Angiosperm Phylogeny Group Classification Prof. B. Ravi Prasad Rao
Angiosperm Phylogeny Group Classification Prof. B. Ravi Prasad Rao
INTRODUCTION
Andrea Caesalpino (1583), proposed the first methodical arrangement of plants into
15 ‘higher genera’ on the basis of the structure of trunk and fructification. In the mid 18th
century, Linnaeus published ‘Species Plantarum’ with a complete list of 7000 species of
flowering plants known at that time which were classified under 1000 genera and 23 classes.
Contributions of Linnaeus and his immediate successors have formed a strong base for the
current day classification. Linnaeus and other botanists till the end of 20 th century (e.g.
Bentham & Hooker, 1862-83) classified flowering plants for ease of the identification.
Systematic research in the past 150 years and particularly during the last four decades have
greatly influenced our views on the classification of plants especially angiosperms. Cladistic
information strongly points to the fact that simplistic division of angiosperms into monocots
and dicots do not reflect phylogenetic history. During 1990s, reconstruction of flowering
plant phylogeny took a great step forward. Rapidly accumulating DNA sequences,
particularly from the plastid gene rbcL provided new and informative sets of data. Cladistic
analysis of these was also much improved through the development of phylogenetic theory
Group (APG), came together to propose a classification based on sound phylogenetic data
sets in 1998. Till 2010, three versions of classifications have been proposed by this group that
was published in 1998, 2003 and 2009, each superseding the previous systems. Peter F.
1
Stevens, one of the authors of all three of the APG classifications maintains a web site,
APweb, hosted by the Missouri Botanical Garden, which regularly update the information
regularly since 2001 (Stevens, 2001 onwards). This web site is a prime source of information
With the efforts of 29 botanical systematists around the world, a new system of
classification has been proposed under the umbrella of Angiosperm Phylogeny Group
succinctly known as APG 1998. This system is based on sound phylogenetic principle of
more strongly on molecular studies with reference to DNA sequences of two chloroplast
genes (cpDNA; atpB and rbcL) and one gene coding for ribosomes (nuclear ribosomal 18s
DNA).
APG 1998 recognized 462 families, which were grouped into 40 monophyletic orders
eudicots, core eudicots, rosids including eurosids I and eurosids II, asterids including
eausterids I and euasterids II. The focus was on orders and less on families. Many families
were not classified to order because their positions were either uncertain or unknown. In this
classification, there are 81 unplaced families, 11 placed towards the beginning, 25 towards
the end and 45 in the informal groups. Alternative options are provided for some groups, in
which a number of families can either be regarded as separate taxa or can be merged into a
phylogenetic tree of flowering plants and establishment of major groups or clades within. The
2
authors chose to adopt a broad approach in defining the limits of orders, resulting in the
monophyletic clade, but the dicots are placed in separate groups, some are basal to monocots
and the remaining is considered as eudicots or 'true dicots'. Various monocot taxa are placed
in between primitive angiosperms and eudicots, thus overcoming the problem of paraphyly.
APG II (2003)
After five years since the publication of APG classification, considering further
classification (APG II) was proposed in 2003. The APG II classification recognized 457
families (5 less than APG 1998) and 45 orders (5 more than APG 1998). Within 457
families, there are 55 optional segregates (presented in square brackets), thereby considering
minimum number of families as 402. Of the 45 orders, 44 are placed in 11 informal groups
which were considered more or less monophyletic. Contrary to APG 1998 which has 81
unplaced families, in APG II, this number has been reduced to 40. The list of unplaced
families in the beginning has been reduced to 4 and uncertain families towards the end to 9.
To fill further gaps in APG II and to develop a much more stabilized classification,
with recommendations of different scientist groups around the world, a revised and updated
version of APG was published in October, 2009 by a team of 8 scientists in the name of APG
III. This classification followed Backlund & Bremer (1998) principles of rank-based
APG III recognizes 413 families. Except ten families, viz., Dasypogonaceae,
3
Metteniusaceae, Oncothecaceae, Cynomoriaceae and Apodanthaceae, rest of the 403 families
and Chloranthales (covering 8 families) are unplaced, i.e. not included under any clade, and
kept in the begining. The remaining 55 orders are assigned to 11 clades or groups:
magnoliids, monocots, commelinids, eudicots, core eudicots, rosids, fabids, malvids, asterids,
Cynomoriaceae, Apodanthaceae and the genera Gumillea Ruiz & Pav., Petenaea Lundell and
abstract of APG III classification with respect to groups, orders and families is presented in
Table 1 and the complete list of 413 families as per LAPG III sequence and along with
number of genera and species for each family, world distribution and representative genera
(including type genus) can be consulted through Rao and Prasanna (2010).
Trochodendrales, Vitales and Zygophyllales are recognized in APG III. Many genera and
families which were unplaced in APG and APG II classifications are now included in
Newly segregated families of APG III for genera previously treated under other APG-
(both Brassicales), Linderniaceae and Thomandersiaceae (both Lamiales). Many families that
were in square brackets in APG II are now treated clearly: Brassicaceae s.l. have been split
4
into 3 families viz., Brassicaceae s.s., Capparaceae and Cleomaceae. However, in some cases,
been clarified, for which additional orders have been proposed that were not previously
recognized.
Of the 413 families of APG III, 121 are monotypic, represented by a single genus and
28 of them are represented by single species. As per the current update information given in
the APG website (Stevens, 2001 onwards), the largest family is Orchidaceae (27,800
species/880 genera). Other dominant families are: Asteraceae (23,600 species / 1620 genera),
Perusal of the literature has revealed that of the 413 families, 259 are represented in
India. The distribution pattern of the families indicate that 5 families are endemic to
Madagascar, 4 to Cape Province of South Africa and Chile each, 3 to New Caledonia, one
family each for Somalia, Tasmania, China, Fiji islands and Mexico. It is interesting to note
that 24 families are endemic to Australian continent and surrounding islands. African
A formal classification of the land plants that is compatible with the APG III
classification was proposed by Chase & Reveal (2009). For detailed information on orders
and family delimitations, readers are advised to consult APWeb (Stevens, 2001 onwards) and
for molecular and allied datasets used for separation of different clades, orders and families in
5
APG III vis-a-vis BENTHAM AND HOOKER CLASSIFICATION
An attempt has been made to compare the positions of families as treated in Bentham
and Hooker’s classification with that of 413 families of APG III. Thirty five families, (mostly
monotypic) have been described after the publication of Genera Plantarum (Bentham and
Hooker, 1862-83). Significant changes in APG III treatment of families vis-à-vis Bentham
and Hookers’s system include the following. Liliaceae s.l. is split into 14 families and many
respectively (Refer to Table 2 for details). It is interesting to note that APG like that of
Haston et al. (2007), developed a linear sequence of families (called as LAPG) based
on APG II classification and provided a list of 479 families. Haston et al. (2009) revised
LAPG for the new APG III classification as LAPG III and considered 413 families. Although
the methodology followed has been questioned by Hawthorne et al. (2008), it is concluded
6
that in the absence of any obviously better way of generating a linear sequence from a
phylogenetic tree, APG III can be considered a viable system with options open to modify the
methodology to ensure stability. It is pertinent to note that many European herbaria have
agreed to adopt LAPG. The LAPG sequence has been accepted by RBG Kew, RBG
Edinburgh, the Natural History Museum (London), the Musée National d’Histoire Naturelle
(Paris), Conservatoire et Jardin Botaniques (Geneva) and the National Herbarium Nederland
In APG III, families are alphabetically arranged in clades. In LAPG, families within
the clades are reorganized. With respect to unplaced orders and families at the beginning,
orders sequence in LAPG is similar to that of APG III, but families are re-organised. In
Magnoliid clade, Magnoliales are followed by Laurales, vice-versa in APG III. In the clade
Monocots, orders are in the same sequence as that of APG III. In Commelinids, Zingiberales
are followed by Poales, vice-versa in APG III. Unplaced Dasypogonaceae is retained in the
same clade, but kept in between Zingiberales and Poales, which were kept at the beginning in
unchanged. In Eudicots, orders are in the same sequence like that of APG III. Even unplaced
Sabiaceae is retained in the same position between Ranunculales and Proteales. In Core
Eudicots, orders are in the same sequence like that of APG III. Dilleniaceae, the unplaced
family is retained in the same position after Gunnerales. Cynomoriaceae considered as taxa of
uncertain position and kept at the end in APG III is included in core eudicots in LAPG. There
5 unplaced families kept at the beginning in APG III, 3 families are kept at the beginning and
7
Currently some of the problems of placement of families and genera have been
research/APweb.
8
REFERENCES