Identification of Canada and Cackling Goose

David Allen Sibley

One of the most significant changes for birders in the 45th supplement of the AOU checklist (Banks et al 2004) was the split of Canada Goose into two species. While widely anticipated - the highly variable Canada Goose has often been considered more than one species in the past - the split has thrust a complex and poorly understood field identification problem into the forefront. I have tried to pull together some of the basic information about subspecies and variation to begin the approach to identification of these species.

I must stress that this is a draft, and should not be considered anything more than my personal notes on the problem of Goose identification. It stresses the problems in eastern North America, and includes lots of unanswered questions. I hope to add more detail in the future. Please direct any comments or discussion to me or to the ID-Frontiers listserv.

What is the "Cackling Goose"?
First, to clear up some confusion about the names of the species and subspecies: The former broad Canada Goose has been divided into a large-bodied, interior- and southern-breeding species, and a small-bodied tundra-breeding subspecies. The large-bodied group is still known as Canada Goose (Branta canadensis) while the small-bodied group takes the name Cackling Goose (Branta hutchinsii). This means that the English name Cackling Goose, which has in the past been more or less restricted to the smallest subspecies (the far western B. c. minima) is now the species name for all four of the small subspecies. This new species takes the scientific name of the earliest-named subspecies and becomes Branta hutchinsii.

When reading published accounts of subspecies, it is important to bear in mind that in the confusing history of the taxonomy of Canada Geese, the subspecies have been grouped and named in various ways. Names such as leucopareia (which is now restricted to the Aleutian population) have been applied to much broader groups in the past. The names attached to individual specimens in past decades might not be the same names that would be used in the current subspecific arrangement… beware.

The subspecies defined below should be taken cautiously as merely one interpretation of the information at hand (which is patchy and incomplete, despite decades of study).
 

• B. c. canadensis - Atlantic Canada Goose - large, light colored, with white at base of neck relatively clear and extending to back and forming a sharp line with black neck
• B. c. interior - Hudson Bay Canada Goose - Similar in size to canadensis, mantle somewhat browner and darker, with dark continuing uninterrupted to black of neck, feather edges on dorsum average darker, contrasting less with black of neck, breast light to medium grayish towards slaty gray.
• B. c. maxima - Giant Canada Goose - nearly extirpated in early 1900s, but now reintroduced and common. Very similar to moffitti and merged by Palmer 1976. Often very difficult to distinguish from moffitti. Very large, appearing rather pale overall, especially on underparts, white on cheeks tends to extend somewhat farther up sides of head than on canadensis.
• B. c. moffitti- Moffitt's or Great Basin Canada Goose - Similar to maxima in size and color, may have white markings on forehead, intermountain birds may have dark throat stripe.
• B. c. parvipes - Lesser Canada Goose - reportedly intergrades with taverneri in w AK, with hutchinsii throughout tundra, with maxima and moffitti to south. Merged with taverneri by Palmer 1976; but mtDNA shows little or no interbreeding. Medium-size, similar in color to moffitti, with pale to dusky breast.
• B. c. occidentalis - Dusky Canada Goose - a medium large goose, dark overall, underparts chestnut to dark chocolate brown (except white vent), feather borders on dorsum rather narrow, reddish brown; upper breast dark, merging with neck.
• B. c. fulva - Vancouver Canada Goose - sometimes merged with occidentalis (Palmer 1976) but disjunct range, moderate differences in morphology, and differences in mtDNA. Slightly larger than occidentalis, and darker on average, but otherwise similar in appearance.

• B. h. hutchinsii - Richardson's (or Hutchins's) Cackling Goose - reportedly intergrades with parvipes throughout range in NWT and Nunavut, but this is uncertain. Small and rather light, pale breast.
• B. h. taverneri - Taverner's (Alaska) Cackling Goose - may intergrade with parvipes in interior AK. Merged by Palmer with parvipes but has unique mtDNA. Similar to leucopareia but slightly larger and lighter in color, with rounder head. Similar to parvipes but breast slightly darker.
• B. h. minima - Cackling Cackling Goose - Smallest, with small bill and short neck but relatively long legs; variable color and pattern but typically quite dark brown with purplish cast on breast, bill stubby, straight to convex culmen, nail less elongated than leucopareia, white cheeks more extensive than leucopareia.
• B. h. leucopareia - Aleutian Cackling Goose - includes asiatica (extinct). Larger than minima, with paler breast usually gray-brown to dark brownish; white collar usually complete and rather thick with blackish feathering at base of neck; head rather square profile; bill short, tapering to narrow tip and somewhat pointed nail; white cheek patches somewhat more restricted, nearly always black throat stripe.
  

Summarizing the subspecies
It may be helpful to simplify the eleven subspecies for field identification purposes. Basically all of the geese nesting in the lower 48 United States and southern Canada are uniformly large and pale. Thus moffitti, and maxima are essentially the same. The two large subspecies nesting in eastern Canada are also relatively large and pale - canadensis and interior. These two may be distinguishable as flocks of slightly smaller, differently colored birds when they arrive in the fall among the large resident geese in the south, but essentially they too are simply large Canada Geese. The two Pacific Northwest subspecies - fulva and occidentalis - are dark like so many other birds in that area. The Dusky Canada Goose B. c. occidentalis, which has a very limited range and small population, can be fairly small, and nearly overlaps in size with Taverner's Cackling Goose, but most should be distinguishable by size. This leaves Lesser Canada Goose to be compared with the various Cackling Geese.

Cackling Cackling Goose B. h. minima is strictly a Pacific population, with few if any acceptable records east of the Rockies (they are fairly common in captivity). It is also the smallest and shortest-billed, with no overlap in size with Lesser Canada Goose. Aleutian Cackling Goose B. h. leucopareia is a rare goose with a very limited range and few records east of the Sierra-Cascade mountains. Taverner's Cackling Goose B. h. taverneri nests in Alaska and apparently winters mainly in Washington and Oregon, and must be rare east of the Rockies. Richardson's Cackling Goose B. h. hutchinsii nests in Arctic Canada and winters mainly in Texas and Mexico. Thus Richardson's and Taverner's Cackling Geese are the heart of the problem. Distinguishing both of these from Lesser Canada Goose will require a better sense of the variation in all three subspecies, since bill measurements overlap.

Identification Tips

• Watch for subflocks - Within any flock of geese one can find family groups and other subflocks sticking close together. Watching for these groups can aid identification by highlighting birds that are traveling together, presumably from the same nesting area and quite possibly related. In these cases it might be possible to identify the male (larger and more alert) the female, and the young of a family group, and once the age and sex is known impressions of size and color are somewhat more meaningful. Identifying a flock will usually be easier than identifying an individual bird.
• Beware of impressions and prejudices - It is possible for two people to look at the same flock of geese and come away with very different impressions of the "average" appearance of those geese. It is too easy to look at a flock, pick out a few darker or paler birds, and say "Yes, this subspecies is normally darker/paler than others", while another observer could come away with the opposite impression, guided by preconceptions or just by random chance. Even if the conclusion is accurate - "this subspecies averages darker" - that observation is of little use without more information on the range of variation and the percentage of birds that do not match the average. Such "average" statements are particularly useless when we are confronted by a solitary odd goose. A thorough survey of each flock, with some guidelines for judging color, would be a more reliable and useful exercise, and that kind of study has been sorely lacking in Canada Goose identification.
• Know what you are comparing to - since identification requires judging subtle differences in average size and color, it is critical to know your local flocks so that you can judge the relative size of an interesting bird against known subspecies.
  

Bill size
Actual and proportional bill length may be the single most useful feature to study when trying to identify Canada and Cackling Geese (see Figure 1). In general the smaller subspecies have shorter, stubbier, proportionally smaller bills, while the larger subspecies have longer, proportionally larger bills. Unfortunately there seems to be overlap between the smaller populations of Canada Goose and larger subspecies of Cackling Goose.

Figure 1. Mean culmen length of 11 subspecies of Canada and Cackling Geese (from Mowbray et al 2002 and Leafloor et al 1998). The four on the left are included in the new Cackling Goose, while the nine on the right make up the new, restricted, Canada Goose. The two samples on the far right from Akimiski Island are the subspecies interior, and show data from Leafloor et al., demonstrating that captive-raised birds grew larger than similar birds raised in the wild there.

Another way of looking at these numbers, and perhaps a better indication of how useful bill length might be for distinguishing subspecies in the field, is shown here in figure 2.

Figure 2. Culmen length for males and females of each subspecies, showing range of variation (plus or minus one standard deviation; from Mowbray et al. 2002). Note that these measurements come from multiple sources, and may not be fully reliable because of different methods used or because of the particular population measured. At least these numbers show that there is overlap between some parvipes and the slightly smaller taverneri and hutchinsii

Size
Size and proportions (including bill size) are very useful for identification. See Figure 3 for a graph of average body mass of the eleven subspecies. Getting a sense of overall body size often requires watching a bird from many different angles as it moves around. One bird may appear smaller when standing in a low spot (or vice versa), seeming to change size when it takes a few steps. A view from head-on or tail-on, rather than from the side, is often the best chance to judge body size. Also be aware of differences between males and females (males average from 3 to 8% larger than females), immatures and adults (first-winter birds average much lighter weight than adults), as well as differences within each subspecies (discussed below).

Compounding the problem of distinguishing subtle size differences, it has been shown that environmental factors (probably nutrition) can dramatically influence the overall size of individual geese. Thus birds raised in environments with reduced quality and/or quantity of food (such as some northern goose nesting colonies that are overcrowded) fail to reach their full potential size, and grow into relatively small geese regardless of their genetic background. Leafloor et al (1998) report that captive-raised birds from Akimiski Island in James Bay have culmen length averaging 5 to 8% larger than wild birds from the same nesting area.

While this is certainly a large enough difference to be visible in the field, and should be considered when judging size of geese, it may have been overstated in recent identification discussions. Judging from the data here, the difference between the "runt" wild-raised birds from Akimiski Island and their better-nourished counterparts is not enough to cause confusion of these small Canada Geese B. c. interior with Cackling Geese B. h. hutchinsii, for example (see Table 2). The four to six percent difference in size is about the same as the difference between males and females within each subspecies and is less than the natural variation within each subspecies (one standard deviation equals plus or minus about 4 to 8%, with further variation geographically; data from Mowbray et al 2002). Leafloor et al report that the "runt" birds from Akimiski Island match the size of the same subspecies from 600 km north in Manitoba, tracking the gradual decrease in size northward. Leafloor et al do report some differences in proportions, primarily that while captive-raised adults have culmen length 5 to 8% larger, tarsus length is 11 to 14% larger than wild birds. This is not enough data from which to draw any firm conclusions, at least it shows that the tarsus becomes proportionally smaller than the bill in "runt" wild birds, and suggests that these birds might actually appear proportionally large-billed. 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.

A website shows photos from Michigan of a banded goose from Akimiski Island that is smaller than its associates. One photo shows the bird looking very small, while another shows the expected subtle difference. The accompanying discussion implies that Akimiski Island birds are small and dark, but no color differences are mentioned by Leafloor et al (1998). I don't know if there is evidence of such "environmental darkening".

Neck length
Geese have a remarkable ability to extend or retract their neck, and within a flock you will see birds change from very short (relaxed) necks to very long slender (alert) necks. In flight the neck is always extended and reliable comparisons can be made, but sitting or swimming birds must be watched for some time to develop a sense of the true neck length. Related to this, with the head feathers fluffed in a relaxed position the head may appear rounder and the bill smaller. This combined with the shortened neck makes it easy to jump to the conclusion that you are looking at a Cackling Goose.

Overall color
The overall color varies with subspecies from dark chocolate brown or slaty-brown to pale gray-brown. The most useful details to focus on are the breast color, the upperparts color, and the color and contrast of the pale tips of the upperparts feathers. I think the upperparts are more likely to be useful for identification than the breast color. Having said that, each subspecies shows a range of color variation, and all descriptions of color will mention average differences and one subspecies "usually" showing something. One need only spend a few minutes looking at any flock of large Canada Geese to appreciate the potential for variation in color. In general birds are darker in the west and paler in the east, but even within the darkest subspecies (B. h. minima and B. c. fulva) some individuals can be quite pale-breasted. The extent of this variation has not been adequately described, and it is likely that small and variable differences in color, such as between B. h. hutchinsii and B. c. interior, will be of limited use for identification of individual birds.

White collars
A white collar at the base of the neck below the dark "neck sock" is a conspicuous feature of some Canada and Cackling Geese. It tends to be more common in the smaller subspecies and is most conspicuous in the dark-breasted ones, since it contrasts more with a dark breast than with a pale one. A survey by Marquardt (1961) of presumed B. c. parvipes and B. h. hutchinsii in the southern Central flyway (Colorado, Oklahoma, and Texas) found that nearly 50% of adults and 25% of immatures had white collars, even though Palmer (1976) and Mowbray et al (2002) report that these two subspecies seldom show a white collar. Even in B. h. leucopareia, the Aleutian Cackling Goose that is characterized by a broad white collar, Palmer (1976) reports a survey of nesting birds on Buldir Island, Alaska in which 47 birds had a white collar, 9 had no white collar, and 8 had breasts so pale that a white collar was not discernible. Perhaps with more data the presence of a white collar will have some value as a weak indicator of one subspecies over another, but at this point it seems nearly worthless.

Black chin stripes
The presence or absence of a dark stripe on the chin dividing the white cheek patches has been mentioned in recent discussions as favoring one subspecies or another. I have no personal observations on this, but from the literature it is clear that this feature has been mentioned and repeated, but no comprehensive survey has been done. It is reportedly shown often or "nearly always" by B. c. moffitti and B. h. leucopareia and often by western populations of B. c. interior. It is reportedly shown occasionally or "usually not" by other subspecies, but given the variation and the lack of information, I think it should be considered as having little or no value for field identification.

Voice
There are significant differences in voice between subspecies, with the smallest ones comprising the new Cackling Goose all having more or less high-pitched squeaky voices (contra Sibley 2000). My impression is that the voice of B. h. minima is higher and more rapid, more cackling, than all others, while the voice of B. h. hutchinsii is simply a higher-pitched version of the typical Canada Goose honk, but this needs further study. One presumed B. h. hutchinsii that I studied at Cape May NJ gave a call slightly higher but essentially indistinguishable from the surrounding B. c. canadensis or interior; while occasional very high-pitched calls come from flocks of large Canada Geese. Whitford (1998) and Mowbray et al (2002) describe the voice in detail, with important points relevant to field identification. The typical "h-ronk" call is given only by males. Females give a higher-pitched and shorter "hrink" or "hrih". Pitch also changes depending on the position of the neck, and the duration of the call varies depending on context. Dominant individuals are about 60 times more vocal than submissive flockmates! Taken together, this information suggests that the sounds produced by a flock can be a useful clue, but for various reasons voice is unlikely to be useful for identifying single birds.

Wing shape
Subtle differences in wing shape and posture may be useful for identification. B. h. minima has long wings that appear to be pushed forward in flight, while B. h. hutchinsii has long wings that are slightly more swept back and pointed than in larger B. canadensis and it tends to hold the wings more arched above the body. Thus these two small Cackling Geese have different wing shapes, each distinguishable from Canada Goose and from each other. More study is needed on other subspecies, particularly B. c. parvipes, before this feature can be used to identify subspecies.

The identification problem
It is relatively easy to distinguish the largest birds from the smallest ones simply by size and bill proportions. Since all four subspecies of Cackling Geese are small, the major problem of distinguishing the species involves the smallest subspecies of Canada Goose - B. c. parvipes. This subspecies is variable in size and color across its wide range from Hudson Bay to Alaska. At the southern edge of its range it reportedly meets and interbreeds with the largest subspecies of Canada Goose, while at the northern edge of its range across northern Canada and northern and central Alaska it reportedly interbreeds with Cackling Geese of the subspecies B. h. hutchinsii and B. h. taverneri.

Distinguishing B. h. hutchinsii from large Canada Geese is relatively easy. In general hutchinsii is much smaller with a short, stubby bill, relatively round or even square head with steep forehead (vs wedge-shaped head), and relatively short neck. They appear paler silvery gray on the upperparts and upperwing coverts in some lighting conditions, with the feathers having a more conspicuous dark subterminal bar and a broader and more conspicuous pale tip. The breast color averages slightly darker than large eastern Canadas, although many are pale-breasted, and the breast often has a golden wash. The voice is slightly higher-pitched, and the wing shape in flight is different.

On the other hand, distinguishing B. h. hutchinsii from smaller individuals of Lesser Canada Goose B. c. parvipes may be much more difficult. Many Lesser Canada Geese are noticeably larger than hutchinsii, with proportionally large bills approaching the large Canada Geese, but Lessers from the northern parts of their breeding range are reportedly smaller than in the south. More study is needed to figure out the characteristics of small Lesser compared to hutchinsii.

Taverner's is simply poorly known, with a range in western and northern Alaska between the range of Lesser Canada Goose and true Cackling (B. h. minima). It reportedly differs from Lesser in mtDNA, suggesting little interbreeding, but differences in appearance are few and subtle, so that field studies generally claim widespread interbreeding and the entire population on the North Slope of Alaska is listed by Mowbray et al as taverneri/parvipes. Clearly, this is another topic for further study.

Finally, as a cautionary note, Pearce et al (2000) used genetic testing to assess the accuracy of identifications of Dusky Canada Geese B. c. occidentalis at hunter check stations in Washington. They found that one-third to one-half of all the geese identified as Dusky by breast color and bill length (measured in the hand) were in fact not true Duskys. And if that's the level of accuracy achieved by experienced people with birds in hand, what can we expect to do in the field? This is an irresistible challenge for some of us, but until we have a better sense of the variation and limits of these subspecies all identification should be approached with a healthy dose of skepticism.

Literature cited
Banks, R. C., C. Cicero, J. L. Dunn, A. W. Kratter, P. C. Rasmussen, J. V. Remsen, J. D. Rising, D. F. Stotz. 2004. Forty-fifth supplement to the American Ornithologists' Union Check-list of North American Birds. Auk 121:985-995.

Leafloor, J. O., C. D. Ankney, and D. H. Rusch. 1998. Environmental effects on body size of Canada Geese. Auk 115:26-33,

Marquardt, R. E, 1961. Albinism in the small white-cheeked geese. Auk 78:99-100

Mowbray, T. B., C. R. Ely, J. S. Sedinger, and R. E. Trost. 2002. Canada Goose (Branta canadensis). In The Birds of North America, No. 682 (A. Poole and F. Gill, eds.). The Birds of North America, Inc., Philadelphia, PA.

Palmer, R. S., ed. 1976. Handbook of North American Birds; volume 2, Waterfowl (Part 1). Yale University Press.

Pearce, J. M., B. J. Pierson, S. L.Talbot, D. V. Derksen, D. K. Kraege, and K. T. Scribner. 2000. A Genetic Evaluation of Morphology Used to Identify Harvested Canada Geese. Journal of Wildlife Management 64(3): 863-874.

Sibley, D. A. 2000. National Audubon Society The Sibley Guide to Birds. Chanticleer Press, NY.

Whitford, P. C. 1998. Vocal and visual communication of Giant Canada Geese. Pp. 375-386 in Biology and management of Canada Geese (D. H. Rusch, M. D. Samuel, D. D. Humburg, and B. D. Sullivan, eds.). Proc. Intl. Canada Goose Symp., Milwaukee, WI.

Other references:
Letter from Ian Mclaren to ID-Frontiers listserv on July 23, 2004
Information on Canada Geese compiled by Angus Wilson
Summary of Canada Geese in Utah by Mark Stackhouse -
Oregon Department of Fish and Wildlife publication about Canada Goose identification -

Figure 4. Map showing basic range of subspecies of Canada and Cackling Goose. Cackling Goose breeding range in darker pink-red; Canada Goose breeding range in paler orange-red. Winter range of all subspecies in blue, gray lines connect some breeding populations to their principal winter range. Many populations engage in a mid-summer molt migration, traveling up to hundreds of miles north of their breeding range. (based on Mowbray et al 2002 and some comments from Ian Mclaren).