Temporal integration of melatonin infusion duration: signal averaging versus frequency dependence

Day length affects somatic and reproductive physiology of Siberian hamsters via regulation of the duration of nocturnal pineal melatonin secretion. Nightly 'long' (e.g. 12 hr) or 'short' (e.g. 6 hr) melatonin signals inhibit or stimulate gonadal growth, respectively. When long and short signals are presented in combination, however, neuroendocrine mechanisms exhibit a frequency-dependent response, stimulating gonadal growth only if short signals are presented every second night or more frequently. The present experiments further assessed formal models for the temporal integration of melatonin signals changing abruptly in duration from night to night. Photo-inhibited Siberian hamsters were housed in constant light and infused subcutaneously with various combinations of nightly short or long melatonin signals according to one of the several regimes that varied the frequency of short melatonin signal occurrence, average duration of the nightly melatonin signal, or both. Six weeks of nightly alternating short and long signals yielded different gonadal responses depending on the average melatonin signal duration. Moreover, when average melatonin signal duration was held constant between groups, gonadal stimulation was independent of the frequency of the constituent melatonin signals except when the duration of the short signal was reduced to 3 hr. Thus, neuroendocrine mechanisms do not solely categorize melatonin signals as either long or short but attend also to the duration of each component signal. In the majority (six of seven) of infusion regimes, reproductive responses to chimeric patterns of long and short melatonin signals were compatible with a simple signal-averaging mechanism.