by Andrea Cattaneo from Nature 521, 164–165 (14 May 2015) doi:10.1038/521164a
For most galaxies, the shutdown of star formation was a slow process that took 4 billion years. An analysis of some 27,000 galaxies suggests that ‘strangulation’ by their environment was the most likely cause.
In humans, death by strangulation is a slow process that takes about four minutes. During this time, the victim uses up oxygen in the lungs but keeps producing carbon dioxide, which remains trapped in the body. On Nature, Peng et al.(1 present evidence of an analogous slow ‘strangulation’ process that ends the formation of stars in many galaxies by disrupting the supply of gas that accretes onto those galaxies from the environment. Instead of building up CO2, the strangled galaxies accumulate metals — elements heavier than helium — produced by massive stars.
Almost 90 years ago, Edwin Hubble classified galaxies into three morphological types: spirals, ellipticals and lenticulars (Fig. 1). In spirals, stars form a disk and turn around in circles like the horses on a cosmic merry-go-round. Ellipticals are the wrecks of galaxy crashes, in which stars move chaotically in all directions. Lenticulars form an intermediate type between the two. Most spirals are blue because they contain young blue stars, but elliptical and lenticular galaxies contain little or no cold gas to make new stars, and so only old red stars remain. Historically, astronomers have been more interested in the morphologies of galaxies than in their colour. Attention switched to colour after the Sloan Digital Sky Survey (SDSS) measured the spectra of hundreds of thousands of galaxies.
The SDSS demonstrated that blue (star-forming) and red (passive) galaxies form distinct populations(2). Since then, various hypotheses have been put forward to explain what causes galaxies to transition from one type to the other. Most revolve around two ideas. The first is that the gas in ellipticals and their surroundings is too hot to make stars and does not cool efficiently. The second is that the gas that could cool and make stars is kept hot or blown away by phenomena linked to the growth of supermassive black holes, which are found at the centres of all ellipticals(3). The main motivation for considering violent expulsion scenarios comes from computer simulations of the formation of ellipticals in galaxy mergers. These simulations need a mechanism that gets rid of gas to avoid forming blue cores(4), which are rare in real ellipticals.
Peng et al. present evidence that the formation of stars in most passive galaxies ended through a slow strangulation process. The authors compared the metal content of the stars in approximately 23,000 passive galaxies from the SDSS with that of a control sample of about 4,000 star-forming galaxies, also from the SDSS. They discovered that the metal content of the former is systematically larger than that of the latter, at least for galaxies that have stellar masses up to 100 billion times the mass of the Sun, the limit mass (M*) above which galaxies become scarce. This constitutes evidence for galactic ‘suffocation’, in the same way that high levels of CO2 in the blood of a corpse suggest suffocation.
From the difference in the metal content of the stars of passive and star-forming galaxies, Peng et al. inferred a delay of 4 billion years (or 2 billion years for galaxies close to M*) between the time that gas stopped being supplied and the time that star formation ended. This delay is consistent with the mean age difference between passive and star-forming galaxies (about 4 billion years at all masses).
As any forensic scientist will tell you, suffocation does not imply strangulation. But the difference in metal content is higher for galaxies in groups and clusters than it is for isolated galaxies, suggesting that crowded environments strangle galaxies by disrupting the accretion of gas onto them. This disruption might occur either through ram pressure (the pressure exerted on a body moving through a fluid medium) or through tidal forces.
The slow shutdown of star formation inferred by Peng et al. from observations of galaxies with masses lower or equal to M* contrasts with the fast shutdown behaviour of much larger galaxies, such as giant ellipticals. The stars of giant ellipticals have a low iron content because they were made on a short time span (less than 0.3 billion years for a galaxy of mass greater than 3 M*)(5) — that is, there was not enough time for many type Ia supernova explosions, the source of iron. Because there are many more red galaxies below M* than there are above, Peng et al. are correct to argue that strangulation is the main mechanism for star-formation shutdown. But the different chemical properties of low- and high-mass red galaxies imply that they must have formed through different routes. Morphological differences back this interpretation: red galaxies with masses above M*/3 are all ellipticals or lenticulars; but below M*/3, 40% of them are red spirals(6), as would be expected if strangulation has occurred.
There is also a difference between ellipticals and lenticulars. Most elliptical galaxies that have stellar masses in the range of 1 M* to 2 M* were already in place at a redshift of between 2 and 3, a period when the Universe was about one-fifth of its present age. Lenticulars appeared more gradually, replacing a pre-existing population of spiral and irregular galaxies (M. Huertas-Company, personal communication). The ratio of ellipticals to lenticulars is approximately 3:5 for galaxies of approximately 1.5 M*, but lenticulars should predominate in the mass range explored by Peng and colleagues. Further evidence for the presence of two evolutionary tracks has been obtained by comparing the evolution of the star-formation rate of galaxies with the evolution of the galaxies’ size(7).
Cosmological models(8) of the formation and evolution of galaxies predict two mechanisms by which star formation shuts down. In these models, galaxies more massive than 1.5 M* reside at the centres of groups and clusters. They grow extremely rapidly until, at a redshift of about 3 (which corresponds to when the Universe was one-quarter of its current size), they attain the critical mass at which infalling gas is effectively shock-heated. Some violent phenomenon then quenches star formation. However, the models also predict that galaxies below 0.6 M* become red much later (at a redshift of less than 0.5) and much more gradually, in most cases because they stop accreting gas after becoming part of a group or a cluster. Thanks to Peng and colleagues’ work, this second theoretical prediction is now an observational fact.
(1) Peng, Y., Maiolino, R. & Cochrane, R. Nature 521, 192–195 (2015).
(2) Baldry, I. K. et al. Astrophys. J. 600, 681–694 (2004).
(3) Cattaneo, A. et al. Nature 460, 213–219 (2009).
(4) Di Matteo, T., Springel, V. & Hernquist, L. Nature 433, 604–607 (2005).
(5) Thomas, D., Maraston, C., Bender, R. & Mendes de Oliveira, C. Astrophys. J. 621, 673–694 (2005).
(6) Bell, E. F., McIntosh, D. H., Katz, N. & Weinberg, M. D. Astrophys. J. Suppl. 149, 289–312 (2003).
(7) Barro, G. et al. Astrophys. J. 765, 104 (2013).
(8) Cattaneo, A., Woo, J., Dekel, A. & Faber, S. M. Mon. Not. R. Astron. Soc. 430, 686–698 (2013).