Вы находитесь на странице: 1из 32


Joint Astronomy Conference

Supplement to
The Messenger 134
December 2008

Garching bei

Chemical Evolution 21– 25 July

München, Germany

Dwarf Galaxies
and Stellar Clusters
Review Articles

Editors: Francesca Primas, Jeremy Walsh, Achim Weiss

Conference Supplement

Special Report on the MPA/ESO/MPE/USM 2008 Joint Astronomy Conference

Chemical Evolution of Dwarf Galaxies and Stellar Clusters

held in Garching bei München, Germany, 21–25 July 2008

Francesca Primas 1 interesting and possibly helpful in under- small dark matter halos, which address
Achim Weiss 2 standing the origin of the abundances in the ‘missing satellites problem’ and help
both classes. to explain the detailed star formation
­histories of dwarf galaxies in the Local
ESO Indeed, the meeting turned out to be very Group.
Max-Planck-Institut für Astrophysik, lively and stimulating, with many interest-
Garching, Germany ing new results presented. So many that All the presentations, including all the
several review speakers mentioned that posters, are publicly available and linked
their presentations, especially in terms from the conference website http://www.
It is our pleasure to celebrate the success of number of known dwarf galaxies, were mpa-garching.mpg.de/mpa/conferences/
of our 2008 summer conference with the up-to-date ‘only’ until the week before garcon08/. This has also been a major
publication of this special supplement the meeting, clearly demonstrating the accomplishment, and we would like to
issue, in which we have collected the arti- incredible pace at which new ultra-faint thank all ­participants and presenters pub-
cles of at least the majority of the review galaxies are being discovered from the licly for having been so responsive to ­­
talks presented at the meeting. Sloan Digital Sky Survey. our calls for papers and presentations.
Special thanks go to the review speakers
The 2008 MPA/ESO/MPE/USM Joint All the major areas were covered by who made it into this supplement issue.
Astronomy Conference focused on the at least one review talk, followed by many We know that the deadline was very tight,
chemical signatures and evolution of invited and contributed presentations. but our (now achieved) goal was to pub-
dwarf galaxies and stellar clusters. It took This issue collects most of the review lish this supplement as close as possi-
place in Garching at the end of July articles: from Mario Mateo’s opening talk, ble to the time of the ­conference. Finally,
(21–25 July), and it was attended by 148 to Raffaele Gratton’s and Kim Venn’s our warm thanks go to all the students
participants. There were no Proceedings, presentations on the chemical signa- and technical/admin­istrative supporters
as the Local Organising Committee had tures of globular clusters and dwarf gal- who have contributed to the success of
decided to collect articles from the review axies respectively; to Santi Cassisi’s this meeting.
speakers instead and have them pub- and Francesca D’Antona’s reviews on
lished ‘somehow’ in the ESO Messenger. how these abundances can be inter- Enjoy the reading!
This ‘somehow’ has now become a spe- preted in terms of stellar evolution models
cial supplement to the December 2008 and how they could be connected; and Francesca and Achim
issue of the ESO Messenger. last, but not least, the concluding remarks
by Ken Freeman.
The choice of the scientific topic of this
year’s Joint Astronomy Conference in Further, at the meeting, Eva Grebel inves-
Garching was driven by the current inten- tigated the links among ages, kinematics,
sive work in determining stellar abun- metallicities and other properties of dwarf
dances in galactic stellar systems (nota- galaxies and presented the dynamical
bly globular clusters) and Local Group and chemical evolution of an isolated
dwarf galaxies. Many of these projects ­system with the properties of a self-gravi-
are being actively pursued with the latest tating three-component dwarf galaxy
instruments, and have revealed surpris- consisting of gas, stars and dark matter.
ing results. Abundances and kinematics Also, we heard a lot about the formation
are now routinely measured for hundreds of stellar systems, in the nice reviews
of stars per galaxy/cluster, thanks to given by Pavel Kroupa and Oleg Gnedin,
the latest generation of multiplex facilities. the latter broadcast via video-connection
Our mapping of the Local Group is basi- from Gnedin’s home institution. Kroupa
cally changing on a month-to-month talked about the early evolution of dense
basis (sometimes even more often), and stellar systems, its dependence on mass,
the recent discovery of several ultra-faint and discussed some hitherto poorly
dwarf galaxies clearly offers new horizons understood scaling relations in the transi-
to explore. tion region between star clusters and
dwarf galaxies. Gnedin presented an
As globular clusters and dwarf galaxies overview of the dynamical evolution of
form a mass sequence, and possible globular clusters and dwarf galaxies over
connections between the two classes of cosmological timescales, with an empha-
stellar systems have always been pro- sis on Local Group systems. He also
posed (e.g., globular clusters as the cores described current ideas on the formation
of former dwarf galaxies), a confronta- of massive star clusters in the first sev-
tion and comparison of cluster and dwarf eral gigayears after the Big Bang, as well
galaxy chemical evolution appeared to be as the latest models of star formation in

2 The Messenger 134 | Supplement – December 2008

Conference Supplement

The Complex Evolution of Simple Systems

Mario Mateo of the low luminosity end of the popula- galaxies, including the Milky Way. Dwarf
University of Michigan, Ann Arbor, USA tion of spheroid systems (Figure 1 of Kor- galaxies and globular clusters must play
mendy, 1985). In the past two and a half a central role in the hierarchical paradigm
decades, we have come to realise that for a fundamental reason. These objects
The simplicity and extreme ages of the simple appearance of dwarf spheroi- comprise the smallest and oldest sys-
globular clusters and dwarf galaxies dal (dSph) galaxies belies a rich range tems surrounding present day galaxies.
imply that these systems may be useful of population, kinematic, environmental But small and old things must, at the very
windows to the earliest era of galaxy and chemical properties that are funda- least, be contemporaneous with the
formation. Recent discoveries of local mentally at odds with the simple para- ­hierarchical ‘building blocks’ that we now
dwarfs have, in some ways, begun to digm summarised above. More recently, believe drove the formation of larger
blur the distinctions between the two some globular clusters – the very em- ­systems. Some of today’s systems may
types of systems. However, it remains bodiment of simple stellar populations – even be identical to some of these early
clear that the two types of systems have been observed to exhibit some structures, but, due to chance, have
arose from fundamentally different con- bizarre properties that reveal unexpected not yet merged into larger galaxies. In
ditions in the early Universe. Globular similarities to dwarf galaxies, blurring these respects, the local dwarfs and
clusters result from ‘intense’ (what is the distinction between these two types globular clusters are identifiable fossils of
often referred to as ‘efficient’) star for- of stellar systems. the era of active galaxy formation, an era
mation processes, possibly related drastically unlike the present. Can we
to major merging, while dwarf galaxies Even if we acknowledge that low lumi- interpret the messages that these fossils
represent regions of much more ‘se- nosity spheroidal systems and their cous- contain?
date’ (low efficiency) star formation, ins, the low luminosity dwarf irregular
possibly independent of significant con- ­galaxies, are intriguing in their own right, This paper is based on my opening talk
tributions from merging. I review spa- then it is their role in bigger questions at the very successful MPA/ESO/MPE/
tial, kinematic and chemical results that of structure formation that makes their USM conference, “Chemical Evolution of
support this interpretation. study particularly compelling. Over the Dwarf Galaxies and Stellar Clusters”,
course of my astronomical career – and, held in Garching in late July 2008 (and
really, it has not been that long! – the skillfully organised by Achim Weiss and
What could be simpler? Collect enough pendulum regarding the paradigm of gal- ­Francesca Primas, to whom I extend
gas – and possibly dark matter – in one axy formation has swung completely my thanks). I thought that I knew enough
region of space so that, even in the pres- from one extreme to the other. The mon- about both dwarf galaxies and globular
ence of Universal expansion and a hot olithic model, first expounded in detail clusters to contrast their properties effec-
cosmic background radiation field, it be- by Eggen, Lynden-Bell & Sandage et al. tively. Although there remain many funda-
comes Jeans unstable. If dense enough, (1962), has swung to models that incor- mental differences between dwarfs and
the cloud forms individual stars that ul- porate fundamentally hier­archical proc- clusters that I outline below, the confer-
timately settle into a dynamically quasi- esses, inspired by Searle & Zinn (1977), in ence did reveal some unexpected traits
stable system in which the stars are dis- which small structures form first, then that they share. This may have muddied
tributed as expected in simple dynamical merge to build up larger systems. Today, our understanding in some areas that
models (e.g., King, 1966). Really, what the hierarchical paradigm is unquestion­ some people – at least me! – felt were
could be simpler? For decades, astrono- ably the more popular, and rightly so. We converging to a fairly broad consensus.
mers were certain that this basic picture see direct evidence for mergers, most Perhaps the most telling example is the
accounted for the properties of globular spectacularly in the form of streams and fact that there are now serious discus-
clusters and, quite possibly, dwarf sphe- tidally disrupted dwarfs that are clearly sions about how we can conclusively dis-
roidal galaxies, the recognised denizens contributing to the populations of local tinguish clusters from low luminosity
dwarfs near the various parameter inter-
Figure 1. An image showing a globular faces that, until relatively recently, com-
cluster (upper right), a dSph galaxy
fortably separated the two classes of
(lower right) and a galaxy, Fornax, that
contains a (small) population of glob­ objects. When the validity of the classifi-
ular clusters (the numbered objects cation of (some) clusters and dwarfs are
in the image to the left). Even here, being called into question, you know
the considerably higher surface bright-
things are getting pretty interesting. I will
ness of the globular clusters, com-
pared to the galaxies, is evident. follow a similar outline in this paper that I
used in my talk, but, where possible,
incorporating some of the exciting new
results and ideas that arose at the con-

The Messenger 134 | Supplement – December 2008 3

Conference Supplement Mateo M., The Complex Evolution of Simple Systems

Clusters and galaxies: fundamentally strongly favour the remote halo. Only a percolating galaxies have yet died a natu-
­different few recently discovered, very low lumi- ral death by consuming all their gas
nosity systems, streams, or possibly un- in star formation (though in the light of
There is no question that some of the bound shreds exist within an effective our lack of information regarding the
­distributions of properties of dwarf galax- ‘no-fly’ zone out to about 70 kpc (Fig- internal kinematics of Tucana and Cetus,
ies and clusters overlap, for example ure 3). This distribution suggests that the the two outliers in Figure 5, we do not
luminosity, baryonic mass, even kinemat- smallest dwarfs are strongly influenced know if these may be examples of such
ics. At the conference there was exten- by tidal effects that either disrupt or trans- deceased dIrr galaxies). Overall, the spa-
sive discussion whether cluster and gal- form them drastically inside this zone tial distribution and populations of dwarfs
axy sizes overlap, given recent claims (e.g., Mayer et al., 2006). The Magellanic apparent in Figures 4–5 are consistent
(for an example in the recent literature, Clouds are an obvious exception, but with a model in which tidal effects drive
see van den Bergh, 2008) that there is a recent proper motion measurements the evolution of these systems.
“gap” between about 20 –100 pc in the (Kallivayalil et al., 2006; Piatek et al., 2008)
distributions of half-light radii in the two suggest they are passing by the Milky The really obvious distinction between
types of systems, with the clusters com- Way for the first time. If true, they are not clusters and dwarfs, however, is most
prising the more compact population. really yet part of the Galaxy and, given strikingly apparent when we consider
Recent highly successful searches for their comparatively large masses, have their surface brightnesses, a point that
new dwarfs and clusters have begun to not yet interacted strongly with the Milky has been known for quite some time
populate the gap. The diffuse clusters Way. What is also clear (Figure 4) is (Kormendy, 1985). Although sizes, lumi-
of M31 (Mackey et al., 2006) and the very that the population of dwarf galaxies itself nosities and even kinematic properties
lowest luminosity dwarfs (Martin et al., changes with Galactocentric distance. overlap among the two populations, the
2008) found in the past few years (and Between 70–250 kpc, the dwarf popula- surface brightnesses do not; this is
months!) have certainly begun to reveal a tion is dominated by spheroidal systems. graphically apparent in Figure 1, where
possible overlap in the distributions of Beyond this outer radius, dwarf irregu- the globular clusters of one dwarf, For-
half-light radii. Indeed, in some cases it is lar (dIrr) galaxies (again, the Magellanic nax, a comparatively high surface bright-
becoming a problem to know what clas- Clouds excepted) dominate the popula- ness example of its class, are readily
sification to apply to individual systems. tion of Local Group dwarfs. This basic ­evident due to their elevated surface
Some discussions of this distinction in segregation of dwarf/cluster properties brightnesses compared to the field stars
classifying specific systems have been a with Galactocentric distance has been in the galaxy. In his excellent review at
bit arbitrary, while others have aimed at commented upon for some time in the lit- the meeting, Oleg Gnedin emphasised
determining more objective criteria appli- erature (van den Bergh, 1994). this point and noted that this strongly
cable to low luminosity systems. This is suggests fundamentally different modes
not easy: some systems have integrated Figure 5 illustrates the distribution of the of star formation in dwarf galaxies and
luminosities comparable to those of ages of the youngest populations in in- globulars.
­individual red giant stars, which can lead dividual galaxies as a function of Galacto-
to large Poisson uncertainties in their centric distance. Only some of the more I would go just a bit further with this idea.
luminosities and structural properties, as critical, recently discovered galaxies, The clear segregation of dwarfs by type,
nicely illustrated by Martin et al. (2008). Leo T and And XVIII for example, are in- and of dwarf galaxies from clusters (Fig-
cluded in this figure, but there is no ures 2–4), suggest that the modes of star
Despite the overlap in some of their prop- ­qualitative change to Figure 5 if all of the formation were of relatively differing im-
erties, there is little question, in my view, ‘newer’ dwarfs are included. The dis­ portance at different stages of the forma-
that dwarf galaxies and globulars are fun- tribution of the youngest ages reveals the tion of the Milky Way. One can imagine,
damentally different sorts of stellar sys- same effect we saw more qualitatively for example, that the initial overdensity in
tems. One of the most obvious hints of in Figure 4. Unless we have been very the matter distribution that grew eventu-
this comes from their relative distributions lucky with the ensemble of dwarf galax- ally into our Galaxy consisted of a lot of
around the Milky Way. Figures 2–4 show ies in the outskirts of the Local Group, gas, undifferentiated and tidally disturbed
the spatial distributions of Galactic glob­ Figure 5 implies that such galaxies have dark halos, but perhaps a few or none of
ular clusters (data from Bill Harris’s online probably been forming stars at the low the classical independent hierarchical
compilation of GC properties) and dwarf rates we see today over much of their building blocks we might imagine to have
satellite galaxies of the Milky Way (data history. I like to call this process ‘percola- preceded the formation of the Milky Way.
from Mateo, 1998), supplemented fully tion’, and it seems to represent the mode At such times, star formation was proba-
with recent data for systems discovered of star formation one can expect in low bly driven more by strong gas interac-
through July 2008). Around the Milky Way mass, gas-rich and tidally undisturbed tions (cloud–cloud collisions, wind-driven
(Figures 2–3) it is clear that the distribu- systems. Many of the dIrr galaxies of the shocks, supernova compression) than
tions of clusters and dwarf galaxies are outer Local Group can continue to form mergers of mature, star-bearing systems.
almost mutually exclusive. Whereas the stars as they do now far into the future. This scenario would favour intense star
median Galactocentric distance of globu- The lack of any clear examples of ‘red formation and considerable supernova
lar clusters is smaller than the Sun’s, the and dead’ systems with clear disc kine- chemical signatures.
objects we consider to be dwarf galaxies matics suggests, moreover, that no such

4 The Messenger 134 | Supplement – December 2008

2#22) +DN3

4,@ ((
2DF ( "5M (
  +DN (
+DN ((
"5M"5M(( (((
+DN 5
+DN (5
 4,@4,@( (((
 2DW 6H 4,H
!NN #Q@ 'DQ 'DQ
4,@ ((!NN (((
2DF 22)
( ((
9  ,55
9  +,"
 l   8
l l
 l  l 
 l  7 
 7  l 
8 l l  
l  l 

Figure 2. The distribution of inner halo objects as a Figure 3. The distibution of objects in the outer
function of Galactic X, Y and Z coordinates. Grey Galactic halo. Individual dwarfs are shown as green
points are globular c ­ lusters, while the green points dots and are labelled by name. The projected posi-
represent the closest of the recently discovered tions in the x-y plane are shown at the bottom of the
dwarfs, as well as the Sgr dwarf. Note that nearly all plot as open circles. The asymmetric distribution
the objects in the inner halo are globulars; their apparent here is largely due to the uneven sampling
median distance from the Galactic Centre is smaller imposed by the SDSS.
than the Sun’s (as shown by the ring of the Solar
­Circle in projection along the bottom of the plot). The
open dots are the positions of the globular clusters
projected down to the x-y plane.


2DW! 2DW


9  (" -&"

MC 75((( ##.
/DF ("
l 4*2

l  7
l l  
l  l

8 l  l


Figure 4. The distribution of all galaxies in the Local Figure 5. A plot of the ages of the youngest popula-
Group. Green dots are the dSph galaxies; blue dots tions as a function of Galactocentric (or, where
the dIrr galaxies. Orange diamonds represent the appropriate, M31-centric) distance (RGC) for Local
so-called transition galaxies (see Mateo, 1998), while Group dwarfs. The symbols are the same as in
three remote dSph galaxies are shown as pink ­Figure 4. Note the clear progression from dSph, to
squares (Cetus, Tucana and And XVIII). The large red transition, to dIrr galaxies with increasing RGC.
dots are M31, the Milky Way and M33. The projected Star formation is preferentially truncated near mas-
distribution in the x-y plane of all the galaxies ex- sive parent galaxies.
cept the dSph systems is shown at the bottom of the
plot with open symbols. Note that while the cluster-
ing of the dSph galaxies is obvious in this plot, the
projected distribution reveals that the dIrr and transi-
tion systems are not at all clustered on M31 or the
Milky Way.

The Messenger 134 | Supplement – December 2008 5

Conference Supplement Mateo M., The Complex Evolution of Simple Systems

Globular clusters probably represent this galaxies that do contain globular clusters tinction between the high intensity star
form of star formation for two reasons. are among the more massive systems. formation mode characterised by globu-
First, they likely could only form this way; This may just be telling us of the hierar- lar clusters and the low intensity mode of
the Antennae galaxy (Figure 8) repre- chical processes, similar to the gas-rich dwarfs, implying that in the low intensity
sents a local example where a violent phase described above for the Milky Way, mode the formation of very massive stars
merger is actively producing massive, that were going on early in the formation may be strongly suppressed, not merely
compact clusters in the interfaces where process of these galaxies. How the mas- statistically unusual. If the initial mass
dense clouds are interacting (Whitmore, sive nuclei of some dwarf galaxies fit into function (IMF) is a purely statistical distri-
2002). Second, structures that may have this picture is unclear, but the complex bution, then a series of N star-forming
formed at lower density have long ago star formation history of w Centauri, a events that produce M solar masses of
been disrupted and become part of the fascinating topic of this meeting, may be stars each should produce as many mas-
overall field of the Galactic Bulge. a clue that these objects form in some sive stars as a single event that produces
hybrid manner in which distinct, intense M × N solar masses in stars. The abun-
Throughout the meeting we heard of this bursts of star formation can occur over dances of the a-elements in dwarfs sug-
mode of star formation in globular clus- an extended timescale. gests this is not the case and that high-
ters as being ‘efficient’, but we do not mass star formation has a minimum
know if that efficiency should be defined threshold in some key regulating parame-
as the fraction of gas in an initial star- Chemistry in dwarf galaxies ter (Star formation intensity? Overall
forming event that is converted into stars. mass of the star-forming region?). In this
I would suggest that the term ‘intense’ A fundamental topic of this meeting con- respect, a comparison between the
is a better label for this mode and it can cerned the chemical evolution of dwarfs a-abundances of the field stars and of
even be quantified by a star formation and clusters. Here again, clusters and the members of the globular clusters in
rate. We know that for some clusters with dwarfs reveal some telling differences. Fornax would be most interesting.
little or no age spread (not some of the Perhaps the best known is the well-
perplexing cases presented at the meet- established lack of enhancement (relative Another key chemical difference between
ing) that such rates would have exceeded to iron) of a-elements of the stars in globular clusters and dwarf galaxies
10 –100 MA/yr! dwarf galaxies compared to metal-poor is apparent when we consider their mean
globular clusters (e.g., Pritzl et al., 2005). abundances as a function of baryonic
By contrast, dwarf galaxies represent Canonical chemical evolution models content. It is well known that among glob-
cases where, for the most part, truly in- attribute this enhancement to Type II SNe ular clusters there is no relation between
dependent hierarchical structures were during the peak star formation epochs cluster luminosity and mean abundance.
able to form stars and even evolve chem- in a given population. The distinction be- The full range of chemical abundances
ically over a significant amount of time. tween dwarfs and clusters appears to exhibited by clusters are apparent when
A few of these – Sagittarius is a clear suggest that, while globular clusters were one considers only the most luminous
example – may subsequently dissolve either self-enriched in a-elements or clusters (e.g., 47 Tuc, (MV, [Fe/H]) = (– 9.4,
into a larger parent galaxy, but the mode formed from material already enriched in – 0.8) and NGC 2419 (– 9.6, – 2.1)) or the
in which the hierarchical process pro- these elements, dwarfs formed neither least luminous clusters (e.g., NGC 5053
ceeds is quite different from that which from such gas nor did they produce much (– 6.7, – 2.3) and Ter 3 (– 4.5, – 0.7)). This
must have occurred early on in a massive additional a-enriched gas that could suggests that their mean abundances
galaxy such as ours. Moreover, these ­pollute subsequent generations. What is were dictated by largely external proc-
independent structures produced stars in perplexing about this is that for most esses or initial conditions and not by self-
a much milder manner. If the remote dIrr globular clusters, the star formation enrichment by their own stars during the
galaxies are representative, then, left to epoch was very short, so the window for periods when they formed their stars. We
themselves such galaxies form stars at a self-enrichment is short. The fact that were reminded of some intriguing clus-
rate of more than 105 times lower than massive stars form first in most models of ters during the meeting (w Centauri, long
that which must have occurred in typical star formation helps this self-enrichment known to be unusual, but also NGC 2808,
globular clusters. Dwarfs clearly repre- to occur. However, in contrast, star for- M54 and others), but it seems that even
sent the ‘low intensity’ mode of star for- mation is demonstrably extended in na- in these cases the chemical anomalies
mation. ture in most dwarf galaxies and so if any that are present may reflect processes
SNe enriched the gas in these systems, that either occurred outside the clusters,
A curious implication of this has to do it would reveal itself to us today in the or before they were born. This topic led
with the few dwarfs that have clusters detailed a-abundances. One has to con- to animated discussion during the confer-
(e.g., Fornax, Sagittarius, NGC 6822 in clude that no Type II SNe occurred dur- ence!
the Local Group). If the two modes of star ing any of the star formation events in
formation – high and low intensity – are dSphs, or that the gas was mixed in such By contrast, dwarf galaxies reveal a
traced by globular clusters and the gen- a way as to concentrate the enhance- strong correlation of luminosity and mean
eral field star formation in dwarf galaxies, ment quite nonuniformly and in stars that chemical abundance, a key point first
respectively, then these galaxies must we have, by chance, not observed yet. noted some time ago by Skillman et
have experienced both. Generally, dwarf This represents another important dis- al. (1989). Figure 6 shows a modern com­

6 The Messenger 134 | Supplement – December 2008

    0.0 0.0

Nebular [O/H] – 8.86

l   l   – 0.5 – 0.5

CMD Average [Fe/H]; σ ≤ 0.25


l  l  –1.0 –1.0

l  l  –1.5 –1.5

Stellar [α/H]
l   l   – 2.0 – 2.0

l   l   – 2.5 – 2.5

l   l   – 3.0 – 3.0
l  l l l l – 20 –16 –12 –8 –4
,5 MV

Figure 6. The age-metallicity relation for dwarf galax- Figure 7. The age-metallicity relation for dwarf
ies of the Local Group (blue dots) and Galactic glob- ­g alaxies of the Local Group from high quality CMD
ular clusters (grey dots). The mean metallicities of and spectral analyses (s [O/Fe] ≤ 0.25 mag; blue dots),
the galaxies are from CMD analyses, with no cuts on nebular abundances (on a scale set by the solar
the quality of the measurements. Do the dwarfs tran- [O/Fe] value; red dots), and from stellar a-abun-
sition into the clusters in this diagram? dances (green dots). The realm of globular clusters is
denoted by the grey dashed lines (see Figure 6).
Here it is more apparent that the galaxy abundance
trend seems to pass through, and is independent of,
the cluster distribution.

pilation of the chemical abundances – Figure 8. Dwarf galaxies seem to have

followed comparatively sedate star
determined from photometric indices
formation histories, devoid of much
such as colour-magnitude diagrams contamination by supernovae (left).
(CMDs) – and luminosities for the many Globular clusters seem to have formed
Local Group dwarfs for which good in far more violent conditions (right).
Together, they may allow us to piece
data exist. Many details went into pro-
together many of the key elements of
ducing this plot, but for our purposes the galaxy formation.
abundances plotted can be assumed
to approximate the modes of the distribu-
tion of abundances in the individual
­galaxies. For reference, the most lumi-
nous galaxy plotted in Figure 6 (and all
subsequent plots of this type) is the
Large Magellanic Cloud (LMC); the least
luminous objects are all recent discover-
ies, some of which are considerably less
luminous than the Pleiades, a few less tion to our understanding of the basic consistent with the recent findings that
luminous than a ­single asymptotic giant properties of these galaxies. Figure 7 up- the a-elements are not generally en-
branch (AGB) star! Figure 6 reveals dates the luminosity-metallicity relation hanced in the stars of these galaxies. In
the ­classical trend of lower metallicity with of Figure 6 with these spectroscopic Figure 7 there is now no hint of a satu­
decreasing luminosity, but with what abundance measurements, both of stars ration at low abundance, a result that is
­appears to be a saturation setting in at a and H ii regions where applicable. Again, almost entirely due to the new abun-
mimimum metallicity of about – 2.2. The the details of this plot are complicated, dances of the faintest dwarfs from Kirby
latter point has been noted by Helmi et al. but the points represent careful means of et al. (2008) that we heard quite a lot
(2006) as evidence for the existence of a spectroscopic estimates from stars or about at the conference.
floor in the metallicities of dwarf galaxies. gas for each galaxy, often from multiple
sources. The resulting distribution now Figure 7 is astonishing in many respects.
We can do better, however. In recent reveals a remarkably tight, essentially lin- Some galaxies plotted here still have
years there has been a very impressive ear relation between Log L and [Fe/H] gas and are forming stars, so their abun-
effort to obtain spectral abundances over a factor of more than a million in gal- dances and luminosities are changing.
of individual stars in dwarfs; we heard axy luminosity! Note too that the [O/H] The implication is that dwarfs evolve
about many of these studies at the meet- abundances obey the same trend with no in such a way that they remain on the
ing. Collectively, this work represents shift, a distinction from previous results L-[Fe/H] relation at all times. The overall
a fantastic and extremely important addi- (Mateo, 1998; Grebel et al., 2006) and are relation also suggests that self-enrich-

The Messenger 134 | Supplement – December 2008 7

Conference Supplement Mateo M., The Complex Evolution of Simple Systems

ment is important for dwarfs as they oc- some of the iron abundances of the indi- References
cur along a line that is consistent with vidual stars measured in these galaxies
Dekel, A. & Silk, J. 1986, ApJ, 303, 39
internal, but truncated chemical enhance- and ones that exhibit a-enhancement are Eggen, O. J., Lynden-Bell, D. & Sandage, A. R. 1962,
ment (Dekel and Silk, 1986). However, 3.0 or lower. Boosting O abundances ApJ, 136, 748
we know that many of these dwarfs have by a factor of three requires adding only Grebel, E. K., Gallagher, J. S. & Harbeck, D. 2006, in
Chemical Abundances and Mixing in Stars in
complex star formation histories, includ- about three Earth masses of that ele-
the Milky Way and its Satellites, ESO Astrophysics
ing many that have no gas today. Thus, ment to a solar-mass star! Such a tiny Symposia
the classical ideas that SNe blew out gas enhancement would be hard to detect in Helmi, A. et al. 2006, ApJ, 651, L121
in these systems to halt their star forma- higher metallicity stars, but is apparent Kallivayalil, N. et al. 2006, ApJ, 638, 772
King, I. R. 1966, AJ, 71, 64
tion and chemical enhancement is sim- in very metal-poor objects. Are we seeing
Kirby, E. N. et al. 2008, ApJ, 685, L43
ply wrong, a conclusion consistent with the (faint) chemical echoes of the very Kormendy, J. 1985, ApJ, 295, 73
the a-abundances summarised above. first supernovae of the first stars that en- Mackey, A. D. et al. 2006, ApJ, 653, L105
Instead, the evolution implied by Figure 7 riched the gas from which dwarf galaxies Martin, N. F. et al. 2008, ApJ, 684, 1075
Mateo, M. L. 1998, ARAA, 36, 435
is of a classical closed box (although per- formed?
Mayer, L. et al. 2006, MNRAS, 369, 1021
haps underaffected by SN ii enrichment Piatek, S., Pryor, C. & Olszewski, E. W. 2008, AJ,
compared to, say, a region in a massive 135, 1024
disc galaxy such as the Milky Way) that Small objects, big implications Pritzl, B. J., Venn, K. A. & Irwin, M. 2005, AJ, 130,
either continues to form stars at, gener-
Searle, L. & Zinn, R. 1978, ApJ, 225, 357
ally, very low rates from self-enriched gas Spatial distribution. Chemistry. I have only Skillman, E. D., Kennicutt, R. C. & Hodge, P. W.
(the dIrr systems), or implies a system focused on two major areas in which 1989, ApJ, 347, 875
in which the gas was removed rather sud- clusters and galaxies clearly differ, point- van den Bergh, S. 2008, ArXiv e-prints, 807,
denly before it could all be cycled into ing out that these differences imply
van den Bergh, S. 1994, AJ, 107, 1328
stars (the dSph galaxies). The latter proc- ­further fundamental distinctions in how Whitmore, B. C. 2002, Extragalactic Star Clusters,
ess is likely to be external (e.g., Mayer et these objects formed. There are other 207, 367
al., 2006), since the chemical signatures important distinctions, dark matter con-
of SN winds appear to be absent. tent perhaps being the most significant.
We heard talks that addressed these
I have intentionally kept silent in this dis- other areas of contrast between clusters
cussion regarding the distribution of and galaxies. We heard that there may
metallicities in dwarf galaxies. Kirby et al. be a significant number of clusters that
(2008) comment on this, particularly evolved in galaxy environments, leading
for the lower luminosity systems plotted to unusual internal age and metallicity
in Figure 7. The key point that emerges distributions. This points to a common
is that all of these galaxies appear to pos- origin, at least in some cases, between
sess a significant range of chemical clusters and their parent galaxies, and
abundances, sometimes only a factor of seems to have produced some clusters
2–3, sometimes up to a factor of 20, with populations reminiscent of galactic
but not obviously correlated with lumi- systems rather than the unimodal popula-
nosity. My guess is that these abundance tions we are used to seeing in most other
spreads reflect the extended nature of clusters.
star and chemical evolution in these gal-
axies, and (possibly in addition to the To reiterate a point that I made at the
first effect) inhomgeneities of the chemi- start, what we do know is that all the lo-
cal properties of the gas from which cal dwarf galaxies, and most of the glob-
these systems formed (see below). We ular clusters, were around at the very
also heard at the conference about ex- ­earliest eras of star formation in the Uni-
tremely intriguing results regarding verse. In globular clusters we generally
a-abundances of stars in some of the see these ancient populations directly,
least luminous galaxies known. In a although some clusters clearly formed at
number of these cases, it appears that later times as we heard from summa-
there are a-element abundances of ries of recent Hubble Space Telescope/
about 0.5 dex above solar, similar to that Advanced Camera for Surveys (HST/ACS)
seen in halo stars, but not in more lumi- ages for globulars. Among the galaxies,
nous dwarf galaxies. If these enhance- we find that there are no examples of any
ments reveal enrichment due to very early dwarf systems that do not contain an
Type II SNe, why do we see these only ‘ancient’ population of stars. These little
in the lowest luminosity dwarfs? Note that systems remain the closest survivors
from Figure 7 the mean abundances of today that witnessed the era of star and
these galaxies are around 2.5 or lower; galaxy­formation so long ago.

8 The Messenger 134 | Supplement – December 2008

Conference Supplement

Abundances in Globular Cluster Stars:

What is the Relation with Dwarf Galaxies?

Raffaele Gratton Two good examples are provided by the formed. Interestingly, at a given luminos-
INAF – Osservatorio Astronomico dSph galaxy Sagittarius, whose tail can ity, the dSph metallicity is a lower enve-
di Padova, Italy be followed around the Milky Way, along lope to the GC metallicities. This effect
an entire great circle (Belokurov et al., has potential implications for the connec-
2006) and by the GC Pal 5 (Odenkirchen tion between GCs and dSphs, and merits
In the last few years increasing evi- et al., 2001). This phenomenon of stellar further examination.
dence has accumulated for some loss has two consequences: (i) a (per-
chemical evolution within globular clus- haps small) fraction of stars in the Galac- Geisler et al. (2007) made a fairly exten-
ters. The evidence is much clearer for tic halo should originate in these environ- sive comparison between the element-to-
the most massive clusters. Many au- ments; and (ii) the observed populations element abundance ratios observed in
thors have proposed that (at least the of GCs and dSphs represent the surviv- dSphs and field halo stars. They found
most massive) globular clusters may ing components of wider original popula- that dSph stars have very peculiar abun-
be closely related to the nuclei of dwarf tions. However the average properties dances of O, a-elements, Na (which are
galaxies. I review recent results on the of the original populations might be quite all underabundant) and s-process ele-
chemical inhomogeneities in globular different from those of the survivors. This ments (which are overabundant). These
clusters and discuss the perspectives distinction has important implications for abundances are all indicators of very
opened up by these results. the so-called missing satellites problem. slow star formation, as expected in these
low density environments. Differences
Since dSphs and GCs may contribute from typical halo stars imply that only a
Globular clusters and dwarf galaxies to the halo population, it is interesting to very minor fraction of the halo stars
compare their chemical composition may have come from the present dSphs.
The distinction between globular clus- to that of field halo stars. Early results for There are, however, a few halo field
ters (GCs) and dwarf galaxies is based dSphs were very discouraging (see the stars with compositions similar to that of
mainly on their structural properties. A discussion in Geisler et al., 2007). How- stars in the Magellanic Clouds and in the
useful recent discussion can be found ever, very recently Kirby et al. (2008) most massive dSphs (such as Sagittarius:
in Dabringhausen et al. (2008). GCs and found that the metallicity distribution of Mottini & Wallerstein, 2008; Sbordone et
Ultra-Compact Dwarf (UCD) galaxies the most metal-poor field stars agrees al., 2006; Letarte et al., 2006).
seem to define a unique sequence in the reasonably well with that of stars in ultra-
mass versus central density plane, sug- faint dSphs. For GCs, we may compare Figure 1 compares the abundances of
gesting a similar formation scenario, while the well-known bimodal distribution of GC stars with those of field stars from the
Dwarf Spheroidals (dSphs) are clearly abundances of GCs with the results work by Carretta et al. presented in the
separated. Although GCs and UCD gal- obtained by Ivezic et al. (2008) for a large last part of this review. In general, GCs
axies seem to form a continuum, they number of halo and thick disc stars have a composition similar to that of the
­differ significantly in their two-body relax- observed in the Sloan Digital Sky Survey Galactic halo, save for the O-Na anticor-
ation time. For GCs this is shorter than (SDSS). They found that the metallicity relation (see last section). The P (likely
a Hubble time, while for UCDs it is longer. distribution function of stars out of the ­primordial, see below) population in GCs
As a consequence, GCs are relaxed Galactic plane can be described by has a composition virtually identical to
objects, while UCDs are not. This has im- the sum of a mod­erately metal-poor disc that of field stars.
portant implications for their mass-to- and of a more metal-poor halo. These
light ratio. There is scarce evidence of two components may well be traced in Finally, we may compare the age–metal-
dark matter in both GCs and UCD galax- the GC metallicity distribution function. licity relation for Milky Way GCs (Rosen-
ies, in agreement with expectations The disc GCs correspond to the moder- berg et al., 1999; De Angeli et al.,
based on their size. However GCs have a ately metal-poor disc at |z| = 0.8–1.2 kpc 2005) with that for dSphs like Sagittarius
lower mass-to-light ratio, probably be- while the halo ones correspond to the ­(Mottini & Wallerstein, 2008) and Sculptor
cause of dynamical evolution (since they metal-poor halo at |z| = 5–7 kpc. The spe- (Tolstoy et al., 2003). The differences
are relaxed objects): in fact, due to en- cific frequency of GCs is however much are obvious: the metallicity rose very fast
ergy equipartition and hence mass seg- larger in the halo component than in the in the Milky Way, and reached a solar
regation, they selectively lose faint low disc one. value within 2 Gyrs; it increased much
mass stars. While these differences are more slowly in the dSphs, being still
on the whole clear, classification of An interesting property of dwarf galaxies below one tenth solar after several Gyrs.
­borderline objects, like w Centauri, is not is that their (mean) metallicity depends on
obvious. However, the presence of a con- luminosity: Kirby et al. (2008) provided a
tinuum of properties may be used to good version of this relation for the case Are GCs the nuclei of mostly dissolved
improve our understanding of the mecha- of dSphs. This relation fits with the con- dwarf galaxies?
nisms that lead to the formation of GCs. cept that dSphs make their own metals.
On the other hand, the metallicity of GCs Most GCs are extremely homogeneous
Both GCs and dwarf galaxies are known is fairly independent of luminosity (and in terms of the Fe-peak elements, with
to lose stars to the general field, as clear- mass), suggesting that they inherited the star-to-star variations no larger than 10 %
ly shown by the presence of tidal tails. metallicity of the medium in which they (Gratton et al., 2005; Carretta et al.,

The Messenger 134 | Supplement – December 2008 9

Conference Supplement Gratton R., Abundances in Globular Cluster Stars

&"Rl45$2 %+ ,$2 &"Rl45$2 %+ ,$2




l   l  



l   l  

:3H (%D<
: K%D<


l   l  




l   l  

l   l l  l l    l   l l  l l   
:%D'< :%D'<

Figure 1. Abundances as a function of the metallicity and estimated abundances for about for extratidal stars from M54 in the field
[Fe/H] for GC stars (Carretta et al.: red-filled symbols)
1200 stars. Their data allowed the sample of Sgr, although this might rather be due
compared with those of field halo stars from the
compilation by Venn et al. (2004), as grey circles. to be cleaned of foreground Milky Way to contamination by Sgr stars in the outer
Only O, Na and other a-elements are shown. For O interlopers, and M54 stars to be sepa- radial bins. Finally Bellazzini et al. con­
and Na, abundances of the primordial population in rated from the Sgr ones. The most impor- sidered the birthplace of M54, simulating
GCs are also indicated, as cyan-filled symbols.
tant result is that the Sgr galaxy has a its possible past orbit within Sgr. They
nucleus (Sgr N), even without considering concluded that M54 might have formed
2007a). On the other hand, significant and M54. The centres of Sgr N and M54 as far as several kpc from the nucleus of
occasionally large abundance variations co­incide to within 2 arcsec (0.2 pc) and Sgr, and then have sunk towards the cen-
in elements produced by SNe have been 0.8 km/s in radial velocity. Both M54 tre of the galaxy due to dynamical fric-
found. Given the large kinetic energy ([Fe/H] ~ –1.6) and Sgr N ([Fe/H] ~ – 0.6) tion. The other Sgr GCs are too small and
injected into the interstellar medium by have a spread in metallicity. Bellazzini et too far from the centre of Sgr for dynami-
SNe, deep potential wells and thus al. then considered the stellar distribution cal friction to have been important.
large masses are required to explain simi- on the sky and the radial velocity disper-
lar abundance spreads. This conclu- sion for the two populations. They found These results can be used to speculate
sion leads directly to the idea that at least that M54 and Sgr N stars have different on the origin of the nuclei of dwarf gal­
some of the GCs were nuclei of presently runs of velocity dispersion with radius: for axies. Bellazzini et al. concluded that the
dissolved dwarf galaxies. M54 the velocity dispersion decreases simultaneous presence of M54 and Sgr N
with radius in a “mass follows light” fash- suggests that the nuclei of dwarf galax-
M54, which is located at the core of the ion (Gilmore et al., 2007) like typical GCs; ies may form both from infall of GC(s)
disrupted dSph in Sagittarius (Sgr), plays Sgr N however has the same flat veloc- to the centre of the galaxy, and from in
a fundamental role in this respect. Very ity dispersion as the inner regions of Sgr, situ formation by the accumulation of gas
recently, Bellazzini et al. (2008) published likely influenced by dark matter and in at the centre of the potential well and its
an interesting study of M54 and its envi- agreement with Navarro-Frenk-White subsequent conversion into a stellar over-
ronment. They measured radial velocities models. Some evidence was also found density.

10 The Messenger 134 | Supplement – December 2008

Multiple populations in GCs est MS is more metal-rich ([Fe/H] ~ –1.2) ing, but their HB is discontinuous, which
than the redder one ([Fe/H] ~ –1.6), but again suggests multiple populations.
M54 is unique because we can, unlike this implies a higher He-content (Y ~ 0.4 However, detailed studies of 47 Tuc do
the nuclei of other galaxies, resolve rather than 0.25)! Comparison of the pop- not show obvious multiple populations.
the individual stars in its nucleus. How- ulations in the various sequences sug-
ever, multiple populations are seen in gests that the He-rich MS is connected
other GCs. The most famous example is to the extreme Blue Horizontal Branch The O-Na anticorrelation
w Centauri. While the wide red giant (BHB).
branch (RGB) and abundance spread of Discovered in the 1970s, the O-Na anti-
this GC have been known since the Multiple populations are also observed in correlation is probably the most charac-
1960s, the first extensive study of the other GCs. NGC 2808 is one of the most teristic feature of GCs. This anticorrela-
abundance distribution was conducted luminous GCs. Carretta et al. (2006) tion was extensively studied among RGB
by Suntzeff & Kraft (1996), who found found a large spread in the O-Na anticor- stars by Kraft, Sneden and co-workers
clear indications for a huge mass loss. As relation, but no spread in Fe-peak abun- (see, e.g., Kraft, 1994) in the 1990s. As
more sophisticated instrumentation dances. The horizontal branch (HB) shown by Denisenkov & Denisenkova
became available, a clear separation of is discontinuous, with a well-populated (1990) and Langer et al. (1993), this is evi-
the RGB into various sequences was Red Horizontal Branch (RHB), and an ex- dence for material processed through
found by Ferraro et al. (2004), showing tended BHB, but few RR Lyrae stars. high temperature H-burning in some of
that the distribution of stars with metallic- Piotto et al. (2007) found that there are the GC stars (but not in the field stars).
ity is not continuous, but that it shows three MSs; they can be explained by The O-Na (and equally the Mg-Al) anti-
evidence of various episodes of star for- ­different He-contents (Y = 0.25, 0.30 and correlation is present in all GCs for which
mation; notably, Pancino et al. (2002) 0.37). There is no splitting of the SGB and adequate data are available, and it is pri-
observed a metal-rich population, with of the RGB, indicating similar age and mordial, indicating pollution from other
[Fe/H] ~ – 0.6. metallicity for the three populations. The stars, as demonstrated by Gratton et al.
distribution of stars amongst these popu- (2001), who found that it also exists
The composition of rather large samples lations suggests that the RHB is con- among MS stars (see also Carretta et al.,
of RGB stars in w Centauri has been nected to the He-poor population, and 2004; and Ramirez & Cohen, 2002).
studied by Norris & Da Costa (1995) and the extended BHB to the He-rich one. While, in general, elements heavier than
Smith et al. (2000). The metal-rich pop- Al seem to have constant abundance
ulation is very rich in s-process elements, NGC 1851 is somewhat less massive. ratios, Yong et al. (2008) found evidence
requiring prolonged star formation. The The HB is discontinuous, with a well-pop- for small variations in NGC 6752, al-
age-metallicity relation from the subgiant ulated RHB, and an extended BHB, but though this result needs confirmation.
branch (SGB) and turn-off stars has been few RR Lyraes. Milone et al. (2008) found
obtained by Stanford et al. (2006), us- that there are two SGBs; the magnitude There are two main hypotheses for the
ing 4 m ground-based photometry and difference corresponds to about 1 Gyr, polluting stars. Decressin et al. (2007)
­spectroscopy, and showing a spread but can also be explained by a spread in proposed that rotating massive stars
of several Gyrs. Progress in instrumenta- the abundances of CNO elements. On (> 20 MA) lose material through a dense,
tion (HST/ACS photometry, spectroscopy the other hand, there is no splitting of the low velocity circumstellar disc. This mech-
with 8 m telescopes) has allowed multiple upper RGB or the MS (implying a similar anism is active on a short timescale
SGB sequences to be distinguished metallicity and He content for the two (~ 10 7 yrs), resembling more a ‘prolonged
­(Villanova et al., 2007) and demonstrated sequences). Both the population and the star formation’ episode rather than two
that the age-metallicity relation is not central concentration suggest that the distinct episodes of star formation, and
monotonic, with old metal-rich and young- RHB is connected to the younger SGB, may even produce values of Y = 0.4;
er metal-poor sequences. Element-to- and the extended BHB to the older SGB. however there are difficulties in avoiding
element abundance trends among SGB The chemical composition of NGC 1851 variations in [Fe/H], because of the con-
stars were found to be similar to those has been studied by Yong & Grundahl temporaneous explosion of core collapse
among RGB stars. (2007), who found no variation in Fe, an SNe, and in producing clear sequences.
extended Na-O anticorrelation and varia- Alternatively, stars with mass 5–8 MA,
However, the most exciting results con- tions of Ba and La correlated with Na; which undergo hot-bottom burning dur-
cern the splitting of the Main Sequence this latter finding suggests some contri- ing their AGB phase, are considered by
(MS) described by Bedin et al. (2004). bution by thermally pulsing asymptotic ­Ventura et al. (2001). This mechanism is
w Centauri has at least two MSs: a bluer giant branch (AGB) stars. active on a longer timescale (~ 10 8 yrs),
and a redder. The bluer one contains and it is a real case of ‘two episodes of
a quarter of the stars, which fits with the Looking at other massive clusters, wide star formation’; there is no problem with
fraction of stars that are more metal-rich; RGBs have also been found in massive [Fe/H] being constant, but apparently
the redder contains three quarters of clusters in M31 (Meylan et al., 2001; Y = 0.4 cannot be produced, and some
the stars and fits with the more metal- Fuentes-Carrera et al., 2008). In the Milky tuning of convection and mass loss is
poor fraction (Suntzeff & Kraft, 1996). Way, NGC 6388 and NGC 6441 are dif­ required to reproduce the observed abun-
Piotto et al. (2005) confirmed that the blu- ficult to study due to differential redden- dance pattern. Both mechanisms require

The Messenger 134 | Supplement – December 2008 11

Conference Supplement Gratton R., Abundances in Globular Cluster Stars

 Figure 2. Maximum temperature of stars along the

horizontal branch (HB) versus interquartile range
(IQR) of the O-Na abundance distribution.

+NF3 L@W

  GIRAFFE, this work

          UVES, this work
(01:.-@< &(1 %%$ (01:.-@< 45$2 Literature

a very large primordial population that is to retain the original unpolluted stars. In Carretta, E. et al. 2006, A&A, 450, 523
Carretta, E. et al. 2007a, A&A, 464, 927
then subsequently lost by the GC. A pos- fact, Carretta et al. also found that there
Carretta, E. et al. 2007b, ApJ, 671, L125
sible piece of evidence in favour of the are at least three populations in GCs: pri- Dabringhausen, J., Hilker, M. & Kroupa, P. 2008,
AGB hypothesis is given by the observa- mordial population (P), intermediate pop- MNRAS, 386, 864
tion of multiple turn-offs in LMC clusters ulation (I), and extreme O-poor popula- D’Antona, F. et al. 2005, ApJ, 631, 868
De Angeli, F. et al. 2005, AJ, 130, 116
(Mackey et al., 2008), with age spreads tion (E). P and I populations are present in
Decressin, T. et al. 2007, A&A, 464, 1029
~ 0.1–0.3 Gyr. These clusters are, how- all GCs, while the E population is pres- Denisenkov, P. A. & Denisenkova, S. N. 1990, SvAL,
ever, less massive than typical GCs. ent in only a few GCs. E and P popula- 16, 275
tions are correlated with the IQR, while Ferraro, F. et al. 2004, ApJ, 603, L81
Fuentes-Carrera, I. et al. 2008, A&A, 483, 769
D’Antona et al. (2005) made an important the I population is anticorrelated with the
Geisler, D. et al. 2007, PASP, 119, 939
breakthrough by recognising that the IQR. Notably, the three groups have the Gilmore, G. et al. 2007, ApJ, 663, 948
spreads in the He abundances imply dif- same [Fe/H] to within ~ 0.01 dex. Gratton, R. G. et al. 2001, A&A, 369, 87
ferent evolutionary masses, and thus the Gratton, R. G. et al. 2005, A&A, 442, 947
Harris, W. E. 1996, AJ, 112, 1487
likely location of the stars along the HB. We conclude by noting that the evidence
Ivezic, Z. et al. 2008, ApJ, 684, 287
Extensive data for many GCs are required for the chemical evolution of GCs is now Kirby, E. N. et al. 2008, arXiv:0807.1925
to confirm the relation between HBs well established, although the details of Kraft, R. 1994, PASP, 106, 553
and O-Na anticorrelation. The availability the evolutionary processes that give rise Langer, G. E., Hoffman, R. & Sneden, C. 1993, PASP,
105, 301
of FLAMES on the VLT has made such to this situation are not yet clear. Massive
Letarte, B. et al. 2006, A&A, 453, 547
a study possible (the Na-O anticorrela- GCs are very likely to have a close rela- Mackey, A. D. et al. 2008, ApJ, 681, L17
tion and HB (Naaah) survey) that was pre- tion with UCDs, and even more probably Meylan, G. et al. 2001, AJ, 122, 830
sented at this meeting by Carretta. with the nuclei of dwarf galaxies. Very Milone, A. et al. 2008, ApJ, 673, 241
Mottini, M. & Wallerstein, G. 2008, AJ, 136, 731
GIRAFFE and UVES spectra were ob- important progress has been made re-
Norris, J. & Da Costa, G. S. 1995, ApJ, 447, 680
tained for over 1200 giants in 19 GCs. cently thanks to the ACS camera on HST Odenkirchen, M. et al. 2001, ApJ, 548, L165
A homogeneous analysis was performed and the ESO VLT GIRAFFE and UVES Pancino, E. et al. 2002, ApJ, 568, L101
and the GIRAFFE spectra provide good spectrographs. We await new and excit- Piotto, G. et al. 2005, ApJ, 621, 777
Piotto, G. et al. 2007, ApJL, 661, L53
statistics for Na and O; in addition UVES ing results from further use of these pow-
Ramirez, S. V. & Cohen, J. G. 2002, AJ, 123, 3277
spectra yield abundances for several erful instruments in the near future. Recio-Blanco, A. et al. 2006, A&A, 452, 875
­elements. To define the extension of the Rosenberg, A. et al. 1999, AJ, 118, 2306
O-Na anticorrelation, Carretta et al. con- Sbordone, L. et al. 2007, A&A, 465, 815
Acknowledgements Smith, V. et al. 2000, AJ, 119, 1239
sidered the interquartile range (IQR),
Stanford, L. et al. 2006, ApJ, 647, 1075
which they found to be correlated with I wish to thank Eugenio Carretta and Angela Suntzeff, R. P. & Kraft, R. 1996, AJ, 111, 1913
the maximum effective temperature of ­B ragaglia, who are responsible of most of the new Tolstoy, E. et al. 2003, AJ, 125, 707
stars on the HB (see Figure 2), confirm- results presented in this discussion, and the meeting Venn, K. A. et al. 2004, AJ, 128, 1177
organisers for financial support. Ventura, P. et al. 2001, ApJ, 550, L65
ing an earlier finding (Carretta et al.,
Villanova, A. et al. 2007, ApJ, 663, 296
2007b). The IQR is also correlated with Yong, D. et al. 2008, ApJ, 684, 1159
cluster luminosity, which is itself corre- References Yong, D. & Grundahl, F. 2007, ApJ, 672, L29
lated with the presence of hot stars on
the HB, as noticed by Recio-Blanco et al. Bedin, L. et al. 2004, ApJ, 605, L125
Bellazzini, M. et al. 2008, AJ, 136, 1147
(2006). This finding may be explained Belokurov, V. et al. 2006, ApJ, 642, L137
by an increased ability for massive GCs Carretta, E. et al. 2004, ApJ, 610, L25

12 The Messenger 134 | Supplement – December 2008

Conference Supplement

Evidence for Sub-Populations in Globular Clusters:

Their Properties and Relationship with Cluster Properties

Santi Cassisi 1 Up until a few years ago the interpretation commonly known as ‘the second pa-
Angela Bragaglia 2 of the available colour-magnitude dia- rameter problem’, and we have not yet
Raffaele Gratton 3 grams (CMDs) of GCs in the framework of achieved a full understanding of their
Antonino Milone 4 theoretical stellar evolution fully sup- ­origin. In recent times the existence of a
Giampaolo Piotto 4 ported the view of GCs as SSP (see, for possible link between the HB morphol-
Alvio Renzini 3 instance, King et al., 1998). However ogy and the peculiar chemical patterns
there is a growing body of empirical find- has been suggested (D’Antona & Caloi,
ings that severely challenges this tradi- 2004): matter processed via high tem­
INAF – Astronomical Observatory of tional view. In fact, accurate spectro- perature H-burning should also be en-
Teramo, Italy scopic surveys of sizable samples of GC riched in He content, and it is known that
INAF – Astronomical Observatory of stars made after 1980 have revealed a change in the initial abundance of He
Bologna, Italy that GCs show a peculiar pattern in their produces remarkable changes in stellar
INAF – Astronomical Observatory of chemical abundances: while they are properties at the HB stage.
Padova, Italy generally – with very few exceptions –
Department of Astronomy, University of very homogeneous in the abundances of Thus, several independent observational
Padova, Italy Fe-peak elements, GC stars show very findings seem to suggest that, at least
clear anticorrelations between the abun- in some GCs, there is a sizable fraction of
dances of C and N, Na and O, Mg and Al stars that have formed from material that
An increasing number of both photo- (for a full review of this issue we refer must have undergone nuclear process-
metric and spectroscopic observations to Gratton et al., 2004) that are not pre- ing by a previous generation of stars. In
over the last decade have shown the dicted by canonical stellar models. It this context, it is clear that the fundamen-
existence of distinct sub-populations is worth noting that this pattern is charac- tal question is whether we can find direct,
in some Galactic globular clusters. teristic of GC stars, since, in field stars, straightforward evidence for the exist-
This evidence severely challenges the the observed trend for C and N abun- ence of multi-populations in some GCs?
paradigm of globular clusters as the dances is consistent with the theoretical
­prototypes of single, simple stellar pop- evolutionary predictions. This occur- Over the last few years, the availability
ulations. In this review, we briefly sum- rence strongly supports the idea that the of high quality photometry on deep HST
marise the main empirical findings col- GC environment must play a role in the images and multi-object spectroscopy
lected so far and discuss the proper­ties appearance of these chemical peculiari- has provided this evidence. So far, the
of these sub-populations and their pos- ties. More importantly, accurate spec­ presence of different sub-populations
sible relationship with global cluster troscopic measurements of both dwarf has only been proved in a few clusters,
properties. and giant stars in GCs show that the but the search is still ongoing and there
observed chemical pattern is primordial, are other GCs that are thought to host
as it is also present in unevolved stars multiple populations. In the following, we
The scientific framework (Gratton et al., 2001), and, does not only briefly summarise the evidence collect-
involve the envelopes of the stars, i.e., it ed so far for multi-populations, and dis-
Globular clusters (GCs) still occupy a piv- is not a simple pollution effect (Cohen et cuss the possible link with the global
otal role in current astrophysical research. al., 2002). properties of the parent GC.
As the oldest population II objects for
which accurate ages can be inferred, Although there is no doubt that the ob-
they place an important constraint on the served anticorrelations are due to the The chief suspects: direct, observational
age of the Universe and, in turn, on cos- fact that a fraction of the GC stars have evidence for multi-populations in GCs
mology. In addition, the analysis of the formed from matter that has been pol-
chronology of the Galactic GC system luted with the yields of high temperature The GCs in which indisputable, direct
can provide fundamental information on H-burning (see Salaris et al., 2002, for ­evidence of the presence of multiple pop-
the formation process of the Galaxy. a review), there is still some debate about ulations has been found, are: w Centauri,
For a long time GCs have represented an what is responsible for this pollution: NGC 2808, NGC 1851, NGC 6121, NGC
ideal laboratory for testing and calibrating asymptotic giant branch (AGB) stars 6388 and M54. The last is considered
stellar evolutionary models. In addition, (Ventura et al., 2001) or fast-rotating mas- to be the compact nucleus of the dwarf
since they are regarded as consisting of sive stars (Decressin et al., 2007). galaxy Sagittarius, currently being ac-
coeval and chemically homogeneous creted by the Milky Way. However it is still
stars, they have been considered as the Many GCs also show a very peculiar Hor- under debate which of the sequences
prototype of Simple Stellar Populations izontal Branch (HB), with the presence observed in the CMD of M54 represents
(SSP). In consequence, GCs have been of an extended blue tail (Recio-Blanco et the true cluster population and which
employed as an ideal template for check- al., 2006) and/or a clumpy distribution are due to stars belonging to Sagittarius
ing the accuracy and reliability of the of stars characterised by the presence of dwarf (Siegel et al., 2007). For this rea-
population synthesis tools that are used one or more gaps (Ferraro et al., 1998; son, we will not discuss this GC further.
to retrieve the properties of unresolved Piotto et al., 1999). These peculiarities in
stellar populations. the horizontal branch morphology are

The Messenger 134 | Supplement – December 2008 13

Conference Supplement Cassisi S. et al., Evidence for Sub-Populations in Globular Clusters

w Centauri Figure 1. The colour-

magnitude diagram
of w Centauri (from
The observational evidence collected ­V illanova et al., 2007).
over the last 40 years indicates that this
GC is the most peculiar object among 

Galactic GCs in terms of structure, kine-

matics and stellar content. It is the only
known GC showing a clear metallicity
spread and it is the most massive one. In
the last decade, both extensive spectro-
scopic and photometric surveys on large
L %6

samples of giant stars have shown that

the distribution of stars in metallicity, 
as well as in colour, along the red giant
branch (RGB) is clearly multi modal
­(Pancino et al., 2000; Sollima et al., 2005;
and references therein), as shown in
­Figure 1. Specifically, Pancino et al. (2000)
have shown the presence of a peculiar
RGB (the so-called RGB-a), associated
with a metal-rich population ([Fe/H] ~ 
– 0.6) that corresponds to about 5 % of
the whole cluster stellar population.
­Sollima et al. (2005) have identified three L %6 lL %6
metal-intermediate components (–1.3 <
[Fe/H] < –1.0) in addition to the dominant
population ([Fe/H] ~ –1.6). and spectroscopic properties of the dou- ous investigations have obtained quite
ble MS is that the blue MS is populated different, if not contradictory, results con-
Accurate observational analysis has by stars with a high helium content of cerning the age-metallicity relation: in
shown that some significant differences Y ~ 0.38 (Norris, 2004; Piotto et al., 2005; some cases no age difference is obtained
do exist among the metal-rich, the Lee et al., 2005). for the different sub-populations; in other
­metal-intermediate and the metal-poor cases the metal-rich component is found
components, as far as both the spatial One of the main problems with this sce- to be younger by about 2–4 Gyrs with
dis­tribution (Pancino et al., 2003) and nario is identifying the mechanism re- respect to the more metal-poor compo-
the kinematical properties (Ferraro et al., sponsible for this huge production of nent; finally independent studies have
2002) are concerned, although Pancino helium. So far, various helium producers found the existence of a significant age
et al. (2007) found no evidence of a dif­ have been proposed, such as AGB stars spread among the metal-poor and
ference in the rotational pattern among (e.g., Ventura & D’Antona, 2008, and metal-intermediate components, with the
the various sub-populations. Sollima et ­references therein), massive rotating stars metal-rich one being the older sub-pop­
al. (2005) obtained the same result for the (Decressin et al., 2007) or even popu­ ulation. It is worth noting that these differ-
radial velocity distribution, but they also lation III stars (Choi & Yi, 2007). However, ent results could partially be accounted
discovered that the metal-rich compo- the properties of He-enriched popula- for by the fact that different regions of the
nent shows a larger velocity dispersion, tions appear to leave only the AGB sce- clusters are sampled in the various analy-
thus appearing kinematically warmer than nario as viable, even if it still faces some ses, and that the presence of a popula-
the metal-intermediate sub-population. quantitative difficulties (Renzini, 2008). tion gradient in w Centauri is well-estab-
lished. In ad­dition, part of the difference
The most surprising recent result was the Even just a few years ago, only rough in the age results can be attributed to
discovery by Bedin et al. (2004; but see estimates of the relative ages of the the different theoretical frameworks, dis-
also Anderson, 1997) that, over a range of sub-populations hosted by w Centauri tance and reddening estimates used in
at least two magnitudes, the Main Se- had been obtained, using broad- and the various analyses.
quence (MS) splits into a red sequence narrowband photometry (see for in-
and a blue sequence. Follow-up spec­ stance, Hughes et al., 2004, and refer- It is evident that a detailed study of the
troscopic analysis from GIRAFFE on the ences therein). Recently, however, more chemical abundances and a more
VLT (Piotto et al., 2005) leads to even detailed analyses have been performed ­complete photometric sampling of the
more intriguing results: at odds with any by taking advantage of the most up- different sub-populations identified
expectations from canonical stellar mod- to-date photometric and spectroscopic in w Centauri is badly needed in order to
els, the bluer sequence is more metal- observational facilities (Sollima et al., understand the complex star formation
rich than the red one. Until now, the only 2005; Stanford et al., 2006; Villanova et history of this cluster better.
plausible explanation of the photometric al., 2007; Calamida et al., 2008). The vari-

14 The Messenger 134 | Supplement – December 2008

NGC 2808 Figure 2. The triple main
sequence of NGC 2808.
:,'<l  The inset shows the
This GC is one of the most peculiar Ga- 8 
­c omparison with suita-
lactic clusters in many respects: for a 8 
ble sets of isochrones
long time the morphology of its HB has  8  computed for different
 assumptions on the
been known to be highly bimodal, with

L %6
­initial helium content
the presence of one or more gaps in the (from Piotto et al., 2007).

stellar distribution (Bedin et al., 2000);
it presents – together with M13 – the

strongest O-Na anticorrelation among the 
Galactic GC population. In addition, $!l5 
L %6

­Carretta et al. (2006) have shown that it is     

possible to identify three different sub- L %6 lL %6
groups of RGB stars on the basis of their 
O abundances: O-normal, O-poor and
super O-poor components. On the basis
of the possible existence of a correlation
between O and He content, one can 
hypothesise the existence in this GC of
three distinct sub-populations of stars,
each characterised by a different He con-
tent. A direct comparison between the
star counts for the different stellar groups     
along the HB, and the sub-groups iden­ L %6 lL %6
tified by Carretta et al. (2006), shows the
presence of a rather straightforward cor- branch (SGB) region (see Figure 3) in its ­ haracterised by a strong CNONa anti-
respondence between the stars along the CMD. If the brightness difference be- correlation pattern, could properly
RGB and their progeny on the HB: the tween the two SGBs were due only to account for the observed SGB splitting,
red HB stars would be those that formed an age difference, the two star forma- without invoking any significant age
with O-rich/Na-poor/He-poor composi- tion episodes should have occurred with ­difference. Interestingly enough, this
tion, while the other two groups of RGB a time delay of about 1 Gyr. However, working scenario seems to be supported
stars would contribute to forming the Cassisi et al. (2008) have shown that the by spectroscopic measurements (Hesser
stars that populate the hottest portion of presence of two stellar populations in et al., 1982) indicating the presence of
the HB. this GC, one with a normal a-enhanced two groups of stars (CN-strong and
heavy element distribution, and one CN-weak) and by the recent works by
However, the most amazing result was
the discovery by Piotto et al. (2007) that
the MS of NGC 2808 is split into three  
loci. The unique scenario found for inter-
preting this occurrence is to assume that
the three MSs correspond to stars with
three different He contents (see Figure 2).
The self-consistency of this interpreta-
tion with the empirical evidence collected
for both RGB and HB stars is really in- 
L %6

Figure 3. The colour-

magnitude diagram of
This GC, which clearly hosts multiple stel- NGC 1851 zoomed
lar populations, is the most massive one around the sub-giant
after w Centauri. branch. The solid lines
represent the iso-
  chrones for a stellar
population with extreme
NGC 1851 CNONa anticorrelations
and ages of 9 and
10 Gyr. The dashed lines
Accurate HST/ACS photometric data correspond to iso-
have provided indisputable evidence that chrones for a population
this cluster hosts at least two distinct with a normal a-en-
sub-populations (Milone et al., 2008): hanced chemical com-
      position and ages of 10
there is a clear splitting in the sub-giant L %6 lL %6 and 11 Gyr.

The Messenger 134 | Supplement – December 2008 15

Conference Supplement Cassisi S. et al., Evidence for Sub-Populations in Globular Clusters

Yong ­­& ­Grundahl (2008) and Calamida et tents, with some spread. How this sce- smaller than that of any other GC hosting
al. (2007). nario might be associated with the recent multi-populations.
discovery of two SGB sub-populations is
NGC 1851 is considered to be the proto- an issue that has not yet been addressed.
type of bimodal HB GCs. It is intriguing A link with cluster properties?
to note that the star counts along the two
different SGBs are in remarkably good NGC 6121 It has already been noted that the five
agreement with those of stars belonging GCs that so far show evidence of the
to the two main groups of HB stars. Accurate spectroscopic data collected presence of multi-populations, are among
Therefore, one is tempted to look for a with FLAMES and UVES at the VLT have the ten most massive clusters in the Gal-
link between the different SGB sub-pop- recently provided the evidence (Marino axy. One may expect that the most mas-
ulations among the two groups of RGB et al., 2008) that this cluster shows an sive clusters are more successful in re-
stars with distinct CN abundances and extended Na-O anticorrelation, and that taining the nuclear-processed ejecta of a
the HB stellar sub-populations. In any two distinct groups of stars with sig­ first generation of stars, from which a
case, it is evident from both the observed nificantly different Na and O content are second (or further) stellar generation sub-
MS width (Milone et al., 2008) and from present. In addition, a tight correla- sequently formed. However, one has
HB synthetic models (Salaris et al., 2008) tion between the NaO and the CN abun- also to note that NGC 1851 is not such a
that the He enhancement, if any, be- dances seems to exist. The coupling massive cluster, and the situation be-
tween the two sub-populations has to be of the spectroscopic data with accurate comes even more puzzling when consid-
very small (less than 0.03). photometric evidence has also shown ering also the case of NGC 6121, which is
that the two sub-populations with differ- a GC with a small mass. The only way
Direct spectroscopic measurements of ent Na abundances occupy distinct po- to reconcile this discrepancy relies on the
the SGB and HB stars, as well as studies sitions (have different colours) along the plausible assumption that the actual
of the mass-loss efficiency among the RGB when the U-band (likely influenced mass of many GCs is just a – sometimes
RGB stars in NGC 1851 (see Salaris et al., by NH- and CN-bands) is included (see minor – fraction of the initial total mass,
2008, for a discussion on this issue), are Figure 4). However, due to the depend- as a consequence of tidal shocks with
mandatory. ence of the result on the adopted photo- the Galactic disc.
metric band, we caution that this empiri-
cal finding may not be a genuine proof It is also worth noting that almost all clus-
NGC 6388 of the presence in this cluster of distinct ters hosting multi-populations are also
sub-populations. We note that the mass characterised by a high central velocity
This GC and its twin cluster NGC 6441 of this cluster is an order of magnitude dispersion. So one could be tempted
are two extremely peculiar clusters. De-
spite their high metallicity – larger than
that of 47 Tuc – they show a bimodal HB,
extending towards very hot effective

­temperatures, a tilt in brightness (Rich et
al., 1997; Busso et al., 2007), and a
Na-O anticorrelation is present in NGC
6388 (and to a smaller extent also in
NGC 6441).

Although both clusters are affected by

severe differential reddening, Piotto (2008) 
has been able to highlight the presence
of two distinct SGB loci. No clues about 

a possible MS splitting have been col-

lected owing to the limitations imposed
by the reddening. On account of the 

close similarities between the two twin


clusters, it is plausible that the same  

­outcome could also apply to NGC 6441.
Figure 4. The colour-
magnitude diagram of
It is worth noting that the peculiar HB mor-  NGC 6121 and the dis­
phology of NGC 6388 (and NGC 6441) l       tribution of Na abun-
has been interpreted (Caloi & D’Antona, :-@%D< dance from Marino et al.
(2008). The different col-
2007; Busso et al., 2007) as due to the ours identify the stars
presence of multiple HB sub-populations      
belonging to the two dif-
characterised by distinct initial He con- 4m! ferent Na groups.

16 The Messenger 134 | Supplement – December 2008

to look for a link between the presence of Gratton, R., Sneden, C. & Carretta, E. 2004,
ARAA, 42, 385
multi-populations, a high total (initial?)
Hesser, J. E. et al. 1982, AJ, 87, 1470
mass and high central velocity disper- Hughes, J. et al. 2004, AJ, 127, 980
sion. An interesting working hypothesis King, I. R. et al. 1998, ApJ, 492, L37
would be to associate multi-popula- Lee, Y.-W. et al. 2005, ApJ, 621, L57
Marino, A. F. et al. 2008, A&A, in press
tions with the presence of an intermedi-
ate mass black hole (IMBH). Milone, A. P. et al. 2008, ApJ, 673, 241
Miocchi, P. 2007, MNRAS, 381, 103
An observable fingerprint of the presence Norris, J. E. 2004, ApJ, 612, L25
Noyola, E., Gebhardt, K. & Bergmann, M. 2008,
of an IMBH would be a small slope in
ApJ, 676, 1008
the radial density distribution in the core Pancino, E. et al. 2000, ApJ, 534, L83
region that would affect the surface Pancino, E. et al. 2003, MNRAS, 345, 683
brightness profile (Baumgardt et al., Pancino, E. et al. 2007, ApJ, 661, L155
Piotto, G. et al. 1999, AJ, 118, 1727
2005; Trenti et al., 2007). Some empirical
Piotto, G. et al. 2005, ApJ, 621, 777
(Noyola et al., 2008) and theoretical Piotto, G. et al. 2007, ApJ, 661, L53
(Miocchi, 2007) indications have been Piotto, G. 2008, Mem. Soc. Astr. It., 79, 334
collected that seem to support this possi- Recio-Blanco, A. et al. 2006, A&A, 452, 875
Renzini, A. 2008, MNRAS, in press
Rich, R. M. et al. 1997, ApJ, 484, L25
The expected effects of the presence Salaris, M., Cassisi, S. & Pietrinferni, A. 2008, ApJ,
of an IMBH in the core region of a stellar 678, L25
Salaris, M., Cassisi, S. & Weiss, A. 2002, PASP, 114,
cluster are: (1) the BH would act as a
‘heat source’ in the central regions; (2) it Siegel, M. H. et al. 2007, ApJ, 667, L57
would strongly enhance the mass loss of Sollima, A. et al. 2005, ApJ, 634, 332
RGB stars passing close by; (3) it could Stanford, L. M. et al. 2006, ApJ, 647, 1075
Trenti, M. et al. 2007, MNRAS, 374, 857
trigger multiple star formation bursts. It
Ventura, P. & D’Antona, F. 2008, MNRAS, 385, 2034
is evident that these processes – if really Ventura, P. et al. 2001, ApJ, 550, L65
occurring – would allow many of the fea- Villanova, S. et al. 2007, ApJ, 663, 296
tures observed in the GCs hosting multi- Yong, D. & Grundahl, F. 2008, ApJ, 672, L29
populations to be explained.

The observational framework is becom-

ing more and more complex, but the new
empirical findings are of pivotal impor-
tance to shed light on the formation and
early evolution of GCs. Therefore, we
are now on the right path to piecing the
jigsaw puzzle together.


Anderson, J. 1997, PhD thesis, University of ­

California, Berkeley
Baumgardt, H., Makino, J. & Hut, P. 2005, ApJ, 620,
Bedin, L. R. et al. 2000, A&A, 363, 159
Bedin, L. R. et al. 2004, ApJ, 605, L125
Busso, G. et al. 2007, A&A, 474, 105
Calamida, A. et al. 2007, ApJ, 670, 400
Calamida, A. et al. 2008, in preparation
Caloi, V. & D’Antona, F. 2007, A&A, 463, 949
Carretta, E. et al. 2006, A&A, 450, 523
Cassisi, S. et al. 2008, ApJ, 672, L115
Choi, E. & Yi, S. K. 2007, MNRAS, 375, L1
Cohen, J. G., Briley, M. M. & Stetson, P. B. 2002,
AJ, 123, 2525
D’Antona, F. & Caloi, V. 2004, ApJ, 611, 871
Decressin, T. et al. 2007, A&A, 464, 1029
Ferraro, F. R. et al. 1998, ApJ, 500, 311
Ferraro, F. R., Bellazzini, M. & Pancino, E. 2002,
ApJ, 573, L95
Gratton, R. G. et al. 2001, A&A, 369, 87

The Messenger 134 | Supplement – December 2008 17

Conference Supplement

Linking Chemical Signatures of Globular Clusters to

Chemical Evolution

Francesca D’Antona clude carbon depletion and nitrogen presence of abundant lithium, due to be-
Paolo Ventura enhancement in a fraction of the cluster ryllium decay, in very high luminosity
INAF – Astronomical Observatory of stars, and it has been shown that the M giants that are above the highest lumi-
Rome, Italy increase in nitrogen is not only due to CN nosity limit for carbon stars. More re-
cycling, but must also include N produc- cently, the possibility that the anomalous
tion at the expense of oxygen (see, e.g., gas comes out from fast-rotating massive
The majority of Globular Clusters (GC) Cohen et al., 2002; 2005), thus relating stars has been explored by the Geneva
show chemical inhomogeneities in the Na-O with the C-N anticorrelations. group (e.g., Decressin et al., 2007), but
the composition of their stars, appar- this model has not yet been fully explored
ently attributable to a second stellar If neither supernovae that are due to the from the dynamical point of view.
generation in which the forming gas is core collapse of massive stars (SNe II),
enriched by hot CNO-cycled material nor supernovae that occur much later in In some recent work, the AGB enrich-
processed in stars belonging to a the life of the clusters (SNe Ia) had a ment scenario, although appealing for the
first stellar generation. We review the role to play in the process of self-enrich- dynamical reasons quoted above, has
reasons why the site of the nucleo­- ment, the most obvious source is often been considered inadequate to ex-
syn­thesis can be identified with hot- the massive AGB stars, as already pro- plain the features of the second stellar
bottom burning in the envelopes of posed in the 1980s, notably by Cottrell & generation (SG). This is attributable to the
massive asymptotic giant branch (AGB) Da Costa (1981). The hot CNO-processed fact that the results of stellar modelling
stars and super-AGB stars. The analy- matter of the AGB envelopes is injected of massive AGB stars obtained by differ-
sis of spectroscopic data and photo- into the cluster at low velocity by stellar ent groups differ greatly from each other.
metric signatures, such as the horizon- winds and planetary nebula ejection, and
tal branch morphology, shows that the the stellar remnants are quiet massive
percentage of ‘anomalous’ stars is 50 % white dwarfs (WDs). The very high tem-
or more in most GCs examined so far. peratures at the bottom of the convective Figure 1. Luminosity (upper) and temperature at the
If anomalies are the rule and not the envelopes (Tbce ) of these stars are exem- bottom of the con­vective envelope T bce (lower) in
exception, then they clearly are closely plified in Figure 1, and no peculiar extra models of 6, 5, 4 and 3 M A with metallicity Z = 0.001.
related to the dynamical way in which mixing needs be invoked to obtain the The abscissa is the current mass of the evolutionary
track, which decreases due to mass loss (from the
GCs form and survive. We show a pos- necessary very hot CNO processing. Hot- Ventura and D’Antona models). The very high T bce
sible solution obtained by a hydrody- bottom burning (HBB) was recognised values allow hot-bottom burning in all models down
namical model followed by the N-body as an important physical process in these to 4 M A, and only marginally in those of 3 M A. More
evolution of the two stellar populations, stars in the 1970s, and 3 He burning massive stars have faster nucleosynthesis and evo-
lution, a smaller number of thermal pulses and third
and propose that most GCs survive with 4 He, followed by the non-instantane- dredge-up episodes, so their evolution is able to
thanks to the formation of the second ous mixing of the resulting 7Be (Cameron- account for the chemistry of the second generation
stellar generation. Fowler hypothesis) could explain the stars.

The most massive AGB stars as sites of
nucleosynthesis of CNO, 23 Na, 27Al and  
+NF++ � 

The presence of the CN dichotomy, and
of the Na-O and Mg-Al anticorrelations  
among the stars of practically all GCs
examined in the Milky Way, show very
clearly that the matter of ‘anomalous’
stars must have been processed through
the hot CNO cycle, i.e. it has not only 
experienced CN, but also ON cycling.
This process occurs deep in stellar interi- 
ors during the H-burning stage or at
the ‘hot’ bottom of the convective enve- 
3ABD  *

lopes of massive asymptotic giant branch

stars. The so-called metals (iron, but also 
calcium and the other heavy elements)
do not show significant star-to-star varia- 
tions in most GCs (Gratton et al., 2004),
so that any process of production of the 
anomalous gas does not involve super-     
nova ejecta. Associated ‘anomalies’ in- ,, �

18 The Messenger 134 | Supplement – December 2008

The uncertainty in the nuclear cross-sec-
tions has often been emphasised in the
work of, e.g., Lattanzio, Karakas, Izzard
and Ventura. In particular, the very impor-
tant sodium yield is made up by a series 
of events: it first increases in the envelope
due to the second dredge-up (e.g., Iben
& Renzini, 1983); then increases due to
HBB of the 22 Ne dredged up in this same
event; afterwards, it increases in the third

dredge-up episodes. If T bce is high, so-

dium is destroyed by p-captures. Nota-
bly, however, the 23 Na(p,a) 20 Ne and,
especially, the 22 Ne(p,g) 23 Na cross-sec-
tions are very uncertain in the range of
temperatures of interest for HBB, by up
to a factor 10 3 for the latter (compare
Hale et al., 2002 with the ‘standard’ 
NACRE cross-sections found in Angulo
et al., 1999). By choosing a high rate
for this cross-section, the 23 Na yield of
massive AGBs, where HBB is actively
depleting oxygen, is high and compatible l     
with the observations. Also the 27Al pro- :.%D<
duction, by proton capture on 25 Mg, 26 Mg
and 26 Al, is crucially dependent on the schematically in Figure 2: the three diag- Figure 2. Schematic representation of the observed
Na-O anticorrelation (dots). The diagonal lines re-
relevant cross-sections, and selection of onal lines represent the Na and O yields
present schematic yields of AGB models, having de-
the highest possible rates in the NACRE (and their possible uncertainty) for AGBs creasing mass from left to right. The middle line
compilation allows the observed Mg-Al as a function of the AGB mass, decreas- ­represents yields that can be consistent with obser-
anticorrelation to be reproduced today. ing from left to right. It is clear that the vations; the lower line represents the yields ob-
tained by adopting smaller cross-sections for the
anticorrelation shown by observations 22
Ne(p,g) 23 Na reaction and the highest line repre-
The highest uncertainty in modelling (dots) cannot be explained by the occur- sents yields that can be obtained with a less efficient
is recognised to be the treatment rence of star formation in pure matter convection model. The dashed cone region includes
of ­convection: the discrepancy, e.g., be- expelled from AGBs of decreasing mass. abundances obtained by diluting the yields given
by the two open squares, along the chosen yield
tween the yields obtained by Karakas The AGB matter must be diluted, at some
line, with different amounts of pristine material hav-
& Lattanzio (2007) and ours is mainly due level, with pristine matter, providing ing the composition of the cone vertex. It is evi-
to this modelling (Ventura & D’Antona, ­values of Na and O intermediate between dent that a satisfactory interpretation of the data
2005). We adopt a very efficient convec- the AGB starting yield and the pristine requires such a dilution model, and cannot simply
rely on the ‘pure’ abundances of the ejecta. In this
tion model, proposed by Canuto et al. value. This is exemplified by the Na-O
scheme, the abundance points outside the cone
(1996), the Full Spectrum of Turbulence area within the cone drawn in the figure. (smaller O abundances, found only in red giants) re-
model, or FST. This results in a higher The Na-O anticorrelation thus requires quire some extra-mixing process in order to be ex-
Tbce and more efficient nuclear process- two events: (1) the minimum mass con- plained.
ing, higher luminosities, higher mass- tributing to the SG should not have so-
loss rates and consequently a lower num- dium abundance too high with respect to the anomalous stars. This may favour
ber of third dredge-up episodes. Thus the observed values; (2) the AGB matter the AGB scenario, indicating that a lim-
the oxygen reduction is not nullified by must be diluted with pristine gas. ited number of third dredge-up episodes
the third dredge-up, and the sodium actually play a (small) role. The most in-
is not increased too much by the dredge- A related observational and modelling teresting indications in favour of the mod-
up of 22 Ne (the ultimate product of 14 N problem is that of the CNO total abun- el came recently from the spectroscopic
burning in the helium intershell) and its dances. If they are constant among and photometric analysis of the clus-
consequent burning by p-captures. ­normal and anomalous stars, this im- ter NGC 1851: this is the only cluster for
plies that the AGB evolution must have which the total CNO variation may reach
Nevertheless, the sodium and oxygen suffered only a few episodes of third a factor ~ 4, as shown by Yong and col-
yields in the matter expelled from AGB dredge-up. Or it might indeed imply that laborators in this meeting, and the giants
stars are directly correlated: lower initial the matter is only CNO cycled, as can also show an s-abundance spread, un-
masses, with smaller Tbce, longer AGB happen preferentially in the evolution of like all other GCs (Yong & Grundahl,
lifetimes and a higher number of third massive stars. The CNO data of Carretta 2008). Are we witnessing a GC in which
dredge-up episodes, have higher oxygen (2005) seem to indicate a small, but un­- the formation of the SG was slightly de-
and sodium abundances. This is shown equivocal, overabundance of total CNO in layed with respect to the other clusters,

The Messenger 134 | Supplement – December 2008 19

Conference Supplement D’Antona F., Ventura P., Linking Chemical Signatures of GC to Chemical Evolution

so that it also involved the ejecta of ables; and the HB of the two metal-rich et al., 1999), and lose mass as ‘normal’
AGBs suffering some more third dredge- clusters NGC 6441 and NGC 6388 (Caloi (but quite massive and luminous)
up processing? The CNO total abun- & D’Antona, 2007; Busso et al., 2007). AGB stars, but different in core composi-
dance variation in NGC 1851 is probably The presence of strongly enhanced heli- tion (ONe versus CO) and core mass
also related to the splitting of the sub- um in peculiar HB stars has been con- (> 1.05 MA). Although full models through
giant branch and to its bimodal horizontal firmed in NGC 2808 and NGC 6441 by the super-AGB thermal pulse phase
branch (HB) morphology, as presented in spectroscopic observations (see, e.g., are not yet available, we can foresee that
other reviews. Moehler et al., 2004; 2007). In addition, the sum of the CNO abundances can
an unexpected feature has recently ap- remain close to the initial value as a result
We finally touch on the problem of the peared from photometric data, confirming of the efficiency of third dredge-up being
lithium abundance of unevolved stars in the interpretation of helium differences limited, because the helium luminosity
GCs. In spite of the difficulties of observ- amongst GC stars, viz. the splitting of the during the thermal pulses is weak, as
ing this element in low luminosity turn-off main sequence in NGC 2808. After first shown again by Siess (2007b). These are
stars, it has been known for many years indications from a wider-than-expected the premises that make super-AGBs
that 7Li in GCs varies from star to star, colour distribution found by D’Antona good candidates for the formation of the
much more than observed in halo stars. et al. (2005) from archival Hubble Space extreme helium population harboured in
In NGC 6752, Pasquini et al. (2005) have Telescope (HST) data, recent HST ob- the most massive GCs (Pumo et al., 2008).
shown that lithium is anticorrelated servations by Piotto et al. (2007) leave no
with sodium. Na-normal, Li-normal stars doubt that there are at least three differ- The fate of super-AGBs depends on the
do exist, as in the halo, and these should ent populations in this cluster. This finding competition between mass-loss rate
in fact be the first generation (FG) stars. came after the first discovery of a pecu- and core growth: if mass-loss wins, they
Other stars then formed from mixing of liar blue main sequence in w ­Centauri by evolve into massive ONe white dwarfs;
the pristine matter with Na-rich, Li-poor Bedin et al. (2004), also interpreted if the core grows until it reaches the
material, thus providing the anticorre­ in terms of a very high helium content Chandrasekhar mass, they evolve into
lation. But Pasquini et al. (2008) have re- (Norris, 2004; Piotto et al., 2005). The electron capture supernovae (ecSNe),
cently shown that two stars in NGC 6397, above-mentioned cases can be consid- electrons being captured on the Ne
both very similar and ‘normal’ in lithi- ered as extreme, in the sense that no nuclei. Thus a fraction of the super-AGBs
um, have oxygen abundances differing explanation had been attempted for them may explode as supernovae, but these
by ~ 0.6 dex. This may indicate that before the hypothesis of helium-enriched events are at least a factor ten less en-
the O-poor matter is Li-rich too, and this populations. ergetic than the SNe II (≤ 10 50 erg) and
is possible only if the SG forms from mat- also a factor ten less frequent than SNe Ia.
ter including AGB ejecta, and not mas- The helium yield of AGBs is in part due to Consequently, the epoch during which
sive star ejecta. The lithium yield of AGBs the second dredge-up, in part to the ef- super-AGBs evolve is probably the quiet-
is extremely dependent on the mass-loss fect of the third dredge-up episodes. As est period in the cluster lifetime, per-
rates, but models providing yields close the number of these third dredge-ups turbed at most by ecSN explosions. This
to the standard population II abundance must be small in order to preserve the stage is not energetic enough to alter
are very reasonable. quasi-constancy of CNO, the main effect either the gas evolution or its chemistry,
must be due to the second dredge-up. as the whole core remains locked by
While standard AGB models do not reach the remnant neutron stars. It has recently
Helium variations: super-AGBs as the helium yields larger than Y ~ 0.35, and been proposed that practically all the
site of production of extreme helium thus seem unable to explain the larger neutron stars (NS) present today in GCs
­populations Y values of w Centauri and NGC 2808, are born from ecSNe. Due to their lower
the super-AGB models by Siess (2007a) energy output, it is also probable that the
All models of self-enrichment show that show yields Y ~ 0.36–0.38, and sug- newborn NS receives a proportionally
the ejected material must be enriched gest these stars as candidates for the smaller natal kick, which allows it to re-
in helium with respect to the pristine one. progeny of the extreme GC population. main bound to the cluster (Ivanova et
The higher helium content has been rec- al., 2008). If there are no energy sources
ognised to have a strong effect on the HB Between the stars that evolve into core- (SN II or SN Ia) able to expel the gas
morphology, possibly helping to explain collapse supernovae and the minimum ejected at low velocity by stars, this gas
some features (gaps, hot blue tails, sec- mass for carbon ignition (below which can collapse into the cluster core and
ond parameter) that have defied alterna- stars evolve into CO white dwarfs), the form new stars with the chemistry of the
tive explanation (D’Antona et al., 2002). A super-AGB stars ignite carbon off-cent­- ejecta.
variety of problems along these lines re in semi-degenerate conditions, but are
have been examined in recent years: the not able to ignite hydrostatic neon-burn-
extreme peculiarity of the HB in the mas- ing in the resulting ONe core. Conse- What is the percentage of second
sive cluster NGC 2808 (D’Antona & Caloi, quently, degeneracy increases in the ­generation stars?
2004); the second parameter effect in core, and these stars may undergo ther-
M13 and M3 (Caloi & D’Antona, 2005); mal pulses, as first shown in models We point out that helium variations
the peculiar features in the RR Lyrae vari- by Iben’s group in the 1990s (e.g. Ritossa ­produce appreciable differences in the

20 The Messenger 134 | Supplement – December 2008

location of the main sequence of clus- Is the second generation necessary for means of N-body simulations, starting
ters only if they are very large and uniform the survival of globular clusters? with a highly concentrated SG, and an
(D’Antona & Caloi, 2004; Salaris et al., FG extended to the tidal radius. The FG is
2006). Small helium spreads can be A back-of-the-envelope computation is also expanding, due to the SN II explo-
revealed from the HB morphology, which enough to realise that the matter forming sions and consequent mass loss. If the
amplifies any small total mass decrease, the SG stars far exceeds the wind matter initial FG was already mass-segregated
by increasing the stellar Teff location on contained in massive AGBs, if the initial (as massive clusters seem to be obser­
the HB, but helium spread in the turn-off mass function (IMF) of the system is more vationally), the heating and the expansion
and main sequence stars remains hid- or less standard and we assume that due to the loss of SN ejecta are aug-
den in the observational errors. Therefore, the FG low mass stars we see today rep- mented by the preferential removal of the
we should not regard the clusters with resent the low mass end of the IMF. The mass from the inner regions of the clus-
split main sequences (w Centauri and initial population from which we need ter. D’Ercole et al. (2008) find that early
NGC 2808) as typical examples of clus- to collect AGB winds, massive enough to cluster evolution and mass loss can lead
ters with multiple stellar populations: they produce a populous SG, must have been to a significant loss of FG stars. In Fig-
are examples of clusters also harbour- at least a factor ten more massive than ure 3 we show that the number ratio of
ing an extreme population identified by its today’s cluster mass. This requirement SG to FG stars (fMS ) may not only reach
blue main sequence, corresponding to lends support to the idea that GCs are values consistent with observations
Y ~ 0.38–0.40. either the compact nuclei of dwarf galax- (0.5–1.5), but that there may also be evo-
ies (Bekki & Norris, 2006) or are formed lutionary routes leading to the loss of
D’Antona & Caloi (2008) have examined within dwarf galaxies that are afterwards most of the FG population and leaving an
the HB features of about 15 clusters and dispersed. SG-dominated cluster. Thus this model
have shown in most cases that the higher shows that the clusters that survive might
helium abundances remain confined be- A different point of view is assumed in preferentially be those in which a sub-
low Y ~ 0.32. Nevertheless, the percent- the recent work by D’Ercole et al. (2008): stantial SG has had time to form.
age of SG stars – defined now as all stars they start with a massive FG cluster,
with Y larger than the ‘standard’ Big and follow the hydrodynamic formation of
Bang abundance Y ~ 0.24 – is generally the SG. After the Type II supernova epoch, Back to the helium inhomogeneity
larger than 50 %! D’Antona & Caloi (2008) which cleared the cluster of its pristine
also pose the question of whether GCs gas, the low velocity winds of super- We have seen that, from the chemical
with only a blue HB (the classic ­second AGBs, and then of massive AGBs, collect point of view, the less extreme anomalies
parameter effect) should be explained by in a cooling flow in the innermost regions require mixing of the AGB ejecta with
assuming that these are clusters formed of the cluster, where they form SG stars. pristine gas in order to be explained. On
only from second generation stars. One The cluster emerging from the hydro­ the contrary, the dynamical model de-
of the inter­esting cases is NGC 6397, dynamical simulations is one with an SG scribed above is based on the fact that
the small cluster with an HR diagram that strongly concentrated at the inner core the SNe II have fully cleared the clus-
has always been regarded as a perfect of a more extended FG population. The ter of its pristine gas. A way to solve this
example of a simple stellar population, initial mass of the FG stars needs to problem has been approached in the
especially ­following the HST proper mo- be large enough to provide enough stel- model by D’Ercole et al. (2008): if the SNe
tion observations by King et al. (1998) lar mass return and to form a substantial have a preferential direction of ejection,
and most recently by Richer et al. (2006). number of SG stars. Con­sequently, the the ejected matter clears out a cone, and
Nevertheless, only three scarcely evolved FG stars are initially the dominant stellar leaves a torus of pristine gas in the out-
stars out of 14 are nitrogen ­normal population with a total mass that must be skirts of the cluster within the tidal radius.
­(Carretta et al., 2005), leading us to sus- about ten times larger than the total mass This pristine gas may be re-accreted (in
pect that the material from which most of the SG stars. The SG formation ends fact the hydro simulation shows that it
stars formed is CNO processed and thus when SNe Ia begin to explode regularly in is re-accreted) and mixes with the AGB
belonging to the SG. This occurrence had the cluster. As SN Ia explosions occur ejecta, providing the desired solution.
already been noticed by Bonifacio et al. when CO WDs reach the Chandrasekhar The model provides us with a very simple
(2002), with reference to the paradox that mass by accretion in binary systems, this solution to the problem of the three sep­
nitrogen-rich stars had almost normal epoch certainly begins some time after arate helium populations in NGC 2808. In
lithium content. The question remains the birth of massive CO WDs, so that the most massive clusters, the super-
whether NGC 6397 is an SG-only cluster. both super-AGBs and the most massive AGB winds are the first to be collected in
AGBs can contribute to the cooling flow, the cluster core, and the first SG stars
We are finally confronted with the real consistent with the scenario outlined are formed by ‘pure’ super-AGB ejecta,
question: how does a GC form? Is it pos- above. In 100 Myr (at most) a cluster with and with a homogeneous, very high, he-
sible to form a cluster with FG and SG two dynamically separated components lium content as the observations require.
stars in equal proportions? Is it possible has been formed. After a while, not only have the ejecta a
to have a cluster made up only of SG smaller helium content, since they come
stars? All GCs so far examined appear to The dynamical evolution of the com- from less massive AGBs, but they also
contain an SG! posite FG plus SG cluster is followed by become diluted by the pristine matter at

The Messenger 134 | Supplement – December 2008 21

Conference Supplement D’Antona F., Ventura P., Linking Chemical Signatures of GC to Chemical Evolution

standard helium Y ~ 0.24. A result of 

such a simulation, obtained without any
tuning of parameters, is shown in Fig-
ure 4, compared with the helium distribu- 
tion inferred for the stars in NGC 2808
(D’Antona & Caloi, 2008). 

Much more will no doubt be learnt about

GC formation in the near future, but at 
least one of the most difficult problems –
E ,2

the mass budget and the loss of the FG

stars – might be on the verge of being
fully understood.


Francesca D’Antona is grateful to Vittoria Caloi,
­A nnibale D’Ercole and Enrico Vesperini for the long-
lasting exchange of ideas and work that made 
this review possible. Francesca D’Antona is deeply
indebted to Achim Weiss for presenting the talk on            
her behalf. S,XQ

Figure 3 (above). Time evolution of the ratio of the Figure 4 (below). The empty histogram represents
References number of second generation (SG) to first generation the number versus helium content distribution
(FG) main sequence stars with 0.1 < M/M A < 0.8, (N(Y)) for stars in NGC 2808, derived by D’Antona &
Angulo, C. et al. 1999, Nucl. Phys. A, 656, 3 f MS, for ­different N-body simulations (D’Ercole et al., Caloi (2008) on the basis of the features of the hori-
Bedin, L. R. et al. 2004, ApJ, 605, L125 2008). Depending on the initial expansion velocity zontal branch and main sequence. Three distinct
Bekki, K. & Norris, J. E. 2006, ApJ, 637, L109 assumed for the FG (due to the mass loss of SNe II populations are present. The hatched red ­histogram
Bonifacio, P. et al. 2002, A&A, 390, 91 that are more or less concentrated in the cluster represents the second generation (SG) formation in a
Busso, G. et al. 2007, A&A, 474, 105 core), the SG could even become the dominant clus- dynamical model in which it is assumed that some
Caloi, V. & D’Antona, F. 2005, A&A, 121, 95 ter population. pristine gas of the first generation (FG) is in a torus at
Caloi, V. & D’Antona, F. 2007, A&A, 463, 949 the periphery of the cluster following the SN II
Canuto, V. M., Goldman, I. & Mazzitelli, I. 1996, ApJ, epoch. In this model there is a first phase of SG for-
473, 550 mation in the core of the cluster when only the
Carretta, E. et al. 2005, A&A, 433, 597 super-AGB winds are present, followed by a phase
Siess, L. 2007b, Why Galaxies Care About AGB
Cohen, J. G., Briley, M. M. & Stetson, P. B. 2002, AJ, during which the winds are diluted by pristine matter
Stars: Their Importance as Actors and Probes,
123, 2525 being re-accreted again. This two-phase pattern
in ASP Conf. Series, 378, 9
Cohen, J. G. & Meléndez, J. 2005, AJ, 129, 303 produces a gap in the N(Y) distribution. No attempt
Ventura, P. & D’Antona, F. 2005, A&A, 431, 279
Cottrell, P. L. & Da Costa, G. S. 1981, ApJ, 245, L79 has been made here to fit the two SG ­p opulations to
Yong, D. & Grundahl, F. 2008, ApJ, 672, L29
D’Antona, F. et al. 2002, A&A, 395, 69 the data (adapted from D’Ercole et al., 2008).
D’Antona, F. & Caloi, V. 2004, ApJ, 611, 871
D’Antona, F. & Caloi, V. 2008, MNRAS, 390, 693
D’Antona, F. et al. 2005, ApJ, 631, 868
Decressin, T. et al. 2007a, A&A, 464, 1029
D’Ercole, A. et al. 2008, MNRAS, in press %&- 2&EQNLOTQD
Gratton, R. G., Sneden, C. & Carretta, E. 2004, RTODQ &!DIDBS@
ARA&A, 42, 385 
Hale, S. E. et al. 2002, Phys. Rev. C, 65, 5801
Iben, I. & Renzini, A. 1983, ARA&A, 21, 271
Ivanova, N. et al. 2008, MNRAS, 386, 553
Karakas, A. & Lattanzio, J. C. 2007, PASA, 24, 103 2&EQNL &!
King, I. R. et al. 1998, ApJ, 492, L37 @MCOQHRSHMDL@SSDQ
Moehler, S. et al. 2004, A&A, 415, 313 
Moehler, S. et al. 2007, A&A, 475, L5

Norris, J. E. 2004, ApJL, 612, L25

Pasquini, L. et al. 2005, A&A, 441, 549
Pasquini, L. et al. 2008, A&A, 489, 315
Piotto, G. et al. 2005, ApJ, 621, 777
Piotto, G. et al. 2007, ApJ, 661, L53

Pumo, M. L., D’Antona, F. & Ventura, P. 2008, ApJ,
672, L25
Richer, H. B. et al. 2006, Science, 313, 936
Ritossa, C., Garca-Berro, E. & Iben, I. J. 1999, ApJ,
515, 381
Salaris, M. et al. 2006, ApJ, 645, 1131

Siess, L. 2007a, A&A, 476, 893

22 The Messenger 134 | Supplement – December 2008

Conference Supplement

Chemical Signatures in Dwarf Galaxies

Kim A. Venn 1 vey (SDSS; e.g., Belokurov et al., 2007) different from the results of today’s more
Vanessa M. Hill 2 data, have properties that are only now sophisticated (and physically accurate)
being uncovered (see presentations at models.
 niversity of Victoria, British Columbia,
U conferences/garcon08). Studies of the chemical abundances of
Canada stars in dwarf galaxies are more recent.
Observatoire de la Côte d’Azur, Nice, How these different types of galaxies are Shetrone et al. (1998) determined the first
France related to one another raises an interest- detailed chemistries for stars in the Draco
ing series of questions. Are dwarf gal­ dwarf galaxy. Ironically, they were mainly
axies related to protogalactic fragments, interested in using the stars in Draco
Chemical signatures in dwarf galaxies the low mass systems that formed in a to address the pattern of deep mixing
describe the examination of specific L Cold Dark Matter (LCDM) Universe that that is seen in red giant stars in globular
elemental abundance ratios to investi- later merged to build up the large spirals clusters, but never in similar field stars
gate the formation and evolution of that we see today? Are the dwarf sphe- in ­the Galaxy, but quickly realised the
dwarf galaxies, particularly when com- roidal and Sloan galaxies that we find in potential of this work for examining galaxy
pared with the variety of stellar popu­ the halo of the MWG related to the gas- formation. Previous to Shetrone’s work,
lations in the Galaxy. Abundance ratios rich dwarf irregular and gas-poor tran­ elemental abundances were determined
can come from H ii region emission sition galaxies that are further away and only for H ii regions, planetary nebulae,
lines, planetary nebulae, or supernova often isolated? One way to address these and bright supergiants in the Magellanic
remnants, but mostly they come from questions is to search for similarities Clouds (e.g., Olszewski et al., 1996). This
stars. Since stars can live a very long in the chemical patterns of the stars in can tell us about the end point of the
time, for example, a 0.8 MA star born at these galaxies. chemical evolution of these galaxies, but
the time of the Big Bang would only not of the initial conditions, nor the inter-
now be ascending the red giant branch, The build-up of the chemical elements is vening steps, because all of these objects
and, if, for the most part, its quiescent unique to each galaxy, depending on are young. Some carbon abundances
main sequence lifetime had been un­- their mass, initial conditions, star forma- had been determined for stars in the Ursa
eventful, then it is possible that the sur- tion histories and gas infall and outflow Minor dwarf galaxy (Suntzeff, 1985),
face chemistry of stars actually still properties. Since dwarf galaxies do not but no other elements were examined. It
resembles their natal chemistry. De- exchange gas with one another (other is impressive that we have gone from four
tailed abundances of stars in dwarf gal- than through major merging events that lone stars in one dwarf galaxy to hun-
axies can be used to reconstruct their leave only one galaxy remaining), then the dreds of stars in five dwarf galaxies in
chemical evolution, which we now find chemical evolution of each dwarf galaxy less than a decade, and with more stars
to be distinct from any other compo- is independent. So how does the chemis- and galaxies on the way.
nent of the Galaxy, questioning the as- try in each dwarf galaxy differ, or are they
sertion that dwarf galaxies like these all the same? Looking at the stars in the We are into the decade of large samples
built up the Galaxy. Potential solutions MWG suggests that they are not all going of stars in other galaxies with detailed
to reconciling dwarf galaxy abundances to be the same. chemical abundance determinations.
and Galaxy formation models include These datasets allow us to: (1) character-
the timescale for significant merging ise a wider variety of stellar populations;
and the possibility for uncovering differ- Chemical signatures: from the Milky Way (2) examine similarities and differences
ent stellar populations in the new ultra- to Local Group (dwarf) galaxies with respect to MWG stellar populations,
faint dwarfs. as well as between the various dwarf
The first studies of the chemical evolution ­galaxies; (3) constrain nucleosynthesis
of the stars in a galaxy occurred in the and stellar yields in the models; (4) exam-
The Local Group 1970s by Beatrice Tinsley and collabora- ine supernova (SN) feedback and reion­
tors (e.g., Tinsley, 1979). In these studies, isation effects on the evolution of dwarf
The Local Group seems to be more they collected the detailed elemental galaxies; (5) ­disentangle age and met­
crowded now than it was ten years ago. abundances in metal-poor stars in the allicity effects in the analysis of the red
Although we knew about the three large Galaxy and tried to model the build-up giant branch from colour-magnitude
spiral galaxies (the Milky Way – hereafter of the elements to the present day, as- ­diagrams – which can still be a compli-
MWG, M31, and M33), and we knew suming that the Galaxy formed from a cated procedure; and (6) couple metal-
about dwarf galaxies with smaller masses monolithic collapse (Eggen, Lynden-Bell licity and kinematic information to ex-
(the Magellanic Clouds and other smaller & Sandage, 1962). The achievements of amine variations in time and location of
satellites of the MWG and M31, as well as these early models are impressive. Look- star formation events, and/or galaxy
the isolated dwarfs), what we did not ing at Figures 3 and 6 from Tinsley (1979), interactions. Ultimately, all of these indi-
know was that there were also extremely an examination of the rise in s-process vidual questions are important in making
faint and low mass dwarf galaxies lurking elements through the evolution of asymp- comparisons with the MWG stellar pop­
within our midst. These galaxies, mostly totic giant branch (AGB) stars is not too ulations, testing galaxy formation scenar-
discovered in the Sloan Digital Sky Sur- ios and LCDM cosmology.

The Messenger 134 | Supplement – December 2008 23

Conference Supplement Venn K. A., Hill V. M., Chemical Signatures in Dwarf Galaxies

Chemical signatures in the Milky Way 

There are five dwarf galaxies to date with  

detailed chemistries determined for a

large sample of stars. These include the 
three dwarf spheroidal galaxies (Sculp-
tor, Fornax and Carina), as well as two l  
dwarf irregular galaxies (LMC and Sagit-
tarius). Most of these abundance analy-
ses have been from high resolution spec- 
tra taken at the VLT with the FLAMES
and UVES spectrographs. Examination of

the ­colour-magnitude diagrams for these 

five galaxies show that each has had a
very different star formation history (e.g., l
­Tolstoy et al., 2003; Smecker-Hane et
al., 2002; Bellazzini et al., 2006). The sim-
plest assumption is that they have had l   l l  l l   
very different chemical evolution routes :%D'<
from one another as well. But, do any of
them have a chemical history similar to ess elements (i.e., those that form through Figure 1. Comparison of the change in the calcium
and barium abundances relative to changes in the
any of the stellar populations in the Gal- slow neutron capture during the ther-
iron abundances of stars in the Galaxy. The long red
axy (e.g., halo, thin disc, retrograde stars, mal pulsing phases of an AGB star, e.g., lines are a guide to the yields of these elements dur-
etc.)? yttrium, strontium, barium). The reason ing the chemical evolution of the Galaxy. For [Ca/Fe]
that the s-process and r-process elements the high values at low [Fe/H] represent the yields
from massive star nucleosynthesis and Type II
All the chemical elements are of interest are particularly useful is not only their
­supernovae, while the downward slope after [Fe/H] =
in this examination, though some give ­different nucleosynthetic sites and time­ –1.0 represents the contribution only to iron from low
more information than others. a-elements scales for enrichment, but also that the mass Type Ia supernovae. For [Ba/Fe], the rising line
are particularly important when com- yields from these sites are metallicity at low [Fe/H] values represents the increasing yield
of barium over iron from Type II supernovae of higher
pared with iron group elements, be- dependent. Their dependence on metal-
metallicity, while the flat line after [Fe/H] = –1.8 re-
cause they have different nucleosynthetic licity (the seeds for these processes) presents the coincidentally similar yields of barium,
sources: iron forms during the ­surface makes the build up of these elements from intermediate mass AGB stars, and iron, from
detonation of a white dwarf in a Type Ia strongly coupled to the star formation his- low mass Type Ia supernovae.
supernova explosion; a-elements are tory of the host galaxy, which builds up
those that form through a-captures dur- over a non-unique timescale. Thus, varia-
ing nucleosynthesis (i.e., the capture of tions in star formation history, as well as
a helium nucleus, e.g., oxygen, magne- variations in SN feedback or gas infall/ Calcium, as an a-element, forms in mas-
sium, silicon, calcium, during quiescent outflow, can be probed with s-process/ sive stars. The high calcium/iron ratio
helium burning in the core of massive r-process element ratios and alpha/ at low metallicities suggests that massive
stars). Thus the alpha/iron ratio is similar r-process ratios. stars form both calcium and iron and
to examining the yields from hydrostat- deposit these elements back into the
ic burning in massive stars versus those In Figure 1, the abundances of calcium interstellar medium with the ratio seen on
from explosive nucleosynthesis in low (a-element) and barium (r-process and the plot. Eventually the low mass stars
mass stars. Differences in the star forma- s-process element) are compared to iron also evolve and explode as Type Ia super-
tion history of a galaxy will show up for stars in the Galaxy, as compiled by novae. These events create iron-group
as differences in the alpha/iron ratios. Venn et al. (2004). The element patterns elements without any calcium, causing
for these two elements are not the same. the calcium/iron ratio to decrease (form-
Of course, the simplest interpretation of When the iron abundance is quite low, ing a ‘knee’ in Figure 1). Taken togeth-
the alpha/iron ratios in stars can be com- then the calcium abundance does not er, the suggestion is that the Galaxy was
plicated by SN feedback, gas infall, or quite scale in the same way, such that enriched to a metallicity 1/10th solar
other events in the evolution of a galaxy. there appears to be an overabundance of ([Fe/H] = –1.0; location of the knee) by
Thankfully these additional processes calcium relative to the iron deficiency. massive stars alone, therefore rather
can have a different influence on the Barium, however, is even more deficient quickly since massive stars have very
abundances of other elements. Other than iron at low metallicities. These short lifetimes (< 1 Gyr), and after this
elements worth examining include ratios are plotted with respect to the Sun time the lower mass stars were able to
­r-process elements (i.e., those that form (which is at [Fe/H] = 0, and on a logarith- contribute iron. This also means that
through rapid neutron capture during mic scale). Why do these elements have the metallicity scale along the x-axis in
Type II supernova collapse, e.g., euro- different patterns? Figure 1 is absolutely not linearly related
pium, neodymium, gallium) and s-proc- to age! The first Gyr of our Galaxy’s

24 The Messenger 134 | Supplement – December 2008

chemical evolution is represented by metallicity stars in the Galaxy. Thus, the al. (2005) – green open symbols. While
most of the figure, and the past 12 Gyr stars in the Galaxy do not resemble those the calcium abundances are lower than in
by a tiny portion on the righthand side. in the LMC. The stars in the Galaxy sam- the Galaxy at intermediate metallicities,
ple are from the thick and thin disc, and they are quite similar to those of the very
Barium is an element that has contribu- it had been proposed that the discs of metal-poor Galactic stars, overlapping
tions from both massive stars during our Galaxy could have formed through the Galactic stars near [Fe/H] = –1.8. The
Type II supernovae and rapid neutron the merger of an LMC-sized dwarf galaxy barium abundances are not significantly
capture, as well as intermediate mass (e.g., Abadi et al., 2003). However, if that different from Galactic stars. These imply
stars during slow neutron capture hap- were true then the chemistry of this vir- some similarities in the early chemical
pening in the AGB phase. That the bar- tual galaxy must have been quite different evolution of the stars in the Sculptor dwarf
ium/iron ratio rises at the lowest met- from that of the LMC itself. galaxy to the MWG halo stars, however
allicities shows that Type II supernovae it also shows the power of having more
contribute less barium than iron, how- Chemical abundances for the Sagittarius than one element to examine. While there
ever as iron goes up, then more barium is (Sgr) and Fornax dwarf galaxies from are similarities in the most metal-poor
made and the contribution of barium Sbordone et al. (2007; cyan points in Fig- stars, variations in star formation histo-
increases. At a metallicity near 1/50th ure 2), Letarte et al. (2007; blue filled cir- ries, gas infall rates, or supernova feed-
solar ([Fe/H] = –1.8) then the barium/iron cles; 2006; blue open squares) and also back yields have affected the a-element
ratio is flat, implying that the yields of a few stars from Shetrone et al. (2003; abundances (calcium) at the time when
these elements are the same, but from blue open symbols) are also unlike the the galaxy had reached intermediate
what source? In the Galaxy, it appears stars in the MWG. Both also show lower metallicities. These results are similar to a
to be a coincidence that the barium yield calcium and higher barium than Galac- sample of stars analysed by Koch
from the s-process in the AGB stars tic stars, though Sgr probes stars at et al. (2008; magenta points in Figure 2)
is similar to the iron yield from Type Ia higher metallicities than the LMC, while in the low mass Carina dwarf galaxy.
supernovae (which contribute the iron at Fornax samples stars at slightly lower
this metallicity). metallic­ities. Sgr and Fornax have had
similar masses to the LMC, but the fact
Additionally, the position of the knee in that their chemical abundance ratios Figure 2. Calcium and barium abundances for stars
the barium plot differs from that of the ­differ from one another and those of the in the dwarf satellites of the Galaxy, including the
LMC (red), Sgr (cyan), Fornax (blue), Sculptor (green)
calcium plot in Figure 1 ([Fe/H] = –1.8 for LMC suggests that other processes and Carina (magenta). Open and closed circles are
barium/iron, rather than –1.0 for calcium/ have been important, and that the mass for field stars, open squares are results from stars in
iron). Since AGB stars can include higher of a dwarf galaxy alone does not fully globular clusters in the dwarf galaxies. Representa-
mass stars than Type Ia supernovae, determine its chemical evolution. tive error bars are shown for some stars in Sculptor.
Clearly each dwarf galaxy has had a different chem­
then these stars will evolve and contrib- ical evolution from the others and from the stellar
ute their products at earlier times (or The Sculptor dwarf galaxy has a lower populations in the Galaxy. This effect can be seen in
lower metallicities). mass than the LMC, which is reflected in the [Ca/Fe] plot by a variation in the metallicity
the lower metallicity of the majority of its when Type Ia supernovae start to contribute iron (the
knee where the ratio slopes downwards), and in the
stars as determined by Hill et al. (2008; [Ba/Fe] plot where the yields of s-process barium are
Chemical signatures in dwarf galaxies green points in Figure 2), with a few stars higher from the AGB stars (the high values occurring
from Shetrone et al. (2003) and Geisler et at high metallicities).
How do these abundance ratios, dis-
cussed in the MWG context, look in 
dwarf galaxies? Are they the same as in
the MWG? We might expect the ratios  
to be the same if the MWG formed from

ongoing merging of small dwarf galax- 

ies to the present epoch. The past dec-
ade of work on the chemistry of stars l  
in nearby dwarf galaxies has shown that
this is not the case, and that dwarf galax-
ies have their own unique chemical pat- 
terns and chemical signatures.

In Figure 2, the calcium and barium 

abundances are shown in the five dwarf
galaxies that have had their chemistries l
determined from large samples of stars.
The LMC stars analysed by Pompéia
et al. (2008; red) show lower calcium and l   l l  l l   
higher barium abundances than similar :%D'<

The Messenger 134 | Supplement – December 2008 25

Conference Supplement Venn K. A., Hill V. M., Chemical Signatures in Dwarf Galaxies

Interpreting chemical signatures in dwarf the same metallicity as that for Sculptor Fenner et al. analysis was the degeneracy
galaxies (in spite of their significantly different in the alpha/iron ratios between SFH and
masses and SFHs), whereas that for For- supernova feedback (see their Figure 2);
The most impressive result from the cal- nax is at a lower metallicity, [Fe/H] ~ – 2.0. the same pattern in alpha/iron ratios is
cium/iron and barium/iron ratios in Figure The Sgr remnant is not well sampled at possible no matter how different the star
2 is simply that each dwarf galaxy has low metallicities, but examination of the formation histories (they examined a SFH
its own chemical evolution. This was ex- slope suggests that the knee occurs at that ends at the moment of reionisation
pected since each has its own unique higher metallicity than in Sculptor, [Fe/H] compared with a SFH that continues to
colour-magnitude diagram, indicating sig- ~ –1.0. The current data for Carina has intermediate ages, see their Figure 1), so
nificantly different star formation histories too large a scatter to say much about the long as the supernova feedback is ad-
(SFH). Precise interpretation of effects metallicity at which low mass stars be- justed to the data. This brings into ques-
of the SFH versus other chemical evolu- gan to contribute iron; this could be due tion the value of the alpha/iron ratio as
tion parameters is complicated (see fur- to differences in the abundance analysis an indicator of chemical evolution! Fortu-
ther discussion below). At the very least, compared to the other analyses which nately, the heavy elements are not de-
we can see that the dwarfs with higher were done more homogeneously, or it generate in these two parameters; when
luminosities (and presumably masses) could be astrophysical and reflect a true the supernova feedback is adjusted to
that have had more ongoing star forma- and very large scatter in the Carina abun- fit the alpha/iron data, then the barium/
tion, as reflected in the distribution of dances due to its complex SFH. More iron ratio predictions are not the same.
stars in their colour-magnitude diagrams, data on this galaxy would certainly be This shows the power and necessity of
do have more stars at higher metallici- interesting. Looking at the knee in the having many different chemical elements
ties (e.g., the LMC versus Sculptor). While barium/iron ratios is similarly interesting. avail­able for chemical evolution model-
metallicity does not have to scale with ling of real systems. While their model
mass, it is reasonable to expect that gal- Each galaxy has similar barium/iron ratios with the continuous SFH fits the data from
axies with more baryons will be able to at the low metallicities sampled, which Shetrone et al. (2003) better than their
form more stars over time, and therefore suggests similar r-process yields with the other models, the larger sample size that
build up their metallicities to higher levels exception of Fornax, which may have we now have from Hill et al. (2008) no
than those with less gas. had a higher r-process contribution (or longer fits that model. Lanfranchi et al.
retention of r-process elements from (2008) have examined the chemical evo-
The two abundance ratios examined in Type II supernovae). However, the bar- lution of the heavy element abundances
Figures 1 and 2 tell us that the build up of ium/iron ratios are certainly higher in For- in six dwarf galaxies, including Sculptor,
these elements has occurred differently nax, the LMC and the Sgr remnant at Carina and the Sgr remnant, yet did
in each galaxy as well, e.g., new contri- higher metallicities, [Fe/H] > –1.0, when not predict the increasing abundances
butions from lower mass stars happened the barium abundance is mainly due to at high metallicities seen in Figure 2. New
at different metallicities in each galaxy s-process contributions from AGB stars. chemical evolution modelling of the
(we cannot say at different ages, because This pattern is not seen in the Galaxy, Sculptor and other dwarf galaxies are
there is no universal age-metallicity rela- nor the Sculptor dwarf galaxy. AGB yields now necessary.
tionship). This can be seen in the different are metallicity dependent – at lower met-
positions of the knee for each element allicities, there are fewer iron seeds for New modelling of dwarf galaxies is being
and in each galaxy. The knees them- slow neutron capture, thus first s-process carried out (e.g., Jablonka, this confer-
selves are difficult to see precisely be- peak abundances (such as yttrium) are ence). Marcolini et al. (2008) have used a
cause they occur at metallicities that sacrificed for the second and third 3D hydrodynamical simulation to exam-
were not well sampled in all of the dwarf s-process peak abundances (such as ine the chemical properties of the inner
galaxies, other than Sculptor. Examining barium and lead), e.g., Travaglio et al. regions of dwarf galaxies. They find that
only the calcium/iron ratio shows that (2004); thus the higher barium/iron abun- the stars in the inner region are relatively
the most metal-poor stars in each galaxy dances suggest that lower metallicity iron-rich and alpha-poor (as observed),
have similar alpha/iron ratios to the AGB stars contributed to these higher but the 3D aspects of the models show
metal-poor stars in the Galaxy, however, s-process yields and that iron is primarily that this pattern differs from the outer
at intermediate and high metallicities from the lower mass stars. We also regions and also that the kinematics of
then the alpha/iron ratios are lower by note that these three galaxies have had the outer regions are hotter (as observed
varying degrees. Thus, the contribution to more vigorous star formation rates at re- by e.g., Battaglia et al., 2006). This model
iron from low mass stars occurs at a dif- cent times (< 5 Gyr), thus the combina- does not currently examine the heavy
ferent metallicity in each galaxy; in Sculp- tion of their SFHs and chemical evolution elements.
tor, it occurs near [Fe/H] = –1.8, in Fornax has affected their recent AGB yields.
and the LMC it occurs before [Fe/H] =
–1.5, but we do not have data on a suffi- Chemical evolution models for the Sculp- Chemical comparisons between the
cient number of stars below that metallic- tor galaxy have been published by ­Fenner dwarf galaxies and the MWG
ity to be more certain. However, looking et al. (2006) and Lanfranchi et al. (2006)
at the slopes of the calcium/iron ratios with similar results to one another. One Only the metal-poor Galactic halo has
suggests that the LMC knee could be at of the most interesting results from the any chemical signatures in common with

26 The Messenger 134 | Supplement – December 2008

the dwarf galaxies; at the lowest metal- The third option for the difference in the References
licities in these systems, the alpha/iron MDF between the well-studied dwarf
Abadi, M. G. et al. 2003, ApJ, 597, 21
and heavy/iron abundances are in good spheroidal galaxies and the MWG is the Battaglia, G. et al. 2006, A&A, 459, 423
agreement (with the possible exception unknown characteristics of the old stel- Belokurov, V. et al. 2007, ApJ, 654, 897
of Fornax where the r-process barium/ lar populations in the isolated dwarf irreg- Bellazzini, M. et al. 2006, A&A, 446, 1
Brook, C. et al. 2007, ApJ, 661, 10
iron ratios may be higher; see Figure 2). ular galaxies. Although these galaxies
de Blok, W. J. G. & Walter, F. 2000, ApJ, 537, 95
If this is true, then it appears that the ear- are more luminous, contain gas and have Demers, S., Battinelli, P. & Kunkel, W. E. 2006, ApJ,
liest stages of star formation yield simi- current star formation, their old popula- 636, 85
lar results in all systems and/or that the tions have been unexplored chemically Eggen, O., Lynden-Bell, D. & Sandage, A. R. 1962,
ApJ, 136, 748
metal-poor halo of the Galaxy built up because of their distance and thus the
Fenner, Y. et al. 2006, ApJ, 646, 184
from the accretion of small dwarf galax- faintness of their red giant stars. The only Geisler, D. et al. 2005, AJ, 129, 1428
ies at the earliest epochs, before the detailed analyses from calcium triplet Helmi, A. et al. 2006, ApJ, 651, L121
dwarf galaxies had any significant chemi- spectroscopy of red giants in a dwarf Hill, V. M. et al. 2008, in prep.
Kirby, E. N. et al. 2008, ApJ, 685, 43
cal evolution of their own. irregular galaxy include ~ 20 stars in
Koch, A. et al. 2008, AJ, 135, 1580
NGC 6822 (Tolstoy et al., 2001) and ~ 80 Lanfranchi, G. A., Matteucci, F. & Cescutti, G. 2006,
Another way to test this proposition is stars in the Wolf-Lundmark-Melotte Gal- A&A, 453, 67
from the shape of the metallicity distribu- axy (WLM; Leaman et al., 2008). The Lanfranchi, G. A., Matteucci, F. & Cescutti, G. 2008,
A&A, 481, 635
tion function (MDF) of the most metal- analysis of the stars in WLM included 13
Leaman, R., et al. 2008, ApJ, submitted
poor stars. Helmi et al. (2006) compared old stars, which is a very small number, Letarte, B., Hill, V. M. & Tolstoy, E. 2007,
the metal-poor tails of the MDFs of the but enough to suggest that the old popu- EAS Publications Series, 24, 33
Galaxy and four dwarf spheroidal galax- lation may have spheroidal kinematics Letarte, B. et al. 2006, A&A, 453, 547
Marcolini, A. et al. 2008, MNRAS, 386, 2173
ies, but found the dwarf galaxies all have unlike the younger populations or the H i
Olszewski, E. W., Suntzeff, N. & Mateo, M. 1996,
similar and sharper declining tails (their gas that is rotationally supported. ARAA, 34, 511
Figure 3). The conclusion was that the Pompéia, L. et al. 2008, A&A, 480, 379
metal-poor Galactic halo could not come The calcium triplet data for the red giants Sbordone, L. et al. 2007, A&A, 465, 815
Schoerck, T. et al. 2008, arXiv:0809.1172
from the accretion of dwarf spheroidal in NGC 6822 did not include stars with
Shetrone, M. D., Bolte, M. & Stetson, P. B. 1998, AJ,
galaxies. There are three other possible old ages (from isochrones), however simi- 115, 1888
interpretations. ­Schoerck et al. (2008) lar work on the carbon stars (AGB stars Shetrone, M. D. et al. 2003, AJ, 125, 684
have rescaled the Galactic MDF for a with intermediate ages) by Demers et al. Simon, J. & Geha, M. 2007, ApJ, 670, 313
Smecker-Hane, T. et al. 2002, ApJ, 566, 239
minor observational bias and selection (2006) has come to the same conclusion,
Suntzeff, N. 1985 in ESO Workshop on Production
function of the Hamburg/ESO Survey that the young stars have disc-like kine- and Distribution of C, N, O Elements, European
sample and suggest that the new Galac- matics, but not the AGB stars which have Southern Observatory, 83
tic MDF has a sharper metal-poor tail, in spheroidal kinematics. Even the H i disc Tinsley, B. M. 1979, ApJ, 229, 1046
Tolstoy, E. et al. 2003, AJ, 125, 707
good agreement with the dwarf spheroi- is peculiar in NGC 6822; de Blok & Walter
Tolstoy, E. et al. 2001, MNRAS, 327, 918
dal galaxies. The rescaled MDF is nor- (2000) have shown there is a hole in the Travaglio, C. et al. 2004, ApJ, 601, 864
malised and compared to the dwarf gal- H i distribution on one side of the disc and Venn, K. A. et al. 2004, AJ, 128, 1177
axies at [Fe/H] = – 2.0, a value that may an apparent overdensity on the other side
be too high since significant chemical (which they suggest could be the core
evolution can occur in the dwarfs by the of a recently merged dwarf). Could all of
time this metallicity is reached. However the dwarf irregular galaxies really be
normalising and comparing the MDF at dwarf spheroidal galaxies that have had a
[Fe/H] = – 2.5, as was done by Helmi recent merger with a gas-rich system (or
et al. (2006) has a less significant effect even just an H i filament), where the gas is
on the original MDF and thus improves in the orbital plane of the merger? Simu-
the comparison with the dwarfs, but still lations by Brook et al. (2007) do suggest
does not eliminate the inconsistencies. that polar ring galaxies could be a natu-
The second option relies on the contri­ ral and common occurrence in the evolu-
bution of the newly discovered ultra-faint tion of the dwarf galaxies. It is exciting
dwarf galaxies (e.g., Belokurov et al., that this is a testable prediction through
2007). The majority of the stars in these examination of the metallicities and kin­
galaxies are more metal-poor than the ematics of the Local Group dwarf irregu-
majority of stars examined in the other lar galaxies, and that these galaxies,
dwarf galaxies (e.g., Simon & Geha, since they formed and evolved in relative
2007; Kirby et al., 2008). Characterising isolation, could be different to the dwarf
these galaxies and examining their MDFs spheroidals and provide new information
could provide a missing link in our cur- on the nature of unperturbed dwarfs at
rent testing of the hierarchical accretion early epochs.
of small systems in the early stages of
galaxy formation.

The Messenger 134 | Supplement – December 2008 27

Conference Supplement

Chemical Evolution of Dwarf Galaxies and

Stellar Clusters: Conference Summary

Kenneth C. Freeman spite their wide range of star formation irregular galaxies, although rotation has
Research School of Astronomy and histories and Galactic locations. The recently been discovered for the first time
Astrophysics, The Australian National present low star formation rates in the in a dSph galaxy (Scl). The globular clus-
University, Canberra, Australia star-forming dSph and dIrr galaxies ters are unlike any of the known clusters
may contribute to their similar abun- that are currently forming stars within the
dance-luminosity relations. In contrast, Galaxy. Some of the younger clusters
The summary begins by considering globular clusters do not show such a in the Large Magellanic Cloud (LMC) and
the contributions on the differences relation. This important discovery indi- other galaxies are, however, very similar
­between globular clusters (GCs) and cates that the dSph galaxies make their in mass and structure to the old Galactic
dwarf spheroidal (dSph) galaxies. Then own heavier elements during their GCs. We do not yet understand the con-
I ­discuss globular clusters: the topics (mostly) extended star formation histo- ditions that are needed to form the globu-
include multiple sequences in the ries, while the GCs primarily inherit their lar clusters.
colour-magnitude plane, light element heavier elements from the gas out
abundance anomalies, globular clus- of which they formed. The tight relation
ter systems and other issues. The next between the stellar luminosity of dSph Globular clusters
section is devoted to dwarf galaxies, galaxies and their chemical abundance
summarising new results on their kine- also suggests that tidal stripping of Multiple populations
matics, masses and baryon content, stars at later times may not be impor-
their star formation histories, and their tant for determining the low stellar Several of the (mainly) more luminous
chemical abundance ranges and chem- ­luminosities of the fainter dSph galax- Galactic GCs show multiple sequences in
ical signatures. The last section dis- ies: for example, dSph galaxies with their colour-magnitude distributions. In
cusses some of the other interesting low [Fe/H] and low stellar luminosities w Centauri, five giant branches and two
related points that came up during the today probably did not have signifi- main sequences are seen. The bluer main
meeting. cantly higher stellar masses in the past. sequence is more metal-rich, its stars
– dSph galaxies have a range of M/L are more centrally concentrated in the
ratios extending up to very large values, cluster, and it is believed to be He-rich,
Differences between globular clusters indicating that most of their mass is in with Y ~ 0.4. There is evidence that the
and dwarf spheroidal galaxies the form of dark matter. Globular clus- more metal-rich stars in w Centauri are
ters have low M/L ratios and appear to younger. NGC 2808 has three main se-
We can summarise the important differ- be baryonic. quences, with three different inferred He
ences between GCs and dSph galaxies –d  Sph galaxies have low baryonic den­ abundances, and the cluster shows a
as follows: sities and their associated stellar re- strong extension of the stellar distribu-
– dSph galaxies overlap with GCs in laxation times are greater than a Hubble tion in [Na/O]. Double subgiant branches
absolute magnitude (Mv), but they have time. For most globular clusters, the are seen in the clusters NGC 1851,
much lower stellar surface densities, relaxation times are shorter than a NGC 6388 and M54, and the subgiant
larger half-light radii (rh ) and are more Hubble time. splitting in NGC 1851 appears to be
elliptical in shape. In the (Mv, rh ) plane, associated with chemical abundance dif-
the classical globular clusters and dSph There are clear environmental effects on ferences.
galaxies are well separated, although the incidence of the nearby dwarf gal­
some of the recently discovered fuzzy axies and GCs as a function of Galacto­ Although most of the multiple-popula-
GCs fall in the gap. centric radius Rg. The inner halo (Rg < tion clusters are relatively massive, not all
– dSph galaxies have a wide range of 30 kpc) is inhabited mainly by globular of the massive clusters show multiple
star formation histories, while the stars clusters. Then, out to Rg = 100 kpc, most populations: 47 Tuc is an example of
in GCs are more nearly coeval. of the satellite systems are dSph galax­- a massive cluster with a single popula-
– dSph galaxies have a variety of chemi- ies with primarily old stellar populations. tion.
cal signatures, as seen most clearly Going out further to Rg = 300 kpc, the
in their wide range of stellar a-element dSph galaxies have mostly extended star The multiple main sequences are be-
abundances [a/Fe] as a function of formation histories. Even further out, lieved to require different He abun-
[Fe/H] within individual dSph galaxies. dwarf irregular galaxies dominate the sat- dances, with values as high as Y ~ 0.4 to
In contrast, most GCs have a relatively ellite population, presumably due to generate the main sequence splitting
tight and homogeneous distribution of some kind of interaction of the satellites and the associated horizontal branch
stellar abundances. with the parent galaxy. morphology. The observed abundances
– dSph galaxies are now found to follow in NGC 2808 provide a constraint on
a fairly tight abundance-luminosity rela- We still do not know much about how the the enrichment scenario: it must avoid
tion, extending down to the least lumi- dSph galaxies and GCs fit into the over- enriching the stars in CNO and heavier
nous of the newly discovered systems. all picture of galaxy formation and evolu- elements like Fe and the a-elements.
This abundance-luminosity relation is tion. The nature of the progenitors of Current ideas include pollution by high
shared by the dwarf irregular galaxies today’s dSph galaxies is not understood. temperature H-burning in a first gener­
(dIrr), dSph and transition objects, de- They are probably not like today’s dwarf ation of massive asymptotic giant branch

28 The Messenger 134 | Supplement – December 2008

(AGB) stars or rapidly rotating massive Mg-Al relations and the long-known CNO tion peaks at a similar abundance to the
stars. At least two generations of star for- anomalies. These relations are seen field stars. The clusters show a fairly wide
mation are needed in such clusters, and down to the main sequence in some range of [a/Fe], but the [a/Fe] values ap-
the main sequence splitting requires the clusters, so they are believed to be pear uncorrelated with metallicity or clus-
process to generate discrete levels of Y, imprinted on the cluster stars at birth. ter age.
not just a spread in Y.
Again, the primary idea is that pollution The GC system of M31 includes clusters
Other ideas include the enhanced He by high temperature H-burning in a first that are more extended than Galactic
coming from first star enrichment, and generation of massive AGB stars or GCs of similar absolute magnitude. These
the possibility that at least some GCs ­rapidly rotating massive stars is respon­­- clusters appear to be of intermediate
are the nuclei of primodial dwarf galaxies sible. At least two generations of star luminosity (– 8 < Mv < – 5) and it is these
in which the He had gravitionally settled ­formation in GCs are again needed. The extended clusters that fall in the gap
in their dark matter mini-halos. details are still far from understood. between the Galactic GCs and dSph gal-
axies in the (Mv, rh ) plane.
At this stage, the enrichment scenario The observation that the Na-O anticorre-
that led to the main sequence splitting lation is not seen in the halo field stars Some of the old GCs in elliptical galaxies
in w Centauri and NGC 2808 is far from may place a limit on the contribution of appear to be extremely a-enhanced –
understood. dissolving GCs to the Galactic stellar halo. by more than 0.5 dex. The corresponding
However it is possible that those GCs yields require very rapid enrichment
Some of the intermediate-age globular that are not going to survive are mostly on Myr timescales only by very massive
clusters in the LMC also show multiple destroyed quite quickly, on timescales stars (> 20 MA).
subgiant branches. NGC 1846 is an ~ 50 Myr. If the source of the Na-O anti-
example. This cluster has an age of about correlation takes longer to act (e.g., if AGB
2 Gyr: if its subgiant splitting is due to stars are involved), then the dissolving Dwarf spheroidal galaxies
an age difference between the branches, clusters would not have suffered the light
the corresponding age difference is element evolution and so would not affect The dark matter content
about 300 Myr. Such an age difference the Na-O properties of the Galactic halo.
could be associated with the merger A large amount of new kinematical data
of binary clusters or with multiple star for- The idea that large and inhomogeneous has been acquired for several of the
mation episodes. The much younger clusters like w Centauri are the surviv- most recently discovered low luminosity
LMC cluster, NGC 1850, with an age of ing nuclei of accreted and stripped galax- dSph galaxies. Like most dSph galax-
about 90 Myr, provides encouragement ies has been around for about 20 years ies, these faint systems are dominated
for both scenarios: it is a binary system and may be relevant to these problems of dynamically by their dark matter, with M/L
and is surrounded by an envelope of chemical inhomogeneity. The nuclei of ratios between about 100 and 1000. The
gas which would provide fuel for a later low luminosity spiral galaxies are much kinematical data allow the mass of dark
episode of star formation. like massive GCs in velocity dispersion, matter within the region populated by
mass, surface density and sub-solar met- stars to be estimated. It turns out that the
allicity, and some show direct spectro- dark matter mass within a standard
Light element anomalies scopic evidence for continuing episodic radius of 300 pc is almost the same for
star formation. For the nuclei of star-form- all known dwarf spheroidals, at about
Most globular clusters appear to be ing galaxies, star formation in the sur- 10 7 MA, although these galaxies have
homogeneous in their internal stellar rounding galaxy can provide multiple stellar masses from about 10 2.5 –10 7 MA.
[Fe/H] and a-element distributions, generations of chemical enrichment. The The corresponding virial dark mass for
but this is not the case for the lighter ele- problem is to get the enriched gas into these systems is probably much larger,
ments. All GCs with adequate stellar the nucleus: this is likely to be a sporadic ~ 10 9 MA.
abundance data show a marked anticor- dynamically driven process, deliver-
relation of stellar [Na/Fe] with [O/Fe] ing discrete levels of enrichment at a few Dynamical analysis of the velocity disper-
within individual clusters, with spreads in particular times. sion profiles and the stellar surface
[Na/Fe] and [O/Fe] exceeding 1 dex in ­density distributions indicates that the
some systems. This anticorrelation is not dark halos of dSph galaxies have
seen among the field stars with similar Globular cluster systems cores (rather than cusps), with central
[Fe/H] abundances, so the Na-O anomaly densities 0.1 MA pc – 3.
is clearly related to the cluster environ- The relation between the chemical prop-
ment, but how is not yet understood. The erties of globular cluster systems and The stellar mass-metallicity relation for
extent of the [Na/O] spread is related to the underlying stellar population of the the dSph galaxies described above in-­
the maximum effective temperature of the parent galaxy remains poorly understood. dicates that the present baryon content
horizontal branch stars in the cluster. New results on the cluster system of was established very early in the lumi-
Other light-element abundance relations NGC 5128 show that most of its GCs are nous life of these systems. Why do the
within clusters include the Na-Li and old and their metallicity distribution func- baryon masses vary so widely (by ~ 10 5),

The Messenger 134 | Supplement – December 2008 29

Conference Supplement Freeman K. C., Conference Summary

although their dark matter masses within ated by AGB star evolution and reflecting – New methods have been developed to
300 pc are so similar? Were the baryons differences in the gradual rise of s-proc- measure accurate element abund-
in the faintest systems lost during their ess abundances with time. ances of extragalactic globular clusters
early evolution, or were they never ac- from high resolution spectra of their
quired? It may be that, at such low dark Stars in the low luminosity Hercules dSph integrated light.
halo masses, the acquisition of baryons show near-solar [Ca/Fe], but an unusually – New data on stellar ages and metallici-
is a stochastic process. high [Mg/Ca] ratio, in excess of +1, sug- ties in the disc of the LMC, at radii
gesting chemical enrichment by just one in the range 3–8 degrees from the LMC
or two high mass SNe II with progenitor centre, show that the age-metallicity
Star formation history and chemical masses ~ 35 MA. relation is very similar in each of the
­evolution fields.
Recent abundance studies of the ultra-
The dSph galaxies show a great range of faint dSph galaxies show that their metal-
star formation histories. All appear to licity distributions extend down below
have old populations. Several show very [Fe/H] = – 3. The metal-poor end of the
clear multiple episodes of star forma- stellar metallicity distribution function
tion in their stellar colour-magnitude dis- (MDF) for the dSph systems now looks
tributions. As mentioned above, there more like that for the Galactic halo. The
is a marked environmental effect in the dSph galaxies appear to follow a stellar
distribution of systems with different luminosity-metallicity relation extending
star formation histories: the incidence of down below [Fe/H] = – 3, although their
more recent star formation increases dark matter masses within 300 pc radius
with Galactocentric radius. The relevant are all very similar. This indicates that
enviromental factor is not yet clear: is their diverse chemical properties are evo-
it tidal interaction, ram pressure stripping, lutionary rather than acquired.
or photoevaporation? In those dSph
­galaxies with extended star formation The distribution of stellar abundances
histories, where does the gas come from along the Sgr stream shows a strong
to fuel the extended star formation? ­gradient away from the core of the Sgr
system. The overall abundance signa-
New colour-magnitude diagrams for large tures (a, s-process elements) look more
numbers of stars in several of the dSph like those in the Large Magellanic Cloud
galaxies (Draco, Ursa Minor, Sextans) (LMC) than in other dSph galaxies.
delineate their different evolved star pop-
ulations and show similar rich blue strag-
gler sequences. Blue stragglers are often Other related issues
interpreted in terms of binary star merg-
ers: the implications of these new data –N  ew stellar abundance data in the Ga-
are not yet clear. lactic Bulge show that the mean abun-
dance decreases with increasing
Detailed abundance data for many ele- Galactic latitude. This finding may argue
ments are now available for large ­samples against the formation of the Bulge by
of stars in several dSph galaxies. Each secular processes.
galaxy shows a wide spread in [Fe/H], – The ACS survey of Galactic globular
and some show abundance gradients. clusters provides a new age-metallic-
The relationship between [Fe/H] and the ity relation for the clusters. Clusters with
abundances of other elements differs [M/H] < –1.3 are all old. Clusters with
from galaxy to galaxy. The most metal- [M/H] > –1.3 split into two families. One
poor stars are mostly a-enriched, but the family is all old, as old as the metal-
[a/Fe] ratio begins to decrease with in- poor clusters. The other family, which
creasing [Fe/H] at different [Fe/H] abun- includes several clusters associat-
dances from galaxy to galaxy, reflecting ed with recent likely accretion events, Right. A colour image of
the Galactic globular
the different star formation timescales. shows a clear age-metallicity rela- cluster NGC 6397. The
This is unlike the rather well-defined tion with age decreasing as [M/H] image is a composite
[a/Fe]-[Fe/H] relation for the stars in the ­increases. of exposures in the B-,
Solar Neighbourhood. Similar diversi- –T  he metal-poor halo of M31 extends V- and I-bands obtained
with the Wide Field
ties are seen in s-process element abun- out to a radius of at least 165 kpc, with Imager (WFI) at the
dances within the dSph galaxies, gener- a mean abundance of ~ –1.4 at a radius 2.2-m MPG/ESO tele-
of 100 kpc. scope.

30 The Messenger 134 | Supplement – December 2008

Conference Supplement

ESO Messenger Supplement Contents

Review articles of the conference
“Chemical Evolution of Dwarf F. Primas, A. Weiss – Chemical Evolution of Dwarf Galaxies and
Galaxies and Stellar Clusters” Stellar Clusters 2
www.mpa-garching.mpg.de/mpa/ M. Mateo – The Complex Evolution of Simple Systems 3
conferences/garcon08 R. Gratton – Abundances in Globular Cluster Stars: What is the Relation with
Dwarf Galaxies? 9
ESO Headquarters S. Cassisi et al. – Evidence for Sub-Populations in Globular Clusters:
Karl-Schwarzschild-Straße 2 Their Properties and Relationship with Cluster Properties 13
85748 Garching bei München F. D’Antona, P. Ventura – Linking Chemical Signatures of Globular Clusters to
Germany Chemical Evolution 18
Phone +49 89 320 06-0 K. A. Venn, V. M. Hill – Chemical Signatures in Dwarf Galaxies 23
Fax +49 89 320 23 62 K. C. Freeman – Chemical Evolution of Dwarf Galaxies and Stellar Clusters:
information@eso.org Conference Summary 28

Printed by Back Cover Picture: A colour composite Wide Field Imager B-, V- and I-band
Peschke Druck image of the Galactic globular cluster exposures. See ESO Release 44/08 for
Schatzbogen 35 w Centauri, which displays abun­­- more details.
81805 München dance anomalies and multiple main
Germany sequences, and emerged as a refer-
ence object of its class from the confer- Front Cover: Based on the conference
© ESO 2008 ence reviews. The image was formed announcement poster designed by
ISSN 0722-6691 from Max-Planck/ESO 2.2-m telescope ESO/L. Calçada.