The present study was carried out at ICAR-National Dairy Research Institute, Eastern Regional Station, Kalyani, West Bengal. The altitude of the city is 9.75 meter above mean sea level, latitude and longitude position being 22°58′30″N and 88°26′04″E, respectively. The weather of Kalyani is hot and humid; the maximum ambient temperature in summer goes up to 39^oC and minimum temperature in winter comes down to about 8^oC. The average annual rainfall is 1000-2000 mm, most of which is received from early June to September.

The experimental animals were comprised of 40 Jersey crossbred cows distributed over 1^st to 3^rd lactations. Experimental observations were taken for 9 months (from September to May). Cows were divided into two groups (20 in each). Group-wise average age, parity, lactation stage, milk yield were similar in both the categories. One group (T₁: Treatment group) was kept in Thermo-comfortable shed (Fig. 1 & 2) having thatched roof with ridge ventilation, more central height and 50% of the open paddock is of soil floor and rest 50% made up of plain cement concrete (PCC). The control group (T₀) cows were kept in existing traditional shed (Fig. 3 & 4) having asbestos sheet as a roof material, less central height and entire floor (100%)of open paddock made up of PCC. The floor of covered area of both T₀ and T₁ shed was made up of PCC.

All the feeding management practices and the feed ingredients were same for the both groups of lactating herd. Concentrate mixture, seasonal green fodder (ad libitum) and straw were made available to complete the nutrient requirements of all the lactating animals. The amount of concentrate requirement was calculated for every animal according to their body weight and milk production. Concentrate feed offered@1.5 kg/day as the maintenance (basal) diet between 8 to 8.30 A.M. and rest of the remaining amount was provided during milking. Clean and wholesome water was supplied throughout the whole day.

Milking was done by semi-automatic machine milking (DeLaval India), twice a day during morning from 6.00 to 8.00 AM and evening from 2.30 to 4.30 PM. The milk was weighed in kg and recorded for individual cow. Before milking, the animals were cleaned and groomed. Udders of cows were thoroughly washed with clean water before the milking. Towels soaked with antiseptic solution were used for wiping of teats and udder just before attaching the teat-cups.

The milk parlour behaviours of cows were recorded during milking operations 1m distance from behind without disturbing them. All the behavioural parameters were observed at weekly intervals by a single observer throughout the study period to avoid personal error variation. All the 40 animals were covered in one week duration during both morning and evening time milking by that single observer i.e., in one day including both time milking all the parameters considered were observed in 5-6 animals on an average on daily basis as all the parameters considered in the study were observed in weekly interval and repeatedly done for 9 months duration.

Temperament of each animal was observed at the milking parlour during the whole process of preparation for milking and actual milking with semi-automated milking machine. Temperament score was evaluated under 1-5 scale as per procedure of Gergovska et al. (2012) and Prasad and Jayalaxmi (2014).

Milk yield (kg) was recorded by a digital weighing balance on routine basis and milk composition (fat, protein, solids not fat) was evaluated by using milkoscreen equipment (indiFoss, Indifoss Analytical Pvt Ltd).

It is the time taken by cows to cross a specified distance in milk parlour after their release from milking. Central passage was marked at 0.5 m intervals. The time interval to cover the measured specific distance was recorded by using stop watch. The parlour leaving speed was calculated as the total distance covered divided by time taken and denoted as meter per second (m/s).

Exit score indicated patterns of movement of cows after releasing from parlour i.e. how animals are coming out of milk parlour at end of milking operations. It was done on a scale of 1 to 3 as described by Lanier and Grandin (2002).

The use of cattle in this experiment was approved by the local institutional animal ethical committee vide institute approved project-NDRI/IRC Project code B-40 and it is an observational study.

The data were analyzed using SPSS software (2007, 16.0 versions). The statistical methods used to analyze the data were General Linear Model and Chi-Square test. The level of significance was considered at 5% (P<0.05).

The overall least squares mean ± SE of temperament score, exit score, parlour leaving speed, milking durations, and milk flow rate (MFR) of Jersey crossbred cows were 2.13±0.04, 1.19±0.03,0.48±0.01 m/s, 229.34±4.48 seconds and 939.63±15.84 g/min, respectively. The mean values of milking behavioural features and milk ability of Jersey crossbred cows of T₁ and T₀ groups’ were given in Table 1.

Jersey crossbred cows kept in Thermo-comfortable sheds (T₁) showed more docile behaviour and exhibited less temperament score as compared to those kept in existing shed and the differences were significant (P<0.01). Milk flow rate, milking durations and exit score were higher in cows of T₁ shed as compared to animals of T₀ shed, however, differences were non-significant. Stepping behaviour shown by cows of existing shed was significantly (P<0.05) higher as compared to Thermo-comfortable shed. Stepping during milking is one of the most important components of milking behaviours shown in the parlour,because it helps in identifying the anxious cows. It gives an idea about nervousness in animals while milking operations is being carried out. More the numbers of stepping while milking more agitated is the cow. Thus, cows of T₀ group exhibited more nervous behaviour in the parlour than T₁ group.

Row-wise means different superscripts differ significantly for groups; Column-wise means different superscripts differ significantly for season (Uppercase Significant P<0.01; Lowercase Significant P<0.05).

The overall least squares mean ± SE of milk yield of Jersey crossbred experimental cows was 6.74 ± 0.03 kg /day. The variations in milk yields of Jersey crossbred cows between groups and seasons were presented in Table 2.

Findings of present study revealed thathousing comfort significantly influenced the daily milk yield of cow. Cows maintained at Thermo-comfortable shed (T₁) produced 190 g more milk /day, which was 2.86% higher than that of cows maintained in traditional shed (T₀). Findings suggested that allocation of proper micro-environmental (temperature, humidity, ventilation etc.) conditions through comfortable housing designs enhanced milk yield (Table 2) and also modulated milking behaviours in Jersey crossbred cows (Table 1).

The overall least squares mean ± SE of morning milk compositions such as fat, solids not fat (SNF) and protein per cent was 4.74±0.04, 9.12±0.01 and 3.95±0.02, respectively. The season and group-wise mean values of morning milk composition were given in Table 3. Present study revealed that fat, SNF and protein percent in milk of cows in experimental shed (T₁) were higher compared to that of cows in control shed (T₀) and values differed significantly. The overall least squares mean ± SE of evening milk compositions such as fat, SNF and protein per cent was 5.48±0.04, 9.14±0.02 and 3.96±0.02, respectively.

Seasonal variations of evening milk composition of Jersey crossbred cows between 2 housing patterns were given in Table 4. In the evening session of milking, fat % were found almost similar and did not differ significantly between two housing systems. Fat, SNF and protein % were found higher during evening session of milking as compared to morning session. Daily milk yield during evening was less as compared to morning. In general, milk compositions showed inverse relationship with milk yield.

Row-wise means different superscripts differ significantly for groups; Column-wise means different superscripts differ significantly for season (A, B Significant P<0.01).

Row-wise means different superscripts differ significantly for groups; Column-wise means different superscripts differ significantly for season (A, B Significant P<0.01; a, b significant P<0.05).

Incidences of different types of eliminative behaviours during milking operations give an idea about nervousness of animal while milking. Table 5 revealed that incidence of defecation, urination and bellowing (vocalization) in milking parlour was less in animals kept in T₁ shed compared to that of T₀, although the differences were non-significant (P>0.05). It indicated that if cows are provided favourable housing environment, they tend to become less nervous during milking operations and might perform better. Moreover, less nervous temperament helps in proper handling and restraining of cows during milking operations. Less defecation and urination facilitate clean milk production and maintaining hygiene of milking parlour.

Milking behaviour of dairy cows had been studied in relation to age, stage of lactation, daily milk yield, temperament of cattle during milking, parity etc. (Dickson et al., 1970, Kumar et al., 2019). Milking behaviour is mainly governed by temperament of cows and measured by expression of temperament score. Other behavioural features are entry order in milking parlour, side preference of milk parlour, exit score, fight speed/score, kicking during milking operation, stepping during milking operation, defecation, urination, vocalization etc. Milking behaviour (temperament) of cows has serious effect on their production efficiency.

Row-wise means different superscripts differ significantly for groups; Column-wise means different superscripts differ significantly for season (A, B Significant P<0.01; a, b significant P<0.05).

Housing of dairy cattle affects milk production and cows’ behaviour in the milking parlour. Overcrowded barns (Wierenga and Hopster, 1990; Leonard et al., 1996; Fregonesi et al., 2007) and long waiting time for getting access to milking parlour (Gomez and Cook, 2010) had been reported to affect the cows eating time availability and post milking lying behaviour. Like present study, pattern of housing of lactating dairy cattle have worse adverse effects on their behaviour and milk yield (Bencsik et al., 2006b; Fregonesi et al., 2007). Changes in housing pattern of lactating cow drops the milk yield for temporary period and huge decline in milk production was observed in older cows as compared to first lactation cows. The results obtained in the present study revealed that housing comfort of lying; feeding, resting and standing byre modulated the cows’ expressions of temperament in milking parlour. The cows of thermo-comfortable shed were more docile, calm and less nervous than those kept in traditional shed and obtained less staying comfort. Thus, the influence of shelter comfort was not only limited to the performance and behaviours in living byre, but also itsimpact was extendedup to their expressivity of behaviour in milking byre. Hence, present results indicated that behavioural rhythm of dairy cows could possibly be altered in favourable directions through shelter management interventions.

In tropical climatic conditions environmental factors are the most important determinants to declined milk yield of cows. Environmental factors such as temperature, relative humidity, solar radiation, and air movement and their interactions often limit the performance of dairy cows (Mandal et al., 2002; West, 2003; Upadhaya et al., 2009). Mandal et al. (2016) reported that under tropical environment Jersey crossbred cows showed a reduction of 170 g milk per cow per day during high stressful conditions.

Improvement in microenvironment of the shed through false roof reduced heat loads and helped maintaining normal physiological indices of cows and thereby showed more daily milk yield per cow (Mandal et al., 2018; Sahu et al., 2018; Sahu et al., 2019). Like present study, housing comfort provided in various forms improved the productivity of dairy cattle by protecting them from extreme climate (Bharambe et al., 2013; Kamal et al., 2014; Patil et al., 2014). In Holstein cows, heat stress reduced daily milk yield by 21% as the THI values went from 68 in the spring to 78 in the summer. Lower milk yields were recorded for confined Holstein cows in Mediterranean climate during spring as compared to summer (Bouraoui et al., 2002).

Lambertz et al. (2014) reported that fat percentage of bulk milk samples was negatively influenced by increasing THI values and the decrease was more pronounced in WG (warm loose housing with grazing) and WI (warm loose housing without grazing) compared with CG (cold loose housing with grazing) and CI (cold loose housing without grazing). In accordance, the protein percentage decreased with increasing THI (P < 0.001). Fat corrected milk was low in WG and WI compared with CG and CI (P < 0.05) in Holstein Friesian cows (Lambertz et al., 2014). Fat and SNF (%) were higher in animals of HF and Jersey breeds having shaded housing as compared to non-shaded housing (Collier et al. 1981). Heat stress significantly reduced milk fat content from 3.58% during the spring to 3.24% during the summer (Bouraoui et al., 2002). Milk protein percentage significantly (P<0.05) decreased as a result of summer heat stress (2.96 vs. 2.88%, respectively for the spring and summer).

In present study, milk compositions were decreased in the cows kept at traditional shed due to stress and uncomfortable micro environment in comparison to cows kept at thermo-comfortable shed. Because of favourable micro-environment in the specially designed experimental shed cows might have performed as per their potency; none the less, cows were able to arrest deterioration of milk quality / composition and quantity. Results indicated that if cows were provided comfortable environment through housing, they could perform to their potentials and also express more favorable milking behavioural features and created differences with less protected ones. Major implication of the present study was that milking temperament, tendency of nervousness and related milk parlour behaviours of dairy cows could be maneuvered in favourable directions through thermo-protective shelter management interventions.