Basic data from wildfowl counts

Abstract

It was desired to test the assumptions that errors due to the spacing of the counts, to the mobile nature of the birds and to the estimation of numbers will cancel out. The detailed results of the study are being published in the Proceedings of the 12th International Ornithological Congress 1958. The mean value of near daily counts gives a good measure of the population level within the month. The count made on the Count Date was found to err, on the average, by between a quarter and a half of this true monthly mean value. Some species are more likely to deviate in this way, due to more rapid fluctuations in their population. This instability is measured by a statistic (V), the coefficient of variation. The assumption that movements of birds between waters would cancel out was tested by comparing the values of V obtained from two waters considered first as separate units and then as one unit. A number of phenomena, particularly weather, affect stability but, being largely unpredictable, cannot be allowed for in planning.