• For Contributors +
• Journal Search +
Journal Search Engine
ISSN : 1225-8504(Print)
ISSN : 2287-8165(Online)
Journal of the Korean Society of International Agriculture Vol.31 No.3 pp.268-277
DOI : https://doi.org/10.12719/KSIA.2019.31.3.268

# Identification of Association between Quantity and Quality Traits by Sex for Increasing Income of Pig Farmers

Tae Wan Kim
Swine Science and Technology Center, Gyeongnam National University of Science & Technology, Jinju 52725, Korea
Corresponding author (Phone) +82-55-751-3218 (E-mail) twkim@jinju.ac.kr
April 1, 2019 September 17, 2019 September 20, 2019

## Abstract

This study was carried out to identify the influence of meat quantity trait variable on meat quality trait variable depending on sex. There was a significant difference between castration and female in most meat trait parameters. Back-fat thickness (BFT) in castration was significantly thicker than that of female (P<0.05). In addition, meat quantity trait variables were 53.6% correlation with meat quality variables using canonical correlation analysis. Furthermore, BFT had more influence on the linear combination of meat quality trait variables than carcass weight (CW) irrespective of castration and female. Especially, BFT maintained a positive relationship with fat content and drip loss in the meat quantity trait variable set, but had a negative correlation with water holding capacity, collagen content, moisture content, protein content, cooking loss and shear force. Ministry of agriculture, food and rural affairs makes the first grade determination based on the pig carcass weight and back-fat thickness. The highest grade, the grade 1+, is based on pig scalding, with a carcass weight of 83-93 kg and a back-fat thickness of 17-25 mm. Therefore, since pig grading is based on CW and BFT as the primary criterion, it is very important to obtain a high grade through proper management of CW and BFT to increase income of pig farmers.

# 양돈농가 소득증대를 위한 돼지의 성별 육량 및 육질특성 연관성 규명

김 태완
경남과학기술대학교 양돈과학기술센터

## 초록

Consumption of meat in composition ratio of food consumption has been increased according to increase in national income, and consumption trend has been increased for preference of high grade meat as shifted from quantity to quality. Among them, the import volume of fresh pork has significantly increased due to expansion of free trade agreements with various countries worldwide. As a result of these factors, since domestic consumers are increasingly required for higher quality of fresh pork, producers to strengthen international competitiveness are urgently demanded not only to reduce production costs but also to secure production technology capabilities of higher grade quality.

Many countries of pig production have made a lot of efforts to reduce fat mass and improved the leanness of pork (Razmaite et al., 2011). However, pork with the thin back-fat thickness (BFT) composed of majorly lean meat presents deficient juiciness and taste due to soft fat, subcutaneous fat separation, excessive juicy and lower taste (Kempster et al., 1986). Hence Korea has developed a technique for production of more marbled meat due to Korean to prefer to the roasted pork. Chemical composition of carcasses from growing pigs is highly associated with BFT and carcass weight (CW), but correlations between muscle quality and carcass characteristics are relatively low (Aziz and Ball, 1995). The cholesterol content of the 3-way cross-breed(LY×D) pig was significantly higher in female than in the castration (Kim et al., 2006). In addition, Jin et al. (2005) reported that as the market weight was heavier, the carcass weight, dressing rate and back-fat thickness increased, but carcass grade was significantly lower. While consumers are prepared to pay depending on meat quantity and quality to determine price (Adzitey and Nurul, 2011).

Korea is promoting the production of high quality pork by establishing the grade standards of pig carcass according to the regulations of Livestock Act and paying different prices according to the grades. Therefore, it is very important to obtain a high pig carcass grade through the proper management of CW and BFT in order to increase the income of pig farmers .

Consequently, it has an important meaning in development of pork industry to analyze direct correlation between CW or BFT to present meat quantity of pork and various variables related to quality characteristics of meat. Especially, comprehensive consideration of meat quantity and meat quality characteristic variables of pork may become useful information for pig farmer to produce pig meat, and this is an important indicator that is directly linked to farmer management improvement through increased incomes of piggery enterprise.

Canonical correlation analysis (CCA) developed by Harold Hotelling in 1935 is a useful multi-variate technique to analyze relationship between the set of independent variables, which are composed of one or more variables, and another set of dependent variables, which are composed of one or more variables (Laessig and Duckett, 1979;Sahin et al., 2011;Kim et al., 2017).

Therefore, in this study, CW and BFT among pork meat traits were evaluated on the meat quality traits of pork, along with the effect of sex difference, via application of CCA method. The results of the study can help improve the management through income increase by providing pig producers with information on appropriate CW and BFT management techniques for the production of high quality pork meet to consumer needs.

## MATERIAL AND METHODS

### Sampling and quality measurement

In this study, 492 pigs of American Berkshire reared at Dasan genetics in Korea were slaughtered at the slaughter plant, and meat characteristics of pork were measured with carcass weight and back-fat thickness after slaughter of pigs. In addition, in order to analyze the meat quality traits of pork, longissimus dorsi was collected after 24 hours post slaughter, brought to a laboratory in refrigerated condition, and then measured for water holding capacity (WHC), collagen content (CC), fat content (FC), moisture content (MC), protein content (PC), drip loss (DL), cooking loss (CL) and the shear force (SF) according to methods employed in the previous studies (AOAC, 2000;Kristensen et al., 2001).

### Statistical analysis

In this study, Canonical correlation analysis (CCA) was performed to examine the relationships between two sets of the traits, the quantity traits and quality traits in pork of American Berkshire, by using PROC CANCORR procedure of SAS 9.1 statistical package. In addition, the analysis of variance was calculated with quantity and quality traits in pork by sex. Data were expressed as mean±SD (standard deviation). Duncan Test was carried out to ascertain significant differences between samples (P<0.05). Pearson correlation coefficients were used to describe the relationship between the different factors examined.

## RESULTS

### Status of quality evaluating and auction price by grade

In Korea, the grade of pork is preferentially evaluated by CW and BFT. Then, final determination is made grade 1+, 1, 2, and others according to the state of fattening and pork belly, the color and texture of meat, and the color and quality of fat. According to KAPE, pigs slaughtered with scalding were much higher than skinning (Table 1). grade 1+ pigs accounted for 29.6 % of the total pigs slaughtered in 2018, on a scalding basis, female (32.4%) was higher than castration (27.5%).

It is possible to analyze the optimal CW and BFT of the pigs to contribute to the increase of farmers income through statistics on the auction price by sex and CW. Although there is a difference between female and castrated pigs, the ranges of CW with the highest auction price were 76 to 87 kg. And the BFT was evaluated as 19.0 to 20.5 mm (Table 3). Therefore, it is suggested that females should be slaughtered with CW of 83.5 kg and BFT of 19.2 mm, a castrated pigs with CW of 77.8 kg and BFT of 20.3 mm.

### Meat quantity traits and quality traits by sex

Variance was analyzed by sex difference to investigate the characteristics of meat quality and quantity of pork (Table 4).

According to the analysis result, CW and BFT in castration showed the measured values higher than those of female. However, CW did not show a significant difference, but BFT in castration was significantly thicker than that of female (P<0.05). In this study, DL and CL in castration were detected by higher values than those of female (P<0.05). On the contrary, WHC, CC, FC, MC, PC and SF in female showed higher values than those of castration (P<0.05).

### Relationship between meat quantity and quality traits

The simple correlations were examined among all the examined pork traits from 492 pigs of American Berkshire (Table 5).

The highest correlation was found between MC and FC (0.793), then was between PC and FC (0.634), but there were negative relationships between them (P<0.01). Except them, however, most of the correlation coefficients represent low value.

On the other hand, when the correlation between the meat quantity and quality characteristics was examined by Pearson correlation, the heavier CW, the thicker BFT and the more DL occurred, but CW had a decreased relationship with CL. In addition, the thicker BFT, the more FC and DL were generated, but BFT were presented by negative correlations with CC, MC, PC and CL (P<0.01).

### Results analyzed from CCA

The summary of results given from CCA for the two pairs of canonical variate was shown in Table 6. The maximum number of canonical function to be extracted equals the number of variable in the smallest canonical variates (Dattalo, 2014). The first canonical correlation of 0.536 represented the highest possible correlation between any linear combination of the meat quality traits variables (Wt1) and any linear combination of the meat quantity traits variables (Vt1) in pork. The results indicated a highly significant relationship between the canonical variables (P<0.01). The coefficients of CCA were 0.430 and 0.503 for the first pair of castration and female, respectively. These results presented a significant relationship between meat quantity and quality characteristic variable pairs (P<0.01).

In Table 5, the meat quantity characteristic variables (CW and BFT) were very low correlations with the meat quality characteristic variables. However, in Table 6, canonical correlations presented to have relatively high values. The standardized canonical coefficients were given for the all pair of canonical variables (Vt1 and Wt1) in Table 7.

The canonical variates to represent the optimal linear combinations of dependent variables and independent variables can be defined by using the standardized canonical coefficients as below,

$V t 1 = 0.084 C W + 0.969 B F T W t 1 = − 0.103 W H C − 0.231 C C + 1.928 F C + 0.957 M C 0.524 P C + 0.348 D L − 0.302 C L + 0.023 S F$

The coefficients for the meat quantity traits variables showed BFT rather than CW to contribute heavily the first canonical variable (Vt1) as 0.969. On the other hand, the coefficients for the meat quality traits variables showed the degree of FC, MC and PC to contribute heavily for the first canonical variable (Wt1) as 1.928, 0.957 and 0.524, respectively. These results also showed similar patterns in analysis by sex. However, the coefficient values of CW (0.560) and DL (0.684) had higher values in castration, whereas the value of PC (0.330) maintained lower.

A canonical loading gives product-moment correlation between the original variable and its corresponding canonical variate. It reflects degree which a variable is represented by a canonical variate (Yaprak et al., 2008;Tahtali et al., 2012). Canonical loadings of the original variables with their canonical variables were shown in Table 8.

In terms that variables with larger canonical loadings were more contributed to the multivariate relationships between the meat quality and quantity trait variables in pork, CW in the loadings for the meat quantity traits was more influence than BFT in formation of the meat quality trait variables (Wt1). In addition, the loadings for FC and MC were more influence than other meat quality traits in formation of the meat quantity trait variables (Vt1). These results showed a similar trend in the case of female. However, both CW and BFT in castration made a similar contribution to formation of meat quality characteristic variables. Conversely, it was analyzed that FC and DL in castration made a larger contribution than other variables to form the meat quantity characteristic variable.

Fig. 1. illustrated the related structure of the first canonical correlation on Wt1 and Vt1. Canonical correlation of the original variables was matched with their canonical variables (Table 8), and the proportion of correlations between the pair of canonical variables Wt1 and Vt1 was 53.6%. It indicates to show high correlation between the meat quality trait variables (Wt1) and the meat quantity trait variables (Vt1).

FC provided the relatively high positive contribution to canonical variate Vt1, whereas DL had the relatively low positive relationship, according to the cross loadings for the first pair of canonical variables (Table 9). However, MC, PC, CL, CC, SF and WHC had negative contribution to canonical variate Vt1. BFT due to the cross loadings for the first pair of canonical variables provided the relatively high positive contribution to canonical variate Wt1, BFT depending to these results was analyzed to have a relatively high influence on the linear combination of meat quality characteristic variables. When BFT is high, FC and DL are high, whereas MC, PC, CL, CC, SF and WHC were low.

### Analysis of redundancy index

The redundancy index indicates total proportion of variance to be shared by two sets of original data variables. It is designed to overcome the inflated correlations (Laessig and Duckett, 1979). The first pair of canonical variable Vt1 was explained by 57.7 % of total variation in the meat quantity component trait from the canonical redundancy analysis, whereas the first canonical variable Wt1 was explained by 16.6 % (Table 10). In addition, 16.3 % of total variation in the meat quality trait variable set was explained by first pair of canonical variable Wt1, while the redundancy index of 0.047 for first canonical variable was explained by 4.7 % of canonical variable Vt1, where RI (0.047) was derived by multiplying the shared variance of the variate (0.163) and the squared canonical correlation (0.288). Therefore, the result of canonical redundancy analysis was explained by 4.7 % and 16.6 % of variance, indicating that neither of the pair of canonical variables is a good overall predictor for the opposite set of variables.

## DISCUSSION

The difference of pig’s CW and BFT affect quality characteristics of pork. To distinguish these differences, we examined quality and quantity traits depending on overall and sex of pork, and correlation between each variable. Furthermore, we applied CCA to confirm correlation between meat quantity and quality characteristic variable sets.

Firstly, the results according to sex differences in pork quantity and quality characteristic variables were shown in Table 4. The difference in BFT is mainly affected by sex (Muhlisin et al., 2014). In this study, BFT in castration was 0.14 mm thicker than that of female (P<0.05). This result was confirmed by similar result in Korean slaughter sexuality of the whole pig in 2018 (KAPE, 2019). BFTs of castration and female in pork scalding have 23.2 and 20.4 mm, respectively, which BFT in castration maintains 2.8 mm thicker than that of female.

DL and CL in castration also had higher values than those of female (P<0.05). In previous studies, although contradictory research results were derived for DL values (Unruh et al., 1996;Franco et al., 2014), in this study it turned out that sex was affected in DL. The result of CL was conflicted with the previous study (Muhlisin et al., 2014), which CL in castration is lower than that of female. However, WHC, CC, FC, MC, PC and SF in female showed higher values than those of castration (P<0.05). FC and intramuscular fat of castrations maintain higher than those of females in meat (Latorre et al., 2003;Muhlisin et al., 2014). Therefore, female to have higher values of the traits such as WHC, CC, MC, PC and SF related with meat quality appears better characteristic than castration in the quality characteristics associated with taste of meat (Kim et al., 2017).

In this way, the difference in the eating quality characteristics of pork to prefer Korean consumers is displayed depending on the results of the discriminately expressed protein and hormone between castration and female. However, since mean BFT in the castrated swine is 0.14 mm thicker than that of sow (Jo et al., 2010), we suggest that producer make a relatively large effort to breeding management so that pig is not accumulated excessively in BFT of castration. In other words, we suggest that it is necessary of improvement for a direction to have an appropriate level of BFT rather than a thin BFT improvement.

In order to analyze two or more variables at the same time, there are various tools including multiple regression analysis, multiple analysis of variance, principal component analysis, factor analysis, multiple analysis of covariance, cluster analysis, multiple discriminant analysis, structural equation modeling, conjoint analysis and canonical correlation analysis (CCA). Among these techniques, CCA are able to predict simultaneously the relationship between the two sets of variables including more than one variables in each (Akbas and Takma, 2005;Kim et al., 2017). In this study, CCA was applied to examine the relationship between the set of meat quantity trait variables, such as CW and BFT, and the set of meat quality trait variables, such as WHC, CC, FC, MC, PC, DL, CL and SF, in American Berkshire.

Canonical correlations between the pair of canonical variables were examined by 53.6% and 23.0%, respectively, and these maintained a significantly high relationships (P<0.01). In particular, from the first pair of canonical variates (Table 8), FC and DL had a positive relationship with set of the meat quantity trait variables, while WHC, CC, MC, PC, CL and SF had a negative relationship with them. These results were nearly similar with the results of cross loading. FC according to cross loadings for the first pair of canonical variables provided the relatively high contribution to canonical variate of the meat quantity traits, while MC and PC maintained an inverse relationship (Table 9). In addition, a redundancy index of 0.166 for the first canonical correlation suggests that 16.6% of the variance in the meat quantity variables set (CW and BFT) is accounted by the meat quality variables set (WHC, CC, FC, MC, PC, DL, CL and SF) in Table 9.

The increased slaughter weight has the advantage to reduce the overhead costs for producers, slaughterers and processors by increasing carcass yields to improve meat to bone ratio and to reduce chilling and processing losses (Ellis and Bertol, 2001). Furthermore, pork quality at heavier weight maintains higher values in juiciness, flavor and tenderness to be attributable to differences in intramuscular fat composition (Piao et al., 2004). However, it has also been shown that each 10 kg increment per 100 kg live weight leads to a slightly lower average daily gain, a significant deterioration in feed efficiency, reduced lean deposition and poorer meat quality (Correa, 2005). In addition, as pig gets bigger body weight with a large amount of fat accumulation, BFT becomes rapidly thicker and fat cakes increase incidence. This not only leads to a decrease in incomes of producers but also can be a subject of repelling by consumers due to an increase in production cost and a decrease in quality.

Therefore, we suggest that pig farmers should maintain adequate BFT to produce high-quality pork for consumer needs while saving costs, because of production management of pig according to pig's sex is greatly important in increasing income.

## 적 요

상업적 영농을 하는 양돈농가는 돼지의 생산비를 줄이면서 소비자들이 선호하는 고품질의 고기를 생산하는 것이 소득증 대에 있어서 무엇보다 중요한 일이다. 이에 농림축산식품부는 돼지도체를 1+, 1, 2 및 등외등급으로 판정하고, 그 결과에 따라 농가수취가격을 차등화함으로써 고품질 돼지고기 생산을 촉진하기 위해서 돼지고기 도체등급제도를 시행하고 있으며, 돼지도체의 체중과 등지방두께를 기준으로 1차 등급판정을 한 다. 가장 높은 등급인 1+ 등급은 돼지의 탕박 기준으로 도체 중이 83-93 kg, 등지방두께가 17~25 mm이다. 축산물등급판정 통계연보에 의하면, 2018년 경락단가가 가장 높은 평균 도체 중량은 암퇘지가 83.5 kg, 거세돼지가 77.8 kg이었으며, 평균 등지방두께는 암퇘지가 19.2 mm, 거세돼지가 20.3 mm인 것으 로 조사되었다.

본 연구는 거세돼지와 암퇘지에 대한 분산분석, 육량변수와 육질변수 사이의 개별 상관관계, 그리고 육량변수세트와 육질 변수세트 사이의 정준상관분석을 수행하였다. 먼저 분산분석 의 결과는 다음과 같이 요약할 수 있다.

첫째, 돼지의 육량 및 육질특성변수들에 있어서 성별 차이 를 분석한 결과는 등지방두께는 거세돼지가 암퇘지보다 0.14㎜ 두꺼웠다(P<0.05).

둘째, 육즙감량과 가열감량은 거세돼지가 암퇘지보다 많았 다 (P<0.05).

셋째, 보수력, 콜라겐함량, 지방함량, 수분함량, 단백질함량 및 전단가는 암퇘지가 거세돼지보다 높았다 (P<0.05).

육량특성과 육질특성 사이의 상관분석 결과는 다음과 같이 요약할 수 있다.

첫째, 도체중이 무거울수록 등지방두께가 두껍고, 육즙감량 이 많이 생기지만, 가열감량은 줄어드는 관계를 갖는다 (P<0.01).

둘째, 등지방두께가 두꺼울수록 지방함량이 늘어나고, 육즙 감량이 많이 생기지만, 콜라겐함량, 수분함량, 단백질함량 및 가열감량과는 음(-)의 상관관계를 갖는 것으로 나타났다 (P<0.01).

셋째, 수분함량과 지방함량의 상관관계가 가장 높았으며 (0.793), 이어서 단백질함량과 지방함량의 상관관계가 높았지 만 (0.634), 이들 함량 간에서는 서로 음(-)의 관계를 보이고 있다 (P<0.01).

마지막으로 정준상관분석의 결과는 다음과 같이 요약할 수 있다.

첫째, 육량특성변수 짝과 육질특성변수 짝 사이에 유의한 관 계가 있는 것으로 나타났다 (P<0.01).

둘째, 등지방두께(BFT)가 높으면 지방함량과 육즙감량은 높 은 반면, 수분함량, 단백질함량, 가열감량, 콜라겐함량, 전단가 및 보수력이 낮은 것으로 나타났다.

이와 같은 결과에 따르면, 소비자들은 지방함량이 많으면서 수분손실이 많은 돼지고기와 단백질함량과 콜라겐함량이 적은 돼지고기는 기피할 수 있다는 점을 간과하여서는 안 될 것이 다. 특히 국내 돼지도체등급은 1차적으로 돼지도체중량과 등지 방두께를 평가한 이후, 육질항목을 평가하여 최종등급이 판정 되고 있기 때문에 우선적으로 육량지수인 도체중량과 등지방 두께를 적절하게 관리하는 것이 소비자의 요구를 충족하면서 농가의 생산비 절감 및 농가수취가격상승에 의한 소득증대로 이어지는 중요한 요인이 된다는 사실에 주목할 필요가 있다.

## Figure

Correlations between the quantity (V1) and quality (W1) factors, and among their canonical variables

## Table

Status of quality evaluating by sex of pig in 2018

Average auction price(KRW) per 1 kg according to grade of pork carcass (scalding) by year

Average auction price according to carcass weight of pig (scalding) by sex

Quantity traits of carcass and quality traits of pork by sex

Pearson correlation matrix for the quantity traits of carcass and the quality traits of pork

Descriptive statistics from canonical correlation analysis

Standardized canonical coefficients for the quantity and quality variables of pork

Redundancy analysis for the pair of canonical variables

## Reference

1. Adzitey, F. and H. Nurul.2011. Pale soft exudative (PSE) and dark firm dry (DFD) meats: causes and measures to reduce these incidences - a mini review. International Food Research Journal 18: 11-20.
2. Akbas, Y. and C. Takma.2005. Canonical correlation analysis for studying the relationship between egg production traits and body weight, egg weight and age at sexual maturity in layers, Czech J Anim Sci. 50: 163-168.
3. AOAC.2000. Official Methods of Analysis of AOAC International. 17th ed. Assoc. Off. Anal. Chem., Gaithersburg, MD.
4. Aziz, N. N. and R. O. Ball.1995. Effects of backfat thickness and carcass weight on the chemical composition and quality of the meat from culled sows. Can. J. Anim. Sci. 75: 19l-196.
5. Cankaya, S. and G. T. Kayaalp.2007. Estimation of relationship between live weight and some body measurements in German farm × hair crossbred by canonical correlation analysis, Hayvansal Uretim 48 (2): 27-32.
6. Correa, J. A. , Faucitano, L. , Laforest, J. P. , Rivest, J. , Marcoux, M. and C. Gariepy.2006. Effects of slaughter weight on carcass composition and meat quality in pigs of two different growth rates, Meat Science 72: 91-99.
7. Dattalo, P. V. 2014. A demonstration of canonical correlation analysis with orthogonal rotation to facilitate interpretation. Virginia Commonwealth University, Richmond, Virginia. pp.3-5.
8. Ellis, M. and T. M. Bertol.2001. Effects of slaughter weight on pork and fat quality. In Proceedings of the 2nd international virtual conference on pork quality, November-6 December 2001, Concordia, Brazil. pp.213-224.
9. Franco D. , Vaquezand, J. A. and J. M. Lorenzo.2014. Growth performance, carcass and meat quality of Celta pig crossbred with Duroc and Landrance genotypes, Meat Sci. 96: 195-202.
10. Jo, H. , Kim, B. W. , Sun, D. W. , Park, J. C. , Park, C. H. and J. G. Lee.2010. Effect of backfat thickness and carcass weight on carcass quality grade in commercial pigs. Journal of Agriculture & Life Science 44: 53-60.
11. KAPE (Korea Institute of Animal Products Quality Evaluation).2019. 2018 Animal Products Grading Statistical Yearbook. Sejong Korea.
12. Kempster, A. J. , Dilworth, A. W. , Evans, D. G. and K. D. Fischer.1986. The effects of fat thickness and sex on pig meat quality with special reference to problems associated with overleanness: Butcher and consumer panel results. Anim. Prod. 43: 517-533.
13. Kim, I. S. , Jin, S. K. , Song, Y. M. , Park, K. H. , Kang, S. M. , Ha, J. H. , Kim, I. J. , Park, Y. S. and J. H. Kim.2006. Quality Characteristics of Pork by Sex on Crossbred Pigs. The Korean Society of International Agriculture 18: 34-39.
14. Kim, T. W. , Kim, I. S. , Kwon, S. G. , Hwang, J. H. , Park, D. H. , Kang, D. G. , Ha, J. , Kim, S. W. and C.W. Kim.2017. Identification of relationship between pork colour and physicochemical traits in American Berkshire by canonical correlation analyses, Animal Production Science 57: 1179-1185.
15. Kristensen, L. and P. P. Purslow.2001. The effect of ageing on the water-holding capacity of pork: role of cytoskeletal proteins. Meat Science 58: 17-23.
16. Laessig, R. E. and E. J. Duckett.1979. Canonical correlation analysis: potential for environmental health planning. Am. J. Public Health 69: 353-359.
17. Latorre, M. A. , Lázaro, R. , Gracia, M. I. , Nieto, M. and G. G. Mateos.2003. Effect of sex and terminal sire genotype on performance, carcass characteristics, and meat quality of pigs slaughtered at 117 kg body weight. Meat Sci. 65: 1369-1377.
18. Muhlisin, Panjono, Lee, S. J. , Lee J. K. and S. K. Lee.2014. Effects of Crossbreeding and sex on the Carcass Traits and Meat Quality of Korean Native Black Pig and Duroc Crossbred, Asian Australas. J. Anim. Sci. 27: 1019-1025.
19. Piao, J. R. , Tian, J. Z. , Kim, B. G. , Choi, Y. I. , Kim, Y. Y. and I. K. Han.2004. Effects of sex and market weight on growth performance carcass characteristics and pork quality of market hogs. Asian-Australasian Journal of Animal Sciences 10: 1452-1458.
20. Razmaite, V., Ribikauskene, D. and A. Stimbirys.2011. Effects of carcass weight on quality of major carcass cuts, their composition, and meat in Lithuanian slaughter pig population, 2011,Acta Veterinaria (Beograd) 61: 259-268.
21. Safari, P. , Danyali, S. F. , Honarnejad, R. and M. Esfahani.2013. Study of relationship between oil quality traits with agromorphological traits in peanut genotypes by canonical correlation analysis. International Journal of Biosciences 3: 1-10.
22. Sahin, M. , Cankaya, S. and A. Ceyhan.2011. Canonical correlation analysis for estimation of relationship between some traits measured at weaning time and six-month age in merino lambs, Bulgarian J. Anim. Sci. 17: 680-686.
23. Jin, S. K. , Kim, I. S. , Hur, S. J. , Hah, K. H. and B. W. Kim.2005. Physico-chemical Characteristics with Market Weight in Korean Native and Landrace Crossbred Pigs. The Korean Society of International Agriculture 17: 182-187.
24. SAS.2011. SAS User’s Guide, SAS Inst. Inc., Cary, NC.
25. Sharma, S. 1996. Applied multivariate techniques, John Wiley and Sons, Inc., Canada, pp.391-404.
26. Tahtali, Y. , Cankaya, S. and Z. Ulutas.2012. Canonical correlation analysis for estimation of relationships between some traits measured at birth and weaning time in Karayaka lambs, Kafkas Univ vet Fat Derg 18: 839-844.
27. Unruh, J. A. , Friesen, K. G. , Stuewe, S. R. , Dunn, B. L. , Nelssen, J. L. , Goodband, R. D. and M. D. Tokack.1996. The influence of genotype, sex, and dietary lysine on pork subprimal cut yields and carcass quality of pigs fed to either 104 or 127 kilograms. J. Anim. Sci. 74: 1274-1283.
28. Ventural, H. T. , Lopes, P. S. , Peloso, J. V. , Guimaraes, S. E. F. , Carneiro, A. A. P. S. and P. L. S. Carneiro.2011. Canonical correlation analysis of the association between carcass and ham traits in pigs used to produce dry-cured ham. Genetics and Molecular Biology 34: 451-455.
29. Yaprak, M , Koycegiz, F. , Kutluca, M. , Emsen, E. and H. W. Ockerman.2008. Canonical correlation analysis of body measurements, growth performance and carcass traits of red Karaman lambs. J. Anim. Vet. Adv. 7: 130-136.