Managing nutrient interactions for improved agronomic efficiency under contrasting soil carbon levels in smallholder farms

Much of Sub –Saharan Africa is plagued by food insecurity due to low and variable soil fertility recurrent droughts and limited access to external inputs among other reasons. Variable levels of soil fertility caused result from diverse soil fertility management methods as well as parent materials of origin leading to inconsistent responses to fertilizer and manure, within and across fields and farms. It was hypothesised that soil organic carbon can be broadly used to define soil fertility domains for fertilizer use efficiencies. An analysis of the profitability of farmer management practices using a decision support tool, Nutrient Expert for Hybrid Maize (NE), was also carried out. Soil fertility was broadly categorised into three domains based on soil organic carbon (SOC) as an indicator of fertility as follows: field type1 ≤4 g C kg-1 soil, field type 2, >4–6 g C kg-1 soil and field type 3 >6 g C kg-1 soil. Nutrient omission experiments were set out to ascertain attainable yields, indigenous nutrient supplies and nutrient responses in Dendenyore ward, Hwedza district over two seasons, 2011/12 and 2012/13. The experimental design was a randomized complete block with plot sizes of 4.5 x 5 m2 and three replicates. Data was analysed by GENSTAT 13 using analysis of variance (ANOVA) to examine site effects and treatment on maize productivity and site means were separated by Tukey’s 95 % confidence interval. Treatments used were i) zero fertilizer control, ii) Nitrogen (N), Potassium (K), (iii) N, Phosphorus (P), Sulphur (S), iv), PKS and v) NPKS. Maize productivity was significantly influenced by nutrient management across all sites, with site mean yields on soils with >4g C kg-1 soil (Field types 2 and 3) ranging 3–3.2 t ha-1 significantly higher than for Field Type 1 which were less than 1 t ha-1. Crop yield responses to N and P ranged from 1.2–2.35 t ha-1 and 0.71–2.10 t ha-1, respectively with the highest responses on soils with >4g C kg-1 soil. Response to K was not significant. Across all sites indigenous N and P supplies were not enough to support a yield of one tonne of maize per ha without external nutrient input. The indigenous N supplies ranged from 8.5 kg to 27.6 kg N ha-1, whilst indigenous P supplies were 2.2–11.7 kg P ha-1. The second set of experiments involved applications of N, P, K and S alone or in combinations with manure, lime and micronutrients in a complete randomized block design with plot sizes 4.5 x 5 m2 and three replicates. Data was analysed using GENSTAT 13 with site, treatment and year as fixed effects in a general ANOVA. Sole manure and a zero fertilizer control were the other treatments. Combinations of manure and fertilizer yielded the highest ranging from 2.67–5.55 t ha-1 across sites. Addition of sole manure on soil with 4 g C kg-1 soil there was a significant increase to 2 t ha-1. The liming effect was only evident on soil with 4g C kg-1 soil) must be targeted for application of fertilizers to increase productivity in smallholder farms. The hypothesis that SOC can be used as an indicator of soil fertility status was accepted. Analysis with NE clearly showed that socio-economic status and soil fertility management practices were linked thus soil fertility and resource endowment cannot be separated. Smallholder farmers must target the more fertile soils and build up soil fertility on the soils with