Other formats

    TEI XML file   ePub eBook file  


    mail icontwitter iconBlogspot iconrss icon

Tuatara: Volume 27, Issue 2, December 1984

Population Genetics and — the ‘Third View’ of Evolution *

page 121

Population Genetics and
the ‘Third View’ of Evolution


The logical structure of population genetics is discussed and its validity in the study of evolution is evaluated. It is argued that there are very serious difficulties with the accurate performance of genotype/form transformations. Alternatively, I suggest that evolution is a problem of individual form transformations not the genetic changes in population. Population genetics, according to the view expressed here, has a far more limited role in biological studies than that which it presently occupies.

Keywords: evolution, population genetics, Weiss, transformations, the “third view.”


The pages of Tuatara have recently been livened by a debate on the role of population genetics in courses on evolution. Saiff and Macbeth (1983) express the view that there is no need to teach population genetics in introductory courses on evolution, although advanced courses may include it as a “matter of history”. Hewitt (1983) has disagreed and argued that Saiff and Macbeth quote authors out of context and instead maintained “…that the failings of population genetic theory argue for more effort rather than less.”

There is currently substantial debate on aspects of evolutionary theory and since I am interested in this debate, and conduct courses in evolution, I would like to make the following comments.

Population Genetics

Lewontin (1974) has provided a succinct outline of the logical structure of population genetics theory. This can be represented as:
where G1 and G1 represent the genetic description of a population at times t and t + δ t, and F1, F2 and G2 represent form and genotype population descriptions of states during transformation (Fig. 1).

Lewontin (1974) describes these states as:

  • T1: a set of epigenetic laws that give the distribution of form that results from the development of various genotypes in various environments.

  • T2: the laws of mating, of migration, and of natural selection that transform the phenotypic array in a population within the span of a generation.

  • T3: an immense set of epigenetic relations that allow inferences about the distribution of genotypes corresponding to the distribution of forms, F2.

  • T4: the genetic rules of Mendel and Morgan that allow us to predict the array of genotypes in the next generation produced from gametogenesis and fertilization, given an array of parental genotypes.

Essentially then, population genetics attempts to map a set of genotype distributions into a set of form distributions. It provides a transformation in population form space, and then maps these new forms back into populations of genotypes where a final transformation occurs to produce the genotypic distribution in the next generation. page 122
Fig. 1 The logical structure of population genetics (modified from Lewontin, 1975).

Fig. 1 The logical structure of population genetics (modified from Lewontin, 1975).

Let us critically consider these transformations in turn:

T1. The path from the gene to form is, at least, a complex one. Perhaps it is best to examine the views, not of population geneticists who do not immediately concern themselves with this problem, but of a developmental biologist. Paul Weiss (1950) views this transformation as involving a number of levels (Fig. 2). He remarks “In dealing with the relations between genetics and development it is well to keep in mind that an organism is constituted like a system of Chinese boxes, in which larger ones enclose smaller ones in a descending series of magnitudes.” (Weiss, 1950). It seems likely that this was on Fritjof Capra's mind (1982 p. 108) when he said recently of genetic reductionism, “It ignores the fact that the page 123 organisms are multi-levelled systems, the genes being embedded in the chromosomes, the chromosomes functioning within the nucleus of their cells, the cells incorporated in the tissues, and so on. All these levels are involved in mutual interactions that influence the organisms development and result in wide variations of the “genetic blueprint”. Certainly it is clear that genes, the environment and events in development interact to produce form (Suzuki et al. 1981 p20), and hence developmental processes themselves should also be included in any such transformation.

Interactions also occur within levels. For example non-allelic genes commonly interact and such interactions affect form. Similar interactions also occur between chromosomes. De Betham Anderson (1984), for example, has shown that in New Zealand lowland grasshopper Phaulacridium marginale the presence of heterochromatic B chromosomes affects rates of crossing-over in the autosomes. Similar interactions have been recorded in a number of cases (e.g. Naranjo and Lacadena, 1980: Miklos and Nankivell, 1976).

It seems then that there are a multitude of interactions within and between Weiss's levels. Hence it appears likely that it is difficult to provide an accurate T1 transformation and that T1 for generation 1 may well be different for T1 in generation 2 etc. As Lovtrup (1977) has argued in a multi-level system (like embryogenesis itself) any deviation at a lower level will result in larger deviations at higher levels. So over a large number of generations huge errors would be expected to occur.
Fig. 2 Paul Weiss's view of the organism as a multi-level hierarchical system (modified from Weiss, 1950).

Fig. 2 Paul Weiss's view of the organism as a multi-level hierarchical system (modified from Weiss, 1950).

page 124

T2 The transformational laws which operate to change form distributions within any generation represents a complex array of phenomena—the “laws of mating”, for example. What are these laws? In virtually all analyses mating is considered to occur at random! Patterns of migration are rarely considered. In fact Sermonti and Catastini (1984) have recently suggested that the “classic case” of natural selection, that of industrial melanism, is the result of a process of migration of moths, and the individual choice of environment by melanic and light-coloured moths.

T3 The ability to accurately infer the genotype distributions corresponding to the distribution of forms is dependent on an accurate knowledge of the T1 transformation. Indeed T3 is a reverse transformation of T1. Any errors in T1 will also result in T3 errors.

T4 The genetic rules of Mendel in a very real sense simply allow us to infer the movement of chromosomes at meiosis. A consequence of this movement is that certain predictable ratios of form arise in a limited number of cases. However as Lewontin (1983) has remarked many characters do not “Mendelise”. Also the phenomenon of meiotic drive will inevitably complicate this transformation in some organisms. So the plain fact is then that, in many cases population geneticists will encounter situations where genetic rules will not allow an accurate T4 transformation.

Milkman (1983) has argued that a central demand we must make of any population genetics theory is that “…it should be able to predict changes in allelic frequencies and phenotypic values…” This task is a difficult one to say the least. If organisms are indeed as multi-levelled as suggested by Weiss, accurate genotype/phenotype transformations will be difficult to perform. Hence any accurate prediction of changes of gene frequencies seems highly unlikely.

Finally in an interesting paper Kempthorne (1983) has recently critically reviewed the state of population genetics theory. After making some important criticisms Kempthorne agrees that “We have been quite unable, except under very limited assumptions, to develop theory of long-term changes, which is what we need if we are to apply the ideas to evolution.” He finally remarks “The overall task is extremely difficult but it has to be attacked.” This is so only if one considers that the whole populatior genetics approach is valid.


Historically population genetics has grown out of attempts to reconcile Darwinian views with Mendelism which was originally regarded as a competing theory (Bowler, 1983). This is reflected in Kempthorne's (1983) remark “… the aim of population genetics theory is to give an empirically validated theoretical basis for the process described in verbal terms by Darwin … “Hence as Hitching (1982) has pointed out, population geneticists consider evolution to be their specific domain. This is partly because commonly accepted definitions of evolution have been framed in neo-Darwinian terms. Evolution is almost exclusively regarded as changes in gene frequencies. Kempthorne (1983) remarks on this point, “The whole problem of evolution revolves around the production of new genetic types and the fitness of the genetic types in the population”. However, as is commonly agreed, even by neo-Darwinists, it is the organism's form which is important. This is surely a view with which most working biologists are sympathetic. The emphasis on genes by population geneticists is rationalised page 125 then by arguing that genes have some direct and causal relationship to form production, and population genetics theory then attempts to understand the transformational pathways which summarise this. However if these transformations cannot be made accurately then we cannot even use a “genetic shorthand” to represent form distributions. Indeed, I agree with Goodwin (1979) that neo-Darwinism, through its agent of population genetics, does not even address this problem.

In contrast however, if we see evolution as a change in form rather than a strictly genetic phenomenon, a different view arises. Genes are surely a component in such a system, but evolution is not a population problem, and most especially not a population genetics problem. It is a problem of the origin of form and therefore can only be understood in terms of the laws governing form transformation (Lambert and Hughes, 1984;). These laws are quite distinct from those regarding the origin of genetic variation.

Where then is population genetics left? I do not think that it is left as a purely historically interesting science. I think that it has a place but a far more modest one. Population genetics can be useful, for example, in the detection of morphologically very similar species (e.g. Lambert, 1982), and perhaps it has limited uses in aspects of human genetics. It is simply that population genetics is not a tool to investigate evolution. This essentially represents a third view (in the sense of Macbeth's (1976) third position) of evolution, the first being creationism and the second being neo-Darwinism. This view has many similarities with Hitching's (1982) “New Biology” where, it is argued the study of development will reveal fundamental laws of form and transformation (Goodwin, 1982), as opposed to the neo-Darwinian view that the study of genetic variation (usually populations of adult individuals) is sufficient to explain evolution.


Bowler, P. J. 1983: The Eclipse of Darwinism. The John Hopkins University Press, Baltimore.

Capra, F. 1982: The Turning Point: Science, Society and the Rising Culture. Fontana, London.

De Betham Anderson L. 1984: A Cytogenetic Analysis of the B Chromosomes of Phaulacridium marginale (Orthoptera: Acrididae). Unpublished MSc Thesis, University of Auckland.

Goodwin, B. C. 1979: On morphogenetic fields. Theoria to Theory 13: 109-114.

Goodwin, B. C. 1982: Evolution and development. J. theor. Biol. 97: 43-55.

Hewitt, G. 1983: The role of population genetics in our understanding of evolution. Tuatara 26(2): 73-75.

Hitching, F. 1982: The Neck of the Giraffe or where Darwin went wrong. Pan Books, London.

Hughes, A. J. and D. M. Lambert, 1984: Functionalism, Structuralism and “ways of seeing”. J. theor. Biol: in press.

Kempthorne, O. 1983: Evaluation of current population genetics theory. Amer. Zool., 23: 111-121.

Lambert, D. M. 1982: A population genetical study of the African mosquito Anopheles marshallii (Theobald). Evolution 37(3): 484-495.

Lambert, D. M. and A. J. Hughes. 1984: The misery of functionalism Rivista di Biologia: in press.

Lewontin, R. C. 1974: The Genetic Basis of Evolutionary Change. Columbia University Press, New York.

Lewontin, R. C. 1983: Gene, organism and environment. In Bendall, D. S. ed. Evolution from Molecules to Men. Cambridge University Press, Cambridge.

Lovtrup, S. 1977: The Phylogeny of the vertebrata. John Wiley and Sons, London.

Macbeth, N. 1976: A third position in the textbook controversy. The American Biology Teacher., 38(8): 495-496.

Miklos, G. L. G. and R. N. Nankivell. 1976: Telomeric Satellite DNA functions in regulating recombination. Chromosoma 56: 143-167.

Milkman, R. 1983: Population Genetic Theory: Another View. Amer. Zool., 23: 123-125.

Naranjo, T. and J. R. Lacadena. 1980: Interaction between wheat chromosomes and rye telomeric heterochromatin on meiotic pairing of chromosomes pair 1R of Rye in Wheat-Rye derivatives. Chromosoma 81: 249-261.

Saiff, E. I. and N. Macbeth. 1983: Population Genetics and Evolutionary Theory. Tuatara 26(2): 71-72.

Sermonti, G. and P. Catastini. 1984: On industrial melanism, Kettlewell's missing evidence. Rivista di Biologia 77: 35-52.

Suzuki, D. T., Griffiths, A. J. F. and R. C. Lewontin. 1981: An Introduction to Genetic Analysis W. H. Freeman and Company, San Francisco.

Weiss, P. 1950: Perspectives in the field of morphogenesis. Quart. Rev. Biol. 25: 177-198.

* Publication No. 7 from the Evolutionary Genetics Laboratory, University of Auckland. D. M. Lambert's research is funded by the Auckland University Research Committee.