You are not currently logged in.
Access your personal account or get JSTOR access through your library or other institution:
If You Use a Screen ReaderThis content is available through Read Online (Free) program, which relies on page scans. Since scans are not currently available to screen readers, please contact JSTOR User Support for access. We'll provide a PDF copy for your screen reader.
Sampling error in measuring temporal density variability in animal populations and communities
Mikko Mönkkönen and Jouni Aspi
Annales Zoologici Fennici
Vol. 35, No. 1 (1998), pp. 47-57
Published by: Finnish Zoological and Botanical Publishing Board
Stable URL: http://www.jstor.org/stable/23735586
Page Count: 11
Since scans are not currently available to screen readers, please contact JSTOR User Support for access. We'll provide a PDF copy for your screen reader.
Preview not available
Variation in animal numbers is an interesting ecological problem in both theoretical and applied research. Recent research has shown that there is a myriad of problems involved in measuring variability in animal populations, which have not been addressed in most empirical studies on fluctuations of animal densities. Therefore, we actually know less about variability in animal populations and communities than we think. It is rarely possible to accurately sample entire populations and our estimates of variability usually come from spatially restricted samples of counts drawn from local populations. The observed variability not only reflects variability in population density or size but also involves a sampling variance component. Sampling variance occurs principally due to inexactness of the counts (i.e. all individuals present in the sampling unit do not enter into samples) and spatial variance (the size of the sampling unit is inadequate to capture the dispersion pattern of individuals in the field). Many samples are affected by both of these sources of error and in most cases we are unable to separate their effects. Sampling variance usually affects the variability estimates and particularly besets small samples. When comparisons are to be made in temporal variability between communities, species, populations or sites, great care must be taken to mitigate the effects of sampling variance. If the counts are replicated in space or time then sampling error can fairly simply be estimated and removed. Even in the absence of replication, statistical methods exist allowing estimation of the sampling variance. These methods are only applicable if we are prepared to make assumptions about the distributions of the counts. We exemplify one of these methods by considering a classical case of latitudinal gradients in density variability in animal communities. We finally discuss recent studies, the results of which might be artefacts arising from sampling variance.
Annales Zoologici Fennici © 1998 Finnish Zoological and Botanical Publishing Board