Adam Kocsis

Current climate change is expected to have a definite effect on the global marine biota and will likely lead to not only local, but also global extinctions. Species distributions rearrange
during global warming; marine species track the isotherms of their thermal niches and studies suggest that low-latitude species will be more affected by local extinctions (extirpations). Predictions on the geographic patterns
of complete extinctions are lacking, although past mass extinctions are often invoked as analogues for possible future scenarios. Inference on the causes of mass extinctions is often based on recorded geographic patterns of
species extinctions, but the relationship between warming and these patterns is based on assumptions and thought experiments, rather than spatially explicit models that consider Earth’s geometry, stochastic processes
and multitudes of species.
Project SPex addresses this issue by simulating extinction scenarios, which I organize around the central hypothesis that extensive warming leads to pronounced geographic patterns of extinctions,
preferentially affecting lower latitudes. To assess this and associated hypotheses, I will construct a high-performance modelling framework of species distributions with cellular automata, and simulate spatially explicit biotic
responses to warming with increasing system complexity: in theoretical settings first, and then using data of recorded, warming-related mass extinction scenarios.
With the cellular automaton approach, assigned temperature niches can be used to limit species distributions, while other influencing variables can be modelled as random processes that expand
or contract geographic ranges of thousands of virtual species. Both recent (OBIS, Aquamaps) and fossil (Paleobiology Database) biotic data will be used to constrain the models that will also incorporate continent reconstructions
and general circulation modelling results. Abiotic input data will be used to reconstruct possible scenarios of hyperthermals, such as the Permian/Triassic, Triassic/Jurassic and Pliensbachian/Toarcian events, as well as the
Paleocene-Eocene Thermal Maximum. Patterns of future extinctions will also be assessed using modelled abiotic parameters of simplified scenarios beyond RCP8.5. Simulation patterns will also be contrasted with extirpation and
invasion patterns of gridded fossil data. Thus, the project intends to integrate past mass extinctions and future settings by adding invasions and extirpations to the past and species extinctions to predicted future scenarios.

Outlining and understanding the geographic structuring of biodiversity is a major challenge both for modern and past ecosystems. Different approaches to delineate biogeographic units
are currently applied, which result in vastly different patterns. In order to objectively define biogeographic provinces, I propose a two-year research project to develop quantitative methods that outline biogeographical units
based on marine organismic occurrence data. The so-defined units will allow the assessment of between-unit, beta-level diversity patterns over the Phanerozoic as well. This will enable the scrutinization of hypotheses such as that continent configuration drives global marine beta diversity patterns, and that beta diversity drops in post-extinction recovery ecosystems.
The project will constitute method development and testing on simulated data to rigorously assess their capacity and increase their accuracy as well. The proposed project is divided into four discreet phases, each built on
the results of the previous one. Research will start with the analysis of biogeographic patterns in modern oceans, which will be followed by the analyses of the fossil record in individual time slices. The project will be
concluded with the outlining of quantitatively defined, traceable biogeographic units over the Phanerozoic. Success in the development of the proposed methodology will allow the analysis of the biogeographic structure in marine
settings based on a reproducible partitioning scheme.