Exploring the Interaction of Altitude, Breedand Management on Physicochemical Properties of Milk in Admixed Dairy Cattle of Muheza District

Dairy cow is a nutrient rich food containing essential nutrients such as fat, protein, lactose, minerals, and vitamins. Breed, agroecology, and management influence its quality, particularly physiochemical composition. This study assessed the physiochemical composition of milk from different breeds raised under varying altitudes and management in Muheza district, Tanga, Tanzania. A cross-sectional study involving structural questionnaires, face-to-face group discussions, and milk sampling was conducted among 207 farmers managing 400 crossbred dairy cows (197 in lowlands, 203 in highlands) data were analyzed using Crosstabs, Student’s t-test, chi-squire, and Analysis of Variance (ANOVA), in Statistical Package for Social Sciences (SPSS). Mean value of fat, protein, lactose, Ph and density in highlands and lowlands milk were 3.38%, 3.22%, 3.61%, 6.65, 1.028 g/cm³ and 2.89%, 3.03%, 3.56%, 6.79, 1.028 g/cm³, respectively. Milk fat, protein, and Ph significantly differed (p < 0.05) with altitude, while lactose and density did not. Altitudes interacted with breed, pasture type and parity affecting fat, protein, and Ph. Despite differences across groups, all milk components were within FAO, WHO and EAC Standards. Farmers are encouraged to improve herd genetics toward Guernsey and Jersey breeds due to their superior milk fat and protein content. 

Introduction

Milk is a vital biological fluid produced by mammals following parturition and serves as the primary source of nutrition for offspring. It is predominantly composed of water, along with essential nutrients such as proteins, carbohydrates (primarily lactose), fats, vitamins, and minerals [1]. Milk is a nutritive food that helps people with their daily intake of vital nutrients [2]. It contains good quality protein, is rich in a supply of healthy short-chain fatty acids, has a variety of macro and micro minerals, and is rich in bioactive peptides [3]. Milk composition and physicochemical properties are known to be affected by several factors, including animal-related issues such as parity, stage of lactation, health status, breed, and genetics [4]–[6]. Cows fed fresh green forage have been reported to have much higher unsaturated fatty acid proportions, with more polyunsaturated fatty acids compared to those fed silage [7]. Agro-ecology and season have a significant effect on the physicochemical properties of raw cow milk [8]. Similarly, the environment and altitude greatly influence milk composition and physicochemical properties [9]. Altitude has an impact on atmospheric pressure, temperature, solar radiation, and feed availability, and significantly interacts with animal physiology, resulting in changes in milk quality and composition [10], [11].

The physicochemical composition of cow’s milk is the most important parameter in the dairy industry [12]. The physicochemical properties of milk, including pH, ash, water, total dissolved solid (TDS) percentages, viscosity, refractive index, proteins, minerals, and vitamins, affect storage shelf-life and industrial processing, particularly mixing and homogenization, fluid flow, sterilization, and freezing [13], [14]. According to the East African Standards and other scientific works, the quality of milk should contain a minimum value of 2.6% fat, 3.5% protein, 7.71% solids-non-fat (SNF), specific gravity (SG) 1.030, and maximum value of total bacterial count 1.3 × 106 CFU/mL and 0.17% titrable acidity (TA) [15], [16], [8]. Research has shown that altitude affects the milk composition in several ways. Reference [17] found higher fat content in milk from cows kept at high altitudes (above 1200 m) compared to those at lower altitudes, both fed with alpine pastures. Correspondingly, [18] reported higher fat but lower protein and lactose levels in dairy cows maintained at higher altitudes than in those maintained at lower altitudes. Reference [19] observed an increase in fat, protein, urea, and somatic cell counts in milk from high-altitude farms compared with low-altitude farms. Reference [20] reported changes in milk composition across different altitudes, as cows at higher altitudes produced milk with higher fat content and lower total solids than those at lower altitudes. Variations in milk composition, apart from being influenced by altitude, are caused by other factors such as management, feed intake, and breed adaptations that contribute equally to changes in milk composition at high altitudes [21]–[23].

Tanga’s dairy innovation system is a robust animal production system utilizing improved technologies, adoption of zero-grazing, cross-breeding, access to inputs, and disease control practices [24]. Muheza District in the Tanga Region plays a significant role in dairy production, particularly through its milk supply to Tanga Fresh, a major milk processor in the northern part of Tanzania [25]. Maintaining milk quality, especially in raw milk production systems, is important for ensuring high-quality milk products, safety, and consumer confidence [26]. In tropical countries, particularly Tanzania, several researchers have explored the effects of breed, nutrition, and management on milk characteristics [27], [16], but none have assessed the influence of altitude on the physicochemical properties of milk. Therefore, this study examined the effects of altitude on cow dairy milk at low and high altitudes in Muheza District, Tanzania.

Materials and Methods

Ethical Considerations and Approval

The research was permitted by the Ministry of Livestock and Fisheries through the Ethics Review Board of the Tanzania Livestock Research Institute (TALIRI) and given Ethical Approval Reference Number; TLRI/CC.21/061. Permission to perform research was first submitted to the Tanga Regional Administrative Secretary (RAS) and thereafter to the Muheza District Executive Director (DED) and the Head of Agriculture, Livestock and Fisheries Department. Livestock officers and farmers were introduced to the research objectives and asked to provide consent before participating in the research.

Study Area

This study was conducted in Muheza District, Tanga Region, which is located on the northeastern coast of Tanzania. The district has two distinct agro-ecological zones, highlands and lowlands. The classification of these zones is primarily based on agro-ecological factors including altitude and temperature. The lowland area, which extends between 5°10′S latitude and 38°47′E longitude, lies at an altitude of approximately 200 m–900 m above sea level and has a temperature range of 25°C–32°C. Most of the open areas in the lowland zone are covered by sub-tropical forests, and receive an average annual precipitation of approximately 1000 mm, with two rainy seasons: short rains from October to December and long rains from February to May. The Highland zone, which is positioned at altitudes of 900–1060 m above sea level, is located between 5°01′S latitude and 38°48′E longitude and has a temperature range of 20°C and 28°C. The highland area receives annual rainfall between 1000 mm and 1800 mm, with a bimodal rainfall pattern characterized by two rainy seasons of short rains from October to December and long rains from February to May.

Sample Size Determination

The sample size of the study was determined according to [28], and the formula for estimating the sample size when the population is unknown is given by:

$$\begin{array}{r}n={\left(\frac{Z\times SD}{d}\right)}^2\end{array}$$

where n is the required sample size and Z is the standard normal variate at the desired normal confidence interval (1.96, 95% confidence level). Standard deviation (SD) was estimated standard deviation (assumed to be 0.5% or 50% for maximum variability) and d was the acceptable margin of error considered (0.05% or 5%). Therefore, through calculation:

$$\begin{array}{r}(n={\left(\frac{1.96\times 0.5}{0.05}\right)}^2)\end{array}$$

A minimum sample size of 384 milk samples was estimated; however, to increase the accuracy, precision, and balanced representation of the study sites, the study included 400 milk samples (203 crossbreed dairy cows from the highland zone and 197 crossbreed dairy cows from the lowland zone). The crossbred dairy cattle were selected based on their accessibility, willingness to participate, and complete breeding and management records.

Study Animals and Their Management

A total of 400 crossbred dairy cows (203 from the highlands and 197 from the lowlands) were included in this study. All animals were maintained under the farmer’s management system. All animals in the highlands were reared under a zero-grazing (indoor rearing) system, while in the lowlands, 159 cows were reared under an intensive system, five under a semi-intensive system, and 33 under an extensive system. Breed composition of animals was determined at the farm through pedigree data as previously described elsewhere [29], [30] and body condition scores were estimated under the 1–5 scale as used by other researchers, including [31], [32]. Most of the animals kept in both zones were crossbred cattle (crosses of indigenous cattle with either Friesian, Ayrshire, Jersey, or Guernsey breeds with different blood compositions). The majority of animals were fed on natural pastures (i.e., naturally occurring grasses, legumes, and other species), with a minority feeding on improved pastures. The lactating and pregnant animals were supplemented with home-compounded concentrates. In both areas, animal husbandry advice and veterinary services were provided by extension officers in specific villages and wards.

Study Design and Data Collection Method

The study employed semi-structured questionnaires and face-to-face interviews. The questionnaire was designed to cover a range of parameters relevant to dairy farming, including milk production levels, feed sources, animal houses, breeding method, age, and parity. Face-to-face focus discussion/interviews were conducted with dairy farm owners within each agroecology. These discussions assisted in collecting additional information on topics that were not captured in the questionnaire but were important, including feed availability, feeding practices, animal health, and milking practices.

Milk Sampling and Evaluation

Milk samples were collected from lactating crossbred dairy cows in the morning during the milking period. Before collection, the plastic bottles (50 mL polyethylene bottles) were washed with 30% HNO₃and thoroughly rinsed with distilled water. During milk sampling, two to three strips of milk were discarded, followed by collection of approximately 30 mL of milk. The collected milk samples were immediately placed in a cool box packed with ice, transported to the TALIRI-Tanga Laboratory, and stored in a deep freezer at temperatures between 0°C and −10°C until analysis. Before assessment, the samples were warmed to room temperature (20°C) and analyzed using a Lactoscan Milk Analyzer (Lactoscan S50) for important parameters such as fat%, protein%, lactose%, density (g/cm³), and pH according to the manufacturer’s instructions, as previously described by [33].

Statistical Analysis

The data collected during the study were stored in Microsoft Excel. The data were then cleaned, coded, and exported to the Statistical Package for the Social Sciences (SPSS, version 27) for analysis. The results are summarized as descriptive statistics: percentages, means, and standard errors. An independent test (t-test) was used to compare the mean values obtained between the study agro-climatic locations, whereas one-way Analysis of Variance (ANOVA) was used to assess mean differences between breeds, types of pasture, feed supplement provision, cow age, parity, and body condition score in the highland and lowland zones. The Least Significant Difference (LSD) test was used to assess variations in the mean values of the parameters based on altitude. In all analyses, a confidence level of 95% was observed, and a p-value less than 0.05 was considered significant. Moreover, the chi-square test was used to determine significant differences in sociodemographic parameters between the highlands and lowlands of the study area.

Results

Animal Management Practices

Smallholder dairy farmers in both lowlands and highlands kept crossbred cattle (the mixture crosses of Friesian, Ayrshire, Jersey, Guernsey, and indigenous cattle with different blood compositions). The majority of the farmers from the highlands (90%) and a moderate number in the lowlands (46.2%) kept few animals (one to five) while employing a zero-grazing system (100% in the highlands vs. 90.7% in the lowlands). Approximately 78% and 74.4% of farmers in the highlands and lowlands, respectively, predominantly fed their cattle with natural pastures (Panicum maximum, Masai grass, Native Bracharia and Napier grasses) grown on communal/open lands. A minority of farmers in the highlands (22.0%) and lowlands (25.6%) cultivated improved pastures (Guatemala, Desmodium and improved Napier grasses) for their animals. Homemade concentrate supplementation for lactating and pregnant animals was common in both highland (100%) and lowland (90.7%) farmers. Concrete floors with good roofs were more common in the lowlands (76.7%) than in the highlands (54.1%). Regarding animal breeding methods in use, artificial insemination was noted to be the favored (85.8%) breeding method for the lowland zone, whereas natural mating was common (84.2%) in the highland zone.

Physicochemical Properties of Milk

The physicochemical properties of raw cow’s milk in the two agroecological zones (highland and lowland) are presented in Tables I and II. Most milk composition changed markedly with agroecological zones, feeding practices, breed, parity, body condition, and age of the animals. The average values of milk fat, protein, lactose, density, and pH obtained from dairy farms in the study area were 3.14% ± 0.09%, 3.13% ± 0.06%, 3.59% ± 0.04%, 1.028 ± 0.001 g/cm³ and 6.72 ± 0.02, respectively. Three milk parameters (fat, protein, and pH) differed significantly among agroecological altitudes, whereas two parameters (lactose and density) were not significantly (p > 0.05) affected by altitude differences. The interaction of altitude with type of pasture fed to cows, breeds, and parity showed a significant (p < 0.05) influence on fat content, while age and body condition score of cows had no effect (p > 0.05) on milk fat content (Tables III–V). Protein, which is the most important component of milk, statistically differed (p < 0.05) between agroecological zones, where cows kept in the highland produced milk with higher protein content (mean 3.22 ± 0.057) than cows from the lowland (mean 3.03 ± 0.059). Similarly, breed and parity had a significant (p < 0.05) effect on protein content. Guernsey cows produced milk with a higher protein content than other breeds, and multiparous cows tended to produce milk with a high protein content. The percentage of lactose content in milk was not significantly (p > 0.05) affected by altitude, although the amount of lactose in milk was higher in milk produced by cows kept in the highlands. Similarly, the interaction of altitude with breed, feeding (supplementation and type of pasture), parity, and age did not significantly (p > 0.05) affect lactose content. However, the interaction of altitude with body condition score showed a significant (p = 0.027) influence on lactose content composition across milk collected from both highland and lowland areas. No significant (p > 0.05) variations in milk density were observed between altitudes, although only the parity of cows interacted positively with agroecological zones, with cows in the highlands producing slightly denser milk (1.027 g/cm³ ± 0.002 g/cm³) on average than cows in the lowlands (1.025 g/cm³ ± 0.002 g/cm³). Milk pH, which is a very important parameter for grading milk quality, was significantly different (p = 0.031) across the two agroecological zones studied. In addition, we report a significant (p < 0.05) influence of parity, body condition score of the cows, and feeding on milk pH across the studied animals.

Table I. The Descriptive Statistics of Milk Components Collected from Crossbred Dairy Cattle Kept in High and Low Altitudes of Muheza District, Tanzania

Location	Variable	Mean	Standard deviation	Minimum	Maximum
Highland N = 203	Fat %	3.38	0.97	1.84	4.77
	Protein %	3.22	0.35	2.44	3.95
	Lactose %	3.61	0.29	3.19	4.11
	Density gm/cm3	1.0277	0.006	1.0146	1.0463
	pH	6.65	0.15	6.36	6.90
Lowland N = 197	Fat %	2.89	1.04	1.31	4.49
	Protein %	3.03	0.38	2.34	3.7
	Lactose %	3.56	0.25	3.22	4.12
	Density gm/cm3	1.0276	0.0049	1.0187	1.0402
	pH	6.79	0.12	6.60	6.91

Table II. Physiochemical Parameters of Raw Milk Collected from Highland and Lowland Dairy Cows

Parameters	Geographical location (Mean ± SE)
	Highlands (n = 203)	Lowlands (n = 197)	p-value
Fat %	3.38 ± 0.157	2.89 ± 0.161	0.031
Protean %	3.22 ± 0.057	3.03 ± 0.059	0.022
Lactose %	3.61 ± 0.043	3.56 ± 0.044	0.429
Density g/cm	1.028 ± 0.001	1.028 ± 0.001	0.880
pH	6.65 ± 0.022	6.79 ± 0.022	0.031

Fig. 1. Physicochemical parameters of raw milk by geographical location as shown in Fig. 1.

Table III. The Effects of Breed and Altitudes on Milk Physiochemical Properties

Breed	Location	Fat (%)	Protein (%)	Lactose (%)	pH	Density (%)
Holstein	Highland	2.64 ± 0.094 a	3.00 ± 0.046 a	3.60 ± 0.056 a	6.74 ± 0.022a	1.028 ± 0.001a
	Lowland	1.99 ± 0.100 b	2.75 ± 0.049 b	3.64 ± 0.060 a	6.81 ± 0.024b	1.028 ± 0.001a
Ayrshire	highland	3.85 ± 0.225 a	3.22 ± 0.110 a	3.90 ± 0.133 a	6.67 ± 0.053a	1.027 ± 0.003a
	Lowland	3.34 ± 0.259 b	2.96 ± 0.127 b	3.49 ± 0.154 a	6.87 ± 0.061b	1.027 ± 0.003a
Jersey	Highland	4.41 ± 0.318 a	3.49 ± 0.156 a	3.30 ± 0.189 a	6.54 ± 0.075a	1.033 ± 0.004a
	Lowland	4.37 ± 0.318 b	3.44 ± 0.156 b	3.52 ± 0.189 a	6.91 ± 0.075b	1.026 ± 0.004a
Guernsey	Highland	4.48 ± 0.130 a	3.61 ± 0.064 a	3.59 ± 0.077 a	6.48 ± 0.031a	1.027 ± 0.002a
	Lowland	3.87 ± 0.120 b	3.40 ± 0.059 b	3.48 ± 0.071 a	6.72 ± 0.028b	1.027 ± 0.001a

Table IV. The Effects of Altitude and Feed on Milk Physicochemical Characteristics

Category	Variable	Location	Fat (%)	Protein (%)	Lactose (%)	pH	Density (%)
Pasture type	Natural	Highland	3.15 ± 0.170	3.18 ± 0.065	3.59 ± 0.048	6.68 ± 0.023	1.028 ± 0.001
		Lowland	2.82 ± 0.179	2.82 ± 0.179	3.60 ± 0.051	6.78 ± 0.025	1.027 ± 0.001
	Natural + Improved	Highland	4.23 ± 0.321	4.23 ± 0.321	3.70 ± 0.091	6.53 ± 0.044	1.028 ± 0.002
		Lowland	3.08 ± 0.30	3.08 ± 0.304	3.47 ± 0.086	6.82 ± 0.042	1.029 ± 0.002
Supplements	Yes	Highland	3.43 ± 0.161	3.25 ± 0.058	3.62 ± 0.044	6.64 ± 0.022	1.028 ± 0.001
		Lowland	2.97 ± 0.210	3.07 ± 0.076	3.56 ± 0.058	6.81 ± 0.029	1.028 ± 0.001
	No	Highland	2.48 ± 0.712	2.81 ± 0.256	3.57 ± 0.196	6.8 ± 0.097	1.029 ± 0.004
		Lowland	2.77 ± 0.252	2.98 ± 0.091	3.58 ± 0.069	6.76 ± 0.034	1.027 ± 0.001

Table V. The Effects of Animal Characteristics and Altitudes on Milk Physiochemical Properties

Category	Variable	Location	Fat (%)	Protein (%)	Lactose (%)	pH	Density (%)
Parity	1	Highland	2.33 ± 0.332	2.90 ± 0.143	3.63 ± 0.108	6.75 ± 0.055	1.029 ± 0.002
		Lowland	1.54 ± 0.332	2.75 ± 0.143	3.81 ± 0.108	6.85 ± 0.055	1.024 ± 0.002
	2	Highland	3.95 ± 0.245	3.33 ± 0.106	3.71 ± 0.080	6.60 ± 0.041	1.025 ± 0.002
		Lowland	2.93 ± 0.182	3.05 ± 0.078	3.55 ± 0.059	6.76 ± 0.030	1.028 ± 0.001
	≥3	Highland	3.80 ± 0.218	3.29 ± 0.094	3.63 ± 0.071	6.62 ± 0.036	1.028 ± 0.001
		Lowland	3.45 ± 0.226	3.14 ± 0.097	3.47 ± 0.073	6.8 ± 0.038	1.029 ± 0.001
BCS	<2.5	Highland	2.43 ± 0.291	3.02 ± 0.125	3.41 ± 0.094	6.77 ± 0.047	1.029 ± 0.002
		Lowland	1.84 ± 0.291	2.90 ± 0.125	3.63 ± 0.094	6.77 ± 0.047	1.029 ± 0.002
	2.5–3.5	Highland	3.67 ± 0.153	3.25 ± 0.066	3.68 ± 0.049	6.61 ± 0.025	1.027 ± 0.001
		Lowland	3.33 ± 0.156	3.11 ± 0.067	3.53 ± 0.050	6.78 ± 0.025	1.027 ± 0.001
	>3.5	Highland	3.21 ± 0.412	3.45 ± 0.177	3.56 ± 0.133	6.71 ± 0.066	1.033 ± 0.003
		Lowland	1.61 ± 0.476	2.71 ± 0.205	3.75 ± 0.154	6.89 ± 0.077	1.027 ± 0.003
AGE	<4 years	Highland	2.59 ± 0.354	2.97 ± 0.146	3.68 ± 0.108	6.73 ± 0.055	1.027 ± 0.002
		Lowland	2.11 ± 0.306	2.87 ± 0.126	3.76 ± 0.094	6.83 ± 0.048	1.024 ± 0.002
	4–7 years	Highland	3.94 ± 0.185	3.28 ± 0.076	3.66 ± 0.057	6.60 ± 0.029	1.028 ± 0.001
		Lowland	3.25 ± 0.177	3.13 ± 0.073	3.49 ± 0.054	6.77 ± 0.028	1.028 ± 0.001
	>7 years	Highland	2.82 ± 0.240	3.25 ± 0.099	3.51 ± 0.074	6.70 ± 0.037	1.028 ± 0.001
		Lowland	2.52 ± 0.328	2.90 ± 0.135	3.58 ± 0.100	6.80 ± 0.051	1.032 ± 0.002

Discussion

Smallholder dairy farming using a zero-grazing system is common in smallholder farmers in Tanzania, where land for grazing is limited, with pasture obtained through cut-and-carry from roadsides and non-cultivated areas such as riverbanks, dam edges, flood plains, and fallow lands [34]. In this study, the cut-and-carry or zero-grazing system was common, practiced by 100% of farmers in the highlands and 90.7% of farmers in the lowlands. Cut-and-carry using natural pastures has also been reported by other researchers for the majority of smallholder farmers in Tanzania [34], [35]. The reason for the zero grazing system being practiced by 100% of farmers at high altitudes could be the geographical landscape, which does not support dairy grazing. Native pastures, such as Panicum, Brachiaria and Napier grasses, which are dominant in the study area, have the potential to increase cattle productivity in many areas of Africa, particularly in East Africa [36], [37]. Pastures are more common and provide the majority of pastures for smallholder dairy animals.

Determination of milk composition is important for determining the nutritive value, processability, and quality of the final products [38], [39]. Differences in the physiochemical properties of milk have been reported among different cow breeds and individual cows.

The same breed, and this is partially attributed to the genetic variations between cows, environmental and managemental factors [40], [41], [6]. In this study, milk components, such as fat and protein, were notably higher in milk collected from animals kept at high altitudes than in those kept at low altitudes. The opposite was true for milk pH, which was relatively lower in lowlands but higher in highlands. Interestingly, the milk density was not affected by altitude. Significant changes in some milk parameters with altitude have also been reported by other researchers [11]. 17 reported a higher fat content in milk from cows maintained at high altitudes than at lower altitudes, both of which were maintained in alpine pastures. Similarly, [19] reported higher fat and protein content in milk from high-altitude farms than from low-altitude farms. However, [18] reported higher fat but lower protein and lactose contents in milk from dairy cows exposed to higher altitudes than those at lower altitudes. Although the lactose content in the studied altitude zones was not statistically different (p > 0.05), a numerically higher lactose content at high altitudes compared to the low altitudes observed is in agreement with the findings reported by [11], who explored the effects of altitude on milk composition. The higher lactose content at high altitudes, although not statistically significant in this study, may be attributed to many factors, including high-altitude environmental stress, feed intake, breed adaptations, and management practices, all of which affect the physiological activities, amount, and quality of animal products; however, further studies are recommended to validate this observation.

Breed and genetic characteristics of the animal significantly affect the concentration and ratio of fat, protein, lactose, and milk [42], [43]. In this study, breed and parity significantly (p < 0.05) influenced protein content, where Guernsey cows produced milk with a higher protein content than other breeds, and multiparous cows produced milk with a higher protein content. Similar observations have been reported previously [44], [42]. Milk pH and acidity are indirect indicators of milk quality and udder health [45]–[47]. The overall mean milk pH reported in this study is within the range recommended by the FAO, WHO, and East African Standards [48], [16], [49], [50]. Although an insignificant difference was observed between cattle breeds, milk pH was higher in milk samples collected at low altitudes (6.79 ± 0.022) than in those collected at high altitudes (6.65 ± 0.022). [51] reported no significant differences between Norwegian cattle breeds, in contrast to [52], who reported significant differences in milk pH among different multi-breed dairy herds.

Conclusion

Notably, higher levels of milk properties, such as fat, protein, and lactose, were observed in milk obtained from high-altitude areas than in milk collected from low-altitude areas. However, the mean values of each milk component were within the range recommended by the FAO, WHO, and East African Standards. Notably, altitude alone did not selectively influence milk physicochemical properties, likely due to interactions with other factors such as animal diet and genetics. Thus, further research is required to clarify these mechanisms and evaluate their effects on milk quality in the dairy production chain.

Acknowledgments

The author wishes to express sincere gratitude to the Tanzania Livestock Research Institute (TALIRI) and TEAGASC through the Maziwa Faida project for their generous financial support, which made this study possible. Appreciation is also extended to Sokoine University of Agriculture (SUA), particularly the staff of the Department of Veterinary Surgery and Theriogenology, for their technical support and guidance throughout the research period.

Special thanks go to the Office of the District Executive Director of Muheza, through the Department of Agriculture, Livestock and Fisheries, for their invaluable assistance in facilitating access to farmers and field logistics.

Conflict of Interest

The authors declare that they do not have any conflict of interest.