PRE_FATE.speciesClustering_step2: Choose clusters and select determinant species
In MayaGueguen/RFate: Vegetation modelling with the FATE model

Description Usage Arguments Details Value Author(s) See Also Examples

View source: R/PRE_FATE.speciesClustering_step2.R

This script is designed to obtain functional groups by : 1) selecting the number of clusters to be kept from an object obtained with the PRE_FATE.speciesClustering_step1 function ; 2) refining these groups by identifying determinant species in each of them.

PRE_FATE.speciesClustering_step2(
  clust.dendrograms,
  no.clusters,
  mat.species.DIST
)

`clust.dendrograms`	a dendrogram, or a `list` of dendrograms (one for each `GROUP` value). Such an object can be obtained with the `PRE_FATE.speciesClustering_step1` function.
`no.clusters`	an `integer`, or a `vector` of `integer` (one for each `GROUP` value), with the number of clusters to be kept
`mat.species.DIST`	a `dist` object, or a `list` of `dist` objects (one for each `GROUP` value), corresponding to the distance between each pair of species. Such an object can be obtained with the `PRE_FATE.speciesDistance` function.

This function allows to obtain a classification of dominant species into Plant Functional Groups (PFG), and the determinant species based on these PFG.

What is the difference between dominant and determinant species ?

Dominant species are species representative of an environment or a studied area, in terms of number of releves or abundance values. They can be found with the PRE_FATE.selectDominant function of this package. These dominant species are used to build PFG with the PRE_FATE.speciesClustering_step1 function.
Once PFG are built, determinant species are defined as refined subsets of dominant species within each PFG.
The process is detailed below :
- each dominant species is assigned to a PFG
- within each PFG :
  - for each species, compute its mean distance to the other species within the PFG (sp.mean.dist)
  - calculate the mean value of all these mean distances (allSp.mean)
  - calculate the deviation values around this mean value (allSp.min and allSp.max)
  - determinant species are the ones that are included between those deviation values

A list containing one vector, one data.frame with the following columns, and two ggplot2 objects :

determ.sp

the names of all determinant species

determ.all

(determinant and non-determinant species)

PFG: ID of the plant functional group (GROUP + ID.cluster)
GROUP: name of data subset
ID.cluster: cluster number
species: name of species
ID.species: species number in each PFG
sp.mean.dist: species mean distance to other species of the same PFG
allSp.mean: mean(\text{sp.mean.dist}) within the PFG
allSp.min: mean(\text{sp.mean.dist}) - 1.64 * sd(\text{sp.mean.dist}) within the PFG
allSp.max: mean(\text{sp.mean.dist}) + 1.64 * sd(\text{sp.mean.dist}) within the PFG
DETERMINANT: TRUE if determinant species, FALSE otherwise

plot.distance

ggplot2 object, representing the distribution of mean distances between species for each functional group

plot.PCO

list of ggplot2 objects, representing the PFG within the functional space

One ‘PRE_FATE_CLUSTERING_STEP_2_distantSpecies_PCO.pdf’ file is created containing two types of graphics :

distantSpecies: to visualize in each PFG the distribution of mean distance of each species to other species, and non-determinant species which are outside the distribution
PCO: to visualize in each PFG the distribution of species, with and without non-determinant species

Maya Guéguen

cutree, quasieuclid, dudi.pco, PRE_FATE.speciesDistance, PRE_FATE.speciesClustering_step1, PRE_FATE.speciesClustering_step3

## Load example data
data(DATASET_Bauges_PFG)

## Species dissimilarity distance (niche overlap + traits distance)
tab.dist = DATASET_Bauges_PFG$dom.dist_total
str(tab.dist)
as.matrix(tab.dist[[1]])[1:5, 1:5]


## Build dendrograms -------------------------------------------------------------------------
sp.CLUST = PRE_FATE.speciesClustering_step1(mat.species.DIST = tab.dist)
names(sp.CLUST)


## Number of clusters per group
plot(sp.CLUST$plot.clustNo)
no.clusters = c(5, 5, 9)

## Find determinant species ------------------------------------------------------------------
sp.DETERM = PRE_FATE.speciesClustering_step2(clust.dendrograms = sp.CLUST$clust.dendrograms
                                             , no.clusters = no.clusters
                                             , mat.species.DIST = tab.dist)
names(sp.DETERM)

str(sp.DETERM$determ.sp)
str(sp.DETERM$determ.all)

## Species names
sp.NAMES = DATASET_Bauges_PFG$sp.names
sp.NAMES$species = paste0("X", sp.NAMES$species)
determ = merge(sp.DETERM$determ.all, sp.NAMES, by = "species", all.x = TRUE)
str(determ)

plot(sp.DETERM$plot.distance)
plot(sp.DETERM$plot.PCO$Chamaephyte)
plot(sp.DETERM$plot.PCO$Herbaceous)
plot(sp.DETERM$plot.PCO$Phanerophyte)