Version classiqueVersion mobile

Forest landscapes of the southern western Ghats, India

B.R. Ramesh
Rajan Gurukkal

Part II. Analysing The Landscape: The Anamalai Region (Nelliyampathi Hills)

5. Richness of Vertebrates and Spatial Features of Selected Large Mammals and Arboreal Fauna

M. Balasubramanian, Pierre Couteron et P.V. Karunakaran

Texte intégral

1Relating geographic distribution of animal species with spatial patterns of environmental features is one of the major goals of landscape ecology. It carries considerable considerable significance in terms of conservation planning. Abundance and distribution of animal species are known to strongly depend on the spatial patterns of suitable vs unsuitable habitats. For many species, habitat suitability is in fact not defined by the properties of a particular vegetation type, but rather by the characteristics of the mosaic integrating several vegetation types and/or land use units. Further, the scale at which the mosaic is perceived by a given animal species is determined by the size of its home range. Hence, it would be unrealistic to venture carrying out an assessment of landscape suitability for hundreds of animal species having body sizes and home range requirements differing by several orders of magnitudes along with contrasted activity patterns and life history traits.

2Consequently, we adopted a two-level approach, which aimed at providing overall information about the “faunal wealth” of the landscape units under study as well as detailed information about a limited number of “scheduled” charismatic species and/or of wider ecological significance (rare, endangered and threatened and/or endemic). The overall assessment of faunal richness was carried out on the basis of secondary information. Fieldwork was then undertaken to get firsthand information on variations of species’ abundances across the study area and to analyze such variations in relation to vegetation types, landscape patterns, and level of human disturbance. Many other fauna species with wider ecological significance can be brought into conservation planning after acquiring management inputs based on a thorough knowledge about their abundance. Hence we selected eleven, emblematic species with very diverse life history characteristics for an in-depth study of the abundance and distribution in the study area.

Overall richness of vertebrate fauna and their status

  • 1 IUCN. 2002. Red List 2002 Information Kit. (See search-expe (...)
  • 2 CITES. 2005. Appendices I, II and III.
  • 3 Government of India (GOI). 2005. The Wildlife (Protection) Act, 1972, New Delhi.

3A compilation of information on five vertebrate faunal classes was carried out with the help of field surveys and secondary sources (Shaji et al, 2000; Gopi, 2000; Daniels, 2001; Islam and Rahmani, 2004). The species were assigned conservation and management status based on IUCN (International Union for the Conservation of Nature and Natural Resources)1 criteria, Appendices of CITES2 (Convention on International Trade in Endangered Species) and schedules of W (P) A, 1972 (Wildlife (Protection) Act) 1972)3. The detailed information of all the vertebrate fauna recorded in the study area is given in the CD provided at the back cover of the book.

4Kerala being part of the southern Western Ghats almost represent more than 75 % of the vertebrate species belonging to mammals, birds, reptiles and amphibians recorded from the Western Ghats (Figure 5.1). The study area situated in the Western Anamalai region is one of the major biogeographic regions within the Western Ghats, represent more than 36 % of the vertebrate fauna recorded from Kerala, particularly birds, account to almost 70 % of the species for the State (Figure 5.1). Among reptiles, 38 species of the recorded species in the study area are snakes.

5Among the endemic vertebrate fauna of the Western Ghats found in the landscape units, fishes accounts to more number of species (53 %) followed by amphibians (50 %) and reptiles (40 %) (Table 5.1). The presence of extensive water bodies such as reservoirs and network of perennial streams in the study area is one of the main factors responsible for high rate of endemism in aquatic and semi aquatic group of fauna. In mammals only six species are endemic viz Nilgiri langur, lion tailed macaque, Nilgiri tahr, Nilgiri marten, brown palm civet and jungle striped squirrel.

Figure 5.1 Proportion of vertebrates fauna found in the landscape units in
comparison with Kerala and the Western Ghats

Table 5.1 Details of vertebrate fauna found in the landscape units

Table 5.1 Details of vertebrate fauna found in the landscape units

6The conservation and protection status of the vertebrate fauna according to ICUN, CITES and W (P) A is summarized in Table 5.1. The IUCN classification system is primarily intended to assign the status to the species that are at high risk of extinction. The fundamental data used in developing IUCN criteria is population status and distribution pattern of species. In the study area 141 species fall under 4 categories of IUCN. Of this, only 7 are critically endangered, which are represented by 1 mammal (wild dog), 3 birds (Nilgiri wood pigeon, broad tailed grass warbler and white bellied shortwing) and 3 fishes (Ticto barb, Osteochilichthys longidorsalis Pethiyagoda and Danio malabaricus).

7W (P) A, which is a national legal frame work, provides legal status to the threatened species as “schedules (I-VI)” in order to protect them from hunting and over exploitation. The species come under the first 4 schedules (I to IV) are prohibited from hunting, schedule V for vermin and VI for plant species that are not to be over exploited. In the study area, 335 species come under scheduled (I to IV) species, of which, birds account to 239.

8CITES referred to an international agreements between the governments to ensure that the trade in specimens of wild animals and plants does not threaten their survival. The species covered by CITES are listed in three appendices according to the degree of restriction in trade they need. Appendix I include species threatened with extinction where trade is permitted only in exceptional circumstances. For species listed under appendix II, trade must be controlled in order to avoid incompatible utilization with the survival and appendix III contains species which are protected at least in one country which asked CITES Parties for assistance in controlling the trade. In the study area 85 species represent one or other appendices of CITES.

9In spite of legal status assigned under IUCN, CITES and W (P) A to the species that are threatened, we found inconsistency in giving prioritization for conservation. For example, 15 mammal species found in the study area were enlisted in the threat categories (critically endangered, endangered and vulnerable) of IUCN whereas only 12 species have been given the most prioritized legal status (Schedule I). Contradictory to this altogether only four species of birds are listed under the categories such as critically endangered and vulnerable, 27 species were listed in Schedule I of W (P)A. Among reptiles, out of the 16 and 12 species listed in threat categories of IUCN and appendices of CITES respectively, only one species (Indian rock python) has been accorded schedule I status. In the case of amphibians and fishes though considerable number of species are listed under both IUCN categories and CITES appendices, no species have been placed in the first three schedules of the W (P)A. These incongruities in conservation and protection status may be an issue, which needs to be immediately structured and harmonized.

Density and distribution of selected fauna

10To study the variation in abundance and distribution patterns of fauna, eleven species belonging to mammals and birds were selected. These species are either confined to specific habitat or cover wider area with multiple habitats as their home range. Moreover, all the selected species fulfill one or other criteria of IUCN, CITES or W (P)A. The selected species are grouped under ungulates (elephant, gaur, sambar and Nilgiri tahr) predators (tiger, leopard and sloth bear) and arboreal fauna (Nilgiri langur, Malabar giant squirrel, lion tailed macaque and great Indian hornbill).


Field estimation of species density

11Distance sampling is a widely used methods for estimating the density/abundance of biological populations. One of the main methods in distance sampling is line transects. Preliminary survey indicated that estimating abundance through direct sightings method (using line transect) could be difficult for all the selected species due to dense vegetation and undulating terrain. Thus it was decided to use direct sightings along with indirect evidences for arboreal animals (such as Nilgiri langur, lion tailed macaque, Malabar giant squirrel and great Indian hornbill) and only indirect evidences for other species.

12A specific effort was devoted to Nilgiri tahr for which all potential locations of populations in the landscape units were identified and visited to collect information on population through direct and indirect evidences. The extent of habitat, availability of fodder species and the threats faced by both species as well as habitat were also recorded.

13A total of 82 transects were laid in different habitat viz evergreen, semi-evergreen, moist deciduous and plantations in the landscape units (Figure 5.2). The locations of all the transects and individual sightings were recorded using Geographical Positioning System. The length of transects varied from 1 to 5 km, and mean length is 2.6 km, depending on the terrain and habitats. While traversing through the transect line, both direct sightings of animals as well as the indirect evidences were recorded. On sighting a species or group of species, information such as species, number of animal (s), time of sighting, perpendicular distance from the transect, activity of the animal (s) and the micro habitat where the animal (s) being sighted were recoded. The indirect evidences such as dung piles/pellets/scats, scrapes/scratches and pugmarks (tiger and leopard) were noted down along with the species, perpendicular distance from the transect and habitat of the object (s) sighted. In the case of tigers, the size of the pugmark was also noted down to distinguish the individuals present in the region.

Figure 5.2 locations of trasects for estimating density and distribution of selected fauna

Analysis of transect data

  • 4 Source: Wildlife Census (2000), Kerala. Report published by Kerala Forests and Wildlife De (...)

14The data collected through the transect method were analyzed to estimate species density using Distance Version 5 (Beta) 4 (Thomas et al, 2005). Transects which fall on the same habitat was pooled and the data were segregated based on the species and microhabitat. The dung densities of elephants in each habitat were used to calculate the actual elephant density using the decay and defecation rates4. The dung and pellet density of gaur and sambar respectively were also tabulated but no density inference was attempted. All these results were spatially represented with respect to vegetation types based on one-year data.

15For other species, occurrences were too scarce to allow a systematic analysis using the Distance software. Both direct sightings and indirect records were nevertheless analyzed with respect to habitat and/or geographic distribution between the landscape units.



Elephant (Elephas maximus)

16The Asian elephant is an endangered species (IUCN) and included in Appendix I of the CITES. The distribution is limited by both the need for daily access to water, and by a preference for feeding on grass (Sukumar, 1996). Due to its large size and extensive spatial requirement, the elephant can be considered a "flagship" species, whose conservation could potentially promote maintenance of biological diversity and ecological integrity on a large scale. The evidences from the transects indicated that they are distributed throughout the landscape units (Figure 5.3a).

17The estimated dung density and the elephant density in each vegetation types for both the landscape units are given in Tables 5.2 and 5.3. The overall dung density and elephant density in LU13 is 843 dung piles/km2 and 0.96 animals/km2 respectively, whereas in LU16, it is 701 and 0.72. But such a difference in elephant density between the landscape units was not signifycant according to the 95 % confidence intervals provided by the Distance software for the estimated densities of dung piles (Table 5.2).

18Furthermore, the relationship between vegetation types and estimated densities was fairly consistent for the two landscape units. In both of them, densities tended to be higher in evergreen forests (EVG) than in moist deciduous forests (MDF) and teak plantation (TP), but the difference was not significant in LU13 (Table 5.2). Semi-evergreen forest (SEVG) showed contrasted patterns of estimated densities in the two units, but this may not be interpretable due to the low extent of this vegetation type all over the study area.

Table 5.2: Details of elephant dung analysis in LU13 and LU16

Table 5.2: Details of elephant dung analysis in LU13 and LU16

Table 5.3 Estimated density of elephant in LU13 and 16

Table 5.3 Estimated density of elephant in LU13 and 16

Figure 5.3 The distribution of selected mammals based on direct and indirect evidences

Figure 5.4 The density of elephants based on vegetation types

19In spite of contrasting levels of human activity and disturbance in the two landscape units, the overall relationship between elephant densities and vegetation is not varying much. However, the absolute density of elephants proved slightly higher in the LU13 (Figure 5.4), which shares longer boundaries with human habitation and cultivated areas. In LU16, the density is largely medium, except in the areas with predominantly teak plantation (Parambikulam) where the density is relatively low.

Gaur (Bos gaurus)

20The gaur is classified as a vulnerable species (IUCN) and included in Appendix I of CITES. Populations of gaur in these tracts are adversely affected due to habitat loss or degradation and epidemics (Choudhury, 2002). The data on dung distribution from the transects indicate that the animal is found throughout the study area except near the human settlements (Figure 5.3b).

21The estimation of dung density in the study area (Table 5.6) revealed that maximum density was found in the evergreen forests interspersed with patches of grasslands (GL) (582 dung piles/km2) followed by semi-evergreen (329/km2) and moist deciduous forests (271/km2) and low in teak plantations (259/km2). The dung density according to vegetation types was classified into three categories (low: < 300 dung piles/km2; medium: 300 to 500 dung piles/km2; high: > 500 dung piles/km2), which are depicted in the Figure 5.5. The higher densities in the evergreen forests could be due to the availability of fodder and sheltering effect of the vegetation compared to the moist deciduous and teak plantations.

Table 5.4 Details of gaur dung analysis

Table 5.4 Details of gaur dung analysis

Figure 5.5 The dung density of gaur based on vegetation types

Sambar (Cervus unicolor)

22The largest and the most widely spread Indian deer, the sambar, found in all the habitat types, from moist deciduous to evergreen forests (Prater, 1971). This species is one among the main prey species for tiger, panther and wild dog. The evidences of sambar from the observations indicated that they are found throughout the landscape units except nearby human settlements (Figure 5.3c). The pellet density analysis in different vegetation types revealed that the abundance is highest in evergreen forests (3.15 pellets/km2) and teak plantations (2.91/km2), while semi-evergreen (1.5/km2) and moist deciduous forests (1.13/km2) display low density (Table 5.5). The pellet density was classified into three categories based on vegetation viz low: < 1.5 pellets/km2; medium: between 1.5 and 3 pellets/km2; high: >3 pellets/km2 (Figure 5.6).

Table 5.5 Details of sambar pellet analysis

Habitat Number of transects Total length (km) No. of samples Density of pellets/km2
EVG+GL 49 110.50 212 3.15
SEVG 4 7.00 21 1.5
MDF 36 65.10 147 1.13
TP 14 33.05 169 2.91
Overall 103 215.65 549 1.30

23The higher density of pellets in the evergreen forests may be due to the preference of habitat for feeding and resting. Moreover, the deciduous forests and monocultures are often getting burnt and subsequently resulted in higher ground biomass just after the rain, which generally attracts the herbivores only during a particular period.

Nilgiri tahr (Hemitragus hylocrius)

24The Nilgiri tahr is an endangered ungulate (IUCN) found in the high altitude mountains of southern Western Ghats and endemic to the hill ranges of Tamil Nadu and Kerala. They inhabit the montane grasslands with steep ridges and rocky cliffs. Although a viable population of the species is found in Eravikulam National Park, Kerala, isolated small populations exist in patches of habitats all along the southern Western Ghats, especially in Kerala (Rice, 1984).

Figure 5.6 The dung density of sambar based on vegetation types

25The landscape units have 12 small isolated populations spread over most of the forest divisions (Figure 5.7). Among these, the largest populations are found in Karimalagopuram (between 30-40 individuals), Padagiri (Hilltop) (30-40) and Manjakkallan (30-50). In areas such as Charpapadam, Pandimudi and Minampara, the population ranges between 10 and 15 individuals and in the remaining areas less than 10 individuals (Table 5.6).

26The important fodder species found in all the locations include Heteropogon contortus, Andropogon lividus, Arundinella purpurea, Ischaemem indicum, Tripogon spp, etc.

27The major threat to the tahr populations in the landscape units is poaching. Cattle grazing and alteration of habitat for commercial and forestry plantations exert severe impact on these scattered populations. Since most of these grassland patches are situated near to the estates where surveillance infrastructures are not adequate the poaching pressure is more on the species. Repeated man-made fire, which is a common phenomenon in these areas, also reduces the availability of fodder species. Hence sustenance of these populations depends on the protection strategies implemented in these areas.

Table 5.6 Nilgiri tahr populations in the landscape units

Division Location Longitude Latitude Population1
Parambikulam Karimalagopuram 76° 44′ 40′′ 10° 21′ 58′′ 30-40
Parambikulam Vengoli 76° 48′ 56′′ 10° 25′ 19′′ 6-10
Parambikulam Pandaravarai 76° 49′ 33′′ 10° 29′ 48′′ 5-10
Vazhachal Charpapadam 76° 35′ 20′′ 10° 21′ 37′′ 10-15
Chalakudy Pandimudi 76° 31′ 38′′ 10° 24′ 06′′ 10-15
Malayattoor Sulimala 76° 51′ 17′′ 10° 13′ 21′′ <5
Malayattoor Vagirian 76° 49′ 41′′ 10° 16′ 01′′ 5-10
Malayattoor Manjakkallan 76° 53′ 02′′ 10° 15′ 26′′ 30 - 50
Nemmara Minampara 76° 43′ 46′′ 10° 32′ 08′′ 10-15
Nemmara Padagiri (Hilltop) 76° 39′ 21′′ 10° 30′ 09′′ 30-40
Nemmara Kottangadi estate uphill 76° 43′ 00′′ 10° 30′ 37′′ 5-10
Nemmara Korakkunnumala 76° 41′ 21′′ 10° 18′ 43′′ 5-10

1Estimation based on indirect evidence

Figure 5.7 Distribution of Nilgiri tahr, lion tailed macaque and great Indian hombill in the landscape units


Tiger (Panthera tigris)

28The tiger is one of the most charismatic and evocative species on earth. It is also one of the most threatened species (IUCN) listed in Appendix I of CITES. Tiger has been an important flagship species for biodiversity and wildlife conservation in India over the past thirty years. As top predator, they keep populations of wild ungulates in check, thereby maintaining the balance between herbivores and the vegetation upon which they feed. Tiger populations are threatened throughout their range, either directly from poaching, or from habitat and prey loss (Karanth, 2001). In many places, they struggle for survival with burgeoning human populations competing for similar resources of food and shelter. Habitat destruction reduces both tiger and its prey. As a result tiger, though they are very shy, are forced to move into human settlements in search of food, where they are more likely to get killed.

29Though the total number of tigers in the landscape units is unknown, it was estimated that Parambikulam has around 7 to 8 individuals. However, a large number of indirect evidences were also found in the adjacent areas especially in Vazhachal, Kollathirumedu and Sholayar Ranges (Figure 5.8a) The observations on individual pugmarks sighted in different regions of the landscape units indicate that there could be a minimum of 20 individuals in the entire study area.

Leopard (Panthera pardus)

30The leopard has the most widespread distribution of all the cats found in India (Prater, 1971). Leopards are generalist predators and are able to live in any habitat where there is sufficient cover and prey, ranging from open savanna to wet evergreen forests and even on the outskirts of urban areas. They are placed on Appendix I of CITES and Schedule II of W (P) A, 1972. The indirect evidences collected during the study revealed that the species is distributed throughout, including regions nearby human settlements (Figure 5.8b). Cattle lifting by leopards in the surrounding areas of the landscape units is frequent. However, this species is known to be highly adaptive and to change its habits consistently with the changes that take place in the habitat.

Figure 5.8 Location of evidences of tiger, leopard, Nilgiri langur and Malabar giant squirrel

Sloth bear (Melursus ursinus)

31The sloth bear is an endemic and vulnerable (IUCN) species of the Indian subcontinent listed in Appendix I of CITES and Schedule I of the W (P) A, 1972. They are found in a variety of habitats ranging from wet evergreen forest to deciduous and degraded scrub forests. The distribution of sloth bear in the study area is basically decided by the availability of food such as ants, termites, honey and fruits. Degradation of the habitat in the form of overgrazing, tree felling, fire, conversion and reclamation for other uses, and over-extraction of forest resources, which are essential for sloth bear survival, appear to be occurring throughout the bear ranges (Yoganand et al, 1999). Hence most of the sloth bear populations in India are declining (Garshelis et al, 1999). The indirect evidences of species in the landscape units revealed that they are abundant in Parambikulam WLS and in the evergreen forests of Vazhachal and Nemmara divisions (Figure 5.3d). It was also observed that their movements were very rare near human settlement and regions dominated with reed brakes.

Arboreal species

Nilgiri langur (Trachypithecus johnii)

32The Nilgiri langur, an endemic species to the rainforests of the Western Ghats of Tamil Nadu, Kerala, and Karnataka, is a threatened species (IUCN). They inhabit in wet, moist as well as relatively dry forests (Singh et al, 1997). This species is sympatric with the lion-tailed macaque (Singh et al, 2000), common langur, and bonnet macaque, (Poirier, 1969). The distribution of Nilgiri langur in the study area indicated that though they exist in almost all habitats, they are more abundant in evergreen forests (Figure 5.8c) and less in forests lying near to the human habitations.

Lion tailed macaque (Macaca silenus)

33Lion tailed macaque (LTM), considered as one of the most important habitat specialist primates of India is endemic to the Western Ghats. The distribution of the species is restricted in the tropical evergreen forests of Kerala, Karnataka and Tamil Nadu. This endangered species (IUCN) is listed in Schedule I of W (P) A, 1972. LTM is mainly arboreal, although it does occasionally descend to the ground. Historically the populations of LTM have been reduced in the Western Ghats due to capturing for pet trade, zoos, and oriental medicine (Kumar, 1995). Their population in the present range is severely fragmented and isolated due to conversion of medium elevation evergreen forests into reservoirs and forestry and commercial plantations (Umapathy and Kumar, 2000). Studies proved that the isolated populations have low birth rate and lower proportion of immature animals. Given this characteristic of the species it does not have the ability to recover from even low levels of poaching (Kumar et al, 1995).

34In the landscape units the LTM troops have been observed in evergreen forests of Parambikulam, Vazahchal, Malayattoor and Chalakudy divisions (Figure 5.7). Few sightings from the abandoned tea estates and adjoining forest patches in Nelliyampathi hills revealed the existence of isolated populations in the man-modified habitats, though they receive very low protection. The presence of extensive patches of almost undisturbed medium elevation evergreen forests (Cullenia-Mesua-Palaquium type) in the Parambikulam-Vazhachal region provides an ideal habitat for the species, since they mostly feed on fruits and young leaves of Cullenia exarillata (Ramachandran and Joseph, 2000).

Malabar giant squirrel (Ratufa indica)

35Malabar giant squirrel is a vulnerable species (IUCN), though commonly found throughout the forests in the State. The species is in Schedule II of the W (P) A, 1972 and listed in Appendix II of CITES. In the study area the species has been sighted throughout the landscape (Figure 5.8d). This arboreal animal is fairly common in all types of habitats viz, moist deciduous, semi-evergreen, and evergreen forests. Very low abundance was observed in the regions near human settlements.

Great Indian hornbill (Buceros bicornis)

36Great Indian hornbill is a large frugivorous bird of the evergreen forests of the Western Ghats of southern India. It is a near threatened species (IUCN). The nesting and foraging habitat studies conducted elsewhere in the State (Harikumar et al, 2001) revealed that the hornbills depend on pristine, undisturbed and mature stands (old growth) of evergreen forests. Heavy destruction of primary evergreen forests coupled with poaching of adults and squabs from nests over the past few decades have been coincident with the declines in number and range of this species (Divya and Kannan, 1997).

37Within the study area they were mainly sighted in the primary evergreen forests of Vazhachal and Malayattoor divisions (Figure 5.7). The forests of these tracts are characterized with tall tress (more than 35 m height compared to the partially and moderately disturbed evergreen forests (Table 5.7). Few sightings were also recorded from Nelliyampathi, Parambikulam and Chimmony in the similar forest type. Studies on the diet of hornbill (Divya and Kannan, 1997; Harikumar et al, 2001) have proved that figs (Ficus spp.) constitute the main species. It was found that Ficus is one of the dominant plant genus in the landscape unit with 21 species. Among the species, Ficus nervosa, a tall tree, is fairly common in the study area and recorded from many sample plots (>50 % of recorded ficus individuals). Other less frequent species are Ficus amplissima, F. beddomei, F. microcarpa, F. talbotii, F. tsjahela and F. virens.

Table 5.7 Number of trees with different height categories in various types of evergreen forests.

Table 5.7 Number of trees with different height categories in various types of evergreen forests.

38The dominant canopy or emergent tree species where hornbills were frequently sighted are Aglaia jainii, Antiaris toxicaria, Aphanamixis polystachya, Calophyllum polyanthum, Cullenia exarillata, Dysoxylum malabaricum, Mesua ferrea, Myristica dactyloides and Palaquium ellipticum.

Comparison of geographical distributions of selected species

39We compared spatial patterns of abundance of the different scheduled species in order to appraise their relative habitat selectivity and their tolerance to human disturbance. Transect-based measures of abundance cannot be directly compared across species, since they rely on very diverse field records (e.g., direct vs indirect evidences), with no direct link to absolute densities. However, we make the assumption that, for a given species, indirect measures of abundances are fairly comparable through space due to fairly equivalent observation efforts all over the study area (Figure 5.2). From this, comparing spatial distributions of scheduled species can make sense, provided that abundance measures are standardized to unit prior to inter-species comparisons. Furthermore, a synthetic picture of inter species comparisons can be achieved using a multivariate approach and we used Correspondence Analysis - CA (Legendre and Legendre, 1998), which integrates the standardization of measures of species abundances.

40All the scheduled species have been included in the analysis except Nilgiri tahr, which has restricted distribution. A total of 752 spatial samples (XY coordinates) consists of 3219 observations (both direct and indirect) were used for the analysis (Table 5.8). For elephant, gaur and sambar, all the sightings in an individual transect have been considered as a single sample “located” at the XY coordinates of the transect mid-point. For other species, each observation was taken as a sample. The entire study area was divided into grid cells (geographical quadrates) of 0.05° x 0.05° and a contingency table cross classifying samples in terms of species and grid cells were computed. Such a table featuring 110 grid cells (rows) and 11 species (columns) was eventually submitted to CA in order to compare species relative abundances within the study area.

Table 5.8 Animal species considered for correspondence analysis with sample size

Species Number of samples Total number of observation
Elephant 82 1586
Gaur 82 575
Sambar 82 552
Great Indian Hornbill 33 33
Leopard 49 49
Lion tailed macaque 30 30
Malabar Giant Squirrel 155 155
Nilgiri langur 90 90
Sloth Bear 82 82
Tiger 51 51
Grand Total 752 3219

41The analysis yielded two prominent axes (featuring together 38 % of the table variance or inertia). The first one sorted out a limited number of grid cells (right-side of Figure 5.9a and b) which are characterized by a high relative density of species that were observed in small and specific areas of the landscape units, such as great Indian hornbill, lion-tailed macaque, Nilgiri langur and Malabar giant squirrel. In contrast, the majority of grid cells was located towards the left-side of the axis which was characterized by the relative importance of species with broad distribution within the LUs, namely, elephant, gaur and sambar. Overall, this axis can be interpreted (Figure 5.9b) as ranking the ten scheduled species on the basis of the geographic amplitude that they displayed in the area, that is elephant, gaur, sambar, leopard (broad amplitude), sloth bear, tiger, and Nilgiri langur (intermediate), and the rest of species (limited to very small amplitude).

Figure 5.9 Results of CA: a) Plot of grid cells against scores along the
two main CA axes (with mention of the results of the clustering,
see text). b) Same for species (LTM: lion-tailed macaque;
MGS: Malabar ground squirel)

42We used CA scores to classify (k-means clustering method; Legendre and Legendre, 1998) grid cells into four classes expressing relative abundances of animal species. The first cluster (Figure 5.9a and 5.10) was linked to the relative abundance of elephant and leopard (in fact relative scarcity of all other species), the second (Fig 5.9a and 5.10) was related to the relative abundance of gaur and sambar, while the third cluster (Figure 5.9a and 5.10) was defined as the relative abundance of sloth bear, tiger, and Nilgiri langur. It is the second axis which contributed to separate the first cluster from the two others, thereby revealing subtle variations among grid cells characterized by the relative abundance of species with broad to intermediate geographic amplitude.

43The last cluster (Figure 5.9a and 5.10) was also characterised by a strong relative frequency of the two latter species with lion-tailed macaque, Malabar giant squirrel and great Indian hornbill, i.e. species which exhibited the smallest geographic amplitudes. Among these species, lion-tailed macaque and great Indian hornbill are known to be habitat-specific. This is particularly well-established regarding the great Indian hornbill (Divya and Kannan, 1997; Harikumar et al, 2001; Ramachandran and Joseph, 2000; Umapathy and Kumar, 2000).

44The map of the distribution of the different types of clusters provides an interesting, synthetic pictures of the overall patterns of space utilization by the ten scheduled species. The grid cells from LU13, in the Western part of the study area, belonged mostly to clusters 1 and 2 characterized by an over-representation of species with wide to intermediate geographic amplitude. On the other hand, the North-Eastern part, displayed species of intermediate to limited ecological amplitude (Figure 5.10).

Figure 5.10 Geographical distribution and relative abundance of selected faunal species in the landscape units


45The study area is very rich in faunal biodiversity especially the vertebrates. The study reveals that the distribution pattern of the species selected for study were not uniform throughout the landscape units.

46The distribution of all selected animals except leopard and Malabar giant squirrel showed a distinct pattern. The density of elephant, gaur and sambar are found to be higher in the places far from human habitations. Grasslands (vayals), a prime habitat for herbivores, were found to be used more intensively by gaur and other herbivores. But sedges, which are unpalatable to most of the herbivores, infest these vayals.

47The distinct patterns in the distribution of animals between the LUs are attributable to several factors. The degraded nature of the habitat and constant interaction of people in the lower reaches of the LUs, and especially in LU13, explains the low density of selected animals. Poaching of Nilgiri langur, which is prevalent in some part of the study area has considerably reduced the population. Distribution of predators such as tiger and leopard is mostly influenced by the distribution of prey species including gaur and sambar, whose abundance is higher in distant to settlements. Moreover, tiger in general avoids human dominated region, whereas leopard is highly adaptive to the local habitat conditions. The distribution of Nilgiri tahr, lion-tailed macaque and great Indian hornbill is restricted to certain areas in the landscape units, which has been explained by its niche characterized by primary evergreen forests and unique habitats (grasslands with cliffs).


1 IUCN. 2002. Red List 2002 Information Kit. (See search-expert.php to search for information on a particular species, including the current IUCN rating.)

2 CITES. 2005. Appendices I, II and III.

3 Government of India (GOI). 2005. The Wildlife (Protection) Act, 1972, New Delhi.

4 Source: Wildlife Census (2000), Kerala. Report published by Kerala Forests and Wildlife Department and Kerala Forest Research Institute.

Table des illustrations

Légende Figure 5.1 Proportion of vertebrates fauna found in the landscape units incomparison with Kerala and the Western Ghats
Fichier image/jpeg, 26k
Titre Table 5.1 Details of vertebrate fauna found in the landscape units
Fichier image/jpeg, 34k
Légende Figure 5.2 locations of trasects for estimating density and distribution of selected fauna
Fichier image/jpeg, 40k
Titre Table 5.2: Details of elephant dung analysis in LU13 and LU16
Fichier image/jpeg, 57k
Titre Table 5.3 Estimated density of elephant in LU13 and 16
Fichier image/jpeg, 43k
Légende Figure 5.3 The distribution of selected mammals based on direct and indirect evidences
Fichier image/jpeg, 94k
Légende Figure 5.4 The density of elephants based on vegetation types
Fichier image/jpeg, 47k
Titre Table 5.4 Details of gaur dung analysis
Fichier image/jpeg, 33k
Légende Figure 5.5 The dung density of gaur based on vegetation types
Fichier image/jpeg, 50k
Légende Figure 5.6 The dung density of sambar based on vegetation types
Fichier image/jpeg, 47k
Légende Figure 5.7 Distribution of Nilgiri tahr, lion tailed macaque and great Indian hombill in the landscape units
Fichier image/jpeg, 46k
Légende Figure 5.8 Location of evidences of tiger, leopard, Nilgiri langur and Malabar giant squirrel
Fichier image/jpeg, 110k
Titre Table 5.7 Number of trees with different height categories in various types of evergreen forests.
Fichier image/jpeg, 24k
Légende Figure 5.9 Results of CA: a) Plot of grid cells against scores along thetwo main CA axes (with mention of the results of the clustering,see text). b) Same for species (LTM: lion-tailed macaque;MGS: Malabar ground squirel)
Fichier image/jpeg, 32k
Légende Figure 5.10 Geographical distribution and relative abundance of selected faunal species in the landscape units
Fichier image/jpeg, 42k


© Institut Français de Pondichéry, 2007

Conditions d’utilisation :


Rechercher dans OpenEdition Search

Vous allez être redirigé vers OpenEdition Search