LDPlot: A Comparative Live-Dead Barplot

In MJKowalewski/PaleoFidelity: Analyze and Visualize Agreement between Live and Dead Communities

A Comparative Live-Dead Barplot

Description

LDPlot function generates a comparative live-dead plot to visualize congruence in relative abundance and rank order of the most common live and dead species, genera, functional groups or other enumerated variables.

Usage

LDPlot(
  live,
  dead,
  tax.names,
  toplimit = 10,
  barwidth = 250/toplimit,
  col1 = "black",
  col2 = "gray",
  arr.col = "black",
  arr.lty = 1,
  arr.lwd = 1,
  cex.axis = 0.7,
  tck = -0.02,
  cex.lab = 0.8,
  cex.names = 0.95,
  cex.label = 1,
  cex.stat = 0.9,
  font.names = 3,
  cor.measure = "spearman",
  report = F,
  iter = 100
)

Arguments

live	A vector of integers with counts of live specimens by species/other units
dead	A vector of integers with counts of dead specimens by species/other units
tax.names	A vector with a list of names of species/other units
toplimit	A numerical value (default = 10) defining the number of top species (or other units) to be plotted
barwidth	A numerical value (default = 250 / toplimit) defining bar width (the thickness of lines used to represent bars)
col1	A character string (default = 'black') defining the color of bars for species/variables shared by live and dead data
col2	A character string (default = 'gray') defining the color of bars for species/variables unique to either live or dead data
arr.col	A character string (default = 'black') defining the color of arrows
arr.lty	An integer or character string specifying type of line used for arrows (default = 1)
arr.lwd	A numerical value specifying the width of arrow lines (default = 1)
cex.axis	A numerical value (default = 0.7) defining font size of tick mark labels
tck	A numerical value (default = -0.02) defining length of tick marks for x axis
cex.lab	A numerical value (default = 0.8) defining font size for x-axis label
cex.names	A numerical value (default = 0.95) defining font size for names of species/other units
cex.label	A numerical value (default = 1) defining font size for 'Live' and 'Dead' titles
cex.stat	A numerical value (default = 0.9) defining font size for correlation coefficient estimate
font.names	A numerical value (default = 3) defining font style
cor.measure	A character string (default='spearman') defining correlation measure (passed on to cor function) used to estimate live-dead correlations. Possible values are 'pearson', 'spearman', 'kendall', 'all'. Reported coefficients include all taxa regardless of how many taxa are included on the plot
report	Logical (default = FALSE) to print a summary output
iter	Numerical (default = 100) number of iterations for randomization which simulates perfect fidelity by resampling pooled live+dead species counts. Applicable only when the argument "report" is set to TRUE

Details

LDPlot function produces a dual barplot that compares rank abundance of one live sample with one dead sample for the top "n" most common species, genera, or other units (e.g., Kowalewski et al. 2003). The L-D comparison can apply to single samples or pooled data. Because names of species/other units can vary in length and the number of plotted variables can span a wide range of values, margin widths and cex argument may need to be customized. When species/other units are tied in rank, they are plotted in an arbitrary order. A warning is returned when more than 50

Value

A single plot. In addition, when the argument "report" is set to TRUE, a list with the following components is produced:

top.live	a list of top live species, including proportions and presence/absence in dead data
top.dead	a list of top dead species, including proportions and presence/absence in live data
summary	summary of species/taxa numbers in live, data, and pooled data
sample info	samples sizes in live, data, and pooled data
cor.coeff	observed coefficients of correlation/association
expected.coeff	expected coefficients of correlation/association estimated via randomization under the null model that the compared live and dead samples came from the same underlying population (perfect fidelity)
p.values	p.values for perfect fidelity null model based on percentile estimates of the resampling distribution
randomized.r	coefficients of correlation/association produced by the randomization model

References

Kowalewski, M., Carroll, M., Casazza, L., Gupta, N., Hannisdal, B., Hendy, A., Krause, R.A., Jr., Labarbera, M., Lazo, D.G., Messina, C., Puchalski, S., Rothfus, T.A., Sälgeback, J., Stempien, J., Terry, R.C., Tomašových, A., (2003), Quantitative fidelity of brachiopod-mollusk assemblages from modern subtidal environments of San Juan Islands, USA. Journal of Taphonomy 1: 43-65.

Examples

temp.par <- par(mar=c(3,6,1,6))
out1 <- LDPlot(live=colSums(FidData$live), dead=colSums(FidData$dead),
tax.names=colnames(FidData$live), toplimit=15, barwidth = 21,
col1 = 'green1', col2 = 'red1', arr.col = 'green4', arr.lty = 1,
report=TRUE, iter=100)
par(temp.par)
hist(out1$randomized.r[,2], main='', xlim=c(-1,1),
xlab=bquote("spearman"~rho), ylab='number of replicate samples')
arrows(out1$cor.coeff[2], 10, out1$cor.coeff[2], 0, length = 0.15, lwd=2)
text(out1$cor.coeff[2], 12, bquote(rho ==.(round(out1$cor.coeff[2],3))))

MJKowalewski/PaleoFidelity documentation built on Aug. 25, 2024, 8:27 p.m.