If I remember, I think there's work on the development of nanowires or nanoparticles for the isolation of microbial strain or virus

If you want to find rare strains present in one environment but absent in another, you should better use a large number of selective media and many samples for both variants (I suggest above 100). Other ways are mostly linked to molecular study without bacteria isolation.

What do you mean of without bacteria isolation.

See profil of Oualid Baghriche (University of constantine)