Long-Range Correlations in the Fossil Record and the Fractal Nature of Macroevolution

Abstract

Recent studies on the fossil record time series has suggested that there is consistent evidence for self-similarity i.e. long-range correlations with power-law behavior. The existence of such fractal strucutes means that, when looking at a given time frame, some basic properties remain the same if a change of scale is performed. In other words, there is no characteristic time scale, as we could expect if some type of periodic or other low-dimensional dynamics were present. A possible explanation for such long-range order is a dynamical process operating at all scales, as it is the case for systems in the neighborhood of critical points. In this paper these results are further explored by extending previous data analysis and examining the relevance of recent theoretical approaches to the statistical features of the fossil record. The presence of long-range correlations is shown through Hurst analysis using non-interpolated data series from the Fossil Record 2 database. As shown in previous studies, such correlations span over hundreds of millions of years and are compared with a simple model of large-scale evolution displaying self-organized criticality.