Botanical Studies (2006) 47: 443-452.

*

Corresponding author: E-mail: renuka@kfri.org; Tel: 0487-

2699-061; Fax: 0487-2699-249.

INTRODUCTION

Rattans are spiny climbing palms belonging to the

subfamily Calamoideae of the Arecaceae (Uhl and

Dransfield, 1987). They comprise about 600 species in

13 genera distributed in equatorial Africa, South Asia,

Southern China, the Malay Archipelago, Australia and the

Western Pacific. Rattans form one of the major non-wood

forest products in international trading. Approximately

700 million people trade or use rattan for different

purposes worldwide mainly for furniture and cottage

industries. The global trade and subsistence value of rattan

and its products is estimated at over US$ 7,000 million

per annum (Pabuayan, 2000). Due to overexploitation,

habitat degradation and low regeneration capacity, the

rattan resources of the world are under serious threat. It is

estimated that around 117 species of rattans are treated as

threatened to some degree (Walter and Gillet, 1998).

In India, rattans comprise about 60 species in four

genera¡Xviz. Calamus, Daemonorops, Korthalsia, an d

Plectocomia¡Xdistributed in the Western Ghats, North

Eastern states, and the Andaman and Nicobar Islands.

Calamus rivalis and C. metzianus are two slender

diameter rattans found in the Western Ghats and Sri

Lanka. Calamus metzianus was first described by Schlecht

(cf. Beccari, 1908) based on specimens collected by Rev.

Metz from the Canara district of Karnataka, India. It was

later discovered in Nilambur of Kerala, India (Renuka

and Bhat, 1987). Calamus rivalis was originally described

from Sri Lanka by Thwaites and was validly published by

Trimen (Beccari, 1908). This species was also distributed

in Ashramam (Kollam) and Chertala (Alapuzha) of India.

Beccari (1908) suggested that C. metzianus represented

the continental form of C. rivalis and this species could be

distinguished from C. metzianus by a larger fruit size and

distinctly channeled fruit scales.

The purposes of the present study are to examine

the patterns of phenetic structure and levels of genetic

variation within and among populations of these two

species using random amplified polymorphic DNA

(RAPD) markers and morphological characters. The

RAPD protocol is relatively quick, easy to perform, and

requires no sequence information prior to analysis and

only a nanogram quantities of DNA (Williams et al.,

1990).

TAXONOMY

Taxonomic reconsideration of Calamus rivalis Thw. ex

Trim. and C. metzianus Schlecht (Arecaceae) through

morphometric and molecular analyses

V.B. SREEKUMAR

1

, C. RENUKA

1,

*, T.B. SUMA

2

, and M. BALASUNDARAN

2

1

Division of Forest Ecology and Biodiversity Conservation, Kerala Forest Research Institute, Peechi, Thrissur, 680 653,

Kerala, India

2

Division of Sustainable Natural and Plantation Forest Management, Kerala Forest Research Institute, Peechi, Thrissur,

680 653, Kerala, India

(Received November 19, 2004; Accepted March 31, 2006)

ABSTRACT.

Molecular and morphometric analyses of different Indian and Sri Lankan populations of two

rattan species (Calamus rivalis and C. metzianus) were carried out using Random Amplified Polymorphic

DNA (RAPD) markers and morphological characters. Multivariate analysis of RAPD and morphological data

failed to separate these populations into two distinct species. RAPDs generated a total of 117 markers with

10 decamer primers, of which 95 percent were polymorphic. The percentage of polymorphic loci between

populations varied from 26.5 to 68.38, and genetic distance between populations ranged from 0.05 to 0.28.

The genetic variation within populations was greater than the variation among populations. Both the RAPD

and morphometric data suggest the existence of a single species, and hence C. rivalis can be merged to C.

metzianus.

Keywords: Calamus; Dendrogram; Genetic diversity; RAPD.

444

Botanical Studies, Vol. 47, 2006

MATERIALS AND METHODS

Morphological characters

Two populations of C. rivalis from the Western Ghats

of India and three from Sri Lanka and two populations of

C. metzianus from the Western Ghats were included in

this study (Table 1). Herbarium specimens were collected

from all these populations. The voucher specimens

were deposited in the KFRI herbarium. The details of

OTUs (Operational Taxonomic Units), included for the

morphometric analysis, are available on request. The

characters were chosen on the basis of differences among

taxa in the vegetative and reproductive parts. Only the

mature plants were considered for the study, and a total of

32 qualitative and quantitative characters were selected.

Qualitative characters were scored as 0 for absence and

1 for presence. Multiple herbarium sheets from a single

collection were grouped into a single OTU; 107 such

OTUs were used for the analysis.

RAPD analysis

Sample collection.

Juvenile leaves were collected

from two populations in Honaver and Nilambur for

C. metzianus and from five populations in Chertala,

Ashramam of India and from Watareke, Yagirala, and

Matugama in Sri Lanka for C. rivalis. A total of 60

individual samples were selected for the analysis (Table 1).

DNA Extraction

. The leaf specimens were collected

in plastic bags and chilled at -20¢XC until DNA was

extracted. For Sri Lankan samples, dried materials were

used for DNA isolation. Total DNAs were extracted from

1 g of the leaf tissues using the modified CTAB protocol

(Doyle and Doyle, 1990).

RAPD Reaction.

PCR-RAPD analysis was carried

out according to the Williams et al. (1990) protocol using

ten primers, viz. OPAU02, OPA04, OPA18, OPAW07,

OPAW09, OPAW10, OPAW20, OPE02, OPE18 and

OPB15 (Operon Technologies, Alameda, CA), which

were selected out of the 65 primers screened, based on the

repeatability of their DNA band profiles. Amplifications

were performed on a PTC-100 Thermal cycler (MJ

Research Inc., USA) in 25 £gL reaction mixtures containing

50-100 ng of template DNA, 200 £gM each of dATP, dTTP,

dCTP and dGTP, 3 units of Taq DNA polymerase, 1 £gL

(20 pm) of each primer, and 5 £gL Taq buffer with 1.5

mM MgCl

2

(Genei, Bangalore, India). The amplification

was performed with 45 cycles, each of 60 s denaturation

(94

o

C), 60 s annealing (36

o

C), and 120 s extension (72

o

C).

The last cycle was followed by incubation for 10 min at

72

o

C.

Electrophoresis of the amplified products.

The

RAPD products were electrophoresed on 1.5% agarose

gel (Sigma, USA) in TBE buffer (40 mM Tris-borate,

1 mM EDTA, pH 8.0). The gel, after the completion of

electrophoresis, was stained with ethidium bromide, and

the bands were compared with DNA markers [100-bp

DNA ladder and Low range DNA ruler (Genei)]. The

gels were documented using a Kodak Digital Science

Electrophoresis Documentation and Analysis System 120

(Kodak, USA).

Data analysis

Morphological characters.

Cluster analysis and

ordination were performed using NTSYS (Rolf, 2000).

For cluster analysis, the standardized data were used to

compute a distance matrix based on Average taxonomic

distances, Manhattan distances, and Euclidian distances

before being subjected to the unweighted pair-group

method with an arithmetic averages (UPGMA) clustering

algorithm. To test the consistency between the resulting

phenogram and the original data, co-phenetic correlation

coefficients were calculated for each phenogram and data

matrix pair. The phenogram with the highest co-phenetic

correlation coefficient (r) was reported here.

Table 1. Details of selected populations of C. metzianus and C. rivalis and comparison of various genetic diversity measures.

Population

Country Latitude (oN) Longitude (oL) No of plants

h npl ppl

RAPD Morphology

C. metzianus

Honaver, Western Ghats India 14o08¡¦

74o30¡¦ 10

15

0.1637 47 40.17%

Nilambur, Western Ghats India 11o17¡¦

76o28¡¦ 10

20 0.2458 80 68.38%

C. rivalis

Ashramam, Western Ghats India 08o53¡¦

76o35¡¦ 10

18

0.2086 66 56.41%

Chertala, Western Ghats

India 09o42¡¦

76o19¡¦ 10

15

0.2326 69 58.97%

Watareke

Sri Lanka 06

o

51¡¦ 80o03¡¦ 10

13

0.1633 54 46.15%

Yagirala

Sri Lanka 06o28¡¦

80o11¡¦

5

13 0.1085 31 26.50%

Matugama

Sri Lanka 06o31¡¦

80o07¡¦

5

13

0.1331 38 32.48%

Mean

0.1793 55 47.00%

h = Nei¡¦s (1973) gene diversity; npl = Number of polymorphic loci; ppl = Percentage of polymorphic loci.

SREEKUMAR et al. ¡X Morphological and RAPD analyses of

Calamus

445

For ordination analyses, we carried out principal

coordinate analysis (PCOA) for a mix of qualitative and

quantitative data and principal component analysis (PCA)

for quantitative data. Additionally, from the standardized

fruit data, average Euclidean distance was calculated for

each population pair to test the correlation between the

average Euclidean distance and the genetic distance using

a mantel test.

RAPD analysis.

RAPD products were scored for

presence (1) and absence (0) of bands. The data matrices

were analysed using the Popgene, Version 1.31 package,

and a pairwise comparison of populations was made

(Yeh et al., 1999). The genetic diversity parameters

within populations, viz. number of polymorphic loci and

gene diversity, were determined. Genetic differentiation

between the analysed populations was calculated

according to Nei (1973).

Genetic distances (Nei, 1978) between all populations

were obtained from Popgene, Version 1.31, and the

resulting distance matrix was then used to construct an

unrooted phenetic tree of the different populations using

the Fitch Program of PHYLIP, Version 3.5 (Felsenstein,

1993). To evaluate the correlation between genetic

distance and geographic distance, the product moment

correlation coefficients were calculated between the

genetic and geographic distance matrices, and significance

levels of the correlation between these matrices were

estimated by a mantel test using TFPGA software,

Version 1.3. Analysis of molecular variation (AMOVA;

Excoffier et al., 1992) was used to estimate variance

components and to test the significance of the partitioning

of RAPD variation among regions and among and within

populations.

Cluster analysis and principal coordinate analysis were

carried out using the NTSYS software package, Version

2.1 (Rolf, 2000). Genetic similarities based on Jaccard¡¦s

coefficient (Jaccard, 1908) were calculated among pairs

using the SIMQUAL option and ordered in a similarity

matrix. The similarity matrix was run on sequential,

agglomerative, hierarchical, and nested clustering (SAHN)

(Sneath and Sokal, 1973) using UPGMA (unweighted

pair group method with arithmetic average). Cophenetic

correlation was calculated to measure goodness of fit.

Principal co-ordinate analysis (PCO) was performed using

the following modules of the NTSYS program: STAND,

SIMINT, DECENTER and EIGEN to identify the number

of groups based on eigen vectors. Three-dimensional

ordination provided an additional representation of genetic

relationships among the individuals of populations.

RESULTS

Morphological characters

The phenogram constructed using Manhattan distances

showed the least distortion with a co-phenetic correlation

coefficient, r = 0.8713. In euclidian distances, r = 0.86829,

and in average taxonomic distances, r = 0.8682 (data not

shown). The OTUs fell into three major clusters. The first

and second clusters were composed of a complex series

of closely nested clusters made up of a majority of the

plants of the Western Ghats (Honaver, Nilambur, Chertala

and Ashramam) and all of the OTUs from Matugama

of Sri Lanka. The third cluster included all remaining

Sri Lankan OTUs and some plants from Chertala. The

first cluster (Hon 1 to Nil 11) consisted mainly of OTUs

from Chertala and Ashramam with which the OTUs from

Honaver and Nilambur overlapped. The second cluster

(Nil 10 to Hon 13) includes the plants from Nilambur,

Ashramam, Honaver, Chertala, and Matugama. The OTUs

from Nilambur, Honaver and Chertala were found to be

scattered among different clusters. The majority of OTUs

from Ashramam were clustered together with OTUs from

Nilambur. Among Sri Lankan specimens, OTUs from

Matugama formed a distinct group linked with some

OTUs of Chertala, Honaver, Watareke, and Yagirala..

The remaining OTUs, from Watareke and Yagirala, are

scattered together and intermingled with OTUs from

Chertala.

In ordination, the first principal components account for

38.58% of the total variation, with leaf sheath spine length,

number of male rachilla/partial inflorescence, number

of female rachilla/partial inflorescence, fruit apical beak

length, number of fruits per rachilla, inflorescence caudex

length, and fruit diameter having the highest loadings

(Table 2). The second principal component explained

22.25% of the total variation, with stem diameter, nature

of leafsheath spine and seed length having the highest

loadings. The third principal component explained 9.17%

of the total variation, with leaflet length, inflorescence,

and caudex length having the highest loadings.

In Principal Coordinate Analysis, the first three

principal co-ordinates explained a total of 69% of the total

variation (Table 3) and failed to disclose any major gap in

the pattern of morphological variation (Figure 1), clearly

indicating the lack of distinct phenetic structure. However,

OTUs from geographic regions do tend to fall together in

a given proportion of the three dimensional factor space,

as happens in cluster analysis.

RAPD analysis

Primer utility.

Sixty-five primers from Kit OPA,

OPB, OPE, OPAU and OPAW (Operon Technology, USA)

were initially screened. Out of these, ten were chosen for

further analysis based on the number and reproducibility

of amplified products. A total of 117 markers were

obtained, with molecular sizes ranging from 100 bp to

2,500 bp. Polymorphism was very pronounced, with 95%

polymorphic markers across all primers.

The percentage of polymorphic loci (ppl) varied from

26.5 in the Matugama population to 68.38 in the Chertala

population. The Chertala population had highest gene

diversity index (h = 0.25) and the Matugama population of

Sri Lanka the lowest (h = 0.11) (Table 1).

446

Botanical Studies, Vol. 47, 2006

Table 2. Eigen vector coefficient for principal component analysis. Character loading values are for the first three Principal

Component (PC) axes.

No. Characters

PC 1

PC 2

PC 3

1. Leafsheath spine length

0.8674

0.2061

0.1302

2. Female rachillae length

0.8515

-0.0476

0.1849

3. Inflorescence caudex length

0.8120

-0.1423

0.3063

4. Fruit diameter

0.7826

0.0027

-0.2615

5. Number of fruits per rachillae

0.7012

-0.0399

-0.0737

6. Number of male rachilla

0.6809

-0.0919

-0.3409

7. Number of female

0.6809

0.1761

-0.2767

8. Internodal length

0.5943

0.4375

-0.0841

9. Rachillae prophyll length

0.5345

0.1184

0.1657

10. Leaflet length

0.5021

0.1985

0.3648

11. Stem diameter

0.3242

0.5600

0.0715

12. Seed length

0.1669

0.5111

-0.2912

13. Partial Inflorescence length

-0.1342

0.3417

-0.6384

14. Number of vertical rows of scales

-0.4868

0.3799

-0.3524

Eigen values

9.147

3.920

2.935

Percentages

28.584

12.251

9.174

Cumulative percentage

28.584

40.835

50.010

Figure 1. Principal coordinate analysis (PcoA) of 32 morphological characters. (H: Honaver; N: Nilambur; A: Ashramam; C:

Chertala; W: Watareke; Y: Yagirala; M: Matugama).

SREEKUMAR et al. ¡X Morphological and RAPD analyses of

Calamus

447

The estimated average genetic distances between

populations ranged from 0.05 (between Yagirala and

Watareke) to 0.28 (between Yagirala and Honaver).

Partitioning of variation within and between populations

using an analysis of molecular variance (AMOVA)

showed that 65.74% of the genetic variability existed as

variation between populations (P<0.001; Table 4). When

all the populations of the Western Ghats were treated

as one region and the Sri Lankan as another region, the

AMOVA result showed that the percentages of variation

attributable to the differences between regions, among

populations within regions, and among individuals within

populations were 18.77% (P<0.001), 21.36%, ( P<0.001)

and 59.87% (P<0.009; Table 4).

Mantel test.

The mantel test did not reveal any

correlation between fruit morphological characters and

genetic distance (r = 0.0156; p = 0.4800) or between the

presence/absence of fruit scales and genetic distance (r =

0.0974; p = 0.6050). Geographical and genetic distances

were significantly correlated (r = 0.7327; p = 0.0090).

The unrooted phenetic tree (Figure 2) constructed based

on the genetic distance showed a clear distinction between

the Indian and Sri Lankan populations. Two clusters were

observed within the Indian populations (between Honaver/

Chertala and between Nilambur/Ashramam). Among the

Sri Lankan populations, Yagirala and Watareke formed a

single cluster, which in turn was linked to the Matugama

population.

Cluster analysis.

The UPGMA dendrogram,

constructed based on Jaccard¡¦s similarity coefficients,

depicted the genetic clustering of all the sixty genotypes

analysed (data not shown). The cophenetic correlation

coefficient between the dendrogram and the original

distance matrix was 0.84. Three major clusters were

observed in the dendrogram. The first cluster consisted

of the individuals Honaver and Chertala with a similarity

of 0.57. The second cluster associated the individuals of

Ashramam and Nilambur with a similarity of 0.56. Two

individuals of Chertala are clustered with individuals of

Ashramam and Nilambur. The third cluster consists of

entire Sri Lankan populations, which share a similarity of

0.61. The individuals of Watareke and Matugama were

clustered first (level 0.72) and then joined with Yagirala at

a level of 0.66.

Principal co-ordinate analysis (PCOA).

The

PCOA chart (data not shown) clearly separated Sri Lankan

populations from Indian populations. The first coordinate

axis comprised the Nilambur and Chertala populations,

and the populations of Honaver and Ashramam were

concentrated in the second axis. The first coordinate

accounted for the 13% of total variance while the second

and third accounted for 11.2% and 6.3%, respectively.

DISCUSSION

The multivariate analyses of morphological and

molecular data indicate that all the samples belong to a

single species. In the RAPD and morphometric analysis,

Table 3. Varia tion e xpla ine d by the firs t th ree p rinci pal

coordinates.

Axis 1 Axis 2 Axis 3

Eigen values

9.14708 4.92040 2.9357

Percent of variance 38.5846 22.2513 9.1743

Cumulative percentage 38.5846 60.8359 69.010

Table 4. Hierarchical analysis of molecular variance (AMOVA).

Variance component

d.f

SSD

MSD Variation variance %

P

One region

Among population

6

405.05 67.50

6.49

34.26% <0.001

Within population

53

660.60 12.46

12.46

65.74% <0.001

Two region

Region

1

153.75 153.75

3.90

18.77% <0.001

Among population

5

251.30 50.26

4.44

21.36% <0.001

Within population

53

660.60 12.46

12.46

59.87% <0.999

Figure 2. Unrooted phenetic tree for all populations bas ed

on genetic distance values produced from F ITCH method of

PHYLIP (Felsenstein, 1993). Numbers at nodes indicate the

bootstrap values.

448

Botanical Studies, Vol. 47, 2006

if the populations sampled had been two distinct species,

the unrooted phenetic tree and UPGMA cluster analysis

would have separated them into two distinct groups. In the

RAPD dendrogram, populations of Nilambur and Honaver

representing C. metzianus mixed with populations of

C. rivalis. Moreover, the RAPD analysis was unable to

identify any species-specific markers. Morphologically,

the distinction between C. metzianus and C. rivalis is

that the latter has larger fruits and a fruit scale channel

in the middle. The other morphological features are

shared by both species. The Mantel test results indicated

overall independence between the fruit morphological

data and the genetic distances measured using RAPD

analysis. The lack of correlation between the genetic

distance and the presence/absence of fruit scales was

further confirmed by the UPGMA dendrogram and PCO

analysis, suggesting further that both C. rivalis and C.

metzianus cannot be distinguished on the basis of fruit

scales. Hence, the morphometric studies revealed a high

degree of morphological similarity among populations

and that no distinct phenetic gap existed between the

two species; rather, the morphological characters were

found to overlap. The OTUs from Nilambur and Honaver

corresponding to two different populations of C. metzianus

were found to be intermixed with the remaining OTUs

representing different populations of C. rivalis. The lack

of distinct phenetic structure in ordination analyses also

indicates a high similarity between these two species.

Therefore, the results of both morphometric and RAPD

analyses did not provide any support to separate C.rivalis

and C. metzianus into distinct species. Hence, C. rivalis

should be merged to C. metzianus.

Genetic diversity

Genetic richness can be assessed by estimating

the genetic diversity parameters (viz. percentage of

polymorphic loci and gene diversity index) (Yeh, 2000).

The percentage of polymorphic loci was 95%. The high

levels of genetic polymorphism have been documented in

other rattans (Bon et al., 1996) and palms such as Euterpe

edulis (Cardoso et al., 2000). In this study, the amount of

gene diversity obtained was found to be 0.18, which is

similar to figures reported for outcrossing, herbaceous, and

insect pollinated species (Cardoso et al., 2000). Among the

seven analysed populations, the genetic diversity measures

were highest in the Chertala population (0.25), followed

by Nilambur (0.23) and Ashramam (0.21). The higher

gene diversity of the Chertala population may be due to its

high population density. Hence, an evolutionary force like

genetic drift can operate on a large number of genotypes,

resulting in the maintenance of a great of amount of gene

diversity. This population can be considered a hot spot of

genetic variation and an important reservoir of potentially

useful genes, and it hence merits a high priority by

conservation managements.

The results of AMOVA on RAPD data showed that

the variance component among populations was 34.26%,

and the variance component within populations was

65.74% of the total variation when all the populations

were treated as one region. A similar pattern was

also observed in the studies of Grevillea barklyana

(Hogbin et al., 1998), and E. grandis (Grattapaglia et

al., 1997). Studies on pines (Hamrick and Godt, 1989)

and Santalum album (Suma and Balasundaran, 2003),

on the other hand, indicated a reverse trend, in which

most of the genetic diversity was found to reside among

populations rather than within them. Autogamous species

are generally supposed to allocate most of their genetic

variability among populations, which results in little

intra-populational genetic variability. This is the case

with Oryza (Buso et al., 1998), and Gentianella (Fischer

and Matthies, 1998), though Hordeum spontaneum is an

exception (Baum et al., 1997). Long-lived, woody, late-

successional organisms typically harbor the highest levels

of genetic variation within populations (Hamrick and

Loveless, 1989; Hamrich and Godt, 1989). In this study,

as populations were divided into two regions, with all

the Western Ghats populations in one region and the rest

grouped in another, the results of AMOVA indicated that

the percentages of variation attributable to the differences

between regions, among populations within regions, and

among individuals within populations were 18.77%,

21.36% and 59.87%, respectively. The breeding system is

one of the main factors determining the genetic structuring

of plant populations (Hamrick and Godt, 1989), and the

considerable amount of intra-populational variation in

these Calamus populations reflects its allogamous mode of

reproduction.

The seven populations in this study covered a wide

geographic range in latitudes and geographical distances.

The most genetically similar populations (Yagirala and

Watareke) were geographically separated by a distance

of only 25 km while the genetically distant populations

(Honaver and Yagirala) were separated by about 980

km. The test of correlation between the genetic and

geographic distance matrices using the mantel test

revealed a significant positive correlation with r = 0.7327

(p = 0.0090), which was partly supported by the UPGMA

dendrogram and PCO analysis. Both analyses revealed the

same tendency for the individuals to group according to

geographic localities.

Diminishing forest areas coupled with overexploitation

has seriously affected the wild populations of rattan,

and its conservation is a major concern. Detailed

knowledge about the genetic diversity within and among

populations is essential to developing gene banks for ex-

situ conservation. The populations such as Honaver,

Matugama, and Yagirala analyzed in the present study are

under continuous threats of degradation, mainly due to

habitat alteration and irregular harvesting of stems for the

furniture and handicraft industries. As a preliminary step in

the ex-situ conservation of rattans, germplasm collections

and seed stands representing all the populations have to be

maintained.

SREEKUMAR et al. ¡X Morphological and RAPD analyses of

Calamus

449

CONCLUSIONS

The present study based on random amplified

polymorphic DNA analysis and morphometric

studies shows that C. metzianus and C. rivalis are

indistinguishable. Since C. metzianus was reported first,

this name will stand. Apparently, RAPD fingerprints

can contribute significantly to our understanding of the

diversity within the different populations of C. metzianus

and C. rivalis.

Taxonomic treatments

Calamus metzianus Schlecht in Linnaea 26: 727. 1853;

Becc. in Hook. f., Fl. Brit. India 6: 462. 1892; Becc.

in Ann. Roy. Bot. Gard. Calcutta 11: 82, 221. t. 67.

1908. Renuka, Rattans of Western Ghats 53, t. 16.

1992 [Figure 6.10; Plate 4]¡XTYPE: India, Karnataka,

Mangalore, 1847, R.T. Hokanaker, 1906 (holotype,

FI).

Calamus rudentum Mart., Hist. Nat. Palm. 3: 340.

1823-1853.

Calamus rivalis Thw. ex Trim. in J. Bot. 23: 268 (1885),

syn. nov.

Clustering, slender cane, climbing to a height of 10-15

m. Stem diameter with sheath 1-2 cm, without sheath to

1 cm. Leaf sheath green, armed with solitary, yellowish,

horizontal or slightly deflexed spines to 10 mm long;

knee present; ochrea short, deciduous. Leaves 1.5 m

long; petiole short or absent; rachis acutely trigonous,

bifaced and naked above; leaflets 40 ¡Ñ 2 cm, regular,

linear, acuminate to a slender apex, lanceolate, on the

upper surface mid vein with out spines or spinulose at

the apex, ciliate beneath from centre upward, margin

spinulose, upper leaflets small, bristly, terminal pair free

at base. Inflorescences flagellate; male inflorescence 3 m.

long; partial inflorescence 20-30 cm long, rachillae 3 cm

long; female inflorescence 3 m long, partial inflorescence

70 cm long, rachilla 5-10 on each side, terminating

in a filiform caudate Appendix, 8 cm long, primary

spathe very long; narrow, closely sheathing, armed with

numerous reflexed scattered spines, obliquely truncate

at the mouth, prolonged in to a short triangular point;

secondary spathe very narrow tubular infundibuliform;

2.5 cm length, spiny and narrow at base; male flowers 4

mm long, acute, calyx veined; female flowers 3 mm long,

ovate; involucre cupular, unequal margin; fruiting perianth

not pedicelliform. Fruit very broadly ovoid, 7-8 mm

broad and 11 mm long; scales on 21 vertical rows, very

faintly channeled, straw yellow, violet when ripe, margin

erosely toothed; seed ovate, 9 mm broad and 5 mm thick

Endosperm not ruminate; embryo basal.

Distribution. India, Sri Lanka. Restricted to plains

along the backwaters and coasts and in sacred groves.

Phenology. Flowering: September-October; Fruiting:

March-April.

Specimens examined. INDIA. KERALA: Malappuram

Dist., Pattakkarimbu, Nilambur, 1988, Babu 4081, 4082,

4083 (KFRI); Nilambur, Perrie 49362, 49363, 49364,

49365, 49366 (CAL); Nilambur, 1984, Renuka 3061,

4031, 4032, 4033 (KFRI); Nilambur, 2002, Sreekumar

8451, 8452 (KFRI). Kollam Dist., Ashramam compound,

1999, Renuka & Sasidharan 3443 (KFRI); Ashramam,

2001 Sreekumar 22446 (KFRI). KARNATAKA: Uttara

Kannada Dist., Honaver, Renuka 5884 (KFRI); Honaver,

Renuka 5884 (KFRI); Honaver, Renuka 7555 (KFRI);

Honaver, 2000, Anto & Sreekumar 22401 (KFRI);

Honaver, 2002, Sreekumar 8444 (KFRI); SRI LANKA:

Watareke, Neela de Zoysa 6814 (KFRI).

Acknowledgements. We are grateful to Dr. J.K. Sharma,

Director, KFRI, for providing the facilities for this work.

Funding was provided by the Ministry of Environment and

Forests, Government of India. We express our gratitude

to the Kerala and Karnataka Forest Department for

helping with the fieldwork. Special thanks to K. Dinesh

for his indispensable assistance in the field. We sincerely

thank Dr. B.M.P. Sinhakumara, Department of Forestry,

University of Sri Jayawardhanapura and his students for

collecting the required Sri Lankan rattan specimens and

sending them promptly to the Kerala Forest Research

Institute.

LITERATURE CITED

Beccari, O. 1908. Asiatic palm s Lepidocaryeae. Part I. The

species of Calamus. Annals of the Royal Botanic Garden,

Calcutta.

Baum , B.R., E . Nevo, D.A Johns on, and A. Beile s. 1997.

Genetic diversity in wild barley (Hordeum spontaneum C.

Koch) in the Near East: a molecular analysis using random

ampl ifi ed polymorphic DNA (RAPD) ma rke rs . Ge ne.

Resources Crops Evol. 44: 147-157.

Bon, M. C., B.A.H. B ars i , and H.I. Holy. 1996 . Gene ti c

Variability of two Rattan Species from Isozyme Markers.

Proceedings of IUFRO Syposium, Chiangimai, Thailand.

Buso, G.S.C., P.H. Rangel, and M.E. Ferreira. 1998. Analysis of

genetic variability in South American wild rice populations

(Oryza glumaepatula) with isozymes and RAPD markers.

Mol. Ecol. 7: 107-117.

Cardoso, S.R.S., N.B. Eloy, J. Proven, M.A. Cardoso, and P.C.G.

Ferreira. 2000. Genetic differentiation of Euterpe edulis

Mart. Populations estimated by AFLP analysis. Mol. Ecol.

9: 1753-1760.

Doyle, J.J. and J.L. Doyle. 1990. Isolation of DNA from fresh

plant tissues. Focus 12: 13-15.

Excoffier, L., P.E. Smouse, and J.M. Quattro. 1992. Analysis of

molecular variance inferred from metric distances among

DNA haplotypes: application to human mitochondrial DNA

restriction data. Genetics 131: 479-491.

Felsenstein, J. 1993. P HYLIP (Phylogeny Inference Package)

450

Botanical Studies, Vol. 47, 2006

Version 3.5. Distributed by the author. Department of

Genetics, University of Washington, Seattle, WA.

Fischer, M. and D. Matthies. 1998. RAPD variation in relation

to population size and plant fitness in the rare Gentianella

germanica (Gentianaceae). Amer. J. Bot. 85: 811-819.

Grattapaglia, D., A.Y. Clam pi, R, Collevatti, A.M.M. Reis,

B.M. Walter, F.B. Gandara, and R.P.V. Brondani. 1997.

Development and application of genomic tec hnologies

for genetic conservation of forest trees. Proceedings, IF S

workshop on "Recent advances in biotechnology for tree

c onse rva tion and m anage ment ", 15 to 19 Se ptem ber,

Florianopolis, Brazil.

Hamrick, J.L and M.D. Loveless. 1989. The genetic structure

of tropical tree populations: association with reproductive

biology. In J . Bo c k a nd Y.B . L i nh a rt ( e ds .), T h e

Evolutionary Ecology of plants. West view press, Boulder,

Co., pp. 130-146.

Hamrick, J.L. and M.J.W. Godt. 1989. Allozyme diversity in

plant species. In A.D.H. Brown, M.T. Clegg, A.L. Kahler,

and B.S. Weir (eds.), Plant Population Genetics, Breeding

and Germplasm Resources. Sinauer, Sunderland. MA., pp.

43-63.

Hogbin, P.M., D.J. Ayre, and R.J . W helan. 1998. Ge netic

va r ia t io n a nd re p rod uc t i ve s u c ce s s of ro ad v e rge

populations of the rare shrub Grevillea barklyana

(Proteaceae). Heredity 80: 180-186.

J accard, P. 1908. Nouvelles recherche s sur la s dis tribution

florale. Bull. Soc. Vand. Sci. Nat. 44: 223-270.

Ne i, M. 1 973. An al ys is of ge ne dive rs it y in s ub divi de d

populations. Proc. Natl. Acad. Sci. USA 70: 3321-3323.

Nei, M. 1978. Estimation of average heterozygosity and genetic

distance from a small number of individuals. Genetics 89:

583-590.

Pabuayan, I. 2000. Addressing rattan technology needs for Asia.

Paper presented at the XXI IUFRO World Congress 2000.

Kuala Lumpur, Malaysia.

Renuka, C. and K.M. Bhat. 1987. On the occurrence of Calamus

metzianus Schlecht in Kerala. Curr. Sci. 56: 313-314.

Rolf, J. 2000. NTSYSpc version 2.0 Exeter Software, Setauket,

New York.

S neath, P.H.A. and R.R. Sokal. 1973. Numerical Taxonomy.

Freeman, San Francisco.

Suma, T.B. and M. Balasundaran. 2003. Isozyme variation of

five sandal provenance in India. Aust. J. Bot. 51: 243-249.

Uhl, N.W. and J . Drans field. 1987. "Genera P alm arum; a

classification of palms based on the work of Harold, E.,

Moor E.Jr." Allan Press, Lowrence, KS.

Walter, K.S. and H.J. Gillett. 1998. IUCN Red list of threatened

plants. IUCN, Gland and Cambridge.

Williams, J.G.K., A.R. Kubelik, K.J. Livak, J.A. Rafalski, and

S .V. Tingley. 1990. DNA polym orphis ms amplified by

arbitrary primers are useful as genetic markers. Nucl. Acids

Res. 18: 6531-6535.

Yeh, F.C. 2000. Population genetics. In A. Young, D. Boshier,

a nd T. Bo yle . (eds .), F ore st Cons erv ati on Gene ti cs .

Pr inc ipl es a nd P ra ct ice . CS IR O P ubli s hing . Unit ed

Kingdom.

Yeh, F.C., R. Yang, and T. Boyle. 1999. POP GENE Version

1.31: Microsoft Windows based Freeware for Population

Genetic analysis . A quick User¡¦s guide. A J oint Project

Developed by Centre for International Forestry Research

and University of Alberta, Canada.

SREEKUMAR et al. ¡X Morphological and RAPD analyses of

Calamus

451

Appendix. List of qualitative and quantitative characters used in the study. (Characters abbreviations included in the bracket)

No. Character

Measurements/Status

1. Stem diameter (stemdia)

cm

2. Internodal length (intle)

cm

3. Leafsheath spine length (lesle)

cm

4. Leafsheath spine (lesp)

Sparse (0); Dense (1)

5. Leafsheath spine hairs (lesph)

Absent (0); Present (1)

6. Petiole (peole)

Absent (0); Present (1)

7. Leaf length (lflen)

m

8. Leaflet length (lfletl)

cm

9. Leaflet width (lfwid)

cm

10. Leaflet upper midvein cilia length (lfumcl)

cm

11. Leaflet lower midvein cilia length (lflmcl)

cm

12. Terminal leaflet length (terleng)

cm

13. Terminal leaflet width (terlwid)

cm

14. Partial Inflorescence length (parlen)

cm

15. Number of male rachilla/partial inflorescence (nomlrach)

Number

16. Number of female rachilla/partial inflorescence (noflrach)

Number

17. Number of fruits per rachillae (nofrurach)

Number

18. Male rachillae length (mlrachl)

cm

19. Female rachillae length (flrachl)

cm

20. Inflorescence caudex length (infcaudel)

cm

21. Rachillae prophyll length (rachprol)

cm

22. Rachillae prophyll width (rachprowid)

cm

23. Fruit apical beak length (fruapl)

cm

24. Fruit diameter (frudia)

cm

25. Fruit length (frulen)

cm

26. Fruit scales length (fruscal)

cm

27. Fruit scales width (fruscalwi)

cm

28. Fruit scales margin (frusclmar)

Reddish-Brown (0); White (1)

29. Number of vertical rows of scales in fruit (novrowscl)

Number

30. Fruit scales channel (fruscales)

Absent (0); Present (1)

31. Seed length (sedlen)

cm

32. Seed width (sedwid)

cm

452

Botanical Studies, Vol. 47, 2006