A. Gargiulo

The different physical mechanisms that drive the star formation histories of giant and dwarf galaxies

We present an analysis of star formation and nuclear activity in galaxies as a function of both luminosity and environment in the fourth data release of the Sloan Digital Sky Survey. Using a sample of 27 753 galaxies in the redshift range 0.005 < z < 0.037 that is ≥90 per cent complete to M r = -18.0, we find that the Ha equivalent width, EW(Hα), distribution is strongly bimodal, allowing galaxies to be robustly separated into passively evolving and star-forming populations about a value EW(Ha) = 2 A. In high-density regions ∼70 per cent of galaxies are passively evolving independent of luminosity. In the rarefied field, however, the fraction of passively evolving galaxies is a strong function of luminosity, dropping from 50 per cent for M r ≤ -21 to zero by M r ∼ -18. Indeed for the lowest luminosity range covered (-18 < M r < -16) none of the ∼600 galaxies in the lowest-density quartile is passively evolving. The few passively evolving dwarf galaxies in field regions appear as satellites to bright (≥ L*) galaxies. We find a systematic reduction of ∼30 per cent in the Ha emission from dwarf (-19 < M r < - 18) star-forming galaxies in high-density regions with respect to field values, implying that the bulk of star-forming dwarf galaxies in groups and clusters are currently in the process of being slowly transformed into passive galaxies. The fraction of galaxies with the optical signatures of an active galactic nucleus (AGN) decreases steadily from ∼50 per cent at M r ∼ -21 to ∼0 per cent by M r ∼ -18 closely mirroring the luminosity dependence of the passive galaxy fraction in low-density environments. This result reflects the increasing importance of AGN feedback with galaxy mass for their evolution, such that the star formation histories of massive galaxies are primarily determined by their past merger history. In contrast, the complete absence of passively evolving dwarf galaxies more than ∼2 virial radii from the nearest massive halo (i.e. cluster, group or massive galaxy) indicates that internal processes, such as merging, AGN feedback or gas consumption through star formation, are not responsible for terminating star formation in dwarf galaxies. Instead the evolution of dwarf galaxies is primarily driven by the mass of their host halo, probably through the combined effects of tidal forces and ram-pressure stripping.

Shapley Optical Survey – II. The effect of environment on the colour–magnitude relation and galaxy colours

We present an analysis of the effects of environment on the photometric properties of galaxies in the core of the Shapley supercluster at z = 0.05, one of the most massive structures in the local universe. The Shapley Optical Survey (SOS) comprises archive Wide Field Imager (WFI) optical imaging of a 2.0deg 2 region containing the rich clusters A3556, A3558 and A3562 which demonstrate a highly complex dynamical situation including ongoing cluster mergers. The B - R/R colour-magnitude relation has an intrinsic dispersion of 0.045 mag and is 0.015 ± 0.005 mag redder in the highest-density regions, indicative of the red sequence galaxy population being 500-Myr older in the cluster cores than towards the virial radius. The B - R colours of galaxies are dependent on their environment, whereas their luminosities are independent of the local density, except for the very brightest galaxies (M R < -22). The global colours of faint (?M* + 2) galaxies change from the cluster cores where ∼90 per cent of galaxies lie along the cluster red sequence to the virial radius, where the fraction has dropped to just ∼20 per cent. This suggests that processes directly related to the supercluster environment are responsible for transforming faint galaxies, rather than galaxy merging, which should be infrequent in any of the regions studied here. The largest concentrations of faint blue galaxies are found between the clusters, coincident with regions containing high fractions of ∼L* galaxies with radio emission indicating starbursts. Their location suggests star formation triggered by cluster mergers, in particular the merger of A3562 and the poor cluster SC 1329-313, although they may also represent recent arrivals in the supercluster core complex. The effect of the A3562-SC 1329-313 merger is also apparent as a displacement in the spatial distribution of the faint galaxy population from both the centres of X-ray emission and the brightest cluster galaxies for both systems. The cores of each of the clusters/groups are marked by regions that have the lowest blue galaxy fractions and reddest mean galaxy colours over the whole supercluster region, confirming that star formation rates are lowest in the cluster cores. In the cases of A3562 and SC 1329-313, these regions coincide with the centres of X-ray emission rather than the peaks in the local surface density, indicating that ram-pressure stripping may have an important role in terminating any remnant star formation in galaxies that encounter the dense intracluster medium (ICM) of the cluster cores.

The SDSS-GALEX viewpoint of the truncated red sequence in field environments at z∼ 0

We combine GALEX near-UV photometry with a volume-limited sample of local (0.005 2u This contamination is greatest at faint magnitudes (Mr> 19) and in field regions where as many as three-quarters of red sequence galaxies are star-forming, and as such has important consequences for following the build-up of the red sequence. We find that the NUV r colour instead allows a much more robust separation of passively-evolving and star-forming galaxies, which allows the build-up of the UV-selected red sequence with redshift and environment to be directly interpreted in terms of the assembly of stellar mass in passively-evolving galaxies. We find that in isolated field regions the number density of UV-optical red sequence galaxies declines rapidly at magnitudes fainter than Mr� 19 and appears completely truncated at Mr� 18. This confirms the findings of Haines et al. (2007) that no passively-evolving dwarf galaxies are found more than two virial radii from a massive halo, whether that be a group, cluster or massive galaxy. These results support the downsizing paradigm whereby the red sequence is assembled from the top down, being already largely in place at the bright end by z�1, and the faint end filled in at later epochs in clusters and groups through environment-related processes such as ram-pressure stripping or galaxy harassment.

On the origin of the scatter around the Fundamental Plane: correlations with stellar population parameters

We present a Fundamental Plane (FP) analysis of 141 early-type galaxies in the Shapley supercluster at z = 0.049 based on spectroscopy from the AAOmega spectrograph at the Anglo-Australian Telescope and photometry from the WFI on the European Southern Observatory/MPI 2.2-m telescope. The key feature of the survey is its coverage of low-mass galaxies down to σ ∼ 50 km s ―1 . We obtain a best-fitting FP relation r e ∝ σ 1.06±:0.06 0 (I) e ―0.82±0.02 in the R band. The shallow exponent of σ 0 is a result of the extension of our sample to low velocity dispersions. Considering the subsample of σ 0 > 100 km s ―1 galaxies, the FP relation is r e ∝ σ 1.35 (I) ―0.81 e , consistent with previous studies in the high-luminosity regime. We investigate the origin of the intrinsic FP scatter, using estimates of age, metallicity and α/Fe. We find that the FP residuals anticorrelate (>3σ) with the mean stellar age in agreement with previous work. However, a stronger (>4σ) correlation with α/Fe is also found. These correlations indicate that galaxies with effective radii smaller than those predicted by the FP have stellar populations systematically older and with α overabundances larger than average, for their σ. Including α/Fe as a fourth parameter in the FP, the total scatter decreases from 0.088 to 0.075 dex and the estimated intrinsic scatter decreases from 0.068 to 0.049 dex. Thus, variations in α/Fe account for ∼30 per cent of the total variance around the FP, and ∼50 per cent of the estimated intrinsic variance. This result indicates that the distribution of galaxies around the FP are tightly related to the enrichment, and hence to the time-scale of star formation. Our results appear to be consistent with the merger hypothesis for the formation of ellipticals which predicts that a significant fraction of the scatter is due to variations in the importance of dissipation in forming merger remnants of a given mass.

Shapley Optical Survey – I. Luminosity functions in the supercluster environment

We present the Shapley Optical Survey, a photometric study covering a ∼2-deg 2 region of the Shapley supercluster core at z ∼ 0.05 in two bands (B and R). The galaxy sample is complete to B = 22.5 (>M* + 6, N gal = 16588) and R = 22.0 (>M* + 7, N gal = 28008). The galaxy luminosity function (LF) cannot be described by a single Schechter function due to dips apparent at B ∼ 17.5 (M B ∼ -19.3) and R ∼ 17.0 (M R ∼ -19.8) and the clear upturn in the counts for galaxies fainter than B and R ∼ 18 mag. We find, instead, that the sum of a Gaussian and a Schechter function, for bright and faint galaxies, respectively, is a suitable representation of the data. We study the effects of the environment on the photometric properties of galaxies, deriving the galaxy LFs in three regions selected according to the local galaxy density, and find a marked luminosity segregation, in the sense that the LF faint end is different at more than 3σ confidence level in regions with different densities. In addition, the LFs of red and blue galaxy populations show very different behaviours: while red sequence counts are very similar to those obtained for the global galaxy population, the blue galaxy LFs are well described by a single Schechter function and do not vary with the density. Such large environmentally dependent deviations from a single Schechter function are difficult to produce solely within galaxy merging or suffocation scenarios. Instead the data support the idea that mechanisms related to the cluster environment, such as galaxy harassment or ram-pressure stripping, shape the galaxy LFs by terminating star formation and producing mass-loss in galaxies at ∼M* + 2, a magnitude range where blue late-type spirals used to dominate cluster populations, but are now absent.

On the central stellar mass density and the inside-out growth of early-type galaxies

Recent theoretical and observational studies on the assembly of early-type galaxies (ETGs) point towards an inside-out growth of their stellar mass characterized by extended low-mass-density haloes grown around compact and dense cores. Models can form ETGs at high-z as compact spheroids that then grow in size through dry minor mergers. Dry mergers would affect mainly the outskirts of the galaxy, enlarging the size (i.e. the effective radius), keeping the inner parts and the total stellar mass nearly unchanged. Hence, the central stellar mass density will not change with time, in contrast to the stellar mass density within the effective radius, which should decrease with time as the effective radius increases. Some previous observations are interpreted as supporting inside-out growth, as the central stellar mass density of high-z ETGs is found to be similar to that of local ETGs. In this paper we derive the central stellar mass density within a fixed radius and the effective stellar mass density within the effective radius for a complete sample of 34 ETGs morphologically selected at 0.9 2. Furthermore, we show that the small scatter of the central mass density of ETGs compared with the large scatter of the effective mass density is simply a peculiar feature of the Sersic profile and hence is independent of redshift and of any assembly history experienced by galaxies. Thus, it has no connection with the possible inside-out growth of ETGs. Finally, we show a tight correlation between the central stellar mass density and the total stellar mass of ETGs in the sense that the central mass density increases with mass as . This implies that the fraction of the central stellar mass of ETGs decreases with the mass of the galaxy. These correlations are valid for the whole population of ETGs considered, independently of their redshift, suggesting that they originate in the early phases of their formation.

Spatially resolved colours and stellar population properties in early-type galaxies at z ∼ 1.5

We present F850LP − F160W colour gradients for 11 early-type galaxies (ETGs) at 1.0 Re, enclosing the whole galaxy, we have found that the fraction of high-z ETGs with negative F850LP − F160W colour gradients rises up to 100 per cent. For each galaxy, we investigate the origin of the radial colour variation with an innovative technique based on the matching of both the spatially resolved colour and the global spectral energy distribution (SED) to predictions of composite stellar population models. In fact, we find that the age of the stellar populations is the only parameter whose radial variation alone can fully account for the observed colour gradients and global SEDs for six ETGs in our sample, without the need of radial variation of any other stellar population property. For four out of these six ETGs, a pure metallicity variation can also reproduce the detected colour gradients. None the less, a minor contribution to the observed colour gradients from the radial variation of star formation time-scale, abundance of low- to high-mass stars and dust cannot be completely ruled out. For the rest of the sample, our analysis suggests a more complex scenario whereby more properties of the stellar populations need to simultaneously vary, likely with comparable weights, to generate the observed colour gradients and global SEDs. Our results show that, despite the young mean age of our galaxies (<3– 4 Gyr), they already exhibit significant differences among their stellar content. We have discussed our results within the framework of the widest accepted scenarios of galaxy formation and conclude that none of them can satisfactorily account for the observed distribution of colour gradients and for the spatially resolved content of high-z ETGs. Our results suggest that the distribution of colour gradients may be due to different initial conditions in the formation mechanisms of ETGs.

Colour gradients in normal and compact early‐type galaxies at 1 < z < 2

We have derived colour gradients for a sample of 20 early-type galaxies (ETGs) at 1 < zspec < 2 selected from the GOODS-South field. The sample includes both normal ETGs (13) having effective radii comparable to the mean radius of local ones and compact ETGs (7) having effective radii from two to six times smaller. Colour gradients have been derived in the F606W-F850LP bands ( UV-U rest-frame) taking advantage of the ultradeep HST-ACS observations covering this field and providing a spatial resolution of about 0.8 kpc at the redshift of the galaxies. Despite of the narrow wavelength baseline covered (1000 u sampling approximatively the emission dominated by the same stellar population, we detect significant radial colour variations in 50 per cent of our sample. In particular, we find five ETGs with positive colour gradients (cores bluer than the external regions), and five galaxies with negative colour gradients (cores redder than the external regions), as commonly observed in the local Universe. These results show that at 1 < z < 2, when the Universe was only 3-4 Gyr old, ETGs constituted a composite population of galaxies whose different assembly histories have generated different stellar distributions with the bluest stellar population either in the center or in the outskirts as well as throughout the whole galaxy. Moreover, we find that compact galaxies seem to preferentially show a blue cores while moving towards normal galaxies, central stellar populations become progressively redder. Nonetheless, the narrow baseline covered together with the low statistics still prevent us to be conclusive about a possible physical connection between colour gradients and the degree of compactness of high-z ETGs.

Spectral detection of multiple stellar populations in z~1 early-type galaxies

We thank the anonymous referee of this paper for providing constructive comments that improved the manuscript. This work has received financial support from PRIN-INAF (1.05.09.01.05).

ACCESS - IV. The quenching of star formation in a cluster population of dusty S0s

We present an analysis of the mid-infrared (MIR) colours of 165 70-μm-detected galaxies in the Shapley supercluster core (SSC) at z= 0.048 using panoramic Spitzer/MIPS 24- and 70-μm imaging. While the bulk of galaxies show f70/f24 colours typical of local star-forming galaxies, we identify a significant subpopulation of 23 70-μm-excess galaxies, whose MIR colours (f70/f24 > 25) are much redder and cannot be reproduced by any of the standard model IR spectral energy distributions (SEDs). These galaxies are found to be strongly concentrated towards the cores of the five clusters that make up the SSC, and also appear rare among local field galaxies, confirming them as a cluster-specific phenomenon. Their optical spectra and lack of significant ultraviolet emission imply little or no ongoing star formation, while fits to their panchromatic SEDs require the far-IR emission to come mostly from a diffuse dust component heated by the general interstellar radiation field rather than ongoing star formation. Most of these 70-μm-excess galaxies are identified as ∼L* S0s with smooth profiles. We find that almost every cluster galaxy in the process of star formation quenching is already either an S0 or Sa, while we find no passive galaxies of class Sb or later. Hence the formation of passive early-type galaxies in cluster cores must involve the prior morphological transformation of late-type spirals into Sa/S0s, perhaps via pre-processing or the impact of cluster tidal fields, before a subsequent quenching of star formation once the lenticular encounters the dense environment of the cluster core. In the cases of many cluster S0s, this phase of star formation quenching is characterized by an excess of 70-μm emission, indicating that the cold dust content is declining at a slower rate than star formation. We suggest that the excess 70-μm emission during quenching is due to either (i) a reduction of the star formation efficiency as proposed within the morphological quenching scenario or (ii) a 2–3 times increase in the dust-to-gas ratio or metallicity of the remaining interstellar medium, as predicted by chemical evolutionary models of galaxies undergoing ram-pressure stripping or starvation.