In-16FEvaluation of Site-Specific Management for Indiana Soybean Production Systems
Principal Investigator: Sylvie M. Brouder, Agronomy Department, PurdueUniversity
Objectives
The overall objective of this project is to provide scientifically sound information to Indiana soybean producers on the use of currently available and prototype precision agriculture technologies for plant nutrient management. Specific objectives include
1)Determining the spatial structure of plant nutrient status and yield and characterizing the relationships among these plant parameters and the soil chemical and physical parameters;
2)Examining the adequacy of 2.5 or 5 acre grid sampling schemes as well as various zones strategies for assessing soil chemical properties in fields with pronounced soil heterogeneity, including
a)Comparing the optimum sampling and interpretation/interpolation strategies for assessment of P, K, organic matter, pH, SMP buffer pH and lime requirement for the purposes of managing fertilizer inputs.
b)Comparing novel approaches to soil fertility characterization (e.g. resin-based technology, prototype sensors for on-the-go pH and K determination at field moisture content, alternative laboratory extraction protocols) to standard soil collection/extraction protocols for predicting the spatially-variable nutrient needs in a field;
3)Demonstrating to Indiana soybean producers, agri-business personnel and crop consultants of the relative merits of site-specific versus whole field management using established fertilizer recommendation, currently-available technologies and best management protocols for input, and
4)Creating and distributing educational information on precision technology selection and decision support to Indiana soybean producers.
RELATED PROJECTS IN 2002
This USB/PPI-FAR supported project has been used to leverage funds to support additional research on the core objectives outlined above and to address other applied research questions related to the implementation of precision technologies to enhance productivity and/or profitability while protecting the environment (See Appendix A3). In 2002, these projects included:
- Geospatial Technology and Corn Nitrogen Management: Evaluation of Production and Environmental Benefits. Specific objectives are to
- Determine whether current university recommendations for N management applied to management zones within a field enhance N use efficiency when compared to university recommendations for whole field N management. Management zones will be delineated based on a previously-collected, extensive field database that includes soil type, organic matter, topography and drainage, yield maps and remotely sensed images, and geo-referenced crop nutrient removal.
- Determine the accuracy of post-harvest N deficiency/sufficiency maps drawn from the integration of yield monitor data with grain quality monitoring data collected with near infrared reflectance sensors.
- Profitability of Variable Rate Lime Application: Comparative Assessment of VR and Whole Field Management. Specific objectives are the
- Characterization of the best sampling strategy and intensity to describe manageable, infield variability in soil pH,
- Comparison of soybean yields and profitability of VR to “whole field” application of lime in selected fields with pronounced variability in lime requirement, and
- Creation of educational materials and decision aids for Indiana soybean producers to assist them in making profitable technology choices in lime management.
General Materials and Methods
The experimental design for the primary field study followed the general protocol of the regional program. The main study site was located at Davis Purdue Agriculture Center (DPAC) in east central Indiana. Two fields, Field R (35 Acres) and Field V (40 acres), were selected based on preliminary soil test data collected in 1995 that indicated that within field variability for several soil factors were significant. According to the published soil survey, both fields are dominated by a Pewamo silty clay loam soil but contain significant areas of Blount silt loam (Fld R: 39%; Fld. V: 30%; Fig. 1) and Glynwood silt loam (Fld. R: 10%; Fld. V: 8.5%). These fields are conventionally tilled and were planted to corn in 1995, soybean in 1996, corn in 1997 and 1998, soybean in 1999, and an annual corn soybean rotation thereafter. Yield monitor data is available for the 1996 through 2002 crops.
For the purposes of objective 1a (above), a historical, spatially intensive soil chemical properties database collected from seven contiguous fields at the Northeast Purdue Agricultural Center (NEPAC) were added to this study. The NEPAC fields (total of 120 acres) are in the Boyers-Shoals-Kalamazoo association and the Morely-Rawson or Morely-Glynwood association. Also, in fall 2000, soil samples were collected from two additional field locations at DPAC. Fields D (31 acres) and F (15 acres) are also in the Blount-Pewamo soil association. The soil association and field sampling information for all locations are presented in Table 1. At NEPAC, only limited cropping history information are available for the years prior to sample collection and no yield monitor data are available for these fields. For DPAC, Fields D and F, cropping history (primarily corn-soybean rotation) and inputs records are available and yields have been collected with a yield monitor since 1995.
Soil Sampling and Site Assessment:
Composite Soil Samples: Fields R and V were divided into half-acre grid units and a stratified, systemic, unaligned pattern was used to select the location of soil sampling points. A half-acre grid was selected to enable us to evaluate the efficacy of less dense sampling strategies that are being used commercially (1-acre, 2-5-acre, sampling by soil type or zones delineated by other data layers). There were 61 sampling points in Field R and 72 sampling points in Field V. Soil samples were collected on April 24, 1998. Fields D and F were sampled on a regular 0.25-acre grid pattern (center point sample location) in April 2000. Ten core composites of 0-4 inches and 4-8 inches were collected at each sample point from a 300 sq. ft area centered on the grid-point location. The NEPAC historical dataset was collected in 1992 (0-8 inch sample depth/ 15 core composite) on a 0.25-acre grid pattern (center point strategy). Collected soils were analyzed by A and L Great Lakes Laboratories, Inc. for the following analytes:
Organic matter, P, K, Mg, Ca, pH, buffer pH, CEC, and percent base saturation for K, Mg, Ca, and H.
Order 1 Soil Survey: An Order 1 Soil Survey was completed in fall 2001 on Fields R and V. An Order 1 Soil Survey is also available for all fields at the NEPAC location. This intensive survey has not yet been conducted on DPAC Fields D and F. The preliminary report on the DPAC R and V Order 1 Soil Survey was reported in the November 2001 Newsletter of the PurdueUniversitySiteSpecificManagementCenter (
Plant Root Simulator Probes:In a collaborative effort with IMC AgriBusiness and Western Ag Innovations Inc., Plant Root Simulator (PRS) probes were also used to quantify soil nutrient availability. The PRS probes contain anion or cation exchange membranes encased in a plastic form. The probes are inserted in soil at field moisture, allowed to equilibrate for 24 hours, and then retrieved and extracted in the laboratory. In both fields V and R, we inserted 5 pairs (a distance of 2 to 4 inches separated a probe pair) of cation and anion probes at every sample location. At ten locations in Field R, we inserted 10 pairs of probes, and at probe retrieval we collected individual soil cores (0-4in and 4-8in) from between the cation and anion probes of a pair.
On-the-go pH Sensing:Between the 1998 and 1999 crops, a small section of Fields V and R (7 acres) were used in a preliminary field verification trial of a prototype for an on-the-go soil pH sensor. The prototype used the Sensorex 450CD electrode and an automated soil sampling system that puts the sensor in contact with the soil. The soil sampling mechanism was developed as the Master of Science project of Viacheslav Adamchuck, a student of Dr. Mark Morgan in the Purdue Department of Agricultural Engineering. The duration of a sampling cycle was 8 seconds, permitting a sample density of 1 sample every 12 yards when the tractor was travelling at 3 mph. Laboratory evaluations of the pH electrode and of a similar, ion selective K (ISK) electrode are reported below. The ISK electrode is a novel approach to the routine assessment of K bioavailability regardless of whether it is used in the field or laboratory setting. Thus, the method itself required evaluation for repeatability and reproducibility to develop a performance objective for measurements.
Treatment Assignment and Fertilizer Application:
Treatment plots were established (Fields R and V only) between the 1998 and 1999 growing season. Sets of 90 x 300 ft grids units were identified in each field. Three adjacent grid units were grouped together as a replicate. Treatments within a replicate included a control (no P, K, or lime addition), a whole field treatment (P, K, and lime additions based on the mean soil test values for the whole field), and a variable rate treatment (P, K, and lime additions based on soil test levels at the core of the plot to be treated; Figure 2). The P and K applications were made in the spring of 1999. Lime applications were made in spring 2001.
The lime application included an in-field study of equipment performance both along and across the application swath. Equipment calibration involved numerous off-site passes where adjustments were made in the rate, spinner speed and baffle setting of the spinner spreader. (For a description of the field facilities and the equipment available at the DPAC, see the farm website at Finally, in order to assess the precision of the application equipment on-the-go, both across and along the application swath, 8 plots (either VR or WF treatments) were selected for concentrated characterization of lime application rate. Within each plot, 9 kitty litter pans were placed across the path of the lime applicator (spinner spreader) at 25, 50, and 150 ft from the georeferenced point where a rate change occurred (Figure 1). The lime collected in each pan was weighed and reserved for analysis of particle size distribution.
Figure 1. Diagram of lime application study treatment assignment (a) and locations within a treatment plot where 9 kitty litter pans were placed across the applicator pass.
Repeat Soil Sampling in Spring 2000 (Field R and V only):
In the spring of 2000, soil samples (0-4 and 4-8 in depth) were collected at the geo-referenced center points of each of the P/K fertilizer treatment plots as well as from the original half-acre grid sample locations. These samples were analyzed as described above.
Plant Sampling and Yield Monitor Data:
In Fields R and V, chlorophyll meter readings were made at greensilks on earleaves (1998, 2000) and at R1 on fully expanded soybean trifoliates (1999, 2001) on ten plants. Earleaves and trifoliates were then collected for analysis of nutrient content. In all years, these sampling activities were conducted at each of the original soil sample points from the half-acre grid. In 1999 - 2001, after the establishment of treatment plots, both leaf samples and 10 whole-plant composite samples were also collected from the center of each of the treatment plots.
At harvest, a yield monitor was used as fields were combined. In 1998, a subsample of grain was pulled as the combine crossed each of the original soil sample points. In 1999 -2001, the sub-sample was pulled from the grain stream as the combine traveled up the center of each of the fertilizer treatment plots. Three separate subsamples were collected from each treatment plot. In 2000 (soybean) and 2001 (corn), grain samples were collected from the soil sampling locations in Fields D and F. The grain samples are being analyzed for total nutrient content (digestion and extraction) and for grain quality parameters (protein and oil; NIR). Both critical level/sufficiency range and M-DRIS approaches will be used to evaluate the plant tissue data. Yield monitor data from this year as well as from all previous years since 1995 will be analyzed for spatial structure and yield stability through time and space.
Equipment:
For a description of the field facilities and the equipment available at the DPAC, see the farm website at
Results and Discussion
The focus of this year's field activities and data analysis has been on
- Completion of the characterization of spatial structure in soil test pH, SMP Buffer pH and lime requirements (LR) in eleven farm fields on two farms,
- Completion of the comparative evaluation of soil collection and interpolation strategies for the purposes of making profitable lime applications,
- Examination of equipment performance for VR lime applications, and
- Development of a protocol for using the ion-selective K electrode to rapidly measure solution-phase K in the laboratory.
- Completion of the characterization of spatial structure in soil test pH, SMP Buffer pH and LR. (Preliminary results reported last year.)
For all fields in the study, calculated whole-field average pHs indicate that a whole-field area composite sampling strategy might not have identified a need for lime. Arithmetic field means ranged from 6.26 to 6.66 while logarithmic means were from 0.15 to 0.30 pH units lower (data shown last year). The values for all individual fields exceeded 5.8, the recommended critical value for identifying lime need when subsoils are not strongly acid. Reviews of crop response to lime suggest maximum yields for corn and soybean in the Midwestern United States are attained at pHs of approximately 6.6 but yield depressions associated with lowering pH to 6.0 are considered minimal (< 1.5% and < 10% for corn and soybean, respectively). For the purposes of these analyses, we assumed subsoil acidity was not restrictive and a pH of 6.0 would optimize productivity for corn and soybean.
Within field variation in pH was substantial, however, and all fields contained areas where lime application would be recommended to optimize crop productivity as well as areas where soil pH neared or exceeded neutrality and lime application would be considered not only unnecessary but potentially undesirable. For a given field, the percent of observations with pH ≤ 5.8 ranged from 0 to 21. As soil pH levels decrease from 6.0 to 5.5 yield reductions can be anticipated to increase marginally for corn (3%) but more dramatically for soybean (20%). Thus, many of the study fields represent management units where it would be theoretically logical to consider VR lime. A majority of the fields had over a quarter of the sample points requiring < 1.0 Mg ha-1 of agricultural limestone while > 4 Mg ha-1 were required for another quarter of the sample locations.
All fields exhibited some degree of spatial autocorrelation as identified by semivariance analysis. Anisotropic analysis revealed no strong, direction-dependent trends in the data and therefore all reported data are for isotropic variogram models (data shown last year. When 0.1 ha and 0.2 ha (0.25 and 0.5 acre) grid data were examined, relationships between semivariance of pH or LR and inter-sample distance could be satisfactorily fit with either an exponential or spherical model (Table 1). Spatial structure accounted for between 29 and 73% of the observed variation. This percentage assumes the variance at the minimum intersample distance represents the nugget effect. Nuggets (y-intercepts) estimated by modeling were one-tenth these values but, since no samples were collected at distances closer than 30-m, the accuracy of these modeled values cannot be confirmed. Early work in Indiana farm fields on core-to-core variability over short distances (15 cores evenly spaced over 5 m transects) reported variances ranging from 0.02 to 0.17 suggesting true nuggets are likely higher than our modeled values.
With one exception, we found range parameter values (95% of the sill value for exponential models) for pH and LR were less than 100 m (330 ft), results in agreement with Mueller et al. (2001) who examined similar soil types. Thus, our data support a general recommendation against point sampling strategies using grids > 1 ha (2.5 acres), as they will likely be too sparse to produce useful information on manageable variability in pH.
- Completion of the comparative evaluation of soil collection and interpolation strategies. (Preliminary results reported last year.)