Give each animal what it needs, and know it before feeding it

My work fits in one sentence: give each animal what it needs, no more and no less, and be able to predict it before feeding it. The question behind it started in my master's at ESALQ: why do two identical steers, from the same pen, eating the same ration, fatten differently? Feeding tables describe the average of a population. I wanted to understand the individual.

The “before” is the hard part. After feeding, anyone can measure. Predicting takes knowing what the feed really supplies, not what the table says; what the animal really needs, which changes with size, body composition and environment; and a mathematical model that does the arithmetic honestly about its own uncertainty. And excess is an error too. Overfeeding is not generosity: it is nitrogen in the groundwater, methane in the atmosphere and money on the ground. Precision has always been an economic agenda and an environmental agenda at the same time.

That program runs through several lines of work:

• Nutrient requirements and feed supply for cattle, sheep and goats: the Cornell Net Carbohydrate and Protein System (CNCPS) family of models, its tropical adaptation, and the Ruminant Nutrition System.
• Growth and body composition: predicting gain, carcass composition and days to finish for individual animals.
• Rumen function: fiber digestion and passage, ruminal pH and volatile fatty acids, in vitro gas production.
• Model evaluation: how to tell whether a model is adequate for the use it is put to.
• Systems thinking: system dynamics and games that teach how structure produces behavior.

Most of the software is free at the Mathematical Nutrition Models website, with documentation and the papers behind each model.

Scientific productivity is not speed; it is accumulated infrastructure: students, collaborators, tools. The investment that looks slowest in the first ten years, training people, building tools, keeping collaborations alive, is exactly what decides the next twenty.

I coordinated a Brazil–United States consortium, funded by CAPES and FIPSE, that brought Brazilian undergraduates to the United States. Since 2011, every summer, I take American students to Brazil to learn about ruminant production there; in Pirassununga, Viçosa, Botucatu and elsewhere, Brazilian institutions have welcomed us. Many of the visiting scientists in my group have been Brazilian. A collaboration with the University of Sassari, in Sardinia, began in 2005 and continues today. In 2022 the American Society of Animal Science gave me the Bouffault International Animal Agriculture Award for this international work.

What I tell students, in six rules, is on the home page; the longer story is in my mini memoir (PDF).

Why publish in peer-reviewed scientific journals? Publication in a journal:

• gives credibility to scientific work, because it is reviewed by peers (experts in the field), and lets you address their concerns, comments and suggestions;
• lets other researchers verify and validate the work, independently and without bias;
• builds trust in the research, because peers can cite it and follow the research program;
• recognizes and documents ideas and concepts, especially original discoveries;
• connects your research to the research of your peers;
• allows criticism and improvement of everyone's research in an organized, coordinated way;
• measures productivity through citation indexes (e.g., the h-index and citation impact), although this has been criticized.

Journals where I have published →

A list of books for those who want to understand the current scientific knowledge in ruminant nutrition, nutrition modeling and contemporary issues in animal production, plus some food-for-thought books.

Recommended books →   My iBooks →