Journal of Biogeography (J. Biogeogr.) (2006) 33, 1929–1944
ORIGINAL
ARTICLE
Historical biogeography of
Rhododendron section Vireya and the
Malesian Archipelago
Gillian K. Brown1,2*, Gareth Nelson1 and Pauline Y. Ladiges1
1
School of Botany, The University of
Melbourne, Vic., 3010 Australia, 2Australian
National Herbarium, Centre for Plant
Biodiversity Research, CSIRO Plant Industry,
Canberra, 2601 Australia
ABSTRACT
Aim Vireya rhododendrons are distinctive and easily recognizable by their general
form; however, they are virtually circumscribed geographically, predominantly
distributed throughout the biogeographically intriguing Malesian Archipelago.
Hypotheses of the evolutionary relationships of the group have been proposed but
the biogeography of vireyas has not been analysed based on molecular phylogeny.
Recently, the first detailed molecular phylogenetic investigation of section Vireya
was completed based on cp- and nrDNA sequence data, therefore making this
cladistic biogeographic study of vireya rhododendrons possible.
Location Malesia, Australia, Solomon Islands, Taiwan, Himalayas, north
Vietnam and south China.
Methods Based on distribution maps, areas of endemism were determined for
the biogeographic region of Malesia. Area relationships were analysed based on a
recent molecular phylogeny of species in section Vireya. The method of paralogyfree subtree analysis was applied.
Results Individual distribution maps were produced for 74 species of
Rhododendron section Vireya. Species clades with bootstrap support proved to
be biogeographically informative. Major clades correspond to three regions:
eastern Malesia, western/middle Malesia and Taiwan/north Vietnam/south China.
Within eastern Malesia, Australia, New Guinea, the Bismarck Archipelago and
Solomon Islands are related. In western Malesia, northern Philippines, Borneo,
southern Moluccas and north and west Sulawesi are related. These areas are more
distantly related to Sumatra, the Malay peninsula, Java, Bali, Palawan, Lesser
Sunda islands and the southern Philippines. The position of the Himalayas is
equivocal and part of a basal polytomy in the summary area cladogram.
Main conclusions Two alternative hypotheses are proposed for the evolution of
vireya rhododendrons based on the pattern of area relationships. The first
hypothesis is that the vireyas are an old group, with ancestors present on
Gondwana, rifting north in the Cretaceous. The second alternative hypothesis is
that vireyas are a young group that has dispersed eastwards from India to
Australia and the Solomon Islands since the current Malesian islands formed.
*Correspondence: Gillian K. Brown, School of
Botany, The University of Melbourne, Vic.,
3010, Australia.
E-mail: browngk@unimelb.edu.au
Keywords
Cladistic biogeography, distribution maps, Malesia, paralogy-free subtree
analysis, Rhododendron section Vireya, Southeast Asia.
The Malesian Archipelago has been of interest to naturalists
and scientists alike since 1869, when Alfred Russel Wallace
recognized a biotic discontinuity between the Indo-Malay and
Australian bird and mammal faunas (Michaux, 1991). What
became to be known as Wallace’s line separates the islands of
Bali and Lombok, and Borneo and Sulawesi, and runs to the
ª 2006 The Authors
Journal compilation ª 2006 Blackwell Publishing Ltd
www.blackwellpublishing.com/jbi
doi:10.1111/j.1365-2699.2006.01548.x
INTRODUCTION
1929
G. K. Brown, G. Nelson and P. Y. Ladiges
south-east of the Philippines (George, 1981). Subsequent to
Wallace, in the late 1800s and early 1900s, Huxley, Lydekker
and Weber proposed other lines to separate the Australian and
Asian faunas (George, 1981). All of these lines fall in the
middle of the Malesian Archipelago in the region bounded by
the Sunda and Sahul continental shelves: the Lesser Sunda
Islands, Moluccas, Sulawesi, and Philippines. This region was
aptly named ‘Wallacea’ by Dickerson in 1928 (Cox, 2001).
The Malesian Archipelago continues to be of the great
interest to biogeographers today, with a large number of
studies based on floral and faunal taxa endemic to Malesia and
its neighbouring regions (e.g. Whitmore, 1987; Ladiges et al.,
1991; Hall & Holloway, 1998; Evans et al., 2003; Heads, 2003;
Ladiges et al., 2003). Most studies have concentrated only on
taxa that are distributed in either the east or the west of
Malesia and that do not span the entire archipelago, as in fact
do many taxa, including rhododendrons.
Vireya biogeography
Rhododendron L. is a species-rich genus of flowering plants
(family Ericaceae Juss.), comprising over 1000 species. It is
currently divided into eight subgenera, which include various
sections and subsections (Sleumer, 1980; Chamberlain et al.,
1996). Rhododendron subg. Rhododendron is the lepidote
(scaly) group, based on the presence of scales on leaves and
other organs. Section Vireya (Blume) Copel. f. is the largest of
three sections within this subgenus, characterized traditionally
by seeds with tailed appendages at both ends, capsule valves
that twist on opening, placentas that separate as thread-like
structures and by their predominantly Malesian distribution.
The last revision of section Vireya was by Sleumer (1966).
An analysis of the biogeography of vireyas based on
molecular phylogeny has yet to be performed. Several authors
have proposed tentative explanations for the origin and
evolution of species in particular areas of Malesia. Copeland
(1929) noted that taxa in the Philippines appear to be related
to species on neighbouring islands, including Borneo and New
Guinea, and speculated about their possible origins. van
Balgooy (1987) made similar observations for the flora of
Sulawesi, concluding that taxa came from northern, southern
and eastern routes, rather than from the west, despite being
geographically close to Borneo. Stevens (1982) suggested that
diversification of Ericaceae in New Guinea has occurred since
the late Miocene (c. 10 Ma) at the earliest, and was dependent
on the arrival of only a few ancestors from West Malesia. Based
on interspecies promiscuity among vireyas, Williams & Rouse
(1997) also considered the section to be of a relatively young
age.
Specht (1988) considered Rhododendron to be an old genus.
He hypothesized that rhododendrons were on Gondwana
before its break-up, and that the lepidote group (subg.
Rhododendron), including sect. Vireya, dispersed northwards
from Australia into the archipelago, while the elepidote
rhododendrons (all other subgenera) moved north with India
and dispersed outwards from there. Irving & Hebda (1993)
1930
proposed that, in their early history, rhododendrons extended
more or less continuously from North America to Europe and
Greenland, and into China and north-east Asia, with their
present-day, more restricted distribution a result of climatic
deterioration over time. They hypothesized that the vireyas are
relatively young, evolving in the last few million years, because
‘they are concentrated in one essentially continuous region’,
and the terranes where they are found are of recent origin.
Irving & Hebda (1993) acknowledged that their assumptions
and hypotheses were speculative as they had no specialized
knowledge of rhododendrons, and instead based their ideas
solely on geology, climate and the ecology of present-day
species.
The recent panbiogeographic study of Heads (2003) is the
only biogeographic study of the vireyas that takes the
taxonomy of the section into account, although it is not based
explicitly on phylogenetic relationships because no cladistic
analysis was available until now. Heads (2003) used the Flora
Malesiana classification (Sleumer, 1966) and all subsequent
taxonomic revisions to map generalized distributions of 276
species of vireya in their subsections and series. He related
centres of endemism to tectonic history. Heads (2003)
indicated that this approach related ‘the characters underlying
the taxa, rather than the taxa per se’.
The first detailed phylogenetic investigation of section
Vireya was only recently completed, and was based on DNA
sequence data (Brown et al., 2006, in press). Because biogeographic studies should be based on sound systematics
(Humphries & Parenti, 1999), it is understandable that no
detailed cladistic biogeographic investigations have been
completed for the section.
In our study we have prepared distribution maps for species
of Vireya, including all those represented in the recent
molecular phylogenetic analyses (Brown et al., 2006, in press).
We use these distribution maps to delineate areas of
endemism, superimposing the areas on the molecular taxon
cladogram. The cladogram summarizes all clades (nodes) that
were resolved with support, based on chloroplast DNA
(psbA-trnH and trnT-trnL) regions. The method of paralogyfree subtree analysis (Nelson & Ladiges, 1996) is applied to
analyse the relationships of the geographic areas of endemism
in the region of Malesia and to infer the evolutionary history of
Rhododendron section Vireya.
METHODS
Molecular phylogeny
A summary molecular phylogeny based on cpDNA sequences
was produced based on parsimony and Bayesian analyses in
Brown et al. (2006). A total of 65 species were sequenced for
this summary phylogeny (Fig. 2); 58 of these species were
sequenced for both the trnT-trnL and the psbA-trnH intergenic
spacer regions, while the other 7 species (marked with an
asterisk in Fig. 2) were only sequenced for the trnT-trnL
intergenic spacer due to amplification difficulties (Brown
Journal of Biogeography 33, 1929–1944
ª 2006 The Authors. Journal compilation ª 2006 Blackwell Publishing Ltd
Historical biogeography of Vireya rhododendrons
et al., 2006). Bootstrap values (> 50%) indicating support for
nodes are shown.
Species maps
Distribution maps were available in the literature for only
about 20 species of Vireya (see references in Heads, 2003).
Maps of all species in the summary molecular phylogeny were
produced using the ANH map program (version 2.0, 13
November 1997, Canberra, Australia). The species included in
the summary phylogeny from outside the section Vireya were
not mapped; instead, their distributions for the cladistic
biogeographic analyses were taken from the literature
(Chamberlain et al., 1996).
To cover the total known distribution of each species,
locality information was obtained from herbarium specimens,
and supplemented with locality information from the literature (data available on request; Sleumer, 1958, 1973; Argent &
Madulid, 1995, 1998; Takeuchi, 2000). Specimens from several
herbaria were examined for locality information: Arnold
Arboretum, Harvard University (A), Australian National
Herbarium (CANB), Royal Botanic Garden Edinburgh (E),
Nationaal Herbarium Nederland, Leiden, University branch
(L), and the New York Botanical Garden (NY). Where the
latitudes and longitudes were not recorded on the specimen
label, geocodes of these localities were estimated for the closest
known location from a number of sources: PNG (Papua New
Guinea) localities list (1991–present, Australian National
Herbarium, unpubl. data); the Australian National Herbarium
Specimen Information Register (ANHSIR) data base; IBIS
(Indonesia Biodiversity Information System) Herbarium Bogoriense, Indonesia; MicrosoftÒ EncartaÒ Interactive World
Atlas, 2001; Orchids of Sarawak (Beaman et al., 2001); Global
Gazetteer (http://www.calle.com/world); NIMA Geographic
Names
Database
(GNDB;
http://gnpswww.nima.mil/
geonames/GNS/index.jsp); Taiwan mountain localities
(http://www.indexmundi.com/taiwan).
because no species in this study extends further east into this
island arc), north and west Sulawesi (J), southern Philippines
(K), northern Philippines (L), Palawan (M), north and south
Borneo (N), Java and Bali (O), north and south Sumatra (P),
Malay Peninsula (Q), Taiwan (R), north Vietnam and south
China (S), and the Himalayas (T).
Paralogy-free subtree analysis
The method of paralogy-free subtree analysis removes paralogous (redundant) nodes and utilizes only the data relevant to
cladistic biogeography (Nelson & Ladiges, 1996). Geographic
paralogy is revealed by geographic distributions that are
duplicated or found to be overlapping among related taxa; it is
analogous to paralogy of molecular systematics (Nelson &
Ladiges, 1996; Ladiges, 1998; Humphries & Parenti, 1999;
Crisci, 2001). Different causes, such as dispersal, sympatric
speciation and imprecise characterization of geographic areas,
can result in geographically paralogous nodes (Nelson &
Ladiges, 1996). Humphries & Parenti (1999) considered that
paralogy-free subtrees ‘should lead to straightforward ways of
expressing area interrelationships’, and Crisp et al. (1999)
commented that the method ‘has the considerable advantage
of simplicity in concept and application’.
Each terminal taxon of the molecular phylogeny was
replaced by a list of the geographic areas in which it occurs.
Paralogy-free subtrees (with no area duplication) were found
by inspection and coded as characters for a parsimony analysis.
Characters were binary coded, with missing areas coded as
question marks. Multiple parsimonious trees are recovered
because of the number of question marks in the data matrix,
which result in resolved nodes that are unsupported by data.
Rather than compute a strict consensus tree, the shortest tree
with the least resolution (the minimal tree) was found (Nelson
& Ladiges, 1996). The program Hennig86 (Farris, 1988), with
option ‘dos equis’, was used for the parsimony analysis and
search for the minimal tree. The minimal tree summarizes the
relationships of the areas to one another (area cladogram).
Areas of endemism
Recognizing areas appropriate for the level of study can be
problematic (Platnick, 1991) and a number of ways to best
define areas have been proposed (e.g. Humphries & Parenti,
1986; Axelius, 1991; Harold & Mooi, 1994; Morrone &
Carpenter, 1994). We recognized areas based on geological
information (Hall, 2002), previous biogeographic studies of
the region (de Boer & Duffels, 1996; Ridder-Numan, 1998;
Heads, 2001), and the distributions of the species included in
these analyses.
Twenty biogeographic areas, A–T, were found throughout
the Malesian Archipelago and neighbouring regions (Fig. 1):
Bismarck Archipelago and the Solomon Islands (A), Papuan
Peninsular (B), north-eastern Australia (C), New Guinea
craton (D), central New Guinea (E), northern New Guinea
(F), Vogelkop Peninsula (G), south Moluccas (H), Lesser
Sunda Islands (I) (restricted to the islands to the west of Flores
RESULTS
Species maps
The distributions of 74 species were mapped (Appendix S1 in
Supplementary Material), and are summarized below.
Borneo
Nine of the mapped species are endemic to the island of
Borneo (Appendix S1a, b, d, i, o, p, y, ll & tt): R. abietifolium
Sleumer, R. acuminatum Hook.f., R. alborugosum Argent &
J.Dransf., R. burttii P.Woods, R. edanoi ssp. pneumonanthum
(Sleumer) Argent, R. ericoides Low ex. Hook.f., R. intranervatum Sleumer, R. lowii Hook.f. and R. polyanthemum Sleumer.
Four of these taxa are restricted to the Gunung (Mount)
Kinabalu area, which is located in the Malaysian state of Sabah:
Journal of Biogeography 33, 1929–1944
ª 2006 The Authors. Journal compilation ª 2006 Blackwell Publishing Ltd
1931
G. K. Brown, G. Nelson and P. Y. Ladiges
T
R Taiwan
S
Indochina
L
Philippines
M
K
Malay
Peninsula
Moluccas
Q
Sulawesi
J
N
Sumatra
Borneo
P
Bismark
Archipelago &
Solomon Islands
New Guinea
G
F
H
D
E
O
Java
A
B
I
Lesser Sunda Islands
C
Australia
Figure 1 Map of Malesia and surrounding region. Islands and 20 areas of endemism are outlined and labelled. The political division
on the island of New Guinea is indicated; the west Irian Jaya (Indonesia) and the east PNG (Papua New Guinea). The 20 areas of endemism
used in the paralogy-free subtree analysis are: Bismarck Archipelago and the Solomon Islands (A), Papuan Peninsular (B), north-eastern
Australia (C), New Guinea craton (D), central New Guinea (E), northern New Guinea (F), Vogelkop Peninsula (G), south Moluccas (H),
Lesser Sunda Islands (I), north and west Sulawesi (J), southern Philippines (K), northern Philippines (L), Palawan (M), Borneo (N),
Java and Bali (O), Sumatra (P), Malay Peninsula (Q), Taiwan (R), north Vietnam and south China (S) and the Himalayas (T). Two
areas outlined with a stippled line are defined based on geological information and previous biogeographic studies; no species of
vireya in this study are found in defined areas, hence the areas are excluded.
R. abietifolium, R. acuminatum, R. lowii and R. polyanthemum
(Appendix S1a, b, ll & tt). Rhododendron ericoides has also
been recorded from G. Kinabalu, although it is not endemic
there, and is also recorded from the G. Mulu National Park in
Sarawak (Malaysia; Appendix S1p). Rhododendron burttii has
been collected from both Sabah and Sarawak, the two
Malaysian states on the island of Borneo (Appendix S1i).
The other three endemic Bornean species are distributed in
Sarawak and the Indonesian province of Kalimantan (Appendix S1d, o & y): R. alborugosum, R. edanoi ssp. pneumonanthum and R. intranervatum.
Rhododendron javanicum (Blume) Benn. and R. malayanum
Jack (Appendix S1bb & oo), while not endemic to Borneo,
have been collected from numerous locations on this island.
1932
Java
Only one endemic species of vireya is mapped for the island of
Java (Appendix S1e): R. album Blume. Several of the widespread species, however, are also recorded from Java (Appendix S1l, bb, oo, aaa & vvv): R. citrinum (Hassk.)Hassk.,
R. javanicum, R. malayanum, R. retusum (Blume) Benn. and
R. zollingeri J.J.Sm.
Lesser Sunda Islands
None of the mapped species is endemic to the Lesser Sunda
Islands. Several of the widespread taxa — R. citrinum, R. javanicum and R. zollingeri (Appendix S1l, bb & vvv) — are found
Journal of Biogeography 33, 1929–1944
ª 2006 The Authors. Journal compilation ª 2006 Blackwell Publishing Ltd
Historical biogeography of Vireya rhododendrons
on the island of Bali, while one of these species has also been
collected further east on the islands of Flores and Lombok
(Appendix S1vvv): R. zollingeri. One possible explanation for
the wide distribution of R. zollingeri in the Lesser Sunda
Islands is dispersal, although further study of populations
across the Lesser Sunda Island is required to test dispersal and
vicariance hypotheses.
Malay Peninsula
Only one mapped species is endemic to the Malay Peninsula:
R. robinsonii Ridl. It is found throughout the western and
central states of Selangor, Pahang and Perak (Appendix S1ccc).
Rhododendron jasminiflorum var. heusseri (J.J.Sm.) Sleumer,
R. javanicum and R. malayanum (Appendix S1aa, bb & oo),
while not endemic to the Malay Peninsula, have also been
collected there.
Moluccas
Rhododendron meliphagidum J.J.Sm. and R. ruttenii J.J.Sm. are
endemic to the Moluccas; both found only in the Manusela
National Park on the island of Ceram (Appendix S1pp & fff).
Rhododendron malayanum, which is widespread across the
archipelago, has also been recorded from Ceram (Appendix
S1oo).
New Guinea
Twenty-nine species endemic to the island of New Guinea,
which includes both the republic of Papua New Guinea (PNG)
and the Indonesian province of Irian Jaya, have been mapped
(Appendix S1h, j, k, m, n, s, t, u, v, w, x, z, ee, ff, hh, ii, nn, qq,
ss, zz, eee, hhh, iii, jjj, lll, nnn, rrr, ttt & uuu): R. baenitzianum
Lauterb., R. carringtoniae F.Muell., R. christi F.Först, R. commonae F.Först, R. culminicola F.Muell., R. gardenia Schltr.,
R. goodenoughii Sleumer, R. gracilentum F.Muell., R. herzogii
Warb., R. hyacinthosmum Sleumer, R. inconspicuum J.J.Sm.,
R. inundatum Sleumer, R. konori Becc., R. laetum J.J.Sm.,
R. leptanthum F.Muell., R. leucogigas Sleumer, R. maius
(J.J.Sm.) Sleumer, R. multinervium Sleumer, R. phaeochitum
F.Muell., R. rarum Schltr., R. rubineiflorum Craven, R. saxifragoides J.J.Sm., R. solitarium Sleumer, R. spondylophyllum
F.Muell., R. superbum Sleumer, R. tuba Sleumer, R. vitis-idaea
Sleumer, R. womersleyi Sleumer and R. zoelleri Warb.
Of these species, 17 have been found only in the east
(PNG) (Appendix S1h, j, k, m, t, u, w, hh, qq, ss, zz, eee, iii,
jjj, lll, nnn & ttt), predominantly distributed throughout the
central mountain ranges, but also on the Huon Peninsula, in
the northern Torricelli Mountains, and on the islands of
New Britain, New Ireland and Goodenough Island. In
contrast, only three species have distributions that are
restricted to the west of the island (Irian Jaya): one in the
central region, one in the Cyclops Mountains to the north
and the other on the Vogelkop Peninsula (Appendix S1z, ff
& ii). The bias between east and west distribution is possibly
a collecting artefact, with Irian Jaya not as well explored as
PNG. Nine of the endemic species are distributed across the
whole island, extending from the Morobe Peninsula, through
the central PNG mountain range and central Irian Jaya, to
the Vogelkop Peninsula (Appendix S1n, s, v, x, ee, nn, hhh,
rrr & uuu).
Rhododendron loranthiflorum Sleumer and R. luraluense
Sleumer are also found in New Guinea, but are not endemics,
their distribution extending eastwards to the Solomon Islands
(Appendix S1kk & mm).
Philippines
Six of the mapped species are endemic to the islands of the
Philippines (Appendix S1g, dd, ww, ddd, mmm & sss):
R. apoanum Stein, R. kochii Stein, R. quadrasinaum var.
rosmarinifolium (Vidal) Copel.f., R. rousei Argent & Madulid, R. taxifolium Merr. and R. williamsii Merr. Rhododendron apoanum is restricted to the southern island of
Mindanao (Appendix S1g), while four species, R. quadrasinaum var. rosmarinifolium, R. rousei, R. taxifolium and
R. williamsii, are found only on the islands in the north:
Luzon, Mindoro, Biliran and Sibuyan (Appendix S1ww, ddd,
mmm & sss). Rhododendron kochii is distributed throughout
the northern and southern Philippines, being found on the
islands of Luzon, Mindanao, Mindoro and Negros (Appendix S1dd).
Two other species are mapped, although they are not
restricted to the Philippines (Appendix S1bb & vvv):
R. javanicum and R. zollingeri. In the Philippines, both
species are found on the island of Luzon, while R. javanicum
has also been recorded from the islands of Mindanao and
Palawan.
Sulawesi
Eight of the mapped species are endemic to the island of
Sulawesi (Appendix S1f, r, gg, rr, vv, xx, bbb & ppp):
R. alternans Sleumer, R. eymae Sleumer, R. lagunculicarpum
J.J.Sm., R. nanophyton Sleumer, R. pudorinum Sleumer,
R. radians J.J.Sm., R. rhodopus Sleumer and R. vanvuurenii
J.J.Sm. Two of these — R. eymae and R. nanophyton
(Appendix S1r & rr) — are restricted to G. Rantemario in
the south, the highest mountain on the island, reaching
c. 3400 m a.s.l. Rhododendron alternans, R. lagunculicarpum
and R. pudorinum are also restricted to the south of the
island (Appendix S1f, gg & vv), while R. rhodopus and
R. vanvuurenii occur in both southern and central Sulawesi
(Appendix S1bbb & ppp). Rhododendron radians, the other
Sulawesi endemic, is more widespread, and is found in
the southern, central and northern parts of the island
(Appendix S1xx).
Rhododendron javanicum, R. malayanum and R. zollingeri
(Appendix S1bb, oo & vvv) have also been collected in south
and central Sulawesi, although they are not endemic to the
island.
Journal of Biogeography 33, 1929–1944
ª 2006 The Authors. Journal compilation ª 2006 Blackwell Publishing Ltd
1933
G. K. Brown, G. Nelson and P. Y. Ladiges
Sumatra
Four species endemic to the island of Sumatra are mapped
(Appendix S1c, uu, yy & kkk): R. aequabile J.J.Sm., R. pubigermen J.J.Sm., R. rarilepidotum J.J.Sm. and R. sumatranum
Merr. Rhododendron rarilepidotum and R. sumatranum are
restricted to the northern part of the island (Appendix S1yy &
kkk), while R. pubigermen occurs in the north and south, on
the west coast (Appendix S1uu). The other species endemic to
Sumatra, R. aequabile (Appendix S1c), has been recorded only
from the mid-west coast to the southern tip of the island.
While not endemic to the island, R. citrinum, R. jasminiflorum var. heusseri, R. javanicum, R. malayanum and R. retusum
(Appendix S1l, aa, bb, oo & aaa) have also been collected from
Sumatra.
Locations outside Malesia
Two species are known from northern Queensland, Australia,
and both were mapped (Appendix S1jj & qqq): R. lochiae F.
Muell. and R. viriosum Craven. The distributions of these two
species do not overlap, with R. viriosum (Appendix S1qqq)
distributed in the Cook district, and R. lochiae (Appendix S1jj)
found further south on Mount Bartle Frère, Bell Peak North
and the Bellenden Ker Range.
Two vireyas from the Himalaya region are mapped:
Rhododendron santapaui Sastry et al. (Appendix S1ggg) restricted to Assam (India), and R. vaccinioides Hook.f. (Appendix S1ooo), widespread across Bhutan, Burma, eastern Tibet,
India and Nepal.
The other two species mapped and located outside Malesia
are: R. kawakamii Hayata (Appendix S1cc) from Taiwan and
R. euonymifolium H.Lév (Appendix S1q) from north Vietnam
and southern China (Guizhou and Guangxi provinces).
Area analysis
Figure 2 is the summary cladogram for sampled taxa within
Rhododendron subg. Rhododendron (node 0) based on DNA
sequence data (Brown et al., 2006). Two species (R. maddenii
and R. lindleyi, node 4) classified in section Rhododendron
were included in the molecular analyses, and in some trees
nested within section Vireya, although the consensus tree
placed them at the basal polytomous node. All other species in
the cladogram are from section Vireya. The clades at nodes 2
and 3 are subsect. Pseudovireya and the clade at node 1 is
termed informally ‘Euvireya’. The Euvireya group consists of
two main clades of geographic significance: node 5 includes
eastern Malesian species and node 6 western and middle
Malesian species. In each of these main clades, there are species
endemic to particular areas, with repeated patterns indicating
geographic paralogy.
Paralogy-free subtree analysis based on Fig. 2 produced 13
area subtrees (Fig. 3). The nodes of these subtrees were coded
as 23 characters in a data matrix (Table 1) with an all-zero
outgroup, which, when analysed in Hennig86 with exhaustive
1934
branch swapping (mb*, bb*), produced an overflow of trees
(7108+) of length 23, CI ¼ 1.00 and RI ¼ 1.00. Of these
equally parsimonious trees, the one with the minimal number
of resolved nodes (the ‘minimal tree’) is shown in Fig. 4. The
position of some areas is unresolved (due to widespread taxa
or lack of endemics). Hence, areas I, M, O, Q, R, S and T –
Lesser Sunda Islands, Palawan, Malay Peninsula, Taiwan, north
Vietnam/south China and the Himalayas – were deleted from
the data matrix shown in Table 1, resulting in a new, reduced
matrix of 20 characters for the fourteen remaining areas and an
all-zero outgroup (Table 2).
The analysis of this reduced data matrix with implicit
enumeration (ie*) resulted in 124 most parsimonious trees of
length 20, CI ¼ 1.00 and RI ¼ 1.00. The strict consensus tree
(length 20) shows the same area relationships as the minimal
tree for the first analysis (Fig. 4) that included all areas.
Node 1 of the area cladogram (Fig. 4) is a polytomy,
including the Himalayas (T) and branches leading to three
nodes. Node 2 relates Taiwan (R) to the area of north
Vietnam/south China (S). The position of these areas in the
polytomy is minimal; however, they can be moved up the tree
without altering its length (indicated by the dashed line,
Fig. 4). This is possible because of a lack of informative
characters relating the areas of Taiwan and north Vietnam/
south China to the rest; the only subtree to contain information regarding the relationships of these areas is subtree 13
(Fig. 3).
Node 3 relates the eastern areas, including Bismarck Archipelago and Solomon Islands (A), north-eastern Australia (C)
and all five New Guinea areas (B, D, E, F, G). North-eastern
Australia (C) and the New Guinea craton (D) are differentiated
from the Papuan Peninsular (B), Central New Guinea (E),
northern New Guinea (F), the Vogelkop Peninsula (G) and the
Bismarck Archipelago and Solomon Islands (A), which form a
group at node 4 (Fig. 4). These relationships are shown by
subtrees 1 to 6 (Fig. 3).
Node 5 relates the western and middle Malesian areas of
southern Moluccas, Lesser Sunda Islands, north and West
Sulawesi, southern Philippines, northern Philippines, Palawan,
Borneo, Java and Bali, Sumatra, and the Malay Peninsula
(areas H to Q). Subtrees 7 to 13 (Fig. 3) show the relationships
within this group (Fig. 4). The position of the Lesser Sunda
Islands (I), Palawan (M), Java and Bali (O) and the Malay
Peninsula (Q) is minimal; however, this group of areas can
move up the tree without affecting tree length because area
relationships are based solely on widespread taxa in the case of
Lesser Sunda Islands (I) and Palawan (M), or in the case of the
Malay Peninsula (Q) there is insufficient information to
resolve the area’s position. The subtrees include information
regarding the relationship of Java and Bali (O) to other western
Malesian areas, with Java and Bali (O) more closely related to
Sumatra (P) than to the areas of eastern Malesia (A–G; subtree
11, Fig. 3). However, the position of area O in the minimal tree
remains uncertain (shown as the dashed line in Fig. 4) because
its relationships to the other areas at node 5 are not
determined.
Journal of Biogeography 33, 1929–1944
ª 2006 The Authors. Journal compilation ª 2006 Blackwell Publishing Ltd
Historical biogeography of Vireya rhododendrons
87
9
91
10
87
11
86
12
98
13
73
14
74
5
‘Euvireya’
88
24
98
1
19
61
20
96
25
87
93
15
28
26
98
21
0
29
92
97
6
22
27
96
72
16
23
78
17
87
18
98
7
63
2
Pseudovireya
100
8
100
3
55
4
R. herzogii*
R. inundatum*
R. tuba
R. culminicola
R. carringtoniae
R. leucogigas
R. leptanthum
R. gracilentum
R. luraluense
R. loranthiflorum
R. spondylophyllum
R. inconspicuum
R. saxifragoides
R. christi
R. commonae
R. baenitzianum
R. rarum
R. goodenoughii
R. rubineiflorum
R. konori
R. laetum
R. zoelleri
R. lochiae
R. hyacinthosmum
R. maius
R. multinervium
R. solitarium
R. superbum
BE
E
B
BEFG
B
F
BEF
BE
A
AB
B
BEG
E
BDE
BE
EF
EF
B
DE
BEG
G
BEFG
C
B
BE
E
E
AE
R. ruttenii
R. radians
R. taxifolium
R. intranervatum
R. polyanthemum
R. sarcodes
R. burttii
R. alborugosum
R. jasminiflorum
R. edanoi
R. vanvuurenii
R. pudorinum
R. alternans
R. rhodopus*
R. zollingeri
R. lagunculicarpum
R. williamsii
R. robinsonii
R. javanicum
R. rarilepidotum
R. aequabile
R. sumatranum*
R. album
R. citrinum
R. acuminatum
R. malayanum*
H
J
L
N
N
K
N
N
PQ
N
J
J
J
J
IJLO
J
L
Q
JKLMNOPQ
P
P
P
O
OP
N
HJNOPQ
R. meliphagidum*
R. nanophyton
R. ericoides
R. quadrasianum
R. retusum
R. kawakamii
R. euonymifolium
R. vaccinioides
R. santapaui
R. maddenii*
R. lindleyi
H
J
N
L
OP
R
S
T
T
S
T
Section Rhododendron
Figure 2 Taxon area cladogram. The summary phylogeny used for the paralogy-free subtree analysis is shown. Bootstrap values > 50% are
shown above the node they support. Node numbers are shown beneath the node. * indicates that a species was only sequenced for the trnTtrnL intergenic spacer region due to amplification difficulties for the psbA-trnH spacer. Letters to the right of the species name correspond to
the areas of endemism in Fig. 1. Areas: Bismarck Archipelago and the Solomon Islands (A), Papuan Peninsular (B), north-eastern Australia
(C), New Guinea craton (D), central New Guinea (E), northern New Guinea (F), Vogelkop Peninsula (G), south Moluccas (H), Lesser
Sunda Islands (I), north and west Sulawesi (J), southern Philippines (K), northern Philippines (L), Palawan (M), Borneo (N), Java and Bali
(O), Sumatra (P), Malay Peninsula (Q), Taiwan (R), north Vietnam and south China (S), and the Himalayas (T).
Journal of Biogeography 33, 1929–1944
ª 2006 The Authors. Journal compilation ª 2006 Blackwell Publishing Ltd
1935
G. K. Brown, G. Nelson and P. Y. Ladiges
Subtree 1:
1
Subtree 7:
D
5
1
L
19
N
C
9
B
HIJKLMNOPQ
ABCDEFG
D
5
Subtree 8:
1
K
20
C
25
10
N
EFG
ABCDEFG
Subtree 9:
Subtree 3:
D
5
1
29
ABCDEFG
B
F
HIJKLMNOPQ
Subtree 10:
1
27
ABCDEFG
D
5
C
12
F
Subtree 11:
1
23
O
HIJKLMNOPQ
ABCDEFG
Subtree 12:
D
5
13
1
17
HJOPQ
A
ABCDEFG
HIJKLMNOPQ
Subtree 13:
2
OP
7
Subtree 6:
18
5
14
H
D
J
C
N
B
L
EG
8
HIJKLMNOPQ
R
S
The relationships and order of differentiation of areas
within the clade at node 5, which can be discussed with
confidence, are as follows (Fig. 5). The earliest geographic
segregation within node 5 relates the southern Philippines
(K) to the rest of the western and middle Malesian areas
(areas H, J, L, N, P, Q; Fig. 5c). The area of Sumatra (P) is
the next area to be differentiated in the western Malesian
region (Fig. 5d). The relationship between the northern
Philippines (L) and Borneo (N) is unresolved (node 7) in
relation to the southern Moluccas (H) and north and west
Sulawesi (J; node 8, Fig. 4), the latter two (H & J) inferred
to be sister areas.
1936
N
C
B
1
P
BE
Subtree 5:
1
Q
JKLMNOP
Subtree 4:
1
ILO
J
C
11
PQ
B
HIJKLMNOPQ
1
H
J
Subtree 2:
1
24
E
Figure 3 Area subtrees. Thirteen area subtrees identified by the paralogy-free subtree
analysis are shown. The node numbers
(indicated below the node) relate to the
nodes of the taxon area cladogram (Fig. 2).
Areas: Bismarck Archipelago and the Solomon Islands (A), Papuan Peninsular (B),
north-eastern Australia (C), New Guinea
craton (D), central New Guinea (E), northern
New Guinea (F), Vogelkop (G), south
Moluccas (H), Lesser Sunda Islands (I),
north and west Sulawesi (J), southern
Philippines (K), northern Philippines (L),
Palawan (M), Borneo (N), Java and Bali (O),
Sumatra (P), Malay Peninsula (Q), Taiwan
(R), north Vietnam and south China (S) and
the Himalayas (T).
DISCUSSION
Cladistic biogeography endeavours to understand the relationships of geographic areas based on the biotic patterns presented
in phylogenetic trees (Humphries & Parenti, 1999). It assumes
that the phylogenies of taxa, and the nodes within it, contain
information regarding the history of the areas that they inhabit
(Ladiges, 1998; Ebach & Edgecombe, 2001). The history of areas
elucidated via cladistic biogeography can be the result of
vicariant or dispersal events. Nevertheless, evidence from other
sources such as geological or climatological studies is required to
infer which type of event has occurred.
Journal of Biogeography 33, 1929–1944
ª 2006 The Authors. Journal compilation ª 2006 Blackwell Publishing Ltd
Historical biogeography of Vireya rhododendrons
Table 1 Nodes of area subtrees coded as a character matrix. Areas not represented in a subtree are coded by a question mark. The tree is
rooted by an all zero outgroup
Area
Characters
Outgroup
Bismarck Archipelago and
Solomon Islands (A)
Papuan Peninsular (B)
North-eastern Australia (C)
New Guinea craton (D)
Central New Guinea (E)
Northern New Guinea (F)
Vogelkop Peninsula (G)
Southern Moluccas (H)
Lesser Sunda Islands (I)
North and West Sulawesi (J)
Southern Philippines (K)
Northern Philippines (L)
Palawan (M)
Borneo (N)
Java and Bali (O)
Sumatra (P)
Malay Peninsula (Q)
Taiwan (R)
North Vietnam/south China (S)
Himalayas (T)
0
?
0
?
0
?
0
?
0
?
0
?
0
?
0
?
0
1
0
1
0
?
0
?
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
?
0
?
0
?
1
1
1
1
?
?
0
0
0
0
0
0
0
0
0
0
?
?
?
1
0
0
1
?
?
0
0
0
0
0
0
0
0
0
0
?
?
?
1
1
1
1
1
1
0
0
0
0
0
0
0
0
0
0
?
?
?
1
0
0
1
1
1
0
0
0
0
0
0
0
0
0
0
?
?
?
1
1
1
?
1
?
0
0
0
0
0
0
0
0
0
0
?
?
?
1
0
0
?
1
?
0
0
0
0
0
0
0
0
0
0
?
?
?
1
1
1
1
1
?
0
0
0
0
0
0
0
0
0
0
?
?
?
1
0
0
1
1
?
0
0
0
0
0
0
0
0
0
0
?
?
?
1
1
1
?
?
?
0
0
0
0
0
0
0
0
0
0
?
?
?
1
0
0
?
?
?
0
0
0
0
0
0
0
0
0
0
?
?
?
1
1
1
1
?
1
0
0
0
0
0
0
0
0
0
0
?
?
?
1
0
0
1
?
1
0
0
0
0
0
0
0
0
0
0
?
?
?
0
0
0
0
0
0
1
?
1
?
1
?
1
?
?
?
?
?
?
0
0
0
0
0
0
1
?
1
?
0
?
0
?
?
?
?
?
?
0
0
0
0
0
0
?
?
?
1
?
?
1
?
1
1
?
?
?
0
0
0
0
0
0
?
?
?
0
?
?
1
?
1
1
?
?
?
0
0
0
0
0
0
?
1
1
?
1
?
?
1
?
?
?
?
?
0
0
0
0
0
0
?
?
1
1
1
1
1
1
1
1
?
?
?
0
0
0
0
0
0
?
?
?
?
?
?
?
1
1
?
?
?
?
0
0
0
0
0
0
1
?
1
?
?
?
1
1
1
1
?
?
?
?
?
?
?
?
?
1
?
1
?
1
?
1
1
1
?
0
0
?
?
?
?
?
?
?
1
?
1
?
1
?
1
0
0
?
0
0
?
?
?
?
?
?
?
0
?
0
?
0
?
0
0
0
?
1
1
?
Geological history of Malesia
The geological history of Malesia is complex, and somewhat
controversial, and has been discussed in considerable detail in
the literature (e.g. Audley-Charles, 1987; Michaux, 1991; de
Boer & Duffels, 1996; Hall, 2001; Metcalfe, 2001; Morley, 2001;
Hall, 2002). A detailed plate tectonic reconstruction of southeast Asia and the south west Pacific in the Cenozoic (last
60 Myr) was presented most recently by Hall (2002). It
includes a comprehensive outline of the geological history of
Malesia, based on present and past plate motions, seismic and
volcanic activity, and palaeomagnetic and isotopic data.
The Malesian Archipelago is bounded by three major plates:
the Pacific to the east, the Indo-Australian to the west and
south, and the Indo-China to the north (Michaux, 1991). The
archipelago formed as a result of Australia and India rifting
away from Gondwana, moving northward and colliding with
the Asian plate. India rifted from Australia and Antarctica c.
130 Ma, and collided with the Asian continent c. 50 Ma
(McLoughlin, 2001; Hall, 2002). Australia later rifted from
eastern Antarctica (rifting initiated at c. 96 Ma but was not
complete until c. 35 Ma; McLoughlin, 2001), and moved
northwards into the Tethys Ocean colliding with the Asian
margin later than India (Audley-Charles, 1981). Both Australia
and India continue to move northwards today (Hall, 2002).
As Australia moved north, slivers of continental crust were
occasionally sliced off as microcontinents; these also moved
north ahead of the main continental margin. The Bird’s Head
of New Guinea is one such microcontinent, although the
location from where it was sliced remains in dispute; Hall
(2002) infers that it is of north-western Australian origin, while
Pigram & Panggabean (1984 in de Boer & Duffels, 1996)
believe it to be of east Queensland or central New Guinea
origin. Several of the islands, or island groups — Moluccas,
New Guinea, Philippines and Sulawesi — are considered to be
of composite origin (Audley-Charles, 1981; Hall, 2002). In
addition to microcontinents, areas comprise also of components that are remnants of volcanic island arcs — central
Philippines; north, central and south-eastern New Guinea; and
the Bismarck Archipelago (de Boer & Duffels, 1996) — or are
part of the continental shelf, e.g. Borneo, Java, Malay
Peninsula, southern New Guinea, Palawan and Sumatra (Hall,
2002). The western Malesian islands of the Sunda shelf are also
known as Sundaland.
Environmental changes, including changes in sea-level,
degree of seasonality of precipitation or temperature, all
influence biogeographic patterns. Climatic conditions can also
be affected by geological events; for example, Hall (2002)
suggests that orogeny in northern New Guinea during the
Pliocene–Pleistocene, is likely to have contributed to increased
aridity in Australia and increased rainfall in New Guinea. Some
studies have investigated the past climate of the Malay
Archipelago, with many focusing on Quaternary changes
(e.g. Whitmore, 1981a; Morley, 1982; Morley & Flenley, 1987;
Newsome & Flenley, 1988; van der Kaars, 1991; UrushibaraYoshino & Yoshino, 1997).
Morley & Flenley (1987) investigated Neogene and Quaternary environmental changes in the Malesian Archipelago
Journal of Biogeography 33, 1929–1944
ª 2006 The Authors. Journal compilation ª 2006 Blackwell Publishing Ltd
1937
G. K. Brown, G. Nelson and P. Y. Ladiges
Figure 4 Minimal Tree. Biogeographic area
relationships as shown in the minimally resolved most parsimonious tree (length 23,
CI ¼ 1.00, RI ¼ 1.00). Dashed lines indicate
areas that are in their minimal position but
have the ability to move up the tree without
altering the length of the tree. Analysis of the
data excluding these areas resulted in a strict
consensus tree (length 20, CI ¼ 1.00,
RI ¼ 1.00) showing the same relationships
(solid lines). Node numbers (italics) are
shown above the node they relate to, while
numbers below the node (grey) indicate the
nodes from the taxon area cladogram (Fig. 2)
that support each biogeographic node. Areas:
Bismarck Archipelago and the Solomon
Islands (A), Papuan Peninsular (B), northeastern Australia (C), New Guinea craton
(D), central New Guinea (E), northern New
Guinea (F), Vogelkop Peninsula (G), south
Moluccas (H), Lesser Sunda Islands (I),
north and west Sulawesi (J), southern
Philippines (K), northern Philippines (L),
Palawan (M), Borneo (N), Java and Bali (O),
Sumatra (P), Malay Peninsula (Q), Taiwan
(R), north Vietnam and south China (S) and
the Himalayas (T).
Table 2 Nodes of area subtrees coded as a character matrix with equivocal areas deleted. Areas not represented in a subtree are coded by a
question mark. The tree is rooted by an all zero outgroup
Areas
Characters
Outgroup
Bismarck Archipelago and
Solomon Islands (A)
Papuan Peninsular (B)
North-eastern Australia (C)
New Guinea craton (D)
Central New Guinea (E)
Northern New Guinea (F)
Vogelkop Peninsula (G)
Southern Moluccas (H)
North and West Sulawesi (J)
Southern Philippines (K)
Northern Philippines (L)
Borneo (N)
Sumatra (P)
0
?
0
?
0
?
0
?
0
?
0
?
0
?
0
?
0
1
0
1
0
?
0
?
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
?
1
1
1
1
?
?
0
0
0
0
0
0
1
0
0
1
?
?
0
0
0
0
0
0
1
1
1
1
1
1
0
0
0
0
0
0
1
0
0
1
1
1
0
0
0
0
0
0
1
1
1
?
1
?
0
0
0
0
0
0
1
0
0
?
1
?
0
0
0
0
0
0
1
1
1
1
1
?
0
0
0
0
0
0
1
0
0
1
1
?
0
0
0
0
0
0
1
1
1
?
?
?
0
0
0
0
0
0
1
0
0
?
?
?
0
0
0
0
0
0
1
1
1
1
?
1
0
0
0
0
0
0
1
0
0
1
?
1
0
0
0
0
0
0
0
0
0
0
0
0
1
1
?
1
1
?
0
0
0
0
0
0
1
1
?
0
0
?
0
0
0
0
0
0
?
?
1
?
1
1
0
0
0
0
0
0
?
?
0
?
1
1
0
0
0
0
0
0
?
1
?
1
?
?
0
0
0
0
0
0
?
1
1
1
1
1
0
0
0
0
0
0
1
1
?
?
1
1
?
?
?
?
?
?
1
1
?
1
1
0
and found that glacial periods were prominent during the
Pleistocene, and therefore land connections between mainland
Asia and Sundaland are probable throughout the last 2 Myr.
1938
Geological data confirm this, although Hall (2002) suggests
that the region between the Malay Peninsula and western
Borneo has been elevated throughout the Cenozoic. These
Journal of Biogeography 33, 1929–1944
ª 2006 The Authors. Journal compilation ª 2006 Blackwell Publishing Ltd
Historical biogeography of Vireya rhododendrons
Figure 5 Relative timing of the differentiation of biogeographic areas. The order of differentiation is based on the nodes of the minimally
resolved tree (Fig. 4); grey areas are those areas that are differentiating at the same time. After an area has been distinguished it is shown as
an open black symbol; stippled lines show those areas that are differentiated but can move position in the minimal tree without altering the
tree length. Letters corresponding to the areas of endemism (Fig. 1) are indicated. Areas: Bismarck Archipelago and Solomon Islands (A),
Papuan Peninsular (B), north-eastern Australia (C), New Guinea craton (D), Central New Guinea (E), northern New Guinea (F), Vogelkop
Peninsula (G), south Moluccas (H), Lesser Sunda Islands (I), north and West Sulawesi (J), southern Philippines (K), northern Philippines
(L), Palawan (M), Borneo (N), Java and Bali (O), Sumatra (P), Malay Peninsula (Q), Taiwan (R), north Vietnam and south China (S)
and the Himalayas (T). (a) The four clades of the polytomy (node 1) – Himalayas (T), Taiwan and Vietnam and south China (R&S), western
and middle Malesia (H–Q), and eastern Malesia (A–G) – differentiated from each other. (b) Taiwan (R) and Vietnam and south China (S)
differentiate from each other, as do the areas A, B, E–G from areas C and D in eastern Malesia (nodes 2 and 3). (c) Southern Philippines (K)
differentiates first from the other western and middle Malesian areas; areas I, M and O (Lesser Sunda Islands, Palawan and Java & Bali)
are indicated as differentiating at the same time as K, although timing of their differentiation may be later (node 5). (d) Sumatra (P)
completes differentiation from the main western Malesian area; the Malay Peninsula is also indicated to differentiate at this time,
although timing of its differentiation may be later (node 6). The areas Moluccas (H), Sulawesi (J), Borneo (N) and northern Philippines (L)
are not yet differentiated from each other. (e) Borneo (N) and northern Philippines (L) differentiate (node 7); Sulawesi (J) and Moluccas
(H), and all the eastern areas (A–G) are yet to differentiate; the timing and order of these remaining events are unknown. (f) All areas of
endemism have completed differentiation (nodes 4 and 8), as defined in Fig. 1.
Journal of Biogeography 33, 1929–1944
ª 2006 The Authors. Journal compilation ª 2006 Blackwell Publishing Ltd
1939
G. K. Brown, G. Nelson and P. Y. Ladiges
glacial periods resulted in fluctuations between wet and dry
conditions. During the last glacial period conditions in the
Malesian region and northern Australia were thought to have
been about 2 °C cooler and drier than the present (Whitmore,
1981a; Morley, 1982; van der Kaars, 1991). Such conditions
would have allowed the montane forests, such as those
inhabited by species of Rhododendron, to occupy a greater
area than they do today, with the present-day atypical
conditions increasing both sea level and the forest limits
(Whitmore, 1981a).
More recently, the vegetation in the region has been
influenced by human activities, with forest clearing evident
over the last 8000 to 7000 years to the present-day (Morley,
1982; Newsome & Flenley, 1988; Haberle et al., 1991).
Interpreting the Vireya area pattern
The most striking result from the Vireya analysis is the
identification of major clades corresponding to eastern Malesia
and western/middle Malesia. The close relationship found
between Australia and New Guinea for Vireya is repeated for
many other groups, such as cicadas, butterflies, tea trees,
eucalypts and genera of Proteaceae (Holloway, 1987; Crisp
et al., 1995; de Boer & Duffels, 1996; Brown et al., 2001;
Ladiges et al., 2003).
In contrast to this striking overall result, relationships
between areas within the two major regions are unclear. Area
relationships within New Guinea remain unresolved based on
the vireyas, there being no consensus among other studies (de
Boer & Duffels, 1996; van Welzen et al., 2001), particularly
regarding the relationship of the Vogelkop Peninsula. The
relationships between the areas in western and middle Malesia
are also inconsistent across a variety of organisms, with almost
every combination of sister areas being reported (Schuh &
Stonedahl, 1986; Holloway, 1987; van Welzen, 1992; Ruedi,
1996; Repetur et al., 1997; Ridder-Numan, 1998). The results
presented here do little to resolve the problem (Fig. 4).
The relationships of sister areas Taiwan and north Vietnam/
south China is uncertain, being placed in the polytomy at node
1 of the area cladogram (Fig. 4), although they may be related
to the Himalayas (Croizat, 1968; Ridder-Numan, 1998;
Denduangboripant et al., 2001; A.L. Denton & B.D. Hall,
unpubl. data). Croizat (1968) suggested that a classic
biogeographic role of Formosa (Taiwan) is as an ‘appendage
to the Sino-Himalayan Domain’, and that it is part of the
track: Nepal–north Burma and Thailand/south China–
Formosa–Luzon.
Determining the age of the clades and the underlying
processes that have led to the evolution of section Vireya is
difficult. Nodes of the phylogeny are not dated using molecular
dating algorithms because of a lack of precise calibration
points for these data and evidence of variable DNA substitution rates. Based on cpDNA Milne (2004) attempted to
estimate the divergence age of Rhododendron subsection
Pontica, ‘a group with a tertiary relict distribution’. He found
that only his synonymous matK mutation data set met the
1940
assumption of a molecular clock, with other regions violating
assumptions.
Geological data, biogeographic patterns and fossils are used
to argue two contrasting hypotheses.
Hypothesis 1: Vireyas are a Gondwanan group
Section Vireya is an old group, with ancestors present on
Gondwana before India rifted north in the Cretaceous
(130 Ma). As the islands of Malesia formed and moved into
their present position, vireya rhododendrons dispersed further
into Malesia; hence today both areas (Himalayas and Australia) are related to Malesia. Ancestral taxa that rifted north on
India were probably similar to section Rhododendron or were
Pseudovireya-like, while the ‘Euvireya’-like vireyas were most
likely of Australian rain forest ancestry.
Based on geological evidence, node 6 (Figs 4 & 5d) —
Sumatra, Malay Peninsula, northern Philippines, Borneo,
south Moluccas and north and west Sulawesi — may be at
least 60 Ma because the region between the Malay Peninsula
and western Borneo has been hypothesized to have been
elevated throughout the Cenozoic (Hall, 2002). Therefore,
nodes 1, 3 and 5 would be at least that age, but arguably older.
Similar scenarios have been suggested for the establishment
of ‘primitive’ angiosperms in tropical Asia–Australasia (Morley, 2001) and for several other plant groups, including genera
of Proteaceae (Whitmore, 1981b), Nastus Juss. (Whitmore,
1981b), and a number of palms (Dransfield, 1981). Specht
(1988) hypothesized that ancestral Rhododendron taxa were
present on Gondwana before its break up. He supposed that
the ancestor of the lepidote rhododendrons (subgenus Rhododendron) remained on the Australian plate and expanded from
the northern part into Malesia and then mainland Asia c.
15 Ma, after the Australia plate made contact with Sundaland.
Specht also considered that the ancestor of the elepidote
rhododendrons (all subgenera except subgenus Rhododendron)
was on the Indian plate as it rifted northwards, and then
spread to the north and east, after India collided with the Asian
mainland, where it diversified and reached its present-day
distribution.
What does the fossil record tell us of the minimum age of
rhododerndrons? The oldest macrofossil assigned to the genus
Rhododendron dates from the Paleocene (65–55 Ma) with fossil
leaves and seeds found in north-west Greenland and at Ash
Quarry in Newbury, England (Collinson & Crane, 1978). The
oldest fossil flower assigned to the Ericales dates from the
Turonian (c. 90 Ma), in the mid-Cretaceous (Nixon & Crepet,
1993). Rhododendrons were more abundant throughout the
Eocene and Miocene, with more, widespread records of
Rhododendron fossils recorded for this period: pollen found
in Germany, and leaves in Alaska, Austria (Köflach, Stoob and
Hausruck), USA (Vermont), Japan and eastern China
(Zhejiang; Collinson & Crane, 1978; Zetter & Hesse, 1996).
No fossils of Rhododendron have been reported in Australia,
although rain forest habitat, including those areas of
north-eastern Australia where the two Australian species
Journal of Biogeography 33, 1929–1944
ª 2006 The Authors. Journal compilation ª 2006 Blackwell Publishing Ltd
Historical biogeography of Vireya rhododendrons
Rhododendron lochiae and R. viriosum occur today, was
originally established in the Cretaceous, and are thought to
have remained largely unaltered in terms of genera present
since then (Barlow & Hyland, 1988; Specht, 1988). Relictual
angiosperms have survived to the present day in these rain
forests because of a combination of processes: these areas were
not exposed to aridity as much of the continent was, and as a
result of Australia drifting northwards, rain forest patches were
generally not affected by the cooling of the climate throughout
the Neogene (Morley, 2001). Tropical rain forests were also
present on India as it drifted northwards in the early Eocene
(Morley, 2001) with environments suitable for Rhododendron.
These ancestral tropical rain forests of north-eastern
Australia most likely extended north to New Guinea, on the
Australian craton, as argued for the sister relationship of
Eucalyptopsis C.T. White and Stockwellia D.J. Carr, S.G.M.
Carr & B. Hyland in the eucalypt group (Ladiges et al., 2003).
Similar patterns — Atherton related to New Guinea — have
also been found in other Myrtaceae genera (e.g. Thaleropia
Peter G. Wilson, Wilson, 1993) and in the Proteaceae
(Telopea Sol. ex Baill. and Alloxylon P.H. Weston and Crisp,
Crisp et al., 1995). Barlow & Hyland (1988) also hypothesized
that ‘Tertiary New Guinea’ was part of the Australian refugial
area.
An old Gondwanic element in tropical floras has been
supported by various authors. Barlow & Hyland (1988)
considered that strong taxonomic links existed between the
tropical floras of Africa, Asia and Australia, and Croizat (1968)
considered that the Pacific ‘massings’ of taxa, such as the
Columba L. pigeons and Ficus drupacea Thunb., were old (at
least early Cretaceous in age) and stemmed from ancestors
originally bound with the Indian Ocean, India, Australia and
Africa, therefore suggesting Gondwana. Heads (2003) presents
a detailed argument for Mesozoic and Cenozoic events shaping
the biogeographic differentiation of vireyas and other taxa in
Malesia. Heads asserts that the simplest explanation for highly
speciose groups such as the rhododendrons, is that ‘there is no
point centre of origin and the groups have developed by
vicariance in immobilism of an already global ancestral
complex’ (p. 435, Heads, 2003).
Hypothesis 2: Vireyas are young
Section Vireya is a young group, which dispersed eastwards
from India to Australia and the Solomon Islands since the
islands of Malesia were in, or close to, their present-day
positions. Two main radiation-dispersal events into Malesia
are likely to have occurred, representing the two lineages
resolved in the phylogeny, Pseudovireya and ‘Euvireya’ (Brown
et al., 2006, in press). There was only one eastward radiation,
into New Guinea, Australia and the Solomon Islands, followed
by mass speciation leading to the present-day diversification,
possibly related to recent orogenic events in New Guinea,
opening up a myriad of new niches and habitats for the vireyas
to exploit. Although vireyas are morphologically diverse, there
is little variation in the DNA regions between related taxa,
particularly those within the eastern Malesian clade whose
relationships are not well resolved. Weak breeding barriers
exist between many Vireya taxa and most subsections of
section Vireya (Williams & Rouse, 1997). This promiscuity was
taken by Williams Rouse to be an indicator of a young group.
Based on an estimated age of Ericaceae on a calibrated
Angiosperm phylogeny and fossil data (Magallón et al., 1999;
Wikström et al., 2001), Milne (2004) concludes that the
subgenus Rhododendron (represented by only one species in his
analysis) is 46–32 Ma. That result infers that section Vireya is
younger, although note our earlier reservations of this analysis.
If hypothesis 2 is correct, then based on geological evidence
node 5 (Fig. 4) could be at least 60 Ma, but node 3 (Fig. 4)
would only be c. 10 Ma, when the islands were close enough
for vireyas to island-hop to their present distribution (Hall,
2002). Vireyas would have had to have dispersed over many
hundreds of kilometres of unsuitable habitat unless continuous, or almost continuous, corridors existed. Barlow & Hyland
(1988) argued that throughout the late Quaternary during
climate and vegetation fluctuations, pockets of rain forest were
close enough to allow exchange of flora between the forests of
New Guinea and Australia. The small, light weight seeds of
Vireya could be dispersed by wind, but the distances that they
can travel and remain viable are yet to be determined.
A similar hypothesis has been put forward for genera of the
Dipterocarpaceae (Whitmore, 1981b), and even though some
of these genera, such as Vatica L. and Hopea Roxb., extend to
New Guinea, none extends into Australia or the Solomons, as
do vireya rhododendrons.
CONCLUSIONS
The molecular phylogenetic analysis of the vireya rhododendrons has revealed a major clade divergence that correlates
with a distinct biogeographic pattern: one major clade
restricted to the east of Wallace’s line and another to the
west. Based on geographic pattern, presence of taxa in relictual
rain forests that include ancient angiosperms (e.g. north-east
Australia), and fossil minimal ages, it can be argued that the
vireyas are an old Gondwanan group. The alternative hypothesis that the group is young relies on accepting that low
molecular distances between taxa within clades reflects a young
age, which in turn requires long-distance dispersal to explain
distribution patterns. It may be that deep divergences within
the vireyas have an old history but diversification within clades
is more recent.
ACKNOWLEDGEMENTS
We would like to thank the following people and organizations: the curators of the following herbaria for access to
Rhododendron collection material A, BO, CANB, E, L, MEL
and NY; The Baker Foundation, The University of Melbourne
(MRS) and CSIRO for funding; Mr Lyn Craven and Dr Frank
Udovicic for their support; The Centre for Plant Biodiversity
Research (CSIRO Plant Industry) and the School of Botany,
Journal of Biogeography 33, 1929–1944
ª 2006 The Authors. Journal compilation ª 2006 Blackwell Publishing Ltd
1941
G. K. Brown, G. Nelson and P. Y. Ladiges
The University of Melbourne, for access to research equipment
and facilities to complete the project; Mr Stephen Teo, Mr Ken
Cox and Pak Ramadhanil for help tracking down locality
information; Dr Rogier de Kok for his help in translating the
Dutch labels; Dr David Frodin for assistance locating the
authority of Ficus drupacea; and Dr Michael Heads for
comments on an early draft of the manuscript. Sequencing
for this project was conducted at the Australian National
Herbarium, CSIRO Plant Industry, Canberra.
REFERENCES
Argent, G. & Madulid, D. (1995) Rhododendron sarcodes: a new
species from the Philippines. The New Plantsman, 2, 156–
161.
Argent, G. & Madulid, D. (1998) Rhododendron rousei (Ericaceae): a beautiful new species from the Philippines. The
New Plantsman, 5, 25–31.
Audley-Charles, M.G. (1981) Geological history of the region
of Wallace’s line. Wallace’s line and plate tectonics (ed. by
T.C. Whitmore), pp. 24–35. Clarendon Press, Oxford.
Audley-Charles, M.G. (1987) Dispersal of Gondwanaland:
relevance to evolution of the angiosperms. Biogeographical
evolution of the Malay Archipelago (ed. by T.C. Whitmore),
pp. 5–25. Clarendon Press, Oxford.
Axelius, B. (1991) Areas of distribution and areas of endemism. Cladistics, 7, 197–199.
van Balgooy, M.M.J. (1987) A plant geographical analysis of
Sulawesi. Biogeographical evolution of the Malay Archipelago
(ed. by T.C. Whitmore), pp. 94–102. Clarendon Press,
Oxford.
Barlow, B.A. & Hyland, B.P.M. (1988) The origins of the flora
of Australia’s wet tropics. Proceedings of the Ecological Society
of Australia, 15, 1–17.
Beaman, T.E., Wood, J.J., Beaman, R.S. & Beaman, J.H. (2001)
Orchids of Sarawak. Natural History Publications, Kota
Kinabalu, Sabah.
de Boer, A.J. & Duffels, J.P. (1996) Historical biogeography of
the cicadas of Wallacea, New Guinea and the West Pacific: a
geotectonic explanation. Palaeogeography, Palaeoclimatology,
Palaeoecology, 124, 153–177.
Brown, G.K., Udovicic, F. & Ladiges, P.Y. (2001) Molecular
phylogeny and biogeography of Melaleuca, Callistemon and
related genera (Myrtaceae). Australian Systematic Botany,
14, 565–585.
Brown, G.K., Craven, L.A., Udovicic, F. & Ladiges, P.Y. (2006)
Phylogeny of Rhododendron section Vireya (Ericaceae) based
on evidence from two non-coding regions of cpDNA. Plant
Systematics and Evolution, 257, 57–93.
Brown, G.K., Craven, L.A., Udovicic, F. & Ladiges, P.Y. (in
press) Phylogenetic relationships of Rhododendron section
Vireya (Ericaceae) inferred from the ITS nrDNA region.
Australian Systematic Botany.
Chamberlain, D.F., Hyam, R., Argent, G., Fairweather, G. &
Walter, K.S. (1996) The genus Rhododendron: its classification
and synonymy. Royal Botanic Garden Edinburgh, Edinburgh.
1942
Collinson, M.E. & Crane, P.R. (1978) Rhododendron seed from
the Palaeocene of southern England. Botanical Journal of the
Linnean Society, 76, 195–205.
Copeland, H.F. (1929) Philippine Ericaceae I: the species of
Rhododendron. The Philippine Journal of Science, 40, 133–179.
Cox, C.B. (2001) The biogeographic regions reconsidered.
Journal of Biogeography, 28, 511–523.
Crisci, J.V. (2001) The voice of historical biogeography.
Journal of Biogeography, 28, 157–168.
Crisp, M.D., Linder, H.P. & Weston, P.H. (1995) Cladistic
biogeography of plants in Australia and New Guinea:
congruent patterns reveals two endemic tropical tracks.
Systematic Biology, 44, 457–473.
Crisp, M.D., West, J.G. & Linder, H.P. (1999) Biogeography of
the terrestrial flora. Flora of Australia (ed. by E. Orchard),
pp. 321–367. ABRS/CSIRO, Collingwood.
Croizat, L. (1968) The biogeography of the tropical lands and
islands east of Suez-Madagascar: with particular reference to
the dispersal and form-making of Ficus L., and different other
vegetal and animal groups. Atti dell’ Istituto Botanico Laboratorio Crittogamico dell’ Università di Pavia, Serie 6, 4, 1–400.
Denduangboripant, J., Mendum, M. & Cronk, Q.C.B. (2001)
Evolution in Aeschynanthus (Gesneriaceae) inferred from ITS
sequences. Plant Systematics and Evolution, 228, 181–197.
Dransfield, J. (1981) Palms and Wallace’s line. Wallace’s Line
and plate tectonics (ed. by T.C. Whitmore), pp. 43–56.
Clarendon Press, Oxford.
Ebach, M.C. & Edgecombe, G.D. (2001) Cladistic biogeography: component-based methods and paleontological
application. Fossils, phylogeny, and form – an analytical
approach (Topics in geobiology) (ed. by J.M. Adrian, G.D.
Edgecombe and B.S. Lieberman), pp. 235–289. Kluwer
Academic/Plenum Publishers, New York.
Evans, B.J., Brown, R.M., McGuire, J.A., Supriatna, J.,
Andayani, N., Diesmos, A., Iskandar, D., Melnick, D.J. &
Cannatella, D.C. (2003) Phylogenetics of fanged frogs:
testing biogeographical hypotheses at the interface of the
Asian and Australian faunal zones. Systematic Biology, 52,
794–819.
Farris, J.S. (1988) Hennig86, Ver. 1.5. Distributed by the
author, Port Jefferson Station, New York.
George, W. (1981) Wallace and his line. Wallace’s Line and
plate tectonics (ed. by T.C. Whitmore), pp. 3–8. Clarendon
Press, Oxford.
Haberle, S.G., Hope, G.S. & de Fretes, Y. (1991) Environmental
change in the Baliem Valley, montane Irian Jaya, Republic of
Indonesia. Journal of Biogeography, 18, 25–40.
Hall, R. (2001) Cenozoic reconstructions of SE Asia and the
SW Pacific: changing patterns of land and sea. Faunal and
floral migrations and evolution in SE Asia-Australasia (ed. by
I. Metcalfe, J.M.B. Smith, M. Morwood and I. Davidson),
pp. 35–56. A.A. Balkema Publishers, Lisse.
Hall, R. (2002) Cenozoic geological and plate tectonic evolution of SE Asia and the SW Pacific: computer-based
reconstructions, model and animations. Journal of Asian
Earth Sciences, 20, 353–431.
Journal of Biogeography 33, 1929–1944
ª 2006 The Authors. Journal compilation ª 2006 Blackwell Publishing Ltd
Historical biogeography of Vireya rhododendrons
Hall, R. & Holloway, J.D. (1998) Biogeography and geological
evolution of SE Asia. Backhuys Publishers, Leiden, The
Netherlands.
Harold, A.S. & Mooi, R.D. (1994) Areas of endemism: definition
and recognition criteria. Systematic Biology, 43, 261–266.
Heads, M. (2001) Birds of paradise, biogeography and ecology in
New Guinea: a review. Journal of Biogeography, 28, 893–925.
Heads, M. (2003) Ericaceae in Malesia: vicariance biogeography,
terrane tectonics and ecology. Telopea, 10, 311–450.
Holloway, J.D. (1987) Lepidoptera patterns involving Sulawesi:
what do they indicate of past geography? Biogeographical
Evolution of the Malay Archipelago (ed. by T.C. Whitmore),
pp. 103–118. Clarendon Press, Oxford.
Humphries, C.J. & Parenti, L.P. (1986) Cladistic biogeography.
Clarendon Press, Oxford.
Humphries, C.J. & Parenti, L.P. (1999) Cladistic biogeography:
interpreting patterns of plant and animal distributions.
Oxford University Press, Oxford.
Irving, E. & Hebda, R. (1993) Concerning the origin and
distribution on Rhododendrons. Journal of the American
Rhododendron Society, 47, 139–146, 152–162.
van der Kaars, W.A. (1991) Palynology of eastern Indonesian
marine piston-cores: a Late Quaternary vegetational and
climatic record for Australasia. Palaeogeography, Palaeoclimatology, Palaeoecology, 85, 239–302.
Ladiges, P.Y. (1998) Biogeography after Burbidge. Australian
Systematic Botany, 11, 231–242.
Ladiges, P.Y., Humphries, C.J. & Martinelli, L.W. (1991)
Austral biogeography. CSIRO Australia, Melbourne.
Ladiges, P.Y., Udovicic, F. & Nelson, G. (2003) Australian
biogeographical connections and the phylogeny of large
genera in the plant family Myrtaceae. Journal of Biogeography, 30, 989–998.
Magallón, S., Crane, P.R. & Herendeen, P.S. (1999) Phylogenetic pattern, diversity and diversification of eudicots.
Annals of the Missouri Botanical Garden, 86, 297–372.
McLoughlin, S. (2001) The breakup history of Gondwana and
its impact on pre-Cenozoic floristic provincialism. Australian Journal of Botany, 49, 271–300.
Metcalfe, I. (2001) Palaeozoic and Mesozoic tectonic evolution
and biogeography of SE Asia-Australasia. Faunal and floral
migrations and evolution in SE Asia-Australasia (ed. by I.
Metcalfe, J.M.B. Smith, M. Morwood and I. Davidson), pp.
15–34. A.A. Balkema Publishers, Lisse.
Michaux, B. (1991) Distributional patterns and tectonic
development in Indonesia: Wallace reinterpreted. Australian
Systematic Botany, 4, 25–36.
Milne, R.I. (2004) Phylogeny and biogeography of Rhododendron subsection Pontica, a group with a tertiary relict distribution. Molecular Phylogenetics and Evolution, 33, 389–
401.
Morley, R.J. (1982) A palaeoecological interpretation of
10,000 year pollen record from Danau Padang, Central
Sumatra, Indonesia. Journal of Biogeography, 9, 151–190.
Morley, R.J. (2001) Why are there so many primitive angiosperms in the rain forests of Asia–Australasia? Faunal and
floral migrations and evolution in SE Asia–Australasia (ed. by
I. Metcalfe, J.M.B. Smith, M. Morwood and I. Davidson),
pp. 185–200. A.A. Balkema Publishers, Lisse.
Morley, R.J. & Flenley, J.R. (1987) Late Cainozoic vegetation
and environmental changes in the Malay Archipelago.
Biogeographical evolution of the Malay Archipelago (ed. by
T.C. Whitmore), pp. 50–59. Clarendon Press, Oxford.
Morrone, J.J. & Carpenter, J.M. (1994) In search of a method
for cladistic biogeography: an empirical comparison of
component analysis, Brooks parsimony analysis, and threearea statements. Cladistics, 10, 99–153.
Nelson, G. & Ladiges, P.Y. (1996) Paralogy in cladistic biogeography and analysis of paralogy-free subtrees. American
Museum Novitates, 3167, 1–58.
Newsome, J. & Flenley, J.R. (1988) Late Quaternary vegetational history of the Central Highlands of Sumatra. II.
Palaeopalynology and vegetational history. Journal of Biogeography, 15, 555–578.
Nixon, K.C. & Crepet, W.L. (1993) Late Cretaceous fossil
flowers of Ericalean affinity. American Journal of Botany, 80,
616–623.
Pigram, C.J. & Panggabean, H. (1984) Rifting of the northern
margin of the Australian continent and the origin of some
microcontinents in eastern Indonesia. Tectonophysics, 107,
331–353.
Platnick, N.I. (1991) On areas of endemism. Austral biogeography (ed. by P.Y. Ladiges and C.J. Humphries), pp. vii–viii.
CSIRO Australia, Melbourne.
Repetur, C.P., van Welzen, P.C. & de Vogel, E.F. (1997)
Phylogeny and historical biogeography of the genus
Bromheadia (Orchidaceae). Systematic Botany, 22, 465–477.
Ridder-Numan, J.W.A. (1998) Historical biogeography of
Spatholobus (Leguminosae–Papilionoideae) and allies in SE
Asia. Biogeography and geological evolution of SE Asia (ed. by
R. Hall and J.D. Holloway), pp. 259–277. Backhuys Publishers, Leiden, The Netherlands.
Ruedi, M. (1996) Phylogenetic evolution and biogeography of
Southeast Asian shrews (genus Crocidura: Soricidae).
Biological Journal of the Linnean Society, 58, 197–219.
Schuh, R.T. & Stonedahl, G.M. (1986) Historical biogeography
in the indo-pacific: a cladistic approach. Cladistics, 2, 337–
355.
Sleumer, H. (1958) The genus Rhododendron L. in Indochina
and Siam. Blumea Supplement, 4, 39–59.
Sleumer, H. (1966) Ericaceae 1. Rhododendron. Flora Malesiana (ed. by C.G.G.J.v. Steenis), pp. 469–656. Wolters-Noordhoff Publishing, Groningen, The Netherlands.
Sleumer, H. (1973) New species and noteworthy records of
Rhododendron in Malesia (Ericaceae). Blumea, 21, 357–
376.
Sleumer, H. (1980) Past and present taxonomic systems of
Rhododendron based on macromorphological characters.
Contributions toward a classification of Rhododendron: Proceedings International Rhododendron Conference (ed. by J.L.
Luteyn and M.E. O’Brien), pp. 19–26. The New York
Botanical Garden, New York.
Journal of Biogeography 33, 1929–1944
ª 2006 The Authors. Journal compilation ª 2006 Blackwell Publishing Ltd
1943
G. K. Brown, G. Nelson and P. Y. Ladiges
Specht, R.L. (1988) Origin and evolution of terrestrial plant
communities in the wet-dry tropics of Australia. Proceedings
of the Ecological Society of Australia, 15, 19–30.
Stevens, P.F. (1982) Phytogeography and evolution of the
Ericaceae of New Guinea. Biogeography and ecology of New
Guinea (ed. by J.L. Gressitt), pp. 331–354. Dr W. Junk
Publishers, The Hague.
Takeuchi, W. (2000) Rhododendron loranthiflorum (Ericaceae)
from mainland New Guinea: a distributional record and
new subspecies. Edinburgh Journal of Botany, 57, 333–337.
Urushibara-Yoshino, K. & Yoshino, M. (1997) Palaeoenvironmental change in Java island and its surrounding areas.
Journal of Quaternary Science, 12, 435–442.
van Welzen, P.C. (1992) Lepidopetalum Blume (Sapindaceae):
taxonomy, phylogeny, and historical biogeography. Blumea,
36, 439–465.
van Welzen, P.C., Turner, H. & Roos, M.C. (2001) New
Guinea: a correlation between accreting areas and dispersing
Sapindaceae. Cladistics, 17, 242–247.
Whitmore, T.C. (1981a) Palaeoclimate and vegetation history.
Wallace’s Line and plate tectonics (ed. by T.C. Whitmore),
pp. 36–42. Clarendon Press, Oxford.
Whitmore, T.C. (1981b) Wallace’s line and some other plants.
Wallace’s Line and plate tectonics (ed. by T.C. Whitmore),
pp. 70–80. Clarendon Press, Oxford.
Whitmore, T.C. (1987) Biogeographical evolution of the Malay
Archipelago. Clarendon Press, Oxford.
Wikström, N., Savolainen, V. & Chase, M.W. (2001) Evolution
of the angiosperms: calibrating the family tree. Proceedings of
the Royal Society of London, Series B, Biological sciences, 268,
2211–2220.
Williams, E.G. & Rouse, J.L. (1997) Evolutionary history and
speciation in section Vireya. Proceedings of the 1994 Pacific
Region International Rhododendron Conference. Burnie,
Tasmania, Australia (ed. by N. Jordan), pp. 52–60. Australian Rhododendron Society, North West Tasmania Branch,
Burnie, Tasmania.
Wilson, P.G. (1993) Thaleropia, a new genus for Metrosideros
queenslandica (Myrtaceae) and its allies. Australian Systematic Botany, 6, 251–259.
1944
Zetter, R. & Hesse, M. (1996) The morphology of pollen tetrads and viscin threads in some Tertiary, Rhododendron-like
Ericaceae. Grana, 35, 285–294.
SUPPLEMENTARY MATERIAL
The following supplementary material is available for this
article from http://www.Blackwell-Synergy.com:
Appendix S1 Individual species distribution maps (5° intervals).
BIOSKETCHES
Gillian Brown recently completed her PhD at CSIRO Plant
Industry and The University of Melbourne, where she worked
on the phylogeny and biogeography of Rhododendron section
Vireya. Her research to date has focused on phylogeny and
biogeography of large genera within three plant families:
Myrtaceae (Melaleuca and Callistemon), Ericaceae (Rhododendron) and the Leguminosae (Acacia and genera of the tribe
Ingeae).
Gareth Nelson is an Honorary Professorial Fellow at the
School of Botany in Melbourne, after retiring from the
Ichthyology Department, Natural History Museum, New
York. He is well known for his contributions to the development of cladistics and vicariance biogeography.
Pauline Ladiges FAA is Head of the School of Botany at The
University of Melbourne. Her research interests are phylogenetic systematics and historical biogeography of Australian
plants, and she is best known for her work on the eucalypts.
With Gareth Nelson, she has developed the biogeographical
method of subtree analysis.
Editor: Malte Ebach
Journal of Biogeography 33, 1929–1944
ª 2006 The Authors. Journal compilation ª 2006 Blackwell Publishing Ltd