Ultramassive dense early-type galaxies: Velocity dispersions and number density evolution since z = 1.6

Abstract

Aims: We investigate the stellar mass assembly history of ultramassive (M⋆≳ 10^11 M☉) dense (Σ = M⋆/2πRe^2 > 2500 M☉ pc^-2) early-type galaxies (ETGs, elliptical and spheroidal galaxies) selected on basis of visual classification over the last 9 Gyr. Methods: We traced the evolution of the comoving number density ρ of ultramassive dense ETGs and compared their structural (effective radius Re and stellar mass M⋆) and dynamical (velocity dispersion σe) parameters over the redshift range 0 < z < 1.6. We derived the number density ρ at 1.6 <z< 1 from the MUNICS and GOODS-South surveys, while we took advantage of the COSMOS spectroscopic survey to probe the intermediate redshift range [0.2-1.0]. We derived the number density of ultramassive dense local ETGs from the SDSS sample taking all of the selection bias affecting the spectroscopic sample into account. To compare the dynamical and structural parameters, we collected a sample of 11 ultramassive dense ETGs at 1.2 < z < 1.6 for which velocity dispersion measurements are available. For four of these ETGs (plus one at z = 1.91), we present previously unpublished estimates of velocity dispersion, based on optical VLT-FORS2 spectra. We probe the intermediate redshift range (0.2 ≲ z ≲ 0.9) and the local Universe with different ETGs samples. <BR /> Results: We find that the comoving number density of ultramassive dense ETGs evolves with z as ρ(z) ∝ (1 + z)^(0.3 ± 0.8) implying a decrease of ~25% of the population of ultramassive dense ETGs since z = 1.6. By comparing the structural and dynamical properties of high-z ultramassive dense ETGs over the range 0 ≲ z < 1.6 in the [Re , M⋆, σe] plane, we find that all of the ETGs of the high-z sample have counterparts with similar properties in the local Universe. This implies either that the majority (~70%) of ultramassive dense ETGs already completed the assembly and shaping at ⟨ z ⟩ = 1.4, or that, if a significant portion of dense ETGs evolves in size, new ultramassive dense ETGs must form at z < 1.5 to maintain their number density at almost constant. The difficulty in identify good progenitors for these new dense ETGs at z ≲ 1.5 and the stellar populations properties of local ultramassive dense ETGs point towards the first hypothesis. In this case, the ultramassive dense galaxies missing in the local Universe could have joined, in the last 9 Gyr, the so colled non-dense ETGs population through minor mergers, thus contributing to mean size growth. In any case, the comparison between their number density and the number density of the whole population of ultramassive ETGs relegates their contribution to the mean size evolution to a secondary process.