What's Cropping Up? Blog

Articles from the bi-monthly Cornell Field Crops newsletter

Precision Ag Project Examines Topographical and Soil Nutrient Influences on Yield

Authored by: Savanna Crossman, Research Coordinator, CCA
Statistical Analysis by: Margaret Krause, Cornell University PhD Student

*This article is part of an ongoing series.  Previous articles can be found at www.nycornsoy.org/research*

Model Validation Underway in NYS

This Precision Ag Research Project debuted its variable rate seeding prescription model on over 500 acres of corn and soybean this spring.  The prescriptions were designed to maximize yield using each field’s specific soil, topographical, and historical characteristics.   Though this project is focused on using the highest resolution grid management, the data can also be used to guide growers at various levels of precision ag adoption.

Field Level Management

Growers in the Northeast know the challenges of farming the highly varied soils of the region.  It is not atypical for one grower to be managing loams, mucks, and gravelly soils across their farm or even across one field.  Loamy ground is traditionally treated as the best ground, where growers might try to drive up seeding rates to increase their yields.  Other, less productive fields, are usually sown with lower seeding rates as an insurance policy against flooding or other expected hindrances to yield.

Figure 1. The yield of P0216 for Field 8 at the four experimental seeding rates.

Figure 1. The yield of P0216 for Field 8 at the four experimental seeding rates.

The data collected over the past four years of this project are beginning to suggest that this traditional knowledge does not always hold true.  For example, Field 8 located in Sackets Harbor, is considered by the grower to be highly productive land.  The Pioneer recommended seeding rate for the hybrid P0216 is 30,800 sds/ac in order to reach the typical yield in this field.  In 2015, this field yielded the same at 27,000 sds/ac as it did at 42,000sds/ac (Figure 1).

In this field, driving up the seeding rate did not result in driving up yields.  If the grower had planted the entire field at 27,000 sds/ac instead of the recommended rate, the grower would have saved $13.30/acre in seed cost or about $1000 in this 75 acre field.

This emphasizes the importance of knowing the productivity of the land and managing each field individually.  There are dramatic gains to be made if growers can begin to manage the extreme variations of the landscape instead of the falling back on whole farm averages.  Once a grower has become comfortable with this type of management, thinking at a subfield level will yield a further economic advantage.

Grid Level Management

Grid level management is the highest resolution of management being used in the Northeast.  This involves using a combination of precision ag technologies such as grid soil sampling, multi-hydraulic drive planters, and RTK GPS.

Figure 2. As-planted data layer displaying elevation acquired via RTK-GPS (field: SW 1).  Credit: M. Krause

Figure 2. As-planted data layer displaying elevation acquired via RTK-GPS (field: SW 1). Credit: M. Krause

For this case study, the analysis is focused on a 117 acre field in Union Springs, NY.  The field, SW1, exhibits large variations in the topographical landscape with elevations ranging from 512 to 638 feet above sea level and slopes as high as 14.3 degrees (Figure 2).  Given these known factors, it is expected that topography will have a significant impact on yield as well as the variation in soil characteristics.  To investigate these hypotheses the field was sown with a split planter containing the corn hybrids P0216AM and P0533AM1 in 2015.  In the same year, fourteen soil characteristics were measured and geo-located on ½ acre grids across the field.

The hybrid P0533AM1 planted at 27,000 sds/ac resulted in the highest average yield while P0216AM planted at 42,000 sds/ac resulted in the lowest average yield.  However, the planting rate was only able to explain 3.1% of the variation in yield here.  This serves as strong indication that other factors in the field are contributing to the variation in yield.

As previously mentioned, elevation and slope were expected to have a signification effect on variation in yield.  In order of importance, slope, elevation, and curvature exhibited significant correlations with yield and were able to explain 15.5% of yield variation.

The model then added the fourteen soil characteristics to the topographic factors and was able to explain 45.6% of the yield variation.  The typical nutrients and soil properties, organic matter, P, K, Ca, and Mg, that are known to support corn growth exhibited significant positive correlations with yield.

Figure 3. Correlations between soil characteristics evaluated via grid soil sampling and topographical characteristics at SW 1.

Figure 3. Correlations between soil characteristics evaluated via grid soil sampling and topographical characteristics at SW 1.

The strongest relationships indicated that higher levels of magnesium (r=0.54) and pH were found in areas of higher elevation, most likely due to shallower depth to bedrock (Figure 3).  The bedrock is more susceptible to weathering and releases nutrients as it broken down by physical disintegration and chemical decomposition.  Given the soil type and location, it is likely that the bedrock in this area is limestone dominated by the dolomite mineral (CaCO3 × MgCO3).  During the weathering processes, the bedrock releases magnesium and calcium, as well as other minerals, into the soil.  Given their stronger charge than other cations, Mg2+ and Ca2+ take the place of other nutrients on negatively charged soil particles.  This results in a more basic pH and higher Mg and Ca concentrations on hilltops and side slopes where the depth to bedrock is lower.

Most other soil nutrients were found in lower concentrations in areas of high slope or elevation.  This aligns with the general knowledge that soil, and thus soil nutrients, can be easily eroded away from the tops of hills and along steep hillsides.  Not surprisingly, phosphorus was the most prone to exhibit a negative relationship with slope.  Phosphorus can easily be lost from the soil profile from water runoff or soil erosion. (Figure 3)

High ground and hillsides tended to yield the highest in 2015, which was expected given the above average rainfall throughout the year.  It is likely that water was pooling in the low flat areas of the field, negatively impacting root growth and nutrient uptake, thus accounting for lower yields.  When compared to a year with dry to average rainfall, it would not be surprising to see these relationships switch.

Figure 4. The distribution of planting rates predicted to be optimal at SW1.

Figure 4. The distribution of planting rates predicted to be optimal at SW1.

Relationships such as these that are heavily driven by climactic conditions make multiyear testing and historical data of the utmost importance when creating a variable rate seeding prescription.  Based on the data available, the model generated a seeding prescription that calls for much lower seeding rates than the grower’s typical practice of 35,000 sds/ac (Figure 4).  This prescription is estimated to save the grower approximately $1800 in seed cost without sacrificing yield, however, the 2016 research harvest will test the true performance of the model.

The extreme variability of the Northeast has presented challenges to growers for decades, and finally the research and technology is here to capitalize on this variability.  The data analysis has already begun to call into question agronomic theories that were once considered fact.  Most notably, the demonstration that the most productive ground can support equally high yields at seeding rates four to five thousand below standard practice.   The analysis emphasizes how yield is driven by the interaction between many factors such as elevation, slope, and soil nutrients.  SW1 demonstrated how topographic factors can drive yield in field and also how it can influence the spatial distribution of soil nutrients.  As the project begins to test these relationships through model validation, it remains critical that the geographic reach of testing locations continues to expand.  Any interested parties are encouraged to contact Savanna Crossman at 802-393-0709 or savanna@nycornsoy.com .

 

Comments are closed.

Subscribe

Follow this blog

Get a weekly email of all new posts.

Skip to toolbar