Advertisement

Effect of midday pasteurizer washing on thermoduric organisms and their progression through Cheddar cheese manufacturing and ripening

Open AccessPublished:October 13, 2021DOI:https://doi.org/10.3168/jds.2021-20446

      ABSTRACT

      Thermoduric bacteria are known to affect the quality of Cheddar cheese, with manifested defects including slits, weak body, and blowing. Thermoduric bacteria are likely to increase in numbers during cheese-making, as in-process conditions are conducive to proliferation. The present study was conducted to track thermoduric bacterial progression during an 18- to 20-h Cheddar cheese production run and during ripening when the pasteurizer was washed at midway through the production day. This study also correlated a broad range of chemical changes to the growth of thermoduric bacteria during ripening. Three independent cheese trials were performed at 3.5- ± 0.5-mo intervals. Samples were drawn in duplicates at 4 different times of the day: at the start of the run (vat 1), prior to a midday wash of the pasteurizer (vat 20), after the midday wash of the pasteurizer (vat 21), and at the end of the run (vat 42) for raw milk, pasteurized milk, and cheese. Cheeses were also tested during ripening for 6 mo. Results showed that raw milk total bacterial counts comprised 0.24% thermoduric mesophiles (TM) and 0.12% thermoduric thermophiles (TT). The thermoduric thermophilic bacterial counts increased by log10 1.23 during the pasteurizer run of 9 to 10 h, indicating a buildup of thermoduric thermophilic bacteria during the pasteurization process itself. Midday washing reduced thermophilic counts by log10 1.36, as evident by pre- and post-midday wash counts. However, a thermophilic buildup during post-midday wash was again noticed near the end of the 20-h run. We found that TT bacteria decreased in the first 60 d of ripening, whereas TM bacteria increased during the same period. However, TT bacteria increased later during 60 to 180 d of ripening. Bacillus licheniformis was the most frequently isolated bacteria in this study and was recovered at all production stages sampled during the cheese-making and ripening. We observed a significant increase in the level of orotic and uric acids in the vat made at the end of the day. No significant difference in the overall chemical composition, proteolysis, sugar, or other organic acids was observed in cheese made at the start versus the end of the production run.

      Key words

      INTRODUCTION

      The cheese market is growing every year in the United States. Cheese production in the United States reached 0.5 billion kilograms in May 2018 and has increased by 1.4% since May 2017. Cheddar cheese accounts for 29% of total cheese in the United States (
      • USDA
      Dairy product production—July 21. Cheese Market News, the Weekly Newspaper of the Nation's Cheese and Dairy/Deli Business.
      ). Despite maintaining good sanitary conditions in cheese manufacturing plants, thermoduric bacteria are a constant matter of concern for cheese quality (
      • Hull R.R.
      • Toyne S.
      • Haynes I.N.
      • Lehmann F.L.
      Thermoduric bacteria: A re-emerging problem in cheesemaking.
      ) and extended shelf-life, as the export of cheese has increased. The lengthy processing hours and increasing capacity (vat numbers and vat sizes) of cheese plants are 2 prominent factors that may lead to substantial financial losses if any spoilage occurs in the final product (
      • Hull R.R.
      • Toyne S.
      • Haynes I.N.
      • Lehmann F.L.
      Thermoduric bacteria: A re-emerging problem in cheesemaking.
      ;
      • Lauzon N.
      Method for controlling microbiological contamination in a heat exchanger while processing a food product.
      ). Milk processing plants rely on the pasteurization step for assuring product quality and safety. Thermoduric bacteria, however, by definition (
      • Frank J.F.
      • Yousef A.E.
      Tests for groups of microorganisms.
      ), can survive the pasteurization process (72°C/16 s). Additionally, endospore formers have much higher heat resistance and largely survive common heat treatments. Thermoduric thermophile (TT) bacteria, of concern in the dairy industry, are mainly of the genus Bacillus. Obligate or facultative thermophiles with growth temperature ranges of 40 to 75°C and 35 to 70°C, respectively, may proliferate in cheese manufacturing. Although thermophilic bacilli are not pathogenic, their spores can germinate into vegetative cells and may result in off-flavor and other defects (
      • McGuiggan J.T.M.
      • McCleery D.R.
      • Hannan A.
      • Gilmour A.
      Aerobic spore-forming bacteria in bulk raw milk: Factors influencing the numbers of psychrotrophic, mesophilic and thermophilic Bacillus spores.
      ). Several cheese defects related to flavor and texture, early and late blowing of cheese (
      • Dasgupta A.P.
      • Hull R.R.
      Late blowing of swiss cheese: Incidence of Clostridium tyrobutyricum in manufacturing milk.
      ;
      • Klijn N.
      • Nieuwenhof F.F.
      • Hoolwerf J.D.
      • van der Waals C.B.
      • Weerkamp A.H.
      Identification of Clostridium tyrobutyricum as the causative agent of late blowing in cheese by species-specific PCR amplification.
      ;
      • Ortakci F.
      • Broadbent J.R.
      • Oberg C.J.
      • McMahon D.J.
      Late blowing of Cheddar cheese induced by accelerated ripening and ribose and galactose supplementation in presence of a novel obligatory heterofermentative nonstarter Lactobacillus wasatchensis.
      ;
      • Sutariya S.
      • Sunkesula V.
      • Bhanduriya K.
      • Jhanwar A.
      A novel organism Lactobacillus wasatchensis: Growth, detection, gassing defects in cheese, control strategy and future research opportunities: A review.
      ), soft-body (
      • Hull R.R.
      • Roberts A.V.
      • Mayes J.J.
      Association of Lactobacillus casei with soft-body defect in commercial mozzarella cheese.
      ), phenolic flavor, gray spot, and biogenic amines (
      • Hull R.R.
      • Toyne S.
      • Haynes I.N.
      • Lehmann F.L.
      Thermoduric bacteria: A re-emerging problem in cheesemaking.
      ) have been linked with thermoduric bacteria and the proteolysis associated with them. Cheddar cheese manufacturing involves the combination of 7 stages: pasteurization, starter addition, renneting, whey drainage (cheddaring), salting, pressing, and ripening. The temperatures of the cheese vat, drainage belt, salting chamber, and cheese tower generally vary from 32 to 37°C, providing favorable conditions for thermoduric bacteria growth (
      • Arce A.
      • Ustunol Z.
      Effect of microencapsulated ferrous sulfate particle size on Cheddar cheese composition and quality.
      ). Curd usually takes 2.5 h to complete these stages. The temperature profile and length of production runs (∼20 h) provide ample opportunities for bacterial growth and colonization on the inner surfaces of processing equipment (
      • Arce A.
      • Ustunol Z.
      Effect of microencapsulated ferrous sulfate particle size on Cheddar cheese composition and quality.
      ).
      Another source of contaminants is biofilm and its presence on improperly cleaned surfaces. Bacterial biofilm formation in the dairy and food industries has been the focus of reviews (
      • Flint S.H.
      • Bremer P.J.
      • Brooks J.D.
      Biofilms in dairy manufacturing plant—Description, current concerns and methods of control.
      ;
      • Brooks J.D.
      • Flint S.H.
      Biofilms in the food industry: Problems and potential solutions.
      ;
      • Burgess S.A.
      • Lindsay D.
      • Flint S.H.
      Thermophilic bacilli and their importance in dairy processing.
      ). Food, including milk, can form deposits on equipment walls during processing, promoting bacterial attachment. The attached bacteria further grow and colonize by forming a polymeric matrix that houses multispecies populations of bacteria known as biofilm on product contact surfaces. These biofilms can protect bacteria in adverse conditions and make them resistant to cleaning and sanitation. Biofilm can be divided into 2 categories. “Process biofilms” refers to bacterial attachment and aggregation on product contact surfaces in process and biofilms that form in a dairy processing environment (
      • Flint S.H.
      • Bremer P.J.
      • Brooks J.D.
      Biofilms in dairy manufacturing plant—Description, current concerns and methods of control.
      ;
      • González-Rivas F.
      • Ripolles-Avila C.
      • Fontecha-Umaña F.
      • Ríos-Castillo A.G.
      • Rodríguez-Jerez J.J.
      Biofilms in the spotlight: Detection, quantification, and removal methods.
      ). The availability of concentrated nutrients within the interface of biofilms further favors the growth of both pathogenic and spoilage-causing bacteria. The persistence of biofilms is thus a major concern for milk processors. Biofilms can increase the level of thermoduric bacteria and spore formers in the final product. Biofilms that survive cleaning in place (CIP) can also contaminate the subsequent batches (
      • Flint S.H.
      • Bremer P.J.
      • Brooks J.D.
      Biofilms in dairy manufacturing plant—Description, current concerns and methods of control.
      ). In-process biofilms in the pasteurizer regeneration section have been discussed in previous studies (
      • Knight G.C.
      • Nicol R.S.
      • McMeekin T.A.
      Temperature steps changes: A novel approach to control biofilm of Streptococcus thermophilus in a pilot plant-scale cheese milk pasteurization plant.
      ). It has been found that 16 h of pasteurizer run can increase the bacteria in milk by log10 6.0/mL. Therefore, cheese manufacturers have adopted the practice of quickly washing the pasteurizer in the middle of the run (midday wash) to clean the regeneration section and reduce contamination in subsequent vats (
      • Lauzon N.
      Method for controlling microbiological contamination in a heat exchanger while processing a food product.
      ). This practice results in a loss of productivity but helps to control cheese quality.
      Cheese ripening involves complex biochemical reactions such as lactose hydrolysis, proteolysis, and lipolysis, resulting in characteristic flavor and texture development through the production of different peptides and organic acids (
      • Marth E.H.
      Microbiological and chemical aspects of cheddar ripening. A review.
      ). A diverse population of bacteria (starter or nonstarter) and their enzymes govern all the biochemical reactions. To date, not many published studies are available that can link thermoduric bacteria with any specific biochemical changes or related defects. In fact, we did not find any studies that demonstrate the changes in thermoduric bacteria population load at different stages of processing during cheese-making and ripening. This would be crucial information that would help cheese manufacturers to reduce and control thermoduric bacteria-originated cheese defects. Considering the magnitude of financial losses that can occur to a cheesemaker, it is vital to understand the growth dynamics and progression behavior of thermoduric spoilage bacteria in cheese-making. In this study, the focus is to investigate the changes in thermoduric bacteria population during continuous cheese manufacturing in a typical 20-h run and ripening. The effects of a midday wash in controlling thermoduric bacteria counts are also evaluated. This study further investigates the differences between the chemical composition in Cheddar cheese manufactured over the production day (samples drawn from the first vat, before a midday washing of the pasteurizer, after the midday washing of the pasteurizer, and from the last vat of the day). The hypothesis is that there will be increased thermoduric bacteria or more chemical changes between cheeses prepared at the start of the day versus the end of the day. We also studied thermoduric bacteria contaminants from the production environment and product contact surfaces.

      MATERIALS AND METHODS

      A commercial Cheddar cheese plant in the Midwestern United States was chosen for this study. This plant processes 42 vats of cheese every day (40,000 kg of milk per vat) with a midday wash of the pasteurizer. The starters used for this study were mesophilic Lactococcus lactis ssp. cremoris and Lactococcus lactis ssp. lactis along with adjunct mesophilic lactobacilli. Rennet was used as the coagulation agent. After setting of curd at 32°C, it was cut and cooked to 37°C for 18 min. The whole mass was pumped to a stirred curd belt (SCB). The SCB residence time was 90 min at 37°C. The target pH at the end of the curd belt was 5.45. Salt belt temperature was 37°C, and the curd took 30 min to cross it. A midday wash of the pasteurizer was carried out between vat 20 and vat 21 (after 10-h run). Only the pasteurizer was cleaned by a short CIP process of 15 min. The CIP steps included initial flushing of 2.0 min, followed by circulating 1% caustic for 8 min (80°C), 1.0 min flushing, 0.7% acid for 2.0 min (75°C) circulation, and 2.0-min final rinse. The plant ran for 10 more hours before starting the complete CIP cycle.

      Sampling Scheme During Cheese-Making Process

      Samples of 100 mL (or 100 g) were aseptically drawn from several stages of cheese manufacture (raw milk, pasteurized milk, fresh cheese block), at 4 time intervals during processing: the first vat of the day (vat 1, 0 h), the vat immediately before the midday pasteurizer wash (vat 20, 10 h), the vat immediately after the midday pasteurizer wash (vat 21, 11 h), and the last vat of the day (vat 42, 20 h). Duplicate samples were collected in 3 trials, performed in the months of August, November, and March. The samples were kept in a refrigerator and analyzed within 4 to 6 h of sampling. One 18.1-kg cheese block was collected from the cheese tower for each vat of interest. Upon arrival at the laboratory, the 18.1-kg block was cut longitudinally into 2 equal portions, vacuumed packed, and stored in a cheese cooler set to 4°C. At d 0, a longitudinal section was cut into ∼3-inch-wide slices, vacuum-packed, and numbered in nonsequential order for later analysis. This cheese was analyzed at d 0, 15, 60, 120, and 180.

      Biofilm Sampling of Cheese-Making Equipment and Plant Environment

      To evaluate the presence of biofilms in the cheese plants, we decided to sample product contact surfaces and the manufacturing environment. Sampling was performed using the swabbing method (
      • Graham T.
      Sampling Dairy and related products.
      ) on cheese vats, SCB, and the cheese tower. Swab samples were also taken from the regeneration and heating sections (before and after CIP) of the milk HTST pasteurizer. Rinse samples were collected after the CIP process to test for surviving bacteria [total viable counts (TVC) and thermoduric bacteria] in the post-CIP samples. Environmental swabs of floor areas were taken from the raw milk silo area, cheese vat area, SCB area, and cheese tower area, following the same procedure. Swabs (3M Quick Swab) containing Letheen medium and buffer components were used. For each sampling area, 5 spots of 50-cm2 area were swabbed. All swabs were transported to the laboratory in iceboxes at 4°C and analyzed within 4 to 6 h of sampling.

      Microbiological Analysis

      Liquid samples were diluted serially to get the desired dilutions. For cheese, 11 g of samples were taken and mixed in 99 mL of citrated or phosphate buffer using a stomacher (Seward, model 400). These stomached samples and swabs from the environment and product contact surface were transferred into sterile test tubes under a biosafety hood for further analysis. All samples were evaluated for TVC, thermoduric mesophile (TM) bacteria, and TT bacteria, on plate count agar at 32°C for 24 h, 32°C for 24 to 48 h, and 55°C for 24 to 48 h, respectively (
      • Frank J.F.
      • Yousef A.E.
      Tests for groups of microorganisms.
      ). For TM and TT bacteria, the samples were heated to 63°C for 30 min and allowed to cool (<10°C) before plating. After incubation, the colonies were counted and expressed as log10 cfu per gram or milliliter.

      Identification of Bacterial Isolates

      Based on different morphology, 131 thermoduric bacterial isolates from raw milk, pasteurized milk, cheese, product contact surfaces, and environment swabs were selected for identification. One set of isolates was sent for genetic identification to the Food Science Laboratory, Cornell University (Ithaca, NY), where identification was performed with PCR using a portion of the rpoB sequences according to previously documented procedures (
      • Huck J.R.
      • Woodcock N.H.
      • Ralyea R.D.
      • Boor K.J.
      Molecular subtyping and characterization of psychrotolerant endospore-forming bacteria in two New York State fluid milk processing systems.
      ). Identifications were based on phylogenetic comparisons of a 635-bp region of the rpoB gene. Raw sequence data were edited using Sequencher sequence analysis software 5.4.5 (Gene Codes Co.). The remaining isolates were further grown on 5% sheep blood agar and identified by MALDI-TOF at the Animal Disease Research and Diagnostic Laboratory, South Dakota State University, Brookings (
      • Buehner K.P.
      • Anand S.
      • Djira G.D.
      • Garcia A.
      Prevalence of thermoduric bacteria and spores on 10 Midwest dairy farms.
      ).

      Chemical Analysis

      Protein was measured by the Kjeldahl method (
      • Hool R.
      • Barbano D.M.
      • Budde D.
      • Bulthaus M.
      • Chettiar M.
      • Lynch J.
      • Reddy R.
      Chemical and physical methods.
      ) with Sharp's solution used for noncasein nitrogen (NCN) analysis (
      • Kosikowski F.
      • Mistry V.V.
      Cheese and Fermented Milk Foods.
      ) and 12% trichloroacetic acid (TCA) for NPN. Organic acids [citric acid, propionic acid, uric acid, orotic acid, and sugar (lactose)] were analyzed as described by
      • Upreti P.
      • Buhlmann P.
      • Metzger L.E.
      Influence of calcium and phosphorus, lactose, and salt-to-moisture ratio on cheddar cheese quality: pH buffering properties of cheese.
      ,
      • Upreti P.
      • McKay L.L.
      • Metzger L.E.
      Influence of calcium and phosphorus, lactose, and salt-to-moisture ratio on Cheddar cheese quality: Changes in residual sugars and water-soluble organic acids during ripening.
      .

      Total Protein

      Total protein was determined by a micro-Kjeldahl method using a 0.75-g cheese sample (
      • Hool R.
      • Barbano D.M.
      • Budde D.
      • Bulthaus M.
      • Chettiar M.
      • Lynch J.
      • Reddy R.
      Chemical and physical methods.
      ). One Kjeltab (Alfie Packers Inc. No TT-50), which provided 5 g of K2SO4, 0.15 g of CuSO4, and 0.15 g of TiO2, was added to the sample with 12 mL of concentrated sulfuric acid. Samples were digested at 420°C for 1 h. After cooling, samples were distilled using a Kjeltec 2200 (Foss). The distillate was captured in 30 mL of 4% boric acid solution. The boric acid containing the distillate was titrated using 0.1 N H2SO4.

      pH 4.6-Soluble Protein

      Primary proteolysis was measured by finding the amount of pH 4.6-soluble protein in the sample. This fraction is also referred to as NCN. Cheese (0.75 g) was weighed into a 50-mL plastic centrifuge tube and 20 mL of Sharp's 33 solutions (
      • Kosikowski F.
      • Mistry V.V.
      Cheese and Fermented Milk Foods.
      ) was added. Exact weights for all reagents were recorded to the fourth decimal place. The sample was homogenized using a high-shear Omni Mixer. Samples were centrifuged at 4,337 × g for 10 min and then cooled in a 4°C refrigerator for 10 min. The floating thin fatty layer was discarded, and the sample was filtered through Whatman filter paper no. 2. The extraction procedure was repeated an additional time, with the second filtrate being combined with the first. Filtrate sample (20 mL) was analyzed with the same Kjeldahl method described previously, with the addition of H2O2 to the digestion tube as an anti-foaming agent.

      12% TCA-Soluble Protein

      A 12% TCA-soluble protein was used to measure secondary proteolysis and is also referred to as nonprotein nitrogen (NPN). To determine its content in the samples, 1.5 g of cheese was weighed into a 50-mL plastic centrifuge tube, and 20 mL of 12% TCA solution was added. Exact weights for all reagents were recorded to the fourth decimal place. The sample was homogenized using a high-shear Omni Mixer. Samples were centrifuged at 4,337 × g for 10 min and then cooled in a 4°C refrigerator for 10 min. The floating fatty layer was discarded, and the sample was filtered through Whatman filter paper. The extraction procedure was repeated an additional time, with the second filtrate being combined with the first. Filtrate sample (20 mL) was analyzed with the same Kjeldahl method described earlier, with the addition of H2O2 to the digestion tube as an anti-foaming agent.

      Organic Acids and Sugars

      Samples were prepared based on the method described by
      • Upreti P.
      • Buhlmann P.
      • Metzger L.E.
      Influence of calcium and phosphorus, lactose, and salt-to-moisture ratio on cheddar cheese quality: pH buffering properties of cheese.
      ,b). Cheese (5 g) was homogenized with 10 g of 0.013 N H2SO4 that was warmed to 65°C before addition. Homogenization was performed for 90 s in a 50-mL plastic centrifuge tube using a high-shear Omni Mixer. After homogenization, samples were centrifuged at 4,337 × g for 10 min. After cooling for 10 min at 4°C, the fat layer was removed, and the samples were filtered through Whatman filter paper no. 4. Approximately 0.5 mL of filtrate was further filtered using a Microcon Centrifugal Filter (Millipore Corp.) with a 3-kDa molecular weight cut-off. The tubes were centrifuged at 14,000 × g for 15 min. This second filtrate was directly injected into the HPLC system. Samples were analyzed using the HPLC program outlined by
      • Amamcharla J.K.
      • Metzger L.E.
      Development of a rapid method for the measurement of lactose in milk using a blood glucose biosensor.
      . The HPLC system (Beckman Coulter) had 2 detectors: a refractive index detector (RI2031, Jasco Corporation) and a UV detector (System Gold 168 detector, Beckman Coulter), set to read at 210 nm and 280 nm, respectively. The column used was a 300- × 7.8-mm ion-exchange column (ROA-Organic 32 acid, Phenomenex Inc.) maintained at 65°C. Sulfuric acid (0.013 N) was used as the mobile phase.

      Statistical Analysis

      Statistical analyses were performed using SAS version 9.1 (

      SAS Institute Inc. 2003. User's Guide: Statistics. Version 9.1.

      ). Microbiological counts were logarithmically transformed before calculating means. The generalized linear model was used to analyze the variance. The ANOVA and significance were indicated by P < 0.05.

      RESULTS AND DISCUSSION

      Progression of TVC, TM, and TT at Different Steps of Cheese-Making

      Raw Milk

      One of the major sources of thermoduric bacteria in cheese is the raw milk used. Hence, raw milk of good microbial quality, with low SCC and absence of antibiotic residues, is recommended as the starting material for good-quality cheese (
      • Farkye N.Y.
      Cheese technology.
      ). In the present study, raw milk supplies were analyzed over 20 h. The filling and emptying of raw milk silos were a part of the daily milk reception routine in the plant. Raw milk samples were analyzed for TVC, TM, and TT (Table 1). The TVC of raw milk varied from log10 3.10 to 5.07 cfu/mL and was observed to increase as the day progressed (P < 0.05; vat 1 to vat 42). The average TVC in raw milk samples was log10 4.38 cfu/mL, of which log10 1.76 were TM bacteria and log10 1.44 were TT bacteria, which translates to 0.24% TM and 0.11% TT (calculated on original cfu/mL), respectively. These numbers provide us information about the proportion of thermoduric bacteria present in raw milk. This finding is in agreement with previous work reported by
      • Buehner K.P.
      • Anand S.
      • Djira G.D.
      • Garcia A.
      Prevalence of thermoduric bacteria and spores on 10 Midwest dairy farms.
      , who reported that raw milk in the Midwest region contains log10 1.85 cfu/mL of TM and log10 0.91 cfu/mL of TT. Further data analysis shows that TM and TT counts did not differ (P > 0.05) during the day. This finding agrees with
      • Celestino E.L.M.
      • Iyer H.
      • Roginski H.
      The effects of refrigerated storage on the quality of raw milk.
      ,
      • O'Connell A.
      • Ruegg P.L.
      • Jordan K.
      • O'Brien B.
      • Gleeson D.
      The effect of storage temperature and duration on the microbial quality of bulk tank milk.
      , and
      • VanderKelen J.J.
      • Mitchell R.D.
      • Laubscher A.
      • Black M.W.
      • Goodman A.L.
      • Montana A.K.
      • Dekhtyar A.M.
      • Jimenez-Flores R.
      • Kitts C.L.
      Short communication: Typing and tracking Bacillaceae in raw milk and milk powder using pyroprinting.
      , who suggested that thermoduric bacteria do not multiply during refrigerated storage and that their presence in bulk tank milk is most likely due to contamination from the environment and milking equipment. Thus, it can be concluded that, despite undefined mixing of raw milk in silos from different milk sources at different time intervals of the day, thermoduric bacteria counts were not significantly different (P > 0.05) and remained the same for the 20-h run of this study. This indicates that a subsequent increase in thermoduric bacteria would be the result of in-process multiplication or concentration.
      Table 1Total viable and thermoduric bacteria count during different stages of cheese-making (log10 cfu/mL or log10 cfu/g)
      Mean and SE of 3 trials sampled in replicates plated in duplicates. Midday cleaning was performed on pasteurizer between vat 20 and vat 21.
      Sampling pointVat 1, 0 hVat 20, 10 hVat 21, 11 hVat 42, 20 hAverage
      Total viable counts
       Raw milk3.10
      Means in the same row (Vat 1–Vat 42) within total viable counts, thermoduric mesophiles counts, and thermoduric thermophiles count having common superscripts do not differ (P < 0.05).
      ± 0.72
      4.47
      Means in the same row (Vat 1–Vat 42) within total viable counts, thermoduric mesophiles counts, and thermoduric thermophiles count having common superscripts do not differ (P < 0.05).
      ± 0.70
      4.87
      Means in the same row (Vat 1–Vat 42) within total viable counts, thermoduric mesophiles counts, and thermoduric thermophiles count having common superscripts do not differ (P < 0.05).
      ± 0.37
      5.07
      Means in the same row (Vat 1–Vat 42) within total viable counts, thermoduric mesophiles counts, and thermoduric thermophiles count having common superscripts do not differ (P < 0.05).
      ± 0.27
      4.38
      Means in the Average column with common superscripts do not differ (P < 0.05).
      ± 0.44
       Pasteurized milk1.93
      Means in the same row (Vat 1–Vat 42) within total viable counts, thermoduric mesophiles counts, and thermoduric thermophiles count having common superscripts do not differ (P < 0.05).
      ± 0.67
      3.13
      Means in the same row (Vat 1–Vat 42) within total viable counts, thermoduric mesophiles counts, and thermoduric thermophiles count having common superscripts do not differ (P < 0.05).
      ± 0.03
      2.90
      Means in the same row (Vat 1–Vat 42) within total viable counts, thermoduric mesophiles counts, and thermoduric thermophiles count having common superscripts do not differ (P < 0.05).
      ± 0.20
      2.50
      Means in the same row (Vat 1–Vat 42) within total viable counts, thermoduric mesophiles counts, and thermoduric thermophiles count having common superscripts do not differ (P < 0.05).
      ± 0.20
      2.62
      Means in the Average column with common superscripts do not differ (P < 0.05).
      ± 0.21
       Fresh cheese6.87
      Means in the same row (Vat 1–Vat 42) within total viable counts, thermoduric mesophiles counts, and thermoduric thermophiles count having common superscripts do not differ (P < 0.05).
      ± 0.87
      6.47
      Means in the same row (Vat 1–Vat 42) within total viable counts, thermoduric mesophiles counts, and thermoduric thermophiles count having common superscripts do not differ (P < 0.05).
      ± 0.57
      6.77
      Means in the same row (Vat 1–Vat 42) within total viable counts, thermoduric mesophiles counts, and thermoduric thermophiles count having common superscripts do not differ (P < 0.05).
      ± 0.83
      7.07
      Means in the same row (Vat 1–Vat 42) within total viable counts, thermoduric mesophiles counts, and thermoduric thermophiles count having common superscripts do not differ (P < 0.05).
      ± 0.60
      7.05
      Means in the Average column with common superscripts do not differ (P < 0.05).
      ± 0.15
      Thermoduric mesophiles counts
       Raw milk2.03
      Means in the same row (Vat 1–Vat 42) within total viable counts, thermoduric mesophiles counts, and thermoduric thermophiles count having common superscripts do not differ (P < 0.05).
      ± 0.13
      1.57
      Means in the same row (Vat 1–Vat 42) within total viable counts, thermoduric mesophiles counts, and thermoduric thermophiles count having common superscripts do not differ (P < 0.05).
      ± 0.27
      1.73
      Means in the same row (Vat 1–Vat 42) within total viable counts, thermoduric mesophiles counts, and thermoduric thermophiles count having common superscripts do not differ (P < 0.05).
      ± 0.33
      1.70
      Means in the same row (Vat 1–Vat 42) within total viable counts, thermoduric mesophiles counts, and thermoduric thermophiles count having common superscripts do not differ (P < 0.05).
      ± 0.20
      1.76
      Means in the same row (Vat 1–Vat 42) within total viable counts, thermoduric mesophiles counts, and thermoduric thermophiles count having common superscripts do not differ (P < 0.05).
      ± 0.10
       Pasteurized milk2.27
      Means in the same row (Vat 1–Vat 42) within total viable counts, thermoduric mesophiles counts, and thermoduric thermophiles count having common superscripts do not differ (P < 0.05).
      ± 0.07
      2.06
      Means in the same row (Vat 1–Vat 42) within total viable counts, thermoduric mesophiles counts, and thermoduric thermophiles count having common superscripts do not differ (P < 0.05).
      ± 0.03
      2.00
      Means in the same row (Vat 1–Vat 42) within total viable counts, thermoduric mesophiles counts, and thermoduric thermophiles count having common superscripts do not differ (P < 0.05).
      ± 0.20
      2.23
      Means in the same row (Vat 1–Vat 42) within total viable counts, thermoduric mesophiles counts, and thermoduric thermophiles count having common superscripts do not differ (P < 0.05).
      ± 0.03
      2.14
      Means in the same row (Vat 1–Vat 42) within total viable counts, thermoduric mesophiles counts, and thermoduric thermophiles count having common superscripts do not differ (P < 0.05).
      ± 0.06
       Fresh cheese3.47
      Means in the same row (Vat 1–Vat 42) within total viable counts, thermoduric mesophiles counts, and thermoduric thermophiles count having common superscripts do not differ (P < 0.05).
      ± 0.03
      3.43
      Means in the same row (Vat 1–Vat 42) within total viable counts, thermoduric mesophiles counts, and thermoduric thermophiles count having common superscripts do not differ (P < 0.05).
      ± 0.07
      2.83
      Means in the same row (Vat 1–Vat 42) within total viable counts, thermoduric mesophiles counts, and thermoduric thermophiles count having common superscripts do not differ (P < 0.05).
      ± 0.35
      2.63
      Means in the same row (Vat 1–Vat 42) within total viable counts, thermoduric mesophiles counts, and thermoduric thermophiles count having common superscripts do not differ (P < 0.05).
      ± 0.67
      3.09
      Means in the Average column with common superscripts do not differ (P < 0.05).
      ± 0.29
      Thermoduric thermophiles counts
       Raw milk1.57
      Means in the same row (Vat 1–Vat 42) within total viable counts, thermoduric mesophiles counts, and thermoduric thermophiles count having common superscripts do not differ (P < 0.05).
      ± 0.23
      1.43
      Means in the same row (Vat 1–Vat 42) within total viable counts, thermoduric mesophiles counts, and thermoduric thermophiles count having common superscripts do not differ (P < 0.05).
      ± 0.07
      1.33
      Means in the same row (Vat 1–Vat 42) within total viable counts, thermoduric mesophiles counts, and thermoduric thermophiles count having common superscripts do not differ (P < 0.05).
      ± 0.06
      1.43
      Means in the same row (Vat 1–Vat 42) within total viable counts, thermoduric mesophiles counts, and thermoduric thermophiles count having common superscripts do not differ (P < 0.05).
      ± 0.08
      1.44
      Means in the Average column with common superscripts do not differ (P < 0.05).
      ± 0.05
       Pasteurized milk1.60
      Means in the same row (Vat 1–Vat 42) within total viable counts, thermoduric mesophiles counts, and thermoduric thermophiles count having common superscripts do not differ (P < 0.05).
      ± 0.20
      2.83
      Means in the same row (Vat 1–Vat 42) within total viable counts, thermoduric mesophiles counts, and thermoduric thermophiles count having common superscripts do not differ (P < 0.05).
      ± 0.03
      1.47
      Means in the same row (Vat 1–Vat 42) within total viable counts, thermoduric mesophiles counts, and thermoduric thermophiles count having common superscripts do not differ (P < 0.05).
      ± 0.17
      2.53
      Means in the same row (Vat 1–Vat 42) within total viable counts, thermoduric mesophiles counts, and thermoduric thermophiles count having common superscripts do not differ (P < 0.05).
      ± 0.03
      2.11
      Means in the Average column with common superscripts do not differ (P < 0.05).
      ± 0.19
       Fresh cheese2.63
      Means in the same row (Vat 1–Vat 42) within total viable counts, thermoduric mesophiles counts, and thermoduric thermophiles count having common superscripts do not differ (P < 0.05).
      ± 0.44
      2.20
      Means in the same row (Vat 1–Vat 42) within total viable counts, thermoduric mesophiles counts, and thermoduric thermophiles count having common superscripts do not differ (P < 0.05).
      ± 0.67
      2.27
      Means in the same row (Vat 1–Vat 42) within total viable counts, thermoduric mesophiles counts, and thermoduric thermophiles count having common superscripts do not differ (P < 0.05).
      ± 0.83
      2.80
      Means in the same row (Vat 1–Vat 42) within total viable counts, thermoduric mesophiles counts, and thermoduric thermophiles count having common superscripts do not differ (P < 0.05).
      ± 0.60
      2.48
      Means in the Average column with common superscripts do not differ (P < 0.05).
      ± 0.14
      a–d Means in the same row (Vat 1–Vat 42) within total viable counts, thermoduric mesophiles counts, and thermoduric thermophiles count having common superscripts do not differ (P < 0.05).
      e–g Means in the Average column with common superscripts do not differ (P < 0.05).
      1 Mean and SE of 3 trials sampled in replicates plated in duplicates. Midday cleaning was performed on pasteurizer between vat 20 and vat 21.

      Pasteurized Milk

      Pasteurization aims to eliminate pathogens from milk and reduce the overall bacterial load. Raw milk is subjected to pasteurization as part of the cheese-making process. Although thermoduric bacteria survive pasteurization, TVC is generally reduced by at least log10 1 (
      • Grappin R.
      • Beuvier E.
      Possible implications of milk pasteurization on the manufacture and sensory quality of ripened cheese.
      ). In this study, the average TVC of raw milk was 4.38 log10/mL, which was reduced significantly to 2.62 log10/mL (P < 0.05; Table 1) by pasteurization (72°C for 16 s). Pasteurized milk TVC counts varied from 1.93 to 3.13 log10/mL over a day. This is in agreement with
      • Ranieri M.L.
      • Boor K.J.
      Short communication: Bacterial ecology of high-temperature, short-time pasteurized milk processed in the United States.
      , who reported similar counts in raw milk that varied from log10 1.0 to 2.26 cfu/mL. Pasteurization makes milk more vulnerable to bacterial growth, as it partially or entirely deactivates peroxidase enzymes and thus the milk's antibacterial defense system (
      • Grappin R.
      • Beuvier E.
      Possible implications of milk pasteurization on the manufacture and sensory quality of ripened cheese.
      ), as well as reducing the competitive heat-sensitive bacterial population. This is in contrast with findings reported by
      • Hileman J.L.
      Thermoduric bacteria in pasteurized milk. A review of literature.
      , who concluded that bacteria multiplied faster in raw milk than in pasteurized milk. Therefore, in our study, it is vital to know the surviving population of bacteria (i.e., thermoduric bacteria) and their contaminants through the pasteurization process and their effect on cheese quality. In this study, pasteurized milk comprised an average of log10 2.14 cfu/mL of TM and log10 2.11 cfu/mL of TT (Table 1), which are 33.11% and 31.0% of TVC, respectively (calculated on original cfu/mL), indicating the presence of thermoduric bacteria that are known to survive pasteurization. This is in agreement with
      • Patel G.B.
      • Blankenagel G.
      Bacterial counts of raw milk and flavor of the milk after pasteurization and storage.
      , who reported that the thermoduric bacterial counts in milk were below log10 3.0 cfu/mL. The effect of process run time (across vat population) and the effect of pasteurizer midday wash will be subsequently discussed in a separate section.

      Fresh Cheese

      Cheese contains a diverse pool of microorganisms. The addition of starter culture in the pasteurized milk during cheese-making increases the TVC population significantly (P < 0.05), to an average of log10 7.05 cfu/g in the d-0 cheese block (Table 1). The thermoduric bacteria carried over from pasteurized milk were also examined in this study. Cheese blocks' average TM and TT counts (log10 3.09 cfu/g and log10 2.48 cfu/g, respectively) were significantly higher (P < 0.05) than pasteurized milk, indicating the in-process multiplication of thermoduric bacteria and their concentration along with milk solids (Figure 1A). The TM and TT bacteria account for 0.01% and 0.003% (calculated on original cfu/mL) of total bacterial population on d-0 cheese, respectively, as the starter dominates the TVC. To date, no data is available on thermoduric bacterial population changes in cheese.
      Figure thumbnail gr1
      Figure 1Graphs and bar charts show means ± SE of 3 independent experiments performed during cheese-making and ripening. Four cheese vats were sampled: at the start of the run (vat 1), prior to midday washing of the pasteurizer (vat 20), immediately after midday washing of the pasteurizer (vat 21), and at the end of the run (vat 42). (A) Overall population changes of thermoduric bacteria at different stages of Cheddar cheese manufacturing. (B) Effects of midday pasteurizer washing on thermoduric bacteria counts in pasteurized milk. Changes in total viable count (TVC; C), thermoduric mesophiles (TM; D), and thermoduric thermophiles (TT; E) during ripening from 15 to 180 d.
      To understand the effect of run time on cheese microbial count, it is important to discuss the pasteurized milk carryover of thermoduric microflora. In our finding, across the vats (Figure 1B and Table 1), it was observed that the pasteurized milk had similar TM counts over the entire day [from vat 1 (log10 2.27 cfu/mL) to vat 20 (log10 2.06 cfu/mL) and vat 21 (log10 2 cfu/mL) to vat 42 (log10 2.23 cfu/mL); P > 0.05], but the TM population increased in cheese samples. The increase from pasteurized milk to cheese was as follows: for vat 1, from log10 2.27 to log10 3.47; vat 20, from log10 2.06 to log10 3.43; vat 21, from log10 2 to log10 2.83; and vat 42, from log10 2.23 to log10 3.63. Overall, the average increase was to log10 3.09 cfu/g (fresh cheese). These increases in counts were significant (P < 0.05) and can be linked to few possible factors. First is the process, wherein cheese curds took 2.5 h to transit from the cheese vat to the tower, which had a favorable temperature range (32–37°C) for the multiplication of TM bacteria (
      • Arce A.
      • Ustunol Z.
      Effect of microencapsulated ferrous sulfate particle size on Cheddar cheese composition and quality.
      ). Second is the concentration of TM while turning milk into curd as a process of cheese-making (Figure 1A), and third is the presence of curd or whey film on the SCB and cheese tower during the 20-h run, further supporting the growth and persistence of TM. The third possibility can be excluded based on data presented in Table 2, which indicates that the SCB and tower did not show the presence of thermoduric bacteria. The presence of thermoduric bacteria in the cheese environment and product contact surfaces was also documented and will be discussed later in this paper.
      Table 2Total viable counts and thermoduric bacteria (log10 cfu/g or log10 cfu/cm2) present in cheese environment and contact surface area
      Mean and SE of 3 trials sampled in replicates plated in duplicates. ND = not detectable: <10 cfu/cm2. CIP = cleaning in place, midday washing of pasteurizer between vats 20 and 21.
      SampleTotal viable countThermoduric bacteria
      MesophilesThermophiles
      Cheese environment (swabs)
       Raw milk silo area3.37 ± 0.531.32 ± 0.58ND
       Cheese vat area2.04 ± 0.56NDND
       Stirred curd belt area3.27 ± 0.63NDND
       Tower area3.18 ± 0.62NDND
      Contact surface area (swabs)
       Cheese vatNDNDND
       Stirred curd belt3.80 ± 0.60NDND
       Tower3.90 ± 0.60NDND
      HTST pasteurizer (swabs)
       Pre-CIP milk HTST regeneration section2.7 ± 0.331.2 ± 0.471.9 ± 0.13
       Pre-CIP milk HTST heating plates2.2 ± 0.45ND1.00 ± 0.21
       Post-CIP milk HTST regeneration1.30 ± 0.771.30 ± 0.741.30 ± 0.77
       Post-CIP milk HTST heatingNDNDND
      CIP rinse (liquid)
       Milk HTST pasteurizer—caustic cycleNDNDND
       Milk HTST pasteurizer—acid cycleNDNDND
       Milk HTST pasteurizer—sanitizer cycleNDNDND
      1 Mean and SE of 3 trials sampled in replicates plated in duplicates. ND = not detectable: <10 cfu/cm2. CIP = cleaning in place, midday washing of pasteurizer between vats 20 and 21.
      By contrast, the TT bacteria population in pasteurized milk increased over time (vat 1, 1.6; vat 20, 2.83; vat 21, 1.47; and vat 40, 2.53). This increase was significant between vat 1 versus vat 20 and vat 21 versus vat 42 (P < 0.05). It is important to mention here that we conducted a midday wash between vats 20 and 21. The average TT counts for all the vats was log10 2.11 cfu/mL (Table 1). Thermoduric thermophilic bacteria in the cheese blocks from vat 1 to vat 42 were similar in counts (P > 0.05). The average TT counts (for all vat cheeses) were log10 2.48 cfu/g (Table 1). Final Cheddar cheese blocks (d 0) from different vats were similar (P > 0.05) in TVC, TM bacteria count, and TT bacteria counts (Table 1). We hypothesized that the cheese made in the last vat before cleaning (vat 20 and vat 42) was likely to have a higher number of thermoduric bacteria counts, but our results showed otherwise. This will be discussed further, in the later section on ripening.

      Cross-Contamination Potential due to the Presence of Thermoduric Bacteria in Cheese Manufacturing Environment and on Contact Surfaces

      The results related to biofilm formation in the production environment, milk HTST pasteurizer (before and after CIP), product contact surface, and CIP rinse are presented in Table 2, where counts below 10 cfu/cm2 have been denoted as not detectable (ND). Table 2 shows the absence or non-detection (ND) of TT and TM in the manufacturing plant, except in the raw milk silo area. This could be attributable to the result of strict hygiene practices and good manufacturing practices employed in cheese plants. The raw milk silo area showed the highest thermoduric bacteria count of log10 1.32 cfu/cm2, compared with other process areas (ND). The product contact surfaces, including the cheese vat, SCB, and cheese tower, also showed low numbers (ND) of TT and TM. Low numbers of TM and TT in the product contact area and environment suggest that these contaminants may not affect the cheese quality. Only a few previous studies have linked biofilm populations to cheese defects.
      • Agarwal S.
      • Sharma K.
      • Swanson B.G.
      • Yüksel G.Ü.
      • Clark S.
      Nonstarter lactic acid bacteria biofilms and calcium lactate crystals in Cheddar cheese.
      correlated biofilms with cheese defects. Similarly, the attachment of both vegetative cells and spores of thermophilic bacilli and thermoduric streptococci to stainless steel has been demonstrated even in the absence of a conditioning film (
      • Knight G.C.
      • Nicol R.S.
      • McMeekin T.A.
      Temperature steps changes: A novel approach to control biofilm of Streptococcus thermophilus in a pilot plant-scale cheese milk pasteurization plant.
      ). Table 2 data also shows the thermoduric bacteria counts in a different area of the pasteurizer before and after CIP. The thermoduric bacteria count of the pasteurizer regeneration section was higher than that of the heating section plates. Regeneration section TM and TT counts were log10 1.2 and log10 1.9 cfu/cm2, respectively, and those of the heating section were ND and log10 1 cfu/cm2. This data shows that thermoduric biofilms were found to be more prominent in the regeneration section of the pasteurizer and less prominent in the heating section. According to
      • Lauzon N.
      Method for controlling microbiological contamination in a heat exchanger while processing a food product.
      , the bacterial population could reach a higher number in the regeneration section in 10 to 12 h or less and could start to detach from the surface and contaminate the milk in its path. This also supports our finding of a higher number of TT in vat 20 (10 h) and vat 42 (20 h) pasteurized milk. Furthermore, Table 3 shows that the isolates sampled from environment and product contact surfaces were identified as Bacillus, and these biofilms were predominating with gram-positive Bacillus licheniformis. These gram-positive bacilli are not novel to milk and milk products and are of concern to the pasteurizer.
      • Flint S.H.
      • Bremer P.J.
      • Brooks J.D.
      Biofilms in dairy manufacturing plant—Description, current concerns and methods of control.
      also summarized that gram-positive bacteria could proliferate in the regeneration of the pasteurizer and increase in number to log10 6 cfu/cm2, and may serve as a source of cross-contamination.
      Table 3Identification of thermoduric bacteria isolates during cheese processing and ripening
      Source, sampling point, and isolatesTotal isolatesNo. of isolates by speciesPercentage of total
      Cheese-making
       Raw milk22
      Bacillus licheniformis1568.2
      Bacillus subtilis418.2
      Bacillus clausii14.5
      Bacillus aerophilus14.5
      Staphylococcusspp.14.5
       Pasteurized milk15
      B. licheniformis853.3
      B. subtilis426.7
      Bacillus cereus320
       Cheese ripening70
      B. licheniformis5375.7
      B. subtilis1217.1
      B. cereus57.1
      Environment5
      B. licheniformis240
      B. subtilis120
      B. cereus120
      Staphylococcusspp.120
      Product contact surface19
       Stirred curd belt
      B. licheniformis631.6
       Cheese tower
      B. licheniformis315.8
       Milk HTST pasteurizer
      B. licheniformis526.3
       Regeneration plates
      B. licheniformis526.3
      Total131
      While discussing the presence of thermoduric bacteria on product contact surfaces, it is important to consider the CIP protocol. The CIP system is designed to clean all chemical and microbial deposits present in the pasteurizer, and its efficiency is generally measured by analyzing rinse quality. This study also analyzed the CIP rinse, and the data are shown in Table 2. The CIP rinse (milk HTST pasteurizer, at the end of caustic, acid, and sanitize cycles) had very low counts (ND) of TM and TT, indicating that CIP is effective. On the contrary, the post-CIP milk HTST pasteurizer swab for the regeneration section showed TM and TT counts of 1.3 and 1.3 cfu/cm2. The presence of bacterial biofilm on milk HTST regeneration plates after CIP indicates the inadequacy of CIP rinse for use as an indicator of cleanliness. The CIP was not able to clean the thermoduric biofilms, especially in the regeneration section of the pasteurizer. This is also supported by
      • Brooks J.D.
      • Flint S.H.
      Biofilms in the food industry: Problems and potential solutions.
      , who reported that some of the cleaning and disinfection protocols were not able to remove all the thermoduric spores. In our study, milk HTST pasteurizer swabs, especially from the regeneration section, were found to have high counts of B. licheniformis (Table 3). Thus, it is possible that after biofilm formation in the regeneration section of the HTST pasteurizer, cells, and sometimes spores, release or the biofilms break and are sloughed off, which leads to contamination of pasteurized milk and subsequent cheese (
      • Brooks J.D.
      • Flint S.H.
      Biofilms in the food industry: Problems and potential solutions.
      ). Bacillus population of log10 3 cfu/cm2 has also been reported to result in defects (
      • Brooks J.D.
      • Flint S.H.
      Biofilms in the food industry: Problems and potential solutions.
      ). However, in our studies, the levels were found to be lower (<log10 2 cfu/cm2) due to the midday wash of the pasteurizer.

      Effect of Midday Wash on Thermoduric Bacteria During Cheese Manufacturing and Ripening Processes

      It is well known that biofilms on food contact surfaces and in the food processing environment can promote the growth of bacteria and contaminants (
      • Brooks J.D.
      • Flint S.H.
      Biofilms in the food industry: Problems and potential solutions.
      ). Long run times offer abundant opportunities to form biofilms (
      • Lauzon N.
      Method for controlling microbiological contamination in a heat exchanger while processing a food product.
      ), which involve the steps of reversible attachment, irreversible attachment, early development of biofilm architecture, maturation, and final dispersion of biofilm (
      • Srey S.
      • Jahid K.I.
      • Ha S.
      Biofilm formation in food industries: A food safety concern.
      ). This study investigated the effect of midday washing on the quality of pasteurized milk and the subsequent cheese produced from it. As we have discussed, the raw milk thermoduric bacteria did not change significantly (P > 0.05) over the 20-h run. In this study, pasteurized milk from the first vat of the day (vat 1, 0 h), the vat immediately before the midday pasteurizer wash (vat 20, 10 h), the vat immediately after the midday pasteurizer wash (vat 21, 11 h), and the last vat of the day (vat 42, 20 h) was sampled for TVC, TM, and TT. The TM counts of pasteurized milk were not found to be significantly different (P > 0.05) between the vats (0–10 h and 11–20 h of the run). This concludes that the pasteurization process did not significantly (P > 0.05) support the growth of TM (Figure 1B) for each 10-h run. By contrast, TT counts of pasteurized milk showed significant differences (P < 0.05) between vats. In the first vat of the day (vat 1), the TT population was log10 1.6 cfu/mL, which increased to log10 2.83 cfu/mL after the 10-h pasteurizer run (vat 20), and reduced to log10 1.47 cfu/mL after the midday wash (vat 21; Figure 1B). This reduction of 1.47 log10/mL indicates that the midday wash somewhat controlled TT counts, which can also be validated by comparing pre- and post-midday wash counts presented in the previous section. This finding is in agreement with
      • Lehmann F.L.
      • Russell P.S.
      • Solomon L.S.
      • Murphy K.D.
      Bacterial growth during continuous milk pasteurisation.
      and
      • Lauzon N.
      Method for controlling microbiological contamination in a heat exchanger while processing a food product.
      , who reported that thermoduric bacteria grow in the regeneration section of the pasteurizer due to favorable growth temperature in the pasteurized milk side. On the contrary,
      • Scott S.A.
      • Brooks J.D.
      • Rakonjac J.
      • Walker K.M.R.
      • Flint S.H.
      The formation of thermophilic spores during the manufacturing of whole milk powder.
      reported that the pasteurization has very little influence on thermophile counts.
      • Scott S.A.
      • Brooks J.D.
      • Rakonjac J.
      • Walker K.M.R.
      • Flint S.H.
      The formation of thermophilic spores during the manufacturing of whole milk powder.
      also pointed out that their conclusions could be due to the shorter run in their process. Another documented study by
      • Knight G.C.
      • Nicol R.S.
      • McMeekin T.A.
      Temperature steps changes: A novel approach to control biofilm of Streptococcus thermophilus in a pilot plant-scale cheese milk pasteurization plant.
      found that Streptococcus thermophilus increased up to log10 6.0 cfu/mL during 16 h of operation. An increase in the thermoduric bacteria population of pasteurized milk up to log10 5.0 cfu/mL has been reported to cause yeasty taste and openness in cheese (
      • Knight G.C.
      • Nicol R.S.
      • McMeekin T.A.
      Temperature steps changes: A novel approach to control biofilm of Streptococcus thermophilus in a pilot plant-scale cheese milk pasteurization plant.
      ). In the present study, the fresh cheese thermoduric bacteria population did not exceed log10 3.5 cfu/mL during the 20-h run, due to the midday wash of the pasteurizer. We observed that biofilms were prominent only in the regeneration section (Table 2), as previously reported by
      • Hinton A.
      • Trinh K.
      • Brooks J.
      • Manderson G.
      Thermophile survival in milk fouling and on stainless steel during cleaning.
      .
      • Hinton A.
      • Trinh K.
      • Brooks J.
      • Manderson G.
      Thermophile survival in milk fouling and on stainless steel during cleaning.
      also found that fouled surfaces could have a thermophile bacterial population up to 105 cfu/cm2.
      • Knight G.C.
      • Nicol R.S.
      • McMeekin T.A.
      Temperature steps changes: A novel approach to control biofilm of Streptococcus thermophilus in a pilot plant-scale cheese milk pasteurization plant.
      also reported that biofilms could also be attached as dead cells to unfouled stainless-steel surfaces after 15 min of cleaning, rather than being removed. Our study suggested that the midday wash helped break the bacterial buildup cycle and helped to control cross-contamination of cheese with thermoduric bacteria.

      Changes During Cheese Ripening

      The TVC of cheese ripened for 180 d was observed in this experiment. As the cheese aged, the TVC (sum of starter and nonstarter bacteria) decreased from log10 8.01cfu/g (15-d-old cheese samples) to log10 5.84 cfu/g (6-mo-old cheese samples; Figure 1C). This result is similar to the work of
      • Banks J.M.
      • Williams A.G.
      The role of the nonstarter lactic acid bacteria in Cheddar cheese ripening.
      , who reported that the starter bacteria inactivate during the ripening of cheese.
      The TM bacteria population of milk at the end of pasteurization varied from log10 2 to 2.27cfu /mL (Table 1). The overall cheese TM population ranged from log10 2.77 to 3.47cfu/g during the entire 180-d ripening period (Figure 1D). The TM population showed an increasing trend in the first 60 d of ripening. This indicates that the first 60 d of ripening are essential in determining cheese quality. If spoilage-causing thermoduric bacteria predominated in the first 60 d, then these cheeses would have shown noticeable defects. We did not find any other previous studies showing thermoduric bacterial growth during cheese ripening.
      The TT bacteria population of milk at the end of pasteurization varied from log10 1.47 to 2.83 cfu/mL, as discussed previously (Table 1). The TT population was observed to decrease during the first 60 d (Figure 1E), after which the TT population increased until 180 d of ripening. The increase was not significant when compared with data from 15 d to 180 d (P > 0.05). The only other study that we found related to thermoduric bacteria growth in Cheddar cheese was performed by
      • Lehmann F.L.
      Thermoduric-thermophilic bacteria in continuous cheese making.
      . The author positively correlated the cheese graded at 180 d, between increasing TT bacteria numbers and run time without a midday wash. In a similar trend, we also found an increasing trend of TT during ripening from 60 to 180 d. Populations of TT bacteria did not increase in the initial phase but showed an increasing trend later (P > 0.05). One possible reason could be that cheese block temperature varied in the mesophilic range (
      • Sattin E.
      • Andreani N.A.
      • Carraro L.
      • Fasolato L.
      • Balzan S.
      • Novelli E.
      • Squartini A.
      • Telatin A.
      • Simionati B.
      • Cardazzo B.
      Microbial dynamics during shelf-life of industrial Ricotta cheese and identification of a Bacillus strain as a cause of a pink discolouration.
      ) during the early days of ripening. This is a favorable factor for the growth of mesophilic bacteria and makes these the predominating microflora. However, once the temperature attains the psychrotrophic range and the presence of competitive microflora (lactic acid bacteria, LAB) reaches a low number (
      • Grappin R.
      • Beuvier E.
      Possible implications of milk pasteurization on the manufacture and sensory quality of ripened cheese.
      ), TT with psychrotrophic growth capabilities take over and begin multiplying, resulting in a high number of TT during the later stages of ripening. This study found that the TT population did not go beyond log10 2.93 cfu/g (Figure 1E) until the end of normal ripening. These findings agree with
      • Sattin E.
      • Andreani N.A.
      • Carraro L.
      • Fasolato L.
      • Balzan S.
      • Novelli E.
      • Squartini A.
      • Telatin A.
      • Simionati B.
      • Cardazzo B.
      Microbial dynamics during shelf-life of industrial Ricotta cheese and identification of a Bacillus strain as a cause of a pink discolouration.
      , who found a similar trend of aerobic spore counts and concluded that the later increase in cheese bacteria count may be attributed mostly to heat-resistant bacteria (thermoduric bacteria and spores).
      Figure 1, Figure 1 show no differences observed in cheese TM and TT counts across the vats (vat 1 vs. vat 20, and vat 21 vs. vat 42) in 180 d of ripening. Thus, midday wash (between vat 20 and vat 21) seems a promising means to control the growth of thermoduric bacteria growth during cheese ripening.
      • Lehmann F.L.
      Thermoduric-thermophilic bacteria in continuous cheese making.
      also reported that an increase in TT positively correlates with the cheese graded at 180 d. In this study, we found that the TT population did not go beyond log10 3 cfu/g in cheese during ripening (Figure 1E). This is likely due to the midday wash of the pasteurizer, which flushes the contaminating biofilm after 10 h of run time and results in ripened cheese with low TT counts and no defects noticeable at the end of 180 d. At the end of the 180-d ripening period of this study, none of the cheeses investigated exhibited body defects, as determined by a trained analyst. This lack of defects was not surprising and also supported our conclusion that midday wash and plant hygiene practices restrict the growth and attachment of thermoduric bacteria and any related noticeable defects.

      Identification of the Isolates

      Identification via PCR and MALDI-TOF (Table 3) revealed that B. licheniformis was the predominant thermoduric bacteria in raw milk, pasteurized milk, Cheddar cheese, environmental swabs, and food contact surfaces. The raw milk thermoduric bacteria isolates showed 68.2% B. licheniformis, 18.2% Bacillus subtilis, 4.5% of Bacillus clausii, Bacillus aerophilus, and Staphylococcus spp. These results were similar to the previous findings of
      • Banykó J.
      • Vyletělová M.
      Determining the source of Bacillus cereus and Bacillus licheniformis isolated from raw milk, pasteurized milk, and yoghurt.
      and
      • Buehner K.P.
      • Anand S.
      • Djira G.D.
      • Garcia A.
      Prevalence of thermoduric bacteria and spores on 10 Midwest dairy farms.
      , who have also isolated many Bacillus species from raw milk, with predominance of B. licheniformis. Most of the thermoduric bacteria present in pasteurized milk belonged to the genus Bacillus. Our identification showed that 53.3% B. licheniformis, 26.7% B. subtilis, and 20% Bacillus cereus constituted the thermoduric bacteria population in pasteurized milk. The presence of Bacillus in pasteurized milk was also reported by
      • Ranieri M.L.
      • Huck J.R.
      • Sonnen M.
      • Barbano D.M.
      • Boor K.J.
      High temperature, short-time pasteurization temperatures inversely affect bacterial numbers during refrigerated storage of pasteurized fluid milk.
      and
      • Martin N.H.
      • Ranieri M.L.
      • Murphy S.C.
      • Ralyea R.D.
      • Wiedmann M.
      • Boor K.J.
      Results from raw milk microbiological tests do not predict the shelf-life performance of commercially pasteurized fluid milk.
      ,
      • Martin N.H.
      • Ranieri M.L.
      • Wiedmann M.
      • Boor K.J.
      Reduction of pasteurization temperature leads to lower bacterial outgrowth in pasteurized fluid milk during refrigerated storage: A case study.
      . This carryover of the thermoduric bacteria population in the final Cheddar cheese comprised 75.7% B. licheniformis, 17.1% B. subtilis, and 7.1% B. cereus. Not many studies have reported on the presence and effect of B. licheniformis on Cheddar cheese quality.
      • Banykó J.
      • Vyletělová M.
      Determining the source of Bacillus cereus and Bacillus licheniformis isolated from raw milk, pasteurized milk, and yoghurt.
      mentioned that B. licheniformis was the dominant member of the microbial community isolated from the dairy samples they examined. In the current study, the presence of B. licheniformis biofilms in the regeneration section, SCB, and cheese tower clearly indicates its growth and multiplication during cheese-making (Table 3). Environmental samples were also positive for B. licheniformis, B. subtilis, and B. cereus.

      Chemical Changes in Ripening Proteolysis

      Noncasein and nonprotein nitrogen, as a ratio of total protein, were considered markers of primary proteolysis and secondary proteolysis, respectively. We expected to see changes in proteolysis as the thermoduric bacteria population changed in cheese. The NCN/protein and NPN/protein percentage values increased significantly (P < 0.05) over time (15–180 d; Figure 2, Figure 2). We did observe an increase in TM for the first 60 d (Figure 1D) and TT from 2 to 180 d (Figure 1E), with an overall decreasing trend for TVC (Figure 1C). The NCN/protein and NPN/protein percentage values did not change significantly between vats (P > 0.05), which is consistent with changes in TM and TT populations across the vats. This, however, was contrary to the hypothesis that there would be an increase in proteolysis and thermoduric bacteria in vats made later in the day compared with those made earlier in the day. This result could very well be a result of the midday wash. Figure 2A represents NCN/protein percentage data organized by vat over time, and Figure 2B represents the NPN/protein percentage data in the same manner. The NCN/protein percentage ranged from an average of 9.2% at 15 d after manufacture to 24.9% after 180 d. The NPN/protein percentage values averaged 4.3% at 15 d and 14.0% after 180 d. Previous reports have indicated that NPN/protein percentage values are half of NCN/protein percentage values (
      • Lau K.Y.
      • Barbano D.M.
      • Rasmussen R.R.
      Influence of pasteurization of milk on protein breakdown in cheddar cheese during aging.
      ). This trend is consistent with the findings of this experiment. At 180 d after manufacture, NCN/protein percentage values ranged from 24.2 to 25.9%. This agrees with previously published values of 6-mo-old Cheddar (
      • O'Keeffe R.B.
      • Fox P.F.
      • Daly C.
      Contribution of rennet and starter proteases to proteolysis in cheddar cheese.
      ). The NPN/protein percentage values at 6 mo ranged from 13.3% to 14.8%, also agreeing with other published works.
      Figure thumbnail gr2
      Figure 2Cheese proteolysis over 6-mo ripening period across the cheese vats. Four cheese vats were sampled: at the start of the run (vat 1), prior to midday washing of the pasteurizer (vat 20), immediately after midday washing of the pasteurizer (vat 21), and at the end of the run (vat 42). (A) Primary proteolysis (NCN/protein, %); (B) secondary proteolysis (NPN/protein, %).

      Sugars and Organic Acids

      Lactose concentration decreased over the 180-d ripening period, as is expected in Cheddar because of LAB metabolism (data not shown), with values averaging 0.47% at 15 d and 0.24% at 180 d. It is interesting to note that the thermoduric bacteria population did not change across the vat made from 0 to 10 h and 11 to 20 h, and, at the same time, we found no significant differences in organic acid and sugar contents (P > 0.05) across the vats analyzed. The exception to this was orotic and uric acids, as significantly higher levels (P < 0.05) of both were present in the last vat of the day, compared with the other 3 vats sampled (Figure 3, Figure 3). Uric and orotic acids are endogenous to cow milk, being products of ruminant metabolism.
      • Upreti P.
      • McKay L.L.
      • Metzger L.E.
      Influence of calcium and phosphorus, lactose, and salt-to-moisture ratio on Cheddar cheese quality: Changes in residual sugars and water-soluble organic acids during ripening.
      linked the increase in orotic acid to calcium, phosphorus, lactose, and ripening time, and uric acid to lactose, salt and moisture, and ripening time, respectively. We also note that these acid contents reduce with the washing of curd, and the increased acid content was also more pronounced at the end of ripening, indicating that they could be a result of some nonstarter LAB and thermoduric metabolism (
      • Upreti P.
      • McKay L.L.
      • Metzger L.E.
      Influence of calcium and phosphorus, lactose, and salt-to-moisture ratio on Cheddar cheese quality: Changes in residual sugars and water-soluble organic acids during ripening.
      ). Because the pasteurizer was flushed midday and the cheese vats were cleaned after each curd transfer, we hypothesized that the accumulation of these acids, calcium, phosphorus, and lactose occurred on the SCB and in the cheese tower, and resulted in increased orotic acid and uric acid in samples from the final vat during ripening. No references were found to explain this accumulation over the production day or its relation to thermoduric bacteria.
      Figure thumbnail gr3
      Figure 3Production of orotic acid (A), uric acid (B), citric acid (C), and propionic acid (D) during ripening across the cheese vats. Four cheese vats were sampled: at the start of the run (vat 1), prior to midday washing of the pasteurizer (vat 20), immediately after midday washing of the pasteurizer (vat 21), and at the end of the run (vat 42).
      Over 180 d, levels of orotic, uric, citric, and propanoic acid all showed significant changes (P < 0.05; Figure 3A–D). The citric acid concentration was significantly different between 30 and 60 d and 60 and 120 d, with an overall downward trend in concentration over the entire ripening period (Figure 3C). Citrate is an intermediate product of LAB metabolism, and its levels are likely to fluctuate with the concentration of other organic acids and microbial ecology as cheese ripens. Propionic acid showed a significant difference in concentration between the last 2 analysis time points, 120 and 180 d, compared with 15 d. It exhibited an overall trend to increase over the 180 d, agreeing with work previously conducted by
      • Upreti P.
      • McKay L.L.
      • Metzger L.E.
      Influence of calcium and phosphorus, lactose, and salt-to-moisture ratio on Cheddar cheese quality: Changes in residual sugars and water-soluble organic acids during ripening.
      .

      Bacillus spp. Concerning Cheese-Making and Ripening

      Bacilli are aerobic and facultative anaerobic spore-forming bacteria that pose persistent quality issues in the milk industry. Their heat-resistant spores and the ability of vegetative cells to produce extracellular enzymes cause spoilage in milk and milk products. Bacillus spp. are quite common in the farm environment and may contaminate milk via various sources such as inadequate udder hygiene, soil, feed, dust, and other sources during storage and processing. Species such as B. licheniformis, B. subtilis, and B. cereus are most commonly isolated from raw milk (
      • Janštová B.L.J.
      • Lukášová J.
      Heat resistance of Bacillus spp. spores isolated from cow's milk and farm environment.
      ). In our study, we also found a similar thermoduric bacteria profile in final cheese and environment samples (Table 3). Bacillus licheniformis was the most frequently isolated bacteria in this study and was recovered at all production stages sampled in cheese manufacturing.
      • Janštová B.L.J.
      • Lukášová J.
      Heat resistance of Bacillus spp. spores isolated from cow's milk and farm environment.
      did some work on the heat resistance of Bacillus spp. The author reported that B. licheniformis was more heat resistant than other serotypes of this genus when 21 spores were compared with the spores of 18 isolates of B. subtilis and 6 of B. cereus. Those authors also assigned B. cereus the second most heat resistivity.

      CONCLUSIONS

      We can conclude from this study that the long hours of the continuous cheese-making process provide mesophilic and thermophilic growth opportunities (temperature, time, and concentration) for thermoduric bacteria. Introducing a midday wash of the pasteurizer during the cheese-making process appeared to be effective in restricting the growth of thermoduric bacteria and did not result in any unusual biochemical changes or noticeable defects. Bacillus licheniformis was the most frequently isolated bacteria in this study and was recovered at all production stages sampled during the cheese-making as well as from the environment. Other species, such as B. subtilis and B. cereus, were also isolated. This study also provides quantitative data on progression of thermoduric bacteria during typical cheese manufacturing and ripening. This information can be used to assist cheese manufacturers in their continuing efforts to improve the quality and microbial safety of cheese.

      ACKNOWLEDGMENTS

      We acknowledge Dairy Management Inc. (Rosemont, IL) for funding this project and the Dairy Research Institute (Rosemont, IL) for their administrative support. Infrastructural support from the Agricultural Experimental Station at South Dakota State University (Brookings) is also acknowledged. We thank Seema Das, from the Animal Disease Research and Diagnostic Laboratory, and Gemechis Djira, from the Department of Mathematics, South Dakota State University (Brookings), for their assistance. The authors have not stated any conflicts of interest.

      REFERENCES

        • Agarwal S.
        • Sharma K.
        • Swanson B.G.
        • Yüksel G.Ü.
        • Clark S.
        Nonstarter lactic acid bacteria biofilms and calcium lactate crystals in Cheddar cheese.
        J. Dairy Sci. 2006; 89 (16606716): 1452-1466
        • Amamcharla J.K.
        • Metzger L.E.
        Development of a rapid method for the measurement of lactose in milk using a blood glucose biosensor.
        J. Dairy Sci. 2011; 94: 4800-4809
        • Arce A.
        • Ustunol Z.
        Effect of microencapsulated ferrous sulfate particle size on Cheddar cheese composition and quality.
        J. Dairy Sci. 2018; 101 (29729915): 6814-6822
        • Banks J.M.
        • Williams A.G.
        The role of the nonstarter lactic acid bacteria in Cheddar cheese ripening.
        Int. J. Dairy Technol. 2004; 57: 145-152
        • Banykó J.
        • Vyletělová M.
        Determining the source of Bacillus cereus and Bacillus licheniformis isolated from raw milk, pasteurized milk, and yoghurt.
        Lett. Appl. Microbiol. 2009; 48 (19187503): 318-323
        • Brooks J.D.
        • Flint S.H.
        Biofilms in the food industry: Problems and potential solutions.
        Int. J. Food Sci. Technol. 2008; 43: 2163-2176
        • Buehner K.P.
        • Anand S.
        • Djira G.D.
        • Garcia A.
        Prevalence of thermoduric bacteria and spores on 10 Midwest dairy farms.
        J. Dairy Sci. 2014; 97 (25200773): 6777-6784
        • Burgess S.A.
        • Lindsay D.
        • Flint S.H.
        Thermophilic bacilli and their importance in dairy processing.
        Int. J. Food Microbiol. 2010; 144 (21047695): 215-225
        • Celestino E.L.M.
        • Iyer H.
        • Roginski H.
        The effects of refrigerated storage on the quality of raw milk.
        Aust. J. Dairy Technol. 1996; 51: 59-63
        • Dasgupta A.P.
        • Hull R.R.
        Late blowing of swiss cheese: Incidence of Clostridium tyrobutyricum in manufacturing milk.
        Aust. J. Dairy Technol. 1989; 44: 82-87
        • Farkye N.Y.
        Cheese technology.
        Int. J. Dairy Technol. 2004; 57: 91-98
        • Flint S.H.
        • Bremer P.J.
        • Brooks J.D.
        Biofilms in dairy manufacturing plant—Description, current concerns and methods of control.
        Biofouling. 1997; 11: 81-97
        • Frank J.F.
        • Yousef A.E.
        Tests for groups of microorganisms.
        in: Standard Methods for the Examination of Dairy Products. 17th ed. American Public Health Association, 2004: 229-234
        • González-Rivas F.
        • Ripolles-Avila C.
        • Fontecha-Umaña F.
        • Ríos-Castillo A.G.
        • Rodríguez-Jerez J.J.
        Biofilms in the spotlight: Detection, quantification, and removal methods.
        Compr. Rev. Food Sci. Food Saf. 2018; 17 (33350156): 1261-1276
        • Graham T.
        Sampling Dairy and related products.
        in: Standard Methods for the Examination of Dairy Products. 17th ed. American Public Health Association, 2004: 81-87
        • Grappin R.
        • Beuvier E.
        Possible implications of milk pasteurization on the manufacture and sensory quality of ripened cheese.
        Int. Dairy J. 1997; 7: 751-761
        • Hileman J.L.
        Thermoduric bacteria in pasteurized milk. A review of literature.
        J. Dairy Sci. 1940; 23: 1143-1160
        • Hinton A.
        • Trinh K.
        • Brooks J.
        • Manderson G.
        Thermophile survival in milk fouling and on stainless steel during cleaning.
        Food Bioprod. Process. 2002; 80: 299-304
        • Hool R.
        • Barbano D.M.
        • Budde D.
        • Bulthaus M.
        • Chettiar M.
        • Lynch J.
        • Reddy R.
        Chemical and physical methods.
        in: Standard Methods for the Examination of Dairy Products. 17th ed. American Public Health Association, 2004: 480-510
        • Huck J.R.
        • Woodcock N.H.
        • Ralyea R.D.
        • Boor K.J.
        Molecular subtyping and characterization of psychrotolerant endospore-forming bacteria in two New York State fluid milk processing systems.
        J. Food Prot. 2007; 70 (17969618): 2354-2364
        • Hull R.R.
        • Roberts A.V.
        • Mayes J.J.
        Association of Lactobacillus casei with soft-body defect in commercial mozzarella cheese.
        Aust. J. Dairy Technol. 1983; 38: 78-80
        • Hull R.R.
        • Toyne S.
        • Haynes I.N.
        • Lehmann F.L.
        Thermoduric bacteria: A re-emerging problem in cheesemaking.
        Aust. J. Dairy Technol. 1992; 47: 91-94
        • Janštová B.L.J.
        • Lukášová J.
        Heat resistance of Bacillus spp. spores isolated from cow's milk and farm environment.
        Acta Vet. Brno. 2001; 70: 179-184
        • Klijn N.
        • Nieuwenhof F.F.
        • Hoolwerf J.D.
        • van der Waals C.B.
        • Weerkamp A.H.
        Identification of Clostridium tyrobutyricum as the causative agent of late blowing in cheese by species-specific PCR amplification.
        Appl. Environ. Microbiol. 1995; 61 (7487024): 2919-2924
        • Knight G.C.
        • Nicol R.S.
        • McMeekin T.A.
        Temperature steps changes: A novel approach to control biofilm of Streptococcus thermophilus in a pilot plant-scale cheese milk pasteurization plant.
        Int. J. Food Microbiol. 2004; 93 (15163587): 305-318
        • Kosikowski F.
        • Mistry V.V.
        Cheese and Fermented Milk Foods.
        3rd ed. F. V. Kosikowski LLC, 1997
        • Lau K.Y.
        • Barbano D.M.
        • Rasmussen R.R.
        Influence of pasteurization of milk on protein breakdown in cheddar cheese during aging.
        J. Dairy Sci. 1991; 74: 727-740
        • Lauzon N.
        Method for controlling microbiological contamination in a heat exchanger while processing a food product.
        (N. Lauzon, asignee. Pat. no. US10010090B2.)
        • Lehmann F.L.
        Thermoduric-thermophilic bacteria in continuous cheese making.
        Aust. J. Dairy Technol. 1992; 47: 94-96
        • Lehmann F.L.
        • Russell P.S.
        • Solomon L.S.
        • Murphy K.D.
        Bacterial growth during continuous milk pasteurisation.
        Aust. J. Dairy Technol. 1992; 47: 28-32
        • Marth E.H.
        Microbiological and chemical aspects of cheddar ripening. A review.
        J. Dairy Sci. 1963; 46: 869-890
        • Martin N.H.
        • Ranieri M.L.
        • Murphy S.C.
        • Ralyea R.D.
        • Wiedmann M.
        • Boor K.J.
        Results from raw milk microbiological tests do not predict the shelf-life performance of commercially pasteurized fluid milk.
        J. Dairy Sci. 2011; 94 (21338787): 1211-1222
        • Martin N.H.
        • Ranieri M.L.
        • Wiedmann M.
        • Boor K.J.
        Reduction of pasteurization temperature leads to lower bacterial outgrowth in pasteurized fluid milk during refrigerated storage: A case study.
        J. Dairy Sci. 2012; 95 (22192227): 471-475
        • McGuiggan J.T.M.
        • McCleery D.R.
        • Hannan A.
        • Gilmour A.
        Aerobic spore-forming bacteria in bulk raw milk: Factors influencing the numbers of psychrotrophic, mesophilic and thermophilic Bacillus spores.
        Int. J. Dairy Technol. 2002; 55: 100-107
        • O'Connell A.
        • Ruegg P.L.
        • Jordan K.
        • O'Brien B.
        • Gleeson D.
        The effect of storage temperature and duration on the microbial quality of bulk tank milk.
        J. Dairy Sci. 2016; 99 (26947309): 3367-3374
        • O'Keeffe R.B.
        • Fox P.F.
        • Daly C.
        Contribution of rennet and starter proteases to proteolysis in cheddar cheese.
        J. Dairy Res. 1976; 43: 97-107
        • Ortakci F.
        • Broadbent J.R.
        • Oberg C.J.
        • McMahon D.J.
        Late blowing of Cheddar cheese induced by accelerated ripening and ribose and galactose supplementation in presence of a novel obligatory heterofermentative nonstarter Lactobacillus wasatchensis.
        J. Dairy Sci. 2015; 98 (26298753): 7460-7472
        • Patel G.B.
        • Blankenagel G.
        Bacterial counts of raw milk and flavor of the milk after pasteurization and storage.
        Milk Food Technol. 1972; 35: 203-206
        • Ranieri M.L.
        • Boor K.J.
        Short communication: Bacterial ecology of high-temperature, short-time pasteurized milk processed in the United States.
        J. Dairy Sci. 2009; 92 (19762798): 4833-4840
        • Ranieri M.L.
        • Huck J.R.
        • Sonnen M.
        • Barbano D.M.
        • Boor K.J.
        High temperature, short-time pasteurization temperatures inversely affect bacterial numbers during refrigerated storage of pasteurized fluid milk.
        J. Dairy Sci. 2009; 92 (19762797): 4823-4832
      1. SAS Institute Inc. 2003. User's Guide: Statistics. Version 9.1.

        • Sattin E.
        • Andreani N.A.
        • Carraro L.
        • Fasolato L.
        • Balzan S.
        • Novelli E.
        • Squartini A.
        • Telatin A.
        • Simionati B.
        • Cardazzo B.
        Microbial dynamics during shelf-life of industrial Ricotta cheese and identification of a Bacillus strain as a cause of a pink discolouration.
        Food Microbiol. 2016; 57 (27052696): 8-15
        • Scott S.A.
        • Brooks J.D.
        • Rakonjac J.
        • Walker K.M.R.
        • Flint S.H.
        The formation of thermophilic spores during the manufacturing of whole milk powder.
        Int. J. Dairy Technol. 2007; 60: 109-117
        • Srey S.
        • Jahid K.I.
        • Ha S.
        Biofilm formation in food industries: A food safety concern.
        Food Control. 2013; 31: 572-585
        • Sutariya S.
        • Sunkesula V.
        • Bhanduriya K.
        • Jhanwar A.
        A novel organism Lactobacillus wasatchensis: Growth, detection, gassing defects in cheese, control strategy and future research opportunities: A review.
        Asian J. Dairy Food Res. 2020; 39: 91-97
        • Upreti P.
        • Buhlmann P.
        • Metzger L.E.
        Influence of calcium and phosphorus, lactose, and salt-to-moisture ratio on cheddar cheese quality: pH buffering properties of cheese.
        J. Dairy Sci. 2006; 89 (16507688): 938-950
        • Upreti P.
        • McKay L.L.
        • Metzger L.E.
        Influence of calcium and phosphorus, lactose, and salt-to-moisture ratio on Cheddar cheese quality: Changes in residual sugars and water-soluble organic acids during ripening.
        J. Dairy Sci. 2006; 89 (16428613): 429-443
        • USDA
        Dairy product production—July 21. Cheese Market News, the Weekly Newspaper of the Nation's Cheese and Dairy/Deli Business.
        • VanderKelen J.J.
        • Mitchell R.D.
        • Laubscher A.
        • Black M.W.
        • Goodman A.L.
        • Montana A.K.
        • Dekhtyar A.M.
        • Jimenez-Flores R.
        • Kitts C.L.
        Short communication: Typing and tracking Bacillaceae in raw milk and milk powder using pyroprinting.
        J. Dairy Sci. 2016; 99 (26585475): 146-151