Conference on Applied Statistics in Agriculture

STATISTICAL CONSIDERATIONS WHEN USING HYSTERESIS TO ESTIMATE INTERNAL HEAT LOAD IN DAIRY COWS

Water is often used to manage heat stress in dairy cattle. Sprinklers are often placed over the feed bunk or used while cattle are waiting to be milked, however in this experiment cattle were given control over water with a cow-activated shower. Previous studies have focused on how wetting can lower body emperature or reduce respiration rates. An alternative way to investigate this management practice is to examine internal heat loads. Internal heat load can be quantified by fitting a hysteresis loop to daily field data. The hysteresis loop is formed by a phase diagram of body temperature versus an environmental input. Internal heat load is the area inside the loop. The area can be estimated using a number of environmental measures. In this paper three environmental measures are considered: ambientair temperature, the temperature-humidity index and the heat-load index. The two stage harmonic least squares methodis used to estimate internal heat load. Then a Bayesian MCMC model is used to predict internal heat load using the environmental inputs and test the effectiveness of allowing shower access on internal heat load reduction. Voluntary use of a shower reduces internal heat load and the strength of this effect increases with the degree of the heat challenge.