Articles

Print Email

RNames, a stratigraphical database designed for the statistical analysis of fossil occurrences – the Ordovician diversification as a case study

Björn Kröger and Kari Lintulaakso

Plain Language Abstract

RNames is an open access relational database, in which opinions are collected about the relative age of rock units (stratigraphic units) published in peer reviewed scientific papers. Currently nearly 400 references are compiled that express nearly 25,000 opinions on more than 4000 named units. The database is constructed in such a way that all opinions can be used by an algorithm to bin the units in a variety of unified time scales. The resulting lists or tables with time-binned stratigraphic units can be matched with other databases that contain palaeontological data, such as fossil occurrences, because these databases are based on the same named rock units. This matching of databases allows for a simultaneous calculation of estimates, such as palaeobiodiversity in different time frames. Here we chose the rock units that belong to the Ordovician Period (485-444 m.y.a.) as a case study and demonstrate that the choice of the time binning is crucial for the shape of the resulting palaeobiodiversity trends. Our estimates confirm earlier estimations, which indicated that the diversity increase during the Ordovician was strongest within the Middle Ordovician (467-458 m.y.a.).

Resumen en Español

RNames, una base de datos estratigáfica diseñada para el análisis estadístico de ocurrencias fósiles – la diversificación ordovícica como un caso de estudio

RNames (http://rnames.luomus.fi/) es una base de datos de acceso libre que relaciona entre sí unidades estratigráficas que son consideradas equivalentes temporales o que se superponen temporalmente. RNames también es una herramienta para correlacionar entre unidades estratigráficas. La estructura de la base de datos permite un amplio rango de consultas y aplicaciones. Actualmente se encuentran disponibles tres algoritmos los cuales calculan un conjunto de tablas de correlación con unidades estratigráficas ordovícicas almacenadas en rodajas cronoestratigráficas de alta resolución (Pisos Globales Ordovícicos, Rodajas de Pisos o Stage Slices, Rodajas Temporales o Time Slices).

La facilidad de disponibilidad de unidades estratigráficas almacenadas de modo diferente y el potencial de crear nuevos esquemas son las principales ventajas y objetivos de RNames. Las distintas unidades estratigráficas almacenadas por intervalo temporal pueden ser combinadas con otras bases de datos y permitir análisis simultáneos actualizados de estimaciones estratigráficamente restringidas en distintos esquemas. Ejemplificamos estas nuevas posibilidades con nuestros datos ordovícicos compilados y analizamos colecciones fósiles de la Base de Datos Paleobiology en base a tres esquemas almacenados distintos. Las curvas de diversidad presentadas son las primeras curvas globales marinas a nivel subpiso para el Ordovícico. Una comparación entre las curvas revela que las diferencias en la partición en rodajas temporales tienen un gran efecto en la forma de la curva. A pesar de incertidumbres en las diversidades del Ordovícico Temprano y Tardío, nuestros cálculos confirman las estimaciones previas de que la diversificación ordovícica tuvo un clímax global durante el piso Darriwiliano.

Palabras clave: base de datos relacional; estratigrafía; almacenamiento temporal; palaeobiodiversidad; GOBE

Traducción: Diana Elizabeth Fernández

Résumé en Français

RNames, une base de données stratigraphique pour l'analyse statistique des occurrences fossiles – la diversification ordovicienne comme cas d'étude

RNames (http://rnames.luomus.fi/) est une base de données relationnelle en accès libre, reliant les unités stratigraphiques qui sont considérées équivalentes ou qui se recoupent temporellement. RNames est aussi un outil pour corréler les unités stratigraphiques entre elles. La structure de la base de données permet une grande variété de requêtes et d'applications. Actuellement, trois algorithmes sont disponibles et calculent un ensemble de tables de corrélation avec des unités stratigraphiques ordoviciennes classées dans des tranches chronostratigraphiques à haute résolution (« Global Ordovician Stages, Stage Slices, Time Slices »). L'accès facilité aux unités stratigraphiques classées de différentes manières et la possibilité de créer des nouvelles combinaisons représentent les principaux avantages et buts de RNames. Des unités stratigraphiques combinées sous différentes catégories temporelles peuvent être couplées avec d'autres bases de données et ainsi permettre des analyses simultanées à jour des estimations contraintes stratigraphiquement sous différents schémas. Nous illustrons ces nouvelles possibilités avec nos données compilées pour l'Ordovicien et analysons les collections fossiles de la « Paleobiology Database » en se basant sur les trois différents schémas de classement. Les courbes de diversité présentées sont les premières courbes de diversité marines globales au niveau du sous-stade pour l'Ordovicien. Une comparaison des courbes révèle que les différences de définition des tranches temporelles ont un effet majeur sur la forme de la courbe. Malgré des incertitudes sur les diversités à l'Ordovicien ancien et récent, nos calculs confirment les estimations précédentes que la diversification ordovicienne a atteint son point culminant pendant le stade du Darriwilien.

Mots-clés : base de données relationnelle ; stratigraphie ; classement par tranches de temps ; paléobiodiversité ; GOBE

Translator: Antoine Souron

Deutsche Zusammenfassung

RNames, eine stratigrafische Datenbank zur statistischen Analyse von Fossil-Vorkommen – die ordovizische Diversifizierung als Fallstudie

RNames (http://rnames.luomus.fi/) ist eine frei zugängliche, relationale Datenbank, die zeitgleiche oder zeitlich überlappende stratigraphische Einheiten miteinander verbindet. Zusätzlich kann mit RNames zwischen den stratigrafischen Einheiten korreliert werden. Die Struktur der Datenbank erlaubt eine große Auswahl an Eingaben und Anwendungen. Momentan sind drei Algorithmen verfügbar, die einen Satz Korrelationstabellen mit ordovizischen stratigraphischen Einheiten kalkulieren und mit Zeit-Bins in hochauflösenden chronostratigraphischen Schichten. (Globale ordovizische Stufen, Stufenschichten, Zeitschichten).

Die einfache Verfügbarkeit der verschieden gebinnten stratigrafischen Einheiten und das Potential neue Programme zu erstellen, sind die Hauptvorteile von RNames. Verschiedene Zeit-binned stratigrafische Einheiten können verschiedenen anderen Datensätzen zugeordnet werden und erlauben simultane aktuelle Analysen stratigrafisch gebundener Schätzungen in verschiedenen Programmen. Wir veranschaulichen diese neuen Möglichkeiten mit unseren kompilierten ordovizischen Daten und analysieren Fossilsammlungen der paläobiologischen Datenbank mit drei unterschiedlichen Binning-Schemata. Die gezeigten Diversitäts-Kurven sind die erste Teilstufen-Ebene, globale, marine Diversitätskurven für das Ordovizium. Ein Vergleich zwischen den Kurven zeigt auf, dass unterschiedliches Zuordnen der Zeitscheiben einen großen Einfluss auf die Form der Kurve hat. Trotz Unsicherheiten bei früh-und spätordovizischen Diversitäten, bestätigen unsere Berechnungen frühere Annahmen, dass ordivizische Diversifikation während des Darriwiliums global den Höhepunkt erreicht hatte.

Schlüsselwörter: relationale Datenbank; Stratigraphie; Zeit Binning; Paläobiodiversität; GOBE

Translator: Eva Gebauer

Arabic

Translator: Ashraf M.T. Elewa

 

 

author1Björn Kröger. Finnish Museum of Natural History, University of Helsinki, P.O. Box 44, Fi-00014, Helsinki, Finland, This email address is being protected from spambots. You need JavaScript enabled to view it.

Björn Kröger is a curator of the Palaeontological Collections of the at Finnish Museum of Natural History, Helsinki, Finland. He works primarily on the palaeobiology and taxonomy of early cephalopods. His research interests include macroevolutionary processes of the Paleozoic. Currently his research focus is on early Paleozoic reef evolution and processes of diversification during the Ordovician.

divider

author2Kari Lintulaakso. Finnish Museum of Natural History, University of Helsinki, P.O. Box 44, Fi-00014, Helsinki, Finland, This email address is being protected from spambots. You need JavaScript enabled to view it.

Kari Lintulaakso is a PhD student at the University of Helsinki and has MSc of biosciences from the same university. He specializes in recent mammals and his main interest is linking current mammalian community structures with key environmental factors that can be used in palaeoclimatological and environmental studies. In these studies he exploits his over 20 years of work experience with databases, including data mining, data integration, and ETL (Extract, Transform, Load). 

 

TABLE 1. Correlation coefficients between series of proportional change (Δt/t-1) of subsampled diversity based on stratigraphic opinions with one-time bin versus two-time bin minimum resolution. Generalised differencing was used to reduce effects of autocorrelations (McKinney, 1990; Alroy, 2000; Novack-Gottshall and Miller, 2003). Abbreviations: StS: Stage Slices, Bergström et al. (2009); TS: Time Slices, Webby et al. (2004).

  Pearson Spearman
TS resolution one / TS resolution <two 0.940 0.918
StS resolution one /StS resolution <two 0.951 0.818
Stage resolution = one /Stage resolution <2 0.988 0.886
 

FIGURE 1. Simplified structure of the RNames Database (rnames.luomus.fi/). The database contains eight related tables (blue and red objects) of which the object “Relations” is central. In “Relations” correlated stratigraphic units are listed by reference. Three output tables (yellow objects) list time binned stratigraphic units based on a search algorithm that uses “Relations” via R-Package RMySQL (the scripts are available under github.com/bjoekroe/RNames). Global Stages after Cooper et al. (2012). Abbreviations: ID, identifier; StS, Stage Slice (Bergström et al., 2009); TS, Time Slice (Webby et al., 2004)

 figure1

FIGURE 2. Structure of algorithm for time binning of stratigraphical units of the RNames Database (available under github.com/bjoekroe/RNames). Time bins are selected via three correlation routes (colour codes) and six rules resulting in six tables with referenced bins from which only those are selected which are most precise (i.e., range through lowest number of bins). Abbreviations: bio.unit, biostratigraphic unit; non-bio. unit, non-biostratigraphic unit. Colour code: red, correlation exclusively based on biostratigraphy; orange; correlation indirectly based on biostratigraphy; yellow, correlation based on direct or indirect assignments to time bins. -> arrow refers to referenced relations in RNames.

figure2 

FIGURE 3. Example of the time binning of the Plectodina tenuis conodont-zone into Stage Slices (StS, Bergström et al., 2009) based on records of referenced relations in RNames and on the binning algorithm (see text and Figure 2). The P. tenuis zone is binned into the StS Sa2 time bin because three references relate the P. tenuis Zone exclusively to the Baltoniodus alobatus conodont zone, which is related to StS Sa2, based on reference 9 (Ainsaar et al., 2010). (see text for further explanation). Red circles denote selected references. Red lines denote selected relations. 1Webby et al. (2004); 2Cooper et al. (2012); 3Sweet (1984); 4Saltzman et al.(sup); 5Lehnert et al. (2005); 6Goldman et al. (2007); 7Sell et al. (2015); 8 Korén et al. (2006), 9Ainsaar et al. (2010); 10Bergström et al. (2009); 11Sennikov et al. (2014); 12Kanygin (2010); 13 (Bergström et al., 2012).

 figure3

FIGURE 4. Ordovician genus-level diversity trends of PaleobioDB occurrence data, based on three different time binning approaches. 1. Total mean standing diversity (after Cooper, 2004). 2. Rarefied diversity with time bins of < 100 collections culled, with quota 600. Diamonds, two-time-bin resolution; triangles, one-time bin resolution; stars, all collections. Red, Global Stages after Cooper et al. (2012), green; Stage Slices, Bergström et al. (2009); blue, Time Slices, Webby et al. (2004). Error bars reflect 95% confidence interval.

 figure4

FIGURE 5. Quality of PaleobioDB data used for diversity calculations. 1. Number of collections available per time bin. 2. Mean stratigraphic range of collections through time bins. Diamonds, two-time-bin resolution; triangles, one-time bin resolution; squares, all collections. Red, Global Stages after Cooper et al. (2012), green; Stage Slices, Bergström et al. (2009); blue, Time Slices, Webby et al. (2004).

figure5