Factors Affecting Habitat Selection and Feeding Habits of Smooth Coated Otter (Lutra perspicillata) in Periyar Tiger Reserve, Kerala

Abstract

Otters are the top predators of the aquatic ecosystems. Despite their important place in the food web and wide distribution, the knowledge on their ecology is relatively scanty as compared to other carnivores. sediments. nutrients, energy and biota, dams interrupt most of the ecological processes of the river. This includes changes in the fish communities because of changes in water current and depth. The aim of this study was to look at how otters adapted themselves for living in a man modified aquatic system. The factors affecting habitat selection and feeding habits of smooth-coated otter was studied in Periyar Tiger Reserve, Kerala during November 2000 to April 2001. The importance of different habitat parameters influencing habitat selection was assessed using ordination technique. The food habit was examined by analyzing the spraints (scats) using both frequency of occurrence and score-bulk estimate methods. The habitat of the animal was visualized as a multidimensional space in which the animal selected a few combinations of parameters./Foraging sites with high densities of fish, areas that offer den sites. The signs that the animal leaves on the banks while using an area were identified as -holts, communal spraint sites, grooming sites and footprints. Presence of these signs were considered as sure indicators of intensive use of an area by the otters. Subsequently, 96 random plots of 250 x 10m (2.5 km2 area) were placed along the banks 'of the reservoir and each plot was characterized by means of 17 habitat variables. Of these variables, three viz. water depth, distance to vegetation cover (escape cover) and numbers of streams joining the lake were outside of the plot. Each of such plots was accordingly assigned to the category of either used or unused by the animal based on the presence or absence of evidences in the plots.The complex data collected from the plots was reduced to three uncorrelated axes using Principal Components Analysis sacrificing 300/0 of the variability of the system the plots had a position, which characterized them in terms of variety and variability. Then each of the plots was marked as either used or unused based on the evidences found in the plot. It was found in the newly created three-dimensional space, the used and unused areas were segregated in distinct regions. The coordinates of the used plots indicated that the otters selected less rocky areas of the shallower parts of lake, which are narrower and slanting. The number of streams joining these areas was more than other stretches of the lake and the vegetation on the banks was dense. The otters did not use the deeper and wider areas of the lake. Other unused areas were steep and rocky parts of the reservoir where no streams joined. The higher fish congregation in the areas of mouths of streams must be the reason why otters' used those areas more in comparison to other areas. The low water depth made the dives of otters in these areas more effective while foraging. The thick vegetation cover of the banks provided them shelter and narrower areas helped them in patch fishing i.e. reducing the chance of fish to escape by forming an arch around the shoals of fish. Mann - Whitney 'U' Test, confirmed this observed pattern.Then a logistic regression model was developed using animal presence I absence as the dependent variable and the habitat variables as covariates. The model selected only rockiness and number of streams from the complex set of variables to predict the probabilities of otter presence in each of the 94 plots chosen for characterizing the otter habitats. Out of the 70 plots, which showed positive signs of otters, 95.71 % were classified in the region of higher probability of occurrence. But out of the 24 plots, which showed no signs of otter occurrence, only 54.17% were correctly classified in the region of lower probability of occurrence. This shows that some of the areas, which showed no signs of otter presence, were potential areas but at the time of this study they were not using these areas for unknown reasons. This may also be due to the inherent problems of using spraints and tracks to assess the degree of use of an area by otters. Feeding habits of otter were studied by analyzing 94 spraints collected from around the Periyar lake. Prey remains were identified using standard references prepared with identified body parts for 12 fish species that were collected during the study period. Two methods were used to estimate the proportions of different prey items consumed viz. frequency of occurrence method and score-bulk estimate method. The former method takes into account the presence or absence of prey items with no consideration of its quantity whereas the later includes the quantity of each of the prey item consumed, by including proportion of each prey item in the spraint and total weight of the spraint. It was observed that in the frequency of occurrence method the major prey items were under estimated and minor items were over estimated, as it did not take into consideration, the proportion of each item individual The number of prey items in a spraint varied between 1 and 7 in a single scat, the average in lower water levels in the reservoir (December, January and February) being 3.21 (varied between 1 and 5) and during higher water levels (March and April) it was 2.73 (varied between 1 and 7). Fish was the major prey item (96.02%), followed by frogs - --(l.08%), crabs (l.07%), birds (1.07%) and insects (0.76%). During the study period. remains of six fish species were found in the spraints.Among these, Tilapia (Oreochromis mossambicus) was the major prey of otter (51.54% of the total diet) followed by catfish (Heteropneustes tossilis), which formed 21.27 % of the diet, curmuca barb Gonoproktopetrus curmuca; 12.37%) and European carp (Cyprinus caprio; 9.16%). A linear relationship was stablished between the total length of the fish and the length of the vertebrae. For this purpose a total of 43 fish belonging to different size classes were used and 215 measurements of vertebrae were taken. From the scat of otters, vertebrae were sorted out and lengths of the corresponding fish were found out. Otters seemed to be consistently taking more fish of size 8.0 cm to 15.0 cm (29% each), followed by 5.0 cm to 8.0 cm class (23.70/0) and 14.0 to 17.0 cm class (14.4%). Attempt to quantify fish availability in the lake to determine the preference failed because of the hindrance caused by submerged tree stumps that damaged the fishing nets. From this study it was concluded that the otters in Periyar lake have adapted to a man modified condition utilizing most common exotic fish species of the lake and selecting those stretches of the lake where the physical constraints of diving did not interfere much with their foraging success. They were also tolerant to a great deal of anthropogenic disturbances such as presence of motorboats and humans. This shows the remarkable adaptability of the species. Furthermore 60 % of the diet was found to be comprised of exotic fish. Thus the changes in the natural habitat both in terms of food and shelter has been successfully compensated by adopting alternate means for survival. As the entire catchment of Mullayar and Periyar are well inside the park boundaries, the pollution and anthropogenic disturbances are minimum in the lake. This may be the reason why Periyar is still supporting a healthy population of otters even though they are getting exterminated from many of its former distribution ranges. So the otter population Periyar Tiger Reserve should be considered as a precious possession of the park.