A further insight into the environmental factors determining potential grain size in malt barley under Mediterranean conditions

摘要

High grain size is a primary target in malt barley especially under Mediterranean conditions where thin grains production is very common. In fact, market value depends on grain size due to the positive relationship between grain size and malt extract yield, which in turn determines the amount of produced beer. Using a large and unbiased data set (n = 241) derived from 25 different environments (Years x Locations) and 36 state-of-the-art European malt barley varieties, we aimed to: i) identify the critical period in which plump grains (% grains > 2.5 mm) fraction presents the highest response to rainfall, ii) verify if plump grains fraction could be predicted with simple and easily measurable soil and meteorological variables, and iii) identify the main environmental drivers of plump grains determination. It was demonstrated that both environmental and genetic effects are important in determining grain size in two-rowed malt barley. Data were analyzed by partial least square regression (PLSR) models that were calibrated and cross-validated for plump grains prediction. The use of PLSR modelling defined the critical periods when rainfall was most influential in determining plump grains fraction. It was demonstrated that the sum of rainfall during the period extending 6 days pre-anthesis (approximately 100 GDD) and 20 days (approximately 390 GDD) post-anthesis was the best predictor of plump grains (hereinafter referred to as "rainfall during grain filling period - RGF"). PLSR models were then developed using the entire data set and all possible combinations of the available meteorological (Tmean, Tmin, Tmax, MaxT, D > 30, wind speed and RGF) and soil (P, K, pH, soil organic content, sand, clay, silt) variables. The highest coefficient of determination in cross-validation (R-cv(2) = 0.74) and the lowest RMSEcv (12.58 %) was achieved when K, clay, silt and RGF were the only variables remained in the analysis. These results may lead to a better understanding of crop-environmental interactions enabling for better crop management, enhancing crop models predictability and assisting malt barley industry in decision making and production planning.