Table Of ContentJOURNAL
OF THE
BOMBAY NATURAL HISTORY SOCIETY
August 2001 Vol.98 No. 2
ESTIMATING THE DENSITY OF PORCUPINES IN SEMI-ARID
SARISKAVALLEY, WESTERNINDIA'
DiwakarSharma2
(With three text-figures)
Key words: Indirect estimates, direct estimates, dung decay, pellet groups, defecation
The population of the Indian crested porcupine (Hystrix indica) was estimated over a
periodof3months(May-July 1989)insemi-aridSariskavalleyintheSariskaTigerReserve,
Rajasthan. The methods included count ofpellet groups in four transects each in three
vegetation types, and direct counts ofanimals in a vehicle at night. The data collection
using these two approaches was repeated 17 times. Calculation of density for indirect
estimatesinvolveddataonthepelletdecayrateandthedefecationrate(ofcaptiveanimals).
The animal density was estimated to be 8.8 ±2.4 animals/ sq. km through direct count,
whilepellet group count provided an estimate of12.4 animals/ sq. km with a variation of
0.9 to 24.9 animals/ sq. km in different habitats. The influence ofsome factors such as
possibleuseoflatrinesitesandslowdecompositionrateondensityestimatesarediscussed.
Introduction feeds on crops (Alkon 1983, Gutterman 1987,
1988) and on the bark of trees (Choudhry and
To estimate the population density of Ahmad 1975, SharmaandPrasad 1992). Inspite
animals in the wild, quantifying faecal matter ofits economic importance, few estimates ofits
abundance, decay and defecation rate is population density in the wild are available.
considered to be useful and convenient (Neff Beingnocturnal,smallinsizeandshyofhumans,
1968, Rowland etal. 1984). Although anumber itishardtoobserveandverydifficulttoestimate
of studies have been conducted elsewhere its population. Therefore, methods involving
(Putman 1984, Koster and Hart 1988), in India indirect evidence can be very useful for
onlyafewstudieshaveestimatedthepopulation estimation ofitspopulation. Duringthe summer
of wild animals using this method (Sale et al. of 1988, the last of three consecutive drought
1990, Dekker etal. 1991). years,comparativelylargenumbersofporcupines
The Indian crested porcupine Hystrix were seen in Sariska valley in the Sariska Tiger
indica Kerris a nocturnal animal thatreportedly Reserve. This was a unique opportunity to
estimate itspopulationusing direct sighting and
'AcceptedJanuary,2001 faecal abundance.
2WildlifeInstituteofIndia,P.B.#18,Chandrabani The aim of this study was to estimate
DehraDun248001,Uttaranchal,India. porcupinedensityusingpelletgroupdensity, and
PresentAddress:GujaratEcologicalSociety,
5,GoldenApartment,Subhanpura, to compare the results with that obtained using
Vadodara390023,Gujarat,India. direct observations.
JOURNAL BOMBAYNATURAL HISTORYSOCIETY, 98(2), AUG.2001 161
ESTIMATINGTHEDENSITYOFPORCUPINES
Study Area temperature up to 47 °C) and monsoon (July-
October). Average annual rainfall is c. 650 mm,
Sariska Tiger Reserve (27° 20' N and 76° mostofwhich fallsbetweenJulyandSeptember.
25' E) is a dry deciduous forest (Champion and ThestudywasconductedbetweenMayand
Seth 1968) located in the semi-arid tract of July 1989 in the Sariska valley (Fig. 1). The
Rajasthan in western India. The weather is dry vegetationofthestudysitesconsistedofZiziphus
and hot during the day throughout the year, scrubland, mixed woodland and Ziziphus
barring the monsoon when it is hot and humid. woodland (Sharnia and Prasad 1992). The hills
There are three distinct seasons: winter surroundingthe valleyhadforests dominatedby
(November-February, with night temperature as Anogeissuspendula and were little used by the
low as 3 °C), summer (March-June, day porcupine(SharmaandPrasad 1992). Therewas
Fig. 1: Sariska valley showing the study sites
162 JOURNAL, BOMBAYNATURAL HISTORYSOCIETY, 98(2), AUG. 2001
ESTIMATINGTHEDENSITYOFPORCUPINES
very little ground cover and litter in the three Where,
vegetation types during the study period. Y = mean pellet group density (no./sq. km)
The major mammals in the Sariska r= mean decay rate per day
National Park include tiger Panthera tigris D = mean defecation rate per day
( ),
leopard P pardus), nilgai Boselaphus Pelletgroupdensity: Porcupinesdefecate
( . (
tragocamelus), sambar(Cei~vus unicolor), chital spindle shaped pellets, which are about
(Axisaxis),wildpig(Susscrofa)andfour-homed 1-1.5 cm in diameter and 3-5 cm in length. The
antelope (Tetracerus quadricornis). The density pellet number may range from 5-25 in a pellet
of animals, particularly of ungulates, is higher group. Pellets groups were counted in strip
in Sariska valley, due to better habitat, than in transects in the three vegetation types, selected
other areas ofthe Tiger Reserve. under stratified random sampling. Four strip
m
transects 500 x 2 wereplacedparallelto each
m
Methods other at 100 intervals in each vegetationtype
(Fig. 1). Each transect was visited 17 times
Direct Density Estimates during May-June. The number ofpellet groups
Since porcupines are nocturnal, they were thatwereavailableattheendofthestudyperiod
countedatnightusingvehicletransects(Rodgers after their accumulation and decay in each
1991), along a 10 km long road (Sariska to vegetation type was used to calculate the pellet
Kalighati), in Sariska valley. In June and July, group density i.e. the number ofpellet groups
the census was done 17 times, between 2000 to per sq. km.
2030hrs,by fourobservers, inajeep drivenata Pellet decay rate: Investigations of the
uniformspeed of20km/hr. The animals seen in decay rate ofpellets were carried out from the
theheadlights ofthe vehicle were recorded. The first week ofMay 1989 to the second week of
m
fieldcoveredbytheheadlightswas 10 oneither July 1989. Fresh pellet groups were located
side of the road. Porcupine density (D) was between May 10 and June 30. Eight groups of
calculatedusingthefixed-widthtransectmethod freshpellets encountered were marked, counted
(Rodgers 1991). andtheirconditionmonitoredonceaweek. Two
N periods i.e. summer (May to mid-June) and
D = pre-monsoon(mid-Junetomid-July)werechosen
2.L.B to compare the seasonal decay rates.
Where, The following equation was used (Barnes
N =number ofanimals observed, and Jensen 1987, Sukumar et al. 1991) to
=
L length ofthe transect in km, and estimate the decay rate for pellet.
B = width on one side ofthe transect in km.
ln(N)-ln(N
= t tl)
Indirect Density Estimates r
Porcupine density (N) based on pellet t
group count, defecation and decay rate was Where,
calculatedusing the following equation (Barnes r= mean decay rate ofpellet groups per day,
and Jensen 1987) N =number ofpelletgroups ofone week,
Y.r N = the number ofpellet groups ofprevious
N = wee]k, and
D t= time (in days).
JOURNAL BOMBAYNATURAL HISTORYSOCIETY, 98(2), AUG.2001 163
,
ESTIMATINGTHEDENSITYOFPORCUPINES
Defecation Rate: Two porcupines were Table2
observed in the National Zoological Park, New SUPMOMRCEURP(IMNAEYPTELOLMEITDD-JEUCNAEY)AINND
Delhi for 15 days, one in Sakkarbagh Zoo, PRE-MONSOON(MID-JUNETOMID-JULY)
Junagadh, for 19 days (in 1992) and another in
SayajibaugZooinVadodarafor21 days(in2000) S.No. Time(days) Condition
Summer Pre-monsoon
and their pellet groups counted. The defecation
ratewascalculatedasthenumberofpelletgroups 1 < 1 < 1 Fresh,moist,coveredwith
defecatedperday. mucous,intact.
Data was analysed using computer 2 1-7 1-7 Intact,butdry.
3 8-30 8-15 Drywithcracksover
software. Non-parametric statistical tests were thesurface;morethan75%
conducted for significance of results. pelletsdistinguishable.
Kruskal-Wallis tests were used to determine 4 30-45 15-30 Crackswidenandcolour
significance of variation among the three becomesdark;about50%
ormorestilldistinguishable.
vegetation types, while Mann-Whitney U 5a 30-45 Allpelletsdisintegrate.
statisticswereusedtodeterminethesignificance 5b >45 Lessthan50%
ofvariationbetween the two seasons. distinguishable.
Table1 decay rates of different pellet groups within a
DIRECTANDINDIRECTDENSITYESTIMA1uSFOR
season. Trampling by large mammals such as
PORCUPINEINDIFFERENTHABITATS
wild pig, sambar, chital and nilgai accelerated
Vegetationtype Pellet Animal Shrub the process. On the other hand the pellets not
density density density
(No./km2 trampled by large mammals were present even
)
(No./km2) Direct Indirect (No./ha) e6n0vidraoynsmenatf.teNrobteeirnmgitedaecfteicoantweadsionbsaermvoedisitn
Ziziphus
scrubland 9975 2.35±2.16 24.9 411 ±114 summer.
Ziziphus The pellet decay rate is shown in Figs 2
woodland 350 6.56±1.04 0.9 137±63
and 3. Table 3 shows the decay rate for various
Mixedwoodland 5075 17.6±4.02 12.7 79±55
Average 5133 8.8±2.4 12.8 206±77 pellet groups monitored during summer and
pre-monsoon. The average decay rate during
summer was 0.005 per day. This rate increased
Results
to more than twice, 0.011 per day during
Pellet Group Density pre-monsoon. The difference in the mean decay
There was a significant difference in the rates between summer and pre-monsoon was
pellet group density among the three vegetation significant(M-WtestsZ=2.31, 2 tailedp<0.02,
types (K-W tests x2=16.15, p<0.001, n=34) n = 27,19). For calculation ofanimal density,
(Table 1). on,ly the decay rate during summer (May-June)
are considered, as pellet group densities were
Pellet Decay Rate estimated only in this season.
The pellet condition was categorised into
five agegroups (Table 2). Pellets observedfrom Defecation Rate
June onwards (afterpre-monsoon showers) also Inthewild,thepelletswerespindleshaped
disintegrated through the same categories, but and discrete; while in the New Delhi zoo the
the decay was faster in the second category. pellets in a group were not as discrete. In
Results indicated a large variation in the SakkarbaghZoo(Junagadh)andSayajibaugZoo
164 JOURNAL, BOMBAYNATURAL HISTORYSOCIETY, 98(2), AUG. 2001
ESTIMATINGTHEDENSITYOFPORCUPINES
Days
——
PelletGroup 1 PelletGroup 2 PelletGroup 3
Fig. 2: Decay rate ofporcupine pelletsduring summer
Fig. 3: Decay rateofporcupine pelletsduring pre-monsoon
JOURNAL BOMBAYNATURAL HISTORYSOCIETY, 98(2), AUG.2001 165
ESTIMATINGTHEDENSITYOFPORCUPINES
DECAYRATESOFVTAaRblIeO3USPELLETGROUPS roadwas similartothatawayfromit. Therefore,
MONITOREDDURINGSUMMERAND it was presumed that porcupine movement and
PRE-MONSOONPERIODS distribution along the roads would be similar to
Pellet Summer Pre-monsoon those away from the road. The small width (20
(Maytomid-June) (mid-Junetomid-July) m) of the transect precluded the chances of
GroupNo. Pellet Decay Pellet Decay missing animals on it. Within each vegetation
group rate group rate type, density estimates from vehicle census at
size (perday) size (perday)
night had low variation, except in the case of
1 17 0.005 5 0.010 Ziziphus scrubland. Further, the overall low
2 8 0.004 23 0.012
3 11 0.011 11 0.014 variation in average density of the animals in
4 10 0.020 the study area was considered good to compare
5 11 0.007
Mean 0.005 0.011 indirect estimates with it.
The higher decay rate in pre-monsoon
(Vadodara), because of near natural diet, the season (mid-June to July) was due to the action
droppings were similar to those in the wild and ofdungbeetles andtermites. This startedwithin
were included for calculations. The diet a week afterthe first shower inthebeginning of
(includingaccesstowater)incaptivityisdifferent June. However, on hard (red soil) and bare
from that in the wild, hence the defecation rates groundthedisintegrationbytermitestooklonger
should be taken only as an approximation. If time.
possible, the defecationrate shouldbeestimated Freshpellets were not monitored after the
innaturalorsemi-naturalconditions. Defecation onset of monsoon in July because of rapid
rate ofporcupine in captivity varied from once disintegration. Often within a day or two about
to thrice a day. The average defecation rate was 50% or more of the pellets changed into an
2 ±0.15 per day (n = 40). amorphousmass.Further,denseundergrov±hdid
notpermit an easy search offresh pellets. Since
Ai^imal Density theweatherisdryfor8months,thedatacollected
The results showed a difference between duringthisperiodisconsideredtobemoreuseful.
the direct and indirect estimates of density in The observed differences in density
each vegetation type (Table 1). The overall estimates for the two methods were perhaps
density estimated by direct observation was 8.8 relatedto the shrub coverinthe three vegetation
±2.4 animalspersq.km,whiletheestimateusing types. Earlier results (based on debarking,
indirect method was 12.9 animals per sq. km. Sharma and Prasad 1992), show that the
porcupinespreferred mixedwoodland,Ziziphus
Discussion woodland followed by Ziziphus scrubland as
habitats. Onewould,therefore, expecttheanimal
The road used for vehicle census at night abundance and pellet density to follow this
passed through three different vegetation types, pattern. While the direct counts conformed to
each of which differed from the others in thispattern,thepelletdensityandindirectcounts
structure, composition, density andutility to the followed the reverse trend. This suggests that
porcupine (Sharma and Prasad 1992). This while the porcupines preferred to feed in mixed
resultedinvariationinporcupinedensityamong woodland and Ziziphus woodland, they spent
thevegetationtypes.Duetotheimpactofdrought substantial time in the dense undergrowth of
and livestock grazing, the vegetation along the Ziziphus scrubland. In Ziziphus scrubland, the
166 JOURNAL, BOMBAYNATURAL HISTORYSOCIETY, 98(2), AUG. 2001
1
ESTIMATINGTHEDENSITYOFPORCUPINES
shrub density was the highest, followed by only, while in captivitythe animals defecatedin
Ziziphus woodland and mixed woodland (Table the corners of the cages. This indicates the
1). Dense shrub provides escape cover for the possibilityofuseoflatrinesitesbytheseanimals.
porcupine (Prater 1993). Ifso, can this method be used for other animals
This pilot study attempted to validate the that use latrine sites?
methods used for indirect animal density Moreover, the semi-arid climate seems to
estimates. Eberhadt andEtten (1956), Etten and have played an important role in substantially
Bennet (1965), Neff(1968) and Putman (1984) slowingdownthedecayrateofpellets. Pelletsof
have comprehensively reviewed the limitations chitalandsambardefecated inthepreviousyear
andsourcesoferrorinsucha study. Asrevealed have been observed in Gir Wildlife Sanctuary
in this case, the differences in density estimates and National Park in Gujarat by this author,
between direct and indirect evidence strongly indicating their slow decay rate in similar
emphasise the need for aproper stratification of conditions. These aspects require more research
habitat,preferablyafterreconnaissance,followed and,perhaps, the inclusionofa correctionfactor
by adequate sampling (Grieb 1958). The results inthecalculationofdensitiesinsuchconditions.
indicate that to estimate the population of
nocturnalanimalssuchasporcupine, oneshould ACKNOWLEDGEMENTS
study all habitats including those that may be
avoided by the animal (Neu etal. 1974). I thank the Chief Wildlife Warden,
Even though most of these factors were Rajasthan, forpermissionto work inthe Sariska
takenintoaccount(Sharma 1989)thedefecation Tiger Reserve and Mr. H.S. Panwar, former
behaviour of porcupine may have played an Director, Wildlife Institute of India for logistic
important role in the indirect density estimates. andfinancial support. Dr. AjithKumar, Dr. S.P.
The droppings were clumped, andundernatural Goyal,M.V.KattiandT.R. ShankarRamangave
conditions were observed at certain locations critical suggestions on the manuscript.
ReferENCES
Alkon,P.U.(1983): ForagingecologyofIndianCrested ofIUCN/SSCAsianElephantSpecialistGroup,pp.
Porcupine{Hystrixindica)inNegevhabitats.Final 54-64.
Report.IsraelAcademyofScience. 14pp. Eberhardt, L. & R.C. Van Etten(1956): Evaluation of
Barnes, R.F.W. & K.L. Jensen (1987): How to count thepelletgroupcountsforcensusingdeer.J. Wildl.
elephants in forest. IUCNAfrican Elephant and Manage. 20:70-74.
RhinoSpecialistGroupTechnicalBulletin I: 1-6. Etten,R.C.Van&C.L. Bennet(1965):Somesourcesof
Champion,H.G.&S.K.Seth(1968):Arevisedsurveyof errorinusingpelletgroupcountsforcensusingdeer.
the forest types of India. The Manager of J. Wildl. Manage. 29:723-729.
Publications,Govt,ofIndia,Delhi,404pp. Grieb,J.R.(1958):Wildlifestatistics.ColoradoGame&
Choudhry,M.I.&A.Ahmad(1975):Trialsofpoisonous FishDepartment.96pp.
gases and baits against porcupines. Pale J. For. Gutterman, Y. (1987): Dynamicsofporcupine(Hystrix
25(2):46-50. indica) digging: Their role in the survival and
Dekker, A.J.F.M., S. Dawson & A.A. Desai (1991): An renewalofgeophytesandhemicryptophytesinthe
indirectmethod forcounting Asian Elephants in Negev Desert Highland. Israel J. Botany 36:
forests. In: Censusing Elephants in Forest. 133-143.
Proceedings ofan International Workshop. (Eds Gutterman, Y. (1988): An ecological assessment of
Ramaknshnan, U.,J.A. Santosh,&R. Sukumar). porcupine activity in a desert biome. In:
Southern India, 2-10 January, 1991. Technical Ecophysiologyofdesertvertebrates. Eds Ghosh,
&
ReportNo.2.AsianElephantConservationCentre P.K. I.Prakash,ScientificPublications,Jodhpur.
JOURNAL, BOMBAYNATURAL HISTORYSOCIETY, 98(2), AUG.200 167
.
ESTIMATINGTHEDENSITYOFPORCUPINES
Pp.289-363. Sale, J.B., A.J.T. Johnsingh & S. Dawson (1990):
Koster,S.H.&J.A.Hart(1988): Methodsofestimating Preliminary Trials with an Indirect Method of
ungulatepopulationsintropicalforests.Afr.J.Ecol. Estimating Asian Elephant Numbers. Report for
26: 117-126. IUCN/SSCAsianElephantSpecialistGroup.
Neff, D.S. (1968): Thepelletgroupcounttechnique for Sharma, D. (1989): Spatial and temporal patterns in
biggametrend,censusanddistribution:Areview. debarking by Indian Crested Porcupine Hystrix
(
J. Wildl.Manage. 32:597-614. indicaKerr)in SariskaNational Park, Rajasthan.
Neu,C.W.,C.R.Byers&J.M.Peck(1974):Atechnique M.Sc. Dissertation, Wildlife Institute of India,
foranalysisofutilizationavailabilitydata.J. Wildl. Dehradun,88p.
Manage. 38:541-545. Sharma, D. & S.N. Prasad(1992): Tree debarking and
Prater,S.H.(1993):TheBookofIndianAnimals.Fourth habitatusebyporcupine(Hystrixindica Kerr) in
impression. Bombay Natural History Society, SariskaNationalParkinWesternIndia.Mammalia.
Bombay,pp.215-217. 56(3):351-361
Putman,R.J.(1984):Factsfromfaeces.MammalReview Sukumar, R., K.S. Verman, J.A. Santosh & U.
14(1):79-97. Ramakrishnan (1991): Methods of estimating
Rodgers,W.—A.(1991):TechniquesforWildlifeCensusin elephant numbers: A review. In: Censusing
India AfieldManual.WildlifeInstituteofIndia Elephants in Forests. Proceedings of an
Dehradun. 82p. International Workshop (Eds Ramakrishnan, U.,
Rowland,M.M.,G.C.White&E.M.Karlen(1984):Use J.A. Santosh, & R. Sukumar). Southern India, 2-
ofpelletgroupplotstomeasuretrendsinestimates 10January, 1991,Technical ReportNo. 2. Asian
in deerand elk populations. Wild. Soc. Bull. 12: ElephantConservationCentreofIUCN/SSCAsian
147-155. ElephantSpecialistGroup,pp2-29.
168 JOURNAL, BOMBAYNATURAL HISTORYSOCIETY, 98(2), AUG. 2001
