Effect of fatty acids source on growth performance, carcass characteristics, plasma urea nitrogen concentration, and fatty acid profile in meat of pigs fed standard- or low-protein diets

Thirty six Landrace × Yorkshire barrows with 18.6 kg of initial body weight were used to evaluate three sources of fatty acids: crude soybean oil, conjugated linoleic acid (CLA), and soybean soapstock in standard crude protein (CP) and low-protein diets for starter (21 d; 205, 160 g kg-1 CP), growing (28 d; 160, 145 g kg-1 CP), and finishing (29 d; 140, 125 g kg-1 CP) phases. Growth performance, carcass characteristics, plasma urea nitrogen concentration and fatty acid profile in meat were evaluated. The reduction of CP diminished average daily gain, feed:gain ratio, longissimus muscle area and plasma urea nitrogen concentration in nursery pigs; reduced longissimus muscle area and plasma urea nitrogen concentration in growing pigs; increased average daily feed intake, and reduced lean meat percentage and plasma urea nitrogen content in finishing pigs. It also increased c9, t11 and c11, t9 CLA isomers and total lipids and lowered eicosapentaenoic and docosahexaenoic acids concentrations in semimembranosus muscle; linolenic acid decreased with low-protein diets in longissimus and semimembranosus muscles; the oil type affected the concentration of c9, t11 and c11, t9 CLA isomers and total saturated fatty acids in semimembranosus muscle; CLA increased individually and total saturated fatty acids, reduced linoleic and docosapentaenoic acids, and increased total lipids in longissimus muscle. These results indicate that decreasing CP changes the profile of fatty acids. The soybean soapstock can replace crude soybean oil in pig diets; while conjugated linoleic acid does not improve response of pigs fed standard- or low-protein diets.


Introduction
The proper addition of synthetic amino acids to sorghum-soybean meal diets formulated to lower content of crude protein (CP) than the suggested by NRC (1998) does not adversely affect the growth performance of pigs (Myer & Gorbet, 2002;Figueroa et al., 2003) and reduces the excretion of N excess from the standard diets (Kerr et al., 2003a). However, these diets have negative effects on carcass traits, and there is less lean meat gain and greater accumulation of adipose tissue (Figueroa et al., 2002;Gómez et al., 2002b). The addition of conjugated linoleic acid (CLA) to pig diets could be an alternative to reduce the negative effects of low-protein diets (LPD), because CLA has a lipolytic effect on adipose tissue (Mersmann, 2002).
Due to the high cost of energy and protein ingredients used in animal feed it is necessary to use alternative feedstuffs. Soybean soapstock (SS) is a byproduct of the crude soybean oil (CSO) extraction and, if the SS is properly processed, it can be added to animal feed (Bruce et al., 2006;Dumont & Narine, 2007). This byproduct is a cheap source of energy and can be used in pig diets because its cost can be up to half of the CSO. However, due to its high concentration of free fatty acids, it has lower intestinal absorption due to the inadequate formation of micelles, which affects its digestibility and energy value (Mateos et al., 1996). It is therefore important to assess its use in LPD where a marginal deficiency of either nutrient can reduce the productive response. The objective of this study was to evaluate the effect of conjugated linoleic acid, soybean soapstock, and crude soybean oil as energy sources on growth performance, carcass characteristics, plasma urea nitrogen concentration, and fatty acid profile of meat from pigs fed standard or low-protein diets.
Palabras clave adicionales: aditivos alimenticios; calidad de carne de cerdo; manejo alimenticio. Fatty acids source in standard or low-protein diets for pigs samples of blood were kept on ice until centrifugation at 2,500 rpm (1286 g) for 15 min; the supernatant was transferred to polypropylene tubes and stored at -20°C until determination of plasma urea nitrogen (PUN) concentration (Chaney & Marbach, 1962). At the beginning and the end of each stage, backfat thickness and longissimus muscle area were measured with a real time ultrasound (Sonovet 600, Medison Inc., Cypress, CA, USA). These data and the initial and final body weights were used to calculate the fat free lean gain and the lean meat percentage in the carcass using the NPPC (1991) equation. The CP content was determined in diets (AOAC, 1990) for each stage of growth.
All pigs were slaughtered at the end of the finishing phase, obtaining samples in warm carcass of approximately 100 g from semimembranosus and longissimus muscles. The samples were ground and stored at -20°C until determination of fatty acid profile. The total lipid content (923.07 method), the fatty acid profile (saturated, unsaturated, polyunsaturated, and isomers of CLA) in the oils (Table 2) and in muscle tissue samples (994.10 method), and the lipid extract (methylated with a complex of 20% methanol trifluorite boron in a methanol solution) were determined (AOAC, 2000). Fatty acid profile of oil and muscle tissue samples were analysed by gas chromatography using a DB23 column (30m × 0.25 mm id) in a Varian 3400 gas chromato-  Ingredient/Treatment  T1 T2 T3 T4 T5 T6  T1 T2 T3 T4 T5 T6  T1 T2 T3 T4 T5   6.6 6.6 6.6 6.6 6.6 6.6 Threonine 6.5 6.5 6.5 6.5 6.5 6.5 5. graph equipped with an autosampler and a flame ionization detector. The carrier gas used was N at a flow of 30 mL min -1 . The operating temperatures were: column 230°C; injector 150°C; detector 300°C. The retention times were compared with methylated fatty acids standards (Supelco 18919-1 AMP). The data were analysed for ANOVA using the GLM procedure (SAS, 2002). The model included the effects of protein level and oil type as main factors and their interaction; the initial weight of pigs was used as covariate in the analysis of variables where this factor had a significant effect. Multiple mean comparisons were performed using Tukey's test; the alpha level for determination of significance was 0.05.

Starter period
The protein level affected (p < 0.05) ADG and final weight ( Table 3). The highest value was observed in pigs fed standard CP diet. The main factors did not affect the ADFI (p > 0.05), but there was an interaction between CP level and oil type (p < 0.006) for pigs fed standard CP diet (p < 0.05) with CSO (1.38 kg d -1 ) or CLA (1.71 kg d -1 ). The FGR was 8.5% higher in pigs fed LPD (p < 0.05). The interaction (p < 0.01) between CP and oil type affected the final backfat thickness, with the greatest thickness in pigs fed standard CP diet and CLA (p < 0.05). The final longissimus muscle area was greater (p < 0.05) in pigs fed the standard CP diet than in pigs fed LPD (1,424 vs. 1,297 mm 2 ). The pigs fed SS had the highest values (1,409 mm 2 ) of longissimus muscle area (p < 0.05) and the lowest (1,283 mm 2 ) was with CSO, irrespective of the CP level.

Growing period
The protein level or oil type did not affect growth performance or carcass characteristics (p > 0.05, (Table 4). However, there was a trend (p = 0.06) to reduce the longissimus muscle area when the CP decreased in the diet. The plasma urea nitrogen concentration was 14% lower in pigs fed the LPD, indicating a reduction of the amino acid excess or a better balance of them in the diet.

Finishing period
The ADFI increased by 290 g d -1 (p < 0.05) when pigs were fed LPD (Table 5). The fat free lean gain was higher (404 g d -1 ; p < 0.05) in pigs fed CSO, and the lowest value (334 g d -1 ) was observed in pigs fed CLA. The diet did not change backfat thickness or longissimus muscle area (p > 0.05). Lean meat percentage was higher (p = 0.051) in pigs fed standard CP diet. The plasma urea nitrogen concentration was 14.2% lower in pigs fed LPD; the interaction between CP × Oil type showed a trend (p = 0.059) to affect the concentration of plasma urea nitrogen concentration, with lower values in pigs fed CSO.

Semimembranosus muscle
Pigs fed LPD showed the highest values (p < 0.05) of CLA isomers (c9, t11 and c11, t9) and eicosaenoic acid (Table 6), and also showed the lowest concentration (p < 0.05) in linolenic, eicosapentaenoic and docosahexaenoic acids. Docosapentaenoic acid decreased (p = 0.075) and the total lipid level increased (p = 0.062) in pigs fed LPD. The CLA diets increased (p < 0.05) the isomers c9, t11 and c11, t9 of CLA. The eicosaenoic acid concentration was similar in pigs fed with CSO and SS, and less with CLA (p < 0.05). The total saturated fatty acids concentration was higher in pigs fed CLA, and less with CSO (p < 0.05). The interaction between CP level and type of oil affected the concentrations of the cis10-pentadecaenoic, cis10-heptadecaenoic, and cis-vaccenic acids (p < 0.05).

Growth performance
The reduction of CP level lowered growth performance of pigs during the starter period, although the CP can be reduced by 40 (Hansen et al., 1993a), 55 (Le , and up to 60 g kg -1 (Kerr  al., 1995), adding synthetic amino acids for this phase of growth.
In the growing stage, the decrease of CP did not affect the growth performance in pigs fed sorghum- (Hansen et al., 1993b) or corn-soybean meal with 140 (Tartrakoon et al., 2004), 130 (Tuitoek et al., 1997a) and 120 g CP kg -1 (Kerr et al., 1995(Kerr et al., , 2003aFigueroa et al., 2002). However, Kendall et al. (1998) found that reducing the CP up to 122 g kg -1 decreased ADG and feed efficiency; this effect can be avoided by the addition of synthetic valine and isoleucine to diet (Le Bellego et al., 2001;Zervas & Zijlstra, 2002). The variability of the results in growing pigs may be due to: 1) the genetic potential of pigs, because there was a great variation of genotypes of pigs used between studies; 2) the feedstuffs of the diets (corn, sorghum, or wheat-soybean meal), because the amount and type of soluble carbohydrates in the diet influence the growth performance of pigs (Atakora et al., 2003); 3) the variation in the initial weight of pigs; and 4) the duration of the evaluation period.
In finishing pigs, reducing the CP in the diet increased ADFI, but did not affect ADG and FGR, as in pigs fed corn-soybean meal diet with 30 g kg -1 less CP (Kerr et al., 2003b), supplemented with synthetic amino acids (Kerr et al., 1995). However, this result has not been consistent since Tuitoek et al. (1997a) and Panetta et al. (2006) indicated that ADG and feed efficiency are adversely affected with a similar reduction in the CP.
Addition of CLA to diets does not have an effect on growth performance of pigs. This result differs from other studies where CLA addition improved some growth response variables (Ramsay et al., 2001;Wiegand et al., 2002;Sun et al., 2004;Lauridsen et al., 2005;Martin et al., 2008a).
Diets supplemented with SS maintained growth performance at similar levels as with CSO or CLA (Starkey et al., 2002a). In finishing pigs fed 60 g kg -1 of SS for 70 d improved ADG and feed efficiency (Starkey et al., 2002b).

Carcass traits
The CP levels analysed in this research did not influence carcass characteristics even when CP was reduced by 40 g kg -1 in sorghum-soybean meal (Myer & Gorbet, 2002) or corn-soybean meal (Le Kerr et al., 2003a) diets supplemented with synthetic amino acids. This amino acids addition is necessary because of deficiency of some of them, such as valine and isoleucine, may produce a greater accumulation of lipids (Kerr et al., 1995;Atakora et al., 2003) and decreases protein in body tissues (Gómez et al., 2002b), reflected in less fat free lean gain and longissimus muscle area in carcass (Figueroa et al., 2002). It has been suggested that a greater reduction of CP in the diet could negatively affect the growth performance and carcass composition (protein and lipid deposition; Tuitoek et al., 1997b).
Addition of CLA increased longissimus muscle area in starter pigs (Wiegand et al., 2002), but had no effect on carcass traits in growing and finishing pigs (Eggert et al., 2001;Averette-Gatlin et al., 2002;Martin et al., 2008a), although some reports have indicated an increase in carcass quality adding 5 to 20 g kg -1 of CLA in growing (Ramsay et al., 2001) and growing-finishing pigs (Lauridsen et al., 2005), and a decrease in backfat thickness (Thiel-Cooper et al., 2001).
Supplementing SS increased longissimus muscle area in starter pigs, and improved fat-free lean gain in finishing pigs. Therefore, the inclusion of SS to meet the energy requirement for fattening pigs fed sorghumsoybean meal diets is a good choice because there is no negative effect on growth response of pigs and it is a cheaper ingredient than CSO.

Plasma urea nitrogen concentration
The reduction of plasma urea nitrogen concentration in pigs fed LPD evaluated in this study indicated a lower N excretion in urine (Gómez et al., 2002a;Tartrakoon et al., 2004) and lower production of ammonia, which is proportional to the reduction of CP in the diet (Powers et al., 2007). Plasma urea nitrogen content and N excretion were linearly reduced (Figueroa et al., 2002) up to 60% when the CP decreases by 40 g kg -1 , with the risk of affecting negatively growth performance and carcass characteristics (Kendall et al., 1998). This problem could be solved by adding limiting essential amino acids : See Table 4. (Zervas & Zijlstra, 2002) and adjusting energy concentration (Herr et al., 2000;, because, with this type of diet, there is a lower energy requirement for deamination, transamination and urea synthesis (Gómez et al., 2002b). It is important not reducing too much the CP to prevent negative effects on pigs response, but it is a cheaper strategy to control the production of gases in intensive pig production (Hayes et al., 2004). The addition of CLA did not change plasma urea nitrogen concentration, which was observed in growing pigs fed 10 g kg -1 of CLA (Ramsay et al., 2001) or pigs that weighed between 105 and 153 kg fed 7.5 g kg -1 of CLA (Corino et al., 2008). This indicates that the main effect of CLA is on lipid metabolism, not on productive variables.

Fatty acid profile in meat
The accumulation of CLA isomers in intramuscular fat of longissimus and semimembranosus muscles has also been reported mainly for longissimus muscle (Eggert et al., 2001) and this concentration is directly related to CLA concentration in the diet (Ramsay et al., 2001;Joo et al., 2002). Isomers of CLA Table 6. Effect of crude protein concentration and oil source on total lipids and fatty acid profile (g kg -1 ) in pork semimembranosus muscle  -Cooper et al., 2001;Lauridsen et al., 2005;Martin et al., 2008b). It could also be a greater accumulation of the isomer t10,c12 (Ramsay et al., 2001) due to the proportion of CLA isomers in the product added to the diet. The increase in total saturated fatty acids and the reduction of total monounsaturated fatty acids in intramuscular fat has also been reported in pigs fed 10 g kg -1 CLA replacing sunflower oil (Eggert et al., 2001), corn oil (Averette-Gatlin et al., 2002) or CSO (Wiegand et al., 2002). Dietary addition of CLA increased the saturated fatty acids both individually and total concentration in meat of longissimus muscle, and reduced linolenic and docosapentaenoic acids. A similar concentration (2.5 g kg -1 ) of CLA used in growing pigs resulted in a decrease of oleic, linoleic, linolenic and arachidonic acids in fat tissue, but only decreased linolenic acid in muscle tissue (Ramsay et al., 2001). This indicates that prolonged periods of treatment may : See Table 6. affect the amount and type of fat in the meat of fattening pigs. The increase in saturated fatty acids (such as stearic acid) and the reduction of unsaturated fatty acids (such as oleic acid) in muscle from pigs fed CLA may be due to the inhibition and downregulation of the enzyme Δ-9-steroil-CoA desaturase (Ramsay et al., 2001), which is involved in synthesis of monounsaturated fatty acids (Lee et al., 1998), reducing its activity in response to higher concentrations of linolenic acid, increasing lipogenesis and the activity of enzymes such as acetyl-CoA-carboxylase (Kouba & Mourot, 1997). For this reason it was important to analyse the fatty acid profile of the oils used in this study (Table 2) and to verify the presence of linolenic acid, because, with the reduced activity of the enzyme Δ-9-esteroil-CoA desaturase, the endogenous synthesis of CLA is inhibited, enhanced by some precursors in the diet such as the C18:1 trans (Gläser et al., 2000).
The changes in fatty acids composition due to the higher CLA levels increased the ratio of saturated: unsaturated fatty acids in intramuscular fat, which could improve the quality of meat, such as water holding capacity that increases when the CLA concentration is lower (Joo et al., 2002). The reduction of this fatty acid is observed in some cases, but has a negative impact on the nutritional value of meat .
Total lipids concentration tended to increase in longissimus muscle of pigs fed CLA. In this regard, the addition of 5 g kg -1 of CLA in substitution of sunflower oil does not change the content of intramuscular fat (Lauridsen et al., 2005), even with 20 g kg -1 of inclusion (Joo et al., 2002). However, in finishing pigs increased intramuscular fat (3.6 vs 2.6%) with 10 g kg -1 of CLA in the diet; but with 20 g kg -1 , the concentration was similar as in pigs fed sunflower oil (Martin et al., 2008a). With 12.5 g kg -1 of CLA increased the marbling score compared to CSO (Wiegand et al., 2002). The variability of results could be due to the concentration of CLA used in the diet and the length of treatment period (Sun et al., 2004). Furthermore, the concentration of intramuscular fat is reduced when CLA increases in the diet to 8.3 g kg -1 in growing-finishing pigs (Thiel-Cooper et al., 2001).
The fatty acid profile of meat in pigs fed CSO or SS was similar, probably due to the low level of addition. This indicates that the oil added was mainly used to meet the energy requirements of pigs, so, the accumulation of intramuscular fat derived from dietary oil was minimal.
The LPD in this experiment did not affect total lipids concentration in meat; this result did not agree with previous research Doran et al., 2006). The increased lipid accumulation occurs when pigs are fed LPD without meeting the requirement of essential amino acids (Kerr et al., 2003b). Also, when pigs are fed LPD, the concentration of ω-6 and ω-3 acids is reduced , as is the linolenic acid in longissimus muscle and the linolenic, eicosapentaenoic, docosapentaenoic, and docosahexaenoic acids in semimembranosus muscle. Another effect of LPD is the increase of saturated and monounsaturated fatty acids , and has been hypothesised that a lower concentration of CP could stimulate the expression of lipogenic enzymes in muscle, such as esteroil-CoA desaturase, which increases de novo synthesis of fatty acids (Doran et al., 2006). This premise is supported by the increase in the enzyme fatty acid synthase and the expression of acetyl-CoA carboxylase, involved in the biosynthesis of fatty acids, which increases in response to LPD (Doran et al., 2006). However, in this study, there was an increase of myristic, oleic, and eicoesaenoic acids, indicating that the addition of amino acids was adequate for pigs.
As conclusions, reduction of CP up to 160 g kg -1 for starter pigs adversely affects growth performance; however, 145 g kg -1 in growing pigs and 125 g kg -1 in finishing pigs are suitable for an acceptable productive performance and carcass characteristics, and reduces plasma urea nitrogen concentration. The addition of conjugated linoleic acid to standard diets or to lowprotein diets for fattening pigs does not improve growth performance or carcass characteristics, therefore, their incorporation in the diet could be based on its cost. Due to the cheaper cost of soybean soapstock, it can be used to replace crude soybean oil in diets without affecting the response of pigs.