ArticlePDF Available

Estimating the population structure of Javan rhinos (Rhinoceros sondaicus) in Ujung Kulon National Park using the markrecapture method based on video and camera trap identification


Abstract and Figures

The population structure of the Javan rhino (Rhinoceros sondaicus) in Ujung Kulon National Park (NP) in Banten, Indonesia was assessed using visual identification and mark-recapture estimation. The software program CAPTURE was used for selecting the best fit estimator for the mark-recapture calculation and yields M(th) as the best model. The software results delivered a mean estimation of 32 rhinos (a minimum of 29 and maximum of 47 rhinos) with a 95% confidence level based on the dataset obtained from April 2008 to September 2009. The visual identification suggests that the current population in Ujung Kulon NP is male biased by a 3:2 sex ratio of males versus females. The demography shows that the population consists of mainly adult individuals that have a tendency of 1% population growth per year.
Content may be subject to copyright.
90 Pachyderm No. 49 January–June 2011
Estimating the population structure of Javan rhinos (Rhinoceros
sondaicus) in Ujung Kulon National Park using the mark-
recapture method based on video and camera trap identification
Adhi Rachmat Sudrajat Hariyadi1*, Agus Priambudi2, Ridwan Setiawan3, Daryan Daryan4,
Asep Yayus5 and Hendra Purnama6
1,3WWF Indonesia, Jl Halmahera No. 9 Villa Admiral Lippo Carita km 8, Labuan-Pandeglang 42264, Banten,
2,4,5,6Ujung Kulon NP Authority, Jl Perintis Kemerdekaan No. 51 Labuan – Pandeglang Banten Indonesia 42264
*Corresponding author:
The population structure of the Javan rhino (Rhinoceros sondaicus) in Ujung Kulon National Park (NP) in
Banten, Indonesia was assessed using visual identication and mark-recapture estimation. The software pro-
gram CAPTURE was used for selecting the best t estimator for the mark-recapture calculation and yields
M(th) as the best model. The software results delivered a mean estimation of 32 rhinos (a minimum of 29
and maximum of 47 rhinos) with a 95% condence level based on the dataset obtained from April 2008 to
September 2009. The visual identication suggests that the current population in Ujung Kulon NP is male
biased by a 3:2 sex ratio of males versus females. The demography shows that the population consists of
mainly adult individuals that have a tendency of 1% population growth per year.
Key words: Javan rhino, population, mark-recapture, population estimation, camera trap
La structure de la population du rhinocéros de Java (Rhinoceros sondaicus) dans le parc national d’Ujung
Kulon à Banten en Indonésie a été évaluée en utilisant l’identication visuelle et une estimation de capture-
marquage-recapture. Le logiciel CAPTURE a été utilisé comme le meilleur modèle pour sélectionner le
meilleur estimateur propre au calcul et aux rendements M (th) de capture-marquage-recapture. Les résultats
du logiciel ont donné une estimation moyenne de 32 rhinocéros (un minimum de 29 et un maximum de 47
rhinocéros) avec un niveau de conance de 95% d’après la série de données obtenues d’avril 2008 à septem-
bre 2009. L’identication visuelle suggère que la population actuelle dans le parc national d’Ujung Kulon
est biaisée en faveur des mâles avec un rapport de 3 mâles contre 2 femelles. La démographie montre que la
population se compose principalement d’individus adultes qui ont une tendance de croissance de 1% par an.
Pachyderm No. 49 January–June 2011 91
The main indicators of success in conserving an en-
dangered wildlife species are the population size and
the net population growth. The Indonesian govern-
ment set the growth of the Javan rhino population at a
rate of 3% annually in order to achieve the conserva-
tion goal set in the Indonesian Rhino Conservation
Strategy (PHKA, 2007). Unfortunately, the existing
condition of the habitat (dense forest) and the scarce
distribution Javan rhinos do not allow the application
of census methods to obtain exact counts such as those
like direct counts, which can be implemented for deer
and pheasants (Lewis, 1970), or using aerial counts
as done for wildlife in the Serengeti-Mara region
(Talbot & Stewart, 1964). Fortunately, the distribu-
tion of Javan rhinos has been extensively studied
(Hoogerwerf, 1970; Muntasib, 2002), so there is no
immediate need for conducting a patch occupancy
survey, which would tend to have a bias due to the
inability of this method to accurately detect the pres-
ence of an animal in a given survey area if not done
repeatedly (Mackenzie & Royle, 2005). Due to the
difculties in applying a direct count/exact count,
the existing conditions for Javan rhinos should only
rely on a relatively accurate estimation of population
structure, instead of on exact counts.
One method that offers relatively accurate
population estimates of wildlife species is the mark-
recapture calculation based on visual identication, as
done on tiger populations in India (Karanth & Nichols,
2002). The use of camera traps for visual identication
of rhinos in Ujung Kulon NP was initiated by Grifth
(1993), and was followed by Yahya (2002) to study the
distribution and by Hariyadi et al. (2010) to study the
behaviour of the Javan rhinos directly in their habitat.
The Javan rhino is known to have frequent
wallowing behaviour because moisture/water is
required to ensure the integrity of their epidermis
(Shadwick et al., 1992), as well as for thermoregulation
(Schenkel & Schenkel-Hullinger, 1969). Failure to
wallow will lead to dryness that could eventually lead
to pathological conditions and pain in the epidermal
tissue (Munson et al., 1998). The need to wallow for
both male and female Javan rhinos in Ujung Kulon
NP is indicated by the daily wallowing behaviour
recorded in previous observations (Schenkel et al.,
1969; Hoogerwerf, 1970; Sajudin, 1984). There is
no known sexual dimorphism in the epidermal tissue,
and no differences in wallow frequencies between
male and female Javan rhinos have been reported.
The differences of wallow frequencies between males
and females are assumed to be very little based on
ndings from Yahya (2002), so they should show
similar wallowing frequencies and probabilities. If
this is true, there will not be a sex bias to take into
account with the survey design.
In using mark-recapture estimation of the
population size, it is important to determine the
population closure during the survey period.
Population closure is dened using two categories—
the demographic closure and the geographic closure
(White et al., 1982). Since the movement trends
of Javan rhinos have been previously recorded in
Yahya (2002) and Setiawan and Yahya (2002), the
geographic closure for the Javan rhino population
can be dened. However, due to the occurrence of
births and mortality during the survey period, the
demographic closure principles may not be fullled in
this study (White et al., 1982), so the analysis must be
carefully designed to ensure that demographic closure
is accounted for in order to select the appropriate
model for mark-recapture estimation.
Material and methods
Estimation of Javan rhino population size using
mark-recapture calculation requires identication of
each individual rhino. Considering the difculties of
physically capturing and marking each rhino, it is
agreed that the less invasive method to calculate is
through individual identication and differentiating
Javan rhinos from photos and/or video. The use of
automatic video recording devices (video traps) is an
approved method for estimating the population size of
the elusive species (Karanth & Nichols, 2002) such as
Javan rhino. Therefore, 34 DVREye automatic record-
ing devices were used for this purpose in the peninsula
of the Ujung Kulon NP from April 2008 to September
2009. Rhino habitat in Ujung Kulon NP consists of
a peninsula of 30,000 ha that is dominated mainly
by lowland rainforest, coastal and mangrove forest.
These 34 cameras were systematically placed
in the study area and 1 camera was placed in the
vicinity of holes at the height of 1.5 to 2.5 metres
above ground. These cameras were secured onto a
tree with 10 to 20 degree downward angles to record
any activity in the wallow hole. Wallow holes were
selected using several criteria such as: type of wallow
holes (temporary or permanent), the numbers of
rhinos known to use a particular hole and the size
of the wallow (area in m2 and the depth of water
Estimating Javan rhino population in Ujung Kulon National Park
92 Pachyderm No. 49 January–June 2011
and mud). Camera placement was mainly focused
on permanent wallows that were utilized by one or
more rhinos, while the area and depth of the wallow
holes were brought in as supporting information used
to categorize the wallow types. Selection of wallow
holes (camera spots) was made based on a grid system
to ensure geographical representation of the rhinos
whereby one wallow was selected within each sector
of the grid and only one camera was placed at each
wallow. Based on rhino movement and home range
of 1.4 to 3.8 km per day (Muntasib, 2002), the survey
area was divided using grids of 4 km2 to cover the
distance that rhinos travel every day. These grids
divided the known rhino habitat of the peninsula into
50 grid squares; 35 out of these 50 grid squares were
then selected based on the high rhino occurences
recorded in a previous study (Muntasib, 2002; Yahya,
2002). The video trap camera placement locations are
shown in Figure. 1.
Based on previous observations of Javan rhino
behaviour by Yahya (2002), it is safe to assume
that female and male Javan rhinos wallow at
approximately the same frequency; thus the camera
placements still allow all individuals to have a non-
zero probability of capture. This was further assessed
by comparing recapture rates of males and females.
The survey was conducted within a 10-year period to
ensure that the survey covered twice the length of the
rhinos’ reproductive cycle, as mentioned in Hariyadi
et al. (2008). The reproductive cycle was estimated
at 3 years with additional 2 years for mother-calf
afliation period; therefore, the period of 10 years
should represent 2 repetitions of the reproductive
cycle within the population. It is also safe to assume
that geographical closure is met, as no rhinos migrated
into or out of the rhino habitat during the survey. The
camera coverage (sample area) also represents the
known rhino home ranges based on Muntasib (2002)
and Yahya (2002); thus the areas without known
rhino home ranges were not sampled. Since there
are no migrations of rhinos into or out of the survey
areas, we conclude that migration does not violate
the demographic closure set up in the survey design.
However, the closure assumption will be tested using
Figure 1. A map of Ujung Kulon peninsula showing the sampling grid representing the known geographic
locations of the Javan rhinos.
Hariyadi et al.
Pachyderm No. 49 January–June 2011 93
CAPTURE software (White et al., 1978) to determine
the most suitable model for estimating the population
of the Javan rhinos.
Each colour represents one of three teams assigned
to installing the video trap camera equipment. These
video traps were placed to record rhino activities in
the selected wallow holes for 15 to 20 days at the
same locations. The installment date was marked by
assigning a person to walk in front of the camera while
holding a signboard containing the date to be recorded,
and the camera trap removal/data retrieval date was
marked in a similar manner. The period between the
installment and data retrieval dates was dened as a
survey period. Each of these survey periods represents
an ‘occasion’, which is the parameter repeated in the
calculation of mark-recapture.
Individual marking and identification were
done by observing and comparing morphological
features and using parameters developed by Grifth
(1993). This method relied on indicators such as:
horn shape and size, neck and eye folds, ear shape,
footprint size, as well as distinct features (scars, birth
marks) to differentiate individual Javan rhinos. At
least three parameters must be employed to make
a positive differentiation among individual rhinos
captured in the photos, while other morphological
features (scars, necks, neck folds, etc.) are used for
detailed differentiation among individuals found
within approximately the same habitat range, or
individuals within the same age class or sex. Special
attention was given to very young calves travelling
with their mothers, as they may indicate recent births.
The calves’ ages were estimated using a comparison
of body size between the calves and their mothers.
Calves estimated at 1 or 1.5 years-old that were not
sighted in the previous surveys would be categorized
as newborn. All newborn rhinos were tabulated for
calculating the birth rate of the population. Examples
of the differentiation method are presented in Fig.
2 and 3. Each identied rhino will be given a code
according to the sex, age group (calf, subadult, adult),
grid number where it was rst recorded and a unique
individual number to differentiate rhinos that may
be detected in the same grid square. With this code,
each individual rhino can be recorded, identied
and re-identified on different occasion(s). Each
video clip with rhino footage was analyzed using a
computer capable of running the VLC media player
programme, a software that allowed for considerably
accurate identication of rhinos based on visual and
morphological features mentioned above.
In addition to the individual identification
parameters described above, the rhinos were classied
into four different age classes based on their overall
body and horn size. The rst category consists of adult
rhinos; both male and female rhinos must have a horn,
though that of the male is typically larger. The second
category consists of subadults with relatively smaller
bodies and horns. The third category represents
calves with very small bodies—normally without
any distinct horn, and most of their time is spent with
their mother.
Analysis was done by identifying, marking
(coding) and recording each rhino that was captured
on video within the survey period (occasion). Rhino
data collection was implemented from April 2008 to
September 2009 to ensure demographic closure. In
May, June and September 2008, as well as February
and April 2009 video trap devices were not operated
due to maintenance and repair. Calculation for the
Javan rhino population estimate was performed using
the Lincoln–Peterson formula and using CAPTURE
software (White et al., 1978), which were known
to deliver accurate mark-recapture calculations. To
comply with the CAPTURE software requirements,
rhino detection was represented in a binary system,
whereby ‘1’ marks presence and ‘0’ marks absence
during each occasion. Data was processed using
analysis pattern x-matrix with a statistical population
estimation at 95% confidence level. Minimum
numbers of rhinos were determined by identifying
individual Javan rhinos that were recorded until
September 2009.
In order to study the population trend for the
past 10 years, results from this year’s analysis were
compared to those of 2000, 2004 and 2009, which
were analysed in the same manner by WWF and
Ujung Kulon NP authorities using the same camera
locations from January to December in each year, but
using different brands of cameras. This comparison is
expected to illustrate the population structure in 2000,
2004 and 2009, as well as describe the differences that
may have occurred (differences in age structures and
sex ratio that might indicate population dynamics).
All births from 2000 and 2009, with the addition of
records from 2010, as well as mortality (based on
ndings of rhino carcasses or remains) were put into
a table to calculate the actual population growth of
Javan rhino in Ujung Kulon NP. Surveys in 2000,
2004 and 2009 were conducted regularly every
month, so birth ndings should represent the actual
population trend. However, rhino deaths may not
Estimating Javan rhino population in Ujung Kulon National Park
94 Pachyderm No. 49 January–June 2011
reect the actual mortality rate, as all mortality nds
were opportunistic.
Throughout the survey periods, 27 rhinos were
identied using CAPTURE (White et al., 1978).
Examples of identication and marking of the Javan
rhinos are shown in Fig. 2 and 3. These rhinos and
their occurrences in each occasion are summarized
in Table 1. Recapture rate of females is 0.15 while
that of males is 0.17. The calculated closure analysis
values (Z=-2.533 and P=0.00565) suggest that the
closure assumption is violated. Further calculations
for model selection using CAPTURE software’s
‘Goodness of Fit’ models resulted in the 1.00 criteria
for M(th), suggesting that it is the best suited estimator
for mark-recapture, given the capture trend from the
dataset. The M(th) model calculates a mean estimation
of 32 rhinos with a standard error of 4.2381. Further
analysis shows that with 95% condence level it can
be ascertained that the Javan rhino population during
the survey period of June–September 2009 was be-
tween 29 to 47 rhinos. The Lincoln-Peterson formula
calculation, using data from the same survey period
to enable comparison between the two estimation
methods, reveals the mean of 41 individual rhinos
with standard error of 19.07, while manual identica-
Figure 2. The video capture equipment enabled the team to identify adult rhinos by the presence of the horn
in the male (B) and lack of distinctive horn in the female (A). Differentiating sex using horn presence can only
be applied for rhinos within the same age class (subadult or adult).
Figure 3. It is possible to differentiate rhinos based on the horn shape as a primary parameter. Note the blunt
horn (cone type) on the male rhino in B, while the male in A has a sharper (funnel type) horn tip (white arrows).
The rhino in B has a rounded jaw and longer prehensile (upper) lip while the rhino in A has a ‘square’ jaw as
indicated with a black arrow. Male A has continuous neck folds with no skin protrusions, while male B has
broken neck folds with conspicuous protrusions downwards from the neck (marked with a dotted oval).
Hariyadi et al.
Pachyderm No. 49 January–June 2011 95
Figure 4. Pie chart showing the composition of males
and females in the Javan rhino population surveyed
between April 2008 to June 2009.
Figure 5. The age structure of Javan rhinos in the
peninsula of Ujung Kulon NP surveyed from April 2008
to July 2009. The large percentage of calves indicates
th e breeding capability of this population.
Estimating Javan rhino population in Ujung Kulon National Park
Table 1. Summary of Javan rhinos identied through video trap implementation between April 2008 and
September 2009, with observation periods (occasions) represented by the months. Individual detection on
each occasion is marked as ‘1’, while no detection is marked as ‘0’
ID Rhino Apr Jul Aug Oct Nov Dec Jan Mar May Jun Jul Aug Sept
FADUB1401 0 0 1 10 0 010 100 0
M,Cal,B14,2 0 0 1 10 0 010 100 0
FADUB1403 0 0 0 10 0 000 000 0
FADUB2604 1 0 1 00 0 000 000 0
MCalB2605 1 0 1 00 0 000 000 0
FADUB2606 1 1 0 00 0 000 000 0
FADUB4407 0 0 0 00 0 110 000 0
FADUB5208 1 0 0 00 0 001 001 0
MSADB5209 1 0 0 00 0 000 000 0
MCal B5210 1 0 1 01 0 000 0 01 0
FADUB5211 0 0 0 01 0 000 100 0
MADUB5212 0 0 0 00 1 000 011 0
MADUB5513 0 0 0 00 1 000 010 0
FADUB5614 0 0 0 00 0 010 000 0
MCal B5615 0 0 0 00 0 010 0 00 0
MADUB3516 0 0 0 00 0 010 110 1
MADUB2117 0 0 0 00 0 010 110 0
MSADB2018 0 0 0 00 0 000 100 0
MADUB3519 0 0 0 00 0 000 110 0
MADUB3520 0 0 0 00 0 000 010 0
MCalB5221 0 0 0 01 0 000 010 0
FADUB4522 0 0 0 00 0 000 000 1
MADUB5723 0 0 0 00 0 000 000 1
MADUB5024 0 0 0 00 0 000 001 0
FSAB1725 0 0 0 00 0 000 010 0
FADUB1726 0 0 0 00 0 000 011 1
MCalB1727 0 0 0 00 0 000 011 1
96 Pachyderm No. 49 January–June 2011
tion from video footage yields 27 individual rhinos.
There were a total of 11 females and 16 male
rhinos identied from video capture, resulting in a
female to male sex ratio of 2:6 (41% females and
59% males), as shown in Fig. 4, while there are 18
adults, 3 subadults and 6 calves as shown in Fig. 5.
This result (2009) is compared with the results
from 2004 to study the population dynamics and show
the population composition over time (Fig. 6). The
comparison shows that the sex ratio does not differ
Figure 6. Comparison of
Javan rhino’s sex ratio (A)
and age composition (B).
The sex ratio recorded in
2004 and 2010 is consistent,
while there is an increase of
of adults and calves from
2004 to 2009.
Table 2. Compilation of birth ndings recorded from camera/video trap implementations from
2000 to 2010 with mortality based on carcass ndings throughout the period
2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010
Birth 2 3 4 32
Mortality 111 1 11 3
Hariyadi et al.
Pachyderm No. 49 January–June 2011 97
much between 2004 and 2009, but there is a difference
in the age structure of the population between these
periods of observation.
In order to study the population trend based on
birth rate (derived from camera trap and video trap
identications) and mortality rate (based on nding
of carcasses and remains etc.), all data from 2000 to
2009 were compiled and shown in Table 2. Based
on this data, the average birth rate is 1.4 births per
year (Standard Deviation: 1.5776), and the average
mortality rate is 0.9 deaths per year (Standard
Deviation: 0.8755) (Fig. 7).
Estimating the population of the Javan rhino using the
mark-recapture technique can be applied as an option
that is less invasive and possibly more discriminative
than the existing footprint count method. However,
this method relies on the accuracy of identication and
differentiation of individual rhinos as a prerequisite.
Detection of false positives will yield an inaccurate
estimation from the CAPTURE software, so photos
of individuals that cannot be identied cannot be used
for such analysis. Both the Lincoln-Peterson formula
and CAPTURE software will only yield accurate
results based on the accurate identication of rhinos
from photos and videos. Three-way identication
(identication made and agreed by three persons) is
considered as a valid evaluation procedure.
Similar recapture rates of females and
males (0.15 and 0.17 respectively) suggest
that sex bias using camera observation
due to differences in wallow frequencies
between male and female is negligible. When
comparing the results from CAPTURE and
Lincoln-Peterson, it is noted that there are
differences in standard error values that may
be attributed to the accuracy of each formula.
CAPTURE produces a mean population
estimate at 32 rhinos with Standard Error
4.2381, while calculation using Lincoln-
Peterson formula produces a mean estimation
of 41 rhinos with standard error 19.07. The
smallest standard error is produced from the
use of CAPTURE software, which suggests
a modest uncertainity in the population
estimate compared to the Lincoln-Peterson formula.
Based on the above calculations, and manual
identication from rhino photos, it can be concluded
that with a 95% degree of condence that the rhino
population on the peninsula of Ujung Kulon NP
during the period of April 2008 to September 2009
was between 29 and 47 rhinos. Closure test suggests
that the closure assumption is violated. This may
be true due to the births and mortalities of rhinos
throughout the survey period causing the demographic
closure to be violated; however, the geographic
closure is met during the surveys. Furthermore,
violation to the closure assumption is common when
a survey is carried out over a long period of time
where there is also a possibility of time dependent
wallowing behaviour due to the reduction of wallow
holes in the dry season.
In comparing the identication results on the
age structure between 2004 and 2009 it is clear that
although the sex ratios do not differ between the two
periods, the age structures do differ. The differences
in age structure may be attributed to the growth of
2004’s subadult individuals into adult individuals in
2010, thus indicating a shift towards an adult-biased
population. However, the increase in the percentage of
calves indicates that this population is still capable of
producing offspring, although almost all of the 2009
newborns are males, which pushed the sex ratio to
Descriptive analysis in comparing the birth
and mortality rates from 2000 to 2010 shows a net
population growth of ve rhinos within the 10-year
period (0.5 net growth per year or approximately
Figure 7. Comparison of average birth and mortality
of Javan rhino from 2000 through 2010.
Estimating Javan rhino population in Ujung Kulon National Park
98 Pachyderm No. 49 January–June 2011
1% population growth). Although not signicant,
and still much lower than the targeted 3% annual
population growth (PHKA, 2007), the population
might potentially be growing. However, the mortality
recorded in 2010 (three deaths) is relatively higher
than average, so further investigation on the cause of
death in 2010 is necessary. Since poaching has been
stopped, other causes of death of adult rhinos will
need to be carefully examined. The possibility of
disease must not be overlooked. This should enable
population managers to prevent an increase of the
mortality rate within these populations in the future.
In addition to studying the cause of mortality, such
as diseases and other abnormalities, the stagnant
population for the past 10 years suggests that the
carrying capacity of the Ujung Kulon peninsula
may have been reached. Therefore, future efforts to
save the Javan rhino must include: improving the
existing habitat quality (increasing food plant quality
and abundance), locating sites outside the current
geographical distribution to be designated as a second
habitat, and the translocation of individual rhinos to
the second habitat to start a new population. Having
another population of Javan rhinos outside their
current distribution in Ujung Kulon NP will greatly
increase the chance of survival of this species.
Implementation of camera/video trap surveys ap-
pear to be a good method for monitoring Javan rhino
population in Ujung Kulon NP, as it enables identica-
tion of individual rhinos. Accurate identication is a
prerequisite in using mark-recapture analysis, as the
identication serves marking of individual specimens.
Judging from the standard error produced from using
CAPTURE program and Lincoln-Peterson estimation
for mark-recapture it can be concluded that the former
yields lower uncertainity in the estimation, as it pro-
duces a much smaller standard error than the latter.
Based on this long-term study to monitor the popula-
tion of Javan rhinos (Rhinoceros sondaicus) in Ujung
Kulon NP, the rhino population size is estimated at
a minimum of 29 and a maximum of 47 individuals.
Despite the male-biased sex ratio the population is
capable of reproduction, but due to the the mortality
rate the net population growth is only 0.5 individual
per year (which corresponds to 1% population growth
per year between the years 2000 and 2010). In order
to achieve the target of 3% net growth of the Javan
rhino population, the birth rate needs to increase and,
importantly, the mortality rate needs to be reduced.
The research is made possible in co-operation with
Ujung Kulon NP Authority, Indonesian Ministry
of Forestry, WWF, International Rhino Foundation
(IRF), Asian Rhino Project (ARP), and WWF-AREAS
(Asian Rhino and Elephant Action Strategy). Also
thanks to Abishek, Stephan Wulfraat and Arnaud Lyet
for the valuable input on the statistical and editorial
aspects of mark-recapture estimation.
Grifths, M. (1993). The Javan Rhino of Ujung Kulon
An Investigation Of Its Population And Ecology
Through Camera Trapping. Indonesia: WWF.
Hariyadi, A.R.S., Santoso, A., Setiawan, R. and
Priambudi, A. (2008). ‘Automatic Camera
Survey for Monitoring Reproductive Pattern
and Behaviour of Javan Rhinoceros (Rhinoceros
sondaicus) in Ujung Kulon National Park,
Indonesia’. Proceeding of the 3rd International
Meeting on Asian Zoo/Wildlife Medicine and
Conservation (AZWMC).
Hariyadi, A.R.S., Setiawan, R., Daryan, Y.A. and
Purnama, H. (2010). ‘Preliminary behaviour
observation of the Javan Rhinoceros (Rhinoceros
sondaicus) Based on Camera Trap Surveys
in Ujung Kulon National Park’. Pachyderm
Hoogerwerf, A. (1970). Udjung Kulon the land of the
last Rhinoceros. Leiden: EJ Brill.
Karanth, K.U. and Nichols, J.D. (2002). Monitoring
Tigers and Their Prey: A Manual for Researchers,
Managers, and Conservationists in Tropical Asia.
Bangalore, India: Centre for Wildlife Studies.
Lewis, J.C. (1970). ‘Wildlife Census Method: A
résumé’. Journal of Wildlife Diseases. 6:356–364.
Mackenzie, D.I. and Royle, J.A. (2005). ‘Designing
occupancy studies: general advice and allocating
survey efforts’. Journal of Applied Ecology
Hariyadi et al.
Pachyderm No. 49 January–June 2011 99
Munson, L., Koehler, J.W., Wilkinson, J.E. and
Miller, R.E. (1998). ‘Vesicular and Ulcerative
Dermatopathy resembling Supercial Necrotic
Dermatitis in Captive Black Rhinoceroses
(Diceros bicornis)’. Vet Pathol 35:31–42.
Muntasib, E.K.S.H. (2002). Penggunaan Ruang
Habitat Oleh Badak Jawa (Rhinoceros sondaicus,
Desm 1822) Di Taman Nasional Ujung Kulon.
Disertasi Fakultas Kehutanan Institut Pertanian
Bogor (IPB).
PHKA (2007). Strategy and Action Plan for the
Conservaion of Rhinos in Indonesia 2007-2017.
Republic of Indonesia: Ministry of Forestry,
Sajudin, H.R. (1984). Studi Perilaku dan Populasi
Badak Jawa (Rhinoceros sondaicus Desmarest
1822) di Ujung Kulon. Tesis Fakultas Biologi
Universitas Nasional.
Schenkel, R., Schenkel-Hullinger, L. and Ramono,
W.S. (1969). ‘Area management for the Javan
rhinoceros (Rhinoceros sondaicus Desm.), A Pilot
study’. The Malayan Nat. J. 31:253–275.
Schenkel, R. and Schenkel-Hullinger, L. (1969). The
Javan rhinoceros in Ujung Kulon Nature Reserve:
its ecology and behavior. Acta Tropica 26:98–135.
Shadwick, R.E., Russel, A.P. and Lauff, R.F. (1992).
‘The Structure and Mechanical Designs of
Rhinoceros’. Dermal Armor. Phil. Trans. R.Soc.
London 337:419–428.
Setiawan, R. and Yahya, A. (2002). Studi Populasi dan
Distribusi Badak Jawa (Rhinoceros sondaicus,
Demarest, 1822) Melalui Perhitungan Koleksi
Feses dan Tapak di Taman Nasional Ujung Kulon.
Balai Taman Nasional Ujung Kulon & WWF
Indonesia Project ID0091.07.
Talbot, L.M. and Stewart, D.R.M. (1964). ‘First
Wildlife Census of the Entire Serengeti-Mara
region, East Africa’. The Journal of Wildlife
Management 28(4):815–827.
White, G.C., Anderson, D.R., Burnham, K.P. and Otis,
D.L. (1978). User’s Manual for Program Capture.
Logan Utah: Utah State University Press.
White, G.C., Anderson, D.R., Burnham, K.P. and Otis,
D.L. (1982). Capture-Recapture and Removal
Methods for Sampling Closed Population. Los
Alamos New Mexico: Los Alamos National
Yahya, A. (2002). Studi Populasi Badak Jawa
(Rhinoceros sondaicus) Melalui Kamera Penjebak
Infra Merah Di Taman Nasional Ujung Kulon.
Indonesia: WWF.
Estimating Javan rhino population in Ujung Kulon National Park
... Questions addressed by using camera traps are often related to animal ecology, behavior and conservation [1]. For example, camera traps have been used to study niche separation [2], competitive exclusion [3], population structure [4,5], density estimation with [6,7] and without individual recognition [8,9], abundance estimation [10], foraging behavior [11], biodiversity [12], activity patterns [13] and habitat use [14,15]. Camera traps can replace other study methods or add to direct observations, track inventories, knowledge of local inhabitants or genetic surveys [16][17][18][19][20]. ...
... Although pictures are more commonly used in camera trap studies and are easier to process, videos provide more detailed information, especially behavioral. They are used in determining competitive exclusion [3], observing time budget [38], observing behavior and determining population structure [4,39] and to study nest predation [40]. Videos are also more appealing to the general public when used as an awareness tool. ...
... The use of the filter protocol must be considered with knowledge of the species, species community and the study design. For example: when camera traps are used to perform capture mark recapture analysis on individually recognizable individuals [4,48,49] it is probably not acceptable to miss a recording since a single recording may have a large impact on results. For this study, a loss of 5% to 20% of the beaver footage was tolerable, since beavers were recorded frequently and the data were collected to determine average activity patterns. ...
Full-text available
Camera traps have proven very useful in ecological, conservation and behavioral research. Camera traps non-invasively record presence and behavior of animals in their natural environment. Since the introduction of digital cameras, large amounts of data can be stored. Unfortunately, processing protocols did not evolve as fast as the technical capabilities of the cameras. We used camera traps to record videos of Eurasian beavers (Castor fiber). However, a large number of recordings did not contain the target species, but instead empty recordings or other species (together non-target recordings), making the removal of these recordings unacceptably time consuming. In this paper we propose a method to partially eliminate non-target recordings without having to watch the recordings, in order to reduce workload. Discrimination between recordings of target species and non-target recordings was based on detecting variation (changes in pixel values from frame to frame) in the recordings. Because of the size of the target species, we supposed that recordings with the target species contain on average much more movements than non-target recordings. Two different filter methods were tested and compared. We show that a partial discrimination can be made between target and non-target recordings based on variation in pixel values and that environmental conditions and filter methods influence the amount of non-target recordings that can be identified and discarded. By allowing a loss of 5% to 20% of recordings containing the target species, in ideal circumstances, 53% to 76% of non-target recordings can be identified and discarded. We conclude that adding an extra processing step in the camera trap protocol can result in large time savings. Since we are convinced that the use of camera traps will become increasingly important in the future, this filter method can benefit many researchers, using it in different contexts across the globe, on both videos and photographs.
... The register can provide essential information on survival, fecundity and migration, which are determinants of population changes in response to local conditions and it is a basic tool that can inform decisions on how to respond to events such as an outbreak of disease, poaching, pollution or an increase in human pressure. For example, in Ujung Kulon, Hariyadi et al. (2011a) survey photos showed some rhinos with excessive salivation and in undernourished condition. Griffith estimated 37 to 58 rhinos with a median of 48, and an even sex ratio of 1:1. ...
... No specifics on sex ratios etc. were given by these authors. During an 18 month census carried out between April 2008 and September 2009, Hariyadi et al. (2011a) assessed the population at 29–47 rhinos with a mean of 32 rhinos, providing evidence of the first population decline in 25 years. They further estimated a sex ratio of 3:2 (male:female), while average birth rate was estimated at 1.4 births per year and the average mortality rate as 0.9 deaths per year. ...
... They further estimated a sex ratio of 3:2 (male:female), while average birth rate was estimated at 1.4 births per year and the average mortality rate as 0.9 deaths per year. Hariyadi et al. (2011a) did not expound on the causes of such low numbers, but pointed out that they found three rhinos that had died from illness. Four International Rhino Foundation (IRF) annual reports, from 2010 to 2013 inclusive, estimated there were up to 44 rhinos and that population numbers were stable. ...
... R. s. inermis formerly occurred in north-east India, Bangladesh and Myanmar and went extinct in the early 1900s ( Rookmaaker, 1980). R. s. sondaicus formerly inhabited Thailand, Malaysia, Java and Sumatra but recent camera-trapping surveys identified a minimum of 35 individuals comprising the last population of this subspecies, found in the 123,051 ha Ujung Kulon National Park, Indonesia ( Hariyadi et al., 2011;Van Strien et al., 2008). R. s. annamiticus formerly occurred in Lao, Cambodia, eastern Thailand and Vietnam (Groves, 1967;Khan, 1989;Van Strien et al., 2008). ...
... The protection and expansion of this population is the utmost priority for conservation of this critically endangered species. A recent paper using video-trapping mark-recapture methods showed a minimum of 29 animals were identified in Ujung Kulon National Park over an 18 month period between 2008 and 2009 ( Hariyadi et al., 2011), and recent unpublished data by Ujung Kulon National Park suggest a minimum of 35 individuals have now been identified using camera and video traps; both being considerably lower than past estimates. ...
... For similar studies, photo identification of Indonesian rhinos estimated the mean population size with a standard error of 19.07% (Hariyade et al. 2011). In snow leopards, a population estimation using photographs showed an error of 12.5% (Johansson et al. 2020). ...
Full-text available
The objectives of this study were to select morphological keys for the identification of individual endangered long-tailed gorals through analysis of photographic data and to use these morphological keys to determine the number and population composition of gorals living in the Osaek Region of Seoraksan National Park. Amongst 8149 photos taken using 73 cameras in the Osaek Region, 2057 photos of faces and horns were analysed. The presence and absence of horns, shape of the horns, proportion of the ring to the length of the horn and facial colour pattern were selected as morphological keys to identify individual gorals. To verify the accuracy of the morphological keys for identifying gorals, a blind test was performed on gorals residing in the sanctuary of the Yanggu Goral Restoration Center. The test revealed that the population and age of gorals were discerned correctly by the morphological keys, but there was a 12.5% error in discriminating between sexes in gorals aged over 10 years. Fifty-six gorals were identified from 2057 pictures, based on the morphological keys and methods developed in this study. The population of 56 individuals consisted of 43 individuals aged over 2 years (subadult or adult) and 13 offspring aged less than 2 years, with a ratio of 3.3:1. Of the total 56 individuals, 45% were adults aged 10 years or older, 18% were adults aged 3–10 years, 7% were subadults aged 2–3 years, 23% were offspring aged less than 2 years and 7% were individuals aged 2 years or older, whose age and sex could not be confirmed. The sex ratio of males to females was 1.17:1, with a corrected sex ratio of 1:1 considering the 12.5% error rate for gorals aged over 10 years, amongst the 39 gorals aged over 2 years.
... Meanwhile, the mortality rate in javan rhino was higher than black rhino but showed lower than white rhino. Hariyadi et al. (2011) obtained the average birth rate and mortality rate in javan rhino at UKNP (from 2000 to 2010) were 1.4 births/year and 0.90 deaths/year respectively. The average birth rate and mortality rate in this study (from 2010 to 2019) were 3 births/year and 1.3 deaths/year respectively and higher than the previous study. ...
Full-text available
ABSTRACT: Javan rhino (Rhinoceros sondaicus) is one of the rare animals with critically endangered status. The Ujung Kulon National Park (UKNP) is one of javan rhino habitat in Indonesia. Every year the monitoring program for javan rhino in UKNP was performed to identify the animals. This research was aimed to analyze the population structure of javan rhino based on the records data from 2011 to 2019 in UKNP. Research showed that the natural increase (NI) and birth rate (BR) values were 17.34% (moderate) and 67.33% (high) respectively. The net return rate (NRR) value was 15.38% (male) and 14.28% (female). The NRR value in the present study was lower than 100% and caused by less number of animals in a population for 30 years of breeding length. Despite this, the inbreeding rate of javan rhino at UKNP in 2019 was 0.01 (low). It was concluded that the natural increase of javan rhino at UKNP showed a good parameter but the population number needs to be increased.
... This estimate is considerably lower than track surveys during this time period and would suggest almost a doubling to our 2013 estimate. A follow-up camera survey carried out only in a portion of the previous survey area in 2009, estimated 32 rhinos (29-47, 95%;Hariyadi et al. 2011), however this would suggest a biologically unrealistic increase between 2009 and 2013. More likely was that the limited number of cameras and useable photos/videos for individual identification in 2009 led to an underestimate of the population. ...
Full-text available
The Javan rhino (Rhinoceros sondaicus) is one of the most threatened mammals on Earth. The only remaining individuals live as part of a small population isolated in a single protected area, Ujung Kulon National Park, Java, Indonesia. Despite almost a century of studies, little is known about the factors that affect Javan rhino demography and distribution. National park officials require such information to identify conservation strategies and track the success and failures of these efforts; translocating selected individuals to establish a second population has been considered, but the risks must be weighed. We show that the 2013 global population of Javan rhinos was 62 individuals, which is likely near the site's carrying capacity. Our analysis of rhino distribution indicates that tsunamis are a significant risk to the species in Ujung Kulon, justifying the risks of establishing additional populations. Continued individual-based monitoring is needed to guide future translocation decisions. This article is protected by copyright. All rights reserved
... Rhinoceros sondaicus inermis of eastern India (Bengal and Assam), Bangladesh and Myanmar went extinct in the early 1900s (Khan, 1989; Rookmaaker, 1980). Rhinoceros sondaicus sondaicus, of Thailand and Malaysia to Sumatra and Java (Indonesia), was, by the 1930s, considered so rare that a special reserve was established, Ujung Kulon National Park, Java, where at least 35 persist: the taxon's only surviving population (Hariyadi et al., 2011; Loch, 1937). The third subspecies, Rhinoceros sondaicus annamiticus, was formerly widespread and at least locally abundant in Vietnam, Lao PDR, Cambodia and eastern Thailand (Corbet and Hill, 1992; Rookmaaker, 1980 ). Historical information on abundance and distribution is incomplete. ...
Full-text available
Ujung Kulon National Park (UKNP) is a natural world heritage site located at the western tip of Java Island and on the edge of the Indian Ocean. This area of 122,956 ha has the potential for tsunami hazard originating from Mount Anak Krakatau and the Sunda Arc subduction zone. Almost no residents live in the UKNP region, however, it is the only place on earth that remains a habitat for the Javan rhino ( Rhinoceros sundaicus ) whose population is less than 100. This study aims to discuss the potential for tsunami hazards in UKNP originating from the earthquakes in Indian Ocean. The shallow water equation model was used to simulate the generation and propagation of tsunami waves. A total of 50 numerical gauges with intervals of 4 km were used as assessment points placed along the 190 km coastline of UKNP. From the simulation it can be seen that the height of the tsunami reaches 12.9 m around the coast of UKNP causes this area is vulnerable to tsunami hazards. This information can be used as consideration for the management of the UKNP area so that it can continue to preserve flora and fauna, especially to avoid the extinction of the Javan rhino.
A monitoring project of the Javan rhino was conducted so as to understand the extent to which the growth of this population has succeeded. Monitoring was conducted by making use of camera traps, which were strategically placed by using a stratified sampling method based on the area of concentration of Javan rhino. The population size of Javan rhino in 2013 was a minimal 58 individuals consisting of 8 calves and 50 sub adults or adults with a sex ratio of 35 males: 23 females. The birth rate was recorded at 13.79% while the mortality rate was 3.45%. We also recorded 4 new calves in 2013. © 2015 International Union for Conservation of Nature and Natural Resources. All rights reserved.
Full-text available
Past studies on tropical carnivores and other secretive animals relied on indirect evidence of animal presence such as tracks, scats, or scrapes. While such evidence can be useful for basic studies, using remotely-triggered camera traps offer researchers more reliable evidence of animal presence and, with appropriate study design and analysis, provides an array of opportunities to investigate carnivore ecology. We present an overview on camera trap uses for the study and conservation of wildlife, with a particular focus on tropical carnivores. Our goals are to promote proper and effective application of camera trapping and related analyses. We highlight major research avenues, give relevant examples and lessons learned from published material and from our own experiences, and review available resources for implementation, from preparation and camera trap fi eld set up, to data management, analysis, and presentation of results. Our review considers sampling design with respect to target species or groups of species, the state variable(s) of interest, what constitutes a sample, sample size needed, collection of supporting data (independent variables), reducing bias/minimising error, and data collection schedule. We also highlight some available camera trap database management packages and available statistical packages to analyse camera trapping data. We discuss presenting fi ndings to a wider audience so results become useful in the conservation and management of species. Finally, we discuss future development of camera trapping technology and related techniques for the study and conservation of carnivores in the tropics.
Full-text available
The behaviours of ten critically endangered Javan rhinoceros (Rhinoceros sondaicus) were observed using video camera traps on the peninsula of Ujung Kulon National Park (06o 38’30’–06o 52’30’ south and 105o 12’00’–105o 37’30’ east), and were organized as descriptive lists of activities (ethogram). Behavioural data were analyzed by examining the length of time each individual rhino displayed a specific activity (duration). Duration of each activity was calculated as a proportion within a total observation time (length of rhino observation in video recording). In addition to duration, the frequency of each activity was recorded. A quantitative analysis summarizing duration and frequency of activities will be used as baseline information about Javan rhino behaviours that can enrich our knowledge of this reclusive species. The results from this study suggest that the use of video trap equipment for quantifying the behaviour of Javan rhinoceros is promising
Full-text available
The histopathology, clinical presentation, and epidemiology of a cutaneous and oral mucosal disease affecting 40 black rhinoceroses (Diceros bicornis) at 21 zoological parks (50% of the captive US population) were investigated. Twenty-seven biopsies were examined from recent lesions, and clinical information was available from 127 episodes. The cutaneous lesions began as plaques that progressed to vesicles, bullae, or ulcers. Lesions waxed and waned in individual cases. Lesions were predominantly bilaterally symmetrical, affecting pressure points, coronary bands, tips of the ears and tail, and along the lateral body wall and dorsum. Oral lesions were first noticed as ulcers and were present on the lateral margins of the tongue, palate, and mucocutaneous junctions of the lips. All recent lesions had similar histopathologic findings of prominent acanthosis, hydropic degeneration of keratinocytes in the stratum spinosum, spongiosis, intraepithelial vesicles, and parakeratosis without dermal inflammation. Chronic lesions were ulcerated. No pathogens were identified by culture or electron microscopy. Most episodes coincided with stress events (transportation, sudden cold temperatures, intraspecific harassment, estrus, advanced pregnancy) or concurrent diseases (toxic hepatopathy, hemolytic anemia, respiratory or urinary tract infections). Affected rhinoceroses usually were lethargic and had weight loss. Affected rhinoceroses also had lower hematocrit, serum albumin, and cholesterol values than captive healthy or wild rhinoceroses. The clinical patterns and histopathologic findings are similar to those of superficial necrolytic dermatitis in dogs and necrolytic migratory erythema in humans. The high prevalence of this skin disease in captive black rhinoceroses under many circumstances suggests that their epidermis is acutely sensitive to any disruption of metabolic homeostasis. We propose that metabolic changes secondary to a stress response from maladaptation or nutritional inadequacy of captive diets may contribute to the development of this disease in rhinoceroses without hepatopathies.
Full-text available
The collagenous dermis of the white rhinoceros forms a thick, protective armour that is highly specialized in its structure and material properties compared with other mammalian skin. Rhinoceros skin is three times thicker than predicted allometrically, and it contains a dense and highly ordered three-dimensional array of relatively straight and highly crosslinked collagen fibres. The skin of the back and flanks exhibits a steep stress-strain curve with very little 'toe' region, a high elastic modulus (240 MPa), a high tensile strength (30 MPa), a low breaking strain (0.24) and high breaking energy (3 MJm-3) and work of fracture (78 kJm-2). By comparison, the belly skin is somewhat less stiff, weaker, and more extensible. In compression, rhinoceros skin withstands average stresses and strains of 170 MPa and 0.7, respectively, before yielding. As a biological material, rhinoceros dorsolateral skin has properties that are intermediate between those of 'normal' mammalian skin and tendons. This study shows that the dermal armour of the rhinoceros is very well adapted to resist blows from the horns of conspecifics, as might occur during aggressive behaviour, due to specialized material properties as well as its great thickness.
The first aerial wildlife census of the entire 15,000-square-mile Serengeti--Mara region of East Africa was conducted between May 18 and 31, 1961. Careful counts were made of wildebeests (Gorgon taurinus) and zebras (Equus burchelli); final totals were 239,516 and 171,873, respectively. Less detailed counts or estimates were made of other species: buffaloes (Syncerus caffer), 21,832; elands (Taurotragus oryx), 6,400-9,600; elephants (Loxodonta africana), 1,157; hartebeests (Alcelaphus buselaphus), 2,100+; lions (Panthera leo), 550-700; black rhinoceroses (Diceros bicornis), 83+; topis (Damaliscus lunatus), 19,877; and Thomson's gazelles (Gazella thomsonii), 480,000-800,000. Slow, high-winged monoplanes were used, at elevations of 400-700 feet, flying 1- to 5-mile-long parallel strips, with the authors counting 150-400 yards on either side. Large herds were counted from color photographs. Maps of species distribution, topography, and vegetation are presented. The significance of the census results to research and management is discussed, and previous counts from parts of the area are reviewed. The census confirmed that a significant irregular interchange of zebras and wildebeests occurs between the Kenya and Tanganyika areas. The seasonal wildlife migration was shown to be an irregular dispersion over a vast area of animals apparently following available supplies of water and preferred food. Royal Air Force photographs of the major game concentrations provided a check on the accuracy of the census.
The problem of estimating animal abundance is common in wildlife management and environmental impact asessment. Capture-recapture and removal methods are often used to estimate population size. Statistical Inference From Capture Data On Closed Animal Populations, a monograph by Otis et al. (1978), provides a comprehensive synthesis of much of the wildlife and statistical literature on the methods, as well as some extensions of the general theory. In our primer, we focus on capture-recapture and removal methods for trapping studies in which a population is assumed to be closed and do not treat open-population models, such as the Jolly-Seber model, or catch-effort methods in any detail. The primer, written for students interested in population estimation, is intended for use with the more theoretical monograph.
Summary 1. The fraction of sampling units in a landscape where a target species is present (occu- pancy) is an extensively used concept in ecology. Yet in many applications the species will not always be detected in a sampling unit even when present, resulting in biased estimates of occupancy. Given that sampling units are surveyed repeatedly within a relatively short timeframe, a number of similar methods have now been developed to provide unbiased occupancy estimates. However, practical guidance on the efficient design of occupancy studies has been lacking. 2. In this paper we comment on a number of general issues related to designing occu- pancy studies, including the need for clear objectives that are explicitly linked to science or management, selection of sampling units, timing of repeat surveys and allocation of survey effort. Advice on the number of repeat surveys per sampling unit is considered in terms of the variance of the occupancy estimator, for three possible study designs. 3. We recommend that sampling units should be surveyed a minimum of three times when detection probability is high (> 0·5 survey − 1 ), unless a removal design is used. 4. We found that an optimal removal design will generally be the most efficient, but we suggest it may be less robust to assumption violations than a standard design. 5. Our results suggest that for a rare species it is more efficient to survey more sampling units less intensively, while for a common species fewer sampling units should be surveyed more intensively. 6. Synthesis and applications . Reliable inferences can only result from quality data. To make the best use of logistical resources, study objectives must be clearly defined; sampling units must be selected, and repeated surveys timed appropriately; and a sufficient number of repeated surveys must be conducted. Failure to do so may compromise the integrity of the study. The guidance given here on study design issues is particularly applicable to studies of species occurrence and distribution, habitat selection and modelling, metapopulation studies and monitoring programmes.
Wildlife census problems and sampling schemes are discussed. Basic census methods are presented. Examples of inventory techniques utilized for various wildlife species are mentioned.
Penggunaan Ruang Habitat Oleh Badak Jawa (Rhinoceros sondaicus, Desm 1822) Di Taman Nasional Ujung Kulon
  • E K S H Muntasib
Muntasib, E.K.S.H. (2002). Penggunaan Ruang Habitat Oleh Badak Jawa (Rhinoceros sondaicus, Desm 1822) Di Taman Nasional Ujung Kulon. Disertasi Fakultas Kehutanan Institut Pertanian Bogor (IPB)
Ministry of Forestry Studi Perilaku dan Populasi Badak Jawa (Rhinoceros sondaicus Desmarest 1822) di Ujung Kulon
  • Indonesia Republic
  • H R Sajudin
Republic of Indonesia: Ministry of Forestry, Sajudin, H.R. (1984). Studi Perilaku dan Populasi Badak Jawa (Rhinoceros sondaicus Desmarest 1822) di Ujung Kulon. Tesis Fakultas Biologi Universitas Nasional
Area management for the Javan rhinoceros (Rhinoceros sondaicus Desm.), A Pilot study
  • R Schenkel
  • Schenkel
  • L Hullinger
  • W S Ramono
Schenkel, R., Schenkel-Hullinger, L. and Ramono, W.S. (1969). ‘Area management for the Javan rhinoceros (Rhinoceros sondaicus Desm.), A Pilot study’. The Malayan Nat. J. 31:253–275