Predicting the spatial suitability distribution of Moringa oleifera cultivation using analytical hierarchical process modelling

Abstract

Moringa oleifera Lam, often grows well under cultivation in the tropics and sub-tropics. It does well in sandy to clayey soils and tolerates low rainfall. The plant is well-known for its nutritional and medicinal properties, hence it is fast gaining popularity in South Africa and the rest of the world. The objective of this study was to predict the suitable areas for cultivating M. oleifera in South Africa using climate and edaphic variables that significantly affect its growth and development. We used an Analytical Hierarchical Process (AHP) and Geographic Information System (GIS) weight function to assign suitability weights to criteria and sub-criteria that affect the plant's growth and a predictive cultivation suitability map. Area under the curve (AUC) was used to evaluate the model's performance. The Analytical Hierarchical Process indicated that the most influential variable determining M. oleifera cultivation were, minimum temperature, soil texture, annual rainfall, mean temperature and soil pH, respectively. Further, the results showed that approximately 16.5% (200,837 km2) of South Africa land area has optimal growth conditions, 17.8% (216,758 km2) suitable conditions, 46% (560,794 km2) less suitable conditions and 19% (240,699 km2) not suitable conditions for cultivating M. oleifera. The area under the curve (AUC) metric of our suitability model suggested that the map is 81% accurate for predicting the spatial suitability of cultivating M. oleifera in South Africa. The results also confirm that the use of AHP model with GIS weight function is useful for explicit identification of sites for M. oleifera cultivation for maximum production output. The results of this study can be useful information for the land-use policy makers and farmers for informed decision regarding the cultivation of M. oleifera in South Africa.