Evaluation of the accuracy of simple body measurements for live weight prediction in growing-finishing pigs

Ochoa Zaragoza, Luis Enrique

Permalink

https://hdl.handle.net/2142/16207 Copy

Description

Title

Evaluation of the accuracy of simple body measurements for live weight prediction in growing-finishing pigs

Author(s)

Ochoa Zaragoza, Luis Enrique

Issue Date

2010-05-19T18:40:36Z

Director of Research (if dissertation) or Advisor (if thesis)

Ellis, Michael

Department of Study

Animal Sciences

Discipline

Animal Sciences

Degree Granting Institution

University of Illinois at Urbana-Champaign

Degree Name

M.S.

Degree Level

Thesis

Keyword(s)

• live weight prediction
• linear body measurements
• photographic image analysis

Abstract

Two studies were carried out to evaluate the accuracy of simple body measurements to predict the live weight of growing-finishing pigs. Study I used 72 barrows from a Landrace-based line; Study II used 72 barrows from each of a Landrace-based (the same line as used in Study I) and a Duroc-based line. Study I was carried out between 57.5 ± 7.1 and 126.6 ± 7.45 kg BW; Study II between 40.6 ± 4.9 and 126 ± 8.4 kg BW. In both studies, pigs were weighed every 2 wk and various body dimensions were taken on the live animal and on dorsal- and lateral-view photographic images of the pig taken at the time of weighing. Stepwise regression analysis was used to develop equations to predict live weight from body measurements. The highest R2 values were obtained for regression equations based on live animal measurements such as chest circumference [R2= 0.95; Residual Standard Deviation (RSD) = 5.7 kg], and flank circumference (R2= 0.94; RSD = 6.5 kg). Regression equations based on live-animal measurements generally gave higher R2 values than those based on measurements on the photographic images; e.g., in Study I the equation based on shoulder height gave R2 of 0.84 (RSD = 10.5 kg) for the measurement taken on the live animal compared to R2 of 0.26 (RSD = 18.2 kg) for the same measurement taken on the lateral image. Combining measurements to calculate body surface areas or volumes gave little improvement in R2 when those for the respective individual measurements were already high. Estimates of weighing period, and genotype biases for weight prediction were small, although, significant (P > 0.05) for the prediction equations from both studies. The results of this study suggest that regression equations based on simple body measurements can be used to accurately predict live weight of growing-finishing pigs.

Graduation Semester

2010-5

Permalink

http://hdl.handle.net/2142/16207

Copyright and License Information

Copyright 2010 Luis Enrique Ochoa Zaragoza

Owning Collections