Discrimination of dominant forest types for Matschie's tree kangaroo conservation in Papua New Guinea using high‐resolution remote sensing data

Abstract

Matschie's tree kangaroos (Dendrolagus matschiei) are arboreal marsupials endemic to the Huon Peninsula in Papua New Guinea (PNG). Primarily because of an increase in hunting pressure and loss of habitat from agricultural expansion, D. matschiei is currently listed as endangered by the International Union for the Conservation of Nature. This paper reports the results of our study to compare the capabilities of Landsat‐7 Enhanced Thematic Mapper Plus (ETM+) and Satellite Pour l'Observation de la Terre (SPOT)‐4 multispectral image data at discriminating dominant forest types at a remote research location in PNG. Nearest‐neighbour vegetation plots were established from July to August 2004 to obtain detailed information about the vegetative communities and guide class assignments. Forests were separated into four distinct habitat types with Dacrydium nidulum dominant forests being the most widespread and the most accurately classified. The comparative results indicated that Landsat‐7 and Spot‐4 had similar classification accuracies but the results were low because of the complex structure and heterogeneity of the forest communities and the limited spatial/spectral resolutions of the satellite data sources. This research provides an improved result compared to past research and provides detailed information towards the future conservation of Matschie's tree kangaroo habitat in PNG.

Acknowledgements

This research was supported by Conservation International and The Conservation, Food, and Health Foundation, and was carried out in the Department of Natural Resources Science at the University of Rhode Island. Many thanks go to Phil Shearman, Director of the Remote Sensing Centre at the University of PNG, and Leonardo Salas, Animal Population Biologist at the Wildlife Conservation Society. We also thank Kaigube (Dick) Fazang and Simon Sennart from the Lae Herbarium for their invaluable assistance in the field, without whose help in collecting and identifying the flora, this project would not have been possible. Thanks also to the Director and staff of the Lae and Queensland Herbariums, who tirelessly dealt with the mountain of specimens and helped in refining the final lists and authors. And last but not least, we thank Mae and Bege, the local landowners/experts in all aspects of flora and fauna, who facilitated our stay and tirelessly worked and dragged us around the mountain to do our research. For access to original and/or processed data, contact the primary author (Jared Stabach).