The baryonic mass-size relation of galaxies. II. Implications for the evolutionary paths between star-forming and passive galaxies

TL;DR

This study integrates a large passive galaxy sample with the SPARC star-forming set to map structural relations in the baryonic and stellar mass–size planes across morphologies. Using Bayesian linear fits on $M_{\rm bar}$, $R_{50,\rm bar}$, and $\Sigma_{50,\rm bar}$ (and their stellar analogs), it identifies four distinct passive families (Es/bulges, S0 disks, non-nucleated dwarfs, nucleated dwarfs) and shows that star-forming high-surface-density disks overlap with S0 disks, while star-forming low-surface-density disks offset from passive dwarfs in baryonic planes but align with dwarfs in stellar planes. The results imply that spirals can become lenticulars as they exhaust gas, whereas dwarfs evolve into passive dwarfs primarily through external gas removal rather than internal consumption; UDGs appear as an extension of non-nucleated dwarfs without requiring substantial stellar expansion. Overall, the work clarifies plausible evolutionary pathways and highlights the critical role of environment in driving gas removal, with future wide-field HI and NIR surveys anticipated to broaden the parameter space and environmental contexts.

Abstract

The baryonic mass-size relation of galaxies links the total baryonic mass (stars plus gas) to the baryonic half-mass radius. In the first paper of this series, we showed that star-forming galaxies from the SPARC sample follow two distinct relations in the baryonic mass-size plane: one defined by high-surface-density (HSD), star-dominated, Sa-to-Sc galaxies, and one defined by low-surface-density (LSD), gas-dominated, Sd-to-dI galaxies. In this second paper, we study the structural relations between baryonic mass, half-mass radius, and mean surface density to constrain possible morphological transformations between star-forming and passive galaxies. We complemented the SPARC sample with $\sim$1200 passive galaxies that are nearly devoid of gas: ellipticals (Es), lenticulars (S0s), dwarf ellipticals (dEs) or dwarf spheroidals (dSphs), and the so-called `ultra-diffuse galaxies' (UDGs). Our results can be summarised as follows: (1) passive stellar components follow four distinct relations at high statistical significance, namely (i) ellipticals plus bulges, (ii) S0 disks, (iii) non-nucleated dwarfs (dEs, dSphs, UDGs), and (iv) nucleated dEs; (2) star-forming HSD disks (mostly Sa to Sc) overlap with S0 disks within 2$σ$ in the baryonic relations and within 1$σ$ in the stellar ones, so present-day spirals may simply evolve into lenticulars as they run out of gas; (3) star-forming LSD disks (mostly Sd to dI) are offset from non-nucleated passive dwarfs at more than 3$σ$ in the baryonic relations, but the two galaxy populations overlap within 1$σ$ in the stellar relations, suggesting that non-nucleated passive dwarfs may form from star-forming dwarfs only after gas removal; (4) UDGs extend the sequence of non-nucleated dEs/dSphs and may originate from the most diffuse star-forming LSD galaxies with no need for a substantial expansion of the stellar component.

Figures (7)

• Figure 1: Baryonic structural relations for passive galaxies: $M_{\rm bar}$ vs $R_{\rm 50, bar}$ (top left), $M_{\rm bar}$ vs $\Sigma_{\rm 50, bar}$ (bottom left), and $R_{\rm 50, bar}$ vs $\Sigma_{\rm bar}$ vs (bottom right). See Sect. \ref{['sec_method']} for references to the various datasets of Es, S0s, dEs/dSphs, and UDGs. We also considered the 'classical' bulges of Sa-Sc galaxies. The mean uncertainties of our sample galaxies are indicated in the top corner of each panel. For all these objects, the baryonic structural relations are virtually the same as the stellar ones. The grey-dashed lines show the selection criteria of UDGs from VandeK2015-UDG.
• Figure 2: Corner plots of the mass-size relations of Es and bulges, S0 disks, HSD baryonic disks, and HSD stellar disks. The panels show the 2D posterior probability distributions of pairs of fitting parameters, and the 1D marginalised probability distribution of each fitting parameter (histograms). In the 2D distributions, the contours correspond to the $1\sigma$, $2\sigma$ and $3\sigma$ regions, respectively. In the histograms, the dashed lines correspond to the median values.
• Figure 3: Same as Fig. \ref{['fig_corner_HSD']} but for non-nucleated dEs/dSphs/UDGs, nucleated dEs(N), baryonic LSD disks, and stellar LSD disks.
• Figure 4: Same as Fig. \ref{['fig1']} but adding star-forming disks from the SPARC database. As we show in Paper I, star-forming disks define two distinct sequences: one formed by star-dominated HSD galaxies (mostly Sa-Sc) and one by gas-dominated LSD galaxies (mostly Sd-dI).
• Figure 5: Same as Fig. \ref{['fig2']} but considering stellar quantities rather than baryonic ones. The main difference occurs for gas-dominated LSD disks, which move closer to the sequence defined by passive dwarfs (dEs, dSphs, UDGs).