Many of the fossils left by the strange “Ediacara biota” that thrived during the Ediacaran Period (about 635 to 540 million years ago) look like ribbed pancakes, armored worms, and fat fern fronds. Despite finding these fossil deposits worldwide, paleontologists have struggled to place the unfamiliar creatures on the tree of life or even to agree on where and when they lived. Now, researchers are turning to data science tools to make sense of Ediacaran fossil data accumulated from decades of paleontological research.
For the first time, researchers used a technique called network analysis, which reveals connections in complex systems, to analyze Ediacaran fossil collections and to match them to specific environments and times. The analysis provides evidence that two extinction events potentially linked to disturbances in the global carbon cycle occurred during this period. The extinctions created an opening for the subsequent Cambrian explosion, when most major groups of animals appeared for the first time in Earth’s history. Drew Muscente (formerly of Harvard University, now at University of Texas, Austin, USA), and Andrew Knoll (Harvard University, USA), with DCO researchers Ahmed Eleish, Anirudh Prabhu, Feifei Pan, Peter Fox (all at Rensselaer Polytechnic Institute, USA), and DCO Executive Director Robert Hazen (Carnegie Institution for Science, USA), report these findings in a new paper in Nature Communications.
“The Ediacaran Period was the time in Earth’s history when life got both large and complex,” said Muscente. Though animals with hard skeletons began to appear during this period, many of the organisms had soft bodies that left only imprints on the ocean floor or tracks in the mud to record their movements. “These organisms left behind a great fossil record that has been the subject of intense debate over the last century. Many of the fossils are quite problematic and we’re not entirely sure what they represent.”
Previous studies of Ediacaran fossils have placed them into four main groups that tend to be found together, called the Avalon, White Sea, Nama, and Miaohe clusters. But scientists weren’t sure whether the clusters represent different ages, types of marine habitats, or means of fossilization.
To make sense of the wealth of fossil data in existing databases and the scientific literature, researchers applied network analysis, which uses algorithms to identify, quantify, and visualize underlying relationships in the data. Scientists have used similar techniques for various applications in the social, life, and Earth sciences. For example, researchers have used data from social media and phone records to map terrorist networks and visualize disease outbreaks. By finding connections among the location, age, and formation conditions of the different fossil types, the researchers could see underlying patterns that weren’t visible when the data sat in tables and two-dimensional graphs.
The analysis showed that two of the clusters, Miaohe and Avalon, represent communities restricted to specific marine environments and pathways of fossilization. The Nama and White Sea clusters occupied habitats that were similar to each other, and their fossils were preserved in the same way. However, the authors show that those clusters are indicators of distinct geological layers, meaning that those communities lived at different times in the Ediacaran. White Sea organisms lived during the Ediacaran biota’s heyday, prior to an extinction event around 550 million years ago. The Nama cluster, in contrast, represents a community that thrived until the next extinction event some ten million years later – around 540 million years ago.
“We sometimes talk about communities like they’re organisms that can change and evolve in time,” said Muscente. “Extinctions have always been an important factor in setting the pace of change in communities of marine life.”
The network analysis also allowed the researchers to quantify the extent of the extinctions and to use statistics to show that their conclusions are based on enough data that they likely represent real phenomena and not just gaps in the fossil record. “While people have been able to quantify species loss, it’s been much more challenging to understand what the ecological dimensions of these extinction events were and network analysis enabled us to visit that in a quantitative way,” said Knoll. “Thanks to computation and the internet, we’ve arrived at this point where we can literally take the fruits of 100 years of geological and paleontological labor and look at them in new ways to see general patterns.”
The Ediacaran was a tumultuous time in Earth’s history, which started around 635 million years ago with the end of a “Snowball Earth-type” glaciation, when glaciers encased most of the planet. Then, between 580 and 550 million years ago, the Earth System experienced a mysterious global shift in carbon isotopes called the “Shuram excursion,” perhaps the largest disturbance to the global carbon cycle in the history of the planet. The researchers suspect that the extinctions may have been caused by environmental disturbances affecting the carbon cycle and competition between the Ediacara biota and early animal life. Whatever the cause, the second extinction wiped out the Ediacara biota, allowing animals to succeed evolutionarily and diversify during the Cambrian period.
Next, Muscente plans to use network analysis to look at marine fossil communities that lived during other major events in Earth’s history, such as the Paleocene–Eocene Thermal Maximum (PETM), a global warming event that occurred around 56 million years ago. The PETM and other ancient warming events may provide insights into how marine communities will respond to climate change in the future.
The current study grew out of a project supported by the Keck Foundation to develop data resources and tool kits to understand the coevolution of the geosphere and the biosphere in deep time. The project also received support from the DCO through the Data Science team, which had previously used network analysis to investigate the evolution of minerals with Hazen’s group. Through the project, several early career scientists received training in data science techniques so that they could take a fresh look at their respective fields.
“It’s nice to change the face of a scientific field by using existing data in really quite new and innovative ways,” said Fox, who is the DCO Data Science Team Lead, “There are tools and methods out there to take on these bigger-scale problems.”
Main image: A fossil of Charnia, a genus of Ediacaran organisms with branching, segmented ridges that look like fern fronds. Credit: Verisimilus at English Wikipedia