Annals of Botany 85: 607±612, 2000
doi:10.1006/anbo.2000.1109, available online at http://www.idealibrary.com on
A Study of Genetic Variation and Relationships within the Bamboo Subtribe
Bambusinae using Ampli®ed Fragment Length Polymorphism
J I N P H A N G LO H {, R U T H K I E W {, O H N S E T {, L E O N G H U AT G A N {
and Y I K - Y U E N G A N *{
{School of Science, National Institute of Education, Nanyang Technological University, 469 Bukit Timah Road,
Singapore 259756 and {Singapore Botanic Gardens, 1 Cluny Road, Singapore 259569
Received: 22 June 1999
Returned for revision: 15 September 1999
Accepted: 6 January 2000
Taxonomic and systematic studies of the woody bamboos are traditionally based on ¯oral morphology, which
can cause problems in identi®cation due to the lack of, or infrequent, ¯owering. Limited studies have been conducted using molecular techniques to overcome this problem. In this study, we used ampli®ed fragment length
polymorphisms (AFLPs) to conduct a study of four genera of bamboos (Bambusa, Dendrocalamus, Gigantochloa and
Thyrsostachys) in the subtribe Bambusinae. AFLP analysis using eight primer combinations was carried out on
15 species of bamboo. Results showed that AFLPs distinguish the di€erent species by their unique banding patterns.
Unique AFLPs were detected in 13 of the 15 species examined. The six Bambusa species examined separated into two
clusters. The six Gigantochloa species studied formed a discrete cluster diverging from one of the Bambusa clusters,
while Thyrsostachys was less similar to the Bambusa clusters. The similarity index between B. lako and G. atroviolacea
was the highest, suggesting that B. lako is more appropriately included within the genus Gigantochloa rather than the
genus Bambusa. The two Dendrocalamus species examined were very di€erent with D. brandisii clustering within one
of the Bambusa clusters and D. giganteus appearing as a very distant species. These results support the contention that
critical study of the genus Dendrocalamus is required. The use of AFLPs for identi®cation of particular bamboo
species, as well as for the study of relationships within the subtribe, will be useful for industrial purposes and for
# 2000 Annals of Botany Company
systematic studies.
Key words: Bamboo, Bambusinae, Bambusa, Gigantochloa, Dendrocalamus, Thyrsostachys, AFLP, diversity, AFLP
markers.
I N T RO D U C T I O N
The bamboo subfamily, Bambusoideae, is one of the ®ve
subfamilies of the grasses (Poaceae) (Soderstrom and
Ellis, 1986). The Bambusoideae are currently divided
into the tribes Bambuseae and Olyreae; woody bamboos
(Bambuseae) consist of approx. 77 genera and 1030 species
worldwide (Drans®eld and Widjaja, 1995). The supertribe
Bambusodae is further subdivided into nine subtribes, one
of which is Bambusinae, consisting of ten to 13 genera,
found mostly in tropical Asia (Drans®eld and Widjaja,
1995). Bamboos are vital to many Asian economies, having
important uses ranging from domestic items to rural
housing and raw materials for industry (Drans®eld and
Widjaja, 1995). However, overexploitation and genetic
erosion of bamboo species have made it necessary to
collect germplasm for conservation purposes. In line with
the collection of germplasm, greater attention is needed
in the classi®cation and identi®cation of bamboos (Rao and
Rao, 1995).
Basic knowledge of the biology and genetics of bamboo
is severely lacking. This is a direct result of the unusual life
cycle of bamboo. Among bamboo species, the vegetative
growth phase varies from 1 year to as long as 120 years, and
some species have never been known to ¯ower (Janzen,
* For correspondence. Fax ‡65-469-8928, e-mail [email protected]
0305-7364/00/050607+06 $35.00/00
1976). Identi®cation of sterile plants is therefore problematic as taxonomic studies of bamboos have traditionally
depended heavily on in¯orescence and ¯oral morphology
because: (1) vegetative characters are often environmentally
in¯uenced, which makes them less constant for systematic
purposes (Wu, 1962); (2) characters that delimit species may
be more subtle and not available for study; and (3) bamboo
clones found in Asia are selected for economic value
and are widely distributed without proper identi®cation
at the species level. The application of modern molecular
techniques is therefore of great assistance in species
identi®cation.
However, application of molecular techniques for the
study of genetic diversity in bamboo has been limited.
Studies include the use of restriction fragment length polymorphisms (RFLP) in Phyllostachys (Friar and Kochert,
1991, 1994), isozyme analysis of a limited selection of
bamboos from ®ve genera (Heng et al., 1996), chloroplast
DNA phylogeny of Asian bamboos (Watanabe, Ito and
Kurita, 1994) and world bamboos (Kobayashi, 1997), and
the use of chloroplast rpl16 intron sequences in determining
phylogenetic relationships within the genus Chusquea
(Kelchner and Clark, 1997).
A novel PCR-based assay for plant DNA ®ngerprinting,
ampli®ed fragment length polymorphisms (AFLPs), has
been developed recently, which reveals signi®cant levels of
# 2000 Annals of Botany Company
608
Loh et al.ÐAFLP Analysis of Bamboo (Bambusinae)
DNA polymorphism (Vos et al., 1995). It is a robust and
reliable genetic molecular marker assay and the number of
polymorphisms detected per reaction is much higher than
that revealed by RFLP or random ampli®ed polymorphic
DNA (RAPD) analysis. Genomic DNA is digested with
two endonucleases and site-speci®c adapters are then
ligated to the DNA fragments. Primers complementary to
the adapters and to the restriction sites are designed with
selective nucleotides added to the 30 ends of the primers;
thus only DNA fragments with nucleotides ¯anking
the restriction sites that match the selective nucleotides of
the primer are ampli®ed during PCR. Resolution of the
resulting DNA fragments on standard sequencing gels
allows the detection of AFLPs.
In this paper, we demonstrate the use of AFLPs in
bamboo identi®cation as well as determining genetic
diversity and relationships between bamboo species in the
subtribe Bambusinae.
M AT E R I A L S A ND M E T H O D S
Plant materials
Samples from fully expanded leaves of bamboo plants were
collected from the Singapore Botanic Gardens (Table 1)
and voucher specimens deposited at the herbarium of
Singapore Botanic Gardens. At least three independent leaf
samples were collected for each species, in order to account
for any arti®cial ampli®cations. The leaves were surface
sterilized using the procedure of Zhang, Wendel and Clark
(1997) which was essential to remove fungi, which are
closely associated with bamboos.
T A B L E 1. List of bamboos studied
No.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Accession No.
Species/Cultivar
EG12/17/95/4060
EG12/17/95/4062
EG12/223/94/3250
EG10/00/3249
EG12/00/127.1
EG10/00/8422
EG12/17/95/4065
EG12/00/1255.3
EG12/00/1264.2
EG10/00/1256.1
EG10/00/1265.1
E10/00/1263.1
EG12/43/95/4720.2
EG12/55/95/813
E10/00/1258.1
Bambusa longispiculata Gamble
Bambusa textilis `gracilis' McClure
Bambusa tulda Roxburgh
Bambusa vulgaris `vittata' A. RivieÁre
Bambusa multiplex `riviereorum' R. Maire
Bambusa ventricosa McClure
Bambusa lako Widjaja
Gigantochloa ridleyi Holttum
Gigantochloa rostrata KM Wong
Gigantochloa scortechinii Gamble
Gigantochloa verticillata (Willd.) Munro
Gigantochloa atroviolacea Widjaja
Dendrocalamus giganteus Wallich
Dendrocalamus brandisii (Munro) Kurz
Thyrsostachys siamensis Gamble
T A B L E 2. Sequence of primers used
Name/Abbreviation
Enzyme
Type
Sequence (50 ±30 )
GYY101/EA ‡
GYY102/EA ÿ
GYY103/MA ‡
GYY104/MA ÿ
EcoRI
EcoRI
MseI
MseI
Adapter ‡
Adapter ÿ
Adapter ‡
Adapter ÿ
CTCGTAGACTGCGTACC
AATTGGTACGCAGTCTAC
GACGATGAGTCCTGAG
TACTCAGGACTCAT
GYY105/E-A
GYY107/E-AAC
GYY108/E-AAG
GYY109/E-ACA
GYY110/E-ACT
GYY111/E-ACC
GYY112/E-ACG
GYY113/E-AGC
GYY114/E-AGG
EcoRI
EcoRI
EcoRI
EcoRI
EcoRI
EcoRI
EcoRI
EcoRI
EcoRI
Primer ‡ 1
Primer ‡ 3
Primer ‡ 3
Primer ‡ 3
Primer ‡ 3
Primer ‡ 3
Primer ‡ 3
Primer ‡ 3
Primer ‡ 3
GACTGCGTACCAATTCA
GACTGCGTACCAATTCAAC
GACTGCGTACCAATTCAAG
GACTGCGTACCAATTCACA
GACTGCGTACCAATTCACT
GACTGCGTACCAATTCACC
GACTGCGTACCAATTCACG
GACTGCGTACCAATTCAGC
GACTGCGTACCAATTCAGG
GYY106/M-C
GYY115/M-CAA
GYY116/M-CAC
GYY117/M-CAG
GYY118/M-CAT
GYY119/M-CTA
GYY120/M-CTC
GYY121/M-CTG
GYY122/M-CTT
MseI
MseI
MseI
MseI
MseI
MseI
MseI
MseI
MseI
Primer ‡ 1
Primer ‡ 3
Primer ‡ 3
Primer ‡ 3
Primer ‡ 3
Primer ‡ 3
Primer ‡ 3
Primer ‡ 3
Primer ‡ 3
GATGAGTCCTGAGTAAC
GATGAGTCCTGAGTAACAA
GATGAGTCCTGAGTAACAC
GATGAGTCCTGAGTAACAG
GATGAGTCCTGAGTAACAT
GATGAGTCCTGAGTAACTA
GATGAGTCCTGAGTAACTC
GATGAGTCCTGAGTAACTG
GATGAGTCCTGAGTAACTT
Loh et al.ÐAFLP Analysis of Bamboo (Bambusinae)
DNA extraction
Plant DNA was extracted using the CTAB method
(Reichardt and Rogers, 1993).
AFLP analysis
609
coecient was converted to genetic diversity estimates
(GDEs; Table 4), which in turn were used for UPGMA
cluster analysis (Fig. 3).
An interesting observation was the clustering of Gigantochloa atroviolacea with Bambusa lako (Timor Giant Black).
These two species showed the highest similarity index.
AFLP analysis was carried out following Vos et al. (1995)
except that the EcoRI primers used were not radioactively
labelled as in the original protocol. Instead, a modi®ed
silver staining method (Promega Corporation, Madison,
USA) was used (Loh et al., 1999). The ®nal PCR products
were run on a 6% denaturing polyacrylamide gel in
1 TBE bu€er. The primer and adapter sequences are
shown in Table 2.
Data analysis
For the diversity analysis, bands were scored as present
(1) or absent (0). A square symmetric matrix of similarity
was then obtained using Jaccard's similarity coecient
[a/(nÿd)], where a is the number of fragments in common
between two cultivars, n is the total number of fragments
scored and, d is the number of fragments absent in both
cultivars (Dudley, 1993). Genetic diversity estimates
(GDEs) were then calculated as 1 minus Jaccard's similarity
coecient and used for cluster analysis using the
unweighted pair group method of arithmetic averages
(UPGMA) technique.
R E S U LT S A N D D I S C U S S I O N
Species identi®cation using AFLP markers
The degree and nature of polymorphism revealed by AFLP
are depicted in Figs 1 and 2. The eight AFLP primer
combinations used in this study generated 603 polymorphic
bands (mean ˆ 75.3 per pair) across 15 species of bamboo,
along with a total of 43 monomorphic bands. This
compares favourably with RFLP analysis of Phyllostachys
in which Friar and Kochert (1994) detected 380 polymorphic bands using 43 probe±enzyme combinations on
12 bamboo clones i.e. an average of approx. nine polymorphisms per probe±enzyme combination. The AFLP
analysis of eight primer combinations on 15 species of
bamboo was used to identify unique molecular markers for
each species if possible. The relatively small number of
primer combinations tested yielded unique molecular
markers for all species except Bambusa tulda and Gigantochloa verticillata (Table 3). Examples of unique molecular
markers identi®ed are also depicted in Figs 1 and 2. The
molecular markers identi®ed could be used to generate
speci®c probes for the di€erent species.
Genetic diversity between bamboo species
The 15 bamboo species examined belong to four genera,
all in the subtribe Bambusinae. AFLP data from the
analysis of eight primer combinations were also used to
determine relationships between species within the subtribe.
The similarity index generated using Jaccard's similarity
F I G . 1. AFLP pro®le generated using primer pair 19 (E-ACA, MCAG) for the 15 bamboo species. Lane 1, Bambusa longispiculata; lane
2, B. textilis; lane 3, B. tulda; lane 4, B. vulgaris; lane 5, B. multiplex;
lane 6, B. ventricosa; lane 7, B. lako; lane 8, Gigantochloa ridleyi; lane 9,
G. rostrata; lane 10, G. scortechinii; lane 11, G. verticillata; lane 12,
G. atroviolacea; lane 13, Dendrocalamus giganteus; lane 14, D. brandisii;
lane 15, Thyrsostachys siamensis.
610
T A B L E 3. Number of unique molecular marker bands speci®c for each bamboo species detected upon AFLP analysis using eight primer combinations
Total number
of unique
markers per
primer pair
EcoRI
MseI
B. longispiculata
B. textilis
B. tulda
B. vulgaris
B. multiplex
B. ventricosa
B. lako
G. ridleyi
G. rostrata
G. scortechinii
G. verticillata
G. atroviolacea
D. giganteus
1
10
19
28
37
46
55
64
AAC*
AAG
ACA
ACC
ACG
ACT
AGC
AGG
CAA*
CAC
CAG
CAT
CTA
CTC
CTG
CTT
4
5
±
1
±
±
1
4
1
3
±
1
1
3
2
±
±
±
±
±
±
±
±
±
±
±
±
±
±
2
1
1
±
±
1
±
±
3
5
±
±
±
±
±
1
±
±
1
±
1
±
1
±
1
±
1
±
3
±
3
2
2
±
±
±
2
±
1
±
±
2
±
3
±
±
±
±
1
±
1
±
±
±
±
±
±
±
±
±
±
1
±
±
±
±
±
±
±
7
1
4
2
4
7
5
1
3
±
4
3
2
1
3
±
5
2
1
1
3
3
16
15
17
10
13
18
20
19
15
11
±
4
9
2
4
10
5
5
±
1
25
19
18
128
Total
D. brandisii T. siamensis
*EcoRI: EcoRI-adapter based primer; the selective nucleotides added at the 30 end are indicated.
*Mse: MseI-adapter based primer; the selective nucleotides added at the 30 end are indicated.
T A B L E 4. Mean of the AFLP-based pairwise genetic diversity estimates (GDEs) between 15 species of bamboo using eight primer combinations
B. longispiculata
B. textilis
B. tulda
B. vulgaris
B. multiplex
B. ventricosa
B. lako
G. ridleyi
G. rostrata
G. scortechinii
G. verticillata
G. atroviolacea
D. giganteus
D. brandisii
T. siamensis
B. longispiculata
B. textilis
B. tulda
B. vulgaris
B. multiplex
B. ventricosa
B. lako
G. ridleyi
G. rostrata
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
0.585
±
±
±
±
±
±
±
±
±
±
±
±
±
±
0.617
0.568
±
±
±
±
±
±
±
±
±
±
±
±
±
0.598
0.607
0.187
±
±
±
±
±
±
±
±
±
±
±
±
0.581
0.448
0.647
0.635
±
±
±
±
±
±
±
±
±
±
±
0.620
0.581
0.130
0.257
0.668
±
±
±
±
±
±
±
±
±
±
0.591
0.547
0.502
0.547
0.598
0.486
±
±
±
±
±
±
±
±
±
0.609
0.782
0.508
0.524
0.615
0.539
0.503
±
±
±
±
±
±
±
±
0.600
0.604
0.485
0.488
0.613
0.508
0.497
0.266
±
±
±
±
±
±
±
GDEs represent 1 ÿ Jaccard's Similarity Coecient.
G. scortechinii G. verticillata G. atroviolacea D. giganteus
0.591
0.509
0.541
0.579
0.680
0.524
0.520
0.468
0.443
±
±
±
±
±
±
0.594
0.566
0.504
0.558
0.637
0.482
0.492
0.491
0.472
0.316
±
±
±
±
±
0.584
0.552
0.501
0.551
0.621
0.701
0.145
0.495
0.480
0.486
0.457
±
±
±
±
0.736
0.742
0.820
0.819
0.763
0.795
0.755
0.766
0.773
0.775
0.727
0.738
±
±
±
D. brandisii
T. siamensis
0.670
0.679
0.605
0.623
0.743
0.575
0.605
0.609
0.591
0.535
0.500
0.584
0.777
±
±
0.645
0.642
0.639
0.639
0.659
0.650
0.613
0.647
0.637
0.605
0.619
0.610
0.789
0.619
±
Loh et al.ÐAFLP Analysis of Bamboo (Bambusinae)
Primer
pair
Loh et al.ÐAFLP Analysis of Bamboo (Bambusinae)
0.8
0.4
611
0
B. tulda
B. ventricosa
B. vulgaris
G. atroviolacea
B. lako
G. verticillata
G. scortechinii
G. ridleyi
G. rostrata
D. brandisii
B. longispiculata
B. multiplex
B. textilis
T. siamensis
D. giganteus
F I G . 3. UPGMA cluster analysis of AFLP data generated by eight
primer combinations for 15 species of bamboo depicting patterns of
genetic diversity. Scale depicts genetic diversity estimates (GDEs).
F I G . 2. AFLP pro®le generated using primer pair 55 (E-AGC, MCTG) for the 15 bamboo species. Lane 1, Bambusa longispiculata;
lane 2, B. textilis; lane 3, B. tulda; lane 4, B. vulgaris; lane 5,
B. multiplex; lane 6, B. ventricosa; lane 7, B. lako; lane 8, Gigantochloa
ridleyi; lane 9, G. rostrata; lane 10, G. scortechinii; lane 11, G. verticillata; lane 12, G. atroviolacea; lane 13, Dendrocalamus giganteus; lane
14, D. brandisii; lane 15, Thyrsostachys siamensis.
Widjaja (1997) distinguished G. atroviolacea from sterile
plants of Timor Giant Black based on vegetative
morphology, and renamed Timor Giant Black Bambusa
lako. However, she noted that ¯owers and in¯orescences are
needed for con®rmation of its generic identity. This is a
classic example showing the need for molecular techniques
which do not require reproductive structures for precise
classi®cation. Our results and the UPGMA analysis
indicate, from the high similarity index shared between
G. atroviolacea and Timor Giant Black, that the two species
are genetically very similar and suggest that Timor Giant
Black belongs to the genus Gigantochloa.
AFLP results also support the position of Thyrsostachys
within the Bambusinae, as a genus distant from Bambusa
and Gigantochloa as suggested by the molecular work of
Watanabe et al. (1994).
In the cluster analysis, the Bambusa species form two
distinct clusters. Bambusa tulda, B. ventricosa and B. vulgaris
form one cluster, while B. longispiculata, B. multiplex and
B. textilis form the other. The two di€erent clusters for
Bambusa species suggest that the genus Bambusa is
polyphyletic and highlights the potential of AFLP techniques in assessing the variation and relationships within
the genus Bambusa.
The genus Gigantochloa is genetically less diverse with all
species examined being placed together in a single cluster,
diverging from the B. tulda, B. ventricosa, B. vulgaris
cluster. The molecular study of Watanabe et al. (1994) also
found a very close relationship between the genera Bambusa
and Gigantochloa.
612
Loh et al.ÐAFLP Analysis of Bamboo (Bambusinae)
However, the two Dendrocalamus species examined were
very di€erent, with D. brandisii falling within the Bambusa
cluster and D. giganteus showing the least genetic similarity
to any of the species of subtribe Bambusinae examined. The
relationship between species within the genus Dendrocalamus is still not understood. One line of approach to this
problem has been to split it into several genera (see Li,
1997) but currently the consensus is to use Dendrocalamus
in the broad sense (Li, 1997; Wong, 1995). At one extreme,
some species are closely similar to species of Bambusa
and there appear to be intermediate species (Li, 1997).
Molecular biology will be a powerful tool for resolving this
problem.
In our study, the very di€erent positions of the two
Dendrocalamus species studied con®rms the wide range of
variation within Dendrocalamus: D. brandisii clustered close
within the Bambusa cluster. While D. giganteus clustered
very distantly from the rest of the Bambusa and Gigantochloa species.
For Peninsular Malaysian species, Wong (1995) drew
attention to two distinct groups within the genus Dendrocalamus, based on a combination of vegetative characters,
in¯orescence morphology and ¯owering behaviour. Wong
(1995) also cited the work of Chou and Hwang (1985),
whose study on the chromatographic separation of phenolic
compounds and electrophoretic isozyme patterns in Dendrocalamus distinguished four distinct groups within the
genus. As Wong (1995) noted: `Dendrocalamus requires
further critical study' and molecular biology holds great
potential in assessing the relationships of Dendrocalamus
species.
In conclusion, AFLP has been shown to be useful in
studying the genetic similarity between bamboo species and
genera as well as identifying molecular markers speci®c for
particular species.
AC K N OW L E D G E M E N T S
This research was funded by the Academic Research Fund,
National Institute of Education, Nanyang Technological
University, Singapore, RP12/98/GYY. We thank the
Director, Singapore Botanic Gardens, for permission to
collect the leaf samples.
L I T E R AT U R E C I T E D
Chou CH, Hwang YH. 1985. A biochemical aspect of phylogenetic
study of Bambusaceae in Taiwan III. The genera Arthrostylidium,
Chimonobambusa and Dendrocalamus. Botanical Bulletin Academia Sinica 26: 155±170.
Drans®eld S, Widjaja EA. 1995. Plant resources of South-East Asia
No. 7 Bamboos. Leiden: Backhuys Publishers.
Dudley JW. 1993. Molecular markers in plant improvement: manipulation of genes a€ecting quantitative traits. Crop Science 33:
660±668.
Friar E, Kochert G. 1991. Bamboo germplasm screening with nuclear
restriction fragment length polymorphisms. Theoretical and
Applied Genetics 82: 697±703.
Friar E, Kochert G. 1994. A study of genetic variation and evolution of
Phyllostachys (Bambusoideae: Poaceae) using nuclear restriction
fragment length polymorphisms. Theoretical and Applied Genetics
89: 265±270.
Heng HP, Yeoh HH, Tan CKC, Rao AN. 1996. Leaf isozyme
polymorphisms in bamboo species. Journal of the Singapore
National Academy of Science 22: 10±14.
Janzen DH. 1976. Why bamboos wait so long to ¯ower. Annual
Reviews in Ecological Systematics 7: 347±391.
Kelchner SA, Clark LG. 1997. Molecular evolution and phylogenetic
utility of the chloroplast rpl16 intron in Chusquea and the
Bambusoideae (Poaceae). Molecular Phylogenetics and Evolution
8: 385±397.
Kobayashi M. 1997. Phylogeny of world bamboos analysed by
restriction fragment length polymorphisms of chloroplast DNA.
In: Chapman GP, ed. The bamboos. Linnean Society Symposium
Series. UK: Linnean Society of London, 227±234.
Li DZ. 1997. The Flora of China Bambusoideae ProjectÐproblems
and current understanding of bamboo taxonomy in China. In:
Chapman GP, ed. The bamboos. Linnean Society Symposium
Series. UK: Linnean Society of London, 61±81.
Loh JP, Kiew R, Kee A, Gan LH, Gan YY. 1999. Ampli®ed fragment
length polymorphism (AFLP) provides molecular markers for the
identi®cation of Caladium bicolor cultivars. Annals of Botany 84:
155±161.
Rao AN, Rao VR. 1995. Patterns of variation in bamboo. In: Williams
JT, Rao IVR, Rao AN, eds. Genetic enhancement of bamboos and
rattan. New Delhi: International Network for Bamboo and
Rattan, Appendix 5.
Reichardt MJ, Rogers SJ. 1993. Plant DNA Isolation using CTAB. In:
Ausubel FM, Brent R, Kingston RE, Moore DD, Seidman JG,
Smith JA, Struhl K, eds. Current protocols in molecular biology.
USA: John Wiley and Sons, Supplement 22.
Soderstrom TR, Ellis RP. 1986. The position of bamboo genera and
allies in a system of grass classi®cation. In: Soderstrom TR, Hilu
KW, Campbell CS, Barkworth ME, eds. Grass systematics and
evolution. USA: Smithsonian Institution Press, 225±238.
Vos P, Hogers R, Bleeker M, Reijans M, van de Lee T, Hornes M,
Frijters A, Pot J, Peleman J, Kupier M, Zabeau M. 1995. AFLP: a
new technique for DNA ®ngerprinting. Nucleic Acids Research 23:
4407±4414.
Watanabe M, Ito M, Kurita S. 1994. Chloroplast DNA phylogeny of
Asian bamboos (Bambusoideae, Poaceae) and its systematic
implication. Journal of Plant Research 107: 253±261.
Widjaja E. 1997. New taxa in Indonesian bamboo. Reinwardtia 11:
57±152.
Wong KM. 1995. The bamboos of Peninsular Malaysia. Malaysia:
Forest Research Institute Malaysia (FRIM).
Wu MCY. 1962. Classi®cation of Bambuseae based on leaf anatomy.
Botanical Bulletin Academia Sinica 3: 83±107.
Zhang WP, Wendel JF, Clark LG. 1997. Bamboozled again! Inadvertent
isolation of fungal rDNA sequences from bamboos (Poaceae:
Bambusoideae). Molecular Phylogenetics and Evolution 8: 205±217.