2025MNRAS.536.1018L

Observations with the JWST have revealed a high abundance of bright galaxies at redshift, $z\gtrsim 12$, which has been widely interpreted as conflicting with the Lambda cold dark matter model. In Cowley et al., predictions were made - prior to the JWST observations - for the expected abundance of these galaxies using the Durham semi-analytic galaxy formation model, GALFORM, which is known to produce a realistic population of galaxies at lower redshifts including the present day. Key to this model is the assumption of a 'top-heavy' initial mass function of stars formed in bursts (required to explain the number counts and redshift distribution of submillimetre galaxies). Here, we compare the rest-frame ultraviolet luminosity functions derived from JWST observations with those predicted by the Cowley et al. model up to $z=14$ and make further predictions for $z=16$. We find that below $z\sim 10$, the Cowley et al. predictions agree very well with observations, while agreement at $z\gtrsim 12$ requires extending the model to take into account the time-scale for the growth of obscuring dust grains at these very early times and its dependence on gas metallicity. We trace the evolution of these galaxies from $z=14$ to $z=0$ and find that their descendants typically reside in haloes with a median mass $2.5\times 10^{13}\, h^{-1}\, \mathrm{{\rm M}_{\odot }}$. The stellar masses of the descendants range from $3.2\times 10^{6}\, h^{-1}\, \mathrm{{\rm M}_{\odot }}$ to $3.2\times 10^{11}\, h^{-1}\, \mathrm{{\rm M}_{\odot }}$. Although these galaxies were all central galaxies at $z=14$, over half of their descendants end up as satellites in massive haloes.