Assessing Environmental Effects upon Modeling the Individual Lactation Curve of a Local Goat Population in a Pastoral System.

The present study aims to use Wood's model to determine the parameters of individual lactation curves in a local goat population and their factors of variation under a pastoral system. A total of 137,927 records from 432 local goats were collected to assess the impact of litter size, year and month of kidding, herd and the age of the dam on lactation curve parameters. Wood's model parameters were estimated using non-linear regression, and individual curves were fitted. The characteristics of the lactation curves were computed. The initial yield (A), rate of increase (B) and rate of decline (C) parameters in Wood's model for local goats were 730 g, 0.26 and 0.09 respectively. The values of peak milk production (PP), peak date (PD) and persistency (PC) were 931.88 g/d, 23.39 days and 91.50%, respectively. Persistence was higher in goats with simple births, while peak production increased by around 0.3 kg for each additional kid. The curve parameters "A" and "C" differed according to the herd and month of kidding (p < 0.05); the age of the dam only had an impact on parameter "A" (p < 0.01). Parameter "B" was not significantly influenced by any of the factors considered (p > 0.05). Correlation coefficients among lactation curve characteristics were ranged from -0.20 to 0.89. Due to a significant negative correlation, selecting for parameter "A" may have an adverse effect on parameter "B", resulting in a shorter time to reach peak production and less persistency, but an increase in peak production among goats. The curves derived from Wood's model suggest that the shape of the curve may serve as a basis for herd management planning and to improve local population potentialities.

On the modelling of weights of kids to enhance growth in a local goat population under Tunisian arid conditions: the maternal effects.

To provide the local goat population with an adequate breeding scheme under an arid environment, this study aimed to evaluate the use of alternative models attempting to dissect the additive genetic (AG) and permanent environmental (PE) components of direct and maternal effects on weights of kids up to weaning. Records of 903 local kids over a period of 16 years were used in this study. Data were split into four groups corresponding to four periods along weight recording. Periods 1/2/3/4 contained weights in an interval with upper-lower limits of 1-20/25-60/65-120/125-150 days of age. Models including or ignoring maternal genetic or permanent environmental effects were fitted for all traits. For all periods, the best models were those including the AG component for both direct and maternal effects and the direct PE effect. Heritability estimates of both the direct and maternal effects ranged from low (0.02 for maternal heritability in P1) to moderate (0.17 for direct and maternal heritability in P2 and P3). Period 1 showed the lowest values for heritability of both direct and maternal effects, with also the largest estimate of the ratio of residual to total variance (around 0.2) compared with the other periods, with decreasing ratios as age increased (from 0.13 for P2 to 0.07 for P4). Both direct and maternal estimated breeding values (EBVs) showed high correlations for models fitting direct AG and PE (DGP) effects. For direct EBVs (DEBVs), correlations were above 0.99, indicating that the same animals are expected to be selected under any model that includes those components, regardless of the maternal effects included. For maternal EBVs, correlations were also high, but slightly lower than for the DEBVs between models including DGP effects and maternal genetic effects. Overall, our recommendation for genetic evaluations of direct and maternal effects in this population raised in extensive and harsh conditions is to use weight records preferably collected during the period of high milk production of dams, for which direct and maternal effects are expected to show full expression. Complete pedigrees and several generations of dam-progeny recording are needed to obtain a proper separation of environmental and genetic components.

Genetic analysis of live weight of local kids to promote genetic evaluations in the arid areas of Tunisia.

Goat meat production, a widely extended activity in the more arid areas of Tunisia, relies on local breeds. These breeds are well adapted to produce under harsh conditions but have a very small size and low productivity. The aim of this study was to establish the basis for future genetic evaluations to improve growth potential of this local stock. A total of 13,095 body weights and pedigree of 945 kids in the caprine herd of the Arid Areas Institute of Médenine were used. Random regression (RR) and multiple trait (MT) models were analyzed and compared. All models included effects of age and weight of dam, age, sex and type of birth of the kid, and year × month of recording, plus random direct and maternal additive genetic and permanent environmental effects. RR and MT models behave similarly, with RR showing slightly better goodness of fit. Heritability estimates for direct (ranging from 0.15 to 0.4) and maternal (0.05 to 0.3) effects showed that efficient selection for weight is feasible in this population. Estimated correlations between ages were high (> 0.8) for direct effects across all ages and low (down to 0.2) for weights taken at distant ages for maternal effects. Estimated genetic correlations between direct and maternal components revealed an antagonistic relationship, especially at early ages. Recording of at least one weight in the first month of age of the kids to evaluate the maternal capacity and a later weight to evaluate direct effects on weight is recommended for genetic evaluations under field conditions.