Variation in bioavailable zinc and iron in Zimbabwe and their effects on crop productivity and quality.

Abstract

Dietary micronutrient deficiencies remain prevalent in southern Africa, especially among rural households consuming staple plant-based diets. Despite evidence of improved micronutrient concentration in staple grains from micronutrient fertilizers, the heterogeneity of farming systems in southern Africa presents major challenges in understanding factors affecting soil micronutrient supply and grains micronutrient quality. The aim of this study was to determine the impact of geospatial variation in soil zinc (Zn) and iron (Fe) concentrations and soil fertility management strategies on grain Zn and Fe concentrations of cereals and cowpea in Zimbabwe. Survey and field experimental approaches were employed from 2014 to 2018 in two Districts varying in agro-ecology in eastern Zimbabwe (Hwedza: 650-800 mm annum-1 and Mutasa: >1000 mm rain annum-1 ). Soil and crop surveys were conducted in 2015-16 to determine factors governing geospatial variation of bioavailable soil Zn and Fe in two contrasting soil types (clayey and sandy) and two field productivity levels (most and least productive fields). An inherited hierarchical sampling design consisting of random effects, fixed effects and covariates was employed. Of the surveyed farms (n=175), >80% had plant-available soil Zn concentration below the critical limit of 0.8 mg kg-1 . In contrast, only 15% of the farms had low plant-available soil Fe concentration of <5 mg kg-1 . Using a linear mixed effects (LME) model, the between-Districts variance component (treated as a random effect) for plant-available soil Zn concentration (0.034) was much smaller than the variance components for farms within Wards (0.293) or fields within farms (0.351). Thus, farmer soil fertility management and short-range spatial effects are more important in predicting soil and grain Zn concentration than broader spatial variation, although the latter can still identify broad spatial trends. Plant-available soil Fe concentration was affected more by soil geochemical factors (soil type and soil pH) than farmer soil fertility management. Crop type had the largest effect on grain Fe concentration of crops grown on farmers’ fields with small grains (sorghum and finger millet) and cowpea consistently having larger grain Fe concentrations of 78.1, 62.3 and 43.7 mg kg-1 , respectively, than maize (28.0 mg kg-1 ). Zinc fertilizer use under integrated soil fertility management (ISFM) significantly improved crop productivity and grain quality of cowpea grown under a legume-maize rotational sequence on contrasting soil types, with a more response to Zn fertilizers and organic matter on sandy soils than on red clay soils. In a different set of field experiments, at sites in Hwedza and Mutasa, nitrogen (N) fertilizers increased grain Zn concentration in field-grown maize, but not in cowpea, when N was applied in combination with soil and foliar Zn fertilizers, than when Zn fertilizers were applied alone. For example, application of soil and foliar Zn fertilizers with mineral N at 45 kg ha-1 produced the largest grain Zn concentration of 39.3 mg kg-1 compared with 30.4 and 29.9 mg kg-1 when N fertilizer was applied as organic and combinations of organic and mineral N, respectively. This implies that fields belonging to resource constrained and intermediate farmers could attain improved grain Zn when N is applied as mineral fertilizer. A maize grain Zn concentration of only 27.2 mg kg-1 was attained when N was not used. At larger N application rates of 90 kg ha-1 , comparable mean maize grain Zn concentrations of 36.3, 36.1 and 38.0 mg kg-1 were attained when N was applied as sole mineral, organic N and combinations of mineral and organic N fertilizer, respectively. Resource-endowed farmers, with the capacity to apply optimal rates of N, could therefore have a wide range of N fertilizer application forms and application strategies to choose from with respect to increasing grain Zn concentration. There were larger increases in grain Zn concentration of maize grown in Hwedza than in Mutasa when different N management strategies were employed with external Zn fertilizer. These differences in crop responsiveness to Zn iii fertilization between Hwedza and Mutasa indicate that agro-ecological differences, which could also include variations in landscape and soil type, can affect the optimal farmer decisions for improving grain Zn concentration. Grain Fe concentration was ~30% greater across all co-applied Fe and N fertilizer treatments than in crops collected from nearby farmer fields. A significant effect of N form on grain Fe concentration was observed for one finger millet landrace, where mineral N fertilizer application increased grain Fe concentration to a greater extent than other N forms, but not for the other crop types (maize and cowpea) in this study. Whilst good soil fertility management is essential for yield and grain quality, the effects on grain Fe concentration are less consistent than those seen previously for Zn. Human dietary deficiencies in Fe and Zn are affected by inherent variation in soil nutrient composition, soil geochemical factors and farmer management strategies employed on farm level. This study shows that it is possible to improve micronutrient supplies through improved soil fertility management practices (i.e organic nutrient resource use), crop type choice and a better understanding of agro-ecological effects on micronutrient fertilizer use. There is therefore a need to inform agriculture and nutrition policies to offer knowledge on these locally adaptable technologies which could ensure inclusive improvement of micronutrient supply in different smallholder social contexts and farming systems