Author Title Institute Sheikh Muhammad Abdur Rashid Population ecology and management of Water Monitors, Varanus salvator (Laurenti 1768) at Sungei Buloh Wetland Reserve, Singapore Thesis (Ph.D.) National Institute of Education, Nanyang Technological University Year 2004 Supervisor Diong, Cheong Hoong
SUMMARY The 87-hectare Sungei Buloh Wetland Reserve is Singapore's (N 1'42' 53.5"; E 103'43' 30.8") only wetland reserve. The Reserve is a mangrove-dominated coastal wetland that is internationally recognised for its rich biodiversity and its importance as a feeding ground and staging post for migratory birds that use the East-Asian Australian Flyway. The water monitor Varanus salvator inhabits this Reserve. Using GIS and aerial photographs, the habitats in the Reserve were mapped and quantified for the first time. Seven habitat types were identified and described for this study on the ecology of the V. salvator population at the Reserve. These are: mangroves, secondv vegetation, grassland, freshwater ponds, mudflats and back swamps, brackish ponds, and dykes (raised embankments). In a mark-recapture study, 181 monitor lizards were captured by live-trapping and marked with PIT-tags to estimate density and to study lizard movements. The total number of captures and recaptures (ranging from 1-8 captures or recaptures) totalled 301. The population is essentially closed and the monitor density was estimated with a closedsystem model [M(h), M(bh), M(th)] to be 3.9 + 0.25 lizards per hectare. Lizards were sexed successfully by X-radiography after an infusion of a radio-opaque marker, barium sulphate, into their cloacal sacs. The sex ratio of all lizards in the sample approximated the 1: 1 male to female ratio, but the sex ratio was skewed towards females in small-sized (subadult) lizards which formed 31% of the population, and males, in adult-sized lizards
which comprised 69% of the population. Principal Components Analysis showed body size difference between sexes, with the MANOVA for PC1 (=size) showing significant difference between sexes (F = 78.16, p < 0.05) and the shape components (PC2 - PC22) showing significant difference in shape between sexes (F = 3.001, p < 0.05). Of the 21 morphological characters described in V. salvator in this population, a set of four characters, viz., Snout-Vent Length, Tail Ixngth, Total Length and Hip Girth were significantly larger in males than females. A total of six monitors were radiotracked using the 172 MHz frequency for 195 k 47 days. Home range estimates (for all radiolocation fixes) averaged 0.209 + 0.09 ha by the minimum convex polygon method, 0.181 :k 0.07 ha by the Jenrich-Turner bivariate elliptical model, and 0.133 k 0.03 ha by the modified Kernel method. The home ranges overlapped considerably. Monitors used trees and burrows as activity centres in their home ranges. Activity centres varied from 1 to 3, the size of an activity centre ranged from 7 to 36 m2. One-way ANOVA on arc-sine transformed data showed that the proportion of time monitors spent in each of the seven habitat types and burrows was significantly different (p < 0.05); the proportion of time monitors spent in mangroves and freshwater ponds, as shown by Tukey's test, was significantly greater than the time lizards spent in tidal mudflats, secondary vegetation, brackish ponds, grasslands, and dykes. Given similar environmental conditions, water monitors spent more time in the freshwater ponds than in mangroves; brackish water ponds were the least selected of the habitats. Monitors do not travel every day, the daily movement rate varied widely among individual monitors, averaging 0.72 5 1.2 day over 504 lizard days. The daily distance
travelled averaged 325.2 5 327 m (0-1927.2 m), with a mean daily travel distance ranging from 236.1 to 409.8 m. There was no significant difference in the daily distances travelled between the sexes (p > 0.05) or between size-classes (p > 0.05). Water monitors were late-risers, emerging from their night shelters usually after 10:OO hrs when the air temperature was already higher than their body temperature. Four major activity types were classified; basking, resting, foraging and movement, with a unimodal activity-pattern in the afternoons; this activity peak varied with environmental conditions. The daily amount of time the monitors spent in each of these different activities was significantly different (F = 457.14, p i 0.05). On average, monitors spent 60.2 f 11.2% of their daytime "resting" to thermoregulate, 13.8 f 10.8% on '%asking", 21.8 f 12.2% of the time on "foraging" activity, and 2.2 f 1.1% of the time for non-foraging related "movements". Mangroves were the most utilised habitat by monitors, and it was used mostly for "resting". The lizard body temperatures determined from implanted transmitters varied significantly among the habitats (F = 26.4, p i 0.05). Tukey's test for multiple comparisons showed the lowest body temperature of the monitors, 27.8 f 1.6" C, was obtained when lizards were in their burrows and the highest temperature, 31.4 f 2.7" C, when lizards were on dykes. Body temperatures also varied significantly among activities (F-value = 33.27, p i 0.05); the body temperature was lowest during "resting" (28.7 5 2.9" C; 22-31' C), which supported the thermoregulation hypothesis, and highest when lizards were
"foraging" (29.9 * 2.4; 23-37' C). The mean daily body temperature of the lizards averaged 29.2? 2.8' C (22-37O C); the lizard body temperatures remained hlgher than the ambient during the early mornings and late afternoons. Lizard activities were significantly dependent on the time of the day (X2 = 51.261, df = 9, p < 0.05). No significant difference was found in the daily distances travelled by lizards of either sex (F-value = 0.37,~ > 0.05) or between size-class (F-value = 0.51, p > 0.05). The diets of the Varanus salvator population at the Reserve comprised 78 prey taxa which were identified from regurgitated stomach contents and direct field observations on lizard foraging activities. Dietary composition showed that crustaceans, mostly Episesaima versicolor, and E. singaporeme (Family Grapsidae) were the most frequently consumed prey item (42%), followed by fish (21%), insects (10%), mammals (9.2%), reptiles (7.5%), birds (5.5%), amphibians (2.4%), and other miscellaneous items (2.4%). Crabs and fish formed the major volume of the food contents (> 60%) while the remaining < 40% of the diet comprised insects (including larvae), reptiles (including snake eggs), birds and mammals. The dietary groups, Arthropoda and Pisces ranked first and second in the Importance Value of all diet categories. No difference was found in the dietary composition between sexes, but large-sized monitors fed more frequently on larger-sized prey items. A MANOVA of the major factors (sex, size class) showed significant difference among the prey volume in the three monitor size-class (p < 0.05). Several foraging behaviours are described. A schematic construction of the dietary items of V. salvator in this population showed that the monitor lizard is a top predator in the mangrove ecosystem in the Wetland Reserve.
Field data suggest that the V. salvator population breeds throughout the year and the breeding activity was probably influenced by rainfall. Female monitors were mostly gravid or engaged in reproductive activities during the wet season, from October to March, with some continuing to breed at other times of the year. Local environmental conditions or flooding also influenced lizard breeding activities. The smallest gravid female captured, measured 530 mm SVL. Female egg nests were significantly more abundantly distributed in sandy soil than in sandy-clayey soils (X2 = 8.62, df = 2, p < 0.05), but there was no difference in the distribution of nests on the top surface and the vertical edges of dyke. Nest depth averaged 27 + 7 cm (n = 28). Mean clutch size averaged 13.1 + 4.1 eggs (n = 14), with clutch size ranging fiom 5 to 25 eggs; relative clutch mass was 15.1 + 2.7% (n = 3). Clutch size was correlated with female body size (p < 0.05, R' = 70%). The average length and width of freshly laid eggs were 68.6 f 4.1 X 38.8 + 2.1 mm; egg mass averaged 38.8 + 2.lg (n = 29). The average incubation period was 214.9 f 6.9 days (n = 7). Newly hatched neonates averaged 135 f 5.62 mm SVL; 35.32 k 3.25 g in body weight (n = 15). The study showed that Varanus salvator plays an important ewlogical role in the coastal mangrove ecosystem at the Sungei Buloh Wetland Reserve. The population of water monitor is not a threat to the resident and migratory avian communities. The monitors, together with the birds and other wildlife have an ecological niche of its own, and the
monitor population in particular, seemed to have adapted well to living in a wetland habitat with high human visitor traffic. Dogs were a major source of monitor lizard mortality in the population. Several resource management plans were discussed in the final chapter of this thesis.