5.3.5. PNe probe the histories of nearby galaxies

A wealth of data exist for large samples of PNe in the Magellanic Clouds, both in the optical and in the UV (Monk et al. 1988, Boroson & Liebert 1989, Meatheringham & Dopita 1991a, b, Vassiliadis et al. 1992, Leisy & Dennefeld 1996, Vassiliadis et al. 1996, 1998). PNe in the Magellanic Clouds represent a statistically significant sample at a common distance, suffering little extinction along the line of sight, and sufficiently bright to allow the measurement of diagnostic lines from various ions.

The oxygen abundances of PNe in the Magellanic Clouds span a relatively small range: log O/H + 12 = 8.10 ± 0.25 from a compilation of 125 objects for the LMC, log O/H + 12 = 7.74 ± 0.39 from a compilation of 48 objects for the SMC reanalyzed in a homogeneous way by Stasinska et al. (1998). If one considers only the high luminosity sample (L_{[O III]} > 100 L), the spread is smaller and the mean abundance is significantly larger: 8.28 ± 0.13 (40 objects) for the LMC, 8.09 ± 0.11 (11 objects) for the SMC. This has been interpreted as due to the fact that, as a class, high luminosity PNe have progenitors of higher masses, therefore younger and made of more chemically enriched gas. The mean oxygen abundance in the high luminosity class compares well with that from H II regions in the Magellanic Clouds: 8.35 ± 0.06 for LMC, 8.03 ± 0.10 for SMC (Russell & Dopita 1992). This indicates that the oxygen abundance in luminous PNe is a very good proxy of the present day ISM oxygen abundances.

Dopita et al. (1997) have produced self consistent photoionization models to fit the observed line fluxes between 1200 and 1800 Å for 8 PNe in the LMC. With these models they obtain not only the elemental abundances, but also the temperatures and luminosities of the central stars. This allows them to place the objects in the HR diagram and derive the central star masses and post-AGB evolution times by comparison with theoretical tracks for post-AGB stars of various masses (the choice of H-burning or He-burning track for each object is made by the requirement of consistency with the observed expansion age of the nebula). Assuming the initial-final mass relation of Marigo et al. (1996), Dopita et al. (1997) are able to estimate the masses of the progenitors. This allows them to trace the age-metallicity relationship in the LMC. As a proxy of metallicity, they use the sum of the abundances from the -process elements Ne, S, Ar (in order to alleviate any doubts that might come from the use of O whose abundance can be slightly affected by mixing processes). They find that the LMC experienced a long period of quiescence, followed by a short period activity within the past 3 Gyr which multiplied its metallicity by a factor 2. A further study is under way by the same autors to include 20 additional PNe in the LMC and 10 PNe in the SMC.

PN spectroscopy is now possible with relatively high signal-to-noise even in more distant galaxies. For example, observations of 28 PNe in the bulge of M31 and 9 PNe in the companion dwarf galaxy M32 allowed to obtain T_e-based abundances for these objects (Richer et al. 1999). The oxygen abundances of the PNe observed in the bulge of M31 are found to be very similar to those of the luminous PNe in the Galactic bulge (the comparison, in order to be meaningful, must be done on nebulae with similar luminosities, since the oxygen abundances has been shown to depend on luminosity in the Magellanic Clouds and the Galactic bulge). One finds log O/H + 12 = 8.64 ± 0.23 for the M31 bulge sample and 8.67 ± 0.21 for the high luminosity PNe in the Galactic bulge (Stasinska et al. 1998). Jacoby & Ciardullo (1999) obtained spectroscopic data on 12 PNe in the bulge and 3 in the disk of M31. They span a larger luminosity range than Richer et al. (1999) who were mainly interested in bright PNe. For the three objects in common with Richer et al. (1999), the oxygen abundances are in excellent agreement. Yet, for their entire sample, Jacoby & Ciardullo (1999) find log O/H + 12 = 8.50 ± 0.23 which is significantly lower than the value found by Stasinska et al. (1998), possibly because of the larger range of PNe luminosities in their sample.

The data on M32 by Richer et al. (1999) confirm the suggestion by Ford (1983) that the PNe in M32 are nitrogen rich. It seems unlikely that all the luminous PNe have high enough central star masses to undergo second dredge up, and this finding suggests that in M32 nitrogen was already enhanced in the precursor stars.

Other local group galaxies have smaller masses and therefore contain only a few PNe. Abundance data exist for PNe in NGC 6822, NGC 205, NGC 185, Sgr B2, Fornax (see references in Richer & McCall 1995 and Richer et al. 1998).

Richer & McCall (1995) compared the oxygen abundances from PNe in diffuse ellipticals and dwarf irregulars. They found that diffuse ellipticals have higher abundances than similarly luminous dwarf irregulars. This seems consistent with the idea that diffuse ellipticals would be the faded remnants of dwarf ellipticals. However, when considering also the O/Fe ratios, obtained by combining stellar abundance measurements, they conclude that diffuse ellipticals and dwarf ellipticals have had in fact fundamentally different star formation histories.

Combining the data on PNe in these dwarf spheroidals galaxies with those on PNe in M32 and in the bulge M31 and of the Milky Way, Richer et al. (1998) have shown that the mean oxygen abundance correlates very well with the mean velocity dispersion. Since the oxygen abundance of luminous PNe is a good proxy of the oxygen abundance in the ISM at the time when star formation stopped, this implies that there is a correlation between the energy input from supernovae and the gravitational potential energy. Such a correlation arises naturally if chemical evolution in these systems is stopped by Galactic winds.

The oxygen abundances found in the elliptical galaxy NGC 5128 (Centaurus-A) by Walsh et al. (1999) show a mean value of about 8.4, i.e. smaller than the mean value determined for the bright PNe in M31. This result is somewhat difficult to understand for such a massive galaxy, unless the most metal rich stars do not produce observable PNe. This possibility is known as the the AGB manqué phenomenon (see e.g. Greggio & Renzini 1990), by which intermediate mass stars do not reach the top of the AGB due to intense stellar winds.