Subspecies: Views are changing, what about you?

[tristanmcknight]

A fun example to illustrate this: one of my labmates from grad school did a study on Hercules beetles where he basically showed that if you apply a consistent objective metric (both DNA or morphology worked) then the variation across Neotropical Dynastes should either all be called species or all be called subspecies, but instead we’ve inherited a patchwork system where the northern variants are called species (because US-based scientists were ok with splitting) but the southern variants are all called subspecies (because the South American taxonomists didn’t want to break up the parent name that dated to Linnaeus). And this bias has affected the narrative, because biogeographers therefore assumed that there was a radiation (because more “species”) when the group expanded north, instead of realizing the south actually had more diversity!

https://academic.oup.com/sysbio/article/65/4/685/1753141

I would push back that “per standard definition species should have a genetic difference of more than 2%”. The standard barcode distance just helps us recognize there is a gap between groups, but the life history of the species can affect whether that gap signifies populations or species (plus the culture of the taxonomists, of course). I don’t know any speciation scientists who actually use 2% or 5% as a hard and fast rule. I guess maybe the bacteria and virus people (and some fungi people, now that I think of it), but species are pretty much meaningless for them anyway…

[AdamWargon]

@AdamWargon

Like (probably) many iNatters, I can clean up any Jeopardy! category about nature or science

@tristanmcknight

US-based scientists were ok with splitting) but the southern variants are all called subspecies (because the South American taxonomists didn’t want to break up the parent name that dated to Linnaeus). And this bias has affected the narrative, because biogeographers therefore assumed that there was a radiation (because more “species”) when the group expanded north, instead of realizing the south actually had more diversity!

This could be a trick question on one of those Jeopardy-style gameshows for nature lovers. The question seems to be asking about biodiversity (subspecies radiation), but the correct answer is actually about the culture of taxonomists in that country!

[alternative]

Whew it certainly is a can of worms.

Coming from Plantae:

I personally like ssp. for plants, but not all ssp. have the same naming patterns, and that is what truly irks me. Also, when considering or classifying a ssp. is it really a subspecies, or a new hybrid or a mutation of an existing species? are we classifying ssp. by morphology or genetic lineage?

For example: I have been observing and studying a subspecies of Quercus sinuata (Common name: Durrand’s or bastard oak). This ssp. is called “Quercus sinuata var. brevibiloba.” However, its common name is “White Shin Oak,” completely different from the other’s common name(s).

It seems to be a subspecies based upon leaf morphological traits and a different geographic distribution, different soil types, and slopes than those of the true species.

__________________________________________________________________

Another plant I can think of that has ssp. based on morphological traits is the highly morphing Toxicodendron genus.

There is Toxicodendron radicans (Eastern), Toxicodendron rydbergii/Toxicodendron radicans ssp. negundo (Midwestern - looks like Acer negundo saplings), and Toxicodendron radicans var./ssp.verrucosum (Texas poison ivy - looks like acer granditatum - big tooth maple)

These are morphological variations across geographical areas; these species all converge in central Texas, but they are all separated by habitats. Eastern will be found more heavily in lowlands and wetlands; Midwestern can be found in facultative wetland adjacent upland/sloped spaces, and Texas will often be found in facultative mostly upland plateaus/ridges and slopes.

It just gets really complex, and if it’s just the same Species with different morphologies in different habitats and not genetically distinct, then maybe we should rethink how they are classified.

_________________________________________________________________

A third naming issue i’ve come across is ssp. with _sp__ “X” ____________

for example I recently observed “Canyon Muhly” which seems like a distinct but similar species of “Muhlenbergia lindheimeri” - in my area. However, it’s ssp. name is “Muhlenbergia X involuta” which does not allude to a ssp, but a hybrid. so that is what is confusing to me mostly.

___________________________________________________

The last one!

Eastern redbud: Cercis canadensis

Texas redbud: Cercis canadensis var. Texensis

I think the differences for this one is length of flowering time and geographic distribution, different soil types. I don’t even know if it’s genetically distinct, just that we ID it under this ssp. in my region, it gets classified as “Texas redbud” west of the balcones escarpment fault lines, and it gets classified as regular Eastern redbud when an ID is east of the fault lines .

Does anyone else find similar issues with plants in their regions?

[upupa-epops]

I was looking into asters and goldenrods recently and there were a couple species with descriptions like “the nominate variety is fuzzy but there’s a less common variety without hairs”. I like that both varieties have formal names, and assuming it’s actually a binary trait it’s useful for learning them and having a label to apply to the different traits. I like that way of doing it for major traits that are variable but not very correlated with range or reproductive isolation (e.g. light/dark morphs in birds, which don’t get any formal taxonomic recognition).

[alternative]

I also like to classify the variation, but having to type out the full name is maddening XD

Maybe variations should be more like check boxes rather than through naming,

because what if someone comes upon something that sits between traits of two distinct ssp.? does that mean we create a new ssp, or does it get classified with an existing one based on % of traits?

[upupa-epops]

On eBird if you see a Willow or Alder Flycatcher, or an Allen’s or Rufous Hummingbird, but aren’t able to identify them, you can choose the “slash” options “Willow/Alder Flycatcher” and “Allen’s/Rufous Hummingbird”. For subspecies they automatically lump them into similar groups and you can’t even identify down to subspecies level if it’s part of a subspecies group with extremely similar-looking subspecies.

In both eBird and iNaturalist someone has to manually set up a “taxon” for each slash or subspecies or hybrid. I like the idea of having the option to make an identification where you indicate that it’s a slash/hybrid and then can type in any 2+ species that you think are involved. Then if I see a Slate-coloured Junco I could identify it as Junco hyemalis hyemalis/carolinensis without a curator having to create a subspecies pairing for every conceivable pair of subspecies that could look similar to each other.

But that would introduce a lot of extra complexity in other ways, so I don’t foresee a system like that actually being introduced. As I mentioned earlier the best way to do that currently is just with observation fields (for example this field has options for the major junco groups).

[cs16-levi]

Good points and thanks for some more threads to look through.

I think an important note for people to understand about making comments here is that I am only referring to the field-identifiable subspecies, which in most cases have less taxonomic confusion than the others. Not none, but less, far less, and I intend to exclude scientific names for the controversial cases and mention the chaos. I will also include a recommendation to leave them for experts, but note them as the populations or groups in the observation details.

@rcavasin

iNaturally, disagreements arose over precisely what constitutes a valid intergrade. Once you start looking carefully at members of this species, you notice a great deal of subtle variation, and no two people seem to agree on what constitutes a “pure” example of either subspecies. Everyone seemed to have different, and often conflicting ideas about which subtle differences indicated that one was dealing with an intergrade. Taken all together, these various criteria would suggest that most individuals in Ontario (which is much larger than many countries) could be interpreted as an intergrade. This may actually be reflection of biological reality, but once you go down that road, the subpecies taxonomy becomes largely pointless.

Again, I can only speak for birds, but in all the resources I’ve read and the situations I’ve encountered, intergrades should almost never be identified from photographs; it serves little purpose and does more harm than good. I see you have learned this the hard way (P.S. I liked the pun)

@upupa-epops

On eBird if you see a Willow or Alder Flycatcher, or an Allen’s or Rufous Hummingbird, but aren’t able to identify them, you can choose the “slash” options “Willow/Alder Flycatcher” and “Allen’s/Rufous Hummingbird”. For subspecies they automatically lump them into similar groups and you can’t even identify down to subspecies level if it’s part of a subspecies group with extremely similar-looking subspecies.

Thanks for mentioning this; I intended to, but ended up forgetting in the midst of other things. In a perfect world I like to only include the field identifiable subspecies that match with eBird’s acceptance, but that could get tricky…

@elsemikkelsen

Making a guide focused on within-species variation is a cool idea, but be careful not to overemphasize them as discrete categories.

I appreciate the response, you make a lot of great points. I do intend to frame the book around variation, including only the expected subspecies that are identifiable in the field. My approach is to take the pressure off the birder to make a definitive diagnosis in every single observation, but instead teach them to appreciate the variation visiting their feeders or in the fields they’ve been birding in their whole lives and the amount of information they don’t know they could be adding to their eBird checklists, etc.

[rcavasin]

@cs16-levi

Again, I can only speak for birds, but in all the resources I’ve read and the situations I’ve encountered, intergrades should almost never be identified from photographs; it serves little purpose and does more harm than good. I see you have learned this the hard way (P.S. I liked the pun)

Just to illustrate how we ended up there, the species in question is Limenitis arthemis. If you look at the two subspecies arthemis (nominate) and astyanax, you’ll note that they are radically different in appearance. Given the (apparent) ease with which these two subspecies could be discriminated, it seemed “safe” to identify them on iNat. It had been thought that intergrades were relatively rare, but we quickly learned otherwise. Most intergrades lean heavily towards astyanax, and historically, these were identified as astyanax. Here on iNat, thanks to the discussions around community ID, questions arose as to what constitutes an intergrade between these two subspecies. Depending on what one took to be a “pure” astyanax (and opinions varied), many minor variations were claimed to indicate an intergrade. Because of the history behind this group (ie. that the two subspecies were once considered separate species), there was a mindset that any ID left at the species level was “incomplete” because it wasn’t in one of the 2 historical pigeon holes (either ssp arthemis or ssp astyanax). We felt that using the 3rd pigeonhole of “intergrade” might be useful, as it would allow us to map out the region of overlap between the two subspecies. At some point, someone pointed out that an alternative taxonomy would drop the nominate subspecies “arthemis”, and instead treat individuals that fit that phenotype as intergrades between the Northern/Western subspecies rubrofasciata and the Southern subspecies astyanax. This alternative taxonomy would create a much larger intergrade zone, and relegate most L. arthemis in Ontario to the intergrade category. Nevertheless, I personally found this alternative taxonomy more intuitively satisfying than the traditional one. Even with a case like this, where there are two obviously different subspecies, identifying them as such STILL ended up descending into unproductive hairsplitting.

In summary, I think this constitutes a classic example of the proverbial “road to hell” being paved with good intentions.

[cs16-levi]

Yeah, I’ve heard butterfly spp. identification frequently ends in pointless pigeonholes. With birds, I avoid intergrade identifications and usually avoid areas where intergrades are abundant.

[antsbyleo]

I’d use them more if it weren’t for the fact that getting ants to species is an accomplishment in itself. I don’t really know how you would get it to subspecies without a home DNA sequencer.

[cs16-levi]

Sheesh, yeah doing a quick look I cannot see how any subspecies would be relevant for the average identifier. Seems like subspecies from a photo is almost always impossible. The subspecies differences usually rely on the exact number of hairs on a specific segment, the texture of their shell, or minor jaw structures, etc.

Seemingly a few decades ago, entomologists named a new “subspecies” every time they found an ant that was slightly darker or a litlle larger. Of course, this is getting cleaned up but thousands of irrelevant taxa were created. Wow, what a headache.

[petezani]

I found this review an interesting take on subspecies.

https://onlinelibrary.wiley.com/doi/full/10.1002/ece3.9069

[rcavasin]

@cs16-levi

Seemingly a few decades ago, entomologists named a new “subspecies” every time they found an ant that was slightly darker or a litlle larger.

I think this kind of thing is still happening. Just because sophisticated DNA analysis is possible, it doesn’t mean it is being applied every time a new subspecies is proposed. It only becomes a problem when identifiers decide to take them seriously and use them in their identifications.

As with many things (but not everything), the solution is to ignore them.

[birdwhisperer]

@cs16-levi I talk about this a lot to users, and I figure my perspective as a “subspecies” identifier may help.

@cs16-levi

(1) What is a subspecies? (for the purposes of a user-friendly guide)

The definition of a subspecies has definitely changed over the years. From an archaic point of view, I have perceived this as any population that occupies a geographically distinct area and exhibits traits that could make it diagnosable, but not quite what we might define as a “species”. You use Yellow-rumped Warblers in your second question, and I think that’s a good place to start. Audubon’s Warbler was described as Sylvia auduboni in 1837, and I’m not sure when it was subsumed into Myrtle other than the fact it was lumped by the AOU in the 1980’s. But rather striking birds like that were identified, because a yellow throat and plain head is rather distinct from white throat and dark auricular patch.

In the late 1800’s and early 1900’s, many began describing subspecies based on measurements or subtle color tone differences. Hermit Thrush is a good example. I will provide the Birds of the World descriptions for id of the “guttatus” or Western Lowland group.

guttatus: Medium-small (Aldrich 1968, wing chord < 91 mm), with base of bill somewhat constricted; mantle moderately dark gray-brown to brownish-olive; uppertail coverts dull rufous; flanks washed brownish gray; breast whitish; and breast spotting moderate.

nanus: Like C. g. guttatus, but mantle both darker brown and more rufescent, flanks darker gray, rump washed dull rufous, and breast buffier and more heavily spotted.

verecundus: Like C. g. nanus, but base of bill slightly swollen, mantle darker brown, and flanks darker brownish gray.

vaccinius: Like C. g. verecundus, but mantle dusky chocolate brown and breast spotting heavier and blacker.

jewetti: Small (wing chord generally <86 mm), with base of bill constricted; mantle pale, warm grayish brown; crown darker brown; uppertail coverts cinnamon; flanks pale grayish; and breast spotting moderate.

slevini: Like C. g. jewetti, but mantle duller grayish brown and breast spotting sparser.

oromelus: Similar to C. g. verecundus, but mantle moderately pale brownish gray (lacking a rufous wash) and flanks pale grayish.

That’s seven taxa representing individuals from the Aleutian Islands to northeastern California. Here is a side-by-side comparison of Hermit Thrushes in July, the left one from Kodiak Island, and the right in Jackson County, Oregon. They should be representing guttatus and oromelus, the endpoints of a cline from pale to dark thrushes.

That is it. That is the difference. This is why I find a lot of merit in the eBird ISSF groups, because it groups together subspecies that are likely not valid or simply diagnosable in the field. Here’s a comparison of the guttatus (lowland) group from Oregon and an auduboni (montane) group in Colorado.

It’s not a strong difference, but it’s a difference nevertheless, and knowing how to identify it leads to my discovery of finding guttatus in Idaho. eBird Checklist - 6 Oct 2023 - Camas NWR–Headquarters - 30 species

@cs16-levi

(4) Should you identify subspecies based on range only?

There is a Birding magazine article that I published with Doug Robinson that discusses this very question. It was our statement that “Probability suggests options, but field marks establish criteria for correct identifications. Do not assume an identification based on location unless you know definitively that there are no other options.”

I still stand by this statement for the most part, but I think it’s heavily contextual based. If it is June/July in North America, I think you can get away with just about any ssp id for birds. I just returned from a 4-week trip to Ecuador, and I am using distribution ids for most of my bird ssp because the tropics are a “stable” ecosystem, so there’s not really a reason for birds to move around when resources are constantly available. And any movements we do see are altitudinal movements, not movements between the west and east Andes ridges.

I’m getting a little bit winded but if there are any other questions, I can try and answer them.