2007 Pilot Project Milk Results Summary

There were 181 samples of raw milk collected from 20 farms: 122 cow, 32 goat and 27 sheep milk samples. Samples were collected aproximately every two weeks from June to December, 2007 and tested for the following:

• Total bacteria (PLC)
• Thermoduric (heat tolerant) bacteria (Past. or LPC
• Psychrotrophic (cold adapted) bacteria (PI)
• Coliform bacteria (Total Coli)
• E. coli bacteria (E. coli)
• Staphylococcus aureus bacteria (Staph)
• Somatic cells (SCC)
• Butterfat (BF)
• Protein
• Lactose
• Other solids

To evaluate the risk of undesirable bacteria growing during cheesemaking, to the extent that they could lower cheese quality and pose a health risk to consumers, two standard limits for bacteria were used.

Counts below the “target limit” indicate that the highest quality milk is being used and successful high quality cheese production is assured from the raw milk side, which assumes that the cheesemaking and aging processes maintain this quality by preventing additional contamination from sources other than the raw milk itself.

Counts above the “critical limit” indicate that the milk may be unsuitable for high quality cheesemaking to the extent that it poses a risk to the quality of the cheese being produced.

Counts between the target limits and critical limits indicate that there is a trend upwards, do not pose a serious quality risk, and require action to find the source of contamination. Counts above critical limits require immediate action to correct a problem and lower the levels of bacteria so as to be in compliance and reduce the risk of producing poor quality and potentially unhealthy cheese.

Some of these critical limits are based on FDA and State standards, e.g., total bacteria and somatic cell counts. Others are taken from The Dairy Practices Council Guidelines and point the farmer towards specific areas to find the source of contamination as follows:

• Past. counts – equipment cleaning and sanitation
• PI – all sources (an indicator of reduced shelf-life in fresh dairy products, e.g. fluid milk, lactic cheeses, sour cream, and buttermilk
• Coliform – animal health and parlor hygiene
• E. coli – fecal coliform from manure
• Staph. – animal health and fat deposits on equipment (water temp. too low)
• SCC – animal health, e.g. mastitis

The pilot project used the following standards of maximum levels for the different groups of bacteria:

The results for the different milks are presented below.

COW MILK

GOAT MILK

SHEEP MILK

DISCUSSION

These results clearly show that the counts of total, coliform, E. coli, and psychrotrophic bacteria were higher than the other bacteria counts in the three types of milk.

Total Bacteria (PLC):

• While cow mik had 30% of samples above the target limit of 10,000/ml only 5% were above the critical limit (Federal limit) of 100,000/ml.
• Goat and sheep milks had about the same percentage of counts higher than the target limit (44% and 37%) but there were more sheep milk (18%) than goat milk samples (10%) above the critical limit (Federal limit).

The Dairy Practices Council’s Guidelines for Raw Milk Quality Tests states, “The PLC gives an overall indication of the total bacteria present in a milk sample. When sanitation is good and cooling is adequate, bacteria counts of fresh producer samples are often less than 5,000/ml. and generally represent the natural flora of the animal’s milk and minor levels of contamination. Under ideal conditions, counts can be less than 1,000/ml. Counts in excess of 10,000/ml indicate that improvements in production and/or milk-handling practices are warranted.”

The Dairy Practices Council’s Guidelines for Troubleshooting On-Farm Bacteria Counts in Raw Milk states, “When raw milk bacteria counts of fresh samples exceed 5,000/ml, there is a reason. Usually this involves poor production and handling practices and, while unusual, improper collection, handling, or testing of samples may result in high counts.”

Coliforms and E. coli:

• Cow milk and sheep milk had the same percentage of samples (50%) with coliform counts above the target limit of 10 CFU/ ml. The goat milk had a lower percentage (28%).
• The three milks were about the same in the percentage (15% for cow , 19% for goat, and 11% for sheep) of coliform counts above the critical limit of 100 CFU/ ml.
• Cow milk had a higher percentage of samples (45%) with E. coli above the target limit of 1 CFU/ ml than did the goat milk (28%) and sheep milk (22%).
• The three milks were about the same in the percentage (12, 9, and 11%) of E. coli counts above the critical limit of 10 CFU/ ml.
• Most of the high coliform and E. coli counts in the cow milk came from two farms. There was a correlation between high coliform counts in the milk and the cheese for the farm that made raw milk cheese but not for the other one that made pasteurized milk cheese.

The Dairy Practices Council’s Guidelines for Raw Milk Quality Tests states, “Coliforms are associated with fecal and/or environmental contamination. Counts in raw milk produced under good production methods should be less than 50-100 /ml, while counts of less than 25 are easily obtainable. This test may be used as an indication of production methods, since excessive levels have been associated with poor pre-milking hygiene, dropped milking units, dirty equipment, and, in some cases, coliform mastitis.”

The Dairy Practices Council’s Guidelines for Troubleshooting On-Farm Bacteria Counts in Raw Milk states, “This test may be used as an indication of unsanitary production methods and/or mastitic infection. With good management practices coliform counts of less than 10/ml are achievable and desirable.”

I have also seen high coliform counts during periods of wet, muddy weather when it is hard to keep animals clean. E. coli counts are associated with fecal contamination so dirty flanks and udders, dropped milking units, and manure in the parlor will be sources of high counts.

Psychrotrophic Bacteria (PI):

• Goat milk had a higher percentage (72%) of psychrotrophic counts above the target limit of 10,000 CFU/ ml than the cow milk (45%) or sheep milk (41%).
• The percentage of psychrotrophic counts above the critical limit of 50,000 CFU/ ml were the lowest in cow milk (28%). Goat and sheep milk psychrotrophic counts were about the same (40 and 36%).

The Dairy Practices Council’s Guidelines for Raw Milk Quality Tests states, “The PI should always be compared to a fresh PLC (total bacteria count) and PI results should be less than 3-4 times the PLC. Desirable results are 25,000/ml or less. There is no legal limit, although values of 25,000 to 50,000/ml are often used as targets when the test is used in premium programs. When the PI is high compared to the fresh PLC, poor production practices, such as inadequate cleaning; omitted sanitization; poor udder washing procedures; or marginal cooling are indicated. The most common psychrotrophic bacteria are capable of growing and increasing in numbers during prolonged refrigerated storage of raw milk, especially if temperatures are marginal. Generally, these bacteria do not survive pasteurization, but they may produce heat stabile enzymes that may degrade dairy product quality.”

It does not seem that the higher percentage of goat milk samples with counts above the target limit was very significant in the context of the above statement. The counts above the critical limit of 50,000/ml may have been from samples of milk that was stored for longer than 40 hours, which is generally the cut off point for “fresh milk.” It does not seem that the PI is an important indicator of milk quality for raw milk cheese unless the count is very high, e.g. >100,000/ml.

The Dairy Practices Council’s Guidelines for Troubleshooting On-Farm Bacteria Counts in Raw Milk states, “Causes of high PI counts include dirty animals, poor udder sanitation practices, slow cooling or temperatures above 40 °F, failure to thoroughly clean equipment after each use, neglecting to sanitize equipment prior to use, a contaminated water supply, improperly drained milking equipment and worn rubber or plastic parts. Industry field personnel may be very discouraged by the first results of PI counts. Initially, counts may exceed 1,000,000/ml and a large percentage may be over 100,000/ml. Adopting a PI Count program requires time, patience and some changes in farm sanitation practices.”

Obviously the PI is an important focus for processors of fresh dairy products. The sources of high PI counts are multiple. While we have not focused on the PI count as being as important as the coliform, thermoduric, Staph. and SCC, it still has relevance to those who make cheese because a high count indicates that a better job can be done in producing and storing the raw milk. The main point of the PI seems to be that it can show that contamination has occurred even if the PLC (total bacteria) is low because “prompt cooling of milk and other procedures may yield low counts, disguising unclean milking equipment or poor production practices.” The PI brings these issues to the forefront.

Thermoduric Bacteria (Past. or LPC)

• Counts were mostly below the target limit of 300 CFU/ml for all milks; there were 10% above the limit for cow and sheep milks and none for goat milk.
• Only the cow milk had samples above the critical limit of 2,000 CFU/ml, which amounted to 2.5%.
• The high counts in cow milk were mostly from one farm but there didn’t seem to be a correlation to the cheese quality made at this farm.

The Dairy Practices Council’s Guidelines for Raw Milk Quality Tests states, “Levels in raw milk following laboratory pasteurization and pasteurized milk should be less than 250-300/ml. The natural flora of the animal’s milk and most mastitis bacteria generally don’t survive pasteurization. Excessive LPC values have been associated with dirty equipment (especially areas that are persistently neglected in cleaning or often left insufficiently cleaned), old rubber parts, and poor pre-milking hygiene procedures.”

The Dairy Practices Council’s Guidelines for Troubleshooting On-Farm Bacteria Countsin Raw Milk states, “Most thermoduric bacteria will not grow at refrigeration temperatures of 40 °F or below. However, thermoduric psychrotrophs do exist and may cause spoilage in pasteurized milk.”

This last statement reminds me of problems I have encountered making cheese in workshop situations using pasteurized milk, which has been stored cold for 1-3 days after pasteurizing. The curd does not form as quickly or as firmly. I have overcome this problem by using more rennet but it is never quite the same. Obviously the surviving bacteria are releasing enzymes that are chopping up the protein thereby making the milk less able to curdle well.

Somatic Cell Count (SCC)

• Goat milk had the highest percentage of samples (62%) higher than the target limit of 400,000/ml, followed by sheep milk (33%), and cow milk (18%).
• Goat milk had the highest percentage of samples (33%) higher than the critical limit of 1,000,000/ml, followed by sheep milk (15%) and cow milk (4%).

It is not surprising to have found such a large number of goat milk samples with high SCC counts since this can occur in healthy goats. SCC can increase naturally during late lactation or due to aging. The Dairy Practices Council’s Guidelines for Troubleshooting On-Farm Bacteria Counts in Raw Milk states, “Because goats have a different physiology than cows, a special stain (Pyronin Y-Methyl Green) and DMSCC procedure is required to be used to determine the number of cells in milk in dairy goats when automated screening tests exceed the 1,000,000 SCC/ml limit. For sheep, the same methods and limits as used for cows currently apply, although the validity of these requirements is in the process of being studied.”

For cows and sheep, “milk is considered ‘abnormal’ when the number of somatic cells is excessive. While the legal limit under the PMO is 750,000 SCC/ml, counts in excess of 200,000 – 300,000/ml are considered to be above the level expected in a healthy herd. High numbers of somatic cells in bulk milk suggest mastitis problems in the herd and may result in products of lower quality and yield.” Researchers have shown a direct linear correlation between the decrease in cheese yield and the increase in SCC after the SCC rises above 250,000/ml/.

Staph. Aureus

• A sample of goat milk was the only sample that exceeded the critical limit of 2,000/ml.
• Only two samples of goat milk, one of cow and one of sheep milk exceeded the target of 500/ml.
• Cheese made from these milks did not contain any Staph. toxin.

Correlations Between Milk Quality and Cheese Quality

• Coliform and E. coli but not affecting cheese quality and not a risk to public health

ANALYSIS

The levels of colform and E. coli were higher than I expected. However, although there did seem to be some correlation between levels of coliform and E. coli in the milk and the cheese in specific cases, the levels of coliform had to be higher than 10,000/ml for there to be survival in the raw milk cheese. For levels in the range of 100 to 10,000/ml, a moderate to fast acidification, e.g. to pH 5.4 in 8 hours, seems to supress the growth of coliform bacteria. Of course the rate of coliform growth is affected by type of cheese being made; more traditional types such as tomme may have slower acidification rates than this and therefore be less able to reduce coliform growth significantly. Any cheese made with a cooking step that exceeds 110 °F will eliminate coliform bacteria. The aging process will also affect coliform survival so, as long as levels do not get above 10,000/ml in the milk, it seems likely that there will be far less in the cheese after aging is completed.

The levels of Staph. and thermoduric bacteria in all milks and SCC in cow milk were very low. The levels of psychrotrophic and total bacteria were higher than I expected. I found it surprising that so much milk being made into farmstead cheeses was not in compliance with state standards.

The milk tests were useful in helping to produce high quality milk, which should translate into high quality cheeses. The correlations between the test results and the quality and safety of the cheeses were not strong. The coliform bacteria were the only group with any significant effect on cheese quality and even this did not seem to have much to do with cheese quality or pose a public health risk.

The only group of bacteria that is not accounted for in this testing program is the spore-forming, thermoduric Clostridia bacteria that are responsible for late gas development in cheese (after two weeks). These are not differentiated in the “thermoduric” count so that one can be sure that the milk poses a risk for this defect to occur. Testing for butyric acid (or the spores themselves) in the fermented feeds will provide an indication that these bacteria may get into the milk and then the cheese. Many farmers regularly test their fermented feeds for butyric acid content because it can be a health risk for animals to consume it.

Achieving these accepted standards for raw milk production (critical limits) help the farmstead cheesemaker remain in compliance with the FDA and state regulations for licensed dairy farms. The important question is whether the target limit, which is more stringent than the critical limit, actually assures the production of higher quality cheese. In the final analysis, it is important for each farmer to understand the profile of their milk as it relates to the types of cheeses they make and set the limits where they provide the most benefit. Having the milk tested at least every two weeks is very helpful in shaping this profile and providing information to maintain compliance. I personally like working with the two limits because I can be working on keeping below the critical limit before it is ever exceeded. However, I make raw milk cheeses that are mostly uncooked or cooked at low temp., e.g. below 100 °F, and have slow acidifications, e.g. reaching pH 5.2 in 12-24 hours. It is very important to regularly test milks being purchased from other farmers and establish a profile. Some of the samples with higher counts in this study came from purchased milks.

Many of the pilot project participants were able to lower high counts within a period of two to four weeks by finding situations pertaining to animals, equipment, environment, hygiene, and sanitation that could be changed or improved. In some cases the process took longer. From the comments to the quick survey it seemed that the regular milk testing was an important benefit. However, the sampling and shipping were not always easy to coordinate. I advise you all to try and stick with at least one milk sample every month from your farm and any farms that you buy milk from (assuming that your state is doing one sample per month as well).

back to top