From experience –attempting to do this for stream inverts– nothing in the world of eDNA is straightforward!  I'm probably telling you what you already know, but any semi-quantitative e-system that is going to be sufficiently accurate for reliable use requires the testing of many potential targets and many mixtures of different compositions before you can ever get near using the system in the field.

We published a paper in Conservation Genetics Resources that may be of interest to you.  We utilized a metabarcoding approach targeting fish specific COI fragments in a bait trade application.

Article Meta-genomic surveillance of invasive species in the bait trade