Variable Fertilizer Application Based on Yield Goal, Soil Fertility, and Soil Map Unit

Soil fertility and soil productivity are variable within farm fields. Over or under application of a nutrient in parts of every field can result from conventional fertilizer application. This study was conducted to determine whether yield and net returns could be increased by considering field variability when applying fertilizer. The influence of soil fertility variation was considered with soil sampling on a grid pattern and with detailed soil maps made of each field. The fields, located in eastern North Dakota, were mainly a mixture of Haploborols, Calciaquolls, and Argialbolls. Wheat ( Triticum aestivum L.) and barley ( Hordeum vulgare L.) were grown. In 1989, variable rate fertilization based on grid soil sampling did not result in a greater yield of barley (32.1 bu/acre) than the conventional method (30.0 bu/acre). In 1990, the variable rate application yielded more barley (73.7 bu/acre) than the conventional method (67.7 bu/acre). The conventional method, however, resulted in the greatest net returns in both experiments. In 1991, the variable rate method based on a 50‐ft soil sampling grid yielded more wheat (37.3 bu/acre) than the method based on soil mapping units (33.8 bu/acre). The greatest net return ($114/acre) was obtained by using the average soil test by map unit and a yield goal of 50 bu/acre. Yields of barley and wheat were increased by consideration of soil nutrient variability but the increased cost of grid soil sampling resulted in less net return than conventional fertilization and sampling methods. Research Question It is well known that fertility levels often vary widely within a field due to management, soil, and other factors. This study examines the feasibility of taking soil variability into consideration to increase small grain yield in the northern Great Plains. Literature Summary Several recent studies indicate that yield and fertilizer use efficiency may be increased by adjusting fertilizer inputs based on soil and fertility spatial variability. The popular literature presents many hypothetical situations where variable fertilizer application should increase or maintain yield while using the same amount, or a smaller amount, of fertilizer. Study Description Three trials were conducted on production fields in eastern North Dakota from 1989 to 1991. The soils used developed from till in a sub‐humid climate. Five to nine soil map units were present in each field. All soils were very high in K and medium to high in P. Nitrate‐N in the top 2 ft of soil varied from 20 to over 200 lb/acre at planting time. In all years, growing season rainfall was adequate to very good. The crops grown were ‘Robust’ barley (1989 and 1990) and ‘Grandin’ hard red spring wheat (1991). Treatments in 1989 and 1990 were: Control (no N fertilizer applied). Conventional method (composite soil sample and a field yield goal). Variable method (soil sample on a 50‐ft grid and a field yield goal). Variable method (grid soil sample on a 50‐ft grid and estimated productivity of each soil map unit). In 1991, treatments consisted of taking soil samples on a 50‐ft, 150‐ft, or 250‐ft grid in order to determine the most efficient grid size. In addition, some of the 1989 and 1990 treatments were used in 1991. Applied Questions Can crop yield be increased by using grid soil sampling and yield goals for each soil mapping unit? In 2 of the 3 yr of this study, grid soil sampling and N fertilizer application based on a 50‐ft grid increased yields. In 1 of the 3 yr (1991), the greatest yield was obtained using a combination of grid soil sampling and yield goals for each individual soil map unit. In 1990, the best yield was obtained by using grid soil sampling and a single yield goal for the whole field. Were net returns improved by taking into consideration the spatial variability of fields? In 1990, the most profitable treatment was N fertilization based on a composite soil sample and a yield goal for the field, which is the conventional method used by farmers (Table ). While N fertilizer application based on grid soil sampling increased yield, it increased costs more than returns. As the grid size decreased from 250 ft to 50 ft, the number of samples increased from 0.7 to 17.4 samples/acre. The most profitable treatment in 1991 was N fertilization based on a composite soil sample and a yield goal for each soil map unit (Table ). Influence of soil sampling grid size on yield, costs, and net returns per acre. Sampling grid Soil samples Sampling and testing costs Barley yield Net returns ft samples/acre $/acre bu/acre $/acre 1990 None Composite 0.54 67.7 161.94 50 17.4 65.00 73.7 111.88 1991 None Composite 1.95 36.2 121.13 250 0.7 2.78 34.1 113.16 150 1.9 7.73 36.5 116.37 50 17.4 65.00 37.3 61.82 Barley @ $2.40/bu and wheat @$3.40/bu. Cost of fertilizer and additional spreading cost not included. Based on one composite soil sample of 20 subsamples per soil map unit.