ABSTRACT The objective of this study was to evaluate residual feed intake (RFI), residual intake and body weight gain (RIG) and their relationship with other traits of efficiency, performance, and economic analysis of sheep. Lambs (n = 102) were evaluated during 56 days and a herd consisting of 500 ewes was simulated with the mean of dry matter intake (DMI) and final body weight of the lambs, the same as that of the experimental lambs. The lambs were fed hay-based diet of Tifton 85 (Cynodon spp.), corn, and soybean in the voluminous:concentrate ratio of 35:65. Residual feed intake and RIG were correlated with DMI, feed conversion ratio, and feed efficiency. Residual intake and body weight gain were positively correlated with average daily gain, relative growth rate, and Kleiber's rate. The most efficient production systems (lower RFI and higher RIG) had lower costs and higher profit margins. The net present value (NPV) and internal rate of return (IRR) were higher in systems with more efficient lambs. In systems with less efficient lambs, NPV and IRR were negative and lower, respectively. Efficient animals for RFI and RIG showed satisfactory performance and better economic results.

RESUMO: O estudo da avaliação genética do consumo alimentar residual ajustado para a gordura (CARFat) é importante para o uso apropriado da eficiência alimentar em programas de seleção. Objetivou-se analisar a influência da seleção para CARFat sobre características de carcaça e de desempenho, frente à estimação dos parâmetros genéticos. Foram analisados os dados de cinco provas de eficiência alimentar, com 677 machos Nelore. A avaliação genética foi realizada por Inferência Bayesiana, com modelo animal, em análises uni e bicaracterística. Foram analisadas as variáveis: ingestão de matéria seca, ganho de peso diário, CARFat, área de olho de lombo, espessura de gordura, espessura de gordura subcutânea na picanha, marmoreio e acabamento. As médias das distribuições a posteriori, estimadas em cada uma das análises, foram usadas para a estimação das herdabilidades e das correlações. As características estudadas apresentaram altas estimativas de herdabilidades, devendo responder bem à seleção. O CARFat apresentou correlações fenotípicas próximas a zero com as características de carcaça, e correlações genéticas negativas. A variabilidade genética aditiva observada para CARFat demonstra que a seleção para essa característica poderá promover ganhos genéticos nas futuras gerações, obtendo animais eficientes na utilização de alimentos e, de acordo com as correlações genéticas e fenotípicas encontradas, sem mudanças negativas significativas às características da carcaça.

ABSTRACT: The study of the genetic evaluation of residual feed intake adjusted for fat (RFIFat) is important for the appropriate use of feed efficiency in selection programs. The objective was to analyze the influence of selection for RFIF at on carcass and performance traits by estimating various genetic parameters. Data were analyzed from five tests of feed efficiency, which were conducted with 677 Nellore males. Genetic evaluation was performed by Bayesian inference using an animal model via single- and two-trait analyses. Variables analyzed were dry matter intake, average daily gain, RFIFat, rib eye area, back fat thickness, rump fat thickness, marbling score, and subcutaneous fat thickness. The posterior mean distributions estimated at each analysis were used to estimate heritability of the traits and to perform various correlations. The studied traits showed high heritability estimates, and they should respond well to selection. The RFIFat presented a phenotypic correlation with carcass traits (which was next to zero), and there was also a negative genetic correlation. Additive genetic variability for RFIFat showed that selection for this trait can promote genetic gains in future generations, resulting in animals that are efficient in terms of nutrient use, and according to the genetic and phenotypic correlations, with no significant negative changes to carcass traits.

Objetivou-se verificar a divergência genética entre linhagens de codornas de corte e seus cruzamentos para as características de desempenho através da análise multivariada. Foram utilizadas quatro linhagens de codornas de corte em um sistema de cruzamentos dialélicos completos, proporcionando a formação de 16 grupos de progênies, sendo quatro parentais puros: L1, L2, L3 e L4; seis grupos genéticos mestiços F1: G12, G13, G14, G23, G24 e G34; e seis grupos genéticos mestiços F1 recíprocos. Foram avaliadas as seguintes variáveis: peso corporal médio ao nascimento, aos 28, aos 35 e aos 42 dias de idade; o consumo médio de dieta do nascimento aos 35 e aos 42 dias de idade; a conversão alimentar do nascimento aos 35 e aos 42 dias de idade; e o ganho em peso do nascimento aos 35 e aos 42 dias de idade. O desempenho dos grupos genéticos foi avaliado por meio da análise de variância multivariada e a primeira variável canônica, usando os testes do maior autovalor de Roy e da união-interseção de Roy para as comparações múltiplas. O estudo da divergência genética foi feito por meio da análise por variáveis canônicas e pelo método de otimização de Tocher. As três primeiras variáveis canônicas explicaram 88,82% da variação entre os grupos genéticos. A divergência genética entre os grupos genéticos avaliados permitiu a formação de quatro grupos: 1 - L1, G12, G23, G21 e G41; 2 - G13 e G31; 3 - L4 e G43; 4 - G32; 5 - G14 e G24; 6 - G42 e 7 - L2, L3 e G34, identificando a L4 como a linhagem que mais influiu na formação de novo grupos.

The objective was to evaluate the genetic divergence between lineages of quail and their crosses to the performance characteristics through multivariate analysis. Four lines of quails were used in a complete diallel system, provided training to 16 groups of progenies, four parental pure: L1, L2, L3 and L4; six genetic groups crossbred F1: G12, G13, G14, G23, G24 and G34; and six reciprocal F1 crossbred genetic groups. The variables were evaluated: mean body weight at birth, at 28, at 35 and 42 days of age; the average consumption of diet from birth to 35 and 42 days of age; feed conversion birth at 35 and 42 days of age; and the weight gain from birth to 35 and 42 days of age. The performance of genetic groups was assessed by multivariate analysis of variance and the first canonical variable, using the tests of the largest eigenvalue of Roy and Roy union-intersection for multiple comparisons. The study of genetic diversity was done by canonical variate analysis and by the optimization method of Tocher. The first three canonical variables explained 88.82% of the genetic variation between groups. The genetic divergence between the analyzed genetic groups allowed the formation of four groups: 1 - L1, G12, G23, G21 and G41; 2 - G13 and G31; 3 - L4 and G43; 4 - G32; 5 - G14 and G24; 6 - G42 and 7 - L2, L3, identifing L4 G34 as the lineage that most influenced the formation of new groups.

Machos e fêmeas, Santa Inês puros (SI) e cruzados de matrizes Santa Inês com reprodutores das raças Texel (TxSI), Ile de France (IFxSI) e Bergamácia (BxSI), terminados em confinamento e abatidos aos 15, 25, 35 e 45kg de peso vivo. Após o abate, foram tomados os pesos das vísceras e analisado o crescimento alométrico. O pulmão dos machos BxSI cresceu em taxas mais elevadas que SI e TxSI. A tráqueia/esôfago dos machos SI e TxSI cresceu mais lentamente que dos IFxSI e BxSI. Os rins dos machos BxSI cresceram mais rápido comparado aos dos TxSI e IFxSI. O fígado das fêmeas TxSI cresceu em taxas mais elevadas quando comparadas às fêmeas dos outros grupos genéticos. Os machos e fêmeas do grupo TxSI apresentaram taxas de crescimento maiores para rúmen/retículo, respectivamente, comparado aos grupos SI e BxSI.

Males and females, including purebred Santa Inês lamb (SI) and crosses between Santa Inês dams and Texel (TxSI), Ile de France (IFxSI) and Bergamasca (BxSI) sires were feedlot finished and slaughtered at 15, 25, 35 and 45kg live weight. After slaughter, the following non- carcass components were weighed and calculated the allometric growth. Lungs of BxSI males grew faster than those of SI and TxSI. Trachea/esophagus in SI and TxSI males grew slower than IFxSI and BxSI. Kidneys BxSI males grew faster than TxSI and IFxSI. Livers of female TxSI lamb grew more rapidly than those of females in the other genetic groups. In both males and females, the rumen/reticulum grew faster in the TxSI group than SI and BxSI groups.

Data from 2439 goats of the Saanen, Alpine, Anglo Nubian and Toggenburg breeds recorded from 1976 to 2009 by the Association of Goats and Sheep Breeders of Minas Gerais were used in principal component analysis. After consistency of data, six morphological variables (thorax perimeter, body length, withers height, height, width and length of the rump) and 12 variables related to breed standard score and fitness (breed characteristic, head, palette and topline, feet and legs, dairy type, body capacity, udder, rear and front ligament, udder texture, teat and final score) were analyzed. Based on the magnitude of the eigenvalue (lower than 0.7), eleven variables considered redundant were discarded, resulting in reduced costs of technician labor to evaluate the animals. Maintenance of records on height, length, rump width, breed characteristic, dairy type, front ligament and udder texture is recommended.

The objective of this study was to estimate (co)variance components for weight at 550 days, average daily gain and an index with both traits, and to compare alternatives for evaluation and ranking of Nellore young bulls in pasture performance tests. The heritability estimates were 0.73, 0.31 and 0.44 for weight at 550 days, average daily gain and index, respectively. Animals were ranked according to their predicted breeding values or the phenotypic deviations in relation to the mean of the test. Although the correlations between breeding values and phenotypic deviations were high, there were differences in the number of animals selected in common when the selection criteria were the predicted breeding values or the phenotypic deviations. Mixed models are more appropriate than the least squares method and should be utilized in the evaluation of young bulls in performance tests.

Objetivou-se com este trabalho estimar os parâmetros genéticos e fenotípicos dos parâmetros da curva de produção de ovos, pelo o modelo de polinômios segmentados ("x p"- idade das aves ao pico de postura, "P"- nível de produção no pico, "s" - taxa de decréscimo semanal na produção de ovos após o pico e "t" - tempo entre o início da postura e o pico de produção de ovos). Foram utilizados dados da produção de ovos semanal da 25ª até 64ª semanas de idade, provenientes de 2.398 matrizes de uma linhagem de frangos de corte. A estimação dos componentes de covariâncias, herdabilidade e correlações para esses parâmetros da curva de postura foram obtidas por máxima verossimilhança restrita. As tendências genéticas foram estimadas com as médias dos valores genéticos dos parâmetros da curva em função do ano de nascimento das matrizes. As médias das herdabilidades foram de baixa a média magnitudes, 0,25 (x p), 0,18 (P), 0,17 (s) e 0,10 (t). As correlações genéticas entre o parâmetro t e os demais parâmetros da curva foram positivas, variando de 0,21 até 1,00, sendo menor entre t e s e maior entre os parâmetros x p e t. Entre os parâmetros x p e P, a correlação também foi positiva, de magnitude média (0,39). As correlações foram negativas entre os parâmetros s e x p e s e P, apresentando-se menor entre s e x p (-0,17) e maior entre s e P (-0,91). As tendências genéticas encontradas para os parâmetros indicaram uma redução no tempo entre o início e pico de produção e idade ao pico de postura. O nível de produção apresentou uma queda no valor genético ao longo do período avaliado. Pode-se observar possível ganho genético para os parâmetros da curva de produção de ovos, contudo, deve-se atentar para as correlações genéticas positivas entre os parâmetro x p e P, o que pode levar a um aumento do nível de produção e a um aumento na idade ao pico de produção. O mesmo caso acontece com os parâmetros x p e s, indicando que as aves mais precoces ao pico poderão apresentar menor persistência.

The objective of this study was to estimate genetic parameters and phenotypic parameters of egg production curve, using the segmented polynomial model( " x p " - the peak age of the birds in attitude, "P" - at peak production level, "s" - rate of decline in weekly egg production after peak, and "t" - time between the onset of lay and peak egg production). We used data from the weekly egg production from 25 to 64 weeks of age from 2,398 arrays of a lineage of broilers chickens. Estimation of covariance components, heritability and correlations for these parameters curve posture were obtained by restricted maximum likelihood. Genetic trends were estimated with the average breeding values the of parameters of the curve as a function of birth year of the arrays. The mean heritabilities were low to medium magnitude, 0.25 (x p), 0.18 (P), 0.17 (s) e 0.10 (t). Genetic correlations among the parameter "t" and the other parameters of the curve were positive ranging from 0.21 to 1.00, being lower between "t" and "s", and higher between the parameters x p and t. Between the parameters x p and P the correlation was also positive, of an average magnitude (0.39). Correlations were negative between parameters "s" and "x p", and "s" and "P", presenting itself lower between "s" and "x p" (-0.17) and higher between "s" and "P" (-0.91). The genetic trends found for the parameters of the model indicated a reduction in time between the production's onset and peak, and age at production's peak. The level of production showed a decline in breeding value over the period evaluated. It can be observed a potential genetic gain for the egg's production parameters' curve; however, it must be paid special attention to the positive genetic correlations between the parameter x p and P. What can lead to an increase in the level of production and may lead to an increase in age at peak production. The same applies to the parameters x p indicating that birds with earlier peaks may present lower persistence.

The objective of this study was to compare different random regression models, defined from different classes of heterogeneity of variance combined with different Legendre polynomial orders for the estimate of (co)variance of quails. The data came from 28,076 observations of 4,507 female meat quails of the LF1 lineage. Quail body weights were determined at birth and 1, 14, 21, 28, 35 and 42 days of age. Six different classes of residual variance were fitted to Legendre polynomial functions (orders ranging from 2 to 6) to determine which model had the best fit to describe the (co)variance structures as a function of time. According to the evaluated criteria (AIC, BIC and LRT), the model with six classes of residual variances and of sixth-order Legendre polynomial was the best fit. The estimated additive genetic variance increased from birth to 28 days of age, and dropped slightly from 35 to 42 days. The heritability estimates decreased along the growth curve and changed from 0.51 (1 day) to 0.16 (42 days). Animal genetic and permanent environmental correlation estimates between weights and age classes were always high and positive, except for birth weight. The sixth order Legendre polynomial, along with the residual variance divided into six classes was the best fit for the growth rate curve of meat quails; therefore, they should be considered for breeding evaluation processes by random regression models.

Performance, carcass, non-carcass and commercial cuts and components of Texel × Santa Inês crossbred lambs, managed in confinement and fed diets based on soybean oil, soybeans and a conventional diet, with or without the use of monensin (78 ppm dry matter basis) were evaluated. Thirty-six Texel × Santa Inês lambs (18 males and 18 females) were managed in intensive systems. Animals were slaughtered after 87 days of confinement, and performance, carcass characteristics, weight and percentages of carcass and non-carcass components were evaluated. The animals fed the control diet had heavier carcass and parts than animals fed soybean, while the oil diet did not differ from the controls in most parameters. The animals fed soybean showed lower intake kg dry matter (DM), organic matter (OM), crude protein (CP), neutral detergent fiber (NDF) and metabolizable energy (ME) compared with animals fed the control diet, increased ether extract (EE) intake in kg, % body weight (BW) and metabolic weight (MW) and did not differ from the soybean oil diet. Animals receiving monensin had lower DM intake, OM, CP, EE, NDF, gross energy (GE) regardless of the expression, % kg BW, or % PM, than the animals that did not receive the additive. Males produced better and had heavier cuts than the females; the latter deposited subcutaneous fat earlier. Animals that received oil or soybean showed greater body weight and small intestine percentage. Soybean oil intake did not improve performance, carcass weights or parts of Santa Ines × Texel lambs in confinement. Soybeans at 15% dry matter reduced energy intake and lamb performance. The use of monensin at 78 ppm on a dry matter basis is not recommended for lambs in confinement, especially if associated with oil or soybeans that may harm animal performance.

This experiment aimed to determine the influence of diets with inclusion of protected fat and vitamin E on performance, yield and carcass characteristics of feedlot lambs with different weights. Thirty-two non-castrated Santa Ines lambs were fed diets with a ratio of 40% forage and 60% concentrate ad libitum, with presence or absence of protected fat and/or vitamin E, in a total of four diets. Two weights of early containment were also considered: between 20 and 25 kg and between 30 and 35 kg. All animals were slaughtered at 84 days of confinement. Animals fed diets without addition of protected fat, regardless of the use of vitamin E, had the highest intakes of dry matter, crude protein, neutral detergent fiber and ash, and decreased intake of ether extract. The variables investigated did not affect daily weight gain and total gain. Feed conversion was better for the lighter confined animals not fed protected fat. The heavier feedlot lambs fed diets with vitamin E showed higher cold carcass. The warm carcass for the lighter animals confined fed with vitamin E, and the heavier ones, fed with protected fat and vitamin E, showed the best yields of cold carcass. Objective measures of the carcass cold had the highest mean for heavier feedlot lambs. The addition of fat in the diet reduces the intake of dry matter and increases the ether extract. Although the inclusion of vitamin has no effect on intake of nutrients, it protects the carcasses from losses during cooling, and weight differences at containment directly reflect the measures of the carcasses.

A criação de tilápia-do-Nilo (Oreochromis niloticus) e o seu filé apresentam um mercado vigoroso e expansão no Brasil e no mundo. Portanto, um modelo de regressão obtido a partir de parâmetros corporais externos, de fácil observação, que prediga o rendimento e o peso do filé possibilita selecionar indivíduos com maior potencial produtivo. Objetivou-se com este trabalho obter um modelo que permita predizer o rendimento e o peso do filé, baseado em medidas morfométricas e do peso. Foram avalidos 890 indivíduos dos quais os caractéres mensurados foram: comprimento total (CT), comprimento-padrão (CP), comprimento da cabeça (CC) e do tronco (CTR), altura da cabeça (AC1), altura lombar (AL), altura caudal (AC2), espessura lombar (EL) e caudal (EC), peso corporal (PESO), peso de filé (PF) e rendimento de filé (RF). Foram realizadas análises de regressão linear, com PF e RF como variáveis dependentes. O R² para o modelo do PF foi 0,80, demonstrando que as medidas corporais são confiáveis para se estimar o peso do filé. Já para o modelo do RF, o R² foi de 0,16, não sendo adequado para se estimar o rendimento do filé, devendo ser desenvolvidos novos estudos para se obter um modelo mais confiável.

The rearing of Nile tilapia (Oreochromis niloticus) and its fillet have shown a vigorous market expansion in Brazil and worldwide. Therefore, a regression model obtained from external physical parameters, of easy observation, which can predict the yield and weight of fillet, making it possible to select individuals with higher yield potential. The objective of this work was to obtain a model that allows a prediction of fillet yield and weight, based on morphometric measurements and weight. Data were collected on 890 individuals and the measured characters were: total length (TL), standard length (CP), length of the head (CC) and the stem (CTR), height of the head (CA1), lumbar height (AL), height flow (CA2), lumbar thickness (EL) and throughput (EC), body weight (PESO), fillet weight (FW) and yield of fillet (RF). A linear regression analysis was performed, with PF and RF as dependent variables. The R² for the model of the FP was 0.80 - a satisfactory result - suggesting that the body measurements are reliable for estimating fillet weight. For the RF model, R² was 0.16, which is not adequate to estimate fillet yield, requiring further studies in order to obtain a reliable model.

Avaliaram-se o desempenho e as características de carcaça de 36 cordeiros machos não-castrados, Santa Inês puros e mestiços Dorper × Santa Inês e Texel × Santa Inês, manejados em três sistemas (intensivo semi-intensivo e extensivo) e abatidos com peso médio de 38,41 ± 1,24 kg. Os melhores ganhos de peso foram observados nos cordeiros criados em sistema intensivo, que apresentaram também os maiores pesos e rendimentos de carcaça. Poucas influências foram constatadas sobre as medidas de carcaça e dos cortes comerciais, sobretudo para maiores pesos nos cortes provenientes do manejo intensivo, proporcionado pelo maior peso da carcaça. O percentual relativo dos cortes em relação à carcaça fria praticamente não variou entre os sistemas. O sistema intensivo proporcionou melhor desempenho, independentemente do grupo genético. A utilização das raças Texel e Dorper em cruzamento com Santa Inês eleva o rendimento das carcaças.

It was evaluated the performance and the carcass characteristics of 36 non-castrated male lambs, Santa Inês purebreed and crossbreeds from Dorper × Santa Inês and Texel × Santa Inês, managed in three different systems (intensive, semi-intensive and extensive), slaughtered at 38.41 ± 1.24 kg. It was observed the highest weight gain for lambs reared in intensive system, which also showed the greatest carcass yield. The management systems did not influence carcass measures of animals in intensive system much because of the greater weight of the carcass. The intensive system provided the best performance, regardless the genetic group. The use of Texel and Dorper breeds in crossings with Santa Inês increases carcass yield.

Utilizaram-se 103 cordeiros, machos e fêmeas, Santa Inês puros (SS) e mestiços com Texel (TS), Ile de France (FS) e Bergamácia (BS) confinados individualmente e abatidos aos 15, 25, 35 ou 45 kg de peso vivo. Após abate e resfriamento da carcaça, foram separados os cortes comerciais pescoço, costela/fralda, costeleta, lombo, paleta e perna, os quais foram dissecados em músculo, osso e gordura. O crescimento alométrico de cada tecido da meia-carcaça foi avaliado em relação ao peso da carcaça fria e os tecidos da perna, do lombo e da paleta em relação ao peso do corte. O desenvolvimento do músculo na meia-carcaça nos machos Santa Inês, Santa Inês × Texel e Santa Inês × Ile de France e nas fêmeas Santa Inês e Santa Inês × Texel foi semelhante ao do peso vivo vazio (coeficiente alométrico (b) = 1), enquanto, nos machos Bergamácia e nas fêmeas Ile de France e Bergamácia, foi precoce (b<1). O desenvolvimento do músculo foi precoce (b<1) na paleta dos machos TS e BS e das fêmeas FS e BS; no lombo de machos TS e fêmeas SS, TS e BS; e na perna dos machos BS e fêmeas TS e BS. O desenvolvimento do osso nos cortes e na meia-carcaça foi precoce (b<1) em todos os grupos, com exceção dos ossos da paleta dos machos SS, TS e BS e das fêmeas TS; e dos ossos da perna dos machos BS, com desenvolvimento semelhante (b=1) ao desenvolvimento do corte. O desenvolvimento da gordura nos cortes e na meia-carcaça foi tardio (b>1) em todos os grupos, com exceção da gordura da paleta nos machos BS, cujo desenvolvimento foi semelhante ao do corte (b=1).

This study was conducted using 103 male and female lambs from the following genetic groups: Santa Inês (SS) purebred and Santa Inês × Texel (TS), Santa Inês × Ile de France (FS), and Santa Inês × Bergamacia (BS) crossbred. The animals were slaughtered at the following body weights: 15, 25, 35 or 45 kg. After slaughter and cooling of the carcasses, the following cuts were made: neck, rib/flank, short ribs, loin, blade, and hind limb. The composition in muscle, bone and fat of cuts and ½ carcass were evaluated. The alometric growth of each tissue of the loin, shoulder and leg in relation to the cut weight, and of each tissue of the ½ carcass in relation to the cold carcass weight, was measured. The allometric growth of muscles in the ½ carcass, for BS males and FS and BS females, was early (b<1). The alometric growth of the bone in the cuts and ½ carcass was early (b<1), except for shoulder bone in SS, TS and BS males, and TS females; and leg bone in BS males, with isogonic allometric growth (b=1). The development of fat in the cuts, and ½ carcass was late (b>1), except shoulder fat in BS males, with isogonic growth (b=1). The growth rate of shoulder muscles in SS and FS males, SS and TS females was isogonic (b=1); for TS and BS males, FS and BS females, it was early (b<1). The muscle development in the loin from SS, FS and BS males, and females FS, was isogonic (b=1); for TS males and SS, TS and BS females, it was negative heterogonic (b<1). The allometric growth of the muscle in the loin from SS, TS and FS, and SS and TS females, was isogonic (b=1); and for BS males and TS and BS females, it was negative heterogonic (b<1).

Objetivou-se com esse trabalho comparar estimativas de componentes de variâncias obtidas por meio de modelos lineares mistos Gaussianos e Robustos, via Amostrador de Gibbs, em dados simulados. Foram simulados 50 arquivos de dados com 1.000 animais cada um, distribuídos em cinco gerações, em dois níveis de efeito fixo e três valores fenotípicos distintos para uma característica hipotética, com diferentes níveis de contaminação. Exceto para os dados sem contaminação, quando os modelos foram iguais, o modelo Robusto apresentou melhores estimativas da variância residual. As estimativas de herdabilidade foram semelhantes em todos os modelos, mas as análises de regressão mostraram que os valores genéticos preditos com uso do modelo Robusto foram mais próximos dos valores genéticos verdadeiros. Esses resultados sugerem que o modelo linear normal contaminado oferece uma alternativa flexível para estimação robusta em melhoramento genético animal.

The objective of this study was to compare Gaussian and Robust linear mixed models for the estimation of variance components by REML and Gibbs Sampling, using data from fifty simulated populations consisting of 1,000 animals distributed in 5 generations. Two levels of fixed effect and three hypothetical phenotypic values for a trait, with different levels of contamination were used in the simulations. Additive and residual variance estimates were similar for both REML and Bayesian inference using the Gaussian and Robust model. The best estimates of residual variance in the presence of contaminants were obtained by the Robust model. Estimates of heritability were similar for all models, but regression analyses indicated that predicted genetic values obtained by the robust model were more similar to real breeding values. These results suggest that the contaminated normal linear model is a flexible alternative for robust estimation in animal breeding.

Informações de 56.413 nascimentos de fêmeas da raça Nelore, ocorridos entre 1950 e 2000, em diferentes regiões do País, sob diferentes condições de criação e de manejo, foram usadas para estudar os componentes de (co)variância e os parâmetros genéticos (herdabilidade e correlação genética) dos pesos das vacas ao desmame (P205), a um ano de idade (P365) e ao sobreano (P550), dos pesos do bezerro ao desmame (P205b), a um ano de idade (P365b) e ao sobreano (P550b), bem como da idade ao primeiro parto (IPPM), da vida produtiva (VPM) e da longevidade, todas em meses (LONGM), do intervalo de partos em dias (INTP), da eficiência reprodutiva (ER), dos índices de produção (IP), de produção metabólico (IPM), de produção somado (IPS) e de produção metabólico somado (IPMS). As estimativas dos componentes de (co)variância foram realizadas por máxima verossimilhança restrita assumindo modelo animal. Analisaram-se os efeitos genéticos diretos, maternos e de ambiente permanente, além de efeitos fixos de fazenda, regime alimentar, condição de criação, ano de nascimento, estação de nascimento e sexo do bezerro. As estimativas de herdabilidade apresentaram-se similares entre as diferentes análises, sendo mais altas (0,24 a 0,75) para P205, P365, P550, P205b, P365b, P550b, IP, IPM, IPPM, VPM e LONGM e mais baixas (0 a 0,10) para ER, INTP, IPS e IPMS. Algumas correlações entre efeito genético materno e as características de crescimento foram negativas. De modo geral, correlações negativas entre LONGM e características de crescimento evidenciam o antagonismo genético entre maturidade precoce e longevidade. Entre LONGM e características de reprodução (IPPM) ou produção (índices de produção), as estimativas foram positivas, evidenciando que a fêmea que permanece no rebanho por mais tempo é aquela que possui os melhores índices reprodutivos e produtivos.

In order to estimate the (co)variance components, heritabilities and genetic correlations of longevity and herd life in females, informations of 56,413 births of females from a Nellore cattle herd, from 1950 to 2000, under different rearing and management conditions, were used. It was studied from calf: weaning weight (W205C), yearling weight (W365C), weight in 550 days (W550C), from cows: weaning weight (W205), yearling weight (W365) and weight in 550 days (W550), age at first calving (AFC), calving interval (CI), herd life (RL), longevity (LONG), reproductive efficiency (RE), index production (IP), index of metabolic production (IMP), sum of index production (SIP), sum of index metabolic production (SIMP). The analysis were performed using the MTDFREML software, estimating (co)variance components, assuming animal model. Direct genetic, maternal and permanent environmental effects were analyzed. Fixed effects of farm, feeding program, rearing condition, birth calf year, birth calf season and calf sex, were also analyzed. Heritability estimates showed similar results among the different analyses, higher (0.24 to 0.75) for W205, W365, W550, W205C, W365C, W550C, IP, IMP, AFC, RL, LONG and lower (0 to 0.10) for RE, CI, SIP and SIMP. Some correlations between maternal genetic effect and the growth traits were negative. In general, estimates of genetic and phenotypic correlations among LONG and growth traits were negative, suggesting genetic antagonism between LONG and early maturity, between the LONG and reproductive and productive traits were positive, suggesting the significant effect of reproductive and productive traits on LONG.