mult_abundances: Fit a unimodal gambin model to multiple species abundance...
In gambin: Fit the Gambin Model to Species Abundance Distributions

Description Usage Arguments Details Examples

Fits the unimodal gambin model to the SADs from multiple sites and returns the standardised and unstandardised alpha values.

1	mult_abundances(mult, N = 100, subsample = NULL, r = 3)

`mult`	Either a matrix, dataframe or list containing the species abundance data of a set of sites. In the case of a matrix or dataframe, a given column contains the abundance data for a given site (i.e. columns are sites and rows are species; each cell is the abundance of a given species in a given site). In the case of a list, each element in the list contains the abundance data (i.e. a vector of abundances) for a given site.
`N`	The number of times to subsample the abundance data in each site to calculate mean standardised alpha.
`subsample`	The number of individuals to sample from each site before fitting the gambin model. The default is subsample = NULL, in which case subsample is set to equal the number of individuals in the site with the fewest individuals.
`r`	The number of decimal points to round the returned alpha values to (default is r = 3)

Because the alpha parameter of the gambin model is dependent on sample size, when comparing the alpha values between sites it can be useful to first standardise the number of individuals in all sites. By default, the mult_abundances function calculates the total number of individuals in each site and selects the minimum value for standardising. This minimum number of individuals is then sampled from each site and the gambin model fitted to this subsample using fit_abundances, and the alpha value stored. This process is then repeated N times and the mean alpha value is calculated for each site. The number of individuals to be subsampled can be manually set using the subsample argument. The function returns a list, in which the first two elements are the mean standardised alpha values for each site, and the raw unstandardized alpha values for each site, respectively. The full set of N alpha values and X2 P-values for each site are also returned. As an input, the SAD data can be in the form of a matrix or dataframe, or a list. A matrix/dataframe is only for when each site (column) has abundance data for the same set of species (rows). For example, an abundance matrix of bird species on a set of islands in an archipelago. A list should be used in cases where the number of species varies between sites; for example, when comparing the SADs of samples from different countries. In this case, each element of the list contains an abundance vector from a given site.

At present, the mult_abundances function only fits the unimodal gambin model.

#simulate a matrix containing the SAD data for 10 sites (50 sp. in each)
mult <- matrix(0, nrow = 50, ncol = 10)
mult <- apply(mult, 2, function(x) ceiling(rlnorm(length(x), 0, 2.5)))

#run the mult_abundances function and view the alpha values
mm <- mult_abundances(mult, N = 10, subsample = NULL)
mm[1:2]
plot(mm$Mean.Stan.Alpha, mm$Unstan.Alpha)

#simulate a list containing the SAD of 5 sites (with varying numbers of sp.)
mult2 <- vector("list", length = 5)
for (i in 1:ncol(mult)){
 dum <- sample(mult[, i], replace = TRUE)
 rm <- round(runif(1, 0, 5), 0)
 if (rm > 0){
   rm2 <- sample(1:length(dum), rm, replace = FALSE)
   dum <- dum[-rm2]
 }
 mult2[[i]] <- dum
}

#run the mult_abundances function on the list
mm2 <- mult_abundances(mult2, N = 5, subsample = NULL)
mm2[1:2]

Minimum number of individuals in a site = 295
Subsampling 295 individuals from each site
 
Warning message:
In mult_abundances(mult, N = 10, subsample = NULL) :
  Some X2 P-value less than 0.05
$Mean.Stan.Alpha
 [1] 0.757 0.657 0.176 1.116 0.737 0.482 0.487 0.859 1.126 0.818

$Unstan.Alpha
 [1] 0.597 0.474 0.466 0.736 0.777 0.407 0.411 0.823 1.126 0.458


Minimum number of individuals in a site = 115
Subsampling 115 individuals from each site
 
Warning message:
In mult_abundances(mult2, N = 5, subsample = NULL) :
  Some X2 P-value less than 0.05
$Mean.Stan.Alpha
 [1]  1.053  2.022 19.703  0.950  0.939  0.911  0.757  1.180  1.233  2.473

$Unstan.Alpha
 [1] 0.586 0.617 0.524 0.843 0.939 0.542 0.605 1.559 1.088 0.793