Quantitative analysis of gametic imprinting effects on productive and reproductive performances of Markhoz goat

Abstract

In this study, in order to assess the relative importance of gametic imprinting effects for the genetic variation of growth and reproduction traits of Markhoz goat, a data set were obtained from a total of 4983 individuals on productive and reproductive traits of this breed (collected from 1993 to 2017) was analysed. The studied productive traits included body weight (BW) of Markhoz kid at birth and 4 (WW), 6 (W6), 9 (W9), and 12 (W12) months of age, average daily weight gains (g day−1) in the first two growth phases (i.e. from birth to weaning, ADG0–4; and from weaning to 6-month age, ADG4–6) and their corresponding Kleiber ratios (i.e. KR0–4 and KR4–6). In addition, reproduction traits of interest were the liter size at birth (LSB) and at weaning (LSW), total litter weight at birth (TLWB) and at weaning (TLWW) per does kidding, litter mean weight per kid born (LMWKB) and per kid weaned (LMWKW). Likelihood ratio tests and Akaike’s Information criterion indicated that gametic imprinting effects were significant only for BW, W9 and KR0–4. Both paternal and maternal gametic imprinting effects had a similar influence on these traits, accounting for between 8 % (W9) and 10 % (BW and KR0–4) of the total phenotypic variance. For these traits the model with the second-best fit was models including paternal gametic imprinting effects, which indicates crucial paternal gametic imprinting effects in the analyzed traits in the production context. For parentally gametic influenced traits of Markhoz kid, including these effects significantly decreased the direct additive genetic variances (20–94 % depending on the trait) and these reduction were higher when paternal and maternal gametic variance evaluated together than when evaluated separately. Generally these results indicate that when gametic imprinting effects are of importance, but not accounted for, heritability estimates are biased upward and subsequently the realized efficiency of selection is reduced. It was concluded that for more precise estimates of genetic parameters, extracting of overly influential observations from the dataset is necessary.