Prediction of The Spread of Acacia nilotica Invasive Species Using Generalized Linear Models

Sahid Agustian Hudjimartsu (Department of Computer Science, Faculty of Math and Natural Sciences, Bogor Agricultural University, Dramaga Campus, Bogor, Indonesia 16680)
Yeni Herdiyeni (Department of Computer Science, Faculty of Math and Natural Sciences, Bogor Agricultural University, Dramaga Campus, Bogor, Indonesia 16680)
Lilik Budi Prasetyo (Department of Forest Conservation and Ecotourism, Faculty of Forestry, Bogor Agricultural University, Dramaga Campus, Bogor, Indonesia, 16680)
Agung Siswoyo (Bromo Tengger Semeru National Park, The Ministry of Environment and Forestry, East Java, Indonesia)

MANAJEMEN HUTAN TROPIKA Journal of Tropical Forest Management
Vol. 23 No. 3 (2017)

Acacia nilotica is an invasive species in the Baluran National Park. The spread of A. nilotica is a serious threat that can disrupt the function of savanna ecosystems in providing food for herbivores and eventually led to a decrease in wildlife populations of herbivorous. The growth of A. nilotica invasive species is influenced by some environmental factors. In order to prevent the spread of A. nilotica, a model can be formed to prodict the species distribution based on the environmental factors namely generalized linear models. This research used the data from Siswoyo’s research in 2014, which consisted of environmental data and distribution of A. nilotica. The results showed that there were five environmental variables that significantly affect on the spread of A. nilotica invasive species. The variables included elevation, surface temperature, normalized difference vegetation index, normalized difference moisture index, and the distance from the river, with p-value less than 0.05. The accuracy of the prediction model on the spread of A. nilotica using GLM was 97.22% and the Area Under Curve (AUC) value was 0.981.