Determination of ileal digestible and apparent metabolizable energy contents of expeller-extracted and solvent-extracted canola meals for broiler chickens by the regression method

The present study was conducted to determine ileal digestible energy (IDE), metabolizable energy (ME), and nitrogen-corrected ME (MEn) contents of expeller- (EECM) and solvent-extracted canola meal (SECM) for broiler chickens using the regression method. Dietary treatments consisted of a corn–soybean meal reference diet and four assay diets prepared by supplementing the reference diet with each of canola meals (EECM or SECM) at 100 or 200 g/kg, respectively, to partly replace the energy yielding sources in the reference diet. Birds received a standard starter diet from day 0 to 14 and the assay diets from day 14 to 21. On day 14, a total of 240 birds were grouped into eight blocks by body weight and randomly allocated to five dietary treatments in each block with six birds per cage in a randomized complete block design. Excreta samples were collected from day 18 to 20 and ileal digesta were collected on day 21. The IDE, ME, and MEn (kcal/kg DM) of EECM or SECM were derived from the regression of EECM- or SECM-associated IDE, ME and MEn intake (Y, kcal) against the intake of EECM or SECM (X, kg DM), respectively. Regression equations of IDE, ME and MEn for the EECM-substituted diet were Y = −21.2 + 3035X (r2 = 0.946), Y = −1.0 + 2807X (r2 = 0.884) and Y = −2.0 + 2679X (r2 = 0.902), respectively. The respective equations for the SECM diet were Y = 20.7 + 2881X (r2 = 0.962), Y = 27.2 + 2077X (r2 = 0.875) and Y = 24.7 + 2013X (r2 = 0.901). The slope for IDE did not differ between the EECM and SECM whereas the slopes for ME and MEn were greater (P < 0.05) for the EECM than for the SECM. These results indicate that the EECM might be a superior energy source for broiler chickens compared with the SECM when both canola meals are used to reduce the cost of feeding.


Background
Canola meal is a by-product of canola seed oil extraction and is a good source of essential amino acids for broiler chickens leading to its frequent use in poultry diet formulation as a protein source (Newkirk 2009). There are two extraction methods for obtaining canola oil from canola seed. Solvent extraction is the most common method and uses solvent to improve oil-extraction efficiency, resulting in a meal with <5 % residual oil. On the other hand, the expeller extraction is suggested to be less efficient for oil extraction because the oil is extracted only mechanically, thus leaving more oil (8-15 %) in the meal compared with solvent extraction method (Spragg and Mailer 2007). Moreover, it does not leave any solvent residues which would remain in the solvent-extracted meal. In addition, processing temperature and moisture contents also differ depending on extraction method. The processing conditions for expeller-extracted canola meal (EECM) are less in moisture (<12 % vs. 15-18 %) and higher in temperature (up to 160 °C vs. 95-115 °C) than for solvent-extracted canola meal (SECM), respectively (Newkirk 2009).
Processing conditions for canola oil extraction as well as residual oil in the meal can affect the nutritive values of canola meal (Woyengo et al. 2010b;Khajali and Slominski 2012). It has been reported that EECM had greater amino acid (AA) digestibility and metabolizable energy (ME) than SECM fed to growing pigs (Woyengo et al. 2010a;Maison and Stein 2014). However, there is a dearth of studies which evaluated the nutritive values of EECM for broilers. Woyengo et al. (2010b) reported that doubly extracted EECM compared with SECM fed to broilers had more standardized ileal digestible AA and N-corrected apparent ME (AME n ) by using the difference method. Toghyani et al. (2014) reported that ileal digestible energy (IDE), apparent ME (AME), and AME n values vary depending on processing conditions and chemical composition of EECM fed to broilers. In view of the dearth of data, the objective of the present study was to determine IDE, ME and ME n of single-extracted EECM and SECM fed to broiler chickens using the regression method.

Methods
All protocols for the experiment were reviewed and approved by the Purdue University Animal Care and Use Committee.

Animals and experimental diets
Day-old male broiler chicks of the Ross 308 (Aviagen, Huntsville, AL, USA) strain were obtained from a local hatchery, tagged with identification numbers, and housed in electrically heated battery cages (model SB 4 T, Alternative Design Manufacturing, Siloam Springs, AR, USA) in an environmentally controlled room. Battery brooder temperatures from day 0 to 7 and day 7-14 were kept at 35 and 32 °C, respectively. All birds received a mash standard broiler starter diet from day 0 to 14 (Table 1). On day 14, the birds were weighed individually and 240 birds were grouped into eight blocks by body weight and randomly allocated to five dietary treatments in each block with six birds per cage in a randomized complete block design using the Experimental Animal Allotment Program of Kim and Lindemann (2007). Birds were provided ad libitum access to water and experimental diets from day 14 to 21 and the battery brooder temperatures were maintained at 27 °C.
The analyzed chemical composition of EECM and SECM used in the present study are presented in Table 2. Dietary treatments consisted of a corn-SBM reference diet and four assay diets. In the reference diet (Table 3), corn, SBM, corn starch, and soybean oil were used as energy yielding sources. The four assay diets were prepared by supplementing the reference diet with each of canola meals (EECM or SECM) at 100 or 200 g/kg, respectively, to partly replace the energy yielding sources in the reference diet. The ratio of the energy yielding sources remained constant in all treatments to enable determination of the energy value of expeller extracted and solvent extracted canola meals by the regression method (Kong and Adeola 2014).

Sample collection and chemical analyses
Excreta samples were collected twice daily from day 18 to 20. During collection, waxed paper was placed in trays under the cages, and excreta on the paper were collected. The excreta samples were pooled per cage over the 2 days and stored in a freezer at −20 °C. On day 21, all birds were euthanized by asphyxiation with carbon dioxide and ileal digesta were collected from the distal two-thirds of ileum by gently rinsing with distilled water. The collected ileal digesta from six birds within a cage were pooled and stored in the freezer at −20 °C. At the completion of the experiment, ileal digesta and excreta samples were thawed and placed in a forced-air oven at 55 °C for 96 h and ground using a mill grinder (Retsch ZM 100, Retsch GmbH & Co., Haan, Germany). Gross energy (GE) of diets, excreta and ileal digesta samples was determined in adiabatic bomb calorimeter (Parr 1261, Parr Instruments Co., Moline, IL, USA) using benzoic acid as a calibration standard. Dry matter analysis of samples was conducted by drying the samples in a drying oven at 105 °C for 24 h (method 934.01; AOAC 2005). Nitrogen contents of the diets, excreta, and ileal digesta were determined using the combustion method (Model FP2000, Leco Corp., St. Joseph, MI) with EDTA as a calibration standard. Chromium concentrations in the diets, ileal, and excreta samples were determined using the method of Fenton and Fenton (1979).

Calculations
The coefficients of ileal digestibility and metabolizability of DM, N and energy in the experimental diets were calculated using the index method with chromic oxide as  an indigestible index (Kong and Adeola 2014). The IDE and AME contents of experimental diets were then calculated as the product of respective coefficients and the gross energy (kcal/kg). The AME was corrected to zero N retention using the factor of 8.22 kcal/g of N (Hill and Anderson 1958). The IDE, ME or ME n of the test ingredients (SECM and EECM) were calculated using the regression method described in the study by Adeola and Ileleji (2009). Gross energy of the assay diets was corrected for non-energy yielding fractions and then the substitution rates of the canola meals were corrected for energy contributions of energy-yielding ingredients and canola meals in the assay diet. The energy contributioncorrected substitution rate was multiplied with dietary IDE, ME or ME n intake to determine EECM-or SECMassociated corresponding energy intake (kcal). This was then regressed against the intake of EECM or SECM (kg DM) to determine the IDE, ME or MEn (kcal/kg DM) of the canola meal samples using slopes of the regression lines.

Statistical analysis
Data for the ileal digestibility and total tract retention of energy were analyzed using the GLM procedures of SAS (SAS Institute Inc., Cary, NC, USA). The model included diet and block as the independent variables and individual cage served as the experimental unit. The orthogonal polynomial contrast was used to examine the relationship between energy utilization response criteria and graded concentrations of either EECM or SECM. The EECMor SECM-associated IDE, ME or MEn intake (kcal) was regressed against the intake of EECM or SECM for cage of six birds, respectively. Slopes and intercepts derived from 3 cages of 0, 100, or 200 g of canola meal substitution in each block were generated using the SLOPE and INTERCEPT functions in Microsoft Office Excel 2010 (Microsoft Corp., Redmond, WA, USA), respectively.
The intercept and slope data were analyzed as a one-way ANOVA in a completely randomized design using canola meal type as the independent variable and the intercept or slope as the dependent variable with 1 df for canola meal type and 14 df for the error term (Adeola and Ileleji 2009). In this analysis, experimental unit was a block of 3 cages of 0, 100, or 200 g of canola meal type substitution. Statistical significance level was set at 5 %.

Results
The CP contents in EECM and SECM were analyzed to be similar (398.7 vs. 402.0 g/kg, on DM basis), which was also reflected in the analyzed values of AA for the respective ingredients. The EECM had greater GE (5789 vs. 4895 kcal/kg, on DM basis) and ether extract (EE) (138.7 vs. 24.3 g/kg, on DM basis) contents than SECM whereas the value of neutral detergent fiber (NDF) was greater in the SECM than in EECM (265.3 vs. 233.8 g/kg). Regardless of the canola type, the body weight gain, feed intake, and feed efficiency of the birds were not influence (P > 0.05) by the supplementation of canola meal to the reference diet (Table 4). The data presented in Table 5 show the ileal digestibility and total tract retention of DM, N and energy of experimental diets used in the current study. There were linear (P < 0.05) and quadratic (P < 0.05) decreases in ileal digestibility of DM and energy as well as IDE as the EECM level in the diets increased from 0 to 200 g/kg. As the SECM substitution into the reference diet increased from 0 to 200 g/kg, all the response criteria measured linearly increased (P < 0.05) with the exception of the N retention which was not affected by the substitution of SECM.
The regression of IDE, ME and ME n intake associated with either EECM or SECM against the intake of substituted canola meals depicted in Figs. 1 or 2, respectively. Regression equations of IDE, ME and ME n for the EECM

Table 4 Growth performance of broilers fed the experimental diets containing expeller-extracted canola meal (EECM) or solvent extracted canola meal (SECM) levels at 0, 100, 200 g/kg
Values are means of five replicate cages with six birds per cage SEM standard error of the mean a Linear (L) and quadratic (Q) contrasts for the EECM b Linear (L) and quadratic (Q) contrasts for the SECM  Comparisons of slopes and intercepts for the IDE, ME, ME n regressions between the SECM and EECM are presented in Table 6. The slopes for ME and ME n were greater (P < 0.05) for the EECM than for the SECM, whereas slope for IDE did not differ between the EECM and SECM.

Discussion
The general steps of solvent-extraction method include flaking, cooking, pressing, and solvent-extraction of seeds followed by desolventizing, toasting, and drying of the extracted meal whereas the steps for expellerextraction method only include flaking, cooking, and pressing of seeds (Spragg and Mailer 2007). The difference in extraction method could influence the composition of two types of canola meal. In the current study, the EECM had 18.4 and 471 % greater GE and EE, respectively, than SECM, which could be due to the difference in extraction method because an additional oil-extraction process by solvent is performed for the SECM. Woyengo et al. (2010b) reported greater GE (5199 vs. 4812 kcal/ kg on DM basis) and EE (12.03 vs. 5.54 %, on DM basis) contents in EECM compared with SECM. Moreover the GE and EE contents in EECM used in the previous study was slightly less compared with the present study, which could be attributed to the number of extractions performed (single extraction for the present study vs. double extraction for the previous study) to maximize oil recovery.
The greater NDF content in the SECM compared with the EECM observed in the current study is in agreement with the study by Woyengo et al. (2010b). One of common Maillard reaction products is neutral detergent insoluble nitrogen (Khajali and Slominski 2012) which falls in NDF fraction (van Soest 1994). Maillard reaction can occur during both oil extraction processes due to high temperature of the expeller pressing process for the EECM and desolventizing and toasting processes for the SECM, but the extent of the Maillard reaction may be less for the EECM than for the SECM because the expeller pressing process is faster compared with desolventizing and toasting processes (Landero et al. 2012). The 903 kcal/kg DM greater GE in EECM than in SECM (5798 vs. 4895) is fully accounted for by the 114 g/kg DM greater EE in EECM than in SECM (138.7 vs. 24.3) and the 31.5 g/kg DM lower NDF in EECM than in SECM (238.8 vs. 265.3) using gross energies of 9.2 and 4.5 kcal/g for EE and NDF.
Regardless of the type of canola meals, the influences by the addition of graded canola meal were not observed. These results might be attributed to the short experimental period in which the effect of treatments on the growth performance may not be obvious (Woyengo et al. 2010b).
Substitution of two canola meals at 100 or 200 g/kg of diet for the energy supplying ingredients (corn, soybean meal, soybean oil, and cornstarch) in the reference diet linearly decreased ileal DM and energy digestibility as well as the IDE of diet. This may be attributed to Table 5 Dry matter (DM), energy, and nitrogen (N) digestibility and total tract retention of broilers fed the experimental diets containing expeller-extracted canola meal (EECM) or solvent extracted canola meal (SECM) levels at 0, 100, 200 g/ kg SEM standard error of the mean, IDE ileal digestible energy, AME apparent metabolizable energy, AME n nitrogen-corrected AME a Linear (L) and quadratic (Q) contrasts for the EECM b Linear (L) and quadratic (Q) contrasts for the SECM  son 1990). The physical barrier of the cell walls of fiber can encapsulate potentially available nutrients. Furthermore, the viscous properties of fiber may interfere with the digestion process and thereby reduce the digestibility of other nutrients (Choct and Annison 1992;Steenfeldt 2001). Consequently, the fiber components have negative impact on the digestibility of nutrients in the broiler chickens and lead to lower energy utilization. High dietary fiber may increase the passage rate of digesta (Khajali and Slominski 2012), which adversely affect the digestibility of nutrient.
In the current study, the EECM compared with SECM had 35 or 33 % more ME (2807 vs. 2077 kcal/kg) or ME n (2679 vs. 2013 kcal/kg) estimated by the regression method, respectively. Using the difference method, Woyengo et al. (2010b) also reported greater AME (3039 vs. 2005 kcal/kg, on DM basis) and AME n (2694 vs. 1801 kcal/kg, on DM basis) contents in EECM than in SECM fed to broilers. This in part could result from the difference in the fat content between the EECM and SECM. The concentration of ether extract in EECM used in the present study or the study by Woyengo et al. (2010b) was 471 (13.87 vs. 2.43 %) or 217 % (5.54 vs. 12.03 %) greater than in SECM, respectively. The significant differences in ME or ME n between the SECM and EECM were observed in the present study whereas the difference in the IDE was not significant. The reason for these results was not clear. Speculatively, difference in the fermentation of undigested dietary fibrous contents might have contributed to such difference in ME values. Because microbes in the ceca of broiler chickens are able to ferment undigested dietary fiber contents Fig. 1 Regression of energy (IDE, ME or ME n ) intake (Y, kcal) associated with expeller-extracted canola meal (EECM) intake against the intake of EECM (X, kg DM) for birds fed experimental diets from day 14 to 21. Regression equations of IDE, ME and ME n for the EECM-substituted diet were Y = −21.2 + 3035X (r 2 = 0.946), Y = −1.0 + 2807X (r 2 = 0.884) and Y = −2.0 + 2679X (r 2 = 0.902), respectively

Fig. 2
Regression of energy (IDE, ME or ME n ) intake (Y, kcal) associated with solvent-extracted canola meal (SECM) intake against the intake of SECM (X, kg DM) for birds fed experimental diets from day 14 to 21. Regression equations of IDE, ME and ME n for the SECM-substituted diet were Y = 20.7 + 2881X (r 2 = 0.962), Y = 27.2 + 2077X (r 2 = 0.875) and Y = 24.7 + 2013X (r 2 = 0.901), respectively Table 6 Comparison of two types of canola meal for intercepts and slopes of the regressions in the determination of ileal digestible energy (IDE), metabolizable energy (ME) and nitrogen-corrected ME (ME n ) of expeller-or solvent extracted canola meal SEM standard error of the mean

Item
Intercept ( entering from the ileum into short chain fatty acids which could be absorbed by broiler chickens and used as energy (Sugahara et al. 2004).

Conclusion
In conclusion, the results showed that the EECM used in the current study had more fat as well as less NDF and consequently more GE content compared with the SECM. In addition, the ME and ME n contents of the EECM were greater than those of the SECM, which indicates that the EECM might be a superior energy source for broiler chickens compared with the SECM when both canola meals are used to reduce the cost of feeding.