Precision breeding for piglet survival and the potential use of stress-related gastrointestinal microbiota biomarkers to improve porcine wellness
Item statusRestricted Access
Embargo end date15/02/2024
Nguyen, Tuan Quoc
In the UK, piglet survival rate from birth to weaning has reduced from 82.9 to 81.0% per litter from 2014 to 2021 due to a substantial increase in the average number of piglets per litter from 13 to 15. This reduction is associated with a high economic loss and substantial animal health and welfare issues. Therefore, there is great interest in the improvement of piglet survival rates. Piglet survival is a complex trait, controlled by direct genetic effects, namely genes of the piglets that contribute to their survivability and maternal genetic effects, i.e., the genes of the sows that encode the maternal characteristics, such as uterine capacity, placental efficiency, colostrum production and maternal behaviours. Additionally, the size of the litter, the piglet birth weight, and their variations have been reported as the main factors for piglet survival. Subsequently, I focused on stress, a health and welfare issue that influences both the sows and the piglets. Stress can be detrimental to pig production and reproduction, and can alter the pig’s gastrointestinal microbiota composition. There were five main research objectives within this thesis: The first objective was to estimate genetic parameters of peri- and postnatal piglet survival and birth weight at piglet level and their associations with litter size to obtain insight into direct and maternal effects of these traits and their correlations, and to compare these to genetic estimates at sow level. The second objective was to estimate the selection responses on piglet survival, piglet birth weight and litter size from an experiment that aimed to improve piglet survival during the nursing period by selecting, in tandem, on maternal and then direct genetic effects of the trait. The third objective was to estimate genetic parameters of the residual standard deviation of the litter size within each multiparous sow and piglet birth weight within litter and explore the potential of using canalised selection for litter size and birth weight to improve piglet survival. The fourth objective was to identify the microbial biomarkers in the pig’s gastrointestinal and faecal samples that are associated with induced social stress to potentially use in breeding programs, dietary interventions, husbandry practice and management to improve the health and welfare of pigs. The fifth objective was to investigate the influence of induced social stress on the interactions between the pigs’ microbiota and their performance traits. In the first chapter of this thesis, I presented a comprehensive summary of the research area about genetic evaluations on piglet survival, the influence of direct and maternal genetic effects as well as the impact of litter size and birthweight on piglet survival. Thereafter, I included a section on the influence of stress on pig’s health and productivity, including traits such as average daily gain, feed intake and feed conversion ratio, and the pig’s gastrointestinal microbiota. Subsequently, I summarized the literature about the associations between the pigs’ growth performance and their microbiota. I also presented the thesis research objectives which aligned with following chapters in the thesis. In chapter two, I presented a study on the genetic factors that controlled piglet survival and the genetic associations between piglet survival and litter size and piglet birthweight. The objective was to develop a genetic statistical model to evaluate genetic parameters of piglet survival and to evaluate the selection response from a selection scheme to improve piglet survival. Data were from a unique selection experiment on piglet survival, which included information on individual birth weight, piglet survival at birth and weaning of 22,483 piglets from 1765 litters. There were two datasets: the first included data at individual piglet level, using binary piglet survival observation (i.e., dead or alive) at two timepoints, survival at birth, survival during the nursing period (SVNP), and individual piglet birthweight, and the second included data at the sow level, consisting of traits at litter level, such as survival rates, average birth weight and litter size. A Bayesian approach using a threshold multivariate model considering both the direct genetic effect of the piglets and the maternal genetic effect of the sows was applied on the piglet dataset, whereas a Bayesian multivariate linear model was used on the sow dataset. Briefly, the results suggested that all traits had significant heritability and therefore can be improved by breeding. However, there were negative associations between the direct and maternal genetic effects of survival traits as well as between survival traits and litter size, which should be considered when developing selection schemes to improve piglet survival. I then estimated the breeding values of all traits and calculated the corresponding selection responses. The results showed that selecting for piglet survival was highly successful, with significant improvement in the target trait, piglet survival during the nursing period, especially when selecting for the maternal genetic effect. There were favourable correlated responses, such as improvement in piglet survival at birth and piglet birthweight. One of the most notable results of this study was that selection for direct genetic effect and selection for both direct and maternal genetic effects could magnify the negative association between direct and maternal genetic effects in piglet survival and thus reduced its overall genetic improvement. This study was published in the Genetics Selection Evolution journal in March of 2021. Piglet birth weight and sow litter size uniformity can be achieved by the selection to reduce their variations. In chapter three, I explored the opportunity of using canalised selection to enhance the uniformity of litter size and birthweight to improve piglet survival. In the same dataset as chapter two, the new trait of environmental standard deviation of litter size of the sow over its parities was introduced. The results showed that selecting for lowering environmental standard deviation of litter size led to favourable responses, for example, increased litter size and piglet survival rate per litter during the nursing period, decreased standard deviation of piglet birth weight within litter, while having no effect on average piglet birth weight per litter. Since selecting for increased piglet survival rate per litter during the nursing period, for reduced environmental standard deviation of litter size and for reduced standard deviation of piglet birth weight improved SVNP, a weighted combination of these three traits could be the optimal solution to improve piglet survival without reducing litter size. In the fourth chapter, I investigated the influence of social stress on the gastrointestinal microbiota, to identify potential microbial biomarkers for stress in pigs. A social stress treatment including reduced floor (1m2/pig) and feeder space allowance (30cm/pig) and weekly regrouping was applied to 19 pigs for 4 weeks. A control group of 19 pigs had 2m2/pig floor and 60cm/pig feeder space allowance and were in a stable group without regrouping. The applied stressors showed significant influence on the stress group indicated by enhanced cortisol levels, increased lesion scores, and decreased growth performance. I used partial least squares discriminant analysis models based on microbiota profiles generated from the gastrointestinal tract of the pigs to identify the most important microbes for the discrimination between treatment groups. The models resulted in successful classification rates of more than 74%. The results highlighted that there were significant shifts in microbiota abundances in the gastrointestinal tract of the pigs between the stress and control groups. Potentially pathogenic microbial genera were found increased in abundance in the stress group, as well as genera previously reported to be linked with stress and mental disorder in humans. On the other hand, potentially beneficial genera, such as short chain fatty acids producers and fibre-fermenters were depleted in the stress group. The results further support the previously proposed idea of a microbiota-gut-brain axis, though which stress may have triggered a physiological response regulated by the hypothalamic-pituitary-adrenal axis, leading to altered cortisol levels, and changes in the pig’s intestinal and faecal microbiota composition. Growth performance was strongly associated with the gastrointestinal microbiota composition. Therefore, it is of interest to know if and how stress can affect the associations between microbiota and growth performance in pigs. In the fifth chapter, I used partial least squares (PLS) models to predict each performance (feed conversion ratio, average daily gain, daily feed intake, abbreviated as FCR, ADG, and DFI, respectively) based on the gastrointestinal microbial abundances. The Variable Importance in Projection scores and the regression coefficients from the PLS models were used to identify potential biomarkers and directions of their association with the performance traits. The results highlighted that stress altered the associations between the microbiota and performance traits. For example, some beneficial genera had enhanced abundances in pigs with higher DFI, ADG and lower (less feed per kg growth) FCR in the control group, but showed unfavourable associations with DFI, ADG and FCR in the stress group. In contrast, seven other microbial genera were found to be depleted in animals with higher DFI, ADG and lower FCR in control, but were enriched in pigs with higher DFI, ADG and lower FCR in the stress group. These findings provide growing evidence for the pivotal role of stress on the association between microbiota composition and growth performance and underline the importance of considering the factor stress when identifying biomarkers for performance traits in pigs. The final chapter includes a summary of the results of all previous chapters, and a general discussion about opportunities to improve piglet survival by selection, by considering litter size and birth weight information. The importance of biomarkers related to stress based on gastrointestinal microbiota of pigs is discussed and its potential use to improve survival, health and welfare in piglets, sows, and growing pigs.