When I Report Subspecies

This is a topic that I've been wanting to share for awhile and I hope you can bare with my bickering and complaining. But here's the topic that keeps stabbing my mind, identification of my sightings to the subspecies level.

There are about 10,738 species of birds in the world and about 20,000 subspecies. Though a lot of these subspecies are different in terms of size, there's still a good portion of them being different morphically. I want to show a few examples.

Most people across the US and Canada knows what a Dark-eyed Junco is. However, there are many subspecies that look a good deal different from each other. Most people in eastern US know the Slate-colored Junco (Junco hymealis hymealis) which are gray overall. However, birdwatchers in western US see the Oregon Junco (J. h. oreganus Group) which have black heads and orange sides. Even people in southwestern US and Mexico see different Dark-eyed Juncos, namely the Red-backed Junco (J. h. dorsalis) have a pale gray body, a black mask and a rich brown-red back.

Point I'm trying to get across is, these subspecies are easy to identify and record in iNat (well, with the exception of the Oregon Junco Group) and I think we should. And that's what I always do when I submit an avian species to iNat, I identify down to the subspecies level. Every time I identify to subspecies if the species has one.

The problems I've been observing though is the fact that most birders or 'experts' on iNat, do not even bother to dive into the rich diverse world of birds. I've stopped counting the occasions when I've had naturalists correct my identification to the species level. When this first started happening to me on this site, I guess you can say I was holding grudges on those naturalists that did that. One of my thoughts were, "I identified to the subspecies level and you really took the time to click the 'Add Id' button and typed the species name I just suggested, just a higher rank." To me I still don't find the logic to that. Such as, if I say I saw an Interior Western Screech-Owl and you know for a fact its Western Screech-Owl, just hit 'agree' for the love of science. Is it really that hard?

And I know, that probably sounds rude and all but I am trying to get you to know how I feel when I try to get more valuable information. And here's the excuse, comeback or whatever you may call it you would say to me. How do you know you're right? And so you know, I do agree with that statement. I can be wrong at times. An example being the Red-tailed Hawk. The differences morphically between subspecies is a lot more subtle than the Dark-eyed Junco. For Red-tailed Hawks, namely Eastern (Buteo jamaicensis borealis), Western (B. j. calurus) and Northern (B. j. albeticola) have their unique "traits" between each other but they have overlaps in features, meaning that you have the observe all the traits of the individual to identify it to the subspecies level. And you can find any of these subspecies virtually anywhere in Canada/US, especially in the fall migration/winter.

Lucky for us however, the Red-tailed Hawk is just one of the few species where subspecies range overlap and a lot of them can be simply identified to the subspecies level through range, and for migratory birds, their summer range. Federally licensed birders have banded birds for years across the continent and their research has shown that subspecies that differ usually by can size can be identified to the subspecies level by range. It's actually what I did when I banded for a summer job.

But we also need to keep in mind of what's in the future of avian taxonomy. Nearly everyone who birds or have posted birds use the Biological Species Concept (BSC), which identifies subspecies by reproductive isolation. However, some believe the Phylogenetic Species Concept (PSC) to be more correct, as it identifies a species through range, characteristics and plumage differences. Let's show an example of the two concepts at work. A person using BSC will simply call a Song Sparrow and leave it at that. However, PSC believes that even subspecies are species and should be identifiable, so an observer using PSC would call it either Rusty Song (Melospiza melodia morphna) or Eastern Song (M. m. melodia) based on range, plumage and vocal differences.

And as we learn more and more about birds and their taxonomy, it appears that the Phylogenetic Species Concept is becoming more and more logical. If it would suddenly become our way of thinking of bird species, we'll suddenly see an increase of 8,000 species of birds. PSC also affects mammals too and I think if they applied it to them, I'm should the White-tailed and Mule Deer would be split.

Point I'm trying to get across is, if we suddenly see an increase of bird species, are we prepared to identify those birds to the new species level. I think it's better to be educated. So if you see an observation of mine with an id to the subspecies level, I want you to think about it before typing in 'Great Horned Owl' or whatever. If you feel that uncomfortable about agreeing with me, tell me. Send me a message and I will try my best to explain the differences in subspecies for that bird. I am willing to help educate. There's a lot of work to do on iNat, let's get going!

Posted on June 01, 2019 10:14 PM by birdwhisperer birdwhisperer


Such as, if I say I saw an Interior Western Screech-Owl and you know for a fact its Western Screech-Owl, just hit 'agree' for the love of science. Is it really that hard?

If the next IDer doesn't know how to identify to subspecies, they shouldn't be blindly agreeing with IDs they're not confident about. There's nothing wrong with adding a coarser species identification. If you want to offer up details for identifying to subspecies, you're welcome to of course.

Posted by bouteloua almost 4 years ago (Flag)

@bouteloua That's the whole point of this journal entry. If the avian taxa world goes to Phylogenetic Species Concept, we need to learn about subspecies. I also encourage those who do place the stricter species identification to state their reasoning. I've had one naturalist who will say every time he corrects my id he says 'I am unfamiliar with subspecies.' Point is, my purpose is to educate naturalist on subspecies and to show how I feel.

Posted by birdwhisperer almost 4 years ago (Flag)

It goes deeper than just users not bothering to deal with subspecies. The iNaturalist programme is also against subspecies.

The identotron/compare/suggestions tools on the ID bars and Identification panel, specifically do not allow one to select or see subspecies. Suggestions are only provided to species level. People using these tools to collaborate, or remind themselves, or to explore the current ID, are not exposed to the subspecies and dont even know that they exist. And when they do, these tools are useless.
It would be nice to be able to use these tools to subspecies/variety level.

When users add a courser ID to a species, it is an "agreement" and is not reported if one has switched "dont report on agreements" in one's preferences. But I get dozens of "disagreements" on my dashboard daily, where people have posted species in agreement with the subspecies (and vars - I do plants too) I have identified. This sometimes comes in batches: today it was over a dozen IDs to Lacyon pictus, which someone obviously "Identified en block", all in support of my IDs as Lacyon pictus pictus.
It would save me a lot of time if this worked properly.

One ID to species, puts the species name in the "headline ID" for the observation: it becomes needs ID species. But this does not work for subspecies. For subspecies, one needs two identifications to species level for this to apply. Even though it is a refined ID in exactly the same philosophy as any other taxon rank.

Posted by tonyrebelo almost 4 years ago (Flag)

@fogartyf This discussion is right up your alley. Dark-eyed Junco subspecies.

Posted by vermfly almost 4 years ago (Flag)

Eh, most avian subspecies are not field identifiable. Subspecies for North American birds, in particular, are typically holdovers from 50-100 years ago and haven't been properly vetted by modern taxonomic methods because the AOS moved away from them decades ago, so a good proportion of them probably aren't valid. In most cases, the best one can do is to make an educated guess based on location during breeding season. I think that's fine, but it's inherently tautological and doesn't add much of anything to the data. For example, anyone who is looking at the data is going to assume all individuals reported to species in the breeding range are the expected subspecies unless there is evidence to the contrary. For example, if I am looking at records of Nashville Warbler in the Sierra Nevada, I'm going to assume they are all subspecies ridgwayi even though most are not reported beyond species level.

There are some exceptions that are well-studied and visually distinct (e.g., Red-tailed Hawks mentioned above, although even some of its subspecies are in question, and Yellow-rumped Warbler).

Outside of breeding season or in areas where subspecies may overlap, identifying to subspecies level is just adding false precision to the data (which is something that some of us have worked hard to clean up for taxa like Dark-eyed Junco in iNat). There are a heap of other species plagued by false precision to the subspecies level in iNat but they are slowly being addressed. A lot of times, the problem is the common names used in iNat and confusion over subspecies versus subspecies groups.

A good place to start for understanding what subspecies can be visually identified in the field is the Clement's checklist that is used by both eBird and iNat. If the subspecies is not given its own "group (monotypic)" line, it likely is not field identifiable. Migratory subspecies in polytypic groups should probably never be identified to subspecies away from breeding territories. During the breeding season and for non-migratory taxa, I think it's fine to make a guess as outlined above. Great Horned Owl is a good example - Clement's lumps all but one (a distinctive population from Patagonia) into the same polytypic group. There are some average differences in plumage between some of the other subspecies, but likely none of them are distinctive enough that they can be identified from a photograph on any logic besides "It is subspecies X because it occurs in place Y".

@birdwhisperer you will be happy to know that the BSC hasn't ever been used in modern taxonomy :) It hangs on in biology textbooks as an easy and straightforward way to introduce students to the species concept and taxonomy, but it's an outdated oversimplification that isn't really used in science. Most modern taxonomic decisions typically are based on measures of gene flow between populations and supported by life history traits. Unfortunately, very few studies are being done examining the subspecies level in birds. It's expensive and isn't generally considered a research priority except in cases where there is conservation concern or there is evidence that a subspecies should be considered for elevation to species status (even the latter is often not enough reason to motivate a study).

Posted by fogartyf almost 4 years ago (Flag)

Greetings All: I recently stumbled upon this journal/blog post. I'm guessing I'm one of the offending subspecies deniers, so thought I'd contribute a comment....  First, I'll say that when I first started birding in the 70s, I too was enamored with the subspecies described--and given their own accounts--in Gabrielson and Jewett's (1940) Birds of Oregon. I loved the recognition of finer distinctions of geographic variation in species, especially when they coincided with different ecoregions. It all made sense and was really cool! As I explored around the state I looked to see the differences mentioned by G&J for some of these species/subspecies. Some were apparent, some were very clinal, some were not really discernible in the field.  When Dave Marshall, Alan Contreras and I were working on "Birds of Oregon: A General Reference" (2003), we recruited M. Ralph Browning to review and contribute subspecies information for that book. Most of Browning's information came from his notes over many years working at the Smithsonian Museum of Natural History--with birds laying on a table in front of him--as well as relevant publications. He revised a few things from AOU 1957 and agreed or disagreed with various other published accounts. Right or wrong, we were glad to record his viewpoints in the book.  Along comes eBird.... When I began using eBird and doing eBird review, at some point (I don't remember the year) we had to address people's use of the available subspecies (or subspecies groups) in their checklists. I personally never use subspecies in eBird except for well-marked taxa that can be ID'd based on visual appearance. Unfortunately, many people choosing subspecies in eBird simply choose the subspecies that ought to be where they are without really being able to ID the subspecies in the field or photo (there are various reasons for this). This approach seems less common in iNat, at least with birds, but it still occurs. Which brings us to the present topic.
  In his comment, Frank did a very nice job outlining various points of view that I also share, but I'll just add a few of my own words/observations.
 @birdwhisperer, it appears to me that the majority of your submissions are not "identified" (or identifiable) to subspecies based on the appearance of the birds in the photos, but rather on what is expected at the location. I can say without question that some photos--e.g. distance silhouetted photos of Red-tailed Hawk--are not identified to subspecies based on evidence in the photo, but rather only on location.
  In my opinion, labeling to subspecies based solely or mostly on location is not science. It is not a data point documenting a verified location for that subspecies. It is an assumption. So "for the love of science" I do not just hit "agree." There are a few species that have readily photo-identifiable subspecies and intergrades, and other times I just get lazy (it is much easier to just hit the agree button for a subspecies ID), but for most species, photos are inadequate to visually confirm subspecies, so I don't. Lastly, I have yet to see you explain how you identified a subspecies based on information available in the photo. For those instances where that is possible, you might get more buy-in for the subspecies. In any case...
  I think it would be worth your while to read through Frank's words again. He offers a lot of good information, perspective, and advice. I for one want to contribute scientifically valid data to iNat (and repositories further down the data stream), not assumptions or guesses. To say it another way, I do want to contribute subspecies observations when they can be verified by the photos, but I don't want to add spurious or tautological "data" into the data set.
  @birwhisperer, from what I can see from your submissions and occasional comments, you are deeply interested in and are enjoying the critters you are observing and studying, and are interested in promoting knowledge about them. I share those same desires and am very happy to support you in these endeavors, but I think we need to limit what we say about observations, and how we ID them, to the evidence available to us in the photos (or recordings). Not all will agree, and there are always exceptions and different issues with different taxa that I don't intend to address in this "short" comment, but in general that's how I look at it.

Best Regards, Matt

Posted by umpquamatt over 3 years ago (Flag)

@umpquamatt I've actually mellowed out from my rant over the summer but it has not stopped me from identifying subspecies. However, the insight from the conservation above led me to create this list for my local area for birding. Pretty much it's every single subspecies of avian species that can be found in... well northeastern Oregon and I spent a lot of time making sure I got every possible one, including breeding, wintering, migrators from western Canada and vagrants depending on how nomadic the species is.


And probably the biggest thing I learned is, is that I'm more sketchy on making a call outside of the breeding example. Example being today, I added all my American Pipits photos to iNat and I know that the first link below is subspecies alticola because it's the expected breeder in Montana, it's May and it exhibits the mostly unstreaked breast known in the subspecies. However in second link, I'm scared to subspecies id the pipit at Cold Springs NWR, OR because of lack of discerning features between subspecies in winter plumage. So it could be pacificus (which is what I suspect) but I can't say it's not rubescens or alticola due to time and location.


And the same thing occurred when you corrected my Savannah Sparrow from a couple days ago. I would've identified it to subspecies level but to me it's too hard to discern if it's the breeding nevadensis subspecies or the migrating kodiak subspecies and maybe even to a lesser exact, a possible coastal vagrant crassus ssp.

As for the Red-tailed Hawks, I initially identify by location and then follow it up with binoculars looking for features since they bring it up closer than my camera and I don't have to deal with shutter. And also the fact I know that after 5 years of watching Red-tails in the area, everything is going to Western except for a dozen or so Harlan's a year in the winter and one or two questionably Eastern/Northern annually.

Posted by birdwhisperer over 3 years ago (Flag)

  You may have mellowed emotionally from your rant, but it appears that you continue to naively and presumptuously label individuals with subspecies IDs, many with no justification other than location, and some with no visual features available for assessment (e.g. blurry, or silhouettes).  Thus, the "information" that you contribute to iNat--at the subspecies level--is no information at all, and in my opinion often either wrong or unjustified except by probability of location, which again, is not empirical data. It serves only to add to a list of taxa for yourself, at the expense of the empirical reliability of iNat data. I hope you will consider the implications of this.

Even the ones you claim to be IDing based on visible features, it appears to me you are unaware of (as most of us would be) the variation within and among subspecies, and you (like many of us) have no reliable way to distinguish them. I'm not talking Audubon's and Myrtle Warblers, or Yellow and Red-shafted Flickers (though even some of them can be difficult), I'm particularly talking about things like Savannah Sparrows, Horned Larks, but many others as well. What resources are you using to ID subspecies. I don't think there is anything generally available that would allow great confidence in IDing most individual subspecies, without great study and confirmation from others who have the field and/or museum experience. 

  Hoping perhaps you might listen to input from someone else, I asked Randy Moore (Randall S. Moore, Oregon State University Streaked Horned Lark researcher) to take a look at all Streaked Horned Larks in iNat. As you see, he has commented on your reports. I intend to ask M. Ralph Browning for his opinion on other species where I think you unjustifiably ID subspecies. I don't know if he will comment directly on iNat, but I will pass on his comments.

Most descriptions of subspecies that I am familiar with are averages or trends. Variation is rarely described for subspecies, or even for most species, but the differences at the subspecies level are of course relatively even more subtle. This should engender caution and tremendous study at museums and scientific literature if you want to be fully versed on the appearance of subspecies of different ages, sexes, and seasons. I am totally with you on being interested in subspecies, but the goal of labeling most photos to subspecies is completely unrealistic given the resources available and scant experience we have. Awesome when we can do it, but the expectation should be that it is few and far between.

I have not yet seen a reasoned and thoughtful response from you to points made by Frank or myself. I think it would be worth your while to digest and process our comments and in time (do not rush) articulate your thoughts on these matters. Some specific vocabulary for study: tautological, empirical.

Best Regards,Matt

Posted by umpquamatt over 3 years ago (Flag)

Well, I guess if you're so passionate about not identifying by range, I'd recommend considering these options next time you submit:

Black-billed Magpie -> Black-billed/Eurasian Magpie
Northern Shrike -> Great Gray/Northern Shrike
Sooty Grouse -> Blue Grouse (Dusky/Sooty Grouse)
Pacific-slope Flycatcher -> Western Flycatcher (Pacific-slope/Cordilleran Flycatcher)
Willow Flycatcher -> Traill's Flycatcher (Willow/Alder Flycatcher)
California Scrub-Jay -> Western Scrub Jay (California/Woodhouse's Scrub-Jay)
Western Sandpiper -> Western/Semipalmated Sandpiper
Pectoral Sandpiper -> Pectoral/Sharp-tailed Sandpiper
Ruddy Duck -> Ruddy/Andean Duck
Western Meadowlark -> Western/Eastern Meadowlark
Eurasian Collared-Dove -> Eurasian/African Collared-Dove

Posted by birdwhisperer over 3 years ago (Flag)

Sean,  Really. Take a few minutes, days, weeks, to breath a little. You clearly have taken no time to consider and try to understand the advice given to you, as you respond with silly attacks/challenges rather than address the points made.

The species pairs you have listed are a mixed bag: some would be nearly impossible to discern in the field while others are very easily distinguished. How we should treat the indistinguishable ones is a good topic. I'll be happy to discuss that and other topics with you further when you make a sincere effort to consider the points already made by Frank and myself.

Best Regards,Matt

Posted by umpquamatt over 3 years ago (Flag)

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