This study investigated the effects of in ovo administration of equol (Eq) on post-hatch growth and hepatic lipid metabolism in broiler chickens. Fertilized eggs (146 eggs/group) were injected with 0 μg (control, Con), 20 μg (low dose, L) and 100 μg (high dose, H) Eq in the albumen on the 7th day of incubation. Except a trend increase in the weight of total fat (P = 0.09), Eq had no effect on growth or liver weight in broilers at 49 days of age. Males presented higher liver and BWs and lower total fat and relative liver weights than females (P < 0.01). However, there were no significant effects of Eq or Eq–gender interactions on growth performance or tissues weight (P > 0.05). With respect to lipid parameters in the serum, the results showed that female broilers presented higher triacyglycerol (TG) and low-density lipoprotein cholesterol concentrations than males, whereas there was no gender difference in serum total cholesterol (TC) or high-density lipoprotein cholesterol (HDLC) concentration (P > 0.05). Eq administration significantly decreased serum TG and TC but increased HDLC concentrations in serum of broilers at 49 days of age (P < 0.05), whereas there were no interactions between gender and Eq (P > 0.05). To elucidate the mechanism behind the significant changes of serum TG and TC levels, the expression of genes involved in lipid metabolism in the liver was investigated in female chickens using reverse transcription-PCR. Carnitine palmitoyl transferase I (CPTI) messenger RNA (mRNA) was significantly upregulated by 20 and 100 μg Eq (P < 0.05). High-dose Eq significantly decreased fatty acid synthase (FAS) and enhanced cholesterol-7alpha-hydroxylase (CYP7A1) mRNA levels in the liver (P < 0.05). Eq had no significant effects on acetyl-CoA carboxylase, sterol regulatory element binding protein-1c, malic enzyme, low-density lipoprotein receptor or 3-hydroxy-3-methylglutaryl coenzyme A reductase mRNA in the liver (P > 0.05). These results in female broilers suggest that Eq decreased blood TG by upregulating CPTI and downregulating FAS mRNA expression in the liver, and that high serum cholesterol levels stimulated CYP7A1 gene transcription in the liver.

Meat and milk produced by ruminants are important agricultural products and are major sources of protein for humans. Ruminant production is of considerable economic value and underpins food security in many regions of the world. However, the sector faces major challenges because of diminishing natural resources and ensuing increases in production costs, and also because of the increased awareness of the environmental impact of farming ruminants. The digestion of feed and the production of enteric methane are key functions that could be manipulated by having a thorough understanding of the rumen microbiome. Advances in DNA sequencing technologies and bioinformatics are transforming our understanding of complex microbial ecosystems, including the gastrointestinal tract of mammals. The application of these techniques to the rumen ecosystem has allowed the study of the microbial diversity under different dietary and production conditions. Furthermore, the sequencing of genomes from several cultured rumen bacterial and archaeal species is providing detailed information about their physiology. More recently, metagenomics, mainly aimed at understanding the enzymatic machinery involved in the degradation of plant structural polysaccharides, is starting to produce new insights by allowing access to the total community and sidestepping the limitations imposed by cultivation. These advances highlight the promise of these approaches for characterising the rumen microbial community structure and linking this with the functions of the rumen microbiota. Initial results using high-throughput culture-independent technologies have also shown that the rumen microbiome is far more complex and diverse than the human caecum. Therefore, cataloguing its genes will require a considerable sequencing and bioinformatic effort. Nevertheless, the construction of a rumen microbial gene catalogue through metagenomics and genomic sequencing of key populations is an attainable goal. A rumen microbial gene catalogue is necessary to understand the function of the microbiome and its interaction with the host animal and feeds, and it will provide a basis for integrative microbiome–host models and inform strategies promoting less-polluting, more robust and efficient ruminants.

Increased activity improves broiler leg health, but also increases the heat production of the bird. This experiment investigated the effects of early open-field activity and ambient temperature on the growth and feed intake of two strains of broiler chickens. On the basis of the level of activity in an open-field test on day 3 after hatching, fast-growing Ross 208 and slow-growing i657 chickens were allocated on day 13 to one of the 48 groups. Each group included either six active or six passive birds from each strain and the groups were housed in floor-pens littered with wood chips and fitted with two heat lamps. Each group was fed ad libitum and subjected to one of the three temperature treatments: two (HH; 26°C), one (HC; 16°C to 26°C) or no (CC; 16°C) heat lamps turned on. Production and behavioural data were collected every 2 weeks until day 57. For both strains, early open-field activity had no significant effects on their subsequent behaviour or on any of the production parameters measured, and overall, the slow-growing strain was more active than the fast-growing strain. Ambient temperature had significant effects on production measures for i657 broilers, with CC chickens eating and weighing more, and with a less efficient feed conversion than HH chickens, with HC birds intermediate. A similar effect was found for Ross 208 only for feed intake from 27 to 41 days of age. Ross 208 chickens distributed themselves in the pen with a preference for cooler areas in the hottest ambient temperature treatments. In contrast, the behaviour of the slow-growing strain appeared to be relatively unaffected by the ambient temperature. In conclusion, fast-growing broilers use behavioural changes when trying to adapt to warm environments, whereas slow-growing broilers use metabolic changes to adapt to cooler ambient temperatures.

The global importance of grasslands is indicated by their extent; they comprise some 26% of total land area and 80% of agriculturally productive land. The majority of grasslands are located in tropical developing countries where they are particularly important to the livelihoods of some one billion poor peoples. Grasslands clearly provide the feed base for grazing livestock and thus numerous high-quality foods, but such livestock also provide products such as fertilizer, transport, traction, fibre and leather. In addition, grasslands provide important services and roles including as water catchments, biodiversity reserves, for cultural and recreational needs, and potentially a carbon sink to alleviate greenhouse gas emissions. Inevitably, such functions may conflict with management for production of livestock products. Much of the increasing global demand for meat and milk, particularly from developing countries, will have to be supplied from grassland ecosystems, and this will provide difficult challenges. Increased production of meat and milk generally requires increased intake of metabolizable energy, and thus increased voluntary intake and/or digestibility of diets selected by grazing animals. These will require more widespread and effective application of improved management. Strategies to improve productivity include fertilizer application, grazing management, greater use of crop by-products, legumes and supplements and manipulation of stocking rate and herbage allowance. However, it is often difficult to predict the efficiency and cost-effectiveness of such strategies, particularly in tropical developing country production systems. Evaluation and on-going adjustment of grazing systems require appropriate and reliable assessment criteria, but these are often lacking. A number of emerging technologies may contribute to timely low-cost acquisition of quantitative information to better understand the soil–pasture–animal interactions and animal management in grassland systems. Development of remote imaging of vegetation, global positioning technology, improved diet markers, near IR spectroscopy and modelling provide improved tools for knowledge-based decisions on the productivity constraints of grazing animals. Individual electronic identification of animals offers opportunities for precision management on an individual animal basis for improved productivity. Improved outcomes in the form of livestock products, services and/or other outcomes from grasslands should be possible, but clearly a diversity of solutions are needed for the vast range of environments and social circumstances of global grasslands.

The combined effects of lucerne (Medicago sativa L.) extract supplementation and ammonia treatment of rice straw (Oryza sativa, variety Thaibonnet) on the ruminal digestion of cell wall components were investigated in six continuous culture systems using a randomised complete block design. Data were fitted to second-order polynomial models. Untreated rice straw had higher contents of ash-free cell wall residues (CWR; 763 v. 687 g/kg dry matter (DM)) and non-cellulosic sugars (191 v. 166 g/kg DM) than treated rice straw. Ammoniation preferentially removed xylose, which resulted in a lower xylose-to-arabinose ratio (5.1 v. 5.8). In absence of lucerne supplementation and ammoniation, degradability coefficients were 0.54, 0.46, 0.58, 0.54, 0.42 and 0.60 for cellulose–glucose, xylose, arabinose, galactose, mannose and uronic acids, respectively. Both factors had significant effects on the microbial degradation of structural polysaccharides. With lucerne extract at an optimal level, ammonia treatment increased ash-free cell wall degradation by more than 10%. The degradability coefficients were increased by ammoniation without any significant interaction with lucerne extract, except for glucose, whose degradability was mostly influenced by lucerne extract in a curvilinear way. The comparison of regression coefficients in cell wall and CWR models suggested that ammoniation improved the degradabilities of xylose, galactose and mannose by partly solubilising the corresponding hemicelluloses and by improving the susceptibility of the remaining fraction to microbial attack, whereas it increased the degradability of arabinose only by favouring microbial attack.

This study investigates the feasibility to predict individual methane (CH4) emissions from dairy cows using milk mid-infrared (MIR) spectra. To have a large variability of milk composition, two experiments were conducted on 11 lactating Holstein cows (two primiparous and nine multiparous). The first experiment aimed to induce a large variation in CH4 emission by feeding two different diets: the first one was mainly composed of fresh grass and sugar beet pulp and the second one of maize silage and hay. The second experiment consisted of grass and corn silage with cracked corn, soybean meal and dried pulp. For each milking period, the milk yields were recorded twice daily and a milk sample of 50 ml was collected from each cow and analyzed by MIR spectrometry. Individual CH4 emissions were measured daily using the sulfur hexafluoride method during a 7-day period. CH4 daily emissions ranged from 10.2 to 47.1 g CH4/kg of milk. The spectral data were transformed to represent an average daily milk spectrum (AMS), which was related to the recorded daily CH4 data. By assuming a delay before the production of fermentation products in the rumen and their use to produce milk components, five different calculations were used: AMS at days 0, 0.5, 1, 1.5 and 2 compared with the CH4 measurement. The equations were built using Partial Least Squares regression. From the calculated R 2 cv, it appears that the accuracy of CH4 prediction by MIR changed in function of the milking days. In our experimental conditions, the AMS at day 1.5 compared with the measure of CH4 emissions gave the best results. The R 2 and s.e. of the cross-validation were equal to 0.79 and 5.14 g of CH4/kg of milk. The multiple correlation analysis performed in this study showed the existence of a close relationship between milk fatty acid (FA) profile and CH4 emission at day 1.5. The lower R 2 (R 2 = 0.76) obtained between FA profile and CH4 emission compared with the one corresponding to the obtained calibration (R 2 c = 0.87) shows the interest to apply directly the developed CH4 equation instead of the use of correlations between FA and CH4. In conclusion, our preliminary results suggest the feasibility of direct CH4 prediction from milk MIR spectra. Additional research has the potential to improve the calibrations even further. This alternative method could be useful to predict the individual CH4 emissions at farm level or at the regional scale and it also could be used to identify low-CH4-emitting cows.

The efficiency of nitrogen utilization will be highest when the amino acid (AA) supply approaches the requirement of the animal. With the availability of different crystalline AA, it is theoretically possible to formulate low-protein diets for growing pigs in which seven AA are co-limiting for performance. In such a diet, the concentration of Lys, Met, Met + Cys, Thr, Trp and Val and a seventh AA would exactly match the requirement. To determine the extent to which low-protein diets can be used, it is important to have reliable information about the requirements for these AA. Isoleucine is often considered the seventh-limiting AA in diets for growing pigs; however, information about the Ile requirement is limited and sometimes conflicting. The purpose of this study was to carry out a meta-analysis of the available literature information to determine the Ile requirement in growing pigs. A total of 46 Ile dose–response experiments were identified that used at least four concentrations of Ile in the diet. Because of differences in experimental design, both the Ile concentration and the response criteria were standardized. In 13 dose–response experiments, there was no indication of a response to an increasing Ile concentration. For the other 33 experiments, a response to the increasing Ile concentration was observed and the Ile requirement estimates ranged from 53% to 114% of that of the National Research Council (1998). An Ile concentration below the requirement resulted in important reductions in both feed intake and growth. A 10% reduction in the Ile concentration (below the requirement) resulted in a 15% reduction in feed intake and a 21% reduction in daily gain. The use of blood products in the diet was the main factor determining whether a response to the Ile concentration was observed or not. Blood meal and blood cells are protein sources with a very low Ile concentration, but with high or very high concentrations of Leu, Val, Phe and His. Some of these AA compete with Ile for catabolic pathways or transport across the blood–brain barrier, thereby potentially increasing the requirement for Ile. In diets without blood products, the Ile requirement appears to be lower than the currently recommended requirement. On the basis of the outcome of this study, we recommend a Ile : Lys requirement ratio of at least 50% on a standardized ileal digestible basis.

This study evaluated different female-selective genotyping strategies to increase the predictive accuracy of genomic breeding values (GBVs) in populations that have a limited number of sires with a large number of progeny. A simulated dairy population was utilized to address the aims of the study. The following selection strategies were used: random selection, two-tailed selection by yield deviations, two-tailed selection by breeding value, top yield deviation selection and top breeding value selection. For comparison, two other strategies, genotyping of sires and pedigree indexes from traditional genetic evaluation, were included in the analysis. Two scenarios were simulated, low heritability (h 2 = 0.10) and medium heritability (h 2 = 0.30). GBVs were estimated using the Bayesian Lasso. The accuracy of predicted GBVs using the two-tailed strategies was better than the accuracy obtained using other strategies (0.50 and 0.63 for the two-tailed selection by yield deviations strategy and 0.48 and 0.63 for the two-tailed selection by breeding values strategy in low- and medium-heritability scenarios, respectively, using 1000 genotyped cows). When 996 genotyped bulls were used as the training population, the sire’ strategy led to accuracies of 0.48 and 0.55 for low- and medium-heritability traits, respectively. The Random strategies required larger training populations to outperform the accuracies of the pedigree index; however, selecting females from the top of the yield deviations or breeding values of the population did not improve accuracy relative to that of the pedigree index. Bias was found for all genotyping strategies considered, although the Top strategies produced the most biased predictions. Strategies that involve genotyping cows can be implemented in breeding programs that have a limited number of sires with a reliable progeny test. The results of this study showed that females that exhibited upper and lower extreme values within the distribution of yield deviations may be included as training population to increase reliability in small reference populations. The strategies that selected only the females that had high estimated breeding values or yield deviations produced suboptimal results.

A total of 72 male weaned pigs were used in a 110-day study to investigate the effect of feeding genetically modified (GM) Bt MON810 maize on selected growth and health indicators. It was hypothesised that in pigs fed Bt maize, growth and health are not impacted compared with pigs fed isogenic maize-based diets. Following a 12-day basal period, pigs (10.7 ± 1.9 kg body weight (BW); ∼40 days old) were blocked by weight and ancestry and randomly assigned to treatments: (1) non-GM maize diet for 110 days (non-GM), (2) GM maize diet for 110 days (GM), (3) non-GM maize diet for 30 days followed by GM maize diet up to day 110 (non-GM/GM) and (4) GM maize diet for 30 days followed by non-GM maize diet up to day 110 (GM/non-GM). BW and daily feed intake were recorded on days 0, 30, 60 and 110 (n = 15). Body composition was determined by dual energy X-ray absorptiometry (n = 10) on day 80. Following slaughter on day 110, organs and intestines were weighed and sampled for histological analysis and urine was collected for biochemical analysis (n = 10). Serum biochemistry analysis was performed on days 0, 30, 60, 100 and 110. Growth performance and serum biochemistry were analysed as repeated measures with time and treatment as main factors. The slice option of SAS was used to determine treatment differences at individual time points. There was no effect of feeding GM maize on overall growth, body composition, organ and intestinal weight and histology or serum biochemistry on days 60 and 100 and on urine biochemistry on day 110. A treatment × time interaction was observed for serum urea (SU; P < 0.05), creatinine (SC; P < 0.05) and aspartate aminotransferase (AST; P < 0.05). On day 30, SU was lower for the non-GM/GM treatment compared with the non-GM, GM and GM/non-GM treatments (P < 0.05). On day 110, SC was higher for the non-GM/GM and GM/non-GM treatments compared with non-GM and GM treatments (P < 0.05). Overall, serum total protein was lower for the GM/non-GM treatment compared with the non-GM/GM treatment (P < 0.05). The magnitude of change observed in some serum biochemical parameters did not indicate organ dysfunction and the changes were not accompanied by histological lesions. Long-term feeding of GM maize to pigs did not adversely affect growth or the selected health indicators investigated.

This review aims to present the different effects produced by a post-weaning intake limitation strategy on the growing rabbit, now largely used by French professional rabbit breeders. Although a quantitative feed restriction leads to slower growth, feed conversion (FC) is improved, particularly when the rabbits are again fed freely, as compensatory growth occurs. This better FC or the healthy rabbit is because of better digestion resulting from slower passage through the intestine, whereas the digestive physiology is slightly modified (morphometry of the intestinal mucosa, fermentation pattern, microbiota). Meat quality and carcass characteristics are not greatly affected by feed restriction, except for a lower dressing-out percentage. One of the main advantages of limiting post-weaning intake of the rabbit is to reduce the mortality and morbidity rate due to digestive disorders (particularly epizootic rabbit enteropathy syndrome). The consequences for animal welfare are debatable, as feed restriction probably leads to hunger, but it reduces the incidence of digestive troubles after weaning. However, the growing rabbit adapts very well to an intake limitation strategy, without any aggressive behaviour for congener. In conclusion, restriction strategies could improve profitability of rabbit breeding, but they should be adapted to any specific breeding situation, according to the national market, feed prices, etc.

A total of eight Simmental heifers (114 ± 3.2 days old and weighing 118 ± 3.8 kg BW) were used to study the effects of feeding method on intake and animal behaviour in a crossover design experiment. Treatments consisted of feeding concentrate and chopped barley straw as (1) choice (CH; concentrate and straw in separate feedbunks) or (2) total mixed ration (TMR; concentrate and straw in one feedbunk). Feeds were offered on an ad libitum basis, but always maintaining a concentrate to straw ratio of 90 to 10. The experiment was performed in two 21-day periods, and sampling was carried out in the last week of each period. At the end of each period, treatment was changed for heifers; hence, the final number of animals per treatment was eight. Intake was recorded over 7 consecutive days. BW was recorded at the beginning and the end of the experiment and on day 21 of each experimental period. Barley straw was coarsely chopped with a chopping machine. Once chopped, all the straw was handled for particle size separation using the 2-screen Penn State Particle Separator and only material of more than 8 mm was used to feed the heifers. Animal behaviour was video-recorded for 24 h on day 2 and day 6 of each experimental period. Concentrate intake and total dry matter intake of heifers fed with the CH feeding method were higher (P < 0.01 and P < 0.05) than when fed with TMR (5.1 and 5.3 v. 4.7 and 5.0 kg dry matter (DM)/day, respectively). Conversely, barley straw was consumed in higher amounts in heifers fed with the TMR feeding method (0.3 v. 0.2 kg DM/day, respectively; P = 0.001). The total NDF intake was similar in both treatments. In contrast, NDF intake from barley straw and physically effective NDF intake were higher in heifers fed with the TMR feeding method than when fed with CH. Feeding method used to feed heifers did not affect the consumption of the different kinds of barley straw particles and eating and drinking behaviours but affected ruminating behaviour. Heifers fed TMR spent more time ruminating than heifers fed concentrate and barley straw separately (376 v. 287 min/day, respectively; P < 0.01). TMR as the feeding method in intensive beef production systems could be a good approach to promote roughage intake.

Increasingly more studies are raising concerns about the increasing consumption of meat and the increasing amount of crops (cereals and oilseeds in particular) used to feed animals and that could be used to feed people. The evolution of this amount is very sensitive to human diets and to the productivity of feed. This article provides a 2050 foresight on the necessary increase in crop production for food and feed in three contrasting scenarios: diets with no animal products; current diets in each main region of the world; and the average diet of developed countries extended to the whole world. We develop empirical aggregate production models for seven world regions, using 43 years and 150 countries. These models realistically account for the contribution of feed from food plants (i.e. plants that would be edible for humans) and of grassland to animal products. We find that the amount of edible crops necessary to feed livestock in 2050 is between 8% and 117% of today's need. The latter figure is lower than that in comparable foresight studies because our models take into account empirical features occurring at an aggregate level, such as the increasing share of animal production from regions using less crop product per unit of animal product. In particular, the expected increase in animal production is estimated to occur mostly in Sub-Saharan Africa and Asia, where the amount of feed from food crops required per unit of animal product proves to be lower than that in other areas. This 117% increase indicates that crop production would have to double if the whole world adopted the present diet of developed countries.

The aim of this work was to study the effects of mastitis induced by intramammary lipopolysaccharide (LPS) challenge on milk oxidative stability, as well as to understand the underlying biochemical processes that cause such changes. LPS challenge was associated with nitric oxide burst from the surrounding mammary epithelial cells and consequently induced nitrosative stress that was induced by the formation of NO2• from nitrite by lactoperoxidase. This response was associated with an ∼3-fold increased formation of hazardous compounds: nitrotyrosines, carbonyls and lipid peroxides. We sustained the involvement of xanthine oxidase as a major source of hydrogen peroxide. In consistent with previous findings, catalase has been shown to play a major role in modulating the nitrosative stress by oxidizing nitrite to nitrate. The current hygienic quality criteria cannot detect mixing of low-quality milk, such as milk with high somatic cells, and nitrite with high-quality milk. Thus, development of an improved quality control methodology may be important for the production of high-quality milk.

For centuries, entire male pigs have been castrated to reduce the risk of boar taint. However, physical castration of pig is increasingly being questioned with regard to animal welfare considerations. Immunization against gonadotrophin releasing hormone (GnRH) provides an alternative to physical castration. Using the currently available commercial product (Improvac®; Pfizer Animal Health), a two-dose regimen of a GnRH vaccine is administered. After the second vaccination, a substantial increase in feed consumption has been reported, which may be associated with increased body fatness and decreased feed efficiency when compared with unvaccinated entire male pigs. The aim of the present study was to investigate the effect of a feed restriction on these traits and on the behaviour of 120 group-housed entire males (five pigs/pen) vaccinated against GnRH. The first vaccination was performed at 62 days of age and the second (V2) at 130 days of age. Pigs were slaughtered in two batches 4 to 5 weeks after V2. They were either offered feed ad libitum over the 22 to 114 kg BW range (AL treatment) or ad libitum up to a maximum of 2.50 (R2.50 treatment) or 2.75 kg/day per pig (R2.75 treatment). Behavioural observations and skin lesion scoring were conducted 1 week before V2, and 1 and 3 weeks after V2. At slaughter, the volumetric lean meat content was measured using an X-ray computed tomography scanner. Between V2 and slaughter, the average feed intakes for the R2.75 and R2.50 treatments were 15% and 22% lower than the average AL feed intake (3.20 kg/day), respectively. Feed restriction was associated with a reduced average daily gain after V2 (846, 932 and 1061 g/day in the R2.50, R2.75 and AL groups, P < 0.01) but had no effect on the feed conversion ratio (3.00 kg feed/kg BW gain on average, P = 0.62). No difference was observed in the lean meat content (71.8%, 70.7% and 70.4% in the R2.50, R2.75 and AL groups, P = 0.14), despite a reduced backfat thickness measured in restrictively fed pigs (12.0, 13.0 and 13.6 mm in the R2.50, R2.75 and AL groups, P < 0.01). Higher skin lesion scores were observed 3 weeks after V2 in R2.50 and R2.75 pigs than in the AL ones (scores 33.4, 27.7 and 25.5, respectively, P = 0.04). These results, combined with an unimproved feed efficiency and no marked change in carcass characteristics, suggest that immunologically castrated pigs should not be restrictively fed during the late finishing period.

Intrinsic postzygotic reproductive isolation is thought to result from the substitution of multiple harmless or beneficial genetic differences between species that are incidentally deleterious when combined in species hybrids, causing hybrid sterility or inviability. Genetic variability for hybrid sterility or inviability phenotypes is, however, rarely assessed in natural populations. Here, we assess variation for Drosophila simulans-encoded maternal factor(s) that cause lethality in D. simulans–Drosophila melanogaster F1 hybrid females. First, we survey genetic variability in the strength of D. simulans-mediated maternal effect hybrid lethality among 37 geographic and laboratory isolates. We find abundant variability in the strength of maternal effect hybrid lethality, ranging from complete lethality to none. Second, we assess maternal effect hybrid lethality for a subset of wild isolates made heterozygous with two so-called hybrid rescue strains. The results suggest that the D. simulans maternal effect hybrid lethality involves a diversity of alleles and/or multiple loci.

Epistasis is an important feature of the genetic architecture of quantitative traits, but the dynamics of epistatic interactions in natural populations and the relationship between epistasis and pleiotropy remain poorly understood. Here, we studied the effects of epistatic modifiers that segregate in a wild-derived Drosophila melanogaster population on the mutational effects of P-element insertions in Semaphorin-5C (Sema-5c) and Calreticulin (Crc), pleiotropic genes that affect olfactory behaviour and startle behaviour and, in the case of Crc, sleep phenotypes. We introduced Canton-S B (CSB) third chromosomes with or without a P-element insertion at the Crc or Sema-5c locus in multiple wild-derived inbred lines of the Drosophila melanogaster Genetic Reference Panel (DGRP) and assessed the effects of epistasis on the olfactory response to benzaldehyde and, for Crc, also on sleep. In each case, we found substantial epistasis and significant variation in the magnitude of epistasis. The predominant direction of epistatic effects was to suppress the mutant phenotype. These observations support a previous study on startle behaviour using the same D. melanogaster chromosome substitution lines, which concluded that suppressing epistasis may buffer the effects of new mutations. However, epistatic effects are not correlated among the different phenotypes. Thus, suppressing epistasis appears to be a pervasive general feature of natural populations to protect against the effects of new mutations, but different epistatic interactions modulate different phenotypes affected by mutations at the same pleiotropic gene.

Genomic selection refers to the use of dense, genome-wide markers for the prediction of breeding values (BV) and subsequent selection of breeding individuals. It has become a standard tool in livestock and plant breeding for accelerating genetic gain. The core of genomic selection is the prediction of a large number of marker effects from a limited number of observations. Various Bayesian methods that successfully cope with this challenge are known. Until now, the main research emphasis has been on additive genetic effects. Dominance coefficients of quantitative trait loci (QTLs), however, can also be large, even if dominance variance and inbreeding depression are relatively small. Considering dominance might contribute to the accuracy of genomic selection and serve as a guide for choosing mating pairs with good combining abilities. A general hierarchical Bayesian model for genomic selection that can realistically account for dominance is introduced. Several submodels are proposed and compared with respect to their ability to predict genomic BV, dominance deviations and genotypic values (GV) by stochastic simulation. These submodels differ in the way the dependency between additive and dominance effects is modelled. Depending on the marker panel, the inclusion of dominance effects increased the accuracy of GV by about 17% and the accuracy of genomic BV by 2% in the offspring. Furthermore, it slowed down the decrease of the accuracies in subsequent generations. It was possible to obtain accurate estimates of GV, which enables mate selection programmes.