On the Origins of New Forms of Life

6.6: Peripheral Isolates and Speedy Gradualism



(Continued from the previous page)

Peripheral Isolates. It is a key assumption of neo-Darwinian theory that mutations occur at random so that they are spread fairly evenly over time. An even pattern of mutation such as this does little or nothing to explain the abrupt appearance of new forms seen in the fossil record (as discussed on the previous page. If evolutionary change is to be abrupt, mutation must be concentrated in a single, brief interval of time. But conventional theory does offer explanations, however unsatisfactory they might be.

One commonly proposed mechanism supposes that rapid change can occur in "peripheral isolates."1 Here, a small sub-population of individuals is pictured as breeding in isolation from others of their own kind. Due to the small size of population, some variants (i.e., alleles) of certain genes are lost from the reproductive process by chance. The supposed result is the production of a new type of organism with a new set of traits distinguishing it from the old one. These new traits are generally imagined to be advantageous, permitting the new type to compete with and rapidly supplant preexisting types. Speaking of himself in the third person, Ernst Mayr (1982: 617), who first proposed the peripheral isolates scenario (Mayr 1963), comments that his explanation

was ignored by paleontologists until used by Eldredge and Gould (1972) [who, Mayr says … ] accepted Mayr's interpretation that such new species had originated somewhere in an isolate (peripheral or not) and were able to spread far and wide if they were successful.2 This interpretation of the 'introduction of new species' (as Lyell called it 150 years earlier) agrees well with the fossil record.

But no population treated as a species is actually known to have had its origin via peripheral isolation. Or, at the very least, if any such populations are known, they are far less numerous than the many forms treated as species now known to have been produced by stabilization processes. Alleged examples of populations produced in peripheral isolation are typically anecdotal. Alternative hypotheses accounting for their origins are not excluded on any logical basis. For example, the distinctive traits of an equatorial penguin might plausibly be explained as a matter of peripheral isolation. Nevertheless, other origin stories (e.g., migration, hybridization, environmental effects on phenotypic plasticity) might seem equally plausible. It seems that in all such cases the actual mode of origin is attributed to peripheral isolation, but is actually unknown and seems never to have been documented.

Moreover, logically speaking, we have no reason to suppose that peripheral isolation should be able to produce organisms that are genuinely new — Alleles may be lost from a population under such circumstances, but no new ones will be introduced because the only genetic changes that can occur are due to ordinary meiotic recombination within a chromoset. Even with artificial or natural selection, no new alleles would appear (because selection is an eliminative process, not a creative one). Unless a point mutation occurs in a gene — an exceedingly rare event — no new alleles can be introduced into a closed population. In particular, in the case of the peripheral isolates scenario, it is safe to assume no mutations whatsoever occur (since this process is supposed to be rapid, there is no time). If alleles are selected against and lost, the only possible result is a decrease in genetic variability: Any new set of alleles present in a supposedly new type of organism generated by such a process, then, would be a mere subset of those alleles already seen in the preexisting population. As Mayr himself says (1963: 393)

Many of the peripheral populations, particularly the more isolated ones, are established by a single fertilized female or a small group of founders which carry only a fraction of the total genetic variability of the species.

Therefore, there seems to be no real potential for the evolution of any new traits under such circumstances. Even if genetic divergence and significant morphological novelty could be produced in such peripheral populations, any organisms thus produced would almost certainly still be able to mate freely with individuals of the preexisting type. What, then, would prevent the preexisting population from simply reabsorbing the new population as soon as the two came back into contact? Moreover, the process described in the peripheral isolates scenario is an extreme form of inbreeding (the mating of closely related organisms). It is well known that inbreeding tends to bring out deleterious, or even lethal, traits. So it seems unlikely such a process would be likely to produce a superior new type of organism capable of displacing preexisting types.

tiger face
Peripheral isolates: How long would it take a small number of tigers breeding in isolation to become something other than tigers? Photo: Steven Bennett

Consideration of a concrete example suggests the shortcomings of the peripheral isolates scenario. Suppose ten caged tigers were allowed to breed. Suppose the number of tigers was kept small in every generation. How long would it be before those tigers suddenly became something other than tigers? How long would it be before they became physiologically incompatible in matings with normal tigers, so that interbreeding of the two types produced hybrids of low fertility? If, after its sudden appearance, this supposedly new type of organism were released from its cage, would it be likely to spread and quickly replace normal tigers? At the very least, it is fair to say that we have no reason to suppose the typical form treated as a species comes into being in such a manner.

Speedy Gradualism. Another attempt to resolve the contradiction between neo-Darwinian theory and the fossils is the "speedy-gradualism" argument: If change occurs rapidly during a single time period, but not for millions of years thereafter, the pattern might seem saltational. But what if this "short" period of change lasted 20,000 years? Wouldn't it be reasonable to simply change the meaning of "gradual" and say that while 20,000 years is not a million years, it's still a very long time, plenty of time for gradual selection and everyday genetic phenomena to take their course and create a new type of organism?

This argument was proposed as an explanation of one of the best-documented cases of saltation, the Turkana mollusks described by P. G. Williamson (1981). The speedy gradualists say some shift in the environment forced Williamson's mollusks to change, that the changes occurred "gradually" during a relatively brief (20,000-year) transition period, with the environment stabilizing thereafter, maintaining the new types unchanged for millions of years. But this is not the picture painted in the Origin. There, Darwin explicitly states his views:

natural selection can act only by taking advantage of slight successive variations; she [i.e., Nature] can never take a leap, but must advance by the shortest and slowest steps.3

In this kind of evolution, variation arises randomly over time, and is not concentrated in a single, brief interval. But the fact that fossil types appear to come into being abruptly is not the most serious difficulty confronting neo-Darwinism. As Williamson (1981) points out,

The principal problem [with the available paleontological data] is morphological stasis. A theory is only as good as its predictions, and conventional neo-Darwinism, which claims to be a comprehensive explanation of the evolutionary process, has failed to predict the widespread long-term morphological stasis now recognized as one of the most striking aspects of the fossil record.

Moreover, Williamson says the appearance of new types of mollusks was "initially accompanied by a major increase in phenotypic variance." But orthodox theory says individual variation does not increase in response to changes in the environment. On the other hand, an initial increase in variation actually can be the expectation under stabilization theory. In particular, in the case of recombinational stabilization, such variability is expected. The more distant the cross initiating the process, the higher the expected level of variation, as long as later-generation hybrids can be produced (Soliman 1992: 199). Later stages of the process are characterized by increasing uniformity because many of the variants produced are too infertile and/or inviable to meet the challenge of continued existence. They are displaced by the few that do. NEXT PAGE >>



1. Bush (1975); Mayr (1963); Mayr and Ashlock (1991).

2. See Gould (1980a: 183–184); Eldredge (1995: 119).

3. Darwin (1859: 194).

Most shared on Macroevolution.net:

Human Origins: Are we hybrids?

On the Origins of New Forms of Life

Mammalian Hybrids

Cat-rabbit Hybrids: Fact or fiction?

Famous Biologists

Dog-cow Hybrids

Georges Cuvier: A Biography

Prothero: A Rebuttal

Branches of Biology

Dog-fox Hybrids