multiconstrained {BiodiversityR}

Pairwise Comparisons for All Levels of a Categorical Variable by RDA, CCA or Capscale

Description

This function implements pairwise comparisons for categorical variable through capscale, cca, dbrda or rda followed by anova.cca. The function simply repeats constrained ordination analysis by selecting subsets of data that correspond to two factor levels.

Usage

multiconstrained(method="capscale", formula, data, distance = "bray"
    , comm = NULL, add = FALSE, multicomp="", contrast=0, ...)

Arguments

method

Method for constrained ordination analysis; one of "rda", "cca", "dbrda" or "capscale".

formula

Model formula as in capscale, cca or rda. The LHS can be a community data matrix or a distance matrix for capscale.

data

Data frame containing the variables on the right hand side of the model formula as in capscale, cca or rda.

distance

Dissimilarity (or distance) index in vegdist used if the LHS of the formula is a data frame instead of dissimilarity matrix; used only with function vegdist and partial match to "manhattan", "euclidean", "canberra", "bray", "kulczynski", "jaccard", "gower", "morisita", "horn" or "mountford". This argument is only used for capscale in case that the LHS of the formula is a community matrix.

comm

Community data frame which will be used for finding species scores when the LHS of the formula was a dissimilarity matrix as only allowed for capscale. This is not used if the LHS is a data frame.

add

Logical indicating if an additive constant should be computed, and added to the non-diagonal dissimilarities such that all eigenvalues are non-negative in underlying Principal Co-ordinates Analysis; only applicable in capscale.

multicomp

Categorical variable used to construct the contrasts from. In case that this variable is missing, then the first explanatory variable of the formula will be used.

contrast

Return the ordination results for the particular contrast indicated by this number (e.g. with 5 levels, one can choose in between contrast 1-10). In case=0, then the first row of the anova.cca results for all contrasts is provided.

...

Other parameters passed to anova.cca.

Details

This function provides a simple expansion of capscale, cca and rda by conducting the analysis for subsets of the community and environmental datasets that only contain two levels of a categoricl variable.

When the choice is made to return results from all contrasts (contrast=0), then the first row of the anova.cca tables for each contrast are provided. It is therefore possible to compare differences in results by modifying the "by" argument of this function (i.e. obtain the total of explained variance, the variance explained on the first axis or the variance explained by the variable alone).

When the choice is made to return results from a particular contrast (contrast>0), then the ordination result is returned and two new datasets ("newcommunity" and "newenvdata") are created that only contain data for the two selected contrasts.

Value

The function returns an ANOVA table that contains the first rows of the ANOVA tables obtained for all possible combinations of levels of the first variable. Alternatively, it returns an ordination result for the selected contrast and creates two new datasets ("newcommunity" and "newenvdata")

Author(s)

Roeland Kindt (World Agroforestry Centre)

References

Legendre, P. & Anderson, M.J. (1999). Distance-based redundancy analysis: testing multispecies responses in multifactorial ecological experiments. Ecological Monographs 69: 1-24.

Anderson, M.J. & Willis, T.J. (2003). Canonical analysis of principal coordinates: a useful method of constrained ordination for ecology. Ecology 84: 511-525.

Examples

## Not run: 
library(vegan)
library(MASS)
data(dune)
data(dune.env)
multiconstrained(method="capscale", dune~Management, data=dune.env,
    distance="bray",add=TRUE)
multiconstrained(method="capscale", dune~Management, data=dune.env, 
    distance="bray", add=TRUE, contrast=3)

## End(Not run)