Sampling Methods for Aerial Censuses of Wildlife Populations

No Thumbnail Available
Jolly G. M
Journal Title
Journal ISSN
Volume Title
While sampling of one kind or another has commonly been used for ground counts, usually because of the practical difficulty of covering larger areas, counts from the air have usually been based on a complete census. Among aerial surveyors who have used sampling methods, "transect sampling" has been favored. In this a straight flight path is followed and counts made along a transect of known width on either one side or both sides of the aircraft. Siniff and Skoog (1964) draw attention to' the possible Sources of error in transect -sampling, and for their own investigation of caribou used stratified random sampling, selecting at random from each of six 'strata varying numbers of sampling units; units were each four square miles and those sampled were completely covered by adjacent transects. Petrides (1953) Db served six sample strips, while Schultz and Muncy (1957) selected three sample lines (not strictly at random) on each of four areas of interest. Banfield et al.(1955) laid out their flight tracks as a systematic grid to' give even coverage. Bergerud (1963)also used regularly spaced transects, sometimes covering less than 1/ 40th of the survey area; he notes the non-validity of sampling errors calculated from systematic samples.
East African Agricultural And Forestry Journal, XXIV (SPECIAL), p. 46-49