Proposal (473) to South American Classification Committee


Separate Aratinga hockingi and A. alticola from A. mitrata


Background: Chapman (1921) described a new subspecies (alticola) of Aratinga mitrata from higher elevations in the Urubamba valley of Cusco department, Peru. This taxon is distinguished from nominate mitrata (type locality unknown, but described in Tschudi’s Unters. Fauna Peruana, so presumably from ‘Peru’; more on this below) by less red overall, a larger bill, and reportedly higher elevational preferences.  Fjeldså and Krabbe (1990) suggested that alticola may deserve species status separate from mitrata based on the fact that they are found in close proximity but separate by elevation. Arndt (2005) took the situation a step further by separating not only alticola from mitrata, but also naming a new form, hockingi (type locality in Huánuco, Peru), as well as naming an additional two subspecies of mitrata from the northern and southern ends of its distribution (chlorogenys and tucumana, respectively). These taxa were described primarily on plumage, measurements, and, to a lesser extent, soft part colors.


Arndt (2005) spent several paragraphs addressing the situation of the type specimen and locality of the name “mitrata”.  As stated above, Tschudi did not specify either. Desfayes (1994) established a lectotype specimen and the locality probably was from somewhere near Chanchomayo, dpto. Pasco (based on Tschudi’s collecting route fide Stevens and Traylor 1984 and Zimmer 1930).  Arndt surmised that the lectotype is a juvenile by its plumage, and admitted it may be best assigned to what he has named “hockingi”.  However, the original type description shows what he considered a ‘typical’ mitrata with considerably more red on the face than the lectotype shows, and therefore the name applies to such birds.  The suggestion I gathered from this is that Arndt believes the illustration should be held as the type (no specimen he could locate from Tschudi’s collection appeared to be the model of this illustration), and the lectotype is not truly the type of the name mitrata, I believe this reasoning is sound.  He further went on to say that the illustration represents a bird from a population from southern Peru, not Chanchomayo, and suggested that the type locality may not be Chanchomayo after all.


Arndt (2005) started his paper relating his sighting of a flock of Aratinga mitrata from the Utcubamba valley (Amazonas dept) and saying that they must represent a new subspecies (which he names A. m. chlorogenys in this paper). This observation was then followed by the news from Peter Hocking that locals in the area have two names for the red-and-green Aratingas (something Arndt noted that Zimmer [1930] also mentioned from his field work in Huánuco) of the area and that they sound different, but travel in mixed flocks with A. mitrata. This latter comment spurred Arndt to describe the second form as A. \ hockingi.


Aratinga hockingi is found from dpto. Amazonas south to dpto. Cusco.  A. alticola is found from the Mantaro valley (Junin/Huancavelica depts.) south and east to Cochabamba dept, Bolivia (fide Arndt 2005).  Aratinga mitrata (sensu Arndt) is found from dpto. Amazonas south to Argentina, and thus completely overlaps with both hockingi and alticola.  That is to say, neither of Arndt’s newly separated/described species is found at a site where A. mitrata is not (I say this based not only on his map, but also from evidence available to me from known localities).


Discussion: I am curious why Arndt (2005) largely dismissed any potential confusion that may be added by Aratinga wagleri (of the frontata/minor group) while discussing the mitrata complex.  He only mentioned wagleri to say that hockingi may be considered more allied to Colombian/Venezuelan (nominate) wagleri than are Peruvian populations of wagleri. That’s all well and good, but why not mention how to distinguish hockingi and alticola from juvenile frontata/minor?  Are the measurements so different?  Arndt only cited Forshaw (1989) to show how much larger the maximum wing and tail measurements of frontata are. Arndt (2005) did not provide a range, nor did he question any differences between his own measurement style and that presented in Forshaw, meaning that we don’t know how compatible these measurements are.  Furthermore, Arndt (2005) never noted whether A. wagleri frontata/mitrata could be easily confused with mitrata (sensu lato) when non-adults are involved.  This begs the question: how certain can we be that Arndt’s A. hockingi does not involve non-adult specimens of wagleri frontata or wagleri minor (or, for that matter, non-adult A. mitrata)?  This is particularly important considering how much Arndt’s argument depended on living birds and specimens that were captive (and thus of unverified provenance and, in some cases I suspect, age).  The same question can be asked of his reasoning for separating A. mitrata and A. alticola.  Arndt mentioned that individuals of his “new species” were reportedly observed in flocks of individuals otherwise identified as A. mitrata.  I don’t have a great deal of experience with Aratingas in the Utcubamba, but in the Mantaro valley, I have seen A. wagleri (subspecies unknown, but identified by the red on the underwing) and A. mitrata in separate, monospecific flocks.  In the Urubamba Valley (Cusco dept), I have watched for individuals matching A. alticola, but have not been able to verify any yet, certainly not in “monospecific” flocks (nor have I heard of anyone else being able to do so).  In fact, the validity of the name “alticola” has been questioned (T. Schulenberg, in litt.) in so much that it may simply be applied to non-adult individuals of A. mitrata.  I cannot find evidence to refute this claim.


Arndt’s use of “local knowledge” as the foundation of his taxonomic reassessment of this complex strikes me as very dangerous.  Arndt (citing Hocking) relayed local knowledge of vocal differences between A. hockingi and A. mitrata, but no evidence is available to back it up.  Anyone who knows parrots will realize that they have large, varied vocabularies and are capable of learning, so using voice to separate populations must be done using homologous vocalizations, and must be done very carefully.  Probably only typical “flight calls” would hold any real, useful differences that could be used to separate Aratingas.  From personal experience, I find most members of the South American red-and-green Aratinga group (e.g., A. leucophthalma, A. wagleri, A. erythrogenys, and A. mitrata) to have effectively indistinguishable flight calls (for examples see:


Aratinga leucophthalma:


Aratinga wagleri:


Aratinga erythrogenys:


Aratinga mitrata:



No doubt Arndt knows parrots, and he may be onto something with the naming of the taxa in this paper. However, I have great reservations about using the evidence presented in Arndt (2005) as proof positive that there are multiple biological species in the red-fronted Aratinga group present in the same valleys.  Based on the map and localities presented within Arndt (2005), up to three species of red-and-green AratingasA. wagleri, A. hockingi, and A. mitrata—in dptos. Amazonas and Huánuco, potentially four—A. wagleri, A. mitrata, A. hockingi, and A. alticola—in the Mantaro valley (I say “potentially” as I have seen both A. wagleri and A. mitrata in very close proximity along the Mantaro farther NW within Huancavelica than Arndt’s “Anco” site for alticola), and three—A. mitrata, A. hockingi, and A. alticola—in dpto. Cusco. Such site diversity within a clade of parrots is unprecedented!


Plumages of parrots can vary notoriously within populations due to age, wear, molt, individual variation, health, etc. Arndt (2005) appears to have made efforts to account for this variation, but four specimens at LSUMZ from Zangudacocha, Huancayo dept. (the type locality of A. m. chlorogenys), show variation that approaches moderately red-faced “nominate” birds from Bolivia, throwing into question whether chlorogenys is a valid taxon or the end of a cline?  Given this, the relative lack of more explicit comparisons to non-adult A. wagleri frontata and A. w. minor when describing A. hockingi, and the incredible claim of three or even four sympatric red-and-green Aratingas within several intermontane valleys in Peru, I remain very wary of the accepting the validity of the names A. hockingi and A. alticola based on what evidence Arndt (2005) has presented.


More critical studies, particularly involving molecular studies and larger series of modern specimens (of wild, not captive, individuals) should shed light on whether these populations are in fact acting as biological species or not.


As a final, pedantic, point, Arndt (2005) explicitly stated that he used the PSC: “According to the phylogenetic species concept, which has been applied in this paper, this sympatric occurrence requires a division into two separate species [i.e.; mitrata and alticola].” This statement suggests an imperfect understanding of the tenants of the PSC and BSC. In fact, the PSC and BSC would likely not treat this situation differently, as long as the populations are both breeding and not overlapping only seasonally when not breeding, in which case the BSC might yet consider the two populations as conspecific, as it does with overlapping migratory and resident populations, e.g. Vireo olivaceus. The PSC would consider the taxa as species as long as they are reciprocally monophyletic and diagnosable.


Recommendation: Whereas Arndt may be found in the long run to be correct in the awarding of species status to “hockingi” and “alticola” (and perhaps also to the validity of the names “chlorogenys” and “tucumana”), the evidence he provided in Arndt (2005) to support his arguments falls short of what most taxonomists would require.  Therefore, I recommend that the SACC vote NO to this proposal, at least until it can be reevaluated with molecular evidence.  If the SACC prefers to separate the cases of A. hockingi and A. alticola, I can modify the proposal to accommodate two separate votes.




ARNDT, T. 2005. A revision of the Aratinga mitrata complex, with the description of one new species, two new subspecies, and species-level status of Aratinga alticola. J. Orn. 147: 73-86.


CHAPMAN, F.M. 1921. The distribution of bird life in the Urubamba Valley of Peru. Smithsonian Institution. Bull US Natl Mus 117. 138 pp.


DESFAYES, M. 1994. Catalogue des types du Muse´ e d’histoire naturelle de Neuchaˆ tel. Bull Soc Neuchaˆ tel Sci nat 117:79–95.


FJELDSÅ, J., AND N. KRABBE. 1990. Birds of the High Andes. Zoological Museum, Univ. Copenhagen, Copenhagen, Denmark.


FORSHAW, J.M. 1989. Parrots of the world, 3rd edition. Lansdowne, Melbourne.


STEPHENS, L. AND M. A. TRAYLOR. 1983. Ornithological Gazetteer of Peru. Museum of Comparative Zoology, Cambridge, Massachusetts.


ZIMMER, J. T. 1930. Birds of the Marshall field Peruvian expedition, 1922–1923. Field Museum of Natural History Zoological Series 15.


Daniel Lane, 22 November 2010



Comments from Remsen:  “NO.  Such an unprecedented situation requires far more rigorous evidence that presented by Arndt, especially associating the allegedly differing vocalizations with individuals that have the plumage characters of the taxa.  Of all the characters that I would consider dangerous to use in green parrots, variation in the amount of red in the plumage, particularly the head, would be at the top of the list given individual variation within a species.”


Comments from Robbins:  “NO. Given the considerable age-related plumage variation within some Aratinga taxa, Arndt 's (2005) results are questionable.  Before any taxonomic changes are made a thorough assessment is needed that includes voucher specimens with associated audio and molecular data.”


Comments from Zimmer: “NO, for all of the reasons stated by Dan, Van and Mark.”


Comments from Thomas Arndt: “I have long thought if I should react to Daniel Lane’s proposal about Aratinga hockingi and Aratinga alticola. But rarely I have read a proposal that summarizes and discusses a paper in such an unbalanced way. As a result, the first voters justify their votes by questioning parts of the paper that – if at all – never played an important roll.


“The paper is based on the analysis of 130 skin specimens, some hypotheses and information are added. Therefore, one would expect that someone would check at least part of the A. hockingi and A. alticola museum material. I don’t know why Mr. Lane did not try this but instead seemed to concentrate on a “potential confusion that may be added by Aratinga wagleri (of the frontata/minor group)”.  I am sorry that my paper did not compare immature frontata or minor, but it seemed not necessary to me because immature frontata and minor already have some red feathers to the bend of wing and therefore cannot be misidentified.”


“The proposal is misleading in other aspects, too. It indicates that “Aratinga hockingi is found from dpto. Amazonas south to dpto. Cusco” and “A. alticola is found from the Mantaro valley (Junin/Huancavelica depts.) south and east to Cochabamba dept, Bolivia.”  Mr. Lane uses this to explain later that possibly three up to four red Aratinga might be found in one area what seems to be questionable. Neither the text nor the map in the paper allow this interpretation. With respect to hockingi and alticola, I wrote about a possible zone of overlap but that there is no evidence at present.


“Both taxa are only known from certain localities, and at the moment one only can speculate how the real distribution is.  The occurrence of the red green Aratinga in the Andes is especially poorly understood and still needs a lot of research.  Last year I visited the Anco area and found exactly the birds I identified in the paper as alticola.  This population differs clearly from A. mitrata mitrata, which was not found. According to the locals (sorry for using local knowledge) Aratinga wagleri minor was present there (there was even an immature of it in a local house). This species, however, is evidently missing in the Chinchao area, which I also visited, from where hockingi and mitrata skins exist.  I, too, expect that usually only two red and green Aratinga could be found in an area, although I wouldn’t be surprised if there are exceptions.


“Mr. Lane’s didactic discussion about using PSC or BSC is quite useless seeing the current puzzling situation.  He gives the reason for not using BSC himself when explaining “as long as the populations are both breeding and not overlapping only seasonally when not breeding, in which case the BSC might yet consider the two populations as conspecific”.  There is no safe knowledge about breeding or seasonally overlapping in case of alticola or hockingi.


“In his discussion, Mr. Lane focuses on my use of local knowledge or captive birds by using phrases like “Arndt’s use of ‘local knowledge’ as the foundation of his taxonomic reassessment of this complex” or “Arndt (citing Hocking) relayed local knowledge of vocal differences between A. hockingi and A. mitrata” or “considering how much Arndt’s argument depended on living birds and specimens that were captive (and thus of unverified provenance and, in some cases I suspect, age)”.  I am sorry, but when reading the paper again, I couldn’t find any hint that I was using local knowledge as a foundation of my taxonomic reassessment or that I was using decisive arguments depending on living birds. The reason why I started the research has been information from local people and living birds I have seen, but I don’t regard this as reprehensible.


“A comment based on living birds indeed can be found concerning whether frontata and minor should be separated from wagleri and whether hockingi might be closer related to wagleri than to frontata/minor.  But this was not used as an argument to prove the species status of hockingi.


“I wonder why Mr. Lane seems to question the existence of mixed flocks of two Aratinga species at feeding locations.  Perhaps he has seen only monospecific flocks in the Urubamba Valley. With an experience of 30 years of field experience concentrating on parrots, I can assure you that this is nothing unusual, including for red and green Aratinga and especially not while feeding in a maize field (that is what exactly was described).


“It’s also strange that the proposal concentrates on the vocal differences between red and green parrots in a separate paragraph. In the introduction one will find the sentence about the first information I got from Peter Hocking about a second Aratinga species in the Utcubamba valley: it’s call was also said to be quite different to that of the Mitred Conure.


“I could go on with these samples. Altogether I can understand everyone who is very cautious with a new red and green Aratinga species in the Andes, and it is clear that a lot of additional research is necessary. But the main problem with Mr. Lane’s proposal is that he was not able or willing to examine the skin material. I am quite sure that his comments about the validity of alticola would have been different (as well as those who have questioned this taxon), including those on chlorogenys. The type locality is not Zangudacocha, Huancayo dept., as mentioned in the proposal, but rather San Pedro, southern Chachapoyas.  The skins that Mr. Lane examined come from the southern end of chlorogenys distribution, and one could expect that the amount of red is a bit more extensive there.


“Originally I wanted to publish photos of skins in my paper, but one of the revisers proposed that paintings might be more accurate (which is what I found, too). Perhaps it helps to understand the taxa a bit better by viewing these photos:


Additional comments from Dan Lane: “I am sorry that Dr. Arndt seems to think that my proposal is unbalanced. That was not my intention. At its foot, I think the reader will find that I do not reject the possibility that A. hockingi and A. alticola are good species, simply that the Arndt (2005) paper does not offer enough evidence for me to feel that the case for this taxonomy is acceptable. I stand by that statement. Should further evidence show that one or both of these taxa are viable species, I’ll be happy to recant and support the proposal that they be accepted by SACC. Given the difficulties that Aratinga parakeets offer, I think that a molecular component will be very instrumental in verifying or falsifying Arndt’s proposed taxonomy.


“First, Arndt points out my error in listing “Zapatagocha” as the type locality of A. m. chlorogenys, and I apologize for being sloppy here. Nevertheless, that the LSU specimens in question are from a locality that Arndt (2005) lists as one inhabited by A. m. chlorogenys, but show characters more similar to A. m. mitrata, either requires clinal variation (not truly ‘step-wise’) within the populations of mitrata, or that the sample size (22, versus 65 of nominate) reviewed by Arndt (2005) was not large enough to account for the full variation present within ‘chlorogenys’ and that those populations are not so divergent from nominate mitrata. Without more evidence, it will be hard to answer this question. However, photos available here:
and here:

suggest that “nominate-like” birds are not unusual even around Pedro Ruiz (to the north of the type locality for chlorogenys). But to be fair, the name ‘chlorogenys’ is not the one that is in question in this proposal; I simply feel that this is a case that illustrates the difficulty of use of plumage characters with which to diagnose a taxon in the group within the group, particularly with a small series of specimens. I will say Arndt may be onto something in suggesting that there is a cline of less red to more red from north to south within A. mitrata.


“Regarding the claim that I have not viewed the specimens in question -- in fact I have. In November and December 2003, I visited the collections in Lima (MUSM; called ‘MJPL’ by Arndt 2005) and New York (AMNH) that provide several of the key specimens mentioned in Arndt (2005), albeit before the paper was published. Admittedly, my eye was not as practiced as Dr. Arndt’s, and I was not necessarily aware of what characters to look for (nor that there were potentially three taxa in question rather than simply A. mitrata and A. alticola), but I did make notes on the specimens at the two collections. The following are comments that I made (while viewing specimens) that I think are illuminating:


MUSM collection:

“Birds identified on the tag as Aratinga mitrata alticola:

Amazonas: Pedro Ruiz (some birds also mention"(Chusgon)"... these birds are *all* wing-clipped [meaning they were captives prior to collection], so maybe [Hocking] got this locality from the people at the [bird] market in Pedro Ruiz?)
MUSM 24068 (F, juv), 25297 (F), 24070 (?, juv), 24067 (M, juv). ["juv" was written on the tag by P. Hocking, the preparator]

--All these birds lack red on the tarsus, most are drabber ("darker"? certainly, I'd say more worn) green than birds identified as "mitrata mitrata" from the same site, and these individuals have restricted, more clean-cut red on the forehead. Females seem to have a shallower bill than males, but this seems to overlap broadly with "mitrata mitrata."

Birds id'ed on the tag as Aratinga mitrata mitrata or without subspecies designation:

Amazonas: Pedro Ruiz (same caveat as above)
MUSM 24205 (F), 25296 (M), 24066 (F), 24072 (M), 24071 (M), 24069 (M, juv), 24204 (F), 24400 (M).
--Of these specimens, only 25296, 24205, and 24204 have their primaries intact (i.e., shot in the field?). Compared to the "alticola" above, these birds all seem to have more extensive red on the forehead with random patches on face, breast, and neck, and all (except 24069) have red tarsi.

Huánuco: Bosque Tapicayog: MUSM 73792 (F)
              Zapatogocha: 5706 (?), 5708 (?).
              Bosque Taprag: 4284 (M)
--All have "alticola"-like plumage with no red on tarsi, restricted red on the forehead, and virtually no red flecking on head etc (except very little on cheek of 4284).

Apurímac: Bosque de Paraguay: 5716 (F)
              Puente del Río Pampas: 5705 (F)
              Puente Pasaje: 24301 (M)
Ayacucho: Pampa de Chumbes (Río Pampas): 5701 (?, juv), 5702 (?), 5713 (M), 5714 (M)
Cusco: Pacaypata (Río Apurímac): 24299 (M), 24300 (F)
--All look like "mitrata mitrata" with the exceptions of 5701 (which lacks red tarsi, and has restricted red on the forehead), and 5713 has no red on tarsi, but has more extensive red onto face, some random flecks past orbit area, etc.

So, I suspect that there is at least a link between restricted, "clean-cut" red on the forehead and no red on the tarsi. This may be an immature character of mitrata (sensu lato), and may explain "alticola." However, I am interested in the uniformity of these characters in the Huánuco specimens (although I think one should be wary of the small sample size).”


“None of MUSM’s specimens are identified as A. alticola by Arndt (2005), but rather as A. hockingi. The specimens listed above were identified in Arndt (2005) as A. m. chlorogenys (MUSM 24204, 24205, 24066, 24071, 24072), A. m. mitrata (MUSM 24299, 24300, 24301, 5702, 2705, 5714, 5716), and A. hockingi (4284, 5701, 5706, 5708, 5713, 24067, 24068, 24069, 73192 [which I list above as ‘73792’]). Of this last series, 24067 and 24068 were identified at MUSM as ‘alticola’, presumably because of the green tarsi. So, this leaves me wondering: what do the juvenile and immature plumages of Aratinga mitrata (sensu stricto) look like? Of the specimens listed above, MUSM 24067, 24068, 24069, and 24070 are noted on the tag (by Hocking) as ‘juv.’ and these are the specimens Arndt (2005) mentioned as ‘clearly identified juveniles’ of A. mitrata chlorogenys. In addition, MUSM 5701 and 5713 are also considered ‘juv’ A. m. mitrata by Arndt (2005); again 5701 is so identified on the label. Unfortunately, the photos (slides) I took of these specimens are no longer available to me, so I can’t comment now on the condition of the plumage to see if I agree with Hocking and Arndt’s assessments of age (assuming the reasoning for the ‘juvenile’ designation are assessable characters still visible on the specimens). However, I am interested to note that no other specimens are explicitly considered ‘non-adults’ by Arndt (2005), and his characters used to assess age of A. mitrata (sensu stricto) involve the amount of red around the eye (in all A. mitrata descriptions in the paper, the presence of red around the orbit excludes A. alticola and A. hockingi; I must assume that any without red around the orbit must, therefore, be exclusively alticola or hockingi?). In short, I conclude that the argument that A. alticola or A. hockingi do not simply represent non-adult plumages of a widely phenotypically variable A. mitrata has not really been addressed by Arndt (2005). This (in my eyes at least) undermines the conclusion that alticola and hockingi are valid given the present evidence. I will further offer the following photo as evidence (weak as it is) to support my alternate hypothesis:

“Visit the following webpage <> and scroll down to ‘October 24 2010: Mitred Parakeet (Aratinga mitrata) Established Exotic In Miami Florida Area.’

It shows three Aratingas in Miami, Florida (USA), where A. mitrata is feral. To me, the photo shows an adult A. mitrata flanked by two offspring… showing characters of A. hockingi/alticola. I cannot say what the source population was for these birds, but if Arndt (2005) is right about the geographic variation within the species, I’d venture to say between southern Peru and Argentina (perhaps there are records that could verify the source?). I highly doubt that A. hockingi or A. alticola are also feral in Miami, but I cannot rule out hybridization between an A. mitrata and some other Aratinga. But for all intents and purposes, I’d say that the photo suggests that it’s possible that birds with ‘hockingi/alticola’ features may be nothing more than non-adult A. mitrata. If Arndt (2005) is correct in his description of the clinal variation of extent of red in A. mitrata populations from north to south, that would explain why “hockingi/alticola”-plumaged birds are found in the north and center of A. mitrata’s range, whereas he did not report any birds of this plumage in the southern part (i.e., Argentina).


“Another question addressed in a rather short comment by Arndt (2005) is why A. hockingi and A. alticola are not the same taxon? Again, it is the width and shape of the red forehead patch that is the main character that he appears to use to distinguish them. Aratinga alticola is also reported to have a ‘shorter tail’ than A. hockingi (I will note that sample sizes used in measurements for either of these ‘taxa’ are <10). The illustrations in Figure 1 show leg color differences, but given the few specimens (at least at AMNH and MUSM, and I would predict even fewer among the older specimens housed in European museums) that had soft part colors available -- not to mention the variability in how preparators see colors and the presumed age and individual variation potential among the birds -- Arndt (2005) did not draw much attention to this character, and rightly so.


“From my visit to AMNH:

“Identified as Aratinga ___ (in the A. wagleri series), but actually A. mitrata
AMNH 234328 (male) PERU: La Lejia, N Chachapoyas, 9000 ft; 25 March 1925 (Watkins)
     234329 (female) ""
     234327 (male)   ""
AMNH 235434 (male) PERU: San Pedro, S Chachapoyas, 8600-9400 ft; 2 Feb 1926 (Watkins)
     235433 (male)   ""

“Interestingly, whereas all specimens [in A. wagleri series] have rose-red foreheads (deeper crimson towards bill), none from Chachapoyas has red on the leading edge of the wing. 235433 and 235434 both have scattered red on head, breast. Chachapoyas birds are deeper green, and have restricted red (not so rose-red) and it is around eye. Chachapoyas birds' bills appear slightly deeper, green is duller and darker.

A. mitrata
A. m. alticola

AMNH 129138 (female) PERU: Cusco, ca 11,000ft; 16 Nov 1914 (Watkins)* [*=photographed]
     129137 (female) "" Tag: feet black, bill horn, iris red-brown. *
Sootier green above, faded red foreheads, no red on tarsi (see photos).

A. m. mitrata
AMNH 474337 (male) PERU: Chanchomayo, 1200m; 1902 (Schunke)*
     474338 (male) ""  *
     474340 (male) PERU: Cushi, Libertad, 1820m; 9/03 (Hoffmanns)
     156058 (female) ""  *
     144996 (male) PERU: San Miguel, foot of Machu Picchu; 20 July 1916 (Chapman and Cherrie)
     144995 (female) "", 5000 ft; 8 July 1916
     166513 (?)      "", 6000 m; 29 June 1915 (Heller and Yale)
     144997 (male) PERU: Santa Rita, Urubamba Canyon; 20 June 1916 (Chapman and Cherrie)
     144994 (male) PERU: Torontoy, Urubamba Canyon, 7800 ft; 7 July 1916 (Chapman and Cherrie)
     166512 (male)   "", 8000 ft; 28 April 1915 (Heller and Yale)
     474341 (female) PERU: Vitoc, Garita del Sol; 24 April 1893 (Kalinowski)*

“Most of these specimens (except "alticola" and 156058) have red thighs, although it is very little in some (e.g., 474341). Red on face variable, with the most on birds from Urubamba valley, least on Chachapoyas, Chanchamayo, Vitoc, and 144994. No red on leading edge of wing, but occasional bird will have red on upperwing.”


“Arndt (2005) maintained the two ‘alticola’ specimens as that taxon (asserting that one, presumably 129138, is a juvenile, although I am not certain I fully understand why — something about ‘bluish-green’ body plumage), but AMNH 144994, 156058, and 414340 were reidentified by him as A. hockingi. The Chachapoyas birds are the ones Arndt (2005) identified as A. mitrata chlorogenys (AMNH 235434 designated as the type of this taxon). I must give him credit for being the first to recognize and publish the presence of this species in Amazonas department.


“Meanwhile, back to MUSM specimens:

Aratinga wagleri:
Apurímac: Huayro: 5703 (F), 5707 (M)
Apurímac: Puente del Río Pampas: 5704 (M)
Apurímac: C'cnoq: 24281 (M), 24280 (F)
Apurímac: Pasaje: 25437 (?)
Cusco: Sumbayo cerca Pacaypata (Río Apurímac): 24289 (F)
“--These birds have red on the leading edge of the wing ([the leading-edge primary coverts], but not actually the alula), some on the very bend of the wing, and red tarsi. The red (orangey-red) on the crown reaches the orbit on most specimens. Only on 24280 is the red entirely lacking on the tarsi, the wing, and the red on the crown does not meet the orbit. This bird, presumably, is a juv and sounds like it agrees with that [which Tom Schulenberg] described. The crown color and the small size agree with wagleri minor, however, not with mitrata.

Ica; cerca la mina El Condor (Valle Pisco): MUSM 5715 (F), 5712 (M)
“--These frontata are apparently a juv (5715) and adult (5712). The former *lacks* red leading edge to the primary coverts, red at the bend of the wing, and red on the tarsi. Also, the red on the forehead of 5715 does not touch the orbit, but is the correct color to be wagleri. For this plumage to be present on a specimen from outside the range of any other Aratinga would suggest to me that these characters are not of hybrid origin, but are indicators of age in the species. I saw no other A. wagleri frontata that exhibited these characters in the MUSM collection.”


“An additional note that may be of interest here: here at LSUMZ we have three specimens of A. wagleri minor that lack most red on the bend of the wing and the tarsi (but have red foreheads). LSUMZ 75108 (Amazonas; Balsas) has *no* red on the bend of either wing, 80367 (Cajamarca; W of Balsas) has only two red feathers on the left wing (none on the right), and 91665 (La Libertad; Chagual on Marañon drainage) has two red feathers on the right wing, and one on the left wing. Of these specimens, only 91665 has the red reaching the orbit skin, contra Arndt’s (2005) comment that “…[A. w.] frontata and [A. w. ] minor from Ecuador and Peru have the red frontal band extending to the eye area and in addition, have red thighs, edges of wing and under-wing coverts.” Yet all three of these specimens are too small to be misidentified A. mitrata-types (e.g.: ‘A. hockingi’). This variation in the amount of red on the face within Peruvian A. wagleri populations adds more doubt to how useful this character is in separating A. mitrata-types into A. mitrata (sensu stricto) and A. hockingi and A. alticola rather than simply being age-related.


“Given the plumages of MUSM 24280 and 5715, and the locality of 5715 (which, I think, eliminates any possibility of a misidentified ‘mitrata’ type), and LSUMZ 75108, 80367, and 91665, it is clear that A. wagleri cannot be discounted as a possible source of confusion with the newly named A. hockingi solely by the presence or absence of red in the bend of the wing. This is the source of my dismay in my original proposal at the lack of comments in Arndt (2005) regarding more in-depth discussion in the separation of the taxa. There appear to be features that reliably separate the mitrata-types and the wagleri-types (color of red on crown, measurements), but these were not satisfactorily outlined in Arndt (2005). I am not necessarily arguing here that A. hockingi is based on misidentified specimens of A. wagleri, but rather that a description dealing with such difficult birds should make a point of clarifying the separation of the newly named taxon from any potential confusion species (particularly those which are potentially sympatric!).


“One thing I must concede is that the red-and-green Aratinga diversity in the Utcubamba valley (Amazonas department) is higher than I would have expected. There is a Hocking specimen of A. leucophthalma from Pedro Ruiz (with unclipped primaries) at MUSM, and additional support from comments during a discussion about Aratingas on BirdingPeru listserv



“Thus, it appears that within the area there are A. mitrata (sensu lato) and A. leucophthalma. I should add that in my original proposal, I did not say that Arndt (2005) proposed that three or four red-and-green Aratinga occur together syntopically, but rather in the same valleys (thus, arguably sympatric), and his maps (in addition to distributional information available to me of A. wagleri and now A. leucophthalma) do indeed strongly suggest this, although he did not necessarily say so explicitly in the text.


“I wonder now if the ‘local knowledge’ originally received by Hocking of two red-and-green Aratinga in the Utcubamba valley didn’t, in fact, involve A. leucophthalma? Local knowledge is chronically unreliable; confidence in the accuracy of where locals acquired specimens/live captive birds, how different sounds are, and how frequently events happen is usually low, in my experience of now more than 15 years in the field in Peru and elsewhere in South America. Furthermore, misunderstandings or ‘leading questions’ in interviews can add yet more error. These are primarily my concerns regarding Arndt’s use of ‘local knowledge’ in his paper. In his response above, Arndt states that he did not use ‘local knowledge’ or live birds to reach any of the conclusions of his paper (other than as anecdotal starting points for the investigation), but in his Results section, he bases his criteria for the identification of A. hockingi on ‘two adult birds, which had been captured by the local people.’ It is not clear if these two individuals are the same two live birds he mentioned in the Introduction, or if they became specimens. This, and the fact that most specimen material at MUSM (which are clearly important to his case) were obviously captive birds prior to collection, weakens his argument.


“I concentrate on vocalizations among red-and-green Aratingas in my proposal because I am desperately looking for potential ‘isolation mechanisms’ that might explain the high diversity of species in single valleys (Utcubamba, Huallaga, Mantaro, Apurímac, and Urubamba) that Arndt (2005) suggests (implicitly if not explicitly) must exist. I can only assume that likely isolation mechanisms would include: plumage characters, voice, and preferred breeding areas.

“Plumage is highly variable within each species, and it seems that among sympatric species of this group, the presence or absence of red on the wing of adults is perhaps the most important isolation mechanism (and is visible a long way away, which is how A. wagleri and A. leucophthalma may be able to coexist with A. mitrata, if only marginally—the three species generally separate by habitat and elevation). Red on the tarsi is variable (by age) and (in my experience) is bloody hard to see at any distance (the parrots, obviously, are more likely to see it close up than are human observers, but for humans to confirm the absence of red on the tarsi without capturing or collecting the individual is virtually impossible, hampering an observer’s ability to use this as a good field character when identifying Aratinga). Extent of red on the head is also highly variable within all forms, begging the question of how large a role does it truly play?


“Voice, so important a character among birds (including syntopic parrots such as Amazona and Aratinga elsewhere), does not seem to pan out given present evidence. Within the red-and-green Aratinga complex, the members are probably not safely distinguishable by voice and, hardly surprisingly, are largely allopatric. I’d wager that widely sympatric species such as A. mitrata, A. alticola, and A. hockingi would likely offset common vocalizations so they would be less similar, facilitating the detection and identification of A. hockingi and A. alticola (by human observers) and the cohesion of flocks. So far, there is no support for this hypothesis other than Arndt’s (2005) anecdotal comment in the introduction of the paper. Now with birders searching for these species, I’d expect that recordings documenting these differences would come to light should these differences be real. However, I have yet to see evidence of birders detecting and documenting A. hockingi or A. alticola, much less confirming vocal differences. Where are these species?


“Breeding habitat has the least potential as a successful mechanism in my opinion, because the red-and-green Aratingas of these valleys are largely (wholly?) restricted to nesting on cliffs, a limited resource, so they are likely to overlap greatly at breeding cliffs. In addition, neither Arndt (2005), nor anyone else, has provided any real evidence to suggest that there truly is separation of breeding areas. Arndt (2005) provided conjecture that A. hockingi and A. alticola are birds that breed more in cloud forest than does A. mitrata (sensu stricto), using anecdotal evidence from specimen localities, but these specimens’ breeding condition is not noted, and Arndt admits that Aratinga are capable of moving widely as they track food resources. If such habitat niche-partitioning was real, then I’d expect that A. hockingi would be the common red-and-green Aratinga found, for example, at Abra Patricia in dpto. San Martín, near the Utcubamba valley, but with humid cloud-forest habitat. However, the red-and-green Aratinga reported there is A. leucophthalma [see comments posted to BirdingPeru listserv]!


“As for the PSC and BSC issue, I was simply pointing out that although Arndt (2005) stated that he evoked the PSC as his species concept in this paper, it is clear he used the BSC, which (I should have further pointed out, but did not) is also supported by the fact that he named subspecies for A. mitrata … something which is at odds with the tenants of the PSC (under the PSC, the concept of subspecies is unacceptable). As I stated in my original proposal, this was a pedantic sidebar only, but something I thought worthy of note.


“So in conclusion, in Dr. Arndt’s response, he claims that I did not review Aratinga specimens prior to my proposal comments (I did), that I do not sufficiently address the matter of validity of A. hockingi and A. alticola (I hope that what I provide above does this better), that he used ‘local knowledge’ only anecdotally (not entirely true), that he was justified in excluding discussion of A. wagleri minor and frontata from the paper because they are easily separated from the taxa in question (they are not, based on the characters he invokes), and that overall my proposal was unbalanced. Well, perhaps I made some heavy-handed statements in the proposal, and I apologize if I offended him, but I believe that my conclusion in the original proposal — namely that the evidence presented in Arndt (2005) is not enough to support the recognition of A. hockingi and A. alticola as species separate from A. mitrata — is still sound. Again, I will say that I think a molecular study will be very helpful, if not in fact necessary, to lay my concerns to rest. If Dr. Arndt can convince me otherwise, I’d be happy to capitulate.”


Comments from Thomas Arndt: Skepticism is one thing, but fighting for being right until one party “capitulates” is different and unhelpful. Mr. Lane’s technique is simple: questioning everything. Therefore, discussing with him or even convincing him is probably impossible. How should one respond to someone whose concern is that misunderstandings or ‘leading questions’ in interviews could add error to the paper? Or how should one assess when Mr. Lane feels that a sample size of 22 birds is not large enough or identifies birds in the following picture:

( as “nominate-like” birds?


“If he had been willing to have a closer look to the above-mentioned photo (which is a bit dark but after brightened very clear), he would have seen that the four birds are very typical chlorogenys.  Even if one takes in account that observers often are confused due to the variation depending on the birds’ individual, sex and age variation, it is difficult to misidentify these birds as A. mitrata mitrata.  Perhaps I should remind Mr. Lane that a few years ago he still misidentified similar chlorogenys representatives from Chachapoyas as Aratinga wagleri when examining the skins in AMNH. I wonder why he changed his mind so completely.


“It is impressive to see Mr. Lane’s documentation to prove that he reviewed Aratinga specimens prior to his proposal comments, and I apologize for assuming that he did not. But in spite of his enormous work his extensive statements can be reduced to the question if alticola and hockingi are representatives of immature mitrata or minor or not.  Apparently he feels very uncertain about what immature mitrata and minor look like. Therefore, he uses every available specimen - MUSM 24280 (minor, F) from Apurímac or MUSM 5715 (frontata, F) from La Mina El Condor (Valle Pisco) – as an argument to prove that his doubts are justified.  It does not bother him that he already knows that the color of the crown (according to him) is “orange-red” or the size does not fit (in frontata) with hockingi.  At least he is “not necessarily arguing (…) that A. hockingi is based on misidentified specimens of A. wagleri”.  But he continues “that a description dealing with such difficult birds should make a point of clarifying the separation of the newly named taxon from any potential confusion species (particularly those which are potentially sympatric!)”. It is self-evident that he can ask for a better clarification, but as I already explained: it seemed not necessary to me. Therefore, I am grateful that Mr. Lane supported and clarified this by discussing the above-mentioned specimens.


“Mr. Lane also wonders how immature mitrata look. In this case, he found the answer by using pictures from the internet ( To him the photo shows “an adult A. mitrata flanked by two offspring… showing characters of A. hockingi/alticola.” Yes, he is right in identifying two immature A. mitrata which have a typical small darkish red forehead mixed with small green feathers at the borders and extending to the periophthalmic ring.  The coloration and shape is typical for a good number (not for all!) immature A. mitrata and applies to all populations, also for Argentine birds (e.g. it is more than likely that the adult bird in the picture represents a bird imported from Argentina, from where in fact most captive mitrata were exported).  I am mentioning this because Mr. Lane proves again that he has a limited knowledge of adult and immature plumage when he erroneously expects that immature A. mitrata tucumana to resemble adults (or already have a greater extension of red to the head) and reckons “that this would explain why ‘hockingi/alticola’-plumaged birds are only found in the north and center of A. mitrata’s range, whereas (the paper) did not report any birds of this plumage in the southern part (i.e., Argentina)”.


“I can assure him that I was aware of this immature plumage, which only shows characters of A. hockingi/alticola at first sight but differs clearly when thorough examined.


“I can understand that Mr. Lane in his long response tries to justify his original comments, but to go into each of his arguments brings us nowhere. It is also not really clear in his response: Does he want the committee to vote about the quality of the paper or the validity of the taxa? What I am missing, too, is how he would treat the existing alticola and hockingi skins (again here the link with the samples: Just dismiss them as immatures of mitrata or wagleri does not work and questioning everything just to call it into question is anything but helpful.


“Yes, a molecular study would be helpful, although it is not a solution for everything, and I doubt that the South American Classification Committee is willing to reject all future proposals not supported by genetic analysis of mtDNA.”


Comments from Stiles:  “NO.  I find Lane’s doubts reasonable and not wholly resolved by Arndt; what is needed is better analysis of good series of wild-taken specimens with full data and associated genetic samples and recordings of vocalizations to resolve this one.”


Comments from Pacheco:  NO. ‘Extraordinary claims require extraordinary evidence’. Após ler os questionamentos de Dan Lane torna-se evidente concluir que o trabalho de Arndt não imunizou os resultados das interpretações alternativas, em boa parte mais esperadas.”


Comments from Nores:  “NO.  Regardless of whether Arndt is right or not in relation to the newly described species and subspecies, the evidence presented in Arndt (2005) as proof is insufficient to show that there are multiple biological species in the red-fronted Aratinga group.  A recent paper by Agnolin (2009) shows clear evidence that the subspecies tucumana described by Arndt in 2006 is not valid.  Moreover, it is widely accepted that the information provided by the local people must not be taken in account in scientific works.  As suggested by Lane, I think that it is recommendable that the SACC vote NO to this proposal, at least until it can be reevaluated with molecular analysis.”


Comments from Jaramillo:  “NO.  Dr. Arndt makes great points, but the basic issue here is that there is the need to clarify the extent of sympatry, if any, between all of these forms. To clarify the vocal differences and document them, and to clarify the issue of individual variation that may be involved here. I do think he is on to some greater pattern in variation that is present here, but he has not convinced me of what that pattern exactly is and how it should affect taxonomy of this group.”


Additional comments from Remsen: I looked at Agnolin (2009)* after seeing Manuel’s comments and found that that author considered tucumana Arndt, 2006, to represent individual variation in the extent of red in A. m. mitrata … in other words, precisely the problem pointed out above with recognizing hockingi as a separate species or alticola as a valid taxon.  Also, as noted by Lane above, I wonder if hockingi and alticola aren’t the same thing – if you look at the very nice plate in Arndt’s paper, they differ primarily in the subtle extent of red on the forecrown, which would seem to be potentially within the range of individual and age variation within any of the valid taxa.


* Agnolin, F. L.  (2009) Sobre el complejo Aratinga mitrata (Psittaciformes: Psittacidae) en el noroeste argentino. Comentarios sistemáticos.  Notúlas Faunísticas, 31, 1-5.


Dan Lane addendum 22 June 2013:

In light of some recent experiences in the Rio Urubamba valley (Cusco dept.) and Rio Utcubamba valley (Amazonas dept.) in June 2011 and June 2013, respectively, I have had the chance to further solidify my thoughts on the status of Aratinga ‘hockingi’ and A. ‘alticola’ with respect to A. mitrata. In both localities, I was able to detect, mixed with flocks of birds with undeniable A. mitrata (sensu Arndt 2005) plumage characters, birds that had characters of ‘hockingi’ and ‘alticola’. In both cases, these individuals were decidedly not paired, as were most typical A. mitrata-types. In addition, I made recordings of these flocks, and no particularly unusual vocalizations are detectable-- but to be fair, it was impossible to separate the vocalizations of the ‘hockingi’ and ‘alticola’ birds from the rest of the flock. Despite scanning as many flocks as I could while visiting these sites, I was not able to detect flocks composed solely of A. ‘hockingi’ or A. ‘alticola’-type birds.

            In addition to these observations, in June 2011, I systematically photographed each Aratinga mitrata/hockingi/alticola specimen present at the Museo de Historia Natural de la Universidad Mayor de San Marcos in Lima, Peru, to have a record of their plumage features. One thing that struck me as I investigated the specimens identified by Arndt (2005) as A. hockingi and A. alticola was the uniformity of the ages of feathers of the body plumage. That is to say, there was no indication that these birds were undergoing typical adult body molt as is exhibited by most Neotropical parrots I know. In general, after a parrot attains the first complete plumage cycle, it starts a raptor-like molt strategy in its body plumage in which there is a slow, but constant, state of feather replacement (Howell 2010). This strategy results in a motley appearance, as patches of feathers of different ages and wear conditions contrast with one another. Such a motley appearance is completely lacking in the specimens of A. hockingi and A. alticola, and to me this suggests that these individuals must still be in their first complete plumage cycle, i.e., they are juveniles! This difference in feather age, wear, and (presumably) structure would also explain the different color green these individuals exhibit.

            These two bits of data—the fact that birds with A. hockingi and A. alticola plumage features are present as unpaired birds within flocks of A. mitrata and that they lack adult molt patterns—suggest to me that the only reasonable conclusion that can be reached is that the names ‘hockingi’ and ‘alticola’ refer to variations within the juvenile plumages of A. mitrata, and have no taxonomic standing whatsoever. Those arguing otherwise have the burden of proof on their shoulders and will require extensive data showing that these birds are maintaining themselves as separate taxonomic entities from their A. mitrata cohorts and have a decidedly unusual molt strategy.


Literature cited:


ARNDT, T. 2005. A revision of the Aratinga mitrata complex, with the description of one new species, two new subspecies, and species-level status of Aratinga alticola. J. Orn. 147: 73-86.

HOWELL, S. N. G. 2010. Peterson Reference Guide to Molt in North American Birds. Houghton Mifflin.