Elemental Abundance Variations and Chemical Enrichment from Massive Stars in Starbursts. I. NGC 4214

This spectroscopic study of NGC 4214 is part of a project to determine the extent to which the winds and supernovae from massive stars contribute to the short-term, localizedchemical enrichment of the interstellar medium (ISM) in low-metallicity galaxies. Long-slit optical spectroscopy at 82 distinct spatial locations covering multiple starburst knots reveals no significant localizedO, N, or He abundance differences that might be attributed to the winds of massive stars. We do find large-scale(200 pc) variations in the oxygen abundance. The southernmost, and probably youngest, starburst region exhibits higher O abundances by 0.095±0.019 dex and correspondingly lower N/O (0.108±0.038 dex) than the rest of the bright emission-line regions. This difference is consistent with O pollution at the locations of the young starburst, possibly from recent supernovae. The surveyed regions exhibit an anticorrelation between N/O and O/H consistent with O pollution, but no significant correlations between Ne/O and O/H. If, as expected from nucleosynthesis models, O and Ne are produced predominantly in the same short-lived massive stars, and N is produced predominantly in less massive, longer lived stars, then these trends are both consistent with the O pollution hypothesis. We show that observed internal N/O and O variations within NGC 4214, NGC 5253, and NGC 3125 are consistent with theoretical chemical evolution predictions during a phase of O production associated with massive stars early in a starburst. While uncertainties in the electron temperature, T(O++) are the dominant source of error, we show that the observed abundance trends are not consistent with those expected from temperature uncertainty effects. The N/O and He/H ratios are relatively insensitive to temperature errors and should serve as good indicators of abundance inhomogeneities wherever they may exist.