We utilize GIZMO , coupled with newly developed sub-grid models for Population III ( Pop III ) and Population II ( Pop II ) , to study the legacy of star formation in the pre-reionization Universe .
We find that the Pop II star formation rate density ( SFRD ) , produced in our simulation ( \sim 10 ^ { -2 } M _ { \odot } { yr ^ { -1 } Mpc ^ { -3 } } at z \simeq 10 ) , matches the total SFRD inferred from observations within a factor of < 2 at 7 \lesssim z \lesssim 10 .
The Pop III SFRD , however , reaches a plateau at \sim 10 ^ { -3 } M _ { \odot } { yr ^ { -1 } Mpc ^ { -3 } } by z \approx 10 , remaining largely unaffected by the presence of Pop II feedback .
At z =7.5 , \sim 20 \% of Pop III star formation occurs in isolated haloes which have never experienced any Pop II star formation ( i.e .
primordial haloes ) .
We predict that Pop III-only galaxies exist at magnitudes M _ { UV } \gtrsim - 11 , beyond the limits for direct detection with the James Webb Space Telescope ( JWST ) .
We assess that our stellar mass function ( SMF ) and UV luminosity function ( UVLF ) agree well with the observed low mass/faint-end behavior at z = 8 and 10 .
However , beyond the current limiting magnitudes , we find that both our SMF and UVLF demonstrate a deviation/turnover from the expected power-law slope ( M _ { UV,turn } = -13.4 \pm 1.1 at z =10 ) .
This could impact observational estimates of the true SFRD by a factor of 2 ( 10 ) when integrating to M _ { UV } = - 12 ( - 8 ) at z \sim 10 , depending on integration limits .
Our turnover correlates well with the transition from dark matter haloes dominated by molecular cooling to those dominated by atomic cooling , for a mass M _ { halo } \approx 10 ^ { 8 } M _ { \odot } at z \simeq 10 .