Amborella trichopoda helps reveal features of the first flowering plants

The flowering plants, or angiosperms, are the largest group of plants, containing over 300 000 living species. All these species descend from a single, most recent common ancestor (MRCA), which probably lived some 225 to 140 million years ago. Its exact date is still rather uncertain! Unfortunately, we don’t have a fossil of the MRCA of the living angiosperms, and even if we did, without a series of fossils linking it to living groups, we might not recognize it as such. However, we can also use living species to reconstruct the characters of the first flowering plants. For this, we first need to make an accurate evolutionary tree from the DNA sequences of living angiosperms. We then map their character states onto this tree to deduce the most likely character states in the angiosperm MRCA. Characters that can be reconstructed in this way include the morphology, anatomy, physiology and even ecology of the first angiosperms. The angiosperm family tree (Fig. 1) turns out to be very asymmetric: its first few bifurcations lead, on one side, to groups containing very few species. Indeed, the likely most basal branch of the angiosperm tree leads to the single living species Amborella trichopoda. This shrub, endemic to the forest understory of the tropical island of New Caledonia, is thus the probable sister of all ~300 000 other living angiosperms. Together with the representatives of the next two early-diverging angiosperm lineages, Nymphaeales (water lilies and their close relatives) and Austrobaileyales, Amborella has a special importance for the reconstruction of character states in early angiosperms.

A scheme of the angiosperm phylogenetic tree, showing the basal position of Amborella. — Fig 1. A scheme of the angiosperm phylogenetic tree, showing the basal position of *Amborella*. All genera (names in italics) in the ANA grade (ANA for Amborellales, Nymphaeales and Autrobaileyales) are shown, as are the families (names ending in “aceae”) and orders (names ending in “ales”) to which they belong. Approximate numbers of species are shown in brackets.

Two papers featured on the cover of the March issue of AoB (Fig. 2) use data from Amborella and other early-diverging angiosperms to try to reconstruct features of the angiosperm MRCA. In the first of these papers, by Fogliani et al., the conclusion seems clear. The first angiosperms appear to have possessed seeds in which a dormancy mechanism that relied on both morphological and physiological components was present. This mechanism functioned to delay the germination of seeds and thereby optimize their chances of survival. This type of dormancy is still present in Amborella and in the genus Trithuria from the second earliest diverging angiosperm order Nymphaeales.

Amborella flowers and drupes (fleshy indehiscent stone-fruits), — Fig. 2. *Amborella* flowers and drupes (fleshy indehiscent stone-fruits), as described by Angeret al. and Fogliani *et al.,* from the cover of the March issue of AoB. The flowers illustrated are unusual for *Amborella* as they show a bisexual phenotype that arises in some otherwise male individuals in the population.

In the second paper, Anger et al. attempt to deduce the type of breeding system in the first angiosperms. Did the flowers of early angiosperms contain both male and female reproductive organs, as is the case for the majority of angiosperms living today, or were these plants dioecious, a state in which only male or only female flowers are produced on each individual? Interestingly,Amborella is rather unusual among basal angiosperms in being dioecious. This character state results in an ambiguous situation in which it is impossible at present to conclude whether the MRCA of living angiosperms was dioecious or produced flowers of both sexes. However, Anger et al. suggest a method through which further work may shed light on this question. These authors show that a seed-grown population of Amborella contains a 1:1 ratio of males to females, and that the sex of individuals is stable between flowering seasons. They conclude from these observations that dioecy in Amborella must be determined by a pair of segregating sex chromosomes. Anger et al. go on to suggest bioinformatics-based methods that could be used to date the origin of chromosomal sex-determination in Amborella. This date-of-origin should help to refine future reconstructions of the breeding system of the MRCA of living angiosperms.

At present, Amborella is the only early-diverging angiosperm whose genome has been entirely sequenced. However, improvements in DNA sequencing technology make it likely that several genomes from other early-diverging angiosperms will soon become available. In addition, we can expect that complete genomes of a greater number of gymnosperms (conifers and their allies) will soon be sequenced. Gymnosperms are the closest living relatives to the angiosperms, and form a vital external reference point for studies of the origin of the angiosperms. All these extra molecular data should help identify the characteristics of early flowering plants. They may also shed light on the more difficult question of how the first flowering plants evolved from a currently unknown, but presumably gymnosperm-like ancestor.

Charlie Scutt studied plant sciences at Reading and Durham universities before doing post-doctoral research in Leeds and Lyon. He is currently a Director of Research of the French CNRS, based in the Ecole Normale Supérieure de Lyon. He works mostly on the origin and early evolution of the flowering plants.