0801.1843/ms.tex
1: %\documentclass[12pt,preprint]{aastex}
2: \documentclass{emulateapj}
3: 
4: \input{macros_vdbosch}
5: 
6: %\slugcomment{Submitted to the Astrophysical Journal Letters}
7: \begin{document}
8: 
9: 
10: \title{New Constraints on the Efficiencies of Ram-Pressure Stripping\\
11: and the Tidal Disruption of Satellite Galaxies}
12: 
13: \author{X. Kang\altaffilmark{1}, Frank C.  van den Bosch\altaffilmark{1}}
14: 
15: \altaffiltext{1}{Max-Planck-Institute for  Astronomy, K\"onigstuhl 17,
16: D-69117, Heidelberg, Germany}
17: %\email{kang,vdbosch@mpia.de} 
18: 
19: 
20: \begin{abstract}
21:   Using data from the Sloan  Digital Sky Survey (SDSS) it has recently
22:   been  shown that the  red fraction  of satellite  galaxies increases
23:   with  stellar  mass. Semi-analytical  models,  however, predict  red
24:   satellite fractions  that are independent of stellar  mass, and much
25:   higher than observed. It has  been argued that this discrepancy owes
26:   to  the fact  that the  models  assume that  satellite galaxies  are
27:   instantaneously stripped  of their hot gas reservoirs  at the moment
28:   they are accreted  into a bigger halo.  In this  letter we show that
29:   the fraction  of red satellites  can be brought in  better agreement
30:   with  the  data by  simply  decreasing  this  stripping efficiency.  
31:   However, this  also results  in a red  fraction of  massive centrals
32:   that  is much  too low.   This  owes to  the fact  that the  massive
33:   centrals  now accrete  satellite galaxies  that are  bluer  and more
34:   gas-rich.  However, if a  significant fraction of low mass satellite
35:   galaxies is tidally disrupted before being accreted by their central
36:   host galaxy,  as suggested by  recent studies, the red  fractions of
37:   both centrals  and satellites can be reproduced  reasonably well.  A
38:   problem remains  with the red  fraction of centrals  of intermediate
39:   mass, which is likely to  reflect an oversimplified treatment of AGN
40:   feedback.
41: \end{abstract}
42: \keywords{galaxies:  general --  galaxies: formation  -- intergalactic
43: medium}
44: 
45: 
46: \section{Introduction}
47: \label{sec:intro}
48: 
49: Recently it has become   clear that semi-analytical models  for galaxy
50: formation   predict red satellite   fractions that are   much too high
51: (Weinmann \etal  2006; Baldry \etal  2006).  It  has been argued  that
52: this owes to  an oversimplified  treatment  of ram-pressure  stripping
53: (Gunn \&  Gott 1972) of  the hot gaseous  halos of satellite galaxies.
54: In all  semi-analytical models it is  assumed that the  entire hot gas
55: reservoir  of  a galaxy is  instantaneously stripped  the moment it is
56: accreted into a larger halo  (i.e., the moment  it becomes a satellite
57: galaxy).
58: 
59: Although  this was  originally motivated  by  the fact  that the  vast
60: majority  of  (faint)  satellite  galaxies  in clusters  are  red  and
61: deficient  in cold  gas  (e.g., Binggeli,  Tammann  \& Sandage  1987),
62: detailed analytical and hydrodynamical simulations have shown that the
63: typical  time  scale  for  this stripping  process  (sometimes  called
64: `strangulation') ranges  from $\sim 1$  Gyr to 10 Gyr  (e.g., Balsara,
65: Livio  \& O'Dea 1994;  Mori \&  Burkert 2000;  Bekki, Couch  \& Shioya
66: 2002; Mayer \etal 2006; McCarthy \etal 2008).
67: 
68: In this letter we study the impact of a prolonged strangulation on the
69: colors of satellite and central  galaxies as function of stellar mass.
70: To that extent we implement a new recipe for ram pressure stripping of
71: hot gaseous halos of satellite galaxies in a semi-analytical model for
72: galaxy formation, and    compare the  red  fractions of    central and
73: satellite galaxies  to   data   obtained from  a  SDSS    galaxy group
74: catalogue.
75: 
76: \section{Semi-Analytical Model}
77: \label{sec:sam}
78: 
79: In order to model the colors of central  and satellite galaxies we use
80: the semi-analytical model of galaxy formation of Kang \etal (2005) and
81: Kang, Jing \&  Silk (2006; hereafter K06),  and we refer the reader to
82: these two papers for  details.  Briefly, dark  matter merger trees are
83: extracted  from a high  resolution $N$-body  simulation (with subhalos
84: resolved, Jing \&  Suto 2002) and populated  with galaxies following a
85: semi-analytical description.  Galaxies    are assumed to form  at  the
86: centers  of their dark matter  halos, using  recipes for cooling, star
87: formation, super-nova feedback, radio-mode  AGN feedback, and spheroid
88: formation that are  all very similar  to those of Croton \etal (2006).
89: When  a  dark matter halo   with its  associated `central'  galaxy  is
90: accreted  by  a larger halo, and  thus  becomes a subhalo, its central
91: galaxy becomes  a   `satellite'  galaxy.  The  orbital    evolution of
92: satellite galaxies  is tracked  using  the  orbital evolution of   its
93: corresponding subhalo, up to the point where  the subhalo is dissolved
94: in the simulation.  From  that  point on   the satellite galaxy   (now
95: called an `orphan  galaxy') is merged  with the central galaxy in  the
96: host halo after a dynamical friction  time computed using the standard
97: Chandrasekhar prescription.  Throughout this  paper, we adopt the same
98: model parameters as   in K06,  for  which  the  model reproduces   the
99: present-day   luminosity   function,  and   yields    a bimodal  color
100: distribution in  reasonable agreement with the data  (see below).  The
101: only changes we make here regard the treatment of the stripping of hot
102: gas around satellite galaxies.
103: %
104: \begin{figure}
105: \begin{center}
106: \includegraphics[width=1.0\columnwidth,height=0.75\columnwidth]{f1.eps}
107: \caption{The  distribution   of  model     galaxies as  function    of
108:   $^{0.1}(g-r)$ color and  the `observed' stellar  mass, obtained from
109:   the relation between stellar mass-to-light  ratio and color of  Bell
110:   \etal (2003).  The dashed line corresponds to  $^{0.1}(g-r) = 0.81 +
111:   0.15(\log[M_{\ast}/(h^{-2}M_{\odot})]-10)$ and  is used to split the
112:   model galaxies in red and blue populations (cf. Fig.7 in vdB07).}
113: \label{fig:cm}
114: \end{center}
115: \end{figure} 
116: 
117: In the standard model it is assumed that the hot gas associated with a
118: halo is instantaneously stripped as soon  as the halo is accreted by a
119: larger  halo  (i.e., as  soon  as  it  becomes a  subhalo).   However,
120: hydrodynamic simulations  indicate that the  ram-pressure stripping of
121: the hot gas  is not instantaneous, and may even  not be complete after
122: as much  as 10  Gyr.  Although the  characteristic time scale  of this
123: stripping process varies from one  simulation to the other, and ranges
124: from $\sim 1$  Gyr to 10 Gyr (see  \S\ref{sec:intro}), all simulations
125: suggest that the stripping  rate is approximately constant.  Motivated
126: by these findings  we model the rate at which  the subhalo is stripped
127: of its hot gas as
128: %
129: \begin{equation}\label{rate}
130: {{\rm d}M_{\rm hot} \over {\rm d}t} = -(1-f_{\rm hot})
131: {M_{{\rm hot},0} \over \tau}\,.
132: \end{equation}
133: %
134: Here $M_{{\rm hot},0}$ is the  hot halo gas mass  of the satellite  at
135: accretion, $\tau$ is the characteristic time  scale for stripping, and
136: $f_{\rm hot}$ is the fraction of hot gas left after one characteristic
137: time scale.  In addition to being stripped, the hot gas reservoir of a
138: satellite galaxy also decreases  due to cooling.   In order to compute
139: the cooling rate of the hot gas associated with a satellite galaxy, we
140: assume that the cooling radius and cooling  time of the hot gas remain
141: fixed at their values  at the time  of accretion.  We also assume that
142: the hot gas  remains associated with  the orphan galaxy even after its
143: subhalo is dissolved in the N-body simulation (we will discuss this in
144: \S\ref{sec:res}).  Finally, we   ignore any  potential  dependency  of
145: $f_{\rm hot}$ or $\tau$ on the satellite mass, the host halo mass, the
146: orbit  of  the subhalo,  or the  density  distribution of the  hot gas
147: associated    with the host  halo.  These    assumptions are certainly
148: oversimplified.  However,  the main aim of this  paper is  not to give
149: the  most accurate  description of the  rate at  which the hot  gas of
150: satellite galaxies is stripped, but merely  to investigate how changes
151: in $\tau$ and  $f_{\rm  hot}$ impact on the    colors of central   and
152: satellite galaxies.
153: 
154: \section{results}
155: \label{sec:res}
156: 
157: In a  recent paper, van den  Bosch \etal (2007;  hereafter vdB07) used
158: the SDSS galaxy group catalogue  of Yang \etal (2007) to determine the
159: red fractions of  central and satellite galaxies as  function of their
160: stellar masses.  Here  central galaxies have been defined  as the most
161: massive  group  members (in  terms  of  their  stellar masses),  while
162: satellite galaxies are all group members that are not centrals.  Using
163: the fact  that the  color-stellar mass relation  of SDSS  galaxies is
164: clearly bimodal, vdB07 split the  galaxies in blue and red populations
165: using the stellar-mass dependent cut,
166: %
167: \begin{equation}\label{colcut}
168: ^{0.1}(g-r) = 0.76 + 0.15 \left( \log[M_{\ast}/(h^{-2}M_{\odot})]-10\right)\,.
169: \end{equation}
170: %
171: Here  $^{0.1}(g-r)$ is  the  $(g-r)$ color  $K$-corrected to  redshift
172: $z=0.1$,  and  $M_{\ast}$  is  the  stellar mass  computed  using  the
173: relation between  stellar mass-to-light ratio and color  of Bell \etal
174: (2003; eq.~[2] in Yang \etal 2007).
175: %
176: 
177: 
178: For our model galaxies, we compute the  $^{0.1}(g-r)$ colors using the
179: stellar  population synthesis code of Bruzual  \& Charlot (2003) for a
180: Salpeter IMF\footnote{We  have  verified   that our results   are  not
181:   sensitive   to  the choice  of  the  IMF.}.  In  order  to model the
182: photometric errors  in  the SDSS, we add  a  random, Gaussian error of
183: 0.05 magn  to both the $^{0.1}g$  and $^{0.1}r$-band magnitudes of our
184: model  galaxies.  Finally,   for   each model galaxy   we  compute the
185: `observed'  stellar  mass  using  the  same  relation  between stellar
186: mass-to-light ratio and color as  used for the data. The color-stellar
187: mass relation thus obtained for  our fiducial model  (with $\tau = 0$,
188: corresponding to  instantaneous stripping of the hot  gas) is shown in
189: Fig.~\ref{fig:cm}. It clearly shows a bimodal color distribution, with
190: a red sequence  that extends to   more massive galaxies  than the blue
191: sequence. This owes mainly to our inclusion  of AGN feedback (see also
192: Croton  \etal 2006 and Bower \etal  2006), and is in good, qualitative
193: agreement with the data (cf. Fig.~7 in vdB07). Upon closer inspection,
194: though, we find that the colors of the red sequence and the bimodality
195: scale are $\sim 0.05$ magn redder than for the  SDSS. It is unclear at
196: present what the origin of this color offset is. In order to limit its
197: impact on our  analysis, we split our model  galaxies in red  and blue
198: populations using  eq.~(\ref{colcut}) but  with  a zero-point  that is
199: $0.05$ magn redder (indicated by the dashed line in Fig.~\ref{fig:cm})
200: %
201: \begin{figure*}
202: \begin{center}
203:   \includegraphics[scale=0.9]{f2.eps}
204:   \caption{The fraction   of  red   centrals  (left-hand    panel) and
205:     satellites (right-hand panel) as  function of stellar mass.   Open
206:     symbols correspond to the results of vdB07, obtained from the SDSS
207:     group  catalogue   of  Yang \etal   (2007), while  different lines
208:     correspond to the results obtained  from our semi-analytical model
209:     for  different values   of the   characteristic   time  scale  for
210:     strangulation, as  indicated.  Errorbars   are obtained  using the
211:     jackknife  technique and,  for  clarity,  are  only shown  for the
212:     fiducial model with instantaneous stripping (i.e, $\tau=0$)}
213: \label{fig:fred}
214: \end{center}
215: \end{figure*}
216: %
217: 
218: Fig.\ref{fig:fred}  compares  the  red  fractions, $f_{\rm  red}$,  of
219: centrals  (left-hand  panel)  and  satellites  (right-hand  panel)  as
220: function  of  `observed'  stellar  mass obtained  from  our  fiducial,
221: instantaneous stripping  model (solid line) with the  results of vdB07
222: (open symbols). Note that this model predicts a red satellite fraction
223: that is $\sim 0.85$, almost  independent of stellar mass, and in clear
224: disagreement with the data.   As demonstrated in Weinmann \etal (2006)
225: and  Baldry \etal (2006)  the semi-analytical  models of  Croton \etal
226: (2006)  and  Bower  \etal  (2006),  which  also  assume  instantaneous
227: stripping of  the hot gas, suffer  from exactly the  same problem.  As
228: for the centrals, the model  roughly predicts the correct red fraction
229: at  the  massive  end,  which  is  mainly  due  to  the  inclusion  of
230: radio-model AGN  feedback.  However, for centrals  with $M_{\ast} \lta
231: 10^{11}  \Msun$, the red  fraction predicted  by the  model is  a poor
232: match to the data. At the low mass end ($M_{\ast} < 10^{10} \Msun$),
233: it is upwards and inconsistent  with the data. These low-mass galaxies
234: are  too  red because  the  cold gas  are  consumed  rapidly in  their
235: earlier,  low-mass progenitors.   Observational work  (Kennicutt 1998)
236: shows that stars form efficiently only in galaxies with surface density
237: higher than a threshold, so  star formation should be more inefficient
238: in low-mass progenitors,  but this process  is ignored in  our model.
239: Another disagreement  is seen at $M_{\ast} \sim  10^{10} \Msun$, where
240: model galaxies are  too blue. We will discuss  this in \S\ref{sec:concl}
241: and in this  section we focus on the model  predictions at the massive
242: end ($M_{\ast} \sim 10^{11} \Msun$).
243: 
244: 
245: As  discussed   in  \S\ref{sec:intro},  the   over-prediction  of  red
246: satellites  could  be due  to  the  inaccurate  treatment of  hot  gas
247: stripping  from satellite  galaxies.  To  test this,  we  now consider
248: models with  less efficient stripping. Seen  from eq.(\ref{rate}) that
249: this can be  obtained by increasing $\tau$ or  $f_{\rm hot}$, and both
250: are found to  have similar trends on model  predictions.  Here we show
251: only  results by  changing $\tau$.   The  dotted and  dashed lines  in
252: Fig.~\ref{fig:fred} show  the results obtained  using eq.~(\ref{rate})
253: with $f_{\rm  hot}=0.3$ and $\tau =  3$ Gyr and  10 Gyr, respectively.
254: The effects  of a prolonged stripping  are twofold.  First  of all, as
255: expected,  it reduces the  red fraction  of satellite  galaxies.  Note
256: that  there is relatively  little difference  between the  models with
257: $\tau =  3$ Gyr and 10 Gyr,  and that neither model  can reproduce the
258: detailed stellar mass dependence of $f_{\rm red}$ observed.  Secondly,
259: both models predict  a much lower red fraction  among the most massive
260: central  galaxies,  ruining  the  good  agreement  achieved  with  the
261: fiducial model.  This  owes to the fact that  the central galaxies now
262: `accrete'  more blue  satellite galaxies  with significant  amounts of
263: fresh cold  gas which triggers a  starburst in the  central galaxy. In
264: principle,  we  could improve  the  fit of  the  red  fraction of  the
265: satellites at the  low mass end by making  $f_{\rm hot}$ and/or $\tau$
266: depend on  the stellar  mass of the  satellite galaxy, but  this would
267: only worsen the red  fraction of the (massive) centrals.  Furthermore,
268: since  the massive centrals  are now  too blue  (on average),  the red
269: fraction of  the massive satellites  (which were centrals  before they
270: were accreted)  is now also  too low.  Clearly, simply  prolonging the
271: stripping of the  hot gas of satellite galaxies  creates more problems
272: than it solves.
273: 
274: 
275: %
276: \begin{figure*}
277: \begin{center}
278: %\includegraphics[scale=0.95]{newf3.eps}
279: \includegraphics[width=1.8\columnwidth,height=1.15\columnwidth]{f3.eps}
280: \caption{Same as   Fig.~\ref{fig:fred}, but showing   the results  for
281:   models in  which we change certain properties  of satellite galaxies
282:   once they  become orphans (i.e., once their  associated subhalos are
283:   dissolved  in  the N-body  simulation).   In  the  upper panels  the
284:   reservoirs  of hot  gas  (solid lines)  and  both hot  and cold  gas
285:   (dashed  lines)  are  removed  once the  satellite  galaxies  become
286:   orphans.  In the  lower panels it is assumed that  50 percent of all
287:   satellite galaxies with $M_{\ast}  < 10^{10}\Msun$ (solid lines) and
288:   with $M_{\ast} >  0$ (dashed lines) are tidally  disrupted once they
289:   become orphans.  All  models shown here model the  hot gas stripping
290:   of  satellite galaxies using  eq.~(\ref{rate}) with  $\tau=3$Gyr and
291:   $f_{\rm hot}=0.3$.}
292: \label{fig:fred1}
293: \end{center}
294: \end{figure*}
295: 
296: 
297: How can  we  prevent the   massive galaxies  from   becoming too blue?
298: Clearly,  we need  to  prevent  the  (massive) central galaxies   from
299: accreting orphan galaxies that are too  blue and that have significant
300: amounts of cold  gas. As a first   attempt we assume that  a satellite
301: galaxy is completely stripped of its hot gas once it becomes an orphan
302: galaxy (i.e., when its subhalo in the numerical simulation dissolves).
303: The resulting   red fractions are shown   as solid lines  in the upper
304: panels of Fig.\ref{fig:fred1}. A comparison  with the dotted lines  in
305: Fig.~\ref{fig:fred}    shows that this  has  little  impact on the red
306: fractions.  If we assume, however, that the satellite is also stripped
307: of its cold gas once it becomes an orphan  we obtain the results shown
308: by  the dashed     lines in Fig.\ref{fig:fred1}.     Although in  good
309: agreement  with the data  at the  massive  end, we don't consider this
310: model very physical.
311: 
312: Recently, a number  of studies have argued that  the lack of evolution
313: in  the massive  end  of the  galaxy  stellar mass  function, and  the
314: presence of a significant  amount of intra-cluster light, suggest that
315: a significant  fraction of  satellite galaxies is  (tidally) disrupted
316: before it has the chance of  being accreted by its central host galaxy
317: (e.g.  Mihos et al.  2005; Zibetti  et al.  2005; Monaco et al.  2006;
318: Conroy,   Wechsler   \&  Kravtsov   2007).    The   lower  panels   of
319: Fig.\ref{fig:fred1} show the red  fractions obtained if we assume that
320: a  certain fraction of  satellite galaxies  is tidally  disrupted once
321: they become orphans. The solid lines correspond to a model in which 50
322: percent of all  orphans with a stellar mass  $M_{\ast} < 10^{10}\Msun$
323: are tidally disrupted.  This model yields red satellite fractions that
324: are  in  good agreement  with  the  data,  without causing  a  serious
325: under-prediction  of  the  red  fraction  of  massive  centrals.   For
326: comparison, the  dashed lines correspond to  a model in  which 50\% of
327: {\it all} orphan galaxies  are tidally disrupted, independent of their
328: stellar  mass.  This  causes  a drastic  under-prediction  of the  red
329: fractions of massive centrals and  satellites, which owes to fact that
330: the  number  of major  mergers  experienced  by  a central  galaxy  is
331: effectively reduced  by a factor  two.  Consequently, the  majority of
332: massive  centrals  are  now  disk  dominated and  the  associated  AGN
333: feedback (assumed to be proportional  to the mass of the spheroid, see
334: K06) is not sufficiently efficient to suppress the cooling of hot gas.
335: 
336: \section{Conclusions}
337: \label{sec:concl}
338: 
339: We have used a semi-analytical  model of galaxy formation to study the
340: fractions of red centrals and satellites as functions of their stellar
341: mass.  In  agreement with Weinmann  et al.  (2006)  and Baldry et  al. 
342: (2007) we find that the  models severely over-predict the red fraction
343: of satellite  galaxies.  This owes  to an oversimplified  treatment of
344: ram-pressure  stripping  of  the  hot gas  associated  with  satellite
345: galaxies.  In  particular, all semi-analytical models  assume that the
346: hot gas reservoir  of a galaxy is instantaneously  stripped the moment
347: it  is accreted  into a  bigger halo  (i.e., the  moment it  becomes a
348: satellite  galaxy).  We have  demonstrated that  the fractions  of red
349: satellites can  be brought in better  agreement with the  data, if the
350: efficiency  of  this  stripping  process  is  significantly  reduced.  
351: However, the implication is that massive centrals also become bluer to
352: the extent that  their red fraction is much too low.  This owes to the
353: fact that the satellite galaxies that are accreted by the centrals are
354: now bluer and more gas rich.  This problem can be prevented if roughly
355: half  of  all satellite  galaxies  with  a  stellar mass  $M_{\ast}  <
356: 10^{10}\Msun$  are tidally disrupted  prior to  being accreted  by the
357: central host galaxy. This is in qualitative agreement with the lack of
358: evolution in the massive end  of the galaxy stellar mass function, and
359: with the presence of a significant amount of intra-cluster light.
360: 
361: 
362: All models presented  here predict a red fraction  of central galaxies
363: with $M_{\ast}  \sim 10^{10} \Msun$ that is  significantly too low.
364: As shown by Baldry et al.  (2006), the semi-analytical model of Croton
365: et al.  (2006) suffers from  exactly the same problem, while the model
366: of Bower et al. (2006) seems  to be in better agreement with the data.
367: The  main difference  between the  model of  Bower et  al.   and those
368: presented here and in Croton et al. is the way in which the efficiency
369: of AGN feedback  scales with galaxy properties. In  fact, for galaxies
370: with $M_{\ast}  \sim 10^{10} \Msun$, QSO activities  are commonly seen
371: across cosmic  time, and it  has been shown that QSO wind can  also quench
372: their star  formation (e.g., Di  Matteo,  Springel  \&
373: Hernquist 2005).  As already eluded  to in Weinmann et al.  (2006), it
374: thus seems that  the red fractions of central  galaxies as function of
375: their stellar  mass yield important  constraints on the  efficiency of
376: AGN  feedback  (or  any  other  mechanism that  can  quench  the  star
377: formation  in  central galaxies,  e.g.,  Birnboim,  Dekel \&  Neistein
378: 2007). Although  a study of  radio AGN or  QSO feedback is  beyond the
379: scope  of this  paper, we  emphasize that  a modification  of  the red
380: fraction of central galaxies will  also modify that of the satellites.
381: After all,  the former are the  direct progenitors of  the latter.  In
382: particular,  a model  that  would  predict a  higher  red fraction  of
383: centrals with $M_{\ast} \sim  10^{10}M_{\odot}$ requires an even lower
384: efficiency of ram pressure stripping of the hot gas of satellites than
385: obtained here.
386: 
387: 
388: 
389: 
390: \begin{thebibliography}{}
391: 
392: \bibitem[Baldry et al. 2006]{Bal06}
393: Baldry, I.K., Balogh, M.L., Bower, R.G., Glazebrook, K., Nichol, R.C.,
394: Bamford, S.P., \& Budavari T., 2006, MNRAS, 373, 469
395: 
396: \bibitem[Balsara et al. 1994]{Bal94}
397: Balsara, D., Livio, M., \& O$^{,}$Dea, C.P., 1994, ApJ, 437, 83
398: 
399: \bibitem[Bekki et al. 2002]{Be02}
400: Bekki, K., Couch, W.J., \& Shioya, Y., 2002, ApJ, 577, 651
401: 
402: \bibitem[Bell et al. 2003]{Be03}
403: Bell, E.F., McIntosh, D.H., Katz, N.,\& Weinberg, M.D., 
404: 2003, ApJS, 149, 289
405: 
406: \bibitem[Binggeli et al. 1987]{Bin87}
407: Binggeli, B., Tammann, G.A., \& Sandage, A., 1987, AJ, 94, 251
408: 
409: \bibitem[Birnboim et al. 2007]{Bi07}
410: Birnboim, Y., Dekel, A.,\& Neistein, E., 2007, MNRAS, 380, 339
411: 
412: \bibitem[Bower et al. 2006]{Bow06} 
413: Bower, R.G., Benson, A.J., Maldon, R.,  Helly, J.C.,  Frenk, C.S.,  
414: Baugh, C.M.,  Cole, S.,  \& Lacey, C.G.,2006, MNRAS, 370, 645
415: 
416: \bibitem[Bruzual et al. 2003]{BC03}
417: Bruzual, G., \& Charlot, S., 2003, MNRAS, 344, 1000
418: 
419: \bibitem[Conroy et al. 2007]{Co07}
420: Conroy, C., Wechsler, R.H., \& Kravtsov, A.V., 2007, ApJ, 668, 826
421: 
422: \bibitem[Croton et al. 2006]{cr06}
423: Croton, D.J., et al., 2006, MNRAS, 365, 11
424: 
425: \bibitem[Di Matteo et al. 2005]{Di05}
426: Di Matteo, T., Springel, V., \& Hernquist, L., 2005, Nature, 433, 604
427: 
428: 
429: \bibitem[Gunn \& Gott 1972]{GG72}
430: Gunn, J.E., \& Gott, J.R.I., 1972, ApJ, 176, 1
431: 
432: \bibitem[Jing et al. 2002]{Jing02}
433: Jing, Y.P., \& Suto, Y., 2002, ApJ, 574, 538
434: 
435: \bibitem[Kang et al. 2005]{kang05}
436: Kang, X., Jing, Y.P., Mo, H.J, \& B\"orner, G., 2005, ApJ, 631, 21
437: 
438: \bibitem[Kang et al. 2006]{kang06}
439: Kang, X., Jing, Y.P., \& Silk, J., 2006, ApJ, 648, 820 (K06)
440: 
441: \bibitem[Kennicutt 1998]{Ken98}
442: Kennictt, R. C., 1998, ApJ, 498, 541
443: 
444: \bibitem[Mayer  et  al.   2006]{My06} 
445: Mayer,  L.,  Mastropietro,  C., Wadsley, J., Stadel, J., \& 
446: Moore, B., 2006, MNRAS, 369, 1021
447: 
448: \bibitem[McCarthy  et  al. 2007]{Mc07} 
449: McCarthy,  I.G., Frenk,  C.S., Font, A.S., Lacey, C.G.,  Bower, R.G., 
450: Mitchell, N.L., Balogh, M.L.,  \& Theuns, T., 2008, MNRAS, 383, 593
451: 
452: \bibitem[Mihos et al. 2005]{Mi05} 
453: Mihos, J.C., Harding, P., Feldmeier, J.,\& Morrison, H., 
454: 2005, ApJ, 631L, 41
455:  
456: \bibitem[Monaco et al. 2006]{Mo06}
457: Monaco, P., Murante, G., Borgani, S., \& Fontanot, F., 
458: 2006, ApJ, 652L, 89
459: 
460: \bibitem[Mori \& Burkert 2000]{MB00}
461: Mori, M., \& Burkert, A., 2000, ApJ, 538, 559
462: 
463: \bibitem[Springel et al. 2001]{Springel01}
464: Springel, V., White, S.D.M., Tormen, G., \& Kauffmann, G., 
465: 2001, MNRAS, 328, 726
466: 
467: \bibitem[Yang et al. 2007]{Y07}
468: Yang, X., Mo, H.J., van den Bosch, F.C., Pasquali, A., Li,C., 
469: \& Barden, M., 2007, ApJ, 671, 153
470:  
471: \bibitem[van den Bosch et al. 2007]{v07}
472: van den Bosch, F.C., et al., preprint, (arXiv:0710.3164) (vdB07)
473: 
474: \bibitem[Weinmann  et al.  2006]{W06} 
475: Weinmann,  S.M., van  den Bosch, F.C., Yang, X.,  Mo, H.J., 
476: Croton, D.J., \& Moore,  B., 2006, MNRAS, 372, 1161
477: 
478: \bibitem[Zibetti et al. 2005]{Zi05}
479: Zibetti, S., White, S.D.M., Schneider, D.P., \& Brinkmann, J., 
480: 2005, MNRAS, 358, 949
481: 
482: \end{thebibliography}
483: 
484: \end{document}
485: 
486: