The history of the IMF in starburst regions is reviewed. The IMFs are no longer believed to be top-heavy, although some superstar clusters, whether in starburst regions or not, could be.
Introduction: History of Starburst IMFsEarly starburst observations suggested the luminous mass from young massive stars is comparable to the dynamical mass from the rotation curve (see reviews in Telesco 1988; Scalo 1990; Zinnecker 1996; Leitherer 1999). This implied there was a deficit in low mass stars. The nearby starburst galaxy, M82, was one of the best cases. Rieke et al. (1980Rieke et al. ( , 1993 modeled M82's gas mass, luminous star mass, rotation curve mass, 2.2 µm flux, Lyman continuum flux, CO index for supergiants, and the Brα/Brγ ratio, giving an extinction A V = 25 mag. They concluded that only the usual IMF models with lower mass limits M L > 3 − 6 M ⊙ worked. Kronberg et al. (1985) used the Lyman continuum flux from radio emission in M82 to determine the star formation rate, and used the total gas mass combined with an efficiency estimate to conclude that the IMF is top-heavy compared to the Miller & Scalo (1979) IMF. Bernlohr (1992 fit the same M82 properties as Rieke et al., plus the heavy element abundance and FIR line ratios, and concluded that either the IMF slope is shallower than the Scalo (1986) IMF by 1, or there is a lower mass cutoff greater than 1.5-2