What are the criteria for determining a subspecies? (1 Viewer)

Gentoo · Sep 6, 2009

Just curious what it takes to constitute a subspecies. Some species have some differences regionally but are not considered subspecies. Herring Gulls in North America are an example I can think of. There are differences in wing tip patters and often in size between eastern and western NA populations. However unlike similar differences in Europe where there are two subspecies, the NA birds are not recognized as subspecies. Likewise with the NA and Asian populations of Glaucous-Winged Gull.

So what does it take for two populations to be considered subspecies?

njlarsen · Sep 6, 2009

I did not go to some textbook for this, so there may be details where I am wrong, however, this is my understanding:

A subspecies is a geographically defined population that differs in some defined way from another population. If it is impossible to say where one population ends and the other starts (because there is a very wide cline) then there really is one population and they are the same subspecies, even if the most distant birds are very different.

If there is 500 miles with type A, 100 miles of gradual change, and 500 miles with type B, then you have two subspecies, but if there is 1100 miles with gradual change, then you don't.

It is, however, my understanding that there is a certain level of judgment call involved in determining whether two populations are different or the same subspecies.

Niels

Gentoo · Sep 7, 2009

njlarsen said:
I did not go to some textbook for this, so there may be details where I am wrong, however, this is my understanding:

A subspecies is a geographically defined population that differs in some defined way from another population. If it is impossible to say where one population ends and the other starts (because there is a very wide cline) then there really is one population and they are the same subspecies, even if the most distant birds are very different.

If there is 500 miles with type A, 100 miles of gradual change, and 500 miles with type B, then you have two subspecies, but if there is 1100 miles with gradual change, then you don't.

It is, however, my understanding that there is a certain level of judgment call involved in determining whether two populations are different or the same subspecies.

Niels

I see what you're saying but I will use another gull as an example. Western Gulls occur in two subspecies; a northern and a southern. There is a broad cline as you move through the range. Near the "step" in central California they are much closer in appearance than birds from say San Diego and Oregon. Perhaps there's something I am missing?

Richard Klim · Sep 7, 2009

I think Niels has described very well how subspecies should be defined, in an ideal world.

In practice, ornithologists have named large numbers of populations on very dubious grounds. But as long as the name and description are satisfactory, the subspecies enters the annals of ornithological record.

The mechanism for subsequently dismissing subspecies of questionable value seems to be informal and subjective. Birding literature is littered with comments such as 'subspecies x, y and z, but variation clinal', or 'subspecies a and b defined, but of questionable diagnosability', etc.

Some authors make a good effort to critically assess the merits of defined subspecies and assertively state which should be synonymised, eg, Shirihai et al 2001 (Sylvia Warblers), Alström et al 2003 (Pipits & Wagtails), Kirwan et al 2008 (Birds of Turkey), some authors of HBW (but this is extremely variable). But in most cases the same old subspecies are perpetuated (perhaps with just a mildly critical comment about diagnosability), often even resurrecting examples which have been clearly discredited elsewhere. And it’s unfortunate that AOU has ducked the issue of subspecies since the Check-list 5th Edition (1957) - hopefully that’s soon to change. Perhaps one problem is that opinions on subspecies validity are often expressed in popular birding literature rather than in peer-reviewed papers.

Probably the most respected world listing of subspecies is Dickinson 2003 (H&M3), which is taken as a baseline by many authors, and already synonymises many names. I’m hoping that when the 4th edition is published in a couple of years time, Edward and his team will have been bold enough to considerably further rationalise the recognised subspecies.

Richard

wintibird · Sep 7, 2009

Richard Klim said:
In practice, ornithologists have named large numbers of populations on very dubious grounds. But as long as the name and description are satisfactory, the subspecies enters the annals of ornithological record.

Some obviously just wanted to place their name...

Others found a different population but were unaware that the variation is clinal. Take your gull, Gentoo. If one scientist describes a bird from the south another one can happily describe a northern bird as a different subspecies. Later they will found that the variation is clinal and that things are not so easy as they first seemed to be.
Like with defining a species we will never be able to do it properly. Evolution is going on, some now clinal variations may develop into different species. Other variations may dissapear. And all this processes are in different stages in different species right now. The scientific naming system is imperfect by definition.

André

njlarsen · Sep 7, 2009

As Richard says, there is a certain level of subjectivity in how subspecies are defined, and to my mind, a certain level of subjectivity in more or less all levels of taxonomy. However, a good example of how subspecies could be refined is the recent paper on Blue-crowned Motmot (see this thread).

I also agree with Andre that our taxonomy system is trying to draw sharp lines when in reality we are looking at cases that together comprise a continuum. This corresponds a little to trying to fit a round peg in a square hole -- it will never be perfect. This is one of the reasons that a certail level of subjectivity is built into the system.

If I have understood your Western Gull complex correctly, this is one example where there never will be one correct answer; or at least not within our lifetime, because it might become a more clearcut case over the next 1000 generations

.

Niels

Gentoo · Sep 8, 2009

wintibird said:
Some obviously just wanted to place their name...
Others found a different population but were unaware that the variation is clinal. Take your gull, Gentoo. If one scientist describes a bird from the south another one can happily describe a northern bird as a different subspecies. Later they will found that the variation is clinal and that things are not so easy as they first seemed to be.
Like with defining a species we will never be able to do it properly. Evolution is going on, some now clinal variations may develop into different species. Other variations may dissapear. And all this processes are in different stages in different species right now. The scientific naming system is imperfect by definition.

André

I get what your saying. This makes perfect sense. Birds, unlike mammals, seem to evolve quickly, sometimes very quickly. House Sparrows have already developed regional differences in the US and then there's that case with a certain Darwins Finch which after just 15 years, grew larger with a different bill and song than the population they came from. So I guess it could be that once we decide on a population, the line can become blurred in less than a lifetime.

Gentoo · Sep 8, 2009

If I have understood your Western Gull complex correctly, this is one example where there never will be one correct answer; or at least not within our lifetime, because it might become a more clearcut case over the next 1000 generations

Good points. This gull was debated as possibly being conspecific with Glaucous-Winged Gull because they hybridize. But if they use this logic alone, then all large gulls are the same species. Also genetic evidence suggest that relationships may be different than we expected among these and other gulls.

beakgeek · Dec 23, 2009

Gentoo said:
Good points. This gull was debated as possibly being conspecific with Glaucous-Winged Gull because they hybridize. But if they use this logic alone, then all large gulls are the same species. Also genetic evidence suggest that relationships may be different than we expected among these and other gulls.

All of this is dependent on which species concept you apply - Biological vs Phylogenetic.

I worked in Bob Zink's lab at the U of MN where a lot of species definitions are based on phylogenetic work. What is intriguing is that some species that we thought to be sister taxa are in fact not based on where the split in the tree took place.

I'm curious what genetic evidence/research you are referring to on the gull relationships. Was the Douglas A. Bell's work?

It is a well done study basing hybridization on morphology and phylogeny.

Mysticete · Dec 23, 2009

subspecies are arbritrary, way way way more arbitrary than species. Even in the 50's, before the rise of other species concepts, people argued whether the subspecies concept was valid at all.

I have been doing some extensive literature reading over the last few weeks for the taxonomy paper I am coauthoring (pinnipeds, not birds): Best quote I have so far found and which rather nicely reflects my belief I have pasted below

"subsumed into a black hole of terminal taxonomic neglect and ambiguity, better known as the subspecies category" Wiley and Mayden, 2000

njlarsen · Dec 24, 2009

Living on an island probably biases my view of a lot of things, including my approach where I tend to think about the problems when populations are truly non-overlapping. In those cases, there is often a good reason to think in terms of subspecies. For example, our yellow warbler is yellow with a chestnut cap, the one in the next island (Martinique) has a chestnut head. Now, unless this is seen as a cause for splitting into species, then subspecies distinction seems in order. Saying the difference is clinal when in fact the cline would have to be positioned in the Martinique Channel, seems to me like nonsense.

Edit: I am not saying that the critique of subspecies as the term is currently used is wrong; just that I do see a use for the term.

Cheers
Niels

mb1848 · Dec 24, 2009

Here is a good example of how to do subspecies with a very confusing species. Song Sparrows. Patten et al take 51 subspecies and make them into 25:
http://journals.cambridge.org/actio...392&jid=SYS&volumeId=7&issueId=01&aid=3760384 .
They use Patten & Unitt's 75% Rule.
http://www.biosurvey.ou.edu/patten/Auk2002.pdf .

Richard Klim · Dec 24, 2009

mb1848 said:
Here is a good example of how to do subspecies with a very confusing species. Song Sparrows. Patten et al take 51 subspecies and make them into 25

I fully agree that many subspecies require rationalisation, and it's good to see an example of proposed synonymisation justified via a peer-reviewed paper. As I commented earlier (post #4), proposals for subspecies synonymisation are more often made in books etc, and are frequently disregarded by later authors.

In this case Byers et al 1995 (and Dickinson 2003) had already reduced the number of subspecies to 39; and Arcese, Sogge, Mark & Patten 2002 (BNA Online) - following Patten 2001 - reduced this further to 24. Patten & Pruett 2009 restores zacapu, which had been synonymised with adusta.

Richard

trundlebob · Jan 20, 2010

Richard Klim said:
I fully agree that many subspecies require rationalisation, and it's good to see an example of proposed synonymisation justified via a peer-reviewed paper. As I commented earlier (post #4), proposals for subspecies synonymisation are more often made in books etc, and are frequently disregarded by later authors.

Richard

I was wondering. Sometimes subspecies are rationalised but then make a bold reappearance after some time in the dark. The Dutch lumping of Taiga Bean Goose into monotypic A fabalis dismissed both johanseni and middendorffii as invalid. After some years of it not even being a subspecies, middendorffii resurfaced ... but this time accepted by the same authority as a species in its own right, separate from fabalis. Wrong guesses turn subspecies rationalisation into a rather precipitous game ...

Richard Klim · Jan 20, 2010

trundlebob said:
I was wondering. Sometimes subspecies are rationalised but then make a bold reappearance after some time in the dark. The Dutch lumping of Taiga Bean Goose into monotypic A fabalis dismissed both johanseni and middendorffii as invalid. After some years of it not even being a subspecies, middendorffii resurfaced ... but this time accepted by the same authority as a species in its own right, separate from fabalis. Wrong guesses turn subspecies rationalisation into a rather precipitous game ...

Yes, that's a very good example indeed of a rather brief and premature removal. But at least the name middendorffii remained available (although 'in the dark') to be resurrected if ever justified. The other extreme would require continued recognition in ornithological literature of all the taxa that have ever been named just in case they ever warranted elevation to species rank.

My main concern is the retention of subspecies that are so questionable that they are not truly in everyday use except within faithfully replicated lists in reference works (or can only ever be identified by assumption based on location).

Richard

trundlebob · Jan 20, 2010

My main concern is the retention of subspecies that are so questionable that they are not truly in everyday use except within faithfully replicated lists in reference works (or can only ever be identified by assumption based on location).

Agreed; I suspect there are many cases where single (or groups of) subspecies could be confidently coagulated, or even dismissed. But there are perhaps as many cases (geese, gulls, shrikes are amongst the recently popular) where low priority geographically-based subspecies may be more useful ... not perhaps royalty in their own right, but as lower value cards that help validate or illustrate taxonomic reshuffles. Didn't such groups at least briefly court favour as ESUs or BSUs or something? I worry slightly that subspecies may sometimes be too easily dismissed on a care-less basis; the recognition of many forms of White-fronted Goose, for example, is of limited use to compilers of world species lists, but could be the meaning of life to a full-time goose researcher... ie could, much like middendorffii, albicans be a latent split?

Richard Klim · Jan 20, 2010

trundlebob said:
But there are perhaps as many cases (geese, gulls, shrikes are amongst the recently popular) where low priority geographically-based subspecies may be more useful ... not perhaps royalty in their own right, but as lower value cards that help validate or illustrate taxonomic reshuffles. Didn't such groups at least briefly court favour as ESUs or BSUs or something? I worry slightly that subspecies may sometimes be too easily dismissed on a care-less basis; the recognition of many forms of White-fronted Goose, for example, is of limited use to compilers of world species lists, but could be the meaning of life to a full-time goose researcher... ie could, much like middendorffii, albicans be a latent split?

Fair points, and I agree that any synonymisation should always be carefully considered.

Re Anser albifrons 'albicans', an earlier thread failed to establish the rationale for associating that name with the NE Siberian population:
http://www.birdforum.net/showthread.php?t=152366

Richard

What are the criteria for determining a subspecies? (1 Viewer)

More options

Gentoo

Guest

njlarsen

Gallery Moderator

Gentoo

Guest

Richard Klim

-------------------------

wintibird

André Weiss

njlarsen

Gallery Moderator

Gentoo

Guest

Gentoo

Guest

beakgeek

Well-known member

Mysticete

Well-known member

njlarsen

Gallery Moderator

mb1848

Well-known member

Richard Klim

-------------------------

trundlebob

New member

Richard Klim

-------------------------

trundlebob

New member

Richard Klim

-------------------------

Similar threads

Users who are viewing this thread