An island-wide status of Sumatran tiger (Panthera tigris sumatrae) and principal prey in Sumatra, Indonesia
University of Delaware
A multilateral effort to establish priorities for global tiger conservation identified 76 Tiger Conservation Landscapes throughout the 13 tiger range countries, 12 are in Sumatra, Indonesia. Despite this designation of conservation landscapes, the status of the Sumatran tiger population is still in question, as existing range-wide density estimates were derived using findings from disparate approaches. Meanwhile, habitat fragmentation and loss continue to threaten the integrity of these landscapes, while demand for tiger body parts, prey depletion, and human–tiger conflict have been documented as causing a rapid decline in tiger numbers. In Chapter 1, we jointly analyze 29 camera trap datasets from 16 sites collected between 1999 and 2017 in a single multi-session model to allow estimating parameters across sites and sessions and tested a variety of models with different covariates for movement parameter (σ), detection probability (g0), and the density (D). We found that Sumatran tiger densities were significantly higher in lowland habitat and under protected status, that adult male tigers moved significantly further than adult females, and that the Sumatran tiger ranged over larger areas in montane habitats. In Chapter 2, we analyzed an animal sign-based detection/non-detection dataset collected along transects in 389 grid cells,17 by 17 km each, between 2007 and 2009, in 60% of the remaining tiger landscapes in Sumatra, Indonesia. We explored the effect of environmental and anthropogenic factors on the occupancy of Sumatran tiger, their main prey, and poaching and logging, using a multi-species occupancy model. We found that the occupancy of Sumatran tiger, sambar deer, and barking deer were higher in grid cells with higher percent of forest cover. Tiger and wild pig preferred lower elevations while barking deer preferred higher elevations. We found positive correlations between the predicted occupancy of Sumatran tiger and that of the sambar deer and barking deer. National parks tended to have a positive effect on tiger presence, although poaching and illegal logging were still widespread throughout the surveyed areas, often occurring in conjunction to each other. In Chapter 3, we modelled landscape connectivity in a human–dominated landscape in West Sumatra Province, Sumatra, Indonesia. We first applied maximum entropy modelling using program Maxent v3.4.1 to identify nine core tiger habitat areas in West Sumatra Province and then developed a resistance layer, i.e., landscape friction for the Sumatran tiger to traverse through, based on roads, elevation, slope, and land cover types. We next applied circuit theory and least-cost path analysis to predict structural connectivity between Sumatran tiger core habitat areas based on our landscape resistance layer. We identified a total of 13 linkages (1,978 km2) connecting nine core areas in West Sumatra Province; mostly skewed to gentler slopes and at higher elevation. The linkages were predominantly secondary forests, shrub, and agriculture, indicating their importance as tiger’s structural corridors in a human-dominated landscape. We recommend: a) the application of the multi-session models with environmental covariates on the existing Sumatran tiger density datasets that are not yet incorporated into our study and for future island-wide tiger population assessment, b) to use the occupancy parameters we produced as the first data points against which past conservation interventions should be evaluated, and c) to implement a hands-on training program on habitat connectivity assessment for members of relevant government bodies so that connectivity analysis can be done for the rest of Sumatran tiger landscapes.
Barking deer, Circuit scape, Habitat connectivity, Occupancy, Sambar deer, Wild pigs