makeVirtualSpecies: Generates virtual species

In BlasBenito/SDMworkshop: Streamlining the Development of Species Distribution Models

Description

Generates virtual species from virtual ecological niches. The niche definition can be either provided by the user, or generated randomly.

Usage

makeVirtualSpecies(
variables,
niche.parameters = NULL,
seed = NULL,
species.type = "multiplicative",
max.n = NULL
)

Arguments

variables	A raster brick or stack with three or more environmental variables.
niche.parameters	A named list containing the parameters of the gaussian functions representing the niche of the species. Each slot name must be a layer name of the brick/stack variables. The slot must contain a vector with two numbers. The first represents the mean of a normal distributión, and the second represents the standard deviation of the same normal distribution. These numbers must be in the same units of the variables layer with the same name of the slot. If this argument is set to NULL, then a random set of niche parameters is generated by the function based on an uncorrelated set of predictive variables.
seed	Integer determining a random seed. Only relevant when niche.parameters = NULL. Added to allow reproducibility in the generation of random virtual species.
species.type	String, one of "additive" or "multiplicative". This is an argument of the function generateSpFromFun, which is used internally. It defines how the overall environmental suitability is computed from the response functions of each predictor. These are summed when species.type = "additive", and multiplied when species.type = "multiplicative", which is the default behavior.
max.n	Numeric integer, maximum number of presence records to extract if possible, otherwise the function returns as many presence records as possible.

Details

This function relies on the functions generateSpFromFun, convertToPA ,sampleOccurrences of the virtualspecies package, so please, cite Leroy et al. (2016) if you find this function useful.

Note that for simplicity, the settings of the functions mentioned above have been simplified. Particularly, the function convertToPA is configured with the parameters PA.method = "probability", prob.method = "linear", a = 1, and b = 0. This setting is the safest one when niche.parameters is not provided, and therefore randomized virtual taxa are being generated.

The function sampleOccurrences is configured as follows: type = "presence only"; correct.by.suitability = TRUE.

Please, check the help files of these three functions to better understand how they work!

Value

A named list with 5 slots:

• niche.dimensions: Names of the variables used to define the niche of the virtual species.

• niche.parameters: Either the same niche.parameters used as input, or the niche definition generated randomly by the function if niche.parameters = NULL.

• niche.plot: A ggplot with the density plots of the presence records (represented in blue) and the background (represented in yellow), and a habitat suitability map with presence records.

• suitability.raster: A habitat suitability raster generated from the niche parameters by the function generateSpFromFun.

• observed.presence: A dataframe with the coordinates x and y (in the same coordinate reference system as variables) of the selected presences.

Author(s)

Blas Benito <[e-mail]>. The functions generateSpFromFun, convertToPA, and sampleOccurrences are authored by Boris Leroy <[e-mail]>, with help from C. N. Meynard, C. Bellard & F. Courchamp

References

Leroy, B., Meynard, C.N., Bellard, C. and Courchamp, F. (2016), virtualspecies, an R package to generate virtual species distributions. Ecography, 39: 599-607. doi:10.1111/ecog.01388

Examples

## Not run: 
data("europe2000")

#niche parameters list
niche.parameters <- list(
 bio12 = c(500, 250),
 bio5 = c(240, 50),
 bio6 = c(10, 30),
 human_footprint = c(0, 30),
 topo_slope = c(0, 2),
 landcover_veg_herb = c(100, 35)
)

#making the virtual species
vs <- makeVirtualSpecies(
 variables = europe2000,
 niche.parameters = NULL,
 max.n = 200,
 species.type = "multiplicative",
 seed = NULL
)

## End(Not run)

BlasBenito/SDMworkshop documentation built on March 4, 2020, 4:16 a.m.