Before the launch of the Kepler Space Telescope, models of low-mass planet formation predicted that convergent type I migration would often produce systems of low-mass planets in low-order mean-motion resonances. Instead, Kepler discovered that systems of small planets frequently have period ratios larger than those associated with mean-motion resonances and rarely have period ratios smaller than those associated with mean-motion resonances. Both short-timescale processes related to the formation or early evolution of planetary systems and long-timescale secular processes have been proposed as explanations for these observations. Using a thin disk stellar population's Galactic velocity dispersion as a relative age proxy, we find that Kepler-discovered multiple-planet systems with at least one planet pair near a period ratio suggestive of a second-order mean-motion resonance have a colder Galactic velocity dispersion and are therefore younger than both single-transiting and multiple-planet systems that lack planet pairs consistent with mean-motion resonances. We argue that a nontidal secular process with a characteristic timescale no less than a few hundred Myr is responsible for moving systems of low-mass planets away from second-order mean-motion resonances. Among systems with at least one planet pair near a period ratio suggestive of a first-order mean-motion resonance, only the population of systems likely affected by tidal dissipation inside their innermost planets has a small Galactic velocity dispersion and is therefore young. We predict that period ratios suggestive of mean-motion resonances are more common in young systems with 10Myr<~{tau}<~100Myr and become less common as planetary systems age.

Cone search capability for table J/AJ/167/55/table1 (Separations from resonances in libration widths)

Cone search capability for table J/AJ/167/55/table2 (*Separations from resonance in units of |{delta}_res| (Equation 6))