Crop models simulate the growth of agricultural crops and help predicting yields. They are used to estimate the impact of projected climate change on agricultural production and to virtually test possible adaptation strategies. While the direct consequences of climate change are often communicated to the broader public as an increasing sum of hot days and tropical nights, or in the form of prolonged droughts, crop models translate the simulations of climate models into figures relevant for production.
These models are based on the principles of physics and they describe mathematically the physiological processes in a plant and the processes of water and nutrient transport in the soil. If physics was to work the same all over the world, so the idea, then also crop models could be applied to all cropping systems in the world. However, experience shows that crop models are not easily transferable from one region to another. Especially in the description of the soil processes, the models show deficits. For example, models can very well simulate the release of nitrogen – the main soil-derived nutrient for plants – from soil organic matter in the temperate climates of Europe and North America, but fail to work in subtropical climates.
Scientists of the Leibniz Centre for Agricultural Landscape Research (ZALF) in Müncheberg, Germany, have been tracking this fact for a long time. Recently, they have moved one step further: in a study published today in "Scientific Reports", the scientists report on a comprehensive analysis of worldwide experiments on the mineralisation of humus in agricultural soils. It turns out that in hot and humid climates a much larger proportion of humus is not turned over by microorganisms than in temperate or dry climates. "The rapid weathering of the rock under hot and humid conditions has over thousands of years produced mostly clayey soils with high levels of iron and aluminium oxides. In these soils, organic compounds are much more strongly bound and protected against degradation by microorganisms," explains study leader and lead author Dr. Claas Nendel of ZALF. This has consequences for the nutrition of plants, because with less humus turnover less nutrients are released. "At the moment, crop models do not yet make that difference," reports Dr. Nendel, "so that in the simulations they overestimate the nutrient supply of tropical and subtropical crops." This affects the model-based estimation of yields for smallholder farming systems in Africa or Southeast Asia, which often manage without mineral fertilizers. Dr. Nendel's research group at ZALF is now working on developing the corresponding algorithms for crop models, so that also smallholder farming can soon be represented more realistically in simulations.
Article in Nature.
