Predictive modelling of soil aluminium saturation as a basis for liming recommendations in vineyard acid soils under Mediterranean conditions
Derechos de accesoopenAccess
MetadatosMostrar el registro completo del ítem
Cita bibliográficaAngel Olego, Miguel, Miguel De Paz, J., Visconti, F., E. Garzon, J. (2014). Predictive modelling of soil aluminium saturation as a basis for liming recommendations in vineyard acid soils under Mediterranean conditions. Soil Science and Plant Nutrition, 60(5), 695-707.
Soil acidification is a process of degradation that becomes more pronounced as a result of various human activities, but can be controlled through appropriate soil management. Calcium, magnesium and phosphorus deficiencies along with aluminium (Al) toxicity are considered the major constraints to plant growth in acid vineyard soils. The main aim of this work was to develop a model for liming amendment recommendation in acid vineyard soils using two liming materials, dolomite and sugar foam. These were used at three doses: 900, 1800 and 2700kgha(-1) of calcium carbonate equivalent (CCE). Seven soil properties, namely pH in water, pH in 1M potassium chloride (KCl), phosphorus content, base saturation, calcium, magnesium, potassium and aluminium exchangeable contents, were monitored at two soil depths (0-30 and 30-60cm) during 3 years. The association among the soil properties, and with the soil acidity, was investigated through principal component analysis. This resulted in the selection of the aluminium saturation in effective cation exchange capacity (Al%ECEC) as the soil property to be modelled. According to the results of a subsequent analysis of variance (ANOVA), the Al%ECEC strongly depends on the dose (in CCE content) of the liming material independently of its dolomite or sugar foam nature. Besides, the dose effect is different depending on the soil depth and the sampling time. As a result, two quadratic models, one per soil depth and for the time of leaf drop stage, have been proposed to make liming recommendations in acid vineyard soils. These quadratic empirical models are comparable with the known linear Cochrane model using an f value between 1.5 and 2 in the range of doses studied, i.e. able to drop the exchangeable aluminium down to 50%. However, the models proposed in this work further provide (i) different dose recommendations for the arable and deeper soil layers, and (ii) confidence intervals for minimum and maximum additions of liming materials and, specifically, for these important soils dedicated to the growing of vines under Mediterranean conditions.