Ibarra, Ecuador
Ibarra, Ecuador
Ibarra, Ecuador
Ibarra, Ecuador
Ibarra, Ecuador
Ibarra, Ecuador
Pontificia Universidad Católica del Ecuador Sede Ibarra
Ibarra, Ecuador
Quito, Ecuador
Ibarra, Ecuador
Introduction. This study aimed to evaluate the quality of milk produced by six cattlemen’s associations in small, isolated farming communities of Carchi, Ecuador. It involved a herd of 814 cows and lasted eight consecutive months. Another aim was to propose a suitable methodology for milk quality evaluation. Study objects and methods. All milk samples were analyzed for total solids, protein, fat, acidity, density, total bacterial count (TBC) and somatic cell count (SCC). Each sample was subjected to an extended qualitative methylene blue reduction test (MBRTe) for which 10 mL of milk, with 0.5 mL of methylene blue, was incubated at 37°C for 24 h. Results and discussion. As a result, we obtained the following types of clots: MBRTe-I (homogeneous solid/liquid clot), MBRTe-II (lumpy clot), MBRTe-III (gaseous clot) and MBRTe-IV (lumpy + gaseous clot). The study showed significant differences in the quality of milk between different associations, suggesting that some of them did not comply with good practices of milking, handling and storage of fresh milk. The quality of milk was classified as good in one association, as regular in another association, and as low in four associations. The MBRTe classified 37% of the samples as MBRTe-I, 18% as MBRTe-II, 14% as MBRTe-III and 12% as MBRTe-IV. Of the MBRTe-I samples, 95% showed the TBC and SCC values of first quality milk. The MBRTe-II had the TBC values of first quality milk, but exceeded the SCC, while the MBRTe-III had good SCC values, but exceeded the TBC. Finally, the MBRTe-IV samples exceeded the permissible levels of both TBC and SCC. Conclusion. It was proved that the MBRTe can help milk producers evaluate the quality of milk and alert them to the possible presence of mastitis in the herd. The MBRTe is a reliable and cheap method that is quick and easy to perform.
Dairy industry, raw milk, dairy cattle, microorganisms, somatic cells
INTRODUCTION
Ecuador produces between 5.5 to 5.8 million liters
of milk on a daily basis and this production has been
steadily growing in recent years [1]. About 75% of milk
is produced in the Andean region, mostly by small
associations of farmers, far from large urban centers [2].
The quality of milk determines the quality of dairy
products. It refers to the content of microorganisms
(pathogenic or not) and somatic cells, as well as the
presence of antibiotics and medicines [3]. Milk quality
is guaranteed by the health of the herd, as well as good
management and milking practices (GMMP). To check
the microbiological quality of raw milk, dairy producers
commonly use the counts of total bacteria (TBC),
somatic cell counts (SCC), and the methylene blue
reduction test (MBRT) [4].
The presence of somatic cells in milk has been
mainly related to the increase of white cells (leukocytes)
as a result of an immune system’s response to mastitis.
It is a livestock disease caused by the inflammation
of the udder due to the action of pathogenic microorganisms
such as Staphilococcus aureaus, Streptococcus
141
Pérez-Lomas M. et al. Foods and Raw Materials, 2020, vol. 8, no. 1, pp. 140–148
dysgalactiae and Streptococcus agalactiae. Mastitis
alters the chemical composition of milk and decreases its
yield [5, 6].
However, the TBC and SCC are usually carried out
in accredited laboratories located in urban areas, far
from small farmers and their associations. They are
expensive for milk producers and, moreover, the latter
have to wait quite long for the results before they can
improve the microbiological quality of their milk.
The MBRT, on the other hand, is an old but effective
method which has been correlated, with some success,
with the total microbial load and, therefore, with the
microbiological quality of milk [7–9]I. It is a simple and
fast method, although sometimes it lacks the expected
precision.
The goals of this work were to evaluate the quality of
milk produced by a group of cattlemen’s associations in
the province of Carchi, Ecuador, and to suggest a cheap,
fast, and reliable alternative methodology that would
allow the associations to evaluate the quality of their
milk in situ.
STUDY OBJECTS AND METHODS
Herd size and geographic location. The study
involved six cattlemen’s associations (A–F) located in
the Andean province of Carchi in Ecuador. There were
11 small farmers in Association A, 27 in Association
B, 16 in Association C, 20 in Association D, 15 in
Association E and 19 in Association F. Their milk was
sampled for eight months, from October 2016 to May
2017. As a result, 709 samples were taken from a herd
of 814 milking cows (34 from Association A, 235 from
Association B, 50 from Association C, 120 from
Association D, 230 from Association E, and 145 from
I ISO 4833-2:2013. Microbiology of the food chain – Horizontal
method for the enumeration of microorganisms – Part 2: Colony
count at 30°C by the surface plating technique. Geneva: International
Organization for Standarization; 2013.
Association F). All the samples were analyzed for total
solids, total protein, fat, acidity and density. The total
bacteria and somatic cells were also counted.
Physicochemical and microbiological properties
of the samples. The determinations of total solids, total
protein, fat, acidity and density, as well as somatic cells
(SC) and total bacteria counts (TBC) were performed
in an accredited laboratory of the Phyto- and Zoo-
Sanitary Regulation and Control Agency of Ecuador
(AgroCalidad) (www.agrocalidad.gob.ec) located in
Tumbaco (Quito, Pichincha, Ecuador) [10]II,III,IV,V. The
somatic cell count (SCC, SC/mL) was performed in a
Foosmatic™7 (Foss, Hilleroed DK-3400, Denmark)
according to the standard procedureVI. The total
bacterial count (TBC, CFU/mL) was performed in a
BactoScan™FC+ (Foss, Hilleroed DK-3400, Denmark),
obtaining values equivalent to those that would be
obtained from a standard plate count (SPC)VII.
Standard and extended methylene blue reduction
test. The standard methylene blue reduction test
(MBRT) and the 24 h extended methylene blue reduction
II ISO 6731:2010 [IDF 21:2010]. Milk, cream and evaporated milk:
determination of total solids content (reference method). Geneva:
International Organization for Standarization; 2010. 5 p.
III ISO 8968-1:2014 [IDF 20-1:2014]. Milk and milk products –
Determination of nitrogen content – Part 1: Kjeldahl principle and
crude protein calculation. Geneva: International Organization for
Standarization; 2014. 18 p.
IV ISO 1211:2010 [IDF 1:2010]. Milk – Determination of fat content
– Gravimetric method (Reference method). Geneva: International
Organization for Standarization; 2010. 18 p.
V ISO/TS 11869:2012 [IDF/RM 150:2012]. Fermented milks –
Determination of titratable acidity – Potentiometric method. Geneva:
International Organization for Standarization; 2012. 7 p.
VI ISO 13366-2:2006 [IDF 148-2:2006]. Milk – Enumeration
of somatic cells – Part 2: Guidance on the operation of fluoroopto-
electronic counters. Geneva: International Organization for
Standarization; 2006. 13 p.
VII ISO 4833-1:2013. Microbiology of the food chain – Horizontal
method for the enumeration of the microorganisms – Part 1: Colony
count at 30°C by the pour plate technique. Geneva: International
Organization for Standarization; 2013. 9 p.
Table 1 Types, characteristics, and possible causes of clots obtained from the extended methylene blue reduction test (MBRTe)
Type Classification Characteristics Causes Image
MBRTe-I Solid or liquid
homogenous clot*
Homogeneous clot, with an acidic
odor and taste, without cracks or
fissures, white in color, without
or with few bubbles
Presence of Lactobacillus spp.
or antibiotics in milk
Figure 1a
MBRTe-II Clumped
heterogeneous clot
Heterogeneous clot with lumps,
with a whitish, yellowish serum,
or other abnormal colors
Produced by germs with bitter tastes and
unpleasant odors; mastitic milk at the end
of lactation; or milk cooled for a long time
Figure 1b
MBRTe-III Gaseous
heterogeneous clot
Heterogeneous clot with bluish
shades and numerous bubbles
and gaseous grooves
Coliform bacteria; milk obtained and preserved
in poor sanitary conditions or refrigerated for
a long time
Figure 1c
MBRTe-IV Clumped + gaseous
heterogeneous clot
Heterogeneous clot with lumps
and numerous bubbles or gaseous
furrows
Combined action of coliform bacteria and somatic
cells; milking and conservation of milk
without complying with the GMMP; mastic
milk or milk refrigerated for a long time
Figure 1d
* If it is liquid, check the presence of antibiotics or substances that can inhibit microbial growth (such as detergents, pesticides, etc.)
142
Pérez-Lomas M. et al. Foods and Raw Materials, 2020, vol. 8, no. 1, pp. 140–148
test (MBRTe) were carried out in the associations’ own
laboratories. For the quantitative MBRT, 10 mL of a
sample was incubated at 37°C with 0.5 mL of methylene
blue and the time (in hours) for the blue coloration to
disappear was measured according to the technique
described in ISO 4833-2:2013I. We used the methylene
blue reagent produced by Merck KGaA (Darmstadt,
Germany). The samples were incubated at 37°C in a
conventional water bath (Thermo Scientific™ TSGP10,
Waltham, Massachusetts, USA).
The procedure of the qualitative MBRTe was similar
to that of the MBRT, but the samples were incubated for
24 h. As a result, we obtained clots of the following four
types (Table 1, Fig. 1): a homogeneous solid or liquid clot
(MBRTe-I); a heterogeneous lumpy clot (MBRTe-II);
a heterogeneous gaseous clot (MBRTe-III), and a heterogeneous
lumpy + gaseous clot (MBRTe-IV).
The MBRT is, therefore, a quantitative test
(measured in hours), while the MBRTe is a qualitative
test (one of the four possible sample types after 24 h
incubation with methylene blue).
Statistical Analysis. The statistical analysis was
applied using the free statistical package R version 3.6.1
(2019-07-05).
RESULTS AND DISCUSSION
We analyzed 709 samples for eight continuous
variables (SCC, TBC, MBRT, fat, protein, total solids,
density and acidity) and three categorical variables:
(1) eight dates (Oct-16, Nov-16, Dec-16, Jan-17,
Feb-17, Mar-17, Apr-17, and May-17);
(2) six associations (A, B, C, D, E, and F); and
(3) four MBRTe clots (MBRTe-I, MBRTe-II, MBRTe-III,
and MBRTe-IV).
The Lilliefors test (a normality test based on the
Kolmogorov-Smirnov test) [11–13] was used to explore
the continuous variables, and none of them showed a
normal distribution of the samples (P < 0.05).
Recently, a similar finding has been reported in a
study conducted to determine the quality of milk (total
bacterial and somatic cell counts) among small livestock
producers where the values obtained did not follow a
normal distribution [14]. This is probably due to the nonhomogeneity
of the samples, the differences between
the producers with respect to compliance with good
practices, as well as uncontrolled factors that fall outside
the framework of the studies.
The Kruskal-Wallis rank-sum test was performed
to establish the influence of categorical variables over
continuous variables [15, 16]. A pairwise comparison
with the Wilcoxon nonparametric rank-sum test was
used to determine which of the associations or MBRTe
types differed from each other (P < 0.05) for each
specific continuous variable [17]. Different letters
near each of the magnitude values showed significant
differences (P < 0.05).
In this study, the values of acidity, total solids,
MBRT, SCC and TBC differed significantly (P < 0.05)
between the associations, while density, protein and fat
concentrations were not different (P > 0.05) (Fig. 2).
As we can see in Fig. 2, only Association A, which
fully implemented the GMMP, showed a better quality
of fresh milk during the whole period. Association B,
which began to implement the GMMP during the study
period, achieved high quality in the final months of the
study. Associations C, E and F are still in the process
of organizing their quality assurance system, and their
results oscillate between regular and low quality. Finally,
Association D always had contamination problems and
showed poor quality milk, so all of their work protocols
need revising.
The values of SCC, TBC, MBRT, fat, total solids,
and acidity were significantly influenced by the type of
MBRTe (P < 0.05), whereas there were no differences
(P > 0.05) for the protein content and density (Fig. 3).
The significant difference (P < 0.05) observed in
the fat content between the MBRTe-II and MBRTe-III
samples (Fig. 3c) could be due to a high concentration
of somatic cells and a low concentration of total
bacteria in the MBRTe-II sample group. In fact, the
method of fat determination presupposes the addition
of sulfuric acid which causes the breakdown of somatic
cells incorporated into milk fat. It was also possible
that exogenous bacteria species that contaminated
milk, which were present in the MBRTe-III samples,
exerted a greater lipolytic effect on the fat and lowered
its concentration in milk, compared to the rest of the
MBRTe samples.
In Fig. 3c, we can observe an increase in acidity and
a decrease in total solids when moving from MBRTe-I
(а) (b) (c) (d)
Figure 1 Four types of clots. (а) MBRTe-I (homogeneous
solid/liquid clot), (b) MBRTe-II (lumpy clot), (c) MBRTe-III
(gaseous clot), and (d) MBRTe-IV (lumpy + gaseous clot)
143
Pérez-Lomas M. et al. Foods and Raw Materials, 2020, vol. 8, no. 1, pp. 140–148
to MBRTe-IV. This trend seems to be associated with
a combined increase in the microbial load and somatic
cells in these groups, enhancing the presence of organic
acids and therefore leading to higher acidity, and a
decrease in carbon sources, such as lactose, leading to
lower total solids.
When comparing the magnitudes of the qualitative
MBRTe and the quantitative MBRT with the SCC and
TBC values in a Kruskal-Wallis rank-sum test, we can
see that unlike the MBRT, which is only significantly
influenced (P < 0.05) by TBC, but not SCC (P > 0.05),
the qualitative variable of MBRTe correlates significantly
(P < 0.05) with both the TBC and SCC values.
In the proposed MBRTe test, some of the samples
incubated with and without the presence of methylene
blue had a similar behavior and formed the same type of
clot after 24 h. This finding suggests that the presence
of the methylene blue dye does not play the same role as
it does in the MBRT test. However, there is a need for
more detailed experiments to corroborate the influence
or necessity of this dye in the MBRTe test. They need
to use the same samples, incubate them under the same
conditions for 24 h and then observe the type of clot
forming after that time.
(c) (d)
(а) (b)
Figure 2 Average values of (a) acidity (°D), (b) total solids (g/100 mL), (c) MBRT (h), (d) SCC (SC/mL), and (e) TBC (CFU/mL)
for each association during eight months of the study. The dashed red lines represent the values that delimit the thresholds of good
quality, regular quality and poor quality of milk or the minimum acceptable values by the Ecuadorian standards. Different letters
mean statistically significant differences (P < 0.05)
(e)
×
×
144
Pérez-Lomas M. et al. Foods and Raw Materials, 2020, vol. 8, no. 1, pp. 140–148
When correlating the MBRT values with the TBC, or
vice versa, we can observe similar correlations to those
previously reported by other authors [8, 9], although
with somewhat lower correlation coefficients R² (Fig. 4).
Thus, the qualitative MBRTe not only would allow
us to assess the microbiological quality of milk samples
through TBC values, but it could also detect a healthy
dairy herd (< 310 000 SC/mL) or the presence of mastitis
in its preclinical (310 000 ≤ SCC ≤ 700 000 SC/mL) or
clinical (> 700 000 SC/mL) stages, which is impossible
to do with the standard MBRT test.
Fig. 5 shows the distribution of the MBRTe samples
Figure 3 Relationship between average values of (a) TBC + SCC, (b) MBRT, (c) Fat (g/100 mL), and (d) Total solids (g/100 mL) +
Acidity (°D) and the MBRTe-types. Equal letters mean no significant differences (P < 0.05) according to the Wilcoxon
nonparametric rank-sum test [17]
(c) (d)
Figure 4 Correlation between (a) MBRT vs log10 (TBC) and (b) log10 (TBC) vs MBRT. The dashed red lines represent the values
that delimit milk quality thresholds
(а) (b)
(а) (b)
SCC
log10, TBC
145
Pérez-Lomas M. et al. Foods and Raw Materials, 2020, vol. 8, no. 1, pp. 140–148
Figure 5 Distribution of MBRTe samples in each association (A‒F) showing the herd health and compliance with good practices
(A) (B) (C)
(D) (E) (F)
Figure 6 Correlation between the MBRTe and TBC + SCC (CFU or SC/mL, respectively). (a) MBRTe-I: n = 265 (37% of the total);
(b) MBRTe-II: n = 130 (18% of the total); (c) MBRTe-III: n = 99 (14% of the total); (d) MBRTe-IV: n = 84 (12% of the total)
(а) (b)
(c) (d)
×
× ×
× ×
×
×
×
146
Pérez-Lomas M. et al. Foods and Raw Materials, 2020, vol. 8, no. 1, pp. 140–148
in each association. As we can see, the associations with
the highest proportions of MBRTe-I, compared to the
other types, had the best quality milk.
As we can see in Fig. 5, Association A provided
better milk quality than the rest of the associations. The
lowest-quality milk was produced by Association D.
This means that the health of its herd and the procedures
for milking, handling, and storage of fresh milk should
be reviewed.
When correlating the TBC and SCC values with the
qualitative MBRTe, we observed that the MBRTe could
adequately predict not only the samples with a high
bacterial load, but also those with a significant presence
of somatic cells. The latter might indicate preclinical or
clinical mastitis in the herd (Fig. 6).
As we can see in Fig. 6a, more than 94% of the
MBRTe-I samples had SC values below 310 000 SC/mL
and TBC values below 300 000 CFU/mL. Fig. 6b shows
that more than 94% of the MBRTe-II samples had TBC
values below 300 000 CFU/mL, while 57% of them
had SC values between 310 000 and 700 000 SC/mL,
which could indicate a preclinical condition of mastitis.
Moreover, 42% of the MBRTe-II samples had SC values
of over 700 000 SC/mL, which suggests the presence
of mastitis in at least part of the dairy herd. Of the
MBRTe-III samples, 97% had SC values below
310 000 SC/mL, which indicates a healthy dairy herd,
without mastitis problems.
However, as we can see in Fig. 6c, 32% of those
samples showed moderate values of microbial contamination
(between 300 000 and 600 000 CFU/mL)
and 67% of them had high values (> 600 000 CFU/mL).
These data suggest that the samples came from a
healthy dairy herd, but the GMMP were not followed
properly. Finally, all the MBRTe-IV samples (Fig. 6d)
showed moderate to high values of both TBC and SCC,
suggesting a dairy herd with mastitis problems and bad
management and milking practices. Such a product
cannot be recommended for direct consumption – it
has to be carefully pasteurized before being used in the
manufacture of dairy products.
Likewise, we analyzed a possible relationship between
the physicochemical and hygienic-sanitary properties
of the samples and the month in which these samples
were taken (from October 2016 to May 2017). For this,
a Kruskal-Wallis rank sum test was applied to each of
the measurements made to each sample and the month
of sampling. We found that the TBC, MBRT, protein
and fat contents did not depend on the months in which
the samples were taken. However, the determinations
of density, total solids and SCC, in at least a couple of
months, were influenced by the month of sampling. To
determine the significance (P < 0.05) of these differences,
we performed a pairwise comparison using the Wilcoxon
nonparametric rank-sum test and the Bonferroni
method (Fig. 7), as we did with the previous categorical
variables (type of association and MBRTe) [17].
The differences associated with the month in which
the samples were analyzed could be explained by some
uncontrolled factors in the experiments. These include
variations in the periods of rain, which could influence
the type and abundance of the grass consumed by the
dairy herd, and changes in the management of the herd,
as well as milking and storage of fresh milk. Also,
Figure 7 Dependence of average values of (a) SCC (SC/mL), (b) density (g/mL), (c) total solids (g/100 mL), and (d) acidity (°D) on
the month of sampling. Equal letters mean no significant justified (P < 0.05)
(а) (b)
(c) (d)
×
147
Pérez-Lomas M. et al. Foods and Raw Materials, 2020, vol. 8, no. 1, pp. 140–148
possible measures taken by the associations to deal
with mastitis problems may have been reflected in the
SC values, as well as the time they were taken.
To sum up, we can say that the MBRTe correctly prequalified
fresh milk and, therefore, allowed us to suggest
possible industrial uses for it and set fair market prices
(Table 2).
We all know of difficulties that small cattlemen’s
associations have with assessing the microbiological
quality of milk and detecting sub-clinical mastitis in
real time to continuously improve the quality of milk
delivered to the industry and consumers. Accredited
laboratories that perform somatic cells and total
bacterial counts, as well as the methylene blue reduction
test (MBRT), are located in provincial cities or capitals,
far away from the rural areas where most of the small
farmers’ associations are, at least in Ecuador [16]VIII.
This means that the farmers’ associations usually have
to wait a few days (an average of 3 days) for the test
results. Thus, they cannot quickly identify individual
producers that affect the milk quality of the whole
association to take prompt corrective measures.
Moreover, the cost of such analysis in Ecuador,
including transportation (for a distance of ~ 50 km), is
approximately $9.56 per sample. In contrast to that,
the qualitative MBRTe takes only one day and costs
approximately $0.46 per sample. In addition, it is easy
to perform and its interpretation is straightforward and
simple: fresh milk is pre-qualified as good (MBRTe-I),
intermediate to good (MBRTe-II), poor to intermediate
(MBRTe-III), and poor (MBRTe-IV).
The qualitative MBRTe would allow us not only
to know if the association follows good practices of
milking, handling and storage of milk, but also to
examine the health of the dairy herd, as far as mastitis
is concerned. In addition, it is a cheap test since it
requires only a conventional thermostatic bath, the blue
methylene reagent, and a set of common glass tubes. It
is significantly cheaper than modern equipment for the
detection and counting of somatic cells.
The above makes the MBRTe suitable for small
associations of livestock farmers that are isolated from
cities and towns where accredited laboratories are
generally located.
VIII NTE INEN 9:2012. Norma Técnica Ecuatoriana. Servicio
Ecuatoriano de normalización. Leche cruda. Requisitos [Ecuadorian
Technical Standard. Ecuadorian Normalization Service. Raw milk.
Requirements]. Quito: INEN; 2012. 7 p.
Additionally, the MBRTe can be applied not only
to raw fresh milk collected from all the farmers in the
association, but also from individual farmers who are its
members. This last feature could help identify individual
cattlemen who own dairy herds with preclinical or
clinical mastitis or those who do not comply with good
practices of milking and handling of fresh milk. By
doing so, the association can make a corrective plan to
improve the microbiological quality of raw fresh milk
in the near future and establish better market prices for
its producers.
CONCLUSION
In this work, we evaluated the quality of milk
produced by six dairy associations of small farmers in
the province of Carchi for eight consecutive months.
We determined the hygienic and sanitary status of
milk and dairy herd, respectively. The study found an
adequate correlation between the quality of milk and
the farmer’s compliance with good practices of milking,
handling and storage of fresh milk. Thus, it served to
encourage some of the associations to comply with these
good practices.
We demonstrated a relationship between the
qualitative MBRTe and somatic cells and total bacteria
counts. As a result, we proposed the MBRTe to the
cattlemen’s associations in the Ecuadorian highlands
to pre-qualify milk collected from both the entire
association and individual farmers. Also, the proposed
methodology can be useful for isolated ranchers, away
from accredited labs, to check the quality of their milk
by themselves. This test can identify the presence
of sub-clinical or clinical mastitis and inadequate
management of milking, handling, storage and
transportation of fresh milk. The results can be used to
make appropriate improvement plans to correct these
deficiencies and enhance the quality of milk.
CONTRIBUTION
Mayra Pérez-Lomas collected the data. Milton
Cuaran-Guerrero, Lucía Yépez-Vásquez, Holger
Pineda-Flores, Jimmy Núñez-Pérez and Rosario Espin-
Valladares contributed the data and analysis tools. José
Pais-Chanfrau conceived and designed the analysis.
Edmundo Recalde-Posso performed the analysis. Luis E.
Trujillo-Toledo wrote the paper in cooperation with José
M. Pais-Chanfrau.
CONFLICT OF INTEREST
The authors declare that there were no conflicts of
interest during the elaboration of this work or later,
during the preparation of the manuscript.
ACKNOWLEDGEMENTS
We would like to thank the engineers Luis Aldean
and Vanessa Bastidas from the Alpina S.A. Foundation
(www.alpina.com.co) for the support given to this
research.
1. Juan Pablo Grijalva: Exportar leche, una buena solución [Juan Pablo Grijalva: Export milk, a good solution] [Internet]. [cited 2019 Oct 01]. Available from: https://www.revistalideres.ec/lideres/exportacion-leche-ganaderos-lacteosconsumo.html.
2. Un tercio de la producción láctea se dedica al queso [A third of dairy production is dedicated to cheese] [Internet]. [cited 2019 Oct 01]. Available from: https://www.revistalideres.ec/lideres/ecuador-produccion-lactea-queso.html.
3. Esguerra JC, Cassoli LD, Munera-Bedoya OD, Ceron-Munoz MF, Machado PF. Milk quality: milking personnel associated factors. Revista Mvz Cordoba. 2018;23(1):6461-6473. DOI: https://doi.org/10.21897/rmvz.1241.
4. Rodrigues MX, Lima SF, Canniatti-Brazaca SG, Bicalho RC. The microbiome of bulk tank milk: Characterization and associations with somatic cell count and bacterial count. Journal of Dairy Science. 2017;100(4):2536-2552. DOI: https://doi.org/10.3168/jds.2016-11540.
5. Jamali H, Barkema HW, Jacques M, Lavallee-Bourget EM, Malouin F, Saini V, et al. Invited review: Incidence, risk factors, and effects of clinical mastitis recurrence in dairy cows. Journal of Dairy Science. 2018;101(6):4729-4746. DOI: https://doi.org/10.3168/jds.2017-13730.
6. Dalen G, Rachah A, Norstebo H, Schukken YH, Reksen O. The detection of intramammary infections using online somatic cell counts. Journal of Dairy Science. 2019;102(6):5419-5429. DOI: https://doi.org/10.3168/jds.2018-15295.
7. Thornton HR, Hastings EG. Studies on oxidation-reduction in milk: The methylene blue reduction test. Journal of Dairy Science. 1930;13(3):221-245. DOI: https://doi.org/10.3168/jds.S0022-0302(30)93520-5.
8. De Silva SASD, Kanugala KANP, Weerakkody NS. Microbiological quality of raw milk and effect on quality by implementing good management practices. Procedia Food Science. 2016;6:92-96. DOI: https://doi.org/10.1016/j.profoo.2016.02.019.
9. Boor KJ, Wiedmann M, Murphy S, Alcaine S. A 100-year review: Microbiology and safety of milk handling. Journal of Dairy Science. 2017;100(12):9933-9951. DOI: https://doi.org/10.3168/jds.2017-12969.
10. Official methods of analysis of AOAC International, 21th Edition. Washington DC: The Association of Official Analytical Chemists, 2019.
11. Lilliefors HW. On the Kolmogorov-Smirnov test for normality with mean and variance unknown. Journal of the American Statistical Association. 1967;62(318):399-402. DOI: https://doi.org/10.1080/01621459.1967.10482916.
12. Lilliefors HW. On the Kolmogorov-Smirnov test for the exponential distribution with mean unknown. Journal of the American Statistical Association. 1969;64(325):387-389. DOI: https://doi.org/10.1080/ 01621459.1969.10500983.
13. Dallal GE, Wilkinson L. An analytic approximation to the distribution of lilliefors test statistic for normality. American Statistician. 1986;40(4):294-296. DOI: https://doi.org/10.2307/2684607.
14. Arrieta G, Gomezcaceres L, Albis D, Calderon-Rangel A, Rodriguez V. Quality of raw milk for human consumption at two locations in Sucre (Colombia). Revista Mvz Cordoba. 2019;24(3):7355-7361. DOI: https://doi.org/10.21897/rmvz.1829.
15. Kruskal WH, Wallis WA. Use of ranks in one-criterion variance analysis. Journal of the American Statistical Association. 1952;47(260):583-621. DOI: https://doi.org/10.1080/01621459.1952.10483441.
16. Choi W, Lee JW, Huh MH, Kang SH. An algorithm for computing the exact distribution of the Kruskal-Wallis test. Communications in Statistics-Simulation and Computation. 2003;32(4):1029-1040. DOI: https://doi.org/10.1081/sac-120023876.
17. Wilcoxon F. Individual comparisons by ranking methods. Biometrics Bulletin. 1945;1(6):80-83. DOI: https://doi.org/10.2307/3001968.
18. García-Martínez EM, Fuentes-López A, Fernández-Segovia I. Determinación de la calidad higiénica de la leche mediante la medición indirecta del tiempo de reducción del azul de metileno o prueba de la reductasa microbiana [Determination of the hygienic quality of milk by indirect measurement of the reduction time of methylene blue or microbial reductase test]. València: Universitat Politècnica València; 2014. 7 p. (In Spanish).
