Residue management strategies for the rainfed N-deprived maize-legume cropping systems of central Mozambique

Abstract

Identifying best-fit residue management strategies for Central Mozambique rainfed N-deprived maize-legume cropping (LEINA) systems managed under conservation agriculture is critical to improve resource use efficiency and yields. Therefore, understanding under what circumstances positive and negative responses from carbon rich residue mulches are likely to occur and what drives farmers’ decision on the adoption of technologies is critical to design relevant and feasible recommendations that fit farmers complex farming circumstances. Here I used household survey data, field experimentation and cropping systems modelling to: 1) characterize farm diversity and understand how it affects smallholder farmers resource use attitudes, management decisions and the likelihood to engage in sustainable intensification (SI) practices; 2) identify best-fit residue management strategies that are more likely to improve resource productivity and yields; and 3) develop simple rules of thumb to match residue management solutions to N-deprived smallholder farming systems. Household characterization data showed that, rather than farm size and labour availability, farmers resource use attitudes and household capacity to meet its annual food security and income generation targets where key farm differentiation factors, therefore critical to map farm typology feasible intensification options. Using a heuristic decision tree to group smallholder farmer into a new set of mutually exclusive farm typologies that highlight existing intragroup diversity among Mozambican small-scale farmers, three key typologies where identified. The resource endowed and innovative farmers (Type B); resource endowed but change resistant farmers (Type - A2); and resource constrained (Type - A1) farmers, who can also be maize self-sufficiency (Type-A1a) and the food insecure (Type-A1b). For the already maize self-sufficient Type-B and Type-A2 farms, the challenge is to improve agronomy to generate enough surplus and income from maize. Mechanization and labour qualification are critical for these groups. For resource constrained Type-A1 farmers, the primary challenge towards intensification is to generate enough income to improve the household maize self-sufficiency and income in the system. Considering its high market value and further contribution to soil N-availability, legume production is a feasible investment alternative. Nevertheless, two key intensification traps were identified: the household maize self-sufficiency and soil fertility perception, and land availability perception trap (extensification). The traps are perception based mind-sets that reflect farmers understanding of their circumstances and the strategies they use to materialize their annual food security and income generation targets. Field trials and model-assisted analyses (64 years) of carbon rich residues effects on a cereal (maize) and a legume (cowpea) crop, at different levels of N supply, and three soils of contrasting water holding capacity (WHC) showed that for Central Mozambique rainfed N-deprived systems, residues are more likely to (i) improve the yield, WUE and NUE of the unfertilized legume crop across all tested environments; and (ii) that the application of carbon rich residues reduces maize yields, NUE and WUE in a high WHC sandy clay loam (cSaCL) soil, and on the wet lowland sandy loam textured soil (SaL). In high WHC soils, only a 5% chance of positive yield response from carbon rich mulched soils was simulated at 0 (ND) and 23 kg N/ha (LNA). At 92 kg N/ha (HNA) benefits from carbon rich residues are expected only in 20% of the seasons in high WHC soils. However, such benefits are only attained during the driest years. In the dry and low WHC fine sandy soil (fSa), positive responses to residue application were simulated in 85-90% of the seasons in both cowpea and maize. In terms of resource productivity, maize proved to be more responsive to N-application, especially on the wet and moisture rich SaL soil. Here, maize response to N-application was attributed to better nitrogen uptake (NUpE) and translocation efficiency (NUtE) due to high incrop moisture regimes. In the drier fSa, poor soil water availability led to poor NUpE and consequently lower maize NUE and yields. For cowpea, yields at HNA were hindered by poor N-translocation into grain. Despite maize providing the best responses to N-fertilizer application in mulch free systems, yield penalties from residue applications in continuous maize cropping systems. The findings from this study showed that, in the rainfed N-deprived systems of central Mozambique, the overall performance of maize-legume cropping systems under residues is governed by two critical interactions: 1) a crop and soil water induced residue response trade-off; 2) a residue modified nitrogen driven WUE and NUE trade-off. Understanding these two responses is key to validate locally feasible resource management strategies crucial to the effective validation of CA systems. However, because resource access and management differs across households, an interactive and personalized socio-technical intervention that takes into account each group biophysical and socioeconomic circumstances is required to successfully involve farmers in co-designing group specific technological interventions centred on the improved use of locally available resources and those that farmers can afford to mobilize and use.