Abstract
I show that recent observations of r-process abundances in metal-poor stars are difficult to explain if neutron star mergers (NSMs) are the major r-process sources. In contrast, such observations and meteoritic data on Hf182 and I129 in the early solar system support a self-consistent picture of r-process enrichment by supernovae (SNe). While further theoretical studies of r-process production and enrichment are needed for both SNe and NSMs, I emphasize two possible direct observational tests of the SN r-process model: gamma rays from decay of r-process nuclei in SN remnants and surface contamination of the companion by SN r-process ejecta in binaries.Comment: 5 pages, to appear in ApJ