Is increase in post-learning paradoxical sleep modified by cueing?

Rats were submitted to three sessions of an active avoidance conditioning. The conditioned stimulus (CS) was either an ear shock or a tone. Sleep was measured after each session. Experimental animals received non-awakening ear shocks during post-learning paradoxical sleep (PS). When ear shocks were the CS and were subsequently used as cues during PS, PS increases were observed, but they displayed 2 unusual characteristics: (1) they were due to longer average size of PS phases rather than to higher number of phases as usually; (2) PS did not return to baseline level even when learning seemed to be achieved. By contrast, when ear shocks were not associated with conditioning, animals shocked during PS exhibited the usual PS increases.