The evolution of early-type galaxies selected by their spatial clustering

Aims. We present a new method that uses luminosity or stellar mass functions combined with clustering measurements to select samples of galaxies at different redshifts likely to follow a progenitor-to-descendant relationship. Because the method uses clustering information, we refer to galaxy samples selected this way as clustering-selected samples. We apply this method to infer the number of mergers during the evolution of MUSYC early-type galaxies (ETGs) from z similar or equal to 1 to the present-day. Methods. The method consists in using clustering information to infer the typical dark-matter halo mass of the hosts of the selected progenitor galaxies. Using Lambda CDM predictions, it is then possible to follow these haloes to a later time where the sample of descendants will be that with the clustering of these descendant haloes. Results. This technique shows that ETGs at a given redshift evolve into brighter galaxies at lower redshifts (considering rest-frame, passively evolved optical luminosities). This indicates that the stellar mass of these galaxies increases with time and that, in principle, a stellar mass selection at different redshifts does not provide samples of galaxies in a progenitor-descendant relationship. Conclusions. The comparison between high-redshift ETGs and their likely descendants at z = 0 points to a higher number density for the progenitors by a factor 5.5 +/- 4.0, implying the need for mergers to decrease their number density by today. Because the luminosity densities of progenitors and descendants are consistent, our results show no need for significant star-formation in ETGs since z = 1, which indicates that the needed mergers are dry, i.e. gas free.