[R-sig-eco] Adonis and Random Effects

[Jari Oksanen]

On 04/02/2013, at 09:14 AM, Erin Nuccio wrote:

> 
> Is adonis capable of modeling random effects?  

No, adonis does not know random effects.

The significance tests in adonis() are based on permutations. The key question for random effects is: "can you design a permutation matrix that treats some variables like they were random and others like they were fixed." If your answer is "yes", you can have random effects, and if you use the R-Forge versions of vegan (now 2.1-25), you can input your own permutation matrices. If your answer is "I don't know", then there is no way of handling random effects. 

You cannot expect lme4 extensions to formula to work in other packages. We can parse many fixed effect formulae to a model matrix, but there is a long way of getting permutations to work in some specific way for variables tagged as random factors. 

Cheers, Jari Oksanen

> I'm analyzing the impact of a treatment on the microbial community in a split-plot design (2 treatments per plot, 4 plots per grassland, 3 grasslands total). I would like to quantify how much of the variance is due to the Treatment versus the Grassland.  It seems like Grassland should be a random effect, since there are thousands of grasslands, and I'm only looking at 3.
> 
> I have tried to use the notation that works with lme4, and it's not working for me (see below for formula and error messages).  If adonis can't do random effects, are there any alternatives?  Or, considering my goal, are there any other programs I should look into?  Any suggestions would be highly appreciated!
> 

-- 
Jari Oksanen, Dept Biology, Univ Oulu, 90014 Finland