Abstract

We present an analysis of optical spectroscopically identified active galactic nuclei (AGN) down to a cluster magnitude of M-star + 1 in a sample of six self-similar Sloan Digital Sky Survey galaxy clusters at z similar to 0.07. These clusters are specifically selected to lack significant substructure at bright limits in their central regions so that we are largely able to eliminate the local action of merging clusters on the frequency of AGN. We demonstrate that the AGN fraction increases significantly from the cluster centre to 1.5R(virial), but tails off at larger radii. If only comparing the cluster core region to regions at similar to 2R(virial), no significant variation would be found. We compute the AGN fraction by mass and show that massive galaxies (log(stellarmass) >10.7) are host to a systematically higher fraction of AGN than lower mass galaxies at all radii from the cluster centre. We attribute this deficit of AGN in the cluster centre to the changing mix of galaxy types with radius. We use the WHAN diagnostic to separate weak AGN from 'retired' galaxies in which the main ionization mechanism comes from old stellar populations. These retired AGN are found at all radii, while the mass effect is much more pronounced: we find that massive galaxies are more likely to be in the retired class. Further, we show that our AGN have no special position inside galaxy clusters - they are neither preferentially located in the infall regions nor situated at local maxima of galaxy density as measured with Sigma(5). However, we find that the most powerful AGN (with [O III] equivalent widths -10 angstrom) reside at significant velocity offsets in the cluster, and this brings our analysis into agreement with previous work on X-ray-selected AGN. Our results suggest that if interactions with other galaxies are responsible for triggering AGN activity, the time lag between trigger and AGN enhancement must be sufficiently long to obfuscate the encounter site and wipe out the local galaxy density signal.