CONFORMITY OF REFRIGERATED RAW MILK FROM FAMILY PRODUCTION UNITS OF SOUTHERN ESPÍRITO

Dairy farming is an important activity in Brazil and Espírito Santo state, with small properties accounting for 80% of production. Despite of this economic importance, data about quality of the milk produced and the hygienic-sanitary conditions are still scarce. Then, the objective of this study was to evaluate the characteristics of refrigerated raw milk produced in family production units in southern Espírito Santo to verify compliance with Brazilian legal standards. Three collections were carried out in 29 community expansion tanks, for a total of 87 samples of refrigerated raw milk. Analysis of standard plate count (SPC), psychrotrophic bacteria count (PBC), somatic cell count (SCC), titratable acidity, density, fat, total dry extract (TDE), non-fat dry extract (NFDE), residues of antibiotics (β-lactam and tetracycline), cadmium and lead levels were performed. Of the 87 samples, 66% presented non-standard values for SPC, and 38% had non-standard values for SCC. Eighty five percent of the samples presented non-standard results for NFDE, 10.3% for density and 2.3% for titratable acidity. All samples complied with the legislation regarding residues of antibiotics, Cd and Pb levels. These results indicated failures in the raw milk obtainment and storage chain from family production units of southern Espírito Santo, Brazil.


Introduction
The quality of refrigerated raw milk is of great importance to the industry, as it directly affects the quality and durability of dairy products, as well as risks to consumer health (1,2) .In Brazil, in 2002, the Ministry of Livestock and Food Supply (MAPA) approved Normative Instruction (NI) number 51, with physicochemical parameters (titratable acidity, fat content, relative density and non-fat dry extract) and microbiological parameters (standard plate count and somatic cell count) for refrigerated raw milk, with deadlines for meeting the requirements based on region in the country (3) .Considering that the great majority of dairy farmers in Brazil found difficult to comply with the legislation, in 2011, the Federal Government approved the Normative Instruction number 62 that extended the deadlines for compliance for standard plate count and somatic cell count (4) .Even currently, studies have shown that producers still find difficult to meet such standards (5,6) .
Despite the quality of the milk to be of great importance, it should be noted that there are very few published studies with this type of data from the southern region of Espírito Santo, Brazil.Therefore, the objective of the work was to carry out microbiological and physicochemical analysis of the raw refrigerated milk coming from family production units in southern Espírito Santo to verify compliance with the current legal standards.

Materials and Methods
Twenty-nine community expansion tanks that collected milk from family production units in thirteen cities in the southern of Espírito Santo were selected (Alegre, Apiacá, Atílio Vivácqua, Bom Jesus do Norte, Cachoeiro de Itapemirim, Castelo, Ibitirama, Jeronimo Monteiro, Mimoso do Sul, Muqui, Presidente Kennedy, Rio novo do Sul, and São José dos Calçados).Refrigerated raw milk samples were collected in sterilized glass bottles and transported to the laboratory in isothermal boxes from March to September 2016 and submitted to microbiological analysis of standard plate count (SPC), psychrotrophic bacteria count (PBC) and somatic cell count (SCC).Physicochemical analysis of titratable acidity, fat, relative density, total dry extract (TDE) and non-fat dry extract (NFDE) were conducted.Analysis of antibiotic residues of β-lactam and tetracycline and the presence of cadmium (Cd) and lead (Pb) were also carried out.Three samples were collected in each tank, for a total of 87 samples.The collections were carried out on different days and at intervals of at least one week.
SPC and PBC were performed on 3M ™ Petrifilm ™ Plates (Sumaré, Brazil) for aerobic counts, according to the manufacturer's recommendations.One milliliter aliquot of milk at the proper dilution was transferred to the plate.The SPC plates were incubated at 35 °C for 48 hours.For PBC plates were incubated at 7 °C for 10 days.SCC were performed using the IDEXX Somaticell SCC Test Kit (São Paulo, Brazil) according to the manufacturer's recommendations.SPC and SCC results were compared to Normative Instruction 62/2011 of MAPA (4) .In Brazil, currently, there is no legislation that determines PBC in raw milk.According to literature reports, it is recommended to avoid the use of raw milk with psychrotrophic counts above 5.7 log of colony-forming units per milliliter (cfu.mL - 1 ) in the manufacture of dairy products (7) , so this value was used to compare our results.
Physicochemical analysis of titratable acidity, relative density, fat, TDE and NFDE were performed according to the methodology prescribed in Normative Instruction 68/2006 of MAPA (8) .
The presence of β-lactam and tetracycline residues in milk was verified using a test kit for antibiotics, BetaStar® Combo (Neogen, Esher, England) according to the manufacturer's recommendations.
For Cd and Pb analysis milk samples were digested in a MARS 6 (CEM Corporation, Matthews, USA) using MARSX press tubes.A total of 0.5 g of lyophilized raw milk sample was added, and 10 mL of 65% PA nitric acid (Vetec, Rio de Janeiro, Brazil) was added according to the methodology described by Klimek et al. (9) .The digestion was performed between 180 °C and 210 °C, with a ramp time of 20 to 25 minutes, and a hold time of 15 minutes, as recommended by the manufacturer for organic samples.After digestion, the samples were diluted to 25 mL with milli-Q water.Calibration curves were prepared from standard solutions of 1000 mg L -1 of Cd and Pb (Merck, Darmstadt, Germany), with a maximum concentration of 1.0 mg mL -1 .The samples were analysed in a flame atomic absorption spectrophotometer (Varian, São Paulo, Brazil), model Spectra 220 FS.Data were tabulated, and descriptive statistics are presented in table form.The results of the analysis of raw milk were compared with the current legislation: Normative Instruction 62/2011 and Decree number 9.013/2017 of MAPA (4,10) and, 55.871/1965 and Ordinance number 685/1998 of Ministry of Health (11,12) .

Results and Discussion
SPC of raw milk is related to hygiene and sanitation practices in milking, including the hygiene of equipment and utensils, cleanliness of milking equipment, water quality used in hygiene processes and quality of the storage conditions (2) , assiduity of milker, poor hygiene, absence of ceilings, inadequate milk cooling, storage for more than 48 hours, and sanity of the herd (7,13) .SPC of refrigerated raw milk from the expansion tanks ranged from less than 4.0 to 7.7 log of cfu mL -1 (Table 1).NI 62/2011 of MAPA sets a maximum SPC of 3.0x10 5 cfu mL -1 or 5.5 log of cfu mL - 1 for the south-eastern region of Brazil (4) .Of the 87 samples evaluated, 66% (57) had SPC above the legal standard (Table 1).A similar result was obtained in four community tanks in the city of Alegre-ES, Brazil by Nascimento Neta et al. (14) wherein 50% (02) of the tanks had SPC above legislation recommendations.Different results were found by Almeida et.al. (15) who obtained non-standard SPC in all of 326 raw milk samples from northern of Minas Gerais, Brazil.The authors associated this result with possible management failures, such as unsatisfactory sanitary practices in the hygiene of utensils and water quality.
Pinto et al. (7) analysed raw refrigerated milk of 33 individual expansion tanks, 12 communities tanks, and one industrial silo in Zona da Mata Mineira.They found that SPC in the industrial silo was higher than allowed by legislation.However, the same was not true for samples from individual expansion tanks.The authors attributed this difference to possible additional contamination and bacterial multiplication during transportation and storage in the industrial silo.
According to Marcondes et al. (16) , the lower the production of the herd, the greater is the SPC.As most of the producers participating in this research were small and with low productivity animals, it was expected that a great part of the samples presented high values for SPC.The increase in SPC occurs due to faults in hygienic-sanitary conditions during milking (17) .These failures are probably due to the lower incidence of trained workers in small farms (18) .
The presence of psychrotrophic microorganisms in refrigerated raw milk may cause problems in reducing the durability of dairy products.This occurs because psychrotrophic microorganisms cause deterioration, and some producing enzymes (proteolytic and lipolytic) also cause degradation of dairy products.Many of these are heat resistant and cause changes in the products during their storage (7,19) .
Considering previous research (7) and that raw milk can be storage at the dairy industry for some hours before being processed, the value of 5.7 log of cfu mL -1 was chosen to be the acceptable maximum value for PBC in this study.The PBC of the raw milk from the expansion tanks ranged from less than 3.0 to 7.1 log of cfu mL -1 .Of the 87 samples evaluated, 16% (14) presented PBC above 5.7 log of cfu mL -1 (Table 1).
Fox (20) reports that problems with yield and off-flavour in cheeses due to the activity of proteolytic psychrotrophic appear when counts are above 6.0 log of cfu mL -1 .Muir et al. (21) concluded that in order to not affect the quality of dairy products, the psychrotrophic count in milk should be less than BARROS, L.R. et al.
or equal to 6.3 log of cfu mL -1 .
Nascimento Neta et al. (14) found that 25% (01) of the four community expansion tanks evaluated in the city of Alegre-ES, presented psychrotrophic bacterial count above 5.7 log of cfu mL -1 .
Ângelo et al. (22) studied four individual and four collective tanks in São João Nepomuceno-MG, Brazil.Two of the community tanks presented a psychrotrophic bacteria count above 5.0 log of cfu mL -1 , value that was considered high by the authors.
Low counts of psychrotrophic bacteria, as observed in the most samples of our study, can be explained by factors such as low initial contamination of raw milk or the collection of samples when the milk had little storage time under refrigeration, leading to insufficient time for the multiplication of psychrotrophic microorganisms.As in the present study, high SPC were observed we believed that the time that milk remained under refrigeration was not sufficient for the multiplication of psychrotrophic microorganisms since the bacterial initial populations was generally high.
High SCC in milk negatively influences its quality, altering its constituents and reducing the durability of dairy products made with that milk.All this generates losses for the dairy industry.Increasing in SCC indicate inflammation in the mammary gland that leads to reduced volume produced, raise proteolytic and lipolytic activity of milk, changes in membrane permeability with consequent loss of components for bloodstream.This results in changes in the milk composition, especially in lactose, casein and mineral salts affecting the yield of dairy products.In addition, it has a public health concern, once mastitis can be caused by pathogenic microorganisms as staphylococci and streptococci (20,23) .
All this causes damage to the dairy industry, since they occur in changes in the chemistry of milk, especially in lactose and mineral salts, in the yield of dairy products.In addition, it has a public health problem, since mastitis can be published by pathogenic microorganisms, for example, of the genus Staphylococcus.
The samples of refrigerated raw milk from expansion tanks presented SCC ranging from 5.0 to 6.3 log of somatic cells mL -1 (Table 1).The NI 62/2011 sets a maximum value of 5.0x10 5 somatic cells per milliliter or 5.7 log of somatic cells mL -1 for the south-eastern region of Brazil (4) .Of the 87 refrigerated raw milk samples, 38% (33) had SCC (Table 1) above the maximum allowed by the legislation (4) .
Results founded by others researchers were better then ours.Angelis et al. (24) founded that only 11% (02) of the 18 raw milk samples obtained from manual or mechanical milking in Argirita-MG, Brazil were outside of the standard.Alves et al. (25) collected 32 samples of milk from refrigerated tanks in Colorado do Oeste-RO, Brazil between 2010 and 2013, and verified the counts within the standard set by the current legislation for all samples.
The somatic cell count is related to mastitis, which is an inflammation of the mammary glands often caused by pathogenic and/or deteriorating bacteria.Blood leukocytes spread to the mammary glands to fight against the infectious agent, causing somatic cell elevation (26,27) .
The incidence of mastitis in lactating animals is influenced by several factors, such as climate (temperature and humidity), genetics, age, lactation period, animal handling and nutrition (28) .Climatic factors, such as high temperatures and high humidity (rainy season), contribute to the increased Conformity of refrigerated raw milk from family production units of southern Espírito Santo incidence of mastitis and consequently to increasing the SCC in milk (29,30) .
Considering that the sampling period was almost completely during the dry season (low humidity), a lower incidence of mastitis was expected in the herds.Moreover, the years of 2015 and 2016 were marked by a severe drought in the state of Espírito Santo.
In instances of mastitis, diseased animals may be in the process of being diagnosed or treated, and the milk produced from that animal must be discarded.Cows with higher milk production tend to have higher SCC values.Besides, some factors may increase predisposition to mastitis, such as stress due to the retention time of the animals, the time interval between the disinfection of the teats and the beginning of milking and maintenance of calves at foot of the cow (31) .Titratable acidity is a rapid and quantitative analysis used to evaluate the quality of raw milk (33) .Bacteria present in milk ferment lactose producing lactic acid, which increases the titratable acidity (34) .Therefore, this test is widely used to check the quality of raw milk.In the analysis of titratable acidity, only 2.3% (02) of the samples fell outside the range recommended by the legislation (i.e., 0.14-0.18g of lactic acid 100 mL -1 ) (Table 2).
For relative density at 15°C all values fell between 1.026 and 1.032 g mL -1 (Table 2).Approximately 10% (09) of the samples were below the range stipulated by the legislation (i.e., 1.028 -1.034 g mL -1 ) (4) .The density is strongly affected by the NFDE.The relationship is directly proportional; that is, a reduction in the NFDE causes a reduction in the relative density.As in our work was observed low level of NFDE for 85% of samples this factor could explain this finding.Another factor that affects the relative density is the fat content; but they are inversely related, when the fat content increases, the density decreases (34,38,40) .In the state of Paraná, Brazil 5.4% (04) of the raw milk samples tested were outside of the standard for relative density (39) .According to the authors, samples with low relative density occurred because of fraud by the water addition.However, it is known that other factors can cause variations in milk density.In addition, the values above the range occurred because the samples had a fat content below 3.0%, which is the minimum legally allowed (39) .
Milk composition as well as its fat content can be influenced by several factors, such as nutrition, race, lactation stage, season and animal health, but the most influential factor is nutrition (32) .
In the analysis of the fat content of refrigerated raw milk, the values ranged from 3.0 to 4.7% (Table 2).All 87 samples presented values within the standards set by the legislation (i.e., ≥ 3.0%) (10) .
Total dry extract (TDE) corresponds to all components of milk except water.Among them, the most relevant are fat and protein, which directly affects the yield of dairy products and are therefore of great interest to the industry (35) .TDE of raw milk ranged from 10.9 to 13.8%.The NI 62/2011 does not stipulate values for TDE in refrigerated raw milk, but Decree 9013 article 248 of March 29, 2017, states that raw milk must contain at least 11.4% TDE (10) .Seven percent (06) of the samples had TDE values below 11.4% (Table 2).In analysis of twenty samples of refrigerated raw milk from bulk tank trucks in the southern of Rio de Janeiro, Brazil, Paula et al. (36) did not observe any TDE values below 11.4%.Variations in TDE may occur due to herd genetics and health, lactation stage, animal age, diet and season.Variations in milk composition are common among different breeds.Among the components, fat usually suffers the most variation.In the first three months of lactation, the levels of fat, protein and lactose are lower, after that period levels increase gradually.The older the animal, the lower the productivity of the mammary glands and the greater the chances of inflammation.The diet given to the animal influences the composition of the milk, since the ingested nutrients will be absorbed and used in the formation of milk components.High temperatures cause a decrease in the fat content of the milk; in addition, stress due to high temperatures causes the reduction of dry matter intake by the animals, affecting milk composition (37,38) .
The NFDE is composed mainly of lactose and protein and therefore also affects the yield of dairy products.In addition to acidity and density, NFDE is also used to detect milk fraud.Eighty five percent (74) of the samples of refrigerated raw milk had levels below those recommended by the legislation (i.e., ≥ 8.4%) (Table 2).The values ranged from 7.3 to 9.2%.Molina et al. (41) analysed samples from 21 producers of raw milk from Itaqui-RS, Brazil.They found that 62% (13) of the samples were outside the legal standard.The authors stated that variations in NFDE may be due to variations in animal feeding or variations caused by mastitis.
In this study the NFDE was the parameter with the highest percentage of samples that were outside of the limits set by the legislation.The breed and age of the animal, stage of lactation, feeding and sanity of the herd can alter the levels of NFDE in milk (42) .Considering that in 2015 and 2016, the State of Espírito Santo faced the worst drought in the last 40 years (43) we believed that the results for NFDE have been influenced by the water crisis.Due to the drought, some farmers gave up dairy farming, and many found difficult to feed their herds, causing great variations in the feeding of the animals and altering the composition of the milk and consequently its quality.
The use of antibiotics in lactating animals is common for the treatment of mastitis, however, the period required to prevent residues of the drug in milk should be respected (44) .None of the samples tested were positive for β-lactam or tetracycline antibiotics.This result shows that milk producers probably did not add milk from cows being treated with this type of drug to the expansion tank or that the levels of these antibiotics in the milk were below the detection limit of the test used.Nascimento Neta et al. (14) carried out two collections in four community expansion tanks in the city of Alegre-ES, Brazil and did not detect residues of antibiotics in refrigerated raw milk.Different results were found by Souza et al. (45) , in which 6.7% of 112 samples of raw milk from six cities in the state of Rio Grande do Norte, Brazil presented positive results for residues of several antimicrobials.The presence of antibiotics in milk not only causes damage to the industry of fermented dairy products, but also is dangerous to consumer health, due to the selection of resistant bacterial strains,