This thesis delves into the investigation of star-forming galaxies and quenched systems at high redshifts, exploring their evolution and properties throughout cosmic time. Firstly, I utilise the foreground lensing of massive galaxy clusters in the Hubble Frontier Fields to probe the high-redshift evolution of the main sequence of star-forming galaxies. I use the BEAGLE SED-fitting code to derive stellar masses, SFRs and redshifts from galaxies within the ASTRODEEP catalogue. I fit a fully Bayesian hierarchical model of the main sequence over redshifts 1.25 < z < 6 while explicitly modelling the outlier distribution. My results agree with an increase in normalisation of the main sequence to high redshifts that follows the redshift-dependent rate of accretion of gas onto dark matter halos. We additionally measure the slope and intrinsic scatter of the star-forming main sequence. We find that the sampling of the SED provided by the combination of filters (Hubble + ground-based Ks-band + Spitzer 3.6 and 4.5 μm) is insufficient to constrain the stellar mass and SFR over the full dynamic range of the observed main sequence, even at the lowest redshifts studied. Whilst this filter set represented (prior to the launch of JWST) the best sampling of high-redshift galaxy SEDs out to z > 3, I show that measurements of the main sequence to low masses and high redshifts were still strongly dependent on the priors employed in SED fitting (as well as other fitting assumptions).

In the first field targeted by JADES, the statistics were not large enough to extend the full main sequence analysis to JWST-based datasets. I therefore continued to study high-redshift star-forming and quenched galaxies with smaller projects, more suitable for the first deep set of spectroscopy obtained by JADES.

Dust, often one of the most poorly constrained parameters in SED fitting, drives the motivation for my second project. As part of the JADES survey, utilising data obtained with the JWST/NIRSpec Micro-Shutter Assembly, I directly explore dust attenuation in the star forming galaxy population. This is achieved by analysing Balmer decrement measurements for a sample of 51 galaxies at redshifts 4 < z < 7. Leveraging 28-hour long exposures and the efficiency of the prism disperser (but also using information from the medium-resolution gratings), I was able to probe directly the low-mass end of the galaxy population, reaching stellar masses as low as 10⁷ M$\odot$. I find that the correlation between Balmer decrement and stellar mass is already established at these high redshifts, indicating a rapid build up of dust in moderately massive galaxies at such early epochs. The lowest-mass galaxies in our sample (1 - 3 x 10⁷ M$\odot$) display a remarkably low Balmer decrement of 2.88, consistent with case B recombination and little or no dust. I further compare the Balmer decrement to continuum-derived star-formation rates, finding tentative evidence of a correlation, which likely traces the underlying connection between SFR and the mass of cold gas.

Finally, based on deep JWST/NIRSpec spectroscopy, I report the discovery of a quiescent galaxy at z = 2.34 with a stellar mass of only 9.5 x 10⁸ M$\odot$. This is the least massive quiescent galaxy found so far at high redshift. I use a Bayesian approach to model the spectrum and photometry, and find the target to have been quiescent for 0.6 Gyr with a mass-weighted average stellar age of 0.8 - 1.7 Gyr (dominated by systematics). The galaxy displays a colour gradient (redder towards the edge), consistent with outside-in, environment-driven quenching. Based on a combination of spectroscopic and robust (medium- and broad-band) photometric redshifts, I identify a galaxy overdensity near the location of the target (5-σ above the background level at this redshift). The overdensity contains three spectroscopically confirmed, massive, old galaxies. The presence of these evolved systems points to accelerated galaxy evolution in overdensities at redshifts z > 2 (in agreement with previous works). In projection, our target lies only 35 pkpc away from the most massive galaxy in this overdensity (spectroscopic redshift z = 2.349) which is located close to the overdensity’s centre. This suggests the low-mass galaxy was quenched by environment, making it the earliest evidence for environment-driven quenching to date.