Introduction
Feed ingredients, such as barley and wheat, vary substantially in quality. If the variation in ingredient quality is ignored, the quality of finished feed may vary, resulting in changed feed intake and/or reduced growth performance. Feed ingredient quality is thus a nutritional and economic concern in the pork industry, especially the quality differences that are related to energy intake or protein deposition. Feed ingredient quality is related to its chemical composition and not to physical characteristics (bushel weight). Thus, ingredient chemical characteristics should be analyzed regularly. Analyses can be considered an investment to ensure a predictable growth performance.
Variation in ingredient quality has become increasingly important for the pork industry, because minimizing the differences between actual and calculated quality of finished feed helps to achieve a predictable performance. Diets are formulated using least-cost diet formulation, and safety margins have been included to guarantee minimum dietary nutrient levels. These margins could be reduced if ingredient quality is monitored properly. Analyses or predictions of nutrients with the greatest impact on diet cost or performance (energy, amino acids) is the most effective way to manage variation in ingredient quality, and will likely provide a high return on investment.

Variation in feed quality
Feed ingredient quality is now defined as digestible nutrient content rather than total nutrient content. Most swine nutritionists in western Canada use digestible energy (DE) and digestible amino acids to describe ingredient quality. The nutritional value of ingredients is a function of their digestible nutrient content and the pig’s voluntary feed intake (VFI).
For cereal grains, DE is the most valuable nutrient. Variations in the DE content of western Canadian wheat (range of 630 kcal/kg) and barley (range of 450 kcal/kg) can be large, due to changes in the grain’s energy digestibility. In the last decade, we have related changes in the DE content of wheat and barley to changes in their fibre or non-starch polysaccharide (NSP) content (Zijlstra et al. 1999; Fairbairn et al., 1999). Based on the relations, we have developed equations that predict the DE content for barley and wheat based on their chemical characteristics.

Prediction of wheat DE. Our best equation for the prediction of wheat DE content on a dry matter (DM) basis was based its fibre (neutral detergent fibre; NDF) and protein (CP) content:
DE (DM) = 3,584 + 38.3 x CP (DM) – 16.0 x NDF (DM) (R2 = 0.75)

Prediction of barley DE. Our best equation for the prediction of barley DE content on a DM basis was based on its fibre (acid detergent fibre; ADF) content:
DE (DM) = 3,918 – 92.8 x ADF (DM) (R2 = 0.85)

Use of prediction equations. Theoretically, prediction equations are only valid for the range of chemical characteristics used to create the equation. Some wheat samples from the 2002-harvest have a unique combination of chemical characteristics (CP and NDF) that have not been measured before. Thus, the existing equations to predict the DE content of wheat based on CP and NDF may not be valid for a large portion of the 2002-harvest. The 15 wheat samples used to create the original prediction equations (¨) have a lower CP and NDF contents than some samples from the 2002-harvest (à) (Figure 1). Thus, the existing prediction equations should be used cautiously. Barley and wheat samples from the 2002-harvest should be analysed for their chemical characteristics and DE content in grower pigs to validate and improve upon these prediction equations.

Origin of variation in DE. Barley and wheat available in western Canada vary substantially in their nutritional value. For example, the DE content of 20 barley samples ranged from 2,700 to 3,100 kcal/kg, a range that cannot be explained by the samples’ gross energy (GE) content (Figure 2). The range in wheat or barley DE content is therefore mostly due to a change in energy digestibility. A current hypothesis is that a reduction in DE content or energy digestibility is partly due to higher fibre levels. Consequently, low quality grain samples might therefore be more responsive to treatments that enhance energy digestibility.

Voluntary feed intake. Apart from ranges in DE content, a range of feed intake has been described for Australian wheat in weaned pigs (Cadogan et al. 2000) and western Canadian wheat in broiler chicks (Swift et al. 1998). A range of energy digestibility exists for western Canadian wheat (Zijlstra et al. 1999); however, the range in energy digestibility is not related to changes in feed intake in weaned pigs (Cadogan et al. 2000) or broiler chicks (Swift 1998). The existence of a range of feed intake and energy digestibility in western Canadian wheat fed to weaned pigs has not been confirmed to date. A recent study using 12 Canadian wheat varieties organized into 2 varieties in 6 wheat classes did not show a range in feed intake or growth performance for weaned pigs. However, the wheat samples included in this study did not differ much in fibre content and thus quality (Zijlstra et al. 2003).

Economic implications of ingredient quality
The economic implication of variation in feed ingredient quality was calculated for two scenarios.
Scenario 1. A change in quality will change the economic value of the ingredients. For wheat, the economic impact of a 5% change in the digestible lysine (dLYS) or DE content was calculated for grower pigs. Lysine-HCl (78% dLYS; $2.45/kg) and canola oil (8,800 kcal DE/kg; $0.89/kg) were considered as purified sources of dLYS and DE. Assuming that wheat has a mean nutritional value of 0.29% dLYS and 3,425 kcal DE/kg, a 5% difference in these nutrients will be 0.15 kg dLYS in a tonne of wheat and 171 kcal DE in a kg of wheat. The economic value of the 5% difference is per tonne of wheat is:

For dLYS: (0.15/780) * $2,450 = $0.47/tonne
For DE: (171/8,800) * $890 = $17.29/tonne

The higher economic value for the change in DE compared to digestible lysine verifies that the greatest cost-pressure with least-cost diet formulation is against dietary DE content (Zijlstra et al. 2001).

Scenario 2. If changes in energy content are ignored, pig performance may be affected. An unaccounted for reduction in dietary DE content and thus DE intake may reduce gain if feed intake is not increased, or may reduce feed efficiency if feed intake is increased to maintain the DE intake. The following estimates of costs were made for a 7% reduction in DE intake, using data from an energy intake study.

A reduction in gain of 70 grams per day results in pigs that are 8 kg lighter at slaughter after 16 weeks in the grower-finisher barn. The lower body weight at marketing would result in a loss of $10.56 per pig sold, assuming a market price of $1.50/kg and an average index of 110.

An increase in feed intake of 7%, to compensate for the reduced dietary DE content, may increase feed conversion by 0.2 kg feed per kg gain, which would increase feed costs by $2.95 per pig sold.

Change in diet formulation
Reaching a predictable performance is important to maximize net income. But to reach a predictable performance the actual diet DE content should be close to the calculated diet DE content. Thus, diet formulation must be adjusted with a decline in feed ingredient quality. Thus, low quality grains can be used effectively in swine diets, considering:

Ingredient DE content has been predicted using chemical analyses (or perhaps NIRS);

The ingredient specifications have been altered during least-cost diet formulation;

The diets have been re-formulated to ensure that the dietary DE content required for a predictable performance is met. For example, if the predicted barley DE content is lower than the average book value for barley DE, additional energy in the form of canola oil or tallow should be added to the diet to ensure that that actual dietary DE content is at the desired level.
Only then can the grower-finisher pig achieve the energy intake required to maintain protein deposition.

Value-added processing
As discussed previously, the variation in barley or wheat DE content is mostly due to a change in energy digestibility and not a change in gross energy. A decrease in energy digestibility has been linked to an increase in fibre content (Zijlstra et al. 1999; Fairbairn et al., 1999). Therefore, ingredient or diet processing techniques that increase energy digestibility might add nutritional and economic value to low quality wheat or barley. Value-added processing of low quality ingredients might include reducing particle size by grinding more finely (Oryschak et al 2002) or supplementation of fibre-degrading enzymes (Zijlstra et al. 2000).
Regular grinding results in a mean particle size of 700 to 900 microns; however, there may be merit in reducing the particle size of low quality grain to 600 microns or below. Due to the higher fibre content in low quality wheat or barley, supplementation of enzymes to digest fibre might be more beneficial in ingredients with a low or normal DE content compared to ingredients with a high DE content. The impact of value-added processing should be validated using the 2002-harvest, although some indication can be given for the expected improvement in DE content using value-added processing (Zijlstra et al. 1998),

Summary
The 2002-harvest in western Canada was poor. The amount harvested was below average and contained a large proportion of low quality grains. This manuscript focused on determining wheat and barley quality and the possible actions to take after identification of low quality grain. Possible actions include changing diet formulations and/or using value-added processing.

Implications
The quality of low quality grains, especially their DE content, may be predicted using equations based on chemical characteristics. The existing prediction equations should be used cautiously for now, and be validated using the 2002-harvest. Changing diet formulation might allow the use of low quality ingredients by introducing a larger portion of energy-providing ingredients. Value-added processing might further increase the effective use of low quality wheat or barley in swine diets, because their DE content will be increased and a lesser amount of expensive energy-providing ingredients will have to be included in the diet to meet the desired dietary DE content. Low quality wheat or barely may form a large proportion of the diet; thus, accurate prediction of their DE content is important. Further flexibility may be introduced by reducing the proportion of low quality grains in the diet and by including a number of other ingredients.