In dairy cattle, feelings of negative emotions by daily encountered potential threats such as an unfamiliar object, person, or more dominant herd mate, may lead to the development of extreme physiological and/or behavioural adjustment or abnormalities. Such adjustments could compromise welfare, reduce appetite and productivity. Measurement of laterality indices has been used to indicate the emotional state of dairy cows. Cows that prefer to pass an unfamiliar person in a laneway on the right side (from the cow’s perspective, R cows) are considered more anxious than those preferring to pass on the left (L cows). This is because cows that pass on the right side see the threat with their left eye which is better connected to the right brain hemisphere where the flight or fight centre is located. However, measurement of a laterality index is time consuming, labour intensive, requires specialised infrastructure, and is distressful to the cattle.
Infrared thermography has the potential to measure emotional changes in a remote and non-invasive way. It measures the emissions of radiated heat from the external body surfaces which varies on the basis of subsurface blood flow. The pattern of radiated heat is displayed as a thermogram (thermal image) of pixels varying in colours or shades that indicate different infrared temperatures (IRT). When a cow experiences negative emotion, there is an enhanced secretion of adrenaline, which constricts blood vessels, hence, increasing blood flow within the body system. Consequently, the central hypothesis tested in this PhD thesis was that the IRT of key external body surfaces would be positively associated with R cows, and negatively with milk productivity.
The first experiment (Chapter 2) in this thesis evaluated thermographic, behavioural and lactation parameters of left (less anxious; n = 15) and right (anxious; n = 16) lateralised lactating dairy cows. In agreement with the hypothesis, there were positive associations between IRT of eyes, and coronary band of the forelimbs, with laterality; the R cows had higher IRT temperatures at these sites than the L cows (P < 0.05). In subsequent regressions between IRT measures with behavioural responses and lactation parameters, it was identified that maximum IRT measures were mostly associated with right laterality (P < 0.05), and with more predicting variables and a higher consequent R2 than average IRT. Moreover, there were no differences between the right and left eye, and the right and left coronary band of the forelimb IRT (P > 0.10). Therefore, among the range of IRT measures, in further investigations, the maximum IRT of both eyes averaged ((left+ right eye)/2) and both limbs averaged ((left+ right coronary band of forelimbs)/2) was of focus. Statistical power analysis also indicated that reliable IRT estimates for a group of cows could be made by measuring either 14-16 cows for 2 consecutive or 10-12 cows for 3 days, with two thermograms taken in rapid succession per cow at each daily session (Chapter 3).
In a second experiment (Chapter 4), important measures of the first experiment were evaluated along with flight speed, crush score, and rectal temperature on 50 cows in 3 repeated periods, once in a month apart across 3 consecutive months. Potential confounders identified in the first experiment, such as days in milk (DIM) and temperature humidity index (THI), were also more carefully considered. Results of the second experiment showed that IRT of both eyes and cows’ sniffing behaviour in the forced lateralisation test were positively associated (P = 0.05) in period 1, and rectal temperature was positively associated with both flight speed (P = 0.001) and crush score (P = 0.01) in period 3. Cow waiting time before being milked was negatively associated with limb IRT (P < 0.05) in each of the 3 months and positively associated with the ratio of eyes to limb IRT(P < 0.05) in 2 of the 3 months, whereas no such associations were detected with rectal temperature. In the analysis across months, there were associations between IRT and behavioural indicators: limb IRT was negatively related to slow to medium walking, and the ratio of IRT of eyes to limbs was positively associated with a vertical, rather than horizontal, tail. IRT of eyes showed a tendency for the negative relationship with flight speed (P = 0.07). No associations were detected between laterality and IRT or rectal temperature. Differences in the DIM, milk yield, milk fat content and somatic cell count of cows evaluated in both experiments were implicated as factors that could confound investigations into the relationship between IRT and laterality. In comparison, IRT measures exhibited a greater potential to predict behavioural and consequently productive responses of cows than rectal temperature. The IRT measures had higher adjusted R2 (eyes 86 %; limbs 78 %) than rectal temperature (63 %) in the corresponding regression models for each, and across periods. The repeated periods’ design was also for consideration of the repeatability of the behavioural and IRT measures. Despite IRT measures not showing potential as a predictor of laterality index, IRT, at least for both eyes, was shown to be a repeatable measure in itself, as was the case with laterality index, flight speed and waiting time (P ≤ 0.001).
In both experiments, there were negative associations between IRT of both limbs and waiting time in the dairy prior to milking, and between IRT of both eyes and milk fat content (P < 0.05). This indicates the potential for lower limb IRT to be used to predict relaxed cows and the efficacy of a cooling strategy for the herd. These relaxed cows could produce more milk fat content which might be due to the milk let down process. Consequently, in future experiments, the link between IRT measures and or waiting time and oxytocin level deserve further investigation.
Therefore, it is concluded that IRT of eyes and coronary band of forelimbs of lactating dairy cows, could predict their behavioural responses of emotions and consequent milk productivity in a better way than conventionally used rectal temperature and or other behavioural measures. The IRT of eyes showed potentiality to predict negative emotional responses as measured by flight speed and other behaviours such as sniffing. Lower limb IRT indicated the potential for the use of these measures to non-invasively select relaxed cows and to evaluate the cooling strategy for a herd. The sampling strategy of IRT measures, outlined in this thesis will further improve handling requirements and cow welfare in routine monitoring of the herd.