Paper short abstract:

A number of studies have proposed causal links between African climate change and events within hominin evolution. Here I discuss a number of biases inherent to both palaeoclimate archives and the hominin fossil record which may limit our ability to test these hypotheses.

Paper long abstract:

In recent decades, a number of studies have compared Miocene to Pleistocene palaeoclimate records with various aspects of the hominin fossil record, including speciation, extinction and species diversity. When palaeoclimate and evolutionary events are temporally coincident, a causal relationship is often assumed, and hypotheses are put forward to explain the possible mechanisms for these relationships. However, different palaeoclimate archives can show distinct climate trends, resulting in a patchy understanding of African climate history. I will demonstrate this using examples of inter-annual to supra-orbital rainfall and vegetation proxies from southern and eastern Africa.

The hominin fossil record is notoriously incomplete, and this limits the degree to which events within hominin evolution can be meaningfully compared with palaeoclimate proxies. Using a compilation of hominin fossils and stratigraphic horizons, I will present confidence intervals for the stratigraphic ranges of each early hominin species. This analysis shows that the First and Last Appearance Datum (FAD and LAD) for each species has a typical uncertainty of about a million years. I will also show that hominin species diversity is governed by the number of fossil-bearing localities, and needs to be corrected for sampling bias prior to comparison with climate proxies. Although the hominin fossil record is gradually improving, it will take decades of new discoveries to see a significant improvement in the quality of the dataset. In the meantime, proxies for the hominin fossil record, such as bovids or stone tool industries, may provide more reliable and fruitful comparisons with temporally constrained climatic events.