3.2 Star Formation History
The most important parameter of the chemical enrichment is expected to be the star formation history (SFH). Since massive stars dominate the production of most elements, the metal production rate will to a first approximation be directly proportional to the star formation rate. Likewise, in a simple chemical evolution picture, the metallicity is a function of the fraction of gas turned into stars.
Star formation (SF) seems to occur in different modes: one being relatively undramatic with continuous star formation at a regular pace while in ``bursting'' galaxies, star formation may be dominated by a small number of short intense bursts of star formation separated by extremely long intervals of time. While the latter scenario is mainly observed in blue compact galaxies it is believed to occur also in dwarf and giant elliptical galaxies. An intermediate picture is that of ``gasping'' SF, characterised by extended star formation episodes separated by moderately long periods of less active SF, which probably is the most realistic picture for many dwarf irregular galaxies. While a starburst produces a lot of metals, the supernovae and stellar winds may eject gas into the intergalactic medium on short time scales. If, on the other hand, the SF is continuous, the energetic feedback from dying stars will have less influence on the ISM. Different scenarii for SF regulation have been implemented with stochastic self-propagating star formation, self regulated star formation and also a gas density threshold. Moreover interactions, mergers and stripping may play an important rôle in regulating the SF.
Discrete star-bursting behaviour in dwarf galaxies may strongly affect the abundance ratio of elements such as N, Fe and C which partly come from longer lived intermediate mass stars as compared to O (Garnett 1990, Gilmore and Wyse 1991, Richer and McCall 1995). In particular, this could explain the tendency of C/O and Fe/O to increase with O/H especially at low O/H, but this awaits more accurate measurements of these ratios in extremely metal-poor galaxies. Moreover, since the net yield for many elements appears to be metallicity dependent (Maeder 1992), element production by subsequent generations of stars will depend on how fast the ISM is enriched and hence the star formation history. In addition, variations in element ratios could affect nucleosynthesis and thus the net yields for a given stellar generation.
A last note concerns ``starbursts'': This notion is frequently used in the literature to describe regions/galaxies with varying degree of active star formation. A proper definition of a starburst is that it involves an unsustainably high star formation rate (SFR) in terms of the gas consumption timescale or the timescale to build up the observed stellar mass (i.e., the time averaged SFR is much lower than the present). Many galaxies have SFRs fluctuating with time, but this does not necessarily imply that the SFR is unsustainable over a Hubble time.