115
Cey. J. Sci. (Bio. Sci.) 35 (2):115-136, 2006
A COMPARATIVE WOOD ANATOMICAL STUDY OF THE GENUS
DIOSPYROS L. (EBENACEAE) IN SRI LANKA
B. K. L. Wickremasinghe1* and Tissa R. Herat2
Department of Botany, The Open University of Sri Lanka, Nawala, Nugegoda, Sri Lanka.
2
Faculty of Applied Sciences, The Rajarata University of Sri Lanka, Mihintale, Sri Lanka.
Accepted: 03 August 2006
1
ABSTRACT
A comparative wood anatomical study of 28 Diospyros L. species described by earlier workers to be
present in Sri Lanka was carried out. The diagnostic features of wood anatomical characters of the taxa
concerned were evaluated by employing standard statistical methods such as Principal Component Analysis,
Cluster Analysis and Discriminant Analysis. The Sri Lankan fraction of the genus revealed homogeneity with
respect to their wood anatomical characters with a dependency range on environmental factors such as water
availability. Further based on the wood anatomical characters an attempt was made to establish levels of
specialization in order to understand the evolutionary relationships that exist within the species concerned.
Even though the evolutionary relationships within the genus are vague, D. malabarica and D. atrata could be
considered as possible ancestral forms from which the Sri Lankan endemic species of the taxon evolved.
Key words: Diospyros, comparative wood anatomy.
INTRODUCTION
The history of the “ebony” woods goes back to
the Tutankhamun’s period, the most famos
Egyptian pharaoh ca. 3300 year before (Hora,
1981). Further, it is reported that, Sri Lanka (then
Ceylon) was reputed for the best quality “ebony”
woods throughout the world from very ancient times
(Alston, 1931; Kostermans, 1977). However, proper
scientific names of such taxa have been dated from
the period of Dutch occupation in parts of Sri Lanka.
For instance, Hermann (1717) reported a species,
Highulhaenda [now known as D. ferrea (Willd.)
Bakh.à which Burmann (1737) recognized as
Highulhaenda folio myrti. Further, Linnaeus (1747)
cited Higulhaenda folio myrti of Burmann in his
Flora Zeylanica which according to Kostermans
(1981) in the present concept is D. ferrea (Willd.)
Bakh. Thwaites (1864) described 11 new species of
Diospyros L. from Sri Lanka namely, D. acuminata
Thw., D. acuta Thw. , D. affinis Thw., D. attenuata
Thw., D. insignis Thw., D. moonii Thw., D.
oblongifolia Thw. , D. oocarpa Thw. , D.
oppositifolia Thw., D. quaesita Thw. He recognized
two varieties within D. embryopteris [now known as
D. malabarica (Desr.) Kostel] var nervosa and
atrata which were considered endemic to Sri Lanka
and also recognized and described a new species,
Macreightia oblongifolia Thw. of the genus
Macreightia of De Candolle (1844), which
Kostermans (1981) later named as Diospyros
oblongifolia (Thw.) Kosterm. Further, Thwaites
(1864) recognized three varieties of Maba buxifolia
var. ebenus Thw., var. microphylla Thw., and var.
*Corresponding author’s e mail: bkwic@ou.ac.lk
angustifolia Thw. and he was of the opinion that
these three varieties were closely related or
connected together by intermediate forms thereby
representing a variable species with broad limits.
Kostermans (1981) named these varieties as
Diospyros
ebenoides
Kosterm.,
Diospyros
nummulariifolia
Kosterm.
and
Diospyros
rheophytica Kosterm. respectively.
Trimen (1893) recognized two genera, Maba
Forst. and Diospyros L., and incorporated the then
known taxa of Macreightia DC. within the genus
Maba Forst. Wright (1904) published a
comprehensive monograph of the Sri Lankan
Diospyros L. and his work became a land mark
study for the genus in Sri Lanka. He described
morphological, anatomical and other taxonomical
characters of 20 representative species of Diospyros
L. in Sri Lanka. However, the genus Maba Forst.
was excluded from the Ebenaceae. Worthington
(1959) recognized 13 species of Sri Lankan
Diospyros L., based on their morphological
characters, of which three species were considered
endemic to Sri Lanka, viz., D. crumenata Thw., D.
quaesita Thw. and D. ovalifolia Wight. However, D.
ovalifolia Wight is also reported from South India.
(Kostermans, 1981; 1977) described two new
endemic species, D. koenigii Kosterm, a species
found only in Gannoruwa forest in Kandy District,
and D. zeylanica Kosterm from Kanneliya forest
near Hiniduma, Galle District.
However,
Kostermans (1981) in the Revision of the Ceylon
Flora excluded this latter taxon stating that this
B. K. L. Wickremasinghe and Tissa R. Herat
plant represents probably a Stemonoporus
(Dipterocarpaceae). Kostermans (1977) incorpora- ted other taxa recognized by Thwaites as Maba and
Macreightia within Diospyros as described in first
paragraph. In addition Kostermans (1981), described
yet another variety of D. insignis, D. insignis var.
parvifolia. Further he included Diospyros discolor
Willd. (D. blancoi A.DC) [Velvet apple (E)], a
species from the Philippines, now occasionally
cultivated and naturalized in parts of the island,
within the Flora of Sri Lanka totaling the number of
species to thirty two. Jayasuriya (1998) added a
further endemic species, D. pemadasai.
All the revisions of the genus mentioned above
have been based mainly on field observations and
on morphological characters. Apart from the
anatomical observations made by Wright in 1904
with then available facilities, no major anatomical
study has been undertaken to determine the species
limits to gain a better understanding of the Sri
Lankan fraction of the genus.
The present study is an attempt to investigate
the wood anatomical characters and ecological
parameters, with the idea of understanding the
species limits of the Sri Lankan taxa of the genus
Diospyros L. and comparing such findings with
finding of other workers.
MATERIALS AND METHODS
Fresh wood samples of 28 species (Table 1),
out of the 32 reported by Kostermans (1981) were
collected from the field along with voucher
herbarium specimens in triplicate. Collection of
specimens was made to include wet, dry and
intermediate zones. Further, species which were
reported to be strictly confined to certain places
were collected (Fig. 1). Four species were excluded
since they were reported to be very rare or extinct
(Kostermans, 1981). Collected specimens were
identified by consulting Trimen (1893), Kostermans
(1981) and comparing with the specimens deposited
at the National Herbarium, Royal Botanic Gardens,
Peradeniya (PDA). Wood samples were taken from
straight branches as far as possible, to avoid
inclusion of tension or pressure wood. An attempt
was made to collect mature wood with maximum
diameter whenever possible.
Collected wood samples were cut into pieces of
ca. 1 x 1 x 1 cm cubes and preserved in 70%
alcohol. Radial, tangential and transverse sections
were taken using a Sliding Microtome (E. Leitz
Wetzla, Germany) at thickness ranging from 10 - 15
µm., stained in Safranin and Fast Green FCF series
116
following Sass (1958) and mounted in Canada
balsam.
Wood pieces of the size of match sticks were
macerated using a solution of 30% hydrogen
peroxide, distilled water and glacial acetic
acid(1:4:5), then washed in running water for at least
24 hours, dehydrated, stained and mounted in
Canada balsam following Johansen’s method
(1940).
Characters of wood for comparative anatomy
were selected according to Tippo (1964) and
Metcalf and Chalk (1989). Length of the vessel
elements was measured including the tails, based on
Chalk and Chattaway (1934). Tangential vessel
diameters, vessel wall thicknesses and vessel
frequencies were measured using cross sections.
Individual pores were counted following Wheeler
(1986). Measurements of vessel frequencies were
based on 25 counts in an area of 2.8 mm2 field of
view.
Ray types were described according to Kribs
(1935). The terminology used for microscopic
features was based on the International Association
of Wood Anatomists’ list of microscopic features for
hardwood identification (IAWA Committee, 1989).
Tangential wood sections were used to measure ray
heights and ray frequencies while the lengths of
fibers were measured using macerated tissues.
Relative cell sizes and vessel frequency are given in
accordance with Chattaway (1932).
For quantitative data macerated wood, cross
sections, radial sections and tangential sections
were measured under the low power (x 10) and high
power (x 40) of an Olympus PM – 10 DAD light
microscope.
Annual rainfall and mean annual temperatures
were obtained from the Meteorological Department,
Colombo. Altitudes of the areas where specimens
were collected were determined using the altimeter
supplemented with topographic sheets of the Survey
Department (1992). Agro-ecological zones (Fig. 1)
identified were based on the soil map of Sri Lanka
(Irrigation Department, 1988).
Statistical Analysis
The basic statistical analysis was done
according to procedures outlined by Snedecor and
Cochran (1967) and Sokal and Rohlf (1981). For
each sample from each species a mean value with
95% confidence interval of wood anatomical
features was calculated. These values were tabulated
with the species description citing mean ranges of
Wood anatomical study of Diospyros
117
Figure
1.
Distribution of Diospyros in Sri Lanka [modified after Kostermans ( 1981) ]. D. acuminata (2) D. acuta
(3) D. affinis (4) D. atrata (5) D. attenuata (6) D. chaetocarpa (7) D. discolor (8) D. ebenoides (9) D. ebenum
(10) D. ferrea (11) D. hirsuta (12) D. insignis (13) D. insignis var. parvifolia (14) D. koenigii (15) D.
malabarica (16) D. melanoxylon (17) D. montana (18) D. moonii (19) D. nummulariifolia (20) D.
oblongifolia (21) D. oocarpa (22) D.oppositifolia (23) D.ovalifolia (24) D. quaesita (25) D. racemosa (26) D.
rheophytica (27) D. thwaitesii (28) D. walkeri.
B. K. L. Wickremasinghe and Tissa R. Herat
species and maximum and minimum intervals
within parenthesis. For statistical analysis, mean
values of each species were tabulated. Mean values
derived were then used for fundamental statistical
analysis such as mean comparison and analysis of
variance. Further, Pearson product movement
correlation was sought for the quantitative data to
find out any interdependencies of anatomical
characters. The same statistical techniques were
employed to trace the relationships between
ecological parameters such as altitude, mean annual
rainfall and mean annual temperature. The
correlation obtained was used to calculate r2
(coefficient of determination) to find dependency of
the anatomical features over ecological parameters.
The extent of influence of ecological parameters on
wood anatomical features was sought in order to
ascertain the degree of dependency between the
two. The collected specimens were assigned to the
climatic zones following Mueller-Dombois (1968)
and further assigned to agro-ecological groups
based on the soil map of Sri Lanka (Irrigation
Department, 1988). Subsequently, mean vessel
element lengths and mean fiber lengths were
plotted separately with agro-ecological zones to
detect whether there are any ecological patterns
reflected. The codes used in investigating
quantitative and qualitative characters and
numerical codes used in qualitative analysis are
given in Tables 2.A and 2.B.
Multivariate analytical techniques, adopted by
Sneath and Sokal (1973), Gauch (1986), Burley
and Miller (1989) were attempted for further
analysis. Principal Component Analysis (PCA) and
Cluster Analysis (CA) were attempted in order to
examine the taxonomic patterns of the data and to
generate a classification system. In the course of
statistical analysis, the species were coded with a
number and an acronym as shown in Table 1.
The original data were converted into
standardized scores i.e. Z-scores (with 0 mean and
standard deviation of 1) in order to give an equal
weight to all the measurements. Phenetic similarity
or dissimilarity was also calculated.
The mesomorphy index (Carlquist and De
Buhr, 1977) was calculated for the specimens since
this index is reported to be related to the habitat
preference. Another index, conductivity suggested
by Carlquist (1984), was also calculated for each
specimen because this index is believed to be a
measure of vulnerability of vessels to extreme
tensile forces created by extreme drought and cold
conditions. Mesomorphy index and conductivity
values were calculated by using the following
118
equations:
Mesomorphy =
Conductance =
Vessel length x Vessel diameter
──────────────
Vessels per mm2
(Vessel diameter)4 x 0.0001
──────────────
Vessels per mm2
All the computations and statistical analyses
were done using the statistical program package
SPSS/PC+ Statistics TM 4.0 and Windows version 6,
(SPSS Inc., 1993).
RESULTS AND DISCUSSION
The occurrence of growth rings was not
consistent among the species of
Sri Lankan
fraction of Diospyros L. Certain species showed
very faint and discontinuous growth rings marked
only by rather thick-walled fibers (Table 3).
However, a number of non-endemic dry zone,
intermediate and low country wet zone species
lack in growth rings meanwhile some non-endemic
dry, intermediate species and endemics of low
country wet zone species showed discontinuous
growth rings indicating that there was no
considerable relationship between endemicity or
with the ecological origin of the species with
growth ring formation. The occurrence of growth
rings in tropical woods is, generally, believed to
depend on the environmental factors such as
photoperiod, temperature and availability of water
(Vliet, 1979; Kramer, 1964; Fahn et al., 1985).
Therefore, it is evident that fluctuations in
temperature, rainfall and altitudinal changes seem
to be of little influence on the presence or absence
of growth rings within the Sri Lankan species of the
genus Diospyros L.
Sri Lankan climate lacks the seasonality.
However, wet and dry spells are well-recognized
all over the island which is directly related to the
mean annual rainfall. Therefore, occurrence of
growth rings may be attributed to the soil moisture
availability which in turn influences the variation in
the plant growth.Further, whether the formation of
growth rings is associated with the photoperiodism
or is influenced by auxins as pointed out by Kramer
(1964) is yet to be ascertained and require in-depth
investigations.
All the wood examined was non-storied.
Individual vessel pores were oval to circular in
cross sectional view. The pores were solitary or
distributed in radially arranged multiples of varying
number of pores (Fig. 2a, d, g).
Wood anatomical study of Diospyros
119
Table 1. Species of Diospyros L. collected for the study and identified according to Kostermans (1981).
endemicity, zonal distribution and acronyms used in the study.
Species
D. acuminata (Thw.) Kosterm.
D. acuta Thw.
D. affinis Thw.
D. atrata Alston.
D. attenuata Thw.
D. chaetocarpa Kosterm.
D. discolor Willd.
D. ebenoides Kosterm.
D. ebenum Koenig.
D. ferrea (Willd.) Bakh.
D. hirsuta L.
D. insignis Thw.
D. insignis var. parvifolia Kosterm.
D. koenigii Kosterm.
D. malabarica (Desr.) Kostel
D. melanoxylon Roxb.
D. montana Roxb.
D. moonii Thw.
D. nummulariifolia Kosterm.
D. oblongifolia (Thw.) Kosterm.
D. oocarpa Thw.
D. oppositifolia Thw.
D. ovalifolia Wight
D. quaesita Thw.
D. racemosa Roxb.
D. rheophytica Kosterm.
D. thwaitesii Beddome
D. walkeri Wight
Endemicity
E
E
NE
E
E
E
NE
E
NE
NE
E
NE
E
E
NE
NE
NE
E
E
E
NE
E
NE
E
NE
E
E
E
Zonal Distribution
W
W
D,I
W
W
W
W,I
I
D,I
D,I
W
W
W
W
D,I
D
D
W
D
W
D,I
W
D,I
W
W
W
W
W
Acronym
Code
DACUM
DACUT
DAFFI
DATTR
DATTN
DCHAE
DDISE
DEBNO
DEBNU
DFERR
DHIRS
DINSI
DINVP
DKOEN
DMALA
DMELA
DMONT
DMOON
DNUMM
DOBLO
DOOCA
DOPPO
DOVAL
DQUES
DRACE
DRHEO
DTHWA
DWALK
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
E =Endemic, NE = Non-endemic, D = Dry zone, W = Wet zone, I = Intermediate zone
Within the dry zone species such as
D. montana Roxb., D. oocarpa Thw., D. affinis
Thw. and D. ovalifolia Wight, the pores
were arranged in multiples of 3-11 in a single
group (Fig. 2g).
In general, the dry zone
species possess higher number of vessel groups
per unit area than the intermediate and the wet
zone species.
Such observations agree with
Carlquist (1984) was states that the increased
number of groups of vessels indicates a xeric
tendency of the habitat of a species.
Metcalf and Chalk (1989) stated that, the
number of vessels per square millimeter could be
very significant as a taxonomic character only if
very low or very high numbers of vessels are
encountered. The present study reveals that the
non-endemic dry zone species posses higher
number of vessels per square millimeter than the
endemic wet zone species.
B. K. L. Wickremasinghe and Tissa R. Herat
120
Table 2. A. Codes used in the study for wood anatomical characters.
Character Code
Description
Quantitative characters
TVEL
VDIA
VWTH
DFPO
NGPA
NSPA
GSRT
FLEN
FDIA
RPMM
RHMM
RHCE
RWMM
RWCE
Vessel element length in µm
Vessel Diameter in µm
Vessel Wall Thickness in µm
Distribution frequency of pores/mm2
Number of Groups of Pores per area/mm2
Number of Solitary Pores per area/mm2
Groups to Solitary Pores ratio
Fiber length in µm
Fiber Diameter in µm
Rays per millimeter
Ray Height in millimeter
Ray Height in number of cells
Ray Width in millimeter
Ray Width in number of cells
Qualitative characters
VEDI
PERP
PITS
TYLO
GRRI
FIBR
RAUB
TVRP
PADS
CRYS
Vessel Distribution
Perforation plates of vessel elements
Pits whether minute or vestured
Tyloses present or absent
Absence or presence of Growth Rings
Fibers thick or thin
Rays, uniseriate or biseriate
Type of vessel-ray pits
Axial parenchyma distribution
Crystals, Present or Absent
However, such a relationship cannot be
generalized for all the dry and wet zone species
examined since some of the species, for instance D.
rheophytica, a species which grows along the banks
of rivers and rivulets in the low country wet zone,
showed a higher value that does not fit into place
along with the idea expressed above. Considering
the percentage values of vessel elements per area,
Wright (1904) attempted to correlate the higher
percentages of vessel elements with thin-leaved
trees and lower percentages with thick-leaved Sri
Lankan species of Diospyros. From an ecological
point of view, Wright (1904) was of the opinion
that vessel distribution frequencies of the dry zone
and the wet zone species could be related to the
habitat. The results of the present study agree with
Wright (1904) on the basis that the dry zone species
show high frequencies of vessel distribution per
area, compared with the wet zone.
Working with the Sri Lankan Diospyros,
Wright (1904) stated that there was no difference in
the vessel dimensions between species growing in
Mannar where more arid conditions occur,
compared with those growing in Adams Peak
Wilderness where the climatic conditions are quite
different, especially in relation to the moisture
availability. The tangential vessel diameters of the
wood samples studied ranged from 48.4 to 107.2
µm. On this basis, the observations of the present
study agree with Wright (1904). However, Metcalf
and Chalk (1989) stated that vessel diameter can
vary with samples which are taken from different
places of the same tree. Further, Aloni and
Zimmermann (1983) have pointed out that the size
and frequency of vessels vary along the plant axis
in response to concentration gradient of auxins.
Kramer (1964) showed that wood formation
depends to a great extent on the water availability
of the habitat. As such vessel element diameters are
of limited value in the delimitation of the taxa
under consideration. However, these could be used
to draw the generic limits of the Sri Lankan
Diospyros.
Wood anatomical study of Diospyros
Bailey (1957) and more recently Baas (1976;
1982) believed that the length of vessel elements
and other morphological characters such as
perforation plates and types of pits reflect the level
of specialization of a taxon and further recognized
the evolutionary trends of vessel element lengths
within angiosperm taxa which came to be known as
Bailian trends. Metcalf and Chalk (1989) also
stated that the vessel element length is more
significant as a measure of phylogenetic
specialization rather than as a diagnostic character
for a taxon. It is the general opinion of Bailey
(1957), Baas (1976; 1982) and Metcalf and Chalk
(1989) that the less specialized plant taxa have
121
longer vessel elements than the specialized forms.
On this basis, it is interesting to note that the nonendemic species of Sri Lanka have a shorter vessel
element length than the endemic species found at
wet and higher altitudes, while the taxa of the
intermediate zone exhibit vessel element lengths of
medium range. On this basis, the endemic
Diospyros species of Sri Lanka are less specialized
than the non-endemic taxa found in the dry zone.
Whether these observations indicate a retrogressive
evolutionary trend with respect to the Sri Lankan
endemic taxa is interesting, however, and needs
further investigation before conclusive decisions
are made.
Table 2. B. Code words and numerical codes used in qualitative analysis.
Character
(Abbreviations
in parentheses)
Character State
Code
Vessel Distribution (VEDI)
Exclusively solitary
Radial groups of 3-4
Radial or oblique
Tangential arrangement
Pore clusters
11
12
13
14
15
Perforation plate (PERP)
Oblique
Transverse
Minute
Vestured
Present
Absent
Present
Absent
Thick-walled
Thin-walled
Present
Absent
Homogenous
Heterogenous
Vertically oriented
Horizontally oriented
Similar to
Intervessel pits
11
12
11
12
11
12
11
12
11
12
11
12
11
12
11
12
Diffuse
Paratracheal
Apotracheal
Vertical
Present
Absent or very rare
11
13
12
14
11
12
Pits (PITS)
Tyloses (TYLO)
Growth Rings (GRRI)
Fibers (FIBR)
Trachieds (TRAC)
Rays (RAUB)
Type of Vessel-Ray Pits(TVRP)
Parenchyma distribution
(PADS)
Crystals (CRYS)
13
B. K. L. Wickremasinghe and Tissa R. Herat
122
Figure 2. Wood anatomical features of the genus Diospyros (a) Cross sectional view of D. ebenum x
105. (b) Radial longitudinal section of D. ebenum x 660, (c) Tangential longitudinal section of D.
ebenum x 660, (d) Cross sectional view of D. moonii x 105, (e) Radial longitudinal section of D.
moonii x 660, (f) Tangential longitudinal section of D. moonii x 660, (g) Cross sectional view of D.
montana x 105, (h) Radial longitudinal section of D. montana x 660 and (i) Tangential longitudinal
section of D. montana x 660.
The thickness of vessel element wall ranged
from 4.32 - 8.65 µm. The vessel wall thickness of
the dry zone species is greater than that of Wet
zone areas. Whether such characters observed in
the dry zone taxa in relation to the water
availability is an adaptive measure of conservation
of water resources, needs further studies. Further, it
is interesting to note that D. rheophytica Kosterm, a
species restricted to the wet zone river banks, has
the highest vessel wall thickness among the Sri
Lankan species. Thus it is possible that further
investigations on these lines may show a
relationship between the vessel wall thickness and
the availability of water.
The terminology of wood fibers is
controversial and yet to be resolved (Carlquist,
1986a; 1986b; Baas 1986). However, Metcalf and
Wood anatomical study of Diospyros
Chalk (1989) stated that non-septate fibers do not
possess many features of taxonomic interest while
septate fibers are often characteristic of large
families and are useful as a taxonomic character.
The wood fibers observed in the study are nonseptate and ranged from 376.00 (302.64 - 444.36)
to 2055.00 (1564.66 - 2545.39) µm in length. The
distribution of fiber length shows that the fiber
length alone as a taxonomic character cannot be
related to the level of specialization within the
123
genus studied. An index derived from the ratio
between fiber length and vessel element length
approaches to one in less specialized woods while
higher values indicate higher levels of
specialization in secondary xylem in angiosperms
(Baretta-Kuipers, 1976). Results indicated a general
tendency in which the non-endemic dry zone
species of the genus are more specialized than
thoseof endemic wet zone species (Table 3).
Table 3. Occurrence of growth rings and the ray histology and ratio between fiber length and vessel
element length in Sri Lankan species of Diospyros L.
Code
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
Species
Diospyros acuminata (Thw.)
Kosterm.
Diospyros acuta Thw.
Diospyros affinis Thw.
Diospyros atrata Alston
Diospyros attenuata Thw.
Diospyros chaetocarpa Kosterm.
Diospyros discolor Willd.
Diospyros ebenoides Kosterm.
Diospyros ebenum Koenig.
Diospyros ferrea (Willd.) Bakh.
Diospyros hirsuta L.
Diospyros insignis Thw.
Diospyros insignis var.parvifolia
Kosterm.
Diospyros koenigii Kosterm.
Diospyros malabarica (Desr.)
Kostel.
Diospyros melanoxylon Roxb.
Diospyros montana Roxb.
Diospyros moonii Thw.
Diospyros nummularifolia
Kosterm.
Diospyros oblongifolia (Thw.)
Kosterm.
Diospyros oocarpa Thw.
Diospyros oppositifolia Thw.
Diospyros ovalifolia Wight
Diospyros quaesita Thw.
Diospyros racemosa Roxb.
Diospyros rheophytica Kosterm.
Diospyros thwaitesii Beddome.
Diospyros walkeri (Wight) Guerke
Character
Growth Rings
Ray Histology
FVRAT
absent
absent
present
absent
present
absent
absent
present
present
present
absent
absent
uniseriate
uniseriate
uniseriate
uniseriate
uniseriate
uniseriate
uniseriate
uniseriate
uniseriate
multiseriate
uniseriate
multiser iate
1.6
2.1
2.3
2.5
2.0
2.0
1.9
2.6
2.4
2.5
1.9
2.1
present
present
multiser iate
uniseriate
2.0
2.2
absent
present
present
absent
uniseriate
uniseriate
uniseriate
multiseriate
2.8
2.3
3.2
1.8
present
uniseriate
2.1
present
present
present
present
present
absent
absent
present
absent
uniseriate
uniseriate
uniseriate
uniseriate
uniseriate
uniseriate
uniseriate
uniseriate
uniseriate
2.2
2.6
1.8
2.1
2.1
2.4
1.7
1.4
1.7
B. K. L. Wickremasinghe and Tissa R. Herat
The characteristic features of wood ray tissues
are of importance in depicting an evolutionary
sequence of specialization within angiosperms
(Kribs, 1935). Barghoorn (1940) investigated the
ontogenetic development and phylogenetic
specialization of rays within the angiosperms and
agreed with Kribs (1935) and stated that,
heterogeneous Type I of Kribs’ which consists of a
number of layers of procumbent cells with four
upright or square marginal cells occurred in the
extinct taxa of Benetitales, Pteridospermae,
Cycadales and the primitive dicotyledonous
species. On the other hand Kribs’ Heterogeneous
Type III which exhibits procumbent cells with a
single row of upright or square marginal cells is
considered as a specialized form of rays. However,
Barghoorn (1941) cautioned that when considering
variations of ray structure for taxonomic purposes,
ontogenic variations which might occur during
development stages, should also be taken into
account. Summarizing all the facts already
mentioned on rays, Metcalf and Chalk (1989)
further added that uniseriate rays are diagnostically
important because of their restricted occurrence and
that this feature is particularly characteristic of
certain taxa that are phylogenetically advanced but
it tends to be typical of individual genera rather
than particular species. The present study reveals
that, almost all the Sri Lankan Diospyros species
possess the exclusively uniseriate rays (Table 3)
and a few of them possess rays with multiseriate
portions as wide as the uniseriate portions (Fig. 2
c, f) and this may be used to split the genus in to
two broad groups. However, Wright (1904)
recognized three groups on the same basis, (a).
medullary rays are one cell in tangential breadth,
(b) medullary rays are more than one cell in
tangential breadth, and (c) presence of special
radial groups of parenchyma.
The present study partly agrees with that of
Wright (1904) on the basis that the Sri Lankan taxa
belong to exhibit the first two categories. However,
the third category with special radial groups sensu
Wright (1904) was not clearly distinguishable from
the findings of the present study. Based on these
lines, it is observed that the Sri Lankan species of
Diospyros under consideration have xylem rays of
the Kribs’ Heterogeneous Type I and or
Heterogeneous Type II. Kribs Heterogeneous Type
III which is considered to be the most advanced
type (Barghoorn 1940; 1941a), was not observed in
any Sri Lankan species studied. Further, Kribs’
Homogenous types of rays which are considered
more
evolutionarily
advanced
than
the
Heterogeneous types were also not observed. On
this basis, the Sri Lankan taxa cannot be considered
as a highly advanced group. With respect to the Sri
124
Lankan species of Diospyros the taxa posses Kribs’
Heterogeneous Type I rays as observed in the two
endemic species D. atrata Alston and D. quaesita
Thw., are less specialized than the taxa that posses
Kribs’ Heterogeneous Type II rays observed in the
non-endemic species D. affinis Thw., D. ebenum
Koenig., D. oocarpa Thw., and the endemics D.
attenuata Thw., D. ebenoides Kosterm., D. hirsuta
L., D. koenigii Kosterm., D. moonii Thw., D.
oblongifolia (Thw.) Kosterm., D. thwaitesii
Beddome.
However, six non-endemic taxa namely, D.
discolor Willd., D. malabarica (Desr.) Kostel., D.
montana Roxb., D. ovalifolia Wight, D. racemosa
Roxb., D. melanoxylon Roxb. and the six endemic
taxa namely, D. acuminata (Thw.) Kosterm., D.
acuta Thw., D. insignis var. parvifolia Kosterm., D.
oppositifolia Thw., D. rheophytica Kosterm. and D.
walkeri (Wight) Guerke possess both Kribs' ray
Types I and II. With respect to evolutionary
advancement, the above mentioned taxa could be
considered as intermediates. There seem to be no
significant evolutionary trends within the Sri
Lankan taxa of the genus with respect to
Heterogeneous ray Types of Kribs. The nonendemic taxa with Kribs’ ray Type II are restricted
to the dry zone while the endemics of the same
types are found in the wet zone. Further, the
endemic taxa with a mixture of Heterogeneous
Type of Kribs I and II are distributed in the wet
zone while the non-endemics mainly occur in the
dry zone. On this basis, the Heterogeneous ray
Types seem to have some relationship with
environmental factors which needs further
investigation.
Kribs in 1937 regarded the absence of axial
parenchyma as a primitive character (Metcalf and
Chalk, 1989).
Metcalf and Chalk (1989),
summarizing the work on axial parenchyma
distribution by Kribs further stated that the number
of cells in a single strand of parenchyma could be
considered as an index of level of specialization.
Based on the classification of axial parenchyma by
IAWA (1989), Sri Lankan Diospyros species
showed banded parenchyma of reticulate,
scalariform or marginal types only. Carlquist
(1975) considered the presence or absence of axial
parenchyma as an ecological variation rather than
any evolutionarily significant feature. On this basis
the presence of axial parenchyma within the Sri
Lankan taxa is of limited value in the delimitation
of the taxa concerned.
The vessel elements of all the species studied
possessed simple perforation plates, end-walls
transverse or slightly oblique, with short tails.
Wood anatomical study of Diospyros
Hence, it is clear that there is no value of such
characters in delimitation of the species. Intervessel
pits of all the species observed were minute,
alternate, vestured, circular or oval in shape (Fig. 3
a). Vessel-ray pits were also similar to intervessel
pits in size and shape throughout the ray cells (Fig.
3 c), in all the specimens examined except D.
ferrea (Willd.) Bakh., which possess large vesselray pits with much reduced borders and
horizontally or vertically oriented as well as with
normal pits (Fig. 3 d). On this basis, D. ferrea
(Willd.) Bakh. stands out from the other species of
Diospyros. Based on morphological characters
Kostermans (1981) maintained D. ferrea (Willd.)
Bakh. as a distinct species where as White (1993)
believed that, based on wood anatomical characters
125
this taxon should be placed under a separate
section of the genus Diospyros. According to the
differences found with respect to vessel-ray pits, it
could be suggested that D. ferrea (Willd.) Bakh. be
separated into a Section of the genus as suggested
by White (1993).
The formation of tyloses has been considered
as an indication of evolutionary primitiveness of
angiosperms (Bonsen and Kucera, 1990). Tyloses
were observed most of the species studied except
D. chaetocarpa, D. acuta, D. affinis, D. attenuata,
D. discolor, D. hirsuta, D. insignis, D. moonii and
D. oppositifolia. In general, the dry zone species of
Diospyros possess more tyloses than the wet zone
species.
Figure 3. Wood anatomical characteristics of the genus Diospyros, (a) Intervessel pits of D.
insignis x 1320, (b) Crystals in rays of D. ebenum. x 1320, (c) Vessel -ray pits of D. ebenum x
1320 and (d) Vessel- ray pits of D. ferrea x 1320.
B. K. L. Wickremasinghe and Tissa R. Herat
Chattaway (1955; 1956) and IAWA (1989),
have pointed out that, the absence or presence and
if present, the type and location of crystals are of
taxonomic importance. Prismatic or rhombohedral
crystals were observed in ray cells (Fig. 3 b) in all
the species of Sri Lankan Diospyros taxa studied.
The abundance of crystals varied within the
species as well as within specimens from the same
species. However, in D. attenuata Thw., D.
insignis var. parvifolia Kosterm., D. malabarica
(Desr.) Kostel., D. rheophytica Kosterm. and D.
walkeri (Wight) Guerke crystals were found very
rarely. Crystals were mostly chambered. The
occurrence of tyloses and crystals is of limited
value in delimitation of the Sri Lankan species of
the genus concerned.
Multivariate statistical analyses were applied
in the present study to trace the possible
relationships between anatomical and ecological
features. Sneath and Sokal (1973) and Sokal and
Rohlf (1981) have shown the importance
of multivariate statistical techniques in numerical
Taxonomy. Using multivariate techniques,
Jacobseni (1979), Robertse et al. (1980) Khidir
and Wright (1982) and Hedren (1990)
successfully solved the problems in variation
between Allium cernuum and Allium stellatum,
wood anatomy of South African Acacia,
systematics of Graminae and the African
complex of Justicia striata respectively. The
same approach has been made by Somaratne
and Heart (2001) and Pathirana and Heart
(2004) in establishing the relationships between
species of the genus Calophyllum and the genus
Garcinia and to elucidate the possible
relationships among the species of Diospyros in
the present study.
The dendrogram derived from Cluster
Analysis (CA) based on quantitative wood
anatomical features of the Diospyros L. species
showed no considerable grouping tendency within
the genus (Fig. 4). This finding indicates that
taking wood anatomical features one at a time and
taking as whole characters simultaneously led to
different conclusions. However, based on the
results of cluster analysis, it could be stated that
wood anatomical features seem to be of limited
value in delimitation of taxa within the genus
Diospyros L. if they are treated individually.
Further analysis with other vegetative anatomical
data from leaf, young stem and node might be of
importance.
In the Principle Component Analysis (PCA),
the characters such as pore distribution frequency
126
(DFPO), number of vessel groups per area
(NGPA), vessel element length (TVEL) and fiber
diameter (FDIA) are heavily loaded along the
PCA Axis one and ray height in number of cells
(RHCE) and in millimeters (RHMM), and ratio of
vessel group to solitary (GSRT), highly influenced
the axis two (Table 4). The total percentage
variation obtained by the analysis is 71.4 % for
the data recorded. The percentage variability
explained by axis one and two is 42.3 %. The rest
of the components are less significant and could be
neglected. Further Table 4 indicates that these
wood anatomical characters which are heavily
loaded along the components could be considered
as important features in separation of species
within the genus. The scatter plots derived from
the PCA with the idea of tracing the grouping
tendency within the genus are shown in the Fig. 5
and 6.
The scatter plots PC 1 Vs PC 2 (Fig. 5)
indicate a weak tendency of grouping of
specimens. Species such as D. insignis and
D. walkeri were separated as two species. The
analysis showed that the above mentioned species
are to a certain extent, related wood anatomically
and further, closely related to D. quaesita Thw.
This agrees with Kostermans’ (1981) who
observed similarity between these two species
based on the colour of the wood and Trimen’s
(1893) observation based on fruit characteristics.
The occurrence of plumbagin, Scopaletin and a
tri-oxygenated Coumarin in these two taxa, further
confirmed their close relationship (Jayasinghe P.,
1992, Unpublished data).
The specimens of D. oppositifolia fall into an
isolated position in scatter plots (No. 22 in Fig. 5
Fig. 6). This is the only species which possesses
opposite leaves within the genus except
D. melanoxylon which possesses sub-opposite
leaves. In geographical distribution, it is highly
restricted only to the top of southwest face of
Haycock mountain in Galle district, at the altitude
of 700 m. Wood anatomically, it is difficult
to establish any
living relationships between
D. oppositifolia and the rest of the species within
the genus.
A group of species which consists of D.
oocarpa, D. montana and D. ebenum (21, 17, 9 in
Fig. 6) could be considered as closely related
species, share the same habitats and are of
arborescent habit. This could be attributed to the
adaptability of the wood anatomical features to a
particular habitat.
Wood anatomical study of Diospyros
127
Rescaled Distance Cluster Combine
0
5
10
15
20
25
┼─────────┼─────────┼─────────┼─────────┼─────────┼
DATTR
DEBNU
ATTNU
DDISC
DHIRS
DACUT
DEBNU
DCHAE
DEBNU
DOVAL
DATTR
DOPPO
DINSI
DFERR
DFERR
DACUM
DINVP
DMOON
DOOCA
DOPPO
DMONT
DRACE
DINSI
DKOEN
DMALA
DMALA
DOBLO
DWALK
DWALK
DOBLO
DOOCA
DNUMM
DACUM
DOBLO
DDISC
DOOCA
DQUES
DINSI
DEBNU
DWALK
DEBNO
DHIRS
DMONT
DMALA
DMELA
DOVAL
DINVP
DRACE
DRACE
DEBNO
DMALA
DRHEO
DTHWA
DAFFI
DATTR
DAFFI
DFERR
─┐
─┤
─┼─┐
─┤ ├─┐
─┘ │ │
─┬─┘ │
─┘
├─────┐
─────┤
│
───┬─┘
│
───┘
│
───┐
│
───┼─────┐ │
───┘
├─┤
───┬───┐ │ │
───┘
├─┘ │
───────┘
├───┐
─────────┐ │
│
─────────┤ │
│
─┬─┐
│ │
│
─┘ │
│ │
│
───┤
│ │
│
───┼───┐ │ │
│
───┘
├─┼─┘
│
───┬─┐ │ │
│
───┘ ├─┘ │
│
─────┘
│
│
─┬───┐
│
│
─┘
├─┐ │
│
─────┘ │ │
│
───┐
│ │
│
───┼─┐ │ │
├─┐
───┘ │ │ │
│ │
─┬─┐ ├─┤ │
│ │
─┘ │ │ │ │
│ │
───┤ │ │ │
│ │
─┐ ├─┘ │ │
│ │
─┼─┤
├─┘
│ │
─┘ │
│
│ │
───┤
│
│ │
───┘
│
│ ├───┐
─────┬─┤
│ │
│
─────┘ │
│ │
│
───────┘
│ │
│
─┬─┐
│ │
│
─┘ ├─────┐
│ │
├───────────┐
───┘
├─────┘ │
│
│
───┬─────┘
│
│
│
───┘
│
│
├───────┐
─────────────────┘
│
│
│
───────────┬─────────┘
│
│
───────────┘
│
├───┐
─────┬───────────────────────────┘
│
│
─────┘
│
├───┐
─────────────────────────┬───────────────┘
│
│
─────────────────────────┘
│
│
─────────────────────────────────────────────┘
│
─────────────────────────────────────────────────┘
Figure 4. Dendrogram based on 13 wood anatomical characters of 58 Diospyros L. specimens.
B. K. L. Wickremasinghe and Tissa R. Herat
128
Table 4. Principal Component Analysis of 13 quantitative wood anatomical characters of the genus
Diospyros L.
───────────────────────────────────────
Component
PC1
PC2
PC3
PC4
PC5
───────────────────────────────────────
% variability
explained
25.8
16.5
11.3
10.0
7.8
──────────────────────────────────────
Cumulative%
25.8
42.3
53.6
63.6
71.4
───────────────────────────────────────
DFPO
.85444 .01995 -.27438
.26724
-.01394
NGPA
.82585 -.07640 -.22702
.26563
.03874
TVEL
-.80365 -.10458 -.03101
.18217
.13401
FDIA
-.68512 .22392 -.08122
.21299
-.15597
RHCE -.08323 .88622 .08822
-.06011
.07625
RHMM
-.14091 .83826 .21480
.10138
.14959
GSRT -.36092 -.50267 .35952
.11301
.15506
RPMM
.09078 .13689 .78069
.12307
-.27920
FLEN
-.16790 .14691 .69696
.06438
.14092
NSPA
.30555 .35471 -.54965
.28028
-.33151
VDIA
-.05679 .05974 -.10736
-.84012
-.04924
VWTH
.41839 -.32269 -.02985
-.43411
.33830
RWMM
.03499 .17232 -.00043
.03343
.88867
───────────────────────────────────────
Figure 5. PCA Scatter Diagram produced by plotting the first PC against the second PC for wood
anatomical characters.
Wood anatomical study of Diospyros
129
Figure 6. PCA Scatter diagram produced by plotting the first PC against the third PC for wood
anatomical characters.
D. ovalifolia which is widely distributed in the
dry zone falls into a separate cluster (23 in Fig. 5
and Fig. 6) which is isolated from the rest of
species. PCA of wood anatomical characters clearly
indicates that D. ovalifolia is easily distinguished
from the rest of the species of the genus. Further
analysis shows that the rest of the species form a
large cluster and reflects their wood anatomical
similarities.
The results of the Discriminant Analysis (DA)
of quantitative wood anatomical data (Table 5)
showed that, 8 out of 13 characters are significant
at P < 0.05, indicating that these characters are
more important in separating the species. Based on
the outcome of the Discriminant Analysis, three
discriminant scores were calculated for each
specimen. Using the discriminant scores derived,
two scatter plots were produced viz. discriminant
function one with discriminant function two and
discriminant function one with discriminant
function three respectively (Fig. 7 and Fig. 8). The
specimens of D. atrata and D. malabarica are
grouped together showing their wood anatomical
similarity (4 and15 in Fig. 5). D. atrata is an
endemic and restricted to the intermediate and the
wet zone of the island. According to Kostermans
(1981), D. malabarica is widely distributed in the
dry and the intermediate parts of Sri Lanka and also
in India, Thailand, Indo-China and Western
Malaysia. Based on the above facts, it could be
tentatively concluded that D. malabarica would be
the basic stock from which the other Diospyros
species arose through the endemic D. atrata.
Separate grouping pattern of D. oocarpa and
D. montana, isolated positioning of D. ovalifolia
and the same grouping pattern of D. walkeri were
also reflected in both PCA and DA. The rest of the
species falls into a single large group (Fig. 7 and
Fig. 8) showing the ambiguity of their
interrelationships.
Results of the Discriminant Analysis of
Qualitative data showed that out of 13 characters 7
were significant at P < 0.05 in the separation of the
species within the genus (Table 6). The results
indicate that intervascular pitting does not
significantly vary within the genus and could be
neglected in the analysis. Scatter plots derived from
the Discriminant Analysis clearly showed that there
is no particular grouping pattern in the data
concerned.
The results of the analysis with wood
anatomical data and ecological origin assigned to
the specimens showed that pore distribution
frequency, fiber to vessel length ratio, number of
groups per area, vessel element length and vessel
wall thickness significantly (P < 0.05) vary with the
ecological condition of the habitat (Table 7). The
scatter plots derived from this analysis reflected
B. K. L. Wickremasinghe and Tissa R. Herat
only the dependence of wood anatomical characters
on ecological conditions of the habitat. The
intermediate zone specimens showed a tendency
of merging with the dry zone specimens rather
than the wet zone specimens. Based on the
above findings it could be concluded that the wood
130
anatomical variation within the genus is of less
importance in delimitation of the taxa. Further,
these results are in compliance with findings which
have already been discussed under
individual
characters.
Figure 7. Scatter diagram produced by plotting Discriminant Function one against Discriminant
Function 2 for quantitative wood anatomical characters
Figure 8. Scatter diagram produced by plotting Discriminant Function one against Discriminant
Function three or quantitative wood anatomical characters.
Wood anatomical study of Diospyros
131
Table 5. Discriminant Analysis of 13 quantitative wood anatomical characters of the genus Diospyros
L. Shows significance of quantitative wood anatomical characters used in Discriminant analysis
(Wilks' Lambda (U-statistic) and univariate F-ratio with 27 and 29 degrees of freedom).
───────────────────────────────────────
Variable
Wilks' Lambda
F
Significance
───────────────────────────────────────
DFPO
.09906
9.7682
.0000
FDIA
.21212
3.9894
.0002
FLEN
.49479
1.0967
.4026
NGPA
.16245
5.5377
.0000
NSPA
.35452
1.9556
.0398
RHCE
.19779
4.3563
.0001
RHMM
.31332
2.3540
.0130
RPMM
.73652
0.3842
.9927
RWMM
.67344
0.5208
.9539
TVEL
.20776
4.0958
.0002
VDIA
.40418
1.5834
.1137
VWTH
.31791
2.3045
.0149
GSRT
.38641
1.7056
.0808
───────────────────────────────────────
Table 6. Discriminant Analysis of 13 qualitative wood anatomical characters of the genus Diospyros
L.. Shows significance of qualitative wood anatomical characters used in Discriminant analysis (Wilks'
Lambda (U-statistic) and univariate F-ratio with 27 and 29 degrees of freedom).
Variable Wilks' Lambda
F
Significance
CRYS
.29530
2.5631
.0073
FIBT
.41912
1.4886
.1477
GRRI
.25743
3.0982
.0018
PABA
.00333 321.0351
.0000
PADI
.07250 13.7418
.0000
PEPL
.36352
1.8806
.0492
PITS
is a constant.
PIVE
.67857
0.5088
.9593
RAGE
.50893
1.0364
.4609
RASE
.60455
0.7026
.8202
TRAC
.22814
3.6339
.0005
TYLO
.24449
3.3191
.0010
VEDI
.24359
3.3353
.0010
─────────────────────────────
Oever et al., (1981) and Carlquist and
Hoekman (1985) working with Symplocaceae
and Staphylaceae respectively showed significant
correlations among wood anatomical characters,
such as vessel wall thickness to vessel diameter,
vessel diameter to distribution frequency, vessel
element length to fiber length and tracheary
element length and vessel element length to ray
height.
B. K. L. Wickremasinghe and Tissa R. Herat
132
Table 7. Discriminant Analysis of wood anatomical characters against ecological zones of the genus
Diospyros L. Shows significance quantitative wood anatomical data in relation to ecological origin of
specimens ( Wilks' Lambda (U-statistic) and univariate F-ratio with 2 and 54 degrees of freedom).
Variable
Wilks' Lambda
F
Significance
DFPO
.53284
23.6723
.0000
FDIA
.86836
4.0931
.0221
FLEN
.96153
1.0802
.3468
FVRAT
.77686
7.7554
.0011
NGPA
.58267
19.3388
.0000
NSPA
.91300
2.5727
.0857
RHCE
.96929
.8555
.4307
RHMM
.92733
2.1159
.1304
RPMM
.90015
2.9949
.0584
RWCE
.90975
2.6786
.0778
RWMM
.97369
.7296
.4868
TVEL
.67061
13.2620
.0000
VDIA
.93179
1.9764
.1485
VWTH
.76149
8.4569
.0006
Somaratne and Heart (2001) showed a higher
significant correlation at P < 0.001, among vessel
element length, fiber length and tracheary element
lengths and significant correlation at P < 0.01
between height of ray and vessel element lengths;
vessel diameter and fiber diameter in Calophyllum
L. species in Sri Lanka.
The correlation coefficients between fiber
diameter and vessel element length and fiber
diameter and number of vessel groups per area
were significantly correlated at P < 0.05 in
the present study. A correlation coefficient at P <
0.05 level was observed for fiber diameter
and vessel wall thickness. These findings do
not support the findings of the above mentioned
work for the genus and further studies
are needed to make conclusions. Evidence is
available to a certain extent that there is a
relationship between wood anatomical variation
and latitudes and latitudes (Metcalf and
Chalk 1989; Graff and Baas 1974; Oever et al.,
1981).
genus Diospyros L. in relation to altitude
isnegligible. This may be due to the narrow
altitudinal distribution of the species studied as all
the specimens collected occurred in the area below
700 m in elevation. Graff and Baas (1974) pointed
out that the elevation differences of the specimens
should be at least more than 2000 m to obtain a
significant dependency of wood anatomical
characters on altitude. Significant correlation
coefficients were obtained for mean annual rainfall
and vessel element length (P < 0.05) and mean
annual rainfall and vessel distribution frequency (P
< 0.05). A strong correlation coefficient has also
been observed among wood anatomical characters
with mesomorphy and conductance.
The results of the study showed a decrease in
vessel element length and vessel diameter with
decrease of water availability (Fig. 9 ). These are
in agreement with findings of Carlquist and De
Buhr (1977). Figure 10 shows the relationships
among mesomorphy indices, conductance and
mean annual rainfall.
Wood anatomical study of Diospyros
Figure 9. The relationship among rainfall, vessel element length and vessel diameter.
Figure 10. The relationship between rainfall, mesomorphy and conductivity.
133
B. K. L. Wickremasinghe and Tissa R. Herat
It is evident that there is a clear relationship
between mesomorphy index and also with
conductance i.e. an increase in mean annual rainfall
causes an increased mesomorphy index and
conductance. As these findings agree with the idea of
Carlquist and De Buhr (1977), both mesomorphy
indices and conductance values could be used to
ascertain the ecological origin of a species of the
genus under consideration.
The relationships among agro-ecological zones
versus mesomorphy and conductance values also
show a linear pattern. Carlquist (1984) pointed out
that vessel grouping and presence of vasicentric
tracheids play an important role in water conductive
efficiency and believed that in case of air embolism
of vessel elements, the neighboring vasicentric
tracheids which will resist the spreading of air
bubbles along the lumen could be of importance in
maintaining the function of the disabled vessel
element. Present study reveals that, vessel grouping is
greater in dry zone specimens rather than wet zone
species. Further, appearance of tracheids is sporadic
and it could be believed that these structures are of
certain value in compensating for the disabled vessel
elements. Lack of tracheids may be considered as a
contributive factor which increases vessel grouping
of the dry zone species of the genus. Further, it is
observed that in the dry zone species, vessel groups
consist of vessels with different diameters.
Functionally this arrangement of vessels will be a
mechanism to maintain the continuous water flow in
case of air embolism of larger vessels which are
susceptible to such conditions. This structural
organization of the vessel arrangement could be
correlated with the drought resistance ability of
some species of the Diospyros taxa of Sri Lanka.
Further studies on these lines are required for better
understanding of the ecological adaptive ability of
wood anatomical characters.
As a whole, it should be pointed out that the
wood anatomical characters of the taxa of the genus
Diospyros are highly correlated with the
environmental conditions of the habitat, especially
with water availability. Characteristic features of
vessel elements and fiber lengths cannot be used as
an indication of evolutionary advancement of the Sri
Lankan endemics of the genus. Such evolutionary
trends cannot be traced among the endemic taxa in
order to establish evolutionary relationships between
the endemic and non-endemic Sri Lankan Diospyros
species. Further studies on comparative anatomical
studies of other vegetative and reproductive parts
may be of importance in delimitation of Diospyros in
Sri Lanka.
ACKNOWLEDGEMENTS
134
Authors thank the colleagues who helped
during field collections and in preparation of the
manuscript, Mr. S. Somaratne for statistical
anmalysis and the laboratory staff of the
Department of Botany for laboratory work.
REFERENCES
Alston, A.G.H. (1931).
Supplement to the
Trimen’s Handbook to the Flora of Ceylon, Part
VI. Dalau & Co. Ltd. London.
Aloni, R., and Zimmermann, M.H. (1983). The
Control of vessel size and density along the Plant
Axis. Differentiation 24: 203-208.
Baas, P. (1976). Some functional and adaptive
aspects of vessel member morphology. In: P. Baas,
A.J. Bolton & D.M. Catling (Eds.), Wood Structure
in Biological and Technological Research, Leiden
University, Leiden Botanical Series No. 3: 157181.
Baas, P. (1982). Systematic, Phylogenetic, and
Ecological Wood anatomy. In History and new
Perspectives in Wood Anatomy. P. Bass, (Ed.),
Martinus Nijhoff / Dr. W. Junk Publisher, The
Hague, Pp.23-58.
Baily, I. W.
(1957). The potentialities and
limitations of wood anatomy in the study of the
phylogeny and classification of angiosperms.
Journal of Arnold Arboratum 38: 243-251.
Baretta-Kuipers, T. (1976). Comparative wood
anatomy of bonnetiaceae, Theaceae and Guttiferae.
In
Wood Structure in Biological and
Technological Research, P. Bass, A.J. Bolton and
D. M. Catling (Eds.). Leiden University, Leiden
Botanical Series No. 3: 76-101.
Barghoorn, E. S. Jr. (1940). The Ontogenic
development and phylogenetic specialization of
rays in the xylem of Dicotyledons I, The primitive
ray structure. American Journal of Botany 27(10):
918-928.
Barghoorn, E. S. Jr. (1941a). The Ontogenic
development and phylogenetic specialization of
rays in the xylem of dicotyledons II, Modification
of the multiseriate and uniseriate rays. American
Journal of Botany 28 (4): 273-282.
Wood anatomical study of Diospyros
Bhargavan, S. and Ramachandran, M. (1984).
Diospyros trichophylla Alston. (Ebenaceae) - A
new record for India. Indian Journal of Forestry.
Botanical. Survey of India 7 (3): 247-248.
Bonsen, K.J.M. and Kucera, L. J. (1990). Vessel
occlusion
in
plants:
Morphological
and
evolutionary Aspects. International Association of
Wood Anatomists Bulletin II (4): 393-399.
Burley, J. and Miller, R. B. (1989). The
application of statistics and computing in wood
anatomy. In Anatomy of Dicotyledons, Vol.II, 2nd
ed., C.R. Metcalf & L. Chalk, (Eds), Clarendon
Press, Oxford, Pp.74-179.
Burmann, J. (1737). Thesaurus Zelanicus exhibiens
Plants in insule Zeylana Nascentes. Janssoniowassbergions & Salonem Schunten. Amsterdam,
Pp.121.
Carlquist, S. (1975). Ecological strategies of xylem
evolution. University of California Press. Burkley,
California.
Carlquist, S. and De Buhr, L. (1977). Wood
anatomy of Penaceae (Myrtales) : Comparative,
phylogenetic,
and
ecological
implications.
Botanical Journal of Linnaean Society 75: 211277.
Carlquist, S. (1984). Vessel grouping in
Dicotyledon wood : Significance and relationship
to imperforate tracheary elements. Aliso 10 (4):
505-525.
Carlquist, S. (1986a). Terminology of imperforate
tracheary element. International Association of
Wood Anatomists Bulletin. n.s. 7(1) : 78-81.
Carlquist, S. (1986b). Terminology of imperforate
tracheary element. International Association of
Wood anatomists Bulletin n.s. 7(2): 168-170.
Carlquist, S. and
Hoekman, D. A. (1985).
Ecological wood anatomy of the woody Southern
Californian Flora. International Association of
Wood Anatomists Bulletin. n.s., 6 (4): 319-347.
Chalk, L. and Chattaway, M. M. (1934). Measuring
the length of vessel members. Tropical Woods 40:
19-26.
Chattaway, M. M. (1932). Proposed standards for
numerical values used in describing woods.
Tropical Woods 29 : 20-28.
135
Chattaway, M. M. (1955). Crystals in woody
tissue. Part I. Tropical Woods. 102 : 55-74.
Chattaway, M. M. (1956). Crystals in woody
tissue. Part II. Tropical Woods 104 : 100 -124.
De Candolle, A. (1844). Ebenaceae, Prodromus
Systematis Naturalis. Regni vegetablis. Sumptibus
Fortin, Masson et Sociorum, Parisiis; 8: 209-243.
Fahn, A., Burley, J., Longman, K.A., Mariaux, A.,
and Tomlinson P.B. (1985). Possible contribution
of the wood wnatomy to the determination of the
age of tropical trees, In ; A. Fahn, Werker, E., and
P. Baas (eds.), Wood Anatomy and Identification
of Trees and Shrubs from Israel and Adjacent
Regions, The Israel Academy of Science and
Humanities, Pp.31-54.
Gauch, H.G. Jr. (1986). Multivariate Analysis in
Community Ecology (Cambridge Study series in
Ecology).
Cambridge
University
Press,
Cambridge, Pp. 298.
Graff, N. A. van der and Bass, P. (1974). Wood
anatomical variation in relation to latitude and
altitude. Blumea 22: 101-121.
Hedren, M. (1990) The Justicia eleganta complex
in tropical Africa (Justicia. Sect. Hannieria,
Acantheceae). Nordic Journal of Botany (Section of
Tropical Botany) 10: 357-398.
Hermann, P. (1717). Musaeum Zelanicum, sive
catalogus plantarum. Lugduni Batavorum, Isaac
Severinus, Pp.21
Hora, B. (1981). The Oxford Encyclopaedia of
trees of the world. Oxford University Press.
IAWA Committee (1989). IAWA list of
microscopic features for hard wood Identification
In; E.A. Wheeler, P. Baas and P.E. Gasson (Eds.),
IAWA Bull. n.s. 10 (3): 219-332.
Irrigation Department (Land Use Division). (1988).
Soil Map of Sri Lanka (Agro-Ecological Regions),
Department of Cartography ITC, Irrigation
Department, Colombo.
Jacobseni, T. D. (1979). Numerical analysis of
variation between Allium cernum and Allium
stellatum (Liliaceae). Taxon 28 (5/6): 517-523.
Jayasuriya, A. H. M. (1998). A new species of
Diospyros (Ebanaceae). Journal of South Asian
Natural History 3 (1): 55-58.
B. K. L. Wickremasinghe and Tissa R. Herat
Johansen, D. A. (1940). Plant microtechnique.
McGraw-Hill, USA.
Kdhir, W. H. and Wright, K. (1982). Systematics
of Graminae: A cluster analysis study. Taxon 31
(1): 9-36.
Kostermans, A. J. G. H. (1977). Notes on
Ceylonese Ebony Trees (Ebenaceae). Ceylon
Journal of Science (Bio. Sci.) 12 (2): 89-109.
Kostermans, A. J. G. H. (1981). Ebenaceae. In:
M.D. Dassanayake and F.R. Fosberg, (Eds.), A
Revised HandBook to the Flora of Ceylon, Vol. III:
1-54, Amerind, New Delhi.
Kramer, P. J. (1964). The Role of Water in Wood
Formation. In M. H. Zimmermann, (ed.), The
Formation of Wood in Forest Trees Academic
Press, New York, Pp.519-532.
Kribs, D. A. (1935). Salient lines of structural
ever, L. van der, Bass, P. and Zandee, M. (1981).
Comparative wood anatomy of Symplocus and
latitude and altitude of provenance. International
Wood Anatomists Bulletin n.s. 2 (1): 2-24.
Pathirana, P.S.K. and Heart, T.R. (2004).
Comparative Vegetative Anatomical Study of the
Genus Garcinia L. (Clusiaceae / Guttiferae) in Sri
Lanka. Ceylon Journal of Science (Bio. Sci.) 32:
39-66.
136
Sokal, R. R. and Rohlf, F.R. (1981). Biometry. The
Principles and Practice of Statistics in Biological
Research, 2nd ed. W. H. Freeman & Co., New
York, Pp.880.
Somaratne, S. and Heart, T.R. (2001). Comparative
vegetative Anatomical Study of the Genus
Calophyllum L. (Clusiaceae) in Sri Lanka. Ceylon
Journal of Sciences (Bio. Sci.) 28: 51-80.
SPSS Inc. ( 1990). SPSS Statistical Data Analysis
(SPSS/PC+ Statistics(TM) 4.0 for the IBM
PC/XT/AT and PS/2). SPSS Inc., Michigan
Avenue, Chicago, Illinois, USA, Pp.B1-E6.
Survey Department ( 1992). Topographic Sheets
(one inch) of Sri Lanka. Revised in 1973, Reprint
1992. Survey Department, Colombo.
Thwaites, G.H.K. (1864). Enumeratio Plantarum
Zeylaniae, Dalau & Co. Ltd. London, Pp.178-183.
Tippo, O. (1964). The role of wood anatomy in
phylogeny. American Midland Naturalist 36: 362372.
Trimen. H. (1893). A Hand-Book to the Flora of
Ceylon, Part III (Indian reprint, 1984). Bishen
Singh Mahendra Pal Sing, Dehradun. 87-102.
Vliet, G. J. C. M. van (1979). Wood anatomy of
Combretaceae. Blumea 25: 141-223.
Robertse, P. J., G. Venter and Janse van Rensberg
(1980). The wood anatomy of the South African
Acacia. International Wood Anatomists Bulletin
n.s. 1 (3): 93-103.
Wheeler, E.A.
(1986). Vessels per Square
Millimeter or Vessel Group per Square Millimeter?
International Association of Wood Anatomists
Bulletin. n.s. 7 (1) : 73-74.
Sass, E. J. (1958). Plant Microtechnique, 3rd ed.,
Iowa State University Press, Ames, Iowa, U.S.A.
White, F. (1993). Twenty two new and little-known
Species of Diospyros (Ebenaceae) from New
Caledonia with comments on section Maba.
Bulletin de Museum Naturel Histoire, Paris, 4e
Ser., 14, Section B, Adansonia, no 2, 179-222.
Sneath, P. H. A. and
Sokal, R.R. (1973).
Numerical Taxonomy. W. H. Freeman & Co., San
Francisco, Pp.573.
Snedecor, G. W. and Cochran, W.G. (1967)
Statistical Methods, 6th ed. Iowa State University
Press, Ames, Iowa, U.S.A, Pp.593.
Worthington, T.B. (1959). Ceylon Trees. The
Colombo Apothecaries Co. Ltd., Colombo.
Wright, H. (1904). The genus Diospyros in Ceylon;
Its Morphology, Anatomy and Taxonomy. Annals
of Royal Botanic Gardens, Peradeniya. II: Part I.
1-106, Part II. 132-210.