Content uploaded by Morton L. Isler
Author content
All content in this area was uploaded by Morton L. Isler on May 06, 2016
Content may be subject to copyright.
Accepted by T. Chesser: 7 Nov. 2009; published: 7 Dec. 2009 61
ZOOTAXA
ISSN 1175-5326 (print edition)
ISSN 1175-5334 (online edition)
Copyright © 2009 · Magnolia Press
Zootaxa 2305: 61–68 (2009)
www.mapress.com/zootaxa/Article
Species limits in antbirds (Aves: Passeriformes: Thamnophilidae): an evaluation
of Frederickena unduligera (Undulated Antshrike) based on vocalizations
MORTON L. ISLER1, PHYLLIS R. ISLER1, BRET M. WHITNEY2, KEVIN J. ZIMMER3,
& ANDREW WHITTAKER4
1Department of Vertebrate Zoology
⎯
Birds, National Museum of Natural History, Smithsonian Institution, Washington, DC 20560,
USA. E-mail: antbird@cox.net
2Louisiana State University Museum of Natural Science, 119 Foster Hall, Baton Rouge, LA 70803, USA.
E-mail: ictinia@earthlink.net
3Los Angeles County Museum of Natural History, 900 Exposition Blvd., Los Angeles, CA 90007, USA. E-mail: kjzimmer@charter.net
4Museu Paraense Emílio Goeldi, Caixa Postal 399, Belém, Pará, Brazil. E-mail: Andrew@birdingbraziltours.com
Abstract
The Undulated Antshrike, Frederickena unduligera, a resident of lowland humid forests of western Amazonia, is
somewhat poorly known as a result of its low population density and retiring habits. In recent years, vocalizations of F.
unduligera have been recorded sufficiently to permit geographic analysis. Vocalizations supply valuable measures of
species limits for antbirds, whose vocalizations are thought not to be learned. Based on vocalizations, we found that
populations of the Undulated Antshrike are more appropriately considered as two biological species. Three populations
considered subspecies of F. unduligera await further evaluation pending the acquisition of additional vocal recordings,
specimens, and tissue for genetic analysis.
Key words: speciation, Neotropics, Frederickena, systematics, vocalizations, Thamnophilidae
Introduction
The Undulated Antshrike, Frederickena unduligera (Pelzeln 1869), was considered a monotypic species until
Zimmer (1944) compared plumages of a small series of specimens. Zimmer described three new subspecies,
primarily on the basis of female plumage characters, an arrangement which has been followed ever since (e.g.,
Peters 1951). Frederickena unduligera is a low density species throughout its range, restricted to large tracts
of terra firme forest where it favors isolated patches of dense forest undergrowth, especially light-gap tangles
around tree falls and stream edges (Zimmer and Isler 2003) and thick, impenetrable palm (often
Lepidocaryum tenue and occasionally Leopoldinia pulchra) undergrowth in sandy and poorly drained forest
spaces (pers. obs. and M. Cohn-Haft, pers. comm.). Specimens of F. unduligera are rare in collections, the
birds vocalize relatively less frequently than do most other antshrikes, and consequently its geographic range
is poorly known. F. u. unduligera is known only from the right bank of the upper and middle Rio Negro in
Amazonas, Brazil; F. u. fulva has been thought to be restricted to south-central Colombia near the base of the
Andes, eastern Ecuador, and northeastern Peru north of the Rio Amazonas and Rio Marañón (but see below);
F. u . d iversa is said to occur from the south bank of the Rio Amazonas in Peru south to southeastern Peru,
extreme southwestern Brazil, and northwestern Bolivia; and F. u. pallida occupies the Purus-Madeira
interfluvium in Amazonas, Brazil (Fig. 1).
Vocalizations of the Thamnophilidae appear not to be learned, to be an important isolating mechanism in
the family, and consequently, to provide a “yardstick” (sensu Mayr and Ashlock 1991) for estimating
ISLER ET AL.62 · Zootaxa 2305 © 2009 Magnolia Press
reproductive isolation and species status under the Biological Species Concept (BSC) (Isler et al. 1998,
Johnson et al. 1999, Baptista and Kroodsma 2001, Helbig et al. 2002, Isler et al. 2005, Remsen 2005).
Recently obtained recordings of vocalizations provide a more rigorous basis than was heretofore available for
reassessing species limits in the complex. That analysis is the objective of this paper.
Methods
Recordings of vocalizations were compiled from the Macaulay Library (ML, Cornell Laboratory of
Ornithology, Ithaca) and from unarchived contributions in the inventory maintained by the Islers (see
Appendix). A total of 85 recordings was examined. We reviewed every recording to identify the number and
sex of individuals vocalizing and to label every vocalization as either loudsong (following Willis 1967) or
call. Calls were not classified further because of the paucity of recordings and of information regarding the
context in which they were given.
RAVEN, version 1.2 (Bioacoustics Research Program, Cornell Laboratory of Ornithology) was used to
make a spectrogram of every vocalization type delivered by each individual of either sex on every recording.
All clearly delineated spectrograms were examined visually for character states (e.g., differences in note
shape) that might distinguish one population from the others. Spectrograms shown in figures were selected to
express centrality in measurements (e.g., the mean number of notes in loudsongs) and typicality in note shape
and were made by exporting RAVEN files into CANVAS, version 9.0.4 (ACD systems, Victoria, British
Colombia).
In visual examination, it became clear that some loudsong vocal characters included in other studies (e.g.,
change in pace of notes) did not differ diagnostically across populations, and vocal characters of loudsongs
employed in this study were limited to: (1) number of notes, (2) overall duration, (3) overall pace (notes/s), (4)
note shape, (5) duration of middle note, (6) duration of interval following middle note, (7) highest frequency
of second, middle, and penultimate notes, (8) lowest frequency of second, middle, and penultimate notes, (9)
bandwidth (frequency range) of the second, middle, and penultimate notes, and (10) change in highest
frequency. These characters appeared independent from one another in that variation in one character could
occur without resulting in a different outcome for another character. Given the possibility, however, that some
characters might be linked by common ancestry, we required pairs of measurements expressing diagnostic
character states to have correlation coefficients < 0.80, an arbitrary level but one that, in our experience, falls
well below that of obviously correlated vocal measures.
Measurements were made of all loudsong recordings of individuals of pallida and unduligera but were
restricted to geographically representative samples of individuals of fulva and diversa. Sample sizes reflect
number of individuals, not number of vocalizations measured. Quantitative measures were obtained from
spectrograms projected on a 43 cm screen using default settings of Raven 1.2 (Charif et al. 2004), except that
the display was set to smooth, overlap was adjusted from 50% to 93.7% depending on recording quality, and
contrast was adjusted according to recording intensity with care taken to retain all elements of the
vocalization. Cursor measurements were typically at scales of 0.12 sec/cm and 1.5 kHz/cm. All available calls
were examined visually to identify differences in note shape. Diagnostic differences had to be discrete, non-
overlapping character states that have the potential for unambiguous signal recognition (Isler et al. 1998,
1999). Ranges of samples of continuous variables could not overlap, and the likelihood that ranges would not
overlap with larger sample sizes was estimated by requiring the means ( ) 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:
a + taSDa b – tbSDb
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, which are not distributed normally. Therefore, a non-parametric
bootstrap simulation was used to examine statistical significance. We compared the difference between the
Zootaxa 2305 © 2009 Magnolia Press · 63
FREDERICKENA UNDULIGERA SPECIES LIMITS
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 significance assigned according to the rules of this distribution.
FIGURE 1. Geographic ranges. Locations documented by specimens, vocal recordings, or published sight records are
assigned to geographic sectors (Isler 1997) and are identified by open circles. Circles in sectors in which loudsong
recordings were available are filled in solidly. Except where they follow rivers, dotted lines that separate taxa are
symbolic because actual geographic limits are unknown. Circles numbered 1 through 4 are type localities. 1 =
Frederickena u. unduligera. 2 = F. fulva. 3 = F. u. diversa. 4 = F. u. pallida. Circles numbered 5 identify sites south of the
Río Marañón in which loudsongs having the characteristics of F. fulva were recorded (see text). Circles numbered 6
denote sites where populations remain to be identified by vocal recordings (see text). Nomenclature follows
recommendations of this paper.
We assessed taxonomic status based on vocal distinctions. We recommend species status for populations
that differed sufficiently in their vocalizations to meet guidelines developed in studies of sympatric species
ISLER ET AL.64 · Zootaxa 2305 © 2009 Magnolia Press
(Isler et al. 1998). In general, 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. For brevity, we use current
subspecies names to refer to the four named populations. As an adjunct to the vocal analysis, we examined
specimens at the Louisiana State University Museum of Natural Science and the Carnegie Museum of Natural
History and compared plumage and mensural characteristics to subspecies descriptions by Zimmer (1944) of
specimens in the American Museum of Natural History.
Results
Vocalizations were categorized as either loudsongs or calls. Loudsongs (following Willis 1967) were
consistently patterned multiple-note vocalizations that were often repeated at regular intervals. Calls were
single notes, continuous tracings on spectrograms.
Loudsongs of all populations (Fig. 2) consisted of clear, evenly spaced whistles that initially grew louder
but otherwise were relatively constant in intensity. The shape of the notes differed, however, between fulva
and the remaining populations. Notes of fulva were upslurred whereas notes of the other populations were
rounded into an inverted U, a difference that was readily audible to the human ear. In addition, the pace (notes/
s) of fulva loudsongs was significantly slower ( = 2.89, SD = 0.30, range = 2.46–3.49, n = 11) than that of the
other three populations combined ( = 4.27, SD = 0.46, range = 3.49–4.93, n = 16). Finally, as measured by
dividing the highest frequency of the penultimate note by the highest frequency of the middle note, notes of
the second half of loudsongs of fulva maintained near constant frequency ( = 0.99, SD = 0.02, range = 0.97–
1.03, n = 11) whereas those of other populations rose in frequency ( = 1.09, SD = 0.03, range = 1.03–1.14, n
= 16). As apparent in Figure 1, terminal notes of diversa loudsongs typically displayed a wider bandwidth
than those of pallida and unduligera, but differences did not meet our standards for diagnosability.
An adequate inventory of recordings of calls was available only for fulva. Three types of fulva calls were
identified (Fig. 3). Whistles (N = 5) began with a sharp upscale element and were relative long in duration
(0.65–1.3 s) while dropping sharply in frequency (peaks ~3500–4000 Hz dropping to 1500–2000 Hz, once to
2500). A second type of call was part descending whistle, part snarl (N = 6). The whistle segment included an
abrupt upscale element at the start and a sharp frequency drop like the straight whistle but then dropped even
more sharply to become a snarl that was usually relatively flat in pitch. The third type of call (N = 4) began
with the abrupt upscale element but immediately became a snarl that typically dropped and then flattened in
frequency.
Few recordings of calls were available for the other populations (Fig. 3). Although these calls showed a
similarity to those of fulva in consisting of descending whistles and snarls, differences among them suggested
that, with larger samples, calls of diversa, pallida, and unduligera might be distinct from fulva and possibly
from one another.
Recorded calls were available for only one individual of nominate unduligera. This snarl lacked the
introductory element found in calls of fulva; one example appeared to be two parted.
Three apparently different calls were recorded from four individuals of pallida. The frequency shift in the
pallida whistle (N = 2) was flatter than those of fulva (peaks descending ~3700 to 3000 HZ) and the initial
upscale element was longer and more clearly audible than that of fulva. Whistles from another location lacked
any initial upscale element, were short (0.45 s), flat, and at a lower frequency (2600 to 2300 Hz). Finally, a
perfectly flat (3200 Hz), short (0.3 s) whistle dropped into a short snarl.
We had recordings of calls of only two individuals of diversa. They differed from one another and from
calls recorded for fulva. The first was a slightly descending (~3700 to 3200 Hz) whistle dropping to snarl. The
whistle showed some suggestion of an initial upscale element, and the snarl portion sometimes started at a
lower frequency (2000 Hz), appearing as a disconnected element. The second was a snarl lacking an
introductory element. The initial portion of this snarl had a wide band width but the emphasis was at ~2500–
2000 Hz; the frequency dropped sharply and became flat, similar to the snarl recorded for the nominate form.
Zootaxa 2305 © 2009 Magnolia Press · 65
FREDERICKENA UNDULIGERA SPECIES LIMITS
FIGURE 2. Loudsongs of Frederickena unduligera and F. fulva. (A) F. u. unduligera São Gabriel da Cachoeira,
Amazonas, Brazil (R. Parrini ISL-MISC.007:49). (B) F. u. pallida Tupana Lodge, Amazonas, Brazil (A. Whittaker ISL-
AW.020:33). (C) F. u. diversa Reserva Natural Palmari, Amazonas, Brazil (B. Whitney ISL-BMW.209:38). (D) F. f ulva
Yuturi Lodge, Napo, Ecuador (N. Krabbe ISL-NK.006:06).
As noted in the Introduction, relatively few specimens of Frederickena unduligera are available in
collections. Examination of female specimens (on which subspecies definitions were primarily based) at two
museums containing a substantial portion of the specimens available in the United States did not provide
additional insights. In general, as described by Zimmer (1944), specimens of fulva were most reddish-brown
(fulvous) and heavily barred and those of pallida palest and least barred, but an unusual amount of plumage
variation was found among individuals in female plumage, even among those collected at the same location.
Data were unavailable to indicate whether this variation was age related, and it is possible that some
specimens sexed as female might have been subadult males. Of particular interest were two female specimens
in the Carnegie Museum from Manacapuru and Tonantins which lie in a region of uncertain affinity (see Fig.
1). Neither specimen clearly had the plumage characteristics of fulva or unduligera, and the more fulvous-
backed and heavily barred specimen came from the eastern locality, Manacapuru, contrary to what one might
expect if there were clinality in plumage coloration towards fulva.
Discussion
On the basis of the vocal analysis described above, in combination with morphological distinctions described
by Zimmer (1944), we recommend that the Frederickena unduligera complex be considered as two biological
species, as follows:
ISLER ET AL.66 · Zootaxa 2305 © 2009 Magnolia Press
Frederickena unduligera (Pelzeln) Undulated Antshrike
Frederickena fulva Zimmer Fulvous Antshrike
Differences in pattern and quality of loudsongs of F. fulva and F. unduligera were first noted by Parker et
al. (1991) who urged further investigation. Diagnostic loudsong distinctions in note shape, overall pace of
notes, and change in frequency provide three independent vocal characters, consistent with the extent of vocal
differentiation documented for closely related sympatric species of thamnophilid antbirds (Isler et al. 1998).
The more fulvous plumage coloration and more extensive barring of females of F. fulva as compared with
females of F. unduligera were described by Zimmer (1944). The English name of Fulvous Antshrike for F.
fulva refers to the female plumage, which is appropriate in this species group and in a family in which
differences among species are often more pronounced in females than in males (Hellmayr 1929).
FIGURE 3. Calls of Frederickena unduligera and F. f ulva. (A) F. u. unduligera snarl. Parque Nacional de Jaú,
Amazonas, Brazil (S. Borges ISL-AW.020:29). (B–D) F. u. pallida. (B) whistle. Tupana Lodge, Amazonas, Brazil (A.
Whittaker ISL-AW.020:28). (C) short whistle. Igarapé Santa Maria, Amazonas, Brazil (B. Whitney ISL-BMW.184:55).
(D) whistle-snarl. Tupana Lodge, Amazonas, Brazil (A. Whittaker ISL-AW.020:21). (E–H) F. u. diversa. (E) whistle.
Parque Nacional da Serra do Divisor, Acre, Brazil (B. Whitney ISL-BMW. 126:07). (F) whistle-snarl. same recording.
(G) snarl. same recording. (H) rattle. same location (B. Whitney ISL-BMW. 126:04). (I–K) F. ful va . (I) whistle. Kapawi
Lodge, Pastaza, Peru (B. Whitney ISL-BMW.157:08). (J) whistle-snarl. same recording. (K) snarl. same location (B.
Whitney ISL-BMW.153:08).
Zootaxa 2305 © 2009 Magnolia Press · 67
FREDERICKENA UNDULIGERA SPECIES LIMITS
Loudsongs recorded in northern Peru at 15 km NE of Jirillo, San Martín (06o03’S, 76o44’W), and 84 km
WNW of Contamana on the Río Cushabatay, Loreto (07o09’S, 75o44’W), displayed vocal characteristics
typical of F. fulva. These extend the range of F. fulva south of the Río Marañón and west of the Río Ucayali
(Fig. 1). North of the Amazon River, the eastern limits of the geographic range of F. fulva and the western
limits of F. u. unduligera are unclear. Specimens have been collected along the north bank of the Amazon
(Solimões) at Tonantins (Carnegie Museum), Codajás (Swedish Museum of Natural History), and
Manacapuru (Carnegie Museum) (west to east in Fig. 1), but no vocal recordings from the region are
available..
Confirmation of the validity and clarification of the geographic limits of named subspecies (diversa and
pallida) of F. unduligera was not possible with available materials. Their reconsideration requires additional
specimens, ideally including tissue, and vocal recordings from throughout the range of F. unduligera to
investigate possibilities of clinality and introgression. As noted above, the limited available specimen
information presents a confusing picture, and the same holds for vocal data. For example, a single recording
of an adult male from just south of Humaitá, Amazonas, Brazil, might have been expected to align well with
those of pallida from lower in the Madeira/Purús interfluvium, but it appears to be closer to recordings of
diversa. Further vocal recording should place an emphasis on obtaining recordings of calls. The existing,
limited sample of recordings suggests the possibility that calls differ diagnostically among populations. In one
sympatric pair of Hypocnemis antbirds, calls were shown to differ whereas loudsongs were indistinguishable
(Isler et al. 2007).
Frederickena fulva and F. unduligera are low-density species that appear to be indicators of high-quality
lowland forest (Zimmer and Isler 2003). This attribute, combined with the results of our analysis, indicates
that independently evolving populations are geographically restricted and suggests that their continued
conservation depends on the maintenance of multiple large protected areas in western Amazonia.
Acknowledgments
The analysis was made possible by the contributions of the recordists listed in the Appendix. We are grateful
to M. Cohn-Haft for providing field observations. We appreciate the continuing support of the Macaulay
Library, Cornell Laboratory of Ornithology (G. Budney). J. V. Remsen, Jr., an anonymous reviewer, and editor
R. T. Chesser made helpful suggestions for improving the manuscript.
References
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.
Helbig, A.J., Knox, A.G., Parkin, D.T., Sangster, G. & Collinson M. (2002) Guidelines for assigning species rank. Ibis,
144, 518–525.
Hellmayr, C.E. (1929) On heterogynism in Formicarian birds. Journal für Ornithologie Festschrift für Ernst Hartert
1929, 41–70.
Isler, M.L., Isler, P.R. & Brumfield, R.T. (2005) Clinal variation in vocalizations of an antbird (Thamnophilidae) and
implications for defining species limits. Auk, 122, 433–444.
Isler, M.L., Isler, P.R. & Whitney, B.M. (1998) Use of vocalizations to establish species limits in antbirds (Passeriformes;
Thamnophilidae). Auk, 115, 577–590.
Isler, M.L., Isler, P.R. & Whitney, B.M. (2007) Species limits in antbirds (Thamnophilidae): the Warbling Antbird
(Hypocnemis cantator) complex. Auk, 124, 11–28.
Johnson, N.K., Remsen Jr., J.V. & Cicero, C. (1999) Resolution of the debate over species concepts in ornithology: a new
comprehensive biological species concept. In Adams, N.J. & Slotow, R. H. (Eds.) Proceedings 22nd International
Ornithological Congress, Durban. BirdLife South Africa, Johannesburg, pp. 1470–1482.
Mayr, E., & Ashlock, P.D. (1991) Principles of systematic zoology. McGraw-Hill, New York.
ISLER ET AL.68 · Zootaxa 2305 © 2009 Magnolia Press
Parker, T.A. III, Castillo V., A., Gell-Mann, M. & Rocha O., O. (1991) Records of new and unusual birds from northern
Bolivia. Bulletin of the British Ornithologists’ Club, 111, 120–138.
Pelzeln, A. von. (1869) Zur Ornithologie Brasiliens, Resultate von Johann Natterers Reisen in den Jahren 1817 bis
1835. Abth. 2. Wien, A. Pichler's Witwe & Sohn.
Peters, J.L. (1951) Check-list of birds of the world, vol. 7. Museum of Comparative Zoölogy, Cambridge, Massachusetts.
Remsen, J.V. Jr. (2005) Pattern, process, and rigor meet classification. Auk, 122, 403–413.
Willis, E.O. (1967) The behavior of Bicolored Antbirds. University of California Publications in Zoology, 79, 1–132.
Zimmer, J.T. (1944) Studies of Peruvian birds. XLIX. Notes on Frederickena and Octhoeca. American Museum
Novitates, 1263, 1–5.
Zimmer, K.J. & Isler, M.L. (2003) Family Thamnophilidae (Typical 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. 448–
681.
Appendix. Recordings examined
The following list identifies recordings used in the study by taxon, country, state or department, recording location, and
recordist. Nomenclature follows recommendations of this paper. Allocation of recordings to subspecies of F.
unduligera follow current assumptions of geographic limits (e.g., Zimmer and Isler 2003). Numbers following the
recordist name identify the number of cuts per recordist per location. Acronyms for recording archives: ML
=Macaulay Library, Cornell Laboratory of Ornithology, Ithaca; ISL = recordings not yet archived in an institutional
collection but that have been copied into the inventory maintained by MLI and PRI. Many of these unarchived
recordings either are in the process of being archived or will eventually be archived by the recordists.
Frederickena u. unduligera: (2 recordings; 2 localities). Brazil: Amazonas: Parque Nacional de Jaú (Borges 1 ISL), right
bank Rio Negro opposite São Gabriel da Cachoeira (Parrini 1 ISL).
Frederickena unduligera pallida: (15 recordings; 3 localities). Brazil: Amazonas: 40 km S Humaitá (Whitney 2 ISL),
Igarapé Santa Maria (Whitney 2 ISL), Tupana Lodge (Whittaker 9 ISL, Zimmer 2 ISL).
Frederickena unduligera diversa: (14 recordings; 8 localities). Brazil: Acre: Boca de Tejo on left bank R. Juruá
(Whittaker 1 ISL), Parque Nacional da Serra do Divisor (Whitney 5 ISL), Amazonas: Barro Vermelha on left bank
R. Juruá (Whittaker 1 ISL), Reserva Natural Palmarí (Whitney 1 ISL). Bolivia; La Paz: Alto Madidi (Parker 1 ISL).
Peru: Cusco: Camisea (Christian 1 ISL), Madre de Dios: Explorer’s Inn (Parker 3 ISL), Pantiacolla Lodge (Lambert
1 ISL).
Frederickena fulva: (54 recordings, 17 localities). Ecuador: Morona-Santiago: Santiago (Robbins 3 ML), Napo: Coca (J.
Rowlett 1 ISL), Lagoa Añangu (Krabbe 2 ISL), La Selva Lodge (Behrstock 1 ISL, Krabbe 2 ISL, Michael 1 ISL, G.
Rosenberg 1 ISL, R. Rowlett 1 ISL), Maxus Road (Krabbe 3 ISL, Zimmer 1 ISL), Sacha Lodge (Arvin 1 ISL),
Tiputini Biodiversity Station (Arvin 1 ISL, Zimmer 7 ISL), Volcán Sumaco (Whitney 2 ISL), Yuturi Lodge (Krabbe
1 ISL), Pastaza: Kapawi Lodge (Whitney 2 ISL), Sucumbios: Cuyabeno (J. Rowlett 1 ISL, Whitney 2 ISL). Peru:
Loreto: El Dorado (Whitney 2 ISL), El Tigre (Alvarez 3 ISL), Intuto (Alvarez 2 ISL), Quebrada Sucusari (Parker 2
ML, J. Rowlett 1 ISL, Whitney 8 ISL), Rio Cushabatay, 84 km WNW Contamana (Lane 1 ISL), San Martín: Jirillo
(Valqui 1 ISL).
