'Oregon Junco' taxonomy

'Oregon Junco' taxonomy

Until recently, there were some issues with the taxonomy of Dark-eyed Junco (Junco hyemalis) on iNat. Those issues have now been corrected, but there are still a lot of records in the database that are attributed to the incorrect subspecies or that do not have sufficient evidence to assign a subspecies. Hopefully this post helps explain the issue.

First, brown-backed Dark-eyed Junco are often referred to as 'Oregon Junco' in field guides and other resources. Birds with this phenotype belong to one of eight well-described subspecies:

J. h. montanus
J. h. oreganus
J. h. pinosus
J. h. pontilis
J. h. shufeldti
J. h. thurberi
J. h. townsendi
J. h. mearnsi

Prior to the recent fix, the common name 'Oregon Junco' was applied only to the ssp J. h. oreganus in the iNat database, and most users were using that taxon for any brown-backed birds, regardless of their actual subspecies. In reality, many individuals cannot be assigned to subspecies from photos, especially in winter when 3 or more subspecies can occur in some areas. Some individuals with exceptional photos can be assigned to subspecies, especially records from the breeding season when the subspecies are geographically separated.

Below is a short description of the subspecies and their ranges. Note that many of the descriptions overlap, especially among the paler and more variable subspecies, and that intergrades likely are common near the boundaries of ranges. Much of this information is referenced from Cornell's 'Birds of North America'. Generally, females are not identifiable to subspecies. The subspecies are ordered from north to south in terms of their breeding range.

J. h. oreganus

Breeding range: Coastal from SE Alaska to southern British Columbia. Some individuals are year-round residents in this area and may move to lower elevations and feeders for the winter.
Wintering range: Mostly the Pacific northwest, but can be found anywhere in the west.
Male description: The darkest and most contrasting subspecies in general. Back is dark reddish brown, flanks dark reddish brown, hood typically black.

J. h. montanus

Breeding range: Central British Columbia south along the east slope of the Cascades through Oregon, extending east into NW Idaho and NW Montana.
Wintering range: primarily in lower elevation mountainous areas in Arizona, Colorado, and New Mexico but can be found virtually anywhere in the western half of the United States.
Male description: Difficult to ID and variable. Intermediate between paler southern breeders and darker northern breeders, very similar to ssp. shufeldti. Back is dark grayish brown, flanks are cinnamon from, hood varies from slate colored to nearly black. Typically less red overall than ssp. oreganus.

J. h. shufeldti

Breeding range: Pacific coast from southern British Columbia south to central Oregon.
Wintering range: Primarily along Pacific coast in breeding range and south through southern California.
Male description: Very similar to ssp. montanus and similar to ssp. oreganus but paler. Back is brownish with less red than ssp. oreganus, flanks are cinnamon brown, hood is typically black.

J. h. thurberi

Range: Nearly all mountainous regions in California, with the exception of the coast from San Francisco to Monterrey County (where it is replaced by ssp. pinosus). Also breeds in coastal southern Oregon and some areas of far western Nevada. Established in low elevation exotic Eucalyptus plantings in some areas of coastal southern California. Resident over most of its range, but descends to lower elevations in winter especially in the Sierra Nevada.
Male description: This is probably the most reported subspecies to iNat, given its year-round proximity to large population centers in California. Overall a relatively pale subspecies with minimal reddish tones. Back is brownish, sometimes with a pinkish tone, flanks are cinnamon brown, hood is blackish but variable. Sierra birds and northern populations typically are darker than birds from southern California.

J. h. pinosus

Range: San Francisco south to Monterrey County. Mostly a non-migratory resident but may move to lower elevations in winter.
Male description: A pale-headed, brightly colored subspecies. Back is bright reddish brown, flanks are bright cinnamon, hood is slate-colored or blackish. Typically brighter and more red in plumage than neighboring ssp. thurberi populations.

J. h. pontilis

Range: Resident in a small area around Laguna Hanson in northern Baja California.
Male description: Overall a dully colored bird, similar to ssp. thurberi but with pinkish flanks.

J. h. townsendi

Range: Resident in mountains of Baja California, with some individuals moving to lower elevations in winter.
Male description: Possibly the most distinct subspecies in the Oregon Group and intermediate between ssp. thurberi and 'Pink-sided Junco' (ssp. mearnsi). Overall a very dull and pink bird. Back is dull brown, flanks are pale pinkish brown, hood is the grayish and paler than most other subspecies. Sexual dimorphism is weak in this subspecies.

J. h. mearnsi (often referred to as 'Pink-sided Junco)

Breeding range: Mountainous areas from southern Alberta south through Idaho, Montana, and NE Wyoming.
Winter range: Utah east through Nebraska and south through Texas to the Chihuahuan Desert.
Male description: A distinctive subspecies, typically field identifiable from other brown-backed subspecies. Back is dull brown, infused with pink, flanks are pinkish brown, hood is grayish with dark lores. Females similar to males.

Posted on March 01, 2018 06:31 PM by fogartyf fogartyf


So I wonder what the breeding birds in the Albertan Rockies are...they are apparently not J. h. mearnsi, but "Oregon" in nature.

Posted by silversea_starsong almost 6 years ago

Thanks for doing this and clearing up where many of us have gone wrong in the past.

Posted by finatic almost 6 years ago

Yeah this is a really useful reference...thanks for taking the time to type it up for us!

Posted by silversea_starsong almost 6 years ago

I think the Albertan Rockies birds probably are either shufeldti or montanus - those to are really similar and have sometimes been treated as the same subspecies. They probably aren't field identfiable except by range. Wouldn't be surprised if they also intergrade with Pink-sided a bit in that area, especially closer to the U.S. border.

Posted by fogartyf almost 6 years ago

Thanks for doing this. Aside from birds in the hand, how many iNaturalist observations of "Oregon" type juncos in the field do you think can be safely identified to subspecies based on the photos? Any? Is it ok to id summer birds based on breeding range?

eBird has an "oreganus" group to keep track of the Oregon subspecies which are not safely identifiable. Is there a way to include an ad-hoc group like this on iNaturalist so that id's can be refined? Classifying a junco on iNaturalist as an "Oregon" type junco is still useful and provides more information than leaving the bird unclassified.

Do you know where iNaturalist gets its subspecies from? I'm assuming maybe the Clements checklist? I don't think AOU tracks subspecies.

Posted by andy71 almost 6 years ago

I think some individuals can be safely ID'd to subspecies from photos. My approach has been to only label individuals on the breeding range that have good photos that fit the classic 'type' for that subspecies. I think it's pretty safe to assume a bird in July in the Sierra Nevada that fits the description of thurberi is a thurberi, for example.

I don't think it's possible to include groups like that on iNat, or it's at least discouraged. There was some discussion on the google groups about this for the similar issue of having a taxon for Rufous/Allen's Hummingbird (again like eBird) and the decision was to remove that option and to only have genus and species options. Doing so for Oregon Juncos would be even more muddled, as I don't know if it's clear that the 8 subspecies form a monophyletic group and you might need to include Pink-sided subspecies to achieve monophyly.

I'm not sure that I agree that classifying birds as 'Oregon-type' is all that informative or useful. It's a convenient label that observers can quickly recognize, but its evolutionary, ecological, and taxonomic significance is questionable given what I mentioned above.

Yes, iNat subspecies (and species) come from Clement's. Your correct that the AOS doesn't include subspecies and hasn't for decades.

Posted by fogartyf almost 6 years ago

If we do this then it may be useful to make fields for the groups ("Oregon" group, for instance) so that when observations are kicked to species level this data is still available.

Posted by silversea_starsong almost 6 years ago

Thanks again. Seems like this is really a problem created by the field guides.

You've probably already thought of this, but it would be awesome to have an iNaturalist "guide" for the "real" Junco subspecies. It's pretty hard to find definitive photos of the subspecies within the Oregon group online. I'm not sure where most users would go to even begin.

Posted by andy71 almost 6 years ago

A guide would be good to do at some point - maybe when I find some free time for it!

iNat is probably a great resource for images of most of the subspecies (unfortunately, most of them are misclassified right now!). There certainly are plenty of excellent images for the widespread ssp (not sure about the two Baja ssp).

Posted by fogartyf almost 6 years ago

Frank, re your comment on Rufous/Allen's Hummingbird.. this is what is known as a "slash" in eBird, meaning one of two possible species. They are common for superficially similar species like Rufous and Allen's or Downy/Hairy Woodpecker. iNat is right to not support these, nor does Cornell count them in their data analyses.

Subspecies groups are different, and like @andy71 I believe they are useful and should be a part of this community. An non-Slate-Coloured observed in eastern North America, for example, would be of interest to many.

Posted by dserp almost 6 years ago

That's why I think fields may be an acceptable compromise for going ahead with the individual subspecies taxa.

Posted by silversea_starsong almost 6 years ago

@xyz the issues are similar, though, in that they involve creating an artificial grouping for multiple taxa that can be difficult to distinguish (in one case, those are species and in the other subspecies, but the taxonomic level of such a grouping seems irrelevant). In eBird, the slash doesn't have to only be for two species and many of these groupings don't include slashes (e.g. diurnal raptor sp or eagle sp or [oreganus group]). They just represent the fact that the unit of classification is paraphyletic or polyphyletic with respect to the existing taxonomy. eBird has these type of groupings at almost every taxonomic level, including subspecies for DEJU.

I don't have any control over whether subspecies groups exist or are allowed in iNat and can only point to past discussions where they have been removed or discouraged. They currently do not exist for any subspecies of DEJU and haven't existed, as far as I've seen, so this is an effort to clean up the data we have within the framework that iNat provides.

Posted by fogartyf almost 6 years ago

It's worth noting that both Rising (Sparrows of the United States and Canada) and Pyle (Identification Guide to North American Birds, Part I) sort several subspecies into the "oreganus" group which mirrors the original (pre-1973) split . Most of these groups are actually pretty straight-forward to sort to group. This is also the way Fox Sparrows are handled. I agree that using the trinomial for the group name AND one of the subspecies names (in most cases the type specimen from the original species description) is inelegant, but it is useful as a house-keeping mechanism. Fox Sparrows will almost certainly be split in the foreseeable future and the jury is still out on whether some or all of the juncos should be re-split. This is the logic both Pyle and Rising (as well a Zink and the BBL) use in dealing with recognizable forms. As a matter of house-keeping iNaturalist should negotiate a mechanism for sorting to recognizable group, either adopting the mechanism used by Pyle and Rising or something similar to the mechanism used by eBird.

Posted by mikepatterson almost 6 years ago

Thanks for doing this Frank. Fox Sparrows are even worse.
But I'm not sure this couldn't be accomplished with something like "Dark-eyed 'Oregon' Junco (ssp. oregonus, etc.)" only because these are not separable from photos other than inferring ssp based on date, which isn't an identification issue, but one that could be done all at once when iNat finally improves bulk editing.
If I'm interpreting what you're doing, thurberi and the others will be lumped as just J. hyemalis; this is ok, but it does reduce expression of specificity conferred by retaining "Oregon-type". It's substituting one type of vagueness for another, basically. To say "well, they're not technically oreganus" is almost a semantic objection since the people who keyed in the names obviously weren't taxonomists and it's basically a typo (annoying, but we all know what they meant). I do see a lot of value retaining an "oregonus-type" designation in iNat.

Posted by dcoopercem almost 6 years ago

Dan, I agree we are losing information by not recording something about the phenotype of the juncos. But as Frank says this is the taxonomic system that iNat set up.

Getting a bit off topic here but, I wonder if iNat could allow multiple subspecies tags that aren't necessarily monophyletic. (Not just for Juncos but for all species). So if you see an Oregon type junco in California in winter you would probably include tags for ssp. oreganus, montanus, shufeldti and thurberi. Others identifiers could include their own list of tags. Then we would be following the taxonomy and retaining some information on subspecies. Maybe all subspecies are included by default when you ID something and identifiers can remove subspecies to refine. Curious what you all think of this.

Posted by andy71 almost 6 years ago

Ugh, I just looked at my own Fox Sparrow records and realized I screwed them up.

Posted by vermfly almost 6 years ago

The problem as I see it is iNat was created for the general public, not for taxonomists or purists although the latter count among its user base. The general public uses field guides and apps to assist with identification, resources which have long used non-monophyletic groupings. When a typical user sees a junco, IDs it as an Oregon-type Junco in their Sibley Guide, and comes to iNat to record it, we should provide a way for them to enter that data, whether as its own taxon or as a field or tag. More data is always better even if it is not taxonomically pure.

Posted by dserp almost 6 years ago


Posted by dcoopercem almost 6 years ago

A little of topic, but if there are several ssp. that can be placed into a larger sub-group within a species, like Oregon type Junco, then doesn't that warrant that sub-group being elevated to species? 'Oregon' Juncos are clearly distinct and identifiable in the field, even if they are multiple different ssp. Another problem with the Dark-eyed Juncos is some sub-groups are clearly sexually dimorphic, like Oregon, and Slate-colored, and other groups, like Gray-headed,Red-backed and Pink-sided, are not sexually dimorphic, which implies a distinct breeding difference. The argument against spliting is largely based on DNA evidence they only separated during the last Ice Age and therefore have not had enough time to evolve into new species, which is based on the assumption evolution happens at a slow constant rate, but a large amount of evidence suggests evolution occurs in rapid bursts, with often long periods of stagnation during stable regimes. So what types belong to which potential new species? Orgeon's clearly form a clade, as do Slate-colored. Cassiar are probably an inter-grade population between them. Red-backed and Gray-headed are another obvious clade, but where do the Pink-sided go? I do think 'Oregon' type Junco does need to still exist on iNaturalist, in the event a split ever does happen, those records won't be un-assigned to a species.

Posted by bryanto almost 6 years ago

@bryanto Not necessarily. A group of taxa must demonstrate more than just visual distinction to be warranted as its own species entirely. The specifics are debated, of course, as we see with yellow-rumped warbler and (in the past) with northern oriole.

No one has agreed with my suggestion yet but the only solution here, if we go forth with the full subspecies listing, is to use fields. Observations can be filtered based on fields, just as they can be filtered on taxa.

Posted by silversea_starsong almost 6 years ago

Thanks for this, and your work to clean up the database

Posted by gyrrlfalcon almost 6 years ago

Cool, Frank!
Some birds have had taxon swaps last year, like Downy and Hairy woodpeckers. Which were in the genus Picoides and now in the genus Dryobates. And the Wigeons and Allies are now in the genus Mareca. Pretty interesting! This obviously has to do with taxonomy, for people who don't know. Though I hope to be a taxonomist. Taxon swaps happen when; the science and process of naming living organisms, is a field that is constantly changing. When our scientific understanding of animal species and their relationships changes, it may mean that scientific names change as well. ... Some species have come to be known by multiple scientific names. Most books like the Sibley Field Guide to Birds are not updated with taxon changes. Subspecies are the different "Types" of the birds that occur in different ranges, though Red Fox Sparrow has many different subspecies I personally have seen P.i. iliaca which is the wintering subspecies of Fox Sparrow that occurs in Connecticut. Dark-eyed Juncos often intergrade with other Dark-eyed Junco subspecies. Connecticut has had the Oregon Junco, though not sure if it was J.h. montanus etc. Connecticut also had a Cassiar Junco J.h. cismontanus reported by my friend Nick.

Posted by bluejay2007 about 5 years ago

Seems to me that the inability to identify to "oreganus" group or similar means that a large number of observations are going to be identified as oreganus ssp. in error by default. Is the taxonomy for inaturalist unable to handle a taxon grouping between species and subspecies, or is there another reason why this isn't reflected?

Posted by markrobinson 6 months ago

Add a Comment

Sign In or Sign Up to add comments

Gracias al apoyo de:

¿Quiere apoyarnos? Pregúntenos cómo escribiendo a snib.guatemala@gmail.com