Radial gradients in chemical composition are characteristic of spiral galaxies. Searle (1971) studied the systematic variation of the [N II], [O II], and [O III] emission-line intensities of GEHRs in several spiral galaxies as a function of galactocentric distance. His analysis indicated radial decreases in O/H and N/O. The latter was consistent with the suggestion by Peimbert (1968) that the strong [N II] emission from the nuclei of M51 and M81 reflected a high abundance of nitrogen. Searle's work was followed by other observational studies and by analyses involving computer models of the nebular ionization and thermal structure (e.g., Smith 1975; Shields & Searle 1978). As results for a substantial number of galaxies became available (e.g., McCall, Rybski, & Shields 1985; Zaritsky, Kennicutt, & Huchra 1994, ZKH), analysis of the systematics of abundances became possible. Radial gradients may be fit by an exponential in R / R0, where R0 is the isophotal radius. Barred spirals have gradients substantially shallower than normal spirals, but similar overall abundances (Martin & Roy 1994; ZKH). Low surface brightness spirals have relatively low abundances for their mass (McGaugh 1994). Local abundances increase with the local surface surface brightness or surface mass density (McCall 1982; Edmunds & Pagel 1984; Vila-Costas & Edmunds 1992).
The Milky Way has a gradient in O/H of -0.06 dex/kpc (HW, and references therein). This is supported by studies of H II regions and planetary nebulae (Maciel & Köoppen 1994). Carraro, Ng, & Portinari (1998) note that a variety of theoretical models for the chemical evolution of the Galaxy match the radial gradient at the present time, but have quite different predictions for the time evolution of the gradient. They conclude from the available stellar data, in particular for open clusters, that the gradient has changed relatively little with time. This conclusion seems in harmony with the fact that non-barred spirals have similar gradients, which might not be expected if gradients changed greatly during the course of galactic evolution.
Space observatories have made possible the measurement of carbon emission lines in H II regions. Results for spiral and irregular galaxies have revealed an unexpected trend in which C/O increases substantially with increasing O/H (Garnett et al. 1999; Garnett 2001b; and references therein). This may be explained in terms of the effect of stellar winds on the evolution of massive stars. The winds carry off more mass at higher abundances, causing the evolution of the stellar core to be arrested and the yield of oxygen to be decreased, relative to that of carbon (e.g., Carigi 1996).