A New Species of Pygmy-Owl (Strigidae:Glaucidium) from the Pacific Slope of the Northern Andes

Home , Northern pygmy owl, Pacific Slope, Pygmy owl

TheAuk 116(2):305-315,1999

A NEW SPECIES OF PYGMY-OWL (STRIGIDAE:GLAUCIDIUM) FROM THE PACIFIC SLOPE OF THE NORTHERN ANDES

MARK B. ROBBINS1'3 AND F. GARY STILES2 'Division of Ornithology, Natural History Museum, University of Kansas, Lawrence,Kansas 66045, USA; and 2Instituto de Ciencias Naturales, Universidad Nacional de Colombia,Apartado 7495, Bogota, Colombia

ABSTRACT.-We describe a new species of pygmy-owl that is restricted to very wet cloud forest at 1,400 to 2,000 m in elevation along the Pacific slope of the western Andes of Colom- bia and Ecuador. This taxon had been long overlooked as a result of its morphological similarity and geographic proximity to the Andean Pygmy-Owl (Glaucidium jardinii), but it is vocally very distinct and exhibits subtle but consistent morphological differences as well. Vocally, the new species is most similar to the Costa Rican-Panamanian form G. costaricanum and to the nominate subspecies of Northern Pygmy-Owl (G. gnoma), but again, consistent differences exist in vocalizations and morphology. Glaucidium costaricanum has long been considered a subspecies of G. jardinii, but at least one recent author treated it as a subspecies of G. gnoma. Genetic, vocal, ecological, and morphological data indicate that G. costaricanum should be elevated to species level, and it along with the new species are considered allospecies within the Northern Pygmy-Owl superspecies complex. Received 20 July 1998, accepted 2 December 1998.

RECENT STUDIES OF THE TAXONOMY of New cific slope near the Colombia/ Ecuador border World pygmy-owls (Glaucidium)have revealed (ca. 4 km SSW of Chical, provincia Carchi; that the diversity of this group has been grossly 00?54'N,78?12'W) that unequivocally belong to underestimated(Vielliard 1989, Konig 1991,Hei- the same taxon as the bird netted near Mindo drich et al. 1995,Howell and Robbins1995, Rob- the previous year. Although Robbins recog- bins and Howell 1995).This is not surprisinggiv- nized that the voice of this undescribed taxon en that morphological and vocal variation in was distinct from that of the upper-elevation these owls are on the same order as the equally Andean Pygmy-Owl (G. jardinii), its affinities cryptic,but better-studiedEmpidonax flycatchers remained unclear. Robbins sent tape record- (Stein 1958, Johnson 1980, Lanyon and Lanyon ings of the Chical birds to Stiles, who recog- 1986,Johnson and Marten1988). However, unlike nized that they were nearly identical to those the diurnal, relativelynumerous, and highly vo- of a bird he had recorded and collected on 3 cal Empidonax,pygmy-owls often occur in low June 1992 in subtropical cloud forest at Alto de densitiesin tall, humid forestand are notoriously Pisones, Risaralda, Colombia (Fig. 1). Stiles difficultto locate and collect. also noted that the voice of these owls was most On 27 an avifaunal in- August 1987, during similar to that of the isolated Costa Rica/ Pan- ventory conducted by the Academy of Natural ama form G. jardinii costaricanum, and in fact Sciences of Philadelphia and the Museo de had tentatively classified his Colombianbird as Ciencias Naturales, Quito, Fred and Jody Shel- costaricanumon that basis. This form had not don and Robbins mist netted a male pygmy- been considered even subspecifically distinct owl (ANSP 180178) along a Pacific-sloperidge from nominate jardiniiof the high Andes until of relatively pristine, humid subtropical forest at 1,550 m above the town of Mindo, provincia Kelso's (1937) evaluation. Since that time, cos- Pichincha,Ecuador. No definitive identification taricanumhas been treated as a subspecies of was made of the specimen at the time for lack jardinii (Peters 1940, Ridgely and Gwynne of comparative material. The following year, 1989, Stiles and Skutch 1989, Sibley and Mon- however, Robbins recorded the voice and col- roe 1990, AOU 1998). However, based on mo- lected a male and female (ANSP 181043-4) at lecular studies, Heidrich et al. (1995) treated 1,725 m in humid subtropical forest on the Pa- costaricanumas a subspecies of the Northern Pygmy-Owl (G. gnoma) of North America, a possibility adverted on the basis of voice by 3 E-mail: [email protected] Konig (1991). Meanwhile, Robbins located 305 306 ROBBINS AND STILES [Auk, Vol. 116

FIG. 1. Adult male Glaucidium nubicola from Alto de Pisones, Colombia (ICN 31200). Note the mininal or complete lack of pale spotting or barring on the back, sides of the chest, and flanks that distinguishes this species from the similar appearing G. costaricanum, G. jardinii, and G. bolivianum. Photograph by F. Gary Stiles.

three additional specimens of the new South vincia Carchi, approximately 4 km SSW Chical, American taxon, all identified as nominate jar- S side Quebrada San Jose of the Rio Blanco, dinii, in North American museums. 00?54'N, 78?12'W, elevation 1,725 m, 16 August Heidrich et al. (1995) compared the mito- 1988; collected by Mark B. Robbins. Recordings chondrial DNA sequences (cytochrome-b gene) of voice deposited at the Library of Natural of the Mindo bird with those of a number of Sounds, Cornell Laboratory of Ornithology, other Glaucidium taxa, including costaricanum. Ithaca, New York (LNS 49158, 49185, 49187). Their findings and further molecular work by Diagnosis.-Typical Neotropical pygmy-owl Nathan H. Rice and Robbins (see below) con- in appearance, especially similar to temperate firmed that the Pacific-slope bird in fact repre- forms G. [gnomal costaricanum, G. jardinii, and sents a distinct, undescribed species and is the recently described Yungas Pygmy-Owl (G. most closely allied to the Northern Pygmy-Owl bolivianum; Konig 1991). The new owl averages complex (including costaricanum) rather than to heavier in body mass and has a shorter tail than G. jardinii. We propose to call this new owl: the other taxa (Table 1), and it has much less extensive pale spotting or barring on the back, Glaucidium nubicola sp. nov. sides of the chest, and flanks than costaricanum, Cloud-forest Pygmy-Owl jardinii, and bolivianum. The black marks on the hindneck (i.e. the so-called "false eyespots") Holotype.-Academy of Natural Sciences of are bordered by white in all non-rufous indi- Philadelphia, No. 181044, adult male (testes 3 viduals, in contrast with the rufous borders of x 2 mm; no bursa of Fabricius), Ecuador, pro- costaricanum. See also Song. April 1999] New Pygmy-Owl from the Andes 307

Descriptionof holotype.-Capitalized color de- scriptions and numbers follow Smithe (1975, 1981). Crown, face, and auricular area brown (closest to Hair Brown 119A) with white spots bordered by Sepia (119). Concealed white nu- chal collar bordered with Sepia (119). Back, scapulars, upperwing coverts, and rump dark brown (closest to Sepia 219) washed with dark bO ?IN O rufous-brown. Scapulars and upperwing coverts with bold white spots, most tinged with pale rufous. Primaries and secondaries with less rufous-brown wash than dorsum and with 7:5 conspicuous, irregularly shaped whitish spots (many with pale rufous tinge) on outer webs, and larger, concealed white spots on inner o +l + +l +l + webs, forming an indistinct band. Tailblackish cllfo o. ? c z (closest to Sepia 119) with five incomplete white bands appearing as irregularly shaped spots. Chin, sides of upper throat, and center of chest white. Center of abdomen and crissum white with buffy brown wash. Center and lower part of throat dark brown (closest to Sepia > E n - -Ln N 0 m 219). Sides of chest rufous brown (closest to 2:~~~~ O~I NO 't "O Raw Umber 223, but with more rufous tinge) o= rt cf o6 with a few inconspicuous white spots extend- 0 _ ing through the abdominal region as streaks. S _ Abdominal streaks becoming darker in color +I +I +I (closest to Sepia 219). Irides yellow, bill green- ish-yellow, tarsi and toes yellow, claws black. See cover for depiction of the holotype. Measurementsof holotype.-Wing chord (un- flattened) 92.0 mm, tail length 46.8 mm, bill

O) H O length (from cere to tip) 11.3 mm, body mass 80 g. =~~~~0 n U N N Etymology.-We have chosen the specific ep- ithet nubicola,which is Latin for "cloud inhab- iting," because the bird is restricted to very hu- ON .O mid cloud forests. gf e NNN _~~~~~~oo U N Ln Specimensexamined: Skins.-Localities are de- picted in Figure 2. Glaucidium nubicola: Co- lombia:Risaralda, Alto Pisones, Mistrato,1,740 m, 05?25'N,76?02'W, 1 male (ICN 31200,Figure +l + + + 1; see Acknowledgments for acronyms of insti- E S ooRatrirAe-i tutions); Valle, San Antonio, 2,000 m, 03?30'N, 76?38'W, 1 male (MVZ 138178). Ecuador: Pi- _U chincha, 3 males (ANSP 181078; CMC 24953, 29972); Carchi, 1 male (ANSP 181044 [holotype]), 1 female (ANSP 181043). In the following lists, locality names are given only to the level of state, province, or department. Glauci- dium costaricanum: Costa Rica: San Jose, 2 males (UCR 2001; LSUMZ 52589); Cartago, 1 male (LSUMZ 153994), 1 female (AMNH 308 ROBBINS AND STILES [Auk, Vol. 116

FIG. 2. Northwestern South America, showing localities at which Glaucidium nubicola has been recorded. The southernmost symbol represents three different sites in provincia Pichincha, Ecuador. See Specimens Examined for specific locality data.

485467); 1 unsexed (AMNH 801545). Panama: diumjardinii: Colombia: Narifio, 1 male, 1 fe- Chiriqui, 3 males (LSUMZ 160719; USNM male (FMNH 287981, 249563);Cauca, 2 males, 485737; AMNH 768851); unknown Panama lo- 1 female (ANSP 17383, FMNH 101287, 102214); cality, 1 male (USNM 154246). Glaucidiumgno- Huila, 1 male (FMNH 102226). Ecuador: Car- ma gnoma: Mexico: Chihuahua, 6 males, 4 fe- chi, 2 males (ANSP 184581, 184582);Imbabura, males (MVZ 143683-4, 119358-9, 119360-1, 2 males (LSUMZ 112509-10); Pichincha, 2 119364, 119378-9, 135144); Veracruz, 1 male males (ANSP 66864, AMNH 485553); Napo, 1 (FMNH 187123); Coahuila, 1 male (KU 31582); male (AMNH 179926);Cotopaxi, 1 male (ANSP Jalisco, 1 male (MVZ 115393); Michoacan, 6 183103); Zamora-Chinchipe, 1 male (ANSP males, 3 females (FMNH 102555-7, 102791-2; 185159). Peru: Piura/Cajamarca, 7 males, 3 fe- MVZ 115391-2, 150017; KU 35073). Glauci- males (LSUMZ 75125, 97581-3, 87284-6, April 1999] New Pygmy-Owl from the Andes 309

97578-80). Glaucidiumbolivianum: Peru: Ama- holotype in having a more rufous-brown wash zonas, 2 males, 2 females (LSUMZ 87287-8, to the entire plumage. The Mindo female 87291; ANSP 118137); San Martin, 1 male (ANSP 181043) has a more extreme rufous (LSUMZ 104450),1 female (LSUMZ 104449); La wash, matching rufous-morphindividuals of G. Libertad, 3 males (LSUMZ 91699-701); Huan- jardiniiand G. bolivianum,with more extensive uco, 2 males, 3 females (FMNH 299047,296610; spotting on its back than the other paratypes, LSUMZ 79676, 73828, 118959); Pasco, 5 males, forming irregular bands. Although ICN 31200 1 female (LSUMZ 128154-7, 105661,128158); was not compared directly with the other par- Apurimac, 1 male (FMNH 299519); Ayacucho, atypes, it agrees closely with the description of 1 male, 1 female (LSUMZ 68817, 69362); Cuzco, the holotype, differing in minor details of col- 3 males, 1 female (FMNH 299517, 311744, oration as follows: dorsum nearest to Raw Um- 311742; LSUMZ 80431); Puno, 1 male, 2 females ber (223), washed anteriorly and laterally with (98096-8). Bolivia: La Paz, 4 males, 4 females rufous-brown (near Mars Brown, 223A). Pale (LSUMZ 90534-6, 101693-7); Cochabamba, 3 spotting on scapulars and remiges whitish, males (LSUMZ 123557-8, 123560). The holo- more or less strongly tinged with Pinkish Buff type of G. jardinii, from the Pichincha area, Ec- (121D) to pale rufous. Sides of breast rich uador, apparently is no longer extant (E. Pas- brown (nearest Mars Brown, 223A), nearly im- quet pers. comm., Museum National d'Histoire maculate; ventral streaks are blackish-brown Naturelle, Paris). centrally (Sepia, 219) shading through Mars Specimens examined: Tape-recordings.-Glau- Brown to Pinkish Buff (121D) on borders giv- cidium nubicola: Colombia: Risaralda, Alto de ing the impression of dark streaking on a buffy Pisones (F. G. Stiles); Narifio, Reserva Rio Nam- background; center of lower breast and abdo- bi, 1,300 m, 01?18'N, 78?05'W (P. Coopmans). men white. Five pale tail-bands occur in all but Ecuador: Pichincha, Mindo (R. Jonsson); Carchi the rufous-morph bird (ANSP 181043), which (holotype; LNS 49158, 49185, 49187). Glauci- has six. Rufous morphs of most (all?) Glauci- dium costaricanum: Costa Rica: Cerro de la dium tend to have more tail bands than do Muerte (UF 884A). Glaucidium gnoma: Mexico: brown or gray morphs (unpubl. data). Sinaloa (LNS 200201, 53150); Colima (LNS Song.-Like virtually all New World Glauci- 57735-7); Coahuila (Hardy et al. 1990); Nuevo dium,the song of nubicolais composed of struc- Le6n (LNS 53441, 56681). Glaucidium jardinii: turally simple notes best expressed as hollow Colombia: Cundimarca (FSU 1352, 1354). Ec- whistles or toots. Although note structure is uador: Loja (LNS 57016) Zamora-Chinchipe similar among taxa, distinct differences exist in (LNS 78948). Glaucidium bolivianum: Peru: La internote length and interval both within and Libertad, Cumpang (LNS 17332, 17347); Cuzco among superspecies groups (Howell and Rob- (R. A. Behrstock); Pasco (LNS 40140, 40144, bins 1995). Because these two features have 40162). Spectrograms in Konig (1991) of the proven taxonomically informative, we used above taxa were also examined. Skins and vo- them to characterizepygmy-owl song patterns. cal material examined of Subtropical (G. par- The length and spacing of notes clearly dis- keri), Colima (G. palmarum), Central American tinguish nubicolafrom costaricanumand gnoma (G. griseiceps), Amazonian (G. hardyi), and Least (Table2). The song of nubicolaconsists of notes (G. minutissimum) pygmy-owls are listed in of constant duration (notes average 161 ? SD Robbins and Howell (1995). of 28.8 ms), delivered in pairs (occasionally in trios). Within note pairs, intranote pauses av- REMARKS eraged 207.9 ? 18.0 ms, whereas pauses between pairs (internote) averaged 344.8 ? 21.0 Individual variation.-The two oldest (ca. 60 ms (Table2). This distinctive pattern of paired years) specimens (CNC 24953, 29972) of nubi- whistles (Fig. 3A-C) serves as a putative syn- cola are somewhat paler than the more recent apormorphy of nubicola,costaricanum, and gno- specimens, likely a result of fading. The Valle, ma and distinguishes this clade from all other Colombia, bird (MVZ 138178) is very similar to Glaucidium(Konig 1991, Howell and Robbins the holotype but lacks suffusion of buffy- 1995). Although Miller (1963) followed the brown wash on the abdomen and crissum. The standard taxonomic treatment of Andean pyg- Mindo male (ANSP 180178) differs from the my-owls at that time (all high elevation birds 310 ROBBINS AND STILES [Auk, Vol. 116

TABLE 2. Selected song characters (xt ? SD in ms) of montane pygmy-owls. Sample sizes are smaller than the number of individuals and songs studied because measurements were restricted to individuals vocal- izing under natural circumstances (i.e. no playback) and for which recordings were sufficiently clear to permit accurate measurements.

Species Internote interval Note length Intranote interval G. gnoma (n = 4) 579.9 ? 96.0 106.2 ? 9.2 321.6 ? 36.5 G. costaricanum (n = 1) 276.2 98.6 185.0 G. nubicola (n = 2) 344.8 + 21.0 161.0 ? 28.8 207.9 ? 18.0 G. jardinii (n = 3) 202.4 ? 30.0 116.0 ? 11.7 G. bolivianum (n = 3) 548.0 ? 72.1 196.7 ? 34.6

2.0 A A A A A A A A A A A 1.0 0 f .AA U H A 0.0

b 2.0 1.0.fv # AA AftAf AA A A4 ftf

o.o-i * w * * * . * ^ * * I 2.0 N 1.0t # a * as n a. C 0.0 -T______I ~~~~~d 2.0

e 2.0 -

1.0o- - 1

0.0-'''-- '''-- ''-''' 0.0 2.0 4.0 6.0 Seconds FIG. 3. Sound spectrograms of primary song of selected pygmy-owls. Y-axis represents frequency in ki- lohertz (kHz). X-axis represents time in seconds. (A) Glaucidium gnoma: Mexico, Colima (LNS 57735), recorded by D. Delaney. (B) G. costaricanum: Costa Rica, Cerro de la Muerte (FSM 884A), recorded by Stiles. (C) G. nubicola: Ecuador, provincia Pichincha, recorded by R. Jonsson. (D) G. jardinii: Ecuador, provincia Za- mora-Chinchipe (LNS 78948), recorded by Robbins. (E) Glaucidium bolivianum: Peru, departamento Cuzco, recorded by R. Behrstock. April 1999] New Pygmy-Owl from the Andes 311 considered as jardinii), he astutely noted that the more distantly relatedjardinii group, which the San Antonio, Colombia, birds had a song has long obscured their systematic relation- similar to nominate gnoma.Although nubicola ships. Presumably,these striking plumage sim- was long confused with jardiniion the basis of ilarities result from convergent adaptations for morphological similarity and geographic prox- color-matchingin very humid montane forests, imity, its note pattern is clearly aligned with whereas the duller, grayer gnoma principally the gnomagroup, and not the jardinii complex inhabits drier pine-oak forests. (i.e. jardiniiand bolivianum).Members of the lat- Breedingand molt.-All three recent Ecuador ter group typically give songs with evenly birds were taken in August; neither the two spaced notes (Fig. 3D-E). males nor the female had enlarged gonads (tes- Morphometrics.-The Cloud-forest Pygmy- tes, 3 X 2 mm; ovary, 7 x 3 mm; ova, <1 mm), Owl has a significantly shorter tail and heavier and all were in fresh plumage. The Mindo bird body mass than gnoma and costaricanumand had no body molt, and both El Chical birds had members of the Andean pygmy-owl complex moderate body molt, heaviest on the nape of (Table1). the female; the male was replacing the right Distribution.-Glaucidiumnubicola may occur sixth rectrix, but otherwise all birds had fresh continuously along the Pacificslope of the An- flight feathers. The nonbreeding status of these des from northwestern Colombia (Antioquia) birds was also reflected in the infrequency of to southwestern Ecuador near the Peruvian their vocalizations; no singing was heard at border, as do many other subtropical taxa of Mindo, and only occasional songs were heard similar cloud-forest habitats (Chapman 1926, at dawn at El Chical. Robbins and Ridgely 1990). The known eleva- The male taken on 1 December at San Anto- tional range of G. nubicolais about 1,400 to nio, Colombia, had small (3.5 mm) testes and 2,000 m; above these elevations it is replaced by exhibited no molt (Miller 1963). The Alto de Pi- C. jardinii. sones specimen from 4 June had testes 6.5 x 4 Habitat.-All seven sites (Fig. 2) where G. nu- mm, no molt, and was in fairly fresh plumage. bicolahas been found are (or were) in wet pri- At this locality on 1 to 3 June 1992, Stiles heard mary cloud forest on steep slopes. The Mindo an individual singing persistently in the im- bird was netted about 2 m above the ground mediate vicinity of the collection site, as well as along a ridge in wet primary cloud forest. Al- persistent singing from a ridgetop some 300 m though the forest near Gualea, only about 20 away on 14 to 17 April 1993;on 16 April, a bird km north of Mindo, has been affected by de- at this site responded strongly to Stiles's whis- forestation, remnant patches on the steepest tled imitation, countersinging and approach- slopes indicate that it also held wet cloud forest ing to within 10 m. Much less persistent sing- (M. Robbins pers. obs.). Both Carchibirds were ing was noted by Stiles in late June 1991 at La found in wet primary cloud forest on steep Planada;on three mornings on the same ridge- slopes. The vegetation and topography at Rio top in early July 1992, he heard only one short Nambi, Alto de Pisones, and Reserva Natural bout of song. Forthese reasons, we infer that G. La Planada are apparently quite similar to the nubicola probably breeds principally between Ecuadorean localities (Salaman and Stiles Februaryand June. Given that the peak breed- 1996). At San Antonio, Valle, Colombia, Miller ing season of small birds at La Planada occurs (1963) collected a singing male from about 45 in the brief dry season in Juneand July (G. Can- m up in the crown of a primary forest tree. Nev- tillo pers. comm., G. Stiles pers. obs.), earlier ertheless, annual rainfall undoubtedly differs breeding by G. nubicola may permit young to between Alto de Pisones and Mindo, given the fledge at a time when potential prey (fledglings negative gradient in rainfall north to south of small birds) are most abundant. along the Colombia and Ecuador Pacific slope Behaviorand ecology.-Similar to other humid (IGAC1989, G6mez 1990). forest pygmy-owls, almost nothing is known Pygmy-owls in humid habitats typically about the natural history of G. nubicola. Most of have darker plumage coloration. Although nu- the handful of observations of such Glaucidium bicolaand costaricanumare phylogenetically re- are of birds being mobbed by passerines; in lated to northern gnoma, their plumage color fact, Miller (1963) watched hummingbirds and and patterns are remarkablysimilar to those of tanagers mobbing a singing male nubicola for 312 ROBBINS AND STILES [Auk, Vol. 116

90 min. About mid-morning on 9 August, the nubicolafrom Mindo (ANSP 180178; =frozen nonvocalizing El Chical female (ANSP 181043) tissue number LSUMZ 12188). These authors was flushed from her perch several meters demonstrated (Heidrich et al. 1995: figure 35) above the ground in moderately dense vegeta- that the closest living relatives of nubicolaare tion along a steep forest trail. Along the same costaricanum (=their "gnoma") and gnoma trail during mid-morning on 16 August, the (=their "californicum"from British Columbia, nonvocalizing male holotype flushed from veg- Canada) and not jardinii.Levels of genetic dif- etation and flew to a perch about 3 m above the ferentiation among these taxa suggested that ground. all three should be treated as species, a conclu- The bird from Mindo had insect fragments in sion supported by the vocal data (see above). its stomach, and the stomach of the female from Inclusion of additional taxa and analyzing 450 El Chical was empty. The holotype had insect base pairs of the cytochrome-bgene have cor- fragments and a small lizard, about 55 mm in roborated and extended these findings (N. H. length, in its stomach. The stomach of the San Rice and Robbins unpubl. data). The latter an- Antonio male held orthopterans and one he- alyses included nominate gnoma(from Michoa- mipteran (Miller 1963). The Alto de Pisones can and Oaxaca,Mexico), an additional sample bird was captured in a mist net along a sharp of costaricanum(LSUMZ frozen tissue 28887, ridge crest, and apparently was attracted by Panama), and a nubicola (feathers from the distress calls of an Ornate Flycatcher(Myiotric- above MVZ 138178) that Heidrich et al. (1995) cus ornatus)in the net. This pygmy-owl also lacked. Preliminary results indicate that these had a 3.5-cm cicada in its stomach. Hence, nu- three taxa form a well-supported and resolved bicolaappears to have the mixed diet of inver- clade (74%of 500 bootstrap replications).Nom- tebrates and small vertebratestypical of the ge- inate gnomawas the sister taxon to costaricanum nus. in 99% of 500 bootstrap replications, with nu- Although members of different species bicola as sister to the gnoma / costaricanum clade. groups (Robbins and Howell 1995, Rice and The molecular data, coupled with morpho- Robbins unpubl. data), the west-slope G. nubi- logical differentiation, vocal differences, dis- cola appears to be replaced ecologically on the tinct habitat preferences, and geographic iso- eastern slope of the Andes by G. parkeri.Both lation meet the requirementsfor species under have been recorded in the same elevational the phylogenetic species concept (Cracraft range (ca. 1,400 to 2,000 m; both likely occur 1983, Zink and McKitrick 1995). From a con- lower) and are replaced by G. jardinii above servation perspective, the significant differenc- 2,000 m. es in these charactersalso qualify these taxa as Systematic affinities-Several nocturnal spe- evolutionary significant units (Ryder 1986, Mo- cies are renowned for their limited plumage ritz 1994). differentiation, even among distantly related Because these differentiated populations are taxa (e.g. caprimulgids and Otus).Perhaps even allopatric, their classificationunder the biolog- more than in these latter groups, Neotropical ical species concept (Mayr 1963, 1969) is am- pygmy-owls show minimal differentiation bivalent because there is no indication on how among taxa in plumage color and pattern and individuals would interact if they came into body size. However, this lack of plumage di- contact. Because molecular data indicate that vergence is accompanied by consistent varia- costaricanumand gnomaare sister taxa, we focus tion among taxa in the patterns and tempo of on their potential interaction. First, even if cos- primary vocalizations. Indeed, quantitative taricanumand gnomacame into close geograph- studies of pygmy-owl vocalizations (Konig ic proximity (the closest populations are sepa- 1991, Howell and Robbins 1995) provided the rated by about 550 km), their dramatic ecolog- first indications that speciation in these owls ical differences would preclude gene flow; had been grossly underestimated. nominate gnoma is primarily an inhabitant of With the advent of molecular tools, relation- drier pine-oak habitat, whereas costaricanumis ships among these taxa are being further elu- found in wet montane cloud forest. Second, cidated. Heidrich et al. (1995) examined 300 even if they were to come into contact in a mon- base pairs of the mitochondrial cytochrome-b tane ecotone, we predict that the dramatic gene in selected pygmy-owl taxa, including the plumage and size differences (Table1; note es- April 1999] New Pygmy-Owlfrom the Andes 313 pecially differences in body mass) would pre- surmise that the gnoma/ costaricanum/ nubicola vent or impede hybridization. Plumage color ancestor expanded from the northern Andes and pattern and morphological differentiation into Central America during a glacial period between costaricanumand gnoma are greater when cooler climate lowered the elevational than is found between species belonging to dif- limits of montane forest life zones. Warming ferent superspecies complexes. Indeed, we re- during an ensuing interglacial period restrict- emphasize that until recently costaricanumwas ed the Central American birds to the highest considered to be an undifferentiated form of mountains of the Cordillera de Talamancaof the distantly related G. jardiniicomplex. Final- Costa Rica and Panamaand the CordilleraCen- ly, as pointed out by Howell and Robbins tral of Costa Rica. Subsequent glacial cycles (1995), consistent differences exist in internote may have permitted this population to expand length and interval among Glaucidiumspecies; northward, occupying the mountains of north- those observed between costaricanumand nom- ern Central America and Mexico and eventu- inate gnomawere about eight times more pro- ally North America proper. An additional re- nounced than those between nominate gnoma cent example of the long-recognized faunal af- and the distantly related G. bolivianum(Table finity between the Costa Rica-Chiriquihigh- 2). Except for plumage coloration (see above), lands and the western Andes (Chapman 1917, molecular, vocal, and morphological differen- 1926; Slud 1964) is the recently described Cho- tiation between nubicolaand costaricanumare co Vireo (Vireomasteri) of the western Andes, equal or greater than between costaricanumand whose closest living relative is probably the nominate gnoma(Heidrich et al. 1995; Tables 1 Yellow-winged Vireo (V carmioli)of the Costa and 2). Thus, we believe these three taxa should Rica-Chiriqui highlands (Salaman and Stiles be considered species under any species defi- 1996). nition. Conservation.-WesternEcuador has suffered We recommend the English name Costa Ri- heavy deforestation during the past 50 years, can Pygmy-Owl for G. costaricanum.Further- with only 15% of the original lowland and more, dramatic differences in voice among montane forests remaining by the late 1980s nominate G. gnoma,G. g. californicum,and G. g. (see Dodson and Gentry 1991). The type local- hoskinsii suggest that at least three species ity surely has been denuded completely, be- should be recognized within the currentNorth- cause most of the slopes already had been ern Pygmy-Owl (Phillips et al.1964, Hardy et cleared when Robbins et al. worked the site in al. 1990, Konig 1991, Heidrich et al. 1995, How- 1988, and loggers with chainsaws were rapidly ell and Webb 1995). Although these authors felling the remaining patches of forest. In Ec- have suggested various taxonomic treatments uador, the upper reaches of the Reserva Ecolo- for the forms currently within gnoma,it seems gica Cotocachi-Cayapas,provincia Imbabura; premature to make changes given the apparent nearly all of the private 3,000 ha of Reserva vocal variation among forms in the United Biologica Los Cedros, provincia Esmeraldas/ States (see Howell and Webb 1995). Molecular Imbabura;and the lower portions of the private and vocal work in progress should clarify spe- 4,500 ha Reserva Maquipucuna, provincia Pi- cies limits within gnoma (R. C. Faucett pers. chincha (Best et al. 1996) contain ratherpristine comm.). At present, we consider the G. gnoma habitat for nubicolaand several other threat- complex to comprise a superspecies with at ened and near-threatenedbird species (Collar least three allospecies: nubicola,costaricanum, et al. 1992, Stattersfieldet al. 1998). and gnoma, leaving open the question of rec- Although human encroachmenton the Pacif- ognition of additional species among the forms ic slope of Colombia has been slower than at currently included within the latter taxon. other Andean sites because of high rainfall, de- Biogeography.-The ancestor of these taxa forestation rates nonetheless have increased probably inhabited the moist forested slopes of dramatically in the past two decades. Major the northern Andes during the late Pliocene road construction in western Colombia is now and early Pleistocene. With the permanent for- providing access to once remote areas and is re- mation of the Middle American landbridge sulting in the same scenario that led to the de- three to five million years ago (Malfait and forestation of western Ecuador (Salaman and Dinkleman 1972, Pindell and Dewey 1982), we Stiles 1996). Viable populations of nubicola 314 ROBBINS AND STILES [Auk, Vol. 116

should occur in at appropriateelevations in the LITERATURECITED following Colombian reserves: Parque Nacion- al Los Farallones, Valle; and Parque Nacional AMERICAN ORNITHOLOGISTS' UNION. 1998. Check- list of North American birds, 7th ed. American Munchique,Cauca; Rio Nambi and Community Ornithologists' Union, Washington, D.C. La Planada Nature Reserves, Narifno. BEST, B. J., T. HEIJNEN, AND R. S. R. WILLIAMS. 1996. A guide to bird-watching in Ecuador and the GalApagos Islands. Biosphere Publications, West ACKNOWLEDGMENTS Yorkshire, United Kingdom. CHAPMAN, F. M. 1917. The distribution of bird-life in We are indebted to the Tracy Pedersen for exqui- Colombia: A contribution to a biological survey site frontispiece. Bob Behrstock, Paul Coopmans, of South America. Bulletin of the American Mu- and Robert Jonsson generously shared their pygmy- seum Natural History 36:1-729. owl recordings. Nathan Rice provided relevant mo- CHAPMAN, F M. 1926. The distribution of bird-life in lecular data regarding this group. Carla Cicero and Ecuador. Bulletin of the American Museum Nat- Ned Johnson kindly furnished copies of Alden Mill- ural History 55:1-784. er's field notes at the MVZ. Neal Woodman measured COLLAR, N. J., L. P. GONZAGA, N. KRABBE, A. MAD- a specimen of costaricanum in the Costa Rican mu- RONO NIETO, L. G. NARANJO, AND T. A. PARKER seum. We thank Eric Pasquet of Museum National III. 1992. Threatened birds of the Americas. In- d'Historie Naturelle, Paris, for determination of the ternational Council for Bird Preservation, Cam- status of the jardinii holotype. ANSP inventory work bridge, United Kingdom. was conducted in conjunction with the Museo Ecu- CRACRAFT, J. 1983. Species concepts and speciation atoriano de Ciencias Naturales, Quito. The Minister- analysis. Current Ornithology 1:159-187. io de Agricultura, Quito, provided authorization for DOBSON, C. H., AND A. H. GENTRY. 1991. Biological work in Ecuador. Stiles's field work at Alto Pisones extinction in western Ecuador. Annals of the was conducted under the auspices of a contract to the Missouri Botantical Garden 78:273-295. Insituto de Ciencias Naturales (ICN) for biological GOMEZ, N. 1990. Atlas del Ecuador. Geografia y inventories in Risaralda from the Corporaci6n Au- Economia. Ediguias C. Limidad, Quito, Ecuador. t6noma Regional de Risaralda; Gonzalo Andrade, HARDY, J. W., B. B. COFFEY, JR., AND G. B. REYNARD. Diana Mora, Milton Munioz, and Carlos Alberto Ve- 1990. Voices of the New World owls (Strigifor- negas provided assistance and companionship in the mes: Tytonidae, Strigidae). ARA Records, field. Ned Johnson, Claus Konig, Jeff Marks, Town Gainesville, Florida. Peterson, Van Remsen, Tom Schulenberg, and Na- HEIDRICH, P., C. KONIG, AND M. WINK. 1995. Biak- than Rice made valuable comments on the manu- ustik, Taxonomie und molekulare Systematik script. We thank the following people and institu- amerikanischer Sperlingskauze (Strigidae: Glau- tions (listed alphabetically by institution) for access cidium spp.). Stuttgarter Beitrage zur Naturkun- de A to specimen material under their care: Robert Rid- Serie (Biologie) 534:1-47. HOWELL, S. N. G., AND S. WEBB. gely and David Agro (Academy of Natural Sciences 1995. A guide to the birds of Mexico and northern Central [ANSPI, Philadelphia), George Barrowclough and America. Oxford University Press, New York. Paul Sweet (American Museum of Natural History HOWELL, S. N. G., AND M. B. ROBBINS. 1995. Species [AMNH], New York), Robert Kennedy (Cincinnati limits in the Least Pygmy-Owl (Glaucidium min- Museum Center [CMC], Cincinnati), Greg Budney utissimum) complex. Wilson Bulletin 107:7-25. (Library of Natural Sounds, Cornell University, Ith- INSTITUTO GEOGRAFiCO AGUSTiN CODAZZI. 1989. At- aca, New York), David Willard (Field Museum of las basico de Colombia, 6th ed. Instituto Geo- Natural History Chicago), David Stead- [FMNH], grafico Agustin Codazzi, Bogota. man and Tom Webber (Florida Museum of Natural JOHNSON, N. K. 1980. Character variation and evo- History [UFI, University of Florida, Gainesville), Van lution of sibling species in the Empidonaxdifficil- Remsen and Steven Cardiff (Louisiana State Univer- is-flavescens complex (Aves: Tyrannidae). Uni- sity [LSUMNSJ, Baton Rouge), Raymond Paytner, Jr. versity of California Publications in Zoology (Museum of Comparative Zoology [MCZ], Harvard), 112:1-151. Ned Johnson and Carla Cicero (Museum of Verte- JOHNSON, N. K., AND J. A. MARTEN. 1988. Evolution- brate Zoology [MVZ], University of California), ary genetics of flycatchers. II. Differentiation in Gary Graves (National Museum of Natural History, the Empidonax difficilis complex. Auk 105:177- Smithsonian Institution), Bernal Rodrfquez-H. 191. (Universidad de Costa Rica [UCRI, San Jose), and KELSO,L. 1937. A Costa Rican race of Jardine's Pyg- Sam Sumida (Western Foundation of Vertebrate Zo- my Owl. Auk 54:304. ology [WFVZ], Camarillo, California). KONIG, C. 1991. Zur Taxonomie und Okologie der April 1999] New Pygmy-Owl from the Andes 315

Sperlingskauze (Glaucidium spp.) des Andenrau- na of an Upper Tropical cloud forest in south- mes. Okologie der Vogel 13:15-76. western Ecuador. Proceedings of the Academy LANYON,W. E., AND S. M. LANYON. 1986. Generic of Natural Sciences of Philadelphia 142:59-71. status of Euler's Flycatcher: A morphological RYDER,0. A. 1986. Species conservationand system- and biochemical study. Auk 103:341-350. atics: The dilemma of subspecies. Trends in MALFAIT,J. A., AND M. G. DINKLEMAN. 1972. Cir- Ecology and Evolution 1:9-10. cum-Caribbean tectonic and igneous activity SALAMAN, P. G. W., AND F.G. STILES. 1996. A distinc- and the evolution of the Caribbean plate. Geol- tive new species of vireo (Passeriformes:Vireon- ogy Society of America Bulletin 83:251-272. idae) from the WesternAndes of Colombia.Ibis MAYR, E. 1963. Animal species and evolution. Har- 138:610-619. vard University Press, Cambridge, Massachu- SIBLEY,C. G., AND B. L. MONROE,JR. 1990.Distri- setts. bution and taxonomy of birds of the world. Yale MAYR, E. 1969. Principles of systematic zoology. Mc- University Press, New Haven, Connecticut. Graw-Hill, New York. SLUD, P. 1964. The birds of Costa Rica: Distribution MILLER, A. H. 1963. Seasonal activity and ecology of and ecology. Bulletin of the American Museum Natural the avifauna of an American equatorial cloud History 128:1-430. SMITHE, F.B. 1975. Naturalist'scolor forest. University California Publications in Zo- guide. American Museum Natural History, New York. ology 66:1-78. SMITHE, F. B. 1981. Naturalist'scolor guide, part III. MORITZ, C. 1994. Applications of mitochondrial American Museum Natural History, New York. DNA analysis in conservation: A critical review. STATTERSFIELD,A. J.,M. J. CROSBY,A. J. LONG, AND Molecular Ecology 3:401-411. D. C. WEGE.1998. Endemic bird areas of the PETERS,J. L. 1940. Check-list of birds of the world, world: Priorities for biodiversity conservation. vol. 4. Museum of Comparative Zoology, Cam- BirdLifeInternational Conservation Series No. 7, bridge, Massachusetts. Cambridge,United Kingdom. PHILLIPS, A., MARSHALL, J. AND G. MONSON. 1964. STEIN, R. C. 1958. The behavior, ecological, and mor- The birds of Arizona. University of Arizona, phological characteristicsof two populations of Press, Tucson. the Alder Flycatcher, Empidonax traillii (Audu- PINDELL, J., AND J. F. DEWEY. 1982. Permo-Triassic re- bon). New YorkState Museum Science Service construction of Western Pangea and the evolu- Bulletin No. 371. tion of the Gulf of Mexico/Caribbean region. STILES,F. G., AND A. F.SKUTCH. 1989. A guide to the Tectonics 1:179-211. birds of Costa Rica. Cornell University Press, RIDGELY, R. S., AND J. A. GWYNNE. 1989. A guide to Ithaca, New York. the birds of Panama, with Costa Rica, Nicara- VIELLIARD, J. 1989. Uma nova especie de Glaucidium gua, and Honduras, 2nd ed. Princeton Univer- (Aves, /Strigidae) da Amazonia. Revista Brasi- sity Press, Princeton, New Jersey. leira Zoologia 6:685-693. ROBBINS, M. B., AND S. N. G. HOWELL. 1995. A new ZINK, R. M., AND M. C. MCKITRICK.1995. The debate species of pygmy-owl (Strigidae: Glaucidium) over species concepts and its implicationsfor or- from the eastern Andes. Wilson Bulletin 107:1-6. nithology. Auk 112:701-719. ROBBINS, M.B., AND R. S. RIDGELY. 1990. The avifau- Associate Editor: J. S. Marks