Utilisation of reactive lysine from meat and bone meals of different ash content by growing-finishing pigs Kirsi Partanen Department ofAnimalScience, PO Box 28, FIN-00014 University ofHelsinki, Finland. Current address: Agricultural Research Centre ofFinland, Animal Production Research, Animal Nutrition, FIN-31600 Jokioinen, Finland, e-mail: kirsi.parlanen@mtt.fi Hilkka Siljander-Rasi,Timo Alaviuhkola Agricultural Research Centre ofFinland, Animal Production Research, Pig Husbandry, Tervamdentie 179, FIN-05840 Hyvinkää, Finland Nina van Gilse van der Pals Department ofAnimal Science, PO Box 28, FIN-00014 University of Helsinki, Finland A growth experiment was conducted using 50 pigs (25-100 kg) to evaluate the use of meat and bone meals of different ash content as a substitute for soyabean meal (SBM) for growing pigs and the potential of l-fluoro-2,4-dinitrobenzene (FDNB) reactive lysine in diet formulation. The control diet consisted of barley and SBM. For test diets, either 33 or 67% of SBM was replaced with meat and bone meal of low (ML, 205 g ash/kg) or high (MH, 349 g ash/kg) ash content. SBM, ML33, ML67, MH33 and MH67 diets contained 7.8, 7.8, 8.2, 7.8 and 7.9 g FDNB-reactive lysine/feed unit (feed unit is equivalent to 9.3 MJ NE), respectively. For these diets, average daily live weight gains (ADG) were 859, 830, 805, 854 and 813 g/d with feed conversion ratios of 2.25, 2.40, 2.41, 2.31 and 2.44 feed units/kg, respectively. Pigs fed the SBM diet grew faster (PcO.Ol) and utilised feed more effi- ciently (PcO.001) than those offered ML and MH diets. The ADG decreased (PcO.001) with increas- ing meat and bone meal dietary inclusion. These results indicate that FDNB-reactive lysine is unsuit- able for diet formulation as it may be incompletely absorbed. Faecal digestibilities of nutrients in the experimental diets were determined at live weights of 30, 44, 64 and 85 kg, respectively. Mean di- gestibility of crude protein (CP) was 74, 74, 68, 75 and 72%, while that of crude fat (CF) was 44, 55, 51, 48 and 41%, for SBM, ML33, ML67, MH33 and MH67 diets, respectively. Faecal digestibilities of CP and CF increased with live weight, with the largest increase being observed between 44 and 65 kg. Increased replacement of SBM with ML or MH increased back fat oleic acid content (PcO.Ol), decreased back fat firmness (P<0.05) but had little influence on palatability. Key words: animal by-products, carcass quality, fatty acids, performance © Agricultural and Food Science in Finland Manuscript received August 1997 Vol. 7(1998): 1-11. AGRICULTURAL AND FOOD SCIENCE IN FINLAND Partanen, K. et al. Utilisation ofreactive lysine by growing-finishing pigs ntroduction Meat and bone meal is considered to be a good source of supplemental protein, calcium and phosphorus and is commonly used at low levels in commercial pig diets. However, growth rates and efficiency of feed utilisation are often re- duced when increasing levels of dietary soya- bean meal are replaced with meat and bone meal (Evans and Leibholz 1979, Cromwell et al. 1991). Impaired performance may result from reduced quality of dietary protein, excessive mineral intake or depressed diet palatability. The quality of protein in meat and bone meal is influenced by the raw materials used and the processing methods and conditions (temperature, pressure and duration of heating) applied (Bat- terham et al. 1986a, 1986b, Donkoh et al. 1994). Protein of soft offal is typically highly digesti- ble and has a more desirable amino acid profile than that of bone or connective tissue. However, application of heat causes a number of reactions within the protein structure. This results in di- gestibility and availability being reduced, and can lead to amino acids in meat and bone meal being destroyed (Batterham et al. 1986a, 1986b). Lysine is the most sensitive amino acid to heat damage, but other amino acids, e.g., methionine, cystine and tryptophan are also susceptible (Pa- padopoulos 1989). It is generally accepted that the ileal amino acid digestibility assay provides an accurate es- timate of lysine availability for unheated feed- stuffs, while for heatedprotein sources it can lead to overestimates (Moughan et al. 1991, Batter- ham 1992). Overestimates can arise during the acid hydrolysis step of the conventional amino acid analysis causing some of the heat modified residues reverting back to lysine. While modi- fied lysine residues may also be partially ab- sorbed from the tract they are of little or no nu- tritional value to the pig. In the case of heated protein sources, the accuracy ofdiet formulation can be improved by using lysine availability de- termined by the slope-ratio assay (Batterham 1992). However, this method is slow and expen- sive, and values of availability may vary depend- ing upon adopted response criteria (Leibholz 1992). Therefore, alternative methods are need- ed for routine evaluation of lysine availability in feeds of highly variable composition, such as meat and bone meal. One approach is to assess availability by chemical analysis. In the case of lysine, residues with an e-amino groupfree to react with chemi- cal reagents are practically the most important source of available lysine. Carpenter (1960) has developed a method where following reaction with l-fluoro-2,4-dinitrobenzene (FDNB), and subsequent acid hydrolysis, e-fluoro-dinitroph- enyl lysine (reactive lysine) is assessed. Despite some disadvantages, the method is still among the most useful in practice, because it allows largeranges of potential protein heat damage to be assessed (Hurrell and Carpenter 1974). Meatand bone meal fat content can vary from a few grams to over 150 g/kg, depending on the method of fat separation (mechanical or solvent extraction). Eating quality and fatty acid com- position of pork are sensitive to changes in the quantity and quality of dietary fat (Miller et al. 1990, Madsen et al. 1992). High levels of die- tary animal fat increase the content of oleic acid in body fat (Mortensen et al. 1983) which may have a negative influence on fat firmness (Mad- sen et al. 1992). Consequantly, high levels of meat and bone meal in a diet can have undesira- ble effects on the technical and eating quality of pork. The aim of this study was to evaluate two meat and bone meals of different ash content as a protein source for growing pigs and the use of FDNB-reactive lysine in diet formulation. Ap- parent faecal digestibilities of dietary nutrients were determined to allow net energy supply from the experimental diets to be estimated. Since the meat and bone meals studied had a relatively high fat content, the sensory characteristics of meat and adipose fatty acid profile were also assessed. 2 AGRICULTURAL AND FOOD SCIENCE IN FINLAND Material and methods Animals, diets and measurements A performance study was conducted using 50 pigs ofFinnish Landrace, Large White or cross- bred origin (20:20; 10) using a randomised com- plete block design. The initial live weight of the pigs was approximately 25 kg. A block was formed from five pigs of the same sex and litter origin. Within a block, pigs were randomly as- signed to five dietary treatments. Pigs were housed individually in concrete floor pens. They were weighed in the beginning of weeks 1,3, 6, 9 and 12 of the experiment and then weekly un- til they reached the minimum slaughter weight of 98 kg or had been in the experiment for 15 weeks. Two meat and bone meals were manufactured in a traditional batch dry rendering system with approximately a two hour cooking cycle and ster- ilisation under pressure at about 125°C for 20- 30 minutes. Fat was removed mechanically af- ter rendering and meals were ground. The raw material consisted of cattle and swine offal and bones in different proportions, as indicated by the meal ash content. Meals with low and high ash content (Table 1) were designated as ML (205 g ash/kg) and MH (349 g ash/kg), respec- tively. The control diet consisted of barley and soya- bean meal (SBM). In the other four diets, ei- ther 33 or 67% of SBM was replaced with an equal amount of digestible crude protein derived from ML or MH (Table 2). The diets were for- mulated to have an equal amount of FDNB-re- active lysine per feed unit with the addition of L-lysine-HCI when required. The net energy content of ingredients was calculated using di- gestibility coefficients determined in previous studies for barley (Partanen et al. 1992), SBM and similar meat and bone meals (Partanen 1994). Dietary content of threonine, methionine and cystine, minerals and vitamins were calcu- lated to satisfy the Finnish feeding recommen- dations (Salo et al. 1990). Pigs were fed twice Table 1.Chemical composition ofbarley, soyabean and meat and bone meals (g/kg). Barley Soyabean Low ash High ash meal meat and meat and bone meal bone meal Dry matter 883 887 984 974 Crude protein 120 420 534 449 Crude fat 20 32 175 15920 32 175 159 Ash 21 55 205 349 Lysine 4.3 26.4 25.6 18.4 FDNB-lysine 4.1 23.1 19.7 15.7 Threonine 4.0 16.4 18.7 13.04.0 16.4 18.7 13.0 Isoleucine 4.0 18.5 16.5 11.2 Leucine 8.0 31.9 32.0 22.5 Valine 5.6 20.1 23.5 17.1 Phenylalanine 5.9 20.6 18.1 13.5 Tyrosine 2.9 13.4 12.8 8.52.9 13.4 12.8 8.5 Histidine 2.8 11.3 10.1 7.2 daily according to a restricted feeding scale (1.0- 2.8 feed units/d). Daily feed allowance was in- creased by 0.2 feed units/week in the beginning and by 0.1 feed units/week after week nine of the experiment. Faecal digestibilities of nutrients and energy were determinedusing acid insoluble ash (AIA) as a marker. Grab samples were collected after morning feeding from three randomly selected blocks of barrows for five days during weeks 3, 6, 9 and 12of the experiment. During collection periods, no bedding was used and the pens were cleaned prior to morning feeding. Only the top part of faeces was collected from the floor im- mediately after defecation. Determined digesti- bility coefficients were used to recalculated die- tary net energy supply according to Tuori et al. (1995), with one feed unit being equal to 9.3 MJ NE. Pigs were slaughtered at a commercial slaughter house and cold carcass weight was measured after overnight chilling. The left half of the cold carcass was partitioned into valuable cuts (fore-end and shoulder, loin, and ham) which were dissected into lean (including bones) and fat (including skin) to determine the propor- tion of lean in the valuable cuts and whole car- 3 Vol. 7(1998): 1-11. AGRICULTURAL AND FOOD SCIENCE IN FINLAND Table 2. Formulation and chemical composition of experimental diets (g/kg) containing soyabean or meat and bone meal. Protein source SBM ML MH Substitution level, % 33 67 33 67 Ingredient composition: Barley 808.8808.8 808.8808.8 808.8 Barley starch 14.027.4 19.523.3 Soyabean meal 150.0100.0 50.0100.0 50.0 Low ash meat and bone meal 47.094.0 - High ash meat and bone meal - - 52.5105.0 L-lysine-HCI 0.61.1 1.61.1 1.4 DL-methionine 2.12.1 2.22.1 2.0 Calcium carbonate 9.08.0 7.06.0 Monocalcium phosphate 18.59.0 - - - Trace element premix l 2.02.0 2.02.0 2.0 Vitamin premix2 5.05.0 5.05.0 5.0 NaCl 4.03.0 2.03.0 2.5 Chemical composition: Dry matter 885 887 889 883 889 Crude protein 156 160 162 159 159 Crude fat 27 34 41 34 41 Crude fibre 49 45 47 45 40 Ash 49 48 48 48 59 Calcium 8.89.4 10.710.2 15.2 Phosphorus 7.16.7 6.46.5 9.4 Lysine 3 7.98.2 8.58.0 7.8 FDNB-lysine 1 7.37.4 7.67.3 7.2 Methionine+cystine 3 6.46.5 6.86.5 6.5 Threonine 1 5.75.8 5.85.6 5.4 Gross energy, MJ/kg 16.216.5 16.7 16.4 16.4 Net energy, FU/kg 4 0.930.95 0.930.94 0.91 SBM = soyabean meal; ML = low ash meat and bone meal; MH = high ash meat and bone meal ' Composition per kg diet: 5000 IU vitamin A, 750 1 U vitamin D,, 15 mg vitamin E, 25 mg vitamin C, 0.25 mg vitamin K, Img thiamine, 2.5 mg riboflavin, 4mg pyridoxine, 0.015 mg vitamin Br , smg nicotinic acid, and 5 mg pantothenic acid. : Composition per kg diet: 26 mg Fe, 92 mg Zn, 26 mg Mn, 26 mg Cu, and 0.12 mg Se. ' Calculated from ingredient amino acid composition, except for methionine+cystine which were based on published values (Partanen 1994). 4 Calculated according to Tuori et al. (1995) based on measured digestibility coefficients. cass. About 100 g of back fat and 100-150 g meat were taken from the loin posterior to the last rib from 6 randomly selected blocks (3 gilts, 3 bar- rows) for fatty acid analysis and organoleptic grading, respectively. Samples were frozen pri- or to analysis. Few samples were lost due to tech- nical problems. Chemical analyses and organoleptic ■• 5 du 6 Chemical composition offeed components, feeds and faeces was determined according to stand- ard AOAC methods (1984). Crude fat was deter- 4 Partanen, K. et al. Utilisation ofreactive lysine by growing-finishing pigs mined after hydrolysis in 4 M HCI. The amino acid composition of feed components was ana- lysed by ion-exchange chromatography with post-column O-phthalaldehyde detection. Be- fore analysis, samples were hydrolysed in 6 M HCI for 22 h at 110°C.FDNB-reactive lysine was measured according to Carpenter (1960). Die- tary amino acid composition was calculated from that of its components. Since methionine and cystine were not determined, values reported by Partanen (1994) were used in the calculations. Phosphorus in feeds was determined by color- imetry (Tayssky and Shorr 1953) and calcium was measured by atomic absorption spectrosco- py. AIA was measured according to Van Keulen and Young (1977). Fat was extracted from feeds with methanol and chloroform (Karow et al. 1984), esterified with hexane, sodium methoxide and calcium chloride and analysed by gas chromatography. In order to analyse the fatty acid profile of adi- pose tissue, skin was removed from a 100 g sam- ple of back fat. After grinding and melting (80°C), fatty acids were saponified with 0.5 M KOH-methanol solution and esterified with BF3 - methanol reagent. Methyl esters of fatty acids were extracted with hexane and analysed by gas chromatogaraphy. The values are given as per- centages of total fatty acids. All analyses were performed in duplicate. For organoleptic evaluation, samples of mus- culus longissimus dorsi were thawed and fried as described by Partanen et al. (1992). A trained test panel of three to five members graded fried samples for tenderness, juicinessand flavour by using a scale of 1 to 7, with 7 being the highest and 1 the poorest grade. Statistical analyses Data were analysed by the GLM procedure of SAS (1985). Performance and carcass data were subjected to a least-squares analysis of variance (Snedecor and Cochran 1989) using the model: Y = u + T. + B + e.., U 1 J u where T and B are the effects of treatment i■ j and block], respectively, and £ y is the error term. A split- plot design was applied for the statis- tical analysis of digestibility data using the model: Yjjki =H+T, + B j +e„+ W k + (W*T) ik + (W*B) jk + Eijkl’ where T, B and W,. are the effects of treat-r i k ment i, block j and week of sampling k, respec- tively, and e.. is the main plot error and £. jk) the sub-plot error. Four orthogonal contrasts were formed to allow the following comparisons: Cl = SBM vs. ML and MFI diets, C 2 = 33 vs. 67% of SBM replaced with ML or MFI, C 3 = ML vs. MH diets, and C 4 = interaction C 2 x C3. Only contrasts which are statistically significant (P<0.05) are presented in the tables. Results The chemical composition of the main dietary ingredients is shown in Table 1. The proportion ofFDNB-reactive lysine of total lysine was 88, 77 and 85% for SBM, ML and MH, respective- ly. Chemical composition of the diets is shown in Table 2. SBM, ML33, ML67, MH33 and MH67 diets contained 7.8, 7.8, 8.2, 7.8 and 7.9 g FDNB-reactive lysine/feed unit, respectively. One pig in treatment MH67 was removed from the experiment due to a bleeding ulcer. The pig ate and grew well for the first 12 weeks of the experiment. Consequently, performance data of weeks 1-12 are included in the calculations. All other pigs remained in good health through- out the study. Faecal digestibilities oforganic matter, crude protein (CP) and crude fat (CF) of the experi- mental diets are given in Table 3. CP digestibil- ity decreased with increasing inclusion of meat and bone meal (P<0.01) and was lower in ML than MH diets (P<0.05). CF digestibility was lower in MH than ML diets (P<0.01) and tended to decrease with increasing level of meat and 5 Vol. 7(1998): 1-11. AGRICULTURAL AND FOOD SCIENCE IN FINLAND Partanen, K. et al. Utilisation ofreactive lysine by growing-finishingpigs Table 3. Apparent faecal digestibilities (%) of nutrients in diets containing soyabean and meat and bone meal. Protein source SBM ML MH Significance 1 Substitution level, % 33 67 33 67 SEM Wk CI C 2 C 3 Organic matter (mean) 81.381.4 79.281.4 79.40.21 *** * *** week 3 80.180.1 77.880.6 79.00.42 week 6 80.380.7 78.680.5 78.10.42 week 9 82.183.1 80.182.1 79.50.42 week 12 82.681.8 80.582.3 80.70.42 Crude protein (mean) 74.274.2 68.475.3 72.10.66 *** ** * week 3 70.970.9 62.772.5 70.01.31 week 6 70.772.4 66.173.0 69.51.31 week 9 76.877.1 71.477.3 73.31.31 week 12 78.476.3 73.378.5 75.41.31 Crude fat (mean) 44.154.5 51.147.6 41.21.05 * ** week 3 43.253.2 47.746.9 40.62.10 week 6 42.253.7 51.946.7 35.42.10 week 9 46.558.9 52.448.8 44.12.10 week 12 44.652.3 52.348.2 44.52.10 SBM = soyabean meal; ML = low ash meat and bone meal; MH= high ash meat and bone meal 1Wk: effect of week; Contrasts: Cl: SBM vs. ML and MH diets, C2: 33 vs. 67% of SBM replaced with ML or MH; C3: ML vs. MH diets; Significance: *** (PcO.OOI), ** (P<0.01), * (P<0.05). bone meal in the diet (P=0.06). Mean pig live weight at the beginning of weeks 3,6, 9 and 12, when the faecal digestibilities were determined, was 30, 44, 64 and 85 kg, respectively. Both lin- ear and cubic effects of time were significant for digestibility of all nutrients, except ash. The larg- est increase in nutrient digestibilities was ob- served between live weights of 44 and 65 kg. Pigs fed SBM diet grew faster (PcO.Ol), uti- lised feed more efficiently (P<0.01) and tended to have a higher meat percentage in carcass (P=0.08) compared with pigs fed ML and MH diets (Table 4). Average daily live weight gain (ADG) was depressed (PcO.001) and the feed conversion ratio (FCR) tended to improve (P=0.07) with increasing replacement of SBM with ML or MH. A slightly lower (P=0.09) ADG was observed for ML than MH diets, but there were no differences in FCR. No significant dif- ferences were observed in back or side fat thick- ness between treatments. Fatty acid profile of back fat was clearly in- fluenced by dietary treatment (Table 5). The con- tent of oleic acid (C 18:1) increased with increas- ing level of meat and bone meal in the diet (P<0.05), whereas that of linoleic and linolenic acid decreased (PcO.ool and P