Using on-farm data to validate crop management recommendations and implementation

Although farmers are encouraged to collect field-specific crop management data to optimize on-farm decisions, farm-derived data are often discounted as a source of information for making broader management and environmental evaluations. This study was conducted to illustrate the use of on-farm data for validating management recommendations and implementation. Data collected by farmers and crop management consultants from 1997 to 2000 from 1119 central Pennsylvania fields were used to evaluate the impact of soil series, potentially available added N (PAAN), and rainfall on corn (Zea mays L.) yield, N required to produce a unit of grain, and PAAN in excess or deficit of crop requirement. Results showed that although observed yields generally followed the same relative ranking as the established soil series yield capabilities, they were consistently lower by 13 to 30%. Also, the amount of PAAN per kilogram of corn grain observed was relatively consistent with PAAN recommendations for the most productive soils and about 25% higher for the least productive soils. Actual PAAN was on average 28 kg ha^-1 higher than PAAN recommended for observed yields. The higher actual PAAN could be due to the difference between observed yields and established yield capabilities used by farmers to determine PAAN recommendations. Because data also showed that rainfall had a significant effect on observed yields, the inability to predict rainfall was another major cause for the difference between actual and recommended PAAN. The analysis showed the database was useful for evaluating corn yields and N management and confirmed that N practices are generally consistent with recommendations.