This is the third and last installment in my "revisiting the Fisher-Wrigt
controversy" series The first installment is
here; the second is here. For a little context, I repost the introductory paragraph from the
first and second posts. (This is another long post; just a warning.)
R. A. Fisher, J. B. S. Haldane, and Sewall Wright are the architects of theoretical population genetics.
Between 1918 and 1932, these three theorists ushered in the field and
set the stage for the period of the history of evolutionary biology
usually called the "evolutionary synthesis."
It's well known that from 1929 until Fisher's death in 1962 that Fisher
and Wright were engaged in a sometimes heated controversy over their
alternative qualitative interpretations of their quantitative models.
In 1985, Will Provine published "The R. A. Fisher—Sewall Wright
Controversy" in Oxford Surveys of Evolutionary Biology (Provine
1985). In that paper, Provine discusses three key disputes between
Fisher and Wright: (1) evolution of dominance, (2) their general
evolutionary theories, and (3) evolution of the Scarlet Tiger Moth, Panaxia dominula.
Now, Provine's biography of Wright published in 1986 is a fuller
treatment of the controversy (Provine 1986). However, I'm in the
process of writing, basically, a new version of Provine's paper in
which I revisit each of the debates of his 1985 essay. In this post, I
look at Fisher's and Wright's dispute over their general evolutionary theories .
I think I don't quite do justice to the debate with this post, but I do
think it's a fair overview. And it does presume some familiarity with
Fisher's and Wright's general evolutionary theories.
The debates between Fisher and Wright during the late 1920s and 1930s, over dominance and their general evlutionary theories, were largely theoretical. However, in 1947, Fisher, with the ecological geneticist E. B. Ford published an experimental paper aimed at discrediting Wright's Shifting Balance Theory and substantiating Fisher's natural selection theory (Fisher and Ford 1947). Fisher and Ford's paper describes and analyzes data from what was at the time a fairly novel field experimental technique, the capture and release protocol, used in populations of the moth Panaxia dominula. Fisher and Ford argued via their experimental results that even in small(ish) populations ( between 1,000 and 10,000) -Wright's assumed norm- genetic drift -Wright's most important evolutionary factor according to Fisher and Ford- was evolutionarily inefficacious. Fisher and Ford argue further that natural selection, even in smallish populations, is the driving factor of evolution.
The capture and release protocol that Fisher and Ford describe in their 1947 paper was carried out between the years 1939-1946. Moths, mainly in Cothill, Oxfordshire, England, were captured (and re-captured), marked inconspicuously with paint (or not if re-captured), scored for phenotype of interest (or as a re-capture), and released if unharmed. The purpose of Fisher and Ford's capture and release protocol was to collect data over time for fluctuations in frequency of genes of interest by scoring particular phenotypes, here wing coloration patterns. The Scarlet Tiger moth is easily identified by its wing coloration. The forewing is black with iridescent green structural coloring and a pattern of green or white spots. The hind wing is usually bright scarlet with black markings. In Fisher and Ford's Panaxia study, form dominula, f. medionigra, and the very rare f. bimacula refer to particular patterns of wing coloration that were assumed to correspond to dominant homozygous (AA), heterozygous (Aa), and recessive homozygous (aa) genotypes respectively (color plates can be found in Fisher and Ford 1947). Breeding studies had been done by E. A. Cockayne (1928), Ford (1940), and H. B. D. Kettlewell (1942) which provided the genotype-phenotype correspondence. (Note: Sorry that there is no photo here. In my experience, the Typepad blogging software is poor at handling images.)
Fisher and Ford argued their data showed the fluctuations in the frequencies of the f. medionigra genes were too large from year to year to be due to genetic drift. Essentially, their specific argument was that even though population size was sufficiently small for genetic drift to be effective, drift nevertheless was not a factor (because the gene frequency changes were too high to merely be due to chance). For the years the population was studied, the average population size was in the range of 3,200-4,000 moths, with approximately 11% overall being f. medionigra, and a total gene frequency change of approximately 6% (Fisher and Ford 1947: 150, 164). Fisher and Ford ultimately inferred that because changes in gene frequencies in the moth populations were not due to random genetic drift, they must be due to natural selection.
In 1948, Wright published a critique of Fisher and Ford's study (Wright 1948). Wright objected on several grounds. First, Fisher and Ford had misinterpreted the role Wright had assumed for random genetic drift. They attributed to him more of a role than he, himself, had attributed. Second, their inference that selection must be the cause of the changes in gene frequencies in the populations of the moths was not justified experimentally. Fisher and Ford provided no direct evidence that selection is the cause; they only infer it after rejecting drift. Wright's paper drew an acerbic attack from Fisher and Ford published in 1950 (Fisher and Ford 1950). Wright (1951) again responded. The substance of the disagreement, after Wright (1948), is the problem of interpreting Wright's view of the role of genetic drift in evolution. These exchanges have all been thoroughly discussed by Provine (1986). What I am about to discuss has not been discussed (at least not the work since 1962; but see Provine's 1986 brief discussion of Ford's summary of the work in the 1964 and 1975 editions of his Ecological Genetics, Ford 1964; 1975).
The experimental work on Panaxia has continued in the spirit of Fisher and Ford's pioneering study. In fact, the work on Panaxia comprises one of the longest-running ecological genetics field experiments, lasting around 60 years, in the history of the scientific field, with the most complete scientific review having been completed by L. M. Cook and David A. Jones for the years 1939-1995 (Cook and Jones 1996). Basically, the ongoing Panaxia work has been, by and large, taken as 60 years of replication of Fisher and Ford's 1947 findings. (Certainly Cook and Jones agree with this assessment.) Interestingly, the only direct observational evidence of selection (e.g., observation of a long-term environmental perturbation such as predation, see e.g., Endler's 1986, chapter 3 methods of detection) acting to change gene frequencies of the f. medionigra genes in the Panaxia work is due to P. M. Sheppard working with Cook and with Ford (Sheppard 1951; Sheppard and Cook 1962; Ford and Sheppard 1969). The selectionist interpretation of the Panaxia work has largely been based on the same eliminative inference (i.e., the inference from observed gene frequency fluctuations independent of a perturbation) Fisher and Ford made in 1947 (in spite of Wright's 1948 criticism of it).
Between 1993 and 1997, however, Cyril A. Clarke, David Goulson, and Denis F. Owen published a series of experimental and review papers criticizing the broad-ranging Panaxia work (Owen and Clarke 1993; Owen and Goulson 1994; Goulson and Owen 1997). Interestingly, the biologists started out with the intention of replicating Fisher and Ford's (1947) results. However, each of the biologists argue ultimately there are good reasons reject the selectionist interpretation of the data: Selection is not the cause of the changes in gene frequencies. Genetic drift is not the cause either.
Owen and Clarke in 1993 reported and analyzed data from a combined reared and wild capture and release protocol with Panaxia in 1991-1992 in Cothill and two other Oxfordshire fens (viz., Dry Sandford, North Hinksey). During their protocol, they noted f. medionigra was extremely variable. Owen and Clarke were suspicious of this variability, prompting them to change their scoring technique. Rather than scoring the moths simply for the three forms, following Fisher and Ford (1947), they complicated their technique, scoring the phenotype by more specific patternings of wing coloration (Owen and Clarke 1993: 396-397). On the standard model, moths with a small or absent spot combined with either a yellow, black, or absent hindwing spot are scored as f. medionigra. Owen and Clarke scored only moths with a small or absent forewing spot in combination with a black hindwing spot as f. medionigra; the others were scored as f. medionigra-like. Using the revised scoring model, the frequency of the apparent f. medionigra gene frequencies that Owen and Clarke reported were, they claim, some of the largest on record (Owen and Clarke 1993: 398). Depending on which phenotypes are scored as f. medionigra, with the low frequency scoring only f. medionigra and the high frequencies all-inclusive, the frequency estimates were 0.4%-49% for their 1991 reared population, 0.4-2.7% for their 1992 wild population, and a remarkable, so they claim, 2.7%-41.9% for their 1992 reared population. This last is four times as high as any population since 1939.
Owen and Clarke found the variability in their data, tied directly to the variability in the form of the moth, impossible to reconcile without looking more closely at the causes of the variability. Although they could not definitively say, they suggested the extreme variability of wing coloration phenotypes was due to an environmental factor (this effect is well understood by developmental biologists but perhaps less so by ecologists). Temperature fluctuations, which would cause an overall darkening in the wing color pattern would make f. dominula look more like f. medionigra and f. medionigra look more like the rare f. bimacula. Owen and Clarke believed such an environmental effect would account for the exaggerated representation of f. medionigra on the standard scoring model. Owen and Clarke raised the issue that Fisher and Ford's original study, as well as the subsequent studies, may be compromised due to probable scoring errors related to temperature effects on the expression of the wing coloration phenotypes in the moths. That is, the gene frequency data may be corrupted because some phenotypes may have been scored as the wrong genotype. Such scoring errors, argued Owen and Clarke, would skew the results and their analysis significantly. Owen and Clarke noted something interesting: Kettlewell had done temperature controlled breeding experiments in the lab in 1943-1944 (Kettlewell 1943/1944). He had shown temperature fluctuations caused differences in expression of the wing coloration phenotype. Now, Kettlewell's experiments were done before Fisher and Ford's 1947 paper. Yet, Fisher and Ford, at least according to the published record, did not mention Kettlewell's findings. (The paper of Kettlewell's that Fisher and Ford do cite is not his temperature effects paper.)
In 1994, Owen and Goulson published a paper demonstrating temperature fluctuations cause changes in the expression of wing coloration during pupal development of the moth (Owen and Goulson 1994). Genes that under mild, or normal temperatures express to look like f. dominula will under more extreme temperatures express to look like f. medionigra. To show this, Owen and Goulson had done lab breeding experiments in controlled temperature environments on Panaxia (larvae) they had brought in from Cothill and North Hinksey. The wings of moths raised in temperatures below 12°C and above 24°C darken. They then argued there is no reliable way of tracking (in nature) changes in gene frequencies in Panaxia by observing and scoring phenotypes: Temperature fluctuations significantly affect the phenotype, belying the genotype. Ultimately, Owen and Goulson concluded the Panaxia work is not good support for the prevailing selectionist interpretation of the Panaxia data.
There were no temperature data records in Fisher and Ford's studies; similarly for the subsequent ones. So, there is really no way to know whether temperature fluctuations really played any role in the variability of wing coloration in the moths during Fisher and Ford's or subsequent studies. Cook and Jones (1996), supporters of Fisher and Ford, claim in their review of the Panaxia work that there is nothing to worry about: They statistically analyzed weather data, independently recorded, and concluded temperature fluctuations likely did not affect expression of the wing coloration trait (Cook and Jones 1996, 1625). So, the long-standing support for natural selection by the Panaxia work, according to Cook and Jones, goes undiminished by Owen and Goulson's 1994 temperature work.
By 1997, Goulson and Owen had done additional temperature experiments and had re-scored museum specimens of Panaxia originally scored by Fisher, Ford, their workers and subsequent workers (Goulson and Owen 1997; cf. Clarke et al. 1991). Goulson and Owen found errors in scoring had occurred in the Panaxia capture and release studies according to the Owen-Clarke revised scoring technique. Cook and Jones, in 1996, never comment on this possibility. Yet, it had been argued by Owen and Clarke (1993) that such errors were likely. Further, Goulson and Owen's newer temperature experiments further substantiate their claim the expression of wing coloration phenotypes in Panaxia is all but controlled by fluctuations in temperature. Goulson and Owen's ultimate assessment of the Panaxia work and its contribution to the Fisher-Wright controversy is this (Goulson and Owen 1997: 616-617): Due to the severe problems with the Panaxia experimental methodologies, the results of the work must be rejected as a way of settling Fisher's and Wright's disagreement over the role of drift in evolution. As I see matters, Wright's (1948) original critique of Fisher and Ford's (1947) argument that the elimination of drift is sufficient to establish selection, ignored from the moment Wright raised it, is substantiated.
Jones, in 2000, published a paper that weakens Goulson and Owen's (1997) conclusion (Jones 2000). What Goulson and Owen lacked was a capture and release study of gene frequency fluctuations in the moth that included the relevant experimental check for temperature fluctuations. Jones (2000) published that study. Jones captured and released moths in Cothill, Oxfordshire during the years 1995-1999. Jones followed Owen and Clarke's (1993) scoring model. And, importantly, Jones monitored temperatures were monitored in precisely the locations in the Cothill Fen that the larvae of the moths pupate. For an average year-to-year population size in the range of 3,100-5,000 moths, Jones recorded gene frequency fluctuations in f. medionigra in the range 0.73% to 2.62%. Jones further found air temperature reached lows and highs exceeding the boundaries at which Owen and Goulson (1994) reported darkening of the wings in the laboratory. Temperatures in the Cothill Fen litters ranged from 4°C to 33°C (Jones 2000: 580). However, the larvae and pupae never experienced those extremes for a prolonged period of time during any month between 1995 and 1999. The monthly mean was well within the 12°C to 24°C range for normal wing development (Jones 2000: 584).
Jones' (2000) gene frequency fluctuation data is consistent with the data collected since 1939. However, Jones never explicitly attributes the fluctuations he recorded between 1995 and 1999 to natural selection. Rather, Jones' purpose was to cast doubt on Owen and Goulson's (1994) criticism of the historical Panaxia work. My view is Jones has done just that. But he has not managed to remove all doubt cast upon the selectionist interpretation of the historical work. Again, Wright's (1948) original criticism of Fisher and Ford's (1947) argument for selection by elimination of drift stands. The Panaxia work is not a strong case to adjudicate between the relative roles of selection and drift in evolution.
Jones' (2000) paper is, so far as I am aware, the most recent in the ongoing Panaxia work. It is hard to say whether his assessment of the work will conclude the debates. Approximately 60 years have been devoted to the Panaxia work in efforts to resolve Fisher's and Wright's original disagreement over the role of random genetic drift in evolution. The first 10 years of work were hotly debated by Fisher, Wright and their immediate associates (e.g., Ford, Sheppard). About 40 years hence have been taken up with the assumption that the continued work has successfully replicated Fisher and Ford's demonstration of the primacy of selection in (even smallish) populations of Panaxia in England. It has only been in the last 15 years that work has been done to confute the selectionist interpretation. And that work did not start with that intention; Owen and Clarke (in particular) thought they were just going to replicate the long-standing selectionist results (Owen and Clarke 1993: 393).
References
Clarke, C., F. Clarke, and D. Owen (1991), "Natural Selection and the Scarlet Tiger Moth, Panaxia dominula: Inconsistencies in the Scoring of the Heterozygote, f. medionigra", Proceedings of the Royal Society of London B 244: 203-205.
Cockayne, E. A. (1928), “Variation in Callimorpha dominula”, Entomological Record 40: 153-160.
Cook, L. M. and D. Jones (1996), “The medionigra gene in the Moth Panaxia dominula: The Case for Selection”, Philosophical Transactions of the Royal Society of London B 351: 1623-1634.
Endler, J. (1986), Natural Selection in the Wild. Princeton: Princeton University Press.
Fisher, R.A. and E. B. Ford (1947), “The Spread of a Gene in Natural Conditions in a Colony of the Moth, Panaxia dominula, L.”, Heredity 1: 143-174.
Fisher, R.A. and E. B. Ford (1950), “The Sewall Wright Effect”, Heredity 4: 117-119.
Ford, E. B. (1940), “Genetic Research in the Lepidoptera”, Annals of Eugenics 10: 227-252.
Ford, E. B. (1964), Ecological Genetics. Welwyn Garden City, UK: The Broadwater Press, Inc.
Ford, E. B. (1975), Ecological Genetics. 4th edition. London, UK: Chapman Hall.
Ford, E. B. and P. M. Sheppard (1969), “The Medionigra Polymorphism of Panaxia dominula”, Heredity 24: 561-569.
Goulson, D. and D. Owen (1997), “Long-Term Studies of the medionigra Polymorphism in the Moth Panaxia dominula: A Critique”, Oikos 80: 613-617.
Jones, D. A. (2000), "Temperatures in the Cothill Habitat of Panaxia (Callimorpha) dominula L. (The Scarlet Tiger Moth)", Heredity 84: 578-584.
Kettlewell, H. B. D. (1942), “A Survey of the Insect Panaxia (Callimorpha) dominula”, Proceedings of the South London Entomological History Society: 1-49.
Kettlewell, H. B. D. (1943/1944), “Temperature Experiments on the Pupae of (1) Heliothis peltigera Schiff., and (2) Panaxia dominula Linn.”, Proceedings of the South London Entomological Natural History Society: 69-81.
Owen, D. and C. Clarke (1993), “The medionigra Polymorphism in the, Panaxia dominula (Lepidoptera Arctiidae): A Critical Re-Assessment”, Oikos 67: 393-402.
Owen, D. and D. Goulson (1994), “Effect of Temperature on the Expression of the medionigra Phenotype of the Moth Panaxia dominula (Lepidoptera: Arctiidae)”, Oikos 71: 107-110.
Provine, W. B. (1985), “The R.A. Fisher–Sewall Wright Controversy and its Influence Upon Modern Evolutionary Biology”, in R. Dawkins and M. Ridley (ed.), Oxford Surveys in Evolutionary Biology, Vol. 2. New York: Oxford University Press: 197-219.
Provine, W. B. (1986), Sewall Wright and Evolutionary Biology. Chicago: University of Chicago Press.
Sheppard, P. M. (1951), “A Quantitative Study of Two Populations of the Moth, Panaxia dominula, L.”, Heredity 5: 349-378.
Sheppard, P. M. and L. M. Cook (1962), “The Manifold Effects of the Medionigra Gene in the moth, Panaxia dominula and the Maintenance of Polymorphism”, Heredity 17: 415-426.
Wright, S. (1948), “On the Roles of Directed and Random Changes in Gene Frequency in the Genetics of Populations”, Evolution 2: 279-294.
Wright, S. (1951), “Fisher and Ford on the “Sewall Wright Effect”“, American Scientist 39: 452-458, 479.
Recent Comments