Observational evidence for the coevolution of galaxy mergers, quasars, and the blue/red galaxy transition

We compile a number of observations to estimate the time-averaged rate of formation or buildup of red sequence galaxies, as a function of mass and redshift. Comparing this with the mass functions of mergers and quasar hosts, and independently comparing their clustering properties as a function of redshift, we find that these populations trace the same mass distribution, with similar evolution, at redshifts 0 < z less than or similar to 1.5. Knowing one of the quasar, merger, or elliptical mass/luminosity functions, it is possible to predict the others. Allowing for greater model dependence, we compare the rate of early-type buildup with the implied merger and quasar triggering rates as a function of mass and redshift and find agreement. Over this redshift range, observed merger fractions can account for the entire bright quasar luminosity function and buildup of the red sequence at all but the highest masses at low redshift (greater than or similar to 10(11) M-circle dot at z less than or similar to 0.3) where "dry'' mergers appear to dominate. This supports a necessary prediction of theories where mergers between gas-rich galaxies produce ellipticals with an associated phase of quasar activity, after which the remnant becomes red. These populations trace a similar characteristic transition mass, possibly reflecting the mass above which the elliptical population is mostly (greater than or similar to 50%) assembled at a given redshift, which increases with redshift over the observed range in a manner consistent with suggestions that cosmic down sizing may apply to red galaxy assembly as well as star formation. These mass distributions as a function of redshift do not uniformly trace the all/red/blue galaxy population, ruling out models in which quasar activity is generically associated with star formation or is long lived in "old'' systems.