Effect of local and large-scale environments on nuclear activity and star formation
M. Argudo-Fernández, S. Shen, J. Sabater, S. Duarte Puertas, S. Verley, X. Yang
AAstronomy & Astrophysics manuscript no. AGNpairs_160513_ac © ESO 2018October 3, 2018
Effect of local and large-scale environments on nuclear activity andstar formation
M. Argudo-Fernández , , S. Shen , , J. Sabater , S. Duarte Puertas , S. Verley , , and X. Yang , Key Laboratory for Research in Galaxies and Cosmology, Shanghai Astronomical Observatory, Chinese Academy of Sciences, 80Nandan Road, Shanghai, China, 200030 Universidad de Antofagasta, Unidad de Astronomía, Facultad Cs. Básicas, Av. U. de Antofagasta 02800, Antofagasta, Chile Key Lab for Astrophysics, Shanghai, 200234, China Institute for Astronomy, University of Edinburgh, Edinburgh EH9 3HJ, UK Instituto de Astrofísica de Andalucía (CSIC) Apdo. 3004, 18080 Granada, Spain Departamento de Física Teórica y del Cosmos, Universidad de Granada, 18071 Granada, Spain Instituto Universitario Carlos I de Física Teórica y Computacional, Universidad de Granada, 18071 Granada, Spain Center for Astronomy and Astrophysics, Shanghai Jiao Tong University, Shanghai 200240, China IFSA Collaborative Innovation Center, Shanghai Jiao Tong University, Shanghai 200240, ChinaReceived 2 February 2016; accepted 18 May 2016
ABSTRACT
Context.
Active galactic nuclei (AGN) is one of the main drivers for transition from star-forming disk to passive spheroidal galaxies.However, the role of large-scale environment versus one-on-one interactions in triggering di ff erent types of AGN is still uncertain.We present a statistical study of the prevalence of the nuclear activity in isolated galaxies and physically bound isolated pairs. Aims.
For the purpose of this study we considered optically and radio selected nuclear activity types. We aim to assess the e ff ect ofone-on-one interaction on the fraction of AGN and the role of their large-scale environment. Methods.
To study the e ff ect of one-on-one interaction on the fraction of AGN in isolated galaxy pairs, we compare with a sample ofisolated galaxies homogeneously selected under the same isolation criterion. We examine the e ff ect of the large-scale environment bycomparing with control samples of single galaxies and galaxy pairs. To quantify the e ff ects of local and large-scale environments weuse the tidal strength parameter. Results.
In general we found no di ff erence in the prevalence of optical AGN for the considered samples. For massive galaxies, thefraction of optical AGN in isolated galaxies is slightly higher than that in control samples. Also the fraction of passives in highmass isolated galaxies is smaller than in any other sample. Generally, there is no dependence on optical nuclear activity with localenvironment. On the other hand, we found evidence that radio AGN are strongly a ff ected by the local environment. Conclusions.
Optical AGN phenomenon is related to cold gas accretion, while radio AGN is related to hot gas accretion. In thiscontext, there is more cold gas, fuelling the central optical AGN, in isolated systems. Our results are in agreement with a scenariowhere cold gas accretion by secular evolution is the main driver of optical AGN, while hot gas accretion and one-on-one interactionsare the main drivers of radio AGN activity.
Key words. galaxies: active – galaxies: formation – galaxies: evolution – galaxies: interactions – radio continuum: galaxies
1. Introduction
The environment in which a galaxy resides plays an importantrole on its formation and evolution. Galaxies su ff er intrinsic andsecular evolution processes (i.e. nature processes), but they arealso exposed to the influences of their local and large-scale en-vironments (i.e. nurture processes) (Casado et al. 2015). Start-ing from the well known morphology-density relation (Dressler1980), other properties such as stellar mass, size, or colourare influenced by environmental processes (Peng et al. 2010;Grützbauch et al. 2011; Calvi et al. 2012; Peng et al. 2012;Boselli & Gavazzi 2014). Moreover, nuclear activity is somehowa ff ected by galaxy environment (Kau ff mann et al. 2004; Choiet al. 2009; Sabater et al. 2013).Active galactic nuclei (AGN) also plays an important rolein galaxy formation and evolution. Observations have led tothe interpretation that the main driver for transition from star-forming disk to passive spheroidal galaxies is through AGN(Kau ff mann et al. 2003a). A galaxy is revealed as an AGN when its central black hole (BH) grows through mass accretion, lib-erating in the process a huge amount of energy (see Salpeter1964; Lynden-Bell 1969; Soltan 1982; Fabian 1999; Alexander& Hickox 2012). There is some evidence that galaxies are morelikely to host an AGN when they are interacting with a neighbour(e.g. Ellison et al. 2011). Since the most evolved and massivegalaxies reside in clusters at the present day (Abell 1958; Baldryet al. 2004), the environment associated to the large-scale struc-ture (LSS) is therefore also likely to have a significant e ff ect onthe triggering of AGN activity. However the physical processesresponsible for triggering AGN, as well as the role of the envi-ronment on the fuelling of BHs, are still uncertain.Is AGN activity connected to environment? There has beensome disagreement in the literature regarding the connection be-tween environment and nuclear activity. Di ff erences are mainlydue to sample selections and the diverse definitions of environ-ment. The statistical studies developed by Sabater et al. (2013,2015) suggest that large-scale environment and galaxy interac-tions play a fundamental but indirect role in AGN activity (by Article number, page 1 of 13 a r X i v : . [ a s t r o - ph . GA ] M a y & A proofs: manuscript no. AGNpairs_160513_ac influencing the gas supply) and that the dependence on AGN lu-minosity is minimal. On the other hand, the results from deep im-ages by Hong et al. (2015) suggest that luminous AGN activity isassociated with galaxy merging. These di ff erences could be ex-plained since the results of the first studies, based on Sloan Digi-tal Sky Survey (SDSS, York et al. 2000; Strauss et al. 2002) sev-enth data release (DR7; Abazajian et al. 2009) data, might con-sider merger systems as single galaxies and also be a ff ected bythe well know fibre-collision problem for close objects (Strausset al. 2002).Even if it can be di ffi cult reconciling the results from onestudy with the results from another study, there is some consen-sus that secular processes may be much more important in driv-ing the black hole growth than previously assumed (McAlpineet al. 2015; Sabater et al. 2015), where the request for AGN trig-gering is an abundant supply of central cold gas, regardless of itsorigin. In the case of high luminosity AGN, major mergers ap-pear to be the main driver (Ellison et al. 2011; Ramos Almeidaet al. 2012; Kaviraj et al. 2015; Manzer & De Robertis 2014;Satyapal et al. 2014; Chiaberge et al. 2015).There are two modes of AGN activity: the ’quasar-mode’(Shakura & Sunyaev 1973), or ’cold-mode’, and the ’radio-mode’ (Hine & Longair 1979), or ’hot-mode’. In ’cold-mode’the AGN radiates across a broad multi-wavelength range (in-cluding optical and radio), while in ’hot-mode’ AGN are mainlydetectable in radio, due to the emission of their jet. The two ac-cretion modes have di ff erent associated feedback e ff ects in thehost galaxy (Kau ff mann et al. 2003a; Best et al. 2005a), and theyare powered by accretion of di ff erent material that can be con-nected to the environment (Hardcastle et al. 2007; Tasse et al.2008), but the precise origin of these di ff erences e ff ects remainsunclear (Best & Heckman 2012).The aim of the present study is to accurately measure thefraction of optical and radio AGN activity that is triggered by ex-ternal or internal processes. With the purpose of determining theimportance of secular evolution versus one-on-one interactionsin triggering di ff erent types of AGN activity, we select samplesof both isolated galaxies and isolated pairs. In this study we fo-cus on isolated systems because any di ff erence in the AGN frac-tion would be directly related to the addition of one companion.Also, by carefully selecting control samples of galaxy pairs andsingle galaxies, we go one step further and explore the e ff ect ofthe large-scale environment versus one-on-one interactions onthe AGN prevalence.This study is organised as follows. In Sect. 2 we describe thesamples used in this work as well as the selected AGN classifi-cation methods and the parameters used to quantify the environ-ment. We present our results in Sect. 3 and the associated discus-sion in Sect. 4. Finally, a summary and the main conclusions ofthe study are presented in Sect. 5. Throughout the study, a cos-mology with Ω Λ = . Ω m0 = .
3, and H =
70 km s − Mpc − is assumed.
2. Data and methodology
To study the e ff ect of one-on-one interaction on the fraction ofAGN, we compare isolated galaxy pairs with a sample of isolatedgalaxies homogeneously selected under the same isolation crite-rion. To understand the e ff ect of the large-scale environment, weselect control samples composed of single galaxies and galaxypairs that can be found in any environment, from clusters orgroups to voids. We use the sample of isolated galaxies and isolated pairs com-piled by Argudo-Fernández et al. (2015b) from the SDSS-DR10(Ahn et al. 2014). Isolated galaxies and central galaxies in thepairs (the brightest of the pair by definition) are selected in avolume limited sample redshift range 0 . ≤ z ≤ . ≤ m r ≤ .
7, where m r is the SDSS model magnitude inthe r -band. This criterion allows the second galaxy in the pairto be at least 2 orders of magnitude fainter within the range ofspectroscopic completeness of the SDSS main galaxy sample at m r , Petrosian < .
77 mag (Strauss et al. 2002).Argudo-Fernández et al. (2015b) used a three dimensionalisolation criterion, based on projected distances on the sky andredshift. The systems are isolated with no neighbours in a vol-ume of 1 Mpc projected distance within a line-of-sight veloc-ity di ff erence of ∆ (cid:51) ≤
500 km s − . After defining isolation,they followed a similar method as in Argudo-Fernández et al.(2014) to identify the physically bound isolated pairs. For thecentral galaxy in isolated pairs, Argudo-Fernández et al. (2015b)found an over-density of neighbours in the 2D distribution of ∆ (cid:51) and distance d . This over-density indicates that those neigh-bours are likely to be physically connected. The distribution of ∆ (cid:51) for those neighbours follows a Gaussian distribution. Theneighbour galaxies within ∆ (cid:51) ≤ σ (160 km s − ) show also atendency to be located within the first 450 kpc from the cen-tral galaxy. Conversely, neighbour galaxies at higher ∆ (cid:51) and d would be associated to the underlying large-scale distribution ofgalaxies (Argudo-Fernández et al. 2015b). They found 3702 iso-lated galaxies, hereafter SIG (SDSS-based Isolated Galaxies),and 1240 isolated pairs physically bound at projected distancesup to d ≤
450 kpc within ∆ (cid:51) ≤
160 km s − , hereafter SIP(SDSS-based Isolated Pairs). The SIG and SIP samples repre-sent about 11% and 7% of the galaxies in the local Universe( z ≤ . ∆ (cid:51) of the SIP sample is d (cid:39)
215 kpc and ∆ (cid:51) (cid:39)
65 km s − . The average stellar mass ratio in isolatedpairs ( M (cid:63) B M (cid:63) A , where A corresponds to the central galaxy and B tothe faintest galaxy in the pair) is M (cid:63) B M (cid:63) A (cid:39) .
30, in the range0 . (cid:46) M (cid:63) B M (cid:63) A (cid:46) .
00 (see Argudo-Fernández et al. 2015b, forfurther details).
The control samples are based on the catalogue of groups com-piled by Yang et al. (2007), which is based on the NYU-VAGC(Blanton et al. 2005) updated with SDSS-DR7 data. Yang et al.(2007) developed a halo-based group finder that is optimized forgrouping galaxies that reside in the same dark matter halo, in-cluding isolated galaxies in small mass haloes. Group membersare complete to M r ≤ − . z ≤ . Stellar masses in the CIG and SIP samples come from the fitting tothe spectral energy distribution on the five SDSS bands using the routinekcorrect (Blanton & Roweis 2007) and the relation between the stellarmass-to-light ratio and color of Bell et al. (2003).Article number, page 2 of 13. Argudo-Fernández et al.: E ff ect of local and large-scale environments on nuclear activity and star formation Fig. 1.
Upper panel:
Distribution of the stellar masses of the SIGand SDSS single galaxies (red solid and magenta dash-dotted his-tograms, respectively), and for central SIP and SDSS pairs (greendashed and blue dotted histograms, respectively). The vertical blackdashed lines correspond to the selected stellar mass range for the studyat 10 . ≤ log(M (cid:63) ) ≤ . (cid:12) ]. Lower panel:
Distribution of the red-shift of the SIG and SDSS single galaxies (red solid and magenta dash-dotted histograms, respectively), and for central SIP and SDSS pairs(green dashed and blue dotted histograms, respectively), for galaxieswithin the stellar mass limited sample ( N T ). criteria, we selected one-member groups in Yang et al. (2007),hereafter the SDSS single sample, as a control sample to com-pare with SIG galaxies.The average projected separation and ∆ (cid:51) of SDSS pairs is d (cid:39)
200 kpc and ∆ (cid:51) (cid:39)
85 km s − , similar to the SIP sam-ple, where the 90% of the pairs show projected separations d (cid:46)
450 kpc and ∆ (cid:51) (cid:46)
200 km s − . The stellar mass ratioalso spans a similar range with an average value M (cid:63) B M (cid:63) A (cid:39) . ≤ m r ≤ . . ≤ z ≤ . For the purpose of this study we considered optically and radioselected nuclear activity types. In particular, we used publishedstellar masses and AGN classifications for galaxies in the SDSS-DR7 from Sabater et al. (2013), hereafter SBA13 classification.SBA13 used information about total stellar masses fromKau ff mann et al. (2003b) and optical AGN classification fromBPT diagnostic (Baldwin et al. 1981; Kau ff mann et al. 2003a).Data based on optical spectra, i.e. the stellar masses and the cor-rected emission-line fluxes used to built BPT diagrams, weredrawn from the Max Plank Institute for Astrophysics and JohnsHopkins University (MPIA-JHU ; Kau ff mann et al. 2003b;Tremonti et al. 2004; Salim et al. 2007) added value catalogue(Brinchmann et al. 2004).Radio AGN galaxies in SBA13 are considered if they areclassified as radio AGN by Best & Heckman (2012) with a ra-dio luminosity brighter than L . ≈ W m − Hz − (seeSBA13, for more details). The radio AGN classification in Best& Heckman (2012) is based on radio-continuum data from theNational Radio Astronomy Observatory (NRAO) Very Large Ar-ray (VLA) Sky Survey Condon et al. (NVSS, 1998) and the FaintImages of the Radio Sky at Twenty centimetres (FIRST, Beckeret al. 1995) data bases, and follows the techniques presented inBest et al. (2005b).Even if we take care on selecting galaxies in the controlsamples within the same volume limited as isolated galaxiesand isolated pairs, it is well known that AGN fraction dependsstrongly on stellar mass (Kau ff mann et al. 2003a; Peng et al.2010, 2012). Henceforth, to have a robust control sample forthe comparisons, we selected galaxies in the SDSS single andSDSS pairs samples with similar stellar mass and redshift dis-tributions than the SIG and SIP samples, respectively (withKolmogorov-Smirnov p-value greater than 0.80, which ensuresthat the distributions of the two samples are the same). To haveenough numbers of objects in each mass bin for each sample,we also considered galaxies with stellar masses within the range10 . ≤ log(M (cid:63) ) ≤ . (cid:12) ] (see the upper panel in Fig. 1).The final number of galaxies, in the stellar mass range consid-ered in this study, with available SBA13 classification is shownin the first row of Table 1.Due to the number of galaxies in our samples, we only sepa-rate nuclear activity into optical AGN, radio AGN, star-forming Stellar masses in the group catalogue were computed fitting the spec-tral energy distribution on the five SDSS + JHK bands using the routinekcorrect (Blanton & Roweis 2007) and the relation between the stellarmass-to-light ratio and color of Bell et al. (2003) (see Yang et al. 2007,for further details). Available at Note that stellar masses from Argudo-Fernández et al. (2015b) andYang et al. (2007) are calculated slightly di ff erent, therefore in order touse a consistent source, we consider stellar masses in SBA13.Article number, page 3 of 13 & A proofs: manuscript no. AGNpairs_160513_ac
Table 1.
Number of galaxies of each type in each sample.
Type SIP SDSS pairs SIG SDSS singleTotal 764 2251 2299 6867Optical AGN 387 1009 1153 3027LINER 152 454 482 1081Seyfert 44 96 123 266TO 191 459 548 1680SFN 169 564 587 1777Passive 208 678 559 2063Radio AGN 10 39 11 22HERG 1 2 0 0LERG 8 37 11 22
Notes.
Meaning of the di ff erent types: Total – total number of galaxiesin each sample; Optical AGN – galaxies classified as LINER, Seyfert,or transition objects (TO); SFN – star forming nuclei galaxies; Passive– galaxies with no optical nuclear activity; Radio AGN – galaxies clas-sified as HERG or LERG radio AGN with L . ≥ W m − Hz − . nuclei (SFN), and passive galaxies (in case that no optical nu-clear activity is detected). Optical AGN classification coverstransition objects (TO), Seyfert (Seyfert 1 not included), andlow-ionization nuclear emission-line region (LINER; Heckman1980) galaxies. In the case of radio AGN, this includes low-excitation (LERG) and high-excitation (HERG) radio galaxies.The number of galaxies in each sample, classified in each typeand subtypes of nuclear activity, is shown in Table 1. Note thatBest & Heckman (2012) classified radio galaxies into HERG andLERG when such classification was possible. Given that thereis only one SIP galaxy without classification we rejected thisgalaxy in the present study. Besides, the three HERG galaxiespresent in the sample were discarded during the comparisons.Henceforth, with the term radio AGN we will be consideringonly LERG radio AGN.Note that, to have a statistically significant number of galax-ies, we do not impose any limit on the [O III ] emission line lu-minosity. Some low luminosity AGN (usually LINERs) at higherredshifts could be classified as passive if their emission is notstrong enough to be detected. However, the overall e ff ect is min-imized if the redshift distribution of the samples is relatively sim-ilar as it is in our case (see lower panel in Fig. 1).Note also that AGN classification based on purely emission-line BPT diagrams are a ff ected by uncertainties (Sabater et al.2012). According to Rosario et al. (2016), these uncertaintiesare specially important for massive main-sequence local galax-ies that might be misclassified as passives. We have checked thatless than the 4% of the total number of galaxies in each sam-ple in our study would be a ff ected. Given this number, we donot expect any change in the observed trends for SFN galaxies.The results of Sabater et al. (2012) suggest that at least someof these misclassified galaxies could be classified as LINERs orSeyferts if the uncertainties were taken into account. However,given that our samples follow the same classification criteria anda relatively similar distribution in mass and redshift, the possiblee ff ect on the comparison between samples will be minimised. Argudo-Fernández et al. (2015b) also provided the isolation de-gree for isolated galaxies and central galaxies in isolated pairs.They quantified the influence of both, local and large-scale envi-
Fig. 2.
Upper panel:
Comparison of Q pair versus Q LSS , isolated pairs(green stars), SDSS single galaxies (magenta circles) and SDSS pairs(blue pluses). Note that for isolated galaxies (red triangles) there is notavailable Q pair but we consider Q pair = Q LSS for comparison purposes.The black dashed line represents the line where Q pair = Q LSS . Lowerpanel:
Schema of the sample definition, total number of galaxies ( N T ) ineach sample and number of galaxies in common ( N C ), and environmentdefinition for the SIG and SIP samples (red and green circles, respec-tively), and for control samples (blue and magenta ellipses for SDSSpairs and SDSS singles respectively). The arrows in the axis indicatesthe direction to higher values of the tidal strength. ronments, using the tidal strength parameter (Verley et al. 2007;Sabater et al. 2013; Argudo-Fernández et al. 2013, 2014).To study the influence of the large-scale environment on theAGN fraction for isolated galaxies and isolated pairs, we se-lected the tidal strength exerted by all the galaxies in the LSS up Article number, page 4 of 13. Argudo-Fernández et al.: E ff ect of local and large-scale environments on nuclear activity and star formation to 5 Mpc, Q LSS (Eq. 1). For isolated pairs, we study the influenceof the companion using the local tidal strength Q pair (Eq. 2).Galaxies in the LSS are defined within a volume of5 Mpc projected distance and line-of-sight velocity di ff erence of ∆ (cid:51) ≤
500 km s − . Then, for each galaxy, i , in the LSS at aprojected distance d LS S i , the total tidal strength on the isolatedgalaxy (SIG), or the central (brightest) galaxy in isolated pairs(SIP), is: Q LSS ≡ log (cid:88) i M LS S i M (cid:32) Dd LS S i (cid:33) , (1)where M is the stellar mass and D is the estimated diameterof the SIG / SIP galaxy. Stellar masses for galaxies in the LSS( M LS S i ) were calculated by fitting the spectral energy distribu-tion, on the five SDSS bands, using the routine kcorrect (Blanton& Roweis 2007).We quantify the e ff ect of the local environment as the tidalstrength a ff ecting the central (brightest) galaxy in a galaxy pair.Then, similarly to the definition of the Q LSS but only consideringtwo galaxies, the local tidal force exerted by the faintest B galaxyon the brightest (central) A galaxy is: Q pair ≡ log M B M A (cid:32) D A d AB (cid:33) , (2)where d AB is the projected physical distance between the galax-ies of the isolated pair.We follow the same methodology to estimate the tidalstrength for control samples, i. e. we estimate Q LSS for SDSSsingle galaxies, and Q pair for SDSS pairs. A scheme of the envi-ronment for the four samples is shown in Fig. 2. For comparisonpurpose, Q pair for SDSS single galaxies is estimated consideringtheir first nearest neighbour. In the case of the SIG sample, as thenearest neighbour is as least at 1 Mpc away, the value of the tidalstrength exerted by this neighbour is practically the same as theone exerted by the LSS, we therefore consider Q pair = Q LSS . Q LSS for the control samples is estimated as for SIG and SIP,considering their LSS up to 5 Mpc. The greater the value of Q ,the less isolated from external influence the galaxy. Therefore,as it is schematically shown in the lower panel of Fig. 2, theSIG and SIP samples have the same degree of isolation with re-spect the LSS, while control samples extend a broader range oflarge-scale environments. With respect to the local environment,SIG and SDSS singles have a similar range, while the e ff ect isstronger for central galaxies in the SIP and SDSS pairs samples.Then at fixed stellar mass, higher values of Q pair are related tocloser pairs. Since there is a strong dependence of AGN withstellar mass, it is recommendable to made a separated study indi ff erent stellar mass bins (SBA13).The black dashed line in the upper panel of Fig. 2 corre-sponds to the line where Q pair = Q LSS . When a galaxy is locatedon this line, the contributions by its local and large-scale environ-ment on the total tidal strengths are the same. Central galaxies inthe SIP sample are located above the line, which means that theirtidal strengths are dominated by their close environment. In fact,more than 95% of the total tidal strength in SIP galaxies is dueto the companion galaxy in the pair (Argudo-Fernández et al.2015b). In general SDSS pairs are also located above the line, D = α r , where r , the Petrosian radius containing 90 % of thetotal flux of the galaxy in the r -band, is scaled by a factor α = .
43 torecover the D (Argudo-Fernández et al. 2013). but there are some pairs below the line. These are pairs locatedin high density environments, mainly in clusters, surrounded bymassive and nearby galaxies which are likely a ff ecting their evo-lution.
3. Results
We aim to study the prevalence of AGN in the four samples pre-viously selected. To do this we compare the relative fraction ofeach type of AGN in isolated galaxies and in physically boundisolated pairs, with respect to the ones found in the control sam-ples. Due to the strong dependence of the prevalence of AGNwith mass of the host galaxy, both in optical (Kau ff mann et al.2003a) and radio (Best et al. 2005a), we fix a stellar mass rangein each step of the study. We considered galaxies in each sam-ple with stellar masses within 10 . ≤ log(M (cid:63) ) ≤ . (cid:12) ],as explained in Sect. 2, and we divided our studies in di ff erentstellar mass bins. The relative fraction of optical nuclear activity (SFN, opticalAGN, and passive galaxies) for each sample is shown in the up-per panels in Fig. 3. We considered seven stellar mass bins toobserve the possible trends for each sample. Note that we takecare of having a significant number of galaxies in each bin. Errorbars are given by considering binomial distribution .The fraction of passive galaxies (right panel) increases withstellar mass, while the general trend for SFN galaxies (middlepanel) is to decrease with stellar mass. The general trend forthe fraction of optical AGN up to log(M (cid:63) ) (cid:46) . (cid:12) ] is toincrease with stellar mass (left panel). We only find significantdi ff erences between samples for massive galaxies. The fractionof optical AGN is still increasing at higher masses for isolatedgalaxies while it starts to decrease in the remaining samples. Ac-cordingly, the fraction of passive isolated galaxies is lower athigher masses.Lower panels in Fig. 3 shows the relative fraction of eachoptical AGN type (LINER, Seyfert, and TO). When consid-ering optical AGN subtypes, we find that the prevalence ofLINER follows the general trend of passive galaxies. On theother hand, the fraction of TOs follows the general trend ob-served for star-forming galaxies. Again, there is a break pointat log(M (cid:63) ) (cid:46) . (cid:12) ] where we start to observe significantdi ff erences between the samples. At high stellar masses there isa higher fraction of TOs and a lower fraction of LINERs SIPgalaxies. Given that we have a low number of radio AGN galaxies in eachsample (see last row in Table 1), we compare the activity be-tween systems composed of one galaxy (SIP and SDSS singlegalaxies) and galaxy pairs (SIP and SDSS pairs). The fractionof radio AGN is shown in Fig. 4. The fraction of radio AGNincreases steeply with the stellar mass. The prevalence of radioAGN in pairs is significantly higher than in single galaxies forthe most massive galaxies. e f = (cid:113) f (1 − f ) N T , where f is the relative fraction for a total number of N T galaxies in each sample. Article number, page 5 of 13 & A proofs: manuscript no. AGNpairs_160513_ac
Fig. 3.
Upper panels:
Fraction of SFN (left panel), optical AGN (middle panel), and passive galaxies (right panel) with respect to stellar mass. Thefraction in SIG ( N T = N T = N T = N T = Lower panels:
Fraction of each AGN subtype with respect to stellar mass for SIG ( N T = N T = N T = N T = It would be possible to lower the radio detection limit tostudy the radio nuclear activity for a wider range of stellarmasses (Best et al. 2005a). The downside is that the closer theUniverse less volume and less of each type galaxies, thereforethe number of galaxies would be not statistically significant.
After studying the prevalence of optical nuclear activity as afunction of mass, we investigate its relation with the local andlarge-scale galactic environments. We take into account the ef-fect of the mass dividing the samples into three stellar massbins: low-mass galaxies (10 . ≤ log(M (cid:63) ) < . (cid:12) ]),intermediate-mass galaxies (10 . ≤ log(M (cid:63) ) < . (cid:12) ])and high-mass galaxies (11 . ≤ log(M (cid:63) ) ≤ . (cid:12) ]). As in-troduced in Sect. 2, we use Q pair to quantify the influence of thecompanion, and Q LSS to quantify the e ff ect of the LSS. Giventhe low statistics for radio AGN and AGN subtypes, we performthis part of the study on the fraction of optical AGN, SFN, andpassive galaxies in each sample. According to Argudo-Fernández et al. (2014, 2015b), about 99%of the total tidal strength is due to the e ff ect of the physicallybound companions. Therefore, to investigate the dependence ofnuclear activity with the local environment we restrict our studyto the SIP and the SDSS pairs samples. The fraction of nuclear activity, segregated in stellar massbins, with respect to the Q pair for central galaxy in the SIP andSDSS pairs samples, is shown in the left and right columns inthe Fig. 5, respectively. Higher values of the Q pair correspond toa stronger interaction between the two galaxies in the pair. Ingeneral we do not see any trend in the fraction of SFN, AGN, orpassive galaxies with the local environment. Moreover, there areno significant di ff erences between the SIP and SDSS samples inthe area with common values of Q pair . We further explore thedependence on the local environment and discuss these resultsin Sect. 4. As introduced in Sect. 2, we use the Q LSS to explore the e ff ectof the large-scale environment on the fraction of nuclear activity.We do not find significant di ff erences in the trends for the SIGand SIP samples, as well as for the control samples. Henceforth,for a clearer analysis of the results, we focus on the comparisonbetween SIG and SDSS single galaxies to explore the e ff ect ofthe large-scale environment.Similarly to Fig. 5, Fig. 6 shows the fraction of optical nu-clear activity, segregated in stellar mass bins, with respect to Q LSS , in this case for the SIG and SDSS single galaxies. In gen-eral, there is no clear dependence on the large-scale environmentin the nuclear activity for SDSS single galaxies. However thereis a strong e ff ect in isolated galaxies, with di ff erent trends de- Article number, page 6 of 13. Argudo-Fernández et al.: E ff ect of local and large-scale environments on nuclear activity and star formation Fig. 4.
Fraction of radio AGN (LERG) with respect to stellar mass.Red circles represent the join sample of SIG and SDSS single galaxies( N T = N T = pending on stellar mass. We discuss these trends in more detailsin Sect. 4.
4. Discussion
We study the optical and radio AGN prevalence in each sampleto explore the di ff erent mechanisms triggering nuclear activity.In particular, we will relate the obtained results to the di ff erentsources of gas that can fuel AGN. Depending on the nature (envi-ronment) of the studied samples, these mechanisms could be: 1)the internal and slow mass loss caused by the central BH (inter-nal secular evolution), 2) the external and slow prolonged gas in-fall or the galaxy harassment (environmental secular evolution),or 3) the external and fast ram-pressure striping or galaxy merg-ers. The general trend of optical AGN and passive galaxies is to in-crease with stellar mass, while the fraction of star forming galax-ies decreases from low-mass to massive galaxies, as it is shownin the upper panels in Fig. 3. These monotonic trends with stellarmass are expected because of the ’downsizing’ e ff ect.We find that, in general, there is no di ff erence in the frac-tion of optical AGN for the four samples. This result pointsout that optical AGN is independent of the local environment,and is therefore in agreement with the general consensus thatsecular evolution is a main mechanism triggering optical AGNin the Local Universe (Coziol et al. 2011; Sabater et al. 2012,2013; Hernández-Ibarra et al. 2013; Sabater et al. 2015; Pulatovaet al. 2015; Hernández-Ibarra et al. 2016). As a consequence, the prevalence of optical nuclear activity is independent to the addi-tion of one single companion: the central galaxy in an isolatedpair has no di ff erence with an isolated galaxy. We discuss thediscrepancies with other studies when considering the e ff ects ofthe galaxy pair in Sect. 4.2.1.However, we observe significant di ff erences for higher stel-lar mass bins ( M (cid:63) > . M (cid:12) ). This e ff ect is in agreementwith Melnyk et al. (2015), which claim that the environmentalinfluence is notable in the highest mass galaxies. For massivegalaxies, the fraction of optical AGN in the SIG and SIP samplesis slightly higher than in control samples. In particular, the frac-tion of optical AGN for massive isolated galaxies is higher thanin any other sample, in discrepancy with Melnyk et al. (2015),where the fraction of AGN in pairs is only a little higher than inisolated galaxies.According to the redshift distribution of the samples (lowerpanel of Fig. 1), we consider that the trends observed in the frac-tion of AGN subtypes (see lower panels in Fig. 3) are real. Oth-erwise the trends would be roughly constant in case that a highfraction of low luminosity AGN at low redshift were misclas-sified as LINERs. These observed trends suggest that low-massAGNs are dominated by TOs, while high-mass AGNs are domi-nated by LINERs.SBA13 interpreted the time sequence from interaction induc-ing star formation to passive galaxies (Li et al. 2008; Wild et al.2010) in order from TO, then Seyfert, and then LINER. Even ifour statistic for Seyfert-like galaxies is small, we can interpretour result in light of this sequence. According to our results, thetransition from TO to LINER is slower in isolated pairs. Unfor-tunately we do not have a statistically meaningful number of SIPpairs ( N T = , ,
152 LINERs, Seyferts, and TOs, respec-tively) to explore this result as a function of the Q pair .These results suggest that the black holes of massive( M (cid:63) > . M (cid:12) ) isolated galaxies are still growing while sim-ilar mass isolated pairs, SDSS pairs, and SDSS single galaxieshave quenched their activity. This value is similar to the transi-tion mass between hot and cold modes of gas accretion in sim-ulations by Kereš et al. (2009). We can therefore conclude thatcold gas accretion by secular evolution is su ffi cient to explainthe optical nuclear activity for more massive galaxies. As sug-gested by SBA13, the decrease of the prevalence of optical AGNand LINER for massive galaxies in denser environments can beexplained by the striping of cold gas and its warming. In the exploration of the possible sources of gas fuelling radioAGN and its connection to the environment, it is crucial to dis-criminate between the di ff erent radio AGN modes (Croton et al.2006; Hardcastle et al. 2007; Tasse et al. 2008). In fact, SBA13found opposite trends in HERG and LERG radio AGNs withrespect to the local density, when LERGs show a clear increasewith density. As it is explained in Sect. 2.3, in this study we focuson LERG radio AGNs. It is important to note that the nature ofradio AGN is also sensitive to redshift evolution (Karouzos et al.2014; Cowley et al. 2016; Magliocchetti et al. 2016). Since thegalaxies with SBA13 classification are restricted to the narrowredshift range 0 . ≤ z ≤ .
08 and the samples follow a similarredshift distribution, we do not expect a bias in our comparisoncaused by the possible redshift evolution.Generally, radio nuclear activity is strongly related to thestellar mass and the density of galaxies (SBA13). Hence, wedo not expect to find a high fraction of radio AGN in low den-sity clusters. This dependence with the galaxy density (large-
Article number, page 7 of 13 & A proofs: manuscript no. AGNpairs_160513_ac scale environment) is even observed in isolated galaxies. Sabateret al. (2008, 2012) do not find any high luminosity radio AGNin isolated galaxies in the AMIGA ( A nalysis of the interstellar M edium of I solated GA laxies, Verdes-Montenegro et al. 2005)sample. In fact, the fraction of radio AGN in isolated galaxies issmaller than expected for galaxies with same stellar mass.A recent study of radio nuclear activity for local galaxies(Kaviraj et al. 2015) discarded AGN to be triggered by internalmass loss (secular processes). In relation to the large-scale en-vironment, they also dismissed that radio AGN are fuelled bycluster-scale cooling flows since their radio detections preferen-tially lie outside clusters. They therefore conclude that the mostimportant trigger for ’cold-mode’ radio AGN is galaxy merg-ing. In this regard, Ramos Almeida et al. (2012) and Chiabergeet al. (2015) found strong evidence that mergers are the trigger-ing mechanism for the radio-loud AGN phenomenon. Note thatthe radio-loud AGNs studied are distinct from the LERGs.The fraction of radio AGN for single galaxies ( N T =
33) andgalaxies in pairs ( N T =
45) in our study is shown in Fig. 4. Thedependence of radio AGN with stellar mass is so strong that,even if the number of radio AGN detections in our samples issmall, we can see significant di ff erences with the addition ofone single companion. These di ff erences reflect that not only theLSS a ff ects the radio AGN activity, we also found evidence thatradio AGN is a ff ected by the local environment. This result isin agreement with Pace & Salim (2014), whose find that radioAGN (LERGs) tend to be located in dense environments. Theyalso claim that they are fuelled by the accretion of small quan-tities of hot halo gas. Therefore, while the e ff ect of interactionsis minimal in triggering optical AGN activity, it seems to havea strong connection with radio AGN activity. Unfortunately, wedo not have a statistically significant sample to further explorethe dependence on the local and large-scale environments as foroptical AGN in this study.Note here that the di ff erences between the two samples couldbe reduced if we consider halo masses instead of stellar masses.In fact, Ellison et al. (2015) found that radio AGN (LERGs) arenot fuelled by mergers, since they do not find an excess on thefraction of radio AGN (LERGs) when matching control samplesin halo mass or D4000. Moreover, the results of Fig. 4 agreewith the results of Ellison et al. (2015), whose find an excess ofLERGs in pairs if only stellar-mass and redshift are matched. There has been some disagreement in the literature regardingthe connection between environment and nuclear activity. Inthis context, we explore the e ff ect of local and large-scale en-vironments on the AGN activity in isolated galaxies and phys-ically bound isolated pairs, selected under a three-dimensionalisolation definition. To investigate this, we computed the tidalstrengths Q pair and Q LSS as explained in Sect. 2. The di ff erentnature of the di ff erent environment definitions for isolated sys-tems and the control samples are shown in Fig. 2. By definition,control samples dominate the range of higher values of the Q LSS ,while the systems of galaxies in pairs lie in the area of higher Q pair . The values of Q pair for SDSS pairs are mainly larger than thevalues for SIP galaxies, as it is shown in Fig. 2. This is expectedsince there are more SDSS pairs at smaller projected distances and higher mass ratio than in isolated pairs. Nevertheless, there isan area with common values of the Q pair between -2.5 to -0.5 (seeFig. 5). We do not observe relevant di ff erences in this commonarea. In general, we do not observe any significant trend as afunction of Q pair (see Fig. 5), neither as a function of ∆ Q , whichis defined as Q pair − Q LSS and quantifies the extent that a givengalaxy is dominated by its closest neighbour or its large-scaleenvironment.On the contrary, Ellison et al. (2011) found an increase inthe AGN fraction in most tight pairs and stated that optical AGNmight be triggered by close interactions. A possible explanationfor the discrepancy is the fairly wide projected sepatarions ofthe SDSS pairs sample ( d (cid:39)
200 kpc on average) in compari-son to the AGN excess seen in Ellison et al. (2011) at distances d (cid:46)
50 kpc. To check this we further explore the fraction ofoptical AGN as a function of the projected separation and themass ratio between the two galaxies in pairs. Even if we do nothave a large statistical sample of isolated pairs with projectedseparation smaller than 50 kpc ( N T = N T = ff erence between the SIP or the SDSS pairssamples with projected separation. This discrepancy might thencome from the di ff erent definitions of the local environment. Forthe purpose of the present study, galaxy pairs are located in lowdensity environments. By definition, the SIP sample is very wellisolated from the large-scale environment, with the first nearestneighbour at projected distances larger than 1 Mpc. In the caseof SDSS pairs, we selected groups in Yang et al. (2007) with twogalaxies within the same dark matter halo to avoid the case wherethe same central galaxy can be included in two or three di ff erentpairs. We therefore confirm our previous result that local envi-ronment has not a principal role in triggering optical AGN.The parameters described in Eqs. 1 and 2 quantify the tidale ff ects onto the central (brightest) galaxy in galaxy pairs. Inthis sense, high values of these parameters, in relation to theAGN fraction, are mainly associated with the gas stripping or theshape distortion of the central galaxy. Nevertheless, AGN prop-erties are more associated with the gas accretion of the centralgalaxy. In this sense, an alternative definition of the Q pair ,i.e. Q pair , B ≡ log (cid:18) M A M B (cid:16) D B d AB (cid:17) (cid:19) , is more closely related to trigger thegas accretion to the central galaxy. To better identify the di ff er-ent contributions it is worth to further explore the tidal strengtha ff ecting the faintest galaxy in the galaxy pairs. However, thisshould be carried out in a separated study since only 45% of theSIP has SBA13 classification for the two members in the massrange considered in the present study (the typical mass ratio is ∼ /
100 Argudo-Fernández et al. 2015b).
In contrast to SDSS single galaxies, we find a strong dependenceof the LSS on the optical nuclear activity and star formation inisolated galaxies. Moreover, the observed trends are di ff erent de-pending on galaxy mass (see Fig. 6).According to Argudo-Fernández et al. (2015b), SIG galaxiesmainly belong to the outer parts of filaments, walls, and clus-ters, and generally di ff er from the void population of galaxies. Infact, only one third of SIG galaxies are located in voids. Usingthe code for data visualisation LSSGALPY (Argudo-Fernándezet al. 2015a), we checked that galaxies with low values of Q LSS https://github.com/margudo/LSSGALPY Article number, page 8 of 13. Argudo-Fernández et al.: E ff ect of local and large-scale environments on nuclear activity and star formation Fig. 5.
Fraction of optical nuclear activity with respect to the Q pair environmental parameter. Low-mass galaxies (10 . ≤ log(M (cid:63) ) < . (cid:12) ])are represented by red circles, intermediate-mass galaxies (10 . ≤ log(M (cid:63) ) < . (cid:12) ]) are represented by orange squares, and high-massgalaxies (11 . ≤ log(M (cid:63) ) ≤ . (cid:12) ]) are represented by green triangles. The fraction of SFN, optical AGN, and passive SIP galaxies ( N T = N T = Q pair bin is shown in tables for each sample at the bottom of each panel. The dashed area inthe figures corresponds to the range with common values of Q pair between the two samples, from -2.5 to -0.5. Error bars are given consideringbinomial distribution. are mainly located in void regions, and galaxies with higher Q LSS are more related with denser structures, as filaments or walls. SeeAppendix A for more details.This means that the di ff erences that we found between SIGgalaxies and SDSS singles, with respect to high values of the Q LSS , are directly related to the location of the galaxies in theoutskirts or inside clusters, respectively. According to this, thegeneral trend for the fraction of passive isolated galaxies is todecrease from voids to denser regions (e. g. clusters, filaments).However, the fraction is smaller than that of similar mass SDSSsingle galaxies located in the same regions.The fraction of optical AGN for high-mass SIG galaxies in-creases with denser large-scale environment (massive SIG galax-ies have more accretion time and therefore they have more gas).We interpret this result as the central black hole in massive iso-lated galaxies being fuelled by cold gas from the LSS. We wouldnot see an increment in SDSS single galaxies because of thewarming of the cold gas in denser environments, where the ef-fect of the stripping of di ff use gas or strangulation, ram-pressurestripping, and galaxy harassment is present. (von der Lindenet al. 2010; Peng et al. 2015).On the contrary, the fraction of optical AGN for low-massSIG galaxies decreases from voids to denser regions (e.g. clus-ters, filaments). In comparison to the trend for passive galaxies, this is directly translated into an important increment of star-forming SIG galaxies in the outskirts of clusters. Given that thefraction of star-forming SIG galaxies with low values of Q LSS is smaller, our results contrast to those of Liu et al. (2015).They found that the fraction of star-forming galaxies in voidsis significantly higher than that in walls. Note that only 14% ofvoid galaxies in Pan et al. (2012) are found in the SIG sample(Argudo-Fernández et al. 2015b), therefore void galaxies spana large range of local environments. In the future, we will ex-plore these di ff erences further by considering the possible e ff ectof morphology or stellar populations.We previously discussed the fact that the fraction of opti-cal AGN in the SIG sample continues increasing, while peaksat M (cid:63) (cid:39) . M (cid:12) for the other samples (see Fig. 3). FromFig. 6, we see that this continue increase mainly occurs at high Q LSS . It is interesting that, on the contrary, the AGN fraction ofthe low-mass SIG galaxies even decreases. Such a di ff erent massdependent worth a further detailed study.The di ff erent mass- and environment-dependence behavioursbetween isolated and control samples suggest that a halo / massor a simple Q LSS parameter is not enough to characterise theenvironmental e ff ects of galaxies in more complicated (small-scale) environments. Article number, page 9 of 13 & A proofs: manuscript no. AGNpairs_160513_ac
Fig. 6.
Fraction of optical nuclear activity with respect to the Q LSS environmental parameter. Low-mass galaxies (10 . ≤ log(M (cid:63) ) < . (cid:12) ])are represented by red circles, intermediate-mass galaxies (10 . ≤ log(M (cid:63) ) < . (cid:12) ]) are represented by orange squares, and high-massgalaxies (11 . ≤ log(M (cid:63) ) ≤ . (cid:12) ]) are represented by green triangles. The fraction of SFN, optical AGN, and passive SIG galaxies ( N T = N T = Q LSS bin is shown in tables for each sample at the bottom of each panel. Thedashed area in the figures corresponds to the range with common values of Q LSS between the two samples, from -5.5 to -3.5. Error bars are givenconsidering binomial distribution.
5. Summary and conclusions
In this work we study the e ff ect of the environment on the frac-tion of optical and radio nuclear activity. In particular, we in-vestigate the e ff ect of both, local and large-scale environmentson nuclear activity and star formation, for the first time usingthree-dimensional isolated galaxies and physically bound iso-lated pairs (SIG and SIP galaxies, respectively). Besides, usingthe tidal strength parameters Q pair and Q LSS we are able to quan-tify separately the e ff ect of one-on-one interactions from the ef-fect of the large-scale environment where the galaxy resides.Control samples of single galaxies and pairs selected from theSDSS were used for comparison.Our main conclusions are the following:1. The prevalence of optical AGN is found to be independent tothe addition of one single companion (see Fig. 3).2. For massive galaxies, the fraction of optical AGN in iso-lated galaxies and isolated pairs is slightly higher than inthe control samples (see the central upper panel in Fig. 3).Moreover, the fraction of massive passive isolated galaxiesis smaller than in any other sample (see the right upper panelin Fig. 3). These results suggest that the black holes of mas-sive (log(M (cid:63) ) > . (cid:12) ]) isolated galaxies are still grow- ing while similar mass isolated pairs, SDSS pairs, and SDSSsingle galaxies have already quenched their activity.3. Local environment has not a principal role in triggering opti-cal AGN (see Fig. 5). Cold gas accretion by secular evolutionis su ffi cient to explain the optical nuclear activity for moremassive galaxies (see the middle left panel in Fig. 6).4. In contrast to local environment, we find a strong dependenceof the optical nuclear activity and star formation on the LSSfor isolated systems (see Fig. 6). In particular, the fractionof AGN for high-mass SIG galaxies increases with denserlarge-scale environment. We interpret this result as the cen-tral black hole in massive isolated galaxies being fuelled bycold gas from the LSS.5. Regarding to radio nuclear activity (LERG), we find that notonly the galaxy mass and large-scale environment a ff ects it,but that radio AGN are also strongly a ff ected by the localenvironment (see Fig. 4).Optical AGN is related to cold gas accretion, while radioAGN is related to hot gas accretion. In this context, there is morecold gas, fuelling the central optical AGN, in isolated systems.Overall, our results are in agreement with a scenario where coldgas accretion by secular evolution is the main driver for opticalAGN, while hot gas accretion and one-on-one interactions arethe main drivers of radio AGN activity. Article number, page 10 of 13. Argudo-Fernández et al.: E ff ect of local and large-scale environments on nuclear activity and star formation Acknowledgments
The authors acknowledge the anonymous referee for his / her verydetailed and useful report, which helped to clarify and improvethe quality of this work.MAF is grateful for financial support from PIFI (fundedby Chinese Academy of Sciences President’s International Fel-lowship Initiative) Grant No. 2015PM056 and from CONICYTFONDECYT project No. 3160304. This work was partly sup-ported by the Strategic Priority Research Program ”The Emer-gence of Cosmological Structures” of the Chinese Academyof Sciences (CAS; grant XDB09030200), the National NaturalScience Foundation of China (NSFC) with the Project Num-ber of 11573050 and 11433003, and the ”973 Program” 2014CB845705. This work was partially supported by MInisteriode Economia y Competividad and by FEDER (Fondo Europeode Desarrollo Regional) via grants AYA2011-24728, AYA2013-47742-C04-01, AYA2014-53506-P, and from the ”Junta de An-dalucía” local government through the FQM-108 project.This research made use of astropy , a community-developedcore python ( ) package for Astron-omy (Astropy Collaboration et al. 2013); ipython (Pérez &Granger 2007); matplotlib (Hunter 2007); numpy (Walt et al.2011); scipy (Jones et al. 2001); and topcat (Taylor 2005).Funding for SDSS-III has been provided by the Alfred P.Sloan Foundation, the Participating Institutions, the NationalScience Foundation, and the U.S. Department of Energy O ffi ceof Science. The SDSS-III web site is http: // / . References
Abazajian, K. N., Adelman-McCarthy, J. K., Agüeros, M. A., et al. 2009, ApJS,182, 543Abell, G. O. 1958, ApJS, 3, 211Ahn, C. P., Alexandro ff , R., Allende Prieto, C., et al. 2014, ApJS, 211, 17Alam, S., Albareti, F. D., Allende Prieto, C., et al. 2015, ApJS, 219, 12Alexander, D. M. & Hickox, R. C. 2012, New A Rev., 56, 93Argudo-Fernández, M., Duarte Puertas, S., Verley, S., Sabater, J., & Ruiz,J. E. 2015a, LSSGALPY: Visualization of the large-scale environmentaround galaxies on the 3D space, Astrophysics Source Code Library, recordascl:1505.012Argudo-Fernández, M., Verley, S., Bergond, G., et al. 2015b, A&A, 578, A110Argudo-Fernández, M., Verley, S., Bergond, G., et al. 2014, A&A, 564, A94Argudo-Fernández, M., Verley, S., Bergond, G., et al. 2013, A&A, 560, A9Astropy Collaboration, Robitaille, T. P., Tollerud, E. J., et al. 2013, A&A, 558,A33Baldry, I. K., Glazebrook, K., Brinkmann, J., et al. 2004, ApJ, 600, 681Baldwin, J. A., Phillips, M. M., & Terlevich, R. 1981, PASP, 93, 5Becker, R. H., White, R. L., & Helfand, D. J. 1995, ApJ, 450, 559Bell, E. F., McIntosh, D. H., Katz, N., & Weinberg, M. D. 2003, ApJS, 149, 289Best, P. N. & Heckman, T. M. 2012, MNRAS, 421, 1569Best, P. N., Kau ff mann, G., Heckman, T. M., et al. 2005a, MNRAS, 362, 25Best, P. N., Kau ff mann, G., Heckman, T. M., & Ivezi´c, Ž. 2005b, MNRAS, 362,9Blanton, M. R. & Roweis, S. 2007, AJ, 133, 734Blanton, M. R., Schlegel, D. J., Strauss, M. A., et al. 2005, AJ, 129, 2562Boselli, A. & Gavazzi, G. 2014, A&A Rev., 22, 74Brinchmann, J., Charlot, S., White, S. D. M., et al. 2004, MNRAS, 351, 1151Calvi, R., Poggianti, B. M., Fasano, G., & Vulcani, B. 2012, MNRAS, 419, L14Casado, J., Ascasibar, Y., Gavilán, M., et al. 2015, MNRAS, 451, 888Chiaberge, M., Gilli, R., Lotz, J. M., & Norman, C. 2015, ApJ, 806, 147Choi, Y.-Y., Woo, J.-H., & Park, C. 2009, ApJ, 699, 1679Condon, J. J., Cotton, W. D., Greisen, E. W., et al. 1998, AJ, 115, 1693Cowley, M. J., Spitler, L. R., Tran, K.-V. H., et al. 2016, MNRAS, 457, 629Coziol, R., Torres-Papaqui, J. P., Plauchu-Frayn, I., et al. 2011, Rev. MexicanaAstron. Astrofis., 47, 361Croton, D. J., Springel, V., White, S. D. M., et al. 2006, MNRAS, 365, 11Dressler, A. 1980, ApJ, 236, 351Ellison, S. L., Patton, D. R., & Hickox, R. C. 2015, MNRAS, 451, L35Ellison, S. L., Patton, D. R., Mendel, J. T., & Scudder, J. M. 2011, MNRAS, 418,2043Fabian, A. C. 1999, MNRAS, 308, L39Grützbauch, R., Conselice, C. J., Varela, J., et al. 2011, MNRAS, 411, 929 Hardcastle, M. J., Evans, D. A., & Croston, J. H. 2007, MNRAS, 376, 1849Heckman, T. M. 1980, A&A, 87, 152Hernández-Ibarra, F. J., Dultzin, D., Krongold, Y., et al. 2013, MNRAS, 434,336Hernández-Ibarra, F. J., Krongold, Y., Dultzin, D., et al. 2016, MN-RAS[ arXiv:1509.02186 ]Hine, R. G. & Longair, M. S. 1979, MNRAS, 188, 111Hong, J., Im, M., Kim, M., & Ho, L. C. 2015, ApJ, 804, 34Hunter, J. D. 2007, Computing In Science & Engineering, 9, 90Jones, E., Oliphant, T., Peterson, P., et al. 2001, SciPy: Open source scientifictools for Python, [Online; accessed 2016-01-15]Karouzos, M., Im, M., Kim, J.-W., et al. 2014, ApJ, 797, 26Kau ff mann, G., Heckman, T. M., Tremonti, C., et al. 2003a, MNRAS, 346, 1055Kau ff mann, G., Heckman, T. M., White, S. D. M., et al. 2003b, MNRAS, 341,33Kau ff mann, G., White, S. D. M., Heckman, T. M., et al. 2004, MNRAS, 353,713Kaviraj, S., Shabala, S. S., Deller, A. T., & Middelberg, E. 2015, MNRAS, 452,774Kereš, D., Katz, N., Fardal, M., Davé, R., & Weinberg, D. H. 2009, MNRAS,395, 160Li, C., Kau ff mann, G., Heckman, T. M., White, S. D. M., & Jing, Y. P. 2008,MNRAS, 385, 1915Liu, C.-X., Pan, D. C., Hao, L., et al. 2015, ApJ, 810, 165Lynden-Bell, D. 1969, Nature, 223, 690Magliocchetti, M., Lutz, D., Santini, P., et al. 2016, MNRAS, 456, 431Manzer, L. H. & De Robertis, M. M. 2014, ApJ, 788, 140McAlpine, K., Prandoni, I., Jarvis, M., et al. 2015, Advancing Astrophysics withthe Square Kilometre Array (AASKA14), 83Melnyk, O., Karachentseva, V., & Karachentsev, I. 2015, MNRAS, 451, 1482Pace, C. & Salim, S. 2014, ApJ, 785, 66Pan, D. C., Vogeley, M. S., Hoyle, F., Choi, Y.-Y., & Park, C. 2012, MNRAS,421, 926Peng, Y., Maiolino, R., & Cochrane, R. 2015, Nature, 521, 192Peng, Y.-j., Lilly, S. J., Kovaˇc, K., et al. 2010, ApJ, 721, 193Peng, Y.-j., Lilly, S. J., Renzini, A., & Carollo, M. 2012, ApJ, 757, 4Pérez, F. & Granger, B. E. 2007, Computing in Science and Engineering, 9, 21Pulatova, N. G., Vavilova, I. B., Sawangwit, U., Babyk, I., & Klimanov, S. 2015,MNRAS, 447, 2209Ramos Almeida, C., Bessiere, P. S., Tadhunter, C. N., et al. 2012, MNRAS, 419,687Rosario, D. J., Mendel, J. T., Ellison, S. L., Lutz, D., & Trump, J. R. 2016,MNRAS, 457, 2703Sabater, J., Best, P. N., & Argudo-Fernández, M. 2013, MNRAS, 430, 638Sabater, J., Best, P. N., & Heckman, T. M. 2015, MNRAS, 447, 110Sabater, J., Leon, S., Verdes-Montenegro, L., et al. 2008, A&A, 486, 73Sabater, J., Verdes-Montenegro, L., Leon, S., Best, P., & Sulentic, J. 2012, A&A,545, A15Salim, S., Rich, R. M., Charlot, S., et al. 2007, ApJS, 173, 267Salpeter, E. E. 1964, ApJ, 140, 796Satyapal, S., Ellison, S. L., McAlpine, W., et al. 2014, MNRAS, 441, 1297Shakura, N. I. & Sunyaev, R. A. 1973, A&A, 24, 337Shen, S.-Y., Argudo-Fernández, M., Chen, L., et al. 2016, Research in Astron-omy and Astrophysics, 16, 007Soltan, A. 1982, MNRAS, 200, 115Strauss, M. A., Weinberg, D. H., Lupton, R. H., et al. 2002, AJ, 124, 1810Tasse, C., Best, P. N., Röttgering, H., & Le Borgne, D. 2008, A&A, 490, 893Taylor, M. B. 2005, in Astronomical Society of the Pacific Conference Se-ries, Vol. 347, Astronomical Data Analysis Software and Systems XIV, ed.P. Shopbell, M. Britton, & R. Ebert, 29Tremonti, C. A., Heckman, T. M., Kau ff mann, G., et al. 2004, ApJ, 613, 898Verdes-Montenegro, L., Sulentic, J., Lisenfeld, U., et al. 2005, A&A, 436, 443Verley, S., Leon, S., Verdes-Montenegro, L., et al. 2007, A&A, 472, 121von der Linden, A., Wild, V., Kau ff mann, G., White, S. D. M., & Weinmann, S.2010, MNRAS, 404, 1231Walt, S. v. d., Colbert, S. C., & Varoquaux, G. 2011, Computing in Science &Engineering, 13, 22Wild, V., Heckman, T., & Charlot, S. 2010, MNRAS, 405, 933Yang, X., Mo, H. J., van den Bosch, F. C., et al. 2007, ApJ, 671, 153York, D. G., Adelman, J., Anderson, J. J. E., et al. 2000, AJ, 120, 1579Zhao, G., Zhao, Y.-H., Chu, Y.-Q., Jing, Y.-P., & Deng, L.-C. 2012, Research inAstronomy and Astrophysics, 12, 723 Article number, page 11 of 13 & A proofs: manuscript no. AGNpairs_160513_ac
Appendix A: LSSGALPY
LSSGALPY (Argudo-Fernández et al. 2015a) is a tool for theinteractive visualization of the large-scale environment aroundgalaxies on the 3D space based on Python language. The toolallows one to easily compare the 3D positions of a sample (orsamples) with respect to the locations of the LSS galaxies in theirlocal and / or large scale environments. For the purpose of thisstudy, we compared the position of galaxies in the SIG sampleaccording to three di ff erent ranges of values of their Q LSS (see asnapshot of the tool in Fig. A.1).We observe that most of the SIG galaxies with Q LSS ≤ − . − . < Q LSS ≤ − .
5, and specially with Q LSS > − .
5, are distributed along the LSS. This means thatthese SIG galaxies mainly belong to the outer parts of filaments,walls, and clusters, and generally di ff er from the void populationof galaxies. Available at https://github.com/margudo/LSSGALPY
Article number, page 12 of 13. Argudo-Fernández et al.: E ff ect of local and large-scale environments on nuclear activity and star formation F i g . A . . I n t e r ac ti v e D v i s u a li s a ti on s o f t w a r e : M o ll w e i d e p r o j ec ti on . M o ll w e i d e p r o j ec ti ono f t h e L SS f o r g a l a x i e s ( b l ac kpo i n t s ) i n t h e r e d s h i f t r a ng e . < z < . a ss ho w n i n t h e b l u e b a r s i n t h e l o w e r p a r t o f t h e fi gu r e . R e dd i s k s r e p r e s e n t S I G g a l a x i e s w it h Q L SS ≤− . w it h i n t h e s a m e r e d s h i f t r a ng e . G r ee nd i s k s r e p r e s e n t S I G g a l a x i e s w it h − . < Q L SS ≤− . w it h i n t h e s a m e r e d s h i f t r a ng e . B l u e d i s k s r e p r e s e n t S I G g a l a x i e s w it h Q L SS > − . w it h i n t h e s a m e r e d s h i f t r a ng e . T ogu i d e t h ee y e , a w e dg e d i a g r a m , f o r L SS g a l a x i e s w it h i n - a nd2d e g r ee s i nd ec li n a ti on , i ss ho w n i n t h e r i gh tl o w e r p a r t o f t h e fi gu r e . C o l ou r c od eacc o r d i ng t o t h e r e d s h i f t fr o m z = ( b l u e ) t o z = . (r e d ) . T h e r e d r i ng i n t h e po l a rr e p r e s e n t a ti on c o rr e s pond s t o t h e s e l ec t e d r e d s h i f t r a ng e i n t h ece n t r a l M o ll w e i d e p r o j ec ti on ..