Mechanistic models of global biodiversity
File(s)
Author(s)
Hintzen, Rogier Eduard
Type
Thesis
Abstract
In a rapidly changing world, society needs ways to understand the distribution of biodiversity and predict its future trajectory. Mechanistic simulation models provide an alternative to traditional statistical correlative methods in explaining the patterns of biodiversity and their underlying processes. I frame the research in mechanistic models and their potential practical applications in conservation using a computational linguistics approach to assess 35,000 full text publications and their citations. I show that conservation and ecology have been steadily drawing apart in topics they discuss. However, there remains cross-dialogue: conservation maintains a strong interest in trophic interactions, community ecology and macroecology in particular. This motivates my focus on global-scale mechanistic models of complete ecological communities. I first study the predictions of an existing general ecosystem model: the Madingley model and test their relationship to empirical data. I find that general trends – the Bergman cline in average body size and allometric scaling of the pace of life with body size – match fairly well. However I also discover that the pace of life, on the whole, is far too fast and small organisms are far too vulnerable to local extinction, leading to suggestions for further model development. I then present a prototype integrated model which nests the Unified Neutral Theory of Biodiversity in subsets of Madingley model organisms, allowing the model to report on a metric of broad ecological and conservation importance: species diversity and associated patterns in range size and local richness across trophic levels and a body size spectrum. Lastly, I develop an eco-evolutionary simulation showing how varying abundance and speciation rate can cause stable latitudinal gradients. I find that the evolutionary process that affects the relative fitness along this gradient can change the origin of lineages, but not the pattern of richness.
Version
Open Access
Date Issued
2019-09
Date Awarded
2019-12
Copyright Statement
Creative Commons Attribution NonCommercial Licence
Advisor
Rosindell, James
Sponsor
Natural Environment Research Council (Great Britain)
Publisher Department
Life Sciences
Publisher Institution
Imperial College London
Qualification Level
Doctoral
Qualification Name
Doctor of Philosophy (PhD)