Sociodemographic characteristics of the respondents
Sociodemographic characteristics of direct suppliers are as represented in Table 2. There was no significant difference (p > 0.05) in gender, age of respondents in all categories, marital status in terms of being married or living with spouse, and level of education of respondents at the elementary and middle school level. However, there was a significant difference (p < 0.05) in the farming system practised in the counties where most farmers (92.3%) in Bomet practiced semi-intensive farming while 83.3% and 100% of farmers in Nakuru and Nyeri respectively practised intensive farming.
Table 2 Sociodemographic Characteristics (percent respondents) of direct suppliers in Bomet, Nakuru, and Nyeri Counties
Characteristic
|
Bomet
|
Nakuru
|
Nyeri
|
Total
|
Gender
|
|
|
|
|
Male
|
61.5
|
44.4
|
65.4
|
57.5
|
Female
|
38.5
|
55.6
|
34.6
|
42.5
|
Age (years)
|
|
|
|
|
18-35
|
26.9
|
27.8
|
11.5
|
22.1
|
35-50
|
42.3
|
38.9
|
38.5
|
39.9
|
>50
|
34.6
|
33.3
|
50.0
|
38.0
|
Education level
|
|
|
|
|
No formal education
|
3.8
|
22.2
|
0.0
|
8.2
|
Elementary
|
30.8
|
38.9
|
50.0
|
39.7
|
Middle school
|
50.0
|
38.9
|
19.2
|
37.0
|
High school
|
11.5
|
0.0
|
30.8
|
15.1
|
University
|
0.0
|
0.0
|
0.0
|
0.0
|
Marital status
|
|
|
|
|
Married
|
80.8
|
77.8
|
73.1
|
76.7
|
Single
|
19.2
|
0.0
|
11.5
|
12.3
|
Divorced
|
0.0
|
0.0
|
3.8
|
1.4
|
Widow/er
|
0.0
|
22.2
|
11.5
|
9.6
|
Farm system
|
|
|
|
|
Intensive
|
7.7
|
83.3
|
100
|
60.3
|
Semi-intensive
|
92.3
|
16.7
|
0.0
|
39.7
|
Milk handling and hygienic practices
Different handling and hygiene practices were observed in farmers in the three counties (Figure 4). There was a significant difference (p < 0.05) in the type of milking containers used by farmers in the three counties. Most used plastic containers for milking and transportation of milk with Nakuru recording the highest where 56% of the farmers used plastic. There was also a significant difference (p < 0.05) in cleaning of the sheds, use of reusable cleaning cloths to clean udders, and setting aside cows with mastitis in the three counties. While the farmers in Nyeri and Nakuru applied more hygienic practices in cleaning the cow shed and udders in comparison to farmers in Bomet, about half of the farmers in Bomet set aside mastitis cows, while none of their counterparts in Nyeri and Nakuru did the same. These practices are important in that they directly affect milk quality. There was no significant difference (p > 0.05) in refrigeration of milk by farmers and the time they took to deliver milk to the processors.
In Bomet, there was a significant association (Χ2 = 16.759, p < 0.05) between the level of education of the respondents and the type of milking can they used. Additionally, in Bomet there was a significant association (Χ2 = 17.944, p < 0.05) between the type of farming practised by the respondents and cleaning of sheds.
Cleaning practices for milk containers and udder were similar in Nakuru and Nyeri where (almost) all farmers always cleaned their containers and used a cleaning cloth for the udder (Table 3). This was different in Bomet which brought about a significant difference (p > 0.05) in the cleaning of milk containers in the three counties. It was found that all the farmers in Nyeri always cleaned milking containers however; most of them (41.4%) used water only to clean the containers while 20.7% used water with disinfectants. Cleaning of udders varied significantly (p < 0.05) in the three counties where almost half (43.5%) of farmers in Bomet used their bare hands to clean udders while most farmers in Nakuru and Nyeri used a cleaning cloth.
Table 3 Cleaning Practices of Farmers (percent respondents) in the Various Counties
County
Hygienic practice
|
Bomet
|
Nakuru
|
Nyeri
|
Total
|
Source of water
|
|
|
|
|
Tap/ pipe
|
10.7
|
21.4
|
80.2
|
35.6
|
Well
|
50.2
|
46.5
|
11.3
|
35.6
|
River
|
39.1
|
32.1
|
8.5
|
26.0
|
Frequency of cleaning
Milking containers
|
|
|
|
|
Always
|
80.7
|
100
|
100
|
93.2
|
Most often
|
19.2
|
0.0
|
0.0
|
6.8
|
Cleaning udder
|
|
|
|
|
Hand
|
42.3
|
0.0
|
9.2
|
17.8
|
Cleaning cloth
|
57.7
|
100
|
80.8
|
82.2
|
In Nyeri, there was a significant association (Χ2 = 2.41, p < 0.05) between the gender of respondents and cleaning of milking containers.
Knowledge on hygiene and milk handling practices
Table 4 shows that 38.4% of all interviewed farmers found it ok to feed Spoiled feed to their cows. This practice was also found by Kiama et al. (2016) in Kenya, where farmers commonly fed Spoiled maize and food to their animals thereby increasing the risk of exposure of mycotoxins to the animals. There was no significant difference (p > 0.05) on the dangers of feeding spoilt feed to cows in the counties. Notably, more than half (58.6%) of farmers in Nyeri said that it was okay to feed Spoiled feed to cows which would be an issue of concern on milk quality (Table 4). There was no significant difference (p > 0.05) in hygienic milking and delivering milk promptly as ways of ensuring milk did not spoil in the three counties. There was also no significant difference (p > 0.05) in knowledge of mastitis where most farmers in the counties knew the disease and could detect it in cows. There was a significant difference (p < 0.05) in adulteration and density as causes of milk rejection in the three counties. Most farmers (34.6%) in Bomet thought that milk adulteration would lead to rejection on delivery while 44.5% and 41.3% in Nakuru and Nyeri, respectively, thought that addition of water to alter the density would lead to milk rejection on delivery.
Table 4 Knowledge on Hygiene and Handling Practices by Farmers (percent respondents) in the various counties
County
Knowledge on hygiene
|
Bomet
|
Nakuru
|
Nyeri
|
Total
|
Is it okay to feed spoilt feed to cows
|
|
|
|
|
Yes
|
30.4
|
32.0
|
58.6
|
38.4
|
No
|
69.6
|
68.0
|
41.4
|
61.6
|
How do you ensure that milk does not
get spoiled during storage
|
|
|
|
|
Boiling
|
15.4
|
14.6
|
19.2
|
15.1
|
Cover container
|
7.6
|
19.7
|
42.3
|
23.3
|
Deliver promptly
|
34.6
|
5.9
|
15.4
|
19.2
|
Hygienic milking
|
3.8
|
17.6
|
3.8
|
8.2
|
Store in a cool place
|
15.4
|
5.9
|
3.8
|
8.2
|
Do Nothing
|
23.1
|
36.3
|
15.4
|
26.0
|
What are the causes of milk rejection on delivery
|
|
|
|
|
Acidity
|
3.8
|
5.6
|
3.8
|
4.1
|
Organoleptic (smell, temperature, visible foreign
particles)
|
17.3
|
5.6
|
3.8
|
9.6
|
Low density (water addition)
|
15.4
|
44.5
|
41.3
|
30.1
|
Adulteration (using other substances except
water)
|
34.6
|
12.5
|
23.2
|
23.3
|
Others
|
11.5
|
11.5
|
12.5
|
17.8
|
Don’t know
|
17.3
|
9.2
|
15.4
|
15.1
|
Can you detect mastitis in cows
|
|
|
|
|
Yes
|
95.7
|
84.0
|
86.2
|
89.0
|
No
|
4.3
|
16.0
|
13.8
|
11.0
|
In Nakuru, there was an association (Χ2 = 26.809, p< 0.05) between the level of education of respondents and their knowledge on the causes of milk rejection on delivery.
Microbial quality of milk in different collection channels
Microbial quality of milk from Bomet county
In the cooperatives channel, After Cooler milk samples from Coop 2 recorded the highest TVC (8.1 log cfu/ml), while After Cooler samples from Coop 1 had the lowest counts of 6.8 log cfu/ml (Table 5). There was, however, no significant difference (p > 0.05) among the samples along the channel. The level of TVC in all counties exceeded the 6.3 log cfu/ml set standard (EAC, 2018).
Milk samples from direct suppliers had the highest S. aureus counts (5.3 log cfu/ml) while First Bulk milk samples from Coop 1 had the lowest counts (3.5 log cfu/ml). There was a significant difference (p < 0.05) in the counts in milk samples supplied directly by farmers and those from Coop 1. However; there was no significant difference (p > 0.05) in milk samples supplied directly by farmers and those from Coop 2. Apart from First Bulk and After Cooler milk samples of Coop 1 which met the set standard of 4.7 log cfu/ml (EAC, 2018), the rest exceeded the standards.
Table 5 Microbial quality of milk in Bomet County along the collection channels
Microbial attribute
Collection channels
|
S. aureus
|
E. coli
|
L.monocytogenes
|
TVC
|
Direct suppliers
|
5.315±0.6b
|
3.268±1.2b
|
5.641±0.8a
|
7.612±0.6a
|
Coop 1 First bulk
|
3.518±0.0a
|
2.739±0.1ab
|
6.622±0.0a
|
6.851±0.0a
|
Coop 1 After Cooler
|
3.643±0.1a
|
0±0a
|
6.874±0.0a
|
6.771±0.0a
|
Coop 2 First Bulk
|
5.058±0.0ab
|
3.498±0.0b
|
6.005±0.0a
|
7.924±0.0a
|
Coop 2 After Cooler
|
4.475±0.1ab
|
5.071±0.0b
|
6.249±0.0a
|
8.095±0.0a
|
Coop 2 After Transport
|
4.198±0.0ab
|
4.541±0.1b
|
6.848±0.0a
|
7.686±0.0a
|
Results are mean of duplicate samples ± standard deviation; TVC = Total Viable Counts
Means with the same letters in superscript in the same column are not significantly different at p<0.05
E. coli counts varied significantly depending on the collection channel with milk from After Cooler samples of Coop 2 recording the highest counts (5.1 log cfu/ml) while After Cooler samples from Coop 1 had the lowest counts (0 log cfu/ml). There was no significant difference (p > 0.05) between First Bulk and After Cooler milk samples from Coop 1 (p < 0.05). All samples along the channel met the set standards of 4 log cfu/ml (EAC 2018) with the exception of After Cooler samples of Coop 2.
There were no significant (p > 0.05) variations in L. monocytogenes counts along the collection channels. Milk samples from direct suppliers had the lowest counts (5.6 log cfu/ml), while After Cooler samples from Coop 1 had the highest counts (6.9 log cfu/ml). It was noted that from the two cooperatives, After Cooler and After Transport milk samples had higher counts than First Bulk milk samples.
Microbial quality of milk from Nakuru County
Total viable counts varied significantly (p < 0.05) where After Transport samples from Coop 4 had the highest counts (9.5 log cfu/ml) while After Transport samples from Coop 3 had the lowest counts of 7.4 log cfu/ml (Table 6). All samples exceeded the set standards of 6.3 log cfu/ml (EAC 2018).
Table 6: Microbial quality of milk in Nakuru in the along the collection channels
Microbial attribute
Collection Channel
|
E. coli
|
L.monocytogenes
|
S. aureus
|
TVC
|
Direct Suppliers
|
3.948±1.2ab
|
5.789±0.5b
|
4.734±1.1a
|
8.378±1.0abc
|
Traders
|
4.469±0.8abc
|
4.605±2.5b
|
5.11±1.2a
|
9.13±0.3bd
|
Coop 3 First bulk
|
6.01±0.0abc
|
7.039±0.0b
|
6.258±0.0a
|
9.444±0.0abcd
|
Coop 3 After Cooler
|
6.348±0.0ac
|
6.585±0.0b
|
6.276±0.0a
|
9.193±0.0abcd
|
Coop 3 After Transport
|
3.379±0.1a
|
0±0.0a
|
3.726±0.1a
|
7.391±0.0a
|
Coop 4 First Bulk
|
4.911±0.0abc
|
5.885±0.1b
|
6.05±0.0a
|
8.325±0.0ab
|
Coop 4 After Cooler
|
5.94±0.1abc
|
6.017±0.0b
|
6.082±0.0a
|
8.927±0.0abcd
|
Coop 4 After Transport
|
5.949±0.0abc
|
6.626±0.1b
|
6.299±0.0a
|
9.458±0.0abcd
|
Results are mean of duplicate samples ± standard deviation; TVC = Total Viable Counts
Means with common letters in superscript in the same column are not significantly different at p<0.05
There were no significant (p > 0.05) variations in S. aureus counts along the collection channels. After Transport milk samples from Coop 4 had the highest counts (6.3 log cfu/ml) together with After Cooler and First Bulk samples from Coop 3 (6.3 log cfu/ml). After Transport samples from Coop 3 had the lowest counts (3.7 log cfu/ml) and the only one that met the set standard of 4.7 log cfu/ml.
There were no significant (p > 0.05) variations in E. coli counts along the collection channels. After Cooler milk samples from Coop 3 had the highest counts (6.3 log cfu/ml) while After Transport samples from the same cooperative had the lowest counts (3.4 log cfu/ml). Milk samples from direct suppliers, traders and Coop 3 After Transport are the only ones that met the set standards of 4 log cfu/ml.
There was a significant (p < 0.05) variation in After Transport samples from Coop 3 with the rest of the samples in L. monocytogenes counts along the collection channels. First Bulk milk samples from Coop 3 had the highest counts (7.0 log cfu/ml), while After Transport samples from the same cooperative had the lowest counts (0 log cfu/ml).
Microbial quality of milk in Nyeri county
There were no significant (p > 0.05) variations in TVC along the collection channels (Table 7). First Bulk Milk samples from Coop 5 had the highest counts (9.4 log cfu/ml) while First Bulk samples from Coop 7 had the lowest counts (8.3 log cfu/ml). All samples exceeded the set standard of 6.3 log cfu/ml (EAC 2018).
Table 7 Microbial quality of milk in Nyeri along the collection channels
Microbial attribute
Collection Centre
|
E. coli
|
L.monocytogenes
|
S. aureus
|
TVC
|
Direct Suppliers
|
5.068±1.7a
|
5.552±1.3a
|
4.463±0.7a
|
8.537±0.6a
|
Coop 5 First Bulk
|
6.291±0.0a
|
7.135±0.0a
|
6.052±0.0b
|
9.438±0.0a
|
Coop 5 After Transport
|
7.172±0.0a
|
5.148±0.1a
|
6.035±0.0b
|
9.365±0.0a
|
Coop 6 First Bulk
|
4.017±0.1a
|
5.611±0.1a
|
6.394±0.0b
|
9.304±0.0a
|
Coop 6 After Transport
|
3.952±0.1a
|
5.724±0.0a
|
5.208±0.0ab
|
9.32±0.0a
|
Coop 7 First Bulk
|
3.239±0.3a
|
5.707±0.0a
|
4.536±0.1ab
|
8.284±0.1a
|
Coop 7 After Cooler
|
3±0a
|
8.017±0.0a
|
4.573±0.0ab
|
8.442±0.0a
|
Coop 7 After Transport
|
4.677±0.1a
|
7.851±0.0a
|
4.806±0.0ab
|
9.037±0.0a
|
Results are mean of duplicate samples ± standard deviation; TVC = Total Viable Counts
Means with common letters in superscript in the same column are not significantly different at p<0.05
Milk samples from direct suppliers varied significantly (p < 0.05) with samples from Coop 5. First Bulk samples from Coop 6 had the highest S. aureus counts (6.4 log cfu/ml), while samples from direct suppliers had the lowest counts (4.5 log cfu/ml). Milk samples from direct suppliers, Coop 7 First Bulk and After Cooler are the only ones that had counts below the set standard of 4.7 log cfu/ml (EAC 2018) while the rest exceeded the set standards.
There were no significant (p > 0.05) variations in E. coli counts along the collection channels. After Transport samples from Coop 5 had the highest counts (7.2 log cfu/ml), while After Cooler samples from Coop 7 had the lowest counts (3 log cfu/ml). Milk samples from direct suppliers, Coop 5 First Bulk and Coop 5 After Transport exceeded the set standard of 4 log cfu/ml (EAC 2018) while the rest met the set standards.
There were no significant (p > 0.05) variations in L. monocytogenes counts along the collection channels. After Cooler milk samples from Coop 7 had the highest counts (8.0 log cfu/ml) followed by After Transport samples from the same cooperative (7.9 log cfu/ml), while After Transport samples from Coop 5 had the lowest counts (5.1 log cfu/ml).
Cooperatives 5 and 6 had no coolers and they recorded higher TVC, S. aureus, and E. coli counts than Coop 7 which had a cooler. On the other hand, Coop 7 had higher L. monocytogenes counts than Coop 5 and 6.
Comparison of milk quality across the studied counties
Table 8 General Microbial quality of milk in different counties
Microbial attribute
Collection Channel
|
TVC
|
E. coli
|
L. monocytogenes
|
S. aureus
|
Bomet
|
7.588±0.6a
|
3.253±1.3a
|
5.783±0.8a
|
5.132±0.7b
|
Nyeri
|
8.641±0.6b
|
4.973±1.7b
|
5.744±1.3a
|
4.656±0.8a
|
Nakuru
|
8.72±0.8b
|
4.449±1.2b
|
5.298±1.9a
|
5.092±1.2b
|
Results are mean of duplicate samples ± standard deviation; TVC = Total Viable Counts
Means with common letters in superscript in the same column are not significantly different at p<0.05
Nakuru County recorded the highest mean TVC of 8.7 log10 cfu/ml, Nyeri had the highest E. coli mean counts of 4.97 log10 cfu/ml and Bomet recorded the highest mean counts of 5.13 and 5.78 log10 cfu/ml for S. aureus and L. monocytogenes respectively (Table 8).