2012A&A...546A.118V

How mass assembly occurs in galaxies and which process(es) contribute to this activity are among the most highly debated questions in galaxy formation and evolution theories. This has motivated our survey MASSIV (Mass Assembly Survey with SINFONI in VVDS) of 0.9<z<1.9 star-forming galaxies selected from the purely flux-limited VVDS redshift survey. We evaluate the characteristic size and stellar mass of 46 MASSIV galaxies at 1<z<1.6 and use the internal dynamics obtained with the SINFONI integral field spectrograph mounted at the Very Large Telescope, to derive the stellar mass-size-velocity relations. We use the Kennicutt-Schmidt formulation to estimate the gas content and compute its contribution to the total baryonic mass in MASSIV galaxies. For the first time, we derive the relations between galaxy size, mass, and internal velocity, and the baryonic Tully-Fisher relation, from a statistically representative sample of star-forming galaxies at 1<z<1.6. We find a dynamical mass that agrees with those of rotating galaxies containing a gas fraction of ∼20%, that is perfectly consistent with the content derived using the Kennicutt-Schmidt formulation and corresponds to the expected evolution. Non-rotating galaxies have more compact sizes for their stellar component, and are less massive than rotators, but do not have statistically different sizes for their gas-component. Discs of ionized gas have irregular, clumpy distributions, but the simplistic assumption of exponential profiles is verified. We measure a marginal evolution in the size-stellar mass and size-velocity relations in which discs become evenly smaller with cosmic time at fixed stellar mass or velocity, and are less massive at a given velocity than in the local Universe. This result is inconsistent with previous reports of an abnormal evolution in the galactic spin. The scatter in the Tully-Fisher relation is smaller when we introduce the S₀₅ index, which we interpret as evidence of an increase in the contribution to galactic kinematics of turbulent motions with cosmic time. We report a persistently large scatter for rotators in our relations, that we suggest is intrinsic, and possibly caused by complex physical mechanism(s) at work in our stellar mass/luminosity regime and redshift range. Our results consistently point towards a mild, net evolution of these relations, comparable to those predicted by cosmological simulations of disc formation. In a conflictual picture where earlier studies reported discrepant results, we place on firmer ground the measurement of a lack of any large evolution of the fundamental relations of star-forming galaxies in at least the past 8Gyr and evidence that dark halo is strongly coupled with galactic spectrophotometric properties.