A total of 43 km of transect was done to estimate scat abundance. Scat abundance was generally low with an overall abundance of 1.28 scats/km. This can be explained by the absence of signs of civets in the sampled area, which generally accounts for the major part of scats in other survey sites. In total 55 scats were collected, with three scats of mongoose and 45 of cats with an abundance of 0.07 and 1.05 scats/km respectively. Cats were thus the most abundant species in this area. The interesting finding here is the presence of Nilgiri marten (not seen, but reported by others). Seven scats, probably of this species, were found along the transects, giving an abundance of 0.16 scats/km. The unusual greater abundance of cat scats may be due to the difficulties in distinguishing these with those of the jackals, which is common in this area. A total of 92.5 km was walked to record direct sightings with one sighting each of brown and stripe-necked mongoose. The shola-grasslands had an overall abundance of 1.41/km, and the wattle plantation 0.96. Cats were the most abundant. All the animals had a higher abundance in shola-grasslands than in plantations. The high abundance of scats in plantations may be due to two reasons: either interspersion of plantations with natural habitats, or misidentification with jackal scats, which were common in plantations.

The reserved forests in Moyar valley were surveyed, covering 50 km for estimating scat abundance and 69 km for direct sightings. Only one vegetation type, the dry thorn forest, was surveyed here. The abundance of scats was relatively low (0.26 scats/km), with those of cats being the most abundant (0.12 scats/km). There was only one sighting of a ruddy mongoose.

A total of 63 km were surveyed in this Sanctuary for estimating scat abundance, and 81 km for direct sightings. Three vegetation types were surveyed. Overall abundance was low (0.22 scats/km), those of mongoose being the most abundant (0.16 scats/km). Among the three vegetation types, the dry deciduous forest had the highest abundance (0.4 scats/km), especially for mongoose (0.32 scats/km). The abundance of scats in the moist deciduous forest and teak plantations was very low (0.08 and 0.11 scats/km respectively). Even though 81 km was covered for direct sighting, no small carnivore was seen.

This Sanctuary has several vegetation types of which scrub forest, dry deciduous forest, moist deciduous forest and plantations were included in the survey, which covered 105 km for estimation of scat abundance and 111 km for direct sightings. The abundance of scats, for all vegetation types together was low (0.18 scats/km), mongoose scats being the most abundant (0.1 scats/km). The scrub forest had the highest abundance (0.42 scats/km), both mongoose and cats being equally abundant (0.18 scats/km each). No scats were seen in 4 km of transects in the teak plantations. There were four sightings of small carnivores, one each of ruddy mongoose, jungle cat, stripe-necked mongoose, and small Indian civet.

The reserved forests in Sigur plateau were surveyed for a total of 52 km for estimating scat abundance, and 73 km for direct sightings. Scrub forests covered the entire area. The overall scat abundance was 0.42 scats/km, that of mongoose being the most abundant (0.25 scats/km) followed by cats (0.15 scats/km) and civets (0.02 scats/km). There was only one sighting, that of a ruddy mongoose.

The mixed deciduous forests in Pillur were surveyed for a total distance of 45 km for estimating scat abundance, and 67 km for direct sightings. The overall abundance was 0.65 scats/km, civets being the most abundant (0.33 scats/km) followed by mongoose (0.2 scats/km) and cats (0.07 scats/km). Two common mongoose were sighted.

Comparison among sites

There were considerable differences in the abundance of scats among the sites surveyed. It was highest in Silent Valley National Park (3.81 scats/km) followed by New Amarambalam (1.61 scats/km), and least in Mudumalai Wildlife Sanctuary (0.18 scats/km) and Wayanad Wildlife Sanctuary (0.22 scats/km). There were also considerable differences among the sites in the abundance of scats of the four taxa. Thus, in the more moist forests in Silent Valley National Park and New Amarambalam Reserved Forests civet scats were the most abundant (3.55 and 1.4 scats/km respectively), whereas in the drier Wayanad Wildlife Sanctuary, Mudumalai Wildlife Sanctuary and Sigur Reserved Forests mongoose scats were most abundant (0.16, 0.10 and 0.25 scats/km respectively). In the high altitude Mukkurthi National Park the abundance of Nilgiri marten scat was 0.16 scats/km. It was the only site from where scats of Nilgiri marten were recorded. In contrast to civets, mongoose and marten showed considerable differences among the sites. The lesser cat scats were more or less equally abundant in all sites. In some sites (Siruvani Hills, Moyar Valley and Pillur Reserved Forests) the scats of mongoose, lesser cats, and civets were more or less equally abundant, while that of Nilgiri marten were absent. The above differences among the nine sites were by and large due to differences in the major vegetation or habitat types among them to which the small carnivores seemed to be sensitive. In Silent Valley and New Amarambalam the vegetation is mostly wet evergreen forest, in Mukkurthi it is shola-grassland, in Siruvani Hills dry and moist deciduous forests, and in other areas mostly dry deciduous (Wayanad, Mudumalai and Pillur) or thorn and scrub forests (Sigur and Moyar Valley).

Factors affecting scat abundance

Nine different vegetation or habitat types (including plantations), which were surveyed for scat abundance, were grouped into six vegetation types by pooling the similar ones, e.g. thorn and scrub forests. Data from various survey sites were pooled to estimate scat abundance in different vegetation types (Table 3).

Table 3. The abundance of scats (no./km) of mongoose, civets, cats and marten in six vegetation categories that were surveyed in the Nilgiri Biosphere Reserve. Number of scats is given in parenthesis.

Wet and semi-evergreen forest had the highest abundance of scats, with about 53.8% of the total number of scats. Moist deciduous forests had 12.7% of the scats, followed by dry and mixed deciduous forests (11.7%), dry thorn and scrub forests (10.5%), montane shola-grasslands (9.5%) and plantations (1.7%). Mongoose were the most abundant in the drier forests, civets in the wet evergreen and marten in montane forests and wattle plantations. Cats were almost equally abundant in all forest types. Plantations had very low abundance of all the taxa.

Wet and semi-evergreen forests had the greatest abundance of scats (2.78 scats/km) among the different vegetation types, followed by montane shola-grassland (1.48 scats/km), moist and dry deciduous forests (0.38 scats/km), dry thorn and scrub forests (0.36 scats/km) and plantations (0.15 scats/km). There were differences among the vegetation types in the abundance of various species groups. In the wet and semi-evergreen forests, civets were the most abundant (2.46 scats/km). In the shola-grassland, cats were the most abundant (1.20 scats/km) followed by marten (0.20 scats/km). In the moist, dry and mixed deciduous forests, mongoose were the most abundant (0.19 scats/km), followed by civets and cats. In the dry thorn and scrub forests mongoose were the most abundant followed by cats.

 

Altitude

Sampling effort was highest in 400-1200 msl, and most of the scats were collected from this altitude class (Fig. 3). All Nilgiri marten scats were collected from above 2000 m. Where as mongoose and cat scats were found above 2000 m, those of civets were not. Mongoose scats were found from all altitude classes, unlike the other taxa. Overall, there was a significant difference in the distribution of scats of the four taxa with reference to altitude (X2=263.4, df=9, p<0.001), but this was because the marten was found only in the higher altitudes.

Proximity to water

Apart from being sources of water, streams and other water bodies are also the sources of food (fishes, amphibians, invertebrates) to small carnivores. Fruit abundance is also greater along riparian forests, especially in dry forests. There was a significant difference in the distribution of scats of the various taxa with reference to the distance from water (X2=101.1, df=9, p<0.001). While civet scats were found closer to water (nearly 80% within 200 m), those of cats were found farther (nearly 75% more than 200 m away, (Fig. 4). Mongoose scats were mostly at medium distance from water (nearly 75% at 100-500 m), while those of marten were found equally at all distance classes. This difference among species persisted even when the analysis was done for each vegetation type separately.

Fig. 3. The percentage of scats of mongoose, civets,  cats and martens found in four different altitude classes.

Fig. 4. The percentage of scats of mongoose, civets,  cats and martens found in four distance classes of water.

Disturbance

The extent of habitat disturbance at each scat location was grouped into three classes (low, moderate and high) for analysis. Disturbance was estimated in the field taking into consideration disturbance to the vegetation around the site from which scat was collected, distance to the nearest village, presence of cattle dung etc. The effect of disturbance was also evaluated with reference to canopy cover and weed cover. Overall, there was a significant difference in the distribution of scats of various taxa with reference to the three disturbance categories (X2=68.0, df=6, p<0.001). Most of the civet (72.4%) and marten (79.2%) scats were found in the low disturbance areas, whereas those of the cats (54.4%) and mongoose (54.1%) were found in moderately disturbed areas (Fig. 5). In the highly disturbed areas scats of mongoose were more abundant than that of the other species.

Canopy cover

----------------- The civets and  martens scats  were found in  the low disturbance  areas, whereas  those of the cats  and mongoose were  found in moderately  disturbed areas. In  the highly disturbed  areas scats of  mongoose were  more abundant  than other species. -----------------

Canopy cover varies among vegetation types (e.g. between wet evergreen and dry deciduous), but within each it is a good indicator of human disturbance (Fig. 6a). When all the vegetation types were combined, there was a significant difference in the distribution of scats with reference to canopy cover (X2=63.7, df=12, p<0.001). Cat scats and, to a lesser extent, mongoose scats were more in the low canopy cover areas compared to civet scats which were more common in the high canopy cover areas. Scats of the Nilgiri marten, found only in the shola-grasslands, were also more common in the low canopy cover areas. An analysis of the effect of canopy cover in each vegetation type separately, however, gave mixed results. In the wet evergreen forests and associated forests, civets scats were most abundant and in moist deciduous and scrub forests the above results were repeated; civet scats were more abundant in sites with high canopy cover, while cat and mongoose scats were more common in the low canopy cover sites. However, in the dry and mixed deciduous forests civet scats were equally distributed while cat and mongoose scats were more in the high canopy cover  areas.

Weed cover

The distribution of scats differed significantly with reference to weed cover (Fig. 6b). The scats of all taxa were more common in the low weed cover areas. However, scats of civets were relatively less in high weed cover areas compared to cats and mongoose.

DISCUSSION

----------------- The scats of all  taxa were more  common in the  low weed cover  areas. However,  scats of civets  were relatively  less in high weed  cover areas  compared to cats  and mongoose. -----------------

The survey results based on scats are constrained by the tentative identification of the scats. Personal observation shows that there is a large scope for misidentification of scats of cats and mongoose. Another constraint was that the relative abundance of scats in an area is assumed to be indicative of the abundance of animals. The defecation behaviour of the various groups of animals is largely unknown and the rates of defecation, sites of scat-marking, communication system and other characters may vary among the different groups. The results of this survey should be viewed bearing these constraints in mind.

 

Fig. 5. The percentage of scats of mongoose, civets,  cats and martens found in three disturbance classes.

The distribution of scats shows considerable variation among sites and altitude, primarily because of the variation in vegetation types. There are few reports of other similar community studies. Using track plots, Mukherjee (1998) found that jungle cats in Sariska Tiger Reserve used dense scrub forests more than hill forests, open scrub forests and grasslands. In Huai Kha Wildlife Sanctuary in Thailand, civets used the dry evergreen and mixed deciduous forests over dry deciduous forests (Rabinowitz 1991). Greater abundance of civet scats in the wet evergreen forests and in areas with greater canopy cover in all vegetation types is not surprising since civets, especially palm civets, are to a large extent frugivorous and also adapted for arboreal life. It is also true that southeast Asia, where closed forest is the primary habitat, was a major centre of speciation in the case of civets. In contrast, the arid areas in Africa have been the major centre of radiation in the case of mongoose. Mongoose in India are also primarily adapted to drier forests and hence the greater abundance of their scats in such forests. However, the sub-specific endemics of the Western Ghats (the stripe-necked mongoose and brown mongoose) are both absent from the drier parts of the Western Ghats. The brown mongoose has not been, in fact, reported from vegetation types other than wet and semi-evergreen forests, even though the sub-species in Sri Lanka is frequently seen in tea estates (Prater 1971). Stripe-necked mongoose has been reported from wet and semi-evergreen and dry deciduous forests especially near water bodies. Lesser cats are known to occur in low densities compared to other small carnivores, but more ubiquitous.

----------------- The sub-specific  endemics of the  Western Ghats  (e.g. stripe-necked  mongoose and  brown mongoose)  are both absent  from the drier  parts of the  Western Ghats. -----------------

The very low abundance of small carnivores in plantations should be a matter of concern. High abundance of large herbivores such as gaur and sambar has been reported from teak plantations (Karanth 1993). Thus, small carnivores are far more adversely affected by habitat conversion than the larger mammals. The primary reason would be low diversity of the fruiting trees, invertebrates and rodents in the plantations.

The survey shows considerable variation among the taxa in their tolerance to habitat degradation and other forms of human disturbance. Civets seem to be most adversely affected, primarily because of their dependence on tree cover. This might be particularly true for palm civets which are primarily frugivorous. In contrast, mongoose and cats can not only tolerate but might even increase in abundance following moderate disturbance. Scats of both taxa are more abundant than that of civets in areas with low canopy cover and weed cover. Both these taxa are more carnivorous than civets, and might benefit from increased abundance of invertebrates and rodents owing to moderate habitat disturbance. Increased abundance of rodents following habitat fragmentation and disturbance has been reported from the Western Ghats (Kumar et al. 1998).

Fig. 6. The percentage of scats of mongoose, civets, cats and martens found in (a) four canopy cover  classes; and (b) four weed cover categories.

The above analysis and conclusions are based on scats identified only up to the family level. There might be considerable differences among the species within the family which we are unable to examine at present owing to the problems of identification of scats up to species level. Such differences might include habitat preferences as well as response to disturbance. For example, among civet there are clear differences between species at the subfamily level. Of the two species of palm civets (Subfamily: Paradoxurinae), the common palm civet is most like to be confined to the drier forests whereas the brown palm civet seems to be mostly confined to the wet-evergreen and semi-evergreen forests (Ashraf et al. 1993). These two species are therefore parapatric rather than sympatric.

Similarly, of the two true civets (Subfamily: Viverrinae) the highly endangered Malabar civet is most likely restricted to the forests of the foothills, while the small Indian civet also occurs in the Western Ghats proper. Sympatry between these two is most likely restricted to the foothills. Between common palm and small Indian civets there could be differences in their uses of areas with varying canopy cover. 

----------------- The small carnivores  are far more adversely  affected by habitat  conversion than the  large mammals. The  primary reason would  be the low diversity  of fruiting trees,  invertebrates and  rodents in the  plantations. -----------------

Mongoose also showed a similar difference in habitat preference. The common Indian mongoose and ruddy mongoose are sympatric throughout Western Ghats, and even central India, but restricted to the drier and more open forest. In contrast, the brown and stripe-necked mongoose were restricted to the more moist and closed forests. The pattern of habitat selection among the small cats is less known. However, all the four species reported to occur in the Western Ghats are widely distributed in the Indian subcontinent. The jungle cat is known to be tolerant of disturbance and human presence, and might even benefit from it due to increased abundance of rodents which are their major diet (Mukherjee 1998). The fishing cat is highly dependent on water bodies and thus, even though widely distributed, might be a microhabitat specialist. Nothing much is known about habitat preferences of the leopard cat and rusty spotted cat, except that both are widely distributed and have been reported to occur in dry as well as moist forests.

ACKNOWLEDGEMENTS

We gratefully acknowledge funding from Ministry of Environment and Forests, Government of India; research permits from Kerala and Tamil Nadu Forest Departments, and facilities provided by Salim Ali Centre for Ornithology and Natural History.

REFERENCES

Ashraf, N.V.K., Kumar, A. & Johnsingh, A.J.T (1993) A survey of two endemic civets of the Western Ghats: The Malabar civet (Viverra civettina) and the brown palm civet (Paradoxurus jerdoni). Oryx, 7, 109-114.

Champion, H.G. & Seth, P.K. (1968) A revised survey of the forest types of India. Government of India Press, Nasik, Delhi.

Gadgil, M. & Sukumar, R. (1986) Scientific programme for the Nilgiri Biosphere Reserve: Report of a workshop, Bangalore, March 7-8, 1986. Technical report no. 8, ENVIS Centre for Ecological Sciences, IISC, Bangalore.

Karanth, K.U. (1993) Predatory-prey relationships among large mammals of Nagarhole National Park (India). Ph.D. thesis. Mangalore University, India

Kumar, A., Sivaganesan, N., Umapathy, G. & Prabhakar, A. (1998) A study on the management of fragmented rainforests of the Western Ghats for the conservation of fauna with special emphasis on small mammals. Final Report submitted to the Ministry of Environment and Forests, Government of India.

Mukherjee, S. (1998) Habitat Use by Sympatric Small Carnivores in Sariska Tiger Reserve, Rajasthan, India. Ph.D. Dissertation, Saurashtra University, Gujarat, India.

Prater, S.H. (1971) The Book of Indian Animals. Bombay Natural History Society. III edition, Oxford University Press. Bombay.

Rabinowitz, A.R. (1991) Behaviour and movements of sympatric civet species in Huai Kha Wildlife Sanctuary, Thailand. Journal of Zoology (London), 223, 281-298.