Annu. Rev. Astron. Astrophys. 1994. 32: 531-590 Copyright © 1994 by Annual Reviews. All rights reserved

4. POPULATION III STARS

There is some confusion in the literature because the term "Population III" has been used in two distinct contexts. It has been applied to describe: 1. the stars that generate the first metals; and 2. the stars hypothesized to provide the dark matter in galactic halos. In either case, the stars only warrant a special name if they are definitely distinct from Population II stars, i.e. if they form at a distinct epoch or if the initial mass function (IMF) of the first stars is bimodal (with distinct populations of stars forming in different locations). We will see that this may not be the case for stars of type I, but it probably is for stars of type 2. If one requires both kinds of "Population III" stars, it is not obvious which ones come first. One could envisage situations in which the dark objects form before, after, or contemporaneously with the stars that make the first metals.

4.1 Population III as the First Metal Producers

Stars of type 1 must exist because heavy elements can only be generated through stellar nucleosynthesis. However, the most natural assumption is that these are merely the ones at the high mass end of the Population II mass spectrum, since in this case they would generate the first metals because they evolved fastest. This is already sufficient to explain most of the abundance characteristics of Population I and II stars (Truran 1984, Wheeler et al 1989, Rana 1991, Pagel 1992). At one point, there appeared to be a metallicity cut-off of order 10^-5 below which no stars were found (Bond 1981); this suggested that the first stars were more massive than those forming today. However, the evidence for the cut-off has now gone away: Beers et al (1992) find that the Z distribution for Population II stars extends well below 10^-6, and there exists one object with Z = 6 × 10^-7 (Bessel & Norris 1984). In any case, the number of low-Z objects is not necessarily incompatible with the assumption that the IMF has always been the same (Pagel 1987), so there is no obvious reason for supposing that the first stars were qualitatively different from Population II. However, one cannot be sure that there are not abundance anomalies at some level (cf Kajino et al 1990, Suntzeff 1992). This is important because, if the dark baryons are in the remnants of massive stars, one might expect some nucleosynthetic consequences.