Subspecies as a "correction"

I’ve been posting for a year or so - primarily birds, and butterflies but also mammals. Often my ID will be “corrected” by someone listing the subspecies. I understand the principle and I’m quite familiar with the subspecies of North American birds. But with posts of African mammals, it seems there is conciderable geographic overlap of subspecies - Cape Buffalo vs African Buffalo, for instance.

Southern Africa is known for its high endemicity. As far as I know this due to epochs of low rainfall make the dense forests of central Africa receded allowing movement to the south. Then when next wet epoch occurs it isolates species long enough for some degree of speciation


Just curious, is there a reason that it comes across as a “correction” and not a “refinement” – or would that be saying the same thing?


Any given ID is not really a “correction”, but rather the expression of what that identifier believes the observation is of. It may or may not shift the Community ID, depending on what other IDs have been made.

1 Like

@kiwifergus, I’m not sure whether the question is more about African subspecies in particular, or about the way subspecies IDs are treated. Also I’m not sure who you are responding to, since I had deleted my post due to this uncertainty.

my apologies! I used the reply at the bottom of the topic, which to me is a reply to the general discussion. I do find it frustrating that discourse drops the “reply to” indicator when you reply directly to the last post, which in this case makes it unclear if it is a reply to the last post, or the topic in general…


I learned something new. I never noticed the reply button at the bottom of the topic. Thanks.


I certainly appreciate others “refining” an identification to subspecies, and I will almost always “agree” with the refinment. I just find it amazing that there are those who are so knowledgable o be able to identify to subspecies especially when there is overlap in range and very little if any clearly identifiable differences in morphology.

1 Like

As I stated, your identification represents your understanding of what it is. If you realised afterward that it was that subspecies, possibly because their identification brought your attention to the characters that differentiate it, then yes, “Agree” to the ID. It’s not encouraged by the site, but if you recognise the identifier and have a high level of confidence in their ID skills, then you can “Agree” to add weight to their ID. It’s not a good idea though to just agree with the ID because it is better than yours, especially for range based IDs that might in turn feedback into supporting range based IDs!

1 Like

The addition of a subspecies to the species ID is somewhat controversial for certain taxa. Some users like providing them, when the geographic race can be clearly identified, whereas others go out of their way to ID only at the species level due to uncertainty or opposition to the subspecies concept. I tend to use subspecies when that taxon is clear and widely recognized and especially in cases when the subspecies is being considered for possible elevation to species level.

1 Like

Also, as noted in other forum topics, subspecies IDs can affect the conservation status, and possible automatic obscuring, applied to an observation (for better or worse!).

Yes, I agree. In cases where there was overlap in the ranges of subspecies I generally stayed with just species. But where there was a clear distinction in ranges, I’d rend to go with the refined subspecies ID - An example was a refinement of my observation of a Redwinged Starling seen at Blyde River Canyon, SA, it was suggested that it was of the subspecies morio - the Southern Redwinged Starling. Looking up the ranges of the 2 ssp - rueppellii, is restricted to south Sudan to central Ethiopia and north Kenya, morio, Uganda and Kenya to Botswana and south South Africa - it seemed clear that morio was the correct subspecies & I made the refinement.


There is some real concern that using range as the basis for a subspecies assignment, especially in species as mobile as birds leads to a feedback loop.

This is discussed here as well as several other threads on the forum.

Range can be used properly to assume a subspecies identity, I’d even go so far as to accept a statement that it can be used to assume with a high degree of probability. But the only way to ‘know’ that it is the particular subspecies is to be able to visualize and validate the morphological or other separator keys that define it.


This is a problem with some species as well which have been defined based on genetics but which cannot be reliably separated morphologically, especially from photos. Three species in the Sceloporus undulatus (fence lizard) complex in western US are now recognized, including on iNat, but species-level IDs are being made based on distribution which is not well-defined. The resulting dot maps on iNat are basically a reflection of where we think these species occur based on a limited number of confirmed locations derived from genetic analyses. So the IDing based on photos becomes somewhat circular: we think Species A is in this area so that’s the ID we assign a record, which gets reinforced over time as records from that area accumulate.

1 Like

This is a touchy issue within the herp community. There are those that are big proponents of identifying taxa to subspecies whenever they can and others who never do. My personal philosophy is to agree with whatever level of ID the original record owner chose (unless they have chosen clearly the wrong subspecies).

The problem we get into is when you try to label each and every animal to a subspecies when by definition a good number of them should be expected to be intergrades and not assignable to a single subspecies identification. Some choose to ID it to the subspecies it “looks more like”, but that is an obfuscation IMHO.


With every photo in iNat we are identifying based on what it “looks more like”, by virtue of them being photos! There is always a chance of misidentification on that basis, and even for the ones we think are very easy and clear-cut now, who knows what splits and taxa re-arrangements will alter the situation in the future.

I believe the identification “belongs” to the identifier, and should represent their own understanding of what species/taxa is represented in the photo/observation. If you have a high level of confidence in the ID put by the observer (or other identifier) then by agreeing with them contrary to your own understanding, you are adding weight to their ID, not putting your own ID. I think this is acceptable even though it is not following iNat guidelines/policy, as it effectively brings a “crowdsourced reputation system” into play, but also fits with the community ID model, so it is best of both worlds!

For me, the question is “what are the chances of strays, and what are the implications of strays… how will the population interact with the stray, and what effect would that gene insertion do to that population at large and over time.” If occurrence records showed the stray, would it be excluded from studies as being an outlier? So for most situations, yes there is a chance it could be the other ssp, but even if it were it might not change anything.

In a census, for instance, if someone identifies themselves as the majority religion even though they are not, it is an inaccuracy but that one error doesn’t alter the “big picture”. Yes, it is better to get the data accurate, but it doesn’t break the data in all situations where that happens.

1 Like

I’m always one who always wants to identify to subspecies when possible, especially when it’s birds. And as the debates explained above, I’ve received mixed opinions on this method of identification, mostly on the negative side of things because I “identify by range”. However, that’s only partially true because here’s how I identify.

I will add all possible subspecies to a list which varies among families, examples being if a species is more stationary like woodpeckers or chickadees, I add any subspecies to my list if they live in or in adjacent counties in my area. If they are something more mobile like hawks or sparrows, I may add all subspecies in my state or my entire tri-state area, especially if there’s subspecies that breed in western Canada or Alaska.

And for more realistic examples, I know for a fact that 99.9% of ALL my Red-tailed Hawks are going to be Western (calurus) subspecies because of first, range and second plumage but I know Harlan’s winter in my area and if I pay attention to thin patagials or heavy bellybands, I may get borealis or abieticola in migration season.

Another example being, in range I should be getting the Interior subspecies annectens Steller’s Jay in my backyard but yet they don’t have the white streaking on the forehead, throat or supercilium of the subspecies, so I know it’s the Coastal subspecies frontalis because the forehead is blue, despite the fact they should live a mountain range over in the west. So I use range to identify but when I know a secondary subspecies is very well possible, I always doublecheck.