1: %% Modified 25 July 2007
2: %%
3: \documentclass{emulateapj}
4: %\documentclass[12pt,preprint]{aastex}
5:
6: %% manuscript produces a one-column, double-spaced document:
7:
8: %% \documentclass[manuscript]{aastex}
9:
10: %% preprint2 produces a double-column, single-spaced document:
11:
12: %% \documentclass[preprint2]{aastex}
13:
14: %% Sometimes a paper's abstract is too long to fit on the
15: %% title page in preprint2 mode. When that is the case,
16: %% use the longabstract style option.
17:
18: %% \documentclass[preprint2,longabstract]{aastex}
19:
20: %% If you want to create your own macros, you can do so
21: %% using \newcommand. Your macros should appear before
22: %% the \begin{document} command.
23: %%
24: %% If you are submitting to a journal that translates manuscripts
25: %% into SGML, you need to follow certain guidelines when preparing
26: %% your macros. See the AASTeX v5.x Author Guide
27: %% for information.
28:
29: \newcommand{\vdag}{(v)^\dagger}
30: \newcommand{\myemail}{jsilverman@mpe.mpg.de}
31:
32: %% You can insert a short comment on the title page using the command below.
33:
34: \slugcomment{Published in ApJ, 2008, 675, 1025}
35:
36: %% If you wish, you may supply running head information, although
37: %% this information may be modified by the editorial offices.
38: %% The left head contains a list of authors,
39: %% usually a maximum of three (otherwise use et al.). The right
40: %% head is a modified title of up to roughly 44 characters.
41: %% Running heads will not print in the manuscript style.
42:
43: \shorttitle{Evolution of AGN host galaxies}
44: \shortauthors{Silverman et al.}
45:
46: %% This is the end of the preamble. Indicate the beginning of the
47: %% paper itself with \begin{document}.
48:
49: \begin{document}
50:
51: %% LaTeX will automatically break titles if they run longer than
52: %% one line. However, you may use \\ to force a line break if
53: %% you desire.
54:
55:
56: \title{The Evolution of AGN Host Galaxies: From Blue to Red and
57: the Influence of Large-Scale Structures}
58:
59: %% Use \author, \affil, and the \and command to format
60: %% author and affiliation information.
61: %% Note that \email has replaced the old \authoremail command
62: %% from AASTeX v4.0. You can use \email to mark an email address
63: %% anywhere in the paper, not just in the front matter.
64: %% As in the title, use \\ to force line breaks.
65:
66: \author{
67: J.~D.~Silverman,\altaffilmark{1}
68: V.~Mainieri,\altaffilmark{1,2}
69: B.~D.~Lehmer,\altaffilmark{3}
70: D.~M.~Alexander,\altaffilmark{4}
71: F.~E.~Bauer,\altaffilmark{5}
72: J.~Bergeron,\altaffilmark{6,12}
73: W.~N.~Brandt,\altaffilmark{3}
74: R.~Gilli,\altaffilmark{7}
75: G.~Hasinger,\altaffilmark{1}
76: D.~P.~Schneider,\altaffilmark{3}
77: P.~Tozzi,\altaffilmark{8}
78: C.~Vignali,\altaffilmark{9}
79: A.~M.~Koekemoer,\altaffilmark{10}
80: T.~Miyaji,\altaffilmark{11}
81: P.~Popesso,\altaffilmark{2,12}
82: P.~Rosati,\altaffilmark{2,12}
83: \& G.~Szokoly\altaffilmark{1}
84: }
85:
86: %% Notice that each of these authors has alternate affiliations, which
87: %% are identified by the \altaffilmark after each name. Specify alternate
88: %% affiliation information with \altaffiltext, with one command per each
89: %% affiliation.
90:
91: \altaffiltext{1}{Max-Planck-Institut f\"ur extraterrestrische Physik,
92: D-84571 Garching, Germany} \altaffiltext{2}{European Southern
93: Observatory, Karl-Schwarzschild-Strasse 2, Garching, D-85748, Germany}
94: \altaffiltext{3}{Department of Astronomy \& Astrophysics, 525 Davey
95: Lab, The Pennsylvania State University, University Park, PA 16802,
96: USA} \altaffiltext{4}{Department of Physics, University of Durham,
97: South Road, Durham, DH1 3LE, UK} \altaffiltext{5}{Columbia
98: Astrophysics Laboratory, Columbia University, Pupin Labortories, 550
99: W. 120th St., Rm 1418, New York, NY 10027, USA}
100: \altaffiltext{6}{Institut d'Astrophysique de Paris, 98bis Boulevard,
101: F-75014 Paris, France} \altaffiltext{7}{Istituto Nazionale di
102: Astrofisica (INAF) - Osservatorio Astronomico di Bologna, Via Ranzani
103: 1, 40127 Bologna, Italy} \altaffiltext{8}{INAF - Osservatorio
104: Astronomico di Trieste, via G. B. Tiepolo 11, 34131 Trieste, Italy}
105: \altaffiltext{9}{Dipartimento di Astronomia, Universit\'a degli Studi
106: di Bologna, Via Ranzani 1, 40127 Bologna, Italy}
107: \altaffiltext{10}{Space Telescope Science Institute, 3700 San Martin
108: Drive, Baltimore, MD 21218, USA} \altaffiltext{11}{Department of
109: Physics, Carnegie Mellon University, Pittsburgh, PA 15213, USA}
110: \altaffiltext{12}{Based on observations made at the European Southern
111: Observatory, Paranal, Chile (ESO programs 170.A-0788, 171.A-3045,
112: 072.A-0139)}
113:
114: %% Mark off your abstract in the ``abstract'' environment. In the manuscript
115: %% style, abstract will output a Received/Accepted line after the
116: %% title and affiliation information. No date will appear since the author
117: %% does not have this information. The dates will be filled in by the
118: %% editorial office after submission.
119:
120: \begin{abstract}
121:
122: We present an analysis of 109 moderate-luminosity ($41.9\leq
123: \log~L_{0.5-8.0~{\rm keV}}\leq 43.7$) AGN in the Extended $Chandra$
124: Deep \hbox{Field-South} survey, which is drawn from 5,549 galaxies
125: from the \hbox{COMBO-17} and GEMS surveys having $0.4 \leq z \leq
126: 1.1$. These obscured or optically-weak AGN facilitate the study of
127: their host galaxies since the AGN provide an insubstantial amount of
128: contamination to the galaxy light. We find that the color
129: distribution of AGN host galaxies is highly dependent upon (1) the
130: strong color-evolution of luminous ($M_V<-20.7$) galaxies, and (2) the
131: influence of $\sim10~ {\rm Mpc}$ scale structures. When excluding
132: galaxies within the redshift range $0.63 \leq z \leq 0.76$, a regime
133: dominated by sources in large-scale structures at $z=0.67$ and
134: $z=0.73$, we observe a bimodality in the host galaxy colors. Galaxies
135: hosting AGN at $z\ga0.8$ preferentially have bluer (rest-frame
136: $U-V<0.7$) colors than their $z\lesssim0.6$ counterparts (many of
137: which fall along the red sequence). The fraction of galaxies hosting
138: AGN peaks in the ``green valley'' ($0.5<U-V<1.0$); this is primarily
139: due to enhanced AGN activity in the redshift interval $0.63\leq z \leq
140: 0.76$. The AGN fraction in this redshift and color interval is 12.8\%
141: (compared to its `field' value of $7.8\%$) and reaches a maximum of
142: $14.8\%$ at $U-V\sim0.8$. We further find that blue, bulge-dominated
143: (S\'{e}rsic index $n>2.5$) galaxies have the highest fraction of AGN
144: ($21\%$) in our sample. We explore the scenario that the
145: evolution of AGN hosts is driven by galaxy mergers and illustrate that
146: an accurate assessment requires a larger area survey since only three
147: hosts may be undergoing a merger with timescales $\lesssim 1$ Gyr
148: following a starburst phase.
149:
150: \end{abstract}
151:
152: %% Keywords should appear after the \end{abstract} command. The uncommented
153: %% example has been keyed in ApJ style. See the instructions to authors
154: %% for the journal to which you are submitting your paper to determine
155: %% what keyword punctuation is appropriate.
156:
157: %% Authors who wish to have the most important objects in their paper
158: %% linked in the electronic edition to a data center may do so in the
159: %% subject header. Objects should be in the appropriate "individual"
160: %% headers (e.g., quasars: individual, stars: individual, etc.) with the
161: %% additional provision that the total number of headers, including each
162: %% individual object, not exceed six. The \objectname{} macro, and its
163: %% alias \object{}, is used to mark each object. The macro takes the object
164: %% name as its primary argument. This name will appear in the paper
165: %% and serve as the link's anchor in the electronic edition if the name
166: %% is recognized by the data centers. The macro also takes an optional
167: %% argument in parentheses in cases where the data center identification
168: %% differs from what is to be printed in the paper.
169:
170: \keywords{quasars: general, galaxies: active, galaxies: evolution,
171: X-rays: galaxies, large-scale structure of universe}
172:
173: %% From the front matter, we move on to the body of the paper.
174: %% In the first two sections, notice the use of the natbib \citep
175: %% and \citet commands to identify citations. The citations are
176: %% tied to the reference list via symbolic KEYs. The KEY corresponds
177: %% to the KEY in the \bibitem in the reference list below. We have
178: %% chosen the first three characters of the first author's name plus
179: %% the last two numeral of the year of publication as our KEY for
180: %% each reference.
181:
182: \section{Introduction}
183:
184: There has been remarkable evidence found in the last few years that
185: the evolution of supermassive black holes (SMBHs) and galaxies are
186: inextricably linked. For instance, ultraluminous infrared and
187: submillimeter galaxies \citep{al05}, with prodigious rates of star
188: formation, show a high fraction of AGN activity. Current models of
189: galaxy evolution \citep[e.g.,][]{di05,hop05,sp05b} demonstrate that
190: merger-driven feedback from an AGN may quench star formation,
191: contributing to the order-of-magnitude decline in the cosmic SFR and
192: QSO emissivity \citep{bo98,fr99,me04} observed between $z\sim1$ and the
193: present day. This self-regulated growth may then establish the
194: correlation between the mass of the central black hole and the stellar
195: velocity dispersion of the host bulge \citep[][]{fe00,ge00,tr02}.
196:
197: A clear imprint of this scenario may be evident in the properties
198: (e.g., color and morphology) of the host galaxies of AGN caught at
199: early epochs in their evolution. Much effort has been undertaken to
200: detect the hosts of nearby bright QSOs
201: \citep[e.g.,][]{ba97,ja04a,mc04} though this endeavor remains
202: challenging. Surveys of obscured AGN, selected through various means
203: (e.g., optical emission lines, X-rays, or reradiated infrared
204: emission), offer the potential to measure cleanly the ages, stellar
205: populations, and star-formation rates of these host galaxies
206: \citep[e.g.,][]{st06}. \citet{ka03} clearly show that the hosts of
207: narrow emission line AGN ($z<0.3$) in the Sloan Digital Sky Survey
208: (SDSS) are predominately massive, early-type galaxies and that those
209: with the most-luminous AGN have significant star formation with
210: younger mean stellar ages. Deep X-ray surveys \citep[see][for a
211: review]{br05} are generating significant samples of obscured AGN
212: closer to the peak ($z\sim1$) of the global star-formation and merger
213: rate of galaxies \citep{hop06b,kart07}. Recently, \citet{na07} have
214: shown that the host galaxies of moderate-luminosity AGN with
215: $0.6<z<1.4$, found in the Extended Groth Strip, have a broad range of
216: optical colors that span the same region of the color-magnitude
217: relation as luminous ($M_B<-20.5$) galaxies. The prevalence of AGN
218: host galaxies within the region separating the blue and red galaxy
219: populations may lend support for the importance of AGN feedback since
220: this location is thought to represent a transitional phase in the
221: evolution of galaxies.
222:
223: It is of much interest to constrain observationally whether
224: environment plays a significant role in the growth of SMBHs as
225: expected by merger-driven accretion models. Observational support for
226: the importance of mergers is mainly based on circumstantial evidence:
227: (1) AGN predominately reside in massive, early-type galaxies
228: \citep[e.g., ][]{ka03}, of which many are massive ellipticals, well
229: thought to be the end-product of a major-merger between gas-rich disk
230: galaxies, (2) ultraluminous infrared galaxies, that are
231: morphologically disturbed in almost all cases, have a high fraction
232: ($\sim50\%$) of AGN activity and large molecular gas concentrations in
233: their nuclei \citep[see ][]{sa96}. Studies of low-redshift ($z<0.3$)
234: AGN from the SDSS indicate that the fraction of galaxies harboring AGN
235: is independent of environment \citep[e.g.,][]{mi03} even in the cores
236: of massive clusters. For the most-luminous narrow-line AGN in the
237: SDSS \citep{ka04}, an AGN-fraction--density relation has been shown to
238: exist analogous to the SFR-density relation with a higher AGN fraction
239: in low-density (i.e., `field') environments. Both studies provide
240: evidence counter to the expectation that SMBH accretion is induced by
241: galaxy mergers. Hard X-ray selected surveys, most effective at
242: identifying obscured accretion at higher redshifts, also present
243: disparate views on the relationship between AGN activity and their
244: environments. \citet{ge07} present preliminary results that
245: demonstrate that X-ray selected AGN at $z\sim1$ `avoid' underdense
246: regions and those with blue ($U-V\lesssim1$) hosts are found in denser
247: environments, hinting at a connection to star formation. On the
248: contrary, \cite{gr05} find that AGN in the $Chandra$ Deep Fields show
249: no evidence for an environmental dependency based on similar AGN host
250: morphologies and near-neighbor counts to the non-active galaxy
251: population. The AGN fraction in clusters \citep{ma06,ma07}, though
252: higher than previously determined \citep{dr83}, is not significantly
253: different than that in the `field'; a different picture may emerge at
254: higher redshifts since \citet{east07} find significant evolution of
255: the AGN fraction in clusters at $z\sim0.6$. There is also evidence
256: that larger scale ($\gtrsim$0.1~Mpc) structures may play an important
257: role in AGN fueling. \citet{li06} have measured the cross-correlation
258: function of narrow-line AGN in the SDSS with a well-controlled sample
259: of non-active galaxies, and find that AGN are primarily not associated
260: with major mergers of galaxies, and preferentially lie within the
261: densest peaks of the underlying dark-matter distribution.
262: \citet{gi03} find that 30\% of the X-ray selected AGN in the 1 Ms
263: CDF-S are located within narrow redshift slices at $z=0.67$ and
264: $z=0.73$ and spread across the full $Chandra$ field-of-view
265: ($17\arcmin\times17\arcmin$), which corresponds to a physical scale of
266: 7.3 Mpc at $z=0.7$. Similar structures are evident in the CDF-N
267: \citep{ba03} that may indicate the importance of large-scale ($\sim10$
268: Mpc) structures to trigger mass accretion onto SMBHs.
269:
270: The Extended $Chandra$ Deep Field-South (\hbox{E-CDF-S}) is an ideal
271: survey field to use for investigating the properties of galaxies
272: harboring AGN and the role of environment due to its remarkable
273: multi-wavelength coverage. We have completed a 1 Ms $Chandra$ Legacy
274: program \citep[P.I.: W. N. Brandt;][]{le05} that covers a wide area
275: (0.33 deg$^2$; three times area of the CDF-S) at the depths required
276: to detect moderate-luminosity ($L_{\rm X}\sim10^{43}$ erg s$^{-1}$)
277: AGN, including those with significant obscuration, out to the quasar
278: epoch ($z\sim2.5$). The optical coverage of the \hbox{E-CDF-S} is
279: extensive, providing large galaxy samples with a wealth of
280: multiwavelength data. For example, the COMBO-17 survey \citep{wo04}
281: has imaged the field with 12 narrow-band and 5 broad-band optical
282: filters thus providing magnitudes, accurate photometric redshifts
283: ($\sigma_z \approx$ 0.03) and galaxy SED types, that further aid in
284: the identification of the X-ray source counterparts. Optical
285: morphological information is available from the {\it HST} Advanced
286: Camera for Surveys (ACS) observations via the GEMS \citep{ri04,ha07},
287: GOODS \citep{gi04} and the {\it HST} Ultra Deep field
288: \citep[UDF;][]{beck06} projects. Over 1000 spectroscopic redshifts
289: are available via the CDF-S \citep{sz04}, VVDS \citep{lef04}, K20
290: \citep{mi05}, and GOODS \citep{va05,va06} surveys.
291:
292: In this paper, we investigate the location of moderate-luminosity AGN,
293: in the \hbox{E-CDF-S}, on the color-magnitude diagram and their
294: relation to the underlying galaxy population using our current catalog
295: of X-ray selected AGN with either spectroscopic or photometric
296: redshifts. The \hbox{E-CDF-S} contains two prominent redshift spikes
297: \citep{gi03} enabling us to determine the influence of $\sim$~10~Mpc
298: structures on the overall color-magnitude distribution. We discuss
299: how our results fit in with the morphological properties of the sample
300: and the impact on galaxy evolution models that incorporate AGN
301: feedback to quench star formation effectively. Throughout this work,
302: we assume $H_0=70$ km s$^{-1}$ Mpc$^{-1}$, $\Omega_{\Lambda}=0.7$, and
303: $\Omega_{\rm{M}}=0.3$.
304:
305: \section{Data}
306:
307: We select a parent sample of galaxies in the \hbox{E-CDF-S} using
308: published catalogs from the COMBO-17 and GEMS surveys. The $Chandra$
309: observations, covering the equivalent sky area, enable identification
310: of those galaxies that harbor moderate-luminosity AGN in a manner that
311: is least biased against obscuration. Follow-up optical spectroscopy
312: of these X-ray sources with the VLT facilitates the identification of
313: AGN. As detailed below, our selection is tuned to generate a sample
314: of galaxies hosting AGN for which the optical emission is dominated by
315: the host galaxy; thus no further removal of AGN light (i.e., cleaning)
316: is required for this study. By restricting ourselves to an initial
317: optically-selected sample of galaxies, we aim to measure the fraction
318: of galaxies harboring AGN as a function of their intrinsic properties.
319:
320: \subsection{Parent galaxy population}
321:
322: COMBO-17 provides a highly complete sample of galaxies over the full
323: \hbox{E-CDF-S} \citep{wo04} with well-known intrinsic properties
324: (i.e., magnitudes and colors). The survey provides reliable object
325: classifications (e.g., Galaxy, QSO, Star) and photometric redshifts by
326: fitting synthetic template optical spectra to the observed magnitudes
327: over a broad wavelength range (3500--9300 {\rm \AA}). The source
328: catalog contains 8,565 objects with aperture magnitudes
329: $R_{ap}\leq24$, a limit at which there are photometric redshift errors
330: of $\delta_z/(1+z) < 0.1$. This magnitude limit ensures that our
331: sample is representative of galaxies of all colors with $M_V\lesssim
332: -21$ and $z<1$, a limit that will be shown to be relevant for this
333: study. Full details including source detection, photometry, object
334: classification, and redshift estimation can be found in \citet{wo04}.
335: Morphological information is provided by {\it HST}/ACS imaging of the
336: field in both the F606W and F850LP filters (hereafter referred to
337: $V_{606}$ and $z_{850}$, respectively) from GEMS. Here, we use the
338: S\'{e}rsic indices ($n$) given in \citet{ha07} to discriminate between
339: bulge and disk-dominated galaxies.
340:
341: We select a sample of 5,549 galaxies with photometric redshifts of
342: $0.4\leq z\leq 1.1$ from the GEMS $z_{850}$-band selected catalog of
343: \citet{ca05} that has been cross-referenced to objects in COMBO-17
344: ($R_{ap}\leq24$). This ensures that the sample can be used for
345: morphological studies and allows an upper limit to be placed upon the
346: optical contribution from each AGN to the color of its host galaxy.
347: The sample includes only those objects that are classified as a
348: `Galaxy' and excludes those identified as a `QSO'. This selection
349: effectively avoids the presence of luminous type 1 AGN that severely
350: dilute the emission from their host galaxies. The lower redshift
351: limit is set to 0.4 because this volume within the \hbox{E-CDF-S} is
352: too small at lower redshifts to generate statistically useful samples
353: of luminous ($M_V\lesssim-21$) galaxies and hence those with AGN.
354: Only two AGN have been unambigously identified at $z<0.4$ that have
355: $L_{\rm X}\sim10^{42}$ erg s$^{-1}$. The redshift cutoff at 1.1 is
356: the limit for which COMBO-17 no longer provides accurate source
357: classification, and redshifts are susceptible to large uncertainties.
358: We primarily use the rest-frame optical magnitudes ($M_U$, $M_V$; Vega
359: magnitudes), publicly available from COMBO-17 \citep{wo04}, derived
360: from synthetic galaxy templates to measure intrinsic luminosity and
361: color. To compare with theoretical models ($\S$~\ref{text:mergers}),
362: we utilize the rest-frame $u$ and $r$ (SDSS) also provided by
363: COMBO-17. The errors \citep[$1\sigma$; ][]{wo04} on rest-frame
364: magnitudes are typically 0.11 ($M_U$) and 0.15 mag ($M_V$).
365: Throughout this work, a representative error ($1\sigma$) on color (U-V)
366: is determined to be 0.19 from the quadrature of the errors of the
367: individual rest-frame magnitudes; this is most likely an overestimate
368: since these magnitude errors are correlated. \citet{be04} report a
369: typical error on U--V to be $\sim0.1$ mag.
370:
371: \subsection{AGN identification}
372:
373: We compile a sample of AGN, based on their high X-ray luminosities, in
374: the \hbox{E-CDF-S} by matching the 762 {\it Chandra} point sources
375: given in \citet{le05} to the available optical and near-infrared
376: catalogs. We do not consider here the fainter X-ray sources solely
377: detected in the 1~Ms CDF-S to maintain a fairly uniform sensitivity
378: across the entire \hbox{E-CDF-S}. Redshifts are available for 362
379: (48\%) of the X-ray sources through either spectroscopic or
380: photometric techniques. We significantly improve upon the 97
381: spectroscopic redshifts available in the literature
382: \citep{sz04,lef04,mi05,va05,va06} with 95 additional redshifts
383: acquired by observations with VIMOS \citep{lef03} on the VLT through
384: the ESO programs 072.A-0139 (P.I.: J. Bergeron; 65 redshifts;
385: Silverman et al. 2007) and 171.A-3045 (GOODS; P.I.: C. Cesarski; 30
386: redshifts; Popesso et al., in preparation). Our VIMOS observations
387: are unique since we have acquired spectra over a wide wavelength range
388: (3600--9500 ${\rm \AA}$) for many targets by observing with both the
389: LRblue and MRred grisms. With integration times reaching 5 hr, we are
390: able to detect faint spectral features (e.g., [O II], Mg II, Fe II) in
391: many of the host galaxies of obscured or optically-faint AGN that
392: enable a redshift measurement. A detailed discussion of the VIMOS
393: observations will be provided in a subsequent paper (Silverman et
394: al. 2007, in preparation). The X-ray luminosity of each AGN is
395: determined from their observed broad-band (0.5-8.0 keV) flux, given in
396: \citet{le05}, with a $k$-correction based on a power-law spectrum
397: (photon index $\Gamma=1.9$) and no correction for intrinsic
398: absorption. Here, we restrict the luminosity range to $41.9\leq
399: \log~L_{0.5-8.0~keV}\leq43.7$. The lower luminosity limit is chosen
400: to provide a robust sample free of any galaxies having significant
401: X-ray binary or diffuse X-ray emission; luminous starburst galaxies in
402: the $Chandra$ Deep Fields are typically a factor of $\sim10$ fainter
403: \citep[see Fig. 4 of ][]{ba02}. We also minimize potential
404: luminosity-dependent effects since lower luminosity AGN would only be
405: detected over a small fraction of the redshift range ($z =$~0.4--1.1)
406: considered here. The $Chandra$ observations of the E-CDF-S are
407: capable of detecting AGN at this lower luminosity limit up to
408: $z\sim0.85$. Our upper limit is motivated by \citet{si05} who show
409: that the optical emission, associated with X-ray selected AGN at
410: $\log~\nu l_{\nu}<43.3$ at $E = 2$~keV is primarily due to their host
411: galaxy since there is a strong departure of these AGN from the known
412: $l_{\rm opt}-l_{\rm X}$ relation for more luminous X-ray selected AGN.
413: We convert this limit to a broad-band (0.5--8.0 keV) X-ray luminosity
414: of $10^{43.7}$ erg s$^{-1}$ assuming the same spectrum as given above.
415: Furthermore, our adopted upper limit is similar to that of
416: \citet{na07} who demonstrate that the host galaxies of
417: moderate-luminosity AGN, at X-ray fluxes equivalent to those detected
418: in the \hbox{E-CDF-S}, contribute the majority of the total optical
419: emission. As detailed below, we have measured upper limits to the AGN
420: contribution, using the {\it HST}/ACS imaging of the field, and show
421: that the optical emission, for the majority of our sample, is
422: dominated by the host galaxy. Optical spectra provide a final check
423: on the AGN contribution with most sources clearly lacking any faint
424: broad emission lines or a rising blue continuum.
425:
426: A cross-correlation of the X-ray and optical catalogs has indicated
427: that 109 of the 5,549 galaxies have AGN with $L_{\rm X} \approx
428: 10^{41.9}$--$10^{43.7}$ erg~s$^{-1}$. Of these, spectroscopic
429: redshifts were available for 54 of these sources, 16 (7 at $z>1$) of
430: which were from our VIMOS observations. Rest-frame magnitudes for 12
431: galaxies were rederived (C. Wolf, private communication) using
432: COMBO-17 tools since their spectroscopic redshifts differed
433: substantially from their photometric redshifts ($\Delta z > 0.1$) that
434: rendered their rest-frame optical magnitudes published by COMBO-17 as
435: inaccurate. This is not surprising since seven of them have
436: spectroscopic redshifts of $z\sim1$, a regime where photometric
437: redshifts for galaxies in COMBO-17 show a higher dispersion \citep[see
438: Figure 6 of][]{wo04}, and lack strong continuum features (e.g.,
439: 4000~\AA~break) most useful for adequate photometric redshift
440: estimates. Our AGN sample is primarily radio-quiet since only nine
441: have a radio-loudness ($R$) greater than 10 ($R \equiv L_{6~{\rm
442: cm}}/L_{4400~{\rm\AA}}$) based on 20cm VLA detections (Tozzi et al. in
443: preparation) that reach a flux limit of 8.5 $\mu Jy$, optical emission
444: attributed to the AGN (25\%; see the following section) based on our
445: $R$-band optical magnitudes, and spectral indices ($L_{\nu} \propto
446: \nu^{-\gamma}$; $\gamma_{radio}=0.8$, $\gamma_{opt}=0.5$). We refer
447: the reader to \citet{ro07} for further details on the radio properties
448: of X-ray selected AGN in the E-CDF-S. We further note that an
449: additional 31 AGN have redshifts and luminosities within our range of
450: interest though fell out of the sample due to their faint optical
451: magnitudes ($R_{ap}>24$). Our final sample of 109 AGN allows us to
452: compare the properties of their host galaxies to the underlying galaxy
453: population.
454:
455: \subsection{AGN contribution to host-galaxy emission}
456: \label{contamination}
457:
458: We are confident that the optical emission, from these 109 galaxies
459: hosting AGN, is dominated by star light and not strongly influenced by
460: the AGN based on the following arguments. In Figure~\ref{agn_sample},
461: we show that there is no correlation between X-ray luminosity and
462: either rest-frame magnitude ($M_V$; Pearson correlation coefficient
463: $r=-0.036$) or color ($U-V$; $r=-0.233$). This is consistent with the
464: results of \citet{na07} who show that only a weak correlation exists
465: for X-ray sources at similar low X-ray and optical fluxes. The lack
466: of any correlation in our sample is due to the fact that we
467: deliberately excluded the luminous ($\log L_{\rm X} \gtrsim 44$) AGN
468: and any objects classified by COMBO-17 as a ``QSO''.
469:
470: \begin{figure}
471:
472: \includegraphics[scale=0.42,angle=0]{f1a.eps}
473: \includegraphics[scale=0.42,angle=0]{f1b.eps}
474: \caption{(a) Rest-frame absolute magnitude ($M_V$) and (b) optical
475: color ($U-V$) as a function of X-ray luminosity for our sample of 109
476: AGN (filled symbols). The type of marker denotes AGN within three
477: separate redshift intervals (circles: $z=$~0.4--0.63; squares:
478: $z=$~0.63--0.76; triangles: $z=$~0.76--1.1). Open symbols show higher
479: luminosity AGN not included in our sample. A typical error bar of
480: size $\pm1\sigma$ is placed in the upper right corner in both panels.}
481: \label{agn_sample}
482: \end{figure}
483:
484: We determine conservative upper limits upon the AGN contribution to
485: the total (galaxy + AGN) light using the {\it HST}/ACS $V_{606}$ and
486: $z_{850}$-band images provided by GEMS. These images, available from
487: MAST (Multi-mission Archive at Space Telescope), have been through the
488: $mark1$ processing as described in \citet{ca05}, and are rescaled to
489: 0.03\arcsec~pixel$^{-1}$. A few AGN are not included in our analysis
490: due to their proximity to the ACS field edge. Extended optical
491: emission is clearly evident for all 109 galaxies hosting AGN. Optical
492: counts are measured without subtracting the background in circular
493: apertures of two different sizes positioned at the centroid of the
494: optical emission: (1) a small aperture with a radius of 3 pixels
495: (0.09$\arcsec$) that contains $50\%$ of the flux from an unresolved
496: point source \citep[see][]{ja04b} and corresponds to a physical scale
497: of 0.48 ($z=0.4$) to 0.74 ($z=1.1$) kpc, and (2) a larger aperture
498: with a radius of 25 pixels (0.75$\arcsec$; covering a physical scale
499: of 4.03--6.13 kpc). The size of the larger aperture is set to that
500: implemented for flux extraction by COMBO-17 \citep{wo04}. The counts
501: within the small aperture provide a firm upper limit to the AGN
502: contribution since we make no attempt to remove emission of stellar
503: origin. For comparison, we measure counts in equivalent regions for
504: the entire sample of 5,549 galaxies and 67 QSOs, which were identified
505: by COMBO-17 and are not included in our parent galaxy sample.
506:
507: The number distribution of the AGN host galaxies as a function of the
508: ratio of counts between these two apertures is shown in
509: Figure~\ref{contamination} for both the $V_{606}$-band (Fig.~2a) and
510: $z_{850}$-band (Fig.~2b). Based on the $V_{606}$-band measurements,
511: the mean ratio of counts of our AGN sample (solid histogram) is $0.14$
512: and $81\%$ of them have a ratio less than 0.2. The distribution is
513: similar to that of the galaxies (dotted histogram), though shifted
514: slightly by 0.04 (the difference of their median values), most likely
515: due to the presence of optically-faint AGN. The host galaxy
516: distribution is noticeably offset from that of the optically-selected
517: QSOs, identified by spectral template fitting with no preference for
518: point-like sources \citep{wo04}, that have a mean ratio of 0.48, and
519: 60\% of them have a ratio greater than 0.5 most indicative of
520: unresolved point sources. A number of low count ratio QSOs are
521: evident due to the fact that COMBO-17 can recognize lower luminosity
522: Seyfert-1 galaxies as QSOs if strong AGN features are present.
523: Similar results are found for the $z_{850}$-band measurements, which
524: suggests that color gradients are dominated by the host galaxies and
525: not the underlying AGN. Since our sample contains a large fraction of
526: obscured AGN, there is a slightly larger contribution from AGN
527: emission in the $z_{850}$ band compared to the $V_{606}$ band, whereas
528: the opposite is true for the quasars.
529:
530: \begin{figure}
531: \includegraphics[scale=0.55,angle=0]{f2.eps}
532: \caption{AGN contribution to the total (AGN + galaxy) light. The
533: abscissa is the ratio of source counts in {\it HST}/ACS images between
534: a circular aperture of radius 0.09\arcsec~and 0.75\arcsec. In panel
535: a, the $V_{606}$-band count-ratio distribution is shown for 104 AGN
536: (solid histogram), 5521 galaxies (dotted), and 67 (dashed)
537: optically-selected QSOs from COMBO-17. In panel {\it b}, the same is shown
538: for the $z_{850}$-band (103 AGN, 5513 galaxies, 66 QSOs). QSOs from
539: COMBO-17 illustrate the typical count ratios for unresolved point
540: sources. The galaxy number distribution in both panels has been
541: scaled down to match the numbers of AGN host galaxies. The majority
542: of the AGN host galaxies have $\gtrsim 80\%$ of their total optical
543: emission (in both the $V_{606}$ and $z_{850}$ bands) outside the
544: circular aperture of radius 0.09\arcsec\ and have count ratios similar
545: to the non-AGN galaxy population.}
546: \label{contamination}
547: \end{figure}
548:
549: The maximum amount that an AGN could shift its host-galaxy rest-frame
550: $U-V$ color bluewards is estimated. We assume that an AGN contributes
551: all the flux in the small ($r=0.09\arcsec$) aperture, the count ratio
552: as detailed above is 0.2, and the background signal is negligible.
553: This corresponds to an AGN contributing an additional amount of flux
554: equal to 0.25 times the total host galaxy flux. We further consider
555: an AGN to add flux to the $U$-band only. While rest-frame UV emission
556: is substantially diminished at $\lambda<4000~{\rm \AA}$ for a typical
557: galaxy, broad-line AGN \citep{vanden06,ga06} have a strong rise in
558: their continuum at these wavelengths effectively driving their color
559: bluewards with increasing contrast between the AGN and its host-galaxy
560: emission as evident in Figure~\ref{contamination} with the $V_{606}$
561: band distribution in panel {\it a} shifted higher than that in panel
562: {\it b}. We note that the rest-frame $U$-band is observed with the
563: $V_{606}$ band for objects with $z\sim0.7$. Based on these
564: assumptions, color offsets for 81\% of our AGN sample are within
565: \hbox{0.24} mag. Furthermore, we do not expect our AGN to reach this
566: value since a significant amount of stellar emission is present in all
567: cases within the small aperture and AGN light is also likely to be
568: emitted at longer wavelengths. We note that if color offsets do reach
569: this level, the contamination is still insufficient to displace AGN
570: hosts from one region to another (e.g, ``green valley'' to the red
571: sequence). It is more likely that the AGN contribute an additional
572: $\sim4\%$ to the total host galaxy flux based on the shift between the
573: median count ratio of the AGN host galaxies and the full galaxy
574: population described above. In $\S$~\ref{discussion}, we utilize
575: optical spectra, which are available for 46 AGN, to further confirm
576: that the AGN host galaxies provide most of the optical light (see
577: Fig.~\ref{examples_sf} and~\ref{examples_valley}), thus supporting our
578: general conclusions.
579:
580: \section{Rest-frame colors of moderate-luminosity AGN}
581:
582: We were motivated by recent studies \citep{sa04,bo06,na07} to
583: investigate further the color-magnitude (rest-frame $U-V$ versus
584: $M_V$) relation of galaxies hosting moderate-luminosity X-ray selected
585: AGN. Since the host galaxies in our sample contribute most of the
586: optical emission, as demonstrated in $\S$~\ref{contamination}, we do
587: not need to remove the AGN component from the total (host + AGN)
588: optical emission. We use the rest-frame $U-V$ (i.e., $M_U-M_V$) color
589: up to $z\sim1$, following \citet{be04}, to quantify the color
590: distribution with a greater sensitivity to the continuum slope across
591: the 4000 ${\rm \AA}$ break than is provided by rest-frame $U-B$ color.
592: We note that \citet{wo04} caution that the rest-frame measurement of
593: $M_V$ at $z>0.7$ is based on an extrapolation of the best-fit galaxy
594: template outside of the observed optical passbands in COMBO-17. For
595: all results presented here, we have confirmed that the same basic
596: conclusions are drawn by using the $M_U$ and $M_B$ rest-frame absolute
597: magnitudes, thus removing the possibility that our results are
598: dependent on the assumed spectral template (i.e., $k$-correction).
599:
600: In Figure~\ref{color_mag}a, we plot the rest-frame colors ($U-V$) as a
601: function of absolute $V$-band magnitude ($M_V$) for our galaxies,
602: including those hosting AGN in the redshift interval $0.4\leq z \leq
603: 1.1$. We confirm past results \citep{ba03,bo06,na07} with better
604: statistics: (1) a high fraction (80\%) of moderate-luminosity AGN
605: reside in the most-luminous ($M_V<-20.7$) galaxies, (2) the rest-frame
606: colors of AGN host galaxies have a broad distribution, over the range
607: $0 < U-V < 1.5$, with no apparent evidence for a color bimodality, as
608: is distinctively evident in the underlying population of galaxies
609: \citep{be04}, and (3) the majority (60\%) of AGN host galaxies have
610: bulge-dominated morphologies \citep[S\'{e}rsic index $n>2.5$;
611: ][]{bl03,mc05} as marked by small blue dots. Here, we further find
612: that 31--44\% of the AGN with $M_V<-20.7$ have blue colors, depending
613: on the chosen division ($0.6<U-V<0.8$) between blue and red galaxies
614: \citep[dashed line in Fig.~\ref{color_mag}a; see][]{sa04}, thus
615: associating them with star-forming galaxies. This is not surprising
616: since we have a fair number of AGN at $z>0.8$, an epoch for which the
617: mean star-formation rate of galaxies has increased by an order of
618: magnitude compared to the present value
619: \citep[e.g.,][]{ho06,no07,zh07}. These results are not strongly color
620: biased since both red and blue luminous ($M_V<-20.7$) galaxies, out to
621: $z\sim1$, mainly fall above the magnitude limits
622: \citep[$R_{ap}\lesssim24$; ][]{be04} shown by the blue lines in
623: Figure~\ref{color_mag}a ($z=0.8$: solid; $z=1.0$: dash-dotted). We
624: note that some red ($U-V>1.0$), less-luminous ($-20.7<M_V<-21.3$)
625: galaxies at $z>0.9$ may be missed.
626:
627: \begin{figure*}
628:
629: \includegraphics[scale=0.70,angle=90]{f3.eps}
630: \caption{(a) Rest-frame optical color ($U-V$) as a function of
631: rest-frame absolute magnitude ($M_V$) for 109 galaxies hosting AGN
632: compared to their parent sample of 5,549 galaxies with $0.4\leq z \leq
633: 1.1$. Galaxies in the parent sample are marked with small black
634: filled circles; those hosting X-ray selected AGN are highlighted by
635: large red circles. The slanted blue lines denote the approximate
636: limits for galaxies with $R_{ap}\le24$ at $z=0.8$ (solid) and $z=1$
637: (dash-dotted). The division between red and blue galaxies implemented
638: by \citet{be04} is shown by the dashed line. AGN hosts classified as
639: bulge-dominated (S\'{e}rsic index 2.5--8) galaxies by the {\it
640: HST}/ACS morphology \citep{ha07} are marked with blue filled circles
641: at the centers of the red circles. AGN with X-ray hardness ratios
642: larger than 0.2 (see $\S$~\ref{text:mergers}) are marked by open green
643: squares in both panels. The typical error ($\pm1\sigma$) bar is shown
644: in the bottom right corner. (b) Rest-frame $U-V$ color versus
645: redshift for the 2,044 most luminous ($M_V<-20.7$) galaxies. The
646: vertical solid lines denote the redshift interval $0.63\leq z \leq
647: 0.76$ with the dotted lines marking the redshift spikes at $z=0.67$
648: and $z=0.73$. In this panel only, smaller blue dots mark the AGN with
649: spectroscopic redshifts.}
650:
651: \label{color_mag}
652: \end{figure*}
653:
654: It is apparent that the color distribution of these AGN clearly
655: depends on redshift as shown in Figure~\ref{color_mag}b. Here, we
656: only consider luminous galaxies ($M_V<-20.7$) to minimize any color
657: bias. In the redshift interval $0.4<z<0.63$ (see
658: Fig.~\ref{color_histo}a), the AGN tend to have red colors ($U-V>0.7$)
659: with a mean color $\langle U-V \rangle$ = $0.86\pm0.10$. Within the
660: same redshift interval, the galaxies without AGN are also
661: preferentially red. We find that 70\% (14 of 20) of the AGN and 62\%
662: (239 of 386) of the galaxies have these colors with most associated
663: with the red sequence. In contrast, the mean color of AGN hosts in a
664: higher redshift interval ($0.76<z<1.1$; Fig.~\ref{color_histo}c) is
665: $\langle U-V \rangle$ = $0.49\pm0.08$ with 29\% (9 of 31) of the AGN
666: having $U-V>0.7$. A similar fraction (27\%; 211 of 790) of luminous
667: galaxies at these redshifts has these colors. The variance of each
668: color distribution is similar ($s^2\sim0.2$). Again, it is as
669: expected that the majority of AGN host galaxies at $z>0.76$ have blue
670: colors ($U-V<0.7$) since blue galaxies dominate the luminous
671: population at these redshifts \citep{wo03,be04}. The null hypothesis,
672: from a Kolmogorov-Smirnov (K-S) test \citep[see ][]{pr93}, has a
673: probability of 2.4\% that the distribution of host galaxy colors in
674: Figure~\ref{color_histo}a and~\ref{color_histo}c could be drawn from
675: the same parent population. We have implemented further K-S tests to
676: determine whether the AGN distributions could be drawn from the
677: underlying galaxy population. In Table~\ref{stats}, we give the
678: results that show that the color distribution of AGN host galaxies in
679: Figure~\ref{color_histo}a ($P_{\rm K-S}=0.79$) and \ref{color_histo}c
680: ($P_{\rm K-S}=0.21$) resembles that of the overall galaxy population.
681: {\it We conclude, based on these tests, that the host galaxies of
682: moderate-luminosity AGN follow a similar passive evolution, or aging,
683: as the underlying galaxy population migrates from blue to red colors
684: with cosmic time}. We note that AGN with spectroscopic redshifts,
685: shown by a small blue dot in Figure~\ref{color_mag}b, confirm this
686: trend.
687:
688: \begin{figure*}
689: \includegraphics[scale=0.65,angle=90]{f4.eps}
690: \caption{Rest-frame color ($U-V$) histogram of galaxies that host AGN
691: (solid line) with absolute magnitude $M_V<-20.7$ in three redshift
692: intervals: $0.4<z<0.63$ (a), $0.63\leq z \leq0.76$ (b), $0.76<z<1.1$
693: (c). For comparison, we have plotted, in all panels, the distribution
694: of galaxies (dotted line) above this absolute magnitude and
695: renormalized to match the number of AGN. Note that the peak of the
696: distribution shifts from red to blue with increasing redshift for both
697: galaxies hosting AGN and the underlying galaxy population.}
698: \label{color_histo}
699: \end{figure*}
700:
701: \begin{deluxetable*}{lllrc}
702: \tabletypesize{\small}
703: \tablecaption{Statistical comparison of color distributions\label{stats}}
704: \tablewidth{0pt}
705: \tablehead{\colhead{Note}&\colhead{\# AGN}&\colhead{\# galaxies}&\colhead{K-S}&\colhead{Redshift range}}
706: \startdata
707: AGN-galaxy (Fig.~\ref{color_histo}a)&20&415&79.4\%&$0.4<z<0.63$\\
708: AGN-galaxy (Fig.~\ref{color_histo}b)&36&669&6.6\%&$0.63 \leq z \leq 0.76$\\
709: AGN-galaxy (Fig.~\ref{color_histo}c)&31&957&20.6\%&$0.76 < z < 1.1$\\
710: AGN (Fig.~\ref{color_histo}a)-AGN (Fig.~\ref{color_histo}c)&--&--&2.4\%&-----\\
711: AGN (Fig.~\ref{color_histo}a)-AGN (Fig.~\ref{color_histo}b)&--&--&13.4\%&-----\\
712: AGN (Fig.~\ref{color_histo}b)-AGN (Fig.~\ref{color_histo}c)&--&--&0.2\%&-----\\
713:
714: \enddata
715: \end{deluxetable*}
716:
717: \section{Large-scale influences}
718:
719: As previously mentioned, recent studies \citep{di05,cr06,hop06a} have
720: attributed the truncation of star formation and eventual redward
721: migration of galaxies to merger-induced AGN feedback that effectively
722: populates the red sequence with massive galaxies \citep{be04,fa06}.
723: This scenario may explain the fair number of AGN host galaxies in our
724: sample residing in the ``green valley'' (see Fig.~\ref{color_mag}a) if
725: they are preferentially located in overdense regions.
726:
727: In contrast to studies that characterize the environment in terms of
728: density local to AGN, we are utilizing the fortuitous structures in
729: the \hbox{E-CDF-S} to search for large-scale effects. To do so, we
730: specifically isolate a redshift interval ($0.63\leq z \leq 0.76$)
731: dominated by two redshift spikes \citep[Fig.~\ref{color_mag}b,
732: vertical dotted lines at $z=0.67$ and $z=0.73$, each with $\delta
733: z<0.02$;][]{gi03} evident in the 1~Ms CDF-S area, which appear to
734: extend over the larger \hbox{E-CDF-S} area (Silverman et al. 2007). A
735: larger fraction of AGN activity in galaxies within these large-scale
736: structures was reported by \citet{gi03} albeit with limited
737: significance ($2\sigma$). Ideally, one would like to consider even
738: narrower redshift intervals ($\delta z \sim 0.02$) to select galaxies
739: cleanly within these spectroscopic redshift spikes, but we are
740: restricted here by photometric redshift errors. Approximately 80\% of
741: the galaxies within this redshift interval have errors in redshift
742: ($\sigma_z$) less than 0.07\footnote{Photometric redshift errors are
743: based on equation 5 of \citet{wo04} and a magnitude limit of 23.5.}, a
744: factor of two smaller than the chosen bin width ($\Delta z=0.14$).
745:
746: It is evident from Figure~\ref{color_mag}b that the majority of AGN
747: within this redshift interval have colors ($0.5<U-V<1.0$) placing them
748: in the ``green valley''. Many of the AGN have spectroscopic
749: redshifts, shown by the small blue dots, that confirm their presence
750: within the narrow redshift spikes ($\Delta z < 0.02$) and their
751: rest-frame $U-V$ colors that place them in the ``green valley'' (67\%;
752: 12 of 18 AGN). We plot in Figure~\ref{color_histo}b the $U-V$ color
753: distribution of both AGN and galaxies with $M_V<-20.7$ that fall
754: within this redshift interval. We have scaled the galaxy distribution
755: to match the number of AGN. The color distribution of the host
756: galaxies is clearly different from the underlying galaxy population.
757: {\it We find that the host galaxies of moderate-luminosity AGN are
758: preferentially located at intermediate colors $(U-V\sim0.8)$ between
759: the red and blue galaxy populations, and skewed toward blue colors}.
760: A K-S test
761: \footnote{The results of additional tests between the color
762: distributions of AGN hosts (AGN-AGN) in the three redshift intervals
763: are reported in Table~\ref{stats}. We note that their significance is
764: limited given the lack of a large comparison sample that effectively
765: improves the AGN-galaxy comparisons.} gives a probability of 6.6\%
766: that the AGN distribution could be randomly drawn from the underlying
767: galaxy population and suggests that the distribution of these 36 AGN
768: is different from that of the 669 underlying galaxies.
769:
770: There does not appear to be any strong selection effects that could be
771: responsible for the color distribution of AGN host galaxies within
772: this redshift interval. Essentially, all of the AGN have optical
773: luminosities well above the limit at z=0.8 shown in
774: Figure~\ref{color_mag}a by the solid, blue line. By isolating
775: luminous ($M_V<-20.7$) host galaxies, there is clearly no color bias.
776: This is an important point since the $4000~{\rm \AA}$ break is moving
777: through the $R$-band filter at $z\sim0.7$ and red galaxies can
778: potentially fall below the flux limit more rapidy than those with
779: bluer colors. There should not be any problems associated with the
780: use of COMBO-17 filters at these redshifts since both rest-frame $M_U$
781: and and $M_V$ fall within the observable window, and multiple filters
782: sample the galaxies' spectral energy distribution below and above the
783: $4000~{\rm \AA}$ break. Both surface brightness dimming and the use
784: of a fixed photometric aperture can induce systematic changes in color
785: with redshift. An increase in the contrast between a bulge and a
786: disk, due to surface brightness dimming as a function of redshift,
787: would typically redden its rest-frame color; this effect is of most
788: concern when comparing colors over a wider redshift baseline.
789: \citet{be04} address aperture effects associated with COMBO-17
790: photometry and state that an induced color gradient is only evident at
791: $z\lesssim0.4$, below the redshift range of our sample, with a
792: potential color offset of $\sim0.1$ mag. There does exists the
793: possibility that the color distribution of AGN hosts is the result of
794: red and blue galaxies having roughly equivalent numbers at $z\sim0.7$,
795: coupled with our small AGN sample and sensitivity to cumulative biases
796: though our statistical test suggests otherwise.
797:
798: \section{AGN fraction}
799:
800: \label{text:agn_fraction}
801:
802: We measure the fraction of galaxies that host AGN as a function of
803: color and large-scale environment. We follow the technique discussed
804: in $\S$~3.1 of \citet{le07} to determine the AGN fraction for our
805: parent population of 5,549 galaxies. This method properly accounts
806: for the spatially varying sensitivity limits of the {\it Chandra}
807: observations of the \hbox{E-CDF-S} \citep[see Fig.~17 of][]{le05}.
808: The necessity of this approach is demonstrated in
809: Figure~\ref{sample_limits}, which shows the limiting X-ray luminosity
810: as a function of redshift for the entire galaxy sample and the
811: measured X-ray luminosities of those galaxies harboring AGN. In
812: addition to the Malmquist bias, there is almost an order-of-magnitude
813: spread in limiting luminosity over the full redshift range. To
814: illustrate these effects, the fraction of galaxies that could host a
815: detectable AGN with $\log~L_{\rm X}=41.9$ is $\approx 23\%$; this
816: fraction rises steeply to $\approx 80\%$ when considering a limiting
817: luminosity of 42.5. To account for this selection effect, we
818: determine the contribution of each AGN separately to the total
819: fraction. The AGN fraction ($f$; see equation{~\ref{fraction} below)
820: and associated error ($\sigma$; see equation~\ref{fraction_error}) are
821: a sum over the full sample of AGN ($N$) with $N_{\rm gal,i}$
822: representing the number of galaxies capable of hosting the $i$th
823: detectable AGN with X-ray luminosity $L^{i}_{\rm X}$.
824:
825: \begin{equation}
826: f=\sum_{i=1}^{N}\frac{1}{N_{\rm gal,i}}
827: \label{fraction}
828: \end{equation}
829:
830: \begin{equation}
831: \sigma^2\approx\sum_{i=1}^{N}\frac{1}{{N_{\rm gal,i}^{2}}}
832: \label{fraction_error}
833: \end{equation}
834:
835: \begin{figure}
836: \includegraphics[scale=0.42,angle=0]{f5.eps}
837: \caption{The full band X-ray luminosity as a function of redshift for
838: the 5,549 galaxies in our parent sample. The filled circles are the
839: X-ray detections (AGN); the open circles are upper limits. The spread
840: of a factor of six in the value of the limits at a given redshift is
841: produced by the variations in survey sensitivity with location; for a
842: given redshift, the objects with the lowest limits are located at the
843: center of each ACIS pointing and those with the highest limits are
844: located at the edge of the ACIS array. The horizontal line denotes
845: the minimum luminosity for inclusion in our AGN sample.}
846: \label{sample_limits}
847: \end{figure}
848:
849: We have calculated the fraction of galaxies harboring AGN in bins of
850: color for the entire sample (Fig.~\ref{agn_fraction}a). Using a bin
851: width of $\Delta (U-V) = 0.4$, we find that the fraction rises from
852: $\sim0\%$ at the bluest colors to $\sim5\%$ at $U-V\sim0.4$ (i.e., red
853: end of the blue-galaxy population), has a peak in the ``green valley''
854: ($U-V\sim0.8$) with a value of $\sim10\%$, and then declines to
855: $\sim4-5\%$ at $U-V\sim1.3$ (i.e., along the red sequence). In
856: Table~\ref{table_fraction}, we list the fraction of luminous
857: ($M_V<-20.7$) galaxies hosting AGN that we measure for various
858: subsamples of galaxies. We note that the AGN fraction in the `field'
859: (i.e., $0.4\leq z<0.63$ and $0.76<z \leq 1.1$) is $5.9 \pm 1.0\%$ (51
860: AGN from effectively 1157 galaxies) when considering all colors in one
861: bin. Remarkably, the distinct enhancement of the AGN fraction in the
862: ``green valley'' is due to an overdensity of AGN in the redshift
863: interval (0.63--0.76) that contains two prominent redshift spikes. In
864: Figure~\ref{agn_fraction}b, we show the fraction of galaxies hosting
865: AGN in (solid line) and out (dotted line) of this narrow redshift
866: interval. For an intermediate color interval $0.5<U-V<1.0$, the AGN
867: fraction within this narrow redshift range (solid line) is
868: $12.8\pm2.9\%$ (21 AGN, 173 galaxies), that reaches $\sim15$\% at
869: $U-V\sim0.8$, while those in the `field' (dotted line) have a fraction
870: of $7.8\pm2.5\%$ (14 AGN, 246 galaxies). The significance of this
871: higher AGN fraction in large-scale stuctures is reported here at the
872: $2.6\sigma$ ($>99\%$) confidence level based on the difference and
873: errors of these measurements. This analysis improves upon the
874: significance of enhanced AGN activity within large-scale structures
875: previously reported by \citet{gi03}. A less conspicuous peak is
876: apparent in the `field' sample though not sufficiently significant
877: given the current sample size. We find an overall higher AGN fraction
878: in the 'field', compared to \citet{na07}, by a factor of $\sim$1.8
879: though a similar $relative$ fraction as a function of color
880: (Fig.~\ref{agn_fraction}b; dotted line); the AGN fraction appears to
881: be rather flat ($\sim6\%$) across a wide range of color ($0.3 \lesssim
882: U-V \lesssim 1.5$) that includes the red sequence and the valley but
883: then drops off for blue galaxies beyond the top ($U-V<0.2$) of the
884: `blue cloud'.
885:
886: \begin{figure}
887: \includegraphics[scale=0.67,angle=90]{f6a.eps}
888: \includegraphics[scale=0.67,angle=90]{f6b.eps}
889: \caption{Fraction of galaxies ($M_V<-20.7$) hosting AGN as a function
890: of their rest-frame optical color ($U-V$) for the entire sample (a).
891: In panel $b$, we have split the sample into those in (solid) and out
892: (dashed) of the redshift interval $0.63 \leq z \leq 0.76$. Errors
893: (1$\sigma$) are shown for all values with a slight displacement of the
894: dashed set for visual purposes only. A suggestive enhancement
895: ($\sim$~two~$\times$ with significance at the $2.6\sigma$ level) of AGN
896: activity is evident for host galaxies residing in the ``green valley''
897: that is mainly attributed to activity in large-scale structures.}
898: \label{agn_fraction}
899: \end{figure}
900:
901: \begin{deluxetable*}{rllccl}
902: \tabletypesize{\small}
903: \tablecaption{Fraction of luminous ($M_V<-20.7$) galaxies hosting AGN\label{table_fraction}}
904: \tablewidth{0pt}
905: \tablehead{\colhead{Index}&\colhead{Redshift\tablenotemark{a}}&\colhead{$U-V$}&\colhead{S\'{e}rsic}&\colhead{Fraction}&\colhead{Notes}\\
906: &\colhead{range}&\colhead{range}&\colhead{indices}&\colhead{\%}}
907: \startdata
908: 1&I+II+III&---&---&$6.4\pm0.8$&Full sample\\
909: 2&I+II+III&$<0.5$&---&$4.6\pm0.9$&Blue cloud\\
910: 3&I+II+III&0.5--1.0&---&$9.8\pm1.9$&``Green valley''\\
911: 4&I+II+III&$>1.0$&---&$6.0\pm1.5$&Red sequence\\
912: 5&I+III&---&---&$5.9\pm1.0$&`Field sample'\\
913: 6&I+III&0.5--1.0&---&$7.8\pm2.5$\\
914: 7&II&0.5--1.0&---&$12.8\pm2.9$&Redshift spikes\\
915: 8&I+II+III&---&0--2.5&$1.9\pm0.5$&disks\\
916: 9&I+II+III&---&0--1.5&$1.0\pm0.4$&purely disks\\
917: 10&I+II+III&---&2.5--8&$10.6\pm1.7$&bulges\\
918: 11&I+II+III&$<0.7$&2.5--8&$21.3\pm5.0$&blue spheroids\\
919: 12&I+II+III&$>0.7$&2.5--8&$7.9\pm1.7$&red spheroids\\
920:
921: \enddata
922: \tablenotetext{a}{I: $z=$~0.4--0.63; II: $z=$~0.63--0.76; III: $z=$~0.76--1.1}
923: \end{deluxetable*}
924:
925: \section{Discussion}
926: \label{discussion}
927: \subsection{Global evolution of host-galaxy colors}
928:
929: It is useful to ask what would be the observed color distribution of
930: the hosts of AGN as a function of redshift up to $z\sim1$ if we simply
931: assume that AGN reside in the most luminous galaxies, and hence the
932: most massive, as corroborated by many findings
933: \citep[e.g.,][]{ka03,be05}. Based on the luminosity function of
934: galaxies from the \hbox{COMBO-17} \citep{wo04} and DEEP2 \citep{fa06}
935: surveys, it is evident \citep[Fig.~6 of][]{fa06} that luminous
936: ($M_B< -21$) galaxies at $z\lesssim0.6$ are predominately red (i.e.,
937: early-type). At $z\gtrsim0.6$, a transition occurs where the majority
938: of luminous galaxies are blue (i.e., late-type, star-forming) as a
939: result of (1) the order-of-magnitude increase in the mean
940: star-formation rate \citep[e.g., ][]{ho06}, across all mass scales
941: \citep[e.g.,][]{no07,zh07}, and (2) strong depletion of galaxies along
942: the red sequence \citep{be04,be07} up to $z\sim1$. In
943: Figure~\ref{galaxy_distributions}, we show the distribution of
944: rest-frame $U-V$ colors for redshift-selected galaxy populations in
945: our sample. The blue peak ($U-V\sim0$) of the luminous ($M_V<-20.7$)
946: galaxy distribution is most prominent in the highest redshift bin
947: ($z=$~0.76--1.1; dotted histogram), while the red peak is more fully
948: populated in the lowest redshift bin ($z=$~0.4--0.63; dashed
949: histogram).
950:
951: We plausibly expect to see a similar redshift dependence in the color
952: distribution of galaxies hosting AGN with redshift. Observational
953: evidence does exist to support this scenario. Host-galaxy studies,
954: pioneered by {\it HST}, show that early-type galaxies represent the
955: majority of the hosts of QSOs at low redshifts \citep[$z<0.3$;
956: e.g.,][]{ba97}. At similar redshifts, \citet{ka03} demonstrate that
957: the hosts of over 20,000 narrow line AGN from the SDSS are mainly
958: early-type, bulge-dominated systems. The majority of the most
959: luminous, type II QSOs from the SDSS \citep{za06} and radio galaxies
960: \citep{mc04} appear to also have similar early-type hosts. At higher
961: redshifts ($z>0.5$), there is now evidence that a large fraction of
962: AGN hosts have blue colors and hence significant star formation
963: \citep{ja04b,sa04,wi06}. However, there is little evidence that the
964: host galaxies of AGN resemble the blue, star-forming galaxies that
965: dominate the galaxy population at $z>0.8$ as pointed out by
966: \citet{ba03} in reference to X-ray selected samples.
967:
968: \begin{figure}
969: \includegraphics[scale=0.67,angle=90]{f7.eps}
970: \caption{Color distribution of all luminous ($M_V<-20.7$) galaxies in three
971: redshift intervals (dashed: $z =$~0.4--0.63; solid: $z =$~0.63--0.76; dotted:
972: $z =$~0.76--1.1). All have been normalized by the number of objects in each
973: interval.}
974: \label{galaxy_distributions}
975: \end{figure}
976:
977: In our sample, the color distribution of luminous ($M_V<-20.7$) AGN
978: host galaxies in the `field' (i.e., outside the redshift interval
979: containing prominent redshift spikes) appears to support the above
980: scenario. In Figure~\ref{color_histo}, we see that the distribution
981: (solid histogram) has both a red (Fig.~\ref{color_histo}a) and a blue
982: (Fig.~\ref{color_histo}c) peak for the $0.4 < z < 0.63$ and $0.76 < z
983: < 1.1$ redshift bins, respectively; these peaks are coincident with
984: those of the underlying galaxy populations (dotted histogram). Since
985: the AGN sample is limited in size, this bimodality is only evident
986: when considering AGN over a wide redshift range and not at each epoch
987: as characteristic of the color distribution of large samples of
988: galaxies \citep[e.g.,][]{st01,bl03,ba04,ba06,ca07}. The colors of the
989: host galaxies at $z\lesssim0.6$ are predominately red ($U-V>0.7$;
990: Figure~\ref{color_mag}b) and lie on or close to the red sequence.
991: Optical spectra are available for five of the reddest ($U-V>1.0$) AGN
992: host galaxies and each has a strong 4000~\AA~break characteristic of
993: an old stellar population. At higher redshifts ($z\gtrsim0.8$), the
994: AGN\footnote{A handful appear to be associated with a less prominent
995: redshift spike identified by \citet{gi03} at $z\sim1.04$.} are more
996: prevalent in blue (i.e., star-forming) galaxies. Nine of these
997: high-redshift AGN with $U-V<0.4$ have optical spectra that confirm
998: that their colors are dominated by stellar emission. In
999: Figure~\ref{examples_sf}, we show representative optical spectra from
1000: our VLT program that are characterized by weak 4000~\AA~breaks, fairly
1001: flat blue continua, and no optical signatures of an embedded AGN in
1002: all but one case (Source \# 333). To summarize, we find that the
1003: color distribution of AGN host galaxies has a redshift dependency that
1004: reflects that of the underlying luminous galaxy population. It is
1005: also apparent that AGN in the `field' (Fig.~\ref{agn_fraction}b;
1006: dotted line) are equally likely to reside in galaxies over a broad
1007: range in color with the exception of those at the blue ($U-V<0.2$)
1008: end. This result suggests that AGN activity in the `field' is
1009: primarily dependent on mass\footnote{Using equation 2 of
1010: \citet{bell05}, we confirm that estimates of stellar mass of AGN host
1011: galaxies in our sample, based on their rest-frame colors, represent
1012: the high end of the galaxy mass distribution since we find a mean of
1013: $\sim1.3\times10^{10}$ M$_{\sun}$ and 90\% of our sample with $M
1014: \gtrsim 3.9\times10^{9}$ M$_{\sun}$, in agreement with
1015: optically-selected AGN host galaxies \citep{sa04}.}, given the strong
1016: dependency on observed luminosity and weaker in the luminous, blue
1017: (i.e., star-forming) galaxy population. In $\S~\ref{morphology}$, we
1018: demonstrate that the lack of an AGN-star formation connection only
1019: applies to disk-dominated galaxies.
1020:
1021:
1022: \begin{figure*}
1023: \includegraphics[scale=0.75,angle=0]{f8.eps}
1024: \caption{Examples of blue ($U-V<0.4$) galaxies hosting AGN at $z>0.8$
1025: that have been identified by our VLT/VIMOS observations. The {\it
1026: HST}/ACS $z_{850}$-band images are shown with a log scaling. The optical
1027: spectra that cover a wavelength range of 6000--8500~\AA~(right column)
1028: have been binned by 2 pixels. The flux scale is erg cm$^{-2}$
1029: s$^{-1}$ ${\rm \AA^{-1}}$. The source numbers follow those presented in
1030: the \citet{le05} X-ray catalog.}
1031: \label{examples_sf}
1032: \end{figure*}
1033:
1034: \subsection{Enhanced AGN activity in large-scale structures}
1035:
1036: The enhancement of galaxies hosting AGN, in large-scale structures
1037: reported in \citet{gi03} and substantiated further in this
1038: investigation, appears to indicate the physical scale most nuturing
1039: for accretion onto SMBHs. The fraction of galaxies hosting AGN
1040: (Fig.~\ref{agn_fraction}b) is significantly enhanced within a redshift
1041: interval ($0.63 \leq z \leq 0.76$) dominated by two prominent redshift
1042: spikes ($z=0.67$ and $z=0.73$) that are overpopulated with both AGN
1043: \citep{gi03} and galaxies \citep{ci02,ad05}. The dimensions of these
1044: structures are estimated to be $\sim$10 Mpc in transverse
1045: extent\footnote{Silverman et al. (2007) demonstrate that both
1046: structures do extend beyond the central 1 Ms region and provide new
1047: measures of their angular size.} and $\sim37$ ($z=0.67$) and $\sim28$
1048: ($z=0.73$) Mpc in depth (R. Gilli, private communication). These two
1049: structures have been shown \citep{gi03,ad05} to be varied in their
1050: dynamical state, though both have signs of ongoing collapse
1051: \citep{ad05}. The 'wall' or 'sheet' at $z=0.67$ is described as a
1052: loose structure, not yet virialized, and with an embedded compact
1053: feature, possibly an infalling group. The $z=0.73$ 'wall' is more
1054: compact with a central dense core due to the presence of a cluster
1055: \citep{ci02,gi03}, and appears to have significant surrounding
1056: substructure \citep{ad05} that may eventually collapse into a massive
1057: Virgo-type cluster in the local universe. The difference in galaxy
1058: populations within these two structures appears to reflect their
1059: dynamic state following the well-known SFR-density relation: galaxies
1060: in the $z=0.67$ structure (Fig.~\ref{color_mag}b) are mainly blue
1061: while the $z=0.73$ structure is dominated by red (i.e., evolved)
1062: galaxies.
1063:
1064: Most remarkably, the colors of the host galaxies of these
1065: moderate-luminosity AGN\footnote{It is worth noting that \citet{ge07}
1066: recently reported a color-dependency for AGN residing in denser
1067: environments with rest-frame colors ($U-B\sim0.8$) coincident with
1068: blue galaxies.}, within the redshift interval $z=$~0.63--0.76,
1069: preferentially fall in the ``green valley'' (Fig.~\ref{color_histo}b).
1070: The color distribution is asymmetric, with many hosts closer to the
1071: red sequence than the blue galaxy peak. As evident in
1072: Figure~\ref{color_mag}b, this color profile is composed of a broad
1073: distribution of galaxies residing in the $z=0.67$ redshift spike and a
1074: more compact, red ($U-V\sim0.8$) group in the $z=0.73$ redshift spike.
1075: It appears that the AGN follow a SFR-density type relation similar to
1076: the galaxies. In Figure~\ref{examples_valley}, we show example
1077: optical spectra and $HST$ images of four AGN that have $z=$~0.63--76
1078: and colors placing them in the ``green valley''. All have no evidence
1079: for an embedded AGN (e.g., H$\beta$ or Mg II emission lines) and a
1080: relatively mild 4000~\AA~break. The spectrum of source \#85 is
1081: characteristic of an E+A galaxy \citep[i.e., poststarburst;
1082: ][]{dr83,bl04} based on the absorption-line strength (EW=6.7 ${\rm
1083: \AA}$) of H$\delta$ and weak [OII] emission. E+A galaxies are thought
1084: to have undergone a starburst phase $\sim1$ Gyr ago based on the
1085: strong contribution from A stars and lack of O and B stars indicative
1086: of very recent ($\sim0.01$ Gyr) or ongoing star formation. This
1087: object is only one of fifteen, over the full redshift range, that fall
1088: within the ``green valley'' and has a spectrum characteristic of an
1089: E+A galaxy. Therefore, most of these galaxies have not had a major
1090: starburst episode within 1--2 Gyr in their past. Most spectra seem to
1091: be characteristic of a galaxy, with gradual ongoing star formation,
1092: which is best fit by the Type 2 and 3 average galaxy templates shown
1093: in Figure 3 of \citet{wo03}, effectively separating blue and red
1094: galaxies. A larger sample is required to measure statistically the
1095: fraction of AGN hosts with poststarburst signatures that have been
1096: shown to be common in AGN from the SDSS \citep{vanden06}. {\it We
1097: conclude that within this narrow redshift interval, containing two
1098: overdense structures, the enhancement of AGN activity within the ``green
1099: valley'' signifies an important link between the evolution of SMBHs
1100: and their host galaxies though not overwhelmingly in poststarburst
1101: systems.}
1102:
1103: \begin{figure*}
1104:
1105: \includegraphics[scale=0.60,angle=0]{f9.eps}
1106: \caption{Examples of AGN hosts with $0.5<U-V<1.0$ and spectroscopic
1107: redshifts placing them in the vicinity of the redshift spikes
1108: ($z\sim0.7$). The {\it HST}/ACS $V_{606}$-band images are shown with a log
1109: scaling. Optical spectra are described in the caption of
1110: Figure~\ref{examples_sf}. The source numbers follow those presented in the
1111: \citet{le05} X-ray catalog.}
1112: \label{examples_valley}
1113: \end{figure*}
1114:
1115:
1116: \subsection{Optical morphology of AGN hosts}
1117:
1118: \label{morphology}
1119:
1120: We highlight the fact that AGN hosts show signs of a bimodal
1121: distribution in their colors but not in their morphological properties
1122: \citep{bo06}. To illustrate, we have marked those AGN in
1123: Figure~\ref{color_mag}a that have a radial surface brightness profile
1124: (i.e., S\'{e}rsic index, $n$), provided by \citet{ha07},
1125: characteristic of a bulge-dominated galaxy ($2.5<n<8$). The
1126: S\'{e}rsic profile of a typical disk-dominated (e.g., late-type)
1127: galaxy is that of an exponential function ($n=1$) while that of a
1128: bulge-dominated galaxy (e.g., early-type) is the $r^{1/4}$-law (de
1129: Vaucoulers profile; $n=4$). We find that there are many
1130: bulge-dominated galaxies with bluer colors (i.e., ongoing star
1131: formation) than a typical red-sequence galaxy and there are a few
1132: disk-dominated galaxies that are red. The {\it HST}/ACS images of the
1133: red disk-dominated galaxies hosting AGN show that most (5 out of 7
1134: with $U-V>1.1$) of them are highly-inclined or edge-on spirals that
1135: probably have significant dust extinction. To investigate further, we
1136: plot in Figure~\ref{morph_color}, the S\'{e}rsic index versus
1137: rest-frame color ($U-V$) for our luminous ($M_V<-20.7$) galaxies and
1138: AGN. We further require here that the science flag provided by
1139: \citet{ha07} is 1 for all galaxies that guarantees that GALFIT
1140: \citep{pe02} converged; this condition is met for 71 of the 85
1141: luminous host galaxies with morphological parameters given in
1142: \citet{ha07}. The rest-frame color distribution of this cleaned
1143: sample is equivalent to that used in previous analyses since the
1144: removed hosts span the full range of color. Symbol types for the AGN
1145: correspond to the three redshift intervals used throughout this work.
1146: First, it is evident that the requirement for AGN host galaxies to be
1147: bulge-dominated \citep[e.g.,][]{ka03} continues up to $z\sim1$
1148: \citep{gr05} since 75\% have $n>2.5$ and the fraction\footnote{The
1149: method to determine the AGN fraction (Table~\ref{table_fraction}; Index
1150: 8-12), using our cleaned sample (science flag=1), is described in
1151: $\S$~\ref{text:agn_fraction} with $n$ replacing $U-V$.} of galaxies
1152: hosting AGN as a function of $n$ sharply rises above this value
1153: (Fig.~\ref{agn_fraction_morph}a; Table~\ref{table_fraction}). Second,
1154: AGN hosts preferentially become bluer with redshift but retain their
1155: bulge-dominated morphology. We conclude that the blue colors of AGN
1156: hosts up to $z\sim1$ are not indicative of star formation in
1157: disk-dominated systems.
1158:
1159: \begin{figure}
1160: \includegraphics[scale=0.50,angle=90]{f10.eps}
1161: \caption{Morphology--color relation of luminous ($M_V<-20.7$) galaxies
1162: with (large colored symbols) and without (small black dots) AGN. The
1163: large symbols denote the following redshift ranges: $z =$~0.4--0.63
1164: (blue circles), $z =$~0.63--0.76 (red boxes), $z =$~0.76--1.1 (green
1165: triangles). Errors (1$\sigma$) associated with $n$ are those reported
1166: in \citet{ha07}. The bar in the upper left corner is the mean
1167: $\pm1\sigma$ error on $U-V$ for the AGN host galaxies. The horizontal
1168: line divides those galaxies that have either a disk-dominated
1169: ($n<2.5$) or bulge-dominated ($n>2.5$) morphology as defined in
1170: \citet{bl03}. The vertical lines highlight the ``green valley''.}
1171: \label{morph_color}
1172: \end{figure}
1173:
1174: It is clear that the fraction of bulge-dominated galaxies hosting AGN
1175: is higher for those with blue colors ($U-V<0.7$); this is shown in the
1176: top left quadrant of Figure~\ref{morph_color}. We measure the AGN
1177: fraction (see Table~\ref{table_fraction}), as detailed in
1178: $\S$~\ref{text:agn_fraction}, and find that $21.3\pm5.0\%$ of the
1179: galaxies with $n>2.5$ and $U-V<0.7$ harbor moderate-luminosity AGN.
1180: This is over three times greater than that found for the entire sample
1181: irrespective of color or morphology ($6.4\pm0.8$\%;
1182: $\S$~\ref{text:agn_fraction}). In contrast, the AGN fraction of
1183: purely disk-dominated galaxies ($n<1.5$) is much lower
1184: ($1.0\pm0.4$\%). We conclude that blue, bulge-dominated galaxies have
1185: a relatively-high AGN fraction, possibly due to a combination of their
1186: massive bulges and gas content.
1187:
1188: \begin{figure}
1189: \includegraphics[scale=0.65,angle=90]{f11.eps}
1190: \caption{Fraction of galaxies ($M_V<-20.7$) hosting AGN as a function
1191: of their morphology (i.e., S\'{e}rsic index $n$) for the entire sample
1192: (a). In panel $b$, we have split the sample into those in (solid) and
1193: out (dashed) of the redshift interval $0.63 \leq z \leq 0.76$ as done
1194: in Figure~\ref{agn_fraction}. AGN clearly reside in bulge-dominated
1195: (Panel $a$; $n\gtrsim3$) galaxies with those in large-scale structures
1196: (Panel $b$; solid line) possibly having a higher fraction with faint
1197: disks that effectively softens their S\'{e}rsic index ($n\sim3$).}
1198: \label{agn_fraction_morph}
1199: \end{figure}
1200:
1201: We further illustrate the color-morphology of our AGN host galaxies by
1202: displaying in Figure~\ref{bulges} {\it HST}/ACS color images of six
1203: AGN host galaxies in the GOODS area with $n>2.5$ sorted by rest-frame
1204: $U-V$ color. Clearly, the most prominent feature is the bright bulge
1205: for all host galaxies. The bluest host galaxy J033213.2$-$274241,
1206: seems to be undergoing a merger. It appears that faint disks may be
1207: prevalent in AGN residing within the ``green valley''; this is evident
1208: in Figure~\ref{examples_valley} as well. Star-forming substructures
1209: may be present in the form of arcs or spiral arms in a few host
1210: galaxies (J033226.8$-$274145, J033246.4$-$275414). For comparison,
1211: the bottom right panel shows a typical red-sequence elliptical galaxy
1212: (\hbox{J033209.7$-$274248}). In Figure~\ref{disks}, we display color
1213: images of disk-dominated ($n<2.5$) AGN host galaxies. Bright,
1214: star-forming regions are ubiquitous in the form of knots that appear
1215: to lie within spiral arms in most examples.
1216:
1217: \begin{figure*}
1218: \includegraphics[scale=1.0,angle=0]{f12.eps}
1219: \caption{Color {\it HST}/ACS postage-stamp images of bulge-dominated
1220: ($n>2.5$) host galaxies located in the GOODS region. Colors correspond to ACS
1221: $B_{435}$ (blue), $V_{606}$ (green), and $z_{850}$ (red) bandpass images. In
1222: each postage-stamp image, we indicate the source name (top), the redshift and
1223: rest-frame $U-V$ color of the galaxy (lower right), and a vertical line of
1224: length $1.5\arcsec$ for scaling reference. The images have been sorted by
1225: rest-frame $U-V$ color such that the bluest source is located in the upper left
1226: panel and the reddest source is shown in the lower right panel.
1227: }
1228: \label{bulges}
1229: \end{figure*}
1230:
1231: \begin{figure*}
1232: \includegraphics[scale=0.85,angle=0]{f13.ps}
1233: \caption{Color {\it HST}/ACS postage-stamp images of disk-dominated
1234: ($n<2.5$) host galaxies located in the GEMS area. Colors correspond
1235: to ACS $V_{606}$ (blue), ($V_{606}$+$z_{850}$)/2 (green) and $z_{850}$
1236: (red) bandpass images. The labels are described in Figure~\ref{bulges}.}
1237: \label{disks}
1238: \end{figure*}
1239:
1240: We remark that a more thorough analysis of the morphological
1241: structures of these host galaxies and their relation to the general
1242: galaxy population is required and beyond the scope of this paper. We
1243: highlight the following complications that are relevant for
1244: interpreting these results and will be taken into consideration in a
1245: future analysis: (1) S\'{e}rsic fits using de Vaucoulers profiles have
1246: been shown to have significant uncertainties \citep[e.g.,][]{ha07},
1247: (2) the presence of an embedded, optically-faint AGN is likely to
1248: affect the measure of structural parameters, such as $n$, (3) accurate
1249: discrimination between bulge and disk-dominated galaxies is difficult
1250: since S\'{e}rsic indices are known to have significant overlap for
1251: these populations \citep[e.g.,][]{sa06}, and (4) one must consider
1252: redshift-dependent surface brightness effects. A fair number of AGN
1253: host galaxies have intermediate S\'{e}rsic indices ($1.5<n<3$;
1254: Figure~\ref{morph_color}) and evidence of faint disks as described
1255: above. A careful consideration of these issue will enable us to
1256: assess the dependence of host-galaxy morphology with environment
1257: (Figure~\ref{agn_fraction_morph}b}). There is suggestive evidence
1258: that S\'{e}rsic indices of hosts within large-scale structures
1259: (Figure~\ref{morph_color}; red squares) are shifted to lower values
1260: ($n\sim3$). The frequency of bulge-dominated disk galaxies among the
1261: hosts of AGN may be an important test of merger-based evolution
1262: models.
1263:
1264: \subsection{Are AGN driving the evolution of the general galaxy population?}
1265: \label{text:mergers}
1266:
1267: Numerical simulations \citep{sp05b} demonstrate that accreting SMBHs
1268: may be a necessary ingredient in the formation of massive elliptical
1269: galaxies from the mergers of gas-rich spiral galaxies. AGN feedback
1270: can potentially suppress star formation \citep{di05} and drive (i.e.,
1271: accelerate) galaxies onto the red sequence \citep{cr06,hop06a}. The
1272: short timescales ($<$ 1 Gyr) for an AGN to provide significant
1273: feedback to expel the gas may then instill the color bimodality of
1274: galaxies \citep{sp05a} since galaxies with intermediate colors are
1275: thought to be evolving rapidly. Strong observational evidence
1276: supporting this scenario has been elusive. We now explore whether our
1277: AGN sample with well determined host colors and luminosities offers
1278: further insight into the AGN-galaxy connection.
1279:
1280: The color distribution of AGN host galaxies in our study further
1281: substantiates the aformentioned model of galaxy evolution. The
1282: fraction of galaxies harboring AGN is significantly enhanced
1283: (Fig.~\ref{agn_fraction}) in the ``green valley'', a region on the
1284: color-magnitude diagram where galaxies are thought to migrate from
1285: blue to red. This higher incidence is associated with galaxies
1286: residing in a redshift interval $0.63 \leq z \leq 0.76$ dominated by
1287: two large-scale structures that are overdense and have an ongoing assembly
1288: of substructure (i.e., galaxy groups and clusters). Within this same
1289: redshift interval, the color distribution of the underlying galaxy
1290: population (Fig.~\ref{galaxy_distributions}) has rapidly evolved as
1291: shown by (1) a significant reduction in the numbers of blue galaxies,
1292: a dominant population at higher redshifts ($z>0.76$), (2) a redward
1293: shift of the blue peak, and (3) a dramatic emergence of a red sequence
1294: compared to previous epochs ($0.76<z \leq 1.1$; dotted line). This
1295: narrow redshift interval ($0.63 \leq z \leq0.76$), with an elapsed
1296: time of 0.73 Gyr, offers a compressed window of the more passive
1297: galaxy evolution that occurs over longer timescales of $\sim$5 Gyr
1298: ($0.4<z<1.1$) as illustrated in Figure~\ref{color_mag}b.
1299:
1300: In addition, numerical simulations demonstrate that the color
1301: evolution of a merger event slows upon approach to the red sequence
1302: \citep{sp05a}. In Figure~\ref{color_tracks}, we show the
1303: color-magnitude relation for galaxies and AGN within the redshift
1304: interval $0.63\leq z \leq 0.76$ with the evolutionary tracks of
1305: mergers, including SMBH feedback, with virial velocity $v_{\rm
1306: vir}$=113, 160, 226, and 320 km s$^{-1}$ overplotted \citep{sp05a}.
1307: The fact that these models do not account for dust reddening should
1308: not severely impact our subsequent findings. Here, we use SDSS
1309: photometric bands ($u$, $r$) to utilize these models and have
1310: converted the photometry to the Vega system. To account roughly for
1311: redshift evolution from $z\sim0.7$ to the present, we have shifted the
1312: model magnitude $M_r$ by $-1$ based on values given for $M^{\star}_B$
1313: in the literature \citep[see Table 5 of ][]{fa06}. The evolutionary
1314: tracks have ages up to $\sim$5.5 Gyr\footnote{It is just a coincidence
1315: and most likely of no physical significance that these large-scale
1316: structures in the CDF-S are present at $z\sim0.7$, a redshift singled out
1317: by \citet{sp05a}, as a possible formation epoch of elliptical galaxies
1318: based an elapsed time of 5.5 Gyr to complete a merger sequence.} with
1319: the first data point corresponding to 1 Gyr after the initial
1320: starburst phase. Based on these model curves, over 1 Gyr has elapsed
1321: since the starburst phase concluded for almost all AGN hosts. The
1322: majority of AGN hosts cover a broad range of age between 1--4 Gyr and
1323: virial velocity between 113 and 226 km s$^{-1}$. As shown in
1324: Figure~\ref{color_tracks}, a large fraction of the merger sequence
1325: ($t\approx$~2--4 Gyr) is spent on approach to the red sequence that
1326: should impart an asymmetry in the color distribution of the hosts of
1327: AGN. In Figure~\ref{color_histo}b, there may be signs of this since
1328: the distribution is skewed with many hosts falling near the blue edge
1329: of the red sequence, although a larger sample is required to a make a
1330: definitive claim. This effect coupled with our sample size may
1331: explain the small numbers of host galaxies with clear poststarburst
1332: ($t \lesssim$~1--2 Gyr) signatures in their optical spectra
1333: (Fig.~\ref{examples_valley}). We further speculate that the AGN
1334: hosts, situated at the entrance to the red sequence, may either redden
1335: with age or further merge with galaxies already on the red sequence
1336: \citep[i.e., `dry mergers';][]{vd05,be06} in rich environments that
1337: can effectively move them to redder colors and high
1338: luminosities/masses since there are few massive ($v_{\rm vir}>226$ km
1339: s$^{-1}$) and luminous progenitors that could populate the luminous
1340: end ($M_{r} \lesssim -22.5$) of the red sequence. We conclude that
1341: the color distribution of AGN hosts further substantiates a
1342: coevolution scenario due to mergers and interactions that are
1343: effectively nurtured in $\sim10$ Mpc scale structures.
1344:
1345: \begin{figure}
1346:
1347: \includegraphics[scale=0.55,angle=0]{f14a.eps}
1348: \includegraphics[scale=0.65,angle=0]{f14b.eps}
1349: \caption{(a) Color-magnitude relation of galaxies ($0.63 \leq z \leq 0.76$)
1350: associated with large-scale structures and AGN-influenced model
1351: evolutionary tracks from \citet{sp05a}. Galaxies hosting AGN are
1352: further marked by an open circle. Theoretical data for merging
1353: galaxies is overplotted for galaxies with virial velocities as
1354: shown. The large filled circles start from 1 Gyr after the inital
1355: starburst phase and each spacing corresponds to 0.5 Gyr. (b) $HST$
1356: $V_{606}$-band image of source \#66 ($z_{phot}=0.69$;
1357: $L_{\rm X}=1.5\times10^{42}$ erg s$^{-1}$) that is the only AGN host galaxy
1358: shown above ($M_r=-23.47; u-r=0.32$; $n=3.56\pm0.018$) with a
1359: possible merger timescale less than 1 Gyr.}
1360: \label{color_tracks}
1361: \end{figure}
1362:
1363: The role of major mergers in triggering AGN has met observational
1364: scrutiny even up to $z\sim1$, where the merger rate is expected to be
1365: higher. Morphological studies \citep{gr05,pi07} have yet to find
1366: X-ray selected AGN at $z>0.5$ in such disturbed systems. Here, we
1367: confirm the findings of \citet{na07} that hard (i.e., obscured) AGN do
1368: not populate a distinct region of the color-magnitude diagram as
1369: predicted by numerical simulations of merging galaxies that trigger a
1370: pre-quasar (optically hidden) phase \citep{hop05}. We have further
1371: marked those AGN in Figure~\ref{color_mag} with hardness ratios
1372: \hbox{$HR=(H-S)/(H+S)$} indicative of X-ray absorption ($HR>-0.2$;
1373: $\Gamma=1.9$ with \hbox{$N_{\rm H}=10^{22}$~cm$^{-2}$} at $z=0$). The
1374: hardness ratio is a measure of the relative numbers of observed X-ray
1375: counts in the soft ($S$; 0.5--2.0 keV) and hard ($H$; 2--8 keV) energy
1376: bands. We see that hard sources are present over a wide range of
1377: rest-frame colors. This is not unexpected since AGN are known to
1378: usually have parsec-scale molecular tori \citep{an93} with substantial
1379: absorbing columns that bear no relation to the presence of star
1380: formation. The lack of mergers may reflect the limited sample used to
1381: date, since a pair of galaxies undergoing a major merger at $z\sim0.7$
1382: may only be resolved by {\it HST} for less than $\sim$1 Gyr of the
1383: merger cycle \citep[see Fig.~11 of ][]{sp05b}. As is evident in
1384: Figure~\ref{color_tracks}, the galaxy samples in the \hbox{E-CDF-S}
1385: are severely limited for the region on the color-magnitude diagram
1386: likely to have merger events between massive ($v_{\rm vir}>160$ km
1387: s$^{-1}$) galaxies on a timescale of $\sim$1 Gyr after the initial
1388: starburst has ended. Three hosts out of our entire sample of 109 AGN
1389: are located on the color-magnitude diagram in a region with $v_{\rm
1390: vir}>160$ km s$^{-1}$ and $\tau < 1$ Gyr which may be indicative of a
1391: major merger. As shown in Figure~\ref{color_tracks}a, one (\#66) of
1392: them falls within our narrow redshift interval $0.63 \leq z \leq 0.67$
1393: with $M_r=-23.47$ and $u-r=0.32$. In Figure~\ref{color_tracks}$b$, we
1394: show the $HST$ $V_{606}$-band image that exemplifies a complex
1395: morphology in the nuclear region and tidal features on scales of
1396: $\sim$10 kpc (1\arcsec~=~7.1 kpc at $z=0.69$) that most likely arise
1397: from a major merger of massive galaxies. A second example
1398: (J033213.2$-$274241), shown in the top left panel of
1399: Figure~\ref{bulges}, also shows signs of interaction. We surmise that
1400: larger area {\it HST} surveys of deep extragalactic fields, such as
1401: COSMOS \citep{sc06} with an area coverage of 1.8 deg$^2$ (5.4~$\times$
1402: the area of the \hbox{E-CDF-S}) will provide improved statistics to
1403: assess adequately the role of mergers in triggering SMBH accretion.
1404:
1405: Finally, we note an alternative to the merger scenario in which
1406: galaxies may be more favorable to AGN activity simply due to the
1407: presence of a massive bulge and disk that provide the two required
1408: ingredients \citep[i.e., a SMBH and a reservoir of gas for
1409: accretion;][]{ka06}. This is possible since a high fraction
1410: ($\sim21\%$) of blue spheroids ($n>2.5$; $U-V<0.7$) host
1411: moderate-luminosity AGN and there is a slight enhancement of AGN
1412: activity in the `field' for galaxies in the ``green valley''
1413: (Fig.~\ref{agn_fraction}b; dotted line). HST images of host galaxies
1414: residing within the ``green valley'' (Fig.~\ref{examples_valley})
1415: appear to have bulges and faint disks. If this were the case,
1416: however, we would then not expect such a strong increase in the AGN
1417: fraction as a function of environment.
1418:
1419: \section{Summary}
1420:
1421: We identified a sample of 109 X-ray selected AGN in the \hbox{E-CDF-S}
1422: with moderate luminosities ($41.9\leq \log~L_{0.5-8.0~{\rm keV}}\leq
1423: 43.7$) to investigate the rest-frame colors of their host galaxies.
1424: These AGN have been selected from a parent sample of 5,549 galaxies
1425: from COMBO-17 and GEMS with $0.4 \leq z \leq 1.1$. Optical spectra
1426: are available for 48\% of the sample; these provide assurance of the
1427: accuracy of the photometric redshifts and show that no strong AGN
1428: signatures are present, confirming the stellar nature of their
1429: rest-frame colors.
1430:
1431: We find that the broad distribution of host-galaxy colors of
1432: moderate-luminosity AGN is due to both (1) the strong color evolution
1433: of the underlying luminous ($M_V<-20.7$) and bulge-dominated ($n>2.5$)
1434: galaxy population and (2) an enhancement of AGN activity in
1435: large-scale structures. We draw the following three main points:
1436:
1437: \begin{itemize}
1438:
1439: \item The host galaxies of X-ray selected AGN have a color bimodality
1440: when excluding a redshift interval $0.63 \leq z \leq 0.76$, which contains two
1441: redshift spikes at $z=0.67$ and $z=0.73$. Galaxies hosting AGN at $z
1442: \lesssim 0.6$ are preferentially red with many falling along the red
1443: sequence. At $z\ga0.8$, a distinct, blue population of host galaxies
1444: is prevalent with colors similar to the star-forming galaxies.
1445:
1446: \item The fraction of galaxies hosting AGN has a prominent peak in the
1447: ``green valley'' that is primarily due to enhanced AGN activity in
1448: large-scale structures. Within the redshift interval $0.63 \leq z
1449: \leq 0.76$, the AGN fraction reaches a peak value of $\sim15\%$ at
1450: $U-V\sim0.8$. Over the color interval $0.5<U-V<1.0$ we find the AGN
1451: fraction to be $12.8\pm2.9\%$, significantly ($2.6\sigma$) higher than
1452: that measured in the `field' (i.e., over all other redshifts;
1453: $7.8\pm2.5\%$).
1454:
1455: \item We find that AGN continue to preferentially reside in luminous
1456: bulges up to $z\sim1$. A large fraction (75\%) of AGN host galaxies
1457: with $M_V<-20.7$ have S\'{e}rsic indices $n>2.5$, even those at
1458: $z\sim1$ that primarily have blue rest-frame colors. Blue,
1459: bulge-dominated galaxies ($n>2.5$, $U-V<0.7$) are most hospitable for
1460: AGN activity based on a measured AGN fraction of $21.3\pm5.0\%$ (four
1461: times the fraction of the `field' sample).
1462:
1463: \end{itemize}
1464:
1465: The overabundance of AGN associated with the redshift spikes found in
1466: the \hbox{E-CDF-S} and their special location in the color-magnitude
1467: relation highlight the importance of environment, on large scales
1468: ($\sim10$ Mpc), to influence the evolution of AGN and their host
1469: galaxies. The richness of these structures (i.e., galaxy overdensities,
1470: group/clusters in early stages of formation) alludes to mergers as a
1471: dominant mechanism to trigger AGN activity, quench star formation and
1472: to drive subsequent migration of galaxies from the blue cloud to the
1473: red sequence. We compare the color-magnitude relation of our AGN host
1474: galaxies with evolutionary tracks of merging galaxies from
1475: \citet{sp05a} that incorporate AGN feedback. Our AGN host galaxies
1476: have colors and morphologies (i.e., bulge-dominated) indicative of
1477: evolved systems that had undergone a starburst phase $\approx$ 1--4
1478: Gyr, and a possible major merger before being observed. These
1479: timescales qualitatively agree with optical spectra that do not show
1480: overwhelming starburst signatures. These timescales also naturally
1481: explain why so few X-ray selected AGN host galaxies appear to be
1482: undergoing major mergers. Larger area $HST$ imaging surveys such as
1483: COSMOS will statistically sample these rare events with timescales
1484: less than 1 Gyr.
1485:
1486: The \hbox{E-CDF-S} is a unique survey field with a fortunate alignment
1487: of large-scale structures for such studies. Our findings exemplify
1488: the complexities that must be disentangled to determine underlying
1489: relationships between AGN and their host galaxies. Much optical and
1490: near-infrared followup is forthcoming in the \hbox{E-CDF-S} that will
1491: further our understanding of the connection between coevolution of AGN
1492: and galaxies.
1493:
1494: \acknowledgments
1495:
1496: We are especially grateful to the referee for providing insightful
1497: comments that strengthened the overall content of this work. We also
1498: thank L.~Guzzo, K. Iwasawa, A. Merloni, K. Nandra, and I. Strateva for
1499: helpful discussions and suggestions. We also recognize the
1500: contribution of the GEMS group that provided an early version of their
1501: catalog, V. Springel for supplying the model galaxy tracks, and
1502: C.~Wolf for computing updated photometry. Support for this work was
1503: provided by NASA through $Chandra$ Award Number G04-5157A (BDL, WNB,
1504: DPS). RG, CV and PT acknowledge partial support by the Italian Space
1505: agency under the contract ASI--INAF I/023/05/0.
1506:
1507: "Some of the data presented in this paper were obtained from the
1508: Multimission Archive at the Space Telescope Science Institute
1509: (MAST). STScI is operated by the Association of Universities for
1510: Research in Astronomy, Inc., under NASA contract NAS5-26555. Support
1511: for MAST for non-{\it HST} data is provided by the NASA Office of Space
1512: Science via grant NAG5-7584 and by other grants and contracts."
1513:
1514: %% To help institutions obtain information on the effectiveness of their
1515: %% telescopes, the AAS Journals has created a group of keywords for telescope
1516: %% facilities. A common set of keywords will make these types of searches
1517: %% significantly easier and more accurate. In addition, they will also be
1518: %% useful in linking papers together which utilize the same telescopes
1519: %% within the framework of the National Virtual Observatory.
1520: %% See the AASTeX Web site at http://www.journals.uchicago.edu/AAS/AASTeX
1521: %% for information on obtaining the facility keywords.
1522:
1523: %% After the acknowledgments section, use the following syntax and the
1524: %% \facility{} macro to list the keywords of facilities used in the research
1525: %% for the paper. Each keyword will be checked against the master list during
1526: %% copy editing. Individual instruments can be provided in parentheses,
1527: %% after the keyword, but they will not be verified.
1528:
1529: Facilities: \facility{VLT(VIMOS)}.
1530:
1531:
1532: \begin{thebibliography}{}
1533: \bibitem[Adami et al.(2005)]{ad05} Adami, C. et al. 2005, A\&A, 805, 818
1534: \bibitem[Alexander et al.(2005)]{al05} Alexander, D. M., Smail, I., Bauer, F. E., Chapman, S. C., Blain, A. W., Brandt, W. N., Ivison, R. J. 2005, \nat, 434, 7034
1535: \bibitem[Antonucci(1993)]{an93} Antonucci, R. 1993, ARA\&A, 31, 473
1536: \bibitem[Barger et al.(2003)]{ba03} Barger, A.J. et al. 2003, AJ, 126,632
1537: \bibitem[Bahcall et al.(1997)]{ba97} Bahcall, J. N., Kirhakos, S., Saxe, D. H. 1997, \apj, 479, 642
1538: \bibitem[Balogh et al.(2004)]{ba04} Balogh, M. L., Baldry, I. K., Nichol, R., Miller, C., Bower, R., Glazebrook, K. 2004, \apj, 615, 101
1539: \bibitem[Baldry et al.(2006)]{ba06} Baldry, I. K., Balogh, M. L., Bower, R., Glazebrook, K., Nichol, R. C., Bamford, S. P., Budavari, T. 2006, \mnras, 373, 469
1540: \bibitem[Bauer et al.(2002)]{ba02} Bauer, F.E., Alexander, D.M., Brandt, W.N., Hornschemeier, A.E., Vignali, C., Garmire, G., Schneider, D. P. 2002, \aj, 124, 2351
1541: \bibitem[Beckwith et al.(2006)]{beck06} Beckwith, S.~V.~W., et al. 2006, \aj, 132, 1729
1542: \bibitem[Bell et al.(2004)]{be04} Bell, E. F., et al. 2004, \apj, 608, 752
1543: \bibitem[Bell et al.(2005)]{bell05} Bell, E. F., et al. 2005, \apj, 625, 23
1544: \bibitem[Bell et al.(2006)]{be06} Bell, E. F., et al. 2006, \apj, 640, 241
1545: \bibitem[Bell et al.(2007)]{be07} Bell, E. F., et al. 2007, \apj, 663, 834
1546: \bibitem[Best et al.(2005)]{be05} Best, P. N., Kauffmann, G., Heckman, T. M., Brinchmann, J., Charlot, S., Ivezic, Z., White, S. D. M. 2005, \mnras, 362, 25
1547: \bibitem[B\"{o}hm et al.(2006)]{bo06} B\"{o}hm, A., Wisotzki and the GEMS team 2006, ASP conference Series, astro-ph/0612310
1548: \bibitem[Boyle \& Terlevich(1998)]{bo98} Boyle, B. J. \& Terlevich, R. J. 1998, MNRAS, 293, L49
1549: \bibitem[Blake et al.(2004)]{bl04} Blake, C. et al. 2004, MNRAS, 355, 713
1550: \bibitem[Blanton et al.(2003)]{bl03} Blanton, M. R., et al. 2003, \apj, 594, 186
1551: \bibitem[Brandt \& Hasinger(2005)]{br05} Brandt, W. N., Hasinger, G. 2005, ARA\&A, 43, 827
1552: \bibitem[Cassata et al.(2006)]{ca07} Cassata, P. et al. 2007, \apjs, in press, astro-ph/0701483
1553: \bibitem[Caldwell et al.(2005)]{ca05} Caldwell, J. A. R. et al. 2005, submitted to ApJS, astro-ph/0510782
1554: \bibitem[Cimatti et al.(2002)]{ci02} Cimatti, A. et al. A\&A, 392, 395
1555: \bibitem[Croton et al.(2006)]{cr06} Croton, D. J. et al. 2006, MNRAS, 365, 11
1556: \bibitem[Di Matteo et al.(2005)]{di05} Di Matteo, T., Springel, V., Hernquist, L. 2005, \nat, 433, 604
1557: \bibitem[Dressler \& Gunn(1983)]{dr83} Dressler, A., Gunn, J. E. 1983, \apj, 270, 7
1558: \bibitem[Eastman et al.(2007)]{east07} Eastman, J., Martini, P., Sivakoff, G., Kelson, D. D., Mulchaey, J., Tran, K.-V. 2007, \apj, in press, arXiV:0706.0209
1559: \bibitem[Faber et al.(2006)]{fa06} Faber, S. M. et al. 2006, \apj, in press, astro-ph/0506044
1560: \bibitem[Ferrarese \& Merritt(2000)]{fe00} Ferrarese, L., Merritt, D. 2000, \apj, 539, L9
1561: \bibitem[Franceschini et al.(1999)]{fr99} Franceschini, A., Hasinger, G., Miyaji, T., Malquori, D. 1999, MNRAS, 310, L5
1562: \bibitem[Gavignaud et al.(2006)]{ga06} Gavignaud, I. et al. 2006, A\&A, 457, 79
1563: \bibitem[Gebhardt et al.(2000)]{ge00} Gebhardt, K. et al. 2000, \apj, 539, L13
1564: \bibitem[Georgakakis et al.(2007)]{ge07} Georgakakis, A. et al. 2006, \apj, 660, L15
1565: \bibitem[Giacconi et al.(2002)]{gi02} Giacconi, R., et al. 2002, \apjs, 139, 369
1566: \bibitem[Giavalisco et al.(2004)]{gi04} Giavalisco,~M., et al. 2004, ApJ, 600, L93
1567: \bibitem[Gilli et al.(2003)]{gi03} Gilli, R. et al. 2003, \apj, 592, 721
1568: %%\bibitem[Gilli et al.(2005)]{gi05} Gilli, R. et al. 2005, A\&A, 430, 811
1569: \bibitem[Grogin et al.(2005)]{gr05} Grogin, N. A. et al. 2005, \apj, 627, 97
1570: \bibitem[H\"aussler et al.(2007)]{ha07} H\"aussler, B. et al. 2007, \apjs, in press, arXiv:0704.2601V2
1571: \bibitem[Hopkins \& Beacom(2006)]{ho06} Hopkins, A. M., Beacom, J. F. 2006, \apj, 651, 142
1572: \bibitem[Hopkins et al.(2005)]{hop05} Hopkins, P. F. et al. 2005, \apj, 630, 705
1573: \bibitem[Hopkins et al.(2006a)]{hop06a} Hopkins, P. F. et al. 2006, \apjs, 163, 1
1574: \bibitem[Hopkins et al.(2006b)]{hop06b} Hopkins, P. F., Somerville, R. S., Hernquist, L., Robertson, B., Li, Y. 2006, \apj, 652, 864
1575: \bibitem[Jahnke et al.(2004a)]{ja04a} Jahnke, K., Kuhlbrodt, B., Wisotzki, L., \mnras, 352, 399
1576: \bibitem[Jahnke et al.(2004b)]{ja04b} Jahnke, K. et al. 2004b, \apj, 614, 568
1577: \bibitem[Kartaltepe et al.(2007)]{kart07} Kartaltepe, J. S. et al. 2007, \apjs, in press, arXiv:0705.2266
1578: \bibitem[Kauffmann et al.(2003)]{ka03} Kauffmann, G. et al. 2003, \mnras, 346, 1055
1579: \bibitem[Kauffmann et al.(2004)]{ka04} Kauffmann, G. et al. 2004, \mnras, 353, 713
1580: \bibitem[Kauffmann et al.(2006)]{ka06} Kauffmann, G. et al. 2007, \mnras, in press, astro-ph/0609436
1581: \bibitem[Li et al.(2006)]{li06} Li, C., Kauffmann, G., Wang, L., White, S. D. M., Heckman, T., Jing, Y. P. 2006, MNRAS, 373, 457
1582: \bibitem[Lehmer et al.(2005)]{le05} Lehmer, B. D. et al. 2005, ApJS, 161, 21
1583: \bibitem[Lehmer et al.(2007)]{le07} Lehmer, B. D. et al. 2007, \apj, 657, 681
1584: \bibitem[Le Fevre et al.(2003)]{lef03} Le Fevre, O., Vettolani, G., Maccagni, D., and the VVDS consortium, 2003, The Messenger 111, 18
1585: \bibitem[Le Fevre et al.(2004)]{lef04} Le Fevre, O., et al. 2004, A\&A, 428, 1043
1586: \bibitem[Martini et al.(2006)]{ma06} Martini, P., Kelson, D. D., Kim, E., Mulchaey, J., Athey, A. 2006, \apj, 644, 116
1587: \bibitem[Martini et al.(2007)]{ma07} Martini, Mulchaey, J., Kelson, D. D. 2007, \apj, 664, 761
1588: \bibitem[McIntosh et al.(2005)]{mc05} McIntosh, D.H. et al. 2005, \apj, 632, 191
1589: \bibitem[McLure et al.(2004)]{mc04} McLure, R. J. et al. 2004, \mnras, 351, 347
1590: \bibitem[Merloni et al.(2004)]{me04} Merloni, A., Rudnick, G., Di Matteo, T. \mnras, 354, 37
1591: \bibitem[Mignoli et al.(2005)]{mi05} Mignoli, M. et al. 2005, A\&A, 437, 883
1592: \bibitem[Miller et al.(2003)]{mi03} Miller, C. J., Nichol, R. C., Gomez, P.L., Hopkins, A. M., Bernardi, M 2003, \apj, 597, 142
1593: \bibitem[Nandra et al.(2007)]{na07} Nandra, K. et al. 2006, \apj, 660, L11
1594: \bibitem[Noeske et al.(2007)]{no07} Noeske, K. G. et al. 2007, \apj, 660, L43
1595: \bibitem[Peng et al.(2002)]{pe02} Peng, C.Y., Ho, L.C., Impey, C. D., Rix, H.-W. 2002, \aj, 124, 266
1596: \bibitem[Pierce et al.(2007)]{pi07} Pierce, C. et al. 2006, \apj, 660, L19
1597: \bibitem[Press et al.(1993)]{pr93} Press, W. H., Teukolsky, S. A., Vetterling, W. T., \& Flannery, B. P. 1993, Numerical
1598: Recipes in C: The Art of Scientific Computing, 2nd edn.(Cambridge: Cambridge University Press)
1599: \bibitem[Rix et al.(2004)]{ri04} Rix, H. -W. et al 2004, ApJS, 152, 163
1600: \bibitem[Rovilos et al.(2007)]{ro07} Rovilos, E., Georgakakis, Georgantopoulos, I., Koekemoer, A., Mobasher, B., Goudis, C. 2007, A\&A, 466, 119
1601: \bibitem[Sanchez et al.(2004)]{sa04} Sanchez, S. F. et al. 2004, \apj, 614, 586
1602: \bibitem[Sanders \& Mirabel(1996)]{sa96} Sanders, D. B., Mirabel, I.F. 1996, ARA\&A, 34, 749
1603: \bibitem[Sargent et al.(2006)]{sa06} Sargent, M~T. et al. 2006, \apjs, in press, astro-ph/0609042
1604: \bibitem[Scoville et al.(2006)]{sc06} Scoville, N. et al. \apjs, in press, arXiv:asto-ph/0612306
1605: \bibitem[Silverman et al.(2005)]{si05} Silverman, J. D., et al. 2005, \apj, 618, 123
1606: \bibitem[Springel et al.(2005a)]{sp05a} Springel, V., Di Matteo, T., Hernquist, L. 2005a, ApJ, 620, L79
1607: \bibitem[Springel et al.(2005b)]{sp05b} Springel, V., Di Matteo, T., Hernquist, L. 2005b, \mnras, 339, 289
1608: \bibitem[Strateva et al.(2001)]{st01} Strateva, I. et al. 2001, \aj, 122, 1861
1609: \bibitem[Sturm et al.(2006)]{st06} Sturm, E. et al. 2006, \apj, 642, 81
1610: \bibitem[Szokoly et al.(2004)]{sz04} Szokoly, G. P., et al. 2004, \apjs, 155, 271
1611: \bibitem[Tremaine et al.(2002)]{tr02} Tremaine, S. et al. 2002, \apj, 574, 740
1612: \bibitem[Vanden Berk et al.(2006)]{vanden06} Vanden Berk, D.~E. et al. 2006, \aj, 131,84
1613: \bibitem[Vanzella et al.(2005)]{va05} Vanzella, E. et al. 2005, A\&A, 434, 53
1614: \bibitem[Vanzella et al.(2006)]{va06} Vanzella, E. et al. 2006, A\&A, 454, 423
1615: \bibitem[van Dokkum et al.(2005)]{vd05} van Dokkum, P. G. 2005, \apj, 130, 2647
1616: \bibitem[Wisotzki et al.(2006)]{wi06} Wisotzki, L., Jahnke, K., Sanchez, S. F., Schramm, M., 2006, NewAR, 50, 829
1617: \bibitem[Wolf et al.(2003)]{wo03} Wolf, C. et al. 2003, A\&A, 401, 73
1618: \bibitem[Wolf et al.(2004)]{wo04} Wolf, C. et al. 2004, A\&A, 421, 913
1619: \bibitem[Zakamska et al.(2006)]{za06} Zakamska, N. et al. 2006, \aj, 132, 1496
1620: \bibitem[Zheng et al.(2007)]{zh07} Zheng, X. Z., et al. 2007, \apj, 661, L41
1621: \end{thebibliography}
1622:
1623: \clearpage
1624:
1625: %% Use the figure environment and \plotone or \plottwo to include
1626: %% figures and captions in your electronic submission.
1627: %% To embed the sample graphics in
1628: %% the file, uncomment the \plotone, \plottwo, and
1629: %% \includegraphics commands
1630: %%
1631: %% If you need a layout that cannot be achieved with \plotone or
1632: %% \plottwo, you can invoke the graphicx package directly with the
1633: %% \includegraphics command or use \plotfiddle. For more information,
1634: %% please see the tutorial on "Using Electronic Art with AASTeX" in the
1635: %% documentation section at the AASTeX Web site,
1636: %% http://www.journals.uchicago.edu/AAS/AASTeX.
1637: %%
1638: %% The examples below also include sample markup for submission of
1639: %% supplemental electronic materials. As always, be sure to check
1640: %% the instructions to authors for the journal you are submitting to
1641: %% for specific submissions guidelines as they vary from
1642: %% journal to journal.
1643:
1644:
1645: %% This example uses \plotone to include an EPS file scaled to
1646: %% 80% of its natural size with \epsscale. Its caption
1647: %% has been written to indicate that additional figure parts will be
1648: %% available in the electronic journal.
1649:
1650: \end{document}
1651: