5.3. Relative Abundance Patterns
Figure 19 shows the variation of log N/O with radius in NGC 2403 for our HII region sample. The weighted least-squares fit to the data yields a gradient of -0.032 ± 0.005 dex/kpc (to be compared with the O/H gradient of -0.201 ± 0.009 dex/kpc). This is very similar to the gradient found by Víchez et al. (1988) for N/O in M33.
When discussing relative abundance gradients in spiral galaxies, one has to be careful to distinguish between gradients in terms of galactic radii, and gradients relative to metallicity baselines. I will try to follow my own advice. The rather shallow gradient in N/O is at odds with the notion that N is a purely secondary element: in such a situation N/O should vary linearly with O/H and the N/O radial gradient should have the same slope as the O/H radial gradient. A shallower slope is expected if N has a primary nucleosynthesis component (Matteucci & Tosi 1985; Díaz & Tosi 1986), as predicted in some models for the third dredge-up on the asymptotic giant branch of intermediate mass stars (e.g., Renzini & Voli 1981). Such a primary component has been invoked to explain the nearly constant N/O vs. O/H observed in dwarf irregular galaxies (Section 4.3.2). The N/O gradient can be explained in this picture if there is a primary N component at a level log N/O = -1.4, similar to the mean value observed in the dwarf irregulars.
However, there is a possible systematic uncertainty in the gradient due to the poorly known correction for N+2, which is predicted to be the dominant ionization state of N in many cases (e.g., Garnett 1990). Far-infrared spectroscopic measurements of [N III] and [O III] in Galactic HII regions (e.g., Lester et al. 1987) suggest that higher N/O ratios are obtained from these lines than from the optical [N II] and [O II] lines. Until the optical/IR discrepancies for nitrogen are resolved, the true gradient for N in NGC 2403 (and in spiral galaxies in general) remains uncertain.