Hi, I'm Diego Navarro, from the Stein Monogastric Nutrition Laboratory at the University of Illinois, and I will talk about the chemical composition and physicochemical characteristics of feed ingredients and their effects on in vitro ileal and total tract digestibility of dry matter.

Prediction equations have traditionally been used to predict the energy content of feed ingredients where analyzed components did not add to 100%. This may result in erroneous predictions because not all energy contributing components in the feed are accounted for. Therefore, we must account for all energy contributing components in the feed.

An example of this is, in this publication, they have analyzed corn and 10 sources of DDGS and generated prediction equations with high R2 values. However, when we look at the feed analysis of the first 3 sources of DDGS that were used to generate these equations, there was an incomplete account of the composition of these ingredients. And so there may have been components that contain energy that are not accounted for.

Therefore, the objective of this study were: 1) To test the hypothesis that total gross energy can be calculated from all energy containing components in the feed ingredient; and 2) To determine correlations between physicochemical characteristics and in vitro apparent ileal digestibility, or IVAID, and in vitro apparent total tract digestibility, or IVATTD, of dry matter.

There were ten test ingredients used in this experiment: two sources of cereal grains, that are corn and wheat; two sources of oilseed meals, that are soybean meal and canola meal, four sources of by products and coproducts that are corn distillers dried grains with solubles, corn germ meal, copra expellers, and sugar beet pulp. We also included two synthetic sources of fiber, that are solka floc (or purified cellulose) and pectin, to represent extremes of insoluble and soluble fiber, although these ingredients are not usually included in commercial diets fed to pigs.

This graph summarizes the concentration of total dietary fiber, or TDF, that is divided into insoluble dietary fiber, or IDF, represented by the orange bars, and soluble dietary fiber, or SDF, represented by the blue bars. We observed a greater concentration of TDF in our byproducts, mostly IDF, with a small portion being SDF. And as expected, synthetic cellulose is purely insoluble, and pectin is soluble fiber. I would like to point out that we only analyzed 50% of SDF in pectin because it contained about 40% glucose as a carrier.

This next slide summarizes the analyzed nutrient compostion of the ten feed ingredients, with the different colored bars representing a nutrient. On the x axis we have our feed ingredient, and on the y axis we have percentage units. The blue bar represents moisture, yellow bar represents total amino acids, the green bar represents acid hydrolyzed ether extract, the black bars represent ash, the orange represent TDF, the white bars represent total starch including resistant starch, and the pink bars represent sugars, oligosaccharides, as well as fructo-oligosaccharides that were analyzed in wheat, DDGS, corn germ meal, and sugar beet pulp. Most ingredients were analyzed to 100% or greater except for synthetic cellulose. Those that were added to greater than 100% may be due to some overlaps in analytical methods used, in particular the carbohydrate analyses.

Gross energy was calculated using this modified equation using Atwater factors, and is summarized in this slide. The concentrations of acid hydrolyzed ether extract was multiplied by a factor of 39.36 MJ/kg. Total amino acids was multiplied by a factor of 23.45. The total amino acids instead of crude protein was used because energy is derived from amino acids and not from non protein nitrogen. Total starch, fructo-oligosaccharides, and non-starch polysaccharides were multiplied by a factor of 17.58. Glucose, fructose, sucrose, stachyose, and raffinose were multiplied by a factor of 15.49. And lignin was multiplied by a factor of 29.13. The factor we used for lignin is not an Atwater factor but is the average gross energy of four commercially available sources of lignin that were derived from forages.

Using this modified equation, we compared the analyzed and the calculated gross energy values of the feed ingredients. On the x-axis we have feed ingredients and on the y-axis is concentration of energy in MJ/kg. The orange bars represent the analyzed gross energy and the blue bars represent the calculated gross energy from all energy-containing components in the feed ingredient. The values in green and red indicate the percentage of analyzed gross energy that was accounted for in the calculated gross energy. The differences between calculated and analyzed values were within 5% for most ingredients. The differences between the analyzed and the calculated values may be due to inaccuracies or overlaps in quantifying the concentration of each energy-contributing component, as the nutrient composition of some ingredients were greater and some less than 100%. Another possible reason is that the Atwater factors that were used may not always be applicable to every ingredient.

To sum this up, we were able to calculate the gross energy close to the actual analyzed value for gross energy in the feed ingredients. However, our calculated gross energy slightly differed from our analyzed gross energy. And this may be due to some inaccuracies, which include the energy value assigned to lignin because this value may not be applicable to feed ingredients as it was derived from lignin found in forages. Another source of inaccuracy may be that because total amino acids was used, there may have been fractions of crude protein that may contain energy that was overlooked.

With that, I want to move on to the physicochemical characteristics of fiber that are relevant to animal nutrition.

First is bulk density, and this is defined as the weight of feed divided by the volume it occupies. The bulk density ranged from around 660 g/L in DDGS, copra expellers and sugar beet pulp to about 780 g/L in soybean meal. Lower bulk density may result in gut fill and reduced nutrient intake which may translate into lower growth performance. But bulk density can be manipulated by feed processing.

Next is water binding capacity, which is the quantity of water that is bound with the application of an external force. The water binding capacity ranged from around 1.2 g/g in the cereal grains with mostly IDF, up to around 3-4 g/g in the byproducts that contained some level of SDF. 

Swelling is defined as the spreading out of molecules by incoming water. Swelling ranged from around 2.5 L/kg dry matter in corn to 8 L/kg dry matter in sugar beet pulp. It was expected that pectin have the greatest swelling capacity because it is purely SDF.

The last one is viscosity, and it is defined as the fluid’s resistance to flow. As expected, pectin had the greatest viscosity due to its high concentration of soluble dietary fiber. All other ingredients had relatively lower viscosity due to lower concentrations of soluble dietary fiber compared with pectin but generally the greater the concentration of soluble dietary fiber the higher the viscosity.

Moving on to the in vitro digestibility procedure that was used in this experiment. The procedure used to determine IVATTD was a 3-step procedure modified from Boisen and Fernandez, where the first and second steps simulated the digestion process in the stomach and the small intestine, respectively. And the third step simulated the degradation of soluble dietary fiber in the hindgut. For IVAID, the same procedure was used but was discontinued after the second step.

The results of the in vitro digestibility of dry matter is summarized in this graph. The orange bars represent IVAID and the blue bars represent IVATTD. We expected to observe cellulose to have the lowest digestibility because it is purely insoluble dietary fiber and pectin to have the highest digestibility because it is mostly soluble dietary fiber and glucose. We also expected to observe an increase in digestibility of dry matter from IVAID to IVATTD due to the additional third step which simulated hindgut fermentation of soluble fiber. We also observed a greater increase in digestibility of dry matter from IVAID to IVATTD in ingredients with greater soluble dietary fiber:total dietary fiber ratios except in corn, and we suspect this is because of resistant starch that was not degraded until the third step of digestion.

This table shows the correlations between fiber fractions and physical characteristics of feed ingredients. We observed that soluble dietary fiber was correlated with swelling and viscosity. We also observed that bulk density was negatively correlated with the concentration of neutral detergent fiber. Swelling was also correlated with water binding capacity.

We observed no correlation between the concentrations of acid detergent fiber and neutral detergent fiber and the in vitro digestibility of dry matter. However, we observed that IDF and TDF were negatively correlated with both IVAID and IVATTD of dry matter, with a stronger correlation using IDF than TDF.

In conclusion, we observed that IDF and TDF were more strongly correlated with IVAID and IVATTD of dry matter and therefore, TDF and IDF may be more appropriate in evaluating digestibility than NDF and ADF. We also observed no correlation between the physical characteristics of feed ingredients and IVAID or IVATTD of dry matter, and therefore conclude that these parameters may not influence digestibility of dry matter in vitro.

I would like to acknowledge Agrifirm Innovation of the Royal Dutch Agrifirm for financial support of this project.

Thank you for listening, and if you'd like to know more about this topic or nutrition in general, I would encourage you to visit our website at nutrition.ansci.illinois.edu.