Cosmic Supernova Rates and the Hubble Sequence

We compute the type Ia, Ib/c and II supernova (SN) rates as functions of thecosmic time for galaxies of different morphological types. We use fourdifferent chemical evolution models, each one reproducing the features of aparticular morphological type: E/S0, S0a/b, Sbc/d and Irr galaxies. Weessentially describe the Hubble sequence by means of decreasing efficiency ofstar formation and increasing infall timescale. These models are used to studythe evolution of the SN rates per unit luminosity and per unit mass asfunctions of cosmic time and as functions of the Hubble type. Our resultsindicate that: (i) the observed increase of the SN rate per unit luminosity andunit mass from early to late galaxy types is accounted for by our models. Ourexplanation of this effect is related to the fact that the latest Hubble typeshave the highest star formation rate per unit mass; (ii) By adopting a Scalo(1986) initial mass function in spiral disks, we find that massive single starsending their lives as Wolf-Rayet objects are not sufficient to account for theobserved type Ib/c SN rate per unit mass. Less massive stars in close binarysystems can give instead a significant contribution to the local Ib/c SN rates.On the other hand, with the assumption of a Salpeter (1955) IMF for all galaxytypes, single massive WR stars are sufficient to account for the observed typeIb/c SN rate. (iii) Our models allow us to reproduce the observed type Ia SNrate density up to redshift z~1. We predict an increasing type Ia SN ratedensity with redshift, reaching a peak at redshift z >= 3, because of thecontribution of massive spheroids.Comment: ApJ, accepted for publication. 17 pages, 11 figure