Title: Nature and Nurture? Comparing Ly$α$ Detections in UV Bright and Fainter [O III]+H$β$ Emitters at $z\sim8$ With Keck/MOSFIRE

Abstract: The 100% detection rate of Ly$\alpha$ in a sample of four luminous $z\sim8$ galaxies with red Spitzer/IRAC colors suggested objects with unusual ionizing capabilities and the presence of early (re)ionized bubbles in a neutral era. Whether such bubbles are a reflection of enhanced ionizing properties (nature) or an overdense environment (nurture), however, remains an open question. Here we aim to distinguish between these hypotheses with a spectroscopic search for Ly$\alpha$ in five fainter galaxies from the CANDELS-GOODS fields with Keck/MOSFIRE, selected on a similar IRAC excess and UV magnitudes that reflect reduced clustering effects. We detect $>4\sigma$ line emission in only two targets, placing them at redshifts of $z_{\rm Ly\alpha}=7.1081$ and $z_{\rm Ly\alpha}=7.9622$, with rest-frame EWs of 16-17 A that are $\sim1.5\times$ weaker than those of their bright counterparts. Comparing line detections we find a reduced rate of Ly$\alpha$ emission of $0.40^{+0.30}_{-0.25}$ in the faint sample compared to $1.00^{+0.00}_{-0.44}$ for the bright sample. The lower rate of the faint sample is well-matched by the predicted number of detections derived from simulations of a mostly neutral IGM and an intrinsic EW$_{0,Ly\alpha}$ distribution from $z\sim6$ galaxies. However, even with an extreme EW model, it is challenging to match the detection rate of the luminous objects. SED-fitting of the faint galaxies indicates generally young, metal- and dust-poor systems, albeit with reduced star-forming properties compared to their luminous counterparts. We argue the enhanced detection rate in UV bright galaxies is likely a byproduct of both extreme ionizing properties as well as environment, where nature and nurture both play an important role. Ultimately, UV emission lines and direct measurements of clustering with JWST are required to resolve the physical nature of this puzzling population.