The sensitivity of Lick indices to abundance variations

We present results of model atmosphere/line formation calculations which quantitatively test how the 21 classical and four higher-order Balmer-line Lick/IDS indices (Worthey et al. 1994; Worthey & Ottaviani 1997) depend on individual elemental abundances (of carbon, nitrogen, oxygen, magnesium, iron, calcium, natrium, silicon, chromium, titanium) and overall metallicity in various stellar evolutionary stages and at various metallicities. At low metallicities the effects of an overall enhancement of alpha-elements are also investigated. The general results obtained by Tripicco & Bell (1995) at solar metallicity are confirmed, while details do differ. Tables are given detailing to which element every index reacts significantly, as a function of evolutionary stage and composition. This work validates a number of assumptions implicitly made in the stellar population models of Thomas, Maraston & Bender (2003), which utilized the results of Tripicco & Bell (1995) to include the effects of element abundance ratios variations. In particular, these computations confirm that fractional changes to index strengths computed at solar metallicity (and solar age) can be applied over a wide range of abundances and ages, also to model old stellar populations with non-solar abundance ratios. The use of metallicity-dependent response functions is, however, required for the proper modelling of the Balmer-line indices. While H beta still responds only moderately to abundance ratio variations, the higher-order Balmer indices H gamma and H delta display very strong dependencies at high metallicities. As shown in Thomas, Maraston & Korn (2004), this result allows to remove systematic effects in age determinations based on different Balmer-line indices. (abridged)