摘要
样线法是在大范围内估计野生动物种群密度的优良方法之一。我们在本文中说明了使用以平均垂直距离作为有效样带宽度来计算种群密度的简单数学公式一般不会得到可靠的结果。即使野外实施方式正确 ,使用该数学计算式估计出的密度结果也往往过于偏高。因此 ,我们建议最好使用计算机软件DISTANCE调查野生动物种群密度 ,并能熟悉样线距离取样法的理论基础。
Line transects are one of the best ways to estimate density of wildlife populations over large areas. However, density estimates will be unreliable if using mathematical procedures that, although simple and easy to use, do not correspond with reality. We provide theoretical and empirical evidence that using a simple “mean distances” approach (e.g., Sheng et al., 1992; Gao et al. , 1997) in which the mean of observed perpendicular distances is equated with the effective strip width (i.e., D=ns/2LW, in which D =estimated density of animals, n =number of animals seen, s =mean group size, L =length of transect line, and W =mean perpendicular distance of animals seen) is unlikely to yield reliable results. This “mean distances” approach will be approximately true only when the true (underlying) detection probability follows a negative exponential distribution (and even then, does not allow calculation of variance). However, if the half normal detection function is true, this “mean distances” approach can be expected to be positively biased by 57%. In empirical tests of line transect estimators, the “mean distances” approach overestimated true density by 41% to 81%. Alternative mathematical formulations to this “mean distances” approach, incorporated into program DISTANCE, are much less likely to be seriously biased. Using these alternative approaches also forces investigators to consider seriously issues of sample size (whereas the “mean distances” approach will produce an estimate even when n =1, in which case the researcher really has no data on how detection varies with distance). We urge researchers to familiarize themselves with the line transect theory (Buckland et al., 1993), and to use program DISTANCE. Equally importantly, we urge researchers to minimize assumption violations of distance sampling, and to follow rigorous random sampling protocols.
出处
《动物学报》
SCIE
CAS
CSCD
北大核心
2002年第6期812-818,共7页
ACTA ZOOLOGICA SINICA
基金
RobertM .Lee基金会
刘国烈基金会资助项目~~
