DRAFT KOMODO DRAGON , TERRESTRIAL WILDLIFE AND HABITAT MONITORING GUIDELINE

KOMODO DRAGON , TERRESTRIAL WILDLIFE AND HABITAT MONITORING GUIDELINE IN THE KOMODO NATIONAL PARK

INTRODUCTION

In archipelagoes such as Indonesia, a nation with extraordinary high levels of biodiversity and endimicity, and where human perturbation is not uncommon, even within protected areas, overcoming the intrinsic biogeographic variation of managing such biodiversity is likely to be another challenge for conservation efforts. Komodo National Park (KNP) is a world heritage site, this area has a high biodiversity of tropical marine and the only potential habitat of giant endemic reptile Komodo dragon (Varanus komodoensis). Like most of National Parks in Indonesia, wildlife management in KNP restrain by less of reliable monitoring data base, because of limited source of fund, logistic, and local staff capability (Jessop et al., 2004). This guideline provides methods which technically can be adopted and used by KNP management to collect high accuracy and reliable database.

OBJECTIVE

The major outcomes of this guideline are to increase knowledge of differences among island within

KNP with respect to prey availability and its influence on the demographic features of Komodo dragon populations. Second, collection of the appropriate life-history information that can be used to build specific population models that could assist in the management and conservation of key terrestrial species within KNP. Third, increasing the capacity of staff from KNP to effectively monitor wildlife populations within the park.

Specific Aims to research:

1. Life history and demographic differences in Komodo dragon populations across 4 islands within KNP.

Mark-Recapture studies should be used to assess demographic variation among island populations with respect to the following demographic parameters:

a) Population abundance of Komodo dragons at 10 study sites across four islands.

b) Age specific growth rates of Komodo dragons at 10 study sites.

c) Spatial ecology of Komodo dragons at 10 study sites

d) Assessing annual female reproductive rates across specific island sites

e) Hatchling production to assess variation in recruitment

f) Sex ratio.

Also recommend for further studies

f) Age specific survival rates of Komodo dragons at 10 study sites.

g) Parasitological (health and desease) of Komodo dragon

2. Prey density and diversity among four islands within Komodo National Park.

These monitoring activities will quantify temporal and spatial difference of large ungulate prey within Komodo National Park. Key criteria gathered from this monitoring include the following parameters. Knowledge of these parameters is essential for determining interactions between Komodo dragon and their prey.

a) Spatial and insular differences in density of Timor deer in KNP.

b) Spatial and insular differences in density of wild pigs in KNP.

c) Spatial and insular differences in density of water buffalo in KNP.

d) Spatial and insular differences in density of rats in KNP.

e) Spatial and insular differences in density of tokay geckos in KNP.

3. Assessment of habitat and other key and endangered species

a) Assessment of food resource (grass) availability

b) Active nest of Megapode birds

c) Active nest and direct population counting by vantage count for Cockatoo

d) Monitoring diversity of birds

e) Assessment the presence of exotic/invasive species

METHODS

Field Methods

1. Life history and demographic differences in Komodo dragon populations across 4 islands within KNP.

Mark Recapture is a very effective method that most scientists implement to collect information on population ecology, age specific growth rates and spatial ecology (inter valley or inter islands migration) of Komodo dragon. Radio telemetry method is applied to assess more detail spatial ecology, i.e. movement pattern, home range, and behavior. Nest survey method that implemented transect grids method is very useful to assess annual female reproductive rates. Nest fencing method also useful to assess variation in recruitment by counting hatchling production from active nests.

Mark recapture

Komodo dragons are captured by baited trap, noose or by hand. These methods are extremely effective for capturing all size classes of monitor above yearlings, which are largely arboreal. This trapping technique requires a 300 cm x 50 cm x 50 cm long box traps baited with goat meat (≈ 0.5 kg). Distance between traps is recommended between 200 m and 700 m from each other, depending on topographical and vegetation. Traps are positioned in shaded areas in order to avoid overheating of trapped individuals and are checked twice daily.

Following capture, Komodo dragons are restrained with rope and their mouths taped shut. Several morphological characters, including head length, and snout to vent length (SVL) are measured using calipers and a fiberglass tape. Body mass is obtained using digital scales. Komodo dragons are permanently identified using passive integrated transponders (i.e. PIT tags- Trovan ID100) inserted in to their left hind leg.

Mark recapture data then are entered into a central data base (Excel) and then transferred to demographic programs including MARK which will enable estimation of key demographic processes including population growth.

Radio telemetry

Radio telemetry is applied to assess a fine scale spatial ecology, especially to determine movement pattern and home range of Komodo dragon. This method requires transmitters, receivers, and antennas to be able to locate monitored animals. Transmitters are attached to selected animals either by harness, duct taped, or glued to the base of tail as it is considered the best site for placement. Attaching transmitters on juveniles by means of a harness was not feasible, thus use of duct tape or glue are consider the most feasible. Individuals are located either by direct observation or triangulation techniques (White & Garrot 1990). During tracking, certain parameters for habitat type or tree visited by animals recorded as follows; habitat type, tree species, breast height diameter, and tree height. Data will be calculated by means of a computer program ESRI ArcView 3.2 (ESRI 1999) with X-Tolls and Animal Movement program (Hooge et al. 2003).

Nest surveys by implementing transect grids

Field method implemented to inventory Komodo dragon nesting sites is consisted of intensive focal sampling across consecutive transect grids. This method involved multiple observers (5–8) walking at intervals of approximately 25 m apart along a series of parallel transects marked with projected GPS way points. The length and number of transects in each valley was defined by the prevailing topography of the valley. The purpose of these comprehensive transects was to identify and mark (with GPS point) all potential Komodo nesting sites and all megapode nests within each valley up to an elevation of 100 m above sea level. Once all nest are located, following annual monitoring are not require another intensive focal sampling, enough by checking the status of all marked nests and identify whether it is active or not.

Komodo dragons are known to use three types of nest and categorized as follows:

1) Ground nests- consisting of deep sloping horizontal burrows constructed in the ground.

2) Hill nests- typically consisting of large excavations resulting in one or more tiered platforms across the face of the hill. Into these excavations females would dig an egg chamber alongside a number of decoy chambers. These nests are situated in open savanna grassland which covers most low hillsides.

3) Mound nests- Komodo dragons utilized mound nests constructed by orange footed scrub fowl. Active Scrub-fowl mound nests are distinguished from active Komodo mound nests chiefly by the amount of debris and recent diggings that had occurred, particularly during August and September. This is fairly easy to determine as Orange-footed Scrub-fowl nest earlier in the year, with eggs recorded from January until April (Lincoln, 1974), plus megapode nests tend to incorporate vegetative debris into the mound and the chambers into which the birds oviposit (Frith, 1956; Jones et al., 1995).

Komodo dragon nests are identified by the presence of large chambers up to 2 meters long sloping into a nest. These nesting chambers are distinguished from resting chambers (Auffenberg, 1981) by the presence of multiple decoy chambers. Komodo dragon nests are confirmed active by the presence of recent digging activity by females (beginning in August) or by repeated observations of the female in association with the nest (August through November). Inactive Komodo dragon nests are confirmed by the absence of recent digging activity or female guarding the nest throughout the nesting season. These inactive nests are known to be used by Komodo dragons due to observations by park rangers (prior to the current season) of female digging and nest attendance activities or due to changes in structural characteristics, particularly the size and number of chambers in the nest. The density of active and inactive Komodo dragon nests is analyzed by dividing total nest number for each category by the area searched as calculated by shape polygons using Arc view 3.1 (ESRI). As an index of nest dispersion, the mean nearest neighbor measurement was calculated between valleys as the average distance to the closest neighbor from each active nest in a survey location.

Komodo dragon active nests monitoring should be undertook during early nesting season (August-September) each year. Monitored nests on Komodo and Rinca can refer to Jessop (2007) data.

Nest fencing

Komodo dragon nests are confirmed active by the presence of recent digging activities by females (beginning in August) or by repeated observations of the female in association with the nest (August through November). Active Komodo nest will be guarded by associated female that laid her eggs in the nest for about 3-4 months (August-December). Once the female that guarding active nest is leaving, by late December, nest need to be fenced by 1,5 meter metal-sheeting plate. This fence is constructed to avoid emerged hatchling escapes before counted and measure, also to give protection to the hatchlings from being attacked by predators. Once nest-fence established, the nest should be checked twice daily. When hatchlings emerge from nest, all hatchlings are needed to capture, measure, and permanently marked. After all emerged hatchlings are released from the nest, to assess fecundity nest should be dug, to find and count the number of eggs and compared to number hatchlings that emerged.

Genetic and health studies

For further study, sex ratio and age specific survivorship of Komodo dragon by mean of genetic analysis and long term mark recapture method is highly recommended. Blood samples (300 µl) are collected from the caudal vein of new individuals (please refer to previous work of Jessop et al 2002-2006 of CRES ZSSD to identify marked and unmarked animals), using a 21 g needle and 3ml syringe, to enable further genetic sexing analysis of dragons. These blood samples should be stored and transported by staff of Balai Komodo National Park prior to further analysis at Genetic lab in Indonesia, i.e. LIPI.

Individual and population health of Komodo dragons is an important point that also necessary to monitor. Parasitological condition should be monitored by mean of fecal analysis and direct observation for thick that exists on the Komodo skin.

2. Prey density and diversity among four islands within Komodo National Park.

Assessment of density in large prey

Three species of large ungulate prey including the Timor deer (Cervus timorensis florensis), Wild pig (Sus scrofa) and Water buffalo (Babulus bubalus) are monitored by implementing indirect survey techniques (reviewed in Thompson et al. 1998) based on faecal counts: estimates from these techniques should be less influenced by the tendencies of prey to avoid people or be missed in forest. Counts of the standing crop of ungulate pellets or faecal pellet groups have been widely used to estimate the relative or absolute abundance of many ungulate species (Bennett et al., 1940; White, 1992; Thompson et al., 1998).

An indirect index of prey density is calculated using pellet counts on linear transects. Within each site between 20 and 49 permanent linier transects were randomly positioned and orientated (refer to Jessop 2007). Pellet groups are tallied from 30 sample plots placed across each 150 meter long transect. Each plot is a circle with a radius of 1 m and encompasses an area of 3.14 m². All deer pellet groups within the plot were recorded. A group is standardized as a dense aggregation of pellets exceeding 40 pellets; groups below 40 are counted as individuals then divided by the mean pellet count (taken from counting 60 intact pellet groups). Pellet groups that are greater than 50% inside the plot area are counted as an entire group. To standardize seasonal differences, in pellet density it is important to conduct all pellet surveys across the 10 sites in late September and early October of 2006.

To calculate means bootstrapping technique is necessary to operate (Manly 1997), 95% confidence intervals (‘CI’) and CVs for the plot-based estimates of faeces abundance (per ha) for each large prey species at each site. Bootstrap estimates were based on 10 000 samples. The CV was:

Assessment of density in small prey

Rat (Rattus ratus)

Rats are captured by operating Elliot traps spaced at 10 meter intervals along randomly positioned trap-lines at each study site. Trap-lines placed with at least 200 m apart to reduce the possibility of animals being sampled by more than one trap-line. Trapped rats are individually identified, measured and released at the point of capture. Newly captured animals were given a unique mark by ear tagging. On their first capture during a trapping session animals are weighed and sexed. The head and body length is taken as being from the tip of the nose to the middle of cloaca, the tail length from the middle of cloaca to the tip of the tail. Tails with a terminal scar were assumed to be shortened and were excluded from measurement.

To assess differences in prey density among the five islands the plot counts, distance and mean number of rats per trap night should be undertook. The four sites on both Komodo and Rinca islands are pooled and used to infer a total island sample. Comparison of island means for each of the five species are analyzed by parametric and non-parametric analysis of variance depending on data meeting the assumptions of normality and homogeneity of sample variance. To discriminate significant differences among islands appropriate post- hoc methods (Tukey’s test and Dunn’s method) was used to identify subgroups.

Tokay Gecko (Gekko gecko)

Gekko gecko are monitored by using line transect technique. During the day prior to each survey mark out transect lines with fishing line marked for every five meters with flagging tape. All gecko surveys begin after dark, using powerful head-mounted 6V spotlights conducted by three people. Geckos counted by slowly walking along the transect line, searching every tree, shrub and vine mat within sight, on both sides of, and directly above the string. As often as possible walk off the transect line a few meters on either side; to give a wider range of viewing angles and the ability to more carefully search within trees directly above the transect line. When a gecko is sighted, measure the perpendicular distance at ground level directly beneath the gecko to the transect line (to within 0.1 m), and estimate the geckos height above ground to the nearest meter.

To estimate the density of geckos, use conventional Distance sampling method analysis. In this method the number of geckos located within the survey area are modelled as a function of perpendicular distance of the detected lizard from the line (Buckland et al. 2001). Data analysed using the program DISTANCE 4.2 release 1 (Buckland et al. 2001). DISTANCE is freeware available at http://www.ruwpa.st-and.ac.uk/distance/, and is widely used for the analysis of line transect data.

3. Assessment of habitat and other terrestrial key with focus on endangered species

Vegetation monitoring

Vegetations are monitored by plot method; permanent plots are placed in each represented habitat composition on each of 10 study sites. A 20×20 m permanent plot is established to estimate tree density with number of plots repetition depends on habitat size. Seedlings and saplings should be estimate by established subplots. Grass as the main food for herbivore animal is also monitored by using similar plot method. Density of grass is estimated by measuring 10 1×1 meter permanent plots in each of 10 study sites. Morphological characters like tree and grass species, DBH, height and canopy cover are measured by plastic measuring tape.

Active nest of Orange-footed Scrub-fowl (Megapodius reindwardt)

Orange-footed Scrub-fowl build conspicuous incubation mounds (Jones et al. 1995, Palmer et al. 2000). For each mound located, we recorded the location, elevation, status (active or inactive), overhead vegetation cover (0-25, 26-50, 51-75, or 76-100%), adjacent vegetation type (open forest, closed forest, savanna, or grasslands), and soil type (loamy, sandy, rocky, or gravelly). Inventoring incubation mounds is implemented by intensive focal samplings across consecutive transect grids with multiple observers (5 – 8) walking at 25-m intervals along parallel transects. The length and number of transects in each valley were determined by topography. Following annual monitoring will not require another intensive focal sampling, enough by check the status of all marked nest and identify whether its active or not. Structural characters of each mound are also need to recorded, including length, width, height, number of chambers excavated in each mound, and adjacent habitat type.

Active scrub-fowl mounds are those used for breeding during breeding season, and are distinguished from inactive mounds by evidence of recent digging, incorporation of new leaf litter, and, in some instances, the presence of adults at a nest or the presence of their tracks. Inactive mounds are those not being used in that breeding season and ranged from mounds with egg chambers containing old leaf litter to flattened mounds with no evidence of activity and covered in grass. The density of active Orange-footed Scrubfowl nests is calculated by dividing total nest number for each category by the area searched as calculated by shape polygons using Arcview 3.1 (ESRI). As an index of nest dispersion, the mean nearest neighbor measurement is calculated within valleys as the average distance to the closest neighbor from each nest in a survey location.

Due to the difficulty of measuring egg predation directly, we can use an index of predation based on the presence of fresh excavations into the egg chambers of active scrubfowl nests. Predators are identified by their tracks and associated burrowing as either Komodo dragons or wild pigs (Sus scrofa). Excavation by predators is likely to be repaired by scrubfowl, so observed excavations are likely made in the week preceding the survey.

Active nest and direct population counting by vantage count for Yellow-crested Cockatoo (Cacatua sulphurea)

Active nests survey

Nest survey is carried out by systematic searches (Mexquida, 2004) across consecutive transect grids, in which multiple observers (3-5 persons) walked at ≈ 25 meter intervals along parallel transects. The length and number of transects in each valley is defined by the prevailing topography of the valley. Nest searching is carried out across the valleys and including hills up to 60 meters elevation. Active nests are indicated by the present of young(s) in the nest and parents guarding the nesting location.

Once an active nest was located, data should be taken to record characteristics including location (GPS position), elevation, adjacent vegetation type (Open forest, closed forest, savanna), and nesting tree species. Structural characters of each nest are also recorded including tree DBH, tree height, and nest height. Nest locations are marked by means of GPS Garmin Etrex Vista (Garmin). Tree and nest height are measured by means of Suunto Clinometer (Suunto, Finland). To avoid disturbance to the occupants of the nests, nest parameters, i.e width, length, and depth, did not measured. To analyze the spatial distribution pattern, the nests were mapped and nearest neighbor-distances were calculated using the computer program of ArcView 3.1 (ESRI).

Population and Density Estimates

The Yellow-crested Cockatoo population estimate using direct counting of vantage point method in each valley (Bibby et al., 1992). Direct counting of vantage points method is carried out from hills, which provides observers a well suit observation points to observe the whole valleys and feasible to count all individuals sighted. To assess the density of this species within each valley, divide the highest number of the birds counted by the size of the valley. Sizes of the valleys are calculated by creating polygons, based on GPS points that collected during the field study, and covered the entire studied valley area on the map using the computer program of ArcView 3.2 (ESRI).

Monitoring diversity of birds

Inventory and monitoring of avifaunal diversity, can implement intensive focal sampling across permanent line transect. The length of each transect is 1 km, and number of transects in each valley are determined by topography and size of valley. Observer walking slowly across transect line, birds identified base on field guide. Duration of observation, number of observer must be recorded. Observation should be done minimum twice, early in the morning and afternoon (when the birds are most active).

Assessment the presence of exotic/invasive species

Field methods used to inventory presence of exotic/invasive species consisted of intensive focal sampling across consecutive transects grids. This method involved multiple observers (5–8) walking at intervals of approximately 25 m apart along a series of parallel transects marked with projected GPS way points, and then record all the exotic/invasive species (e.g. cactus, dogs, cats). The length and number of transects in each valley was defined by the prevailing topography of the valley. The purpose of these comprehensive transects was to try and identify all presence of exotic/invasive species within each valley. Once exotic species identified and located, further monitoring should be undertook on the same location. Further, elimination efforts should be consider preventing disturbances to the native wildlife and habitat in the Komodo National Park.