This function performs the same calculations as 'indval' or 'indvals', but does so using matrix operations on a set of binary group assignments in matrix form. Function 'indvalps' calculates indicator value scores using a modification of Dufrene and Legendre (1997), whereby relative abundance is computed as total relative abundance across groups rather than as mean relative abundance.
1 2 3
A site-by-permutation matrix of binary (0,1) assignments indicating group membership for successive randomizations of rows of 'spp' according to a ranks of an environmental variable.
A site-by-taxon matrix of taxa counts at each sampling location.
Although the 'indval' function is reasonably efficient for a single calculation, when repeated during permutations (default = 250) and again during for each bootstrap replicate (default = 500), small differences in processing time can quickly become unwieldy. Even with the matrix operation, the permutation accounts for > 3/4 of processing time for most data sets (due to the fact that it is repeated for each bootstrap replicate).
The output matrix does not include information about the group membership of indval maxima because this information is not used in the permutation procedure (i.e., only the distribution of IndVal magnitudes is relevant).
Modification of the original IndVal method was developed to address potential for bias in change point identification for highly skewed samples. This and the function 'indvals' are run when the argument 'ivTot' in the original TITAN function call is TRUE. It is expected that 'indval' and 'indvalp' will be used more commonly.
A matrix (ivmax) of IndVal maxima across each partition with nrow equal to the number of permutations and ncol equal to the number of taxa in 'spp'.
M. Baker and R. King
Dufrene, M. and P. Legendre. 1997. Species assemblages and indicator species: the need for a flexible asymmetrical approach. Ecol. Mon. 67:345-366.
Baker, ME and RS King. 2010. A new method for detecting and interpreting biodiversity and ecological community thresholds. Methods in Ecology and Evolution 1(1): 25:37.