Dear Dan, to my information the population of Herpsilochmus rufimarginatus in Bahia, Brasil belongs to subspecies scapularis. Do you have a specific reason to attribute your recording to the nominate subspecies? Kind regards, Erik
Laziness, mostly. I see you are right. Thanks for the correction!
Since the type locality of 'scapularis' is 'Bahia' (I would presume this means from around Salvador), this means that the other northeast Brazilian cuts on XC of this species should be relabeled, as well. Other cuts of scapularis would likely be XC80200, XC80214, XC80215, XC80220 (the song of this taxon seem very like nominate). Meanwhile, XC80216, XC80219, XC7949, and XC13399 must be another taxon by the difference in voice. To me, they sound very much like H. r. frater (type locality Sarayacu, Ecuador)... frater has a most remarkable circum-Amazonian distribution, but it sounds rather uniform all around! As an aside, the two cuts from Panama should be exiguus (type locality Cana, Darien).
Herpsilochmus rufimarginatus rufimarginatus is one of the few subspecies that I thought I could specify on XC without adding the caveat "ssp asumed from location".
In the recordings on XC from eastern Brazil, the birds in the north sing quite differently from those in the south. The most southern examples of the northern form are from Murici and Flexeiras in Alagoas and the most northern example of the southern form is XC82519 from Cardeal da Silva and XC303810 from Bonito, both in Bahia. I have always assumed that the northern birds were scapularis and the southern birds were rufimarginatus, relying on the vocal difference.
The fact that the type locality of scapularis is 'Bahia' does not of itself mean that the recordings from central and southern Bahia are scapularis, not rufimarginatus. The type locality of scapularis could be in northern Bahia and the dividing line between the two subspecies could be somewhere between the type locality and Bonito and Cardeal da Silva.
I don't have access to the descriptions of these two subspecies. How are they distinguished?
Cardeal da Silva, Bahia., Brazil
Jeremy,
You may find this thesis of interest.
https://repositorio.ufrn.br/jspui/handle/123456789/18132
Shaun
...and, of course, there is Peter Boesman's note (http://www.hbw.com/node/931929) which I believe led to the split of nominate (from all other subspecies) in Volume 2 of the HBW/BLI Illustrated Checklist of the BIrds of the World (which I have not yet seen).
Sorry to suggest dissent here, but I'm afraid Peter's comments in his HBW note are flawed by not checking the actual identification of the cuts he included under "scapularis". The type locality of that name is "Bahia". "Bahia" as a type locality of early named taxa (18th and 19th centuries) refers to the environs around the present day city of Salvador (then known by the name "Bahia"). Collectors would have been working in the forests around that city. So, given that XC's Bahia recordings are largely from south of that city, Jeremy could well be right that there is a changeover in song type within Bahia state around Salvador and north. However, that doesn't pan out if you listen to recordings from Bahia immediately north of Salvador, such as XC82519 and this one I found on Wikiaves: http://www.wikiaves.com/383366&tm=s&t=s&s=10886&p=1 These cuts suggest that the 'nominate' songtype encompasses Salvador, so that the name scapularis almost certainly is a nominate-sounding taxon. Unless there is evidence to suggest that plumage characters distinguish scapularis from nominate, it is likely that the two are synonymous (Hellmayr and Zimmer do not delve into the name scapularis, the former synonymized it with nominate and the latter said he didn't have Peruvian material of H. rufimarginatus, and therefore didn't discuss the species in depth at all). I point out that this appears to be the result reached by Da Silva in the dissertation Shaun has linked to above.
Vocally, birds from Alagoas and Pernambuco clearly are something apart from birds in Bahia and south on the Atlantic coast of Brazil. There appears to be no name available for them, but since the name "frater" is applied to birds from the Andes to at least Maranhao state in NE Brazil, it seems reasonable to consider that that name may well apply to birds from Alagoas and Pernambuco. By voicetype, it is clear that the name "scapularis" cannot be applied to birds from those states. Da Silva, in his dissertation, appears to be on the same scent, and he proposes that this population is a new species (!), but doesn't give a name (perhaps saving that for a formal published description).
In any event, this is yet another case that makes me wish that XC contributors would be a bit more careful about what subspecific names they use when uploading and identifying cuts! This case, in particular, may well have caused HBW/Birdlife to use the wrong name in their taxonomy thanks to a group of recordings that have been mis-named to subspecies. Unless the contributor makes the effort to research the subspecies name, rather than simply using those names given by IOC or HBW, compounded error such as this case will arise more commonly than I think many would expect!
Cardeal da Silva, Bahia., Brazil
Hi Dan, I just want to comment that in my Ornithological Note I did not go that far as checking type localities etc. (given this is just one out of some 800 cases I tackled), but I DID clearly indicate there was some issue here that needed further research, I quote "This is assuming that scapularis occurs only south to S Alagoas, while birds further south are nominate (different from HBW Vol. 8 !)." Apart from giving the right names (which obviously is important, but in quite some cases requires real detective work to find out which name has priority for which taxon), I think everybody agrees that there is a clear vocal break, with the southern birds being vocally distinct. Greetings, Peter.
Some further input re. Dan's general comment about subspecies assignation. During my use of XC recordings for the above-mentioned Ornithological Notes, I have indeed stumbled on several hundreds (!) of recordings for which the subspecies name was clearly not correct. My own interim conclusion was that while identification to species level was on XC very reliable, in fact much more so than other on-line databases (thanks to the forum members!), subspecies assignation was definitely not reliable (despite the effort of some forum members such as Erik Eggenkamp to correct as many as possible). There is no easy solution to this issue. Few recordists have access to literature other than e.g. HBW, let alone they can check the original type description, and even then this doesn't tell us always the exact range for every taxon. Fortunately, there are many cases where things are more straightforward. At the other hand, the mapping tools of XC motivate people to assign up to subspecies level, as this is a very interesting feature. As Dan says, some caution is needed... Hopefully at some point, the map tools of Rolf de By will create an alert whenever a subspecies name is given from a locality which does not fit the known range for that taxon! A request for our wish list :-) Greetings, Peter
Thank you, Shaun, for pointing me to Marcelo da Silva's dissertation and Peter Boesman's note. Can you or anyone else send me a pdf of Tobias, J.A., Seddon, N., Spottiswoode, C.N., Pilgrim, J.D., Fishpool, L.D.C. & Collar, N.J. (2010). Quantitative criteria for species delimitation. Ibis 152(4): 724–746?
I see no reason to alter the assignment of my recordings of Herpsilochmus rufimarginatus from Bahia to subspecies rufimarginatus but how should we treat recordings from Alagoas and Pernambuco? Marcelo da Silva's thesis that scapularis is a junior synonym of rufimarginatus and that the birds in Alagoas and Pernambuco belong to a new species seems to me persuasive. I'd be inclined to classify these recordings as 'ssp. nov.' till the species is described, with a note outlining the taxonomic situation. The alternative of continuing to call them scapularis seems to fly in the face of the evidence and to leave the subspecies blank would be rather a cop out.
What do people suggest?
I respectfully disagree with Peter's statement that "Few recordists have access to literature other than e.g. HBW." I wrote an entire XC article dedicated to the topic complete with links to Peters' checklist, Hellmayr's Birds of the Americas, Zimmer's papers, and other references. If someone can get online to upload recordings, they have access to these invaluable references. It is a case of whether they are willing to do the research or not.
Answering Jeremy's question in the last comment above: how about simply not assigning a subspecific taxon at all? It is entirely acceptable and does not threaten to misinform users. I don't understand why more contributors don't do it.
Hi Dan, Sorry, forgot about your excellent XC article! Then indeed it boils down to only specifying subspecies when one has done proper research or when there is no uncertainty, the latter obviously being somewhat subjective. Greetings! Peter
To follow up on this: I only just realized that HBW Alive has acted on Peter's report and has split nominate H. rufimarginatus from the remainder of that group, and awarded the latter the name scapularis, which as we have just demonstrated, is not the correct name for any of the populations therein!! This illustrates the danger of applying taxonomic names without more carefully investigating to which population that name correctly belongs.
Excellent discussion!! This truly is science acting in it's purest form. This wouldn't have happened if we just stuck to the specific level on XC.
Dan is of course correct when calling for caution when assigning subspecies, although this is a case in which a subspecies previously assumed to be one (and generally treated as such e.g. in HBW 2003), now seems to be rather two, primarily based on voice. There are several more cases in Brazil. E. g. what is presently known as Rufous Gnateater subspecies Conopophaga lineata vulgaris clearly has two vocal groups meeting just north of Sao Paolo city. Here as well, a split into two different taxa seems required. It is thanks to the many recordists that such findings become obvious. Indeed science in action!
I think it's more accurate to say that the wrong subspecies name was applied to a population rather than to say that "one subspecies now seems to be two." Unless the original description specifically stated that birds from Alagoas south to Espiritu Santo were all called scapularis, it was a misapplication of the name to the northern population by subsequent authors. This is the very nut of the problem, and the reason that it is important not to compound error when labeling birds with subspecific names, but rather to be sure that you are using the correct one... or simply not to use one if you are unsure. Now, someone will have to explain to HBW/Birdlife that they have to use the name "frater" for the Northern Rufous-winged Antwren because "scapularis" is actually a taxon of Southern Rufous-winged Antwren...
In this particular case, Maximilian zu Wied-Neuwied wrote in his fantastic work "Beiträge zur Naturgeschichte von Brasilien" (1830-1831) on p. 1086 about scapularis (translated into English): <>. "Sertão" refers to the semi-arid region in Northeastern Brazil comprising parts of the states of Alagoas, Bahia, Pernambuco, Paraíba, Rio Grande do Norte, Ceará, Maranhão, Piauí, and parts of northern Minas Gerais. It was thus not immediately around Salvador, since this coastal town lies in another subregion (Zona da Mata). We will thus have to prove that the northern song-type does not occur in Bahia to exclude the possibility that scapularis was given to that population, unless the type specimen allows to do so...
Oops, the text between <> brackets disappeared ! OK, I'll translate again: "This species lives in the large woods of interior Brazil. I found it in the Sertao of the Bahia province."
Just rereading this conversation, I note that the second to the last comment above by Peter suggests that the type locality of scapularis would be from the interior of Bahia, not the coast around Salvador, and therefore it could be, conceivably, correctly applied to the NE population that also occurs in the interior semi-arid habitats of "Alagoas, Bahia, Pernambuco, Paraíba, Rio Grande do Norte, Ceará, Maranhão, Piauí, and parts of northern Minas Gerais". However, the XC map for H. rufimarginatus does not show the species to be found in the "Sertao" of Paraiba, Pernambuco, Alagoas (it does not appear to exist in Ceara, Rio Grande do Norte, or Piauí), and the only cuts from the interior of Bahia--including in the background here (the northernmost interior localities in the state for which recordings are available): XC80910, XC283321--are still of the southeastern song type, representing the nominate group. However you cut it, it appears that the name scapularis cannot be applied to birds from the northeast shoulder of Brazil!
Yellow-olive Flatbill (Tolmomyias sulphurescens pallescens (ssp assumed from location)) · call, song
Looks like stuff for an article!
Marcelo da Silva has already written it - his dissertation, refenced above. He splits Herpsilochmus rufimarginatus into H. frater (in Amazonia), a sp. nov. (eastern Brazil north of the Rio São Francisco) and H. rufimarginatus (eastern Brazil from Bahia south). He shows that the sp. nov. is distinguishable from Amazonian H. frater and that scapularis is a junior synonym of rufimarginatus.
The HBW has got it hopelessly wrong, as Dan has pointed out. It splits H. rufimarginatus into two species, with rufimarginatus restricted to SE Brazil and H. scapularis in the rest of South America. This leaves the populations in eastern Brazil that sing two completely different songs in the same subspecies, H. s. scapularis.
I do hope that da Silva will publish this somewhere. I need to reread the dissertation to understand what the distinctions are that mark the new species in the NE. It seems to fit better within the frater group to me.
According to the cataloguing data, the dissertation is available in the library of the Universidade Federal do Rio Grande do Norte. Although the work is admittedly not easily accessible to the scientific community, doesn't this constitute publication? If da Silva had named the sp. nov. wouldn't this have been a valid description of a new species?
Gentlemen,
this comes late after this thread was started/ended, but just to inform that in the currently ongoing XC map review process I have just updated our maps to reflect present understanding of above Herpsilochmus taxa distributions, and I have uploaded now a new map for H. rufimarginatus for IOCv9.1.
XC maps currently reflect IOC v9.1 and the above work will take us into v11.1. We hope to make subversion steps from 9.1 upwards one at the time in the coming weeks.
The new map may take a few days before Google has cleared its cache and the old version has disappeared. When we push XC to IOC v.11.1 the split nominate/frater will be effectuated also.
With best, Rolf
And for those interested: I also made a brief comparison of call notes, and found these to be quite ditinctive as well for both groups. See:
https://birdsoftheworld.org/bow/species/ruwant4/cur/introduction#vocal
https://birdsoftheworld.org/bow/species/ruwant3/cur/sounds
greetings! Peter
Log in to reply to this topic.
2013-01-29