Trop Anim Health Prod DOI 10.1007/s11250-014-0624-3
REGULAR ARTICLES
Effects of management practices on yield and quality of milk from smallholder dairy units in urban and peri-urban Morogoro, Tanzania Kejeri A. Gillah & George C. Kifaro & Jorgen Madsen
Accepted: 2 June 2014 # Springer Science+Business Media Dordrecht 2014
Abstract A longitudinal study design was used to assess the management, chemical composition of cows’ milk and quantify the microbial load of raw milk produced at farm level. Data were collected between December 2010 and September 2011 in Morogoro municipality. Milk samples were collected once every month and analysed for butter fat (BF), crude protein (CP), total solids (TS) and solids non-fat (SNF). Total bacterial count (TBC) and coliform counts (CC) were normalized by log transformation. The average milk yield was 7.0 l/ day and was not influenced by feeding systems and breeds. Dairy cows owned by people who had no regular income produced more milk than government employees and retired officers. Means of BF, TS, SNF and CP were similar in different feeding systems. Wet season had significantly higher TBC (5.9 log10 cfu/ml) and CC (2.4 log10 cfu/ml) but feeding systems had no effect. Stocking density influenced TBC but not CC. It can be concluded that dairy cows produced low milk yield and its quality was poor.
Keywords Coliform count . Feeding systems . Milk quality . Stocking density . Total bacterial count
K. A. Gillah (*) Ministry of Livestock Development and Fisheries, P.O. Box 9152, Dar es Salaam, Tanzania e-mail: [email protected] G. C. Kifaro Department of Animal Science and Production, Sokoine University of Agriculture, P.O. Box 3004, Morogoro, Tanzania J. Madsen Department of Large Animal Sciences, University of Copenhagen, Groennegardsvej 2, 1870, Frederiksberg Copenhagen, Denmark
Introduction Feeding strategies in urban and peri-urban areas of developing countries are diverse as a result of different factors such as availability of feed resources, social categories, type of animal species and household income (FAO 2001). These factors may influence milk yield and its composition. Milk is among the most valuable and nutritious food for humans and young mammals but, improper handling contaminates it with microorganisms and renders it unsafe for human consumption or unfit for processing (Ali and Abdelgadir 2011). Generally, fresh milk drawn from a healthy cow contains less than 103 cfu/ml bacteria load, but once it is stored at ambient (30 to 35 °C) temperature, the bacterial load may increase tremendously (Lingathurai et al. 2009). The microbial load of milk determines quality and shows the hygienic level in the milk production chain. Factors such as use of unclean milking utensils, poor hygiene of cattle shed, and lack of good quality water for cleaning purposes influence the number and types of microbes in milk immediately after milking (Ali and Abdelgadir 2011). Understanding the microbial load of milk right from the producers and factors that may influence the levels of microbial load is very important to safeguard milk consumers. However, information on variations of milk yield, chemical composition, and microbiological qualities of raw milk in relation to feeding systems, breeds, parity, and stage of lactation of crossbred cattle in periurban areas is scant. The present study was designed to assess the effect of management practices on milk yield and quality of raw milk produced in urban and periurban dairy units of Morogoro, Tanzania.
Trop Anim Health Prod
Materials and methods Study area Data were collected between December 2010 and September 2011 in Morogoro municipality (4 ° 49′ S and 37 ° 40′ E). Altitude is about 528 m, rainfall is bimodal with 800– 1,200 mm per annum, and mean maximum and minimum temperatures are 32.4 and 14.8 °C, respectively (URT 2007a). Sampling procedure and description of the feeding systems A longitudinal study design, which covered the wet (January to May 2011) and dry (June to September 2011) seasons, was adopted and involved 60 smallholder dairy farmers who had at least one lactating cow and practiced zero or partial zero feeding systems. Milk yield was recorded from 123 cows partitioned into three stages of lactation: early (7 to 90 days), mid (91 to 180 days) and late (181 to 210 days). Cattle sheds floor areas were measured to determine stocking densities, which were calculated by dividing the shed floor area by 6.7 m2, the standard area required by one cow (FAO 1998). The cattle sheds, which were regularly cleaned with water or manure removal by scraping, were assessed as being in good and poor hygienic conditions, respectively. Cows were hand milked in the morning and evening and kept in sheds using either zero or partial zero feeding systems. Zero feeding is a system in which cattle do not graze in the fields but are totally confined in a shed where feed and water are brought to them while in partial zero feeding, cattle are allowed to graze on the field for some hours and given supplementary feeds when confined in the sheds. Vegetations from natural pastures were the major source of feeds while concentrates were used as supplementary feeds to lactating cows during milking time. A structured questionnaire was used to obtain information on types of farm labour, feeds, milking hygiene and practices. Dairy units were regularly visited in order to assess the hygienic condition of the cattle sheds, collect milk and feed samples. Milk samples collection and chemical determination Milk was recorded once a week and morning milk samples were collected every 2 months, kept in ice packed cool boxes and transported to Sokoine University of Agriculture laboratory, where microbial count was done within 24 h. Part of the milk samples were deep frozen in the laboratory pending further analysis. Milk samples were analysed for butter fat percent using Gerber method while crude protein was calculated from nitrogen content from Kjeldahl method analysis (AOAC 1995) multiplied by a factor of 6.38. The percentage total solids were determined by the reference method of oven drying at 105 °C
in acid-washed sand for 12 h. Solids that were not fat were calculated as the difference between total solids and butter fat percentages. The grading of butter fat percentage was based on the Tanzania Dairy Industry Regulations (URT 2007b). Determination of total bacterial count and Coliform count The test procedure for total bacterial count followed the guidance given by International Dairy Federation standard methods (IDF 1992). Normal saline was prepared by dissolving 8.5 g of sodium chloride in 1 l of distilled water. Then, 9 ml of normal saline was transferred in several 20-ml test tubes and securely stoppered with cotton wool and aluminium foil for sterilization. Diluents were sterilized in an autoclave at 121 °C for 15 min, After sterilization, the tubes were removed from the autoclave, left to cool at room temperature and kept in the refrigerator. About 28 g of Nutrient agar (oxoid) was dissolved in 1 l of distilled water, to prepare agar media and autoclaved at 121 °C for 15 min and then left to cool to 45 °C. For each raw milk sample, tenfold (10−1–105) serial dilutions were prepared in a sterile normal saline solution using a sterile pipette and tips. One millilitre of 103 to 105 dilution of milk was pipetted into well-labelled and sterilized petri dishes in duplicates. Then, 20 ml of sterile standard plate count agar was poured into each petri dish, mixed thoroughly by gently tilting, and left to solidify on a level surface. The plates were then turned upside down and incubated at 37 °C for 48 h. Countable colonies between 30 and 300 colony forming units (cfu) per plate were chosen for counting with an aid of colony counter and computed following guidelines by Maturin and Peeler (2001). Total bacterial counts of the milk were interpreted and graded according to the Tanzania Bureau of Standards guidelines (TBS 1983). Violet red bile agar (VRBA) medium was prepared by dissolving 41.5 g of VRBA powder in 1,000 ml of distilled water and mixed thoroughly. The mixture was then heated in a Pyrex bottle with frequent agitation to dissolve the powder and after that, the mixture was cooled to 45 °C. It was then poured into sterile petri dishes containing the sample and left to solidify. After cooling and solidification, VRBA plates were incubated at 37 °C for 24 h. All samples were cultured in duplicate and VRBA plate with countable dark red colonies (Hall et al. 1967), between 15 and 150 cfu⁄plate were chosen for counting with the aid of a colony counter. Coliform counts were interpreted and graded according to Tanzania Bureau of Standards guidelines (TBS 1983). Statistical analyses Total bacteria count and coliform count data were converted into logarithm of the number of colony forming units per millilitre of raw cow milk samples (log10 cfu/ml). Descriptive statistics of general linear model (GLM) of SAS (2002) were
Trop Anim Health Prod
used to describe herd size and composition, stocking density, milk yield and qualities. Differences in sample sizes among continuous variables were taken care of by using GLM for generalized least squares procedures. Chi-square test was used to analyse associations between possible combinations of categorical variables and values of P
REGULAR ARTICLES
Effects of management practices on yield and quality of milk from smallholder dairy units in urban and peri-urban Morogoro, Tanzania Kejeri A. Gillah & George C. Kifaro & Jorgen Madsen
Accepted: 2 June 2014 # Springer Science+Business Media Dordrecht 2014
Abstract A longitudinal study design was used to assess the management, chemical composition of cows’ milk and quantify the microbial load of raw milk produced at farm level. Data were collected between December 2010 and September 2011 in Morogoro municipality. Milk samples were collected once every month and analysed for butter fat (BF), crude protein (CP), total solids (TS) and solids non-fat (SNF). Total bacterial count (TBC) and coliform counts (CC) were normalized by log transformation. The average milk yield was 7.0 l/ day and was not influenced by feeding systems and breeds. Dairy cows owned by people who had no regular income produced more milk than government employees and retired officers. Means of BF, TS, SNF and CP were similar in different feeding systems. Wet season had significantly higher TBC (5.9 log10 cfu/ml) and CC (2.4 log10 cfu/ml) but feeding systems had no effect. Stocking density influenced TBC but not CC. It can be concluded that dairy cows produced low milk yield and its quality was poor.
Keywords Coliform count . Feeding systems . Milk quality . Stocking density . Total bacterial count
K. A. Gillah (*) Ministry of Livestock Development and Fisheries, P.O. Box 9152, Dar es Salaam, Tanzania e-mail: [email protected] G. C. Kifaro Department of Animal Science and Production, Sokoine University of Agriculture, P.O. Box 3004, Morogoro, Tanzania J. Madsen Department of Large Animal Sciences, University of Copenhagen, Groennegardsvej 2, 1870, Frederiksberg Copenhagen, Denmark
Introduction Feeding strategies in urban and peri-urban areas of developing countries are diverse as a result of different factors such as availability of feed resources, social categories, type of animal species and household income (FAO 2001). These factors may influence milk yield and its composition. Milk is among the most valuable and nutritious food for humans and young mammals but, improper handling contaminates it with microorganisms and renders it unsafe for human consumption or unfit for processing (Ali and Abdelgadir 2011). Generally, fresh milk drawn from a healthy cow contains less than 103 cfu/ml bacteria load, but once it is stored at ambient (30 to 35 °C) temperature, the bacterial load may increase tremendously (Lingathurai et al. 2009). The microbial load of milk determines quality and shows the hygienic level in the milk production chain. Factors such as use of unclean milking utensils, poor hygiene of cattle shed, and lack of good quality water for cleaning purposes influence the number and types of microbes in milk immediately after milking (Ali and Abdelgadir 2011). Understanding the microbial load of milk right from the producers and factors that may influence the levels of microbial load is very important to safeguard milk consumers. However, information on variations of milk yield, chemical composition, and microbiological qualities of raw milk in relation to feeding systems, breeds, parity, and stage of lactation of crossbred cattle in periurban areas is scant. The present study was designed to assess the effect of management practices on milk yield and quality of raw milk produced in urban and periurban dairy units of Morogoro, Tanzania.
Trop Anim Health Prod
Materials and methods Study area Data were collected between December 2010 and September 2011 in Morogoro municipality (4 ° 49′ S and 37 ° 40′ E). Altitude is about 528 m, rainfall is bimodal with 800– 1,200 mm per annum, and mean maximum and minimum temperatures are 32.4 and 14.8 °C, respectively (URT 2007a). Sampling procedure and description of the feeding systems A longitudinal study design, which covered the wet (January to May 2011) and dry (June to September 2011) seasons, was adopted and involved 60 smallholder dairy farmers who had at least one lactating cow and practiced zero or partial zero feeding systems. Milk yield was recorded from 123 cows partitioned into three stages of lactation: early (7 to 90 days), mid (91 to 180 days) and late (181 to 210 days). Cattle sheds floor areas were measured to determine stocking densities, which were calculated by dividing the shed floor area by 6.7 m2, the standard area required by one cow (FAO 1998). The cattle sheds, which were regularly cleaned with water or manure removal by scraping, were assessed as being in good and poor hygienic conditions, respectively. Cows were hand milked in the morning and evening and kept in sheds using either zero or partial zero feeding systems. Zero feeding is a system in which cattle do not graze in the fields but are totally confined in a shed where feed and water are brought to them while in partial zero feeding, cattle are allowed to graze on the field for some hours and given supplementary feeds when confined in the sheds. Vegetations from natural pastures were the major source of feeds while concentrates were used as supplementary feeds to lactating cows during milking time. A structured questionnaire was used to obtain information on types of farm labour, feeds, milking hygiene and practices. Dairy units were regularly visited in order to assess the hygienic condition of the cattle sheds, collect milk and feed samples. Milk samples collection and chemical determination Milk was recorded once a week and morning milk samples were collected every 2 months, kept in ice packed cool boxes and transported to Sokoine University of Agriculture laboratory, where microbial count was done within 24 h. Part of the milk samples were deep frozen in the laboratory pending further analysis. Milk samples were analysed for butter fat percent using Gerber method while crude protein was calculated from nitrogen content from Kjeldahl method analysis (AOAC 1995) multiplied by a factor of 6.38. The percentage total solids were determined by the reference method of oven drying at 105 °C
in acid-washed sand for 12 h. Solids that were not fat were calculated as the difference between total solids and butter fat percentages. The grading of butter fat percentage was based on the Tanzania Dairy Industry Regulations (URT 2007b). Determination of total bacterial count and Coliform count The test procedure for total bacterial count followed the guidance given by International Dairy Federation standard methods (IDF 1992). Normal saline was prepared by dissolving 8.5 g of sodium chloride in 1 l of distilled water. Then, 9 ml of normal saline was transferred in several 20-ml test tubes and securely stoppered with cotton wool and aluminium foil for sterilization. Diluents were sterilized in an autoclave at 121 °C for 15 min, After sterilization, the tubes were removed from the autoclave, left to cool at room temperature and kept in the refrigerator. About 28 g of Nutrient agar (oxoid) was dissolved in 1 l of distilled water, to prepare agar media and autoclaved at 121 °C for 15 min and then left to cool to 45 °C. For each raw milk sample, tenfold (10−1–105) serial dilutions were prepared in a sterile normal saline solution using a sterile pipette and tips. One millilitre of 103 to 105 dilution of milk was pipetted into well-labelled and sterilized petri dishes in duplicates. Then, 20 ml of sterile standard plate count agar was poured into each petri dish, mixed thoroughly by gently tilting, and left to solidify on a level surface. The plates were then turned upside down and incubated at 37 °C for 48 h. Countable colonies between 30 and 300 colony forming units (cfu) per plate were chosen for counting with an aid of colony counter and computed following guidelines by Maturin and Peeler (2001). Total bacterial counts of the milk were interpreted and graded according to the Tanzania Bureau of Standards guidelines (TBS 1983). Violet red bile agar (VRBA) medium was prepared by dissolving 41.5 g of VRBA powder in 1,000 ml of distilled water and mixed thoroughly. The mixture was then heated in a Pyrex bottle with frequent agitation to dissolve the powder and after that, the mixture was cooled to 45 °C. It was then poured into sterile petri dishes containing the sample and left to solidify. After cooling and solidification, VRBA plates were incubated at 37 °C for 24 h. All samples were cultured in duplicate and VRBA plate with countable dark red colonies (Hall et al. 1967), between 15 and 150 cfu⁄plate were chosen for counting with the aid of a colony counter. Coliform counts were interpreted and graded according to Tanzania Bureau of Standards guidelines (TBS 1983). Statistical analyses Total bacteria count and coliform count data were converted into logarithm of the number of colony forming units per millilitre of raw cow milk samples (log10 cfu/ml). Descriptive statistics of general linear model (GLM) of SAS (2002) were
Trop Anim Health Prod
used to describe herd size and composition, stocking density, milk yield and qualities. Differences in sample sizes among continuous variables were taken care of by using GLM for generalized least squares procedures. Chi-square test was used to analyse associations between possible combinations of categorical variables and values of P
