Robotic milking and free fatty acids B. A. Slaghuis *, K. Bos, O. de Jong, A.J. Tudos 2, M.C. te Giffel 2 and K. de Koning Applied research of the Animal Sciences Group of Wageningen UR, PO Box 276, 829 PK Lelystad, the Netherlands. 2 NIZO food research, PO Box 20, 670 BA Ede, the Netherlands Abstract With the introduction of automatic milking (AM) systems, increased levels of free fatty acids (FFA) in milk were observed, which might result in off-flavours in milk and dairy products. The aim of this study was to investigate the factors contributing to elevated FFA levels: influence of the milking frequency, technical parameters of the milking system, and finally, farm management aspects. Milking frequency was studied in a Latin square design with milking intervals of 4, 8 and 2 hours and showed increased FFA -levels for the shorter intervals. Technical factors were studied in a laboratory study using milking machine components of AM-systems and of conventional systems. With susceptible milk, differences in increase in FFA levels were found, but results were difficult to interpret. Tests will be repeated. Some FFA problems remained after solving technical and milking frequency problems. Indications were found that feed composition and feeding regime might influence susceptibility of milk for FFA-formation. These farm management aspects are subject of ongoing research. Introduction On the introduction of milk lines and farm milk tanks in the seventies, problems with elevated levels of free fatty acids (FFA) were observed, resulting in off-flavours in milk and dairy products. Therefore, measurement of FFA was integrated in the quality control system in the Netherlands (determination twice a year, in spring and autumn). The problems of elevated FFA levels were tackled practically by avoiding blind pumping, reducing air leakage and improving maintenance of the milking equipment. Especially the introduction of a national milking machine maintenance program contributed largely to overcome problems with FFA. When milking three or more times a day instead of two times is applied, increased milk yields, decreased fat and protein contents and increased free fatty acid levels have been reported (Ipema and Schuiling, 992, Jellema, 986, Klei et al., 997, Svennersten-Sjauna et al, 2002). With the introduction of AM systems, the problem of increased FFA levels (also known from farms milking 3 times per day) occurred again (Klungel et al., 2000,Vorst and Koning, 2002). This problem seems due to the higher average milking frequency with AM systems and to technical features of this system compared to conventional milking systems. As elevating FFA levels are a result of lipolysis, the possible mechanism of lipolysis is explained below. Possible mechanism of lipolysis Lipolysis is enzymatic hydrolysis of milk fat by milk lipoprotein lipase (mlpl), causing accumulation of FFA. The milk fat present in milk fat globules is protected
against the action of mlpl by the milk fat globule membrane. If this membrane is disrupted by e.g. agitation, fatty acids can be split from the glycerol part by mlpl. But in untreated milk, lipolysis also occurs due to spontaneous lipolysis. Jellema (986) defined susceptibility as FFA level in untreated milk after 24 hours storage at 4 C. Late lactation milk and milk from cows milked three times per day or more is more susceptible to spontaneous lipolysis (Jellema,986). The milk fat globule is not disrupted, but some factors (activators) present in milk may favour the interaction of mlpl with milk fat resulting in higher degree of FFA. Activators are related to blood serum components, as addition of blood serum to raw milk increases FFA levels (Jellema, 975,986). Inhibitors, that inhibit contact between mlpl and milk fat globule membrane, have been determined as proteose pepton component 3 fractions (Cartier et al., 990). Several treatments, referred to as activation treatments, enhance lipolysis (induced lipolysis) and are often used to study lipolysis of milk fat by endogenous mlpl. Treatments which cause activation include agitation, homogenisation, temperature changes and the addition of blood serum or heparin to milk. Milk from individual cows differs in susceptibility to these treatments. Correlations between spontaneous and induced lipolysis have been reported (Jellema, 986; Chazal and Chilliard, 987; Cartier and Chilliard, 989). The mechanisms that promote lipolysis are not fully understood. The aim of this study was to investigate the factors contributing to elevated FFA concentrations: influence of the milking frequency, technical parameters of the milking system, and finally, farm management aspects. Material and methods Milking frequency In the first study spontaneous lipolysis (Cartier and Chilliard, 990) was determined in a Latin square design experiment with 2 cows, three periods of four days, three milking intervals (4, 8 and 2 hours corresponding to 6, 3 and 2 times milking a day) and 4 cows per group (Table.). Cows were selected to form equal groups (parity and lactation stage). Cows were milked according to this scheme for 4 days and on day 4 all cows were milked conventionally and sampled. <Table.> Milk samples were divided into two sub samples: one was inactivated by hydrogen peroxide immediately (0,02%, Jellema, 979) to stop lipolysis, the other one was stored at 5 C ± C for 24 hours and then inactivated with hydrogen peroxide. Difference between the sub samples is defined as spontaneous lipolysis (Cartier and Chilliard, 990). Samples were analysed according to the BDI-method (IDF,99). Technical parameters of the AM-system Fresh milk, susceptible for lipolysis (from 4-5 cows at the end of lactation), was divided in two parts. One part was passed through a reconstructed AM-system (two brands) and milk was sampled before and after passing the system. The other part of milk was passed through a conventional system and sampled before and after passing the system. The milk was mixed and used again for a repetition of the experiment. Farm management In a study on variation of FFA levels in farm tank milk of 2 farms (4 conventional (milking twice a day), 4 conventional milking (three times a day), 4 AM; Table 2.) were determined monthly and in two periods of 4 days from every bulk tank. 2
Groups of farms were selected as high (>0.80 mmol/00 g fat during the last year) or low (<0.70 mmol/00 g fat). <Table 2.> The effect of farm management aspects was studied at 8 farms using the same brand of AM-system and the same cooling system and 6 farms milking three times a day (Table 3.). Half of the farms were classified as high (FFA >0.75 mmol/ 00 g fat during at least the last year) and the other half as low (<0.70 mmol/00 g fat). <Table 3.> Bulk tank milk was sampled and analysed for FFA. Detailed farm (including milk frequencies), robotic and management information regarding feeding, housing conditions and animal health was obtained by a questionnaire during a visit at the farm and correlations among these factors with FFA levels were calculated. Results and discussion Milking frequency Analysis of variance of log transformed FFA showed slight differences between intervals after 0 hour storage and significant differences after 24 hours of storage. (Table 4.) <Table 4.> The increase after 24 hours storage, defined as spontaneous lipolysis (Cartier and Chilliard, 990) was significant for the different milking intervals. Initial levels of FFA were rather low compared to other studies. As all bulk milk samples contain milk of 3 days old and the increase in FFA is the highest in the first 24 hours (data not published), no hydrogen peroxide was used anymore during the rest of the experiments and no sub samples were taken anymore. Technical parameters of the AM-system The milk used in test with brand was not susceptible enough, because initial FFA level was low (mean FFA: 0.36 mmol/00 g fat). Selection of susceptible milk was not properly. The milk used for test 2 with brand 2 was susceptible enough to show difference in FFA increase (mean FFA: 0.52 mmol/00 g fat), but different initial FFA levels in milk passing the conventional and in milk passing the reconstructed AM system made results difficult to interpret. Tests will be repeated. In AM systems more air is needed to transport the same amount of milk in comparison with conventional milking, so FFA increase may be expected. Farm management aspects Preliminary results from the variation study indicate that previous selection, based on the FFA levels of the previous year, of farms in high (>0,80 mmol/00 g fat) and low (<0,70 mmol/00 g fat) FFA level groups was in accordance with the selection for farms milking conventionally and robotically (Figure.). Twice a year determination of FFA was a good indication for these farms. All farms milking three times a day had high levels of FFA on a regular base (>0,80 mmol/00 g fat) and FFA levels were not in accordance to previous results. But the previous results were based on two FFA determinations: one in spring and one in autumn. So more variation in FFA during the year was found for these type of farms. Two of the high 2x farms and one of the high 3x farms were tied stalls with long high milk lines, indicating that FFA levels are more elevated with elevation of milk and long milk lines. 3
On one high and one low 3x farm, FFA increased when cows were pastured. Ferlay et al. (2002) found a higher lipolysis for milk of pasturing cows than for milk from cows feeding a diet rich in concentrate or a corn silage-based diet. The cows grazing on pasture were slightly underfed. Increased lipolysis due to underfeeding in cow milk were observed previously (Jellema, 975). So for two farms there might be a feeding problem, especially in summer. On one high AM farm, the software was changed, resulting in a decrease in the number of failed milkings. As a result FFA levels decreased. This may have been a technical problem related with excessive air inlet. For the other high AM farm no clear explanation could be found, but feeding experiments showed a decrease in FFA levels. <Figure.> Selection of farms, based on FFA levels in the previous year, in the study on farm management aspects seemed to be adequate for 3x milking per day but not for AMsystems (Figure 2.). Farms milking 3x with high FFA levels had a lower fat content in the milk than farms with low FFA levels. This has to be studied in more detail. For AM farms the selection was made on the same brand and cooling systems, all quite recently installed, but FFA results for at least one year were known. After solving some starting problems most of the FFA levels decreased and no significant difference was found between high and low FFA on AM farms. On the farm with the highest FFA levels, the milking frequency was not adequately adjusted. Cows producing less (7 kg per milking) were milked too often. <Figure 2.> Conclusions Increased FFA levels were due to increased milking frequencies (both for AM and conventional: 3x milking per day) and probably to milking machine components of AM systems, but tests must be repeated. Milking frequencies seemed to be of more importance than technical parameters of the AM system, because FFA levels for farms milking three times per day and AM systems were comparable. However, technical aspects cannot be excluded. Compared to conventional milking, the air/milk ratio is higher, probably resulting in more disruption of milk fat globule membranes. Apart from milking frequencies and technical parameters, management aspects probably play a role. Feeding might be a factor of importance and could be studied in more detail. Also more fundamental research is needed regarding the susceptibility of cows. References Cartier, P., Chilliard, Y. 989. Lipase redistribution in cows milk during induced lipolysis. I. Activation by agitation, temperature change, blood serum and heparin. Journal of Dairy Research 56:699-709. Cartier, P., Chilliard, Y., Paquet, D., 990. Inhibiting and activating effcts of skim milks and proteose-peptone fractions on spontaneous lipolysis and purefied lipoprotein lipase activity in bovine milk. Journal of Dairy Science 73:73-77. Chazal, M.P., Chilliard, Y., Coulon, J.B., 987. Effect of nature of forage on spontaneous lipolysis in milk from cows in late lactation. Journal of Dairy Research 54: 3-8. Ferlay, A., Martin, B., Pradel, Ph., Chilliard, Y., 2002. Effect of the nature of forages on lipolytic system in cow milk. In: Proceedings Congrilait 2002, 26 th IDF World Dairy Congress, 24-27 september 2002, Paris, France. poster B4-38. 4
IDF, 99. Determination of free fatty acids in milk and milk products. IDF Bulletin No. 265. Brussels, International Dairy Federation. Ipema, A.H., Schuiling,E., 992. Free fatty acids; influence of milking frequency. In: Ipema, A.H., A.C. Lippus, J.H.M. Metz and W. Rossing (editors) Proceedings of the International Symposium on Prospects for Automatic Milking, Wageningen, EAAP series 65: 49-496. Jellema, A., 975. Susceptibility of bovine milk to lipolysis. Netherlands Milk and Dairy Journal 29:45-52. Jellema, A., 986. Some factors affecting the susceptibility of raw cow milk to lipolysis. Milchwissenschaft 4:553-558. Klei,L.R., Lynch, J.M., Barbano, D.M., Oltenacu, P.A., Lednor, J., Bandler, D.K., 997. Influence of milking three times a day on milk quality. Journal of Dairy Science 80: 427-436. Klungel, G.H., Slaghuis, B.A., Hogeveen,H., 2000. The effect of the introduction of automatic milking systems on milk quality. Journal of Dairy Science 83:998-2003. Svennersten-Sjauna, K., Persson, S., Wiktorsson, H., 2002. The effect of milking interval on milk yield, milk composition an raw milk quality. In: Proceedings of the first North American Conference on Robotic Milking, Wageningen Pers: V43 V48. Vorst, Y., Koning, K. de, 2002. Automatic milking systems and milk quality in three European countries. In: Proceedings of the first North American Conference on Robotic Milking, Wageningen Pers: V-V2. Acknowledgements The work was funded by the European Commission within the EU project "Implications of the Introduction of Automatic Milking on Dairy Farms (QLK5-2000-3006)" as part of the EU-programme "Quality of Life and Management of Living Resources". The content of this publication is the sole responsibility of its publisher, and does not necessarily represent the views of the European Commission nor any of the other partners of the project. Neither the European Commission nor any person acting on behalf of the Commission is responsible for the use, which might be made of the information presented above. The authors thank the farmers for their cooperation during this investigation. Table I. Latin square design of spontaneous lipolysis study, with groups of cows defined as A, B and C. Period 2 3 Milking interval 2x per day A B C 3x per day B C A 6x per day C A B Table II. Number of farms in design of study on variation of FFA levels FFA level high low Type of farm Conventional 2x per day 2 2 Conventional 3x per day 2 2 AM 2 2 Table III. Number of farms in design of study on farm management. FFA level high low remarks Type of farm Conventional 3x per day 3 3 different brands, same cooling system 5
AM 4 4 same brand, same cooling system Table IV. Effect of milking interval on FFA contents (mmol/00 g fat) and standard deviations (between brackets) in raw milk after 0 hours and 24 hours storage after sampling. Interval (times/day) 6 3 2 0 hours at 5 C 0.20 a (0.04) 0.9 ab (0.07) 0.5 b (0.07) 24 hours at 5 C.23 a (0.98) 0.7 b (0.47) 0.42 c (0.3) increase in 24hours 0.97 a (0.98) 0.49 b (0.45) 0.25 c (0.26) a,b,c statistically significant difference on the same row P<0.05) 2 FFA (mmol/00 g fat),5 0,5 0 high high low low high high low low high high low low 2x 2x 2x 2x 3x 3x 3x 3x AM AM AM AM farm type and selection Figure I. Mean FFA and standard deviations on twelve farms based on monthly sampling and two times two weeks sampling of every farm milk tank (5-6 samples in two weeks). Results were based on 9 2 samples per farm.,4 FFA (mmol/00 g fat),2 0,8 0,6 0,4 0,2 0 3x 3x 3x 3x 3x 3x AM AM AM AM AM AM AM AM h h h l l l h h h h l l l l farm type and selection 2 Figure II. FFA in farm tank milk from 6 farms milking 3x per day and from 8 farms with an AM-system. Results of two samplings ( and 2) per farm within six weeks in autumn. Selection was based on high (h) and low (l) FFA-levels in at least the last year. 6