Vascular Plant Classification in botanic gardens – an international approach

In this item James Wearn provides an overview for botanists on the Vascular Plant Classification Committee (VPCC), of which he is the Secretary, and the recent agreement achieved for the rearrangement of plant specimens in herbaria and other collections.

20 Sep 2010

  • Close Thanks for liking this page. Tell us why by adding a comment at the bottom.
Euphorbia pulcherrima and Rafflesia pricei

Unexpected relationships of parasitic plants recognised by APG III. Rafflesia pricei (right) placed in Malpighiales with species such as Euphorbia pulcherrima (left).

The addition of an extension to the Herbarium, Library, Art and Archives at Kew, designated as ‘Wing E’, provided an opportunity not only for expansion, but also for systematic rearrangement of the vast dried and spirit collections of plants (some seven million specimens).

The use of several disparate, and outdated, systems of collection arrangement across different herbaria is not optimal for international research. However, considering the large numbers of specimens that require rearrangement, it is not difficult to understand why the sequence had not been updated as classification had moved forward.

It was found that several other European institutions were also undergoing building extensions, or would be able to re-arrange their collections in the near future and representatives from these organisations became the core membership of the Vascular Plant Classification Committee (VPCC). Participation of multiple organisations was welcomed in order to reach a widely accepted vascular plant classification system ― a groundbreaking scientific and political accomplishment.

Formation of the Vascular Plant Classification Committee

 The process of forming the VPCC was initiated through a “Workshop on vascular plant classification system to be adopted by Kew” on 27 June 2008, in which Kew botanists discussed possible classification systems and criteria for assessing possible options to be embraced by Kew. Discussions were subsequently held with Royal Botanic Garden Edinburgh and the Department of Botany at the Natural History Museum in London.

The workshop comprised a discussion of systems adopted by other organisations and criteria for assessing possible options to be embraced by Kew. Coupled with further internal and external submissions, this led to the prospect of preparation of an agreed version of a vascular plant classification for lycophytes, pteridophytes, gymnosperms and angiosperms that would be used throughout Kew with regard to family circumscription and the sequence to be followed when arranging the collections in the Herbarium

An international concern

The international consortium that became the VPCC was formed during external discussions with other European herbaria and museums, further enhanced by input from extra-European (non-Committee) experts including fern expert Alan Smith (Berkeley, California) and the creator of the Angiosperm Phylogeny website Peter Stevens (Missouri Botanical Garden). The Committee included members from six institutes (see membership list at the end of this article).

Cytinus hypocistis and Muntingia calabura.
Cytinus hypocistis (right) is now in Malvales with species such as Muntingia calabura (left).

The Royal Horticultural Society and several smaller herbaria were also interested in the outcome of the VPCC discussions. Therefore, the VPCC solicited opinion on the most appropriate templates for the four groups of vascular plants externally too, in order to make the most informed decisions by involving a large community of expert botanists.

Significant progress

The VPCC, tasked with reaching an agreement on family classification and circumscription and herbarium sequences for all vascular plants, reached a consensus in early 2009 following three meetings and many email and telephone communications. The chance to capitalise on an opportunity for an international agreement had been seized. This was momentous. 

It should be noted that generic matters were not the focus of the VPCC and discussion was restricted to family level or above. 

Lycophytes & Gymnosperms

The agreed classification of lycophytes and gymnosperms will follow Mabberley (2009), except for the removal of class Gnetopsida from the Pinopsida (a reversion to its separate position in Mabberley 1997).

David Mabberley commented: “the evidence used for the position of Gnetopsida in Plant-book edition 3 is still controversial, with different analyses coming up with different results, so a VPCC agreement to revert to the arrangement in Plant-book edition 2 is acceptable, but with the three families [Gnetaceae, Ephedraceae and Welwitschiaceae] in the sequence of the later edition”. Mabberley’s classification is not a single system, but a condensation of modern works. 


The classification of Smith et al. (2008), a slight modification of Smith et al. (2006), will be used as the template for pteridophytes.


At the time of the VPCC meetings, the Angiosperm Phylogeny Group were incorporating significant modifications into the previously published APG II (APG 2003) classification as a result of new molecular and morphological studies. A draft version of APG III was available during the process (APG 2009). 

Mark Chase, Keeper of the Jodrell Laboratory at Kew, drew particular attention to the benefits of grouping taxa to decrease the number of families. This aids teaching and often maximises diagnostic field characters. He said that circumscriptions should allow the user to gain information with higher levels in a tree combining taxa. Thus, having monogenic families leads to redundancy in the classification. This was agreed by the Committee, although there are cases in which monogeneric families are the best solution (e.g. Nelumbonaceae, Acoraceae).

The circumscriptions of Mabberley (2008) were used as a benchmark. Deviations from this were from the newest research included in the updated angiosperm phylogeny (APG 2009) and subsequently in Mabberley (2009). Thus, the new APG III was adopted for angiosperm classification. As the changes to angiosperm classification have now been published, I have included here only some of the most significant alterations to family circumscription and placement accompanied by typical examples.

1. Asparagales

Xanthorrhoeaceae has been circumscribed in its broad sense (to include Hemerocallidaceae and Asphodelaceae). This enlarged family is a change from the ‘unpopular’ APG II with narrower circumscriptions. The heterogeneity of this family is recognised. The type genus is Xanthorrhoea, native to Australia and commonly called ‘grass trees’ due to their distinctive pachycaul form.

2. Myrtales

Neither a broad nor a narrow circumscription of Penaeaceae is currently robustly supported. More supporting evidence would be required in order to lump families previously recognised separately in APG II and now maintained in APG III. This family includes the genera Penaea and Saltera, both common small, ericoid shrubs with colourful inflorescences found in the fynbos flora of southern Africa.

3. Malpighiales

An unexpected relationship was revealed, that of placement of the unique, parasitic plants of the Rafflesiaceae within Euphorbiaceae s.l. such that Peraceae must be recognised. The Rafflesiaceae family is characterised by the charismatic type genus Rafflesia, bearing the largest flowers of all angiosperms (c. 1 m in diameter), although somewhat resembling fungal basidiomes. Rafflesia species have become a focus for conservation due to their rarity and dependency upon a single host genus (the vine Tetrastigma, Vitaceae). Additionally, their popularity with ecotourists has provided an incentive for local efforts in Southeast Asia.

4. Geraniales

Alternatives of (a) accepting Geraniaceae s.s.; (b) combining with Ledocarpaceae; or (c) combining with Ledocarpaceae and Vivianiaceae were discussed at length within the Committee and with Peter Stevens. The combination of Geraniaceae and Ledocarpaceae with Vivianiaceae (as in Kubitzki 2007) was also considered. It was agreed that pending further knowledge, Ledocarpaceae and Vivianiaceae, as poorly-known groups, should be maintained, although sister to Geraniaceae. Geranium is a large and widespread genus of plants, cultivated widely for their ornamental appeal. Conversely, members of the closely-related family Ledocarpaceae, including Wendtia, restricted to Chile and Argentina, are considerably less eminent.

Herbarium rearrangement

Once templates had been chosen, the practicality of the task of organising multidimensional phylogenetic trees into a linear sequence to be implemented in herbarium collections was considered. Reorganisation of Kew’s Herbarium will take approximately two years to be completed, as it is a significant undertaking to reorder millions of specimens into a usable linear sequence within a multi–floored building totalling five wings. Additionally, several facets comprise the collections, including spirit-stored material, dried specimens on sheets, carpological material in boxes and various other supplementary assemblages such as illustrations and monographs in the Herbarium, which all require careful consideration.

Haston et al. (2007) and Hawthorne & Hughes (2008) were used as a starting point for issues pertinent to the organisation of a linear sequence. The Committee contemplated the degree of arbitrariness from tree to sequence order and was aware that many optima exist, although a single version needed to be, and was, agreed. Throughout the sequence, smaller groups have been placed before larger sister taxa.

One of the main issues arising was the placement of the monocots, either after the Magnoliales or at the end of the entire herbarium sequence. The opinion that eudicots and monocots are sister taxa was put forward and so, in theory, they could be swivelled either way. Thus, there would be no strong reason to put them earlier or last in the sequence. It was also noted that other institutes, especially smaller herbaria, which already have the monocots placed at one end, would benefit from this flexibility more than those where monocots are kept physically separate. At RBG Edinburgh, monocots are placed in the middle between Magnoliales and eudicots in line with the paper by Haston et al. (2007) and this was the agreed position for herbaria where a complete separation was not in place.  


Article by James Wearn, Research Assistant to the Keeper, Herbarium, Library, Art & Archives, Kew 

Full list of references for this article

Donate now - support science and research at Kew

By making a donation to Kew today you can help our scientists to find out more about the fascinating world of plants, break new ground and inspire generations of young people to get to know plants better.

Our scientific programmes are focused on understanding plants and conserving the world's plant life and habitats at risk. Plants are essential to life on earth. In a world where the sustainability of the planet’s rich biodiversity is becoming less certain, Kew’s science work is ever more critical. Find out how your donation can make a difference.

Give now and support Kew’s vital science and research work

Scientific Research & Data


Browse Kew News

No comments on 'Vascular Plant Classification in botanic gardens – an international approach'

See your favourite reasons to visit