Hi @Jana_Greta,
That's certainly an important criteria to keep in mind. Do the primers target the same region or different areas? Does each sample have 2 regions sequenced or different samples have different regions. There's been a few posts about how to deal with this that may be of help. See this discussion for example with lots of other useful links.
So this relates to the area where your forward and reverse reads approaching from opposite directions meet and 'overlap'. The overlap region is what is used to merge your reads into one longer read and perhaps correct some errors. This area depends on your primers and how long the sequencing cycles was. In your case you have 2x250 cycles so a total of 500 bp read. You need to calculate based on the total size of your amplicons how much of the 500bp is overlap region. Your truncating parameters should factor this into consideration. For a more thorough explanation see here.
I was actually referring to the --p-n-reads-learn parameter in the dada2 plugin which is basically dictating how many reads should the algorithm use to train its error models, but don't worry about right now, that suggestion no longer is important in this case and the default settings should be plenty enough.