Prioritization of conservation areas using species distribution modeling based on current data of butterflies in Johor

• Main Author: Awg Abdul Rahman, Aqilah
• Format: Thesis
• Language: English; English; English
• Published: 2019
• Subjects: QL Zoology
• Online Access: http://eprints.uthm.edu.my/53/1/24p%20AQILAH%20AWG%20ABDUL%20RAHMAN.pdf; http://eprints.uthm.edu.my/53/2/AQILAH%20AWG%20ABDUL%20RAHMAN%20WATERMARK.pdf; http://eprints.uthm.edu.my/53/3/AQILAH%20AWG%20ABDUL%20RAHMAN%20COPYRIGHT%20DECLARATION.pdf; http://eprints.uthm.edu.my/53/
• Institution: Universiti Tun Hussein Onn Malaysia

Rapid disappearance of habitats and species justifies the need to study species range and distributions so that high-priority conservation areas can be identified. Species distribution modeling (SDM) is an approach that can overcome the time and budget constraints of protected area management. This research aims at providing an efficient method for stakeholders to develop conservation strategies using butterflies as environmental surrogates. The objectives are to (i) document butterfly diversity in Johor, (ii) map out their distribution using geographic information system (GIS), (iii) determine the best model with variables that affects the distribution of butterflies in Johor, (iv) predict the distribution in less sampled areas using SDM, and (v) determine strategies for prioritization of conservation areas based on the results obtained from objective (iv). Butterfly presence data were obtained from field samplings from ten sites throughout Johor, published literature, reports, and reference collections. A total of 2347 presence records comprising of 447 species were collated in Johor. This accounts for 42% of the overall diversity in Peninsular Malaysia. MaxEnt was then used to model and map the potential distribution of butterflies based on land cover, vegetation indices, land use, elevation, and bioclimatic layers as variables. Nine models were formulated and compared based on the area under curve (AUC) for receiver operating characteristic (ROC) values, percentage of habitat suitability, and variable complexity. Model 6 was chosen as the best model with distance to forest, temperature, precipitation, and distance to road being the highest contributing variables. Forest-dwelling butterflies consistently showed the best model performance. Based on the habitat suitability map generated, the high to low priority ranking of conservation clusters are as follows: Endau–Rompin–Labis forest complex, Gunung Ledang forest complex, Pulai forest complex, Kluang forest reserve, and Maokil–Air Hitam–Bukit Inas forest complex. Implementation of MaxEnt in Malaysia can be improved through capacity building, fostering better communication between stakeholders and integration into site-based management plans.