2015MNRAS.446..169M

We present an analysis of the clustering of high-redshift galaxies in the recently completed 94 deg2 Spitzer South Pole Telescope Deep-Field survey. Applying flux and colour cuts to the mid-infrared photometry efficiently selects galaxies at z ∼ 1.5 in the stellar mass range 10¹⁰-10¹¹M_☉, making this sample the largest used so far to study such a distant population. We measure the angular correlation function in different flux-limited samples at scales >6arcsec (corresponding to physical distances >0.05 Mpc) and thereby map the one- and two-halo contributions to the clustering. We fit halo occupation distributions and determine how the central galaxy's stellar mass and satellite occupation depend on the halo mass. We measure a prominent peak in the stellar-to-halo mass ratio at a halo mass of log(Mhalo/M_☉) = 12.44±0.08, 4.5 times higher than the z = 0 value. This supports the idea of an evolving mass threshold above which star formation is quenched. We estimate the large-scale bias in the range bg = 2-4 and the satellite fraction to be fsat ∼ 0.2, showing a clear evolution compared to z = 0. We also find that, above a given stellar mass limit, the fraction of galaxies that are in similar mass pairs is higher at z = 1.5 than at z = 0. In addition, we measure that this fraction mildly increases with the stellar mass limit at z = 1.5, which is the opposite of the behaviour seen at low redshift.