Comparison of principal component analysis method and multiple regressions in estimating the weight of fattening camels

Authors

1 Animal Science Research Department, Yazd Agricultural and Natural Resources Research and Education Center, AREEO, Yazd, Iran

2 Animal Science Research Department, Qom Agricultural and Natural Resources Research and Education Center, AREEO, Qom, Iran

3 Asisstant professor, Animal Science Research Department, Golestan Agricultural and Natural Resources Research and Education Center, AREEO, Gorgan, Iran.

4 Animal Science Research Department, Yazd Agricultural and Natural Resources Research and Education Center, AREEO, Yazd, Iran.

Abstract

Background and objectives: Camel breeding is one of the sources of income for people on the edge of arid and desert areas in many parts of the world. Traits related to camel growth, including birth weight, weaning weight, daily weight gain and one-year-old weight are considered as the main economic traits for camel owners. To manage and improve the genetic and phenotypic values of these traits, weight recording of animals is essential. Recording of camels, especially their weight is associated with many difficulties due to their restless temperament and large size. Using efficient mathematical methods can greatly solve this problem. Various approved efficient mathematical methods have been proposed to predict the weight of camels based on their body dimension, and their effectiveness has been proven. So, the present study was conducted to compare the efficiency of the Principal Component Analysis and multiple regressions in estimating the weight of fattening camels from its body dimensions.
Materials and methods: In order to compare the efficiency of principal components analysis and multiple regressions in estimating the weight of fattening camels based on body dimensions, the records of 220 fattening camels of Bafgh station in Yazd were used. For this purpose, new born camels were fed for a 9 months period using standard diets. During the period, each of the fattening camels was weighed and different body sizes including body length(BL), whither height(WH), breast girth(BG), abdomen with(AW), hump height to the ground(HH), muzzle girth(MG), neck length(NEL), whither to pin length(WPL), tail length(TL), pelvic width (PW), abdomen to hump height (ABH) and the head length(HL) were measured. The body dimensions of the camels were recorded using a tape measure and their body weight was recorded using a scale. Then, the data were analyzed using the principal component analysis and multiple regressions. In order to fit the predictive models, body weight of camels was introduced as dependent variable and body size of camels as independent variables. Analysis of regression models was done using one and multivariable linear models and the best model were selected to estimate the weight of camels based on their body dimensions by Stepwise method. The performance of the above models was evaluated using comparison of the coefficient of determination (R2) of them.
Results: According to the results, the correlation between the body weight of camels with their different body dimensions including BL, WH, BG, AW, HH, MG, NEL, WPL, TL, PW,ABH and HL were 0.93, 0.89, 0.89, 0.89, 0.94, 0.73, 0.89, 0.90, 0.80, 0.85, 0.89 and 0.79 ,respectively. The results showed that among the 6 multiple regression models fitted to estimate the weight of fattening camels, model No. 6 in which head length, body length, breast girth, neck length, muzzle girth and whither height were used as predictive variables had the least error (12.06) and the highest accuracy (0.92) compared to other models. The results showed that the use of the first and second principal component , and both of them in the model could explain 82.1%, 3.73% and 85% of the variance of body weight, respectively. The weight of fattening camels was determined using principal component analysis with an accuracy of 0.93 and an error of 11.54.
Conclusion: The results of the present study showed that in order to estimate the weight of fattening camels, the use of principal component analysis, in addition to simplifying predictive models, has higher efficiency and accuracy as well as less error compared to multiple regressions, and this method can be a suitable alternative to the multiple regression method in predicting the weight of camels.

Keywords

Main Subjects

References

Atta, M. and el-Khidir, O.A. 2004. Use of heart girth wither height and scapuloischial length for prediction of love weight of Nilotic sheep. Small Ruminant Research, 55(1):233-237.
Bahashwan, S., Alrawas, A.S., Alfadli, S. and Johnson, E.S. 2016. Dofari cattle growth curve prediction by different nonlinear model functions. Livestock Research, 27(12).
Bitaraf Sani, M., Harofte, J.Z., Banabazi, M.H., Esmaeilkhanian, S., Naderi, A.S., Salim, N. and Faghihi, M.A. 2021. Genomic prediction for growth using a low-density SNP panel in dromedary camels. Scientific Reports, 11(1): 1-14.
Cannas, A. and Boe, F. 2003. Prediction of the relationship between body weight and body condition score in sheep. Italian Journal of Animal Sciences, 2: 527-529.
FAO. 2019. FAOSTAT, Food and Agriculture Organization of the United Nations, Rome. ISBN 978-92-5-131789-1.
FAO. 2021. Statistical year book world food and agriculture. Food and Agriculture Organization of the United Nations, Rome. ISBN 978-92-5-134332-6.
Faye, B. 2013. Camel Farming Sustainability: The Challenges of the Camel Farming System in the XXIth Century. Journal of Sustainable Development, 6(12): 74-82.
Franicis, J., Sibanda, S., Hermansen, J. and Kristrerrsen, T.E. 2002. Estimating body weight of cattle using linear body measurements. Zimbabwe Veterinary Journal, 33:15.
Iqbal, Z.M., Javed, K., Abdollah, M., Ahmad, N., Ali, A., Khalique, A., Aslamand, N. and Younas, U. 2014. Estimation of body weight from different morphometric measurement in Kali lambs. Journal of Animal and Plant Science, 24(3): 700-703.
Kadim, I.T., Mahgoub, O. and Purchas, R.W. 2008. A review of the growth and the carcass and meat quality characteristics of the one humped camel. Meat Science, 80(3): 555-69.
Khodai, S.A. 2004. The report on camel production systems and the socio-economics of camel herders in the Islamic Republic of Iran. Cardn /Acsad / Camel, 107: 4–10.
Kohler-Rollefson, I., Mundy, P. and Mathias, E. 2001. A Field Manual of Camel Diseases. London, U.K., Intermediate Technology Group Publishing, pp. 254.
Littell, R.C., Freund, R.J. and Spector, P.C. 1991. SAS System for Linear Model. 8th Edition, SAS Institute Inc., Cary.
Mahmoud, M.A., Shaba, P. and Zubairu, U.V. 2014. Live body weight estimation in small ruminant: areview. Global Journal of Animal Scientific Research, 2(2).
Moradi Sharbabak, H., Moghbeli, H., Moradi Sharbabak, M., Miraie Ashtiani, S.R. 2015. Determination of camel weight regression equation using biometrical traits of Yazdi camel breed by multivariate linear regression analysis based on the principal component analysis. Animal Science Journal (Pajuhesh and Sazandegi), 108 : 25-34. (In Persian). 
Ogah, D.M. 2011. Assessing size and conformation of the body of Nigerian indigenous turkey. Slovak Journal of Animal Science, 44 (1): 21-27.
R-Cran. 2019. R-3.5.3 for Windows (32/64 bit). Package source: ggfortify_0.4.14.tar.gz.
Salau, J., Haas, J., Junge, W., Bauer, U., Harms, J. and Bieletzki, S. 2014. Feasibility of automated body trait determination using the SR4K time-of-flight camera in cow barns. Springer Plus, 3: 225.
Salehi, M. and Gharahdaghi, A.A. 2013. Camel Production Potential and Recent Research in Iran. Available online: ttp://agris.fao.org/agris-search (12 Jan 2022).
Snedecor, S.W. and Cochran, W.G. 1989. Statistical method s. 8th Edition, Iowa state University Press. USA.
SPSS. 2016. IBM SPSS Statistics for Windows, Version 24.0. Armonk, NY: IBM Corp (Released 2016).
Tsegaye, D., Belay, B. and Haile, A. 2013. Linear body measurement as predictor of body weight in Haraghe Highland goat under farmer’s environment Ethiopia. Global Veterinarian, 11(5): 649-656.
Yakubu, A., Kingsley, K.O. and Agade, Y.I. 2009. Using factor scores in multiple linear regression models for predicting the carcass weight of broiler chickens using body measurements. Revista UDO Agrícola, 9 (4): 963-967. 2009.
Journal of Ruminant Research
Volume 11, Issue 2
September 2023
Pages 37-48

Files

Share

How to cite

Statistics