rangemod1d: Range Cohesion Model for Ordered (and Non-spatial) Data
In rangemodelR: Mid-Domain Effect and Species Richness

View source: R/rangemod1d.R

rangemod1d

R Documentation

Range Cohesion Model for Ordered (and Non-spatial) Data

Description

The function takes observed site by species matrix and returns expected species richness values of each site

Usage

rangemod1d(
  spmat,
  var = NULL,
  cohesion = T,
  first = FALSE,
  degen = FALSE,
  rsize = c("observed", "unif"),
  reps
)

Arguments

`spmat`	a site by species matrix or data frame with species in columns
`var`	an optional vector containing explanatory variable for constraining the randomization. It should be NULL when absent
`cohesion`	If true, species distributions are without gaps i.e. result is range cohesion, otherwise it is range scatter
`first`	If TRUE, 'var' is used while choosing the first occurrence as well.if 'var' is null, first is always set 'FALSE'
`degen`	If true, each randomized site by species matrix is saved and provided in output
`rsize`	which range sizes to use for simulation, can be an integer vector of same length as number of species(columns) or either 'observed' or 'unif'. See details for explanations
`reps`	number of replicates

Details

Implements simulations used by Rahbeck et.al (2007) to data which are only in form of a site by species matrix and without any spatial information. A list similar to an nb object of spdep can prepared according to order in which the rows (sites) are arranged. A manually prepared list of neighbors for each site can also be used.It is important that each site must have at least one neighbor. 'rsize' provides a vector of range sizes.It can be 'unif' - ranges are drawn from a uniform distribution,between 1 to number of sites or 'observed' - range size of each species is exactly the same as in the observed matrix. Alternatively a it can also be a user specified integer vector, of same length as number of species.

Value

If degen is FALSE, a data frame with four columns for mean, SD and confidence intervals of expected richness

"mod.rich": mean richness of each site
"mod.sd": standard deviation of species richness
"q2.5": lower limit of the confidence interval
"q97.5": upper limit of the confidence interval

If degen is TRUE, then a list containing above data frame and a list of all the randomized matrices

References

Rahbek, C., Gotelli, N., Colwell, R., Entsminger, G., Rangel, T. & Graves, G. (2007) Predicting continental-scale patterns of bird species richness with spatially explicit models. Proceedings of the Royal Society B: Biological Sciences, 274, 165.

Gotelli, N.J., Anderson, M.J., Arita, H.T., Chao, A., Colwell, R.K., Connolly, S.R., Currie, D.J., Dunn, R.R., Graves, G.R. & Green, J.L. (2009) Patterns and causes of species richness: a general simulation model for macroecology. Ecology Letters, 12, 873-886.

Examples

tempmat <- matrix(0,nrow=10,ncol=200,dimnames=list(letters[1:10],1:200))
tempmat <- as.matrix(apply(tempmat,2,function(x){rbinom(nrow(tempmat),1,
                     runif(1,0.1,1))}))
rownames(tempmat) <- letters[1:10]
temp <- rangemod1d(tempmat,cohesion = TRUE,var = NULL,rsize = "observed",reps = 5)
plot(temp[,1],ylim= c(min(temp[,1] -2),max(temp[,1]+2)),pch = 16,ylab = 'Species Richness')
segments(1:10,y0=temp[,1]-temp[,2],y1= temp[,1]+temp[,2])

rangemodelR documentation built on April 3, 2025, 7:08 p.m.