Cuervo_etal2005

8/3/2019 Cuervo_etal2005

1/20

The AukVol. 122 No. 2 April 2005

PUBLISHED BY


2/20

445445

SCYTALOPUS STILESI , A NEW SPECIES OF TAPACULO(RHINOCRYPTIDAE) FROM THE CORDILLERA CENTRAL

OF COLOMBIAA M. C , 1,5C D C , 2 N K , 3L M R 2,4

1Instituto de Biologa, Universidad de Antioquia, A.A. 1226, Medelln, Colombia;2Department of Biology and International Center for Tropical Ecology, University of Missouri-St. Louis,

8001 Natural Bridge Road, St. Louis, Missouri 63121, USA;3Zoological Museum, University of Copenhagen, Universitetsparken 15, DK-2100 Copenhagen, Denmark; and

4Departamento de Ecologa y Territorio, Facultad de Estudios Ambientales y Rurales, Ponticia Universidad Javeriana,Transversal 4 No. 42-00 Piso 8, Bogot, Colombia

A .We describe Scytalopus stilesi , an overlooked species of tapaculo

endemic to Colombia, on the basis of a series of eight specimens taken in 2002 andcomparative analyses of its vocalizations, mitochondrial DNA sequences, and dis-tribution. The new species ranges in the northern half of the Cordillera Central ofthe Colombian Andes in the Departments of Antioquia, Caldas, and Risaralda, incloud forests between 1,420 and 2,130 m above sea level. The song, calls, and femalesong of the new species differ distinctly from those of all other knownScytalopus taxa. Phylogenetic analyses based on sequences of the cytochrome-b gene stronglysuggest affi nities withS. robbinsi of southwestern Ecuador and with two as-yet-undescribed tapaculos from the Colombian Andes.Scytalopus stilesi coexists locallywith, though it is ecologically segregated from,S. atratus, S. latrans, and S. spillmanni.The mid-elevation premontane wet forests to which the new species is restrictedhave been subject to severe deforestation and fragmentation. The species is, how-ever, relatively common in continuous mature-forest remnants, large primary-forestfragments, riparian forests, and tall secondary-forest patches. We employed a geo-graphic information system (GIS) approach to model the potential distribution ofthe new species and assess its conservation status under the International Unionfor the Conservation of Nature (IUCN) criteria.Scytalopus stilesi does not qualify asthreatened according to those criteria, but it should be regarded as near threatened.The new species coexists with numerous threatened bird species that are in need ofmore effective conservation.Received 25 April 2004, accepted 18 November 2004.

Key words: Andes, cytochromeb , new species,Scytalopus , suboscines, systematics.

Scytalopus stilesi , una Nueva Especie de Tapaculo (Rhinocryptidae) de la CordilleraCentral de Colombia

R .Describimos a Scytalopus stilesi , una nueva especie de tapaculoendmica de Colombia, con base en una serie de ocho especmenes colectados en2002 y en anlisis comparativos de sus vocalizaciones, secuencias de ADN mitocon-drial y distribucin. La nueva especie se distribuye en la mitad norte de la CordilleraCentral de los Andes colombianos en los departamentos de Antioquia, Caldas y

The Auk 122(2):445463, 2005 The American Ornithologists Union, 2005.Printed in USA.

5Present address: Department of Biology, University of Puerto Rico, P.O. Box 23360, San Juan, Puerto Rico00931. E-mail: [email protected]


3/20

C .446 [Auk, Vol. 122

S the suboscine genusScytalopus (Rhinocryptidae) have been exceed-ingly diffi cult to resolve, leading ornithologiststo consider its taxonomy unusually diffi cult(Zimmer 1939) and the most complicated of allNeotropical bird genera (Ridgely and Tudor1994). The diffi culty results from subtle varia-tion in plumage and morphology among spe-cies, coupled with substantial age-related andindividual variation within species; the birdsskulking behavior in low and dense understory,which makes them diffi cult to observe and col-lect; lack of specimens with reliable data frommany areas; foxing of old skin specimens,including types; and insuffi cient study of theirvocalizations (Zimmer 1939; Whitney 1994;Krabbe and Schulenberg 1997, 2003).

Increased eldwork in South America overrecent years has conrmed that plumage is nota good indicator of species limits inScytalopus,casting doubt on the reliability of traditional,museum-based taxonomic treatments of thegroup (e.g. Zimmer 1939) and spurring devel-opment of new studies that draw on other typesof data to delimit species (Whitney 1994, Krabbeand Schulenberg 1997, Coopmans et al. 2001).As in other tracheophone suboscines, vocaliza-tions inScytalopus are believed to be innate andhighly stereotyped within populations (Fjeldsand Krabbe 1990; Vielliard 1990; Krabbe andSchulenberg 1997, 2003) and appear to charac-

terize genetically distinct units (Arctander andFjelds 1994). Moreover,Scytalopus distribu-tions are characterized by sharp replacements

of species along altitudinal gradients, andsympatric taxa are o en segregated by habitat(Whitney 1994, Krabbe and Schulenberg 2003).Thus, it is now evident that resolving the clas-sication ofScytalopus requires comprehensiveassessments of populations in terms of vocaliza-tions, elevational distribution, and habitat, ide-ally in combination with genetic analyses.

Extensive new data on vocalizations anddistribution have revealed that the number ofScytalopus species occurring in the northernAndes was grossly underestimated by tradi-tional taxonomy and have led to the descrip-tion of three new species from Ecuador andthe elevation of several previously namedforms to species rank (Krabbe and Schulenberg1997). Although Krabbe and Schulenberg (1997)included some taxa whose ranges extend intoColombia, our understanding ofScytalopus taxonomy and distribution in this countryremains especially complicated (e.g. Krabbeand Schulenberg 2003). As with Ecuadoriantaxa, the number of species occurring inColombia has clearly been underestimated, yetno detailed taxonomic studies of ColombianScytalopus have been undertaken. Currently,some new species remain undescribed becauseof insuffi cient comparative material (e.g.Krabbe and Schulenberg 1997, Cuervo et al.2003), and others may have been overlooked inspite of ornithological eldwork. By integrating

data on vocalizations, distribution, ecology, andmitochondrial DNA variation, we have startedto re-address the species-level taxonomy of

Risaralda, en bosques de niebla entre 1,420 y 2,130 m de elevacin. El canto, losreclamos y el canto de la hembra de la nueva especie dieren distintivamente de losde todos los taxa conocidos del gneroScytalopus. Los anlisis logenticos basadosen secuencias del gen mitocondrial citocromob sugieren fuertemente anidades conS. robbinsi del suroccidente de Ecuador y con dos especies an no descritas de los

Andes colombianos.Scytalopus stilesi coexiste localmente conS. atratus, S. latrans y S. spillmanni , aunque se segrega ecolgicamente de ellos por uso de hbitat. Los bosques premontanos hmedos de elevaciones medias a los cuales est restringidala nueva especie han sido objeto de una severa deforestacin y fragmentacin. Sinembargo, la especie es relativamente comn en remanentes de bosques maduroscontinuos, fragmentos de bosque primario, bosques riparios y parches de bosquesecundario avanzado. Empleamos un anlisis basado en sistemas de informacingeogrca para modelar la distribucin potencial de la nueva especie y evaluarsu estado de conservacin bajo los criterios de la IUCN.S. stilesi no calica comoamenazada de acuerdo con estos criterios, pero debe ser considerada como casiamenazada. La nueva especie coexiste con numerosas especies de aves amenazadasque requieren un nivel de conservacin ms efectivo.


4/20

New Tapaculo from ColombiaApril 2005] 447

Colombian Scytalopus. The present study is therst in a series that, we expect, will ultimatelyresult in a substantially be er understandingof the diversity and distribution pa erns of thisgenus in Colombia.

Over the past ten years, A.M.C., C.D.C.,L.M.R., and others (see Acknowledgments)have repeatedly observed and tape-recordedvocalizations of a Scytalopus tapaculo in mid-elevation humid forests of the Cordillera Centralof the Andes in the Departments of Antioquia,Caldas, and Risaralda (Colombia). Upon exam-ining its vocalizations, N.K. suspected that thistapaculo was a new species. Recent eldwork by A.M.C. in northern Antioquia resulted in col-lection of eight specimens with associated tissuesamples, in addition to numerous recordings ofcalls and songs of this taxon. Vocal, morphologi-cal, ecological, and genetic comparative analy-ses of the available material conrm that thisCentral Andean Scytalopus indeed represents aheretofore undescribed species, which we pro-pose to name:

Scytalopus stilesi , sp. nov.Stiless TapaculoTapaculo de Stiles

Holotype.Instituto de Ciencias Naturales(ICN) at Universidad Nacional de Colombiano. 34569; adult male (testes: 8.6 5.8 mm le ,7.5 5.0 mm right) from Finca Canales, VeredaCajamarca, Municipality of Amal, Department ofAntioquia, Colombia (64925.2N, 75 0537.8W,~1,845 m above sea level). Collected by A.M.C.(eld number AMC 104) on 5 January 2002.Tissue samples deposited in the tissue bank ofInstituto Alexander von Humboldt, Colombia(IAvH-BT, no. 2093); recordings of voice to be

archived in the Macaulay Library of NaturalSounds (MLNS), Cornell Laboratory ofOrnithology, Ithaca, New York. DNA sequenceof a fragment of the cytochrome- b mitochon-drial gene deposited in GenBank (accession no.AY755652).

Diagnosis.A typical rhinocryptid (seecover plate and Fig. 1), assignable to the genusScytalopus by the combination of a fairly sharpand laterally compressed bill with high culmen;nares covered by a conspicuous operculum; short

and somewhat erect feathers in lores; so , short,very concave, and very rounded wings, with 10primaries; relatively short and graduated tail

with the outermost rectrices shorter than the rest;tail much shorter than wing; large and strongfeet, with distinct taxaspidean tarsal scutellation;hind claw strongly curved and shorter than thedigit; overall dark and so plumage, with li le

sexual dimorphism and substantial individualvariation in plumage related to age (Ridgway1911, Krabbe and Schulenberg 2003).

The new species is most readily diagnosed by its vocalizations, the song being a long seriesof short churrs, the three-note call and femaleadvertising song differing distinctly from voicesof all other known Scytalopus taxa (see below).Scytalopus stilesi can also be reliably diagnosedgenetically, on the basis of DNA sequences of thecytochrome- b mitochondrial gene (see below).

Morphologically, the new species may not be reliably distinguishable from most otherScytalopus taxa. However, S. stilesi is likelyto co-occur only with S. atratus , S. latrans , S.spillmanni , and possibly S. vicinior (see below).Scytalopus stilesi can be distinguished from S.latrans on the basis of its larger size, less blackishappearance, and brown wash and barring onthe plumage; from S. atratus on the basis of itssmaller body size and lack of white crown spot;and from S. spillmanni by its generally smallersize and longer bill (see Table 1 and Krabbeand Schulenberg 1997, 2003). Scytalopus stilesiis less dark than some specimens of S. vicinior (e.g. ICN 31207 and 31208 from Risaralda), but not than others (e.g. ICN 34840 from Valledel Cauca), and has a shorter tail (apparentlyno overlap; Krabbe and Schulenberg 1997).

F . 1. Adult Scytalopus stilesi at Otn-Quimbaya Fauna and Flora Sanctuary, Risaralda.(Photo courtesy of Gustavo Londoo.)


5/20

C .448 [Auk, Vol. 122

Juveniles ofS. stilesi are probably not diagnos-able morphologically from those of similar spe-cies (for comparative morphometrics see Table 1and Krabbe and Schulenberg 1997).

Description of holotype.A medium-sizedtapaculo with forecrown, lores, ear coverts, andupper partsincluding crown, mantle, scapu-lars, and backuniformly Blackish NeutralGray 82 (capitalized color names and numbersfollow Smithe 1975). Underparts paler in thechin and throat, between Medium Neutral Gray84 and Light Neutral Gray 85, becoming darkertoward chest, sides, and upper belly (betweenBlackish Neutral Gray and Dark Neutral Gray83). Lower belly, anks, and thighs bright brown (closest to Mikado Brown 121C), withdark gray (Blackish Neutral Gray) barring.Rump somewhat darker brown (betweenMikado Brown and Brussels Brown 121B) and

much less barred than the underparts. Tail andwings (including wing coverts) Black NeutralGray. Irides dark brown, bill black with base ofthe mandible horn, feet and toes grayish brown,and claws pale horn. Stomach contents: smallinsect fragments; relatively abundant subcuta-neous fat. Body mass 24.5 g; bill (from fore edgeof the operculum) 7.5 mm; bill height 4.0 mm; bill width 3.5 mm; tarsus 23.0 mm; tail 43.3 mm;wing (a ened) 64.5 mm; 11 rectrices.

Designation of paratypes.A total of seven

specimens from Antioquia, prepared as con-ventional study skins deposited at the ICN,with tissue samples deposited at the IAvH-BT,

as follows: adult male: ICN 34609, IAvH-BT2227; adult female: ICN 34505, IAvH-BT 2231;adult female: ICN 34610, IAvH-BT 2119; juve-nile male: ICN 34584, IAvH-BT 2229; subadultfemale: ICN 34420, IAvH-BT 2230; adult female:ICN 34615, IAvH-BT 2127; immature female:ICN 34512, IAvH-BT 2228.

Etymology.We take great pleasure innaming this species a er F. Gary Stiles, inrecognition of his outstanding career and themany substantial contributions he has madeto Neotropical ornithology over the past threedecades. Gary has admirably combined the tra-ditions of museum and eld ornithology while based in Neotropical countries (Costa Rica andColombia), where he has forged his seminalstudies on the ecology and taxonomy of birds,most notably hummingbirds. Since his arrival inColombia in 1988, he has deeply inuenced the

development of ornithology as a science in thecountry through his research and teaching, andthrough his pivotal role in the establishmentand growth of the Asociacin Bogotana deOrnitologa and, more recently, the AsociacinColombiana de Ornitologa.

R

Variation within the type series .Considerablevariation exists in the type series, largely

reecting variation related to age. Adult maleICN 34609 resembles the holotype closely, butis generally slightly paler, with subtle brownish

T 1. Body mass (g) and selected morphological measurements (mm) ofS. stilesi specimens. Juvenile male ICN 34584 not included. Measurements are omi ed in cases in which they couldnot be taken accurately.

ICN Body Wing Bill Bill Billnumber mass (at) Tail Tarsus lengtha height b width b

Males

34569 24.5 64.5 43.3 23.0 7.5 4.0 3.534609 21.5 63.0 42.0 22.9 8.1 4.1 3.6

x SD 23.0 2.1 63.75 1.1 42.65 0.9 22.95 0.1 7.8 0.4 4.05 0.1 3.55 0.Females

34420 20.0 58.0 40.0 21.8 7.0 3.8 3.634505 20.5 60.0 40.6 22.0 7.0 3.9 3.434512 21.5 61.0 41.1 22.9 7.8 3.8 3.534610 22.0 58.0 39.8 21.5 7.2 3.534615 22.0 59.0 42.3 21.6

x SD 21.2 0.9 59.2 1.3 40.8 1.0 22.0 0.6 7.25 0.4 3.8 0.1 3.5 0.1a From fore edge of operculum to tip. bAt fore edge of operculum.


6/20


tinges in the crown and wing (on tertials andupper coverts); also, the rump is more con-spicuously barred and darker brown (closestto Brussels Brown), and the brown extendshigher into the lower back than in the type.

Juvenile male ICN 34584 is similar to juvenilesof otherScytalopus species: completely lightly barred brown and gray; head, back, and wings(Brussels Brown) darker than throat, chest, andlower belly (closest to Mikado Brown). Thatspecimen had presumably edged recently, because it had no rectrices and had only a fewgrowing wing feathers, of less than one-quarterof their expected full size. Females tend to bepaler ventrally, closer to Medium Neutral Graythan to the Dark Neutral Gray of males. Mostfemale specimens have brown secondariesand wing coverts; the la er o en show brownmargins and a dark subterminal half moon.Immature female ICN 34420 (see cover) showspaler grays and redder browns overall: throatand chest between Light Neutral Gray (85) andPale Neutral Gray (86); feathers on sides of head,lores, and malars tinged brown; the brown inlower abdomen is paler than in other specimens(closest to Tawny 223D). Most of that specimensplumage shows some barring, which is espe-cially pronounced dorsally, where it is ne inthe head and becomes thicker in the back; thespecimen has an overall scaled appearance. Likemale ICN 34609, all the females show tinges of brown to varying degrees in the back and head,mostly in the crown. The extent of barring in therump and upper tail coverts is variable: in somespecimens (e.g. female ICN 34505), it is verypronounced; whereas in others (e.g. female ICN34512), it is faint. Although sample sizes are lim-ited, males and females do not appear to differin morphometrics or in body mass, yet malestend to have longer wings (Table 1). Number ofrectrices seems to vary among specimens (range1012). In general, no molt was observed on thetype series, but female ICN 34615 had a singleremix and some wing coverts growing.

Vocalizations.We analyzed natural (unpro-voked) songs recorded from 24 different indi-viduals from various localities encompassingmost of the range ofS. stilesi. Bouts of songconsisted of up to 35 or more monotonouschurrs ,the rst o en longest, ranging from 3 to 5 s and

occasionally up to 15 s long, the following one ortwo transitional, all the rest constant, 12 s long,typically with pauses 0.51.0 s between them

(Fig. 2). Eachchurr was composed of a singlenote repeated regularly at a pace of 2330 s1(Fig.3A). As in most congeners, the rst overtone ina song note ofS. stilesi was normally the loud-est. The fundamental note was usually weaker,

and the second overtone barely audible; but inone recording, the fundamental was loudest.The rst overtone ranged in frequency between1.8 and 2.2 kHz. Within eachchurr , the volumeincreased gradually over the rst 0.30.5 s andthen remained constant, to fade and fall slightlyon the last one or two notes, which were alsoslightly lower-pitched than the rest. No geo-graphic variation in the song or calls of the newspecies was observed.

In comparison with the song of the closelyrelated S. robbinsi (see below), the song ofS.stilesi is considerably faster and lower-pitchedand composed ofchurrs rather than being a con-tinuous series, and the individual song notesare single and mainly upstrokes, instead of theusually double strokes and mainly downstrokesof S. robbinsi (Fig. 3B). Song ofS. stilesi also dif-fers distinctly from song ofS. atratus confusus (Fig. 3C),S. spillmanni (Fig. 3D), andS. latrans (Fig. 3E), three taxa whose ranges locally over-lap with or border the range ofS. stilesi.

Calls ofS. stilesi were recorded from at leasteight different individuals. Calls were repeatedat irregular intervals and were a two-note or,more commonly, three-notecu-wi? , cu-cui-wi? ,or cu-wi-wi? , rst note at 2.12.2 kHz, last notedistinctly rising at 2.32.5 kHz, and middle notevariably like the rst or the last note or interme-diate (Fig. 4AC). The calls were lower-pitchedthan the single-note, slowly rising calls ofS. rob-binsi (Fig. 4DF) and also unlike the calls of anyother knownScytalopus.

The single female advertising song recordedwas given in duet with male song (Fig. 5A). Itwas ~13 s long; a series of ~10 sharp downstroke

F . 2. Song bout of 13churrs by S. stilesi.Aranzazu, Caldas, Colombia, 27 August 2003(P. Caycedo).


7/20

C .450 [Auk, Vol. 122

F . 3. Songs of males of variousScytalopus spp. (A) Singlechurr by S. stilesi , Otn-Quimbaya,Risaralda, Colombia, 2 November 1996 (L.M.R.); (B) part of song byS. robbinsi , Buenaventura, El Oro,Ecuador, 15 April 1991 (N.K.); (C) one song phrase byS. atratus confusus , Mampuestos, Antioquia,

Colombia, 30 March 2002 (A.M.C.); (D) part of song byS. spillmanni , Ucumar, Risaralda, Colombia,17 July 2001 (C.D.C.); (E) part of song byS. latrans , Alto El Gallinazo, Antioquia, Colombia, 12 June1994 (N.K.).


8/20


notes; notes rst single, later double andtriple; gradually descending in pitch from 3.0to 2.5 kHz. Notes were given at intervals of1.21.4 s, with the rst interval a li le longer(1.7 s). The female advertising song ofS. robbinsi is higher-pitched and composed only of singlenotes of different quality (Fig. 5B). Femaleadvertising songs ofS. vicinior , S. spillmanni ,S. parkeri, and S. canus opacus are both higher-pitched and faster than that ofS. stilesi (Fig.5CF).

Distribution .Scytalopus stilesi is known from

21 localities on both slopes of the northern halfof the Cordillera Central of Colombia (Fig. 6and Appendix). On the western slope, records

range from southern Risaralda north to west-ern Antioquia. On the eastern slope, recordsare from the northernmost part of Antioquia(Fig. 6 and Appendix). The southernmostlocality known is Otn-Quimbaya Fauna andFlora Sanctuary in the Municipality of Pereira,Risaralda; the northernmost is Santa Gertrudisin the Municipality of Anor, Antioquia (Fig. 6and Appendix); those two localities are sepa-rated by a linear distance of ~280 km. Althoughwe have no records ofS. stilesi from the easternslope of the Cordillera Central south of northern

Antioquia, we believe that the species is likelydistributed continuously in wet premontaneforest along both slopes of the northern half of

F . 4. Calls of two closely relatedScytalopus tapaculos. Scytalopus stilesi: (A) Carolina, Antioquia,Colombia, April 1985 (T. Cuadros); (B) Otn-Quimbaya, Risaralda, Colombia, 22 April 1999(C.D.C.); (C) Cajamarca, Amalfi, Antioquia, Colombia, 9 January 2002 (A.M.C.).Scytalopus robbinsi:(DF) recorded at Buenaventura, El Oro, Ecuador by N.K.: (D) 16 April 1991; (E) 16 November 199(F) 26 September 1990.


9/20

C .452 [Auk, Vol. 122

that mountain range. Before extensive deforesta-tion took place, the wet premontane forests thatS. stilesi inhabits extended continuously along both slopes of the cordillera, from its northernextreme to dryer forest enclaves farther south.On the western slope, those wet premontane for-ests range from central Antioquia south throughCaldas, Risaralda, and Quindo just into Valledel Cauca (Fig. 7); on the eastern slope, theyextend from the northernmost part of the cor-dillera in Antioquia south through Caldas andTolima to its border with Huila (Fig. 7). Thus,the distribution ofS. stilesi may correspond tothe distribution of wet premontane forests inthe northern sector of the Cordillera Central,its limits imposed by marked habitat disconti-nuities. In addition, considering that parapatric

distributions along the Andes are common inthe genus, the range ofS. stilesi may abut thatof anotherScytalopus at its southern extreme. Infact, an unnamed species occurs at similar eleva-tions farther south in the Cordillera Central atthe head of the Magdalena Valley in Huila andCauca (Krabbe and Schulenberg 1997).

We have recordedS. stilesi in wet forests from1,420 to 2,130 m in elevation. Because we have notsampled forests at lower elevations thoroughly,we have not documented the lower elevationallimit for the species, but we expect that it is notfound below 1,200 m. Below that elevation, thepremontane wet forest shi s into a transitionallife-zone with lowland rainforests, where noScytalopus species occur. The upper elevationallimit is not well established either, because we

F . 5. Female advertising songs. (A)Scytalopus stilesi (in duet with male), Otn-Quimbaya,Risaralda, Colombia, 2 November 1996 (L.M.R.); (B)Scytalopus robbinsi , Buenaventura, El Oro,Ecuador, 16 November 1991 (N.K.); (C)Scytalopus vicinior , Mindo, Pichincha, Ecuador, 27 April 1991(N.K.); (D)Scytalopus spillmanni , Ucumar, Risaralda, Colombia, 17 July 2001 (C.D.C.); (E)Scytalopus

parkeri (in duet with male), Gualaceo-Limn road, Morona-Santiago, Ecuador, 18 June 1984 (N.K.); (F)Scytalopus canus opacus , Papallacta Pass, Pichincha-Napo border, Ecuador, 14 October 1983 (N.K.).


10/20


have not surveyed locations between 2,200 and2,400 m extensively. However, within the Otnriver watershed in Risaralda, we have noted thatS. stilesi is absent from forests at 2,430 m at ParqueRegional Ucumar (La Pastora), but the species iscommon up to 2,110 m in the Otn-QuimbayaFauna and Flora Sanctuary, just a few kilometersdown-slope. At least in this region,S. stilesi isreplaced altitudinally byS. spillmanni , which iscommon there at elevations of 2,040 to >2,800 m.

S B

Systematic affi nities.We assessed the

systematic position and genetic distinctive-ness ofS. stilesi as part of an ongoing studyon the phylogeny of the genusScytalopus

based on mitochondrial DNA sequence data.Phylogenetic trees and detailed descriptions ofmethods and analyses will be published else-

where (C. D. Cadena et al. unpubl. data); here,we summarize preliminary results relevant tothe affi nities ofS. stilesi. We focus on analyses based on 284315 base pairs of sequence fromthe cytochrome-b gene generated for fourS.stilesi individuals from different localities andfor multiple additionalScytalopus and outgrouptaxa (Table 2). Sequences taken from the litera-ture vary in length because they were generatedfor different studies that did not target exactlythe same DNA regions. Although this data

set consists of an admi edly small number ofcharacters, it is taxonomically comprehensive,including 21 of the 37 species ofScytalopus

F . 6. Map of western Colombia showing major cities and principal mountain ranges in the coun-try, which are depicted by a contour line at 1,000 m elevation. Black dots along Cordillera Centralrepresent the localities ofScytalopus stilesi, and the star indicates the type locality at Cajamarca,Amalfi (Deptartment of Antioquia).


11/20

C .454 [Auk, Vol. 122

currently recognized (Krabbe and Schulenberg

2003), in addition to several vocally or morpho-logically distinct populations, including a fewas-yet-undescribed species-level taxa. With the

exception ofS. sanctaemartae and S. latebricola , both endemic to the Sierra Nevada de SantaMarta, all the species ofScytalopus known tooccur in Colombia and Ecuador are representedin the data set.

A variety of parsimony- and model-based(i.e. maximum likelihood, Bayesian inference)phylogenetic analyses consistently recovereda monophyletic group formed byS. stilesi , S.robbinsi (a species endemic to the Pacic slopeof the Andes of southwestern Ecuador), andtwo undescribedScytalopus species from theColombian Andes: one from Alto de Pisones,Pacic slope of the Cordillera Occidental,Risaralda (see Cuervo et al. 2003); the otherfrom Finca Merenberg, head of the MagdalenaValley, Huila (see Krabbe and Schulenberg1997). The clade formed by those four specieswas well supported in all analyses, with par-simony bootstrap values >90% and a Bayesianposterior probability of 1.0. The four speciesare very distinct genetically from all otherScytalopus , which indicates that they have beenisolated for a substantial time; the minimumsequence divergence (uncorrectedP distance)observed with otherScytalopus was ~9.5%.Because of the short length of the sequencesavailable, relationships among the four taxacould not be established clearly, and the spe-cic position ofS. stilesi was unresolved orlacked support. In any case, it is noteworthythat all the members of this highly distinctiveclade were recently discovered (S. robbinsi wasdescribed in 1997), highlighting the general lackof understanding of the diversity ofScytalopusin the northern Andes, especially in Colombia.

Affi nities ofS. stilesi to species that occurin different regions and not to those co-occur-ring with it on the same mountain slopes (i.e.S. atratus , S. latrans , andS. spillmanni) supportthe scenario of allopatric speciation proposedfor the genus by Arctander and Fjelds (1994),which seems to be predominant in Andean birds (Garca-Moreno and Fjelds 2000). Theclose relationship ofS. stilesi to species occur-ring on the Andean Pacic slope suggests a bio-geographic connection between the Choc areaand the Cordillera Central of Colombia. Thatconnection is also suggested by several specieswhose distributions are concentrated in the

Pacic slope of the Andes and have been foundto occur in the northern Cordillera Central(A. M. Cuervo et al. unpubl. data). This may be

F . 7. Map of the distribution ofScytalopusstilesi in the Cordillera Central modeled by GISanalyses (see text). Forest clearance is depictedin light gray, and remnants of premontane wetforests in dark gray. Most of the suitable habi-tat for the species is located in Antioquia, andremarkably along the eastern slope in Caldasand southern Tolima, where confirmed recordsare still lacking.


12/20


T 2. Specimens of Scytalopus and rhinocryptid outgroups used for molecular analyses,with GenBank accession numbers of their cytochrome-b sequences. Unless otherwise noted,specimens are referable to the corresponding nominate subspecies. Nomenclature followsKrabbe and Schulenberg (2003).

GenBankTaxon Voucher(s) number LocalityS. aff.bolivianus FMNH 390674 AY755631a Per, Junn, Cordillera Vilcabamba, Ro

PoyeniS. atratus ZMUC 80145 U06168 b Ecuador, Napo, Guacamayos MountainsS. a. confusus ICN 34387 AY755632a Colombia, Antioquia, Anor, MampuestosS. femoralis LSUMZ 106110 U06159 b Per, Pasco, Santa Cruz, 9 km south-

southeast of OxapampaS. micropterus 1 MECN 5798 U06160 b Ecuador, Morona-Santiago, Cutuc,

YapityaS. micropterus 2 IAvH 12377 AY755633a Colombia, Huila, Cueva de los

GucharosS. vicinior 1 ZMUC 125176 AY755634a Ecuador, Pichincha, Obelisque nearMindo

S. vicinior 2 ICN 34840 AY755635a Colombia, Valle del Cauca, La CumbreS. robbinsi ZMUC 80100 U06176 b Ecuador, El Oro, 9 km west of PiasS. chocoensis ANSP 180149 U06158 b Ecuador, Esmeraldas, El PlacerS. argentifrons LSUMZ 135542 AY755636d Costa Rica, km 120 San JosSan Isidro

roadS. spillmanni 1 ZMUC 80012 U06171 b Ecuador, Sucumbos, 10 km east of Santa

BrbaraS. spillmanni 2 ZMUC 80014 U06173 b Ecuador, Imbabura, 15 km southeast of

Apuela, Las DeliciasS. parkeri 1 ZMUC 80054 U06178 b Ecuador, Morona-Santiago, Zapote-Najda

mountainsS. parkeri 2 ZMUC 80055 U06169 b Ecuador, Morona-Santiago, Zapote-Najda

mountainsS. parvirostris 1 LSUMZ 128572 U06161 b Per, Pasco, Playa Pampa, 8 km northwest

of CushiS. parvirostris 2 CBF 2368 U35080c Bolivia, La Paz, Nor YungasS. griseicollis1 IAvH 12586 AY755637a Colombia, Boyac, Villa de LeyvaS. griseicollis2 IAvH 12701 AY755638a Colombia, Cundinamarca, GuascaS. canus opacus ZMUC 125688 AY755639a Ecuador, Loja, JimburaS. affi nis ZMUC 125154 AY755640a Per, Ancash, Quebrada PucuvadoS. schulenbergi 1 CBF 2640 U35081c Bolivia, La Paz, Franz Tamayo, 5 km east

of PelechucoS. schulenbergi 2 CBF 2636 U35082c Bolivia, La Paz, Nor Yungas, 4 km west-

northwest of ChuspipataS. simonsi 1 ZMUC 80031 U35083c Bolivia, Cochabamba, Khasa Punta PampaS. simonsi 2 ZMUC 80029 U35085c Per, Cusco, north of Urubamba below

ChainapuertoS. zimmeri ZMUC 126277 AY755641a Bolivia, Potos, PortilloS. magellanicus 1 AMNH RTC 449 AY755642a Chile, Regin VII, Termas de ChillnS. magellanicus 2 AMNH RTC 451 AY755643a Chile, Regin VII, RalcoS. fuscus 1 AMNH RTC 392 AY755644a Chile, Regin Metropolitana, Cerro El RobleS. fuscus 2 AMNH RTC 497 AY755645a Chile, Regin Metropolitana, Cerro El Roble


13/20

C .456 [Auk, Vol. 122

a general pa ern that was not previously appre-ciated because of inadequate sampling of mon-tane birds in this sector of the Cordillera Central(e.g. Cuervo et al. 2001). In fact, a similar pat-tern is well documented for low-elevation birds,many of which are distributed throughout therainforests of the Choc and extend eastward

along the northern lower foothills of the Andesinto the Nech center of endemism (Cracra1985) and into the mid-Magdalena valley (i.e.the ChocMagdalena biogeographic province;Hernndez-Camacho et al. 1992). Moreover,close relationships between pairs of taxa occur-ring in the Cordillera Central and Cordillera

T 2. Continued.

GenBankTaxon Voucher(s) number LocalityS. latrans 1 ZMUC 80042 U06170 b Ecuador, Sucumbos, between Santa

Brbara and GuanderalS. latrans 2 IAvH 11682 AY755646a Colombia, Cundinamarca, BojacS. l. subcinereus 1 ZMUC 125127 AY755647a Ecuador, Loja, UtuanaS. l. subcinereus 2 ZMUC 125122 AY755648a Ecuador, Azuay, MolleturoS. latrans ZMUC 125112 AY755649a Ecuador, Napo, Guacamayos Mountains (unnamed E race)S. superciliaris 1 MBM 8242 AY755650a Argentina, Tucumn, Taf del ValleS. superciliaris 2 MBM 8487 AY755651a Argentina, Tucumn, Taf del ValleS. stilesi 1 ICN 34569 AY755652a Colombia, Antioquia, Amal, Cajamarca (holotype) S. stilesi 2 ICN 34609 AY755653a Colombia, Antioquia, Amal, Escuela

Las nimasS. stilesi 3 ICN 34512 AY755654a Colombia, Antioquia, Amal,

Bodega ViejaS. stilesi 4 ICN 34584 AY755655a Colombia, Antioquia, Amal,

La SecretaUnnamed species ICN 31209 AY755656a Colombia, Risaralda, Mistrat, PisonesAlto de Pisones

Unnamed species ICN 34841 AY755657a Colombia, Huila, Finca MerenbergFinca Merenberg

Unnamed species LSUMZ 128624 X60945e Per, Pasco, Millpo, Abra Portachuela,Millpo east of Tambo de Vacas

Unnamed species ZMUC 80025 U35084 b Per, Apurmac, north side of NevadoAmpay Ampay Myornis senilis 1 ZMUC 80061 U35087c Ecuador, Loja, Parque Nacional Podocarpus

M. senilis 2 IAvH 11866 AY755658a Colombia, Caldas, Aranzazu, El LaurelPteroptochos tarnii AMNH RTC 467 AY065717f Chile, Regin IX, Provincia Cantn, ChacamoRhinocrypta lanceolata NRM 966793 AY078174g Paraguay, Dept. Boquern, 7 km west of

Dr. P. Peaa Present study b Arctander and Fjelds 1994c Arctander 1995d J. M. C. Silva unpubl. datae Edwards et al. 1991f Irestedt et al. 2002g

Johansson et al. 2002Museum acronyms: AMNH (American Museum of Natural History, New York); ANSP (Academy of Natural Sciences ofPhiladelphia, Philadelphia); CBF (Coleccin Boliviana de Fauna, La Paz); FMNH (Field Museum of Natural History, Chicago);IAvH (Instituto Alexander von Humboldt, Villa de Leyva); ICN (Instituto de Ciencias Naturales, Universidad Nacional deColombia, Bogot); LSUMZ (Louisiana State University Museum of Natural Science, Baton Rouge); MBM (Marjorie BarrickMuseum, University of Nevada, Las Vegas); NRM (Swedish Museum of Natural History, Stockholm); ZMUC (ZoologicalMuseum, University of Copenhagen, Copenhagen).


14/20


Occidental have been reported for muridrodents (Voss et al. 2002) and leptodactylidfrogs (Lynch 1999).

The four individuals ofS. stilesi examinedhad identical haplotypes in the cytochrome-b

region assayed. The mean sequence divergence(P distance) in cytochromeb betweenS. stilesiand the other members of its clade is 5.4%(range: 3.96.7%; Table 3). Assuming a sequencesubstitution rate of 1.62.0% divergence permillion years estimated for that gene (Fleischeret al. 1998, Love e 2004), the separation ofS.stilesi from its closest relatives appears to have been an ancient event, likely taking place in thelate Pliocene, more than ~2 million years ago.During the late Pliocene, the northern Andeswere in an active geological period of majorupli (Gregory-Wodzicki 2000), which presum-ably created opportunities for allopatric differ-entiation through formation of biogeographic barriers that fragmented previously continuouspopulations. We must caution, however, thatthis temporal scenario should be consideredtentative, because a molecular clock has not been calibrated specically forScytalopus andrates of cytochrome-b evolution >2% per millionyears have been estimated for some passerinetaxa (e.g. Warren et al. 2003). Thus, it is conceiv-able that the split ofS. stilesi may have takenplace more recently, possibly in the Pleistocene.

Criteria for recognition as a new species.Vocalizations in suboscine birds are innate.Given the slight variation in plumage inScytalopus , vocalizations very likely functionas the principal species-recognition signal andreproductive isolating mechanism in the group.Indeed, songs and scolds inScytalopus popula-tions are highly stereotypical and appear to bereliable indicators of species limits in the genus,

with vocal differentiation being concordant with

genetic differentiation (Arctander and Fjelds1994; Krabbe and Schulenberg 1997, 2003).Scytalopus stilesi has distinctive vocalizations,which differ from the vocalizations of otherScytalopus species about as much as those

species vocalizations differ from one another; itis genetically distinct; and it retains its integritythroughout a wide expanse of the CordilleraCentral of the Colombian Andes, where it occu-pies a unique habitat and elevational range. Insum, S. stilesi is a segment of an evolutionarylineage than can be considered a species underthe general lineage concept of species (DeQueiroz 1998). In addition, this taxon is fullydiagnosable and reproductively isolated fromother Scytalopus , thus meeting the criteria to beconsidered a new species under the phyloge-netic and biological species concepts (Cracra1989, Johnson et al. 1999).

Potential additional specimens.Given thatS. stilesi is relatively widespread and com-mon within its distribution range, the speciescould have been collected and misidentied inthe past. Thus, we reviewed the catalogues orexamined specimens of several major museumcollections housing Colombian birds. Theonly Scytalopus specimens from the distribu-tion and elevational range ofS. stilesi that wefound corresponded toS. atratus (morphologi-cally distinctive fromS. stilesi) and S. vicinior (very diffi cult to distinguish fromS. stilesi).Considering the lack of well-substantiatedrecords (i.e. specimens linked to recordings ofvocalizations) ofS. vicinior from the CordilleraCentral, it is conceivable that at least some spec-imens referred to that speciestaken on bothslopes in Quindo, Tolima, and Caldas (1,5002,600 m) and housed at the American Museumof Natural History and the Carnegie Museum

are, in fact,S. stilesi. However, because plumage

T 3. Pairwise mitochondrial DNA sequence-divergence amongScytalopus stilesi and its threeclosest relatives. Numbers below the diagonal are uncorrectedP distances; those above arenumbers of transition/transversion substitutions. Comparisons are based on a 284-base-pairregion of the cytochrome-b gene for which sequences were available for all four taxa. FourS.stilesi individuals were sequenced and had identical haplotypes.

Taxon (1) (2) (3) (4)(1) Scytalopus stilesi 11/0 18/1 15/1(2) UnnamedScytalopus , Alto de Pisones 0.0389 13/1 8/1(3) UnnamedScytalopus , Finca Merenberg 0.0671 0.0495 15/2(4) S. robbinsi 0.0565 0.0318 0.0601


15/20

C .458 [Auk, Vol. 122

and overall morphology are not diagnostic, theidentity of those specimens cannot, at present, be established reliably.Scytalopus stilesi and S.vicinior are quite distinct in mitochondrial DNAsequences (i.e. 11% cytochrome-b divergence;C. D. Cadena et al. unpubl. data), so analysesof ancient DNA extracted from the specimenswould probably enable identication. In anycase, eld surveys involving collection of speci-mens and voice recordings should be conductedin Quindo, Tolima, and Caldas to identify con-dently theScytalopus species that occur there.We did not nd any additional specimens thatcould potentially correspond toS. stilesi , andthere are no type specimens ofScytalopus fromthe northern sector of the Cordillera Central.

E C

Habitat .Scytalopus stilesi is restricted to mid-elevation montane forests (between 1,420 and2,130 m) classied as premontane wet forest(sensu Holdridge 1967) or as subandean forest(sensu Cuatrecasas 1958). The forests have mildmean temperatures year-round (17.022.0C).Rainfall pa ern is bimodal, with drier periodstoward the beginning and middle of the year(Wi e 1995); total annual rainfall varies region-ally from 2,500 to 4,500 mm. These mid-elevationforests are characterized by outstanding treeand epiphyte diversity. The most abundant treefamilies are Lauraceae, Moraceae, Clusiaceae,and Myristicaceae. Although there are no domi-nant tree genera, palms such asDyctiocariumlamarckianum, We inia spp., andGeonoma spp., as well as tree ferns (Cyatheaceae), are char-acteristic oristic components at some sites.Shrubs in the Melastomataceae and Rubiaceaeand vascular epiphytes (especially aroids) are

abundant in the understory. Terrestrial brome-liads are locally common, andChusquea bamboois locally abundant on ridges, disturbed patches,gaps, and forest edges. In Antioquia, canopyheight ranges from 6 to 7 m on ridges, 15 to 17 mon slopes, and 22 to 24 m along stream valleys,with sca ered emergent trees of up to 3032 m(Cuervo et al. 2001). In Risaralda, canopy heightat study sites was 21 m, on average, with emer-gent trees 45 m (Renjifo 1999, 2001).

Ecology.Scytalopus stilesi is a common spe-

cies throughout its known range. At RepresaMiraores, Antioquia, it was the fourth mostabundant understory bird in a 6-ha plot (listed

as S. femoralis in Cuadros 1988). Throughout ayear of study, its relative abundance within con-tinuous forest sites in Risaralda was 0.47 indi-viduals per count during 15-min point-counts.In addition, during ~200 days of eldwork inAntioquia,S. stilesi was recorded almost dailyin 14 localities.

Scytalopus stilesi is strongly dependent on for-ests, occurring in continuous forests (>1,000 ha),mature-forest fragments (2.8500 ha), tall sec-ondary forests, riparian forests, and forest edgesalong roads. In contrast, despite intensivesampling, we did not detectS. stilesi in otherhabitats, such as pine, eucalyptus, or shade cof-fee plantations; young second growth; or openareas. However, one individual seemed to holda territory in a live fence composed of a singlerow of trees with no understory, where it has been repeatedly observed and tape-recorded (P.Caycedo pers. comm.). AlthoughS. stilesi occursmostly in understory, below closed canopy-cover, it can also be found in forest gaps andin a variety of environmental conditions withinforest. We did not nd any signicant effect ofmean canopy height, terrain slope, elevation(range: 1,8602,065 m), density of stems inunderstory, or basal area of trees (ANOVAs,P >0.05,n = 8 sites) on the abundance ofS. stilesi estimated during point-counts.The new species distribution partly overlapsthose ofS. atratus confusus ,S. latrans latrans , andS. spillmanni. WithS. a. confusus it co-occurredin six localities in Anor and Amal (Antioquia) between 1,420 and 1,735 m; withS. l. latrans inVa Parque Angelpolis (Antioquia) between1,950 and 2,050 m; and withS. spillmanni inUcumar Regional Park and Otn-QuimbayaFauna and Flora Sanctuary (Risaralda) between2,040 and 2,110 m. In other localities , S. stilesi

was the onlyScytalopus present. WhenS. stilesico-occurred with other tapaculo, they weremarkedly segregated by habitat. For instance,S. a. confusus occupied forest edges, younggaps, steep slopes along streams, shrubbystands whereChusquea bamboo thickets weredominant, and nearby secondary vegetation;whereasS. stilesi occupied less-disturbed areas,typically the understory of tall, closed-canopyforest.

Like otherScytalopus , S. stilesi is much more

o en heard than seen; the species is very diffi -cult to observe as it forages rapidly on or closeto the forest oor, in root tunnels, and in tangles


16/20


of vegetation. Stomach contents invariablyincluded small fragments of arthropods and noplant material. Breeding and reproduction maytake place during the rst half of the year, aswith most passerine birds in the northern sector

of the Cordillera Central (Cuadros 1988, Cuervoet al. 2001). Adult males ICN 34569 and 34609were in breeding condition on 5 January and 14February, respectively. Females ICN 34610 (14February), 34615 (17 February), and 34505 (20February) were reproductively active, as sug-gested by the condition of their gonads (e.g. thelargest ovum of the former measured 2.7 mm).The last two females also had brood patches,an indication of active nesting. An adult wasobserved feeding a edgling on 20 July 1985at Represa Miraores (Cuadros 1988), and inAugust 2002 at Otn-Quimbaya (G. Londoopers. comm.), and a recently edged juvenilemale (ICN 34584) was collected on 9 June 2002.Although songs and calls are heard year-round,some variation in vocal activity has been noted by Cuadros (1988), who observed a peak ofactivity from August to December at RepresaMiraores.

Conservation.The subtropical premontanewet forests of the Cordillera Central have beensignicantly deforested and fragmented sincepre-Columbian times, especially during the19th and 20th centuries (Santa 1993, E er andvan Wyngaarden 2000). The Cordillera Centralis currently the most deforested Andean rangein Colombia; for the most part, its mid-elevationlandscapes are dominated by urban se lements,pastures, coffee plantations, and other types ofproduction systems, and much of the remainingforests are highly fragmented (E er 1998).

To evaluate the conservation status ofS. stilesi according to the criteria of the InternationalUnion for Conservation of Nature (IUCN),we used the method developed by Renjifo etal. (2002). A geographic information system(GIS) was used to model the distribution ofthe species on the basis of four geographic lay-ers: a general ecosystem map of Colombia (1:1500000; E er 1998), an ecosystem map of theAndean region of Colombia for the year 2000 (1:1000000; IAvH 2004), a digital elevation model(90 90 m; Shu le Radar Topography Mission2000; see Acknowledgments), and the locali-

ties geographic coordinates. We generated apotential distribution map for the species basedon these layers, using habitat continuity and

elevation range to dene distribution limits(Fig. 7). The rst step was to delimit a minimumconvex polygon that included all known locali-ties for the species. That polygon was extendedto include entirely continuous blocks of forest

types where the species was recorded, until aclear habitat discontinuity was located (i.e. adryer forest type or elevations above or belowknown range). Working with that new polygon,we measured extent of occurrence and potentialarea of occupancy based on suitable habitat,following IUCN (2001) guidelines. Finally,we measured the extent of habitat loss as theamount of formerly suitable forests that have been replaced by agroecosystems. We estimatedpopulation size on the basis of extent of suitablehabitat to make comparisons with each of theIUCN thresholds that dene levels of threat(IUCN 2001; Fig. 7).

We estimated thatS. stilesi has lost 63% ofits original habitat. That habitat loss has takenplace over centuries, and we believe it is unlikelythat 30% of the remaining habitat will be lostwithin the next 10 years. Moreover, the speciesis not subject to any known selective pressure.For those reasons, the species does not qualifyas threatened according to IUCN criterion A. Inaddition, its extent of occurrence is 38,800 km2 and potential area of occupancy 4,030 km2;therefore, the species does not meet the thresh-olds of IUCN criterion B to be consideredthreatened because of a small, fragmented,and declining distribution (20,000 km2 extentof occurrence or 2,000 km2 area of occupancy).Further, the species has high population densi-ties. For instance, two pairs had permanentterritories in a 2.8-ha forest fragment, which isroughly equivalent to 140 mature individualsper square kilometer. In Risaralda, the spe-cies was found in three forest fragments of7.3 6.2 ha. Assuming there was only one pairper fragment, that would be roughly equivalentto 30 mature individuals per square kilometer.These estimates should not be considered accu-rate, but they are useful for making an educatedguess of overall population size. The guresare quite high for most Neotropical birds (e.g.Cresswell et al. 1999, Robinson et al. 2000), butsome small insectivores a ain high populationdensities; for instance,Hylophylax naevioideshas

densities of 40 mature individuals per squarekilometer in lowland Panam (Willis 1974). Ifwe multiply these estimates forS. stilesi by the


17/20

C .460 [Auk, Vol. 122

extent of suitable habitat (i.e. 4,030 km2), pro- jected population size exceeds the thresholdof 10,000 mature individuals (IUCN criterionC) by more than one order of magnitude. As aresult, even if only 10% of the suitable habitat

were occupied, the species would not qualifyas threatened under criterion C. Further, thespecies does not qualify as threatened under cri-terion D, because it largely exceeds the thresh-olds for being considered vulnerable becauseof a very small (1,000 mature individuals) orrestricted population (


18/20


19/20

C .462 [Auk, Vol. 122

Andean avifauna. Pages 2546in IsolatedVertebrate Communities in the Tropics(G. Rheinwald, Ed.). Bonner ZoologischeMonographien, vol. 46.

G -W , K. M. 2000. Upli history of

the Central and Northern Andes: A review.Geological Society of America Bulletin 112:10911105.

H -C , J., A. H , R. O , T. W . 1992. Unidades

biogeogrcas de Colombia. Pages 105152in La Diversidad Biolgica de Iberoamrica(G. Hal er, Ed.). Acta Zoolgica Mexicana,Xalapa, Mexico.

H , L. 1967. Life Zone Ecology. TropicalScience Center, San Jos, Costa Rica.

I A H [IA H].2004. Mapa de ecosistemas de los Andescolombianos del ao 2000, Escala 1:1000.000.Instituto Alexander von Humboldt, Bogot,Colombia.

I U C N [IUCN]. 2001. IUCN Red List

Categories and Criteria, version 3.1. SpeciesSurvival Commission, International Unionfor the Conservation of Nature, Cambridge,United Kingdom.

I , M., J. F , U. S. J , P. G. P. E . 2002. Systematic relation-ships and biogeography of the tracheophonesuboscines (Aves: Passeriformes). MolecularPhylogenetics and Evolution 23:409512.

J , U. S., M. I , T. J. P , P. G. P. E . 2002. Basal phylogeny ofthe Tyrannoidea based on comparisons ofcytochromeb and exons of nuclearc-myc and RAG-1 genes. Auk 119:984985.

J , N. K., J. V. R , C. C .1999. Resolution of the debate over species

concepts in ornithology: A new comprehen-sive biologic species concept. Pages 14701482in Acta XXII Congressus InternationalisOrnithologici (N. J. Adams and R. H. Slotow,Eds.). BirdLife South Africa, Johannesburg.

K , N., T. S. S . 1997. Specieslimits and natural history ofScytalopustapaculos (Rhinocryptidae), with descrip-tions of the Ecuadorian taxa, includingthree new species. Pages 4788in Studiesin Neotropical Ornithology Honoring Ted

Parker (J. V. Remsen, Jr., Ed.). OrnithologicalMonographs, no. 48.K , N., T. S. S . 2003. Family

Rhinocryptidae (tapaculos). Pages 748787in Handbook of the Birds of the World, vol.8: Broadbills to Tapaculos (J. del Hoyo, A.Ellio , and D. Christie, Eds.). Lynx Edicions,Barcelona, Spain.

L , I. J. 2004. Mitochondrial dating andmixed support for the 2% rule in birds.Auk 121:16.

L , J. D. 1999. Ranas pequeas, la geometrade evolucin, y la especiacin en los Andescolombianos. Revista de la AcademiaColombiana de Ciencias Exactas, Fsicas yNaturales 23:143159.

R , L. M. 1999. Effect of the landscapematrix on the composition and conservationof forest bird communities. Ph.D. disserta-tion, University of Missouri, St. Louis.

R , L. M. 2001. Effect of natural andanthropogenic landscape matrices on theabundance of subandean bird species.Ecological Applications 11:1431.

R , L. M., A. M. F , J. D. A , G. H.K , B. L , E . 2002. Libro Rojode Aves de Colombia. Serie Libros Rojos deEspecies Amenazadas de Colombia. InstitutoAlexander von Humboldt and Ministerio delMedio Ambiente, Bogot, Colombia.

R , R. S., G. T . 1994. The Birdsof South America, vol. 2: The SuboscinePasserines. University of Texas Press, Austin.

R , R. 1911. The birds of North and MiddleAmerica, part V. Bulletin of the United StatesNational Museum, no. 50.

R , W. D., J. D. B , S. K.R . 2000. Forest bird communitystructure in central Panama: Inuence ofspatial scale and biogeography. EcologicalMonographs 70:209235.

S , E. 1993. La Colonizacin Antioquea:

Una Empresa de Caminos. T. M. Editores,Bogot, Colombia.S , F. B. 1975. Naturalists Color Guide.

American Museum of Natural History, NewYork.

V , J. M. E. 1990. Estudo bioacsticodas aves do Brasil: O gneroScytalopus.Ararajuba l:518.

V , R. S., M. G -L , V. P .2002. A new genus for Aepeomys fuscatus Allen, 1912, andOryzomys intectus Thomas,

1921: Enigmatic murid rodents from Andeancloud forests. American Museum Novitates3373:142.


20/20

Cuervo_etal2005

Documents

Transcript of Cuervo_etal2005