1: %%

2: %% accepted, 02/03/00

3: %%

4: 

5: \documentstyle[emulateapj]{article}

6: %\documentstyle[aas2pp4]{article}

7: %%\documentclass[preprint2]{aastex}

8: 

9: \slugcomment{Accepted for publication in ApJ Letters}

10: \lefthead{Burke et al.}

11: \righthead{BCG Evolution and Cluster Selection}

12: 

13: %%

14: %% Personal definitions

15: %%

16: 

17: \newcommand{\ho}[1]{{\rm H$_0 = #1$~km\,\persec\,\permpc}}

18: 

19: %%

20: %% Sample info

21: %%

22: \newcommand{\lx}{\mbox{$L_{x}$}}

23: 

24: \newcommand{\lmineds}{2.3}

25: 

26: \newcommand{\nclus}{76}   

27: \newcommand{\abscal}{0.05}

28: 

29: \newcommand{\ltsim}{\lesssim}

30: \newcommand{\gtsim}{\gtrsim}

31: 

32: \newcommand{\zf}{\mbox{$z_f$}}

33: 

34: \newcommand{\persec}{\mbox{s$^{-1}$}}

35: \newcommand{\permpc}{\mbox{Mpc$^{-1}$}}

36: \newcommand{\lumin}{\mbox{erg \persec}}

37: 

38: %% English/US spelling

39: \newcommand{\centre}{center}

40: \newcommand{\colour}{color}

41: \newcommand{\favoured}{favored}

42: \newcommand{\behaviour}{behavior}

43: \newcommand{\programmes}{programs}

44: 

45: %% hyphenation

46: \newcommand{\semianalytic}{semi-analytic}

47: \newcommand{\semianalytical}{semi-analytical}

48: \newcommand{\xray}{X-ray}

49: 

50: %% punctuation

51: \newcommand{\eg}{\mbox{e.g.}}

52: \newcommand{\etal}{\mbox{et al.}}

53: 

54: %% papers

55: \newcommand{\cmfirst}{Collins \& Mann~\markcite{cm98}(1998; hereafter CM98)}

56: \newcommand{\pcmmain}{\protect\markcite{cm98}CM98}

57: \newcommand{\cmmain}{\markcite{cm98}CM98}

58: 

59: \newcommand{\abkfirst}{Arag\'{o}n-Salamanca, Baugh \& Kauffmann \markcite{abk98}(1998; hereafter ABK98)}

60: \newcommand{\pabkmain}{\protect\markcite{abk98}ABK98}

61: \newcommand{\abkmain}{\markcite{abk98}ABK98}

62: 

63: \newcommand{\deprop}{De Propris \etal~\markcite{depropris99}(1999)}

64: \newcommand{\depropbracket}{De Propris \etal~\markcite{depropris99}1999}

65: 

66: \newcommand{\lynam}{Lynam \etal~\markcite{pdl99}(1999)}

67: 

68: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

69: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

70: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

71: 

72: \begin{document}

73: 

74: \title{Cluster Selection and the Evolution of Brightest Cluster Galaxies}

75: 

76: \author{

77: D. J. Burke\altaffilmark{1,2,5},

78: C. A. Collins\altaffilmark{1,5},

79: R. G. Mann\altaffilmark{3,4,5}}

80: 

81: \altaffiltext{1}{Astrophysics Research Institute, Liverpool John Moores University,

82: 12 Quays House, Egerton Wharf, Birkenhead, CH41 1LD, UK}

83: \altaffiltext{2}{Institute for Astronomy, University of Hawaii,

84: 2680 Woodlawn Drive, Honolulu, HI 96822}

85: \altaffiltext{3}{Astrophysics Group, Blackett Laboratory, 

86: Imperial College, Prince Consort Road, London, SW7 2AZ, UK}

87: \altaffiltext{4}{Institute for Astronomy, University of Edinburgh, 

88: Royal Observatory, Blackford Hill, Edinburgh, EH9 3NJ, UK}

89: \altaffiltext{5}{Visiting Astronomer at the 

90: NASA Infrared Telescope

91: Facility, which is operated by the University of

92: Hawaii under contract from the National Aeronautics

93: and Space Administration.}

94: %%\email{burke@ifa.hawaii.edu,cac@astro.livjm.ac.uk}

95: %%\email{rgm@roe.ac.uk}

96: 

97: %% Abstract

98: \begin{abstract}

99: The K-band Hubble diagram of Brightest Cluster Galaxies (BCGs) is presented

100: for a large, \xray\ selected cluster sample extending out to $z = 0.8$.

101: The controversy over the degree of BCG evolution is shown to be due to

102: sample selection,

103: since the BCG luminosity depends upon the cluster environment.

104: Selecting only the most \xray\ luminous clusters produces a BCG sample 

105: which 

106: shows, under the assumption of an Einstein-de Sitter cosmology, 

107: significantly less mass growth than that predicted by current 

108: \semianalytic\ galaxy formation models, and significant evidence of 

109: any growth only if the

110: dominant stellar population of the BCGs formed relatively recently ($z \leq 2.6$).

111: 

112: \end{abstract}

113: 

114: \keywords{Galaxies: clusters: general --- galaxies: elliptical and lenticular, 

115: cD --- galaxies: evolution --- galaxies: formation}

116: 

117: \section{INTRODUCTION}

118: \label{intro}

119: 

120: The majority of stars in giant ellipticals 

121: found in the cores of rich galaxy clusters are old;

122: photometric and spectroscopic studies of cluster galaxies out to

123: $z \approx 1$ suggest a formation redshift, \zf, greater than 2,

124: with little variation within a cluster, 

125: and that secondary bursts of star formation account for

126: a small fraction of the stellar mass

127: (\eg\ 

128: Arag\'{o}n-Salamanca \etal~\markcite{aecc93}1993;

129: Ellis \etal~\markcite{ellis97}1997;

130: Stanford, Eisenhardt \& Dickinson~\markcite{sed98}1998;

131: van Dokkum \etal~\markcite{vd98}1998;

132: Poggianti \etal~\markcite{morphs99}1999).

133: 

134: However, to understand the process of galaxy formation it is necessary to

135: know where the stars were formed as well as when.

136: In the traditional view of early-type galaxy formation---a

137: ``monolithic'' collapse at high redshift 

138: (\eg\ Eggen, Lynden-Bell \& Sandage~\markcite{els62}1962;

139: Larson~\markcite{larson69}1969)---all the stars

140: were formed in situ, in direct contrast to the merger-driven

141: growth of galaxies predicted by 

142: \semianalytical\ models for hierarchical cosmologies, such as CDM

143: (\eg\ Kauffmann \& White~\markcite{kw93}1993;

144: Baugh, Cole \& Frenk~\markcite{bcf96}1996).

145: Since the ages of the stars are similar in both scenarios,

146: it is the change in mass with look-back time that

147: separates the two 

148: pictures

149: observationally.

150: The current data are inconclusive;

151: for example,

152: the hierarchical models are \favoured\ by the enhanced

153: merger fraction seen in the $z=0.8$ cluster MS1054.4-0321

154: (van Dokkum \etal~\markcite{v99}1999), whilst

155: \deprop\ show 

156: no evidence for mass evolution of bright ellipticals in clusters

157: out to $z \approx 1$.

158: In general it is difficult to follow the evolutionary history of 

159: ellipticals since selection methods can seriously bias the samples,

160: c.f. the discussions of progenitor bias in van Dokkum \& Franx~\markcite{vf96}(1996),

161: the effect of preferential selection of the most massive objects at

162: each epoch in Kauffmann \& Charlot~\markcite{kc98}(1998),

163: and the use of \colour\ selection in Jimenez \etal~\markcite{jimenez99}(1999).

164: 

165: One approach to minimising such problems is to study the evolution of 

166: a particular class of ellipticals---brightest cluster galaxies (BCGs)---because

167: of their unique location, close to the \centre\ of the cluster's

168: potential well.

169: BCGs do not appear to be drawn from the same luminosity function

170: as other cluster galaxies (\eg\ Dressler~\markcite{dressler78}1978), 

171: which suggests that they have a distinct formation history.

172: Knowledge of BCG evolution can therefore provide different

173: constraints on galaxy formation models to studies

174: of the general cluster population.

175: 

176: The K-band Hubble diagram for BCGs has recently been extended

177: to $z \approx 1$ by 

178: both \cmfirst\ and 

179: \abkfirst: these

180: observations provide the opportunity to measure the luminosity

181: evolution of BCGs since evolutionary and pass-band 

182: corrections are insensitive to the recent star-formation history of

183: a galaxy at near-IR wavelengths (\eg\ Bershady~\markcite{bershady95}1995;

184: Madau, Pozzetti, \& Dickinson~\markcite{madau98}1998).

185: The conclusions drawn are contradictory,

186: despite the use of the same cosmology and a common assumption that

187: the stellar populations of BCGs are old and passively evolving;

188: \cmmain\ assert that the stellar populations of BCGs in the most massive clusters

189: have not grown significantly since $z \approx 1$,

190: whilst \abkmain\ argue that their results are in good agreement with 

191: the mass increase of BCGs---by a factor of four in an

192: Einstein-de Sitter cosmology---predicted by 

193: \semianalytical\ 

194: models over the same redshift range.

195: The two samples have almost no overlap---\cmmain\ having used

196: an \xray\ selected cluster catalogue whilst

197: \abkmain\ used a heterogeneous compilation that was mainly optically selected---and

198: it is the aim of the present work to show that the results

199: can be reconciled by considering the properties of the 

200: clusters in the two samples.

201: Section~\ref{data} describes the BCG sample used---

202: an extension of that of \cmmain---and

203: the reduction methods employed, whilst

204: section~\ref{results} presents the results of the analysis

205: and a comparison to those of \abkmain.

206: Throughout this letter

207: an Einstein-de Sitter cosmology with \ho{50} is assumed,

208: and \xray\ luminosities (\lx) are quoted for the 0.3--3.5~keV pass band.

209: 

210: \section{DATA}

211: \label{data}

212: 

213: \subsection{Sample}

214: \label{data:sample}

215: 

216: The data presented here comprise K-band observations of \nclus\ BCGs.

217: This sample, which incorporates that of \cmmain,

218: spans a redshift range of 0.05 to 0.83, and is drawn from the following

219: \xray\ selected cluster catalogues:

220: the Einstein EMSS (Gioia \& Luppino \markcite{gl94}1994; 

221: Nichol \etal\ \markcite{n97}1997; Henry \markcite{h99}1999),

222: the Southern and Bright SHARC catalogues 

223: (Burke \etal\ \markcite{b97}1997; 

224: Romer \etal\ \markcite{bsharc}2000),

225: and the ROSAT NEP Survey (Henry \etal\ \markcite{nep97}1997).

226: \xray\ selection is to be preferred, since both \xray\ luminosity and 

227: \xray\ temperature should be more closely related to cluster

228: mass than optical richness.

229: 

230: The \xray\ luminosity-redshift coverage of the cluster sample is shown

231: by the circles in Figure~1. 

232: The additional symbols show those clusters from \abkmain\ with a 

233: measured \xray\ flux or upper limit,

234: except for Cl~2155+0334 (also known as Cl~2157+0347), which

235: has been removed because photometric and

236: spectroscopic observations show no evidence for a cluster

237: (Thimm \& Belloni~\markcite{tb94}1994; Oke, Postman \& 

238: Lubin~\markcite{opl98}1998), and Cl~0016+16, since it is in both samples.

239: The difference in \xray\ luminosity coverage at $z > 0.5$

240: for the two samples is striking;

241: the implications of this are discussed in section~\ref{results}.

242: 

243: \vspace{2mm}

244: \begin{center}

245: %

246: %%\plotfiddle{fig1.eps}{2in}{-90}{32}{32}{-120}{180}

247: \plotfiddle{fig1.eps}{2in}{-90}{32}{32}{-120}{190}

248: %

249: \begin{minipage}{8.75cm}

250: \small\parindent=3.5mm

251: {\sc Fig.}~1.---Cluster \xray\ luminosity as a function of redshift.

252: The circles indicate the BCG sample presented here; 

253: open for clusters used 

254: in \pcmmain\ and filled for the new observations.

255: Clusters from \pabkmain\ with measured \xray\ fluxes are shown

256: as plus ($+$) symbols and the arrow ($\uparrow$) symbol 

257: represents the $3\sigma$ upper limit for Cl~1603+4329.

258: The dashed line, at $\lx = \lmineds \times 10^{44}$~\lumin, 

259: shows the luminosity used by

260: \pcmmain\ to separate their sample into high- and low-\lx\ clusters.

261: %

262: \par

263: \end{minipage}

264: %

265: \end{center}

266: \vspace{3mm}

267: 

268: The K-band BCG observations were made using the IRCAM3 and UFTI cameras on the

269: UKIRT and NSFCAM on the IRTF, with some of the

270: data being provided by the service \programmes\ of both telescopes.

271: IRCAM3 and NSFCAM are 256x256 pixel InSb devices with a field of

272: view close to 70\arcsec\ by 70\arcsec\ (the IRCAM3 and NSFCAM

273: pixel scales are 0.281 and 0.3~\arcsec\,pixel$^{-1}$

274: respectively), and UFTI is a 1024x1024 pixel HgCdTe array with

275: a pixel scale of 0.091~\arcsec\,pixel$^{-1}$,

276: giving a field of view of 92\arcsec\ by 92\arcsec.

277: The observing strategy is the same as presented in \cmmain:

278: the BCGs were imaged using a jitter pattern and

279: separate sky exposures were taken 

280: for those objects which filled the field of view.

281: 

282: \subsection{Reduction}

283: \label{data:reduction}

284: 

285: The data reduction system improves upon that presented in 

286: \cmmain, and incorporates elements from the methods described

287: in Stanford, Eisenhardt \& Dickinson~\markcite{sed95}(1995) 

288: and Hall, Green \& Owen~\markcite{hall98}(1998).

289: An outline is presented below as the method 

290: will be fully described in a later paper.

291: 

292: The individual frames were masked for bad pixels, dark subtracted

293: and divided by the exposure time.

294: A flat field image was created by median combination of the 

295: object images---or separate sky exposures if these were available---and

296: applied to the object frames. 

297: Masking of cosmic ray events was performed 

298: on the flattened images before they were mosaiced together,

299: which completed the processing of those objects with sky exposures.

300: Otherwise the mosaic---which is substantially deeper than 

301: the individual exposures---was used to create an object mask,

302: which was then applied to the individual images before they were

303: median-combined to form a flat.

304: The flattened exposures were then processed as above to create the

305: final image.

306: 

307: Observations of stars from the UKIRT faint standards 

308: list (Casali \& Hawarden~\markcite{ukirt-fs}1992)

309: were used to calibrate the photometry onto the UKIRT system

310: assuming an extinction of 0.088 mag airmass$^{-1}$, 

311: the median value for K-band observations at Mauna Kea.

312: Comparisons of the results from repeat observations, both

313: within and between observing runs, show that the magnitudes

314: agree to \abscal\ mag.

315: 

316: Aperture magnitudes were measured using a 50~kpc diameter aperture

317: and have been corrected for Galactic absorption using the

318: maps of Schlegel, Finkbeiner, \& Davis~\markcite{schlegel98}(1998):

319: the correction is small, mostly being less than 0.05~mag, but

320: reaching 0.1~mag in several cases.

321: The position of the aperture was chosen so as to maximise the 

322: flux contained within it whilst remaining close to the \centre\ of

323: the cluster \xray\ emission.

324: Those pixels contaminated by stars and obvious non-cluster

325: galaxies were excluded from the calculation, being replaced by values 

326: chosen from regions at the same distance from the aperture \centre.

327: No attempt has been made to remove flux due to other cluster galaxies

328: falling within the aperture, and so the results are

329: directly comparable to those of \abkmain.

330: 

331: \section{RESULTS}

332: \label{results}

333: 

334: The K-band Hubble diagram for the two BCG samples is shown in 

335: Figure~2. The lines show model predictions

336: calculated using the GISSEL96 code (Bruzual \& Charlot \markcite{bc}1993),

337: for a solar-metallicity stellar population with a Salpeter initial mass function:

338: the solid line indicates a no-evolution model for a 10~Gyr old

339: stellar population, whereas the other lines are for stellar populations 

340: which form in an instantaneous burst of star formation at a single 

341: epoch---$\zf = 2$ for the dashed line and $\zf = 5$ for the dotted line---and then 

342: evolve passively.

343: The models have been normalised to match the low-redshift, X-ray selected, 

344: BCG sample of \lynam, following the method used in \abkmain, 

345: assuming a growth curve, $d\,\log{L}$/$d\,\log{r}$, of 0.7

346: for the aperture corrections and a \colour\ of $R-K = 2.6$.

347: 

348: %%\begin{figure*}[b]

349: %%\plotone{fig2.eps}

350: %%\caption[fig2.eps]{

351: %%Magnitude-redshift relation for brightest cluster galaxies

352: %%in the observed K band.

353: %%Filled and open symbols represent those 

354: %%BCGs in high- and low-\lx\ clusters respectively; the

355: %%division is as in Figure~\ref{fig:z-lx}.

356: %%The circles in the top panel indicate the sample presented here,

357: %%whilst the squares are for the BCGs in the

358: %%two $z=1.3$ clusters from Rosati \etal~\protect\markcite{rosati99}(1999),

359: %%where the magnitudes have been measured within 50~kpc diameter apertures

360: %%(P. Rosati~1999, private communication).

361: %%The bottom panel shows the sample of \pabkmain, where 

362: %%the crosses are for those clusters without a measured \xray\ flux.

363: %%The no-evolution prediction, assuming a 10~Gyr old

364: %%stellar population, is shown by the solid line;

365: %%passive-evolution models, in which the stars form at a single

366: %%epoch, are shown as dashed ($\zf = 2$)

367: %%and dotted ($\zf = 5$) lines.

368: %%\label{fig:z-mk}}

369: %%\end{figure*}

370: 

371: \vspace{2mm}

372: \begin{center}

373: %

374: %\plotone{fig2.eps}

375: \plotfiddle{fig2.eps}{4.2in}{0}{42}{42}{-120}{-10}

376: %

377: \begin{minipage}{8.75cm}

378: \small\parindent=3.5mm

379: {\sc Fig.}~2.---Magnitude-redshift relation for brightest cluster galaxies

380: in the observed K band.

381: Filled and open symbols represent those 

382: BCGs in high- and low-\lx\ clusters respectively; the

383: division is as in Figure~1.

384: The circles in the top panel indicate the sample presented here,

385: whilst the squares are for the BCGs in the

386: two $z=1.3$ clusters from Rosati \etal~\protect\markcite{rosati99}(1999),

387: where the magnitudes have been measured within 50~kpc diameter apertures

388: (P. Rosati~1999, private communication).

389: The bottom panel shows the sample of \pabkmain, where 

390: the crosses are for those clusters without a measured \xray\ flux.

391: The no-evolution prediction, assuming a 10~Gyr old

392: stellar population, is shown by the solid line;

393: passive-evolution models, in which the stars form at a single

394: epoch, are shown as dashed ($\zf = 2$)

395: and dotted ($\zf = 5$) lines.

396: %

397: \par

398: \end{minipage}

399: %

400: \end{center}

401: \vspace{3mm}

402: 

403: The result remains qualitatively the same as Figure~6 of \cmmain; BCGs

404: in high-\lx\ clusters form a homogeneous population which is

405: brighter, and has a smaller scatter, than that of low-\lx\ clusters.

406: This can be more clearly seen in Figure~3,

407: which shows the scatter around the model predictions

408: as a function of cluster \xray\ luminosity.

409: It is this relationship between BCG and cluster properties that leads

410: to the contradictory conclusions of \cmmain\ and \abkmain:

411: out of the eleven $z > 0.5$ clusters in the latter sample, nine have

412: \xray\ flux measurements or upper limits, with all but two of these having

413: a low \xray\ luminosity (Figure~1).

414: It is unsurprising that these clusters are not similar

415: to rich, local clusters, as they

416: were discovered on the basis of their optical properties

417: (\eg\ Castander \etal~\markcite{castander94}1994;

418: Holden \etal~\markcite{holden97}1997).

419: The squares in Figure~2 represent the 

420: BCGs in the two $z = 1.3$ clusters discussed by

421: Rosati \etal~\markcite{rosati99}(1999): the high-\lx\ cluster (solid square)

422: was discovered by means of its \xray\ emission,

423: whereas the low-\lx\ cluster (open square) was detected by its galaxy population. 

424: Although based on only two points, this suggests that the correlation 

425: with environment holds at this redshift. 

426: 

427: %%\begin{figure*}[b]

428: %%\plotone{fig3.eps}

429: %%\caption[fig3.eps]{

430: %%Residuals about the model predictions, defined as 

431: %%$\Delta m_k = m_{\rm BCG} - m_{\rm model}$, as a function of 

432: %%cluster \xray\ luminosity.

433: %%The BCGs presented here are shown as circles, the

434: %%sample of \pabkmain\ is shown as in Figure~\ref{fig:z-lx},

435: %%and the squares represent the two clusters from 

436: %%Rosati \etal~\protect\markcite{rosati99}(1999).

437: %%The three panels are for the models shown in

438: %%Figure~\ref{fig:z-mk}:

439: %%a) no-evolution model for a 10~Gyr old stellar population,

440: %%b) formation at a redshift of 2 followed by passive evolution,

441: %%and 

442: %%c) as for b) but with a formation redshift of 5.

443: %%\label{fig:lx-dmk}}

444: %%\end{figure*}

445: 

446: The \semianalytic\ models discussed in \abkmain\ predict a factor of 

447: $\sim $4--5 increase in the stellar masses of BCGs in

448: massive clusters since $z=1$, for an Einstein-de Sitter universe.

449: To test whether the data presented here supports this level of evolution, 

450: a correlation between redshift and the BCG residuals

451: ($\Delta m_k$, \eg\ Figure~4)

452: has been sought.

453: Passive-evolution models with $\zf=2$ and $\zf=5$ have been used to

454: calculate the residuals---since they provide a conservative range for the 

455: formation epoch of massive cluster ellipticals 

456: (\eg\ Ellis \etal~\markcite{ellis97}1997)---and

457: separate fits made to the high- and low-\lx\ cluster subsamples.

458: Since the form of any evolution is unknown a priori, a non-parametric

459: rank-order statistic---Kendall's $\tau$---was used; it also has the advantage that

460: it is insensitive to the choice of normalisation adopted for the Bruzual

461: \& Charlot models.

462: All save one of the fits showed no significant ($>3\sigma$) evidence for

463: evolution; the exception, at a significance of $3.6\sigma$, was the

464: high-\lx\ subsample with $\zf=2$.

465: To find the maximum formation epoch that is still compatible with 

466: evolution of the high-\lx\ subsample, \zf\ was increased 

467: from 2 until the correlation significance dropped below $3\sigma$.

468: Evolution is found only if the stars formed recently ($\zf \leq 2.6$).

469: 

470: \vspace{2mm}

471: \begin{center}

472: %

473: %%\plotone{fig3.eps}

474: %\plotfiddle{fig3.eps}{4.5in}{0}{45}{45}{-130}{-10}

475: \plotfiddle{fig3.eps}{4.1in}{0}{42}{42}{-120}{-10}

476: %

477: \begin{minipage}{8.75cm}

478: \small\parindent=3.5mm

479: {\sc Fig.}~3.---Residuals about the model predictions, defined as 

480: $\Delta m_k = m_{\rm BCG} - m_{\rm model}$, as a function of 

481: cluster \xray\ luminosity.

482: The BCGs presented here are shown as circles, the

483: sample of \pabkmain\ is shown as in Figure~1,

484: and the squares represent the two clusters from 

485: Rosati \etal~\protect\markcite{rosati99}(1999).

486: The three panels are for the models shown in

487: Figure~2:

488: a) no-evolution model for a 10~Gyr old stellar population,

489: b) formation at a redshift of 2 followed by passive evolution,

490: and 

491: c) as for b) but with a formation redshift of 5.

492: %

493: \par

494: \end{minipage}

495: %

496: \end{center}

497: \vspace{3mm}

498: 

499: %%\begin{figure*}[b]

500: %%\plotfiddle{fig4.eps}{2in}{-90}{32}{32}{-120}{180}

501: %%\caption[fig4.eps]{

502: %%Residuals about the $\zf=5$ model for the X-ray selected BCG sample.

503: %%Filled and open symbols indicate BCGs in high- and low-\lx\

504: %%clusters respectively.

505: %%The lines show the expected locus of the residuals for the three

506: %%models shown in Figure~\ref{fig:z-mk}.

507: %%\label{fig:lz-dmk}}

508: %%\end{figure*}

509: 

510: \vspace{2mm}

511: \begin{center}

512: %

513: %\plotfiddle{fig4.eps}{2in}{-90}{32}{32}{-120}{180}

514: \plotfiddle{fig4.eps}{2.1in}{-90}{32}{32}{-120}{190}

515: %

516: \begin{minipage}{8.75cm}

517: \small\parindent=3.5mm

518: {\sc Fig.}~4.---Residuals about the $\zf=5$ model for the X-ray selected BCG sample.

519: Filled and open symbols indicate BCGs in high- and low-\lx\

520: clusters respectively.

521: The lines show the expected locus of the residuals for the three

522: models shown in Figure~2.

523: %

524: \par

525: \end{minipage}

526: %

527: \end{center}

528: \vspace{3mm}

529: 

530: To quantify the amount of evolution allowed by the data, the same

531: parametric form as employed by

532: \abkmain---namely $M(z) = M(0) \times (1+z)^\gamma$---was used to estimate 

533: the growth in the stellar mass content of BCGs.  

534: Fitting for both $\gamma$ and $M(0)$ indicates that, in the

535: high-\lx\ sample 

536: (which best approximates the cluster selection adopted for

537: the \semianalytic\ models), 

538: the typical BCG mass has increased by a factor

539: of $1.9\pm0.3$ (for $\zf=2$) or $1.3\pm0.2$ ($\zf=5$) 

540: between $z=1$ and the present.

541: These growth factors are substantially lower than either 

542: the factor of $\sim$4--5 predicted by the \semianalytic\ models,

543: or the measured values of 4.6 ($\zf=2$) and 3.2 ($\zf=5$),

544: of \abkmain.

545: The growth factor can also be estimated by fitting for $\gamma$ alone if one

546: assumes a low-redshift normalisation for the model predictions.

547: However, this currently involves applying a \colour-correction

548: to low-redshift optical BCG data,

549: which introduces further uncertainty: 

550: applying a single $R-K$ correction to the \xray-selected

551: sample of \lynam\ changes the measured growth factor 

552: of the high-\lx\ sample by less than 20\%,

553: whilst using the normalisation adopted by \abkmain---based

554: on an optically-selected sample---increases the

555: growth factor by 50\%.

556: K-band observations of the \lynam\ 

557: sample are being obtained to circumvent this problem in future

558: work.

559: 

560: \section{CONCLUSION}

561: \label{conclusion}

562: 

563: The K-band luminosities of BCGs are correlated with their environment:

564: clusters with a high \xray\ luminosity contain

565: BCGs which are brighter, and have a smaller scatter, 

566: than those BCGs in clusters with a low \xray\ luminosity.

567: The BCG evolution seen by \abkmain\ has been shown to be

568: an artifact of a selection bias in their cluster sample;

569: at high redshifts, their clusters are systematically less \xray\ luminous

570: than their low-redshift 

571: sample, and so their BCGs are systematically fainter.

572: 

573: Under the assumption of an Einstein-de Sitter universe,

574: non-parametric tests show that the only significant evidence for

575: BCG mass evolution over the range $0.05 \leq z \leq 0.83$

576: occurs when the dominant stellar population formed

577: relatively recently ($\zf \leq 2.6$).

578: Using the same parametric form as \abkmain, the masses of BCGs 

579: in high-\lx\ clusters are found to have, at most, doubled since $z=1$,

580: compared to the factor of $\sim 4$ increase predicted, for BCGs in

581: massive clusters, by the \semianalytic\ models discussed

582: by \abkmain.

583: 

584: \acknowledgements

585: 

586: DJB acknowledges support from PPARC grant 

587: GR/L21402

588: and SAO contract SV4-64008

589: and RGM that from 

590: PPARC at Imperial College and Edinburgh.

591: DJB would like to thank

592: Peter Draper, Tim Hawarden, and Sandy Leggett for useful discussions.

593: We thank the referee, Alfonso Arag\'{o}n-Salamanca, for

594: useful comments that improved the paper,

595: the service \programmes\ of both UKIRT and IRTF for obtaining

596: some of the data presented here,

597: and Piero Rosati and collaborators for providing aperture

598: magnitudes for the two Lynx clusters.

599: The United Kingdom Infrared Telescope is operated by the 

600: Joint Astronomy Centre on behalf of the U.K. Particle Physics and 

601: Astronomy Research Council.

602: 

603: %%\clearpage

604: 

605: \begin{references}

606: %%

607: \reference{abk98}

608: Arag\'{o}n-Salamanca, A., Baugh, C. M., \& Kauffmann, G.  1998, \mnras, 297, 427 (ABK98)

609: %%

610: \reference{aecc93}

611: Arag\'{o}n-Salamanca, A., Ellis, R. S., Couch, W. J., Carter, D. 1993, \mnras, 262, 764

612: %%

613: \reference{bcf96}

614: Baugh, C. M., Cole, S., \& Frenk, C. S. 1996, \mnras, 282, L27

615: %%

616: \reference{bershady95}

617: Bershady, M. A. 1995, \aj, 109, 87

618: %%

619: \reference{bc}

620: Bruzual A., G. \& Charlot, S. 1993, \apj, 405, 538   

621: %%

622: \reference{b97}

623: Burke, D. J., Collins, C. A., Sharples, R. M., Romer, A. K., Holden, B. P.,

624: \& Nichol, R. C. 1997, \apjl, 488, L83

625: %%

626: \reference{cm98}

627: Collins, C. A., \& Mann, R. G. 1998, \mnras, 297, 128 (CM98)

628: %%

629: \reference{ukirt-fs}

630: Casali, M. M., \& Hawarden, T. G. 1992, 

631: The JCMT-UKIRT Newsletter, No. 3, 33

632: %%

633: \reference{castander94}

634: Castander, F. J., Ellis, R. S., Frenk, C. S., Dressler, A.,  

635: \& Gunn, J. E. 1994, \apjl, 424, L79 

636: %%

637: \reference{depropris99}

638: De Propris, R., Stanford, S. A., Eisenhardt, P. R., Dickinson, M.,

639: \& Elston, R. 1999, \aj, 118, 719

640: %%

641: \reference{dressler78}

642: Dressler, A. 1978, \apj, 222, 23

643: %%

644: \reference{els62}

645: Eggen, O. J., Lynden-Bell, D., \& Sandage, A. R. 1962, \apj, 136, 748

646: %%

647: \reference{ellis97}

648: Ellis, R.S., Smail, I., Dressler, A., Couch, W.J., Oemler, A., Butcher, H., \&

649: Sharples, R.M., 1997, \apj, 483, 582

650: %%

651: \reference{gl-94}

652: Gioia, I. M., \& Luppino, G. A. 1994, \apjs, 94, 583

653: %%

654: \reference{hall98}

655: Hall, P. B., Green, R. F., \& Cohen, M.  1998, \apjs, 119, 1 

656: %%

657: \reference{h97}

658: Henry, J. P., et al. 1997, \aj, 114, 1293 

659: %%

660: \reference{h99}

661: Henry, J. P. 1999, \apj, submitted

662: %%

663: \reference{holden97}

664: Holden, B. P., Romer, A. K., Nichol, R. C., \& Ulmer, M. P. 1997, \aj, 114, 1701 

665: %%

666: \reference{jimenez99}

667: Jimenez, R., Friaca, A., Dunlop, J., Terlevich, R., Peacock, J.,

668: \& Nolan, L. 1999, \mnras, 305, L16

669: %%

670: \reference{kc98}

671: Kauffmann, G., \& Charlot, S. 1998, \mnras, 294, 705

672: %%

673: \reference{kw93}

674: Kauffmann, G., \& White, S. D. M. 1993, \mnras, 264, 201

675: %%

676: \reference{larson69}

677: Larson, R. B. 1969, \mnras, 145, 405

678: %%

679: \reference{pdl99}

680: Lynam, P. D., Collins, C. A., James, P. A., B\"{o}hringer, H., \&

681: Neumann, D. M. 1999, preprint (astro-ph/9908348)

682: %%

683: \reference{madau98}

684: Madau, P., Pozzetti, L., \& Dickinson, M. E. 1998, \apj, 498, 106

685: %%

686: \reference{n97}

687: Nichol, R. C., Holden, B. P., Romer, A. K., Ulmer, M. P.,

688: Burke, D. J., \& Collins, C. A. 1997, \apj, 481, 644

689: %%

690: \reference{opl98}

691: Oke, J. B., Postman, M.,  \& Lubin, L. M. 1998, \aj, 116, 549 

692: %%

693: \reference{morphs99}

694: Poggianti, B. M., Smail, I.,  Dressler, A.,  Couch, W. J., Barger, A. J., 

695: Butcher, H., Ellis, R. S., \& Oemler, A., Jr. 1999, \apj, 518, 576 

696: %%

697: \reference{bsharc}

698: Romer, A. K., et al. 2000, \apjs, in print

699: %%

700: \reference{rosati99}

701: Rosati, P., Stanford, S. A., Eisenhardt, P. R., Elston, R.,  Spinrad, H., 

702: Stern, D., \& Dey, A. 1999, \aj, 118, 76 

703: %%

704: \reference{schlegel98}

705: Schlegel, D., Finkbeiner, D., \& Davis, M. 1998, \apj, 500, 525

706: %%

707: \reference{sed95}

708: Stanford, S. A., Eisenhardt, P. R., \& Dickinson, M. 1995, \apj, 450, 512 

709: %%						     

710: \reference{sed98}				     

711: Stanford, S. A., Eisenhardt, P. R., \& Dickinson, M. 1998, \apj, 492, 461

712: %%

713: \reference{tb94}

714: Thimm, G. J., \& Belloni, P. 1994, \aap, 289, L27 

715: %%

716: \reference{v96}

717: van Dokkum, P. G., \& Franx, M., 1996, \mnras, 281, 985

718: %%

719: \reference{vf98}

720: van Dokkum, P. G., Franx, M., Kelson, D. D., \& Illingworth, G. D. 1998,

721: \apjl, 504, L17

722: %%

723: \reference{v99}

724: van Dokkum, P. G., Franx, M., Fabricant, D., Kelson, D. D., \& Illingworth, G. D. 1999,

725: \apjl, 520, L95

726: %%

727: \end{references}

728: 

729: \end{document}

730: