Application of random regression model for quantitative-genetic analysis of the number of live-born piglets in a sow litter

[N. Stojiljković]

96 str.

Biotehničke nauke - Opšte stočarstvo i oplemenjivanje domaćih i gajenih životinja / Biotechnical Sciences - General animal science and breeding of domestic and farm animals

Reproduktivne osobine krmača u Republici Srbiji nisu u velikoj meri zastupljene u odgajivačko-selekcijskim programima, samim tim i biološki potencijal ovih osobina nije dovoljno iskorišćen. Uključivanjem reproduktivnih osobina krmača u selekcijske programe može se uticati pozitivno na ekonomsku isplativost u proizvodnji svinjskog mesa. Uzimajući navedeno u obzir, ovo istraživanje je najviše usmereno na utvrđivanje mogućnosti za genetsko poboljšanje osobina veličine legla krmača, kako bi se povećala produktivnost i ekonomska isplativost pri proizvodnji svinjskog mesa. Tokom izrade ove doktorske disertacije akcenat je stavljen na procenu genetskih parametara u cilju pouzdanije procene priplodne vrednosti što je od izuzetno velikog značaja s obzirom na uticaj ovog parametra na tačnost i pouzdanost procenjenih priplodnih vrednosti grla i posledično na efekte selekcije. Procena genetskih parametara za niskonasledne osobine veličine legla krmača izvršena je primenom različitih metodskih postupaka i modela: metodom mešovitih modela (model ponovljivosti i višeosobinski model) i metodom slučajne regresije. Primenom različitog pristupa osobinama (repetability i multirait) i korišćenih modela za osobinu broj živorođene prasadi u analiziranoj populaciji krmača utvrđen je optimalni model za ocenu aditivne genetske varijanse i kovarijanse kako bi se utvrdila naslednost i povezanost (fenotipska i genetska), a sve sa ciljem stvaranja preduslova za što tačniju procenu priplodne vrednosti i postizanje većih efekata selekcije. U ovom istraživanju analizirana je fenotipska i genotipska varijabilnost za broj živorođene prasadi (BŽP) u leglu krmača. Ova osobina je izabrana jer je u direktnoj i jakoj vezi sa ostalim osobinama veličine legla, kako pri prašenju, tako i pri zalučenju legla, a od nje je direktno zavisna i ukupna godišnja produktivnost krmača, kao i ekonomska efikasnost ove faze proizvodnog ciklusa u svinjarskoj proizvodnji. Analiza je izvršena na osnovu podataka sa tri farme svinja pojedinačno (A, B, C), kao i nakon objedinjavanja podataka u jedan set (ABC). Objedinjavanje podatka i istovremena genetska analiza su bile moguće zbog postojanja genetskih veza između životinja gajenih na analiziranim farmama. Prosečan BŽP na analiziranim farmama iznosio je 16,18 (A), 14,01 (B), 16,88 (C) i 16,06 (ABC). Podaci o plodnosti krmača u periodu od 13 godina (2008-2020) su obuhvatili ukupno 58043 legla sa tri farme, koja su oprasile 18962 krmače. Istraživanjem su bili obuhvaćeni rezultati plodnosti četiri najčešća genotipa plodnih mesnatih rasa svinja na farmama u Republici Srbiji (Landras - L, Veliki Jorkšir - VJ, recipročni melezi - LxVJ i VJxL). Broj krmača obuhvaćenih ovim istraživanjem po farmama kretao se od 3160 (farma B) do 10192 (farma A), dok se broj oprašenih legala kretao od 7823 (farma B) do 36200 (farma A). Nakon objedinjavanja sve tri farme u jedan set podataka (ABC) istovremeno je analizirano ukupno 58043 legala, koje su oprasile 18962 plotkinje. Primena korišćenih metodskih postupaka za analizu genetske varijabilnosti zahtevala je korišćenje podataka o poreklu (pedigre fajlova) koji su kreirani za tri generacije predaka. Pedigre fajl 1 (farma A) sadržao je 15478 individua, pedigre fajl 2 (farma B) sadržao je 4886 individua, pedigre fajl 3 (farma C) 10879 individua, a nakon objedinjavanja sve tri farme u jedan set podataka pedigre fajl (ABC) je sadržao ukupno 23543 individue...

In the Republic of Serbia sow reproductive traits have not been represented in breeding and selection programmes therefore a biological potential of these traits has not been used enough. By including sow reproductive traits in selection programmes the economic profitability in pork production can be positively affected upon. Taking this into account the research is mostly focused on identifying the possibilities of genetic improvement of the traits of sow litter size in order to increase productivity and economic profitability in pig meat industry. During composing this doctoral thesis the focus was on estimation of genetic parameters with the goal of obtaining more reliable evaluation of breeding value which is essential taking into account the impact of this parameter on accuracy and reliability of animal evaluated breeding values and consequently on selection effects. The estimation of genetic parameters for low-heritability traits of litter size in sows was performed by applying different methodological procedures and models: mixed model method (repeatability model and multi-trait model) and random regression method. By applying different treatments (repeatability and multi-trait) and models for a trait number of live born piglets in the analyzed sow population, the most optimal model for the assessment of additive genetic variance and covariance was determined in order to define their heredity and connection (phenotypic and genetic), with the aim of obtaining, as accurate as possible, estimation of breeding value and further achieving greater effects of selection. In this study, phenotypic and genotypic variability of the number born-alive piglets (NBA) in a litter of sows were analyzed. This trait was chosen because it is in a direct and strong relationship with the other traits of litter size, both at farrowing and weaning and the total reproductive efficiency of sows at an annual level, as well as an economic efficiency of this phase of production cycle in pig production that is directly dependent on it. The analysis was performed on the basis of data obtained from three pig farms (A, B, C) and upon combining data from all three farms into one set (ABC). Combining all three farms into one data set was possible due to the presence of genetic links between the animals reared on the analyzed farms. The average NBA on the analyzed farms was 16.18 (A), 14.01 (B), 16.88 (C) and 16.06 (ABC). Data on sow fertility in a 13-year period (2008-2020) included a total of 58,043 litters farrowed by 18,962 sows on all three farms. The research included reproductive characteristics of the four most common fertile meat breed genotypes in pigs reared on the farms in the Republic of Serbia (Landrace-L, Large White-LW, LxLW and LWxL). The number of sows per farm included in this research ranged from 3160 (Farm B) to 10192 (Farm A) being the lowest - 7823 (Farm B) and the highest - 36200 (Farm A). After combining all three farms into one data set (ABC), a total of 58,043 litters farrowed by 18,962 sows were analyzed at the same time. For applying the methodological procedures used in the analysis of genetic variability data the animal origin (pedigree files) created for three generations of ancestors was needed. Pedigree files 1 (PED1), 2 (PED2) and 3 (PED3) contained 15478, 4886 and 10879 individuals, respectively, and upon merging all these three farms into one data set, pedigree file 4 (PED4) with a total of 23543 individuals was formed...

krmače, broj živorođene prasadi, heritabilitet, model ponovljivosti, višeosobinski model, slučajno-regresijski model, priplodna vrednost, model životinje, genetska povezanost

sows, NBA, heritability, repeatability model, multi-trait model, random-regression model, REML, animal model, genetic association

636.4.082(043.3)

Tekst