IS PHYLETIC GRADUALISM A USEFUL CONCEPT? INSIGHTS FROM STRATIGRAPHIC PALEOBIOLOGY AND MORPHOLOGICAL TIME SERIES

Inferences about the tempo and mode of evolution are two distinctive and important contributions of paleobiology to our understanding of macroevolutionary dynamics. Most discussions of this contribution, however, are expressed in terms of microevolutionary mechanisms and population-scale dynamics. In particular, the pattern of tempo known as "phyletic gradualism" appears to be a chimera based on hypothetical extrapolation of directional selection within populations upward to macro-evolutionary time scales. The pervasive inverse correlation between rate of evolutionary change and interval length suggests that unsteady or reversing change is widespread. Furthermore, to qualify as dynamically meaningful, morphological (stratophenetic) time series must be significantly different from those produced by a random walk-like process. Nearly all "gradual" series fail this test, while stasis-like series routinely pass. This asymmetry is partly an artifact of the mathematics of noisy data sets, but is also a reflection of the nature of paleontological time series: they are geological data sets not purely biological data sets, and include not only evolutionary dynamical information but also sedimentologic, taphonomic, and collecting-induced signals. Thus, we frequently arrive at the unsatisfactory conclusion that manifestly adaptive macroevolutionary histories (viewed from the perspective of a functional anatomist) are indistinguishable from random walks. Accordingly, we conclude that paleontological time series actually comprise only two classes: stabilized dynamics (equilibrium and punctuation) and non-stabilized dynamics (pseudo-random walks sensu Raup and, rarely, more strongly directional changes, etc.). Paleobiology may benefit by an alteration of our thinking about the questions that we can ask of the fossil record in light of this taxonomy. Morphological time series (like stratigraphic sections) reflect a world in which chaotic dynamics operate in hierarchically nested ways. By incorporating more sophisticated insights about the stratigraphic and taphonomic components of morphological time series we may arrive at a clearer understand how this complexity connects across processes and scales to generate the stabilized and non-stabilized long-term outcomes that we observe.