Monthly model for genetic evaluation of laying hens II. Random regression

Abstract

1. We investigated the use of monthly production records for genetic evaluation of laying hens, derived from a test day model with random regression in dairy cattle and compared it with other models. 2. Records of 6450 hens, daughters of 180 sires and 1335 dams, were analysed using a model with restricted maximum likelihood (REML): traits considered were monthly and cumulative egg production. Five models were studied: (1) random regression with covariates derived from the regression of Ali and Schaeffer (Canadian Journal of Animal Science, 67: 637-644, 1987) (RRMAS), (2) random regression with covariates derived from quartic polynomial (RRMP4), (3) fixed regression with covariates derived from Ali and Schaeffer (FRM), (4) multiple trait (MTM) and (5) cumulative (CM). 3. The models were compared on the basis of Spearman rank correlations of individual breeding values and sire breeding values estimated from subsets of full-sib split data. The hens (about 10% per generation) which ranked highest on their estimated breeding values from different models were compared phenotypically with their full records. 4. The estimates of heritability resulting from RRMP4 were biased upward from the estimates obtained from MTM, so this model was discarded. The heritabilities for monthly productions from RRMAS and MTM showed a similar pattern. They were high for the 1st month of production, decreased to their lowest value at about month 5 of production and increased again to the end of lay. 5. Spearman rank correlations between animal breeding values estimated by monthly models (RRMAS, FRM and MTM) were high, between 0.91 and 0.98, whereas those between estimates of monthly models and CM were lower, from 0.85 to 0.87. The correlations estimated either from intermittent months of measurements (odd vs even months) or full records were generally high, from 0.93 to 0.99. Information from odd months of production could be sufficient for cost-efficient recording schemes. The RRMAS generally had the highest correlation of sire breeding values between subsets of full-sib records, followed by MTM, RM and CM. Monthly models selected hens with higher productivity than the cumulative model. 6. In conclusion, genetic evaluation based on monthly production may be better than using cumulative production and RRMAS appeared to be the best among the models tested here.

Estimation of genetic parameters for partial egg production periods by means of random regression models

“…The periods P1, P16, and P17 (from 20 to 22, 65 to 67, and 68 to 70 weeks of age, respectively), representing the initial and final production periods, had the lowest production averages, with 10.12 ± 4.12, 10.66 ± 3.87, and 10.15 ± 3.94 eggs produced and greater variation coefficients (VCs) of 40.71, 36.30, and 38.82%, respectively (Table 1). Anang et al (2002) observed similar variations for the initial and final production periods. In the initial period, the phenotypic variation can be attributed to the differences in the age of sexual maturity, whereas the laying persistence variation, hatching occurrence, and natural change might influence the phenotypic values of the final periods.…”

“…The egg production started to decrease from P9 (from 44 to 46 weeks of age; average, 12.85 ± 2.96 eggs). Anang et al (2002) observed that this decrease in the number of eggs began around 36 weeks of age. The lines for which the production reduction occurred at younger ages showed weaker laying persistence.…”

Estimates of the genetic parameters of turkey body weight using random regression analysis

“…Similar issues have been reported when other traits are studied (test-day milk yield) using the RR (El Faro et al, 2008). In other species, RR has been used to estimate the genetic parameters of growth and egg production in quail (Akbas et al, 2004;DioneIlo et al, 2006), laying hens (Anang et al, 2002) and broiler chicken (Banos et al, 2006;Wolc et al, 2009). Kranis et al (2007) fitted RR for egg production in turkeys and illustrated that the use of RR for genetic analysis offered greater accuracy for prediction.…”

“…In our study, efforts to fit a polynomial quadratic led to failure to converge. Convergence problems while estimating variance components with RR have been mentioned previously (Anang et al, 2002).…”

Estimates of genetic parameters for monthly egg production traits in a commercial broiler female line

“…The pattern of genetic variations of these traits has been studied, using single-trait or multiple-trait analyses, treating each performance as different traits without or with covariance among them (Anang et al, 2000;Nurgiartiningsih et al, 2004), repeated-record analyses, considering each performance as repeated measurement of the same trait (Kranis et al, 2007;Wolc et al, 2007), fixed regression (Anang et al, 2001) and random regression analyses (Anang et al, 2002;Luo et al, 2007;Wolc and Szwaczkowski, 2009).…”