Abstract

We examine the origins of the bimodality observed in the global properties of galaxies around a stellar mass of 3 x 10(10) M-circle dot by comparing the environmental dependencies of star formation for the giant and dwarf galaxy populations. The Sloan Digital Sky Survey DR4 spectroscopic data set is used to produce a sample of galaxies in the vicinity of the supercluster centered on the cluster A2199 at z = 0.03 that is greater than or similar to 90% complete to M-r(*) + 3.3. From this we measure global trends with environment for both giant (M-r -20 mag) and dwarf (-19 M-r -17.8 mag) subsamples using the luminosity-weighted mean stellar age and H alpha emission as independent measures of star formation history. The fraction of giant galaxies classified as old (t > 7 Gyr) or passive (EW [H alpha] = 4 angstrom) falls gradually from greater than or similar to 80% in the cluster cores to similar to 40% in field regions beyond 3-4(vir), as found in previous studies. In contrast, we find that the dwarf galaxy population shows a sharp transition at similar to 1(vir), from being predominantly old/ passive within the cluster, to outside where virtually all galaxies are forming stars and old/ passive galaxies are only found as satellites to more massive galaxies. These results imply fundamental differences in the evolution of giant and dwarf galaxies: whereas the star formation histories of giant galaxies are determined primarily by their merger history, star formation in dwarf galaxies is much more resilient to the effects of major mergers. Instead, dwarf galaxies become passive only once they become satellites within a more massive halo either by losing their halo gas reservoir to the host halo or through other environment- related processes such as galaxy harassment and ram pressure stripping.