Comparing dynamical and stellar population mass-to-light ratio estimates

We investigate the mass-to-light ratios of stellar populations as predicted by stellar population synthesis codes and compare those to dynamical/gravitational measurements. In Bell & de Jong (2001) we showed that population synthesis models predict a tight relation between the color and mass-to-light ratio of a stellar population. The normalization of this relation depends critically on the shape of the stellar IMF at the low-mass end. These faint stars contribute significantly to the mass, but insignificantly to the luminosity and color of a stellar system. In Bell & de Jong (2001) we used rotation curves to normalize the relation, but rotation curves provide only an upper limit to the stellar masses in a system. Here we compare stellar and dynamical masses for a range of stellar systems in order to constrain the mass normalization of stellar population models. We find that the normalization of Bell & de Jong (2001) should be lowered by about 0.05-0.1 dex in M/L. This is consistent with a Kroupa (2001), Chabrier (2003), or a Kennicutt (1983) IMF, but does not leave much room for other unseen components