DR. THOMAS J. BRUMM(*2)
DR. CHARLES R. HURBURGH, JR.(*2)
DR. SETH NAEVE(*3)

Quality continues to be an important soybean marketing issue. This report summarizes current knowledge on the following soybean quality topics:

• protein and oil composition of the 2005 US soybean crop,
• the 2005 crop in historical perspective,
• weather conditions affecting the 2005 crop,
• the connection between protein content and amino acid levels,
• the initial results of a study funded by USB to examine foreign material in soybean shipments, and the adoption of AOCS reference methods by GIPSA.

The data and analyses in this report are intended to assist customers in the purchase and use of US soybeans.

Since 1986, Iowa State University (ISU) and the American Soybean Association (ASA) have been surveying the quality of new crop soybean harvests. US soybean producers, representing 30 soybean production states, in response to a mailed request, provided samples of 2005 crop soybeans for analysis.Samples received by November 4, 2005 were analyzed for protein and oil contents using an Infratec near-infrared instrument (Foss North America, Eden Prairie, Minn.). The standard deviation of the Infratec instrument using IowaStateUniversity calibrations relative to wet chemistry is 0.33% points for protein and 0.32% points for oil, on commodity soybean samples.From other sources, data on the yield and physical quality (US Grade factors) of US soybeans have been collected.Data were organized by state and region (groups of states).Weather data for the 2005 growing season were collected to demonstrate the impact on soybean composition.

The United States produced 2.967 billion bushels (80.9 million metric tons) of soybeans in 2005, according to the October 12, 2005 USDA production estimates (USDA, 2005).This is a decrease of five percent from 2004, the largest US soybean crop on record.The average soybean yield was 41.6 bushels per acre, down from 42.2 bushels per acre in 2004.An estimated 71.3

1/ Prepared for the American Soybean Association Quality Mission to Asia, November 14-21, 2005
2/ Assistant Professor and Professor, respectively, Agricultural and Biosystems Engineering Department, Iowa State University, Ames, IA50011
3/ Assistant Professor, Department of Agronomy and Plant Genetics, University of Minnesota, St. Paul, Minnesota5510million acres (28.8 million hectares) of soybeans were harvested, a 3.6% decrease from 2004.Table 1 summarizes production statistics for the 2005 crop, by state and growing region.Production decreases occurred in every soybean growing region, except the Western Corn Belt.

Composition data are given in Table 2.Average US protein and oil contents for 2005 were 34.92% and 19.41% respectively (on a 13% moisture basis).The protein content is approximately 0.5 percentage points below, and the oil approximately 0.8 percentage points above, the long-term US averages of 35.38 % protein and 18.65 % oil. The soybeans from the 2005 crop will produce, on average, 42.5 lbs of 48% protein meal and 11.4 lbs of oil per bushel from soybeans at 13% moisture.(This yield of meal and oil assumes that the meal is formulated following the National Oilseed Processors Association (NOPA) trading rules with a residual oil content of 0.5%.)The variability (standard deviation) within states, regions, and the US was approximately equal to the long-term averages for variability.

Weather conditions varied from ideal to droughty across the major soybean growing regions.Planting and emergence were ahead of schedule in the dry areas, somewhat behind average in the wetter areas.The darker areas on the drought map (Figure 1) show the concentration of dry weather in Eastern Iowa and Central Illinois.

Beginning in late September, there were rains across all growing regions, accompanied by abnormally warm conditions for the season.There was an abundance of light and heat to continue photosynthesis at a very high rate late into the season.Protein production (on an acre basis) was established by midseason by production of vegetative tissues, but continued plant metabolism in late fall resulted in more oil and greater yield.Thus on a percentage basis, the protein content was lower. Harvest moisture contents were generally low, with the possible exception of the far eastern growing areas that were extremely wet late in the year.

Soybean yields, which increased steadily in the 1990s, appeared to have stabilized in recent years.The 2004 and 2005 crops, however, returned to the 1990s trend, due to improved genetics and good growing conditions across many soybean growing regions.

Table 3 shows a combination of USDA production and survey composition data.The yield, and protein and oil data is shown graphically in Figure 2. In the 1990s, yields increased by 0.5 bushels/acre/year, with little change in average protein and oil content.In 2005, there was a yield increase and protein and oil levels were not out of the range of previous data.Breeding programs continue to emphasize yield, apparently without creating quality loss.The net result is a steady increase in the production of protein and oil per unit of area (Figure 3).From a consumer perspective, this has meant a steadier, more abundant supply from the same inputs.

Figure 4 shows the long-term variability (standard deviation) of the protein and oil results in the survey.The average long-term standard deviation for protein is 1.50 percentage points, 0.88 for oil.In 2005, the standard deviations for protein and oil were 1.46 and 0.87 percentage points, respectively.The ratio of the standard deviations of protein to oil did not change significantly from 2004 – the relative variability of the two components remained about the same.

GIPSA collects results from Official soybean export inspections (GIPSA, 2005a).Official inspections establish Grade based on a set of physical factors and, on request, will report protein and oil contents. Historical data is given in Table 4.The majority of inspections (>94%) were for US No. 2 soybeans in 2005.There is some evidence to suggest that the average foreign material (FM) content of US exports has decreased – an average level of under 1.5% since 2002, compared to over 1.6% in the eight years previous to 2002.The GIPSA composition measurements (protein and oil) agree with the ASA-USB-ISU Quality Survey data.This means that exports are generally an average of the US production.

The GIPSA data is not separated by export location.The Gulf South ports are generally served by states along the Mississippi-Ohio-Illinois-Missouri river system.Pacific Northwest ports are more likely to receive grain from the states classified as Western Corn Belt in the survey.Export quality at any port will tend to mirror the quality of production areas that serve it.Over the 19 years of the survey, the percentage of US crop produced in the Western Corn Belt has gradually increased from about 40% to over 50%.This shift is toward areas of potentially lower protein content.The previous data showing a constant level of US average protein and oil continues to be a real credit to the US soybean genetics industry when viewed in the context of this geographic shift in production.

The processing chart in Figure 5 shows the combinations of protein and oil content that will produce 47.5% to 48.5% protein soybean meal.Only once (1997) did US soybeans fall to the left of the optimal area, shown by the shading.Soybeans from individual states and regions often fall to the far right, above 48.5% meal, and the US averages are regularly in the middle of this area.

Some soybean processors do not dehull and therefore produce a lower protein meal (e.g., 44%).The processing chart in Figure 6 shows the combinations of protein and oil content that will produce 43.5% to 44.5% protein soybean meal

It is clear that the oil and protein contents of the 2005 crop were at the high and low edges, respectively, of the historical data.The protein was the next to lowest, but the oil was the highest of the 20 year history.The sum (total of oil and protein) is the third highest in the record.This means that, in terms of total outputs, meal and oil, the soybean value, on average, is quite good.

In the early years of the survey, we developed the concept of estimated processed value to track the combined output of all products in economic terms.In the US, the pricing of meal and oil actually favors oil somewhat (a per pound price ratio – oil/meal – of 2.6 compared to the long term average of 2.3).In total value, therefore, the crop conditions that favored oil development likely created a gain over the value that would have occurred had the oil and protein contents been at their long term averages.

On average, the US crop will easily make 48% meal, although, because of the higher oil and lower protein, there will be less meal per bushel than average.Despite the relatively lower protein content, processors that do not dehull should still be able to make acceptable low-protein meal (44%), while enjoying the benefit of higher oil yields.Included in Figure 6 are the values from three states:Minnesota, Iowa and Illinois.Soybeans shipped through the Pacific Northwest (e.g., Minnesota and Iowa) should still be able to make soybean meal with sufficient protein content.Processors, whether dehulling or not, will benefit from increased oil yield without sacrificing meal quality.

Over the last 10 years, IowaState has been studying amino acid profiles as well as crude protein and oil.There are 5 major essential amino acids that US nutritionists use to balance feed rations; lysine, methionine, cysteine, threonine and tryptophan.These are amino acids that swine and poultry need but cannot make for themselves.Figure 7 charts the combined amount of these amino acids, both as a percent by weight of soybeans and as a percent of the total protein.The percent of protein data is an effective measure of protein quality.

The total amount of essential amino acids did not fall sharply in lower protein soybeans; and the percent of protein data show that the protein quality was actually better in the lower protein soybeans.This means that if rations are balanced on amino acids rather than crude protein, meal from lower protein soybeans can be at least as good if not better than meal from higher protein soybeans.The amino acid to protein ratio is not changed by the processing of soybean meal; protein is just concentrated by the removal of oil.

Grain shipping patterns will interact with the composition data in 2005.Soybeans shipped from US West Coast ports are likely from Western Corn Belt states.Freight rates have changed to favor the Pacific Northwest for grain shipments to Asia.At the same time, several processors in the Western Corn Belt have begun programs to give premiums for soybeans that are above average in protein, oil, or the combination of the two.Specifics of these programs have been given in the 2003 and 2004 survey reports.The impact of these programs is to create natural sorting of soybeans; those most likely to receive premiums tend to go to the local processors offering the incentives, and the rest are available for sales to other buyers (including exporters) that are not offering incentives.Western Corn Belt processors began these programs because of the historic patterns of lower protein.From an import buyer’s perspective, the likely trend will be a very gradual shift in export soybean quality away from those traits that are being given premiums domestically.Similar incentives at export may be needed to send a unified signal to plant breeders and producers that all soybeans need improvement in composition as expectations in domestic and world markets rise.

Foreign Material Study

The United Soybean Board is funding a study at IowaStateUniversity to examine the amount and type of foreign material (FM) in US soybeans.Some anecdotal evidence suggests that the composition of FM has changed in the last ten years due to the advent of Roundup Ready® soybeans.Similar studies have been conducted in the past, most recently in 1991.This study is designed to build on those studies, updating them for current market conditions.Final results are expected in December of 2005.

The project objective is to analyze the level and composition of foreign material at various stages in the soybean market channel, from farms to export elevator, to determine if there are cost efficiencies that could be captured by exporting a more competitive (lower foreign material) product. While there are some samples yet to analyze, some trends are becoming evident.

• The broken beans component changes the most of all the FM components.If soybeans are more broken (more splits) they also have more broken beans in the FM and therefore less of other things in the FM.At the same time, the overall FM percentage was not related to the splits percentage, showing that the total FM percentages are probably being managed by the handler.
• Other grains were the second most variable component, followed by weed seed and other components.The weed seed and other components tended to be a more constant share of the FM, and of the total sample, probably being more fixed at the first point of handling.This is consistent with the inbound elevator data, where the major share of FM is material found in the field (dirt, plant parts, weed seeds), with less other grains and broken soybeans in the FM.
• The 2003-2004 export samples had a larger proportion of the FM as other grains and as pods than did the 1991 study samples.Weed seeds dropped as a contributor to FM, and the other components (dirt, plant parts, and other) remained about the same share of the total FM.

The United States Department of Agriculture, Grain Inspection Packers and Stockyards Administration offers protein and oil testing for export soybeans.The GIPSA test is done on every sublot portion (50,000 bushels) of each hold of a ship, and the averages are reported on the certificate, if requested by the buyer.

GIPSA uses an Infratec near infrared analyzer, as does IowaStateUniversity for the ASA Soybean Quality Survey.The GIPSA calibrations were developed for blended export samples to a greater extent than were the IowaState calibrations. GIPSA’s calibration statistics are given in Table 5.The IowaState calibrations were intended to measure individual samples, such as variety trials or grain lots delivered by producers to the first point of sale.The equivalence of the GIPSA and survey averages indicates that performance of the two NIR calibrations is quite similar in their respective application areas.

NIR instruments must be calibrated to match a reference standard.In world trade, the methods of the American Oil Chemists Society (AOCS) are generally considered universal.The IowaState calibrations are based on the following methods and labs:

• Moisture: AOCS Ac 2-41, IowaState (whole grain oven)
• Protein: AOAC 990.03, Eurofins, Des Moines (combustion)
• Oil: AOCS Ac 3-44, Eurofins, Des Moines (ether extract)

The published reproducibility of the protein and oil reference methods is 0.2 % points when done by AOCS-certified analysts (in the AOCS sample proficiency program).

The GIPSA calibrations are based on in-house adaptations of published methods.The following short descriptions are from the GIPSA website (GIPSA, 2005b):

Air Oven is the reference method for moisture in grains and oilseeds. The moisture value is used to adjust the protein and/or oil and/or starch content to a constant moisture basis. A representative sample is weighed and placed in a standardized, forced-draft oven according to the GIPSA standard operating procedures. After the required drying time, the sample is re-weighed and the moisture content of the sample is calculated. The Air Oven reference laboratory supports the moisture meter instruments used for rapid inspection at field locations performing official testing.GIPSA's Air Oven laboratory is ISO 9001:2000 registered. ISO's worldwide acceptance gives customers of GIPSA increased confidence that our analytical results are accurate and reproducible.

The reference method for analyzing the protein content of grains and oilseeds is combustion nitrogen analysis (CNA). A representative ground sample is weighed and the CNA uses heat and pure oxygen to completely combust (burn) the sample. From the resulting gases, the CNA measures the amount of nitrogen and calculates the "as is" protein content of the ground sample. GIPSA's standard operating procedure follows the AOAC International Method 992.23. The CNA laboratory supports the NIRT instruments used for rapid inspection at field locations performing official testing.GIPSA's CNA laboratory is ISO 9001:2000 registered. ISO's worldwide acceptance gives customers of GIPSA increased confidence that our analytical results are accurate and reproducible.

Solvent Oil Reference Method

The reference method for analyzing the oil content of corn, soybeans, and sunflower seeds utilizes the Soxtec solvent oil extraction method. A representative ground sample is weighed and the oil is extracted using petroleum ether according to the GIPSA standard operating procedures. Upon completion of the extraction process, the "as is" oil content of the ground sample is calculated. This method is a modification of the American Oil Chemists Society Official Method Ac 3-44. The Solvent Oil Extraction laboratory supports the NMR and NIRT instruments used for rapid inspection at field locations performing official testing. GIPSA’s Solvent Oil Extraction laboratory is ISO 9001:2000 registered. ISO’s worldwide acceptance gives customers of GIPSA increased confidence that our analytical results are accurate and reproducible.

The AOCS is now running a harmonization project to verify the alignment of reference laboratories used to create NIR calibrations.Through the first year of sample exchange, the GIPSA laboratory has been in excellent agreement with AOCS-recognized labs.One possible result of this project is the submission of the GIPSA lab protocols as permissible and equivalent method modifications in the AOCS methods book.

It is our view that GIPSA protein and oil tests could be used to enforce contracts and premiums at the time of ship loading with the expectation that these results will be consistent with those in markets where the standard is an AOCS method done by proficient analysts using documented quality control techniques.The accuracy of NIR instruments is within the expected variation of the chemical methods across different laboratories and operating conditions.

• The total US soybean production in 2005 was nearly three billion bushels, despite isolated areas of drought.This was 5% lower than last year’s record production.
• The 2005 US soybean crop is lower in protein (34.9%) and significantly higher in oil (19.4%) than previous years.The variability in protein and oil content was similar to the long-term average variability.Processors should be able to meet target meal protein levels and benefit from higher oil yields.
• The US soybean crop, while lower in protein quantity than soybeans from other countries, has better protein quality, as measured by the amount of five essential amino acids.
• The composition of foreign material (FM) in US soybeans has changed since a 1991 study.The percentage of weed seed and dirt has decreased while the percentage of other crops has increased.
• Work is being done to align GIPSA and AOCS methods for the determination of moisture, protein and oil content in soybeans.
  

GIPSA. 2005a.2005 US Grain Exports: Quality Report (Soybeans).USDA Grain Inspection, Packers and Stockyards Administration, Washington, DC. http://www.gipsa.usda.gov/pubs/04ex/04soybeans.pdf/, accessed November 3, 2005.

GIPSA, 2005b.Reference methods and laboratories.USDA Grain Inspection, Packers and Stockyards Administration, Washington, D.C.http://www.gipsa.usda.gov/tech-servsup/metheqp/refmeths.htm, accessed November 3, 2005.

UNL, 2005.US drought monitor map, September 27, 2005.National Drought Mitigation Center, University of Nebraska-Lincoln, Lincoln, NE.http://drought.unl.edu/dm/archive/2005/drmon0927.htm, accessed October 31, 2005.

USDA, 2005. Crop production 2005. US Department of Agriculture, National Agricultural Statistics Service, Washington, D.C.http://usda.mannlib.cornell.edu/reports/nassr/field/pcp-bb/2005/crop1005.txt, accessed October 23, 2005.