A taxonomic review of the Dark-winged Trumpeter PsoPhia viridis ( Aves : Gruiformes : Psophiidae )

The Dark-winged Trumpeter, Psophia viridis (Gruiformes, Psophiidae) is a Brazilian endemic species and includes three subspecies: Psophia viridis viridis Spix, 1825; Psophia v. dextralis Conover, 1934, and Psophia v. obscura Pelzeln, 1857, as well as P. v. interjecta Griscom & Greenway, 1937, whose validity has been questioned by several authors. These taxa are allopatric in distribution along the south of the Amazon River, although the precise limits of their distribution still remain unknown. This complex has never been taxonomically reviewed and this work aims to test the validity of its taxa based on the Phylogenetic Species Concept. Morphometrical characters and plumage colour patterns were analyzed, and the distribution of the taxa was also revised. In this study, 108 specimens from 41 localities were examined (all types included), with each reliable literature-based locality being included in order to delimit the geographical distribution of the complex. Morphometrical data did not point out significant differences between the taxa, also showing no sexual dimorphism among them. Meanwhile, plumage characters showed consistent and distinct patterns for each of the taxa, except for P. v. interjecta, whose features indicated by authors as diagnosable are the result of individual variation. No clinal variation or intergradation were observed, even at regions close to the rivers headwaters, where supposedly populations could be in contact. It is suggested that the currently accepted subspecies be elevated to the species level, such as: Psophia viridis Spix, 1825, distributed in the Madeira-Tapajós interfluvium, P. dextralis, found in the TapajósTocantins interfluvium, and P. obscura, which occurs from the right bank of the Tocantins River to the west of the State of Maranhão.

zon, from eastern Peru to the Madeira river in the west, and Bolivia.The Dark-winged Trumpeter (P.viridis), is a Brazilian endemic and also occurs south of the Amazon river, but east of the Madeira river to the state of Maranhão (Sherman, 1996).
Depending on the author, the Dark-winged Trumpeter may include three or four taxa.The nominate form occurs in the region between the Madeira and Tapajós rivers, and P. v. dextralis occurs between the Tapajós and Xingu rivers.The putative and debateable P. v. interjecta occurs from the east bank of the Xingu to the west margin of the Tocantins river, being replaced by P. v. obscura from the east bank of the Tocantins to western Maranhão (Peters, 1934;Blake, 1977;Pinto, 1978;Sherman, 1996;Sick, 1997).Spix (1825:66) described Psophia viridis based on a specimen from "Villa Nuova" (= Parintins, Amazonas), the only island locality for Psophia.He describes the purple iridescence in the lower neck, the green back, and calls attention to dark green scapulars, whereas the greater wing-coverts are rusty.Psophia obscura Pelzeln, 1857 was the next to be described, based on three specimens collected by J. Natterer.The type locality given by Pelzeln (1857) was simply "Brasilia", later changed by him to "Pará" (Pelzeln, 1871).Natterer (in Pelzeln, 1857) states that he collected three specimens on January 1835.At that time he was in Belém or nearby (Vanzolini, 1993), which suggests that the type locality should be Belém, as also suggested by Hellmayr & Conover (1942) and Pinto (1978).Pelzeln (1857) pointed that P. obscura differs from P. viridis by its brownish-green mantle and discreet purple iridescence at the base of the neck.
Psophia viridis dextralis Conover, 1934, was described based on an adult male collected by A.M. Olalla on 02 December 1932 at Tauari, Tapajós river, Pará.The diagnostic characters of P. v. dextralis are the greenish olive mantle and that the iridescence of the wings is almost absent and, when not, is greenish.Three years after P. v. dextralis was described, P. v. interjecta was described from an adult male from Cametá, left (west) bank of the Tocantins river (Griscom & Greenway, 1937).Psophia v. interjecta was considered as a combination of characters from P. v. obscura and the nominate form.The only specimen used to describe this taxon was different from two topotypic pairs of P. v. dextralis available to the authors, which led them to consider P. v. interjecta to be a valid taxon (Griscom & Greenway, 1937).Diagnosis is based on the purple iridescence of the wing and on the brown upper part of the mantle, thus separating it from P. v. obscura.Sharpe (1894) considered Psophia obscura to be a full species, while Peters (1934) made it subordinate to P. viridis, as it is treated today.Griscom & Greenway (1937) observed that P. v. interjecta has features of both P. v. obscura and P. v. viridis, while intermediate geographically between P. v. obscura and P. v. dextralis. Haffer (1974) stated that P. v. interjecta is indistinguishable from P. v. dextralis and suggested that it is an intergradation of P. v. dextralis and P. v. obscura, and only subtly different from them, being possibly found in simpatry with both.Although he did not synonymize the two, Haffer's treatment (1974) was followed by many (e.g.Sherman, 1996), yet Pinto (1978) considered P. v. interjecta to be valid while recognizing that its diagnosis from the others remains uncertain.
Since the original descriptions, the taxa belonging to the Psophia viridis complex have not been extensively reviewed.Here, we address this question by examining all of the taxa involved to examine and compare their morphological variation and geographic distribution.
Colors were determined following the catalogs of Smithe (1975;hereafter S) and of Munsell (1994;hereafter M).All specimens were analyzed under natural light.Plumage characters used included some that were used in the past to separate taxa, and included: a) color of the mantle in the upper (proximal to the neck), intermediate, and lower portion (distal from the neck); b) iridescence in the neck (presence -absence); and c) purple in the wing (presence -absence).
Colors of other body parts were not used here due to the inherent variation found in museum specimens.

AnD DIScuSSIOn
We found no morphometric differences between the sexes and among the described taxa (data not shown, available under request).Color patterns resulted in three well defined groups.Birds from the region between the Madeira and Tapajós rivers (24 specimens) have, with one exception, the distal part of the mantle Parrot Green (S 160 and 260).The exception was specimen MNRJ 9645, which has this area Olive Green (S 47).The intermediate portion of the mantle varied between Parrot Green (S 260) and Olive Green, while the proximal portion was Dark Green (S 162A).Again, the exception was specimen MNRJ 9645, which has Very Dark Brown (M 7.5YR/2.5/3).This appears to be an individual variation since another specimen collected 70 km of this specimen (MZUSP 76728) is typical to the remainder of those birds.All specimens from this region have iridescent purple on their wings and neck (Table 1; Figures 1-3a).
A second group (n = 73), collected between the Tapajós and the Tocantins rivers, is very homogeneous in color.The distal portion of the mantle is Olive Green (S 46), with two exceptions (MPEG 51281 and 51284), with the former being Leaf Green (S 146) and the latter being Greenish Olive (S 49).Two other specimens in this same series (MPEG 51282 and 51283) have the distal portion of the mantle Olive Green, in agreement with all the other specimens.iridescence is almost or entirely absent in all specimens analyzed (Table 1; Figures 1-3b, c).
The third group (n = 11), found from eastern Tocantins to Maranhão, has the distal and intermediate portion of mantle consistently Dark Green (S 162A and 262), while the proximal portion is Dark Brown (M 7.5YR/3/2).Wings and neck iridescence is almost or entirely absent on this group (Table 1; Figures 1-3d).
Psophia viridis, according to Spix (1825), has a green mantle and the lower neck is iridescent purple, both considered to be diagnostic features for this species.The colors found in the holotype and in a topotype (MZUSP 10938) are in agreement with the original description and are constant in the specimens between the Madeira and Tapajós rivers.Also, P. viridis Spix, 1825 can be separated from other Psophia by having the distal and intermediate portion of the mantle Par-rot Green and the proximal portion Dark Green.The neck and wings have a conspicuous purple iridescent.Conover (1934) compared a specimen of Psophia viridis collected on the east bank of the Tapajós river (Tauari) with three specimens from Madeira river (including the type) and two P. obscura from the Natural History Museum of Wien to describe P. v. dextralis.Additionally, he refers to a specimen from the Camaraipi River (Pará) that has more in common with P. v. obscura than with P. v. dextralis, suggesting that specimen is perhaps transitional between P. v. dextralis and P. v. obscura (see below).Two topotypes in the MZUSP have the distal portion of the mantle Olive Green, in agreement with the holotype (FMNH 410480).Also, purple iridescence on the wings and neck is almost absent.These characters are consistent in specimens from the area between the Tapajós-Tocantins rivers.Psophia viridis dextralis Conover, 1934   Griscom & Greenway described P. v. interjecta based in a single specimen, and stated that this taxon presented a combination of the characters from P. v. obscura and the nominate form.The characters of P. v. obscura observed on P. v. interjecta would be the discreet iridescence both on the neck and the wings and the mantle darker than observed in P. v. obscura.Griscom & Greenway (1937) do not determine which characters of the nominate form could be found in P. v. interjecta.
Our results show that specimens from the putative range of P. v. interjecta do not possess characters which can diagnose them from P. v. dextralis.Thus, according to morphological characters, P. v. interjecta cannot be considered a valid taxon, and therefore it is an individual variation of P. v. dextralis.Some authors have already recognized this, but also suggested that these may be intermediate between P. v. dextralis and P. v. obscura (Blake, 1977;Pinto, 1978;Sherman, 1996), but our analysis rejects this hypothesis.Although the original description of P. v. interjecta includes extensive blue and iridescent wing coverts ("extensive blue apical spots to the wing-coverts"), only one of the specimens analyzed (MPEG 37205) had this color on the wing coverts and therefore should be considered to be an individual variation as observed in the mantle of some specimens from this same area.Pelzeln (1857) described Psophia obscura based on three individuals without designating a holotype, designating an adult male as the lectotype some years later (NHMW 39401, Pelzeln, 1871).Pelzeln (1857) noted that P. obscura differs from P. viridis due to its smaller bill and tarsus, discreet purple iridescence at the base of the neck and a brownish-green mantle.Here, we show that P. v. obscura can be separated both from P. v. viridis and P. v. dextralis based on the Dark Green (S 162A and S 262) color of the distal and intermediate portion of the mantle, being Dark Brown (M 7.5YR/3/2) in the upper region.The purple iridescence of the wings and the neck, when present, is less noticeable than in P. v. viridis.
We found five distinct and non-overlapping plumage characters that unambiguously diagnose each one of the three groups of specimens established above (see Table 2), and we also noted a lack of clinal variation and putative hybrids.Helbig et al. (2002) state that a taxon is diagnosable if individuals of the same age and sex are different from individuals of the same age and sex of all of the other taxa by at least one qualitative difference, or if one category of age and sex is separated by a complete discontinuity in at least one continuous and variable character (quantitative character) of the same category of age and sex of other related taxa.Based on morphological characters, our findings fulfill this definition and lead us to consider that the former Psophia viridis complex comprises three distinct species, which differ in their mantle color and in the iridescence of the wings and base of the neck.Psophia viridis interjecta Griscom & Greenway, 1937 shows no diagnostic character and should be considered as a junior synonym of P. dextralis Conover, 1934.The diversity of dark-winged trumpeters is represented by the following species:
Diagnosis: distal and intermediate portions of the mantle Parrot Green (S 160; S 260), and upper portion Dark Green (S 162A); purple iridescence of the wings and neck present and conspicuous.
Diagnosis: distinguished from the other taxa by presenting the distal portion of the mantle Olive Green (S 46), and the intermediate and proximal portion of the mantle Very Dark Brown (M 7.5YR/2.5/3e M 7.5YR/2.5/2).The purple iridescence of the wings and the neck is discreet or absent.
Distribution: between the Tapajós and Tocantins rivers, with the northern limit at Portel (Pará, MPEG 40708 and MPEG 40709) and the southern limit at Sete de Setembro river (Mato Grosso, MPEG 14781; Figure 4).
Diagnosis: distinguished from the other species by the distal and intermediate portion of the mantle Distribution: Occurs to the east of Tocantins river and west to Buriticupu (Maranhão, MPEG 37338;Oren 1990Oren , 1991)).The northern limit is Belém, the type locality, and the southern limit is Jacundazinho (Pará, MPEG 36328; Figure 4).
These species of Dark-winged Trumpeters are allopatric.The large rivers of the Amazon basin seems to be efficient barriers and may often prevent contact between related populations, as partially observed with these trumpeters.While complete isolation by rivers is rarely observed (Haffer 1974), the differences in plumage could disappear clinally towards the headwaters, and the river would be less effective as a barrier for dispersal, allowing contact between the populations.Despite the sampling of Dark-winged Trumpeters are far from ideal, most of the specimens were collected along the rivers and we examined specimens collected at the headwaters of some main rivers south of Amazon.We were able to check the constancy of the characters especially on these areas and we did not see any plumage variations associated with latitude, being conservative even at regions where the rivers are narrower and supposedly the taxa could be in contact.We found no evidence for clinal variation, and color patterns appeared consistent along the species' distribution.The Xingu river did not isolate (at least from a morphological point of view) populations of the Dark-winged Trumpeters while the Madeira, Tapajós and Tocantins rivers did, separating populations and eventually preventing gene flow between Psophia viridis, P. dextralis and P. obscura, respectively.
Taxonomic research can also be an important tool for conservation of the Dark-winged Trumpeters.While subspecies are not often considered when addressing the issue of threatened taxa, Psophia obscura is in the Brazilian list of threatened birds (as P. v. obscura, IBAMA, 2003).The Belém center of endemism, where P. obscura is found, is an area with many endemic and/or threatened birds, such as the Golden Parakeet (Guarouba guarouba), among others (Roma, 1996;Silveira & Belmonte, 2005) and is the most degraded and logged area in the Brazilian Amazon.Trumpeters require quite large home ranges and are sensitive to hunting, and hence are much more likely to disappear in disturbed sites.The main conservation strategy for P. obscura is the urgent conservation of the remaining forested areas of the east of Pará and Maranhão (Oppenheimer, 2008).The creation and effective protection of conservation units or private reserves on the remaining preserved forests in the Belém center of endemism is essential not only for this taxon but for the other endemic and threatened species in this region.

Os
jacamins-de-costas-verdes, Psophia viridis (Gruiformes, Psophiidae) são endêmicos da Amazônia brasileira e contam, atualmente, com três subespécies reconhecidas: Psophia viridis viridis Spix, 1825;Psophia v. dextralis Conover, 1934, e Psophia v. obscura Pelzeln, 1857, além de P. v. interjecta Griscom & Greenway, 1937, cuja  Manuscripts are considered on the understanding that they have not been published or will not appear elsewhere in substantially the same or abbreviated form.The criteria for acceptance of articles are: quality and relevance of research, clarity of text, and compliance with the guidelines for manuscript preparation.
Manuscripts should be written preferentially in English, but texts in Portuguese or Spanish will also be considered.Studies with a broad coverage are encouraged to be submitted in English.All manuscripts should include an abstract and keywords in English and a second abstract and keywords in Portuguese or Spanish.
Authors are requested to pay attention to the instructions concerning the preparation of the manuscripts.Close adherence to the guidelines will expedite processing of the manuscript.
Manuscript Form: Manuscripts should not exceed 150 pages of double-spaced, justified text, with size 12 and source Times New Roman (except for symbols).Page format should be A4 (21 by 29.7 cm), with 3 cm of margins.The pages of the manuscript should be numbered consecutively.(1) Title Page: This should include the title, short title, author(s) name(s) and institutions.The title should be concise and, where appropriate, should include mention of families and/or higher taxa.Names of new taxa should not be included in titles.
(  , 1998) or (Rangel, 1890;Silva & Oliveira, 1998a, b;Adams, 2000) or (Silva, pers.com.) or (Silva et al., 1998), the latter when the paper has three or more authors.The reference need not be cited when authors and date are given only as authority for a taxonomic name.
(5) references: The literature cited should be arranged strictly alphabetically and given in the following format: • Journal Article -Author(s).Year.Article title.Journal name, volume: initial pagefinal page.Names of journals must be spelled out in full.
Tables: All tables must be numbered in the same sequence in which they appear in text.
Authors are encouraged to indicate where the tables should be placed in the text.They should be comprehensible without reference to the text.Tables should be formatted with vertical (portrait), not horizontal (landscape), rules.In the text, tables should be referred as Table 1, Tables 2 and 3, Tables 2-6.Use "  For other details of manuscript preparation of format, consult the CBE Style Manual, available from the Council of Science Editors (www.councilscienceeditors.org/publications/style.cfm).
Papéis Avulsos de Zoologia and Arquivos de Zoologia are publications of the Museu de Zoologia da Universidade de São Paulo (www.mz.usp.br).
Always consult the Instructions to Authors printed in the last issue or in the electronic home pages: www.scielo.br/pazor www.mz.usp.br/publicacoes.

FIGure 4 :
FIGure 4: Distribution of dark-winged trumpeters in Brazil.Asterisk refers to the type-locality of each taxa.Points with a central dot refer to records obtained from literature.Circles refers to specimens analyzed of Psophia viridis, triangles to P. dextralis (including P. v. interjecta), and squares to Psophia obscura.Brazilian States are Amazonas (AM), Pará (PA), Rondônia (RO), Mato Grosso (MT), Tocantins (TO) and Maranhão (MA).
The text should be arranged in the following order: Title Page, Abstracts with keywords, body of Text, Literature cited, Tables, Appendices, and Figure captions.Each of these sections should begin on a new page.

Illustrations:
Figures should be numbered consecutively, in the same sequence that they appear in the text.Each illustration of a composite figure should be identified by capital letters and referred in the text as: Fig. 1A, Fig. 1B, for example.When possible, letters should be placed in the left lower corner of each illustration of a composite figure.Handwritten lettering on illustrations is unacceptable.Figures should be mounted in order to minimize blank areas between each illustration.Black and white or color photographs should be digitized in high resolution (300 dpi at least).Use "Fig(s)."for referring to figures in the text, but "FIGURE(S)" in the figure captions and "fig(s)."when referring to figures in another paper.responsability: Scientific content and opinions expressed in this publication are sole responsibility of the respective authors.copyrights: The journals Papéis Avulsos de Zoologia and Arquivos de Zoologia are licensed under a Creative Commons Licence (http://creativecommons.org).

TAbLe 1 :
Plumage patterns found on dark-winged trumpeters

Material and Methods, results, Discussion, conclusion, Acknowledgments, and references at end.
) Abstract: All papers should have an abstract in english and another in Portuguese or Spanish.The abstract is of great importance as it may be reproduced elsewhere.It should be in a form intelligible if published alone and should summarize the main facts, ideas, and conclusions of the article.Telegraphic abstracts are strongly discouraged.Include all new taxonomic names for referencing purposes.Abbreviations should be avoided.It should not include references.Abstracts and keywords should not exceed 350 and 5 words, respectively.Primary headings in the text should be in capital letters, in bold and centered.Secondary headings should be in capital and lower case letters, in bold and centered.Tertiary headings should be in capital and lower case letters, in bold and indented at left.In all the cases the text should begin in the following line.