ArticlePDF Available

Taxonomic evaluation of the Grallaria rufula (Rufous Antpitta) complex (Aves: Passeriformes: Grallariidae) distinguishes sixteen species

Authors:

Abstract and Figures

Populations in the Rufous Antpitta (Grallaria rufula) complex occupy humid montane forests of the Andes from northern Colombia and adjacent Venezuela to central Bolivia. Their tawny to cinnamon-colored plumages are generally uniform, featuring subtle variation in hue and saturation across this range. In contrast to their conservative plumage, substantial vocal differences occur among geographically isolated or parapatric populations. Working within the framework of a comprehensive molecular phylogeny, we reexamined species limits in the G. rufula complex, basing taxonomic recommendations on diagnostic differences in vocalizations and considering identifiable differences in plumage where pertinent. We identified 16 populations for species designation, including seven populations previously described as subspecies and, remarkably, six new species described herein. Within one of these species, we identified less robust vocal differences between populations that we designate as subspecies. Geographic variation exists within another species, but its critical evaluation requires additional material. Taxonomic revisions of groups consisting of cryptic species, like the Grallaria rufula complex, are imperative for their conservation. Rather than widespread species as currently defined, these complexes can comprise many range-restricted taxa at higher risk of extinction given the continuing human pressures on their habitats.
Content may be subject to copyright.
https://doi.org/10.11646/zootaxa.4817.1.1
http://zoobank.org/urn:lsid:zoobank.org:pub:7CBDB6A9-9AF9-495F-A55A-83BF36A4934D
ZOOTAXA
ISSN 1175-5326 (print edition)
ISSN 1175-5334 (online edition)
Accepted by P. Rasmussen: 31 Mar. 2020; published: 21 Jul. 2020
4817
ZOOTAXA
Magnolia Press
Auckland, New Zealand
Zootaxa 4817 (1): 001–074
https://www.mapress.com/j/zt/
Copyright © 2020 Magnolia Press Monograph
Taxonomic evaluation of the Grallaria rufula (Rufous Antpitta) complex
(Aves: Passeriformes: Grallariidae) distinguishes sixteen species
MORTON L. ISLER1,2, R. TERRY CHESSER1,3, MARK B. ROBBINS4, ANDRÉS M. CUERVO5,6,
CARLOS DANIEL CADENA7 & PETER A. HOSNER1,4,8
1Division of Birds, Department of Vertebrate Zoology, Smithsonian Institution, P O Box 37012, Washington, DC 20013, USA.
2
antbird2707@gmail.com; https://orcid.org/0000-0002-0784-7234
3U.S. Geological Survey, Patuxent Wildlife Research Center, 12100 Beech Forest Road, Laurel, MD 20708, USA.
tchesser@usgs.gov; chessert@si.edu; https://orcid.org/0000-0003-4389-7092
4University of Kansas Biodiversity Institute, 1345 Jayhawk Boulevard, Lawrence, KS 66045, USA.
mbrobbins@ku.edu; https://orcid.org/0000-0001-7186-5253
5Instituto de Ciencias Naturales, Universidad Nacional de Colombia, Bogotá, Colombia.
6Museum of Natural Science, Louisiana State University, Baton Rouge, LA 70803, USA.
amcuervom@unal.edu.co; https://orcid.org/0000-0002-4949-0288
7Departamento de Ciencias Biológicas, Universidad de los Andes, Bogotá, Colombia.
ccadena@uniandes.edu.co; https://orcid.org/0000-0003-4530-2478
8Natural History Museum of Denmark and Center for Macroecology, Evolution, and Climate, University of Copenhagen,
Universitetsparken 15, Copenhagen 2100, Denmark.
peter.hosner@snm.ku.dk; https://orcid.org/0000-0001-7499-6224
ISLER ET AL.
2 · Zootaxa 4817 (1) © 2020 Magnolia Press
MORTON L. ISLER, R. TERRY CHESSER, MARK B. ROBBINS, ANDS M. CUERVO, CARLOS
DANIEL CADENA & PETER A. HOSNER
Taxonomic evaluation of the Grallaria rufula (Rufous Antpitta) complex (Aves: Passeriformes: Grallari-
idae) distinguishes sixteen species
(Zootaxa 4817)
74 pp.; 30 cm.
21 Jul. 2020
ISBN 978-1-77670-995-3 (paperback)
ISBN 978-1-77670-996-0 (Online edition)
FIRST PUBLISHED IN 2020 BY
Magnolia Press
P.O. Box 41-383
Auckland 1346
New Zealand
e-mail: magnolia@mapress.com
https://www.mapress.com/j/zt
© 2020 Magnolia Press
All rights reserved.
No part of this publication may be reproduced, stored, transmitted or disseminated, in any form, or by any
means, without prior written permission from the publisher, to whom all requests to reproduce copyright
material should be directed in writing.
This authorization does not extend to any other kind of copying, by any means, in any form, and for any
purpose other than private research use.
ISSN 1175-5326 (Print edition)
ISSN 1175-5334 (Online edition)
GRALLARIA RUFULA COMPLEX Zootaxa 4817 (1) © 2020 Magnolia Press · 3
Table of Contents
Abstract ...................................................................................................3
Introduction ................................................................................................4
Methods. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
........................8
........................9
........................9
........................9
.......................10
.......................11
.......................12
.......................13
.......................14
.......................15
.......................15
.......................16
.......................16
.......................17
.......................17
.......................18
.......................19
.......................20
.......................20
.......................21
.......................22
.......................22
.......................23
.......................24
.......................25
.......................28
.......................29
.......................29
.......................30
Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
PRINCIPAL CLADE A . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Grallaria saltuensis Wetmore, 1946, subspecies elevated to species . . . . . . . . . . . . . . . . . . . . . . . . . . .
Grallaria spatiator Bangs, 1898, subspecies elevated to species . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
PRINCIPAL CLADE B . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Grallaria rufula sensu stricto Lafresnaye, 1843 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Grallaria rufocinerea Sclater & Salvin, 1879 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
PRINCIPAL CLADE C . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Grallaria alvarezi Cuervo, Cadena, Isler, & Chesser, new species . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Grallaria saturata Domaniewski & Stolzmann, 1918, subspecies resurrected and elevated to species .
PRINCIPAL CLADE D. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Grallaria cajamarcae (Chapman, 1927), subspecies elevated to species . . . . . . . . . . . . . . . . . . . . . . . .
PRINCIPAL CLADE E. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Grallaria blakei Graves, 1987 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
PRINCIPAL CLADE F. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Grallaria gravesi Isler, Chesser, Robbins & Hosner, new species . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Grallaria oneilli Chesser & Isler, new species . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Grallaria obscura Berlepsch & Stolzmann, 1896, subspecies elevated to species . . . . . . . . . . . . . . . . .
PRINCIPAL CLADE G. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Grallaria centralis Hosner, Robbins, Isler, & Chesser, new species . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Grallaria ayacuchensis Hosner, Robbins, Isler, & Chesser, new species . . . . . . . . . . . . . . . . . . . . . . . .
PRINCIPAL CLADE H. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Grallaria occabambae (Chapman, 1923), subspecies elevated to species . . . . . . . . . . . . . . . . . . . . . . . .
Grallaria occabambae marcapatensis, Isler & Chesser, new subspecies . . . . . . . . . . . . . . . . . . . . . . . .
Grallaria sinaensis Robbins, Isler, Chesser, & Tobias, new species . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Grallaria cochabambae Bond & Meyer de Schauensee, 1940, subspecies elevated to species . . . . . . .
Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Author contributions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Appendices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ........................32
APPENDIX 1.Vocal Analysis of Principal Clades. . ..............................................................32
APPENDIX 2. Summary of Vocal Comparisons of Species in Different Clades.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
APPENDIX 3. Genetic Distance Matrix for Study Populations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
APPENDIX 4.Recordings Examined. .........................................................................48
APPENDIX 5.Specimens Examined. ..........................................................................50
APPENDIX 6.Photographs ..................................................................................51
Abstract
Populations in the Rufous Antpitta (Grallaria rufula) complex occupy humid montane forests of the Andes from northern
Colombia and adjacent Venezuela to central Bolivia. Their tawny to cinnamon-colored plumages are generally uniform,
featuring subtle variation in hue and saturation across this range. In contrast to their conservative plumage, substantial
vocal differences occur among geographically isolated or parapatric populations. Working within the framework
of a comprehensive molecular phylogeny, we reexamined species limits in the G. rufula complex, basing taxonomic
recommendations on diagnostic differences in vocalizations and considering identifiable differences in plumage where
pertinent. We identified 16 populations for species designation, including seven populations previously described as
subspecies and, remarkably, six new species described herein. Within one of these species, we identified less robust vocal
differences between populations that we designate as subspecies. Geographic variation exists within another species, but
its critical evaluation requires additional material. Taxonomic revisions of groups consisting of cryptic species, like the
Grallaria rufula complex, are imperative for their conservation. Rather than widespread species as currently defined, these
complexes can comprise many range-restricted taxa at higher risk of extinction given the continuing human pressures on
their habitats.
Key words: Andes, systematics, species limits, Grallariidae, Grallaria rufula, Grallaria blakei, Grallaria rufocinerea
.......................12
ISLER ET AL.
4 · Zootaxa 4817 (1) © 2020 Magnolia Press
Introduction
The Rufous Antpitta, Grallaria rufula (Lafresnaye, 1843) is a group of suboscine passerine bird populations that
inhabit humid montane forest and borders of the Andes from northern Colombia and adjacent Venezuela to central
Bolivia, mostly from 2500–3500 m. In addition to the nominate form, six subspecies are generally recognized
(Peters 1951; Krabbe & Schulenberg 2003; Greeney 2018): G. r. obscura Berlepsch & Stolzmann, 1896; G. r. spa-
tiator Bangs, 1898; G. r. occabambae (Chapman, 1923); G. r. cajamarcae (Chapman, 1927); G. r. cochabambae
Bond & Meyer de Schauensee, 1940; and G. r. saltuensis Wetmore, 1946. An additional subspecies, G. r. saturata
Domaniewski & Stolzmann, 1918, was subsumed within G. r. rufula by Cory & Hellmayr (1924). Subspecies des-
ignations were based for the most part on subtle differences in plumage colors, which range from bright rufous or
cinnamon to dull tawny or buff, sometimes with an olivaceous tinge to the upperparts. The paleness of belly plum-
age and presence or absence of a pale eye-ring also vary among populations. In addition, as noted by earlier authors
(e.g., Cory & Hellmayr 1924), substantial individual plumage variation is found within some G. rufula populations,
complicating interpretation of subtle plumage differences among populations. Further complicating plumage com-
parisons, underparts feathers of individuals of some subspecies feature subtle scalloping patterns with pale edging
whose appearance could be influenced heavily by feather wear. When considered geographically, plumage features
of G. rufula seem to vary randomly. Widely separated populations (e.g., G. r. spatiator, G. r. cajamarcae) can ap-
pear more similar to one another than to intervening populations (e.g., G. r. rufula), in “leapfrog” fashion (Remsen
1984).
The rufula complex (Greeney 2018) also includes a deeply rufescent population described as Grallaria blakei
Graves, 1987. This species occupies lower elevations on the same mountain slopes as G. rufula in Peru. Further-
more, molecular studies (Harvey et al. unpublished data; Chesser et al. 2020) recently determined that G. rufocine-
rea Sclater & Salvin, 1879, which is found at lower elevations than parapatric G. rufula populations in Colombia
and northern Ecuador, is embedded within the rufula complex, despite being distinguished from all other popula-
tions by its slate gray underparts.
In the final decades of the twentieth century, field workers began to record vocalizations of G. rufula, noting
substantial structural and tonal differences among populations. Since that time, it has been generally recognized that
multiple species are involved (e.g., Fjeldså & Krabbe 1990; Ridgely & Tudor 1994; Isler & Whitney 2002; Krabbe
& Schulenberg 2003; Schulenberg et al. 2007; López-O. et al. 2014). In addition to providing documentation of
vocal differences among known populations, recent biological exploration has refined and extended the geographic
distribution of the complex, providing specimens and vocalizations that have suggested the existence of new vocally
distinct populations (Schulenberg et al. 2007; López-O. et al. 2014; Hosner et al. 2015; Greeney 2018). Publicly
shared repositories of vocal recordings (e.g., xeno-canto [XC]; Macaulay Library [ML]) now provide recordings
that clarify the distributions of vocal types and support characterization of their repertoires.
Antpitta vocalizations, like those of other suboscines, are understood to be innate and heritable (Kroodsma &
Konishi 1991; Touchton et al. 2014). As such, their vocalizations are considered important in species recognition
and maintenance, and consequently provide an empirical “yardstick” (sensu Mayr & Ashlock 1991) for establishing
taxonomic status of allopatric populations under the Biological Species Concept (BSC; Isler et al. 1998; Johnson
et al. 1999; Baptista & Kroodsma 2001; Helbig et al. 2002; Isler et al. 2005; Remsen 2005). Vocal characters serve
especially well taxonomically when placed in phylogenetic/evolutionary frameworks which by themselves do not
necessarily invoke reproductive isolation (De Queiroz 2007; Sangster 2009). Therefore, vocal patterns, especially
when considered alongside phylogenetic and distributional information, provide a reliable and robust basis for as-
sessing species limits in complexes with subtle plumage variation, such as G. rufula. We here present a taxonomic
revision of the G. rufula complex based on geographic variation in vocalizations supported by patterns of plumage
coloration and by genetic variation described in a companion paper (Chesser et al. 2020).
Methods
We based taxonomic conclusions on diagnostic differences in vocal characters. Plumage differences among popula-
tions were not used to diagnose taxa independently but instead provided supporting information. Perceived plumage
differences among similarly colored populations in the G. rufula complex should be interpreted with caution due
GRALLARIA RUFULA COMPLEX Zootaxa 4817 (1) © 2020 Magnolia Press · 5
to substantial individual variation. For example, the large number of specimens available from localities in Cun-
dinamarca, Colombia, range in color intensity from deep rufous to much lighter rufescent yellow-brown.
We defined study populations (Table 1) based on currently recognized taxa (Peters 1951; Dickinson & Christidis
2014), results of an extensive phylogeographic study (Chesser et al. 2020), published comments regarding possibly
unrecognized populations (e.g., Graves 1987; Hosner et al. 2015), and preliminary vocal analysis. The companion
phylogeographic analysis (Chesser et al. 2020), based on mitochondrial and nuclear DNA sequences from 80 indi-
viduals of the rufula complex, revealed substantial genetic differentiation among populations, including a number of
unnamed populations within recognized subspecies, and identified eight principal clades into which the populations
were organized (Table 1). Principal clades and populations were monophyletic in the accompanying phylogeny
(Fig. 1) and have accrued substantial genetic divergence (> 5% uncorrected pairwise distance in mitochondrial
ND2). Because of the large number of study populations and the substantial genetic and vocal differentiation among
them, we employed a two-step vocal analysis. First we compared vocalizations of study populations within principal
clades, which are listed north to south. Within each clade, we examined spectrograms from available recordings,
selected samples of spectrograms for measurement, compared measurements, and formulated taxonomic proposals
based on diagnostic vocal differences within each clade. Then we compared major vocal attributes distinguishing
proposed species across all clades to ensure that these differences were indeed diagnostic across the entire G. rufula
complex. Consistent distinctions were the basis for taxonomic recommendations.
Vocal analysis and taxonomic recommendations. We compiled audio recordings from archives and unar-
chived contributions in an inventory maintained by MLI. We examined a total of 750 recordings. We reviewed
documentation of each recording to relate it geographically to a study population. We used RAVEN, version 1.5
(Bioacoustics Research Program, Cornell Laboratory of Ornithology) to make spectrograms of each vocalization
type delivered by each individual for as many as 20 recordings for each population, depending on availability.
When available samples were larger than 20, we visually examined RAVEN spectrograms, as well as spectrograms
provided by vocal archives, to identify type of vocalizations and character states (e.g., trills versus note series, note
shape) distinguishing populations. Spectrograms shown in figures were selected to express centrality in measure-
ments (e.g., mean number of notes in long songs) and to portray typical note shapes.
MLI measured vocal characteristics of each spectrogram using RAVEN. Measurements taken depended on
vocal structure (e.g., number of notes) but typically included duration and frequency measures. Note count and
duration were taken from the beginnings of initial and terminal notes in computation of pace (notes/sec). When
overtones were recorded, peak frequency measurements were made at the highest points of notes in the fundamen-
tal. Measurements in tables include mean (ξ) ± standard deviation (where normally distributed) followed by range.
Sample size refers to individuals, not recordings. Total number of recordings available for a population is often
greater than the combined sample sizes of long and short songs, because some recordings could not be measured
with precision and only a subsample was measured for populations with a large number of recordings. However,
spectrograms of all recordings were examined visually. Note shape was determined visually using blind tests.
Following Isler et al. (1998), to be considered diagnostic, differences between study populations had to be dis-
crete, typically non-overlapping character states that have the potential for unambiguous signal recognition. Ranges
of samples of continuous variables could not overlap (although exceptions could be made for outliers in large
samples of 20 or more), and the likelihood that ranges would not overlap with larger sample sizes was estimated by
requiring the means (ξ) and standard deviations (SD) of the population with the smaller set of measurements (a) and
the population with the larger set of measurements (b) to meet the test:
ξa + taSDa ≤ ξbtbSDb
where ti = the t-score at the 97.5 percentile of the t distribution for n – 1 degrees of freedom.
A similar test could not be used for ratios (e.g., pace), which are not distributed normally. Therefore, a non-
parametric bootstrap simulation was used to examine statistical significance. We compared the difference between
the means (DBM) for the two taxa being analyzed and two groups of generated data of the same sample sizes. The
method generated 10,000 sample population pairs, with replacement, and compared the DBM of the two compared
species to the distribution of DBMs of the simulated populations. The result was distributed normally, and signifi-
cance assigned according to the rules of this distribution.
ISLER ET AL.
6 · Zootaxa 4817 (1) © 2020 Magnolia Press
TABLE 1. Principal clades, study populations, number of vocal samples measured, and distribution. BO = Bolivia, CO = Colombia, EC = Ecuador, PE = Peru, VE = Venezuela. E = east
or eastern, N = north or northern, NW = northwest or northwestern, S = south or southern, W = west or western. Additional geographic information is provided in the companion paper
(Chesser et al. 2020). Most study populations occur within known ranges of recognized taxa and are named accordingly. Exceptions to this are recently discovered populations blakei 2,
named based on plumage similarities to blakei 1, and blakei 3, named based on vocal similarities to blakei 2. Following the genetic study of Chesser et al. (2020), populations with “a”
and “b” designations differed genetically by <3%, all other populations by >3% in mtDNA (Appendix 3), except romeroana, a subspecies of rufocinerea, which was not included in the
genetic study.
Principal Clade Study population Species1n2Geographic distribution Elevation (m)3
Clade A saltuensis 1 13 Serranía de Perijá, VE and CO 2500–3250
spatiator 2 9 Sierra Nevada de Santa Marta, CO 2200–2900
Clade B rufula 1 3 7 Eastern Andes, from S Táchira, VE, to N Santander, CO 1850–3050
rufula 1 or 33 10 E slope of Eastern Andes, Boyacá, CO 2800–3800
rufula 3 3 17 Eastern Andes, Cundinamarca and adjoining Meta, CO 2600–3450
rufocinerea 4 12 Central Andes and upper Magdalena valley, S Antioquia to W Huila, CO 1950–3250
romeroana 4 6 E Andean slope from Cauca, CO, S to Sucumbíos, EC 2400–3000
Clade C rufula 4 5 17 Western Andes, CO 2350–3650
rufula 2a 6 6 W slope Eastern Andes in the Iguaque Massif, Boyacá and adjacent Santander, CO 3000–3300
rufula 2b 6 8 Central Andes, Caldas to Tolima, CO 3150–3650
rufula 2c 6 8 Andes in the Colombian Massif, Cauca, and adjacent Huila, CO 3000–3400
rufula 546 39 Nariño and Putumayo, CO, through EC to E Andean slope in Piura and Cajamarca N of R. Marañón and E of R. Huan-
cabamba, PE
2550–3350
Clade D cajamarcae 7 12 Piura, Cajamarca and Lambayeque W and S of R. Huancabamba and R. Marañón, PE 2850–3400
Clade E blakei 1 8 10 E Andean slope in Amazonas and San Martín S of R. Marañón to Huánuco N of R. Huallaga, PE 1700–3000
Clade F obscura 1 9 28 E Andean slope in Amazonas S of R. Marañón to Huánuco N of R. Huallaga, PE 2400–3900
obscura 2 10 10 E Andean slope in Huánuco and Pasco S of the R. Huallaga and N of the R. Perené, PE 2750–3700
obscura 3 11 14 E Andean slope in Junín S of the R. Perené and R. Paucartambo, N of the R. Mantaro, and W of R. Ene, PE 3000–3600
Clade G blakei 2 12 14 E Andean slope from Huánuco S of the R. Huallaga to Junín W of the R. Ene and N of the R. Mantaro, PE 2400–2700
blakei 3 13 15 E Andean slope in Ayacucho W of the R. Apurímac between the R. Mantaro and R. Pampas, PE 2500–3700
Clade H occabambae 1a 14 16 E Andean slope in Junín and Cusco between R. Ene/Apurímac and R. Yanatili valleys, PE 2550–3650
occabambae 1b 14 15 E Andean slope in Cusco E of R. Yanatili valley, PE 2450–3650
cochabambae 1 15 6 E Andean slope in Puno, PE, and extreme western La Paz, BO 2900–3150
cochabambae 2 16 14 E Andean slope from E La Paz to extreme W Santa Cruz, BO 1950–3500
1 Ordinal number of descriptions of 16 species-level taxa recommended in this study.
2 Number of songs extracted for analysis. Spectrograms of a much larger number of recordings (750 total) were examined visually.
3 To nearest 50 m. A few published elevations that are outliers and require verification have been omitted.
4 Populations rufula 5a and rufula 5b of the companion paper (Chesser et al. 2020) were treated as one population because their vocalizations are indistinguishable.
GRALLARIA RUFULA COMPLEX Zootaxa 4817 (1) © 2020 Magnolia Press · 7
FIGURE 1. Results of the phylogeographic analysis of the Grallaria rufula complex, from Chesser et al. (2020). Numbers
above the nodes represent bootstrap support values based on 100 bootstrap replicates.
We recommend biological species status for populations that differed sufficiently in their vocalizations in ac-
cordance with guidelines developed in a study of sympatric species in the closely related Thamnophilidae (Isler et
al. 1998), an approach previously employed to examine species limits in another complex in the Grallariidae (Van
Doren et al. 2018). Following Isler et al. 1998, vocal differences are considered to be of a scale reflective of species
status under the BSC if the analysis revealed three or more diagnostic characters. However, three diagnosable dif-
ferences in vocal characters are a point of reference, not a requirement. Other considerations (biogeographic, mor-
phological, genetic) may at times indicate that fewer vocal characters might be considered an appropriate threshold
for species recognition.
Supporting information. In addition to informing the vocal analysis by breaking the complex into comparable
ISLER ET AL.
8 · Zootaxa 4817 (1) © 2020 Magnolia Press
principal clades, we also employed as supporting information the extent of genetic differentiation, as presented in
the companion paper (Chesser et al. 2020). To provide supporting morphological information, we examined speci-
mens of each study population from the collections of (in alphabetical order by acronym) the American Museum of
Natural History, New York (AMNH); Universidad de los Andes, Bogotá (ANDES); Academy of Natural Sciences
of Drexel University, Philadelphia (ANSP); Carnegie Museum of Natural History, Pittsburgh (CM); Centro de Or-
nitología y Biodiversidad, Lima (CORBIDI); Delaware Museum of Natural History, New Castle County (DMNH);
Field Museum of Natural History, Chicago (FMNH); Instituto Alexander von Humboldt, Villa de Leyva, Colombia
(IAvH-A); Instituto de Ciencias Naturales, Universidad Nacional de Colombia, Bogotá (ICN); University of Kansas
Biodiversity Institute, Lawrence (KU); Natural History Museum of Los Angeles County (LACM); Louisiana State
University Museum of Natural Science, Baton Rouge (LSUMZ); Museum of Southwestern Biology, Albuquerque
(MSB); Museo Universitario de la Universidad de Antioquia, Medellín (MUA-AVP); National Museum of Natural
History, Smithsonian Institution, Washington (USNM); and Universidad del Valle, Cali (UV). In addition to view-
ing entire specimen series at separate institutions, none of which held comprehensive collections, we gathered repre-
sentative specimens (2–4) of each of the 18 geographically coherent lineages for direct visual comparison at USNM.
Morphological measurements were taken in mm. Plumage coloration was categorized in accordance with Munsell
Soil Color Charts (Munsell Color Company, Grand Rapids, Michigan). A hyphen in the Munsell notation signifies
intermediacy. Results were compared within principal clades to identify apparent differences among populations as
supporting information.
Organization of the paper. The results section of the main text presents succinct descriptions of species using
a traditional format whereas supporting data and analysis of principal clades are in appendices. Appendix 1 provides
the supporting vocal analysis for each principal clade, and Appendix 2 compares results across clades. Genetic re-
sults are not duplicated here, but for convenience an uncorrected mtDNA distance matrix for study populations is
included as Appendix 3. Vocal recordings and sources used in the analysis are identified in Appendix 4. Plumage
coloration used as supporting information was compared using samples from study populations identified in Appen-
dix 5. Specimen photographs of examples, in some cases type specimens, are provided in Appendix 6. Some materi-
als (e.g., spectrograms) are duplicated in the results and in Appendix 1, but these are put into different contexts.
Results
Vocal differences along with supporting data were sufficient to recommend that 16 populations be considered dis-
tinct species. Taxonomic descriptions for each species (numbered in Table 1 and in the descriptions) are presented
below. We evaluate limits of three currently recognized species; elevate six taxa from subspecies to species rank;
resurrect a currently synonymized subspecies, elevating it to species rank; and describe six new species. We also
describe a new subspecies within one newly-elevated species. Sequence of authorship of new taxa reflects extent of
participation of individuals involved in collecting or analyzing material leading to their recognition as taxa. Spec-
trograms provided in taxon descriptions are at the same scale except as noted. Spectrogram scales in Appendix 1 are
consistent within vocal type but differ between vocal types for clarity. Archive acronyms are defined in Appendix
4.
Birds in nearly all populations deliver two types of vocalizations, which we classify as long songs and short
songs. In most populations, based on incidence in recordings examined and observations by recordists, short songs
are more commonly delivered than long songs. Some observers consider the short song as the “song” and the long
song as an “alarm call”. However, incidence in recordings also indicates that the relative use of long songs and short
songs differs among populations, and in one well-recorded population (blakei 1) only a long song has been docu-
mented. In light of this finding and of the number of stereotyped vocal characters they contain, both vocalizations
are considered “songs”.
In five populations (rufula 2b, rufula 2c, rufula 4, rufula 5, and cajamarcae) birds delivered extended songs of a
regular pattern consisting of a long series of notes followed by 2–4 short series of notes. These may represent a distinct
vocalization, but too few recordings of extended songs were available, and they were omitted from the analysis.
GRALLARIA RUFULA COMPLEX Zootaxa 4817 (1) © 2020 Magnolia Press · 9
PRINCIPAL CLADE A
This clade includes the two northernmost populations, saltuensis and spatiator, which are restricted to isolated
mountain systems. Trilled short songs distinguish them vocally from taxa in other clades in the complex.
Grallaria saltuensis Wetmore, 1946, subspecies elevated to species
Perijá Antpitta
Includes population designated saltuensis in the analysis.
Diagnosis. Differs from all other taxa in the complex in its olive to buffy-brown upperparts and extensively
white or whitish underparts. A combination of trilled long and short songs (Fig. 2) distinguish G. saltuensis from
all taxa in the complex except G. spatiator. Long songs of G. saltuensis differ from those of G. spatiator in having
fewer notes, slower pace, and less decline in the peak frequency of individual notes. The short song illustrated in
Figure 2 is thought to be given naturally; other variations thought to be given in response to playback of recorded vo-
calizations are included in the analysis. All variations of short songs of G. saltuensis appear to differ from short songs
of G. spatiator by slower pace and lower initial frequencies of note peaks, although sample sizes were limited.
FIGURE 2. Songs of Grallaria saltuensis: (A) long song, RNA Chamicero del Perijá, Cesar, Colombia (Borgmann ML297026);
(B) short song, above El Cinco, Serranía de Perijá, Cesar (Cuervo ML204673/ XC234289).
Distribution. Endemic to the Serranía de Perijá in Zulia, Venezuela, and La Guajira and Cesar, Colombia,
2500–3250 m.
Plumage. Overall plumage of upperparts, including rectrices, remiges (wing edgings paler), crown and auricu-
lars, olive to buffy brown (2.5Y 4/4-10YR 4/3), shading to color of underparts in malar area. Narrow breast band
buffy brown (10YR 7/4–6/4); remaining underparts except flanks but including undertail coverts light grayish white
tinged buffy brown; flanks like upperparts (specimen photographs App. 6, Figs. A13 and A34).
Etymology. The English name, first proposed by Meyer de Schauensee (1950), reflects the mountain range to
which this species is restricted.
Remarks. Grallaria saltuensis is specifically distinct based on vocal characters and supported by plumage
characters and mtDNA genetic distance of >7.5% from all other populations. A combination of trilled long and short
songs differentiates both G. saltuensis and G. spatiator from all other populations. Long songs of G. saltuensis and
G. spatiator differ significantly from one another, and significant differences in short songs are likely to be con-
firmed when larger samples are available. Species status of G. saltuensis and G. spatiator is supported by diagnos-
able plumage distinctions between the olivaceous G. saltuensis and rufescent G. spatiator and an mtDNA genetic
distance of ~7%. The lack of rufescent coloration sets G. saltuensis apart from all other taxa in the complex.
Grallaria spatiator Bangs, 1898, subspecies elevated to species
Sierra Nevada Antpitta
Includes population designated spatiator in the analysis.
Diagnosis. Upperparts and flanks reddish-yellow-brown; anterior underparts paler grading into grayish white
ISLER ET AL.
10 · Zootaxa 4817 (1) © 2020 Magnolia Press
posterior underparts. A combination of trilled long and short songs (Fig. 3) distinguishes G. spatiator from all other
taxa in the complex except G. saltuensis. Trilled long songs of G. spatiator differ from those of G. saltuensis by hav-
ing more notes, faster pace, and a greater decline in frequency of note peaks. Short songs of G. spatiator appear to
differ from those of G. saltuensis by faster pace and higher initial frequencies of note peaks, although sample sizes
were limited.
Distribution. Endemic to Colombia in the Sierra Nevada de Santa Marta in Magdalena, La Guajira, and Cesar,
2200–2900 m.
Plumage. Overall plumage of upperparts, including rectrices, remiges (wing edgings paler), crown and auricu-
lars, dark reddish yellow-brown (7.5Y 4/4–10YR4/3) shading to color of underparts in malar area. Overall plumage
of underparts, including undertail coverts, light grayish white tinged brownish yellow; broad breast band brownish
yellow (2.5Y–10YR 6/8); flanks like upperparts (specimen photographs App. 6, Figs. A14, A33, and A34).
Etymology. The English name Sierra Nevada is a shortened name widely used in Colombia for the Sierra Ne-
vada de Santa Marta, to which this species is restricted. The name was also used by Krabbe et al. (2019).
Remarks. Grallaria spatiator is specifically distinct based on vocal characters, supported by plumage charac-
ters and mtDNA genetic distance of >8.5% from all other populations. See Remarks under Grallaria saltuensis.
FIGURE 3. Songs of Grallaria spatiator: (A) long song, Cuchilla San Lorenzo, Magdalena, Colombia (Coopmans ML68051);
(B) short song, Cuchilla San Lorenzo (Coopmans ML236889).
PRINCIPAL CLADE B
This clade includes rufula 1 and rufula 3 of the Eastern Andes and G. rufocinerea of the Central Andes of Colombia
and extreme northern Ecuador, the latter including the poorly known subspecies G. r. romeroana. Grallaria rufoci-
nerea was not previously associated with the G. rufula complex and was added based on preliminary results from
a phylogenomic analysis of suboscine birds (Harvey et al. unpublished data), a result later confirmed by genetic
analysis of the rufula complex (Chesser et al. 2020). Long songs are a nearly evenly paced series of distinct notes,
and short songs consist of one or two long (>0.5 sec) notes. The two pairs of populations (rufula 1 + rufula 3 and
rufocinerea + romeroana) share this unique combination of vocalizations even though their plumage differences are
among the greatest in the complex.
GRALLARIA RUFULA COMPLEX Zootaxa 4817 (1) © 2020 Magnolia Press · 11
Grallaria rufula sensu stricto Lafresnaye, 1843
Muisca Antpitta
Includes populations designated rufula 1 and rufula 3 in the analysis.
Diagnosis. Overall plumage typically reddish-brown, underparts somewhat paler, but overall intensity variable.
Grallaria rufula sensu stricto is best distinguished by vocalizations (Fig. 4) from all other taxa in the G. rufula
complex except its morphologically distinct sister species, G. rufocinerea; its structurally unique long song, which
is a nearly evenly paced series of similar notes, rather flat in frequency. Grallaria rufula sensu stricto is also dis-
tinguished from all other taxa, including G. rufocinerea, by its two-noted short song, the second note broken into
abrupt vertical elements with a distinct burry quality.
Distribution. Eastern Andes from the Tamá massif in Táchira, Venezuela, and Norte de Santander, Colombia,
south to Cundinamarca and western Meta, Colombia, except for the west slope in the Iguaque Massif in Boyacá and
extreme southwestern Santander, Colombia, 1850–3800 m.
Plumage. Overall plumage of upperparts, including rectrices, remiges (primaries edged paler), crown and au-
riculars, typically reddish-brown (5YR 4/4) but variable and often brighter yellowish-red (5YR 4/6) in the south
and browner (5YR 4/3) in the north; coloration shades to color of underparts in malar area. Overall plumage of
underparts yellowish red-brown (5YR–7.5YR 5/8), becoming slightly paler in center of belly and undertail coverts
in the south; flank coloration like breast. Individuals of the northern population (rufula 1) have the malar a paler
buff, typically with a pale pink tone at the base of the mandible. Birds of the southern population (rufula 3) vary
in intensity of their rufescent tones, particularly in Cundinamarca, Colombia (specimen photographs App. 6, Figs.
A15, A16, A33, and A34).
Etymology. The English name honors the Muisca civilization that occupied the altiplano and slopes of the
Eastern Andes. Muisca culture survives in the contemporary Colombian society of this region.
FIGURE 4. Songs of Grallaria rufula sensu stricto: (A) long song, Quebrada La Vieja, Cundinamarca, Colombia (Athanas
XC32225); (B) short song, Quebrada La Vieja (Cadena IAvH-CSA6294/ML256116).
Remarks. The substantial genetic distance (~5–6%) between samples of populations rufula 1 and rufula 3 is
remarkable given the lack of vocal distinctions between these populations, but seems to agree with subtle pheno-
typic differences that remain to be assessed in detail. A better understanding of populations in the Eastern Andes
will require additional specimens and vouchered vocal recordings from the region between Norte de Santander and
southeastern Boyacá. The type locality of G. rufula was simply “Colombia” (Lafresnaye, 1843), but it seems highly
probable, as suggested by other authors (e.g., Cory & Hellmayr 1924), that the type specimen (photograph App. 6,
ISLER ET AL.
12 · Zootaxa 4817 (1) © 2020 Magnolia Press
Fig. A33) was collected in the Eastern Andes in the vicinity of Bogotá. As confirmation, we compared the holotype
at MCZ against a larger series, and it fell within the variance observed in rufula 3 around Bogotá and Cundinamarca.
Consequently, we have ascribed the name G. rufula to this population.
Grallaria rufocinerea Sclater & Salvin, 1879
Bicolored Antpitta
Includes populations designated rufocinerea and romeroana in the analysis.
Diagnosis. Differs from all other species in the complex in having gray underparts that contrast sharply with its
rufous head, throat, and upperparts. Its long song (Fig. 5), a nearly evenly paced series of similar notes, distinguishes
G. rufocinerea from all other taxa in the complex, except G. rufula sensu stricto. Its single-noted short song distin-
guishes G. rufocinerea from G. rufula sensu stricto and from all other taxa except G. centralis and G. ayacuchensis;
it is distinguished from the latter two species by greater duration and by note shape, which is flat or very slightly
upslurred, as contrasted with G. centralis (inverted U) and G. ayacuchensis (downslurred).
Distribution. Central Andes from central Antioquia, Colombia, south to the upper Magdalena Valley and along
the eastern slope from Cauca, Colombia, to Sucumbíos, Ecuador, 1950–3250 m.
Plumage. The dark gray underparts of G. rufocinerea are unique in the G. rufula complex (specimen photo-
graphs App. 6, Figs. A17 and A18).
Remarks. Species status is supported by vocal as well as plumage characters. Reexamination of the taxonomic
status of romeroana (Hernández & Rodríguez, 1979) requires more vocal data, especially recordings of its long
song, as well as additional plumage comparisons with the nominate. Pending that review, we recommend its main-
tenance as a subspecies of rufocinerea based on the original description. The geographic range of G. rufocinerea is
sympatric/parapatric with that of northern populations of G. saturata (see below).
FIGURE 5. Songs of Grallaria rufocinerea: (A) long song, RN Río Blanco, Caldas, Colombia (van Oosten XC18278); (B)
short song, RN Río Blanco (Geale XC51230).
PRINCIPAL CLADE C. This clade includes rufula 2 of the Central Andes of Colombia with an outlying popu-
lation (rufula 2a) on the western slope of the Eastern Andes, rufula 4 in the Western Andes of Colombia, and rufula
5 of the Andes from southern Colombia to extreme northern Peru. The structure of their short songs is distinct from
those of all other populations.
GRALLARIA RUFULA COMPLEX Zootaxa 4817 (1) © 2020 Magnolia Press · 13
Grallaria alvarezi Cuervo, Cadena, Isler, & Chesser, new species
Chamí Antpitta
Includes the population designated rufula 4 in the analysis.
Diagnosis. Overall plumage deep reddish-brown, deeper in color than other taxa, especially in the throat and
underparts. Vocalizations of G. alvarezi (Fig. 6) are distinguished from those of all other taxa in the complex, except
G. saturata, by a short song that consists of a single note, followed by a longer interval and ending with a trill. Short
songs of G. alvarezi differ diagnosably from those of G. saturata by rounded rather than flat or downslurred notes,
and the trill in the short song of G. alvarezi lacks the interruption found in the southern population of G. saturata.
Peak frequencies of notes in long songs of G. alvarezi decline, whereas those of G. saturata remain nearly constant.
In addition, typical notes of long songs of G. alvarezi are box-shaped and have a narrow band width, whereas those
of G. saturata are typically downslurred and have a wider band width, although less commonly they approach the
shape of G. alvarezi notes, especially in extended songs.
Distribution. Endemic to Colombia in Western Andes from Paramillo, northwestern Antioquia, south to north-
western Cauca, 2350–3650 m.
Holotype. Instituto Alexander von Humboldt (IAvH-A) 13358, prepared as round skin and partial body in fluid,
with frozen tissue samples number IAvH-CT-3997. Adult female mist-netted in upper montane wet forest on 23
August 2004 by M. A. Echeverry, S. Córdoba, S. Sierra (collector field number MAE-431, mist-net format number
89) in Colombia, Dpto. Risaralda, mpio. Pueblo Rico, vda. La Cumbre, PNN Tatamá (5°9’29”N, 76°1’0”W, eleva-
tion 2620–2680 m).
FIGURE 6. Songs of Grallaria alvarezi: (A) long song, Páramo de Frontino, Antioquia, Colombia (Krabbe XC27437); (B)
short song, Alto de Ventanas, Antioquia (Calderón-F. XC104251).
Description of holotype. Adult female. Overall plumage color intense reddish brown (2.5YR 3/4), especially
on crown, nape, mantle, rump and rectrices, and ventrally on throat, sides of neck, upper flanks, and thighs. Facial
area including orbital feathers, lores and auriculars also reddish brown (2.5YR 3/4). The breast shows an even more
intense saturation of reddish brown (close to 2.5YR 3/4 but more saturated), which suggests a diffuse breast band
that grades towards lighter buffy brown (7.5YR 6/6–6/8) at center of belly, particularly at the outer portion of vanes
of contour feathers, with inner portions duskier (close to 7.5YR 5/8), giving an appearance of a scaled pattern in
lower belly; vent and outer tail coverts slightly paler (7.5YR 7/6). Reddish-brown (5YR 3/4–3/6) on wings mostly
due to coloration of outer webs of remiges and wing coverts, whereas inner webs are slightly duskier (5YR 3/2).
Axillaries, lesser and median underwing coverts reddish brown (5YR 5/6–5/8). Greater underwing coverts dusky
brown (close to 5YR 3/2) with rufescent outer borders (7.5YR 6/8). Ovary 9 x 4.1 mm, largest follicle 1.2 x 2.0
mm in postovulatory regression, skull 100% pneumatized, abundant fat, no molt. Stomach, syrinx and other organs
unexamined but saved intact as partial liquid specimen in ethanol (IAvH-A 13358, MAE-431). Soft part coloration
(parentheses in Spanish as originally annotated on label): iris brown (“marrón”), maxilla dusky black (“cuerno
oscuro”), mandible dusky black with buffy base and pale tip (“base crema, punta clara, resto cuerno”), feet/legs
grayish blue (“azul grisáceo”).
Measurements of holotype. Bill length (total) 23.04 mm, exposed culmen 20.5 mm, bill from nares 12.49 mm,
bill width 5.27 mm, bill depth 6.01 mm, wing length (chord) 81.22 mm, flattened wing length 84 mm, tarsus length
43.23 mm, tail length 37.08 mm.
Paratypes. USNM 436485 (female), AMNH 133533 (female), LACM 37383 (female), FMNH 249750 (fe-
ISLER ET AL.
14 · Zootaxa 4817 (1) © 2020 Magnolia Press
male), AMNH 109634 (immature male), AMNH 109635 (female) (specimen photographs App. 6, Figs. A19 and
A34).
Variation in plumage. Most specimens show pale streaking in the centers of some contour feathers, and in
some the intensity of breast feathers suggests a diffuse band.
Etymology. The scientific name honors Colombian ornithologist Mauricio Álvarez Rebolledo, who led sev-
eral biological expeditions during the peak of political instability of the 1990s and early 2000s in Colombia, while
playing an important role in conservation and education. Mauricio pioneered avian sound recording in the country,
establishing the Colección de Sonidos Ambientales at Instituto Alexander von Humboldt. The English name honors
the “people of the mountains,” the Emberá-Chamí indigenous community inhabiting the slopes of northern Western
Andes of Colombia. Chamí means mountain, and Tatamá, the name of the type locality, means “the grandfather of
the rivers” in Emberá language.
Remarks. With larger samples of G. saturata from the Central Andes and from southern Colombia in Nariño, it
appears that additional vocal characters may be found to differ diagnostically between G. alvarezi and G. saturata.
These include the change in pace and intensity in long songs and duration of note intervals in short songs. The rufous
coloration of G. alvarezi is among the darkest and most intense of any population in the complex save G. blakei
and G. centralis; consequently, the plumage of G. alvarezi is distinct from its geographical neighbor G. saturata.
The northern range limit of this new species are the peaks of PNN Paramillo, but determining the southern bound-
ary awaits additional field work. We had no specimens in fresh plumage nor vocal recordings from south of PNN
Tatamá in western Risaralda. Specimens from Farallones de Cali and Cerro Munchique appear to be even darker
than the type series in Risaralda and Antioquia. Field work in the isolated upper montane forests and treeline of the
Western Andes in Cerro Calima, Farallones de Cali, and Cerro Munchique would help elucidate the range limits of
this species, and of G. saturata. A single vocal recording attributed to Cerro Munchique appears to be of G. saturata,
but the precise location of the recording (slope and elevation) is unknown.
Grallaria saturata Domaniewski & Stolzmann, 1918, subspecies resurrected and elevated to species
Equatorial Antpitta
Includes populations designated rufula 2a, rufula 2b, rufula 2c and rufula 5 in the analysis.
Diagnosis. Overall plumage typically reddish-brown, underparts paler, but overall intensity variable. Vocaliza-
tions of G. saturata (Fig. 7) are distinguished from all other taxa in the G. rufula complex except G. alvarezi by
a short song that consists of a single note, followed by a longer interval and ending with an even longer terminal
phrase whose structure appears to vary geographically (discussed in App. 1). Short songs of G. saturata differ diag-
nosably from those of G. alvarezi by flat or downslurred rather than rounded notes. Long songs of G. saturata differ
from those of G. alvarezi in having note peaks at the same frequency or nearly so, whereas peaks of G. alvarezi notes
decline substantially in frequency. Long song note shapes are typically more downslurred and wider in band width
than notes of G. alvarezi.
FIGURE 7. Songs of Grallaria saturata: (A) long song, Cerro Mongus, Carchi, Ecuador (Krabbe XC248386); (B) short song,
Cajanuma, Loja, Ecuador (Krabbe XC250311).
Distribution. Iguaque Massif in the west slope of Eastern Andes in Boyacá and extreme southwestern Santand-
er, Colombia; Central Andes from southern Antioquia to Tolima and in Cauca, Colombia; and the Andes from Nari-
GRALLARIA RUFULA COMPLEX Zootaxa 4817 (1) © 2020 Magnolia Press · 15
ño and western Putumayo, Colombia, south through Ecuador to Piura and Cajamarca, Peru, north of Río Marañón
and east of Río Huancabamba, 2550–3650 m.
Plumage. Overall plumage of upperparts, including rectrices, remiges (primaries edged paler), crown and au-
riculars, typically reddish-brown (5YR 4/4) but variable; specimens from the southern end of the range of rufula 5
(rufula 5b) in Zamora-Chinchipe, Ecuador, and from the northern end of the range of rufula 2 (rufula 2b) in Quindío,
Colombia, are lighter and browner than those from intervening populations at the northern end of the range of rufula
5 (rufula 5a) and from Cauca, Colombia (rufula 2c); coloration shades to color of underparts in malar area. Overall
plumage of underparts typically reddish yellow-brown (7.5YR 4/8) but lighter or darker consistent with upperparts;
center of belly and undertail coverts slightly or substantially paler and yellower (7.5YR 6/6–8/2); flank coloration
like upperparts or breast (specimen photographs App. 6, Figs. A20, A21, A33, and A34).
Etymology. The English name reflects the geographic location of the range of this species, which straddles the
Equator.
Remarks. See remarks under G. alvarezi for possible additional vocal differences from that species. Vocal
analysis in this clade was made difficult by geographically unbalanced availability of recordings and specimens.
Although recordings of 112 individuals were available for rufula 5, only a handful of recordings was available
for the three study populations comprising rufula 2, and nearly all were of short songs. Short songs varied but not
sufficiently to distinguish populations with certainty. Notably, genetic distance was ~4% between rufula 5 and the
rufula 2 populations. Additional field work in Colombia is needed to provide recordings, specimens, and tissue
samples and to determine whether populations are continuous, parapatric or allopatric. Type locality of G. rufula
saturata Domaniewski & Stolzmann, 1918 is San Rafael, Tungurahua, Ecuador, in the geographic range of rufula
5. The surprising occurrence of G. saturata in a small region on the west slope of the Eastern Andes surrounded by
populations of G. rufula sensu stricto to the north, east, and south is discussed in the companion paper (Chesser et
al. 2020).
PRINCIPAL CLADE D.
This clade is limited to cajamarcae of the Andes of northern Peru west of the Río Huancabamba and the upper Río
Marañón. The structure of its short song is distinct from that of all other populations.
Grallaria cajamarcae (Chapman, 1927), subspecies elevated to species
Cajamarca Antpitta
Includes population designated cajamarcae in the analysis.
Diagnosis. Upperparts and throat reddish yellow-brown; breast paler, posteriorly with pale feather edges grad-
ing to a mix of buff and white on the belly and undertail coverts; pale eye-ring. Underparts of cajamarcae are the
palest of all populations and separable in blind tests from neighboring populations. Short song is a slow-paced (~2
notes/sec), short (typically 5–6 notes) series of similarly shaped notes delivered at about the same frequency, which
is a unique vocalization in the complex (Fig. 8). Notes may be slightly downslurred or flat and often lengthen and
then shorten. Long song (Fig. 8) is a trill, similar to others in the complex but distinct in that peaks of the final
notes rise in frequency, a character otherwise found only in long songs of G. cochabambae. The long song of G.
cochabambae differs from that of G. cajamarcae in that peaks of initial notes decline in frequency, whereas peaks
of initial notes of G. cajamarcae are constant in frequency.
Distribution. Endemic to Peru; in Andes of Piura, Cajamarca, and Lambayeque west of the Río Huancabamba
and the upper Río Marañón, 2850–3400 m.
Plumage. Overall plumage of upperparts, including rectrices, remiges (primaries edged paler), crown and au-
riculars, dark reddish yellow-brown (7.5YR 4/4–5/4) with pale eye-ring and shading to color of underparts in malar
area. Overall plumage of underparts light reddish yellow-brown (7.5YR 7/10–6/10) with pale feather edgings,
becoming pale yellowish-buff (2.5Y 8/4) mixed with whitish in center of belly and undertail coverts; flanks dark
yellowish-brown (10YR 4/4) (specimen photographs App. 6, Figs. A22 and A33).
ISLER ET AL.
16 · Zootaxa 4817 (1) © 2020 Magnolia Press
FIGURE 8. Songs of Grallaria cajamarcae: (A) long song, Paja Blanca, Cajamarca, Peru (Lane ML236890); (B) short song,
Celendín-Cajamarca Road, Cajamarca (Zimmer ML236898).
Etymology. The English name reflects the scientific name and the geographic range of this species, which is
primarily confined to Cajamarca, Peru. The name was also used by Krabbe et al. (2019).
Remarks. Grallaria cajamarcae is specifically distinct based on vocal characters, supported by plumage char-
acters and mtDNA genetic distance of >5.5% with all other populations.
PRINCIPAL CLADE E.
This clade includes only blakei 1 of the east slope of the northern Andes of Peru south to the Río Huallaga. The lack
of a short song is unique in the complex.
Grallaria blakei Graves, 1987
Chestnut Antpitta
Includes the population designated blakei 1 in the analysis.
Diagnosis. Overall plumage dusky reddish-brown, anterior underparts similar but paler posteriorly, often with
faint barring. The long song of G. blakei is a lengthy trill (Fig. 9), unlike the frequency-modulated note series of the
apparently parapatric G. gravesi (see below). The long song of G. blakei differs from that of its geographic neighbor,
G. cajamarcae, by initial intervals between notes that shorten in duration (lengthening in G. cajamarcae) and by
lack of frequency change in the final notes (rising in G. cajamarcae). The lack of a short song in 36 recordings of
G. blakei makes it highly probable that absence in its repertoire is not a recording artifact; no other population in the
complex lacks a short song.
FIGURE 9. Song of Grallaria blakei: (A) long song, trail to Hacienda Paty, Huánuco, Peru (Walker ML236896).
Distribution. Endemic to Peru on east Andean slope from Amazonas and San Martín, south and east of the Río
Marañón, south to Huánuco, north of Río Huallaga, 1700–3000 m.
Plumage. Overall plumage of upperparts, including rectrices, remiges (primaries edged paler), crown and au-
GRALLARIA RUFULA COMPLEX Zootaxa 4817 (1) © 2020 Magnolia Press · 17
riculars, dusky reddish-brown (2.5YR 3/4), some slightly lighter and redder (to 2.5YR 4/6), shading to color of
underparts in malar area. Overall plumage of anterior underparts yellowish red-brown (5YR 4/6–5/8); belly barred
reddish-brown over dirty white with pinkish gray (5YR 6/2) tinge; flanks and undertail coverts reddish brown (5YR
4/4–4/6) (specimen photographs App. 6, Fig. A23).
Remarks. Grallaria blakei is distinguished vocally at the species level from neighboring populations, a distinc-
tion that is supported by the dark coloration noted in the description and by an mtDNA genetic distance of 6% or
more from all other populations. Specimens and tissue samples are needed from the northern end of the distribution
of G. blakei in the Cordillera Colán, Amazonas, Peru.
PRINCIPAL CLADE F.
This clade includes obscura 1, obscura 2, and obscura 3, populations occurring on the eastern slope of the Andes in
northern and central Peru south of the Río Marañón to the department of Junín. Both long and short songs contain
frequency modulated notes, unique in the rufula complex. Along with populations in Principal Clade C, the three
obscura populations present the greatest need for additional field study, including the collection of additional vocal
recordings and specimens for the extreme northern population of obscura 1.
Grallaria gravesi Isler, Chesser, Robbins & Hosner, new species
Graves’s Antpitta
Includes population designated obscura 1 in the analysis.
Diagnosis. Upperparts dark reddish yellow-brown; underparts paler, palest on belly and undertail coverts; pale
buff eye-ring. Both long songs and short songs are distinguished from those of all other populations except G. oneilli
and G. obscura by their frequency modulated notes, delivered in a series in long songs and in pairs in short songs
(Fig. 10). The pace of long songs of G. gravesi is faster than that of G. obscura but slower than that of G. oneilli.
Intervals between notes of the long song of G. gravesi increase and then decrease in duration, whereas intervals of
the G. oneilli long song remain nearly constant, and those of G. obscura increase in duration throughout. Two ad-
ditional characters distinguish vocalizations of G. gravesi from G. obscura although not from G. oneilli. Notes in
the second half of long songs of G. gravesi and G. oneilli rise in frequency, whereas those of G. obscura decline.
In addition, initial notes of short songs of G. gravesi and G. oneilli come to a single frequency peak, whereas initial
notes of G. obscura include three peaks (rarely two).
FIGURE 10. Songs of Grallaria gravesi: (A) long song, Bosque Unchog, Huánuco, Peru (Valqui ML236895); (B) short song,
34 km ENE of Huánuco, Huánuco (Cáceres ML168656).
Distribution. Endemic to Peru on east Andean slope in Amazonas and San Martín south and east of the Río
Marañón, south to Huánuco north of Río Huallaga, 2400–3900 m.
Holotype. Louisiana State University Museum of Natural Science (LSUMZ) 104488, tissue number LSUMZ
B–809. Adult female mist-netted in isolated patch of temperate forest and prepared by S. Allen-Stotz on 5 August
1981 at Puerto del Monte, ca. 30 km NE Los Alisos, San Martín, Peru (approximately 77° 28’ W, 07° 32’ S, eleva-
tion 3250 m).
Description of holotype. Adult female. Overall plumage of upperparts, including crown, auriculars, nape,
ISLER ET AL.
18 · Zootaxa 4817 (1) © 2020 Magnolia Press
back, and uppertail coverts, dark reddish yellow-brown (7.5YR 4/4), shading to underparts color in malar area; eye-
ring pale buff. Rectrices and remiges dark reddish yellow-brown (7.5YR 4/4), primaries edged paler. Throat and
breast reddish yellow-brown (7.5YR 5/8), paler on belly and undertail coverts with center of belly whitish; flank
coloration intermediate between upperparts and breast. Soft part colors: iris brown, bill slate, tarsi/feet blue-gray.
Ovarian mass 5 x 4 mm; skull ossified; mass 33.2 g; insect parts in stomach.
Measurements of holotype. Bill length 10.58 mm, bill width 4.77 mm, wing length 81.66 mm, tail length
45.40 mm, tarsus length 41.69 mm.
Paratopotype. LSUMZ 104491 (female).
Paratypes. LSUMZ 74099 (male), LSUMZ 74104 (female), ANSP 176470 (male), FMNH 296697 (female)
(specimen photographs App. 6, Fig. A24).
Etymology. We are pleased to name this species for our friend and colleague Dr. Gary R. Graves, whose orni-
thological contributions include the field work and subsequent analysis that resulted in the recognition of G. blakei.
After describing G. blakei, Gary Graves embarked on a study of the taxonomic issues presented by the G. rufula
complex that culminated in this paper, to which he has provided support.
Remarks. Long songs and short songs of G. gravesi and its sister species, G. oneilli and G. obscura, are unique
in the complex and set them apart from all other populations at the species level. Although fewer vocal differences
distinguish G. gravesi and G. oneilli than the substantial vocal differences distinguishing G. obscura from its sisters,
all three taxa are supported at the species level by plumage distinctions. Comparing plumages of the three taxa, the
lighter coloration of the underparts of the geographically intermediate G. oneilli were separable in blind tests from
the browner G. gravesi and G. obscura. The mtDNA genetic distance was greatest between G. gravesi and G. oneilli
(~5%) and was least between G. gravesi and G. obscura (~3%), but the greatest number of vocal differences were
found between G. obscura and G. gravesi, as well as between G. obscura and G. oneilli.
Grallaria oneilli Chesser & Isler, new species
O’Neill’s Antpitta
Includes population designated obscura 2 in the analysis.
Diagnosis. Upperparts reddish yellow-brown; underparts paler, grading to whitish on center of belly and un-
dertail coverts; pale buff eye-ring. Vocalizations differ from all taxa in the G. rufula complex except G. gravesi and
G. obscura by their frequency modulated notes. Long songs of G. oneilli (Fig. 11) are faster paced than long songs
of G. gravesi and G. obscura. Intervals between notes of the long song of G. oneilli are near-constant in duration,
whereas intervals of long songs of G. gravesi increase and then decrease in duration, and those of G. obscura in-
crease in duration throughout. Two additional characters distinguish vocalizations of G. oneilli from G. obscura,
although not from G. gravesi. Notes in the second half of long songs of G. gravesi and G. oneilli rise in frequency,
whereas notes of G. obscura decline in frequency. In addition, initial notes of short songs of G. gravesi and G.
oneilli come to a single frequency peak, whereas those of G. obscura include three peaks (rarely two).
FIGURE 11. Songs of Grallaria oneilli: (A) long song, Millpo, Pasco, Peru (Schulenberg ML40152); (B) short song, PN Yana-
chaga-Chemillén, Pasco (Vellinga XC62929).
GRALLARIA RUFULA COMPLEX Zootaxa 4817 (1) © 2020 Magnolia Press · 19
Distribution. East Andean slope in Huánuco and Pasco, Peru, south of the Río Huallaga and north of the Río
Perené, 2750–3700 m.
Holotype. Louisiana State University Museum of Natural Science (LSUMZ) 113581, tissue number LSUMZ B–
3510. Adult male found near timberline by MLI and P. R. Isler, mist-netted and prepared as a study skin by J. P. O’Neill
on 21 June 1983 at Bosque Potrero, 14 km W Panao, Huánuco, Peru (76° 05’ W, 09° 59’ S, elevation 3345 m).
Description of holotype. Adult male. Overall plumage of upperparts, including crown, auriculars, nape, back,
and uppertail coverts reddish yellow-brown (7.5YR 4/6), shading to color of underparts in malar area; eye-ring pale
buff. Rectrices and remiges reddish yellow-brown (7.5YR 4/6), primaries edged paler. Throat and breast light red-
dish yellow-brown (7.5YR 6/8), paler on belly with center of belly whitish; undertail coverts color of belly mixed
with pale yellowish-buff (2.5Y 8/4); flank coloration intermediate between upperparts and breast. Throat and breast
substantially paler than that of its sister species, G. gravesi and G. obscura. Testes 2 mm; skull ossified; mass 38 g;
moderate fat; no body molt; stomach packed with insect parts. Soft part colors: iris dark brown, base of mandible
pinkish, feet and tarsi medium gray.
Measurements of holotype. Bill length 11.54 mm, bill width 4.76 mm, wing length 80.97 mm, tail length 46.66
mm, tarsus length 42.00 mm.
Paratopotype. LSUMZ 113582 (male).
Paratypes. LSUMZ 128539 (male), LSUMZ 128541 (female), FMNH 66252 (female), AMNH 174092 (fe-
male immature) (specimen photographs App. 6, Fig. A25).
Etymology. The scientific and English names honor Dr. John P. O’Neill who collected the type specimen. John
O’Neill has dedicated his career to the exploration of ornithologically unknown regions in Peru, extending knowl-
edge of poorly known species, discovering new taxa, and training generations of students in the art and science of
ornithological field work. Newly described species in this paper reflect the modern renewal of ornithological col-
lecting in the Andes, of which John O’Neill was a pioneer.
Remarks. See remarks provided for G. gravesi and G. obscura.
Grallaria obscura Berlepsch & Stolzmann, 1896, subspecies elevated to species
Junín Antpitta
Includes population designated obscura 3 in the analysis.
Diagnosis. Upperparts dark reddish yellow-brown; throat and breast slightly paler; belly even paler with center
of belly and undertail coverts white to pale yellowish-buff ; pale buff eye-ring. Songs are distinguished from songs
of all other populations except G. gravesi and G. oneilli by their frequency-modulated notes, delivered in a series
in long songs and in pairs in short songs (Fig. 12). The vocalizations of G. obscura differ from those of G. oneilli
and G. gravesi in four characters: the pace of G. obscura long songs is slower than that of G. oneilli and G. gravesi;
the duration of intervals between notes of G. obscura long songs increases throughout, whereas those of G. gravesi
increase and then decrease, and those of G. oneilli are nearly constant; notes in the second half of G. obscura long
songs decline in frequency, whereas those of G. oneilli and G. gravesi rise in frequency, and finally, the initial note
of short songs of G. obscura includes three frequency peaks (rarely two), whereas in G. gravesi and G. oneilli the
initial note has a single peak.
FIGURE 12. Songs of Grallaria obscura: (A) long song, Toldopampa, Junín, Peru (Hosner ML171865); (B) short song, Satipo
Road, Junín (Spencer XC47028).
ISLER ET AL.
20 · Zootaxa 4817 (1) © 2020 Magnolia Press
Distribution. Endemic to Peru on the east Andean slope in Department of Junín, south of the Río Perené and
Río Paucartambo, north of the Río Mantaro, and west of Río Ene, 3000–3600 m.
Plumage. Overall plumage of upperparts, including rectrices, remiges (primaries edged paler), crown and au-
riculars, dark reddish yellow-brown (7.5YR 4/4), shading to color of underparts in malar area; eye-ring pale buff.
Throat and breast reddish yellow-brown (7.5YR-5YR 5/8), paler on belly with center of belly whitish; undertail
coverts pale yellowish-buff (2.5Y 8/4); flank coloration intermediate between upperparts and breast (specimen pho-
tographs App. 6, Figs. A26 and A33).
Etymology. The English name is adapted from Cory & Hellmayr (1924) and reflects the limited distribution of
this species.
Remarks. See remarks for G. gravesi. Of the three species of what was known as G. r. obscura, vocalizations
of G. obscura sensu stricto are the most distinct. The type locality (Maraynioc, Junín, Peru) of G. rufula obscura
lies within the geographic range of this species.
PRINCIPAL CLADE G.
This clade includes study populations blakei 2 and blakei 3, which have restricted geographic ranges in central Peru
south of the Río Huallaga from Huánuco to Ayacucho. The molecular analysis showed that they are not closely re-
lated to G. blakei, as had been previously assumed (Chesser et al. 2020). They are distinguished vocally from taxa
in other clades by a combination of a short song of a single unmodulated note and a trill-like long song. Although
blakei 3 is the only member of the G. rufula complex in its geographic range, blakei 2 occurs at lower elevations in
the same mountain ranges as obscura 1 and obscura 2.
Grallaria centralis Hosner, Robbins, Isler, & Chesser, new species
Oxapampa Antpitta
Includes population designated blakei 2 in the analysis.
Diagnosis. Upperparts dark reddish brown; anterior underparts slightly paler grading to pale yellow on lower
belly and pinkish white on undertail coverts. Trilled long songs in combination with short songs consisting of a
regularly repeated unmodulated note (Fig. 13) distinguish G. centralis from all other populations except G. aya-
cuchensis. Vocal differences between G. centralis and G. ayacuchensis include short song note shape (rounded in G.
centralis and downslurred in G. ayacuchensis) and pattern of change in duration of long song notes (notes lengthen
or remain constant in G. centralis and shorten in G. ayacuchensis). The pace of long songs of G. centralis is slower
than that of G. ayacuchensis; although ranges of values do not overlap, means do not differ significantly, presum-
ably because of small sample sizes.
FIGURE 13. Songs of Grallaria centralis: (A) long song, Apalla-Andamarca Road, Junín, Peru (Athanas XC148512); (B) short
song, Río Satipo, Junín (Hosner ML171921).
Distribution. Endemic to Peru on east Andean slope from Huánuco south of the Río Huallaga through Pasco to
Junín west of the Río Ene and north of the Río Mantaro, 2400–2700 m.
Holotype. Centro de Ornitología y Biodiversidad (CORBIDI) AV-PAH653. Adult female shot in secondary
GRALLARIA RUFULA COMPLEX Zootaxa 4817 (1) © 2020 Magnolia Press · 21
forest and prepared as a study skin by P. A. Hosner on 4 October 2008 along the Río Satipo above Calabaza, Junín
(11° 31’ S, 74° 52’ W, elevation 2400 m).
Description of holotype. Adult female. Overall plumage of upperparts dark reddish-brown (5YR 3/4), with
throat and breast grading towards yellowish red-brown (5YR 4/6). Crown, nape, orbital feathers, lores, auriculars,
back, rump, wing, uppertail coverts, and rectrices dark reddish-brown (5YR 3/4). Remiges dark reddish-brown
(5YR 3/4) with the inner vane of each feather slightly duskier. Throat, breast, side of breast, and upper flanks yel-
lowish red-brown (5YR 4/6), grading to antimony yellow (7.5YR 5/8) on center of belly and light buff on lower
belly. Some feathers on belly have faint pale neutral gray mottling. Lower flanks grading to yellowish red-brown
(5YR 4/6); undertail coverts antimony yellow (7.5YR 5/8). Ovary 8 x 6 mm, oviduct slightly enlarged and convo-
luted 2 mm; skull 100% pneumatized; moderate fat; mass 40.5 g; no bursa of Fabricius; stomach: insect parts, also
bolus of insects in its bill; no molt. Iris dark brown, feet/legs slate gray, bill black.
Measurements of holotype. Bill length 12.0 mm, wing length 77.2 mm, tail length 45.7 mm, tarsus length 43.0
mm.
Paratopotypes. KU 133990 (male), KU 1133991 (male), CORBIDI Av-MFOR324 (male).
Paratypes. LSUMZ 170664 (male), 160081 (female) (specimen photographs App. 6, Fig. A27).
Etymology. The scientific name reflects the range of this species near the geographic center of Peru. The Eng-
lish name reflects the Province of Oxapampa, Pasco, Peru, where initial recordings were made by T. S. Schulenberg
(ML 35960, 40063) and specimens (LSUMZ 106081, 170664) were collected by Louisiana State University expedi-
tions in 1982 and 1985.
Remarks. Vocalizations of G. centralis and G. ayacuchensis differ sufficiently to meet our guidelines for dis-
tinguishing them as species, and this conclusion is supported by an mtDNA genetic distance of ~4%. Plumages of G.
centralis and G. ayacuchensis are similar, but that of G. centralis is diagnostically darker and redder, especially on
the upperparts. Holotype also cataloged as University of Kansas Biodiversity Institute (KU) 124422, tissue number
14767.
Grallaria ayacuchensis Hosner, Robbins, Isler, & Chesser, new species
Ayacucho Antpitta
Includes population designated blakei 3 in the analysis.
Diagnosis. Upperparts dark reddish yellow-brown; anterior underparts paler grading to light buff on lower belly.
Trilled long songs in combination with short songs consisting of a regularly repeated unmodulated note distinguish
G. ayacuchensis (Fig. 14) from all other populations except G. centralis. Vocal differences between G. ayacuchensis
and G. centralis include short song note shape (rounded in G. centralis and downslurred in G. ayacuchensis) and
pattern of change in duration of long song notes (notes lengthen or remain constant in G. centralis and shorten in G.
ayacuchensis). The pace of long songs of G. ayacuchensis is faster than that of G. centralis, and although ranges of
values did not overlap, means did not differ significantly, presumably because of small sample sizes.
FIGURE 14. Songs of Grallaria ayacuchensis: (A) long song, Chupón, Ayacucho, Peru (Hosner ML186919); (B) short song,
Chupón (Robbins ML173831).
Distribution. Endemic to Peru on east Andean slope in Ayacucho west of the Río Apurímac between the Río
Mantaro and Río Pampas, 2500–3700 m.
ISLER ET AL.
22 · Zootaxa 4817 (1) © 2020 Magnolia Press
Holotype. Centro de Ornitología y Biodiversidad (CORBIDI) AV-KVG166. Adult male shot by P. A. Hosner in
Chusquea thicket in secondary forest, and prepared as a study skin by Karen Verde-Guerra, on 22 September 2012
at Chupón, Dpto. Ayacucho, Peru (13° 16’ S, 73°30’ W, elevation 3300 m). Voice recorded (ML186926).
Description of holotype. Adult male. Overall plumage of upperparts dark reddish yellow-brown (7.5YR 4/4),
with throat and breast grading toward yellowish red-brown (5YR 5/8). Crown, auriculars, nape, back, rump, wing,
uppertail coverts, and rectrices dark reddish yellow-brown (7.5YR 4/4). Remiges dark reddish yellow-brown (7.5YR
4/4) with the inner vane of each feather slightly duskier. Orbital, lores, throat, breast, side of breast, and upper flanks
yellowish red-brown (5YR 5/8), grading to light buff on lower belly. Some feathers on belly have faint pale neutral
gray mottling. Lower flanks grading to dark reddish yellow-brown (7.5YR 4/4), undertail coverts pinkish white
(7.5YR 8/4). Testes 8 x 4 mm; skull 50% pneumatized; no bursa of Fabricius; stomach: insects; moderate fat, wt:
48.4 g; heavy contour feather molt. Iris dark brown, legs dusky gray, bill black.
Measurements of holotype. Bill length 12.7 mm, wing length 83.7 mm, tail length 41.3 mm, tarsus length 46.3
mm.
Paratopotypes. KU 132736 (male), KU 132701 (male), CORBIDI Av-MBR 8329 (male), CORBIDI Av-
PAH1242 (male).
Paratypes. KU 119996 (female), AMNH 820776 (male) (specimen photographs App. 6, Fig. A28).
Etymology. The scientific and English names reflect its restricted known distribution on the humid eastern
slopes of the Department of Ayacucho, Peru.
Remarks. May also occur in adjacent northeastern Huancavelica. See remarks for G. centralis. Holotype also
cataloged as University of Kansas Biodiversity Institute (KU) 122629, tissue number 29375.
PRINCIPAL CLADE H.
This clade consists of four study populations; two (occabambae 1a and occabambae 1b) in south-central Peru, one
(cochabambae 1) in extreme southern Peru and western Bolivia, and one (cochabambae 2) in western and central
Bolivia. They are distinguished vocally from taxa in other clades by repertoires that include a short song of 2–3
unmodulated notes.
Grallaria occabambae (Chapman, 1923), subspecies elevated to species
Urubamba Antpitta
Includes populations designated occabambae 1a and occabambae 1b in the analysis.
Diagnosis. Nominate form upperparts dark reddish yellow-brown; underparts lighter with pale feather edgings;
undertail coverts pale yellowish-buff. Short songs of two (G. o. occabambae) (Fig. 15) or three (G. occabambae
marcapatensis) (Fig. 16) unmodulated notes distinguish G. occabambae from all but the two-noted short songs
of G. cochabambae and G. sinaensis. Short song notes of G. occabambae differ from those of G. cochabambae
and G. sinaensis in being rounded (as opposed to flat and slightly downslurred) and by longer notes and internote
intervals. Long songs of G. occabambae differ from those of G. cochabambae and G. sinaensis in having inverted
U-shaped notes (as opposed to flat or downslurred notes), in having internote intervals lengthening (as opposed to
lengthening, stabilizing, and shortening) resulting in decelerating pace, and in having a flat or slowly declining peak
frequency pattern (as opposed to peaks decreasing, stabilizing, and increasing); also apparently in overall pace,
although small sample sizes limit diagnosability.
Distribution. Endemic to Peru in extreme eastern Junín (Cordillera Vilcabamba) and Cusco east of the Río Ene
and Río Apurímac and between the Río Tambo to the north and the Río Marcapata to the south, 2450–3650 m. Two
subspecies; nominate subspecies occurs west of the Río Yanatili valley, Cusco.
Plumage. Overall plumage of upperparts, including rectrices, remiges (primaries edged paler), crown and au-
riculars, typically dark reddish-yellow-brown (7.5YR 4/4) but variable, tending toward browner (5YR 4/4) in nomi-
nate form and lighter and brighter (7.5YR 5/6) in G. o. marcapatensis; malar tending to color of breast. Throat,
breast and belly light reddish yellow-brown (7.5YR 6/8) with pale feather edgings. Undertail coverts pale yellow-
ish-buff (2.5Y 8/4) in nominate form and light reddish-yellow-brown (7.5YR 7/8) in G. o. marcapatensis (specimen
photographs App. 6, Figs. A29 and A30).
GRALLARIA RUFULA COMPLEX Zootaxa 4817 (1) © 2020 Magnolia Press · 23
FIGURE 15. Songs of Grallaria o. occabambae: (A) long song, Carrizales, Cusco, Peru (Lane ML236893); (B) short song, San
Luis, Cusco (Krabbe XC47645).
FIGURE 16. Songs of Grallaria occabambae marcapatensis: (A) long song, Abra Acjanaco, Cusco, Peru (Lane ML236892);
(B) short song, upper Manu Road, Cusco (Lane ML236891).
Etymology. The English name is adapted from Cory & Hellmayr (1924).
Remarks. The distinction in number of notes in short songs indicates that occabambae 1a and occabambae 1b
should be considered subspecifically distinct, and a new subspecies is described below. Compared to neighboring
taxa, plumage of G. occabambae is paler and separable from that of Grallaria centralis and G. ayacuchensis, but
similarly colored G. gravesi, G. obscura and G. sinaensis appear to be distinguished from G. occabambae only by
the greater extent of pale edgings on the breast feathers of G. occabambae, a character requiring confirmation in
large samples. Nominate form includes occabambae 1a. It appears that the Río Yanatili valley separates ranges of
subspecies of G. occabambae. The phylogeny (Chesser et al. 2020) includes in occabambae 1b a specimen (MSB
168436) collected on a mountain bordering the Río Yanatili to the east, and the type locality of G. rufula occabam-
bae is the Occabamba Valley = Río Ocobamba, which is a western tributary of the Río Yanatili (Stephens & Traylor
1983). Recordings from the upper reaches of this river are of occabambae 1a.
Grallaria occabambae marcapatensis, Isler & Chesser, new subspecies
Includes the population designated occabambae 1b in the analysis.
Diagnosis. Plumage similar to nominate except upperparts tend to be lighter and brighter, and undertail coverts
similar in color to belly but paler. Short songs of this subspecies are three-noted rather than two-noted as in the nom-
inate form, and the average pace of long songs is significantly slower, although ranges of measurements overlap.
Distribution. Endemic to Peru in Cusco east of the Río Yanatili valley and between the Río Tambo to the north
and the Río Marcapata to the south, 2450–3650 m.
Holotype. Field Museum of Natural History (FMNH) 222152. Adult female collected by C. Kalinowski on 26
July 1953 at Amacho, Cusco, Peru (near Marcapata, 70° 55’ W, 13° 30’ S, elevation 2750 m).
Description of holotype. Adult female. Overall plumage of upperparts, including crown, auriculars, nape,
back, and uppertail coverts, strong reddish yellow-brown (7.5YR 4/6); malar tending to breast color. Rectrices and
remiges dark reddish-yellow-brown (7.5YR 4/4), primaries edged paler. Throat, breast and belly light reddish yel-
low-brown (7.5YR 6/8) with pale feather edgings. Undertail coverts light reddish yellow-brown (7.5YR 7/8).
ISLER ET AL.
24 · Zootaxa 4817 (1) © 2020 Magnolia Press
Measurements of holotype. Bill length 10.91, bill width 4.55, wing length 77.67, tail length 41.91, tarsus
length 39.78.
Paratypes. LSUMZ 78575 (male), MSB 34460 (male), FMNH 429996 (female) (specimen photographs App.
6, Fig. A30).
Etymology. The subspecific name refers to the Marcapata district, where the type specimen was collected.
Remarks. Consistent differences in the number of notes in short songs and the generally slower pace of long
songs of G. o. marcapatensis indicate that the two subspecies of G. occabambae are on independent evolutionary
paths. However, these vocal differences are insufficient to support species status under our criteria, a conclusion
supported by the small mtDNA genetic distance (~1%) between them. Also, possible differences in coloration re-
quire substantiation with larger samples.
Grallaria sinaensis Robbins, Isler, Chesser, & Tobias, new species
Puno Antpitta
Includes population designated cochabambae 1 in the analysis.
Diagnosis. Upperparts dark reddish yellow-brown; underparts paler grading to pale buffy brown on belly and
undertail coverts; pale feather edgings and pencil streaks on belly and lower breast; pale eye-ring. Short songs of
two unmodulated notes distinguish G. sinaensis (Fig. 17) from all other populations except G. cochabambae and the
nominate form of G. occabambae. The shorter internote intervals and second notes of short songs of G. sinaensis
differentiate them from those of both G. occabambae and G. cochabambae, and the flat or slightly downslurred
shape of notes distinguishes short songs of G. sinaensis from the rounded notes of short songs of G. occabambae.
Comparing long songs, the flat or slightly downslurred shape of notes of G. sinaensis differs from the rounded
notes of G. occabambae and the sharply downslurred notes of long songs of G. cochabambae. Long songs of G.
sinaensis are also distinguished from those of G. cochabambae by their slower pace and by the greater extent that
note frequency rises in the terminal half of the song. They differ from long songs of G. occabambae by their falling
and rising frequency pattern and by their note duration pattern, in which notes gradually lengthen, become constant
in length, and then gradually shorten.
Distribution. Department of Puno, Peru, and extreme western La Paz, Bolivia, 2900–3150 m.
Holotype. CORBIDI AV00982. Adult male mist-netted and prepared by A. Nyári on 15 October 2009 below
Sina, Dpto. Puno, Peru (14° 29’ S, 69° 17’ W, elevation 3100 m); audio recording ML148119 in same recording as
the paratopotype.
Description of holotype. Adult male. Overall plumage of upperparts, including rectrices, remiges (primaries
edged paler), crown and auriculars, dark reddish yellow-brown (7.5YR 4/4), shading to color of underparts in malar
area; eye ring pale. Throat and breast light reddish yellow-brown (7.5YR 6/8–6/10); belly and undertail coverts pale
buffy brown (10YR 7/4); pale feather edgings and pencil streaks of belly extend to lower breast; flanks dark reddish
yellow-brown (7.5YR 5/4–5/2). Testes 11 x 5 mm, seminal vesicles enlarged; skull ossified; light fat; 34.2 g; light
body molt; no bursa; stomach filled with insects. Soft body parts: irides dark brown; bill black; tarsi bluish-gray.
Measurements of holotype. Bill length 9.6 mm, wing length 80.7 mm, tail length 40.9 mm, tarsus length 40.1
mm.
FIGURE 17. Songs of Grallaria sinaensis: (A) long song, Río Keara, PN Madidi, La Paz, Bolivia (Tobias XC73603); (B) short
song, Río Keara, PN Madidi (Tobias XC73602).
GRALLARIA RUFULA COMPLEX Zootaxa 4817 (1) © 2020 Magnolia Press · 25
Paratopotype. KU 115475 (male), tissue number KU 21173, audio recording ML148119 in same recording as
the holotype (specimen photographs App. 6, Fig. A31).
Etymology. The scientific name reflects the name of the type locality, and the English name reflects the Peru-
vian department in which the type locality is located.
Remarks. Plumage of G. sinaensis differs from that of G. cochabambae by pale feather edgings and pen-
cil streaks of lower underparts extending to breast, whereas breast of G. cochabambae is somewhat darker and
more uniform in appearance. The distribution in extreme western La Paz, Bolivia, is based on vocal recordings
(XC73602–XC73605) made in 2006 by J. Tobias and N. Seddon, who first noted the vocal distinctiveness of this
population on the upper Río Keara, Madidi National Park (69.059 W, 14.690 S, elevation 3000 m). Holotype also
cataloged as University of Kansas Biodiversity Institute (KU) 115476, tissue number 21234.
Grallaria cochabambae Bond & Meyer de Schauensee, 1940, subspecies elevated to species
Bolivian Antpitta
Includes population designated cochabambae 2 in the analysis.
Diagnosis. Upperparts dark reddish yellow-brown; underparts paler grading to pale buffy brown on center
of belly and undertail coverts; pale feather edgings and pencil streaks on belly; pale eye-ring. Short songs of two
unmodulated notes distinguish G. cochabambae (Fig. 18) from all other populations except G. sinaensis and the
nominate form of G. occabambae. The flat or slightly downslurred shape of short song notes of G. cochabambae
distinguishes them from the rounded notes of those of G. occabambae. The internote interval of short songs of G.
cochabambae is significantly longer than that of G. sinaensis and significantly shorter than that of G. occabambae.
Long songs of G. cochabambae differ from those of G. sinaensis and G. occabambae in their sharply downslurred
note shape and in their significantly faster pace. They differ from long songs of G. occabambae by their falling and
rising frequency pattern and by their duration pattern, in which notes lengthen, become even, and shorten while
intervals remain more constant. They differ from long songs of G. sinaensis by the lesser extent that frequency rises
in the terminal half of the song.
Distribution. Endemic to Bolivia in eastern La Paz and Cochabamba, 2950–3500 m.
Plumage. Overall plumage of upperparts, including rectrices, remiges (primaries edged paler), crown and au-
riculars, dark reddish yellow-brown (7.5YR 4/4), shading to color of underparts in malar area; eye-ring pale. Throat
and breast reddish-yellow-brown (7.5YR 5/8–5/6); belly and undertail coverts pale buffy brown (10YR 7/4), pale
feather edgings and pencil streaks restricted to belly; flanks dark reddish yellow brown (7.5 5/4) (specimen photo-
graphs App. 6, Fig. A32).
Etymology. The English name refers to the country to which this species is endemic. The name was also used
by Krabbe et al. (2019).
Remarks. The type locality (Incachaca, Cochabamba) of G. rufula cochabambae falls within the geographic
range of this species.
FIGURE 18. Songs of Grallaria cochabambae: (A) long song, Chuspipata, La Paz, Bolivia (Rowlett ML236894); (B) short
song, Corioco Road, La Paz (Boesman XC223701).
ISLER ET AL.
26 · Zootaxa 4817 (1) © 2020 Magnolia Press
Discussion
Since the advent of portable audio recorders in the 1970s (Parker 1991), field ornithologists have speculated that
multiple species were embedded within the wide-ranging Rufous Antpitta complex. Vocal recordings, along with
other data, have helped clarify species limits within this taxonomically confusing group. Instead of three species (G.
rufula, G. blakei, and G. rufocinerea, the latter only recently recognized as closely related), we now propose that 16
species be recognized (Figures 19 and 20). Finding this large number of unrecognized species in this complex was
not a total surprise given that other wide-ranging Andean species with conserved plumage (e.g., Scytalopus spp.;
Krabbe & Schulenberg 1997) have proven to be composed of multiple species. Other such complexes remain to be
studied. Species level analysis of such complexes should be based on testable differences in phenotypic characters
(Remsen 2016), but analysis time is a problem. To shorten analytic time, our analysis focused on study populations
within principal clades identified in a molecular phylogeny. After completion of within-clade analyses, vocal char-
acters were confirmed as diagnostic by comparing species across principal clades.
It is enlightening to compare the results of the vocally-based taxonomic analysis with the genetic diversity un-
masked in the companion paper (Chesser et al. 2020) and summarized here in a table of mtDNA genetic distances
(Appendix 3). Notably, of the 18 study populations that differed genetically by >3% in mtDNA, the vocalizations of
16 were judged to be distinct at the species level. The genetic finding that the morphologically distinct but vocally
similar G. rufula and G. rufocinerea are closely related species reinforces the conservative nature of vocalizations
in the complex and their value in taxonomic analysis.
Geographical ranges provided for the taxa described, elevated, or newly extended herein are based on both the
vocal data presented above and the molecular data in the companion paper (Chesser et al. 2020). Species are al-
lopatric with three exceptions that occur on the same mountain slopes but appear to be segregated elevationally: G.
rufocinerea with G. saturata (populations rufula 2b, rufula 2c, rufula 5); G. blakei with G. gravesi; and G. centralis
with G. oneilli and G. obscura. Distributions of most taxa in the rufula complex are delimited by recognized biogeo-
graphic barriers, including deep river canyons (e.g., Marañón, Huallaga, Apurímac) and other major topographical
features (e.g., Táchira Depression, Colombian Massif). However, ranges of some taxa are not so closely associated
with specific geographical features or are occasionally distributed on both sides of a recognized barrier, such as G.
saturata across the Magdalena Valley. These and other biogeographic considerations are reviewed in detail in the
discussion in Chesser et al. (2020), which is based on the results of both papers.
Continued field work, including collection of specimens, tissue, vocalizations, and natural history informa-
tion, is vital for the conservation of potentially threatened populations in the Grallaria rufula complex. In particu-
lar, we expect that field studies of populations assigned to G. saturata will expose additional differences among
them, possibly at the species level, and will allow determination of geographic limits along the Central Andes of
Colombia, where apparent range disjunctions may represent information gaps. Some uncertainty still surrounds
the southern limit of G. alvarezi in the southern Western Andes of Colombia, where additional vocal recordings
and specimens are needed to confirm that its range extends as far south as Cerro Munchique. Further field study
should also assess whether G. o. occabambae and G. o. marcapatensis are truly allopatric, isolated by the Río
Yanatili valley, or whether they are parapatric, or intergrade. Also requiring confirmation is the presence of ap-
parently large geographic gaps between G. o. marcapatensis and G. sinaensis and between G. sinaensis and G.
cochabambae, the only such gaps in the Grallaria rufula complex between western Venezuela and central Bo-
livia. Suitable habitats in these regions have seldom been visited by ornithologists, and these gaps may simply
be the result of insufficient exploration. Lastly, G. rufocinerea is poorly known. No modern specimens or tissues
are available, and large geographic gaps in data prevent assessing the taxonomic status of its two subspecies, and
the extent of range overlap with G. saturata. Continued collection of specimens and vocal recordings is key to
increasing our understanding of known and perhaps currently unknown populations of the G. rufula complex.
Finally, we stress that redefining species limits in the G. rufula complex has important implications for con-
servation. All 16 species-level taxa require status assessments, and many, if not most, are likely to require conser-
vation measures. For example, as previously defined, G. blakei is already considered Near Threatened (BirdLife
International 2019), yet our revised species limits underscore that several species are involved, and all appear to
be of conservation concern. Most species in the complex have narrow elevational ranges and small geographic dis-
tributions, and some are restricted to regions where habitats are extremely threatened. Of special note are the two
northernmost members of the complex, which are endemic to the Serranía de Perijá (G. saltuensis) and the Sierra
Nevada de Santa Marta (G. spatiator) and which are among the most evolutionarily distinct lineages in the group
GRALLARIA RUFULA COMPLEX Zootaxa 4817 (1) © 2020 Magnolia Press · 27
FIGURE 19. Map showing the taxonomy and distribution of the Grallaria rufula complex as revised herein, Part 1: popula-
tions formerly ascribed to G. rufula. Solid circles represent sectors (Isler 1997) that contain locations of genetic samples used
in a companion paper (Chesser et al. 2020) and stars that are sectors containing locations where specimens have been collected
or vocalizations recorded. Sectors may contain more than one genetic or vocal sample or specimen. Lines encompass sectors
attributed to the same taxon and are not projections of geographic range. Some sectors are placed away from boundary lines
(country boundaries, coasts) for clarity which results in discrepancies with geographic range projections which can be found in
the companion paper (Chesser et al. 2020).
ISLER ET AL.
28 · Zootaxa 4817 (1) © 2020 Magnolia Press
FIGURE 20. Map showing the taxonomy and distribution of the Grallaria rufula complex as revised herein, Part 2: G. rufoci-
nerea and populations formerly ascribed to G. blakei. Solid circles represent sectors (Isler 1997) that contain locations of genetic
samples used in a companion paper (Chesser et al. 2020) and stars are sectors containing locations where specimens have been
collected or vocalizations recorded. Sectors may contain more than one genetic or vocal sample or specimen. Lines encompass
sectors attributed to the same taxon and are not projections of geographic range which can be found in the companion paper
(Chesser et al. 2020).
centralis
GRALLARIA RUFULA COMPLEX Zootaxa 4817 (1) © 2020 Magnolia Press · 29
(Chesser et al. 2020). Both Perijá and Santa Marta are known for their high endemism and biotic irreplaceability
(e.g., Le Saout et al. 2013), and for threats to their biotas, including deforestation for legal and illegal crops and
cattle ranching (Fjeldså et al. 2005; Robbins & Nyári 2014; Negret et al. 2019). Despite increased understanding of
the biotic uniqueness of these mountain systems (e.g., Avendaño et al. 2015; Cadena et al. 2016), conservation chal-
lenges remain, and establishment of protected areas is vital. In general, high-elevation species are highly vulnerable
to local effects of climate change (Freeman et al. 2018), and conservation scenarios are likely no different in the rest
of the distribution of the G. rufula complex. Grallaria rufocinerea is already considered vulnerable owing to habitat
destruction and fragmentation and small population size (Renjifo et al. 2014), and the same might be true of other
range-restricted species in the north, including G. alvarezi, and species towards the southern end of the complex,
including G. oneillei, G. obscura, G. ayacuchensis, and G. sinaensis.
Acknowledgments
G. R. Graves initiated the study of the complex, and we appreciate the support he provided in its completion. This
study would not have been possible without the contributions of the many recordists whose names are listed fol-
lowing recording locations in Appendix 4. We appreciate the time they took to contribute to recording archives, and
we especially give thanks to those who went to extra lengths over many years to obtain recordings of the G. rufula
complex or to supply relevant information, including O. Acevedo, K. Borgmann, D. Calderón-F., P. Coopmans
(deceased), O. Cortés, K. Garrett, H. Greeney, P. Hocking, J. Jankowski, O. Janni, N. K. Krabbe, H. Lloyd, D. Ma-
cana, O. Marin, S. Mayer (deceased), C. Merkord, J. Muñoz, R. A. Rowlett, S. Chaparro, A. Spencer, J. Trimbel, T.
Valqui, B. Walker, B. Whitney, Y. Yabar, K. J. Zimmer, and J. Zuluaga. In particular, J. Tobias and N. Seddon kept
us apprised of their discoveries of the population in extreme western Bolivia. D. F. Lane shared his recordings and
his broad experience with populations in Peru and Bolivia. S. Beliew greatly improved the maps, and M. Foster
scoured the manuscript and made many suggestions to improve presentation. T. S. Schulenberg provided a number
of constructive comments that have led to important corrections and suggestions, including the scientific name of
a new species. We thank M. Andersen, A. Urbay-Tello, L. Cueto-Aparicio, K. Verde-Guerra, L. Sánchez-González,
A. Navarro-Sigüenza, R. Boyd, J. Núnez, J. Tiravanti, M. Combe, A.T Peterson, H. Owens, Á. Nyári, W. Wehtje,
W. Náñez, J. Montgomery, and C. González for facilitating and participating in KU fieldwork in Peru 2008–2012.
KU Fieldwork in 2012 was funded by National Geographic Society Committee for Research & Exploration Grant
#8872-11. Field work in Colombia was supported by the Lewis and Clark Exploration Fund, Society of Systematic
Biologists, AMNH Chapman Memorial Fund, IAvH, LSUMNS, and by NSF DDIG Grant DEB-0910285. We ap-
preciate the continuing support of The Macaulay Library, Cornell Laboratory of Ornithology (M. Medler and M.
Young); xeno-canto (W.-P. Vellinga and R. Planque); and Colección de Sonidos Ambientales (formerly Banco de
Sonidos Animales), Instituto Alexander von Humboldt (M. Álvarez, P. Caycedo, and O. Acevedo-Charry). We
are grateful to P. Sweet (AMNH), N. Rice (ANSP), S. Rogers (CM), J. Woods (DMNH), J. Bates and D. Wil-
lard (FMNH), D. Ocampo and S. Sierra (IAvH), K. Garrett (LACM), J. V. Remsen, R. Brumfield, and S. Cardiff
(LSUMNS), C. Witt (MSB), and G. H. Kattan for providing locational data and for the loan of specimens. Three
anonymous reviewers provided stimulating insights that led to amendments to the paper. P. Rasmussen edited and
provided numerous suggestions for improving the readability of the paper. Any use of trade, product, or firm names
is for descriptive purposes only and does not imply endorsement by the U.S. Government.
Author contributions
RTC and MLI conceived this study and its companion paper and worked together on both. MLI collected vocal and
biogeographic data and implemented the vocal analyses, identifying diagnostic vocal characters that distinguish the
17 taxa. PAH and MBR collected type specimens and drafted morphological descriptions of three newly described
Peruvian taxa. AMC and CDC wrote the morphological description of a newly described Colombian taxon and
obtained vocal and distributional data and examined specimens of all the Colombian populations. RTC and MLI
examined specimens from the entire complex, and RTC wrote morphological descriptions of the other three newly
described Peruvian taxa. PAH with input from AMC prepared the photographic inventory. MLI wrote the manu-
script which was improved substantially by contributions from all authors.
ISLER ET AL.
30 · Zootaxa 4817 (1) © 2020 Magnolia Press
References
Avendaño, J.E., Cuervo, A.M., López-O., J.P., Gutiérrez-Pinto, N., Cortés-Diago, A. & Cadena, C.D. (2015) A new species of
tapaculo (Rhinocryptidae: Scytalopus) from the Serranía de Perijá of Colombia and Venezuela. The Auk: Ornithological
Advances, 132, 450–466.
https://doi.org/10.1642/AUK-14-166.1
Bangs, O. (1898) On some birds from the Sierra Nevada de Santa Marta, Colombia. Proceedings of the Biological Society of
Washington, 12, 171–182.
Baptista, L.F. & Kroodsma, D.E. (2001) Avian bioacoustics. In: del Hoyo, J., Elliot, A. & Sargatal, J. (Eds.), Handbook of the
Birds of the World. Vol. 6. Mousebirds to Hornbills. Lynx Edicions, Barcelona, pp. 11–52.
Berlepsch, H. von & Stolzmann, J. (1896) On the ornithological researches of M. Jean Kalinowski in central Peru. Proceedings
of the Zoological Society of London, 1896, 322–388.
https://doi.org/10.1111/j.1096-3642.1896.tb03048.x
BirdLife International (2019) Species factsheet: Grallaria blakei. Available from: http://www.birdlife.org (accessed 2 Septem-
ber 2019)
Bond, J. & Meyer de Schauensee, R. (1940) Descriptions of new birds from Bolivia. Part 3. Mesomyodi. Notulae Naturae, 44,
1–4.
Cadena, C.D., Caro, L.M., Caycedo, P.C., Cuervo, A.M., Bowie, R.C.K. & Slabbekoorn, H. (2016) Henicorhina anachoreta
(Troglodytidae), another endemic bird species for the Sierra Nevada de Santa Marta, Colombia. Ornitología Colombiana,
15, 82–89.
Chapman, F.M. (1923) Descriptions of proposed new birds from Panama, Venezuela, Ecuador, Peru, and Bolivia. American
Museum Novitates, 67, 1–12.
Chapman, F.M. (1927) Descriptions of new birds from northwestern Peru and western Colombia. American Museum Novitates,
250, 1–7.
Chesser, R.T., Isler, M.L., Cuervo, A.M., Cadena, C.D., Galen, S.C., Bergner, L.M., Fleischer, R.C., Bravo, G.A., Lane, D.F. &
Hosner, P.A. (2020) Conservative plumage masks extraordinary phylogenetic diversity in the Grallaria rufula (Rufous
Antpitta) complex of the humid Andes. The Auk: Ornithological Advances, 137, 1–25.
https://doi.org/10.1093/auk/ukaa009
Cory, C.B. & Hellmayr, C.E. (1924) Catalogue of birds of the Americas and the adjacent islands. Pteroptochidae—Conopopha-
gidae—Formicariidae. Field Museum of Natural History, Zoological Series, No. 13, Part 3, 1–369.
https://doi.org/10:2307/1363248.
De Queiroz, K. (2007) Species concepts and species delimitation. Systematic Biology, 56, 879–886.
https://doi.org/10.1080/10635150701701083
Dickinson, E.C. & Christidis, L. (Eds.) (2014) The Howard & Moore Complete Checklist of the Birds of the World. Vol. 2. Pas-
serines. 4th Edition. Aves Press, Eastbourne, 752 pp.
Domaniewski, J. & Stolzmann, J. (1918) Contribution à la connaissance des formes du genre Grallaria Vieillot. In: Comptes
rendus des séances: Classe 4, Sciences biologiques. Société scientifique de Varsowie, Varsowie, pp. 479–484.
Fjeldså, J., Álvarez, M.D., Lazcano, J.M. & León, B. (2005) Illicit crops and armed conflict as constraints on biodiversity con-
servation in the Andes region. Ambio, 34, 205–211.
https://doi.org/10.1579/0044-7447-34.3.205
Fjeldså, J. & Krabbe, N. (1990) Birds of the High Andes. Zoological Museum, University of Copenhagen and Apollo Books,
Svendborg, 876 pp.
Freeman, B.G., Scholer, M.N., Ruiz-Gutiérrez, V. & Fitzpatrick, J.W. (2018) Climate change causes upslope shifts and moun-
taintop extirpations in a tropical bird community. Proceedings of the National Academy of Sciences of the United States of
America, 115, 11982–11987.
https://doi.org/10.1073/pnas.1804224115
Graves, G.R. (1987) A cryptic new species of antpitta (Formicariidae: Grallaria) from the Peruvian Andes. Wilson Bulletin, 99,
313–321.
Greeney, H.F. (2018) Antpittas and Gnateaters. Helm Identification Guides. Helm, London & New York, 496 pp.
Helbig, A.J., Knox, A.G., Parkin, D.T., Sangster, G. & Collinson, M. (2002) Guidelines for assigning species rank. Ibis, 144,
518–525.
https://doi.org/10.1046/j.1474-919X.2002.00091.x
Hernández C., J.I. & Rodríguez M., J.V. (1979) Dos nuevos taxa del genero Grallaria (Aves: Formicariidae) del alto valle del
Magdalena (Colombia). Caldasia, 12, 573–580.
Hosner, P.A., Andersen, M.J., Robbins, M.B., Urbay-Tello, A., Cueto-Aparicio, L., Verde-Guerra, K., Sánchez-González, L.A.,
Navarro-Sigüenza, A.G., Boyd, R.L., Núñez, J., Tiravanti, J., Combe, M., Owens, H.L. & Peterson, A.T. (2015) Avifaunal
surveys of the upper Apurímac River Valley, Ayacucho and Cuzco Departments, Peru: New distributional records and bio-
geographic, taxonomic, and conservation implications. Wilson Journal of Ornithology, 127, 563–581.
https://doi.org/10.1676/14-178.1
Isler, M.L. (1997) A sector-based ornithological geographic information system for the Neotropics. In: Remsen Jr., J.V. (Ed.),
Ornithological Monographs 48: Studies in Neotropical Ornithology Honoring Ted Parker. The American Ornithologists’
GRALLARIA RUFULA COMPLEX Zootaxa 4817 (1) © 2020 Magnolia Press · 31
Union, Washington, pp. 345–354.
https://doi.org/10.2307/40157542
Isler, M.L., Isler, P.R. & Brumfield, R.T. (2005) Clinal variation in vocalizations of an antbird (Thamnophilidae) and implica-
tions for defining species limits. Auk, 122, 433–444.
https://doi.org/10.1093/auk/122.2.433
Isler, M.L., Isler, P.R. & Whitney, B.M. (1998) Use of vocalizations to establish species limits in antbirds (Passeriformes: Tham-
nophilidae). The Auk, 115, 577–590.
https://doi.org/10.2307/4089407
Isler, P.R. & Whitney, B.M. (2002) Songs of the Antbirds: Thamnophilidae, Formicariidae, and Conopophagidae. Cornell
Laboratory of Ornithology, Ithaca. [CD-Rom]
Johnson, N.K., Remsen Jr., J.V. & Cicero, C. (1999) Resolution of the debate over species concepts in ornithology: a new com-
prehensive biological species concept. In: Adams, N.J. & Slotow, R.H. (Eds.), Proceedings 22nd International Ornithologi-
cal Congress, Durban. BirdLife South Africa, Johannesburg, pp. 1470–1482.
Krabbe, N.K., Schulenberg, T.S. & Kirwan, G.M. (2019) Rufous Antpitta (Grallaria rufula). In: del Hoyo, J., Elliott, A., Sar-
gatal, J., Christie, D.A. & de Juana, E. (Eds.), Handbook of the Birds of the World Alive. Lynx Edicions, Barcelona. Avail-
able from: https://www.hbw.com/node/56911 (accessed 11 February 2019)
Krabbe, N. & Schulenberg, T.S. (1997) Species limits and natural history of Scytalopus tapaculos (Rhinocryptidae) of Ecuador,
with descriptions of three new species and notes on extralimital forms. In: Remsen Jr., J.V. (Ed.), Studies in Neotropical
Ornithology Honoring Ted Parker. Ornithological Monographs, 48, pp. 47–88.
https://doi.org/10.2307/40157527
Krabbe, N. & Schulenberg, T.S. (2003) Family Formicariidae (ground antbirds). In: del Hoyo, J., Elliot, A. & Christie, D.A.
(Eds.), Handbook of the Birds of the World. Vol. 8. Broadbills to Tapaculos. Lynx Edicions, Barcelona, pp. 682–731.
Kroodsma, D.E. & Konishi, M. (1991) A suboscine bird (Eastern Phoebe, Sayornis phoebe) develops normal song without audi-
tory feedback. Animal Behaviour, 42, 477–487.
https://doi.org/10.1016/S0003-3472(05)80047-8
Lafresnaye, M.F. de (1843) Quelques nouvelles espèces d’oiseaux. Revue Zoologique, 6, 99.
Le Saout, S., Hoffmann, M., Shi, Y., Hughes, A., Bernard, C., Brooks, T.M., Bertzky, B., Butchart, S.H.M., Stuart, S.N., Bad-
man, T. & Rodrigues, A.S.L. (2013) Protected areas and effective biodiversity conservation. Science, 342, 803–805.
https://doi.org/10.1126/science.1239268
López-O., J.P., Avendaño, J.E., Gutiérrez-Pinto, N. & Cuervo, A.M. (2014) The birds of the Serranía de Perijá: the northernmost
avifauna of the Andes. Revista Ornitología Colombiana, 14, 62–93.
Mayr, E. & Ashlock, P.D. (1991) Principles of Systematic Zoology. McGraw-Hill, New York, 475 pp.
Meyer de Schauensee, R. (1950) The birds of the Republic of Colombia. Part 3. Caldasia, 5, 645–871.
Negret, P., Sonter, J.L., Watson, J.E.M., Possingham, H.P., Jones, K.R., Suarez, C., Ochoa-Quintero, J.M. & Maron, M. (2019)
Emerging evidence that armed conflict and coca cultivation influence deforestation patterns. Biological Conservation,
108176.
https://doi.org/10.1016/j.biocon.2019.07.021.
Parker III, T.A. (1991) On the use of tape recorders in avifaunal surveys. The Auk, 108, 443–444.
Peters, J.L. (1951) Check-list of Birds of the World. Vol. 7. Museum of Comparative Zoology, Cambridge, Massachusetts, 318
pp.
Remsen Jr., J.V. (1984) High incidence “leapfrog” pattern of geographic variation in Andean birds: implications for the specia-
tion process. Science, 224, 171–173.
https://doi.org/10.1126/science.224.4645.171
Remsen Jr., J.V. (2005) Pattern, process, and rigor meet classification. The Auk, 122, 403–413.
https://doi.org/10.1093/auk/122.2.403
Remsen Jr., J.V. (2016) A “rapid assessment program” for assigning species rank? Journal of Field Ornithology, 87, 110–115.
https://doi.org/10.1111/jofo.12142
Renjifo, L.M., Gómez, M.F., Velásquez-Tibatá, J., Amaya-Villareal, A.M., Kattan, G.H., Amaya-Espinel, J.D. & Burbano-Girón,
J. (2014) Libro Rojo de Aves de Colombia. Vol. I. Bosques Húmedos de los Andes y la Costa Pacífica. Editorial Pontificia
Universidad Javeriana and Instituto Alexander von Humboldt, Bogotá, 466 pp.
Ridgely, R.S. & Tudor, G. (1994) The Birds of South America, Vol. 2. The Suboscine Passerines. University of Texas Press,
Austin, Texas, 814 pp.
Robbins, M.B. & Nyári, A.S. (2014) Canada to Tierra del Fuego: Species limits and historical biogeography of the Sedge Wren
(Cistothorus platensis). Wilson Journal of Ornithology, 126, 649–662.
https://doi.org/10.1676/13-162.1
Sangster, G. (2009) Increasing numbers of bird species result from taxonomic progress, not taxonomic inflation. Proceedings of
the Royal Society of London B: Biological Sciences, 276, 3185–3191.
https://doi.org/10.1098/rspb.2009.0582
Schulenberg, T.S., Stotz, D.F., Lane, D.F., O’Neill, J.P. & Parker III, T.A. (2007) Birds of Peru. Princeton University Press,
Princeton, New Jersey, 656 pp.
Sclater, P.L. & Salvin, O. (1879) On the birds collected by the late Mr. T. K. Salmon in the state of Antioquia, United States of
ISLER ET AL.
32 · Zootaxa 4817 (1) © 2020 Magnolia Press
Colombia. Proceedings of the Zoological Society of London, 1879, 486–550.
https://doi.org/10.1111/j.1096-3642.1879.tb02684.x
Stephens, L. & Traylor, M.A. Jr. (1983) Ornithological gazetteer of Peru. Museum of Comparative Zoology, Cambridge, Mas-
sachusetts, 271 pp.
Touchton, J.M., Seddon, N. & Tobias, J.A. (2014) Captive rearing experiments confirm song development without learning in a
tracheophone suboscine bird. PLoS ONE, 9 (e95746), 1–9.
https://doi.org/10.1371/journal.pone.0095746
Van Doren, B.M., Freeman, B.G., Aristizábal, N.A., Álvarez-J., R.M., Pérez-Emán, L., Cuervo, A.M. & Bravo, G.A. (2018)
Species limits in the Rusty-breasted Antpitta (Grallaricula ferrugineipectus) complex. The Wilson Journal of Ornithology,
130, 152–167.
https://doi.org/10.1676/16-126.1
Wetmore, A. (1946) New birds from Colombia. Smithsonian Miscellaneous Collections, 106 (16), 1–14.
Appendices
APPENDIX 1. Vocal Analysis of Principal Clades. Table 1 identifies the eight principal clades and 22 constituent study
populations whose vocalizations are analyzed in this appendix.
Principal Clade A. Consists of study populations saltuensis and spatiator.
Vocal characteristics of the clade.—Both long songs and short songs of saltuensis and spatiator are trills.
Analysis of long songs.—Although similar structurally, long songs of spatiator and saltuensis (Fig. A1) differed in note count,
in pace (also in duration of notes and intervals, which are correlated with pace), and in the extent to which note peak fre-
quencies diminished between the initial and terminal notes (Table A1). Measured songs were recorded without playback.
A shorter long song of spatiator was recorded after playback, but measurements of pace and peak frequency decrease were
consistent. Characteristics of long songs in backgrounds of three additional recordings of saltuensis were consistent with
measurable recordings.
FIGURE A1. Long songs of Principal Clade A populations: (A) saltuensis, RNA Chamicero del Perijá, Cesar, Colombia (Borg-
mann ML297026); (B) spatiator, Cuchilla San Lorenzo, Magdalena, Colombia (Coopmans ML68051).
GRALLARIA RUFULA COMPLEX Zootaxa 4817 (1) © 2020 Magnolia Press · 33
TABLE A1. Selected measurements of long songs of populations of Principal Clade A, presented as mean±SD (range).
Population nNote count Pace (notes/sec) Peak frequency decrease (Hz)1
saltuensis 616±2.6, 12–20 5.2±0.2, 5.0–5.4 530±9, 520–542
spatiator 643±5.2, 36–49 9.8±0.4, 9.3–10.3 1319±152, 1133–1499
1Peak frequency initial note minus peak frequency terminal note.
Analysis of short songs.—Short songs of both saltuensis and spatiator sounded like an abrupt rapid trill. In spectrograms of
saltuensis, individual notes consisted of a rectangular block usually adorned by upslurred or downslurred elements. In four
recordings the block included upslurred elements (Fig. A2A); in two recordings of saltuensis the block was unadorned (Fig.
A2B); and in four recordings the block included downslurred elements (Fig. A2C). Based on a recording that contained two
types of notes, it appeared that type A was given naturally and type B after playback. In recordings of spatiator (Fig. A2D),
blocks included upslurred elements in all four examples. Note peaks in saltuensis recordings rose and fell slightly whereas
notes on spatiator recordings initially peaked at a higher frequency and dropped more sharply than in saltuensis. Ranges of
measurements (Table A2) did not overlap but were not significant, possibly due to the small sample size for spatiator.
FIGURE A2. Short songs of Principal Clade A populations: (A) saltuensis, above El Cinco, Serranía de Perijá, Cesar, Colom-
bia (Cuervo ML204673/XC234289); (B) saltuensis; Páramo de Sabana Rubia, Cesar, Colombia (Calderón-F. XC322051); (C)
saltuensis, Páramo de Sabana Rubia (Schmitt XC234277); (D) spatiator, Cuchilla San Lorenzo, Magdalena, Colombia (Coop-
mans ML236889).
TABLE A2. Selected measurements of short songs of populations of Principal Clade A, presented as mean±SD (range).
Population nPace (notes/sec) Peak frequency of 1st note block (Hz)
saltuensis 927.4±1.7 (25.1–29.5) n=7 2796±356 (2160–3112)
spatiator 435.0±5.1 (30.1–41.9) 3785±118 (3654–3888)
Analysis of Plumage.—The coloration of saltuensis was the least rufous and most distinct of any population in the G. rufula
complex, except rufocinerea and romeroana: upperparts tended toward olive brown (2.5Y 4/4) and underparts were primar-
ily whitish with coloration concentrated primarily in a narrow yellowish-brown (10YR 7/4–6/4) breast band. As a whole,
the color pattern of spatiator was similar but slightly darker and distinctly more rufescent; upperparts were dark reddish
yellow-brown (7.5YR–10YR 4/4) and a broader breast band brownish yellow (10YR 6/8).
Analysis of mtDNA.—Uncorrected pairwise distance between samples of spatiator and saltuensis was 7.5%.
Taxonomic conclusions.—According to our criteria, spatiator and saltuensis are distinct species. Their long songs differ in three
vocal characters. Ranges of short song measurements of two vocal characters do not overlap and may differ statistically
with larger samples of spatiator. Species status is supported by diagnosable distinctions in plumage and by deep genetic
divergence.
Principal Clade B. Consists of study populations rufula 1, rufula 3, rufocinerea, and romeroana.
Vocal characteristics of the clade.—Long songs are a series of distinct, rather flat, similar notes. Short songs consist of one or
two long notes.
Analysis of long songs.—Long songs consisted of a series of similar notes delivered at a steady or slightly accelerating pace (Fig.
A3). Small sample sizes for rufula 1, rufocinerea, and romeroana limited diagnosis, but available data suggested that with
larger samples significant differences might be found in pace and constituent measures of note and interval duration among
ISLER ET AL.
34 · Zootaxa 4817 (1) © 2020 Magnolia Press
these populations (Table A3). Spectrograms of the relatively few recordings available for rufula 1 were indistinguishable
visually from those of the well-recorded rufula 3, and mensural differences between them were not significant. Because of
uncertainty as to where to draw the line between the geographic ranges of rufula 1 and rufula 3, measurements of eastern
slope Boyacá songs, which varied individually in duration of notes and intervals, were separated in the analysis. In nearly
all recordings of rufula 3 from Cundinamarca, the duration of note intervals gradually shortened with the result that pace
accelerated, as can be seen in Figure A3A, but the possible diagnostic value of this characteristic remains to be determined.
Long songs of rufocinerea and romeroana were too few to compare statistically.
FIGURE A3. Long songs of Principal Clade B populations: (A) rufula 3 Quebrada La Vieja, Bogotá, Cundinamarca, Colombia
(Athanas XC32225); (B) rufocinerea, RN Río Blanco, Caldas, Colombia (van Oosten XC18278).
TABLE A3. Selected measurements of long songs of populations of Principal Clade B, presented as mean±SD (range).
Population nPace (notes/sec) Average note duration
(sec)
Average interval duration
(sec)
Average note peak
(Hz)
rufula 3
Cundinamarca
10 0.93 (0.75–1.30) 0.47±0.06
(0.39–0.52)
0.65±0.16
(0.36–0.86)
2651±220
(2097–2827)
rufula 3 or rufula 1
E Boyacá
5 0.82 (0.35–1.58) 0.40±0.12
(0.25–0.55)
1.11±0.75
(0.39–2.38)
2850±238
(2681–3323)
rufula 1
Norte de Santander
3 1.27 (1.09–1.38) 0.29±0.03
(0.25–0.32)
0.54±0.11
(0.41–0.60)
3255±75
(3169–3309)
rufocinerea 2 1.47 (1.44–1.50) 0.24±0.07
(0.19–0.29)
0.47±0.07
(0.41–0.52)
2488±224
(2330–2647)
romeroana 1 1.03 0.38 0.59 2520
Analysis of short songs.—The two-noted short songs of rufula 1 and rufula 3 differed consistently from the single-noted short
songs of rufocinerea and romeroana (Fig. A4). As can be seen in spectrograms, the second notes of rufula 1 and rufula 3
songs were broken into abrupt (~0.025 sec) vertical elements, providing a distinct burry quality. The single notes of ru-
focinerea were sometimes broken briefly (3 of 10 recordings), but these interruptions were irregular and not considered
diagnostic. Duration and frequency measurements of initial notes were indistinguishable statistically (Table A4).
GRALLARIA RUFULA COMPLEX Zootaxa 4817 (1) © 2020 Magnolia Press · 35
FIGURE A4. Short songs of Principal Clade B populations: (A) rufula 3, Quebrada La Vieja, Cundinamarca, Colombia (Ca-
dena IAvH-CSA6294/ML256116); (B) rufocinerea, RN Río Blanco, Caldas, Colombia (Geale XC51230). Short songs of rufula
1 were similar to those of rufula 3, and short songs of romeroana were similar to those of rufocinerea.
TABLE A4. Selected measurements of short songs of populations of Principal Clade B, presented as mean±SD (range).
Population n1st note duration
(sec)
Interval duration
(sec)
2nd note duration
(sec)
1st note peak
(Hz)
2nd note peak
(Hz)
rufula 3
Cundinamarca
10 0.77±0.11
(0.58–0.89)
0.35±0.06
(0.24–0.43)
0.72±0.07
(0.59–0.84)
2413±92
(2286–2540)
2205±187
(2000–2460)
rufula 3 or rufula 1
E. Boyacá
30.65±0.19
(0.44–0.79)
0.40±0.09
(0.31–0.49)
0.66±0.06
(0.61–0.73)
2644±223
(2500–2900)
2106±358
(1800–2500)
rufula 1
Norte de Santander
30.84±0.05
(0.79–0.89)
0.36±0.05
(0.30–0.40)
0.74±0.04
(0.71–0.79)
2682±26
(2654–2704)
2361±23
(2342–2386)
rufocinerea 10 0.99±0.21
(0.71–1.24)
2329±88
(2180–2480)
romeroana 50.96±0.12
(0.79–1.10)
2398±89
(2310–2540)
Analysis of Plumage.—Upperparts of typical rufula 3 were reddish brown (5YR 4/4–4/6), underparts yellowish red-brown
(5YR–7.5 YR 5/8), center of belly paler. No diagnosable plumage differences were found between rufula 3 and rufula 1
although there was a tendency for rufula 1 to be darker and browner. Specimens from Cundinamarca exhibited substantial
variation in depth of plumage coloration, as previously noted (Cory & Hellmayr 1924). The dark gray underparts (chest to
undertail coverts) of rufocinerea and romeroana contrasted with the reddish brown of the remainder of the body, including
their chin and throat, and were unique in the complex. The type specimens of romeroana examined look very much like
typical rufocinerea from Antioquia and Caldas in the Central Andes, except for slightly less rufous upperparts (see Hernan-
dez-Camacho & Rodríguez-M. 1979), but this difference may result from individual variation in conjunction with small
sample size (only 10 specimens of rufocinerea and 2 of romeroana are known to exist worldwide). Both rufocinerea and
romeroana exhibit chestnut throats and pale scaling in central lower belly, characters purported to distinguish romeroana
(Hernandez-Camacho & Rodríguez-M. 1979).
Analysis of mtDNA.—Uncorrected pairwise distances between rufula 1, rufula 3, and rufocinerea were 5–6%.
ISLER ET AL.
36 · Zootaxa 4817 (1) © 2020 Magnolia Press
Taxonomic conclusions.—Diagnostic differences in number and quality of short song notes, as well as striking plumage dif-
ferences, support maintenance of rufula 1/rufula 3 and of rufocinerea/romeroana as distinct species. We cannot identify
significant distinctions between rufula 1 and rufula 3 in available recordings or specimens, and consequently rufula 1 and
rufula 3 are considered populations of the same species despite the 5.5% mtDNA genetic distance between samples from
near the ends of their distribution. Additional recordings and specimens, especially from the region intermediate between
available tissue locations across Santander and in the Sierra Nevada del Cocuy, eastern Boyacá, should shed light on the re-
lationship of rufula 1 and rufula 3 populations. Short songs of rufocinerea and romeroana are indistinguishable, but sample
sizes for long songs are too limited for comparison. Specimens of romeroana other than the type specimens are unavailable
for plumage comparison, and described plumage differences between rufocinerea and romeroana are subtle. Under our
criteria we retain romeroana as a subspecies until further information, especially long song recordings, is obtained.
FIGURE A5. Long songs of Principal Clade C populations: (A) rufula 2c PN Puracé, Cauca, Colombia (Whitney ML236897);
(B) rufula 5, Cerro Mongus, Carchi, Ecuador (Krabbe XC248386); (C) rufula 4 Páramo de Frontino, Antioquia, Colombia
(Krabbe XC27437).
Principal Clade C. Consists of study populations rufula 2a, rufula 2b, rufula 2c, rufula 4, and rufula 5. Northern and southern
tissue samples of rufula 5 were separated by 2% in mtDNA pairwise distance, yet no significant differences were found
between vocal samples. Hence, rufula 5 was treated as a single population in subsequent vocal analyses.
Vocal characteristics of the clade.—Long songs are trills. Short songs include a relatively brief single note, followed by a longer
interval, and ending with an even longer duration terminal phrase that is often broken by abrupt intervals.
Analysis of long songs.—The only populations with a satisfactory sample of long songs (Figure A5; Table A5) were rufula 4
and rufula 5. No recordings of a long song were available for rufula 2a or rufula 2b. A single long song recorded for rufula
2c did not appear to differ diagnosably from long songs of rufula 5. Although long songs of rufula 4 were typically longer
GRALLARIA RUFULA COMPLEX Zootaxa 4817 (1) © 2020 Magnolia Press · 37
and faster-paced than those of rufula 5, measurements of these characteristics overlapped. However, peak frequencies of
notes of long songs of rufula 4 declined over the course of the song, whereas note peaks of long songs in other populations
remained essentially flat. In addition, typical notes of long songs of rufula 4 were box-shaped (sometimes approaching the
shape of a chevron) and had a narrow band width, whereas those of rufula 5 were wider and downslurred, although less
commonly they approached the shape of rufula 4 notes.
TABLE A5. Number of notes and pace of long songs of populations of Principal Clade C, presented as mean±SD
(range).
Population nNote count1Pace (notes/sec) Frequency change (Hz)2
rufula 2c 1 18 10.2 170
rufula 4 644.0±10.1 (29–54) 16.4 (11.3–18.8) 460±138 (319–634)
rufula 5 19 19.4±7.7 (10–36) 10.6 (7.6–14.0) 51±78 (0–220)
1One less than total number of notes was used to compute pace.
2Peak frequency of second note minus peak frequency penultimate note.
Analysis of short songs.—Short songs of all populations had a similar structure and were comprised of a single note, a longer
interval, and a terminal phrase that was most often broken by abrupt intervals but sometimes delivered as a continuous note
(Fig. A6, Table A6).
TABLE A6. Selected measurements of short songs of populations of Principal Clade C, presented as mean±SD (range).
Population nDuration 1st note (sec) Duration interval (sec) Duration final group (sec)
rufula 2a 60.092±0.056 (0.060–0.206) 0.157±0.046 (0.123–0.247) 0.255±0.090 (0.200–0.435)
rufula 2b 80.065±0.008 (0.055–0.083) 0.163±0.013 (0.148–0.189) 0.286±0.034 (0.242–0.337)
rufula 2c 70.092±0.016 (0.068–0.114) 0.155±0.044 (0.118–0.240) 0.309±0.059 (0.235–0.410)
rufula 4 11 0.070±0.009 (0.059–0.092) 0.108±0.016 (0.085–0.128) 0.238±0.044 (0.175–0.294)
rufula 5 60.073±0.012 (0.062–0.096) 0.206±0.020 (0.176–0.228) 0.346±0.061 (0.263–0.422)
In seven recordings of rufula 2a, the duration of the elements and the pattern of the terminal phrase varied. In three high quality
recordings the terminal phrase was continuous (Fig. A6A); in one recording it was broken nearly in half by an abrupt inter-
val (Fig. A6B); and in another recording the initial note and the interval were longer in duration, and the terminal phrase was
broken into nine segments by abrupt intervals (Fig. A6C). The latter pattern was also suggested in spectrograms of three
other recordings whose elements were obscured by reverberation.
In eight recordings of the short song of rufula 2b, the long terminal phrase was broken by abrupt intervals into five segments
that became progressively shorter in duration (Fig. A6D). Individual notes were flat, although a few started or ended with
a sharp uptick or downtick.
In one of seven recordings of rufula 2c, the terminal phrase was broken into ten abrupt bursts of sound (Fig. A6E). In five of
the seven recordings, the terminal phrase had the appearance of a single continuous note, but was sometimes irregularly
broken up or with a hint in the spectrogram that they may contain similar short bursts, suggesting that the continuous ap-
pearance might be the result of reverberation. In the final example the terminal phrase had a clear interval after an initial
element, somewhat resembling that of rufula 5 (see below), but shorter in duration so that the terminal phrase sounded like
a continuous trill to the human ear.
The terminal phrase of short songs of rufula 5 was broken up into typically five (3–6) elements that progressively shortened
in duration (Fig. A6F). The initial elements of the terminal phrase (second and third elements in the song) were separated
by an interval long enough (~0.06 sec) to be distinguished consistently by the human ear from the following trill (n = 33),
thus appearing to differ diagnostically from short songs of all three rufula 2 populations with the possible exception of one
example of rufula 2c.
Short song notes of rufula 4 were rounded or box-shaped (Fig. A6G), whereas those of other populations of Principal Clade C
were flat or slightly downslurred. The downslurred “tail” at the end of notes shown in the rufula 4 illustration was some-
times omitted. The duration of the interval after the first note (Table A6) was significantly shorter in rufula 4 than in rufula
5, although the difference in this measurement between rufula 4 and the other populations was not significant.
ISLER ET AL.
38 · Zootaxa 4817 (1) © 2020 Magnolia Press
FIGURE A6. Short songs of Principal Clade C populations: (A) rufula 2a, SFF Iguaque, Boyacá, Colombia (Córdoba IAvH-
CSA13025); (B) rufula 2a, Morro Negro, Boyacá (López IAvH-CSA9941); (C) rufula 2a, SFF Iguaque (Caycedo ML236886);
(D) rufula 2b, Finca La Martinica, Caldas, Colombia (Cuervo IAvH-CSA30914); (E) rufula 2c, PN Puracé, Cauca, Colombia
(Coopmans ML236888); (F) rufula 5, Cajanuma, Loja, Ecuador (Krabbe XC250311); (G) rufula 4, Alto de Ventanas, Antioquia,
Colombia (Calderón-F. XC104251).
Analysis of Plumage.—Coloration of rufula 4 was distinct from others in the clade, distinguished by the deep reddish brown
(2.5YR 3/4) of upperparts and tail, facial area, throat, breast, and flanks; center of belly was slightly paler buffy brown
(7.5YR 6/6–6/8) with hints of a faint scaled appearance. Specimens of rufula 5 and rufula 2 populations varied in color-
ation, although bellies and undertail coverts of all populations were palter and yellower than breasts, whose coloration
shaded to upperparts color in malar area. Most similar to coloration of rufula 4 were two specimens of rufula 2c from Cauca
(PNN Puracé and Laguna San Rafael) and specimens of rufula 5 from northern Ecuador, which were reddish-brown (up-
perparts 5YR 3/4–4/6, underparts 5YR 4/8). Two southern Ecuador specimens of rufula 5 were lighter and more yellowish
(upperparts 5YR 4/6–4/4, underparts 5YR 5/8) than more northern specimens. Upperparts of two other specimens of rufula
2c from Cauca were brightly colored, yellowish red-brown (closer to 5YR 4/6), whereas two specimens from Quindío were
lighter and browner (closer to 5YR 4/4). Underparts of these rufula 2c specimens were yellowish red-brown (5YR 4/8) but
intensity varied in concert with upperparts; lower belly had some color of undertail-coverts mixed in; undertail coverts were
light reddish yellow-brown (7.5YR 7/6) in Cauca but pale yellowish-buff (2.5Y 8/2) in Quindío. Two specimens of rufula
2b were less rufous especially in the underparts (7.5YR 5/8).
Cerro Munchique, Cauca.—A single recording from Cerro Munchique (slope unknown), in the Western Andes, contained a
GRALLARIA RUFULA COMPLEX Zootaxa 4817 (1) © 2020 Magnolia Press · 39
short song that was indistinguishable from short songs of rufula 2c. A series of poorly recorded notes in the same record-
ing may have been delivered by this individual, although note shapes were unlike long songs of rufula 2 and pace of notes
unlike long songs of rufula 4. With regard to plumage, two specimens from Cerro Munchique were similar to rufula 4
specimens from the northern Western Andes although their coloration was even richer and darker. Conceivably, two rufula
species could occur in this region, and specimens may have been obtained on the west slope and the recordings on the east
slope, or at different elevations on the same slope. Taxonomic placement of Cerro Munchique birds, along with populations
that exist in PNN Farallones and other locations along the Western Andes between known populations of rufula 4 and its
neighbors, awaits additional specimens and recordings, particularly from the region of western Valle del Cauca (e.g., Cerro
Calima) south through western Nariño, and from Nevado del Huila and PNN Puracé south through northern Ecuador.
Analysis of mtDNA.—Uncorrected pairwise genetic distance between rufula 4 and other populations in Principal Clade C was
7.5–8%. Uncorrected pairwise distance between rufula 5 and all rufula 2 populations was 4%; between an individual of
rufula 2b and individuals of the other two rufula 2 populations 1.5%; and between individuals of rufula 2a and rufula 2c
less than 1%.
Taxonomic conclusions.—According to our criteria, rufula 4 is a distinct species. Its long song notes are box-shaped and de-
cline in peak frequency over the course of the song, whereas long song notes in other Principal Clade C populations are
downslurred and note peaks remain essentially flat. Short song notes of rufula 4 are rounded or box-shaped, whereas those
of other Principal Clade C populations are flat or slightly downslurred. The duration of the major interval in short songs
of rufula 4 differs from those of rufula 5 although not from those of populations of rufula 2. Designation of rufula 4 as a
species is supported by diagnosable plumage differences and a high level of genetic divergence with other Principal Clade
C populations. Populations rufula 2 and rufula 5 are maintained in a single species according to our criteria. A sufficient
number of long song examples is available only for rufula 5, and the single long song recorded for rufula 2c is similar to
songs of rufula 5. The only vocal character consistently distinguishing short songs of rufula 5 and rufula 2 is an interval
duration within the terminal phrase of rufula 5 short songs. Pairwise mtDNA distances among populations of rufula 2 are
relatively small, although the difference between rufula 5 and rufula 2 is substantial (4%). In general, available data for
populations of rufula 2 are insufficient, and dense geographic sampling of vocal recordings and specimens are needed
to consider the possibility of differentiating populations taxonomically within this species. Regarding nomenclature, the
Tungurahua, Ecuador, type locality of a currently synonymized subspecies, G. rufula saturata (Domaniewski & Stolzmann
1918), places it in the range of rufula 5.
Principal Clade D. Consists of the study population cajamarcae.
Vocal characteristics of the clade.—Long song is a trill. Short song is a widely spaced series of individual notes.
Analysis of long song.—Long song (Fig. 8) was a distinct trill in which the initial notes lengthened in duration and peaks of the
final notes rose in frequency. Length and pace (Mean ± SD, Range): 18±6 (12–32) notes at 9.0±1.4 (7.1–11.5) notes/sec (n
= 11).
Analysis of short song.—Short song (Fig. 8) was a slow-paced (~2 notes/sec but somewhat variable) series of typically 5–6
(range 3–7) similarly shaped notes that were delivered at about the same frequency. Individual notes were either slightly
downslurred or flat, and they typically lengthened and then shortened in duration.
Analysis of Plumage.—Upperparts were dark reddish yellow-brown (7.5 YR 4/4–5/4) with a pale eye-ring; underparts were light
reddish yellow-brown (5YR 7/10–6/10) with pale feather edges, and belly was extensively whitish.
Analysis of mtDNA.—Uncorrected pairwise genetic distance between cajamarcae and other populations was 5.5–11.5%.
Taxonomic conclusions.—Under our criteria, cajamarcae is a distinct species. The structure of its short song is unique among
the populations of the G. rufula complex. Its trilled long song is distinguished by a combination of a change in pace and a
change in frequency. Species status is supported by plumage distinctions (yellowish-brown coloration, pale eye-ring, and
whitish belly) and by genetic divergence.
Principal Clade E. Consists of the study population blakei 1.
Vocal characteristics of the clade.—Long song is a trill. No short song was found in 36 recordings from across the geographic
range of the population.
Analysis of long song.—Long song (Fig. 9) was a lengthy trill, delivered rapidly. The similar chevron-shaped notes were deliv-
ered at an even pace except for 2–3 initial notes in which note intervals started longer and shortened in duration. Length and
pace (Mean ± SD, Range): 38±13 (23–73) notes delivered at 11.0±1.0 (8.8–12.5) notes/sec (n = 20).
Analysis of short song.—No short song has been recorded for blakei 1, although one or more notes of the long song were deliv-
ISLER ET AL.
40 · Zootaxa 4817 (1) © 2020 Magnolia Press
ered occasionally in inconsistent patterns.
Analysis of Plumage.—Upperparts were dusky-reddish-brown (2.5YR 3/4). Underparts were yellowish red-brown (5YR 4/8–
5/8) becoming barred in the dirty white center of belly.
Analysis of mtDNA.—Uncorrected pairwise genetic distance between blakei 1 and other populations was 5.5–12%.
Taxonomic conclusions.—Under our criteria, blakei 1 is a distinct species consistent in geographic range and plumage descrip-
tion with G. blakei (Graves, 1987). Species status is substantiated by its vocalizations, especially the lack of a short song,
unique in the G. rufula complex, and by the lengthy and evenly paced trill of its long song, and is supported by genetic
divergence.
Principal Clade F. Consists of study populations obscura 1, obscura 2, and obscura 3. Five recordings of four individuals ob-
tained by D. Lane at the northern tip of the range of obscura 1 at Cerro Huicsacunga, Amazonas, Peru, were not included
in the analysis pending acquisition of more data from this region. Vocal measurements of these recordings fell outside the
ranges obtained for obscura 1 (pace of long song was faster and terminal note duration in short songs more abrupt). No
specimens were collected.
FIGURE A7. Long songs of Principal Clade F populations: (A) obscura 1 Bosque Unchog, Huánuco, Peru (Valqui ML236895);
(B) obscura 2, Millpo, Pasco, Peru (Schulenberg ML40152); (C) obscura 3, Toldopampa, Junín, Peru (Hosner ML171865).
TABLE A7. Selected measurements of long songs of populations of Principal Clade F, presented as mean±SD (range).
Population obscura 1 obscura 2 obscura 3
n7 6 5
Pace (notes/sec)14.70 (3.42–6.67) 8.84 (6.67–10.68) 3.07 (2.68–3.39)
Change of pace in second half of song20.721 (0.595–0.873) 0.530 (0.364–0.632) 1.179 (1.085–1.289)
First interval duration (sec)30.079 (0.068–0.088) 0.050 (0.042–0.056) 0.061 (0.049–0.068)
Middle interval duration (sec)30.086 (0.055–0.113) 0.053 (0.044–0.069) 0.095 (0.083–0.115)
Penultimate interval duration (sec)30.068 (0.042–0.096) 0.042 (0.037–0.049) 0.148 (0.110–0.179)
Change in frequency from initial to terminal
notes (Hz)4
-381±285 [-775–(-)86] -331±115 [-431–(-)172] 180±70 (86–282)
1Differences significant between all pairs.
2Computed as duration penultimate note and terminal interval divided by duration middle note and interval; obscura 3
differed significantly from obscura 1 and obscura 2.
3Pattern of change of interval duration differed significantly between all pairs (see text).
GRALLARIA RUFULA COMPLEX Zootaxa 4817 (1) © 2020 Magnolia Press · 41
4 Computed as maximum frequency first note minus maximum frequency penultimate note. Maximum frequency is the
lowest frequency in a note at which maximum power occurs. Difference between obscura 2 and obscura 3 was signifi-
cant.
Vocal characteristics of the clade.—Long songs are a series of frequency-modulated notes. Short song is an abrupt note fol-
lowed by a long frequency-modulated note.
Analysis of long songs.—Pace of long songs of all three obscura populations (Fig. A7, Table A7) differed diagnosably as inter-
vals between notes of obscura 3 increased throughout the song, those of obscura 1 increased and then decreased, and those
of obscura 2 remained nearly constant. Remaining vocal differences only distinguish obscura 3. The pace of notes slowed
down in the second half of songs of obscura 3, whereas that of other populations speeded up. The peak frequency of notes
of songs of obscura 3 descended terminally, whereas peak frequencies in songs of other populations rose, especially in the
second half of the song, although only the mensural difference with obscura 2 was significant (Table A7).
FIGURE A8. Short songs of Principal Clade F populations: (A) obscura 1, 34 km ENE of Huánuco, Huánuco, Peru (Cáceres
ML168656); (B) obscura 2, PN Yanachaga-Chemillén, Pasco, Peru (Vellinga XC62929); (C) obscura 3, Satipo Road, Junín,
Peru (Spencer XC47028).
ISLER ET AL.
42 · Zootaxa 4817 (1) © 2020 Magnolia Press
Analysis of short songs.—Short songs of all populations (Fig. A8) included two frequency-modulated notes, the first one abrupt.
Although durations of notes and intervals did not differ among the three populations (Table A8), the initial note in short
songs of obscura 3 consisted of three typically V-shaped elements (rarely two) whereas initial notes were limited to a single
such element in other populations. Differences in frequency band width suggested by the illustration were not significant
because band width was variable within populations.
TABLE A8. Selected measurements of short songs of populations of Principal Clade F, presented as mean±SD (range).
nDuration initial note (sec) Duration interval (sec) Duration final note (sec)
obscura 1 21 0.03±0.01 (0.001–0.06) 0.17±0.03 (0.05–0.23) 0.65±0.12 (0.42–0.84)
obscura 2 30.03±0.02 (0.002–0.05) 0.14±0.01 (0.13–0.15) 0.77±0.09 (0.67–0.83)
obscura 3 80.06±0.01 (0.005–0.06) 0.17±0.01 (0.16–0.20) 0.74±0.11 (0.54–0.86)
Analysis of Plumage.—The throat and breast of obscura 2 were substantially paler (7.5YR 6/8–7/8) than those of obscura 1
and obscura 3 (7.5YR 5/8 although redder tending towards 5YR 5/8 in obscura 3). Although upperparts of all obscura
populations were close to reddish-brown (7.5YR 4/4) those of obscura 2 were slightly lighter and less reddish and those of
obscura 3 were darker. In blind tests six specimens of obscura 2 were accurately distinguished from six specimens of each
of the other two populations. In a blind test comparing obscura 1 and obscura 3 five of six specimens of each were sorted
correctly.
Analysis of mtDNA.—Uncorrected pairwise distance between obscura 1 and obscura 2 was 5%; between obscura 1 and obscura
3 was 4%, and between obscura 2 and obscura 3 was 3%.
Taxonomic conclusions.—Under our criteria, the three populations are considered specifically distinct based on multiple vocal
characters that distinguish obscura 3 from obscura 1 and obscura 2, even though fewer measured vocal differences between
obscura 1 and obscura 2 are known. Plumage differences between geographically adjacent populations provide support,
and the pair with the least mtDNA genetic distance between them (obscura 2 and obscura 3) was distinguished by the great-
est number of vocal characters. The type locality of G. r. obscura (Maraynioc, Junín) is located in the range of obscura 3.
Principal Clade G. Consists of study populations blakei 2 and blakei 3.
Vocal characteristics of the clade.—Long songs are a trill. Short songs consist of a regular repetition of a single note that was
not frequency modulated.
TABLE A9. Selected measurements of long songs of populations of Principal Clade G, presented as mean±SD (range).
Population nPace (notes/sec) Duration terminal note/ middle note
blakei 2 58.20±0.63 (7.61–9.10) 1.06±0.03 (1.03–1.10)
blakei 3 512.98±1.38 (11.36–14.48) 0.59±0.06 (0.52–0.66)
FIGURE A9. Long songs of Principal Clade G populations: (A) blakei 2, Apalla-Andamarca Road, Junín, Peru (Athanas
XC148512); (B) blakei 3, Chupón, Ayacucho, Peru (Hosner ML186919).
Analysis of long songs.—Note duration in long songs (Fig. A9) remained constant or increased slightly in the terminal half of
songs of blakei 2 but the duration of equivalent notes in songs of blakei 3 decreased (Table A9). The overall pace of notes
GRALLARIA RUFULA COMPLEX Zootaxa 4817 (1) © 2020 Magnolia Press · 43
in songs of blakei 2 was slower and did not overlap the pace in songs of faster blakei 3 (Table A9), a character likely to
prove diagnostic with larger samples. Individual notes of the trills were downslurred. Their peak frequency slowly declined
initially and then typically flattened out. Initial notes shortened in duration, and terminal intervals lengthened. The last three
characteristics did not differ between blakei 2 and blakei 3, but are provided for comparison with those of other popula-
tions.
Analysis of short songs.—Short songs consisted of the repetition of a single note. Note shape in songs of blakei 2 and blakei 3
differed diagnostically (Fig. A10). The duration of notes of blakei 2 was longer, although the difference did not meet the
significance test (Table A10).
TABLE A10. Selected characteristics of short songs of populations of Principal Clade G, presented as mean±SD
(range).
Population nNote shape Note Duration (sec)
blakei 2 9 Inverted U, tending to a block; once split 0.157±0.010 (0.143–0.172)
blakei 3 10 Downslurred, once split 0.128±0.009 (0.115–0.142)
FIGURE A10. Short songs of Principal Clade G populations: (A) blakei 2, Río Satipo, Junín, Peru (Hosner ML171921); (B)
blakei 3, Chupón, Ayacucho, Peru (Robbins ML173831).
Analysis of Plumage.—Coloration of blakei 2 was darker and redder (e.g., upperparts 5YR 3/4) than those of blakei 3 (up-
perparts 7.5YR 4/4). Underparts of both populations had pale feather edgings to lower underpart feathers that provided a
scalloped appearance that was less obvious in blakei 2. Specimens were separable in blind tests. Both blakei 2 and blakei 3
lacked the pale eye ring of Principal Clade H populations.
Analysis of mtDNA.—Uncorrected pairwise distance between blakei 2 and blakei 3 was 4%.
Taxonomic conclusions.—Given diagnosable differences in both types of vocalizations, blakei 2 and blakei 3 are considered
specifically distinct under our criteria. Two vocal characters (change in pattern of duration of long song notes and short
song note shape) distinguish blakei 2 and blakei 3. In addition, ranges of measurements of long song pace and short song
note duration do not overlap although they do not meet probability tests. Consideration as distinct species is supported by
differences in plumage coloration and mtDNA genetic distance. Differences in their elevational ranges are also notable,
suggesting a different evolutionary trajectory or perhaps competitive release.
Principal Clade H. Consists of study populations occabambae 1a, occabambae 1b, cochabambae 1, and cochabambae 2.
Vocal characteristics of the clade.—Long songs are trills. Short songs consist of two or three notes depending on the population.
Notes are not frequency modulated.
Analysis of long songs.—Individual notes in long songs of occabambae 1a and 1b (Fig. A11) were shaped like an inverted U,
differing from those of cochabambae 1 (box-shaped, slightly downslurred) and cochabambae 2 (sharply downslurred).
Peaks of notes in songs of cochabambae 1 and cochabambae 2 decreased, flattened, and increased, a pattern that differed
from the flat or slowly declining frequency of notes of occabambae 1a and 1b. Change in peak frequency of notes over the
last half of the song differed diagnostically between cochabambae 1 and 2 (Table A11). Note and interval duration patterns
of the two populations of occabambae differed from those of cochabambae 1 and 2. Pace differed among all four popula-
tions although not significantly between cochabambae 1 and occabambae 1a, and with a slight overlap of range between
occabambae populations.
ISLER ET AL.
44 · Zootaxa 4817 (1) © 2020 Magnolia Press
FIGURE A11. Long songs of Principal Clade H populations: (A) occabambae 1a, Carrizales, Cusco, Peru (Lane ML236893);
(B) occabambae 1b, Abra Acjanaco, Cusco (Lane ML236892); (C) cochabambae 1 Río Keara, PN Madidi, La Paz, Bolivia
(Tobias XC73603); (D) cochabambae 2 Chuspipata, La Paz (Rowlett ML236894).
Analysis of short songs.—Inverted U-shaped notes of populations of occabambae differed from downslurred or flat notes of
populations of cochabambae (Fig. A12). Short songs of occabambae 1a and occabambae 1b differed in number of notes,
but otherwise they were similar. The duration of the interval between the first two notes differed significantly between the
two populations of cochabambae and between each of them and the two populations of occabambae (Table A12). The
duration of the second note of cochabambae 1 differed significantly from the longer notes of cochabambae 2 and those of
both populations of occabambae.
Analysis of Plumage.—Populations in this clade had pale feather edgings to lower underpart feathers (on belly, extending to
breast in some populations) that provided a scalloped appearance. Although occabambae 1a and occabambae 1b were
GRALLARIA RUFULA COMPLEX Zootaxa 4817 (1) © 2020 Magnolia Press · 45
similar in coloration, undertail coverts differed between occabambae 1a (yellower and paler, 2.5Y 8/4) and occabambae
1b (darker and redder, 7.5R 7/8) in available specimens. Underparts of cochabambae 1 and cochabambae 2 were the palest
in this clade, and specimens of each were separable in blind tests from populations of occabambae and from each other.
Notably, the breast color of cochabambae 2 was darker (7.5YR 5/8–5/6) and more uniform than the breast coloration
(7.5YR 6/8–6/10) of cochabambae 1, which had pale feather edges and pencil streaks on feather shafts extending up from
the belly.
TABLE A11. Selected characteristics and measurements of long songs of populations of Principal Clade H, presented as
mean±SD (range).
Population nFrequency pattern Note and interval duration pattern Pace
(notes/sec)
Frequency peaks
middle note minus
terminal note (Hz)
occabambae 1a 8 Typically dropped
slightly throughout;
occasionally flat
Intervals lengthened resulting in
decelerating pace (notes constant or
shortened slightly)
4.17
(3.62–5.09)
90±54
(0–156)
occabambae 1b 5Same as occabambae
1a
Same as occabambae 1a (except
notes sometimes lengthened)
3.38
(2.93–3.68)
59±55
(0–109)
cochabambae 1 2 Peaks decreased, sta-
bilized, and increased
Both lengthened, became even, and
shortened
5.63
(5.51–5.74)
-177±30
[-198–(-)155]
cochabambae 214Same as cochabam-
bae 1
Same as cochabambae 1; except
intervals lengthened in one example
10.31
(9.34–10.94)
-580±738,
[-1687–(-)211]
1An apparently aberrant song having similar frequency, note, and interval patterns but with notes V-shaped not included.
Analysis of mtDNA.—Uncorrected genetic distance between the two populations of cochabambae and the two populations of
occabambae was 7–7.5%. Uncorrected genetic distance between occabambae 1a and occabambae 1b was 1%; and between
cochabambae 1 and cochabambae 2 was 5.5%.
FIGURE A12. Short songs of Principal Clade H populations: (A) occabambae 1a, Choquequirao, Cusco, Peru (Krabbe
XC47645); (B) occabambae 1b, upper Manu Road, Cusco (Lane ML236891); (C) cochabambae 1 Río Keara, Madidi National
Park, La Paz, Bolivia (Tobias XC73602); (D) cochabambae 2 Corioco Road, La Paz (Boesman XC223701).
ISLER ET AL.
46 · Zootaxa 4817 (1) © 2020 Magnolia Press
TABLE A12. Selected characteristics and measurements of short songs of populations of Principal Clade H, presented
as mean±SD (range).
Population nNote shape Duration (sec)
Note 1 Interval 1 Note 2 Interval 2 Note 3
occabambae 1a 14 Inverted U 0.110±0.014
(0.088–0.138)
0.245±0.020
(0.212–0.269)
0.156±0.015
(0.124–0.179)
occabambae 1b 11 Inverted U 0.115±0.014
(0.097–0.137)
0.312±0.021
(0.278–0.344)
0.147±0.018
(0.125–0.180)
0.110±0.018
(0.085–0.137)
0.102±0.012
(0.086–0.120)
cochabambae 1 3 Downslur
or flat and
downslurred
0.067±0.008
(0.058–0.074)
0.070±0.013
(0.056–0.081)
0.059±0.004
(0.056–0.063)
cochabambae 2 9 Downslur
or flat and
downslurred
0.068±0.015
(0.046–0.082)
0.139±0.014
(0.120–0.163)
0.155±0.014
(0.133–0.172)
Taxonomic conclusions.—Five or six vocal characters distinguished both populations of occabambae from both populations
of cochabambae, placing them in different species in accordance with our criteria. Vocalizations of the two occabambae
populations differed only in the number of notes in short songs. According to our criteria, species status was not indicated,
but subspecific status is appropriate given consistency in number of notes in short song in the large samples examined and is
supported by apparent differences in undertail covert coloration. Although few examples of songs of cochabambae 1 were
available, five vocal characters distinguished cochabambae 1 and cochabambae 2. Under our criteria, we consider them to
be specifically distinct, a conclusion supported by differences in plumage coloration and by deep genetic divergence. The
type locality of G. r. occabambae is within the geographic range of occabambae 1a, and the type locality of G. r. cochabam-
bae is within the geographic range of cochabambae 2.
APPENDIX 2. Summary of Vocal Comparisons of Species in Different Clades. The foregoing analyses
provided the basis for our recommendation that populations in the Grallaria rufula complex be considered to
consist of the 16 species identified in Table A13. These recommendations were based on comparisons within
principal clades. Despite the substantial genetic distances between the principal clades (Chesser et al. 2020),
the question remains whether recommended species in different clades are vocally distinct. The simplest way
to answer this question is to compare vocal structure. If vocal structure differs in its entirety, a sufficient num-
ber of vocal characters will meet our yardstick for species recommendations. Table A13 provides descriptions
of vocal structure to afford comparisons of species in different clades. Comparing pairs of recommended spe-
cies located in different principal clades, all pairs differed in the structure of their long songs, short songs, or
both.
TABLE A13. Vocal structure of long and short songs of Grallaria species.
Principal
Clade
Grallaria species Populations Long song structure Short song structure
1G. saltuensis saltuensis Trill Trill
G. spatiator spatiator Trill Trill
2G. rufula rufula 1, rufula 3 Note series Two long notes, second a trill
G. rufocinerea rufocinerea,
romeroana
Note series Single, long unmodulated note
3G. alvarezi rufula 4 Trill Note, interval, trill
G. saturata rufula 2a, rufula 2b,
rufula 2c, rufula 5
Trill Note, interval, trill
4G. cajamarcae cajamarcae Trill Note series
5G. blakei blakei 1 Trill None
...Continued on the next page
GRALLARIA RUFULA COMPLEX Zootaxa 4817 (1) © 2020 Magnolia Press · 47
TABLE A13. (Continued)
Principal
Clade
Grallaria species Populations Long song structure Short song structure
6G. gravesi obscura 1 Frequency modulated series Two frequency modulated notes,
first shorter
G. oneilli obscura 2 Frequency modulated series Two frequency modulated notes,
first shorter
G. obscura obscura 3 Frequency modulated series Two frequency modulated notes,
first shorter
7G. centralis blakei 2 Trill Single unmodulated note
G. ayacuchensis blakei 3 Trill Single unmodulated note
8G. occabambae occabambae 1a,
occabambae 1b
Trill Two or three unmodulated notes
G. sinaensis cochabambae 1 Trill Two unmodulated notes
G. cochabambae cochabambae 2 Trill Two unmodulated notes
ISLER ET AL.
48 · Zootaxa 4817 (1) © 2020 Magnolia Press
APPENDIX 3. Genetic Distance Matrix for Study Populations. The following genetic distance matrix (Table A14) was computed in an analysis reported
in the companion paper (Chesser et al.2020).
TABLE A14. Uncorrected distance matrix for mtDNA for study populations. Percentages rounded to nearest 0.5%. No shading in a cell identifies pairs placed in different
species. Gray shading identifies pairs whose vocal and supporting differences are insufficient to recommend placing them in different species. Each of these gray shaded cells
is discussed in the text.
Clade 1 2 3 4 5 6 7 8
Pop. spat ru1 ru3 rufo ru4 ru2a ru2b ru2c ru5 caja bla1 obs1 obs2 obs3 bla2 bla3 oc1a oc1b coc1 coc2
salt 7.5 9.5 8.5 9.5 9.5 10.0 10.0 10.0 10.5 11.0 11.5 12.0 12.5 12.0 9.5 9.5 10.0 10.0 11.5 11.5
spat 9.0 9.0 9.0 11.0 10.0 9.5 10.0 9.5 10.0 10.5 12.0 12.0 12.0 11.0 10.5 9.0 8.5 11.0 10.5
ru1 5.5 6.5 11.5 11.0 11.5 11.0 11.5 11.5 12.0 12.0 12.5 12.0 11.0 11.0 10.0 10.0 12.5 12.5
ru3 5.5 10.5 10.5 10.5 10.5 10.0 10.0 11.0 12.0 12.0 11.5 10.5 10.0 10.0 10.0 11.5 12.0
rufo 12.2 10.5 10.0 10.5 10.6 11.6 11.4 12.5 13.0 12.3 10.4 10.7 10.4 10.5 12.2 13.0
ru4 7.5 7.5 7.5 8.0 7.0 9.0 9.0 9.5 9.0 9.0 8.5 8.0 7.5 9.5 9.5
ru2a 1.0 1.5 4.0 5.5 6.0 6.5 7.5 7.0 8.0 7.0 6.0 8.5 9.0 9.0
ru2b 1.5 4.0 5.5 5.5 6.5 8.0 7.0 7.5 7.0 7.0 8.5 8.5 8.5
ru2c 4.0 5.5 6.0 6.5 7.5 7.0 8.0 7.0 6.5 8.0 8.5 8.5
ru5 6.5 6.5 7.0 7.5 7.0 8.5 7.5 7.0 8.5 9.0 9.0
caja 6.5 7.0 8.0 7.5 8.0 7.5 7.5 7.0 9.0 9.0
bla1 7.0 8.0 7.0 9.0 8.0 7.5 7.5 10.0 9.5
obs1 5.0 4.0 9.5 8.5 8.0 8.0 10.0 9.5
obs2 3.0 10.5 9.5 9.0 9.0 10.5 10.5
obs3 10.0 9.0 8.5 8.5 10.0 10.0
bla2 4.0 8.0 8.0 9.0 9.0
bla3 6.0 6.0 7.5 8.5
oc1a 1.0 7.5 7.0
oc1b 7.5 7.5
coc1 5.5
GRALLARIA RUFULA COMPLEX Zootaxa 4817 (1) © 2020 Magnolia Press · 49
APPENDIX 4. Recordings Examined. The following list identifies recordings used in the study by principal
clade, study population, country, state or department, recording location, and recordist. Numbers following
the recordist name identify the number of individuals recorded per archive. Acronyms for recording archives:
CBN = Colección Bioacústica Nariño, Colombia; FSM = Florida State Museum, Gainesville; IAvH-CSA
= Colección de Sonidos Ambientales, Instituto Alexander von Humboldt, Villa de Leyva, Colombia; ML =
Macaulay Library, Cornell Laboratory of Ornithology, Ithaca, New York; XC = xeno-canto database (www.
xeno-canto.org). ISL = recordings not yet archived in an institutional collection but that have been copied into
the inventory maintained by MLI.
Principal Clade A
saltuensis (11 individuals; 3 localities). Colombia: Cesar: above El Cinco (Cuervo 2 XC, 1 ML, 3 ISL, Gallardy 2 XC, Marín 1
ISL), Páramo de Sabana Rubia (Schmitt 1 XC), RNA Chamicero del Perijá (Borgmann 3 ML).
spatiator (23 individuals; 2 localities). Colombia: Magdalena: Cuchilla San Lorenzo (Athanas 1 XC, Boesman 1 ISL, Borgmann
1 ML, Coopmans 1 ML, 3 ISL, Flórez 1 XC, Kirwan 1 XC, Krabbe 6 XC, 1 ISL, Lambert 1 XC, López-Lanús 2 XC, Sarver
1 ML, Schmitt 2 XC), Santa Marta-Lorenzo Road (Boesman 1 XC).
Principal Clade B
rufula 1 (9 individuals; 3 localities). Colombia: Norte de Santander: PNN Tamá (Álvarez 3 IAvH-CSA, Socolar 3 ISL), above
Asiria-Belén (Cuervo 2 ISL); Santander: Finca Contreras (López-Lanús 1 XC).
rufula 1 or rufula 3 (18 individuals; 7 localities). Colombia: Boyacá: Páramo de Bijagual (Macana 1 ISL), RN El Malmo (Zu-
luaga 7 ISL), Páramo de Mamapacha (Macana 1 ISL, 1 XC), RF Nacimiento del Río Boyacá (Cubillos 1 XC), Páramo de
Rabanal (Macana 5 ISL), Páramo de Siscunsi (Macana 1 XC), Soatá (Cortés 2 XC).
rufula 3 (64 individuals; 9 localities). Colombia: location unknown (Stiles 4 FSM); Cundinamarca: BioAndina Reserve (Brad-
ley 2 XC, Calderón-F. 1 XC, Pinto-Gómez 3 XC), Bogotá (Cortes 1 XC, Lopez-Lanus 1 XC), Finca Cárpatos (Chaparro 1
ISL), PN Chingasa (Calderón-F. 1 XC, Coopmans 1 ISL, Cordoba 2 IAvH-CSA, Kaestner 2 ISL, Laverde 5 XC, Schmitt 3
XC, Whitney 4 ISL), RN Encenillo (Acevedo 1 ISL), Páramo de Guasca (Stiles 2 IAvH-CSA), Finca La Calera (Córdoba
12 IAvH-CSA, Laverde 7 IAvH-CSA), Quebrada La Vieja (Athanas 1 XC, Cadena 4 IAvH-CSA, Knapp 2 XC), Loma La
Aurora (Coopmans 3 ISL); Meta: El Calvario (Cuervo 2 ISL).
rufocinerea: (34 individuals; 7 localities). Colombia: Antioquia: Páramo de Sonsón (Florez 1 XC); Caldas: RN Río Blanco
(Boesman 4 XC, Bradley 1 XC, Calderón-F. 1 XC, Cortés 2 XC, Culasso 1 XC, Davis 1 XC, Geale 1 XC, Kaestner 1 ISL,
King 2 XC, Kirwan 2 XC, Lambert 3 XC, van Oosten 1 XC); Quindío: RM EL Mirador (Athanas 2 XC); Riseralda: Santa
Rosa de Cabal (Holmes 1 XC), Santuario Otún-Quimbaya (López-Lanus 1 XC); Tolima: Río Toche (Dreyer 1 XC, López-
Lanús 5 XC); Valle de Cauca: Tenerife (Baldwin 1 XC).
romeroana: (9 individuals; 2 localities). Colombia: Nariño: Laguna La Cocha (Hilty 1 ISL). Ecuador: Sucumbíos: Santa Bárbara
(Ahlman 2 XC, Gelis 1 XC, Nilsson 3 XC, Spencer 1 XC, Vogt 1 XC).
Principal Clade C
rufula 2a (10 individuals; 3 localities). Colombia: Boyacá: SFF Iguaque (Caycedo 1 ML, Calderón-F. 2 XC, Córdoba 3 IAvH-
CSA), Morro Negro (Bravo 1 ISL, López 1 IAvH-CSA), Sotáquira (Caguazango 2 ISL).
rufula 2b (14 individuals; 8 localities). Colombia: Caldas: Finca La Martinica (Calderón-F. 4 ISL, Cuervo 2 IAvH-CSA), PNN
Los Nevados (St-Michel 1 XC), above Manizales (Minns 1 XC), RN Río Blanco (Laverde 1 IAvH-CSA); Caldas/Tolima:
Nevado del Ruiz (Cortés XC); Tolima: Cajamarca (Marín-G. 1 XC), Roncesvalles (Krabbe 1 XC), Tapias (Kaestner 2
ISL).
rufula 2b or rufula 2c (2 individuals; 1 locality). Colombia: Huila: Laguna del Buey (Muñoz 2 XC).
rufula 2c: (4 individuals; 1 locality). Colombia: Cauca: PN Puracé (Calderón-F. 1 XC, Coopmans 1 ML, Whitney, 1 ML, 1
ISL).
rufula 4 (46 individuals; 10 localities). Colombia: Antioquia: Farallones de Citará (Pulgarin 1 ISL), RNA Colibri del Sol (King
2 ISL), Eme (Calderón-F. 1 XC; Schmitt 1 XC), Páramo Frontino (Boesman 1 XC, Krabbe 2 XC, 6 ISL), north of Jardín
(Calderón-F. 1 ML, 2 ISL, Certuche 4 ISL, Guarín 1 XC, Lambert 3 ISL), RNA Loro Orejiamarillo (Lambert 2 XC, Sarver
1 ML, Schmitt 1 XC), Montezuma Road Peak (Calderón-F. 1 ISL, 1 XC), Cerro Plateado (Chaparro 2 ISL), Urrao (King 1
XC, Schmitt 1 XC), Alto de Ventanas (Calderón-F. 2 ISL, 6 XC, Guarín 3 XC); Cauca: Cerro Munchique, slope unknown
ISLER ET AL.
50 · Zootaxa 4817 (1) © 2020 Magnolia Press
(Coopmans 1 ML).
rufula 5: (102 individuals; 43 localities). Colombia: Nariño: Casa Fría (Fernandez-G 2 CBN), Pasto (Laverde 1 ISL); Putumayo:
Páramo Bordoncillo (Jaramillo 1 XC). Ecuador: Azuay: Laguna Llaviuco (Woods 1 XC); Carchi: Cerro Mongus (Krabbe 3
XC, 1 ISL, Robbins 3 ML, Vogt 1 ISL); Páramo El Angel (Krabbe 1 ISL); Napo: Río Anatenorio (Krabbe 1 XC), Cosanga
(Whitney 1 ISL), Cordillera de Guacamayos (Krabbe 2 ISL, G. Rosenberg 1 ISL), Oyacachi (Krabbe 2 XC), Papallacta
(Krabbe 1 ISL, Moore 2 XC); Sucumbíos: Santa Bárbara (Krabbe 1 XC); Esmeraldas: ER Cotacachi-Cayapas (Jahn 1 XC);
Imbabura: Loma Taminanga (Krabbe 1 XC); Pichincha: Volcán Atacazo (Krabbe 4 ISL), west of Corazón (Krabbe 1 ISL,
2 XC), Nono (Kerr 1 ML, Macaulay 1 ML), Volcán Pichincha (Krabbe 3 ISL, 1 XC, Moore 3 XC, Nilsson 1 XC, 1 XC),
Laguna San Marcos (Ordónez D. 1 XC), Yanacocha (Ahlman 1 XC, Athanas 2 XC, Chartier 1 XC, DeFonso 1 XC, King
5 XC, Krabbe 1 XC, Lachmann 1 XC, Leite 1 XC); Cotopaxi: Cerro Parcato (Krabbe 1 ISL, 2 XC); Tungurahua: Cordil-
lera de los Llaganates (Krabbe 1 XC), Volcán Tungurahua (Krabbe 1 ISL, 1 XC); Bolívar: 10 km west of Salinas (Krabbe
1 XC); Cañar: La Libertad (Planque 1 XC); Morona-Santiago: Coangos (Parker 1 ISL), B.P.C. Jima (Ordónez-Delgado 2
XC), Zapote-Najda Mountains (Krabbe 2 ISL); Azuay: Azuay (Aguilar 1 XC), Río Mazán valley (Krabbe 2 XC), Molleturo
(Krabbe 1 ISL, 1 XC); Loja: Acanama (Krabbe 1 ISL, 1 XC, Spencer 1 XC), Cajanuma Valley (Ahlman 1 XC, Nilsson 1
XC, Krabbe 1 ISL, 1 XC), Lagunas de Manú (Ordónez-D. 1 XC), Saraguro (Robbins 1 ML); Zamora-Chinchipe: 25 road
km SSE Jimbura (Robbins 1 ML), PN Podocarpus (Boesman 2 ISL, Whitney 1 ISL, 3 ML), Tapichalaca Reserve (Halfwerk
1 XC, Haribal 1 XC, Krabbe 1 XC, Lambert 3 XC, Winger 1 ML, Zyskowski 1 ML). Peru: Cajamarca: El Espino (Lane 2
ML), Hito Jesus (Mark 1 XC); Piura: Cerro Chinguela (Parker 2 ML, Whitney 1 ISL).
Principal Clade D
cajamarcae (68 individuals; 14 localities). Peru: Piura: Canchaque (Engblom 1 ISL), Cruz Blanca (Lane 1 XC), 33 km SW
Huancabamba (O’Neill 1 ML, Parker 5 ML), Huarmaca (Geale 1 XC); Cajamarca: Balsas-Celendín Road (Zimmer 3 ISL),
Cajamarca (van Oosten 1 XC), 35–50 km E Cajamarca to Celendín (Ahlman 1 XC, Athanas 1 XC, Innes 1 ISL, Zimmer 1
ISL, 1 ML), Celendín (Boesman 3 XC, 2 ISL, Lambert 1 XC, Nilsson 1 ISL, Rowlett 5 ISL, Webster 2 ISL), Chalhuayacu
(Diaz 1 XC), Cruz Conga (Ahlman 1 ZC, Beadle 1 XC, Boesman 1 XC, DeFonso 1 XC, King 2 XC, Lambert 1 XC, Lloyd 3
ISL, Matheve 1 XC, Nelson 1 XC, Schmitt 2 XC, Valqui 1 ISL), Bosque Pagaibamba (Vellinga 7 XC), Bosque Paja Blanca
(Lane 3 ISL, 1 ML, Planque 6 XC, Vellinga 4 XC), Sinchahuelito (Mark 1 XC).
Principal Clade E
blakei 1 (35 individuals; 7 localities). Peru: Amazonas: Leimebamba (Winger 2 ML); Amazonas/San Martín: Pardo de Miguel
or Abra Patricia (Athanas 1 XC, Beadle 1 XC, Geale 1 XC, Innes 1 ISL, King 1 XC, Lambert 1 XC, Matheve 1 XC, Whit-
ney 2 ISL, Zimmer 1 ISL); La Libertad: Cumpang (Parker 4 ML); Huánuco: Cordillera Carpish (Geale 1 XC), Carpish
Tunnel (Boesman 2 XC, Gallardy 1 XC, Krabbe 2 XC, Schmitt 1 XC, Valqui 1 ISL, Whitney 1 ISL), trail to Hacienda Paty
(Parker 1 ML, Schmitt 1 XC, Schulenberg 1 ML, Spencer 1 ML, van den Berg 1 ML, Walker 1 ISL, 1 ML); San Martín:
22–24 km ENE Florida (Lane 3 ISL).
Principal Clade F
obscura 1 (28 individuals; 10 localities). Peru: Amazonas: Abra Barro Negro (Boesman 2 XC, King 1 XC, Lane 1 ISL, Lloyd
1 ISL, van Oosten 1 XC, Zimmer 3 ISL), Leimebamba (Beadle 1 XC, Winger 1 ML), 18–25 km W Leimebamba (Lane 1
ISL), Quintecocha (Vellinga 1 XC); La Libertad: Tayabamba (Parker 1 ML); Huánuco: Bosque Unchog (Boesman 1 ISL,
Krabbe 2 XC, Lees 1 XC, Lloyd 3 ISL, Valqui 1 ISL, 1 ML), Cordillera Carpish (Geale 1 XC), Esperanza (Parker 1 ML),
34 km ENE Huánuco (Cáceres 2 ML), Uchiza Valley (Harvey 1 ML).
obscura 2 (11 individuals; 4 localities). Peru: Huánuco: Bosque Potrero (Isler 2 ML), Monte Potrero (Geale 1 XC); Pasco: Milpo
(Schulenberg 3 ML), PN Yanachaga-Chemillén (Vellinga 5 XC).
obscura 3: (18 individuals; 9 localities). Peru: Junín: Apalla-Andamarca Road (van Oosten 1 XC), Calabaza (Boesman 2 ISL),
Carrizal (Barros/Schmi 1 XC), Carrizales Bridge (Boesman 1 XC), Chilifruta (Athanas 1 XC), Huancayo-Lampo Road
(Lane 1 ISL), Puente Carrizales (Lloyd 2 ISL, Spencer 1 XC), Satipo Road (Hornbuckle 1 XC, Spencer 2 XC), Toldopampa
(Hosner 2 ML, Robbins 3 ML).
Principal Clade G
blakei 2 (14 individuals; 5 localities). Peru: Pasco: Playa Pampa (Schulenberg 1 ML), Cumbre de Ollón (Hornbuckle 2 XC,
Schulenberg 1 ML); Junín: Apalla-Andamarca Road (Athanas 1 XC, Spencer 2 XC), Calabaza (Boesman 3 XC), Río Satipo
GRALLARIA RUFULA COMPLEX Zootaxa 4817 (1) © 2020 Magnolia Press · 51
(Hosner 2 ML, Robbins 2 ML).
blakei 3: (20 individuals; 2 localities). Peru: Ayacucho: Chupón (Hosner 6 ML, Robbins 10 ML), Rumichaca (Hosner 3 ISL, 1
ML).
Principal Clade H
occabambae 1a (41 individuals; 10 localities). Peru: Junín: Cordillera de Vilcabamba Schulenberg Camp 1 (Schulenberg 14
ISL); Cusco: Abra Málaga (Ahlman 1 XC, Begazo 1 ISL, Boesman 2 XC, Donahue 1 ML, Lambert 1 ISL, Lloyd 1 ISL,
Schulenberg 1 ML, R. Yabar 1 ISL, 1 XC), 14–15 km NE Abra Málaga (Krabbe 3 ISL, Lambert 3 XC), Canchaillo (Fjeldså
1 ISL, Parker 3 ML, Walker 1 ISL), Carrizales (Geale 2 XC, Lane 1 ISL, 1 ML, King 1 XC), Choquequirao (R. Yabar 1
ISL), Ollantaytambo-Quillabamba Road (Lane 1 XC), Paccaypata (Hosner 2 ML, Robbins 6 ML), San Luis (Andersen 1
ML, Krabbe 1 XC, Walker 1 ISL), Wiñay Wayna (Cáceres 1 ML).
occabambae 1b (22 individuals; 4 localities). Peru: Cusco: Abra Acjanaco (Lane 1 ML, Schmitt 2 XC, Yabar 1 XC, Zyskowski
2 ML), upper Manu Road (Jankowski 4 ISL, Lane 1 ISL, 1 ML, Nilsson 1 ISL, Lloyd 4 ISL, Yabar 3 ISL), Pillahuata (Lane
1 ML), Wayqecha (Geale 1 XC).
cochabambae 1 (6 individuals; 3 localities). Peru: Puno: Sina (Robbins 1 ML). Bolivia: La Paz: Río Keara, NP Madidi (Tobias/
Seddon 4 XC), 9–10 km E of Pelechuco (Mayer 1 ISL).
cochabambae 2 (27 individuals; 13 localities). Bolivia: Cochabamba: Cañadón (Lane 1 XC), Alto Río Chaparé (Parker 1 ML),
40 km NE Cochabamba (Lane 1 ISL), Corani Aqueduct Road (Kirwan 2 XC), Cuchicancha (Mayer 1 XC), Presa Corani
(King 2 XC), 52 km by road W Villa Tunari (Rowlett 1 ISL); La Paz: Chuspipata (Rowlett 1 ML, Wiedenfeld 1 ML), Co-
rioco Road (Boesman 3 XC), Cotapata (King 3 XC, Mayer 4 XC, Remold 1 ML, Whitney 1 ISL), Nor Yungas (Lane 1 XC);
Santa Cruz: Comarapa (Lambert 1 XC), Siberia Cloud Forest (Kirwan 2 XC).
APPENDIX 5. Specimens Examined. The following list identifies specimens used in the study by principal clade, study
population, country, state or department, location, museum, and specimen number. Acronyms for museums are provided
in the text of the paper.
Principal Clade A
saltuensis: Colombia: Cesar: above Hiroca USNM 373674–5, El Cinco ICN 36771, 36756; La Guajira: Cerro Pintado USNM
373676, ICN 39435.
spatiator: Colombia: Cesar: Río Guatapurí USNM 387368–9; La Guajira: Cuchilla Caracas CM 45155; Magdalena: Cuchilla
San Lorenzo CM 37917, USNM 387370, Cuchilla Cebolleta ICN 23459.
Principal Clade B
rufula 1: Colombia: Norte de Santander: PNN Tamá IAvH-A 8385, 14947; Santander: La Pica CM 59572, Piedecuesta USNM
411893–4, Santa Bárbara IAvH-A 16969–70. Venezuela: Táchira: Páramo de Tamá FMNH 43600–1.
rufula 1 or rufula 3: Colombia: Santander or Boyacá: Peña Blanca CM 59689.
rufula 3: Colombia: Cundinamarca: RN Bioandina IAvH-A 16188, Bogotá FMNH 57240, 57242, Carpanta ICN 31315, 31324,
Chipaque AMNH 121976, Choachí AMNH 121320, El Peñón AMNH 121977, Fómeque IAvH-A 6631, La Calera IAvH-A
12727, 16174, Páramo de Guasca ICN 11816–9, IAvH-A 12688, 12673, San Miguel ICN 11815, Usme ANDES-O 004.
rufocinerea rufocinerea: Colombia: Antioquia: Páramo Sonsón USNM 436486.
rufocinerea romeroana: Colombia: Huila: Cascada Río Bedón IAvH-A 525–6.
Principal Clade C
rufula 2a: Colombia: Boyacá: SFF Iguaque IAvH-A 12296, Páramo El Valle IAvH-A 17056.
rufula 2b: Colombia: Caldas: La Martinica MUA-AVP 556; Quindío: Laguneta AMNH 111998, 112000; Tolima: Toche IAvH-A
17513.
rufula 2c: Colombia: Cauca: Coconuco ANSP 142401, Laguna San Rafael FMNH 255649–50, IAvH-A 2437–8, 7680, Santa
Leticia IAvH-A 2436.
rufula 4: Colombia: Antioquia: Cerro Paramillo AMNH 133533, Páramo Frontino USNM 436485; Cauca: Coast Range AMNH
109634–5; Cerro Munchique FMNH 249750, LACM 37383; Risaralda: PNN Tatamá IAvH-A 13358, Valle del Cauca:
PNN. Farallones UV 3971.
rufula 5: Ecuador: El Oro: Taraguacocha AMNH 167278; Napo: Papallacta AMNH 492263, 492265; Pichincha: Pichincha
ISLER ET AL.
52 · Zootaxa 4817 (1) © 2020 Magnolia Press
USNM 305184; Zamora-Chinchipe: Cordillera Las Lagunillas ANSP 185511. Peru: Sucumbíos: Cocha Seca USNM
614861.
Principal Clade D
cajamarcae: Peru: Cajamarca: Chugur AMNH 235896; Lambayeque: Tres Lagunas MSB 28074, 28142; Piura: Tambo FMNH
222151; Huancabamba LSU 78576–7.
Principal Clade E
blakei 1: Peru: Amazonas: Leimebamba-Las Chilchos Trail FMNH 480820–1; Huánuco: Hacienda Paty LSUMZ 74096–7,
75243; San Martín: Florida LSUMZ 174009.
Principal Clade F
obscura 1: Peru: Huánuco: Esperanza FMNH 296697, Huaylaspampa ANSP 176470, LSUMZ 74099, 74104; San Martín:
Puerto del Monte LSUMZ 104488, 104491.
obscura 2: Peru: Huánuco: Panao FMNH 66252, Bosque Potrero LSUMZ 113581–2; Pasco: Millpo LSUMZ 128539, 128541,
Rumicruz AMNH 174092.
obscura 3: Peru: Junín: Maraynioc AMNH 169728–9; Satipo LSUMZ 127645, 127648–9; Toldopampa KU 113873–5, COR-
BIDI Av-MCFK44, Av-JRSL172, Av-JRSL180.
Principal Clade G
blakei 2: Peru: Huánuco: Chaglla LSUMZ 113578; Pasco: Cumbre de Ollón LSUMZ 106081; Playa Pampa 170664; Junín: R.
Satipo CORBIDI Av-PAH653, Av-MFOR324; KU 113990–1.
blakei 3: Peru: Ayacucho: Ccano KU 112735–6, CORBIDI Av-MJA311, Av-LAGSP-60, Chupon CORBIDI Av-KVG166, Av-
PAH1241, Av-LCA2012-18; KU 122539–40, Punco AMNH 820774–5.
Principal Clade H
occabambae: Peru: Cusco: Abra Bella Vista MSB 34460, Amacho FMNH 222152, Cedrobamba USNM 273165–6, Cordillera
Vilcabamba AMNH 820463, Marcapata LSUMZ 78575, Pillahuata FMNH 429996, Tocopoqueyu USNM 273161–4.
cochabambae 1: Peru: Puno: Sina CORBIDI AV00982, KU 115475, Valcón LSUMZ 98357–8, 98361, 98363.
cochabambae 2: Bolivia: Cochabamba: Incachaca ANSP 140348, 146340, Sihuencas ANSP 170141; La Paz: Cotapata LSUMZ
96069, 96072; Santa Cruz: Santa Cruz DMNH 67057.
APPENDIX 6. Photographs
We gathered a representative series of all rufula complex populations considered in this analysis for direct visual comparison
at USNM. We present photos of examples for qualitative comparison, arranged by population, taken under identical lighting in
three positions (ventral, dorsal, lateral; PAH) and listed in each photo from left to right. We also present photos of the type speci-
men of G. rufula in the context of other rufula complex specimens housed at MCZ (ventral, dorsal, lateral positions; J. Trimble).
Finally, we present photos of several important Colombian specimens for additional comparison (ventral, dorsal, lateral posi-
tions; AMC). Acronyms for museums are provided in the text of the paper.
GRALLARIA RUFULA COMPLEX Zootaxa 4817 (1) © 2020 Magnolia Press · 53
FIGURE A13. Specimens of (1) Grallaria saltuensis: (left) USNM 373674, Serranía de Perijá, Cesar, Colombia; (center)
USNM 373675, Serranía de Perijá; (right) USNM 373676, Serranía de Perijá.
ISLER ET AL.
54 · Zootaxa 4817 (1) © 2020 Magnolia Press
FIGURE A14. Specimens of (2) Grallaria spatiator: (left) CM P45155, Sierra Nevada de Santa Marta, Magdalena, Colombia;
(center) USNM 387369, Río Guatapurí, Cesar, Colombia; (right) CM P37917, San Lorenzo, Magdalena.
GRALLARIA RUFULA COMPLEX Zootaxa 4817 (1) © 2020 Magnolia Press · 55
FIGURE A15. Specimens of (3) Grallaria rufula sensu stricto, study population rufula 1: (left) CM P59572, La Pica, Santand-
er, Colombia; (center) USNM 411893, Hacienda Las Vegas, Santander; (right) FMNH 43601, Páramo de Tamá, Santander.
ISLER ET AL.
56 · Zootaxa 4817 (1) © 2020 Magnolia Press
FIGURE A16. Specimens of (3) Grallaria rufula sensu stricto, study population rufula 3: (far left) FMNH 57240, “Bogota”,
Colombia; (left center) AMNH 121976, Chipaque, Cundinamarca, Colombia; (right center) AMNH 121320, no specific locality,
Colombia; (far right) FMNH 57242, “Bogota”.
GRALLARIA RUFULA COMPLEX Zootaxa 4817 (1) © 2020 Magnolia Press · 57
FIGURE A17. Specimen of (4a) Grallaria r. rufocinerea: USNM 436486, Páramo Sonsón, Antioquia, Colombia.
ISLER ET AL.
58 · Zootaxa 4817 (1) © 2020 Magnolia Press
FIGURE A18. Specimen of (4b) Grallaria rufocinerea romeroana: IAvH-A 525, Cascada Río Bedón, PNN Puracé, Huila,
Colombia (type specimen).
GRALLARIA RUFULA COMPLEX Zootaxa 4817 (1) © 2020 Magnolia Press · 59
FIGURE A19. Specimens of (5) Grallaria alvarezi: (left) AMNH 109685, Popayán, Cauca, Colombia; (center) LACM 29918,
Cerro Munchique, Cauca; (right) FMNH 249750, Cerro Munchique.
ISLER ET AL.
60 · Zootaxa 4817 (1) © 2020 Magnolia Press
FIGURE A20. Specimens of (6) Grallaria saturata, study populations rufula 2b and 2c: (left) AMNH 111998, Santa Isabel,
Quindío, Colombia; (center) AMNH 112000, Santa Isabel; (right) FMNH 24895, Laguna San Rafael, Cauca, Colombia.
GRALLARIA RUFULA COMPLEX Zootaxa 4817 (1) © 2020 Magnolia Press · 61
FIGURE A21. Specimens of (6) Grallaria saturata, study population rufula 5: (left) ANSP 185511, Cordillera Laguanillas,
Zamora Chinchipe, Ecuador; (center) AMNH 167278, Taraguacocha, El Oro, Ecuador; (right) AMNH 492265, Papallacta,
Napo, Ecuador.
ISLER ET AL.
62 · Zootaxa 4817 (1) © 2020 Magnolia Press
FIGURE A22. Specimens of (7) Grallaria cajamarcae: (left) MSB 28074, Tres Lagunas, Lambayeque, Peru; (center) FMNH
222151, Tambo, Piura, Peru; (right) AMNH 235896, 40 miles NW Chugur, Cajamarca, Peru.
GRALLARIA RUFULA COMPLEX Zootaxa 4817 (1) © 2020 Magnolia Press · 63
FIGURE A23. Specimens of (8) Grallaria blakei: (left) FMNH 480821, Leimebamba, Amazonas, Peru; (center) LSUMZ
174009, 22 km ENE La Florida, San Martín, Peru; (right) LSUMZ 75243, trail to Hacienda Paty, Huánuco, Peru.
ISLER ET AL.
64 · Zootaxa 4817 (1) © 2020 Magnolia Press
FIGURE A24. Specimens of (9) Grallaria gravesi: (left) LSUMZ 104488, Puerta del Monte, San Martín, Peru; (center) LSUMZ
74104, between Zapatagocha and Huallaspampa, Huánuco, Peru; (right) LSUMZ 74099, Bosque Cutirraga, Huánuco.
GRALLARIA RUFULA COMPLEX Zootaxa 4817 (1) © 2020 Magnolia Press · 65
FIGURE A25. Specimens of (10) Grallaria oneilli: (left) LSUMZ 113582, Bosque Potrero, Huánuco, Peru; (center) LSUMZ
128541, Millpo, Pasco, Peru; (right): LSUMZ 128539, Millpo.
ISLER ET AL.
66 · Zootaxa 4817 (1) © 2020 Magnolia Press
FIGURE A26. Specimens of (11) Grallaria obscura: (left) AMNH 169728, Maraynioc, Junín, Peru; (center) LSUMZ 127648,
8 km SE Calabaza, Junín; (right) KU 113873, below Toldopampa, Junín.
GRALLARIA RUFULA COMPLEX Zootaxa 4817 (1) © 2020 Magnolia Press · 67
FIGURE A27. Specimens of (12) Grallaria centralis: (left) LSUMZ 106081, Cumbre de Ollón, Pasco, Peru; (right) KU 113990,
below Toldopampa, Junín, Peru.
ISLER ET AL.
68 · Zootaxa 4817 (1) © 2020 Magnolia Press
FIGURE A28. Specimens of (13) Grallaria ayacuchensis: (left) KU 122539, Chupón, Ayacucho, Peru; (center) AMNH 820774,
Puncu, Ayacucho; (right) AMNH 820775, Puncu.
GRALLARIA RUFULA COMPLEX Zootaxa 4817 (1) © 2020 Magnolia Press · 69
FIGURE A29. Specimens of (14a) Grallaria o. occabambae: (left) AMNH 820463, Cordillera Vilcabamba, Cusco, Peru; (cen-
ter) USNM 273163, Tocopoquen, Occobamba Valley, Cusco; (right) USNM 273161, Occobamba Valley.
ISLER ET AL.
70 · Zootaxa 4817 (1) © 2020 Magnolia Press
FIGURE A30. Specimens of (14b) Grallaria occabambae marcapatensis: (left) FMNH 222152, Amacho, Cusco, Peru; (right)
MSB 34460, Abra Bella Vista, Cusco.
GRALLARIA RUFULA COMPLEX Zootaxa 4817 (1) © 2020 Magnolia Press · 71
FIGURE A31. Specimens of (15) Grallaria sinaensis: (left) LSUMZ 98363, Valcón, Puno, Peru; (right) LSUMZ 98358, Val-
cón; right: LSUMZ 98357, Valcón.
ISLER ET AL.
72 · Zootaxa 4817 (1) © 2020 Magnolia Press
FIGURE A32. Specimens of (16) Grallaria cochabambae: (left) ANSP 146340, Incachaca, Cochabamba, Bolivia; (center)
LSUMZ 96097, Cotapata, La Paz, Bolivia; (right) DMNH 67057, Siberia cloud forest, Cochabamba.
GRALLARIA RUFULA COMPLEX Zootaxa 4817 (1) © 2020 Magnolia Press · 73
FIGURE A33. Series of specimens at MCZ: (first, on the far left) Grallaria spatiator: MCZ 105638 (type specimen), Maco-
tama, La Guajira, Colombia; (second) G. rufula sensu stricto (s.s.) MCZ 76736 (type specimen): “Bogota”, Colombia; (third)
G. cajamarcae: MCZ 366082, Llama, Cajamarca, Peru; (fourth) G. cajamarcae: MCZ 366151, Llama; (fifth) G. saturata: MCZ
138487, Pichincha, Pichincha, Ecuador; (sixth) G. rufula s.s.: MCZ 76737, “Bogota”; (seventh) G. rufula s.s.: MCZ 76738,
“Bogota”; (eighth) G. saturata: MCZ 138486, Oyacachi, Napo, Ecuador; (ninth) G. rufula s.s.: MCZ 46073, “Bogota”; (tenth)
G. obscura: MCZ 266701, Maraynioc, Junín, Peru.
ISLER ET AL.
74 · Zootaxa 4817 (1) © 2020 Magnolia Press
FIGURE A34. Series of specimens at ICN and IAvH-A: (first, on the far left): G. rufula s.s., study population rufula 3: ICN
11817, Páramo de Guasca, Cundinamarca, Colombia; (second) G. rufula s.s., study population rufula 1: IAvH-A 16969, vda.
Salinas, Santa Bárbara, Santander, Colombia; (third) Grallaria saturata, study population rufula 2a: IAvH-A 17056, Páramo
El Valle, Cómbita, Boyacá, Colombia; (fourth) Grallaria saturata, study population rufula 2a: IAvH-A 12296, Morro Negro,
SFF Iguaque, Boyacá, Colombia; (fifth) G. alvarezi: IAvH-A 13358, Pueblo Rico, PNN Tatamá, Risaralda, Colombia; (sixth)
G. saltuensis: ICN 36771, vda. El Cinco, Cesar, Colombia; seventh (seventh): G. spatiator: ICN 23459, Cuchilla Cebolleta, San
Pedro de la Sierra, Magdalena, Colombia.