Maize fertilization under conservation agriculture systems on heterogeneous smallholder farms in Zimbabwe.

Abstract

Soil fertility depletion, unreliable rainfall patterns and poor access to nutrient supplements continue to impede crop production in much of Sub-Saharan Africa (SSA). There is increasing need to develop holistic approaches to deal with soil fertility depletion and food security in the region. Farming in Zimbabwe is characterised by blanket fertilizer recommendations that fail to address farm-specific niches which potentially pose as hotspots for crop production. Farmers have developed adaptive mechanisms that are specifically tailored to increase production at the farm level. In light of this argument, the current study sought to gain an improved understanding on the effects of basin-based conservation agriculture (BCA), furrow-based conservation agriculture (FCA), and conventional tillage (CT) on maize grain yields under rain fed conditions and soils with different soil organic carbon (SOC) contents in Murewa district, Zimbabwe. An assessment of farmers’ local knowledge of soil fertility was carried out using a survey during the 2013/2014 cropping season. Yield estimates on 152 fields were quantified using yield cuts on fields that each of the farmers identified as either rich or poor during the 2013/2014 cropping season as well. Another set of experiments was set up as nutrient omission trials on both the rich and poor fields with contrasting SOC in 3 cropping seasons (2013/2014 to 2015/2016) using nitrogen (N), phosphorus (P), potassium (K), cattle manure (M) and their combinations on twenty farms in Study 1, and ten farms in Study 2 under rain-fed conditions in Murewa. Phosphorus fertilizers also contain sulphur (S) but adding an NKS fertilizer combination was not possible due to lack of fertilizers containing S only in Zimbabwe. Grain yield data was analysed in Genstat 14 to test for differences between grain yields from poor and rich fields. Questionnaire data was subjected to analysis of simple proportions in SPSS version 20. Indicators of soil fertility that were consistent with farmers’ perceptions of soil fertility were also identified. Farmers in Murewa district, were grouped into four recommendation domains (RDs), that is the very rich, rich, poor and very poor using draught power, product orientation, mineral fertilizer use, labour use and annual remittance. Host farmer depiction of both rich and poor soils agreed with the laboratory indices (p < 0.001). Poor fields had SOC < 0.4 % and had significantly lower maize grain yields than rich soils (p= 0.001). The poor soils had significantly lower (p < 0.001) bases, % clay, SOC, total N and were more acidic than the rich fields. Most farmers are constrained in inputs and harvested average maize yields of less than 1 Mg ha-1 . Farmers have a deep knowledge and understanding of their farm niches that is passed down generations. Farms had SOC ranging from 0.18% to 0.89%, and clay content from 6 to 15%. Maize grain yields were significantly higher under B-CA compared to both F-CA and CT (P <0.01). For Study 1, nutrient management showed positive effects on yields across the three years (P<0.001) but there were no significant tillage effects observed. Grain yields increased from 0.3 Mg ha-1 for unfertilized control to 4.1 Mg ha-1 for the nitrogen + phosphorus + potassium + manure (NPKSM) treatment. Yield response to N was consistently larger than for P or K, irrespective of soil fertility status. Response to N significantly increased with increase in soil fertility, suggesting higher N use efficiency for soils with higher SOC. Except for NPKSM, no significant yield differences were observed under the residual and additive plots for treatments, when N was added each year. At productivity levels of <4 Mg ha-1 , there was no yield gain in applying both P and K for consecutive years, suggesting that nutrient investments by resource constrained farmers for Year 2 could target only N application. For Study 2, maize grain yields were significantly higher under B-CA compared with both F-CA and CT (P <0.01), iv suggesting a positive effect due to concentration of nutrients in the second year. The consistently larger NPKSM yields highlight the importance of integrated nutrient management, combining mineral and organic sources of nutrients to ensure maize productivity on poor soils in agro ecologies receiving unreliable rainfall.