2007ApJS..173..404B

We select far-infrared (FIR: 60 µm) and far-ultraviolet (FUV: 530 Å) samples of nearby galaxies in order to discuss the biases encountered by monochromatic surveys (FIR or FUV). Very different volumes are sampled by each selection, and much care is taken to apply volume corrections to all the analyses. The distributions of the bolometric luminosity of young stars are compared for both samples: they are found to be consistent with each other for galaxies of intermediate luminosities, but some differences are found for high (>5x10¹⁰ L_☉) luminosities. The shallowness of the IRAS survey prevents us from securing a comparison at low luminosities (<2x10⁹ L_☉). The ratio of the total infrared (TIR) luminosity to the FUV luminosity is found to increase with the bolometric luminosity in a similar way for both samples up to 5x10¹⁰ L_☉. Brighter galaxies are found to have a different behavior according to their selection: the L_TIR/L_FUV ratio of the FUV-selected galaxies brighter than 5x10¹⁰ L_☉ reaches a plateau, whereas L_TIR/L_FUVcontinues to increase with the luminosity of bright galaxies selected in FIR. The volume-averaged specific star formation rate (SFR per unit galaxy stellar mass, SSFR) is found to decrease toward massive galaxies within each selection. The mean values of the SSFR are found to be larger than those measured for optical and NIR-selected samples over the whole mass range for the FIR selection, and for masses larger than 10¹⁰ M_☉ for the FUV selection. Luminous and massive galaxies selected in FIR appear as active as galaxies with similar characteristics detected at z∼0.7.