Effects of genotype and salting time on chemical, physical and sensorial traits of a new pig seasoned meat product ‘Cuore Di Spalla’

Abstract The aim of this work was to determine the chemical, physical and sensorial characteristics of a cured product called ‘Cuore di Spalla’, resulting from two different pig genotypes (G) and two salting times (S). Fourteen Duroc × Large White (D × LW) and 14 Cinta Senese × Large White (CS × LW), reared in the same condition in outdoor and fed commercial mixture. The animals were slaughtered at 10 months of age which corresponds to an average live weight of 143 ± 15 kg for D × LW and 122 ± 15 kg for CS × LW. The right shoulders were salted for 3 days (L) and the left ones for 5 days (H). Physical (colour, texture profile analysis), chemical (moisture; crude protein; IMF; ash; salt; total lipids; fatty acids) and sensorial analyses (by panellists’ group) as well as volatile compounds profile (by SPME-GC-MS technique) were determined at the end of seasoning period. The genotype affected product’s size (1484.3 vs 1857.1 g of final weight for CS × LW and D × LW respectively), chemical characteristics and texture (cohesiveness 0.50 vs 0.53 and springiness 6.73 vs 7.12 for CS × LW and D × LW respectively). Aromatic profile was less affected by the factors considered or at least the panellists were not able to discriminate the effect. The main family of volatile compounds affected by genotype was alcohols. Salting time seemed to affect only the parameters closely related to salt (16.52 vs 15.33% of salt and 5.15 vs 4.35 for saltiness in H vs L respectively). Highlights The genotypes determined differences in size, fat quantity and physical-chemical parameters of the ‘Cuore di Spalla’ product. Very limited differences in the aromatic profile were found. Both salting times led to high salt content in the product. It seems feasible to produce the ‘Cuore di Spalla’ with further reduction of salting times.


Introduction
Local pigs are generally linked to products characterised by high specificity. Various researches pinpointed the differences between products coming from autochthonous pigs compared to improved ones (Franci et al. 2007;Sirtori et al. 2011;Pugliese and Sirtori 2012) and the genetic effect on meat quality is recognised. Mediterranean local pigs are often crossed with selected breeds, as in the production of Duroc Â Iberian (Fuentes et al. 2014) and Large White Â Corsican pigs (Coutron-Gambotti et al. 1998). In Italy, historically, crosses between Cinta Senese sows and Large White boars have been produced by Cinta Senese farmers (Pugliese and Sirtori 2012).
The crossbreeding counterbalance some limits of autochthonous pigs both on in vita performances and on carcase and meat traits (Sirtori et al. 2011).
The use of local breeds to produce dry-cured meats must take into account the qualitative characteristics of the raw material. The greater adipogenicity of the local breeds influences the final characteristics of the products (Sirtori et al. 2011) as well as the production process, including the diffusion of salt.
The dry-cured products produced with the autochthonous breeds are often characterised by high salt content (Hersleth et al. 2011); the highest moisture and the lowest size characterising the products of these breeds favour the penetration of salt. In addition, especially in southern Europe, the traditional seasoning methods are characterised by excessive use of salt (Pugliese and Sirtori 2012), generating urgent worries for human health. Recently, European Union (EU) was forced to implement salt reduction initiatives because of its negative effect on hypertension (Corral et al. 2013). However, salt cannot be reduced without considering its influence on the sensorial and technologic characteristics (Corral et al. 2014); like the development of an optimal texture through proteolysis (Purriños et al. 2012;Harkouss et al. 2015) and the formation of flavours and aromas due to its lipids prooxidant role during curing (Andr es et al. 2002;Purriños et al. 2012). In dry-cured products the penetration of salt is pivotal to ensure product's safety and stability, as well as to develop characteristic sensory attributes (Martuscelli et al. 2017). The NaCl contributes to control the endogenous enzymatic activity and is responsible for the salty taste and typical texture of dry-cured ham (Flores et al. 2012). The curing process consists of a stabilisation phase at low temperature, that includes the curing and resting steps and a further phase of drying-ageing at increasing temperatures. The main goal of the stabilisation phase is to reduce water activity (a w ), thus increasing salt and decreasing water contents (Garcia-Gil et al. 2012). The amount of salt used can be undetermined or exact. The first procedure is mainly used in Spain and part of France and Italy and the product is completely covered by salt for several days. In the second way the exact amount of salt per kg of fresh ham is added on the lean surface and hand-rubbed (Toldr a 2002). Other factors can affect the penetration of the salt in meat; primarily the characteristics of the raw product such as size, moisture, fat, as well as if the product is whole or partitioned (S anchez-Molinero and Arnau 2008; Gou et al. 2008). These factors affect the diffusion process, where the salt absorption is parallel to a water loss caused by differences in concentration and osmotic pressures (Raoult-Wack 1994).
The objective of this study was to evaluate the influence of two genotypes and two different salting times on technological, physical and chemical traits as well as on the volatile compounds (VOCs) profile of a dry-cured pig product.

Samples manufacturing
Twenty-eight castrated male pigs, 14 belonging to Duroc Â Large White (D Â LW) and 14 to Cinta Senese Â Large White (CS Â LW) genotypes, were reared in the same conditions. The animals were slaughtered at the same age to an average live weight of 143 ± 15 kg for D Â LW and 122 ± 15 kg for CS Â LW. From the boneless shoulder, including all muscles and removing subcutaneous fat, the product called 'Cuore di Spalla' was created. The right shoulder of each pig was salted for 3 days (Low salt: L) while the left shoulder for 5 days (High salt: H). The dry-cured period was carried out in a controlled cell for humidity and temperature. After salting, the shoulders were washed, stuffed inside a synthetic casing and rested in a chamber under controlled atmosphere (at 4-6 C and 70-85% relative humidity) for 4 weeks. Subsequently, the shoulders were dried at 14-20 C and 60-80% RH for 5 months, approximately. The authors declare that the animals were bred in 'Le Selve di Vallolmo' farm, following the principles of the declaration of Helsinki and the experiments complied laboratory health and safety procedures.

Colour and chemical analysis
At the end of seasoning time, colour parameters CIE L Ã (lightness), a (redness) and b (yellowness) were measured on a slice of each product, repeated four times, using a Minolta colourimeter CR-200 (Minolta Camera Co., Ltd, Osaka, Japan). Recalibration on white and red plates was performed at the start of each measuring session.
Chemical analyses were performed according to AOAC methods (AOAC 2000) on samples deriving from the whole ground product: (i) moisture by lyophilising to a constant weight; (ii) intramuscular fat (IMF) as ether extract; and (iii) protein using the Kjeldahl method.

Texture profile analysis (TPA)
Texture Profile Analysis (TPA) were carried out on 20mm-thick lean slices cut and accurately carved with a scalpel into 3 cubes of 10 mm per side (dimensions of 10 mm Â 10 mm Â 10 mm for length, width and height, respectively) for each sample. TPA was performed using a texture analyser (Zwick GmbH & Co. KG, Ulm, Germany) with a 1 kN-load cell and a 100-mm-diameter compression plate (modification of the method described by Morales et al. 2007). The samples were compressed twice to 50% of their original height (time ¼ 0 s between the two compression cycles), at a crosshead speed of 1 mm/s and perpendicularly to the fibre-bundle direction. Force-time curves were recorded, and the following parameters were calculated, according to Ruiz-Ramirez et al. (2005): hardness (N), adhesiveness (J), cohesiveness (dimensionless), springiness (mm) and chewiness (J).

Fatty acid composition
Total lipids content was determined using the method of Folch et al. (1957); fatty acid profile of total lipids, using the modified technique of Morrison and Smith (1964). Fatty acids (FAs) methyl esters were analysed by gas chromatography using a Varian 430 apparatus (Varian Inc., Palo Alto, CA, USA) equipped with a flame ionisation detector. FAs separation occurred in a Supelco Omegawax TM 320 capillary column (30-m length; 0.32 mm internal diameter; 0.25 lm film thickness; Supelco, Bellafonte, PA, USA). The chromatographic conditions were an initial temperature of 160 C, which was then increased by 2 C/min until the temperature reached 220 C. One microliter of sample in hexane was injected with the carrier gas (helium) at a constant flow of 1.5 mL min-1 and at a split ratio of 1:20. The detector temperature was set at 260 C. The chromatograms were recorded using computing integrator software (Galaxie Chromatography Data System 1.9.302.952; Varian Inc.). The FAs were identified by comparing the retention time of the FAME with the standard Supelco 37-component FAME mix (Supelco) and were quantified through calibration curves using nonadecanoic acid (C19:0) (Supelco) as an internal standard. Results were expressed as grams of FAs per 100 g of dry product.

Analysis of volatile compounds
The VOCs profile was obtained by SPME-GC-MS technique. An Agilent 7890 GC-chromatograph equipped with a 5975A MSD with EI ionisation was used for analysis. A three-phase DVB/Carboxen/PDMS 75-lm SPME fibre, (Supelco, Bellafonte, PA, USA) was exposed in the head space of the vials at 60 C for 30 min for volatile compound sampling after 5-min equilibration time. A Gerstel MPS2 XL autosampler equipped with magnetic transportation adapter and a temperature controlled agitator (250 rpm with on/cycles of 10 s) was used for ensuring consistent SPME extraction conditions. Chromatographic conditions were column J&W Innovax 30 m, 0.25 mm, ID 0.5 lm DF; injection temperature 250 C, splitless mode, oven programme 40 for 1 min then 2 C/min to 60 C, then 3 C/min to 150 C, then 10 C/min to 200 C, then 25 C/min to 260 C for 6.6 min. Mass spectra were acquired within the M/Z interval 29-350 with an Agilent 5975C MSD spectrometer at a scan speed such to obtain three scans/s. After acquisition, the GC-MS chromatogram was processed by a CLEAR VIEW tm (Markes, UK) algorithm to dynamically subtract the background noise contribution along the chromatograms, defining an average chromatographic peak width of 4 s. VOCs were identified by matching EI mass spectra against NIST 05 or Wiley 07 spectral library and Kovats indices. The peak area of each compound was determined on a specific target ion and confirmed by the matching of two qualifier ions using a maximum threshold ratio of 20%. The peak areas for each analyte were normalised for the appropriate internal standard. A detailed description of the method can be found in Pugliese et al. (2010).

Sensory analysis
Two slices, 2 mm thick, for each cured shoulder were evaluated by a trained panel of 10 members using a descriptive analysis method: 12 traits regarding sensory characteristics, grouped in appearance (visual intramuscular fat, colour intensity and oiliness), texture (hardness), aroma (aroma and after-flavour) and taste (saltiness, after-taste, rancid, chewiness, juiciness and persistence) were evaluated. A 10-cm unstructured scale was used, whose extremes were 'very low' and 'very high'. All sessions were done at 20-24 C in a panel room equipped with fluorescent lighting. Four products were evaluated simultaneously assessing the four treatment combinations (H-CS Â LW; L-CS Â LW; H-D Â LW; L-D Â LW) per session with a total of 14 sessions. Samples order was randomised.

Statistical analysis
Data were analysed using the GLM procedure (SAS 2012) using genotype, salting time and subject as discrete effects: For sensorial data the following model was used: The genotype effect was tested against subject within genotype in both statistical analyses. The residual mean square was used as the error term for the effect of salting time.

Results
Size, colour, chemical traits and texture profile analysis (TPA) Genotype strongly influenced almost all parameters (Table 1). D Â LW products were heavier both at the beginning and at the end of seasoning (p < .01), while seasoning loss (p > .05) was not affected. As regards chemical traits, CS Â LW showed the highest moisture (p < .05) while no difference was recorded in crude protein percentage. D Â LW pigs recorded higher percentage of IMF (p < .01) and lower of Ash and Salt (p < .01) than CS Â LW. Colour parameters were higher in D Â LW for all the three coordinates.
The salting time seems to have affected only salt content, being higher in the H group, and consequently also ash (p < .05). This result mainly concerns the D Â LW genotype within which the difference between H and L was statistically significant (14.8 vs 11.9%).
TPA analysis (Table 2) showed higher values of Cohesiveness (p < .05) and Springiness (p < .01) in D Â LW than CS Â LW. Salting Time did not affect any parameters.

Fatty acid composition
The results of FAs profile are shown in Table 3. D Â LW product showed the highest quantity of total   lipids (p < .01). This result was linked to the whole amount of FAs, higher in the D Â LW group, irrespective if they were saturated (SFAs), monounsaturated (MUFAs) or polyunsaturated fatty acids (PUFAs) (p < .01). As regard Salting Time effect, no significant differences were recorded in FAs profile (p > .05).

Effect of genotype
The differences in shoulder weight for different genotypes are both a consequence of differences in the carcase weight (Franci et al. 2005) and of raw commercial cuts proportion (Franci et al. 2003). At the same age, the improved breeds showed a higher in vivo performance, consequently, they achieved a greater slaughter weight in lesser time (Acciaioli et al. 2002;Sirtori et al. 2011;Pugliese and Sirtori 2012). In this work, the higher slaughter weight of D Â LW compared to CS Â LW (143 vs 122 kg respectively) affected the products' size.
It is well known the importance of genetic effect also on physical and chemical characteristics of meat (Bonneau and Lebret 2010;Pugliese and Sirtori 2012); in the present research, being LW the maternal basis, the influence of the two paternal breeds was high. The increasing contribution of Duroc on IMF content, reported by other authors (Lebret et al. 2011;Sirtori et al. 2011), was confirmed also in comparison with a local breed such as Cinta Senese. IMF, together with moisture and product size, greatly affect final product's salt content, with salt diffusion being positively correlated with water and negatively with IMF ( Skrlep et al. 2016). The IMF increase creates barriers to penetration (Martuscelli et al. 2017), as well as the increase in the product's size causes an extension of the diffusion time. The present work confirmed these influences, since higher values of salt were observed in CS Â LW products, which were also characterised by lower IMF, higher moisture and smaller size than D Â LW. Colour is one of the most important sensorial characteristics. Pork meat colour is influenced by many intrinsic (e.g. breed, gender, age, type of muscle) and extrinsic (e.g. feeding, pre-slaughter handling, slaughtering) factors (Tikk et al. 2008). In this research the higher values recorded for Duroc crossbreeds may be attributable to the paternal breed contribution probably for higher slaughter weight and IMF content. Alonso et al. (2015) reported that an increase in the percentage of Duroc genes affected colour parameters, in particular a value; the same authors reported also that Duroc and LW breeds seem to have higher myoglobin content than other breeds. The slaughter weight increase can lead to higher a and b values because of an increase in myoglobin and IMF content, respectively (Garc ıa-Mac ıas et al. 1996; Latorre et al. 2004;Gali an et al. 2009). IMF values also seem to affect L Ã parameter; Suzuki et al. (2005), on Duroc MS: comparison with corresponding mass spectra in NIST05 or Wyley 7 database; KI: matching with reported Kovats Indices. Peak areas for each analyte were normalised for the appropriate internal standard. The value of each compound is the response ratio to the internal standard to which it is associated.
pigs, found that meat having high IMF resulted lighter; similarly, Gjerlaug-Enger et al. (2010), found positive correlation between L Ã and IMF likely due to the visible fat cells. Another important parameter for the consumers is meat texture, which depends on raw material (Parolari et al. 1988;Morales et al. 2007;Zochowska-Kujawska 2016). Indeed, the differences between genotypes for raw material characteristics, such as IMF and moisture, could have affected the TPA.
As aforementioned, the inclusion of Duroc breed leads to an increase in IMF ( Barton-Gade 1987;Suzuki et al. 2003;Alonso et al. 2015) and, consequently, to an increase of the total lipids content with a possible influence also on fatty acids profile. The higher IMF value found in D Â LW could also be due to the different growth rate ( Candek-Potokar et al. 1998) and precocity (Franco et al. 2014) compared to CS Â LW. Indeed, at the same age, the slaughter weight was higher in D Â LW than in CS Â LW (Correa et al. 2006;Gali an et al. 2009).
Proteolytic and lipolytic enzymes play an important role in the formation of VOCs. VOCs can result from chemical or enzymatic oxidation of unsaturated FAs and from Strecker or Maillard reactions (Toldr a 1998;Toldr a and Flores 1998). Many authors reported aldehydes as the main family in dry-cured products and as the major contributors of overall aroma, due to their low thresholds (Huan et al. 2005;Ram ırez and Cava 2007;Maru sic et al. 2014;Lorenzo and Carballo 2015). However, in this study aldehydes were not the most abundant family. This result could be attributable to several factors, with the low level of oxidation present in the products probably being the main one. Indeed, the present products had a lower maturing time respect to ones reported in other studies (Sunesen et al. 2001;Huan et al. 2005;Ram ırez and Cava 2007;Maru sic et al 2014;Narv aez-Rivas et al. 2014); a second reason could be the low content of PUFAs from which many aldehydes derive (Ram ırez and Cava 2007;Laranjo et al. 2015). The low hexanal value recorded confirms this hypothesis, being Hexanal one of the major oxidation products (Maru sic et al. 2014). This compound derives from oxidation of n-6 FAs such linoleic and arachidonic acids (Ram ırez and Cava 2007). In this case, Hexanal value suggests a low oxidation for both the genotypes (CS Â LW and D Â LW). Lastly, considering the great amount of alcohol observed, the lower level of aldehydes can be explained with their reduction to corresponding alcohols by alcohol dehydrogenase (Sunesen et al. 2001). The genotype effect was found only for Benzaldehyde, being more abundant in D Â LW products, which had also a higher amount of C18:3 content. This confirms that the origin of Benzaldehyde is mainly attributable to degradation of alpha-linolenic acid (Zhao et al. 2017).
Acids can be produced from hydrolysis, degradation of triglycerides and phospholipids, deamination of amino acids, microbial activity (Huan et al. 2005) and aldehydes oxidation ( Skrlep et al. 2016). Long chain acids do not directly affect the flavour of cured products due to their high olfaction thresholds (Muriel et al. 2004). Contrariwise, short chain acids (<6 carbon atoms) have an important role on aroma due to their odour as vinegar, cheese or cucumber (Stahnke 1995) together to low threshold values. CS Â LW showed lower values for Butanoic, Octanoic and n-Decanoic acids and higher value of Hexanoic acid-2-ethyl respect to D Â LW. However, the low capacity of long  Muriel et al. (2004) who indicated alcohols as the most important compounds in Iberian dry-cured loin. The formation of alcohols can be attributed to aldehydes degradation, microbial fermentation for branched alcohols and PUFAs oxidation for straight chain ones (Muriel et al. 2004). Alcohols' flavour becomes stronger as their carbon chain increases (Garc ıa-Gonz alez et al. 2008). In the present study, only few compounds showed difference between the genotypes, although the total alcohols content was almost double in D Â LW.
The esters were formed from the esterification of various acids and alcohols. The contribution of esters to flavour formation is very important, having low odour thresholds (Lorenzo and Carballo 2015) and typically aged meat aroma (Huan et al. 2005) and fruity odour (Zhao et al. 2017). However, genotype appeared to have not affected this group of compounds, being the sporadic differences in individual esters inconsistent.
Ketones, as well as esters, were not affected by genotype and thus, no discrimination is expected on the perception. Nevertheless, Ketones, especially 2ketones are considered to have a great influence on the aroma of meat having an intense odour (Muriel et al. 2004;Maru sic et al. 2014) and a high concentration of ketones is the signal of low product quality (Pastorelli et al. 2003).
No noticeable differences between the genotypes were found for Thiols or Sulphur compounds, as well as for Terpenes and Alkanes. Contrariwise, Furans, that are considered important for desirable aroma due to their low threshold values (Ram ırez and Cava 2007), were more present in D Â LW. This could be attributed to the higher amount of C18:2 as the formation of these compounds starts from linoleic acid (Muriel et al. 2004).
The different aromatic profile of the two genotypes didn't influence the panellists perception, as olfactory and taste parameters did not show any differences, except for saltiness, confirming the higher salt content in CS Â LW recorded with chemical analysis. The genotype effect in the aromatic profile were minimal, especially for compounds with low perceptive threshold values as aldehydes (Pugliese et al. 2010).
As for the other sensorial parameters, colour and IMF seem to contradict the results of the instrumental determination. Carrapiso et al. (2003) concluded that some instrumental measurements are often not particularly useful to assess differences perceived by panellists.
The higher values of juiciness and persistence in CS Â LW than D Â LW are probably attributable to the higher salt content that could accentuate these perceptions in panellists, as reported by Lorido et al. (2015) and Lorido et al. (2016).

Effect of salting time
Generally, the effect of salting time was very low. The high amount of salt found in both treatments (above 8% on wet basis), highlighted the possibility of further reductions to meet the requirements demanded by consumers and EU. Purriños et al. (2012) in a trial with similar salting times on pig shoulder, recorded a percentage of salt content in agreement to the present values. Salting time did not influence physical-chemical parameters (i.e. crude protein, IMF, colour, FAs profile) and the seasoning loss agreeing with other studies (Lorido et al. 2015;Skrlep et al. 2016). Salt content did not significantly affect texture parameters too (even if H group shows tendentially higher values), contrarily to other researches where higher salt percentages increased hardness values (Lorido et al. 2015;Skrlep et al. 2016). The changes in texture caused by increasing salt content are correlated to dehydration, that causes meat hardening (Ruiz-Ramirez et al. 2005) and to proteolysis inhibition (Toldr a et al. 1997;Martin et al. 1998). In this research, the same seasoning loss for both salting times was recorded, so, the lack of significant differences is likely due to the high salt percentage in both products and to the reduced gap between H and L groups for these parameters.
The aromatic profile revealed few differences between salting times. Almost all the compounds showed higher values in H than in L group. This result could be attributed to the salt pro-oxidizing and solubilising role, as reported by Purriños et al. (2012), which found higher amounts of compounds derived from lipid oxidation in products with 5 days of salting time when compared to 3 days of salting. The same authors recorded a higher amount of alcohols in the product with 5 day of salting time; it was likely due to the decrease of the water molecules available for alcohols solubilisation. Even in this work, the amount of alcohols was higher in H than L level. Despite the differences found in some important aroma-active compounds (Benzaldehyde, 1-Octen-3-ol) (Garc ıa-Gonz alez Maru sic et al. 2014), the sensory analysis confirmed the absence of substantial differences between the two salting times.
Panellists only perceived a change in saltiness between the two product types, as reported in other works (Corral et al. 2013;Skrlep et al. 2016). However, these authors also reported other differences, probably due to the different products tested or to different seasoning times.

Effect of interaction G Â S
The interaction effect was minimal. Only D Â LW genotype showed differences between salting times as regards the salt content and the saltiness parameter of the panel test. Probably the effects of greater size, higher fat content and lower value of moisture in D Â LW respect to CS Â LW caused a differentiation between H and L products. In fact, as already mentioned, the salt diffusion is influenced by size, fat and moisture content of the product. The lower size, lower fat content and higher moisture in CS Â LW allowed also to have a full penetration of salt in L level and then to achieve the same results than H.

Conclusions
The study highlighted the genotype effect, probably linked to the intrinsic characteristics of the two paternal breeds (Duroc and Cinta Senese) that have led to differences in size, fat quantity and physical-chemical parameters of the product. The aromatic profile seems to be only slightly affected, with no or unperceivable differences in the sensorial analysis.
The two salting times led to high salt content in both products, even higher than foreseen. Thus, it seems feasible to produce the 'Cuore di Spalla' with further reduction of salting times with positive influence on health aspects providing that safety issues are guaranteed. Saltiness was the only parameters perceived by panellists and no other substantial differences were found for aromatic profile.