The effect of singletons and interval length on interpreting diversity trends from the palaeobotanical record

The reliability and the identification of potential biases are central aspects to providing an accurate robust palaeontological data analysis. In the last decade, older concepts of evolutionary processes and patterns have been revised in the light of new methods and hypotheses. Moreover, new mathematical algorithms have been developed to better interpret evolution and to reduce several biases inherent in the fossil record from the animal viewpoint. This new global study, based on a complete overview of the plant fossil record, uses comparisons with previously established palaeofloristic patterns and employs a set of regression, accumulative and evolutionary analyses to test the influence of two important variables in the inference methods: the role of singleton taxa and the duration of the time units. The regression analyses reveal that the duration of the time units employed does not distort our perception of the number of singletons at the family level. Likewise, the duration as single factor also does not affect the measures of taxonomic diversity, and does not influence representation of the main evolutionary patterns of vascular plants. The analysis reveals that the percapita origination and extinction rates used provide global diversity patterns that diminish the effects of the possible taxonomic problems and preservational biases relating to the controversial nature of singleton taxa due to its restricted record, and they reveal that the main trends and plant turnover is characterized by slight abrupt changes, thus providing a more realistic vision about the dimension and magnitude of the observed evolutionary processes and diversification patterns from the plant fossil record. © Palaeontological Association January 2012.