The diminishing importance of major galaxy mergers at higher redshifts

Abstract: Using mass-selected galaxy samples from deep multiwavelength data we investigate the incidence of close galaxy pairs between z=0.4-2. Many such close pairs will eventually merge, and the pair fraction is therefore related to the merger rate. Over this redshift range the mean pair fraction is essentially constant (evolving as f_pair (1+z)^{-0.4 +/- 0.6}) with about 6+/-1% of massive galaxies having a 1:4 or greater companion within 30h^-1 kpc. Assuming the timescale over which pairs merge is not a strong function of redshift, this implies a similarly constant merger rate (per unit time) out to z=2. Since about three times as much cosmic time passes at z<1 as between z=1-2, this implies that correspondingly more mergers occur in the low-redshift universe. When minor companions (1:10 mass ratio or greater) are included, the pair fraction increases to ~20% and still does not evolve strongly with redshift. We also use a rest-frame color criterion to select pairs containing only quiescent galaxies (major "dry merger" progenitors), and find them to be similarly rare and constant with 4-7% of massive quiescent galaxies exhibiting a nearby companion. Thus, even though other studies find major mergers to be relatively uncommon since z=1, our results suggest that few additional mergers occur in the 1<z<2 range and other mechanisms may be required to explain the mass and size growth of galaxies over this epoch.