Your figure is of a face mounted motor directly bolted to a centrifugal pump. The only reaction torque is during startup and shutdown, and it is shared between the motor and pump. The loads on the motor and pump feet are from the piping, coupling misalignment, pump impeller unbalance, thermal expansion, etc. I once saw a misaligned 60 hp pump vibrate so bad that it broke both pipes within a day after initial startup even though the pump and motor was solidly bolted to a solid concrete base.
If, however, the motor was mounted separately and the only connection to the pump was a flexible shaft coupling, the motor feet would be subject to the actual motor torque. The motor torque is easily calculated from the equation relating torque, power, and RPM.
Also, be advised that the pump feet are not necessarily in the same plane as the motor feet. If the installer bolts the motor/pump assembly to a flat surface, the forces involved can get very high. This is a case where practical realities are greater than simple theory (motor reaction torque) predicts.