4.1. Extremely Red Objects
Extremely red objects (EROs), defined by an optical-infrared color limit, typically using the criteria (R-K) > 5 - 6, or (I-K) > 4 - 5 (e.g., Daddi et al. 2000), were traditionally thought to be early type or dusty galaxies at z > 0.8. Extremely red galaxies are red because of a large 4000 Å break produced by aged, or dusty, stellar populations. EROs are now also found in the near infrared with a (J-K) limit that locates objects at redshifts z > 2. The ERO population is therefore a good one for determining the basic properties of evolved galaxies at high redshifts.
The morphologies of EROs are mixed, with a strong redshift dependence (Figure 7). Systems at z < 1.2 are typically early or late-type galaxies, while those at z > 1.2 are more irregular or peculiar (Moustakas et al. 2004). The spectra of EROs are also mixed, with about half showing signs of evolved stellar populations, with the other half showing emission lines (Cimatti et al. 2002). Morphological studies of EROs demonstrate that a large fraction, perhaps the majority of the K < 20 objects, are disks (Yan & Thompson 2003). It thus appears likely that the ERO definition, far from finding only specific galaxy populations, is sampling all morphological types. This is a good sample for studying galaxies evolving onto the Hubble sequence since Hubble types are the most evolved galaxies at z ~ 0 and are likely also the most evolved at z ~ 1. This has been done by e.g., Moustakas et al. (2004) who studied the redshift distributions of EROs in the GOODS-South field as a function of morphology. Moustakas et al. found that the lower redshift EROs are dominated by regular galaxies, while higher redshift samples at z > 1.5 are dominated by galaxies that cannot be place on the Hubble sequence (Figure 7).
Figure 7. The redshift distributions for extremely red galaxies from the GOODS-South field (Moustakas et al. 2004) showing how the morphological distribution for this popualtion shifts to the 'other' or peculiar types at higher redshifts.
Recently, there have been claims of a new population of red galaxies at z > 2. These are similar to the lower redshift ERO population in that they are identified by a color cut that isolates galaxies based on the Balmer break, but uses the color defined by two infrared filters, normally the J and K bands (e.g., (J-K) > 2.3). These systems are typically at z > 2 and may be quite distinct from the Lyman-break galaxy population (Franx et al. 2003; van Dokkum et al. 2004). The morphologies of these systems have not been studied in detail, partially because there are so few examples, yet bright K-band selected galaxies at z > 2 and with K < 20 often have irregular UV morphologies (Labbe et al. 2003; Daddi et al. 2004) indicating star formation. There are however some hints for spiral structures and more regular compact near infrared morphologies in some of these systems (Labbe et al. 2003; Daddi et al. 2004). These infrared EROs have large stellar masses and are generally consistent, based on clustering analyses and stellar population arguments, with the most evolved systems at high redshift. These galaxies are therefore the best candidates for being the progenitors of evolved galaxies found in dense regions today.