Skip to main content
ARS Home » Research » Publications at this Location » Publication #167875
Title: EVALUATING MODELLING TECHNIQUES FOR CATTLE HEAT STRESS PREDICTION

Author
item BROWN-BRANDL, TAMI
item JONES, DAVID - UNIV OF NEBRASKA
item WOLDT, WAYNE - UNIV OF NEBRASKA

Submitted to: Biosystems Engineering
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 4/12/2005
Publication Date: 7/7/2005
Citation: Brown Brandl, T.M., Jones, D.D., Woldt, W.E. 2005. Evaluating modelling techniques for cattle heat stress prediction. Biosystems Engineering 91(4):513-524.

Interpretive Summary: Extended periods of high temperature and relative humidity and low airflow can cause severe stress in feedlot cattle. Cattle producers must be made aware of these conditions to help deal with the problem. This study reports the efforts to predict stressful conditions using several mathematical methods. Measures to represent the environment include temperature, humidity, wind speed, and solar radiation. Breathing rate was used as the measure of stress on the cattle. Two types of "fuzzy" models and a neural network were compared to traditional math models. Results showed that the neural network and one "fuzzy" model gave better predictions than traditional models, but the second "fuzzy" model did poorly. Additional work is planned, because each model under-predicted the breathing rate at the most severe weather conditions.

Technical Abstract: Researchers have traditionally modeled animal responses by means of statistical models. This study was conducted to evaluate different modeling techniques. One hundred twenty-eight feedlot heifers were observed during a two-month period during the summer of 2002. Respiration rate and surface temperature were taken on a random sample of 40 animals twice a day. Five different models (two statistical models, two fuzzy inference systems [Sugeno and Mamdani] and one neural network) were developed using 70% of this data, and then tested using the remaining 30%. Results showed that the neural network described the most variation in test data (68%), followed by the Sungeno model which described 66%, the regression models described 59 and 62%, while the Mandami model described only 27%. While the neural network model may be a slightly better approach, the researcher may learn more about responses using a fuzzy inference system approach. With all models tested, respiration rate is over-predicted at low stress conditions and under-predicted at high stress conditions. This would indicate that models are lacking a key piece of input data to make an accurate prediction, which is possibly some form of weather history.