VALIDATION OF THE EQUILIBRIUM MODEL FOR GALAXY EVOLUTION TO z ∼ 3 THROUGH MOLECULAR GAS AND DUST OBSERVATIONS OF LENSED STAR-FORMING GALAXIES

We combine IRAM Plateau de Bure Interferometer and Herschel PACS and SPIRE measurements to study the dust and gas contents of high-redshift star-forming galaxies. We present new observations for a sample of 17 lensed galaxies at z = 1.4-3.1, which allow us to directly probe the cold interstellar medium of normal star-forming galaxies with stellar masses of similar to 10(10) M-circle dot, a regime otherwise not (yet) accessible by individual detections in Herschel and molecular gas studies. The lensed galaxies are combined with reference samples of submillimeter and normal z similar to 1-2 star-forming galaxies with similar far-infrared photometry to study the gas and dust properties of galaxies in the SFR-M-*-redshift parameter space. The mean gas depletion timescale of main-sequence (MS) galaxies at z > 2 is measured to be only similar to 450 Myr, a factor of similar to 1.5 (similar to 5) shorter than at z = 1 (z = 0), in agreement with a (1+z)(-1) scaling. The mean gas mass fraction at z = 2.8 is 40% +/- 15% (44% after incompleteness correction), suggesting a flattening or even a reversal of the trend of increasing gas fractions with redshift recently observed up to z similar to 2. The depletion timescale and gas fractions of the z > 2 normal star-forming galaxies can be explained under the "equilibrium model" for galaxy evolution, in which the gas reservoir of galaxies is the primary driver of the redshift evolution of specific star formation rates. Due to their high star formation efficiencies and low metallicities, the z > 2 lensed galaxies have warm dust despite being located on the star formation MS. At fixed metallicity, they also have a gas-to-dust ratio 1.7 times larger than observed locally when using the same standard techniques, suggesting that applying the local calibration of the delta(GDR)-metallicity relation to infer the molecular gas mass of high-redshift galaxies may lead to systematic differences with CO-based estimates.