G. Simm

Linear and threshold model genetic parameters for disease, fertility and milk production in dairy cattle

This study provides estimates of genetic parameters for various diseases, fertility and 305-day milk production traits in dairy cattle using data from a UK national milk recording scheme. The data set consisted of 63891 multiple lactation records on diseases (mastitis, lameness, milk fever, ketosis and tetany), fertility traits (calving interval, conception to first service, number of services for a conception, and number of days to first service), dystocia and 305-day milk, fat and protein yield. All traits were analysed by multi-trait repeatability linear animal models (LM). Binary diseases and fertility traits were further analysed by threshold sire models (TM). Both LM and TM analyses were based on the generalized linear mixed model framework. The LM included herd-year-season of calving (HYS), age at calving and parity as fixed effects and genetic, permanent environmental and residual effects as random. The TM analyses included the same effects as for LM, but HYS effects were treated as random to avoid convergence problems when HYS sub-classes had 0 or 100% incidence. Because HYS effects were treated as random, herd effects were fitted as fixed effects to account for effect of herds in the data. The LM estimates of heritability ranged from 0•389 to 0•399 for 305-day milk production traits, 0•010 to 0•029 for fertility traits and 0•004 to 0•038 for diseases. The LM estimates of repeatability ranged from 0•556 to 0•586 for 305-day milk production traits, 0•029 to 0•086 for fertility traits and 0•004 to 0•100 for diseases. The TM estimates of heritabilities and repeatabilities were greater than LM estimates for binary traits and were in the range 0•012 to 0•126 and 0•013 to 0•168, respectively. Genetic correlations between milk production traits and fertility and diseases were all unfavorable: they ranged from 0•07 to 0•37 for milk production and diseases, 0•31 to 0•54 for milk production and poor fertility and 0•06 to 0•41 for diseases and poor fertility. These results show that future selection programmes should include disease and fertility for genetic improvement of health and reproduction and for sustained economic growth in the dairy cattle industry.

Genetic aspects of common health disorders and measures of fertility in Holstein Friesian dairy cattle

The purpose of this study was to estimate genetic parameters for measures of fertility and several health disorders in dairy cows. Data consisted of 33732 records, of which 9163 were on heifers, on 305-day milk yield, health disorders and inseminations. Measures of fertility were calculated from calving and insemination dates and included calving interval, days to first service and conception to first service. Health disorders included milk fever, mastitis and lameness. Genetic and phenotypic (co)variances were estimated using restricted maximum likelihood. Heritability estimates for both health disorders and fertility traits were low, ranging from 0·003 to 0·080. All genetic correlations between 305-day milk yield and health and fertility traits, in cows and heifers together, were antagonistic implying that selection for milk yield may have caused a deterioration in health and fertility. The unfavourable correlation between milk yield and health and fertility traits, plus the economic importance of the latter, suggests that future breeding goals should be expanded to include some health disorders and fertility.

Genotype and feeding system effects and interactions for health and fertility traits in dairy cattle

The effects of feeding system, genotype and genotype by feeding system interactions on a range of health and fertility traits were investigated in Holstein Friesian cows at the Langhill Dairy Cattle Research Centre. There were two genetic groups: a selection (S) and control (C) line, housed and managed as one herd. Animals from each group were assigned to either a high concentrate (HC) or low concentrate (LC) feeding system and offered approximately 2500 kg and 1000 kg of concentrate per lactation on the HC and LC diets respectively. Feeding system had a significant effect on milk fever, days to first service and days to first heat. Lactation number had a significant effect (P<0.05) on the incidence of mastitis, ketosis, retained placentas, milk fever and lameness and conception at first service. Effects of genotype were investigated first by comparing the two genetic groups and then by regressions of the health and fertility traits on pedigree index for fat plus protein (PI). Significant effects of PI were found for oestrus not observed, conception at first service, days to first heat, calving interval, days open and days to first service. The regression coefficient for mastitis on PI was also significantly different from zero (P<0.05). There were no statistically significant genetic line by feeding system interactions, indicating that the observed line differences applied to both dietary treatments. Heritabilities for the health traits ranged between 0.00 and 0.08 and for the fertility traits they ranged between 0.02 and 0.15. Selection for high genetic merit in this herd, seems to have led to a deterioration in fertility but not health traits (with the exception of mastitis). These results, in line with those from studies in large populations, suggest that as genetic merit for production rises, fertility and some aspects of health are deteriorating.

Estimation of genetic parameters using health, fertility and production data from a management recording system for dairy cattle

The Dairy Information System (DAISY) was developed to record fertility and health information for use in research and to help farmers manage their farms. Data from 33 herds recording health and fertility over a 6-year period were used to study genetic relationships of several health, fertility and production traits. There were 10 569 records from 4642 cows of all parities. These were used to estimate genetic parameters for health: mastitis, lameness and somatic cell score (SCS), for fertility: calving interval, days to first service, conception to first service and for production: 305-day milk, butterfat and protein yields. Heritabilities for these traits were also estimated for the first three lactations. (Co)variances were estimated using linear, multitrait restricted maximum likelihood (REML) with an animal model. Mastitis and lameness were treated as all-or-none traits. The incidence of these diseases increased with lactation number, which may lead to variance component estimation problems, as the mean is linked to the variance in binomial distributions. Therefore, a method was used to fix the within-lactation variance to one in all lactations while maintaining the same mean. The heritability for SCS across lactations was 0·15. Heritabilities for other health and fertility traits were low and ranged between 0·013 and 0·047. All genetic correlations with the production traits were antagonistic implying that selection for yield may have led to a deterioration in health and fertility. The genetic correlation between SCS and mastitis was 0·65 indicating that indirect selection for improvements in mastitis may be achieved using somatic cell counts as a selection criterion. The potential use of linear type scores as predictors of the health traits was investigated by regressing health traits on sire predicted transmitting abilities for type. The results indicate that some type traits may be useful as future selection criteria.

Effects of interaction between genotype and feeding system on milk production, feed intake, efficiency and body tissue mobilization in dairy cows

The objective of this study was to investigate genotype by feeding system interactions in Holstein-Friesian dairy cows. For this purpose, selection (S) and control line (C) cows, housed and managed at the Langhill Dairy Cattle Research Centre, were offered ad lib. complete mixed diets, with proportions (in total DM) of concentrates, silage and brewers grains of either 20:5:75 (LC) or 45:5:50 (HC), over a full lactation. No significant feeding system × genetic line interactions were observed for a number of traits, describing milk production, feed intake, efficiency and body tissue mobilisation, when compared as treatment means (128 heifer lactations and 249 cow lactations). However, regression coefficients of milk yield (P < 0.01) and condition score (P < 0.05) on pedigree index for fat plus protein yield were significantly different between LC and HC. This indicates that G × E might become of importance in the future, with continued selection for fat plus protein yield.

Variance components for residual feed intake in dairy cows

Residual feed intake (defined as energy intake minus predicted energy requirements based on lactational performance, metabolic live weight and live weight change) was investigated in 377 lactations on 204 Holstein Friesian dairy cows, fed complete mixed diets ad libitum. Restricted maximum likelihood (REML), and an individual animal model with a random additive, and a random permanent environmental, effect were used to estimate the values of the genetic parameters of several traits. When energy requirement for each cow was estimated from phenotypic regressions then the h2 for residual feed intake was estimated to be from 0.30 to 0.38, depending on the way of calculating the energy requirements. Estimates for the genetic correlation between residual feed intake and several other traits showed that this relatively high h2 came from high genetic correlations of residual feed intake with live weight change and condition score. When energy requirements were estimated using coefficients based on partial genetic regressions of energy intake on milk energy yield, metabolic live weight and live weight change then the heritability of residual feed intake was only 0.05. The difference between the estimates of heritability for ‘genetic’ residual feed intake and phenotypic residual feed intake was a consequence of (i) the antagonistic genetic and environmental correlations between live weight change and energy intake and (ii) a strong bias downwards in the estimation of the h2 for genetic residual feed intake. It was concluded that there probably is some additive genetic variance for the residual feed intake in Holstein cows. Consequently, measurement of feed intake does provide genetic information additional to that when only production, live weight and live weight change are measured.

A genetic investigation of various growth models to describe growth of lambs of two contrasting breeds

This study compared the use of various models to describe growth in lambs of 2 contrasting breeds from birth to slaughter. Live BW records (n = 7559) from 240 Texel and 231 Scottish Blackface (SBF) lambs weighed at 2-wk intervals were modeled. Biologically relevant variables were estimated for each lamb from modified versions of the logistic, Gompertz, Richards, and exponential models, and from linear regression. In both breeds, all nonlinear models fitted the data well, with an average coefficient of determination (R2) of > 0.98. The linear model had a lower average R2 than any of the nonlinear models (< 0.94). The variables used to describe the best 3 models (logistic, Gompertz, and Richards) included estimated final BW (A); maximum ADG (B); age at maximum ADG (C); position of point of inflection in relation to A (D, for Richards only). The Richards and Gompertz models provided the best fit (average R2 = 0.986 to 0.989) in both breeds. Richards estimated an extra variable, allowing increased flexibility in describing individual growth patterns, but the Akaikes information criteria value (which weighs log-likelihood by number of parameters estimated) was similar to that of the Gompertz model. Variables A, B, C, and D were moderately to highly heritable in Texel lambs (h2 = 0.33 to 0.87), and genetic correlations between variables within-model ranged from -0.80 to 0.89, suggesting some flexibility to change the shape of the growth curve when selecting for different variables. In SBF lambs, only variables from the logistic and Gompertz models had moderate heritabilities (0.17 to 0.56), but with high genetic correlations between variables within each model (< -0.88 or > 0.92). Selection on growth variables seems promising (in Texel more than SBF), but high genetic correlations between variables may restrict the possibilities to change the growth curve shape. A random regression model was also fitted to the data to allow predictions of growth rates at relevant time points. Heritabilities for growth rates differed markedly at various stages of growth and between the 2 breeds (Texel: 0.14 to 0.74; SBF: 0.07 to 0.34), with negative correlations between growth rate at 60 d of age and growth rate at finishing. Following these results, future studies should investigate genetic relationships between relevant growth curve variables and other important production traits, such as carcass composition and meat quality.

On-line application of visible and near infrared reflectance spectroscopy to predict chemical―physical and sensory characteristics of beef quality

The aim of this study was to assess the on-line implementation of visible and near infrared reflectance (Vis-NIR) spectroscopy as an early predictor of beef quality traits, by direct application of a fibre-optic probe to the muscle immediately after exposing the meat surface in the abattoir. Samples from M.longissimus thoracis from 194 heifers and steers were scanned at quartering 48h postmortem over the Vis-NIR spectral range from 350 to 1800nm. Thereafter, samples from M.longissimus thoraciset lumborum were analysed for colour (L(∗), a(∗), b(∗); 48h postmortem), cooking loss (14 days postmortem), instrumental texture (Volodkevitch, 10 days aged meat; slice shear force, 3 and 14 days aged meat) and sensory characteristics. Vis-NIR calibrations, tested by cross-validation, showed high predictability for L(∗), a(∗) and b(∗) (R(2)=0.86, 0.86 and 0.91; SE(CV)=0.96, 0.95 and 0.69, respectively). The accuracy of Vis-NIR to estimate cooking loss and instrumental texture ranged from R(2)=0.31 to 0.54, suggesting relatively low prediction ability. Sensory characteristics assessed on 14 days aged meat samples showed R(2) in the range from 0.21 (juiciness) to 0.59 (flavour). Considering the subjective assessment of sensory characteristics the correlations of Vis-NIR measurements and several meat quality traits in the range from 0.46 to 0.95 support the use of on-line Vis-NIR in the abattoir. Improvement of predictability was achieved if only extreme classes of meat characteristics have to be predicted by Vis-NIR spectroscopy.

Selection for longevity and yield in dairy cows using transmitting abilities for type and yield

A dynamic programming model was used to derive economic values for the goal traits milk, fat and protein yield and longevity. The economic values derived were £3.37 per % cows surviving to complete lactation four (conditional on having a milk record in the first lactation) and £-0.03, £0.60 and £4.04 per kg for milk, fat and protein yield respectively. In terms of genetic standard deviations the weight for protein, fat, milk and longevity were 1.0, 0.21, —0.25 and 0.55, respectively. Using economic values and genetic (co) variances, weights were derived for milk fat, protein and four linear type traits (chosen out of fifteen on the basis of the genetic correlation with longevity): angularity (angular), foot angle (steeper), udder depth (shallower) and teat length (shorter). Three additive indices were derived, assuming that the breeding goal was for: (i) yield only (PIN), (ii) longevity only (LIN) or (ii) yield and longevity, hence economic merit (ITEM). Selection on ITEM is expected to give a 2% higher annual rate of genetic progress compared with selection on PIN. Efficiency of using ITEM was larger than 0.97 compared with the optimum index, when the real individual economic values increased or decreased by a factor 1.5 or 2.0. Weights for ITEM were calculated assuming that predicted transmitting abilities (PTAs) from complete multivariate analysis were used as index measurements. In the practical situation that index measurements came from (i) separate univariate best linear unbiased prediction (BLUP) evaluations or (ii) two multivariate BLUP evaluations (one for type and one for yield), efficiency of ITEM (compared with the optimum index) decreased with decreasing accuracy of the PTAs and with increasing ratio between number of records for type and yield, or vice versa, but remained close to 100%. Only in the (not practical) situation where accurate PTAs for type and inaccurate information for yield were combined, did the efficiency of ITEM drop as low as 0.44, due to a change of sign for udder depth in the optimal index.

Modelling the spread of scrapie in a sheep flock: evidence for increased transmission during lambing seasons

Summary.Presence of scrapie infectivity in the placenta suggests the possibility of increased transmission of scrapie during the lambing season. This hypothesis was explored here using a mathematical model of scrapie transmission dynamics which has previously been successfully used to study several scrapie outbreaks in Scottish sheep flocks. It was applied here to the Langlade experimental sheep flock (INRA Toulouse, France), in which a natural scrapie epidemic started in 1993. Extensive data were available, including pedigree, scrapie histopathological diagnoses and PrP genotypes. Detailed simulations of the scrapie outbreak reveal that the observed patterns of seasonality in incidence can not be accounted for by seasonality in demography alone and provide strong support for the hypothesis of increased transmission during lambing. Observations from several other scrapie outbreaks also showing seasonal incidence patterns support these conclusions.