Taphonomic bias against soft tissue preservation is widely considered to be a barrier to understanding evolutionary relationships and diversification patterns between fossil and recent taxa. A morphologically based phylogenetic analysis using 21 hard part and 23 soft part characters for 16 extant taxa of a monophyletic clade of lacustrine podocopid ostracodes from the Lake Tanganyika system of East Africa was done using PAUP (v. 4.0), yielding 2 trees of 98 steps (CI=0.56). The skewness of tree length distribution reveals significant phylogenetic structure in the data. Nodes are supported by 1 to 11 character state changes, and these character changes are sometimes reversed or paralleled elsewhere, accounting for much of the homoplasy in the reconstructions. Eliminating all hard part characters in subsequent analyses caused the collapse of many branches to polytomies and decreased the agreement of the hard part trees. Additional analyses excluding all soft part characters increased the number of most parsimonious trees, and decreased the resolution of the trees by creating many unresolved polytomies, but produced similar islands of stability as the original analysis.

It is not surprising that this study verifies that the loss of soft and hard part characters reduces the resolution of the analysis of diversity. It is significant, however, that more resolution was lost with omission of the soft parts in comparison to the hard parts, although an approximately equal number of characters were used in each analysis, suggesting that the soft part preservational bias present in the ostracode fossil record may have an appreciable (loss of ~20%) effect on the analysis of their biological diversity. The hard part only tree may be less resolved because those features are more likely to be ecophenotypic and therefore more plastic, which would be consistent with previous studies on ecologically promoted variation in ostracode carapaces and significant with respect to understanding ostracode evolutionary patterns.