Summary: FD is getting very popular, so I figured that would be good to post not only the code (mostly borrowed) I am using: https://github.com/ibartomeus/fundiv, but also what I learnt of it. This posts assume you have read about FD calculation a bit.
I’ve been working with functional diversity (FD) for a couple of years now, and some papers are to come applying this concept to pollinator (and other insect). You can find some help on calculation FD in Owen Petchey‘s webpage and in FD package. However, Petchey’s et al FD is not in an easy to use R function, so I uploaded some wrap up R functions to calculate the principal indexes, along with some null models and other stuff. There is a brief explanation on how to install and use the package there. But first some background: FD calculates the diversity of traits present in the community, and is a cool concept, because at a same number of species, a community can have quite homogenous or diverse traits, and this can be important to depict mechanisms of ecosystem functioning or responses to global change. However I have a hate-love relationship with the concept.
The first obvious one is that which traits are considered will directly affect the conclusions you get. The garbage in -> garbage out concept applies here, but even if you try to select appropriate traits, trait information quality is usually not good for most species, we don’t know what traits are responsible for which functions and any choice of traits is more subjective than a simple tally of species richness (or Phylogenetic diversity). However, if you identify traits that are meaningful for your question (e.g. body size is the usual suspect), can potentially be a very powerful tool.
A second big problem is that there is quite a few metrics out there. Just reading the literature is quite daunting, so I`ll give you my opinion after implementing them in R and using them on a lot of different datasets. First problem is that Functional diversity calculated as a dendrogram (Petchey approach) or as a space based (FD package approach) metric is not equivalent (they are correlated, but not strongly), and this is bothering me. If I have to choose, I like better dendogram based solutions. Why? You have the ability to calculate FD for communities with very few species (even richness = 1 species). This is useful sometimes, especially with simulations of random species removal. You can easily visualize the dendogram, but not a 8D space. For me dendograms are also easier to understand than PCoA’s. Moreover, the original distance matrix used tend to be better described by dendograms than by PCoA’s (in several datasets). Space based calculations crash more often than one would want and you are forced to drop lots of axes to calculate hypervolums (and sometimes is impossible to get some metrics for concrete datasets). Also, Feve and Fdiv are hard to interpret for me, even after reading about them several times. Said so, I have to say that we found Feve to predict fairly good function, and better than dendogram based indexes. Dendrogram based indexes are not perfect either, and specially they need for a better coverage of different complementary metrics, including not only Frich, but also equivalents of Feve and Fdis. I developed a weighted version of Frich (see link to package above), and an evenness one can be easily calculated using phylogenetic imbalance metrics (or see treeNODF!). If anyone is interested in exploring this options further drop me an email.
Bartomeus lab 
Pingback: On Functional Diversity metrics ← Patient 2 Earn
Quick thought on the question which pops up why space-based approaches might be more useful in prediction: think about the n-dimensional hypervolumes of your measured traits as an analogon of the n-dimensional ecological ‘nice space'(*). The ecological factors which a species is able to adapt to (or: is physiologically, and phylogeneticall preadapted to) shape the traits, and (expressable) traits themselves constrain the realised niche of your given species.
From my POV, dendrogram approaches inherently throw away meaningful information. Space based approaches also do this, but to a lesser extent. If you are searching for patterns, don’t reduce your dimensions beyond need. If you want to explain those patterns with covariates, then don’t be surprised that the feedback-loop of covariates to your traits is a good predictor. :)
(*) Just for the fun of it: I’ve got a tiny bit of a background as evolutionary biologist, and my teachers were influenced by Peter Ax. Guess they would start yelling at me for using ‘niche space’ at all, as they would argue that the niche is identical with the species. Diving into that boggles the mind of most people, but I highly recommend digging into the philosophical implications. I’m still boggled, but at the same time, I flinch every time my ecologically trained colleagues use the word ‘niche’, or even worse the expression ‘filling a niche’. It’s like teaching someone to see bad kerning. (cf. xkcd.com/1015 )
Maybe you are right and there is a theoretical expectation for Space based approaches to perform better than dendrograms for prediction, but from a practical point of view, simplifying a complex concept (a.k.a. throwing away information) can be also desirable in some situations. In any case, I don’t think in the current implementations of the indexes dendrogram based approaches throw more information away than space based indexes. With my data (11 different datasets), dendrograms capture better the information of the dissimilarity matrix than PCoA’s. I would love to see a robust study comparing its performance with simulated data and achieving a consensus on how to measure functional richness, diversity and evenness.
It was just a quick thought. :)
I’m all for simplifying. To paraphrase: all models are wrong, but some are more beautiful.
And I agree wholeheartedly that a simulation study would be great. (Sadly, I’m not at all qualified to do this.)
Pingback: Can functional diversity indices predict ecosystem function? | Bartomeus lab