fpcomSims: Functional-phylogenetic community simulations
In funphylocom: Functional traits, phylogenies, communities, simulations

Description Usage Arguments Details Value

Simulations for assessing the relative importance of phylogenetic versus functional information for understanding variation in community composition.

fpcomSims(n, m, p = 0.5, diverg.obs = rep(0, m), diverg.unk = rep(0, m),
  sim.envObserved = as.vector(scale(rnorm(n))),
  sim.envUnknown = as.vector(scale(rnorm(n))), site.names = numnames(n,
  "site"), spp.names = numnames(m, "sp"))

`n`	Number of sites to simulate.
`m`	Number of species to simulate.
`p`	Number between 0 and 1 giving the relative importance of observed versus unknown traits in determining community structure.
`diverg.obs`	A vector with m elements giving the divergence effects on the observed trait for each species.
`diverg.unk`	A vector with m elements giving the divergence effects on the unknown trait for each species.
`sim.envObserved`	A numeric vector of simulated values for the observed environmental variables.
`sim.envUnknown`	A numeric vector of simulated values for the unknown environmental variables.
`site.names`	Character vector of site names.
`spp.names`	Character vector of species names.

Simulations are based on the following procedure:

Phylogeny: A tree is simulated using rcoal with default settings.
Trait evolution: Two traits are simulated under Brownian motion. One trait is called 'observed' and the other 'unknown'. Both traits influence species' probabilies of occurrence, but only the 'observed' trait is included in the output. The idea behind this distinction is that information about the 'unknown' trait will possibly be present in the phylogeny, which is assumed known.
Probabilities of occurrence: For each species at each site, these probabilities depend logistically on two gradients; one is observed and the other is unknown, each corresponding to the observed and unknown traits (see the sim.envObserved and sim.envUnknown arguments). The corresponding traits are the logit-scale slopes of these logistic curves – in other words each species depends on the gradient differently depending on their traits. When p equals one (zero), only the observed (unknown) gradient and trait determine probability of occurrence. When p is between zero and one, both the observed and unknown trait-gradient combinations have some influence.
Occurrence: Each species is present at each site with probability given by the corresponding probability of occurrence.

An object of class fpcomSims with components:

`comm`	An n-by-m matrix of presences and absences
`probs`	An n-by-m matrix of probabilities of occurrence
`traits`	A length-m vector of values for the observed trait
`env`	A length-n vector of values for the observed gradient
`tree`	A phylo object with the phylogenetic tree relating the m species