The microbiology of butter reflects the microflora present in pasteurized cream from which it is made, water added at the time of salting butter, sanitary conditions of process equipment, manufacturing environment, and conditions under which the product is stored. Intrinsic properties of butter, for example, aw, pH, salt content, uniformity of moisture distribution and droplet size, all impact microbiological stability.
The main source of microorganisms in butter made under excellent sanitary conditions is cream. Raw milk may be contaminated with a wide variety of pathogenic and spoilage microorganisms. The microflora of raw milk is related to that found in and on the cow's udder, milk-handling equipment, and storage conditions (Jay, 2000). Proper handling, pasteurization, and storage conditions should result in a predominantly gram-positive microflora in milk. Psychrotrophic Bacillus spp. (United States and Europe) and Clostridium spp. (Europe) have been found in 25-35% and 8% of raw milk samples, respectively (Jay, 2000; IDF/FIL,
The Code of Federal Regulations (21 CFR 58.334) stipulates that pasteurization of cream for butter manufacture will be at or above 85°C for 15 s. This thermal treatment minimizes reactivation of lipase native to milk. Further, after 2 years of frozen storage at - 30°C, resultant butter will still have a score of 92 or grade A. Moreover, there are further benefits to this process. Many microorganisms are inactivated. However, there is a lack of research data to show destruction of enzymes from psychrotrophic bacteria during this thermal exposure. Because finished butter is stable during frozen storage, it is thought that all enzymes were destroyed. Pasteurization of cream from raw milk is designed to eliminate vegetative microbial pathogens and reduce numbers of potential spoilage organisms. In the United States, cream must contain not less than 18% fat. However, heat-resistant microbes and spores of Bacillus and Clostridium will survive. Temperatures between 95 and 112°C are commonly used to inactivate them (Schweizer, 1986). Cream is also heated to inactivate lipases (which cause hydrolytic rancidity in butter), reduce intensity of undesirable flavors by vacuum treatment (e.g., from feed ingredients), activate sulfydryl compounds (which can reduce autooxidation of butter), and liquefy milkfat for subsequent efficient churning (Schweizer, 1986).
Many people in western and northern Europe and a few in the United States prefer the flavor of butter manufactured from microbiologically ripened cream (Pesonen, 1986). Traditionally, pasteurized cream is adjusted to 21°C and inoculated with lactic cultures composed of pure or mixed strains of Lactococcus lactis subsp. lactis, Lc. lactis subsp. cremoris, Leuconostoc mesenteroides subsp. crem-oris, and Lc. lactis subsp. lactis biovar diacetylactis. Ripening occurs for 4-6 h until a pH of about 5 is achieved, and then cream is cooled to stop the fermentation. In this process, spoilage microorganisms are controlled primarily through the bacteriostatic effect of lactic acid produced by the starter culture.
The NIZO method (Kimenai, 1986) for producing a cultured butter is allowed in several countries and is used by many dairies in western Europe. In the NIZO method, starter culture is not added to cream, but instead, a mixture of diacetyl-
rich permeate and starter cultures is worked into butter. Fermentation of partly delactosed whey or other suitable media containing milk components by lactic acid bacteria (i.e., Lactobacillus helveticus) continues for 2 days at 37°C, and then the medium is ultrafiltered to remove proteins and bacteria and to further concentrate the medium (Kimenai, 1986). During ultrafiltration, macromole-cules are removed and concentrated in the retentate, whereas low molecular weight solutes pass through into the permeate stream. The pH of butter made with the permeate from this process is more easily adjusted in the desired range of 4.8-5.3. This permeate can be stored at 4°C for more than 4 months under proper conditions. Advantages cited for this process are numerous (Ki-menai, 1986).
Homofermentative lactic acid bacteria such as Lc. lactis subsp. lactis and Lc. lactis subsp. cremoris are used to produce lactic acid from lactose in dairy products. However, flavor production requires addition of a heterofermentative organism such as L. mesenteroides subsp. cremoris or Lc. lactis subsp. lactis biovar diacetylactis to produce diacetyl (Jay, 2000). Diacetyl, in addition to imparting flavor, inhibits gram-negative bacteria and fungi (Jay, 2000).
The bacterial load of buttermilk is typically greater than that of cream or butter (Milner, 1995). When culture-ripened cream is used to manufacture butter, most starter culture organisms are retained in buttermilk; however, some remain in butter. In several studies, butter made from cultured cream retained 0.5-2.0% of the culture organisms present in cream (Hammer and Babel, 1957). Olsen et al. (1988) found numbers of Listeria monocytogenes were 6.7-15.0 times higher in pasteurized but subsequently inoculated creams than in butter manufactured from the same cream. In an earlier study (Minor and Marth, 1972), Staphylococcus aureus behaved similarly. These organisms are gram positive, and it is unclear how other microorganisms with different cell wall and membrane structures distribute themselves between cream and butter. Diacetyl content of milkfat increases during churning; agitation during churning favors oxidative processes needed for diacetyl production (Foster et al., 1957). The pH of salted butter can prohibit formation of diacetyl (Foster et al., 1957).
F. Moisture Distribution During Churning and Working
From 10 to 18 billion droplets of water are dispersed in 1 g of the water-in-oil emulsion that is butter (Hammer and Babel, 1957). Given the low microbial load expected in pasteurized sweet cream (less than 20,000 cfu/mL) (Jay, 2000), most of the droplets are sterile. This depends on size and degree of dispersion of drop lets and the microbial level in cream (Hammer and Babel, 1957). The diameter of water droplets in conventionally made butter has been reported at <1 to >30 |m (Brunner, 1976).
The number of water droplets greater than 30 |m in diameter is inversely proportional to the time of working during conventional (batch churn) butter manufacture (Hammer and Babel, 1957). A consequence of uneven distribution of droplets containing microorganisms is a high degree of nonhomogeneity regarding microbial distribution in butter. Inadequate working of the butter in batch churns results in poor dispersion of water droplets and promotes microbial spoilage (Hammer and Babel, 1957; Foster et al., 1957). Further, this defect can be observed on a trier in the form of moisture droplets. The defect is called ''leaky'' butter and results in a reduced score. This implies that availability of nutrients or inhibitor is limited by the fine dispersion of water droplets (Foster et al., 1957). Droplet size ideally is less than 10 |m (Varnum and Sutherland, 1994).
Butter granules may be washed to remove excess buttermilk (Foster et al., 1957); however, this is not often done today. Salt added to butter inhibits microbial growth. However, salt must be distributed evenly in the moisture phase of butter effectively to inhibit microbial growth in water droplets. Insufficient working results in a nonhomogeneous distribution of salt in the water droplets (Hammer and Babel, 1957; Milner, 1995). Salt creates an osmotic gradient between salt granules and buttermilk during working. This tends to cause aggregation of water droplets and can lead to free moisture (''leaky'' butter) and a color defect called ''mottling.'' Adequate working and use of finely ground salt or salt flour can minimize this defect (Varnam and Sutherland, 1994).
The use of brine to salt butter is restricted to products with less than 1% salt, because the brine cannot contain more than 26% salt (w/w). Mostly, slurries of salt in saturated brine solutions containing up to 70% w/w sodium chloride are used. Salt granules used to produce a slurry should be less than 50 |m in diameter. Salt in the slurry should also be of high chemical purity, with insignificant levels of lead (<1 ppm), iron (<10 ppm), and copper (<2 ppm) (Varnam and Sutherland, 1994).
The microbiological quality of water used for washing or for brines is critical to production of a safe and stable product. Water with less than 100 cfu/mL total aerobic count when plates are incubated at 22°C and less than 10 cfu/mL total aerobic count when plates are incubated at 37°C has been deemed to be acceptable (Murphy, 1990). Formerly, wash water was chilled and chlorinated at 10 ppm 2 h before use to control microflora. Little if any butter washing is done today.
Listeria survive in a saturated brine solution held at 4°C for 132 days (Mitscherlich and Marth, 1984). Thus, brines used to salt butter must be free of Listeria. Water is frequently contaminated with pseudomonads, and consequently care must be taken to insure water and brines used are free of these bacteria. The most common form of spoilage in butter occurs with species of Pseudomonas (Jay 2000; Milner, 1995). Addition of salt to butter lowers the freezing point so that psychrotrophic microorganisms present may be able to grow at less than 0°C. Some psychrotrophic organisms multiply in salted butter stored as low as -6°C (Hammer and Babel, 1957).
Distribution of salt in the moisture phase of butter has less impact on growth of yeasts and molds on the surface of butter as compared to bacteria (Hammer and Babel, 1957). Humid conditions appear to have a greater impact on mold growth than does the material on which they grow. Bacterial spoilage may occur in areas of butter with low salt in large droplets of moisture (poor working).
Varnam and Sutherland (1994), Kimenai (1986), and Munro (1986) have provided more detailed descriptions of continuous butter manufacturing processes.
In batch operations, butter is loaded directly from the churn into hoppers and wheeled to packaging machines. Handling butter this way exposes it to air, workers, plant environment, and ambient temperatures that may accelerate spoilage. Control of the microbiological quality of air in the packaging room is therefore important. HEPA (High Efficiency Particulate Arrester) quality air with the filtration after temperature modification is desired. Practices that result in standing water on the floor or residual and spilled product facilitate growth of environmental contaminants. Practices that aerosolize contaminants often produce unacceptable levels of microbiological contamination in the air. Thus, maintaining dry conditions in the plant is preferred. Numerous approaches can be taken to monitor microbiological air quality, which include sedimentation, impaction on solid surfaces, impingement in liquids, centrifugation, and filtration (Hickey et al., 1992). Air quality is particularly important in butter produced from continuous-type churns that may incorporate up to 5% air into the product (if a vacuum deaerator is not used) (Varnam and Sutherland, 1994). Most whipped butter does not have processing room air incorporated but instead uses purified compressed nitrogen gas. Gases used must be of acceptable microbiological quality.
Personnel hygiene is critical at this point of butter manufacture, because contaminants from hands, mouth, nasal passages, and clothing may be transmitted to butter during packaging. Few continuous churns are arranged to discharge product directly into the receiving hopper of packaging machinery (Varnam and Sutherland, 1994). However, to ensure uninterrupted operation, it is common to transfer butter to a butter boat (open) or covered silo. Covered silos minimize the risk of further contamination from the plant environment. Screw augers in the bottom of the boat or silo move butter to the suction side of a rotary positive displacement pump which moves butter from the boat or silo to packaging equipment. Direct packaging into consumer-size containers is preferable over bulk packaging, because such butter must be reworked and repackaged before sale. Such reworking increases the risk of contamination and subsequent spoilage of butter (Milner, 1995).
Cardboard boxes lined with vegetable parchment, parchment aluminum foil laminate, or a variety of plastic films are typically used for bulk packaging of butter (Varnam and Sutherland, 1994). Polyethylene is the preferred material based on its physical properties (low density, high impact, cost effectiveness, absence of copper, and near sterile condition). Parchment, which supports mold growth under humid conditions, is still frequently used (Varnam and Sutherland, 1994). Retail butter packs are typically wrapped in parchment, waxed parchment, or foil/parchment laminate and overwrapped with a cardboard container. Odors in storage refrigerators will permeate and ultraviolet rays from light will penetrate parchment wraps more rapidly than other wrappers and result in oxidized flavor. Individual butter packs, for example, continentals, cups, and chips, used in restaurants and food service are made at the time of packaging by appropriate highspeed equipment.
Research conducted using the following pathogenic microorganisms has shown their growth in butter products: L. monocytogenes in butter at 4 and 13°C (made from inoculated cream) (Olsen et al., 1988), S. aureus in lightly salted (1% w/ w) whey cream butter at 25 and 30°C (Halpin-Dohnalek and Marth, 1989b), and inoculated whipped butter at 25°C (Halpin-Dohnalek and Marth, 1989a). L. innocua (not a pathogen but frequently associated with L. monocytogenes in environmental samples) was found in butter by Massa et al. (1990).
The incidence of documented food poisoning associated with butter is low. This is partially attributed to widespread use of pasteurization at elevated temperatures. Postpasteurization environmental contamination of cream or butter represents the greatest risk to butter contamination and spoilage. Several outbreaks of staphylo-coccal intoxication related to butter have been reported in the United States (Centers for Disease Control, 1970, 1974, 1977). In one instance, gastrointestinal illness developed in 24 customers and employees of a department store restaurant and was traced to whipped butter manufactured from whey cream (Centers for
Disease Control, 1970). The same butter used to manufacture the implicated whipped product also resulted in one case of gastroenteritis. This butter contained 10 ng of staphylococcal enterotoxin A/g. In 1977, more than 100 customers of pancake houses in the Midwest became ill after consumption of whipped butter (Centers for Disease Control, 1977).
The two principal types of microbial spoilage of butter are surface taint and hy-drolytic rancidity (Jay, 2000). Both conditions can be caused by growth of Pseudomonas spp. Some Pseudomonas spp. are psychrotrophic (Kornacki and Gabis, 1990) and produce proteases and lipases which may survive pasteurization (Cousin, 1982) and which hydrolyse protein and fat, respectively. P. putrifaciens can grow on butter surfaces at 4 to 7°C and produce a putrid odor within 7-10 days (Jay, 2000). This odor may result from liberation of certain organic acids, especially isovaleric acid (Jay, 2000).
Rancidity, the second most common spoilage defect, is caused by both microbial and nonmicrobial lipases, which degrade milkfat to free fatty acids. P. fragi and sometimes P. fluorescens are associated with this defect (Jay, 2000). Mold growth on butter also can cause hydrolytic rancidity for the same reasons (Irbe, 1993). Molds that can cause this defect in butter include some in the genera Rhizopus, Geotrichum, Penicillium, and Cladosporium (Irbe, 1993). Less common spoilage defects include malty flavor, skunk-like odor, and black discoloration. These defects are caused by Lc. lactis var. maltigenes, P. mephitica, and P. nigrifaciens, respectively. Other microbially induced color changes may result from surface growth of various fungi that produce colored spores (Jay, 2000). Heat-resistant proteases and lipases produced by pseudomonads that may grow during storage of raw milk or cream may result in spoilage of butter after manufacture even though spoilage organisms may have been destroyed by pasteurization.
The necessity for milk, cream, and wash water to be of high microbial quality and the importance of pasteurization to public health have been described. Yeasts and molds are particularly resistant to dry conditions when compared to bacteria. Unlike bacteria, many of these fungi can grow at water activities (aw) below 0.84. A few can grow below an aw of 0.65 (Troller and Christian, 1978). A study was reported in which molds would not grow on butter held at or below 70% humidity (Hammer and Babel, 1957). Therefore, to prevent growth of osmotolerant yeasts and molds, a humidity of 60% or less should be maintained in the processing environment.
Ineffective sanitation of processing equipment could result in product contamination from equipment such as piping, pumps, silos, or other equipment (Hammer and Babel, 1957). In our experience, the backplate of older positive displacement pumps (e.g., from pasteurized cream storage tanks) may be neglected during sanitation and become a microbial growth niche, which in turn provides an inoculum to the product stream. Stress cracks in double-walled, insulated tanks can also provide a source of product contamination when the insulating material between walls becomes wet. Further, published data validating effective cleaning and sanitation on continuous churns through use of microbiological swabs are lacking.
Personal hygiene of employees working with butter is also important. Cross contamination from hands, mouths, nasal passages, and clothing must be precluded (Hammer and Babel, 1957). Handling butter in restaurants may also result in cross contamination of a product; for example, when 1-lb prints are divided with knives used for cutting meat or when whipped butter is scooped with improperly sanitized equipment (Halpin-Dohnalek and Marth, 1989a).
Was this article helpful?