With typical sizes of a few kiloparsecs, bulges in nearby galaxies would be very difficult to resolve spatially at intermediate to high redshifts even with the best instruments on board of Hubble Space Telescope. In addition, the morphologies of galaxies are known to deviate from the standard Hubble sequence from redshift ∼1 onwards (e.g. Elmegreen et al., 2008), so we should probably not think of bulges at high-redshift in the same way we think of them in the local Universe. Nevertheless knowing the conditions, in terms of rotational support, of the galaxies that will eventually lead to lenticular and spiral galaxies nearby, can help us understand the kind of progenitors that will host the variety of bulges we see today.
In the light of the large amount of pseudobulges observed in the nearby Universe, a logical question to ask is: do we see the signatures of secular evolution in bulges at high-z? Numerical simulations reproducing the clumpy galaxies from redshift z ∼ 1 suggest that bulge kinematics is not very different from the values observed for pressure-supported systems, with (V/σ) values below 0.5 (e.g. Bournaud et al., 2007, Elmegreen et al., 2008). This is likely due to the turbulent nature of clumps merging at the centre of galaxies (e.g. Ceverino et al., 2012). Note, however, that the merging and migration of clumps towards the inner regions is an internal process, as it takes place in the disk of galaxies. The physical conditions, in terms of gas supply, for bulge formation at high redshifts are very different from the ones observed in the local Universe. Secular evolution takes place at a much faster pace at high-z.
Integral-field observations of galaxies at increasing redshifts confirm the turbulent nature of disks, as revealed by the systematically high velocity dispersion values (e.g. Newman et al., 2013, Wisnioski et al., 2014). Nevertheless, galaxies show a wide range of kinematic properties: from well behave rotating disks, to dispersion dominated systems, and galaxies with chaotic motions (e.g. Yang et al., 2008, Genzel et al., 2008, Wisnioski et al., 2011, Buitrago et al., 2014). Recent results from the KMOS3D survey (Wisnioski et al., 2014) show that most galaxies, in the main star forming sequence, between redshifts 1 and 2 are rotationally-supported. When combined with other datasets, they measure an evolution of the ionised-gas velocity dispersion which is consistent with the observed changes in the gas fractions and specific star formation rates of galaxies as a function of redshift. This results favours an 'equilibrium' model where the amount of turbulence of a disk is defined by the balance between gas accretion and outflows.
The physical conditions between redshifts 1 and 4 appear to be particularly favourable for the formation of bulges, and yet it appears that it cannot be the only channel to build the (pseudo)bulges observed in the nearby Universe. Mergers seem to be required too (e.g. Ceverino et al., 2014). To complicate the issue further, the analysis of the star formation histories of different types of bulges (e.g. Seidel et al., 2015) suggest that at least 60% of the stellar mass of those bulges formed at redshifts beyond 4 (see Figure 9). All these results together indicate that bulge formation most likely happens in a two stage process (e.g. Obreja et al., 2013), with an initial period of rapid build-up (with possible influence of mergers) and a secondary phase (between redshifts 1 and 2) of high star formation activity that would lead to the younger pseudobulge components we see today.
Figure 9. Relative light (top row) and mass (bottom row) fractions of young, intermediate and old stellar populations as a function of radius present in three galactic bulges studied in Seidel et al. (2015). Uncertainties in the analysis are indicated in the top left corner. Shaded regions mark the regions where the average light and mass fractions of this study are computed. More than 60% of the stellar mass in those bulges was already in place beyond z ∼ 4.