Observational evidence of galaxy assembly bias

TL;DR

This study provides observational evidence that the clustering of luminous red galaxies depends on their stellar-mass assembly history, supporting the existence of galaxy assembly bias in large-scale structure formation.

Contribution

It presents the first direct observational evidence linking galaxy assembly history with clustering properties, demonstrating galaxy assembly bias in a homogeneous galaxy population.

Findings

Fast-growing LRGs assemble most of their mass by z~5

Slow-growing LRGs reach similar mass at z~1.5

Fast-growing LRGs are more strongly clustered and in denser environments

Abstract

We analyze the spectra of 300,000 luminous red galaxies (LRGs) with stellar masses $M_* \gtrsim 10^{11} M_{\odot}$ from the SDSS-III Baryon Oscillation Spectroscopic Survey (BOSS). By studying their star-formation histories, we find two main evolutionary paths converging into the same quiescent galaxy population at $z\sim0.55$. Fast-growing LRGs assemble $80\%$ of their stellar mass very early on ($z\sim5$), whereas slow-growing LRGs reach the same evolutionary state at $z\sim1.5$. Further investigation reveals that their clustering properties on scales of $\sim$1-30 Mpc are, at a high level of significance, also different. Fast-growing LRGs are found to be more strongly clustered and reside in overall denser large-scale structure environments than slow-growing systems, for a given stellar-mass threshold. Our results imply a dependence of clustering on stellar-mass assembly history (naturally connected to the mass-formation history of the corresponding halos) for a homogeneous population of similar mass and color, which constitutes a strong observational evidence of galaxy assembly bias.