Introduction

A global trend in agriculture consists in producing safe foods for human consumption. It is an ongoing concern to all parts of the dairy industry -and farms are not an exception- to guarantee milk quality (^{Calderón et al., 2006}). Excellent milk should conform to food safety and quality regulations (Minprotección, Colombia, 2006). To succeed on this, the following parameters should be obtained: protein >3.2%, fat >3.5%, total solids >12.2%, mesophilic count <50,000 CFU/mL, somatic cell count (SCC) <100,000 SC/mL (Calderón et al., 2006), and absence of foreign substances or inhibitors. Prior to pasteurization, high counts of mesophilic and somatic cells favor the release of proteolytic enzymes, which decrease the shelf life of the product (Calderón et al., 2006) affecting milk yield and physicochemical parameters during milk coagulation processes (Calderón et al., 2012).

In regard to public health, there are obvious risks related to the consumption of raw milk. The Colombian government has issued a regulatory framework to rule the market and consumption of raw milk in order to minimize health risks for humans (^{Calderón et al., 2006}; Minprotección, 2006; Minprotección, 2011).

Mesophilic bacterial count, which grow between 20 and 40 °C is an indicator of the microbiological quality of raw milk (^{Chacón, 2004}). Nevertheless, a lower number of mesophilic counts do not assure the lack of pathogens or their toxins.

Somatic cells count is indicative of inflammation of the mammary gland, which results from the attack of pathogens or other factors, such as trauma. SCC consist of epithelial cells or desquamation cells (2%) and leukocyte migration to the epithelium. When infection develops, the diapedesis phenomenon increases, which is also a non-specific immune response. This phagocytic response against infectious agents helps repairing the alveolar epithelium and return to normal function. The former action is mediated by proteases, lipases and phospholipases. These enzymes and bacterial inhibitors are incorporated into the milk and accelerate the decomposition of proteins and fats. Consequently, somatic cells counts increase with the severity of the infection. Hence, somatic cells quantification is a key parameter to evaluate udder health and milk quality (^{Cerón et al., 2007}; ^{Calderón et al., 2012}). According to the above, the general objective of this study was to assess somatic cell count, physicochemical and microbiological quality of raw milk produced in northwestern Colombia.

Material and methods

Location

This study comprised five areas in northwestern Colombia, including the following municipalities: Zone 1: Agustin Codazzi, San Diego, La Paz, and Valledupar. Zone 2: Betania, Bosconia and El Copey. Zone 3: Curumaní and Becerril. Zone 4: Nueva Granada and Santa Ana. Zone 5: Urumita and Villanueva. Zones 1, 2, and 3 are located in Cesar Province, Zone 4 in Magdalena province, and Zone 5 in La Guajira Province.

Sample size

A convenience non-probability sampling method was used to take 186 milk samples from a farm devoted to milk and beef production (dual-purpose system).

Selection criteria

The farms were providers of raw milk for a well-known dairy processing company. They were managed under a dual-purpose system and used cooling tanks to preserve milk in individual or collective tanks from daily milking until dispatch to the milk processing plant.

Sampling

Each tank (individual and collective) was stirred for 10 minutes. Then, samples were manually collected with a sterile stainless steel sampler. 20 mL of raw milk was transferred to a labeled sterile bottle, which was then stored under refrigeration (4 °C).

Physicochemical parameters

Determination of the percentages of protein, fat, and total solids was done in an equipment with infrared technology (MilkoScan™ FT 120; Foss, Hillerød, Denmark).

Microbiological parameters

Count of mesophilic bacteria in milk was performed in an equipment with flow cytometry technology (BactoScan™ FC; Foss, Hillerød, Denmark).

Somatic cell count (SCC)

SCC was performed by fluorescence using an optical cell counter (DeLaval, 2005. Tumba, Sweden).

Statistical analysis

A descriptive analysis of SCC, physicochemical and microbiological variables was performed. Comparisons were made between types of milk storage tanks (individual and collective) and areas of study, using the test of Kruskal Wallis. The SAS software (SAS Institute Inc, Cary, NC, USA) was used.

Results

Table 1 shows the number of farms by area and Table 2 shows the descriptive statistics analysis of SCC, physicochemical and microbiological parameters of raw milk.

Table 1 Number of farms by area of study in Northwestern Colombia. 

Table 2 Overall quality of raw milk in Northwestern Colombia. 

Physicochemical quality of raw milk in each geographic region is shown in Figure 1.

Figure 1 Physicochemical quality of raw milk in Northwestern Colombia. 

Microbiological quality and SCC of raw milk in each geographic region are shown in Figure 2.

Figure 2 Mesophilic (CFU/mL) and somatic cell count (SC/mL) in raw milk in Northwestern Colombia . 

Table 3 shows the distribution of mesophilic count of raw milk in the dairy processing plant. 

Table 4 shows the distribution of somatic cell count, which is an indicator of udder health and mastitis in cattle from farms located in Northwestern Colombia.

Table 4 Distribution of somatic cell count in raw milk from farms in Northwestern Colombia. 

Table 5 shows the distribution of SCC, physicochemical and microbiological quality of raw milk by type of storage tank (individual or collective) collected at a dairy processing plant in Valledupar (Colombia).

Tabla 5 Distribution of somatic cell count, physicochemical and microbiological quality of raw milk collected at a dairy processing plant in Valledupar (Colombia), depending on the type of storage tank (individual or collective) . 

1= individual tank 2= collective tank.

n: number of samples; CFU: colony forming units; NFS: non-fat solids; TS: total solids.

Discussion

The average percentage of protein was 3.43 ± 0.13%, with a minimum of 3.13 and maximum of 3.92% (Table 2). This is lower than the value reported by ^{Calderón et al. (2007}), which was 3.6 ± 0.42%, and greater than 3.22 ± 0.09 and 3.28 ± 0.23% determined by Calderón et al. (2011) and Calderón et al. (2012), respectively. ^{Álvarez et al. (2012}) in Mexico found a lower value (3.29%), and ^{García et al. (2013}) in Venezuela reported a similar percentage (3.49%) under comparable management conditions. ^{Sandoval et al. (2011}) in Paraguay concluded that high SCC decreases the percentage of protein, and ^{Barbosa et al., (2013}) found that high SCC is associated with protein reduction in Gyr cattle (although, this situation did not occur in this study, since correlation was low and not significant, r: 0.06; p>0.05).

Average percentage of fat determined in this study was of 4.17 ± 0.34%, with a minimum of 2.75% and a maximum of 5.4% (Table 2). According to Decree 617 (Ministry of Social Protection, Colombia, 2006), raw milk must contain at least 3% fat. This average is higher than the 3.64 ± 0.46 and 3.70 ± 0.23% reported by ^{Calderon et al. (2011}) and Calderon et al. (2012), respectively, in milk from cows without subclinical mastitis. These values could be due to racial types (Bos indicus x Bos taurus), which were adapted to the environmental conditions of the area (^{Briñez et al., 2008}). ^{Barbosa et al. (2013}) reported a high SCC value not associated with the reduction in fat percentage of Gyr cattle in Brazil. ^{Cervantes et al. (2013}) concluded that cow nutrition has an important effect on milk composition, mainly on fat content.

The average NFS was 8.70 ± 0.14%, with a minimum value of 8.15% and a maximum of 9.01% (Table 2). This value is higher than the minimum established by law in Colombian (Ministry of Social Protection, 2006) and the reported by ^{Calderón et al. (2007}) and Calderon et al. (2011) (8.38 ± 0.48 and 8.59 ± 0.16%, respectively) for cows without subclinical mastitis. Higher values (9.56%) have been reported in Venezuela (^{Briñez et al., 2003}). In Brazil, in semi-specialized milk production systems a value of 8.58 ± 0.25% was obtained for NFS percentage (^{Rosa et al., 2012}). ^{Alvarez et al. (2012}) determined that NFS percentage does not significantly change during the year. Moreover, ^{Barbosa et al. (2013}) determined that high SCC was associated with a reduction of NFS in Gyr cattle.

The average TS percentage was 13.16 ± 0.39%, with a minimum of 11.90% and maximum of 14.55% (Table 2). This value classifies as ordinary milk (^{Calderón et al.,2006}) and is higher than the current rule accepted in Colombia (11.30%; Ministry of Social Protection, 2006), and higher than 12.06 ± 0.87% reported by Calderón et al. (2007) and 11.45 ± 0.75% found by Calderon et al. (2012). A smaller percentage of TS has been reported in Brazil (12.37%; ^{Ribas et al, 2004}.). ^{Barbosa et al. (2013}) found that SCC was associated with reduced milk protein percentage in Gyr cattle. ^{Sandoval et al. (2011}) showed that high SCC are related to reduced TS content.

Mesophilic count, SCC, and physicochemical parameters are shown in Figures 2 and 3, for each of the areas sampled. Average mesophilic count was 305,276 ± 698,206 CFU/mL, with a minimum of 8,000 and maximum of 6,796,000 CFU/mL (Table 2). This average is higher than 235,450 CFU/ mL reported by ^{Calderon et al. (2012}) from a farm equipped with a cooling tank in Monteria (Colombia). A high mesophilic count could be due to hygienic conditions of the stables and milking sites, lack of proper teat sanitation practices, inadequate cleaning routine, improper teat disinfection, poor disinfecting containers used in milking conditions, which were not evaluated in this study. In these situations, good manufacturing practices are required (^{Villoch, 2010}). ^{Ruiz et al. (2012}) determined that mesophilic count ranged between 4,500 and 2,103,500 CFU/mL in eastern Antioquia (Colombia), with an average of 115,932 CFU/mL. They proposed to improve the former variables to enhance the bacteriological quality of milk. In different Colombian regions, ^{Vásquez et al. (2012}) established that 93.9 % showed mesophilic count up to 200,000 CFU/mL and out of these, 84.49% had values under 100,000 CFU/mL, as required by the European Community and the United States. The 47.53% of samples in the above report displayed mesophilic count below 25,000 CFU/ mL, which inferred an excellent microbiological quality. ^{Cervantes et al. (2013}) in Hidalgo (Mexico) suggested that reducing the exposure time of the milk to environmental factors such as temperature, humidity, dust, and other pollutants will improve its bacteriological quality.

In agreement with the Colombian resolution # 000017 of 2012, only 68% (132/186) of milk producers in this study should receive bonus for mesophilic counts lower than 200,000 CFU/mL (Table 3). In a study regarding temperature and spread of bacteria, ^{Ruiz et al. (2012}) reported milk of excellent quality from San Pedro de los Milagros in northern Antioquia, with a mesophilic count of 4,500 CFU/mL. ^{Calderon et al. (2012}) estimated an average of 160,346 ± 213,354 CFU/mL and found differences between non-refrigerated (235,450 ± 252,559 CFU/mL) and refrigerated milk (74,514 ± 125,222 CFU/mL). The reason for these differences was the use of cooling systems, which inhibit bacterial growth, extends the storage time in the farm and reduces transportation costs (DeLaval, 2006).

On the other hand, high bacteria load in milk decreases the shelf time of the product and deteriorates its organoleptic and nutritional quality. In addition, it interferes with lactic acid fermentation, enzymatic degradation or coagulation and promotes casein proteolysis (^{Signorini et al., 2008}). ^{Silanikove et al. (2010}) admitted that high counts of aerobic mesophilic bacteria (AMB), total coliform bacteria (TCB) and psychrophilic bacteria could be a source of transmission of foodborne diseases.

Table 5 shows the distribution of physicochemical quality of mesophilic and somatic cell counts by tank type (individual or collective). There was a significant difference between individual and collective tanks. Probably, these were due to the material used to build the collective tanks, which may not be suitable for the food industry. Besides, other factors such as the container itself and prolonged milk transportation could be reasons for those differences, although they were not evaluated in this study.

The average SCC in this study was 523,207 ± 209,191 SC/mL with minimum and maximum values of 180,000 and 1,7015,000 SC/mL, respectively (Table 2). ^{Calderón et al. (2011}) obtained a SCC of 230,000 ± 5,650 SC/mL, which was assumed as low for cows without subclinical mastitis. It was determined that high SCC values greatly affect the physicochemical quality of milk and diminish the production of coastal cheese by 5.58%. Additionally, Calderon et al. (2012) reported a SCC of 345,133 ± 302,241 in northwestern Colombia. In other regions, such as northern Antioquia, SCC less than or equal to 400,000 SC/mL have been reported (^{Posada et al., 2010}). This parameter has been accepted by the dairy industry; nonetheless, it is still an evidence of subclinical mastitis, loss of quality, and low milk yield. ^{Vásquez et al. (2012}) established an average SCC of 642,000 SC/mL and 39.54% of the milk samples did not exceed international standards for SCC. Finally, those authors concluded that these numbers should be reduced in order to compete in international markets. ^{Noro et al. (2006}) obtained an average SCC of 390,000 SC/mL in Brazil. While ^{Ortiz and Vera (2006}) (Peru) concluded that the best technological levels for processed milk result in lower SCC. Moreover, ^{Revelli et al. (2011}) reported an average SCC of 407,000 ± 230,000 SC/mL. Meanwhile, ^{Sandoval et al. (2011}) in Paraguay established that 47.5% of their samples had less than 200,000 SC/mL.

Only 1.14% (2/176) of the studied farms in this work had SCC less than or equal to 200,000 SC/mL (Table 4). SCC less than or equal to 200,000 SCC can be considered as good, although with subclinical mastitis; SCC greater than 400,000 SC/mL are denotative of subclinical and clinical mastitis. Therefore, it is necessary to implement preventative measures to control bovine mastitis. ^{Sears and Mc Carnthy (2003}) concluded that SCC below 400,000 SC/mL is indicative of farms with good management practices, but they do not emphasize on mastitis control.

The physicochemical variables in raw milk presented an excellent compositional quality. However, its microbiological quality is deficient in spite of the fact that milk was kept in cold conditions. Farms should implement quality assurance programs, which allow production of excellent quality milk.

Conflicts of interest

The authors declare they have no conflicts of interest with regard to the work presented in this report.

Notes

To cite this article: Suàrez GJ, Pomarez JV, Rangel AC, Rodrìguez VR, Angulo LM, Garay OV. Raw milk quality in Northwestern Colombia. Rev Colomb Cienc Pecu