Design and Results of a Waterfowl Breeding Population Survey for Wyoming

Abstract

The waterfowl breeding population survey for Wyoming was redesigned to concentrate on areas of high waterfowl density and to resemble more closely the continental waterfowl survey in terms of design, procedures, and estimates of precision. The redesigned survey was conducted and evaluated in 1993 and in 1994. Ten strata in areas of high expected waterfowl density were sampled systematically with east-west aerial strip transects 400 m wide. To determine the effect of surveying areas of scarce waterfowl habitat, 6 transects in 1993 and 10 in 1994 were placed randomly outside the 10 strata of high expected waterfowl density. Including population estimates from this low density area increased state-wide mallard (Anas platyrhynchos) estimates by 110 and 81% in 1993 and 1994, and increased CVs of mallard estimates from 18.3 to 25.2% in 1993 and from 17.3 to 21.2% in 1994. Computation of optimal allocation indicated that, to achieve the greatest precision of state-wide estimates, sampling effort in the low density area should be increased almost threefold. In 1993, surveys of 2 strata in narrow intermountain river valleys had problems with maintaining flight altitude and counting waterfowl. In these strata, adherence to desired altitude was increased in 1994 by redefining transect boundaries and placing transects in a southwest-northeast orientation. Density of all ducks in one of the river valley strata increased from 2.75 to 6.91 ducks/km 2 from 1993 to 1994. To investigate use of 200-m-wide transects surveyed by only 1 observer, data from 1 side of each 400-m-wide transect were omitted. Attaining a CV of 10% for estimates of all waterfowl species with transects 200 m wide would require flying over twice as many kilometers as with transects 400 m wide. We do not recommend waterfowl surveys with 200-m-wide transects flown by only 1 observer.