Physicochemical Properties and Microbial Contamination of Raw Milk Along the Value Chain in Muheza District, Tanzania

Physicochemical properties and microbial contamination are the key   determinants of milk quality. Quality milk must be free from abnormal odors, colors, and harmful microorganisms. This study aimed to assess the physicochemical properties and microbial contamination of raw milk along the value chain in the Muheza District, Tanzania. A total of 300 milk samples (216 from cow teats, 42 from farmers’ containers, 16 from milk collection centers, and 26 from milk vendors) were collected from five wards: Amani, Genge, Ngomeni, Pande Darajani, and Mkuzi in Muheza district. Physicochemical parameters were analyzed using a Lactoscan Milk Analyzer, while standard procedures were employed for microbial count assessment to determine Total Bacterial Count (TBC) and Total Coliform Count (TCC). The results showed that the physicochemical and microbial parameters met the East African Community standards. However, 19% of the samples from farmers’ and vendors’ containers showed abnormal color, and 31% from farmers’ containers and 26% from vendors’ containers showed abnormal smell. The fat content of milk samples from cow teats (3.31% ± 0.80%) was significantly lower (p < 0.05) than that of samples from vendor containers (3.77% ± 0.14%) and collection centers (3.81% ± 0.44%). The pH of milk samples from vendors (6.68 ± 0.05) was significantly higher (p < 0.05) than that of milk from farmers’ containers (6.59 ± 0.20). Furthermore, milk samples from vendors (1.78% ± 1.26%) and collection centers (1.70% ± 1.34%) contained more water than the milk samples from cow teats (0.89% ± 0.68%) and farmers’ containers (1.03% ± 1.12%). Some raw milk samples had TBC and TCC levels that exceeded the recommended standards of 20 and 4, respectively. Overall, milk from various sources is of an acceptable quality. Regular milk quality control should be institutionalized, emphazizing the improvement of hygienic practices among the actors in the value chain to reduce contamination risks. 

Introduction

Raw cow milk is a highly nutritious, natural source of nourishment that’s essential for newborn growth and a valuable part of the adult diet [1], [2]. Milk, often informally called "white gold," is rich in important nutrients, including micro and macronutrients, which are vital for mammalian development. Milk is a unique composition of proteins, fats, carbohydrates, vitamins, and minerals that make it an indispensable component of a healthy diet, benefiting both human and animal health [3]. Worldwide, cow's milk is the most widely consumed, accounting for approximately 81% of total milk production, while buffalo milk accounts for 15%, and milk from goats, sheep, and camels accounts for 4%. The consumption of milk solids in developing countries is projected to increase from 10.7 kg to 12.6 kg per person by 2030 [4]. Furthermore, the global demand for high-quality raw milk and dairy products is expected to increase owing to population growth, urbanisation, economic development, and the need for diverse food sources to meet nutritional requirements [5]. In Tanzania, approximately 70% of milk is produced by small-scale farms, where it is either consumed locally or sold through informal channels such as small traders, vendors, and kiosks [6]. Informal milk markets pose significant risks of contamination owing to insufficient regulatory oversight and limited inspections, which can compromise milk quality and safety [7].

The quality of milk sold along the milk value chain is often questionable and frequently compromised due to a high risk of contamination and deterioration of its physicochemical and microbial properties at various stages [2]. This raises significant concerns regarding milk safety and shelf life, particularly in developing countries like Tanzania. Microbial contamination primarily arises from poor hygienic practices during milking, handling, and storage, as well as contact with contaminated equipment, bedding materials, and lactating animals [8], [9]. Although raw milk is considered sterile within a healthy cow's udder, its rich nutritional composition makes it an ideal medium for microbial proliferation under suitable conditions [10], [11]. The microbial load is a key determinant of milk quality across the chain, with the presence of pathogenic bacteria such as Escherichia coli O157:H7, Campylobacter jejuni, Salmonella, Clostridium sp., Staphylococcus aureus and Listeria not only compromising its safety and shelf life but also undermining the dairy industry’s reputation and economic stability [2]. Furthermore, pathogens pose serious public health risks and contribute to foodborne outbreaks [12], [13]. The physicochemical parameters, including lactose, protein, fat content, density, and pH, with the microbial indicators such as total bacterial count and total coliform count, are critical determinants of milk quality. According to East African standards, raw cow milk should have a minimum fat content of 3.25%, a protein content of 3%, a density ranging from 1.026 g/mL to 1.032 g/mL, and a pH between 6.6 and 6.9 for human consumption. Microbial safety standards recommend that the TBC should not exceed 2,000,000 CFU/mL and that the TCC should remain below 50,000 CFU/mL [14]. Physicochemical parameters and microbial load were significantly influenced by hygienic practices employed along the chain. In Muheza District, milk production is predominantly conducted by small-scale dairy cattle farmers. The marketing stage involves multiple vendors, a limited number of milk collection centers, a processing factory, and a diverse consumer base. However, information on milk quality from different sources remains elusive. Milk produced and handled under unhygienic conditions presents serious risks to human health, highlighting the importance of ongoing quality assessment. This study was conducted to assess and compare the variation in physicochemical properties and microbial contamination of raw cow milk sampled from cow teats, farmers’ containers, vendor containers, and bulk milk tanks at milk collection centers across five wards in Muheza District, including the evaluation of adulteration as an indicator of handling practices affecting milk quality.

Materials and Methods

Description of the Study Area and the Dairy Production System

The study was conducted in Muheza District (Fig. 1), which is one of the 11 districts in the Tanga Region. The district’s center is located along the northeastern coast of Tanzania at latitude 4° 54′ 18′′ S and longitude 38° 55′ 23′′ E. Covering an area of 1497 km², the district has a tropical savanna climate characterized by a bimodal rainfall pattern where long rains occur from March to May, while short rains fall between October and December, with annual precipitation ranging from 1000 mm to 1800 mm and temperatures between 20°C and 32°C. Smallholder dairy farming is prevalent, with the majority of farmers practising a semi-intensive system in which cattle graze on natural pastures and receive supplemental feed, including cut grasses and concentrates after grazing. Few farmers, particularly in peri-urban areas, have adopted a zero-grazing (cut-and-carry) system, where cattle are housed and fed exclusively in confinement. The dairy cattle population primarily consists of Friesian, Ayrshire, and Jersey crosses with indigenous Zebu breeds, yielding an average of 10 litres of milk per cow per day [15].

Fig. 1. Map of Muheza district showing the study area.

Study Design

This cross-sectional study was conducted between March 2024 and May 2024. The study started by identifying and selecting smallholder dairy farmers (producers), vendors, and collection centers (sellers) among the sources of value chains in Muheza District. This study targeted five wards: Amani, Genge, Mkuzi, Ngomeni, and Pande Darajani. The selection criteria for participants at each node were based on the availability of milk from dairy cattle farmers and the active operation of milk vendors and collection centers during the sampling period. At the production level, milk samples were collected from cow teats (directly milked) and farmers’ containers. At the marketing level, samples were collected from vendor containers and bulk tanks at collection centers.

Sampling and Sample Collections

A simple random sampling technique was used to obtain participants from smallholder dairy farmers, milk vendors, and collection centers. A total of 300 milk samples were collected, with 216 obtained directly from teats, 42 from farmers' containers, 26 from milk vendors, and 16 from milk collection centers. Before sampling, the milk was assessed for color and smell using standard procedures that included several assessors to minimize subjectivity [16]. Milk odor was assessed by smelling it after opening the lid, and milk color was then evaluated (visually) inside the container before stirring. Subsequently, the milk was stirred thoroughly and poured into a clean transparent container for the final visual inspection of any color deviation from the normal milk color. Approximately 15 mL of raw milk was collected in duplicate from cow teats, farmers’ containers, vendors, and collection centers for subsequent physicochemical and microbial analyses. During sampling, milk was thoroughly mixed to ensure uniformity. Each sample was placed in a screw-capped Falcon tube, which was marked and stored in a cooling box with ice packs. Within four hours, the samples were delivered to the dairy laboratory at the Tanzania Livestock Research Institute (TALIRI) in Tanga. Upon arrival, the samples were placed in a freezer set to a temperature of −18°C for 24 h before undergoing physicochemical analyses.

Analysis of Physicochemical Properties

The physicochemical parameters of raw milk were evaluated at the Tanzania Livestock Research Institute, Tanga Laboratory, using a Lactoscan Milk Analyzer (Model S50, serial number 035462, Milkotronic Ltd, Bulgaria). Before analysis, each milk sample was gently stirred to 2–3 times to ensure homogeneity. A 15 mL milk sample was poured into the sample holder and positioned in the analyzer’s recess. The analyzer measured the density (g/ml), pH, protein content (%), lactose (%), and added water (%) content, according to the manufacturer's guidelines [17].

Total Bacterial Count (TBC) Analysis

The total bacterial count was evaluated at the Microbiology Laboratory of the Sokoine University of Agriculture (SUA) to assess microbial contamination in milk samples. The procedure followed the protocol described previously [16]. Briefly, 1 mL of each milk sample was aseptically transferred into a sterile tube containing 9 mL sterile peptone water and mixed thoroughly. The sample was then serially diluted up to 10¹⁰ dilution, and 0.1 mL in each dilution was inoculated onto nutrient agar (Oxoid) in Petri dishes. The plate was allowed to solidify and then incubated aerobically at 37°C for 24 h–48 h. After incubation, colony enumeration was performed using a colony counter, and the average number of colonies was used to calculate the bacterial colony-forming units per mL (CFU/mL). The results were interpreted based on East African Community (EAC) standards [14].

Total Coliform Counts (TCC) Analysis

The TCC was determined following the dilution procedure used for TBC. For each dilution, 0.1 mL of raw milk was inoculated on MacConkey agar (Oxoid) using the spread plate method. The plates were incubated at 37°C for 24 h–48 h. Colony enumeration was performed using a colony counter, and the results were interpreted according to the East African Community (EAC) standards [16].

Statistical Analysis

All data were recorded using Microsoft Excel. A contingency test using the chi-square statistic determined the association of categorical variables (milk color, odor, and microbial contamination grade) with milk sources. Quantitative variables (milk density, pH, protein, fat, and added water content) were analyzed using two-way analysis of variance (Two-way ANOVA) to evaluate the effects of wards and milk sources. Before analysis, TBC and TCC data were transformed into the log₁₀ value. Because the interaction effects between wards and milk sources were not statistically significant for milk quality parameters, one-way ANOVA was performed using JMP Pro 18® (SAS Institute) to compare milk sources, including cow teats, farmers’ containers, milk vendors, and milk collection centers. Post hoc pairwise comparisons were conducted using Tukey’s HSD test at a significance level of p < 0.05. Categorical variables are reported as frequencies and percentages, while quantitative variables are presented as mean ± standard error.

Results

Physicochemical Properties of Raw Milk along the Milk Value Chain

The organoleptic parameters of raw milk samples collected from various sources along the milk value chain largely met the standard requirements, as presented in Table I. The milk color varied significantly (p < 0.05) among sources, with most samples showing a typical yellowish-white color, which is normal for milk. However, milk odor was significantly different among the milk samples from various sources (p < 0.05). A higher percentage of samples with undesirable odors was observed in milk obtained from vendors (31%) and farmers’ containers (26%) than in samples from cow teats (4%) and milk collection centers.

Table I. Physical (Organoleptic) Characteristics of Raw Milk Collected from Different Milk Sources along the Milk Value Chain in Muheza, Tanzania

Physical parameter	Category	Milk sources				Chi-Square	p-value
		Cow teats	Farmers’ containers	Milk collection centre	Milk vendor
Colour	Normal milk	197 (91)	34 (81)	16 (100)	21 (81)	7.80	0.0503
	Abnormal milk	19 (9)	8 (19)	0 (0.0)	5 (19)
Smell	Normal smell	207 (96)	31 (74)	16 (100)	18 (69)	36.68	0.0001
	Bad smell	9 (4)	11 (26)	0 (0.0)	8 (31)

Table II shows the physicochemical parameters of the raw milk samples from various milk sources along the milk value chain. The mean fat content, protein content, pH, and added water showed significant variations, whereas lactose content and density showed non-significant differences (p > 0.05) among the milk sources. The mean fat content of milk collected directly from cow teats (3.31% ± 0.801%) was significantly lower (p < 0.05) than that of milk obtained from vendors (3.77% ± 0.149%). Furthermore, the pH level of milk from vendors (6.68 ± 0.055) was significantly higher (p < 0.05) than that from farmers’ containers (6.59 ± 0.204). The percentage of added water was higher (p ≤ 0.05) in milk from vendors and collection centers than in milk from teats (0.89% ± 0.682%) and farmers’ containers (1.03% ± 1.118%).

Table II. Physicochemical Parameters of the Raw Milk Collected from Different Milk Sources along the Milk Value Chain in Muheza, Tanzania

Milk quality parameter	Milk sources				p-value
		Mean ± Standard error
	Cow teats	Farmers’ containers	Milk collection centre	Milk vendors
Fat (%)	3.31 ± 0.50b	3.52 ± 0.11ab	3.81 ± 0.19ab	3.77 ± 0.14a	0.0028
Lactose (%)	3.51 ± 0.64	3.57 ± 0.71	3.69 ± 0.57	3.61 ± 0.70	0.6267
Protein (%)	3.01 ± 0.03b	3.03 ± 0.06b	3.26 ± 0.10a	3.21 ± 0.08a	0.0137
Density (g/ml)	1.028 ± 0.19	1.028 ± 0.45	1.028 ± 0.73	1.027 ± 1.57	0.3151
pH	6.64 ± 0.01ab	6.59 ± 0.20b	6.61 ± 0.03ab	6.68 ± 0.02a	0.0376
AddedWater(%)	0.89 ± 0.05b	1.03 ± 0.13b	1.70 ± 0.21a	1.78 ± 0.16a	0.0001

Microbial Contamination of Raw Milk along the Milk Value Chain

Microbial contamination along the milk value chain was classified according to the East Africa Community (EAC) microbial standards (Table III). The distribution of milk samples across different contamination grades (I, II, III, and IV) varied significantly (p < 0.05) among the sources. A large proportion of milk samples from cow teats (85%), farmers’ containers (69%), milk collection centers (88%), and milk vendors (96%) were within grade I (<200,000 CFU/mL) for TBC. However, a small proportion of samples exceeded the recommended microbial limits, with grade IV contamination (>2,000,000 CFU/mL) observed in 6% of cow teat samples and 17% of samples from the farmers’ containers. For TBC, a large proportion of milk samples also met grade I (Very Good), with 95%, 90%, 81%, and 88% of samples from cow teats, farmers' containers, milk collection centers, and milk vendors, respectively, within the acceptable limits. Moreover, a few samples (1%–5%) from cow teats and farmers’ containers exceeded the grade III TCC limit (>50,000 CFU/mL). No samples from milk collection centers or vendors surpassed the EAC recommendations. There were no significant differences in mean TBC and TCC levels among the sources (Table IV). However, samples from farmers’ containers showed the highest mean TBC (4.71 ± 0.28 Log₁₀ CFU/mL) and TCC (4.33 ± 0.97 Log₁₀ CFU/mL), indicating increased microbial contamination at this stage of value chain.

Table III. Grades for Microbial Contamination of Raw Milk Samples Collected from the Different Milk Sources along the Milk Value Chain in Muheza, Tanzania

Microbial parameters (CFU/ml)	Grades	Milk sources				Chi-Square	p-value
		Cow teat	Farmers containers	Milk collection centre	Milk vendors
TBC	I	185 (85)	29 (69)	14 (88)	25 (96)	17.68	0.0390
	II	15 (7)	3 (7)	2 (12)	0 (0)
	III	4 (2)	3 (7)	0 (0)	0 (0)
	IV	12 (6)	7 (17)	0 (0)	1 (4)
TCC	I	206 (95)	38 (90)	13 (81)	23 (88)	13.36	0.0377
	II	8 (4)	2 (5)	3 (19)	3 (12)
	III	2 (1)	2 (5)	0 (0)	0 (0)

Table IV. Microbial Contamination in Raw Milk Samples Collected from the Different Milk Sources along the Milk Value Chain in Muheza, Tanzania

Microbial parameter (Log10CFU/ml)	Milk sources				p-value
	Mean ± Std. Error
	Cow teats	Farmers containers	Milk collection centre	Milk vendors
TBC	4.31 ± 0.13	4.71 ± 0.28	4.15 ± 0.59	4.41 ± 0.54	0.7020
TCC	3.91 ± 1.01	4.33 ± 0.97	3.63 ± 0.80	4.19 ± 0.51	0.7175

Discussion

Based on the analysis of physicochemical properties and microbial contamination of raw milk conducted in this study, milk handled by the majority of dairy value chain actors in Muheza District, Tanzania has acceptable quality. This is the standard stipulated by the East African Community. However, notable variations in quality attributes can be observed among different milk sources within the dairy chain, particularly in terms of microbial contamination, added water, and pH stability.

Physicochemical Properties of Raw Milk along the Milk Value Chain

Physical (organoleptic) assessment revealed that a higher percentage of milk samples from all sources were free from abnormal odors, discoloration, and foreign materials, indicating relatively good handling practices across the value chain. However, a significantly higher percentage of samples with undesirable odors was detected in milk from vendors' and farmers’ containers than in milk from cow teats and collection centers. This variation could be attributed to contamination from unclean containers and the influence of animal feed types, medication residues, or acaricides, as previously reported [18]. Contamination with cow dung or milk from mastitis-infected animals may also have contributed to the abnormal color observed in a few samples. The presence of fewer abnormal characteristics in milk collected directly from cow teats aligns with expectations, as this milk has minimal exposure to external contaminants [2].

The fat content across all sampling points complied with EAC standards and aligned with previous reports, indicating that fresh milk typically contains 3%–4% fat [14], [18], [19]. The significantly lower fat content observed in milk from cow teats compared to vendors can be explained by the natural progression of fat concentration during the milking process, with milk drawn first having a lower fat content than milk drawn later [19]. Fat is one of the most variable components of milk [20]. Its fluctuation is consistent with previous studies [21], [22], which highlighted how genetic and non-genetic factors, including milking intervals, lactation stage, feeding regimes, seasonal variations, and completeness of milking, affect milk fat percentage because fat tends to be retained in the udder during milking [23]. Similar variations were observed between different sampling points in the value chain [5], [22].

The pH of milk samples varied significantly, with milk from farmers’ containers showing values slightly below the range recommended by EAC standards (6.6–6.8), which indicated the onset of acidification [18]. This could be attributed to the bacterial activity resulting from insufficient cooling and unsanitary procedures, particularly with milking and storage equipment. Although reduced pH is often associated with spoilage and a bitter taste [24]–[26], however, such variations are not always directly linked to higher acidity but may reflect differences in chemical composition and management practices [21]. The higher pH values observed in vendors’ milk compared to farmers’ containers suggest that milk quality may partially recover during subsequent handling, possibly because of better storage conditions at later stages of the value chain.

The added water was significantly higher in milk from vendors and collection centers than in samples from cow teats and farmers’ containers, suggesting that adulteration occurs primarily at later stages of the value chain. This practice is employed to increase milk volume and profits, thereby negatively impacting the nutritional value of milk and introducing potential microbial hazards, consistent with findings reported previously [2], [27]. Water adulteration affects processing quality and product consistency, and undermines consumer trust and safety [28], [27]. Detection of water adulteration is crucial for maintaining quality control across the milk value chain and the dairy industry [28]. These findings highlight the need for targeted interventions at specific points in the chain.

Microbial Contamination of Raw Milk along the Milk Value Chain

Total Bacteria Counts of all sampling points were within EAC standards (<6.3 Log₁₀CFU/mL). Most samples from cow teats, farmers’ containers, collection centers, and vendors fell within grade I. The results for TBC and TCC observed in this study agree with the findings reported previously [16], [20], [29]. Aluminium containers with wide necks facilitate proper cleaning and contribute to the relatively low microbial contamination. However, variations in contamination levels across the value chain could be attributed to factors such as inadequate container cleaning, poor milk handling hygiene, extended storage times, and uncontrolled temperatures during transportation [2], [30]. The highest mean TBC was observed in the farmers’ containers, suggesting that this stage may represent a critical control point for bacterial contamination of raw milk.

Total Coliform Counts did not differ significantly along the value chain and remained within the EAC upper limit of 4.69 log₁₀ CFU/mL. A large proportion of samples from cow teats, farmers’ containers, milk collection centers, and vendors met the grade I standard for TCC. These coliform levels were also lower than those reported in similar studies [10]. Coliforms in milk suggest inadequate hygienic conditions, most commonly caused by fecal contamination, the use of contaminated water, and insufficiently cleaned equipment. The detection of coliforms in samples from cow teats suggests udder contamination from fecal matter or inadequate washing practices [2], [11], [30]. Beyond safety concerns, high coliform contamination significantly reduces the shelf life of dairy products [5].

The relatively low microbial contamination observed throughout the value chain reflects improved hygiene practices, including proper barn cleanliness, clean milking utensils, hand washing before milking, proper udder cleaning with warm water, and the use of appropriate milk containers. The study observed that cows were milked in their cowsheds, and recommendations for optimal practices included cleaning the animal house before milking, washing the udder with warm water, drying it with a clean towel, and using suitable disinfectants for teat dipping after milking [2], [20]. Additionally, the short storage time at the farm level, with farmers promptly delivering to milk collection centers, minimized bacterial proliferation. These practices have been demonstrated to improve milk quality throughout the value chain [20], [22]. The microbial contamination of milk from vendors and collection centers could be attributed to the bulking of milk from different sources and the poor cleanliness of the storage equipment. Furthermore, an irregular electric power supply can disrupt the cold chain, which may lead to higher levels of microbial contamination in the milk [3], [31].

Conclusion and Recommendation

In conclusion, this study found that while the variation in physicochemical properties met East African Community standards, concerns remain about water adulteration, low fat content, and pH deviations, particularly at the vendor and collection center levels. Although microbiological contamination is within regulatory limits, it is influenced by hygiene practices, storage conditions, and transportation, which can pose potential risks to milk quality and safety. Overcoming these challenges requires the monitoring and enforcement of adulteration, hygienic measures, and proper storage. Strengthening quality control services, ensuring cold chain integrity, and promoting appropriate milking and handling procedures can further mitigate contamination risk. Furthermore, training stakeholders in hygienic milk handling and the risks of poor practices will help ensure consumer safety and maintain product integrity. Continuous monitoring of milk quality at all stages of the value chain is critical for ensuring safety and meeting industrial standards. Furthermore, future studies are proposed to evaluate seasonal variability and explore the causes of contamination in depth, including stakeholder perceptions and practices.

Acknowledgment

The authors are grateful for the financial assistance provided by the Tanzania Livestock Research Institute (TALIRI) in collaboration with the Agriculture and Food Development Authority of Ireland, TEAGASC, under the MAZIWA FAIDA Project. Furthermore, we would like to express our profound gratitude to the milk industry stakeholders as well as extension officials, who played a significant role in this study. We also express our sincere thanks to the staff of TALIRI-Tanga for their invaluable collaboration and support throughout the research process.

Conflict of Interest

The authors declare that they have no conflict of interest.