Novel Technique Offers New Look at Ancient Diets

Scale-sensitive fractal analysis. a-e, Relative area is calculated by dividing the area of a surface, calculated using triangles of a given scale in a virtual tiling algorithm (a, b, c), by the projected area of the surface (d). Relative area can then be plotted against scale in a log-log plot (e). Asfc30 is a scale-sensitive measure of roughness and is the slope of the steepest part of the curve fitted to the plot of relative area over scale, multiplied by -1,000.

A new technique for studying the microscopic marks caused by wear on teeth has generated insights into dietary divergences between some of our human ancestors, allowing scientists to better understand the evolutionary path that led to modern-day human diets. A paper describing the technique, written by Robert Scott and Peter Ungar at the University of Arkansas and colleagues including Alan Walker, Evan Pugh Professor of Anthropology and Biology at Penn State, will be published in the 4 August 2005 issue of the journal Nature.

Walker pioneered the field of dental microwear in 1978, when he described in the journal Science his use of a scanning electron microscope, the best tool available at that time. With this tool, Walker and his research team were able to distinguish broad categories of diet in mammals from many epochs in Earth's history. The pits and scratches found on the teeth offer a visual history of the type of food consumed by the tooth's owner. Pits indicate a diet of hard, brittle foods, like nuts and seeds, while scratches imply a diet of tough foods, like leaves and possibly meat.

But limitations of the scanning electron microscope, including its two-dimensional images and its reliance on human interpretation of these images, limited the advances that could be made in the field of dental microwear. "Now with the development of scanning confocal microscopes," Walker says, "we can get image in 3 dimensions automatically and we can process the data automatically with no human intervention--a major step forward. The confocal microscope and engineering software allow the researchers to take three-dimensional coordinates of the entire tooth and form a detailed image of the surface. When these images are combined, they can use fractal analysis to examine patterns in the tooth wear.

Microwear texture analyses. a-d, Meshed axonometrics of digital elevation models (left), bivariate plots of relative area versus scale (middle), and rosette plots of normalized relative length at 1.8 µm by orientation (right) for scans on representative specimins of Alouatta palliata (NMNH 543117) (a), Cebus apella (NMNH 518433) (b), Australopithecus africanus (Sts 61) (c), and Paranthropus robustus (SK 16) (d). Steeper best-fit lines for the steepest order of magnitude on the relative area-scale curves evince higher Asfc values, indicating greater complexity (for example, b and d). More clumped rosettes (for example, a and c) have higher values of epLsar, indicating greater anisotropy of the wear fabric.

"Our new techniques are allowing us to get beyond simple dichotomies and helping us understand the processes by which dietary evolution is working," said Peter Ungar, professor of anthropology at the University of Arkansas. The research team includes Ungar; Walker; Robert Scott, a postdoctoral fellow at the University of Arkansas; and other colleagues at the Worcester Polytechnic Institute, State University of New York at Stony Brook, and the Johns Hopkins University School of Medicine. The researchers examined microscopic wear on the teeth of two species of ancient hominids - Australopithecus africanus, which lived between 3.3 and 2.3 million years ago, and Paranthropus robustus, which lived between 2 and 1.5 million years ago.

Traditional examinations of these ancient teeth--counting pits and lines on a black and white electron micrograph image--suggested that A. africanus ate tough foods and P. robustus dined on hard, brittle fare. However, the researchers used a new technique developed by Ungar and his colleagues that combines engineering software, scale-sensitive fractal analysis, and a scanning confocal microscope to create a reproducible texture analysis for teeth that tells a more complete story.

The researchers looked at both roughness, or complexity, and directionality in the teeth they examined. "Since food objects interact with teeth, we have different kinds of complexity in different diets. Directionality also correlates with diet," Scott said. Hard foods like nuts and seeds tend to lead to more complex tooth profiles, while tough foods like leaves lead to more scratches, which corresponds with directionality.

Anisotropy and complexity. a, b, Bivariate plots of epLsar1.8 versus Asfc for Alouatta palliata and Cebus apella (a), and Australopithecus africansus and Paranthropus robustus (b). The values plotted are means based on the four scans witth adjoining edges from each specimen. Credit: Nature

The analysis showed that the two species had significant amounts of overlap in their diets and that while P. robustus had more complexity in its tooth wear, indicating that it ate more hard and brittle foods than A. africanus, it ate tough foods as well. The researchers believe that this discovery indicates that the species frequently ate the same types of foods, but that in times of scarcity or seasonal changes, P. robustus changed its diet to include foods that differed from those of A. africanus. "The difference in their evolution in terms of diet is not driven by their preferences, but by scarcity," Ungar said. "It gives you a whole new way of thinking about dietary adaptation."

The researchers credit the new method of examining microscopic wear on teeth with allowing them to gain new insights into dietary evolution. "The old technique was subject to observer error, so we couldn't get a handle on whether the variation we observed was real or the result of observer error in data acquisition," Ungar said. "The new technique is free of subjective observer error, so the variation we see is real. This allows us to actually look at within-species variation. We can finally get beyond 'these differed' and start to understand how much they differed and overlapped, and what this means in terms of their adaptations and evolution."

"This technique does the same thing as finding new fossils," Scott said. The researchers examined the teeth, approached the pits and scratches with a new technique, and drew new conclusions from the data. "We can say things that we could never say before," he said. The project was funded by two grants from the National Science Foundation.