THE WILSON BULLETIN A QUARTERLY MAGAZINE OF ORNITHOLOGY Published by the Wilson Ornithological Society VoL. 106, No. 1 March 1994 Pages 1-188 Wilson Bull., 106(1), 1994, pp. 1-17 THE ENDEMIC VIREO OF FERNANDO DE NORONHA (VIREO GRACILIROSTRIS) Storrs L. Olson^ Abstract.—The Noronha Vireo (Vireo gmcilirostris) is endemic to the small oceanic island ofFernando de Noronha offthe easternmost tip ofBrazil. Although derived from the Red-eyed Vireo (F. olivaceus) complex, the Noronha Vireo isdifferentiated stronglyin coloration, plumage pattern, and morphology and fully merits recognition as a distinct species. It is a smallerbirdwith a much more rounded wing, longer, more slenderbill and amoreelongatedtailandtarsus.Theseappeartobespecializationsforgleaningsmallinsects fromfoliage,particularlytheundersidesofleaves.Thebirdsareabundantwhereappropriate habitatismaintained. Thefewavailabledataonreproductiveandmoltcycles,nesting,and vocalizationsin V. gracilirostrisare summarized. Received 11 November 1992, accepted24 March 1993. The archipelago ofFernando de Noronha is the easternmost extension ofland in the Neotropics, lying 345 km eastofthe eastern tipofmainland Brazil (3°50'S, 32°25'W). It consists ofone main island with a string of minor rocks and islets at its northeastern end and various other scattered stacks. The total land area is 18.4 km^. The island is volcanic in origin and before its discovery in 1503 probably was almost entirely forested. The avifauna consists of the usual complement of tropical seabirds, an as yet undescribed extinct flightless rail (Olson 1982), the Eared Dove {Zenaida auriculata), and the easternmost populations in the world of tyrant flycatcher(Tyrannidae) and vireo (Vireonidae). The flycatchergen- erally is considered to be an endemic subspecies of the Large Elaenia {Elaeniaspectabilisridleyana)(Traylor 1979),whereastheNoronhaVireo is a highly distinctive endemic species, Vireo gracilirostris. Little has been written about V. gracilirostris, and apart from a sketch ofthe bill (Sharpe 1890:478), a black-and-white photograph ofa live bird ' Dept, ofVertebrate Zoology, National Museum ofNatural History. Smithsonian Institution, Wash- ington, D.C. 20560. 1 2 THE WILSON BULLETIN • Vol. 106, No. 1, March 1994 (NacinovicandTeixeira 1989), and photographs ofthe skull and sternum (Barlow and Bortolotti 1989), the species has not otherwise been illus- trated. I here bring together the scattered literature on this species, my own observations made on the island nearly 20 years ago, and further illustrative material. HISTORY OF OBSERVATIONS AND COLLECTING Opportunitiesforstudyingvireoson Fernandode Noronhahavebeen limited(Table 1). The first recorded observation ofthe species is that ofMoseley (1892:68), a naturalist on the ChallengerExpeditionwhonoted“asmallwarbler{Sylvia), withgreenishbrownplum- age” when the Challenger visited the island in 1873. Unfortunately, the expedition was refusedpermissiontoconductinvestigationsontheisland,andnospecimenswerecollected. In 1887, the biology ofFernando de Noronha was investigatedcomprehensively forthe RoyalSocietyby H. N. Ridley(1890a,b). Fivespecimensofvireosweretakenatthistime, upon which Sharpe (1890:478) based the original description of V. gracilirostris. Ridley (1888:44) noted only that the vireo was common and “frequents the cashew-nut tree and the cocoa-nut palms. It is a small green bird, like a Willow Wren [Phylloscopus trochilus], very active, but by no means difficult to shoot. We never succeeded in findingits nest.” The species was next encountered by Nicoll (1904:38, 1908:16), who was on the island in 1903. Henotedthatthespecieswasfairlyabundant,andhelikenedittoaEurasianReed Warbler {Acrocephalus scirpaceus) in its actions. On 15 October 1912, Murphy (1915:50) encountered“manyofthesegreenlets... inthefigtreesandinthethicketsnearthebeach” and collected a pairin fresh plumage. TheBlossomExpeditionoftheClevelandMuseumofNaturalHistorycollectedextensively inthe South Atlanticfrom 1923 to 1926. FernandodeNoronhawasvisitedin 1926, when 40 specimens ofvireos were obtained, among other species. The expedition was poorly equipped, however, and experienced many difficulties with personnel and provisions. The only publication dealingspecificallywith the expeditionis a popularaccountbythe leader Simmons(1927),andtheonlyscientificinformationnowretrievableisfromspecimenlabels and the fieldcatalog. The bulk ofthe collection is at the Peabody Museum ofYale Univ., but parts ofit have been ratherwidely dispersed. ThereseemstohavebeennofurtherornithologicalexplorationofFernandodeNoronha until my sojourn in 1973 (Olson 1982). Oren (1982, 1984) visited the island in 1980 and again in 1982. Nacinovic and Teixeira (1989) record ornithological observations made on three visits to Fernando de Noronha in the 1980s. SYSTEMATICS AND MORPHOLOGY Vireo gracilirostris Sharpe Noronha Vireo “small warbler{Sylviay' Moseley, 1892:68. Vireo gracilirostris Sharpe 1890:478 (orig. descr., fig. ofbill).—Nicoll, 1904:38.—Nicoll, 1908:16.-Hellmayr, 1935:144.-Pinto, 1944:401.-Santos, 1948:177.-Warrenand Harrison, 1971:212.-01son, 1982:482.-Oren, 1982:13.-Oren, 1984:36.-Sick, 1984:644.—Nacinovic and Teixeira, 1989:723 (photograph of live bird).—Barlow and Bortolotti, 1989:1536-1537, 1540-1545 (skull and sternum figured), 1547.— Ridgelyand Tudor, 1989:150. Olson • NORONHA VIREO 3 Table 1 SourcesofObservationsand Specimensof Vireo gracilirostrisat Fernando de Noronha Expedition Duration N' References Challenger 1-2 Sept. 1873 0 Moseley 1892 Royal Society 14 Aug.-24 Sept. 1887 5 Ridley 1888, 1890a, b Valhalla 20-25 Dec. 1903 5 Nicoll 1904, 1908 Daisy 15 Oct. 1912 2 Murphy 1915 Blossom 18 March-26 April 1926 40 Simmons 1927 Smithsonian 6 July-18 Aug. 1973 22 Olson 1982 Museu Goeldi 16 Nov. 1980 17 Oren 1982, 1984 1-13 Dec. 1982 Museu Nacional 25 Sept.-l Oct. 1983 7 Nacinovic and Teixeira 1989 8-22 June 1986 25-30 Sept. 1988 'N=numbercollected. Vireosylvagracilirostris.—M\xrph.y, 1915:50.—Murphy, 1936:148. Vireo olivaceus gracilirostris.—MQyQV de Schauensee, 1966:424.—Mayr and Short, 1970: 72.—Orenstein and Barlow, 1981:4, 20, 32. Localpeopleontheislandcallthe vireo ""sibito."' Although Oren (1984) spelled the name ^"sebito,'" Nacinovic and Teixeira (1989) use "'sibito/' which is the spelling I was given. This name, according to Oren (1984), is used in northeastern Brazil for various nondescript birds. As the preceding synonymy shows, the Noronha Vireo has almost al- ways been treated as a distinct species. In his original description, Sharpe (1890) ventured that “there is no doubt that the Fernando Noronha bird comes nearest to V. magister"" a conclusion repeated by Sick (1984), probably on Sharpe’s authority only. Vireo magister is a vicariant form oftheBlack-whiskeredVireo(K altiloquus)thatoccursinYucatan, Belize, and Grand Cayman. In discussing V. gracilirostris, which they considered to stand “apart on so many counts that it amply deserves recognition as a full species,” Ridgely and Tudor (1989:150) stated erroneously that “since Hellmayr [1935] this form has been considered merely an insular race of V. oli- vaceus.'' Although Hellmayr synonymized species on many occasions, this was not one ofthem, the deed apparently having been done first by Meyerde Schauensee (1966) but merely in a compilation with no system- atic revisionary study. What Hellmayr (1935:144, footnote) actually said is still pertinent to- day: 4 THE WILSON BULLETIN • Vol. 106, No. 1. March 1994 “This peculiar species is quite distinct from the continental V. vires- cens[=olivaceus] chivi, fromwhichitdiffers, inadditiontocoloration, by proportionately much longer tail, very much slenderer as well as longer bill, and shorter first primary, the latter being either equal to, or a little longer than, the seventh. In general coloration it is indeed not unlike V. magister, though it may be readily distinguished from itby much shorterwings, slenderertarsi, much smallerand slenderer bill, the absence ofthe dusky loral spot and the grayish suffusion on thesidesofthebody,etc. Icannotbelievethatanygeneticrelationship really exists between these birds. .” . . Plumage.—Tht subdued coloration and slender proportions of V. gra- cilirostrisgive ita strong superficial resemblance to an Old World warbler (Sylviidae), as noted by several early naturalists. The species has been described in some detail by Sharpe (1890) and Hellmayr (1935), and additional specimen measurements are found in Oren (1982) and Naci- novic and Teixeira (1989). I will therefore confine my remarks to direct comparisons with its presumed closest relative, V. olivaceus chivi (I use the term here in a collective sense to mean all of the South American populations ofthe complex, exclusive of V.flavoviridis). Compared to V. o. chivi (Frontispiece), the dorsum of the Noronha Vireoisbrownishanteriorlyratherthangreen, andthegraycapislacking, sothatthecrownisessentiallythe samecolorastheback. Mostspecimens in North American collections are from the indifferently prepared, and exceedinglywornandfaded, seriestakenbytheBlossomexpedition. These give the appearance ofa very brownish or grayish bird, whereas in fresh plumage the lower back, rump, and margins of the flight feathers are decidedly green. There is some individual variation, as one specimen in the series I obtained is much grayer above, with little brown or green in the plumage. Theblackdorsalborderofthe superciliary stripeofV. o. chiviis lacking in V. gracilirostris, the superciliary itself is less pronounced and buffy rather than whitish, and the dark preocular spot is brownish rather than blackish. The underparts of V. gracilirostris are washed with buff, palest on the lowerbelly but not white with greenish flanks as in V. o. chivi. The underwing and under tail coverts are yellow in V. o. chivi but bufly in V. gracilirostris, although the crissum is yellowish buffin some individuals. One specimen that I obtained is still mainly in the lax, flufly juvenile plumage, with the crown, back, and secondarycoverts a rich rustybrown, rathersimilarto thatinthe V. olivaceusgroup ingeneral butmore reddish than injuveniles of V. olivaceus itself. Soft-part colors.—y[y annotations indicate that the iris is brown, the T Olson • NORONHA VIREO 5 Fig. 1. Comparison ofexternal morphology of Vireogracilirostris(top in each pair, on left in dorsal view ofbills) with that of V. olivaceus chivi (bottom in each pair, on right in dorsal view ofbills) showing bills in lateral and dorsal views, wingtips in dorsal view, and tails in ventral view. Scale = 2 cm. upper mandible brownish-horn, the lower mandible whitish, and the feet light bluish-gray. External morphology.— \\q more obvious distinctions of V. gracili- rostris in external morphology and proportions are shown in Figs. and 1 3 and Table 2. The longer and obviously more slender bill (both in width 6 THE WILSON BULLETIN • Vol. 106, No. 1, March 1994 Fig. 2. Comparison ofskeletal elements of Vireo olivaceus chivi (USNM 558010—on the left in each pair) with V. gracilirostris (USNM 491946—on the right in each pair). A, skulls in dorsal view; B, mandibles in dorsal view; C, sterna in ventral view; D, coracoids inventralview;E,pelvesindorsalview;F,wingelements;G,legelements.Allfiguresabout % natural size. and depth) give rise to the specific name, and the tail is much longer than in the V. chivi group. As noted above, Hellmayr (1935) remarked on the shorter first primary, to which may be added that the entire wing tip (distance between the tips ofthe secondaries and tips ofthe primaries) is shorter (Fig. IE, F). Despite this and the fact that the pectoral girdle and wing elements are smaller in V. gracilirostris, the overall wing size, as indicated by the chord length, does not seem to be as correspondingly reduced, as there is considerable overlap in measurements. The differences in wing shape are shown in Fig. 3. The wing in V. gracilirostris is broad and rounded, whereas in the highly migratory V. olivaceus the wing is long and pointed. The differences are exaggerated here by contrasting the most migratory form with perhaps the most sed- entarymemberofthe V. olivaceuscomplex. Itislikelythat V. gracilirostris does not differ as much in wing shape from the more sedentary forms of the V. o. chivi group. Unfortunately, no spread wings were available for any ofthese taxa. By simply tracing the outline ofthe two specimens shown in Fig. 3 on graph paper, I found that the wing area was identical (27.0 cm^). This probably indicates that wing area cannot be reduced below a certain Olson • NORONHA VIREO 7 Fig. 3. Comparison ofwing shape in Vireo olivaceus (left) and V. gracilirostris (right). Although the wing shape is quite different in these two species, with that in the highly migratory V. olivaceusbeing very long and pointed, the surface area was identical in these two specimens. Scale = 2 cm. amount without adversely affecting arboreal foraging, so that V. gracili- rostrismaintainsthesamewingareawhilehavingamuchsmallerpectoral girdle and associated musculature, which is advantageous in a sedentary, insular species. — Osteology andBortolotti(1989)comparedaspectsoftheskel- . eton of V. gracilirostris with other members ofthe F. olivaceus complex (but not V. chivi or V. Jlavoviridis), among which it was almost always the most divergent and in one principal component analysis (Barlow and Bortolotti 1989: fig 4) was widely separated from the other taxa. Addi- tional data and analyses are provided here in Table 2 and Figs. 2 and 4. Such standard indicators of overall size as cranium width and femur length show V. gracilirostris to be, on average, a slightly smaller bird but with considerable overlap. All elements ofthe wing and pectoral girdle are much smaller in V. gracilirostris, but so too is the pelvis. In plotting the combined lengths ofsternum and coracoid versus combined lengths ofthe wing elements (Fig. 4b), V. gracilirostris falls out on the same slope as V. olivaceus, which seems to indicate that it is simply a smaller bird, rather than having a disproportionately reduced pectoral girdle. Barlow and Bortolotti (1989) found that relative to the total length ofthe wing elements, the humerus in V. gracilirostris is shorter and the carpometa- carpus longer than in related taxa, whereas the ulna remains the same. In absolute measurements the lengths of the tail, culmen, and tarsometa- tarsus are greater in V. gracilirostris, with no overlap. Bill width in the skeleton is consistently smaller, and bill depth in skins is likewise small in V. gracilirostris, although with some overlap. In sum, compared to V. olivaceus (including V. o. chivi), the Noronha Vireo is a smaller bird with a longer, more slender bill, and much longer 8 THE WILSON BULLETIN • Vol. 106, No. 1, March 1994 tail and tarsus. The wing is rounded rather than pointed, but retains the same surface area, possibly in part through relative elongation of the carpometacarpus. —Unfortunately, I was unable to obtain mass data for the spec- imens I collected; the only such information comes from Nacinovic and Teixeira (1989), who gave the masses ofseven individuals of V. gracili- rostrisas ranging from 1 1.5 to 20 g (mean 16.5). This seems a rathergreat disparity for a small non-migratory passerine, particularly as one bird at the low end ofthe range (12 g) was noted as being very fat. The mean mass of seven individuals of V. o. chivi from Peru was 13.8 g, and mi- gratory V. o. olivaceusfrom Pennsylvania ranged from 12.0 to 25.1 gwith a mean of 16.7 g (Dunning 1992). ECOLOGY AND BEHAVIOR — Distribution andabundance. Vireogracilirostrisisconfinedtothe main island ofFernando de Noronha anddoes notoccuron any ofthe adjacent islets, the largest ofwhich is Ilha Rata, where the elaenia exists but not the vireo. The birds are generally distributed throughout the island wher- ever there is forest, or at least scrub, but are absent from large areas in the center ofthe island that have been cleared for airport runways, fields, etc. The greatest numbers occurred in the forest around the famous pho- nolitic plug known as Morro do Pico, and in the forest that covered the western quarter ofthe island, where birds were truly abundant. My only attempt to census vireos took place on the morning of 18 July 1973 from about 08:30 to 10:30 h while walking along the woodland trailatthewestern end oftheisland fromthebase ofMorro DoisAbragos to the lighthouse at Alto da Bandeira, a distance of 2 km. During this time, I saw 31 vireos and counted 93 singing males. Although the birds areafavoritetargetofchildrenwhokillthemtoeatorpurelyforrecreation (Nacinovic and Teixeira 1989), this is probably a minor source ofmor- tality, and the species should not be in any danger as long as existing forested areas ofthe island are preserved. Feeding and general behavior.—The Noronha Vireo is a curious and tame bird that allows close approach by humans and is generally quite tolerant of the presence of conspecifics. Birds that seemed to work too closetoeachotherwouldsometimessnaptheirbillsaudiblyatoneanother and move apart. I twice observed chases followed by weak singing by the “victor,” but usually there was little aggressive behavior between birds, whichoftenforagedin proximityto oneanotherinconsiderablenumbers. Birds in pairs, presumably mates, scold human intruders with great fre- quency while approaching closely. 2 Olson • NORONHA VIREO 9 Table 2 Skinand Skeletal Measurements (mm) of Vireogracilirostrisand V. olivaceus V.gracilirostris V.olivaceus Measurement Range Mean Range Mean 1. Wingchord 60.2-66.5 62.9 62.4-73.5 68.1 2. Tail length 56.3-66.3 60.1 44.4-55.9 50.4 3. Culmen length 14.3-16.2 15.1 11.7-14.0 12.9 4. Bill depth 3.3-3.9 3.5 3.7-4.7 4.1 5. Cranium length 16.1-17.3 16.8 16.9-18.6 17.8 6. Cranium width 13.3-14.2 13.7 13.6-14.8 14.2 7. Bill length 16.2-17.7 16.7 14.0-16.6 15.6 8. Bill width 5.4-5.9 5.6 6.2-7. 6.7 9. Pelviswidth 8.0-9.0 8.6 9.1-10.5 10.1 10. Sternum length 11.3-12.3 12.0 14.9-18.0 16.9 11. Carina depth 3.4-4.6 4.0 5.3-6.4 5.9 12. Coracoid length 12.3-13.1 12.7 14.9-16.8 16.0 13. Humerus length 13.6-14.8 14.5 15.6-17.5 16.6 14. Ulna length 16.4-17.9 17.4 19.5-22.0 21.0 15. Carpometacarpus length 8.5-9.2 9.0 10.2-11.9 11.3 16. Femurlength 13.2-14.5 14.2 14.1-15.9 15.1 17. Tibiotarsus length 26.3-28.5 27.6 23.8-27.5 25.8 18. Tarsometatarsus length 19.6-21.3 20.6 16.7-18.9 18.1 Notes: Measurements 1-4 are from skins, the specimensofV. olivaceusbeingofSouth American chivigroup. For Mmeeaassuurreemmeennttss51-a1n8dar2e,fNro=ms2k2e;leftoornsm,eNasu=re8mfeonrtesac3hatnadxon4,.NThe=s9kefloertonL.sogrfacVi.liorloisvtarciseuasnidncNlud=ef1o3urfoIr',oI.',ool.ivcahcievau.s fromNorthAmericaand4fromSouthAmericathatarelabelledasbeingofthechivigroup,althoughoneisprobablya misidentifiednorthernmigrant.Whennotastandardmeasurementorself-evident,themanneroftakingeachmeasurement isspecified as follows: 4. At anterior margin ofexternal nostril. 5. From nasofrontal hinge to posteriormost extentof braincase.6.Greatestwidth. 7.Fromnasofrontalhingetotip. 8.Atposteriormarginofbonynostril.9. Acrossantitro- chanters. 10.Frommidlineofmanubrialforktoposteriormargin. 11.Fromtheventralsternalplatetotipofcarina. 12. Fromheadtoexternaldistalangle. 17. Includingcnemialcrest. Oren (1984) wrote that the vireo was flexible in procuring food, which is always small arthropods, and may forage from the tops ofthe trees to the ground and in leaves, on trunks, or in inflorescences. He also noted thatithabituallyhangsheaddown. Analysisofstomachcontentsreported by Nacinovic and Teixeira (1989) revealed a variety of insect remains (Coleoptera, Hymenoptera, Orthoptera, and Trichoptera) and a few small fruits ofUlmaceae. In my experience, although the birds did show some variability in feeding behavior, most individuals spent more time foraging on the un- dersides ofleaves than in any other feeding activity. In their most char- acteristic pose, the birds would grasp the edge ofa leafin their feet and bend over upside-down to feed from the undersurface. One bird stayed in this position for 15-16 sec, gleaning insects the entire time. Another 10 THE WILSON BULLETIN • Vol. 106, No. 1, March 1994 Depth Bill Culmen Length Osteological Wing Length Length Tail Wing Chord Femur Length Length Wing Osteological Osteological Leg Length Femur Length