Uchiyama et al. propose a thought-provoking synthesis between the fields of behavioral genetics and cultural evolution. We suggest that the principles laid out by Uchiyama et al. can be extended to the levels of the family, the lineage, and even the individual. These extensions expose further challenges in assessment and interpretation of heritability even when cultural clusters are well defined.

Uchiyama et al. point out, in their third and fourth key points, that cultural effects on phenotypes can be directional and predictable, and can become stronger as ecological and cultural selection becomes stronger. These points are well demonstrated, respectively, by the case of sunscreen use and the expression of skin cancer, and by the increased use of sunscreen by Australians of European ancestry. They further highlight that cultural clustering, and particularly hidden clustering, may hinder proper interpretation of heritability scores. We propose that further challenges stem from the same fundamental principle that Uchiyama et al. propose when considered at the level of the lineage, family, or individual.

Even within a tight mono-cultural society, cultural vertical transmission is strong for many traits (Chen, Cavalli-Sforza, & Feldman, Reference Chen, Cavalli-Sforza and Feldman1982), and different families may vary in their choice of available cultural tools or knowledge. This between-family variation may influence the expression of related phenotypes in a consistent and directional manner, masking or unmasking genetic effects (Feldman et al., Reference Feldman, Christiansen and Otto2013). For example, a lineage with a history of skin cancer may use sunscreen and other protection more than the average family, thus reducing the risk of cancer via this cultural practice. That is, when genetic variation has an observable phenotypic impact, families may differentially implement cultural practices that mask or unmask this genetic variation as a function of the genetic variants that each family carries. This effect can be exacerbated by the emergence of personal genomics services and precision medicine. Because familial structure is correlated with genetics, the challenges extend beyond the consideration of hidden cultural clusters (as discussed by Uchiyama et al.). In particular, differential practice of cultural behaviors that are designed to alter genetically influenced phenotypes as a function of those phenotypes' expression is expected to render the effects of genetics and culture inseparable. Teasing apart the influence of these factors then becomes impossible without manipulation experiments, which are unethical and unfeasible in humans.

Additional challenges arise when considering phenotypes expressed gradually or repeatedly during the lifetime of an individual. In such cases, if cultural practices designed to enhance or inhibit the phenotype are available, different individuals are likely to use them differentially and in a manner that is highly directional, possibly masking or unmasking genetic effects and deflating or inflating heritability scores. For example, if a certain individual exhibits symptoms that place her at high risk for autoimmune disorders if excessively exposed to UV, she may remain indoors or use protective measures above and beyond the average use in her cultural cluster, thereby reducing both heritability scores and phenotypic variation. In contrast, gifted young athletes may invest in their training well above the average and thus perform better than would be expected based on their genetic makeup. Such individual-based effects can accordingly increase or decrease heritability scores.

Finally, we would like to point out that the masking or unmasking of genetic effects on the phenotype by cultural practices may usefully be viewed as a form of differential niche construction (Odling-Smee, Laland, & Feldman, Reference Odling-Smee, Laland and Feldman2013). In this view, different cultural clusters may alter the selective pressures that they experience as a function of the perceived severity of these pressures (Ihara & Feldman, Reference Ihara and Feldman2004). The same extension that we presented above holds here: Differential application of cultural practices at the family or individual level, when matched with the selection differential they would have experienced otherwise, can offset the selection differential and diminish its effect.