7.2. Galaxy interactions and nuclear activity
Huge starbursts require galaxy interactions and mergers (e.g. Sanders & Mirabel 1996). For AGN of low luminosity, external triggering appears less necessary, since only 0.01 M/yr is required for a Seyfert like NGC 1068 for example, during 108 yr. However, once interactions drive gas to the nucleus, some activity must be revived. Time-scales may be the reason why the actions are not simultaneous. Large-scale gas has to be driven at very small scales in the center, and the whole process requires several intermediate steps.
Even for the good starburst/interactions correlations, there are exceptions for the low-luminosity samples. Interacting galaxies selected optically (not IRAS galaxies) are often not enhanced in star formation (Bushouse 1986, Lawrence et al 1989). Only the obviously merging galaxies, like the Toomre (1977) sample, are truly a starbursting class; it is difficult to reveal a progression along a possible evolutionary sequence (Heckman 1990). There are too many determining parameters: geometry, distance, gas content, etc...
A complication comes from the time-scales involved: the starburst phase is short, of the order of a few 108yr, similar only to the end of the merging phase. It is more the presence of morphological distortions than the presence of nearby companions that is correlated with activity. More than 50% of ULIRGS possess multiple nuclei (Carico et al 1990, Graham et al 1990).
Are Seyfert galaxies preferentially interacting? According to Dahari (1984), 15% have close companions, compared to 3% in the control sample. But the Seyferts with or without companions have the same H or radio power (Dahari 1985), although they may be more infrared bright with companions (Dahari & DeRobertis 1988, McKenty 1989). According to Keel et al (1985), there are 5% of Seyferts in control sample, and 25% in the close pairs of Arp Atlas. But it is possible that the Arp Atlas galaxies suffer from selection effects. Bushouse (1986) on the contrary finds a deficiency of Seyferts in interacting galaxies.
In summary, if the environment influence is evident for powerful QSOs, Radio-galaxies and BlLacs, it is not so significant for Seyferts (deRobertis et al 1996).
Surprising observations also involve LSB (Low Surface Brightness Galaxies) that are about 6 times less bright than HSB. Their low evolution state is generally attributed to their isolated environment. A large fraction of active nuclei have been reported in their category: 65% in about 50 LSB instead of 1% expected for these low luminosities (Knezek & Schombert 1993, Sprayberry et al 1995). Are there unknown selection effects?