Starbursts superimposed on old populations: spectral evolution of the composite system over $3 \times {10}^{9}$ yr

We simulate in this paper the occurrence of star formation events superimposed on old populations, during the last $$3 \times {10}^{9}$$ yr, by combining star cluster integrated spectra of different ages with those of red strong-lined galaxy nuclei. As the young star clusters have metallicities in the range $$-0.5\lt[\text{Z}/{\text{Z}}_{\odot}]{\leqslant} 0$$, our simulations address in particular the case of starbursts induced by galaxy interactions bringing fresh, low-metallicity gas to the host galaxy. We follow the evolution of the composite spectrum for burst to old population mass ratios of 10, 1 and 0.1 per cent. We present also the effects of the starburst on the restframe BVRI magnitudes and colours, as well as on the equivalent widths ( W_λ) for a set of metallic and Balmer lines. If a burst amounts to 10 per cent of the galaxy mass, the underlying old population of the galaxy will remain undetectable, even in the near infrared, during $$5\times {10}^{7}$$ yr. When this strong burst reaches intermediate ages $$({10}^{9}\,\text{yr}\,\lt\,t\,\lt\,5\times {10}^{9}$$ it still contributes ≈ 20 per cent of the total visible flux. For a 1 per cent mass burst, the underlying galaxy should be marginally detectable, in terms of near-infrared absorption features, during the H II region phase. However, because of the a priori unknown burst duration, any attempt to identify the old population by means of integrated observations will be difficult owing to the prompt occurrence of the red supergiant phase. At intermediate ages, the 1 per cent mass burst has become barely visible. Finally a 0.1 per cent mass burst affects the galaxy spectrum during $$2\times {10}^{7}$$ yr only. It is virtually inconspicuous in the visible and near-infrared ranges as soon as $$t\,{\geqslant}{10}^{8}$$yr. We discuss a series of spectral configurations of interest, whenever the burst and galaxy contributions are of comparable importance, depending on the burst age and strength. These results hold true in the more general frame of star formation, and are not restricted to interactioninduced events.