fingerprint.regression {pez}

Regress trait evolution against trait ecology (following Cavender-Bares et al. 2004)

Description

Calculates traits' phylogenetic inertia and regresses this against trait similarity among co-existing species (sensu Cavender-Bares et al. 2004 Figure 6)

Usage

fingerprint.regression(
  data,
  eco.rnd = c("taxa.labels", "richness", "frequency", "sample.pool", "phylogeny.pool",
    "independentswap", "trialswap"),
  eco.method = c("quantile", "lm", "mantel"),
  eco.permute = 1000,
  evo.method = c("lambda", "delta", "kappa", "blom.k"),
  eco.swap = 1000,
  abundance = TRUE,
  ...
)

## S3 method for class 'fingerprint.regression'
print(x, ...)

## S3 method for class 'fingerprint.regression'
summary(object, ...)

## S3 method for class 'fingerprint.regression'
plot(
  x,
  eco = c("slope", "corrected"),
  xlab = "Community Trait Similarity",
  ylab = "Phylogenetic inertia",
  ...
)

Arguments

data

comparative.comm for analysis

eco.rnd

null distribution with which to compare your community data, one of: taxa.labels (DEFAULT), richness, frequency, sample.pool, phylogeny.pool, independentswap, trialswap (as implemented in picante)

eco.method

how to compare distance matrices (only the lower triangle;), one of: lm (linear regression), quantile (DEFAULT; rq), mantel (mantel)

eco.permute

number of permutations for ecological null model (eco.rnd); default 1000

evo.method

how to measure phylogenetic inertia, one of: lambda (default), delta, kappa, blom.k; see phy.signal.

eco.swap

number of independent swap iterations to perform (if specified in eco.rnd; DEFAULT 1000)

abundance

whether to incorporate species' abundances (default: TRUE)

...

additional parameters to pass on to model fitting functions and plotting functions

x

fingerprint.regression object

object

fingerprint.regression object

eco

plot the observed slopes (DEFAULT: slope), or the median difference between the simulations and the observed values (corrected)

xlab

label for x-axis (default "Ecological Trait Coexistence")

ylab

label for y-axis (default "Phylogenetic inertia")

Details

While the term ‘fingerprint regression’ is new to pez, the method is very similar to that employed in Cavender-Bares et al. 2004 Figure 6. For each trait, the phylogenetic inertia of species traits is regressed against their co-occurrence in the community matrix. Note that Pagel's lambda, delta, and kappa, and Blomberg's K, can be used, unlike the original where a mantel test was employed. Moreover, note also that Pianka's distance (as described in the manuscript) is used to measure species overlap.

Note

Like eco.xxx.regression, this is a data-hungry method. Warnings will be generated if any of the methods cannot be fitted properly (the examples below give toy examples of this). In such cases the summary and plot methods of these functions may generate errors; perhaps using traceback to examine where these are coming from, and consider whether you want to be working with the data generating these errors. I am loathe to hide these errors or gloss over them, because they represent the reality of your data!

WDP loves quantile regressions, and advises that you check different quantiles using the tau options.

Author(s)

Will Pearse and Jeannine Cavender-Bares

References

Cavender-Bares J., Ackerly D.D., Baum D.A. & Bazzaz F.A. (2004) Phylogenetic overdispersion in Floridian oak communities. The Americant Naturalist 163(6): 823–843.

Kembel, S.W., Cowan, P.D., Helmus, M.R., Cornwell, W.K., Morlon, H., Ackerly, D.D., Blomberg, S.P. & Webb, C.O. Picante: R tools for integrating phylogenies and ecology. Bioinformatics 26(11): 1463–1464.

Pagel M. Inferring the historical patterns of biological evolution. Nature 401(6756): 877–884.

See Also

eco.xxx.regression phy.signal

Examples

data(laja)
data <- comparative.comm(invert.tree, river.sites, invert.traits, river.env)
fingerprint.regression(data, eco.permute=10)
plot(fingerprint.regression(data, permute=10, method="lm"))