In dam rearing, where dairy cows nurse their calves for several weeks while being milked, the farmer may loose considerable amounts of harvested milk, partly due to disturbed alveolar milk ejection. Therefore the objective of this thesis was to improve the machine gained milk yield in dam rearing. The influence of acoustic, olfactory and manual stimulation in the parlour on cows with free or no calf-contact were investigated. Beside parameters of milk let-down, agitation behaviour, heart rate (HR) and heart rate variability (HRV) were assessed to analyse if dams are stressed during milking. This could be a reason for impaired alveolar milk ejection. The second approach was to reduce mother-calf-contact: free, half-day and no cow-calf-contact were compared.
In a preliminary study (Chapter 2), it was investigated whether cows behaviourally respond to calf hair presented in a tandem milking parlour, and whether this is affected by suckling the own calf or not. Discrimination between hair of the own calf in a thin cloth bag (‘own’), hair of an alien calf (‘alien’) and a control cloth bag without calf hair (‘no’) was tested among 17 multiparous and 6 primiparous cows with free calf-contact (‘contact’) and
13 multiparous and 4 primiparous cows separated within 12 h after parturition from their calves (‘control’). Both groups were milked twice daily in a tandem milking parlour, where they were individually tested in six consecutive milkings (trials) starting between the 12th and 20th day of lactation. Two of three olfactory stimuli were presented simultaneously. Sniffing or licking of the stimuli during the first minutes of milking (response duration in % of total observation time) and number of trials with any response (frequency of responses) were recorded. Data were analysed using non-parametric tests. Calf hair (‘own’ or ’alien’) elicited responses in 60% of the animals at least once, but altogether there were only overt responses in 23% of trials. Significant differences in responsiveness towards the different stimuli were found in terms of frequency of responses for all cows (n = 28 without missing data, p = 0.003). Response duration differed significantly for all responsive multiparous cows (n = 12, p = 0.049) and in tendency for all responsive heifers (n = 8, p = 0.061) and for responsive ‘contact’ cows and heifers (n = 11, p = 0.034). In all these cases, responses were highest for ‘own’, intermediate for ‘alien’ and lowest for ‘no’. In the post hoc tests, no significant differences between ‘own’ and ‘alien’ could be detected. Despite low response rates to the presented olfactory stimuli in general, it can be concluded that the responsive multiparous cows and ‘contact’ heifers were able to perceive the presented calf odour and preferred to sniff/lick those stimuli compared to a stimulus with ‘no’ odour.
Out of the cows of the pilot study 15 dams and 22 ‘control’ animals took part in the next experiment (Chapter 3). Both were conducted at the research farm of the Thünen Institute, Trenthorst, where cows were loose housed in two similar cubicle pens with one calf creep available each. In this study, dams and ‘control’ cows were compared during milking concerning machine collected milk yield, machine-on time, milk flow characteristics, milk fat content, somatic cell score (SCS), agitation behaviour, HR and HRV. During three consecutive weeks (26th to 50th d in milk) in each week one of three treatments were conducted during milking in the parlour and results were compared with those from routine milking with vibration stimulation in the same week: played-back calls of calves before milk feeding (acoustic), hair of the own calf in a thin cloth bag (olfactory), teat massage following pre-milking and udder cleaning (in total 60 sec, manual). Mixed models were applied. Over all treatments machine collected milk yield (-9.9 kg per milking), fat content (-0.66%) and milk flow characteristics of dams were lower than in ‘control’ animals (all tests: p < 0.0001, effect size r > 0.70). SCS as indicator of udder health did not differ between groups (p = 0.4111, r = 0.13). There was no impact of ‘contact’ on rumination, stepping, kicking, HR and some parameters of HRV (RMSSD, SDNN, HF%) in the parlour. Dams showed a tense head position (p = 0.0007, r = 0.56) and defecated (p = 0.0125, r = 0.50) at more milkings than cows without calf contact. On the other hand, some characteristics of HRV differed between ‘contact’ animals and the ‘control’ (LF%, LF/HF; p < 0.05, r > 0.30), indicating a higher vagal activity in dams. Reason for this may be a generally higher vagal activity due to suckling, which could also result in higher gut motility and therefore a higher defecation frequency. None of the treatments had great impact on the animals. Manual stimulation increased the mean milk flow during the main milking phase. However, this is possibly due to technical differences compared to vibration stimulation without effects on harvested milk. Acoustic stimulation led to lower SCS compared to routine milking, but only in dams (interaction: p = 0.0023). This result is hard to interpret. In conclusion, it was not possible to enhance milk let-down in dams with free calf-contact through acoustic, olfactory and manual stimulation.
As the applied stimuli did not enhance milk let-down in dams, in a third experiment at the research farm of the University of Kassel the influence of different durations of cow-calf-contact during the day was assessed (Chapter 4). After full-day cow-calf-contact for three days post partum in a calving pen, 11 cows had half-day calf-contact between morning and afternoon milking (ca. 10 h 45 min, ‘half-day’) and 13 cows had free calf-contact
(24 h, ‘free’). Both groups were housed each in a deep litter pen with a calf creep. Cows of the ‘no contact’ group were separated from their calf half a day post partum and were moved into a third deep litter area one day after birth. Control calves were moved to individual calf igloos and were group-housed after the first week of life in a deep litter system. They were fed a maximum of 2x3 L heated whole milk d-1 from teat buckets. All cows were milked twice daily. After nine weeks of ‘nursing’, calves with dam-contact were moved to a pen, were they could see, but not suckle or touch the mother, and were trained to drink whole milk from a teat bucket (2x3 L d-1, ‘in sight+milk feeding’). In the 11th - 12th week of life these calves were housed with the ‘no contact’ calves, where they no longer could see their mothers, and gradual weaning started (4 - 2 L d-1, ‘out of sight+weaning’). Data assessment ended two weeks after weaning (13th - 14th week of life, ‘post weaning’). The influence of contact-treatment and experimental phase on machine milk yield, milk content, SCS and percentage of milkings with SCC >100,000 cells ml-1 were analysed with mixed models. ANOVA was used to analyse the effect of treatment on the average daily milk yield of the lactation (220 - 305 days in milk), calving-interval and daily weight gain of calves during ‘nursing’ and ‘in sight+milk feeding’. After the latter phase all male calves were sold for fattening, resulting in a decreased sample size. Incidences of mastitis and calf data during ‘out of sight+weaning’ and ‘post weaning’ were analysed for potential differences using non-parametric tests.
Daily milk yield during ‘nursing’ was on average 9.9 kg lower in ‘half-day’ compared to ‘no contact’ cows (p = 0.0054, r = 0.48) and 3.6 kg higher compared to ‘free’ cows (p = 0.0576,
r = 0.32). Nearly 80% of the machine milk yield of ‘half-day’ cows could be harvested after the separation from the calf overnight, during morning milking. Over the whole lactation, ‘half-day’ milk yields were in tendency higher than ‘free’ (p = 0.0889, r = 0.31) and not significantly different from ‘no contact’ (p = 0.2193, r = 0.23). In both dam rearing groups milk fat content was about 1 percent point lower during the ‘nursing’ phase. The lower machine milk yield and fat content indicates a disturbed alveolar milk ejection in both groups with calf-contact. This, however, did not negatively affect parameters of udder health. Somatic cell counts did not differ between treatments (p > 0.1), however mastitis incidences were high in all treatments. Protein content was 0.15 - 0.30 percent points higher and lactose content 0.17 - 0.33 percent points lower in dam rearing groups than in ‘no contact’ animals during ‘nursing’. After ‘nursing’ ended, machine milk yield and fat and lactose contents of dams increased. Protein content in ‘half-day’ and ‘no contact’ differed between phases, but remained stable in ‘free’ cows. Calving interval was not affected by calf-contact (p = 0.714). Weight development of dam-reared calves was similar. ‘No contact’ calves’ gained significantly less weight during ‘nursing’ (p < 0.0001) and more during ’in sight+milk feeding’ than ‘half-day’ calves (p < 0.05). During the following weeks, weight gain of dam reared calves increased, but numerically remained under the level of ‘no contact’ calves (p > 0.1). Two weeks after weaning, however, dam reared calves’ body weight was still higher than the body weight of restrictedly fed calves (p < 0.05). In conclusion ‘half-day’ contact helped to reduce milk losses during dam rearing, while calf development did not differ from ‘free’ contact.