EDP Sciences logo
Subscriber Authentication Point
Search
Menu
Home All issues Volume 689 (September 2024) A&A, 689 (2024) A40 Full HTML
Open Access
Issue
A&A
Volume 689, September 2024
Article Number A40
Number of page(s) 10
Section Cosmology (including clusters of galaxies)
DOI https://doi.org/10.1051/0004-6361/202348530
Published online 30 August 2024

© The Authors 2024

Licence Creative CommonsOpen Access article, published by EDP Sciences, under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

This article is published in open access under the Subscribe to Open model. Subscribe to A&A to support open access publication.

1. Introduction

At low redshifts, galaxies in the Universe exhibit diverse star formation activity, leading to their classification into two main populations. The first group consists of ‘star-forming’ galaxies characterised by ongoing star formation, young stellar populations, and blue colours, signifying their evolving nature. In contrast, the second category comprises ‘quenched’ galaxies, which have ceased to form young stars, displaying older stellar populations and redder colours. This dichotomy is reflected in the so-called colour–magnitude diagram (CMD, Baldry et al. 2004a). The latter population typically displays spheroidal or elliptical morphologies and is frequently found in dense environments, such as groups and clusters (Binggeli et al. 1988; Strateva et al. 2001; Baldry et al. 2004b). This indicates that environment can play an important role in galaxy evolution (Boselli et al. 2014; Boselli & Gavazzi 2014), and this role can be particularly significant for less-massive galaxies, such as dwarfs (Binggeli et al. 1988). For this reason, it is usual to characterise an environment with the percentage of quenched galaxies it has, a percentage that depends on the galaxy mass and its environment, since both properties can play a role (Peng et al. 2010; Thomas et al. 2010; Vulcani et al. 2012; Romero-Gómez et al. 2024).

For less-massive galaxies (M⋆ < 1010 M⊙) in high-density environments, there are physical mechanisms that can stop the star formation by removing the gas (Lisker 2009; Cortese et al. 2021), such as ram-pressure stripping (Gunn & Gott 1972; Quilis et al. 2000) or strangulation Larson et al. (1980). Other environment-related processes that can play a less significant role are harassment (Moore et al. 1998; Aguerri & González-García 2009) or even strong interactions with other galaxies, which can transform massive galaxies into dwarfs (Moore et al. 1999; Mastropietro et al. 2005). In general, for satellite dwarf galaxies falling into massive halos like those of groups and clusters, most of their gas is removed at the first infall before the pericentre of their orbit (Boselli et al. 2022). On the other hand, galaxies more massive than 1010 M⊙ can also suffer strong tidal interactions and mergers that dramatically change the morphology; although in terms of stellar populations, they usually evolve according to their own internal properties, and are not strongly influenced by environment (Zheng et al. 2019; Davies et al. 2019).

The extent of the environmental influence depends on several factors, the most important being the mass and the dynamical state of the cluster. Fornax is a compact and relatively relaxed cluster, with a lower mass than others, where ram pressure is expected to have a relatively weak effect, in contrast to more massive clusters, such as Coma or Virgo, which display a more turbulent nature. In particular, the dynamical state of Fornax suggests that tidal forces may play a more significant role in stripping gas from galaxies, as outlined by Serra et al. (2023). Conversely, massive clusters such as Virgo, which are characterised by a complex substructure that indicates an unstable and young dynamical state, experience hydrodynamic interactions and galaxy harassment, contributing to the formation of massive early-type galaxies (Boselli et al. 2014). Remarkably, for less massive galaxies in Virgo, the increased density of the intracluster medium facilitates the removal of the gaseous component of late-type galaxies through ram pressure stripping events.

As far as observations are concerned, to study galaxy evolution, we rely on star formation histories (SFHs), which tell us how much mass (or other related quantities, such as integrated light) has formed over time. Using SFHs, we can infer the impact of and the dependency on the different properties of a galaxy. In particular, we can investigate when a given galaxy formed its last stars, or, in other words, when it was quenched. At this point, the definition of quenching becomes important as do the properties used to define it, which could depend on the redshift, wavelength, or even star-formation-rate tracers used in observations (Speagle et al. 2014). As observational samples can also be affected by selection effects, spectral resolution, and integrations (Salim et al. 2007; Pintos-Castro et al. 2019), the picture that we get from observations is not complete, and therefore comparison with theoretical data could help us to decrease some uncertainties and better understand the complex process of galaxy evolution.

The aim of the present study is to infer the fraction of quenched galaxies as a function of galactic stellar mass from observational data. To this end, we computed the time at which 90% of the stellar population of the galaxies was in place. We then compared these observational results with state-of-the-art cosmological simulations of galaxies.

This is paper V of this series, which is based on the analysis of spectroscopic data of dwarf galaxies, and is organised as follows: Section 2 summarises the sample selection and spectroscopic observations. In Section 3, we explain the methodology used to analyse the data and obtain the quenching times and probabilities that we present in Section 4. We then discuss the implications of our results in Section 5, and finally, in Section 6 we summarise our findings and conclusions. Throughout this paper, we adopt a ΛCMD cosmology with Ωm = 0.3, ΩΛ = 0.7, and H0 = 70 km s−1 Mpc−1.

2. Data

Below we give a brief description of the samples used for this work: The SAMI-Fornax Dwarfs Survey, the ATLAS3D project, and the Fornax3D project. This constitutes a sample of 90 galaxies spanning a stellar mass range of 107 − 12 M⊙. More details of the dwarf galaxy sample are provided in Romero-Gómez et al. (2023; hereafter Paper III), while full details of the ATLAS3D and Fornax3D sample can be found in Cappellari et al. (2011a) and Sarzi et al. (2018), respectively.

2.1. The SAMI-Fornax Dwarfs Survey

The Fornax Cluster, at α(J2000) = 3h38m30s; δ(J2000) = −35°27′18″, with a mean recessional velocity of 1454 km/s, is the second closest galaxy cluster to us, at a distance of 20 Mpc (Blakeslee et al. 2009). Within this distance, it is the second most massive cluster, with a virial mass of 7 × 1013 M⊙, and is also a relatively small cluster with a virial radius of only 0.7 Mpc (Drinkwater et al. 2001). Fornax is a dynamically evolved cluster (Churazov et al. 2008), with a velocity dispersion of 318 km/s (Maddox et al. 2019). The cluster contains about 1000 known galaxies (Venhola et al. 2021), among which are found a significant population of dwarf galaxies.

For the spectroscopic observations of selected objects, we used the Multi-Object Integral-Field Spectrograph (SAMI), which is mounted on the 3.9 m Anglo-Australian Telescope (AAT) at the Sydney-Australian Astronomical Observatory (AAO). The primary targets were 62 early-type, dE or dS0 galaxies, and the kinematic analysis was performed on 38 of them (Scott et al. 2020). After spectroscopic inspection, 39 dEs were selected. They are located within the virial radius of the Fornax cluster, except for one associated with the Fornax A group. Some dwarf galaxies show strong emission lines in their spectra, and these were excluded from the main analysis, leaving a final sample of 31 galaxies. This sample was also analysed in Paper III and in Romero-Gómez et al. (2024; hereafter Paper IV). This sample of galaxies forms the low-mass tail of the sample, spanning a stellar mass range from 107.22 to 109.3 M⊙.

2.2. The ATLAS3D project

To compare with massive galaxies, we include two samples, the first of which is located inside the Virgo cluster, and the second is in Fornax. The Virgo Cluster is the closest cluster to us, at a distance of only 16.5 Mpc, and has a recessional velocity of ∼1150 km/s (Mei et al. 2007). The cluster has a complex structure with centres around several massive galaxies, such as M86, M49, and M87, of which the former is the most massive (Schindler et al. 1999). As M87 is also the source of the peak emission in X-ray observations (Böhringer et al. 1994), it is usually considered to be the main centre. Virgo is also a relatively massive and large cluster, with a virial mass of 7 × 1014 M⊙ (Karachentsev et al. 2014) and a virial radius of 1.5 Mpc (Kim et al. 2014). Compared to Fornax, Virgo is a dynamically young cluster (Aguerri et al. 2005), with a velocity dispersion of 753 km/s (Boselli & Gavazzi 2006). Its massive galaxies are well-studied systems, and as we know from the literature, the giants are quenched early on, mostly independently of the cluster environment (Heavens et al. 2004; Panter et al. 2007; Carnall et al. 2019; McDermid et al. 2015).

To obtain information on the massive galaxies of the Virgo cluster, we reanalysed data from the ATLAS3D Project1 (Cappellari et al. 2011a), in a similar way to in our previous work. The sample consists of 260 early-type galaxies with a mass of between 1010 and 1012 M⊙; some of these belong to the Virgo cluster and some are in the field. We divided the sample into clusters and non-clusters based on the local density parameter Cappellari et al. (2011b). For this study, we only used those galaxies that fulfil the cluster membership. We also excluded galaxies with low signal-to-noise ratio, emission lines, or other issues according to McDermid et al. (2015), leaving a final sample of 50 galaxies. The ATLAS3D galaxies used here were analysed in Paper III and Paper IV. This sample can be considered a representative sample of massive early-type galaxies.

2.3. Fornax3D project

From the literature, we know that massive galaxies are relatively insensitive to the actions of the environment. Thus, one might expect massive galaxies from Virgo or Fornax clusters to behave similarly despite the differences between the two clusters. Nonetheless, to investigate the validity of these expectations, we also added a sample of massive galaxies from the Fornax cluster to compare with the galaxies from Virgo and, on the other hand, with the dwarf galaxies in Fornax.

Fornax3D (Sarzi et al. 2018) is a survey that focuses on most of the massive galaxies within the virial radius of the Fornax cluster. Using the Multi Unit Spectroscopic Explorer (MUSE) instrument on the Very Large Telescope (VLT), the survey successfully collected spectroscopic data for a comprehensive sample of 30 galaxies with a limited magnitude of MV ≤ −17 mag. For the purposes of this paper, we use position and velocity data for each galaxy from Sarzi et al. (2018), along with stellar mass information derived from Iodice et al. (2019) and Liu et al. (2019). We also incorporate the star formation histories (SFHs) of a subset of the galaxies studied in Fahrion et al. (2021). These authors studied 25 objects, from which we had to remove a couple of galaxies from Virgo, a few dwarfs that were already in our data, one galaxy that is outside of the phase-space range of the models, and a number of other low-mass dwarfs. This leaves us with a sample of ten galaxies that are also part of the Fornax3D project. The selected galaxies are intermediate-mass galaxies that have stellar masses ranging from 109.5 to 1011 solar masses; these are not as massive as those in Virgo, but are more massive than the dwarfs of our main sample. Fahrion et al. (2021) examined galaxy spectra at different apertures; however, for consistency with our treatment of the other datasets, we focus on integrated spectra at one effective radius. It is worth noting that the methods used by these latter authors to extract SFHs from the spectra are very similar to our own.

3. The quenching time of the galaxies

To obtain the galaxy properties investigated in this study, we have to work with information extracted from their SFHs. For a detailed description of the process we used to obtain these SFHs, see Paper IV; here we provide just a brief summary.

For all the samples, the dEs from Fornax, the galaxies from Fornax3D, and the giants from Virgo, the stellar population properties are obtained using full-spectral-fitting techniques, from which we derived the SFHs. The data in all cases come from integral field spectroscopy instruments, but for the purposes of this paper, we collapsed all available data into a single spectrum per galaxy. In the case of the Fornax dwarfs, Eftekhari et al. (2021) analysed the radial profiles and concluded that the velocity dispersion is constant as a function of radius. In this sense, it is safe to assume that the stellar populations do not change much with radius, as dwarfs typically have relatively uniform trends (Koleva et al. 2011; den Brok et al. 2011; Ryś et al. 2015; Bidaran et al. 2023). There are some exceptions that can be observed in local clusters. For instance, in some cases, dEs can be identified by the existence of blue nuclei, as noted by Lisker et al. (2006).

For massive galaxies, other studies have also shown that they are characterised by mostly flat age and [Mg/Fe] ratio profiles (Parikh et al. 2024). Even so, in the case of the Virgo galaxies, we tested in Paper III the resulting stellar population properties of the integrated spectra at different radii. After careful consideration, we decided to include the entire integrated spectrum of each galaxy in our study, because there is no significant difference between the observed results.

In previous studies of the SAMI galaxies, results were obtained using a spectral range of between 4700 and 5400 Å, which is similar to that of the ATLAS3D galaxies, of namely 4800 to 5300 Å. However, for our analysis of the dwarfs, we decided to use the red wavelengths of the spectra as well, adding the range between 6300 and 6800 Å. As we show in Paper IV, this extended wavelength range helped us to improve the results of the kinematics necessary to obtain the stellar populations, but it does not have any impact on the populations.

All the fitting and analysis of the galaxy spectra is made with the code https://www-astro.physics.ox.ac.uk/ mxc/software/ (Cappellari & Emsellem 2004; Cappellari 2017). The spectrum of each galaxy is fitted with a combination of single stellar population (SSP) models from the http://miles.iac.es/ library (Falcón-Barroso et al. 2011; Vazdekis et al. 2010; Sánchez-Blázquez et al. 2006). The analysis of the Fornax3D galaxies in Fahrion et al. (2021) uses the same code and SSP models. The age of these models can be seen as the time at which the different stellar populations formed, and therefore the SFH is composed of the weights of the SSP at different times. The quality of our spectra in both dwarfs and giants ensures that the https://www-astro.physics.ox.ac.uk/mxc/software/ fit is good. The time resolution of our fit is about 1–2 Gyr, which is the maximum allowed by the data. This means that we do not have sufficient time resolution to distinguish individual bursts, and for this reason, we focus on the overall shape of the reconstructed SFHs, rather than on small features that may appear (see Paper IV for more details).

3.1. Quenching and infall times

The SFH is often presented as a function of galaxy age, showing the mass of stars formed at a given point in the galaxy’s evolution. For our research, we are interested in the amount of mass formed up to a given point in time, and so we used the cumulative SFH. This function can be easily constructed as the cumulative sum of the resulting weights of the fit. This allows us to calculate, for any given time, the percentage of a galaxy’s stellar mass that has been formed relative to the total stellar mass at the present time. In particular, we are interested in knowing when a given galaxy reached a fraction equal to 90% of its present mass. This parameter, the t90, can be considered as a proxy of the quenching time of the galaxy, when the star formation has entirely ceased (see Weisz et al. 2014; Ferré-Mateu et al. 2018; Collins & Read 2022). For example, as we are working in look-back time, the smaller the t90 is, the longer the galaxy has been forming stars; it is important to note here that the t90 parameter is dependent on the time resolution of the SFHs.

The second timescale that we define is the infall time, tinf, a property that does not depend on the internal properties of the galaxy, but only on the environment. This time is defined as the moment when a galaxy that is falling into the cluster crosses the virial radius for the first time. As we cannot go back in time with the observed galaxies, we required simulations to help us with this. In Pasquali et al. (2019), the YZiCS simulations (Choi & Yi 2017) are used to study the cluster infall time of galaxies in halos from 5.3 × 1013 M⊙ to 9.2 × 1014 M⊙. To do this, these latter authors used the phase space, which is a representation of the position of the galaxies within their parent halo. This is achieved by plotting the peculiar velocities normalised by their host (cluster) velocity dispersion as a function of their projected distance normalised by the host virial radius R200. Thus, the phase space gives us some constraints on the infall time of the galaxies. However, in phase space, there can be potential biases due to projection problems. Therefore, given the position of a galaxy in the diagram, a range of infall times can be derived, each of which is associated with a probability. In this sense, it is reasonable to think that the closer to the centre a galaxy is, the longer it should have been within the host. Therefore, Pasquali et al. (2019) divided the phase space into eight zones using quadratic curves, and each of these zones is represented by the mean infall time of all the galaxies within. The different zones can be seen in the phase space of Figure 1. Using these models and the position of our galaxies in phase space, we obtain the infall time of each galaxy in our sample as the mean of the corresponding region of phase space in Fig. 1. This means that this time is an approximation, in the sense that, given the position, it is the most likely time according to the models. Even though the models use galaxies slightly more massive than the dwarfs of our sample, the tinf from the models are not very different from the corresponding infall times for galaxies in our mass range. Therefore, as we do not actually know the infall time of a given galaxy, we assume it is the mean tinf of its zone given by Pasquali et al. (2019).

thumbnail Fig. 1.

Plots showing the distribution of the galaxies in the phase space, and some of their properties. The top panel shows the phase space for the SAMI-Fornax, ATLAS3D, and Fornax3D samples. This phase space is divided into eight different zones – according to Pasquali et al. (2019) – that represent the infall time of the galaxies in those zones. For a better contrast, we marked zones 2, 4, and 6 with black lines. From green to blue, the infall times in Gyr change according to the top colour bar. The middle panel shows the stellar mass of the galaxies as a function of projected distance, with a horizontal black dashed line that separates the Fornax’s dwarfs from the rest of the galaxies. In the bottom panel, we present the t90 of the galaxies as a function of projected distance, with a dashed red line fitting the trend for giants of Virgo, a red line for the Fornax3D galaxies, and an orange line fitting the trend for dwarfs. These relations of t90 with the environment were highlighted in Paper IV. In all panels, we show the SAMI-Fornax dwarfs with circles, while the massive galaxies for Virgo and Fornax are marked with crosses and triangles, respectively. All symbols are coloured with t90, the value of which is indicated by the vertical colour bar.

3.2. Environmental quenching probability

It is difficult to determine whether or not a galaxy has been quenched by its environment. For example, we cannot be 100% sure that a host halo is responsible for quenching its satellites; however, if the quenching time of a galaxy is shorter than its infall time, then there is a higher probability that the quenching is related to the environment. With this premise, and taking into account that our dwarf galaxies are quenched by selection, we defined the probability that they were quenched in the cluster, PQ, that is, the environmental quenching probability. To compute this value, we examine whether or not

t 90 − t inf < 0 . $$ \begin{aligned} t_{90} - t_{\rm inf} < 0. \end{aligned} $$(1)

If so, then the infall time is longer than the quenching time, both measured in look-back time. Then, using the errors computed for t90 in Paper IV and for tinf in Pasquali et al. (2019), we ran Monte Carlo simulations to check how many times the condition of Equation (1) was fulfilled. Thus, we interpreted the percentage as the probability of a galaxy being quenched by its environment. Under these conditions, the limit to consider whether or not a galaxy has been influenced by its environment is t90 – tinf = 0. It is important to note that we are assuming that the quenching time we are measuring is primarily influenced by the environment. We have not considered the possibility of mass quenching, or of galaxy mergers, which are more appropriate for massive galaxies and could also stop star formation activity (Peng et al. 2010). In Section 5, we investigate these possibilities further and discuss how excluding them may affect our results.

4. Results

Figure 1 displays the phase-space distribution of all the galaxies within their respective clusters. We note that in the case of the ATLAS3D sample, we used only those galaxies in a cluster environment (for more details see Paper III). The figure reveals intriguing insights into their respective distributions. The phase space is divided into eight distinct zones, according to Pasquali et al. (2019), with zone 1 representing the region closest to the centre, while the zones progress outward with increasing distance. When considering the overall trend, it becomes evident that both dwarfs and massive galaxies exhibit distinct patterns within the phase space. The percentages of the galaxies in each of the zones are given in Table 1. We see that the dEs galaxies in Fornax demonstrate a relatively uniform distribution, with a significant proportion of objects occupying zones 2 to 5 in Fig. 1, with between 18% and 25% in each zone. For the small sample of Fornax3D galaxies, we see that all of them are between zones 2 and 5 with the majority, 40%, in zone 3. On the other hand, the giant galaxies in Virgo exhibit a more concentrated distribution, with 75% of objects found between zones 2 and 4. Additionally, there is a notable difference in the outermost cluster regions, with the dwarf sample showing a notable presence in zone 8, which contains 11% of objects. Of the giants, only 12% are in zone 5, and only one giant galaxy is in zone 7. The virialised region, zone 1, is less populated. Both clusters show similar numbers, 4% and 9% for dwarfs in Fornax and giants in Virgo, respectively. This is expected to some extent; as can be seen in Fig. 5 of Pasquali et al. (2019), the highest density of objects in the models is in zones 2 and 3. This is because, while galaxies are orbiting the cluster, the passage through the central parts is relatively fast, making it less probable that we observe them in those regions.

Table 1.

Table with the percentages of galaxies in each zone of the phase-space. The simulated galaxies are divided into two samples according to their stellar mass in order to easily compare with the dwarf and giant samples.

Figure 1 also shows the distribution of stellar mass as a function of projected distance. This does not reveal any relation between these two properties, but it clearly shows that dwarfs and massive galaxies are well separated in stellar mass. All the dEs are below 109.5 M⊙, and all the giants are above. For the t90 of the galaxies, as pointed out in the literature (Sandage 1986; Gavazzi & Scodeggio 1996; Romero-Gómez et al. 2024), the values are strongly correlated with stellar mass, in the sense that massive galaxies formed within the 2–3 Gyr following the Big Bang, while the dwarfs took more time to form. The relation between t90 and distance only exists for dwarf galaxies, as already pointed out in the literature (Michielsen et al. 2008), showing that those dEs closer to the centre of the cluster were quenched faster.

4.1. Galaxy quenching

As described in the previous section, the difference between t90 and tinf gives us the definition of the environmental quenching probability. Considering a galaxy to be quenched by the environment if t90 is equal to or lower than tinf, after analysing the distribution of PQ as a function of t90 – tinf we find that the limiting condition of quenching, t90 – tinf = 0, corresponds to PQ = 0.497 ± 0.093. We find that 3±3% of the Virgo giants, 36 ± 9% of dwarfs, and 20 ± 13% of the intermediate-massive galaxies in Fornax have probabilities higher than this limit, and so in total, only 18 ± 5% of the whole sample is compatible with being quenched by its present-day environment. Narrowing our focus to the Fornax cluster, dwarfs play a dominant role in the statistics, with about 31 ± 7% of Fornax galaxies showing that they are compatible with quenching. Notably, there are differences between the samples of massive galaxies, with those within the Fornax cluster showing a higher propensity for quenching influenced by their surroundings.

Given that t90 is related to stellar mass, it would seem logical to assume that PQ is also related. To explore this possibility, in Fig. 2 we present a compelling view of the relationship between quenching probability and stellar mass. The results reveal a distinct trend between quenching probability and stellar mass. For high mass, the probability remains consistently low, nearly zero, until reaching stellar masses of around 1010 M⊙. These galaxies have been quenched by internal processes. At this point, the probability starts to rise up to PQ = 0.4, which is more or less maintained until 108 M⊙. Upon closer examination of the Fornax3D galaxies, most of them have PQ < 0.5, which means that their environment most likely did not play a major role in the quenching. For lower stellar masses, we reach the sample of dwarf galaxies in Fornax; here we note that some galaxies have almost zero probability, while for some others we see that the quenching probability increases almost exponentially and reaches almost PQ = 1.

thumbnail Fig. 2.

Relation between the environmental quenching probability and the stellar mass. In the top panel, we show a histogram with the percentage of quenched galaxies in different mass bins, separated into observed galaxies in red and simulated galaxies in orange. In the central panel, we present the individual values of PQ as a function of the stellar mass, with symbols colour coded according to t90. In the bottom panel, we show the distribution of PQ as a function of the projected distance. For the observational sample, the symbols are the same as in Fig. 1, and the background points are the simulated galaxies from the Illustris Project. For all the observed galaxies, we include a black line that represents the mean PQ in different mass bins, and a blue dotted line that shows the same fit but only for the galaxies of the Fornax cluster. For the simulated galaxies, the mean PQ is represented by the dashed black line.

Comparing the galaxies between all the samples, we find that the massive galaxies predominantly exhibit low quenching probabilities. Approximately 94% of the giants fall below PQ = 0.2. In contrast, the dwarf sample displays a different distribution, with around 68% of the dwarfs surpassing a quenching probability of 0.2. These findings emphasise that the quenching probability of dwarfs and giants is different, showing that, as expected, low-mass galaxies are much more sensitive to their environment and are more likely to be quenched by it. Our findings are also consistent with the idea that giants are self-quenching, and do this independently of their environment (as can be seen in Peng et al. 2010; Pasquali et al. 2019; Smith et al. 2019).

4.2. Quenching of simulated galaxies

To further investigate this and to test the consistency of our findings, we compared our results with those of simulated galaxies. For the simulated sample, we used the results obtained from the IllustrisTNG2 project (Marinacci et al. 2018; Naiman et al. 2018; Nelson et al. 2018; Pillepich et al. 2018a; Springel et al. 2018), which comprises cosmological magnetohydrodynamical simulations conducted within different comoving volumes. These simulations incorporate various physical processes relevant to galaxy formation and evolution, such as star formation and evolution, gas heating and cooling, black hole growth through mergers and accretion, and feedback mechanisms, such as galactic winds and AGN activity. Detailed information about the physics implemented in the TNG simulations can be found in Weinberger et al. (2017) and Pillepich et al. (2018b). All TNG runs assume a flat ΛCDM cosmology with parameters derived from the Planck Collaboration XIII (2016) results. In our study, we specifically focused on the TNG50 simulation (Nelson et al. 2019; Pillepich et al. 2019), which provides the highest resolution within the suite, a mass particle of 8.5 × 104 M⊙.

To obtain a sample of simulated galaxies that is as similar as possible to our samples of observed galaxies, we first selected only those halos with a virial mass of as low as the Fornax cluster, and no more massive than the Virgo cluster: 13.5 < log(M200/M⊙) < 15. These halos are slightly more massive than Fornax, but with this selection, we can ensure that the simulated clusters represent a good statistical sample since each cluster contains a large number of galaxies. Then, as our galaxies do not show strong signs of active star formation, we imposed a condition on the gas-to-star particle ratio, and selected galaxies with a maximum of 20% gas, so that they can be considered early-type. Finally, regarding resolution, as we are interested in the SFHs, we selected galaxies with 50 star particles or more, to ensure sufficient resolution (Martínez-García et al. 2023). These filters resulted in a sample of 2677 simulated galaxies with 6.1 < log(M⋆/M⊙) < 11.3.

After the selection, we employed the same methodology to compute the PQ as in the observed galaxies. Briefly, we obtained the SFH and computed the t90 ; we used this as the quenching time, placed the galaxies in the projected phase-space in order to use Pasquali et al. (2019) zones to get the tinf, and then used those values to compute the PQ for each simulated galaxy. The different percentages of the distribution in the phase-space can be seen in Table 1, while the relation between PQ and stellar mass is presented in Fig. 2. As for the giants and dwarfs, the simulated galaxies are also concentrated in zones 2 to 4. However, contrary to the observed sample, which is barely present in the virialised region, zone 1, 22% of the simulated galaxies are in this zone. Although there are some simulated galaxies with high PQ, looking at Fig. 2 we see that the mean trend of the simulations gives a completely different result. The majority of the galaxies below 1010 M⊙ are not compatible with being quenched by their current environment: only 5 ± 1% are quenched by the present-day environment. These opposing results raise some interesting questions about the evolution of galaxies in clusters. To look deeper into the differences, we divided the simulated sample into galaxies with a stellar mass of greater than 109.5 M⊙ and those with a smaller mass. The simulated galaxies that are more massive than this boundary, which are similar to the giant sample, are more concentrated in the central regions and are also more quenched, 11 ± 3% of them are compatible with our quenching conditions. On the other hand, the less massive simulated galaxies, which can be compared to the SAMI-Fornax dwarfs, are as quenched as the giants from ATLAS3D, and only 5 ± 1% of them are quenched. This behaviour is visible in Fig. 2, where we fit a line to the mean values of different mass bins and the probability goes down with decreasing mass.

5. Discussion

The findings presented in our study shed light on the relationship between the probability of local environmental quenching and stellar mass for observed galaxies.

5.1. Infall time and the goodness of the sample

When we incorporate the simulated galaxies into Fig. 2, the first thing we notice is that the extent of PQ is relatively similar for both observed and simulated galaxies. This indicates that the quenching times reproduced by the simulated physical mechanism are in the same range as those of the observed galaxies. Regarding the mean PQ of the simulations, we observe that giant simulated galaxies (M⋆ > 109.5 M⊙) exhibit slightly higher quenching probabilities in the simulations, which gradually increase as the stellar mass decrease down to ∼1010 M⊙. This aligns with our observational results and suggests a degree of consistency between the two datasets. However, a notable departure from the observed trend emerges when examining less massive galaxies in the simulations. Contrary to the observations, the mean probability for a galaxy to be quenched decreases towards lower-mass galaxies. This discrepancy can be attributed to the presence of a significantly larger number of dwarf galaxies with higher quenching times in the simulated sample; that is, dwarfs that were quenched quickly after the Big Bang. As a result, these galaxies are generally somewhat bluer. These high-quenching-time dwarf galaxies exert a downward influence on the mean quenching probability, leading to the observed decline. In the SAMI-Fornax sample, we find dEs with similar t90, but in such small numbers that they do not have a significant impact on the mean trend.

To further investigate the results of the simulations, we tried to replicate the U-shape of Paper III and Paper IV with the simulated data. This is a special shape related to galaxy evolution; it appears when studying properties that depend on other internal and external properties. One such property is the abundance ratio [α/Fe], which can be obtained as part of the different stellar population properties resulting from the full-spectrum fit. The α elements, such as Mg and C, are produced in Type II supernovae, whereas Fe is predominantly produced in Type Ia supernovae. The former marks the end stage of stars with relatively short lives, while the latter originates from progenitors with significantly longer lives. Therefore, α-element abundance values can serve as a proxy for the timescales of star formation in a galaxy (Peletier 1989; Worthey et al. 1992). When comparing the [α/Fe] values of dwarfs and giants as a function of stellar mass, they form a U-shaped curve. The relation for massive galaxies is mostly linear, indicating that they evolve based on their internal properties. For galaxies with a stellar mass of lower than ∼108, the relation begins to rise, and the higher values of [α/Fe] correspond to low-mass galaxies that are closer to their host. This demonstrates how the evolution of dwarf galaxies can be influenced not only by their internal properties but also by their environment, which can rapidly suppress the star formation of less massive galaxies. In accordance with this, our previous study (Paper IV) demonstrates that another parameter associated with galaxy evolution, t90, also exhibits a U-shaped distribution. In general, this distribution shows how dwarf galaxies with a stellar mass of ∼108 M⊙ are formed more slowly, showing the lowest star formation rates. In Fig. 3, we replicate the U-shape from Paper IV between t90 and stellar mass, and include the simulations in the picture for comparison. The figure also includes dwarf galaxies from the Local Group in order to show how the relation behaves for lower masses, as well as to demonstrate how the U-shape appears even for galaxies in different environments.

thumbnail Fig. 3.

Figure showing the relation between t90 and stellar mass, as presented in Paper IV. The orange diamonds are dwarf galaxy satellites of the Milky Way or Andromeda; these galaxies are included in the plot simply to show how the U-shape continues for less massive galaxies. The red circles are the dwarfs from the SAMI-Fornax sample. The giant galaxies from the ATLAS3D project are represented by empty or filled black dots depending on whether they are in the field or a cluster, respectively. In addition to the figure in Paper IV, we have added the Fornax3D galaxies as red triangles. For all the observed galaxies, we include the mass-binned mean of t90 with a dashed grey line and a grey shadow shows the standard deviation. The simulated galaxies from IllustrisTNG (https://www.tng-project.org/) are the cyan points, while the blue line is the mass-binned mean of these galaxies.

Although weak and diluted, the simulated galaxies also show a possible U-shape. However, this simulated U-shape has the minimum t90 centred around 1010 M⊙, two orders of magnitude higher than for the observations. This suggests that the physical mechanisms acting in the simulations are making intermediate-mass dwarfs older than the observations, while the giant galaxies are younger than they should be. Our findings are consistent with the analysis that Joshi et al. (2021) made of the cumulative SFHs of dwarfs in TNG50 across a variety of environments. These authors described the weak U-shape as a decrease in t90 with satellite mass and then a flattening around 109−10 M⊙. This and the behaviour of the SFHs are strongly related to stellar mass, as found by Joshi et al. (2021), and are also dependent on the mass of the host, the position of the star inside it, and the accretion time. These authors conclude that the phase space made by Rhee et al. (2017) can be used to condense all of this, separating the satellite galaxies into ancient, intermediate, and recent infallers.

Similar to Pasquali et al. (2019), Rhee et al. (2017) parameterised the phase-space into different zones and for each one calculated the probability that a galaxy will be an ancient, intermediate, or recent infaller, along with the corresponding infall time. Using the probabilities to compute mean infall times for each zone results in similar numbers to those of Pasquali et al. (2019), but separating them into infall groups allows the infall time to be older. For example, the galaxies closer to the centre in Fig. 1 are given an infall time of 5.42 Gyr, but as they are probably ancient infallers, their infall time could be 8.35 Gyr according to Rhee et al. (2017). In Joshi et al. (2021), the accretion time can be as old as 12 Gyr, which is more similar to the infall times of the ancient infallers.

Ding et al. (2023) investigated galaxy infall times, with a particular focus on intermediate-massive galaxies within the Fornax cluster. The methodology they used involves determination of the galaxy infall time by establishing a robust correlation with the cold-disc age, a relationship calibrated using the TNG50 cosmological simulation. These authors found that the resulting times were statistically aligned with the positions of the galaxies in the phase space. Their study has three dwarfs in common with the sample studied here: FCC143, FCC 182, and FCC 301. However, in each case, the cold disc fraction is at most 0.07, meaning that their results cannot be compared with ours. In any case, Ding et al. (2023) used the phase space division outlined by Rhee et al. (2017), which, as mentioned, has infall times larger than those from Pasquali et al. (2019). With those numbers, the PQ would indeed also be larger, giving us a greater number of galaxies that are compatible with environmental quenching.

This highlights how the probability of environmental quenching may vary depending on the measurement method used and means that the results presented here are dependent on the models used. However, it is important to note that the accuracy of these models depends on the correlation between the position in phase space and the difference between recent and old infallers. Therefore, the reliability of this correlation within the adopted models is crucial for the robustness of the results.

With all this in mind, the results using the zones and times provided in Pasquali et al. (2019) are to be taken with caution because we assume that the infall time of a galaxy is equal to the mean for that zone, and not its real infall time. The results would be unlikely to change significantly if we were to use the real tinf, because if we were to have a sufficient number of galaxies, they would be closer to the mean. This means the fractions that we measure are probably very noisy because of the relatively low number of objects in our observational sample. Thus, our observational sample may not represent a good statistical sample and we could have a selection bias in the dwarfs regime. In this regard, the observational sample is just a single line of sight and could be affected by projection effects and cluster asymmetry. This suggests that, with more observations, the statistics would be compatible with the simulations. To test this hypothesis, we tried to replicate the stellar mass distribution of the observational data with the simulations. Using mass bins of 0.25 M⊙, we randomly selected simulated galaxies so that the ratio between bins and the total number of galaxies was equal in both simulated and observational samples. Once the distribution was the same in both samples, we replicated Fig. 2 with the randomly selected galaxies. The trend between PQ and stellar mass was almost identical, and therefore the target selection for the observation does not seem to have any notable impact on the results.

The problem could instead be that the simulations do not produce realistic quenching for a chosen environment (see Xie et al. 2020). The observed fractions of quenched satellite galaxies are difficult to reproduce using theoretical models of galaxy formation and evolution. In the past, state-of-the-art models tended to overestimate these fractions, especially in the low-mass regime (Weinmann et al. 2006; Baldry et al. 2006). The quenching times predicted in these models were only about 1 Gyr, which is significantly shorter than the several gigayears inferred from the observational data (Wetzel et al. 2013). Comparison of observational data with semi-analytic models leads to similar underestimation of the quenching time (De Lucia et al. 2012), though is clear that the timescales needed decrease with increasing stellar mass (Hirschmann et al. 2014). This difference could be a fundamental source of disagreement between our results and the observed and simulated galaxies.

5.2. Quenching time and preprocessed galaxies

To study the relation between quenching time and preprocessing in the TNG50 simulations, Joshi et al. (2021) defined the quenching time as the moment when the star formation rate of a galaxy is 1 dex below the corresponding star-forming main sequence. The results of these latter authors showed that using t90 as an approximation of the quenching is reasonably valid, although there could be a delay between the time when galaxies assemble 90% of their stellar mass and the moment of quenching. This is caused by the small amount of star formation left in the galaxies.

This bias arises from the fact that our results are based on the assumption that the quenching is dictated by the environment, and therefore a galaxy is quenched after its infall into a cluster or group (Muzzin et al. 2008; Fillingham et al. 2015). From the study of massive galaxies, we know that the mass formation inside a galaxy can expel the gas contributing to the quenching of the galaxy. Because of this, we are biased if there has been a process of mass quenching in the galaxy. This may suggest that dwarf galaxies are not quenched by the environment.

To investigate this possibility, Peng et al. (2010) studied galaxies in different environments using data from surveys such as SDSS (York et al. 2000) and zCOSMOS (Scoville et al. 2007). The authors concluded that the quenching of massive galaxies was primarily due to mass-quenching processes rather than environmental processes. However, for satellite galaxies, the results of Peng et al. (2010) indicated that between 30% and 70% of them are quenched when they fall into a larger halo. These percentages are consistent with the findings presented here for Fornax dwarfs, which suggest that 36 ± 9% of them are most likely quenched by the environment (Paper III). According to Serra et al. (2023), ram-pressure stripping alone is not sufficient to quench less massive galaxies in the Fornax cluster, and tidal forces are also necessary. This finding is consistent with the observations of Kleiner et al. (2023), who studied the HI content of Fornax galaxies. These authors detected HI in less than ∼5% of their sample and found that most of the HI-detected dwarfs were not close to either massive galaxies or the cluster centre. Kleiner et al. (2023) concluded that massive galaxies also play an active role in removing gas from dwarfs. Looking at the position of our dwarf galaxies in Fornax (see Fig.1 in Paper III), most of them are close to galaxies with stellar masses of greater than 1010 solar masses. In addition, some of our dwarfs have a much longer t90 than their tinf obtained from the phase space. This adds another possible mechanism leading to environmental quenching and introduces the possibility that some dwarfs may have fallen into the cluster as part of a group, arriving already quenched by the effect of that group, a phenomenon known as preprocessing (Bidaran et al. 2022). There is also the possibility that some galaxies located on the edges of the phase space and with long quenching times may be ‘backsplash’ galaxies (Sales et al. 2007). These galaxies fell into the cluster a long time ago, but after passing near the core, their orbits created a slingshot effect that threw them to the outskirts of the cluster. During this process, the environment can quench the galaxy. However, it is difficult to distinguish these galaxies from those that are falling into the environment through observations (Pimbblet 2011), and their study is more common in ΛCDM cosmological simulations of clusters (Haggar et al. 2020).

To investigate the effect of environment on the quenching of galaxies in simulations, a detailed study of various quenching pathways was conducted by Donnari et al. (2021) using TNG300. These are the lowest resolution versions of the TNG models, meaning that the stellar mass of galaxies is 109−12 M⊙. Among the pathways studied by Donnari et al. (2021), preprocessing generates an important fraction of quenched galaxies. For example, in massive hosts, M200> 1014.5 M⊙, only half of the low-mass satellites are quenched in that environment. These numbers increase as the stellar mass of the galaxy decreases, which suggests that the numbers could be even higher for the TNG50 galaxies of 106−7 M⊙. In general, around 30% of the satellites found in clusters experienced quenching before becoming part of their current host at z = 0. This would mean that the quenching probability obtained from the phase space at z = 0 is not a reliable way to decide whether or not a galaxy has been quenched by its current environment. Donnari et al. (2021) also noted that, compared to observations, the number of quenched galaxies in TNG is higher, which they argued is because of a combination of different physical mechanisms. In the IllustrisTNG simulations, this problem is known as the over-quenching effect (Angthopo et al. 2021). It means that environmental quenching is not only due to one host, and the quenching process could start in a subgroup and finally halt quickly after arriving at that host. Looking at the definition of the quenching probability, the aforementioned process would also add galaxies with low PQ in Fig. 2. In general, Donnari et al. (2021) found that the fraction of quenched low-mass satellites is higher in more massive hosts, such as clusters, and also higher near the host centre, which is expected in the environmental quenching scenario; while galaxies with M⋆ > 1010 M⊙ are only quenched by internal processes, such as AGN feedback, independently of host mass, position, or infall time, agreeing well with our results in Paper IV and Fig. 2 of this paper.

6. Summary and conclusions

In this paper, we present a study of the probability that galaxies have been quenched by their present-day environment. To this end, we used the t90 obtained from the SFHs as an approximation of the quenching time and derived the infall time of each galaxy from the parametrisation of the phase space. As observational datasets, we used a sample of 31 dEs from the SAMI-Fornax and 50 ETGs of the ATLAS3D project. Also, for comparison, we analysed the same properties in a similar sample of simulated galaxies from the Illustris TNG50.

  1. Our results demonstrate robust agreement with previous findings in the literature. Specifically, more massive galaxies evolve on their own, while less-massive galaxies, such as dwarfs, can experience quenching due not only to their internal properties but also to environmental processes.

  2. Analysing the environmental quenching probability as a function of stellar mass, we obtain a relation in which less massive galaxies are more likely to be quenched by the current environment. From our sample of SAMI-Fornax dwarfs, only 36 ± 9% of them are compatible with being quenched by the environment, while for the ATLAS3D giant galaxies, the percentage falls to 3 ± 3%.

  3. The simulated galaxies of IllustrisTNG3 show a different behaviour, with the quenching probability decreasing with decreasing stellar mass. This corresponds to massive simulated galaxies being younger than the observational ones, while less massive simulated galaxies are older than our observed sample.

  4. The observed disparity between the trends of less massive galaxies in the simulated and observed samples highlights the intricate interplay of various physical processes governing galaxy evolution. Though we have gained valuable insights into the quenching of dwarf galaxies, further investigation is essential to discern the specific mechanisms responsible for the distinct behaviours observed in both our simulations and real-world data. Such investigations could more deeply probe the role of feedback mechanisms and the influence of the surrounding environment on quenching processes, contributing to a more comprehensive understanding of galaxy evolution.

  5. Our comparison between observations and simulations underscores the importance of considering sample selection, volume, the quenching definition, and projection effects. Caution must be exercised when relying on phase-space parametrisation results, especially with smaller sample sizes that may not yield statistically robust conclusions. However, this was tested and did not appear to be a major issue. Future studies could address these limitations by exploring larger sample sizes and employing alternative analysis methods. Such endeavours will provide a more robust foundation for interpreting and contextualising the implications of our findings.

In conclusion, our analysis of observed galaxies combined with simulations reinforces the known trends between local quenching probability and stellar mass but also highlights important inconsistencies between observations and simulations. While the observed galaxies exhibit a gradual increase in quenching probability for less massive galaxies, the simulated galaxies show a contrasting decline. This disparity underscores the need for comprehensive theoretical frameworks and a deeper understanding of the intricate processes that shape the evolution of galaxies, particularly for the dwarf population.

Data Availability

The reduced data underlying this article will be available through Romero-Gómez et al. (2024) in CDS. The raw data is publicly available in the AAT data archive.

Acknowledgments

JRG and JALA are supported by the Spanish Ministry of Education, Culture and Sports under grant AYA2017-83204-P and by the Spanish Ministerio de Ciencia e Innovación y Universidades by the grant PID2020-119342GB-I00. We would like to thank Ignacio Martín Navarro and Katja Fahrion for kindly giving us their SFHs of the Fornax3D galaxies. For the analysis, we have used Python http://www.python.org; Matplotlib (Hunter 2007), a suite of open source python modules that provide a framework for creating scientific plots; and Astropy, a community-developed core Python package for Astronomy (Astropy Collaboration 2013).

References

  1. Aguerri, J. A. L., & González-García, A. C. 2009, A&A, 494, 891 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  2. Aguerri, J. A. L., Gerhard, O. E., Arnaboldi, M., et al. 2005, AJ, 129, 2585 [CrossRef] [Google Scholar]
  3. Angthopo, J., Negri, A., Ferreras, I., et al. 2021, MNRAS, 502, 3685 [Google Scholar]
  4. Astropy Collaboration (Robitaille, T. P., et al.) 2013, A&A, 558, A33 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  5. Baldry, I. K., Glazebrook, K., Brinkmann, J., et al. 2004a, ApJ, 600, 681 [Google Scholar]
  6. Baldry, I. K., Balogh, M. L., Bower, R., Glazebrook, K., & Nichol, R. C. 2004b, Am. Inst. Phys. Conf. Ser., 743, 106 [Google Scholar]
  7. Baldry, I. K., Balogh, M. L., Bower, R. G., et al. 2006, MNRAS, 373, 469 [Google Scholar]
  8. Bidaran, B., La Barbera, F., Pasquali, A., et al. 2022, MNRAS, 515, 4622 [NASA ADS] [CrossRef] [Google Scholar]
  9. Bidaran, B., La Barbera, F., Pasquali, A., et al. 2023, MNRAS, 525, 4329 [NASA ADS] [CrossRef] [Google Scholar]
  10. Binggeli, B., Sandage, A., & Tammann, G. A. 1988, ARA&A, 26, 509 [NASA ADS] [CrossRef] [Google Scholar]
  11. Blakeslee, J. P., Jordán, A., Mei, S., et al. 2009, ApJ, 694, 556 [Google Scholar]
  12. Böhringer, H., Briel, U. G., Schwarz, R. A., et al. 1994, Nature, 368, 828 [CrossRef] [Google Scholar]
  13. Boselli, A., & Gavazzi, G. 2006, PASP, 118, 517 [Google Scholar]
  14. Boselli, A., & Gavazzi, G. 2014, A&ARv., 22, 74 [NASA ADS] [CrossRef] [Google Scholar]
  15. Boselli, A., Voyer, E., Boissier, S., et al. 2014, A&A, 570, A69 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  16. Boselli, A., Fossati, M., & Sun, M. 2022, A&ARv., 30, 3 [Google Scholar]
  17. Cappellari, M. 2017, MNRAS, 466, 798 [Google Scholar]
  18. Cappellari, M., & Emsellem, E. 2004, PASP, 116, 138 [Google Scholar]
  19. Cappellari, M., Emsellem, E., Krajnović, D., et al. 2011a, MNRAS, 413, 813 [Google Scholar]
  20. Cappellari, M., Emsellem, E., Krajnović, D., et al. 2011b, MNRAS, 416, 1680 [Google Scholar]
  21. Carnall, A. C., McLure, R. J., Dunlop, J. S., et al. 2019, MNRAS, 490, 417 [Google Scholar]
  22. Choi, H., & Yi, S. K. 2017, ApJ, 837, 68 [NASA ADS] [CrossRef] [Google Scholar]
  23. Churazov, E., Forman, W., Vikhlinin, A., et al. 2008, MNRAS, 388, 1062 [Google Scholar]
  24. Collins, M. L. M., & Read, J. I. 2022, Nat. Astron., 6, 647 [NASA ADS] [CrossRef] [Google Scholar]
  25. Cortese, L., Catinella, B., & Smith, R. 2021, PASA, 38, e035 [NASA ADS] [CrossRef] [Google Scholar]
  26. Davies, L. J. M., Robotham, A. S. G., Lagos, C. d. P., et al. 2019, MNRAS, 483, 5444 [Google Scholar]
  27. De Lucia, G., Weinmann, S., Poggianti, B. M., Aragón-Salamanca, A., & Zaritsky, D. 2012, MNRAS, 423, 1277 [NASA ADS] [CrossRef] [Google Scholar]
  28. den Brok, M., Peletier, R. F., Valentijn, E. A., et al. 2011, MNRAS, 414, 3052 [Google Scholar]
  29. Ding, Y., Zhu, L., van de Ven, G., et al. 2023, A&A, 672, A84 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  30. Donnari, M., Pillepich, A., Joshi, G. D., et al. 2021, MNRAS, 500, 4004 [Google Scholar]
  31. Drinkwater, M. J., Gregg, M. D., & Colless, M. 2001, ApJ, 548, L139 [Google Scholar]
  32. Eftekhari, F. S., Peletier, R. F., Scott, N., et al. 2021, MNRAS, 497, 1571 [NASA ADS] [Google Scholar]
  33. Fahrion, K., Lyubenova, M., van de Ven, G., et al. 2021, A&A, 650, A137 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  34. Falcón-Barroso, J., Sánchez-Blázquez, P., Vazdekis, A., et al. 2011, A&A, 532, A95 [Google Scholar]
  35. Ferré-Mateu, A., Alabi, A., Forbes, D. A., et al. 2018, MNRAS, 479, 4891 [CrossRef] [Google Scholar]
  36. Fillingham, S. P., Cooper, M. C., Wheeler, C., et al. 2015, MNRAS, 454, 2039 [NASA ADS] [CrossRef] [Google Scholar]
  37. Gavazzi, G., & Scodeggio, M. 1996, A&A, 312, L29 [NASA ADS] [Google Scholar]
  38. Gunn, J. E., & Gott, J. R. I. 1972, ApJ, 176, 1 [Google Scholar]
  39. Haggar, R., Gray, M. E., Pearce, F. R., et al. 2020, MNRAS, 492, 6074 [NASA ADS] [CrossRef] [Google Scholar]
  40. Heavens, A., Panter, B., Jimenez, R., & Dunlop, J. 2004, Nature, 428, 625 [CrossRef] [Google Scholar]
  41. Hirschmann, M., De Lucia, G., Wilman, D., et al. 2014, MNRAS, 444, 2938 [NASA ADS] [CrossRef] [Google Scholar]
  42. Hunter, J. D. 2007, Comput. Sci. Eng., 9, 90 [NASA ADS] [CrossRef] [Google Scholar]
  43. Iodice, E., Sarzi, M., Bittner, A., et al. 2019, A&A, 627, A136 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  44. Joshi, G. D., Pillepich, A., Nelson, D., et al. 2021, MNRAS, 508, 1652 [NASA ADS] [CrossRef] [Google Scholar]
  45. Karachentsev, I. D., Tully, R. B., Wu, P.-F., Shaya, E. J., & Dolphin, A. E. 2014, ApJ, 782, 4 [NASA ADS] [CrossRef] [Google Scholar]
  46. Kim, S., Rey, S.-C., Jerjen, H., et al. 2014, ApJS, 215, 22 [Google Scholar]
  47. Kleiner, D., Serra, P., Maccagni, F. M., et al. 2023, A&A, 675, A108 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  48. Koleva, M., Prugniel, P., De Rijcke, S., & Zeilinger, W. W. 2011, MNRAS, 417, 1643 [NASA ADS] [CrossRef] [Google Scholar]
  49. Larson, R. B., Tinsley, B. M., & Caldwell, C. N. 1980, ApJ, 237, 692 [Google Scholar]
  50. Lisker, T. 2009, Astron. Nachr., 330, 1043 [NASA ADS] [CrossRef] [Google Scholar]
  51. Lisker, T., Glatt, K., Westera, P., & Grebel, E. K. 2006, AJ, 132, 2432 [Google Scholar]
  52. Liu, Y., Peng, E. W., Jordán, A., et al. 2019, ApJ, 875, 156 [NASA ADS] [CrossRef] [Google Scholar]
  53. Maddox, N., Serra, P., Venhola, A., et al. 2019, MNRAS, 490, 1666 [Google Scholar]
  54. Marinacci, F., Vogelsberger, M., Pakmor, R., et al. 2018, MNRAS, 480, 5113 [NASA ADS] [Google Scholar]
  55. Martínez-García, A. M., del Pino, A., Łokas, E. L., van der Marel, R. P., & Aparicio, A. 2023, MNRAS, 526, 3589 [CrossRef] [Google Scholar]
  56. Mastropietro, C., Moore, B., Mayer, L., et al. 2005, MNRAS, 364, 607 [Google Scholar]
  57. McDermid, R. M., Alatalo, K., Blitz, L., et al. 2015, MNRAS, 448, 3484 [Google Scholar]
  58. Mei, S., Blakeslee, J. P., Côté, P., et al. 2007, ApJ, 655, 144 [Google Scholar]
  59. Michielsen, D., Boselli, A., Conselice, C. J., et al. 2008, MNRAS, 385, 1374 [NASA ADS] [CrossRef] [Google Scholar]
  60. Moore, B., Lake, G., & Katz, N. 1998, ApJ, 495, 139 [Google Scholar]
  61. Moore, B., Lake, G., Quinn, T., & Stadel, J. 1999, MNRAS, 304, 465 [NASA ADS] [CrossRef] [Google Scholar]
  62. Muzzin, A., Wilson, G., Lacy, M., Yee, H. K. C., & Stanford, S. A. 2008, ApJ, 686, 966 [NASA ADS] [CrossRef] [Google Scholar]
  63. Naiman, J. P., Pillepich, A., Springel, V., et al. 2018, MNRAS, 477, 1206 [Google Scholar]
  64. Nelson, D., Pillepich, A., Springel, V., et al. 2018, MNRAS, 475, 624 [Google Scholar]
  65. Nelson, D., Pillepich, A., Springel, V., et al. 2019, MNRAS, 490, 3234 [Google Scholar]
  66. Panter, B., Jimenez, R., Heavens, A. F., & Charlot, S. 2007, MNRAS, 378, 1550 [Google Scholar]
  67. Parikh, T., Saglia, R., Thomas, J., et al. 2024, MNRAS, 528, 7338 [CrossRef] [Google Scholar]
  68. Pasquali, A., Smith, R., Gallazzi, A., et al. 2019, MNRAS, 484, 1702 [Google Scholar]
  69. Peletier, R. F. 1989, PhD thesis, University of Groningen, The Netherlands [Google Scholar]
  70. Peng, Y.-J., Lilly, S. J., Kovač, K., et al. 2010, ApJ, 721, 193 [Google Scholar]
  71. Pillepich, A., Nelson, D., Hernquist, L., et al. 2018a, MNRAS, 475, 648 [Google Scholar]
  72. Pillepich, A., Springel, V., Nelson, D., et al. 2018b, MNRAS, 473, 4077 [Google Scholar]
  73. Pillepich, A., Nelson, D., Springel, V., et al. 2019, MNRAS, 490, 3196 [Google Scholar]
  74. Pimbblet, K. A. 2011, MNRAS, 411, 2637 [NASA ADS] [CrossRef] [Google Scholar]
  75. Pintos-Castro, I., Yee, H. K. C., Muzzin, A., Old, L., & Wilson, G. 2019, ApJ, 876, 40 [Google Scholar]
  76. Planck Collaboration XIII. 2016, A&A, 594, A13 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  77. Quilis, V., Moore, B., & Bower, R. 2000, Science, 288, 1617 [Google Scholar]
  78. Rhee, J., Smith, R., Choi, H., et al. 2017, ApJ, 843, 128 [Google Scholar]
  79. Romero-Gómez, J., Peletier, R. F., Aguerri, J. A. L., et al. 2023, MNRAS, 522, 130 [CrossRef] [Google Scholar]
  80. Romero-Gómez, J., Aguerri, J. A. L., Peletier, R. F., et al. 2024, MNRAS, 527, 9715 [Google Scholar]
  81. Ryś, A., Koleva, M., Falcón-Barroso, J., et al. 2015, MNRAS, 452, 1888 [CrossRef] [Google Scholar]
  82. Sales, L. V., Navarro, J. F., Abadi, M. G., & Steinmetz, M. 2007, MNRAS, 379, 1475 [NASA ADS] [CrossRef] [Google Scholar]
  83. Salim, S., Rich, R. M., Charlot, S., et al. 2007, ApJS, 173, 267 [NASA ADS] [CrossRef] [Google Scholar]
  84. Sánchez-Blázquez, P., Peletier, R. F., Jiménez-Vicente, J., et al. 2006, MNRAS, 371, 703 [Google Scholar]
  85. Sandage, A. 1986, A&A, 161, 89 [NASA ADS] [Google Scholar]
  86. Sarzi, M., Iodice, E., Coccato, L., et al. 2018, A&A, 616, A121 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  87. Schindler, S., Binggeli, B., & Böhringer, H. 1999, A&A, 343, 420 [Google Scholar]
  88. Scott, N., Eftekhari, F. S., Peletier, R. F., et al. 2020, MNRAS, 497, 1571 [Google Scholar]
  89. Scoville, N., Abraham, R. G., Aussel, H., et al. 2007, ApJS, 172, 38 [NASA ADS] [CrossRef] [Google Scholar]
  90. Serra, P., Maccagni, F. M., Kleiner, D., et al. 2023, A&A, 673, A146 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  91. Smith, R., Pacifici, C., Pasquali, A., & Calderón-Castillo, P. 2019, ApJ, 876, 145 [Google Scholar]
  92. Speagle, J. S., Steinhardt, C. L., Capak, P. L., & Silverman, J. D. 2014, ApJS, 214, 15 [Google Scholar]
  93. Springel, V., Pakmor, R., Pillepich, A., et al. 2018, MNRAS, 475, 676 [Google Scholar]
  94. Strateva, I., Ivezić, Ž., Knapp, G. R., et al. 2001, AJ, 122, 1861 [CrossRef] [Google Scholar]
  95. Thomas, D., Maraston, C., Schawinski, K., Sarzi, M., & Silk, J. 2010, MNRAS, 404, 1775 [NASA ADS] [Google Scholar]
  96. Vazdekis, A., Sánchez-Blázquez, P., Falcón-Barroso, J., et al. 2010, MNRAS, 404, 1639 [NASA ADS] [Google Scholar]
  97. Venhola, A., Peletier, R. F., Salo, H., et al. 2021, A&A, accepted, [arXiv:2111.01855] [Google Scholar]
  98. Vulcani, B., Poggianti, B. M., Fasano, G., et al. 2012, MNRAS, 420, 1481 [NASA ADS] [CrossRef] [Google Scholar]
  99. Weinberger, R., Springel, V., Hernquist, L., et al. 2017, MNRAS, 465, 3291 [Google Scholar]
  100. Weinmann, S. M., van den Bosch, F. C., Yang, X., & Mo, H. J. 2006, MNRAS, 366, 2 [NASA ADS] [CrossRef] [Google Scholar]
  101. Weisz, D. R., Dolphin, A. E., Skillman, E. D., et al. 2014, ApJ, 789, 147 [Google Scholar]
  102. Wetzel, A. R., Tinker, J. L., Conroy, C., & van den Bosch, F. C. 2013, MNRAS, 432, 336 [Google Scholar]
  103. Worthey, G., Faber, S. M., & Gonzalez, J. J. 1992, ApJ, 398, 69 [Google Scholar]
  104. Xie, L., De Lucia, G., Hirschmann, M., & Fontanot, F. 2020, MNRAS, 498, 4327 [NASA ADS] [CrossRef] [Google Scholar]
  105. York, D. G., Adelman, J., Anderson, J. E. J., et al. 2000, AJ, 120, 1579 [Google Scholar]
  106. Zheng, Z., Li, C., Mao, S., et al. 2019, ApJ, 873, 63 [NASA ADS] [CrossRef] [Google Scholar]

All Tables

Table 1.

Table with the percentages of galaxies in each zone of the phase-space. The simulated galaxies are divided into two samples according to their stellar mass in order to easily compare with the dwarf and giant samples.

In the text

All Figures

thumbnail Fig. 1.

Plots showing the distribution of the galaxies in the phase space, and some of their properties. The top panel shows the phase space for the SAMI-Fornax, ATLAS3D, and Fornax3D samples. This phase space is divided into eight different zones – according to Pasquali et al. (2019) – that represent the infall time of the galaxies in those zones. For a better contrast, we marked zones 2, 4, and 6 with black lines. From green to blue, the infall times in Gyr change according to the top colour bar. The middle panel shows the stellar mass of the galaxies as a function of projected distance, with a horizontal black dashed line that separates the Fornax’s dwarfs from the rest of the galaxies. In the bottom panel, we present the t90 of the galaxies as a function of projected distance, with a dashed red line fitting the trend for giants of Virgo, a red line for the Fornax3D galaxies, and an orange line fitting the trend for dwarfs. These relations of t90 with the environment were highlighted in Paper IV. In all panels, we show the SAMI-Fornax dwarfs with circles, while the massive galaxies for Virgo and Fornax are marked with crosses and triangles, respectively. All symbols are coloured with t90, the value of which is indicated by the vertical colour bar.
In the text
thumbnail Fig. 2.

Relation between the environmental quenching probability and the stellar mass. In the top panel, we show a histogram with the percentage of quenched galaxies in different mass bins, separated into observed galaxies in red and simulated galaxies in orange. In the central panel, we present the individual values of PQ as a function of the stellar mass, with symbols colour coded according to t90. In the bottom panel, we show the distribution of PQ as a function of the projected distance. For the observational sample, the symbols are the same as in Fig. 1, and the background points are the simulated galaxies from the Illustris Project. For all the observed galaxies, we include a black line that represents the mean PQ in different mass bins, and a blue dotted line that shows the same fit but only for the galaxies of the Fornax cluster. For the simulated galaxies, the mean PQ is represented by the dashed black line.
In the text
thumbnail Fig. 3.

Figure showing the relation between t90 and stellar mass, as presented in Paper IV. The orange diamonds are dwarf galaxy satellites of the Milky Way or Andromeda; these galaxies are included in the plot simply to show how the U-shape continues for less massive galaxies. The red circles are the dwarfs from the SAMI-Fornax sample. The giant galaxies from the ATLAS3D project are represented by empty or filled black dots depending on whether they are in the field or a cluster, respectively. In addition to the figure in Paper IV, we have added the Fornax3D galaxies as red triangles. For all the observed galaxies, we include the mass-binned mean of t90 with a dashed grey line and a grey shadow shows the standard deviation. The simulated galaxies from IllustrisTNG (https://www.tng-project.org/) are the cyan points, while the blue line is the mass-binned mean of these galaxies.
In the text

Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.

Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.

Initial download of the metrics may take a while.