Sign up for email alerts of new Fluid Journal issues!
Fluid Journal : Summer 2015
7 The Fluid Journal Summer 2015 of precipitation during the crop growing season. Field. The 22-acre field is located at Colorado State University’s Agricultural Research Development and Education Center in Fort Collins, Colorado. Soil at this site is classified as a fine- loamy, mixed, super-active, mesic Aridic Haplustalf. Based on soil samples, texture was classified as a sandy clay loam. Slope. Field slope is lower than two percent in a single plane gradient. Site history is one of continuous maize production for ten years with conventional tillage. Population. Corn hybrid Dekalb 4620 was planted at a population of 20,000, 27,000, 34,000, 41,000, and 48,000 plants/A (depending on plant population treatment strips) on April 29th 2014. Seeds were planted using a precision planter in long strips crossing the entire field. The sequence of the population strips was randomly assigned. Fertilizer. Monoammonium phosphate (100 lbs/A of 11-52 -0) was applied in early spring (April 1), followed by a fluid fertilizer (UAN 32%) rate of 75 lbs of N/A at planting and 150 lbs of N/A at growth stage V8 of the crop. Irrigation was supplied with a sprinkler irrigation system to compensate crop evapotranspiration, using the web-based irrigation scheduler eRams (www.eRams. com). Vacuum planter. Plant population targets were programmed in a 6 rows Monosem (NG+3 Series) precision vacuum planter. This planter is equipped with sensors that monitor the actual seed rate at every location of the field and create an “as-applied” map of plant population (Figure 2). This map was used rather than the target map to analyze the data. Harvesting. Corn was harvested on October 30th at corn maturation with a 6-rows Case IH combine harvester equipped with a yield monitor. Data analysis Cleaning. Yield data were cleaned to remove outliers using an algorithm rejecting all data above and below the average plus or the average minus three times the standard deviation. Cluster. Yield was clustered in two classes based on productivity potential. Groups. The low group was below average, while the high group was above yield average. Quadratic function was used to model the relationship between yield and plant population, forcing the intercept to 0 on the basis that at plant population zero the yield has to be zero. The maximum of the quadratic function was considered as the optimal plant population to maximize yield. Soil properties were monitored at the location of each yield data point using geographic information system software. The soil properties investigated were: • Percent sand, clay, and organic matter • Cation exchange capacity • Soluble salts, nitrates, phosphorus, potassium, magnesium, and calcium. T-test. A Student’s t-test was used to compare soil properties between the low and high yield data with a level of significance of 0.05. Results Grain yield ranged from 100 to 215 bu/A. Quadratic relationship between plant population and yield was strong for the whole dataset, which is consistent with observations in other studies on the effect of seed rate on yield. Applied plant population (1000 pl/ac) 16-23 24-30 31-35 36-40 41-45 46-49 50-55 Figure 2. Applied map of the plant population. “Seed rate potent way to measure variability in soil properties.” Figure 1. A field showing spatial variability in soil properties as measured by grid soil sampling.