The CO Tully-Fisher relation and implications for the host galaxies of high-redshift quasars

The integrated line width derived from CO spectroscopy provides a powerful tool to study the internal kinematics of extragalactic objects, including quasars at high redshift, provided that the observed line width can be properly translated to more conventionally used kinematical parameters of galaxies. We show, through the construction of a K-s-band CO Tully-Fisher relation for nearby galaxies spanning a wide range in infrared luminosity, that the CO line width measured at 20% of the peak intensity, when corrected for inclination and other effects, successfully recovers the maximum rotation velocity of the disk. The line width at 50% of the peak intensity performs much more poorly, in large part because CO lines have a wide range of profiles, which are shown to vary systematically with infrared luminosity. We present a practical prescription for converting observed CO line widths into the stellar velocity dispersion of the bulge (sigma(*)) and then apply it to a sample of low-redshift (z P 0.2) and high-redshift (1.4 less than or similar to z less than or similar to 6.4) quasars to study their host galaxies. Nearby quasars roughly fall on the correlation between black hole mass and bulge stellar velocity dispersion established for inactive galaxies, but the host galaxies of the high-z quasars systematically deviate from the local M-center dot-sigma(*) relation. At a given sigma(*), high-z quasars have black hole masses larger by a factor of similar to 4 relative to local galaxies, suggesting that early in the life cycle of galaxies the development of the bulge lags behind the growth of the central black hole. An alternative explanation for these observations, which currently cannot be ruled out rigorously, is that high-redshift quasars are preferentially viewed at face-on orientations.