Friday, August 29, 2008

A Swan Showing Bill Characteristics Possibly Intermediate Between Whistling Swan and Bewick's Swan in Connecticut.

On 30 March 2008, I was heading back east from photographing a very interesting oriole in Cannan. I stopped by a small farm pond off of RT 4 in Goshen, Litchfield County, Connecticut to check for Tundra Swans that had been reported there. I immediately located three swans, an adult and two immatures, at the far end of the pond. I moved my car to a side road, East Street South, and attempted to get photos.

This closer position allowed for a better inspection of the birds and I was struck by how much yellow was present on the bill of the adult. A few days prior, I observed four Tundra Swans on a small pond in Eastford, Windham Co., and while the birds were distant, I made notes as to how difficult it was to make out any yellow on the bill. This bird’s yellow patch stood out remarkably well.


Tundra Swan, 30 March, Goshen,
Litchfield Co., CT

I had a discussion with Greg Hanisek, Louis Bevier, and Nick Bonomo about the amount of yellow this bird showed. We discussed the possibility that this could be a Bewick’s Swan, the Asian subspecies corresponding to the North American Whistling Swan. Tundra Swan is comprised of two subspecies, Bewick’s Swan, Cygnus columbianus bewickii, and Whistling Swan, Cygnus columbianus columbianus. Bewick’s is best distinguished from Whistling Swan by the large yellow patch at the base of the bill, often covering the entire basal third of the upper mandible. Whistling Swan typicall shows a small patch of yellow just ahead of the eye on the upper mandible. This patch is small and often difficult to see in the field. Occasionally, there is no yellow present. We all agreed that there was more yellow than we usually see but certainly less yellow than is shown by typical Bewick’s Swan.


Tundra Swan, one of three present on a small
pond in Goshen, Litchfield Co., Connecticut
on 30 March 2008.


Enlargement of the bill showing unusually
large yellow marking.

In July of 2008, I was re-reading an old issue of North American Birds and noted a photograph of Tundra Swans from Washington. These birds showed a similar amount of yellow to that of the March Connecticut bird. In the text, reference was made to an article published in the Auk, the Journal of the American Ornithologist’s Union, wherein a method of quantifying the yellow in the subspecies of Tundra Swans was described. This article also discussed the hybridization of these two subspecies at their natural contact zone. The authors captured a large number of each subspecies and took profile photographs of the bills. They then traced the upper mandible , delineating the yellow area from the black. This tracing was laid over graph paper and the number of yellow squares was compared to the black squares.

Their findings showed that the mean amount of yellow in Whistling Swans was 3.1%. Three percent of their sample of Whistling swans showed no yellow at all and 4.3% had more than 10% yellow on the bill. The largest amount of yellow in a known Whisltling Swan was 15.8%.

Bewick’s swans were divided into three bill-types: Darky, showing a pattern similar to the Goshen bird (17.9% of their sample), Pennyface, with the central black surrounding a yellow spot (19.1% of their sample), and Yellow Neb, the basal third of the upper mandible all yellow (63% of their sample). The mean amount of yellow on the Darky-type Bewick’s Swan Bills was 31.5%. The darkest known Bewick’s Swan bill was 22.9% yellow.

I decided to try this method on photos I secured of the Goshen bird. The best photo I had showed the bird in near profile and I admit that this could build in inaccuracy to my measurement. I enlarged the photo and made a tracing. I laid this over graph paper to quantify the yellow compared to the black. The details of the procedure described in the Auk were lacking and I had to decide how to count squares that showed both yellow and black. I decided to count them as both, feeling that this would eliminate any judgment on my part and make them neutral in the results (I don’t know if this is accurate mathematically but it seemed to work for me).

I counted 38 squares of yellow (including yellow and black) and 135 squares of black (including yellow and black). If any square at the edge of the mandible contained black it was counted as a black square. If any square contained yellow and also contained white as it was against the white of the face, it was called yellow. These numbers worked out to a bill that was 21.96% yellow and 78.03% black. This fell into what the authors described as an intermediate area.



Tracing made from photo diagramming the
distribution of black and yellow on the upper mandible.

Troubled by the assumptions built into my methods, I consulted with Louis Bevier. He told me of a computer program that quantifies an selected area electronically. This program, NIH Image, was installed on his computer and I forwarded him my photograph. On his three attempts, he got measurements of 18.2%, 18.7%, and 18.9% for the amount of yellow. Using his 18.7% as a working average, my square counting method was only 3.26% off. The measurements Louis secured are still in the intermediate range between the two subspecies.

So what does all this mean? Well, the authors of the article in the Auk believe that, using these methods, out of range Tundra Swans can be reasonably assigned to a subspecies. Birds falling in the intermediate range may pertain to intergrade swans, swans from the contact zone between Bewick’s and Whistling. The authors site wild-occurring family groups with each adult belonging to different subspecies.

The Goshen bird? There is not enough yellow to call it a Bewick’s Swan. Is there enough yellow to eliminate a pure Whistling Swan? I am not sure. Certainly, the amount of yellow as quantified by the various methods puts the bird in the intermediate range. If I could have had the bird in a perfect profile I might be more ready to make that type of determination. The Goshen bird showed more yellow than was known from the experience of several long-time eastern birders and a few west coast birders as well. One birder in the west said that a small percentage of the Whistling Swans he sees every year in northwestern United States shows a similar amount of yellow as the Goshen bird. Are these intergrade swans…swans with both Whistling and Bewick’s genetics? Geographically, it would seem possible. There is no way to be certain but the work of Evans and Sladen published in the Auk seems to suggest that this is a real possibility.


Roland Limpert of the Maryland Department of Natural Resources, comments on the Goshen swan, “…The bird in the photo appears to be a Tundra Swan to me based on the amount of yellow. In banding Tundras on the East Coast (Maryland & North Carolina) we would catch birds with a similar amount of yellow not that infrequently and in analyzing more the face photos that we took and banding records we tended to find the Tundras with the most yellow came from breeding grounds on the North Slope and western Alaska. I would say that the bird in the photo was mostly like a Tundra Swan that came from western Alaska.”

This is notable because that origin is where one might expect to see gene flow between the two forms.

I guess I would just call it a really interesting bird.

Remember when, while back in school, you said you didn’t think you’d ever need math in the real world?


Whistling Swan with a typical amount of yellow
on bill. Photo courtesy of Ian Gereg, Avian Director,
Livingston Ripley Waterfowl Conservancy, Litchfield CT.


Enlargement of the bill on the bird shown above.


Typical Bewick's Swan Bill pattern.
Some can show significantly less yellow
than this. Photo courtesy of Ian Gereg,
Avian Director, Livingston Ripley
Waterfowl Conservancy, Litchfield, CT.


Figure 7 Another typical Bewick's Swan
Bill pattern. Photo courtesy of Ian Gereg,
Avian Director, Livingston Ripley Waterfowl
Conservancy, Litchfield, CT.

I would like to thank Louis Bevier, Steve Mlodinow, Dave Shealer of the Waterbird Society, R. Michael Erwin of the USFWS, and Roland Limpert of the Maryland Wildfowl Trust for their help in discussing the identification question, Louis for being a bigger geek than me as he actually owned the quantification software, Ian A. Gereg, Avian Director, Livingston Ripley Waterfowl Conservancy for images of both subspecies, and Joy Mark, Assistant Professor of Mathematics, for showing me the errors of my “gazzinta’s”.


Literature Cited

Evans, Mary E. and William J. L. Sladen, A comparative Analysis of the Bill Markings of Whistling and Bewick’s Swan and Out-of range Occurrences of the Two Taxa., The Auk 97: 697-703, October 1980

Lesser Canada Goose, Branta canadensis parvipes, in Connecticut



It seems good things happen when I chase Greater White-fronted Geese. Remember the Taverner’s Cackling Goose? Well, on 23 March 2008, I was trying to get pictures of a beautiful Greenland-race Greater White-fronted Goose, Anser albifrons flavirostris, which was offering very confiding views on Mirror Lake, the small pond on the University of Connecticut campus in Storrs, Tolland County. While maneuvering for closer photos, I noticed a very small Canada-type Goose resting on the water with the other geese. Thinking it was a Richardson’s Cackling Goose, Branta hutchinsii hutchinsii, I snapped one photo and continued my quest for flavirostris pictures. Only later, while I was processing my images, did I notice something wrong with this identification.

Richardson’s Goose, our expected form of Cackling Goose, shows a short, stubby bill that usually appears convex along the culmen. It sometimes looks rather bumpy and warty (my description, not really in the literature). Hutchinsii also shows a square and rather blocky head shape. This bird showed a bill that was smooth and shaped like a smaller version of a Canada Goose bill. It was long and slender, not at all what I am accustomed to seeing on Cackling Goose. The bird was small, though, as small as a Cackling Goose. In plumage, it seemed identical to the other Canada Geese on the lake.

I started wondering what a Lesser Canada Goose, B. c. parvipes, would look like. A check of on-line images seemed to indicate that this could be what I photographed. I once again sent images to my trio of western goose experts, Steve Mlodinow, Bruce Deuel, and Larry Semo. Mlodinow deferred to the other two as they have more experience with this form.

Again, Bruce and Larry confirmed my identification as Lesser Canada Goose, stating that the head and bill shape and size of the bird were much better for this form than for any Cackling Goose.

Larry Semo states, “Your bird looks like a dead ringer for parvipes to me. Though neck size and head shape could suggest Richardson's, the longer, concave bill and the length of the body relative to the head/neck size/shape would indicate parvipes.”.

Quoting from the Birds of North America Online account for Canada Goose,
“B. c. parvipes (Cassin, 1852); type locality Veracruz, Mexico. Lesser Canada Goose. Breeds throughout boreal regions in widely scattered areas along stream banks, river islands, and beaver (Castor canadensis) ponds, from the central interior of Alaska east through n. Yukon, n. Northwest Territories (Mackenzie River delta), s. Victoria I., Queen Maud Gulf, and e. Nunavut, south to n. British Columbia, n. Alberta, n. Saskatchewan, extreme n. Manitoba to nw. Hudson Bay (Am. Ornithol. Union 1957, MacInnes 1963, Salter et al. 1980, Semenchuk 1992, Smith 1996). Intergrades with B. c. taverneri in western interior of Alaska, with B. c. hutchinsii throughout the tundra portion of its range, and to the south with B. c. maxima and B. c. moffitti (Palmer 1976, Johnsgard 1978); considered morphologically inseparable from B. c. taverneri and treated as a single subspecies by Palmer (1976). Analysis of leg-band recovery data (Migratory Bird Banding Laboratory) indicates that small Canada Geese originating from interior Alaska (presumed B. c. parvipes) and the North Slope of Alaska (B. c. taverneri / B. c. parvipes complex?) winter primarily in e. Washington and Oregon, while small Canada Geese originating in far w. Alaska (presumed B. c. taverneri) winter primarily in w. Washington and Oregon, and n. California (CRE). Others consider B. c. parvipes to winter from w. Washington (Smith et al. 1997) and Oregon (Gilligan et al. 1994) south to extreme n. California (Tule and Lower Klamath Lakes; Small 1994) n. Mexico (Howell and Webb 1995), se. Colorado, ne. New Mexico, nw. Texas, w. Oklahoma (Am. Ornithol. Union 1957, Grieb 1970, Jarvis and Cornely 1988, Small 1994, Jarvis and Bromley 1998). A medium-sized goose, similar in overall shape and color to B. c. moffitti, with pale to dusky breast.”.
As far as I know, Lesser Canada Goose is rarely, if ever, documented in Connecticut and the only other record I am aware of is a sighting by Frank Gallo of a small flock including a Richardson’s Cackling Goose. I have not seen his photos but they should be examined to try and confirm this record. It is likely that hunters take this form during the goose hunting season and perhaps a search of the DEP Game Division database will provide a better idea of its occurrence in the state.
Again, thanks go to the West Coast Anser-men, Steve Mlodinow, Larry Semo, and Bruce Deuel for their information and support and to the vast number of people who post images on the web and thereby allow me to spend even more hours chasing birds at the Geek level.

Literature cited:

The Birds of North America Online (A. Poole, Ed.). Ithaca: Cornell Laboratory of Ornithology; Retrieved from The Birds of North America Online database: http://bna.birds.cornell.edu/BNA/; AUG 2005.

Thursday, August 28, 2008

Taverner's Cackling Goose, Branta hutchinsii taverneri, in Connecticut.


It was my birthday. November 30, 2007. I was going to spend the day birding to see what fortune this day had to offer. After killing my battery trying to photograph Horned Larks at Hammonassett Beach State Park in Madison and spending too much time trying to persuade a state park employee to jump my battery ( I was finally successful though it took an amazing amount of pleading), I decided to head inland to the Lyman Farm Orchards and pond in Middlefield to see and the photograph a Greater White-fronted Goose reported earlier in the week by Greg Hanisek. Well, my luck was holding. The White-fronted Goose was nowhere to be found. Just before leaving, I was scanning the flock of Canada Geese there and noticed a small billed goose amongst the flock. Thinking I had found the Richardson’s Cackling Goose, Branta hutchinsii hutchinsii, that had also been previously reported, I pulled over, set up my camera, and attempted to get photos. I was amazed at how difficult it was to keep track of this goose. Normally this smaller form sticks out from the larger Canada Geese. This bird, while smaller was not really small. Likewise, Richardson’s Geese often appear frostier in the field and stand out from the darker Canadas. Again, this bird did not. I noticed it was darker breasted as well. Typically, Richardson’s Cackling Goose is the palest breasted of the Cackling Goose forms. This bird also showed a white collar between the black neck stocking and the darker breast. While most of the forms of Cackling Goose can show this, I have never noted it on any Richardson’s Goose that I have seen. I kept studying the bird as I took photos ( I would end up with about 200 images of the bird!!!). I noticed that the head was shaped differently from what I have come to expect with Richardson’s Goose. Rather than appearing square-headed with a steep forehead angle where the stubby little bill meets the head, this birds profile was more rounded, the forehead seeming to continue from the slope of the bill. The neck seemed longer as well. The rounded head on this longer neck gave me a very different impression of this bird compared to the many Richardson’s Geese I have seen. As the light was fading I was getting ready to leave but noticed one more thing that seemed strange to me. I was getting glimpses of the underside of the birds head as it fed and I kept thinking I was seeing a smudgy looking dark stripe from the base of the bill to the back of the white patch in the throat. I attempted to get some images of this but the time and the light made it tough. This gular stripe is present in most Taverner’s Geese and some other forms of Cackling Geese as well. I have never noted one on a Richardson’s Cackling Goose.
On the way home I started thinking about what else this bird could have been other than a weird Richardson’s Cackling Goose. I remembered somewhere in the dusty old archives of my memory that Taverner’s Goose, which I know primarily from its appearance in the British pages at Sufbirds.com, shows a sloping head shape. Not having the proper references with me, I made good use of my hands-free cell phone adaptor and called Nick Bonomo as he and I had recently been discussing this form of goose. I wanted to run my observation by him and see what he thought. After our conversation and after returning home to check references, I was more sure of my ID as Taverner’s Goose. I sent the images to three West Coast birders and experts on the Cackling Goose complex, I asked Steve Mlodinow, Bruce Deuel, and Larry Semo to comment on my images and written description. These guys bird the West Coast and Rocky Mountain West and have opportunity to see all the forms of Cackling Geese and all the smaller forms of Canada Geese ( this will be important to a future Birding Geek note ). In fact, I had been collaborating with Steve Mlodinow by supplying images for an upcoming article he is doing on the identification of the forms of Cackling Geese. All of them responded that they were sure that this bird was, indeed, Branta hutchinsii taverneri, or Taverner’s Cackling Goose. All of their analysis cited the bill and head shape, the size and neck length, plumage coloration, the darker breast with collar, and the dark gular stripe as being very supportive of this identification. Richardson’s Cackling Goose was eliminated in the same ways and Lesser Canada Goose, B.c. parvipes was similarly eliminated ( again, pay attention for a future Bird Geek Diary!).

Interestingly, earlier, in October 2007, a Taverner’s Cackling Goose was identified and photographed in Amhurst, MA by James P. Smith. This bird appears very similar to the Connecticut bird and questions can be raised if in fact it is the same individual. To my knowledge, this is the first documented record of this subspecies for Connecticut, possibly the first record of any sort.

I am quoting from The Birds of North America on-line account for this subspecies:

“B. c. taverneri Delacour, 1951; type locality Colusa, CA. Taverner’s (or Alaska) Goose. Breeds in low tundra vegetation along shorelines of major rivers and small braided streams, shorelines of small tundra ponds, and on islands in tundra ponds and lakes of Yukon-Kuskokwim Delta, AK; breeding on the delta restricted to inland areas beyond the influence of tides, and extending inland to the north slope (Johnson et al. 1979, Jarvis and Bromley 1998). ..
…Winters primarily in Willamette River valley of w. Oregon, lower Columbia River valley of n.-central Oregon and s.-central Washington, and n. California (Johnson et al. 1979, Gilligan et al. 1994, Jarvis and Bromley 1998, CRE). “
This account includes the Cackling Geese as tundra-form Canada Geese. Branta canadensis taverneri is now considered Branta hutchinsii taverneri in a recent split by the American Ornithologist’s Union.(A.O.U.)

I want to thank Nick Bonomo for answering his cell-phone and wondering along with me, Steve Mlodinow, Larry Semo, and Bruce Deuel for their work in sorting out this very tough ID question, and that unidentified park employee for finally agreeing that jumping my battery was not a crime.

Literature cited:

The Birds of North America Online (A. Poole, Ed.). Ithaca: Cornell Laboratory of Ornithology; Retrieved from The Birds of North America Online database: http://bna.birds.cornell.edu/BNA/; AUG 2005.