Thursday, 16 June: Julius Kotir
Seminar summary by Dr Jane O'Sullivan, Convenor, GCI Food Security Seminar Series
The future of water for agriculture in the Volta River Basin in West Africa: modelling biophysical and socio-economic influences.
The Volta River catchment is a major river system in West Africa, covering some 385,000 km2 including most of Ghana and Burkina Faso, and part of Mali, Cote d’Ivor, Togo and Benin. Lake Volta is the largest man-made lake in the world. The basin is home to 15 - 20 million people of diverse ethnicity, about 10 million of whom are in Ghana. While 70-90% are subsistence farmers, fishing and small-scale mining are other important livelihoods. Although rainfall averages over 1000 mm, agricultural productivity is low, infrastructure is poor, and land degradation and desertification is progressing. Land clearing, climate change, rapid population growth and rural-urban migration all pose challenges to resource management.
Julius Kotir’s PhD study focused on the Volta catchment within Ghana. It sought to develop an integrated system dynamic model that can help policy makers understand the state of water resources in the basin, and to explore the implications of management decisions to enhance agricultural production and food security.
This involved identifying the key biophysical and socio-economic factors influencing water availability and agricultural production, developing a conceptual model capturing the systemic feedback loops, quantifying the impacts of factors in a segment of the model, and using this to simulate scenarios employing different strategies, up to 2050.
The model was developed using participatory engagement with local experts (farmers and community leaders) to identify drivers, and technical experts (natural and social scientists) to workshop the system connections. The entire model contains 96 parameters, 57 main equations and 9 graphical functions. Sub-sectors of the model, on population, water resources and agricultural production, were operationalised using stock-and-flow diagrams and calibrated using historical data, enabling simulations to be run.
Three policy scenarios were simulated, in comparison with the “business as usual” (b.a.u.) scenario:
- Water infrastructure development – including a 50% increase in reservoir capacity and availability of surface water;
- Cropland expansion – including increasing total arable area by 30%;
- Dry conditions – including a 30% decline in precipitation and 50% decline in available surface water.
Effects of these scenarios on crop yields, agricultural water demand and farm incomes were presented.
Our discussion explored the systems of land tenure and their impact on land management. In most of the study area, land is owned by the community, not by individuals. It was speculated whether greater mechanisation of agriculture could substantially increase production. The difficulty in defining “best practice” makes it hard to project impacts of technological change.
The scale at which the model is applied also influences the value of its predictions. When modelling the average impact of factors over large areas, differential impacts are neglected and trade-offs are largely hidden. While this study took the basin as a whole, the model could be applied to smaller, more uniform areas or communities, to produce outputs of greater relevance to decision-making. At any scale, boundary effects should be considered: what are the impacts of changes within the study area on people and ecosystems beyond it? This model is making a start towards exploring these dynamics.