Thinking on bean goose taxonomy certainly divides the authorities, there are currently one, two or three species, Tundra, Taiga and Middendorff's, depending on which view you subscribe to. Plus a pick-n-mix of as many as five forms, fabalis, johanseni, middendorffii, rossicus and serrirostris, though johanseni usually fails to get the nod being considered an intergrade population. But from the Japan perspective Middendorff's or Taiga Anser (fabalis) middendorffii and Tundra A. (serrirostris) serrirostris are the only two of real concern. That said, I understand a Yamashina Institute for Ornithology publication intriguingly states wintering Taiga Bean Geese in Shimane, not far to our west, are Taiga Bean Geese but not middendorffii! Is there more to Japan's bean geese than meets the eye? Would these Shimane birds be morphologically distinct from middendorffii elsewhere in Japan? Could these birds also occur at Lake Biwa?
The geese on Lake Biwa are invariably in the Kohoku area, along the north east shore of the lake or in the north west corner where a small group can sometimes be found. Just outside Kansai, Katano Kamoike (Ishikawa) to the north also hosts a wintering flock. These middendorffii flocks are very aquatic in habits and on Lake Biwa they feed by upending for tubers from the lake bottom, coming onto small islands to sleep. I never see them on fields next to the lakes. That large flocks of Greater White-fronts (with a very small number of Lessers) can be found grazing wheat fields in Fukui prefecture between these Shiga and Ishikawa middendorffii locations serves to underline how strictly dependant on bodies of water these geese are.
Because they are invariably on the water when active makes judging relative size more difficult compared to assessing birds on mainly flat arable land. Judging relative leg length is out of the question obviously, neck length/thickness is variable depending on whether the geese are stretching to the lake bottom or loafing, the feathers wet and flattened or not. So head shape with bill size and structure, which is in any case widely considered crucial in separating Taiga and Tundra, is the most consistently apparent feature when viewing this flock. It's no surprise to me that out-of-range individuals in Asia or North America come with identification issues because I see such variability in bill size and shape within "my" flock that it suggests either the Biwako flock is a mixed taxon group or bill structure is less reliable than might be hoped for.
I'm often in the Kohoku area late morning/early afternoon when the geese aren't very active, so visits later in the day recently enabled me to get slightly better than usual digiscoped shots. With the following images I'm trying to show how much variation there is within a flock considered to be Middendorff's Bean Goose. Most were taken on three dates in January and February this year (2014) with a few earlier images going back to 2009.
Not quite square on, but nevertheless a fairly good portrait of a bird amongst closer-than-usual geese. The bill, if folded back, is about 90% of the head length. It's very deep but there is a forehead step, the culmen is straight and it angles towards the elongated tip from a point just distal of the nostril. The cutting edge of the upper mandible is smoothly arched with the highest point about mid-way and no sharp turndown at the base. It has a conspicuous drooping lower mandible, most obvious close to the highest point of the arching upper, there is little of the lower mandible visible at the base (maybe because of a slight tilt to the head). The orange bill band is limited to a neat ring behind the nail. A nice portrait, but how typical is it?
The following images are of a group of birds actively feeding. Could this account for the very long, thin necks as they stretch to the bottom of the lake? Perhaps slicked down wet feathers produce the very thin appearance while also emphasising the bill size? Whether or no, the necks look too thin for the bulk of the body and these bills are grotesquely large and very deep, even bulbous, towards the tip, I almost expect to find an elastic band circling the head to hold them in place like a child's false nose, the real bill hidden beneath. Compared to the previous bird the bill looks longer and angles less steeply towards the tip.
These geese are very long-bodied, thin-necked with large, deep-tipped bills. Impressive, but ungainly.
If the difference between the above types is slight, a few geese differ more obviously. The following bird (three images) is the only one of the type I've seen at reasonably close range. It has a slightly convex culmen, the cutting edge is straighter with a more distinct downturn towards the base and a deeper base to the lower mandible. The short tip extension lacks depth but then expands noticeably on both mandibles producing a deep base to the lower, this continues into a jowly chin. Thus the whole head and bill looks a cohesive triangular unit. Compare the total bill depth at the base to the overall length, the grotesque-billed birds have a bill twice as long as the depth whereas the following bird's bill depth would reach to the nail. None of these geese are particularly small which might militate against Tundra even though the bill shape is radically at odds with birds that seem to be classic middendorffii.
The following two images, taken in early December 2012, seem to show three examples of the same triangular head and bill type. I noticed these three geese (in the foreground) swim about 300m to join the flock of well proportioned birds I'd been watching. They all have deep-based, slightly shorter bills that slope more directly into the fore-crown. At the time I put it down to individual variation despite three such birds together leaving niggling doubt. The other birds all have the same basic appearance except one or two of the huge-, bulbous-tipped bill types and a narrow-, black-billed very small goose at the back towards the right (far right in the second image). More of that later.
The following goose (two images) has the same basic short, deep bill and overall triangular head/bill appearance. The only subtle difference is deep-based lower mandible has a distinct kink towards the tip creating a drooped look.
Coming back to the small (largely) black-billed bird(s), I managed to get reasonable shots this year. Only 15 months after the images above, it could well be the same returning individual. It is small but proportionately not as short-bodied as a Tundra might look. The bill shape doesn't seem to be a scaled down version of the larger birds if this were just an exceptionally small individual, it's very duck-like tapering to a thin tip and it's narrow throughout its length and the cutting edge is straight from tip to base unlike any other goose here.
The final bird, present in January 2009, really does seem to be a Tundra Bean Goose and is quite different to any other goose I've seen at Biwako, though it also has less obvious orange in the bill. It was extremely compact sleeping, much shorter-bodied than the other geese. The head was chunky, almost square, with a steep forehead and stubby, small-based bill with a straight culmen and prominent grin line.