Spatiotemporal Analytical Modelling for Paleobiology

Introduction

Radical changes in the Earth’s biome as a consequence of climate change will fundamentally affect human society and its relation to the natural world. How can we best model species dynamics to make predictions for the future? What are the main drivers underlying these dynamics, and how are these changing as we enter the Anthropocene?

Historical Context

A great compass for biotic changes we will see in coming decades is the study of changes the Earth has experienced before. This includes:

• Alterations in the distribution of terrestrial and marine mammals
• Dynamic changes in the range and connectivity of forests

There is now a wealth of historic and prehistoric records documenting these processes, including:

• Pollen and fossil records
• Ancient genomes
• Sedimentary DNA

Research Proposal

Here, I propose a research programme called Spatiotemporal Analytical Modelling for Paleobiology (STAMP) to link these disparate types of paleobiological records with the methodological tools of spatiotemporal process analysis.

Empirical Research Areas

I will focus on three empirical research areas:

1. Reconstruction of megafauna species ranges across the late Pleistocene and Holocene
2. Reconstruction of boreal paleo-forest dynamics
3. Study of the historical resilience and mobility of arctic marine mammals

Computational Approaches

These empirical projects will be supported by two computational approaches to inference:

• One based on descriptive spatiotemporal modelling
• Another based on dynamic simulation-based inference

Expected Outcomes

STAMP will produce a new conceptual paradigm for thinking about paleobiological data and create a statistical methodology that will be widely applicable to palaeontologists, geneticists, paleo-botanists, and ecologists worldwide. Ultimately, it will empower the next generation of paleo-scientists with powerful tools to use the living past as a spatial roadmap into the future.