Corn co-products in broilers

The expansion of the ethanol industry has increased the availability of a variety of corn co-products as feedstuffs for livestock and poultry. Of the many co-products produced, distillers dried grains with solubles (DDGS) is the most commonly utilized in broiler diets as a source of dietary energy and protein. Recently, advanced biorefining technologies have allowed for the additional extraction of corn oil from DDGS during processing. The removal of oil has the potential to alter both the nutrient profile and energy value of DDGS. Recent studies have investigated the effect of oil extraction on nitrogen-corrected apparent metabolizable energy and standardized ileal amino acid digestibility of DDGS in broilers. The altered nutrient profile of reduced-oil DDGS may impact broiler performance and carcass characteristics. In order to mitigate the negative effects of nutrient variability, accurate values of nutrient content are critical. Prediction equations have been suggested as a means to more quickly and efficiently evaluate the energy value of DDGS. However, the use of these equations in a practical application requires an understanding of the error associated with each model in a prediction setting. Therefore, validation of future prediction models is recommended. Corn co-products are the remnants of the wet- or dry-milling processes used in ethanol production. Wet-milling involves soaking the corn in order to separate its components prior to fermentation into ethanol. This process results in co-products such as corn germ, corn germ meal, corn gluten feed, and corn gluten meal. Dry-milling utilizes the whole corn kernel in the fermentation process, producing distillers dried grains as well as distillers solubles. These co-products may be combined in various blends of DDGS. Because of their high fiber content, these co-products are mostly fed to ruminant animals. However, poultry producers have increasingly utilized corn co-products at low dietary inclusion levels as a cost-effective substitute for portions of traditional sources of energy and protein. Because these feedstuffs are products of different processes, substantial nutrient variation exists both between and within each product type. Such variation may limit the utility of co-products in broiler diets.

Nutrient Content and Variability
The inadvertent miscalculation of nutrient content may create substantial difficulties during diet formulation. The use of artificially low nutrient values in feed formulation results in needlessly increased feed costs through the inclusion of additional sources of each nutrient perceived as limiting in the diet. In contrast, the overestimation of the nutrient content of a feedstuff may result in marginal or deficient diets that have the potential to negatively impact live performance.

Corn Germ and Corn Germ Meal
The germ of the corn kernel is removed during wet milling and may be further processed to obtain corn oil and corn gluten meal. Corn germ contains the majority of the oil in the corn kernel, as well as protein and fiber. Kim et al. (2008) evaluated a dehydrated corn germ sample, which contained 14.9 % crude protein (CP), 19.1 % fat, and had a TME_n value of 4,336 kcal/kg. Rochell et al. (2011) evaluated two samples of corn germ with an average CP content of 16.6 % and a determined average AME_n value of 3,308 kcal/kg.Corn germ meal contains the residual components of the germ after oil extraction is completed. Rochell et al. (2011) determined the AME_n of a single sample of corn germ meal to be 1,991 kcal/kg.

Corn Gluten Feed and Corn Gluten Meal
From each bushel of corn converted to ethanol through wet-milling, approximately thirteen pounds of corn gluten feed (CGF) and three pounds of corn gluten meal (CGM) are produced. Corn gluten feed contains bran, germ meal, as well as dehydrated solubles from the steeping process. Rochell et al. (2011) determined the AME_n of one sample of corn gluten feed to be 1,746 kcal/kg. Corn gluten meal is a combination of corn germ meal and the dried residual components of corn after the starch, germ, and bran have been separated. Despite containing approximately 60 % crude protein (CP), CGM is deficient in lysine, tryptophan, and arginine. Furthermore, the high leucine content of CGM may cause issues with branch chain amino acid imbalances in broilers. Corn gluten meal has been suggested as a good source of available methionine in broilers. Although these corn co-products have nutrient profiles compatible with inclusion in broiler diets, currently they are not commonly utilized in poultry. Approximately 10 % and 32 % of CGF and CGM are exported each year, and the remaining supply is predominantly utilized for cattle feed as well as by the petfood industry.

Distillers Dried Grains with Solubles
Distillers dried grains with solubles are commonly utilized in broiler diets as source of energy, crude protein, phosphorus, and amino acids. The value of DDGS as a replacement for protein sources is primarily dictated by its amino acid profile and CP content. The amino acid profile of DDGS is very similar to that of corn, and is therefore particularly limiting in Lys. Lysine contents varying from 0.72 to 1.02 % (CV=17.3 %) have been reported. Of the essential amino acids, Lys is the most variable in DDGS due to the potential for heat damage during the processing and drying of DDGS. During the heating stages of DDGS production, reducing sugars present in the DDGS and the ε-amino group of Lys may react to form dark-colored Maillard reaction products, reducing the bioavailability of Lys. The CP content of conventional DDGS has remained relatively consistent over time. Cromwell et al. (1993) reported a range from 26.0 to 31.7 % CP (CV = 5.3) among nine DDGS sources, including two from beverage alcohol producers. Belyea et al. (2004) determined a range from 30.8 to 33.3 CP for 235 samples from a single ethanol plant over a five year period. The similarity of these results indicates that the CP content of DDGS is relatively invariable both among and within sources.

In contrast, the carbohydrate composition of DDGS is highly variable. Total carbohydrate values for DDGS typically range from 50 to 60 % on a DM-basis. The majority of the starch content of corn is utilized by yeast during the fermentation process, resulting in DDGS with low starch content and relatively high amounts of non-starch carbohydrates (NSC). The specific compounds that make up the NSC fraction are difficult to quantify due to vast differences in their nutritional, chemical, and physical characteristics. In DDGS, NSC may be quantified on the basis of their solubility in neutral or acidic detergent solutions by the method of Van Soest et al. (1991). Neutral detergent fiber (NDF) is primarily composed of portions of the cell wall, including hemicellulose (HC), cellulose, and lignin. Acid detergent fiber (ADF) contains the residual fiber fractions after the removal of hemicellulose. The carbohydrate compounds quantified by NDF and ADF are relatively indigestible by poultry and have been associated with reductions in ME content of feedstuffs. Stein and Shurson (2009) reported an average starch content of 7.3 % among 46 samples of DDGS with a range of 3.8 to 11.4 %. In these samples, NDF ranged from 20.1 to 32.9 % with an average of 25.3 % and ADF ranged from 7.2 to 17.3 % with an average value of 9.9 %. Meloche et al. (2013) reported starch content ranging from 0.8 to 3.9 % with an average of 2.3 % among 15 DDGS samples. For these samples, NDF ranged from 27.0 to 51.0 % with an average of 35.7 % and ADF ranged from 7.7 to 15.8 % with an average of 11.9 %. These reports further emphasize the extensive variability of fiber fractions in DDGS. In addition to NDF and ADF, NSC may be also be characterized on the basis of their solubility. The total dietary fiber (TDF) system allows for the additional recovery and quantification of soluble cell wall fractions that are lost in the determination of NDF. Total dietary fiber contains both insoluble and soluble fiber components. The former includes structural components such as HC, cellulose, and lignin, whereas the latter includes pectins, arabinans, β-glucans, and resistant starches. Insoluble fiber is largely indigestible by poultry, and therefore acts as a diluent of dietary energy. Soluble fiber compounds have the potential to adhere water, producing viscous gels in the gastrointestinal tract. Increased digesta viscosity obstructs the action of digestive enzymes, reducing the digestion and absorption of other nutrients. High levels of soluble fiber cause a reduction in ME digestibility that far surpasses the energy dilution effect of the indigestible fiber itsel. Meloche et al. (2013) determined a range in TDF content from 28.9 to 37.8 % TDF with an average of 33.4 % among fifteen samples of DDGS. Oil content also contributes to the feeding value of DDGS as a source of dietary energy. Typically, the ether extract (EE) content of DDGS ranges from 9 to 12 %. These results might suggest that oil content is relatively constant among conventional DDGS sources, particularly in comparison with other more variable nutrients such as carbohydrates. However, modifications of the conventional dry-grind method have recently been developed to allow for the extraction of specific non-fermentable corn fractions in DDGS. Increased demand for corn oil as a feedstock for biodiesel production has created an economic incentive for ethanol producers to implement post-fermentation oil extraction strategies. It has been estimated that approximately 8 5 % of ethanol producers in the U.S. will have implemented oil extraction processes in their facilities by 2014. The removal of an additional 3 to 6 % EE results in a concentrating effect on other nutrients in reduced-oil DDGS. Oil-extracted DDGS samples typically have EE contents lower than 9.0 %, with some samples as low as 2.0 to 4.0 % EE . Data are limited on the ME value of reduced-oil DDGS for poultry; however, Rochell et al. (2011) reported a significantly lower AME_n content of 2,146 kcal/kg (DM basis) for one sample of reduced oil DDGS. Furthermore, studies in swine have indicated that extracted oil may be more digestible than intact oil. Therefore, extraction may also preferentially remove more readily available oil, in addition to reducing overall oil content. As reduced-oil varieties of DDGS become more common, compensating for the additional effects of oil extraction on the variability of ME and other nutrients in DDGS will become increasingly critical to diet formulation.

Broiler performance
The differences in AME_n and SIAAD values among DDGS with variable oil content may pose difficulties in diet formulation, as compensation must be made for variation in nutrient content to avoid negative effects on broiler performance. Dozier and Hess (2014) investigated the effects of feeding diets containing low, moderate, and high oil DDGS at moderate (5 %, 7 %, and 9 %) and high (8 %, 10 % and 12 %) inclusion rates in the starter, grower, and finisher diets, respectively. In this experiment, three sources of DDGS with varying ether extract content (6.06, 8.80, and 11.59 %; DM basis) were fed to broilers from 1 to 33 days of age. Apparent ME_n and digestible amino acid values of the 3 DDGS sources used in formulation were determined by Perryman et al. (2014), as described above. Diets varying in DDGS source did not affect BW gain or feed conversion ratio of broilers from 1 to 33 days of age. Broilers receiving diets formulated to the moderate inclusion rate of DDGS (P≤0.001) grew faster and more efficiently from 1 to 25 and 1 to 33 d than birds provided diets containing the high inclusion rate of DDGS. Significant interactions (P≤0.02) occurred for 1 to 25 d feed conversion, carcass yield, total breast meat yield, wing yield, and abdominal fat percentage. Broilers fed diets containing high-oil DDGS with the moderate inclusion rate had improved 1 to 25 d feed conversion, carcass yield, total breast meat yield, and less abdominal fat compared with the other dietary treatments. These data indicated that oil content of DDGS may affect carcass characteristics in conjunction with inclusion rate. Inclusion rate of DDGS had a more pronounced effect than DDGS source on growth performance and processing yields of broilers from 1 to 33 d of age. Given the reported impact of oil extraction on the nutrient content of DDGS and the potential for negative effects on bird performance and carcass characteristics, it may be necessary to conduct a thorough chemical analysis of each source of DDGS to be used in formulation. However, in vivo assays for the determination of ME are both time-consuming and costly. Therefore, alternative methods for accurately and rapidly determining the ME content of DDGS may prove a valuable asset to poultry producers.

Prediction equations
Previous research has demonstrated the development of equations that predict the metabolizable energy (ME) value of corn co-products based on their chemical composition. The importance of measures of fiber in the prediction of ME has been reported in poultry and in swine. The weak relationship between EE and ME may be attributed to the predominance of fiber fractions as a percentage of DDGS composition. These fractions not only dilute dietary ME, but also may negatively affect the ME of other nutrients within the feedstuff due to the formation of gels in the gastrointestinal tract. The dual effects of fiber on ME create a stronger relationship between the concentration of fiber and the predicted ME value. Although these equations have high associated R² values for the dataset used in model-building, there is no guarantee that each model will perform similarly on an external set of DDGS samples. Meloche et al. (2014) applied the equations of Rochell et al. (2011) and Meloche et al. (2013) to a set of 15 additional DDGS samples. A comparison of the predicted and actual AME_n revealed substantial root mean square error (RMSE) values (335, 381, 488, and 502 kcal/kg, respectively) for the four equations evaluated. These errors were not consistent with the expected predictive potential of these equations based on R² values and may preclude the application of these models in a practical setting. Furthermore, the inconsistency of performance for these equations indicated that thorough validation of future models for the prediction of AME_n is warranted in order to assess the associated errors of prediction.

Conclusions
Co-products of the ethanol industry may be cost-effective sources of energy and protein in poultry diets. However, successful integration of these products into poultry diets may require careful consideration during formulation in order to combat excessive variability in nutrient content. Prediction equations may be useful in ranking DDGS samples by their relative energy values, although the error associated with such equations exceeds the allowable error for feed formulation. Little information exists concerning the effects of commercial diets containing corn co-products, other than DDGS, on live performance in a practical setting. Additional research concerning these less common products may allow for more effective inclusion in poultry diets.