0711.3792/ms.tex
1: \documentclass[useAMS,usenatbib]{mn2e}              
2: \usepackage{epsfig}
3: \bibliographystyle{mn2e} 
4: \citestyle{mn2e} 
5: \input{macro1}
6: 
7: \topmargin -2.0cm
8: 
9: %%%%%%%%%%%
10: \title[Clustering of star-forming  galaxies] 
11: {Interaction-induced star formation in a complete sample of 
12: $10^5$ nearby star-forming galaxies}  
13: \author[Li et al.]
14: {Cheng Li$^{1,2}$\thanks{E-mail: leech@mpa-garching.mpg.de},
15: Guinevere Kauffmann$^{2}$,  
16: Timothy M. Heckman$^{3}$,
17: Y. P. Jing$^{1}$, 
18: \newauthor Simon D. M. White$^{2}$
19: \\  
20: ${^1}$ MPA/SHAO  Joint  Center  for  Astrophysical Cosmology 
21:        at Shanghai Astronomical Observatory, 
22:        Nandan Road 80, Shanghai 200030, China \\ 
23: ${^2}$ Max Planck Institut f\"ur Astrophysik, 
24:        Karl-Schwarzschild-Strasse 1, 85748 Garching, Germany \\
25: ${^3}$ Department of Physics and Astronomy,
26:        Johns Hopkins University, Baltimore, MD 21218
27: }
28: 
29: 
30: 
31: %%%%%%%%%%%%%%%%%%%%%%%%%%
32: \begin{document}
33: 
34: \graphicspath{{figs/}}
35: 
36: \date{Accepted ........ Received ........; in original form ........}
37: 
38: \pagerange{\pageref{firstpage}--\pageref{lastpage}} \pubyear{2007}
39: 
40: \maketitle
41: 
42: \label{firstpage}
43: 
44:  
45: \begin  {abstract}  We  investigate  the clustering  properties  of  a
46: complete sample  of $10^5$ star-forming  galaxies drawn from  the data
47: release 4 (DR4) of the Sloan  Digital Sky Survey.  On scales less than
48: 100 kpc , the amplitude  of the correlation function exhibits a strong
49: dependence  on the  specific star  formation rate  of the  galaxy.  We
50: interpret this as the signature  of enhanced star formation induced by
51: tidal interactions.   We then explore  how the average  star formation
52: rate in a galaxy is enhanced as the projected separation $r_p$ between
53: the galaxy and its companions decreases.  We find that the enhancement
54: depends strongly on $r_p$, but  very weakly on the relative luminosity
55: of  the companions.   The enhancement  is  also stronger  in low  mass
56: galaxies than  in high mass galaxies.   In order to  explore whether a
57: tidal  interaction  is not  only  sufficient,  but  also necessary  to
58: trigger  enhanced star formation  in a  galaxy, we  compute background
59: subtracted  neighbour counts  for  the galaxies  in  our sample.   The
60: average number of close neighbours around galaxies with low to average
61: values of  SFR/$M_*$ is close to  zero.  At the  highest specific star
62: formation rates, however, more than 40\% of the galaxies in our sample
63: have  a  companion within  a  projected  radius  of 100  kpc.   Visual
64: inspection  of  the  highest  SFR/$M_*$  galaxies  without  companions
65: reveals that more than 50\%  of these are clear interacting or merging
66: systems.  We conclude that tidal interactions are the dominant trigger
67: of enhanced star formation  in the most strongly star-forming systems.
68: Finally, we find clear evidence  that tidal interactions not only lead
69: to  enhanced star  formation in  galaxies, but  also  cause structural
70: changes such as an increase in concentration.
71:  
72: 
73: \end {abstract}
74: 
75: \begin{keywords}  galaxies:  clustering   -  galaxies:  distances  and
76: redshifts -  large-scale structure of  Universe - cosmology:  theory -
77: dark matter
78: \end{keywords}
79: 
80: \section {Introduction}
81: 
82: It has been known for  more than thirty years that galaxy interactions
83: lead  to enhanced  star formation.   \cite{Toomre-Toomre-72} pioneered
84: the use of numerical simulations to study the interactions of galaxies
85: and suggested that gas may be  funnelled to the central regions of the
86: systems as a result of the strong tidal forces that operate during the
87: encounter. This  gas is then able  to fuel a burst  of star formation.
88: Since  then, there  have  been many  studies,  both observational  and
89: theoretical,  that   have  examined  the   relationship  between  star
90: formation and galaxy interactions.
91: 
92: Most   early  observational   studies  adopted   broad   band  colours
93: \citep[e.g.][]{Larson-Tinsley-78},    H$\alpha$   equivalent   widths,
94: \citep[e.g.][]{Keel-85,Bushouse-86,Kennicutt-87},          far-infrared
95: luminosities   \citep[e.g.][]{Bushouse-Werner-Lamb-88},  or  molecular
96: (CO) emission \citep{Young-86, Sanders-86, Solomon-Sage-88, Tinney-90,
97: Young-96}  as  indicators  of   star  formation.   These  studies  all
98: demonstrated  that  galaxy interactions  are  statistically linked  to
99: enhanced   rates    of   star   formation    \citep[see   the   review
100: of][]{Keel-91,Struck-99}.
101: 
102: Recent  studies of star  formation in  interacting galaxies  have been
103: based on redshift surveys such as the Center for Astrophysics redshift
104: survey
105: \citep[CfA2;][]{Barton-Geller-Kenyon-00,Woods-Geller-Barton-06},    the
106: Two Degree Field Redshift Survey \citep[2dFGRS;][]{Lambas-03}, and the
107: Sloan                 Digital                Sky                Survey
108: \citep[SDSS;][]{Nikolic-Cullen-Alexander-04,Woods-Geller-07,Ellison-08}.
109: These  studies have  also  provided observational  evidence that  star
110: formation is enhanced as a consequence of tidal interactions.  Most of
111: these studies have also demonstrated that the degree of enhancement is
112: a strong function of the projected separation between the two galaxies
113: as well  as their difference  in redshift.  In addition,  some studies
114: investigated   how    galaxy   properties   such    as   concentration
115: \citep{Nikolic-Cullen-Alexander-04},          luminosity         ratio
116: \citep{Woods-Geller-Barton-06,Woods-Geller-07},   stellar  mass  ratio
117: \citep{Ellison-08},  colour, and AGN  activity \citep{Woods-Geller-07}
118: depend on separation.
119: 
120: Although  most studies  have supported  the picture  that interactions
121: induce  star formation, there  have been  number of  dissenting papers.
122: For     example,     \citet{Bergvall-Laurikainen-Aalto-03}    analyzed
123: optical/near-IR  observations of  a sample  of  59 interacting/merging
124: systems and concluded that they  do not differ very much from isolated
125: galaxies   in   terms   of   their  global   star   formation   rates.
126: \citet{Brosch-Almoznino-Heller-04} found that interaction-induced star
127: formation is not significant for  dwarf galaxies.  A more recent study
128: by \citet{Smith-07}  analyzed Spitzer mid-infrared (MIR)  imaging of a
129: sample  of 35  interacting galaxy  pairs selected  from the  Arp Atlas
130: \citep{Arp-66}.   They compared  the  global MIR  properties of  these
131: systems  with those  of normal  spiral  galaxies.  The  MIR colors  of
132: interacting  galaxies were  found to  be redder  than  normal spirals,
133: implying enhancements  to the  specific SFRs of  a factor  of $\sim$2.
134: However, in contrast to results from previous investigations, they did
135: not   find   any   evidence   that   the   enhancement   depended   on
136: separation. This may be due to the small size of their sample and fact
137: that   the   galaxies   were   selected  to   be   tidally   disturbed
138: \citep{Smith-07}.
139: 
140: On  the   theoretical  side,  $N$-body  simulations   that  treat  the
141: hydrodynamics      of     the      gas     \citep{Negroponte-White-83,
142: Barnes-Hernquist-92,   Mihos-Hernquist-96,   Springel-00,  Tissera-02,
143: Meza-03,  Kapferer-05,  Cox-06}  have demonstrated  that  interactions
144: between galaxies can bring gas from the disc to the central regions of
145: the  galaxy,  leading  to   enhanced  star  formation  in  the  bulge.
146: Recently, \citet{DiMatteo-07}  investigated star formation  in a suite
147: of several hundred numerical  simulations of interacting galaxies with
148: different gas fractions,  bulge-to-disk ratios and orbital parameters.
149: Their work confirmed that  galaxy interactions and mergers can trigger
150: strong nuclear starbursts.  However, the authors pointed out that this
151: is not always the case, because strong tidal interactions at the first
152: pericenter passage  can remove a large  amount of gas  from the galaxy
153: disks, and this  gas is only partially re-acquired  by the galaxies in
154: the last phase of the merging event.
155: 
156: In   summary,    although   it   is   now    well   established   that
157: interactions/mergers   between  galaxies   {\em   can}  enhance   star
158: formation, a number of important questions remain to be answered:
159: \begin{itemize}
160: \item  Are interactions  not  only sufficient  but  also necessary  to
161: enhance star formation?
162: \item Do interactions {\em always} trigger enhanced star formation?
163: \item How does the enhancement  in star formation depend on parameters
164: such as the separation between  the two galaxies and their mass ratio?
165: Does the enhancement also depend on properties such as stellar mass or
166: galaxy morphology?
167: \end{itemize}
168: 
169: To answer these questions, we adopt three different methods to analyse
170: a sample  of $\sim10^5$ star-forming  galaxies selected from  the Data
171: Release 4 (DR4) of the  SDSS.  First, we compute the cross-correlation
172: between star-forming galaxies and a reference sample of galaxies drawn
173: from  the DR4.   In the  standard  model of  structure formation,  the
174: amplitude of the correlation function  on scales larger than a few Mpc
175: provides a  direct measure of the  mass of the dark  matter haloes that
176: host the galaxies.  As we  will show, the amplitude of the correlation
177: function on  scales less than $\sim 100$  kpc can serve as  a probe of
178: physical processes such as  mergers and interactions.  We then compute
179: the  average  enhancement in  star  formation  as  a function  of  the
180: projected  separation between  two  galaxies and  we  explore how  the
181: enhancement  depends on  galaxy properties  such as  stellar  mass and
182: concentration index.   Finally, we compute counts  around our galaxies
183: as a function of separation and explore how this changes as a function
184: of  the specific  star formation  rate  SFR/$M_*$. This  allows us  to
185: investigate  whether  the  majority  of galaxies  with  specific  star
186: formation  rates  above   some  critical  threshold  are  experiencing
187: merger-induced starbursts.   In a  separate paper, we  explore whether
188: AGN activity  is also triggered  by tidal interactions using  the same
189: set of analysis techniques.
190: 
191: Throughout this paper, We assume a cosmological model with the density
192: parameter $\Omega_0=0.3$ and  a cosmological constant $\Lambda_0=0.7$.
193: To avoid the $-5\log_{10}h$ factor,  a Hubble constant $h=1$, in units
194: of  $100\kms{\rm Mpc}^{-1}$,  is  assumed throughout  this paper  when
195: computing absolute magnitudes.
196: 
197: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
198: \section{Samples}
199: 
200: \subsection {The SDSS Spectroscopic Sample}
201: 
202: The data analyzed  in this study are drawn from  the Sloan Digital Sky
203: Survey (SDSS).  The survey goals are to obtain photometry of a quarter
204: of  the sky and  spectra of  nearly one  million objects.   Imaging is
205: obtained    in     the    {\em    u,    g,    r,     i,    z}    bands
206: \citep{Fukugita-96,Smith-02,Ivezic-04}  with a  special  purpose drift
207: scan camera  \citep{Gunn-98} mounted  on the SDSS  2.5~meter telescope
208: \citep{Gunn-06}  at Apache  Point Observatory.   The imaging  data are
209: photometrically    \citep{Hogg-01,Tucker-06}    and    astrometrically
210: \citep{Pier-03} calibrated,  and used  to select stars,  galaxies, and
211: quasars  for follow-up fibre  spectroscopy.  Spectroscopic  fibres are
212: assigned to  objects on  the sky using  an efficient  tiling algorithm
213: designed to optimize  completeness \citep{Blanton-03}.  The details of
214: the survey strategy can be found in \citet{York-00} and an overview of
215: the data pipelines and products  is provided in the Early Data Release
216: paper  \citep{Stoughton-02}. More details  on the  photometric pipeline
217: can be found in \citet{Lupton-01}.
218: 
219: Our parent sample for this  study is composed of 397,344 objects which
220: have  been  spectroscopically  confirmed  as galaxies  and  have  data
221: publicly     available     in      the     SDSS     Data     Release~4
222: \citep{Adelman-McCarthy-06}.   These  galaxies are  part  of the  SDSS
223: `main'  galaxy   sample  used   for  large  scale   structure  studies
224: \citep{Strauss-02}  and have  Petrosian  $r$ magnitudes  in the  range
225: $14.5 < r < 17.77$ after correction for foreground galactic extinction
226: using  the  reddening  maps  of  \citet{Schlegel-Finkbeiner-Davis-98}.
227: Their redshift  distribution extends from $\sim0.005$ to  0.30, with a
228: median $z$ of 0.10.
229: 
230: The SDSS spectra are obtained with two 320-fibre spectrographs mounted
231: on  the SDSS  2.5-meter telescope.   Fibers 3  arcsec in  diameter are
232: manually plugged  into custom-drilled  aluminum plates mounted  at the
233: focal plane of  the telescope. The spectra are  exposed for 45 minutes
234: or until a fiducial signal-to-noise  (S/N) is reached.  The median S/N
235: per pixel  for galaxies in the  main sample is  $\sim14$.  The spectra
236: are  processed by  an automated  pipeline, which  flux  and wavelength
237: calibrates  the   data  from  3800  to   9200~\AA.   The  instrumental
238: resolution  is   R~$\equiv  \lambda/\delta\lambda$  =   1850  --  2200
239: (FWHM$\sim2.4$~\AA\ at 5000~\AA).
240: 
241: \subsection{Star-forming galaxies}
242: 
243: Our   sample  of  star-forming   galaxies  is   drawn  from   the  DR4
244: spectroscopic    sample    using    the    criteria    described    in
245: \citet{Brinchmann-04}. In order for a galaxy to be securely classified
246: as  star-forming, we  require  that the  four  emission lines  [OIII],
247: H$\beta$,  H$\alpha$ and  [NII] all  be detected  with signal-to-noise
248: greater than 3 and that the ratios [OIII]/H$\beta$ and [NII]/H$\alpha$
249: have   values   that   place   them   within   the   region   of   the
250: \citet[][BPT]{Baldwin-Phillips-Terlevich-81}   diagram   occupied   by
251: galaxies in which  the primary source of ionizing  photons is from HII
252: regions rather  than an AGN. We refer  to this sample as  the high S/N
253: star-forming class.   In certain cases, we supplement  the sample with
254: the low S/N star-forming class defined by Brinchmann et al.  These are
255: the  galaxies that  are  left over  after  all the  AGN  and high  S/N
256: star-forming galaxies  have been  removed, and they  have S/N  $>2$ in
257: H$\alpha$.   Star formation rates  can still  be estimated  from their
258: emission line  strengths, but  the errors on  these estimates  will be
259: significantly larger than for the high S/N sample.
260: 
261: The  reader  is  referred  to  \citet{Brinchmann-04}  for  a  detailed
262: description of  how star formation  rates are derived for  the various
263: samples. We  will be  making use of  the specific star  formation rate
264: $SFR/M_*$ estimated within the  3 arsecond SDSS fibre aperture.  These
265: star formation rates  are more accurate than the  total star formation
266: rates derived  by Brinchmann  et al, because  they depend only  on the
267: emission line fluxes measured from the spectra and they do not involve
268: any uncertain colour corrections.  The disadvantage of the fibre-based
269: specific star formation  rates is that they are  only sensitive to the
270: emission from the inner region of the galaxy, which includes one third
271: of the total light on average.
272: 
273: \subsection{Reference Samples}
274: 
275: We   work  with   two  different   reference  samples:   (i)   a  {\em
276: spectroscopic}  reference  sample,  which   is  used  to  compute  the
277: projected  cross-correlation function $w_p(r_p)$  between star-forming
278: galaxies  and  reference  galaxies,   and  (ii)  a  {\em  photometric}
279: reference  sample,  which  is   used  to  calculate  counts  of  close
280: neighbours  around  star-forming  galaxies.    We  use  the  New  York
281: University Value  Added Galaxy  Catalogue (NYU-VAGC) to  construct the
282: reference  samples. The  original  NYU-VAGC is  a  catalogue of  local
283: galaxies    (mostly    below    $z\approx   0.3$)    constructed    by
284: \cite{Blanton-05} based on the SDSS DR2. Here, we use a new version of
285: the  NYU-VAGC ({\tt  Sample dr4}),  which is  based on  SDSS  DR4. The
286: NYU-VAGC is described in detail in \cite{Blanton-05}.
287: 
288: The reference samples  are exactly the same as  used in \cite{Li-06b}.
289: In  short,  the  spectroscopic  reference  sample  is  constructed  by
290: selecting from {\tt  Sample dr4} all galaxies with $14.5 <  r < 17.6 $
291: that are identified as galaxies  from the Main sample, in the redshift
292: range    $0.01\leq   z\leq0.3$,    and   with    absolute   magnitudes
293: $-23<M_{^{0.1}r}<-17$.   The spectroscopic  reference  sample contains
294: 292,782   galaxies.   The   photometric  reference   sample   is  also
295: constructed  from {\tt  Sample  dr4} by  selecting  all galaxies  with
296: $14.5<r<19$.   The resulting sample  includes 1,065,183  galaxies.  In
297: certain cases, we will work  with photometric reference samples with a
298: range of differing limiting magnitudes.
299: 
300: 
301: \section{Cross-correlation functions}
302: 
303: \begin{figure} 
304: \centerline{\psfig{figure=f1.ps,clip=true,width=0.5\textwidth}}
305: \caption{Projected  redshift-space 2-point  cross-correlation function
306: $w_p(r_p)$  between  star-forming galaxies  and  the reference  galaxy
307: sample.   Different  lines correspond  to  star-forming galaxies  with
308: different specific star formation rates.   See the text for a detailed
309: description.}
310: \label{fig:wrp}
311: \end{figure}
312: 
313: \begin{figure*}      
314: \centerline{\psfig{figure=f2.ps,clip=true,width=\textwidth}}
315: \caption{Similar to  Figure~1, but  in different intervals  of stellar
316: mass as indicated at the top  of the figure.  The symbols are the same
317: as in  Figure~\ref{fig:wrp}, except that a power  law corresponding to
318: $\xi(r)=(r/5h^{-1}Mpc)^{-1.8}$ is  additionally plotted in  each panel
319: as a long-dashed line.}
320: \label{fig:wrp_smass}
321: \end{figure*}
322: 
323: \begin{figure*}
324: \centerline{\psfig{figure=f3.ps,clip=true,width=\textwidth}}
325: \caption{Same  as Figure~\ref{fig:wrp_smass},  except  that the  three
326: $SFR/M_\ast$ samples in each panel are matched in concentration.}
327: \label{fig:wrp_smass_c}
328: \end{figure*}
329: 
330: \begin{figure*}
331: \centerline{\psfig{figure=f4.ps,clip=true,width=\textwidth}}
332: \caption{The projected  2PCCF $w_p(r_p)$  normalized by the  power law
333: corresponding to a  real-space 2PCF of $\xi(r)=(r/5h^{-1}Mpc)^{-1.8}$,
334: as  measured  at  different  physical  scales and  as  a  function  of
335: $SFR/M_\ast$.   Dashed lines  are for  high S/N  star-forming galaxies
336: only, while  solid lines  show results for  the sample  including both
337: high and low S/N star-forming galaxies.}
338: \label{fig:wrp_sfr}
339: \end{figure*}
340: 
341: Our methodology for computing correlation functions has been described
342: in detail  in our  previous papers \citep{Li-06a,Li-06b}.   We present
343: here a  brief description  and the reader  is referred to  the earlier
344: papers for details.  Random samples are constructed that have the same
345: selection  function  as  the  reference  sample.   The  redshift-space
346: two-point  cross-correlation function  (2PCCF)  $\xi(r_p,\pi)$ between
347: star-forming  galaxies and  the  reference sample  is then  calculated
348: using  the  estimator   presented  in  \citet{Li-06b}.   Finally,  the
349: redshift-space projected 2PCCF  $w_p(r_p)$ is estimated by integrating
350: $\xi(r_p,\pi)$  along the line-of-sight  direction $\pi$  with $|\pi|$
351: ranging from  0 to 40  $h^{-1}Mpc$.  We have also  corrected carefully
352: for the effect of fibre collisions  and a description and tests of the
353: method  are given  in \citet{Li-06b}.   The errors  on  the clustering
354: measurements  are estimated using  the bootstrap  resampling technique
355: \citep{Barrow-Bhavsar-Sonoda-84}.
356: 
357: We first  compute $w_p(r_p)$ for  our sample of high  S/N star-forming
358: galaxies from the SDSS DR4. In order to study how this depends on star
359: formation rate (SFR),  we rank all the high  S/N star-forming galaxies
360: according to the values of their specific star formation rates (SSFR),
361: $SFR/M_\ast$,  and define  subsamples of  'high SSFR'  and  'low SSFR'
362: galaxies  as   those  contained  within  the  upper   and  lower  25th
363: percentiles  of the distribution  of this  quantity.  The  results are
364: shown in  Figure~\ref{fig:wrp}.  The dashed  (dotted) line corresponds
365: to  the high  (low) SSFR  subsample, while  the solid  line  shows the
366: result for the sample as a whole.
367: 
368: Figure~\ref{fig:wrp} shows that galaxies with higher $SFR/M_\ast$ have
369: stronger  clustering on  scales smaller  than 0.1  Mpc and  the effect
370: becomes stronger at smaller  projected separations.  As pointed out by
371: \citet{Li-06b},  the clustering  amplitude  of galaxies  depends on  a
372: variety  of  galaxy  properties,  including stellar  mass  and  galaxy
373: structure.  If we  wish to  isolate the  effect of  the  specific star
374: formation  rate, it is  important that  we make  sure that  the galaxy
375: samples  that  we  study  are   closely  matched  in  terms  of  other
376: properties,  so that  the effect  on the  star formation  rate  can be
377: isolated. We have thus divided  all the high S/N star-forming galaxies
378: into  four subsamples  according to  $\log_{10}(M_\ast/M_\odot)$.  For
379: each  subsample  we repeat  the  clustering  analysis  as above.   The
380: results  are  shown in  Figure~\ref{fig:wrp_smass}.   The four  panels
381: correspond to different  intervals of $\log_{10}(M_\ast/M_\odot)$.  To
382: guide  the eye,  a power  law corresponding  to a  real-space  2PCF of
383: $\xi(r)=(r/5h^{-1}Mpc)^{-1.8}$  is plotted  as a  long-dashed  line in
384: each panel.   We see  that the amplitude  of $w_p(r_p)$  increases for
385: galaxies  with larger  stellar masses.   This is  consistent  with our
386: previous findings about the  mass dependence of galaxy clustering.  We
387: also see that the difference  in clustering between galaxies with high
388: and low SFR/$M_*$ on scales smaller than 0.1 Mpc is most pronounced in
389: the lowest stellar mass interval.   Next, in each of the four $M_\ast$
390: intervals, we  match the  three $SFR/M_\ast$ samples  in concentration
391: parameter $C$ by requiring that the distribution of $C$ is exactly the
392: same as in each of these samples. The $w_p(r_p)$ measurements for such
393: matched  samples  are   shown  in  Figure~\ref{fig:wrp_smass_c}.   The
394: results are very similar to those shown in the previous figure.
395: 
396: We  conclude that  the  small scale  clustering  dependences shown  in
397: Figure~\ref{fig:wrp} are  genuinely related to  the differing specific
398: star   formation   rates   of    the   galaxies   in   the   different
399: samples.  Galaxies  with the  highest  specific  star formation  rates
400: apparently have  an excess of companions  on scales less  than 100 kpc
401: when compared  to the average  star-forming galaxy. The fact  that the
402: increase in clustering occurs only  on very small scales suggests that
403: the  excess   star  formation  is   being  triggered  by   {\em  tidal
404: interactions} with these  companions.  Another intriguing result shown
405: in these figures is that  galaxies with low $SFR/M_\ast$ are {\em less
406: clustered} on small  scales. This suggests that there  might be a {\em
407: continuous}  trend  linking average  number  of  close neighbours  and
408: $SFR/M_\ast$.
409: 
410: To investigate  this in more  detail, we calculate how  the clustering
411: amplitude depends  on $SFR/M_\ast$ at a variety  of different physical
412: scales.  The results are shown in Figure~\ref{fig:wrp_sfr} (red dashed
413: lines).  One problem with the  high S/N star-forming sample is that it
414: does not  extend to $SFR/M_\ast$  values much below $\sim  -10.5$.  To
415: extend our analysis to lower values, we include the sample of {\em low
416: S/N}  star-forming  galaxies  defined  by  \citet{Brinchmann-04}.   As
417: discussed in section  2.2, the star formation rates  in these galaxies
418: are  estimated  from  the  H$\alpha$  line luminosity,  but  the  dust
419: correction is quite uncertain because H$\beta$ is not usually detected
420: with high S/N.   Results where the low S/N  star-forming galaxies have
421: been    included   are    plotted    as   black    solid   lines    in
422: Figure~\ref{fig:wrp_sfr}.  As can be seen, $\log SFR/M_*$ extends down
423: to values around $\sim -11$ for this sample.
424: 
425: On  scales larger than  100 kpc,  there is  very little  dependence of
426: clustering amplitude on specific star formation rate for $\log SFR/M_*
427: >  -10$.   At lower  values  of  $SFR/M_*$,  the clustering  amplitude
428: increases.   This is a  manifestation of  the strong  relation between
429: star  formation and local  density or  environment.  It  is well-known
430: that galaxies  located in dense,  massive structures such  as clusters
431: have  lower  specific star  formation  rates  than ``field''  galaxies
432: \citep[e.g.][]{Kauffmann-04}. It  is currently accepted  that after a
433: galaxy  is  accreted onto  a  larger structure,  such  as  a group  or
434: cluster, its star formation rate  will decline, either because its gas
435: is  removed by  processes such  as ram-pressure  stripping,  or simply
436: because no further  gas accretion takes place and  the galaxy runs out
437: of the fuel to make new stars.
438: 
439: On scales less than 100 kpc, the dependence of clustering amplitude on
440: specific star formation  rate is more complicated. At  values of $\log
441: SFR/M_*$  less  than -10,  we  see  the  same increase  in  clustering
442: amplitude  that we  saw on  larger  scales. This  may appear  somewhat
443: surprising at first. In a recent paper, however, \citet{Barton-07} use
444: cosmological  simulations  to  show  that a  substantial  fraction  of
445: galaxies selected as ``close pairs'' from surveys such as SDSS or 2DF,
446: do in  fact reside in very  massive dark matter halos.   Based on this
447: work, we conjecture that the rise in clustering amplitude seen at all
448: separations  at low values  of SFR/$M_*$  rate is  the result  of star
449: formation shutting down  in galaxy groups and clusters.   At values of
450: $\log SFR/M_*$ greater than -10 and at separations less than $\sim 50$
451: kpc,  the clustering  amplitude  shows a  strong  and continuous  {\em
452: increase} towards larger values of  $SFR/M_*$.  This is a clear signal
453: that mergers or interactions play an important role in triggering {\em
454: enhanced} star formation in galaxies.
455: 
456: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
457: \section{Star formation enhancement functions}
458: \label{sec:enhancement}
459: 
460: \begin{figure*}
461: \centerline{\psfig{figure=f5.ps,clip=true,width=\textwidth}}
462: \caption{Star  formation enhancement  as a  function of  the projected
463: separation  $r_p$  (top  panels)  and  as a  function  of  the  scaled
464: separation  $r_p/R_{90}$ (bottom panels),  for all  the high  S/N star
465: forming galaxies  (left panels) and for galaxies  in different stellar
466: mass ranges  (right panels).  All the errors  are estimated  using the
467: Bootstrap  resampling  technique.   The  dashed lines  in  each  panel
468: indicate  the  variance  between  10  realizations in  which  the  sky
469: positions of  the star-forming galaxies  are randomized. See  the text
470: for details.}
471: \label{fig:sfef}
472: \end{figure*}
473: 
474: \begin{figure*}   
475: \centerline{\psfig{figure=f6.ps,clip=true,width=\textwidth}}
476: \caption{Star  formation  enhancement   as  a  function  of  projected
477: separation  $r_p$  (top  panels)  and  as a  function  of  the  scaled
478: separation  $r_p/R_{90}$  (bottom panels).  In  the left-hand  panels,
479: different  symbols connected  by solid  lines correspond  to reference
480: samples with  different limiting magnitudes (as  indicated), while the
481: magnitude    of   star-forming   samples    is   kept    constant   at
482: $r_{SFG}=17.6$.  In the  right-hand  panels, the  reference sample  is
483: always  limited  at $r_{pho}=19.5$  but  the  magnitude  limit of  the
484: star-forming  sample is changed  (as indicated).   All the  errors are
485: estimated using the Bootstrap  resampling technique.  The dashed lines
486: in each panel  indicate the variance between 10  realizations in which
487: the sky  positions of the  star-forming galaxies are  randomized.  See
488: the text for details.  }
489: \label{fig:sfef_fainter}
490: \end{figure*}
491: 
492: In this section, we probe  the relationship between star formation and
493: galaxy interactions by quantifying the enhancement in star formation as
494: a function of the projected  separation between two galaxies.  We also
495: study how  the enhancement depends  on the physical properties  of the
496: main galaxy.  We compute how the average value of $SFR/M_\ast$ changes
497: as a  function of the projected  distance to the  neighbours. From now
498: on, we restrict our attention to the sample of high $S/N$ star-forming
499: galaxies. These  lie in low density environments  where processes such
500: as ram-pressure stripping, gas starvation  etc should play a much less
501: important role (see Figure 4).
502: 
503: The  neighbours  of a  galaxy  are  identified  using the  photometric
504: reference sample.   The advantage of  using the photometric  sample is
505: that the result is not affected by incompleteness (e.g.  the effect of
506: fibre collisions).  However, the disadvantage is that some fraction of
507: the  close neighbours  will not  be  true nearby  systems, but  rather
508: chance  projections of  foreground  and background  galaxies that  lie
509: along the line-of-sight. We correct  for this as follows: We count the
510: number  of  companions  in  the  photometric  reference  sample  at  a
511: projected  physical distance  $r_p$ for  each galaxy  with a  high S/N
512: measure of  the specific  star formation rate  $\log(SFR/M_\ast)$.  We
513: also generate  10 random  samples that have  the same geometry  as the
514: photometric reference  sample by randomizing  the sky position  of the
515: photometric  objects and keeping  all the  other quantities  (e.g. the
516: magnitudes) fixed. We use these random catalogues to estimate the mean
517: number of projected companions  expected at random around each galaxy.
518: The  true number of  companions at  separation $r_p$  is given  by the
519: difference  between the  observed and  the projected  random companion
520: count.  We  then calculate a weighted average  specific star formation
521: rate at projected distance $r_p$  by weighting each galaxy by its true
522: companion number.  The  enhancement in $\log(SFR/M_\ast$), $E_X(r_p)$,
523: is  defined as  the difference  between the  weighted average  and the
524: unweighted one.  This can be written as
525: \begin{equation}
526: E_X(r_p) = \frac{\sum_i^NX_i[n_{o,i}(r_p)-n_{p,i}(r_p)]}
527: {\sum_i^N[n_{o,i}(r_p)-n_{p,i}(r_p)]}-\frac{\sum_i^NX_i}{N},
528: \end{equation} where $X_i=\log(SFR_i/M_{\ast,i})$ is the specific star
529: formation rate of the i'th galaxy, and $n_{o,i}$ and $n_{p,i}$ are the
530: observed and projected random companion counts as described above.
531: 
532: We  first consider all  high S/N  star-forming galaxies  with $r$-band
533: apparent magnitude in the  range $14.5<r<17.6$.  To begin, we restrict
534: the photometric reference sample  to galaxies with $r$-band magnitudes
535: $r<19.0$. In order to ensure that we are finding similar neighbours at
536: all  redshifts, we  only  consider neighbouring  galaxies that  are
537: brighter   than   $r_{SFG}+1.4$   mag.   The  result   is   shown   in
538: Figure~\ref{fig:sfef}.  The  errors   are  estimated  using  Bootstrap
539: resampling techniques. The dashed  lines indicate the variance between
540: 10 samples in which we randomize the sky positions of the star-forming
541: galaxies.
542: 
543: \begin{figure*}
544: \centerline{\psfig{figure=f7.ps,clip=true,width=\textwidth}}
545: \caption{The     same     as      the     bottom-left     panel     of
546: Figure~\ref{fig:sfef_fainter},  but for  the low  mass  (the left-hand
547: panel) and the high mass (the right-hand panel) subsamples separately.
548: To  guide  the  eye, the  result  for  the  whole  sample in  case  of
549: $r_{SFG}<17.6$ and $r_{pho}<19.0$ is plotted as solid black circles in
550: every panel.  }
551: \label{fig:sfef_smass_fainter}
552: \end{figure*}
553: 
554: \begin{figure*}
555: \centerline{\psfig{figure=f8.ps,clip=true,width=\textwidth}}
556: \caption{Similar to  the previous  plot but for  star-forming galaxies
557: with different  concentration indices, as indicated  above each panel,
558: and for the  case of $r_{SFG}<17.6$ and $r_{pho}<19.5$  only. The blue
559: triangles  are for the  concentration subsamples  and the  black solid
560: circles are for the whole sample.}
561: \label{fig:sfef_c}
562: \end{figure*}
563: 
564: 
565: The top left  panel of Figure~\ref{fig:sfef} gives the  result for the
566: sample as a whole. On scales larger than a Mpc or so $E_X$ is constant
567: at a  slightly but significantly  negative value. This is  because the
568: average in equation 1 is  pair-weighted, and galaxies in massive halos
569: have  lower  specific star  formation  rates  than  average, but  more
570: "companions" at  large $r_p$ than  average as a result  of large-scale
571: bias effects. On  scales below about 100 kpc,  $E_X$ increases sharply
572: and reaches values corresponding to a factor of about two at projected
573: separations  less   than  20   kpc.   In  the   top  right   panel  of
574: Figure~\ref{fig:sfef}, we  plot results for galaxies  divided into two
575: different ranges in stellar mass.   These results show that there is a
576: strong  dependence of star  formation enhancement  on galaxy  mass, in
577: that star formation  in small galaxies is more  strongly enhanced at a
578: given projected separation.
579: 
580: In the bottom  panels, we scale the projected  separation $r_p$ by the
581: physical size of  the galaxy.  We use $R_{90}$,  the radius containing
582: 90\%  of  the  total  $r$-band  light, to  calculate  a  {\em  scaled}
583: projected  separation  and recompute  the  enhancement  function as  a
584: function of  this scaled quantity.  One  can see that  the results are
585: quite similar.   Star formation is  enhanced at separations  less than
586: $\sim 10$  times the optical  radius of the  galaxy and the  effect is
587: stronger for lower mass systems.
588: 
589: We  now  investigate  the  importance  of the  relative  mass  of  the
590: companion galaxy in determining the  degree to which star formation is
591: enhanced  in  the primary  galaxy  by  analyzing  galaxy samples  with
592: different limiting  magnitudes.  We first keep the  magnitude limit of
593: star-forming  sample  constant  at  $r_{SFG}=17.6$, and  explore  what
594: happens if we  change the limiting magnitude of  the reference sample.
595: The results are plotted as circles for $r_{pho}<18.5$ and as triangles
596: for     $r_{pho}<19.5$     in      the     left-hand     panels     of
597: Figure~\ref{fig:sfef_fainter}.   The  result  shown  in  the  previous
598: figure is plotted as squares.   Next, we fix the limiting magnitude of
599: the  reference sample  at $r_{pho}=19.5$,  but decrease  the magnitude
600: limit of  the star-forming sample.  The maximum allowed  difference in
601: magnitude between  the star-forming galaxy  and its companion  is also
602: increased accordingly.  Results are shown for $r_{SFG}<16.5, 17.0$ and
603: $17.6$.  in the right-hand panels of Figure~\ref{fig:sfef_fainter}.
604: 
605: \begin{figure}
606: \centerline{
607: \psfig{figure=f9.ps,clip=true,width=0.5\textwidth}}
608: \caption{Distribution of both  high and low S/N star-forming galaxies
609: in  the plane  of stellar  mass versus  specific star  formation rate,
610: coloured by  concentration index  $R90/R50$ measured in  the $z$-band.
611: The color coding  of $R90/R50$ is shown in the  bar at the right-hand.}
612: \label{fig:con_contour}
613: \end{figure}
614: 
615: We  see  from Figure~\ref{fig:sfef_fainter}  that  the star  formation
616: enhancement  depends  very  little  on  the  mass  ratio  between  the
617: star-forming galaxy and its companion.  There are small changes in the
618: expected  direction  (i.e.  there  is  slightly  less enhancement  for
619: companions  with  lower  relative   mass),  but  to  first  order  the
620: enhancement  function remains remarkably  constant for  different mass
621: ratios.    In  Figure~\ref{fig:sfef_smass_fainter},   we   divide  the
622: star-forming  sample into  two  different stellar  mass intervals  and
623: explore if our results change.   We find that the enhancement function
624: has very little dependence on the mass ratio of the companion for both
625: low mass and high mass star-forming galaxies.
626: 
627: Finally,  we investigate  the enhancement  function for  galaxies with
628: different  structural   properties.   We  divide  all   the  high  S/N
629: star-forming  galaxies  into   different  intervals  of  concentration
630: parameter  $C$ and  repeat the  analysis described  above for  each of
631: these subsamples.  In Figure~\ref{fig:sfef_c}, we plot the results for
632: star-forming galaxies  with $r_{SFG}<17.6$ and  for reference galaxies
633: with $r_{pho}<19.5$.  We  see that the  star formation
634: enhancement does depend  on $C$, in that the  galaxies with larger $C$
635: values are  more strongly enhanced. One possible  explanation for this
636: effect  is that  interaction-induced starbursts  occur when  gas flows
637: into  the  core of  a  galaxy, causing  it  to  become more  centrally
638: concentrated \citep{Sanders-Mirabel-96}.
639: 
640: In Figure~\ref{fig:con_contour}, we  investigate how the concentration
641: index of a star-forming galaxy depends on its location in the plane of
642: specific  star formation  rate versus  stellar mass.   We see  that at
643: fixed  stellar mass, the  average concentration  index is  highest for
644: galaxies that are  currently experiencing both higher-than-average and
645: lower-than-average  rates of  star formation.   One  interpretation of
646: this plot is  that tidal interactions cause gas to  flow from the disk
647: to the  nucleus and  this triggers  a starburst at  the centre  of the
648: galaxy. The formation  of new stars in the  central regions causes the
649: concentration index to  increase. The starburst is then  followed by a
650: period of relative quiescence, which lasts until the galaxy is able to
651: accrete more gas into its disk.  Formation of stars in the disk brings
652: the galaxy  back into  the "central plane"  occupied by  galaxies with
653: $\log  SFR/M_*  \sim -9.5$  in  Figure~\ref{fig:con_contour}.  In  the
654: bottom-right corner  of the plot, both stellar  mass and concentration
655: are high,  but the specific star  formation rate is low.   This is the
656: regime of early-type galaxies.
657: 
658: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
659: \section{Close neighbour counts}
660: 
661: \begin{figure*}
662: \centerline{
663: \psfig{figure=f10a.ps,clip=true,width=0.33\textwidth}
664: \psfig{figure=f10b.ps,clip=true,width=0.33\textwidth}
665: \psfig{figure=f10c.ps,clip=true,width=0.33\textwidth} }
666: \caption{Average counts of galaxies  in the photometric sample (panels
667: from left  to right: $r_{lim}  < 18$, $19$,  and $20$) within  a given
668: projected  radius  $R_p$ from  the  star-forming galaxies.   Different
669: symbols  are  for  star-forming  galaxies in  different  intervals  of
670: specific star formation rate, as indicated.}
671: \label{fig:counts}
672: \end{figure*}
673: 
674: In  this section, we  investigate whether  tidal interactions  are not
675: only  a sufficient, but  also a  necessary condition  for a  galaxy to
676: experience enhanced  star formation.  We count the  number of galaxies
677: in the photometric sample in  the vicinity of the star-forming galaxies
678: and make a statistical correction for the effect of chance projections
679: by subtracting the average count around randomly placed galaxies.
680: 
681: In  Figure~\ref{fig:counts} we plot  the average  correlated neighbour
682: count  (i.e. after statistical  correction for  uncorrelated projected
683: neighbours)   within   a  given   value   of   the  projected   radius
684: $R_p$.  Results  are  shown  for  high S/N  star-forming  galaxies  in
685: different intervals  of specific star formation rate.  We have trimmed
686: each  subsample  so that  they  each  have  the same  distribution  in
687: redshift  and  in stellar  mass  $M_*$.   Panels  from left  to  right
688: correspond  to  photometric  reference  samples that  are  limited  at
689: $r=18.0,  19.0$ and  $20.0$.   The star-forming  sample  always has  a
690: limiting magnitude of 17.6.
691: 
692: \begin{figure}
693: \centerline{\psfig{figure=f11.ps,clip=true,width=0.5\textwidth}}
694: \caption{Same as the  right-hand panel of Figure~\ref{fig:counts}, but
695: for 289 galaxies  that have the highest specific  star formation rates
696: ($\log_{10}(SFR/M_\ast)>-8.8$).   Results are  shown  only for  scales
697: below 100kpc.}
698: \label{fig:counts_highest_bin}
699: \end{figure}
700: 
701: \begin{figure*}
702: \centerline{\psfig{figure=f12a.ps,clip=true,width=\textwidth}}
703: \centerline{\psfig{figure=f12b.ps,clip=true,width=\textwidth}}
704: \caption{SDSS  optical image  for  20 star-forming  galaxies that  are
705: included  in   the  subsample  of  highest   specific  star  formation
706: rate($\log_{10}(SFR/M_\ast)>-8.8$).  Images are  shown for 10 galaxies
707: that are classified as "mergers" (top panels) and for 10 galaxies that
708: are classified as "no mergers" (bottom panels).}
709: \label{fig:images}
710: \end{figure*}
711: 
712: Figure~\ref{fig:counts} shows that  the counts around the star-forming
713: galaxies with  different specific star  formation rates match  well on
714: large scales.  On scales smaller than $\sim 100$ kpc, there are strong
715: trends  in  the number  of  neighbours  as  a function  of  SFR/$M_*$;
716: galaxies with  higher star formation rates  are more likely  to have a
717: near neighbour.
718: 
719: It is interesting  that the average number of  close neighbours around
720: galaxies with  low-to-average values of $SFR/M_\ast$ is  close to zero
721: on scales  less than  20-30 kpc.   On scales less  than 100  kpc, only
722: around  3\%  of  the galaxies  in  the  lowest  SFR/$M_*$ bin  have  a
723: companion.  This implies that tidal interactions that do not result in
724: enhanced  star formation are  a rare  occurrence.  This  is consistent
725: with the findings  of \citet{DiMatteo-07} who find that  about 85\% of
726: their simulated sample of  interacting galaxies show an enhancement in
727: star formation by a factor $>$ 2.  As the specific star formation rate
728: increases, the average number of close neighbours also rises.  Fifteen
729: percent  of galaxies  in our  highest SFR/$M_*$  bin have  a companion
730: within  20-30 kpc  and  this rises  to a  value  close to  30\% if  we
731: consider  companions  within 100  kpc  from  the  primary galaxy.   In
732: Figure~\ref{fig:counts_highest_bin},  we  show   the  result  for  289
733: galaxies with  the very highest  specific star formation rates  in our
734: sample.  The fraction of galaxies that have a companion within 100 kpc
735: increases to values close to 40\%.  However, it is still true that not
736: {\em every} star-bursting galaxy in our sample has a close companion.
737: 
738: How can we  explain those galaxies with very  high $SFR/M_\ast$ but no
739: close neighbours?  We have  visually examined the SDSS $r$-band images
740: of  160  star-forming  galaxies  that  are  included  in  the  highest
741: $SFR/M_\ast$  subsample but  have  no companions  within  50 kpc.   We
742: classified the  systems according  to whether or  not they  show clear
743: signs of  mergers or interactions,  including double nuclei  and tidal
744: tails.   Three  of   us  (CL,  GK  and  Roderik   Overzier)  did  this
745: independently to make sure that  we obtained the same answer.  We find
746: that more  than half  of such galaxies  show clear evidence  of recent
747: mergers.   In Figure~\ref{fig:images},  we show  some examples  of our
748: classifications.  We  thus conclude that,  at least for  galaxies with
749: very  high  $SFR/M_\ast$, interactions  or  mergers  are the  dominant
750: mechanism for triggering and enhancing their star formation.
751: 
752: \section{Effect of rich environments}
753: 
754: \begin{figure*}
755: \centerline{\psfig{figure=f13.ps,clip=true,width=\textwidth}}
756: \caption{Left: Ope circles show  the fraction of star-forming galaxies
757: that have  at least one companion  in the photometric  sample within a
758: given projected  radius $R_p$. The  solid line shows the  result after
759: the number of companions is corrected using random samples. Right: the
760: fraction  of   "isolated"  (red)  and   "non-isolated"  (blue)  paired
761: galaxies, classified according to  whether they have companions in the
762: spectroscopic reference sample with  the projected separation $r_p$ in
763: the range  100 $h^{-1}$kpc  $< r_p <  $ $R_p$ and  velocity difference
764: smaller than 500 km s$^{-1}$.  The parent sample of paired galaxies is
765: selected from all the high  S/N star-forming galaxies by requring that
766: a paired galaxy  has at least one companion  within a projected radius
767: of  50 $h^{-1}$kpc  in  the photometric  reference  sample limited  at
768: $r_{pho}=19.0$. Solid/dashed lines (open circles/squares) show results
769: obtained with the corrected (observed) companion numbers.}
770: \label{fig:pfraction}
771: \end{figure*}
772: 
773: \begin{figure*}
774: \centerline{\psfig{figure=f14.ps,clip=true,width=\textwidth}}
775: \caption{Distribution of specific star formation rate for the "isolated"
776: (solid) and "non-isolated" (dashed) paired galaxies. In the right-hand panel,
777: the two samples are matched closly in stellar mass.}
778: \label{fig:sfr_hist}
779: \end{figure*}
780: 
781: 
782: Fig.~\ref{fig:sfef}   (\S~\ref{sec:enhancement})    shows   that   the
783: enhancement  in  $\log(SFR/M_\ast)$  on  large  scales  is  not  zero.
784: Rather, it  is constant at  a small but significantly  negative value.
785: We   have  attributed   this  large-scale   bias  to   the  well-known
786: anti-correlation  of the  star formation  rates in  galaxies  with the
787: richness of their local  environment. The question then arises whether
788: the  star formation enhancements  that we  compute may  not be  a true
789: reflection  of the effect  of galaxy-galaxy  interactions, but  may be
790: biased because some close pairs  are not real interacting systems, but
791: are associated with group/cluster environments.
792: 
793: In order  to address this problem,  we have selected a  sample of 9052
794: paired galaxies from all the  high S/N star-forming galaxies. A galaxy
795: is defined  to be  paired if it  has at  least one companion  within a
796: projected radius of 50 $h^{-1}$kpc in the photometric reference sample
797: limited at  $r_{pho}=19.0$. We use  the the photometric  sample rather
798: than the  sepectroscopic one  to select pairs,  so that  the resulting
799: sample is  not biased to  pairs with approximately equal  masses.  The
800: maximum  pair separation  is chosen  to yield  a sample  in  which the
801: contamination   by  chance   projections  is   negligible.    This  is
802: demonstrated in the  left-hand panel of Fig.~\ref{fig:pfraction}.  The
803: fraction of star-forming galaxies that  have at least one companion in
804: the  photometric  sample within  a  given  projected  radius $R_p$  is
805: plotted as a function of $R_p$ as open circles, and is compared to the
806: result (the  solid line) after  the number of companions  is corrected
807: with  the help  of  random samples  (see \S~\ref{sec:enhancement}  for
808: details). As can be seen,  at separations smaller than 50 $h^{-1}$kpc,
809: the correction is negligible.
810: 
811: Next,  We  classify the  paired  galaxies  as  either ``isolated''  or
812: ``non-isolated''  according to  whether  they have  companions in  the
813: spectroscopic reference sample with  projected separation $r_p$ in the
814: range 100 $h^{-1}$kpc $< r_p < $ $R_p$ and velocity difference smaller
815: than 500 km s$^{-1}$. A galaxy is isolated if this ring-like region is
816: completely empty.  In contrast, a  non-isolated galaxy is  required to
817: have at least one companion in  the annulus that is brighter than that
818: galaxy    in     the    $r$-band.    The     right-hand    panel    of
819: Fig.~\ref{fig:pfraction} shows how the  fraction of these two types of
820: paired galaxies changes as  the maximum separation $R_p$ increases. We
821: see that  the sample is  always dominated by isolated  pairs. However,
822: even a  small fraction of  non-isolated galaxies may still  change the
823: enhancement function significantly,  because the quantities defined in
824: equation 1  are pair-weighted and galaxies in  richer environment have
825: more companions. In the  left-hand panel of Fig.~\ref{fig:sfr_hist} we
826: compare  the distribution  of specific  star formation  rates  for the
827: isolated and the non-isolated  samples. The right-hand panel shows the
828: results after  the two samples are  matched in stellar  mass with each
829: other (This  is important because massive galaxies  are more clustered
830: than  less massive  galaxies  \citep{Li-06a}).  As  can  be seen,  the
831: effect of  a rich  environment is quite  small.  The  average specific
832: star formation  rate of the isolated  sample differs from  that of the
833: non-isolated sample by  $\la$ 0.03 dex only.  This  is consistent with
834: the findings of \citet{Balogh-04} who showed that the main effect of a
835: dense  environment on  large  scales  is that  the  {\em fraction}  of
836: emission-line  galaxies  decreases,   but  that  the  distribution  of
837: equivalent  widths  of  H$\alpha$  among the  star-forming  population
838: remains virtually  constant.  Here we  show that the same  thing holds
839: for galaxies with  close pairs. We thus conclude  that the large-scale
840: environment does  not affect the measured  star formation enhancements
841: in our  sample of high S/N  star-forming galaxies on  scales below 100
842: $h^{-1}$kpc.
843: 
844: \section{Summary and Discussion}
845: 
846: We  find  that  the  clustering  amplitude of  high  S/N  star-forming
847: galaxies  depends strongly  on  the specific  star  formation rate  on
848: scales less than 100  kpc. The clustering amplitude increases smoothly
849: as a function of SFR/$M_*$ and the increase in amplitude is largest at
850: the smallest  projected separations.   We interpret this  behaviour as
851: the signature of  tidal interactions, which lead to  inflow of gas and
852: an enhancement in star formation  in the two interacting galaxies.  At
853: low  values of  SFR/$M_*$, the  clustering amplitude  again increases.
854: The increase occurs on {\em all scales} and probably reflects the fact
855: that star formation  in galaxies switches off after  they are accreted
856: onto larger structures such as groups and clusters.
857: 
858: We have explored how the  average star formation rates of galaxies are
859: enhanced  as   a  function  of  the  projected   separation  of  their
860: companions.   The  enhancement is  a  strong  function of  separation,
861: increasing  from zero  at  $r_p >  100  kpc$ to  factors  of 1.5-4  at
862: $r_p=20$ kpc.   We find  that the enhancement  at given  separation is
863: stronger  for  lower mass  galaxies.   Remarkably,  we  find that  the
864: enhancement has almost no dependence on the relative luminosity of the
865: companions.
866: 
867: The tidal force  between two objects is expected  to scale as $d^{-3}$
868: and $m/M$,  where $d$ is  the separation and  $m/M$ is the  mass ratio
869: between the two  objects, so it is perhaps not  surprising that we see
870: stronger star  formation enhancement as a function  of separation than
871: as a  function of  mass ratio.  Nevertheless,  in this study,  we find
872: that the effect  of a companion that is 3  magnitudes fainter than the
873: primary galaxy is  very similar to that of a companion  that is only a
874: factor of 3  less luminous than the primary.   This is quite startling
875: and is worthy of further investigation.
876: 
877: In  order   to  explore  whether  tidal  interactions   are  not  only
878: sufficient, but also a necessary condition for enhanced star formation
879: in a  galaxy, we have computed background  subtracted neighbour counts
880: around the galaxies in our sample.  We find that the average number of
881: galaxies around galaxies  with low values of SFR/$M_*$  is very small.
882: At the very  highest specific star formation rates,  more than 40\% of
883: the galaxies in our sample  have a companion within a projected radius
884: of 100 kpc.  Visual inspection of the high  SFR/$M_*$ galaxies without
885: companions reveals that more than  50\% of these are clear interacting
886: or merging systems.  We thus  conclude that tidal interactions are the
887: primary mechanism for inducing the  highest rates of star formation in
888: galaxies in the local Universe.
889: 
890: Finally, we find clear evidence  that tidal interactions not only lead
891: to  enhanced star  formation,  but also  cause  structural changes  in
892: galaxies.  Many of  the  most strongly  star-forming  galaxies in  our
893: sample  have   concentration  indices  similar  to   those  of  normal
894: early-type galaxies.  We note that the concentration index is measured
895: for the $r$-band  light and not for the stellar  mass, so more careful
896: analysis  is  needed  before  one  can definitely  conclude  that  the
897: interactions  will result in  the formation  of a  galaxy with  a high
898: bulge mass fraction. Nevertheless, we conclude that our results are in
899: general  accord  with  the  theoretical  picture  first  laid  out  by
900: \citet{Toomre-Toomre-72},  which showed  that galaxy  interactions can
901: lead to the growth of bulges and spheroids in galaxies.
902: 
903: \section*{Acknowledgements} 
904: 
905: CL is  supported by the Joint Postdoctoral  Programme in Astrophysical
906: Cosmology  of  Max  Planck  Institute for  Astrophysics  and  Shanghai
907: Astronomical Observatory.  CL and YPJ are supported by NSFC (10533030,
908: 10643005, 10633020),  by the Knowledge Innovation Program  of CAS (No.
909: KJCX2-YW-T05), and by 973  Program (No.2007CB815402).  We are grateful
910: to the referee for his/her comments which have helped to improve the paper,
911: and Roderik Overzier for  visually examining and classifying the images
912: of our  galaxies.  CL, GK and  SW would like to  thank the hospitality
913: and stimulating atmosphere of the  Aspen Center for Physics while this
914: work was being completed.
915: 
916: Funding for  the SDSS and SDSS-II  has been provided by  the Alfred P.
917: Sloan Foundation, the Participating Institutions, the National Science
918: Foundation, the  U.S.  Department of Energy,  the National Aeronautics
919: and Space Administration, the  Japanese Monbukagakusho, the Max Planck
920: Society, and  the Higher Education  Funding Council for  England.  The
921: SDSS Web  Site is  http://www.sdss.org/.  The SDSS  is managed  by the
922: Astrophysical    Research    Consortium    for    the    Participating
923: Institutions. The  Participating Institutions are  the American Museum
924: of  Natural History,  Astrophysical Institute  Potsdam,  University of
925: Basel,   Cambridge  University,   Case  Western   Reserve  University,
926: University of Chicago, Drexel  University, Fermilab, the Institute for
927: Advanced   Study,  the  Japan   Participation  Group,   Johns  Hopkins
928: University, the  Joint Institute  for Nuclear Astrophysics,  the Kavli
929: Institute  for   Particle  Astrophysics  and   Cosmology,  the  Korean
930: Scientist Group, the Chinese  Academy of Sciences (LAMOST), Los Alamos
931: National  Laboratory, the  Max-Planck-Institute for  Astronomy (MPIA),
932: the  Max-Planck-Institute  for Astrophysics  (MPA),  New Mexico  State
933: University,   Ohio  State   University,   University  of   Pittsburgh,
934: University  of  Portsmouth, Princeton  University,  the United  States
935: Naval Observatory, and the University of Washington.
936: 
937: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
938: \begin{thebibliography}{}
939: 
940: \bibitem[\protect\citeauthoryear{{Adelman-McCarthy},     {Ag{\"u}eros},
941: {Allam}, {Anderson},  {Anderson}, {Annis}, {Bahcall},  {Baldry} \& {et
942: al.,}}{{Adelman-McCarthy}           et~al.}{2006}]{Adelman-McCarthy-06}
943: {Adelman-McCarthy}   J.~K.,   {Ag{\"u}eros}   M.~A.,  {Allam}   S.~S.,
944: {Anderson}  K.~S.~J., {Anderson} S.~F.,  {Annis} J.,  {Bahcall} N.~A.,
945: {Baldry} I.~K., {et al.,} 2006, \apjs, 162, 38
946: 
947: \bibitem[\protect\citeauthoryear{{Arp}}{{Arp}}{1966}]{Arp-66}     {Arp}
948: H.,  1966,   {Atlas  of  peculiar   galaxies}.   Pasadena:  California
949: Inst.~Technology, 1966
950: 
951: \bibitem[\protect\citeauthoryear{{Baldwin},        {Phillips}       \&
952: {Terlevich}}{{Baldwin}    et~al.}{1981}]{Baldwin-Phillips-Terlevich-81}
953: {Baldwin} J.~A., {Phillips} M.~M., {Terlevich} R., 1981, \pasp, 93, 5
954: 
955: \bibitem[\protect\citeauthoryear{Balogh et al.}{2004}]{Balogh-04} 
956: Balogh M., et al., 2004, MNRAS, 348, 1355 
957: 
958: \bibitem[\protect\citeauthoryear{{Barnes}  \& {Hernquist}}{{Barnes} \&
959: {Hernquist}}{1992}]{Barnes-Hernquist-92}  {Barnes}  J.~E., {Hernquist}
960: L., 1992, \araa, 30, 705
961: 
962: \bibitem[\protect\citeauthoryear{{Barrow},         {Bhavsar}        \&
963: {Sonoda}}{{Barrow}  et~al.}{1984}]{Barrow-Bhavsar-Sonoda-84}  {Barrow}
964: J.~D., {Bhavsar} S.~P., {Sonoda} D.~H., 1984, \mnras, 210, 19P
965: 
966: \bibitem[\protect\citeauthoryear{{Barton},     {Arnold},    {Zentner},
967: {Bullock}  \& {Wechsler}}{{Barton}  et~al.}{2007}]{Barton-07} {Barton}
968: E.~J.,  {Arnold} J.~A., {Zentner}  A.~R., {Bullock}  J.~S., {Wechsler}
969: R.~H., 2007, ArXiv e-prints, 708
970: 
971: \bibitem[\protect\citeauthoryear{{Barton},         {Geller}         \&
972: {Kenyon}}{{Barton}   et~al.}{2000}]{Barton-Geller-Kenyon-00}  {Barton}
973: E.~J., {Geller} M.~J., {Kenyon} S.~J., 2000, \apj, 530, 660
974: 
975: \bibitem[\protect\citeauthoryear{{Bergvall},      {Laurikainen}     \&
976: {Aalto}}{{Bergvall}       et~al.}{2003}]{Bergvall-Laurikainen-Aalto-03}
977: {Bergvall} N., {Laurikainen} E., {Aalto} S., 2003, \aap, 405, 31
978: 
979: \bibitem[\protect\citeauthoryear{{Blanton},  {Lin}, {Lupton}, {Maley},
980: {Young},  {Zehavi} \&  {Loveday}}{{Blanton} et~al.}{2003}]{Blanton-03}
981: {Blanton} M.~R., {Lin} H.,  {Lupton} R.~H., {Maley} F.~M., {Young} N.,
982: {Zehavi} I., {Loveday} J., 2003, \aj, 125, 2276
983: 
984: \bibitem[\protect\citeauthoryear{{Blanton},   {Schlegel},   {Strauss},
985: {Brinkmann},   {Finkbeiner},  {Fukugita},   {Gunn},   {Hogg}  \&   {et
986: al.,}}{{Blanton}     et~al.}{2005}]{Blanton-05}    {Blanton}    M.~R.,
987: {Schlegel}  D.~J., {Strauss} M.~A.,  {Brinkmann} J.,  {Finkbeiner} D.,
988: {Fukugita} M., {Gunn}  J.~E., {Hogg} D.~W., {et al.,}  2005, \aj, 129,
989: 2562
990: 
991: \bibitem[\protect\citeauthoryear{{Brinchmann},   {Charlot},   {White},
992: {Tremonti},   {Kauffmann},   {Heckman}  \&   {Brinkmann}}{{Brinchmann}
993: et~al.}{2004}]{Brinchmann-04}  {Brinchmann} J., {Charlot}  S., {White}
994: S.~D.~M., {Tremonti} C., {Kauffmann} G., {Heckman} T., {Brinkmann} J.,
995: 2004, \mnras, 351, 1151
996: 
997: \bibitem[\protect\citeauthoryear{{Brosch},        {Almoznino}       \&
998: {Heller}}{{Brosch} et~al.}{2004}]{Brosch-Almoznino-Heller-04} {Brosch}
999: N., {Almoznino} E., {Heller} A.~B., 2004, \mnras, 349, 357
1000: 
1001: \bibitem[\protect\citeauthoryear{{Bushouse}}{{Bushouse}}{1986}]{Bushouse-86}
1002: {Bushouse} H.~A., 1986, \aj, 91, 255
1003: 
1004: \bibitem[\protect\citeauthoryear{{Bushouse},        {Werner}        \&
1005: {Lamb}}{{Bushouse}  et~al.}{1988}]{Bushouse-Werner-Lamb-88} {Bushouse}
1006: H.~A., {Werner} M.~W., {Lamb} S.~A., 1988, \apj, 335, 74
1007: 
1008: \bibitem[\protect\citeauthoryear{{Cox},    {Jonsson},   {Primack}   \&
1009: {Somerville}}{{Cox} et~al.}{2006}]{Cox-06}  {Cox} T.~J., {Jonsson} P.,
1010: {Primack} J.~R., {Somerville} R.~S., 2006, \mnras, 373, 1013
1011: 
1012: \bibitem[\protect\citeauthoryear{{Di Matteo},  {Combes}, {Melchior} \&
1013: {Semelin}}{{Di  Matteo}  et~al.}{2007}]{DiMatteo-07}  {Di Matteo}  P.,
1014: {Combes} F.,  {Melchior} A.~., {Semelin} B.,  2007, ArXiv Astrophysics
1015: e-prints
1016: 
1017: \bibitem[\protect\citeauthoryear{{Ellison},   {Patton},   {Simard}  \&
1018: {McConnachie}} {{Ellison}  et~al.}{2008}]{Ellison-08} {Ellison} S.~L.,
1019: {Patton} D.~R., {Simard} L., {McConnachie} A.~W., 2008, submitted
1020: 
1021: \bibitem[\protect\citeauthoryear{{Fukugita},    {Ichikawa},    {Gunn},
1022: {Doi},          {Shimasaku}         \&         {Schneider}}{{Fukugita}
1023: et~al.}{1996}]{Fukugita-96}  {Fukugita}   M.,  {Ichikawa}  T.,  {Gunn}
1024: J.~E., {Doi}  M., {Shimasaku} K.,  {Schneider} D.~P., 1996,  \aj, 111,
1025: 1748
1026: 
1027: \bibitem[\protect\citeauthoryear{{Gunn},       {Carr},      {Rockosi},
1028: {Sekiguchi},   {Berry},  {Elms},  {de   Haas},  {Ivezi{\'c}}   \&  {et
1029: al.,}}{{Gunn}   et~al.}{1998}]{Gunn-98}  {Gunn}   J.~E.,   {Carr}  M.,
1030: {Rockosi} C.,  {Sekiguchi} M.,  {Berry} K., {Elms}  B., {de  Haas} E.,
1031: {Ivezi{\'c}} {\v Z}., {et al.,} 1998, \aj, 116, 3040
1032: 
1033: \bibitem[\protect\citeauthoryear{{Gunn},     {Siegmund},    {Mannery},
1034: {Owen},  {Hull},  {Leger},   {Carey},  {Knapp}  \&  {et  al.,}}{{Gunn}
1035: et~al.}{2006}]{Gunn-06}  {Gunn}  J.~E.,  {Siegmund}  W.~A.,  {Mannery}
1036: E.~J.,  {Owen}  R.~E., {Hull}  C.~L.,  {Leger}  R.~F., {Carey}  L.~N.,
1037: {Knapp} G.~R., {et al.,} 2006, \aj, 131, 2332
1038: 
1039: \bibitem[\protect\citeauthoryear{{Hogg},  {Finkbeiner},  {Schlegel} \&
1040: {Gunn}}{{Hogg}  et~al.}{2001}]{Hogg-01}   {Hogg}  D.~W.,  {Finkbeiner}
1041: D.~P., {Schlegel} D.~J., {Gunn} J.~E., 2001, \aj, 122, 2129
1042: 
1043: \bibitem[\protect\citeauthoryear{{Ivezi{\'c}},  {Lupton},  {Schlegel},
1044: {Boroski},  {Adelman-McCarthy}, {Yanny},  {Kent},  {Stoughton} \&  {et
1045: al.,}}{{Ivezi{\'c}}  et~al.}{2004}]{Ivezic-04}  {Ivezi{\'c}} {\v  Z}.,
1046: {Lupton}  R.~H., {Schlegel} D.,  {Boroski} B.,  {Adelman-McCarthy} J.,
1047: {Yanny} B.,  {Kent} S., {Stoughton} C., {et  al.,} 2004, Astronomische
1048: Nachrichten, 325, 583
1049: 
1050: \bibitem[\protect\citeauthoryear{{Kapferer},   {Knapp},   {Schindler},
1051: {Kimeswenger} \&  {van Kampen}}{{Kapferer} et~al.}{2005}]{Kapferer-05}
1052: {Kapferer}  W., {Knapp}  A.,  {Schindler} S.,  {Kimeswenger} S.,  {van
1053: Kampen} E., 2005, \aap, 438, 87
1054: 
1055: \bibitem[\protect\citeauthoryear{{Kauffmann},    {White},   {Heckman},
1056: {M{\'e}nard},      {Brinchmann},     {Charlot},      {Tremonti}     \&
1057: {Brinkmann}}{{Kauffmann}  et~al.}{2004}]{Kauffmann-04} {Kauffmann} G.,
1058: {White} S.~D.~M.,  {Heckman} T.~M., {M{\'e}nard}  B., {Brinchmann} J.,
1059: {Charlot} S., {Tremonti} C., {Brinkmann} J., 2004, \mnras, 353, 713
1060: 
1061: \bibitem[\protect\citeauthoryear{{Keel}}{{Keel}}{1991}]{Keel-91}
1062: {Keel}  W.~C., 1991,  in {Combes}  F., {Casoli}  F., eds,  Dynamics of
1063: Galaxies  and  Their  Molecular  Cloud Distributions  Vol.~146,  {Star
1064: Formation and Galaxy Interactions}.  pp 243--+
1065: 
1066: \bibitem[\protect\citeauthoryear{{Keel}, {Kennicutt}  Jr., {Hummel} \&
1067: {van   der   Hulst}}{{Keel}   et~al.}{1985}]{Keel-85}  {Keel}   W.~C.,
1068: {Kennicutt} Jr. R.~C., {Hummel} E.,  {van der Hulst} J.~M., 1985, \aj,
1069: 90, 708
1070: 
1071: \bibitem[\protect\citeauthoryear{{Kennicutt} Jr., {Roettiger}, {Keel},
1072: {van der  Hulst} \& {Hummel}}{{Kennicutt} et~al.}{1987}]{Kennicutt-87}
1073: {Kennicutt}  Jr.  R.~C., {Roettiger}  K.~A.,  {Keel}  W.~C., {van  der
1074: Hulst} J.~M., {Hummel} E., 1987, \aj, 93, 1011
1075: 
1076: \bibitem[\protect\citeauthoryear{{Lambas},   {Tissera},   {Alonso}  \&
1077: {Coldwell}}{{Lambas}    et~al.}{2003}]{Lambas-03}    {Lambas}   D.~G.,
1078: {Tissera}  P.~B., {Alonso}  M.~S., {Coldwell}  G., 2003,  \mnras, 346,
1079: 1189
1080: 
1081: \bibitem[\protect\citeauthoryear{{Larson}  \&  {Tinsley}}{{Larson}  \&
1082: {Tinsley}}{1978}]{Larson-Tinsley-78} {Larson}  R.~B., {Tinsley} B.~M.,
1083: 1978, \apj, 219, 46
1084: 
1085: \bibitem[\protect\citeauthoryear{{Li},  {Kauffmann},  {Jing}, {White},
1086: {B{\"o}rner}   \&  {Cheng}}{{Li}   et~al.}{2006a}]{Li-06a}   {Li}  C.,
1087: {Kauffmann}  G.,  {Jing}  Y.~P.,  {White} S.~D.~M.,  {B{\"o}rner}  G.,
1088: {Cheng} F.~Z., 2006a, \mnras, 368, 21
1089: 
1090: \bibitem[\protect\citeauthoryear{{Li},  {Kauffmann},  {Wang}, {White},
1091: {Heckman} \& {Jing}}{{Li} et~al.}{2006b}]{Li-06b} {Li} C., {Kauffmann}
1092: G., {Wang} L., {White} S.~D.~M., {Heckman} T.~M., {Jing} Y.~P., 2006b,
1093: \mnras, 373, 457
1094: 
1095: \bibitem[\protect\citeauthoryear{{Lupton},    {Gunn},    {Ivezi{\'c}},
1096: {Knapp}  \&  {Kent}}{{Lupton}  et~al.}{2001}]{Lupton-01} {Lupton}  R.,
1097: {Gunn}  J.~E., {Ivezi{\'c}}  Z., {Knapp}  G.~R., {Kent}  S.,  2001, in
1098: {Harnden} Jr.  F.~R., {Primini}  F.~A., {Payne} H.,  eds, Astronomical
1099: Data  Analysis Software  and  Systems X  Vol.~238,  {The SDSS  Imaging
1100: Pipelines}.  pp 269--+
1101: 
1102: \bibitem[\protect\citeauthoryear{{Meza},   {Navarro},  {Steinmetz}  \&
1103: {Eke}}{{Meza}  et~al.}{2003}]{Meza-03}  {Meza}  A.,  {Navarro}  J.~F.,
1104: {Steinmetz} M., {Eke} V.~R., 2003, \apj, 590, 619
1105: 
1106: \bibitem[\protect\citeauthoryear{{Mihos}  \&  {Hernquist}}{{Mihos}  \&
1107: {Hernquist}}{1996}]{Mihos-Hernquist-96} {Mihos} J.~C., {Hernquist} L.,
1108: 1996, \apj, 464, 641
1109: 
1110: \bibitem[\protect\citeauthoryear{{Negroponte} \& {White}}{{Negroponte}
1111: \&   {White}}{1983}]{Negroponte-White-83}  {Negroponte}   J.,  {White}
1112: S.~D.~M., 1983, \mnras, 205, 1009
1113: 
1114: \bibitem[\protect\citeauthoryear{{Nikolic},         {Cullen}        \&
1115: {Alexander}}{{Nikolic}      et~al.}{2004}]{Nikolic-Cullen-Alexander-04}
1116: {Nikolic} B., {Cullen} H., {Alexander} P., 2004, \mnras, 355, 874
1117: 
1118: \bibitem[\protect\citeauthoryear{{Pier},      {Munn},      {Hindsley},
1119: {Hennessy},      {Kent},     {Lupton}      \&     {Ivezi{\'c}}}{{Pier}
1120: et~al.}{2003}]{Pier-03} {Pier} J.~R.,  {Munn} J.~A., {Hindsley} R.~B.,
1121: {Hennessy} G.~S., {Kent} S.~M.,  {Lupton} R.~H., {Ivezi{\'c}} {\v Z}.,
1122: 2003, \aj, 125, 1559
1123: 
1124: \bibitem[\protect\citeauthoryear{{Sanders}  \& {Mirabel}}{{Sanders} \&
1125: {Mirabel}}{1996}]{Sanders-Mirabel-96}   {Sanders}   D.~B.,   {Mirabel}
1126: I.~F., 1996, \araa, 34, 749
1127: 
1128: \bibitem[\protect\citeauthoryear{{Sanders},    {Scoville},    {Young},
1129: {Soifer},     {Schloerb},     {Rice}     \&     {Danielson}}{{Sanders}
1130: et~al.}{1986}]{Sanders-86} {Sanders}  D.~B., {Scoville} N.~Z., {Young}
1131: J.~S.,  {Soifer} B.~T.,  {Schloerb} F.~P.,  {Rice}  W.~L., {Danielson}
1132: G.~E., 1986, \apjl, 305, L45
1133: 
1134: \bibitem[\protect\citeauthoryear{{Schlegel},      {Finkbeiner}      \&
1135: {Davis}}{{Schlegel}        et~al.}{1998}]{Schlegel-Finkbeiner-Davis-98}
1136: {Schlegel} D.~J., {Finkbeiner} D.~P., {Davis} M., 1998, \apj, 500, 525
1137: 
1138: \bibitem[\protect\citeauthoryear{{Smith},     {Struck},     {Hancock},
1139: {Appleton},         {Charmandaris}         \&         {Reach}}{{Smith}
1140: et~al.}{2007}]{Smith-07}  {Smith} B.~J.,  {Struck}  C., {Hancock}  M.,
1141: {Appleton} P.~N.,  {Charmandaris} V.,  {Reach} W.~T., 2007,  \aj, 133,
1142: 791
1143: 
1144: \bibitem[\protect\citeauthoryear{{Smith},       {Tucker},      {Kent},
1145: {Richmond},  {Fukugita}, {Ichikawa},  {Ichikawa},  {Jorgensen} \&  {et
1146: al.,}}{{Smith} et~al.}{2002}]{Smith-02} {Smith} J.~A., {Tucker} D.~L.,
1147: {Kent} S., {Richmond} M.~W.,  {Fukugita} M., {Ichikawa} T., {Ichikawa}
1148: S.-i., {Jorgensen} A.~M., {et al.,} 2002, \aj, 123, 2121
1149: 
1150: \bibitem[\protect\citeauthoryear{{Solomon}   \&  {Sage}}{{Solomon}  \&
1151: {Sage}}{1988}]{Solomon-Sage-88}  {Solomon} P.~M., {Sage}  L.~J., 1988,
1152: \apj, 334, 613
1153: 
1154: \bibitem[\protect\citeauthoryear{{Springel}}{{Springel}}{2000}]{Springel-00}
1155: {Springel} V., 2000, \mnras, 312, 859
1156: 
1157: \bibitem[\protect\citeauthoryear{{Stoughton},   {Lupton},  {Bernardi},
1158: {Blanton},  {Burles},  {Castander},  {Connolly}, {Eisenstein}  \&  {et
1159: al.,}}{{Stoughton}    et~al.}{2002}]{Stoughton-02}   {Stoughton}   C.,
1160: {Lupton}   R.~H.,  {Bernardi}  M.,   {Blanton}  M.~R.,   {Burles}  S.,
1161: {Castander}  F.~J., {Connolly}  A.~J., {Eisenstein}  D.~J.,  {et al.,}
1162: 2002, \aj, 123, 485
1163: 
1164: \bibitem[\protect\citeauthoryear{{Strauss},    {Weinberg},   {Lupton},
1165: {Narayanan},   {Annis},  {Bernardi},   {Blanton},   {Burles}  \&   {et
1166: al.,}}{{Strauss}     et~al.}{2002}]{Strauss-02}    {Strauss}    M.~A.,
1167: {Weinberg}  D.~H.,  {Lupton}  R.~H.,  {Narayanan} V.~K.,  {Annis}  J.,
1168: {Bernardi} M.,  {Blanton} M., {Burles}  S., {et al.,} 2002,  \aj, 124,
1169: 1810
1170: 
1171: \bibitem[\protect\citeauthoryear{{Struck}}{{Struck}}{1999}]{Struck-99}
1172: {Struck} C., 1999, \physrep, 321, 1
1173: 
1174: %\bibitem[\protect\citeauthoryear{{Struck}}{{Struck}}{2006}]{Struck-06}
1175: %{Struck} C., 2006, ArXiv Astrophysics e-prints
1176: 
1177: \bibitem[\protect\citeauthoryear{{Tinney},  {Scoville},  {Sanders}  \&
1178: {Soifer}}{{Tinney}     et~al.}{1990}]{Tinney-90}    {Tinney}    C.~G.,
1179: {Scoville} N.~Z.,  {Sanders} D.~B.,  {Soifer} B.~T., 1990,  \apj, 362,
1180: 473
1181: 
1182: \bibitem[\protect\citeauthoryear{{Tissera},
1183: {Dom{\'{\i}}nguez-Tenreiro},  {Scannapieco}  \&  {S{\'a}iz}}{{Tissera}
1184: et~al.}{2002}]{Tissera-02}               {Tissera}              P.~B.,
1185: {Dom{\'{\i}}nguez-Tenreiro} R., {Scannapieco} C., {S{\'a}iz} A., 2002,
1186: \mnras, 333, 327
1187: 
1188: \bibitem[\protect\citeauthoryear{{Toomre}   \&  {Toomre}}{{Toomre}  \&
1189: {Toomre}}{1972}]{Toomre-Toomre-72}  {Toomre}  A.,  {Toomre} J.,  1972,
1190: \apj, 178, 623
1191: 
1192: \bibitem[\protect\citeauthoryear{{Tucker},     {Kent},     {Richmond},
1193: {Annis}, {Smith},  {Allam}, {Rodgers}, {Stute}  \& {et al.,}}{{Tucker}
1194: et~al.}{2006}]{Tucker-06} {Tucker} D.~L., {Kent} S., {Richmond} M.~W.,
1195: {Annis}  J., {Smith}  J.~A., {Allam}  S.~S., {Rodgers}  C.~T., {Stute}
1196: J.~L., {et al.,} 2006, Astronomische Nachrichten, 327, 821
1197: 
1198: \bibitem[\protect\citeauthoryear{{Woods}   \&   {Geller}}{{Woods}   \&
1199: {Geller}}{2007}]{Woods-Geller-07} {Woods} D.~F., {Geller} M.~J., 2007,
1200: \aj, 134, 527
1201: 
1202: \bibitem[\protect\citeauthoryear{{Woods},          {Geller}         \&
1203: {Barton}}{{Woods}    et~al.}{2006}]{Woods-Geller-Barton-06}    {Woods}
1204: D.~F., {Geller} M.~J., {Barton} E.~J., 2006, \aj, 132, 197
1205: 
1206: \bibitem[\protect\citeauthoryear{{York},  {Adelman},  {Anderson}  Jr.,
1207: {Anderson},   {Annis},  {Bahcall},   {Bakken},  {Barkhouser}   \&  {et
1208: al.,}}{{York}  et~al.}{2000}]{York-00}  {York}  D.~G.,  {Adelman}  J.,
1209: {Anderson} Jr.  J.~E., {Anderson} S.~F., {Annis}  J., {Bahcall} N.~A.,
1210: {Bakken} J.~A., {Barkhouser} R., {et al.,} 2000, \aj, 120, 1579
1211: 
1212: \bibitem[\protect\citeauthoryear{{Young},  {Allen}, {Kenney}, {Lesser}
1213: \&  {Rownd}}{{Young} et~al.}{1996}]{Young-96}  {Young}  J.~S., {Allen}
1214: L., {Kenney} J.~D.~P., {Lesser} A., {Rownd} B., 1996, \aj, 112, 1903
1215: 
1216: \bibitem[\protect\citeauthoryear{{Young},     {Kenney},     {Tacconi},
1217: {Claussen},  {Huang}, {Tacconi-Garman},  {Xie}  \& {Schloerb}}{{Young}
1218: et~al.}{1986}]{Young-86} {Young} J.~S.,  {Kenney} J.~D., {Tacconi} L.,
1219: {Claussen}  M.~J.,  {Huang}  Y.-L.,  {Tacconi-Garman}  L.,  {Xie}  S.,
1220: {Schloerb} F.~P., 1986, \apjl, 311, L17
1221: 
1222: \end{thebibliography}
1223: 
1224: 
1225: \bsp
1226: \label{lastpage}
1227: 
1228: \end{document}
1229: 
1230: