2007ApJ...668..756V

We study the mass-to-light ratio (M/L) evolution of early-type galaxies using dynamical modeling of resolved internal kinematics. This makes fewer assumptions than fundamental plane (FP) studies and provides a powerful new approach for studying galaxy evolution. We focus on the sample of 25 galaxies in clusters at z~0.5 modeled in Paper I. For comparison, we compile and homogenize M/L literature data for 60 nearby galaxies that were modeled in comparable detail. The nearby sample obeys log(M/L)_B=Z+Slog(σ_eff/200 km/s), where Z=0.896±0.010, S=0.992±0.054, and σ_eff is the effective velocity dispersion. The z~0.5 sample follows a similar relation, but with lower zero point. The implied M/L evolution is Δlog(M/L)/Δz=-0.457±0.046(random)±0.078(systematic), consistent with passive evolution following high-redshift formation. This agrees with the FP results for this sample by van Dokkum & van der Marel, and confirms that FP evolution tracks M/L evolution, which is an important verification of the assumptions that underlie FP studies. However, while we find more FP evolution for galaxies of low σ_eff(or low mass), the dynamical M/L evolution shows little correlation with σ_eff. We argue that this difference can be plausibly attributed to a combination of two effects: (1) evolution in structural galaxy properties other than M/L, and (2) the neglect of rotational support in studies of FP evolution. The results leave the question open as to whether the low-mass galaxies in the sample have younger populations than the high-mass galaxies. This highlights the general importance in the study of population ages for complementing dynamical measurements with broadband colors or spectroscopic population diagnostics.