Nearly all current simulations predict that outcomes of the star formation process, such as the fraction of stars that form in bound clusters (Γ), depend on the intensity of star formation activity (ΣSFR) in the host galaxy. The exact shape and strength of the predicted correlations, however, vary from simulation to simulation. Observational results also remain unclear at this time, because most works have mixed estimates made from very young clusters for galaxies with higher ΣSFR with those from older clusters for galaxies with lower ΣSFR. The three blue compact dwarf (BCD) galaxies ESO 185-IG13, ESO 338-IG04, and Haro 11 have played a central role on the observational side because they have some of the highest known ΣSFR and published values of Γ. We present new estimates of Γ for these BCDs in three age intervals (1–10 Myr, 10–100 Myr, 100–400 Myr), based on age-dating, which includes Hα photometry to better discriminate between clusters younger and older than ≈10 Myr. We find significantly lower values for Γ(1–10 Myr) than published previously. The likely reason for the discrepancy is that previous estimates appear to be based on age–reddening results that underestimated ages and overestimated reddening for many clusters, artificially boosting Γ(1–10 Myr). We also find that fewer stars remain in clusters over time, with ≈15%–39% in 1–10 Myr clusters, ≈5%–7% in 10–100 Myr clusters, and ≈1%–2% in 100–400 Myr clusters. We find no evidence that Γ increases with ΣSFR. These results imply that cluster formation efficiency does not vary with star formation intensity in the host galaxy. If confirmed, our results will help guide future assumptions in galaxy-scale simulations of cluster formation and evolution.
Read full abstract