BioOne sees sustainable scholarly publishing as an inherently collaborative enterprise connecting authors, nonprofit publishers, academic institutions,research libraries, and research funders in the common goal of maximizing access to critical research.

Distribution, Occupancy and Activity Patterns of Goral (Nemorhaedus goral)and Serow (Capricornis thar) in Khangchendzonga Biosphere Reserve, Sikkim,IndiaAuthor(s): Tapajit Bhattacharya, Tawqir Bashir, Kamal Poudyal, Sambandam Sathyakumar andGoutam Kumar SahaSource: Mammal Study, 37(3):173-181. 2012.Published By: Mammal Society of JapanDOI: http://dx.doi.org/10.3106/041.037.0302URL: http://www.bioone.org/doi/full/10.3106/041.037.0302

BioOne (www.bioone.org) is a nonprofit, online aggregation of core research in the biological, ecological,and environmental sciences. BioOne provides a sustainable online platform for over 170 journals and bookspublished by nonprofit societies, associations, museums, institutions, and presses.

Your use of this PDF, the BioOne Web site, and all posted and associated content indicates your acceptance ofBioOne’s Terms of Use, available at www.bioone.org/page/terms_of_use.

Usage of BioOne content is strictly limited to personal, educational, and non-commercial use. Commercialinquiries or rights and permissions requests should be directed to the individual publisher as copyright holder.

Mammal Study 37: 173–181 (2012)

© The Mammal Society of Japan

Distribution, occupancy and activity patterns of goral (Nemorhaedus

goral) and serow (Capricornis thar) in Khangchendzonga Biosphere

Reserve, Sikkim, India

Tapajit Bhattacharya1,2

, Tawqir Bashir1

, Kamal Poudyal1,2

, Sambandam Sathyakumar1,

*

and Goutam Kumar Saha2

1

Wildlife Institute of India, P.O. Box 18, Chandrabani, Dehradun 248 001, Uttarakhand, India

2

Department of Zoology, University of Calcutta, 35, Ballygunge Circular Road, Kolkata 700019, West Bengal, India

Abstract. We assessed the distribution, occupancy, and activity patterns of two rupicaprids viz.,

Himalayan goral Nemorhaedus goral and Himalayan serow Capricornis thar in the western part of

the Khangchendzonga Biosphere Reserve, Sikkim, using camera traps during 2009–2010. Goral had

the highest photo-capture rate (# photo/100 days) of 6.37 ± 3.02 in temperate habitats (n = 169)

followed by 1.82 ± 1.27 in subalpine habitats (n = 41). Serow had the highest photo-capture rate of

1.65 ± 0.88 in subalpine habitats (n = 53) followed by 0.58 ± 0.34 in temperate habitats (n = 19). The

estimated detection probability was 0.57 for goral and 0.46 for serow. Detection probabilities were

negatively related to human presence. Occupancy of goral (0.27) was slightly lesser than serow

(0.30). Denser tree cover, warmer aspect and sites far away from tourist trails were the best predictors

for the occupancy of goral. Denser tree cover, higher elevation and warmer aspect were the best

predictors for the occupancy of serow. Spatial separation between these two species was not clear

although different activity peaks were observed. To ensure the survival of these species, protection

measures are required to keep their habitats free from anthropogenic activities.

Key words: activity patterns, eastern Himalaya, goral, occupancy, serow.

The Rupicaprids or goat-antelopes, a general name for

goral (Nemorhaedus spp.), serow (Capricornis spp.) and

takin (Budorcas taxicolor), hold an intermediate position

between ‘goats’ on the one side and ‘antelopes’ on the

other (Prater 1980; Sathyakumar 2002). These mountain

ungulates have a more or less goat-like build, goat-like

teeth, short tails and relatively small cylindrical horns

present in both sexes (Prater 1980). The goral is a cliff-

dwelling, sexually monomorphic mountain ungulate

with size 65–70 cm (shoulder high) and weighing about

25–30 kg (Prater 1980). The Himalayan goral is repre-

sented by two sub-species: the gray goral (N. goral

bedfordi) of the western Himalaya and the brown goral

(N. goral goral) inhabiting the eastern Himalaya

(Sathyakumar 2002). Goral in general are solitary

(Schaller 1977; Green 1987) but group size can vary

from 1 to 12 (Vinod and Sathyakumar 1999).

In appearance, the serow resembles a goral and is

generally a solitary animal (Schaller 1977; Prater 1980;

Nowak and Paradiso 1983). Both sexes are similar in

appearance and of about equal size (Schaller 1977). An

adult serow measures about 100 to 110 cm at its shoul-

ders and weighs about 91 kg on average (Prater 1980).

The Himalayan serow is found along the Himalayan

range of India, Nepal and Bhutan and also in Bangladesh

(Schaller 1977; Prater 1980; Nowak and Paradiso 1983;

Shackleton and Lovari 1997). It inhabits steep, rugged,

inaccessible and densely forested areas of the Himalaya.

Serow prefers damp and thickly wooded gorges and

occurs between 1,500–4,000 m (Schaller 1977; Prater

1980; Sathyakumar 2002). It is also seen on open cliffs

and rocky slopes.

These two near threatened (IUCN 2010) goat-antelopes

of Himalaya inhabit the heavily forested habitats

(Sathyakumar 1994; Mishra et al. 1994) or open habitats

at higher elevations, such as sub-alpine rhododendron

scrub, alpine meadow and grassland (Green 1985).

Complex terrain, steep topography and dense vegetation

*To whom correspondence should be addressed. E-mail: ssk@wii.gov.in

Mammal Study 37 (2012)174

impede field research and monitoring activities. Tradi-

tional field survey methods based on direct observation,

such as transect counts and behavioral observations, are

hence difficult to undertake due to the inaccessibility of

remote areas, lack of visibility in dense vegetation and

the species extreme sensitivity to human disturbance

(Sathyakumar 1994, 1997, 2002). In mountainous land-

scapes, a key factor that has limited quantitative habitat

modeling for forest antelope is the difficulty in estimat-

ing density or relative abundance (Bowkett et al. 2007).

This is because of low detection rates from methods

based on direct sightings (Feer 1995) and methodologi-

cal problems extrapolating from indirect signs such as

dung or tracks (Bowland and Perrin 1994; Struhsaker

1997; Lunt et al. 2007). However, camera-trapping has

shown excellent potential for studying elusive animals in

comparison with more traditional methods (Jácomo et al.

2004) and has proven useful for forest antelope in the

Udzungwa Mountains of Tanzania, where it has a high

detection efficiency and has recorded species otherwise

undetected (Rovero and Marshall 2004; Rovero et al.

2005; Bowkett et al. 2006). In this study, we used camera-

traps to detect goral and serow in dense, steep and very

moist eastern Himalayan forest. Previous studies on

these species in Indian and Nepal Himalaya (Green

1987; Lovari and Apollonio 1993, 1994; Cavallini

1992; Mishra et al. 1994; Sathyakumar 1994; Vinod

and Sathyakumar 1999; Aryal 2008; Bhattacharya and

Sathyakumar 2008) indicated that they can be present

in a wide range of elevation (1,000–4,000 m), and

depicted them as very shy, forest dwelling, steep slope

preferring animals which occasionally use the forest

openings (Mishra 1993; Pendharkar 1994; Sathyakumar

1994; Mishra and Johnsingh 1996; Aryal 2008). All the

above mentioned studies indicated that the geographical

factors such as terrain characteristics and ecological

factors such as forest type and vegetation cover and level

of anthropogenic activities may determine the abundance

of goral and serow. Resource-selection probability func-

tions and occupancy models are powerful methods of

identifying areas within a landscape that are highly used

by a species (MacKenzie 2005, 2006). In this study, we

modeled the essential habitat characters for occupancy

of goral and serow in an intricate valley of Eastern

Himalaya.

Apart from detection, camera traps have provided

detailed information on the occurrence and activity

patterns of relatively secretive mammals (Azlan and

Sharma 2006; Kawanishi and Sunquist 2008). As goral

and serow are reported to share similar type of habitats

(Sathyakumar 2002), assessment of their activity pat-

terns may elucidate about their likely differential strategy

of resource use. Studies on the behavioral patterns of

these ungulates indicated about their crepuscular and

probable nocturnal activity (Cavallini 1992; Lovari and

Apollonio 1994). In this study, using the camera-trap

data, we also observed their activity patterns and

assessed if any temporal difference was present.

Methods

Study area

Ecological information on rupicaprids in Indian

Himalaya, particularly from Eastern Himalaya is very

scanty (Sathyakumar et al. 2011). The geographical

position of Eastern Himalaya denotes the convergence

of three biogeographic realms, vis., Palaearctic, Africo-

tropical and Indo-Malayan (Mani 1974). This special

biogeographic position may have consequence the rec-

ognition of this area as one of the global biodiversity

hotspots (Myers et al. 2000) and also as one among the

important Global 200 Ecoregions (Olson and Dinerstein

1998). Sikkim, a tiny mountainous state of India is the

westernmost part of this biodiversity rich region. In spite

of being a biodiversity rich state, there has been only one

scientific study available on mammalian assemblage

from this state (Sathyakumar et al. 2011). Till date, only

the presence of these Rupicaprids has been reported

(Sharma and Lachungpa 2002) from Sikkim, whereas

almost all the ecological information about these impor-

tant prey species of large carnivores of subalpine and

temperate forests is still unknown.

The study was carried out in Khangchendzonga

Biosphere Reserve (BR) in Sikkim, India (Fig. 1), from

February 2008 to August 2010. The Khangchendzonga

BR is one of the most significant biodiversity hotspots of

India, having varying ecozones from temperate to arctic

(1,220–8,586 m). The varying elevation within an aerial

distance of just 42 km makes this park a unique natural

heritage hotspot in the world. The Khangchendzonga

BR encompasses temperate, subalpine and alpine habi-

tats (1,000 to 5,000 m) as well as rocky slopes, glacial

moraines and permafrost areas (> 5,000 m) with diverse

slope and aspect categories, along with a range of wild-

life use. The area of Khangchendzonga BR has been

divided into seven watersheds (Fig. 1) or river sub-

systems, namely, Lhonak (15%), Zemu (23%), Lachen

(5%), Rangyong (36%), Rangit (6%), Prek (8%) and

Bhattacharya et al., Goral and serow in Khangchendzonga 175

Churong (7%). The Prek Chu valley (27°37'N, 88°12'E–

27°21'N, 88°17'E) opens up in the upper reaches with a

total area of 182 km2

. The highest elevation and the low-

est elevation of the Prek Chu are 6,691 m (summit of

Pandim) and 1,200 m respectively, with a mean of 3,562

m (Tambe 2007). The following elevation classes are

observed: 1,200–2,000 m (5%), 2,001–3,000 m (13%),

3,001–4,000 m (25%), 4,001–5,000 m (44%) and 5,001–

6,991 m (13%) in the study area. The annual rainfall

ranges from 1,750 mm to 2,250 mm, with the mean

being around 2,230 mm (Tambe 2007). The Prek Chu

has a typical monsoon climate and can be divided into

the following habitat classes-mixed sub-tropical and

mixed temperate (17%), subalpine and krummholtz

(36%), alpine meadows (5%), rocky area and snow cover

(41%) and water bodies (1%) (Fig. 2).

Camera trapping

Based on the knowledge acquired through reconnais-

sance surveys and trail walks (n = 22, 1.5 to 7 km) for

ungulate evidences carried out in 2008 (Bhattacharya et

al. 2010; Sathyakumar et al. 2011), intensive camera

trapping was carried out in Prek Chu from February

2009 to August 2010. Prek Chu was divided into 1 km2

blocks using Geographic Information System (GIS)

tools. For simplicity, the area was categorized into three

different survey zones according to the elevation, vis.,

temperate (1,200–3,000 m), sub-alpine (3,000–4,000 m)

and alpine (above 4,000 m) and the camera traps were

deployed corresponding to the area coverage of the sur-

vey zones and their accessibility (13 blocks in temperate,

18 blocks in subalpine and 13 blocks in alpine). One

camera unit was placed in each block for at least 1

month, and was then moved to a different block of differ-

ent habitat. Due to difficult navigation in the field, more

than one camera units were occasionally placed in one

block (15 blocks). Within each survey zone, cameras

were placed in likely animal-use areas and > 500 m from

other cameras. Twenty seven cameras were deployed at

71 sites in 44 blocks (Fig. 3). The camera trapping was

done continuously in all the seasons (winter: January–

March; spring: April–May; summer: June–September;

Fig. 1. Location of the study area: Khangchendzonga Biosphere

Reserve in Sikkim, India showing the different watersheds including

Prek Chu catchment.

Fig. 2. Major habitat categories in Prek Chu Catchment,

Khangchendzonga Biosphere Reserve, Sikkim, India.

Mammal Study 37 (2012)176

autumn: October–December); however, similar intensity

of camera trapping could not be maintained in alpine

zone during winter and in temperate zone during peak

monsoon (July–August). Four models of infrared-

triggered camera units: two DeerCam (Deercam Scout-

ing Camera, Non Typical, Inc., Park Falls, WI, USA),

two Wild-view (wildview xtreme2, Grand Prairie, Texas,

USA), 18 Stealthcam (Stealthcam, LLC, Grand Prairie,

Texas, USA) and five Moultrie (Moultrie Feeders, Ala-

baster, Alabama, USA) were used. Head-on, oblique

and side-view camera configurations were used to obtain

photographs at varying body orientations (Blomqvist

and Nystrom 1980; Jackson et al. 2006). Since the study

species were rare and the area being vast, the strategy

was to survey more sampling units less intensively rather

than less sampling units more intensively (MacKenzie

and Royle 2005). All the camera units were set with a

1 min delay between photographs, and set for 24-h

monitoring. Camera units were attached to trees 15–30

cm above the ground and 3–5 m from a trail or point

where animal movement might be expected. A time/date

stamp accompanied each photograph and at each sample

site we recorded GPS location, elevation, slope, aspect

and habitat, including forest type and percentage of

vegetation cover (ground cover along with understorey

and canopy cover). The number of camera trap-days

was calculated from the date of deployment till the date

of retrieval (if the memory card was not full) or till the

date of the final photo.

Monitoring of camera traps was done at least twice

a month which included changing the batteries and

memory card. Due to the topography and remoteness

of the area all field activities were carried out in the

form of field expeditions i.e., camping in different areas

of the Prek Chu.

Analytical methods

Photographic encounter rates were calculated for each

species for each habitat zone. Photographic encounter

rate was calculated as the number of photo captures of

a species divided by the number of trap-days per site.

Trap-days were computed as the number of 24-h periods

from deployment of camera until the film/memory card

was used up or the camera was retrieved. Instances

where the same species were captured by the same

camera more than once within 1 hour were excluded

from trap rate calculation (Bowkett et al. 2007). This

was a compromise between scoring the same individual

multiple times and missing individuals (Rovero et al.

2005) and is more conservative than other published

studies (e.g., Kinnaird et al. 2003). All detections for

each species were summed for each camera site, multi-

plied by 100, and divided by the total sampling effort

for that sample point (number of camera-days). We

used this photographic rate to compare detection of each

species in different habitats and tested the significance

using Mann-Whitney U test.

The sampling record at each site was divided into five

consecutive 5-day segments based on the date stamp on

the photographs. A detection matrix of each species

was established following the approach proposed by

MacKenzie et al. (2002). We excluded sites where the

sampling was less than five 5-day segment (41 sites were

included). An occupancy model (program PRESENCE,

v. 2.2; Hines 2006; MacKenzie et al. 2006) was used to

estimate the site occupancy rate (ψ) and detection prob-

Fig. 3. Map of study area showing locations of camera traps in

1 km × 1 km grids in different habitats of Prek Chu Catchment,

Khangchendzonga Biosphere Reserve, Sikkim, India.

Bhattacharya et al., Goral and serow in Khangchendzonga 177

ability (p) relative to seven sampling variables and two

detection variables (Table 1). Geographical orientation

of mountain ranges are in north-south direction in

Sikkim Himalaya (Tambe 2007); hence, only the eastern

aspects get exposed to the sun in the forenoon and regu-

lar cloudy weather in afternoon prevent the western

aspects to get the warmth of the afternoon sun. This

typical orientation of mountain ranges and the cloudy

weather have resulted the categorization of eastern

aspects as warmer and western aspects as cooler. The

detection variable ‘human signs’ included every possible

indirect evidence of human such as plastics, discarded

shoes or clothes or any other such objects and direct

evidence such as photo capture. Akaike Information

Criterion (AIC; Akaike 1973) values were used to rank

the occupancy models and all the models whose ΔAIC

< 2 were considered as equivalent models. The summed

model weight of each covariate in these models was used

to determine the most influential variables for each

species. The sign of logistic coefficient of each variable

(positive or negative) was used to determine the direction

of influence of the variable. The time and date printed

on the photographs has been used to determine the daily

activity pattern of individual species (Pei 1998). We

used a Daily Activity Index (DAI) of 2-hour durations to

examine the daily activity level:

DAI = No. of photographs within a duration

× 100/Total no. of photographs

Rayleigh uniformity test and Watson U2

test (Mardia

and Jupp 2000) using Oriana 3.13 (1994–2010 Kovach

Computing Services) was applied to determine unifor-

mity and the significance of differences in the daily

activity patterns between species.

Results

A sampling effort of 6,278 camera-days across 71

sample sites was achieved in the three survey zones

(1,407 camera-days in temperate, 3,061 camera-days in

subalpine, 1,810 camera-days in alpine), resulting in

4,517 photo captures (2,668 wild animal, 1,849 domestic

animals and human) 284 of which contained goat-

antelopes. During the 2-year survey, 42 mammal species

were recorded (Sathyakumar et al. 2011), seven of which

were ungulates (n = 468 photo captures). Brown goral

and Himalayan serow were the most frequently detected

ungulates for both temperate and subalpine habitats

within the low and mid-elevation range (1,200–3,700 m).

Blue sheep Pseudois nayaur and musk deer Moschus

sp. only occurred over 3,700 m in alpine meadows and

sub-alpine shrub habitats respectively. Barking deer

Muntiacus muntjak and wild pig Sus scrofa occurred

only < 2,500 m in lower temperate habitat. Himalayan

tahr Hemitragus jemlahicus was detected very rarely

(n = 6 photo captures) in subalpine and alpine habitats.

Between the goat-antelopes, aggregation behaviour was

recorded only in case of goral as twenty five photo cap-

tures contained two adult individuals and three photo-

graphs contained a group of three adults. Only one

photograph of serow contained two individuals.

Goral was detected over the broadest elevation range

(1,730–3,670 m), with the highest photo-capture rate

6.37 ± 3.02 (SE) in temperate (n = 169 captures) fol-

lowed by 1.82 ± 1.27 in subalpine (n = 41 captures).

Serow was detected over the elevation range 2,310–

3,700 m, and had the highest photo-capture rate 1.65 ±

0.88 in subalpine (n = 53) followed by 0.58 ± 0.34 at

temperate habitat (n = 19). Mann-Whitney U test showed

that there was no significant difference in photo-capture

rates between subalpine and temperate habitats in case of

both goral (P = 0.68) and serow (P = 0.07).

Goat-antelopes were detected at 12 sites (9 sites each

for goral and serow). The estimated site occupancy rates

of both species were slightly higher than the naive esti-

mates (i.e., the proportion of sites where the species was

detected at least once, 0.24 for both goral and serow),

although the estimated detection probabilities of both

Table 1. Variables used to estimate the site occupancy rates and

detection probabilities of goral and serow in the occupancy model

Abbreviation Name Description

Sampling variables

V Vegetation cover (%) Numeric (Range 0–80%)

E Elevation Numeric (Range 1,830–4,520 m)

A Aspect Categorical (Warm – NE, E,

SE, S (denoted as 0); Cold – N,

NW, W, SW (denoted as 1))

T Trekking trail Categorical (present, absent)

C Conifer forest Categorical (present, absent)

B Broadleaved forest Categorical (present, absent)

S Slope Categorical (steep > 30°

denoted as 0, gentle ≤ 30°

denoted as 1)

Detection variables

H Human presence Human sign (including direct

and indirect evidences) present

or absent

R Season Rainy (May–September) or

dry (October–April)

Mammal Study 37 (2012)178

species were greater than 0.40 (Table 2), suggesting that

our sampling duration (at least 30 days) was long enough

to detect the species at each site when they were present.

Results of occupancy modeling showed that detection

probability of both serow and goral were negatively

related with human presence at camera site (Table 3).

The estimation of occupancy rate of goral (0.27) was

slightly lesser than that of serow (0.30). Denser tree

cover, warmer aspect and distant sites from regularly

used tourist trails were the best predictors for the occu-

pancy of goral. Denser tree cover, higher elevation and

warmer aspect were determined as the best predictors for

the occupancy of serow (Table 3).

The DAI for goral and serow (182 and 60 photo-

graphs, respectively) confirmed that both species can be

found during daytime as well as at night, however, the

activity peaks differed. Goral showed an early morning

activity peak at 0400–0600 h (Rayleigh Z = 3.73, P =

0.02; indicating that the data was not uniformly distrib-

uted). Serow was active during the daytime with a peak

at morning (0600–0800 h), during afternoon, evening

and throughout the night; from 1600 h to 2200 h the

DAI of serow were same and highest (13.33). Rayleigh

Z test showed that the data was more or less uniformly

distributed (Rayleigh Z = 1.02, P = 0.36). The difference

between DAI of goral and serow (Fig. 4) was significant

(Watson’s U2

= 0.196, P < 0.05).

Discussion

During reconnaissance survey and trail sampling in

2008, we got only eight and four visual encounters of

goral and serow respectively, and the dung densities

calculated for different seasons were also very low

Table 2. The top models for predicting site occupancy (Est. ψ) and detection probability (Est. p) of goral and serow in Prek Chu watershed of

Khangchendzonga Biosphere Reserve

Models ΔAIC AIC weight No. Par. (–2LL) Est. ψ (± 1 se) Est. p (± 1 se) Est. of c-hat

Goral

ψ (VA), p (H) 0 0.2634 5 80.11 0.2671 ± 0.0492 0.5474 ± 0.0191 0.68

ψ (VAT), p (H) 0.21 0.2372 6 78.32 0.2544 ± 0.0491 0.5682 ± 0.0178 0.78

ψ (V), p (H) 0.78 0.1783 4 82.89 0.2585 ± 0.0439 0.5599 ± 0.0184 0.75

ψ (VATB), p (H) 1.72 0.1115 7 77.83 0.2551 ± 0.0491 0.5689 ± 0.0178 0.79

ψ (EVAT), p (H) 1.89 0.1024 7 78 0.2555 ± 0.0492 0.5678 ± 0.0178 0.72

ψ (VAT), p (HR) 1.99 0.0974 7 78.1 0.2557 ± 0.0493 0.5552 ± 0.0175 0.71

Serow

ψ (EVAC), p (H) 0 0.284 7 59.87 0.3046 ± 0.0667 0.4452 ± 0.0237 0.91

ψ (EVAC), p (HR) 0.08 0.2729 8 57.95 0.3151 ± 0.0660 0.4496 ± 0.0253 0.67

ψ (EVTA), p (HR) 0.28 0.2469 8 58.15 0.2661 ± 0.0591 0.4706 ± 0.0242 0.83

ψ (EVTA), p (H) 1.5 0.1342 7 61.37 0.2690 ± 0.0594 0.4668 ± 0.0233 0.93

Table 3. Summed model weight (Σ), Average β value with standard error (SE) and sign [positive (+), negative (–) and (*

) if significant] of each

sampling variable in the equivalent models listed in Table 2

Variables (Abbreviation)

Goral Serow

Σ model weight Average β value (SE) Σ model weight Average β value (SE)

Sampling variables

Tree % (V) 0.99 +2.1 (0.91)* 0.94 +14.38 (4.51)*

Elevation (E) 0.1 +0.56 (1.02) 0.94 +11.08 (3.74)*

Aspect (A) 0.81 –1.89 (0.7)* 0.94 –5.83 (2.53)*

Trekking trail (T) 0.55 –1.74 (0.62)* 0.38 –4.82 (2.15)*

Conifer forest (C) NA NA 0.56 +4.97 (2.39)*

Broadleaved forest (B) 0.11 –0.94 (1.32) NA NA

Detection variables

Human presence (H) 0.99 –1.61 (0.55)* 0.94 –2.77 (1.01)*

Season (R) 0.10 –0.19 (0.42) 0.52 0.74 (2.13)

NA = Not appeared in top models.

Bhattacharya et al., Goral and serow in Khangchendzonga 179

(Bhattacharya et al. 2010). Visual encounters of goat-

antelopes were very rare due to dense cover and steep

terrain (Bhattacharya et al. 2010; Sathyakumar et al.

2011). Our result showed that these species are active at

night also, thus camera trapping proved to be the most

efficient method to detect these mountain ungulates in

Khangchendzonga BR. Of the seven ungulate species

recorded, goral and serow showed the coverage of

broadest elevation range. Other studies on the ecology

and distribution of these goat-antelopes also indicated

broad elevation range for goral (Green 1985; Cavallini

1992; Mishra 1993; Sathyakumar 1994; Vinod and

Sathyakumar 1999) in western Himalaya. All of these

studies indicated that goral can be found from very low

elevation (500 m) to the tree line (4,000 m), similarly

during this study, goral photographs had been captured

from the lowermost part of the study area (1,730 m) to

the subalpine forests (3,700 m). Aryal (2008) showed

that serow preferred 2,500–3,500 m altitude range in

central Himalaya of Nepal. Similarly, in this study most

of the serow photographs were captured in this elevation

range. No significant difference between photo-capture

rates of goral in subalpine and temperate habitats indi-

cated the similar level of presence (or use) of the species

in both habitats. Photo-capture rates of serow indicated

no significant difference between subalpine and temper-

ate habitats but the significance level (P = 0.07 at 95%

level of significance) may also hint for its presence more

in subalpine (53 captures) than in temperate habitats

(19 captures).

Our findings from the occupancy based models

(summed model weights) for serow indicated that dense

conifer forests in upper elevation zone of Prek Chu may

be occupied more by the species. The summed model

weight of elevation (Table 3) was low and depicted that

dense broadleaved forests of the temperate zone may be

occupied more by goral. Thus, findings from both photo

captures and occupancy based models indicated a ten-

dency for spatial separation between goral and serow

(altitudinal) in terms of forest type specific distribution

although the distinction was not very prominent.

Absence of other ungulates in the subalpine and upper

temperate region; as blue sheep and musk deer inhabited

the alpine and alpine-subalpine edge and barking deer

and wild pig confined to the lower temperate zones,

goral and serow may co-occur in the above mentioned

unoccupied zones. Being goat-antelopes, goral and

serow may have similar food and covers requirements

(Green 1985, 1987; Sathyakumar 1994, 2002; Awasthi et

al. 2003) as they were detected in the same temperate

and subalpine forests of the study area. Although the

spatial separation between these two species was not

clear from our findings but the different activity peaks

(goral was active at dawn and morning and serow was

active at afternoon and night) and a significant difference

in daily activity indices may indicate probable temporal

separation between these rupicaprids. Detailed studies

on behavior using intensive camera trapping in specific

locations and on the food habits through fecal pellet anal-

ysis of these goat-antelopes in eastern Himalayan region

may be helpful to properly understand the complex eco-

logical perspectives of these least studied goat-antelopes.

Studies on habitat use patterns of goral in western

Himalaya (Sathyakumar 1994; Mishra and Johnsingh

1996; Vinod and Sathyakumar 1999; Bhattacharya and

Sathyakumar 2008) described its preference for open

patches, but unlike the western Himalayan light wooded

or open habitats such as scattered trees and scrub

(Sathyakumar 1994), the eastern Himalayan habitats are

denser and thus in our findings dense forest cover was

one of the best predictors as depicted by the occupancy

models. Both goral and serow preferred warmer eastern

aspects of the study area but the specific reason for this

preference was not clear from the results of this study.

As depicted by the studies in western Himalaya, both

the species showed preference for steep slopes and we

considered steepness of slope as a site covariate in the

occupancy model (Table 1), but more or less uniform

steepness (> 30°) of the study area may result the omis-

sion of slope as a variable from best predictor models of

occupancy of goat-antelopes though presence of cliffs

and/or steep slope is important determinant of these rupi-

caprids habitat preference (Bhattacharya et al. 2010).

Fig. 4. Daily activity patterns of goral and serow in Prek Chu catch-

ment area of Khangchendzonga Biosphere Reserve.

Mammal Study 37 (2012)180

According to the result of occupancy modeling, dis-

tant sites from the regularly used trekking trails were best

occupied by goral and same was true for the occupancy

of serow. Presence of human at the camera sites clearly

affected the detection probability of both species nega-

tively. Both species are shy and mostly solitary in nature

(Sathyakumar 2002) thus avoidance of human presence

was quite normal. But at the same time, presence of

human inside the National Park either for tourism or for

livestock grazing was frequently observed during the

study period. In the temperate zone, proximity to the

villages resulted in presence of more livestock (pack

animals such as horse, dzo and cattle) and human inside

the dense broadleaved forest (dominated by Quercus

spp. and Castanopsis spp.) and in some cases at potential

goral habitats. In the subalpine conifer forests (domi-

nated by Abies spp. and Rhdodendron spp.), only anthro-

pogenic activity recorded was tourism and its impact was

probably confined only along the trekking trails. Regu-

lation of human and livestock entry inside the temperate

zone of the Biosphere Reserve may be helpful in protect-

ing the goral habitat. Strict vigil is also necessary to

keep the subalpine habitats intact and free from the

adverse impacts of tourism for the survival of serow.

Acknowledgments: We are grateful to the Department

of Forests, Environment and Wildlife Management,

Government of Sikkim for granting us permission to

work in Sikkim. We thank the Wildlife Institute of

India, Dehradun for providing us the grants and support.

We thank Mr. Sukbahadur and Mr. Sukraj for their help

during field work. We thank the reviewers for their

comments on the manuscript.

References

Akaike, H. 1973. Information theory and an extension of the maxi-

mum likelihood principle. In (B. N. Petrov and F. Csadki, eds.)

Proceedings of 2nd International Symposium on Information

Theory, pp. 267–281. Akademiai Kiado, Budapest.

Aryal, A. 2008. Status and Conservation of Himalayan Serow

(Capricornis sumatraensis. thar) in Annapurna Conservation

Area of Nepal. BRTF Nepal: A Report Submitted to the Rufford

Small Grant for Nature Conservation, UK and the People’s Trust

for Endangered Species, UK, 57 pp.

Awasthi, A., Uniyal, S. K., Rawat, G. S. and Sathyakumar, S. 2003.

Food plants and feeding habits of Himalayan ungulates. Current

Science 85: 719–723.

Azlan, M. J. and Sharma, D. S. K. 2006. The diversity and activity

patterns of wild felids in a secondary forest in peninsular Malaysia.

Oryx 40: 36–41.

Bhattacharya, T. and Sathyakumar, S. 2008. Abundance, Group-

sizes and Habitat use patterns of Himalayan tahr (Hemitragus

jemlahicus) and goral (Nemorhaedus goral) in Chenab valley,

Chamoli district (Uttarakhand). The Indian Forester 134: 1359–

1370.

Bhattacharya, T., Bashir, T., Poudyal, K., Sathyakumar, S., Bisht, S.

and Saha, G. K. 2010. Distribution, relative abundance and

habitat use by mountain ungulates in Prek Chu catchment,

Khangchendzonga Biosphere Reserve, Sikkim, India. In (J. E.

Granados, P. Fandos, R. Cadenas de Llano and M. Festa, eds.)

Mountain Ungulates 2009. A selection of edited papers from the

V World Conference on Mountain Ungulates. Galemys 22

(Special Issue): 149–170.

Blomqvist, L. and Nystrom, V. 1980. On identifying snow leopards

by their facial markings. In (L. Blomqvuist, ed.) International

Pedigree Book of Snow Leopards, Vol. 2, pp. 159–167. Helsinki

Zoo, Helsinki.

Bowkett, A. E., Lunt, N., Rovero, F. and Plowman, A. B. 2006.

How do you monitor rare and elusive mammals? Counting

duikers in Kenya, Tanzania and Zimbabwe. In (E. Zgrabczyñska,

P. Ćwiertnia and J. Ziomek, eds.) Animals, Zoos and Conserva-

tion, pp. 21–28. Zoological Garden in Poznan, Poland.

Bowkett, A. E., Rovero, F. and Marshall, A. R. 2007. The use of

camera-trap data to model habitat use by antelope species in the

Udzungwa Mountain forests, Tanzania. African Journal of

Ecology 46: 479–487.

Bowland, A. E. and Perrin, M. R. 1994. Density estimate methods

for blue duikers Philantomba monticola and red duikers

Cephalophus natalensis in Natal, South Africa. Journal of

African Zoology 108: 505–519.

Cavallini, P. 1992. Survey of the goral Nemorhaedus goral

(Hardwicke) in Himachal Pradesh. Journal of Bombay Natural

History Society 89: 302–307.

Feer, F. 1995. Seed dispersal in African forest ruminants. Journal of

Tropical Ecology 11: 683–689.

Green, M. J. B. 1985. Aspects of the ecology of the Himalayan Musk

Deer. Ph.D. Dissertation. Cambridge University, Cambridge,

292 pp.

Green, M. J. B. 1987. The Conservation Status of the Leopard, Goral

and Serow in Bangladesh, Bhutan, Northern India and Southern

Tibet: A Report Prepared by the IUCN Conservation Monitoring

Centre for the United States Fish and Wildlife Service. IUCN

Conservation Monitoring Centre, Cambridge, 26 pp.

Hines, J. E. 2006. PRESENCE v 2.2—Software to estimate patch

occupancy and related parameters. USGS-PWRC. Available at:

http://www.mbr-pwrc.usgs.gov/software/ PRESENCE.html

IUCN 2010. IUCN Red List of Threatened Species. Version 2010.4.

www.iucnredlist.org.

Jackson, R. M., Roe, J. D., Lichuk, R. and Hunter, D. O. 2006. Esti-

mating snow leopard population abundance using photography

and capture-recapture techniques. Wildlife Society Bulletin 34:

772–781.

Jácomo, A. T. A., Silveira, L. and Diniz-Filho, J. A. F. 2004. Niche

separation between the maned wolf (Chrysocyon brachyurus), the

crab-eating fox (Dusicyon thous) and the hoary fox (Dusicyon

vetulus) in central Brazil. Journal of Zoology 262: 99–106.

Kawanishi, K. and Sunquist, M. E. 2008. Food habits and activity pat-

terns of the Asiatic golden cat (Catopuma temminckii) and dhole

(Cuon alpinus) in a primary rainforest of Peninsular Malaysia.

Mammal Study 33: 173–177.

Kinnaird, M. F., Sanderson, E. W., O’Brien, T. G., Wibisono, H. T.

and Woolmer, G. 2003. Deforestation trends in a tropical

landscape and implications for endangered large mammals.

Conservation Biology 17: 245–257.

Lovari, S. and Apollonio, M. 1993. Notes on the ecology of gorals in

Bhattacharya et al., Goral and serow in Khangchendzonga 181

two areas of southern Asia. Revue d’Ecologie: La Terre et la Vie

48: 365–374.

Lovari, S. and Apollonio, M. 1994. On the rutting behaviour of the

Himalayan goral Nemorhaedus goral (Hardwicke, 1825). Journal

of Ethology 12: 25–34.

Lunt, N., Bowkett, A. E. and Plowman, A. P. 2007. Implications of

assumption violation in density estimates of antelope from dung-

heap counts: a case study on grey duiker (Sylvicapra grimmia) in

Zimbabwe. African Journal of Ecology 45: 382–389.

MacKenzie, D. I. 2005. What are the issues with presence-absence

data for wildlife managers? Journal of Wildlife Management 69:

849–860.

MacKenzie, D. I. 2006. Modeling the probability of resource use: the

effect of, and dealing with, detecting a species imperfectly.

Journal of Wildlife Management 70: 367–374.

MacKenzie, D. I. and Royle, A. 2005. Designing occupancy studies:

general advice and allocating survey effort. Journal of Applied

Ecology 42: 1105–1114.

MacKenzie, D. I., Nichols, J. D., Lachman, G. B., Droege, S., Royle,

J. A. and Langtimm, C. A. 2002. Estimating site occupancy rates

when detection probabilities are less than one. Ecology 83:

2248–2255.

MacKenzie, D. I., Nichols, J. D., Royle, J. A., Pollack, K. H., Bailey,

L. L. and Hines, J. E. 2006. Occupancy Estimation and Model-

ing. Academic Press, New York, 324 pp.

Mani, M. S. 1974. Biogeography of the Himalaya. In (M. S. Mani and

W. Junk, eds.) Ecology and Biogeography in India, pp. 664–681.

B.V. Publishers, The Hague.

Mardia, K. V. and Jupp, P. E. 2000. Directional Statistics. John

Willey and Sons. Inc., Chichester, 460 pp.

Mishra, C. 1993. Habitat Use by Goral (Nemorhaedus goral bedfordi)

in Mujhatal Harsang Wildlife Sanctuary, Himachal Pradesh,

India. Wildlife Institute of India, Dehradun, 55 pp.

Mishra, C. and Johnsingh, A. J. T. 1996. On habitat selection by

the goral Nemorhaedus goral bedfordi (Bovidae, Artiodactyla).

Journal of Zoology 240: 573–580.

Mishra, C., Raman, T. S. and Johnsingh, A. J. T. 1994. Survey of

Primates, Serow and Goral in Mizoram. Wildlife Institute of

India, Dehradun, 36 pp.

Myers, N., Mittermier, R. A., Mittermier, C. G., da Fonseca, G. A. B.

and Kent, J. 2000. Biodiversity hotspots for conservation

priorities. Nature 40: 853–858.

Nowak, R. M. and Paradiso, J. L. 1983. Walker’s Mammals of the

World, 4th ed. Johns Hopkins University Press, Baltimore/

London, 1362 pp.

Olson, D. and Dinerstein, E. 1998. The Global 200: A representation

approach to conserving the Earth’s most biologically valuable

ecoregions. Conservation Biology 12: 502–515.

Pei, J-C. 1998. An evaluation of using auto-trigger cameras to record

activity patterns of wild animals. Taiwan Journal of Forest

Science 13: 317–324.

Pendharkar, A. 1994. Habitat use, Group size and Activity Pattern of

Goral (Nemorhaedus goral) in Simbalbara Sanctuary (Himachal

Pradesh) and Darpur Reserved Forest (Haryana), India. Wildlife

Institute of India, Dehradun, 59 pp.

Prater, S. H. 1980. The Book of Indian Animals. Bombay Natural

History Society, Oxford University Press, Oxford, 324 pp.

Rovero, F. and Marshall, A. R. 2004. Reliability of line transect

techniques for estimating abundance of forest antelopes: a case

from the Udzungwa Mountains, Tanzania. Tropical Zoology 17:

267–277.

Rovero, F., Jones, T. and Sanderson, J. 2005. Notes on Abbott’s

duiker (Cephalophus spadix True 1890) and other forest ante-

lopes of Mwanihana Forest, Udzungwa Mountains, Tanzania, as

revealed from camera-trapping and direct observations. Tropical

Zoology 18: 13–23.

Sathyakumar, S. 1994. Habitat Ecology of Major Ungulates in

Kedarnath Musk Deer Sanctuary, Western Himalaya. Ph.D

Dissertation. Saurashtra University, Rajkot, 244 pp.

Sathyakumar, S. 1997. The Elusive Serow: Surviving under Threat

from Humans. Frontline, Chennai, pp. 71–72.

Sathyakumar, S. 2002. Species of the greater Himalaya. In (S.

Sathyakumar and Y. V. Bhatnagar, eds.) ENVIS Bulletin: Wild-

life and Protected Areas, pp. 1–6. Wildlife Institiute of India,

Dehradun.

Sathyakumar, S., Bashir, T., Bhattacharya, T. and Poudyal, K. 2011.

Assessing mammal distribution and abundance in intricate eastern

Himalayan habitats of Khangchendzonga, Sikkim, India. Mam-

malia 75: 257–268.

Schaller, G. B. 1977. Mountain Monarchs: Wild Sheep and Goats of

the Himalaya. University of Chicago Press, Chicago, 425 pp.

Shackleton, D. M. and Lovari, S. 1997. Classification adopted for the

Caprinae survey. In (D. M. Shackleton, ed.) Wild Sheep and

Goats and Their Relatives: Status Survey and Conservation

Action Plan for Caprinae, pp. 9–14. IUCN, Gland, Switzerland

and Cambridge, UK.

Sharma, T. R. and Lachungpa, U. 2002. Status, distribution and

management of mountain ungulates in Sikkim. In (S. Sathyakumar

and Y. V. Bhatnagar, eds.) ENVIS Bulletin: Wildlife and Pro-

tected Areas, pp. 38–43. Wildlife Institute of India, Dehradun.

Struhsaker, T. T. 1997. Ecology of an African Rain Forest: Logging

in Kibale and the Conflict Between Conservation and Exploita-

tion. The University of Florida Press, Gainesville, 456 pp.

Tambe, S. 2007. Ecology and Management of the Alpine Landscape

in the Khanchendzonga National Park, Sikkim Himalaya. Ph.D

Dissertation. FRI University, Dehradun, 232 pp.

Vinod, T. R. and Sathyakumar, S. 1999. Ecology and Conservation of

Mountain Ungulates in Great Himalayan National Park, Western

Himalaya, Final Report (FREEP-GHNP) Vol. 3. Wildlife Insti-

tute of India, Dehradun, 92 pp.

Received 22 December 2011. Accepted 23 March 2012.