Effect of the amount of chestnuts in the diet of Celta pigs on the fatty acid profile of dry-cured lacon

M.C. de Jesús, R. Domínguez, J. Cantalapiedra, A. Iglesias and J.M. Lorenzo* Departamento de Anatomía y Producción Animal. Universidad de Santiago de Compostela, 27002, Lugo, Spain Centro Tecnológico de la Carne de Galicia, Rúa Galicia N° 4, Parque Tecnológico de Galicia, San Cibrao das Viñas, 32900 Ourense, Spain Farm Counselling Services. Consellería do Medio Rural, Xunta de Galicia, 27004, Lugo, Spain Corresponding author: jmlorenzo@ceteca.net


INTRODUCTION
The chestnut (Castanea sativa Mill.) is a Geographically Protected Identity (G.P.I.) (Official Journal of the European Communities, 2010) in the NW of Spain, the main production area.In order to increase the quality of meat and meat products, it might be beneficial to recover some of the more traditional ingredients used in pig feeding such as chestnuts.In Galicia, the extensive rearing of pigs, similar to montanera in Iberian pigs, is difficult due to the small size of the farms; thus, feeding pigs with these products in large scale production can only be carried out by the use of concentrated diets (Díaz et al., 2009).
Chestnuts are characterized by a high moisture content (over 50%), high levels of starch (57 g•100 g −1 dry matter -d.m.), a low protein level (5.8 g•100g −1 d.m.) and low fat content (3 g•100 g −1 d.m.) (Pereira- Lorenzo et al., 2006).These fruits also have a significant content of polyphenols, with gallic acid and ellagic acid being predominant among the hydrolysable and condensed tannins (Gonçalves et al., 2010).In this sense, Barreira et al. (2008) evaluated the in vitro antioxidant activities of the extracts from chestnut flower, leaf, skin and fruit and found that chestnut skins had antioxidant properties, particularly against lipid peroxidation.Knowledge of the effect of providing chestnuts in controlled rearing conditions is scarce (Bermúdez et al., 2012;Coutron-Gambotti et al., 1998;Lorenzo et al., 2013;Lorenzo et al., 2014;Pugliese et al., 2013).
The Celta pig breed was the typical pig breed raised on farms in Galicia (NW of Spain) until the beginning of the 20 th century.However, with the arrival of improved breeds and their crosses, Celta pig production was displaced so that the Celta pig is included in the Official Catalogue of Cattle Breeds of Spain as being in danger of extinction (Real Decreto 2129/2008).The Celta carcasses, which are often very fatty, are used in the production of dry-cured meat products, which have a high value on the market and great quality (Lorenzo et al., 2012).There is a growing demand for products derived from autochthonous breeds.This has permitted an increment in the census of the Celta pig, increasing from 623 animals in 2001 to 4882 at the end of 2012 (Carril et al., 2013).
Dry-cured lacon is a traditional cured meat product made in the NW of Spain from the foreleg of the pig, which is cut at the shoulder blade-humerus joint, following a very similar manufacturing process as the one used in the production of dry-cured ham (Purriños et al., 2011).In the Galicia region, this product has been awarded a Geographically Protected Identity (Official Journal of the European Communities, 2001).
At the present time, chestnuts are underutilized and this situation contrasts with the high current prices of commercial concentrates for animal feed.
The use of chestnuts in the fattening feed, in an extensive management system, would allow for a reduction in production costs and put quality products on the market, differentiated by a high added value and with healthier fat (Bermúdez et al., 2012).In this regard, the smaller sized chestnuts and the by-products of this industry (pieces of chestnuts and chestnut meal) can be used in the elaboration of concentrated diets for pigs.So, the aim of this work was to study the effect of the inclusion of chestnuts in the pigs' diet on the fatty acid profile of Celta dry-cured lacon.

Experimental design and animal management
The experiment was conducted on 24 Celta pigs.Experimental farms were located in Lugo (Spain).The pigs were reared in an extensive system with a forest of 960 m 2 (40 m 2 per animal) composed of Eucalyptus spp.and Castanea sativa trees.Camping tents for rest and a pool for swimming were available in the area.The animals were fed a standard piglet diet for 3 months and then they were randomly divided into three groups.The first group was fed with a commercial diet as control (C diet) and it was composed basically of cereals (barley, soy and wheat), the second group was fed a diet containing 15% dried chestnuts (CH15) and the third group was fed a diet containing 25% dried chestnuts (CH25).The chemical and fatty acid compositions of the diets are shown in Table 1.The feed composition was determined according to AOAC (2005) procedures.The animals were feed ad libitum and slaughtered by electrical stunning and exsanguinations at a commercial abattoir (Castro Riberas de Lea, Lugo, Spain) when they were 12 months old.The carcasses attained an average weight of 103.58±5.86,118.66±6.97 and 105.50±8.12kg, P<0.01 for C, CH15 and CH25 groups, respectively.

Samples
After the refrigeration period (24 h at 4 °C), lacon pieces were extracted from the left side of the carcass.A total of 24 lacon pieces (8 if each group; 1 per animal) were used.The raw pieces were dry salted with an excess of coarse salt.A heap was formed consisting of alternating layers of lacon pieces and layers of salt.The pieces (4.68±0.32,5.40±0.61and 4.65±0.87kg, P<0.05 for C, CH15 and CH25 groups, respectively) were totally covered with salt and they remained in the heap for five days.The temperature of the salting room was in the range 2-5 °C and 80-90% relative humidity.After the salting stage, the pieces were taken from the heap, brushed, washed to remove salt from the surface, and transferred to a post-salting chamber where they stayed for 14 days at 2-5 °C and 85% relative humidity.After the post-salting stage the pieces were transferred to a room at 12 °C and 60% relative humidity where drying-ripening took place for 30 days.After this period, the samples were skinned, deboned, and the infraspinatus, triceps brachiilateralis and deltoideus muscles were extracted.The samples were minced, vacuum-packed and stored at −30 °C for no longer than four weeks until analysis.

Analysis of fatty acid methyl esters
Lipids were extracted from the muscles following the Bligh and Dyer (1959) method.Methyl esters of the samples were produced according to the method of Morrison and Smith (1964).Fifty milligrams of the extracted lipids were esterified with 4 mL of borontrifluoride (13%) in methanol.The sample was heated at 100 °C during 30 minutes and fatty acid methyl esters (FAMEs) were extracted with 2 mL of hexane and stored at −80 °C until chromatographic analysis.
Separation and quantification of FAMEs was carried out using a gas chromatograph, Carlo Erba Instruments MFC 500 auto/HRGC/MS (Milan, Italy) equipped with a flame ionization detector and using a Supelco DB-23 fused silica capillary column (30 m, 0.32 mm i.d., 0.25 μm film thickness, Supelco Inc., Bellafonte, PA, USA).Chromatographic conditions were as follows: initial oven temperature 170 °C (held for 2 min), first ramp at 3.5 °C•min −1 to 210 °C and second ramp at 2.5 °C•min −1 to 250 °C (held for 5 min).The injector and detector were maintained at 250 °C.Helium was used as carrier gas at a constant flow-rate of 2.0 mL•min −1 .The split ratio was 1:50 and 1 μL of solution was injected.Nonadecanoic acid methyl ester (C19:0 ME) at 0.3 mg•mL −1 was used as internal standard and added to the samples prior to fat extraction and methylation.Individual FAMEs were identified by comparing their retention times with those of authentic standards.Data regarding FAME composition were expressed as g•100 g −1 of total identified FAMEs.The proportion of saturated (SFA), monounsaturated (MUFA) and polyunsaturated (PUFA) fatty acid contents, PUFA/SFA, n6/n3, and the hypocholesterolemic/ Hypercholesterolemic ratio were calculated.

Statistical analysis
All statistical analyses were performed using IBM SPSS Statistics 19 software (IBM, Chicago, IL, USA).After verification of the normal distribution and constant variance of data, significant differences were determined using one-way analysis of variance (ANOVA) to analyze the effect of the inclusion of chestnuts in the diet on the fatty acid profile of Celta dry-cured lacón.The least squares means (LSM) were separated using Duncan's test.All statistical tests of LSM were performed for a significance level of P<0.05.Correlations between variables (P<0.05) were determined by a correlation analysis using Pearson's linear correlation coefficient.
To evaluate the relation between variables a factorial analysis of the fatty acids with significant differences (P<0.05)among the three groups was carried out.A principal component analysis (PCA) was used as extraction method and was performed on the correlation matrix.In order to verify the capacity of the fatty acid analysis as a tool for diet feed type discrimination in dry-cured lacon a stepwise discriminate analysis was done.An "a priori" equal probability for a sample to be in one group independently of the group size was considered.The criterion for the selection of variables was Wilk's lambda (F to enter and out value of 3.84 and 2.71, respectively).

RESULTS AND DISCUSSION
Table 2 shows the fatty acid composition (expressed as g/100 g of total fatty acids) of the intramuscular fat of lacon samples from pigs fed the C, CH15 and CH25 diets.Intramuscular fat did not show significant differences (P<0.05)among groups (19.80±0.74 vs. 21.26±0.68 vs. 18.66±0.80g•100 g −1 of dry matter for C, CH15 and CH25 feed, respectively) (data not shown).Significant (P<0.05)differences among treatments were observed regarding saturated fatty acid (SFA), monounsaturated fatty acid (MUFA) and polyunsaturated fatty acid (PUFA).Celta lacon (from CH15 and CH25 diets) presented significantly (P<0.001)lower percentages of SFA and significantly (P<0.001) higher proportions of PUFA than those from the control fed animals.This outcome is in agreement with those reported by Bermúdez et al. (2012) who found lower SFA contents in Celta hams from pigs fed with a diet of chestnuts only (5 kg/ animal a day) in the three months prior to slaughter.A similar result was observed by Pugliese et al. (2013) in fresh back fat from Cinta Senese pigs fed with chestnuts during the fattening period and by Domínguez et al. (2015) who also found the lowest SFA values in longissimus dorsi and psoas major fat in Celta pigs fed only with chestnuts during the finishing diet.
Statistical analysis showed that the inclusion of chestnuts in the diet increased (P<0.001)PUFA contents (16.84 vs. 24.24 vs. 24.44 g•100 g −1 from lacon for the C, CH15 and CH25 groups, respectively).An increase in omega 6 (15.22 vs. 20.39 vs. 19.28g•100 g −1 from lacon for the C, CH15 and CH25 groups, respectively; P<0.001) and omega 3 (1.05 vs. 2.11 vs. 1.79 g•100 g −1 from lacon for the C, CH15 and CH25 groups, respectively; P<0.001)The amount of PUFAs was appreciated as being increased due to chestnuts in the diet.This fact agrees with the results described by Coutron-Gambotti et al. (1998), who reported higher contents of PUFA in the biceps femoris muscle in pigs fed with chestnuts.These authors explained the higher PUFA contents in the muscles of pigs fed with chestnuts on the basis of differences in the mechanisms of the de novo synthesis of fatty acids from carbohydrates or of the desaturation and elongation of both endogenous and dietary fatty acids.In contrast, Temperan et al. (2014) did not find differences in PUFA contents between pigs fed with chestnuts and commercial feed while Bermúdez et al. (2012) and Domínguez et al. (2015) observed that the inclusion of chestnuts in the finishing diet decreased PUFA contents.Within omega 6 PUFAs only linoleic acid (C18:2n6) presented significant differences among groups.Lacon from the CH15 and CH25 groups showed higher (P<0.001)amounts of C18:2n6 (19.31 and 18.29 g•100 g −1 , respectively) than lacon from the C group (14.40 g•100 g −1 ).These findings agree with the results obtained by Pugliese et al. (2013).In contrast, Bermúdez et al. (2012) and Domínguez et al. (2015) found a decrease in this fatty acid as the amount of the chestnuts in the diet increased.Regarding omega 3 fatty acids (C18:3n3, C20:3n3, C20:5n3, C22:5n3 and C22:6n3), five out of six omega 3 PUFAs were affected by diet.In all cases, the increase in chestnuts in the diet also increases the amount of n3 PUFAs.α-Linolenic acid (C18:3n3) was the major n3 fatty acid.The Lacon from the CH15 group presented the highest (P<0.001)amounts of C18:3n3 (1.53 g•100 g −1 ), followed by the lacon from the CH25 group (1.12 g•100 g −1 ) and by the lacon from the C group (0.81 g•100 g −1 ).These results agree with those reported by Bermúdez et al. (2012) in ham and Domínguez et al. (2015) in subcutaneous fat, who found an increase in C18:3n3 when the chestnut was included in the diet.However, Domínguez et al. (2015) and Temperan et al. (2014) did not find differences in the amount of this fatty acid in intramuscular fat between chestnut and control diets.The C18:2n6 and C18:3n3 contents in the porcine tissues are directly related to the dietary contents of these two fatty acids because they cannot be synthesised in the tissues (Mitchaothai et al., 2007;Teye et al., 2006).Therefore, the higher contents of C18:2n6 and C18:3n3 in the lacon from animals fed with chestnut diets was due to the higher content of these fatty acids in the CH15 and CH25 diets.As mentioned above, also eicosatrienoic acid (C20:3n3), eicosapentaenoic acid (C20:5n3; EPA), docosapentaenoic acid (C22:5n3; DPA) and docosahexaenoic acid (C22:6n3; DHA) amounts increased as the chestnut was included in the diet.The C18:3n3 acid is the precursor of the synthesis of the other omega 3 PUFAs, although they can be incorporated from the diet.In the present research, the diets presented similar values of C20:3n3, EPA, DPA and DHA, therefore, the highest content of these fatty acids in the lacon from CH15 and CH25 is related to their greater synthesis from C18:3n3.
The higher conversion of the essential fatty acids into n-3 long-chain PUFAs than into n-6 PUFAs could be explained because they share the enzymes involved in their metabolism.These findings agree with the results obtained by Domínguez et al. (2015), Raes et al. (2004) and Rey et al. (2004), who found that there is competition between the C18:2n6 and C18:3n3 for incorporation into the lipid molecules as well as for the enzymes involved in the processes of elongation and desaturation.Moreover, in the present study there was a preferential elongation and desaturation of C18:3n3.Domínguez et al. (2015) also found a preferential incorporation of omega 3 PUFAs compared to omega 6 PUFAs into animal tissues.The high PUFA and lower SFA amounts observed in the lacon from chestnut groups indicate its suitability for healthier diets, due to human diets rich in PUFA, particularly n3 PUFA such as C18:3n3, EPA, DPA and DHA, which have been reported to have positive effects on health (Lopez-Huertas, 2010).
In order to establish which group presented the healthiest fat, the PUFA/SFA ratio (P/S) and the PUFA n6/n3 ratio were calculated.In relation to P/S, a value above 0.4 is recommended for healthy foods and diets (UK Department of Health, 1994) although, the high proportion of PUFA in and of itself is not necessarily healthy if it is not balanced in relation to the n6/n3 ratio.Simopoulos (2004) reported that high n6/n3 PUFA ratios promote several kinds of pathogenesis, including CVD, cancer and inflammatory and autoimmune diseases.In our work, the lacon from chestnut-fed animals had higher values (P<0.001) of P/S ratio (0.76 and 0.80 from CH15 and CH25, respectively) than animals from the C group (0.46).I n addition, the lacon from chestnut-fed animals also presented the best n6/n3 ratios (9.89 and 10.43 from CH15 and CH25, respectively) in comparison with those observed in the C group (13.25).Although CH15 and CH25 lacón showed lower values of n6/n3, ratios obtained in the three groups were higher than the maximum levels recommended by different authors and institutions, such as the Bri tish Nutrition Foundation (1992) (n6/ n3 ratio=6) or Simopoulos (2004) (n6/n3 ratio=4).
The ratio of hypocholesterolaemic/Hypercholesterolaemic (h/H) fatty acids is also an effective parameter for measuring the healthiness of a fat, based on the functional effects of fatty acids on cholesterol metabolism (Santos-Silva et al., 2002).The percentage of fatty acids considered as hypocholesterolaemic was higher in the CH15 (66.26 g•100 g −1 of fat) and CH25 (65.40 g•100 g −1 of fat) groups than in the C group (60.77g•100 g −1 of fat), while the amount of hypercholesterolaemic fatty acids showed the opposite behavior (25.69, 22.16 and 22.13 g•100 g −1 of fat from C, CH15 and CH25 groups, respectively) (data not shown).Consequently, the h/H ratio was significantly higher (P<0.01) in chestnut-fed lacon (3.00 and 2.95 from the CH15 and CH25 group, respectively) than in the control lacon (2.24).
Finally, the nutritive value (C18:0+C18:1n9)/ C16:0), which indicates the healthiness of the diet with regard to its lipid content, represents an important part of the fatty acids as a whole.Within the components of this ratio, C16:0 leads to increased blood cholesterol, while C18:0 does not affect cholesterol levels and C18:1n9 helps to decrease them.In the present study, a statistical analysis did not show differences among the three groups.Although there were no significant differences, slightly higher values were obtained for lacon from chestnut-fed pigs (2.45 and 2.38 from CH15 and CH25 groups, respectively) than for the control lacon (2.23).
As a result, from a health point of view, the P/S, n6/n3 and h/H ratios revealed that the lacon obtained from animals fed with chestnuts (with 15 and 25% inclusion of chestnut in the diet) was healthier than lacon from animals fed a commercial feed.

Factor analysis using principal component
The objective of this factor analysis (FtA) was to obtain a reduced number of principal components which would explain the variability of the selected compounds.When this data set was used (20 parameters), the first four principal components were chosen (94.16% of the total variance) because the eigenvalues were higher than 1, and therefore, they explain a higher percentage of variance than each original variable.All variables presented communality higher than 0.824, which indicates that they were well represented by the four factors.A varimax rotation was carried out to minimize the number of variables that influence each fatty acid, and then, to facilitate the interpretation of the results.The first principal component (PC1) that explained the higher percentage of variance (28.29%) was positively associated with C17:1n7, C18:2n6, C18:3n3, PUFA, Σn3, Σn6 and P/S ratio and negatively correlated with C18:0 (Figure 1).The PC2 that explained the 26.60% of the total variance was positively related to the h/H ratio and nutritive value and negatively linked to C16:0, SFA and the Σn6/Σn3 ratio.The PC3 explained 25.47% of the total variance and was positively correlated to C15:0, C20:3n3, FIGURE 1. Projection of the variables in the rotated space defined by the first three principle components.
Figure 1 shows the projection of the variables in the rotated space defined by the first three principle components.

Discriminant analysis
Using a new matrix of data integrated by the standardized reduced original variables, which were selected from the previous FtA, a canonical discriminant analysis (CDA) was developed (Figure 2).This data set was subjected to the CDA according to the diet (C, CH15 and CH25).The leave-one-out cross validation was used to validate the results.Table 3 shows the classification parameters such as number of eigenvalue and canonical correlation.These classification results revealed a correct aggregation for the first two discriminant factors defined.From the data set (20 parameters chosen after FtA) subjected to discriminant analysis, 4 variables (C17:1n7, C20:3n3, C20:5n3 and C22:5n3) were retained at the end of the stepwise discriminant analysis and were linearly combined to form canonical discriminant functions.When chestnut was presented in the diet in different proportions (CH15 and CH25 groups) it was still possible to discriminate among groups.
When results obtained from function discriminant functions of classification (CAN1) were plotted against results obtained from function CAN2 on axes of coordinates for each meat sample, a good discrimination among groups according to the diet was observed (see Fig 2).The first two CAN accounted together for 100% of the total variance, with the first CAN (CAN1) explaining 85.1% of total variability and the second CAN (CAN2) explaining the remaining 14.9% of total variability.The CAN1 variable allowed segregation of the C samples from the CH15 and CH25 ones, while the CAN2 allowed segregation of the CH15 from the CH25 lacon (Fig. 2).After cross-validation, the discriminant analysis correctly attributed each lacon to its original group with an accuracy of 87.5% for the C group and of 87.5% and 75% for CH15 and CH25 groups, respectively.Effect of the amount of chestnuts in the diet of Celta pigs on the fatty acid profile of dry-cured lacon • 9 Grasas Aceites 67 (1), January-March 2016, e119.ISSN-L: 0017-3495 doi: http://dx.doi.org/10.3989/gya.0492151 According to these coefficients, the parameters which mostly accounted for the segregation of CAN1 were C20:3n3 and C22:5n3, while the variables accounting for group segregation of CAN2 were C17:1n7 and C20:5n3.As reported in Table 2, C20:3n3 and C22:5n3 contents were lower in C lacon samples compared to CH15 and CH25, while C20:5n3 percentages were higher in CH25 lacon samples compared to CH15 ones.

CONCLUSIONS
The inclusion of chestnuts in the formulation of feed decreases the SFA and MUFA contents in the intramuscular fat of lacon, while the total PUFA content increases (in particular n3 PUFAs).According to P/S, n6/n3 and h/H ratios lacon from chestnut-fed animals was healthier than those obtained from pigs fed with commercial feed.In this regard, lacon from CH15 and CH25 also had higher contents of some specific fatty acids related to health benefits (such as C18:3n3, EPA, DPA or DHA).The discriminant analysis selected 4 variables (C17:1n7, C20:3n3, C20:5n3 and C22:5n3) and calculated two discriminating functions which allowed for prediction of the amount of chestnut in the diets.When chestnut was presented in the finishing diet (CH15 and CH25 groups) it was still possible to discriminate between groups.Therefore, the use of chestnuts in the diet (at both levels: 15% and 25% of chestnuts) allow us to obtain healthier dry cured products (in this study, lacon).

FIGURE 2 .
FIGURE 2. Score plot on the plane of two CAN discriminant functions obtained to predict the type of diet.

TABLE 2 .
Effect of diet on fatty acid contents (mean±standard error) of intramuscular fat of dry-cured "lacon" from Celta pig breed

TABLE 3 .
CAN classification parameters and results according to diet(C, CH15 and CH25)

Data set (20 parameters chosen after FtA) Eigenvalues Function Eigenvalue % of Variance Cumulative % Canonical correlation
Cross validation is done only for those cases in the analysis.In cross validation, each case is classified by the functions derived from all cases other than that case; b 83.3% cross-validated grouped cases correctly classified.