First, for wanting to bike the dikes at Pt. Mouillee SGA on Dec 28th. Second, for deciding to spend 6 hours scanning a video to count 50,000 Canvasbacks on Lake Erie...
The recent thaw and re-freeze meant that the dikes at Pt. Mouillee would be free of snow and would be rideable. I was hoping to refind some of the Short-eared Owls that we've seen the past few days, so I loaded up the bike and headed to the Mouillee Creek entrance. It was chilly (30F), but winds were practically nil, so biking was quite pleasant. Unfortunately, there was little to no activity going on inside the Units.
The heavy construction trucks continue to haul gravel for the dike that is being built in Cell 5, so I headed along the Banana Unit toward the north end of Cell 5 toward Lake Erie. Cells 4 and 5 were now mostly open from recent rains, and Lake Erie was open.
The raft was too far away to do a cluster count. I decided that I could at least grab a video sweep of the lake and try to estimate from the video. After all, most of the ducks were just tiny black dots in the distance. Binoculars would not help. So, I decided I needed to take a digiscoped video at 40X magnification. Problem was, I still couldn't see the front of the duck raft!
I continued walking until I was almost at the junction of the North Causeway and the Banana Unit. From there, I could scope into the Huron River near Riverside Park, and sweep outward into the lake.
It took me 10 minutes to record first the Tundra/Mute Swans, then Mallard/Black Duck flocks, then Lesser Scaup / Redhead flocks, then finally the Canvasbacks.
No sooner had I just completed taking the video that I heard off in the distant south a chopper. I turned to see a large green military helicopter heading northeast from across the Units in Pt. Mouillee. Suddenly, I could see Lake Erie beginning to boil as thousands of the Canvasbacks began to take to the air. I put the scope on the distant flock and tried to capture some the birds taking off. It was pretty cool to watch.
I would then make the insane decision to go home and try to count all of the ducks / swans that I recorded in the 10-minute video. Placing a sheet of acetate over the computer monitor, I used a marker to dot-out all of the birds I was counting in single fields-of-view (FOV). The blurriness of the video in some places made things difficult, but after awhile you begin to know what is a duck and what isn't. Mute Swans and Tundra Swans tended to stay in separate flocks, so I counted swans and identified them based on the visible-colored bills that were showing among a group (I never saw mixed Tundra/Mute Swans in the video). Mallard/Black Ducks were a bit more challenging: I counted all ducks as Mallard except for the obvious Black Ducks, and where there were clusters of Black Ducks only, I counted all ducks as Black. I believe the numbers are pretty accurate. Redhead and Scaup (I'm presuming Lesser based on the relatively few birds I could identify) tended to stay on the outskirts of the Canvasback raft, and were found closer-in than out. Common Goldeneye and Bufflehead tended to stay on the outsides of the raft, as well.
It would take better than 6 hours to scan the video FOV-by-FOV, which meant advancing frame-by-frame (at 4K) until a single duck passed across the screen from right edge to left edge before counting all of the ducks in the FOV. Luckily, the Canvasback tended to be relatively spaced uniformly, so I could count 100 birds, then use my thumb-finger spacing to estimate the rest of the birds in the frame.
Fields-of-view grew from clusters of 100 (111 + 118 + 107...) to clusters of 500 (450 + 500 + 500 + 550) to as many as 1200 in a single FOV (here I would have to break the raft into upper, middle and lower thirds, with numbers of 750 + 325 + 250 being the norm). Numbers in FOV then gradually decreased into 500's, then 300's, then 100's before the last FOV numbered 80 birds. Incredibly, when I added up all of the numbers of Canvasbacks I got 49,631 birds - I should've just stayed w/ my original 50,000 estimate.
Interestingly enough, I conducted a similar estimation game a few years ago w/ a large raft of Canvasbacks on the Detroit River. I asked folks to estimate the number of ducks on the river in a ¼-mile stretch and found that 95% of guesses were 90% or better accurate when an estimate was made in less than 30 seconds!
If I do this again, I'll try to use the Nikon D500 and 300/2.8 VRII to take a video. It'll be much clearer and sharper.
Pte. Mouillee SGA, Monroe, Michigan, US
Dec 28, 2016 9:00 AM - 11:30 AM
Comments: Clearing skies, 30F made for a nice bike ride through the SGA; few birds to be seen inland; another story on Lake Erie, where a ½-mile long raft of Canvasbacks number 50,000 birds; Digiscoped video to show size (see https://youtu.be/_yWHje-kLhc and blog post describing in detail method for counting - http://jerryjourdan.blogspot.
Canada Goose (Branta canadensis) 12
Mute Swan (Cygnus olor) 547 See video: https://youtu.be/_yWHje-kLhc birds were seen on Huron River at mouth of Lake Erie and stretched for hundreds of yards; separate from Tundra Swans that were also present.
Tundra Swan (Cygnus columbianus) 788
Gadwall (Anas strepera) 10
American Black Duck (Anas rubripes) 797 Mixed among Mallard; see video: https://youtu.be/_yWHje-kLhc hand-counted from video frame-by-frame; number may be low, as mostly black males were counted and a few females added;
Mallard (Anas platyrhynchos) 2755 Mixed w/ Black Ducks along ice in mouth of Huron River mouth; see video: https://youtu.be/_yWHje-kLhc hand-counted from video frame-by-frame
Canvasback (Aythya valisineria) 49631 A raft of 50,000 birds at the mouth of Huron River was exclusively Canvasbacks. Took video through scope (see https://youtu.be/_yWHje-kLhc); count made based on field-of-view by field-of-view counting; took 6 hours to scan 10 minute video frame-by-frame.
Redhead (Aythya americana) 361
Ring-necked Duck (Aythya collaris) 2
Greater Scaup (Aythya marila) 2
Lesser Scaup (Aythya affinis) 10637 A raft of Lesser Scaup was separate from the Canvasbacks was located at the south end of Pt. Mouillee; its size was approximately 20% of the Canvasback flock, so estimate of 10,000 was made; the remaining birds were counted in the Canvasback video;
Bufflehead (Bucephala albeola) 234
Common Goldeneye (Bucephala clangula) 134
Hooded Merganser (Lophodytes cucullatus) 2
Common Merganser (Mergus merganser) 2
Pied-billed Grebe (Podilymbus podiceps) 1
Great Blue Heron (Ardea herodias) 1
Northern Harrier (Circus cyaneus) 3
Bald Eagle (Haliaeetus leucocephalus) 8
Red-tailed Hawk (Buteo jamaicensis) 2
Ring-billed Gull (Larus delawarensis) 573
Herring Gull (Larus argentatus) 60
Great Black-backed Gull (Larus marinus) 1
Downy Woodpecker (Picoides pubescens) 1
American Kestrel (Falco sparverius) 1
Horned Lark (Eremophila alpestris) 2
American Tree Sparrow (Spizelloides arborea) 35
American Goldfinch (Spinus tristis) 12
View this checklist online at http://ebird.org/ebird/view/
This report was generated automatically by eBird v3 (http://ebird.org)