Use this URL to cite or link to this record in EThOS:
Title: Use of computed tomography based predictors of meat quality in sheep breeding programmes
Author: Clelland, Neil
ISNI:       0000 0004 6059 4648
Awarding Body: University of Edinburgh
Current Institution: University of Edinburgh
Date of Award: 2016
Availability of Full Text:
Access from EThOS:
Full text unavailable from EThOS. Please try the link below.
Access from Institution:
One of the main drivers influencing consumers in the purchasing of red meat is the level of visible fat, and this is particularly important in lamb, with lamb often being perceived as fatty. Consumer-driven preference for leaner meat, coupled with the meat processing industries preference for a reduction in carcass fat, increasing lean meat yield and reducing waste, have led to continued selection for lean growth and reduced fatness in several meat producing species The perception of lamb being fatty could be directly targeted in isolation by reducing overall fat levels, however there are related effects on meat (eating) quality, and the combined improvement and consistency of meat (eating) quality and the reduction of overall fatness is more complicated. It is apparent that fat content plays a significant role in meat (eating) quality. Generally four major fat depots are recognised in animal carcasses, these are: subcutaneous (under the skin); internal organ associated; intermuscular (between muscles and surrounding muscle groups); and intramuscular (marbling, between muscle fibres), the latter generally regarded as having the greatest association with meat (eating) quality. X-ray computed tomography (CT) can measure fat, muscle and bone in vivo in sheep and CT predictions of carcass composition have been used in commercial UK sheep breeding programmes over the last two decades. Together with ultrasound measures of fat and muscle depth in the loin region, CT measured carcass fat and muscle weights have contributed much to the success of breeding for leaner carcasses and increased lean meat yield. Recently it has also been considered that x-ray computed tomography provides the means to simultaneously estimate IMF and carcass fat in vivo. Thus the aim of this project is to investigate the use of two and three-dimensional x-ray computed tomography techniques in the estimation of meat (eating) quality traits in sheep, and to further investigate the genetic basis of these traits and the possibility of their inclusion into current breeding programmes. The primary approach was the use of two-dimensional x-ray computed tomography, determining the most accurate combination of variables to predict IMF and mechanical shear force in the loin. The prediction of mechanical shear force was poor with accuracies ranging from Adj R2 0.03 – 0.14, however the prediction of IMF in the loin was more promising. CT predicted carcass fat weight accounted for a moderate amount of variation in IMF (R2 =0.51). These accuracies were significantly improved upon by including other information from the CT scans (i.e. fat and muscle densities, Adj R2 >0.65). Average muscle density in a single or multiple scans accounted for a moderate amount of the variation in IMF (Adj R2 = 0.51-0.60), and again accuracies R2 >0.65 were achieved, independent of CT-measured fat areas or predicted fat weights. Similar results were achieved with the use of three-dimensional CT scanning techniques (Adj R2 0.51 – 0.71), however, there was a dramatically increased requirement for image analysis when compared to two-dimensional techniques, and the increase in accuracy was not significant. This suggests that the current method of two-dimensional image capture is sufficient in the estimation of IMF in vivo in sheep. The prediction equations developed as part of this work were applied across divergent breed types (Texel, Scottish Blackface and Texel cross mule), to investigate the transferability of the prediction equations directly across to other breeds of sheep. As part of this study, the IMF levels across the breed types and sexes were also compared and found that IMF was significantly affected by breed type (P<0.001) with Scottish Blackface lambs having higher levels of IMF when compared to Texel cross mule lambs, and the lowest levels of IMF were in the purebred Texel lambs at the same liveweight or similar levels of carcass fatness. Sex also had a significant effect on IMF across breeds (P<0.001) with females having higher levels of IMF at similar levels of both carcass fat and liveweight, and within breed, females had significantly higher levels of IMF in both the purebred Texel and Scottish Blackface lambs, when compared at similar levels of carcass fat and liveweight (P<0.05). Using the models previously developed in purebred Texel to predict IMF in the Scottish Blackface and Texel cross mule, accuracies were found to be R2 = 0.57 – 0.64 and R2 = 0.37 – 0.38 respectively. Providing evidence that the equations are transferable across to some breeds more successfully than others, however, given that there is currently no method of accurately estimating IMF in vivo, accuracies across to both breeds are acceptable. The genetic parameter estimation was unsuccessful using the same research-derived dataset as previously employed in the study. However the ambition was always to investigate the genetic relationships between traits in a large industry dataset, exploiting the wealth of commercial CT information available. These investigations were considerably more successful, and among the first to present genetic parameters of novel CT-derived IMF estimates. The results found moderate heritability estimates of h2 0.31 and 0.36 for the final selected prediction equations, with clear indications that one model not including CT predicted carcass fat or any other fat measures, was more independent of these measures and the two separate prediction methods were highly genetically correlated with each other (rg = 0.89). The results from this study show that not only is it possible to accurately estimate IMF levels in the loin of Texel sheep using CT scanning, but that, until breed specific predictions are developed, the methods developed in this study are transferable across some breed types. The results also show that CT predicted IMF is heritable, independent of overall fatness and has the potential to be included in current breeding programmes. These findings can now be used to develop breeding programmes which enable breeders to make the best use of CT scanning technology to improve carcass composition while maintaining or possibly improving aspects of meat (eating) quality.
Supervisor: Lambe, Nicola ; Bunger, Lutz ; Knott, Sara Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available
Keywords: lean meat yield ; fat content ; meat quality ; carcass composition ; X-ray computed tomography ; marbling ; lamb