We address the temporal organization of circadian and ultradian rhythms (URs), crucial for understanding biological timekeeping in behavior, physiology, metabolism, and alignment with geophysical time. Using a newly developed five-step wavelet-based approach to analyze high-resolution time series of metabolism in yeast cultures and spontaneous movement, metabolism, and feeding behavior in mice, rats, and quails, we describe a dynamically coherent pattern of rhythms spanning over a broad range of temporal scales (hours to minutes). The dynamic pattern found shares key features among the four, evolutionary distant, species analyzed. Specifically, a branching appearance given by splitting periods from 24h into 12h, 8h and below in mammalian and avian species, or from 14 h down to 0.07h in yeast. Scale-free fluctuations with long-range correlations prevail below ~4 h. Synthetic time series modeling support a scenario of coexisting behavioral rhythms, with circadian and URs at the center of the emergent pattern observed.