Combined application of the EM-DEA and EX-ACT approaches for integrated assessment of resource use efficiency, sustainability and carbon footprint of smallholder maize production practices in sub-Saharan Africa

The goal to improve food security in sub-Saharan Africa (SSA) through domestic, resource efficient and low carbon agriculture is importance. Interventions to produce more food could impact the resource-base and lead to increase in greenhouse gas (GHG) emissions from agroecosystems. Unfortunately, existing methods are limited in analyzing small-scale agricultural systems, and this situation is an obstacle to decision making which aims at sustainable agriculture. In this paper, we showcase the recently developed Emergy-Data Envelopment Analysis (EM-DEA) approach to assess the resource use efficiency (RUE) and sustainability in maize production systems in Ghana, SSA. Using the Agricultural Production Systems sIMulator (APSIM), five land use and resource management scenarios were modeled to represent practices as decision making units (DMUs) in small-scale maize systems. The carbon footprint of the systems was assessed using an approach, which we adapted from the FAO Ex-Ante Carbon balance Tool (EX-ACT). The overall trend of the results showed that the yield, total emergy, GHG emissions and carbon footprint all increased with increase in urea application intensity. However, the relationship between the yield and urea intensity was not always linear. A system that used more renewable or fewer resources to produce a yield equal to that of its peer was considered more efficient and sustainable in relative terms. In particular, the business-as-usual scenario (12 kg/ha/yr NPK input to rainfed maize system, i.e. Extensive12) was inefficient when compared to the four contrasting scenarios. The ecological intensive scenario (20 kg/ha/yr urea input to rainfed maize-legume intercropping system, i.e. Intercrop20) achieved the greatest marginal yield, better RUE and sustainability. The high input scenario (100 kg/ha/yr urea input plus supplemental irrigation to maize monoculture, i.e. Intensive100) produced the greatest yield, but the demand for purchased inputs as well as GHG emissions and carbon footprint were greatest. The no external input scenario (0 kg/ha/yr urea input to rainfed maize system, i.e. Extensive0), and the moderate input scenario (50 kg/ha/yr urea input plus supplemental irrigation to maize monoculture, i.e. Intensive50) showed the greatest and least yield gaps relative to Intensive100, respectively. Based on these results and trade-off analysis, it was evident that Intercrop20 and Intensive50 were the two best case scenarios. As such, land use policy that aims at sustainable agriculture could recommend Intercrop20 and Intensive50 for implementation in low and high input maize production systems, respectively. Comparison between our results and other existing empirical studies revealed similarities that confirm our results. We conclude that the information derived using the EM-DEA and EX-ACT approaches could be useful when making informed decisions that aim at sustainable agriculture. Despite the limitation caused by scarcity of data, the use of the EM-DEA approach led to inclusive information on RUE and sustainability of the DMUs. Hence, the EM-DEA approach represents a way forward to better assess energy footprint in agricultural land use as a whole.