survey Ital. J. Food Sci., vol. 27 - 2015 1 - Keywords: cow milk, derogations, somatic cell count, total bacterial count - END OF THE DEROGATIONS TO REGULATION (EC) 853/2004 FOR COW’S MILK IN ITALY G. BOLzONI*, A. MARCOLINI and E. BUFFOLI National Reference Center for Bovine Milk Quality IZSLER, Via Bianchi 9, 25100 Brescia, Italy *Corresponding author: Tel. 0039 030 2290541, email: giuseppe.bolzoni@izsler.it AbstrAct Derogations for somatic cell and total bacterial count limits had allowed non-compliant milk to be used for cheesemaking in Italy. commercial and health considerations prompted a decision to implement a program to gradually repeal the derogations. In this study, we report the statistical evaluation of the situation in 2007-2008, the outcomes of the program to close the derogation and observations of its effects during its implementation from 2010-2013 in the Lombardy region. the introduction of a progressive decrease of the limit allowed regulators to minimize the negative im- pact on production levels by focusing on the management of the most non-compliant farms first. 2 Ital. J. Food Sci., vol. 27 - 2015 INtrODUctION After the adoption of European regulations for food safety (reg. Ec 178/2002, 852-853- 854/2004), Italian farmers were still able to sell cow’s milk that was non-compliant in somatic cell count (scc) and total bacterial count (tbc), indicators of presence of udder pathogens and insufficient hygiene during production and stor- age, respectively, because of the derogation of article 10 of the regulation (Ec) 853/2004. specifically the derogation allowed for cow’s milk with geometric means exceeding the legal limits (400,000 cells/ml for scc and 100,000 cFU/ml for tbc) to be used for the production of cheese with ripening periods of at least 60 days. this derogation was based on the knowledge that potential risks linked to high scc and tbc can be significantly reduced or even eliminated dur- ing the production processes and ripening peri- ods (Annex IV of the reg. Ec 854/2004). In fact, many stages of processing have antimicrobial ef- fects including: the cooking of curd, the acidifi- cation of curd, the salting of cheese and the re- duction of free water. However, over the years un- favorable opinions about the use of non-compli- ant milk have increased in Italy. Aged dairy prod- ucts, like Grana Padano and Parmigiano reggiano cheese, represent the uniqueness and tradition of Italian raw milk cheeses. Given that they are some of the most popular cheeses on the interna- tional market, is the use of the “worst” milk ap- propriate for the “best” cheeses of Italy? It became clear that the presence of the dero- gation did not promote improvement in the qual- ity of Italian milk. Furthermore, according to the principles of the community regulations, a der- ogation has to be considered “temporary” and contingent on specific issues. A derogation, if it affects food safety, must always provide a time limit or an exit strategy that will lead to conform- ity with the other nations. For these reasons It- aly began a gradual process to repeal the dero- gation. the project was developed in 2008, af- ter which it was communicated to the Europe- an commission (notification number 134/2010) and it was formalized by the “Agreement between the Government, regions and autonomous Prov- inces of 09 september 2010”. the goals of this study were to conduct a preliminary evaluation of the problem based on data from 2005 to 2008, to create a plan to phase out the derogation and to evaluate the results of its application over 2 years, from January 2011 to June 2013, in the Lombardy region (an area responsible for more than 40% of the national milk production). MAtErIALs AND MEtHODs the data come from the analysis of the milk quality payment system instituted in the Lom- bardy region. the system requires at least 2 sam- ples per month from each farm. the samples are taken from farms by trained and qualified dairy industry operators. For statistical evaluation and calculation of the geometric mean we selected the farms with continuous production and active par- ticipation in the milk quality payment system dur- ing the 2-year study period (roughly 4600 farms out of 6,000 total active farms in the region). the majority of farms can be characterized by rear- ing Holstein Friesian (85%) or brown swiss (15%) cows in loose housing with cubicles, with milk- ing parlor, fed with mixed ration of corn silage, hay and concentrate. On these farms the average herd size is 70 cows and the average milk yield is 9,400 kg per cow per year. tbc was determined with bactoscan Fc and the scc was obtained using Fossomatic 5000 (Foss, DK). For descriptive statistics (frequen- cy distribution of farms’ geometric means of scc and tbc) the free “r” software environ- ment was used. rEsULts AND DIscUssION Analysis of compliance the distribution of the farms’ yearly scc and tbc geometric means during the year 2008 are represented in Figures 1 and 2, respectively (5,200 farms). the “yearly” geometric means were calculated for each farm from 24 or more sam- ples per year. From the figures it is evident that the scc situation was more critical than that of tbc with respect to regulation 853/2004 limits (bErtOccHI et al., 2012), (bOLZONI et al., 2007). In table 1 an evaluation of the farms’ scc roll- ing geometric means from 2005 to 2007 is pre- sented. the “rolling” geometric means were cal- culated over periods of 2 or 3 months for tbc and scc, respectively. the data show that 44% of the farms were consistently under the scc limit, while 29% of farms exceeded it one or more times but returned under the limit within the 3-months observation period; the remain- ing 27% of the farms were still non-compliant after the observation period. Further analysis of the last group of non-compliant farms (Figure 3) revealed that only a small fraction of these farms returned to compliance shortly after the 3 months of observation. Most of the farms spent a long time in non-compliance and even some farms never became compliant. these were the farms that were able to avoid compliance by commercializing their non-compliant milk in a geographic area where most of the milk is used for aged cheese. The end of the derogation: potential effects According to the abovementioned observa- tions, it could be hypothesized that closure of the derogation would cause economic problems Ital. J. Food Sci., vol. 27 - 2015 3 table 1 - somatic cell count compliance among 4,595 farms from 2005 to 2007. Rolling geometric mean (cells/mL) Farms (n) Farms (%) Always < 400,000 2,032 44% > 400,000 with recuperation < 90 days 1,313 29% > 400,000 without recuperation < 90 days 1,250 27% Fig. 1. Fig. 2. Fig. 3. for about 30% of the farms based on their scc values and 4% of the farms based on their tbc values (data not shown). to explore these ef- fects, we performed a statistical simulation on the 2008 data to quantify the effect of a grad- ual repeal of the derogation and its effects on farm production. the results of the simulation for scc are presented in table 2. Program to repeal the derogation A regional program was developed with a se- ries of decreasing temporary limits. Given that controlling scc levels is known to require both medium- and long- term actions, it was ex- pected that some of the farmers would become proactive with control measures and improve- ments in order to reach compliance in time for the more restrictive future limits. the plan was approved by the Ministry of Health and then it was expanded into a national program with the “Agreement between the Government, regions 4 Ital. J. Food Sci., vol. 27 - 2015 table 3 - temporary limits for compliance during the program to close the derogation (geometric means calculated over pe- riods of 2 or 3 months, respectively, for tbc and scc). PERIOD Total Bacterial Count Somatic Cell Count (Rolling GM calculated over 2 months) (Rolling GM calculated over 3 months) January 2011 - June 2011 < 200,000 < 700,000 July 2011 - June 2012 < 100,000 < 600,000 July 2012 - June 2013 No derogation < 500,000 From July 2013 No derogation < 400,000 Abbreviation: GM, geometric mean. and Autonomous Provinces”. It banned the use of non-compliant milk for human consumption and set temporary limits for tbc and scc as shown in table 3. currently milk with rolling ge- ometric means > 400,000 for scc or > 100,000 for tbc continues to be used in the production of cheeses with over 60 days of ripening. table 2 - Geometric means of the compliance of 4,669 farms at different somatic cell count limits in 2008. Somatic cell limit % of farms always % of farms over the limit % of farms over the limit (cells/mL) under limit with recuperation in 90 days without recuperation in 90 days 700,000 89 8.4 2.7 600,000 83 12.1 4.6 500,000 73 19.6 7.7 400,000 50 32.1 18 Fig. 4. Program application and results ten-year trends in scc and tbc levels in the Lombardy region are shown in Figures 4 and 5, respectively. From these graphs it is possible to deduce a preliminary and general trend of de- cline in the two parameters. In particular, in fig- Ital. J. Food Sci., vol. 27 - 2015 5 ure 5, the low percentage of samples over 100,000 UFc/ml (colored bars) and the low and decreas- ing value of the yearly regional mean (blue line), indirectly confirm the very low number of non- compliant farms in tbc during the last year. the number of samples used to generate the yearly Fig. 5. means varies from a high of 165,000 in 2003 to a low of about 100,000 in 2013 which reflects the significant reduction in the number active farms in the region during this period. Fig. 6 presents a specific assessment of the effect of the progressive application of the pro- Fig. 6. 6 Ital. J. Food Sci., vol. 27 - 2015 gram where the blue line indicates the distribu- tion of the farms’ geometric means in the first year (2011) while the blue bars indicate the sit- uation in second year (2012). the difference be- tween the two years is particularly evident in some areas (e.g. 200,000-250,000 scc). It is in- teresting to note that there is a decrease in the percentage of farms presenting with 400,000- 500,000 cells/ml in 2012 even though the limit was still set at 700,000-600,000 cells/ml. this again suggests that some farmers took preven- tive actions early with an eye to the more restric- tive limits of 2013. Figure 7 presents the same comparison for the tbc. In this case, it is not possible to appreciate graphically the differenc- es between the two years because of the mini- mal variations of the data across the board. One further analysis was performed on scc to com- pare the first semester of 2010 (before the start of the program) and the first semester of 2013 and the data are presented in Figure 8; Figure 9 shows the same comparison between the first semesters of 2012 and 2013. It is evident that the percentage of geometric means with a lower scc range increased over time while the high- er scc values decreased. For example, in Fig- ure 8 the percentages of farms in the ranges of 300,000-350,000 and 350,000-400,000 cells/ ml both decreased while the percentage of farms in the ranges of 100,000-150,000 and 150,000- 200,000 cells/ml both increased almost equal- ly. Even though the data come from thousands of different farms, the combination of these rep- resentations enabled us to notice a shift to the right in the distribution and through it the pos- Fig. 7. itive effects of the introduction of this progres- sive program. Its effects will be fully appreciated, of course, only in 2014 when the program will have been finished for over 6 months. However we still hypothesize that the impact on compliant productivity will be minimal due to the progres- sive shift of the majority of the farms to adopt- ing long-term practices that should ensure their continued conformity within scc limits (KEL- LY et al., 2009; NOrMAN et al., 1995; sHUKKEN et al., 2003). the number of farms with tbc over the limit has already been negligible since 2011, likely because corrective actions against tbc can be effective in a very short time (KEL- LY et al., 2009). cONcLUsIONs the decision to repeal the derogations for scc and tbc in raw milk was made for several rea- sons but particularly because of the need to stim- ulate improvements in the quality of milk and traditional dairy products. statistical evaluation of the data from the previous years (2005-2007) suggested that a sudden closure of the deroga- tions, especially for scc, would create difficulties for farmers with downstream repercussions on the dairy industry as well as the Health Authori- ty (regional veterinary service in Italy). the intro- duction of a progressive decrease of the limits al- lowed regulators to minimize these consequences and focus on the most non-compliant producers. At the same time, it allowed farmers to make im- provements and preventive measures during the Ital. J. Food Sci., vol. 27 - 2015 7 24-month program, without excessive conflict or serious effects on farm productivity. the progressive reduction of the scc limit also allowed the regional veterinary service to split the management of the non-compliant farms, giving priority to solving the most serious and Fig. 8. significant problems first (starting from the first and highest limit of 700,000 cells/ml). the is- sue of non-compliance in tbc has been nearly resolved; since 2012 the cases of tbc non-com- pliance appear only occasionally and are often quickly resolved. Fig. 9. 8 Ital. J. Food Sci., vol. 27 - 2015 rEFErENcEs Agreement between the Government, regions and autono- mous Provinces of trento and bolzano on the transitional use of raw bovine milk not meeting the criteria laid down in Annex III, sec. IX, of reg. (Ec) n. 853/2004 as regards plate count and somatic cell count for the manufacture of cheeses with an ageing or ripening period of at least sixty days – GU General series n. 250 of 10/25/2010 the r Project for statistical computing. Available at: http:// www.r-project.org/ bertocchi L., Zanardi G., bolzoni G. Daga s. andVismara F. 2012. relation between dairy cattle welfare and bulk milk somatic cell count. Presented at XXVII World bui- atrics congress, Lisbon, Portugal, June 3-8 bolzoni G., Varisco G., Marcolini A., benicchio s. and Ghilar- di c. 2007. Pagamento del latte in base alla qualità in Lombardia: ci sono le premesse per un nuovo sviluppo. Il Mondo del Latte 61 (5): 37-43 Kelly P.t., O’sullivan K., berry D.P., More s.J., Meaney W.J., O’callaghan E.J. and O’brien b. 2009. Farm management factors associated with bulk tank somatic cell count in Irish dairy herds. Ir Vet J. 62(suppl 4): s45-s51 Paper Received November 19, 2013 Accepted May 30, 2014 Kelly P.t., O’sullivan K., berry D.P., More s.J., Meaney W.J., O’callaghan E.J. and O’brien b. 2009. Farm management factors associated with bulk tank total bacterial count in Irish dairy herds during 2006/2007. Ir Vet J. 62(1):36-42. Norman H.D., Miller H.D., Wright J.r. and Wiggans G.r. 2000. Herd and state means for somatic cell count from dairy herd improvement. JDs. 83 (12): 2782-2788 regulation (Ec) N. 178/2002 - 28 January 2002 - Laying down the general principles and requirements of food law, establishing the European Food safety Authority and lay- ing down procedures in matters of food safety regulation (Ec) N. 852/2004 - 29 April 2004 - on the hy- giene of foodstuffs regulation (Ec) N. 853/2004 - 29 April 2004 - laying down specific hygiene rules for on the hygiene of foodstuffs regulation (Ec) N. 854/2004 - 29 April 2004 - laying down specific rules for the organization of official controls on products of animal origin intended for human con- sumption schukken Y.H., Wilson D.J., Welcome F., Garrison-tikof- sky L. and Gonzalez r.N. 2003. Monitoring udder health and milk quality using somatic cell counts. Vet res. 34:579-596.