Corn Stalk Nitrate Test (CSNT); Developing a Sampling Protocol Based on within-Field Spatial Variability and Yield Zones.

Nitrogen (N) over-fertilization can decrease crop profitability and cause loss of N to the environment. The corn stalk nitrate test (CSNT) is an end-of-season assessment tool that can be used to identify where N management changes can be made in future years. New York’s nutrient management policy for regulated farms allows producers to override land-grant university recommendations in favor of an adaptive management approach that uses yield assessments and CSNT results. Current recommended sampling density is 2.2 stalks ha^-1 with each field being represented by one CSNT value. The aim of this study was to (1) describe and relate within-field CSNT and yield spatial variability, and (2) to develop a yield-based and targeted CSNT sampling protocol. Five corn silage fields were sampled post-harvest for cornstalks in 2013. Whole field CSNT variability was quantified using a regular grid sampling protocol with 100 stalks. Optimal sampling density was estimated by simulating sampling densities from 0.4 to 11 stalks ha^-1. Short-range CSNT variability was studied using ten stalks in six randomly selected rows within a field. Yield was recorded using forage yield monitors, three yield zones (low, medium and high) were created, and CSNT variability was determined for each zone. In 2014 (3 fields), CSNT sampling locations were chosen based on previous year yield zones (3 locations per zone, 17 stalks per location) and stalks were harvested following a circle pattern. Whole-field and short-range CSNT variability were high, suggesting that the currently recommended sampling protocol of 2.2 stalks ha^-1 is insufficient. Variability in CSNT was similar among yield zones in 2013 (high yields) but lower in low yielding zones than in high yielding zones in 2014 (lower yields overall that year). Sampling protocols that take into account CSNT spatial variability and/or other within field information (yield, soil properties, etc.) can improve usefulness of a CSNT-based adaptive management approach.