I'll offer one idea. Whenever the battery is disconnected from the circuit, all capacitors become discharged. And this includes the capacitor we call the mosfet gate. When the battery is plugged in and power comes up, you can get a significant GS differential - enough to turn on the mosfet - because the source rises immediately, but the gate can only come up through the high-value resistor. The capacitor is there to transfer the source voltage immediately to the gate. A capacitor acts as a low-impedance pathway to a rapidly changing signal. So for a very brief period, it's as if the source and gate are tied together, so there can be no GS differential, and that keeps the mosfet off. Then after that, the resistor takes over and maintains the same situation. So basically, it's there to prevent the mosfet from turning on by itself when you first connect power to the circuit.
That's just one guess, and I don't know if it makes sense. In any case, you are substantially increasing the effective gate capacitance, so subsequent transitions would tend to be slow. That's probably not an issue in this situation.