The diverse evolutionary paths of simulated high-

Abstract

Massive quiescent galaxies have much smaller physical sizes at high redshift than today. The strong evolution of galaxy size may be caused by progenitor bias, major and minor mergers, adiabatic expansion, and/or renewed star formation, but it is difficult to test these theories observationally. Herein, we select a sample of 35 massive, compact galaxies (M* = 1–3 × 10[superscript 11] M⊙, M*/R[superscript 1.5] > 10[superscript 10.5] M⊙/kpc[superscript 1.5) at z = 2 in the cosmological hydrodynamical simulation Illustris and trace them forwards to z = 0 to uncover their evolution and identify their descendants. By z = 0, the original factor of 3 difference in stellar mass spreads to a factor of 20. The dark matter halo masses similarly spread from a factor of 5 to 40. The galaxies’ evolutionary paths are diverse: about half acquire an ex situ envelope and are the core of a more massive descendant, a third survive undisturbed and gain very little mass, 15 per cent are consumed in a merger with a more massive galaxy, and a small remainder are thoroughly mixed by major mergers. The galaxies grow in size as well as mass, and only ∼10 per cent remain compact by z = 0. The majority of the size growth is driven by the acquisition of ex situ mass. The most massive galaxies at z = 0 are the most likely to have compact progenitors, but this trend possesses significant dispersion which precludes a direct linkage to compact galaxies at z = 2. The compact galaxies’ merger rates are influenced by their z = 2 environments, so that isolated or satellite compact galaxies (which are protected from mergers) are the most likely to survive to the present day.