astro-ph0307545/ms.tex
1: \documentclass[12pt,preprint]{aastex} 
2: %\documentstyle[12pt,/home/propris/final/emulateapj]{article} 
3: %\documentstyle[12pt,aaspp4]{article} 
4: %\documentstyle[11pt,/home/propris/final/emulateapj]{article} 
5: %\usepackage[/home/propris/final/emulateapj5] 
6: %\tighten 
7: 
8: %\slugcomment{\shortstack[r]{Submitted to {\sl The Astrophysical Journal}}} 
9: 
10: \begin{document} 
11: 
12: \title 
13: {The $K$--selected Butcher-Oemler Effect} 
14: 
15: \author 
16: {Roberto De Propris} 
17: 
18: \affil 
19: {Research School of Astronomy \& Astrophysics, Australian National 
20: University, Weston Creek, ACT, 2611, Australia} 
21: \email{propris@mso.anu.edu.au } 
22: 
23:  
24: \author 
25: {S. A. Stanford\altaffilmark{1,2}} 
26: \affil 
27: {Physics Department, University of California at Davis, Davis, CA 95616} 
28: \email{adam@igpp.ucllnl.org} 
29: \altaffiltext{1} 
30: {Institute of Geophysics and Planetary Physics,  Lawrence Livermore National 
31: Laboratories,  Livermore, CA, 94550} 
32:  
33: \author 
34: {Peter R. Eisenhardt\altaffilmark{2}} 
35: \affil 
36: {MS 169-327, Jet Propulsion Laboratory, 
37: California Institute of Technology, 
38: 4800 Oak Grove Drive, Pasadena, CA, 91109} 
39: \email{prme@kromos.jpl.nasa.gov} 
40:  
41: \author 
42: {Mark Dickinson\altaffilmark{2}} 
43: \affil 
44: {Space Telescope Science Institute, 3700 San Martin Drive, Baltimore, MD, 21218} 
45: \email{med@stsci.edu} 
46:  
47:  
48: \altaffiltext{2}{Visiting Astronomer, Kitt Peak National Observatory, National 
49: Optical Astronomy Observatories, which is operated by the Association of 
50: Universities for Research in Astronomy, Inc. (AURA) under cooperative agreement 
51: with the National Science Foundation.} 
52:  
53:  
54: \begin{abstract} 
55:  
56: We investigate the Butcher-Oemler effect using samples of galaxies
57: brighter than observed frame $K^*+1.5$ in 33 clusters at $0.1 \lesssim
58: z \lesssim 0.9$.  We attempt to duplicate as closely as possible the
59: methodology of Butcher \& Oemler.  Apart from selecting in the
60: $K$-band, the most important difference is that we use a brightness
61: limit fixed at 1.5 magnitudes below an observed frame $K^\ast$ rather
62: than the nominal limit of rest frame $M(V) = -20$ used by Butcher \&
63: Oemler. For an early type galaxy at $z=0.1$ our sample cutoff is 0.2
64: magnitudes brighter than rest frame $M(V)=-20$, while at $z=0.9$ our
65: cutoff is 0.9 magnitudes brighter.  If the blue galaxies tend to be
66: faint, then the difference in magnitude limits should result in our
67: measuring lower blue fractions.  A more minor difference from the
68: Butcher \& Oemler methodology is that the area covered by our galaxy
69: samples has a radius of 0.5 or 0.7 Mpc at all redshifts rather than
70: $R_{30}$, the radius containing 30\% of the cluster population. In
71: practice our field sizes are generally similar to those used by
72: Butcher \& Oemler.
73: 
74: We find the fraction of blue galaxies in our $K$-selected samples to
75: be lower on average than that derived from several optically selected
76: samples, and that it shows little trend with redshift.  However, at
77: the redshifts $z < 0.6$ where our sample overlaps with that of Butcher
78: \& Oemler, the difference in $f_B$ as determined from our $K$-selected
79: samples and those of Butcher \& Oemler is much reduced. The large
80: scatter in the measured $f_B$, even in small redshift ranges, in
81: our study indicates that determining the $f_B$ for a much larger 
82: sample of clusters from $K$-selected galaxy samples is important.
83: 
84: As a test of our methods, our data allow us to construct
85: optically-selected samples down to rest frame $M(V) =-20$, as used by
86: Butcher \& Oemler, for four clusters that are common between our
87: sample and that of Butcher \& Oemler.  For these rest $V$ selected
88: samples, we find similar fractions of blue galaxies to Butcher \&
89: Oemler, while the $K$ selected samples for the same 4 clusters yield
90: blue fractions which are typically half as large.  This comparison
91: indicates that selecting in the $K$-band is the primary difference
92: between our study and previous optically-based studies of the Butcher
93: \& Oemler effect.
94: 
95: Selecting in the observed $K$-band is more nearly a process of selecting
96: galaxies by their mass than is the case for optically-selected samples.
97: Our results suggest that the Butcher-Oemler effect is at least partly due to
98: low mass galaxies whose optical luminosities are boosted.  These lower
99: mass galaxies could evolve into the rich dwarf population observed in 
100: nearby clusters.
101:  
102: \end{abstract} 
103: \keywords{galaxies:formation and evolution --- galaxies: clusters} 
104:  
105: \section{Introduction} 
106:  
107: In the late 1970's a variety of new imaging technologies were being 
108: tried out, with an order of magnitude or more better sensitivity than
109: photographic plates. Butcher \& Oemler's (1978) observations of two
110: galaxy clusters at $z \sim 0.4$ using the ISIT Vidicon on the KPNO 2.1~m
111: altered the landscape for studies of distant galaxies, providing the first
112: clear indication of dramatic changes in galaxy properties, and in the 
113: unexpectedly recent past. Butcher \& Oemler (1984; hereafter BO84) confirmed
114: the reality of the effect by addressing a number of concerns about 
115: systematic effects in the analysis procedure, and by broadening the sample
116: both in number and redshift.  The Butcher-Oemler (BO) effect---the discovery 
117: that galaxy clusters at $z > 0.2$ contain a higher fraction of blue 
118: galaxies ($f_B$) than do nearby galaxy clusters---inspired an 
119: empirical approach to galaxy evolution which continues to the present 
120: day. 
121:   
122: The origin of the BO effect is important to discern. Larson, Tinsley, 
123: \& Caldwell (1980) suggested  that S0s could be later type disks transformed 
124: by the cluster environment into earlier types.  The fraction of blue galaxies 
125: in $z \sim 0.4$ clusters is qualitatively similar to the excess of lenticulars 
126: (S0) in nearby clusters relative to those at $z \sim 0.5$ \citep{dre97}, 
127: supporting the idea that infalling field spirals are transformed into S0s. 
128: Morphological transformation from spiral to S0 may be induced by 
129: starbursts occurring upon cluster infall and leading to the rapid 
130: exhaustion of gas reservoirs \citep{dg83,bar96,pog99}, by truncation 
131: of normal star formation after infall \citep{abr96,mor98}, or by a 
132: gradual decline in star formation as the dark haloes surrounding 
133: spirals are removed and gas supplies can no longer be replenished by 
134: cooling and infall \citep{ltc80,bal99}. These mechanisms also provide 
135: natural explanations for the origin of the morphology-density relation 
136: \citep{dre80} and agree with the apparent preference of the blue 
137: galaxies for the outskirts of clusters (BO84). The starburst
138: explanation is supported by spectroscopy of the blue galaxies 
139: indicating that they are seen during or shortly after an episode of 
140: star formation \citep{cs87,co94,co98} (hereinafter CS87, C94, and C98 
141: respectively). Finally, $HST$ images indicate that the blue galaxies 
142: are predominantly normal late-type disks, with some tendency to be 
143: involved in interactions (Lavery \& Henry 1988, Oemler et al. 1997,
144: C94, C98). 
145: 
146: On the other hand, \cite{ros97} and C98 conclude that many of
147: the most actively star forming blue galaxies are actually low luminosity
148: systems, temporarily brightened by starbursts, which will fade from view and
149: evolve not into S0s but into dwarf galaxies. In this bursting dwarf
150: scenario the optical luminosities of such galaxies are boosted to
151: enter the luminosity cut defined by BO84.  This raises the possibility
152: that at least part of the BO effect is a result of photometric
153: selection.  In BO84 selection is carried out in optical passbands,
154: which could magnify the effect of even minor starbursts (even though
155: the BO84 passbands correspond approximately to the rest-frame $V$
156: band).
157: 
158: Our understanding of the physical mechanisms responsible for the BO effect
159: remains contentious.  Part of the problem is simply that there is a wide
160: range in measured blue fractions at a given redshift.  This could be 
161: explained by the larger problem that the cluster samples usually used 
162: in studies of the BO effect do not consist of the same kind of clusters 
163: over the redshift range of interest.  Andreon \& Ettori's (1999) analysis 
164: of the clusters in BO84 using X-ray data indicate that the latter's 
165: results are biased.  They found that the $L_x$ of the BO84 sample 
166: increases with redshift, whereas there is no evidence for evolution in 
167: the X-ray luminosity function up to $z \sim 0.8$.  When \cite{ae99} add 
168: X-ray selected clusters to the BO84 sample, the trend of the blue 
169: fraction with redshift is much reduced.  \cite{maca00} found similar 
170: results, underscoring the need for a study of the BO effect using a 
171: well-defined sample of clusters, over a large redshift range.  Further
172: illustrating the importance of sample selection, \cite{margo01} found
173: a strong correlation between cluster richness and the blue fraction, in
174: the sense that $f_B$ is higher for poorer clusters, using a large
175: sample of 295 Abell clusters. 
176: 
177: Selection of galaxy samples in the near-infrared as opposed to the 
178: optical should result in samples more representative by stellar {\it mass} 
179: \citep{aes91,gpb96,sed98}. \cite{sed98} and \cite{dep99} have shown that
180: the luminosities and colors of galaxies selected in the $K$ band are not
181: strongly dependent on their environment. Therefore $K$-selected samples 
182: can be used to constrain the masses of the blue galaxies causing the BO 
183: effect and investigate if these objects evolve into S0's or dwarfs. 
184: 
185: We present here a study of the BO effect in distant clusters using 
186: galaxies selected in the $K$ band. Our sample, observations and data
187: reduction are described in the next section. The analysis and the main
188: results are presented in Section~\ref{secana}. We discuss our findings
189: in Section~\ref{secdisc}. For consistency with BO84, we adopt a 
190: cosmology with $H_0 = 50$ km s$^{-1}$ Mpc$^{-1}$ and $q_0=0.1$. 
191: 
192: \section{Data} \label{secobs} 
193: 
194: Our sample is the heterogeneous set of clusters which were studied for luminosity 
195: function evolution by \cite{dep99}.  For this sample, we have $JK$ and 
196: two optical bands of imaging which reach at least to 1.5 magnitudes below an 
197: evolving $K^\ast$ at the 5$\sigma$ level.  The optical bands change with 
198: redshift, bracketing the 4000 ${\rm \AA}$ break as closely as possible. 
199: Catalogs for these clusters were generated from the $K$-band images. 
200: The observations, data reduction, and photometry for this sample are 
201: presented in \cite{sta02}. 
202: 
203: \section{Analysis} \label{secana}
204: 
205: In our analysis we are attempting to duplicate the methodology of BO84 
206: as closely as possible.  To that end, we begin with a summary of the 
207: cluster sample and procedures used by BO84.  Their sample consisted of 
208: 33 clusters spanning the range $0.003\ {\rm (Virgo)} < z < 0.54$ drawn 
209: mostly from the Abell catalog.  The data used by BO84 came from three main 
210: sources.  Except for CL0016+16 (photographic plates were obtained by 
211: Koo 1981), the most distant clusters were observed with the ISIT 
212: Vidicon camera on the Kitt Peak 2.1~m telescope in the $V$ and $R$ 
213: bands covering areas of $\sim$6.2 arcmin$^2$ down to a limit of $R 
214: \sim 22.0$.  The photometry for their intermediate redshift clusters 
215: was obtained using $J$ and $F$ plates on the 4~m telescopes at KPNO 
216: and CTIO and covered areas of 55 arcmin$^2$ down to a red magnitude of 
217: 22.  The photometry for the low redshift clusters was obtained from 
218: $J$ and $F$ plates done at the Palomar 1.2~m Schmidt and covered areas 
219: of radius 1.5 Mpc. 
220: 
221: BO84 defined `blue' galaxies as (i) objects lying within a radius 
222: containing 30\% of the cluster population (R$_{30}$); (ii) brighter 
223: than a no-evolution $M_V=-20$; and (iii) bluer by 0.2 magnitudes in 
224: rest frame $B-V$ than the red sequence defined by the cluster E/S0 
225: galaxies.  BO84 derived the slope and intercept of the color-magnitude 
226: relation from their data or from the unevolved slope of the relation in 
227: nearby clusters \citep{sv78} and calculated their $B-V$ color offset 
228: using spectral energy distributions from \cite{cww80} and assuming no 
229: evolution. 
230: 
231: Because of the smaller size of infrared detectors, we are generally 
232: unable to cover enough field to trace the surface density distribution 
233: of galaxies and derive $R_{30}$ directly from our data. Therefore we 
234: adopt circular apertures of radius 0.5 and 0.7 Mpc centered on the 
235: brightest cluster galaxy.  The $R_{30}$ listed in Table 1 of BO84 are 
236: between $\sim$1 and $\sim$5 arcmin with an average of 2.2 arcmin for 
237: the clusters at $z > 0.2$ (see our Table~\ref{fbcomp} for the $R_{30}$ 
238: of the comparison clusters). A radius $= 0.7$ Mpc is very similar to 
239: this average $R_{30}$ in the redshift range where our sample overlaps 
240: with that of BO84.
241: 
242: As for the magnitude limit for our samples, first we need to convert
243: from the optical band used in BO84 to the $K$-band we are using for
244: galaxy selection. We calculate that $M_V=-20$ corresponds to
245: $M_K=-23.05$ for early-type galaxies in the present epoch, based on
246: data from the $UBVRIzJHK$ survey of the Coma cluster by \cite{eis02a}.
247: Our data on the distant clusters do not reach $M_K=-23.05$ in all
248: cases.  We choose to use a magnitude cut at $K^*+1.5$, where $K^*$ has
249: been determined from our data on each cluster \citep{dep99}.
250: Table~\ref{tab1} gives these $K^\ast$, along with both the limiting
251: $K$ actually used with our data in calculating blue fractions, the
252: observed frame $K$ magnitude that is equivalent to the nominal
253: $M(V)=-20.0$ for an elliptical galaxy assuming pure $k-$correction
254: (using the models of Poggianti 1997) of the observed colors of Coma
255: early-types, and the observed frame $K$ equivalent to the {\it actual}
256: limits in $M(V)$ used by BO84 (see Table~\ref{fbcomp}). For $0.1 < 
257: z < 0.3$ our limiting magnitudes are somewhat brighter than those used 
258: in BO84, a point we will return to later.  While different
259: from the fixed absolute magnitude cut used by BO84, our variable
260: magnitude limit accounts for the passive evolution seen in cluster
261: galaxies (e.g., Stanford et al. 1998).  The single absolute magnitude limit
262: used by BO84 translates into a cut relative to $M^\ast$ that changes
263: with redshift, thus possibly including variable proportions of giant
264: and dwarf galaxies into their samples, whereas our magnitude limit
265: is more likely to sample similar populations at all redshifts.
266: 
267: To determine which galaxies in our samples defined by the magnitude
268: and area limits described above are blue we follow the methodology of
269: BO84.  First we correct the galaxy colors for the color-magnitude
270: correlation of E/S0 galaxies by fitting the optical and optical-IR
271: colors with a robust linear least squares routine with brightest
272: cluster galaxies excluded from the fits. This algorithm minimizes
273: least absolute deviation (rather than its square) with iterative
274: rejection (Applied Statistics algorithm \# 132) and therefore reduces
275: the influence of outliers. We carry out this fit within the 0.5 Mpc
276: region to maximize the strength of the cluster red sequence.  We
277: inspect the fit by eye and, occasionally, intervene manually to
278: produce a more acceptable relation (i.e. one where the mode of the
279: marginalized color distribution is as close as possible to 0).  The
280: fit should be done only to the early-type galaxies but we do not have
281: morphological information on all the clusters in our sample.  The
282: color-magnitude diagrams and the best fits are shown in Figure 1 for
283: both the optical-IR and the pure optical colors.
284: 
285: In order to set a color boundary that defines the blue galaxies for
286: our chosen passbands, we need to transform the rest frame $(B-V)=0.2$
287: offset into a difference in the observed colors ($B-R$, $g-R$, $V-I$,
288: $R-I$, $R-K$ and $I-K$ as described below) at the redshift of each
289: cluster. We follow the procedure outlined by BO84 and \cite{sma98} and
290: use E, Sa and Sc spectral energy distributions from \cite{pog97}.  We
291: find that a mixture of 55\% Sa and 45\% Sc yields a spectrum with a
292: color difference $\Delta(B-V)=0.2$ from an E model.  We use this mixed
293: spectrum to calculate the corresponding $\Delta\ (color)$ (for our
294: observed colors) at $z=0$. We then use the $k-$ corrections given in
295: \cite{pog97} (for no-evolution models) to calculate a color difference
296: as a function of redshift (i.e., we compute the no-evolution colors
297: for the E model and for the Sa+Sc mixture). For example, the
298: difference corresponding to an offset of $(B-V)=0.2$ in the observed
299: $V-I$ at $z=0$ is 0.21 mag. For a solar metallicity elliptical in
300: e.g.\ GHO1601+4253 the $k-$correction in $V$ is 1.77 mag and 0.41 in
301: $I$ (Poggianti 1997); for the composite Sa+Sc spectrum the
302: $k-$correction is 1.24 in $V$ and 0.21 in $I$. Therefore the
303: $k-$corrected difference corresponding to $(B-V)=0.2$ is 0.54
304: magnitudes in $V-I$ at $z=0.54$.
305: 
306: We remove contamination from field galaxies by using the SPICES survey
307: \citep{spices2,eis02b}. SPICES is an imaging survey of four fields
308: (Cetus, Lynx, Pisces, and SA57) in $BRIzJK$.  Objects in these fields
309: were selected in the $K$ band down to $K = 20.0$, where the
310: completeness level is $80$\%. These objects were photometered in the
311: same manner as was done for the clusters in our sample \citep{sta02}.
312: Before being used for background correction, color distributions for
313: the field galaxy samples are corrected for the color-magnitude
314: relations derived for each cluster.  We have no $g$ or $V$ data in the
315: SPICES field survey and so need to interpolate magnitudes for those
316: bands for the clusters which have $g$ or $V$ data.  As part of the
317: analysis of SPICES data, photometric redshifts have been determined by
318: A.\ Connolly using the methods of \cite{cs99} and \cite{csa00}.  This
319: process yields spectral types as well photometric redshifts. Using
320: these spectral types and the \cite{cww80} spectral energy
321: distributions, we may calculate colors in any band for the objects in
322: the SPICES catalogs.  As a test of this procedure, we show a
323: comparison of interpolated and measured colors in Figure~\ref{fig2}. 
324: The average of the differences amounts to only a few hundredths of 
325: a magnitude with an rms varying from 0.15 to 0.3 mags (as shown in 
326: the figure).  This indicates that the interpolated $g$ and $V$ band 
327: colors of the field galaxies in the SPICES sample are reasonably accurate. 
328: The above rms values are typically 2--3 times smaller than the actual 
329: offsets in the observed colors that are used in determining which are 
330: the blue galaxies when calculating Butcher-Oemler fractions.
331: 
332: The marginal color distributions for the clusters and the field
333: sample, normalized by the relative areas and with the appropriate
334: magnitude limits, are shown in Figure 3.  The partial galaxies in the
335: field histograms are due to the normalization to the area and
336: magnitude limit of each cluster. As we state above, we have checked
337: that the mode of the color distributions are close to 0 as
338: expected. This does not always work perfectly both because measurement
339: scatter in the colors tends to place fainter objects towards the
340: red, and because field galaxies are in the color-magnitude diagrams
341: biasing the fits to the cluster color-mag sequences.  The
342: distributions for M0906 are a special case: note they have modes
343: which are significantly different from 0.  This
344: appears to be due to the fact that this cluster is composed of two
345: clumps in redshift space, each with its own color-magnitude relation
346: \citep{ell01}.  
347: 
348: The arrow in each panel of Figure 3 indicates the color
349: corresponding to $(B-V)_{rest}= $0.2 mag bluer than the early-type
350: c--m relation. The blue fraction is defined as the ratio $N_b/N$,
351: where $N_b$ is the field-corrected number of galaxies bluer than the
352: BO84 color limit, and $N$ is the total number of cluster members (in a
353: statistical sense, after subtraction of the field population).  The
354: error in this quantity is derived from
355: 
356: $${\sigma^2 (f_B) \over f_B^2} = \Big({\sigma^2 N_b \over N^2_b}\Big)+
357: \Big({\sigma^2 N \over N^2}\Big)$$
358: 
359: and
360: 
361: $$N = N{\rm (all)} - N{\rm (field)}$$
362: $$N_b = N_b {\rm (all)} - N{\rm (field)}$$
363: 
364: \noindent where $N{\rm (all)}$ and $N_{\rm(field)}$ are Poisson variables
365: (including an extra contribution from clustering which is computed 
366: from the four separate background fields of SPICES) and $N_b{\rm(all)}$ and $N_b
367: {\rm(field)}$ are binomial variables. 
368: The correction for field contamination is statistical which means that
369: the corrections can be too large resulting, on some
370: occasions, in blue fractions that are negative. In these cases we have
371: indicated that the value is an upper limit in Figures~\ref{fig4} and
372: \ref{fig5} below.
373: 
374: We plot blue fractions as a function of redshift for the 0.7 Mpc radii
375: for both optical-infrared and optical-optical colors in
376: Figure~\ref{fig4}, and compare our results for the optical colors with
377: the previous compilations by BO84 and \cite{rs95} in
378: Figure~\ref{fig5}. The derived $f_B$'s for our $K$-selected samples
379: are tabulated in Table~\ref{tabn2}. A fit to our data shows that $f_B$
380: is about 10\% at all redshifts we consider (albeit with
381: large scatter).  
382: 
383: To see if there are any systematic differences between our blue
384: fractions and those reported by BO84, we have attempted to make direct
385: comparisons with BO84.  For this purpose, there are four clusters in
386: both our sample and that of BO84 for which we have adequate data to
387: make useful comparisons: Abell 1942, CL0024+1654, 3C~295, and
388: CL0016+16.  For these clusters, we have selected samples in both the
389: $K$-band and also in an optical band, $R$ or $I$, depending on the
390: cluster redshift.  Our samples are chosen over the same areas as in
391: BO84, and to the same magnitude limits actually used by BO84, to the
392: extent that these magnitude limits can be determined from the
393: literature.  The areas and magnitude limits for these comparisons are
394: summarized, along with the resulting blue fractions, in
395: Table~\ref{fbcomp}.  For 3 of the 4 clusters we find blue fractions in
396: our optically-selected samples similar to those calculated by BO84.
397: The exception is CL0016+16, the highest redshift cluster in BO84 at $z
398: = 0.54$, for which we determined a significantly higher value of
399: $f_B$.  However, the $f_B = 0.02 \pm 0.07$ found by BO84 is far below
400: the value of $f_B$ predicted by the Butcher-Oemler effect for the
401: cluster's redshift.  For all four clusters the $f_B$ that we found in
402: our $K$-selected sample in this comparison are lower than those found
403: from the optically-selected samples, even when the area and effective
404: magnitude limits are the same.  So these tests indicate that selecting
405: in the $K$-band is the most important factor determining the generally
406: lower values of $f_B$ that we find from our $K$-band selected samples.
407: 
408: \section{Discussion} \label{secdisc} 
409: 
410: We find that: (i) our infrared selected blue fractions are generally
411: lower than optically selected $f_B$ and (ii) our data show no strong
412: trend in $f_B$ with redshift.  However, at the redshifts $z < 0.6$
413: where our sample overlaps with that of Butcher \& Oemler, there is
414: little if any significant difference in $f_B$ as determined from our
415: $K$-selected samples and those of Butcher \& Omeler, given the larger
416: scatter in the $f_B$ found both by us and by Butcher \& Oemler.  
417: 
418: These points need further clarification. Our sample of clusters is 
419: heterogeneous and is somewhat biased to rich clusters at higher 
420: redshifts, for which blue fractions could be lower, as these objects 
421: are more likely to be more dynamically evolved. \cite{ae99} show that 
422: the BO effect weakens when X-ray selected clusters are added to the set 
423: of clusters in BO84. However, we have shown that the same degree of 
424: passive evolution is present in at least the early-type galaxies in our 
425: sample \citep{sed98,dep99} independent of the X-ray luminosity of the 
426: cluster. 
427: 
428: The $K$-band selection procedure adopted will identify galaxies with
429: colors of Sb's or later types, as long as they are above our magnitude
430: limit (i.e.\ are sufficiently massive).  At the redshifts we are
431: considering, the $R$ band passes through the rest frame $B$ and $U$
432: passbands, whereas $I$ samples the $V$ and $B$ bands, so our optical
433: colors should be sensitive even to minor episodes of star formation in
434: otherwise old populations.  We show a comparison of blue fractions as
435: derived in the optical-$K$ and optical-only colors in
436: Figure~\ref{fig6}.  There is no evidence of large systematic
437: differences in the derived blue fractions in these colors.  On average
438: the difference in blue fractions between optical-$K$ and purely
439: optical colors is $0.07 \pm 0.10$. Our $K$ selected sample is roughly
440: equivalent to selection by stellar mass and therefore our results
441: indicate that the star formation likely causing the BO effect 
442: takes place among relatively low mass objects as well as in normal
443: spirals \cite{dressler94}, \cite{oe97}.
444: 
445: A plausible interpretation of our results is that there are two
446: components to the classical BO effect.  One component, which is seen
447: by our $K$-selected samples giving rise to the $\sim$10\% values of
448: $f_B$ that we find, is the massive galaxies, with colors approximately 
449: equivalent to those of Sb's, that are present at a roughly constant 
450: fraction at all redshifts; these objects are likely to be field spirals 
451: whose star formation history is modified by infall into the cluster. 
452: The second component, which is missing from our infrared selected data, is
453: responsible for both the larger blue fractions of optically-selected
454: samples and their redshift dependence. If this is correct, these
455: objects have $K-$band luminosities lower than $K^* + 1.5$ and are
456: likely to be a population of starbursting dwarfs (or low luminosity
457: spirals), subject to luminosity boosting in the optical bands
458: \citep{bar96}. The first population would be identified with the
459: `normal' spirals of C94 and C98, whereas the dwarfs would constitute
460: the more extreme starburst and post-starburst objects.  These dwarfs
461: are unlikely to evolve into S0's and instead contribute to the rich
462: populations of dwarf galaxies observed in nearby clusters.  As dwarfs
463: are sampled from the more steeply rising power-law regime of the
464: luminosity function ($\alpha \sim -1.3$), the exact placement of an
465: optical magnitude limit with respect to $M^\ast$ strongly affects the
466: number of dwarfs included in a sample and thus can result in a large
467: increase in observed $f_B$. This is qualitatively consistent with the
468: results of C98, where the greater part of the blue population
469: corresponds to relatively low mass late type spirals and irregulars.
470: 
471: There is considerable evidence that low luminosity galaxies show 
472: greater star forming activity at moderate redshifts.  The Canada-France 
473: Redshift Survey detected a population of bright objects with strong OII 
474: emission at $z > 0.2$ \citep{ham97}. The blue excess field population 
475: at $z=0.5$ consists of dwarfs (kinematically: Mall\'en-Ornelas et al. 1997).
476: The luminosity function of dwarfs in two distant clusters, MS
477: 2255.7+2039 at $z=0.29$ \citep{nfg00} and CL1601+54 at $z=0.54$
478: \citep{dfn01}, appears to steepen with redshift, suggesting a
479: brightening of the dwarf population associated with star formation
480: episodes. Such objects would then fade to become part of the rich dwarf
481: populations observed in nearby clusters \citep{wil97}. This is
482: consistent with the `harassment' scenario of \cite{mlk98} whereby low
483: mass spirals are transformed into spheroidals via interactions with the
484: cluster tidal field. A possible interpretation then would account for
485: the Butcher-Oemler effect as a cluster counterpart of the faint blue
486: field galaxies, coupled with an increase in cluster infall at larger
487: redshift, as in the scenario presented by \cite{ell01}.
488: 
489: Although our data only provide information on the bright, massive
490: component of the BO effect, we observe that there is no redshift at
491: which the blue fractions peak strongly in Figure~\ref{fig4}. If the
492: blue galaxies in our $K$-selected samples are mostly normal spirals 
493: falling into clusters, this suggests that there is no preferred epoch 
494: of cluster merging and that spiral and lenticular populations may be 
495: built up gradually by infall of small groups at least since $z \sim 1$. 
496: 
497: To conclude, we need to caution the reader that our results are based on
498: a heterogeneous sample of objects, which is likely to contain significant
499: observational biases. One obvious example is that the fit in Fig.~5 is
500: weighed to low blue fraction by the small number of clusters at $z > 0.7$
501: which have low $f_B$. If we exclude these from our fit, we still find a
502: shallower slope than BO84 but the discrepancy is much reduced. The higher
503: redshift clusters are those most likely to be massive and to have low
504: blue fractions because more highly evolved. This points to the the need
505: to acquire more data, over wider fields of view and to fainter limits,
506: and for a larger sample of clusters at high redshift, in order to confirm
507: the results presented here.
508: \acknowledgments 
509: 
510: The authors would like to thank NOAO for a generous allocation of 
511: observing time to this project, and the staffs at Kitt Peak and Cerro 
512: Tololo for their help with the observing. We wish to thank Warrick J. Couch 
513: for having read the manuscript and substantially improved it by 
514: his comments and advice. We also would like to thank the referee,
515: Augustus E. Oemler, for his careful and cooperative reports which
516: have made this paper substantially better. This research has made use 
517: of the NASA/IPAC Extragalactic Database (NED) which is operated by the 
518: Jet Propulsion Laboratory, California Institute of Technology, under 
519: contract with the National Aeronautics and Space Administration. 
520: Support for this work was provided by NASA through grant number 
521: AR--5790.02--94A from the Space Telescope Science Institute, which is 
522: operated by the Association of Universities for Research in Astronomy, 
523: Inc., under NASA contract NAS5-26555.   Portions of the 
524: research described here were carried out at the Jet Propulsion 
525: Laboratory, California Institute of Technology, under a contract with 
526: NASA.  Work performed at the Lawrence Livermore National Laboratory is 
527: supported by the DOE under contract W7405-ENG-48. 
528: 
529: \begin{thebibliography}{100} 
530: \expandafter\ifx\csname natexlab\endcsname\relax\def\natexlab#1{#1}\fi 
531: 
532: \bibitem[{{Abraham} {et~al.}(1996)}]{abr96} 
533: {Abraham}, R. {et~al}, 1996, \apj, 471, 694 
534: 
535: \bibitem[{{Andreon} \& {Ettori}(1999)}]{ae99} 
536: {Andreon}, S., \& {Ettori}, S. 1999, \apj, 516, 647 
537: 
538: \bibitem[{{Aragon-Salamanca} {et~al.}(1991){Aragon-Salamanca}, {Ellis}, 
539: \& {Sharples}}]{aes91} 
540: {Aragon-Salamanca}, A., {Ellis}, R.~S. \& {Sharples}, R.~M. 1991, \mnras, 
541: 248, 128 
542: 
543: \bibitem[{{Balogh} {et~al.}(2000){Balogh}, {Navarro}, \& {Morris}}]{bal99} 
544: {Balogh}, M. L., {Navarro}, J.~F., \& {Morris}, S.~L. 1999, \apj, 527, 54 
545: 
546: \bibitem[{{Barger} {et~al.}(1996){Barger}, {Aragon-Salamanca}, {Ellis}, 
547: {Couch}, {Smail}, \& {Sharples}}]{bar96} 
548: {Barger}, A.~J., {Aragon-Salamanca}, A., {Ellis}, R.~S., {Couch}, W.~J., 
549: {Smail}, I. \& {Sharples}, R.~M. 1996, \mnras, 279, 1 
550: 
551: \bibitem[{{Butcher} \& {Oemler}(1978)}]{bo78} 
552: {Butcher}, H., \& {Oemler}, A. 1978, \apj, 226, 559 
553: 
554: \bibitem[{{Butcher} \& {Oemler}(1984)}]{bo84} 
555: {Butcher}, H., \& {Oemler}, A. 1984, \apj, 285, 426 
556: 
557: \bibitem[{{Coleman} {et~al.}(1980){Coleman}, {Wu}, \& {Weedman}}]{cww80} 
558: {Coleman}, G.~D., {Wu}, C.-C., \& {Weedman}, D.~W. 1980, \apjs, 43, 393 
559: 
560: \bibitem[{{Connolly} \& {Szalay}(1999)}]{cs99} 
561: {Connolly}, A.~J. \& {Szalay}, A.~S. 1999, \aj, 117, 2052 
562: 
563: \bibitem[{{Couch} \& {Sharples}(1987)}]{cs87} 
564: {Couch}, W.~J., \& {Sharples}, R.~M. 1987, \mnras, 229, 483 
565: 
566: \bibitem[{{Couch} {et~al.}(1994){Couch}, {Ellis}, {Sharples}, \& 
567: {Smail}}]{co94} 
568: {Couch}, W.~J., {Ellis}, R.~S., {Sharples}, R.~M. \& {Smail}, I. 1994, \apj, 
569: 430, 121 
570: 
571: \bibitem[{{Couch} {et~al.}(1998){Couch}, {Barger}, {Smail}, {Ellis}, \& 
572: {Sharples}}]{co98} 
573: {Couch}, W.~J., {Barger}, A.~J., {Smail}, I., {Ellis}, R.~S. \& {Sharples}, 
574: R.~M. 1998, \apj, 497, 188 
575: 
576: \bibitem[{{Csabai} {et~al.}(2000){Csabai}, {Connolly}, {Szalay}, \& 
577: {Bud\'avari}}]{csa00} 
578: {Csabai}, I., {Connolly}, A.~J., {Szalay}, A.~S. \& {Bud\'avari}, T. 2000, 
579: \aj, 119, 69 
580:   
581: \bibitem[{{D\'ahlen} {et~al.}(2001){D\'ahlen}, {Fransson}, \& 
582: {N\"aslund}}]{dfn01} 
583: {D\'ahlen}, T., {Fransson}, C. \& {N\"aslund}, M. 2001, \mnras, 330, 167 
584: 
585: \bibitem[{{De~Propris} {et~al.}(1999){De~Propris}, {Stanford}, {Eisenhardt}, 
586: {Dickinson}, \& {Elston}}]{dep99} 
587: {De~Propris}, R., {Stanford}, S.~A., {Eisenhardt}, P.~R., {Dickinson}, M., \& 
588: {Elston}, R. 1999, \aj, 118, 719 
589: 
590: \bibitem[{{Dressler}(1980)}]{dre80} 
591: {Dressler}, A. 1980, \apj, 236, 351 
592: 
593: \bibitem[{{Dressler} \& {Gunn}(1983)}]{dg83} 
594: {Dressler}, A. \& {Gunn}, J.~E. 1983, \apj, 270, 7 
595: 
596: 
597: \bibitem[{{Dressler} {et~al.}(1994){Dressler}, {Oemler}, {Butcher}, \&
598: {Gunn}}]{dressler94}{Dressler}, A., {Oemler}, A., {Butcher}, H., \& 
599: {Gunn}, J.E.\ 1994, \apj, 430, 107
600: 
601: \bibitem[{{Dressler} {et~al.}(1997)}]{dre97} 
602: {Dressler}, A. {et~al.} 1997, \apj, 490, 577 
603: 
604: \bibitem[{{Eisenhardt} {et~al.}(2003){Eisenhardt}, {De~Propris}, {Gonzales}, 
605: {Stanford}, {Dickinson} \& {Wang}}]{eis02a} 
606: {Eisenhardt}, P. R., {De~Propris}, R., {Gonzales}, A., {Stanford}, S.~A., 
607: {Dickinson}, M. \& {Wang}, M., in preparation 
608: 
609: \bibitem[{{Eisenhardt} {et~al.}(2003){Eisenhardt}, {Elston},
610: {Stanford}, {Stern}, {Wu}, {Connolly} \& {Spinrad}}]{eis02b} 
611: {Eisenhardt}, P.~R., {Elston}, R., {Stanford}, S.~A.,
612: {Stern}, D., {Wu}, K.~L., {Connolly}, A.~J. \& {Spinrad}, H.
613: 2003, in preparation 
614: 
615: \bibitem[{{Ellingson} {et~al.}(2001){Ellingson}, {Lin}, {Yee} \& {Carlberg}}]
616: {ell01}
617: {Ellingson} E., {Lin} H., {Yee} H. K. C. \& {Carlberg} R. 2001, \apj, 547, 609
618: \bibitem[{{Gavazzi} {et~al.}(1996){Gavazzi}, {Pierini}, \& {Boselli}}]{gpb96} 
619: {Gavazzi}, G., {Pierini}, A. \& {Boselli}, D. 1996, \aap, 312, 297 
620: 
621: \bibitem[{{Hammer} {et~al.}(1997)}]{ham97} 
622: {Hammer}, F. {et~al.} 1997, \apj, 481, 49 
623: 
624: \bibitem[{{Kauffmann}(1995)}]{kau95} 
625: {Kauffmann}, G. 1995, \mnras, 274, 161 
626: 
627: \bibitem[{{Koo}(1981)}]{koo81}Koo, D.\ 1981, \apj, 251, L75
628: 
629: \bibitem[{{Larson} {et~al.} (1980){Larson}, {Tinsley}, \& {Caldwell}}]{ltc80}
630: {Larson}, R.~B., Tinsley, {B.~M.} \& {Caldwell}, C.~N. 1980, ApJ, 237, 692 
631: 
632: \bibitem[{{Lavery} \& {Henry}(1988)}]{lh88} 
633: {Lavery}, R.~J., \& {Henry}, J.~P. 1988, \apj, 330, 596 
634: 
635: \bibitem[{{Mall\'en-Ornelas} {et~al.}(1999){Mall\'en-Ornelas}, {Lilly}, 
636: {Crampton} \& {Schade}}]{mao99} 
637: {Mall\'en-Ornelas}, G., {Lilly}, S.~J., {Crampton}, D. \& {Schade}, D. 
638: 1999, \apj, 518, L83 
639: 
640: \bibitem[{{Margoniner} \& {de~Carvalho}(2000)}]{maca00} 
641: {Margoniner}, V.~E., \& de Carvalho, R.~R. 2000, \aj, 119, 1562 
642: 
643: \bibitem[{{Margoniner} {et~al.}(2001){De Carvalho}, {Gal}, \&
644: {Djorgovski}}]{margo01}{Margoniner}, V.E., {De Carvalho}, R.R., {Gal},
645:   R., \& {Djorgovski}, S.G. 2001, \apj, 548, L143 
646: 
647: \bibitem[{{Metevier} {et~al.}(2000){Metevier}, {Romer}, \& {Ulmer}}]{mru00} 
648: {Metevier}, A.~J., {Romer}, A.~K. \& {Ulmer}, M.~P. 2000, \aj, 119, 1090 
649: 
650: \bibitem[{{Moore} {et~al.}(1998){Moore}, {Lake}, \& {Katz}}]{mlk98} 
651: {Moore}, B., {Lake}, G., \& {Katz}, N. 1998, \apj, 495, 139 
652: 
653: \bibitem[{{Morris} {et~al}(1998){Morris}, {Hutchings}, {Carlberg}, {Yee}, {Ellingson}, {Balogh}, {Abraham}, \& {Smecker-Hane}}]{mor98} 
654: {Morris}, S.~L., {Hutchings} J.~B., {Carlberg}, R.~G., {Yee}, H.~K.~C., 
655: {Ellingson}, E., {Balogh}, M.~L., {Abraham}, R.~G., {Smecker-Hane}, T.~A. 
656: 1998, \apj, 507, 84 
657: 
658: \bibitem[{{N\"aslund} {et~al.}(2000){N\"aslund}, {Fransson}, \& 
659: {Huldtgren}}]{nfg00} 
660: {N\"aslund}, M., {Fransson}, C., \& {Huldtgren}, M. 2000, \aap, 356, 435 
661: 
662: \bibitem[{{Oemler} {et~al.}(1997){Oemler}, {Dressler}, \& {Butcher}}]{oe97} 
663: {Oemler}, A., {Dressler}, A. \& {Butcher}, H. 1997, \apj, 474, 561 
664: 
665: \bibitem[{{Poggianti}(1997)}]{pog97} 
666: {Poggianti}, B.~M. 1997, \aaps, 122, 399 
667: 
668: \bibitem[{{Poggianti} {et~al.}(1999){Poggianti}, {Smail}, {Dressler}, {Couch}, 
669: {Barger}, {Butcher}, {Ellis}, \& {Oemler}}]{pog99} 
670: {Poggianti}, B,~M., {Smail}, I., {Dressler}, A., {Couch}, W.~J., {Barger}, 
671: A.~J.,{Butcher}, H., {Ellis}, R.~S. \& {Oemler}, A. 1999, \apj, 578, 516 
672: 
673: \bibitem[{{Rakos} \& {Schombert}(1995)}]{rs95} 
674: {Rakos}, K.~D., \& {Schombert}, J.~M. 1995, \apj, 439, 47 
675: 
676: \bibitem[{{Rakos} {et~al.}(1997){Rakos}, {Odell} \& {Schombert}}]{ros97} 
677: {Rakos}, K.~D., {Odell}, A.~D. \& {Schombert}, J.~M. 1997, \apj, 490, 194 
678: 
679: \bibitem[{{Sandage} \& {Visvanathan}(1978)}]{sv78} 
680: {Sandage}, A. \& {Visvanathan}, N. 1978, \apj, 223, 707 
681: 
682: \bibitem[{{Smail} {et~al.}(1998){Smail}, {Edge}, {Ellis}, \& 
683: {Blandford}}]{sma98} 
684: {Smail}, I., {Edge}, A.~C., {Ellis}, R.~S. \& {Blandford}, R.~D. 1998, \mnras, 
685: 293, 124 
686: 
687: \bibitem[{{Stanford} {et~al.}(1998){Stanford}, {Eisenhardt}, \& 
688: {Dickinson}}]{sed98} 
689: {Stanford}, S.~A., {Eisenhardt}, P.~R., \& {Dickinson}, M. 1998, \apj, 492, 461 
690: 
691: \bibitem[{{Stanford} {et~al.}(2002){Stanford}, {Eisenhardt}, {Dickinson}, 
692: {Holden}, \& {De~Propris}}]{sta02} 
693: {Stanford}, S.~A., {Eisenhardt}, P.~R., {Dickinson}, M., {Holden}, B.~P., 
694: \& {De~Propris}, R. 2002, \apjs, 142, 153
695: 
696: \bibitem[{{Stern} {et~al.}(2002)}]{spices2}Stern, D.\ et al. 2002, 
697: \aj, 123, 2223
698: 
699: \bibitem[{{Wilson} {et~al.}(1997){Wilson}, {Smail}, {Ellis}, \& {Couch}}]{wil97} 
700: {Wilson}, G., {Smail}, I., {Ellis}, R.~S., \& {Couch}, W.~J. 1997, \mnras, 284, 915 
701: 
702: \end{thebibliography} 
703: 
704: \clearpage 
705: 
706: \begin{deluxetable}{ccccc} 
707: \tablewidth{0pt} 
708: \tablecaption{Reference magnitudes vs. $z$} 
709: \tablehead{ 
710: \colhead{Redshift} & 
711: \colhead{$K_*$\tablenotemark{a}} & 
712: \colhead{$K_{lim}$\tablenotemark{b}} & 
713: \colhead{$K(BO)$\tablenotemark{c}} & 
714: \colhead{actual $K(BO)$\tablenotemark{d}} 
715: } 
716: 
717: \startdata 
718: 0.15 & $14.84 \pm 0.49$ & 16.3 & 16.5 & 16.5 \\ 
719: 0.20 & $15.16 \pm 0.07$ & 16.7 & 17.0 & 17.0 \\ 
720: 0.25 & $15.64 \pm 0.38$ & 17.1 & 17.5 & 17.5 \\ 
721: 0.32 & $15.74 \pm 0.08$ & 17.2 & 18.0 & 18.0 \\ 
722: 0.40 & $16.50 \pm 0.11$ & 18.0 & 18.5 & 17.8 \\ 
723: 0.46 & $16.38 \pm 0.08$ & 17.9 & 18.8 & 17.7 \\ 
724: 0.54 & $16.85 \pm 0.18$ & 18.4 & 19.2 & 18.2 \\ 
725: 0.69 & $17.2\tablenotemark{e}\pm 0.4$ & 18.7 & 19.1 & \nodata \\ 
726: 0.79 & $17.51 \pm 0.26$ & 19.0 & 20.1 & \nodata \\ 
727: 0.90 & $18.05 \pm 0.25$ & 19.5 & 20.4 & \nodata \\ 
728: \enddata 
729: \label{tab1} 
730: \tablenotetext{a}{Measured observed frame $K^\ast$ from De Propris et 
731: al.\ (1999)} 
732: \tablenotetext{b}{$K$ limiting magnitude used in calculating our blue 
733: fractions from our $K$-selected samples} 
734: \tablenotetext{c}{Limit in observed $K$ band corresponding to $M(V)=-20.0$} 
735: \tablenotetext{d}{Limit in observed $K$ corresponding to actual limit 
736: in $M(V)$ used by Butcher \& Oemler (1984)}
737: \tablenotetext{e}{Interpolated value in Figure 8 of De Propris et
738: al.\ (1999)}
739: \end{deluxetable} 
740: \clearpage 
741: 
742: \begin{deluxetable}{lccccccccc}
743: \tabletypesize{\footnotesize}
744: \rotate
745: \tablewidth{0pt}
746: \tablecaption{Comparisons with BO84}
747: \tablehead{
748: \colhead{Cluster} & 
749: \colhead{BO84 $f_B$}& 
750: \colhead{R(30)}& 
751: \colhead{BO84 limit\tablenotemark{a}} &
752: \colhead{N$_{sample}$ (BO84)\tablenotemark{b}} &
753: \colhead{Optical\tablenotemark{c} $f_B$} & 
754: \colhead{IR\tablenotemark{d} $f_B$} & 
755: \colhead{radius} & 
756: \colhead{$K_{lim}$\tablenotemark{e}} &
757: \colhead{N$_{field}$/N$_{sample}$} \\
758: \colhead{} & \colhead{} & \colhead{arcmin} & \colhead{$M(V)$}&\colhead{} & \colhead{} &
759: \colhead{} & \colhead{arcmin} & \colhead{}
760: }
761: 
762: \startdata
763: Abell 1942 & $0.17\pm 0.05$ & 2.8 & -20.0 & 57 & $0.21\pm 0.05$ & $0.16\pm 0.04$ 
764: & 2.2 & 17.3 & 8.7/65\\
765: CL0024+16  & $0.16\pm 0.02$ & 1.1 & -20.8\tablenotemark{f} & 87 & $0.16\pm 
766: 0.02$&$0.08\pm 0.03$ & 1.1 & 17.8 & 5.7/76 \\
767: 3C~295&$0.23\pm 0.05$&1.0&-21.1\tablenotemark{g}& 45 & $0.27\pm 0.10$ &$0.14\pm 
768: 0.04$&1.0&17.8 & 5.5/26 \\
769: CL0016+16&$0.02\pm 0.07$&1.0&-21.0\tablenotemark{h}& 65 & $0.18\pm 
770: 0.06$&$0.09\pm 0.06$&1.0&18.3 & 9/51\\
771: \enddata
772: \label{fbcomp}
773: \tablenotetext{a}{Actual magnitude limit used by BO84}
774: \tablenotetext{b}{Number of galaxies within R(30) in BO84}
775: \tablenotetext{c}{Blue fraction based on our optically-selected
776: catalogs}
777: \tablenotetext{d}{Blue fraction based on our $K$-band selected catalogs}
778: \tablenotetext{e}{$K$ limiting magnitude used in calculating our blue
779: fractions, which for these comparisons was adjusted so as to
780: reach the same limit in $M(V)$ (for a red envelope galaxy in the
781: cluster) as used by BO84}
782: \tablenotetext{f}{Butcher \& Omeler (1978) indicate that the
783: completness limit is $m_R = 21.8$ for their CL0024+16 data.}
784: \tablenotetext{g}{BO84 indicate that the actual limiting magnitude in
785: the rest frame $V$-band in
786: their study for 3C295 was $M_{lim}=-21.1$}
787: \tablenotetext{h}{BO84 used photometry from Koo (1981), apparently
788: down to $F=23.0$ which is equivalent to $M_V=-21$}
789: \end{deluxetable}
790: \clearpage
791: 
792: 
793: \begin{deluxetable}{lcrrrrrr}
794: \tablewidth{0pt}
795: \tabletypesize{\footnotesize}
796: \rotate
797: \tablecaption{Blue Fractions for the complete cluster sample}
798: \tablehead{
799: \colhead{Cluster ID} & \colhead{Redshift} & \colhead{$f_B (R/I-K)$} & 
800: \colhead{$f_B (R/I-K)$} & \colhead{$f_B$ (optical)} & \colhead{$f_B$ (optical)} & 
801: \colhead{$N_{field}/N_{sample}$\tablenotemark{a}} & 
802: \colhead{$N_{field}/N_{sample}$\tablenotemark{a}} 
803: \\
804: \colhead{} & \colhead{} & \colhead{R=0.5 Mpc} & \colhead{R=0.7 Mpc} & 
805: \colhead{R=0.5 Mpc} & \colhead{R=0.7 Mpc} & \colhead{R=0.5 Mpc} & 
806: \colhead{R=0.7 Mpc}
807: }
808: \startdata
809: A1146 & 0.142 & $0.000 \pm 0.000$ & $0.000 \pm 0.000$ & $-0.020 \pm 0.044$ &
810: $-0.024 \pm 0.046$ & 1.7/38 & 3.3/55 \\
811: A3305 & 0.157 & $0.050 \pm 0.056$ & $0.047 \pm 0.051$ & $0.038 \pm 0.058$ & 
812: $0.067 \pm 0.075$ & 4.0/24 & 7.8/29 \\
813: MS0906.5+1110 & 0.180 & $0.027 \pm 0.029$ & $0.018 \pm 0.019$ & $0.029 \pm 
814: 0.045$
815: & $0.077 \pm 0.050$ & 5.3/42 & 10.3/65 \\
816: A1689 & 0.185 & $0.046 \pm 0.038$ & \nodata & $0.029 \pm 0.025$ & \nodata & 
817: 5.2/91 & \nodata \\
818: A1942 & 0.224 & $0.069 \pm 0.053$ & \nodata & $0.054 \pm 0.051$ & \nodata & 
819: 3.8/47 & \nodata \\
820: MS1253.9+0456 & 0.230 & $0.090 \pm 0.055$ & \nodata & $0.112 \pm 0.069$ & 
821: \nodata & 6.7/59 & \nodata \\
822: A1525 & 0.259 & $0.055 \pm 0.048$ & \nodata & $0.051 \pm 0.061$ & \nodata & 
823: 5.7/40 & \nodata\\
824: M1008-1225 & 0.301 & $0.000 \pm 0.010$ & $ 0.015 \pm 0.023$ & $0.047 \pm 0.054$ &
825: $0.085\pm 0.066$ & 5.2/44 & 10.4/60 \\
826: M1147+1103 & 0.308 & $0.036 \pm 0.040$ & $0.023 \pm 0.024$ & $0.112 \pm 0.092$ &
827: $0.071 \pm 0.062$ & 5.2/33 & 10.3/54\\
828: AC118 & 0.308 & $0.073 \pm 0.054$ & $0.073 \pm 0.044$ & $0.123 \pm 0.076$ & 
829: $0.131
830:  \pm 0.062$ & 5.2/52 & 10.3/77 \\
831: AC114 & 0.312 & $0.050 \pm 0.051$ & $0.026 \pm 0.033$ & $0.148 \pm 0.096$ &
832: $0.155 \pm 0.077$ & 5.1/38 & 10.1/60 \\
833: AC103 & 0.313 & $0.107 \pm 0.078$ & $0.175 \pm 0.088$ & $0.082 \pm 0.064$ & 
834: $0.161 \pm 0.088$ & 5.1/39 & 10.0/57\\
835: MS2137-0234 & 0.313 & $0.014 \pm 0.057$ & $0.024 \pm 0.068$ & $0.177 \pm 0.124$ 
836: & $0.260 \pm 0.139$ & 5.1/27 & 10.1/36\\
837: Abell S0506 & 0.316 & $0.029 \pm 0.034$ & $0.069 \pm 0.047$ & $0.077 \pm 0.066$ &
838: $0.120 \pm 0.073$ & 5.0/38 & 9.8/52\\
839: M1358+6245 & 0.328 & $0.046 \pm 0.050$ & $0.084 \pm 0.056$ & $0.054 \pm 0.062$ &
840: $0.096 \pm 0.056$ & 4.8/39 & 9.4/59 \\
841: CL2244-02 & 0.330 & $-0.014 \pm 0.026$ & $-0.022 \pm 0.029$ & $0.007 \pm 0.061$ 
842: & $0.050 \pm 0.075$ & 4.8/26 & 9.4/36 \\
843: CL0024+16 & 0.391 & $0.081 \pm 0.044$ & $0.077 \pm 0.037$ & $0.153 \pm 0.068$ & 
844: $0.200 \pm 0.068$ & 10.2/88 & 20/119 \\
845: GHO 0303+1706 & 0.418 & $0.065 \pm 0.064$ & $0.086 \pm 0.059$ & $0.174 \pm 
846: 0.096$ & $0.140 \pm 0.067$ & 9.5/46 & 18.5/74\\
847: 3C~313 & 0.461 & $0.010 \pm 0.112$ & $0.114 \pm 0.122$ & $-0.070 \pm 0.083$ &
848: $0.120 \pm 0.124$ & 7.3/19 & 14.3/34\\
849: 3C~295 & 0.461 & $0.194 \pm 0.134$ & $0.178 \pm 0.107$ & $0.240 \pm 0.152$ & 
850: $0.184 \pm 0.109$ & 7.3/29 & 14.4/45 \\
851: F1557.19TC & 0.510 & $-0.021 \pm 0.105$ & $-0.186 \pm 0.189$ & $0.143 \pm 0.132$ & 
852: $0.106 \pm 0.190$ & 10.6/29 & 20.7/34 \\
853: GHO 1601+4253 & 0.539 & $0.027 \pm 0.077$ & $0.038 \pm 0.072$ & $0.020 \pm 
854: 0.078$ & $0.117 \pm 0.094$ & 10.1/33 & 19.8/53\\
855: MS0451.6-0306 & 0.539 & $0.121 \pm 0.066$ & \nodata & $0.197 \pm 0.089$ & 
856: \nodata
857: & 10.1/70 & \nodata\\
858: CL0016+16 & 0.545 & $ 0.000 \pm 0.043 $ & $0.049 \pm 0.051 $ & $0.193 \pm 0.098$ & 
859: $0.160 \pm 0.073$ & 10/52 & 19.5/80 \\
860: J1888.16CL & 0.560 & $0.243 \pm 0.142$ & $0.296 \pm 0.130$ & $0.353 \pm 0.181$ & 
861: $0.416 \pm 0.162$ & 9.7/36 & 19.1/59 \\
862: MS2053-0449 & 0.582 & $0.083 \pm 0.116$ & $ 0.089 \pm 0.085$ & $0.226 \pm 0.168$ &
863: $0.314 \pm 0.142$ & 9.4/30 & 18.4/58 \\
864: 3C 34  & 0.689 & $-0.067 \pm 0.092$ & $0.114 \pm 0.109$ & $0.143 \pm 0.153$ & 
865: $0.398 \pm 0.193$ & 9.3/28 & 18.1/47 \\
866: GHO 1322+3027 & 0.751 & $-0.180 \pm 0.188$ & $0.018 \pm 0.152$ & $-0.200 \pm 
867: 0.231$ & $0.089 \pm 0.173$ & 12.9/26 & 25.2/49  \\
868: MS1137.5+6625 & 0.780 & $ -0.006 \pm 0.099 $ & $-0.040 \pm 0.120 $ & $0.014 \pm 
869: 0.121$ & $-0.101 \pm 0.137$ & 12.6/38 & 24.7/55 \\
870: MS1054.5-0327 & 0.820 & $0.052 \pm 0.065$ & $0.105 \pm 0.057$ & $ 0.013 \pm 
871: 0.076$ & $0.009 \pm 0.069$ & 12.2/54 & 24.1/86\\
872: GHO 1603+4313 & 0.895 & $-0.070 \pm 0.191$ & $-0.363 \pm 0.361$ & 
873: $-0.310 \pm 0.236$ & $ -0.987 \pm 0.653$ & 17.1/33 & 33.5/48  \\
874: GHO 1603+4329 & 0.920 & $-0.297 \pm 0.393$ & $-0.342 \pm 0.251$ & $0.050 \pm 
875: 0.190$ & $0.014 \pm 0.144$ & 15.5/24 & 30.4/54 \\
876: 
877: \enddata
878: \label{tabn2}
879: \tablenotetext{a}{$K-$selected}
880: \end{deluxetable}
881: 
882: \clearpage
883: 
884: \begin{figure} 
885: \figurenum{1} 
886: \caption{Color--magnitude diagrams, with best fits to the
887: color-magnitude relation shown by the solid line, in optical-$K$
888: vs. $K$ and in an optical color ($B-R$, $g-$, $V-I$ or $R-I$)
889: straddling the 4000 \AA\ break vs. $K$ for all sample clusters.
890: The data are for the 0.7 Mpc regions described in the text (unless 
891: otherwise specified) and to the magnitude limit referred to in 
892: Table~\ref{tab1}.}  
893: \epsscale{0.9}
894: \plotone{f1a.eps}
895: \label{fig1} 
896: \end{figure} 
897: \clearpage 
898: 
899: \begin{figure} 
900: \figurenum{1} 
901: \caption{continued} 
902: \plotone{f1b.eps} 
903: \end{figure} 
904: \clearpage 
905: 
906: \begin{figure} 
907: \figurenum{1} 
908: \caption{continued} 
909: \plotone{f1c.eps} 
910: \end{figure} 
911: \clearpage 
912: 
913: \begin{figure} 
914: \figurenum{1} 
915: \caption{continued} 
916: \plotone{f1d.eps} 
917: \end{figure} 
918: \clearpage 
919: 
920: \begin{figure} 
921: \figurenum{1} 
922: \caption{continued} 
923: \plotone{f1e.eps} 
924: \end{figure} 
925: \clearpage 
926: 
927: \begin{figure} 
928: \figurenum{1} 
929: \caption{continued} 
930: \epsscale{0.9} 
931: \plotone{f1f.eps}  
932: \end{figure} 
933: \clearpage 
934: 
935: \begin{figure} 
936: \figurenum{1} 
937: \caption{continued} 
938: \plotone{f1g.eps} 
939: \end{figure} 
940: \clearpage 
941: 
942: \begin{figure} 
943: \figurenum{1} 
944: \caption{continued} 
945: \plotone{f1h.eps} 
946: \end{figure} 
947: \clearpage 
948: 
949: \begin{figure} 
950: \figurenum{2} 
951: \caption{Comparison between interpolated and measured colors for 
952: galaxies in the SPICES survey.} 
953: \plotone{f2.eps} 
954: \label{fig2} 
955: \end{figure} 
956: \clearpage 
957: 
958: \begin{figure} 
959: \figurenum{3} 
960: \caption{ 
961: Marginal color distributions in optical-$K$ and an optical color 
962: straddling the 4000 \AA\  break for galaxies in the cluster fields 
963: (open histogram) and in the background fields (hatched histogram). We have 
964: removed the color--magnitude relations from both the cluster and field 
965: datasets and normalized field counts to the areas covered. Here we show the
966: distributions for the 0.7 Mpc areas we use in our analysis (except for a
967: few clusters where our data cover only 0.5 Mpc; these are identified in
968: the figure). The arrow indicates the $k$-corrected color for galaxies to 
969: be considered `blue' as per the definition of BO84 (see text for 
970: details).  A Gaussian is shown centered at 0 in the marginal color
971: with a width specified by the expected scatter in the observed colors
972: due only to measurement uncertainties. 
973: } 
974: \label{fig3} 
975: \plotone{f3a.eps} 
976: \end{figure} 
977: \clearpage 
978: 
979: \begin{figure} 
980: \figurenum{3} 
981: \caption{continued} 
982: \plotone{f3b.eps} 
983: \end{figure} 
984: \clearpage 
985: 
986: \begin{figure} 
987: \figurenum{3} 
988: \caption{continued} 
989: \plotone{f3c.eps} 
990: \end{figure} 
991: \clearpage 
992: 
993: \begin{figure} 
994: \figurenum{3} 
995: \caption{continued} 
996: \plotone{f3d.eps} 
997: \end{figure} 
998: \clearpage 
999: 
1000: \begin{figure} 
1001: \figurenum{3} 
1002: \caption{continued} 
1003: \plotone{f3e.eps} 
1004: \end{figure} 
1005: \clearpage 
1006: 
1007: \begin{figure} 
1008: \figurenum{3} 
1009: \caption{continued} 
1010: \plotone{f3f.eps} 
1011: \end{figure} 
1012: \clearpage 
1013: 
1014: \begin{figure} 
1015: \figurenum{3} 
1016: \caption{continued} 
1017: \plotone{f3g.eps} 
1018: \end{figure} 
1019: \clearpage 
1020: 
1021: \begin{figure} 
1022: \figurenum{3} 
1023: \caption{continued} 
1024: \plotone{f3h.eps} 
1025: \end{figure} 
1026: \clearpage 
1027: 
1028: \begin{figure} 
1029: \figurenum{4} 
1030: \caption{ 
1031: Variation of the blue fraction with redshift in the 
1032: 0.7 Mpc fields in all clusters  for both the optical$-K$ 
1033: (bottom panel) and the optical color (top panel). The upper
1034: limits (downward pointing arrows) indicate clusters for which
1035: the derived f$_B$ is negative because of overcorrection of the
1036: field population.} 
1037: \plotone{f4.eps} 
1038: \label{fig4} 
1039: \end{figure} 
1040: \clearpage 
1041: 
1042: \begin{figure} 
1043: \figurenum{5} 
1044: \caption{Comparison of blue fractions in our optical colors with BO84
1045: and \cite{rs95} for the R=0.7 Mpc fields. Closed symbols are our data;
1046: open circles show BO84's and open squares \cite{rs95}'s data. Upper
1047: limits as for Figure 4 above. The thick solid line is a linear fit to
1048: our data to show the trend in blue fraction with redshift. The thin
1049: dot dashed line is the fit shown by BO84 to their data, extrapolated
1050: to $z=1$. }  
1051: \plotone{f5.eps}
1052: \label{fig5} 
1053: \end{figure} 
1054: \clearpage 
1055: 
1056: \begin{figure} 
1057: \figurenum{6} 
1058: \caption{Differences in blue fractions between optical-only and 
1059: $R/I-K$ colors as a function of redshift for the R=0.7 Mpc 
1060: aperture. The thick solid line is a straight line fit to these 
1061: differences.}   
1062: \plotone{f6.eps} 
1063: \label{fig6} 
1064: \end{figure} 
1065: \clearpage 
1066: 
1067: \end{document} 
1068: