COBISS: 1.08 Agris category code: Q01
CHARACTERIZATION OF MILK COAGULATION ABILITY IN BULK MILK SAMPLES
Valentina TOFFANIN 1, Massimo DE MARCHI 2, Mauro PENASA 2, Denis PRETTO 2, Martino CASSANDRO 2
ABSTRACT
Basic requirements for cheese-making production are milk coagulation properties (MCP), which can be measured as: (i) rennet coagulation time (RCT, min), (ii) curd-firming time (k20, min), and (iii) curd firmness (a30 mm). The aim of this study was to investigate sources of variation of coagulation properties in bulk milk from dairy cattle farms of northern Italy. A total of 1,570 samples were collected from 436 herds that delivered milk to 4 dairy cooperatives, and analyzed for quality traits and MCP. About 4% of total samples did not coagulate within 30 min. Rennet coagulation time, k20, and a30 averaged 18.83 min, 6.85 min, and 26.97 mm, respectively. Milk coagulation properties were analyzed through a general linear model which included the fixed effects of dairy cooperative, herd nested within dairy cooperative, year and season of sampling, and milk quality traits, inserted as class factors in the model. Dairy cooperative and herd effects were the major sources of variation in explaining the variation of MCP (P < 0.001). Season of sampling was statistically significant for RCT and k20 (P < 0.05), showing the best results in summer, whereas year of sampling was statistically significant only for RCT (P < 0.001). Rennet coagulation time was influenced by titratable acidity and bacterial count (P < 0.05), and k20 and a30 by casein content and titratable acidity (P < 0.01).
Key words: bulk milk / coagulation properties / dairy industry / cheese making / milk acidity
1 introduction
Milk and dairy products are key components for Italian agricultural and food sectors, and in several areas of the country they represent the primary source of income for farmers. At present, 70% of the milk available in Italy is used for cheese production and about 35% is transformed into PDO products (Pieri, 2011). More than 83% of cow milk is obtained in northern Italy, where Veneto Region represents the third producer (Pieri, 2011). In this Region cheese-making is performed by several dairy cooperatives that process the milk collected from dairy farms.
As reported by Summer et al. (2002), basic requirements for cheese-making are milk coagulation properties (MCP), which are: (i) rennet coagulation time (RCT, min), (ii) curd-firming time (k20, min), and (iii)
curd firmness (a30 mm). These parameters can be determined by Formagraph (Annibaldi et al., 1977; Zannoni and Annibaldi, 1981; MacMahon and Brown, 1982). The milk clotting properties are important both with regard to quality and yield of cheese (Wedholm et al., 2006); in fact, a firmer curd at cutting is positively correlated to cheese yield (Aleandri et al., 1989; Martin et al., 1997).
Milk coagulation properties are affected by different factors such as milk quality, breed, season, and herd. Several studies have investigated the relationship between MCP and milk quality traits both from phenotypic and genetic point of view. Jôudu et al. (2008) reported that an increase of milk protein and casein contents reduced RCT. Politis and Ng-Kwai-Hang (1988) and Okigbo et al. (1985b) found an improvement of MCP associated to decreasing values of somatic cell count (SCC). Okigbo et al. (1985a) and Formaggioni et al. (2001) reported that
1	Corresponding author, School in Animal and Food Science, Univ. of Padova, Viale dell'Università 16, 35020 Legnaro (PD), Italy; valentina.toffanin@studenti.unipd.it
2	Dept. of Agronomy, Food, Natural resources, Animals and Environment (DAFNAE), Univ. of Padova, Viale dell'Università 16, 35020 Legnaro (PD), Italy
MCP were better when associated to decreasing values of pH and increasing values of titratable acidity (TA), respectively. Besides milk quality, breed of cow and herd are important sources of variation for MCP. De Marchi et al. (2007) reported relevant differences among five dairy cattle breeds for MCP measured on bulk milk samples. A strong breed effect was also confirmed in individual milk samples (Mariani et al., 1997, 2002; Tyriseva et al., 2002). Finally, the season affects MCP even if contradictory findings are available in literature. Chladek et al. (2011) found that season significantly affected MCP; in particular, these authors reported better RCT in summer than in other seasons, and this trend was confirmed also by Bar-lowska et al. (2012). Opposite results were obtained by Joudu et al. (2008) and De Marchi et al. (2007) showing worst MCP for milk samples collected in summer.
The aim of this study was to investigate sources of variation of MCP in bulk milk destined to cheese production.
2 material and methods
2.1 DATA COLLECTION AND LABORATORY
ANALYSES
A total of 1,570 bulk milk samples were collected between June 2008 and November 2009 from 436 dairy herds (mainly rearing Holstein-Friesian cows) that delivered milk to 4 dairy cooperatives. After collection, samples (100 mL) without preservative were transferred to the Milk Quality Laboratory of Veneto Agricoltura (Thiene, Italy) and assessed for fat, protein, and casein contents (MilkoScan1™ 6000, Foss Electric A/S, Hillerod, Denmark), SCC (Cell Fossomatic 250, Foss Electric A/S, Hillered, Denmark), pH, and TA expressed in Soxhlet-Henkel degrees (°SH/50 ml; Crison Compact D, Crison Instruments SA, Alella, Spain) as proposed by Anonymous (Sauregradbestimmung ..., 1963).
Analysis of MCP was carried out following Cassan-dro et al. (2008); briefly, milk samples (10 mL) were heated to 35 °C and 200 pL of rennet (Hansen standard 190 with 63% of chymosin and 37% of pepsin, Pacovis Am-rein AG, Bern, Switzerland) diluted 1.6% in distilled water, was added to milk. Milk coagulation properties were determined through a computerized renneting meter for 30 min. Measurements recorded were rennet coagulation time (RCT, min), curd-firming time (k20, min), and curd firmness (a30, mm). Rennet coagulation time is the time between the addition of the clotting enzyme to the milk and the beginning of coagulation, k20 corresponds to the time required to achieve 20 mm of firmness, and a30 ex-
presses the millimeters of curd obtained at 30 min from rennet addition.
2.2 STATISTICAL ANALYSES
Milk samples that did not coagulate within 30 min (3.95% of total records) were discarded from the database prior to statistical analysis. An analysis of variance was performed on MCP using the GLM procedure of SAS (SAS, 2008) according to the following linear model:
y.,, = u + DC + H(DC) + YS, + SS, +
' ijklmnopqr 1	i	jv r	k	l
CAS + FAT + TA + SCC + BC + £..,, ,
m	n	o	p	q ijklmnopqr
where y..,, is RCT, L, or a,„; u is the overall mean;
' ijklmnopqr	20	30 r
DC. is the fixed effect of the ith dairy cooperative (i = 1 to 4); H is the fixed effect of the jth herd nested within dairy cooperative (j = 1 to 436); YSk is the fixed effect of the kth year of sampling (k = 2008, 2009); SSj is the fixed effect of the lth season of sampling (l = 1 to 4); CASm is the fixed effect of the mth class of casein content of milk (m = 1 to 5); FATn is the fixed effect of the nth class of fat content of milk (n = 1 to 5); TAo is the fixed effect of the oth class of titratable acidity of milk (o = 1 to 5); SCCp is the fixed effect of the pth class of somatic cell content of milk (p = 1 to 5); BCq is the fixed effect of the qth class of bacterial count (q = 1 to 5); and £..,, is the random
v 1	'	ijklmnopqr
residual effect ~N(0, a2e). The herd effect was tested on the error line of herd within dairy cooperative variance. Significance was set at P < 0.05.
3 results and discussion
Rennet coagulation time, k20, and a30 averaged 18.83 min, 6.85 min, and 26.97 mm, respectively (data not shown), and the coefficient of variation ranged from 0.20 to 0.30, highlighting a good variation of MCP. De Marchi et al. (2007) reported similar values for herd bulk milk samples collected on 5 dairy cattle breeds in northern Italy. The mean values for RCT, k20, and a30 were not close to those recommended by Zannoni and Annibaldi (1981) for cheese-making; these authors reported optimal values of 13 min for RCT, 9 min for k20, and 35 mm for a30. Sixty two samples (3.95% of total records) did not coagulate within 30 min from the beginning of the analysis. Non-coagulating milk samples ranged between 7.5 and 13.2% in Ikonen et al. (2004), Tyriseva et al. (2004), and Cassandro et al. (2008), i.e., they were much higher than findings from our study. A possible explanation of such difference is that we consider bulk milk, whereas previous studies dealt with individual samples. It is likely that
Table 1: Results from ANOVA for milk coagulation properties of bulk milk samples
	Trait1					
	RCT, min		k20, min		a30, mm	
Effect	F	P-value	F	P-value	F	P-value
Dairy cooperative2	25.03	<0.001	9.36	<0.001	25.57	<0.001
Herd (within dairy cooperative)	1.86	<0.001	1.57	<0.001	1.83	<0.001
Year of sampling	19.08	<0.001	1.12	0.290	0.07	0.797
Season of sampling	13.75	<0.001	2.66	0.047	1.51	0.211
Casein, %	0.71	0.585	4.84	0.001	5.88	<0.001
Fat, %	0.58	0.676	1.06	0.376	1.47	0.209
Titratable acidity, °SH/50 mL	14.31	<0.001	4.78	0.001	13.63	<0.001
Somatic cell count, cells/mL	2.31	0.056	1.13	0.339	0.97	0.422
Bacterial count, cells/mL	2.48	0.042	1.70	0.148	1.56	0.183
R2	0.52		0.52		0.52	
RMSE3	3.03		1.65		6.72	
1 RCT = rennet coagulation time; k20 = curd-firming time; a30 = curd firmness; 2 Tested on herd within dairy cooperative variance; 3 RMSE = root mean square error.
the variation in bulk milk is reduced as a consequence of the blending of milk from different cows.
Results from the analysis of variance for MCP are shown in Table 1. The coefficient of determination (R2) was the same for all the studied traits (0.52). The dairy cooperative and herd effects were highly significant (P < 0.001) in explaining the variation of MCP. Year of sampling was statistically significant (P < 0.001) only for RCT, whereas season of sampling was significant for RCT and k20 (P < 0.05). The importance of herd effect on MCP was reported by several authors (Ikonen et al., 2004; Tyri-seva et al., 2004; De Marchi et al., 2007; Vallas et al., 2010; Barlowska et al., 2012). Regarding the effect of season of
sampling on MCP, Barlowska et al. (2012), Chladek et al. (2011), and Hanus et al. (2010) reported the shortest RCT in summer, whereas De Marchi et al. (2007) found shorter RCT and higher a30 in fall.
The effect of milk composition on MCP is reported in Table 1. Titratable acidity and bacterial count significantly (P < 0.05) influenced RCT, whereas SCC approached significance (P < 0.10) for this trait. Casein content and TA significantly influenced (P < 0.01) and a30. Rennet coagulation time decreased with increasing values of TA, and increased with increasing values of SCC (Fig. 1). Better values of k20 were associated to higher content of casein in milk and to higher TA (Fig. 2). Finally, a30 in-
(a)
(b)
21
20. = 19.
5 1S a i?. 16. 1?.
V
I I I I ■
Cl C2 C.3 C4 C5 Titratable acidity, °SH 50niL
Figure 1: Least squares means of rennet coagulation time (RCT, min) across classes of (a) titratable acidity and (b) somatic cell count. Titratable acidity (°SH/50 mL): C1 < 3.00; C2 3.00 to 3.20; C3 3.20 to 3.40; C4 3.40 to 3.60; C5 > 3.60. Somatic cell count: C1 < 100,000; C2 100,000 to 200,000; C3 200,000 to 300,000; C4 300,000 to 400,000; C5 > 400,000.
V. TOFFANIN et al.
(a)
(b)
C2 Ci C4 Casein, %
Figure 2: Least squares means of curd-firming time (k20, min) across classes of (a) casein content and (b) titratable acidity. Casein content: C1 < 237; C2 2.37 to 2.52; C3 2.52 to 2.67; C4 2.67 to 2.82; C5 > 2.82. Titratable acidity (°SH/50 mL): C1 < 3.00; C2 3.00 to 3.20; C3 3.20 to 3.40; C4 3.40 to 3.60; C5 > 3.60.
creased with increasing values of casein content and TA (Fig. 3). These results were previously confirmed by De Marchi et al. (2007) who found lower RCT and k20 values associated to higher TA of bulk milk. The fundamental role of TA in explaining the variation of MCP was also reported by Formaggioni et al. (2001) in bulk milk samples and Okigbo et al. (1985a) in individual milks.
The relationship between MCP and composition and acidity of milk has also been studied at cow level. Favorable association of technological properties with acidity was found by Cassandro et al. (2008) who reported phenotypic and genetic correlations of -0.43 and -0.50 between RCT and TA, and 0.41 and 0.66 between a30 and TA. The same authors estimated phenotypic and genetic correlations of 0.17 and 0.25 between RCT and SCC, and -0.14 and -0.40 between a30 and SCC, in agreement with estimates from a previous study by Ikonen et al. (2004). Similar findings were also reported by Politis and Ng-Kwai-Hang (1988) who detected a worsening of MCP when values of SCC exceeded 500,000
cells/mL. The relationship between casein content and MCP has been investigated by several authors who reported a positive association between these traits: an increase of milk casein content is associated to an increase of the strength of the coagulum (Mariani et al., 1976; Okigbo et al., 1985a; Politis and Ng-Kwai-Hang, 1988; Ikonen et al, 1997; Tyriseva et al, 2003, 2004).
4 conclusions
This study showed the strong influence of dairy cooperative and herd effects on MCP measured on bulk milk. The best MCP were obtained in summer; however, the season effect should be further investigated as contradictory results have been found in literature. Quality of milk affected MCP. In particular, TA strongly affected RCT, k20, and a30, and casein content had a large influence
on k20 and a30.
(a)
(b)

C3 C4 Casein, %
Figure 3: Least squares means of curd firmness (a30, mm) across classes of (a) casein content and (b) titratable acidity. Casein content: C1 < 2.37; C2 2.37 to 2.52; C3 2.52 to 2.67; C4 2.67 to 2.82; C5 > 2.82. Titratable acidity (°SH/50 mL): C1 < 3.00; C2 3.00 to 3.20; C3 3.20 to 3.40; C4 3.40 to 3.60; C5 > 3.60.
5	acknowledgments
The authors thank Lattebusche (Busche, Italy), Latteria di Soligo (Soligo, Italy), Latterie Trevigiane (Vede-lago, Italy), and Latterie Vicentine (Bressanvido, Italy) for technical support, and laboratories of the Breeders Association of Veneto region (ARAV, Padova, Italy) and Veneto Agricoltura (Thiene, Italy) for milk bulk analyses.
6	references
Aleandri R., Schneider J.C., Buttazzoni L.G. 1989. Evaluation of milk for cheese production based on milk characteristics and Formagraph measures. Journal of Dairy Science, 72: 1967-1975
Annibaldi S., Ferri F., Morra R. 1977. Nuovi orientamenti nella valutazione tecnica del latte: tipizzazione lattodinamo-grafica. Scienza e Tecnica Lattiero-Casearia, 28: 115-126 Barlowska J., Litwinczuk Z., Brodziak A., Chabuz W. 2012. Effect of the production season on nutritional value and technological suitability of milk obtained from intensive (TMR) and traditional feeding system of cows. Journal of Microbiology, Biotechnology and Food Sciences, 1: 1205-1220 Cassandro M., Comin A., Ojala M., Dal Zotto R., De Marchi M., Gallo L., Carnier P., Bittante G. 2008. Genetic parameters of milk coagulation properties and their relationships with milk yield and quality traits in Italian Holstein cows. Journal of Dairy Science, 91: 371-376 Chladek G., Cejna V., Falta D., Machal L. 2011. Effect of season and herd on rennet coagulation time and other parameters of milk technological quality in Holstein dairy cow. Acta Universitatis Agriculturae Et Silviculturae Mendelianae Brunensis, 59: 113-118 De Marchi M., Dal Zotto R., Cassandro M., Bittante G. 2007. Milk coagulation ability of five dairy cattle breeds. Journal of Dairy Science, 90: 3986-3992 Formaggioni P., Malacarne M., Summer A., Fossa E., Mariani P. 2001. Milk with abnormal acidity. VI. The role of phosphorus content and the rennet-coagulation properties of Italian Friesian herd milks. Annali della Facoltà di Medicina Veterinaria, Università degli studi di Parma, 21: 261-268 Hanus O., Frelich J., Tomaska M., Vyletëlova M., Gencurova V., Kucera J., Trinacty J. 2010. The analysis of relationships between chemical composition, physical, technological and health indicators and freezing point in raw cow milk. Czech Journal of Animal Science, 55: 11-29 Ikonen T., Morri A., Tyrisevä A.-M., Ruottinen O., Ojala M. 2004. Genetic and phenotypic correlations between milk coagulation properties, milk production traits, somatic cell count, casein content and pH of milk. Journal of Dairy Science, 87: 458-467 Ikonen T., Ojala M., Syväoja E.-L. 1997. Effects of composite casein and beta-lactoglobulin genotypes on renneting properties and composition of bovine milk by assuming an animal model. Agricultural and Food Science in Finland, 6: 283-294
Jöudu I., Henno M., Kaart T., Pussa T., Kart O. 2008. The ef-
fect of milk protein on the rennet coagulation properties of milk from individual dairy cows. International Dairy Journal, 18: 964-967 MacMahon D.J., Brown R.J. 1982. Evaluation of Formagraph for comparing rennet solutions. Journal of Dairy Science, 65: 1639-1642
Mariani P., Summer A., Formaggioni P., Malacarne M. 2002. La qualità casearia del latte di differenti razze bovine. La Razza Bruna Italiana, 1: 7-13 Mariani P., Serventi P., Fossa E. 1997. Contenuto di caseina, varianti genetiche ed attitudine tecnologico-casearia del latte delle vacche di razza Bruna nella produzione del for-maggio grana. Allegato a La Razza Bruna Italiana, 2: 8-14 Mariani P., Losi G., Russo V., Castagnetti G.B., Grazia L., Mori-ni D., Fossa E. 1976. Cheesemaking trials with milk characterized by K-casein A and B variant in the production of the Parmigiano-Reggiano cheese. Scienza e Tecnica Lattiero-Casearia, 27: 208-227 Martin B., Chamba J.F., Coulon J.B., Perreard E. 1997. Effect of milk chemical composition and clotting characteristics on chemical and sensory properties of Reblochon de Savoie cheese. Journal of Dairy Research, 64: 157-162 Okigbo L.M., Richardson G.H., Brown R.J., Ernstrom A. 1985a. Effects of pH, calcium chloride, and chymosin concentration on coagulation properties of abnormal and normal milk. Journal of Dairy Science, 68: 2527-2533 Okigbo L.M., Richardson G.H., Brown R.J., Ernstrom A. 1985b. Coagulation properties of abnormal and normal milk from individual cow quarters. Journal of Dairy Science, 68: 1893-1896
Pieri R. 2011. Il Mercato del Latte. Rapporto 2011. Milano,
Franco Angeli: 232-234 Politis I., Ng-Kwai-Hang K.F. 1988. Effect of somatic cell counts and milk composition on the coagulation properties of milk. Journal of Dairy Science, 71: 1740-1746 SAS Institute. 2008. SAS/STAT Software, Release 9.2. Cary, NC,
SAS Institute, Inc., Sauregradbestimmung nach Soxhlet-Henkel (SH). Titratable acidity evaluation with the Soxhlet-Henkel (SH) method. 1963. Milchwissenschaft, 18: 520 Summer A., Malacarne M., Martuzzi F., Mariani P. 2002. Structural and fuctional characteristics of Modenese cow milk in Parmigiano-Reggiano cheese production. Annali della Facoltà di Medicina Veterinaria, Università degli studi di Parma, 22: 163-174 Tyriseva A.M., Vahlsten T., Ruottinen O., Ojala M. 2004. Non-coagulation of milk in Finnish Ayrshire and Holstein-Frie-sian cows and effect of herds on milk coagulation ability. Journal of Dairy Science, 87: 3958-3966 Tyriseva A.M., Ikonen T., Ojala M.. 2003. Repeatability estimates for milk coagulation traits and non-coagulation of milk in Finnish Ayrshire cows. Journal of Dairy Research, 70: 91-98
Tyriseva A.M., Vahlsten T., Ruottinen O., Ojala M. 2002. Milk coagulation ability and prevalence of noncoagulating milk in Finnish dairy cows. In: 7th World Congress on Genetics Applied to Livestock Production, Montpellier, France. CD-ROM communication No. 09-02 Vallas M., Bovenhuis H., Kaart T., Parna K., Kiiman H., Parna
E. 2010. Genetic parameters for milk coagulation properties in Estonian Holstein cows. Journal of Dairy Science, 93: 3789-3796
Wedholm A., Larsen L.B., Lindmark-Mânsson H., Karlsson A.H., Andrén A. 2006. Effect of protein composition on the cheese-making properties of milk from individual dairy cows. Journal of Dairy Science, 89: 3296-3305
Zannoni M., Annibaldi S. 1981. Standardization of the rennet-ing ability of milk by Formagraph. Scienza e Tecnica Lattie-ro-Casearia, 32: 79-94