A time-series approach for clustering farms based on slaughterhouse health aberration data.

A large amount of data is collected routinely in meat inspection in pig slaughterhouses. A time series clustering approach is presented and applied that groups farms based on similar statistical characteristics of meat inspection data over time. A three step characteristic-based clustering approach was used from the idea that the data contain more info than the incidence figures. A stratified subset containing 511,645 pigs was derived as a study set from 3.5 years of meat inspection data. The monthly averages of incidence of pleuritis and of pneumonia of 44 Dutch farms (delivering 5149 batches to 2 pig slaughterhouses) were subjected to 1) derivation of farm level data characteristics 2) factor analysis and 3) clustering into groups of farms. The characteristic-based clustering was able to cluster farms for both lung aberrations. Three groups of data characteristics were informative, describing incidence, time pattern and degree of autocorrelation. The consistency of clustering similar farms was confirmed by repetition of the analysis in a larger dataset. The robustness of the clustering was tested on a substantially extended dataset. This confirmed the earlier results, three data distribution aspects make up the majority of distinction between groups of farms and in these groups (clusters) the majority of the farms was allocated comparable to the earlier allocation (75% and 62% for pleuritis and pneumonia, respectively). The difference between pleuritis and pneumonia in their seasonal dependency was confirmed, supporting the biological relevance of the clustering. Comparison of the identified clusters of statistically comparable farms can be used to detect farm level risk factors causing the health aberrations beyond comparison on disease incidence and trend alone.

Evaluation of the sustainability of contrasted pig farming systems: development of a market conformity tool for pork products based on technological quality traits.

A market conformity tool, based on technological meat quality parameters, was developed within the Q-PorkChains project, to be included in a global sustainability evaluation of pig farming systems. The specific objective of the market conformity tool was to define a scoring system based on the suitability of meat to elaborate the main pork products, according to their market shares based on industry requirements, in different pig farming systems. The tool was based on carcass and meat quality parameters that are commonly used for the assessment of technological quality, which provide representative and repeatable data and are easily measurable. They were the following: cold carcass weight; lean meat percentage; minimum subcutaneous back fat depth at m. gluteus medius level, 45 postmortem and ultimate pH (measured at 24-h postmortem) in m. longissimus lumborum and semimembranosus; meat colour; drip losses and intramuscular fat content in a m. longissimus sample. Five categories of pork products produced at large scale in Europe were considered in the study: fresh meat, cooked products, dry products, specialties and other meat products. For each of the studied farming systems, the technological meat quality requirements, as well as the market shares for each product category within farming system, were obtained from the literature and personal communications from experts. The tool resulted in an overall conformity score that enabled to discriminate among systems according to the degree of matching of the achieved carcass and meat quality with the requirements of the targeted market. In order to improve feasibility, the tool was simplified by selecting ultimate pH at m. longissimus or semimembranosus, minimum fat thickness measured at the left half carcass over m. gluteus medius and intramuscular fat content in a m. longissimus sample as iceberg indicators. The overall suitability scores calculated by using both the complete and the reduced tools presented good correlation and the results obtained were similar. The tool could be considered as robust enough to discriminate among different systems, since it was tested in a wide range of them. It also can be used to detect improvement opportunities to enhance sustainability of pig farming systems. The final objective of the study was achieved, since the market suitability tool could be used in an integrated sustainability analysis of pig farming systems.

Meta-analysis on the effects of the physical environment, animal traits, feeder and feed characteristics on the feeding behaviour and performance of growing-finishing pigs.

A meta-analysis, using information from 45 experiments on growing-finishing pigs published in 39 manuscripts, was carried out to determine the simultaneous effects of the physical environment (space allowance, group size, flooring conditions, temperature, presence of enrichment), pig traits (initial body weight (BW) for each studied time interval, sex, genetics), feeder characteristics (water provision within the feeder, feeder design (individual/collective), feeder places/pig, presence of feeder protection) and feed characteristics (feed allowance (ad libitum/restricted), net energy content, crude protein (CP) content), as well as their potential interactions, on the feeding behaviour and performance of growing-finishing pigs. The detrimental effect of low temperature on performance was particularly evident for restricted-fed pigs (P < 0.05). At reduced feeder space allowance, a reduction in the percentage of time spent eating was predicted when increasing initial BW, whereas the opposite was predicted for larger feeder space allowances (P < 0.001). The reduction in visit duration to the feeder in higher BW groups became gradually more important with increasing feeder space allowance (P < 0.01), whereas the increase in the ingestion rate and average daily feed intake (ADFI) with increasing initial BW became smaller with increasing feeder space (P < 0.05). The model predicted a reduction in feed conversion ratio (FCR) with increasing group size (P < 0.05) and floor space allowance (P < 0.01) and on solid floors with or without bedding (P < 0.05). In comparison with other feeders, wet/dry feeders were associated with more frequent but shorter feeder visits (P < 0.05), higher ingestion rates (P < 0.001) and higher ADFI (P < 0.10). The use of protection within individual feeders increased the time spent feeding (P < 0.001), reduced the number of visits per day (P < 0.01), the ingestion rate (P < 0.001) and FCR (P < 0.01) in comparison with other feeder types. Sex modulated the effect of the number of feeder places/pig on FCR (P < 0.05), with a gradual reduction of FCR in entire males and females when increasing feeder space allowance. Genetics tended to modulate the effect of diets' CP content on FCR (P < 0.10). Overall, these results may contribute to the improvement of the welfare and performance of growing-finishing pigs by a better knowledge of the influence of the rearing environment and may help optimize the feeding strategies in current production systems.

Quantitative assessment of the effects of space allowance, group size and floor characteristics on the lying behaviour of growing-finishing pigs.

To obtain quantitative information that can be later used in animal welfare modelling, the relationship between the lying behaviour of growing-finishing pigs (initial body weight (BW) between 19 and 87 kg) and different factors related to the housing conditions, with a potential negative effect on their welfare, was studied by means of a meta-analytical approach. Data from 22 experiments reported in 21 scientific publications were collected. The space allowance, expressed on an allometric basis by means of a k-value (m2/BW0.667), the group size (n) and the floor characteristics (fully and partly slatted v. non-slatted floor), as well as their significant two-way interactions were used as fixed effects, and the experiment was used as a random factor to take into account the interexperiment effect. Further regression analyses were performed on the predicted values of observations in order to improve the adjustment of data. A significant quadratic relationship was established between space allowance (k-value, P < 0.05; squared k-value, P < 0.01) and the percentage of time spent lying. A significant interaction between the k-value and the floor type was also found (P < 0.05), showing that the relationship between space allowance and lying behaviour is affected by the presence or absence of slats. Threshold k-values were obtained using broken-line analyses, being about 0.039 for slatted floors and almost double for non-slatted floors. Compared to other studies, these values suggest that the ability to rest as space availability decreases may be compromised before a reduced performance becomes apparent. Group size did not show a significant effect. Additional information should be added to the model, as further data become available, to adjust the proposed parameters as well as to try to include the effect of other important aspects such as that of ambient temperature.

Breeding for societally important traits in pigs.

Pig breeding programs traditionally focus on the genetic improvement of production and reproduction traits that have a clear economic value. Because citizens and consumers increasingly attach value to traits that have little or no direct relationship to production costs or to the price of a product, pig breeding organizations want to pay more attention to societally important traits, such as the welfare and health of pigs, the ecological effects of pork production, and the healthiness and sensory quality of pork. Most societally important traits have an economic and a noneconomic value and are sufficiently heritable for effective genetic selection, although many genetic and phenotypic (co)-variances still have to be estimated. However, it often is not clear to a pig breeding organization how it may deal with the noneconomic value of breeding-goal traits. In this study, a retrospective selection-index method is proposed to obtain the proper weights for societally important traits in the breeding goal. First, the genetic-progress space for each breeding-goal trait is explored by increasing the weights, in a stepwise manner, to each societally important trait in the breeding goal, starting from zero. Subsequently, a pig breeding organization can adopt the resulting genetic-progress scenario that it considers most sustainable or most acceptable. The weights underlying the adopted scenario are considered to be the proper breeding-goal weights. The noneconomic value of each societally important breeding-goal trait is found by deducting its economic value from its obtained weight and is thereby expressed in monetary units. In addition to obtaining weights for societally important breeding-goal traits, the proposed method offers the possibility to estimate the societal costs of selecting for economic traits only, as well as the societal benefits and the economic costs of selecting for traits with a noneconomic value. The method is therefore a useful tool for the development of sustainable breeding goals. An example has been worked out for a sow-line breeding program.

The importance of a high feed intake during lactation of primiparous sows nursing large litters.

The objective of this study was to investigate whether nursing a large number of piglets has negative effects on lactation and postweaning performance of primiparous sows and whether a greater lactation feed intake can prevent possible negative effects. Data were recorded on 268 ad libitum-fed sows of three genotypes (G1, G2, and G3) in an experiment where litter size was standardized to 8, 11, or 14 piglets during a 4-wk lactation. Compared to G1 and G2, G3 sows were heavier (P < 0.05) and leaner (P < 0.05) at weaning of their litters, lost similar amounts of BW and backfat, and their piglets grew faster (P < 0.05). Compared to G1, feed intake during lactation was higher for G3 sows (P < 0.05), and their risk of a prolonged weaning-to-estrus interval was lower (P < 0.01). Daily feed intake by sows was not affected by litter size in G1 and G3, but it was quadratically affected in G2 (P < 0.05), with a maximum at 10.8 piglets. Backfat loss of the sows increased linearly with litter size (P < 0.05) in G1 and G3. In G2, backfat loss increased only at litter sizes > 9.8 piglets (P < 0.01). Body weight loss of the sow and litter weight gain increased linearly with litter size (P < 0.001). Per extra piglet nursed, sows had a 23% (P < 0.01) higher probability of a prolonged weaning-to-estrus interval. A higher daily feed intake during lactation reduced tissue loss of the sow, increased litter weight gain (P < 0.01), and reduced the probability of a prolonged weaning-to-estrus interval (by 42% per extra kilogram; P < 0.01). Sows with a lower daily body weight loss during first lactation had a larger second litter (1.28 piglets/kg; P < 0.01), and their probability of a prolonged weaning-to-estrus interval was reduced by 61% per kilogram (P < 0.001). With increasing litter size, it is therefore recommended to reduce body weight loss during lactation by stimulating daily feed intake and by genetic selection.

Protein and lipid deposition rates in male broiler chickens: separate responses to amino acids and protein-free energy.

Two experiments of similar design were conducted with male broiler chickens over two body weight ranges, 200 to 800 g in Experiment 1 and 800 to 1,600 g in Experiment 2. The data were used to test the hypothesis that protein deposition rate increases (linearly) with increasing amino acid intake, until energy intake becomes limiting for protein deposition rate. Additional amino acid intake above this point would be deposited less efficiently. An increase in energy intake would increase lipid deposition rate, but should, at low amino acid intakes, not affect protein deposition rate. Each experiment consisted of 18 treatments: two levels of protein-free energy (energy(pf)) intake, combined with nine amino acid to energypf ratios. Protein was balanced for amino acid content and lysine was the first limiting amino acid in the diet. Protein deposition rate increased with additional amino acid intake. No evidence was found that energy(pf) intake limited protein deposition rate at high amino acid intake. Extra intake of energy(pf) increased lipid deposition rate, which was independent of amino acid intake. Where amino acid intake was limiting, additional intake of energy(pf) had generally no effect on protein deposition rate. The marginal efficiency of amino acid utilization for protein deposition did not depend on body weight. The facts are relevant to the modeling of the growth of broiler chickens.

Disease incidence and immunological traits for the selection of healthy pigs. A review.

Disease is a major issue in animal production systems and society demands that the use of medicines and vaccines be reduced. This review describes the breeding approaches that could be used to improve disease resistance and focuses especially on their application to pigs. Disease reduction by genetic means has certain advantages through cumulative and permanent effects, and direct and indirect selection methods are available. Direct selection for disease incidence requires, besides a unique pig identification and disease registration system, challenge routines that are inconvenient in intensive pig production. Indirect selection for the expression of immune capacity may be an alternative but requires detailed knowledge of the different components of the immune system. There is ample opportunity for genetic improvement of the immune capacity because immune traits show substantial genetic variation between pigs. We therefore conclude that indirect selection via immune traits is very interesting, also for practical implementation, and that there is an urgent need for knowledge, within lines, about the genetic relationships between immune capacity traits and resistance to specific diseases or to disease incidence in general. Furthermore, knowledge about the relationship between immune system traits and production traits is needed as well as knowledge about the effect of selection on the epidemiology of disease at a farm/population level and on the host-pathogen interaction and coevolution.

Selection for growth rate or against back fat thickness in pigs is associated with changes in growth hormone axis plasma protein concentration and mRNA level.

Selection for increased growth rate or decreased back fat thickness results in concomitant changes in endocrine and metabolic status. Growth hormone (GH) changes in blood plasma concentration related to selection for growth rate and fat deposition were reported in pigs. The molecular mechanisms regulating selection-induced changes in GH plasma concentration remain largely unknown. We investigated selection-associated changes in GH axis parameters in 2 pig lines selected for increased growth rate (F-line), or decreased back fat thickness (L-line), respectively. First, we investigated selection-associated changes in GH pulse parameters. In both selection lines we found each generation a declining GH peak maximum concentration and area under the GH curve. GH pulse width was not associated with generation number. In both lines generation number was associated with a declined pulse interval, indicating that the number of pulses per day increased on average with 1 pulse per 24 h per generation. Second, plasma concentration of GH axis related Insulin-like growth factor-I (IGF-I) and insulin were investigated. Plasma IGF-I concentration was not associated with generation number in the F-line. Mean plasma insulin concentration declined each generation in both lines. Third, we investigated changes in GH and Pit-1 mRNA levels. In both selection lines GH and Pit-1 mRNA levels increased approximately 50% each generation. The high SD of the GH mRNA levels in both lines may suggest that the GH mRNA levels are pulsatile in vivo. We postulate a molecular mechanism that may explain how selection is associated with increased GH mRNA levels and GH pulse numbers, while lowering GH release per pulse.

Disease-induced variability of genetic correlations: ascites in broilers as a case study.

Breeding against a production disease is complicated by multiple relationships between productivity, disease, and environment. Ascites in broilers is such a disease. The combination of the reasonably well understood etiology (a physiological/pathological cascade due to inadequate oxygen supply) and the practical relevance makes ascites a relevant case for demonstrating and partly revealing these complex relationships. Chickens (n = 2,788) were tested in an ascites-challenging (cold) environment. Genetic analysis of mortality and pathology in combination with performance and physiological traits (especially blood gas traits) revealed ample opportunities for selection against ascites expression. The genetic correlation structure indicated that different mortality traits and pathology traits roughly represent one common characteristic. Direct selection against pathology is more effective than selection on the basis of growth or blood gas traits. The observed negative correlation (-0.26) between productivity and ascites was unexpected. From the etiology of ascites (inadequate supply of oxygen relative to the demand), a positive (unfavorable) correlation was expected. To demonstrate that the actual disease occurrence caused this apparent contradiction, the data from the undiseased subpopulation were reanalyzed. In the undiseased subpopulation, the genetic correlation between productivity and ascites was positive (0.29). This discrepancy was confirmed by comparing regression of ascites expression on actual performance with regression of ascites on independently assessed performance breeding values. The lability of the genetic correlation was explained from complex interactions between productivity, disease susceptibility, and actual occurrence of the disease. The revealed mechanism can be generalized to other production-related diseases and results in systematically lower genetic correlations between disease and productivity. It was inferred that genetic correlations between productivity and such diseases will always be prone to the demonstrated environmental sensitivity, which complicates index selection against production-related diseases.

Evaluation of between-line variation for within-line selection against ascites in broilers.

Within-line opportunities for selection against ascites were studied in a data set comprising a 10-line comparison. The study attempted to reveal whether contrasts between lines provide reliable candidate traits for within-line selection. Mortality was chosen as the reference trait. As no pedigree information was available, a trait was required that related mortality to the nonmortality data. By principal component analysis, such a trait (ASC_INDIC = ascites indicator) was developed from pathology data. The composite trait ASC_INDIC ranked lines well for their mortality figures (r = 0.96), from which it was concluded that ASC_INDIC represents an underlying continuous ascites trait. Between lines, blood gas traits seemed to be the most promising traits. Within lines, performance traits appeared to be highly correlated to ascites. Comparison of within-line variation to between-line contrasts revealed considerable differences. The high correlation of the blood gas traits with mortality was not present within lines. However, although the magnitude was considerably reduced, the nature of the blood gas traits in their relationship to ascites was similar within and between lines. The study primarily demonstrates that contrasts between lines carry systematic but limited information for within-line coherence. Therefore, line contrasts must be interpreted with care when aiming to study genetic variation and coherence within lines.

Messenger ribonucleic acid expression of the MyoD gene family in muscle tissue at slaughter in relation to selection for porcine growth rate.

Livestock meat production capacity is related to muscle fiber numbers and growth. Muscle fibers develop during early embryonic development from proliferating and differentiating myoblasts. Post-natal muscle growth requires satellite cell proliferation and differentiation. Myoblast and satellite cell proliferation and differentiation is regulated by the genes of the MyoD gene family (myogenin, myf-5, myf-6, and MyoD1). Our aim was to study the mRNA expression of these genes in postnatal muscle tissue in relation to porcine selection for growth rate or leanness. Five boars from a line selected for fast growth (F-line) and five boars from a line selected against backfat thickness (L-line) were slaughtered, and biopsies were taken from 12 muscles. Between-line effects, within-line effects in relation to the performance of the pigs, and muscle-specific effects were studied. Comparing the F-line with the L-line revealed significantly greater myogenin, myf-5, and MyoD1 mRNA expression in some muscles of the F-line. The expression of myf-6 showed a tendency for the opposite effect in some muscles. Muscles were ordered by their muscle-specific growth rate (b-value). Within-line evaluation of the data revealed a systematic muscle effect for the myf-6 expression level in the F-line because higher b-values correlated with increased myf-6 expression level. Backfat thickness was negatively related to myogenin expression in the F-line. A relationship was found between myogenin:MyoD1 mRNA expression ratio and meat color/muscle fiber type composition in the L-line. Furthermore, the myogenin:MyoD1 ratio was greater in muscles from F-line boars than in muscles from L-line boars, which relates to the difference between the lines in muscle fiber type. We conclude that the mRNA levels of the MyoD genes in porcine muscle tissue at slaughter showed different relationships to selection for growth rate when evaluated between selection lines and within selection lines.

Fibre area and capillary supply in broiler breast muscle in relation to productivity and ascites.

The objective in this study was to examine the capillary supply in the broiler breast muscle in relation to productivity and the incidence of ascites. Muscle samples (pectoralis superficialis) were collected from 24 pairs of broilers of a commercial paternal strain. The pairs consisted of full sib broilers (brothers and sisters). Of each pair one broiler had a high arterial pressure index, as an indication of ascites, and the other was a healthy chick of the same sex. Data collected were, among others, body weight, arterial pressure index, haematocrit and weight of breast muscle and lungs. Pectoral muscle samples were histologically analysed with respect to capillary density, fibre area and number of capillaries per fibre. Ascitic chickens (measured as high values for heart hypertrophy) had a lower number of capillaries per fibre and tended to have reduced fibre area, although they had a similar capillary density in the pectoral muscle in comparison with healthy chickens. The expected increase in capillary density was not observed. Normally capillary densities decrease in proportion to the increase in fibre size. So ascites has unfavourable relations with capillary traits. Fast growth was not significantly associated with capillary density. However, chickens with a higher percentage breast muscle had a lower capillary density. This implies that high muscularity might be a risk for a diminished oxygen supply to the breast muscle.

Effect of recombinant porcine somatotropin on growth and carcass quality in growing pigs: interactions with genotype, gender and slaughter weight.

Effects of recombinant porcine somatotropin (rpST) on growth, lean tissue growth, feed intake, feed conversion, lean tissue feed conversion, backfat thickness and lean percentage were examined in 96 growing pigs. The experiment used barrows and gilts from the genotypes Duroc, F1 (Dutch Yorkshire x Dutch Landrace) and Pietrain. Half the pigs received 14 mg rpST i.m. twice each week starting at 60 kg; others received a placebo. Pigs had ad libitum access to a diet containing 2,162 kcal net energy and 182 g crude protein per kilogram and were slaughtered at either 100 or 140 kg live weight. From 60 to 100 and from 100 to 140 kg, live weight responses to rpST averaged as follows: daily gain, +4.5 and +19.9%; feed intake, -4.4 and +3.5%; feed conversion, -8.4 and -13.9%; backfat thickness, -13.8 and -22.8%; lean percentage, +4.4 and +8.7%; lean tissue growth rate, +8.6 and +35.8%; and lean tissue feed conversion, -13.1 and -24.9%. No gender x rpST interaction was detected. However, a genotype x treatment interaction was significant for backfat thickness at both slaughter weights, showing a higher response to rpST in Duroc than in Pietrain and F1. Growth performance was improved more by rpST in F1 and Pietrain than in Duroc, especially at higher weights, but carcass traits were improved more by rpST in Duroc. The response to rpST in lean tissue growth rate from 60 to 100 kg was highest in fatter animals (Duroc, barrows), whereas from 100 to 140 kg, response in lean tissue growth rate to rpST was highest in leaner animals (Pietrain, F1, gilts).