synchrony (version 0.3.8)

community.sync: Compute community-wide synchrony and its significance via Monte Carlo randomizations

Description

Compute community-wide synchrony and its the significance via Monte Carlo randomizations. If all species fluctuate in perfect unison, the community-wide synchrony will be 1. If species undergo uncorrelated fluctuations, the community-wide synchrony will be 1/S. The Monte Carlo randomizations are performed by shuffling the columns of the community matrix independently. This function also returns the mean correlation between the columns of the matrix.

Usage

community.sync (data, nrands = 0, method = c("pearson", "kendall", "spearman"), 
                alternative = c("greater", "less"), type = 1, quiet = FALSE, ...)

Arguments

data

community matrix in wide format where each row contains the abundance at each time step and each column corresponds to a different species.

nrands

number of randomizations to perform (default is 0)

method

Method to compute mean correlation between columns? Options include pearson, kendall, and spearman. Default is pearson

alternative

Alternative hypothesis. Options are less and greater. Default is greater

type

Randomization method. The type=1 method randomly shuffles each column of the data matrix, thus destroying both the autocorrelation structure of each column and the cross-correlation between columns. The type=2 method shifts each column of the data matrix by a random amount, thus preserving the autocorrelation structure of each column but destroying the cross-correlation between columns (Purves and Law 2002). Default is type=1

quiet

Suppress progress bar when set to TRUE. Default is FALSE

…

Other parameters to cor function.

Value

Returns a named list containing:

obs

the observed community synchrony

meancorr

the mean correlation between the columns of the matrix

rands

the community synchrony value the randomizations. This variable is only returned if nrands > 0.

pval

p-value of observed community synchrony. This variable is only returned if nrands > 0.

alternative

Alternative hypothesis. This variable is only returned if nrands > 0.

Details

Loreau and de Mazancourt (2008) show that community-wide synchrony \(\varphi\) can be quantified by computing the temporal variance \(\sigma_{x_T}^2\) of the community time series \(x_T(t)=\sum{x_i(t)}\) and the sum of the temporal standard deviation of the time series across all species \(\left(\sum{\sigma_{x_i}}\right)^2\) such that: \(\varphi=\frac{\sigma_{x_T}^2}{\left(\sum{\sigma_{x_i}}\right)^2}\)

References

Loreau, M., and C. de Mazancourt. 2008. Species synchrony and its drivers: Neutral and nonneutral community dynamics in fluctuating environments. The American Naturalist 172:E48-E66.

Purves, D. W., and R. Law. 2002. Fine-scale spatial structure in a grassland community: quantifying the plant's eye view. Journal of Ecology 90:121-129.

Examples

# NOT RUN {
# Community matrix for 20 species undergoing random fluctuations 
comm.rand=matrix(runif(100), nrow=5, ncol=20)
community.sync(comm.rand, nrands=20)$pval
# Community matrix for 20 species undergoing synchronized fluctuations 
comm.corr=matrix(rep(comm.rand[,1], 20), nrow=5, ncol=20)
community.sync(comm.corr, nrands=20)$pval
# On "real" data
data(bird.traits)
community.sync(bird.traits, nrands=20)$pval
# }