1. Fast rhythms in discharges of individual phrenic (PHR) motoneurons were studied by spectral and interval analyses; and they were compared, using coherence analysis, with similar rhythms in whole-PHR nerve discharge. The purpose of this study was to ascertain the origin of the two rhythms, manifested as distinct spectral peaks, in PHR motoneuron and nerve discharge: medium-frequency oscillations (MFO, usual range 20-50 Hz); and high-frequency oscillations (HFO, usual range 50-100 Hz). 2. In paralyzed artificially ventilated cats, unit recordings were taken from 1) 26 isolated single PHR fibers (in 8 sodium pentobarbital-anesthetized cats) and 2) 27 identified PHR motoneuron somata in the spinal cord (in 5 decerebrate cats). Simultaneous whole-PHR activity was monophasically recorded from the contralateral PHR nerve for 1 and from both PHR nerves for 2. 3. The signals were subjected to time- and frequency-domain analyses. The latter included a novel application of coherence analysis to the study of population synchrony. 4. The autospectra of all PHR units showed prominent MFO peaks in the frequency range of the nerve MFO spectral peaks, as well as harmonic peaks, indicating the presence of this type of fast rhythm in the units' discharges. Spectral analysis of the augmenting PHR activities in different segments of the inspiratory (1) phase showed that the frequency of unit MFO and of nerve MFO rose during the course of I. Further, cycle-triggered histogram and interval analysis indicated that the frequencies of unit MFO autospectral peaks were very close to the peak firing rates of the units during the portion of I analyzed. Thus unit MFO spectral peaks reflected the rhythmic and augmenting discharges of the motoneurons, and similar nerve MFO peaks reflected the superposition of individual motoneuron discharges. 5. The coherences of motoneurons' MFOs to nerve MFOs were low or zero, indicating that only partial and weak MFO correlations occurred within the PHR motoneuron population. 6. In those cats (n = 11) that had clear PHR nerve HFO spectral peaks, about one-half of the recorded PHR motoneurons had HFO, as indicated by HFO peaks in the unit autospectra and/or the unit-nerve coherences. 7. For motoneurons having HFO, the coherence between unit and nerve HFOs was substantial, particularly when the latter were strong, indicating HFO correlations among a number of PHR motoneurons. 8. In the light of theoretical considerations on the generation of aggregate rhythms from superposition of unitary rhythms, these observations indicate the following. 1) PHR nerve MFO arises from uncorrelated or weakly correlated MFOs in PHR units having augmenting discharge rates. As a consequence, nerve MFO frequency increases in the course of I, together with unit MFO frequencies; and the nerve MFO is manifested as a broad spectral deflection with a maximum in the band of peak firing rates of the units. 2) PHR nerve HFO arises from correlated, common-frequency HFOs in a subpopulation of PHR units, which are due to HFO inputs from medullary I neurons; it is thus manifested as a sharp, usually dominant autospectral peak. 9. Thus the MFOs reflect the rhythmic and augmenting discharge of individual PHR motoneurons that results from the augmenting drive of supraspinal origin; and the main MFO peak in the PHR nerve spectrum is an indicator of the peak firing rates in the population. On the other hand, HFOs in PHR motoneurons and nerves are a manifestation of the common medullary (system) HFO, which most likely arises from interactions between neurons of the central I pattern generator.
