Annu. Rev. Astron. Astrophys. 1998. 36: 189-231 Copyright © 1998 by Annual Reviews. All rights reserved

3.3. Other Global Influences on Star Formation Rate

What other global properties of a galaxy influence its SFR? It is plausible to expect the mass, bar structure, spiral arm structure, or environment to be important, and empirical information on all of these are available.

3.3.1. LUMINOSITY AND MASS     Gavazzi & Scodeggio (1996), Gavazzi et al (1996) have compiled UV, visible, and near-IR photometry for over 900 nearby galaxies, and they found an anti-correlation between the SFR per unit mass and the galaxy luminosity, as indicated by broadband colors and H EWs. At least part of this trend seems to reflect the same dependence of SFR on Hubble type discussed above, but a mass dependence is also observed among galaxies of the same Hubble type. It is interesting that there is considerable overlap between the color-luminosity relations of different spiral types, which suggests that part of the trends that are attributed to morphological type may be more fundamentally related to total mass. A strong correlation between B - H color and galaxy luminosity or linewidth has been discussed previously by Tully et al (1982), Wyse (1983). The trends seem to be especially strong for redder colors, which are more closely tied to the star formation history and mean metallicity than the current SFR. More data are needed to fully disentangle the effects of galaxy type and mass, for both the SFR and the star formation history.

3.3.2. BARS     Stellar bars can strongly perturb the gas flows in disks and trigger nuclear star formation (see next section), but they do not appear to significantly affect the total disk SFRs. Figure 3 plots the H EW distributions separately for normal (SA and SAB) and barred (SB) spirals, as classified in the Second Reference Catalog of Bright Galaxies. There is no significant difference in the EW distributions (except possibly for the Sa/SBa galaxies), which suggests that the global effect of a bar on the disk SFR is unimportant. Ryder & Dopita (1994) reached the same conclusion based on H observations of 24 southern galaxies.

Tomita et al (1996) carried out a similar comparison of FIR emission, based on IRAS data and broadband photometry for 139 normal spirals and 260 barred Sa-Sc galaxies. They compared the distributions of L_FIR / L_B ratios for Sa/SBa, Sb/SBb, and Sc/SBc galaxies and concluded that there is no significant correlation with bar structure, consistent with the H results. There is evidence for a slight excess in FIR emission in SBa galaxies, which could reflect bar-triggered circumnuclear star formation in some of the galaxies, though the statistical significance of the result is marginal (Tomita et al 1996).

Recent work by Martinet & Friedli (1997) suggests that influence of bars on the global SFR may not be as simple as suggested above. They analyzed H and FIR-based SFRs for a sample of 32 late-type barred galaxies and found a correlation between SFR and the strength and length of the bar. This suggests that large samples are needed to study the effects of bars on the large-scale SFR and that the structural properties of the bars themselves need to be incorporated in the analysis.

3.3.3. SPIRAL ARM STRUCTURE     Similar tests have been carried out to explore whether a strong grand-design spiral structure enhances the global SFR. Elmegreen & Elmegreen (1986) compared UV and visible broadband colors and H EWs for galaxies they classified as grand design (strong two-armed spiral patterns) and flocculent (ill-defined, patchy spiral arms), and they found no significant difference in SFRs. McCall & Schmidt (1986) compared the supernova rates in grand-design and flocculent spirals and drew similar conclusions. Grand-design spiral galaxies show strong local enhancements of star formation in their spiral arms (e.g. Cepa & Beckman 1990, Knapen et al 1992), so the absence of a corresponding excess in their total SFRs suggests that the primary effect of the spiral density wave is to concentrate star formation in the arms, but not to increase the global efficiency of star formation.

3.3.4. GALAXY-GALAXY INTERACTIONS     Given the modest effects of internal disk structure on global SFRs, it is perhaps somewhat surprising that external environmental influences can have much stronger effects on the SFR. The most important influences by far are tidal interactions. Numerous studies of the global H and FIR emission of interacting and merging galaxies have shown strong excess star formation (e.g. Bushouse 1987, Kennicutt et al 1987, Bushouse et al 1988, Telesco et al 1988, Xu & Sulentic 1991, Liu & Kennicutt 1995). The degree of the SFR enhancement is highly variable, ranging from zero in gas-poor galaxies to on the order of 10-100 times in extreme cases. The average enhancement in SFR over large samples is a factor of 2-3. Much larger enhancements are often seen in the circumnuclear regions of strongly interacting and merging systems (see next section). This subject is reviewed in depth by Kennicutt et al (1998).

3.3.5. CLUSTER ENVIRONMENT     There is evidence that a cluster environment systematically alters the star formation properties of galaxies, independently of the well-known density-morphology relation (Dressler 1984). Many spiral galaxies located in rich clusters exhibit significant atomic gas deficiencies (Haynes et al 1984, Warmels 1988, Cayatte et al 1994), which presumably are the result of ram pressure stripping from the intercluster medium, combined with tidal stripping from interactions with other galaxies and the cluster potential. In extreme cases, the gas removal would be expected to affect the SFRs as well. Kennicutt (1983b) compared H EWs of 26 late-type spirals in the Virgo cluster core with the field sample of Kennicutt & Kent (1983) and found evidence for a 50% lower SFR in Virgo, comparable to the level of HI deficiency. The UV observations of the cluster Abell 1367 by Donas et al (1990) also show evidence for lower SFRs. However, subsequent studies of the Coma, Cancer, and A1367 clusters by Kennicutt et al (1984), Gavazzi et al (1991) showed no reduction in the average SFRs and, if anything, a higher number of strong emission-line galaxies.

A comprehensive H survey of eight Abell clusters by Moss & Whittle (1993) suggests that the effects of cluster environoment on global star formation activity are quite complex. They found a 37-46% lower H detection rate among Sb, Sc, and irregular galaxies in the clusters, but a 50% higher detection rate among Sa-Sab galaxies. They argued that these results arise from a combination of competing effects, including reduced star formation from gas stripping as well as enhanced star formation triggered by tidal interactions. Ram-pressure-induced star formation may also be taking place in a few objects (Gavazzi & Jaffe 1985).