JPNyers 1976 Journal 53 GRIDS 2-3, Bairou, Alaska 25 August 0745-1055 censused both grids. 39°, 5% cloudy, light ENE wind growing through morning. GRID TOTALS (each 25 ha) #2 Albulis dominica a j 9 C. alpina a j 14 Ph. fulicarius a j 0 Stercorarius pom. a j 1 Calcavis a j 2 Plectrophenax a j 1 #3 Albulis dominica a j 0 C. alpina a j 2 C. melanochor a j 0 Ph. fulicarius a j 0 Stercorarius pom. a j 1 Asio flammeus a j 1 Calcavis a j 3 Obviously, a change has occurred since the last census, and indeed perhaps since yesterday. There were relatively few C. alpina as compared to grid censuses throughout late July and August. Petrels were virtually absent on grid 2—in fact I saw none even flying until beginning #3, and they were not very evident even there. Dovekets take the show, with 41 occurring in one flock, spread over 2 units in grid 2. There were a number of flocks flying about, and I may have under estimated the total. Certainly it is not an overestimate, probably on either grid. This morning is the first fogless morning (and night) for 5 days. Did the birds leave with the clear night skies? GRID 4 Painted more stakes this pm. Of interest was that dunlins have not disappeared from other areas. My impression is that dunlins are dispersed in a series of discrete flocks which remain in local sites. This morning's data reflects that the flock using the grids have largely gone. But this afternoon I found the usual group present between the goppie and Grid 4. They seemed while running transects that there are consistent birds there, as well, birds that use a local area even though no longer nesting. While this is nothing astonishing in itself, the length of time that flocks remain—perhaps several weeks—surprises me