The relationship between rhythms of circadian periodicity (i.e., from 20 to 28 h) and ultradian periodicity (i.e., less than 20 h) was studied in the burrowing decapod Nephrops norvegicus at the level of its cardiac activity. Animals were kept over a month under constant darkness (DD) interrupted by a few days of light-darkness regime (LD) at the beginning of the experiment. Time series (beats per 10 min) were subdivided into stages of similar numbers of days. A general mean waveform was computed per stage by averaging 24 h segments of different time series per corresponding 10 min intervals. Marked fluctuations were observed at the beginning of tests in DD, being disrupted during animals' exposure to LD. Fluctuations progressively recovered over following stages of DD. The activity part of a rhythm (α) was computed in mean waveforms of different time series per each stage. Resulting values were averaged at corresponding stages. A significant increment of mean α was observed from DD to LD, decreasing over the following prolonged DD exposure. Periodogram analysis was used to assess periodicities of time series at each stage. Fourier analysis was undertaken to assess the transformation of cardiac rhythms over consecutive stages not only in terms of periodicity but also as amplitude. Both analyses showed the presence of different circadian and ultradian (i.e., 12 h and 18 h) rhythms varying in their amplitude at different stages of testing. A preponderance of time series with ultradian periodicities took place in the first stage of DD. Under LD, the number of time series showing 18 h periodicity increased, but their amplitude of fluctuation was lower compared to the previous stage. In contrast, the circadian periodicity present in the first stage of DD disappeared in LD, to be restored over prolonged DD exposure. Present results suggest that a disruption of the circadian rhythm in cardiac activity generated ultradian periods when controlling oscillators became uncoupled. Results are discussed in the context of the ecology of the species, and a model based on the phase decoupling of circadian oscillators is presented to account for the generation of ultradian 12 h and 18 h periodicities.