The accretion of gas on to galaxies as traced by their satellites

We have compiled a large sample of isolated central galaxies from the Sloan Digital Sky Survey, which do not have a neighbour of comparable brightness within a projected distance of 1 Mpc. We use the colours, luminosities and surface brightnesses of satellite galaxies in the vicinity of these objects to estimate their neutral gas content and to derive the average total mass of HI gas contained in the satellites as a function of the projected radius from the primary. Recent calibrations of merging time-scales from N-body simulations are used to estimate the rate at which this gas will accrete on to the primary galaxies. Our estimated accretion rate falls short of that needed to maintain the observed level of star formation in the primaries by nearly two orders of magnitude. Nevertheless, there are strong correlations between the total mass of gas in satellites and the colours and specific star formation rates of primary galaxies of all stellar masses. The correlations are much weaker, if we consider the total stellar mass in satellites, rather than their total gas mass. We ask why star formation in the primaries should be correlated with gas contained in satellites at projected separations of an Mpc or more, well outside the virial radius of the dark matter haloes of these systems. We suggest that gas-rich satellites trace an underlying reservoir of ionized gas that is accreted continuously and that provides fuel for the ongoing star formation in galaxies in the local Universe.