Wednesday, January 22, 2020

The costs of outbreeding: a few more points

Although the house mouse (Mus musculus) is a single species, fertility is reduced in crosses between geographic populations (Wikipedia - Bolid74)

It's widely believed that mating between two species never produces fertile offspring. Untrue. Interspecific mating sometimes produces fertile offspring, and the genomes of closely related species often show gene flow from one to the other. It would be more accurate to say that such crossings are usually compromised in one way or another—incompatibilities arise during development, which lead to miscarriage, sterile offspring, or less viable offspring. 

There is a similar misunderstanding about mating within a species. It isn't true that mating within a species is uniformly possible. Even below the species level, genetic incompatibilities can arise between different geographic populations. Keep in mind the arbitrariness of all these terms: “geographic population,” “race,” “subspecies,” and “species.” These are points on a continuum of increasing reproductive isolation and genetic incompatibility. 

Various problems have been observed in crossings between geographic populations of the same species (Presgraves 2010; Turner et al. 2011). In the case of crossings between two subspecies of the house mouse (M. m. domesticus and M. m. musculus), the problems have no single cause. There is not only reduced fertility but also reduced immune function, as shown by higher loads of intestinal parasites in hybrids. Even the reduction in fertility has multiple causes: "The type and severity of fertility defects observed depends on the geographic origin of the strains and also varies among individuals within regions. This variability suggests that multiple genetic incompatibilities contribute to hybrid male sterility" (Turner et al. 2011). Multiple loci seem to be responsible, and incompatibilities that cause sterility likely act in combination with each other and with other incompatibilities (Turner et al. 2011).

In general, as genetic distance increases, so does the risk of faulty interactions between different gene loci:

The Dobzhansky-Muller model proposes that populations diverging independently in allopatry will accumulate differences through drift or selection, but will maintain coadaptation between divergent loci within each single population. If secondary contact and hybridization occurs between these populations, these divergent loci from the previously allopatric populations may interact deleteriously in hybrids leading to lowered hybrid fitness (Leppälä and Savolainen 2011)

Icelandic and Danish studies on outbreeding

The above model is consistent with two recent studies on humans, one in Iceland and the other in Denmark. The Icelandic study used a database of couples born between 1800 and 1965. In the authors' own words, "the advantage of using the Icelandic data set lies in this population being small and one of the most socioeconomically and culturally homogeneous societies in the world, with little variation in family size, use of contraceptives, and marriage practices" (Helgason et al. 2008). The authors found that fertility rose and then fell with increasing genetic distance between husband and wife. Specifically, the relationship was:

[...] an n-shaped curve from the relatively low reproductive success of couples related at the level of second cousins or closer, to the maximum for couples related at the level of third and fourth cousins, after which there is a steady decrease in reproductive success with diminishing kinship between spouses. A similar picture emerges when the number of grandchildren per couple is examined (Helgason et al. 2008)

The Icelandic study was criticized in a tweet by Jayman:

Doubt it. The Icelandic study is likely confounded in various ways so I don't think genetic relatedness is the thing directly affecting fertility.

He was retweeted by hbdchick, but neither of them identified the "various" confounds. Nor did I find this criticism in the 140 papers that cited the study on Google Scholar. So we're not in the realm of common knowledge.

The original paper did identify a possible confound. From 1800 to 1965, there was a decrease in mean fertility and an increase in marriages with more distantly related individuals. The relationship between fertility and kinship may thus be an historical confound:

These results are based on couples born during a period of almost 200 years, in the course of which there was a marked decline both in the mean fertility and in kinship between couples. (Helgason et al. 2008)

To eliminate the confound, the authors broke the data down by 25-year intervals:

Nonetheless, the same general relationship between kinship and reproductive outcome was observed within each 25-year subinterval (fig. S2). We evaluated the correlation between the standardized variables of kinship and reproductive outcome for all couples and for each time interval separately (Table 2), adjusting for the impact of geographical differences in the kinship and fertility of couples within Iceland (10). Each test revealed as significant association with kinship, with correlation coefficients of 0.063 (P = 1.5 × 10-129) for the number of children, 0.045 (P = 3.6 × 10-66) for the number of children who reproduced, and 0.042 (P = 7.6 × 10 -58) for the number of grandchildren. (Helgason et al. 2008)

There might be other confounds. That's why the Danish study is important. It went through a number of socioeconomic factors: "education, family income, urbanicity, mother's age at first birth, and six variables representing proximity to kin [maternal radius, i.e., distance between mother's and child's birthplace; paternal radius; and presence of each grandparent in the child's birth parish or neighboring parish" (Laboriau and Amorim 2008). Controlling for those factors did not change the findings.

Davenport's study on outbreeding in Jamaica

Jayman also criticized Davenport's study, apparently in reference to the finding that between 5 and 30% of the "brown" children did worse than predicted.

As for mixed race children, you have to keep in mind that people don't enter those relationships randomly - i.e., also confounded

Jayman is arguing that the parents of the brown children were atypical, apparently in the sense of being below-average. In its early days, Jamaica did have many poor whites, but most of them left because of competition from slave labor. European ancestry entered the Jamaican population largely from landowners and businessmen, particularly people of Scottish descent:

Jamaica has more people using the Campbell surnames than the population of Scotland itself, and it also has the highest percentage of Scottish surnames outside of Scotland. Scottish surnames account to about 60% of the surnames in the Jamaican phone books. The first Jamaican inhabitants from Scotland were exiled "rebels". Later, they would be followed by ambitious businessmen who spent time between their great country estates in Scotland and the island. As a result, many of the slave owning plantations on the island were owned by Scottish men, and thus a large number of mixed-race Jamaicans can claim Scottish ancestry. (Wikipedia 2020).

Today, "brown" Jamaicans are generally middle class, and they were even more so during the time of Davenport's study. It is indeed strange to intimate that his biracial subjects were the offspring of below-average whites and blacks.


Davenport's study probably corresponds to the highest degree of outbreeding possible. Even in that case, adverse mental effects were observable only in a minority of offspring. The most widespread effect is probably lower fertility and perhaps a higher risk of testicular cancer. Joffe (2009) suggests that an increase in outbreeding might explain a century-long decrease in semen quality and a corresponding increase in testicular cancer:

One implication of the proposed pathogenesis is that as D&D [duplications and deletions of genetic material during meiosis] accumulate, mating between individuals who are genetically unalike would be associated with lower reproductive success as pairing at the start of meiosis would be more likely to be impaired. On the face of it, this contradicts the earlier observation that hybrid vigour would tend to increase fertility in unrelated individuals. Yet these two ideas may be compatible: both extremes—genetic similarity (inbreeding) and genetic distance (D&D accumulation)—could decrease fertility, so that an intermediate degree of relatedness would be associated with the highest degree of fertility. This could explain the evidence from Iceland that the greatest reproductive success, measured as the number of grandchildren, was observed in couples who were third or fourth cousins (Helgason et al., 2008). Comparable findings have been reported from elsewhere, for example Denmark (Labouriau and Amorim, 2008).

I realize this may be a sensitive subject. If you feel offended, please read something else.


Davenport, C.B. and M. Steggerda. (1928). Race Crossing in Jamaica. Washington: Carnegie Institution, Publication no. 395.  

Helgason, A., S. Pálsson, D.F. Guðbjartsson, þ. Kristjánsson, K. Stefánsson. (2008). An association between the kinship and fertility of human couples. Science 319(5864): 813-816.   

Joffe, M. (2009). What has happened to human fertility? Human Reproduction 25(2): 295-307. 

Labouriau, R., and A. Amorim. (2008). Comment on "An Association Between the Kinship and Fertility of Human Couples." Science 322(5908): 1634   

Leppälä, J. and O. Savolainen. (2011). Nuclear-cytoplasmic interactions reduce male fertility in hybrids of Arabidopsis lyrata subspecies. Evolution 65(10): 2959-2972

Presgraves, D.C. (2010). The molecular evolutionary basis of species formation. Nature Reviews Genetics 11:175-180.  

Turner, L.M., D.J. Schwahn, and B. Harr. (2011). Reduced male fertility is common but highly variable in form and severity in a natural house mouse hybrid zone. Evolution 66(2): 443-458.  

Wikipedia (2020). Jamaica


Luke Lea said...

Certainly an interesting hypothesis. That extreme inbreeding can be deleterious is of course well known. That extreme outbreeding can also be deleterious (as deleterious?) is a novel idea, at least to me, an interested layman. Still the reasoning behind it seems plausible, so I am inclined to think it is likely true.

As usual, Peter Frost is a master of expository prose. Without equal in fact.

Luke Lea said...

So hybrid vigor is only true within limits, is another way of saying the same thing. True or false?

Anonymous said...

Admixture studies of IQ show no outbreeding hit. Given that IQ is extremely polygenic and for example reacts strongly to inbreeding, that makes it unlikely that there is any broad negative effect of outbreeding.

Studies on a prevalence of behavioral problems of mixed-race kids are better explained by parents with unusual behavior (marrying outside of the own group, defying social expectations ...).

Maybe the sperm quality / fertility thing is real. Though I really wonder why there would be no hit on IQ.

Bonner Tal

M said...

IQ is quite resilient. We already know (Dutton, Woodley) that mutational load increases the probability to suffer from things like autism or being asymmetric but doesn't affect IQ.

Peter Frost said...


Since the mechanisms of inbreeding depression and outbreeding depression are different, you could have outbreeding benefits and disadvantages acting at the same time. What matters is the sum total of all the negative and positive factors.


Unfortunately, admixture studies are not confined to first generation crosses. So they're irrelevant to this topic. With each passing generation, the negative effects of crossing should be weeded out, if only because the most dysfunctional crosses have reduced fertility.

In addition, the negative effects are not spread equally across the first generation. Most offspring seem normal. Davenport concluded that 5% show clear signs of dysfunctional mental performance, and his study indicated that another 25% showed some signs of dysfunctionality. We would need a larger sample to quantify the magnitude of these negative effects.


Hmmm ... no.


Ernst Mayr's textbook came out in 1963, and it was quite open about the possible adverse effects of outbreeding. Since the 1960s, discussion of outbreeding depression has been more or less limited to nonhumans. If you do a Google Scholar search for "outbreeding depression," it comes back with 8740 papers,includng 505 since 2019.

The genetic distance between Europeans and Amerindians is much less. If there are negative effects they would be smaller than in Davenport's study.

Anonymous said...

@M: Single denovo mutations increase the risk of high-functioning autism, that is autism without IQ hit. But mutational load in the sense of having many negative mutations has a big effect on IQ. Which is why people with autism usually have pretty low IQ. Greg Cochran estimates up to half of the variance in IQ being due to mutational load (In a recent interview.).

@Peter: But there are studies with first generation mixed kids, which usually land squarely between the two parental groups in terms of IQ. And 25% having some noticeably dysfunction should influence the mean.

Of course it could be the case that there is simultaneously a positive effect of outbreeding, where recessive mutations are much less likely to be homozygous, which cancels the negative effect on average. If that is the case I would expect the effect of incompatible genes to lead to big hits, while lower homozygosity leads to more gradual effects.

If lower fertility requires these "big hits", this would explain lower fertility, no average effect on IQ and more dysfunctional people.

Bonner Tal

J said...

I remember my plant genetics class, the hybrid vigor effect was largest in crossing two inbred, incestous lines. Those are inexistent among wild human populations. The lower fertility phenomenon in humans may be marginal. We are all interfertile.

Peter Frost said...


"And 25% having some noticeably dysfunction should influence the mean."

Apparently not, if we go back to Davenport's research. He found that the "Browns" were, on average, intermediate between the Whites and the Blacks. These are "tail" effects that were most obvious in a 5% core group, i.e., on all of the mental tests.


Even different species can be interfertile. What matters is the degree of interfertility, and that can vary even within a single species.

V E Lane "VEL - The Contemporary Heretic" said...

Peter Frost: “Since the mechanisms of inbreeding depression and outbreeding depression are different, you could have outbreeding benefits and disadvantages acting at the same time. What matters is the sum total of all the negative and positive factors.”

Does this mean that the optimal degree of outbreeding might be different for different traits? In other words, might the level of outbreeding that produces the highest level of fertility in couples be different to that which produces the highest, say, intelligence, athletic ability or physical attractiveness in offspring? Or is “the sum total of all the negative and positive factors” always the same for each trait?

Or does it vary on an individual basis i.e. depending on which specific genes an individual offspring inherits from each parent?

If the latter (i.e. it varies individually by offspring depending on the specific genes they get from each parent), could this explain why IQs for biracial offspring tend to be roughly intermediate between those of the two parent races? In other words, if 25% of crosses show some dysfunction, could this be balanced by another 25% performing above average. (Though this would mean that the variance in IQ would be higher among biracial offspring, something I don’t recall reading.)


)))Munchkin Person))) said...

Beethoven had a spanish grandmother, Chopin was half french, Ingmar Bergman was half wallon... the % of crossed-nationals among ''men' of genius has been higher than the normal population.

But i don't think select neonazi sociopaths can be described as ''eugenics'' at all..

Peter Frost said...

V.E. Lane,

What you say sounds plausible. Davenport found that different cognitive tests produced different results. I assume you're referring to the Minnesota Adoption Study. The biracial sample size was probably too small to detect an effect that is consistently noticeable in 5% of the subjects.

Three data points don't make an argument. And aren't you getting a bit carried away with your rhetoric?

)))Munchkin Person))) said...

Three data points don't make an argument. And aren't you getting a bit carried away with your rhetoric?''

To answer questions your IQ don't look higher...

I gave to you some examples... Among the biggest geniuses of humanity, the % of cross-national background SEEMs higher. Yes, i no have any significant evidence about it but it's possible, even because Lombroso perceived this phenomenon in XIX's.

Creative personality is well known to be internally constrasting. Highly creative individuals must have a complex personality or character and mix between different behavioral trends may help.

And the second part of my comment

No, select smart sociopaths is not a true or philosophically correct eugenics... sorry

Anonymous said...

I realize this may be a sensitive subject. If you feel offended, please read something else.

Peter, you're a racialist who sympathizes with Salter's views. Not that there's anything wrong with that, but obviously you have certain biases regarding this topic. It's not just those who disagree with you as you seem to imply.

Sean said...

Dawkins's 1982 book The Extended Phenotype, discussed the abundant evidence of outbreeding being delirious in every animal studied and he said it applies to humans. The 2012 book Science of Love and Betrayal by the eminent anthropologist Robin Dunbar discusses it in relation to humans. He says (for your genes) you are the perfect mix so they wants you to mate with someone as much like yourself as is possible without suffering inbreeding depression, and that sweet spot appears to be a third cousin.

tomR said...

1) Outbreeding is absolutely necessary part of a "genius strategy". Genius strategy as explained by Edward Dutton:

2) The cost of outbreeding can be incured at a very fundamental level, eg. compatibility between mitochondrial and nuclear genes. It was mentioned by Nick Lane in his books (eg. "Power, Sex, Suicide: Mitochondria and the Meaning of Life"). He studied the origins of multicellularity - a time when a bacterium became incorporated as mitochondria. Such bacteria could have as much 1500 genes, most of which were moved do the nucleus, while still being used for mitochondrial purposes. If the genes that are in the nucleus don't cooperate well with genes that remained in the mitochondria - an organism has a big problem with fitness.