The evolution of early-type galaxies in clusters from z ~ 0.8 to z ~ 0: the ellipticity distribution and the morphological mix

Abstract

We present the ellipticity distribution and its evolution for early-type galaxies in clusters from z~ 0.8 to the current epoch, based on the WIde-field Nearby Galaxy-cluster Survey (0.04 ≤ z ≤ 0.07) and the ESO Distant Cluster Survey (0.4 ≤ z ≤ 0.8). We first investigate a mass-limited sample and we find that above a fixed mass limit (M_(*) ≥ 10^(10.2) M_⊙), the ellipticity (ε) distribution of early-type galaxies notably evolves with redshift. In the local Universe, there are proportionally more galaxies with higher ellipticity, hence flatter, than in distant clusters. This evolution is due partly to the change in the mass distribution and mainly to the change in the morphological mix with z among the early types, the fraction of ellipticals goes from ~70 per cent at high z to ~40 per cent at low z). Analysing separately the ellipticity distribution of the different morphological types, we find no evolution both for ellipticals and for S0s. However, for ellipticals a change with redshift in the median value of the distributions is detected. This is due to a larger population of very round (ε < 0.05) elliptical galaxies at low z. In order to compare our finding to previous studies, we also assemble a magnitude-'delimited' sample that consists of early-type galaxies on the red sequence with −19.3 > M_(B)+ 1.208z > −21. Analysing this sample, we do not recover exactly the same results as the mass-limited sample. This indicates that the selection criteria are crucial to characterize the galactic properties: the choice of the magnitude-'delimited' sample implies the loss of many less-massive galaxies and so it biases the final conclusions. Moreover, although we are adopting the same selection criteria, our results in the magnitude-'delimited' sample are also not in agreement with those of Holden et al. This is due to the fact that our and their low-zsamples have a different magnitude distribution because the Holden et al. sample suffers from incompleteness at faint magnitudes.

Files

Additional details