Evaluation Of Boar Tain Using Fast And On-line Measurement Technique

Increasing public concern for animal welfare has led to European pork production stakeholders declaring a ban on surgical castration of male piglets from 2018.The ban of castration is linked to the risk of impaired consumer acceptance of pork as a result of off-flavors what is referred to as boar taint which is primarily caused by the accumulation of androstenone(5a-androst-16-en-3-one)and skatole(3-methylindole).Currently no rapid technical method is available for routine analyses of boar taint in or at the slaughter line and that copes with industrial speed constraints at modern abattoirs slaughter up to 1200 pigs per hour.Sensory evaluation is the only way can be performed in-line or at-line by trained assessors.However,even trained experts,are limited in their capability to identify components of odor mixtures from a chemosensory background.Noresearch on boar taint detection could be found in China.Starting with evaluating the performance of sensory analysis,we found disagreement between sensory and chemical analysis.We investigated the possibility of using spectral methods(Raman and NIR)to detect boar taint.Relationship between fatty acids and boar taint,and the distribution of boar taint in fat layers were also studied.The aim of this thesis was to provide theory basis and technique for boar taint evaluation at slaughter line of entire male pigs.The contents and results are as follows:1.Comparisonbetween sensory evaluation and chemical analysis of boar taintThe aim of this chapterwas to compare sensory evaluation by trained assessors(n=10)with chemical analysis of two compounds(androstenone and skatole)in backfat.Backfat samples were randomly obtained from slaughterline for sensory evaluation and chemical analysis.Chemical analysis result was taken as "golden standard" to assess the performance of sensory evaluation.Two different threshold scenarios for chemical analysis were androstenone ?1.5 ?g/g melted backfat and skatole ?0.2?g/g backfat or androstenone?2.0?g/g and skatole?0.25?g/g.For sensory evaluation,the average score of ten assessors' ?2 or ?3(scale from 0-5)was taken as tainted samples.Uncertainty range which was defined as ±2 standard deviations of the measurement result was also considered in our study.The diagnose accuracy of the test method was evaluated by sensitivity and specificity.Results showed that best sensitivity and specificity(67.9%and 94.9%)were obtained when higher threshold for chemical(androstenone>2.0?g/g and skatole? 0.25?g/g)and sensory(average score?3)analysis were chosen.Excluded uncertainty range,sensitivity and specificity were improved to 74.3%and 96.3%.Disagreement between sensory evaluation and chemical analysis always exist no matter which threshold was used.2.Relationship between fatty acids and boar taint including sensory and chemical analysis resultThe aim of this chapter was to analyze the relationship between fatty acids and boar taint,and also analyze the effects of the fatty acid composition on the disagreement between sensory evaluation and chemical analysis of boar taint.The 143 samples were selected for investigation according to their sensory evaluation conducted with a 10-person sensory panel,and the chemical analysis(GC-MS)of androstenone and skatole.A GC-FID quantification of fatty acid was conducted.The correlation analysis revealed that several fatty acids weresignificantly correlated with androstenone and skatole.C10:0 and C24:0 were significantly correlated to average sensory score which improved the regression coefficient when they were added to the model using A/S to predict averaged sensory score.SFA was positively correlated with skatole.With the increase of androstenone,MUFA decreased significantly while PUFA increased significantly.PCA indicated that fatty acids couldbe used to classify samples according to their chemical analysis result or raising locations.However,PCA demonstrated no significant effects of the fatty acid composition on the disagreement between chemical analysis and sensory evaluation of boar taint.3.Distributionof boar taint and fatty aicds in fat inner and outer layersThe aim of this chapterwas to assess the distributionof boar taint and fatty acids in fat inner layer(IL)and outer layer(OL).Samples were selected according to their boar taint level including balanced number of BH,HA,HS and BL.Samples were divided into inner layer and outer layer for boar taint and fatty acids analysis.Difference analysis was conducted with ANOVA was conducted for boar taint and fatty acids in inner and outer layer.Correlation analysis was used to evaluate relationship between boar taint and fatty acids in IL and OL.Results indicated that no significant difference of boar taint between IL and OL.IL was more saturated than OL,but hadless MUFA.Better correlation between fatty acids and boar taint in inner layer was found.Difference of SFA,MUFA and PUFA among BH,HA,HS,and BL in IL wasbigger than OL.4.Detection of boar taint using Raman spectroscopy and NIR spectroscopyThe chapter was to explore whether Raman spectroscopy can be used to detect boar taint.The backfat samples were collected randomly used for chemical analysis and Raman analysis of androstenone and skatole.Different preprocessing methods were used on spectral optimization.PLSR was used to quantify boar taint.Results showed that the highest R2cv for androstenone and skatole were 0.57 and 0.45,together with RPD were 1.10 and 0.99 which imply Raman can not be used to the quantization of androstenone and skatole.Then chemical result was converted into binary variables(0=no boar taint/untainted;1=boar tainted),then PLS-DA was used to qualify boar taint.Classify accuracy was 68.5%which can not satisfy on-line using.Uncertainty of chemical analysis and unbalanced distribution of samples are possible reasons contributing to the undesirable result.Then we used backfat samples which were selected according to their levels of androstenone and skatole as determined by gas chromatography and their sensory score by a trained panel.A model using only spectra obtained at the inner layer resulted in the highest classification accuracy for boar taint(81%of samples correctly classified).The discrimination was shown to reflect differences in the degree of fatty acid saturation between tainted and untainted boars.In conclusion,the findings suggest that with further development Raman spectroscopy may be used to classify boar taint.But the findings indicated that the application of NIR to classify boar taint is limited since only 68.3%of samples can be correctly classified...