Basal angiosperm relationships

Lecture © Marc McPherson
BIOL 606 Session, University of Alberta, March 8, 2000

The study of angiosperm evolution would be enhanced by knowing which seed plant group gave rise to the angiosperms and which extant angiosperm groups are the most basal. Historically and currently the most basal angiosperms have been envisioned as members of the subclass Magnoliidae (Crane et al, 1995). The Magnoliidae are believed to have retained ancestral characteristics such as large bisexual flowers with many free parts as well as some groups with small reduced unisexual flowers (Crane et al, 1995; Takhtajian, 1954). During the mid-Cretaceous the angiosperms spread worldwide and diversified over a short time span. All groups of extinct and extant gymnosperms have been suggested to be the ancestor to the angiosperms (Crane et al, 1995). A comparison of potential outgroups and the use of the fossil record have led to conflicting views about the homology of many seed plant characteristics (Crane et al, 1995). Molecular studies have converged on conflicting rootings for the angiosperms (Parkinson et al, 1999; Qui et al, 1999; Soltis et al, 1999; Graham and Olstead, in press). Hence, the root of the angiosperm tree remains elusive. Phylogenetic trees derived from molecular data show a long branch leading to the angiosperms from the other seed plants, suggesting that the angiosperms are highly divergent from other seed plants (Mathews and Donoghue, 1995). Long-branch attraction has been suggested as an explanation for the conflicting molecular analyses. Long-branch attraction occurs when "multiple hits" or repeated substitutions at homologous base positions of a nucleotide sequence cause highly divergent taxa to be erroneously attracted to one another (Hills et al, 1996). Thus, molecular analyses can be misled by convergent evolution. Hillis et al (1996) suggested that maximum likelihood methods might decrease long-branch attraction problems.

Placing the root for an unrooted phylogenetic tree requires the use of information about events that have occurred prior to speciation of the group under study. Traditionally, outgroups have been used to root trees, but recent studies have shown the utility of paralogous (not homologous) genes in rooting trees (Brown and Doolittle, 1995). Paralogous genes are genes that have duplicated prior to speciation. The phylogeny based on one gene can be rooted using the other gene (Brown and Doolittle, 1995). The tree of life has been rooted using aminoacyl-tRNA synthetase genes, which are believed to have diverged prior to the split of the prokaryotes and the eukaryotes (Brown and Doolittle, 1995).

A similar technique using duplicate phytochrome (PHY) genes has been used to root the angiosperm tree. The genes PHY A and PHY C are believed to have duplicated after the angiosperms split from the other seed plants but prior to the radiation of the living angiosperms (Mathews and Donoghue, 1999). The phytochrome analysis found the root of the angiosperms to be at the branch leading to Amborella (Mathews and Donoghue, 1999). Several molecular studies also suggest that Amborella is the most basal angiosperm (Graham and Olmstead, in press; Qui et al, 1999; Soltis et al, 1999; Parkinson et al, 1999). Amborella is a weedy, woody, vessel-less, angiosperm with unisexual flowers borne on separate plants (dioecious). Tentative hypotheses can be drawn about characteristics of the ancestor to the angiosperms if Amborella retained these characteristics from the common ancestor (Mathews and Donoghue, 1995).


Brown, J. R. & Doolittle, W. F. 1995. Root of the universal tree of life based on ancient aminoacyl-tRNA synthetase gene duplications. Proc. Natl. Acad. Sci. USA, 92: 2441-2445.

Crane, P. R., Friis, E. M. & Pedersen, K. R. 1995. The origin and early diversification of angiosperms. Nature, 374: 27-33.

Graham S. W. & Olmstead R. In press. Utility of 17 chloroplast genes for inferring the phylogeny of the basal angiosperms.

Mathews, S. & Donoghue, M.J. 1999. The root of angiosperm phylogeny inferred from duplicate phytochrome genes. Science, 286: 947-950.

Hillis, D. M., Moritz, C. & Mable, B. K. 1996. Molecular systematics 2nd ed. Sinauer Associates, Sunderland, Mass. USA.

Parkinson, C. L., Adams, K. L. & Palmer, J. D. 1999. Multigene analysis identify the three earliest lineages of extant flowering plants. Current biology, 9 (24): 1-4.

Soltis, P. S., Soltis, D. E. & Chase, M. W. 1999. Angiosperm phylogeny inferred from multiple genes as a tool for comparative biology. Nature, 402: 402-403.

Takhtajian, A. L. 1954. Origins of angiospermous plants. Translated by O. H. Gankin and edited by G. L. Stebbins. The American institute of biological science. Washington, D.C.

Yin-Long Qui, Lee, J., Bernasconi-Quadroni, F., Soltis, D. E., Soltis, P. S., Zanis, M., Zimmer, E. A., Chen, Z., Savolainen, V. & Chase, M. W. 1999. The earliest angiosperms: evidence from mitochondrial, plastid, and nuclear genomes. Nature, 402: 404-407.


Rapporteur: Mark Wilson

Answers to questions indicated the following:

The vessel-less condition is seen in Amborella, some water lilies, etc. The potential angiosperm sister-group Gnetales has vessels. Overall, the conclusion that the last common ancestor of crown angiosperms were vessel-less is weak. For unisexual flowers, the conclusion seems more robust, since early fossil flowers are unisexual also. However, Chloranthaceae have very simple (some would say reduced) flowers, with e.g. a single stamen. They have been in various positions on recent molecular trees.

What are the earliest fossils flowers like? The recently described Lower Cretaceous flower from China has unexpected features. It is difficult to assign early angiosperm pollen to taxa, because the pollen is very simple. The oldest pollen in a flower is probably from the Aptian (mid Early Cretaceous) of eastern USA.

Jonathan Perry led off the main part of the discussion by asking whether the authors of the focal paper had achieved their goals. The consensus was that they had. The main assumptions of the paper, the timing of the duplication and the independent evolution of the different copies, seem to be valid. The two halves of the tree seem fairly similar.

Some possible weaknesses include the lack of a bootstrap value for the Amborella node(s). [This isn't the only node without a bootstrap; maybe the value was less than 50%]. Is the branch connecting the two sides of the tree long or short? One would like to know. One could get sequences and re-analyze to learn more. The omission of Sorgum (because it changes the results drastically) is unfortunate. The conflicting conclusions about what is the second most basal taxon are also unfortunate, as they somewhat weaken the case for Amborella being basal.

Rates of change in sequences clearly differ greatly, so using amount of change to date splitting points would not be wise. Even rates of synonymous mutations vary widely. A different rooting location might give a tree with slightly less glaring inequality in rates. Generation times might vary enough to account for some of the rate variations; most of the rapidly evolving lines seem to be herbaceous, while woody plants seem to have evolved more slowly.

Do studies involving hundreds of species suffer from sub-optimal solutions owing to computational requirements? Yes in general, though in some cases, adding more data (more genes) can help; perhaps the stronger phylogenetic signal with more data allows the algorithm to zero in on optimal structure more easily.

What are the chances of finding new, more basal, living taxa? They are slim. Best places to look include New Caledonia, where Amborella is found (except that red tape and geography make collecting difficult), Australia, New Guinea, and the Falkland Islands.

Has our view of ancestral angiosperms changed greatly thanks to the conclusion that Amborella is basal? Magnoliaceae were beginning to doubted as basal many years ago. The character states of Amborella are relevant, but not conclusive evidence concerning the character states at the basal node of crown angiosperms. That is, just because a vessel-less woody shrub with unisexual flowers from New Caledonia is the basal angiosperm doesn't mean that the first angiosperms were vessel-less woody shrubs with unisexual flowers from New Caledonia.

Could gene conversion cause problems? In some cases, yes, but it doesn't seem to be a problem in this case.