[R-sig-ME] Williams test from StatsCharrms package after linear mixed model with factors as explanatory variables (mixed anova)

[Inka Willms]

Hello, I am working on the analysis of studies that have to be analyzed exactly as described in an OECD Guideline. It is written, that after a mixed anova (lmer(as.numeric(fluorescence)~ as.factor(treatment) + (1|as.factor(replicate)) + (1|Replicate:Treatment))
a williams test like described in the StatsCharrms package has to be conducted, to receive Information about the influence of the chemical on the test subjects. 

It is not possible to use the williams option from the glht function of the multcomp package, as that is not the correct one ( not the one required for ecotoxicology). 
My problem is that I do not know on which output of the mixed anova I can conduct the williams test. I would really appreciate your help!