calc_evoregions: Defining biogeographic regions based on phylogenetic turnover

In GabrielNakamura/Rrodotus: Tools for Historical Biogeography analysis

Defining biogeographic regions based on phylogenetic turnover

Description

Defining biogeographic regions based on phylogenetic turnover

Usage

calc_evoregions(
  comm,
  phy,
  max.n.clust = NULL,
  method.dist = "bray",
  tresh.dist = 0.05,
  method.clust = "kmeans",
  stat.clust = "BIC",
  n.iter.clust = 1e+07,
  criterion.clust = "diffNgroup",
  seed = NULL
)

Arguments

comm	Species occurrence matrix. Assemblages are rows and species are columns
phy	phylogenetic tree of class phylo containing the species in comm.
max.n.clust	Integer value to be used in find.clusters. Indicates the maximum number of clusters to be tried.
method.dist	Character. The method to be used to compute phyogenetic distances among assemblages. Dissimilarity index, as accepted by vegdist (Default "bray").
tresh.dist	A scalar informing the threshold value to select eigenvectors based on the amount of variation of each eigenvector. Default is 0.05 (eigenvector with at least 5% of variation). See details.
method.clust	Character indicating the grouping algorithm to be used in cluster analysis. Options avalible are "kmeans" (default) or "ward".
stat.clust	Character to be used in find.clusters indicating the statistic to be computed for each model. Can be "BIC" (default), "AIC" or "WSS".
n.iter.clust	Integer to be used in find.clusters function of adegenet package to indicate the number of iterations to be used in each run of K-means algorithm
criterion.clust	Character string matching "diffNgroup" (default), "min", "goesup", "smoothNgoesup", or "goodfit", indicating the criterion for automatic selection of the optimal number of clusters. See criterion argument in find.clusters for an explanation of these procedures.
seed	default NULL. Set a seed to the find.clusters, which provides the same names for cluster groups

Details

evoregions performs biogeographical regionalization analysis, differently from other methods, evoregion uses a phylogenetic turnover metric based on fuzzy sets, therefore accounting for characteristics of evolutionary history, e.g tree imbalance, that is not accounted by other metrics of phylogenetic turnover.

Value

A list of length four containing:

• PCPS A matrix with PCPS vectors

• cluster_evoregions A vector indicating the region in which each assemblage was classified

See Also

find_max_nclust to decide the maximum number of clusters to be used

Examples

## Not run: 
data(akodon_sites) # occurrence data 
data(akodon_newick) # phylogenetic tree

akodon_pa <- akodon_sites %>% 
    dplyr::select(-LONG, -LAT)
    
spp_in_tree <- names(akodon_pa) %in% akodon_newick$tip.label
akodon_pa_tree <- akodon_pa[, spp_in_tree]

regions <- calc_evoregions(comm = akodon_pa_tree, phy = akodon_newick)
site_region <- regions$cluster_evoregions # classification of each community in regions

## End(Not run)

GabrielNakamura/Rrodotus documentation built on Aug. 31, 2023, 2:13 p.m.