Michal Kowalewski

High-resolution estimates of temporal mixing within shell beds: the evils and virtues of time-averaging

This study quantifies the fine structure of time-averaging by using large samples of dated shells collected from within individual strata. Time-averaging results in both good and bad news for interpreting bioclastic deposits. Nine samples of shells were collected from four Holocene cheniers on the Colorado Delta (Gulf of California) and 165 shells of the bivalve Chione fluctifraga were dated using 14 C calibrated amino acid racemization (D-alloisoleucine/L-isoleucine). The age range of shells within samples averages 661 years and, in seven out of nine samples, exceeds 500 years. The sample standard deviation ranges from 73 to 294 years and averages 203 years, far exceeding the dating errors (<100 years) and potential variation in the life span of Chione (<10 years). Time-averaging is homogeneous among strata within cheniers but varies significantly among cheniers. Age-distributions of dated shells indicate that at 50-year resolution, the samples provide a continuous and uniform record for the entire interval. The actual sample completeness (63.6%) is very close to that predicted by simulations of sampling a 100% complete, uniform record (67.3%). The bad news is that, no matter how carefully collected, data from shell beds may not be suitable for studying processes on timescales shorter than 10 2 to 10 3 years; explanations for faunal change that invoke reasoning or models derived from a strictly ecological point of view may rarely be justifiable. Also, notable differences in temporal resolution between the shell beds of seemingly identical origin imply that paleontological patterns (e.g., species diversity) may be affected by cryptic variation in time-averaging. The comparison of our data with time-averaging estimates obtained from other cheniers at coarser sampling resolutions indicates that pooling of samples (analytical time-averaging) can significantly reduce the temporal resolution of paleontological data. The good news is that shell beds can record the optimal type of time-averaging: where paleobiological data are a time-weighted average of the faunal composition from the spectrum of environments that existed during the entire interval of time. Samples from single strata provide a long-term record that is representative of the predominating environments. Within the range of 14 C dating, shell beds can provide a complete, high-resolution record, and thus may offer exceptional insights into the environmental and climatic changes of the last 40 thousand years.

A fossil record full of holes: The Phanerozoic history of drilling predation

The evolutionary history of drilling predation, despite a long and rich fossil record (Precambrian–Holocene), contains a 120 m.y. gap (Late Triassic–Early Cretaceous). Drilled bivalve and brachiopod shells from Jurassic deposits of Hungary, India, and four localities documented in the literature indicate that drillers may have existed continuously throughout the Mesozoic. They may have been descendants of Paleozoic predators, unknown Mesozoic carnivores, or precursors of modern drillers. A literature database suggests three major phases in the Phanerozoic history of drilling predators: (1) the Paleozoic phase (latest Precambrian–Carboniferous) dominated by rare to moderately frequent drillings in brachiopods and sessile echinoderms; (2) the Mesozoic phase (Permian–Early Cretaceous) with very rare, or even facultative, drillers that had little impact on marine benthic communities, but nevertheless may have been present continuously; and (3) the Cenozoic phase (Late Cretaceous –Holocene) dominated by frequent gastropod drillings in mollusks.

Time-Averaging, Overcompleteness, and the Geological Record

Time-averaging makes events that happened at different times appear to be synchronous in the geological record. It can generate overcompleteness by concentrating objects (e.g., fossils, clasts) from a longer time interval into a unit that formed during a shorter interval. Time-averaging results both from extrinsic agents (e.g., vertical mixing, condensation, analytical pooling of data) and the intrinsic properties of the fossilized objects (preservational potential and abundance). Variation in intrinsic and extrinsic factors results in disharmonious time-aver aging: different taxa preserved together may be time-averaged over different time-scales. Because extrinsic and intrinsic factors vary in time and space, the geological record is characterized by vertical and lateral variations in time-averaging. The importance of time-averaging depends on the time-aver aging threshold, which is determined by the time-scale of the process of interest. Significant time-aver aging occurs when the scale of temporal mixing exceeds the time-scale of the process. Insignificant time-averaging occurs under opposite conditions. Accordingly, the consequences of time-averaging can be viewed as a function of the time-averaging threshold. Insignificant time-averaging eliminates noise by erasing short-term fluctuations and enhancing persistent signals. Significant time-averaging reduces resolution by the cumulative averaging of short-term signals. It may also generate false patterns by making diachronous events appear to be synchronous.

Taphofacies analysis of recent shelly cheniers (beach ridges), northeastern baja california, Mexico

SummaryThis report presents the results of taphofacies analyses of shelly cheniers (mollusk-dominated lag-concentrations) from the tidal flats of northeastern Baja California, Mexico. The three generations of moderm (formed during last 70 years), submodem (younger than 1,500 BP), and subfossil (5,000–2,000 BP) cheniers can be distinguished by their position relative to the shoreline, their topography, and the radiocarbon-age of their shells. The generations differ in the duration and complexity of their taphonomic history. Sixty-one samples from nine localities were collected to test the utility of the taphofacies approach for studying chenier-type shell deposits. The three chenier generations, although all dominated by the bivalve molluskMulinia coloradoensis, differ significantly in their taxonomic composition due to taphonomic and/or biologic factors. The taphofacies analysis included 4,334 specimens ofM. coloradoensis described by nine taphonomic variables. Univariate analysis of those variables indicated that the shells that accumulated in the cheniers are little-affected by biological processes (bioerosion, encrustation), and moderately affected by physical processes (fragmetation, cracking, peeling, edge preservation). Only the luster features of shells (external luster, internal luster, and internal features) vary substantially and consistently with chenier age —a result of subaerial weathering. Multivariate taphofacies analysis discriminates the three generations of cheniers even when the poorly preservable luster variables are excluded from the analysis. This suggests that taphofacies discrimination is possible for fossil cheniers. The shells collected from the chenier surface have substantially poorer preservation than shells from the subsurface, indicating that taphonomic degradation in the chenier plain environment is a surface phenomenon. Chenier plain shelly assemblages are taphonomically distinct from assemblages formed in other marine environments: they have a very low frequency of macroscopically recognizable bioerosion and encrustation. The existence of preservable taphonomic differences between the cheniers that differ in their age (i.e., duration of preburial history), suggests that fossil lag concentrations may be useful in detecting incompleteness gradients along stratigraphic boundaries. A ‘taphonomic clock’—a correlation between a ‘time-sincedeath’ and shell preservation—was found only for luster features, taphonomic attributes that are unlikely to be preserved in the fossil record.

Morphometric analysis of predatory drillholes

Abstract Quantitative studies of post-Jurassic predatory drillholes assume that drill holes were made by either naticid or muricid gastropods. However, other potential predators can leave similar traces, and the traditional method of discrimination (based on differences in vertical shape of drillholes: cylindrical vs. countersunk) is not always applicable. Poor preservation or thin shells may prevent identification. A multivariate morphometric approach, including principal component analysis (PCA), canonical variate analysis (CVA) and size-free discriminant analysis, is suggested here as a supplementary method. I test the approach by analyzing 98 drillholes made in the valves of the thin-shelled bivalve Mulinia coloradoensis collected from subfossil deposits in northeastern Baja California. Using traditional criteria 10 (10.2%) of 98 drillholes were not correctly identified. Morphometric analyses indicate that in planar view muricid and naticid drillholes do not differ in shape. However, they perfectly discriminate in size. The cylindrical drillholes (Oichnus simplex) attributed to muricid predators are significantly smaller than countersunk ones (Oichnus paraboloides) made most likely by naticid gastropods. The results do not necessarily undermine the validity of the traditional ichnotaxonomy: muricids tend to produce smaller drillholes than naticids. Nevertheless, the vertical shape of a drillhole may be a function of both the type of predator and the size of hole drilled. More than half of the unidentified drillholes (6 out of 10) from an independent group characterized by a more oval shape (octopod drilling?). The traditional method, supported by the morphometric methods employed here,is a powerful tool for analysis of drillholes. It may be valuable not only as a tool for analysis of muricid and naticid drillholes but also as a method for comparative analysis of drillholes of unknown source found throughout the Phanerozoic fossil record.

Phenetic discrimination of biometric simpletons: Paleobiological implications of morphospecies in the lingulide brachiopod Glottidia

The extreme morphological simplicity of lingulide brachiopod shells makes them particularly useful for investigating the species-level taxonomic resolution of the fossil record as well as the relationships between taxonomy, morphological complexity, and evolutionary rates. Lingulides have undergone little change in shell morphology and have had low taxonomic diversity since the Paleozoic. Is this pattern an evolutionary phenomenon or an artifact of the shell9s simplicity? Multivariate methods were used to establish morphogroups among seven populations of four extant species of Glottidia. Six characters (three shell dimensions and three internal septa) were measured for 162 specimens from field and museum collections. All populations follow similar allometric trajectories: internal septa display positive allometry and shell dimensions display negative allometry. The allometric pattern may reflect D9Arcy Thompson9s Principle of Similitude. Principal component analysis does not reveal any distinct clusters in Glottidia morphospace but suggests that some differences independent from ontogeny exist among the populations. Size-free canonical variate analysis indicates the presence of five size-invariant groups that are statistically distinct. Bootstrap-corrected error rates indicate that four specimens are enough to classify a sample correctly at alpha = 0.05 and eight specimens at alpha = 0.01. The groups are consistent with neontological classification with the exception of two populations of G. pyramidata identified by discriminant analysis as two distinct groups. The size-free morphogroups reflect geographic separation rather than ontogenetic or substrate differences among the populations. Despite the morphological simplicity of the shell, size-free multivariate analysis of Glottidia delineates groups that offer taxonomic resolution comparable with the neontological classification. The method offers a promising tool for identifying natural morphogroups on the basis of few morphological characters. Moreover, the agreement between neontological taxonomy and the morphogroups suggests that the size-free approach can be applicable for evaluating the reality of the low diversity and turnover rates observed in the fossil record of lingulide brachiopods (= Family Lingulidae). Assuming that the neontological species of Glottidia are biologically meaningful, this study shows that morphological simplicity of lingulides does not necessarily result in taxonomic underresolution. Our analysis, as well as several previous case studies, suggests that taxonomic diversity and turnover rates do not have to be dependent on the morphological complexity of preservable parts. In many cases, when rigorous quantitative methods are employed, the differences in the rates of morphological evolution may be a real evolutionary phenomenon and not artifacts of morphological complexity.

Taphonomy of a living fossil; the lingulide brachiopod Glottidia palmeri Dall from Baja California, Mexico

Glottidia palmeri Dall, a Recent lingulide brachiopod, inhabits intertidal flats of the Colorado Delta plain in Baja California, Mexico. Live lingulides occur in patches across the intertidal zone. Patches are dominated by single cohorts with unimodal size-frequency distributions. The oldest cohorts appear to be between four and five years old. Dead specimens of G. palmeri occur in three modes: (1) In situ shells remaining after death in their burrows. These undergo decay and maceration, and blacken, soften, and eventually disintegrate. They persist up to at least a year and represent a slightly time-averaged, but strictly in place record of a single patch: (2) Surface shells removed from their burrows be-. fore or soon after death. These are deposited on the surface as single valves and undergo quick mechanical disintegration. They disappear within days, or at most weeks, and are a synchronous but spatially mixed record of different patches coexisting on the flat: (3) Beach shells represented by dried shells with mummified soft bodies. These specimens are washed out by storms and deposited on supratidal beach ridges. They disappear within weeks to months and are a synchronous, but spatially mixed record of different patches. This study offers criteria for the recognition and interpretation of preservation modes in fossil lingulides. Moreover, it confirms previous notions that Recent lingulide brachiopods have a very low fossilization potential and suggestions that high to catastrophic sedimentation rates are needed to preserve them. Thus, large-scale aspects of lingulide history (e.g., global taxonomic diversity, stratigraphic ranges) must be poorly recorded in the fossil record. At the same time, extensive time-averaging or significant transport of lingulide shells are highly unlikely. Therefore, on those rare occasions when lingulide shells are preserved, they record outcropscale, short-term aspects of their history with high fidelity. The literature on the Paleozoic lingulides suggests that they had a higher fossilization potential than their post-Paleozoic relatives. Thus, this study is most applicable to Mesozoic and Cenozoic lingulides. Most importantly, this study suggests that the fossil record of the lingulides may be affected by a taphonomic megabias-their decrease in diversity and ecologic importance after the Paleozoic may be, in part, a taphonomic artifact.

The Temporal Resolution of Epibiont Assemblages: Are They Ecological Snapshots or Overexposures?

The effects of time averaging on the fossil record of soft‐substrate marine faunas have been investigated in great detail, but the temporal resolution of epibiont assemblages has been inferred only from limited‐duration deployment experiments. Individually dated shells provide insight into the temporal resolution of epibiont assemblages and the taphonomic history of their hosts over decades to centuries. Epibiont abundance and richness were evaluated for 86 dated valves of the rhynchonelliform brachiopod Bouchardia rosea collected from the inner shelf. Maximum abundance occurred on shells less than 400 yr old, and maximum diversity was attained within a century. Taphonomic evidence does not support models of live‐host colonization, net accumulation, or erasure of epibionts over time. Encrustation appears to have occurred during a brief interval between host death and burial, with no evidence of significant recolonization of exhumed shells. Epibiont assemblages of individually dated shells preserve ecological snapshots, despite host‐shell time averaging, and may record long‐term ecological changes or anthropogenic environmental changes. Unless the ages of individual shells are directly estimated, however, pooling shells of different ages artificially reduces the temporal resolution of their encrusting assemblages to that of their hosts, an artifact of analytical time averaging.

Improving with age: the fossil record of lingulide brachiopods and the nature of taphonomic megabiases

Variation in the actual preservation of fossils can be used to detect taphonomic megabiases—large-scale distortions caused by changes in the quality of the fossil record. For example, temporal changes in the taphonomy of lingulide brachiopods suggest that Paleozoic lingulides had a higher fossilization potential than the post-Paleozoic ones. The frequency of stratigraphic occurrences of lingulides covaries with their fossilization potential. Thus, the apparent decline in the importance of lingulides in Phanerozoic communities may be due to changes in their fossilization potential. Taphonomic megabiases may be very common. We distinguish four major types of megabiases: within taxon, static among taxon, dynamic among taxon, and global.

Shell Beds as paleoecological puzzles: A case study from the Upper Permian of the Paraná Basin, Brazil

SummaryMicrostratigraphic, sedimentological, and taphonomic features of the Ferraz Shell Bed, from the Upper Permian (Kazanian-Tatarian?) Corumbataí Formation of Rio Claro Region (the Paraná Basin, Brazil), indicate that the bed consists of four distinct microstratigraphic units. They include, from bottom to top, a lag concentration (Unit 1), a partly reworked storm deposit (Unit 2), a rapidly deposited sandstone unit with three thin horizons recording episodes of reworking (Unit 3), and a shell-rich horizon generated by reworking/winnowing that was subsequently buried by storm-induced obrution deposit (Unit 4). The bioclasts of the Ferraz Shell Bed represent exclusively bivalve mollusks.Pinzonellaillusa andTerraia aequilateralis are the dominant species. Taphonomic analysis indicates that mollusks are heavily time-averaged (except for some parts of Unit 3). Moreover, different species are time-averaged to a different degree (disharmonious time-averaging). The units differ statistically from one another in their taxonomic and ecological composition, in their taphonomic pattern, and in the size-frequency distributions of the two most common species. Other Permian shell beds of the Paraná Basin are simílar to the Ferraz Shell Bed in their faunal composition (they typically contain similar sets of 5 to 10 bivalve species) and in their taphonomic, sedimentologic, and microstratigraphic characteristics. However, rare shell beds that include 2–3 species only and are dominated by articulated shells preserved in life position also occur. Diversity levels in the Permian benthic associations of the Paraná Basin were very low, with the point diversity of 2–3 species and with the within-habitat and basin-wide (alpha and gamma) diversities of 10 species, at most. The Paraná Basin benthic communities may have thus been analogous to low-diversity bivalve-dominated associations of the present-day Baltic Sea. The ‘Ferraz-type’ shell beds of the Paraná Basin represent genetically complex and highly heterogeneous sources of paleontological data. They are cumulative records of spectra of benthic ecosystems time-averaged over long periods of time (102–104 years judging from actualistic research). Detailed biostratinomic reconstructions of shell beds can not only offer useful insights into their depositional histories, but may also allow paleoecologists to optimize their sampling designs, and consequently, refine paleoecological and paleoenvironmental interpretations.