Environmental availability of micronutrients in tropical soils

Trace elements such as zinc (Zn), copper (Cu) and boron (B), are important micronutrients for crop production. Their bioavailability is essential to crops yield quantity and quality in tropical soils from Sub-Saharan Africa (SSA). Blanket fertilizer recommendations including only macronutrients are most common practice in SSA. Alternatively, science-based approaches have been developed for formulating site-specific fertilizer recommendations. Such science-based approaches could be extended to account for micronutrients, based on the knowledge about the processes that affect their availability for plant uptake.Micronutrient bioavailability depends partly on the soil micronutrient status, and in particular on the labile, reactive or potentially available pool that is distributed over the solid and solution phase, through sorption/desorption and precipitation/dissolution. A fraction of this labile pool is present in the soil solution and is therefore directly available for plant uptake. Together, the soluble and potential available pool represent the soil micronutrient environmental availability.The aim of this thesis was to gain insight in the soil chemical processes that control the solid-solution partitioning of Zn, Cu, and B in tropical soils from SSA, and to use this knowledge to develop accessible tools for predicting the soluble concentrations of these micronutrients. This was done using geochemical multi-surface models and partition relations.Geochemical multi-surface models require input data about the amount of reactive surfaces and their corresponding reactivity for ion adsorption. The metal (hydr)oxides and soil organic matter have been previously identified as the most important reactive surfaces for micronutrient adsorption in soils. The first objective of this thesis was therefore to better understand the reactivity of metal (hydr)oxides and soil organic matter in tropical soils from SSA. Soils from the humid tropics are often intensively weathered, resulting in a high abundance of metal (hydr)oxides. These metal (hydr)oxides are mainly present as well-crystallized metal (hydr)oxides while the contribution of oxide nanoparticles (i.e. ferrihydrite-like materials) may be relatively small on a mass basis. Nevertheless, we have shown in this thesis that irrespective of weathering stage, ferrihydrite is a good model oxide to describe the surface reactivity of the natural metal (hydr)oxides in soils. With regard to the soil organic matter content, our results showed that in soils, the mean particle size of the natural oxide fraction explains the organic carbon content found in soils, irrespective of origin, land use and soil depth. According to these results, soil organic carbon is predominantly stored in primary organo-oxide aggregates that are additionally organized by association with larger mineral particles.In addition to the amount and reactivity of the adsorption surfaces, geochemical multi-surface models also require well-parameterized models to describe the ion adsorption processes to these surfaces. Since ferrihydrite was identified as the best proxy for the natural metal (hydr)oxides in tropical soils, the second objective of this thesis was to use a consistent modeling approach for describing the adsorption of Zn, Cu and B to ferrihydrite with the Charge Distribution model in combination with a Multi Site Ion Complexation model.The third objective of this thesis was to apply a multi-surface model for calculating the solid-solution partitioning of Zn, Cu, and B in tropical soils, and to translate these results in easily accessible prediction tools in the form of partition relations.First, the chemical speciation of B in soils was studied with a geochemical multi-surface model that included B adsorption to dissolved and solid organic matter, ferrihydrite, and clay mineral edges. This was done for five temperate and five tropical. In addition, the performance of previously proposed extraction methods for measuring reactive B were evaluated, since this information is needed as input variable for multi-surface modeling calculations. Based on modeling calculations, the reactive B in soils corresponded best to the B measured in a 0.05 M KH2PO4 (pH 4.5) extraction. In general, the multi-surface modeling showed that 68% or more of reactive boron was present in the solution phase for the soils in this study and that the adsorption was dominated by oxides in the tropical soils, while organic matter was the main adsorbent in the temperate soils. When changing the soil pH, B concentration was found to decrease with increasing pH, and both experimental data and modelling suggested that this effect is mainly due to increased binding of B to organic matter.Next, the solid-solution partitioning of Zn, Cu and B was studied for 172 soils from Burundi, Rwanda and Kenya, using extensive soil characterization in combination with multi-surface modelling and two types of Freundlich-like partition relations. The results showed that the generic multi-surface model applied to these tropical soils performs similarly for Zn and Cu as in previous studies on temperate and contaminated soils. The Zn and Cu speciation was dominated by adsorption to soil organic matter, with an increased importance of metal (hydr)oxides with increasing pH. Given its generally low concentrations in these soils, dissolved organic matter was found to be important only for the solution speciation of Cu. The observed and modeled solid-solution partitioning of Cu and B was found to be rather constant among soils, and the soluble concentration was consistently mainly controlled by the reactive concentration. The solid-solution partitioning of Zn was strongly related to the soil pH. The partition relations in which the soluble concentration was optimized, resulted in an average overestimation for the lowest observed concentrations, and an underestimation for highest concentrations of all three elements. Partition relations in which the Zn solid-solution partitioning was optimized, resulted in more robust predictions since the prediction error was not related to the actual measured concentration.Overall, this thesis delivered valuable information on the micronutrient soil status in SSA, and knowledge on the processes that control soil micronutrient availability. Based on this knowledge, a low-cost and reliable method was developed for predicting soluble micronutrient concentrations, based on partition and transfer relations. This method can be used for testing soils for Zn, Cu and B availability and for making regional and national soil maps showing micronutrient availability. The soil micronutrient information obtained by these methods, can be used in future work for establishing micronutrient fertilizer recommendation systems. &nbsp