What is the mechanism that causes more inductive spikes with a higher duty cycle, even if the frequency of the pwm is held constant? Doesn't the motor create voltage kickback only when it is turned off? If that is the case increasing duty cycle doesn't increase the frequency of the motor turning off(since PWM frequency is constant) and so should not increase motor spikes.
I am observing the PWM signal on a scope, and the spikes are more frequent and of larger magnitude. This is my setup: 
The left portion of the circuit controls the duty cycle of the pwm, which is generated on the out pin.