Inclusion of olive by-products in growing diets causes minor effects on meat quality of Iberian pigs fattened in a traditional system

INTRODUCTION

The traditional production system of the Iberian pig includes a finish-fattening period (up to approximately 160 kg with a minimum length of 60 days) known as Montanera, based on an ad libitum intake of acorns and pastures (López-Bote, 1998). The usual preparation of the animals for this final fattening system includes reaching a weight between 92 and 115 kg at 12 months of age, at the end of the growth period before Montanera. These requirements were incorporated into the current Spanish Quality Standard for Iberian Products (BOE, 2014). Therefore, a strong restricted diet is required to slow down the growth rate and prevent the animals for an excessive fatness. However, this restriction causes undesirable feed stress with a consequent negative effect on animal welfare. One of the possible alternatives to minimize this problem is the use of low-energy diets and high fiber supplements that increase the satiety feeling and improve animal welfare.

Indigenous breeds and the use of local feed resources are closely linked issues in integrated and sustainable agriculture and livestock (Rodríguez & Preston, 1997). By-products derived from olive oil production have been proposed for animal feed in the Mediterranean areas (Molina-Alcaide & Nefzaoui, 1996), since they meet the necessary energy and fiber requirements for the animals and they are cheap. Spain is the world’s largest producer of olive oil, making these by-products easy to obtain (Davidson, 2014). In addition to this, it should be noted that the use of agricultural residues would be in accordance with sustainability principles, leading to a reduction in the costs of agricultural and livestock production (De Miguel et al., 2015).

In Iberian pig production, olive-based by-products have been proposed as raw material in feed components for growing and finishing pigs (Benito et al., 1998; Hernandez-Matamoros et al., 2011; Joven et al., 2014; González-Sánchez et al., 2016). According to these studies, pigs fed with olive by-products did not suffer negative growth rate effects or lower final slaughter weight, but they did show a positive effect on the backfat fatty acid profile, with an increase of the percentage of oleic acid and a decrease of the saturated fatty acids. However, to our knowledge, no study has been done on the usefulness of these by-products for the traditional period of feed restriction prior to the Montanera practiced in finishing Iberian pigs. Likewise, it has not been studied whether this type of diet can affect other parameters related to the quality of the meat, such as color, water holding capacity or texture.

The purpose of this study was to explore the suitability of using by-products from olive oil industry supplied in the diet of Iberian Montanera pigs during the growth period by means of two different diets including olive cake. The effects on the growth rate, carcass composition and meat quality traits have been studied.

MATERIAL AND METHODS

Olive by-products description

The by-products of the olive were obtained in several steps. First, the olive oil was extracted, and the remainders of the process was the wet crude olive cake, a semi-liquid paste made of olive pulp, skin, stone and 71% water. After that, the stone was removed, and a second oil extraction was carried out. The resulting pastry was partially dehydrated obtaining the dry olive pulp, which contains skin, pieces of olive stones and a small proportion of olive oil (Alburquerque et al., 2004; Molina-Alcaide & Yáñez-Ruiz, 2008). Olive pulp and crude olive cake were the two by-products used in the formulation of experimental diets of this study, whose analytical composition is showed in Table 1.

Animals, diets, and experimental design

Animal manipulations were performed according to the Spanish Policy for Animal Protection RD 53/2013, which meets the European Union Directive 2010/63/EU about the protection of animals used in experimentation. The experimental protocol was approved by the Bioethical Committee of the Extremadura region.

A total of 44 Iberian pigs were included in the experiment, being 29 and 15 castrated males and females, respectively. The amount of daily feed supplied, health status, growth and fatness were controlled from post-weaning to slaughter. These animals were randomly allocated in three different pens (one group per pen), at 6.5 months of age and 42 kg (SD 8.6 kg) body weight (BW). Two pens housed 15 pigs and the other one, 14 pigs. All the pens had open and covered places, and the total area of each pen was 110 m2 that were adjacent to each other, with the same water availability and spatial orientation. The animals were separated in three groups according to dietary treatment during the growing period: control (C, n=15, 9 males and 6 females), dry olive pulp (DOP, n=14, 9 males and 5 females) and wet crude olive cake (WCOC, n=15, 11 males and 4 females) group, respectively. In Figure 1, a description of by-products, diets and groups is presented. Control group (C) was fed a compound feed (C_CF) formulated to cover the protein and energy requirements of the growth period. The DOP group was fed a compound feed (DOP_CF) containing a 45% of olive pulp by-product. Finally, the WCOC group was fed crude olive cake by-product in silage form to facilitate its conservation, and a specific compound feed (WCOC_CF) as a complement. As the crude olive cake is a semi-solid by-product, it had to be mixed with a solid support in order to transform it into a suitable raw material that can be administered to pigs. Therefore, a mixture with 75% of crude olive cake and 25% of barley straw was made for preparing the silage. Thereafter, it was packed under high pressure resulting in a 42.5% dry matter mix. The DOP_CF, WCOC_CF and silage crude olive cake products were not commercially available, and they were made specifically for this experiment. The analytical compositions of the feed used in each dietary treatment are shown in Table 1. The corresponding ingredients of the different compound feeds supplied to each group are described in Table 2. All compound feeds were provided as pellets.

The animals were fed these diets for 191 days, reaching 95 kg (SD 13.7 kg) of BW at the end of the growth period with 12 months age. Daily weight gain was controlled every three weeks by monitoring individual weight and the food provided was adjusted according to BW. With this objective, the compound feeds were administered in a restricted way (60 and 70% ad libitum for WCOC_CF and C_CF, respectively), while the silage crude olive cake by-product was supplied ad libitum. DOP_CF was supplied ad libitum during the first days, however, the excessive intake of the DOP_CF concentrate, together with a high consumption of water by the animals advised a supplementation with a restricted regimen (60% ad libitum) as well.

Following the traditional livestock uses to take advantage of natural resources, at the end of the growth period all pigs were rearranged into three new batches and distributed in three fences (batches) with oak trees according to their BW and the availability of pasture and acorn on each fence. In the first batch (Mont1), the animals were fed exclusively on acorn and grass. In the second one (Mont2), in addition to the pastures and the scarce acorn available in that fence, a daily high-oleic compound feed was supplied (2 kg per day and pig), specific for the Montanera period (M_CF). The animals of the third batch (Mont3) were fed exclusively with acorns and grass during approximately 35 days, after that (116 days), the amount of acorn decreased so much that it was not enough for feeding and pigs were fed the same M_CF supplement up to slaughter weight (4 kg per day and pig). Pigs from each one of the three experimental groups were allocated in each one of the three fattening groups (Table 3). The average length of this period was 123 days (SD 16 days) and afterwards pigs were slaughtered in three different dates (Slaughter 1 to 3, Table 3) at 162 kg BW (SD 8.8 kg). Samples of acorns and pasture from different fences were collected and analysed (Table 1).

Backfat thickness, at the level of the last lumbar vertebra, and longissimus thoracis et lumborum muscle (LTL) area (LMA) were measured at the end of the growth period using PIE Medical-Aquila real-time ultrasound equipment, with an 18 cm probe and 3.5 MHz (Pie Medical Equipment B.V., Maastricht, The Netherlands). The ultrasound images were stored and subsequently interpreted using AutoCAD 2017 (Auto-CAD Academic Software. Mill Valley, CA, USA: Autodesk In). At the same time, backfat biopsies were taken from the rear part of the body for fatty acid profile analysis, approx. 4-6 cm from the midline of the spine in the dorsal lumbar region, at a depth of about 2 cm. Samples were obtained using a cylindrical biopsy device with a sharpened edge 0.5 cm in diameter under local anesthesia with 2% lidocaine-HCI (Alphacaine; Fendigo, Brussels, Belgium). Pigs were previously tranquilized by intramuscular injection of 2 mg/kg BW of azaperon (Stresnil, Esteve, Barcelona, Spain). After the biopsy, the area was sprayed with oxytretracycline and lidocaine (Veterin Tenicol; Lab. Intervet S.A., Salamanca, Spain). All necessary measures were taken to prevent and avoid discomfort of the animals during and after the process. Biopsy samples were placed in cryotubes, snap frozen in liquid nitrogen, and stored at -80 °C until analysis.

The day before slaughter, the animals were weighted. Pigs were euthanized by exsanguination after electrical stunning. The carcass was dissected, and the weight of the carcass, hams, loins and shoulders were individually recorded at the cutting line. Carcass length and backfat thickness were also measured with a ruler. Samples of backfat and LTL were collected and stored at -20 ºC until quality meat analyses and fatty acid composition were carried out.

Analytical methods

Nutrient analyses

All feed and muscle analyses were performed in duplicate. Dry matter (method 934.01), total ash content (942.05) and crude fiber (962.09) were determined using official methods (AOAC, 2005). The fat content in the feeds was determined by extraction with diethyl ether according to standard procedures (AOAC, 2005). Intramuscular lipids were extracted with a chloroform/methanol mixture (1:2) according to the method described by Bligh & Dyer (1959). Total nitrogen was analysed according to the Dumas procedure using a LECO Truspec CN kit (LECO Corporation, St. Joseph, MI, USA). The crude protein content was calculated using the factor of 6.25.

Fatty acids analyses

The lipids of the subcutaneous fat samples from the biopsies and the carcasses were extracted in a microwave oven following the method described by De Pedro et al. (1997). The backfat was homogenized with chloroform until the fat was completely dissolved, and then the chloroform was collected and evaporated under a stream of nitrogen. The neutral lipid fraction of intramuscular fat (IMF) from LTL was extracted by solid phase extraction according to Kaluzny et al. (1985) with the modifications proposed by Pinkart et al. (1998).

The separation and quantification of fatty acids (FA) from the total lipids of backfat and from the total and neutral lipids of LTL were determined by gas chromatography according to Sandler & Karo (1992). The gas chromatograph (Hewlett-Packard 4890 Series II, Hewlett-Packard, Avondale, PA, USA) was equipped with a split/splitless injector and a flame-ionisation detector. The fatty acid methyl esters (FAMEs) were separated on a polyethylene glycol capillary column (Carbowax 20M) (30 m × 0.25 mm id × 0.25 μm film thickness) maintained at 200 ºC for 20 min. The injector and detector temperatures were 250 °C. Nitrogen was used as a carrier gas at 1.8 mL/min. Individual FAMEs were identified by comparing their retention times with those of reference standard mixtures (Sigma Chemical Co., St Louis, MO, USA). The fatty acid profile was expressed as a percentage of identified fatty acids methyl esters. The percentages of the total sums of saturated fatty acids (SFA), monounsaturated (MUFA) and polyunsaturated (PUFA) were also computed.

Water holding capacity

Water holding capacity (WHC) was determined on LTL by three different methods. Water loss by centrifugation was determined according to the method described by Tejerina et al. (2012) and Zheng et al. (2018). Approximately 2.5 g of sample was weighed and wrapped in preweighed filter paper. The sample was then centrifuged at 4,000g for 20 min at 16 °C. After removing the sample, the filter paper was reweighed. The loss by centrifugation was calculated as the difference in weight between the initial and final weight of the filter and was expressed as the percentage of water loss with respect to the total weight. Measures were done by duplicate.

For the determination of thawing loss (TL), a portion (~ 100 g) of LTL was vacuum-packed in nylon/polyethylene bags and kept frozen at -20 °C. The difference between the initial weight of the fresh sample and the final weight of the thawed samples (day 1 and 30, respectively) was used to calculate TL. Results are presented as g of water/100 g of muscle. The same samples as TL were used for the determination of cooking loss (CL). Samples of LTL were vacuum-packed in nylon/polyethylene bags and cooked by immersion at +80 °C for 45 min in a temperature-controlled water bath (Combes et al., 2004). The difference in weight before and after cooking was used to calculate CL. Results are presented as a percentage.

Instrumental texture analysis

Warner-Bratzler (Stable Microsystems TA.XT Plus, Godalming, UK) analyses (Honikel, 1997) were performed on the same samples as previously measured cooking and thawing losses. The samples were prepared in 1 × 1 × 3 cm slices (width × length × thickness), cut with a Warner-Bratzler blade (HDP/BS) in a perpendicular direction to the muscle fibers and the maximum shear force was recorded. Instrumental determinations were repeated 8 times per sample and the replicate values were averaged.

Instrumental color

Instrumental color (CIE L*, a*, b*) was measured in the cross sections of LTL muscle using a Minolta Chromameter (CR-400, Konica Minolta Corp., Japan) with illuminant D65 and 0° standard observed. The colorimeter was calibrated before use with a white ceramic tile. In addition, the saturation index or chroma (C*= (a*2 + b*2)0.5), which indicates the brightness or vividness of color, and the hue angle (H° = arctg b*/a*), which describes the hue color, were also calculated. Color features were obtained as the average of 3 measurements performed at randomly selected places on each muscle.

Myoglobin content

The concentration of myoglobin content was assayed from the total haem content according to Hornsey (1956). In summary, 2.5 g of LTL muscle were ground in a mincer and homogenized with 10 mL of acetone and 0.25 mL of 37% HCl. After 24 h, the mixtures were filtered and the absorbance of the supernatant was detected at 512 nm against a blank using a spectrophotometer (Thermo Scientific, Sorvall, ST16R). Results are presented as mg of myoglobin per g of muscle.

Statistical analyses

The performance data collected during the growth period and the FA profile of the backfat samples from the biopsies were analysed with a univariate general linear model that included sex (two levels) and diet (three levels) as the only fixed effects. All the traits recorded after slaughter were analysed with the same models, but another effect was also included as a fixed factor that combined the date of slaughter and the diet batch of the individual during the fattening period (4 levels, Table 3). For the analysis of the FA profile at slaughter, the values of the same FAs at the end of the growth period were used as covariates. The weights of the premium cuts were also adjusted by the carcass weight as covariate.

It was not possible to adjust the interaction between the dietary treatment during the growing period and the batch during the Montanera (12 levels) because the number of observations prevents a balanced design and the data structure resulted in missing cells. The analyses were carried out in R studio (2020) using the linear model (lm), anova and Tukey HSD functions. The Fischer test was applied to test the effects of the factors and the Tukey test to make pairwise comparisons. Differences with a p-value lower than 0.05 were considered as significant.

RESULTS

The results on performance after the growth period and on the carcass traits at slaughter are described in Table 4. Carcass weight was not modified by dietary treatments (p > 0.05). Not significant differences were observed for the weights of hams, shoulders and loins. Measurements of loin area and backfat depth at the last and tenth ribs at 90 kg were lower in DOP pigs (p ≤ 0.001) but after Montanera period, backfat depths at 160 kg were very similar in the three diet groups because the depth increase between the two periods was slightly higher for the DOP group (p = 0.065).

Table 5 shows the mean values of the 15 fatty acids analysed from backfat biopsies at 95 kg. Significant differences were observed between the groups of treatment for most of them. The control diet (C) showed higher values for the main saturated fatty acids (ΣSFA, C14:0, C16:0, C18:0). The dry and wet diets including olive by-products (DOP and WCOC) have higher levels of unsaturated FA (ΣPUFA, C18:1, C18:2, C18:3) than C group, however, there were not many differences between them. Mean values of some low abundant FA were significantly different (C17:0, C20:1, C20:2, C20:3 n-3) between DOP and WCOC. For stearic acid (C18:0), DOP group showed lower proportions than WCOC, although the sums of FA (ΣMUFA, ΣSFA and ΣPUFA) were similar.

The results regarding backfat fatty acid profile from slaughtered pigs are also shown in Table 5. The WCOC pigs had higher content of C18:0, C20:0, C20:1 and ΣSFA (p ≤ 0.001 to p ≤ 0.057) and lower content of C18: 1 (p = 0.030) than the DOP pigs, with similar values when compared to group C except for C16:1 (p = 0.013). The change in the fatty acid profile from the beginning to the end of the fattening period (expressed as an increase or decrease in the percentage of fatty acids estimated every 10 days of Montanera) is shown in Figure 2 for the sum of saturated, monounsaturated and polyunsaturated fatty acids. The decrease in ΣSFA and ΣPUFA were significantly different in the control group compared to the experimental diets, more pronounced in the first sum and less in the second.

Some slight differences were found in the fatty acid profile of LTL intramuscular fat, as shown in Table 6. The contents of C18:2, C20:2, C22:3 n-3 and ΣPUFA in the total lipid fraction were higher (p ≤ 0.05) in the WCOD than in the DOP group. However, these values were not different compared to control group, except for C20:2 which showed significant differences. The differences between diets were smaller in the fraction of neutral lipids. The diet seems to modify just C18:2 and ΣPUFA contents, with higher values in WCOC than in the DOP and C groups (p ≤ 0.01).

The chemical composition and quality traits of LTL muscle between groups are shown in Table 7. No significant differences were observed in chemical composition. The average values for IMF, protein, water and ash in this muscle were 5.72, 24.48, 70.08 and 1.05, respectively (expressed as g/100g). Color parameters (L *, a * and b *) and Chroma and Hue angle values were similar for pigs on WCOC and C pigs, but some differences were observed between the DOP and WCOC groups. Pigs fed the WCOC diet showed darker muscles (lower L*, p ≤ 0.05) and higher values of redness (a*, p ≤ 0.000) than muscles from DOP pigs, with clear differences between both also for the Hue angle and for Chroma (p ≤ 0.000). Similarly, higher myoglobin content was observed in WCOC compared to DOP pigs (2.0 vs 1.7 mg/g, p ≤ 0.05), while C pigs showed intermediate values (1.9 mg/g). In the remaining quality traits analysed, the WCOC muscles had lower water losses due to thawing (p ≤ 0.05) than the DOP, while the losses due to cooking and centrifugal force were lower for WCOC, but the differences were not statistically significant. Higher centrifugal losses were observed in C compared to WCOC loins (p ≤ 0.05) and no significant effect of dietary treatment for shear force measurements.

DISCUSSION

In the traditional production system of the Iberian pig that establishes the exploitation of the natural resources of the Montanera, a long period of time must be applied with a strong feed restriction to achieve an adequate size and body composition but with low fat content. The recent legal requirements that regulate the breeding conditions determine that Iberian pigs must be at least 12 months old and around 100 kg BW ±10 kg at the beginning of Montanera period. The Iberian pig has a high intake capacity (Lopez-Bote, 1998) thus this feed restriction prolonged over time could create anxiety to the animals. Among the several alternatives proposed to solve this negative effect on animal welfare, this study has evaluated the increase of the ration using agricultural by-products with high fiber content.

It is worth to highlight the complexity of including this type of by-products in pig´s feeding, due to the need to find an adequate form of administration. The olive pulp was dehydrated and therefore was included in a compound feed (DOP_CF). Olive pulp by-product corresponded to 45% of compound fed, the straw was 25%, and a 15.3% was made up of soybeans, which represented 51.8% of the compound feed without by-products. In the case of the crude olive cake, it was mixed with straw to transform it into a suitable raw material that could be used to fed pigs. In this case, the silage would contribute a 70% of the total intake of dry matter, the remaining 30% was provided by the feed supplied as a complement (WCOC_CF). Therefore, this feed had a 50% of soybean meal. The initial idea was that both diets were similar in the contribution of nutritional principles, but each one had different by-product and supplementation form.

The effects on growth and carcass performance and meat quality of two presentations of the olive oil industry by-products, in dry olive pulp and wet crude olive cake, were analysed and compared with a usual control diet (C).

In a previous work (García-Casco et al., 2017), although individual feed intake measures were not available, it was found that growth was different between diet groups but the final slaughter weight were similar. The average growth during the growing period was higher in the group of pigs fed the WCOC compared to the fed the DOP diet. But during the Montanera the DOP pigs reached the commercial slaughter weight almost at the same time as the control group and shortly after the WCOC pigs, within the usual timeframes for this type of fattening. Growth between both periods was finally balanced and, therefore, no important differences were observed between the dietary treatments. Also, there were less fat deposition in the backfat tissue and a smaller loin area in the DOP than the WCOC and C groups at the end of the growth period. However, in DOP group, the modification that had to be made in the ration supplied during the first weeks (from ad libitum to restriction at 60%), could affect these results. In any case, the lower backfat thickness disappeared after the fattening period, since the thicknesses were similar at slaughter. This compensatory effect was not reflected in a greater weight of the lean tissues since no significant differences were observed in the premium cuts.

The backfat fatty acid profile is one of the main determinants of the organoleptic properties associated with dry-cured Iberian pig products. The processing industries demand a minimum of 53% oleic acid and low values for palmitic (<22%), stearic (<10.5%) and linoleic (<10.5%) acids at slaughter (BOE, 2007). For this reason, the farmer must control the FA profile before the Montanera by providing adequate diets. High SFA levels resulting from some growth diets may not be changed with acorn feeding during Montanera and, therefore, the final quality of the products may be affected. The pre-Montanera feed formulation used for Iberian pigs takes this situation into account and many of them have a high oleic acid content obtained from raw materials such as sunflower seeds. In our study, the animals consumed different amounts of dietary fat with different fatty acid profiles. For those fed the DOP and C diets, a restricted feeding regime was implemented. Pigs fed the DOP diet consumed more fat and richer in oleic acid and MUFA than the control group. The WCOC group consumed a compound feed (WCOC_CF) rich in PUFA and low in oleic acid, but ad libitum consumption of the crude olive cake by-product increased their oleic acid intake and de novo oleic acid deposition. In general, similar FA backfat composition was observed between the diets consumed by the pigs containing olive by-products at 95 kg, but they were different from those fed the control diet. At the end of the growth period, the animals fed with DOP and WCOC diets presented a backfat profile rich in oleic acid and in the essential fatty acids linoleic and linolenic and with lower proportion of SFA compared to the control diet. These observations partially agree with previous findings that reported differences in the backfat fatty acid profile in Alentejano pigs (Martins et al., 2017), a Portuguese breed very close to the Iberian (Muñoz et al., 2018), which were fed with high or low oleic acid feeds in a free-range system. According to Wood et al. (1989), as backfat depth increases, the concentrations of SFA and C18:1 increase and those of C18:2 and C18:3 decrease. In part, this is because the essential FAs come exclusively from the diet and they cannot be synthesized de novo, therefore, the pigs of the WCOC group, fed with enough crude olive cake by-product, have similar proportions of PUFA, SFA and MUFA than those fed the DOP diet, despite the fact that the backfat depth at 95 kg was 41% thicker in the WCOC than in the DOP group.

At the end of the Montanera period, the values of the fatty acids sums ranged from 28.3% to 29.7% for SFA, from 59.5% to 60.6% for MUFA and from 10.6% to 11.1% for PUFA, with minor statistical differences between the diet groups. These values were different from other studies as Niñoles et al. (2007), that showed lower levels of MUFA and PUFA, but closer to those reported by Hernández-Matamoros (2014). In general terms, the differences observed in the FA profile during the growth period were compensated after the fattening period in the free-range regime. The differences in proportions of the four most abundant FAs found at the end of the growth period of diet C with respect to the DOP and WCOC groups, were equalled after the fattening period. Even slightly higher proportions of oleic acid and lower of saturated fatty acids sum were observed at the end of this period in pigs fed the DOP diet. Therefore, the inclusion of olive by-products in the diets maintains the proportions of SFA and MUFA within the desired limits, without an increase in PUFA that usually causes oxidation and rancidity problems in the processed products.

The proportions of fatty acids change as the fat content of the tissues increases (Wood et al., 2008), linked to an increase in C18:1 and MUFA due to the raising of lipogenesis and the decrease in C18:2 and C18:3 by a declining of the direct incorporation from the diet of these essential fatty acids. In the present study, it is noteworthy the lower proportion of PUFA in the loin of pigs fed DOP group comparing with the other two, probably reflecting the lower linoleic and linolenic acid content provided by the DOP diet that the Montanera period is not able to balance. Despite the different lipid profile and fat content of the growth diets, IMF, SFA and MUFA of LTL muscle were similar at slaughter. Martins et al. (2015, 2017) explored the effect of a diet supplemented with oleic acid in 100 kg BW pigs and found changes in the neutral lipid fraction of muscles such as LTL, semimembranosus and biceps femoris, with higher proportions of MUFA and PUFA. However, the IMF percentage did not change. Other authors have previously reported significant differences in the percentage of FA of Iberian pigs fed diets with different saturated and unsaturated fatty acid levels (Cava et al., 1997; Andres et al., 2001). Although in the present experiment there is an absence of data recorded on the muscles at the end of the growth period, feeding during the Montanera period most likely balances the effects on the fatty acid profile caused by the experimental diets supplied during the growth period.

The experimental diets had little effect on the chemical composition of the LTL muscle. A positive relationship between the tenderness or the acceptability of pork by the consumers and the IMF content have been described (Fonti-Furnols et al., 2011) and a IMF percentage lower than 2.5% negatively affects to the acceptability parameters. A drop in the percentage of IMF in lambs fed olive cake has been observed (Mioc et al., 2007) despite the high fat intake provided in the diet. However, those used in the present study did not affect the IMF content in the LTL of the Iberian pig. The values obtained with the by-product diets were similar to the control diet and to those described in other studies (Tejerina et al., 2012). Hence, the experimental diets did not modify this relevant meat quality trait.

Meat color is the first sensory characteristic of pork quality and it is considered a direct determinant of consumer preferences and purchase decisions (Needham & Hoffman, 2015). The results obtained in the color parameters showed that LTL from pigs fed the DOP diet present a different color pattern (especially for a*, Hue and Chroma) than the control and WCOC groups. The differences in redness need attention, since red-cherry color makes the meat more attractive for producers and consumers (Straadt et al., 2013). These results partially agree with those obtained by Joven et al. (2014), who observed changes in meat color in pigs fed increasing amounts of olive cake in the diet. The red color in the muscles is directly related to the presence of the myoglobin pigment. Accordingly, pigs fed WCOC and C diets showed a higher myoglobin content in LTL. The relationship between the red color and the IMF content or even the IMF fatty acid profile, specifically with the MUFA levels, pointed out in other studies (Palma-Granados et al., 2017; Nieto et al., 2019) has not been confirmed by these results.

WHC has a great influence on the sensory traits of meat such as color, juiciness, and tenderness (Tejerina et al., 2012), in addition to its economic and nutritional implications. The previous data available in more commercial and widespread pig breeds have not described any effect due to the inclusion of olive cake in the diet on water retention capacity and tenderness (Joven et al., 2014). However, in this study, small differences in meat water losses between dietary treatments have been observed. In general, the greatest differences were observed between animals from DOP and WCOC groups, with greater water losses in the first one. Furthermore, since WHC and color are closely related (Seiquer et al., 2019), the meat of DOP pigs shows signs of being paler and less red. Again, the negative relationship reported by other authors between IMF and water losses related to a higher moisture content (Gjerlaug-Enger et al., 2010) as well as meat hardness (Ramírez & Cava, 2007; Lorenzo & Carballo, 2015) have not been observed in the present work.

In addition to the few differences found in performance and meat quality traits between the experimental and control groups, some explanation regarding practical concerns is needed. The huge variability of the raw material used in this study is a complicated problem to solve. Previous works describe the high variability in the chemical composition of the olive cake due to the oil extraction process, the year and geographical origin of the olives, or the posterior processing out such as drying or pitting (Molina-Alcaide et al., 2003). The diet must be formulated for a specific raw material, and reformulation could be necessary even in the same pre-Montanera period. The WCOC group was fed straw and ensilage whose composition also changes seasonally. Moreover, the difficulty increases because it is not possible to know in advance the composition of the diet based on acorn and grass. The acorn composition changes depending on a series of uncontrollable external parameters (weather, parasites, etc.) that vary from year to year and even in the same Montanera period (Tejerina et al., 2011). Another concern was the amount of the DOP ration supplied, since in a first moment, it was intended to be supplied ad libitum, however, excessive and unwanted fattening of the animals was observed. Therefore, a restriction on the diet supplied had to be decided during the experiment. A new ration adjustment should be studied in the future to find out if this by-product could be a viable option. Regarding the WCOC group, the lack of a current market makes its preparation difficult and expensive. However, the raw material used is cheap, and it also represents a notable improvement in the sustainability of olive oil production, since it would contribute to solve, in part, the problem of the management and processing of secondary waste from its extraction. The cost of these feeds is a point that must be taken into account, therefore further studies on the cost/benefits of the feeds should be carried out.

The use of a diet based on by-products from olive oil extraction has two important aspects to be considered, as mentioned above, the sustainability of the agricultural environment, and the improvement of animal welfare. Although an animal welfare study has not been properly carried out in our work, other authors have evaluated Iberian pigs fed with rice husk (which is a rice by-product) using the Welfare Quality© test and they obtained favorable results (Matías et al., 2018). According to the conclusion of their work, the high proportion of fiber in the rice husk had a satiating effect on the animals, causing a positive emotional state. In our case, the field observations evidenced a less noisy and nervous behavior when the daily ration of the group fed the WCOC diet was supplied. Thus, continued access to feed in a silage form seems to reduce anxiety produced by the restriction regime.

In conclusion, the use of by-products from the olive industry in feeding of the Iberian pigs during the growth period before the Montanera, does not substantially alter either the growth, carcass composition or meat quality traits. However, the supplementation of olive cake in its dry presentation has shown slight negative effects on some quality parameters (color, water losses or the fatty acid profile in the loin) that should be evaluated in greater depth. A further investigation of by-products effects on sensory traits would be useful before its potential use in pig feeding in order to ensure the maintenance of the high organoleptic quality associated to Iberian pig meat products. By-product supplementation in silage form does not produce any negative effect on quality and could reduce the welfare problems derived from the strong dietary restriction necessary in this type of production. Its use in feeding Iberian pigs in extensive systems, at the expense of adequate economic studies, could be highly recommended.

ACKNOWLEDGEMENTS

The authors would like to thank the staff of the Puerto Lobo farm for their dedication and cooperation and, above all, to its owner, Francisco B. Ramírez, for making it available to us to carry out the experiment.