This summer's meeting season saw five gatherings devoted specifically to the topic of extinction: Chicago's Field Museum symposium in May (see Diamond 1983); a field conference to examine the Cretaceous-Tertiary transition in Montana; a symposium on molluscan extinction at the American Malacological Union meeting in Seattle; a special session at the annual meeting of the Society for the Study of Evolution in St. Louis; and a research conference on the campus of Northern Arizona University in Flagstaff (see also Lewin 1983a,b,c). Several recent books consider the topic (Ehrlich and Ehrlich 1981; Nitecki 1981; Silver and Schultz 1982; Martin and Klein, 1984) and plans are underway to publish the papers presented at the Field Museum, Seattle, and Flagstaff meetings. Our review of the topic is based principally on the themes explored at the St. Louis and Flagstaff meetings.

I. Many believe that sooner or later molecular biology will mechanistically provide the precise genotypic basis for phenotypic change. This in turn will provide the basis for attacking several major evolutionary problems posed by paleontology, including the following:

1. — There are approximately 30 phylum-level taxa in the animal kingdom, and yet in the “fast” version of evolutionary history, many of then arose 600 Myr ago in a very short period of time (of order <30 Myr). In terms of basic body plan and inferred ecology, these sponges, clams, arthropods, etc. were seemingly as different then as they are now. It is as though the rate of morphologic change was much more rapid during that interval than during the nearly 600 Myr since then. Even the “slow” version of evolutionary history (which allows for an additional unfossilized record of about 500 Myr) demands an answer to what has controlled the development of major evolutionary Bauplans.

2. — Some morphologic traits (though probably not individual species) persist for immense periods of geologic time—the general shape of the horseshoe crab, for example, or the bristle patterns of taxa of meiofaunal “worms” living among sand grains. What is the developmental rationale for these persistent traits? In short, what is the molecular basis for genetic and morphologic canalization?

3. — The general pattern of species specialization proceeds unabated overtime, but every once in a while the developmental clock is reset. Organisms which are sexually mature but with simplified or “primitive” morphology arise de novo, and these lineages then form the stock of unspecialized forms from which future specializations evolve. What is the molecular basis for resetting the developmental clock?

4. — Reproductive isolation may arise “accidentally” through genomic changes unrelated to changes in the local environment. To the extent this is true, speciation has a largely accidental component. Knowledge of the rate and type of genomic change is critical for an evaluation of this idea.

The concept of punctuated equilibria was introduced by Eldredge and Gould (1972) as an interpretation of the sequential occurrence of species as actually observed in the fossil record. The pattern of sudden appearance of a new species followed by little or no morphological change during the remainder of the species' existence contrasts with phyletic gradualism, the pattern of slower, more evenly distributed change which was long considered to be the principal mode of species formation (Simpson 1944). The question of which is the dominant pattern of evolutionary change has continued as an actively debated theme in much of the paleontological literature of the past decade.

The punctuated equilibrium hypothesis, largely derivable from Mayr's peripheral isolate model of speciation, is but one of many possible paleontologic patterns of speciation which may be tested against the deep-sea fossil record of planktonic microfossils. The nature of reproduction and variation within a species has a profound effect on its expected tempo and mode of evolutionary change. Biologic data on pelagic plankton and on protistan genetics and reproduction suggest that speciation in pelagic holoplanktonic protists may also be parapatric, “equal” allopatric, or the result of hybridization. Each of these models is shown to make testable predictions of paleontologic pattern. Published records of speciation in planktonic deep-sea microfossil data are compatible with these alternative models. Existing data sets are not yet sufficiently complete to provide strong tests. Future work needs to integrate taxonomic, paleoecologic, biogeographic, and biometric data in studies of individual speciation events. Micropaleontologic studies of morphologic change through time may thereby contribute to our understanding of speciation mechanisms in the pelagic protistan plankton.

Changes in morphological variability within fossil populations of planktonic protists provide insight into the processes responsible for morphological change. Evolution in the radiolarian genera Theocorythium and Lamprocyclas is documented biometrically in Pliocene and Pleistocene core material from the equatorial Pacific and Indian Oceans. The patterns of morphological change in Theocorythium within a single Pacific core could be interpreted as indicating the in situ evolution of T. trachelium trachelium from T. vetulum via intermediate forms during the Pleistocene. However, consideration of biogeographic data shows that this is not the case and supports instead two alternate interpretations: (1) immigration coupled with extensive hybridization and introgression, or (2) multiple polymorphs of a single species changing in frequency along a cline. Analyses of evolutionary change in Theocorythium and the closely related genus Lamprocyclas during the Pliocene suggest that hybridization may be a significant factor in the production of new species. Evolutionary models and phylogenetic analyses should consider that barriers to gene flow may vary significantly with time and permit the exchange of genetic information between populations previously reproductively isolated.

One method of increasing the resolving power of paleontological inquiries is to assess the time distribution, or phenology, of perispeciational change. The rationale for this test is that, while concentration of morphologic change around the time of speciation is consistent with both known microevolutionary mechanisms and punctuated equilibria, punctuated equilibrium theory presupposes that most of the change occurs either before or during initial allopatry to produce morphologically static taxa. Considerable change in the initial phases of speciation could also result from known microevolutionary mechanisms, but if significant change occurs in the neosympatric phase, or at any time after the species have become differentiated, then that change is best explained by known microevolutionary mechanisms.

Amount and rate of morphologic change during each phase of the allopatric speciation process were determined for two cognate species of the radiolarian genus Eucyrtidium. In this case, morphologic change accruing to these species during the neosympatric phase was 2-3 times as great as that which occurred during the initial allopatric phase. Morphologic trends in both these species during the allopatric phase differed significantly from a random walk. The size differential that characterized this speciation event was not the product of a single large step early in the process but of a disproportionate number of small steps in one direction during neosympatry. Therefore, the hypothesis of no statistically significant change following the “first stage” of speciation, which is a major tenet of punctuated equilibrium theory, is falsified for this case.

Polycystine radiolarians axe the most widely distributed (geographically and geologically) of the well-preserved microzooplankton. To correctly interpret the tempo and mode of radiolarian microevolution, speciation and macroevolution, the zoogeography and ecological niches of extant, and paleozoogeography and paleoecological niches of extinct subspecies, species and higher taxa of the studied lineages should be carefully considered. Such studies of the Stichocorys and Lamprocyrtis lineages suggest that allopatric speciation as peripheral isolates is important. Sympatric speciation is also a possible mode of evolution in these lineages. Hybridization may play a role in the evolution of one subspecies and one species. Allopatrically evolved subspecies and species require some time to inhabit their maximum geographical range after evolving in an isolated and peripheral part of that “environment,” whereas the supposedly sympatrically evolved forms do not. Cooling trends and water mass reorganization may be related to the speciation events. Similar, but much less controlled, studies on a macroevolutionary scale (evolution above the species level, essentially the “family” level) suggest that there have been about half a dozen periods of major adaptive breakthroughs resulting in quantum, or macro, evolution. Global coolings and the initiations of “new” cold-water spheres and related water masses appear to have allowed these macroevolutionary steps during the Ordovician, Devonian-Carboniferous, Permian, and early Neogene. Ecological pressure from diatoms and planktonic foraminiferans appears to have been related to macroevolutionary events bounding the Cretaceous-Tertiary transition.

The warm-water planktonic foraminiferal Globorotalia tumida lineage has been studied in a 10-Myr-long stratigraphic sequence (Late Miocene through Recent) from the Indian Ocean to determine long-term evolutionary patterns through the lineage's history, and particularly to study in great detail the evolutionary transition from G. plesiotumida to G. tumida across the Miocene/Pliocene boundary. Sampling resolution was very good, between 5 × 103 and 15 × 103 yr across the Miocene/Pliocene boundary and about 2 × 105 yr otherwise. The test shape was analyzed in edge view, permitting determinations of variation in inflation and elongation of the test. Shape was analyzed quantitatively using eigenshape analysis. This method represents the greatest proportion of variation observed among a collection of shapes by the least number of different shapes. The Late Miocene (10.4-5.6 Myr B.P.) populations exhibited only minor fluctuations in shape that did not result in any net phyletic change. This period of stasis was followed by an 0.6-Myr-long period (between 5.6 and 5.0 Myr B.P.) of gradual transformation of the Late Miocene morphotype (G. plesiotumida) into the Early Pliocene morphotype (G. tumida). The populations were again more or less in stasis in the Pliocene and Pleistocene (5.0 Myr to the present day), so that no major modifications of the newly evolved Early Pliocene morphotype occurred during these 5 Myr. Thus it would appear that the G. tumida lineage, while remaining in relative stasis over a considerable part of its total duration underwent periodic, relatively rapid, morphologic change that did not lead to lineage branching. This pattern does not conform to the gradualistic model of evolution, because that would assume gradual changes throughout the history of the lineage. It also does not conform to the punctuational model, because (1) there was no speciation (lineage branching) in this lineage and (2) the transition was not rapid enough (<1% of the descendant species' duration according to definition). For this evolutionary modality we propose the term “punctuated gradualism” and suggest that this may be a common norm for evolution—at least within the planktonic foraminifera.

A preliminary report is presented documenting the evolution of the Globorotalia crassaformis (Galloway and Wissler) lineage from the Early Pliocene G. cibaoensis (Bermudez) in Hydraulic Piston Core 72/516 on the Rio Grande Rise (western South Atlantic). A new technique, eigenshape analysis, is used to describe the gross changes in morphology, and the use of “Raupian” analysis of expansion rates, angular displacement, position relative to coiling axis, and translation rates of chambers is suggested as a useful vehicle for inferring the geometric causes of observed evolutionary changes. These changes include size increase, decrease in the number of chambers per whorl, inflation, reduction of relative height of aperture, and increase in the peripheral angle. The sequence is judged to be an example of phyletic gradualism.

Changes in shape and in the pattern of the reticulate carapace ornament of 63 representative female ostracode specimens of the deep-sea genus Poseidonamicus from 28 Deep Sea Drilling Project Sites, extending over a geographic distance of 40,000 miles and a geologic age of 40 × 106 yr, have been analyzed by Resistant Fit Theta-Rho Analysis, conventional least-squares Theta-Rho Analysis, and by inspection of homologies in the fossae of the reticulum. Through comparison of the relative rates of morphological change on at least three levels of integrated complexity, it is possible to demonstrate that a sudden evolutionary change in the architectural framework of the carapace can be the product of mechanical accommodation under stress to more gradual and general changes in carapace shape. Certain geometric imperatives related to these mechanical needs determine morphologic stability and consequently rates of evolutionary change. In the western deep South Atlantic (near the Vema Channel on the Rio Grande Rise), an important evolutionary “punctuational event” seems to have taken place at about 14 Myr, whereas its concomitant, more gradual transformation is traceable in the shallower eastern South Atlantic (Walvis Ridge). This evolutionary event in Poseidonamicus probably reflects the relatively sudden invasion of deep Antarctic Bottom Water through the Vema Channel resulting from the formation of the East Antarctic Ice Cap, while the more gradual transition in shallower depths of Walvis Ridge represents a temporary transitory refuge. The problems of recognizing allopatric origins of species in the deep sea compared to in situ sudden transitions caused by biomechanical instability are considered.

The biogeography of differences in average shape of the modern planktonic foraminifer Globorotalia truncatulinoides (d'Orbigny) exhibits systematic relationships to changes in the ocean's surface environment. Comparison of these shape changes, as they exist today in the Southern Hemisphere, with fossil shapes preserved in a Late Pleistocene record from the South Atlantic Ocean, shows that the biogeography of G. truncatulinoides ecophenotypes has changed markedly through time. Beginning at least 700,000 yr ago and continuing up to the present, there has been a gradual but clear migration of certain morphotypes of G. truncatulinoides toward lower latitudes. The history of this migration bears no simple relationship to the cyclic climatic changes that characterize the Late Pleistocene. We conclude that either (1) phenotypic variants of Gr. truncatulinoides reflect some previously unmeasured, gradually changing aspect of Late Pleistocene oceans, or (2) we are witnessing a gradual evolution of the environment preferences of G. truncatulinoides.

To the student of evolution, Malmgren and Kennett's (1981) detailed study of the Globorotalia inflata lineage demonstrates the continuity of the stratigraphic record and the abundance of planktic microfossils in deep-sea sediments. Their work recalls the prediction (Prothero and Lazarus 1980) that planktic microfossils might serve as the laboratory animals of evolutionary studies and indicates the potential value of fossils in determining ancestry. However, for theorists to judge the evolutionary import of Malmgren and Kennett's high-resolution morphometric study, broader consideration of lineage relationships is needed. They portray the G. inflata lineage as a set of five successional taxa in simple ancestor-descendant relationship. Trends, established morphometrically and claimed to be “one of the clearest examples of gradualism in any group of fossils” (Malmgren and Kennett 1981, p. 236), are shown to progressively transform a compressed shell with peripheral keel (G. conoidea) into a globose shell (G. inflata) without a keel. While the succession of taxa given by Malmgren and Kennett has been supported in other studies of western South Pacific sequences (e.g., McInnes 1965; Walters 1965; Kennett 1966; 1973; Jenkins 1971; Scott 1979; Hornibrook 1982), it is not established that the order indicates simple phyletic transformation without branching, as Malmgren and Kennett imply. Crucial evidence concerns the history and status of keeled populations which resemble the ancestral G. conoidea but occur in the Pliocene, associated with advanced members of the G. inflata lineage. These populations are not mentioned by Malmgren and Kennett, yet, if they are also descendants of G. conoidea, their occurrence strongly suggests that lineage divergence occurred. I wish to review evidence in support of this view.

We appreciate Dr. Scott's comments about our paper (Malmgren and Kennett 1981), but wish to clarify some aspects which we hope will help provide a better understanding about the Late Neogene evolution of the Globorotalia (Globoconella) group (Kennett and Srinivasan 1983). We also wish to dispute some of Scott's other criticisms. In effect, Scott's arguments have provided no reason to justify abandonment of our initial contention that this lineage represents one of the best documented cases of phyletic gradualism in any group of fossils.

We support Scott's notion that divergence of the lineage took place near the Miocene-Pliocene boundary, but this divergence is not relevant for our evolutionary study. In fact, the Early Pliocene form resembling Globorotalia conoidea was already recognized by one of us (J.P.K.) 10 yr ago (Kennett 1973). It was distinguished as G. cf. conoidea and not G. conoidea as stated by Scott. In our previous papers (Malmgren and Kennett 1981, 1982) we confined our attention to the evolution that occurred in the primary temperate stock (cool subtropical) water mass—the abundant and dominant forms of Globoconella.