This article was downloaded by: [New York University]On: 11 May 2013, At: 02:59Publisher: Taylor & FrancisInforma Ltd Registered in England and Wales Registered Number: 1072954Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

British Poultry SciencePublication details, including instructions for authors andsubscription information:http://www.tandfonline.com/loi/cbps20

Apparent metabolisable energyvalues of fats for broiler chicksJ. Wiseman a , D. J. A. Cole a , F. G. Perry b , B. G. Vernon b c

& B. C. Cooke ca University of Nottingham, School of Agriculture, SuttonBonington, Loughborough, LE12 5RDb B. P. Nutrition (UK) Ltd, Wincham, Northwich, CW9 6DFc Dalgety Agriculture Ltd, Dalgety House, The Promenade,Clifton, Bristol, BS8 3NJ, EnglandPublished online: 08 Nov 2007.

To cite this article: J. Wiseman , D. J. A. Cole , F. G. Perry , B. G. Vernon & B. C. Cooke(1986): Apparent metabolisable energy values of fats for broiler chicks, British Poultry Science,27:4, 561-576

To link to this article: http://dx.doi.org/10.1080/00071668608416915

PLEASE SCROLL DOWN FOR ARTICLE

Full terms and conditions of use: http://www.tandfonline.com/page/terms-and-conditions

This article may be used for research, teaching, and private study purposes. Anysubstantial or systematic reproduction, redistribution, reselling, loan, sub-licensing,systematic supply, or distribution in any form to anyone is expressly forbidden.

The publisher does not give any warranty express or implied or make anyrepresentation that the contents will be complete or accurate or up to date. Theaccuracy of any instructions, formulae, and drug doses should be independentlyverified with primary sources. The publisher shall not be liable for any loss, actions,claims, proceedings, demand, or costs or damages whatsoever or howsoever causedarising directly or indirectly in connection with or arising out of the use of thismaterial.

British Poultry Science (1986) 27, 561-576

APPARENT METABOLISABLE ENERGY VALUES OF FATS FORBROILER CHICKS

J. WISEMAN, D. J. A. COLE, F. G. PERRY,1 B. G. VERNON1,2

AND B. C. COOKE3

University of Nottingham School of Agriculture, Sutton Bonington, LoughboroughLE12 5RD, 1B. P. Nutrition (UK) Ltd, Wincham, Northwich CW9 6DF, and

3Dalgety Agriculture Ltd, Dalgety House, The Promenade, Clifton, Bristol BS8 3NJ,England

Received for publication 17th June 1985

Abstract 1. Seven fats were included at 30, 60 and 90 g/kg (experi-ment 1) and at 20, 40, 60 g/kg (experiment 2) in a semi-synthetic fat-free diet and in a practical diet respectively.2. Apparent metabolisable energy (AME) was evaluated with 6 repli-cate cages each with three Ross 1 cockerel chicks 8 d old. Diets werefed for 11 d and a total collection of excreta undertaken for the lastfour. There was no significant departure from linearity in the re-sponse of dietary AME to added fat, indicating no interactions be-tween basal diet and added fat.3. In experiment 3 one fat was evaluated at 10 rates of inclusion (10g/kg to 100 g/kg in 10 g increments) in both a semi-synthetic fat-freebasal and a practical basal diet. A significant departure from linearityin the response of dietary AME to added fat was detected but therewas no significant fat×basal diet interaction.4. In experiment 4 twelve commercially available fat blends were eachevaluated at 10 rates of inclusion (15 g/kg to 150 g/kg in 15 gincrements) in a practical basal diet. Significant departures fromlinearity in the responses of dietary AME to added fat were observedwith some of the fats.5. It was concluded that the AME of fats mat be determined frommulti-level assays by interpolating the quadratic relationships derived.

INTRODUCTION

The increasing use of fats in poultry diets of high nutrient density hasmade their evaluation of particular relevance to the efficiency of least cost dietformulation.

2 Present address: Dalgety Agriculture Ltd, Dalgety House, The Promenade, Clifton, BristolBS8 3NJ.

561

Dow

nloa

ded

by [

New

Yor

k U

nive

rsity

] at

02:

59 1

1 M

ay 2

013

562 J. WISEMAN ETAL.

The mechanisms of fat digestion and absorption in poultry (Freeman,1984) indicate the importance of the formation of bile salt micelles before theabsorption of the products of lipolysis (predominantly 2-monoglycerides andfree fatty acids). The differential solubility of these products in bile saltsolution explains the relatively higher absorbability of, for example, mediumchain length fatty acids and long chain length polyunsaturated fatty acids(polar solutes) compared to long chain saturated fatty acids (non-polar so-lutes). Such differences are reflected in the fats' apparent metabolisableenergy (AME) values, which have been shown to be influenced by both chainlength and degree of saturation of the constituent fatty acids (Renner andHill, 1961). However, there are factors, other than structure per se, which areof importance in the evaluation of fats. Thus, Freeman (1984) discussed theability of polar solutes to increase the micellar solubility of non-polar solutes.This mechanism is responsible for the phenomenon commonly referred to assynergism, whereby the presence of dietary polyunsaturated fatty acids mayassist in the absorption of an otherwise poorly utilised long chain saturatedfatty acid. Consequently, the AME determined for a fat such as tallow con-taining a preponderance of saturated fatty acids may be elevated if it isevaluated in conjunction with a basal diet containing ingredients of plantorigin with high proportions of polyunsaturated fatty acids (Sibbald et al.,1961; Artman, 1964; Lewis and Payne, 1966).

The consequences of such interactions on the AME of diets containingadded fat have been considered by Leeson and Summers (1976), who pro-posed a model based on the observation that the ability of polyunsaturatedfatty acids to solubilise saturated fatty acids into the micellar phase waslimited. Thus the effect of synergism on fat absorption would be progressivelysmaller at higher fat concentrations. Furthermore, from studies on themechanism of fat digestion and absorption, it has been observed that there is,particularly in the young chick, a deficiency of bile production (Freeman,1984). Thus it would appear that young birds have a finite capacity to utilisefat. This could also be important for adult birds, who have been reported toshow impaired ability to utilise fats at high concentrations (Shannon, 1971;Kussaibati, 1978).

The net effect of both synergism and the bird's finite capacity forfat utilisation may be that dietary AME values respond curvilinearly toadded fat, with the degree of curvilinearity being dependent upon theconditions of evaluation. However, previous studies have failed to detecta significant departure from linearity in this response (Sibbald and Slinger,1963; Fuller and Rendon, 1979; Mateos and Sell, 1980a; Mateos and Sell,1981a).

Whether or not the AME of a fat is influenced by its dietary concentra-tion is of considerable importance to the animal feedingstuffs industry. Theobjective of the current study was to evaluate the effect on AME of thedietary concentration of different fats obtained from various commercialsources. The study was also designed to assess the effectiveness of differentmethods in yielding AME data of fats that would improve the precision ofdiet formulation.

Dow

nloa

ded

by [

New

Yor

k U

nive

rsity

] at

02:

59 1

1 M

ay 2

013

METABOLISABLE ENERGY OF FATS 563

METHODS AND MATERIALS

Experiment 1

Seven commercially available fats were obtained for investigation. Fattyacid profiles are presented in Table 1. A semi-synthetic basal diet was formu-lated to be free from fat (Table 2) and each of the 7 fats was substituted at 30g/kg, 60 g/kg and 90 g/kg. A proprietary mineral and vitamin premix calcu-lated to supply the requirements of the growing broiler chicken, together withDL-methionine, were added to each of the 22 experimental diets at concentra-tions of 50 g/kg and 2 g/kg respectively.

The AME values of the experimental diets were determined using 6replicate cages containing 3 Ross I male chicks of 8 d initial age. The birdswere housed at random in an environmentally controlled room. Diets were fedfor l i d and the excreta collected quantitatively for the last 4 d. Excretasamples were dried in a forced draught oven at 90 °C for 36 h, ground andstored at 5 °C. Gross energy determinations were performed in duplicate on allfat, diet and excreta samples using a Gallenkamp ballistic bomb calorimeter.

Experiment 2

The same fats as those described in experiment 1 were used, with theexception that blend 3, which was by then unavailable, was replaced by blend 4.A practical basal diet was formulated (Table 3) and fats substituted at 20 g/kg,40 g/kg and 60 g/kg. Minerals were provided by dicalcium phosphate (9-3g/kg), limestone flour (5-2 g/kg) and a commercial trace mineral/vitaminsupplement designed to meet the requirements of the growing broiler chicken(130 g/kg). DL-methionine was added at 0-4 g/kg. The AME values of the 22experimental diets were determined in the same way as in experiment 1.

Experiment 3

Blend 4, a commercial fat blend (Table 1), was substituted into a semi-synthetic fat-free basal diet and a practical basal diet (Tables 2, 3) at concentra-tions of 10 g/kg to 100 g/kg in 10 g/kg increments. The 22 experimental dietswere evaluated for AME in the manner described for experiment 1.

Experiment 4

Samples of soya bean oil, palm acid oil and 10 commercially available fatswere obtained. Analyses are presented in Table 4. The fats were substitutedinto a basal diet (Table 5) in 15 g/kg increments from 15 g/kg to 150 g/kg.All diets were supplemented with macro-elements and a proprietary tracemineral/vitamin premix designed to meet the requirements of the growingbroiler chicken (Table 5). Evaluation of diets was achieved in a series ofexperiments similar to those described for experiment 1. Thus soya bean oiland blend 5 were evaluated in trial 4a; tallow, blends 6 and 7 in trial 4b; blends8, 9 and palm acid oil in trial 4c; blends 10 and 11 in trial Ad and blends 12 and13 in trial 4e. All diets were mixed at the start of each trial rather than at the

Dow

nloa

ded

by [

New

Yor

k U

nive

rsity

] at

02:

59 1

1 M

ay 2

013

TABLE 1

Analyses of fat samples used in experiment 1, 2 and 3

FatExperimentIodine valueUnsaponifiablecontent (g/kg)Free fatty acid (g/kg)Moisture (g/kg)Gross energy(MJAg)Fatty acids (g/kg)

C 14:0C 16:0C 16:1C 18:0C 18:1C 18:2C 18:3C20:lC22:0C 24:1

Lard1,254-27

<100<100<100

3901

13-9266130-5

157-7424-3

90-7<100<100

Tallow1, 249-24

1901910

<100

39-80

27-9255-339-8

1930399948-9

<100<100

Soya oil1,275-99

<100130

<100

3808

106-3<100

38-6251-25211

69-7

Fish oil1, 2

124-67

3 2 04 6 0

<100

39-49

70-8118096-5

<100110-52 0 1

<100205-2255-3

631

Blend 11,266-78

3 2 03920

<100

38-85

26-3228-333-4

126-5377-3135-217-421-6

<100

Blend 21,264-66

5 4 0291-0

<100

38-98

22-3175-8

12-9286-2298-4102-8

10-312 326-9

Blend 31

67-23

6 8 03100

<100

38-97

35-7234-7

19-5178-3326-4

94-4<100

18-7161

Blend 42, 371-42

6 1 03600

<100

38-44

26-52020

2 1 1113-5374-51631

17-220-620-2

en

|

Dow

nloa

ded

by [

New

Yor

k U

nive

rsity

] at

02:

59 1

1 M

ay 2

013

METABOLISABLE ENERGY OF FATS 5 6 5

TABLE 2

Composition of semi synthetic basal(experiment 1)

Fat-free soya protein isolate1

Cellulose2

Cane molassesWheat starchSucrose

Promin DSolka-floc

diet (g/kg)

285790 763-6

16004000

All diets were supplemented with a premix (designedto meet the macro-and micro-mineral and vitaminrequirements of the broiler chick) and DL-methionineat inclusion rates of 50 and 2 g/kg respectively.

TABLE 3

Composition of practical basal diet (g/kg) (experiments 2 and 3)

Maize meal . 129 0Wheat meal 545 0Soya bean meal (440 g/kg crude

protein) 2410Herring meal 850

All diets were supplemented with a premix (see Table2—130 g/kg) dicalcium phosphate (9-3 g/kg), limestoneflour (5-2 g/kg) and DL-methionine (0-4 g/kg).

beginning of the experimental programme to avoid deterioration in quality. Intrial 4a, only one basal diet with no added fat was tested whereas in trials 4b to4e each fat evaluated was associated with a basal diet with no added fat.

RESULTS

The AME values obtained for the diets used in experiments 1 and 2 arepresented in Tables 6 and 7. The AME values for the various fat blends weredetermined by difference. In both experiments and with every fat there was asignificant (P<0-001) linear increase in the AME values of the diets as the fatconcentration increased. The AME values of diets found in experiment 3 arepresented in Table 8, which also includes AME values of the fats determined bydifference. There was a significant (P<0-001) departure from linearity in theresponse of dietary AME to added fat but no fat Xbasal interaction was found.Table 9 contains data obtained from experiments 4a to 4e inclusive on theAME of diets and of fats determined by difference. Analysis of variancerevealed a significant (P<0-001) difference between fats (trials 4a, 4b, 4c, 4e) asignificant effect of fat concentration (trial 4aP<0-001 linear, trial 46P<001quadratic, trial 4c P<0 01 quadratic, trial 4d P<0-001 linear and trial 4eP < 0-001 linear) and a significant fat X concentration interaction (trial 4a

Dow

nloa

ded

by [

New

Yor

k U

nive

rsity

] at

02:

59 1

1 M

ay 2

013

CJi

TABLE 4

Analysis of fats used in experiment 4

FatExperimentUnsaponifiable (g/kg) •Free fatty acid (g/kg)Moisture (g/kg)Gross energy (MJ/kg)Major fatty acids (g/kg)

C12.0C14:0C16:0C16:lC18:0C18:l

,C18:2C18:3C20:0C20:lC22:0C22:l

SoyaBean

oil4a

<100<100<100

38-720

1310

21023605440

6 8 0

Blend54a

4 3 05060

<10038-567

2 0 03 8 0

3060190730

34901610

120290

Tallow46

9 06 0 010039-514

3 8 03390

16013104130

6 3 0

Blend64b

4 3 0432010039069

100140

3030

63042201760

150

Blend74b

3 8 04800

13039020

2 5 02 8 0

2000100

142040301540

180120130

Blend84c

29-05000

10038-873

2 1 03120

1403 2 0

37902130

2 1 0

100

Blend94c

3 8 04000

<10038-621

2 2 02 3 0

1290100

119033103040

100140

170100

PalmAcidoil4c

1707500

<10038-918

1204910

2 1 03820

9 4 0

Blend104rf

2 0 01840

<10039094

1002 1 0

2700

119041001550

1 7 0

Blend11U

1702580

<10039-250

2 0 02600

110042001710

Blend124e

1401480

<10038-927

1302 8 0

2700

6 2 045101610

130100

2-0

Blend134e

2002840

10038-570

4 02 4 0

3240

65041101550

8 04 0

100

1—1

inM

Dow

nloa

ded

by [

New

Yor

k U

nive

rsity

] at

02:

59 1

1 M

ay 2

013

METABOLISABLE ENERGY OF FATS 567

TABLE 5

Composition of basal diet (experiment 4)

Pruteen granules.As in Table 2.

Wheat mealMaize mealJet sploded wheatWhite fish mealSoya bean meal (500 g/kg crude protein)Single cell protein1

All experimental diets were supplementedwith the following:

Dicalcium phosphateSaltLimestone flourDL-methionineLysinePremix2

(g/kg)160-853-6

479-426-8

225-853-6

9 0 01005 0 00-90100

1304

P<0-001 linear, trial 4c P<0001 linear, trial Ad P < 0 0 5 linear and trial 4eP<0001 quadratic).

DISCUSSION

By selecting fats of widely differing chemical structure and basal diets ofdiffering oil content, the objective of experiments 1 and 2 were to attempt toshow that the response of dietary AME to added fat was non-linear and that thedegree of curvilinearity observed was specific to the particular fat.

Calculation of the AME values of the fats (Tables 6 and 7) indicated thatalthough there were differences there were no consistent trends (confirmingthe observations of Muztarret al., 1981). The standard errors associated withthe derived values for the fats were large because of the low concentrationschosen for their evaluation. In addition, regression analysis of the data fromexperiments 1 and 2 revealed no significant departure from linearity in theresponse of dietary AME to added fat. It may be concluded, therefore, that inthese two experiments the AME values of the fats were not influenced by theirinclusion rate. Solving the linear equations (Mateos and Sell, 19806) is analternative means of deriving AME value of fats and standard errors werereduced (Table 10).

Data derived from experiments 1 and 2 suggested that single-level assaysmay not be appropriate for the evaluation of fats. They also failed to confirmthe model of Leeson and Summers (1976) and, because of the high standarderrors, did not product AME values for fats which could be reliably used in dietformulation. However the observations of Freeman (1984) and Leeson andSummers (1976) suggest that the response of dietary AME to added fat shouldnot be linear and this raises the question of the nature of the curve used todescribe the response. In considering the net effect of fatty acid synergism andthe bird's finite capacity for fat utilisation, it seems that the response of dietaryAME to added fat should be adequately described by a quadratic equation

Dow

nloa

ded

by [

New

Yor

k U

nive

rsity

] at

02:

59 1

1 M

ay 2

013

568 J. WISEMAN ETAL.

TABLE 6

AME (MJ/kg) of experimental diets (based on semi-synthetic basal) and of fats determined by difference(experiment 1)

Diet

1

2

3

4

5

6

7

1 f^l^.

Fat(Basal)

Lard

Tallow

Soya oil

Fish oil

Blend 1

Blend 2

Blend 3

Inclusionrate of

fat(g/kg)

0

306090

306090

306090

306090

306090

306090

306090

AME of diet(as fed)(MJAg)13-89

14-4915-4016-33

14-6514-8215-14

14-53150315-62

14-8615-2816-24

14-6914-6115-49

14-3914-8114-84

14-5914-5314-74

1000

SE016

0-21P<0-001 (Linear)

NS (Quadratic)

019P<0 001 (Linear)

NS (Quadratic)

019P<0-001 (Linear)

NS (Quadratic)

011P<0 001 (Linear)

NS (Quadratic)

0-31P<0001 (Linear)

NS (Quadratic)

013P<0-001 (Linear)

NS (Quadratic)

0-14/><0001 (Linear)

NS (Quadratic)

AME of fat(MJ/kg)

33-9391410

39-229-427-8

35-232-9331

46-2371400

40-625-931-7

30-629-224-4

37-224-623-3

SE1

8-74-32-8

8-24-02-7

8-2402-7

6-33120

11-65-73-8

6-73-32-2

7-03-42-2

g/kg added fatwhere VAR,= variance of diet with added fat

VARB=variance of basal dietF = proportion of basal in diet with added fat.

representing a progressive decline in the degree of utilisation of a fat with itsincreasing concentration. No attempt is made to fix an optimum concentration,but merely to characterise the response over the range of inclusion rates likelyto be encountered in practice. As this range is covered by the experimentaltreatments employed, values at particular concentrations are simply derived byinterpolation of the quadratic equation.

A quadratic response of dietary AME to added fat may be evaluated in anexperiment with only three rates of inclusion. However, the use of more thanthree rates is considered more powerful by allowing the curvilinear slope to becharacterised with a greater degree of accuracy (of particular importance in theassessment of the relative values of fats) and by also allowing detection of anysystematic departure from the quadratic model. Accordingly, in experiment 3 acommercial fat blend was evaluated at 10 inclusion rates in either a semi-synthetic fat-free basal diet or practical basal diet. The objective was toestablish whether the linear responses obtained in experiments 1 and 2 were aconsequence of the insensitivity of the design to detect non-linear responses.

Dow

nloa

ded

by [

New

Yor

k U

nive

rsity

] at

02:

59 1

1 M

ay 2

013

METABOLISABLE ENERGY OF FATS 569

TABLE 7

AME (MJ/kg) of experimental diets (based on practical basal) and fats determined by difference (experiment 2)

DietFat(Basal)

Lard

Tallow

Soya oil

Fish oil

Blend 1

Blend 2

Blend 4

Inclusionrate of

fat(gAg)

0

204060

204060

204060

204060

204060

204060

204060

AME of diet(as fed)(MJAg)13-59

141514-7815-35

141214-5514-82

141314-4915-34

140714-5715-20

140914-6314-80

13-9814-3514-51

141514-3914-63

SE016

0-22P<0001 (Linear)

NS (Quadratic)

0-21P<0001 (Linear)

NS (Quadratic)

019P<0001 (Linear)

NS (Quadratic)

0-20P<0001 (Linear)

NS (Quadratic)

019F<0001 (Linear)

NS (Quadratic)

0-20P < 0 0 1 (Linear)NS (Quadratic)

0-20P<00l (Linear)NS (Quadratic)

AME of fat(MJAg)

41-643-342-9

40137-6341

40-636142-8

37-638-140-4

38-639-633-8

33132-628-9

41-633-630-9

SE1

13-56-74-4

1316-54-3

12-3614-0

12 76-34-2

12-36-64 0

12-76-34-2

12-76 34-2

'• Calculated as in Table 6.

TABLE 8

AME (MJ/kg) of experimental diets and of fat, determined by difference (experiment 3)

Dietary inclusionrate of fat(gAg)Basal 0

102030405060708090

100

SemiAME diet SE

13-79 014140614-3914-5414-8014-9114-8815-2415-32150614-97

SEFatsInclusion ratesFats X rates

-syntheticAME fat

40-843-838-83 9 036-23 2 034-532-927-925-6

0 0 60 1 40-24

basalSE1

19-79-86-54-93-93-22-72-42 11-9

P<0-P<0001

Commercial basalAME diet

12-6112-89131813-5313-6413-8513-7813-99141013-901419

001(Linear)

(NS)

SE AME fat0 1 4

40-64 1 143-338-437-43 2 132-331-226-928-4

SE1

19-79-86-54-93-93-22-72-42 11-9

Calculated as in Table 6.

Dow

nloa

ded

by [

New

Yor

k U

nive

rsity

] at

02:

59 1

1 M

ay 2

013

570 J. WISEMAN ETAL.

TABLE 9

AME (MJ/kg) of experimental diets and fats determined by difference

TrialFat

Rateofinclusion(gAg)

0153045607590

105120135150

TrialFat

Levelofinclusion(gAg)

0153045607590

105120135150

TrialFat

Rateofinclusion(gAg)

0153045607590

105120135150

Soya oil

• AME (MJAg)Diet Fat

11-80112120123191253212-940135531377114 186145831509715052

SEFatsRates

Fats X Rates

33129128030-835-233-73 4 53 5 036-233-5

Blend 8

AME (MJAg)Diet Fat

11-57912-24812-60312-54512-98213-27613-60413-7541403114-55014-244

56-245-733-035-034-23 4 132-33 2 03 3 629-3

SEFatsRates

Fats X Rates

Blend 12

AME (MJAg)Diet Fat

11-60212-32512-28212-39912-94713-47113-92313-86814-26414 74115-446

59-834-329-33 4 036-537-43 3 233-834-937-2

Fats

4aBlend 5

AME (MJAg)Diet Fat

12-288 44-312133 22-912-615 29-913041 32-512173 30113673 32-613-605 29014056 30-613-693 25-814-339 28-7

008 P<0001017 P<0001

(Linear)0-24 P<0001

(Linear)

4cBlend 9

AME (MJAg)Diet Fat

11-44312006 49-012-260 38712-637 38012-837 34-712-759 29012-863 27-213-591 31-913-488 28-513-828 29113-935 281

0-07 P<0-001014 P<0001

(Linear)P<0-01

(Quadratic)0-24 P<0-001

(Linear)4«

Blend 13

AME (MJAg)Diet Fat

11-83412093 29112-483 33-512-893 35-413123 33313325 31 713-508 30-413-853 31113-873 28-814-107 28-714075 26-8

SE0 0 6

Rates 014FatXRates 0 19

Tallow

AME (Diet

11-46511-96512-52412-65913-07412-98313-66013-98313-98714-07214-487

FatsRates

MJAg)Fat

44-846-83 8 038-331 73 5 935-432-530-831-6

SE

Fats X Rates

Palm acid oil

AME (MJAg)Diet Fat

11-71911-7831218912-52012-45112-66712-32012-70913-01712-71512-802

FatsRates

16027-429-523-924-418-421-122-519118 9

SE

Fats X Rates

P<0001P<0001P<0001i><0001

(Linear)(Linear)(Quadratic)

4*Blend 6

AME (MJAg)Diet Fat

11-68511-74011-93811-97912-56112-89813-37413-25413-5231335713-920

0-080 1 5

0-27id

Blend

15-420-118-226-327-930-526-62702 4 126-6

P<000]P-C0-001(Linear)P<001

Blend 7

AME (MJAg)Diet Fat

1112011-8931215212-22512-81012-9521306413-43212-47913-59213-864

1I

(Quadratic)NS

10

AME (MJAg)Diet Fat

12-36012-85212-9911325113-7201411314-60914-8171485115-43215-585

0 0 60 1 4

0-20

45-233-432-23 5 035-737-335-83 3 135133-9

P<0001P<0-001(Linear)P<005(Linear)

Blend

62-745-535-739-335-532-73 3 130-829-429-4

11

AME (MJAg)

12-28412-75112-01912-65413-85214-55814-83014-7721513614-95715-840

43-43 5

20-538-442-640-63 6 036-132-13 6 0

Dow

nloa

ded

by [

New

Yor

k U

nive

rsity

] at

02:

59 1

1 M

ay 2

013

METABOLISABLE ENERGY OF FATS 571

TABLE 10

Regression equations derived using AME (MJ/kg) of the compound diets as the dependent variable, and level of fat(g/kg) as the independent variable

FatSoya oil

Fish oil

Lard

Tallow

Bl

B2

B3B4

Experiment12121212121212

Equation(gAg)

> = 13-91+0-0190xJI = 13-54 + 0 0 2 8 2 X^ = 13-95 + 00249xji = 13-56 + 00268x;y=13-79 + 00275xjy=13-58 + 00297x;y=13-96 + 00157x^=13-65 + 00210x;y=1403 + 00131xJ I = 1 3 - 6 5 + 0 0 2 0 6 X}i = 13-99 + 00110x> = 13-64 + 00157xJI = 1 4 0 6 + 0 0 0 8 3 Xv = 13-69 + 0 0168x

RSD0-450-470-280-500-520-550 3 20-480-460-510-310 500-340-48

r2

0-990-970-990-990-990-990-870-960-910-980-900-970-740-95

AME of fat determined by regression

(MJ/kg)32-941 738-940-441-343-329-734-72 7 134-325029-322-430-5

SE1

2-23-51-43-72-55-41-53-52-2

•3 -7

1-53 71-63-5

' The standard error of the estimate of y when x = 1000 at the 5% confidence limit was determined from therelationship:

SE= 1 VAR^X1 (1000-x

when n=number of observations.x = numerical points of x.

VAR^=variance of regression.

Results from experiment 3 (Table 8) confirmed the observations fromexperiments 1 and 2 in that individual fat concentrations are inappropriate inthe evaluation of fats. Very small differences in determined AME values ofdiets with added fat may result in large differences being ascribed to the AMEof fats. Fat AME values determined in this way and taken in isolation from anyobserved trend in the overall response of dietary AME to added fat may notaccurately reflect the influence of the added fat.

Examination of the data from experiment 3 revealed a statisticallysignificant departure from linearity in the response of dietary AME to addedfat with both the semi-synthetic (Fig. 1) and practical basal diets (Fig. 2). Thesenon-linear responses may be attributable to the influence of fatty acidsynergism and/or the bird's finite capacity for fat utilisation. Although the trialwas not designed to separate the two effects it may be that the latter was ofgreater importance as no fatXbasal diet interaction was observed. It is alsopossible that added fat may enhance the utilisation of a diet by slowing down itsrate of passage (Mateos and Sell 1980a; 19816). Thus the AME of the basalcomponent of the diets containing added fat would be higher than would bepredicted from the AME of the basal diet determined with no added fat. As thisincrease in the utilisation of the basal diet would be more evident at lowerinclusion rates of fat (that is, there would-be a limit to the amount ofimprovement attributable solely to this mechanism) and less evident at higher

Dow

nloa

ded

by [

New

Yor

k U

nive

rsity

] at

02:

59 1

1 M

ay 2

013

572 J. WISEMAN ET Ah.

rates, then the overall response of dietary AME to increasing rates of fataddition would, in theory be curvilinear. However, although rates of passagewere not measured during experiment 3, it is improbable that any changeswould account for the responses obtained, as Golian and Polin (1984) haverecently reported that transit time of digesta was not influenced by thepresence of oil in diets for young chicks. In contrast the results of Pesti (1984)do suggest that the AME of maize was improved in the presence of oil,indicating an interaction between the two.

15-5r

Linear, y=14078 + 00129xRSD = 0-234, r=0-89

Quadratic, y = 13-754 + 00344x-0.00022x2RSD = 0-108, r=0-98

13-5 •

0 50 100

Inclusion rate of added fat (g/kg)

FIG. 1.—Effect of inclusion rate of fat (g/kg) into a serai-synthetic basal diet on the AME (MJ/kg).

The data presented in Table 11 were produced by interpolation of thequadratic equations (Figs 1 and 2). It can be seen that the effect of acurvilinear response in dietary AME to added fat is a reduction in thecalculated AME of the fat at higher concentrations. This assumes that therelative contribution from the basal diet is constant. It was concluded that fatscould best be evaluated by multi-level inclusion into a practical basal diet.

Experiment 4 was designed to evaluate a range of commercially availablefats and oils, and to provide comparative data on their AME values whichwould be of use in diet formulation. A multi-level assay was adopted, based onthe observations of experiment 3, with the range of fat contents increased to150 g/kg added fat. It was considered that such a procedure would provideuseful information on the behaviour of fats at high inclusion rates even thoughthese currently rarely exceed 80 g/kg. Determination of the AME of the fats bydifference (Table 9) confirmed the observations of experiment 3 in that

Dow

nloa

ded

by [

New

Yor

k U

nive

rsity

] at

02:

59 1

1 M

ay 2

013

METABOLISABLE ENERGY OF FATS 573

TABLE 11

AME (MJ/kg) of diets derived from quadratic equations and of fat by difference (experiment 3)

Inclusionrateof fat(gAg)Basal 0

102030405060708090

100

Semi-synthetic basal dietAME Diet1

13-7541407714-35714-59414-78714-9371504515109151301510715042

AME Fat3

4 6 043-941-739-637-435-33 3 130-928-826-6

SE4

3-43-12-72-42 01-71-41-21 01 0

PracticalAME Diet2

12-62812-9281319313-42413-61913-77913-90513-995140511407214058

basal dietAME Fat3

42-640-93 9 137-435-633-932-230-428-726-9

SE<

3 22-92-52-21-91-61-31 11 00-9

1 determined from quadratic equation in Fig. la.2 determined from quadratic equation in Fig. 16.8 determined by difference.4 determined as:

VVARA+VARS+F2VARC+2(CVAR^+F(CVARBC+CVARAC)where: VAR̂ =variance of intercept of function.

VARB = variance of linear component of function (103.B).VARC =variance of quadratic component of function (106.C).F •= proportion of added fat in diet.CVAR = covariance, of AS {A, 10\B) etc.

reference to data obtained at only one inclusion rate was not a satisfactorymeans of evaluation.

There was a significant departure from linearity (Table 9) in the responseof dietary AME to added fat, with only some of the fats. Any departure fromlinearity has a profound effect on the derivation of fat AME values. Valuesdetermined by extrapolation of a linear function to 1000 g/kg added fat willdiffer from those determined by interpolation of a fitted quadratic function ata given rate of added fat. If it is accepted that the efficiency of digestion andabsorption of fats decreases with increasing intake (with the degree ofreduction being a characteristic of the fat) then it could be argued that it isappropriate to fit quadratic functions to all the responses obtained, evenalthough statistically significant departures from linearity are not alwaysdetected.

Quadratic equations derived for all fats evaluated in experiment 4 arepresented in Table 12. Diets were grouped for each series (4a, 4b, 4c, 4d, 4e)and a common intercept from the responses of all fats was derived for eachgroup of fats evaluated. Solving individual functions at 30, 60 and 90 g/kgadded fat generated fat AME values which are presented in table 12. It hasbeen argued that it is more appropriate to fit a quadratic model to data of thistype, but, in doing so, two seemingly anomalous results were obtained for soyabean oil and blend 12, which gave positive coefficients for the quadratic termin the equations derived. For soya bean oil, examination of the dietary AMEdata, particularly at the concentrations of 30, 45 and 60 g/kg, revealed that thevalues were lower than anticipated and this could have contributed to the

Dow

nloa

ded

by [

New

Yor

k U

nive

rsity

] at

02:

59 1

1 M

ay 2

013

574 J. WISEMAN ETAL.

14-5r

13-5

12-5

Linear, y = 12-890 + 00143xRSD = 0-195, r=0-93

Quadratic, y= 12-628 + 00317x-0.00018x2RSD = O-1O1, /-=0-98

50

Inclusion rate of added fat (g/kg)100

KIG. 2.—Effect of inclusion rate of fat (g/kg) into a practical basal diet on the AME (MJ/kg).

Experiment Kat4f l

4ft

4c

Ad

4e

TABLE 12

Quadratic regression equations derived from experiment 4

Soya oil

Blend 5

Tallow

Blend 6Blend 7

Blend 8

Blend 9Palm acid

oil

Blend 10

Blend 11

Blend 12

Blend 13

Quadratic equationk x x3

y=\ 1-753 + 0-0207 + 0-00002±0-096 ±0-0034 ±0-00002

>>=ll-753+0-0229-0-00004

>= 11-449 + 00303-0-00007± 0079 ± 0003l±000002

;y=l 1-449 + 0-0204-0-00003y = 11-449 + 0-0237-0-00005

y= 11 •679 + 00255-000005±0078±0-0031 ±000002

-0186-0-00002

31=11-679 + 00167-0-00006

y= 12161 +0 0294-000004±0-196±0-0070±0-00005

>=12161+00258-000001

y=\ 1-770 + 0-0174 + 0-00004±0083 ±00030 ±000002

31=11-770 + 00268-0 00007

0-96

0 9 9

AME (MJ/kg) fat at inclusion

0

0

0

r•99

•98

•98

RSD0179

0179

0177

303 3 0±2-7'33-3

39 6±2-530-934-3

35-8±2-4

rates (g/kg)60

33-6± 2 03 2 0

37-4±1-83 0 032-7

34-3±1-8

9034-2

± 1 430-7

35-2±1-3291311

32-9±1-3

29-6

26-5

28-9

24-7

28-2

22-8

0-363

0153

40-3±5-537-6

30-3±2-336-3

38-9±4-237-2

31-6±1-834 1

37-6±2-936-8

32-8±1-231-9

' See footnote 4, Table 11.

Dow

nloa

ded

by [

New

Yor

k U

nive

rsity

] at

02:

59 1

1 M

ay 2

013

METABOLISABLE ENERGY OF FATS 575

unexpected function. In addition, the somewhat varied nature of the dietaryAME data for blend 12 may have been of relevance in this context.

Nevertheless it is proposed that the procedure is useful in ranking fats interms of their AME. It is also of interest that the relative nutritive value ofdifferent fats is influenced by their concentration, that is, the degree ofcurvilinearity in the response of dietary AME to fat addition is influenced bythe nature of the fat used. Thus it is not possible to extrapolate differencesderived at one rate of inclusion to another. The results of the present study areconsistent with those of Kussaibati (1978) that the rate of decline in nutritivevalue with increasing intake is fat specific.

ACKNOWLEDGEMENTS

The authors are grateful to Jim Craigon for statistical advice, KatherineWragg and Carol Pass for technical assistance and Unichema International Ltdfor their support.

REFERENCES

ARTMAN, N.R. (1964) Interactions of fats and fatty acids as energy sources for the chick. PoultryScience, 43, pp. 994-1004.

FREEMAN, C.P. (1984) The digestion, absorption and transport of fats—non-ruminants, in:WISEMAN, J. (Ed.) Fats in Animal Nutrition, pp. 105-122 (London, Butterworths).

FULLER, H.L. & RENDON, M. (1979) Energetic efficiency of corn oil and poultry fat at differentlevels in broiler diets. Poultry Science, 58, pp. 1234-1238.

GOLIAN, A. & POLIN, D. (1984) Passage rate of feed in very young chicks. Poultry Science, 63,pp. 1013-1019.

KUSSAIBATI, R. (1978) Influence of dietary intake level on the metabolisable energy and thedigestibility of lipids in the growing chicken and the adult cockerel. Proceedings II EuropeanSymposium on Poultry Nutrition, pp. 14-22, Beekbergcn.

LEESON, S. & SUMMERS, J.D. (1976) Fat ME values: the effect of fatty acid saturation. Feedstuffs,48, (No. 46), pp. 26 & 28.

LEWIS, D. & PAYNE, C.G. (1966) Fats and amino acids in broiler rations. 6. Synergisticrelationships in fatty acid utilisation. British Poultry Science, 7, pp. 209-218.

MATEOS, G.G. & SELL, J.L. (1980a) Influence of graded levels of fat on utilisation of purecarbohydrate by the laying hen. Journal of Nutrition, 110, 1894-1903.

MATEOS, G.G. & SELL, J.L. (19806) True and apparent metabolisable energy value of fat forlaying hens: influence of level of use. Poultry Science, 59, pp. 369-373.

MATEOS, G.G. & SELL, J.L. (1981a) Metabolisable energy of supplemental fat as related to dietaryfat level and methods of estimation. Poultry Science, 60, pp. 1509-1515.

MATEOS, G.G. & SELL, J.L. (1981A). Nature of the extra-metabolic effect of supplemental fat usedin semi-purified diets for laying hens. Poultry Science, 60, pp. 1925-1930.

MUZTAR, A.J., LEESON, S. & SLINGER, S.J. (1981) Effect of blending and level of inclusion onthe metabolisable energy of tallow and tower rapeseed soapstocks. Poultry Science, 60,pp. 365-372.

PESTI, G.M. (1984) Influence of substitution method and of food intake on bioassays todetermine metabolisable energy with chickens. British Poultry Science, 25, pp. 495-504.

RENNER, R. & HILL, F.W. (1961) Utilization of fatty acids by the chicken. Journal of Nutrition, 74,pp. 259-264.

SHANNON, D.W.F. (1971) The effect of level of intake and free fatty acid content on the

Dow

nloa

ded

by [

New

Yor

k U

nive

rsity

] at

02:

59 1

1 M

ay 2

013

576 J. WISEMAN ET AL.

metabolisable energy value and net absorption of tallow by the laying hen. Journal ofAgricultural Science, (Cambridge) 76, pp. 217-221.

SIBBALD, I.R., SLINGER, S.J. & ASHTON, G.C. (1961) Factors affecting the metabolisable energycontent of poultry feeds. Variability in the ME values attributed to samples of tallow, andundegummed soybean oil. Poultry Science, 40, pp. 303-308.

SIBBALD, I.R., SLINGER, S.J. (1963) A biological assay for metabolisable energy in poultry feedingredients together with findings which demonstrate some of the problems associated withthe evaluation of fats. Poultry Science, 42, pp. 313-325.

Dow

nloa

ded

by [

New

Yor

k U

nive

rsity

] at

02:

59 1

1 M

ay 2

013