“…Thus, our results demonstrate that the position of DSFGs in the IRX − β plane does not necessarily imply that they are short-lived starbursts, contrary to the claim of Casey et al (2014); instead, DSFGs well above the M99 relation can be steadily star-forming, massive, gas-rich disks as long as they are sufficiently dust-obscured, which is a natural outcome for high-gas-fraction, metal-enriched galaxies. Consequently, high-z DSFGs may represent a heterogeneous population of massive, gas-rich, steadily starforming galaxies, merger-induced starbursts, and obscured AGN (e.g., Hopkins et al 2010;Hayward et al 2011Hayward et al , 2012Hayward et al , 2013bda Cunha et al 2015;Koprowski et al 2016). We do find that all of the simulated galaxies with L IR > 10 13 L (see also Hopkins et al 2010) and, of the lower-luminosity simulated galaxies, those with the highest IRX values (log IRX 2.5) are exclusively mergers, and in the simulations, these HyLIRGs are typically powered by combination of merger-induced starbursts and obscured AGN (Roebuck et al, submitted).…”