NAPC 2001

June 26 - July 1 2001 Berkeley, California

Abstracts, Ka - Ku

Jump to: Kaesler | Kaplan | Karam | Kauffman | Kaufman | Kennedy | Key | Kidwell | Kirchner | Klaus | Kohn | Kosnik | Kowalewski | Kraus | Kurkin

Return to Main Abstracts Page


KAESLER, Roger L., Jill W. Krebs, and Douglas L. Miller, Dept. of Geology, Paleontological Institute, Natural History Museum and Biodiversity Research Center, and Information Technology Services, The University of Kansas, Lawrence, KS, USA

In the late 1960s, classification theory developed two essential points. First, every classification must have a purpose if it is to be effective. Second, a classification can be general purpose only if it incorporates an underlying causality. When a taxonomic classification is also a general-purpose classification, evolution, of course, is the underlying cause that makes it so. The data model that underlies a database is a classification of information, and the same strictures apply to a data model as apply to any classification. Most paleontological databases are means of managing information on collections of fossil specimens. As such, they are likely to be rather special purpose in their scope, but the underlying causal organization of much of their information is, nevertheless, descent with modification. A concept-focused database deals with ideas or such concepts as genera and higher taxa rather than with collections of fossil specimens. A concept-focused database that aims to fulfill a number of needs—to be a general-purpose classification—needs to be organized around a causal influence. The Treatise on Invertebrate Paleontology, whose data are organized taxonomically according to evolutionary causes, is a general-purpose database, albeit one that is intensely textual rather than digital, and is not relational. PaleoBank is an electronic, relational, web-accessible database that is intended to serve ultimately as a digital extension of the Treatise. Its data-capture screens will accept information on taxonomy, morphology, biostratigraphy, paleoecology, biogeography, and bibliography. Our intention is that its use will enhance the value of the Treatise by making its information searchable electronically and available for comparative analysis. Our hope is that it will also be used in research by paleontologists who are not involved in the Treatise project.


KAPLAN, Peter, Museum of Paleontology and Dept. of Geological Sciences, University of Michigan, Ann Arbor, MI, USA

The study of ontogenetic patterns of covariance among morphological traits can yield remarkable insights into the developmental machinery whose evolution produces observed patterns of diversity, evolvability, and directionality ("trends"). Moreover, patterns of developmental integration potentially influence patterns of phenotypic plasticity. E.g., in a well-integrated character complex, the strength of one character's phenotypic response will be governed by the degree to which the response of covarying characters will be "forced" by selection on the one.

The degree of developmental integration can be operationalized as the number of independent parameters required to adequately describe a developmental transformation or a given proportion thereof. This operationalization is made tangible when the characters are anatomical landmark points, and the measure of integration among them is spatial. The fewer the modes of shape change required to adequately describe the developmental transformation of the whole landmark configuration, and the larger the spatial scales at which they appear, the more tightly integrated the development of all anatomical landmarks.

Vermeij has linked functional versatility of a clade to the number of geometric parameters required to adequately describe the clade's variety of form. The parallel between the present formalization of integration and Vermeij's formalization of versatility suggests functional links among integration, plasticity, and versatility. Testing predictions for the evolutionary behavior of species with well-integrated development, however, cannot be extended beyond the species level unless the developmental integration configuration remains stable through speciation events. Although empirical support for this stability is currently lacking, several recent sister-clade comparisons of "evolutionary flexibility" rely tacitly on the stability of developmental integration throughout each clade. The present study of the evolution of morphological integration in the trilobite genus Phacops speaks to just such crucial questions.


KARAM, P. Andrew, Dept. of Environmental Medicine, University of Rochester, Rochester, NY, USA; and Stephen A. Leslie, Dept. of Earth Sciences, University of Arkansas at Little Rock, Little Rock, AR, USA

Most molecular clock models described in the literature assume mutation rates have remained constant with time, while acknowledging this is likely an oversimplification. In fact, most natural environmental mutagens have varied in magnitude throughout the history of life on earth. We started this line of inquiry by examining the effects of changing radiogenic mutation rates over the past 4 Gy, finding that background radiation levels have changed by nearly an order of magnitude in this interval. This was followed by investigations into the effects of changing atmospheric oxygen levels on radiogenic DNA damage because oxygen has a strong modifying effect on radiation damage. We found that low atmospheric oxygen levels early in the history of life helped offset elevated radiation levels, reducing peak radiogenic DNA damage rates to about four times current levels. We also understand that other factors have likely resulted in a background DNA mutation rate that has changed continually, and predictably over the history of life on earth. These include exposure to UV radiation, the development of eukaryotic cells with their attendant oxidative metabolic byproducts, and exposure to cosmic radiation. This paper presents an overview of factors affecting DNA mutations and how these factors have likely changed with time. In some cases, our results are presented quantitatively and, in other cases, semi-quantitatively. We conclude that the area beneath a likely curve for integrated mutations over time is between five and ten times the area beneath a straight line representing modern DNA damage rates, and we suggest that such changes be considered when attempting to reconcile molecular clocks with the fossil record. We also note that this line of inquiry has interesting implications for the field of radiobiology, particularly in assessing the ability of modern organisms to contend with the effects of radiation and other environmental mutagenic agents.


KAUFFMAN, Erle G., Dept. of Geological Sciences, Indiana University, Bloomington, IN, USA

Onshore to offshore paleocommunities react primarily to water depth (shallow vs. deep), energy (active vs. passive), amount of dissolved oxygen in the water column (anoxic, dysoxic, normal), and salt content (subsaline, normal, and hypersaline). Under normal marine conditions, diversity of modern communities follows an onshore to offshore trend of smaller size and moderately-depleted genera, but more species in each genus or genus group. This is a modern icehouse scenario. However, what about the other end of the spectrum, when sea level was elevated by 100 m to 300 m during the Cretaceous greenhouse interval, spreading out the paleobiogeographic units, and eliminating Subarctic/Subantarctic and Arctic/Antarctic conditions? In the Cretaceous of the Western Interior Seaway of North America, depth, oxygen, and salinity had a greater effect than today. Depth regulated oxygen to a greater degree, and oxygen was moderately to highly depleted on transgressions and early regressions. This had a profound effect on community structure, greatly restricting the taxa. Salinity worked in a similar manner. Moving from normal marine to fresh water, number of taxa were moderately decreased, restricted in size, and depleted in diversity, especially generic diversity. Hypersaline species were specifically adapted to elevated salinity, and they were smaller and less numerous than those in subsaline conditions. Both environmental factors and paleocommunities were much more spread out latitudinally due to the protected nature of the basin, the higher Cretaceous sea level, warm climatic conditions, and no permanent ice at the poles. Ocean flow was reversed and moved poleward, carrying larvae and spreading out the paleobiogeographic limits.


KAUFMAN, James H., and Owen M. Melroy, I.B.M. Research Division, Almaden Research Center, San Jose, CA, USA

Ecosystems are evolving dynamical systems in which organisms survive subject to a complex web of interactions. Are ecosystems intrinsically stable or do they evolve to a state of self-organized criticality where mass extinction is an intrinsic property of the dynamics? To study this problem we developed a computer model in which the organisms and their interactions evolve by a "natural selection" process. Our model suggests a third scenario wherein "ecosystems" evolve through a critical state towards an ordered state where global perturbations are required to cause mass extinction. We find evidence for a critical level of diversity at which our "ecosystems" are highly susceptible to extinction. Critical biodiversity is not a point of attraction in the evolution process. Evolution through this critical point to the ordered state is kinetically limited because the susceptibility to extinction is so high near the critical biodiversity. We quantify this behavior in analogy to a physical phase transition and suggest model independent measures for the susceptibility to extinction, order parameter, and effective temperature. We propose that these measures may be applied to natural (real) ecosystems to study evolution and extinction on earth as well as the influence of human activity on ecosystem stability.


KENNEDY, Elizabeth M., Institute of Geological & Nuclear Sciences, Lower Hutt, New Zealand

New Zealand has abundant Late Cretaceous and Tertiary terrestrial sediments, yet relatively little research has been carried out on the leaf fossil floras and their paleoenvironmental implications.

To date three Late Cretaceous and three Paleocene floras have been analyzed using approaches such as Leaf Margin Analysis and multivariate methods. All of these floras are from the South Island of New Zealand: Northwest Nelson (one Late Cretaceous and two early Paleocene assemblages), North Otago (a Late Cretaceous assemblage), Greymouth (a Paleocene assemblage) and the Clarence Valley (a latest Albian-earliest Cenomanian flora).

Paleoclimate estimates from these floras indicate that in the latest Cretaceous, at a paleolatitude of between 50­60°S, the climate was temperate with mean annual temperatures of 12­15°C and growing conditions were generally favorable. In the early Paleocene however, a lowering of diversity and generally smaller leaf sizes suggest less favorable conditions with cooler temperatures. These changes are not dramatic, but they are consistent within the group of floras that have been analyzed. The Clarence flora, which was deposited around the time of the Albian-Cenomanian boundary, has proven to be difficult to interpret in climatic terms. Different foliar methods produced significantly different temperature estimates for this flora and comparison with data from other tools such as general circulation models has proven inconclusive. These results again question how far back in time foliar-based paleoclimate methods can be applied.

Many of the dicotyledonous leaf forms that have been used for these paleoclimate analyses are undescribed species and there is still much to be done on the systematics of the New Zealand K/T macrofloras. Representatives of the Proteaceae and Lauraceae are common elements of both the latest Cretaceous and Paleocene assemblages. Podocarps, araucarians and occasionally ferns are also found in these assemblages, but the dicotyledonous angiosperms are the most dominant feature.


KEY, Heyward M., and Patricia H. Kelley, Dept. of Earth Sciences, University of North Carolina at Wilmington, Wilmington, NC, USA; Gregory P. Dietl, Dept. of Zoology, North Carolina State University, Raleigh, NC, USA; and Thor A. Hansen, Dept. of Geology, Western Washington University, Bellingham, WA, USA

Kelley and Hansen documented "escalation cycles" in naticid gastropod drilling predation from a database of 150,000 specimens (28 Coastal Plain formations). Cycles were punctuated by mass extinctions; drilling increased significantly after the Cretaceous-Tertiary, Eocene-Oligocene and mid-Miocene extinctions, though no such increase followed the Plio-Pleistocene event. We tabulated data on drilling by muricid gastropods from the Kelley-Hansen database to test the hypothesis that cycles occurred in muricid predation. We also collected Recent baseline data on naticid and muricid drilling from a North Carolina mollusc assemblage. Drilling frequencies by epifaunal muricids were much less (0.001 to 0.013 on fossil bivalves) than for (largely) infaunal naticids; samples were dominated by infaunal bivalves and contained few muricids (0­2% of fauna). Drilling frequencies by naticids and muricids were not significantly correlated for bivalve or gastropod assemblages. However, muricid drilling displays its own pattern of cycles of longer periodicity (50 m.y.) than the naticid cycles (maximum 30 m.y.) that do not appear to be correlated with muricid abundance or environmental factors. Muricid drilling on bivalves was low in the Cretaceous (0.002), rose to an Eocene peak (0.013), declined into the Neogene and then increased dramatically in the Plio-Pleistocene (0.012); the Recent bivalve assemblage showed still greater drilling frequencies (0.017). Differences are statistically significant. Muricid drilling on gastropods was low in the Paleogene (0.007) with significantly higher Neogene-Recent values (0.019). However, no clear muricid drilling cycles occur for gastropod prey; patterns are confounded by varying proportions of infaunal and epifaunal taxa. The Recent assemblage displays a high drilling frequency by naticids (0.28), suggesting that a rebound in naticid predation may have followed the Plio-Pleistocene extinction, as previously predicted. In contrast, naticid drilling on Recent gastropods was low (0.044), perhaps because epifaunal prey dominate the assemblage.


KIDWELL, Susan M., Dept. of Geophysical Sciences, University of Chicago, Chicago, IL, USA

Over the past 40 years, the most common approach to quantifying taphonomic preservation potential and bias in estimates of species richness, evenness, etc, has been to compare the composition of the living community with the locally accumulating death assemblage. Marine mollusks have been the focus of the greatest number of such live:dead studies, but it has been difficult to quantify true taphonomic bias because of large among-study differences in environments and in methods of sampling, data collection and analysis. By expanding my search for datasets to the ecological and fisheries literatures, entering sample data for all studies into an electronic database (many authors provided unpublished information to augment their original reports), and using stricter criteria for dataset eligibility (85 habitat-level datasets from 22 studies; only 50% overlap with previous analyses), I can now test for agreement in species rank-order and relative abundance data, and have sufficient statistical power to explore specific effects. Beyond the limiting factor of sample size (datasets having <100 live individuals show wide scatter), I find a large effect from sieve mesh size: 92% of coarse-sieve datasets (>1 mm; 100% of 2 mm datasets), which are composed almost exclusively of adults, show no difference in live and dead species rank-orders when samples are pooled to the habitat level. Average weighted Spearman-r values (Fisher's Zr) are higher among coarse-sieve datasets, and vary with sediment grain size, ranging from 0.5 in gravels, relictual shell-gravels and various high-energy sands, to an impressive 0.8 in muds. This previously unrecognized sieve-size effect makes good sense—choice of sieve-size determines the extent to which data are dominated by the volatile ecological and taphonomic dynamics of larval and early juvenile individuals—and suggests a simple protocol for exploiting molluscan death assemblages in paleo and neontological work. A post-juvenile focus may also be key to isolating high-fidelity data among other metazoan groups.


KIDWELL, Susan M., Dept. of Geophysical Sciences, University of Chicago, Chicago, IL, USA; and M.M.R. Best, Dept. of Geology, University of Toronto, Toronto, ON, Canada

Forty AMS radiocarbon dates of taphonomically-graded bivalve shells from tropical Panama (San Blas Archipelago, 9°N) indicate significant between-facies differences in scales of time-averaging and rates of taphonomic damage. Shells are all from the top 20 cm of unconsolidated seafloor sediment, in >10 m water depth, and from a series of contemporaneous siliciclastic, carbonate, and mixed-composition seafloors. Time-averaging in fine-grained carbonates (micrite, silt/v. fine sand, sandy grassbeds; four sites) is on the order of centuries, whereas in comparable siliciclastics (muds and muddy sands; three sites) it is on the order of thousands to tens of thousands of years; seafloors with intermediate compositions are characterised by intermediate shell-ages (0.5­1 ky maximum; three sites). Multivariate (NMDS) score of taphonomic damage plotted against AMS age indicates that rates of damage accrual (taphonomic clock) are significantly faster for shells in pure-carbonate sediments than in siliciclastic sediments, consistent with environmental differences in the short term (three year) experimental rates of shell deterioration and weight loss found by Best (2000). It is also consistent with pore-water geochemical evidence of carbonate dissolution in carbonates, and of carbonate and other mineral precipitation in siliciclastics, found by our collaborators Ku and Walter (2000; 2001). Thus taphonomic clocks do not keep strict time; rates are much slower in siliciclastics than in carbonates, for reasons both of less bioerosion and more favorable porewater chemistry. Looking across facies, the same damage level signifies dramatically different scales of bias and time-averaging. In a taphonomic Heisenberg principle, the poorest condition assemblages (in pure carbonate sediments) have the shortest scales of time-averaging but have probably suffered the highest proportional loss (bias) in shell content; the best condition assemblages (in siliciclastics) have probably suffered the least proportional loss but have the longest scales of time-averaging (lowest time-resolution).


KIRCHNER, James W., Dept. of Earth and Planetary Science, University of California, Berkeley, CA, USA; and Anne Weil, Dept. of Biological Anthropology and Anatomy, Duke University, Durham, NC, USA

Extracting and interpreting patterns of extinction and origination rates from the fossil record can be challenging; fossil databases generally have coarse resolution and uneven spacing in time, rendering many conventional time-series analysis tools inapplicable. Nonetheless, specialized tools developed for unevenly spaced data can provide insight into the mechanisms linking the biological processes of extinction and diversification.

Cross-correlation of extinction and origination statistics shows that originations significantly lag extinctions by 10 Myr, and this result is robust to the deletion of major events from the record. Yet extinction and origination records are systematically different; extinctions are not autocorrelated over short time scales, while origination statistics are significantly autocorrelated at time scales up to 10 Myr. This suggests that the lag between the two results from an intrinsic property of the process generating originations.

It has been hypothesized that large extinction events are driven by non-biological processes, and thus can occur rapidly, whereas origination rates are subject to biological constraints. Since each new taxon must originate from an already existing one, origination statistics should be autocorrelated, but the fact that this autocorrelation persists up to 10 Myr suggests that evolutionary response is surprisingly sluggish.

Spectral filtering can be used to decompose fossil time series into different time scales of fluctuations. At long wavelengths, originations and extinctions appear closely coupled, possibly due to global factors affecting preservation. The long-wavelength variability of origination rates equals or exceeds that of extinctions. By contrast, over timescales of less than 30 Myr, origination rates are markedly less variable than extinction rates, indicating that originations indeed do not fluctuate as quickly and suggesting that they are more contstrained by biological processes.


KLAUS, James S., Dept. of Geoscience, University of Iowa, Iowa City, IA, USA

Occurrence data of coral species have been used to document a large faunal turnover within the late Pliocene to early Pleistocene (4­1.5 Ma) (Budd and Johnson, 1999). The complex pattern associated with stepwise extinction suggests that while driven ultimately by the closure of the isthmus, selectivity of extinction and resulting turnover was caused by localized environmental changes and linked biological factors.

Roughly 90% of Caribbean coral species present following turnover can be found on modern reef environments; among these the Montastraea annularis species complex is ecologically dominant. By comparing the ecological role of M. annularis-like corals prior to and following faunal turnover insights can be gained into how their ecological/evolutionary history has influenced the structure of modern reef communities. Landmark based morphometric techniques and multivariate methods of community ecology were used to recognize species and study their ecological role in two broad regions of the Dominican Republic; the early Pliocene Arroyo Bellaco locality of the Cibao Valley, and the late Pleistocene six meter terrace exposed on the southern flank of the Island.

The dynamics of early Pliocene Arroyo Bellaco reef communities are in marked contrast to those of the late Pleistocene terraces. Community membership of Arroyo Bellaco is nearly double that found on the terraces, reef facies patterns appear more distinct, and within facies variation in typology is minimal. Five species of M. annularis-like corals were recognized in the Arroyo Bellaco reef communities. Of these five species, none directly correspond to the four species identified within the Pleistocene terraces. All five species identified at Arroyo Bellaco possess similar massive typologies and distribution patterns. Pleistocene representatives show four different typological variations with unique distribution patterns. These results suggest the M. annularis species complex possessed adequate variation during turnover to permit wholesale changes in their morphology and ecological roles.


KOHN, Matthew J., and Jennifer Miselis, Dept. of Geological Sciences, University of South Carolina, Columbia, SC, USA; and Theodore J. Fremd, John Day Fossil Beds National Monument, Kimberly, OR, USA

18O/16O was measured from fossil equid enamel spanning the last ~30 Ma to constrain the paleoclimate history in central/eastern Oregon. These data reveal (1) a long-term systematic depletion in d18O through time, which likely reflects progressive uplift of the Cascade range, and (2) systematic changes in isotope seasonality, the origin of which remains obscure.

Teeth were analyzed from the John Day (19­35 Ma), Quartz Basin (~15 Ma), Mascall (~15 Ma), Juntura (~11 Ma), and Rattlesnake (~7 Ma) Formations, as well as the Glenns Ferry Formation (3.2 Ma) in SW Idaho. A modern horse and cow were analyzed from central Oregon to provide a modern baseline. Median isotope compositions show a decrease in d18O: ~17.5 (V-SMOW) at 27 Ma, ~15 at 15­7 Ma, and ~13 today. A decrease in relative humidity accompanied the d18O decrease: paleosols and paleoflora indicate humidities of ~85% (rainforest) at 27 Ma, ~75% (woodlands) at ~15 Ma, and ~65 (grassland) at 7 Ma. Modern-day humidity is ~55%. Herbivore d18O depends on both local water composition and humidity. A decrease in humidity of ~30% should increase tooth d18O by 3­5. The combined decrease in d18O and humidity thus indicate a decrease in local water compositions by 7.5­9.5. Although global cooling may affect compositions, this large shift more likely reflects progressive uplift of the Cascade Range. Water compositions west of the Cascades are ~8 higher than in central Oregon, because rainout over the range depletes subsequent precipitation in 18O.

Intratooth isotope zoning monitors paleoseasonality, and shows ~5 variability at 27 Ma, ~6­7 at 15­7 Ma, ~2.5 at 3.2 Ma, and ~8 today. The causes of these changes remain enigmatic: perhaps they are caused by the interplay between global fluctuations in mean temperature and changes in the tectonic framework (e.g., connectivity between North and South America, etc.).


KOSNIK, Matthew A., Dept. of Geophysical Sciences, University of Chicago, Chicago, IL, USA; and Peter J. Wagner, Dept. of Geology, Field Museum of Natural History, Chicago, IL, USA

Sampled "diversity" and/or "richness" is effected by both the total number of taxa present, the underlying abundance distribution, as well as the "evenness" of that distribution. This is as true for re-sampling techniques such as rarefaction and bootstrapping as well as for field sampling.

We calculate expected sampled richness given true richness and distribution using geometric, log-normal and Zipf-Mandelbrot distributions which represent a realistic range of hollowness. The first produces the least hollow curve and the last produces the most hollow curves. At the same evenness and true richness, the expected sampled richness is greatest for geometric distributions and lowest for Zipf-Mandelbrot distributions.

We find a set of parameters (defined in terms of relative richness, evenness and distribution) under which comparisons of two samples can be expected to correctly determine that sample A is more "diverse" or "rich" than sample B. The true richness, evenness and distribution all impact our ability to determine the relative richness of sample pairs.


KOWALEWSKI, Michal, Dept. of Geological Sciences, Virginia Polytechnic Institute and State University, Blacksburg, VA, USA; Guillermo A. Avila-Serrano, Facultad de Ciencias Marinas, Universidad Autonoma de Baja California, Ensenada, BC, Mexico; Karl W. Flessa, Dept. of Geosciences, University of Arizona, Tucson, AZ, USA; and Glenn A. Goodfriend, Dept. of Earth and Environmental Sciences, George Washington University, Washington, DC, USA

Because bioclasts undergo extensive temporal mixing, shells scattered in modern coastal and shelf environments are mixtures of individuals that lived anytime between now and the late Pleistocene. When dated using 14C-calibrated amino acid racemization, these shells provide continuous high-resolution records for the last centuries or millennia and can offer direct insights into the pre-industrial history of aquatic ecosystems. They can provide estimates of past shellfish productivity and a benchmark for a quantitative assessment of the current benthos and the efficacy of ecosystem restoration efforts. Also, age structures of dated shells can offer estimates of natural levels of variability in productivity and provide data on multi-centennial dynamics of aquatic ecosystems.

We estimated past productivity of mollusks in the Colorado River Delta prior to upstream irrigation projects that triggered the collapse of the estuarine ecosystem. Over two trillion mollusk shells make up the current shorelines of the delta and range in age from A.D. 950 to 1950 (as indicated by 125 dated valves). A conservative estimate based on these data indicates that, in times of natural river flow, an average density of ~50 mollusks per m2 thrived on the delta. The present abundance of shellfish is almost 20 times lower (3 per m2, 1999­2000). The results testify to the severe loss of productivity resulting from diversion of the river's flow and the inadequacy of its partial resumption (1981 to present). Computer models of the age structure of dated shells suggest that the productivity remained relatively constant between A.D. 950 and 1950.

Paleontology can provide a reference baseline regarding the pre-industrial state of benthic ecosystems. We can use paleoecological techniques to improve our ability to evaluate the current state of aquatic habitats, forecast their future, and guide their restoration.


KRAUS, William F., Digital Biology, Moss Beach, CA, USA

The entertainment industry, through theatrical releases such as "Jurassic Park" and "The Lost World," and television programs such as The Discovery Channel's "Walking With Dinosaurs," has popularized the use of computer graphics as a medium for illustrating both the appearance and behavior of extinct fauna. While visually stunning, these animations are highly scripted due to the fact that (i) the rendering of such photo-realistic images requires significant computational resources making real-time interaction impractical, and (ii) the behaviors and events depicted in the animations are designed a priori to fit into a pre-determined story line.

On the contrary, real-time interaction, achieved primarily through a selective reduction in rendering quality, has been a central theme of the computer games industry. In fact, much of the current effort in game design is focused on increasing the interactivity of games by developing physics-based environments with autonomous characters.

The purpose of this session is to demonstrate how interactive simulations of extinct fauna can be created on desktop computers by synthesizing optimizations taken from the computer games industry with concepts from artificial life research. Specifically, interactive reconstructions of the Burgess Shale fauna, created by combining real-time mesh deformation with a set of behavioral controllers and a simple physics model, will be presented. These simulations can easily be deployed on the World Wide Web, and have the potential to be used for scientific illustration and hypothesis testing.


KURKIN, Andrey A., Paleontological Institute, Russian Academy of Sciences, Moscow, Russia

Anomodont therapsids are known from 40 localities in Eastern Europe. These include basal forms, such as members of the suborder Dromasauria, as well as derived dicynodonts.

In the evolution of Permian anomodonts of Eastern Europe four main stages are apparent: (1) Dinocephalian Superassemblage, characterized by the existence of an endemic fauna of basal anomodonts; (2) the main part of the Sokolki Assemblage (Kotelnich and Ilinscoe Subassemblages); characterized by the appearance of Gondwanan elements in Eastern Europe. Low-skulled dicynodonts are the dominant anomodonts in the fauna; (3) the end of the Sokolki Assemblage (Sokolki Subassemblage); characterized by the dominance of high-skulled dicynodonts; (4) Vyazniki Assemblage; characterized by high-skulled dicynodonts that possess some features typical of Triassic dicynodonts.

These data suggest the existence of a faunal exchange between Gondwana and Eastern Europe at the time of the boundary between the Ulemosaurus svijagensis and the Deltavjatia vjatkensis Zones (= Tapinocephalus-Pristerognathus Assemblage Zone boundary). During this period in Eastern Europe part of the endemic anomodont fauna disappeared, and the basal anomodont Suminia appeared. Dicynodonts also first occur in Eastern Europe at this time. During this period in South Africa the primitive dromasaurs disappeared, and the endothiodonts, which possess some features similar to the venyukoviids of Russia, appeared. After this time, faunal exchange appears to have been rare, and an endemic radiation of Russian dicynodonts occurred. These dicynodonts repeated many of the main evolutionary trends of South African dicynodonts. For example, some members of the Dicynodontidae that appear in the Sokolki Assemblage are convergent with South African forms such as Diictodon or Pristerodon. These taxa may have occupied an ecological role similar to that of the basal anomodont Suminia, and the presence of Suminia in Russia during the period of faunal exchange may explain why Diictodon and Pristerodon failed to disperse to Eastern Europe.