How do milking conditions affect the hygienic quality of raw milk?
Case study from Moroccan dairy farms

M T Sraïri, J Moudnib, L Rahho^* and A Hamama^*

Animal Production Department, Hassan II Agronomy and Veterinary Medicine Institute
PO Box 6202, Rabat, Madinate Al Irfane, 10101, Rabat, Morocco
mt.srairi@iav.ac.ma
Department of Hygiene and Industry of food Products from Animal Origin, Institut Agronomique et Vétérinaire Hassan II
PO Box 6202, Rabat, Madinate Al Irfane, 10101, Rabat, Morocco

Abstract

This study was carried out to evaluate the hygienic quality of raw milk and its relationship to milking conditions in Morocco. A total of 109 bulk milk samples obtained directly after milking from 109 different farms around the city of Rabat-Salé, were examined. Concurrently, the rearing and milking conditions in these farms were investigated. To evaluate the hygienic quality of raw milk, the following microbiological counts were determined: Aerobic Plate count, total and faecal coliforms counts, and Staphylococci count.

The overall results obtained suggest that the raw milk tested is of poor hygienic quality with presence of a great variability among milk samples regarding the hygienic quality. The analysis of rearing and milking practices with multivariate statistical tools allows differentiating between the following three types of milking profiles. Profile 1: one milking per day, calf suckling and irregular cleaning of hands and udder; profile 2: one to two milking per day, calf suckling and regular hands and udder cleaning and profile 3: two milking per day, no calf suckling and regular cleaning of hands and udder. The statistical analyses show that these determined milking profiles have significant links with the faecal coliforms and staphylococci counts in milk, but none with the other counts used to characterize the milk hygienic quality. Thus, it could be concluded that the calf suckling practice may play a role in reducing the fæcal coliforms and staphylococci counts in milk by the elimination of the foremilk (the first milk fraction), which is known to be the most contaminated by bacteria, whereas cleaning practices, under the current hygienic conditions in the farms, seem to be ineffective to generate good hygienic environment. These overall results imply the promotion of measures of elementary hygiene within the farms in Morocco in order to improve the hygienic quality of raw milk in the country.

Key-words: Aerobic plate count, coliforms, hygienic quality, milking conditions, milk hygiene, Morocco, raw milk, staphylococci

Comment les conditions de traite affectent-elles la qualité hygiénique du lait cru ?
Étude de cas à partir de fermes laitières au Maroc

Résumé

Cette étude a consisté en une évaluation de la qualité hygiénique du lait cru de vache et de ses relations aux conditions de traite au Maroc. Au total, 109 étables laitières ont été concernées par l'étude. Dans chacune, un échantillon de lait de mélange (toutes les vaches traites) a été prélevé. Simultanément, les pratiques d'élevage et de traite dans ces exploitations ont été caractérisées. Les résultats ont montré que la qualité hygiénique du lait était généralement mauvaise. Elle se caractérisait aussi par une importante variabilité au niveau de tous les critères utilisés au cours de l'étude (Flore mésophile aérobie totale, coliformes totaux et fécaux, et staphylocoques).

L'analyse des pratiques d'élevage et de traite par des méthodes statistiques multidimensionnelles a permis de distinguer les trois profils de traite suivants: profil 1 : une seule traite quotidienne en présence du veau et avec un lavage des mains et de la mamelle aléatoire ; profil 2 : une à deux traites par jour avec la présence du veau et avec un lavage systématique des mains et de la mamelle, et profil 3 : deux traites quotidiennes sans le veau avec un lavage systématique des mains et de la mamelle. Les analyses statistiques ont montré que ces profils de traite ne généraient des différences significatives qu'au niveau de la contamination du lait en coliformes fécaux et en staphylocoques, mais sans incidence sur les comptages en germes totaux ou en coliformes totaux. Il en a été conclu que les pratiques de traite avec la tétée du veau préalable réduisaient les comptages en coliformes fécaux et en staphylocoques par l'élimination des premiers jets, généralement les plus contaminés, tandis que les pratiques de lavage, dans les conditions actuelles d'hygiène dans les fermes, semblent être inefficaces pour générer un environnement hygiénique adéquat. Dans leur globalité, ces résultats suggèrent qu'une vulgarisation poussée de mesures hygiéniques élémentaires doit d'urgence être instaurée pour améliorer la qualité du lait cru au Maroc.

Mots-clés: coliformes, conditions de traite, flore mésophile aérobie totale, hygiène du lait, lait cru, Maroc, qualité hygiénique, staphylocoques

Introduction

In Morocco, dairy production with high merit cows (Holstein and crosses with local strains) has been intensively encouraged since the seventies, because of an important demand for milk, due to the developing demographic growth and a noticeable change in nutritional habits (Hajji et al 1992). A "Dairy Plan" has been consequently launched in the beginning of this period. Thirty years later, an increase in the annual milk production has been observed, even at the scale of smallholders, due to the numerous government incentives provided to stockmen: Holstein dairy heifers imports, artificial insemination, concentrates use, forages irrigation … (MADR 2003). However, dairy production systems in Morocco, remain dominated by extensive farms (Sraïri 2004), due to drought, scarce fodder availability, herdsmen cultural background (Sraïri and Faye 2004) and weak extension of good farming practices. On the other hand, the milk quality is still evaluated in Morocco on the basis of physical and chemical tests such as milk density and fat content only while studies on milk quality involving hygienic and/or microbiological criteria at the farm level are relatively very few (Hamama and El Mouktafi 1990; Amhouri et al 1998). This lack of updated data on milk hygienic quality in Morocco may be harmful to the whole sector in comparison with other countries of the south of the Mediterranean sea (Tunisia for example) where this element represents a key figure factor for a sustainable dairy channel (Djemali and Kayouli 2003). In fact, these considerations on raw milk hygienic quality and the questions they raise for the farmers and for the safety of the consumers are nowadays frequently investigated in other emerging dairy nations such as Iran or Malaysia (Ehsani et al 2004; Chye et al 2004). Recent preliminary studies conducted on that topic in Morocco show that milk hygienic quality is highly variable from farm to farm (Sraïri et al 2005). It can be assumed that this variability is linked to milking and rearing conditions in farms as shown in other contexts (Michel et al 2001; Tourette et al 2002).

The present study aims to determine raw milk hygienic quality at farm level in Morocco, through the assessment of counts of total microbial flora (Aerobic Plate Count), faecal contamination flora (total and faecal coliforms) and Staphylococci. The other objective of this study is to evaluate the relationship of milk hygienic quality with milking conditions (type of milking, role of the suckling calves, role of hands cleaning and udder washing ...).

Material and methods

Survey of dairy farms studied

A detailed survey of some dairy farms in the region of Rabat-Salé (capital city of Morocco) was conducted during spring 2005. It concerned 109 farms located within 50 km around the city. From each farm, following a brief survey on the farming conditions with the farmer, a single sample of bulk milk representing the whole milk produced was collected.

This survey was performed in order to determine within each farm, the cattle production practices, the hygiene conditions, the feeding strategies, and above all, the milking practices. The following aspects which might affect the hygienic quality of milk were particularly studied:

• average milk yield per cow in the day of milk sampling;
• cattle housing conditions;
• type of milking (manual vs. mechanical);
• hygienic conditions of milking (udder cleaning, use of detergent, teat dipping …);
• presence or not of a suckling calf before milking.

Sampling procedure and laboratory analysis

Milk samples were collected immediately after milking. pH and temperature of milk samples were directly measured at farms using a pH-meter and thermometer (WTW inolab), respectively. For the microbiological analysis of milk, 100 ml samples were aseptically collected in sterilized glass flasks of 250ml capacity using an alcohol flamed scoop. The milk samples were transported at the laboratory in an ice box and then stored at 6-8°C before being analyzed within 24h of sampling.

The aerobic plate count was determined by pour plating appropriate dilutions of the milk samples using Plate Count Agar (Biokar Diagnostics, Beauvais, France) as the medium. Plates were incubated at 32°C for 48h prior to counting colonies as outlined in the International Dairy Federation reference method (IDF 1987).

The total coliform count was determined as described in the IDF reference method (IDF 1974). Growth and appearance on Violet Red Bile Agar (VRBA, Biokar Diagnostics, Beauvais, France) after an incubation period of 24h at 35°C were used for a presumptive count, and growth and gas production in 2% Brilliant Green Broth (Difco, Detroit, MI, USA) were used as the confirmatory test for coliforms. Simultaneously, the faecal coliform count was obtained on VRBA plates incubated for 24h at 44°C, confirming typical colonies from these VRBA plates with growth and gas production in E.C. Broth (Difco, Detroit, MI, USA) over 24h at 44°C.

For the determination of the staphylococci count, appropriate dilutions of the milk samples were surface spread on predried Baird-Parker Agar (Biokar Diagnostics, Beauvais, France) medium and the plates were incubated at 35°C for 48h. Typical black colonies on this medium were counted.

Statistical analyses

Simple descriptive statistical analyses were used to report the variability of the different parameters involved in the evaluation of the milk hygienic quality. In the next step, statistical multivariate tools such as Multiple Factorial Analyses (MFA) and Cluster Analyses (CA) were used to determine the main factors which describe the variability of milk hygienic quality and to establish a typology of milking and rearing conditions. The relationship between those two aspects (milk hygienic quality and milking practices) was established by the use of tables of contingency. Differences between milking profiles and their effects on milk hygienic quality were tested by χ² Pearson test. R statistical package was used (R-Project 2004).

Results

Dairy farms characteristics

Data obtained from the survey showed a wide variability in daily milk productivity per cow with a range of 2 to 21 kg and a mean value of 8.5 ± 4.2 kg of milk. The average number of cows per herd was 3.6 ± 8.9 and varied from 1 to 56 cows. The genetic composition of herds was dominated by Friesian and Holstein breeds and their crosses with local strains, as they represent 89% of total cows.

Because of drought, cows were mainly fed with poor quality roughage (straw, lignified hay) and small amounts of concentrates. Very few farms had green fodder (maize or alfalfa for instance) to feed cows and fewer use ration calculations to determine precise quantities of feed for the herd.

Farms infrastructures were also very variable. Most of farms had a concrete soil in the barn, but only half of them had concrete coverage in housing accommodations.

The milking operation is generally conducted in the housing barn, as only one farm has a milking parlour. Milking is essentially manual (98% of the cases) and it is performed only once a day in nearly 80% of the investigated farms. In fact, a great majority of these prefer to leave the relative small quantities of the afternoon milk to the calves, because it is not worthy to deliver it to milk collection centres.

Milking is carried out, generally, in dark and poor ventilated barns. About 70% of the farmers let the cows suckle the calf before milking them, as they think that these would not be milked easily without this practice. The equipments used to milk the cows are in plastic matter and very few farms (6%) have adopted aluminium vessels. None of the farmers used antiseptic solution for teat dipping.

Hands and recipient cleaning before milking is quite systematic (87%), although it is generally done without any detergent using only water, whereas udder cleaning is relatively less frequent (75%).

Raw milk hygienic quality indicators

Milk samples were collected at an average temperature of 13.4 ± 7.4°C and were found to have an average pH value of 6.84 ± 0.21.

Data presented in Table 1, show that the 109 samples of raw milk collected had an average aerobic plate count (APC) of 42.4 x 10⁶ colony forming unit (cfu)/ml. This figure was very variable as standard deviation was 5-times the average value. The cleanest milk sample had a minimum APC value of 1.3 x 10⁴ cfu/ml and the most contaminated sample had a maximum APC value of 1.5 x 10⁹ cfu/ml.

With regard to total coliforms count, the same variability was observed. The average value in the collected raw milk samples was 4.1 x 10⁵ cfu/ml with a minimum value of less than 30 cfu/ml and a maximum value of 20.8 x 10⁶ cfu/ml. The same trend was noticed for the faecal coliforms with a mean value of 80.3 cfu/ml.

Of the 109 samples of raw milk examined, 76 were exempt from any detectable staphylococci. The average count of staphylococci had a value of 246 cfu/ml.

Based on these results, 3 classes of milk hygienic quality were then identified by the use of cluster analysis. Within these classes, only contaminations of milk by faecal coliforms and staphylococci were statistically significant (Table 2), whereas aerobic plate counts and total coliforms counts were not.

Milk hygienic class 1 which included 64 milk samples, had high counts of faecal coliforms (117cfu/ml in average) and intermediate counts of staphylococci (170 cfu/ml in average). The second class of milk hygienic quality composed of 30 milk samples, showed intermediary counts of faecal coliforms (51 cfu/ml in average) and high contaminations by staphylococci (539 cfu/ml in average). Finally, the third class (15 samples) was characterized by the absence of detectable faecal coliforms and by low counts of staphylococci(123 cfu/ml in average).

Typology of milking profiles

The characterization of conditions of cattle rearing and milking showed that they could be divided into three groups: a) those specifically related to on-farm milk quality management (time between milking and milk delivery, type of milking utensils and use of litter), b) those linked to dairy intensification (milking in presence of the calf, number of milking per day, milk yield per cow per day and type of milking), and c) those regarding milking hygiene (hands washing, udder cleaning, vessel cleaning and milk filtration).

Results showed the dichotomy between farms with an average daily milk yield per cow superior to 10 kg and relatively clean milking practices and farms with more extensive dairy production (less than 5 kg in average as daily milk yield per cow) and poor hygienic practices.

By the use of multivariate tools, a typology of cattle rearing and milking practices has been established (Table 3) and consisted in the following three milking profiles:

Relationship between milking profiles and milk hygienic quality

The relationship between milking profiles and indicators of raw milk hygienic quality was investigated by the use of tables of contingency (Table 4).

The χ² Pearson test was highly significant (P = 0.0014). Results indicated that milking profiles led to highly significant differences in the contamination of milk by faecal coliforms and staphylococci, whereas they had apparently no influence, neither on aerobic plate count, nor on total coliforms count. The high counts of staphylococci (milk samples from the hygienic class 2) were associated (up to 50%) to the milking profile 3, which seem to be the cleanest (hands and udder washing and no presence of the calf). In the opposite, the low counts of staphylococci(milk samples from the hygienic class 3) are equally related (46.7%) to the milking profiles 1 and 2, which are mainly characterized by the presence of the calf during milking. Finally, milk samples in the hygienic class 1 (high faecal coliforms contamination) were almost equally associated to milking profiles number 1 and 2 (45 and 40%).

Discussion

The results obtained during this work on the hygienic quality of raw milk in Morocco indicate that the current situation is critical and needs real improvement. In effect, a great majority of milk samples had very high levels of contamination with total microorganisms and total coliforms, which may reach 100 fold the international standards. Similar findings have been reported in previous studies dealing with milk hygienic quality in the same region (Ounine et al 2004; Sraïri et al 2005). This shows that, in general, milk samples collected are of very bad hygienic quality. These raw milks should be rejected and the dairy industry would refuse to use them with regard to the EU regulation (Journal Officiel des Communautés Européennes 1992).

On another hand, the values found in this study for the milk hygienic quality are quite similar to those reported by Hamama and El Mouktafi (1990) who worked on a similar topic and in the same region of Morocco. Hence, it seems that there has not been any significant improvement in hygiene conditions at dairy farms level, in the last 20 years, despite the efforts made to increase milk productivity and to improve milk quality.

Another significant result of this study is the characterization of the type of relationship between milk hygienic quality and milking practices. Moreover, it seems that this relationship is only valid for specific criteria linked to the environmental hygiene (faecal coliforms) and to the sanitary status of the udder (staphylococci). At the same time, it appears that there is no such steady relationship between milking practices and the aerobic plate count, as this criterion expresses the total of microbial contaminations in milk. This could be explained by the important global microbial contamination of the majority of raw milk samples.

Our results suggest that using a suckler calf before milking the cows might decrease both faecal coliforms and staphylococci contaminations in milk. This is in agreement with previous research findings which emphasize on the role of suckling to eliminate foremilk from the udder, which is generally the most contaminated in bacteria (Nielsen et al 2005). This is particularly true for staphylococci in highly subclinical mastitis infected herds (Woolford et al 1998), which is frequently the case in Moroccan dairy farms (Bouslikhane 1998).

On another hand, our results also show that hands washing and udder cleaning before milking the cows and seem to be ineffective to decrease global milk microbial contamination because these operations are not properly executed. However, the quality of water used in these cleaning operations should be investigated because it was noticed during the survey that less than 20% of the farms had access to disinfected tap water. In fact, the water used comes, in general, from wells and is of doubtful hygienic quality. It was also observed that very few farms use detergents during the cleaning operations of hands and udder before milking the cows. This incomplete cleaning may further hinder the role of elementary hygiene practices as pointed out by Gran et al (2002) in the context of smallholder farms in Zimbabwe. It thus explains the inefficiency of hands and udder washing on improving milk hygienic quality.

Globally, our results are consistent with those found in different dairying contexts around the world. For example, in highly specialised dairy farms in France, Michel et al (2001) insist on the significant incidences of milking conditions on milk contamination by Staphyloccocus aureus. Such differences in milk microbiological quality due to the impact of suckling while milking mammals have also been reported in a camel traditional dairy system in Mauritania (Tourette et al 2002).

Conclusion

• This study confirms the broad variety in the hygienic quality of cows' raw milk collected from dairy farms in Morocco. Globally, it also emphasizes on the poor hygienic quality of milk, as the main indicators used (aerobic plate count, coliforms and staphylococci) are, in average, as much as 100 fold higher than international standards.

• Data clearly indicate that milk hygienic quality, under current conditions in Morocco, is not linked to the number of cows per farm, but it is affected by milking practices, especially calf suckling. This relationship between milk hygienic quality and the presence of a suckler calf is particularly obvious for faecal coliforms and staphylococci, whereas it is not significant for aerobic plate count and total coliforms.

• All together, these results show that urgent measures are needed to ensure safe milk at farm level, by the promotion of good hygiene practices. These would preferably focus on efficient cleaning of vessels, hands, udder and the whole building facilities with appropriate detergents and water. These measures should be applied massively because this criterion remains very poor in the majority of dairy systems.

• Measures to decrease milk presumptive pathogens such as staphylococci should consider the problem of persistence of suckling practice, even in farms with high merit cows (Holstein breed).

• However, to avoid the failure of these promoting measures, a simultaneous step would be to implement a system of milk payment linked to its hygienic quality. To be sustainable, this system should be fair and extended to all milks delivered by farmers.

Acknowledgments

This study was carried out in the frame of project PRFI 02/03 (Projet de Recherche Fédérateur Intégré). The authors would like to thank Mr. Chichi M., Mr. El Ayadi M. and Mrs. Bennani M. for their help in laboratory analyses. The authors are grateful to Mr. Messad S. (CIRAD, Montpellier, France) for his kind assistance during the statistical analyses of the experimental results.

References

Amhouri F, Saïdi B, Hamama A et Zahar M 1998 Qualité microbiologique du lait cru. Cas de la région d'Errachidia. Actes Institut Agronomique et Vétérinaire (Maroc). 18, 31-35.

Bouslikhane M 1998 Tropical theileriosis (Theileria annulata) in Morocco: epidemiological and economic investigation. Unpublished Ph D. thesis. The University of Reading. UK. 209p.

Chye F Y, Aminah A and Mohamed Khan A 2004 Bacteriological quality and safety of raw milk in Malaysia. Food Microbiology 21, 535-541.

Djemali M et Kayouli C 2003 L'élevage laitier en Tunisie. In Djemali M., Guellouz M. Les filières lait en Méditerranée: enjeux pour un future durable. Wageningen Pers. EAAP Publication n° 99, 98-105.

Ehsani M R, Azad E, Dabirian S and Shademani H R 2004 Problems and needs of emerging dairy nations in ensuring safe milk and dairy products. In FAO. A farm to table approach for emerging and developed dairy countries. IDF, FAO/OIE international symposium on dairy safety and hygiene. Cape Town, South Africa. March, 2^nd-5^th, 2004. p. 27-32.

Gran H M, Mutukumira A N, Wetlesen A and Narvhus J A 2002 Smallholder dairy processing in Zimbabwe: hygienic practices during milking and the microbiological quality of the milk at the farm and on delivery. Food Control 13, 41-47.

Hajji N, Jaïdi L et Zouaoui M 1992 Prix et concurrence au Maroc : impact de la libéralisation des prix sur la concurrence (Douze études sectorielles). Editions Najah El Jadida : Casablanca, 425 p.

Hamama A et El Mouktafi M 1990 Étude de la qualité hygiénique du lait cru produit au Maroc. Maghreb Vétérinaire 5, 17-20.

IDF (International Dairy Federation) 1974 Milk bacteria count. International Dairy Federation Standards 73/74. Brussels 1974, p. 7.

IDF (International Dairy Federation) 1987 Milk coliforms count. International Dairy Federation Standards 14/87. Brussels, p. 9.

Journal Officiel des Communautés Européennes 1992 Directive 92/46/CEE du conseil du 16 juin, Bruxelles, L268/17.

MADR (Ministère de l'Agriculture du Développement Rural) 2003 Statistiques de la situation de l'élevage et de l'agriculture au Maroc. Direction de l'Élevage. Rabat, Maroc. 53p.

Michel V, Hauwuy A et Chamba J F 2001 La flore microbienne de laits crus de vache : diversité et influence des conditions de production. Lait 81, 575-592. http://www.edpsciences.org/articles/lait/pdf/2001/05/01-michel.pdf?access=ok

Nielsen N I, Larsen T, Bjerring M and Ingvartsen K L 2005 Quarter health, milking interval and sampling time during milking affect the concentration of milk constituents. Journal of Dairy Science 88, 3186-3200.

Ounine K, Rhoutaisse A et El Haloui N E 2004 Caractérisation bactériologique du lait cru produit dans les étables de la région du Gharb. Al Awamia, 109-110 187-204.

R-Project 2004 Version 2.0.0. Statistical analysis package. Available online www.r-project.org. Downloaded on September 20^th, 2004.

Sraïri M T 2004 Typology of dairy cattle systems in Morocco for the analysis of their performances. Ph D. Thesis, Gembloux Agricultural University, Belgium, 200 p.

Sraïri M T et Faye B 2004 Pratiques d'élevage de bovins laitiers considérées à l'aune du discours technique : quelques exemples à partir du Maroc. Ethnozootechnie 74, 47-58

Sraïri M T, Hasni Alaoui I, Hamama A et Faye B 2005 Relations entre pratiques d'élevage et qualité globale du lait de vache en étables suburbaines au Maroc. Revue de Médecine vétérinaire 156, 155-162.

Tourette I, Messad S et Faye B 2002 Impact des pratiques de traite des éleveurs sur la qualité sanitaire du lait de chamelle en Mauritanie. Revue d'Élevage et de Médecine vétérinaire des Pays tropicaux 33, 229-233.

Woolford M W, Williamson J H and Henderson H V 1998 Changes in electrical conductivity and somatic cell count between milk fractions from quarters subclinically infected with particular mastitis pathogens. Journal of Dairy Research 65, 187-198.