The utility of fossil data in phylogenetic analyses: a likelihood example using Ordovician-Silurian species of the Lophospiridae (Gastropoda: Murchisoniina

Gastropods have a dense fossil record dating back to the Late Cambrian. Intuitively, this would appear to aid phylogenetic reconstructions. However, workers question both whether gastropod shell characters are phylogenetically informative and whether stratigraphic data can be used to test phylogenetic hypotheses. Both questions are addressed with an analysis of 82 species of the Lophospiroidea (= Lophospiridae + Trochonematidae) from the Ordovician and Silurian. Compatibility analyses of 95 shell characters shows that characters are far more compatible than one would expect from homoplasy-saturated data. However, compatibility among the 73 characters that vary among the earliest lophospirids decreases over time, which suggests that later species introduced homoplasy. Simulations using 95 characters and sampling similar to that observed for lophospirids show that parsimony performs poorly relative to methods that incorporate stratigraphie data such as stratocladistics. An alternative approach is used here. The first step is to estimate the likelihood of a hypothesized tree given observed character congruence (i. e. parsimony length) using simulations. The second step uses two different statistical tests to estimate the likelihood of hypothesized trees given observed stratigraphic data. Likelihoods then are combined to evaluate trees. The resulting likelihood tree is nearly 30 steps longer (378.4 versus 350.5), but is considered more likely (given a 350.5-step matrix) than a tree of 350.5 steps. Both trees suggest that budding cladogenesis (where ancestors co-exist with descendants) was the most common pattern of speciation, although the likelihood tree is more emphatic on this point. Both suggest a trend towards increasing numbers of ornate species; however, whereas parsimony suggests the differential diversification of an ornate clade, likelihood suggests a strong tendency for inornate ancestors to have ornate descendants. A genus-level taxonomic revision is provided that is consistent with both trees and that attempts to reflect historical diversity patterns.