUS CINERASCENS slightly larger in bill length than M. c. ciner- The earliest specimens of Ash-throated Fly- ascens. Lanyon ( 1961:428) detected no sig- catchers from San Estkban Island (Townsend nificant divergence in plumage coloration be- 1923) were reported under the name M. c. tween the two races. Ridgway (1907:628) cinerascens. Van Rossem originally (1931) mentioned that the yellow of the underparts referred his material from there to the race of pertinax averaged slightly deeper, but noted cinerascens, although he considered them to .that this character was not entirely constant. be intergrades with the mainland form, in- The difference seems more useful in fresh quietus. The fact that van Rossem did not plumaged fall birds than in samples of the separate the species nuttingi, of which the race breeding populations, and I could not use 90 RICHARD C. BANKS this or any color character in trying to place Soon after the above statements appeared, the birds from San Es&ban Island. Huey ( 1930) described A. f. ignutius from the Measurements of my sample from San central part of the peninsula of Baja Cali- Es&ban are presented in table 1, with the fornia, restricting lumprocephalus to the Cape data given by Lanyon ( 1961) for the two races region proper, as van Rossem (1930a) had for comparison. Both van Rossem ( 1945b) and suggested should be the case. Before long, Lanyon (op. cit. ) noted that Mexican popula- however, Grinnell ( 1931) reported new infor- tions of M. c. cinerascens average somewhat mation about the type specimen of the species smaller than more northerly samples. Dr. which led’ him to apply the subspecific name Lanyon kindly provided’ me with the pertinent flaviceps to the birds of Baja California (in- data for a Mexican series of 46 male and 20 cluding both lumprocephalus and ignutius) female birds, which are included in table 1. and to rename the population of southern Cali- Except for the tail length of males, the fornia and Arizona as A. f. acaciarum. means of the small sample from San Esteban Again in 1931, van Rossem referred the San Island fall between the means of pertinax and Esteban Island Verdins to the Sonoran race cinerascens. With the single exception of bill fraterculus, and only slightly later (1932:X38) length of the few females available, means for he repeated this position, stating, “In the light the San EstCban Island population are closer of more freshly plumaged material taken . . . to those of cinerascens than of pertinar, and on both Tiburon and San Esteban Islands I in general are most like the Mexican cineras- am still of the opinion that this name is the tens. Thus, although the island population is best one to employ. Birds from the two islands somewhat intermediate, it should most prop- are slightly but definitely grayer even than erly be referred to M. c. cineruscens. This Kino Bay intergrades between fracterculus and species gives no indication that the avifauna Auriparus flaviceps acaciurum Grinnell. , , ,” of San Estkban is closely related to that of It is evident from the last sentence quoted that southern Baja California. van Rossem considered the San Esteban birds The southern race pertinux is resident to be most like those of Tiburon Island, and through most of the peninsula, north to lati- closely related to the birds of central and tude 28” with a zone of intergradation north to southern Sonora. at least 29” ( Grinnell 1928; Lanyon 1961). San Later, however, van Rossem (1942) referred Esteban Island is well within the zone where his San Esteban specimens to A. f. flaviceps some tendency toward pertinux would be ex- for the first time, stating about the insular pected in populations which might be con- population “the Verdin is seemingly identical sidered basically cinerascens. with that of the nearest part of the penin- sula and the intermediate San Lorenzo Island. AURIPARUS FLAVXEPS It may be here noted that . . . there is a readily When van Rossem (1930a) described the race distinguishable race of the Verdin in central A. f. fraterculus of central and southern So- Lower California and that lumprocephulus is nora, he inchrded in it the Verdins of Tiburon confined to the Cape region proper.” From and San Esteban Islands, even though they this it is clear that he used the name flaviceps averaged slightly paler than mainland birds. to refer to the birds of central Baja California At that time van Rossem remarked on the re- (Hueys’ ignatius) and not in reference to the semblance of the adults of fraterculus to the birds of the Cape region. In a later study, van birds of the race in southern Baja California, Rossem (1945b) mentioned the San Esteban for which he used the then current name Verdins under the subspecific name fluviceps lamprocephulus, but he considered it to be once again, but it is not clear to which portion possibly a matter of convergence of characters of the Baja California population he con- as a result of the occupation of areas of similar sidered’ the name to apply. climatic conditions. He noted that the young Under the present generally accepted ar- of fraterculus were “zjery different” from those rangement (Miller et al. 1957; A.O.U. 1957) of of the subspecies of the Cape region of Baja races of the Verdin, the name flaviceps applies California, and stated that “the evidence of to birds of Baja California south of about 30”, the juveniles places fraterculus in the ornutus- and the name fraterculus is attached to birds fluviceps series . . . of relatively distant rela- in Sonora south of 30”. The San Esteban pop- tionship to lumprocephalus.” It should be re- ulation, fianked by these two races, is called called that the name fluviceps as used at that fluviceps in both these check-lists but the time referred to the birds of southern Cali- Tiburon birds are included in fraterculus. Re- fornia and Arizona. call that van Rossem ( 193Oa, 1931) considered AVIFAUNA OF SAN ESTBBAN ISLAND, SONORA 91 these two insular populations to be identical POLlOPTILA CAERULEA and that at various times he used both names Van Rossem and Hachisuka (1937) referred for the San Esteban birds. to two specimens of Blue-gray Gnatcatcher I had available a total of nine specimens taken on San Esteban Island in January as from San Esteban Island, three males and six being “unmistakably” of their newly described females, all adult birds. This sample included Sonoran form P. c. grucilis. Later van Rossem all the birds available to van Rossem and two ( 1945b) recanted, and assigned these speci- more recently taken skins. Two of the males mens to obscura, the race of the Cape region were taken in January; all others were April of Baja California. He assumed that these birds. birds were representative of the breeding pop Comparison of the San Esteban material ulation on the island, although no specimens with small seasonally comparable series of had been taken during the breeding season. specimens from both Sonora and the Cape Miller et al. (1957) incorporated the report region of Baja California showed that the of wintering birds in their account of P. c. greatest similarity is with frutercuhs of So- gracilis, but did not mention a breeding pop- nora. The Baja California birds are brighter ulation on San Esteban; it appears that van yellow on the throat and head, slightly grayer Rossems’ 1945 reassessment was overlooked. on the back, and have a richer, darker chest- The two January specimens, one male and nut patch at the bend of the wing than the one female, are still the only material available Sonoran birds. In all these characters, the San for examination. I have compared these birds Esteban birds agree with those from Sonora, with samples of both obscura and amoenissima but they are, as van Rossem (1930a) men- (no grads were available) at the U.S. Na- tioned, slightly paler dorsally. tional Museum, and find that they fit best with Thus, I feel that van Rossems’ original as- the latter. The measurements of these birds signment of the San Esteban population of are within the range of variation of both races Verdins to the race A. f. fraterculus was cor- (Grinnell 1926), but in color they compare rect, and that he erred in later calling it fluui- closely with examples of amoenissima from ceps. The Verdins thus show a relationship the western United States. to the Sonoran rather than to the Baja Cali- It is not definitely established that there is forian avifauna. a breeding population of this species on San Esteban Island. I did not see this gnatcatcher TOXOSTOMA CURWROSTRE there in late March 1962 or in mid-April 1966, When he described a race of Curve-billed and no breeding material is available. The Thrasher, T. c. insularurn, as endemic to San Blue-gray Gnatcatcher is absent as a breeding Es&ban and Tiburon islands, van Rossem bird in the nearby coastal areas of both Baja (1936b) had but one bird from the former California (Grinnell 1928) and Sonora (van island, which he made the type, and two first- Rossem 1945b). At any rate, the specimens year birds from Tiburon. He later (van Ros- available, which are probably wintering in- sem 1932, 194513) reported five additional dividuals, show no indication of a relationship specimens from Tiburon, but no further ma- between the avifaunas of San Esteban Island terial was taken on San Esteban. Van Rossem and southern Baja California. (1945b:195) stated that “It is now possibly CARPODACUS MEXICANUS extinct on San Esteban, for suitable associa- tions are of limited area and investigations On the basis of sight records, Townsend (1923) subsequent to 1930 have failed to find it.” I referred to the House Finch on San Esteban did not find this species on the island during as C. m. frontalis. Van Rossem (1931) listed my visits in 1962 or 1966. One wonders if this three April specimens under the same name. species is or was really a part of the avifauna Later, van Rossem (1945b) referred to winter of San Esteban, or if the single individual specimens of the race front&, but he called known from there may have been a waif from the resident form ruberrimus, the subspecies the larger nearby island. of southern Baja California. He noted that the This is the only one of the geographically single resident male specimen available was variable species in the San Esteban avifauna paler than birds from Cape San Lucas, Baja that van Rossem ( 1942, 1945b:24) considered California, but that five females were quite to be definitely of Sonoran affinities. This is typical of the southern peninsular race. The certainly so, as the species does not occur in A.O.U. Check-list (1957) includes San Esteban Baja California. Island in the range of ruberrimus, but this 92 RICHARD C. BANKS form is expanded to include populations of characterized bangsi, the only race on the southern Sonora which van Rossem considered mainland of Baja California south of San Ig- a distinct race, sonmien& Miller et al. (1957) nacio, as darker than deserticola, and cana as also included son&en& in ruberrimus, but paler. In describing the birds from Carmen assigned the latter to San Esteban Island Island (A. b. carmeni, merged with bangsi by “somewhat doubtfully on [the] basis of one Miller et al. 1957, but kept separate by Banks specimen.” 1963) van Rossem (1945a:244) stated that “in I had available for examination two males series they average slightly grayer and paler and three females from San Esteban Island, than bangsi. At one time I was inclined to including at least the one male and two of place them as nearest to canu . . . .” This is the the females mentioned by van Rossem (194513). only direct comparison that I have found of All were taken 17-18 April 1930, and all but canu with birds from farther south in Baja one (female) are labeled as breeding birds. California. All are quite worn. I have examined a series from San Esteban Neither of the males shows the intensity or Island consisting of seven birds taken 11-12 the extent of red that is typical of ruberrimw, January and 14 birds taken between 22 March or the darkening of other parts of the plumage. and 26 April. These specimens were compared None of the females is dark as are the females with series of destiicola, bangsi, pacifica, of typical ruberrimus, nor are the birds very and grisea taken at comparable seasons (no like members of the race sonoriensis,’ which I April pacifica were available). The paleness am inclined to keep separate from ruberrimus. and grayish coloration of cana is quite distinct, All specimens from San Esteban are within and these features set that population apart the range of variation of, and similar to typical quite sharply from all others. In being gray, individuals of, the widespread race frontalis. cana is similar only to grkea, but the latter is Carpodacus m. frontalis is the form of House very much darker. Of the brown races, pa- Finch which occurs at the latitude of San cifica is the palest although in coloration it is Es&ban Island on both the peninsula and the not greatly different from bangsi, from which mainland (Miller et al. 1957), although it is it is best distinguished by its much greater near its southern limits in both areas (Grin- size. The San Esteban race, canu, is so distinct nell 1928; van Rossem 1945b). Thus the pres- from the other races that I cannot hazard an ence of C. m. frontalis on San Esteban Island opinion on its nearest relative. It bears no does not reveal whether the affinities of the greater resemblance to birds of Baja California avifauna of the island are with Sonora or Baja than of elsewhere. California; certainly it does not indicate a relationship to the avifauna of the Cape region CONCLUSIONS AND SUMMARY of Baja California. Seven species of geographically variable birds AMPHISPIZA BILINEATA have been reported as part of the resident avi- fauna of San Esteban Island in the Gulf of In the original description of the Black- California, Sonora. Two of these, the Curve- throated Sparrow, A. b. cana, of San Esteban billed Thrasher and the Blue-gray Gnat- Island, van Rossem (193Oc:223) characterized catcher, may not actually have breeding pop- it as “palest and grayest of the races; most ulations there, at least at the present time. Of closely resembles Amphispiza bilineata deserti- the seven, only two, the Verdin and the doubt- cola but . . . coloration paler . . . .” In a dis- fully resident thrasher, can be used in a de- cussion of A. b. bangsi in the same paper, he termination of the relationships of the avi- called that race darker than desertioola and fauna, and both show affinities to that of “decidedly darker than canu.” Somewhat later Sonora. The Ash-throated Flycatcher and the (van Rossem 1945a) he noted that “five worn examples from the vicinity of San Igancio east- House Finch are represented by the same ward to the Gulf (lat. 27” ) are seemingly like form on San Esteban and in both Sonora and equally worn deserticota in color . . . .” Still, Baja California at the latitude of the island. he repeated the idea that the San Esteban avi- Neither the Ladder-backed Woodpecker nor fauna showed relationships to southern Baja the endemic Black-throated Sparrow shows California, and noted that “in grayness of greater similarity to the Baja California or coloration cana stands out conspicuously from Sonoran avifauna. The slight balance of evi- deserticolu of the peninsula and still more so dence indicates that the avifauna of San Es- from pacifica of Tiburon Island and the south &ban Island is best considered with that of em Sonora mainland.” Nonetheless, he again Sonora. AVIFAUNA OF SAN ESTEBAN ISLAND, SONORA 93

ACKNOWLEDGMENTS Lower California, Mexico. Trans. San Diego Sot. Nat. Hist. 6:211-212. This project was begun while I was employed LANYON, W. E. 1961. Specific limits and distribu- at the San Diego Natural History Museum and tion of Ash-throated and Nutting Flycatchers. was completed in the Bird and Mammal Lab- Condor 63:421449. oratories of the Division of Wildlife Research, LINDSAY, G. E. 1962. The Belvedere expedition to the Gulf of California. Trans. San Diego Sot. Bureau of Sport Fisheries and Wildlife, at the Nat. Hist. 13:144. U.S. National Museum. For the loan of speci- LINDSAY, G. E. 1966. The Gulf islands expedition mens from San Esteban Island I am indebted’ of 1966. Proc. California Acad. Sci.. Ser. 4. 30: to Richard P. Phillips and Mrs. Suzanne Bond 309-355. MILLER, A. H., H. FRIEDMANN, L. GRISCOM, AND R. T. of the San Diego museum, Thomas R. Howell MOORE. 1957. Distributional check-list of the of the University of California, Los Angeles, birds of Mexico. Part II. Pacific Coast Avifauna and Dean Amadon of the American Museum no. 33. of Natural History. My own collecting was RIDGWAY, R. 1907. The birds of North and Middle America. U.S. Natl. Mus. Bull. 50. Part 4. done under permit from the Direction General TOWNSEND, C. H. 1923. Birds collected in Lower de la Fauna Silvestre of Mexico, on trips spon- California. Bull. Amer. Mus. Nat. Hist. 48:1-26. sored by the Belvedere Scientific Fund and VAN ROSSE~~, A. J. 1930a. The races of Autipurus supported in part by National Science Founda- flaoiceps (Sundevall). Trans. San Diego Sot. Nat. Hist. 6199-202. tion Grant GB-2317. I am particularly grate- VAN ROSSEM, A. J. 1930b. New Sonora races of ful to Mrs. Roxie C. Laybourne for her inde- Toxostomu and Pheugopedius. Trans. San Diego pendent evaluation of some of the San Esteban Sot. Nat. Hist. 6:207-208. Island specimens during the course of the VAN ROSSEM, A. J. 193Oc. Four new birds from study. Wesley E. Lanyon read the account of northwestern Mexico. Trans. San Diego Sot. Nat. Hist. 6:213-226. the Ash-throated Flycatcher and provided in- VAN ROSSEM, A. J. 1931. Report on a collection of formation on the Mexican population. land birds from Sonora, Mexico. Trans. San Diego Sot. Nat. Hist. 6:237-304. LITERATURE CITED VAN ROSSEM, A. J. 1932. The avifauna of Tiburon AMERICAN ORNITHOLOGISTS ’ UNION. 1957. Check- Island, Sonora, Mexico, with descriptions of four list of North American birds. 5th ed. Amer. new races. Trans. San Diego. Sot. Nat. Hist. 7: Ornithol. Union, Baltimore. 119-150. BANKS, R. C. 1963. Birds of the Belvedere expedi- VAN ROSSEM, A. J. 1942. Range extensions of three tion to the Gulf of California. Trans. San Diego Lower California birds. Condor 44:184-185. Sot. Nat. Hist. 13:49-f_%. VAN ROSSEM, A. J. 1945a. Preliminary studies on GRINNELL, J. 1926. A critical inspection of the the Black-throated &arrows of Baia California. gnatcatchers of the Califomias. Proc. California Mexico. Trans. San-Diego Sot. Nat. Hist. 10; Acad. Sci., Ser. 4, 15:49%500. 237-244. GRINNELL, J. 1928. A distributional summation of VAN ROSSEM, A. J. 194513. A distributional survey the ornithology of Lower California. Univ. Cali- of the birds of Sonora, Mexico. Oct. Papers Mus. fornia Publ. Zoijl. 32:1-300. Zoo]. Louisiana State Univ. no. 21. GRINNELL, J. 1931. The type locality of the Verdin. VAN ROSSEM, A. J., AND HACHISUKA, THE MARQUESS. Condor 33:X3-168. 1937. The Blue-gray Gnatcatcher of southern HUEY, L. M. 1930. A new Verdin from central Sonora. Proc. Biol. Sot. Washington 50:109-110.