I want to use MCMCglmm to account for phylogenetic autocorrelation on my GLMM. However I have more than one trait measurement for certain species.
Is there a way to account for intraspecific variation using MCMCglmm such as it can be done in geiger's fitContinuous including a vector of standard errors in SE?
The MCMCglmm package is well suited to incorporating repeated measures of interspecific data. The following code demonstrates how to account for repeated measures by specifying a random effect for species in the MCMCglmm function. The pedigree argument did not work for me, so I use ginverse to specify the covariance structure of the random effect (i.e., the phylogenetic covariance structure).
# Load packages
library(phytools)
library(MCMCglmm)
# Generate data
phy <- pbtree(n=20)
dat <- fastBM(phy, nsim=12)
dat <- stack(data.frame(t(dat)))
colnames(dat) <- c("values", "species")
# Phylogenetic covariance structure of random effects
phy.inv <- inverseA(phy)$Ainv
# Fit model
fit <- MCMCglmm(values ~ 1, random=~species, data=dat, ginverse=list(species=phy.inv))
