We detect AC118 from the center to half the Abell radius (1.5 Mpc) and possibly to 2.0 Mpc.
There is a luminosity segregation among the galaxies in the AC118 cluster and it has been shown in three different ways: by studying the LF dependence on environment, by a radial analysis of the dwarf-to-giant ratio and by comparing the radial profiles of galaxies of different luminosities. While the three methods differ, they are not independent. It is the order in which the grouping is done that changes: galaxies are first grouped spatially and then their luminosity distribution is studied in the LF analysis, while in the two other methods galaxies are first grouped in luminosity and then their spatial distribution is studied.
Any choice of cosmology rigidly moves the upper abscissa of Figs. 2-5 by a fixed amount, and does not change the shape or relative differences of the plotted profiles. Therefore, the detection of a luminosity segregation in AC118 is independent of the choice of the cosmological values.
The segregation concerns mainly the inner 250 Kpc of the cluster (see in particular Fig. 4), while at larger radii all galaxies have the same spatial distribution regardless of the galaxy near-infrared luminosity, up to 2 Mpc away from the cluster center. The segregation consists of an excess, of a factor of 3, of giants galaxies in the cluster core (or in a deficit, of the same factor, of dwarf galaxies). Since the numerical density of dwarfs turns out to be largely constant, the luminosity segregation found seems due to an excess (relative to the number of dwarf galaxies) of giant galaxies in the cluster center, and not due to a deficit of dwarfs in the remaining of the cluster.
Beside AC118, luminosity segregation in the near-infrared has been suggested in the Coma cluster (Andreon & Pelló 2000), although through comparison of heterogeneous data.
The luminosity segregation found in Paper I is here confirmed to hold over a even wider cluster region. With respect to the previous investigation on AC118, we take two more paths for confirming the luminosity segregations: the analysis of the galaxy spatial distribution, and the computation of the radial profile of the giant to dwarf ratio. Our results are in broad agreement with what has been found in similar analyses, but performed at optical wavelengths (Zwicky 1957; Mellier et al. 1988; Driver et al. 1998; Secker et al. 1997; Garilli et al. 1999), or by using the velocity segregation (Chincarini & Rood 1977; Struble 1979; Biviano et al. 1992; Stein 1997), or by analysing the galaxy angular correlation function (Loveday et al. 1995): these studies found that brightest galaxies are more tightly correlated (or have lower velocity dispersions) than the faintest galaxies.
Therefore, there is clear evidence of luminosity segregation. Since the near-infrared luminosity is a good tracer of the stellar mass (Bruzual & Charlot 1993), the segregation found is interpreted a mass-related segregation. The luminosity segregation we found in the near-infrared implies a mass segregation more tightly that under the usual assumption than optical luminosity traces mass: here we show directly that massive galaxies are found preferentially in the cluster center.
A mass-related segregation is a natural expected outcome of a hierarchical scenario of cluster formation, because the clustering strength depends on the halo circular velocity (and therefore on mass) in cold dark matter models (White et al. 1987; Kauffmann et al. 1997). However, the effect has been detected only recently in the simulations (Springer et al. 2001) and a quantitative comparison between observations and simulations awaits a prediction in a more suitable form.
The hostile cluster environment plays a role in shaping the AC118 LF but only at small clustercentric radii (or high density), since outside the cluster core the LF computed at several locations are all compatible with each other and the dwarf-to-giant ratio is constant within the errors.
Acknowledgements
This work is part of a collaboration with M. Arnaboldi, G. Busarello, M. Capaccioli, G. Longo, P. Merluzzi and G. Theureau. A. Wolter and A. Iovino are acknowledged for useful discussions. The near-infrared observations presented in this paper have been taken during the NTT guaranteed time of Osservatorio di Capodimonte. The director of my institute, Prof. M. Capaccioli, is warmly thanked for permitting me a long stay at the Osservatorio Astronomico di Brera, where this work has been prepared. The director of the latter institute, Prof. G. Chincarini, is acknowledged for hospitality. Comments from the referee helped to improve the presentation of this paper.
Copyright ESO 2002