M. citrina (Curtis)
M. citrina (Curtis)
M. linearis var. linearis
Callistemon macropunctatus M. rugulosa
Callistemon montis-zamia M. montis-zamia
M. linearis var. pinifolia
M. williamsii subsp.
Callistemon pungens subsp.
M. linearis var. linearis
Callistemon rugulosus var.
M. viminalis subsp.
Callistemon viminalis subsp.
Callistemon viminalis var.
Melaleuca acacioides subsp.
M. tuberculata var.
Melaleuca calycina subsp.
Melaleuca citrina Turcz.
Melaleuca coccinea subsp.
Melaleuca densa var. pritzelii M. pritzelii
Melaleuca lanceolata subsp.
Melaleuca lanceolata subsp.
Melaleuca lateriflora subsp.
Melaleuca lateriflora var.
Melaleuca minutifolia subsp.
Melaleuca nervosa f. latifolia M. nervosa
Melaleuca nervosa subsp.
Melaleuca styphelioides var.
Melaleuca urceolaris var.
Melaleuca viridiflora var.
Melaleuca viridiflora var. β
M. linearis var. pinifolia
Metrosideros quinquenervia M. quinquenervia
Melaleuca is the basis of several industries in Australia and elsewhere. Predominantly
these industries are based on the extraction of essential oils from the foliage of three
species, Melaleuca alternifolia (but sometimes including minor use of M. linariifolia
and M. dissitiflora), M. cajuputi subsp. cajuputi and M. quinquenervia. An emerg-
ing industry is underway in South-East Asia, especially on potentially acid-sulfate
soils, where trees are being grown primarily for roundwood, and research is being
undertaken into the suitability of their wood for fibre. In view of the challenge to find
novel sources of commercially significant oils, we have been collecting samples over
the past three decades, and have been extracting and analysing the oils. In spite of this
activity, when this book was mooted, information on the oils of about 100 species was
still lacking. Dedicated fieldwork was undertaken in south-western Western Australia,
where most of the unsampled species occurred, and requests were made of individuals
and organisations that might have been in a position to assist. These efforts were suc-
cessful and there are only one species and four infraspecific taxa for which there are
still no data on their essential oils. The majority of the oils reported in this volume have
not previously been described in the scientific literature. As a result of the fieldwork,
several species (e.g. M. halophila, M. hamata, M. ochroma) have been identified as
potentially valuable sources of essential oils and may warrant further investigation of
their oil content and yield.
As it is presently circumscribed, Melaleuca consists of
290 species with 37 infraspecific taxa. Work on the sys-
tematics of the genus in recent years has indicated that
Melaleuca may be made larger due to the inclusion of
several genera presently regarded as distinct. As this work
is at present incomplete, Melaleuca is treated in a conven-
tional concept in this volume, although with the addition
of the species previously known as Callistemon. Many of
the species included in this volume have previously been
treated only skeletally and, for these, detailed descriptions
are being published for the first time. Distribution maps
are included for all species, and colour photographs of the
flowers are included for those we have been able to source.
Apart from essential oils, melaleucas have been used
for a wide range of purposes, from making brushwood
fencing, as ornamental shrubs, as shelterbelt species for
farmland, for rehabilitating salt-affected lands, for street
and park trees, and so on. One interesting potential use
for M. bracteata is as a source of water-soluble betaines—
compounds that act as osmoprotectants against stress in
plants. These compounds may increase stress tolerance in
agricultural crops which in turn would increase national
income from agriculture.
A threat to the future health and genetic diversity of a
substantial number of Melaleuca species in eastern Aus-
tralia is from Puccinia psidii sensu lato (synonym Uredo
rangelii). This exotic pathogen has the common name of
myrtle rust in Australia but it is known as guava or euca-
lyptus rust elsewhere, with origins in Brazil. Myrtle rust
targets species of the family Myrtaceae, including Mela-
leuca. First observed in Australia on the central coast of
New South Wales in 2010, it has now spread from Victoria
to northern Queensland. In susceptible plants, young
torted and severe infection can cause shoots to die, causing
these plants to become stunted after repeated infections.
In the worst cases, death of the whole plant can occur after
repeated destruction of new growth. As this book goes to
press, this disease is of concern to all with an interest in
the conservation and sustainable use of Australian plants
of the family Myrtaceae.
Melaleuca is almost entirely Australian in its distribution yet the first of its species to
be formally described, Melaleuca leucadendra, was based on material from Ambon, in
present-day Indonesia. Georgius Everhardus Rumphius, a merchant with the Dutch
East Indies Company, compiled a detailed account of many of the plants growing in
the Malesian biogeographical region but this was not published until 1741; this impor-
tant work has recently been translated into English and published with annotations
(Rumphius 2011). The plant we now know as Melaleuca leucadendra was called Arbor
alba by Rumphius. Rumphius’ 1741 publication predated the accepted starting point for
the scientific botanical nomenclature of flowering plants and the formal description of
the species occurred in 1754 when the Swedish botanist Carolus Linnaeus gave it the
name Myrtus leucadendra, taking his descriptive data from Rumphius’ work (Linnaeus
1754). Subsequently, Linnaeus realised that his Myrtus leucadendra did not have very
much in common with the other species of Myrtus and in 1767 he described the genus
Melaleuca to accommodate this plant (Linnaeus 1767).
The nomenclature applied to the first endemic Australian
melaleucas to be described was inconsistent due to a lack
of appreciation of the relationships of the species. Doubt-
less this was due to the limited numbers of specimens that
had been collected and consequent uncertainty as to how
the genera of Myrtaceae should be circumscribed. Several
species, such as M. armillaris and M. decora, described in
1788 and 1796, respectively, initially were placed in Metro-
sideros. In other cases, the author recognised a relationship
with Linnaeus’ Melaleuca and the species was placed in
that genus, e.g. M. ericifolia and M. gibbosa, described
in 1797 and 1806, respectively. From the perspective of
having a foundation on which to build new knowledge
of Melaleuca species, George Bentham’s treatment of the
genus in his Australian flora (Bentham 1867) provided the
first comprehensive summary of the species known, and
those species that had been described in Metrosideros were
then brought into Melaleuca. Bentham’s important account
enabled later workers to identify their materials and thus
determine if there were additional species that should be
described. During the next 100 years, many species were
added to the genus, including two from New Caledonia
(M. brongniartii and M. gnidioides) and one from Lord
Howe Island (M. howeana).
Melaleuca was circumscribed by Bentham (1867) in part
as having stamens united in bundles opposite the petals.
onomic history and sy
The related genus Callistemon was distinguished from
he noted that the stamens of Callistemon speciosus (now
M. glauca) were often in bundles. Most of the species
placed in Callistemon are Australian but five species were
also described from New Caledonia, two of which have
stamens in bundles; one wonders why these were not there-
fore placed in Melaleuca. The recognition of Melaleuca and
Callistemon as separate genera had been regarded as being
artificial from the time of the description of the latter.
When describing Callistemon, Robert Brown (1814, p. 547)
wrote: ‘The maximum of Melaleuca exists in the principal
parallel, but it declines less towards the south than within
the tropic, where its species are chiefly of that section
which gradually passes into Callistemon, a genus formed of
those species of Metrosideros that have inflorescences simi-
lar to that of Melaleuca, and distinct elongated filaments’.
Ferdinand Mueller, who was well acquainted with both
genera in the field and in the herbarium, also regarded Cal-
listemon as being artificial (Mueller 1864). Bentham may
in fact have had his own doubts about the distinctness of
Callistemon, for he commented in Flora Australiensis that
Callistemon ‘passes gradually into Melaleuca, with which
F. Mueller proposes to reunite it’ (Bentham 1867, p. 118).
The majority of authors have accepted the two genera as
being distinct but the nineteenth century botanist Henri
Ernest Baillon included Callistemon and two other genera
(Conothamnus and Lamarchea) in Melaleuca, recognis-
ing them at sectional level (Baillon 1876). In 1998, Lyn
Craven and John Dawson transferred the New Caledonian
Callistemon species to Melaleuca, as they considered the
endemic New Caledonian species of the complex should be
placed in the same genus (Craven and Dawson 1998). Cra-
ven (2006) then discussed morphological evidence relevant
to the separation of the two genera and transferred all the
accepted Australian species of Callistemon to Melaleuca.
Studies based on
Surprisingly for a genus of nearly 300 species, few revi-
sionary-level treatments of species groups or of prescribed
geographical regions within Australian Melaleuca have
been published. Perhaps fittingly, in view of the ecologi-
cal and/or economic significance of its several species, the
first was an account of the broad-leaved paperbarks, the
M. leucadendra species group, by Stan Blake (Blake 1968);
these species are a common component of savannah and
woodland communities in northern and north-eastern
Australia, south-eastern Malesia and New Caledonia.
John Carrick and Kosmyn Chorney (Czornij) published a
revisionary-level account of the South Australian species
of Melaleuca in 1979 which gave an insight into arid-zone
species of the genus (Carrick and Chorney 1979). Norm
Byrnes in 1984 published the first part of a concise revision
of the Melaleuca species of northern and eastern Australia
(Byrnes 1984, 1985, 1986). Several of these species were
removed from Melaleuca in 1989 as a result of the resurrec-
tion of Asteromyrtus (Craven 1989). Asteromyrtus is related
to Agonis and Leptospermum and presumably had been
included in Melaleuca only because its stamens are in bun-
dles. Byrnes’ M. punicea was also removed from the genus
because of its anomalous androecium, gynoecium and seed.
Initially the species was placed in Regelia (Barlow 1987a)
but it was subsequently placed in the new genus Petraeo-
myrtus (Craven 1999), as the species was as anomalous in
Regelia as it had been in Melaleuca. However, following
recent investigations using molecular data, M. punicea is
treated in the present volume under Melaleuca (see below).
Bryan Barlow initiated revisionary studies of Melaleuca
in Canberra in the early 1980s. In 1986, Barlow published
the results of his studies of three species complexes: the
M. acacioides, M. tamariscina and M. minutifolia com-
plexes (Barlow 1987b). Subsequently, the M. cuticularis
and M. lanceolata species groups were revised by Barlow
and Kirsten Cowley (Barlow and Cowley 1988) and the
M. fulgens and M. laxiflora species groups were revised
by Cowley and collaborators (Cowley et al. 1990). An
enumeration of the Australian species of Melaleuca sensu
stricto was published by Craven and Brendan Lepschi in
1999; this paper included an identification key—the first
key including all these species of the genus since 1867
(Craven and Lepschi 1999). Four species belonging to
the M. thymoides species group—M. lutea, M. pungens,
M. striata and M. thymoides—were not included in the
1999 enumeration, as Craven, at that time, considered the
species of this group warranted separate generic status;
however, in the present volume they are again included
evidence in classification
Plant classification until recently has been based largely
upon morphological evidence, utilising data from anatomy,
cytology, chemistry (secondary metabolites) and any other
sources where these have been available. The overarch-
ing objective of classification has been to group plants
according to their presumed natural relationships, with
putatively closely related species classified together. The
technological advances that have permitted sequencing
of DNA (deoxyribonucleic acid), together with develop-
ment of computer programs for detecting related species,
have enabled biologists to study the genetic relationships
of plants in detail for the first time. The consequences of
such studies for taxon delimitation, and thus classification,
are varied. In some cases, generic circumscriptions are sup-
ported by analysis of DNA data and there are no changes
to the classification, and hence nomenclature, of a genus.
In other cases, the current classification is not supported,
with part or all of the species sampled falling (or ‘nesting’)
within one or more other genera. There is no automatic
procedure to be adopted in such cases. Typically, if all the
species of a genus or group of genera nest within a single
genus, a decision has to be made to either combine all the
genera into one or to break the original single genus into
several. This clearly has implications for nomenclature. In
other cases, some of the species of a genus are found to nest
within another, with the remaining species still comprising
a distinct group, and in those cases the nested species are
transferred to the other genus. This still has implications
for nomenclature, especially if the name of the genus is
transferred along with the nested genus, as the remaining
species will then require a new generic name.
Species of Melaleuca and the morphologically closely
related genera Beaufortia, Calothamnus, Conothamnus,
Eremaea, Lamarchea, Phymatocarpus and Regelia, and
including M. punicea (Petraeomyrtus), the four species of
the M. thymoides group, and several Australian and New
Caledonian species of Melaleuca that formerly had been
placed in Callistemon, were included by Robert Edwards
and his collaborators in an analysis of chloroplast DNA
data (Edwards et al. 2010). Chloroplast DNA is maternally
inherited, and phylogenies derived from chloroplast DNA
data give a good estimate of the maternal ‘family tree’.
The results of Edwards et al. (2010) showed there were
three major groupings or clades, each of which contained
species of Melaleuca sensu stricto. The whole Melaleuca
group itself formed a well-supported clade relative to the
outgroup taxa included in the analyses. In one of the three
clades are species of the M. leucadendra, M. acacioides,
M. scabra and M. thymoides groups, two New Caledonian
species previously placed in Callistemon, M. punicea,
and representative species of Beaufortia, Calothamnus,
Conothamnus, Eremaea, Lamarchea, Phymatocarpus
and Regelia. In a second clade are representatives of the
M. cuticularis, M. fulgens and M. laxiflora groups, together
with Australian species previously placed in Callistemon
and several species not placed within a particular group-
ing. The third clade contains members of the M. bracteata,
M. cuticularis, M. lanceolata and M. minutifolia groups,
together with other species not allocated to a particular
grouping and the morphologically anomalous M. foliolosa.
The incorporation of nuclear DNA data in phylogenetic
studies enables the paternal contribution to be assessed;
using only the maternal chloroplast data could give a
biased result. The broad structure of the inferred chloro-
plast DNA phylogeny given in Edwards et al. (2010) is in
agreement with the nuclear DNA phylogenies of Ladiges et
al. (1999) and Brown et al. (2001); these were based upon
data derived from a smaller number of species than in the
study of Edwards et al. (2010) but the sampling was drawn
from most of the various genera of the complex.
Current and future
The taxonomic implications of the DNA studies are that
one either includes all the related genera within Melaleuca
or one retains the existing segregate genera and recognises
many new ones, perhaps 10 or more. Given the lack of
distinctive morphological features available to differenti-
ate many of them, such new genera would not be readily
recognisable. Consequently, Melaleuca will be enlarged to
include Beaufortia, Calothamnus, Conothamnus, Eremaea,
Lamarchea, Phymatocarpus and Regelia (L.A. Craven, R.D.
Edwards and K.J. Cowley, in preparation). The species
treated in the present volume are those that accord with
the conventional concept of Melaleuca and also include
M. punicea, the M. thymoides group and those species
previously placed in Callistemon.
Morphological evidence also has been a primary source
of data for developing species’ concepts in Melaleuca. It
is unlikely that this situation will change dramatically in
the foreseeable future but the few studies that have been
made using DNA data have shown that this will be a very
powerful tool for taxonomists. Linda Broadhurst et al.
(2004), utilising nuclear restriction fragment length poly-
morphism (RFLP) data derived from a range of Western
Australian populations of the broombush (M. uncinata)
complex (and including one South Australian popula-
tion from the type locality of M. uncinata), found that
phylo genetic analysis showed the sampled populations
represented seven species of the broombush complex. The
results of the phylogenetic analysis were congruent with
those of a parallel morphological study that encompassed
the whole of the broombush complex in Western Australia
(Craven et al. 2004).
The broad-leaved paperbark group (M. leucadendra
group) is an ecologically important component of vegeta-
tion in northern and eastern Australia. Conventionally it
has been regarded as a taxonomically difficult group and,
in the mid nineteenth century, it was treated by Bentham
(1867) as a single species with numerous varieties. Blake,
with an insight gained from studying the complex in the
field, published a detailed and workable account of the
group that, despite several new species being described
since, remains the most useful guide to the group (Blake
1968). One of the broad-leaved paperbarks, M. quinquen-
ervia, is a common wetlands tree in eastern Australia and
also occurs naturally in Papua province of Indonesia,
woody weed in Florida, United States of America (USA).
This species is harvested for its essential oil, niaouli oil, in
New Caledonia and also in Madagascar, where it is cul-
tivated for this purpose. In a study using DNA sequence
data from two chloroplast and two nuclear regions, Cook
et al. (2008) found that the genome of M. quinquenervia
contained alleles that link the species to several other
broad-leaved paperbark species and that there was regional
sharing of chloroplast haplotypes, indicating introgression
or lineage sorting. This has significance for biological con-
trol studies of M. quinquenervia as it will be necessary to
match the genetics of the weed populations with naturally
occurring populations in Australia when seeking control
agents. A significant conclusion of the Cook et al. (2008)
study was that species boundaries within the complex were
not clear. This work is being extended by Robert Edwards,
drawing upon a comprehensive sampling of the complex
across northern Australia and the results should be of
interest to systematists and others.
The extent to which hybridisation has played a role
in the evolution of Melaleuca species is not known but
it may be important and occurs quite widely across the
genus (Craven 2006). DNA studies may assist in clarifying
species circumscriptions in the bottlebrush (Callistemon)
group, which is at least as difficult taxonomically as the
broad-leaved paperbarks. The bottlebrush M. paludicola
and its putative close relatives M. phratra, M. quercina,
M. sabrina and M. wimmerensis make up one group
worthy of investigation. The red-flowered Queensland
bottlebrushes M. hemisticta, M. lazaridis, M. montis-zamia
and M. pyramidalis are another, as there are populations
in southern Queensland that presently are not definitively
identifiable to one of these species and the whole complex
requires a comprehensive population genetics study of the
type conducted by Broadhurst et al. (2004) on members of
the broombush group. Similarly, the relationship between
M. citrina and M. subulata in southern New South Wales
and eastern Victoria needs investigation. Bill Molyneux
has described Callistemon forresterae, C. genofluvialis,
C. kenmorisonii and C. nyallingensis from this region
(Molyneux 1993, 1994, 1997, 2005) but these are regarded
by one of us (LAC) as being hybrids or hybrid derivatives
between M. citrina and M. subulata and are not accepted
as species in the conventional sense; consequently, they
are not included in the present volume. Apomixis in two
bottlebrush species was studied by James (1958) and this
genetic process may also be a factor in the evolution and
relationships of the bottlebrushes in eastern Australia
It is clear there is need for an infrageneric classification
of Melaleuca, with closely related species being grouped
together in sections and subsections etc. Already there are
nearly 300 species in the genus and with the transfer of the
species flagged above there will be nearly 400; this is too
large a number to be left in an unstructured arrangement.
A preferred classification would be one based upon mor-
phological data alone, but in the case of Melaleuca these are
not sufficient and it will be necessary to incorporate DNA
results in the synthesis. The DNA studies that have been
published to date, while sufficient to support the merger
of the genera of tribe Melaleuceae into a single genus, i.e.
Melaleuca, do not adequately resolve the clades that were
found and contain too few species to permit a classifica-
tion to be prepared. Research presently underway at the
Australian National University, Canberra, using next-
generation sequencing methods and a very much greater
sampling of species across Melaleuca sensu lato should give
information that will guide development of a robust clas-
sification (Mike Crisp and Bo Choi, pers. comm.).
Family and tribe
The genus Melaleuca is in the family Myrtaceae and tribe Melaleuceae.
The generic name is derived from the Greek melas, meaning black or dark, and leucon,
meaning white—apparently alluding to the white branches and black trunk of the first
named species, M. leucadendra, the trunks of which are often blackened by fire.
Species with thick, spongy, peeling bark comprising many papery layers are commonly
referred to as ‘paperbarks’, with some qualifying adjective (e.g. silver-leaved paperbark,
M. argentea). In southern Australia, the common name ‘honey myrtle’ is also well
established for many shrub-sized species (e.g. bracelet honey myrtle, M. armillaris). A
few have distinctive Aboriginal (e.g. winti, M. arcana) or locality (e.g. South Australian
swamp paperbark, M. halmaturorum) names, while some are referred to as ‘tea tree’
(e.g. black tea tree, M. bracteata), a common name shared with many species of Lepto-
spermum. As explained above, this treatment of Melaleuca includes species previously
belonging to the genus Callistemon. They retain their common name of ‘bottlebrush’
which alludes to the resemblance of the flowers and emerging new growth to a kitchen
bottlebrush (e.g. crimson bottlebrush, M. citrina).
The great majority of Melaleuca species are diploid with
ploidy appears to be relatively infrequent in this genus,
with only a few recorded instances of aneuploidy
(2n + 2 = 24), triploidy (3n = 33), tetraploidy (4n = 44)
and hexaploidy (6n = 66) (James 1958; Brighton and
Ferguson 1976; Rye 1979) which have been linked with
hybridisation and apomixis. The apomictic species, M. lin-
earis, may have populations that are diploid, triploid or
tetraploid (James 1958).
the genus Melaleuca
oduction to the genus
Number of species
Melaleuca, as circumscribed in this work, comprises 290
species (327 entities inclusive of infraspecific taxa). As
such, it is the third-largest genus of Myrtaceae after Euca-
lyptus and Syzygium in the Australasian region.
Habit and size
Melaleucas range from woody, multistemmed shrubs to
very large, single-stemmed trees of timber-producing
value. By far the majority of species are shrubs or small
trees less than 10 m tall, 40 of which do not exceed 1 m in
height. These ground-hugging types are found largely in
the south of Western Australia. Fifteen species have been
documented as being over 10 m in height, with seven of
these species exceeding 20 m. Boland et al. (2006) report
that M. cajuputi has been recorded with heights up to
46 m in the Northern Territory and is the tallest tree in
the region; similarly, M. leucadendra has been measured
to 43 m in northern Queensland (Figure 1).
The bark of the majority of species is of the papery type.
separated by thin fibrous layers, which may reach 5 cm in
thickness. The outer layers peel naturally to give a distinc-
tive ragged, torn and unkempt appearance to the lower bole
(Figure 2A). In addition, a substantial proportion (c. 20%)
of Melaleuca species has hard, deeply furrowed, rough bark,
as exemplified by M. bracteata (Figure 2B). There is a fur-
ther 20% of species where the bark is described as fibrous.
There is some overlap in bark types in some species. For
example, M. clarksonii is variously described as displaying
all three bark types in different individuals.
green, bluish-green, grey-green or silvery-grey foliage
unless drought or other stresses (e.g. salt) have stimulated
leaf abscission. Leaves are minute to large. In all, 60% of
recognised species have short (<30 mm long) to very short
(<10 mm) leaves, while the others have medium to long
leaves. Melaleuca leucadendra, with its narrow lanceolate