Can humans distinguish subtle genetic differences, even within the immediate family? Apparently, grandmothers can do so, as explained in Molly Fox, et al., Grandma plays favourites: X-chromosome relatedness and sex-specific childhood mortality, published in Proceedings of the Royal Society B, in October of last year. The story is an excellent example of how evolutionary thinking sheds light on human behavior. As J. Philippe Rushton has theorized, the genes really do matter. People who share a greater percentage of genes favor each other compared to people with less genetic overlap.
The “grandmother hypothesis” has been invoked to explain why women live several decades past the end of their reproductive years. Older, post-menopausal women can enhance their biological fitness by contributing to the survivability of their grandchildren, since there is significant genetic relatedness between grandparent-grandchildren. Thus, the grandmother hypothesis suggests that a long post-reproductive lifespan for human women evolved because of the adaptive value of assisting in the care of their children’s children.
However, due to the specific mechanisms of the inheritance of the sex chromosomes, the genetic relatedness of grandmother to grandchild will differ depending on whether the grandchild is a boy or a girl, and whether the grandmother is the paternal grandmother (father’s mother) or maternal grandmother (mother’s mother). Let me explain.
Girls have two X chromosomes, one from the mother and one from the father, while boys have one X chromosome (from the mother) and one Y chromosome (from the father). These X and Y chromosomes are, in turn, inherited from the grandparents.
Let us consider a maternal grandmother. She has two X chromosomes. One of these two X’s will be passed on to her daughter (the mother); thus the probability that any specific X chromosome gene will be passed from maternal grandmother to mother is 0.5. (See Figure.)
The figure shows that a paternal grandmother (PGM) is more closely related to her granddaughter because the granddaughter carries a more or less exact replica of one of her X Chromosomes (the blue X chromosome in the figure) while her grandson carries neither or her X Chromosomes. Theoretically, she should prefer her granddaughter to her grandson. The maternal grandmother (MGM) is equally related to grandson and granddaughter, so she should not show favoritism. The chromosomes with red and black parts result from recombination (crossover) during meiosis.
The mother has two X chromosomes herself, so the probability of the mother passing any specific X chromosome gene to either a boy or a girl child is likewise 0.5. Therefore, the probability of any specific X chromosome gene being passed on from a maternal grandmother to a grandchild is 0.25 (0.5 x 0.5). From the standpoint of the maternal grandmother, there is no difference between grandson and granddaughter in X-chromosome relatedness and, hence, approximate overall genetic relatedness.
The situation is markedly different for a paternal grandmother, the mother of the grandchildren’s father. Again, she has two X chromosomes, one which can be passed on to her son (the father); thus, there is a 0.5 chance of any X-chromosome gene being passed from paternal grandmother to father. However, the father has only that one X chromosome to be passed on to his own children. If he has a girl (granddaughter), there is a probability of 100% (1.0) that his X chromosome is passed on to his daughter; therefore, one of the two X chromosomes in that girl must be derived from the paternal grandmother.
What about the grandson of a paternal grandmother? His only X chromosome comes from his mother, not his father. Therefore, the paternal grandmother does not contribute an X chromosome to a grandson, implying an X-chromosome relatedness of 0.
Therefore (and this is the important point), a paternal grandmother, all else being equal, is genetically less related to a grandson than to a granddaughter, and less related to a grandson than is a maternal grandmother. Conversely, a paternal grandmother likely is more genetically related to a granddaughter than is a maternal grandmother, given the certainty that the granddaughter possesses an X chromosome from the paternal grandmother.
Further, genetic recombination between the X and Y chromosomes in the formation of sperm is very limited, meaning that a granddaughter inherits from her paternal grandmother a more or less “intact” X chromosome. On the other hand, the X chromosome inherited from the maternal grandmother will contain some amount of genes from the maternal grandfather, due to the extensive recombination between X chromosomes that occurs during the formation of the human egg.
The bottom line is that the highest degree of genetic relatedness is most likely to be between paternal grandmother and granddaughter, the least between paternal grandmother and grandson, with maternal grandmother-grandchildren exhibiting more intermediate levels of relatedness. (See Figure above.)
The authors of this study hypothesized that grandmothers’ investment in their grandchildren would mirror the relative genetic relatedness. They analyzed the effect of the presence of paternal or maternal grandmothers, in the grandchildren’s household or in the grandchildren’s village, to childhood survivability/mortality. The data were obtained from seven populations: Japan, Germany, England, Ethiopia, The Gambia, Malawi, and Canada dating in some cases as far back as the 17th century (when available) and extending to the 21st century.
The findings are remarkable. The presence of the paternal grandmother was detrimental to the survivorship of grandsons. In six out of the seven populations (Gambia being the exception) the presence of the paternal grandmother was better for girls than for boys. In all populations, boys survived better with the maternal, as opposed to the paternal, grandmother; in four out of the seven populations, girls survived better with the paternal, as opposed to the maternal, grandmother. The findings essentially support the authors’ fundamental hypothesis.
What can be the mechanism behind this phenomenon? Here the authors speculate that a “behavioral adaption” may cause women to favor grandchildren of a specific sex; alternatively, the grandchildren may be sending out “signals” that represent relative genetic relatedness (e.g., physical resemblance, pheromones, etc.).
The authors unfortunately tend to dismiss a behavioral mechanism based upon negative results of “questionnaires” that did not observe a relationship between preferential care and genetic relatedness. However, in my opinion, these sorts of social science analyses are very weak and highly dependent upon the types of questions asked and the honesty of those answering the questions.
The authors instead propose a purely genetic explanation in that genes related to phenotypes influencing fitness are likely found on the X chromosome and thus variation in the inheritance of these genes between grandmother and grandchild can influence survivability. For example, if the paternal grandmother carries particularly helpful genes on her X chromosome, these would be present in the granddaughter, enhancing survivability, and absent in the grandson, depressing survivability. However, this explanation, and the related one involving “epigenetic modifications” of X chromosome genes inherited from the paternal grandmother to granddaughter, do not explain why the presence of one or the other grandmother should have such a marked effect on childhood outcomes and why the presence of the paternal grandmother actually seems detrimental to grandson survivability.
I believe that the social science “questionnaire data” do not eliminate the possibility of a behavioral mechanism predicated on the ability of the grandmothers to somehow become aware of the differing levels of genetic relatedness of the grandchildren. I do not suggest any sort of conscious activity — “that grandson of mine is less related to me than his sister, so I’ll ignore him and help her.”
On the contrary, similar to J.P Rushton’s “genetic similarity theory” (GST) of which this phenomenon is likely a variant, the grandmothers’ perception of genetic relatedness is probably more on the unconscious level; there is a greater attraction to and affinity for the more genetically related paternal granddaughter, resulting in a greater investment. Likewise, there may be a subconscious aversion towards a paternal grandson which results in a more careless attitude toward care and provisioning. Averaged over large numbers of people, these subtle differences in behavior would result in variant levels of childhood survivorship vs. mortality. How exactly the genetic differences are being unconsciously detected is as yet unknown.
Related to this, and also likely associated with GST, is a fascinating analysis from the laboratory of Neil Risch, who is one of the few population geneticists honest about the biological reality of race (Risch et al., Ancestry-related assortative mating in Latino populations, Genome Biology, doi: 10.1186/gb-2009-10-11-r132). Risch and colleagues set out to determine whether there was any correlation between the ancestral proportions of spouses in either Mexican or Puerto Rican populations. Assortative mating (like with like) is a proven phenomenon, based on a number of characteristics, including physical appearance, behavior, and other biological attributes. Hispanic populations are well suited for analyzing whether genetic ancestry influences mate choice, since “Latino” groups are heavily admixed and exhibit a wide range of ancestries within single ethnic-national-cultural groups. Both Mexicans and Puerto Ricans are significantly triracial (European-sub-Saharan African-Amerindian), with Mexicans characterized by greater Amerindian and minor African admixture, and Puerto Ricans by significant African and somewhat lower, but still significant, Amerindian ancestry. Socioeconomic status (SES) and geographical origin (Mexico/Puerto Rico or the USA) of the study participants was controlled for, strengthening the findings.
Risch et al. discovered was that among Mexicans there is a strong correlation between spousal European and Native American, but not African, ancestry; in other words, Mexican husbands and wives are well matched in percent European and Native American ancestry, significantly beyond random chance. African ancestry does not correlate significantly in Mexican spouses. For Puerto Ricans, the spousal correlations were significant for European and African, but not Amerindian, ancestry.
It seems that Mexicans are somehow gauging the Amerindian admixture in their potential mates, while Puerto Ricans concentrate on matching levels of African ancestry.
Analysis of the distribution of gene alleles within and between individuals strongly suggests that this “like with like” mating has been going on for quite some time, and possibly was more pronounced in previous generations than today. Given that SES and geographical origin do not seem to play any roles in markers of ancestry, what is the mechanism by which this assortative mating takes place?
Obviously, physical features can serve as a proxy for ancestry; therefore, perhaps, Latinos are hooking up with each other based upon who “looks Whiter” or “looks more Amerindian or African.” However, the authors note that to achieve the fairly high correlations of husband-wife ancestry, a very marked association between physical traits and genetic ancestral proportions is required. The problem is that previous studies of Latin American groups have not shown genotype-phenotype associations of the required levels. Therefore, it is unlikely that physical appearance alone can suffice as an explanation.
Perhaps a number of different factors, including appearance, acting in concert can have a synergistic effect on ancestral identification for assortative mating. One must remember that a number of “Latino” cultures have historically prized higher degrees of White ancestry, and perhaps older family members “advertise” their “whiteness” so as to attract the “right sort” of mates for the younger generation.
Both the grandmother and Latino assortative mating stories are of relevance to us because they describe real-life examples of GST-like behavior. They show the ability of individuals to, at some level, recognize relative differences in genetic similarity.
This obviously has bearing on the recognition of ethnic kin in multiracial, multicultural societies, and the propensity to favor such kin, either consciously or unconsciously. One may speculate that if people can perceive relative genetic ancestry via some sets of cues, and if this influences behavior, then this may explain some of the costs of diversity, including lower levels of societal investment.
Indeed, given that these studies suggest the ability to distinguish genetic differences even within immediate families, it is plausible that differences in ethnic kinship within single ethnic groups can determine societal harmony and progress. For example, one would expect that ethnic-national groups that are relatively genetically homogenous would exhibit higher degrees of investment in collective goods, such as public health insurance. On the other hand, ethnic-national groups that are more genetically heterogeneous, for whatever reason, would exhibit less societal investment and more focus on the family unit and extended kin relationships.
The possibility that relative genetic homogeneity could be part of the explanation for differences in social structure and public goods investment in different European nations is a provocative hypothesis. It is also possible that different ethnic and racial groups exhibit these sorts of behaviors to different degrees; more ethnocentric peoples may be better at detecting subtle differences in genetic kinship than those who are more tolerant.
And what if the mechanism behind the Fox and Risch findings is discovered? Will those opposed to our people’s survival attempt to find ways of short-circuiting this mechanism, perhaps even to misdirect it so as to make Westerners favor aliens over kin to an even greater extent than they already do? Conversely, can we use these discoveries to our advantage? Only time will tell.
Ted Sallis (email him) writes on scientific issues.