The potential impact of climate change on the distribution pattern of Eusideroxylon zwageri (Bornean Ironwood) in Kalimantan, Indonesia
Abstract. Yudaputra A, Fijridiyanto I, Cropper WPJr. 2020. The potential impact of climate change on the distribution pattern of Eusideroxylon zwageri (Bornean Ironwood) in Kalimantan, Indonesia. Biodiversitas 21: 326-333. Eusideroxylon zwageri Teijsm & Binn. is a vulnerable tree species with co...
Guardado en:
| Autores principales: | , , |
|---|---|
| Formato: | article |
| Lenguaje: | EN |
| Publicado: |
MBI & UNS Solo
2020
|
| Materias: | |
| Acceso en línea: | https://doaj.org/article/62d7864730394955bc89c022e72b984c |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| Sumario: | Abstract. Yudaputra A, Fijridiyanto I, Cropper WPJr. 2020. The potential impact of climate change on the distribution pattern of Eusideroxylon zwageri (Bornean Ironwood) in Kalimantan, Indonesia. Biodiversitas 21: 326-333. Eusideroxylon zwageri Teijsm & Binn. is a vulnerable tree species with considerable economic value. The high demand for its wood makes this species vulnerable. Population vulnerabilities of E. zwageri include habitat loss, land-use change, and forest encroachment. An additional potential risk factor is climate change, with a possible increase in temperature of about 2.5 to 10 degrees Fahrenheit over the next century. Climate is considered to be a principal factor that determines the distribution of many species. This study addresses the potential current and future distribution of E. zwageri under climate change. Seven predictor climate variables are selected from 19 climatic variables using VIF (Variance Inflation Factor) to eliminate multicollinearity among variables. The spatial data is prepared using Geographic Information System (GIS). Six species distribution models (RF, SVM, MARS, GAM, GLM) and the ensemble model are applied to understand the potential current geographical distribution of E. zwageri. For risk assessment, the potential future distribution is predicted using the ensemble model only. All models are run using R open-source software. In model evaluation, all models have AUC value >0.80, indicating those models are good predictive models. All predictive models have the TSS >0.60 which means those models having a useful agreement between prediction and real observation. Precipitation seasonality, isothermality, and precipitation of the coldest quarter are the most important model variables that influence the current and future distribution of E. zwageri. Four models (RF, SVM, GAM, GLM) produce similar predictive maps of potential current distribution. MARS produces a slightly different predictive map. The future projection of ensemble model shows that the distribution area is more likely shifted and decreased from the current to 2050 and 2070. |
|---|