From Morphology to Functionality - Development and application of a Bivalve Burrowing Simulation

Burrowing robot

ProjectLeaders
Dr. Peter Eggenberger Hotz
Artificial Intelligence Laboratory
Universität Zürich

Dr. Wolfgang Schatz
Universität Luzern

Abstract
In this interdisciplinary project with scientists from paleontology and computer science, one tries to deepen the understanding of evolution of bivalves using computer simulated evolution. These animals live in different habitats, where they hide in different soils. The differences of these habitats, let them develop different digging behaviors and different shells to hide from predators. In order to understand the different selection pressures leading to different shell shapes as well as digging behavior, we simulate different bivalves in simulation and try to evolve different kinds of bivalves. These evolved results are then produced in a 3D-printer and also tried out with a digging robot consisting of the shells and an actuating mechanisms. The obtained results are then compared with living bivalves as well as with fossils.

The project aims towards two different goals: In Paleontology we hope to discern in simulation the effects of the selection pressure for "good* digging on the shape of the shells and the digging behavior. In computer science we hope to deepen our understanding of evolution in order to construct better artificial evolutionary systems.

Funding:
SNF Project no. 113934

Duration
2008-2010

to top

Combining Genetical, Developmental, and Evolutionary Processes in an Artificial System to Evolve the Morphology of the Precambrian Dickinsonia costata

Dickinsonia costata Abstract
The goal of this project was the development of an artificial system in order to examine the evolutionary relations between genome changes and external morphology. This goal was achieved by combining genetical, developmental, and evolutionary processes to evolve the morphology of the Precambrian fossil Dickinsonia costata. On one hand computer science may provide a reliable tool to investigate aspects of morphology and ecology of a fossil organism, which cannot be gained by the fossil record. On the other hand, paleontology provides an insight into the processes of morphological changes in time (e.g. anagegesis, evolutionary pathways). Therefore, paleontology improves our knowledge about evolutionary processes. In this work collaboration between computer science and paleontology is established using the advantages of one another.
The artificial evolutionary model is based on cells each equipped with a genome. Each gene within the genome is linked to a function or a product (analog of a structural gene) and is regulated by one or several regulatory units. Each cell will perform its linked function (cell division and cell adhesion) depending on the subset of active genes. Cellular interactions implemented as viscoelastic elements are simulated. A differential genetic regulatory network -controlling the developmental processes - forms the core of the artificial evolutionary system. It is coupled to a simulator, which describes intercellular forces.
In the field of paleontology and computer science further applications of the herein proposed artificial evolutionary system are conceivable like investigations on phenotypic plasticity, simulating locomotion of extinct organisms (and therefore a possible linking of fossil traces and its originator), new approaches in development and control of modular robots.

Index Terms
Artificial evolutionary system, computer simulation, Ediacara biota, Dickinsonia costata

ProjectLeaders
Eggenberger Hotz, Peter, Artificial Intelligence Laboratory, Dept. of Informatics, University of Zurich
Wolfgang Schatz, Paleontological Institute and Museum, University of Zurich

In Collaboration with

• Prof. Dr. Rolf Pfeifer, Artificial Intelligence Laboratory, Dept. of Informatics, University of Zurich

• Prof. Dr. R. Wehner, Institute of Zoology, University of Zurich

Student participation (Master-Thesis)

• dipl. zool. Maik Hadorn, Department of Informatics, University of Zurich, Switzerland

Duration
2003-2005 to top

Palaeoecology, fluctuations in conch morphology, palaeobiogeography, extinctions and immigrations of Triassic cephalopods from the Germanic Basin

Abstract
Among the few cephalopod genera from the Muschelkalk, specimens of the ammonoid genera Cerat-ites and Paraceratites outnumber representatives of other genera with respect to diversity, morpho-logical disparity, as well as abundance by far. During the late Anisian and Ladinian, the species of Ceratites and Paraceratites underwent continuous morphological transformations that are reflected in a fairly large number of species and subspecies as well as successive biozones, respectively.
For a stratophenetic analysis of Middle Triassic ammonoids from the German Muschelkalk, data sets of numerous specimens of Ceratites and Paraceratites, were evaluated in scattergraphs as well as canonical discriminant function analyses. Several of the diagrams, which were produced in these analyses, reflect more or less steady changes in conch morphology through geological time, except for some intervals with abrupt and drastic transformations. These morphological discontinuities are synchronous with immigrations of crinoid and brachiopod taxa. This indicates disturbances in the endemic evolution of the ammonoids caused by extinctions and immigrations. At a small scale, this case study demonstrates that a rising sea-level may have boosted the faunal exchange between an open marine and a restricted epicontinental basin, causing a minor regional increase in biodiversity.

Project Leader
Christian Klug, Paläontologisches Institut und Museum, Universität Zürich; Switzerland

Dr. W. Schatz (Paläontologisches Institut und Museum, Universität Zürich; Switzerland)
Dr. D. Korn (Museum für Naturkunde der Humboldt Universität zu Berlin; Germany)
A. G. Reisdorf (Universität Basel, Geologisch-Paläontologisches Institut; Switzerland)

Duration
2003-2005

to top