PhD Thesis
Smallholder dairy farmers produce about 70% of cow milk in Tanzania. As the dairy
industry grows, it is increasingly becoming important to focus on milk quality and safety in
order to produce and retain the existing markets. This study was conducted in order to
assess the milk quality and safety in smallholder dairy farming systems in Mbeya rural and
Mbozi districts. A cross sectional study was conducted using a structured questionnaire,
observation as well as collection of milk samples from smallholder dairy farmers (n=96),
milk collection points (n=18) and milk shops (n=36) during the wet and dry seasons of the
year 2015 to 2016. A total of 192, 36 and 216 fresh and processed bovine milk samples
were collected from individual farmer’s herds, milk collection points and milk shops,
respectively. Data on existing practices, general microbial quality of the milk, prevalence
of Staphylococcus aureus and risk of human exposure to Staphylococcus enterotoxins gene
through raw and ready to consume milk were collected. The results showed that 96.9% of
the respondents washed hands and cleaned the cow udders before milking. About 42.2% of
the respondents used individual cloth for drying the udder and teats while the rest (57.8%)
used one cloth to dry subsequent cows. Milk samples collected from cows kept in barn
with concrete floor were significantly (p<0.001) associated with lower Total Bacteria
Counts (TBC) and Total Staphylococcus Counts (TSC). Frequency of cleaning of cow’s
barn had significantly (p<0.05) lower TBC, TSC and Total Coliform Counts (TCC). The
practice of not fore milking and not discarding first strips of milk was significantly
(p<0.05) associated with higher TBC. Milking practices which included washing of hands,
udder and teat, drying of teats using individual cloth/towel per cow and followed by fore
milking yielded milk with significantly (p<0.05) lower bacteria counts than other practices.
Similarly, cow barns with concrete floors and cleaned twice or more daily produced milk
with lower bacteria counts than cows kept on floors made from timber and mad. The overall mean for bacteria contamination showed that, TBC, TCC and TSC in the milk were
5.86, 3.47, 4.79 and 5.72, 3.40, 4.25 log 10 cfu/ml in Mbeya rural and Mbozi districts,
respectively. Mean TBC and TCC were significantly (p<0.05) higher in the wet than in dry
season. Total bacteria counts increased significantly (p<0.05) from farms to milk shops.
Total staphylococcus counts was comparable (p>0.05) between farms, seasons and
collection points. Fermented and “boiled-milk-served-hot” sold in the milk shops had
significantly (p<0.05) higher TCC in the wet than in the dry season in both districts, while
TSC in fermented and boiled-milk-served-hot was significantly (p<0.05) higher in Mbozi
district. The results of the current study indicated that the 58.7% of raw cow milk produced
and sold in the study area in wet season exceeded the East Africa Standards (EAS) limits
for TBC.
The overall prevalence of S. aureus in milk was 12.4%. At the farm, milk collection points
and milk shops were 3.2, 4.0 and 5.2%, respectively. The antibiotic resistance of S. aureus
isolates showed that, resistance to Penicillin was higher (57.1%) than in other antibiotics
and 57.1% of the isolates were resistant to three or more antibiotics tested in this study.
Staphylococcus aureus isolates demonstrated multiple antimicrobial resistances (MAR)
pattern, particularly with respect to Penicillin, Ampicillin, Erythromycin and Tetracycline.
The results on the prevalence of Staphylococcal Enterotoxins (SEs) gene in S. aureus
isolates showed that 36.4% of the total S. aureus isolates (n=55) had SEs coding genes.
The most frequently observed gene was Sea (32.6%) while Sej was not detected in any of
the isolates. The distribution of the SEs coding genes along the milk market channel
showed that 35, 15 and 50% came from samples collected at farm level, milk collection
points and milk shops, respectively. Hence, 65% of SEs coding genes were detected after
the milk left the farm. Exposure assessment results showed that 1197 litres (L) (90% CI,
987-1 416) of ready to consume milk was sold per day in the study area, of which 860 L (90% CI, 645-1 071) and 337 L (90% CI, 168-530) were boiled-milk-served-hot and
fermented milk, respectively. Out of the ready to consume milk, 490 L (90% CI; 464-516)
of boiled-milk-served-hot was contaminated with SEs gene compared to 77.5 L (90% CI,
67-88) of fermented milk. Number of daily consumers of milk and milk products was 2
394 (90% CI, 2 000 – 2 788).
Exposure assessment shows that the probability of consuming boiled-milk–served-hot and
fermented milk contaminated with SEs coding genes at milk shop was 0.42 (90% CI,
0.071-0.838) and 0.17 (90% CI, 0-0.62), respectively. It was estimated that every day, 363
(90% CI, 341-385) and 58 (90% CI, 49-66) people were likely to consume respectively,
boiled-milk-served-hot and fermented milk contaminated with SEs coding gene in the
study areas. The finding shows that exposure to SEs coding gene for the people who
consume boiled-milk-served-hot was two times more (OR. 2.21 (90% CI, 0.6-6.16) likely
to occur than people who consume fermented milk. Generally, the findings of this study
shows that milk produced and marketed in the study area was contaminated with S. aureus
expressing gene for enterotoxins production and MAR which pose a potential health risk to
milk consumers. The high level of microbial counts found in the sampled cow milk
represent a poor keeping quality of milk and public health risk to the consumer. This
suggests the need for improved hygienic practice at all levels of milk production in the
smallholder dairy production system.