6 Emission properties of galaxies in CGs

UZC labels homogeneously the spectral classification ( $E={\rm emission}$ lines, $A={\rm absorption}$ lines, B=E+A) for each galaxy, thereby allowing a check for possible links between emission properties and membership in CGs.

To test whether samples of Ts and Ms are intrinsically different, the fraction of emission (with or without absorption lines) to absorption galaxies can be compared. This fraction also represents a rough estimate of the incidence of young (or rejuvenated) over old objects, or alternatively of Spirals over Ellipticals. Figure 8 shows the Emission over Absorption (E/A) galaxy ratio for Ts (triangles) and Ms (squares) within each distance class. It is worth underlying that points in Fig. 8 indicate the ratio of the total population of emission galaxies over A galaxies in Ts and Ms.

It emerges that the fraction of emission over absorption galaxies decreases from sample I to IV. This trend towards a larger fraction of galaxies with emission spectra increasing for lower galaxy luminosities was already known to exist both in the optical (Zucca et al. 1997; Ratcliffe et al. 1998; Tresse et al. 1999) and in the near-IR (Mamon et al. 2001). Any comparison of the emission line galaxy fraction with respect to kinematical parameters has to account for this trend which, concerning morphology, was already reported by Tikhonov (1990), Mamon (1990) and by Whitmore (1992). However, Fig. 8 shows that, even when accounting for the decrease of emission line galaxies with redshift, Ts include higher fractions of emission line galaxies than Ms. The luminosity of Ts and Ms member galaxies being similar, the trend of increasing fraction of emission-line galaxies with decreasing multiplicity is probably real. Galaxies in Ts and Ms in sample I display no significant differences, in accordance with kinematical similarities between Ts and Ms in this subsamples.

Given that emission line galaxies are typically field galaxies, the data clearly suggest that Ts are more likely than Ms to be field structures (or to be contaminated by field interlopers) as already indicated by their lower $\sigma _{v}$. To make this point more evident Fig. 8 additionally displays the E/A ratio for Single galaxies and for galaxies in CGs which are ACO subclumps (ACO$_{\rm CG}$). Single galaxies are UZC galaxies which turn out to have no UZC companion(s) within an area of 200 h^-1 kpc radius, and within $\Delta cz=\pm 1000$ km s^-1 and form a plausible comparison sample for CGs on small scales. Among UZC galaxies single galaxies are $\approx$10 times more numerous than CG galaxies. It clearly emerges that CGs, whatever their luminosity, are lacking in gas rich galaxies when compared to single galaxies, and that the deficiency is larger for Ms. At the same time Fig. 8 shows that CGs as a whole display an excess of spiral-rich galaxies when compared to those CGs which have been excluded from the sample because they turned out to be ACO$_{\rm CG}$.

Figure 8: Relative fraction of galaxies displaying Emission spectra over galaxies displaying Absorption spectra for Ts, Ms, for single galaxies in UZC (no neighbours within $r=200~h^{\rm -1}$ kpc) and for CGs that are ACO subclumps, which have been excluded from the main CG sample.

Figure 9: Distribution of emission over absorption galaxy content as a function of CGs velocity dispersion. While in Ms (hatched area) the galaxy content is only modestly related to the velocity dispersion, the fraction of emission galaxies in Ts (solid histogram) turns out to be a strongly decreasing function of velocity dispersion. The hatched line shows the distribution for the whole CG sample.

Our data show the existence of a trend from single galaxies to galaxies in cluster subclumps, in which CGs occupy an intermediate position. Figure 9, displaying the ratio of emission over absorption galaxies (in CGs at distance between 2500 and 10 000 km s^-1) as a function of CG $\sigma _{v}$ confirms that a morphology-velocity dispersion relation holds for the whole sample (hatched line), but also that the trend is induced by the inclusion among the CG sample of Ts (bold line) and specifically of low $\sigma _{v}$ Ts. Accordingly, any process linking the increase of $\sigma _{v}$ to the evolution of the spectral content of CGs is expected to be relevant predominantly in low multiplicity CGs. It is worth pointing out that if most low $\sigma _{v}$ Ts are non-real structures, the morphology-velocity dispersion relation is not retrieved. The morphology-velocity dispersion relation is similar to the morphology-density relation observed in clusters and loose groups (Dressler 1980; Postman & Geller 1984; Whitmore & Gilmore 1991) with the fraction of gas-rich galaxies being a strong signature of multiplicity. The morphology-density relation has previously been shown to hold for HCGs (Mamon 1986; Hickson et al. 1988) with an offset relative to the general Postman & Geller relation, indicating that at given spiral fraction, compact groups appear denser. It might be the inclusion within the sample of several spiral-rich, low multiplicity CGs that induces the offset, given that we find Ts to be even denser than Ms. Again, as for the morphology-velocity dispersion relation, the offset is to be reduced if most spiral rich, low $\sigma _{v}$ Ts are non-physical systems.

If the lower fraction of emission line galaxies in Ms corresponds to a lower fraction of Spirals, one accordingly expects the median $(M/L)_{\rm gal}$ of Ms members to be higher than for Ts galaxies. This could at least partially account for the higher M/L associated with Ms, although it remains uncertain whether the higher $\sigma _{v}$ and early type galaxy content associated with Ms do indeed indicate that these are systems more evolved than Ts. Multiplicity also appears to strongly influences the behaviour of systems in Hickson's sample. Specifically we have shown (Focardi & Kelm 2001) that the observed correlation between morphology and velocity dispersion in HCGs, (Hickson et al. 1988, 1992; Prandoni et al. 1994) just strongly reflects the different dynamical properties of systems with different multiplicity. In summary spectral characteristics indicate that two factors tend to strongly influence the number of emission line galaxies that will be retrieved in a CG sample. One is the fraction of faint galaxies included in the sample, with fainter galaxies being more likely to display emission line spectra. The second is the minimum multiplicity of CGs. The inclusion of Ts strongly biases a sample towards emission spectra galaxies. Combined with the average lower $\sigma _{v}$, interactions between galaxies in Ts are accordingly predicted to be more disruptive than those in Ms, which suggests that perturbation patterns and/or asymmetric rotation curves (Rubin et al. 1991) should be more frequent among Ts.