1: %\documentclass[12pt,preprint,psfig]{aastex}

2: \documentclass[12pt,preprint]{emulateapj}

3: 

4: %\documentclass[12pt,preprint]{aastex}

5: 

6: %\documentclass[manuscript]{aastex}

7: %\documentstyle[emulateapj5,psfig]{aastex}

8: %\usepackage{emulateapj5}

9: %\usepackage{epsf}

10: %\usepackage{epsfig}

11: %\usepackage{apjfonts}

12: 

13: \newcommand{\sub}[1]{_{\rm #1}}

14: \newcommand{\md}{\cal M}

15: \newcommand{\ab}[1]{#1\sub{AB}}

16: \newcommand{\oii}{{\sc [Oii]}}

17: \newcommand{\Mstar}{M\sub{star}}

18: \newcommand{\Rsfr}{\mathcal{R}\sub{SFR}}

19: \newcommand{\gapprox}{_>\atop^\sim}  % math mode only!

20: \newcommand{\ltsima}{$\buildrel<\over\sim$}

21: \newcommand{\lapprox}{\lower.5ex\hbox{\ltsima}}

22: \newcommand{\msun}{M$_{\odot}$}

23: \newcommand{\kband}{{\it K$_s$}-band}

24: 

25: %\setlength{\topmargin}{0.4in}

26: %\flushbottom

27: 

28: % \newcommand{\lapprox}{\ensuremath{\sim\kern-1em\raise 0.65ex\hbox{$< $}}}

29: %\newenvironment{inlinefigure}{

30: %\def\@captype{figure}

31: %\noindent\begin{minipage}{0.999\linewidth}\begin{center}}

32: %{\end{center}\end{minipage}\smallskip}

33: 

34: 

35: \shorttitle{Did New Spheroidals Arise Via Major Mergers?}

36: \shortauthors{Bundy et al.}

37: \slugcomment{accepted to ApJL}

38: 

39: \begin{document}

40: 

41: \title{The Mass Assembly History of Spheroidal Galaxies: Did 

42:  Newly-Formed Systems Arise Via Major Mergers?}

43: 

44: \author{Kevin Bundy\altaffilmark{1}, Tommaso Treu\altaffilmark{2},

45:   Richard S. Ellis\altaffilmark{3}}

46: 

47: %\email{kbundy@astro.caltech.edu, rse@atro.caltech.edu,

48: %  cc@astro.caltech.edu, cooper@astron.berkeley.edu, bjw@ucolick.org,

49: %  jet@astro.caltech.edu, cnaw@ucolick.org}

50: 

51: \altaffiltext{1}{Reinhardt Fellow, Dept.~of Astronomy and Astrophysics,

52:   University of Toronto, 50 St.~George Street, Rm 101, Toronto, ON M5S

53:   3H4, Canada}

54: 

55: \altaffiltext{2}{Alfred P. Sloan Research Fellow, Dept.~of Physics, University of California,

56: Santa Barbara, CA 93106-9530}

57: 

58: \altaffiltext{3}{105--24 Caltech, 1201 E. California Blvd., Pasadena,

59:   CA 91125}

60: 

61: \begin{abstract}

62: 

63:   We examine the properties of a morphologically-selected sample of $0.4

64:   < z < 1.0$ spheroidal galaxies in the GOODS fields in order to

65:   ascertain whether their increase in abundance with time arises

66:   primarily from major mergers. To address this question we determine scaling

67:   relations between the dynamical mass $M_{dyn}$, determined from

68:   stellar velocity dispersions, and the stellar mass $M_*$, determined

69:   from optical and infrared photometry.  We exploit these relations

70:   across the larger sample for which we have stellar masses in order to

71:   construct the first statistically robust estimate of the evolving {\em

72:     dynamical mass function} (MF) over $0<z<1$. The trends observed

73:   match those seen in the stellar mass functions of \citet{bundy05}

74:   regarding the top-down growth in the abundance of spheroidal

75:   galaxies. By referencing our dynamical masses to the halo

76:   virial mass, $M_{vir}$, we compare the growth rate in the abundance of

77:   spheroidals to that predicted by the assembly of dark matter

78:   halos. Our comparisons demonstrate that major mergers do not fully account

79:   for the appearance of new spheroidals since $z \sim 1$ and that

80:   additional mechanisms, such as morphological transformations, are

81:   required to drive the observed evolution.

82: 

83: % The abundance of the most massive

84: % spheroidals in our sample ($4 \times 10^{11}$\msun) shows no evolution

85: % across the redshift range probed, while the number density of the

86: % least massive systems ($2 \times 10^{10}$\msun) grows by at least a

87: % factor of 1.6 and likely by a factor of two.  

88: 

89: % Combined with the

90: % observed mass-dependent evolution in the mass-to-light (M/L) ratios as

91: % inferred from the fundamental plane, our results favor a picture in

92: % which spheroidals are created by the transformation of previously

93: % star-forming galaxies, a process that begins at the highest masses and

94: % becomes effective at lower masses with increasing cosmic time.

95: 

96: \end{abstract}

97: 

98: \keywords{galaxies: evolution --- galaxies: elliptical and lenticular

99:   --- galaxies: kinematics and dynamics --- galaxies: mass function}

100: 

101: \section{Introduction}

102: 

103: Substantial progress has been made in characterizing the

104: primarily old, red, and passively-evolving stellar populations of spheroidal

105: galaxies (defined here to include elliptical and lenticulars) from z$\simeq$1 

106: to the present day \citep[see][for a review]{renzini06}.  Their evolving 

107: abundance contains valuable clues to the processes by which they form,

108: as well as insight into the history of cosmic mass assembly.  In recent 

109: years, many surveys have converged on a broad evolutionary pattern 

110: in which it is claimed that galaxies with the largest stellar masses evolve  

111: predominantly into spheroidals by $z \gtrsim 1$ with little subsequent

112: growth. Meanwhile, intermediate and lower mass systems continue to form 

113: at later epochs  \citep[e.g.,][]{bundy05, bundy06, tanaka05, franceschini06,

114: borch06, pannella06, cimatti06, abraham07}.

115:   

116: Some authors \citep[e.g.,][]{van-dokkum05a} have questioned the absence

117: of growth among massive galaxies, pointing to a high rate of tidal

118: features in local examples, indicative of recent merging. It is likewise

119: argued \citep{drory04, van-der-wel06a, maraston06, kannappan07} that

120: stellar masses derived from optical-infrared photometry may suffer

121: biases that lead to large uncertainties in the inferred growth rate.

122: Beyond such concerns, interpreting the observed evolution in the context

123: of the $\Lambda$CDM paradigm necessitates appealing to potentially

124: uncertain semi-analytic models.  Clearly what is needed is a way to

125: connect observations of galaxy growth directly to the assembly history

126: of the halos in which they reside.

127: 

128: Spheroidal galaxies offer particular advantages in this regard.  First,

129: because of their regular mass profiles, spectroscopic observations can

130: be used to infer the dynamical mass, bypassing luminous quantities.

131: Second, because of their pressure-supported dynamical configurations,

132: they are expected to result, for the most part, from mergers

133: \citep[e.g.,][]{toomre77, springel05b}. Accordingly, their evolving

134: abundances can be compared directly to the expected assembly history of

135: dark matter halos.  In this Letter we develop a technique to estimate

136: dynamical and halo masses for a large sample of field spheroidals and

137: use it to interpret their evolution in the context of halo assembly as

138: predicted in numerical simulations.  We assume $\Omega_{\rm M}=0.3$,

139: $\Omega_\Lambda=0.7$, $H_0=70 h_{70}$ km~s$^{-1}$~Mpc$^{-1}$.

140: 

141: \section{Observations and Mass Estimates}\label{data}

142: 

143: The dynamical mass (hereafter $M_{dyn}$) measurements used in this

144: Letter are based on the analysis presented in \citet{treu05a}, to which

145: we refer for details. Briefly, this sample comprises 165 field

146: visually-classified spheroidals selected from ACS imaging in the

147: northern GOODS field to a magnitude limit of $z_{AB} < 22.5$ and studied

148: with the DEIMOS spectrograph on Keck II.  A subset of 125 galaxies have

149: reliable stellar velocity dispersions, stellar masses, and surface

150: photometry fits, which provide $B$-band effective radii, $R_{eB}$

151: \citep[see][]{treu05a}.  Dynamical masses were derived using the

152: formula, $M_{dyn} = K_v \sigma^2 R_{eB}/G$ where $K_v$ is the virial

153: coefficient and $\sigma$ is the {\it central} velocity dispersion,

154: obtained by increasing the observed dispersion by 10\% to correct to the

155: standard circular aperture of diameter $R_{\rm eB}/4$. Detailed

156: modelling of fiber as well as 3D spectroscopy of nearby ellipticals

157: supports the use of this equation \citep{padmanabhan04, cappellari06}

158: with little uncertainties due to the overall flatness of the velocity

159: dispersion profile.

160: 

161: With the kinematic sample defined in this way, the investigation that

162: follows is based on the much larger sample of galaxies with stellar

163: masses (hereafter $M_*$) in both GOODS fields with $z_{AB} < 22.5$

164: \citep{bundy05}. We refer to that paper for further details but

165: summarize the key features.  For $0.4 < z < 1.0$, spectroscopic

166: redshifts for 633 galaxies from the Keck Team Redshift Survey

167: \citep[KTRS,][]{wirth04} were supplemented with 695 determined using

168: photometric techniques \citep[see][]{bundy05}.  The full sample of

169: morphologically-selected spheroidals includes 393 sources. Using deep

170: \kband\ imaging obtained at Palomar Observatory for GOODS-N and publicly

171: available data in GOODS-S, stellar masses were estimated assuming a

172: Chabrier \citep{chabrier03} initial mass function (IMF) using the code

173: described in \citet{bundy06}\footnote{The results can be adjusted to

174:   those for a Salpeter IMF by adding $\sim$0.25 dex to the derived

175:   masses.}.  For galaxies with photometric redshifts, an additional 0.3

176: dex is added in quadrature to the base uncertainty which is typically

177: 0.2 dex \citep{bundy05,bundy06}.

178: 

179: As discussed in \citet{bundy05}, the GOODS morphological sample is

180: complete in the \kband\ but becomes incomplete for very red galaxies

181: below a certain stellar mass limit as a result of the $z_{AB} < 22.5$

182: limit (see their Figure 4).  By considering synthetic SEDs representing

183: red (in $z - K_s$ color) stellar populations, we determine mass limits

184: of $\log (M_{limit}/M_{\odot}) = [9.8, 10.4, 10.8]$ for redshifts, $z =

185: [0.4, 0.7, 1.0]$.  The completeness limits of the MFs below are

186: determined at the near edge of each redshift interval.  When comparing

187: $M_*$ to $M_{dyn}$ we employ the more strict, volume-limited limits at

188: the far edge of each interval \citep[see discussion in][]{bundy06}.

189: 

190: \section{Linking Dynamical Masses and Stellar Masses}

191: 

192: We begin by restricting our analysis to those 125 spheroidals in

193: GOODS-N for which both $M_{dyn}$ and $M_*$ measurements are

194: available. The aim is to develop scaling relations that can be applied

195: to the larger GOODS-N/S sample. For local galaxies, a fairly tight

196: correlation of the form $M_{dyn} = (M_*)^{\alpha}$ has been observed

197: by several authors \citep[e.g.,][]{padmanabhan04, cappellari06,

198: gallazzi06}.  \citet{gallazzi06} exploited a local sample of

199: early-type systems selected by concentration, and find $\alpha = 1.28

200: \pm 0.03$---the zeropoint is not explicitly solved for.  In this work,

201: we adopt a zeropoint such that $M_{dyn} \equiv M_*$ at $10^{11}$\msun,

202: and also apply it to the \citet{gallazzi06} relation.  This choice

203: corresponds to $K_v=4.8$ and 6.6 at $0.4 < z < 0.7$ and $0.7 < z <

204: 1.0$ and is motivated by the fact that the $K_v$ depends on the mass

205: density profile and, most importantly, on the spectroscopic aperture

206: used to measure $\sigma$. Since the aperture is effectively

207: redshift-dependent (the slit width is fixed at 1\arcsec), assuming a

208: constant $K_v$ could introduce a spurious redshift dependency.

209: Similarly, while there is no evidence that the density profile evolves

210: for high mass spheroidals \citep[e.g.,][]{koopmans06}, potential

211: evolution could mimic changes in the MF if $K_v$ were held fixed. More

212: detailed studies are needed in order to determine the normalization

213: independently.

214: 

215: 

216: % Figure 1

217: \begin{figure}

218: \plotone{f1.eps}

219: \caption{Correlation between $M_{dyn}$ and $M_*$.  The dotted line

220:   illustrates the fitted slope, $\alpha$, which is also indicated in the

221:   top left corner.  A typical error bar is shown and the relation found

222:   by \citet{gallazzi06} is also plotted.  The normalization in each

223:   panel has been set to the same scale, as described in the

224:   text.  \label{relplot}}

225: \end{figure}

226: 

227: 

228: We examine the $M_{dyn}$--$M_*$ relation at $0.4 < z < 0.7$ and $0.7 <

229: z < 1.0$ in Figure \ref{relplot}, restricting the comparison to

230: systems above our completeness limit; the best fit values for $\alpha$

231: are also indicated. Our observed slopes are formally consistent with

232: those of \citet[$1.28\pm0.03$]{gallazzi06} at $z=0$ and

233: \citet[$1.25\pm0.05$]{Rettura06} at $z \sim 1$ \citep[see also]{di-serego-alighieri05}, although some

234: differences may be expected as a result of sample selection and fiber-

235: versus slit-based measures of velocity dispersion.  The fact that

236: $\alpha > 1$ reflects the ``tilt'' of the Fundamental Plane

237: \citep[e.g.][]{faber87,CLR96,renzini06} and is consistent with

238: non-homology \citep[e.g.][]{TBB04} including an increasing dark matter

239: fraction within $R_e$ at higher masses \citep[e.g.][]{padmanabhan04,

240: gallazzi06, bolton07}.  With regard to what follows, the key result in

241: Figure \ref{relplot} is that $\alpha$ shows no measurable evolution

242: across our sample.  In $\S$\ref{section:DMF}, we will show that this

243: leads to dynamical MFs that behave similarly to their stellar mass

244: counterparts.  For consistency, final $M_{dyn}$ estimates for galaxies

245: in the full GOODS sample are determined using the value of $\alpha$

246: determined for the two redshift bins, and the final uncertainty in

247: $M_{dyn}$ is taken to be the sum in quadrature of the uncertainty in

248: $M_*$ and the scatter in Figure \ref{relplot}.

249: 

250: % Figure 2: Mdyn versus Mstar Relations: MLvsL/Mdyn_vs_Mstar.eps -

251: % Made with MLvsL/scripts.pro

252: % -- Black/White

253: \begin{figure}

254: \plotone{f2.eps}

255: \caption{Comparison of evolving dynamical and stellar mass functions.

256:   The top two panels show the MFs in two redshift intervals, while the

257:   bottom panel illustrates the differential growth between them.

258:   Solid points with gray shading trace the stellar mass function of

259:   the full (all types) GOODS $z_{AB} < 22.5$ sample, while crosses

260:   denote the equivalent $z_{AB} < 23.5$ sample.  Asterisks with light red shading

261:   trace the stellar MF of the spheroidal component, while diamonds

262:   with blue shading indicate the dynamical MF.  The fact

263:   that the observed increase with time in the 

264:   spheroidal MFs is more significant than the increase observed for the full

265:   sample (bottom panel), shows that this result is robust to potential

266:   incompleteness.  Isolated error bars

267:   indicate the estimated uncertainty from cosmic variance.

268:   \label{lvm_comp}}

269: \end{figure}

270: 

271: \section{The Dynamical Mass Function}\label{section:DMF}

272: 

273: Using the $M_{dyn}$ estimates derived above for the full GOODS sample,

274: we now construct the dynamical MF and compare with its

275: stellar mass counterpart in the top two panels of Figure \ref{lvm_comp}.

276: We use the $1/V_{max}$ estimator and closely follow the method described

277: in \citet{bundy05}. The solid points with grey shading show the stellar MF for

278: the full sample, while the crosses indicate the full stellar MF that

279: results when the magnitude limit is relaxed to $z_{AB} < 23.5$.  The

280: bottom panel of Figure \ref{lvm_comp} shows the differential growth

281: derived across the two redshift bins.

282: 

283: 

284: % MFN Comparison - stellar, M_dyn-stellar, M_dyn-LK: mfn/dmfn_LvM_comp.eps -

285: % Old Figure (with LvM) Made with mfn/dmfn, /LvM, res_mvm=res_mvm, /comp

286: % Feb14 Figure Made with mfn/dmfn, /MvM, /comp

287: % -- Color, CMYK

288: 

289: Figure \ref{lvm_comp} demonstrates consistency between the stellar MFs

290: and dynamical MFs. Although our adopted zeropoint in the $M_{dyn}$

291: relations ensures a similar vertical normalization, the application of

292: our scaling relations does not mean both MFs should have the same shape.

293: Indeed, small differences are noticeable in our highest mass bin.  The

294: good agreement results in large part from the fact that the

295: $M_{dyn}$--$M_*$ relation does not vary significantly across our sample.

296: Had a variation in $\alpha$ with redshift been present in Figure

297: \ref{relplot}, the resulting dynamical MFs would evolve with respect to

298: the stellar MFs.  The fact the two agree validates earlier stellar mass

299: estimates at $z \sim 1$ and the top-down growth in the abundance of

300: spheroidals derived from those measures.

301: 

302: % Incompleteness is important to understand because it could reduce the

303: % number of measured high-$z$ spheroidals and therefore produce an

304: % apparent evolutionary signal.  In $\S$\ref{data}, we described ways of

305: % estimating mass completeness limits, but incompleteness can also be

306: % understood by considering in Figure \ref{lvm_comp} the discrepancy

307: % between the deeper $z_{AB} < 23.5$ sample (cross symbols) and the full

308: % morphological $z_{AB} < 22.5$ sample (solid points with shading).  We

309: % can eliminate the extreme hypothesis that the observed evolution in

310: % the total MF ($z_{AB} <22.5$) results entirely from missing high-$z$

311: % spheroidal galaxies, since yet greater evolution is seen in the growth

312: % of detected spheroidals (red shading c.f. grey shading in the bottom

313: % panel of Figure \ref{lvm_comp})..  A more likely scenario in which

314: % half of the missing galaxies (below the completeness limit) are

315: % spheroidals \citep[see][]{bundy06} would mean that the spheroidal mass

316: % function---as traced using both stellar mass and dynamical mass---has

317: % increased by a factor of $\sim$2 at the lowest probed masses

318: % ($\sim$10$^{10}$\msun) and, most strikingly, not at all at the highest

319: % ($\sim$3$\times 10^{11}$\msun).

320: 

321: \section{Testing the Role of Mergers}

322: 

323: We now turn to the key question which our analysis can address: is the

324: rising abundance with time of spheroidal galaxies consistent with that

325: predicted by the growth of merging dark matter halos?  To answer this

326: question we adopt a simple, observationally-motivated model to connect

327: $\sigma$ and $M_*$ to the virial mass, $M_{vir}$, of the host halos.

328: Lensing studies of early-type galaxies at $z \approx 0.2$

329: \citet[see][and references therein]{gavazzi07} show that the total mass

330: density profile of spheroidal halos is nearly isothermal within the halo

331: scale radius, $r_s$.  We therefore assume a mass profile that is

332: isothermal for $r \lesssim r_s$ and follows a $\rho \propto r^{-3}$

333: profile \citep[NFW,][]{navarro97} for $r \gtrsim r_s$.  We note that the

334: $\rho \propto r^{-3}$ behavior at large radii ($r \sim r_{vir}$) is consistent with non-NFW

335: scaling laws which can disagree instead at small radii

336: \citep[e.g.,][]{graham06}.  The mass profile assumed here is given by

337: $\rho(r) = \sigma^2 [2 \pi G r^2 (1+r/\gamma r_s)]^{-1}$, where

338: $\sigma$ is the observed central velocity dispersion and we have

339: introduced the scaling parameter, $\gamma$, which can be used to tune

340: the profile shape to observations.  By comparing to \citet{gavazzi07},

341: we find $\gamma = 12$---future observations will constrain $\gamma$ as

342: a function of redshift and mass.  We define $M_{vir} = \frac{4\pi}{3}

343: \Delta(z) \rho_c c^3 r_s^3$, where $\rho_c$ is the cosmic critical

344: density, $\Delta(z)$ is the overdensity parameter

345: \citep[see][]{bryan98}, and $r_{vir}= c r_s$.  The concentration, $c$, is assumed to the

346: follow the relation found in simulations, $c = 9(1+z)^{-1}(M_{vir}/(8.12

347: \times 10^{12}h^{-1} M_{\odot}))^{-0.14}$ \citep{bullock01}.  Using this

348: model and the relations above, we solve for $M_{vir}$ as a function of

349: velocity dispersion and find,

350: 

351: \begin{equation}

352: \log M_{vir} = A_{\sigma} \log \sigma_{200} + B_{\sigma} + C_{\sigma}(1+z),

353: \label{eqn_sigma}

354: \end{equation}

355: 

356: \noindent where $M_{vir}$ has units of $h_{70}^{-1}$\msun,

357: $\sigma_{200}$ is the velocity dispersion in units of 200 km s$^{-1}$,

358: and the fit parameters are given by $A_{\sigma}=3.1$,

359: $B_{\sigma}=13.2$ and $C_{\sigma}=-0.3$.  Equation \ref{eqn_sigma}

360: follows from the model described above and so random uncertainties on

361: the fitting parameters are negligible.  However, systematic errors

362: resulting from tuning this model to observations should reflect the

363: uncertainty found by \citet{gavazzi07} of 0.2 dex.  Using our

364: kinematic sample for which we measure both $\sigma$ and $M_*$, we can

365: also directly compare $M_{vir}$ and $M_*$, finding the following

366: relation:

367: 

368: \begin{equation}

369: \log M_{vir} = A_{*} \log M_{*11} + B_{*} + C_* (1+z)

370: \label{eqn_mstar}

371: \end{equation}

372: 

373: \noindent where a Chabrier IMF was assumed as before and $M_{*11}$ has

374: units of $10^{11} h_{70}^{-2}$\msun.  The best fitting parameters are

375: $A_{*}=1.0 \pm 0.2$ and $B_{*}=12.83 \pm 0.11 \pm \sigma_{B,sys}$,

376: $C_*=-0.2 \pm 0.1$ where $\sigma_{B,sys} = 0.25$ dex and accounts for

377: the systematic uncertainty in Equation \ref{eqn_sigma} (0.2 dex) as well

378: as uncertainty in the stellar IMF (0.15 dex). Equations 1 and 2 are

379: applicable for galaxies with $\log

380: M_*/M_\odot \approx 10-12$ and $\sigma \approx 90-300$ kms$^{-1}$.  Equation

381: \ref{eqn_mstar} gives values of $M_{vir}/M_* = 30^{+25}_{-15}$,

382: consistent with \citet{mandelbaum05} and \citet{gavazzi07}.  We note

383: that our marginally non-zero value of $C_*$, if confirmed, implies a

384: slight increase with time of $M_{vir}/M_*$ ($\sim$30\%) over $0.4 < z

385: < 1.0$ that is consistent with halo growth \citep[see discussion

386: in][]{conroy07} and independent evidence that stellar mass assembly is

387: mostly completed by $z \approx 1$ \citep[e.g.][]{treu05a}. We also

388: note that virial mass and stellar mass turn out to be approximately in

389: linear proportion within this mass range.  Studies with a larger

390: dynamical range in stellar masses and velocity dispersion, and more

391: weak lensing measurements, are needed to measure the slope with higher

392: precision and detect departures from linearity, expected as a result

393: of varying star formation efficiency with halo mass.

394: 

395: Using Equation \ref{eqn_mstar} we estimate halo virial masses for our sample and compare the resulting virial MFs in the

396: top panel of Figure \ref{mvm_evol}.  The dashed line is the total halo

397: mass function predicted at $z \approx 0.5$ (it evolves little over

398: $0<z<1.0$) using the online Millennium Simulation

399: Database\footnote{http://www.mpa-garching.mpg.de/millennium.  The

400:   ``milli-Millennium'' Database covers a volume of $7 \times

401:   10^{5}$~$h_{70}^{-3}$Mpc$^3$, nearly 4 times larger than our largest

402:   sampled volume at $0.7 < z < 0.4$.} \citep{springel05a} and adjusted

403: in number density by $+$0.1 dex to be consistent with the spheroidal MFs

404: (this adjustment is within the expected cosmic variance

405: of 0.15 dex).  The dotted line shows a local ($z \approx 0.1$) estimate constructed by

406: converting the $M_*$ MF for early-types (based on concentration, $C \ge

407: 2.6$) measured by \citet{bell03} into a virial MF using Equation

408: \ref{eqn_mstar}.  Adjustments of $+$0.1 dex were also applied to the

409: local number density and the local mass scale to match the GOODS sample

410: at high masses.  Such corrections are likely needed as a result of

411: cosmic variance, differing selection and mass estimate methods, and

412: photometric uncertainties.  Thus, comparisons between

413: our results and the local MF must be done with caution.

414: 

415: The top panel of Figure \ref{mvm_evol} reveals a similar pattern of

416: evolution as seen in Figure \ref{lvm_comp}.  A useful interpretation of

417: this evolution is given in the bottom panel of Figure \ref{mvm_evol}

418: which plots the $\log$ {\em difference} ($\log \Delta dN/dM$), thereby

419: displaying the halo MF of new spheroidals that have appeared over the

420: time interval separating the two redshift windows and providing a way to

421: compare with the incidence of halo mergers.  We use the MPAHalo

422: milli-Millennium database to identify halos that have experienced a

423: recent merger, at which point two previously separate halos or subhalos

424: become a single halo.  Halo merging often occurs several Gyr after halo

425: accretion, defined by the point at which the two progenitor halos first

426: become associated with the same friends-of-friends (FOF) group.  Because the

427: smaller progenitor halo may experience significant tidal stripping after

428: halo accretion but before merging \citep[see][]{gao04}, we take the

429: merger mass ratio to be the initial ratio before halo accretion.

430: 

431: After integrating over the two time intervals that separate our

432: observations, we overplot two merging scenarios in the bottom panel of

433: Figure \ref{mvm_evol} (the $+$0.1 dex vertical offset is also applied).

434: It is a striking result that major halo mergers (mass ratios greater

435: than 1:3, indicated by dark lines) cannot account for the rise of

436: spheroidal systems over the two epochs probed here.  The light yellow

437: lines (without symbols) show this is still the case when a $\sim$0.5 Gyr

438: delay is added to the merger timescale (as suggested by expected

439: differences in the galaxy and halo merger times).  We note that at the

440: highest masses probed there is an indication that the predicted merger

441: frequency rises above the growth of new spheroidals.  This might be

442: expected since most systems at such masses already exhibit spheroidal

443: morphologies by these redshifts \citep[e.g.,][]{bundy05}, suggesting

444: such interactions represent ``dry'' mergers

445: \citep[e.g.,][]{van-dokkum05a}.

446: 

447: % Figure 3: Dyn. MFN evolution (compared to SDSS): mfn/dmfn_MvM_evol.eps -

448: % Made with mfn/dmfn, /MvM, /evol

449: % -- Color, CMYK

450: \begin{figure}

451: \plotone{f3.eps}

452: \caption{Evolving virial mass functions of halos hosting spheroidals.

453:   The top panel shows MFs in two $z$-bins. The local MF estimated from

454:   \citet{bell03}, is indicated by the dotted line.  The dashed line

455:   shows the total MF of predicted dark matter halos, and the isolated

456:   error bar (top right) denotes the cosmic variance uncertainty.  The

457:   bottom panel shows the MF of newly formed spheroidals determined by

458:   subtracting the MFs (top panel) in adjacent redshift bins.

459:   Overplotted are two halo merger scenarios based on the Millennium

460:   Simulation, integrated over the redshift ranges indicated.  Black

461:   lines with triangle symbols trace $>$1:3 mergers while yellow lines

462:   (without symbols) denote such mergers with a $\sim$0.5 Gyr time

463:   delay.  \label{mvm_evol}}

464: \end{figure}

465: 

466: 

467: \section{Discussion}

468: 

469: By constructing dynamical mass functions for 393 field spheroidals and

470: linking those mass estimates to the halos in which they reside, we have

471: found a significant result in the bottom panel of Figure \ref{mvm_evol}:

472: major merging of dark matter halos as described by $\Lambda$CDM does not

473: occur frequently enough to explain the observed increase in the mass

474: function of spheroidal galaxies.  A check made by appealing to the

475: semi-analytic model of \citet{de-lucia07} shows a similar result when a

476: galaxy's stellar mass is used to define recent mergers and even when

477: $M_*$ is used as the ``accounting variable.'' 

478: 

479: If merging is not the only mechanism that produces newly-formed

480: spheroidals since $z \sim 1$, other physical processes likely play a

481: significant role.  These may include mechanisms for

482: transforming disk and irregular galaxies into relaxed spheroidals, for

483: example secular bulge growth accompanied by disk fading.  Indeed, recent

484: semi-analytic models \citep[e.g.,][]{croton06, bower06, de-lucia07} have

485: required prescriptions for such mechanisms, in addition to mergers, to match the local

486: abundance of spheroidals.  More work is needed to understand these

487: processes and their role in morphological evolution.

488: 

489: \acknowledgments 

490: 

491: We are very grateful to Simon White and Volker Springel for help

492: in deriving results from the Millennium Run Database and also thank Raphael

493: Gavazzi, Andrew Benson, Phil Hopkins, and Karl Glazebrook for useful

494: discussions.  RSE acknowledges the hospitality of Ray Carlberg at the

495: University of Toronto during a sabbatical visit and TT acknowledges

496: support from a Sloan Research Fellowship.

497: 

498: %\bibliographystyle{apj}

499: %\bibliography{references}

500: 

501: \begin{thebibliography}{42}

502: \expandafter\ifx\csname natexlab\endcsname\relax\def\natexlab#1{#1}\fi

503: 

504: \bibitem[{{Abraham} {et~al.}(2007)}]{abraham07}

505: {Abraham}, R.~G. {et~al.} 2007, preprint (astro-ph/0701779)

506: 

507: \bibitem[{{Bell} {et~al.}(2003){Bell}, {McIntosh}, {Katz}, \&

508:   {Weinberg}}]{bell03}

509: {Bell}, E.~F., {McIntosh}, D.~H., {Katz}, N., \& {Weinberg}, M.~D. 2003, \apjl,

510:   585, L117

511: 

512: \bibitem[{{Bolton} {et~al.}(2007){Bolton}, {Burles}, {Treu}, {Koopmans}, \&

513:   {Moustakas}}]{bolton07}

514: {Bolton}, A.~S., {Burles}, S., {Treu}, T., {Koopmans}, L.~V.~E., \&

515:   {Moustakas}, L.~A. 2007, preprint (astro-ph/0701706)

516: 

517: \bibitem[{{Borch} {et~al.}(2006)}]{borch06}

518: {Borch}, A. {et~al.} 2006, preprint (astro-ph/0604405)

519: 

520: \bibitem[{{Bower} {et~al.}(2006)}]{bower06}

521: {Bower}, R.~G. {et~al.} 2006, \mnras, 370, 645

522: 

523: \bibitem[{{Bryan} \& {Norman}(1998)}]{bryan98}

524: {Bryan}, G.~L. \& {Norman}, M.~L. 1998, \apj, 495, 80

525: 

526: \bibitem[{{Bullock} {et~al.}(2001){Bullock}, {Kolatt}, {Sigad}, {Somerville},

527:   {Kravtsov}, {Klypin}, {Primack}, \& {Dekel}}]{bullock01}

528: {Bullock}, J.~S., {Kolatt}, T.~S., {Sigad}, Y., {Somerville}, R.~S.,

529:   {Kravtsov}, A.~V., {Klypin}, A.~A., {Primack}, J.~R., \& {Dekel}, A. 2001,

530:   \mnras, 321, 559

531: 

532: \bibitem[{{Bundy} {et~al.}(2005){Bundy}, {Ellis}, \& {Conselice}}]{bundy05}

533: {Bundy}, K., {Ellis}, R.~S., \& {Conselice}, C.~J. 2005, \apj, 625, 621

534: 

535: \bibitem[{{Bundy} {et~al.}(2006)}]{bundy06}

536: {Bundy}, K. {et~al.} 2006, \apj, 651, 120

537: 

538: \bibitem[{{Cappellari} {et~al.}(2006)}]{cappellari06}

539: {Cappellari}, M. {et~al.} 2006, \mnras, 366, 1126

540: 

541: \bibitem[{{Chabrier}(2003)}]{chabrier03}

542: {Chabrier}, G. 2003, \pasp, 115, 763

543: 

544: \bibitem[{{Cimatti} {et~al.}(2006){Cimatti}, {Daddi}, \& {Renzini}}]{cimatti06}

545: {Cimatti}, A., {Daddi}, E., \& {Renzini}, A. 2006, \aap, 453, L29

546: 

547: \bibitem[{{Ciotti} {et~al.}(1996){Ciotti}, {Lanzoni}, \& {Renzini}}]{CLR96}

548: {Ciotti}, L., {Lanzoni}, B., \& {Renzini}, A. 1996, \mnras, 282, 1

549: 

550: \bibitem[{{Conroy} {et~al.}(2007)}]{conroy07}

551: {Conroy}, C. {et~al.} 2007, \apj, 654, 153

552: 

553: \bibitem[{{Croton} {et~al.}(2006)}]{croton06}

554: {Croton}, D.~J. {et~al.} 2006, \mnras, 365, 11

555: 

556: \bibitem[{{de Lucia} \& {Blaizot}(2007)}]{de-lucia07}

557: {de Lucia}, G. \& {Blaizot}, J. 2007, \mnras, 375, 2

558: 

559: \bibitem[{{di Serego Alighieri} {et~al.}(2005)}]{di-serego-alighieri05}

560: {di Serego Alighieri}, S. {et~al.} 2005, \aap, 442, 125

561: 

562: \bibitem[{{Drory} {et~al.}(2004){Drory}, {Bender}, \& {Hopp}}]{drory04}

563: {Drory}, N., {Bender}, R., \& {Hopp}, U. 2004, \apjl, 616, L103

564: 

565: \bibitem[{{Faber} {et~al.}(1987){Faber}, {Dressler}, {Davies}, {Burstein}, \&

566:   {Lynden-Bell}}]{faber87}

567: {Faber}, S.~M., {Dressler}, A., {Davies}, R.~L., {Burstein}, D., \&

568:   {Lynden-Bell}, D. 1987, in Nearly Normal Galaxies. From the Planck Time to

569:   the Present, ed. S.~M. {Faber}, 175--183

570: 

571: \bibitem[{{Franceschini} {et~al.}(2006)}]{franceschini06}

572: {Franceschini}, A. {et~al.} 2006, \aap, 453, 397

573: 

574: \bibitem[{{Gallazzi} {et~al.}(2006){Gallazzi}, {Charlot}, {Brinchmann}, \&

575:   {White}}]{gallazzi06}

576: {Gallazzi}, A., {Charlot}, S., {Brinchmann}, J., \& {White}, S.~D.~M. 2006,

577:   \mnras, 370, 1106

578: 

579: \bibitem[{{Gao} {et~al.}(2004){Gao}, {White}, {Jenkins}, {Stoehr}, \&

580:   {Springel}}]{gao04}

581: {Gao}, L., {White}, S.~D.~M., {Jenkins}, A., {Stoehr}, F., \& {Springel}, V.

582:   2004, \mnras, 355, 819

583: 

584: \bibitem[{{Gavazzi} {et~al.}(2007){Gavazzi}, {Treu}, {Rhodes}, {Koopmans},

585:   {Bolton}, {Burles}, {Massey}, \& {Moustakas}}]{gavazzi07}

586: {Gavazzi}, R., {Treu}, T., {Rhodes}, J.~D., {Koopmans}, L.~V., {Bolton}, A.~S.,

587:   {Burles}, S., {Massey}, R., \& {Moustakas}, L.~A. 2007, preprint

588:   (astro-ph/0701589)

589: 

590: \bibitem[{{Graham} {et~al.}(2006){Graham}, {Merritt}, {Moore}, {Diemand}, \&

591:   {Terzi{\'c}}}]{graham06}

592: {Graham}, A.~W., {Merritt}, D., {Moore}, B., {Diemand}, J., \& {Terzi{\'c}}, B.

593:   2006, \aj, 132, 2701

594: 

595: \bibitem[{{Kannappan} \& {Gawiser}(2007)}]{kannappan07}

596: {Kannappan}, S.~J. \& {Gawiser}, E. 2007, preprint (astro-ph/0701749)

597: 

598: \bibitem[{{Koopmans} {et~al.}(2006){Koopmans}, {Treu}, {Bolton}, {Burles}, \&

599:   {Moustakas}}]{koopmans06}

600: {Koopmans}, L.~V.~E., {Treu}, T., {Bolton}, A.~S., {Burles}, S., \&

601:   {Moustakas}, L.~A. 2006, \apj, 649, 599

602: 

603: \bibitem[{{Mandelbaum} {et~al.}(2005){Mandelbaum}, {Tasitsiomi}, {Seljak},

604:   {Kravtsov}, \& {Wechsler}}]{mandelbaum05}

605: {Mandelbaum}, R., {Tasitsiomi}, A., {Seljak}, U., {Kravtsov}, A.~V., \&

606:   {Wechsler}, R.~H. 2005, \mnras, 362, 1451

607: 

608: \bibitem[{{Maraston} {et~al.}(2006)}]{maraston06}

609: {Maraston}, C. {et~al.} 2006, preprint (astro-ph/0604530)

610: 

611: \bibitem[{{Navarro} {et~al.}(1997){Navarro}, {Frenk}, \& {White}}]{navarro97}

612: {Navarro}, J.~F., {Frenk}, C.~S., \& {White}, S.~D.~M. 1997, \apj, 490, 493

613: 

614: \bibitem[{{Padmanabhan} {et~al.}(2004)}]{padmanabhan04}

615: {Padmanabhan}, N. {et~al.} 2004, New Astronomy, 9, 329

616: 

617: \bibitem[{{Pannella} {et~al.}(2006)}]{pannella06}

618: {Pannella}, M. {et~al.} 2006, \apjl, 639, L1

619: 

620: \bibitem[{{Renzini}(2006)}]{renzini06}

621: {Renzini}, A. 2006, \araa, 44, 141

622: 

623: \bibitem[{{Rettura} {et~al.}(2006){Rettura}, {Rosati}, {Strazzullo},

624:   {Dickinson}, {Fosbury}, {Rocca-Volmerange}, {Cimatti}, {di Serego Alighieri},

625:   {Kuntschner}, {Lanzoni}, {Nonino}, {Popesso}, {Stern}, {Eisenhardt},

626:   {Lidman}, \& {Stanford}}]{Rettura06}

627: {Rettura}, A., {Rosati}, P., {Strazzullo}, V., {Dickinson}, M., {Fosbury},

628:   R.~A.~E., {Rocca-Volmerange}, B., {Cimatti}, A., {di Serego Alighieri}, S.,

629:   {Kuntschner}, H., {Lanzoni}, B., {Nonino}, M., {Popesso}, P., {Stern}, D.,

630:   {Eisenhardt}, P.~R., {Lidman}, C., \& {Stanford}, S.~A. 2006, \aap, 458, 717

631: 

632: \bibitem[{{Springel} {et~al.}(2005{\natexlab{a}}){Springel}, {Di Matteo}, \&

633:   {Hernquist}}]{springel05b}

634: {Springel}, V., {Di Matteo}, T., \& {Hernquist}, L. 2005{\natexlab{a}}, \apjl,

635:   620, L79

636: 

637: \bibitem[{{Springel} {et~al.}(2005{\natexlab{b}})}]{springel05a}

638: {Springel}, V. {et~al.} 2005{\natexlab{b}}, \nat, 435, 629

639: 

640: \bibitem[{{Tanaka} {et~al.}(2005)}]{tanaka05}

641: {Tanaka}, M. {et~al.} 2005, \mnras, 362, 268

642: 

643: \bibitem[{{Toomre}(1977)}]{toomre77}

644: {Toomre}, A. 1977, in Evolution of Galaxies and Stellar Populations, 401--+

645: 

646: \bibitem[{{Treu} {et~al.}(2005)}]{treu05a}

647: {Treu}, T. {et~al.} 2005, \apj, 633, 174

648: 

649: \bibitem[{{Trujillo} {et~al.}(2004){Trujillo}, {Burkert}, \& {Bell}}]{TBB04}

650: {Trujillo}, I., {Burkert}, A., \& {Bell}, E.~F. 2004, \apjl, 600, L39

651: 

652: \bibitem[{{van der Wel} {et~al.}(2006)}]{van-der-wel06a}

653: {van der Wel}, A. {et~al.} 2006, \apj, 652, 97

654: 

655: \bibitem[{{van Dokkum}(2005)}]{van-dokkum05a}

656: {van Dokkum}, P.~G. 2005, \aj, 130, 2647

657: 

658: \bibitem[{{Wirth} {et~al.}(2004)}]{wirth04}

659: {Wirth}, G.~D. {et~al.} 2004, \aj, 127, 3121

660: 

661: \end{thebibliography}

662: 

663: \end{document}

664: 

665: