Spatio-temporal analysis of yield and weather data for defining site-specific crop management zones

Understanding yield potential and yield-limiting factors is essential for improving profitability and grain yields while avoiding adverse environmental effects. In the USA, grain yield monitors are widely used but the information they provide is rarely used to understand within-field yield variations and associated yield constraints. The objectives of this research were to understand the influence of in-season precipitation on within-field spatio-temporal variation of maize (Zea mays L.) yield and to determine manageable yield variation in two contrasting Major Land Resource Areas of the Mid-Atlantic USA. It does this by assessing the association of grain yield monitor data and in-season precipitation information to be used for variable rate management. Maize yields, as evaluated by baseline functions, were more closely associated with in-season precipitation in the Coastal Plain than in the Piedmont. The study then used standardized principal component analysis (stdPCA) to reveal within-field yield patterns. These varied only under moisture-limited conditions in the Coastal Plain. In the Piedmont, the within-field yield pattern was more consistent under a range of in-season precipitation conditions. In Coastal Plain rainfed fields, the yield predictability increased at the end of June, indicating the possibility of predicting within-field spatial patterns in mid-season. These approaches were successful in deciding whether within-field site- and time-specific management is beneficial for a particular region or field. The presented method, combining stdPCA and geostatistical assessment, is useful in strategizing precision crop management, but do not reveal causes. Detailed soil information and topography could additionally be valuable for understanding constraints to crop yield.