On-Barn Forecasting Beef Cattle Production Based on Automated Non-Contact Body Measurement System.

The main task of selective breeding is to determine the early productivity of offspring. The sooner the economic value of an animal is determined, the more profitable the result will be, due to the proper estimation of high and low productive calves and distribution of the resources among them, accordingly. To predict productivity, we offer to use a systematic assessment of animals by using the main genetic parameters (correlation coefficients, heritability, and regression) based on data such as the measurement of morphological characteristics of animals, obtained using the automated non-contact body measurement system based on RGB-D image capture. The usefulness of the image capture system lies in significant time reduction that is spent on data collection and improvement in data collection accuracy due to the absence of subjective measurement errors. We used the RGB-D image capture system to measure the live weight of mother cows, as well as the live weight and body size of their calves (height at the withers, height in the sacrum, oblique length of the trunk, chest depth, chest girth, pastern girth). Cows and cattle of black-and-white and Holstein breeds (n = 561) were selected as the object of the study. Correlation analysis revealed the main indices for the forecast of meat productivity-live weight and measurements of animals at birth. Calculation of the selection effect is necessary for planning breeding work, since it can determine the value of economically beneficial traits in subsequent generations, which is very important for increasing the profitability of livestock production. This approach can be used in livestock farms for predicting the meat productivity of black-and-white cattle.

Genome-Wide Associative Study of Phenotypic Parameters of the 3D Body Model of Aberdeen Angus Cattle with Multiple Depth Cameras.

In beef cattle breeding, genome-wide association studies (GWAS) using single nucleotide polymorphisms (SNPs) arrays can reveal many loci of various production traits, such as growth, productivity, and meat quality. With the development of genome sequencing technologies, new opportunities are opening up for more accurate identification of areas associated with these traits. This article aims to develop a novel approach to the lifetime evaluation of cattle by 3-D visualization of economic-biological and genetic features. The purpose of this study was to identify significant variants underlying differences in the qualitative characteristics of meat, using imputed data on the sequence of the entire genome. Samples of biomaterial of young Aberdeen-Angus breed cattle (n = 96) were the material for carrying out genome-wide SNP genotyping. Genotyping was performed using a high-density DNA chip Bovine GPU HD BeadChip (Illumina Inc., San Diego, CA, USA), containing ~150 thousand SNPs. The following indicators were selected as phenotypic features: chest width and chest girth retrieved by 3-D model and meat output on the bones. Correlation analysis showed a reliable positive relationship between chest width and meat output on the bones, which can potentially be used for lifetime evaluation of meat productivity of animals.

A Comparative Study of Machine Learning Methods for Predicting Live Weight of Duroc, Landrace, and Yorkshire Pigs.

Live weight is an important indicator of livestock productivity and serves as an informative measure for the health, feeding, breeding, and selection of livestock. In this paper, the live weight of pig was estimated using six morphometric measurements, weight at birth, weight at weaning, and age at weaning. This study utilised a dataset including 340 pigs of the Duroc, Landrace, and Yorkshire breeds. In the present paper, we propose a comparative analysis of various machine learning methods using outlier detection, normalisation, hyperparameter optimisation, and stack generalisation to increase the accuracy of the predictions of the live weight of pigs. The performance of live weight prediction was assessed based on the evaluation criteria: the coefficient of determination, the root-mean-squared error, the mean absolute error, and the mean absolute percentage error. The performance measures in our experiments were also validated through 10-fold cross-validation to provide a robust model for predicting the pig live weight. The StackingRegressor model was found to provide the best results with an MAE of 4.331 and a MAPE of 4.296 on the test dataset.