ANGIOSPERM PLANT DNA C-VALUES DATABASE AND ITS USE

Jodrell Laboratory, Royal Botanic Gardens, Kew, Richmond, Surrey TW9 3DS UK

Bennett Michael D, Cox Antony V, Leitch Ilia J Jodrell Laboratory, Royal Botanic Gardens, Kew, Richmond, Surrey TW9 3DS, UK. Nuclear DNA C-values are used in many fields including: systematics, ecology, cell and molecular biology and plant breeding, so C-value data must be easily available for reference and analysis in a user-friendly form. Information on DNA amounts in different angiosperm taxa is often hard to locate, as published data are widely scattered in diverse journals, while a significant amount is unpublished and unavailable. From 1976 to 1997, Bennett and co-authors published five collected lists of DNA C-values in angiosperms for reference purposes. As the number of such lists rose, finding whether an estimate for a particular species was listed, took longer. To overcome this, in 1995 it was decided to pool all the data from the five separate lists into one combined list, and to make this available in both hard copy and electronic forms. A new pooled list including 4000 C-value estimates for 2802 species from over 300 original sources was produced. The `Angiosperm DNA C-values database', a first limited version of the pooled list (http://www.rbgkew.org.uk/cval/database1.html ) was published electronically in April 1997. This has been well used, with over 4600 successful enquiries logged in its first year. A second, relational, version went live in October 1998. Further improvements are in hand and new versions are planned, including the addition of C-values for gymnosperms. Existence of the relational database has already facilitated broad new analyses for angiosperms, including studies of C-values in the world's worst weeds, and of phylogenetic trends in genome size. It has also allowed major gaps to be identified, giving information needed to target new research. No C-value is known for >66% of families, including many tropical families with over 800 species. There is a major need to estimate angiosperm DNA C-values, and the choice of new materials must now focus strongly on ensuring improved representation.