We investigate whether the satellite luminosity function (LF) of primary galaxies identified in the Sloan Digital Sky Survey (SDSS) depends on whether the host galaxy is in a filament or not. Isolated primary galaxies are identified in the SDSS spect
roscopic sample while potential satellites (that are up to 4 magnitudes fainter than their hosts) are searched for in the much deeper photometric sample. Filaments are constructed from the galaxy distribution by the Bisous process. Isolated primary galaxies are divided into two subsamples: those in filaments and those not in filaments. We examine the stacked mean satellite LF of both the filament and non-filament sample and find that, on average, the satellite LFs of galaxies in filaments is significantly higher than those of galaxies not in filaments. The filamentary environment can increases the abundance of the brightest satellites ($M_mathrm{sat.} < M_mathrm{prim.} + 2.0$), by a factor of $sim 2$ compared with non-filament isolated galaxies. This result is independent of primary galaxy magnitude although the satellite LF of galaxies in the faintest magnitude bin, is too noisy to determine if such a dependence exists. Since our filaments are extracted from a spectroscopic flux-limited sample, we consider the possibility that the difference in satellite LF is due to a redshift, colour or environmental bias, finding these to be insufficient to explain our result. The dependence of the satellite LF on the cosmic web suggests that the filamentary environment may have a strong effect on the efficiency of galaxy formation.
We investigate the luminosity functions (LFs) and projected number density profiles of galactic satellites around isolated primaries of different luminosities. We measure these quantities for model satellites placed into the Millennium and Millennium
II dark matter simulations by the GALFORM semi-analytic galaxy formation model for different bins of primary galaxy magnitude and we investigate their dependence on satellite luminosity. We compare our model predictions to the data of Guo et al. from the Sloan Digital Sky Survey Data Release 8 (SDSS DR8). First, we use a mock light-cone catalogue to verify that the method we used to count satellites in the SDSS DR8 is unbiased. We find that the radial distributions of model satellites are similar to those around comparable primary galaxies in the SDSS DR8, with only slight differences at low luminosities and small projected radii. However, when splitting the satellites by colour, the model and SDSS satellite systems no longer resemble one another, with many red model satellites, in contrast to the dominant blue fraction at similar luminosity in SDSS. The few model blue satellites are also significantly less centrally concentrated in the halo of their stacked primary than their SDSS counterparts. The implications of this result for the GALFORM model are discussed.
