Manuscript On-line prediction of fatty acid profiles in crossbred Limousin 1 and Aberdeen Angus beef cattle using near infrared 2 reflectance spectroscopy 3 4 Running title Prediction of meat fatty acid composition by NIRS 5 N. Prieto1†, D. W. Ross1, E. A. Navajas1, R. I. Richardson2, J. J. Hyslop3, G. Simm1 6 and R. Roehe1 7 8 1Sustainable Livestock Systems Group, Scottish Agricultural College, West Mains Road, 9 Edinburgh EH9 3JG, UK. 10 2University of Bristol, Division of Farm Animal Science, Langford, Bristol, BS40 5DU, 11 UK. 12 3Select Services, Scottish Agricultural College, West Mains Road, Edinburgh EH9 3JG, 13 UK. 14 15 16 17 †CORRESPONDING AUTHOR: Nuria Prieto. Scottish Agricultural College (SAC), 18 Bush Estate, Edinburgh EH26 0PH, UK. Tel.: +44 131 535 3361, fax: +44 131 535 19 3121. E-mail: [email protected]; [email protected] 1 20 Abstract 21 The objective of this study was to examine the on-line use of near infrared 22 reflectance (NIR) spectroscopy to estimate the concentration of individual and groups of 23 fatty acids (FA) as well as intramuscular fat (IMF) in crossbred Aberdeen Angus (AAx) 24 and Limousin (LIMx) cattle. This was achieved by direct application of a fibre-optic 25 probe to the muscle immediately after exposing the meat surface in the abattoir at 48 h 26 post mortem. Samples of M. longissimus thoracis from 88 AAx and 106 LIMx were 27 scanned over the NIR spectral range from 350-1800 nm and samples of the M. 28 longissimus lumborum were analyzed for IMF content and fatty acid composition. 29 Statistically significant differences (P<0.001) were observed in most FA between the 30 two breeds studied, with FA concentration being higher in AAx meat mainly. NIR 31 calibrations, tested by cross-validation, showed moderate to high predictability in LIMx 32 meat samples for C16:0, C16:1, C18:0, trans11 C18:1, C18:1, C18:2 n-6, C20:1, cis9, 33 trans11 C18:2, SFA (saturated FA), MUFA (monounsaturated FA), PUFA 34 (polyunsaturated FA) and IMF content with R2 (SE mg.100 g-1 muscle) of 0.69 (146), CV, 35 0.69 (28), 0.71 (62), 0.70 (8.1), 0.76 (192), 0.65 (13), 0.71 (0.9), 0.71 (2.9), 0.68 (235), 36 0.75 (240), 0.64 (17) and 0.75 (477), respectively. FA such as C14:0, C18:3 n-3, C20:4 37 n-6, C20:5 n-3, C22:6 n-3, n-6 and n-3 were more difficult to predict by NIR in these 38 LIMx samples (R2 = 0.12 to 0.62; SE = 0.5 to 26mg.100 g-1 muscle). In contrast, NIR CV 39 showed low predictability for FA in AAx beef samples. In particular for LIMx, the 40 correlations of NIR measurements and several FA in the range from 0.81 to 0.87 41 indicated that the NIR spectroscopy is a useful on-line technique for the early, fast and 42 relatively inexpensive estimation of FA composition in the abattoir. 43 Keywords: near infrared reflectance spectroscopy, fibre-optic probe, beef, fatty acid, 44 intramuscular fat 2 45 Implications 46 It is widely accepted that the amount and type of fat in meat influence two major 47 components of meat quality notably tenderness and flavour. Additionally, change in 48 fatty acid profiles of meat can reduce the risk of cardiovascular diseases, cancer, and 49 diabetes. Our study shows the ability of NIR spectroscopy to accurately predict fatty 50 acid profile of individual carcasses under abattoir conditions. This information could 51 help to develop a value-based marketing system for meat quality and be used in 52 breeding programmes to genetically improve fatty acids profiles of meat. 53 Introduction 54 The amount and proportion of fatty acids (FA) in beef intramuscular fat (IMF) are 55 key factors that influence technological and sensory meat quality, especially shelf life 56 (lipid and pigment oxidation) and flavour (Elmore et al., 1999; Vatansever et al., 2000; 57 Wood et al., 2003). Individual FA have very different melting points which affect the 58 firmness or softness of the fat in meat, especially the subcutaneous and intermuscular 59 fat, but also the intramuscular fat. Groups of fat cells containing solidified fat with a 60 high melting point appear whiter than liquid fat with a lower melting point, such that fat 61 colour is another aspect of lean meat quality affected by FA composition. Furthermore, 62 the ability of unsaturated FA to rapidly oxidise, especially those containing more than 63 two double bonds, influences the rancidity and colour deterioration of meat and the 64 flavour development during cooking (Wood et al., 2003). 65 Consumers are interested in the fat composition of meat, as scientific evidence 66 suggests that diets high in saturated fat are associated with increased levels of blood 67 total and LDL-cholesterol, which results in increased risk of cardiovascular diseases 68 (Webb and O'Neill, 2008). Coronary heart disease is one of the major public health 3 69 concerns, primarily because it accounts for more deaths than any other disease or group 70 of diseases. Thus, a lower saturated FA (SFA) intake and a higher polyunsaturated FA 71 (PUFA) intake, especially of n-3 FA for an appropriate n-6/n-3 balance, are 72 recommended in order to avoid cardiovascular-type diseases. The consumption of beef 73 in human diets also supplies conjugated linoleic acids (CLA), which are a group of 74 positional and geometrical isomers that are associated with beneficial health properties 75 such as reducing the risk for cancer, atherosclerosis and diabetes (Enser, 2001; Rainer 76 and Heiss, 2004). Additionally, the amount and composition of ruminant IMF depends 77 on factors such as diet (Gatellier et al., 2005; Dannenberger et al., 2007), age or live 78 weight (Okeudo and Moss, 2007) and the genetic origin of the animals (Raes et al., 79 2001), which also explains the increasing interest in defining the FA profile of beef 80 from different breeds. 81 Quantitative chemical techniques for the determination of FA involve extraction of 82 total lipids with diethyl ether, followed by conversion of the fatty acids to their methyl 83 esters and then analysis by capillary gas chromatography, a costly and time-consuming 84 process. 85 Near infrared reflectance (NIR) spectroscopy is a rapid and non destructive method, 86 neither requiring reagents nor producing waste (Osborne et al., 1993; Prieto et al., 87 2009a). Because of these advantages, this technology is being broadly used by the 88 industry research units for large-scale meat quality evaluation to predict the chemical 89 composition (Tøgersen et al., 1999; Cozzolino and Murray, 2002; Alomar et al., 2003; 90 Prieto et al., 2006) as well as physical and sensory characteristics of meat (Park et al., 91 1998; Shackelford et al., 2005; Andrés et al., 2007; Prieto et al., 2008; Prieto et al., 92 2009b). The structure of FA can produce individual spectral characteristics and they are, 93 therefore, very accessible for detection by near infrared spectroscopy (González-Martín 4 94 et al., 2002). Hence, NIR spectroscopy has been applied to study the FA content of 95 Iberian pig fat (De Pedro et al., 1992; García-Olmo et al., 2001), intact pork loins 96 (González-Martín et al., 2005) and ground beef (Windham and Morrison, 1998; Realini 97 et al., 2004; Sierra et al., 2008). 98 As part of a much wider meat eating quality project, this study examined the use of 99 NIR spectroscopy for the on-line estimation of FA (individual and groups) composition 100 and IMF content of beef from Aberdeen Angus and Limousin crossbred cattle, by direct 101 application of a fibre-optic probe to the M. longissimus thoracis immediately after 102 exposing the meat surface in the abattoir. The animals used in the analysis were mostly 103 from our experimental farm where the diet was the same for all animals to detect 104 genetic differences between breeds independent from the diet. 105 Material and methods 106 Animals and meat samples 107 Data were collected on 194 crossbred steers and heifers, whereby 88 and 106 were 108 sired by either Aberdeen Angus (AA) or Limousin (LIM) sires respectively. A total of 109 144 of these animals were obtained from the Beef Research Centre (BRC) situated at 110 the Scottish Agricultural College, Edinburgh. All these 144 animals from the BRC were 111 finishing during the final 2-4 months of their production cycle on similar diets 112 consisting of first cut grass silage and a barley based concentrate (50:50 on a dry matter 113 basis) which was offered ad libitum as a completely mixed ration on a daily basis. The 114 ration analysis averaged 381 g.kg-1 dry matter (DM), 12.0 MJ.kg-1 DM metabolisable 115 energy and 139 g.kg-1 DM crude protein. All animals remained on these diets for a 116 minimum of eight weeks after which they were selected for slaughter according to 117 standard commercial practice (target grades R4L or better). A further 50 crossbred 5 118 Aberdeen Angus (AAx) and Limousin (LIMx) steers and heifers were selected at the 119 abattoir from commercial farms where, although the ration formulation is not known, 120 their ages and slaughter dates suggest that their finishing management was likely to be 121 similar to that at the BRC. All 194 animals were slaughtered in 11 batches from autumn 122 2006 until late winter 2008 (batches 1 to 3, 4 to 8, and 9 to 11 during the autumn and 123 winter of 2006, 2007 and 2008, respectively) where the average live weight was 582 124 and 609 kg and age at slaughter was 546 and 544 days for AAx and LIMx sired beef 125 cattle respectively. 126 The left sides of the carcasses were cut at the 13th rib at 48 h post mortem. NIR 127 measurements were taken on the caudal cut surface of the M. longissimus thoracis. 128 After removing a 125 mm section, the next 25 mm of the M. longissimus lumborum was 129 taken, vacuum packed and frozen for subsequent analysis of FA composition. 130 Fatty acid and intramuscular fat analysis 131 Fatty acids analysis was carried out by direct saponification as described in detail by 132 Teye et al. (2006). Samples were hydrolysed with 2M KOH in water:methanol (1:1) and 133 the FAs extracted into petroleum spirit, methylated using diazomethane and analysed by 134 gas liquid chromatography. Samples were injected in the split mode, 70:1, onto a CP Sil 135 88, 50 m  0.25 mm fatty acid methyl esters (FAME) column (Chrompack UK Ltd, 136 London) with helium as the carrier gas. The output from the flame ionization detector 137 was quantified using a computing integrator (Spectra Physics 4270) and linearity of the 138 system was tested using saturated (FAME4) and monounsaturated (FAME5) methyl 139 ester quantitative standards (Thames Restek UK Ltd, Windsor, UK). All measurements 140 of FA were performed in duplicate, the error between replications being usually 1-2% 141 with a maximum allowance of 5% error. Total IMF content was calculated 142 gravimetrically as total weight of FA extracted. 6 143 Spectra collection 144 NIR measurements were taken at 48 h post mortem by placing the active area 145 scanning head, 63.5 mm in diameter, over the surface of the exposed M. longissimus 146 thoracis and recording a spectrum from 350 to 1800 nm, by means of a NIR 147 spectrophotometer (Qualityspec Pro, ASD Inc., Boulder, Colorado). Twenty replicate 148 measurements were taken by moving and rotating the scanning head around the muscle 149 surface. The sampling error of measurements is reduced due to the large area measured 150 by the NIR head (Downey and Hildrum, 2004). The scanning head incorporated a 151 broad-band light source for tissue illumination and a sampling fibre optic probe that 152 passed the reflected light back to the spectrometer. The spectrometer interpolated the 153 data to produce measurements in one nm steps, resulting in a diffuse reflectance 154 spectrum of 1451 data points. Absorbance data were stored as log (1/R), where R is the 155 reflectance. The twenty reflectance spectra of each sample were visually examined for 156 consistency and then averaged. The instrument was operated by the proprietary software 157 package Indico Pro (ASD Inc., Boulder, Colorado). 158 Data analysis 159 The effect of breed cross (AAx or LIMx) on FA composition measured by chemical 160 analysis was estimated using the General Linear Models (GLM) procedure of the SAS 161 package (SAS, 2003). Calibration and validation of the NIR data were performed using 162 The UNSCRAMBLER program (version 8.5.0, Camo, Trondheim, Norway). The 163 detection of anomalous spectra was accomplished using the Mahalanobis distance (H- 164 statistic) to the centre of the population, which indicates how different a sample 165 spectrum is from the average spectrum of the set (Williams and Norris, 2001). A sample 166 with an H statistic of ≥ 3.0 standardized units from the mean spectrum was defined as a 167 global H outlier and was eliminated from the population. In addition, some samples 7 168 were removed from the initial data set as concentration outliers (T-statistic), which 169 measures how closely the reference value matches the predicted value. Hence, the 170 samples whose predicted values exceed 2.5 times the standard error of estimation were 171 considered as T statistic outliers and excluded from the population. Because NIR spectra 172 are affected by particle size, light scatter and path-length variation, pre-treatments of the 173 spectral data improve the accuracy of calibration. Thus, spectral data pre-treatments 174 such as Multiplicative Scatter Correction (MSC; Dhanoa et al., 1994) and first or second 175 order derivatives (Shenk et al., 1992) were applied to the spectra. MSC reduces 176 multicolinearity and the confusing effects of baseline shift and curvature of the spectra 177 arising from scattering effects because of physical effects. The use of first and second- 178 order derivatives increase the resolution of spectrum peaks, hence highlighting the 179 signals related to the chemical composition of the meat samples (Davies and Grant, 180 1987). Partial least square regression type I (PLSR1) was used for predicting FA 181 concentration using NIR spectra as independent variables. Internal full cross-validation 182 was performed in order to avoid over-fitting the PLSR equations. Thus, the optimal 183 number of factors in each equation was determined as the number of factors after which 184 the standard error of cross-validation no longer decreased substantially. The accuracy of 185 prediction was evaluated in terms of coefficient of determination (R2) and standard error 186 of cross-validation (SE ) (Westerhaus et al., 2004). CV 187 Results and discussion 188 Chemical data 189 Ranges, means, standard deviations (SD) and coefficients of variation (CV) of the 190 FA profile (individual and groups) and IMF content of AAx and LIMx muscle samples 191 are summarized in Table 1. Most authors express FA data on a percentage basis 192 (percentage of the total FA) when used for prediction by NIR spectroscopy. Only Sierra 8 193 et al. (2008) used reference data expressed in absolute concentrations (weight of 194 muscle) in order to perform the NIR calibrations. In this study, FA were also expressed 195 as concentration (mg) in the muscle because in preliminary analyses higher accuracies 196 of prediction were obtained in comparison with FA expressed of total fat. The 197 advantage of using the amount of FA can be expected because the NIR absorbance is 198 based on the amount of molecular bonds in the organic matrix, so correlating the NIR 199 data with FA expressed in absolute concentrations should be more accurate than on a 200 percentage basis. The latter may be equivalent to the ratio of amount of molecular bonds 201 of the FA to the total molecular bonds of all FA. 202 In general terms, the values of concentration for individual and groups of FA as well 203 as the IMF content in both AAx and LIMx beef samples were within the normal range 204 of variation reported by other authors (Enser et al., 1996 and 1998; Sierra et al., 2008). 205 Statistically significant differences (P<0.001) were observed in most individual and 206 groups of FA and IMF content between the two breed crosses studied (Table 1), 207 whereby FA concentration was mainly higher in AAx beef samples, which would be a 208 consequence of these animals having greater total IMF. Nonsignificant differences 209 between breed crosses were found for linoleic, arachidonic, eicosapentaenoic (EPA), 210 docosahexaenoic (DHA), PUFA and n-6 FA, since on a standard diet the variability in 211 supply of these FA would be small and many PUFA, especially the longer chain PUFA, 212 are in the phospholipids fraction which does not vary much as the animal gets fatter. 213 Most individual FA showed a broad range of concentration in both the AAx and LIMx 214 sample populations. In general, those FA with a higher range of concentration were 215 those with a higher presence in meat. Regarding FA groups, SFA and MUFA 216 (monounsaturated FA) showed a high variability of concentration (see Table 1), 217 probably due to the high heterogeneity of total IMF content of the meat samples 9 218 included in this study (ranging from 1186 to 6618 and 774 to 5405 mg FA.100 g-1 219 muscle for AAx and LIMx beef samples, respectively, Table 1). However, PUFA, n-6 220 and n-3 FA showed narrower ranges of concentration, which could be because PUFA 221 are mainly located in membrane phospholipids, strictly controlled by a complex system 222 of enzymes and relatively constant between individuals (Scollan et al., 2006). As 223 animals mature and deposit more fat, the relative proportion of PUFA decreases 224 (Warren et al., 2008). When observing the variability among samples, most individual 225 and groups of FA and IMF content had coefficients of variation (CV) higher than 20% 226 in both AAx and LIMx samples, except for arachidonic, PUFA and n-6 FA which 227 showed CV in the range from 9 to 18% in both breed crosses. 228 The ratios polyunsaturated:saturated (PUFA:SFA) and n-6:n-3 fatty acids are often 229 used to evaluate the nutritional quality of fat. The recommendations of the Department 230 of Health and Social Security UK (1994) are that the ratio n-6:n-3 is less than 4.0 and 231 that the ratio PUFA:SFA is 0.45 or higher. In our study, the ratio n-6:n-3 was lower than 232 4 for samples from both breed crosses (2.8 and 3.1 for AAx and LIMx samples, 233 respectively) but the ratio PUFA:SFA was lower than 0.45 (0.1 and 0.2 for AAx and 234 LIMx samples, respectively). A ratio of PUFA:SFA of around 0.1 has been widely 235 found in meat of ruminants (Enser et al., 1996 and 1998) as microbial biohydrogenation 236 of unsaturated FA in the rumen produces a high proportion of SFA (Scollan et al., 237 2006). Hence, meat has been implicated in causing the imbalanced fatty acid intake of 238 today´s consumers and, for this reason, ways to improve the PUFA:SFA ratio during 239 meat production have been investigated in many studies (Sañudo et al., 2000; Sheard et 240 al., 2000; Scollan et al., 2001; Kouba et al., 2003). 241 [Table 1 near here, please] 242 Spectral information 10