University of Arkansas, FayettevilleScholarWorks@UARK

Theses and Dissertations

12-2018

Optimizing Methods to Determine MetabolizableEnergy Values of Feed Ingredients for BroilersSkyler P. WestUniversity of Arkansas, Fayetteville

Follow this and additional works at: https://scholarworks.uark.edu/etd

Part of the Animal Experimentation and Research Commons, Animal Studies Commons, and thePoultry or Avian Science Commons

This Thesis is brought to you for free and open access by ScholarWorks@UARK. It has been accepted for inclusion in Theses and Dissertations by anauthorized administrator of ScholarWorks@UARK. For more information, please contact scholar@uark.edu, ccmiddle@uark.edu.

Recommended CitationWest, Skyler P., "Optimizing Methods to Determine Metabolizable Energy Values of Feed Ingredients for Broilers" (2018). Theses andDissertations. 3083.https://scholarworks.uark.edu/etd/3083

Page

i

Optimizing Methods to Determine Metabolizable Energy Values of Feed Ingredients for Broilers

A thesis submitted in partial fulfillment of the requirements for the degree of Master of Science in Poultry Science

by

Skyler West University of Missouri

Bachelor of Science in Animal Science, 2016

December 2018 University of Arkansas

This thesis is approved for recommendation to the Graduate Council.

Samuel J. Rochell, Ph.D. Thesis Director Karen Christensen, Ph.D. Committee Member

Charles Maxwell, Ph.D. Committee Member

Page

ii

ABSTRACT

Determination of metabolizable energy (ME) and amino acid (AA) digestibility values of

single feed indigents continues to be two of the most important aspects for successful least-cost

poultry feed formulation. It would be advantageous if a common diet type could be utilized to

determine both ME and AA digestibility values of feed ingredients within a single assay. Two

experiments were conducted to evaluate the effect of basal diet type and excreta collection

method on the ME value of single feed ingredients determined in broiler chicks using the

regression method. In experiment 1, the objective was to investigate the influence of semi-

purified (SP) and corn-soybean meal (CSBM) based diet types on the nitrogen-corrected ME

(MEn) of dextrose when determined by regression and to compare MEn values using the index

and total excreta collection (TC) methods. The dextrose-associated caloric intake was regressed

against the amount of dextrose intake to generate linear regression equations with slopes

corresponding to the MEn value of dextrose within each basal type. The resulting dextrose MEn

values determined using SP basal diets (3,502 and 3,553 kcal/kg) were similar (P > 0.05) to

those determined using CSBM basal diets (3,839 and 3,588 kcal/kg) for index and TC

procedures. In experiment 2, the influence of basal diet type on the ME and MEn values of an

expeller-extruded soybean meal (EE-SBM) generated in broiler chicks using the regression

method were evaluated. Linear regressions of EE-SBM associated MEn intake in kcal against

EE-SBM intake in kg resulted in similar (P > 0.05) MEn values for SP (2,542 kcal/kg) and for

CSBM (2,575 kcal/kg) diets. These results indicate that both total collection and index

procedures may be reliably used to characterize the MEn content of feed ingredients, and that

similar estimates of ingredient MEn can be determined in SP and CSBM diets, potentially

allowing for simultaneous determination of ME and AA digestibility in a single study.

Page

iii

©2018 by Skyler West

All Rights Reserved

Page

iv

ACKNOWLEDGEMENTS

I would like to express my gratitude to my advisors, Dr. Samuel Rochell and Dr. Karen

Christensen, for both having introduced me to research and for going above and beyond to

support me throughout my master’s program. I would like to thank my parents, Eric West and

Paula West, for all their support and encouragement that they have provided me ultimately

leading me down this road. I would like to also thank my mother and father-in-law, Robyn and

Mike Daniels, for providing me with endless amounts of encouragement and support. I am also

very grateful for the advice and wisdom given to me by my grandparents, Paul Klein and Bobbie

Klein. Finally, I could not be more grateful for the endless amount of love, encouragement and

support provided by my wife, Lacey West.

Page

v

TABLE OF CONTENTS

CHAPTER 1: INTRODUCTION ………………..………………………….……………….…...1

LITERATURE REVIEW ……………………………………….…………………...…...4 METABOLIC UTILIZATION OF FEED ENERGY ..…..…….……….……………..…4 PARTITIONING OF FEED ENERGY………………..……………………..……....…...7 DETERMINATION OF DIETARY ME CONTENT …………………………..………..9

METABOLIZABLE ENERGY VALUE OF SOYBEAN MEAL……………………... 15 REFERENCES …………..………………………………...…………………………... 19

CHAPTER 2: INFLUENCE OF BASAL DIET TYPE ON REGRESSION-BASED METABOLIZABLE ENERGY VALUES OF DEXTROSE DETERMINED USING INDEX AND TOTAL COLLECTION METHODS …………………………………………………… 28

ABSTRACT …………………………………………………….……………………… 28 INTRODUCTION …………………………………………….……………………….. 29 MATERIALS AND METHODS…………………………….…………………………. 31 RESULTS AND DISCUSSION……...……………………..……………………….…. 34 REFERENCES……….………………………………………………………………… 45 CHAPTER 3: INFLUENCE OF BASAL DIET TYPE ON METABOLIZABLE ENERGY VALUES OF AN EXPELLER-EXTRUDED SOYBEAN MEAL DETERMINED IN BROILER CHICKS USING THE REGRESSION METHOD ……………………………………………..47

ABSTRACT …………………………………………….……………………………… 47 INTRODUCTION …………………….……………………………………………….. 48 MATERIALS AND METHODS ………………………………………………………. 50 RESULTS ……………………………………………..……………………………….. 54 DISCUSSION ………………………………..………………………………………… 56 REFERENCES ………………………………………………………………………… 68 CONCLUSION ….……………………...…….……………………………………………...… 72 APPENDIX………………….………....………………………………………………………...73

Page

vi

LIST OF TABLES

Table 1.1 Summary of metabolizable energy (ME) values of various feed ingredients using index and total excreta collection (TC) methods ………………………………... 25

Table 1.2 Summary of metabolizable energy (ME) values of various types of soybean meal

(SBM) fed to poultry …..……………...………………………..……………...... 26

Table 2.1 Ingredient composition of experimental diets ...………………..………………... 43

Table 2.2 Growth performance of broilers fed semi-purified (SP) or corn-soybean meal (CSBM) diets containing 0, 22.5, or 45% dextrose from 14 to 21 d of age…...… 45

Table 2.3 Nitrogen retention and analyzed ME and MEn values of experimental diets

containing 0, 22.5, or 45% dextrose fed to broilers from 19 to 21 d of age……………………………………………………….……………...……..… 46

Table 2.4 Metabolizable energy and MEn values of dextrose when included in semi-purified

(SP) and corn-soybean meal (CSBM) broiler diets at 0, 22.5, or 45% using the total excreta collection (TC) and index method1 ………………………….………..… 47

Table 2.5 Intercepts and slopes of regression equations of dextrose-associated MEn intake

(kcal) on dextrose intake (kg) determined using semi-purified (SP) or corn-soybean meal (CSBM) diets and total excreta collection (TC) or index methods………………………………………………………………………..… 48

Table 3.1 Ingredient composition of experimental diets….….……………………...……... 69

Table 3.2 Growth performance of broilers fed experimental diets containing 0, 15, 30, or 45% expeller-extruded soybean meal (EE-SBM) from 14-21 d post-hatch …..... 71

Table 3.3 Nitrogen retention and metabolizable energy values of experimental diets fed to

broilers and containing 0, 15, 30, or 45% expeller-extruded soybean meal (EE-SBM)1 …………………………………………………………………………….72

Table 3.4 Metabolizable energy values of an expeller-extruded soybean meal (EE-SBM) fed

to broilers at an inclusion rate of 0, 15, 30, or 45%……………….…………..… 73

Table 3.5 Regression equations used to determine ME and MEn of an expeller-extruded soybean meal (EE-SBM) in semi-purified (SP) and corn-soybean meal (CSBM) based diets...............................................................................................................74

1 Pa

ge1

CHAPTER 1: INTRODUCTION

The largest economic inputs of live poultry production are feed costs, in particular dietary

energy costs, which represent the largest proportion of total cost for broiler production (Lopez

and Leeson, 2008a). Energy is not classified as a nutrient but rather a property of nutrients that is

released by metabolic processes. The energy content within feed ingredients that is available to

the animal must be determined by the use of bioassays. However, in vivo assays to determine the

available energy content of feed ingredients are inherently variable due to a lack standardization

among various bioassays to determine energy values for feed ingredients. As such, there is a

need for further work to refine feed energy bioassays.

At present, metabolizable energy (ME) is the system typically used in commercial

poultry nutrition to describe the energy requirements and dietary energy content for poultry

(NRC, 1994). Metabolizable energy represents the feed energy utilized by the animal and is

commonly determined with a balance procedure where the difference in ingested gross energy

(GE) and excreta energy output are measured over the same period of time (NRC, 1994). Two

approaches to quantify ME values for feedstuffs include true metabolizable energy (TME) and

apparent metabolizable energy (AME) (Sibbald and Slinger, 1963; Guillaume and Summers,

1970; Sibbald, 1976; Farrell, 1978). The TME assay, developed by Sibbald (1976) often utilizes

adult roosters whereby single test ingredients are precision fed using gavage to provide a rapid,

direct calculation of ME with also accounting for endogenous losses. Conversely, AME assays

usually involve the feeding of two or more complete diets whereby the test diets are created by

substituting a portion of the reference diet with the test ingredient (Anderson et al., 1958; Sibbald

and Slinger, 1963). Additionally, metabolizable energy values are often corrected for nitrogen

(N) retention (AMEn, TMEn) so that all data are on a N equilibrium basis for comparative

2 Pa

ge2

purposes with the assumption that all N retained will be excreta as uric acid (Lopez and Leeson,

2008b).

In ME studies, the estimated ME value of the test ingredient can be influenced by its

inclusion level in the test diet. Relatively high inclusion levels can result in nutritional

imbalances of the test diet; whereas, low inclusion levels can increase variability among

determined ME values (Sibbald and Slinger, 1962; Leeson et al., 1977; Mateos and Sell, 1980).

These factors should also be balanced with a feeding level that is relevant for commercial

practice. The regression method proposed by Potter et al. (1960) has been utilized in ME studies

to allow for a test ingredient to be substituted at multiple levels within the test diet. Therefore,

the regression approach may reduce some variation associated with inclusion level when

compared with studies that only use a single inclusion level (Leeson et al., 1977). In addition to

ME, the regression method has also been used for amino acid (AA) digestibility studies, which

use more purified, highly digestible diet types compared with ME studies so that any undigested

AA can be attributed back to the test ingredient (Rodehutscord et al., 2007). If a common diet

type could be used in the determination of both ME and AA digestibility values for feed

ingredients within a single regression study, then this would align with the philosophy to reduce,

refine, and replace animals for research.

Dextrose has served as a reliable ingredient to evaluate when comparing different

methodological approaches to determine the MEn value of feed ingredients (Anderson et al.,

1958). Using a standard ME value for dextrose, metabolizable energy values for poultry have

been determined by a direct ingredient replacement approach whereby dextrose is used as a

reference ingredient (Rochell et al., 2012; Meloche et al., 2013). However, no experiments have

3 Pa

ge3

been conducted since those of Anderson et al. (1958) to confirm the MEn value of dextrose,

which may vary under different experimental conditions.

The total collection (TC) method whereby total feed energy intake and excreta energy

output are measured in the determination of ME has been predominately used in ME studies

(Sales and Janssens, 2003). However, the index method is an alternative to the TC method and is

based on the use of an indigestible marker, such as titanium dioxide (TiO2), recovered in the feed

and excreta in the determination of ME (Short et al., 1996). The index method negates any errors

associated with inaccurate measurement of feed intake and excreta output, but requires the use of

variable analytical procedures, which can also be a source of error (Sales and Janssens, 2003). At

present, ME studies that compare values generated using both TC and index methods with TiO2

as an indigestible marker are limited for poultry (Smeets et al., 2015).

Soybean meal (SBM) is the most commonly-used proteinaceous ingredient for

nonrumminat animal feeds and is considered the “gold standard” for intact protein sources

(Dozier and Hess, 2011; Ravindran et al., 2014a). Expeller-extruded SBM (EE-SBM) is an

alternative to the conventional, solvent-extracted SBM (SE-SBM) and typically contains

approximately 4 times the amount of oil (Zhang and Parsons, 1993; García-Rebollar et al., 2016)

resulting in an overall greater ME content (Powell et al., 2011). Thus, EE-SBM has been

suggested to be good alternative source of both ME and digestible AA in poultry diets but

published ME values of EE-SBM for broilers are sparse (Powell et al., 2011). Therefore, the

objectives of this thesis are as follows:

1. To investigate the influence of semi-purified (SP) and corn-soybean meal

(CSBM) diet types on the MEn of dextrose when determined by regression, and to

compare ME values generated using the TC and index methods.

4 Pa

ge4

2. To evaluate the influence of SP and CSBM diets on the ME values of a

commercially available EE-SBM determined using the regression method in

growing broilers.

LITERATURE REVIEW

METABOLIC UTILIZATION OF FEED ENERGY

Carbohydrate Utilization

Commercial broiler diets in the U.S. consist of 60-70% corn, which is the primary

energy-providing ingredient for broilers (Cowieson, 2005). Specifically, this energy is derived

from carbohydrates, which exist largely in the form of starch. Amylose and amylopectin

comprise starch granules from plants, which differ by source in structural characteristic affecting

overall digestibility. Starch is readily digestible in the presence of endogenous digestive enzymes

including pancreatic α-amylase and is almost completely digested and absorbed (up to 95%) by

the time it reaches the terminal ileum of the small intestine (Moran, 1985). The release of α-

amylase can be altered as the amount of dietary starch changes (Moran, 1985), allowing the bird

to maximize carbohydrate utilization. Additionally, the high capacity that poultry exhibit for

digesting starch improves with age due to increases in amylase activity (Siddons, 1969;

Krogdahl, 1985; Moran, 1985; Uni et al., 1995).

A major factor influencing carbohydrate utilization by poultry is the presence of complex

polysaccharides, termed non-starch polysaccharides (NSP), that originate from both cereal grains

and plant protein sources (Wagner and Thomas, 1978; Antoniou and Marquardt, 1981; Bedford

et al., 1991; Slominski, 2011). Non-starch polysaccharides from cereals include cellulose, β-

glucans, and arabinoxylans, whereas in SBM, arabinans, arabinogalactans, galactans,

5 Pa

ge5

galactomannans, mannans, and pectic-polysaccharides predominate (Slominski, 2011). In SBM,

total NSP contain both insoluble and soluble NSP, which include both cellulosic and non-

cellulosic polysaccharides, respectively (Karr-Lilienthal et al., 2005). Negative impacts from

NSP have been reported on feed intake, growth performance, and nutrient digestibility

(Jørgensen et al., 1996; Sklan et al., 2003). Soluble NSP increase digesta viscosity which can

have a detrimental influence on digesta flow and utilization of nutrients in broilers (Choct and

Annison, 1990; McNab and Smithard, 1992; Pettersson and Åman, 2007).

A significant amount of feed energy is provided by soybean meal (SBM), which contains

30-40% carbohydrates; however, not all carbohydrates provided by SBM are readily digestible

by poultry (Karr-Lilienthal et al., 2005). Galacto-oligosaccharides (GOS) play a large role in

affecting the ME content of SBM in poultry diets. Raffinose, stachyose and verbascose comprise

between 5 and 7% of GOS in SBM on a DM basis (Grieshop et al., 2003; Karr-Lilienthal et al.,

2005; Knudsen, 2014) and are poorly digested due to the absence of an endogenous α-1,6

galactosidase enzyme in poultry (Gitzelmann and Auricchio, 1965; Carré et al., 1995). It has

been demonstrated that a reduction in SBM oligosaccharides from 3.81 to 0.6% through genetic

modification increased the ME value of SBM from 2,739 to 2,931 kcal/kg (Parsons et al., 2000).

In addition to GOS, β-mannans are linear polysaccharides found in SBM that have been reported

to significantly decrease growth and increase feed:gain ratio in broilers (Ray et al., 1982;

Daskiran et al., 2004). As such, the addition of exogenous β-mannanase may improve the ME of

SBM as well as the growth and feed conversion of poultry (Odetallah et al., 1997; Jackson et al.,

2004; Wu et al., 2005; Zou et al., 2006).

Lipid Utilization

6 Pa

ge6

Lipids are the most concentrated sources of energy in poultry diets and provide essential

fatty acids and aid in the absorption of fat-soluble vitamins. The process of lipid digestion and

absorption mainly occurs in the small intestine of the chicken (Uni et al., 1995). Emulsification

occurs resulting in the formation of lipid droplets, which leads to lipolysis aided by pancreatic

lipase. Monoglycerides from lipolysis are incorporated into mixed micelles with bile salts and

are absorbed by intestinal enterocytes (Bauer et al., 2005). Sell et al. (1986) reported that

improvements in fat utilization with increasing bird age are due to increased bile salt secretions

and increased pancreatic lipase activity. In addition, the fatty acid profile of lipids has been

reported to impact overall fat digestibility among feed ingredients (Lewis, 1989; Whitehead et

al., 1993; Kim et al., 2013b). In general, unsaturated fatty acids are more readily absorbed than

saturated fatty acids although there is an optimal synergistic ratio between the two that leads to

optimal fat utilization (Ketels and Groote, 1989; Lewis, 1989; Zollitsch et al., 1997). Common

supplemental lipid sources in poultry diets include poultry fat and vegetable sources such as

soybean oil and corn oil, which are gaining popularity due to consumer demands for all

vegetable-based animal diets. Thus, ingredients such as EE-SBM that contain more fat may be

especially advantageous for all vegetable-based diets. However, supplemental soybean oil also

has been reported to be more digestible compared with intact soybean oil from SBM that is

incapsulated within the bean (Kim et al., 2013a; Tancharoenrat et al., 2014; Su et al., 2015). In

addition, EE-SBM contains a higher oil content than conventional SE-SBM, which may be

economically beneficial for providing more energy to broiler diets.

Protein Utilization

Protein intake in poultry is driven by the need to meet specific amino acid (AA)

requirements to achieve lean muscle growth and tissue maintenance. In birds, dietary AA serve

7 Pa

ge7

as structural components of skin, feathers, bone matrix, ligaments, as well as muscle and

digestive tissues (Wu, 2009). Dietary supplementation of synthetic AA is a common practice for

commercial poultry diets to balance the profile of AA derived from intact proteins such as SBM

to sustain necessary lean muscle growth while minimizing AA catabolism. However, protein also

contributes a significant amount of energy to the diet through the supply of gluconeogenic

precursors. Glucogenic AA have been reported to increase feed intake and improve energy

balance by increasing the amount of succinyl-CoA to support gluconeogenesis (McNab, 2000).

In addition, in a commercial broiler diet containing 25% SBM and formulated to 3,100 kcal/kg

approximately 18% of the energy would be derived from SBM, some of which is derived from

carbon skeletons of excess AA. Although excess protein and AA can provide some energy, it is

costly from both an economic and physiological standpoint due to the metabolic strain imposed

on the bird to synthesize uric acid to excrete nitrogen (N), a process which is energetically costly

in itself (Musharaf and Latshaw, 1999; Lopez and Leeson, 2008a). Therefore, accurate

determination of ME for high-protein feed ingredients such as SBM is crucial for providing a

diet with a balanced AA profile and that supports efficient energy utilization.

PARTITIONING OF FEED ENERGY

In poultry, the partitioning of feed energy is based on the following measurements: GE,

ileal digestible energy (DE), AME, TME, or net energy (NE). The GE of a feed is measured as

the total amount of heat (i.e., energy) released from its chemical bonds when it is burned

completely in the presence of oxygen using an adiabatic bomb calorimeter. However, GE

provides no indication of the amount of energy within a feedstuff that is actually available to the

animal. Therefore, bioassays to measure the amount of energy utilized by the bird are necessary

for use in feed formulation.

8 Pa

ge8

The GE that is consumed is partly lost via the feces and the retained energy is termed DE

(NRC, 1994). In poultry, feces and uric acid are both voided together as excreta, therefore DE

cannot be quantified without surgical modification that exteriorizes the ureters (Sibbald et al.,

1962) or by collecting ileal digesta. The latter approach is termed ileal digestible energy and is

determined by taking contents from the distal section of the ileum directly after euthanasia

(Sibbald et al., 1960; Bolarinwa and Adeola, 2012; Cao and Adeola, 2016; Zhang and Adeola,

2017).

Following DE, after further losses of nitrogenous and gaseous wastes the energy value

retained by the animal of the feed or feedstuff is the AME. It is termed “apparent” due to energy

in the feces and urine containing endogenous losses from normal body turnover and not just

undigested nutrients (Lopez and Leeson, 2008a). Correction of AME for endogenous energy

losses yields the TME value of a feed or feedstuff. In addition, the ME is generally corrected to

zero N retention (AMEn, TMEn) so that all data can be compared across birds that are under

different states of N retention (Lopez and Leeson, 2008a). Currently, ME is the most commonly-

used measurement to establish dietary energy requirements of poultry and energy content of feed

ingredients in commercial feed formulation (Lopez and Leeson, 2008a).

Although ME is a useful measurement of the gross energy that is available to the bird,

some ME is ultimately lost as heat and therefore ME is not totally reflective of the energy

retained for maintenance, growth, egg production, protein turnover, or fat deposition (Sibbald,

1982; Noblet et al., 1994). Net energy, which can be determined by direct or indirect calorimetry,

necessitates the quantification for heat increment (HI). When the energy from HI is subtracted

from ME, the resulting value is the NE of the feed or feedstuff, which is the portion of energy

utilized by the animal for maintenance and production. The net energy system is theoretically the

9 Pa

ge9

most accurate and applicable measure of energy utilization in animals. Indirect calorimetry is a

tool used to measure HI in animals however, it is expensive and difficult to quantify (Pirgozliev

and Rose, 1999). As a result, values are not available for many commonly-used feed ingredients,

and studies comparing ME and NE values of feedstuffs for poultry are inconsistent (Van der Klis

and Fledderus, 2007; Carré et al., 2014).

DETERMINATION OF DIETARY ME CONTENT

ME Methods for Single Feed Ingredients

Methods to determine ME values of feed ingredients can be determined by the direct

method whereby single ingredients are fed as the sole source of energy, or by the indirect method

whereby the test ingredient is fed as part of a complete diet. In both cases, balance procedures

are often used in which feed intake and excreta output are quantitatively measured over the same

time period (Anderson et al., 1958; Sibbald and Slinger, 1963; Sibbald, 1976). Direct assays

often use fasted adult roosters whereby single ingredients are tube-fed and endogenous losses are

determined to directly quantify TME (Sibbald, 1976). Indirect assays to determine AME are

more commonly used with broilers fed a complete diet. Differences in ME values of single feed

ingredients determined using these two approaches have been reported for various feed

ingredients (Coon et al., 1990; Stefanello et al., 2011; Ravindran et al., 2014b; Zhang and

Adeola, 2017). Apparent metabolizable energy is the most commonly used measurement through

actual measurement of feed intake and excreta output (NRC, 1994; Lopez and Leeson, 2008a).

The substitution approach allows for differences in the determined ME values of the basal and

test diets to be related to the proportions of the test ingredient substituted (Sibbald et al., 1960;

Pesti et al., 1986; Lopez and Leeson, 2008b). Consequently, broiler ME assays are challenged

with the difficulty of determining the ME value of a single ingredient when fed as part of a

10 Pa

ge10

complete diet (Sibbald et al., 1960). Two commonly used methods to calculate the ME of a test

ingredient when fed in a complex diet are the complete-basal and single-ingredient replacement

assays (Anderson et al., 1958; Sibbald and Slinger, 1963).

In the complete-basal replacement approach proposed by Sibbald and Slinger (1963), the

test ingredient is added at the expense of all energy contributing components of the reference

diet. The ME values of the test and reference diets are determined, and the ME of the test

ingredient is calculated by difference (Sibbald et al., 1960; Sibbald and Slinger, 1963). The

reference diet is comprised of a corn-SBM based basal with constant vitamin and mineral

supplementation among all test diets to prevent any micronutrient deficiencies with varying

inclusions of the test ingredient. Sibbald and Slinger (1963) determined that as the inclusion

level of a test material increases within the test diet, the experimental error associated with the

estimated ME values will decrease. The complete-basal replacement approach has been used to

quantify a number of primary and alternative feed ingredients for poultry (Sibbald, 1982;

Ravindran et al., 2014a; Abdollahi et al., 2015; Cao and Adeola, 2016; Zhang and Adeola, 2017).

The single-ingredient replacement approach involves the inclusion of a well-

characterized ingredient such as glucose in the reference diet, which is then replaced by the test

ingredient in the test diets at various inclusion levels (Anderson et al., 1958). A series of

experiments by Anderson et al. (1958) determined the MEn value of glucose to be 3,640 kcal/kg

by substituting graded levels of glucose (0, 5, 10, and 20%) for cellulose in the diet; thus, it was

proposed that glucose be used as a well-established test material in ME assays. These

experiments were based on several assumptions including 1) that the digestibility of cellulose

was zero, 2) that the increase in dietary ME could be attributed solely to increased glucose

inclusion, and 3) that the ME value of glucose is constant under all experimental conditions.

11 Pa

ge11

Although several researchers have used glucose replacement to determine ME of feed

ingredients for poultry (Hill and Anderson, 1958; Pesti et al., 1986; Rochell et al., 2011; Meloche

et al., 2014), studies to compare this approach with other methods of ME determination are

lacking. Therefore, it is important that the ME value of glucose is verified to provide reliable ME

values for test ingredients. Furthermore, its consistent composition and presumably consistent

ME value should make it an ideal ingredient to test variation in ME assay methodology.

In both the complete-basal and single-ingredient replacement assays (Anderson et al.,

1958; Sibbald and Slinger, 1963), the estimated ME value of the test ingredient can be

influenced by its inclusion level in the test diet. Lopez and Leeson (2008b) reported AMEn

values of SBM to be 1,986, 2,286, and 2,477 kcal/kg when SBM was included at 10, 20, and

30%, respectively, at the expense of a complete corn-SBM based basal and fed to broilers from

30 to 33 d of age. On the other hand, Pesti et al. (1986) reported that the AMEn of poultry by-

product meal decreased from 3,330 to 2,970 kcal/kg as the inclusion level increased from 20 to

40% of the diet. Thus, the impact of test ingredient inclusion level on its determined ME value

likely varies among dietary ingredients and with other assay conditions. Although low inclusion

levels have been reported to cause variability among ME values (Mateos and Sell, 1980), high

inclusion levels can induce nutritional imbalances based on the nutritional composition of the

ingredient (Sibbald and Slinger, 1962). Therefore, feed ingredient inclusion levels should be

selected to minimize variability, while maintaining a range that is still commercially relevant.

One strategy to circumvent the influence of test ingredient inclusion levels on resulting

ME values is to use a regression method in which the test ingredient is fed at multiple

concentrations. In this approach, MEn intake (kcals) from the feed ingredient is regressed against

its feed intake (Kg), with the slope corresponding the MEn (kcal/kg) of the ingredient tested.

12 Pa

ge12

(Potter et al., 1960; Sibbald and Slinger, 1962; Short et al., 1999; Berrocoso et al., 2017). This

method has been used to quantify the ME value of several feed ingredients including rice bran,

distillers dried grains with solubles, poultry by-product meal, canola meal, and macadamia nut

cake for broilers (Pereira and Adeola, 1998; Adeola and Ileleji, 2009; Cao and Adeola, 2013;

Berrocoso et al., 2017; Zhang and Adeola, 2017).

Regression approaches have also been utilized to determine AA digestibility values of

feed ingredients (Fan and Sauer, 1995; Short et al., 1999; Kluth and Rodehutscord, 2006;

Rodehutscord et al., 2007). Rodehutscord et al. (2007) suggested that the linear regression model

should be adopted for AA digestibility studies due to the fact that basal endogenous losses are

contained within the intercept of the regression line and do not need to be determined separately.

When compared with ME studies, ileal AA assays utilize more purified, highly digestible

ingredients that are typically used in the basal diet so that undigested AA recovered at the distal

ileum can be attributed to the test ingredient (Short et al., 1999; Kluth and Rodehutscord, 2006;

Rodehutscord et al., 2007). Typically, ME studies use corn-SBM based diets to quantify feed

ingredient ME values; thus, the use of a semi-purified diet to quantify ME may not be reliable.

Adeola and Ileleji (2009) determined the energy value of distillers dried grains with solubles to

be 2,963 kcal/kg when using a SP diet, whereas a lower value of 2,787 kcal/kg was reported

when using a corn-SBM based diet. However, more research is warranted that would allow for

the determination of both ME and AA digestibility values of feed ingredients within a single

study.

Nitrogen Correction

Nitrogen retention can be influenced by both age and bird type; therefore, it is generally

accepted that a correction of ME values to zero N retention minimizes variability across assays

13 Pa

ge13

for feed ingredients (Parsons et al., 1982; Sibbald, 1982; Lopez and Leeson, 2008b). The

correction of energy values for N retention for feed ingredients was first demonstrated by Hill

and Anderson (1958). It was proposed that a correction value of 8.22 kcal/g of N retained should

be applied to ME values because this is the energy required by the bird to expel nitrogen as uric

acid. However, the concept of N retention is questionable due to fact that ME values are

penalized when N is efficiently converted to lean muscle growth, a process for which modern

broilers have been genetically selected. Dale and Fuller (1984) found that the ME value of high-

protein sources such as SBM were influenced the most when corrected for N retention compared

with low-protein ingredients such as cereals. Lopez and Leeson (2008b) confirmed the findings

by Dale and Fuller (1984) and further concluded that correcting for N retention reduced the ME

values of a conventional SBM from 7 to 12%. Therefore, it is advantageous to report N retention

values as well as both ME and MEn values to provide readers with awareness for interpreting

energy values.

Total Excreta Collection vs. Index Method

The method of excreta collection used to determine ME of feeds can be based on total

excreta collection (TC) or use of an indigestible marker (i.e., index method), and both have been

used extensively in previous literature, as reviewed by (Sales and Janssens, 2003). Total excreta

collection is based on relating total feed intake by excreta output whereby the index method is

based on the ratio of an indigestible marker within the feed and voided excreta (Kotb and

Luckey, 1972; Sibbald, 1982). In theory, the TC method is preferred as it measures ME directly

by use of a physical balance procedure; however, in practice, accurate measurement of excreta

output can be difficult to measure due to adherence of excreta to the bird or wire flooring,

contamination of excreta with feathers, excreta loss during emptying of trays to containers, and

14 Pa

ge14

birds excreting away from the tray. This can result in an underestimation of the amount of energy

excreted and overestimation of energy retained (McNab, 2000). Errors associated with

inaccurately measuring feed intake and excreta output are obviated with the use of markers.

However, analytical procedures for marker assays introduce an external source for potential

errors (Sibbald, 1987; Sales and Janssens, 2003). Additionally, it has been recognized in

previous literature that the agreement between ME values determined by TC and those

determined using the index method varies among the use of different markers. These data have

been summarized in Table 1.1.

Short et al. (1996) developed a colorimetric procedure for using TiO2 that facilitated its

use in place of Cr2O3, which was widely used through the 1960’s. The use of Cr2O3 as an

indigestible marker has been questioned, due to substantial losses during feed mixing because of

separation from fibrous particles as well as substantial variation reported among laboratory

procedures (Kane et al., 1950; Vohra and Kratzer, 1967; Vohra, 1972). Additionally, the use of

TiO2 in both AA digestibility and ME studies has resulted in consistently higher recovery rates

with minimal variation among samples due to ease of mixing, homogeneity, and consistent

laboratory methods compared with Cr2O3 procedures (Jagger et al., 1992; Smeets et al., 2015).

To date, only one publication exists in which ME values generated using TiO2 and the TC

method were compared (Smeets et al., 2015). In this paper, Smeets et al. (2015) reported that

AMEn of complete wheat based diets showed an overall similar trend for both TC and index

methods; however, lower AMEn and P-values were reported for the index method using TiO2 as

an indigestible marker. Further research should be conducted to evaluate the variability of TiO2

as an indigestible marker for the index method compared with the more conventional TC

method.

15 Pa

ge15

METABOLIZABLE ENERGY VALUE OF SOYBEAN MEAL

Soybean meal is the most commonly used source of protein in poultry diets worldwide

(Dozier and Hess, 2011; Ravindran et al., 2014a). It contains favorable attributes that

complement cereals, such as corn, due to its high CP content and digestible AA profile. The

nutrient profile of SBM can be influenced by agronomic conditions of the raw soybeans and

subsequent processing conditions (Dozier and Hess, 2011). Indeed, optimal SBM processing

conditions ensure that trypsin inhibitors, which are antinutritional factors inherently present

within soybeans, are properly denatured in the meal (Zhang and Parsons, 1993; Parsons et al.,

2000; García-Rebollar et al., 2016). To produce solvent-extracted SBM, soybeans are cracked

and dehulled, and hexane is used to extract oil from the flaked oilseeds (Dozier and Hess, 2011).

Comparatively, in EE-SBM production, soybeans are cracked, dehulled and forced through die

holes under pressure inactivating trypsin inhibitors followed by expelling (Karr-Lilienthal et al.,

2006). These two processing techniques will result in different amounts of fat and crude protein

content in the resulting meals. Solvent-extracted SBM is expected to contain between 44 to 48%

crude protein with 0.8 to 1.9% fat (NRC, 1994; García-Rebollar et al., 2016), whereas EE-SBM

typically contains 42 to 47% crude protein with up to 10% oil content (Zhang and Parsons,

1993). As such, EE-SBM is a good potential source of both ME and digestible AA in poultry

diets (Powell et al., 2011), but published ME values of EE-SBM for broilers are sparse. The cost

of installing expeller-extrusion plants has been reported to be relatively inexpensive and more

feasible for processing locally grown soybeans (Wang and Johnson, 2001).

The ME of SE-SBM has been well characterized for poultry with average ME values of

2,263 kcal/kg for broilers and 2,770 kcal/kg for roosters reported in Table 1.2. However, the ME

of EE-SBM has mainly been reported for roosters. Zhang and Parsons (1993) showed the TMEn

16 Pa

ge16

content of an EE-SBM processed under various temperatures ranged from 3,125 to 3,239 kcal/kg

DM in cecectomized adult roosters. The average TME value of EE-SBM derived from poultry is

3,210 kcal/kg DM reported in Table 1.2. In addition, a calculated MEn value of 2,882 kcal/kg

DM for EE-SBM has been reported by Powell et al. (2011) for broilers; however, no studies have

reported any in vivo ME values of EE-SBM for broilers. In conclusion, the current body of

literature contains a wide range of values reported for SBM meals that vary across bird type and

origin. Therefore, due to the lack of ME values determined in broilers for EE-SBM more

research is needed to evaluate its feeding value and use in poultry diets.

Page

18

17

Table 1.1 Summary of metabolizable energy (ME) values of various feed ingredients using index and total excreta collection (TC) methods

Reference

Bird type

Estimate1

Ingredient

Marker2

Index

method (kcal/kg)

TC (kcal/kg)

Index vs

TC (kcal/kg)

Vogtmann et al. (1975) Broilers ME Soybean oil AIA 7,950 7,780 +170 Han et al. (1976) Broilers

Broilers Broilers Broilers

MEn MEn MEn MEn

Yellow corn

Barley Wheat SBM

Cr2O3 Cr2O3 Cr2O3 Cr2O3

3,010 2,660 3,080 2,150

3,310 2,500 3,060 2,190

-300 +160 +20 -40

Coates et al. (1977) Broilers Roosters

ME ME

Barley Barley

Cr2O3 Cr2O3

3,050 3,250

3,110 3,300

-60 -50

Halloran and Sibbald (1979)

Broilers

AME

Animal fat

CF

8,050 8,160 -110

Tillman and Waldroup (1988)

Broilers

AMEn

Grain amaranth

AIA

4,474 3,414 +1,060

Smeets et al. (2015) Broilers AMEn Wheat TiO2 3,110 3,264 -154 1AME – apparent metabolizable energy, AMEn – nitrogen-corrected apparent metabolizable energy, MEn – nitrogen-corrected metabolizable energy. 2AIA – acid insoluble ash, Cr2O3 – chromic oxide, TiO2 – titanium dioxide, CF- crude fiber. 3SBM – soybean meal.

18 Pa

ge18

Table 1.2 Summary of metabolizable energy (ME) values of various types of soybean meal (SBM) fed to poultry

Reference

Method1

Bird type

Ingredient2

Estimate

Value3 (kcal/kg DM)

Coon et al. (1990) Direct Direct

Roosters Roosters

SE-SBM EE-SBM

TMEn TMEn

2,794 3,368

Zhang and Parsons, (1993)

Direct Direct Direct

Roosters Roosters Roosters

EE-SBM EE-SBM EE-SBM

TMEn

TMEn

TMEn

3,125 3,265 3,239

Sulistiyanto et al. (1999) Indirect Indirect Indirect

Broilers Broilers Broilers

SE-SBM SE-SBM SE-SBM

TMEn

TMEn

TMEn

2,150 2,540 2,451

Edwards et al. (2000) Direct Direct

Roosters Roosters

SE-SBM SE-SBM

TMEn TMEn

2,172 2,213

Parsons et al. (2000) Direct Direct

Roosters Roosters

SE-SBM SBM-LO

TMEn TMEn

2,739 2,931

Lopez and Leeson (2008b)

Indirect Indirect Indirect Indirect Indirect Indirect Indirect Indirect Indirect Indirect Indirect Indirect

Broilers Broilers Broilers Broilers Broilers Broilers Broilers Broilers Broilers Broilers Broilers Broilers

SE-SBM SE-SBM SE-SBM SE-SBM SE-SBM SE-SBM SE-SBM SE-SBM SE-SBM SE-SBM SE-SBM SE-SBM

AMEn

AMEn AMEn

AMEn

AMEn

AMEn

AMEn AMEn

AMEn

AMEn

AMEn

AMEn

1,962 1,986 1,998 2,037 2,286 2,477 2,170 2,171 2,180 2,186 2,327 2,383

Baker et al. (2011) Direct Direct

Roosters Roosters

SBM-LO SE-SBM

TMEn

TMEn 2,984 2,963

Perryman and Dozier (2012)

Indirect Indirect Indirect Indirect Indirect

Broilers Broilers Broilers Broilers Broilers

SE-SBM SE-SBM SBM-LO SBM-LO SBM-UL

AMEn AMEn

AMEn

AMEn

AMEn

2,073 2,241 2,214 2,435 2,080

Loeffler et al. (2013) Direct Indirect

Roosters Broilers

SE-SBM SE-SBM

TMEn

AMEn 3,740 3,468

Ravindran et al. (2014b) Indirect Indirect Indirect Indirect

Broilers Broilers Broilers Broilers

FF-SBM FF-SBM FF-SBM FF-SBM

AMEn AMEn AMEn AMEn

2,800 3,357 3,366 3,395

1Direct – based on the test ingredient fed as the sole source, Indirect – based on the test ingredient fed as part of a complete diet. 2SE-SBM – solvent-extracted soybean meal, EE-SBM – expeller-extruded soybean meal, FF-SBM – full-fat soybean meal, SBM-LO – low oligosaccharide soybean meal, SBM-UL – ultra-low oligosaccharide soybean meal. 3Average ingredient estimates by bird type are as follows in kcal/kg DM: Broiler SE-SBM AMEn, 2,263; Broiler FF-SBM AMEn, 3,230; Broiler SBM-LO AMEn, 2,325; Broiler SE-SBM TMEn, 2,380; Rooster SE-SBM TMEn, 2,770; Rooster EE-SBM TMEn, 3,249; Rooster SBM-LO TMEn, 2,958.

19 Pa

ge19

REFERENCES

Abdollahi, M. R., B. Hosking, and V. Ravindran. 2015. Nutrient analysis, metabolisable energy and ileal amino acid digestibility of palm kernel meal for broilers. Anim. Feed Sci. Technol. 206:119–125.

Adeola, O., and K. E. Ileleji. 2009. Comparison of two diet types in the determination of

metabolizable energy content of corn distillers dried grains with solubles for broiler chickens by the regression method. Poult. Sci. 88:579–585.

Anderson, D. L., W. Hill, and R. Renner. 1958. Studies of metabolizable and productive energy

of glucose for the growing chick. J. Nutr. 65:561–574. Antoniou, T., and R. R. Marquardt. 1981. Influence of rye pentosans on the growth of chicks.

Poult. Sci. 60:1898–1904. Baker, K. M., P. L. Utterback, C. M. Parsons, and H. H. Stein. 2011. Nutritional value of

soybean meal produced from conventional, high-protein, or low-oligosaccharide varieties of soybeans and fed to broiler chicks. Poult. Sci. 90:390–395.

Bauer, E., S. Jakob, and R. Mosenthin. 2005. Principles of physiology of lipid digestion. Asian-

Australasian J. Anim. Sci. 18:282–295. Bedford, M. R., H. L. Classen, and G. L. Campbell. 1991. The effect of pelleting, salt, and

pentosanase on the viscosity of intestinal contents and the performance of broilers fed rye. Poult. Sci. 70:1571–1577.

Berrocoso, J. D., S. Yadav, and R. Jha. 2017. Nitrogen-corrected apparent metabolizable energy

value of macadamia nut cake for broiler chickens determined by difference and regression methods. Anim. Feed Sci. Technol. 234:65–71.

Bolarinwa, O. A., and O. Adeola. 2012. Energy value of wheat, barley, and wheat dried distillers

grains with solubles for broiler chickens determined using the regression method. Poult. Sci. 91:1928–1935.

Cao, M. H., and O. Adeola. 2016. Energy value of poultry byproduct meal and animal-vegetable

oil blend for broiler chickens by the regression method. Poult. Sci. 95:268–275. Carré, B., J. Gomez, and A. M. Chagneau. 1995. Contribution of oligosaccharide and

polysaccharide digestion, and excreta losses of lactic acid and short chain fatty acids to dietary metabolisable energy value in broiler chickens and adult cockerels. Br. Poult. Sci. 36:611–629.

Carré, B., M. Lessire, and H. Juin. 2014. Prediction of the net energy value of broiler diets.

Animal 8:1395–1401.

20 Pa

ge20

Choct, M., and G. Annison. 1990. Antinutritive activity of wheat pentosans in broiler diets. Biritsh Poult. Sci. 31:811–821.

Coates, B. J., S. J. Slinger, J. D. Summers, and H. S. Bayley. 1977. Metabolizable energy values

and chemical and physical characteristics of wheat and barley. 207:195–207. Coon, C. N., K. L. Leske, O. Akavanichan, and T. K. Cheng. 1990. Effect of oligosaccharide-

free soybean meal on true metabolizable energy and fiber digestion in adult roosters. Poult. Sci. 69:787–793.

Cowieson, A. J. 2005. Factors that affect the nutritional value of maize for broilers. Anim. Feed

Sci. Technol. 119:293–305. Dale, N., and H. L. Fuller. 1984. Correlation of protein content of feedstuffs with the magnitude

of nitrogen correction in true metabolizable energy determinations. Poult. Sci. 63:1008–1012.

Daskiran, M., R. G. Teeter, D. Fodge, and H. Y. Hsiao. 2004. An evaluation of endo-β-D-

mannanase (Hemicell) effects on broiler performance and energy use in diets varying in β-mannan content. Poult. Sci. 83:662–668.

Dozier, W. A., and J. B. Hess. 2011. Soybean meal quality and analytical techniques. Pages 111–

124 in Soybean and Nutrition. H. El-Shemy, ed. Accessed Mar. 15, 2014.

Edwards, H. M., M. W. Douglas, C. M. Parsons, and D. H. Baker. 2000. Protein and energy evaluation of soybean meals processed from genetically modified high-protein soybeans. Poult. Sci. 79:525–527.

Fan, M. Z., and W. C. Sauer. 1995. Determination of apparent ileal amino acid digestibility in

barley and canola meal for pigs with the direct, difference, and regression methods. J. Anim. Sci. 73:2364–2374.

Farrell, D. J. 1978. Rapid determination of metabolisable energy of foods using cockerels. Br.

Poult. Sci. 19:303–308. García-Rebollar, P., L. Cámara, R. P. Lázaro, C. Dapoza, R. Pérez-Maldonado, and G. G.

Mateos. 2016. Influence of the origin of the beans on the chemical composition and nutritive value of commercial soybean meals. Anim. Feed Sci. Technol. 221:245–261.

Gitzelmann, R., and S. Auricchio. 1965. The handling of Soya alpha-galactosides by a normal

and galactosemic child. 36:231–235. Grieshop, C. M., C. T. Kadzere, G. M. Clapper, E. A. Flickinger, L. L. Bauer, R. L. Frazier, and

G. C. Fahey. 2003. chemical and nutritional characteristics of United States soybeans and soybean meals. J. Agric. Food Chem. 51:7684–7691.

21 Pa

ge21

Guillaume, J., and J. D. Summers. 1970. Maintenance energy requirement of the rooster and influence of plane of nutrition on metabolizable energy. Can. J. Anim. Sci. 50:363–369.

Halloran, H. R., and I. R. Sibbald. 1979. Metabolizable energy values of fats measured by

several procedures. Poult. Sci. 58:1299–1307. Han, I. K., H. W. Hochstetler, and M. L. Scott. 1976. Metabolizable energy values of some

poultry feeds determined by various methods and their estimation using metabolizability of dry matter. Poult. Sci. 55:1335–1342.

Hill, F. W., and D. L. Anderson. 1958. Comparison of metabolizable energy and productive

energy determinations with growing chicks. J. Nutr. 64:587–603. Jackson, M. E., K. Geronian, A. Knox, J. McNab, and E. McCartney. 2004. A dose-response

study with the feed enzyme β-mannanase in broilers provided with corn-soybean meal based diets in the absence of antibiotic growth promoters. Poult. Sci. 83:1992–1996.

Jagger, S., J. Wiseman, D. J. A. Cole, and J. Craigon. 1992. Evaluation of inert markers for the

determination of ileal and faecal apparent digestibility values in the pig. Br. J. Nutr. 68:129–139.

Jørgensen, H., X. Q. Zhao, K. E. B. Knudsen, and B. O. Eggum. 1996. The influence of dietary

fibre source and level on the development of the gastrointestinal tract, digestibility and energy metabolism in broiler chickens. Br. J. Nutr. 75:379–395.

Kane, E. A., W. C. Jacobson, and L. A. Moore. 1950. A comparison of techniques used in

digestibility studies with dairy cattle. J. Nutr. 41:583–596. Karr-Lilienthal, L. K., L. L. Bauer, P. L. Utterback, K. E. Zinn, R. L. Frazier, C. M. Parsons, and

G. C. Fahey. 2006. Chemical composition and nutritional quality of soybean meals prepared by extruder/expeller processing for use in poultry diets. J. Agric. Food Chem. 54:8108–8114.

Karr-Lilienthal, L. K., C. T. Kadzere, C. M. Grieshop, and G. C. Fahey. 2005. Chemical and

nutritional properties of soybean carbohydrates as related to nonruminants: A review. Livest. Prod. Sci. 97:1–12.

Ketels, E., and G. D. E. Groote. 1989. Effect of ratio of unsaturated to saturated fatty acids of the

dietary lipid fraction on utilization and metabolizable energy of added fats digestibility and AME of fats in young chicks. Poult. Sci. 68:1506–1512.

Kim, B. G., D. Y. Kil, and H. H. Stein. 2013a. In growing pigs, the true ileal and total tract

digestibility of acid hydrolyzed ether extract in extracted corn oil is greater than in intact sources of corn oil or soybean oil. J. Anim. Sci. 91:755–763.

22 Pa

ge22

Kim, E. J., J. L. Purswell, J. D. Davis, I. E. Loar, and K. Karges. 2013b. Live production and carcass characteristics of broilers fed a blend of poultry fat and corn oil derived from distillers dried grains with solubles. Poult. Sci. 92:2732–2736.

Kluth, H., and M. Rodehutscord. 2006. Comparison of amino acid digestibility in broiler

chickens, turkeys, and Pekin ducks. Poult. Sci. 85:1953–1960. Knudsen, K. E. B. 2014. Fiber and nonstarch polysaccharide content and variation in common

crops used in broiler diets. Poult. Sci. 93:2380–2393. Kotb, A.R. and Luckey, T.D. (1972) Markers in nutrition. Nutrition Abstracts and Reviews 42:8

13-845. Krogdahl, Å. 1985. Digestion and absorption of lipids. J. Nutr. 115:675–685. Leeson, S., K. N. Boorman, and D. Lewis. 1977. Metabolisable energy studies with turkeys:

Nitrogen correction factor in metabolisable energy determinations. Br. Poult. Sci. 18:373–379.

Lewis, D. 1989. Fat improves use of other nutrients in poultry diets. Feedstuffs. 33. Loeffler, T., M. Y. Shim, R. B. Beckstead, A. B. Batal, and G. M. Pesti. 2013. Amino acid

digestibility and metabolizable energy of genetically selected soybean products. Poult. Sci. 92:1790–1798.

Lopez, G., and S. Leeson. 2008a. Review: Energy partitioning in broiler chickens. Can. J. Anim.

Sci. 88:205–212. Lopez, G., and S. Leeson. 2008b. Assessment of the nitrogen correction factor in evaluating

metabolizable energy of corn and soybean meal in diets for broilers. Poult. Sci. 87:298–306. Mateos, G. G., and J. L. Sell. 1980. True and apparent metabolizable energy value of fat for

laying hens: influence of level of use. Poult. Sci. 59:369–373. McNab, J. M. 2000. Rapid metabolizable energy assays. Pages 307–315 in Farm Animal

Metabolism and Nutrition. In: Mello,. CABI International, Wallingford, Oxon, United Kingdom.

McNab, J. M., and R. R. Smithard. 1992. Barley !-glucans: An antinutritional factor in poultry

Feeding. Nutr. Res. Rev. 5:45–60. Meloche, K. J., B. J. Kerr, G. C. Shurson, and W. A. Dozier. 2013. Apparent metabolizable

energy and prediction equations for reduced-oil corn distillers dried grains with solubles in broiler chicks from 10 to 18 days of age. Poult. Sci. 92:3176–3183.

23 Pa

ge23

Meloche, K. J., B. J. Kerr, G. C. Shurson, and W. A. Dozier. 2014. Apparent metabolizable energy and predication equations for reduced-oil corn distillers dried grains with solubles in broiler chicks from 10 to 18 days of age. Poult. Sci. 92:3176–3183.

Moran, E. T. 1985. Digestion and absorption of carbohydrates and events through perinatal

development. J. Nutr. 115:665–674. Musharaf, N. A., and J. D. Latshaw. 1999. Heat increment as affected by protein and amino acid

nutrition. Worlds. Poult. Sci. J. 55:233–240. Noblet, J., H. Fortune, X. S. Shi, and S. Dubois. 1994. Prediction of net energy of feeds for

growing pigs. J. Anim. Sci. 72:344–354. NRC. 1994. Nutrient Requirements of Poultry. 9th ed. Natl. Acad. Press, Washington, DC. Odetallah, N. H., P. R. Ferket, J. L. Grimes, and J. L. Mcnaughton. 1997. Effect of mannan-

endo-1,4-β-mannosidase on the growth performance of turkeys fed diets containing 44 and 48 % crude protein soybean meal. Poult. Sci. 81:1322–1331.

Parsons, C. M., L. M. Potter, and B. A. Bliss. 1982. True metabolizable energy corrected to

nitrogen equilibrium. Poult. Sci. 61:2241–2246. Parsons, C. M., Y. Zhang, and M. Araba. 2000. Nutritional evaluation of soybean meals varying

in oligosaccharide content. Poult. Sci. 79:1127–1131. Pereira, L. F. P., and O. Adeola. 1998. Energy and phosphorus values of sunflower meal and rice

bran for broiler chickens using the regression method. Poult. Sci. 95:2081–2089. Perryman, K. R., and W. A. Dozier. 2012. Apparent metabolizable energy and apparent ileal

amino acid digestibility of low and ultra-low oligosaccharide soybean meals fed to broiler chickens. Poult. Sci. 91:2556–2563.

Pesti, G. M., L. O. Faust, H. L. Fuller, N. M. Dale, and F. H. Benoff. 1986. Nutritive value of

poultry by-product meal. 1. Metabolizable energy values as influenced by method of determination and level of substitution. Poult. Sci. 65:2258–2267.

Pettersson, D., and P. Åman. 2007. Enzyme supplementation of a poultry diet containing rye and

wheat. Br. J. Nutr. 62:139. Pirgozliev, V., and S. P. Rose. 1999. Net energy systems for poultry feeds: a quantitative review.

Worlds. Poult. Sci. J. 55:23–36. Potter, L. M., L. D. Matterson, A. W. Arnold, W. J. Pudelkiewicz, and E. P. Singsen. 1960.

Studies in evaluating energy content of feeds for the chick: I. The evaluation of the metabolizable energy and productive energy of alpha cellulose. Poult. Sci. 35:1166–1178.

24 Pa

ge24

Powell, S., V. D. Naranjo, D. Lauzon, T. D. Bidner, L. L. Southern, and C. M. Parsons. 2011. Evaluation of an expeller-extruded soybean meal for broilers. J. Appl. Poult. Res. 20:353–360.

Ravindran, V., M. R. Abdollahi, and S. M. Bootwalla. 2014a. Nutrient analysis, metabolizable

energy, and digestible amino acids of soybean meals of different origins for broilers. Poult. Sci. 93:2567–2577.

Ravindran, V., M. Abdollahi, and S. Bootwalla. 2014b. Nutrient analysis, apparent metabolisable

energy and ileal amino acid digestibility of full fat soybean for broilers. Anim. Feed Sci. Technol. 197:233–240.

Ray, S., M. H. Pubols, and J. McGinnis. 1982. The effect of a purified guar degrading enzyme on

chick growth. Poult. Sci. 61:488–494. Rochell, S. J., T. J. Applegate, E. J. Kim, and W. A. Dozier. 2012. Effects of diet type and

ingredient composition on rate of passage and apparent ileal amino acid digestibility in broiler chicks. Poult. Sci. 91:1647–53.

Rochell, S. J., B. J. Kerr, and W. A. Dozier. 2011. Energy determination of corn co-products fed to broiler chicks from 15 to 24 days of age, and use of composition analysis to predict nitrogen-corrected apparent metabolizable energy. Poult. Sci. 90:1999–2007.

Rodehutscord, M., M. Kapocius, R. Timmler, and A. Dieckmann. 2007. Linear regression approach to study amino acid digestibility in broiler chickens. Br. Poult. Sci. 45:85–92.

Sales, J., and G. P. J. Janssens. 2003. The use of markers to determine energy metabolizability

and nutrient digestibility in avian species. Poult. Sci. 59:314–323. Sell, J. L., A. Krogdahl, and N. Hanyu. 1986. Influence of age on utilization of supplemental fats

by young turkeys. Poult. Sci. 65:546–554. Short, F. J., P. Gorton, J. Wiseman, and K. N. Boorman. 1996. Determination of titanium dioxide

added as an inert marker in chicken digestibility studies. Anim. Feed Sci. Technol. 59:215–221.

Short, F. J., J. Wiseman, and K. N. Boorman. 1999. Application of a method to determine ileal

digestibility in broilers of amino acids in wheat. Anim. Feed Sci. Technol. 79:195–209. Sibbald, I. R. 1976. A bioassay for true metabolizable energy in feedingstuffs. Poult. Sci.

55:303–8. Sibbald, I. R. 1982. Measurement of bioavailable energy in poultry feedstuffs: a review. Can. J.

Anim. Sci. 62:983–1048.

25 Pa

ge25

Sibbald, I. R. 1987. Estimation of bioavailable amino acids in feedingstuffs for poultry and pigs: A review with emphasis on balance experiments. Can. J. Anim. Sci. 67:221–301.

Sibbald, I. R., and S. J. Slinger. 1962. Factors affecting the metabolizable energy content of

poultry feeds: 10. A study of the effect of level of dietary inclusion on metabolizable energy values of several high protein feedingstuffs. Poult. Sci. 41:1282–1289.

Sibbald, I. R., and S. J. Slinger. 1963. A biological assay for metabolizable energy in poultry

feed ingredients together with findings which demonstrate some of the problems associated with the evaluation of fats. Poult. Sci. 42:313–325.

Sibbald, I. R., S. J. Slinger, and G. C. Ashton. 1962. Factors affecting the metabolizable energy

content of poultry feeds: 5. The level of protein and of test material in the diet 6. A note on the relationship between digestible and metabolizable energy values. Poult. Sci. 41:107–116.

Sibbald, I. R., J. D. Summers, and S. J. Slinger. 1960. Factors affecting the metabolizable energy

content of poultry feeds. Poult. Sci. 39:544–556. Siddons, R. C. 1969. Intestinal disaccharidase activities in the chick. Biochem. J. 112:51–59. Sklan, D., A. Smirnov, and I. Plavnik. 2003. The effect of dietary fibre on the small intestines

and apparent digestion in the turkey. Br. Poult. Sci. 44:735–740. Slominski, B. A. 2011. Recent advances in research on enzymes for poultry diets. Poult. Sci.

90:2013–2023. Smeets, N., F. Nuyens, L. Van Campenhout, E. Delezie, J. Pannecoucque, and T. Niewold. 2015.

Relationship between wheat characteristics and nutrient digestibility in broilers: comparison between total collection and marker (titanium dioxide) technique. Poult. Sci. 94:1584–1591.

Stefanello, C., S. L. Vieira, P. Xue, K. M. Ajuwon, and O. Adeola. 2011. Age-related energy

values of bakery meal for broiler chickens determined using the regression method. Poult. Sci. 95:1582–1590.

Su, Y., Y. She, Q. Huang, C. Shi, Z. Li, C. Huang, X. Piao, and D. Li. 2015. The effect of inclusion level of soybean oil and palm oil on their digestible and metabolizable energy content determined with the difference and regression method when fed to growing pigs. Asian-Australasian J. Anim. Sci. 28:1751–1759.

Sulistiyanto, B., Y. Akiba, and K. Sato. 1999. Energy utilisation of carbohydrate, fat and protein

sources in newly hatched broiler chicks. Br. Poult. Sci. 40:653–659. Tancharoenrat, P., V. Ravindran, F. Zaefarian, and G. Ravindran. 2014. Digestion of fat and fatty

acids along the gastrointestinal tract of broiler chickens. Poult. Sci. 93:371–379.

26 Pa

ge26

Tillman, P. B., and P. W. Waldroup. 1988. Assessment of extruded grain amaranth as a feed ingredient for broilers. 1. Apparent amino acid availability values. Poult. Sci. 67:647–65.

Uni, Z., Y. Noy, and D. Sklan. 1995. Posthatch changes in morphology and function of the small

intestines in heavy- and light-strain chicks. Poult. Sci. 74:1622–1629. Van der Klis, J. D. van der, and J. Fledderus. 2007. Evaluation of raw materials for poultry:

What’s up? Eur. Symp. Poult. Nutr.:123–130. Vogtmann, H., H. P. Pfirter, and A. L. Prabucki. 1975. A new method of determining

metabolisability of energy and digestibility of fatty acids in broiler diets. Br. Poult. Sci. 16:531–534.

Vohra, P. 1972. Evaluation of metabolizable energy for poultry. Worlds. Poult. Sci. J. 28:204–

214. Vohra, P., and F. H. Kratzer. 1967. Absorption of barium sulfate and chromic oxide from the

chicken gastrointestinal tract. Poult. Sci. 46:1603–1604. Wagner, D. D., and O. P. Thomas. 1978. Influence of diets containing rye or pectin on the

intestinal flora of chicks. Poult. Sci. 57:971–975. Wang, T., and L. A. Johnson. 2001. Survey of soybean oil and meal qualities produced by

different processes. JAOCS, J. Am. Oil Chem. Soc. 78:311–318. Whitehead, C. C., A. S. Bowman, and H. D. Griffin. 1993. Regulation of plasma oestrogen by

dietary fats in the laying hen: Relationships with egg weight. Br. Poult. Sci. 34:999–1010. Wu, G. 2009. Amino acids: Metabolism, functions, and nutrition. Amino Acids 37:1–17. Wu, G., M. M. Bryant, R. A. Voitle, and D. A. Roland. 2005. Effects of β-mannanase in corn-

soy diets on commercial leghorns in second-cycle hens. Poult. Sci. 84:894–897. Zhang, F., and O. Adeola. 2017. Energy values of canola meal, cottonseed meal, bakery meal,

and peanut flour meal for broiler chickens determined using the regression method. Poult. Sci. 96:397–404.

Zhang, Y. E., and C. M. Parsons. 1993. Effects of extrusion and expelling on the nutritional

quality of conventional and kunitz trypsin inhibitor-free soybeans. Poult. Sci. 72:2299–2308.

Zollitsch, W., W. Knaus, F. Aichinger, and F. Lettner. 1997. Effects of different dietary fat

sources on performance and carcass characteristics of broilers. Anim. Feed Sci. Technol. 66:63–73.

27 Pa

ge27

Zou, X. T., X. J. Qiao, and Z. R. Xu. 2006. Effect of β-mannanase (Hemicell) on growth performance and immunity of broilers. Poult. Sci. 85:2176–2179.

28 Pa

ge28

CHAPTER 2: INFLUENCE OF BASAL DIET TYPE ON REGRESSION-BASED METABOLIZABLE ENERGY VALUES OF DEXTROSE DETERMINED USING

INDEX AND TOTAL COLLECTION METHODS

S. P. West* and S. J. Rochell*2

* Center of Excellence for Poultry Science, University of Arkansas, Fayetteville, AR 72701

ABSTRACT

The metabolizable energy (ME) value of dextrose has been well-characterized in classic

literature, and as such, it is often included as a component of reference diets used in ME assays.

Furthermore, due to its purified and consistent composition, dextrose serves as a good ingredient

to evaluate when comparing different methodological approaches to determine the ME value of

feed ingredients. Thus, an experiment was conducted to evaluate the effect of basal diet type on

the nitrogen-corrected ME (MEn) value of dextrose determined in broiler chicks using the

regression method based on both index and total excreta collection (TC) procedures. Two basal

diet types included a semi-purified (SP) basal based on corn, casein, and dextrose and a more

practical basal based on corn and soybean meal (CSBM). The dextrose was included at 0, 22.5

and 45% at the expense of all energy-providing ingredients in both SP and CSBM diets.

Titanium dioxide (TiO2) was added at 0.5% to all experimental diets as an indigestible marker

for determination of MEn by the index method. Three-hundred and eighty-four male Cobb broiler

chicks were randomly distributed among 48 battery cages (8 birds/cage) and fed a common

starter diet for 14 d. At 14 d post-hatch, 8 replicate cages of chicks were provided 1 of 6

experimental diets until 21 d post-hatch. Total feed intake and excreta output were measured

over 48 h from 19 to 21 d. All diets and representative excreta samples from each cage were

29 Pa

ge29

analyzed for dry matter, nitrogen, gross energy, and TiO2 concentrations. The MEn of dextrose

was determined by difference for each test diet using both index and TC procedures, and the

dextrose-associated caloric intake was regressed against the amount of dextrose intake to

generate linear regression equations with slopes corresponding to the MEn value of dextrose

within each basal type. Based on overlapping 95% confidence intervals of the resulting slopes,

MEn values of dextrose determined using the TC method were not different between the SP

(3,553 kcal/kg) and CSBM (3,588 kcal/kg) basal diet types. Using the index method, there was a

relatively larger difference between the MEn values of dextrose determined in SP basal diets

(3,502 kcal/kg) and in CSBM basal diets (3,839 kcal/kg), but these values also had overlapping

95% confidence intervals. These results indicate that both the index method using TiO2 as an

indigestible marker and TC procedures may be reliably used to characterize the MEn content of

feed ingredients. Furthermore, it is possible to obtain similar estimates of ingredient MEn in

CSBM and SP diets, potentially allowing for simultaneous determination of MEn in amino acid

digestibility assays that utilize SP diets.

Key words: dextrose, semi-purified diet, basal type, metabolizable energy, broiler

INTRODUCTION

Poultry feed formulation relies on accurate estimates of digestible amino acid and

metabolizable energy (ME) contents within individual feed ingredients. Currently, two different

assays are typically used to generate ME and amino acid digestibility data for feed ingredients,

with ME typically determined using corn-soybean meal (CSBM) diets and amino acid

digestibility determined with nitrogen-free semi-purified basal diets (Kluth and Rodehutscord,

2006; Rodehutscord et al., 2007; Adeola and Ileleji, 2009). Thus, concurrent determination of

ME and amino acid digestibility within a single feeding assay would be advantageous, but ME

30 Pa

ge30

values of feed ingredients have been shown to differ between nitrogen-free semi-purified and

CSBM-based basal diets (Adeola and Ileleji, 2009).

As an alternative to determining amino acid digestibility with nitrogen-free semi-purified

diets, amino acid digestibility data has been determined using a regression method that does not

require the use of nitrogen-free basal type diets (Rodehutscord et al., 2007). Multiple inclusion

levels are utilized with the regression approach, which is also beneficial for ME estimates since

inclusion levels can impact ME (Sibbald and Slinger, 1962; Leeson et al., 1977; Mateos and Sell,

1980). For example, Pesti et al. (1986) reported that the nitrogen-corrected ME (MEn) of poultry

by-product meal decreased from 3,330 to 2,970 kcal/kg as the inclusion level increased from 20

to 40% of the diet. Therefore, the regression approach allows for the determination of feed

ingredient ME values at multiple inclusion levels while simultaneously testing the effect of diet

type.

Dextrose may serve as a reliable ingredient to evaluate when comparing different

methodological approaches to determine the MEn value of feed ingredients (Anderson et al.,

1958). Anderson et al. (1958) determined the dextrose MEn to range from 3,500 to 3,740 kcal/kg

with an average value of 3,640 kcal/kg. More recently, researchers have used the value of 3,640

reported by Anderson et al. (1958) for dextrose when including it within reference diets used in

ME assays (Rochell et al., 2012; Meloche et al., 2013). However, few experiments have been

conducted since those of Anderson et al. (1958) to confirm the MEn value of dextrose, which

may vary under different experimental conditions.

The method of excreta collection used to determine ME of poultry feeds can be based on

quantitative, total excreta collection (TC) or on the use of an indigestible marker (i.e., index

method), and differences in ME values determined using these two methods are inconsistent and

31 Pa

ge31

vary considerably within the literature, as reviewed by Sales and Janssens (2003). The use of

indigestible markers mitigates the need to quantitatively collect excreta, potentially eliminating

error arising from excreta contamination with the TC method. Recently, titanium dioxide (TiO2)

has grown in popularity as an inert marker for determining amino acid digestibility and ME

values for feed ingredients using the index method, but to our knowledge, only one publication

exists in which ME values generated using TiO2 and the TC method have been compared

(Smeets et al., 2015). In this paper, Smeets et al. (2015) reported that MEn of complete wheat

based diets showed an overall similar trend for both TC and index methods; however, lower MEn

and P-values were reported for the index method using TiO2 as an indigestible marker. This

discrepancy may be even greater when the ME value of a single ingredient is determined by

difference between reference and experimental diets. Therefore, the objective of the current

study was to investigate the influence of semi-purified (SP) and CSBM diet types on the MEn of

dextrose when determined by regression and to compare ME values using the TC and index

methods.

MATERIALS AND METHODS

All animal care and experimental procedures were approved by the University of

Arkansas Institutional Animal Care and Use Committee before initiation of the experiment.

Bird Husbandry and Experimental Diets

Three-hundred and eighty-four male Cobb broiler chicks were obtained from a

commercial hatchery and randomly distributed among 48 battery cages with raised wire floors in

thermostatically-controlled rooms. A photoperiod of 23L:1D at an intensity of 30 lux was used

for the duration of the experiment. Birds were fed a common starter diet with ad libitum access

to water from 0 to 14 d post-hatch. Room temperature was set at 330C at placement and

32 Pa

ge32

gradually decreased to 240C by 21 d post-hatch. At 14 d of age, all birds were individually

weighed and sorted to equalize average BW among 6 treatment groups. Experimental treatments

consisted of a 2 x 3 factorial arrangement of diet type (SP or CSBM) and dextrose inclusion at 0,

22.5, or 45%. Titanium dioxide was included at 0.5% in experimental treatments as an

indigestible marker for the index method (Short et al., 1996). In both SP and CSBM diets,

dextrose was added at the expense of all energy-providing ingredients (corn, soybean meal, soy

oil, and supplemental amino acids), which were maintained at a constant ratio in all diets (Fan

and Sauer, 1995). Birds were provided experimental treatments from 14 to 21 d post-hatch.

Feeders and birds were weighed to determine BW gain and feed intake from 14 to 21 d post-

hatch. A 48 h total excreta collection was conducted from 19 to 21 d post-hatch, excreta were

weighed, and a homogeneous sample was carefully collected to exclude contamination with feed

or feathers and frozen for subsequent analysis.

Laboratory Analyses

Frozen excreta samples were thawed, lyophilized, and ground using an electric coffee

grinder before analysis. Feed samples were ground to pass through a 1 mm screen (Perten LM

3100, Perten Instruments, Hägersten, Sweden). Feed and excreta samples were analyzed for dry

matter, gross energy (GE), and nitrogen content to determine MEn. For DM determination, diets

and lyophilized excreta samples were dried at 1050C in a drying oven (Isotemp oven, Fisher

Scientific, Pittsburgh, PA) for 24 hours [AOAC Official methods 934.01 (for dry matter)]. Gross

energy and nitrogen analyses of feed and excreta were conducted at the University of Arkansas

Center of Excellence for Poultry Science Central Analytical Laboratory. Gross energy was

determined in a bomb calorimeter (model 6200, Parr Instruments, Moline, IL) and nitrogen was

determined using the combustion method [(AOAC Official Methods 990.03 (for nitrogen)]. The

33 Pa

ge33

indigestible marker TiO2 was determined using the method proposed by Short et al. (1996) with

slight modifications.

Calculations

The ME and MEn values (kcal/kg) for each dietary treatment were calculated according

to the following equations:

METotalCollection (kcal/kg) = [GEdiet – GEexcreta] / FI

MEIndex (kcal/kg) = GEdiet – [GEexcreta × (TiO2,diet / TiO2,excreta)]

where GEdiet and GEexcreta are the analyzed gross energy values (kcal/kg); TiO2diet and TiO2excreta

are the analyzed TiO2 values (%), and FI is the feed intake (kg). A nitrogen correction factor of

8,220 kcal/kg of nitrogen retained was used to determine the MEn and is based on an estimate of

energy used when 1 kg of tissue nitrogen is catabolized (Anderson et al., 1958).

The ME and MEn of dextrose was determined by the difference method for both diet

types. For the SP and CSBM diets, dextrose was added at the expense of all energy-providing

ingredients of the basal diet. The ME and MEn values (kcal/kg) of dextrose were calculated by

the following equation:

MEn,Dextrose (kcal/kg) = [MEn,diet – MEn,basal × BI] / TI

where MEndiet and MEnbasal are the analyzed MEn values (kcal/kg) of the test and basal diets,

respectively, and TI and BI are concentrations (%) of the test ingredient and basal diet in the

experimental diet.

Dextrose intake (kg) was calculated based on total feed intake and dextrose inclusion

level, and dextrose associated caloric intake (kcal) was calculated from the ME (kcal/kg) of

34 Pa

ge34

dextrose multiplied by the dextrose intake (kg) for each given inclusion level. The dextrose

associated caloric intake (kcal) was regressed against the amount of dextrose intake (kg) and

linear regression equations were generated with slopes corresponding to the MEn value (kcal/kg)

of the dextrose for both SP and CSBM diets (Adeola and Ileleji, 2009).

Statistical Analysis

Six experimental treatments consisted of a 2 x 3 factorial arrangement of diet type (SP or

CSBM) by dextrose inclusion (0, 22.5, or 45%). There were 8 replicate cages of each dietary

treatment. Excreta collection method (index or TC) was considered as an additional factor

resulting in a total of 12 treatments. Growth performance, dietary ME and MEn, and nitrogen

retention data were analyzed as a randomized complete block design using a two-way ANOVA

(PROC MIXED, SAS, 2009) to evaluate the effects of diet type, dextrose inclusion level, method

type, and their interaction, with cage location as the blocking factor. Data are presented as least

squares means of treatment groups and statistical significance was considered at P < 0.05.

Linear regression equations were generated using the general linear model procedure of

SAS (2009) whereby the dextrose associated caloric intake (kcal) was regressed against the

amount of dextrose intake (kg) to yield the MEn value (kcal/kg) of the dextrose. Confidence

intervals (95%) were used to compare the intercepts and slopes for each diet type and excreta

collection method

RESULTS AND DISCUSSION

Palatability, influence on intestinal function, and rate of digesta passage are often key

concerns when feeding SP diets in ingredient evaluation assays (Rochell et al., 2012; O’Neill et

al., 2014). Broiler growth performance was not a primary objective of the current study;

35 Pa

ge35

however, BWG and FI did indicate that the birds generally accepted both the SP and CSBM diets

containing graded levels of dextrose (0, 22.5, or 45%). Interestingly, interactions between diet

type and dextrose level for BWG (P < 0.001), FI (P = 0.034), and FCR (P < 0.001) indicated that

increasing dextrose inclusions generally had a negative impact on these measures for birds fed

the CSBM diets, but less so for birds fed the SP diets (Table 2.2). Ultimately, there was less of a

change in diet texture with increasing levels of dextrose for birds fed the SP diets. Ultimately,

there was a greater change in diet texture with increasing levels (0, 22.5, or 45%) of dextrose fed

to birds for the CSBM group compared with the SP group.

Nitrogen retention, ME, and MEn values for the complete experimental diets are

presented in Table 2.3. Three-way interactions (P < 0.05) were observed for all of these

measurements, indicating interactive effects among diet type, dextrose level and excreta

collection method. Total excreta collection yielded numerically higher values for N retention,

ME, and MEn in all cases, but the magnitude and statistical significance of this difference varied

for each combination of diet type and dextrose level. The greatest difference between values

generated using TC and index methods on MEn was observed for the CSBM-0% dextrose diet

which had a TC-derived value that was 462 kcal/kg higher (P < 0.05) than the index value,

whereas the smallest difference was observed for the CSBM-45% dextrose diet for which the

TC-derived ME value was only 57 kcal/kg higher (P > 0.05) than the index value. Similarly,

Smeets et al. (2015) determined that the apparent MEn values of CSBM diets containing various

sources of wheat were higher with the TC method (3,264 kcal/kg) than when using TiO2 as an

inert marker with the index method (3,110 kcal/kg). Smeets et al. (2015) reported recovered TiO2

values determined by the index method to vary from 85.1 to 87.5 % across dietary treatments.

Therefore, variability in recovered TiO2 values determined by the index method may cause a

36 Pa

ge36

treatment effect across resulting ME values. In previous literature it has been recognized that the

agreement between ME values determined by TC and those determined using the index method

varies among the use of different markers (Han et al., 1976; Halloran and Sibbald, 1979; Tillman

and Waldroup, 1988).

With the exception of the diets containing 45% dextrose, nitrogen retention, ME, and

MEn values for the experimental diets were numerically higher for the SP diets compared with

the CSBM-based diets in all cases, although the magnitude and statistical significance of

differences varied by dextrose inclusion and excreta collection method. This is not surprising

given the fact that the calculated MEn of the SP diets were on average 125 kcal/kg higher than

the CSBM diets (Table 2.1), although the observed difference (main effect of diet type) was

greater than expected (295 kcal/kg; P < 0.05). Based on the observed index-based MEn value

(2,937 kcal/kg) of the basal diet (CSBM-0%) and reference MEn value (3,640 kcal/kg) by

Anderson et al. (1958) of dextrose, there was an expected increase from 2,937 to 3,292 kcal/kg

as dextrose inclusion increased from 0 to 45% in the CSBM diets; however, the expected change

was not observed for TC-based diet MEn values. For SP diets, there were similar trends for both

index and TC-based MEn values as dextrose inclusion increased from 0 to 45 %. As such, SP-0%

basal diet MEn values were relatively higher than CSBM-0% basal diet MEn values; therefore,

the discrepancy in trends between diet type may be attributed to an overestimation of MEn values

observed for the SP basal diets.

The ME and MEn values of dextrose determined by difference for each diet type and

excreta collection method are presented in Table 2.4. Two-way interactions (P < 0.05) between

diet type and dextrose level, and dextrose level and excreta collection method were observed for

both ME and MEn, and a diet type × excreta collection method interaction (P < 0.05) was

37 Pa

ge37

observed for ME but not MEn. Regarding the diet type × dextrose level interaction, the ME and

MEn of dextrose decreased by 764 and 590 kcal/kg in the SP diets and by 116 and 76 kcal/kg in

the CSBM diets, respectively, as its dietary inclusion increased from 22.5 to 45%. Similarly,

Pesti et al. (1986) reported that the apparent MEn of poultry by-product meal decreased by 360

kcal/kg from 3,330 to 2,970 kcal/kg as the inclusion level increased from 20 to 40% of the diet.

It has been shown that low inclusion levels may result in an overestimation of ingredient ME

values whereas higher inclusion levels often result in estimators closer to the true value of the

ingredient tested (Sibbald and Slinger, 1962; Leeson et al., 1977; Pesti et al., 1986). Previous

studies have shown that the impact of the test ingredient inclusion level on its determined ME

value likely varies among dietary ingredients and composition of the basal diet (Sibbald and

Slinger, 1962; Leeson et al., 1977; Mateos and Sell, 1980).

In contrast to the complete diets, ME and MEn values of dextrose determined by using the

TC method were lower than values determined by the index method. A diet type × excreta

collection method interaction (P = 0.04) was observed for ME, whereby the TC-based value was

420 kcal/kg lower than the index method for the CSBM diets, but only 160 kcal/kg lower and not

significantly different from the index value within the SP diets. Additionally, a similar trend (P =

0.09) was observed for the diet type × excreta collection method interaction on MEn. The

difference between TC and index-derived ME values for dextrose may be due to the impact of

inclusion level on excreta collection method when ME value of a single ingredient (dextrose)

being determined by the difference method. Therefore, any error associated with analyzed diet

ME and MEn values may be amplified to a greater effect for dextrose ME and MEn values since

they are calculated by difference from the diet ME and MEn values. For the 22.5% dextrose

diets, the ME and MEn values determined by TC were 479 and 416 kcal/kg lower than values

38 Pa

ge38

determined by the index method, whereas TC values were only 100 and 97 kcal/kg lower than

index-generated ME and MEn values of the 45% dextrose diet, resulting in a dextrose level ×

excreta collection method interaction (P < 0.05). Therefore, larger differences between TC and

index-based ME values were observed for dextrose at the 22.5% inclusion level than at the 45%

level, indicating that there is more variability among ME values associated with lower inclusion

levels.

A linear regression approach based on data from multiple inclusion levels of the test

ingredient may reduce the influence of inclusion level on ingredient ME and MEn values

determined by difference (Potter et al., 1960; Sibbald and Slinger, 1962; Berrocoso et al., 2017).

Thus, linear regression of dextrose-associated MEn intake (kcal) on dextrose intake (kg) was

used to generate slopes that corresponded to MEn values determined in both SP and CSBM diet

types for TC and index methods (Table 2.5). Using the TC method, linear regression equations

determined using SP diets were Y = 3,553X – 35.2 (R2 = 0.99), and Y = 3,588X – 4.70 (R2 =

0.99) when based on CSBM diets, reflecting MEn values for dextrose of 3,553 and 3,588 kcal/kg,

respectively. The 95% confidence intervals for MEn regression slopes (i.e. MEn values)

determined using SP and CSBM diet types were 3,387 to 3,720 and 3,426 to 3,751 kcal/kg,

respectively. Using the index method, linear regression equations determined using SP diets were

Y = 3,502X – 76.7 (R2 = 0.97) and Y = 3,839X – 18.6, (R2 = 0.99) when based on CSBM diets,

reflecting MEn values for dextrose of 3,502 kcal/kg and 3,839 kcal/kg, respectively. The 95%

confidence intervals for regression slopes for both SP and CSBM diet types were 3,241 to 3,763

kcal/kg and 3,735 to 3,945 kcal/kg, respectively. Confidence intervals for regression-based MEn

values of dextrose overlapped when determined for both excreta collection methods and SP and

CSBM diets, indicating no statistical differences in estimates of MEn of dextrose between

39 Pa

ge39

method or diet type. Furthermore, the average of all regression-based ME values (3,502 to 3,839

kcal/kg) for dextrose determined in the current study agrees well with the average reference

value (3,640 kcal/kg) for dextrose determined by Anderson et al. (1958). However, the range

highlights that actual values can vary under different experimental conditions and approaches.

In conclusion, the current study demonstrated the impact of different ME methods on

dextrose as a highly digestible reference ingredient for both ME and AA digestibility studies.

The average dextrose MEn value (3,621 kcal/kg) determined by regression in the current study

agreed well with the average reference dextrose ME value (3,640 kcal/kg) by Anderson et al.

(1958). Additionally, the regression method seemed to alleviate error associated with the use of

multiple inclusion levels and the difference method. The influence of different methodologies on

the dextrose MEn value observed in the current study may be even greater for feed ingredients

that present more complex nutrient interactions. Therefore, future research is warranted to test

the additivity in ME values of various feed ingredients when fed as part of complete diets.

40 Pa

ge40

Table 2.1 Ingredient composition of experimental diets1

Dextrose, % SP diets CSBM diets

Item 0 22.5 45 0 22.5 45 Ingredients, %

Corn 61.22 46.55 31.88 42.00 32.18 22.36 Soybean meal 0.00 0.00 0.00 48.50 37.16 25.82 Soybean oil 1.00 0.76 0.52 5.00 3.83 2.66 Casein 22.80 17.34 11.87 0.00 0.00 0.00 Corn-starch 5.00 3.80 2.60 0.00 0.00 0.00 Dextrose 0.00 22.50 45.00 0.00 22.50 45.00 Solkaflock2 2.40 1.82 1.25 0.00 0.00 0.00 Limestone 0.70 0.70 0.70 0.95 0.95 0.95 Dicalcium phosphate 2.20 2.20 2.20 1.60 1.60 1.60 Titanium dioxide 0.50 0.50 0.50 0.50 0.50 0.50 Vitamin premix3 0.10 0.10 0.10 0.10 0.10 0.10 Mineral premix3 0.10 0.10 0.10 0.10 0.10 0.10 Potassium sulfate 1.80 1.80 1.80 0.00 0.00 0.00 Sodium chloride 0.48 0.48 0.48 0.40 0.40 0.40 Choline chloride 0.20 0.20 0.20 0.10 0.10 0.10 Santoquin 0.02 0.02 0.02 0.02 0.02 0.02 Selenium premix, 0.06% 0.02 0.02 0.02 0.02 0.02 0.02 L-Lys·HCl 0.00 0.00 0.00 0.13 0.10 0.07 DL-Met 0.45 0.34 0.23 0.47 0.36 0.25 L-Thr 0.11 0.08 0.06 0.11 0.08 0.06 L-Arg 0.90 0.68 0.47 0.00 0.00 0.00

Calculated composition AMEn, kcal/kg 3,240 3,282 3,325 3,052 3,157 3,262 CP (N × 6.25), % 26.44 20.11 13.77 26.27 20.13 13.98 Digestible Lys, % 1.74 1.32 0.90 1.55 1.19 0.83 Digestible TSAA, % 1.32 1.00 0.69 1.18 0.91 0.63 Digestible Thr, % 1.15 0.87 0.60 1.03 0.79 0.55 Ca, % 0.85 0.83 0.80 0.88 0.84 0.80 Nonphytate P, % 0.44 0.43 0.42 0.46 0.42 0.38

Analyzed composition DM, % 91.22 90.56 90.89 89.67 89.88 90.52 Gross energy, kcal/kg 4,078 3,879 3,652 4,089 3,959 3,741 CP (N × 6.25), % 29.10 22.30 14.80 25.60 20.20 14.30

1Semi-purified = SP; Practical = CSBM. 2Solka Floc, International Fiber Corp., North Tonawanda, NY. 3Supplied the following per kg of diet: vitamin A, 6,173 IU; vitamin D3, 4,409 ICU; vitamin E, 44 IU; vitamin B12, 0.01 mg; menadione, 1.20 mg; riboflavin, 5.29 mg; d-pantothenic acid, 7.94 mg; thiamine, 1.23 mg; niacin, 30.86 mg; pyridoxine, 2.20 mg; folic acid, 0.71 mg; biotin, 0.07 mg; manganese, 24 mg; zinc, 14.4 mg; selenium, 0.04 mg; copper, 0.68 mg; iodine, 0.47 mg.

41 Pa

ge41

Table 2.2 Growth performance of broilers fed semi-purified (SP) or corn-soybean meal (CSBM) diets containing 0, 22.5, or 45% dextrose from 14 to 21 d of age1

Item, % BWG, g FI, g FCR Two-way interaction means (n = 8)

SP 0% 320bc 449d 1.410c SP 22.5% 292c 543c 1.724ab SP 45% 336bc 529c 1.535bc CSBM 0% 480a 650b 1.330c CSBM 22.5% 480a 699a 1.453bc CSBM 45% 367b 674ab 1.840a SEM 13.8 11.3 0.068

Main effect of diet type (n = 24) SP diets 316b 507b 1.556 CSBM diets 443a 674a 1.541 SEM 7.6 6.2 0.038

Main effect of dextrose level (n = 16) 0% 400a 549b 1.370b 22.5% 386a 621a 1.588a 45% 352b 602a 1.687a SEM 9.4 7.7 0.047

P-value Diet type <0.001 <0.001 0.774 Dextrose level 0.002 <0.001 <0.001 Diet type x dextrose level <0.001 0.034 <0.001

a – dMeans within a column with different superscripts are significantly different (P ≤ 0.05). 1Values represent least squares means with 8 birds in each replicate cage.

42 Pa

ge42

Table 2.3 Nitrogen retention and analyzed ME and MEn values of experimental diets containing 0, 22.5, or 45% dextrose fed to broilers from 19 to 21 d of age1

Item, %

Nitrogen retained, % of nitrogen intake

ME, kcal/kg

MEn, kcal/kg Three-way interaction means2 (n = 8)

SP 0% – Index 65.27cd 3,745cd 3,432cd SP 0% –TC 78.26ab 4,085a 3,722a SP 22.5% – Index 81.76ab 3,852bc 3,579ab SP 22.5% – TC 85.36a 3,928b 3,648ab SP 45% – Index 64.30d 3,492fg 3,354d SP 45% – TC 76.52b 3,685de 3,518bc CSBM 0% – Index 55.26e 3,144h 2,937f CSBM 0% – TC 73.58bc 3,669de 3,399cd CSBM 22.5% – Index 63.86d 3,339g 3,150e CSBM 22.5% – TC 75.75b 3,565ef 3,354d CSBM 45% – Index 73.25bc 3,434fg 3,292de CSBM 45% – TC 79.58ab 3,498f 3,349d SEM 1.907 35.1 32.3

Main effect of diet type (n = 48) SP diets 75.25a 3,798a 3,542a CSBM diets 70.21b 3,441b 3,247b SEM 0.745 13.5 12.5

Main effect of dextrose level (n = 32) 0% 68.09c 3,661a 3,373b 22.5% 76.68a 3,671a 3,433a 45% 73.41b 3,527b 3,378b SEM 0.923 16.7 15.4

Main effect of method (n = 48) Index 67.29b 3,501b 3,291b Total collection 78.17a 3,738a 3,498a SEM 0.745 13.6 12.5

P-value Diet type <0.001 <0.001 <0.001 Dextrose level <0.001 <0.001 0.011 Method <0.001 <0.001 <0.001 Diet type x dextrose level 0.006 <0.001 <0.001 Diet type x method 0.223 0.076 0.060 Dextrose level x method 0.006 <0.001 <0.001 Diet type x dextrose level x method 0.019 0.002 0.003

a – hMeans within a column with different superscripts are significantly different (P ≤ 0.05). 1Values represent least squares means with 8 birds in each replicate cage. 2Semi-purified = SP; corn-soybean meal = CSBM.

43 Pa

ge43

Table 2.4 Metabolizable energy and MEn values of dextrose when included in semi-purified (SP) and corn-soybean meal (CSBM) broiler diets at 0, 22.5, or 45% using the total excreta collection (TC) and index method1

Item, % Dextrose ME, kcal/kg Dextrose MEn, kcal/kg Two-way interaction means2 (n = 8) Diet type x dextrose level SP 22.5% 4,248a 4,110a SP 45% 3,484c 3,520c CSBM 22.5% 3,871b 3,783b CSBM 45% 3,755b 3,707bc SEM 63.9 57.6

Diet type x method SP – Index 3,946ab 3,895a SP – TC 3,786bc 3,735ab CSBM – Index 4,023a 3,921a CSBM – TC 3,603c 3,568b SEM 63.9 57.6

Dextrose level x method 22.5% – Index 4,299a 4,155a 22.5% – TC 3,820b 3,739b 45% – Index 3,669bc 3,662b 45% – TC 3,569c 3,565b SEM 63.9 57.6

Main effect of diet type (n = 32) SP diets 3,866 3,815 CSBM diets 3,813 3,745 SEM 44.4 40.0

Main effect of dextrose level (n = 32) 22.5% 4,060a 3,947a 45% 3,619b 3,613b SEM 44.4 40.0

Main effect of method (n = 32) Index 3,984a 3,908a Total collection 3,695b 3,652b SEM 44.4 40.0

P-value Diet type 0.691 0.217 Dextrose level <0.001 <0.001 Method <0.001 <0.001 Diet type x dextrose level <0.001 0.002 Diet type x method 0.041 0.092 Dextrose level x method 0.004 0.006 Diet type x dextrose level x method 0.102 0.182

a,b,cMeans within a column with different superscripts are significantly different (P ≤ 0.05). 1Values represent least squares means with 8 birds in each replicate cage.

44 Pa

ge44

Table 2.5 Intercepts and slopes of regression equations of dextrose-associated MEn intake (kcal) on dextrose intake (kg) determined using semi-purified (SP) or corn-soybean meal (CSBM) diets and total excreta collection (TC) or index methods1

Item Intercept, kcal2 Slope, kcal/kg2 R2 SP – Index 76.7 (-20.5 – 174.0) 3,502 (3,241 – 3,763) 0.97

SEM 46.9 125.8 P-value 0.116 <0.001

SP – TC 35.2 (-27.0 – 97.4) 3,553 (3,387 – 3,720) 0.99 SEM 30.0 80.4 P-value 0.253 <0.001

CSBM – Index 18.6 (-32.5 – 69.6) 3,839 (3,735 – 3,945) 0.99 SEM 24.6 50.6 P-value 0.460 <0.001

CSBM – TC 4.7 (-74.4 – 83.8) 3,588 (3,426 – 3,751) 0.99 SEM 38.1 78.3 P-value 0.903 <0.001

1Equations were generated by regressing the dextrose associated caloric intake (kcal) against the amount of dextrose intake (kg) to yield the ME or MEn value (kcal/kg) of dextrose based on 8 replicate cages with 8 birds per cage. 2Corresponding 95% confidence intervals are represented within parentheses for intercepts and slopes.

45 Pa

ge45

REFERENCES

Adeola, O., and K. E. Ileleji. 2009. Comparison of two diet types in the determination of metabolizable energy content of corn distillers dried grains with solubles for broiler chickens by the regression method. Poult. Sci. 88:579–585.

Anderson, D. L., W. Hill, and R. Renner. 1958. Studies of metabolizable and productive energy

of glucose for the growing chick. J. Nutr. 65:561–574. Berrocoso, J. D., S. Yadav, and R. Jha. 2017. Nitrogen-corrected apparent metabolizable energy

value of macadamia nut cake for broiler chickens determined by difference and regression methods. Anim. Feed Sci. Technol. 234:65–71.

Fan, M. Z., and W. C. Sauer. 1995. Determination of apparent ileal amino acid digestibility in

barley and canola meal for pigs with the direct, difference, and regression methods. J. Anim. Sci. 73:2364–2374.

Halloran, H. R., and I. R. Sibbald. 1979. Metabolizable energy values of fats measured by

several procedures. Poult. Sci. 58:1299–1307. Han, I. K., H. W. Hochstetler, and M. L. Scott. 1976. Metabolizable energy values of some

poultry feeds determined by various methods and their estimation using metabolizability of dry matter. Poult. Sci. 55:1335–1342.

Kluth, H., and M. Rodehutscord. 2006. Comparison of amino acid digestibility in broiler

chickens, turkeys, and pekin ducks. Poult. Sci. 85:1953–1960. Leeson, S., K. N. Boorman, and D. Lewis. 1977. Metabolisable energy studies with turkeys:

Nitrogen correction factor in metabolisable energy determinations. Br. Poult. Sci. 18:373–379.

Mateos, G. G., and J. L. Sell. 1980. True and apparent metabolizable energy value of fat for

laying hens: Influence of level of use. Poult. Sci. 59:369–373. Meloche, K. J., B. J. Kerr, G. C. Shurson, and W. A. Dozier. 2013. Apparent metabolizable

energy and prediction equations for reduced-oil corn distillers dried grains with solubles in broiler chicks from 10 to 18 days of age. Poult. Sci. 92:3176–3183.

O’Neill, H. V. M., G. A. White, D. Li, M. R. Bedford, J. K. Htoo, and J. Wiseman. 2014.

Influence of the in vivo method and basal dietary ingredients employed in the determination of the amino acid digestibility of wheat distillers dried grains with solubles in broilers. Poult. Sci. 93:1178–1185.

Pesti, G. M., L. O. Faust, H. L. Fuller, N. M. Dale, and F. H. Benoff. 1986. Nutritive value of

poultry by-product meal. 1. Metabolizable energy values as influenced by method of determination and level of substitution. Poult. Sci. 65:2258–2267.

46 Pa

ge46

Potter, L. M., L. D. Matterson, A. W. Arnold, W. J. Pudelkiewicz, and E. P. Singsen. 1960. Studies in evaluating energy content of feeds for the chick: I. the evaluation of the metabolizable energy and productive energy of alpha cellulose. Poult. Sci. 35:1166–1178.

Rochell, S. J., T. J. Applegate, E. J. Kim, and W. A. Dozier. 2012. Effects of diet type and

ingredient composition on rate of passage and apparent ileal amino acid digestibility in broiler chicks. Poult. Sci. 91:1647–53.

Rodehutscord, M., M. Kapocius, R. Timmler, and A. Dieckmann. 2007. Linear regression

approach to study amino acid digestibility in broiler chickens. Br. Poult. Sci. 45:85–92. Sales, J., and G. P. J. Janssens. 2003. The use of markers to determine energy metabolizability

and nutrient digestibility in avian species. Poult. Sci. 59:314–323. Short, F. J., P. Gorton, J. Wiseman, and K. N. Boorman. 1996. Determination of titanium dioxide

added as an inert marker in chicken digestibility studies. Anim. Feed Sci. Technol. 59:215–221.

Sibbald, I. R., and S. J. Slinger. 1962. Factors affecting the metabolizable energy content of

poultry feeds: 10. A study of the effect of level of dietary inclusion on metabolizable energy values of several high protein feedingstuffs. Poult. Sci. 41:1282–1289.

Smeets, N., F. Nuyens, L. Van Campenhout, E. Delezie, J. Pannecoucque, and T. Niewold. 2015.

Relationship between wheat characteristics and nutrient digestibility in broilers : comparison between total collection and marker (titanium dioxide) technique. Poult. Sci. 94:1584–1591.

Tillman, P. B., and P. W. Waldroup. 1988. Assessment of extruded grain amaranth as a feed

ingredient for broilers. 1. Apparent amino acid availability values. Poult. Sci. 67:647–651

47 Pa

ge47

CHAPTER 3: INFLUENCE OF BASAL DIET TYPE ON METABOLIZABLE ENERGY VALUES OF AN EXPELLER-EXTRUDED SOYBEAN MEAL DETERMINED IN BROILER CHICKS USING THE REGRESSION METHOD

S. P. West* and S. J. Rochell*2

* Center of Excellence for Poultry Science, University of Arkansas, Fayetteville, AR 72701

ABSTRACT

Basal diets used to determine ileal amino acid digestibility values of feed ingredients

typically contain more purified ingredients than those used to determine metabolizable energy

(ME) values; however, it would be advantageous if both measurements could be determined

using the same basal type in a single assay. An experiment was conducted to determine if ME

and nitrogen-corrected ME (MEn) values of an expeller-extruded soybean meal (EE-SBM)

generated in broiler chicks using the regression method are influenced by basal diet type. Two

diet types included a semi-purified (SP) basal based on corn, casein, and dextrose and a basal

based on corn and soybean meal (CSBM). The EE-SBM was included at 0, 15, 30, and 45% at

the expense of dextrose in the SP diets and at the expense of all energy-providing ingredients in

the CSBM diets. Five-hundred and four male Cobb broiler chicks were randomly distributed

among 72 battery cages (7 birds/cage) and fed a common starter diet for 14 d. At 14 d post-hatch,

8 replicate cages of chicks were provided 1 of 8 experimental diets until 21 d post-hatch. A 48 h

total excreta collection was conducted from 19 to 21 d to determine the MEn of each

experimental diet. The MEn of the EE-SBM within each diet was determined by the difference

method based on its inclusion level and the MEn value of the dietary components it replaced.

The EE-SBM associated caloric intake was regressed against amount of EE-SBM intake in kg to

48 Pa

ge48

generate linear regression equations with slopes corresponding to the ME or MEn value of the

EE-SBM within each basal type. As EE-SBM inclusion increased from 0 to 45%, MEn values of

the complete SP and CSBM diets decreased linearly from 3,438 to 2,942 and 3,122 to 2,784

kcal/kg, respectively. Linear regression of EE-SBM associated MEn intake in kcal against EE-

SBM intake in kg resulted in the following equations: Y = 2,542X – 17, (R2 = 0.98) for the SP

diets and Y = 2,575X – 33, (R2 = 0.97) for the CSBM diets. The resulting EE-SBM MEn values

determined using SP basal diets (2,542 kcal/kg) were similar (P > 0.05) to those determined

using CSBM basal diets (2,575 kcal/kg). These results indicate that both SP and CSBM basal

diets may be reliably used to characterize the MEn content of EE-SBM for broiler chicks when

using the regression method.

Key words: soybean meal, semi-purified diet, regression, metabolizable energy, broiler

INTRODUCTION

Effective least-cost diet formulation of poultry diets requires accurate estimates of the

metabolizable energy (ME) content within available feed ingredients. However, in vivo ME

assays are inherently variable and there is a lack of standardization among bioassays to

determine ME. Metabolizable energy values are typically determined using precision-fed

roosters or growing broiler chicks. The precision-fed rooster assay involves tube-feeding the test

material as the sole ingredient to cecectomized adult white-leghorn roosters, which allows for a

rapid and direct calculation of ME with no influence of ingredient interactions (Parsons, 1986).

Compared with using adult roosters, a key advantage of using a broiler chick assay is that it

allows for ME determination in a bird with a physiological status that better reflects that of

commercially-grown birds (Renner and Hill, 1960; Sibbald, 1976; Sibbald and Wolynetz, 1985).

49 Pa

ge49

However, broiler ME assays are challenged with the difficulty of determining the ME value of a

single ingredient when fed as part of a more complex diet (Sibbald et al., 1960).

Two commonly used methods to calculate the ME value of a test ingredient when fed in a

complex diet are the complete basal and single-ingredient replacement assays (Anderson et al.,

1958; Sibbald and Slinger, 1963). In the complete basal replacement approach, the test ingredient

is added at the expense of all energy contributing components of the reference diet (Sibbald and

Slinger, 1963). The single ingredient replacement approach involves inclusion of a well-

characterized ingredient such as glucose in the reference diet, which is then replaced by the test

ingredient (Anderson et al., 1958). In both methods, the ME value of the test ingredient can be

determined by difference based on the inclusion level of the test ingredient and ME value of the

reference and test diet. With either approach, the estimated ME value of the test ingredient can

be influenced by its inclusion level in the test diet (Sibbald and Slinger, 1962; Sibbald et al.,

1962; Mateos and Sell, 1980). Therefore, one strategy to circumvent this issue is using a

regression method in which the test ingredient is fed at multiple concentrations whereby the

slope from the regression line corresponds to the ME value of the test ingredient (Potter et al.,

1960; Short et al., 1999).

Ileal amino acid digestibility of feed ingredients can also be determined using a

regression method, but compared with ME assays, more purified, highly-digestible ingredients

are typically used in the basal diet so that undigested amino acids recovered at the distal ileum

can be attributed to the test ingredient (Short et al., 1999; Kluth and Rodehutscord, 2006;

Rodehutscord et al., 2007). Therefore, it would be advantageous if a common basal type could be

used that would allow for determination of both ME and amino acid digestibility values of feed

ingredients within a single study. Adeola and Ileleji (2009) reported that the ME value of

50 Pa

ge50

distillers dried grains with solubles determined with the regression method was 176 kcal/kg

greater when using a nitrogen-free basal diet than when using a corn-soybean meal (CSBM)

basal diet. However, determining ileal amino acid digestibility by regression does not

necessitate the use of nitrogen-free basal diets (Short et al., 1999; Kluth and Rodehutscord, 2006;

Rodehutscord et al., 2007), and a less purified diet may yield ME values similar to those

obtained with a CSBM diet.

Expeller-extruded soybean meal (EE-SBM) is an alternative to solvent-extracted soybean

meal (SE-SBM) and can contain up to 10.3% oil content (Zhang and Parsons, 1993), which is

much higher than average oil content (1.9%) typically found in SE-SBM produced in the United

States (García-Rebollar et al., 2016). As such, EE-SBM is a good potential source of both ME

and digestible amino acids in poultry diets (Powell et al., 2011), but published ME values of EE-

SBM for broilers are sparse. Thus, the objective of this experiment was to evaluate the influence

of basal diet type on the ME and nitrogen-corrected ME (MEn) values of a commercially

available EE-SBM determined using the regression method in growing broilers.

MATERIALS AND METHODS

All animal care and experimental procedures were approved by the University of

Arkansas Institutional Animal Care and Use Committee before initiation of the experiment.

Bird Husbandry and Experimental Diets

Five-hundred and four male Cobb broiler chicks were obtained from a commercial

hatchery and randomly distributed among 72 battery cages with raised wire floors in

thermostatically-controlled rooms. A photoperiod of 23L:1D at an intensity of 30 lux was used

for the duration of the experiment. The birds were fed a common starter diet with ad libitum

51 Pa

ge51

access to water from 0 to 14 d post-hatch. Room temperature was set at 330C at placement and

gradually decreased to 240C by 21 d post-hatch. At 14 d of age, all birds were individually

weighed and sorted to equalize average BW among 9 treatment groups. Experimental treatments

consisted of a 2 x 4 factorial arrangement of diet type (SP or CSBM) and EE-SBM inclusion (0,

15, 30, or 45%) with an additional CSBM diet containing a test SE-SBM rather than EE-SBM

added at 45% to validate the ME assay. In the SP diets, EE-SBM was added at the expense of

glucose whereas all other ingredient inclusion levels were held constant. For the CSBM diets,

EE-SBM was added at the expense of all energy-providing ingredients (corn, soybean meal, soy

oil, and supplemental amino acids) which were maintained at a constant ratio in all diets (Fan

and Sauer, 1995). Birds were provided experimental treatments from 14 to 21 d post-hatch.

Feeders and birds were weighed to determine BW gain and feed intake from 14 to 21 d post-

hatch. A 48 h total excreta collection was conducted from 19 to 21 d post-hatch and a

representative excreta sample was carefully collected to avoid contamination with feed or

feathers. Excreta samples were frozen and stored for subsequent analysis.

Laboratory Analyses

Frozen excreta samples were thawed, lyophilized, and ground using an electric coffee

grinder, and feed samples were ground to pass through a 1 mm screen (Perten LM 3100, Perten

Instruments, Hägersten, Sweden). Feed and excreta samples were analyzed for dry matter, gross

energy (GE), and nitrogen content to determine ME and MEn. For DM determination, diet and

lyophilized excreta samples were dried at 1050C in a drying oven (Isotemp oven, Fisher

Scientific, Pittsburgh, PA) for 24 hours (AOAC Official methods 934.01). Gross energy and

nitrogen analyses of feed and excreta were conducted at the University of Arkansas Center of

Excellence for Poultry Science Central Analytical Laboratory. Gross energy was determined in a

52 Pa

ge52

bomb calorimeter (model 6200, Parr Instruments, Moline, IL) and nitrogen was determined using

the combustion method [(AOAC Official Methods 990.03 (for nitrogen)]. The EE-SBM was also

analyzed for DM, GE, crude fat (CF), and nitrogen content as described above.

Calculations

The ME and MEn values (kcal/kg) for each dietary treatment were calculated according

to the following equations:

ME (kcal/kg) = [GEdiet – GEexcreta] / FI

MEn (kcal/kg) = [GEdiet – (GEexcreta + (Ndiet – Nexcreta × 8,220))] / FI

where GEdiet and GEexcreta are the analyzed gross energy values (kcal/kg); Ndiet and Nexcreta are the

analyzed nitrogen intake and output (kg), respectively, and FI is the feed intake (kg). A nitrogen

correction factor of 8,220 kcal/kg of nitrogen retained was used to determine MEn and is based

on an estimate of the energy required when 1 kg of tissue nitrogen is catabolized (Anderson et

al., 1958).

The ME and MEn of EE-SBM were determined by the difference method for both diet

types. For the CSBM diets, EE-SBM was added at the expense of all energy-providing

ingredients within the basal diet, and the ME and MEn values (kcal/kg) of EE-SBM were

calculated by the following equation:

MEn,EE-SBM (kcal/kg) = [MEn,diet – MEn,basal × BI%] / (TI%)

where MEn,diet and MEn,basal are the analyzed ME or MEn values (kcal/kg) of the test and basal

diets, respectively, and TI and BI are concentrations of the test ingredient and basal diet in the

experimental diet.

53 Pa

ge53

For SP diets, EE-SBM was added at the expense of dextrose and the ME and MEn values

(kcal/kg) of EE-SBM were calculated as follows:

MEn,EE-SBM (kcal/kg) = [MEn,diet – (MEn,basal + (3,640 × DI%))] / (TI%)

The SP dietary treatments were substituted directly for dextrose at four inclusion levels (0, 15,

30, or 45%); therefore, MEn (basal) is the portion of basal contributing energy and DI is the

inclusion level of dextrose in the basal. The energy contribution from dextrose of 3,640 kcal/kg

(Anderson et al., 1958) was subtracted from the control diet based no its inclusion level of 45%

to obtain the remaining basal energy values of 1,968 kcal/kg and 1,800 kcal/kg for ME and MEn,

respectively.

EE-SBM intake (kg) was calculated based on total feed intake and EE-SBM inclusion

level, and the EE-SBM associated caloric intake (kcal) was calculated from the ME (kcal/kg) of

EE-SBM multiplied by the EE-SBM intake (kg) for each given inclusion level. The EE-SBM

associated caloric intake was regressed against the amount of EE-SBM intake and linear

regression equations were generated with slopes corresponding to the ME or MEn value of the

EE-SBM for both SP and CSBM diets (Adeola and Ileleji, 2009).

Statistical Analysis

Eight treatments included a 2 x 4 factorial arrangement of diet type (SP or CSBM) and

EE-SBM inclusion (0, 15, 30, or 45%), and an additional treatment included a diet in which a

test SE-SBM was included at 45% in the CSBM diets (Table 3.1). There were 8 replicate cages

of each dietary treatment. Growth performance, dietary ME and MEn, and nitrogen retention data

were analyzed as a randomized complete block design using a two-way ANOVA (PROC

MIXED, SAS, 2009) to evaluate the effects of diet type, EE-SBM inclusion level, and their

54 Pa

ge54

interaction, and cage location was the blocking factor. A separate two-way ANOVA was used to

analyze ME and MEn values of EE-SBM in a 2 x 3 factorial arrangement of diet type (SP or

CSBM) and EE-SBM inclusion (15, 30, or 45%). Linear and quadratic polynomial contrasts

were also used to assess the effects EE-SBM inclusion level within diet type. Additionally, single

degree of freedom orthogonal contrasts were made between CSBM treatments containing SE-

SBM and EE-SBM at the 45% level to compare these ingredients. Data are presented as least

squares means of treatment groups and statistical significance was considered at P < 0.05.

Linear regressions were conducted using the GLM procedures of SAS (2009) where the

EE-SBM associated caloric intake (kcal) was regressed against the amount of EE-SBM intake

(kg) to yield the ME or MEn value (kcal/kg) of the EE-SBM. Confidence intervals (95%) were

used to statistically compare the intercepts and slopes of regression equations generated for each

diet type.

RESULTS

The EE-SBM used in this experiment contained 4,888 kcal/kg of gross energy, 47.5%

CP, 6.95% EE and 0.4% moisture on a DM basis, compared with 4,534 kcal/kg of gross energy,

52.2% CP, 1.13% EE and 0.9% moisture on a DM basis for the SE-SBM (data not shown).

Growth performance results for broilers fed the experimental diets from 14 to 21 d of age are

presented in Table 3.2. There was an interaction (P < 0.05) between diet type and EE-SBM level

on BWG of broilers, although quadratic responses to EE-SBM level were observed with the

numerically-highest BWG occurring for broilers fed 15% EE-SBM in both diet types (Table

3.2). Feed intake was greater (P < 0.05) for birds fed the CSBM diets than for those fed the SP

diets and decreased (P < 0.05) as EE-SBM inclusion increased, with no interaction observed (P >

0.05) between diet type and EE-SBM level. Diet type did not influence FCR of broilers (P >

55 Pa

ge55

0.05), and FCR decreased (P < 0.05) quadratically (P < 0.001) for SP diets and linearly (P <

0.05) for CSBM diets as EE-SBM inclusion increased from 0 to 45%.

There was a diet type × EE-SBM level interaction (P < 0.05) on nitrogen retention. There

was a quadratic increase (P < 0.001) in nitrogen retention with the numerically highest value

occurring for broilers fed 30% EE-SBM in the SP diet with no response (P > 0.05) observed for

CSBM diets (Table 3.3). The inclusion level of EE-SBM had an overall effect on nitrogen

retention (P < 0.001). Metabolizable energy content of the experimental diets decreased linearly

(P < 0.001) as EE-SBM inclusion increased from 0 to 45% in the SP diets, but decreased

quadratically (P < 0.05) for birds fed the CSBM diets, which led to a diet type × EE-SBM level

interaction (P < 0.05) on ME. Nitrogen-corrected ME of both SP and CSBM diets decreased

linearly (P < 0.001) as the inclusion level of EE-SBM increased. However, the magnitude of the

linear decrease was greater for birds bed the CSBM diets than for those fed the SP diets, leading

to a diet type × EE-SBM level interaction (P < 0.05) for MEn.

The ME and MEn values of the EE-SBM determined by the difference method within

each diet type are presented in Table 3.4. Similar to ME values of the experimental diets, there

were interactions (P < 0.05) between diet type and EE-SBM inclusion level for both ME and

MEn of EE-SBM. Within the CSBM diets, the ME of EE-SBM increased quadratically (P <

0.05) from 2,305 to 2,746 kcal/kg as its inclusion level of increased from 15 to 45%, whereas

there was no effect of inclusion level (P > 0.05) on ME of EE-SBM when determined in SP

diets. As the inclusion level of EE-SBM increased from 15 to 45%, MEn values of EE-SBM

increased linearly (P < 0.05) from 2,293 to 2,538 kcal/kg when determined in SP diets, but

quadratically from 1,921 to 2,506 kcal/kg when determined in CSBM diets. The ME and MEn

values determined for the test SE-SBM included at 45% in the CSBM diet were 2,731 and 2,500

56 Pa

ge56

kcal/kg, respectively and did not differ (P > 0.05) from the ME and MEn values of the EE-SBM

determined at a 45% inclusion level.

Linear regression of EE-SBM-associated ME or MEn intake (kcal) on EE-SBM intake

(kg) was used to generate slopes that corresponded to ME or MEn values determined in both SP

and CSBM diet types (Table 3.5). For ME, the linear regression equation determined using SP

diets was Y = 2,657X – 5.79 (R2 = 0.98), reflecting a ME value of 2,657 kcal/kg for the EE-

SBM. Using data generated in the CSBM diets, the equation was Y = 2,814X – 22.17 (R2 =

0.97), which resulted in an ME value of 2,814 kcal/kg for the EE-SBM. The overlapping 95%

confidence intervals of the regression slopes indicated that the ME values calculated using this

approach were similar between the SP and CSBM diet types. For MEn, the regression equation

for the SP diet was Y = 2,542X – 17.98, (R2 = 0.98), yielding a MEn value of 2,542 kcal/kg for

EE-SBM. For the CSBM diet, the equation was Y = 2,575X – 31.99, (R2 = 0.97), resulting in a

MEn value of 2,575 kcal/kg for EE-SBM. Similar to ME, the 95% confidence intervals of the

MEn regression slopes determined using SP and CSBM diets overlapped, indicating no statistical

differences in estimates of MEn of EE-SBM between diet types.

DISCUSSION

The aim of the current study was to investigate the effect of two diet types on the ME and

MEn values of an EE-SBM determined by the regression method and to contribute to the limited

number of in vivo ME values of EE-SBM reported for broilers. The influence of EE-SBM and

diet type on growth performance was not a primary objective of this study, but it is important to

note that in general the diets were palatable and supported acceptable growth performance of the

birds. Poor palatability and nutrient imbalance are often key concerns when semi-purified diets

are fed (Rochell et al., 2012) or when assays involve feeding a high inclusion level of a single

57 Pa

ge57

ingredient to growing chicks (Sibbald et al., 1962; Mateos and Sell, 1980). Feed intake in the

current experiment did decrease as EE-SBM increased from 15 to 45%. This may have been

partly due to the concurrent increase in CP of the experimental diets, as Jackson et al. (1981)

observed that FI decreased as dietary protein content increased to 36%. Increased levels of

dietary fiber and fat also have been reported to cause a decrease in feed intake due to induced

satiety and a slower rate of passage (Vermeersch and Vanschoubroek, 1968; Mateos and Sell,

1981; Mateos et al., 2012).

In addition to the diets containing EE-SBM, the current study included one additional

CSBM treatment that contained a test SE-SBM rather than EE-SBM at 45% to compare with

previously-published MEn values of this ingredient to validate the experimental approach. The

ME and MEn values of this SE-SBM were determined to be 2,731 and 2,500 kcal/kg,

respectively. Perryman and Dozier (2012) reported the MEn content of two conventional SE-

SBM that were produced in different years but obtained from the same geographical location to

be 2,073 and 2,241 kcal/kg in growing broilers. Similarly, Lopez and Leeson (2008) determined

the MEn value of a SE-SBM at various inclusions levels to range from 2,170 to 2,383 kcal/kg for

broilers. Using adult roosters, Parsons et al. (2000) reported the true MEn (TMEn) value of SE-

SBM to be 2,739 kcal/kg, which was in close agreement with the TMEn value of 2,794 kcal/kg

reported by Coon et al. (1990). Therefore, the MEn value obtained for the SE-SBM in the current

study was somewhat higher than those previously reported for broilers, but lower than the TMEn

values determined in adult roosters. Nonetheless, the MEn of SE-SBM reported herein appears to

be reasonable and indicates that the experimental procedures and difference method employed

were suitable for the primary objective of determining the ME and MEn of EE-SBM.

58 Pa

ge58

Metabolizable energy values are typically corrected to zero nitrogen retention so that

comparisons can be made across experiments in which birds are in different states of nitrogen

utilization (Lopez and Leeson, 2008). In particular, nitrogen retention can vary with bird age,

type (i.e., leghorn versus broiler), physiological state (i.e., growth versus maintenance) and

protein content of the diet and test ingredient (Leeson et al., 1977; Dale and Fuller, 1984; Lopez

and Leeson, 2008). Nitrogen retention was expected to cause large differences in ME and MEn in

the current study, due to high levels of CP within experimental treatments containing 30 and

45% EE-SBM for SP and CSBM diets. On average the MEn values presented in Table 3.4

decreased by 5 and 8.5% when determined by regression for SP and CSBM diets, respectively.

Lopez and Leeson (2008) reported that SE-SBM apparent MEn values were 7 to 12% lower than

the apparent ME values of SE-SBM when determined using a CSBM basal diet. In addition,

Zhang and Adeola (2017) reported 10, 15 and 19% reductions in ME values for peanut flower

meal, cottonseed meal, and canola meal determined in broilers after correcting for nitrogen,

respectively. Although correcting for nitrogen retention resulted in lower MEn values compared

with ME values in the current study there was more variability associated among ME values as

indicated by a larger SEM.

The determined ME value of the test ingredient can be influenced by its inclusion level in

the test diet (Pesti et al., 1986; Lopez and Leeson, 2008). This study utilized a regression

approach that allowed for the EE-SBM to be substituted at multiple levels, which may provide a

more accurate estimate compared with approaches based on a single inclusion level (Leeson et

al., 1977). The variability of determined ME values for ingredients is dependent on the

proportion of the test ingredient substituted. Low inclusion levels have been reported to cause

variability among ME values (Mateos and Sell, 1980); on the other hand, high inclusion levels

59 Pa

ge59

can induce nutritional imbalances based on the nutritional composition of the ingredient (Sibbald

and Slinger, 1962). The MEn content of EE-SBM in the current study was 2,293 kcal/kg (SP) and

1,921 kcal/kg (CSBM) when fed at 15% of the diet compared with 2,538 kcal/kg (SP) and 2,506

kcal/kg (CSBM) feeding at 45% of the diet. Similarly, Lopez and Leeson (2008) reported the

apparent MEn values of SBM to be 1,986, 2,286, and 2,477 kcal/kg when SBM was included at

10, 20, and 30%, respectively, at the expense of a complete corn-SBM based basal and fed to

broilers from 30 to 33 d of age.

Using the regression approach, the ME and MEn values of EE-SBM were found to be

2,657 and 2,542 kcal/kg using the SP diets and 2,814 and 2,575 kcal/kg using the CSBM diets,

respectively. The EE-SBM used in the current study contained (DM basis) 4,888 kcal/kg gross

energy, 47.5% CP, 6.95% EE, and 0.4% moisture and was similar in composition to other EE-

SBM characterized in the literature (Coon et al., 1990; Zhang and Parsons, 1993; Woodworth et

al., 2001; Karr-Lilienthal et al., 2006; Opapeju et al., 2006; Baker and Stein, 2009). Only a few

reports of in vivo MEn values of EE-SBM for poultry are currently available in the literature, and

to our knowledge, there are no published experiments in which the MEn of EE-SBM were

determined in growing broilers. Zhang and Parsons (1993) reported that the TMEn content of an

EE-SBM subjected to various processing temperatures ranged between 3,125 and 3,239 kcal/kg

DM in cecectomized adult roosters. Similarly, Coon et al. (1990) determined the TMEn content

of EE-SBM to be 3,368 kcal/kg DM in adult roosters. Additionally, Powell et al. (2011) used

prediction equations developed by Janssen (1989) to calculate the MEn content of an EE-SBM to

be 2,882 kcal/kg DM.

While it is not surprising that MEn value of 2,559 kcal/kg (average of regression-based

values determined in the CSBM and SP diets) for EE-SBM determined in the current study using

60 Pa

ge60

growing broilers was lower than previously-reported values mentioned above determined in

adult roosters, a higher determined value was expected based on the ether extract content of the

EE-SBM. The MEn of CSBM experimental diets containing the 45% SE-SBM or EE-SBM were

2,735 and 2,784 kcal/kg, which resulted in MEn values of these individual ingredients to be

2,500 and 2,506 kcal/kg when determined by difference at this inclusion level. This is despite the

fact that the EE-SBM contained 5.82 percentage units more ether extract than the SE-SBM. One

potential explanation of the lower than expected MEn content of the EE-SBM could have been

based on the utilization of the lipid within the EE-SBM. In growing pigs, it has been shown that

the digestibility of supplemental soy oil was greater than intact soy oil derived from SBM

(Cervantes-Pahm and Stein, 2008). Therefore, the broilers may not have been able to utilize the

full amount of dietary fat encapsulated in cell membranes or fibrous compounds within the EE-

SBM. In addition, EE-SBM can contain a greater amount of indigestible fiber compared with

SE-SBM due to EE-SBM not dehulled before extruding and expeller processes (Baker and Stein,

2009). However, the cause for the lower than expected ME value of EE-SBM in the current

study remains unknown.

From a practical perspective, basal diets that are comprised of feed ingredients that the

birds will be fed in practice are advantageous. While CSBM basal diets are often used in ME

assays, determination of ileal amino acid digestibility requires more purified, highly-digestible

ingredients to minimize the confounding effects of undigested amino acids in the terminal ileum

that are not attributed to the test ingredient (Short et al., 1999; Kluth and Rodehutscord, 2006;

Rodehutscord et al., 2007). This is often accomplished using a nitrogen-free basal diet, but

regression-based assays to determine ileal amino acid digestibility have utilized less purified

diets that have relatively better palatability and nutritional balance, particularly for essential

61 Pa

ge61

amino acids (Short et al., 1999; Rodehutscord et al., 2007). The SP diets used in the current study

contained 27.26% corn, 14.09% casein, and supplemental Lys, Met, Thr, and Arg to meet dietary

essential AA requirements. Despite differences in ingredient composition SP and CSBM diets

produced similar ME and MEn values for an EE-SBM in the current study. It appears that the SP

diets could be used to generate reliable ME values, but further research is warranted to evaluate

this comparison for various feed ingredients and to test additivity of ME values when fed as part

of complete diets.

One key difference between the CSBM and SP diet-based approaches used in the current

study is the portion of basal diet replaced by the test ingredient. The EE-SBM was included the

expense of all energy-providing ingredients in the CSBM diets and at the expense of dextrose in

the SP diets. It is important to note that the reference value (3,640 kcal/kg DM) determined by

Anderson et al. (1958) used for dextrose may have had an impact on the resulting MEn value of

EE-SBM determined in SP diets if it was not actually 3,640 kcal/kg. In another study by our lab,

the MEn of dextrose fed to broilers at inclusions of 0, 22.5, and 45% ranged from 3,553 to 3,502

kcal/kg and 3,588 to 3,839 kcal/kg for SP and CSBM diet types, respectively. Therefore, it may

be ideal to determine dextrose under the exact experimental conditions used; however, the

average value (3,621 kcal/kg DM) observed agreed well with the average reference value (3,640

kcal/kg DM) determined by Anderson et al. (1958).

In summary, the ME and MEn values of EE-SBM determined for SP and CSBM diets

were dependent on the inclusion of the EE-SBM in the test diets. In our study, similar MEn

values of a commercially available EE-SBM were observed for both SP and CSBM diets,

indicating the potential to utilize a single diet type to determine both ME and AA digestibility

62 Pa

ge62

values within a single study. Further research should be conducted to validate the use of SP diets

to simultaneously measure ME and AA digestibility values of various feed ingredients.

63 Pa

ge63

Table 3.1 Ingredient composition of experimental diets1

Expeller-extruded soybean meal (EE-SBM), % SP diets CSBM diets2

Item 0 15 30 45 0 15 30 45 Ingredients, % as-fed

Corn 27.26 27.26 27.26 27.26 63.97 53.97 43.96 33.96 Soybean meal 0.00 0.00 0.00 0.00 29.03 24.49 19.95 15.41 EE-SBM 0.00 15.00 30.00 45.00 0.00 15.00 30.00 45.00 Soybean oil 1.00 1.00 1.00 1.00 2.50 2.11 1.72 1.33 Casein 14.09 14.09 14.09 14.09 0.00 0.00 0.00 0.00 Dextrose 45.00 30.00 15.00 0.00 0.00 0.00 0.00 0.00 Solkaflock3 5.90 5.90 5.90 5.90 0.00 0.00 0.00 0.00 Limestone 0.85 0.85 0.85 0.85 1.05 1.05 1.05 1.05 Dicalcium phosphate 2.35 2.35 2.35 2.35 1.80 1.80 1.80 1.80 Titanium dioxide 0.50 0.50 0.50 0.50 0.50 0.50 0.50 0.50 Vitamin premix4 0.10 0.10 0.10 0.10 0.10 0.10 0.10 0.10 Mineral premix5 0.10 0.10 0.10 0.10 0.10 0.10 0.10 0.10 Potassium sulfate 1.20 1.20 1.20 1.20 0.00 0.00 0.00 0.00 Sodium chloride 0.48 0.48 0.48 0.48 0.40 0.40 0.40 0.40 Choline chloride 0.20 0.20 0.20 0.20 0.10 0.10 0.10 0.10 Santoquin 0.02 0.02 0.02 0.02 0.02 0.02 0.02 0.02 Se premix, 0.06%6 0.02 0.02 0.02 0.02 0.02 0.02 0.02 0.02 L-Lys·HCl 0.00 0.00 0.00 0.00 0.12 0.10 0.08 0.06 DL-Met 0.29 0.29 0.29 0.29 0.25 0.21 0.17 0.13 L-Thr 0.08 0.08 0.08 0.08 0.05 0.04 0.03 0.02 L-Arg 0.56 0.56 0.56 0.56 0.00 0.00 0.00 0.00

Calculated composition AMEn, kcal/kg 3,199 3,016 2,833 2,650 3,100 2,978 2,856 2,734 CP (N × 6.25), % 16.65 23.55 30.45 37.35 18.81 22.77 26.73 30.69 Digestible Lys, % 1.05 1.43 1.80 2.18 1.05 1.26 1.48 1.69 Digestible TSAA, % 0.79 0.97 1.14 1.31 0.79 0.84 0.89 0.93 Digestible Thr, % 0.70 0.92 1.14 1.36 0.69 0.81 0.92 1.03 Ca, % 0.90 0.90 0.90 0.90 0.89 0.87 0.85 0.85 Nonphytate P, % 0.45 0.45 0.45 0.45 0.45 0.43 0.41 0.40

Analyzed composition DM, % 86.60 89.20 90.90 91.70 89.50 91.70 90.50 91.60 Gross energy, kcal/kg 3,686 3,869 4,059 4,231 3,929 3,981 4,067 4,113 CP (N × 6.25), 16.40 23.30 30.00 36.50 18.60 23.20 27.20 30.50

1Semi-purified = SP; Practical = CSBM. 2CSBM series consist of an additional treatment containing a solvent-extracted SBM at 45% rather than expeller-extruded SBM. 3Solka Floc, International Fiber Corp., North Tonawanda, NY. 4Supplied the following per kg of diet: vitamin A, 6,173 IU; vitamin D3, 4,409 ICU; vitamin E, 44 IU; vitamin B12, 0.01 mg; menadione, 1.20 mg; riboflavin, 5.29 mg; d-pantothenic acid, 7.94 mg; thiamine, 1.23 mg; niacin, 30.86 mg; pyridoxine, 2.20 mg; folic acid, 0.71 mg; biotin, 0.07 mg. 5Supplied the following per kg of diet: manganese, 10.00 mg; zinc, 10.00 mg; copper, 1.00 mg; iodine, 0.1 mg; iron, 5.00 mg; magnesium, 2.7 mg; calcium, 5.50 mg. 6Supplied 0.12 mg of selenium per kg of diet.

64 Pa

ge64

Table 3.2 Growth performance of broilers fed experimental diets containing 0, 15, 30, or 45% expeller-extruded soybean meal (EE-SBM) from 14-21 d post-hatch1

Item2 BWG, g FI, g FCR SP diets

0% 345 542 1.564 15% 416 570 1.327 30% 409 512 1.248 45% 359 506 1.424

CSBM diets 0% 409 599 1.473 15% 416 567 1.354 30% 394 528 1.328 45% 362 509 1.376 45% SE-SBM3 359 543 1.508

SEM 8.9 9.4 0.030 P-value4

Diet type 0.023 0.004 0.698 EE-SBM level 0.001 0.001 <0.001 Diet type x EE-SBM level 0.046 0.563 0.113 SE-SBM vs. EE-SBM 0.829 0.015 0.001 SP-linear 0.304 <0.001 <0.001 SP-quadratic <0.001 0.055 <0.001 CSBM-linear <0.001 <0.001 0.018 CSBM-quadratic <0.001 <0.001 0.200

1Values are least squares means of 8 replicate cages with 7 birds per cage. 2Semi-purified = SP; corn-soybean meal = CSBM. 3SE-SBM = solvent-extracted soybean meal. 4Probability values of the main effects of diet type, interaction of diet type and EE-SBM level, and linear and quadratic contrasts.

65 Pa

ge65

Table 3.3 Nitrogen retention and metabolizable energy values of experimental diets fed to broilers and containing 0, 15, 30, or 45% expeller-extruded soybean meal (EE-SBM)1

Item2

Nitrogen retention, g

ME, kcal/kg

MEn, kcal/kg

SP diets 0% 22.70 3,607 3,438 15% 29.12 3,444 3,236 30% 30.86 3,306 3,078 45% 28.23 3,159 2,942

CSBM diets 0% 26.74 3,293 3,122 15% 30.34 3,126 2,926 30% 28.033 3,135 2,932 45% 28.033 2,983 2,784 45% SE-SBM3 28.95 2,916 2,735

SEM 1.451 26.5 23.4 P-value4

Diet type 0.579 <0.001 <0.001 EE-SBM level <0.001 <0.001 <0.001 Diet type x EE-SBM level 0.047 0.005 0.001 SE-SBM vs. EE-SBM 0.549 0.104 0.178 SP-linear <0.001 <0.001 <0.001 SP-quadratic <0.001 0.773 0.164 CSBM-linear 0.923 <0.001 <0.001 CSBM-quadratic 0.182 0.022 0.306

1Values are least squares means of 8 replicate cages with 7 birds per cage. 2Semi-purified = SP; corn-soybean meal = CSBM. 3SE-SBM = solvent-extracted soybean meal. 4Probability values of the main effects of diet type, interaction of diet type and EE-SBM level, and linear and quadratic contrasts.

66 Pa

ge66

Table 3.4 Metabolizable energy values of an expeller-extruded soybean meal (EE-SBM) fed to broilers at an inclusion rate of 0, 15, 30, or 45%1

Item2 ME, kcal/kg MEn, kcal/kg SP diets

15 % 2,558 2,293 30 % 2,639 2,441 45 % 2,646 2,538

CSBM diets 15 % 2,305 1,921 30 % 2,907 2,622 45 % 2,746 2,506 45% SE-SBM3 2,731 2,500

SEM 99.3 89.4 P-value4

Diet type 0.600 0.258 EE-SBM level 0.001 <0.001 Diet type x EE-SBM level 0.021 0.005 SE-SBM vs. EE-SBM 0.939 0.978 L5 0.472 0.031 Q5 0.728 0.794 L6 0.002 0.446 Q6 0.001 <0.001

1Values are least squares means of 8 replicate cages with 7 birds per cage. 2Semi-purified = SP; corn-soybean meal = CSBM. 3SE-SBM = solvent-extracted soybean meal. 4Probability values of the main effects of diet type, interaction of diet type and EE-SBM level, and linear (L) and quadratic (Q) contrasts. 5Linear (L) and quadratic (Q) contrasts for the SP diets. 6Linear (L) and quadratic (Q) contrasts for the CSBM diets.

67 Pa

ge67

Table 3.5 Regression equations used to determine ME and MEn of an expeller-extruded soybean meal (EE-SBM) in semi-purified (SP) and corn-soybean meal (CSBM) based diets1

Item Intercept, kcal2 Slope, kcal/g2 R2 ME kcal/kg SP diets -5.79 (-46.87 – 35.29) 2,657 (2,515 – 2,799) 0.98 SEM 20.11 69.58

P-value 0.776 <0.001

CSBM diets -22.17 (-76.56 – 32.22) 2,814 (2,637 – 2,991) 0.97 SEM 26.63 86.81

P-value 0.412 <0.001

MEn kcal/kg

SP diets -17.98 (-53.53 – 17.56) 2,542 (2,419 – 2,665) 0.98 SEM 17.41 60.21

P-value 0.310 <0.001

CSBM diets -31.99 (-83.82 – 19.84) 2,575 (2,406 – 2,744) 0.97 SEM 25.38 82.73

P-value 0.217 <0.001

1Equations were generated by regressing the EE-SBM associated caloric intake (kcal) against the amount of EE-SBM intake (kg) to yield the ME or MEn value (kcal/kg) of the EE-SBM. Data are based on 8 replicate cages with 7 birds per cage. 2Corresponding 95% confidence intervals of regression equations intercepts and slopes are given in parentheses.

68 Pa

ge68

REFERENCES

Adeola, O., and K. E. Ileleji. 2009. Comparison of two diet types in the determination of metabolizable energy content of corn distillers dried grains with solubles for broiler chickens by the regression method. Poult. Sci. 88:579–585.

Anderson, D. L., W. Hill, and R. Renner. 1958. Studies of metabolizable and productive energy

of glucose for the growing chick. J. Nutr. 65:561–574. Baker, K. M., and H. H. Stein. 2009. Amino acid digestibility and concentration of digestible and

metabolizable energy in soybean meal produced from conventional, high-protein, or low-oligosaccharide varieties of soybeans and fed to growing pigs. J. Anim. Sci. 87:2282–2290.

Cervantes-Pahm, S. K., and H. H. Stein. 2008. Effect of dietary soybean oil and soybean protein

concentration on the concentration of digestible amino acids in soybean products fed to growing pigs. J. Anim. Sci. 86:1841–1849.

Coon, C. N., K. L. Leske, O. Akavanichan, and T. K. Cheng. 1990. Effect of oligosaccharide-

free soybean meal on true metabolizable energy and fiber digestion in adult roosters. Poult. Sci. 69:787–793.

Dale, N., and H. L. Fuller. 1984. Correlation of protein content of feedstuffs with the magnitude

of nitrogen correction in true metabolizable energy determinations. Poult. Sci. 63:1008–1012.

Fan, M. Z., and W. C. Sauer. 1995. Determination of apparent ileal amino acid digestibility in

barley and canola meal for pigs with the direct, difference, and regression methods. J. Anim. Sci. 73:2364–2374.

García-Rebollar, P., L. Cámara, R. P. Lázaro, C. Dapoza, R. Pérez-Maldonado, and G. G.

Mateos. 2016. Influence of the origin of the beans on the chemical composition and nutritive value of commercial soybean meals. Anim. Feed Sci. Technol. 221:245–261.

Jackson, S., J. D. Summers, and S. Leeson. 1981. Effect of dietary protein and energy on broiler

carcass composition and efficiency of nutrient utilization. Poult. Sci. 1:2224–2231. Janssen, W. M. M. A. 1989. European table of energy values for poultry feedstuffs. 3rd ed.

Beekbergen, Netherlands. Karr-Lilienthal, L. K., L. L. Bauer, P. L. Utterback, K. E. Zinn, R. L. Frazier, C. M. Parsons, and

G. C. Fahey. 2006. Chemical composition and nutritional quality of soybean meals prepared by extruder/expeller processing for use in poultry diets. J. Agric. Food Chem. 54:8108–8114.

Kluth, H., and M. Rodehutscord. 2006. Comparison of amino acid digestibility in broiler

chickens, turkeys, and Pekin ducks. Poult. Sci. 85:1953–1960.

69 Pa

ge69

Leeson, S., K. N. Boorman, and D. Lewis. 1977. Metabolisable energy studies with Turkeys: Nitrogen correction factor in metabolisable energy determinations. Br. Poult. Sci. 18:373–379.

Lopez, G., and S. Leeson. 2008. Assessment of the nitrogen correction factor in evaluating

metabolizable energy of corn and soybean meal in diets for broilers. Poult. Sci. 87:298–306. Mateos, G. G., E. Jiménez-Moreno, M. P. Serrano, and R. P. Lázaro. 2012. Poultry response to

high levels of dietary fiber sources varying in physical and chemical characteristics. J. Appl. Poult. Res. 21:156–174.

Mateos, G. G., and J. L. Sell. 1980. True and apparent metabolizable energy value of fat for

laying hens: Influence of level of use. Poult. Sci. 59:369–373. Mateos, G. G., and J. L. Sell. 1981. Influence of fat and carbohydrate source on rate of food

passage of semipurified diets for laying hens. Poult. Sci. 60:2114–2119. Opapeju, F. O., A. Golian, C. M. Nyachoti, and L. D. Campbell. 2006. Amino acid digestibility

in dry extruded-expelled soybean meal fed to pigs and poultry. J. Anim. Sci. 84:1130–1137. Parsons, C. M. 1986. Determination of digestible and available amino acids in meat meal using

conventional and caecectomized cockerels or chick growth assays. Br. J. Nutr. 56:227–240. Parsons, C. M., Y. Zhang, and M. Araba. 2000. Nutritional evaluation of soybean meals varying

in oligosaccharide content. Poult. Sci. 79:1127–1131. Perryman, K. R., and W. A. Dozier. 2012. Apparent metabolizable energy and apparent ileal

amino acid digestibility of low and ultra-low oligosaccharide soybean meals fed to broiler chickens. Poult. Sci. 91:2556–2563.

Pesti, G. M., L. O. Faust, H. L. Fuller, N. M. Dale, and F. H. Benoff. 1986. Nutritive value of

poultry by-product meal. 1. Metabolizable energy values as influenced by method of determination and level of substitution. Poult. Sci. 65:2258–2267.

Potter, L. M., L. D. Matterson, A. W. Arnold, W. J. Pudelkiewicz, and E. P. Singsen. 1960. The

evaluation of the metabolizable energy and productive energy of alpha cellulose. Poult. Sci. 39:1166–1178.

Powell, S., V. D. Naranjo, D. Lauzon, T. D. Bidner, L. L. Southern, and C. M. Parsons. 2011.

Evaluation of an expeller-extruded soybean meal for broilers. J. Appl. Poult. Res. 20:353–360.

Renner, R., and F. W. Hill. 1960. The utilization of corn oil, lard and tallow by chickens of

various ages. Poult. Sci. 39:849–854.

70 Pa

ge70

Rochell, S. J., T. J. Applegate, E. J. Kim, and W. A. Dozier. 2012. Effects of diet type and ingredient composition on rate of passage and apparent ileal amino acid digestibility in broiler chicks. Poult. Sci. 91:1647–53.

Rodehutscord, M., M. Kapocius, R. Timmler, and A. Dieckmann. 2007. Linear regression

approach to study amino acid digestibility in broiler chickens. Br. Poult. Sci. 45:85–92. Short, F. J., J. Wiseman, and K. N. Boorman. 1999. Application of a method to determine ileal

digestibility in broilers of amino acids in wheat. Anim. Feed Sci. Technol. 79:195–209. Sibbald, I. R. 1976. The true metabolizable energy values of several feedingstuffs measured with

roosters, laying hens, turkeys and broiler hens. Poult. Sci. 55:1459–1463. Sibbald, I. R., and S. J. Slinger. 1962. Factors affecting the metabolizable energy content of

poultry feeds: 10. A study of the effect of level of dietary inclusion on metabolizable energy values of several high protein feedingstuffs. Poult. Sci. 41:1282–1289.

Sibbald, I. R., and S. J. Slinger. 1963. A biological assay for metabolizable energy in poultry

feed ingredients together with findings which demonstrate some of the problems associated with the evaluation of fats. Poult. Sci. 42:313–325.

Sibbald, I. R., S. J. Slinger, and G. C. Ashton. 1962. Factors affecting the metabolizable energy

content of poultry feeds: 5. The level of protein and of test material in the diet 6. A note on the relationship between digestible and metabolizable energy values. Poult. Sci. 41:107–116.

Sibbald, I. R., J. D. Summers, and S. J. Slinger. 1960. Factors affecting the metabolizable energy

content of poultry feeds. Poult. Sci. 39:544–556. Sibbald, I. R., and M. S. Wolynetz. 1985. Relationships between estimates of bioavailable

energy made with adult cockerels and chicks: Effects of feed intake and nitrogen retention. Poult. Sci. 64:127–138.

Vermeersch, G., and F. Vanschoubroek. 1968. The quantification of the effect of increasing

levels of various fats on body weight gain, efficiency of food conversion and food intake of growing chicks. Br. Poult. Sci. 9:13–30.

Woodworth, J. C., M. D. Tokach, R. D. Goodband, J. L. Nelssen, P. R. O’Quinn, D. A. Knabe,

and N. W. Said. 2001. Apparent ileal digestibility of amino acids and the digestible and metabolizable energy content of dry extruded-expelled soybean meal and its effects on growth performance of pigs. J. Anim. Sci. 79:1280–1287.

Zhang, F., and O. Adeola. 2017. Energy values of canola meal, cottonseed meal, bakery meal,

and peanut flour meal for broiler chickens determined using the regression method. Poult. Sci. 96:397–404.

71 Pa

ge71

Zhang, Y., and C. M. Parsons. 1993. Effects of extrusion and expelling no the nutritional quality of conventional and kunitz trypsin inhibitor-free soybeans. Poult. Sci. 72:2299–2308.

72 Pa

ge72

CONCLUSION

In summary, it is evident that feed ingredient ME values are greatly influenced by the

inclusion level of the test ingredient in the diet. The first experiment investigated the influence of

diet type on the MEn of dextrose when determined by regression and to compare both index and

TC procedures. It was concluded that the dietary inclusion level of dextrose influenced the

effects of diet type and excreta collection method on the determined MEn values for dextrose.

The second experiment evaluated the influence of SP and CSBM diets on the ME value of a

commercially available EE-SBM determined by the regression method. It was concluded that the

MEn values of EE-SBM determined for SP and CSBM diets were dependent on the inclusion of

the EE-SBM in the test diets. Therefore, both experiments demonstrated the impact of inclusion

level on single feed ingredient ME values. Furthermore, the regression method was found to

reduce error associated with using multiple inclusion levels when ME values are based on the

difference method. Further research should be conducted to validate these findings by testing the

additivity in ME values of feed ingredients with different nutritional profiles when fed as part of

complete diets and to conclude whether the index or TC method is more suitable when using the

regression method to determine ME values of feed ingredients.

73 Pa

ge73

APPENDIX