Periodic Reporting for period 1 - MindTheGap (Quantifying the completeness of the stratigraphic record and its role in reconstructing the tempo and mode of evolution)
Reporting period: 2023-02-01 to 2025-07-31
Darwin considered the incompleteness of the fossil record as the greatest challenge to the theory of evolution. Earth’s sedimentary archives are so full of gaps that most evolutionary events are lost, and our record of past diversity comprises only rare instances of preservation. Evolutionary biology has struggled to recover conclusive information from the fossil record. The ultimate way of making sense of past and present diversity are evolutionary trees: they allow reconstructing relationships between taxa and contribute to understanding processes that lead to their originations and extinctions. These trees are necessary to identify what environmental conditions increase extinction risk or which traits allow organisms to colonize new environments or adapt to change. However, phylogenetic trees are constructed almost exclusively from genetic and morphological data of living or fossilized organisms. To be included in a tree, a species must be preserved in the first place, then discovered, described, and identified. Similarly, the geological information in what environments and at what time in the Earth’s history these species occurred, which could inform us on the environmental context of past evolutionary events, is disregarded in the construction of phylogenies. Recent studies demonstrate that exclusion of fossil data leads to erroneous estimates of extinction rates and distorts recovered evolutionary relationships. At the same time, we have made substantial progress in understanding the structure of the rock record, and quantitative models predicting the distribution of stratigraphic gaps have become available. This project bridges stratigraphic knowledge and phylogenetic inference to include the breadth of geological expertise in evolutionary reconstructions.
We develop a predictive framework of how gaps in the geological record are distributed and how they affect reconstructions of evolutionary processes at the highest resolution. Long-term trends in past diversity can be predicted from the volume of preserved sedimentary rocks, but until now such predictions addressed only (multi)millenial breaks spanning 1 Myr -10 Myr, corresponding to major tectonic events. Such gaps are easy to detect thanks to their broad geographic range, but too long to reveal mechanisms of evolution. Shorter gaps are much more frequent and more relevant for evolutionary processes, but harder to detect. They represent the same time scale as evolutionary events happening at the population and species level. Resolving them will allow this project to reconcile evolutionary processes observable at human timescales, i.e. gradual shifts in populations, with long-term evolutionary processes accessible to us through the fossil record.
Objectives
1. Identify quantitative constraints on evolutionary reconstructions due to incomplete record
2. Create a toolbox to identify drivers of stratigraphic completeness
3. Reconstruct evolution at the 10 000 - 100 000 years scale from empirical geological data
Impact
Why are evolutionary processes inferred from living organisms always different from those inferred from fossils? MindTheGap will obtain unbiased answers from the fossil record on what mechanisms drive major extinctions and diversifications and how we can use past evolutionary events to predict adaptations. Are the geological and fossil records systematically biased? If yes, how? MindTheGap will focus on carbonates, but the understanding of the structure of the geological record developed in this project will be transferable to other environments. It will reveal how well past conditions are archived in the rock record: what biotic and abiotic environments are we missing from times of low sea level or low sediment production?
We develop a predictive framework of how gaps in the geological record are distributed and how they affect reconstructions of evolutionary processes at the highest resolution. Long-term trends in past diversity can be predicted from the volume of preserved sedimentary rocks, but until now such predictions addressed only (multi)millenial breaks spanning 1 Myr -10 Myr, corresponding to major tectonic events. Such gaps are easy to detect thanks to their broad geographic range, but too long to reveal mechanisms of evolution. Shorter gaps are much more frequent and more relevant for evolutionary processes, but harder to detect. They represent the same time scale as evolutionary events happening at the population and species level. Resolving them will allow this project to reconcile evolutionary processes observable at human timescales, i.e. gradual shifts in populations, with long-term evolutionary processes accessible to us through the fossil record.
Objectives
1. Identify quantitative constraints on evolutionary reconstructions due to incomplete record
2. Create a toolbox to identify drivers of stratigraphic completeness
3. Reconstruct evolution at the 10 000 - 100 000 years scale from empirical geological data
Impact
Why are evolutionary processes inferred from living organisms always different from those inferred from fossils? MindTheGap will obtain unbiased answers from the fossil record on what mechanisms drive major extinctions and diversifications and how we can use past evolutionary events to predict adaptations. Are the geological and fossil records systematically biased? If yes, how? MindTheGap will focus on carbonates, but the understanding of the structure of the geological record developed in this project will be transferable to other environments. It will reveal how well past conditions are archived in the rock record: what biotic and abiotic environments are we missing from times of low sea level or low sediment production?
We developed a novel, Open Source stratigraphic forward model of marine carbonates, CarboKitten.jl (https://mindthegap-erc.github.io/CarboKitten.jl/(opens in new window)) which includes new, bespoke algorithms of sediment transport, insolation-driven carbonate production and subaerial denudation. This model allows all users to create their own stratigraphic records representing different climatic and tectonic conditions. It largely disproves Darwin's concerns that the fossil record cannot be comprehended: the model allows to examine and predict the structure of the geological record, including the positions and duration of gaps, and can be coupled with fossil or paleoclimate proxy data to test hypotheses on what parts of the Earth's history are missing or distorted.
To couple models of the geological records with the fossil and evolutionary records, MindTheGap introduced two research software packages: StratPal and admtools. StratPal (https://mindthegap-erc.github.io/StratPal/(opens in new window)) is the first public implementation of the core methods used in stratigraphic paleobiology, a new field that couples Earth system research with paleobiology. The algorithms proposed in StratPal allow integrating knowledge of stratigraphy into paleobiological datasets, e.g. to test hypotheses on rates of evolution or dynamics of mass extinctions. Strat Pal relies on the second software package developed in MindTheGap, admtools (https://mindthegap-erc.github.io/admtools/(opens in new window)) which introduces a general mathematical representation of the relationship between time and the stratigraphic record and allows testing multiple hypotheses on time contents of any geological record. This is a major addition to previous methodology in stratigraphic and general paleobiology, as generally uncertainty due to the structure of the stratigraphic record had not been accounted for in reconstructing evolution from fossil observations.
MindTheGap refuted the model of diagenetic self-organization, which posed a major challenge to inferring time preservation in carbonate records. We demonstrated that the model does not allow the formation of sedimentary rhythms and that astronomically driven Milankovich cycles are preserved and identified with high certainty even in disturbed geological records.
The Open Source character of all models developed in the project means that they are transparent to the research community, reproducible, and assumptions-explicit.
To couple models of the geological records with the fossil and evolutionary records, MindTheGap introduced two research software packages: StratPal and admtools. StratPal (https://mindthegap-erc.github.io/StratPal/(opens in new window)) is the first public implementation of the core methods used in stratigraphic paleobiology, a new field that couples Earth system research with paleobiology. The algorithms proposed in StratPal allow integrating knowledge of stratigraphy into paleobiological datasets, e.g. to test hypotheses on rates of evolution or dynamics of mass extinctions. Strat Pal relies on the second software package developed in MindTheGap, admtools (https://mindthegap-erc.github.io/admtools/(opens in new window)) which introduces a general mathematical representation of the relationship between time and the stratigraphic record and allows testing multiple hypotheses on time contents of any geological record. This is a major addition to previous methodology in stratigraphic and general paleobiology, as generally uncertainty due to the structure of the stratigraphic record had not been accounted for in reconstructing evolution from fossil observations.
MindTheGap refuted the model of diagenetic self-organization, which posed a major challenge to inferring time preservation in carbonate records. We demonstrated that the model does not allow the formation of sedimentary rhythms and that astronomically driven Milankovich cycles are preserved and identified with high certainty even in disturbed geological records.
The Open Source character of all models developed in the project means that they are transparent to the research community, reproducible, and assumptions-explicit.
MindTheGap yielded unexpected results: like Darwin, we anticipated that the fossil record is heavily distorted and prone to errors. Contrary to our own expectations and those of most researchers in the field, we identified many conditions under which the mode of evolution, paleoclimate, and timing of events such as mass extinctions can be reconstructed with high certainty. We have also discovered that two major assumptions widely assumed to be state of the art, namely sedimentary autocyclicity and diagenetic self-organization, two mathematically challenging phenomena, have been mis-represented in previous models and can be refuted in the form as proposed previously. These findings strengthen our confidence in evolutionary and environmental reconstructions based on the geological record. They also illustrate the acute need for improved standards, scrutiny and training in the field of research model development and implementation, as faulty or irreproducible models can set an entire field back.
The results of MindTheGap pertain primarily to evolutionary biology and stratigraphy and are being disseminated via the International Commission on Stratigraphy, geological surveys and through teaching and public engagement to raise confidence in the theory of evolution.
The results of MindTheGap pertain primarily to evolutionary biology and stratigraphy and are being disseminated via the International Commission on Stratigraphy, geological surveys and through teaching and public engagement to raise confidence in the theory of evolution.