Species vs. Time: In this example, a snail species, A, is isolated by a geographic barrier to form subspecies X and Y. Eventually, this geographic separation leads to genetic isolation, and two distinct new species, B and C, form. © AMNH
Is evolutionary change mostly dependent upon speciation, which occurs when one or more descendant species split from an ancestral one? Do most anatomical changes occur during these events? Or can species evolve as single lineages? And how long does it take the process of speciation to work in geological time?
Conventional wisdom: evolution is gradual and constant
The traditional view in evolutionary biology, which stems largely from Darwin's On the Origin of Species, is that adaptive change, or genetically based anatomical change, will tend to accumulate more or less regularly and gradually throughout a species' history. This assumption underlies the concept of the "molecular clock": that if genetic change on the whole accumulates at an steady rate, genetic "distance," or the amount of genetic difference between two species, should be directly proportional to the amount of time the two have been phylogenetically separate. According to this view, even when lineages divide, the slight differences between the newly separate species will tend to grow steadily and gradually increase over time [Figure 1].
Fossils tell a different story
The fossil record indicates otherwise. New species tend to appear in the fossil record already morphologically distinct from their closest relatives. And once established, species tend not to change significantly or permanently, which in the case of marine invertebrates can amount to a 5-million-year or even 10-million-year history without change.
The birth of a new theory: speciation and punctuated equilibria
In the late 1960s I was conducting my doctoral dissertation research on the evolution of a Middle Devonian trilobite called Phacops rana. Along with many other species of trilobites, mollusks, brachiopods, corals, and other invertebrates, the abundant Phacops rana inhabited the shallow-water inland seas that ran from the present-day Appalachian mountains as far west as Iowa. Many, including P. rana, were derived from species that migrated into North America from Europe and Africa when those continents collided with North America about 380 million years ago. North America lay astride the Equator in those days, and its shallow tropical seas literally teemed with life that has left a rich and dense fossil record. The fauna persisted for some 6 million to 8 million years.
Phacops rana: A fully articulated trilobite and a well-preserved trilobite head from Dr. Eldredge's collection. © AMNH
I had been trained to believe that evolution proceeds gradually, as depicted in Figure 1. Collecting all over the Midwest and up and down the Appalachian mountain chain over several summers, I was dismayed to detect no obvious evolutionary change across 6 million years and thousands of miles. Finally, I stumbled on some changes: very slight but persistent, like the consistent differences Darwin saw in the mockingbirds on different islands in the Galapagos. Back in the laboratory at the American Museum of Natural History, where I was privileged to work as a Columbia University graduate student, I carefully removed the rocky matrix covering my specimens and began to measure and examine the trilobites.
Having learned from another paleontologist that the lenses of the eyes of these kinds of trilobites are arranged into vertical columns holding anywhere from one to more than 15 large, bulging lenses, I counted the lenses on each specimen. One day I suddenly realized that samples from different localities—different places in both time and space—differed mainly in the number of those vertical columns of lenses. Babies would add columns, then stop at the "mature" number: Some stopped at 18; many stopped at 17; and a few stopped at 15.
Then I had an another idea: I simply plotted these numbers on a sequence of maps of the United States, moving from the earliest samples to progressively younger ones. I was delighted to see that a clear pattern emerged, one that reflected NOT gradual change through time, but geographic speciation. Like their European/African ancestors, the earliest samples, from the eastern part of the range, had 18 columns of lenses in the eye.
Interestingly, an early sample from upstate New York seemed to vary; some had 18 and some 17. With one exception, all samples up and down the Appalachians had 17, while all the samples from the shallow limey seas that once covered Ohio, Michigan, and Iowa still had the ancestral 18 columns. It was beginning to look as if "Phacops rana" included several species, closely similar yet distinct.
Evolution of the Trilobite: This sequence shows how the continental seas of North America waxed and waned over a seven million year interval during the Devonian Period (400 million years ago). The white circles indicate some of the more important localities where the trilobite, Phacops rana, were found by Dr. Eldredge. © AMNH
Two million years later, those Midwestern shallow seas dried up. When marine conditions were once again restored, the 18-column species of Phacops was gone. (Presumably a victim of extinction, they never reappeared in the fossil record.) The 17-column species took its place, spreading westward from the Appalachian area where the seas had NOT dried up. A similar event happened some 2 million years later when a 15-column species arose, again probably in the eastern part of the range. Then it, too, spread westward, and the 17-column form became extinct.
"Aha!" I said to myself: This is a case of allopatric speciation. Lineages are splitting into descendant species, with the ancestral species persisting alongside the new, daughter species. And though it is difficult to measure small increments of time in the fossil record, my data seemed to suggest that speciation could be quite rapid—taking, perhaps, anywhere from five to 50,000 years. Later, the daughter species spread, and in the case of Phacops, at least, eventually replaced the ancestral species. But what remained surprising were the long periods of time with little or no change at all. The 17-column species, for example, seemed to have lasted unchanged for nearly 5 million years!
I wrote these results up in a paper in 1971 called "The allopatric model and phylogeny in Paleozoic invertebrates," concluding that speciation has played a critically important role in evolution since the dawn of time. It was beginning to look like evolution is mostly correlated with speciation events.
But what about this unexpected stability? In the following year (1972) I published a longer paper with Stephen Jay Gould, with whom I had worked in graduate school. (Steve was two years ahead of me and was already teaching at Harvard.) We dubbed the phenomenon of non-change "stasis" and called the entire pattern of stasis + speciation "punctuated equilibria" [Figure 2].
Historical connections: the importance of geology in evolution
Gould and I were not the first to point out that patterns of stasis and change in the fossil record disagree with the standard image of evolution as slow, steady, gradual change. Darwin himself, in his earliest notebooks, thought that speciation must come in sudden "jumps." These early "saltational" views were based in part on his experiences collecting fossils, but also on his observations of recent species. Noting, for example, that the two species of ostrich-like rheas in South America meet only in one place and do not seem to interbreed, Darwin initially imagined that the "lesser rhea" probably was derived from the larger "common rhea" in one swift evolutionary jump. Only later in his life, based on his experiences with variation and his vision of how natural selection works in the natural world, did Darwin reject saltationism for the "phyletic gradualism" [Figure 1] that has come down to us as the standard image of evolution at work.
A number of biologists after Darwin thought that geography and isolation underlay speciation, and indeed the entire evolutionary process. The 19th-century German biologist G.J. Romanes, for example, felt that without isolation, evolution would not be possible. Another German biologist, Moritz Wagner, who corresponded at length with Charles Darwin, independently derived the notion of punctuated equilibria in the 1870s, apparently on purely theoretical grounds. Wagner saw evolution as occurring mainly in small populations isolated geographically from the large, well-established populations of the parental species. Such isolated populations, he felt, would undergo rapid evolution; if the fledgling species survived, it might grow in numbers and expand its range—in which case, Wagner surmised, further evolution of that species would grind to a halt?
Yet the importance of geographic variation, isolation, and the emergence of new species was a largely muted, secondary theme in evolutionary biology until the 1930s, when the geneticist Theodosius Dobzhansky and the systematist Ernst Mayr propounded the "biological species concept" and re-established the importance of geographic, or allopatric, speciation in the evolutionary process. But even their concept differed little from the gradual divergence model seen in Figure 1.
Explaining stasis: the biological evidence
Perceived as radically anti-Darwinian, stasis was the most controversial component of punctuated equilibria. Nevertheless, over the 35 years or so since the original publications on punctuated equilibria, stasis has been increasingly accepted as a real, unpredicted phenomenon that must be explained by evolutionary biologists. Species may exhibit lots of molecular variation, yet their morphology is consistently found to vary only within certain limits and to remain essentially stable for millions of years. How could this be?
In the late 1990s, biologist John N. Thompson (see Week 3, Essay 1) and I convened a team of geneticists, ecologists, and paleontologists to study the problem of stasis from an interdisciplinary perspective. We concluded that the most likely cause of stasis is the simple fact that most species are distributed as local populations inhabiting a variety of physical and biotic environments. Especially in species with large-scale distributions (e.g. over a half continent or more), local populations have to contend with different environments (predators, food resources, temperature, and rainfall parameters, etc.). Since species are not homogenous mega-populations, it is statistically unlikely that natural selection will systematically modify an entire species in any one direction, especially over long periods of geological time. Stasis should have arisen as a simple prediction from population genetics as long ago as the 1930s, but its apparent anti-Darwinian nature prevented that from happening.
Despite the fact that the history of life records a single evolutionary process, findings from the fossil record are sometimes difficult to reconcile with genetic evidence. Recently, geneticist Mark Pagel and colleagues found that, on average, 22 percent of the DNA change in animals, plants and fungi can be attributed to punctuational evolution while the rest accumulates gradually. Their results call into question the "molecular clock" assumption of constant, gradual change through time. They also found that there is little evidence for stasis at the molecular level. In contrast, paleontological data suggest that far more than 22 percent of morphological evolutionary change is punctuational, and that stasis is a very common phenomenon. Since molecular data for fossil examples is unavailable, there's no direct way to compare anatomical and morphological data.
Do genes show punctuated equilibrium at work?
Though the molecular data support the paleontological notion of punctuated equilibria to some degree, the differences in results need to be reconciled. This is the hypothesis about the molecular data that requires further testing: Changes in genes that code for protein products—and thus produce morphology—will be found to be disproportionately concentrated at the splitting (speciation) events, while neutral, non-coding changes not subject to natural selection will simply accumulate in a clocklike fashion through time. If further studies find that to be the case, as I anticipate, not only will most coding DNA changes be focused in splitting events, but they will also be locked into stasis in the interim—evidence of punctuated equilibrium at the molecular level.
The fossil record suggests not only that most evolutionary change occurs in speciation events, but also that speciation events themselves are non-randomly clustered in space and especially in time. In the history of life, speciation events seem to follow episodes of environmental disruption, especially when sufficiently large-scale events drive many pre-existing species to extinction. Even before Darwin's day, extinction was considered a reality, predominantly caused by physical events. The half-dozen or so truly global mass extinctions have removed entire groups (e.g. terrestrial dinosaurs and marine ammonoids at the end of the Cretaceous), followed by evolutionary bursts of other groups (e.g. mammals and nautiloids in the Tertiary), usually after a lag of several million years.
Gradualism vs. Punctuated Equilibria: Here are two models of speciation: gradualism, where a species slowly changes over time; and punctuated equilibrium, where morphological changes occur relatively rapidly. © AMNH
Extinctions and evolutionary rebounds: "turnovers" large and small
The same thing happens more regularly on a regional basis: Entire faunas and floras are often found to be locked in stasis, where the individual species show little or no evolutionary change through time. Then an environmental perturbation such as climate change or an asteroid impact disrupts the ecosystems, and if severe enough, drives many component species to extinction at more or less the same time. They are eventually replaced, in part by surviving species from elsewhere—as in the case of the 17-column species of Phacops migrating into vacated habitat in the American Midwest after the ancestral species had disappeared. But sometimes the reconstituted ecosystems are populated by newly evolved species that arose in isolation in the disturbed environment over several hundred thousand years or so. Paleontologist Elisabeth S. Vrba calls these associations of large-scale extinction events of species with subsequent evolutionary bursts "turnover pulses."
Darwinian theory meets the challenge
Darwin added a note to his unpublished 1844 essay on evolution: "Better begin with this: if species really, after catastrophes, created in showers over world, my theory false." Aware that geologists were talking about such turnovers, Darwin felt that they threatened his vision of natural selection slowly modifying species through time.
We know now that most morphological evolution occurs relatively rapidly in conjunction with speciation, and that most speciation events are concentrated into turnover events. Yet Darwin's theory of evolution through natural selection remains essentially sound. All we need do is add concepts of isolation, speciation, and extinction; and understand the conservative action of natural selection producing stasis in stable ecological regimes, to grasp the actual context in which natural selection produces evolutionary change in the history of life.