1998PASP..110..888W

Model and observed spectral feature indices indicate that, in galaxies of all types with velocity dispersions larger than about 225 km.s^–1, [Mg/Fe] progressively drifts greater than zero, until it reaches about 0.3 in the largest ellipticals. For elliptical galaxies that have absorption-line data for more elements, the abundances of Na and N elevate in a similar fashion, relative to both Ca and Fe ([Ca/Fe] ~ 0; [Na, N/Fe] progressively > 0). Titanium may also share this apparently generic light-element behavior.

The abundance pattern in elliptical galaxies matches neither the disk, the halo, nor the bulge of our own Galaxy, although the bulge appears to come the closest. N is depleted in the Galactic bulge but elevated in M31, M31's metal-rich globular clusters, and large elliptical galaxies. If all measured elements are considered, the abundance patterns in our own Galaxy and in external galaxies requires the presence of at least three sources of chemical enrichment whose relative contributions can vary from environment to environment. These three sources of enrichment may correspond to Type Ia supernovae (SNe), Type II SNe, and N-rich mass loss from intermediate-mass stars, but the behavior of O, Ca, Si, Sc, V, and Ti in the Galactic bulge, disk, and halo seems to require at least one additional supernova flavor.

Abundance ratio effects represent a barrier to the estimation of mean ages from integrated light, a barrier that is greater than that of isochrone error. Isochrone grids allowing for the variation of individual elements are needed, but relative changes in isochrone temperatures computed as a function of abundance pattern need to be accurate to roughly 7 K if 5% age estimates are desired.