It will be a combination of things. Like TonyK mentioned in his comments, most of the change will occur during cis-lunar transfer. This is optimal because inter-planetary trajectories can be designed to result in an optimal parking orbit around your target planetary body. One neat trick astrodynamists use is B-plane targeting (a nice AGI article here: http://help.agi.com/stk/index.htm#gator/eq-bplane.htm). I do not know exactly what ISRO did to design these trajectories but I am sure it follows in the same line of reasoning in that since the interplanetary problem is underconstrained, there are multiple parameters engineers can play with to get their desired orbit.
For some reason if the initial inclination of the orbit is not ideal, ISRO could "wait it out" if it is within some margin for the osculating orbit to fall within some tolerance for its orbital parameters. These osculating orbital elements may not be periodic in nature. A good example is polar LEO satellites experiencing something called RAAN drift, where the obliquity of Earth leads to an ever increasing RAAN over time. Obviously if you start off at 0 RAAN, you will naturally end up back there since 360 degrees = 0 degrees, but in a non-spherical body, you will see that these orbital elements will non-linearly effect other parameters including inclination.
Hoping for orbital perturbations to get you into your desired orbit seems a bit naive at first glance (correct me if I am wrong), so you will probably design your entry trajectory to get you where you need to go. Mentioning these perturbations, I think, is neat to know and I am sure very important if you were designing the control systems to maintain your desired orbit.