Biological Conservation 43 (1988) 115-135
Forest Ecosystems and Nature Conservation in Sri Lanka
Zoologisches Institut der Universit~it, Seidlstr. 25, D-8000 Miinchen 2, Federal Republic of Germany
(Received 1 February 1987; revised version received 22 June 1987; accepted 3 July 1987)
A BSTRA CT
Problems related to the conservation of natural forests and the associated faunal communities in Sri Lanka are discussed. Particular reference is made to endemic plant and animal species. The present status of natural forests in Sri Lanka is outlined and the major factors contributing to their decline are described. The most important impacts on the forests are direct or indirect consequences of growth of the human population and associated measures to extend and/or improve land use. The most extinction-prone animal species in Sri Lanka are the larger carnivores such as leopards and eagles, large herbivores (elephant and deer), and habitat specialists, which include most of the endemic vertebrates. Better management and extension of the existing reserves through reafforestation could lower extinction rates of the indigenous and/or endemic plant and animal species.
Ideas of preserving nature in Sri Lanka date back to the advent of Buddhism, about 2500 years BP. Sanctuaries were already established in the 12th century (cf. Crusz, 1973; DeAlwis, 1969), most probably even earlier. Although these ideas are still alive today the basic background of present- day conservationists, as well as of society itself, has changed considerably.
Conservation can be defined as 'the management of human use of the biosphere so that it may yield the greatest sustainable benefit to present generations while maintaining its potential to meet the needs and aspirations of future generations' (IUCN, 1980).
This paper describes the present status and future trends that might 115
Biol. Ctmserl,. 0006-3207/88/$03-50 ~'~ Elsevier Applied Science Publishers Ltd. England. 1988. Printed in Great Britain
116 W. Erdelen
influence the further existence of the forest ecosystems in Sri Lanka. The so- called 'values' of moist tropical forest ecosystems and environmental consequences of their removal have been discussed by Balakrishnan & Easa (1982) and Poore (1979). This paper (1) gives an outline of the present distribution of the 'natural' forests, more or less uninfluenced by man, and the associated vertebrate communities; (2) shows trends affecting Sri Lanka's flora and fauna now or in the near future; and (3) outlines the major factors contributing to the extinction of plant and animal species or populations in Sri Lanka. Finally, a number of suggestions are given to improve conservation measures and to maximise long-term existence of the natural plant and animal communities.
The ideas presented here are the result of almost 4 years of field experience in Sri Lanka in the course of studies on ecology, systematics, and biogeography of lizards (Erdelen, 1984).
THE ENVIRONMENTAL BACKGROUND
Size, relief, climate, and types of natural vegetation in Sri Lanka
Sri Lanka is situated between longitudes E 79 39' and 81 5Y, and latitudes N 5 54' and 9 52'. The island is pear-shaped and has an area of 65 610 km". It is connected to India by a continental shelf and separated from it by the shallow Palk Strait (maximum depth about 15 m, minimum width about 35 km) and the Gulf of Mannar.
The relief of Sri Lanka may be subdivided into three peneplains (Adams, 1929) with average altitudes of 30m, 500m, and 1500-1800m (Domr6s, 1976). The second and third peneplain form the Central Highlands in the southern half of the island, and comprise the Knuckles Mountains in the north-east, the Central Massif with the island's highest mountain, the Pidurutalagala (2524 m), and the Sabaragamuwa ridges in the south-west (Fig. 1).
Temperature, relative humidity and day length show only slight seasonal variations. Mean annual temperatures are between 27C at sea level and 15C at about 1900 m altitude. Humidity is generally high throughout the year (annual averages between 80-85%).
Sri Lanka is commonly subdivided into two climatic subregions, referred to as Dry Zone and Wet Zone (see Fig. 1).
The classification of the vegetation of Sri Lanka is mainly based on the publications of Koelmeyer (1957, 1958), Gaussen et al. (1965, 1968), Fernando (1968) and Mueller-Dombois (1968). Based on Dittus (1977), Crusz & Nugaliyadde (1978) and Crusz (1984, 1986), Sri Lanka can be
Forest ecosystems and nature conservation in Sri Lanka 117
6 4 aU Km
80 81 i i i
Fig. I. Relief and major vegetation types in Sri Lanka, based on Crusz (1984, 1986), Crusz & Nugaliyadde (1978), and Dittus, 1977). A1, A2 = monsoon scrub jungle ('arid zone forest' sensu Dittus (1977); A3 = semi-evergreen forest; A4 = intermediate forest; B = rain forest and grassland below 3000 ft (c. 900 m); C (hatched) = rain forest and grassland between 3000 and 5000 ft (c. 900-1500 m); D (black) = rain forest and grassland above 5000 ft (c. 1500 m). 150 = 150 m isohypse. KN = Knuckles Mountains. Border line between B and A4 more or
less coinciding with border of the Wet Zone (SW) and Dry Zone of Sri Lanka.
subdivided into 7 zones (Fig. 1). For more detailed subdivisions, floristic composit ions and nomenclatures see Gaussen et aL (1968), Peeris (1975), Perera (1975), Grel ler & Ba lasubramaniam (1980), Werner (1984) and Ba lasubramaniam (1985). A discussion o f the origin o f the grasslands o f Sri Lanka is given in Pemadasa (1984).
Distribution of flora and fauna
The flora o f Sri Lanka has 3365 species o f angiosperms (Gunati l leke & Gunati l leke, 1983) and comprises endemics, species which also occur on the
118 IV. Erdelen
TABLE 1 Endemism in the Vertebrates and the Angiosperms in Sri Lanka (Data from Crusz (1986) and Gunatilleke &
Taxon Species Endemic Endemic number species species
Fishes 59 16 27 Amphibians 37 19 51 Reptiles 139 70 50 Birds ~ 237 20 8 Mammals 85 12 14 Total 557 137 25 Angiosperms 3305 830 25
a Only freshwater fishes. Only resident birds.
Indian peninsula, and species which are not found in India but in the Malesian or African region (Abeywickrama, 1959).
Data on non-angiosperms are scanty, an exception possibly being the ferns (e.g. Sledge, 1982), and hardly anything is known about their distribution patterns. Accordingly, they are not treated further here. Similarly, our present knowledge on distribution of most of the invertebrate groups in Sri Lanka is rather poor, and only the vertebrates are treated in detail.
Species numbers and data on endemism for the vertebrates and the angiosperms are given in Table 1. The overall percentage of endemic species is the same for both groups (25%). Within the vertebrate taxa, however, there is considerable variation in the degree of endemism. The more mobile taxa, i.e. the birds and mammals, have fewer endemics than the other three groups. The highest numbers of endemics are found in the amphibians and reptiles. The angiosperms have 27 (Abeywickrama, 1956), fishes 3, amphibians 1, reptiles 9, birds 0 and mammals 4 endemic genera (Crusz, 1986 and pers. comm.; Crusz & Daundasekara, in press).
The geographic distribution of the endemics shows a clear pattern (Fig. 2). The smallest number are found in the relatively dry areas of Sri Lanka (A) and the highest numbers in the lowlands of the Wet Zone (B), followed by zones C (montane forest zone) and D (cloud forest zone). In the angiosperms the pattern is similar and the differences between the Dry and Wet Zones seem to be even more pronounced (Fig. 2). As there were no separate data available for zones C and D these had to be taken as one unit.
A quantitative analysis of the densities of endemic species per vegetational
Forest ecosystems and nature conservation in Sri Lanka 119
~LU Zo m ku ~0.
0' A B C+D
A B C D
Fig. 2. Absolute numbers (black) and percentages (open bars) ofendemic species per taxon in the different vegetation zones A (comprising A1, A2, A3 and A4), B, C and D (for definitions see Fig. 1). Top: Data for the angiosperms (raw data from Gunatilleke & Gunatilleke, 1983). For zones C and D no separate data were available. Bottom: 1 = fishes,
2 = amphibians, 3 = reptiles, 4 = birds, 5 = mammals.
zone would be desirable within the context of conservation. This was carried out by Senanayake (1977) and Senanayake et al. (1977) for the amphibians, lizards, and birds. As pointed out by Crusz (1986 and pets. comm.), one possible source of error in these studies is the use of 'surveyors' areas instead of absolute surface area. This is probably of particular importance in mountainous country.
FACTORS AFFECT ING CONSERVATION MEASURES
Human settlements and land use
The first time large areas were cleared for cultivation (mainly of rice) was during the advent of the Sinhalese, approx. 2500 Be. As can be seen from the
120 W. Erdelen
Fig. 3. Forest cover and major crops in Sri Lanka. Combined after Geiser & Sommer (1982) and the Sri Lanka Map of Vegetative Cover (Survey Department, Sri Lanka). Black = dense natural forest, dots = land under rice cultivation, hatching = rubber, vertical lines = tea, horizontal lines = coconut. Locations for orientations: C = Colombo, J = Jaffna, T = Trincomalee. K = Kanneliya Forest. PR = Pidurutalagala Ridge. PW = Peak Wilder- ness. S = Sinharaja Forest. For information on relief and delimitation of Dry Zone and Wet
Zone see Fig. 1.
Forest ecosystems and nature conservation in Sri Lanka 121
distribution of the so-called 'tanks' (water reservoirs for irrigation purposes (Fig~ 4 in Abeywickrama, 1956)) the natural vegetation in the Dry Zone must have been severely affected. Hence it is assumed by some authors that there are no primary forests left in the Dry Zone (e.g. Holmes, 1951, 1958). This matter is discussed in more detail in DeRosayro (1960).
During the time of the old Sinhalese kingdoms (2250 BP to the beginning of the 16th century) there were only few settlements in the Wet Zone (DeRosayro, 1960). According to Abeywickrama (1956), large-scale plantation in the Wet Zone was introduced in the last 200 years, under the influence of the British. The introduction of coffee in the early 19th century, and later the planting of tea and rubber, caused the most significant impact on the forests in the wet mountainous regions. For construction work and for use as firewood on a large scale exotic trees such as Eucalyptus, Cupressus, and Acacia were planted in the hills. Furthermore, as a consequence of the growing tea industry, the number of settlements in the hills has increased considerably during the last 100 years (for an overview see Werner, 1984).
Since Sri Lanka became independent (1948) these trends have continued. As a result, much of Sri Lanka's area is under cultivation now (Fig. 3, and below).
Present status of natural vegetation and protected areas
Remote sensing under the Sri Lanka/Swiss Satellite Image Interpretation Project, carried out in 1981, gave a figure of 24.9% for the 'natural' forest cover, which amounts to approx. 16 300 km 2 (Fig. 3; for details see Geiser & Sommer, 1982). With more than 50% forest cover in the 1950s Sri Lanka was well above the world average and above the average for the Asian nations (Anon., 1978a). The highest percentage of forest cover today is found in the lowlands of the Dry Zone (Fig. 3, approx. 30%), followed by the Central Hills and the lowland Wet Zone. Large continuous patches of forest are left in the north-western, north-eastern and south-eastern parts only. These forests represent the semi-evergreen type of monsoon forest. The only two larger areas in the Wet Zone, of which only 9% are still under natural forest cover (Gunatilleke & Gunatilleke, 1983), are the Sinharaja Man and Biosphere Reserve (90 km 2) and the Kanneliya Forest (Fig. 3), the latter still being harvested for timber by the Plywoods Corporation. The only large continuous areas of more or less natural forest in the Central Hills are the Peak Wilderness Sanctuary (224 km 2) and the Pidurutalagala Ridge (Fig. 3). There is heavy encroachment at the border of the Peak Wilderness; settlements and shifting cultivation are extending into the sanctuary (Werner, 1985). The present status of the montane and cloud forests along
122 IV. Erdelen
1900 1950 2000
Fig. 4. Trends in human population growth (POP) and forest cover (FC) in Sri Lanka. Population in millions and forest cover in per cent of total land area of Sri Lanka. Data on population growth are from th~Department of Census and Statistics, data on forest cover are from Nanayakkara (1982). Fitting of data points by an exponential regression model:
Yl = c~ ec2t, where y~ = population or forest cover in the year t, c~, c 2 = constants.
. . . .
10 8 6 t. 2 0 0 2 t. 6 8 10
Fig. 5. Age pyramid for Sri Lanka, based on 1981 census (courtesy, Department of Census and Statistics, Colombo) Abscissa: human population in hundred thousands. Ordinate: age
classes. Total census in 1981: 14 847 000.
Forest ecosystems and nature conservation in Sri Lanka 123
the Pidurutalagala Ridge has to be considered critical (Werner, 1984). To sum up, the units with the highest degrees of endemism, viz. the Wet Zone lowlands and the Central Hills (see Fig. 2), nowadays have only very little natural forest cover.
The potentially maximum extent of protected areas in Sri Lanka, including forest reserves, proposed reserves and wildlife reserves, is about 15 693 km 2 (Department of Census and Statistics, 1985, Table 50). This is, however, a rather optimistic estimate as in the greater part of these areas there is already more or less heavy encroachment on the forests, even in forest reserves.
The human population in Sri Lanka is growing rapidly (Fig. 4). Predictions based on an exponential regression model would give an estimate of about 20.8 millions for the year 2000. This is equivalent to a population density of 318 individuals per km 2.
Most problems related to conservation measures are direct or indirect consequences of growth of the human population. From 1900 to 1982 (last census available) the population in Sri Lanka increased from 3.6 to 15-2 millions, which is more than 400%. The age pyramid for Sri Lanka (Fig. 5) is typical for a developing country, the broad base indicating a high percentage of children, as generally found in the fast-growing populations of the Third World. Even strict birth control measures can scarcely prevent further growth due to the lower age classes reaching sexual maturity.
As would be expected, the trend in natural forest cover runs opposite to the population (Fig. 4). According to predictions based on extrapolations of the present trend we should have reached, in 1986, the value of 15 693 km 2 for the total natural forest cover of the island. These are all gross estimates. Firstly the satellite survey estimates natural forest cover for the year 1981, whereas forest reserves or proposed reserves have to a large extent been cleared and replaced with exotic trees since then (Werner, 1984). As a result, the figure of 15 693 km 2 is certainly higher than the area within these reserves actually covered with natural forest. Secondly, very small isolated forest pockets are certainly not mapped by the satellite survey. Although their present status may in many cases be a type of secondary forest and although their total area is probably not very large, they should be taken into account when the total forest cover is estimated. In particular, these fragments are of importance in the lowlands of the Wet Zone and the Central Hills. Thirdly, seemingly natural forest has been mapped under the satellite survey in which cardamom (Elettaria cardamomum, Zingiberaceae) is planted. These forests have a relatively intact canopy but are unable to regenerate due to the removal of the young trees and saplings. As a result, the long-term existence of these montane and cloud forests is threatened. This is of particular importance in the Knuckles Mountains (Fig. 1) where.in 1978 approx. 50%
124 w. Erdelen
of the total area growing this spice was located (Werner, 1984). It should be mentioned that some of Sri Lanka's endemic plants and animals are restricted exclusively to the Knuckles range of mountains. Since 1978 the area where cardamom is grown has increased almost 30%, although there was a slight decrease from 1980 (5300 ha) to 1983 (5000 ha; Sri Lanka socio- economic data, 1984).
From an aerial survey the degree of disturbance of the lower vegetation layers, beneath the crown level, cannot be estimated. Consequently, the survey data should be related to data of forest structure to give realistic estimates of the extension of undisturbed forests.
Generally, in all areas still under more or less natural forest cover there is constant encroachment by the surrounding human population. Less severe disturbance can be expected for the National Parks and the Strict Natural Reserves (for definitions of these terms see DeAlwis, 1969). All other areas are likely soon to be deforested.
The Mahaweli Project
One of the biggest irrigation works carried out in Asia is Sri Lanka's Mahaweli Ganga (river) Project. The pros and cons of this scheme cannot be treated in detail here (see Anon., 1978b; Iriyagolle, 1978), but a brief outline of the scheme and its impact on the natural ecosystems of Sri Lanka can be given.
The purpose of the project is to use the longest river (approx. 330 km), the Mahaweli, and adjoining river basins to irrigate the northern and northeastern parts of the Dry Zone lowlands (Fig. 6). A number of hydroelectric plants were planned, most of them now completed (1986). For particulars of the projects see Mahaweli Development Board (1977).
There is no general agreement on the total area that will be affected by the Mahaweli Scheme. According to a Canada-Ceylon Colombo Plan Project (Andrews, 1971), it amounts to about 10400km 2, which is equivalent to about 16% of the total area ofSri Lanka or 24% of the total area of the Dry Zone. Hoffmann (1978) estimated that almost 50% of the central and northern parts of the island will be affected by the scheme. According to Perera (1984) the gross area of the Mahaweli Project is about 55% of the Dry Zone or 39% of the total area of the island.
The scheme will be located in that part of Sri Lanka still most densely forested (Fig. 3) and an estimated 2630 km 2 of forest will vanish after it is fully implemented.
This figure, in addition to the estimates of present forest cover (see above), shows that there will be virtually no forest left in Sri Lanka which is not yet declared forest reserve (or proposed reserve) or wildlife reserve.
Forest ecosystems and nature conservation in Sri Lanka 125
Fig. 6. Area (hatched) of Sri Lanka expected to be affected by the Mahaweli Scheme (after Mahaweli Development Board, 1977). See Figs 1 and 3 for comparisons with relief, major
vegetation types, and present forest cover of Sri Lanka.
A number of suggestions have been made to lessen the impact in the zone affected by the Mahaweli Project. This area has comparatively little rainfall, dry desiccating winds, high air and soil temperatures during day and night, and a soil moisture deficit for at least six months of the year (Andrews, 1971). However, their successful realisation is doubtful. In particular, it remains to be seen whether the already existing National Parks and the proposed reserves in the area of the scheme can be successfully maintained and effectively protected for a longer period of time. Furthermore, the acceleration of the whole programme, which was initially scheduled for a 30- year period, to five to six years (decided by the Sri Lanka Government in 1977) will bring about a number of problems, all directly or indirectly but negatively affecting the natural forests in the relevant area.
126 w. Erdelen
TAXA-SPECIFIC QUESTIONS RELATED TO CONSERVATION
The main conservation aim should be to preserve as many of the plant and animal species as possible, in particular the indigenous and/or endemic taxa. Specific ecological requirements of some of the flora and fauna are described and whenever possible suggestions for their preservation are made, although knowledge of the biology of most of the species is poor.
The diversity of vegetation on a small scale, recovery of the rain forests after disturbance, and dispersal ability of dipterocarp trees, are of special importance in relation to conservation measures for Sri Lankan plants.
Vegetation A mosaic of different vegetation types exists even within a single climatic zone (Peeris, 1975). Individual hills can be covered with a remarkable, unique flora, as shown for example, by Haycock Mountain in the lowlands of the Wet Zone. This peak contains the highest number of endemic tree species (75%) compared to the other lowland rain forests so far studied in detail in Sri Lanka (Gunatilleke & Gunatilleke, 1984).
A number of other isolated mountains or inselbergs in Sri Lanka are of special importance in this context, for instance, Ritigala in the Dry Zone, which is famous for its medicinal plants, Namunukula Hill in the eastern part of the Central Hills and the rather isolated hills in the vicinity of Monaragala. Furthermore, remarkable differences in floristic composition exist between the Knuckles Mountains and the Central Hills (Werner, 1982).
Analysis of these patterns in vegetation and associated distribution patterns of certain animal taxa might be the only way to understand faunal and floral evolution in Sri Lanka.
Primary rain forest Primary rain forest species in Sri Lanka seem unable to regenerate in gaps inside the forest larger than about 1 ha (Gunatilleke & Gunatilleke, 1983) because of development of secondary scrub. Careful management could allow for speedy regeneration of virgin forest in cleared areas abandoned after cultivation.
It is very important to note that even selective logging can cause large gaps in the forests. Hence, succession following this disturbance might stop at the stage of secondary growth or lead to the former virgin forest only after very long time-spans.
In Sinharaja, the largest area in Sri Lanka still under more or less virgin evergreen tropical lowland forests, most plant species might be classified as
Forest ecosystems and nature conservation in Sri Lanka 127
vulnerable or endangered (Gunatilleke & Gunatilleke, 1985). This is mainly the result of man's impact before 1977 when Sinharaja was declared a Man and Biosphere Reserve.
Dipterocarp forests The characteristic elements of the evergreen lowland rain forests and the hill forests of Sri Lanka, the dipterocarps, are the most important from the viewpoint of endemism. All but one of the 56 Sri Lankan species are endemic (Balasubramaniam, 1985). These forests form a refuge of the once widespread mid-Miocene tropical rain forests (Meijer, 1982), and their conservation should take into account that these trees are poor dispersors, a general feature among the Wet Zone primary forest species (Peeris, 1975), and that their existence is threatened by disturbances of the crown cover of the forests.
Amphibians and reptiles
Three aspects are of special importance in the conservation of the herpetofauna of Sri Lanka: the high degree of endemicity, natural patchiness in geographic distribution, and sensitivity to man-made alterations of habitat in most of the species.
Amphibians and reptiles seem to be less extinction-prone taxa when compared to birds and mammals (Wilcox, 1980). They do, however, show the highest degree of endemicity in Sri Lanka, approx. 50% of the species being endemic (Table 1). However, being less conspicuous they have so far been neglected when conservation measures have been discussed.
The amphibians and reptiles, together with fishes, have a more restricted range of geographic distribution than birds and mammals, as can be seen from the average occurrences in the seven different vegetation zones (as listed above and mapped in Crusz, 1986) per endemic species: the values are 2.0 (fishes), 1.9 (amphibians), 2.2 (reptiles), 4-0 (birds), and 3.2 (mammals). This means, for example, that an endemic fish species is found on the average in two of the seven zones as given above.
Unfortunately, we know very little about the habitat requirements of the Sri Lankan amphibians in relation to their reproductive cycles, in order to assess the degree of vulnerability to man-made changes. It can be assumed, however, that they react rather sensitively to changes in temperature and humidity in their environment due to their low resistance to dry conditions. Most of the endemic amphibians and reptiles are typical forest species which cannot establish populations in disturbed ecosystems. Some, like the agamid lizards Otocryptis wiegmanni, and Ceratophora aspera, are ground-dwellers found in undisturbed patches of forest only, that is in a shady environment.
128 w. Erdelen
Opening up of the forest might completely change microclimatic conditions for species like the three belonging to the endemic frog genus Nannophrys, which seem to occur only under rocks and small stones along forest streams (personal observations).
Furthermore, availability of ponds for reproduction might be restricted to periods with high rainfall in disturbed forest patches with higher evaporation and insolation in the lower vegetation strata compared to undisturbed forests.
Reptiles also have a well-pronounced patchiness in horizontal and vertical distribution. The endemic agamid lizard Cophotis ceylanica has allopatric populations in the highest parts of the Knuckles Mountains and the Central Hills, and according to a survey carried out in 1985 and 1986 seems to have become extremely rare in the Knuckles (Fig. 1). It remains doubtful whether its continued existence is possible under the present conditions where almost no undisturbed forests remain. In Ceratophora, one of the three endemic species is exclusively restricted to the Knuckles, and although it seems to be able to live in cardamom plantations we do not know whether populations are stable or declining.
This natural patchiness in distribution is becoming more pronounced through further fragmentation and a higher degree of isolation of the forest patches. Overall quantitative effects on the amphibians and reptiles are unknown. Likewise, we do not know the critical minimum population sizes needed to prevent extinctions of populations or even species, nor whether these critical sizes have already been reached in some of the forest fragments.
Changes in soil quality due to forest removal have considerable impact on burrowing skinks of the subfamily Scincinae, which live in the leaf litter or in the top layers of the soil. The hitherto decribed nine species (Deraniyagala, 1953; Greer, 1970, Cruz & Daundasekera, in press) are all endemic to Sri Lanka, their closest relatives bein.g found in Africa, Madagascar, and India.
Terborgh (1975) lists four factors which contribute to the resistance of temperate zone bird species to extinction, namely, their capability of migration, acceptance of middle-aged secondary vegetation, high repro- ductive rates, and dense populations.
Most tropical birds lay only two to three eggs per breeding season. The 20 endemic Sri Lankan bird species produce clutches of 2.3 eggs on the average (Henry, 1978). Some of them, however, have two breeding seasons per year.
Evidence for the poor performance of tropical forest birds in disturbed habitats is their inability to compete with non-forest birds, inadequate
Forest ecosystems and nature conservation in Sri Lanka 129
zonation of vegetation structure for niche separation, and their limited dis- persal ability (Terborgh, 1975). Eighty per cent of the Sri Lankan endemic birds are typical forest species requiring undisturbed conditions (Hoffmann, 1984). In addition, the considerable distances which now occur between undisturbed patches of vegetation makes it doubtful whether colonisation will take place. A similar situation was reported for the mammals of the mountain tops of the Great Basin in the United States, where due to the degree of isolation between mountains colonisation ceased (Brown, 1971).
There is evidence that forest disturbance results in the migration of endemic birds to undisturbed patches in the interior of the forest (Kotagama in Saparamadu, 1983). An endemic, the blue magpie Kitta ornata, which tried to invade cleared areas on the border of Sinharaja, seemed unable to cope with the new environment due to high rates of parasitism (90% of the nests) by the Indian koel Eudynamis scolopacea (P. B. Karunaratne, pers. comm.), which normally lays its eggs in the nests of the two crow species found in Sri Lanka and does not occur in dense forest (Henry, 1978).
Hoffmann (1984) lists among the four greatly endangered species of Sri Lankan birds two species of eagles found in the hill forests (mountain hawk eagle Spizaetus nipalensis kelaarti and the rufous-bellied hawk eagle Hieraetus kienerii k&nerii), which need large home ranges and hence very large reserves. Among the eight species listed as endangered by Hoffmann (1984), one is a habitat specialist (Ceylon painted partridge Francolinus pictus), occurring on dry grasslands, and one a hill forest species with very low population densities (Ceylon scaly thrush Zoothera dauma imbricata). All the other six endangered species are forest birds, five of them also endemics.
Generally, the mammals of Sri Lanka show the lowest degree of distributional zonation among the higher vertebrates (see above). Nevertheless, heterogeneity within the group related to the different taxa and/or habitat requirements in some species require specific statements regarding their conservation.
Carnivores, ungulates and primates have their highest species diversities in the Dry Zone (McKay, 1984) and will consequently be profoundly affected by the impact on the vegetation caused by the Mahaweli Scheme. In particular, populations of large carnivores such as the leopard, with an average home range size of 30 km 2 (Eisenberg & Lockhart, 1972), will be most extinction-prone.
Bats and rodents, the two most diverse groups of mammals with 28 (31 according to McKay, 1984) and 21(23) species, respectively (Phillips, 1984)
130 W. Erdelen
have their highest percentage of habitat specialists in zones C and D, including a number of endemics. Many rodent species are rare (McKay, 1984) and the present status of many bat species and their importance in the pollinating system is only poorly understood.
Four species of Cervidae in Sri Lanka require large reserves, and three, i.e. the barking deer Muntiacus muntjak, hog deer Cervus porcinus and sambar C. unicolor, are mostly solitary-living forest species and hence additionally affected by the reduction in forest cover.
The small nocturnal primate, the slender loris Loris tardigradus, has to be considered endangered in Sri Lanka. This species is an extreme specialist dependent on dense understorey vegetation (for details see references in McKay, 1984). Furthermore, its population densities of c. 1 individual per hectare indicate that this species, too, can survive only in large reserves with undisturbed vegetation.
The most conspicuous Sri Lankan mammal is certainly the Asian elephant, the nominate race of which occurs on the island (Elephas maximus maximus). The population was estimated to be about 12 000 at the begin- ning of the last century, occurring all over the island up to altitudes of about 2000m. Since then numbers have changed drastically. In 1970 the population was estimated to be about 3000 individuals, almost exclusively found in the Dry Zone. As this is still the case (McKay, 1984), measures planned under the Mahaweli Scheme will have considerable impact on their populations, particularly because habitats with the highest elephant densities are located in the project area.
For elephants, a specific need exists to establish forest corridors which ideally should coincide with their traditional migration routes. A lack of corridors and the surrounding of the remaining forest patches by human settlements and cultivated areas may ultimately result in extinction of certain herds due to excessive damage of the remaining vegetation and processes of degradation thus initiated (Vancuylenberg, 1977).
Translocation of elephants has not been successful and might ultimately result in overcrowding in the last few protected areas, i.e. mainly the National Parks of Wilpattu in the northwest, and Yala in the southeast of the island. For information on 'pocketed' herds and translocation experiments see Fernando (1983) and Hoffmann (1978).
In order to prevent wildlife in Sri Lanka being restricted to the National Parks only and to improve the general situation as far as nature conservation measures are concerned, the following suggestions (which are
Forest ecosystems and nature conservation in Sri Lanka 131
not mutually exclusive) should be discussed and the necessary measures taken as soon as possible.
(1) Generally, there should be an immediate stop to the clearing of natural forests, whatever the purpose. All areas still under natural or almost natural forest cover should immediately be strictly and totally protected. In particular, areas scheduled for deforestation under the Mahaweli Scheme should be kept to a minimum. Obviously, the best solution would be to leave the remaining forests untouched. If unavoidable, the reserves within the relevant area should be designed according to the criteria already mentioned.
(2) Effective protection of reserves and other so-called protected areas. (3) Small remaining pockets of natural forest should be absolutely
protected and constantly monitored for assessment of their successional status.
(4) A solution should be found for the fuelwood problem without causing further impact on the few remaining natural forests.
(5) Planting of mahogany trees should cease because this process is slowly changing the natural forest into a man-made one.
(6) There should be immediate reafforestation of abandoned tea plantation land with the objective of estabiishing communities of plants and animals closely resembling the original wildlife.
(7) There should be a policy of regeneration of forests in cardamom plantations, and no further extension of these plantations.
(8) Replacement of man-made forests comprising exotic trees (e.g. Pinus) by forests with indigenous tree species.
Education, organization, and research
(1) Education is necessary to improve the consciousness of people and activate their interests in preserving natural ecosystems, in particular forests.
(2) Conservation issues should come under one body, not, as at present, under the Department of Wildlife Conservation and the Forest Department. This body should comprise experts from all fields related to conservation.
(3) Better collaboration between universities and other institutions dealing with nature conservation.
(4) Further research on pocket sizes, feasibility of corridors in relation to habitat requirements, and on the ecology of the Sri Lankan animal and plant species. The research could be carried out under the NARESA-SAREC Zoological Survey of Sri Lanka. A similar survey on the vegetation should
132 w. Erdelen
be set up and geographic distributions of species, at least of some indicator species, should be constantly monitored.
These points give only an outline of the most pressing problems. Their solution requires effective collaboration between the different institutions dealing with the natural forests in Sri Lanka. It is hoped that the alarming situation outlined in this paper will make more people in Sri Lanka aware of the urgent and vital problem the country faces and that this paper will stimulate co-operation between those who still can do something to prevent the worst from happening. However, nature conservation has no future in Sri Lanka unless the population explosion is immediately arrested.
I gratefully acknowledge financial support for my studies in Sri Lanka by the Deutsche Forschungsge .meinschaft (German Science Foundation). For their hospitality, help, and many discussions I wish to thank Prof. Dr Breckenridge, Prof. Dr H. Crusz, Drs Gunatilleke, Dr N. Ishwaran, University of Peradeniya, and Dr (Mrs) W. T. T. P. Gunawardane, Director, National Museum, Colombo. The Department of Wildlife Conservation and the Forest Department of Sri Lanka gave permission to work in protected areas. Mr W. Konle provided assistance in the data analyses. Mr A. Palihawadana and Mr E. B. Wedanda assisted in the field. I am grateful to Prof. Dr Balasubramaniam, Prof. Dr H. Crusz, Prof. Dr J. Jacobs, Dr H. R. Siegismund, Dr W. Werner, and an anonymous referee for constructive criticism of the text.
Abeywickrama, B. A. (1956). The origin and affinities of the flora of Ceylon. Proc. Ann. Sess. Ceylon Ass. Adv. Sci., llth, 2, 99-121
Abeywickrama, B. A. (1959). The evolution of the flora of Ceylon. Proc. Ann. Sess. Ceylon Ass. Adv. Sci., 14th, 2, 217-19.
Adams, F. D. (1929). The geology of Ceylon. Can. J. Res., 1, 425-511 Andrews, J. R. T. (1971). A forest inventory of Ceylon. A Canada-Ceylon Colombo
Plan Project. Colombo, Government Press. Anon. (1978a). Forests--their role in human welfare and economic value. Econ.
Rev., 3(11), 3-15. Anon. (1978b). The Mahaweli Ganga Development Project. Econ. Rev., 4(8 & 9),
3-22. Balakrishnan, M. & Easa, P. S. (1982). Strategies for management of forests and
wildlife in Kerala. In Recent trends in forest management, 1-12. Trivandrum, India, Department of Forests, Government of Kerala.
Balasubramaniam, S. (1985). Tree flora of Sri Lanka. In Proc. Int. Conf. on Timber Technology, ed. by A. Jayatilleke, 58-67. Moratuwa, Sri Lanka.
Forest ecosystems and nature conservation in Sri Lanka 133
Brown, J. H. (1971). Mammals on mountaintops: Nonequilibrium insular biogeography. Am. Nat., 105, 467-78.
Crusz, H. (1973). Nature conservation in Sri Lanka (Ceylon). Biol. Conserv., 5, 199-208.
Crusz, H. (1984). Parasites of endemic and relict vertebrates: A biogeographical review. In Ecology and biogeography in Sri Lanka, ed. by C. H. Fernando, 321-51. The Hague, Martinus Nijhoff.
Crusz, H. (1986). The vertebrates of Sri Lanka: Endemism and other aspects. Rep. Soc. Res. Native Livestock (Jap.), 11, 65-80.
Crusz, H. & Daundasekara, D. M. A. K. (in press). Parasites of the relict fauna of Ceylon. IX. Helminths from burrowing scinine lizards. Annls Parasit. hum.comp.
Crusz, H. & Nugaliyadde, L. (1978). Parasites of the relict fauna of Ceylon, VII. General considerations and first host-parasite checklist. C. R. Soc. Biogeogr., 477, 85-106.
DeAlwis, L. (1969). The National Parks of Ceylon. A guide. Colombo, Government Press.
Department of Census and Statistics (1985). Statistical pocket book of the Democratic Socialist Republic of Sri Lanka. Colombo.
Deraniyagala, P. E. P. (1953). A colored atlas of some vertebrates from Ceylon, 2. Tetrapod Reptilia. Colombo, Government Press.
DeRosayro, R. A. (1960). The nature and origin of secondary vegetational communities in Ceylon. In UNESCO Symposium on the impact of man on humid tropics vegetation, 279-306. Goroka.
Dittus, W. P. J. (1977). The ecology of a semi-evergreen forest community in Sri Lanka. Biotropica, 9, 268-86.
Domr6s, M. (1976). Sri Lanka. Die Tropeninsel Ceylon. Wiss. l_~nderk., 12. Darmstadt, Wiss. Buchgesellschaft.
Eisenberg, J. F. & Lockhart, M. (1972). An ecological reconnaissance of Wilpattu National Park, Ceylon. Smithson. Contrib. Zool., 101, 1-118.
Erdelen, W. (1984). The genus Calotes (Sauria, Agamidae) in Sri Lanka: Distribution patterns. J. Biogeogr., 11, 515-25.
Fernando, A. B. (1983). A new concept in the conservation of Sri Lanka elephants. Loris, 16, 108-11.
Fernando, S. N. U. (1968). The natural vegetation of Ceylon. Colombo, Swabasha Printers.
Gaussen, H. P., Legris, P., Viart, M. & Labroue, L. (1965). International map of the vegetation and of environmental conditions. Colombo, Government Survey Department.
Gaussen, H. P., Legris, P., Viart, M. & Labroue, L. (1968). Explanatory notes on the vegetation map of Ceylon. Colombo, Government Press.
Geiser, U. & Sommer, M. (1982). Up-to-date information on Sri Lanka's forest cover. Loris, 16, 66-9.
Greer, A. E. (1970). A subfamilial classification of scincid lizards. Bull. Mus. Comp. ZooL, 139, 151-84.
Greller, A. M. & Balasubramaniam, S. (1980). A preliminary floristic-climatic classification of the forests of Sri Lanka. Sri Lanka Forester, 14(3 & 4), 163-9.
Gunatilleke, C. V. S. & Gunatiileke, I. A. U. N. (1983). A forestry case study of the Sinharaja rainforest in Sri Lanka. In Forest and watershed development and conservation in Asia and the Pacific, ed. by L. S. Hamilton, 289-357. Boulder, Colorado, Westview Press.
134 W. Erdelen
Gunatilleke, C. V. S. & Gunatilleke, I. A. U. N. (1985). Phytosociology of Sinharaja--a contribution to rain forest conservation in Sri Lanka. Biol. Conserv., 31, 21-40.
Gunatilleke, I. A. U. N. & Gunatilleke, C. V. S. (1984). Distribution of endemics in the tree flora of a lowland hill forest in Sri Lanka. Biol. Conserv., 28, 275-85.
Henry, G. M. (1978). A guide to the birds of Ceylon. Kandy, K. V. G. DeSilva & Sons. Hoffmann, T. W. (1978). The Mahaweli diversion project. Loris, 14, 282--4. Hoffmann, T. W. (1984). National red data list of endangered and rare birds of Sri
Lanka. Colombo, Ceylon Bird Club and Wildlife and Nature Protection Society of Sri Lanka.
Holmes, C. H. (1951). The climate and vegetation of the Dry Zone of Ceylon. Bull. Ceyl. Geogr. Soc., 5, 145-53.
Holmes, C. H. (1958). The broad pattern of climate and vegetational distribution in Ceylon. Ceyl. Forester, N.S., 2, 209-25.
Iriyagolle, G. (1978). The truth about the Mahaweli. Nugegoda, Sri Lanka, published by the author.
IUCN (1980). Worm conservation strategy: Living resource conservation for sustainable development. Gland, IUCN.
Koelmeyer, K. O. (1957). Climatic classification and the distribution of vegetation in Ceylon. Cey. Forester, N.S., 3, 144-63.
Koelmeyer, K. O. (1958). Climatic classification and the distribution of vegetation in Ceylon. Ceyl. Forester, N.S., 4, 265-88.
Mahaweli Development Board (1977). Mahaweli Ganga development in Sri Lanka. Summary reports and projects. Colombo.
McKay, G. M. (1984). Ecology and biogeography of mammals. In Ecology and biogeography of Sri Lanka, ed. by C. H. Fernando, 413-29. The Hague, Martinus Nijhoff.
Meijer, W. (1982). Plant refuges in the Indo-Malesian Region. In Biological diversification in the tropics, ed. by G. T. Prance, 576--84. New York, Columbia University Press.
Mueller-Dombois, D. (1968). Ecogeogi'aphic analysis of a climate map of Ceylon with particular reference to vegetation. Cey. Forester, N.S., 8, 1-20.
Nanayakkara, V. R. (1982). Forests--policies and strategies for conservation and development. Sri Lanka Forester, 15, 75-9.
Peeris, C. V. S. (1975). The ecology of endemic tree species in Sri Lanka in relation to their conservation. PhD thesis, University of Aberdeen.
Pemadasa, M. A. (1984). Grasslands. In Ecology and biogeography ofSri Lanka, ed. by C. H. Fernando, 99-131, The Hague, Martinus Nijhoff.
Perera, N. P. (1975). A physiognomic vegetation map of Sri Lanka (Ceylon). J. Biogeogr., 2, 185-203.
Perera, N. P. (1984). Natural resources, settlements and land use. In Ecology and biogeography of Sri Lanka, ed. by C. H. Fernando, 453-92. The Hague, Martinus Nijhoff.
Phillips, W. W. A. (1984). Manual of the mammals ofSri Lanka. Colombo, Wildlife and Nature Protection Society of Sri Lanka.
Poore, D. (1979). The values of tropical moist forest ecosystems and the environmental consequences of their removal. Sri Lanka Forester, 14, 15-36.
Saparamadu, S. D. (1983). Introduction. In A history of birds of Ceylon, 1, by V. Legge, vii-lxxxvii. Dehiwala, Tisara Prakasakayo Ltd.
Forest ecosystems and nature conseroation in Sri Lanka 135
Senanayake, F. R. (1977). Habitat values and endemicity in the vanishing rain forests of Sri Lanka. Nature, Lond., 268, 568.
Senanayake, F. R., Soul6, M. & Senner, J. W. (1977). Habitat values and endemicity in the vanishing rain forests of Sri Lanka. Nature, Lond., 265, 351-4.
Sledge, W. A. (1982). An annotated check-list of the Pteridophyta of Ceylon. Bot. J. Linn. Soc., 84, 1-30.
Sri Lanka socio-economic data (1984). Colombo, Statistics Department, Central Bank of Ceylon.
Terborgh, J. (1975). Faunal equilibria and the design of wildlife preserves. In Ecological studies, 11. Tropical ecological systems, ed. by F. B. Golley and E. Medina, 369-80. New York, Springer.
Vancuylenberg, B. W. B. (1977). Feeding behaviour of the Asiatic elephant in south- east Sri Lanka in relation to conservation. Biol. Conserv., 12, 33-54.
Werner, W. L. (1982). The upper montane rain forests of Sri Lanka. Sri Lanka Forester, 15, 119-35.
Werner, W. L. (1984). Die H6hen- und Nebelwiilder auf der Insel Ceylon (Sri Lanka). Wiesbaden, Steiner.
Werner, W. L. (1985). Naturreservate im Hochland der Insel Ceylon (Sri Lanka): Peak Wilderness und Horton Plains. Natur u. Museum, 115, 65-76.
Wilcox, B. A. (1980). Insular ecology and conservation. In Conservation biology. An evolutionary-ecologicaiperspective, ed. by M. E. Soul6 & B. A. Wilcox, 95-117. Sunderland, MA, Sinauer Associates.