[R] Nested anova with unbalanced design and corrected sample size for spatial autocorrelation

[=?utf-8?B?RnJhbmNlc2MgTW9udGFu6Q==?=]

Hello all,

This may be a simple question to answer, but I'm a little bit stumped with
respect to the calculation of the F statistics in nested  anovas with
unbalanced design in R.

In my case, I have 11 vegetation transects (with 1000 10cmx10cm squares),
where we estimated shrub cover. We have two different treatments: wildfire
(4 transects) and prescribed burning (7 transects) and we want to compare
the mean shrub cover between the 2 different treatments. I guess that I
have to apply a one-way nested anova (transect number within treatment)
with unbalanced number of samples (4000 in wildfire vs 7000 in prescribed
burning).
Moreover, I have to correct the initial sample size (1000 squares) to a
corrected sample size by spatial autocorrelation (which in fact, makes all
the n different between transects).

Can anyone, please, tell me how to do this in R? Do I need to use lme()?
Or is it possible to do it using aov()?

Thanks a lot!

Francesc

PhD student
University of Barcelona