C_seg: The classical 'C-score' (C_seg) seggregation metric of Stone...

In petrkeil/spasm: Species association measures for various data types

Description

The scaled version divides the raw C-score by the number of possible site combinations, which is 'n(n-1)/2' according to Ulrich & Gotelli (2013). Also note that their scaling formula also includes 'm', the number of species. Thus, Ulrich & Gotelli's (2013) formula for the scaled C-score is actually '4*(C_seg)/(m*n*(m-1)(n-1))'. For two species ('m=2') this reduces to 'C_seg / ((n*(n-1))/2)', which is the formula that I use here for the scaled version. Note: Before the calculation, species (rows) with zero occurrences, and sites (columns) with zero species are removed from m.

Usage

C_seg(m, scale = TRUE, clean = TRUE)

Arguments

m	Community data matrix with species as rows and sites as columns. Can contain either incidences (1/0) or abundances (natural numbers).
scale	Should the raw metric be scaled using the total number of possible site combinations?
clean	Logical. Do you want to remove zero-sum rows or columns? When the index needs the "d" fraction (sites where no species occurs), only zero-sum rows are removed.

Value

A dist object with the pairwise association values.

References

Stone L. & Roberts A. (1990) The checkerboard score and species distributions. Oecologia 85: 74-79.

McNickle G.G. et al. (2018) Checkerboard score-area relationships reveal spatial scales of plant community structure. Oikos 127: 415-426.

Ulrich W. & Gotelli N.J. (2013) Pattern detection in null model analysis. Oikos 122: 2-18.

petrkeil/spasm documentation built on Jan. 15, 2021, 6:08 p.m.