This is a surprisingly interesting question. Can a dark particle 'feel' gravity? Most physicists think that any local experiment in a freely falling lab is independent of the velocity and position of the lab in spacetime (the strong equivalence principle). Here local means confined to the lab, but also small compared to variations in the gravitational field. However, you can setup nonlocal gravitational observatories in a number of ways that preserve the equivalence principle:
1) If the DM critters have a way to do non-local measurements, like measuring tidal forces, then they could measure gravity gradients over time. If they can communicate, they can set up gravity gradient sensors in freefall around stars and infer both the spatial distribution and time evolution of gravitational masses.
2) If the DM critters have clocks, they can measure the gravitational redshift around stars and discover stars. This should work even in a freely falling lab.
3) The DM critters could see gravitational waves passing their freely falling lab and infer changes in gravitational fields. This assumes gravitational waves exist (not a bad assumption).