Investigating the opportunities to improve smallholder rubber production through the development of an integrated decision support system: A Malaysian case study

Abstract

Rubber tree (Hevea brasiliensis) is the primary source of natural rubber, an important raw material for industrial and non-industrial products. Smallholders are the primary producers of rubber. They account for 85% of global rubber production and more than 90% in Malaysia. Increased volatility in global rubber prices has led to the decline in Malaysian rubber production and this has adversely impacted the livelihood of local smallholders. The premise of this thesis is that there is an urgency to help smallholders improve their production and revenue. The thesis explores whether an integrated rubber decision support system can be used to enhance rubber stakeholders understanding of the impact of the interaction between factors on rubber production and facilitate in the formulation of effective smallholder-oriented policies. Decision support systems have been widely applied in agriculture to facilitate the improvement of crop production through enhanced farm management practices and the development of effective intervention strategies. Little attention has been paid to the incorporation of biophysical, socioeconomic and institutional factors in the development of decision support systems to improve smallholder rubber production. This thesis explores this gap by developing an integrated rubber decision support system that incorporates the three key factors and evaluating the model capacity in providing effective interventions to improve smallholder rubber production in Malaysia. The thesis begins by systematically reviewing the factors that influence rubber production and explore how they have been incorporated into the design of existing rubber decision support systems. The limitations of existing models are then analysed to provide insights into the development of the integrated model. Results revealed that despite being a smallholder crop, there is a paucity of research on the influence of socioeconomic and institutional factors on rubber production. Existing models have three main limitations: (1) the extensive data requirement; (2) the lack of incorporation of socioeconomic and institutional factors; and (3) the difficulty of communicating model results to smallholders. Our analysis highlights the improvement strategies for the development of an enhanced rubber decision support systems. The strategies include: (1) refinement in data collection and analysis; (2) inclusion of socioeconomic and institutional factors; and (3) enhancement of user convenience. The first strategy was addressed by assessing yield performance of 37 rubber clones at two major production regions in Malaysia. The results showed that the yield performance differed widely by clone but showed relatively consistent trends across regions. The individual clonal yield performance was then applied to the development of the integrated rubber decision support system. The remaining strategies were addressed using systems thinking. Causal loop modelling was conducted to analyse and explain the underlying feedback mechanisms causing the decline in Malaysian rubber production. The results showed that current policies do not effectively address issues caused by the feedback between biophysical, socioeconomic and institutional factors in the global, national and farm levels within the rubber production system. This method provides a platform to better facilitate engagement with rubber stakeholders, especially smallholders, and inform the formulation of effective smallholder-oriented policies. A well-established rubber decision support system, the Bioeconomic Agroforestry Model, was updated and integrated into a system dynamics model to quantitatively analyse smallholder rubber production. The foundation of this integration was to incorporate a comprehensive set of biophysical, socioeconomic and institutional factors in the development of an enhanced decision support system. Our analysis shows that the integrated model demonstrated a reliable prediction capacity on smallholder rubber production. Finally, the integrated model was used to examine the interactions between key components of the three levels in the rubber production system, and to explore plausible policy scenarios to improve smallholder rubber production. The results showed that the balanced policy scenario, which targets all levels within the system, would significantly improve smallholder rubber production. Based on the analysis, rubber production can be improved by improving rubber export restriction at the global level. This needs to be accompanied by increasing domestic rubber consumption and reducing rubber import at the national level. At the farm level, the following changes are required to improve production: (1) an increase in the government rubber production incentive and improvement in its campaign; (2) an improvement in the collaboration between agencies; and (3) encourage the generation of additional income through the diversification of agricultural activities.A significant contribution of the developed model in this thesis is that it provides a decision support tool for key rubber stakeholders to improve their understanding of the long-term dynamic behaviour of rubber production to formulate effective smallholder-oriented policies to improve smallholder rubber production and assist in identifying areas requiring further research and development. The ability of this model to assess alternative policy scenarios can also promote collaboration between the key stakeholders to improve rubber production. Finally, the integrated rubber decision support system can be adapted by other rubber producing countries to evaluate the performance of their policies at improving local production.