ArticlePDF Available

Plant biogeography and conservation of the south-western hill forests of Sri Lanka

  • Cal Poly Humboldt
THE RAFFLES BULLETIN OF ZOOLOGY 2005 Supplement No. 12: 9–22
© National University of Singapore
I. A. U. N. Gunatilleke, C. V. S. Gunatilleke and M. A. A. B. Dilhan
Department of Botany, University of Peradeniya, Sri Lanka
Email: (IAUNG & CVSG); (MAABD)
(All correspondence to I. A. U. N. Gunatilleke)
The south-west of Sri Lanka, encompassing approximately
20,000 km2, is the only aseasonal ever-wet region in the whole
of South Asia (Ashton & Gunatilleke, 1987). This area is of
considerable biogeographic significance to the South and
South-east Asian region for harbouring an exceptional
diversity of a relict biota distinct from that of the Western
Ghats of India (Ashton & Gunatilleke, 1987; Ayyappan &
Parthasarathy, 2001; Gunatilleke & Ashton, 1987; Morley,
2000; van Welzen et al., 2003). The moist rain forests of the
Western Ghats of peninsular India and the aseasonal rain
forests of south-west Sri Lanka are together considered a
refugium of the relict biota of the former Indian plate (Myers
et al., 2000). This plate rafted in isolation from other continents
for a period of over 25 million years in the mid-Paleocene to
late Eocene Epochs (60–35 million years) during which the
evolution of taxa within families and genera continued. Many
of them have subsequently receded in distribution as a
consequence of later competition from Eurasian taxa and
climatic changes accompanying the changing landscape.
These taxa now persist as isolated relicts, vicariants, or
endemics in climatically favourable enclaves like south-
western Sri Lanka and the Western Ghats of peninsular India
(Ashton & Gunatilleke, 1987; Morley, 2000; Swamy et al.,
In both regions, the natural forests are fast disappearing due
to expanding human populations. In southwest Sri Lanka,
much of the lowland and lower montane rain forests were
cleared in the latter part of the 19th century primarily for
expansion of plantation agriculture and its attendant
anthropogenic influences. Recognising the regional as well
as global significance of these two regions of biogeographic
alliance, southwest Sri Lanka and the Western Ghats of
peninsular India have been declared one of the 25 global ‘Hot
Spots’ of biodiversity as they harbour exclusively in them 0.7
% of the plants and 1.3 % vertebrates listed in the global
literature with their habitats under threat (Myers et al., 2000).
Taking in to consideration the numbers of endemics and
ABSTRACT. – The south-western hill forests of Sri Lanka, harbouring a relict of the Deccan-Gondwana biota,
represent an important element of the Sri Lanka-Western Ghats Biodiversity Hotspot. The tree vegetation of
five of these forests along an altitudinal gradient from 300–1250 m a.s.l was quantitatively analysed. While tree
density over 30 cm gbh increased with increasing altitude, tree size decreased. The canopy, sub-canopy and
understorey tree strata in each forest type were largely dominated by species representing distinct forest
communities along this altitudinal gradient except for a few widespread species. Classification and ordination
analysis of three hill forests along this gradient broadly separated each forest into a distinct floristic/
bioclimatological forest community. These analyses further revealed the presence of ecologically distinct
species assemblages on the ridges, slopes and valleys within each forest type. These results highlight the
presence of both landscape and habitat level floristic heterogeneity in the south-western hill forests of Sri
Lanka, which need to be considered in their conservation management. A high proportion of the tree species
(41%) enumerated in Hinidumkanda, Sinhagala and Tibbottagala are globally threatened. The Rakwana-Deniyaya
hill range in the south-eastern lower montane zone is among the least protected owing to the inadequate
knowledge of their rich biological wealth, which is highlighted in this study. A landscape level integrated
conservation plan for the remaining natural ecosystems along the Gin Ganga (River) from its headwaters in the
Rakwana-Deniyaya hills through the southern Sinharaja forest cluster and the Kanneliya-Hinidumkanda cluster
extending to the Hikkaduwa marine sanctuary seascape at the confluence of the Gin Ganga with the Indian
Ocean is recommended as a model river basin landscape-seascape conservation area for this biodiversity
KEY WORDS. – Sri Lanka, biogeography, conservation, lowland hill forests, biodiversity hotspot.
Gunatilleke et al.: South-western Hill Forests of Sri Lanka
endemic species/area ratios for both plants and vertebrates,
and their habitat loss, the Sri Lanka/Western Ghats region is
also ranked among the eight ‘hottest hotspots’ underlining
the global significance of their conservation.
The south-western hills of Sri Lanka are defined as the hills
and hill ranges in the lowland and lower montane aseasonal
climatic regions, to the south and south-west of the central
massif of the island. The rain forests of the south-western
hills are threatened as a result of unabated habitat destruction.
Consequently, the natural vegetation that remains is often
restricted to isolated peaks such as Kalubowitiana (Fig. 1 )
and to the hill ranges such as Kanneliya, Nakiyadeniya,
Dediyagala, Polgahakanda and Hinidumkanda in the coastal
lowlands, the Sinharaja cluster of forests at middle altitudes
(Fig. 2) and the lower montane forests and table-lands of the
Rakwana-Deniyaya hills at the eastern boundary, which abut
the seasonally dry forests to their east (Figs. 3A–C). The
south-western hills are well separated from and located
southwest of the central massif of the island. Abbey Rock
(1300 m) and Gongala (1358 m) represent the highest peaks of
the Rakwana–Deniyaya hills. The south-western coastal hill
ranges commencing immediately east of Kalutara and forming
an arc up to about 30 km wide and parallel to the coast-line
towards the south-east, terminating just inland of Hulandawa
in Matara District, is considered as the floristically richest
area of Sri Lanka and indeed of all South Asia (area 7 of Fig. 6;
Ashton & Gunatilleke, 1987: 275).
Geology. – The main lithotectonic unit of almost the entire
southwest of Sri Lanka is known as the Highland Complex. It
comprises metamorphosed Precambrian sediments and
gneisses of igneous origin (Cooray, 1997). In the extreme
south-western part of this Highland Complex, the crystalline
metamorphic rocks are considered as a separate group, as
they are predominant in granitic gneisses, relatively rare in
the rest of the Highland Complex that extends to the central
massif and north-eastern lower peneplain (Cooray, 1984). The
main metasedimentary rock types in the south-western region
are hypersthene bearing charnockite gneiss of granitic origin,
inter-banded with garnet-sillimanite-graphite gneisses
(schists), cordierite-bearing gneisses and garnet-quartz-
feldspar granulites and gneisses (Cooray, 1988).
A conspicuous geological feature within this region is the
presence of a metabasic (amphibolite and pyroxine granulites)
band of intrusive rocks, extending over 30 km in the SE-NW
direction, known as the ‘Sinharaja basic zone’. This zone is
characterised by a marked aeromagnetic anomaly due to the
high percentage of magnetite present in these basic rocks
(Cooray, 1984).
Land form. – The land form of the Sabaragamuwa hills, of
which the Rakwana-Deniyaya hills form a part, represents a
series of sub-parallel strike ridges running in a SSE-NNW
direction. These ridges have been etched out of a much larger
metamorphic bedrock structure by differential weathering and
erosion of lithologically less resistant beds (Erb, 1984). Heavy
rainfall (>500 cm yr-1) from both the south-west and north-
east monsoons has resulted in deep weathering and active
Fig. 1. Kalubowitiana kanda: an isolated lowland hill (735 m) to the
south of the Sinharaja World Heritage Site, surrounded at the base
by paddy fields and home gardens.
Fig. 2. The mid-elevational hill forests (300900 m) in Sinharaja
World Heritage site dominated by Mesua-Shorea (Doona) forest
erosion almost to base level in the south and south-west
flowing Nilwala, Gin and Kalu rivers. In contrast, the lower
rainfall to the east has resulted in less weathering and erosion,
giving rise to waterfalls and rapids in tributaries of east-
flowing rivers such as the Walawe.
The Rakwana-Deniyaya hills are characterised by a number
of remnants of ancient structurally controlled erosion surfaces,
including the Handapan Ella plateau at 1250 m above sea
level (a.s.l.) and the Thangamale plateau at 1128 m a.s.l. The
Handapan Ella plains have a central undulating valley, several
km2 in extent, covered by grasslands and forests. It is
surrounded by a rim of mountain ranges clad in windswept
lower-montane forests (Fig. 4). Across these ranges, the peaks
of Ilumbekanda (1192 m), Beralagala (1385 m), Suriyakanda
(1310 m), Kabaragala I (1234 m) and Kabaragala II (1291 m)
stand out. The Sabaragamuwa hills drain mainly along a series
of more or less parallel strike-, fault- and joint-controlled valleys
running in the SSE-NNW direction (Erb, 1984). Weathering of
these hills has been intense. Consequently, extensive
deposits of talus and colluvial material resulting from active
mass-wasting have formed deep residual soils, characteristic
of this region.
Soils. – The modal soils of the entire south-western region of
Sri Lanka, except certain parts of the coastal plains, are the
Red-Yellow Podzolic soils with low base saturation. They
belong to the modern Soil Taxonomic Order Ultisols and
Suborders Udults and Humults (Panabokke, 1996). The
coastal plains, especially the lower floodplains and river deltas,
Fig. 3. A, lower montane forests (1000 1300 m) in the Eastern
Sinharaja region. The canopy is dominated by hill dipterocarp
species (Shorea trapezifolia and S. gardneri). B, lower montane
forest above Ensalwatta tea plantation. The rocky outcrop to the
far right is Gongala (1358 m), the highest peak in the Rakwana-
Deniyaya hill range. C, land-use pattern in lower montane areas in
Deniyaya. The Lower valleys are predominantly home gardens,
lower and middle slopes are planted with tea, and only the upper
crests remain clothed with natural forests.
Fig. 4. The stepping-stone arrangement of rocks at the top of the
Handapan Ella, a waterfall dried out in the dry season.
Gunatilleke et al.: South-western Hill Forests of Sri Lanka
have recently-formed alluvial soils of the Order Entisols
belonging to the Sub Orders Aquents, Fluents and
Psamments. Within the coastal plains, there are isolated
residual hilly areas of Red-Yellow Podzolic soils with soft and
hard laterites (Great Soil Group Plinthudults).
Objectives. – The objectives of this study were to: examine
the variation in structure and floristic composition of the tree
flora in the vegetation of the south-western hill forests of Sri
Lanka with increasing altitude; analyse the community
structure of three hill forests along the altitudinal gradient in
order to provide scientific underpinning for the conservation
of plant communities recognized in the south-western hill
forests; and estimate the number of globally threatened plant
species in the IUCN Red List for Sri Lanka that are harboured
in each study site in order to strengthen the cause for the
conservation of their habitats.
The information gathered on phytosociological surveys of
four hill forests (Gunatilleke & Gunatilleke, 1984, 1985) and
one plateau have been re-examined in this paper so as to meet
the above objectives. At some of these sites (Hinidumkanda,
Sinhagala, Tibbottagala), quantitative data were collected from
20 randomly placed 0.25 ha paired plots. In the remaining
sites, floristic information were collected from fewer and
smaller number of plots during short visits to these sites. The
geographical locations of these study sites and their altitudinal
ranges are given in Table 1 and Fig. 5.
Study sites. – A brief description of the study sites follows:
Hinidumkanda or Haycock (300–668 m a.s.l.) is a prominent
lowland hill forest located in the hinterland of the south-
western coastal plains. It is a member of the broader
Kanneliya-Nakiyadeniya-Dediyagala cluster of forests, but
separated from it by the valley of the Gin Ganga [ = river]. In
the more interior Sinharaja range of forests, is Sinhagala (550–
730 m a.s.l.), also a lowland hill forest, and east of it is
Tibbottagala (670–900 m a.s.l), which represents the upper
limits of the lowland rain forests. Still further east and up the
altitudinal gradient on the broader south-western landscape,
the Sinharaja range extends to the Rakwana-Deniyaya hills,
Suriyakanda to the north and Gongala and Abbey Rock to
the south. This area encompasses Illumbekanda, Handapan
Ella, Suriyakanda, Thangamale Plains and Gongala, and the
natural vegetation represents lower montane evergreen
forests (Fig.5).
Vegetation Sampling. – In Hinidumkanda, Sinhagala and
Tibbottagala, 5–5.5 ha of primary forest per site were
quantitatively sampled (Table 1). In each site 10–12 locations
along the elevation gradient were selected randomly. At a
given location, two plots, each 0.25 ha in extent, adjacent to
each other, were demarcated, when space was available.
Mostly rectangular plots (25×100 m) were used to keep within-
plot heterogeneity at a minimum. Where space was limiting,
especially near the summits and ridge tops, square plots (50×50
m) were selected. In each plot, all trees > 30 cm gbh were
sampled by measuring their girth at 1.3 m above the ground.
In the lower montane forests at Suriyakanda only 0.25 ha was
sampled and trees > 30 cm gbh were enumerated; at Handapan
Ella, as the trees were all much smaller, only 0.1875 ha was
studied and the lower limit of trees enumerated was 15 cm
gbh, compared to those in the rest of the study sites.
Herbarium material from trees enumerated was collected for
identification to their respective species, based on
Dassanayake et al. (1980–2000).
Multivariate Vegetation Analyses. – A vegetation
classification based on Two Way Indicator Species Analysis
(TWINSPAN) and an ordination based on Detrended
Correspondence Analysis (DCA, DECORANA) using PCORD
version 4 software (McCune & Mefford, 1999) were performed,
with species abundance data of Hinidumkanda, Sinhagala and
Tibbottagala as three representative hill forests along the
altitude gradient of south-west Sri Lanka from 300–900 m a.s.l.
The data were analysed for all three sites together as well as
Assessment of threatened species. – To assess the threatened
status of the species identified in the study sites, they were
compared with the threatened Sri Lankan plant species
documented in the IUCN Red List (IUCN, 2003).
Structure and floristic composition of tree vegetation. – A
stratigraphic profile of an idealized broad sector of the island’s
floristically rich south-western landscape, from the coastal
region to the eastern boundary of the Rakwana-Deniyaya
hills, depicts its major geomorphology, soil and natural
vegetation features at a glance (Fig. 5). Salient landscape
level differences of these features from the coastal south-
western lowlands to south-central hills in the Rakwana-
Deniyaya range are also summarized in Fig. 5. The structural
features, floristic composition and the dominant species in
the canopy, subcanopy and understorey tree strata of the
forests studied are given in Tables 1 and 2. A comparison of
structural features, such as density (ha-1) of the forests shows
a steady increase from the south-western lowland hills to the
lower montane plateau (Table 2), with a concomitant decrease
in tree height and girth. At the lower montane altitude, there
is a preponderance of individuals with girth lower than 30 cm.
In the Handapan Ella plains, the density of individuals
between 15 and 29.9 cm gbh was 3,386 trees ha-1. In other
areas, individuals of this size-class were not enumerated and
comparison was therefore not possible.
The dominant tree species in the canopy, subcanopy and
understorey tree strata varied among sites, with the exception
of a few widespread species (Palaquium petiolare,
Anisophyllea cinnamomoides, Myristica dactyloides,
Calophyllum trapezifolium and Garcinia hermonii) (Table
2). In general, the dipterocarp-dominated lower montane
forests shared more species with their lowland counterparts
than with those in the Handapan Ella plateau.
Table 1. Some site details and structural and floristic composition of the tree flora (> 30 cm gbh) of lowland hill forests and lower montane
forests in the southwest of Sri Lanka. Numbers within brackets represent unidentified species.
Site details and Hinidum- Sinha- Tibbotta- Suriya- Handapan-
features of vegetation kanda gala gala kanda Ella
Geographical location (° N, ° E) 6.13, 80.18 6.23, 80.28 6.24, 80.32 6.26, 80.38 6.26, 80.36
Elevation range (m) 300–668 550–730 670–910 950–1310 1250–1300
Area sampled (ha) 5 5.5 5 1.25 0.1875
Density ha-1 764 769 712 848 2341
Number of species 140 (21) 125 (15) 140 (17) 75 (21) 40 (23)
Endemic species (%) 75 75 60 >50 >50
Fig. 5. Above: a diagramatic sketch of the relief profile of hills of southwest Sri Lanka, indicating their geology,
soil and forest types. The elevation at the peak of each hill is given along side its name. Below: geographical
distribution of natural forests in southwest Sri Lanka showing their fragmented nature and the approximate
locations (numbers in map) of some of the hill forests.
Gunatilleke et al.: South-western Hill Forests of Sri Lanka
Table 2. The dominant species in different canopy, sub canopy and understorey tree strata in hill forests in the southwest of Sri Lanka. (Calo.
= Calophyllum, S. gard. = Shorea gardneri, Syzyg.= Syzygium, Elaeoc. = Elaeocarpus). * = species endemic to Sri Lanka. ** = subspecies
endemic to Sri Lanka. L/LH = lowland/lower montane; RF = rain forest.
Hill forest studied
Forest type lowland lowland lowland/lower- lower lower
(hill) RF (hill) RF montane RF montane RF montane RF
Vegetation community Diptero- Mesua- Calo.- S. gard.- Syzyg.-
carpus Shorea Shorea Calo.- Calo.-
(Doona) Cullenia Elaeo.-
Canopy and Sub-canopy dominants:
Dipterocarpus zeylanicus (Dipterocarpaceae)*+
Chaetocarpus castanocarpus (Euphorbiaceae) +
Cyathocalyx zeylanica (Annonaceae) +
Palaquium petiolare (Sapotaceae)*++
Anisophyllea cinnamomoides (Rhizophoraceae)*+++
Myristica dactyloides (Myristicaceae) + + +
Mesua nagassarium (Clusiaceae)** +
Mesua ferrea (Clusiaceae)*+
Shorea worthingtonii (Dipterocarpaceae)*+
Shorea affinis (Dipterocarpaceae)*+
Cullenia rosayroana (Bombacaceae)*+
Cullenia ceylanica (Bombacaceae)*+
Calophyllum trapezifolium (Clusiaceae)*+++
Shorea trapezifolia (Dipterocarpaceae)*++
Shorea zeylanica (Dipterocarpaceae)*+
Shorea gardneri (Dipterocarpaceae)*+
Hydnocarpus octandra (Flacourtiaceae)*+
Elaeocarpus glandulifer (Elaeocarpaceae)*+
Axinandra zeylanica (Crypteroniaceae)*+
Palaquium laevifolium (Sapotaceae)*+
Syzygium revolutum (Myrtaceae) +
Syzygium micranthum (Myrtaceae)*+
Syzygium cordifolium (Myrtaceae)*+
Garcinia echinocarpa (Clusiaceae) +
Agrostistachys coriacea (Euphorbiaceae)*+
Understorey Tree Dominants:
Gyrinops walla (Thymelaeaceae) + +
Garcinia hermonii (Clusiaceae)*+++
Xylopia championii (Annonaceae)*++
Dillenia triquetra (Dilleniaceae) +
Humboldtia laurifolia (Fabaceae) +
Diospyros insignis (Ebenaceae) +
Fahrenheitia zeylanica (Euphorbiaceae) +
Strobilanthes spp. (Acanthaceae) +
Bamboo sp. (Poaceae) +
300 668 m
550 730 m
(*Ecotone) 670
900 m
Suriyakanda 950
1050 m
Handapan Ella
1250 1300 m
Among the lowland hill forests sampled, the number of
tree species enumerated was comparable in Hinidumkanda
and Tibbottagala, but lower in Sinhagala. Their numbers
declined progressively in the lower montane forests, again
partly due to the lower girth limit of 30 cm selected for
comparison among sites. Interestingly, the number of
endemic woody species was highest in the lowland forests
and decreased with altitude.
The canopy/sub canopy layer. – The most abundant
species at Hinidumkanda were Dipterocarpus zeylanicus
(Dipterocarpaceae), Palaquium petiolare (Sapotaceae),
Cyathocalyx zeylanica (Annonaceae), Chaetocarpus
castanocarpus (Euphorbiaceae), Anisophyllea
cinnamomoides (Rhizophoraceae) and Myristica
dactyloides (Myristicaceae) (Table 2). However, both the
structure and the abundant species around the peak of
Hinidumkanda differed from those of the middle and lower
elevations (Gunatilleke & Gunatilleke, 1984). Cullenia
rosayroana (Bombacaceae), Myristica dactyloides,
Stemonoporus canaliculatus (Dipterocarpaceae)
dominated the canopy, subcanopy and the understorey
tree layer of the summit vegetation, respectively.
Endangered species like Diospyros oppositifolia
(Ebenaceae) and Schumacheria angustifolia (Dilleniaceae)
are restricted to the summit of Hinidumkanda. This hill
forest represents the Dipterocarpus and Vitex-Dillenia
(Wormia)-Chaetocarpus-Anisophyllea communities in de
Rosayro’s forest classification (de Rosayro, 1950).
Sinhagala, unlike Hinidumkanda, is dominated by the Mesua-
Shorea (Doona) association, which is the climax forest
community in the mid-elevational lowland wet evergreen (rain)
forests of Sri Lanka (de Rosayro, 1950; Gaussen et al., 1966;
Greller & Balasubramanium, 1993). The species characteristic
of the ridge plots in Sinhagala were Mesua nagassarium and
M. ferrea (Clusiaceae), Humboldtia laurifolia (Fabaceae),
Shorea worthingtonii,S. affinis,S. cordifolia, S. disticha
(Dipterocarpaceae) and Palaquium laevifolium (Sapotaceae).
At middle and lower elevations of this hill forest, the most
abundant canopy species were Anisophyllea cinnamomoides
(Anisophylleaceae), Dillenia triquetra (Dilleniaceae), Shorea
congestiflora (Dipterocarpaceae) and Campnosperma
zeylanica (Anacardiaceae). The highest diversity of genus
Shorea among the sites sampled,represented by five species,
was recorded in this hill forest.
In the lower montane ever-wet forests of Tibbottagala (670 –
900 m a.s.l), the most abundant canopy and subcanopy
species were Calophyllum trapezifolium (Clusiaceae), Shorea
trapezifolia (Dipterocarpaceae), Palaquium petiolare
(Sapotaceae), Anisophyllea cinnamomoides (Anisophyll-
eaceae) and Myristica dactyloides (Myristicaceae) (Table 2;
Gunatilleke & Gunatilleke, 1985). The latter two species are
also abundant in the two lower-elevation forests (Table 2).
In Suriyakanda (950–1050 m a.s.l.), the lower montane rain
forests were dominated by Shorea gardneri (Dipterocarp-
aceae), Calophyllum trapezifolium (Clusiaceae) and Cullenia
spp. (Bombacaceae). Similar forest types have been observed
in sites abutting the tea estates in Rakwana-Deniyaya hills.
These forests represent the Shorea (Doona)-Cullenia-
Calophyllum community of the lower montane notophyllous
evergreen dipterocarp rain forest type of Greller &
Balasubramanium (1993).
In Handapan Ella plateau (1250–1300 m a.s.l.), a forest
resembling the lower montane notophyllous evergreen mixed
rain forest of Jayasuriya et al. (1993) and Greller &
Balaubramanium (1993) dominated by the Syzygium-
Calophyllum-Eleocarpus association was observed. This
climatically and edaphically influenced vegetation on the
plateau was short-statured, sometimes exposed to mist and
during the dry months of the year to strong desiccating winds.
It is heterogeneous in its floristic composition, harbouring
elements from lowland and lower-montane rain forests and to
a lesser degree those from the upper-montane forests. Among
the common canopy tree species recorded in plots as well as
in general field collecting were those of Syzygium (Myrtaceae),
Semecarpus (Anacardiaceae), Memecylon (Melastomataceae),
Symplocos (Symplocaceae), Palaquium (Sapotaceae),
Calophyllum and Garcinia echinocarpa (Clusiaceae).
However, the dominant forest species of Handapan Ella
plateau on shallow soils with impeded drainage were not
shared by those of any other forest type examined in this
study (Table 2). This forest type may be an edaphically
influenced variant in shallow soils. Interspersed with the lower
montane rain forests of Handapan Ella and Thangamale plains
and beyond them at Ensalwatta and Beverly estates (although
to a lesser extent) are grasslands, quaking bogs and marshes,
each with their own characteristic biodiversity, which remain
poorly studied to-date.
Understorey Vegetation.– Among the most abundant
understorey tree species were Garcinia hermonii and
Gyrinops walla (Thymelaeaceae) in the lowland hill forests
of Hinidumkanda. In the entire Sinharaja range Garcinia
hermonii and Xylopia championii (Annonanceae) dominated
the undersorey stratum based on the Important Value Indices
(Gunatilleke & Gunatilleke, 1985). Diospyros insignis
(Ebenaceae) and Fahrenheitia zeylanica (Euphorbiaceae)
dominated the understorey of the lower montane forests in
Suriyakanda (Table 2).
On the Handapan Ella plains, the members of Rubiaceae and
Acanthaceae (Strobilathes spp.) were frequent in the
understorey. Bamboo species along with those of Eriocaulon
and Osbeckia, common also in the open grasslands near rocky
outcrops, extend into the forest interior in some areas. On the
plains as well as its outer and inner forested rim an
exceptionally rich orchid flora, some rare and endangered
elsewhere in the island, is present. The rare and cryptic root
parasite Christisonia (Orobanchaceae), only visible when in
flower, is locally abundant on the forest floor in a few areas.
No less than 55 species of orchids, some rare and threatened,
as well as probable new species of Loranthaceae,
Hymenophyllaceae, Lycopodiaceae, Gesneriaceae and
Eriocaulaceae have also been reported from Handapan Ella
and Ensalwatta forests (Suranjan Fernando, pers. comm.).
Three main streams that flow through the plains converge on
Gunatilleke et al.: South-western Hill Forests of Sri Lanka
a rocky substratum at the north-eastern edge giving rise to a
waterfall, overlooking the Deepdene estate and Rakwana-
Kalwana road down below. On this escarpment, either side of
the water fall, is an extensive stand of Loxococcus rupicola
(an endemic and monotypic genus), a palm species
characteristic of rocky steep habitats (Fig. 6).
Multivariate analyses. – In the TWINSPAN classification
using all 62 plots from the three hill forest sites (Fig. 7), the
first division based on the indicator species Cullenia
ceylanica and Bhesa zeylanica clumped the 22 Sinhagala
plots and the 20 Tibbottagala plots as group 1 (mid-elevational
and lower montane forests) and all 20 Hiniduma Kanda plots
as group 2 (coastal lowland hill forest) based on the indicator
species Cyathocalyx zeylanica,Vitex altissima and
Calophyllum moonii. In further divisions of group 1, 10 of
the Tibbottagala plots, representing those in the high plateau
and the valley, grouped together and the remaining 10 with
the 22 Sinhagala plots. Subsequent divisions of the latter
group separated the Tibbottagala slope plots, from the
Sinhagala valley plots in one branch and the Sinhagala ridge
and upper slope plots in the opposite branch. The successive
divisions of the Hinidumkanda plots in group 2 separated the
valley plots, ridge plots and the upper and lower slope plots
into four clusters. The dendrogram (Fig. 7) gives the details
of the clusters, the indicator species and the Eigen values at
each level of division. By comparing the results of the
TWINSPAN analyses and ordination, the divisions in the
classification were ecologically meaningful at level 4 where
plot clusters in valleys, slopes, ridges summits and plateaux
in respective hill forest separated out (Figs. 7, 8).
The groups of plots in each of the three hill forests
separated out in the two-dimensional ordination (Fig. 8).
Those groups in each study site that emerged in the
classification fall in close proximity to each other, and away
from plots in the other two study sites, with one exception,
where the Sinhagala plots on the ridge and upper slopes
clustered adjacent to the Tibbottagala plots on steep
slopes. In a given site the greatest separation among plots
along these two axes is shown by the Tibbottagala plots,
suggesting that this site has more community diversity
than either the Hinidumkanda or Sinhagala forests. Within
each site, finer scale clustering into ridge, slope and valley
plots were evident from both the ordination and
classification (Figs. 7, 8). The multivariate analyses done
for each site separately showed results similar to that
obtained in the analyses of all three sites considered
together, and is therefore not detailed here.
Globally Threatened species. – In the three study sites,
Hinidumkanda, Sinhagala and Tibbottagala, that were
quantitatively sampled with greater intensity than the other
two sites, 224 woody species were recorded (Table 3). By
comparing these species with those reported in the IUCN
Red List, their threatened status was examined. There are 290
globally threatened Sri Lankan species in the 2003 IUCN Red
List (IUCN, 2003). Among them, 78 are Critically Endangered,
73 are Endangered, 129 are Vulnerable and the remaining 10
belong to the Lower Risk, Extinct and Data Deficient
categories. In the three hill sites studied, 118 (41%) of these
species were recorded in the following categories: 31 Critically
Endangered, 16 Endangered, 66 Vulnerable, and 4 Lower Risk.
The data for individual sites (Table 3; Gunatilleke &
Gunatilleke, 1991) show that the number of threatened species
at Hinidumkanda is 91, which ranks highest, followed by
Sinhagala with 83 and Tibbottagala with 68; the number of
threatened species restricted to each of these sites is 26, 8
and 4, respectively.
Forest types represented. – In the past many attempts have
been made to classify the forests of Sri Lanka using floristic
(de Rosayro, 1950; Gaussen et al., 1966), bioclimatic and
physiognomic (Greller & Balasubramaniam, 1993; Koelmeyer,
1957) characteristics (Table 4). The floristic analysis of the
present study reveals that the Hinidumkanda vegetation
represents both the Dipterocarpus and Vitex-Dillenia
(Wormia)-Chaetocarpus-Anisophyllea communities of the
wet evergreen forest climax formation of de Rosayro’s
classification and the lowland mesophyllous evergreen
dipterocarp rain forests of Greller & Balasubramaniam. The
Sinhagala vegetation, on the other hand, represents the
Mesua-Doona (Shorea) community of de Rosayro; but in
Greller & Balasubramaniam’s classification, the distinction
between these coastal and inland lowland forest communities
of de Rosayro has not been made. The presence of several
abundant species in Tibbottagala (670 – 900), which are also
found in both Hinidumkanda (300–668 m) and Suriyakanda
(950–1050 m), suggests that the Tibbottagala vegetation
represents an ecotone between lowland and lower-montane
forests of this region. In Tibbottagala, Mesua nagassarium
is replaced by Calophyllum trapezifolium as one of the most
abundant canopy species.
Floristically, the Suriyakanda vegetation conforms to the
submontane evergreen forests of Koelmeyer (1957), Doona-
Calophyllum-Syzygium series of Gaussen et al. (1966), and
Fig. 6. A stand of Loxococcus rupicola (Araceae) on steep rocky
slopes near Handapan Ella.
Fig. 7. Classification (Two-Way Indicator Species Analyses) of 62 plots sampled in the hill forests of Hinidumkanda (plots Y1 Y10, Z1
Z10), Sinhagala ( A11 A21, B11 B21) and Tibbottagala ( A33 A42, B33 B42) in southwest Sri Lanka, using abundance data of each
species. The indicator species and the eigen values at each division are also given.
the lower montane notophyllous evergreen dipterocarp rain
forests of Greller & Balasubramaniam (1993). de Rosayro has
not recognised this as a distinct forest type in his
classification. The Handapan Ella vegetation, characterized
by the absence of Shorea gardneri, corresponds to the lower
montane notophyllous evergreen mixed rain forests of Greller
& Balasubramaniam (1993). This distinct forest type has been
overlooked in other forest classifications. The floristic data
in our study support the forest classifications based on
distinct floristic assemblages although no single classification
Plots: Sinhagala 22,
Hiniduma 20 and
Tibottagala 20
4210 11
20 Plots
32 8
42 Plots
albidiflora 1
retusa 1
nagassarium 1
walla 1
0.386 0.290 0.422 0.366
Cyathocalyx zeylanica 1
Vitex altissima 1
Calophyllum moonii 1
Cullenia ceylanica 1
hesa ceylanica 1
beddomei 1
coriacea 1
laurifolia 1
laevifolium 1
worthingtonii 1
alaquium 1
Garcinia sp. 1
lanica 1
quaesita 1
zeylanicum 2
petiolare 4
flavescens 1
Less steep
slopes & +ive
Gunatilleke et al.: South-western Hill Forests of Sri Lanka
High plateau
Less steep
slope plots
Borderline plots
Ridge & steep
slope plots
Steep slope
Axis 2
Axis 1
Fig. 8. Ordination (Detrended Correspondence Analysis) on the 62 plots sampled in the hill forests of Hinidumkanda (plots Y1Y10, Z1
Z10), Sinhagala (A11A21, B11B21) and Tibbottagala (A33 A42, B33B42) in southwest Sri Lanka, using abundance data of each
species. The plot distributions in the ordination diagram are indicated by the plot number used in field sampling.
Table 3. Numbers of species in the south-western hill forests of Hinidumkanda (Hinidum.), Sinhagala and/or Tibbottagala (Tibbott.) in
different categories of the 2003 IUCN Red List. CR = critically endangered, EN= endangered, VU = vulnerable, LR/cd = lower risk:
conservation dependent; RIRL, NIRL = number of species recorded and not recorded in the Red List, respectively.
Study Sites Sampled CR EN VU LR/cd Sub Total: Sub Total: Total
Hinidumkanda only 6 4 16 26 42 68
Sinhagala only 3 2 3 8614
Tibbottagala only 1 1 2 41115
Hinidumkanda & Sinhagala 3 2 10 15 4 19
Hinidumkanda & Tibbottagala 1 2 1 41317
Sinhagala & Tibbottagala 8 614 12 26
Hinidum., Sinhagala & Tibbot. 9 5 28 4 46 19 65
Total in all three sites 31 16 66 4 117 107 224
Individual Study Sites
Total in Hinidumkanda 19 13 55 4 91 78 169
Total in Sinhagala 23 9 47 4 83 41 124
Total in Tibbottagala 19 8 37 4 68 55 123
recognised all the different forest communities encountered
in our survey. Interestingly, the six forests quantitatively
sampled by us, along this altitudinal gradient some 80 km
direct east from the south-west coast represent five different
forest types.
This study emphasizes the need for a quantitative analysis of
a much larger database including those from other forests in
the south-western region of Sri Lanka in order to identify
distinct floristic communities facilitating a review the existing
forest classifications in a more objective manner recognising
at the same time micro-scale habitat variations to capture the
habitats of localised endemics.
Floristic variation. – Both TWINSPAN and DCA of pooled
plots of the three hill forests revealed the presence of distinct
floristic communities akin to those described in literature along
the altitudinal gradient at landscape level. Separation of plots
in each of the three hill forests along axes 1 and 2 in the DCA
provides quantitative evidence for the presence of distinct
forest communities in the coastal lowland (Hinidumkanda),
interior mid-elevational (Sinhagala) and lower montane/
ecotone (Tibbottagala) hill forests confirming previous
qualitative classifications of the vegetation. The only
exception was the overlap of four steep slope plots (A39,
B39, A40 and B40) in Tibbottagala with the ridge and steep
slope plot cluster in Sinhagala (Fig. 8), owing to the higher
proportion of species shared between them.
Both analyses indicated a finer scale separation of floristically
recognizable local communities in valleys, lower and upper
slopes and high plateau or summit plots in each hill forest.
However, the plots on the ridges, summits or plateaux, slopes
and valleys in each hill forest separated from their counterparts
in both analyses suggesting their distinctiveness. This
provides evidence for spatial heterogeneity and ecological
differentiation of tree species along the finer-level climatic
and topographic gradients and consequently, provides the
scientific underpinning for conservation management
planning of these forests.
The presence of distinct species assemblages on the ridges,
slopes and valleys, along the altitudinal gradient of the Mesua-
Shorea (Doona) community in hilly topography has been
further supported by census data of the 25 ha Forest Dynamics
Plot (FDP) in Sinharaja, where all free standing woody plants
> 1 cm dbh have been enumerated and plotted on a contour
map (Gunatilleke et al., 2004). This FDP plot spans an elevation
range of only 150 m; yet it depicts eight different habitats
based on topography, convexity and slope. Further, at least
for some canopy dominant taxa, the ecophysiological
adaptations of their seedlings to micro-climatic gradients of
light, moisture and soil nutrients along the topographic
gradient during early establishment and recruitment phases
have been shown (Ashton et al., 1995; Gunatilleke et al.,
1996,1997,1998). All these studies, including the present study,
show habitat specialization of plants along small scale
gradients in the hill forests of Sri Lanka, which has important
Table 4. The south-western hill forests of the present study according to different forest classifications in the literature.
Forest classification of
Hill forests
sampled in de Rosayro Koelmeyer Gaussen et al. Greller & Balasub-
present study (1950) (1957) (1966) ramaniam (1993)
Hinidumakanda Dipterocarpus Tropical wet Doona-Dipterocarpus Lowland mesophyllous
community evergreen Mesua series evergreen dipterocarp rain
forests forests
Sinhagala Mesua-Doona Tropical wet Doona-Dipterocarpus Lowland mesophyllous
community evergreen Mesua series evergreen dipterocarp rain
forests forests
Tibbottagala Mesua-Doona Tropical wet Doona-Dipterocarpus Lowland mesophyllous
community evergreen Mesua series evergreen dipterocarp rain
forests forests
Suriyakanda Not recognized Submontane Doona-Dipterocarpus Lower montane mesophyllous
evergreen Syzygium series evergreen dipterocarp rain
forests forests
Handapan Ella Not recognized Not Not Lower montane mesophyllous
recognized recognized evergreen mixed rain forests
Gunatilleke et al.: South-western Hill Forests of Sri Lanka
conservation implications considering the fact that many of
these species are relict endemics.
Localized distribution of threatened species. – Comparison
of the species in the IUCN Red List (Table 3) in the three
quantitatively sampled forests, shows that 43% of the 224
species identified were found only in one of the three sites,
depicting very localized distribution of these species,
particularly in Hinidumkanda with as many as 68 species.
With 169 tree species it is also the richest among the study
sites sampled. Among the endangered species in the IUCN
Red list that are highly restricted in their distribution and
enumerated in Hinidumkanda are Shorea dyeri, S. lissophylla
and Diospyros oppositifolia. The results of this study also
show that tree species richness decreases with altitude,
Hinidumkanda ranking highest and Handapan Ella the lowest,
the latter probably being also due to the smaller extent of
forest studied. On the other hand most, if not all, studies in
Sri Lanka, including our own, have so far failed to investigate
the diversity of the herbaceous and lower plant floras,
components of the vegetation known to increase in diversity
with altitude.
The remaining forests in southwest Sri Lanka represent only
a small fraction of less than 9% of their original extent. Today
even these are also highly fragmented, most small in extent,
with receding boundaries due to a multitude of human impacts.
Hinidumkanda is one of the few protected areas that represent
the Dipterocarpus community. High point-endemism and high
habitat specialization in Sri Lanka’s rain forest species
demands that the remaining forests in its south-western
quarter, however disturbed or fragmented they are, need to
be conserved with a sense of urgency. Among these are the
representatives of the Dipterocarpus community in the coastal
belt and hill forests and high plateaux of the Rakwana-
Deniyaya hill range representing the lower montane
notophyllous evergreen mixed/Dipterocarp rain forests.
Compared to these forests types, the Mesua-Doona
community is better protected in the Sinharaja and the
Kanneliya-Dediyagala-Nakiyadeniya groups of forests.
Conservation issues. – In response to criticism mounted
against the first Forestry Master Plan for Sri Lanka (Anon.,
1986) and subsequent assessment of the environmental
implications of the Forestry Sector Development Project,
an Environmental Management component was included
in this project (Anon., 1997). This component was
mandated to evaluate all remaining natural forests in the
country with respect to their importance for biodiversity,
in terms of ecosystem and species diversity, and their value
for soil and water conservation. This study, commonly
referred to as the National Conservation Review (NCR),
inventoried and analysed the data of selected groups of
plant and animal species in 204 forests in the island. Among
the recommendations of the NCR, those related to the
present study are, (a) all montane forests should be strictly
protected by upgrading the status of proposed reserves
to conservation forests, or even national heritage
wilderness areas; and (b) the largest remaining forests in
the wet zone should be designated in their entirety as
conservation forests, inter alia, Sinharaja complex (a group
of 13 forests including Handapan Ella Plains, Diyadawa
and Dellawa) and broader Kanneliya, Nakiyadeniya and
Dediyagala (KDN) complex so that the long-term
conservation of biodiversity could be ensured.
The biological value of the forests in close proximity to the
northern, western and southern borders of the Sinharaja World
Heritage Site (WHS) has been recognised and identified as
conservation forests, based on recommendations of the NCR.
However, forests along the eastern perimeter of the Sinharaja
WHS, where part of this study has focused (Suriyakanda and
Handapan Ella), have yet to receive the attention they merit.
These areas were not adequately surveyed for their biological
wealth even during the NCR. With respect to plants, they
harbour an exceptionally diverse herbaceous and epiphytic
flora including orchids, balsams, ferns, bryophytes, lichens,
fungi & algae, a component of the vegetation in all
ecosystems, which was not studied by the NCR and for that
matter, has almost entirely been neglected in Sri Lankan
biodiversity surveys.
Therefore, relying entirely on the NCR as the source of
quantitative scientific information for purposes of
conservation planning could lead to serious omissions of
areas rich in herbaceous and epiphytic plants which are at
greater threat of extinction considering their sensitivity to
habitat alteration and habitat size decrease. Consequently,
due to the dearth of adequate scientific information,
particularly with respect to the rich flora of herbaceous and
lower-plant taxa, some areas of the Rakwana-Deniyaya hills
are not included in the current protected area system. Hence,
they are in imminent danger of being degraded further through
poaching, timber and firewood harvesting, cardamom
cultivation, illicit gem prospecting and other activities inimical
to biodiversity.
Forest degradation and destruction. – The threats to the
many fragile ecosystems of the Handapan Ella plains appear
to be extensive, particularly in the grasslands and stream
banks, where thousands of gem pits (most ~1 m deep) have
been dug over the years. These illegal encroachers have caused
extensive damage to the forest by cutting wood and poaching
animals in this fragile ecosystem.
In early 2004 there was an even more sinister threat to the
relict forest fragments adjoining the eastern boundary of the
Sinharaja WHS, despite repeated requests and
recommendations made to annex them to the Sinharaja WHS
to increase its conservation value. Instead, these state-owned
forest lands were blocked out and sold for tea and cardamom
cultivation. This irresponsible act, amidst public protests,
including those of scientists familiar with the biological wealth
of the region and its conservation value, has irreparably
damaged this fragile ecosystem, critically endangering some
of the threatened animal and plant taxa exclusive to the
Rakwana-Deniyaya hills.
The forests near the south-eastern boundary of the
Sinharaja WHS abutting the Sinharaja Division of the
Ensalwatta Plantation and also Beverley and Manikkawatta
Estates are on long-term lease to plantation companies.
These appear to be relatively free of any encroachments
for gem prospecting and wood-cutting. The forests
surrounding the Sinharaja Division of the Ensalwatta
Plantation represent Shorea gardneri (Sinhala: Rath Dun)
and Shorea trapezifolia (S: Yakahalu Dun) tree formations
at their very best (Figs. 9, 10). These magnificent forest
stands, traversed by streams and spectacular waterfalls,
harbour abundant wildlife, including a small herd of
elephants, and are an important watershed for the
downstream communities.
Similarly, forests in the highest peaks of the Rakwana-
Deniyaya hills, viz., Abbey Rock (1300 m) and Gongala (1358
m), those near Panilkanda and Aninkanda estates, Naigala,
Kabaragala, Beralagala, Suriyakanda and Kurulugala deserve
to be conserved. Apart from a few explorations for taxonomic
and preliminary ecological study purposes, their conservation
value is yet to be evaluated and documented (Jayasuriya et
al., 1993). Several animal species endemic to the Rakwana-
Deniyaya hills and new to science have been collected from
this region (Bahir & Ng, 2005; Manamendra-Arachchi &
Pethiyagoda, 2005; Pethiyagoda & Manamendra-Arachchi,
Extension of the current eastern boundary of the Sinharaja
WHS to include these forests within an appropriate protected-
area category is proposed once again (Gunatilleke & Ashton,
1987), as an urgent measure to conserve their rich biodiversity.
The presence of a small herd of elephants in the Rakwana-
Deniyaya hills is an important criterion for establishing a
wildlife reserve in this region. Their migratory routes need to
be studied in order to demarcate and establish habitat corridors
connecting the surrounding forests, particularly to the north
of the Rakwana-Deniyaya hills. As a dwindling population of
elephants similar to that inhabiting the Peak Wilderness area,
a conservation management plan for wet zone elephants
should also be considered. The role of the elephant in the
functioning of these complex rain forests is virtually unknown.
Such a study would especially benefit nature-based tourism
in this region.
The issues highlighted in this paper should be addressed
coherently with a sense of urgency, before the next wave of
‘development’ overwhelms Sri Lanka. Conservation planning
for Sri Lanka’s south-western region should not be done
piece-meal or on a short-term basis, but at landscape level
that takes long-term impacts into consideration. Appropriate
corridors and buffer zones including multi-species home
gardens, where appropriate, could be used to link the different
ecosystems and forest types in this region.
The present study provides sound evidence that the south-
western hills and, the Rakwana-Deniyaya range of lower
montane forests in particular, indeed possess high
conservation value. However, they have hitherto remained
relatively unprotected owing to their biological and ecological
value not having been adequately assessed in previous
biological surveys. We hope that this study will serve as a
stimulus for demarcating such high-priority conservation
landscapes (both protected and yet unprotected) along the
Gin Ganga basin from its headwaters in the Gongala area
through southern Sinharaja, KDN complex and Hinidumkanda
and remaining mangroves, together with the Hikkaduwa
Marine sanctuary seascape beyond its confluence with the
Indian Ocean at Gintota as an integrated model river basin
landscape-seascape conservation area. Where feasible,
linking of neighbouring forest areas through restoration
corridors with already available research experience for this
region Gunatilleke (1999) is also strongly recommended. The
stratigraphic profile accompanying the map of the south-west
Sri Lanka in Fig. 5 provides some directions for the initial
identification of the relevant areas for this exercise.
We gratefully acknowledge the assistance of the field and
database management staff of the Sinharaja project, the
Department of Botany at the University of Peradeniya and
the Department of Forest Conservation, Sri Lanka for logistic
Fig. 10. Wind-swept and gnarled crowns of the canopy dominant
monospecific dipterocarp (S. gardneri) stands in eastern Sinharaja.
Fig. 9. Shorea gardneri and S. trapezifolia dominated lower montane
forest stand at1000m altitude near Sinharaja division of the
Ensalwatta tea plantation.
Gunatilleke et al.: South-western Hill Forests of Sri Lanka
support. Mr. Suranjan Fernando is thanked for his valuable
information on the orchids of the eastern Sinharaja region.
Anonymous, 1986. Forestry Master Plan for Sri Lanka. (Main Re-
port). Jakko Poyry Oy International, Helsinki. 238 pp.
Anonymous, 1997. Designing an optimum protected areas system
for Sri Lanka’s natural forests (Volume 1). Unpublished Project
Report by IUCN/WCMC/FAO. 201 pp.
Ashton, P. M. S., C. V. S. Gunatilleke & I. A. U. N. Gunatilleke,
1995. Seedling and growth of four Shorea species in a Sri Lankan
rain forest. Journal of Tropical Ecology,11: 263–279.
Ashton, P. S. & C. V. S. Gunatilleke, 1987. New light on the plant
geography of Ceylon I. Historical plant geography. Journal of
Biogeography,14: 249–285.
Ayyappan, N. & N. Parthasarathy, 2001. Composition, population
structure and distribution of dipterocarps in a tropical evergreen
forest at Varagalaiar, Anamalais, Western Ghats, South India.
Journal of Tropical Forest Science, 13: 311–321.
Bahir, M. M. & P. K. L. Ng, 2005. Description of ten new species
of freshwater crabs (Parathelphusidae: Ceylonthelphusa,
Mahatha, Perbrinckia) from Sri Lanka. In: Yeo, D. C. J., P. K.
L. Ng & R. Pethiyagoda (eds.), Contributions to biodiversity
exploration and research in Sri Lanka. The Raffles Bulletin of
Zoology, Supplement No. 12: 47–75.
Cooray, P. G., 1984. Geology, with special reference to the
Precambrian. In: Fernando, C. H. (ed.), Ecology and biogeogra-
phy in Sri Lanka. Dr. W. Junk Publishers, The Hague. Pp. 1–34.
Cooray, P. G., 1988. The National Atlas of Sri Lanka. Survey De-
partment, Sri Lanka. 124 pp.
Cooray, P. G., 1997. Geology. In: Somasekeran, T., M. P. Perera, M.
B. G. de Silva & H. Godellawatta (eds.), Arjuna’s Atlas of Sri
Lanka. Arjuna Consulting Co. Ltd., Dehiwala, Sri Lanka. 220
Dassanayake, M. D., F. R. Fosberg, & W. D. Clayton (eds.), 1980–
2000. A revised handbook to the flora of Ceylon. Oxford & IBH
Publishing Co. Pvt. Ltd. New Delhi. 14 vols.
de Rosayro, R. A., 1950. Ecological conceptions and vegetational
types with special reference to Ceylon. Tropical Agriculturist,
56: 108–121.
Erb, D. K., 1984. Land forms and drainage. In: Fernando, C. H. (ed.),
Ecology and biogeography in Sri Lanka. Dr. W. Junk Publishers,
The Hague. Pp. 35–64.
Gaussen, H., P. Legris, M. Viart & L. Labroue, 1966. International
map of the vegetation: Ceylon. Ceylon Survey Department.
Greller, A. M. & S. Balasubramanium, 1993. Physiognomic, floristic
and bioclimatological characterisation of the major forest types
of Sri Lanka. In: Erdelen, W., C. Preu, N. Ishwaran & C. Madduma
Bandara (eds.), Ecology and Landscape Management in Sri
Lanka: Proceedings of the International and Interdisciplinary
Symposium. Margraf Verlag, Weikersheim, Germany. Pp. 55–78.
Gunatilleke, I.A.U.N., 1999. Reproductive biology and genetics of
selected rain forest plant species of Sri Lanka: Implications for
conservation. (Presidential address-Section D, 1998), Proceed-
ings of the 54th Annual Session, Part II. Sri Lanka Association for
the Advancement of Science, Colombo, Sri Lanka.
Gunatilleke, C. V. S. & P. S. Ashton, 1987. New light on the plant
geography of Ceylon II. The ecological biogeography of the low-
land endemic tree flora. Journal of Biogeography, 14: 295–327.
Gunatilleke, I.A.U.N. & C.V.S. Gunatilleke, 1984. Distribution of
endemics in the tree flora of a lowland hill forest in Sri Lanka.
Biological Conservation, 28: 275–285.
Gunatilleke, C. V. S. & I. A. U. N. Gunatilleke, 1985. Phytosociology
of Sinharaja—a contribution to rain forest conservation in Sri
Lanka. Biological Conservation,31: 21–40.
Gunatilleke, I. A. U. N. & C. V. S. Gunatilleke, 1991. Threatened
woody endemics of the wet lowlands of Sri Lanka and their
conservation. Biological Conservation,55: 17–36.
Gunatilleke, C. V. S., G. A. D. Perera, P. M. S. Ashton, P. S. Ashton
& I. A. U. N. Gunatilleke, 1996. Seedling growth of Shorea
section Doona (Dipterocarpaceae) in soils from topographical
different sites of Sinharaja rain forest in Sri Lanka. In: Swaine,
M. D. (ed.), Man and the Biosphere Series, UNESCO, Paris.
Parthenon Publishing, Carnforth, U.K., 17: 245–263.
Gunatilleke, C. V. S., I. A. U. N. Gunatilleke, G. A. D. Perera, D. F. R.
P. Burslem, P. M. S. Ashton & P. S. Ashton, 1997. Responses to
nutrient addition among Seedling of eight closely related species
of Shorea in Sri Lanka. Journal of Ecology.85: 301–311.
Gunatilleke, C. V. S., I. A. U. N. Gunatilleke, P. M. S. Ashton, & P. S.
Ashton, 1998 . Seedling growth of Shorea (Dipterocarpaceae)
across an elevational range in southwest Sri Lanka. Journal of
Tropical Ecology,14: 231–245.
Gunatilleke, C.V.S., I. A. U. N. Gunatilleke, A. U. K. Ethugala, N. S.
Weerasekara, P. S. Ashton, P. M. S. Ashton & D. S. A. Wijesundara,
2004. Community ecology in an everwet forest in Sri Lanka. In:
Losos, E. C. & E. G. Leigh, Jr. (eds.), Tropical forest diversity
and dynamism: findings from a large-scale plot network, (eds).
University of Chicago Press, Chicago,. Pp. 119–144.
IUCN, 2003. IUCN Red List of threatened species.
Jayasuriya, A.H.M., A. M. Greller, S. Balasubramaniam, C. V. S.
Gunatilleke, I. A. U. N. Gunatilleke & M. D. Dassanayake, 1993.
Phytosociological studies of mid-elevational (lower montane)
evergreen forests in Sri Lanka. In: Erdelen, W., C. Preu, N. Ishwaran
& C. Madduma Bandara (eds.), Ecology and Landscape Man-
agement in Sri Lanka: Proceedings of the International and Inter-
disciplinary Symposium. Margraf Verlag, Weikersheim, Germany.
Pp. 79–94.
Koelmeyer, K.O., 1957. Climatic classification and distribution of
vegetation in Ceylon. The Ceylon Forester, 3: 144–163.
Manamendra-Arachchi, K. & R. Pethiyagoda, 2005. The Sri Lankan
shrub-frogs of the genus Philautus (Gistel, 1848) (Ranidae:
Rhacophorinae), with description of 27 new species. In: Yeo, D.
C. J., P. K. L. Ng & R. Pethiyagoda (eds.), Contributions to
biodiversity exploration and research in Sri Lanka. The Raffles
Bulletin of Zoology, Supplement No. 12: 163–303.
McCune, B. & M. J. Mefford, 1999. PC-ORD: multivariate analy-
sis of ecological data, version 4. MjM Software design, Gleneden
Beach, Oregon, U.S.A.
Morley, R. J., 2000. Origin and evolution of tropical rain forests.
John Wiley & Sons, Inc. New York.362 pp.
Myers, N., R. A. Mittermeier, C. G. Mittermeier, G. A. B. Da Fonseca
& J. Kent, 2000. Biodiversity hotspots for conservation priori-
ties. Nature, 403: 853–858.
Panabokke, C. R., 1996. Soils and agro-ecological environments of
Sri Lanka. Natural Resources, Energy and Science Authority, Sri
Lanka. 220 pp.
Pethiyagoda, R. & K. Manamendra-Arachchi, 1998. A revision of
the endemic Sri Lankan agamid lizard genus Ceratophora Gray,
1835, with description of two new species. Journal of South
Asian Natural History, 3: 1–50.
Swamy, P.S., S.M. Sundarapandian, P. Chandrasekar & S.
Chandrasekar, 2000. Plant species diversity and tree population
structure of a humid tropical forest in Tamil Nadu, India.
Biodiversity and Conservation, 9: 1643–1669.
van Welzen, P.C., H. Turner & P. Hovenkamp, 2003. Historical
biogeography of Southeast Asia and the Western Pacific, or the
generality of unrooted area networks as historical biogeographic
hypotheses. Journal of Biogeography, 30: 181–192.
... Although the species composition of fauna and flora seems to differ between these elevational zones (e.g. fish: Pethiyagoda 1991; agamid lizards: Biswas & Pawar 2006; pollen: Bonnefille et al. 1999; plants: Gunatilleke et al. 2005; snails: Naggs et al. 2005), variation of biodiversity across the three zones remains less well understood. Recent studies elsewhere , however, have shown that intermediate elevations often harbour the greatest species richness (McCain 2005; Oomen & Shanker 2005; Wiens et al. 2007). ...
As part of a Global Biodiversity Hotspot, the conservation of Sri Lanka's endemic biodiversity warrants special attention. With 51 species (50 of them endemic) occurring in the island, the biodiversity of freshwater crabs is unusually high for such a small area (65,600 km(2)). Freshwater crabs have successfully colonized most moist habitats and all climatic and elevational zones in Sri Lanka. We assessed the biodiversity of these crabs in relation to the different elevational zones (lowland, upland and highland) based on both species richness and phylogenetic diversity. Three different lineages appear to have radiated simultaneously, each within a specific elevational zone, with little interchange thereafter. The lowland and upland zones show a higher species richness than the highland zone while--unexpectedly--phylogenetic diversity is highest in the lowland zone, illustrating the importance of considering both these measures in conservation planning. The diversity indices for the species in the various IUCN Red List categories in each of the three zones suggest that risk of extinction may be related to elevational zone. Our results also show that overall more than 50% of Sri Lanka's freshwater crab species (including several as yet undescribed ones), or approximately 72 million years of evolutionary history, are threatened with extinction.
Full-text available
Tropical island countries are often highly populated and deliver immense ecosystem service benefits. As human wellbeing depends on these ecosystems, proper management is crucial in the resource-rich tropical lands where there is less related research. Though ecosystem service and biodiversity studies are a promising path to inform the ecosystem management for these mostly developing countries, published evidence of using ecosystem service studies in decision making is lacking. The purpose of this study is to provide an overview of ecosystem services and related research in Sri Lanka, examining trends and gaps in how these studies are conceptualized. Out of the considered 220 peer-reviewed articles, the majority of articles (48.2%) were terrestrial and forest related while coastal ecosystems were considered in 33.2% of studies. In most studies, the ecosystem service category studied was provisioning (31.5%) followed by regulatory service (28.7%). Studies investigating and quantifying ecosystem services, pressures on ecosystems, and their management were fewer compared to studies related to biodiversity or species introduction. Moreover, studies investigating the value of ecosystem services and biodiversity to the communities or involvement of stakeholders in the development of management actions regarding the ecosystem services were rare in Sri Lanka, and an intense focus from future studies in these aspects is timely and necessary.
Full-text available
The Sinharaja rainforest in southwestern Sri Lanka is a protected forest in a largely agriculture-dominated landscape. In keeping with global UNESCO global biosphere reserves planning, the Sinharaja is surrounded by a buffer zone of regenerating forest and villages with small tea plots and multi-strata tree gardens (homegardens). Globally, however, conservation planning lacks standards on buffer zone management. We ask what relationships exist between village land use and bird assemblages, which are effective ecosystem indicators. Birds have been little studied across land use and vegetation structure in actively managed, large, protected forest buffer zones. To that end, we ran spatially- and temporally-replicated bird point counts across tree gardens, forest fragments, and tea plots within a Sinharaja village. Tree gardens held a greater abundance of birds across habitat association, conservation concern, diet, and endemic species than forest fragments or tea plots. Forest fragments and tree gardens hosted statistically similar numbers of birds in some subsets, but their species assemblages differed. In tea plots, greater shade tree species richness correlated with greater bird abundance and species richness. Our results support the argument for programs to support complex small-scale tree-based agroforestry embedded in buffer zone regenerating forest.
Full-text available
This study presents a systematic revision of South Asian members of the Tribe Chrysillini Simon, 1901. Genetic and morphological variations were analysed of a “similar-looking” group of species that were initially identified as members of the jumping spider genera Chrysilla Thorell, 1887 and Phintella Strand, in Bösenberg and Strand 1906 to determine their phylogenetic relationships. Results suggest that the assessed morphospecies complex constitute of three evolutionary lineages, two previously unrecognised, which are described and diagnosed as two new genera: Phintelloides gen. n. and Proszynskia gen. n. The third lineage, Phintella, is sister to these proposed genera. The following new species are described: Phintelloides alborea sp. n., P. brunne sp. n., P. flavoviri sp. n., P. flavumi sp. n., P. orbisa sp. n., Phintella argentea sp. n., and P. jaleeli sp. n. Sri Lanka is rich in biodiversity but currently has one of the highest rates of deforestation. Lack of clarity on diversity and distribution of the islands’ biodiversity can lead to underestimations during threat assessments and thus downgrading of conservation needs of individual species.
Full-text available
The residual effects of logging on forest fauna and flora have been well studied in other regions of the world, with many investigations finding that recovery of species richness and abundance can occur within one or two decades after the logging event. In this study, we use ant assemblages to compare logged (>30 years) and unlogged mixed dipterocarp forest in the Sinharaja Forest Reserve, a World Heritage Site in Sri Lanka. Species richness and abundance were higher in the logged forest than in the unlogged forest, but not significantly so. Species assemblages, on the other hand, were significantly different and were associated with different environmental variables in the logged and unlogged forest. The findings from this study corroborate other studies that have shown that species composition in logged forest does not appear to return to unlogged forest composition, even after three to six decades have passed. Since this study was not a before-and-after comparison, it is difficult to confirm whether the differences arise from the residual effects of logging or from the general patchiness of species distribution in tropical forest. However, the cumulative results do suggest that there is a residual effect of logging on ant species composition in this forest, even after more than 30 years of regeneration.
Full-text available
The Red Slender Loris (Loris tardigradus) is one of the three primate species endemic to Sri Lanka. Currently there are two recognized subspecies of the Red Slender Loris, the Sri Lanka Red Slender Loris (L. t. tardigradus) and the Sri Lanka Montane Slender Loris (L. t. nycticeboides). Of these, L. t. tardigradus inhabits rainforests in the southwestern region of the island while L. t. nycticeboides is restricted to the montane zone. Kottawa Arboretum harbors one of the few remaining L. t. tardigradus populations in the country. This study was conducted to determine the population density, habitat selection criteria and to asses the habitat availability of L. t. tardigradus in the Kottawa Arboretum. Using the line transect method, 34 sightings were made over a period of 21 months. Based on these observations the density of L. t. tardigradus in the Kottawa Arboretum and habitat selection criteria was ascertained. The calculated density of L. t. tardigradus in the Kottawa Arboretum is 41 animals/ km2. The average height of trees preferred by L. t. tardigradus is 13.97 m plus or minus 6.02. Most of the time lorises were observed at a height range of 3.5-15 m above the ground level. Average height from the ground level where L. t. tardigradus were observed to occupy the tree was 8.64 m plus or minus 5.00. Of the 50 tree species recorded in the Kottawa Arboretum, L. t. tardigradus was found to utilize only 16 species.
The diversity, density, basal area, population structure and spatial distribution of all dipterocarps (≥ 10 cm dbh) was investigated in a 30-ha permanent plot of tropical evergreen forest at Varagalaiar, Anamalais, Western Ghats, south India. Dipterocarpaceae in our site was represented by three species, Dipterocarpus indicus, Hopea parviflora and Vateria indica. These contributed 2% of tree species richness, 6.8% of stand density and 18.3% of stand basal area. The density of dipterocarps in each hectare of the present study site varied greatly, ranging from 7 to 62 trees with a mean density of 31 trees ha-1. Dipterocarpus indicus was the most abundant species accounting for 59% of dipterocarp density. Of the 51 tree families encountered in the 30-ha plot, dipterocarp species ranked first in basal area because of their greater girth frequency in the larger girth classes. Population structures of the three dipterocarp species obtained based on the number of individuals in each diameter class displayed a greater proportion of individuals in the lower size class. Spatial patterns of all dipterocarp species and their three stem sizes analysed at 1 ha, 0.5 ha and 0.25 ha scales revealed that individuals were dispersed both in clumped and uniform patterns, except V. indica which was randomly distributed in one subplot at the 1-ha scale.
Geologically, the island of Sri Lanka is an extension of Peninsular India and forms part of the Indian Shield, one of the oldest and most stable parts of the earth’s crust. Early ideas on the island’s geology by pioneers like F.D. Adams (1929), J.S. Coates (1935) and N.D. Wadia (1945) were therefore largely influenced by concepts of the evolution of the crust of south India current at the time. It was only after systematic geological mapping of the island began in 1952 that more recent ideas on the island’s geology and geological history began to emerge. Within the last 15 years or so, the completion of the preliminary geological mapping and of the gravity survey of the island, on-shore and off-shore exploration for oil, and individual research have contributed to a greater understanding of the geology of Sri Lanka, though many major problems remain unsolved.
Historically the description and discussion of the landforms of Sri Lanka was restricted almost entirely to minor segments of geological reports and papers, or to specific comments on unusual geomorphologic features with local significance (Erb 1963).
The Deccan Plate has played an important part in the history of Dipterocarpaceae and Dilleniaceae, and of suprageneric taxa notably of Palmae, Monimiaceae, Proteaceae, Myrtaceae, and Bombacaceae, while the southern origin of the plate explains the absence or poor representation of many of the leading families of the montane evergreen forests of the Himalaya. The contemporary plant geography of Ceylon is discussed, with emphasis on the rich endemic flora which is concentrated in the southwest. Floristic provinces are identified within that region, and distributional pattern within them elaborated by using Dipterocarpaceae and Myrtaceae as examples. -from Authors
Endemism is concentrated in the lowland Mixed Dipterocarp forests. Within the wet zone, soils and historical geography play a more important role, and altitude a less important role in floristic differentiation than formerly thought. The endemic flora is found to dominate the mature phase, and comprises a greater proportion of stand density and basal area than its floristic representation would suggest. The capacity of the endemic component to re-invade cleared land is poor, however. Gap phase species, many of which are also endemic, play a more important role in wet zone forests than is the rule in the Far East. Contour mapping of floristic variation within this plot using a nearest neighbour technique revealed an unexpected pattern of past disturbance. -from Authors
Asian rainforests suffer from massive deforestation but remain less studied pantropically. Although Sri Lanka is a severely threatened biodiversity hot spot with high endemicity and unique conservation challenges, few have investigated how fauna respond to environmental perturbations caused by human modification of habitats. We investigated differences in the diversity and community composition of amphibians and butterflies in primary and selectively logged forests (SLF), home gardens and plantations in the Sinharaja lowland tropical forest and surrounding areas. A Bayesian multi-model inference approach was then used to identify environmental variables that explained differences in richness and abundance patterns among sites. Amphibian species richness (SPR) was highest in a forest habitat, and community composition differed between forest and disturbed habitats. In contrast, butterfly SPR increased with increasing disturbance. Butterfly diversity patterns were best explained by climatic variables, while structural habitat parameters were the best predictors of amphibian patterns. Based on research findings, the protection of primary forest habitats is essential for amphibian conservation. Regenerating SLF can be of high conservation value especially if substantial pristine habitat remains conserved in the landscape. Anthropogenic habitats at the periphery of forests require realistic management recommendations to support the species inhabiting, visiting or migrating through them. Different environmental predictors for butterflies and amphibians underscore the importance of including multiple indicator taxa in surveys that are intended to inform habitat management.