Lyman-break galaxies: high mass or low?

We reassess the hypothesis that Lyman-break galaxies (LBGs) at redshifts   z ∼ 3  mark the centres of the most massive dark matter haloes at that epoch. First we reanalyse the kinematic measurements of Pettini et al. and Erb et al. of the rest-frame optical emission lines of LBGs. We compare the distribution of the ratio of the rotation velocity to the central line width, against the expected distribution for galaxies with random inclination angles, modelled as singular isothermal spheres. The model fits the data well. On this basis we argue that the central line width provides a predictor of the circular velocity at a radius of several kpc. Assembling a larger sample of LBGs with measured line widths, we compare these results against the theoretical ΛCDM rotation curves of Mo, Mao & White, under the hypothesis that LBGs mark the centres of the most massive dark matter haloes. We find that the circular velocities are overpredicted by a substantial factor, which we estimate conservatively as  1.8 ± 0.4  . This indicates that the model is probably incorrect. The model of LBGs as relatively low-mass starburst systems, of Somerville, Primack & Faber, provides a good fit to the data.