Star Cluster Formation and Star Formation: The Role of Environment and Star Formation Efficiencies

0.15 MB·English

by  Uta Fritze

AstrophysicsandSpaceScience
DOI10.1007/s•••••-•••-••••-•
Star Cluster Formation and Star Formation:
The Role of Environment and Star Formation Efficiencies
Uta Fritze
8
0
0
2

n
a (cid:13)c Springer-Verlag••••
J

5 Abstract Analyzingglobalstarburstpropertiesinvar- the older SCs are already gone, dissolved and/or faded
1 ious kinds of starburst and post-starburstgalaxies and belowdetection. Itisthereforeimportanttotakethese

relating them to the properties of the star cluster pop- processesintoaccountwhencomparingSCpopulations
]
h ulations they form,I explorethe conditions for the for- indifferentgalaxies. Theydependontheinitialproper-
p mation of massive, compact, long-lived star clusters. ties of individual SCs, their masses, radii, abundances,
-
o The aim is to find out whether the relative amount of stellar IMF, and of the SC population, i.e. the lumi-
r star formation that goes into star cluster formation as nosityfunction,themassfunction,distributionofradii,
t
s opposedto field star formation,and into the formation ages, etc.
a
[ of massive long-livedclusters in particular, is universal That SC formation is an important mode of SF in
or scales with star formation rate, burst strength, star starbursts was shown on the example of the Tadpole
1 formationefficiency,galaxyorgasmass,andwhetheror and Mice interacting galaxies. A pixel-by-pixel anal-
v
not there are special conditions or some threshold for
8 ysis of ACS data (BVRI) with GALEV evolutionary
9 the formation of star clusters that merit to be called synthesis models showed that ∼70 % of the blue light
2 globular clusters a few gigayearslater. is emitted by YSCs as opposed to only ∼ 30 % com-
2
ing from field stars. We estimated that more than 35
. Keywords starformation;starclusterformation;star
1 % of all SF went into the formation of YSCs, not only
0 formation efficiencies; environment
in the main bodies of these two galaxies, but all along
8
0 theirextendedtidaltails(de Grijs et al.2003). Clearly,
: 1 Motivation this analysis needs to be extended to different types of
v
i galaxies, starburst/non-starburst, dwarf/normal, gas-
X Star Cluster (SC) formation is a major or even dom- rich/gas-poor, interacting/non-interacting, in various
r inant mode of all star formation (SF) and occurs in stages of the interaction, etc. to explore the system-
a
very different environments. This immediately raises atics. ThereareindicationsfromMeurer et al.(1995)’s
the question whether young star clusters (YSCs) form- work that the contribution from YSCs to the UV light
ing in different environments are similar or different. ofagalaxyincreaseswithincreasingUVsurfacebright-
SCs are not only interesting in their own right, but ness, which itself is a measure of SF intensity. A burn-
bearconsiderablepowerastracersofSFintheirparent
ing questionthat we arecurrently exploringis whether
galaxies. YSCs trace the spatialdistribution ofSF and
theamountofSFthatgoesintomassivelong-livedSCs
its recent history within a galaxy, old globular clusters
inrelationtotheamountofSFthatgoesintolow-mass
(GCs) trace violent SF phases in their parent galaxy –
short-livedSCsandfieldstars,increaseswithincreasing
all the way back to the very onset of SF, i.e. over a
overall SFR, with bursts strength, or with star forma-
Hubble time. But SCs also fade and dissolve. In ac-
tion efficiency (SFE). If so, the immediate next ques-
tively starforming galaxies,the youngestSCsmaystill
tionis whether this is acontinuousincreaseorrathera
be embedded in their natal dust clouds while part of
threshold effect in the sense that only above a certain
SFintensityorefficiencythemassivelong-livedSCsare
UtaFritze formed that later are called GCs.
UniversityofHertfordshire SCs form in very different environments, in normal
star-forming galaxies, as e.g. M51 or NGC 5236 (cf.
2
Larsen(2004),Mora et al.(2007),indwarfgalaxystar- stages of massive gas-rich mergers with strong nuclear
bursts like NGC 1569(Anders et al. 2004), in interact- (few 100 pc) starbursts. In those ULIRGs, almost all
ing gas-rich galaxies like NGC 4038/39 (the Anten- the molecular gas in the central starburst region is at
nae) (Whitmore et al. 1995, 2005; Anders et al. 2007). the high densities of molecular cloud cores, indicating
Slightly older and intermediate-age SCs are observed that the molecular cloud structure must be very dif-
in post-starburst merger remnants like NGC 7252 ferent from what we know in our Galaxy. The entire
(Whitmore et al.1993;Fritze-v. Alvensleben and Burkert nuclear region is just one supergiant molecular cloud
1995; Miller et al. 1997; Schweizer and Seitzer 1998) core,seriouslyraisingthequestionwhetherthestarand
and dynamically young ellipticals like NGC 1316 SC formation processes can be the same as in normal
(Goudfrooij et al.2001,2004,2007),respectively. They galaxies, not to mention the situation in extended, ex-
can form all over the main body of a galaxy, as e.g. in panding,low-densitytidalstructuresintheoutskirtsof
the Antennae or NGC 7252,in and around a starburst other interacting galaxies.
nucleus, as e.g. in Arp 220 or NGC 6240 (Shioya et al. In any case, before ALMA becomes operational,the
2001; Pasquali et al. 2003), all along some – but YSCs forming in these different types of environments
not all – extended tidal features (Knierman et al. are our best proxy to the molecular cloud structure.
2003; de Grijs et al. 2003; Trancho et al. 2007), as In the Milky Way, molecular cloud cores, molecular
well as in group environments like Stephan’s quintett clouds, and YSCs all feature power-law mass func-
(Gallagher et al. 2001). These environments cover a tions, suggesting scale-free self-similar evolution. Not
huge range in terms of density, kinetic temperature, even for the closest massive merger, the Antennae, can
chemical abundances and it is by no means obvious we presently determine the molecular cloud or cloud
whether or not all these SCs are similar or different, core mass functions (cf. Wilson et al. (2003)). The
individually or as a population. Related questions are masses of YSCs and the shape of their mass function
where,when,andhowGCareformedandwhatayoung is all we can access (Wilson et al. 2006; Anders et al.
GC looks like. Or how to tell apart YSCs into long- 2007). Gao and Solomon (2004) and Solomon et al.
lived and short-lived ones – by mass, concentration, (1992) have shown that for all galaxies – from Blue
mass function, ...? Compact Dwarfs to spirals and ULIRGs – there is a
Currentcluster formation models require exception- tight correlation between SFR, as derived from far-
ally high SF efficiencies SFE := M∗/Mgas >30 % as infrared luminosity, and the mass in molecular cloud
a prerequisite for the formation of massive strongly cores, as derived from the HCN luminosity. They also
bound and long-term stable SCs, i.e. for the forma- find the SF efficiency to be proportional to the mass
tion of young GCs (Brown et al. 1995; Burkert et al. ratioof moleculargas atcore andnormaldensities, i.e.
1996;Elmegreen and Efremov1997;Li et al.2004). On to the ratio between HCN or CS luminosity and CO
aglobalscale,SFefficienciesinnormalspiralandirreg- luminosity. The highest density molecular gas in all
ular galaxies, as well as in starbursting dwarf galaxies these environments is transformed into stars with al-
are of order 0.1 − 3% (Krueger et al. 1995). On the most 100 % efficiency and it is the amount of gas at
smallerscaleofmolecularcloudsinthe MilkyWay,the those high densities that controls SF. The fraction of
SF efficiency is of the same order of magnitude and molecular gas at the highest densities therefore defines
so is the mass ratio between the molecular cloud core the SF efficiency.
and the entire molecular cloud. No GC formation is The highambientpressurebuilding upinthe course
thereforeexpectedinspirals,irregularsorstar-bursting ofmassivegas-richmergerscandriveup SFefficiencies
dwarf galaxies by today. In giant gas-rich interacting by 1−2 orders of magnitude by compressing molecu-
galaxies, on the other hand, SF efficiencies of order lar clouds, increasing their masses and, in particular,
10−50% are reported on global scales, and of order their core mass fractions. Jog and Das (1992, 1996)
30−90% on nuclear scales of a few hundred pc up to have shown that the ISM pressure during mergers can
∼ 1 kpc. In those systems, GC formation should be easilybecome 3−4timeshigherthanthe typicalinter-
possible. The fact that different submm lines (CO(1- nal molecular cloud pressure, raising the SF efficiency
0),HCN(1-0),CS(1-0))tracemoleculargasatdifferent to 70−90%. This leads us to expect that the relative
densities (n≥100,n ≥3·104,n ∼105 cm−3), has al- amount of SF that goes into the formation of massive,
lowed to see that while in the Milky Way and other stronglyboundyoungGCsinrelationtotheamountof
nearby galaxies only a small fraction (0.1−3%) of the SF that goes into field stars and low-mass, short-lived
mass of a molecular cloud makes up its high density clusters is enhanced in massive gas-richmergers.
core, the situation is drastically different in Ultralu-
minous Infrared galaxies (ULIRGs), which all are late
StarandStarClusterFormation 3
2 Results so far ... all the SC parameters they return: age, metallicity,
E(B − V), and mass. Extensive tests with artificial
2.1 Analysis method SCs have shown that UV or U−band observations are
essential for age dating of YSCs and a NIR band is
Before I turn to the results obtained so far for SCs important to obtain accurate metallicities. For YSCs
and SC populations in different environments, I briefly in dusty galaxies four passbands including UV/U and
recall our GALEV evolutionary synthesis models and H or K with observational uncertainties ≤ 0.05 mag
the dedicated analysis tools we use in our analysis in the UV/optical and ≤ 0.1 mag in the NIR allow to
of SC systems. GALEV models in the first place largelydisentangleagesandandmetallicitiesandtoob-
describe the spectral evolution of SCs of various tain ages to ∆ age/age ≤ 0.3 and metallicties to ±0.2
metallicities −1.7≤[Fe/H]≤+0.4 over the age range dex. For intermediate-age SCs or old GCs in dust-free
from 4 Myr through 13 Gyr, including gaseous emis- environments, three passbands, again ranging from U
sion, which significantly affects broad band luminosi- or B through H or K are enough (Anders et al. 2004;
ties and colours during early evolutionary stages (see de Grijs et al. 2003).
Anders and Fritze-v. Alvensleben (2003) for details).
Spectra are then folded with filter functions for any 2.2 Star cluster formation in dwarf galaxy starbursts
desired filter system to yield the photometric evolu-
tion. This is important in order to avoid uncertainties Applying our SED analysis tool to HST WFPC2 and
from a posteriori transformations between filter sys- NICMOS archival data for some 170 compact YSCs
tems. Models well reproduce empirical colour – metal- that we identified in the not apparently interacting
licity calibrations over their range of validity and indi- dwarfstarburst galaxy NGC 1569,we obtained masses
cate significant deviations from their linear behaviour for the bulk of its YSCs in the range 103−104 M⊙.
towards higher metallicties. We showed that transfor- Only a handful of these, including the 3 previously
mationsfromcolourtometallicityaresignificantlyage- knownso-calledSuperStarClusters,havemassesabove
dependentandthattransformationsfromcolourto age afew 105 M⊙, i.e. inthe rangeofGC masses. We con-
are significantly metallicity-dependent (Schulz et al. cludethatthisstronglystarburstingbutnotapparently
2002). The effect of dust absorption is included in interacting dwarf galaxy does not form any new GCs,
GALEV models assuming a starburst extinction law or, at most, very few (Anders et al. 2004).
(Calzetti et al.2000)forarangeofvaluesforE(B−V)
(0 ≤ E(B −V) ≤ 1 mag). GALEV models also in- 2.3 Star cluster formation in the merger remnant
clude the full set of Lick spectral absorption indices NGC 7252
on the basis of empirical calibrations for the indices in
terms of stellar parameters for every individual clus- For the starburst in the massive gas-rich spiral – spi-
ter star T , log g, [Fe/H] as given by (Gorgas et al. ral merger remnant NGC 7252, we could estimate the
eff
1993) and (Worthey et al. 1994). We showed that SF efficiency very conservatively to be at least 35 %.
the transformation from the age-sensitive Lick index This estimate was based on the amount of new stars
H to age is significantly metallicity-dependent and formed during the burst, as obtained from the deep
β
that the transformation from the metallicity-sensitive Balmer absorption lines in the overall spectrum, and
Lick indices (Mgb, Mg , [MgFe], ...) to metallicity is a very generous estimate of the gas mass available in
2
age-dependent for ages ≤ 10 Gyr (Kurth et al. 1999; the two Sc-type progenitor spirals, of which the ample
Lilly and Alvensleben 2006). HI still observed all along the extended tidal tails is
Our analysis methods use the full information from the proof (Fritze-v. Alvensleben and Gerhard 1994,?).
multi-bandimaging(UV, U, B,..., NIR)or/andLick SuchahighSFefficiencyshouldallowfortheformation
spectroscopy available for a SC system, compare them of massive, compact, strongly bound GCs. HST obser-
toalargegridofover100.000GALEVmodelsinterms vations indeed revealed a rich population of compact
of Spectral Energy Distributions (SEDs), Lick indices, SCswithagesintherange600−900Myrandmetallic-
or a combination of both (cf. Anders et al. (2004), ities close to solar (Fritze-v. Alvensleben and Burkert
Lilly and Alvensleben (2006), Lilly & Fritze 2008, sub- 1995). They apparently have survived many inter-
mitted). SEDs, we recall, are sets of magnitudes in a nal crossing times and the most critical phase in their
number of filters from short to long wavelengths, e.g. lives, the infant mortality phase after expulsion of
U ... K. Our analysis tools not only determine the the gas left over at their formation when the first
best fit model but attribute probabilities to all mod- SNe went off, and they are still compact and bound.
els that allow us to determine the 1σ uncertainties for This is particularly impressive since all this happened
4
during the violent relaxation phase that restructured the clear turn-over in the luminosity function could be
their parent galaxy from two spiral disks into a spher- a hint that the amount of SF that goes into massive
ical configuration featuring a de Vaucouleurs profile SCs relative to the amount of SF that goes into low-
(Schweizer 2006). These young GCs have all chances mass SCs might be higher in this gas-rich merger than
to survive another Hubble time. They have masses in in other environments. It will be very interesting to
the range 105−106 M⊙ with cluster W3 even reach- check with ALMA our expectation that the molecu-
ing7−8 M⊙ (Fritze-v. Alvensleben and Burkert1995; lar cloud structure and mass spectrum in this major
Maraston et al.2004). EnoughofthoseyoungGCssur- merger are different from what they are in the Milky
viveduntiltodaytosecurethemergerremnantthetyp- Way, i.e. closer to what is observed in terms of inte-
icalGC specific frequencyofanellipticalgalaxy,which grated light from higher and lower density molecular
istwiceashighwhendefinedintermsofnumberofGCs gas L(CS, HCN)/L(CO) in ULIRGs. NGC 4038/39
inrelationto galaxytotalmass asfor anaveragespiral is currentlya LIRG(LIR >1011 L⊙) andwillprobably
(Zepf and Ashman 1993). I.e. during the strong global furtherincreaseitsSFRclosetofinalmerging. Thereis
starburst accompanying the merger that transformed observationalevidenceforverylargeamountsofmolec-
twobrightScgalaxiesintoadynamicallystillyoungel- ular gas at the level of about twice the total gas mass
liptical, a number of secondaryGCs has formed that is (HI + H ) in the Milky Way (Gao et al. 2001) and for
2
comparable to the number of preexisting GCs in both extremely massive concentrations of it (Wilson et al.
progenitor spirals. 2003)withlowkinetictemperature(Schulz et al.2007).
Shocked gas, on the other hand, is found displaced
2.4 Star cluster formation in the ongoing merger NGC from the regions of high present SF, i.e. most proba-
4038/39 bly due to the collision of the two galaxies (Haas et al.
2005). The exceptionally high magnetic field strength
The ongoing gas-rich spiral – spiral merger NGC
that Hummel and van der Hulst (1986) measured over
4038/39,the Antennae, forms a richYSC system. It is
an extended region also suggests compression of the
notpossibleto tellapartthe YSCs intoshort-livedand
ISM. The fraction of very dense molecular gas as seen
long-lived ones. In any kind of observationally acces-
in L is still low,as wellas the SF efficiency, where-
HCN
sible parameter (mass, half-light radius) or parameter
from Gao et al. (2001) conclude that the bulk of the
combination, these YSCs form a continuous distribu-
starburstis yet to come as the two nuclei merge, prob-
tion. In a very careful analysis of this SC system in
ably driving the Antennae abovethe ULIRG threshold
formation,including conservativeSC identification,ac-
in terms of IR luminosity.
curate aperture corrections for SC sizes and photome-
We speculate that if the turnover in the luminosity
try, careful completeness analysis, extensive statistical
functionwouldreflectaturnoverintheunderlyingmass
tests and likelihood evaluations by Monte Carlo simu-
function,thenthiswouldtieinnicelywiththerecentre-
lations, we could show that the luminosity function of
sult obtained by Parmentier and Gilmore (2005, 2007)
the SC system features a turnover with 99.5 % signif-
that the Milky Way GC system originally must have
icance (Anders et al. 2007). In this respect, the YSC
had a mass spectrum with a turnover around 105 M⊙.
luminosity function in this ongoing merger differs from
luminosity functions of YSCs in dwarf galaxies, spi-
rals, and isolated starbursts, which all are power laws.
3 Conclusions so far ...
It also differs from the power laws found for the mass
functionsofmolecularcloudsandmolecularcloudcores
We so far conclude that starbursts in non-
in undisturbed galaxies. Mass functions of molecular
interacting dwarf galaxies do not form substan-
clouds and molecular cloud cores in gas-rich interact-
tialpopulationsofmassive,long-livedYSCsthat
ing galaxies cannot yet be measured, they will have
could evolve into GC, while massive gas-rich
to await ALMA. Unfortunately, it is not straightfor-
mergers do.
ward to transform the YSC luminosity function into
NGC 7252 is not the only example of a merger
a mass function, since it is not clear how to translate
that no doubt has produced a new generation of GCs.
the completeness limit in luminosity into a complete-
The ∼ 1 − 3 Gyr old merger remnants NGC 3921
ness limit in mass during the rapid luminosity evolu-
(Schweizer et al.1996),NGC34(Schweizer and Seitzer
tion of YSCs. The obvious way to evaluate the mass
2007), and NGC 1316 (Goudfrooij et al. 2001,?, 2004,
function in small age bins suffers from low statisti-
2007) as well feature young GC populations formed
cal significance. Repeating our accurate analysis on
during the mergers. The metallicities of these newly
the ACS data covering a larger FoV with better sam-
formed GCs agree well with expectations on the basis
pling might be a promising way to go. In any case,
StarandStarClusterFormation 5
ofspiralgalaxyISMproperties. Evolutionarysynthesis Our preliminary analysis of the 324 YSCs in those
models predictthese SCs to takeonthe opticalcolours sixofL04’sgalaxiesthathave≥30YSCseach,revealed
of the red-peak GC widely observed in E/S0 galaxies thatasignificantnumberof∼70YSCs withages≥50
by the time the tidal features indicative of the merger Myr have masses ≥105 M⊙. We have chosen a gen-
origin will have vanished. They also predict that they erous lower age limit of 50 Myr to be sure that the
should readily be detectable against other populations YSCs have already survived the most dangerous phase
of red GCs in combined optical and NIR observations in their lives, the infant mortality phase after the first
(Fritze-v. Alvensleben 2004) (see also R. Kotulla, this SNe haveexpelled the left-overgasand the subsequent
volume). dynamical rearrangement to the change in the poten-
No example of a clearly merging gas-rich dwarf tial (cf. Lamers, this volume). And we concentrate
galaxypair,norofanaccretionofagas-richdwarfbyan on YSCs with masses ≥105 M⊙ since those have fair
ellipticalorS0hasasyetbeenstudiedtocheckwhether survivalchances for the forthcoming Gyrs accordingto
those would also give rise to new GC populations. present knowledge (Boutloukos and Lamers 2003) and
they have masses in the range of GCs, hence merit to
be called young GCs.
4 Star cluster formation in normal spirals
4.2 The trouble
Whenitcomestothefirstdetailedanalysisofso-called
Super Star Clusters in normal actively star-forming This result that apparently undisturbed and not cur-
Sbc − Sd type spirals, the situation gets embarrass- rently starbursting Sbc...Sd-type spirals form SCs
ing. Larsen (2004) (L04) presents ground-based and whichhaveallthepropertiesofyoungGCsissurprising
HST multi-band photometric data for a sample of 17 and presents a challenge to our current understanding
non-interactingactively star forming face-onspirals,in of SC and GC formation and evolution. Larsen (2004)
which he identifies between 7 and 149 YSCs that he
argues that these spirals are not currently in any par-
calls Super Star Clusters (SSCs), and that we are cur-
ticularly active state of SF, they look undisturbed and
rently analyzing in the way described above. All of
featurenicediskswithregularspiralstructure. Theage
them arecompactwithradiiinthe rangeof3to 10pc.
distributionsthatweobtainfortheirYSCssupportthis
First of all, I’d like to caution the notion SSC, since
argument.
it only refers to luminosity and not to mass. The term
But how can these undisturbed spirals afford the
SSC goes back to van den Bergh(van den Bergh1971)
high SF efficiencies that current theories for GC for-
who referred to SCs much brighter than the brightest
mation require? And where are the successors of pre-
open clusters known in the Milky Way by that time.
vious generations of this type of SCs? Do normal ac-
Meanwhile, however, we know that other galaxies (e.g.
tively star-forming spirals feature continuous age dis-
the LMC) can have much richer SCs than our Milky
tributions among their GC systems? Is our Milky Way
Way and we have detailed evolutionarysynthesis mod-
(and M31) special in this respect? Can intermediate-
els that show how stronglySCs fade, in particular dur-
ageGCshaveescapedourdetectionsofar? Orarethese
ing their youngest stages. Depending on metallicity, a
actively star-forming galaxies at the same time partic-
SC fades by ∼4−5 mag inV during the firstfew hun-
ularly hostile to their YSCs and destroy them beyond
dred Myr – alone through stellar evolution effects, i.e.
our current estimates?
with the stellar-dynamical mass loss not yet included.
We discuss this in more depth in Fritze et al., in
A very luminous SC therefore need not necessarily be
prep.
extremely massive. Even at relatively modest masses,
In any case, our results require a careful reconsider-
SCs can be very bright and look like SSCs as long as
ation of currently accepted concepts of SC formation,
they are very young. We therefore strongly suggest to
evolution, and destruction.
refer to masses rather than to luminosities for YSCs.
Acknowledgements Itisapleasuretothanktheor-
4.1 YSC masses
ganisers for a very well-organizedand inspiring confer-
enceinaparticularlysplendidlocationandtothankP.
The two brightest of L04’s YSCs have spectroscopic
Anders,R.deGrijs,S.Larsen,andR.Kotullafortheir
masses of 4·105 and 2·106 M⊙ (Larsen et al. 2004).
contributions to the work presented here.
Being so few, these two could, however, be singular
outstanding SCs that formed by some local extreme
compression of molecular gas, e.g. from a supergiant
molecular cloud compressed between expanding shells.
6
References Krueger,H.,Fritze-v.Alvensleben,U.,Loose,H.H.: Astron.
Astrophys.303, 41 (1995)
Anders, P., Bissantz, N., Boysen, L., de Grijs, R., Fritze- Kurth, O.M., Fritze-v. Alvensleben, U., Fricke, K.J.: As-
v. Alvensleben, U.: Mon. Not. R. Astron. Soc. 377, 91 tron. Astrophys.Suppl.Ser. 138, 19 (1999)
(2007) Larsen, S.S.: Astron. Astrophys.416, 537 (2004)
Anders, P., Bissantz, N., Fritze-v. Alvensleben, U., de Larsen, S.S., Brodie, J.P., Hunter, D.A.: Astron. J. 128,
Grijs, R.: Mon. Not. R.Astron. Soc. 347, 196 (2004) 2295 (2004)
Anders, P., de Grijs, R., Fritze-v. Alvensleben, U., Bis- Li, Y., Mac Low, M.M., Klessen, R.S.: Astrophys. J. Lett.
santz, N.: Mon. Not. R.Astron. Soc. 347, 17 (2004) 614, L29 (2004)
Anders, P., Fritze-v. Alvensleben, U.: Astron. Astrophys. Lilly,T.,Alvensleben,U.F.V.: Astron.Astrophys.457,467
401, 1063 (2003) (2006)
Boutloukos, S.G., Lamers, H.J.G.L.M.: Mon. Not. R. As- Maraston, C., Bastian, N., Saglia, R.P., Kissler-Patig, M.,
tron. Soc. 338, 717 (2003) Schweizer, F., Goudfrooij, P.: Astron. Astrophys. 416,
Brown,J.H.,Burkert,A.,Truran,J.W.: Astrophys.J.440, 467 (2004)
666 (1995) Meurer, G.R., Heckman, T.M., Leitherer, C., Kinney, A.,
Burkert, A., Brown, J.H., Truran, J.W.: In: Burkert, A., Robert, C., Garnett, D.R.: Astron. J. 110, 2665 (1995)
Hartmann,D.H.,Majewski,S.A.(eds.)TheHistoryofthe Miller, B.W., Whitmore, B.C., Schweizer, F., Fall, S.M.:
MilkyWayandItsSatelliteSystem.AstronomicalSociety Astron. J. 114, 2381 (1997)
of the Pacific Conference Series, vol. 112, 121 (1996)
Mora,M.D.,Larsen,S.S.,Kissler-Patig,M.: Astron.Astro-
Calzetti,D.,Armus,L.,Bohlin,R.C.,Kinney,A.L.,Koorn- phys.464, 495 (2007)
neef, J., Storchi-Bergmann, T.: Astrophys. J. 533, 682
Parmentier, G., Gilmore, G.: Mon. Not. R. Astron. Soc.
(2000) 363, 326 (2005)
de Grijs, R., Fritze-v. Alvensleben, U., Anders, P., Gal-
Parmentier, G., Gilmore, G.: Mon. Not. R. Astron. Soc.
lagher, J.S., Bastian, N., Taylor, V.A., Windhorst, R.A.: 377, 352 (2007)
Mon. Not. R. Astron.Soc. 342, 259 (2003)
Pasquali, A., de Grijs, R., Gallagher, J.S.: Mon. Not. R.
de Grijs, R., Lee, J.T., Mora Herrera, M.C., Fritze-v. Al- Astron. Soc. 345, 161 (2003)
vensleben,U.,Anders,P.: NewAstronomy8,155(2003)
Schulz, A., Henkel, C., Muders, D., Mao, R.Q., R¨ollig, M.,
Elmegreen, B.G., Efremov, Y.N.: Astrophys. J. 480, 235
Mauersberger, R.: Astron. Astrophys.466, 467 (2007)
(1997)
Fritze-v. Alvensleben, U.: Astron. Astrophys. 414, 515 Schulz,J.,Fritze-v.Alvensleben,U.,M¨oller,C.S.,Fricke,K.J.:
Astron. Astrophys.392, 1 (2002)
(2004)
Schweizer,F.: ArXivAstrophysicse-prints(2006, inpress)
Fritze-v. Alvensleben, U., Burkert, A.: Astron. Astrophys.
300, 58 (1995) Schweizer, F., Miller, B.W., Whitmore, B.C., Fall, S.M.:
Astron. J. 112, 1839 (1996)
Fritze-v. Alvensleben, U., Gerhard, O.E.: Astron. Astro-
phys.285, 751 (1994) Schweizer, F., Seitzer, P.: Astron. J. 116, 2206 (1998)
Schweizer, F., Seitzer, P.: Astron. J. 133, 2132 (2007)
Fritze-v. Alvensleben, U., Gerhard, O.E.: Astron. Astro-
phys.285, 775 (1994) Shioya,Y.,Taniguchi, Y.,Trentham,N.: Mon.Not.R.As-
Gallagher, S.C., Charlton, J.C., Hunsberger, S.D., Zarit- tron. Soc. 321, 11 (2001)
sky,D., Whitmore, B.C.: Astron. J. 122, 163 (2001) Solomon,P.M., Downes, D.,Radford,S.J.E.: Astrophys.J.
Gao,Y.,Lo,K.Y.,Lee,S.W.,Lee,T.H.: Astrophys.J.548, Lett. 387, L55 (1992)
172 (2001) Trancho, G., Bastian, N., Schweizer, F., Miller, B.W.: As-
Gao, Y., Solomon, P.M.: Astrophys.J. 606, 271 (2004) trophys. J. 658, 993 (2007)
Gorgas, J., Faber, S.M., Burstein, D., Gonzalez, J.J., vanden Bergh, S.: Astron.Astrophys.12, 474 (1971)
Courteau, S., Prosser, C.: Astrophys. J. Suppl. Ser. 86, Whitmore, B.C., Gilmore, D., Leitherer, C., Fall, S.M.,
153 (1993) Chandar, R., Blair, W.P., Schweizer, F., Zhang, Q.,
Goudfrooij, P., Alonso, M.V., Maraston, C., Minniti, D.: Miller, B.W.: Astron. J. 130, 2104 (2005)
Mon. Not. R. Astron.Soc. 328, 237 (2001) Whitmore, B.C., Schweizer, F., Leitherer, C., Borne, K.,
Goudfrooij,P.,Gilmore,D.,Whitmore,B.C.,Schweizer,F.: Robert, C.: Astron. J. 106, 1354 (1993)
Astrophys.J. Lett.613, L121 (2004) Whitmore,B.C.,Sparks,W.B.,Lucas,R.A.,Macchetto,F.D.,
Goudfrooij, P., Mack, J., Kissler-Patig, M., Meylan, G., Biretta, J.A.: Astrophys.J. Lett. 454, L73 (1995)
Minniti, D.: Mon. Not. R. Astron.Soc. 322, 643 (2001) Wilson, C.D., Harris, W.E., Longden, R., Scoville, N.Z.:
Goudfrooij,P.,Schweizer,F.,Gilmore,D.,Whitmore,B.C.: Astrophys.J. 641, 763 (2006)
Astron. J. 133, 2737 (2007) Wilson,C.D.,Scoville,N.,Madden,S.C.,Charmandaris,V.:
Haas, M., Chini, R., Klaas, U.: Astron. Astrophys. 433, Astrophys.J. 599, 1049 (2003)
L17 (2005) Worthey,G.,Faber,S.M.,Gonzalez,J.J.,Burstein,D.: As-
Hummel, E., van der Hulst, J.M.: Astron. Astrophys. 155, trophys. J. Suppl.Ser. 94, 687 (1994)
151 (1986) Zepf,S.E.,Ashman,K.M.: Mon.Not.R.Astron.Soc. 264,
Jog, C.J., Das, M.: Astrophys.J. 400, 476 (1992)
611 (1993)
Jog, C.J., Das, M.: Astrophys.J. 473, 797 (1996)
Knierman, K.A., Gallagher, S.C., Charlton, J.C., Huns-
berger, S.D., Whitmore, B., Kundu, A., Hibbard, J.E.,
Zaritsky,D.: Astron. J. 126, 1227 (2003) ThismanuscriptwaspreparedwiththeAASLATEXmacrosv5.2.