Can R substitute PRIMER efficiently and effectively? Quite a simple question, looking to see if anybody has experience using both. I have no knowledge of PRIMER, but wondering if R (currently) has the same capacity for data analysis specifically for ecological data using things like perMANOVAs and NMDS plotting. 
I know it can be done in R (what can't!?) but I'm more concerned if it's a much longer route to the same conclusion.
 A: Yes. I have no experience using PRIMER because I didn't want to pay for it, but I have used the vegan package in R extensively for cluster analysis, NMDS, perMANOVA, CCA and PCA. To give you some idea of what R is able to achieve (at least graphically) using this package, I've included a final image from my thesis (first image) and and alternative version I was playing around with that incorporated ggplot2 (second image but not finalized, I did this one differently and wrong in the early days!). As always with R, there is a load of flexibility that I think may be a bit more limited in PRIMER (just a guess).


A: Yes. I have used both, and these days I only use R. I am going to extend the scope of my answer by providing a quick and dirty comparison of the capabilities of PRIMER v7, CANOCO5, PAST, and R for multivariate analysis:
Things PRIMER v7 can do that I can do in R: calculation of distance matrices according to various metrics, ANOSIM test, SIMPER, MDS and PCA ordinations, cluster analysis, diversity curves, PERMANOVA.
Things PRIMER v7 can do that I have never done in R: SIMPROF, BEST, Bio-Env, BVStep, LINKTREE, RELATE, 2STAGE, MVDISP, DIVERSE, TAXDTEST. It may be possible to perform these tests in R, I just don't know.
Things CANOCO5 can do that can be done in R: all of them, as far as I can see.
Multivariate analyses that PAST can do that can also be done in R: all of them, with the possible exception of seriation.
Generally speaking, in the context of multivariate statistical analyses, R allows to perform all the ordination approaches you can dream of, visualisation of multivariate data, and statistical testing of the correlation between changes in multivariate biological information and environmental variables. Package vegan is the staple for this kind of analyses, but there are others. See here for further detail.
