The get.group command allows you to get a listing of the different groups available for the multi-sample OTU data that are currently stored in memory. This feature should help in using the group option for other commands. For this tutorial you should download and decompress amazondata.zip.
The shared parameter is required.
mothur > get.group(shared=amazon.an.shared) rainforest pasture
This tells you that you have two groups: one for the rainforest sequences and another for the pasture sequences.