In an effort to better understand how the properties of star formation in starburst galaxies depend on various environmental parameters, I present a comparison between two well-known WR galaxies: the interacting galaxy system NGC 1741 in the Hickson Compact Group 31, and the dwarf galaxy He2-10. The high spatial resolution of HST has allowed identification of a large number of starburst knots, or ‘super star clusters’ in these WR galaxies. Broad-band photometry and the latest stellar synthesis models are used to estimate the ages and masses of the super star clusters. The properties of the clusters are then used to compare and contrast the overall star-formation histories in the two WR galaxies.

We present results of X-ray spectroscopy of the prototypical Wolf-Rayet galaxy Henize 2–10 with ASCA. We find that the X-ray spectrum is best described by a model with soft (kT ≃ 0.7 keV) and hard (kT ≃ 3.9 keV) components. The soft component is attributed to the hot gas associated with the superwind, and the hard component with a collection of the young sources (young SNRs and HMXBs) associated with the starburst.

The prospects of utilizing Wolf-Rayet populations in starburst galaxies to infer the stellar content are reviewed. I discuss which WR star features can be detected in an integrated stellar population. Specific examples are given where the presence of WR stars can help understand galaxy properties independent of the O-type 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 WR stars. Different synthesis models for WR populations are compared. Predictions for WR 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 WR atmosphere and evolution models than for deriving the star-formation history and the initial mass function.

We summarise recent developments on synthesis models for massive star populations with a particular emphasis on Wolf-Rayet stars. Quantitative analyses of the stellar content of WR galaxies are reviewed. Comparing observations of WR galaxies from various samples with synthesis models we derive constraints on their burst properties. The observations indicate very short burst periods and are generally compatible with a Salpeter IMF and a large upper mass cut-off. The use of the Hβ equivalent width as an age indicator works well for WR galaxies. We briefly summarise comparisons of stellar populations in super star clusters which also provide useful contraints on evolutionary models for massive stars e.g., at very low metallicities inaccessible in the Local Group. In particular the observed WN and WC populations favour the high mass loss models of Meynet et al. (1994). Finally we review recent work on the origin of nebular He II emission. From the new catalogue of WR galaxies and high excitation HII regions of Schaerer et al. (1999) we find a close relation between the presence of nebular He II and WR stars. The analysis of individual WR galaxies including IZw 18 supports the suggestion of Schaerer (1996) that hot WR stars are responsible for the hard ionizing flux.

We present in this contribution our set of multi-wavelength synthesis models including the evolution of single and binary stars. The main results we have obtained can be summarized as follows: (a) massive close-binary systems will start to experience mass transfer episodes after the first 4Myr of the starburst evolution; (b) as a result of these mass transfer processes, stars of relatively low initial mass can lose completely their envelope and become a Wolf-Rayet star. In this way, the formation of WR stars is extended over longer than 15 Myr, and does not stop at 6Myr as predicted by models including only single stars; (c) WR stars can thus be coeval with red supergiants, which peak at around 10 Myr for solar metallicities; (d) the accretion of mass will originate relatively massive stars at ages for which they should have already disappeared; these stars, together with the WR stars formed in rather evolved clusters, increase the production of ionizing photons, so that the Hβ equivalent width will not drop as rapidly as predicted by models considering only individual stars; and (e) the mass transfer to compact companions will produce an additional source of high-energy radiation in the form of high-mass X-ray binaries, not predicted either by standard synthesis models.

We model the flux-density distribution of hot molecular cores as in-falling envelopes of gas and dust onto a central massive star. The envelopes are heated by both the stellar luminosity and the accretion luminosity. We find that, in order to reproduce the observed fluxes, these objects require high mass accretion rates Ṁ ≃ 10−4−3 M⊙yr−1 infalling onto central late B-type stars. We discuss the implications of this intense accretion phase on the formation of massive stars.

Many galaxies emit a dominant fraction of their bolometric luminosity in the infrared. The small mass-luminosity ratios suggest that these infrared luminosities cannot be sustained by a normal stellar population over a Hubble time. The commonly held explanation is that there has been a burst of star formation sufficiently recently that the full spectrum of the initial mass function up to the highest mass stars is contributing to the luminosity. Starburst galaxies therefore provide laboratories for studying the upper end of the initial mass function.

After a briefly summarizing studies of the IMF in the solar neighborhood and in nearby ‘normal’ galaxies, I will turn to studies of environments in which there appears to be evidence for a somewhat different IMF, viz., in starburst galaxies. I will discuss infrared spectroscopic diagnostics of a sample of the most luminous starburst galaxies and what this implies for the IMFs in these ‘violent’ star formation events.

Detection and analysis of Wolf-Rayet star populations in young starbursts is a powerful tool to reveal the properties of the star-forming events. We present new high S/N optical spectra of a sample of well-known WR galaxies, in which we detected for the first time the spectral signature of WC stars. The detailed comparison of WN and WC star populations derived in these galaxies, with the most recent model predictions, allows us to put tight constraints on starburst parameters, such as the age and duration of the burts, and the IMF.

We have observed nine Wolf-Rayet galaxies at 20,6, 3.6 and 2 cm from the VLA and at Hα from the Wise Observatory, and have mapped the radio spectral index in each galaxy on sub-arcsec scales. We find that more than half the WR sample has distinct regions where the radio spectrum rises between 6 and 2 cm. We believe that this is the radio signature of the youngest super-star-clusters which are otherwise invisible, and which require NLyc ≳ 1052, or equivalent to 750 stars of the 07 subtype. We call these sources radio super-star-clusters.

We report the observation of a prominent Wolf-Rayet feature in the high excitation H II region 74C (Courtés et al. 1993) of the spiral galaxy NGC4258. In order to reproduce the observed HeII λ4686 Å line luminosity, around 38 Wolf-Rayet stars in this region are needed.

We report the detection of a WR feature in the nuclear spectrum of the Seyfert 2 galaxy Mrk 1210. The growing list of WR/Seyfert galaxies strengthens the evidence for a connection between nuclear activity and star-formation.

We present in this contribution the predictions on the multiwavelength spectral energy distribution of our evolutionary population synthesis models including single and binary stellar systems. The high energy computations include the emission associated to X-ray binaries and supernovae remnants, as well as the mechanical energy released into the interstellar medium, which can be partially reprocessed into thermal X-rays. With these components we compute the spectral energy distribution of starburst galaxies from X-ray to radio ranges, and analyze finally the effects of the high energy emission on the H and He ionizing continuum.

Wolf-Rayet galaxies are a subset of blue emission-line galaxies, whose spectra show the signature of large numbers of WR stars, with ages in the range 1-10 x 106 yr. Mid infra-red observations are well suited to the study of their hot, massive stars since this spectral region contains many fine-structure nebular lines which depend very sensitively on stellar content. The observations of ISO have opened-up this window, which we are exploiting through a Guest Observer program with the Short Wavelength Spectrometer (sws), that includes NGC 5253, a nearby (4 Mpc) WR galaxy. NGC 5253 contains several young super-star clusters, no older than a few million years and has been the focus of many recent studies which have shown that the extinction is very high and patchy in the UV and optical. ISO observations of NGC 5253 are particularly important since the interstellar extinction at mid-IR wavelengths is low and because of its remarkably hot and young stellar population.

With an evolutionary synthesis code, I follow the evolution of a starburst composed of massive close binaries (MCBs) and of single stars at various metallicities. Evolutionary tracks for MCBs have been developed in collaboration with D. Vanbeveren and C. de Loore. The models reproduce the observed WR/O ratio in the Magellanic Clouds.

We have obtained two-dimensional spectroscopy of the central ∼10″ × 12″ of the Wolf-Rayet galaxy NGC 4861. We have carried out a preliminary analysis of the data, deriving continua at different wavelengths, and emission-line intensity maps (HeI λ4471, HeII λ4686, Hβ, [O III]λ4959, 5007) directly from the observed spectra.

We report the discovery of two new Wolf-Rayet galaxies: Mrk 1039, and F8208+2816. Two broad WR bumps at 5808Å and 4650Å indicate the presence of WCE and WNL star populations in these two sources. Comparison of the observed EW(HeII λ4686) and EW(CIV λ5808) with recent models of WR populations in young starbursts provide an indication that the stellar initial mass function in some WR galaxies might not be Salpeter-like.

We report our preliminary results on the role of starbursts in LINERS by use of Wolf-Rayet galaxies as a tool. The essence of this approach lies in the different FIR-radio relation of WR galaxies from the pure AGN's.

Preliminary results from a systematic search for WR stars in starburst galaxies are presented. We report new detections of WR spectral features (the ‘WR bump’ around ∼ 4700 Å) in eight objects.

Wolf-Rayet stars have been detected in the NW region of the metal–poor starburst galaxy I Zw 18. The integrated luminosity and FWHM of the bumps at 4650 Å and 5808 Å are consistent with the presence of a few individual stars of WC4 or WC5 type. The unexpected presence of WC stars in such a low-metallicity galaxy could, however, be explained by high mass loss rates, or alternatively favor a binary channel for WR formation. WC stars could also account for the strong and narrow He II 4686Å emission line which peaks co-spatially with the WR bump emission (see Schaerer 1996).

Most of the commonly used tracers of age of starburst regions are based on the presence of massive O-type and Wolf Rayet stars, and hence are applicable only during the nebular and WR phases of starburst. The beginning of the WR phase coincides with the end of the RSG phase. The presence of RSGs in starburst regions can be directly inferred by the Calcium Triplet (CaT) and other characteristic absorption features. These features, accompanied by evolutionary population synthesis modeling, can be used for accurate age determinations of starburst regions. Spatially resolved CaT spectra of a sample of nearby starburst galaxies are being analyzed to determine the age of starbursts in them. Results for the starburst knots of M 82 are presented here.