What's a cackling goose?

I think I'm as confused as I've ever been, and have come to the conclusion that I have no idea where to draw the line in the sand between Cackling and some Canada geese.  

The wonderfully mud-clear nomenclature is ironically representative of the problem.  "Cackling goose" used to refer to a subspecies of Canada goose found in coastal Alaska, a small very dark bird that formerly went by the trinomial Branta canadensis minima.  The form of small goose that migrates through Michigan, the "Richardson's goose," went by the trinomial of Branta canadensis hutchinsii (nice that the bird is named after 2 different people).  Now several of the small subspecies of white-cheeked geese are lumped as Cackling Goose with the scientific name Branta hutchinsii - a full species now goes by the former common name of one of the subspecific forms and the former scientific name of a different subspecific form (and includes a couple others as well).

In Michigan 3 forms of white-cheeked geese occur (to my knowledge):  the non-migratory and introduced "Giant" Canada (maxima subspecies), the migratory "Interior" Canada (interior subspecies), and the "Richardson's" Cackling (hutchinsii subspecies).  A problem comes in that runty "Interior" Canadas are known to occur.  Size is a good screening test to look for candidate cacklers, but is not all a person can go on.  A proportionately small bill is said to be strong support for cackling, but there are other plumage differences (grayer plumage, especially of the breast, in "Richardson's" hutchinsii Cackling as compared to Canada, with more brightly edged scapulars in the "Richardson's" as well).  The scapular edging brightness is shown in the illustration (though not "fieldmarked") in the big Sibley, but is explicitly described in Sibley's web article as a good mark for "Richardsons" Cacklers.

I started my dilemma with some birds recently photographed in Ann Arbor; photos by Laurent Fournier can be viewed here.  There's a large Canada, and 2 smaller white-cheeked geese, one obviously a size bigger than the smallest, but still much smaller than the largest.  To my eyes there are small plumage differences between the smaller birds and the Canada:  their breasts are slightly grayer and there is the suggestion of a white collar at the base of the neck.  To my eyes there is no difference in the edging of the scapulars.  I have always called birds like these runty Canadas.  Contrast those birds with another small Ann Arbor goose viewable through the umichbirders Grovestreet.com web album here photographed by Lyle Hamilton at Avis farms. This bird has quite brightly edged scapulars and an even more prominent collar at the base of the neck, though appears very similar in tone and color to the surrounding Canadas in similar postures.

 This is a bird I photographed on Ford Lake in Ann Arbor in mid-March a few years ago.  I labeled it a runty Canada.  It follows the pattern of the first photo, a smaller bird with a slightly browner breast, but scapulars essentially identical to the surrounding large Canadas.  It does have a collar at the base of the neck, unlike most Canadas.  It was slightly larger than an even smaller bird that was also present that I labeled a cackler (though that I didn't manage great photos of).  The bill is smaller than those of the surrounding birds, but is it small enough?

This is a very similar bird I photographed at the LMC ponds here in Berrien Co in mid-October 2 years ago.  Again, a bird with a brown-gray breast making a whitish collar visible at the base of the neck, a small though not stubby bill, and the same brown color of the back with little edging of the scapulars:

Compare that with these birds in similar lighting from last October, small geese that are much grayer overall with brightly edged scapulars, birds that I'm confident are cackling geese.

The final 2 pics are also from the LMC ponds from last December of 2 cackling geese that show all the differences, small size, short bills, grayer plumage, and brightly edged scapulars (note that the lead bird is more brightly edged than the trail bird)

So what to do with the small brown birds that lack the bright scapular edging?  I think that Lyle's Avis Farms photograph of a Cackler does show that not all Cackling geese in all lighting conditions are as silvery-gray as easier ones.  I don't know, however, that it convinces me that all the brown birds without bright scapular edgings are Cacklers.  

Sibley spends a good amount of time in his web article discussing bill size, going so far as to say, "Actual and proportional bill length may be the single most useful feature to study when trying to identify Canada and Cackling Geese."  He reviews data from a paper and book which I don't have access to (so can't fully analyze the original data).  However, interesting graphs are shown, demonstrating average bill size (including the standard deviation, a measure of how variable a sampled population is).  The general trend is that small birds have small bills and large birds have large bills.  Most of the subspecies groups have a standard deviation of about 5% or so, meaning an average bill size of 50mm long would vary between about 47 and about 53 mm in most birds.  However, there are 3 groups that have virtually no standard deviation (variability) amongst the sampled population: female moffiti Cacklers from the interior West, and male and female hutchinsii Cacklers.  In all groups except the moffiti Cacklers, the sampled males and females show about the same standard deviation (variability) in bill size.  It doesn't make sense then, that in a single subspecies of white-cheeked goose that there is a difference in the variability of the male bill and the female bill.  Therefore, I think we can conclude that the sample size was relatively small and that the females of that group likely do have the same variability as all other forms of white-cheeked geese.  Extending this out, it would also seem to make sense that hutchinsii Cacklers also have some variability in bill size as well.  The significance is that our migratory interior Canada geese were measured as having a bill size of about 51mm with a standard deviation of about 4mm.  That means (based on the standard deviation calculation and assuming a "normal distribution" of the data) about 2/3 of the interior Canada geese will have a bill size between 47 and 55mm.  A hutchinsii Cackling goose would seem to average a bill length of about 38mm with 2/3 having a bill size between 36 and 41mm.  How obvious in the field, therefore, is the difference between a large Cackler at 41mm and a small interior Canada at 47mm.  Even if you can easily see a 6mm difference in the field present between average birds, what of the runty Canadas that are know to occur???  An interior Canada 2 standard deviations from the mean (again assuming a "normal distribution") would have a bill length of 44mm and would comprise about 2% of females (or 1 out of 100 birds in a flock).  Now we're down to a 3mm difference.  

A critical reader (if there's any readers left at this point) could correctly point out that all those numbers do depend on whether white-cheeked goose bills vary on a "normal distribution".  Honestly, I don't know that it really matters.  What matters is that we know that runty Canada geese occur (as well documented by Julie Craves' classic online article).  They can be small birds and, at least in her photos, have a small bill as well.  It would be easy to say that those birds could fall 2 standard deviations from the mean, even if there aren't correspondingly huge-billed birds to make the data distribution truly "normal."  

Sibley, however, interprets the data differently.  In his Figure 1 on the far right side of the table is a comparison between captive-raised interior Canada geese on Akamiski Island and wild interior Canada geese on Akamiski Island.  He observes that captive raised birds are only slightly larger than the wild birds.  HOWEVER, that the paper's comparison is actually between captive-raised birds and average Akamiski birds can also be seen (the bill length of the comparison wild interior Akamiski birds are the same size as the other interior birds in the table).  The paper doesn't therefore appear to give us data on runty interior birds and so does not allow us to conclude that a runty interior Canada goose's bill would be only slightly smaller than a typical interior's.  (Of course it would help to have the actual paper to know that for sure ... perhaps I'll be writing a retraction...).

If, however, a person takes Sibley at his word that, "There is no evidence that the "runt" birds from Akimiski Island take on the short-billed proportions of smaller subspecies, and therefore should not cause confusion with Cackling Goose," then most, if not all, the birds discussed here are cackling geese... ... unless some of the borderline birds are parvipes Canada geese, the smallest Canada goose subspecies, that's much closer in size to hutchinsii Cackling geese than are the larger interiors.  Different range maps show parvipes as either slightly west of hutchinsii Cacklers or actually overlapping them somewhat, so it is conceivable that they could occur in Michigan as well.

At any rate, I haven't decided what to do with the small plain-looking birds.  I'm not convinced that I can truly say whether a bird's bill is stubby enough to be a Cackler's even if it's shorter than most of the Canada's that it's around.  One other potential issue is that most of the birds with nice brightly-edged scapulars are from November and December, most of the plainer birds are from earlier in the year.  I don't know if this implies again a different population of birds, or if possibly moult differences could be involved (though I would think that migrating waterfowl would be migrating with moult pretty much completed).  I think for me the scapulars are the best way to rule in a Cackler but can't decide if lack of bright edges rules them out (which I've typically done in the past).  At any rate, it's 4am and I think I'm finally tired enough to sleep...