Recently, a really important paper by Alex Dunhill of Bristol University, UK, (@AlexDunhill on Twitter) has been published looking at the problems of using rock outcrop area as a sampling proxy for understanding patterns of biodiversity in deep geological time. What this paper highlights is the important distinction between geological and anthropological sampling proxies, and their relative importance in an on-going international series of projects and collaborations in understanding the patterns and processes that influence biodiversity through geological time.
The reason that this topic is so thoroughly studied are two-fold. Firstly, the co-evolution of the Earth and it’s floral and faunal constituents through time is one of the most important questions of our time, relying on integration of molecular systematists, palaeontologists, geologists, zoologists and botanists (and microbiologists, I guess..). Secondly, understanding the responses of organisms during times of high ecological pressure, high global extinction rates, and strong climatic fluctuations is of obvious importance in the modern world, during this time of rapid global climate change and increasing anthropogenic pressure on the environment. Therefore, studies like this are crucial in aiding our understanding the diversity dynamics of extinct, but still highly relevant, groups.
Dunhill rightly picks up on the annoying substitution that occurs indiscriminately between ‘outcrop’ and ‘exposure’:
“The terms outcrop and exposure are often used interchangeably in the literature, although they have distinct, and important, meanings. Outcrop area is defined as outcropping sedimentary rock that may or may not be buried beneath recent superficial deposits, vegetation, or human land use, and essentially equates to geologic map area, whereas exposure strictly refers to sedimentary bedrock that is visibly exposed at the surface.”
Outcrop is the two dimensional areal extent of a particular rock formation on a geological map, correct. The crucial thing to note though, is that this is inferred directly from the exposures of that particular formation (not necessarily sedimentary), and the structural geometry within. When you look at any geological map, you are looking at the inferred outcrop (see below). Usually, there is a cross-section that accompanies a geological map, representing a two-dimensional transect through a particular region of interest. Using these however, it is extremely difficult to predict rock volume, depending on the complexity of the structural system. In fact, to my knowledge, no consistent methods exist for predicting the volume of geological formations on the local or regional scale. It is possible using seismic methods and boreholes however, as predicting the volume of hydrocarbon-bearing rocks is the foundation of the hydrocarbon industry. This is predominantly utilised offshore, and as most fossils collected are from deposits currently on land, not particularly useful currently.
So, exposure is expected to correlate to inferred volume, while outcrop is the areal manifestation of the intersection of topography with three dimensional cross-sectional geometry that has been reconstructed using exposure patterns. Outcrop has been typically used as a sampling proxy in numerous studies, but what is the justification? Although the three are intrinsically interlinked, the correlations between each other will be on a local scale, dependent on the geometry and thickness variations of beds, and any additional structural variations such as fault systems. Currently, this coupling is poorly understood on both a global and regional scale, although some clarity from previous work by Dunhill is being achieved between exposure and outcrop levels.
Exposure is what you can physically touch, and collect fossils from. It is exposure which is expected to correlate with sampling and collecting effort imposing an anthropological bias. Therefore, rock volume and outcrop area actually have nothing to do with apparent biodiversity, as they do not influence sampling methods or effort. This same logic can be applied to number of formations (or members etc.), as these are simply arbitrary taxonomic classifications. They are all constructions based on what we know from exposed rocks. They may be related on a larger scale to sea-level fluctuations, or extraneous climatic variables, but this is still a point of hot debate, framed in the larger question regarding the co-evolution of the Earth and it’s biota.
The crux of this study is that outcrop area does not correspond to the amount of exposed rock that fossils can be obtained from, irrespective of the method used to calculate area. Accordingly, it is an illogical proxy to use to predict, model, or correct past biodiversity with. Further evidence is added to Dunhill’s previous study (based in England and Wales) from New York state that this may hold true on a global scale, but is contradicted by spatial analysis in California and Australia (the man knows how to pick his study areas!). Dunhill suggests that this may be an artefact of the lack of exposure in these regions. He goes further to suggest novel potential correlates with exposure area, such as elevation and proximity to the coast, as well as acknowledging that as his technique only records modern exposed rocks, it ignores temporal variations such as fossils yielded from now closed quarries. This is considered to be a minor influencing factor.
Dunhill correctly and somewhat disturbingly concludes that we may never be able to attain a global and accurate sampling proxy to help recover true patterns of biodiversity through time. But then, why have recent studies recovered a strong correspondence between outcrop area and palaeobiodiversity? The ‘common cause’ is a hypothesis that suggests both diversity and sedimentary rock area are coupled with an external factor such as sea-level. This hypothesis has found little support so far, in the case of dinosaurs, and requires careful consideration in the future, tied strongly to the work by Dunhill. Future work should focus on local scale interactions of sampling proxies and biodiversity patterns, and use this as a foundation for predicting larger scale patterns and processes. This is critical if we are to understand firstly, the evolutionary dynamics that led to the extant diversity we see today, and secondly how these organisms will respond in the future under climatic, geographic and anthropogenic pressure.
Pingback: One small step for digital Palaeontology | Green tea and Velociraptors