The Clustering of Galaxy Groups: Dependence on Mass and Other Properties

We investigate the clustering of galaxy groups and clusters in the SDSS using the Berlind et al. (2006) group sample, which is designed to identify galaxy systems that each occupy a single dark matter halo. We estimate group masses from their abundances, and measure their relative large-scale bias as a function of mass. Our measurements are in agreement with the theoretical halo bias function, given a standard LCDM cosmological model, and they tend to favor a low value of the power spectrum amplitude sigma_8. We search for a residual dependence of clustering on other group properties at fixed mass, and find the strongest signal for central galaxy color in high mass groups. Massive groups with less red central galaxies are more biased on large scales than similar mass groups with redder central galaxies. We show that this effect is unlikely to be caused by errors in our mass estimates, and is most likely observational evidence of recent theoretical findings that halo bias depends on a ``second parameter'' other than mass, such as age or concentration. To compare with the data, we study the bias of massive halos in N-body simulations and quantify the strength of the relation between halo bias and concentration at fixed mass. In addition to confirming a non-trivial prediction of the LCDM cosmological model, these results have important implications for the role that environment plays in shaping galaxy properties.