Gongal'skiĭ K B, Zaĭtsev A S, Savin F A
Zh Obshch Biol. 2009 Nov-Dec;70(6):484-94.
Spatial distribution is one of the main parameters of populations of soil animals. Spatial soil ecology having been developing during last decades bases animal distribution estimates on the geostatistic approach. A simple principle underlying the latter's methodology is that samples placed close to each other have more similarity than those distantly placed, it is usually called autocorrelation. The principles of basic statistics cannot be applied to autocorrelated data. Apiplying variograms, Mantel test, Moran index, and SADIE statistics enables to reveal the size of clusters of both soil parameters and soil animal aggregations. This direction of investigations quite popular in the western literature is just rarely employed by Russian soil ecologists. Statistically correct procedures allow developing field sampling methodology that is vital in applied studies of soil ecology, namely, in bioindication and ecotoxicology of soils, in the assessment of biological resources in terms of abundance and biomass of soil animals. This methodology has a decisive importance in the development of soil biogeography.
空间分布是土壤动物种群的主要参数之一。在过去几十年中不断发展的空间土壤生态学，基于地统计学方法对动物分布进行估算。该方法背后的一个简单原则是，彼此靠近放置的样本比间隔较远放置的样本具有更高的相似性，这通常被称为自相关性。基本统计学原理不能应用于自相关数据。应用变异函数、曼特尔检验、莫兰指数和SADIE统计方法，能够揭示土壤参数和土壤动物聚集的集群规模。在西方文献中颇受欢迎的这一研究方向，俄罗斯土壤生态学家却很少采用。统计上正确的程序有助于开发实地采样方法，这在土壤生态学的应用研究中至关重要，即在土壤生物指示和生态毒理学、根据土壤动物的丰度和生物量评估生物资源方面。这种方法在土壤生物地理学的发展中具有决定性意义。
