2021A&A...654A.132B

The relation between nuclear (≤50 pc) star formation and nuclear galactic activity is still elusive; theoretical models predict a link between the two, but it is unclear whether active galactic nuclei (AGNs) should appear at the same time, before, or after nuclear star formation activity. We present a study of this relation in a complete, volume-limited sample of nine of the most luminous (log L_14–195keV > 10^42.5 erg s^–1) local AGNs (the LLAMA sample), including a sample of 18 inactive control galaxies (six star-forming; 12 passive) that are matched by Hubble type, stellar mass (9.5≤log M*/M_☉≤10.5), inclination, and distance. This allows us to calibrate our methods on the control sample and perform a differential analysis between the AGN and control samples. We performed stellar population synthesis on VLT/X-shooter spectra in an aperture corresponding to a physical radius of ≃150 pc. We find young (≤30 Myr) stellar populations in seven out of nine AGNs and in four out of six star-forming control galaxies. In the non-star-forming control population, in contrast, only two out of 12 galaxies show such a population. We further show that these young populations are not indicative of ongoing star formation, providing evidence for models that see AGN activity as a consequence of nuclear star formation. Based on the similar nuclear star formation histories of AGNs and star-forming control galaxies, we speculate that the latter may turn into the former for some fraction of their time. Under this assumption, and making use of the volume completeness of our sample, we infer that the AGN phase lasts for about 5% of the nuclear starburst phase.