Evidence Against Primordial Mass Segregation in Young Open Clusters: A Bayesian Multiscale Analysis of 3881 Gaia DR3 Clusters

Abstract The origin of mass segregation in open clusters—whether it is primordial (imprinted at birth) or dynamical (evolved over time)—remains a subject of intense debate. We present a comprehensive multi-scale analysis of 3881 open clusters using high-precision astrometry from Gaia DR3 based on a recent comprehensive catalog. To strictly control for observational biases, we implement a rigorous Bayesian deconvolution method that corrects for both spatial “Finger-of-God” elongations and kinematic measurement errors via posterior sampling. We evaluate mass segregation using three independent indicators: the spatial mass segregation ratio (Λ MSR ), the Q -parameter, and a kinematic segregation index ( Q kin ). Our results reveal no statistically significant evidence for primordial spatial mass segregation in very young clusters (<10 Myr). The spatial metrics are strictly consistent with the random expectation baseline ( Q 1 ≈ Q rand = 0.25; Λ MSR ≈ 1.0). Furthermore, we detect a signature of inverse kinematic mass segregation in the youngest systems ( Q kin ≈ 0.23 < Q rand ), indicating that massive stars possess systematically higher tangential velocities than low-mass stars. This kinematic “overheating” prevents massive stars from settling into the cluster center early on. We observe that classical mass segregation only emerges in clusters older than ∼100 Myr, consistent with two-body relaxation timescales. These findings strongly support the scenario where mass segregation is a dynamical product rather than a primordial feature of star formation.