It depends, but in general, it is expected be on the lower end.
The principal task of a "research science manager" is not to "do" the research, but foster an environment that will help other researcher that are actually focused into looking in / for the "dirt" (by dirt I assume you mean the nitty-gritty details of unraveling research problems).
This would be setting high level direction, advising on approach to research problems, as well as the general "bureaucracy" and logistics required to make the team running.
If time allows, a research science manager can still explore some research topics on his own as a research, but that becomes more of an side quest than actual main quest. Of course, they can still yield interesting result.
Finally, a research science manager should not just work in a top down fashion, but also incorporate bottom up feedback, i.e. adapt the overall research team's direction based on the "dirt" that is brought up to him by the dedicated researchers.
The principal task of a "research science manager" is not to "do" the research, but foster an environment that will help other researcher that are actually focused into looking in / for the "dirt" (by dirt I assume you mean the nitty-gritty details of unraveling research problems). This would be setting high level direction, advising on approach to research problems, as well as the general "bureaucracy" and logistics required to make the team running. If time allows, a research science manager can still explore some research topics on his own as a research, but that becomes more of an side quest than actual main quest. Of course, they can still yield interesting result.
Finally, a research science manager should not just work in a top down fashion, but also incorporate bottom up feedback, i.e. adapt the overall research team's direction based on the "dirt" that is brought up to him by the dedicated researchers.