DIRECT OXYGEN ABUNDANCES FOR LOW-LUMINOSITY LVL GALAXIES

We present MMT spectroscopic observations of H II regions in 42 low luminosity galaxies in the Spitzer Local Volume Legacy survey. For 31 of the 42 galaxies in our sample, we were able to measure the temperature sensitive [O III] lambda 4363 line at a strength of 4 sigma or greater, and thus determine oxygen abundances using the direct method. Our results provide the first direct estimates of oxygen abundance for 19 of these galaxies. Direct oxygen abundances were compared to B-band luminosities, 4.5 mu m luminosities, and stellar masses in order to characterize the luminosity-metallicity and mass-metallicity relationships at low luminosity. We present and analyze a Combined Select sample composed of 38 objects (drawn from a sub-set of our parent sample and the literature) with direct oxygen abundances and reliable distance determinations (based on the tip of the red giant branch or Cepheid variables). Consistent with previous studies, the B band and 4.5 mu m luminosity-metallicity relationships for the 38 objects were found to be 12 + log(O/H) = (6.27 +/- 0.21) + (-0.11 +/- 0.01) M-B and 12 + log(O/H) = (6.10 +/- 0.21) + (-0.10 +/- 0.01) M-[4.5] with dispersions of sigma = 0.15 and 0.14, respectively. The slopes of the optical and near-IR L-Z relationships have been reported to be different for galaxies with luminosities greater than that of the LMC. However, the similarity of the slopes of the optical and near-IR L-Z relationships for our sample probably reflects little influence by dust extinction in the low luminosity galaxies. For this sample, we derive a mass-metallicity relationship of 12 + log(O/H) = (5.61 +/- 0.24)+(0.29 +/- 0.03) log(M-star), which agrees with previous studies; however, the dispersion (sigma = 0.15) is not significantly lower than that of the L-Z relationships. Because of the low dispersions in these relationships, if an accurate distance is available, the luminosity of a low luminosity galaxy is often a better indicator of metallicity than that derived using certain strong-line methods, so significant departures from the L-Z relationships may indicate that caution is prudent in such cases. With these new direct metallicities we also revisit the 70/160 mu m color metallicity relationship. Additionally, we examine N/O abundance trends with respect to oxygen abundance and B-V color. We find a positive correlation between N/O ratio and B - V color for 0.05 less than or similar to B - V less than or similar to 0.75: log(N/O) = (1.18 +/- 0.9) x (B - V) + (-1.92 +/- 0.08), with a dispersion of sigma = 0.14, which is in agreement with previous studies.