Agriculture (Jul 2022)
Simulating Spring Barley Yield under Moderate Input Management System in Poland
Abstract
In recent years, forecasting has become particularly important as all areas of economic life are subject to very dynamic changes. In the case of agriculture, forecasting is an essential element of effective and efficient farm management. Factors affecting crop yields, such as soil, weather, and farm management, are complex and investigations into the relation between these variables are crucial for agricultural studies and decision-making related to crop monitoring, with special emphasis for climate change. Because of this, the aim of this study was to create a spring barley yield prediction model, as a part of the Advisory Support platform in the form of application for Polish agriculture under a moderate input management system. As a representative sample, 20 barley varieties, evaluated under 13 environments representative for Polish conditions, were used. To create yield potential model data for the genotype (G), environment (E), and management (M) were collected over 3 years. The model developed using Multiple Linear Regression (MLR) simulated barley yields with high goodness of fit to the measured data across three years of evaluation. On average, the precision of the cultivar yielding forecast (expressed as a percentage), based on the independent traits, was 78.60% (Model F-statistic: 102.55***) and the range, depending of the variety, was 89.10% (Model F-statistic: 19.26***)–74.60% (Model F-statistic: 6.88***). The model developed using Multiple Linear Regression (MLR) simulated barley yields with high goodness of fit to the measured data across three years of evaluation. It was possible to observe a large differentiation for the response to agroclimatic or soil factors. Under Polish conditions, ten traits have a similar effect (in the prediction model, they have the same sign: + or -) on the yield of almost all varieties (from 17 to 20). Traits that negatively affected final yield were: lodging tendency for 18 varieties (18-), sum of rainfall in January for 19 varieties (19-), and April for 17 varieties (17-). However, the sum of rainfall in February positively affected the final yield for 20 varieties (20+). Average monthly ground temperature in March positively affected final yield for 17 varieties (17+). The average air temperature in March negatively affected final yield for 18 varieties (18-) and for 17 varieties in June (17-). In total, the level of N + P + K fertilization negatively affected the final yield for 15 varieties (15-), but N sum fertilization significantly positively affected final yield for 15 varieties (15+). Soil complex positively influenced the final yield of this crop. In the group of diseases, resistance to powdery mildew and rhynchosporium significantly decreased the final yield. For Polish conditions, it is a complex model for prediction of variety in the yield, including its genetic potential.
Keywords
