Development of geospatial model for Metisa plana (Walker) outbreak and outbreak prediction in oil palm plantations in Malaysia

Metisa plana (Walker) is a leaves defoliating insect that is able to cause complete skeletonization and death of the oil palm’s fronds. This insect can cause a loss of USD 2.32 billion for two consecutive years given only 10% of the 5 million hectares of oil palms being infested. Hence, efficient, rigorous control methods should be properly planned. In order to do this, the role of environmental factors on the pests’ population’s fluctuations should be well understood. Nonetheless, the current practices are still leaning towards the conventional approaches that are highly dependent on ineffective, time-consuming in-situ data collection. On the other hand, the utilization of geospatial technologies can be used to obtain data in rapid, harmless, and cost-effective manners. This study utilized the geospatial technologies to i) examine spatial and temporal climatic stresses that cause the outbreak of Metisa plana, ii) to construct the relationship between the geospatial data and Metisa plana outbreak, and iii) to predict the outbreak of Metisa plana in oil palm plantation. Metisa plana census data of larvae instar 1, 2, 3, and 4 were collected approximately biweekly over the period of 2014 and 2015. Moderate Resolution Imaging Spectroradiometer (MODIS) and The Tropical Rainfall Measuring Mission (TRMM) satellite images providing values of land surface temperature (LST), rainfall (RF), relative humidity (RH), and Normalized Difference Vegetation Index (NDVI) were extracted and apportioned to 6 time lags; 1 week (T1), 2 weeks (T2), 3 week (T3), 4 weeks (T4), 5 week (T5) and 6 weeks (T6). Linear relationship between Metisa plana with LST, RF, RH, and NDVI were carried out using the Pearson’s correlation, multiple linear regression (MLR) and multiple polynomial regression analysis (MPR). Artificial neural network (ANN) was then used to develop the best prediction model of Metisa plana’s outbreak. Presence of Metisa plana was influenced by LST, RF and RH, but not NDVI. The LST between 24oC and 28oC showed a strong relationship with Metisa plana, whereby its presence started to decline with LST from 28oC and above. However, the effect of time lag on the presence of Metisa plana was not prominent. The best MLR model was obtained with LST, RF and RH at T4 to T6 with an adjusted R2 = 0.29. The MPR model of LST at T4 to T6 depicted the best fit line with an adjusted R2 = 0.50. The highest accuracy of 95.29% was achieved by models generated by ANN utilizing the relative humidity at T1 to T3. The model generated by combined variables, LST, RF and RH at T4 to T6 was able to predict the presence of Metisa plana with the accuracy by up to 89.95%. Based on the result of this study, the elucidation of Metisa plana’s landscape ecology was possible with the utilization of geospatial technology.

Text FP 2019 6 ir.pdf Download (1MB)

Actions (login required)

View Item