Decision Trees for Predicting the Physiological Responses of Rabbits

Area of Science:

• Animal Science
• Environmental Physiology
• Machine Learning in Agriculture

Background:

• The thermal environment in rabbit houses significantly impacts animal physiology and productivity.
• Maintaining a thermoneutral zone is crucial for rabbit welfare and efficient production.
• Decision-making tools are needed to manage environmental conditions effectively.

Purpose of the Study:

• To develop decision tree models for predicting rabbit physiological responses.
• To correlate environmental variables (temperature, humidity) with physiological indicators (respiratory rate, ear temperature).
• To provide a practical tool for real-time assessment of rabbit welfare.

Main Methods:

• Collected 546 data points from 26 eight-week-old rabbits in a controlled rabbit house environment.
• Utilized environmental variables: dry bulb temperature (t_db) and relative humidity (RH).
• Developed sixty decision tree models to predict respiratory rate (RR) and ear temperature (ET).

Main Results:

• Decision tree models were successfully developed to predict rabbit respiratory rate and ear temperature.
• The ear temperature (ET) prediction model demonstrated superior statistical performance compared to the respiratory rate (RR) model.
• Models effectively link environmental conditions to physiological responses.

Conclusions:

• The developed decision trees offer a practical method for rapid evaluation of rabbit welfare.
• Real-time data from these models can aid producers in making timely environmental management decisions.
• Ensuring satisfactory environmental conditions promotes better rabbit health and productivity.