Our assessment of morphological diversity is influenced by morphological extremes and therefore depends on sample size (taxonomic richness). Rarefaction predicts the morphological diversity that would probably be observed in a sample of reduced size, thereby allowing both compensation for differences in sample size that may be strictly preservational, and analysis of diversity structure, that is, the relationship between morphological and taxonomic diversity. Middle and Late Cambrian trilobites exhibit a diversity structure characterized by many variations on a few morphological themes. In contrast, Middle and Late Ordovician trilobites occupy a larger range of morphospace per unit of species richness. Diversity structure in the Devonian is similar to that in the Middle and Late Ordovician, but the magnitude of morphological diversity is lower in the Devonian, as many fewer species are observed. For blastoids, different aspects of morphological diversity (range of morphospace occupied, number of character states possessed, and number of different regions in morphospace occupied) exhibit different relationships to taxonomic richness. In all cases Permian blastoids are characterized by a diversity structure in which morphological diversity per unit of taxonomic richness is greater than for Devonian blastoids. Changes in morphological diversity in fissiculate blastoids appear to reflect evolution of continuous variation in thecal morphology more than changes in the number of character states. Saunders and Swan's data on Namurian ammonoids illustrate some significant differences in diversity structure among stratigraphic levels, but many apparent differences in morphological diversity are consistent with the possibility that they reflect the sampling of different numbers of species from the same underlying diversity structure. Rarefaction curves are also presented for idealized increases and decreases in diversity, and these are compared to some of the observed changes in trilobites, blastoids, and ammonoids.

Calliostoma ligatum shells inhabited by hermit crabs were weaker than shells inhabited by snails collected at the same locality. When shells of C. ligatum were loaded repetitively to 80% of their predicted failure load, hermit crab-inhabited shells showed an immediate drop in shell strength followed by a progressive further loss of shell strength over the next 18 days. Snailinhabited C. ligatum shells exhibited a decrease in strength after 9 days, but returned to initial values within 18 days of loading.

Hermit crabs thus bear shells significantly weaker than they were when borne by the gastropods that produced them. Reported similarities in vulnerability of gastropods and hermit crabs to shell-crushing predators may be artifacts of the metric (critical size) used to compare vulnerabilities. Hermit crabs probably were a significant factor in the diversification of durophagous predators in the Mesozoic, supplying a prey base identical in size and shape to gastropods but with significantly lower resistance to crushing. The unpredictability of strength in hermit-crab inhabited shells may maintain the apparently inefficient indiscriminate attacks common among durophagous predators.

The genus Homo evolved its pronounced encephalization through postnatal extension of the high rate of brain growth that characterizes all primates in utero. Linked to this extension was delayed development, which represented an enormous ecological sacrifice because it produced the longest postnatal interval of physical helplessness in the Mammalia and forced mothers to carry infants.

Graphs relating brain growth to body growth indicate a pongid pattern of development for gracile australopithecines, implying that infants could cling to mothers whose forelimbs were occupied with climbing. Also present were several postcranial traits that would have made the adults more adept climbers than modern humans. Habitual use of these inherited traits is suggested by the fact that evolution failed to eliminate certain ones, such as short legs and long pedal phalanges, that impeded terrestrial locomotion. Moreover, the intensity of predation by large, swift, social carnivores must have compelled australopithecines to use trees as refuges, in the manner of chimpanzees and baboons; australopithecines probably also gathered some of their food in trees. Gracile australopithecines failed to expand their brain size, experiencing general evolutionary stasis for more than 1.5 m.y. I propose that this stability resulted from these animals' semiarboreal mode of life: First, their postcranial morphology remained compromised by selection pressures to maintain both terrestrial and arboreal adaptations. Second, by requiring that neonates be mature enough to cling to mothers, obligate arboreal activity precluded encephalization of the kind that characterizes Homo; this evolutionary constraint has previously been overlooked.

In contrast to australopithecines, early Homo approached H. erectus in pelvic configuration and brain size. A new brain-body growth curve for early Homo indicates extension of the fetal pattern well into the postnatal interval, implying that neonates were highly immature so that adults had to be fully terrestrial. Homo evolved shortly after the onset of the modern ice age about 2.5 Ma. Fossil pollen and carbon isotopes in paleosols record a contraction of forests in Africa at this time. I propose that this represented a crisis that led to the evolution of Homo by compelling some australopithecine populations to adopt a fully terrestrial existence. Although ecologically difficult, this behavioral restriction finally made possible encephalization through the evolution of delayed development. During the ecological crisis, a large brain evolved in at least one population of gracile australopithecines. An advanced tool industry and cunning behavior were of such great adaptive value for avoiding predators and expanding food resources on the ground that selection for encephalization soon overrode the problems imposed by helpless infants.

The fossil record of antelopes and micromammals provides a test of the idea that environmental forcing opened the way for the evolution of Homo: both of these groups experienced heavy extinction of forest-adapted species about 2.5–2.4 Ma and a rapid proliferation of species adapted to unforested habitats. The transformation of the hominid clade during Plio-Pleistocene time did not follow a simple pattern. Homo may have arisen either by anagenetic transformation of a “bottlenecked” species or by speciation, and it may not have evolved immediately with the onset of climatic change. Furthermore, just as a few forest-adapted antelope species survived the biotic crisis, a small-brained gracile taxon with arboreal adaptations may have persisted to the start of the Pleistocene. Robust australopithecines survived into the Pleistocene, perhaps because a broad vegetarian diet reduced their need to migrate frequently between home bases. With their extinction in mid-Pleistocene time, about the time that savannahs became widespread, only Homo remained.

For a survivorship curve to show a meaningful pattern, it is essential that a suitably homogeneous group is selected for analysis. If taxa have originated at different times in the geological record and have consequently experienced different extinction probabilities during the times that they existed, they do not constitute such a homogeneous group. A correction factor for variations in real-time extinction rates must be employed, as is described here. By application of the “corrected survivorship score,” it is shown that the Paleogene planktonic foraminifera exhibit a very strong survivorship pattern, namely an extinction probability that progressively increases with taxonomic longevity. That is to say, when extinction occurred, it was preferentially selective of older taxa. This is manifested as convexity in the survivorship curves. However, the pattern is degraded by variations in the real-time extinction rate, which causes straightening of the survivorship curves if they are calculated in the usual uncorrected way. If the law of constant extinction is to be tested for a given group, it is necessary either that stochastically constant real-time extinction rates are demonstrated or that variations in the extinction rate are corrected for in the manner described. As a separate issue, it is also essential that biologically meaningful taxa are analyzed. The convexity of planktonic foraminiferal survivorship curves is probably an artifact of the way evolving lineages have been subdivided in the application of typological taxonomy. Consequently, we are still a long way from being able to use survivorship analysis of planktonic foraminiferal data for adequately testing evolutionary models such as the Red Queen's hypothesis.

The post-Paleozoic radiation of echinoids entailed a rapid diversification not only of adult morphology, but also of larval morphology. The timing, order, and phylogenetic distribution of evolutionary transformations in echinopluteus larvae are reconstructed here under maximum parsimony assumptions from a large neontological data base. Many echinoid larval apomorphies apparently evolved within the Paleozoic stem lineage and were subsequently retained during much of the crown-group radiation. This suite of apomorphies includes most (and perhaps all) of the skeletal elements and some features of soft anatomy such as vibratile lobes. Other features apparently arose or were modified during the post-Paleozoic radiation. These include skeletal features such as arm-rod structure and length, and soft structures such as epaulettes and skeletal muscles. Transformational hypotheses of this kind can be supported or rejected with further neontological data, and many can potentially be tested from fossil evidence. Many post-Paleozoic transformations in echinopluteus structure may be adaptive. For example, increases in arm length and flattening of arm ectoderm may increase feeding rate and efficiency, and both types of transformation have occurred several times within the crown group. Relational hypotheses of this nature can be tested through comparative functional studies in extant echinoplutei. Parallel evolutionary losses of feeding in echinoplutei are accompanied by loss or modification of characteristic structures. This suggests that developmental constraints do not fully explain the conservation of these structures in planktotrophic echinoplutei. Comparisons of larval and adult morphology over congruent time intervals demonstrate that the origin of many orders was accompanied by a suite of synapomorphies in larval morphology that was subsequently conserved. Many details of echinopluteus morphology are therefore of taxonomic significance. Intraordinal patterns of larval diversity, however, vary considerably. That larval morphology has diversified independently of adult morphology indicates that mosaic evolution has occurred within the life cycle and suggests that echinoid larvae and adults can and do respond to selection independently. Taken together, these findings underscore the complex ways in which complex life cycles can evolve.

The shovel-tusked gomphotheres are normally portrayed scooping up water plants with their shovellike mandibular tusks. This portrayal is based on speculation about the possible functions of the lower tusks and misinterpretation of mandibular-tusk wear patterns that goes back to the 1920s and 1930s. In addition, most literature concerning shovel-tusker feeding behavior ignores the possibility that the upper tusks may have had a role in feeding, as they do in modern African elephants. Because wear patterns on tusks provide direct evidence as to how tusks were used during feeding, I examined and interpreted patterns on the lower and upper tusks of shovel-tusked gomphotheres on the basis of theoretical models of tusk wear to determine how the tusks were used in feeding.

This examination led to the following conclusions. (1) The genus Amebelodon was characterized by feeding opportunism, gathering food with both its lower and upper tusks in a variety of ways. (2) There is evidence for similar flexibility within the genus Serbelodon. (3) Wear patterns on mandibular tusks of Platybelodon grangeri and Torynobelodon barnumbrowni indicate that these taxa did not practice mandibular shoveling but used these tusks to cut tough vegetation in a specialized fashion. (4) The flaplike trunks (probosces) generally attributed to shovel-tusked gomphotheres in restorations are shown almost certainly to be erroneous; morphological and functional evidence suggests that these animals had trunks similar to those of modern elephants. (5) Strong tusk sexual dimorphism is unknown in the shovel-tusked gomphotheres except for Platybelodon and should not be found in any taxa that relied on their tusks for procuring food. New restorations of both Amebelodon and Platybelodon are provided that are more consistent with the actual evidence bearing on their feeding behavior.

Complex nasal turbinal bones are associated with reduction of respiratory water loss in desert mammals and have previously been described as an adaptation to xeric conditions. However, complex turbinates are found in virtually all mammals. Experimental data presented here indicate that turbinates also substantially reduce respiratory water loss in five species of small mammals from relatively mesic environments. The data support the conclusion that turbinates did not evolve primarily as an adaptation to particular environmental conditions, but in relation to high ventilation rates, typical of all mammals. Complex turbinates appear to be an ancient attribute of mammals and may have originated among the therapsid ancestors of mammals, in relation to elevated ventilation rates and the evolution of endothermy.

Dickinsonia is reconstructed as a benthic polypoid of generally cnidarian design. The oral surface was without tentacles but contained a median oral slit that probably led through a pharynx into an enteron, which was divided into digestive diverticulae by radiating mesenteries; feeding may have been via ciliary tracts. The mesenteries and body wall contained a stiff form of mesogloea. There seems to be no need to postulate a novel constructional grade for this organism, bringing the concept of the Vendozoa into question.

New evidence from fossil plants at the Middle Miocene fossil quarry near Fort Ternan, Kenya, together with that from paleosols, allow reconstruction of a mosaic of early successional woodland (on Dhero paleosols), grassy woodland (on Chogo clay eroded phase and ferruginized nodule variant paleosols) and wooded grassland (on type Chogo and Onuria clay paleosols). This grassy open vegetation was on a high plateau of phonolite at the foot of a carbonatite-nephelinite stratovolcano, which probably supported dry Afromontane forest, alpine meadows, and marsh. This earliest savanna-mosaic vegetation yet documented from Africa, was probably recruited from dry lateritic soils elsewhere in Africa during climatic drying and cooling some 15 Ma. These early grassland ecosystems were very different from Early Miocene forest ecosystems of East Africa, but not altogether like modern grasslands either. Already present were grasses with dense growth and rich in silica bodies, and abundant antelope with moderately high crowned teeth and cursorial limb structure. These mammalian adaptations to grasslands, however, were not nearly so pronounced as they are in modern African grassland faunas, which include zebra and other Asiatic immigrants, as well as antelope. Grasses of the subfamily Chloridoideae and supertribe Panicanae were common in tropical Africa by Middle Miocene time, if not much earlier, but there is not yet evidence so far back in time for the grass supertribe Andropogonae which is now dominant in seasonally arid, overgrazed, and burned African grasslands.

Although much natural historical evidence has been adduced in support of the occurrence of several mass extinctions during the Phanerozoic, unambiguous statistical confirmation of the mass extinction phenomenon has remained elusive. Using bootstrapping techniques that have not previously been applied to the study of mass extinction, we have amassed strong or very strong statistical evidence for mass extinctions (see text for definitions) during the Late Ordovician, Late Permian, and Late Cretaceous. Bootstrapping therefore verifies three of the mass extinction events that were proposed by Raup and Sepkoski (1982). A small amount of bootstrapping evidence is also presented for mass extinctions in the Induan (Triassic) and Coniacean (Cretaceous) Stages, but high overall turnover rates (including high origination) in the Induan and uncertain estimates of the temporal duration of the Coniacean force us to conclude that the evidence is not compelling.

We also present the results of more liberal X2 tests of the differences between expected and observed numbers of familial extinctions for stratigraphic stages. In addition to verifying the mass extinctions identified using bootstrapping, these analyses suggest that several stages that could not be verified as mass extinction stages using bootstrapping (including the last three in the Devonian, and the Norian Stage of the Triassic) should still be regarded as candidates for mass extinction. Further analysis will be required to test these stages in more detail.

To assess the degree to which forest litter reflects the source forest, three 1-ha plots of temperate deciduous forest were mapped and litter accumulating in these forests was sampled. Identity, position, and diameter of all stems 2 cm or larger diameter at breast height are known for each forest. Composition of the leaf litter is governed by two key factors: (1) abscised leaves are deposited primarily on the forest floor directly underneath the canopy that produced them, and (2) the leaf mass of a species is highly correlated with its stem cross-sectional area. These factors produce autochthonous litter samples that correspond closely in composition to the forest within a circle of canopy-height radius or less. Even relatively small litter samples (350 leaves) consistently contained all the common species in the local area. However, the rarer tree species were seldom recovered in the litter samples. Correlation coefficients for litter mass and basal area by species are typically over .80.

These observations have three important implications for interpreting autochthonous compression-fossil assemblages. First, approximate relative abundances of locally dominant and subdominant forest taxa can be obtained from relatively small samples of autochthonous compression-fossil assemblages. Second, representation of rare forest species, even in large fossil samples, will be fortuitous. For this reason, complete species lists and consistent estimates of richness cannot be derived directly from most existing samples of autochthonous compression-fossil assemblages. Third, the strong tendency for leaves to fall beneath the canopy of the tree that sheds them suggests that properly sampled autochthonous fossil leaf assemblages may yield information on crown size of individual trees and the spatial distribution of individuals and species, aspects of vegetational structure that have been thought accessible only in well-preserved “fossil forests” with standing trunks.

Lengths of crinoid stalk segments consisting of multiple columnals (pluricolumnals) from the Fort Payne Formation of south-central Kentucky (Mississippian) were tested for uniformity using simulations. The results indicate that stalk segment lengths are too uniform to have been produced by a random process operating on homogeneous stalks.

The most parsimonious explanation for the uniformity of pluricolumnal lengths is based on analogy with stalks of living isocrinids. The stalks of these crinoids are organized into multicolumnal segments of approximately uniform length: columnals within each segment are connected by “through-going” ligament and “intercolumnal” ligament. The articulation between segments, however, consists of only intercolumnal ligament. Preliminary results from experiments on Recent carcasses indicate that the latter is the primary plane of disarticulation following death.

The preservational pattern of the Mississippian stalks suggests that their ligamentary organization was similar to that of the isocrinids. Since the plane of primary disarticulation in isocrinids corresponds to the autotomy plane, Mississippian crinoids also may have been capable of autotomizing their stalks.

Ecologists attempt to explain species diversity within Recent seabird communities in terms of Recent oceanographic and ecological phenomena. However, many of the principal oceanographic processes that are thought to structure Recent seabird systems are functions of geological processes operating at many temporal and spatial scales. For example, major oceanic currents, such as the North Pacific Gyre, are functions of the relative positions of continents and Antarctic glaciation, whereas regional air masses, submarine topography, and coastline shape affect local processes such as upwelling. I hypothesize that the long-term development of these abiotic processes has influenced the relative diversity and community composition of North Pacific seabirds. To explore this hypothesis, I divided the history of North Pacific seabirds into seven intervals of time. Using published descriptions, I summarized the tectonic and oceanographic events that occurred during each of these time intervals, and related changes in species diversity to changes in the physical environment. Over the past 95 years, at least 94 species of fossil seabirds have been described from marine deposits of the North Pacific. Most of these species are from Middle Miocene through Pliocene (16.0-1.6 Ma) sediments of southern California, although species from Eocene to Early Miocene (52.0-22.0 Ma) deposits are from Japan, British Columbia, Washington, and Oregon. During the history of the North Pacific seabirds, thee were many widespread (global) and local (California Current) oceanographic events, but the underlying physical processes that affected the diversity of seabirds throughout the North Pacific were tectonic changes that led to sequential stages of thermal isolation and refrigeration of Antarctica. Besides these broad-scale phenomena, the uplift of the Isthmus of Panama and the initiation and intensification of coastal upwelling in the California Current directly affected the diversity of seabirds in southern California. For the most part, Middle to Late Miocene and Pliocene seabird faunas from California resembled Recent communities, but with several exceptions: (1) the extreme diversity and abundance, but subsequent extinction of gannets (Morus) and flightless alcids (mancallids); (2) the comparatively low abundance and diversity of cormorants and shags (Phalacrocoracidae) until the Late Pliocene; and (3) the absence of marine Laridae until the Late Pliocene. By affecting both the physical and biological environments, geological factors such as Antarctic glaciation, eustatic changes in sea levels, and local tectonic activities influenced and will continue to influence the structure of seabird systems in the North Pacific.

The class Echinoidea apparently originated during the Ordovician Period and diversified slowly through the Paleozoic Era. The clade then mushroomed in diversity beginning in Late Triassic time and continued expanding into the present. Although this evolutionary history is generally accepted, the taphonomic overprint affecting it has not been explored. To gain a more accurate perception of the evolutionary history of the group, I have compared the diversity history of the family Cidaridae (Echinodermata: Echinoidea) with the preservational style of fossil type species using literature-derived data. The Cidaridae apparently originated in Middle Triassic time and diversified slowly through the Neocomian (Early Cretaceous). Diversity was maintained through the remainder of the Cretaceous and Tertiary Periods, reflecting the diversity history of the subclass. Characterization of the preservational style of type fossil material for the family revealed the following breakdown of preservational states: 60% of species were described on the basis of disarticulated skeletal material, primarily spines; 20% based on intact coronas denuded of spines, apical system, Aristotle's lantern and peristomial plates; 10% based on large coronal fragments; and 10% based on other skeletal elements. This distribution may represent the effect of a disarticulation threshold on the condition of echinoid carcasses before final burial and suggests that preservation of intact specimens may be very unlikely. For cidaroids, previous work has suggested that this threshold is likely to be reached after 7 days of decay.

Comparison of the diversity history of the Cidaridae with the preservation data reveals that characteristic patterns of taphonomic overprint have affected the group since its origination in Middle Triassic time, and the nature of that overprint has changed over time: the early diversity history of the group is characterized by occurrences of fragmented fossil material, with spines predominant; further radiation of the group in mid-Jurassic time coincided with an increase in modes of preservation, ranging between exceptionally well-preserved material and disarticulated skeletal elements. Finally, type material is more rarely described from younger stratigraphic intervals (Miocene–Pleistocene) and consists predominantly of disarticulated skeletal elements and coronal fragments larger than an interambulacrum in size. Intact, denuded coronas are noticeably lacking.

The number of type species of Cidaridae described in each stratigraphic interval has not been consistent during post-Paleozoic time. Middle Triassic, Malm (Upper Jurassic), Senonian (Upper Cretaceous) and Eocene series yielded significantly (α = .05) higher numbers of type specimens per million years, while the Lias (Lower Jurassic), Dogger (Mid-Jurassic), Lower Cretaceous and Paleocene yielded significantly (α = .05) lower numbers of type specimens per million years. This may be the result of a combination of taxonomic, sampling, and geographical biases.

Faunal similarity among regions is often used as a means of identifying regions of endemism in fossil faunas. At least two large-scale taphonomic effects can affect apparent faunal similarity: stratigraphic and facies mismatching. In stratigraphic mismatching, an unconformity represents removal of most or all of a complete assemblage zone in one region, and the constituent taxa are mistakenly interpreted as having never inhabited that region. In facies mismatching, environmental differences between two regions (possibly unrecognized) cause the inference of a barrier that never existed. The two types of mismatching can work in concert if a facies was originally represented in a single stratigraphic interval that has been completely removed from one region. Analysis of faunal similarity via multivariate analysis of individual localities, coupled with comparison of the regions as single samples, may indicate mismatching if the results differ significantly.

We view these two problems as part of a suite of taphonomic effects that are not evident in paleobiological analyses of smaller geographic scope. First, there is ambiguity in the notion of “barrier,” even when a candidate is obvious. Second, barriers in paleobiogeography are often hidden and must be inferred from their effects rather than observed. Third, stratigraphic and facies mismatching produce effects on regional faunas similar to those produced by barriers. Anyone using barriers to explain faunal disruptions should address these three points.

Upper Eocene faunas of central Florida seem taxonomically distinct from those of the remainder of the Gulf Coastal Plain. This has historically been attributed to a known paleogeographic feature, the Suwannee Strait, which acts as a barrier. The amount of dissimilarity of the echinoid faunas is greater than the amount predicted as a result of sampling problems. Comparison of the results of multivariate and whole-region analyses suggests that mismatching of the two faunas, rather than a true barrier, causes the distinction. Principally facies, but also strata are mismatched. Early Late Eocene faunas inhabited terrigenous sands to the north of the strait and carbonate sands to the south and show the highest distinctiveness. Middle Late Eocene faunas inhabited primarily carbonate sands to the south and both carbonate sands and muds to the north. Overall similarity is higher for both local and regional analyses, and the faunas of northern sands are more similar to those of the southern region than to the northern mud faunas. Overall similarity across the strait is highest in the late Late Eocene strata when both regions had carbonate mud and sand facies. The faunas exhibit greater similarity within facies than they do within regions. Upper Late Eocene strata are poorly preserved north of the strait because of post-Eocene erosion.

The function of serrated teeth is analyzed by experimental comparison with the action of artificially made steel blades. Serrated blades cut compliant materials with a grip-and-rip mechanism, whereas smooth, sharp blades cut by concentrating a large downward force on a tiny area.

Tyrannosaurid teeth from the Cretaceous Judith River Formation bear rows of serrations that have thick, rounded enamel caps, gripping slots between neighboring serrations, thick enamel bodies inside the teeth underneath the gripping slots, and a root beneath each serration. In contrast, the carnivorous dinosaur Troodon has teeth with exposed pointed serrations, thin enamel, and possibly serration roots.

Serrations on the teeth of Troodon and the fossil shark Carcharodon, cut compliant materials in the same way as a serrated hacksaw blade. In contrast, the cutting action of tyrannosaurid teeth most closely resembles that of a dull smooth blade. The spaces between the serrations act as minute frictional vises that grip and hold meat fibers; chambers between neighboring serrations receive and retain small fragments of meat, and inevitably would have acted as havens where bacteria could be stored. These spaces may therefore have led to infections in wounds, analogous to those inflicted by the living Komodo dragon or ora. By analogy, the hunting and feeding behavior of tyrannosaurs may have resembled that of the ora.

Serrations and slots are widely distributed among cutting devices in the natural world, and many of these deserve further study. For example, the carnassial teeth of mammalian carnivores cut by a combination of static force at the cutting edge, a crushing or scissoring action at the advancing junction between upper and lower teeth, and by lateral gripping and compression in a slot, like that seen on a much smaller scale in tyrannosaurid serrations. Mammalian teeth operate well only when deployed with sophisticated control over jaw movement, however, and the fine neural control necessary to operate them may have formed the basis for the later development of intelligence in mammals.

Previously, being interested in mammals was largely a matter of being interested in teeth, whereas being interested in reptiles was largely a matter of being interested in everything but teeth. I suggest that the teeth of at least some reptiles are as rich in information as the teeth of any mammals.

This paper presents a new means of interpreting the distribution of taxa among taxonomic lists. Traditionally, “similarity” indices have been used to compare lists, and “association” measures have been used to compare taxonomic distributions. It is argued that when sampling regimes are poorly understood, all similarity and association indices are unjustifiable. However, observations that taxa have overlapping (conjunct) or nonoverlapping (disjunct) distributions are universally meaningful. Because greater sampling can only increase the number of known conjunctions and because long lists serve as conjunctional Rosetta Stones, conjunction data sets can be far more reliable than the lists that generate them. One way to account for patterns of conjunction and disjunction is to create theoretical spaces composed of two distributional boundaries or edges for each taxon in each dimension. A parsimonious arrangement of edges implies distributions that always overlap when taxa are conjunct and infrequently overlap when taxa are disjunct. A procedure involving correspondence analysis is shown to minimize the number of implied disjunctions. The method is used to analyze a set of 271 Miocene large-mammal genus lists from the Great Plains region. Biostratigraphic and radiometric data demonstrate that the “best” arrangement of distributional edges is temporal and corresponds to an age-range chart. The edge sequence is calibrated to the radiometric time scale and used to compute a detailed genus-level diversity curve and to redefine the boundaries between Miocene Land-Mammal Ages and Subages in the Great Plains. Each of the Miocene Ages can be divided into two or three Subages. Diversity levels are apparently low throughout the Early Miocene and rise sharply during the Middle Miocene. Significant drops in diversity occurred at about 17.4, 11.6, 8.7, and 6.5 Ma. The base of the Barstovian corresponds to the base of the Middle Miocene, and the Clarendonian straddles the Middle-Late Miocene boundary.

The kill curve for Phanerozoic marine species is used to investigate large-body impact as a cause of species extinction. Current estimates of Phanerozoic impact rates are combined with the kill curve to produce an impact-kill curve, which predicts extinction levels from crater diameter, on the working assumption that impacts are responsible for all “pulsed” extinctions. By definition, pulsed extinction includes the approximately 60% of Phanerozoic extinctions that occurred in short-lived events having extinction rates greater than 5%. The resulting impact-kill curve is credible, thus justifying more thorough testing of the impact-extinction hypothesis. Such testing is possible but requires an exhaustive analysis of radiometric dating of Phanerozoic impact events.

A phylogeny of Cretaceous Montastraea-like corals was constructed and used to evaluate the importance of differential speciation, selective extinction, and developmental constraints in the evolutionary history of the clade. Colonies assembled from localities across the central and western Tethyan region were subdivided into four stratigraphic levels: (1) Neocomian to lower Albian, (2) upper Albian to Cenomanian, (3) Turonian to Campanian, and (4) Maastrichtian. Ten corallite characters were measured on transverse thin sections of each colony, and analyzed following a three-step procedure: (1) species were recognized using all-inclusive average linkage cluster analyses (UPGMA) and a series of iterative canonical discriminant analyses; (2) species were separated by stratigraphic level and linked between levels by comparative analysis of the resulting discriminant scores; and (3) ancestor-descendant relationships were interpreted within and among adjacent levels using phenetic and cladistic approaches. A composite tree was then examined with respect to biogeography.

The results suggest that a total of 16 species existed during the Cretaceous, only 2 of which extended between stratigraphic levels. Speciation events between levels 1 and 2 were associated with the radiation of “small-corallite” forms from a “large-corallite” form, as its distribution expanded southward into the Tethyan realm. Limited speciation and stasis predominated among species within the clade between levels 2 and 3. Speciation events between levels 3 and 4 were associated with the radiation of predominantly “large-corallite” forms from a “small-corallite” form, as the clade became restricted to the Caribbean. Extinction initially focussed on “small-corallite” forms, but later shifted to “large-corallite” forms. Morphologic change was constrained between two extremes, the lower limit of which involved a minimum corallite diameter possibly set by developmental or ecological constraints, and the upper by limitations to increases in number of septa. The overall net result was that of nonprogressive evolution within the clade.

Dental measurements are commonly used in the diagnosis of fossil elephant species, yet elephant teeth develop slowly, within a highly dynamic context that enhances opportunities for physical deformation (or its subtler manifestation, quantitative phenotypic variation). This paper examines intraspecific variation in elephant teeth and compares it with variability in other mammals (83 species in 7 orders). I conclude that (1) male elephants tend to have slightly larger cheek teeth than females, though the difference is not marked; (2) of the full complement of six cheek teeth per jaw quadrant, no single tooth consistently varies less than the others (so on this basis, for taxonomic decisions no tooth is preferable to the customarily used M3 = tooth VI); (3) single-population samples vary less than more inclusive, geographically heterogeneous samples of elephant teeth; (4) although differential wear and eruption are important sources of variation in dental measurements, complete elephant teeth are consistently more variable in length and width than the cheek teeth of other mammals; (5) variability in dental dimensions of recognized fossil species of elephants is in general not grossly inconsistent with variability noted in modern elephants, but there are some exceptions. With the information and guidelines presented here, consideration of variability, and assessment of the statistical power permitted by available samples, can enhance confidence and precision in the delimitation of species and provide a firmer basis for macroevolutionary inferences.