The Pine Squirrel Conundrum
In North America, if you have ever hiked in the mountains or boreal forests, you've probably encountered one of several species of small squirrels. Often coined as "pine squirrels", the taxonomy of the genus Tamiasciurus is complicated to say the least. Then again, what does have a straightforward taxonomy. Prior to 2016, three species were recognized; the American Red Squirrel (T. hudsonicus) in most of the US and Canada, Douglas's Squirrel (T. douglasii) in the Pacific Northwest, and Mearn's Squirrel (T. mearnsii) in Baja California. However, a genetics paper has shaken the grounds on where we define species in this genus, and now it's caused a roadblock in submitting reports into iNaturalist with the correct species id. I will try to give those interested a full understanding on what's going on, and just my thoughts on these changes.
Every Squirrel Population
Before we start delving into the current issue at hand, we need a clear understanding on what the taxonomy was like in the past. Depending on whose authority you follow, there are 29 different populations of squirrels that have binomial or trinomial name. Twenty-five are included within the American Red Squirrel (sensu lato), three in Douglas's Squirrel, and the Mearn's Squirrel. Here's the list and their approximate distribution as cited in Steele (1998, 1999).
American Red Squirrel
- preblei -- Most of Alaska, Yukon, NW Territories, south into the Canadian boreal forests to approximately Manitoba
- kenaiensis -- South-central Alaska; primarily Anchorage and Denali region
- petulans -- Southeastern Alaska; primarily Chugach and St. Elias Mts
- pictans -- Extreme southeastern Alaska, and Haida Gwaii region
- lanuginosus -- Vancouver Island
- columbiensis -- Interior British Columbia and southern Yukon
- hudsonicus -- Manitoba, Ontario, north peninsula Michigan, and northern Minnesota
- regalis -- Isle Royale Island in Lake Superior
- ungavensis -- Northern Quebec and Newfoundland
- laurentianus -- Southern Quebec from approximately Montreal and east
- gymnicus -- New England, New Brunswick, Nova Scotia, and Quebec on the south side of the Lawrence River
- pallescens -- North Dakota and southern Manitoba
- minnesota -- Minnesota, Iowa, Wisconsin, and northern Illinois
- loquax -- Indiana and lower peninsula Michigan east to New York and New Jersey, including southeastern Ontario and southwestern Quebec
- abieticola -- Applachian Mts from Virginia to Alabama.
- streatori -- Southeastern British Columbia Rockies into eastern Washington and Idaho
- richardsoni -- Alberta Rockies, western Montana, central Idaho, and northeastern Oregon
- baileyi -- Central Montana and Wyoming
- ventorum -- Yellowstone ecosystem of southern Montana south to Watasch Mountains, Utah
- dakotensis -- Black Hills of South Dakota and Wyoming
- fremonti -- Colorado Rockies, extreme southern Wyoming, eastern Utah, and northern New Mexico
- dixiensis -- Mountains of central and southwestern Utah
- mogollensis -- Central Arizona west into New Mexico to extreme south-central Colorado in the Sangre de Cristo
- grahamensis -- Mountains around Tucson, Arizona
- lychnuchus -- Sacramento Mountains, New Mexico
Douglas's Squirrel
- mollipilosus -- Cascade Mtns from southwestern British Columbia south to the California Coast Range to San Francisco; absent on east slope of the Oregon Cascades
- douglasii -- Olympic Mtns, Puget Sound, and Coast Range to southwestern Oregon
- albolimatus -- East slope of Oregon Cascades, plus western Blue Mtns, south through the Sierra Nevadas
Mearn's Squirrel
- mearnsii -- Sierra de Pedro Martir, Baja California
The Genetics and Split
A genetics paper was published (Hope et. al 2016) in which 109 localities were sampled. In was in the author's conclusion that the genus actually represented three species, but not the same three species we recognized prior to the split. Their suggestion was to lump Mearn's Squirrel into Douglas's Squirrel, and treat all the "American Red Squirrels" in the southwest (specifically fremonti, dixiensis, mogollensis, grahamensis, and lychnuchus) as a new species. This analysis was accepted by ASM, the mammalian taxonomic source for iNaturalist, and as such, was installed into the site.
The Issue
You might be wondering, why are there issues? The split should be rather straightforward, all the squirrels from xeric mountain chains should be Fremont's Squirrel, and all mesic mountains should be American Red Squirrel. Or at least, that's what the literature above regarding subspecies distributions claim. Except in Hope (2016), there are 3 localities in which American Red Squirrels are apparently present in the Colorado Rockies, with one location in which they are sympatric with Fremont's Squirrel.
This conflicts with the idea that maybe these squirrels don't interact, a potential reason why they speciated. The Colorado Rockies are only connected the northern chains by the Book Cliffs that in eastern Utah, and if that had been the point of contact, we would be dealing with a small and narrow hybrid zone. But now that we know that American Red Squirrels occur in places like Rocky Mountain National Park and the headwaters of the Arkansas River, there is essentially nothing stopping them from traversing throughout mass connections of mountains.
This is a prime example of the pitfalls in the phylogenetic species concept, it is not a field-friendly species concept. Because we split the Fremont's Squirrel on nothing more than phylogenetics, and now we're in a situation in which we don't know what we're looking at. And this is a common occurrence on the iNaturalist platform. When the species was initially split, all observations in accord to distributions described in older literature were transferred to Fremont's Squirrel. But some observations were reversed due to the above research. Here comes the next question, how do you identify the squirrels, so you can skip past all this confusion?
Simple answer, is you can't under our current knowledge. That is because the Fremont's Squirrel is nearly identical in pelage to the American Red Squirrel. I tried doing some digging myself, and I can't find the diagnoses for the subspecific variation within the pine squirrels. I can only assume that it was based on pelage, but without the literature actually stating so, I can't see there being a realistic way to distinguish Fremont's from American Red Squirrel.
Future Chances Necessary?
I do believe future changes are necessary to completely understand the pine squirrel complex. I do see some glaring issues as a biologist when I see phylogenetic tree that Hope (2016) provided. I truly believe that they only split the Fremont's Squirrel because it created a paraphyletic relationship with the Douglas's Squirrel. What does that mean? It means under the taxonomic at the time, it would be like someone saying you're cousin is the same species as yourself, but your sibling is not. This is simply not true, so you resolve the issue in one of two ones: one, you treat your cousin as a different species, or you treat your sibling as the same species as yourself and your cousin. So rather than two species, you need to make a decision on whether to recognize one or three species. Hope (2016) chose the 3-species route.
Was the three species route the best route to go? Hard to say, since everyone has a different concept on what defines a species. One thing I will mention is that genetic divergence is not very significant with any of the taxa. Hope (2016) reported that the greatest sequence divergence was 3.64% between the Northwestern and Southern Rockies haplogroups. This is in the upper edge of what is generally acceptable for subspecies, but of course, this is all arbitrary and up to interpretation. Though, I will make the note that there's a number of North American birds with a higher divergence rate, but are still maintained as one species, most notable of which being the Marsh Wren (7.9%). But the divergence rate provided should've been grounds for further out factors are keeping or not keeping these two taxa apart.
If we accept the taxonomy as is, we are still faced a rather concerning issue with the taxonomy. As previously mentioned, I believe that Hope (2016) split the Fremont's Squirrel purely because it was paraphyletic, and that they were more closely related to the Douglas's Squirrel than American Red Squirrels. That's fine and appropriate, but lets take a look at the phylogenetic tree, which I shared above. In the first split of the tree, we see the upper branch comprising the Northwestern and Continental mtDNA haplotypes for American Red Squirrels. In the lower branch, we have the Douglas's and Fremont's Squirrels, but there's other taxa mixed in. You guessed it, haplotypes of squirrels still listed in the current taxonomy as American Red Squirrels. Meaning, even with species split, the American Red Squirrel is still paraphyletic!
This is precisely the reason why I don't like phylogenetic delineate species. You fix the problem without fixing the problem. And the problem is, the Red Squirrel is paraphyletic... twice. If the phylogenetic species concept is used properly, the following changes have to be made in order to maintain the species status of the Fremont's Squirrel. One, split the North Pacific Coast and Central Rockies haplotype as their own species. So all Red Squirrels from the Haida Gwaii to the Blue Mountains of Oregon, would be their own species. Two, you need to do something with the Vancouver haplotype. You can treat that as a full species as well, because the genetic difference is not that great. But given that this taxon is sister to Douglas's Squirrels, I see no reason why they wouldn't be considered a subspecies... except they look phenotypically like Red Squirrels, not Douglas's.
I think you can see what I'm getting at now. We were so super-fixated on recognizing Fremont's Squirrel, that we didn't consider other ramifications to justify our claims. In the end, if we are recognize Fremont's, we would have to recognize four species instead of three, plus a subspecies transfer. Or lump all the squirrels into one species due to the inconsistency of morphological delineation. Sure the Douglas's Squirrel is distinct, but if we include the Vancouver population, then we have squirrels that are phenotypically intermediate or identical to the sister species. Therefore, they aren't as distinct as we thought.
Possible Future Lineup
Remember when I listed all the taxa for pine squirrels? Well, I figured I would post a hypothetical taxonomic list based on the phylogenetics presented in Hope (2016), if we were support a four species concept, and synonymizing subspecies that do not have any genetic structure.
American Red Squirrel
- richardsoni "Northwestern" haplotype -- Synonyms: preblei, kenaiensis, petulans, ventorum
- hudsonicus "Continental" haplotype -- Synonyms: baileyi, dakotensis, pallescens, minnesota, regalis, ungavensis, laurentianus, loquax, gymnicus, abieticola
Columbian Red Squirrel
- picatus "North Pacific Coast" haplotype -- Synonyms: columbianus, streatori
- undescribed ssp "Central Rockies" haplotype
Douglas's Squirrel
- douglasii "Western" haplotype -- Synonyms: mollipilosus, albolimatus, mearnsii
- lanuginosus "Vancouver" haplotype
Fremont's Squirrel
- fremontii "Southern Rockies" haplotype
- lychnuchus "Sacramento Mtns" haplotype
- mogollensis "Southwest Sky-Islands" haplotype -- Synonyms: grahamensis
Now you know the pine squirrel dilemma. If you have any additional thoughts or opinions, please feel free to comment below. I do not perpetuate to be an expert in any way, but this is just my take on the issue as a biologist.
