| 要旨トップ | 目次 | 日本生態学会第68回全国大会 (2021年3月、岡山) 講演要旨
ESJ68 Abstract


一般講演(口頭発表) F03-02  (Oral presentation)

Species distribution mapping of mangrove forests using satellite remote sensing【EPAA】

*Mohammad shamim hasan MANDAL(Hiroshima Univ.), Md. KAMRUZZAMAN(Khulna Univ.), Tetsuro HOSAKA(Hiroshima Univ.)

Mangrove ecosystems are among the most productive ecosystems. These magnificent ecosystems are under the threat of climate change, sea-level rise, and frequent cyclones. Mangrove species distribution is shaped by many different bioclimatic factors. In this study, we performed MaxEnt species distribution modeling of Heritiera fomes in the Sundarbans mangrove forest. We used all the Worldclim bioclimatic variables, along with soil salinity, elevation, wind speed and cumulative cyclone impact frequency to identify important bioclimatic factors for species distribution. Out of 27 variables, only 11 variables were used for the final model based on variance inflation factor (VIF). Percent contribution, permutation importance and jacknife test score was used to identify important environmental predictors. Elevation, precipitation in the driest month, precipitation seasonality, soil salinity and annual mean temperature were the most important factors and together contributed about 90% on the distribution model. The final model AUC was 0.84 indicating good model fit. Cyclone frequency had a very low (4%) contribution, which indicates subtle influence towards distribution of H. fomes in the Sundarbans. The model developed here can be used to predict and measure species distribution under future climate change scenarios.


日本生態学会