Lenski's bacterial populations generate large numbers of mutations (some 106 every day in each population, by the estimate of Elena et al., 1996). But the step-dynamics revealed by the finer scheme of sampling — a pattern “predicted ... by a simple model in which successive beneficial mutations sweep through an evolving population by natural selection” (Elena et al., 1996) — presumably occur for two reasons: first, the well-known exponential principle, however intuitively paradoxical for most people untrained in science, that “many generations are required for the beneficial allele to reach a frequency at which it has an appreciable effect on mean fitness, but then relatively few generations are required for that allele to become numerically dominant” (op. cit.); second, and probably more important, the extreme rarity of favorable variants amidst the daily plethora of mutations, leading to “a substantial waiting period before a beneficial mutation even occurs” (op. cit.). Thus, at the proper scale (for resolving the causal mechanism) of sampling every 100 generations, the plateaux of stasis mark the waiting times between favorable sweeps, while punctuations express the sweep itself.
Finally — and to place some substantial empirical weight upon the keystone of my argument for the potential generality of punctuational change — Lenski and his colleagues have greatly increased our understanding of evolution by developing an artificial (in the best and fully positive sense of the word) experimental system purposely reduced to an absolute “bare bones” of Darwinian causal minimalism. With an identical genetic starting point for each replicate, an asexual clonal system that permits no genetic exchange among cells, and an unchanged environment (the regimen of daily transfer by controlled dilutions into a constant growth medium), this experiment leaves only two factors free to work and vary as potential agents of evolutionary change: the paired and essential Darwinian components of new mutations as raw material, and shaping by natural selection among cells that vary as a consequence of these mutations. The fact that punctuational dynamics prevail in this first truly adequate experiment in pure Darwinian minimalism must at least evoke a suspicion — even among biologists who, by custom and faute de mieux, have never questioned gradualism — that this episodic mode might be expressing something important about the general [Page 936] nature of change itself across the varying scales of nature's evolutionary construction.
PUNCTUATION ABOVE THE SPECIES LEVEL. Punctuated equilibrium stands on an “isthmus of a middle state” (to quote Alexander Pope out of specific context, but in proper structural analogy, see page 680) — a speciational bridge linking the microevolutionary history of discrete populations with the macroevolutionary waxing and waning of clades through geological time. I believe that the prevalence of punctuational change on the bridge itself (punctuated equilibrium sensu stricto), combined with a strong case for punctuational dynamics in Darwinian processes stripped to a lean and clean minimality in microevolution (see preceding section), behooves us to consider a generalization across all scales, by suggesting an examination of larger realms beyond the bridge of speciation. I shall therefore discuss potentially instructive examples (not comprehensive statistical generalities, a worthy goal not nearly in current reach) at three expanding levels: (1) consequences of accumulated events of ordinary speciation within the history of individual clades; (2) the origin of phenotypically complex and extensive evolutionary novelties; and (3) the history of biotas in ecological and evolutionary time.
Stasis analogs: trending and non-trending in the geological history OF clades. Do we find cladal patterns, generated by different causal mechanisms, that might be sufficiently “homologous” (see pp. 928–931) with punctuated equilibrium to warrant comparison based on a common deep structure? (I am, of course, not considering or reiterating here (see full discussion, pp. 886–893) the most important and direct impact of punctuated equilibrium upon cladal histories — its ability to explain trends as the differential success of species rather than the extrapolated result of adaptive anagenesis. This section treats other scales and causes of change with meaningful structural parallels to punctuated equilibrium at the species level.) Possible parallels for the punctuational aspect will be treated in subsequent chapters — rapid origin of extensive evolutionary novelties for cladal beginnings (Chapter 10), and patterns of mass extinction for endings (Chapter 12) — but we should also consider the almost entirely neglected analogs of stasis at the cladal level.
An apt comparison for clades may be made to philosophical and sociological reasons for the previous failure of evolutionary biology to study, or even to acknowledge, the phenomenon of stasis as the predominant feature of phyletic history in a large majority of species. Stasis, construed as absence of evolution, once designated a negative result unworthy of a category, or even a name. In a similar way (and I cast no stones from a sinless state, for I have followed this tradition myself throughout my career), the study of trends has consumed nearly all research on the history of clades. And why not? Trends tell stories, and evolution is a narrative science. Western tradition, if not universal human nature itself, has always favored directional tales of conquest and valor (with Darwinian analogs of competition and adaptation), while experiencing [Page 937] great discomfort with the aimless cyclicity of Ecclesiastes, however much we may admire the literary power: “The thing that hath been, it is that which shall be; and that which is done is that which shall be done; and there is no new thing under the sun” (Ecclesiastes 1:9).
But the undeniable salience of trends — a psychological comment about our focus of attention — bears no necessary relationship to the relative frequency or causal weight of this phenomenon in the natural history of clades. How many monophyletic clades feature sustained and substantial trends in major characters of functional importance — and what percentage of characters participates in trends that do exist in such clades? Indeed, we have no idea whatever, for no neutral compilations exist. No one has ever tabulated the number or percentage of non-trending clades within larger monophyletic groups. The concept of a non-trending clade — the higher-level analog of a species in stasis — has never been explicitly formulated at all. If only one percent of clades exhibited sustained trends, we would still focus our attention upon this tiny minority in telling our favored version of the story of life's history.
(Ironically, stable lineages become salient enough to catch our attention only at the extreme that we call “living fossils” — species or lineages supposedly unchanged during such long stretches of geological time that their stability becomes a paradox in a world of Darwinian evolutionary flux and continuity. As a double irony — see pages 817–820 for a full discussion in the light of punctuated equilibrium's different reading — we have also thoroughly misinterpreted this phenomenon under the same gradualistic bias that inspired our notice in the first place! The classical “living fossils” (the inarticulate brachiopod Lingula, the horseshoe crab Limulus polyphemus, the extant coelacanth) are not long-lived as species (Limulus polyphemus, for example, has no fossil record at all), but rather belong to clades with such a low speciation rate that little raw material for cladal trending has been generated over the ages.)
I suspect that most clades, while waxing and waning in species diversity through time, show no outstanding overall directionality. But we do not know because the literature has never recognized, or attempted to tabulate, the frequency of such “Ecclesiastical” clades that change all the time, but “go” nowhere in particular during their evolutionary peregrinations. Paleontologists achieved no sense of the strength of punctuated equilibrium, even though Eldredge and I had formulated the theoretical apparatus, until researchers studied the relative frequencies of stasis and punctuation in entire faunas, or entire clades, by full sampling and with no predisposing bias to favor one kind of result — see discussion on pages 854–874 for this extensive and growing literature. Similarly, we will not know the general fate of clades until we ratchet this methodology one notch higher,
and sample sets of clades not identified by our prior sense of their evolutionary “interest.” Stasis is data at the species level. Non-trending is data at the clade level.
Budd and Coates (1992) broke conceptual ground in devoting an entire paper to such Ecclesiastical non-directionality in the actively evolving and speciating clade of montastraeid corals during 80 million years of Cretaceous [Page 938] time. As a rationale for their study, the authors state an analogy to the lower level phenomenon of punctuated equilibrium: “Just as the study of stasis within species has facilitated understanding of morphologic changes associated with speciation, we show that study of nonprogressive evolution offers valuable insight into how the causes of trends interact and thereby produce complex evolutionary patterns within clades, regardless of their overall direction.”
The central theme of non-trending, identified by Budd and Coates for this large clade of massive, reef-building corals, stands as an empirical pattern in any case, but the (admittedly somewhat speculative) explanation proposed by the authors also builds an interesting framework for regarding such a signal as predictable and unsurprising, rather than anomalous. Their proposal also integrates the two principal Darwinian critiques of this book by attributing a causal pattern generated at the species level (the hierarchical expansion of my first theme) to the effects of architectural or developmental constraint (the structuralist or internalist perspective of my second theme) in channeling the possibilities and directions of natural selection.
Budd and Coates (1992) propose that monstastraeid species vary within a range set by minimal and maximal size of individual corallites on these large colonies. Such a notion does not debar classical trending, for the clade could originate in one small portion of the permitted range, and then strongly trend towards the other domain. But Budd and Coates argue that Cretaceous montestraeids already inhabited the full range, and that each end represented an adaptive configuration continually available and exploited throughout the clade's duration. Therefore, phenotypic evolution fluctuated between the two realized potentials of a fully populated domain of workable solutions.
The authors argue that “large-corallite” species (3.5 to 8.0 mm in diameter) maximize efficiency in removal of sediment, and tend to dominate in turbid waters; while “small-corallite” species (2.0 to 3.5 mm in diameter) prevail in clearer waters of the reef crest. Moreover, large-corallite species derive most nutrition by direct carnivory, whereas small-corallite species tend to feed upon their own symbiotic zooxanthellae. Budd and Coates then advance the claim — the more speculative aspect of their scenario — that montastraeid species remain constrained with this range by limitations at either end: an inability of still smaller corallites to develop and function adequately, and a restriction in septal number and strength that would not grant sufficient bio-mechanical support to still larger corallites.
These two arguments may validate and explain the basis of active non-trending in a persistently vigorous and successful clade. For if such constraints limit the range of corallite size, and if each end enjoys advantages in different environments continuously available in major parts of the habitat, then evolution might oscillate back and forth, with no persistent directional component, throughout cladal history.
Budd and Coates document such a directionless oscillation within the clade's developmental and adaptive boundaries during four successive divisions of Cretaceous time. The transition from interval 1 to interval 2 featured [Page 939] a differential production of small-corallite species from large-corallite ancestors, as well as a southward expansion of the clade's geographic range. Limited and directionless speciation, accompanied by predominant stasis within established species prevailed during intervals 2 and 3. Between intervals 3 and 4, large-corallite species radiated from small-corallite ancestors, and the geographic range of the clade became more restricted. In other words, the general pattern at the end of interval 4 differed little from the initial spread of morphology and geographic range at the outset of interval one, the inception of the study itself.
But the montastraeids remained a vigorous, successful, and evolutionarily active clade throughout Cretaceous times. Who are we to proclaim their pattern “boring” or unworthy of study because the evolutionary history of these corals does not resonate well with human preferences about “interesting” or “instructive” stories? Perhaps we should force ourselves to learn that patterns traditionally shunned for such quirky reasons of human appetite may hold unusual interest and capacity to teach — precisely because we have never sought messages that might challenge our complacencies. The predominant pattern of life's history cannot fail to be instructive — and such non-trending may well mark a norm of this magnitude, even if heretofore hidden in plain sight because we also see with our minds, and conventional concepts can be more blinding than mere ocular failure.
Punctuational analogs in lineages: the pace of morphological innovation. I do not wish to resuscitate one of the oldest canards, and least fruitful themes of evolutionary debate: the claim for truly saltational, or macromutational — that is, effectively one generational, or ecologically “overnight” — origin of new species or morphotypes. This ultimate extreme in punctuational change has never been supported by punctuated equilibrium, or by any sensible modern account of punctuational change in any form. Even if older evolutionists did advocate this mode of change (see my discussions of de Vries on pp. 415–451 and of Goldschmidt on pp. 451–466), they granted no exclusivity to its operation, and they also defended more continuationist, and more structurally plausible, accounts of rapid origin for morphological novelties — as in the developmentally rooted and theoretically sensible concept, based on mutations in “rate genes,” embodied in Goldschmidt's unwisely named “hopeful monster,” in contrast with the speculative scenario that he built upon his later concept of “systemic mutations” — see Schwartz (1999) for an interesting modern retelling and defense of this notion.
Just as punctuated equilibrium scales the geologically abrupt (but ecologically slow and continuous) process of speciation against the long duration of most species in subsequent stasis, punctuational hypotheses at higher levels regard the pacing of substantial phenotypic change in the origin of novel morphotypes as similarly episodic, with origins concentrated in very short episodes relative to periods of stability in basic design during the normal waxing and waning of clades — and perhaps with the ecologies, or the structural and developmental bases, of such episodes belonging to a distinctly different [Page 940] class of circumstances from those that regulate the ordinary pace of flux and speciation during the long-term geological history of most clades and morphotypes. In other words, a timekeeper with a metronome beating at an appropriate frequency for discerning the units and causes of evolution at each scale of nature's hierarchy might recognize an episodically (and rarely) pulsating, rather than an equably flowing, tempo as the dominant signal of change in all realms.
For example, Erdtmann (1986, p. 139) proposed that the active cladogenesis of early Ordovician planktic graptolites (an extinct subphylum of colonial organisms close to the chordate lineage) “operated on two levels: gradualistic change involving species-level and intergeneric clades, and punctualistic (anagenetic) changes operating on supergeneric levels.” He linked the rapid and extensive morphogenetic innovations of the punctuational mode, involving such basic features of colonial form as loss of bithecae and reduction in number of stipes, to major environmental changes marked by rapid eustatic shifts in sea level. Moreover, these punctuational innovations arise by an astogenetic mode (a term for the ontogeny of colonies) different from the developmental basis of most smaller changes that mark the flux of speciation during “normal” geological intervals. The punctuational innovations that produce new developmental patterns begin at the proximal end of the colony — that is, they affect the early ontogeny of the initial organisms of the developing aggregate, thereby pervading the life cycles of both the organism and
the colony, and strongly affecting the global phenotype of the entire structure. The lower-level changes (smaller variations within existing developmental themes), on the other hand, tend to begin at the colony's distal end — that is, they arise late in the ontogeny of older organisms in the colony (for new organisms of the colony arise proximally, pushing older organisms to progressively more distal positions), and then proceed to earlier phylogenetic expression in both the colony's astogeny and the individual organism's ontogenies. These lower-level changes therefore affect only a small, and astogenetically late, portion of the colony's form — hence their much more limited capacity for yielding major morphological change.
This case provides an interesting astogenetic analog to the common claim that heterochronic changes in the early ontogeny of organisms gain a distinctive status among evolutionary mechanisms in their potential for rendering substantial phenotypic change (at a punctuational tempo) with minimal genetic alteration. In fact, the lability inherent in early ontogenetic changes of rate and regulation undergirds most theorizing about qualitatively different categories of evolutionary outcomes based on similar underlying magnitudes of raw genetic alteration — the most promising basis for a dominant punctuational tempo in the history of morphological innovation in evolution.
The Structure of Evolutionary Theory Page 149