Sam Bar Paper

8/6/2019 Sam Bar Paper

1/12

1993 Elsevier Science Publishers B. V. All rights reserved.Leer of ChinlJ. N. Ohta;shi and H.-I. Sheng. editors. 27

cologyof Samhar iriMudumalai Sanctuary, Southern IndiaSurendra VARMAN and R SUKUMAR

entre for Ecological Sciences, Indian Institute of Scienceangalore - 560012, IndiaABSTRACT. The sambar iCeruus unicolor) is the largest and most widely distributed deerndia. This paper looks at ecological correlates of sambar density, habitat utilization, grou p structur

nd composition and reproductive phenology in the tropical deciduous forests of Mudumalaisouthern India. during 1988-1991. Sambar densities as estimated by vehicle-and walkingvaried significantly. with the latter method giving higher and less biased figures. Average

.anba .density in the study area was 8.0 animalJkm2;'this translates into a biomass of 1059 kg/km"6er t ~anthera tigris) showed a high preference for sam bar intheir diet. Group size ranged from on50 individuals, with an average of 3.1 animals. The larger groups tended to form during the weeason, probably due to increased food availability. Sambar preferred moist deciduous forest duringdry seasons and dry deciduous forest with tall grass during the wet seasons and this may b

elated to foraging behaviour. Sex ratio showed a bias in favour of females


2/12

274

SOUTH INDIA MUDUMAlAI W ILDLIFE SANCTVARY

K"RN"TA~AN

t

IL ~-- Zone boundary- - - - Game- roads

Figure 1.Mudwnalai Wildlife Sanctuary - Showing study area with differentzones and game road.

radient, a range of vegetation types from small patches of semi-evergreen forest,hrough moist deciduous and dry deciduous forest, to dry thorn forest can be seen.escriptions of floristics and vegetation structure are available elsewhere (Sharmat al. 1978, Sukumar et al. 1992).tudy MethodsBased on vegetation type, the study area was stratified into a number of zonesere were five major zones: moist deciduous forest (MDF), dry deciduous forest with

all grass (DDFT), dry deciduous forest with short grass (DDFS), riparian forestRPF) and dry thorn forest (DTF). Several sub-zones were identified within each ohese, but we report results summarized for the major zones only. Three seasons, dryJanuary - April), first wet (May - August) and second wet (September - December)ere defined.


3/12

275

walked, covering all the zones proportionate to their areas. Transects were coveredonce in the morning (0700-0930) and once in the evening (1600-1830) each month.Recording of data basically followed line transect methods described by Burnham etal. (1980) and used by Karanth and Sunquist (1992) in a similar forest type insou the rn Ind ia ..For each aniIJlal sighting the perpendicular distance from the cen treof the group to the transect line (or road) was recorded using a range-finder, inaddition to details of group composition. Sambar were classified as adult female,yearling female, hard- and velvet-antler males, yearling males with spike antlers,and fawns. ~Densities were estimated using probability density function theory (we used aFourier Series function) as described byBurnham et al. (1980), using the programme

TRANSECT (White 1987). Densities were computed zone-wise for each season inorder to understand patterns of habitat utilization. Biomass of the population wascomputed using mean animal weights reported in the literature (Schaller 1967,Karanth and Sunquist 1992). To assess the importance of sambar as a prey speciesfor the major predators in the area, all dead sam bar found in the field were aged andsexed whenever possible and the cause of death ascertained. Scats of the majorpredators, dhole tCuon alpinus l.Ieopard (Pantherapardus)and tiger were collected.

RESULTSDensity and BiomassSambar density estimated by vehicle transects was significantly lower than that

estimated bywalking transects during 1991 when both methods were used. As therecould be significant bias in estimates from vehicle counts because sambar may tendto avoid roadsides, the estimates from walking transects would be more unbiased.These indicate a sambar density of7.9 animals/km" (Table ll.Taking a mean sambarweight of 134 kg, the average biomass of sam bar works out to 1058 kg/km! for thestudy area. This constitutes 29% of the biomass of important prey species atMudumalai (Table 2).

Table 1. Density of sambar as determined by vehic1e and walking transects.

Mode Year Dry First Wet Second Wet AnnualD C I D CI D C I D CI

Vehicle 1988 0.07 0.35 1.71 0.44 0.69 0.30 0.94 0.23Vehicle 1989 1.99 0.48 1.92 0.&4 0.51 0.21 1.32 0.33Vehicle 1990 0.87 0.27 0.60 fO.22 0.64 0.29 0.96 0.23Vehicle 1991 1.79 0.48


4/12

276

Table 2. Biomass of prey species in the study .area.

Prey species Prey weight Prey density(kg) (per km2)

Chital 41 31.2Sambar 134 7 . 9Gaur 450 2,4l.angur 9 25.2Total

Prey biorn asstkgikm2)

Prey biomass(~)

12.7910591080227

3645

3.5.029.029.66.4

Table 3. Prey preference of major predators.

Prey spi" ties Prey remains in predator scatWild dog 'k Panther 'k

Chital 424 61 149 5 9Sam bar ?;j9 39 9 8 39Gaur 0 0 0 0l.angur 0 0 3 2

% Prey remains divided by % prey biomass

57514o

Prey preference tratio*)% W. Dog ?an Tiger51 1.7 1.8 1.446 1.3 1.3 1.83 0.0 0.0 0.10 0.0 0.2 0.0

Tiger

Predation on SambarTiger, panther and wild dog are the major predators in Mudumalai. Sambar

chital, gaur and langur are the main prey species for these predators. Of all theseprey species, sambar hair constituted 38.8% inwild dog scat (n = 269), 39.2% inpanther scatIn =98) and 45.5% in tiger scat In e51) (Table 3). The sarnbar biomass,however, constitutes only 29% of the total prey biomass.

Habitat Utilization PatternsThe proportion of the sambar population found seasonally in different habitat

zones in relation to the proportional availability of the zones shows the changinghabitat utilization pattern. (Table 4 and Fig. 2). Sambar clearly utilized the DDFrzone in greater proportion (bya factor of'two) than its availability during the first andsecond wet seasons. But during the dry season there is a distinct shift. to the MDFzone ..The DDFS zone was utilized only in proportion to area during the dry seasonand under-utilized during other seasons. DTF was clearly shunned throughout theyear. The riparian vegetation was again underutilized, but this could be because ovehicular disturbance here along the transect route. Also, this zone adjoins the


5/12

21

Table 4. Habitat utilization pattern of sambar in dit'f'erent vegetative zones.

% sambar indifferent habitatsZone Area

Dry First wet Second wet% % 'l>

DDFTMDFDDFSRPFDTF

31.7533.0518.043.5044.97

24.1625.1013.702.8134.23

21.9062.5017.302.506.80

50.6028.108.802.10

lOAD

56.935.93.90.942.~

7 DRY- WET 2~ll 6NC/) 50 . . ..:) 4a~ 3

C) 2>-co

DDFT MDF DDFSZONES

RPF DTF

Figure 2. Seasonal group size of sambar in different zones.

et season and 44 during the second wet season. The average group size of sarnbarall seasons was 3.1 animals. This varies from one season to another; the mean

roup size of 4.23 animals during the first season differed significantly from thosef other seasons (dry = 2.55, second wet = 2.66) (Table 5a and 5b). The frequencyistribution of group size (Table 6) was significantly different between dry and firstet seasons (G = 81.6, df"" 6, p


6/12

7 8

Table 5". Avel"age group size of sambar- in different seasons

Season No Average Group 3D !l.Iax-.Dry ;l3B 2.55 2.10 19Wet 1 203 4.2Q 5.50 50Wet 2 153 2..60 4..30 44Annual 694 3.06 3.90 50

Table Sb..A,,'erage group size of sambar in different zones in different seasons.

Dry We.t 1 Wet 2Zone No. Mean SO No. Mean SD No.. Mean SDgroup size group size group sizeDDIT Ul 2.1 2,29 108 5.3 6.87 83 2.9 5..00MDF 137 2.5 1.95 110 3.0 2.65 6 4 2.2 2.39RPF 51 2.9 1.74 23 6.7 9.02 14 2.9 4.53DDFS 88 3.0 2.43 30 6..6 10.10 30 2.3 2.28DTF 34 1.9 1.29 19 1.9 1.14 3 1.3 0.47

Table 6. Group .ue frequency of sambar indifferent seasons,Seasons

Group size Dry(%) Wet.l (%) Wet 2 (%) Aimual(%)1 4 0 30 51 402 21 25 23 2 23-4 27 2 3 14 2 35-8 11 9 7 109-16 1 9 3 417-32 0 4 1 1.33-64 0 0 1 0

When we look at the propensity of females and males to form. unisexual groupsis. excludes solitaries and female plus fawn groups) asopposed to bisexual groups,e see that females have a much greater tendency to form unisexual groups (75%,:: 476 individuals) as compared to males (37%, n = 104 individuals).The frequency distributions of groups of different compesition-call-female, feale-plus-fawn, all-male and' bisexual-in relation to group size, is given in Tablethe group size increases the proportion offawns in the group also increases (Fig.


7/12

2 79

60 DRY WET 1 WET 25 0 - I Z Z l r.2> -04 0Z

W:J 3 00W 20a:LL100 2 4 8 16 3 2 64

GRJUP SIZEFigure 3 . Group size frequency ofsamhar in differt!nt seasons,

60

wCJ-cf-ZUJoa:w0 . . .

FM, FEMALEYF: YEARLING FEMALEHAM: HARD ANTLER MALEVAM VELVET ANTLER MALESAM: SPIKE ANTLER MALEShAM: SHED ANTLER MALEF: FAWN

FM YF HAM VAM SAM ShAMSEX & AGE CLASS

F

Figure 4. Proportion of different sex and .Age classes in sambae,

easons with peak during dry season, when overall 66 .7o /c of males were in hardntler. During the first wet season a majority (62 .. 2%) of sam bar were seen in velvetntler but a substantial number of males were also in hard antler (37, 7r; \ t TebleSnd Fig. 6J. These obser-vations indicate that rutting in sam bar is spread over severalduring the second wet and dry seasons, with perhaps a peak during the dry


8/12

280

Table 7. Composition of social groups in sambar.

~li)(edGroup size All-female Female/fawn Female/male

No. % No. 'lc x e . 'C2 59 5 2 27 23 16 143-4 36 2 9 42 3 4 3 9 315-6 10 25 11 2 7 1 9 477 -8 8 1 0 09 -1 9 4 100

All-maleNo. 't-12 106 4o 0

Cf)Zs : 0 . 2LLLL00.15Z0 I- 0.1c r :0o,00.05c r :n,

02 3

4 5 6GROUP SIZE7 8

igw:-e 5. Proportion oCfawns in groups of different sizes.

able 8. Seasonal proportion of hard- and velvet-antlered males in sambar.


9/12

Figure 6. Seasonal proportion of hard and velvet an tIered males.

Table 9. Fawn to female ratio in different seasons.

Season No. fema.les No. fawns Fawn/female ratioDryWet 1Wet 2

40823 4138

965323

0_240 . 2 30.17

DISCUSSIONSam bar generalIy seem to prefer forest with dense vegetation cover over more

open forest and roadside clearings. The substantially higher density estimatearrived at by walking transects than by vehicle counts along roadsides is one clearindication of this preference. Similarly, the highest densities of sarnbar are attainedin moistdeciduous forest or dry deciduous forest with tall grass cover as opposed todrier and more open forest. Sambar density (7 .91k.m2 ) and biomassin Mudumalai areamong the highest reported in Asian forests. Only Johnsingh's( 1983) estimate of 9sambar/krn? (1399 kg/krn'') during wet and 7 sambar/km? (1244 kg/km''] during dryseason in the adjoining Bandipur National Park compares with the Mudurnalaifigures. However, this applies only to a small 23 km2 study area where the largenumber of water holes could have caused the high density. Other estimates includeIs8 kg/km" for Wilpattu National Park, Sri Lanka (Eisenberg and Lockhart 19721,443 kg/km? in Chitwan, Nepal (Seidensticker 1976) and 736 kg/krn" in Nagar hole,


10/12

282

mainly graze when fresh green grass is available ISchaller 1967) as during the firstwet season. During the second wet season the grass stratum in dry deciduous forestbecomes more fibrous and in the dry season is unpalatable and nutritionally poorISukurnar 1989). This change in grass quality could explain the local movement ofsarnbar from one habitat type to another. Grass in the dry deciduous forest with tallgrass is most palatable after the rains during the first wet season. During the dryseason sambar shift. to higher quality forage (both grass and browse! in the moistdeciduous forest. Forest fire may also cause. the movement ofsambar from one areato another (Karanth and Sunquist 1992). Sambar actually frequent fire-burnt areasto lick the ash which is a significant source ofminerals,Animals which live in open areas tend to form large groups as an anti-predatory

strategy. On the other hand, living in groups does not confer any advantage forterrestrial prey animals living in dense cover. Being a deer that prefers relativelydense forest, sambar show a lesser degree ofgrouping; s.scompared to axis deer (A-risaxis!. This is also in accordance with the predator-avoidance hypothesis propoundedbyothers (e.g. Dasmann and Taber 1956, Hirth 1977).ln dense cover a single animalcan be efficiently concealed, whereas this concealing effect is lost with a brgt:r group.

As studies on a variety of other ungulates have shown, the group size and social"organization may also be influenced by relative amounts and distribution of foodresources (Clutton-Brock et al, 1982). The increasein mean group size during thefirst wet season could be due tohigher foodavailability at that time. When food is nota limiting factor, there would be little or no competition among individuals foragingin large groups.An interesting observation is that group sizes recorded by us at Mudumalai arelarger than those reported by others for any region. Maximum number in a grouprecorded by others include 9 at Kanha by Schaller (1967), 15 at Mudumalai byKrishnan (1972) and 35 at Bandipur by Johnsingh (1983). Ganesh Prasanna (1990)observed a grou p of45 sambar at Mudumalai. Krishnan (1972) speculates that theselarger groups are aggregations of individuals from different family groups nearwaterholes or in open areas. However, we have seen large sam bar groups away fromsuch areas. The .larger group sizes has led to the higher sambar densities atMudumalai during our study.The age and sex. composition of the sambar population at Mudumalai is very

similar to that observed in other regions, including at Kanha by Schaller (1967). Theex ratio of2.7 females to one male is also similar to ratios observed elsewhere; 2.4:1t Bandipur(Johnsingh 1983),2.4:1 at Nagarhole (Karanth and Sunquist 1992), 3:1t Kanha (Schaller 1967). Erroneous field classification could also cause female-iased sex ratio. In some cases, spike males and shed-antler males could have beenlassified as females when sighted from a considerable distance. However, this is notto change the sex ratio substantially, and a female-bias would still remain. One


11/12

2 8

higher rate of predation in males, both as fawns and as adults, or unequal sex ratioat birth..The latter explanation isunlikely; it ismore likely that male sambar sufferhigher mortality due to predation or other natural causes as observed inmany otherpolygynous mammals.The tendency for a higher frequency of fawns in larger groups may be an anti-predatory strategy in sambar, Fawns, which are highly vulnerable to predation, may

e better protected in larger groups. The observation that larger sambar groups arenvariably mixed groups of males and females may be partly a chance event, asandom association of individuals is more likely to bring the sexes together as groupize increases. However, this may also be related to reproductive activity and mateIn cervids, changes in antler stage or presence of breeding stags in groups may

ndicate a seasonality in reproduction (Barrette 1987). In sambar the developmentf hard antlers in males, sore patch, territoriality, wallowing and courtship behav-our may indicate the rut (Johnsingh 1983). Johnsingh (1983) reported a higherroportion .of hard-antlered males between November and April at Bandipur, butince hard antlers have been found to be associated with females from October toune, he concluded this to be the period of rot. In Kanha, Schaller (1967) reportedrut spread over a period of seven months with a peak during November andecember. In the high altitude (:>2000m) Eravikulam National Park sambar gaveirth in the post-monsoon, indicating a rut during the pre-monsoon (Rice 1986).rishnan (1972)' states that, based on shedding and regeneration of antlers, thevidence for seasonality in reproduction is confusing. Our own observations onmalesn hard antler and female/fawn ratios also indicates that at best there may be onlyweak seasonality in reproduction.In sambar, the sore patch is said to be one of the manifestations ofthe early part

f rut, and sam bar develop the sore-patch only during certain months (Schaller 1967).e observed sore patches in sarnbar throughout. the year and in almost an agelasses in Mudumalai, although we do not have quantified data on the frequency. Ife consider sore patch as an indicator of rutting season, this observation supportsseasonality in reproduction. Krishnan (1972), however, argues that the sore patchoes not have any sexual significance, because he observed sort patches on hindsith very young fawns. As the hinds were still in the phase oflactation, they couldot have been in breeding condition. In some places like Hazaribagh National Parkt no time were sambar seen with sore patches.Food and other resource availability, predator pressure, and climate are the main

actors which determine the phenology of reproduction in any species. In Mudumalaihere rainfall is spread oyer a period of about eight months, there may not be theistinct seasonality inecological and behavioural attributes ofa species as comparedo the more seasonal habitats found at northern latitudes in India.


12/12

2!J4

colleagues R. Arumugam for the data on predation and Milind Watve for comments OD themanuscript. Trackera Anoadurai and Sivaji provided valuable support in the field. Surendra Vannanalso thanks Dr. N.Ma:ruyama ofthe Department ofEn vironmentaJ Science and Conservation, TokyoNoko Univenlity, Japan; Dr. N. Ohtaishi of the Department of Oral Anatomy, Hokkaido University,Japan; and the Secret!uiat, International Symposium on Deer in China (lSDCI, Department ofBiology, East China Normal University, Shanghai. China, for accepting this paper for presentationat the [SOC and providing necessary funds to attend the symposium in China,

LITERATURE CITEDBarrette, C. 1~7. Mating behaviour and mate choice bywild axil! deer in Sri Lanka. Journal of the

Bombay Natural History Society 84:361-37LBurnham, KP., D. R. Anderson and J. L. Laake. 1980. Estimation of density from line transect

sampling u(biological populations. Wildlife Monographs 72:1-202.Clutton-Brock, T. H-. J , ' . E. Guinness and S. D. Albon. 1982. Red deer - behavior and ecology of two

sexes. The University of Chicago Press, Chicago.Dasmaan, R.F. and R .. D. Taber. 1956.Behavior of Columbian black-tailed deer with reference to

population ecology. JoumalofMammalogy 87:143-164.Eisenberg, J. F. and M. Lockhart. 1972.An ecological reconnaissance ofWilpattu National Park,

Ceylon. Smithsonian Contributions to Zoology 101. Washington, DC.Ganesh Prasanna, A J. 1990. Ecology of the sambar (Cervus unioolor) at Mudumalai Wildlife

. Sanctuary. MS dissertation, Loyola College, University ofMadras.Hirth. D. H. 1977. Social behavior of white-tailed deer in relation to habitat. Wildlife Monographs

63:1-55.ohnai~gh, A. J. T. 1983. Large mammalian prey - predators in Bandipur. Journal of the Bombay

Natural History Socie~y80:1-49.Karanth, K. U. and ~. E. Sunquist. 1992. Population structure, density and biomass of largeherbivores in the tropical forest ofNagarhole, India. Journal ofTropical Ecology 8:21-35.

rishnan, M. 1972. An ecological survey of the larger mammals of peninsular India. Journal of theBombay Natural History Society 67:496--501.

C. 1987. Habitat use by the sambar (Cervus unicolori in Thailand - a case study forKhao- Yai National Park. In Biologyand management ofthe Cervidae (C.M.Wemmer. ed.)pp.28~298. Smithsonian Institution Press. Wash.ington, DC.

ice, C. G. 1986. Observation on predators and prey at Eravikulam National Park, Kerala, Journalof the Bombay Natural History Society 83:283305.

G. B. 1967. The deer and the tiger - a study ofwildlife in India. University ofChicago Press,Chicago.eidenstic.k.er, J. 1976. Ungulate populations in Chitawan Valley, Nepal. Biological Conservation

10:183-210.B . D., B..V. Shetty, K Vivekanandan a~d N. C. Ramachandran. 1978. Flora ofMudumalaiWildlife Sanctuary. Journal ofthe Bombay Natural History Society 76:13-45.R . 1989. The Asian elephant =ecclogy and management. Cambridge University Press.

ukumar, R., H. S. Dattaraja, H. S. Suresh, J. Radhakrishan, R. Vasudeva, S. Nirmala and N. V.Jo.hi. 1992. Long-term monitoring ofvegetation in a tropical deciduous forest in Mudumalai,Southern India. Current Science 62(9):608--616.G. C. 1987. Program TRANSECT - line transect data analysis program. Colorado State

University, Fort Collins, Colorado.

Documents

Sam Bar Paper