After 90 d of experience, pigs were transported to a commercial abattoir, fasted for 12 hours (h) and slaughtered in two different d in which both genders and rearing systems were adequately represented. The pigs were electrically stunned, exsanguinated, scalded, skinned, and eviscerated according to standard commercial procedures. Data collected at slaughter were: slaughter weight (SW) and hot carcass weight (HCW). At 45 minutes (min) post mortem, initial pH (pH45) was measured at the 7^th thoracic vertebra of the carcasses using a portable pH-meter Lutron mod. 212 (Taipei, Taiwan) equipped with a glass electrode. Immediately after, and after making a cut in the m. Longissimus thoracis, CIE L*a*b* colour parameters (Commission International de l'Eclairage, 1976) were measured at three locations of a freshly cut surface of m. Longissimus thoracis at 12^th rib after 30 min blooming using a Lutron RGB-1002 (Taipei, Taiwan) colorimeter (illuminant D65, 10° standard observer) calibrated with black and white tiles. The average of 3 random readings was used to measure L*, a* and b*.

The carcasses were suspended in the air and refrigerated (Cora Refrigeración, Quito, Ecuador) at 4 °C for 24 h, and data collected were: cold carcass weight (CCW), and ultimate pH (pH24h) and meat color in the same place and procedure than previously. Then, carcasses were split down the midline according to standard commercial procedures and the head was removed at the atlanto-occipital junction. In the right side of each carcass, backfat thickness (with calipers with a precision of 0.1 mm) at midline of the carcass at the position of the 1st, 12^th and last ribs (three measurements are made at each location and the final value for backfat was considered the mean of the three estimations). After, 150 g samples of the m. Longissimus thoracis were taken at the level of the 6^th rib for instrumental meat analysis (drip and cooked losses). Afterwards, three samples (of approximately 6 cm thick) of the m. Longissimus thoracis and lumborum (from the 7^th rib to the last lumbar vertebra) were taken, weighed, placed in plastic bags and vacuum packaged and stored at 4ºC for 24h (cranial sample), and 7d (middle sample) and 15d (caudal sample). Aged samples were frozen at -20 ºC until subsequent sensorial meat analysis.

The left side of each carcass was fully dissected into skin, bone, muscle and fat, according to the EU reference method [43]. The carcass composition was calculated as well as the weight of each component from the dissection with respect to the half-carcass.

All groups showed similar values of pH₄₅and pH_24h (TABLE III), darker than that of indoor pigs, which coul probably be due to an enlarged spontaneous activity, leading to an increased mean fiber cross-sectional area. As regards meat color measurements considered to be within a normal pH range for pigs that do not produce Pale Soft and Exudative (PSE) meat, to those recorded in most of the reviewed literature. The average values recorded by Fortina et al. [12] for pH45are higher than those observed in our study (6.48 vs. 6.29 and 6.06 vs. 5.39, respectively), while the pH fall was lower (0.42 vs. 0.89 in the present study), revealing an incomplete post mortem glycolysis associated with pre-slaughter stress. In the present study, the rearing system and gender had no significant effect (P> 0.05) on pH measured both 45 min and 24 h post-mortem. Other authors [14, 38] have also found that sex does not influence the pH. However, in the literature, the effect of sex on ultimate pH remains unclear [3].

The color measurement (L*, a* and b* indexes) corresponded with the dark and red meat of the autochthonous pig breeds (TABLE III). Cie L*45 and b*45 values were lower than those obtained by Estévez et al. [10] in free-range reared Iberian pigs slaughtered at 90 kg live-weight (37.69 and 0.4 vs. 44.99 and 4.26, respectively), whereas the a*45 values were similar (10.87 vs. 10.50). Also, Renaudeau and Mourot [38] in Creole pigs, and Galián et al. [16] found higher levels for L*, a* and b*. In agree with Peinado et al. [32] and Galián et al. [17], it can be seen that at 24 h, all the color parameters had increased (37.7 vs. 50.0, 10.8 vs. 14.8, and 0.4 vs. 3.0), which is normal due to ageing time [34]. Meat color measurements did not differ between rearing systems. It is generally accepted that meat from outdoor reared animals is at 24 h post-mortem, gender significantly affected (P < 0.05) the three chromatic variables assessed. It was apparent that the raw meat from gilts had a lighter, less reddish and more yellowish color when compared with barrows.

The data of the present study did not show such differences in pH_24h in any of the two different rearing systems, which could explain the lack of differences in DL and CL (data not show). As noted Meilgaard et al. [29], in the present study an increase in pH resulted in pork with lower L*, a*, b*, although correlations reached statistical significance only for L * (-0.367). Also, we found that the ultimate pH of m. Longissimus thoracis was negatively correlated to the cooking loss (-0.355), in agree with Purslow et al. [36].

Several studies evaluating all four sexes have had mixed results in relation to pork quality characteristics. Jeong et al. [22] found that a* and shear force differed significantly among the sexes. However, pH, L* and b* did not differ. Gispert et al. [18] drew similar conclusions in that L*, and a* differed, but b* and pH did not differ among the four sexes. Boler et al. [5] observed that pH differed but L*, a*, and b* did not differ. Gispert et al. [18] explained that although there were significant differences in some characteristics, the meat quality was considered to be of good quality among the four sexes. Therefore, the consumer may not be able to detect differences in meat from different sexes when meat quality is acceptable.

TABLE III
EFFECT OF REARING SYSTEM AND GENDER ON PH, COLOR, AND DRIP AND COOKING LOSSES OF LONGISSIMUS THORACIS M. FROM CREOLE BEC PIGS

RMSE =root mean square error; SYS = rearing system; GEND = gender; ns = no significant; * = P< 0.05; ** = P <0.01; ***

= P < 0.001;

a-bMeans in the same row with a common superscript letter do not differ significantly (P>0.05).

In the present study, like other sensory studies, it was necessary to freeze some samples for a short period of time before evaluation. Average values of sensory parameters are showed in TABLE IV. Based on these results, the meat of the animals studied can be defined as moderately tender, with little brightness and color, with high odor and pork flavor, little unctuous on the palate, juicy and with moderate to high persistence of flavor. The evaluation by the sensory panel of the pork loins from the four groups did not show large differences in tenderness and pork flavor. However, significant differences (P < 0.05) were obtained for the attributes meat color, brightness, lard flavor, sweet taste, metallic taste, unctuousness, juiciness and persistence of flavor between rearing systems. Lard flavor, sweet taste, metallic taste, juiciness and persistence of flavor were significantly higher in C pigs when compared with the SC pigs, while meat color, brightness, pork odor and unctuousness were lower. Differences could be due to the characteristics of the rearing system, exercise and/or physical parameters [6, 23], but not be ascribed to a pH difference as no significant differences in pH were found between the different systems and genders (TABLE IV). In contrast to rearing system, few sensory attributes of cooked meat were affected by gender: metallic taste (higher in barrows, P < 0.01), and brightness (higher in gilts, P < 0.01). One experiment comparing sensory attributes [9] of the four sexes observed that, for fresh roasts, frozen chops and roasts, sexes differed in pork flavor, while no differences were found among sexes for juiciness, chewiness, and tenderness

It is widely accepted that the tenderness is the most important factor affecting overall meat acceptability. The tenderness was positively correlated with cooking loss (0.604), not reaching statistical significance other correlations with postmortem meat instrumental traits. Nam et al. [30] reported that tenderness score was correlated with L* and drip loss, and negatively correlated with pH45. Also, cooking loss was positively correlated with brightness (0.542), while showed negative correlations with taste attributes as sweetness (-0.486) and metallicness (-0.483), and juiciness (-0.494), which agrees with Nam et al. [30]. On the other hand, a high correlation between cooking loss and juiciness is reflected in Pillsbury [33]. As the above authors reported, these results suggest that panelists do not prefer pork in terms of taste which has high cooking loss and is perceived as tough meat. So, it may be natural that taste acceptance decreases as cooking loss increase.

The ageing time influenced a number of pork characteristics although its effect had limited benefit for the sensory attributes of m. Longissimus thoracis from Creole BEC pigs. Ageing time had significant effect (P < 0.05) on the tenderness, brightness, unctuousness and juiciness (TABLE IV). The values of the four traits were significantly higher (P < 0.05) as the ageing time increased. Similar results were reported by Juárez et al. [24] who analyze the evolution of the sensory characteristics of samples from m. Longissimus thoracis aged for 2, 7 and 14 d. Their results showed that initial and overall tenderness, juiciness and flavour intensity increased (P < 0.01) from d2 to 14, which they attribute to ageing increases the concentration of flavour precursors, responsible for forming the characteristic meat flavour during cooking. Ngapo et al. [31] found that ageing time had significant effects (P < 0.05) on the sensory quality when grilled or shabushabu, but not when roasted. These authors indicated that grilled pork at the longer ageing period had stronger sweet taste and caramel, more panelists made notes and a greater number of notes were obtained for the pork that had been stored 43 d than for 5 d. The enzymatic action is the most important factor affecting meat tenderization [39]. Juiciness showed a similar evolution along ageing between groups, in contrast to the results obtained by Campo et al. [7].

TABLE IV
SENSORY EVALUATION OF THE LONGISSIMUS THORACIS M. COOKED FROM CREOLE BEC PIGS. EFFECTS OF REARING SYSTEM, GENDER AND AGEING TIME

SEM = standard error of the means

x-zMeans in the same row with a common superscript letter do not differ significantly (P > 0.05).

a-cMeans in the same column and between the same trait with a common superscript letter do not differ significantly (P > 0.05).