Abstract
Background The fossil record provides the unique opportunity to observe evolution over millions of years, but is known to be incomplete. While incompleteness varies spatially and is hard to estimate for empirical sections, computer simulations of geological processes can be used to examine the effects of the incompleteness in silico.
We combine simulations of different modes of evolution (stasis, (un)biased random walks) with deposition of carbonate platforms strata to examine how well the mode of evolution can be recovered from fossil time series, and how test results vary between different positions in the carbonate platform and multiple stratigraphic architectures generated by different sea level curves.
Results Stratigraphic architecture and position along an onshore-offshore gradient has only a small influence on the mode of evolution recovered by statistical tests. Tests fail to identify the correct mode of evolution in the absence of stratigraphic effects, and support for the correct mode decreases with time series length.
Visual examination of trait evolution in lineages shows that rather than stratigraphic incompleteness, maximum hiatus duration determines how much fossil time series differ from the original evolutionary process. Directional evolution is more susceptible to stratigraphic effects, turning it into apparent punctuated equilibrium. In contrast, stasis remains unaffected.
Conclusions
Tests for the mode of evolution should be reviewed critically, as they do not find good support for the correct (simulated) mode of evolution, even for adequate models that generated the data, in the absence of stratigraphic effects, and for exceptionally long time series.
Fossil time series favor the recognition of both stasis and complex, punctuated modes of evolution.
Not stratigraphic incompleteness, but the presence of rare, prolonged gaps has the largest effect on trait evolution. This suggests that incomplete sections with regular hiatus frequency and durations can potentially preserve evolutionary history without major biases. Understanding external controls on stratigraphic architectures such as sea level fluctuations is crucial for distinguishing between stratigraphic effects and genuine evolutionary process.
Competing Interest Statement
The authors have declared no competing interest.
Footnotes
Clarified usage and interpretation of AICc, expanded methods for simulations of trait evolution, emphasized role of different sea level curves, expanded discussion on test results.