The Brazilian Swine and Poultry Research Center (Embrapa Suínos e Aves) maintains a chicken breeding program for meat production since 1985. Two control lines (LLc and PPc) are maintained, whereas two male lines (TT and ZZ) and three female lines (PP, VV and KK) have been selected. This paper reports the genetic gain after 15 generations of combined selection (mass and independent culling levels) in order to develop the commercial broiler stocks Embrapa 021 and Embrapa 022. Selection pressure has been exerted on weight gain, carcass traits and fertility. In addition, female lines have also been selected for egg production, whereas males have been selected for feed efficiency since 1992. All lines have been selected for breast area instead of carcass traits since 1999. The genetic gain was estimated as the deviation between selected lines and the respective unselected lines at 42 days of age. In female lines, body weight improved 504, 548 and 587 g; average breast area increased 27.60; 16.99 and 26.43 cm²; adjusted feed conversion (42-49 d) improved -1.46; -0.97 and 1.76 units, and egg production varied 6.99; 7.12 and -3.43% units for PP, VV and KK, respectively. In male lines, body weight improved 758 and 408 g; average breast area increased 31.95 and 19.38 cm², and adjusted feed conversion improved (42-49 d) -0.99 and 1.26 for TT and ZZ, respectively. This breeding program has been effective to generate genetic gain and to develop two commercial products, Embrapa 021 (standard) and Embrapa 022 (high yield). Nevertheless, feed efficiency is still not satisfactory.

Two paternal broiler lines (LL and LLc, originated from the same stock in 1998) of Embrapa Suinos e Aves genetic program were used. Line LL had been selected for eight generations for rapid growth while LLc had been mated randomly, without selection, as a control line. In the last generation, 30 males and 240 females were used in hierarchical matings with pedigree control to determine sire effect. The same number of birds were used, but matings were at random for LLc. Eggs of both lines were incubated and embryos were collected in different development stages (40, 45, 50, 55 and 60 hours after setting) in order to determine somite number (NOS). The number of evaluated embryos were 1,967 and 498 for LL and LLc, respectively. This study aimed to determine the changes in somite formation during embryonic development that resulted from artificial selection for heavier body weight. In the evaluated generation, the differences for body weight, feed conversion ratio, carcass and breast yields between LL and LLc were 546 g; 0.16 FC units, 2.20 and 2.19%, respectively. NOS was greater in the initial (40 h) and in the final (60 h) stages for LL and in the intermediary stages (45, 50 and 55 h) for LLc, indicating significant differences in the mechanisms of somite formation and development. Considering that both lines had the same genetic origin, such deviation indicates genetic changes due to artificial selection for body weight. The heritability estimates indicated low additive genetic effect for NOS in the different stages. Since the alterations were due to artificial selection for body weight, one could consider the hypothesis of using such genetic marker throughout family selection.

One thousand two hundred and ninety six Arbor Acres females and 144 males were used to study the influence of feeding time on sexual maturity and carcass composition. Treatments were: T1 = feeding at 6:30 am, T2 = 50% feeding at 6:30 am and 50% at 3:30 pm (dual), T3 = feeding at 11:00 am, T4 = feeding at 3:30 pm. Nutrition and management were as recommended to the commercial line. Variables studied were: sexual maturity at 5% production (SM), mean egg production percentage (PEP) body weight (BW), carcass composition, tibia weight (TIBW) and tibia calcium (Ca) and phosphorus (P) contents. Analysis of variance was used, and the means were compared using Student's t test. T2 females reached SM at 25.92 weeks, which was earlier (p<0.01) than the other treatments. PEP was greater (p<0.01) for T1 and T2 than for the other treatments. BW was greater in T1 and smaller in T4 (p<0.05) than in the other treatments. TIBW was greater in T3 than in T1 and T2 (p<0.05), with no effect (p>0.10) on carcass dry matter (DM), ether extract (EE), crude protein (CP) and ash (ASH). DM, EE, TIBW, Ca and P increased (p<0.05) whereas CP and ASH decreased with age. Time of feeding influenced body weight, but not carcass composition. Hens fed twice a day had earlier sexual maturity and similar egg production compared to those fed once at 6:30 am.

This study used 300 females and 30 males with 36 weeks of age from the selected PP and control PPc maternal broiler lines. PP has been selected for heavy body weight (PC) and high egg production for eight generations. Fertile eggs were collected and weighed individually for 4 periods of 5 consecutive days at two-week intervals. In each period, a total of 960 eggs/line were identified and separated in groups of 240 eggs, and stored for later incubation. Embryo weight (PE) was evaluated at 9 (P9), 11 (P11), 13 (P13), 15 (P15), 17 (P17) and 21 (P21) days of incubation. The objective was to estimate the effect of selection on embryo development. Egg weight (PO) was similar between the two lines. The differences in PE were significant from P15 on, resulting in 1.9g of difference in the chick weight, indicating correlated genetic changes in the embryo development, which can be credited to the selection for PC. Changes in PE while PO was kept unaltered modified the correlations between these two traits. Differences were significant from P13 on and estimated correlations for P21 were 0.72 and 0.70 for PP and PPc, respectively. Chick weight corresponded to 70.91% (PP) and 68.48% (PPc) of egg weight. The estimated increase in P21 that resulted from the increase of 1.0g in PO was 0.71 in PP and 0.68g in PPc.

Um modelo polinomial segmentado foi usado em diversas curvas de produção de ovos, numa tentativa de analisar as diferenças entre as curvas de um modo mais descritivo. Foram usadas curvas de produção de diferentes linhagens comerciais e experimentais de poedeiras. Os parâmetros do modelo incluem a idade e nível do pico de produção, a taxa de decréscimo na produção após o pico e o tempo entre o início e o pico de produção. Essas e outras variáveis derivadas pelo modelo foram comparadas entre as curvas usando contrastes. A metodologia usada facilitou a interpretação das curvas e a detecção de diferenças na maturidade sexual, uniformidade e pico de produção. Esse trabalho valida o modelo de polinômios segmentados para uso em pesquisas futuras que tratem de análise de curvas de produção de ovos.

A segmented polynomial model was used on several egg production curves in an attempt to analyze the differences between the curves in a more descriptive manner. Production curves from different commercial and experimental genetic lines of layers were used. The parameters of the model include the age and level of peak production, rate of decrease in production after the peak and time between start and peak of production. These and other derived variables were compared between the curves using contrasts. The methodology used allowed an easier interpretation of the curves, and the detection of differences in sexual maturity, uniformity and peak of production. This work validates the segmented polynomial model for use in future research dealing with analysis of egg production curves.