2016A&A...585A.156O

Polar ring galaxies (PRGs) are composed of two kinematically distinct and nearly orthogonal components, a host galaxy (HG) and a polar ring/disk (PR). The HG usually contains an older stellar population than the PR. The suggested formation channel of PRGs is still poorly constrained. Suggested options are merger, gas accretion, tidal interaction, or a combination of both. To constrain the formation scenario of PRGs, we study the compact stellar systems (CSSs) in two PRGs at different evolutionary stages: NGC4650A with well-defined PR, and NGC3808B, which is in the process of PR formation. We use archival HST/WFPC2 imaging in the F450W, F555W, or F606W and F814W filters. Extensive completeness tests, PSF-fitting techniques, and color selection criteria are used to select cluster candidates. Photometric analysis of the CSSs was performed to determine their ages and masses using stellar population models at a fixed metallicity. Both PRGs contain young CSSs (<1Gyr) with masses of up to 5x10⁶M_☉, mostly located in the PR and along the tidal debris. The most massive CSSs may be progenitors of metal-rich globular clusters or ultra compact dwarf (UCD) galaxies. We identify one such young UCD candidate, NGC3808B-8, and measure its size of r_eff=25.23^+1.43_–2.01pc. We reconstruct the star formation history of the two PRGs and find strong peaks in the star formation rate (SFR, ≃200M_☉/yr) in NGC3808B, while NGC4650A shows milder (declining) star formation (SFR<10M_☉/yr). This difference may support different evolutionary paths between these PRGs. The spatial distribution, masses, and peak star formation epoch of the clusters in NGC3808 suggest for a tidally triggered star formation. Incompleteness at old ages prevents us from probing the SFR at earlier epochs of NGC4650A, where we observe the fading tail of CSS formation. This also impedes us from testing the formation scenarios of this PRG.