IMPACTS OF TAPHONOMIC LOSS AT LOWER TAXONOMIC LEVELS: STEM-WARD SLIPPAGE IN A BIVALVE FAMILY

Recent studies have demonstrated that removing taphonomically susceptible characters during phylogenetic inference systemically leads to taxa appearing erroneously more ‘primitive’, a phenomenon known as ‘stem-ward slippage’. The impact of ‘stem-ward slippage’ has mostly been demonstrated at higher taxonomic levels due to the removal of key soft tissue characters (Sansom [2010] Nature 463: 797-800). However, it is less clear whether stem-ward slippage occurs at lower taxonomic levels, or where hard tissue characters dominate character matrices. Here, we examine the effect of morphological character loss by taphonomic processes on phylogenetic inference in extant members from the bivalve family Carditidae. Based on literature review and examination of museum specimens, we derived four sequential taphonomic stages of shell character loss for Carditids, with the first taphonomic stage containing less robust characters that are quickly lost and the fourth taphonomic stage containing more robust characters that persist for longer. We then sequentially simulated taphonomic character loss on a previously published character matrix (Perez [2019] J. Syst. Palaeo 17: 1359-1395) and used strict consensus trees to examine the impact of taphonomic degradation on topology inference. Removal of taphonomically susceptible characters resulted in an overall decrease in phylogenetic resolution, measured as the proportion of possible nodes recovered in the strict consensus tree, for one third of extant taxa (N = 23). In 26.1% of cases, this also resulted in stem-ward slippage of the degraded taxon. Across all four taphonomic stages, the third taphonomic stage resulted in the greatest increase in the number of taxa shifting positions, with 20.8% of extant species shifting stem-ward after the third taphonomic stage removal. A qualitative analysis showed specific clades demonstrate more signs of stem-ward slippage when characters were removed in taphonomic order than when removed randomly. Our results suggest taphonomic processes can interfere with phylogenetic analyses and historically may have been severely underestimated, warranting further consideration.