THE OVERLOOKED REALITIES OF SAMPLING BIAS IN THE FOSSIL RECORD

Analyses of past global biodiversity are severely hindered by sampling biases and gaps in the fossil record. These have previously been attributed to several geological and biological factors such as rock outcrop area, taxonomic definitions, fossil abundance, among others. In many cases, inferred so-called global diversity curves only reflect regional diversity of well-sampled localities. However, global compilations of fossil occurrences, such as from the Paleobiology Database (PBDB), not only reflect the physical aspects of the fossil record, but also document the historical compilation of this data over more than two centuries. In this study, we find that socioeconomic factors such as wealth, language and security may help to explain spatial variation in the compiled data of the PBDB. Countries with high per capita gross domestic product (GDP; implying greater access to research funds), a population with high English proficiency, and greater political stability were generally found to be the countries from which higher number of collections originated and documented in the PBDB. Similarly, we found that these countries also contributed to a majority of the compiled literature from which the data was extracted. The factors identified in this study influence fossil data collection, access and availability by determining who has access to specimens and the ability to do research. As such, efforts towards achieving an understanding of the “real” past global diversity will benefit from investing in scientific education, and more research and collaborations in countries at a disadvantage, with a focus on improving local geological and paleontological knowledge.