Synthesis Models for Starburst Populations with Wolf-Rayet Stars

Abstract. The prospects of utilizing Wolf-Rayet populations in star-burst galaxies to infer the stellar content are reviewed. I discuss which Wolf-Rayet star features can be detected in an integrated stellar pop-ulation. Specic examples are given where the presence of Wolf-Rayet stars can help understand galaxy properties independent of the O-star population. I demonstrate how populations with small age spread, such as super star clusters, permit observational tests to distinguish between single-star and binary models to produce Wolf-Rayet stars. Dierent synthesis models for Wolf-Rayet populations are compared. Predictions for Wolf-Rayet properties vary dramatically between individual models. The current state of the models is such that a comparison with starburst populations is more useful for improving Wolf-Rayet atmosphere and evo-lution models than for deriving the star-formation history and the initial mass function. 1. Wolf-Rayet Signatures in Young Populations The central 30 Doradus region has the highest concentration of Wolf-Rayet (WR) stars in the LMC. Parker et al. (1995) classify 15 stars within 2000 (or 5 pc) of R136 as WR stars, including objects which may appear WR-like due to very dense winds (de Koter et al. 1997). This suggests that about 1 out of 10 ionizing stars around R136 is of WR type. The WR stars can be seen in an ultraviolet (UV) drift-scan spectrum of the integrated 30 Dor population obtained by Vacca et al. (1995). Their spectrum is reproduced in Fig. 1. The He II 1640 emission is from the WR population, demonstrating that WR signatures can be seen superimposed on the integrated spectrum of the OB population. While Fig. 1 suggests that WR features should be detectable, it is also clear that they will not be very prominent in comparison with O-star features. WR-and O-stars occupy roughly the same part of the Hertzsprung-Russell diagram, therefore they have similar L and Te , whereas WRs have lifetimes of only 10% those of O stars, and they tend to descend from more massive (less frequent) O stars. Taken together, this implies that any spectral tracer of WR stars that is present in O stars as well will not be useful to identify WR stars in an unresolved population