Each read group is aligned to the reference genome separately and all read group alignments that belong to a single aliquot are merged.
It seems that each sample/aliquot was sequenced several times (e.g. different libraries, lanes, runs) and, as such, each one is assigned a different read group that is mapped separately. However, since all aliquots come from the same sample, you want to merge them so you can use all data for your downstream analyses.
For some analysis you need certain amount of minimal reads is order to see what you are looking for. This is the reason why the same sample is sequenced several times, so you can reach this amount. The files can be merged from the fasta file or later in the pipeline of your analysis. In this case, as you say, bam/sam files.