Transcript:

[0:00]

Hello, my name is Dr. Emily Leishman and today I'm going to be talking about mathematical modeling and some ongoing applications in the equine sector.

[0:09]

First, it's best to define: what is a model? A model is an equation or a set of equations that describe the behavior of a system. In animal science, this system is usually a biological system, like a whole animal. Scientists and engineers have been using models for decades to represent bits of the real world. Typically, these models represent a simplified version of the most essential components of a given system, and it's up to the modeler to determine what those essential components are, which depend on the purpose or the overall goal of the model.

[0:48]

When these models are built, they're used to improve our understanding of how the system works, or to predict how the system will behave under different conditions. There are many types of models, but they can be broadly classified into the following groups:

• Dynamic or static: Dynamic models follow a system over time; static models look at a single point in time.
• Mechanistic or empirical: Mechanistic models are based on underlying biology; empirical models are based solely on relationships within the data.
• Stochastic or deterministic: Stochastic models include randomness and are often used for group predictions; deterministic models provide a single prediction value, often for an individual animal.

In reality, models often combine these characteristics.

[2:18]

Why model in the first place? Why not just observe the system in the real world? The average person or experiment can only monitor a limited number of variables at once, often with limited resources or animals. Models allow us to follow many variables simultaneously and observe how they interact, acting as powerful tools to synthesize knowledge.

[3:02]

For example, if we want to understand how one diet influences the growth of a horse, we might find multiple experiments in the literature. However, differences in breed, sex, and even country of origin among studies add complexity, making it hard to interpret results directly. Models can integrate all of this information into one place, connecting the dots between variables and their interactions.

[5:09]

In animal science, models translate research into practice, helping to predict and manage animal performance, optimize systems, reduce costs, and integrate factors like management, health, genetics, and environment. They can evaluate nutrition programs, optimize diets for goals like feed efficiency or athletic performance, adapt nutrition under stress or disease, and provide diagnostic support tools.

[7:13]

Mechanistic models, in particular, are process-based, built using biological principles to understand causation. For example, rather than simply linking feed intake to body weight gain (an empirical approach), a mechanistic model tracks feed intake through digestion, nutrient absorption, metabolite pools (like amino acids), body composition, and growth, relying on detailed biological knowledge.

[10:02]

Mechanistic models have key applications in opportunity analysis, decision support, performance optimization, and understanding biological systems. They can forecast outcomes for scenarios not yet tested in practice, optimize performance or environmental outcomes for different breeds or conditions, and highlight areas where scientific knowledge is lacking, guiding new research. They are also valuable teaching tools, allowing students to explore systems interactively.

[13:34]

In the equine sector, a major ongoing application is a large-scale project to develop a mechanistic digestion and metabolism model for horses. Equine diets often oversupply macronutrients and undersupply micronutrients, which can harm health and the environment. There is a need for accessible models that help end users apply advanced knowledge in practice.

[15:15]

The project’s goal is to develop an equine metabolism and digestion model using models from other species, published equine research, and new experimental work to fill knowledge gaps. The metabolism model will describe post-absorptive nutrient dynamics in mature horses, tracking pools like fatty acids, glucose, glycogen, and amino acids, and their contributions to body composition and growth. Eventually, it will integrate with a digestion model, feedstuff characterization, optimization routines, and modules for exercise, gestation, and lactation.

[18:20]

The end product will be a digital decision-making tool for the equine sector, capable of ration balancing, performance optimization, environmental footprint reduction, and disease risk reduction. For researchers, it will be a platform to store and explore cumulative biological knowledge, generate and test hypotheses, identify knowledge gaps, and train the next generation of professionals in both nutrition and digital systems.

[20:08]

Thank you for taking the time to listen to our ongoing research. We look forward to sharing the results and progress of this project with all stakeholders in the equine industry. If you have any questions about the research, please reach out and we’ll be happy to answer them.