#16-103

Complete

Date Full Report Received

05/14/2019

Date Abstract Report Received

05/14/2019

Investigation

Institution:
Primary Investigator:
Co-Investigators: Francesco Tiezzi

Factors determining the seasonal decrease in reproductive performance (otherwise called ‘stressors’) are known to be high temperature and humidity. These conditions are met during the summer and early fall when reproductive performance is usually depressed. We have proven that genetic variation exists among the sows such that some individuals can cope better with these stressors. This genetic variation in the ability to cope with external stressors such as seasonality can be utilized in selection and mating programs to produce sows that are more robust/tolerant to adverse environmental conditions.
The overall objectives of this project were to:

1. Assess the impact of genetic variation on seasonality in 4 nucleus herds of Landrace and Large White sows located in Illinois, Indiana, Nebraska, Texas, North Carolina, and Georgia.
2. Map regions in the pig genome that contribute to improved tolerance to adverse environmental conditions
3. Build a genomic prediction and mating tool that leverages on additive and non-additive genetic effects to breed for more robust sows.

This study made available new selection and mating tools custom-built mating programs, which will be able to find the best combinations of genetic material available to breed animals able to cope with environmental stressors without having a depressed reproductive efficiency.
Results from objective 1 showed that genetic variation exists among individuals for their ability to cope with high temperature and humidity. The first step to obtain these results was to test several covariates, that would affect reproductive performance. Such performance was expressed in farrowing records as Total Number Born, Number Born Alive, and Average Birth Weight. Climatic Data Center data were downloaded and matched to each farm based on geographical distance. For every litter, a summary of weather conditions was drawn, using records of temperature and humidity, either un-transformed or converted into a temperature-humidity index. From the weather records the environmental covariates were calculated for each week of the litter’s cycle from three weeks before breeding through farrowing, generating a total of 19 windows. This allowed us to target a specific time during which heat stress would be particularly detrimental to litter development. We were able to find 2-3 environmental covariates that would affect the reproductive performance and show a differential response among the genetic families within the population. Surprisingly, relative humidity proved a much stronger stress than temperature. This could be explained by the fact that air conditioning systems in nucleus farm are capable of controlling temperature, but not humidity.
We confirmed that different genetic families within a breed (or population) have different tolerance to those stressors. With the models we implemented, we are capable of running a genetic evaluation for maternal traits that provide the following information:

1- A breeding value for reproductive performance on standard conditions, same as traditional genetic evaluations
2- A breeding value for tolerance to a given stressor, since this stressor is proven to impact such performance
3- A breeding value for a specific farm in a given time of the year, since heat stress conditions in that farm can be predicted, so can the reproductive performance of a given individual or mating.

We believe that our models are robust, and we tested such models using cross-validation. By masking phenotypic information of whole groups of sows, we were able to predict such phenotypes with good accuracy. This works in support of the use of these models in genetic evaluation.
Leveraging on a large number of genotypes individuals provided by the industry partners (Smithfield Premium Genetics and The Maschhoffs LLC) we were also able to map regions of the genome that contribute significantly to the genetic variation of the traits analyzed and to the tolerance to heat stress.
With this study, we were able to provide a basis for the exploitation of genetic and genomic variance for sow tolerance to heat stress. We offer a framework for genetic evaluation, although it seems that each trait and population need specific environmental covariates to be used in the modeling. Companies can efficiently implement these evaluations into their breeding schemes.