2026ApJ...997..153B

A merger origin has been suggested for M83's massive, metal-rich extended H I disk and nuclear starburst. We observe M83's stellar halo to test this idea. We train nearest-neighbor star-galaxy separation on wide-area Subaru imaging with Hubble Space Telescope data to map M83's halo in resolved stars. We find that M83 has an extended, very low-density smooth stellar halo of old and metal-poor [M/H] ∼ -1.15 RGB stars with a mass between 15 and 40 kpc of ${\mathrm{log}}_{10}{M}_{\ast ,15-40,\mathrm{maj}}/{M}_{\odot }=8.02\pm 0.10$. In addition to M83's well-known Northern Stream, our ground-based Subaru imaging reveals a new stream to M83's south, which modeling suggests could be its trailing arm. The combined stream masses are ${\mathrm{log}}_{10}{M}_{\mathrm{stream}}/{M}_{\odot }=7.93\pm 0.10$, with metallicity [M/H] = -1.0 ± 0.2. The stream progenitor was only recently accreted, as its stellar populations suggest that it formed stars until 2.1 ± 1.3 Gyr ago. M83 lies on the stellar halo mass-metallicity correlation seen for other Milky Way-mass galaxies, albeit with low stellar halo mass. We infer a total accreted mass of ${\mathrm{log}}_{10}{M}_{* ,\mathrm{accreted}}/{M}_{\odot }=8.7{8}_{-0.28}^{+0.22}$, with the most massive past merger having ${\mathrm{log}}_{10}{M}_{* ,\mathrm{dom}}/{M}_{\odot }=8.5\pm 0.3$. We identify plausible M83 analogs in TNG-50 with similar stellar halos, finding that while a recent accretion can create a prominent stellar stream, such accretions do not trigger starburst activity, nor do they deliver enough gas to form M83's extended H I disk. We conclude that other nonmerger mechanisms, such as secular evolution or accretion of gas from the intergalactic medium, are likely to be responsible for M83's remarkable properties.