ASBS Newsletter – Book Review

The Evolution and Classification of Flowering Plants
Second edition

written by Arthur Cronquist

(From ASBS Newsletter Number 63, June 1990)

Publisher: The New York Botanical Garden, New York. 1988. 555 pp. ISBN 0-89327-332-5.

What did the first angiosperm look like? Where did it originate? And when? And what were its immediate descendants, the progenitors of the highest natural groups within the angiosperms? These are some of the "big" questions in plant systematics, yet their answers are still largely shrouded in mystery. Enormous energy has gone into the attempt to answer them, and those who have tried to do so have built reputations on their efforts.

Arthur Cronquist is one such. His system of angiosperm classification is influential, especially in this country where it has been adopted for the Flora of Australia. So it behoves Australians especially to examine his rationale.

Taxonomic Metaphors

Never doubt the power of visual metaphors in science (Gould 1990). Four graphic metaphors have encapsulated the history of taxonomic thought by illustrating successive models of relationships (Stevens 1984). In the beginning, Aristotle gave us the ladder (scala naturae), a great unbroken chain of being ascending from inanimate matter through plants and animals to humankind at the summit. Simplistic though it is, scala naturae still pervades popular thinking on the place of humans in the scheme of things (Gould 1990: Chapter 1).

After Linnaeus, taxonomists invoked the web or net, in which relationships were perceived as horizontal, multiple and undirected. The web was the consequence of a superficial level of character analysis, which seemed to show multiple parallelisms between intuitively recognised natural taxa. These taxa seemed polythetic: definable only by an inconsistent set of overlapping morphological characters.

Even before Darwin, some taxonomists used the metaphor of a tree which, after all, described the hierarchy of Linnean classification. Darwin's theory of descent with modification gave the tree a theoretical buttress, but ironically it was mainly zoologists, especially Haeckel, who embraced this botanical metaphor. As Peter Stevens so eloquently relates, botanists accepted Darwinism as an explanation for their systems but continued using polythetic methods.

The unconscious fusion of these disparate elements gave birth to a chimaera: Bessey's cactus (1915; Fig.1). Horizontal relationships of the web and vertical relationships of the tree blended to form grades; extant taxa gave rise to other extant taxa, almost without direction. Gradistic metaphors have taken a variety of visual forms: cacti, blobs, balloons, bubbles, dartboards; but all embodied an imprecise view of relationships, as if higher taxa have evolved along a broad front. The polythetic grade taxa were circumscribed intuitively and explained afterwards in evolutionary terms. Comparison of alternative classifications, having no explicit basis in character analysis, came down to polemics, and the number of alternative classifications proliferated. Stevens (1984: p. 201) provides a telling series of quotations describing this descent into chaos by "evolutionary" systematics. I shall never forget the spectacle of Cronquist and Thorne slugging it out at the International Botanical Congress in Sydney in 1981.

Gilmour and his heirs, the pheneticists, rebelled against the subjectivity of gradism. They rejected any phylogenetic basis for classification, opting instead for the objectivity of overall similarity. Ironically, they resurrected the tree as their visual metaphor, in the form of the unrooted phenogram. This was ironical because it paved the way for the cladists, who reclaimed the tree for phylogenetics, giving it a root by rejecting overall similarity in favour of synapomorphy. Cladistics finally gave an explicit basis for reconstructing the historically unique pattern of evolutionary relationships.

Cronquist leaves the reader of The Evolution and Classification of Flowering Plants in no doubt about the metaphor he accepts for taxonomic methodology (pages 159-163):

"We are all - or nearly all - Besseyans. It is conceivable that some completely new set of concepts will at some future time displace the modified Besseyan concepts now in vogue. ... Yet the prognosis is for evolution rather than revolution."

The Besseyan method which Cronquist favours must "start with the groups which are least modified from the ancestral prototype, rather than those which have undergone most change". Such emphasis on taxa as ancestors typifies gradistic systematics, although it actually comes from residual notions of scala naturae (Stevens 1984: p. 190). Cronquist's diagram of relationships among the subclasses of Magnoliopsida (dicots), on pages 263 and 502 (Fig. 2), confirms his debt to Bessey.

Methods and Criticisms

Cronquist discusses taxonomic methodology, including phylogeny, in the first chapter of his book. A large part of this is a critique of "cladism", which is reproduced essentially unchanged from an earlier paper (Cronquist 1987). His arguments have been refuted by Donoghue & Cantino (1988) and Humphries & Chappill (1988), and I do not intend to repeat the whole debate here. One of Cronquist's strongest criticisms of cladists is against their rejection of paraphyletic taxa. I find this attitude (which is widespread) curious, given that the same people do not hesitate to reject polyphyletic taxa. Yet paraphyletic taxa confound the discussion of relationships among both taxa and characters in exactly the same way as do polyphyletic taxa.

While criticising much of cladistic methodology, Cronquist does accept some aspects, such as the concepts of apomorphy and plesiomorphy. And he finally recommends using Wagner Groundplan-Divergence for reconstructing phylogeny. Wagner's method was original and, when published in 1961, anticipated Hennig's method, which was to become the basis of cladistics after it was translated into English in 1966. Moreover, Farris gave Wagner's name to the first computer algorithm to calculate cladograms using the parsimony criterion, which now underlies the mainstream of cladistic methodology. In an empirical comparison of methods, Churchill et al. (1984) found no essential difference between Wagner Groundplan-Divergence, Wagner [Farris's] parsimony and Hennig's argumentation method. Nevertheless, Wagner Groundplan-Divergence, being a pencil-and-paper method, is limited to small data sets, and has been left far behind by the powerful parsimony algorithms available today. Why then does Cronquist reject parsimony (pages 38-39) while advocating its conceptual but long-superseded predecessor for phylogenetic reconstruction?

Conversely, Cronquist reveals an essentially phenetic approach to systematics on page 2: "taxonomy is a study aimed at producing a system of classification of organisms which best reflects the totality of their similarities and differences". The same view is expressed on page 5, under the heading "Some Principles". This may seem surprising, given Cronquist's claim to a phylogenetic approach. However, as Humphries & Chappill (1988) have explained in detail, emphasising gaps or differences when defining taxa amounts to phenetics, not phylogenetics.

Cronquist makes a common error in suggesting that "taxa are properly established on the basis of multiple correlations of characters" (page 5; emphasis is mine). An hierarchical system is better described by the more general properties of covariation or congruence.

In a fleeting mention of molecular data (page 6), Cronquist completely misses the point by seeing it as basically different from, and more limited than, morphological data. He considers molecular evidence to be useless for delimiting taxa. However, for "procaryotes" it is the only useful data, and has contributed enormously to the understanding of these groups in recent years (Lake 1989).

Origins and Ancestors

Chapters 4 and 5 of the book are devoted to the "big" questions asked at the beginning of this review. Cronquist takes a conventional view of the proto-angiosperm: a magnolioid tree with pinnately-veined leaves and showy strobilus-like bisexual flowers. In this he follows Arber & Parkin (1907) and many modern authorities. The outstanding recent cladistic studies by Doyle & Donoghue (1986) and Donoghue & Doyle (1989), who critically reinterpreted available morphological evidence in extant taxa as well as fossils, have shown this to be the most likely hypothesis. Cronquist correctly rejects the hypothesis of an "Amentiferae" ancestor, in which the unisexual strobili of the Gnetales are homologised with those of hamamelid families with catkins (Wettstein 1907). This homology has not stood up to critical examination, and the presence of vessels in the wood of both groups is now seen as a convergence. The "Amentiferae" belong with the "higher dicots", well removed from the base of the angiosperm tree (Donoghue & Doyle 1989). However, an alternative possibility that is almost as well supported by Donoghue & Doyle's analyses, as well as by the ribosomal RNA sequence data of Zimmer et al. (1989), is that the first angiosperm may have been an herb with palmate leaf venation resembling the waterlilies and basal monocots ("palaeoherbs").

Both cladistic studies have cast serious doubt upon the traditional recognition of monocots and dicots as the fundamental division within the angiosperms. The dicots, especially, appear to be paraphyletic. Cronquist, while acknowledging the latter, accepts the traditional classification, but then he has clearly stated his acceptance of paraphyletic taxa. His scenario for the morphology of the first angiosperms and their possible origins makes a good narrative but is scarcely rigorous science. It will take comprehensive cladistic analyses using all of the available taxa, both fossil and recent, and all of the available evidence, both molecular and morphological, to unravel these mysteries (Donoghue et al. 1989).

Classification System

The second half of this book is a synopsis of Cronquist's (1981) general system of flowering plant families, updated in only minor detail. In fact, Cronquist seems content with his scheme:

"Only a few small changes are introduced, made necessary by advances in knowledge or understanding since 1981. A more substantial revision at this time would be premature; I must leave it to my successors".

The lengthy descriptions in almost every lead of the keys, especially for the higher taxa (subclasses and orders), are symptomatic of the polythetic nature of Cronquist's grade taxa.

Given that Cronquist (1981) expounds his system in much more detail than in this book, I shall not discuss it further. Kanis (1981) analysed and compared Cronquist's system with those of contemporary authors.

Conclusion

To summarise, The Evolution and Classification of Flowering Plants provides a detailed insight into Arthur Cronquist's ideas on the evolutionary process, phylogeny and taxonomic methodology. It is important, given the pervasive influence of Cronquist's system in modern floras and herbaria. I suspect that Cronquist's influence derives more from his system's comprehensiveness and detail, which are almost unique, than from admiration for his methodology. The book is readable and well presented. However, Cronquist's approach is dated, epitomising the unperceived conflict between theory and method which has caused plant systematics to stagnate for a century. I would not teach a course in plant systematics from it.

Acknowledgements

Thanks to Jenny Chappill and Dan Faith for their helpful comments.

References

Arber, E.A. and Parkin, J. (1907). On the origin of angiosperms. Bot. J. Linn. Soc. 38: 29-80.

Bessey, C.E. (1915). The phylogenetic taxonomy of flowering plants. Ann. Missouri Bot. Gard. 2: 109-164.

Churchill, S.P., Wiley, E.O. and Hauser, L.A. (1984). A critique of Wagner Groundplan-Divergence studies and a comparison with other methods of phylogenetic analysis. Taxon 33: 212-232.

Cronquist, A. (1981). An Integrated System of Classification of Flowering Plants. (Columbia Univ. Press: New York).

Cronquist, A. (1987). A botanical critique of cladism. Bot. Rev. 53: 107-128.

Donoghue, M.J. and Cantino, P.D. (1988). Paraphyly, ancestors, and the goals of taxonomy: A botanical defense of cladism. Bot. Rev. 54: 107-128.

Donoghue, M.J. and Doyle, J.A. (1989). Phylogenetic analysis of angiosperms and the relationships of the Hamamelidae. In P. R. Crane and S. Blackmore (eds) Evolution, Systematics and Phylogeny of the Hamamelidae, pp. 17-45. (Clarendon Press: Oxford).

Donoghue, M.J. et al. (1989). The importance of fossils in phylogeny reconstruction. Ann. Rev. Ecol. Syst. 20: 341-460.

Doyle, J.A. and Donoghue, M.J. (1986). Seed plant phylogeny and the origin of angiosperms: an experimental cladistic approach. Bot. Rev. 52: 321-431.

Gould, S.J. (1990). Wonderful Life: The Burgess Shale and the Nature of History. (Hutchinson Radius: London).

Humphries, C.J. and Chappill, J.A. (1988). Systemetics as science: a response to Cronquist. Bot. Rev. 54: 129-144.

Kanis, A. (1981). An introduction to the system of classification used in the Flora of Australia. In A.S. George (ed.) Flora of Australia, Volume 1, pp. 77-111. (AGPS: Canberra).

Lake, J.A. (1989). Origin of the eukaryotic nucleus determined by rate-invariant analyses of ribosomal RNA genes. In B. Fernholm, K. Bremer and H. Jornvall (eds) The Hierarchy of Life, pp. 87-102. (Excerpta Medica: Amsterdam).

Stevens, P.F. (1984). Metaphors and typology in the development of botanical systematics 1690-1960, or the art of putting new wine in old bottles. Taxon 33: 169-211.

Wettstein, R.R. von (1907). Handbuch der Systematischen Botanik, II Band. (Franz Deuticke: Leipzig).

Zimmer, E.A. et al. (1989). Ribosomal RNA phylogenies and flowering plant evolution. In B. Fernholm, K. Bremer and H Jornvall (eds) The Hierarchy of Life, pp. 205-214. (Excerpta Medica: Amsterdam).

Reviewer: Mike Crisp
Australian National Botanic Gardens

(Now at Division of Botany & Zoology
Australian National University)