Cluster galaxies die hard

We investigate how the specific star formation rates of galaxies of different masses depend on cluster-centric radius and on the central/satellite dichotomy in both field and cluster environments. Recent data from a variety of sources, including the cluster catalogue of von der Linden et al., are compared to the semi-analytic models of De Lucia & Blaizot. We find that these models predict too many passive satellite galaxies in clusters, too few passive central galaxies with low stellar masses and too many passive central galaxies with high masses. We then outline a series of modifications to the model necessary to solve these problems: (a) instead of instantaneous stripping of the external gas reservoir after a galaxy becomes a satellite, the gas supply is assumed to decrease at the same rate that the surrounding halo loses mass due to tidal stripping and (b) the active galactic nuclei (AGN) feedback efficiency is lowered to bring the fraction of massive passive centrals in better agreement with the data. We also allow for radio mode AGN feedback in satellite galaxies. (c) We assume that satellite galaxies residing in host haloes with masses below 10¹² h^-1 M_⊙ do not undergo any stripping. We highlight the fact that in low-mass galaxies, the external reservoir is composed primarily of gas that has been expelled from the galactic disc by supernovae-driven winds. This gas must remain available as a future reservoir for star formation, even in satellite galaxies. Finally, we present a simple recipe for the stripping of gas and dark matter in satellites that can be used in models where subhalo evolution is not followed in detail.