An understanding of soil characteristics provides essential information for farmers to take decisions regarding agricultural practises, especially turning of the superficial soil layer. The results of this study indicate that preserving soil structure (no soil turning) generates a stable condition similar to that in native areas and maintains high levels of carbon in the soil. Thus, maintaining soil as undisturbed as possible for crop production is an alternative to preserve systems, to reduce atmospheric carbon (carbon sequestration) and to improve crop and food production.

Ammonia (NH₃) emitted from livestock urine deposited in pasture soils during grazing reduces the nitrogen use efficiency on farms and has negative effects on the environment. We evaluated both the effectiveness and longevity of the urease inhibitors N-(2-Nitrophenyl) phosphoric triamide (2-NPT) and N-(n-butyl) thiophosphoric triamide (nBTPT) in reducing NH₃ emissions from cattle urine added to dairy-grazed pasture soils. The longevity of 2-NPT in reducing emissions was greater than that of the more commonly used inhibitor nBTPT.

Physical and chemical soil properties are factors controlling the fate of pesticides in water resources. We found that atrazine and ametryn were leacher and non-leacher herbicides respectively in these sugarcane soils. Herbicide movement was conditioned by a homogenous soil porosity system and instantaneous and hindered sorption sites. The total leached atrazine recovered could be a threat to drinking water quality.

Understanding how much nitrogen (N) is supplied by the soil to the sugarcane crop through N mineralisation (organic N conversion to plant-available forms) is important for managing N inputs and N loss pathways. We also suggest a model that uses soil properties to predict mineralisation under laboratory conditions. We found that soil %N or %TOC correlated with mineralisable N, however measures representing organic N and CO₂ production improved predictions of mineralisable N over 301 days.

Prediction of plant-available mineral N under field conditions is important for making efficient use of fertiliser and avoiding unnecessarily large applications. We present an approach to predict available N, combining measured soil indices, historical climate information, mineralisation models, and soil temperature and moisture simulations. The approach can incorporate irrigation management and seasonal climate forecasts, with predictions showing similar variability to plant N uptake data, though a thorough validation is required before predictions could form a basis for decision support.

Calcareous soils rich in carbonates (CaCO₃) are causing increased carbon dioxide (CO₂) release from the soil. This study shows that N fertiliser-induced soil acidification is among the causes of CO₂ release from such soils. This carbon loss can be decreased by the combined application of N fertiliser and a nitrification inhibitor dicyandiamide. Dicyandiamide application inhibited nitrification and decreased soil acidification, which decreased the rate of CaCO₃ dissolution in the soil. Consequently, the rate of CO₂ release from the soil decreased.

Crop yields, organic matter, microbial biomass, enzyme activity, and bacterial diversity in the soil increased with crop species in summer tobacco–winter barley–next summer maize, tobacco–fallow monoculture, and tobacco–barley succession systems. The presence of predominant bacteria was more influenced by soil properties than cropping systems in a period of land use.

Clay mineralogy may determine some important properties of soils. Determination of clay type and proportion in calcareous soils with X-ray diffraction is time-consuming and expensive. With use of adaptive neuro-fuzzy inference system (ANFIS) methods, it is possible to estimate clay mineralogy in calcareous soils from other known soil properties.

As high-resolution digital elevation models (DEMs) are increasingly available, we are facing considerable challenges in selecting the optimal DEM. We compared DEMs at resolutions from 1 to 90 m, and assessed their effects on the slope steepness factor calculation and soil loss estimation over a burnt national park in Australia. We found that the most optimal resolution was 5–10 m for the rugged landscapes, and that coarser resolution data resulted in overestimates of slope length and soil loss.