The Powers of Natural Selection

Directional selection is of course the agent of evolutionary change, the simplest situation being the favouring and therefore spread in the population of one extreme form at the expense of other forms. Table 1 shows a number of examples of this collected by Berry (1990) in which the strength of selection has been estimated. As in the Table for stabilising selection (Table 1, Part 3) the very considerable strengths again show the power of natural selection.

In the course of evolution, both stabilising and directional selection have played parts. In very stable environments, notably in the sea, stabilising selection can maintain a group of animals with little or no change over very long periods; in rapidly changing environments, there may be correspondingly rapid changes in anatomy, physiology and way of life, due to directional selection. These great differences in rate of change were already noted by Darwin (1951 – sixth edition of 1872). He correctly observed that “the productions of the land seem to have changed at a quicker rate than those of the sea”. The modern Coelacanth fish Latimeria “is very similar in shape to Macropoma of the Cretaceous” (Kirkaldy, 1967) – c.140-65 million years BC. The tadpole shrimp Triops cancriformis has changed hardly at all since the Jurassic (Schilthuizen, 2001) – c.200-140 million years BC. Darwin himself noted that “the Silurian Lingula differs but little from the living species of this genus” of Brachiopod; but we can go even farther back – “the living Lingula is almost identical with Lingulella of the Cambrian rocks” (Kirkaldy, 1967) – c.550-500 million years BC. Contrast these cases with the fantastic changes in the phyletic line leading to Homo sapiens from the earliest vertebrates in the Ordovician – c.500-430 million years BC.

What applies to whole animals also applies to characters, for instance to proteins. Some proteins have virtually unchanging functions in the most varied animals, and cannot have many of their amino acid residues altered without disturbance of these functions. Stabilising selection will limit and slow down change in such proteins. Others can readily accept such changes to fit in with changing physiologies. Cytochrome c is a conservative protein, so to speak, the fibrinopeptides are readily changing. The time required for a 1% change in aminoacid residues is 1.1 million years for the fibrinopeptides, and 20 million years for cytochrome c (Creighton, 1984).

The whole matter of evolutionary rates has been thoroughly dealt with by the great palaeontologist G.G. Simpson (1944). Here I shall only note that, as Ford observed in the opening quotation of my first part, when change occurs under directional selection it can be very much faster than was supposed in Darwin’s day and for long afterwards. Darwin himself (1951–1872), when he considered the evolution under selection of such a very complex organ as the vertebrate eye, assumed it would take “millions of years”. It has recently been estimated that “with the most pessimistic assumptions at every stage, the shift from a flat patch of light-sensitive cells to an advanced eye could be achieved in a few hundred thousand years”. (Manning, 2001).

Considerable directional changes can of course be achieved far faster than this. The Ashkenazi “Jews” suffer from thirteen unusual genetic disorders. (Milunsky, 1977). Koestler (1977) has shown that these people are mostly descended not from Jews but from Khazar proselytes, though admittedly some people of Jewish ancestry had migrated to the Khazar Empire. The Khazars, of Central Asian origin, first appear as a distinct people in the 5th century AD, so they must have acquired their disorder-laden gene pool (which must have had some compensating advantages, at least in heterozygotes) in the fifteen centuries since then.

A study of heads was made in populations of Poland, where a “relatively high degree of ethnic continuity and stability” makes it “practically certain” that any change in head form has not been due to “population mixture” (Bielicki and Welon, 1971). In the seven centuries since AD 1300 the cephalic index (degree of round-headedness as opposed to long-headedness) has risen by ten points under directional selection, “a very great shift”.