Forest mysteries: using genetics to identify microbial patterns associated with hummingbirds and bats


Post author Jeremy Hsu is a graduate student in the Hadly lab.

Many animals that visit flowers are known to carry microfungal communities; these fungi are important ecologically because they have the potential to alter the attractiveness of nectar to these pollinators, thus influencing plant-animal interactions. The pollinators also act as vectors for the microfungi, helping distribute communities throughout the local ecosystem. In a recent paper in Fungal Ecology, Melinda Belisle, a then-graduate student in the Department of Biology in CEHG faculty member Tad Fukami’s lab, led a team of researchers at Stanford and in Costa Rica to investigate these relationships further and examine how the distribution of microfungi varies in space and time in a local ecosystem.

Hummingbirds, bats, and beyond: the Costa Rican landscape

To investigate this question, Melinda and colleagues focused on two specific pollinators: the hummingbird and bat. Both of these flying animals are known to carry microfungal communities in their mouths, and by dipping the animals’ bills or beaks in a sugar solution, the researchers were able to obtain samples of the microfungi communities that live there. The researchers caught hummingbirds and bats in a specific countryside landscape in southwestern Costa Rica over a span of two years. The location of the research, Coto Brus, was chosen for its diversity of different habitats, ranging from true forest areas to deforested areas now dominated by coffee plantations. The wide diversity of habitat types within this landscape allowed the team to test two of their predictions: first, that there would be differences in which microfungal species are present in various parts of the landscape, and second, that the presence and distribution of these microfungal communities would change over time.

Figure 1. Melinda and colleagues fed hummingbirds sugar water and then were able to test the sugar water solution for the presence of microfungal communities.

Figure 1. Melinda and colleagues fed hummingbirds sugar water and then were able to test the sugar water solution for the presence of microfungal communities.

Genetic identification of microbial communities

After catching the hummingbirds and bats, dipping their mouths in sugar water, and releasing the animals back into the wild unharmed, the researchers then spread the sugar water on a yeast-malt agar plate. This allowed any microfungal communities that were present on the beaks or bills of the hummingbirds and bats (and thus present in the sugar water) to grow into colonies. Following this growth, Melinda and the other researchers extracted DNA from these colonies and amplified and sequenced the large subunit nuclear ribosomal RNA gene. By examining the resulting unique DNA sequences, they could then identify specific species of microfungi present on each of the hummingbird and bat bills and beaks.

Microfungal communities vary in composition across time, but do not correlate with landscape

The researchers analyzed the composition of microfungal species across multiple sampling locations in the landscape over a period of two years. They discovered that microfungal community presence and composition differed substantially between the different sampling periods, including on both a micro- and macro- temporal scale (i.e., short and long amounts of time), confirming one of their hypotheses. However, they did not find any pattern or correlation between microfungal community composition and abundance with the landscape type, and no other spatial correlations were found, contrary to the researchers’ expectations. Thus, their results indicate that microfungal communities change significantly over time but are not directly influenced by the habitat type of hummingbirds and bats.

The authors were surprised that there was no spatial distribution pattern of microfungal communities and that the composition and abundance of these communities did not correlate with landscape type. These findings indicate that there may not be significant limitations on dispersal for microfungi, which would otherwise have resulted in geographic and spatial patterning as well as correlations between microfungal community composition and habitat type. However, the authors did find that this composition varied across both small and large amounts of time, and they speculate that this change could be triggered by differing environmental characteristics, such as variations in food supply and availability for the hummingbirds and bats in different seasons. Variation in diet for the animals could thus lead to differences in which microfungi are able to survive and proliferate in their mouths. To conclude, the researchers propose future work to determine the speed at which microfungal communities change over time and to determine whether these patterns are found in other types of pollinators, such as insects.


Belisle M., Mendenhall C. D., Brenes F. O., Fukami T. Fungal Ecology (2014). DOI: 10.1016/j.funeco.2014.02.007

Paper author Melinda Belisle was a graduate student in Tadeshi Fukami's lab. She is now a Senior Scientist at Exponent, Inc.

Paper author Melinda Belisle was a graduate student in the Fukami lab. She is now a Senior Scientist at Exponent, Inc.

Biochemical constraints on genes involved in early embryonic development.

Sandeep Venkataram is a graduate student in the Petrov lab.

Post author Sandeep Venkataram is a graduate student in the Petrov lab.

Chemical reactions form the foundation of life, yet such elementary activities are rarely considered when trying to understand higher-level processes, such as embryonic development. Nevertheless, as recently shown by Artieri and Fraser (MBE 2014), limitations on the kinetics of gene expression strongly constrain the length of highly expressed transcripts during early embryonic development of fruit flies. Furthermore, this phenomenon appears to be a general feature of fruit fly development as it is evolutionarily conserved across a number of species.

The long and short of mRNA transcription

It has long been known that only a portion of the mRNA molecules are used to produce functional proteins – multicellular species contain many long ‘introns’, which must first be transcribed, then spliced out before translation can occur. Introns can be very long, causing transcription of some mRNA molecules to take significant amounts of time: for example, one 2.3 million bp transcript in humans takes over half a day to be produced. This creates a problem as incompletely transcribed mRNA molecules are degraded when DNA is replicated at the beginning of cell division, and the process must begin anew once division is completed. Together, this implies that cell divisions need to be spaced out long enough apart from each other to produce all of the transcripts necessary for the growth of the cell before the next division occurs.

Studies of fruit fly development have shown that zygotes undergo “syncytial division” at the beginning of development, where the DNA within the zygotic nuclei divide every ~10 minutes for 9 cycles, followed by 4 additional progressively lengthening divisions. While most mRNA in the cell at this time are supplied by the mother (maternal mRNA), this also represents the phase during which the zygote begins producing its own mRNA. The extremely rapid cell divisions led Artieri and Fraser to hypothesize that long mRNA molecules transcribed from the zygotic genome may be underrepresented during these early stages of development. Maternal mRNAs, on the other hand, would be unaffected as they are already present in the cell and do not have to be transcribed.

Transcript length vs. developmental timing

The authors classified embryonically expressed genes as “maternal” or “zygotic” depending on whether or not the gene was present as maternal mRNA in unfertilized embryos using published data. They then obtained multiple developmental mRNA expression timecourses and found that long zygotically expressed genes took longer to reach maximal expression levels than short genes – consistent with their inability to be fully transcribed during early development (Figure 1). Furthermore, they were able to use total RNA expression data to detect the presence of incomplete transcripts, indicating that delay was not due to later transcriptional activation, but rather the incomplete production of transcripts.

Modified from Artieri and Fraser 2014 Figure 2B . Long zygotic genes are underexpressed early in the syncytial division phase relative to short genes, but catch up in expression by the end of the syncytial phase while maternally derived transcripts show no such changes.

Figure 1. Long zygotic genes are underexpressed early in the syncytial division phase relative to short genes, but catch up in expression by the end of the syncytial phase while maternally derived transcripts show no such changes. [Modified from Artieri and Fraser 2014 Figure 2B . ]

Using a published set of developmental mRNA expression timecourses from additional Drosophila species, Artieri and Fraser show that these patterns are consistent across all species examined. Finally, they also observed that the introns present in highly expressed zygotic genes appear to be highly evolutionarily constrained in terms of their lengths when compared to either genes maternally deposited or zygotically expressed during later timepoints. This suggests that natural selection has played a role in limiting the expansion of introns in early expressed zygotic genes, allowing them to escape ‘intron delay’.


In summary, Artieri and Fraser have found evidence that a significant fraction of zygotically expressed transcripts in fruit flies are delayed from reaching their maximal levels of expression due to the rapid cell cycles taking place at the beginning of development. This suggests a simple mechanism for developmental timing of zygotic gene expression: genes that are required early must be short, while genes whose expression is needed at a later time can delay their expression via the presence of long introns. While some evidence for the use of intron length as a regulatory mechanism has recently emerged (Takashima et al. 2011), future experiments will be required to determine how widespread is the effect of selection to maintain long lengths and delayed expression.


Carlo G. Artieri and Hunter B. Fraser Transcript Length Mediates Developmental Timing of Gene Expression Across Drosophila. (2014) Molecular Biology and Evolution doi:10.1093/molbev/msu226

Takashima Y, Ohtsuka T, González A, Miyachi H, Kageyama R. Intronic delay is essential for oscillatory expression in the segmentation clock. Proc Natl Acad Sci U S A. 2011;108:3300-3305.

Paper author Carlo Artieri is a postdoctoral fellow in the Fraser lab.

Paper author Carlo Artieri is a postdoctoral fellow in the Fraser lab.

Genetically blond: Mice shed light on the molecular basis of blond hair in Europeans.

Post author Alicia Martin is a graduate student in the Bustamante Lab.

Post author Alicia Martin is a graduate student in the Bustamante lab.

Global pigmentation variability in the hair, eyes, and skin is among the most striking phenotypic human traits. Differences in genomic regions associated with these traits show some of the strongest signals of selection in the human genome, indicating the importance of pigmentation throughout human evolution. Hair and eye color are especially variable across Europe, and several previous studies have queried the genome to determine where the mutations causing blond hair are located. In a recent paper by HHMI research specialist Catherine Guenther (David Kingsley’s lab) and colleagues, the team explored why blond hair occurs when an allele strongly associated with blond hair in Europeans is found (1).

The impact of mutations near the gene KITLG on skin pigmentation has been debated in previous studies in humans and sticklebacks (2, 3), but one regulatory region has an undeniably strong association with European blond hair (4). The allele most strongly associated with blond hair is 350 kb upstream of KITLG in a highly conserved region. As a starting point to characterize the molecular basis of the blond mutation, the authors looked at fur pigmentation in a Kitl mutant mouse strain (Steel panda or Slpan) with a 65 Mb inversion that includes sequences orthologous to the putatively causal blond hair allele. Compared to the background strain, heterozygotes and homozygotes had increasingly lighter hair. They also had reduced Kitl expression.

Lightening the candidate load

The authors next wanted to narrow down the region that causally regulates blond hair pigmentation in Kitl. The authors narrowed down the potential blond target to a 17 kb window bounding the human association signal (see Figure 1 below).


Figure 1: The human blond-associated region contains a functional hair follicle enhancer.

The most strongly associated SNP in humans was in the center of the window and in the only highly conserved region. They cloned 3 reporter constructs, H1, H2, and H3, into mice. Each clone contained a candidate regulatory region subsetting the 17 kb window driving lacZ expression. Only one of the clones, H2, drove consistent expression of lacZ. The authors repeated their cloning process by further subdividing the region spanned by H2 into two candidate regulatory peaks, HFE and H2b, both spanning conservation peaks. The HFE peak, which contains the most strongly associated blond hair SNP, drove consistent expression in hair follicles and epithelial cells of developing hair and skin, while the H2b peak drove consistent expression in kidneys. Their results indicate that the H2 region is an important gene enhancer, that the HFE region drives expression in the hair follicles, while the H2b region drives expression in kidneys.

The authors next made clones with the H2 region containing the exact ancestral (H2ANC) and derived (H2DER) mutations putatively causing blond hair in Europeans. The H2ANC and H2DER embryos qualitatively looked similar, and further quantification of lacZ expression in keratinocyte cell lines using the smaller 1.9 kb HFE clones showed significantly less expression in HFE-DER compared to HFE-ANC.

Regulating blondness

When the authors wanted to know why the blond allele is regulating expression in hair follicles, they turned to ENCODE. Previous ChIP-seq studies showed that the transcription factor family TCF/LEF strongly bound the region of interest in a colorectal epithelial cell line. The authors discovered a well-conserved LEF binding motif disrupted by the blond allele. LEF1 is a transcription factor expressed during hair follicle development, and Lef1 knockout mice are light-furred, providing a potential mechanism for upstream regulation of Kitl. Consistent with previous LEF1 binding site analyses, the authors showed reduced LEF1 responsiveness from the derived blond allele experimentally.

Finally, the authors integrated a single copy of the hair follicle enhancer into the same genomic location including either the blond allele (BLD-Kitl) or the ancestral allele (ANC-Kitl) into mice. The extra Kitl enhancer darkened both mice, but the mice with the BLD-Kitl insertion had only 79% of Kitl expression as the ANC-Kitl mice, and BLD-Kitl mice were noticeably lighter.


Figure 2: Mouse lines differing at a single base-pair position in the KITLG hair enhancer (HE) show obvious differences in hair color.

The molecular explanation behind the European blond hair allele identified previously in a GWAS is intrinsically very biologically interesting. Many pigmentation associations are in non-coding regions far from canonical pigmentation genes. Strong signals of positive selection have been identified upstream of KITLG as well as near many other pigmentation genes. Some other pigmentation associations also come from highly conserved regions, and this work has provided a framework for dissecting the regulatory function of pigmentation variants.

The implications of this study also extend beyond pigmentation. The time and cost to disentangle the molecular basis of a single GWAS likely means that most GWAS variants or regions will likely not be biologically characterized as carefully any time soon. The driver of blond hair was in a distal regulatory region 350 kb from the gene in an important transcription factor binding site where the consensus sequence does not perfectly match. Additionally, the expression change of the Kitl gene resulting from the causal mutation was by only a fraction of the ancestral case. Diseases without such an obvious phenotypic readout will likely be harder to dissect.

On the other hand, the abundance of publicly available data enabled much of the work presented here. Multiple studies have implicated the GWAS SNP in pigmentation phenotypes, increasing its likelihood of causality. Additionally, the authors were able to identify a potential upstream mechanism of Kitl regulation via the GWAS region with ENCODE ChIP-seq data alone. As larger genomic, interaction, regulatory datasets, etc. become available, our ability to explain phenotypic variation identified in humans with less costly and time-consuming in vivo models increases.


1. C. A. Guenther, B. Tasic, L. Luo, M. A. Bedell, D. M. Kingsley, A molecular basis for classic blond hair color in Europeans., Nat. Genet. 46, 748–52 (2014).
2. S. Beleza et al., R. A. Spritz, Ed. Genetic Architecture of Skin and Eye Color in an African-European Admixed Population, PLoS Genet. 9, e1003372 (2013).
3. C. T. Miller et al., cis-Regulatory changes in Kit ligand expression and parallel evolution of pigmentation in sticklebacks and humans., Cell 131, 1179–89 (2007).
4. P. Sulem et al., Genetic determinants of hair , eye and skin pigmentation in Europeans, 39, 1443–1452 (2007).

Paper author Kate Guenther is a Research Specialist in the Kingsley lab.

Paper author Kate Guenther is a Research Specialist in the Kingsley lab.

Echoes of the Past: Hereditarianism and A Troublesome Inheritance

troublesomeMany researchers have expressed concerns about misrepresentations of human population genetics in a recent popular book by journalist Nicholas Wade: A Troublesome Inheritance (Penguin Press, NY, 2014). A letter signed by 143 scientists, including seven from Stanford, criticized the book in the New York Times Book Review on August 8, 2014. In this post, Prof. Marcus Feldman situates A Troublesome Inheritance in a problematic intellectual tradition, highlighting a number of the book’s major problems.

In 1969, Arthur Jensen1 ignited a decades-long debate when he wrote that it is a “not unreasonable hypothesis that genetic factors are strongly implicated in the average Negro-white intelligence difference.” From this he inferred that educational interventions in communities whose members have lower measured IQ could not succeed.

The errors in Jensen’s choice of data (from Burt2) and statistical methods used to compute a heritability of about 80% for measured IQ were pointed out by numerous geneticists and statisticians. Twenty-five years after Jensen’s incendiary paper, Herrnstein and Murray’s book The Bell Curve3 drew inferences similar to Jensen’s that differences among races and social classes in IQ were genetically based. The Bell Curve elicited a flood of strong criticisms of the data used, the statistical analyses, and the policy inferences.4 Much of the criticism of Jensen and Herrnstein and Murray centers on their interpretation of heritability of IQ. In 1975, Richard Lewontin and I5 stressed the failure of the heritability statistic to do what these authors claim, namely, to show that IQ is largely genetically determined and hence that traits related to IQ, such as educational or economic success, would be impervious to environmental intervention.

As pointed out by Nicholas Wade in the first half of A Troublesome Inheritance6, we are now in a genomic age, where individual differences at the level of DNA can be detected. These chapters present a hodgepodge of historical ideas about race, aggression, and genetics. We are given an inkling of what will come in the last half of the book on page 57: “important aspects of human social behavior are shaped by the genes” and “these behavior traits are likely to vary from one race to another, sometimes significantly so.”

Whereas inferences on the causes of human behavioral variation referred to above were based on correlations between relatives, Wade develops his arguments for the genetic basis of social behaviors in the second half of A Troublesome Inheritance from results on worldwide variation in DNA polymorphisms, namely microsatellite polymorphisms (“The Rosenberg-Feldman studies”7,8) and single nucleotide polymorphisms (another Stanford study; Li et al. 20089), from the Human Genome Diversity Panel.10 Here, as in his previous journalism about these studies, Wade exhibits a complete lack of understanding of their implications. For example, he does not mention the finding, stressed in both studies, that only 5–10% of the worldwide genomic variation is between continental groups, while the vast majority is between individuals within populations.

Using data from 15 protein genes, R. C. Lewontin in 197211 was the first to point out that the overwhelming majority of human genotypic variation is within populations, and that continental “races” differed little genetically. Twenty-five years later Barbujani et al.12 came to the same conclusion from their study of 109 DNA markers. Wade criticizes Lewontin’s conclusion that “racial classification is now seen to be of virtually no genetic or taxonomic significance” as representing Lewontin’s “political stake in the issue.”13

From the data and analyses of worldwide molecular genomic variation7,8,9, Richard Lewontin and I amplified the conclusions of Lewontin and Barbujani et al. as follows14: “The repeated and consistent results on the apportionment of genetic diversity … show that the genes underlying the phenotypic differences used to assign race categories are atypical of the genome in general and are not a reliable index to the amount of genetic differentiation between groups. Thus racial assignment loses any general biological interest. For the human species, race assignment of individuals does not carry with it any general implication about genetic differentiation.”

Even though the between-continent fraction of genetic variation is small, as the reader discovers on leaving the first half of A Troublesome Inheritance, Wade’s erroneous interpretation of its significance for racial differences becomes the basis for his entry into the “speculative arena at the interface of history, economics, and human evolution.”15 In the second half of the book, he claims that differences among continents in economic development, social institutions, and social behaviors are based in genetics. This classic correlation-causation error cannot be excused on the grounds that Wade is just speculating: continents can be distinguished genetically; they also have different economic and social histories. One cannot conclude, as Wade does, that the former causes the latter.

The first paragraph of Chapter 7 summarizes Wade’s process of inference: “Each of the major civilizations has developed institutions appropriate for its circumstances and survival. But these institutions, though heavily imbued with cultural traditions, rest on a bedrock of genetically shaped human behavior. And when a civilization produces a distinctive set of institutions that endures for many generations, that is the sign of a supporting suite of variations in the genes that influence human social behavior.”16 I will focus on two of the studies invoked by Wade to justify his totally unfounded claims that differences in the societies of different continents (which he terms “races” even though in a biological sense they are not understood as such) are due to their genetic differences.

The first is by Gregory Clark, an economic historian who studies changes in interpersonal violence, literacy, the propensity to save, and the propensity to work in the English population from 1200 CE to 1800 CE.17 As Wade puts it18, during this period “the nature of the people had changed.” Between 1200 and 1800 “these behavioral changes in the English population … gradually transformed a violent and undisciplined peasant population into an efficient and productive workforce.” How did this happen? “Clark has uncovered the simple genetic mechanism … the rich had more surviving children than the poor.”19 Wade explains further: “Most children of the rich had to sink in the social scale,”20 and as a result, “Their social descent had the far reaching genetic consequence that they carried with them the same behaviors that had made their parents rich.”21

Against the argument that changing culture may have been involved in the 600-year process, Wade states that these “behaviors emerged gradually over several centuries, a time course more typical of an evolutionary change than a cultural change.”22 To justify his claim that 600 years is enough time to have produced “significant changes in social behavior”23 of the English, Wade leans on experiments by Belyaev, who artificially selected silver foxes for tameness. The strength of this selection was extreme: “typically not more than 4 or 5 percent of male offspring and about 20 percent of female offspring have been allowed to breed.”24 The strongest natural selection on humans is orders of magnitude weaker than this “sufficiently intense”25 artificial selection imposed on the foxes. Few evolutionists would agree that 600 years, that is, about twenty-five generations, is long enough for such significant behavioral changes to be due to human genetic evolution; Here and elsewhere in the book, Wade uses “evolutionary” where it is obvious that he means “genetic”. “Ingrained” is another euphemism he occasionally uses. For example, “Tribal behavior is more deeply ingrained than are mere cultural prescriptions. Its longevity and stability point strongly to a genetic basis.”26 Galton and Pearson would have approved of Wade’s espousal of a genetic basis for class differences; there is more than a whiff of eugenics here.

Wade devotes almost four pages of Chapter 7, the longest chapter in the book, to IQ. After claiming27 that “Intelligence is almost certainly under genetic influence,” he goes on to discuss the relationship between wealth and IQ and invokes the work of Richard Lynn and Tatu Vanhanen, in particular their book IQ and Global Inequality.28 Lynn is notorious for his work as an associate editor of The Mankind Quarterly, described by the famous psychologist Leon Kamin29 as a “vulgarly racist” journal. Lynn’s 1991 paper on IQ of Africans is described by Kamin as “truly venomous racism, combined with scandalous disregard for scientific objectivity.” In 2002, Lynn wrote the nonsensical statement: “The conclusion that there is a true association between skin color and IQ is consistent with the hypothesis that genetic factors are partly responsible for the black-white difference in intelligence … the evidence that a statistically significant correlation is present confirms the genetic hypothesis.”30 In placing so much emphasis on the work of one of the most consistently racist psychologists (whose work was strongly supported by the notorious Pioneer Fund, which also supported William Shockley and was chaired by an even more notorious scholar, J. Philippe Rushton), Wade has chosen to ignore important studies on IQ and environment such as those by Brooks-Gunn et al.31 and Turkeheimer et al.32 Brooks-Gunn found that “adjustments for economic and social differences in the lives of black and white children all but eliminate differences in the IQ scores between the two groups,” suggesting that socio-economic status (SES) might be an important contributor to high heritability estimates. In the same vein, Turkheimer et al. found that heritability of IQ depended strongly on SES: there was a high heritability in higher SES environments but not in low SES environments. In choosing not to mention such studies that find very strong environmental contributions to IQ, while relying on Richard Lynn, Wade cements his hereditarian credentials.

Wade gives the appearance of care in interpreting Lynn and Vanhanen: “It is hard to know which way the arrow of causality may be pointing, whether higher IQ makes a nation wealthier or whether a wealthy nation enables its citizens to do better on IQ tests.”33 However, from his statements about “the strong heritability of intelligence”34 and his belief21, referring back to Clark, that in England “the children of the rich carried with them inheritance for the same behaviors that made their parents rich,” we can only assume that Wade believes there is a genetic basis for both IQ and wealth. His “arrow of causality” has two points, with genetics responsible for both IQ and wealth.

This section of the book is redolent of the claims of Jensen, as well as Herrnstein and Murray, mentioned at the beginning of this review. It also harks back to claims by Taubman35 in the 1970s, based on correlations between relatives, that variation between individuals in wealth has a strong genetic basis. It is most informative to compare Goldberger’s 197736 criticism of Taubman’s analysis with related negative evaluations of studies on heritability of IQ. By invoking Richard Lynn on racial variation in IQ and wealth, Wade departs from his “speculative arena,” leaving us to infer not only that he is a devout hereditarian, but also that he is comfortable with Lynn’s racist worldview.

Wade goes even further than proposing a genetic basis for continental variation in wealth; he would have us believe that differences in economic and political institutions among populations have a genetic basis. He criticizes the book Why Nations Fail by Acemoglu and Robinson37 because “they have ruled out the obvious possibility that variations in human behavior are the cause of good or bad institutions.”38 Variation in institutions is why “a part of the world has grown steadily and vastly richer over the past 300 years.”39 He concludes that a reasonable explanation for this variation “is available in terms of human evolution.”40

Wade is using “evolution” here to mean the production and maintenance of genetic differences, and “variations in human behavior” is his euphemism for racial, and hence (in his understanding) genetic differences. He appears to backtrack slightly in the final chapter, where he poses the paradox “that people as individuals are so similar yet human societies differ so copiously.”41 His resolution of the paradox is that these societal differences “stem from the quite minor variations in human social behavior … that have evolved within each race during its geographical and historical experience.”42 Again, “evolved” must be understood in genetic terms: it is “because of their institutions—which are largely cultural edifices resting on a base of genetically shaped social behaviors—that the societies of the West and East Asia are so different.”43

We can juxtapose Wade’s conclusions on the genetic basis of racial differences in wealth, economies, and institutions with those of Ashraf and Galor44 on a similar topic. Their claim was that the high and low molecular genetic diversity characteristic of African and Native American populations, respectively, “have been detrimental for the development of these regions,” while “the intermediate levels of diversity associated with European and Asian populations have been conducive for development.” Wade’s use of worldwide patterns of human molecular genetic variation to define races and his inference that genetic variation between races explains their economic differences are qualitatively similar to Ashraf and Galor’s thesis. Speculation aside, readers of A Troublesome Inheritance are advised to heed the admonition by Guedes et al.45 concerning Ashraf and Galor: “bold claims on the basis of weak data and methods can have profoundly detrimental social and political effects.”

Wade’s premise is that molecular population genetics has shown sufficient variation between continents to define races. He then argues that these genetic differences are responsible for differences in individual social behaviors that “undergird”46 societal institutions, which themselves differ among races. Echoes of the hereditarian arguments about racial difference in IQ and the reductionist arguments of sociobiology and evolutionary psychology resound in A Troublesome Inheritance. I have no trouble with the existence of human genetic variation. It is Wade’s dangerous interpretation, however speculative, of the meaning of this variation that is indeed troublesome.

Author Marc Feldman is the Burnet C. and Mildred Finley Wohlford Professor in the School of Humanities and Sciences

Marcus W. Feldman is the Burnet C. and Mildred Finley Wohlford Professor in the School of Humanities and Sciences at Stanford and a Founding Director of CEHG.

[Correction: a previous version of this article said that 600 years is about 20 generations; it is actually closer to 25].


  1. Jensen, A. R. (1969) How much can we boost IQ and scholastic achievement. Harvard Educational Review 39: 1–123.
  2. See Kamin, L. J. (1974) The Science and Politics of IQ. Potomac, MD: Lawrence Erlbaum.
  3. Herrnstein, R. J., and C. Murray (1994). The Bell Curve: Intelligence and Class Structure in American Life. New York: Free Press.
  4. Jacoby, R., and N. Glauberman (eds.) (1995) The Bell Curve Debate. New York: Times Books.
  5. Feldman, M. W., and R. C. Lewontin (1975). The heritability hangup. Science 190: 1163–1168.
  6. I abbreviate A Troublesome Inheritance in this list of references as “ATI.”
  7. ATI, pp. 97–99: Rosenberg, N. A., et al. (2002) The genetic structure of human populations. Science 298: 2381–2385.
  8. Rosenberg, N.A., et al. (2005) Clines, clusters, and the effect of study design on the inference of human population structure. PLoS Genet. 1: 660–671.
  9. ATI, p 99: Li, J. Z., et al. (2008) Genome-wide characterization of genetic diversity in human populations. Science 319: 1100–1104.
  10. Cann, H. M., et al. (2002) A human diversity cell-line panel. Science 296: 261–262.
  11. Lewontin, R. C. (1972) The apportionment of human diversity. Evolutionary Biology 6: 381–398.
  12. Barbujani, G., et al. (1997) An apportionment of human DNA diversity. Proceedings of the National Academy of Sciences 94: 4516–4519.
  13. ATI, p. 120.
  14. Feldman, M.W., and R.C. Lewontin (2008) Race, ancestry, and medicine. Pp. 89–101 in Koenig, B.A. et al. (eds.) Revisiting Race in a Genomic Age. Rutgers University Press.
  15. ATI, p. 15.
  16. ATI, p. 150.
  17. Clark, G. (2007) A Farewell to Alms: A Brief Economic History of the World. Princeton, NJ: Princeton University Press.
  18. ATI, p. 154.
  19. ATI, p. 159.
  20. ATI, p. 160.
  21. ATI, p. 160.
  22. ATI, p. 160.
  23. ATI, p. 161.
  24. These experiments are reviewed by Belyaev’s long-time collaborator Lyudmila N. Trut (1999) Early canid domestication: the farm-fox experiment. American Scientist 87: 160–169.
  25. ATI, p. 161.
  26. ATI, p. 177.
  27. ATI, p. 190.
  28. Lynn, R., and T. Vanhanen (2006) IQ and Global Inequality. Augusta, GA: Washington Summit.
  29. Kamin, L. (1995) Lies, damned lies, and statistics. Pp. 81–105 Jacoby, R., and N. Glauberman (eds.) The Bell Curve Debate. New York: Times Books. (page 86)
  30. Lynn, R. (2002) Skin color and intelligence in African-Americans. Population and Environment 23: 365–374. (page 372)
  31. Brooks-Gunn, J., et al. (1996) Ethnic differences in children’s intelligence test scores: role of economic deprivation, home environment, and maternal characteristics. Child Development 67: 396–408.π
  32. Turkheimer, E., et al. (2003) Socio-economic status modifies heritability of IQ in young children. Psychological Science 14: 623–628.
  33. ATI, p. 192.
  34. ATI, p. 203.
  35. Taubman, P. (1976) The determinants of earnings: genetics, family, and other environments: a study of white male twins. American Economic Review 66: 858–870.
  36. Goldberger, A. S. (1977) The genetic determination of income. University of Wisconsin Social Systems Research Institute. Paper 7707.
  37. Acemoglu, D., and J. A. Robinson (2012) Why Nations Fail: The Origins of Power Prosperity and Poverty. New York: Crown.
  38. ATI, p. 196.
  39. ATI, p. 196.
  40. ATI, p. 196.
  41. ATI, p. 240.
  42. ATI, p. 241.
  43. ATI, p. 241.
  44. Ashraf, Q., and O. Galor (2013) The out-of-Africa hypothesis, human genetic diversity, and comparative economic development. American Economic Review 103: 1–46.
  45. Guedes, J. d’A., et al. (2013) Is poverty in our genes? Current Anthropology 54: 71–79.
  46. ATI, p. 126.

Catching the drift of experimental evolution.

Blog author Doc Edge is a graduate student in Noah Rosenberg's lab.

Blog author Doc Edge is a graduate student in the Rosenberg lab.

One picture of science, popular in dramatic depictions of scientific history, shows an isolated theorist working out the implications of a bold idea. In this picture, the theorist relies, somewhat indifferently, on the ingenuity of an unknown, unspecified experimentalist who will someday—perhaps many years later—test the theorist’s ideas against unforgiving reality. It’s trite to point out that this picture isn’t a good representation of actual scientific practice. The relationship between theory and experiment both is and should be more bidirectional, more collaborative, and more nuanced than this, as philosophers of science, sociologists of science, and scientists themselves have pointed out for decades. A recent paper by CEHG graduate student Arbel Harpak (Pritchard lab) and Guy Sella (Stanford PhD 2001 from Marc Feldman’s group, in the less-sunny, pre-CEHG days) is a lesson in how experimentally-motivated theoretical work can inform both future theory and future experiment.

(Neutral) Evolution in a Test Tube

Harpak and Sella take up one of the most interesting contexts for the study of evolution: the serial-transfer experiment. In serial-transfer experiments, microorganisms (usually E. coli or yeast) are allowed to divide in an isolated container with some nutrients. Once the population in a container reaches a pre-specified size (which Harpak and Sella call N2), a sample of a pre-specified size (N1) is taken from the original container and transferred to a new one. This process of growth, sampling, and transfer is then repeated many times.

A schematic of a serial transfer experiment (from Sprouffske et al., 2012) [Note: in a previous version this image was mistakenly attributed to A. Harpak.]

Both experimenters and theorists have focused on understanding selection in serial-transfer experiments. This is sensible: the serial-transfer scenario allows researchers to manipulate the parameters that influence adaptation—population size, selective pressure, etc.—and to study the dynamics of adaptation. Harpak and Sella start from a premise that has been less widely-appreciated in the serial-transfer context: namely, that the other forces of evolution, including drift and demography, are also active in serial-transfer contexts. These selectively-neutral forces are ever-present, running in the background of experiments designed to study selection. It is not that Harpak and Sella claim that neutral forces are as or more important than selective forces in a typical serial-founder experiment—they acknowledge that selective forces likely predominate. But they rightly point out that we cannot fully understand what selection does without understanding the context in which it acts. Population geneticists have known this for a long time, but until now, they have not formally applied this insight to serial-transfer experiments.

Harpak and Sella develop a fully-articulated model for neutral evolution in a serial-transfer context. In population-genetic terms, the outcome of a serial-transfer experiment is the product of repeated periods of growth followed by bottlenecks. They build on the intuitions that population geneticists already have about population growth and bottlenecks to sketch a picture of what serial transfers will do to measurements of genetic diversity. Diversity slowly builds up as the population grows and as more distinct lineages appear in the population. But when a small subset of the population is removed and transferred to a new container, only the diversity that is represented in that small subset will be present in the next phase of the population’s growth. When the experiment involves many cycles of serial transfer, the genetic diversity ends up looking like that of a population with a constant size of N1*log2(N2/N1), where N1 is the number of organisms transferred during each cycle and N2 is the size to which the population grows just before the end of the cycle. This is a neat result—as has been found in other contexts, the genetic diversity is more strongly influenced by extremity of the bottleneck than it is by the maximum size to which the population grows.

Relaxing the Assumptions

This result agrees with population geneticists’ intuitions about demographic changes and genetic diversity. Nonetheless, there are two reasons why it is unrealistic for actual serial-transfer experiments. First, most serial-transfer experiments are not long enough to reach the predicted equilibrium state. Second, the equilibrium prediction assumes that all the microorganisms in the sample divide at the same time, but in reality, there is variation in the length of time that a cell persists before dividing. Harpak and Sella adjust their model to deal with both of these issues. In considering the shorter time scales typical of serial-transfer experiments, they find that the shorter the experiment is, the more the genetic diversity will be dominated by rare variation—genetic differences seen only in a single individual, for example. They also derive results for another version of their model in which the length of time to the next cell division is random. When cells vary in how long they take to divide, the average number of differences between any two cells decreases. This is because cells that divide quickly are likely to leave more offspring in subsequent generations. This means that any two cells in a subsequent generation are more likely to be recently related, since there is an increased chance that they both descend from some quickly-dividing recent ancestor.

Harpak and Sella’s paper is a step toward a fully-realized population-genetic framework for serial transfer experiments and for experimental evolution more generally. We have long known that models of evolution have to account for selectively-neutral forces in addition to selection, and it is only sensible to think of evolutionary experiments in the same way. By responding to a frequently-used experimental setup, Harpak and Sella have done a service for experimental evolution, providing a better understanding of the selectively neutral processes that act in serial transfer experiments. At the same time, Harpak and Sella perform a service that is perhaps even more important for population-genetic theory. Evolutionary experiments are an unprecedented opportunity to learn about the actions of evolutionary forces at a large scale. Moreover, experiments offer researchers the power to control the setting in which evolution takes place and to repeat evolutionary processes. Population-genetic theorists, who until recently have had to make do with the single iteration of evolution that nature has provided, ought to jump at the chance to test their ideas in real-time. Harpak and Sella’s paper is an example of how theorists attending closely to experimental methods can lay the ground for improved experiments and improved theories.


Harpak A, Sella G (2014) Neutral null models for diversity in serial transfer evolution experiments. Evolution. DOI: 10.1111/evo.12454

Sprouffske K, Merlo LM., Gerrish PJ, Maley CC, & Sniegowski PD (2012). Cancer in light of experimental evolution. Current Biology22(17), R762-R771, DOI: 10.1016/j.cub.2012.06.065.

Paper author Arbel Harpak is a graduate student in the Pritchard lab.

Paper author Arbel Harpak is a graduate student in the Pritchard lab.

What’s Sardinia got to do with it? Ancient and modern genomes shed light on the genetic structure of Europe.

Blog author Yuan Zhu is a graduate student in Dmitri Petrov's lab

Blog author Yuan Zhu, formerly a PhD student in the Petrov lab, is now a Research Fellow at the Genome Institute of Singapore.

The Neolithic Revolution is the oldest documented agricultural revolution in human history. More than just the domestication of certain crops and animals, it describes a critical time in human history when hunter-gatherer groups transitioned into sedentary farming communities. This drastic change in lifestyle led to a major shift in living conditions and cultural practices, setting up the necessary prerequisites to support the kind of population density eventually possible in modern society.

In Central Europe, the Neolithic Revolution is thought to have taken place around 8,000-4,000 BC. Historians have long wondered about how farming was introduced and spread across the continents. Was the new practice brought in as novel ideas incorporated by local communities? Did new immigrants bring their lifestyle with them, possibly outcompeting existing hunter-gatherers and eventually displacing them all together? Was it perhaps even more complicated? What happened after?

What Ötzi can tell us

Ancient human remains from around the time of the revolution can yield some insight. Ötzi the Tyrolean Iceman, a 5,300-year-old natural mummy found frozen in the Alps on the border of Italy and Austria, was recently shown (by a group that included CEHG researchers Martin Sikora and Carlos Bustamante) to belong to a Y-chromosome lineage mostly found in contemporary Sardinia [1]. This was surprising information. The Iceman’s life was spent in a narrow range within 60 km of his site of discovery [2]. He was unequivocally local, and clearly a farmer. Yet his lineage has since disappeared from Central Europe, suggesting that demographic scenarios were more complex than expected, and that at some point this Sardinian-like ancestry may have spanned Neolithic Europe.

A). The location of the discovery sites of ancient individuals studied, with hunter-gatherers (HG) represented as circles, and farming (F) individuals represented as squares. B). ADMIXTURE results of modern populations on the left panel, and inferred genetic composition of ancient individuals on the right. [Adapted from Figure 1, Sikora et al. 2014.]

A). The location of the discovery sites of ancient individuals studied, with hunter-gatherers (HG) represented as circles, and farming (F) individuals represented as squares. B). ADMIXTURE results of modern populations on the left panel, and inferred genetic composition of ancient individuals on the right. [Adapted from Figure 1, Sikora et al. 2014.]

Sardinia: a genetic snapshot of the Neolithic?

In a recent paper published in PLOS Genetics, Sikora and colleagues sought to address this hypothesis by making full use of recent advancements in the sequencing of nuclear ancient DNA [3]. However, the Iceman alone was not sufficient to represent a continent. Ancient DNA sequences from six individuals from across Europe, including both farmer and hunter-gatherer individuals, were analyzed by the authors in order to paint a clearer picture of the demographics of Neolithic Europe. Two of the farmers were found in Bulgaria and were previously sequenced using an ancient DNA capture method developed by Sikora’s colleague in the Bustamante lab, Meredith Carpenter [4, and see blog post here]. In addition, Sikora made use of contemporary population SNP data, including sequence data from over 400 modern Sardinians, to provide a solid reference from which to estimate the true genetic affiliation of these ancient humans.

Some of the most interesting results from the analysis came from contrasts between the farmers (Iceman, gok4, and P192-1), the hunter-gatherers (ajv7 and brana1), and modern-day European populations. When the authors applied the clustering algorithm ADMIXTURE to the data, they found that the farmer individuals had significant portions of shared ancestry with modern Sardinians (Southern Europe), a characteristic largely absent in the HG individuals, who showed mainly Northern European (Basque) and Russian affiliated ancestry. Principal component analysis (PCA) and a statistic called the D-test agreed with high confidence—hunter-gatherers looked more Northern European, whereas farmers seemed more Sardinian than any other European group tested. TreeMix, a program that models population splits while allowing for admixture between branches, provided a similar answer when applied to the data from 1000 Genomes and the modern Sardinians, and further suggested a possible admixture scenario involving at least three major events, all of which falls neatly in line with previous work.

Taken together, the data support the authors’ original hypothesis—Sardinian-like ancestry was probably once common in Neolithic Europe. The Iceman, gok4, and P192-1 were discovered in very different locations, and P192-1 in particular was 2,000 years younger than the others, making it even more unlikely that all three were recurrent immigrants from Sardinia (which was thought to be uninhabited by hunter-gatherers prior to the Neolithic), and further suggesting that the lineage may have persisted for a while on the continent. In fact, Sikora and colleagues propose that Sardinia is a “modern-day ‘snapshot’ of the genetic structure of the people associated with the spread of agriculture in Europe.”

A proposed, highly simplified version of recent European demographic history. A). Early hunter-gatherers (closest to modern day Russian/Basque) were B). heavily influenced by an influx of farmers C) who spread across all of Europe and into Sardinia D). and subsequently maintained only in Sardinia due to genetic isolation. [Adapted from Figure 4, Sikora et al. 2014]

A proposed, highly simplified version of recent European demographic history. A). Early hunter-gatherers (closest to modern day Russian/Basque) were B). heavily influenced by an influx of farmers C) who spread across all of Europe and into Sardinia D). and subsequently maintained only in Sardinia due to genetic isolation. [Adapted from Figure 4, Sikora et al. 2014]

Bridging the past and the future with ancient DNA

From here, the story is far from over. In fact, it only gets more complicated, and more work remains to be done. While a simplified model was proposed, the authors note that multiple sources of evidence suggest a far more complex and nuanced recent demographic history for Europe that we have yet to untangle. There are issues with ancient DNA sequences, such as characteristic DNA damage patterns, that are unique to the nature of the data. Potential issues with current methods being unable to handle such underlying patterns forced the authors to analyze every ancient DNA sample against modern populations individually. As with every advance in sequencing technology, with ancient DNA sequencing getting more accurate and accessible, new analytical methods must be developed to take full advantage of the data.


[1] Keller A, Graefen A, Ball M, Matzas M, Boisguerin V, et al. (2012) New insights into the Tyrolean Iceman’s origin and phenotype as inferred by whole-genome sequencing. Nature Communications 3: 698.
[2] Müller W, Fricke H, Halliday AN, McCulloch MT, Wartho J-A (2003) Origin and Migration of the Alpine Iceman. Science 302: 862–866. doi: 10.1126/science.1089837
[3] Sikora M, Carpenter ML, Moreno-Estrada A, Henn BM, Underhill PA, et al. (2014) Population Genomic Analysis of Ancient and Modern Genomes Yields New Insights into the Genetic Ancestry of the Tyrolean Iceman and the Genetic Structure of Europe. PLOS Genetics, DOI:10.1371/journal.pgen.1004353
[4] Carpenter, ML, Buenrostro, JD, Valdiosera, C, Schroeder, H, Allentoft, ME, Sikora, M, Rasmussen, M, et al. (2013). Pulling out the 1%: Whole-Genome Capture for the Targeted Enrichment of Ancient DNA Sequencing Libraries. Am J Hum Genet. 2013 Nov 7;93(5):852-64. doi: 10.1016/j.ajhg.2013.10.002.

Paper author: Martin Sikora was a postdoc in Carlos Bustamante's lab. He is now a group leader at the Centre for GeoGenetics in Copenhagen, Denmark.

Paper author Martin Sikora was a postdoctoral fellow in Carlos Bustamante’s lab. He is now a group leader at the Center for GeoGenetics in Copenhagen, Denmark.