phangorn  package in R could be used for comparing phylogenetic trees. Check the link below for more information about that package: 

http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3035803/

You could have a look in this paper

http://sysbio.oxfordjournals.org/content/64/2/205.full

And choose the measure the most appropriated to your data.

Thank you ..

Many consider that measures of tree similarity based on rearrangement algorithms (such as SPR) are the gold standard for tree comparing techniques (an idea also shared by the paper by Kuhner & Yamato to which you were already refered). In this sense, the software TNT has extremely fast SPR algorithms implemented, which allow to use this measure instead of having to settle with faster but less accurate ones.

However, I have found that, in general, an SPR based metric performs fairly similar to the Robinson-Foulds distance. The RB distance even retains more resolution than the SPR one when comparing very different trees, and is extremely fast and more intuitive.

As for R packages, I have no idea. Sorry.

Hi,

a similar question has been posted last year, you can find links toward numerous tools/R packages that allow to compare phylogenetic trees on this thread:

https://www.researchgate.net/post/Is_there_a_software_that_can_compare_two_phylogenetic_trees_to_each_othe_face_to_face_comparison

Some of my colleagues also recently developped CompPhy: a web-based collaborative platform for comparing phylogenies

http://www.atgc-montpellier.fr/compphy/

http://www.biomedcentral.com/1471-2148/14/253

Concerning the RF and SPR distances, the RF is widely used as it is easy to compute and to understand. However you may have cases where two trees are identical up to a leave placement that will lead to a RF similarity of 0 even thought the two trees are almost identical (e.g. http://doi.ieeecomputersociety.org/cms/Computer.org/dl/trans/tb/2011/04/figures/ttb20110409024.gif). SPR, MAST or quartet distances are less sensitive to such misplacement of few leaves/subtrees. 

I hope this will be helpful.

The SPR distance is hard to compute (NP-hard). The Robinson-Foulds distance is a typical distance that is computable in polynomial time. We have an implementation of it in PhyloNet (http://bioinfo.cs.rice.edu/phylonet), but it's easy to implement.

Luay

The R package Quartet implements bipartiton and quartet comparisons, from which several metrics can be calculated, including RF distances and Quartet distances.

https://cran.r-project.org/web/packages/Quartet/index.html