I want to investigate patterns of evolution (coevolution, host switching, ...) in a host-parasite system using reconciliation methods.

1-should the host and parasite trees be made by a similar marker (e.g. COI for both of them) or different markers (e.g. COI for host vs. 28s for parasite) can be used? 

2- how many markers are necessary?

your reply will be highly appreciated