How to root a phylogenetic tree when calculating phylogenetic diversity in plant community?

Question

[shadowstep]

I randomly draw some species (say 1, 2 and 4) from a 8 common species pool to assemble the communities, then transplant a special species X into the communities. Now I want to measure the effect of phylogenetic diversity of the communities, on the growth performance of species X.

Q1: I first estimated a phylo tree of the common species in communities, without species X, to calculate PD. In the R package picante, even if I used a unrooted tree, to calculate PD of a community with only one common species, the pd() function will return me a value (NOT 0), so what does this value mean? Does a community with only one species really have a phylogenetic diversity?

Q2:Otherwise, what criterion should I use to choose a correct root for a specific phylo tree?

Thanks for you help!