Darwin's Doubt

by Stephen C. Meyer

  As a consequence of this theory, neither Gould nor Eldredge expected to find a wealth of transitional intermediate forms in the fossil record. In their view, the main periods of biological innovation simply occurred too rapidly to leave many fossil intermediates behind.6

  Gould and Eldredge sought to explain the occurrence of the rapid periods7 of change (i.e., the punctuations) as the byproduct of different kinds of evolutionary mechanisms or processes of change. They proposed, first, a mechanism called “allopatric speciation” to explain the rapid generation of new species. Gould, Eldredge, and another early advocate of punctuated equilibrium named Steven Stanley, a paleontologist at Johns Hopkins University, also proposed that natural selection operated on higher levels. Rather than natural selection favoring the fittest individual organisms within a species—as it does in classical Darwinism and neo-Darwinism—these paleontologists proposed that it often selected the most fit species among a group of competing species. Because they thought that speciation occurred more rapidly, and because they thought that natural selection acted on whole species and not just individual organisms, the advocates of punctuated equilibrium theorized that morphological change typically occurs in larger, more discrete jumps than Darwin first envisioned.

  FIGURE 7.1

  Figure 7.1a (left): Stephen Jay Gould. Courtesy Getty Images. Figure 7.1b (right): Niles Eldredge. Copyright © Julian Dufort 2011. Used with permission.

  Thus, in one sense, the theory of punctuated equilibrium, like the artifact hypothesis, sought to explain the absence of the transitional intermediate forms that were expected based on Darwin’s theory. By repudiating Darwinian gradualism, the advocates of punctuated equilibrium sought to account for the absence of transitional forms in the fossil record apart from the artifact hypothesis or, at best, using what they imagined as a more modest version of it. But in repudiating Darwinian gradualism, punctuated equilibrium also represented a radically different view of the pace and mode of evolution—a new theory of evolution that purported to identify a new mechanism of evolutionary change. As historian of science David Sepkoski explains, “Gould and Eldredge proposed a radical revision of this standard [neo-Darwinian] narrative. They argued that the pattern of evolutionary history really was composed of fits and starts, consisting of long periods of evolutionary stasis (or ‘equilibrium’) ‘punctuated’ by shorter periods of rapid speciation.”8

  During the 1970s and 1980s, the theory of punctuated equilibrium, or “punk eek” as it is affectionately known, generated both intense scientific debate and extensive media coverage.9 Critics called the model “evolution by jerks,” leading Gould to reply that proponents of gradualism were offering “evolution by creeps.”10 Though initially Eldredge played more of a role in formulating the theory, Stephen Jay Gould emerged as the leading spokesman for it. As a result of his advocacy of the theory as well as his popular science writing, Gould achieved immense celebrity—celebrity that has in turn secured an enduring place for punctuated equilibrium in scientific awareness.

  So what has become of this bold scientific proposal? Does punctuated equilibrium solve the problems that traditional neo-Darwinism does not? Does it help explain the Cambrian explosion and the missing fossil intermediates that render it so mysterious?

  Wanted: Fast Engine

  Once they decided to take the fossil record at face value, the question for Gould and Eldredge was obvious: What could generate evolutionary change so rapidly? To explain the short bursts or punctuations, Gould and Eldredge proposed a mechanism of rapid speciation to which Stanley added (with their agreement) a new understanding of the mechanism of natural selection.11

  Whereas the neo-Darwinian mechanism of natural selection acting on random mutations necessarily acts slowly and gradually, Gould and Eldredge invoked a process called “allopatric speciation” to explain how new species might arise quickly. The prefix allo means “other” or “different,” and the suffix patric means “father.” Thus, allopatric speciation refers to processes that generate new species from separate parent (or “father”) populations. Allopatric speciation typically occurs when part of a population of organisms becomes geographically isolated—perhaps by the emergence of a mountain range or the shifting of a river’s course—from a larger parent population and then a daughter population changes in response to differing environmental pressures.

  Gould and Eldredge drew on insights from population genetics to explain why new genetic traits were more likely to spread and establish themselves within these smaller subpopulations. Population genetics, a subject to which I’ll return in Chapter 12, describes the processes by which genetic traits change and become fixed in a population of organisms. It teaches that in typically large populations of organisms, it is difficult for a newly arising genetic trait to spread throughout an entire population. Yet for any evolutionary change to occur in a population, new genetic traits must become widespread, or “fixed,” by a process called “fixation.”

  In smaller populations, however, the probability of a newly arising trait becoming fixed is much higher, since the new trait needs to spread to fewer organisms. By way of illustration, consider a bag containing fifty red and fifty blue marbles. Suppose that, by removing individual marbles at random, you seek to change the “population” of mixed-color marbles to one in which all the marbles are red. To produce a completely red “species,” we must generate a population in which all the blue marbles have been eliminated. If someone randomly picks half of the one hundred marbles out of the bag, it is extremely improbable that all of the marbles thus eliminated will be of just one color. Indeed, there is less than 1 chance in 1030 that all of the marbles selected for removal will be blue.12 Conversely, there is an extremely high probability that the remaining batch will still include both blue and red marbles.

  In a much smaller group of, say, eight marbles, evenly divided between four red and four blue, the probability of selecting four blue marbles at random and leaving only red marbles, though unlikely, is not prohibitively small. There’s now a much higher chance—1 in 70—that the remaining population of marbles will be all red.13 By starting with a smaller number of marbles, the probability that random selection will result in a population of uniform color is much higher. In a similar way, the probability of fixing a genetic trait in a population of organisms decreases exponentially with the size of the population.

  In formulating punctuated equilibrium, Gould realized that new species would inevitably have to arise in smaller populations, where random processes could have a greater chance of fixing traits. Prominent among those random processes is one called genetic drift. This occurs when genetic changes spread or disappear randomly through a population, without regard for their effect on survival and reproduction.

  In Gould and Eldredge’s view, allopatric speciation helped to explain how evolution could occur in larger, more discrete jumps than Darwinian gradualism predicts (see Fig. 7.2). As allopatric speciation occurs, it can generate what Gould and Eldredge conceived as sibling or offspring species. They thought that the processes that drive these speciation events occur relatively quickly in smaller populations, thus helping to explain the sudden jumps in the fossil record. As they put it: “Small numbers and rapid evolution virtually preclude the preservation of speciation events in the fossil record.”14 As they envisioned the evolutionary process working, the branches on the tree of life would split off so abruptly that they would appear as virtually “horizontal” lines, producing sudden discontinuities in the fossil record and therefore fewer fossilized intermediates. Eldredge and Gould explained it this way: “The theory of allopatric (or geographic) speciation suggests a different interpretation of paleontological data. If new species arise very rapidly in small, peripherally isolated populations, then the expectation of insensibly graded fossils is a chimera. A new species does not evolve in the area of its ancestors; it does not arise from the slow transformation of all its forbears.” Thus, they concluded, “Many breaks in the fossil record ar
e real.”15

  FIGURE 7.2

  Two views of the history of life. Figure 7.2a (left): The traditional Darwinian picture showing slow, gradual change. Figure 7.2b (right): The history of life as depicted by the theory of punctuated equilibrium showing rapid speciation.

  Gould, Eldredge, and Stanley thought that members of these sibling or offspring species would, subsequent to their origin by allopatric speciation, compete against each other for resources and survival, just as, in neo-Darwinism, individual organisms or siblings may compete to survive and reproduce within a population. In their view, if members of one species succeed over another because of some selective advantages they possess, then that species will survive and predominate, passing on its traits. This process of interspecies or interpopulation competition (as opposed to intraspecies competition) Gould called “species selection.”16

  As Gould himself explained: “I propose, as the central proposition of macroevolution, that species play the same role of fundamental individual that organisms assume in microevolution. Species represent the basic units in theories and mechanisms of macroevolutionary change.”17 Since natural selection then would act upon large differences in overall biological form—differences between whole species as opposed to individuals within species—evolutionary change would take place in bigger, more discrete jumps.18

  Gould and Eldredge thus did not expect the fossil record to document many intermediates. Instead, they thought the “gaps” in the fossil record were “the logical and expected result of the allopatric model of speciation” as well as the closely related mechanism of species selection.19 Species selection made “the species” the unit of selection; the allopatric model of speciation asserted that new species quickly arise from smaller populations of organisms. Both mechanisms implied that fewer fossil intermediates would be preserved. According to punctuated equilibrium, the long “missing” transitional intermediates are not missing after all. In the process of species selection, the species, rather than the individual organism, competes for survival and thus becomes—in the jargon of evolutionary biology—the main “unit of selection” in macroevolution.

  “Punk Eek” and the Fossil Record

  Eldredge and Gould devised the theory of punctuated equilibrium to eliminate the conflict between the fossil record and evolutionary theory. Nevertheless, punctuated equilibrium has its own problems accounting for the fossil record. In particular, the pattern of fossil appearance in the Cambrian period is inconsistent with both the way in which punctuated equilibrium depicts the history of life and with the idea that allopatric speciation and species selection are responsible for that pattern. There are several reasons for this.

  First, the top-down pattern of appearance of Cambrian animal forms that we saw in Chapter 2 contradicts punctuated equilibrium’s depiction of the history of life almost as much as it does the Darwinian picture (see Fig. 2.11). Recall that Darwin thought that the first representatives of the higher taxonomic categories emerged after the first appearance of representatives of each of the lower level taxa—that small differences distinguishing, for example, one species from another should gradually accumulate until they produced organisms different enough to be classified, first, as different genera, then, as different families, and eventually as different orders, classes, and so on. Instead, the first Cambrian animal forms are different enough from each other to justify classifying them as separate classes, subphyla, and phyla from their first appearance in the fossil record (see Fig. 7.3).

  This pattern creates an acute difficulty for the theory of punctuated equilibrium., First, due to the action of allopatric speciation and species selection, advocates of punctuated equilibrium envision morphological change (represented as horizontal distance in Figure 7.2) arising in larger, more discontinuous increments of change. Nevertheless, like neo-Darwinists, they too see phyla-level differences arising from the “bottom up,” starting with lower level taxonomic differences—albeit occurring in increments involving whole new species rather than individuals or varieties within species. Indeed, according to the theory of punctuated equilibrium, allopatric speciation first produces new species in smaller geographically isolated populations. For representatives of higher taxonomic categories to arise, these new species must accumulate new traits and evolve further. For this reason, punctuated equilibrium also expects small-scale diversity and differentiation of new species to precede the emergence of larger-scale morphological disparity and taxonomic differences. It also expects a “bottom-up” rather than a “top-down” pattern of appearance (see Fig. 7.3).

  Second, for species selection to produce many new species, such as those that arise in the Cambrian explosion, a large pool of different species must first exist. The Precambrian fossil record does not document, however, the existence of such a large and diverse pool of competing Precambrian species upon which species selection (via allopatric speciation) might operate.

  Paleontologists Douglas Erwin and James Valentine exposed this problem in 1987 in a seminal paper titled “Interpreting Great Developmental Experiments: The Fossil Record.”20 They questioned the ability of both of the main evolutionary theories of the time—punctuated equilibrium and neo-Darwinism—to explain the pattern of fossil appearance in the Precambrian–Cambrian fossil record.21 Clearly, neo-Darwinism does not explain this pattern. But, as Valentine and Erwin argue, neither does punctuated equilibrium. As they conclude, the mechanism of species selection requires a large pool of species upon which to act. Thus, Valentine and Erwin conclude: “The probability that species selection is a general solution to the origin of higher taxa is not great.”22

  FIGURE 7.3

  This figure shows that the theory of punctuated equilibrium (on left), like neo-Darwinism, anticipates small-scale diversity preceding large-scale disparity in form in contrast to the pattern in the fossil record (shown on right). Punctuated equilibrium also anticipates a “bottom-up,” rather than a “top-down,” pattern.

  The late-Precambrian and Cambrian fossil records present another difficulty for punctuated equilibrium. Though Gould and Eldredge envisioned new traits becoming fixed in small isolated populations where speciation eventually occurs, they envisioned these traits first arising during periods of stasis in the large populations from which the smaller populations later separated. Gould realized that only stable large populations would afford enough opportunities for mutations to generate the new traits that macroevolution requires.23 At the same time, he recognized that these new traits would have a far greater chance of being fixed into small, isolated populations where the random loss of some traits makes the fixation of others more likely (recall the marble example). By relying on large populations to generate new traits and small populations to fix them throughout a population, Gould wanted to provide both a plausible (if finely tuned) mechanism to explain both macroevolutionary change and the absence of fossil intermediates.24 The late University of Chicago paleontologist Thomas J. M. Schopf described the balance this way, under punctuated equilibrium, evolution proceeds “in populations large enough to be reasonably variable, but small enough to permit large changes in gene frequencies due to random drift.”25

  But by relying on the accumulation of new traits within large parent populations, Gould undercut his own rationale for concluding that the fossil record should not preserve many intermediate forms. The reason for this is obvious: if novel genetic traits arise and spread within a large population of organisms, they are more likely to leave behind fossil evidence of their existence. Organisms with new and unique combinations or mosaics of traits represent nothing less than new forms of life. Thus, the process by which Gould envisions new genetic traits arising in large populations implies that new forms of life—some presumably transitional to other forms—should be preserved in the fossil record. Yet the Precambrian fossil record fails to preserve such a wealth of biological experiments during the long periods of relative stability in large populations that Gould’s theory envisions.

  T
he Testimony of Statistical Paleontology

  Studies in statistical paleontology have raised additional questions about whether the fossil record documents enough transitional intermediates to render punctuated equilibrium credible. In Chapter 3, I discussed the work of the statistical paleontologist Michael Foote. Foote used sampling theory to argue that the fossil record provides a reasonably complete picture of the forms of life that have existed on earth and to suggest that paleontologists are unlikely to find the many intermediate forms that neo-Darwinian theory requires.

  Foote has also analyzed the question of whether the fossil record documents the number of intermediate forms that punctuated equilibrium requires. His answer: it depends.

  Foote notes that whether a particular version of evolutionary theory can account for patterns in the fossil record depends upon the kind of mechanism of change it invokes. Neo-Darwinism relies upon a slow, gradually acting mechanism of change and, thus, has difficulty accounting for evidence of sudden appearance. Foote analyzes whether the number of proposed transitional species in the fossil record is consistent with punctuated equilibrium and concludes that it depends upon how fast the mechanisms upon which it relies can generate new forms of life. Though its proponents envision the evolution of new forms of life arising more abruptly than do the neo-Darwinists, they would still expect that the fossil record should have preserved some transitional fossils. Does the fossil record preserve even the relatively few intermediate forms that the theory of punctuated equilibrium implies that it should?

  To answer that question, Foote developed a statistical method of testing the adequacy of different evolutionary models against several variables.26 He observed that for punctuated equilibrium to succeed as an explanation for the data of the fossil record, it needs a mechanism capable of producing major evolutionary change quickly, because only such fast-acting change could account for the relative paucity of transitional forms in the fossil record. As Foote explained (writing with Gould in fact), the adequacy of punctuated equilibrium as an account of the fossil record depends upon the existence of a mechanism “of unusual speed and flexibility.”27