Bias in Aerial Survey

摘要: Aerial censuses of large mammals are inaccurate because the observer misses a significant number animals on transect. The accuracy deteriorates progressively with increasing width transect, cruising speed, and altitude. Methods eliminating bias by refining techniques discussed rejected; there seems to be no technical solution. An alternative strategy is measure correct estimates accordingly. A method suggested for estimating during an aerial census, subsequent analysis returning unbiased estimate density. No direct true density needed little extra effort involved over that required standard survey. J. WILDL. MANAGE. 38(4):927-933 This paper examines effect visibility bias, discusses means which arises, suggests methods it might eliminated from survey population size. is, at best, rough size population. Most efforts refinement have been aimed raising precision combining impeccable design, high sampling intensity, intricate stratification, powerful analysis. trend can traced back Siniff Skoog's (1964) superb census caribou (Rangifer tarandus). Their use stratified random sampling, allocated proportional density, contrasted markedly crudity previously reported surveys. Subsequently, Jolly's (1969a) designs analyses appropriate has encouraged rigorous disciplined application method. Recent papers following this lead tended treat difficulties air largely as constituting problem, being rated successful or otherwise according estimate's error. Tacitly, error was treated rather than its repeatability. Underlying preoccupation often lurked implicit assumption free observers counted all each sampled