A Bayesian approach to prediction of stallion daily sperm output.

This is a study on predicting a stallion’s daily sperm output (DSO) using a statistical model. The research also examined the impact of testicular size on the predictive models for DSO.

Plain Language Overview

The study explores how to predict the daily amount of sperm produced by a stallion, and investigates the influence of the horse’s testicle size on this prediction.

Objective of the Study and Data Collection

• The first objective was to ascertain the precision of predictions for DSO, releasing sperm once a day for 10 days is a standard method to predict DSO—how well this works was examined over four different periods, using 15 different stallions. The statistical method chosen for this investigation is a hierarchical Bayesian model applied via Gibbs Sampling.
• The second objective was to determine how the size of stallions’ testicles, specifically measured in volume (milliliters), affected DSO predictive models. This involved expanding the existing model to include testicular size data.

The Analysis and Findings

• The overall population model showed an initial decrease in total sperm count of approximately 1.54 billion sperm per day until a mean change point at about 4.71 days. After this point, the average DSO was estimated to be approximately 5.28 billion sperm per day.
• For individual stallions, the standard deviation in DSO was around 0.67 billion sperm per day. However, between different stallions, it was larger, about 1.86 billion sperm per day. These figures address the study’s first objective by showing the variance of DSO both within and among different horses.
• When the size of stallion’s testicles was included in the model to address the second objective, an optimal predictive formula of DSO = 0.79 + 0.018 x testicular size (in milliliters) was proposed. This indicates that every milliliter increase in testicle size accounted for an increase of 0.018 billion more sperm cells in the DSO.
• Testicle size accounted for 36.5% of the among-stallion standard deviation in DSO, implying it had significant influence on variability in DSO across different stallions. However, it did not significantly affect the mean number of days needed to achieve the DSO, indicating collection period is consistent regardless of testicle size.