Milking machine development has advanced over the years, often with a focus on rapid milk harvesting and preservation of the teat canal preservation and, hence, cow longevity. High milking vacuum and overmilking increase the risk of inadequate teat conditions and intramammary infections (IMI) (Hillerton et al., Reference Hillerton, Pankey and Pankey2002). Likewise, various claw vacuum levels influence the milking performance and teat condition (Besier and Bruckmaier, Reference Besier and Bruckmaier2016), with short-term damage in the teat being a consequence of circulation impairment and the teat canal openness. Milk from the teat and udder cisterns starts the milk flow. The ACR should detach the cluster once milk flow has decreased below a pre-set level or switch-point (kg/min), affecting the cluster's time on the teats (Jago et al., Reference Jago, Burke and Williamson2010). The same authors reported that the ACR setting influences milking duration without reducing milk yield. When set optimally, an ACR should maximise milking efficiency (cows milked per operator), preserve teat canal integrity and minimise overmilking (Ginsberg et al., Reference Ginsberg, Dzidic, Rasmussen, Poulet, Manninen, Sigurdsson, Tančin and Bruckmaie2018). Extending milking time is the primary cause of overmilking (Rasmussen and Madsen, Reference Rasmussen and Madsen2000). These aspects are discussed in more detail in the online Supplementary File Supporting Literature.
There are no ISO standards for ACR settings with guidelines to apply for setups (Ginsberg et al., Reference Ginsberg, Dzidic, Rasmussen, Poulet, Manninen, Sigurdsson, Tančin and Bruckmaie2018). The authors concluded that combining a switch-point setting with a maximum cluster attachment time during early or peak lactation reduced milking duration with no negative effects. An optimal termination point can be determined in the milk-flow curve decline phase before the overmilking phase (Tančin et al., Reference Tančin, Ipema, Hogewerf and Mačuhová2006).
The study objective was to assess the effects of three ACR switch-point settings on milking and overmilking durations and on vacuum levels during overmilking and peak milk flow, including cow comfort during milking.
Materials and methods
Experimental design
This study was conducted at the Experimental Farm of the University of Pretoria, South Africa (Hillcrest Campus) with a Waikato milking machine (New Zealand, LP) and Afimilk (Afikim, Israel) advanced management system with a low milk line. The mean daily milk yield was 36 kg. All milking machine settings, except the switch-point volumes, were consistent throughout the study. A VaDia device (Biocontrol AS, Rakkestad, Viken, Norway) determined the machine-on time and vacuum levels at the claw, mouthpiece and pulsator chamber, as described in detail in the Online Supplementary materials and methods.
Milking events
The observed milking events included pre-milking preparations and cow discomfort. Weight shifting or stepping, kicking, teat base swelling and ring swelling were recorded during milking. Clinical mastitis, liner slip and cluster fall-off was defined as incomplete information and this s well as inconclusive VaDia graph data (12.1%) were excluded from the final calculations.
Milking machine settings
The pulsation ratio was 65:35 with 60 cycles/min and the milking (system) vacuum was 43.5 kPa with the second-minute teat end vacuum of 37.6 kPa for the average producers with an average fluctuation of 2.1 kPa and a 5 s delay time. For each switch-point setting, measurements occurred during the morning milking of 25 cows for 15 consecutive days. Cow selection was based on teat size and udder conformation, excluding those with teats shorter than 3 cm and asymmetric udders. The three switch-point settings were evaluated in series with a three-day adaption period among test groups. Flow rates where the milking ends at 0.840 kg/min, 0.630 kg/min and 0.504 kg/min were selected. The switch-point setting of 0.504 kg/ml was used for the machine; increasing volumes were chosen based on previous research (Jago et al., Reference Jago, Burke and Williamson2010; Ginsberg et al., Reference Ginsberg, Dzidic, Rasmussen, Poulet, Manninen, Sigurdsson, Tančin and Bruckmaie2018). Guidelines from Tančin et al. (Reference Tančin, Ipema, Hogewerf and Mačuhová2006) determined the start of milking, defined as when vacuum first increased in the mouthpiece chamber (MPC) or short milk tube (SMT) as well as the start of the incline phase (peak flow), start of overmilking (decline phase) and the end of milking (the point where all MPC and SMT traces returned to the non-milking vacuum level or 0 kPa) (online Supplementary materials and methods and Figure S1).
Statistical analyses
Generalised linear models (GLMs) were applied to the cows with behavioural events to indicate cow discomfort from the total milked per day. The binomial distribution and the logit link function defined the required transformation, enabling a linear model (online Supplementary File). The predicted means were compared using Tukey's test at the 5% significance level. The statistical software GenStat® (VSN International, Hemel Hempstead, UK) was used for the data analysis. The three ACR switch-point settings were compared for machine-on time, overmilking duration, total vacuum, vacuum at peak flow, and vacuum during overmilking. The Shapiro–Wilk test evaluated the normality distribution for each factor, which was not normally distributed for any of these the variables. Therefore, the non-parametric Kruskal–Wallis rank-sum test was used to compare the three ACR switch-point settings. The Kruskal–Wallis test indicated significant differences, therefore, this test was followed by the Wilcoxon Rank-sum test to compare the differences among the separate groups (details in online Supplementary File) for machine-on time, overmilking duration, mean total claw vacuum, claw peak flow vacuum, and claw overmilking vacuum.
Results and discussion
Cow comfort during milking
The cows that kicked or shifted their weight (stepping) towards the end of milking were 2.2%, 5.7% and 0%. Ring formation or slight teat base swelling was evident at 0.3%, 2.2% and 1.9% of all cows at switch-point settings of 0.840 kg/min, 0.630 kg/min and 0.504 kg/min, respectively. Back-transformed means of the proportion of cows with behavioural events indicated cow discomfort per setting (Table 1).
Back-transformed means of the proportion of cows with behavioural events per set
Different superscript lower case letters indicated values differed significantly at the 5% level.
Compelling evidence (P < 0.001) indicates that the switch-point setting of 0.840 kg/ml specified the least proportion of kicking and stepping, teat base swelling, and rings post milking (Table 1). This agreed with Cerqueira et al. (Reference Cerqueira, Araújo, Blanco-Penedo, Cantalapiedra, Sørensen and Niza-Ribeiro2017), establishing that overmilking led to significantly more stepping and kicking behaviour during milking. More cows had somatic cell counts exceeding 200 000 cells/ml of milk at this setting.
Machine-on time
Reduced machine-on time may lower the risk of teat canal damage and reduce overmilking without harming milk production, milk components, or yields. These aspects are discussed in more detail in the online Supplementary File Supporting Literature. Various studies, including the current study, agree that when the switch-point settings increased, machine-on time decreased (Jago et al., Reference Jago, Burke and Williamson2010; Besier and Bruckmaier, Reference Besier and Bruckmaier2016; Ginsberg et al., Reference Ginsberg, Dzidic, Rasmussen, Poulet, Manninen, Sigurdsson, Tančin and Bruckmaie2018; Boloña et al., Reference Boloña, Upton and Reinemann2020). In the current study, machine-on time decreased significantly (P < 0.05) by 4.65% when the switch-point setting was increased from 0.504 kg/min to 0.630 kg/min. However, when each of the switch-point setting were compared separately, the machine-on time indicated no significant differences for 0.840 kg/min vs. 0.630 kg/min and 0.840 kg/min vs. 0.504 kg/min (Table 2). As the ACR switch-point settings approached their optimal setting for the parlour, the improvement in machine-on time was reduced.
Comparison of the variables measured by the VaDia for the three switch-point settings of a dairy cow milking machine.
Superscript letters a and b next to the mean values indicate the significant differences between the Wilcoxon and Kruskal–Wallis tests. P-values for the machine-on time were significant at P ≤ 0.05 level of significance; all other variables were significant at the P ≤ 0.0001 level of significance.
Overmilking duration
Significant differences (P < 0.05) were observed when comparing the mean vacuums at peak flow and overmilking for the three switch-point settings. The liner mouthpiece forms a seal at the teat base, constricting the annular ring. Milk flow between the gland and teat cistern is delayed when overmilking increases, leading to congestion and oedema of the teat (Tančin et al., Reference Tančin, Ipema, Hogewerf and Mačuhová2006; Ginsberg et al., Reference Ginsberg, Dzidic, Rasmussen, Poulet, Manninen, Sigurdsson, Tančin and Bruckmaie2018). This study established a significant decrease of 31.7 s and 27.4 s in overmilking when increasing the switch-point setting from the 0.504 and 0.630 kg/min to the 0.840 kg/min, respectively (Table 2). Overmilking, as a percentage of the total machine-on time, increased as switch-point setting decreased. These findings confirm that overmilking correlates with increasing and decreasing milking and parlour efficiency. Ginsberg et al. (Reference Ginsberg, Dzidic, Rasmussen, Poulet, Manninen, Sigurdsson, Tančin and Bruckmaie2018) established that rear teats were overmilked at a lower ACR switch point than front teats. Rear quarters not only produce more milk but have longer increases and plateau while decreasing milk flow phases (Tančin et al., Reference Tančin, Ipema, Hogewerf and Mačuhová2006).
Mean total milking vacuum
A milking vacuum that is too low may cause the claw vacuum to reverse during the a-phase of pulsation, risking milk reflux back from the claw to the teat tip or into the udder. This may lead to cross infections between quarters and teat canal damage (Ambord and Bruckmaier, Reference Ambord and Bruckmaier2010). The claw vacuum increased significantly from 28.2 kPa to 31.0 kPa when the switch point level was lowered from 0.840 kg/min to 0.504 kg/min, respectively. This finding agrees with those of Besier and Bruckmaier (Reference Besier and Bruckmaier2016), who concluded that a high milking vacuum compensates for a non-avoidable vacuum drop, allowing for good milking performance. The latter increases the risk of adverse effects on the teat end, which can be lessened by increasing the ACR switch point setting (Besier and Bruckmaier, Reference Besier and Bruckmaier2016).
Peak flow claw vacuum
This study established a significantly lower vacuum during peak milk flow at the 0.840 kg/min switch point setting than the two lower settings. It differed little from the vacuum levels during mean and overmilking for similar settings. This might be caused by minor overmilking when starting the milking because of inadequate pre-milking preparations. On the other hand, bimodal milk flow was absent, while a low system vacuum in the low-line milking system was observed. Ambord and Bruckmaier (Reference Ambord and Bruckmaier2010) identified a significant claw vacuum drop of 15 ± 0.7 kPa during peak milk flow. These authors recommend that for optimal milking performance and teat conditions, only a reasonable vacuum drop should occur during peak milk flow with a minimum claw vacuum above 30 kPa. The 28.1 kPa at a switch-point setting of 0.840 kg/min was closer to this value, whereas those of the lower switch point settings were higher.
Mean overmilking claw vacuum
The milk flow rate affects the claw vacuum, especially at high flow rates, as the vacuum in the claw approaches that of the system vacuum during overmilking (Rasmussen et al., Reference Rasmussen, Wiking, Bjerring and Larsen2006). According to Besier and Bruckmaier (Reference Besier and Bruckmaier2016), the system vacuum did not affect teat health at the start of milking unless pre-stimulation was inadequate. The current study established a significant decrease in the mean overmilking claw vacuum of 4.3 kPa when lowering the switch-point setting of 0.840 kg/min to 0.630 kg/min (Table 2).
Comparing mean vacuums at peak milk flow and during overmilking
In this study, the mean claw vacuum at peak flow and overmilking differed significantly for the three switch-point settings (Table 2). No other published work could compare the mean vacuums at peak flow and overmilking. Tančin et al. (Reference Tančin, Ipema, Hogewerf and Mačuhová2006) indicated that a decline in the milking phase duration was more important at the quarter level when considering the physiological response of the cow to milking. Milk flow reduction, nearing the end of milking, lessened the chances of flushing out pathogens while increasing the risk of IMI. At the udder level, the fastest milking quarter determines the end of the plateau phase, whereas, at the quarter level, it is determined by the milk availability in that quarter. Boloña et al. (Reference Boloña, Upton and Reinemann2020) established significant within-udder differences in overmilking between individual quarters varied from 1.28 min for the fastest to 0.47 min for the slowest milking quarter.
In conclusion, this study indicated that when machine-on time was decreased, overmilking was shorter. The vacuum during overmilking was lower when an ACR switch-point setting of 0.840 kg/min was used, compared to 0.504 kg/min. Similarly, overmilking and overmilking vacuum decreased when an ACR switch-point setting of 0.840 kg/min and 0.630 kg/min were compared. It was concluded that the best of the three ACR switch-point settings, which was 0.840 kg/min in this study, can lower the risk of teat-end damage because of shorter machine-on and overmilking times. Decreasing teat-end damage could assist with reducing new IMI and mastitis while improving parlour efficiency.
Supplementary material
The supplementary material for this article can be found at https://doi.org/10.1017/S002202992200053X
Acknowledgements
The authors thank the staff of the experimental farm (Hillcrest Campus) for allowing using the facilities and for all their assistance during the study. The National Research Foundation (NRF) (Grant Number: 120319) and Milk South Africa (contract number: PRJ-211-2018) partly supported this research. Our English language editor is Elizabeth Marx from Academic and Professional Editing Services (APES).
