The original question was phrased to hold airspeed constant between the gliding and powered cases, not angle-of-attack. To hold the airspeed constant as we transition from the gliding case to the powered case, we must increase the angle-of-attack to increase the lift coefficient to provide the extra lift needed in the powered case. Now we can be sure that the L/D ratio is almost certainly not remaining constant. Since airspeed is now constant, the change in the required power will be directly proportional to the resulting change in the Drag vector. And the resulting change in the Drag vector depends upon where we are on the L/D versus airspeed curve. If we are in high-speed cruise well above the best L/D speed, the Drag vector will actually be smaller in level flight than in gliding flight. If we are flying slower than best L/D speed, the Drag vector will be larger in level flight than in gliding flight. There is aone special case where we happen to gliding with an angle-of-attack slightly lower than the angle-of-attack yielding the maximum L/D ratio and minimum Drag, and as we transition to horizontal flight, we increase the angle-of-attack and end up with an angle-of-attack slightly higher than the angle-of-attack yielding the maximum L/D ratio and minimum Drag, and the net change of Drag is exactly zero.