Format: ORAL

Richard M. Bateman

Royal Botanic Gardens Kew, Richmond, TW9 3DS, UK (r.bateman@kew.org)

Traditional morphological taxonomy utilises multiple preserved collections that collectively constitute a serendipitous sample of organisms that have been immediately killed and often also dismembered, thereby reducing the number of morphological characters available for subsequent study. The comparative approach typically applied to herbarium specimens prioritises qualitative description above quantitative analysis and cerebral rather than algorithmic assignments of specimens to taxa; it also over-emphasises the holotype – a single point in space and time. The contrasting 'demographic' approach to taxonomy, illustrated here using studies of European orchids, employs a statistically rigorous in situ sampling strategy, and yields large volumes of fully quantified data of two or more kinds obtained from the same spectrum of individual organisms. Each sampled individual is subjected to character-rich morphometric data collection, describing every organ, and to equally detailed population genetic analysis; additional data categories are optional. Each category of data then undergoes multivariate ordination via an algorithm that does not require prior taxonomic assignment of individuals. This approach entails two contrasting forms of reciprocal illumination – the essential element in any truly scientific enterprise. (1) Organisms are aggregated into their source populations, and populations into infraspecific taxa and species, relatively large distances among sets of populations providing the basis of a genuine species circumscription process. (2) Confidence in, and the underlying nature of, the putative species boundaries are then determined through congruence among patterns of similarity in contrasting categories of data, addressed in the context of information gathered on extrinsic properties such as symbiotic relationships, habitat preference and putative geographic distribution. Although the demographic approach consumes considerably more time and resources than traditional taxonomy, it rigorously tests initial species hypotheses and provides far deeper understanding of the resulting species – perhaps even sufficient to predict their behaviour in the face of environmental challenges such as climate change.