Environmental processing of galaxies in H i-rich groups

Hydrogen in its atomic form (H i) is the raw material for condensing into molecular Hydrogen (H₂), the fuel for future star formation, thus it is an essential part of galaxies (Bigiel & Blitz, 2012; Catinella et al., 2018). The H i content in galaxies is often analysed using various scaling relations utilizing galaxy properties such as stellar mass, color, morphology, star formation rate, as well as tracing galaxy environment (Haynes & Giovanelli, 1984; Giovanelli & Haynes, 1985; Dénes et al., 2014; Catinella et al., 2018). Large surveys such as The H i Parkes All Sky Survey (Hipass, Barnes et al. 2001) and the Arecibo Legacy Fast ALFA Survey (Alfalfa, Giovanelli et al. 2005) detected H i in more than 5000 and 30000 galaxies respectively, and have significantly improved these scaling relations.

Mapping the H i gas within and around the galaxies can show its distribution, be used to determine galaxy kinematics, and to study galaxy interactions. The H i features that reveal galaxy interactions, and are often found in galaxy groups, are in the form of tidal tails (Scott et al., 2012), bridges (Jones et al., 2019), rings (Malphrus et al., 1997) and clouds (Ryder et al., 2001), and more of such systems are described in the brief overview by Koribalski (2020). The abundance of such H i features points towards a complex history of galaxy evolution in the group environment.

It is known that in galaxy clusters, the most densest of environments, a galaxy’s evolution is influenced by its environment. Thus, we observe an increased fraction of early-type galaxies (Dressler, 1980), decreased star formation rates (Lewis et al., 2002) and decreased gas content of galaxies (Haynes & Giovanelli, 1984) in high density environments. On the other hand, the evolution of galaxies in the field is thought to be free from environmental effects. Naturally, galaxy groups are in between these extreme environments.

It has been shown that spiral galaxies are more H i deficient in higher density environments, for instance, near the galaxy cluster centre or in groups, than spiral galaxies in the field (Giovanelli & Haynes, 1985; Chung et al., 2009; Solanes et al., 2001; Verdes-Montenegro et al., 2001; Kilborn et al., 2009; Dénes et al., 2014, 2016). The difference in H i content across different environments suggests that the higher density environment contributes to the exhaustion of the H i gas. The depletion of the H i content within clusters is most often connected to ram-pressure stripping (Gunn & Gott, 1972; Chung et al., 2009; Cortese et al., 2011; Brown et al., 2017; Stevens & Brown, 2017). On the other hand, the depletion of the H i content in groups is debated between ram-pressure stripping (Brown et al., 2017), tidal interactions (Yun et al., 1994), starvation (Larson et al., 1980) and viscous stripping (Nulsen, 1982).

The most well-studied type of groups are Hickson Compact Groups (HCG), which are dense galaxy configurations that contain four to 10 galaxies with a projected distance between them on the order of a galaxy size (and smaller) and velocity dispersion of $\sim$200 km s^-1 (Hickson, 1997; Williams et al., 1990; Verdes-Montenegro et al., 2001; Borthakur et al., 2015; Coziol & Plauchu-Frayn, 2007). From observations of the H i content in Hickson Compact Groups (Hickson, 1982; Verdes-Montenegro et al., 2001; Borthakur et al., 2015) it has been found that gas removal is evident as the galaxies are often depleted of their H i content. Verdes-Montenegro et al. (2001) explains the evolutionary path of galaxy groups through a gradual removal of the H i gas from the galaxies. In contrast to HCGs, loose groups are sparse and have a typical galaxy separation of few hundred kpc (Sengupta & Balasubramanyam, 2006; Pisano et al., 2011). Kilborn et al. (2009) found that the H i deficient galaxies within loose groups tend to lie within the 1 Mpc in projected distance from the group centre. Utilizing over 740 galaxy groups from the Alfalfa survey, Hess & Wilcots (2013) found that the fraction of H i-rich galaxies decreases in the central regions of groups as the group optical membership increases. They also found that H i gas depletion starts first in the lowest H i mass galaxies.

It is common to make a distinction between the central galaxy which is most often the brightest (most massive) galaxy within a group (or a cluster) and satellite galaxies, being all other galaxies within the same group (e.g. White & Rees 1978; Skibba et al. 2011; Lacerna et al. 2014. The difference between the central galaxy and satellites is that the central galaxy is theoretically at rest with respect to its halo, while satellite galaxies are moving in the gravitational potential of a halo and can experience environmental effects (White & Rees, 1978; Yang et al., 2007; Brown et al., 2017). Džudžar et al. (2019a) gives the example of a group with two centrals, both at rest within the halo. Recent studies of the H i content of central and satellite galaxies have shown significance of the group mass regime. Janowiecki et al. (2017) have shown that central galaxies within small groups are $\sim$0.3 dex more H i-rich than the central galaxies in isolation. Brown et al. (2017), utilizing the stacking technique, found that satellite galaxies in high mass haloes have a lower H i mass fraction with respect to those in low mass haloes.

An increasing number of observations of the H i gas content show evidence and importance of galaxy environmental processing in groups (Verdes-Montenegro et al., 2001; Kilborn et al., 2005; Pisano et al., 2011; Hess & Wilcots, 2013; Catinella et al., 2013; Brown et al., 2017; Džudžar et al., 2019b). How much environmental processing is important, and to what extent it has progressed in the H i-rich and late-type dominated groups is still an open question.

The Survey of Ionization in Neutral Gas Galaxies (SINGG) imaged a sample of 468 H i detections from the Hipass survey in the R-band and H$\alpha$ narrow-band (Meurer et al., 2006). Inspecting SINGG fields¹¹1Field-of-view$\sim$15′ Sweet et al. (2013) discovered 15 fields that have 4+ emission line galaxies within the 15′ beam of the Parkes H i detections, catalogued them and denoted them ‘Choirs’. Such selection overcomes optical-selection biases, however it introduces selection effects on the distance and H i mass. Due to the field-of-view of the SINGG fields, selection of the Choir groups is prone towards H i-rich systems at distance $\geq$30 Mpc (Sweet et al., 2013).

Galaxies in the Choir groups have similar properties as the single galaxies in SINGG based on their R-band radius, R-band surface brightness, H$\alpha$ equivalent width and specific star formation rate (Sweet et al., 2013). The mean group projected size is around two times larger than that of Hickson Compact Groups and around 10 times smaller than Garcia (1993) groups. Based on the total group H i content, Sweet et al. (2013) found that Choir groups are in an early stage of a group assembly: they have an average star formation efficiency, and they are not significantly H i-deficient.

Sweet et al. (2014) probed the gas-phase metallicity relation of galaxies within H i-rich Choir groups. They found two sub-populations of galaxies: metal-poor dwarfs and metal-rich giants in the metallicity-luminosity plane. Sweet et al. (2014) also found that the metallicity of the dwarf galaxies depends on the group membership and group H i gas content: i) dwarf galaxies in Choir groups have lower metallicity than those in isolation; ii) at low luminosity, metallicity has a higher dispersion which indicates difference in H i content and environment.

Choir groups were also used to find a population of galaxies that are formed from the tidal tail material of the interacting galaxies - tidal dwarf galaxies. Sweet et al. (2014) found three strong tidal dwarf galaxy candidates: J0205-55:S7, J0400-52:S8 and J0400-52:S9. Moreover, Sweet et al. (2016) observed 22 star-forming dwarf galaxies in four Choir groups with DEIMOS multi-object spectroscopy and analysed the galaxy kinematics, finding two new tidal dwarf galaxy candidates: J1051-17:g11 and J1403-06:g1. The rotation curves of the observed dwarf galaxies are disturbed due to recent interactions (Sweet et al., 2016). Moreover, they found that as much as half of the sample may be affected by the tidal interactions with neighbouring galaxies.

In Džudžar et al. (2019a) we analysed the integrated H i properties of 27 galaxies within nine Choir groups and compared them to a sample of isolated galaxies. We explored the H i mass fraction, specific star formation rate, star formation efficiency, H i deficiency, H i mass-size relation and stability parameter of the group galaxies with respect to the isolated galaxies. We found that the majority of central galaxies are within 2$\sigma$ scatter of the isolated galaxies, comparing H i mass fraction, the specific star formation and star formation efficiency. Satellite galaxies were shown to have lower H i mass fractions which is a possible indication that these galaxies experienced tidal stripping (Džudžar et al., 2019a). Furthermore, we determined that seven gas-rich Choir galaxies lie on the f${}_{\textrm{atm}}$-q relation (Obreschkow et al., 2016) indicating that there are no strong environmental influences on these galaxies.

The motivation for this paper is to analyse the environmental processing in the H i-rich Choir groups, using the resolved H i content of member galaxies. For each group from our sample, we present the H i column density maps, kinematic maps and the global environment for each group. We present the sample in the context of the parent SINGG and grandparent Hipass sample, and several galaxy samples from the literature. We discuss the properties of the H i in galaxies and in the tidal streams.

This paper is organised as follows: Section 2 describes our sample of Choir groups and their H i mapping, and auxiliary data used in this work. Section 3 presents our main results in this work: Choir galaxies and their H i-mass fraction in context, the H i distribution in Choir galaxies - highlighting H i-deficient and H i-excess galaxies. This section also presents the mid-infrared properties of Choir galaxies, as well as the Choir groups in context of Hickson compact groups and their evolution. Section 4 presents the global environment around Choir groups. Section 5 discusses our results and summarises our conclusions. Appendix B describes the Choir group properties, Appendix H describes our Choir groups individually. We present H i intensity maps, H i kinematics, H i spectra and global group environment for each Choir group in Appendix I.

Throughout this paper the adopted cosmology is H₀ = 70 km s^-1 Mpc^-1, $\Omega_{\Lambda}$ = 0.7 and $\Omega_{\text{m}}$ = 0.3. Mid-infrared star-formation follows the Cluver et al. (2017) prescription. The WISE photometric calibration is described in Jarrett et al. (2011). Presented H i maps and spectra use velocities in the optical convention. The group distances are based on the multipole attractor model (Mould et al., 2000), as in (Meurer et al. 2006; Sweet et al. 2013 and Džudžar et al. 2019a).

We processed the H i data on the Green II and OzSTAR Swinburne supercomputers. ATCA and VLA data were reduced with the standard procedures in MIRIAD - Multichannel Image Reconstruction, Image Analysis and Display (Sault et al., 1995) and CASA - Common Astronomy Software Applications (McMullin et al., 2007) software packages, and we summarise the procedures below. In Džudžar et al. (2019a) we presented the integrated properties obtained from our H i mapping. In this work we present the full maps of Choir groups including both: moment 0 - the H i total intensity maps; and moment 1 - the velocity field maps.

The ATCA data reduction was carried out as follows. Data were loaded with the task atlod. Raw data were flagged for radio interference using the tasks uvflag, blflag and pgflag. After the flagging we performed bandpass, flux and phase calibrations using the following tasks: mfcal, gpcal, gpboot and mfboot. For bandpass and flux calibration we used the standard calibration source PKS 1934-638, while phase calibrators are given in Table 1 as well as in Table 1 in Džudžar et al. (2019a). After calibrations we performed continuum subtraction with uvlin by fitting it to the line-free channels. This procedure was performed on each of the array configurations and then all visibilities from the same source were combined using invert to form a dirty cube (excluding antenna 6), using Briggs’ robust parameter of 0.5 and a channel width of 5 km s^-1. The obtained dirty data cubes were then cleaned using the task clean, restored using the task restor. The primary beam correction was done with the task linmos. We determined the RMS in the data cube with the tasks imstat and imhist.

We measured the galaxy H i properties using mbspect and obtained the moment 0 and moment 1 maps with the task moment, applying a 3$\sigma$ clipping. We used kvis from the Karma library (Gooch, 1996) for initial visualization, while exported moment maps were visualized using python. We present the H i-intensity maps, H i velocity fields and spectra in Appendix I.

The VLA data reduction was carried out as follows. Data were cleaned from radio interference using task flagdata. Bandpass, flux and phase calibration was performed using the following tasks: setju, gaincal, bandpass, and fluxscale. We then split the data using task split and performed continuum subtraction using line-free channels with the task uvcontsub. We produced the H i data cubes with the task clean, using natural weighting (for J1026-19 and J1059-09) and Briggs’ robust parameter of of 0 (for J1408-21) and a channel width of 5 km s^-1. With the task impbcor we corrected for the primary beam, and we created moment maps with the task immoments. We used viewer for initial visualization, while exported moment maps were visualized using python and presented in Appendix I.

We make use of imaging from the Wide-field Infrared Survey Explorer (WISE; Wright et al. 2010) to obtain integrated mid-infrared photometry in the four WISE bands: W1 (3.4$\mu$m), W2 (4.6$\mu$m), W3 (12$\mu$m) and W4 (22$\mu$m). Sources are measured using the same procedure as detailed in Jarrett et al. (2013); Jarrett et al. (2019), and the resulting photometry used to derive stellar masses, star formation rates and mid-infrared colour properties of the Choir galaxies (see Section 3.4 for details).

We make use of the optical observations of Choir groups which were carried out with the Dark Energy Camera (DECam) on the CTIO Blanco 4-m telescope in the g, r, i and z bands obtained for program AAT/13A/02 (PI S. Sweet). In addition, we use optical imaging from Dark Energy Survey Data Release 1 (g, r and i-band) for two Choir groups: HIPASS J0205-55 (partially covered by DES) and HIPASS J0400-52 (as used in Džudžar et al. 2019b). We also make use of the Digitized Sky Survey (DSS) imaging.

To place our sample of Choir groups in the wider context, we compare it to several galaxy samples from the literature:

• •

  HICAT catalogue (Meyer et al., 2004; Zwaan et al., 2004) from the H i All Sky Survey (Hipass Barnes et al. 2001), which is the parent sample of the SINGG survey. We are using HICAT+WISE catalogue (HI-WISE, $\sim$2300 galaxies) with the updated WISE photometry from Parkash et al. (2018).

• •

  SINGG – Survey of Ionization in Neutral Gas Galaxies (Meurer et al., 2006) comprised of 468 galaxies, and it is a parent sample of Choir groups.

• •

  AMIGA – Analysis of the Interstellar Medium of Isolated GAlaxies (Verdes-Montenegro et al., 2005). Data of the AMIGA sample is obtained combining two catalogues: H i masses from Jones et al. (2018) and stellar masses from Fernández Lorenzo et al. (2013) for a total of 342 galaxies.

• •

  xGASS – The extended GALEX Arecibo SDSS Survey (Catinella et al., 2018), a gas fraction limited survey of $\sim$1200 galaxies.

• •

  HCG – Hickson Compact Groups (Hickson, 1982). We use data of 72 HCGs (Verdes-Montenegro et al., 2001) and their H i properties.

• •

  H i-deficient galaxies from Dénes et al. (2014) and Murugeshan et al. (2019).

• •

  H i-excess galaxies from Lutz et al. (2018).

Within 13 Choir groups presented in this paper, Sweet et al. (2013) catalogued 80 galaxies. We have detected H i emission in 44 galaxies: nine of these are obtained from the literature and 35 with our observations, including seven new detections (hereafter, annotated with A) which are outside of the field of view in SINGG. Therefore, we obtained the H i emission in 50.6% of galaxies, and now we examine the H i non-detections.

In Figure 2 we show the stellar mass versus distance, for Choir galaxies. Choir galaxies detected in the H i are shown with the filled coloured stars (where colour denotes their H i mass), while open circles are Choirs not detected in H i. In the mass regime M${}_{\star}\leq$10⁹ M_⊙, we detect four out of 23 galaxies in H i emission. This is not surprising as the majority of these non-detections are small compact dwarf galaxies. In Figure 2 we also show that we have H i  non-detections in the stellar mass range 10⁹–10¹¹ M$\odot$, well within the range of the H i detected galaxies. These galaxies have a sufficiently high stellar mass, and angular size, to be above our H i detection threshold if they have normal amounts of H i, indicating that they are H i-deficient galaxies - they have on average eight times less H i than the average galaxy of the similar stellar mass (they are marked with the grey diamonds in Figure 2, and see Appendix E for details). These galaxies are: J1250-20:S4 [Irrs], J1059-09:S4 [SB0], J1408-21:S2 [S]; J1408-21:S4 [S0], J1026-19:S6 [S0], J0400-52:S4 [Scd], J0400-52:S5 [SB pec] and J0400-52:S6 [SB pec], morphology adopted from Sweet et al. (2013). Choir J0400-52 has three galaxies that are H i-deficient; as this group is infalling into a cluster (Sweet et al., 2013; Džudžar et al., 2019a), a lower/deficient H i content in these galaxies is not surprising. See Table 6 for details on these H i-deficient galaxies.

The H i mass fraction versus stellar mass relation (M${}_{\mbox{\sc H\thinspace{i}}}$/M_⋆ vs M_⋆) for Choir galaxies is shown in Figure 3. The H i mass fraction decreases with the stellar mass and has a sizeable scatter (Catinella et al., 2013, 2018). It is still an open question what drives the scatter in this relation, whether it is environment or internal galaxy properties, or a combination of both.

We show the H i-mass fraction of Choir galaxies in the wider context, comparing it to parent SINGG (Meurer et al., 2006), grandparent HIPASS sample using HI-WISE data from the (Parkash et al., 2018), AMIGA isolated galaxy sample (Verdes-Montenegro et al., 2005), xGASS sample (Catinella et al., 2018), and to H i-deficient (Dénes et al., 2014; Murugeshan et al., 2019) and H i-excess (Lutz et al., 2018) galaxy samples.

In this section we present and discuss the H i distribution in Choir group galaxies in order to determine whether environmental processing has an impact on it. We highlight six Choir galaxies which are respectively at the largest distance below (three most H i-deficient galaxies) and above (three most H i-excess galaxies) the running median line of the Hipass galaxies, representing most extreme outliers (see Figure 3). We track these galaxies in the following analysis.

In Džudžar et al. (2019a) we have shown that Choir galaxies follow the established H i size-mass relation (e.g. Wang et al. 2016). We present the stellar size-mass relation in Figure 5. We show the median of the stellar mass-size relation of the Choir parent SINGG sample and its 16th and 84th percentile. Moreover, we compare SINGG size-mass relation to the relation for the late-type galaxies from the Galaxy And Mass Assembly (GAMA) survey, where they are overlapping at stellar masses above 10^9.5 M_⊙ (Lange et al. 2015, relation for r-band). We find six (out of 69) Choir galaxies lie outside the SINGG percentiles and GAMA scatter. This infers a potential underlying morphological difference resulting in these outliers.

We compare the ratio of the H i to optical radii of Choir galaxies to see if we can quantify the difference between the H i-excess and H i-deficient galaxies (see Figure 6). We use the effective R-band radii obtained from the SINGG survey (Meurer et al., 2006). The optical disc sizes are often compared to the H i disc sizes measured at 1 M_⊙pc^-2 (Wang et al., 2016), however, in our most H i-deficient galaxies the H i intensity does not reach this density, thus we use a column density of 3$\times$10¹⁹cm^-2 and measure along the minor axis of the synthesised beam to avoid beam smearing effects (some of these measurements are upper limits, as indicated in Table 7, and they are in agreement with analytically derived H i sizes from the H i size-mass relation by Wang et al. 2016). With this approach, we are comparing the extent of the low H i column density in galaxies.

We find that 12 (out of 15) Choir galaxies have 2–10 times larger H i radii when compared to the r${}_{\textrm{R,eff}}$ (see Figure 6). J0205-55:S2, J0400-51:S1 and J1250-20:S1 have the largest ratio of H i radius to r${}_{\textrm{R,eff}}$, of around 20 times, for H i radii measured 3$\times$10¹⁹cm^-2 (corresponding to around 10 times for H i radii measured at 1 M_⊙pc^-2 from Džudžar et al. 2019a). The H i intensity distribution in these galaxies reaches densities larger than 50$\times$10¹⁹ cm^-2 as seen in Figure 8. Visually inspecting DECam optical images (Figure 8), we see that these galaxies have fainter stellar outer discs. Moreover, each of these galaxies have a peculiar morphology that can be connected with an interaction and possible accretion of the H i gas. J0400-51:S1 has visible faint stellar streams in the North, and we discussed in Džudžar et al. (2019b) how these may be related to a merger. J1250-20:S1 had clear ongoing multiple interactions and a stellar stream reaching out of a spiral arm (Džudžar et al., 2019a) and see Figure 21. Furthermore, we find that three out of four H i-excess galaxies are interacting, or show signs of past interactions, thus it is possible that they acquired some of their H i content from mergers.

The most H i-deficient galaxies have small H i discs with respect to their optical discs (see Figure 7). Their H i size is comparable to the synthesized beam size, thus they are not fully resolved. The H i-deficient galaxies that we mapped in H i lack high H i-column densities, when compared to the H i-excess galaxies (Figure 8); their H i-column densities on average reach $\sim$10$\times$10¹⁹ cm^-2. Six out of eight H i-deficient galaxies in our sample show irregular structure (optical or H i) which indicates that they most likely interacted (or currently interacting) with other galaxies within the group. This indicates that the tidal interaction is the primary cause of the H i-deficiency in Choir galaxies. In the next section we discuss individual H i-deficient and H i-excess galaxies.

We analyse the mid-infrared properties of Choir galaxies using “drizzled” WISE imaging Jarrett et al. (2012), in order to determine whether their star-formation activity is impacted by the group environment. The photometry was obtained from reconstructed images from WISE using the ICORE co-addition software Masci (2013). We show the WISE measurements for Choir galaxies in Table 8. For details on the WISE measurements see Jarrett et al. (2012) and Jarrett et al. (2013).

The flux of the W1 band traces the stellar light dominated by old stars, thus is it a good tracer of the galaxy’s stellar mass (Cluver et al., 2014). We obtain the W1 ‘in-band’ (relative to the Sun) luminosity from the W1-W2 colour, and derive the stellar mass based on the prescription of Cluver et al. (2014). The WISE W3 mid-infrared luminosity traces the dust-obscured star-formation activity (Cluver et al., 2017; Jarrett et al., 2019). We determine the star formation rate from the W3 band flux after removing the contribution from evolved stars, based on the prescription of Cluver et al. (2017).

We show the mid-infrared colour-colour (W1-W2 versus W2-W3) diagram of the Choir galaxies in Figure 9. It has been shown that the galaxy activity (e.g. star-forming or AGN) can be roughly determined based on galaxy’s mid-infrared colours (e.g. Jarrett et al. 2019). We see that the majority of Choir galaxies are scattered around the WISE colour-colour sequence (star-forming sequence) which was determined by Jarrett et al. (2019), who also show that W1-W2 colours well above the sequence have excess infrared emission due to host-dust-enshrouded AGN. The increase of the W1-W2 colour towards the right indicates the elevation of the W2 band emission by the hot dust from star-formation (Jarrett et al., 2019). We find several Choir galaxies with elevated W1-W2 colours with respect to the star-forming sequence which may be Seyfert galaxies with excess hot dust emission associated with the AGN torus (see annotations in Figure 9). However, none of them are within the AGN-dominated region (above the horizontal dashed line at W1-W2$\sim$0.8). The most extreme outlier is J0205-55:S8, it is a small compact dwarf galaxy. J0209-10:S1 and J0205-55:A1 are classified as Seyfert 2 galaxies (de Carvalho & Coziol, 1999; Schmitt et al., 2003), and J2318-42a:S1 as Seyfert 1 (Véron-Cetty & Véron, 2006). J1026-19:S2 is an irregular galaxy which is tidally interacting with J1026-19:S1 (see Figure 19), and J1051-17:S2 is a high surface brightness spiral galaxy (Sweet et al., 2013).

Denotation of infrared morphological galaxy classification: spheroids, intermediate discs and star-forming discs are assumed based on the colour, which was suggested in Jarrett et al. (2017). In the ‘Spheroids’ region are expected to reside galaxies with the prominent bulges, such as E or S0 galaxies, most often considered gas-depleted high mass galaxies or low burner dwarf spheroidal galaxies. Choir galaxies in this region are J1059-09:S10 (W2-W3$\sim$0 mag) and J1159-19:S4 (W2-W3$\sim$1 mag) and these are respectively classified as dwarf spheroidal [dS] and dwarf [D] galaxy by Sweet et al. (2013). The ‘Intermediate discs’ are described as Milky Way type galaxies, with prominent bulges and semi-quiescent discs (1.5$<$W2-W3$<$3). Galaxies in this region are possibly undergoing quenching of their star formation (Jarrett et al., 2019). The ‘Star-forming discs’ are galaxies with W2-W3$>$3, and these galaxies have ongoing active star-formation.

In Figure 10 we show the star-formation properties of the Choir galaxies. The dashed lines represent constant specific star-formation rates, and we see that the Choir galaxies are spread between 10^-9 and 10^-11 yr^-1, thus we do not find them to be significantly quenched. We use the HI-WISE data from Parkash et al. (2018) and show the running median of the star formation rate in bins of the stellar mass. We show the running median line in Figure 10 with the blue solid line, while the shaded region shows the 16th and 84th percentile. For comparison we show the best fit line to the SINGS/KINGFISH sample, used for the calibration of the star formation rate from 12$\mu$m (Cluver et al., 2017), and we also show the star-formation biased ‘sequence’ of the the equatorial GAMA G12 field (Jarrett et al., 2017).

Three galaxies with a specific star formation rate of $\sim$10^-11 yr^-1 are: J1159-19:S4 a dwarf galaxy; J1250-20:S4 an Irr galaxy, and J0209-10:S4 an SApec galaxy. All three of these galaxies are experiencing tidal interactions, J1159-19:S4 – as seen in H i, while J1250-20:S4 and J0209-10:S4 as seen in H i and optical.

Based on the group compactness, Choir groups are most similar to HCGs, as shown by the projected separation between the two most luminous galaxy members (Sweet et al., 2013). Therefore, in order to place the Choirs into the broader context of groups, we compare them to Hickson Compact Groups (HCG; Hickson 1982; Verdes-Montenegro et al. 2001). In Figure 11 we show group H i mass versus group deficiency of Choir groups and HCGs. For each Choir, the total group H i mass is taken from Hipass, while the group H i deficiency is taken from Sweet et al. (2013). We find that Choir groups have a high mass and a low deficiency values, while HCGs are predominantly tailing towards lower group H i masses and a higher deficiency values. It is known that HCGs are comprised of galaxies with varying morphologies, and many of these galaxies are early-type galaxies. Based on the morphological type noted in (Hickson, 1982) and available H i data from Verdes-Montenegro et al. (2001) we select HCGs that are comprised of only late-type galaxies: HCG2, HCG16, HCG31, HCG38, HCG80, HCG88, HCG89, HCG100 for comparison. We find that these late-type dominated HCGs are more similar to the Choir groups (see Figure 11).

In Džudžar et al. (2019a), we suggested that the Choir groups would fit into phase 1 and phase 2 of the proposed H i evolutionary scenario of HCGs (Verdes-Montenegro et al., 2001). We place the phase classification for each Choir group in Table 10. In summary the Verdes-Montenegro et al. (2001) evolutionary scenario of the gas content in HCGs is as follows: i) Groups in the phase 1 have an average gas content and a low level of galaxy interaction. ii) Groups in phase 2 contain H i tidal features. iii${}_{\textrm{a}}$) Groups in the phase 3a do not have H i within galaxies (it can be in the intragroup medium). iii${}_{\textrm{b}}$) Groups in the phase 3b contain the H i gas in the form of a cloud that envelopes entire group. Over time, it is likely that majority of galaxies in Choir groups will continue to deplete their gas content through star formation and interactions due to low velocity dispersion (tidal interactions, mergers), and they will become progressively more H i deficient.

Comparing gas depletion timescales from Džudžar et al. (2019a) and the crossing time for Choir groups (see Table 3), we find that the crossing time is on average shorter than the H i depletion time. Therefore, galaxy mergers that will happen in the future will most likely be gas-rich mergers²²2Under assumptions that galaxies will go thought the gravitational potential of the group.. To investigate which Choir group has conditions for galaxy mergers, we use approximated conditions from Pfister et al. (2020) for the galaxy pair mergers. Utilizing the Horizon-AGN simulation, Pfister et al. (2020) formulated a threshold on the projected distance d${}_{\textrm{th}}$ and redshift difference $\Delta$z${}_{\textrm{th}}$ below which galaxy pairs are likely to merge. To adapt the Pfister et al. (2020) relations for Choir groups, we use the group projected radius and radial velocity dispersion. Comparing the Choirs group velocity dispersion and the Pfister et al. (2020) criteria for optimal $\Delta$z${}_{\textrm{th}}$, all Choir groups have optimal condition for galaxy mergers. Whereas, looking at the group projected radii, we find that seven groups do not satisfy imposed criteria, two are on the threshold and six satisfy imposed criteria. Even though this analysis is rough, since groups have more massive gravitational potential than galaxy pairs, with such approach we obtained lower limits on the given conditions for galaxy mergers. Choir groups J1026-19, J1059-09 and J1408-21 are above optimal d${}_{\textrm{th}}$ threshold, however, they contain galaxies that are already interacting, while J0400-52 is infalling into a cluster. To better constrain minimal galaxy separation for mergers, this work requires further studies of Choir group analogues in simulations.

Based on the mentioned parameters, we suggest that some of the Choir groups will go through a phase of compact groups or they will end up as an optical fossil group, depending on the global environment around them. A fossil group is usually defined as a galaxy group with a single dominant elliptical galaxy with an X-ray luminosity of L${}_{\textrm{X,bol}}\geq 10^{42}$ h${}_{50}^{-2}$ erg s^-1 and with at least 2.0 magnitude difference from the second brightest group galaxy (e.g. Jones et al. 2003). We define an optical fossil group, where the magnitude difference between the brightest galaxy and second brightest galaxy is larger than 2 magnitudes, but we do not impose that they are X-ray bright as in the definition of fossil groups by Ponman et al. (1994).

Nine of the Choir groups have galaxies in the neighbourhood and thus they have a possibility of forming a HCG like group in the future, assuming that these galaxies will ‘fall’ towards the same potential well (see Section 4 and Appendix I). These groups are: J0205-55, J0209-10, J0443-05, J1051-17, 1059-09, J1159-19, J1403-06, J2027-51 and J2318-42a.

Choir J1159-19 is a compact group which contains a one armed spiral galaxy and three dwarf galaxies. J1159-19 meets the criteria for an optical fossil group. All galaxies within this group are interacting, as seen in their H i distribution (see Figure 27) and they will most likely merge in the future. A further five Choirs lie in a fairly isolated environment are more likely to create an optical fossil group, similar to the current state of J1159-19. In these cases, there are fewer galaxies that can merge and thus they will not be able to satisfy criteria to be classified as HCGs. These groups are: J0258-74, J1026-19, J1250-20, J1408-21 and J1956-50.

One Choir group, J0400-51, is infalling into a cluster. Džudžar et al. (2019b) proposed that the gas-rich galaxy within this group (ESO156-G029) will be under the effects of ram-pressure once the galaxy approaches closer to the cluster centre. Moreover, we used this galaxy as an example to show that group pre-processing is important in its evolution.

We analyse the wider environment of the Choir groups to determine the environment outside the 15′ of the SINGG field of view. We made a script in python to query for sources around the Hipass detection in NASA/IPAC Extragalactic Database (NED). We use RA, Dec and the recessional velocity V${}_{\mbox{\sc H\thinspace{i}}}$ of Choir groups from Sweet et al. (2013), and make a query typically $\sim$4 degrees around each Choir group. The initial result is sub-sampled for only ‘G’ objects i.e. only galaxies, as there are sources which mark position of groups, triplets etc. The obtained catalogue is cross-matched with the 2MASS Extended Source Image Server to obtain the Ks-band total magnitude in order to make a magnitude limited sample.

We use galaxies with known Ks-magnitude, recessional velocity between $\pm$1000 km s^-1 from the Choir group velocity, and $<$0.5 degree projected separation from the HIPASS centre, to compute the Ks weighted mean position of the group centroid. Using this sample of galaxies we re-compute the projected distances for each galaxy from the group centroid. We use this data to map the galaxies with their velocity difference (with respect to the Choir group) around all Choir groups (see panel c in Appendix I). We also use the projected angular separation versus the velocity difference of each galaxy from the group mean velocity to show Choir caustics curves in order to map the group potential and find potential galaxies that will merge with Choir groups (see panel d in Appendix I).

We find Choirs in low density environments (J0258-74, J1026-19, J1250-20,J1408-21 and J1956-50), Choirs with nearby, possible infalling galaxies (J0205-55, J1051-17) and Choirs which are potentially part of a larger structure (J0400-52, J1159-19 and J2027-51). We present results for each Choir group individually in Appendix H and Appendix I: panel c and d in Figures.

We explore whether the Choir groups have a ‘special’ position in the cosmic web. Janowiecki et al. (2017) suggested that the H i-rich central galaxies in the group environment are H i-rich due to gas accretion from cosmic web filaments, as groups can be positioned in the intersection of those filaments. Kleiner et al. (2017) used the 6 degree Field Galaxy Survey (6dFGS) and the Discrete Persistent Structures Extractor (DisPerSe) to map the filaments in the local Universe. Furthermore, combining filaments and data from Hipass, they found that galaxies near filaments, with M$\star\geq$10¹¹ M$\odot$, have higher H i mass fractions than the control sample. Utilizing catalogue of the filaments by Kleiner et al. (2017), we derive the distances to the nearest filament for Choir galaxy groups, see Table 2.

Exploring 14 Choir galaxy groups (J0400-52 is excluded as it is near galaxy cluster), we find that seven are within $\leq$1 Mpc, one is at $\sim$2 Mpc and six are at $\geq$4 Mpc from the nearest filament. We find only two Choir groups near the intersection of filaments: J1408-21 and J1059-09. Based on the H i-mass fraction, central galaxies in J1408-21 and J1059-09 have average gas content. Furthermore, we find two Choir groups that are more than 10 Mpc away from the nearest filament: J0258-74 and J1956-50. These groups can be classified as being in a void, as distances of $\sim$10 Mpc are considered as voids (Kuutma et al., 2017). The central galaxy in J0258-74 has an average H i-mass fraction, while central in J1956-50 has a higher than average H i-mass fraction. We conclude that Choir groups span a range of global environments, from being in a void to being near an intersection of filaments, thus they are not within a unique position in the cosmic web. While gas accretion from the cosmic web may be occurring in some Choir groups, it is not shown to drastically enhance the H i content as the Choir groups with similar H i properties span all global environments.

In summary, the Choir groups whose H i content we mapped can be classify into several broad categories:

• •

  Isolated groups, embedded in the low density global environment: J0258-74, J1026-19, J1250-20, J1408-21, J1956-50.

• •

  Cluster infalling group: J0400-52.

• •

  Groups with tidal interactions: J1159-19, J1059-09, J1026-19, J1408-21, J1250-20, J0400-52.

• •

  Merging subgroups: J0205-55 (composed of J0205-55a and J0205-55b), J1051-17

• •

  Potentially part of the larger structure: J2027-51, J1159-19.