A timesaving estimation of per‐quadrat species number in grassland communities based on a Poisson‐like model. Issue 2 (8th January 2018)

• Record Type: Journal Article
• Authors: Chen, Jun; Shiyomi, Masae; Bai, Huimin
• Is Part Of: Ecological research. Volume 33:Issue 2(2018)
• Page Start: 427
• Page End: 434
• Publication Date: 2018-01-08
• Subjects: Censored sample -- Labor‐saving -- Small‐scale -- Spatial pattern -- Species richness; Ecology -- Periodicals; Ecology -- Japan -- Periodicals; Écologie; Japon; Ecology; Japan; Ressource Internet (Descripteur de forme); Périodique électronique (Descripteur de forme); Periodicals; 577.05
• DOI: 10.1007/s11284-017-1544-2
• Languages: English
• ISSNs: 0912-3814

Abstract

Abstract: In estimating the number of species per quadrat with a given area, we usually need much time and labor, because we have no simple ways to easily count it using hardware. We devised here a software method to estimate the mean number of species by counting them only partially in the survey, and estimating it mathematically. We classify quadrats into three classes composed of k, k + 1 and more than k + 1 species; or classify quadrats into four classes composed of k, k + 1, k + 2, and more than k + 2 (referred to as "censored sampling"), where k is the minimum number of species per quadrat. We do not need to count more than k + 2 or k + 3 species, respectively. Using only these 3 or 4 numerals we can estimate the mean of number of species per quadrat, the 95% confidence intervals for the mean and know the spatial pattern value such as aggregated, random or uniform pattern of the number of species per quadrat. Using this method, 10–20% time and labor in counting will be saved compared to the full census. The estimates are obtained through the maximum likelihood calculation. Computer programs to estimate the mean number of species per quadrat are attached as ESMs.