The Powers of
Natural Selection

The most famous case of fast directional selection is the phenomenon called industrial melanism (from the Greek word for ‘black’). Since the mid-nineteenth century, over a hundred species of hitherto pale moths in Britain have become predominantly black, not through change in the individuals but through spread in the populations of a black mutant type controlled by a major gene, the black type being dominant (that is, the heterozygote is black). (Ford, 1975, on which this section is mainly based.) This has also occurred in other industrialised regions, on the Continent and in North America. By 1957 more than 40 moth species were involved in and around the large towns on the St. Lawrence, and, over 100 species around Pittsburg. (The reasons for the dominance will be considered in a later section.)

Our understanding of the phenomenon has been made possible by the extensive studies of Kettlewell on the Peppered Moth, Biston betularia. The first black specimen was found in Manchester in 1848; by 1895 about 98% of the population there were black, indicating a selective advantage of blackness of 50%.

The moth spends some time resting on the lichen-covered bark of deciduous trees. Its normal pale form is so beautifully camouflaged on the lichen that it is extremely hard to detect in a photograph. Industrial pollution kills the lichens and blackens the trees with soot. The pale form is now very conspicuous, and the black form – extremely conspicuous on lichen – is now quite well concealed. Conspicuous forms are promptly caught by several species of birds, as was filmed by Tinbergen when he accompanied Kettlewell on one of his field trips.

Kettlewell proved conclusively that the spread of melanism in polluted areas is due to selection by bird predation. He released large numbers of black and pale forms (Biston betularia carbonaria and Biston betularia betularia) in a polluted area near Birmingham and in an unpolluted wood in Dorset. He then recorded the number of specimens of each type recaptured by means of traps with lights (which attract the moths) or by attracting them to caged females (only’ males were used). Obviously the fewer moths recovered the more have been caught by birds. The results are shown in Table 1, simplified from Kettlewell (1955 and 1956). About twice as many pale as black types were recovered in Dorset, and the reverse in Birmingham. Kettlewell proved that the two types do not differ in longevity, attraction to light and females, or tendency to wander away from their territory. His results are therefore conclusive proof of the nature of the selection.

There is a third type of Peppered Moth, called insularia, intermediate in tone between carbonaria and betularia. It is not, as one might think, the heterozygote (which is black, since black is dominant), but due to another allele at the same locus. In Kettlewell’s experiment it gave intermediate recapture values in both regions. It seems to occur in moderate frequencies in still only mildly polluted areas, where some kinds of lichens are still present. (Sheppard, 1975 as well as Ford, 1975.) It may also be mentioned that carbonaria has itself become more completely black since the 19th century, presumably by the selection of modifier polygenes to improve its camouflage on soot.

The importance of melanism is emphasized by the case of the Rosy Minor Moth, Miana literosa, which, having no melanic type, became extinct in the Sheffield area. ‘Only in the mid-l940s was it able to recolonize the centre of the city through a newly arisen melanic.’ (Berry, 1990)

If there is a gradual change in some important variable over a stretch of country, natural selection will produce a corresponding gradual change or gradient in a character of an animal species adapted to the value of this variable. This is because there is naturally a gradient in the relevant selective advantage or disadvantage, but the gradient in the character is smoothed by the flow of genes between links in the chain owing to some intermating between neighbouring populations. But the two ends of the chain may show a considerable difference in the character. This is called a cline (from a Greek word for ‘bend’ or ‘slope’). A gradient in the proportion of two types is called a morph-ratio cline. (Sheppard, 1975; the terms were coined by Sir Julian Huxley.) There is such a morph-ratio cline in the relative frequencies of Peppered Moth carbonaria and betularia types between the highly polluted Mersey area and pollution-free North Wales, the frequencies of melanics gradually declining over this stretch of country. (Sheppard, 1975, Berry, 1990.) For instance, the frequency of carbonaria is 97.3% in central Liverpool, 11.8% at Colwyn, 50 miles from Liverpool, and 3.1% at Bangor, 70 miles away.

When smokeless zones are introduced, reducing pollution and permitting the return of lichens, the frequency of melanics declines, as one would expect. This has been observed at Caldy on the Wirral and at Didsbury near Manchester. (Sheppard, 1975, Berry, 1990, as well as Ford, 1975.)

Recapture experiments enabled Bishop to calculate selection strengths at various points along the Mersey-North-Wales cline. He found that frequencies at the unpolluted end of the cline were higher than would be expected from these selection strength values, which were obviously based on the selection due to conspicuousness and bird predation. It is also generally found that the melanics never reach 100% in polluted areas, and the pale forms never recover 100% frequencies in smokeless zones where pollution is reduced. (Sheppard, 1975, Berry, 1990.) The reason for this is that the heterozygote is physiologically fitter than either homozygote, as first discovered by Ford in 1940. This prevents total elimination of either allele, just as in the warfarin-resistant rats. I shall return to this in a later section.

Because of the predominant action of Westerly winds in Britain, pollution spreads Eastwards from industrial areas. Therefore high melanic frequencies are found in non-industrial rural areas East of industrial areas, for instance in East Anglia. I compared a map by Kettlewell of melanic frequencies in Britain with maps of the infrequency of epiphytic lichens and of high frequencies of human bronchitis and cancer of the trachea, lung and bronchus. (Russell, 1978, 1993.) All these maps showed good agreement.