Currently the world’s energy supply system for the 21st century is under intensive debate. Renewable energy sources, especially biomass, play an important role in the discussion where several scientists have tried to identify high potential areas for the production of biomass for energy. African countries such as Mozambique are currently under review, because of a favourable climate, ample land availability and low population density (Batidzirai et al., 2006). Cassava, a potential crop for the production of bio-ethanol, is currently being discussed as a feasible option for biofuel production by smallholder farmers as well as maize and sorghum. However current (cassava) yields are low and farmers have very limited resources for yield improvement.
This study was developed with the overall goal to explore feasible yields for cassava production for food and fuel in the context of smallholder farming systems in Alto Molócuè, Northern Mozambique with the objectives of i) assessing the heterogeneity between farms by making a rapid farm characterisation, ii) estimating current yields of cassava, sorghum and maize of selected smallholder farmers, iii) explaining current yields of cassava, sorghum and maize, iv) making a yield gap analysis between actual yields from selected crops from field estimations and feasible potential yields simulated using the FIELD model, v) collecting parameter input used in the FIELD model.
Three villages (Mugema, Nacuaca and Gafaria) were selected in the district of Alto Molócuè to represent variability at district level. Experts, key-informants, back ground information, and first approach farm characterisation interviews were used to describe and categorize bio-physical, socio-economic and farm management practices variability found at village, farm and field level. A farm typology based on expert knowledge and K-means clustering was developed to analyse variability found at farm level. A dynamic simulation model FIELD was parameterized with field data, and simulation scenarios were developed from the rapid farm characterisation interviews and informal meetings with farmers. The model was run to study and quantify the cassava yield response of different farm management practices. Due to circumstances, model simulations were only based on soil fertility properties of the sampled fields.
Simulated yield variability between sites was mainly explained by inherent soil properties, such as soil texture (clay %) which determined exchangeable K level for a large extent. Simulated no input yields in Gafaria were found higher compared to Nacuaca because of: a higher clay content, higher N and K level. Simulated yields for Gafaria ranged between an average of 2.0 and 10.8 t/ha and were found lower for Nacuaca: 1.8 and 3.9 t/ha. Within village, yields were grouped into low yielding fields (1st quartile), medium yielding fields (2nd and 3rd quartile) and high yielding fields (4th quartile). Simulations of continuous cultivation showed yields in Gafaria to drop in average by 24% for 2nd & 3rd quartile fields (stabilizing at 3.1 t ha-1), 27% for 4th quartile fields (stabilizing at 8.0 t ha-1), while yields of 1st quartile fields did not change during time and remained at around 2 t ha-1 over a time span of 25 years. Yields in Nacuaca did not change over time for 1st and 2nd & 3rd quartile fields and remained at 1.8 and 2.3 t ha-1. Yield of 4th quartile fields dropped in average by 22% till it stabilized around 3.2 t ha-1 .
Amongst farmers, a large variation of crop residue management was found. In Mugema, Nacuaca and Gafaria 48, 33 and 60% of the respondents of the rapid farm characterisation incorporated crop residues in the soil and 89% of all farmers asked included grass in preparing of the ridges/hills for cassava. However a large part of the same respondents would use fire for clearing of the field: 59%, 80% and 50% in Mugema, Nacuaca and Gafaria. The effect of maize residue application on cassava fresh yield was analysed with the help of the simulation model FIELD. Addition of maize residues had a positive effect on fresh cassava yield with an increase in yield of between 156 – 237kg/ ha in Nacuaca and 189-449 kg/ha in Gafaria over the three quartiles used per ton maize residue applied.
In general most farmers would only consider the use of manure if they would produce horticultures. Generally manure was piled up and removed not to be of further use. Simulated yields comparing no manure and manure application showed an increase of 402-482 kg and 433-1034 kg increase/ ton DM manure applied for Nacuaca and Gafaria subsequently.
Average yield response per kg applied nutrient was highly variable comparing villages and comparing quartiles. Increasing levels of potassium was most effective on soils in Nacuaca, while a higher response (increase kg fresh cassava / kg applied N, P or K ha-1) for phosphorus was found in Gafaria. First and second quartile classified fields had a lower response and remained at a low yield level compared to the higher yielding fourth quartile fields. High increase in yield was obtained with NPK fertiliser application (100:22:83 N:P:K) ranging from 37 to 65 kg yield increase per kg fertiliser added in Gafaria and between 49 and 67 kg yield increase per kg fertiliser added in Nacuaca. Improved fertiliser management was developed to increase yield response per kg fertiliser applied with the help of available nutrients from crop residues and manure. An increase in yield of 6, 21, and 35% was attained for 4th quartile-2nd & 3rd quartile and 1st quartile fields for Nacuaca. For Gafaria yield increase due to improved management was even higher: 22, 34 and 40% yield increase for 4th quartile-2nd & 3rd quartile and 1st quartile fields.
Ranked as the most important crop by 60% and 70% of the respondents in Nacuaca and Gafaria respectively, cassava plays an important role for farmers in the research villages. Only one third of the farmers in Gafaria (26%) sold cassava compared to 42% in Nacuaca and 48% in Mugema. Overall cassava was mostly sold within the community (82% of respondents). When sold cassava was sold as less than half of the produced quantity: 47, 49 and 43% respectively for Mugema, Nacuaca and Gafaria. If farmers would be involved in the new emerging cassava for the bio fuel market, with current production levels, the surplus being sold to neighbours with insufficient cassava production at the moment, would be sold for biofuel purposes. This thesis has discussed ways to increase yields starting from low capital required residue and manure applications to higher capital investments needed if fertiliser would be applied Outgrower schemes for e.g. the production of tobacco are existing and providing farmers with fertiliser under contract production. These existing outgrower schemes could provide with examples and frameworks on how commercial firms can arrange input purchases or input credit for farmers under production contracts.