Justin D. Yeakel bio photo

Justin D. Yeakel

Assistant professor at UC-Merced; Omidyar Postdoctoral Fellow at the Santa Fe Institute

Email Google+ Github Twitter

Recently, our paper on reconstructing and analyzing the structure and function of ecological networks in Ancient Egypt (published in PNAS this past September), recieved two letters to the editor. One letter was written by Dr. Linda Evans, an Egyptologist studying the roles of animals in ancient Egypt, and one was by a research team at the University of Haifa, Israel.


The second letter, by Bar-Oz et al., was not a criticism, mentioning that they have found similar patterns of extinction in paleontological sites in the Levant. One of the primary differences in the Levant is that the large aridification shifts documented in Egypt were not nearly as pronounced, suggesting that alternative mechanisms may have been responsible for animal extinctions. They note that the most likely mechanisms are human-driven. In our analysis of the Egyptian fauna, we can’t distinguish between different potential drivers of change, though the largest changes to the mammalian community co-occur with large aridification events. Certainly, impacts to the mammalian community during these events may also have been driven by humans, perhaps exaggerated by changes in climate.


The letter by Dr. Evans was more critical, and brought up some important points. The primary concern that Dr. Evans raised was that changes to the mammalian community - as depicted in artwork - could have been the result of cultural forces, and may not accurately represent which species were around, and when. This is certainly a valid criticism. One of my favorite examples of this is the California state flag, which depicts a grizzly bear, even though there are no longer grizzly bears in California (sadly). Because we were using artwork to reconstruct animal occurrance, this was a potential issue that we had expected going into the project. To deal with potential cultural bias – or even errors due to preservational bias – we injected a large amount of uncertainty with regard to the timing of species extinctions into our analysis. Importantly, this added uncertainty did not have a significant influence on our results.


Read the letter by Evans here, the letter by Bar-Oz here, and our response to both here.