The divergence between monocots and dicots represents a major event in higher plant evolution, yet the date of its occurrence remains unknown because of the scarcity of relevant fossils. We have estimated this date by reconstructing phylogenetic trees from chloroplast DNA sequences, using two independent approaches: the rate of synonymous nucleotide substitution was calibrated from the divergence of maize, wheat, and rice, whereas the rate of nonsynonymous substitution was calibrated from the divergence of angiosperms and bryophytes. Both methods lead to an estimate of the monocotdicot divergence at 200 million years (Myr) ago (with an uncertainty of about 40 Myr). This estimate is also supported by analyses of the nuclear genes encoding large and small subunit ribosomal RNAs. These results imply that the angiosperm lineage emerged in Jurassic-Triassic time, which considerably predates its appearance in the fossil record (-120 Myr ago). We estimate the divergence between cycads and angiosperms to be -340 Myr, which can be taken as an upper bound for the age of angiosperms.The fossil record shows a vast increase in the numbers and distribution of angiosperm species in the mid-Cretaceous period, around 100 million years (Myr) ago (1). The earliest reliable angiosperm macrofossils are about 120 Myr old, but because these are already clearly divisible into monocotyledonous and dicotyledonous types it seems that the earliest stages of angiosperm evolution evaded fossilization (2, 3). Although it is generally accepted that angiosperms descended from the progymnosperm lineage, there is little agreement as to when they arose or even from which branch of the gymnosperms they stem (4). Since the progymnosperm lineage extends back to at least 370 Myr ago (2), there is an enormous range of time during which angiosperms might have had their beginnings. Theories as to why there are no fossils of progenitor angiosperms fall into two basic types: either angiosperms did not exist until the early Cretaceous and then radiated explosively, or pre-Cretaceous angiosperms lived in habitats so refractory to fossilization that they left no record (3-5). In this paper we attempt to decide between these alternative theories by analyzing plant DNA sequences, which can be used to estimate the date of divergence of monocots and dicots, and hence to provide a minimal age for angiosperms themselves.Despite promising early results from protein sequencing (for example, see ref. 6), molecular data have not been used extensively to investigate plant evolution. An initial application of DNA sequences to studying the origin of angiosperms has recently been made by Martin et al. (7). Their analysis is based on comparison between plants, animals, and fungi of the sequences of the nuclear gene (called gapC in plants) encoding cytosolic glyceraldehyde-3-phosphate dehydrogenase (GAPDH). By using several divergence dates between animal taxa and between the plant, animal, and fungal kingdoms they were able to calculate the rate of evolution of this...
