EOCENE ATHLETA (VOLUTOSPINA) (GASTROPODA:VOLUTIDAE) FROM THE U.S. GULF COAST AND EUROPE: THE CLASSIC CASE OF DIRECTIONAL EVOLUTION IN ATHLETA (VOLUTOSPINA) PETROSUS CONFIRMED BY MAXIMUM LIKELIHOOD MODELS

Gastropods of the genus Athleta (Volutospina) Conrad, 1853 are common and well preserved in Eocene deposits of the U.S. Gulf Coastal Plain and the UK-Paris basin. In 1964, Fisher et al. documented directional evolution within a clade they referred to as the “Athleta petrosa stock”. Four species (A. (V.) petrosus, lisbonensis, tuomeyi, and dalli) were depicted in their phylogenetic tree. I re-evaluated their phylogeny using current quantitative methods of tree construction and added additional co-eval Athleta (Volutospina) species of the Paris Basin and southern U.K. Unsurprisingly, the new tree is markedly different from the original of 1964.

In addition to their phylogeny, Fisher et al. wrote that A. (V.) petrosus seems to exhibit directional trends in multiple traits yet their study lacked robust multivariate statistical techniques and a null model. Whether the trends exhibit random variation, stasis or directional change has never been explicitly tested. Here, I investigate whether trait changes in five Athleta (V.) species are best characterized by Hunt’s models (2006) of directional change, random variation or stasis.

The dataset I assembled contains species from the U.S. Gulf Coast and coeval congenerics from the UK and France, each with a stratigraphic range resolved to the member or bed level, and a character matrix of 30 discrete and 15 continuous characters. I used the package ‘paleoTS’ in R to fit maximum likelihood models corresponding to stasis, random variation (the unbiased random walk), and directional change (the biased random walk) to shape data. The relative fit of models was assessed by comparing their Aikaike’s information criterion (AIC) scores. Remarkably, preliminary results indicate that both stasis and random non-directional change (the unbiased random walk) can be rejected in favor of directional change (the biased random walk) for the A. (V.) petrosus lineage. The remaining four species exhibited good fit of the data to stasis and the unbiased random walk models. Implications for macroevolution and punctuated equilibrium are explored.