Heritability
Heritability estimates provide information as to the extent to which phenotypic trait variation has a genetic basis and the potential for adaptive change in these traits (Visscher et al., 2008). SNPs are intrinsically linked to narrow-sense heritability given h2 is proportional to the product of the number of SNPs and their effect size (Holland et al., 2019). We found a high level of variation in survival, bleaching, and growth attributed to narrow-sense heritability (h2) when pooled at the level of population cross, which varied in juveniles infected with C. goreaui across the two temperatures tested here (see Supplementary Information for further details on S. tridacnidorum and D. trenchii ).
Interestingly, additive genetic variance was only important in influencing survival at 31°C when infected with C. goreaui but not at 27.5°C. Hence, it is likely that there is significant selection on survival at high temperatures forC. goreaui , consistent with the generally overall poorer performance of these symbionts at both temperatures (S. tridacnidorum ) or at higher temperatures in coral juveniles (C. goreaui ) (see Table 6.1 in Quigley, Baker, Coffroth, Willis, & van Oppen, 2018). The shapes of the posterior heritability distributions are informative for generating hypotheses concerning the underlying genomic architecture associated with selective breeding. Bimodal heritability estimates can be driven by single large effect eQTLs (as seen generally for traits with high heritability), whereas multiple loci of small effect size generally result in unimodal distributions (Rudra et al., 2017). The influence of cis and trans co-regulation of the traits may also influence the underlying distribution (Yang et al., 2014). Bimodal heritability (often with wide credibility intervals) estimates can also be driven by high underlying within-population variability even with large sample sizes (mirrored in the high phenotypic variability between genotypes, Quigley et al., 2020), and have been recorded across a range of disease-related expression of genes (Yang et al., 2014). High heritability at ambient temperatures may also be common (Kronenberg et al., 2019), and potentially indicative of core traits under continual selective pressures. Finally, it is likely that not all the SNPs associated with these three phenotypes have been captured although narrow-sense heritability should incorporate additively all the common SNPs associated (Holland et al., 2019) and it will be important to use these methods under a variety of experimental settings to fully develop models of adaptive selection.