Milking dairy cows is labour intensive and time consuming. To reduce costs and save time in milking, novel approaches and technologies are growingly important in Indian dairy herds. The technologies aid in supervision and handling of most of the milking operations, with minimum human interventions (Jacobs and Seigford, Reference Jacobs and Siegford2012). Automatic cluster removers (ACR) were described first by Armstrong et al. (Reference Armstrong, Bickert, Gerrish and Spike1970) but have evolved progressively to sense the desired low flow rate at the end of milking, shutting off the milking vacuum and removing the clusters from the udder (Thompson, Reference Thompson1981 and Spencer, Reference Spencer1989). Benefits of using ACR include reduction in milking labour, improvement in ‘mastitis scores’, and less over-milking so improving teat conditions (Tonelli, 1972; Natzke et al., Reference Natzke, Everett and Bray1982; Rasmussen, Reference Rasmussen2004; Tancin et al., Reference Tancin, Ipema, Hogewerf and Macuhova2006). However, ACR settings may require optimisation of minimum flow level to trigger cluster removal according to breed of animal and farm system (Jago et al., Reference Jago, Burke and Williamson2010a; Besier and Bruckmaier, Reference Besier and Bruckmaier2016; Krawczel et al., Reference Krawczel, Ferneborg, Wiking, Dalsgaard, Gregersen, Black, Larsen, Agenas, Svennersten-Sjaunja and Ternman2017). Major equipment manufacturers offer ACR with options for milk flow threshold. Generally, an ACR threshold for milk flow of 0.2 to 0.4 kg/min is common (Burke and Jago, Reference Burke and Jago2010).
Previously, a cow was considered to be milked out when the milk flow rate reduced to 0.2 kg/min, therefore, commercially available ACRs are often set near to that arbitrary threshold (Sagi, Reference Sagi1978). However, the overall milking efficiency of cows may benefit at higher flow rates under certain conditions (Rasmussen, Reference Rasmussen1993). Clarke et al. (Reference Clarke, Cuthbertson, Greenall, Hannah, Jongman and Shoesmith2004) used higher ACR settings for slow milking cows to optimise milking efficiency. Jago et al. (Reference Jago, Burke and Williamson2010a) used ACR settings of 0.2 kg/min and 0.4 kg/min and reported improved milking efficiency with the higher flow rate when using minimal pre-milking preparations. Besier and Bruckmaier (Reference Besier and Bruckmaier2016), in a study based on milk-flow-dependent vacuum levels, proposed early detachment levels (high flow rate) to improve milking performance and teat condition through the avoidance of a sustained period of low milk flow at the end of milking. Optimal milk flow rate is a cow characteristic (Stewart et al., Reference Stewart, Godden, Rapnicki, Reid, Johnson and Eicker2002; Wieland et al., Reference Wieland, Melvin, Virkler and Nydam2016) and hence may determine the ACR setting for herds of different breeds varying in production and milking potential. Crossbred cows vary in their milking characteristics, lactation curve pattern and persistency of milking differing between high yielding exotic and indigenous cows with low production (Fadlemoula et al., Reference Fadlemoula, Yousif and Abu Nikhaila2007; Jingar et al., Reference Jingar, Mehla, Singh and Roy2014). Therefore, it is probably necessary to apply particular ACR settings based on cow breed, production and milking performance to optimise milkability of such cows. This study investigates the optimal ACR settings for Indian cross bred cows under Indian tropical conditions.
Material and Methods
The study was conducted at the Livestock Research Center, ICAR-National Dairy Research Institute (NDRI), Karnal, Haryana, at an altitude of 250 m above mean sea level. The ambient temperature ranges from a maximum 45 °C during summer to a minimum of 0 °C during winter with a diurnal variation of 15–20 °C. Fifty six crossbred Karan Fries cows (Tharparker X HF) in lactation 1 to 4 were selected for the study. The animals were allocated to three groups based on their level of production. The low yield group (N = 16) produced <12 kg/d; the medium yield group (N = 32) produced 12–18 kg/d; the high yield group (N = 8) produced >18 kg/d. The cows averaged 142 ± 51 d in milk and 422 ± 74.6 kg body weight at the start of the study. All the cows were beyond 60 DIM at the beginning of the study. The cows were loose housed under typical local conditions. During the study period the average maximum temperature was 29.2–46 °C and relative humidity 32–89%. Milking was thrice daily, in morning 0500 to 0700 h, in afternoon 1230 to 1330 h and in evening 1730 to 1830 h in a De Laval, low-line, automated herringbone milking parlour. Teats were prepared by washing, drying and fore stripping. The average time for udder preparations was 1.72 ± 0.01 min for each batch (8 cows) involving two trained milkers. The time for cluster attachment was 1.59 ± 0.02 min/batch of cows. The average lag time between fore-stripping and cluster attachment was 51 s for each cow. Strip yields were determined manually after each milking. Hand stripping of milk was done in a graduated cylinder immediately after the cluster removal without oxytocin injection, with a collection time of 15 s for each quarter. The milking vacuum level was 42 kPa, with a pulsation ratio 65:35 and rate of 1 Hz. The ACR settings tested were 0.1, 0.2, 0.3 and 0.4 kg/min with a 13 s delay time. These settings are commonly used at many commercial dairies and parlours in India. After cluster removal all teats were sprayed with Dipal™ (DeLaval. India).
To minimize the effect of variations in milk yield, each cow had each ACR setting for three days, repeated thrice in reverse order. A one day adjustment time was allowed at each setting before collecting data. Milk electrical conductivity (EC) was measured on-line using the milk meters (DeLaval, MM27BC). Somatic cell counts were determined using the DeLaval cell counter (DCC), with a measuring range from 10 000 to 40 00 000 somatic cells/ml. Sampling of the milk was done during morning milking using a proportional sampler attached to the milk meter.
The data on milkability in terms of total milk yield, session milk yield, yield in the first 2 min, machine-on time, average milk flow rate and peak milk flow rate were generated from the ALPRO management system (DeLaval, ALPRO windows 6.90). The number of cluster slips and cluster reattachments were recorded manually for each cow during each session. On-line EC was measured through milk meter equipped with infrared light based senor. The data on somatic cell counts and post milking strip yield were recorded manually for the individual cows.
Statistical analyses
The entire data set was analysed using the general linear model procedure in SAS 9.3 (SAS Inst., Inc., Cary, NC). The effects of altering the ACR settings were analysed using two-way Analysis of variance (ANOVA) and the means were compared using Duncan's Multiple Range Test (DMRT) for the level of significance.
Results and Discussion
The total milk yield was significantly (P < 0.01) affected by ACR setting in high and medium yield groups, but not in the low yield group, indicating that ACR setting may be important in cows with high production. This was more pronounced at the morning milking. Milk production was similar at the 0.1 and 0.2 kg/min settings, but was lower at 0.4 kg/min, in cows with the highest level of production (P < 0.01) (Table 1). Similar results were reported by Sagi (Reference Sagi1978) and Rasmussen (Reference Rasmussen1993) when comparing 0.2 and 0.4 kg/min take-off threshold levels and measuring milk remaining in udders of Holstein cows. The machine-on time was reduced significantly (P < 0.01) for high yielding cows at 0.4 kg/min, whereas, no effect was observed for the low and medium yielding cows. This differs from the results of Jago et al. (Reference Jago, Burke and Williamson2010a) who, comparing ACR settings of 0.2 and 0.4 kg/min, found a significant reduction in machine-on time at the higher ACR setting without significant changes in milk yield. They reported a trend for increased milk yield at higher ACR settings in dairy cows. Several issues may account for the discrepancy between this study and that of Jago et al. (Reference Jago, Burke and Williamson2010a), including differences in cow whole lactation milk production, milk flow rate, milking frequency, and milking equipment. The cows in the present study averaged 4.90 ± 0.06, 6.38 ± 0.09 and 8.92 ± 0.10 kg/milking respectively, for low, medium and high yielding animals as compared with average milk yield of 12.85 ± 1.76 kg/milking reported by Jago et al. (Reference Jago, Burke and Williamson2010a). Cows milked three times per day have less milk per milking, so it is possible that our cows have a lower milk flow rate, and thereby may require lower ACR setting for complete evacuation of the udder. Such a relationship between degree of udder fill and milk flow was reported by Bruckmaier and Hilger (Reference Bruckmaier and Hilger2001). Stewart et al. (Reference Stewart, Godden, Rapnicki, Reid, Johnson and Eicker2002) conducted a field trial for higher ACR settings in five commercial dairy herds of high producing cows and found that average milking duration per cow decreased, but the effects on milk yield were variable. Magliaro and Kensinger (Reference Magliaro and Kensinger2005) reported that changing the ACR setting affected both milk yield and milking time. The results in the present study on crossbred cows agree with Magliaro and Kensinger (Reference Magliaro and Kensinger2005) but differ from Stewart et al. (Reference Stewart, Godden, Rapnicki, Reid, Johnson and Eicker2002) in terms of milk yield. The ACR setting also had no effect on the average and peak milk flow rates in the crossbred cows. Stewart et al. (Reference Stewart, Godden, Rapnicki, Reid, Johnson and Eicker2002) comment that the instantaneous flow rate of a cow is not altered on changing the ACR setting but the lower milking duration, assuming unchanged total milk yield, suggests an increased average flow rate as the proportion of time spent in low flow is lower. The flow rate determined over the first 2 min after cluster attachment varied significantly (P < 0.01) with level of production, but was not affected by the ACR settings (online Supplementary Fig. S1).
Milkability of crossbred cows at different ACR settings in automated herringbone milking parlour
Means bearing different superscript within a row in small letter and within a column in capital letter differ significantly at 0.01.
The initial milk flow rate in the present study was found to be higher in low yielding cows at the beginning of milking (0–30 s). This may be due to a smaller gland cistern compared with medium and high yielding cows. The cows showed a gradual increase in flow rate with significant differences in the flow rate after 30 s of milking. Magliaro and Kensinger (Reference Magliaro and Kensinger2005) reported that the ACR setting may play a role in setting the milk flow rate based on intra-mammary pressure depending on milking frequency and time since last milking. This supports the findings here with thrice daily milking.
The milk flow rate was bimodal for the low yielding cows over the first 2 min of milking, irrespective of the ACR setting. Bruckmaier and Hilger (Reference Bruckmaier and Hilger2001) explained that bimodality in the milk flow curve was associated with the degree of udder fill and teat stimulation before milking. Strapak et al. (Reference Strapak, Antalík and Szencziova2011) suggested that bimodality is connected with the beginning of the milking process, so bimodality may lead to a prolongation of the incline phase and a decrease in the quantity of milk obtained during first minute of milking. The results of the present study were in agreement to Bruckmaier and Hilger (Reference Bruckmaier and Hilger2001) and Strapak et al. (Reference Strapak, Antalík and Szencziova2011) for the low yielding cows which have less udder fill compared with medium and high yielding cows. No bimodality occurred in the milk flow rate curve at the beginning of milking for the medium and high yielding cows. The study, therefore, suggests that thrice daily milking practice is not suitable for low yielding cows.
The rise in EC (Table 2) may be associated with incomplete milking. Wheelock et al. (Reference Wheelock, Rook and Dodd1965) reported that incomplete milking alters milk composition, lactose, potassium, sodium, chloride, whey protein and casein. In low yielding cows, the ACR setting at 0.1 kg/min was associated with a rise in EC (Table 2) may be more due to over-milking. Stewart et al. (Reference Stewart, Godden, Rapnicki, Reid, Johnson and Eicker2002) reported that over-milking occur on herds milked at lower ACR settings. Similar findings were reported by Lewis et al. (Reference Lewis, Cockroft, Bramley and Jackson2000) with over-milking in experimental herds. The EC values in the current study were similar to those of Zaninelli et al. (Reference Zaninelli, Tangorra, Costa, Rossi, Dell'Orto and Savoini2016) who reported a mean EC value of 7.64 ± 0.07 ms/cm for healthy udders/glands. The results showed higher EC values in milk of healthy udders which is due to higher temperature of milk measured for on-line electrical conductivity, as reported by Henningsson et al. (Reference Hennigsson, Ostergren and Dejmek2005) and Gaspardy et al. (Reference Gaspardy, Ismach, Bajcsy, Veress, Markus and Komlosi2012). The somatic cell counts were similar for the different ACR settings (Table 2). The results agree with Burke and Jago (Reference Burke and Jago2010) who reported that retention of small volumes of milk in the udder post cup-removal do not result in an increased SCC, in the absence of infection. Rasmussen (Reference Rasmussen1993) reported that increasing the ACR threshold level had no effect on the incidence and prevalence of clinical and subclinical mastitis in dairy cows. Similarly, Jago et al. (Reference Jago, Burke and Williamson2010a) reported no significant difference in SCC and culling rate based on changes in ACR setting. The current study suggested that for high yielding cows, ACR settings up to 0.4 kg/min do not affect the quality of milk, however, low yielding cows are affected by ACR settings as reflected by the change in EC.
Milk quality and milking irregularities in crossbred cows at different ACR settings
Means bearing different superscript within row in capital letters differ significantly at 0.01 level and in small letters differ significantly at 0.05 level; ** values significant at the 0.01 level.
More cluster reattachments were required at ACR settings of 0.2 kg/min and higher for high yielding cows which exceeds the goal of <5–10 cluster slips and reattachments per 100 cow milkings (Mein and Reid, Reference Mein and Reid1996). Ruegg et al. (Reference Ruegg, Rasmussen and Reinemann2005) reported a wide variation (0–25%) in unit reattachment rate in a survey of Wisconsin dairy operators, which affected their milking efficiency. Cluster reattachment was required due to incomplete milking, evident from the amount of post milking strip yield, when the remaining milk in the udders were removed manually after clusters were detached automatically. Significantly more milk remained in udders at the higher ACR setting at all levels of production (Table 2). The increase in strip yield at the highest ACR setting in the present study was associated with production losses in these crossbred cows. The results are similar to Burke and Jago (Reference Burke and Jago2010) who reported a strip yield of 0.35 kg for ACR at 0.4 kg/min and 0.19 kg for ACR at 0.2 kg/min, with a decrease of 1.2% milk production at the higher ACR setting. However, Jago et al. (Reference Jago, Burke and Williamson2010b) found no significant effect on production due to altered milking end-points, although there were significant differences in average strip yields.
In conclusion, this study on varying ACR settings for milking of Indian crossbred cows determined that, at high levels of production, a higher ACR flow rate threshold decreased machine-on time and milk yield without affecting milk flow rate. More milking irregularities (cluster slips, reattachments, higher milk conductivity) occurred with the higher ACR settings, suggesting the cows are not milked properly. This may result in lower production at same yield potential. In general, an ACR trigger at 0.2 kg/min could be the most appropriate set point under Indian conditions which could harvest the maximum amount of milk from the udders, avoiding milking related irregularities, and improve the quality of production in crossbred dairy cows.
Supplementary material
The supplementary material for this article can be found at https://doi.org/10.1017/S002202991900030X.
Acknowledgements
The authors express sincere thanks to the Director, NDRI, Karnal, for providing necessary facilities and funding for carrying out the research work. We are also grateful to DeLaval for their technical help and support in altering the milking machine settings for the study.
Conflict of interest
The authors declare no conflict of interest with any person or organization in the conduct and reporting of this research.
