Hello,
we would like to expand our lab (qPCR) with sequencing analysis. We would like to sequence fish populations, and have the following questions:
We assume that each organism, and thus also each fish, sheds a different amount of DNA to the water. Could this influence the sequencing results (we want our results to show relative abundance).
What is the common way in this community to deal with this issue?
Hello!
we would like to expand our lab (qPCR) with sequencing analysis
Nice! 
This is a great question! Yes, we would expect different and varying amounts of genomic shedding that introduces bias. Similarly, we expect bias due to genome size and marker gene copy number.
People have been working to normalize copy number forever, with limited success.
That's the neat part; we don't.
We can compare samples and what's different without ever trying to infer absolute abundances. This works well because all these biases should be consistent sample-to-sample. We can detect if a microbe is increasing in relative abundance over time without ever knowing it's true colony-forming unit in the sample.
Going from counts of amplicons to counts of fish is pretty hard. Is this your goal?
EDIT: Here's a paper to consider: