1 Carnosic acid dietary supplementation at 0.12% rates slows 2 down meat discolouration in gluteus medius of fattening lambs 3 RUNNING TITLE: 4 L. Morán1*, J. M. Rodríguez-Calleja2, R. Bodas1, N. Prieto1, F. J. Giráldez1 5 and S. Andrés1 6 1 7 24346 Grulleros, León, Spain. 8 2 9 24071 León, Spain. Instituto de Ganadería de Montaña (CSIC-Universidad de León). Finca Marzanas. E- Departamento de Higiene y Tecnología de los Alimentos, Universidad de León, E- 10 11 12 13 14 15 16 (Received - Accepted) 17 CORRESPONDING AUTHOR: Lara Morán, Instituto de Ganadería de Montaña 18 (CSIC-ULE). Finca Marzanas. E-24346 Grulleros, León, Spain. Tel. +34 987 307 054 19 Fax +34 987 317 161 E-mail: laramoran@eae.csic.es 1 20 Abstract 21 Thirty-two Merino lambs fed barley straw and a concentrate alone (CONTROL) or 22 enriched with vitamin E (VITE006) or carnosic acid (CARN006; CARN012) were used 23 to assess the effect of these antioxidant compounds on meat quality attributes. The 24 animals were slaughtered after being fed for at least 5 weeks with the experimental 25 diets. The longissimus lumborum samples of VITE006, CARN006 and CARN012 26 groups showed higher values (P<0.001) of L* (lightness) thorough the complete storage 27 period under modified atmosphere when compared to the CONTROL group. Moreover, 28 the VITE006 and CARN012 samples revealed lower discolouration when compared to 29 the CONTROL group, these differences being more apparent in a less colour stable 30 muscle such as gluteus medius (P<0.05 for hue after 14 days of refrigerated storage). 31 Meat sensory traits were not significantly affected by carnosic acid and microbiological 32 analyses were not conclusive at the doses administered. 33 34 Keywords: Microbial spoilage; colour; sensory evaluation; water holding capacity; carnosic acid; rosemary 35 Introduction 36 The shelf life of meat and meat products is seriously shortened by two main factors: 37 microbial spoilage and colour stability, the last one being closely related to lipid 38 peroxidation (Young & West, 2001). Therefore, any finding focused on delaying either 39 of these processes is highly relevant for the meat industry. However, the high quality 40 meat demanded by the consumer in developed countries (Boleman et al., 1997) must be 41 free of chemically synthesized additives, so the addition of synthetic food stabilizers to 42 meat in order to extend the shelf life of these products in the market does not seem to be 2 43 the best choice. A suitable alternative is the inclusion of natural compounds (plant 44 origin) in animal feedstuff, thus avoiding any further manipulation of the meat. 45 In this context, especial attention has been paid to the antimicrobial and antioxidant 46 effects promoted by rosemary (Rosmarinus officinalis L.), a herb commonly used as a 47 flavouring agent. The bioactive properties of this herb have been attributed to the 48 phenolic compounds in rosemary plants (Hernández-Hernández, Ponce-Alquicira, 49 Jaramillo-Flores, & Legarreta, 2009). These compounds have demonstrated 50 antimicrobial and antioxidant activities when added to food as additives (McBride, 51 Hogan, & Kerry, 2007), but they have also shown beneficial effects on eggs, milk and 52 meat products when rosemary is included in the diet of the animal (Botsoglou, Govaris, 53 Giannenas, Botsoglou, & Papageorgiou, 2007; Galobart, Barroeta, Baucells, Codony, & 54 Ternes, 2001; Jordán, Moñino, Martínez, Lafuente, & Sotomayor, 2010; Nieto, Díaz, 55 Bañón, & Garrido, 2010). In this last sense, the main phenolic compound retained in 56 animal tissues after the consumption of rosemary is carnosic acid (Moñino, Martínez, 57 Sotomayor, Lafuente, & Jordán, 2008), so it can be hypothesized that the antimicrobial 58 and antioxidant properties observed in meat quality are mainly due to the increment of 59 this phenolic compound at this level. However, the amount of carnosic acid which must 60 be fed to the animals to have beneficial effects on meat quality has not yet been 61 quantified. It must also be considered that the concentration of phenolic compounds in 62 the plants varies depending on the maturity stage or climatic conditions, mainly 63 conditions of drought (Munné-Bosch, Mueller, Schwarz, & Alegre, 2001). This is the 64 reason why feeding rosemary extracts with a known richness of carnosic acid instead of 65 intact plants will allow recommendations to be established about the amount of 66 rosemary which should be fed to the animals according to its levels of carnosic acid. 3 67 Therefore, the aim of the present study was to investigate the shelf-life extension of 68 meat (antimicrobial properties and colour stabilization) when two different doses of 69 carnosic acid (from rosemary extract) were included in the diet of lambs. Likewise, 70 vitamin E (one of the most frequently used antioxidants in animal nutrition) was 71 included in another group as a positive control. 72 Material and Methods 73 2.1 Animals and diets 74 Two weeks before the commencement of the trial, 32 male Merino lambs were 75 treated with Ivermectin (Ivomec, Merial Labs, Barcelona, Spain) and vaccinated against 76 enterotoxaemia (Miloxan, Merial Labs, Barcelona, Spain). 77 After random stratification on the basis of body weight (average body weight (BW), 78 15.2 ± 0.749 kg), the lambs were allocated to four different groups: a control group 79 (CONTROL), a group fed vitamin E (α- tocopheryl acetate) at a rate of 0.6 g kg-1 of 80 feed concentrate (VITE006, equivalent to 900 IU of vitamin E kg-1 of feed concentrate), 81 a third group fed a similar dose as the previous group (0.6 g carnosic acid kg-1 of feed 82 concentrate, CARN006) of carnosic acid (Shaanxi Sciphar Biotechnology Co., Ltd, 83 Xi'an, China) and the fourth group fed a double dose of carnosic acid (1.2 g carnosic 84 acid kg-1 of feed concentrate, CARN012). The animals were then individually penned. 85 All handling practises followed the recommendations of the European Council 86 Directive 86/609/EEC for the protection of animals used for experimental and other 87 scientific purposes, and all of the animals were able to see and hear other lambs. 88 After 7 days of adaptation to the basal diet comprised of concentrate (50% barley, 89 20% soybean meal, 15% maize, 8% wheat, 4% molasses and 3% mineral premix; 90 chemical composition: dry matter (DM) 888 g kg-1, crude protein (CP) 178 g kg-1 DM, 4 91 neutral detergent fibre (NDF) 134 g kg-1 DM, and ash 56 g kg-1 DM) and barley straw 92 (DM 913 g kg-1, CP 35 g kg-1 DM, NDF 757 g kg-1 DM, ash 55 g kg-1 DM), all of the 93 lambs were fed barley straw and the corresponding concentrate feed alone (CONTROL 94 group) or supplemented with either vitamin E or carnosic acid. The concentrate (35 g 95 kg-1 BW day-1) and forage (200 g day-1) were weighed and supplied in separate feeding 96 troughs at 9:00 a.m. every day, and fresh drinking water was always available. 97 2.2 Slaughter procedure, packaging, storage and sampling 98 The animals were slaughtered on four different days, two lambs per group each day. 99 The lambs were selected each day according to their weight (25 kg intended body 100 weight) and slaughtered by stunning and exsanguination from the jugular vein; they 101 were then eviscerated and skinned. The hot carcass of each lamb was weighed, chilled 102 at 4° C for 24 h and weighed again. The pH value from the longissimus thoracis muscle 103 at the sixth rib was determined in triplicate at 0 h, 45 min and at 24 h post-mortem 104 before the muscle was removed from the carcass, using a pH meter equipped with a 105 penetrating glass electrode (pH meter Metrohm® 704, Switzerland). 106 The left hind leg was removed and stored at -30 ºC until sensory evaluation. 107 Moreover, the longissimus thoracis (LT) et lumborum (LL) and gluteus medius (GM) 108 muscles were removed from the right and left carcass sides. The LT samples were used 109 for chemical analysis in accordance with the methods described by the Association of 110 Official Analytical Chemists (AOAC, 2003), whereas the LL and GM muscles were cut 111 into slices 2.5 cm thick, placed on impermeable polypropylene trays and wrapped with 112 ML40-G bags (Krehalon; Proveedora Hispano Holandesa S.A., Barcelona, Spain), 113 which were immediately modified-atmosphere packaged (MAP) using a tabletop 114 Multivac 115 Wolfertschwenden, Germany). The air in the bags was replaced by a commercial gas A300 packaging machine 5 (Multivac Verpackungsmaschinen, 116 blend intended for red and poultry meats consisting of 35% CO2, 35% O2 and 30% N2, 117 with a gas:meat volume ratio of about 2:5:1. The ML40-G bags had O2 and CO2 118 transmission rates of 20 and 100 ml m-2 24 h-1, respectively, at 23 ºC and 80% relative 119 humidity. All packages were stored under simulated retail display conditions [12 h daily 120 fluorescent illumination (34 W) and 3±1 ºC] and the air temperature was monitored 121 using a Testo175-T2 data logger (Instrumentos Testo S.A., Cabrils, Barcelona, Spain). 122 The meat in these polypropylene trays was used to study the extract-release volume 123 (ERV, on LT muscle), microbial spoilage (on LT muscle), the rate of discolouration (on 124 LL and GM muscles) and the water holding capacity (WHC, on LL muscle). 125 2.3 Colour, extract release volume (ERV) and water holding capacity (WHC) 126 On each sampling day, the concentrations of CO2 and O2 inside each tray were 127 determined using a CheckMate 9900 (PBI Dansensor, Denmark). After opening the 128 packages, a slice of fresh meat from each muscle was measured for colour parameters 129 on days 0, 3, 7, 9 and 14. The L* (lightness), a* (redness) and b* (yellowness) values 130 (Centre Internationale de l'Eclairage, 1986) were used to determine the meat colour of 131 the muscles using a chromameter (Minolta® Chroma Meter 2002, Germany). The hue 132 angle (h*), which defines colour (0° is red; 90° is yellow), was calculated as arctangent 133 (b*/a*), and the chroma (C*), a measure of colour intensity (0 is dull; 60 is vivid), was 134 computed as 135 (ERV) was measured (on days 0, 7 and 14) as previously described (Rodríguez-Calleja, 136 Santos, Otero, & García-López, 2004). Briefly, minced lamb (15 g) was mixed with 60 137 ml of the extraction reagent (0.2 M KH2PO4 and 0.2 M NaOH; pH 5.8) and 138 homogenized for 2 min. The homogenate was filtered through Whatman no. 1 paper and 139 the ERV was recorded as the volume collected in 15 min. (a *2 b *2 ) (Young & West, 2001). Also, the extract release volume 6 140 141 The water holding capacity (WHC) was measured on LL muscle via cooking losses, according to Honikel (1998). 142 2.4 Microbiological analysis 143 Twenty-five grams of LL muscle from each tray (0, 7 and 14 days) was placed into 144 sterile Stomacher bags, rinsed with peptone water (1:5 dilution) and the rinsate was then 145 diluted tenfold. The numbers of total viable bacteria at 4.5ºC (TVB), Pseudomonas spp., 146 lactic acid bacteria (LAB) and Enterobacteriaceae (EC) were determined and 147 confirmed as described elsewhere (Rodríguez-Calleja et al., 2004; Rodríguez-Calleja, 148 García-López, Santos, & Otero, 2005). Briefly, TVB were determined by the pour plate 149 technique on Plate Count Agar (PCA; Oxoid, Basingstoke, UK) incubated at 4.5 ºC for 150 14 days. Pseudomonas spp. numbers were determined after 2 days incubation at 25 ºC 151 on a Pseudomonas agar base (Oxoid) to which a CFC (cetrimide, fucidin, cephaloridine; 152 Oxoid) supplement was added. The oxidase test (Oxidase Touch sticks, Oxoid) was 153 performed on randomly selected colonies and only oxidase-positive colonies were 154 counted as Pseudomonas spp. The LAB colonies were enumerated on overlaid plates of 155 MRS (de Man, Rogosa and Sharpe; Oxoid) agar following 3 days incubation at 30 ºC. 156 Brochothrix thermosphacta was enumerated on streptomycin sulphate cycloheximide 157 thallous acetate agar (STAA, Oxoid) after incubation for 2 days at 25 ºC. Overlaid 158 plates of Violet Red Bile Glucose agar (VRBGA, Oxoid) were used for EC counts after 159 24 h incubation at 37 ºC. 160 2.5 Sensory evaluation 161 The muscle vastus intermedius (VI) of the left hind leg was chosen for sensory 162 evaluation because it is considered a high value retail cut -lamb leg roast, chump- 163 (Johansen, Aastveit, Egelandsdal, Kvaal, & Røe, 2006). Sensory analysis was carried 7 164 out by 24 consumers. The left hind leg was defrosted at 4 ºC for 48 h and the VI was 165 dissected and cut into steaks 20 mm thick. The steaks were cooked a pre-warmed clam- 166 shell grill to an internal temperature of 74º C in the geometric centre of the steak 167 (measured by a Digi-Sense thermocouple probe, Cole-Parmer Instrument Company, 168 Illinois, USA), after which all fat and connective tissue was trimmed and the muscle cut 169 into blocks of 2×1×1 cm. The blocks were wrapped in pre-labelled foil (the blocks from 170 each animal were coded with the same alphabetical letter), placed in a heated incubator 171 and then given to the assessors in a random order chosen by a random number 172 generator. Two triangle tests (ISO 4120:1983) were carried out to discriminate between 173 meat from the lambs fed (1) CONTROL vs. CARN012, and (2) CARN006 vs. VITE006 174 diets. In the triangle test, the assessors were requested to identify which two of three 175 products were most alike and hence to indicate which product was the most different 176 from the other two. Additionally, all consumers participated in two blind preference 177 tests in which they received two meat samples on a coded paper plate. Each consumer 178 tasted and evaluated their preferences for the meat samples from lambs fed (1) 179 CONTROL vs. CARN012, and (2) CARN006 vs. VITE006 diets, and provided the 180 reasons for their choices. 181 2.6 Statistical analysis 182 Microbiological counts were transformed and expressed as log CFU g-1. The basic 183 descriptive statistics for each parameter (mean and standard deviation) were calculated. 184 The pH data from different times were analysed by a one-way analysis of variance. The 185 ERV, WHC, microbial population and colour parameters data (L*, C* and h*) were 186 subjected to a two-way analysis of variance, with treatment and day as the main factors. 187 The Tukey’s honestly significant difference test was used to compare means where the 188 main effects were significant (P<0.05). All analyses were performed using the GLM 8 189 procedure of the SAS package (SAS, 1999). Data from the sensorial analysis were 190 analysed using the chi-square test (Stone & Sidel, 1993) of the FREQ procedure in SAS 191 (SAS, 1999). 192 Results and discussion 193 A previous study (López-Bote, Daza, Soares, & Berges, 2001) described a 194 significant positive effect with low supplementation levels of vitamin E (270 mg kg-1 195 diet) on lipid oxidation in meat; however, the protection effects at different storage 196 times, and particularly the effect on meat surface redness, were optimized at a dietary 197 inclusion of 550-625 mg α-tocopheryl acetate kg-1 diet. Consequently, in the present 198 study, vitamin E was supplied at a rate of 0.6 g kg-1 of feed concentrate (VITE006 199 group). The same dose of carnosic acid was supplied to another experimental group 200 (CARN006) to compare the effectiveness of both components. The last group was fed a 201 double dose of carnosic acid (CARN012) in order to clarify whether this substance 202 could show a dose-dependent effect at the meat level. 203 3.1 Chemical composition 204 The mean and standard deviation (SD) of the chemical data corresponding to the 205 lamb meat samples (LT) are summarized in Table 1. 206 [INSERT TABLE 1 NEAR HERE, PLEASE] 207 As can be observed, chemical composition of LT meat samples seemed to be 208 affected by carnosic acid added to the diet, thus giving rise to higher levels of crude 209 protein content when compared to the VITE006 group (Table 1). 210 3.2 Colour parameters 211 Table 2 summarizes the results of LL and GM muscles when the colour stability was 212 studied after packaging in a controlled gas mixture for 14 days. These muscles have 9 213 been described in bovine as high (LL) and medium (GM) colour-stable muscles 214 (McKenna et al., 2005) due to the different values observed in their metmyoglobin 215 reductase activity (MRA) and oxygen-consumption rate (OCR) parameters (McKenna 216 et al., 2005; Young & West, 2001). 217 [INSERT TABLE 2 AND FIGURE 1 NEAR HERE, PLEASE] 218 As can be observed (Figure 1a), antioxidant treatments (CARN006, CARN012 and 219 VITE006) showed higher L* values in the LL muscle throughout the complete storage 220 period when compared to the CONTROL (P<0.05). However, no differences in L* 221 values were detected among CARN006, CARN012 and VITE006 groups (Figure 1a and 222 Table 2). It has been previously demonstrated that phenolic-rich extracts are iron 223 chelating agents (Saman et al., 2001) thus promoting either lower red blood cells counts 224 or low haemoglobin levels. In fact, lower levels of haemoglobin content were measured 225 in the blood of CARN006, CARN012 and VITE006 lambs when compared to the 226 CONTROL group (data not shown). The inhibition of iron absorption caused by 227 phenolic compounds in the diet of the lambs of the present experiment might have 228 caused lower myoglobin contents and hence higher values of L* in CARN006 and 229 CARN012 groups when compared to the CONTROL lambs. High doses of vitamin E 230 have also demonstrated to be effective in reducing total iron concentrations to normal 231 levels in the brain of mice (Lan & Jiang, 1997). The evidence observed by Lan & Jiang 232 (1997) would be in accordance with the low L*values in the LL of the VITE006 group. 233 Moreover, oxygenation (which involves oxymyoglobin formation) reduces L* values 234 while oxidation (which involves metmyoglobin formation) had an opposite effect, 235 suggesting that lightness is also affected by the state of the myoglobin (Fernández- 236 López, Pérez-Álvarez & Sayas-Barberá, 2000; Linares, Bórnez & Vergara, 2008). This 237 last argument might be the reason why L* was increased in the GM (medium colour10 238 stable muscles) of the CONTROL lambs after 7 days of refrigerated storage (Figure 1b), 239 whereas this parameter was quite stable in the GM of the lambs being fed antioxidants 240 (Figure 1b). Regarding LL muscle, L* was also very steady even in the CONTROL 241 group (Figure 1a), probably due to the high colour stability of this muscle. 242 As far as C* is concerned, the evolution of this parameter during refrigerated storage 243 under MAP was similar in both muscles (Figure 2). Thus, C* evolution could be fitted 244 successfully to a second order polynomial function in all of the treatments either on LL 245 (R2= 0.77-0.95) or GM (R2= 0.77-0.98) muscles. Consequently, the first derivative was 246 calculated for each curve in order to know when the most intense colour (C*max) was 247 reached for each muscle in each group (Fernández-López et al., 2000). Thus, in LL 248 muscle C*max was reached after 6.8 days of refrigerated storage in the CONTROL 249 lambs, and slightly later in the antioxidant groups (7.3, 7.4 and 7.6 days for CARN006, 250 CARN012 and VITE006 groups, respectively). Thereafter, LL discolouration took 251 place, with CARN006, CARN012 and VITE006 groups showing no differences in C* 252 when compared to the CONTROL lambs after 14 days of refrigerated storage (Figure 253 2a). 254 [INSERT FIGURE 2 NEAR HERE, PLEASE] 255 The previous statement is in agreement with the higher values (although not 256 significant) of the h* parameter in the LL muscle after 14 days of refrigerated storage in 257 the CONTROL and CARN006 groups (Figure 3a), which also indicated that the meat 258 become dull faster in these lambs when compared to VITE006 and CARN012 samples 259 (Young & West, 2001). However, it must be stressed that none of these differences in 260 C* and h* values reached a level of significance in LL samples. This might have been 261 due to either a lack of effect of treatment or the high colour stability of LL muscle. 11 262 [INSERT FIGURE 3 NEAR HERE, PLEASE] 263 On the contrary, once again a medium colour-stable muscle such as GM showed 264 more promising results when antioxidants were fed to the lambs. Thus, in GM muscle 265 C*max was reached after 5.5 days of refrigerated storage in the CONTROL lambs, and 266 after 5.9, 7.6 and 7.5 days in CARN006, CARN012 and VITE006 groups, respectively. 267 According to these data meat discolouration would be delayed at least two days in 268 medium-colour stable muscles from lambs being fed the highest antioxidant doses. 269 Thereafter, a more pronounced decrease in C* values was observed in CONTROL 270 samples, thus indicating a faster rate of metmyoglobin accumulation in this group when 271 compared to the others (Figure 2b). This is in agreement with the lower values in h* 272 (P<0.05) observed in VITE006 and CARN012 groups (more intense redness) when 273 compared to the CONTROL and CARN006 groups after 14 days of refrigerated storage 274 (Figure 3b). However it must be stated that at the same doses (CARN006 vs. VITE006), 275 carnosic acid showed a less potent effect than vitamin E on colour of GM muscle, and 276 only the highest dose of carnosic acid (CARN012) tended to show similar effects than 277 vitamin E. 278 3.3 pH, ERV and WHC 279 The pH values of the meat samples are summarized in Table 3. As expected, the pH 280 values significantly decreased in all of the groups (P<0.001) after 45 minutes and 24 h 281 post-mortem; however, no significant differences (P>0.05) between the groups were 282 detected at any time. Accordingly, no differences in WHC among the groups due to pH 283 variations were expected. 284 [INSERT TABLES 3 AND 4 NEAR HERE, PLEASE] 12 285 Table 4 summarizes the ERV values throughout the refrigerated storage period. As 286 can be observed, the ERV values became significantly reduced during the storage 287 period (P<0.001) in all the groups, probably due to increases in the bacterial populations 288 (Jay, 1966). However, no differences among groups were detected at any time. 289 Regarding WHC (Figure 4), the lost of water in the CONTROL group (20.4%) was 290 not statistically different from that measured in the VITE006 (17.5%) and CARN012 291 (17.5%) groups after 7 days of refrigerated storage. However, the endogenous proteases 292 (µ-calpain and m-calpain) have been suggested to be protected by antioxidants against 293 post-mortem oxidation,thus allowing their activity (Huff-Lonergan & Lonergan, 2005), 294 so the water expelled from the intramyofibrillar spaces would be kept within the cells 295 thus reducing drip production (Huff-Lonergan & Lonergan, 2005). Consequently, the no 296 significant differences in the present study might have been due to either the lack of 297 treatment effect or the high intra-group variability observed in WHC values. 298 [INSERT FIGURE 4 NEAR HERE, PLEASE] 299 3.4 Microbiological analysis 300 All of the counts of Brochothrix thermosphacta were under the detection level (<1.4 301 log ufc/g, data not shown). The means of Pseudomonas spp., EC, TVB (4.5ºC) and 302 LAB for each storage time in each treatment are shown in Table 5. 303 [INSERT TABLE 5 NEAR HERE, PLEASE] 304 In agreement with the lack of significant differences in ERV (Table 4), 305 microbiological growth for Pseudomonas spp., EC (Gram-negative), LAB (Gram- 306 positive) and TVB (4.5ºC) did not seem to be significantly affected by dietary carnosic 307 or vitamin E supplementation. It must be stressed that Pseudomonas spp. counts have 308 been found to decrease when rosemary extract is directly added to meat (Zhang, Kong, 13 309 Xiong, & Sun, 2009), thus indicating antimicrobial properties of rosemary extract 310 against one of the common meat spoilage bacteria at refrigerated temperatures, i.e. 311 Pseudomonas spp. However, it must be considered that Pseudomonas spp. is among the 312 most sensitive species to CO2 (Jay, Loessner, & Golden, 2005), so the MAP used in the 313 present study might have inhibited this microorganism, thus hindering a possible effect 314 of carnosic acid. In any case, the lack of significant differences in the counts of the rest 315 of microorganisms (EC, LAB and TVB, Table 5) also suggests that other compounds 316 (but not carnosic acid) might be responsible for the antimicrobial properties of rosemary 317 (Pintore, Usai, Bradesi, Juliano, Boatto, Tomi, Chessa, Cerri & Casanova, 2002). 318 3.5 Sensory evaluation 319 When meat from the lambs fed the CONTROL and CARN012 diets were compared, 320 most of the consumers (17 out of 24, 70.8%, P<0.05, Table 6) could not distinguish 321 between the treatments. Moreover, 62.5% of the consumers preferred the CONTROL 322 samples describing them as more tender and juicy, but this percentage was not 323 statistically different from the percentage of consumers who preferred the CARN012 324 samples (P>0.05). When meat from lambs fed the CARN006 and VITE006 diets were 325 compared, only eight correct judgements (33.3%) were given. Therefore, the consumers 326 could not distinguish between these samples (P<0.10, Table 6). Likewise, the 327 CARN006 samples were preferred in 10 out of 24 cases (P>0.05), and the VITE006 328 samples were judged as more tender and tasty. Hence, in both cases, the meat samples 329 from lambs fed carnosic acid (CARN006 and CARN012) were neither clearly 330 distinguished nor preferred by the consumers when they were compared with those 331 from lambs fed either the CONTROL or VITE006 diets. 332 [INSERT TABLE 6 NEAR HERE, PLEASE] 14 333 Conclusion 334 At the doses used in the present study, it can be concluded that carnosic acid shows a 335 lower antioxidant activity when compared to vitamin E but it seems to be useful to 336 protect meat from discolouration after a long period of time under MAP at refrigerated 337 storage, particularly in low colour stable muscles such as gluteus medius. Moreover, 338 carnosic acid does not significantly affect meat sensory traits. The microbiological 339 analyses were not conclusive at the doses administered in the present study. Future 340 experiments should include higher rates of carnosic acid in the diet of fattening lambs to 341 check the potential of this compound in further delaying the microbial spoilage of meat. 342 Acknowledgements 343 Financial support received from ‘Consejería de Educación de la Junta de Castilla y 344 León’ (Project GR158) is gratefully acknowledged. Raúl Bodas and Nuria Prieto have a 345 JAE-Doc contract and Lara Morán was supported by a JAE-Predoc grant from the CSIC 346 under the programme ‘Junta para la Ampliación de Estudios’. We would also like to 347 thank M. J. Jordán from IMIDA (Murcia, Spain) for the carnosic acid quantification by 348 HPLC of the rosemary extract used in the study. M. Y., A. S. C., and D. D. G. designed 349 the research, M. Y. conducted the research, M. Y. and D. D. G. analysed the data and 350 wrote the paper. D. D. G. had primary responsibility for the final content. 351 References 352 AOAC (2003). Official methods of analysis of the Association of Official Analytical 353 Chemists (17th ed.). Gaithersburg, MD: AOAC International. 354 Boleman, S. J., Boleman, S. L., Miller, R. K., Taylor, J. F., Cross, H. R., Wheeler, T. L., 355 et al. (1997). Consumer evaluation of beef of known categories of tenderness. 356 Journal of Animal Science, 75, 1521–1524. 15 357 Botsoglou, N. A., Govaris, A., Giannenas, I., Botsoglou, E., & Papageorgiou, G. (2007). 358 The incorporation of dehydrated rosemary leaves in the rations of turkeys and 359 their impact on the oxidative stability of the produced raw and cooked meat. Int J 360 Food Sci Nutr, 58(4), 312-320. 361 Fernández-López, J., Pérez-Álvarez, J. A., Sayas-Barberá, E., & Aranda-Catalá, V. 362 (2000). Characterization of the different states of myoglobin in pork using color 363 parameters and reflectance ratios. Journal of muscle foods, 11(3), 157-167. 364 Galobart, J., Barroeta, A. C., Baucells, M. D., Codony, R., & Ternes, W. (2001). Effect 365 of dietary supplementation with rosemary extract and alpha-tocopheryl acetate on 366 lipid oxidation in eggs enriched with omega3-fatty acids. Poultry science, 80(4), 367 460-467. 368 Hernández-Hernández, E., Ponce-Alquicira, E., Jaramillo-Flores, M. E., & Legarreta, I. 369 G. (2009). Antioxidant effect rosemary (Rosmarinus officinalis L.) and oregano 370 (Origanum vulgare L.) extracts on TBARS and colour of model raw pork batters. 371 Meat Science, 81(2), 410-417. 372 373 Honikel, K. O. (1998). Reference methods for the assessment of physical characteristics of meat. Meat Science, 49(4), 447-457. 374 Huff-Lonergan, E., & Lonergan, S. M. (2005). Mechanisms of water-holding capacity 375 of meat: The role of postmortem biochemical and structural changes. Meat 376 Science, 71(1), 194-204. 377 ISO 4120:1983. Sensory analysis-methodology-Triangle test. 378 Jay, J. M. (1966). Response of the extract-release volume and water-holding capacity 379 phenomena to microbiologically spoiled beef and aged beef. Applied 380 Microbiology, 14(4), 492-496. 16 381 382 Jay, J. M., Loessner, M. J., & Golden, D. A. (2005). Modern food microbiology (7 ed.). New York: Springer Science + Bisiness Media. 383 Johansen, J., Aastveit, A. H., Egelandsdal, B., Kvaal, K., & Røe, M. (2006). Validation 384 of the EUROP system for lamb classification in Norway; repeatability and 385 accuracy of visual assessment and prediction of lamb carcass composition. Meat 386 Science, 74(3), 497-509. 387 Jordán, M. J., Moñino, M. I., Martínez, C., Lafuente, A., & Sotomayor, J. A. (2010). 388 Introduction of distillate Rosemary leaves into the diet of the Murciano-Granadina 389 goat: Transfer of polyphenolic compounds to goats' milk and the plasma of 390 suckling goat kids. Journal of Agricultural and Food Chemistry, 58(14), 8265- 391 8270. 392 Lan, J., & Jiang, D. H. (1997). Desferrioxamine and vitamin E protect against iron and 393 MPTP-induced neurodegeneration in mice. Journal of Neural Transmission, 394 104(4), 469-481. 395 Linares, M. B., Bórnez, R., & Vergara, H. (2008). Effect of stunning systems on meat 396 quality of Manchego suckling lamb packed under modified atmospheres. Meat 397 Science, 78(3), 279-287. 398 López-Bote, C. J., Daza, A., Soares, M., & Berges, E. (2001). Dose-response effect of 399 dietary vitamin E concentration on meat quality characteristics in light-weight 400 lambs. Animal Science, 73(3), 451-458. 401 McBride, N., Hogan, S. A., & Kerry, J. P. (2007). Comparative addition of rosemary 402 extract and additives on sensory and antioxidant properties of retail packaged 403 beef. International Journal of Food Science & Technology, 42(10), 1201-1207. 17 404 McKenna, D. R., Mies, P. D., Baird, B. E., Pfeiffer, K. D., Ellebracht, J. W., & Savell, 405 J. W. (2005). Biochemical and physical factors affecting discoloration 406 characteristics of 19 bovine muscles. Meat Science, 70(4), 665-682. 407 Moñino, I., Martínez, C., Sotomayor, J., Lafuente, A., & Jordán, M. (2008). 408 Polyphenolic transmission to Segureno lamb meat from ewes' diet supplemented 409 with the distillate from rosemary (Rosmarinus officinalis) leaves. Journal of 410 Agricultural and Food Chemistry, 56(9), 3363-3367. 411 Munné-Bosch, S., Mueller, M., Schwarz, K., & Alegre, L. (2001). Diterpenes and 412 antioxidative protection in drought-stressed Salvia officinalis plants. Journal of 413 Plant Physiology, 158(11), 1431-1437. 414 Nieto, G., Díaz, P., Bañón, S. L., & Garrido, M. D. (2010). Dietary administration of 415 ewe diets with a distillate from rosemary leaves (Rosmarinus officinalis L.): 416 Influence on lamb meat quality. Meat Science, 84(1), 23-29. 417 Pintore, G., Usai, M., Bradesi, P., Juliano, C., Boatto, G., Tomi, F., Chessa, M., Cerri, 418 R., Casanova, J. (2002). Chemical composition and antimicrobial activity of 419 Rosmarinus officinalis L. oil from Sardinia and Corsica. Flavour and Fragance 420 Journal, 17(1), 15-19. 421 Rodríguez-Calleja, J. M., García-López, M. L., Santos, J. A., & Otero, A. (2005). 422 Development of the aerobic spoilage flora of chilled rabbit meat. Meat Science, 423 70(2), 389-394. 424 Rodriguez-Calleja, J. M., Santos, J. A., Otero, A., & García-López, M. L. (2004). 425 Microbiological quality of rabbit meat. Journal of Food Protection, 67(5), 966- 426 971. 18 427 Samman, S., Sandstrom, B., Toft, M.B., Bukhave, K, Jensen, M., Sorensen, S. S. & 428 Hansen, M. (2001). Green tea or rosemary extract added to foods reduces 429 nonheme-iron absorption. The American Journal of Clinical Nutrition, 73(3), 607- 430 612. 431 432 433 434 SAS, 1999 SAS, SAS/STAT(R) user’s guide (version 8), previous SAS term Publishing, Cary, NC, USA (1999). Stone, H., & Sidel, J. L. (1993). Sensory evaluation practices. San Diego, CA, Academic Press. Talmant. 435 Young, O. A., & West, J. (2001). Meat color. In Y. H. Huy, W. Nip, R. W. Rogers & O. 436 A. Young, Meat science and applications (pp. 36-69). New York: Mercel Dekker 437 Inc. 438 Zhang, H., Kong, B., Xiong, Y. L., & Sun, X. (2009). Antimicrobial activities of spice 439 extracts against pathogenic and spoilage bacteria in modified atmosphere 440 packaged fresh pork and vacuum packaged ham slices stored at 4 °C. Meat 441 Science, 81(4), 686-692. 442 443 19 444 Table 1. Chemical composition of longissimus thoracis muscle (g kg-1 meat) CONTROL CARN006 CARN012 VITE006 P-value Ash 11.3 ± 0.28 11.9 ± 0.65 11.9 ± 0.81 12.0 ± 0.52 0.080 CP 196.8ab ± 4.28 201.2a ± 7.54 201.7a ± 4.56 194.3b ± 3.27 0.020 EE 23.8 ± 6.26 26.3 ± 4.51 24.7 ± 7.86 26.6 ± 4.30 0.753 DM 233.5 ± 6.34 239.5 ± 7.67 238.4 ± 9.50 236.1 ± 4.46 0.373 445 446 447 448 449 450 Mean ± standard deviation; CONTROL group (no antioxidants); CARN006 (0.6 g carnosic acid kg-1 of feed concentrate); CARN012 (1.2 g carnosic acid kg-1 of feed concentrate); VITE006 (0.6 g α-tocopheryl acetate kg-1 of feed concentrate); CP, crude protein; EE, ether extract; DM, dry matter; a, b, c: different superscripts in the same row indicate statistical differences (P<0.05) among treatments 20 451 Table 2. Colour parameters for each treatment and storage day at 4oC under modified atmosphere package (35% CO2, 35% O2 and 30% N2) 452 21 longissimus lumborum TREATMENT DAY P-value gluteus medius L* a* b* L a* b* CONTROL 39.3b ± 1.63 10.3 ± 1.41 9.9a ± 1.08 41.4 ± 1.65 9.8 ± 1.50 10.8ª ± 0.85 CARN006 40.6ª ± 2.19 10.2 ± 1.45 10.3b ± 1.21 41.7 ± 1.43 9.9 ± 1.45 11.0a ± 1.12 CARN012 40.3ª ± 1.50 10.2 ± 1.30 9.8a ± 1.18 41.4 ± 1.71 10.1 ± 1.28 10.6ª ± 0.94 VITE006 40.5ª ± 2.08 10.2 ± 0.85 9.7a ± 1.25 41.3 ± 1.48 10.1 ± 0.87 10.2b ± 1.09 0 39.9 ± 2.27 9.9b ± 1.07 8.6b ± 1.14 41.1 ± 1.70 10.3ª ± 1.31 9.6b ± 1.09 3 40.4 ± 1.83 10.7ª ± 1.17 10.2ª ± 0.56 41.8 ± 1.36 10.3ª ± 1.07 10.9ª ± 0.93 7 40.4 ± 1.55 10.8ª ± 0.85 10.5ª ± 0.60 41.4 ± 1.58 10.5ª ± 0.78 10.9ª ± 0.78 9 40.3 ± 1.89 10.4ab ± 0.96 10.2ª ± 0.71 41.5 ± 1.39 10.0a ± 0.82 11.0a ± 0.59 14 40.0 ± 2.07 9.2c ± 1.55 10.1a ± 1.60 41.5 ± 1.77 8.9b ± 1.67 10.9a ± 1.09 Treatment 0.014 0.956 0.021 0.720 0.542 0.001 Day 0.821 <0.0001 <0.0001 0.585 <0.0001 <0.0001 Treatment*Day 0.999 0.404 0.293 0.432 0.153 0.940 22 453 454 455 Mean ± standard deviation; CONTROL group (no antioxidants); CARN006 (0.6 g carnosic acid kg-1 of feed concentrate); CARN012 (1.2 g carnosic acid kg-1 of feed concentrate); VITE006 (0.6 g α-tocopheryl acetate kg-1 of feed concentrate); a, b, c: different superscripts in the same column indicate statistical differences (P<0.05) among treatments or days 456 457 23 458 Table 3. pH values at 0 h, 45 min and 24 h post- slaughter in longissimus thoracis muscle 459 460 461 462 463 pH 0 h pH 45 minutes pH 24 h CONTROL 6.79a ± 0.198 6.43b ± 0.159 5.84c ± 0.040 CARN006 6.87a ± 0.178 6.49b ± 0.192 5.76c ± 0.061 CARN012 6.74a ± 0.371 6.47b ± 0.211 5.82c ± 0.088 VITE006 6.77a ± 0.308 6.55b ± 0.139 5.83c ± 0.060 464 465 466 467 468 Mean ± standard deviation; CONTROL group (no antioxidants); CARN006 (0.6 g carnosic acid kg-1 of feed concentrate); CARN012 (1.2 g carnosic acid kg-1 of feed concentrate); VITE006 (0.6 g α-tocopheryl acetate kg-1 of feed concentrate); a, b, c: different superscripts in the same row indicate statistical differences (P<0.05) measuring times 469 470 24 471 472 Table 4. Extract release volume (ERV, ml) of lamb meat samples from longissimus thoracis muscle stored at 4 ºC under modified atmosphere package (35% CO2, 35% O2 and 30% N2) during 14 days. 0 days 7 days CONTROL 36.5a ± 3.42 26.4b ± 7.65 CARN006 31.5a ± 5.21 23.4b ± 6.72 CARN012 35.6a ± 8.72 21.1b ± 4.49 VITE006 33.1a ± 10.20 21.9b ± 2.75 473 474 18.6c ± 3.66475 476 18.3b ± 3.99477 478 18.3b ± 4.17 479 14 days 16.6b ± 2.83 480 481 482 25 Mean ± standard deviation; CONTROL group (no antioxidants); CARN006 (0.6 g carnosic acid kg-1 of feed concentrate); CARN012 (1.2 g carnosic acid kg-1 of feed concentrate); VITE006 (0.6 g α-tocopheryl acetate kg-1 of feed concentrate); a, b, c: different superscripts in the same row indicate statistical differences (P<0.05) among measuring times 483 484 Table 5. Effects of carnosic acid and vitamin E on microbial populations in lamb meat samples from longissimus thoracis muscle stored at 4 ºC under modified atmosphere package (35% CO2, 35% O2 and 30% N2) during 14 days 485 Day 0 Day 7 Day 14 486 CONTROL CARN006 CARN012 VITE006 2.69 ± 0.905 2.02 ± 1.110 1.54a ± 0.660 1.29a ± 0.681 3.05 ± 1.053 2.97 ± 0.472 2.60ab ± 0.521 2.59b ± 0.683 4.40 ± 1.403 487 2.97 ± 1.417 3.66b ± 0.999488 3.64b ± 0.830 EC CONTROL CARN006 CARN012 VITE006 1.04 ± 0.614 1.18 ± 0.512 1.42ab ± 0.688 1.09a ± 0.269 1.35 ± 0.296 1.05 ± 0.287 1.01a ± 0.380 1.21a ± 0.387 2.06 ± 1.030 490 1.78 ± 0.409 491 2.10b ± 0.378 2.23b ± 0.440492 TVB at 4.5ºC CONTROL CARN006 CARN012 VITE006 2.20a ± 1.137 1.97 ± 0.959 1.81a ± 0.845 1.66 ± 0.893 2.54b ± 0.795 2.70 ± 0.477 2.53ab ± 0.611 2.17 ± 0.506 3.20b ± 1.797 3.47 ± 1.815 494 4.03b ± 1.446 495 3.60 ± 1.123 Pseudomonas 489 493 496 LAB CONTROL CARN006 CARN012 VITE006 2.74 ± 1.227 2.54 ± 0.900 1.89 ± 0.410 1.71 ± 0.279 2.74 ± 1.233 2.34 ± 0.707 2.39 ± 0.390 1.97 ± 0.839 2.99 ± 2.202 497 3.05 ± 1.261 3.27 ± 1.951 498 2.88 ± 2.212 26 499 500 501 502 503 Mean ± standard deviation; CONTROL group (no antioxidants); CARN006 (0.6 g carnosic acid kg-1 of feed concentrate); CARN012 (1.2 g carnosic acid kg-1 of feed concentrate); VITE006 (0.6 g α-tocopheryl acetate kg-1 of feed concentrate); a, b, c: different superscripts in the same row indicate statistical differences (P<0.05) among refrigerated storage days; Enterobacteriaceae (EC), total viable bacteria (TVB), lactic acid bacteria (LAB) 504 27 505 506 507 508 509 Table 6. Results from sensory analysis carried out to discriminate between cooked meat samples from lambs fed 1) CONTROL and CARN012 and 2) CARN006 and VITE006 Triangle test Correct judgements (n = 24) P-value CONTROL vs. CARN012 7 0.041 CARN006 vs. VITE006 8 0.100 Preference test Preferences for B (n=24) P-value CONTROL (A) vs. CARN012 (B) 9 0.221 VITE006 (A) vs. CARN006 (B) 10 0.414 CONTROL group (no antioxidants); CARN006 (0.6 g carnosic acid kg-1 of feed concentrate); CARN012 (1.2 g carnosic acid kg-1 of feed concentrate); VITE006 (0.6 g α-tocopheryl acetate kg-1 of feed concentrate) 510 28 511 512 Figure 1. Evolution of L* parameter in longissimus lumborum and gluteus medius muscles stored at 4oC under modified atmosphere package (35% CO2, 35% O2 and 30% N2) 513 514 515 Figure 2. Evolution of chroma (C*) in longissimus lumborum and gluteus medius muscles stored at 4oC under modified atmosphere package (35% CO2, 35% O2 and 30% N2) 516 517 518 Figure 3. Evolution of hue (h*) in longissimus lumborum and gluteus medius muscles stored at 4oC under modified atmosphere package (35% CO2, 35% O2 and 30% N2) 519 520 521 Figure 4. Evolution of water holding capacity (WHC) in longissimus lumborum muscle stored at 4ºC under modified atmosphere package (35% CO2, 35% O2 and 30% N2) 522 523 29 524 Figure 1. 525 526 (a) longissimus lumborum 527 528 529 530 531 532 533 534 535 536 537 (b) gluteus medius 538 539 540 541 542 543 544 545 CONTROL group (no antioxidants); CARN006 (0.6 g carnosic acid kg-1 of feed concentrate); CARN012 (1.2 g carnosic acid kg-1 of feed concentrate); VITE006 30 (0.6 g α-tocopheryl acetate kg-1 of feed concentrate) CONTROL group (no antioxidants); CARN006 (0.6 g carnosic acid kg-1 of feed concentrate); CARN012 (1.2 g carnosic acid kg-1 of feed concentrate); VITE006 546 547 Figure 2. (a) longissimus lumborum 548 549 550 551 552 553 554 555 556 557 558 (b) gluteus medius 559 560 561 562 563 564 565 566 567 568 569 570 571 572 31 573 574 Figure 3. (a) longissimus lumborum 575 576 577 578 579 580 581 582 583 584 585 586 (b) gluteus medius 587 588 589 590 591 592 593 594 595 596 597 598 599 32 600 Figure 4. 601 602 603 604 605 606 607 608 609 33