Predicting Rectal Temperature,Respiration Rate And Milk Yield In Dairy Cows Under Hot Climate

The adverse effect of hot climate on body heat balance is considered as one of the greatest challenges in dairy production.Hot climate can trigger a series of heat strain,including hyperthermia,decline of milk yield,decrease of conception rate,and increase of standing time.To mitigate the adverse effect of hot climate,the forecast of body temperature or respiration rate that denotes body heat balance is necessary for developing cooling strategies.Additionally,the forecast of daily milk yield is essential for estimating production loss.To help the farm managing body heat of cows,this study developed new models for predicting deep body temperature or respiration rate with thermal environment factors and other factors associated with body heat.Moreover,we introduced a new panting-related behavior for predicting core body temperature or respiration rate.Furthermore,we investigated the cross correlation between daily milk yield and heat stress indicators under the hypothesis of dynamic system,and provided a new model for forecasting the variation of daily milk yield.The main results are summarized as follows:Firstly,we investigated the relationships between mean rectal temperature（MRT）and heat stress indicators[air temperature（T_a）and temperature-humidity index（THI）].Those relationships were described with broken line models.The results showed that the normal MRT was between 38.4 and38.6°C.There was a linear relationship between MRT and mean respiration rate（MRR）as follows:MRT=0.021×MRR+37.6,R²=0.925.This equation suggested that each 0.1°C increase of MRT denoted about 4.8 bpm increase of MRR.This equation implies that if MRT exceeds 38.6°C（the upper limit of normal value）,MRR will exceed 48 bpm.Secondly,we developed new models to predict MRR with information of milk yield and time blocks.The milk yield had a moderate correlation with MRR（r=0.44）.There were significant different MRR responses among time blocks（P<0.05）.Time blocks and milk yield accounted for as much as 16%of relative importance in total,suggesting their important role in prediction.The quadratic term of air temperature（T_a²）improved the predictive effect and corrected non-linearity and heteroscedasticity problems.The models combining T_a,T_a²,relative humidity,wind speed,milk yield and time blocks were proposed.Compared with THI equations,the proposed models had better performance on suppressing prediction errors,and had a better sensitivity and accuracy on recognizing heat strain.Thirdly,we developed a new approach to estimate rectal temperature and respiration rate based on the body oscillations driven by fast panting.With the chest movement of fast panting,the cows’head and trunk might show oscillations that are in line with the rhythm of chest movements,which is named as panting-related body oscillations（PBO）.The results showed that the MRT for cows with PBO response was 39.3°C,which is about 0.8°C higher than normal MRT,implying that PBO denotes hyperthermia problems.Additionally,the proportion of cows with PBO（Percent＿PBO）was proportional to MRR of herd（MRR=50.64+0.568×Percent＿PBO,R²=0.87）.This proportion can be used to rapidly estimate the magnitude of heat strain of herd.Fourthly,we hypothesize that daily milk yield（MY）and heat stress indicators constitute a dynamic system.In this system,MY was output series and heat stress indicators（THI and T_a）were input series.MY and the maximum daily air temperature（Ta＿max）satisfied stationary hypothesis.The cross-correlation function between MY and Ta＿max（after pre-whitening）suggested that MY had lag response to Ta＿max 1 d later and the lag response persisted 3 d.The transfer function model for MY and Ta＿max is:MY_t=16.90+0.74×MY_t–₁–0.25×Ta＿max_t–₁,R²=0.817.The transfer function model clearly tracked the fluctuation of MY process.The correction factors for DIM intervals improved the predictive effect of transfer function model.The prediction model combining milk yield and time blocks reflected the complex correlation between hot climate and heat strain,which provided a new method for controlling heat strain of cows precisely.The transfer function model provided theoretical foundation for predicting and controlling milk performance in dairy production.