i OPTIMISING THE WAGASHIE (A TRADITIONAL COTTAGE CHEESE) PROCESS AND SENSORY QUALITY BY AKUA BOATEMAA ARTHUR 10443436 THIS THESIS IS SUBMITTED TO THE UNIVERSITY OF GHANA, LEGON, IN PARTIAL FULFILMENT OF THE REQUIREMENT FOR THE AWARD OF MASTER OF PHILOSOPHY IN RADIATION PROCESSING DEGREE (FOOD SCIENCE AND POST HARVEST TECHNOLOGY) DEPARTMENT OF NUCLEAR AGRICULTURE AND RADIATION PROCESSING, SCHOOL OF NUCLEAR AND ALLIED SCIENCES UNIVERSITY OF GHANA July, 2016 University of Ghana http://ugspace.ug.edu.gh ii DECLARATION This thesis is the result of the research work undertaken by Akua Boatemaa Arthur in the department of nuclear agriculture and radiation proessing (N.A.R.P.) of the Graduate Shool of Nuclear and Allied Sciences (GSNAS), University of Ghana, Legon under the supervision of Prof. Wisdom kofi Amoa-Awua and Prof. Victoria Appiah. Except for the references of other research works which have been duly cited, in this dissertation, this work has never been presented either in whole or in part for any other degree in this University or elsewhere. Signature......................................... Date........................................... Akua Boatemaa Arthur (Student) Certified by: Signature......................................... Date............................................... Prof. Wisdom Amoa-Awua (SUPERVISOR) Signature.......................................... Date................................................... Prof. Victoria Appiah (SUPERVISOR) University of Ghana http://ugspace.ug.edu.gh iii DEDICATION I dedicate this work to God almighty, my parents; Mr. and Mrs. Osei-Bonsu, my children; Ebo and Nana, my super supportive sibblings (Serwah and Nketia) and my husband; Ing Philip Bernard Arthur. University of Ghana http://ugspace.ug.edu.gh iv ACKNOWLEDGEMENT My unending gratitude goes to God almighty for his umlimited grace and favour throughout my study. He is my source of inspiration. Also to my parents, Mr and Mrs Osei-Bonsu, my husband; Ing Philip Bernard Arthur and my siblings; Afia Serwah Bonsu and Nketiah Osei- Bonsu, you were very supportive and i am very grateful. This work was carried out as part of the Danida funded project ‘Preserving African food microorganisms for Green Growth. My profound gratitude therefore goes to the green growth team for giving me the opportunity to carry out this project and for the experience acquired. My sincere gratitude goes to my supervisors, Dr Wisdom Kofi Amoa-Awua and Prof. Victoria Appiah for their encouragement, support and guidance throughout my study. My gratitude also goes to Dr. Margaret Owusu, Mrs. Amy Atter and Mr. Theophilus Annan at the Food Research Institute -CSIR for their assistance. I am grateful to Prof. Saalia for his assistance in my experimental design and analysis. To Mrs. Charlottte Oduro - Yeboah and Mr. Papa Toah, I am grateful for their enormous support and assistance. I am also grateful to Dr. Abbey for his assistance during my instrumental texture analysis. My gratitude goes to all the laboratory technicians at the Food microbiology laboratory, Nutrition test kitchen, Chemistry laboratory and the Food Processing laboratory at the Food Research Institute –CSIR for their help and assistance especially, Alex, Yahayah, Auntie Constance Boateng, Auntie Joyce, Auntie Alice, Edna, Solomon, Jemima, Mr Ameh, Frank and all the service and attachment persons that were present at the time of the study. University of Ghana http://ugspace.ug.edu.gh v My gratitude also goes to all the staff of Food Research Institute (CSIR) who took part in the consumer and sensory evaluation for the project, thank you for your time and enthusiasm that you exhibited throughout the activity. Finally my gratitude goes to Mr Adu Gyamfi (BNARI) and the technicians at the Radiation Technology Center (RTC-BNARI) for their help especially, Stanely and Jonathan. University of Ghana http://ugspace.ug.edu.gh vi ABSTRACT Wagashie is a traditional West African cottage cheese produced by the Fulani who are semi- nomadic. It is a good protein source and can replace fish or meat in the diet of low income families in Africa. However, it is a product with high moisture content (60%) which is favourable for the growth of microorganisms and thus has a short shelf life of 3 days; it also has a bland taste with limited patronage. This research was therefore carried out to reengineer wagashie for a larger market with a focus on improving its sensory quality, safety and shelf life. A brief survey was carried out to confirm the wagashie production procedure and identify retailers and producers for collection of samples. The safety of market samples of fresh and fried wagashie samples were determined by assaying for various indicator and pathogenic microorganisms including aerobic mesophiles ,Yeast and moulds , coliform bacteria, E. coli, Staphylococcus aureus Bacillus cereus, Salmonella spp ,Enterococcus, Enterobacteriaceae. Studies were also carried out to replace the traditional coagulant of milk which involves the use of plant extract of Sodom apple (Calotropis procera) with commercial rennet used in industrial cheese production and ferment fresh milk used in the preparation. The traditional method of preparation was also standardised to improve its sensory quality. The process variables of wagashie, which are salt concentration, coagulant and fermentation time, were thus optimised using the Box Behnken design which is a response surface methodology and an affective testing was carried out to evaluate the consumer preference of the product with a nine-point hedonic scale. The sensory profile of the ‘wagashie’ samples were described by a Quantitative Descriptive Analysis by a trained 13 member panel. They evaluated the wagashie samples for desirable and undesirable attributes of wagashie. The rheology of ‘wagashie’ which involves Texture Profile Analysis (TPA) with a texture analyser and colour determination using a chroma photometer were carried out on the improved product which were the rennet coagulated fresh and smoked wagashie samples. University of Ghana http://ugspace.ug.edu.gh vii Chemical analyses were carried out on the samples whereby the protein, ash, fat, free fatty acids (FFA), moisture, pH and Titratable acididty (TTA) of wagashie were determined. The safety of the laboratory prepared wagashie was assessed and shelf life studies were carried out for 5 weeks. The results of the microbiological tests carried out on the market wagashie showed that, Salmonella and Staphylococcus aureus were not detected in the samples whiles Bacillus cereus was detected in low counts in half of the samples. The rest of the microorganisms were found in fairly high counts. The optimum levels of the process variables which was used in standardizing the product were 23 g of salt, 150 g of plant extract and 0 h fermentation for the traditional preparation and 11 g of salt, 5.35ml of commercial rennet and 4 h fermentation for the improved wagashie The results of the consumer preference testing showed that the panelists preferred the non- fermented product to the fermented wagashie for the traditional preparation (Sodom apple extract as coagulant). The same panelist however preferred the fermented product to the non- fermented product for the preparation with commercial rennet as coagulant. After a confirmatory affective test where the wagashie samples were processed by frying and smoking, the panelists rated the acceptability of the traditional non- fermented smoked and the fermented rennet coagulated fried sample significantly higher at p<0.05, followed by the fermented rennet coagulated smoked sample. However, all three samples were rated ‘like- moderately’ on the 9-point hedonic scale. Thus the fresh and smoked samples were considered for the rest of the study due to health concerns raised by consumers which involved the reduction of fat in wagashie. University of Ghana http://ugspace.ug.edu.gh viii In the Quantitative Descriptive Analysis, the panellists scored higher intensities of the desirable attributes in the improved wagashie which were mainly milky aroma, milky taste, cheesy aroma, cheesy taste, yoghurt aroma. The undesirable descriptors which were bitter taste, bland taste, spoilt milk aroma, fermented cassava dough aroma were rated high in the market samples. Generally, there were no significant differences between the plant extract coagulated samples and the rennet coagulated samples prepared in the laboratory. The Texture profile analysis showed that the wagashie samples were ‘hard’, ‘gummy’ and ‘chewable’. The instrumental colour determination showed that the fresh wagashie sample had higher L* values which indicated a lighter colour while the smoked samples had lower L* values which indicated a darker colour which corresponded to the results of the Quantitative Descriptive analysis of the fresh and smoked wagashie samples. The chemical analysis showed that the rennet coagulated smoked sample had the highest protein content of 30.18 g/100g and the highest FFA value of 0.53. The fried samples had the highest fat content of 25.32 g and the highest ash content of 2.0 g and the fermented fresh samples had the highest moisture content of 56.10 g. Generally the FFA values for the samples were low. The safety of the improved wagashie samples was improved with fermentation and smoking, this is because the count for microorganisms reduced when compared with the safety of the market wagashie.The result of the shelf life study showed that preserving the wagashie samples with vacuum packaging and irradiation extended the shelf life of wagashie from 3 days to 3 weeks of storage under ambient conditions. University of Ghana http://ugspace.ug.edu.gh ix TABLE OF CONTENTS DELARATION .......................................................................................................................... ii DEDICATION .......................................................................................................................... iii ACKNOWLEDGEMENT ........................................................................................................ iv ABSTRACT .............................................................................................................................. vi LIST OF TABLES ................................................................................................................... xii LIST OF FIGURES ................................................................................................................. xiv CHAPTER ONE ........................................................................................................................ 1 1.0 INTRODUCTION ................................................................................................................ 1 1.1.0 Justification .................................................................................................................... 4 1.2 Main Objecive .................................................................................................................. 4 CHAPTER TWO ........................................................................................................................ 5 2.0 LITERATURE REVIEW ..................................................................................................... 5 2.1 Traditional food Processing .............................................................................................. 5 2.2.0 Traditional production of ‘wagashie’ a local cottage cheese ........................................ 6 2.2.1 Preservation of wagashie ` ............................................................................................ 7 2.2.2 Safety of wagashie......................................................................................................... 8 2.2.3 Spoilage and Pathogenic microorganisms in soft cheeses (‘wagashie’) ....................... 9 2.2.4 Calotropis procera ........................................................................................................ 10 2.3.0 Milk production in Ghana............................................................................................ 11 2.3.1 Milk consumption in Ghana ........................................................................................ 12 2.3.2 Traditional Fermented milk products in Ghana ........................................................... 12 2.3.3 Milk composition ......................................................................................................... 14 2.3.4 Milk Quality and Safety .............................................................................................. 17 2.3.6 Lactic acid fermentation of milk ................................................................................. 20 2.4.0 Cheese .......................................................................................................................... 21 2.5.0 Cheese Manufacture .................................................................................................... 29 2.5.7 pH of cheese ................................................................................................................ 35 2.6.0 Rheology of cheese ...................................................................................................... 36 2.7.0 Food Product Development ......................................................................................... 38 2.8.0 Packaging and Preservation of cheese ......................................................................... 39 2.8.1 Food Irradiation ........................................................................................................... 41 2.9.1 Descriptive Sensory Analysis ...................................................................................... 43 2.9.6 Principal Composite Analysis (PCA) .......................................................................... 47 University of Ghana http://ugspace.ug.edu.gh x 2.10.0 Experimental Design ................................................................................................. 48 2.10.1 Response Surface Methodology and Box Behnken Design ...................................... 48 2.10.2 Box Behnken Design ................................................................................................. 49 CHAPTER THREE .................................................................................................................. 51 3.0 MATERIALS AND METHODS ....................................................................................... 51 3.1 Breif field study .............................................................................................................. 51 3.2 Sampling of wagashie ..................................................................................................... 51 3.3 Microbiological analysis of the market samples and laboratory prepared samples of wagashie ............................................................................................................................... 51 3.4.3 Modification of the Wagashie process ........................................................................ 58 3.5.0 Fermentation of fresh cow milk using starter culture .................................................. 60 3.6.0 Design of Experiment for ‘wagashie’ preparation ...................................................... 62 3.7.0 Sensory Evaluation ...................................................................................................... 66 3.7.1 Hedonic Sensory Evaluation ....................................................................................... 66 3.7.2 Quantitative Descriptive Analysis ............................................................................... 67 3.8.0 Physicochemical Analyses .......................................................................................... 68 3.8.1 Determination of pH .................................................................................................... 69 3.8.2 Determination of Total Titratable acidity .................................................................... 69 3.8.3 Determination of protein ............................................................................................. 70 3.8.5 Determination of FFA ................................................................................................. 71 3.8.6 Moisture determination................................................................................................ 71 3.8.7 Determination of Ash content of the samples ............................................................. 72 3.8.8 Colour measurement of ‘wagashie’ ............................................................................. 72 3.8.9 Texture Profile Analysis (TPA) of ‘wagashie’ ............................................................ 72 3.9.0 Shelf life study of wagashie ............................................................................................ 73 3.9.1 Irradiation and Packaging of wagashie ........................................................................ 73 3.9.2 Storage of ‘wagashie’ .................................................................................................. 74 3.10 Statistical Analysis ....................................................................................................... 74 CHAPTER FOUR .................................................................................................................... 76 4.0 RESULTS ........................................................................................................................... 76 4.1.0 Market Survey ............................................................................................................. 76 4.1.1 Microrganisms present in market ‘wagashie’ samples ................................................ 76 4.2.2 pH of the market wagashie samples. ........................................................................... 78 University of Ghana http://ugspace.ug.edu.gh xi 4.2.0 Fermentation trials for laboratory preparation of ‘wagashie’Rate of fermentation of raw cow milk with cheese and yoghurt starter cultures........................................................ 78 4.3.0 Optimisation of the ‘wagashie’ process....................................................................... 81 4.3.1.0 Using the response surface methodology to optimise the ‘wagashie’ process prepared with plant extract as coagulant. ............................................................................. 81 4.3.2.0 Using the Response Surface methodology to optimise the wagashie process using commercial rennet as coagulant. .......................................................................................... 89 4.3.0 Sensory Evaluation .................................................................................................... 100 4.3.1 Affective Sensory Evaluation .................................................................................... 100 4.3.2 Quantitative Descriptive Sensory Evaluation ............................................................ 102 4.4.0 Pricipal composite analysis (PCA) ............................................................................ 108 4.5.0 Cluster Analysis ......................................................................................................... 111 4.5.1 Spider plot.................................................................................................................. 114 4.6.0 Proximate composition of ‘wagashie’ ....................................................................... 115 4.8.0 Rheology of Improved ‘wagashie’ ............................................................................ 116 4.9.0 Safety of improved ‘wagashie’ .................................................................................. 119 4.9.1 Shelf life of wagashie ................................................................................................ 120 4.9.4 pH of the wagashie samples during the 5 weeks storage period. .............................. 124 CHAPTER FIVE .................................................................................................................... 126 5.0 DISCUSSION .................................................................................................................. 126 5.1 Safety of wagashie ........................................................................................................ 126 5.2 Improving the quality of wagashie ............................................................................... 129 5.2.1 Affective testing ........................................................................................................ 129 5.2.2 Quantitative Descriptive Sensory Analysis ............................................................... 132 5.3 Chemical analysis of wagashie samples ....................................................................... 134 5.4 Rheology of wagashie .................................................................................................. 136 5.4.1 Colour Determination ................................................................................................ 136 5.4.2 Texture Profile Analysis (TPA) ................................................................................. 136 5.5 Shelf life of wagashie ................................................................................................... 137 5.5.1 pH of wagashie samples during the 5 weeks storage period .................................... 139 CHAPTER SIX ...................................................................................................................... 140 6.0 CONCLUSIONS AND RECOMMENDATIONS........................................................... 140 REFERENCES ....................................................................................................................... 142 APENDDIX ........................................................................................................................... 161 University of Ghana http://ugspace.ug.edu.gh xii LIST OF TABLES Table 2.1: The most commonly used rennet and coagulants and their enzymes 28 Table 2.2: Classification of cheese 38 Table 2.3: The Box Behnken Experimental Design 50 Table 3.1: The Box Behnken Design matrix of variables (k=3) for the optimisation of ‘wagashie’ coagulated with Rennet 63 Table 3.2: Coded and actual levels of the factors for three levels Box Behnken design for ‘wagashie’ using plant extract as coagulant 63 Table 3.3: The Box Behnken Design matrix of variables (k=3) for the optimisation of ‘wagashie’ coagulated with plant extract 64 Table 3.4: The Box Behnken Design matrix of variables (k=3) for the optimisation of ‘wagashie’ coagulated with Rennet (ml) 65 Table 4.1: Mean microbial count on fresh and fried market wagashie samples in g/CFU 77 Table 4.2: Effect of deep frying and aseptic packaging on the microbial count of fresh wagashie obtained from Nima in g/CFU. 77 Table 4.3: Mean scores for the confirmatory affective sensory evaluation 101 Table 4.4: Mean pH and TTA values of the optimised ‘wagashie’ samples after confirmatory affective sensory evaluation 103 Table 4.5: Mean values for wagashie descriptors during the Quantitative Descriptive Analysis 108 University of Ghana http://ugspace.ug.edu.gh xiii Table 4.6: Mean values for the chemical composition of wagashie 116 Table 4.7: Mean values with standard deviations for the colour of improved wagashie 117 Table 4.8: Mean values for the textural characteristics of ‘wagashie’ 119 Table 4.9: Mean microbial counts in the improved ‘wagashie’ in CFU/g 121 Table 4.10: Changes in the mean microbial counts in the rennet coagulated fermented fresh and smoked ‘wagashie’ samples as affected by packaging and irradiation for the 5 weeks storage period in CFU/g (from day 0 to week 2) 123 Table 4.11: Mean pH values of the ‘wagashie’ samples during the 5 weeks storage period 125 University of Ghana http://ugspace.ug.edu.gh xiv LIST OF FIGURES Figure 3.1: Flow diagram for plant extract preparation (coagulant) 56 Figure 3.2 Flow diagram of wagashie production process (Traditional method) 57 Figure 3.3: Flow diagram of wagashie process with Plant extract and rennet for both non- fermented and fermented preparations with cheese and yoghurt cultures 59 Figure 4.1: Mean pH of ‘wagashie’ sampled from the market 78 Figure 4.2: The rate of pH change after 24 hours fermentation in a water bath set at 45oC with 5mls and 10mls of cheese culture 79 Figure 4.3: The rate of pH change after 24 h fermentation of 2 L of pastuerized fresh cow milk in a water bath set at 45oC with 10 ml and 10 ml of cheese culture. 80 Figure 4.4: Response Surface plot representing the effect of Fermentation time and Salt on the score for Texture when the weight of extract is 150 g 82 Figure 4.5: Response Surface plot representing the effect of fermentation time and Salt on the score for Colour when the Extract is 150 g 84 Figure 4.6: Response Surface plot representing the effect of Salt and Fermentation time on the score for taste. 85 Figure 4.7: Response surface plot representing the effect of salt and fermentation time on the score for Overall acceptability. 86 Figure 4.8: Contour plot for texture, taste, colour and overall acceptability of wagashie overlaid on one axis of fermentation time (h) and salt (g) 88 University of Ghana http://ugspace.ug.edu.gh xv Figure 4.9: Mean pH values for wagashie samples coagulated with plant extract combined by Box Behnken Design 89 Figure 4.10: The TTA of wagashie samples combined by Box Behnken Design using plant extract as cogulant 90 Figure 4.11: Response Surface plots representing the effect of salt and Fermentation time on the score for Texture when Rennet is 0.27%. 92 Figure 4.12: Response Surface plots representing the effect of salt and Fermentation time on the score for colour when Rennet is 0.27%. 94 Figure 4.13: Response Surface plots representing the effect of salt and Fermentation time on the score for Taste when Rennet is 0.27%. 95 Figure 4.14: Response Surface plot representing the effect of salt and Fermentation time on the score for overall acceptability when Rennet is 0.27%. 97 Figure 4.15: Contour plot for texture, taste, colour and overall acceptability of wagashie overlaid on one axis of fermentation time and salt at a constant rennet concentration of 5.35ml. 98 Figure 4.16: The mean pH of wagashie samples combined by Box Behnken Design using Rennet as caogulant 99 Figure 4.17: The TTA of wagashie samples combined by Box Behnken Design using Rennet as cogulant 100 Figure 4.18: PCA bi-plot of quantitative descriptive sensory used to describe the sensory attributes of wagashie in their fresh, fried and roasted forms. 110 University of Ghana http://ugspace.ug.edu.gh xvi Figure 4.19: Dendogram from cluster analysis of wagashie samples considering sensory attributes 113 Figure 4.20: Spider plot of the fresh and processed wagashie samples after the quantitative descriptive sensory evaluation 114 University of Ghana http://ugspace.ug.edu.gh 1 CHAPTER ONE 1.0 INTRODUCTION The slow progress in upgrading traditional food processing and preservation techniques in Ghana contributes to food and nutrition insecurity. Simple, low-cost, traditional food processing techniques are the bedrock of small-scale food processing enterprises that are crucial to rural development. By generating employment opportunities in the rural areas, small scale food industries reduce rural-urban migration and the associated social problems. They are vital to reducing post-harvest food losses and increasing food availability. However, rapid growth and development of small-scale food industries in Ghana are hampered by adoption of inefficient and inappropriate technologies, poor management, inadequate working capital and low profit margins. Some successes have however been achieved in upgrading traditional food processing technologies in Ghana including the production of gari and agbelema from cassava, the production of kenkey and banku from maize, the production of dawadawa (a fermented condiment) , production of fura from millet and the traditional cheese-making (wagashie) process from fresh cow milk. The scientific study of traditional cheese making offers a growing understanding of the inherent nature, strength and limitations of traditional food processing and preservation techniques. Cheese-making is one of the oldest methods of preserving excess milk and is a major business worth billions of dollars in many industrialized countries. Cheeses are now unique products in their own right and cheese-making has advanced beyond being merely a food preservation technique. ‘Wagashie’ production thrives mainly in the peri-urban milk producing areas where it provides employment mainly to women and increases the income of fresh cow milk sellers. A higher demand for traditional soft cheese (wagashie) increased income of milk sellers by 54% in Ghana (ILRI, 2006). The terminology with which this product is called has seen some University of Ghana http://ugspace.ug.edu.gh 2 variations in West Africa; it is called woagachi (O‘Connor, 1993) and wara or warankashi (Ogundiwin, 1978) by the people of Benin Republic and Nigeria respectively. The process of making ‘wagashie’ was developed by the nomadic Fulani as a means of preserving excess milk and is based on the milk coagulating properties of the juice from the leaves and stems of the Sodom apple plant (Calotropis procera). The juice which is obtained by crushing Sodom apple leaves or stems is mixed with raw cow milk gently heated in a pot over fire wood or water. The leaves and stems of the Sodom apple plant contain an organic acid called calotropin which has the ability to coagulate milk. Following coagulation, the loose curd pieces are poured into small raffia baskets and allowed to drain from the whey. Others produce ‘wagashie’ by storing milk in the abomasum of slaughtered calves (Sanni et al., 1999). It is either sold in the fresh or fried forms. Wagashie is a highly nutritious food with an excellent source of protein, fat, vitamins and minerals such as calcium, iron and phosphorus. It is used to replace meat or fish, or in combination with them in various food recipes especially for people with low income and can contribute to solving problems related to protein deficiency in diets in Africa (Elkhider et al., 2011). Its low lactose content makes it an acceptable food to many people who suffer from lactose intolerance associated with milk consumption in Africa and Asia due to low levels of intestinal β-galactosidase (lactase) (O‘Connor, 1993).The flower and other parts of the Calotropis procera can be used to cure coughs, catarrh, asthma, stomach pains and headaches. The problem associated with the product is that; Good hygienic practices during milking for wagashie production are not adhered to. Thus the Fulani do not clean the udder of the cow and the equipments used in the milking process. The environment in which the milking is done is also not clean; the milking is done mostly in the kraal with the faeces and urine of the animals. Also, wagashie is not packaged after University of Ghana http://ugspace.ug.edu.gh 3 preparation thus producers and retailers put the curds on metal trays and hand pick them into flexible polyethylene films for consumers. This exposes the product to post production contamination. The shelf life of the product may be affected and it also makes the product unsafe for the consumer. Traditional wagashie is known to have a bland taste as well as a bitter after taste. The bitter taste results from the high non-specific proteolytic activity of Sodom apple which also affects the yield of wagashie and the generation of excessive acid, bitter flavours and green colouration in the product (Mahami et al., 2012). The product has a bland taste mainly because the production process has not been standardised. ‘Wagashie' is referred to as a soft unripened cheese therefore it has a high moisture content of about 50% which makes it highly perishable. Ashaye et al., (2006) observed that, the shelf life does not exceed three days, after the second day of storage, ‘wagashie’ under ambient temperature undergo considerable chemical changes. These changes which are moisture change, proteolysis and Lipolysis are caused by increased activity of the resident lactic acid bacteria and adventitious microbes. The moisture content reduces causing hardening. Proteolysis sets in causing sourness and Lipolysis occurs imparting a rancid aroma to the product. The change in the composition is accompanied by changes in the sensory quality of the product. Traditionally, ‘wagashie’ is preserved in its whey which extends the shelf life to two days or boiled in water to make it tough which can increase the shelf life to 4 days when refrigerated. It is sometimes fried, smoked or dried to enhance its keeping quality. However, all these increase its shelf life by only a few days or a week at best. Soaking ‘wagashie’ in different concentrations of brine ,preservation with biological plant extracts like Afromomum danielli (Ashaye et al., 2006), ginger and garlic (Belewu et al., 2005) extended the shelf life up to fifteen days under ambient conditions. However, the taste was affected. Application of University of Ghana http://ugspace.ug.edu.gh 4 preservatives like propionic acid and sodium benzoate (Joseph and Akinyosoye, 1997) also increased the shelf life up to 15 days but residues of the chemicals in the product altered the taste. 1.1.0 Justification The poor hygienic practices during milking and processing of wagashie by the Fulani’s raise the need for its safety to be assessed.There is also the need to introduce a packaging material for wagashie in order to minimise post production contamination and extend its keeping quality. Standardising the traditional process for wagashie and replacing the extract of sodom apple plant with commercial rennet will eliminate the bitter after taste in the product. Traditional and chemical processing and preservative methods affect the sensory quality of ‘wagashie’ thus there is the need to find an appropriate preservative method which will keep the initial quality of the product and extend the shelf life. 1.2 Main Objecive To improve the sensory quality, safety and preservation of wagashie 1.3 Specific Objectives To assess the safety of wagashie To improve the sensory quality of wagashie To improve on the safety of wagashie To extend the shelf life of wagashie University of Ghana http://ugspace.ug.edu.gh 5 CHAPTER TWO 2.0 LITERATURE REVIEW 2.1 Traditional food Processing Traditional foods continue to play a central role in the eating habits of people in most part of West Africa. High post-harvest food losses, arising largely from limited food preservation capacity, are a major factor constraining food and nutrition security in the developing countries of West Africa, where seasonal food shortages and nutritional deficiency diseases are still a major concern. It is estimated that about 50% of perishable food commodities including fruits, vegetables, roots and tubers and about 30% of food grains including maize, sorghum,millet, rice and cowpeas are lost after harvest in West Africa. Ineffective or inappropriate food processing technologies, careless harvesting and inefficient post-harvest handling practices, bad roads, moribund rail systems, bad market practices and inadequate or complete lack of storage facilities, packing houses and market infrastructures are some of the factors responsible for high post-harvest food losses in West African countries (Aworh, 2008).The capacity to preserve food is directly related to the level of technological development and the slow progress in upgrading traditional food processing and preservation techniques in West Africa contributes to food and nutrition insecurity in the sub-region. Traditional food processing provides a source of livelihood for a large number of traditional food processors in the rural areas and currently in the urban areas (Lartey, 1975). Traditional foods and traditional food processing techniques form part of the culture of the people and its activities constitute a vital body of indigenous knowledge handed down from parent to child over several generations. Some of these techniques include fermentation, drying such as shallow layer sun drying, heat processing such as roasting, parboiling, cooking, milling such as wet milling and dry milling and post harvest operations such as winnowing, threshing, University of Ghana http://ugspace.ug.edu.gh 6 peeling, dehulling (Aworh, 2008). There is a vast array of traditional fermented foods produced in West African Countries .These include foods staple such as gari, kenkey, agbelima ,fufu, lafun and ogi ,fura; condiments such as dawadawa, ogiri (ogili) and ugba (ukpaka); alcoholic beverages such as burukutu (pito or otika),shekete and agadagidi; and the traditional fermented milks and cheese. Lactic acid bacteria and yeasts are responsible for most of these fermentations (Cooke et al., 1987). The fermentation processes for these products constitute a vital body of indigenous knowledge used for food preservation, acquired by observations and experience, and passed on from generation to generation. 2.2.0 Traditional production of ‘wagashie’ a local cottage cheese In West Africa, the Fulani pastoralists process surplus fresh milk into various stable products like West African soft cheese (wagashie), Nono (fermented skimmed milk) ,nyarmie and yoghurt. Wagashie is an unripened cheese consumed in several parts of West Africa mainly Nigeria (wara or warankashi), Benin (woagachie) and Ghana (wagashie). Nomadic Fulani who are pastoralists and are mainly involved in the rearing of cattle from one place to the other are involved in milk production in Ghana. In herding families, only the children, pregnant women and the elderly drink milk regularly while others get milk only on rare occasions due to transportation problem or poor keeping quality of milk if not processed (Belewu et al., 2005). The Fulani women process excess raw milk into a soft, unripened cheese called “wagashie” as a way of preserving the excess milk for short term. Wagashie is mainly consumed at home or sold on the market widely in the Northern and Volta regions of Ghana. Wagashie is also produced in Ghana by small holder groups in the rural and peri- urban communities under “indigenous” conditions using skills based on traditions (Mahami et al., 2012). Wagashie processing involves the use of rudimentary equipment, and mostly University of Ghana http://ugspace.ug.edu.gh 7 starter cultures are not used as processing conditions are not normally standardized or optimized (Belewu et al., 2005). Wagashie is traditionally prepared by heating fresh milk then coagulating it with an extract of stems or leaves from the Sodom Apple plant (Calotropis procera). The formed cheese curd is then poured into perforated calabashes to allow the whey to drain off (ILLR, 2006). The leaves and stem extracts of pawpaw (Carica papaya) can also be used as a coagulant but the extracts from Calotropis procera are preferred to the extracts of pawpaw because cheese processed with Calotropis procera has a sweeter flavor compared to the cheese processed with pawpaw leaf extracts (O’Connor, 1993). Wagashie is described as soft white unripened cheese by (Ogundiwin, 1978); it has also been described as soft, wet, feta-like cottage cheese made from whole milk (Jansen, 1990). It’s similarity with the cottage cheese is related to the fact that they are both classified as soft cheese due to the high moisture content, they do not undergo ripening and they have a curd like texture, however, cottage cheese involves the use of starter culture whiles ‘wagashie’ does not (Ashaye et al., 2006). Improvement of the product has been carried out by (Sanni et al., 1999) focusing on the texture, aroma and the nutritional composition by the use of starter culture (lactococcus lactis). It was discovered that consumers preferred the traditionally prepared ‘wara’ in terms of appearance and texture but liked the aroma and palatability of the improved type. Improving ‘wagashie’ with a different coagulant other than Calotropis procera has however not been done. 2.2.1 Preservation of wagashie ` ‘Wagashie’ provides a useful service in extending the shelf life of milk (Alalade and Adeneye 2006) by serving as an important preservation method for surplus milk in rural areas especially during rainy season when milk is in abundance. (Aworh and Egounlety, 1985) University of Ghana http://ugspace.ug.edu.gh 8 reported the processing technology of ‘wagashie’, its stabilization by heat treatment and the use of chemical additives such as propionic acid and sorbates. (Kèkè et al., 2008) reported a method of preservation of ‘wagashie’ using strains of Lactobacillus plantarum. The preservation of ‘wagashie’ by chemical method has a negative effect on the sensory quality of the product (Kèkè et al., 2008). Usually wagashie is stored in its whey at room temperature (28oC), under this storage condition, wagashie is highly perishable and has a shelf life of 2 to 3 days ((Belewu et al., 2005, Adejunti, 2011). Therefore, ‘wagashie’ production must be protected from spoilage microorganisms from production to the consumer. ‘Wagashie’ is usually fried and used as a meat substitute in stews and soups, or smoke dried to enhance its keeping qualities. Studies by (Appiah, 2000) showed that preserving ‘wagashie’ in different concentrations of NaCl increased the shelf life up to twenty days during storage. However, all these preservation methods increase its shelf life by only a few days or few weeks (Sessou et al., 2013). Moreover the traditional practices for preservation of wagashie extends the shelf life and preserves the sensory quality of the product for a relatively long time but the exhaustive inventories of all these practices are not well documented (Sessou et al., 2013). 2.2.2 Safety of wagashie Soft cheeses are good proteins source with high water content (60%) which is favorable for the growth of microorganisms that affects its quality (Sessou et al., 2013). ‘Wagashie’ is a type of fresh cheese and its production in Ghana has become a common sight without supervision or quality due to increase in population and consumption. ‘Wagashie’ is delivered to the market immediately after processing, under inadequate conditions, poor handling techniques, inappropriate packaging materials and lack of adequate storage facilities, however, dairy products including cheese must be safe, acceptable and meet consumer's satisfaction. The United States Food and Drugs Administration (FDA, 2005) stated that, soft University of Ghana http://ugspace.ug.edu.gh 9 raw milk cheeses can cause serious infectious diseases including Listeriosis, Brucelleosis, Salmonellosis and Tuberculosis. Recently, the consumer desire for healthy microbiologically safe foods has been increased; therefore the importance of production of cheese being properly packed in convenience, smaller size packages, and longer product shelf life is important. ‘Wagashie’ preparation does not involve the use of starter cultures and lack of adherence to hygienic principles in production line leads to contamination in cheese with non-starter lactic acid and Psychotrophic bacteria as investigated by Muehlenkamp-Ulate and Warthesen (1999); and Sousa et al., (2001). Also according to (Ashaye et al., 2006), wagashie has a neutral pH of 6.0 to 6.5 and a low salt content which makes it highly susceptible to the growth of spoilage and pathogenic microbes. 2.2.3 Spoilage and Pathogenic microorganisms in soft cheeses (‘wagashie’) High moisture content and high pH of soft fresh cheeses make them susceptible to pathogenic and spoilage microorganisms like E. coli, L. monocytogenes, Salmonella spp. and Staphylococcus aureus (Kousta et al., 2010). Manufacture of fresh soft cheeses with pasteurized milk is necessary to reduce non essential microorganisms in milk prior to cheese preparation. Addition of salt, good sanitation of the cheese plant, the use of good starters noted by acid production and strict control of storage and processing temperatures are important factors considered in cheese making (Farkye and Vedamuthu, 2002). Gram negative psychrotrophic microorganisms such as pseudomonas spp, Coliform bacteria, yeast and moulds are associated with the spoilage of fresh cheeses. These organisms can cause sliminess, bitter tastes, off flavours and colour defects due to their high proteolytic and lypolitic activities. University of Ghana http://ugspace.ug.edu.gh 10 2.2.4 Calotropis procera Calotropis procera (Sodom apple) is a member of the plant family Asclepiadaceae, a shrub about 6m high and is widely distributed in West Africa and other parts of the tropics (Irvine, 1961). The plant is erect, tall, large, much branched and perennial with milky latex throughout. In Nigeria traditional medicine, C.procera is either used alone or with other herbs to treat common diseases such as fevers, rheumatism, indigestion, cold, eczema and diarrhea also, preparations from latex with honey are also used as antirabies and also in the treatment of toothache and cough (Kew, 1985). The secretion from the root bark is traditionally used for the treatment of skin diseases, enlargement of abdominal viscera and intestinal worms in India (Parrotta, 2001). In Senegal, the milky latex is locally applied in the treatment of cutaneous diseases such as ringworm, syphilitic sores and leprosy (Kew, 1985). Leaf extracts, chopped leaves and latex of C.procera have shown great promise as a nematicide in vitro and in vivo (Khirstova and Tissot, 1995). The potentials of C.procera leaves in water treatment and its ability to reduce total viable count have also been reported (Shittu, et al., 2004). In Ghana, the latex from the leaves and stems of Calotropis procera is used to prepare ‘wagashie’ a local cheese; the extract however causes bitterness in the cheese due to its high proteolytic activitiy (Mahami et al., 2012).The latex is used as an abortifacient, spasmogenic and carminative properties, antidysentric, antisyphilitic, antirheumatic, antifungal, mullusccide, diaphoretic and for the treatment of leprosy, bronchial asthma and skin infection. Different parts of the plant have been reported to possess a number of biological activities such as proteolytic, antimicrobial, larvicidal, nematocidal, anticancer, anti- inflammatory (Basu and Chaudhury, 1991). The flowers have digestive and tonic properties. On the contrary, the powdered root bark has been reported to give relief in diarrhoea and dysentery. The root of the plant is used as a carminative in the treatment of dyspepsia. The root bark and leaves of Calotropis procera are used by various tribes of central India as a University of Ghana http://ugspace.ug.edu.gh 11 curative agent for jaundice, ulcer and leprosy (Shamar et al., 2009). C. Procera is drought- resistant, salt-tolerant to a relatively high degree, and it disperses seeds through wind and animals. It quickly becomes established as a weed along degraded roadsides, lagoon edges and in overgrazed native pastures. It has a preference for and is often dominant in areas of abandoned cultivation especially sandy soils in areas of low rainfall; assumed to be an indicator of over-cultivation (Shamar et al., 2009). 2.3.0 Milk production in Ghana Ghana has about 1.25 million cattle and the West African Shornhorn is the most popular breed selected for high milk production (Otchere and Okantah, 2001). Currently local milk production is conservatively estimated at 36.5 thousand tonnes which are mostly from agropastorial producers. Milk is collected by small holder herdsmen for home consumption and for sale. Milking is done often in the morning in the presence of the calf to induce milk let down (Okantah et al., 1999). Okantah (1990), reported a mean daily partial milk yield of 0.9 kg and 0.7 kg for wet and dry season on the Accra plains, Karbo et al., (1998a) reported similar observations also on the Accra plains. These observations together showed that cattle kept by smallholders are low milk producers. However large quantities of milk are available from several thousands of low yielding cattle in small holder system (Okantah, 1990). The Ministry of Food and Agriculture has therefore initiated a pilot milk collection project on the Accra plains, peri-urban Kumasi and Sekyedumasi in the Ashanti Region. In 1998, the pilot collected 85,587 litres of milk from small holder farmers in peri-urban Accra. In 2001, the annual milk production was estimated at 36.5 thousand tonnes with most of it coming from small holder agropastorial producers. Also, according to the Animal Production Directorate of MOFA, (2003), the Sanga was crossed with Friesian to increase daily milk production to University of Ghana http://ugspace.ug.edu.gh 12 between 6 and 8 L as compared to the local breeds which produced less than 2 L of milk per day. Since the milk production in Ghana is low, the per capital consumption of milk is low. 2.3.1 Milk consumption in Ghana Milk production in 2006 was estimated at 34,000 metric tonnes and an average of 37,195 metric tonnes of liquid milk equivalent (LME) was imported yearly into the country (Government of Ghana (GOG)/Food and Agricultural Organisation (FAO), 2002). In Ghana, few people consume pasteurized raw milk whiles most people consume the condensed full cream evaporated milk. This is because milk production in Ghana is low resulting in a low per capital income milk consumption. There is therefore a shortfall between domestic milk production and consumption. The deficit is made up through imports of milk and milk products. From 1995 to 1999 the total volume of dairy products that was imported into the country was 39,831.4 x 103 t (Otchere and Okantah, 2001). However, in peri-urban Tamale milk producers obtained a fair daily income from the sale of milk annually (Karbo et al., 1998). According to Nsiah-Ababio (1998) who conducted a study revealed that the potential income of milk producers was the same as the income contribution from crop producers. Also a study conducted in the Techiman District of Brong Ahafo revealed that making ‘wagashie’ (cottage cheese) from fresh local milk added a value of 54% to milk compared with sales of the fresh product (SFSP–GTZ–MOFA, 1998). 2.3.2 Traditional Fermented milk products in Ghana Milk is the most abundant fermented animal product in Africa, although the extent to which milk is used in the dairy diet varies to a great extent (Jespersen, 2003). Fermented milk products are very important for people suffering from lactose intolerance, social value and as University of Ghana http://ugspace.ug.edu.gh 13 a means of generating income (Beukes et al., 2001). The art of making these products is handed down from one generation to generation (Caplice and Fitzgerald 1999). In Ghana, apart from ‘wagashie’, various fermented milk products are produced; there is nunu, nyarmie, and recently Burkina. 2.3.2.1 Nunu Traditionally, nunu is prepared by inoculating freshly drawn cow milk with a little of the leftover as a starter and then it is allowed to ferment for about 24 h at room temperature. During fermentation, some of the lactose is converted to lactic acid. At the end of fermentation period, the milk butter is removed by churning for further use, and the remaining sour milk, nono, is a delicious and refreshing beverage (Olalokun, 1996). Fermentation is said to be essentially brought about by various species of bacteria especially members of the genus Lactobacillus and other Lactic Acid Bacteria (LAB), moulds and yeasts and variations in milk composition, bacterial flora and ambient temperatures have been noticed to be responsible for products of varying qualities (O’Connor and Tripachi,1995). 2.3.2.2 Burkina Burkina is also prepared by adding cooked millet to partially fermented milk, sugar and flavour may be added. Milk powder is mostly used for the preparation. The milk powder is mixed with water and heated. It is allowed to cool to room temperature and inoculated with prepared yoghurt or sometimes by natural fermentation and is allowed to occur overnight. After fermentation sugar and flavour is added and sieved steamed millet is added to the milk and served in plastic bottles (personal observation). University of Ghana http://ugspace.ug.edu.gh 14 2.3.2.3 Nyarmie Nyarmie is another fermented product consumed in Ghana by the Fulani tribe. It is produced by pasteurising raw cow milk at 60 to 75 0 C for 30 – 45 mins. The milk is cooled and kept overnight at 28 0C for natural fermentation to occur without the use of starter cultures A thick product is formed; it is then whipped and served. There is however variability among producers. It is preferred to both the fresh and the pasteurized milk and can be left and consumed for 5 days unrefrigerated or refrigerated for several weeks at 4 oC. Freshly produced nyarmie has a pleasant taste and a pH of about 4.2; however, the pH drops to below 4.0 after a day and becomes sour (Obodai and Dodd, 2006). 2.3.3 Milk composition Since the middle of the 19th century, knowledge of the chemistry of milk has been developed and today there is extensive literature on the chemistry of the major and minor constituents of milk, especially cow milk. Variations exist in the composition of milk for various species. The composition of cow milk varies for a number of reasons, e.g. the individuality of the cow, the breed and age, stage of lactation, health of the cow, climatic conditions and herd management which includes feeding and general care (O’Connor, 1993).Milk consists of protein (caseins and whey proteins), lipid, lactose, minerals (soluble and insoluble), minor components (enzymes, free amino acids, peptides) and water.The nitrogenous fraction of cows’ milk typically consists of casein, whey protein and non protein nitrogen (urea, proteose-peptones, peptides) at levels of 78, 18 and 4 g 100 g−1, respectively, of total nitrogen (Law and Tamime, 2010). University of Ghana http://ugspace.ug.edu.gh 15 2.3.3.1 Casein Caseins make up over 80% of the total protein content and they can be further divided into five groups; the alphas1, alphas2, beta, gamma, and kappa caseins. Caseins do not have an organized structure, thus they cannot be denatured by heating. The casein fraction coexists with the insoluble minerals as a calcium phosphate–casein complex. Casein in milk exists in the form of spherical-shaped colloid particles (40–300 nm diameters) known as casein micelles (Fox & Brodkorb, 2008; McMahon & Oommen, 2008). Casein, which is typically present at a level of 2.5 g 100 g−1 in cows’ milk, is the main structural protein of both rennet- and acid-induced milk gels (cheese) (Law and Tamime, 2010). 2.3.3.2 Whey protein Whey is the liquid remaining after milk has been cudled and strained during the manufacture of cheese. It is used to produce ricotta and brown cheeses and is an additive in many processed foods, such as breads, crackers, pastries, and animal feed. A high level of casein to whey protein interaction, induced by high heat treatment of the milk, is highly favoured in the manufacture of yoghurt and smooth-textured cheeses with a high moisture to protein ratio, such as cream cheese and ultra filtration produced Quark (Guinee et al., 1995). High heat treatment of milk is generally undesirable for rennet curd cheeses as denatured protein at levels of ≥25% impedes the ability of the milk to gel on rennet addition, causes deterioration in melt properties of the cheese and reduces the recovery of fat from milk to cheese (Rynne et al., 2004). However, a higher than normal heat treatment that gives a moderate degree of whey protein denaturation may be desirable as a means of modulating the texture of reduced fat cheese (Guinee, 2003; Rynne et al., 2004). University of Ghana http://ugspace.ug.edu.gh 16 2.3.3.3 Lactose Lactose or milk sugar is the major carbohydrate of milk occurring at 4.5 to 4.9% levels. It is a disaccharide composed of glucose and galactose. Lactose is the most abundant of the milk solid, its crystallization is important in the manufacture and utilization of several dairy ingredients. Lactose is a useful source of energy and it promotes the absorption of calcium. However, lack of the enzyme lactase results in the difficulty in digesting lactose, which results in gastrointestinal distress. Some milk products are available that have lactase introduced aseptically into the previously sterile product before packaging. The relative sweetness of lactose is small (20%) when compared to sucrose which is about 100%, (http://www.lactose.com/basic/physiological_properties.html). 2.3.3.4 Minerals Cow milk contains minerals, which comprises Potassium, Calcium, Chlorine, Phosphorus, Sodium, Magnesium, Sulphate and Citric acid (O’Brien et al., 1999c). These minerals are partitioned into varying degrees between the serum (soluble) and the casein (colloidal or insoluble) in native milk at room temperature at a pH of 6.6–6.7. Ash which is the white residue after incineration of a given weight of milk is used as a measure of the mineral content of milk. It is not identical to milk mineral level because of the decomposition and volatilization of certain minerals due to heat (Law and Tamime, 2010).Their concentration is less than 1% in milk, but they are involved in heat stability and alcohol coagulation of milk, age thickening of sweetened condensed milk, feathering of coffee cream, rennin coagulation, and clumping of fat globules upon homogenisation. The calcium level of milk influences the firmness of curd during cheese making University of Ghana http://ugspace.ug.edu.gh 17 2.3.3.5 Lipids Lipids exist in the form of an oil-in-water type of emulsion, with fat globules varying from 0.1 to 22 μm in diameter.The lipid content of milk fat is 97–98%, triacylglycerols 2–1% phospholipids, 0.2–0.4% sterols, and traces of fatty acids, and vitamins A, D, E, and K. The cholesterol content of whole milk (3.3% fat) and skimmed milk is 14 mg/100 ml and 2 mg/100 ml, respectively. The fat in milk exists in the form of dispersed globules surrounded by a lipoprotein membrane (milk fat globule membrane, MFGM) (Keenan & Maher, 2006). Inadvertent damage of the membrane, by manhandling of the milk (e.g. excessive shearing, turbulence), is highly undesirable in cheese manufacture. It leads to free fat in the cheese milk, lower recovery of milk fat to cheese, lipolysis of the fat by lipases that survive pasteurisation treatment, high levels of FFA and undesirable flavours (e.g. bitter, soapiness, metallic), especially in some cheese types (e.g. Emmental, Gouda, Cheddar). 2.3.4 Milk Quality and Safety Dairy product safety is an additional concern related to milk quality. Milk safety hazards are associated with undesirable substances or organisms that contaminate milk and constitute a risk to the health of the consumer (Anon., 2003).Milk quality may be defined under a broad range of characteristics notably; microbial (Pathogenic and non Pathogenic bacteria), chemical, compositional, physicochemical, enzymatic and issues of adulteration. Milk produced under hygienic conditions from healthy cows should contain not more than 50, 000 bacteria per millilitre (O’Connor, 1993). Sources of microbial pathogens that can contaminate milk are endogenous sources, such as the cow and exogenous sources, such as the environment (soil, water, manure or human contact), collection and processing equipment, milk handlers on the farm and in the factory (Tybor and Gilson, 2003). University of Ghana http://ugspace.ug.edu.gh 18 Milk quality usually may also be defined by the somatic cell count and the bacterial count of pre-pasteurized bulk tank milk. The largest factor that influences the somatic cell count of milk is mastitis (Hamman, 2003). The somatic cell count of a cow that is not infected with mastitis is usually less than 200,000 cells/ml and many cows maintain somatic cell count values of less than 100,000 cells/ml. A somatic cell count greater than 200,000 cells/ml is almost always caused by mastitis. High somatic cell count in milk reduces the shelf life of dairy products and diminishes the quality and quantity of milk protein, thereby reducing cheese yields (Barbano, et al., 1991). It is important that milk, whether it is for direct consumption or for the manufacture of dairy products, is of good hygienic quality (O’Connor, 1993). Milk that is of poor hygienic quality will result in poor quality products with low consumer acceptabilityContaminated milk may cause illnesses to humans including tuberculosis, brucellosis, sore throats, diarrhoea and abdominal pains (O’Connor 1993). 2.3.5 Microbes and sources of microbes in milk Milk is an excellent medium for growth of microorganisms; it provides rich nutrients i.e. proteins, fats, lactose, vitamins and minerals for microbes because of the high moisture content and neutral pH. Raw milk can potentially contain pathogenic bacteria such as Salmonella spp., Staphyloccocus spp., E. coli, Bacillus spp., Enterobacteriacea, yeasts, molds and coliform bacteria (Jayarao and Henning 2001). However, recent cheeses produced in industries use pasteurised milk thus reducing the risk to public health. Consideration of pathogenic bacteria is necessary where cheese is manufactured from raw unpasteurised milk or the manufacture of speciality cheese types although pasteurised milk is used. Cheeses can easily be contaminated with pathogenic bacteria through mishandling mostly in the development of smear type cheeses where a complex and different microbial system evolves on the surfaces of unpackaged cheeses with high pH during ripening in humidified University of Ghana http://ugspace.ug.edu.gh 19 atmospheres . Common Salmonella species (spp.) which contain several strains that cause illness in humans are the Enteriditis and Typhimurium and are found in the intestinal tracts of all warm blooded animals including humans can be destroyed by pasteurization (Kongo, 2013). The rate of spoilage of many dairy foods is slowed by the application of one or more of the following treatments: reducing the pH by fermenting the lactose to lactic acid; adding acids or other approved preservatives; introducing a desirable microflora that restricts the growth of undesirable microorganisms; adding sugar or salt to reduce the water activity (aw); removing water, vacuum packaging and freezing (Ledenbach and Marshall, 2009). In cheese production, slow lactic acid production by starter cultures favours the growth and production of gas by coliform bacteria, with coliforms having short generation times under such conditions. In soft, mold ripened cheeses, the pH increases during maturing which increases the growth potential of coliform bacteria (Frank, 2001). Bacterial contamination of raw milk can originate from different sources: air, milking equipment, feed, soil, faeces and grass. The number and types of microorganisms in milk immediately after milking are affected by factors such as animal and equipment cleanliness, season, feed and animal health (Rogelj, 2003). It is hypothesized that differences in feeding and housing strategies of cows may influence the microbial quality of milk. Rinsing water for milking machine and milking equipment also influence the presence of a higher number of micro-organisms including pathogens in raw milk (Bramley, 1990). After milking, milk is cooled, which additionally influence the dynamic of microbial process (Rogelj, 2003). Entry of food borne pathogens through contaminated raw milk into dairy food processing plants can lead to the perseverance and establishment of pathogens in the form of bio films, subsequent contamination of processed milk products and exposure of consumers to the pathogens. Insufficient processing may result in the survival of definite pathogens and such contaminants become a public-health threat (Oliver et al., 2005). The conditions during storage and University of Ghana http://ugspace.ug.edu.gh 20 transport in refrigerated tanks cause the raw milk micro biota to change from predominantly Gram-positive to predominantly Gram-negative organisms as they grow. 2.3.6 Lactic acid fermentation of milk Lactic acid bacteria (LAB) are Gram-positive, usually non-motile, acid tolerant microorganisms. They have complex nutritional requirements and a fermentative metabolism. Phylogenetically the lactic acid bacteria belong to the clostridial branch of the Gram-positive bacteria. They are catalase negative, non spore forming, cocci, cocobacilli or rods that have less than 55 mol% G+C content in their DNA (Stiles and Holzapfel, 1997). Species of lactic acid bacteria (LAB) belong to numerous genus under the family of Lactobacillaceae. Fermentation, also known as bio preservation, is a cheap, widely accessible method of food preservation. Bio preservation with lactic acid bacteria (LAB) is indeed one of the oldest and highly efficient forms of non-thermal processing method. Fermentation is generally considered as a safe and acceptable preservation technology of food and fermentation using LAB can be categorized into two groups based on the raw material used, non-dairy and dairy fermentation. Milk from different mammalian animals can be used in dairy fermentation to produce several products. Milk of cow followed by milk of goat and sheep are the most widely used raw materials to produce particular economic value fermented milk products worldwide (Widyastuti et al., 2014). The presence of LAB in milk fermentation can be either spontaneous or inoculated starter cultures. Milk is one of the natural habitats of LAB (Delavenne, et al., 2013, Fox and Mcsweeny,2004) The most important properties of LAB are their ability to reducing the pH of milk and to generate flavour and texture, by converting milk protein due to their proteolytic activities (Griffiths and Tellez,2013). LAB has the ability to ferment sugars, especially glucose and galactose, to produce lactic acid and aroma substances that give characteristic flavors and tastes to fermented products. LAB also release University of Ghana http://ugspace.ug.edu.gh 21 antimicrobial metabolites called bacteriocins, which are considered safe and natural preservatives, with great potential to be used on their own or synergistically with other methods in food preservation (Kongo, 2013). ). Among the bacteriocins produced by LAB, nisin produced by Lactococcus lactis spp., is the only bacteriocin that has been officially employed in the food industry and its use has been approved worldwide (Zacharof and Lovitt, 2012). Recently LAB isolates from traditional Portuguese raw-milk cheeses, revealed several lactobacilli having antibacterial activity against pathogens such as Listeria monocytogenes, Staphyloccus aureus, Salmonella Newport and E.coli (Kongo, 2013). LAB have a long and safe history of application and consumption in cheese processing (Wood, 1997; Wood & Holzapfel,1995, Caplice & Fitzgerald, 1999 Giraffa et al., 2010) thus being generally regarded as safe (GRAS). Lactic acid bacteria represent the most extensively studied microorganisms for milk fermentation (Olson, 1990 and Maragkoudakis et al., 2006) and imparts the mild acid taste and pleasant fresh characteristics to fermented milk products such as yoghurt and cheese. 2.4.0 Cheese Cheese is a concentrated protein gel, which absorb fat and moisture (Law and Tamime, 2010). According to CODEX STAN 283(1978) Cheese is the matured or not fully matured soft, semi-hard, hard, or extra-hard dairy product, which may be coated, and in which the whey protein to casein ratio does not exceed that of milk by coagulating wholly or partly the protein of milk, skimmed milk, partly skimmed milk, cream, whey cream or buttermilk, or any combination of these materials, through the action of rennet or other suitable coagulating agents, and by partly draining the whey resulting from the coagulation, while respecting the principle that cheese making results in a concentration of milk protein and that therefore, the protein content of the cheese will be definitely higher than the protein level of the blend of the University of Ghana http://ugspace.ug.edu.gh 22 milk supplies from which the cheese was made and by processing techniques involving coagulation of the protein of milk and products obtained from milk which give an end- product with the same physical, chemical and organoleptic characteristics.Cheese is one of the commonest dairy products in the world (Belewu et al.,2012). It is today, a major business worth billions of dollars in many industrialized countries (Aworh, 2008). Cheeses are now unique products in their own right and cheese-making has advanced beyond being merely a food preservation technique (Aworh, 2008). 2.4.1 Ingredients for cheese making 2.4.2 Milk 2.4.2.1 Selection of milk for cheese making Cheese manufacture commences with the selection of milk of high microbiological and chemical quality, cheese milk must be free of antibiotics. In commercial practice, milk for cheese is normally cooled to 4°C immediately after milking and may be held for several days (Fox, 1993). Although raw milk is still used in cheese making, most cheese milk is pasteurized before use. Pasteurization alters the indigenous micro flora and facilitates the manufacture of cheese of more uniform quality, but unless due care is exercised, it may damage the ability of the rennet to coagulate and the curd forming properties of the milk even when properly pasteurized. Pasteurization of cheese milk minimizes the risk of cheese serving as a carrier of food poisoning or pathogenic microorganisms, thus high quality raw milk may be unacceptable for cheese manufacture (Fox, 1993). Milk for cheese production should be free of any visible impurities, must not have any abnormal taste or odour, have a pH of 6.6 or slightly higher at the milking time, must not be contaminated by pathogenic microorganisms which may prove undesirable for the production of cheese. The University of Ghana http://ugspace.ug.edu.gh 23 milk must also contain no foreign substances such as antibiotics, antiseptics and cleaning products. 2.4.2.2 Starter culture Certain cheese varieties require pure cultures of lactic acid bacteria containing organisms with specific functions while for traditional cheeses, a natural fermentation is allowed using whey from the previous lot (Kongo et al., 2013). The recipe will indicate the type and quantity of starter to be used and temperature conditions. 2.4.2.3 Lactic acid bacteria as starter culture in cheese making Cheese making is based on the use of lactic acid bacteria ( LAB) in the form of defined or undefined starter cultures that are recognized to cause a rapid acidification of milk through the production of lactic acid, with a resultant decrease in pH, thus affecting a number of aspects of the cheese making process and in the end cheese composition and quality (Kongo, 2013).The initial productions of cheeses were based on the natural fermentation resulting from the growth of the microflora naturally present in the raw milk and its environment. It is known that while unprocessed milk can be stored for few hours at room temperatures, cheeses may reach a shelf life up to five years depending on the variety (Kongo, 2013). The quality of the cheese is as a result of the microbial load and range of the raw materials. Natural fermentation was later optimized through back slopping, which is inoculating of the raw material with the whey from a previously performed successful fermentation, the end product characteristics depended on the best adapted strains dominan. Fresh cheeses with unlimited shelf life have the primary proteolysis which is performed by the coagulating agents and to a lesser extent plasmin residual coagulants and enzymes from the starter organisms (Sousa et al., 2001). Starter cultures of LAB used in cheese making can be either mesophilic from the University of Ghana http://ugspace.ug.edu.gh 24 genera of Lactococcus and Leuconostoc or thermophilic from the genera of Streptococcus and Lactobacillus (Fox and Mcsweeny, 2004). Among species, Lactococcus lactis (Dias and Weimer 1998, Hannon et al., 2007), Streptococcus thermophilus (Helinck et al., 2004) and Lactobacillus helveticus (Dias and Weimer, 1998) are intensively studied. L. helveticus is specialized in milk species and belong to the member of dairy niche species (Slaterry et al., 2010). Several cheese products are based on L. helveticus as starter culture. It is also known that L. helveticus have significant role in production of specific flavour compounds in Italian cheese types and removing the bitterness in cheese (Gatti et al., 2003, Rossetti et al., 2008). The starter cultures added during the production of fresh cheese are mesophilic group including L. Latis sub sp. lactis and L. lactis sub sp. cremoris with different capacity of producing citrate. Diacetyl is a major product of citrate metabolism of lactococci and is desired in many fresh cheese varieties such as cottage cheese (Fernández et al., 1994). 2.4.2.4 Chemicals Chemicals such as calcium chloride and sodium nitrate are recommended for some varieties of cheese to improve curd quality and prevent the growth of organisms which may cause problems during the ripening or maturing of the cheese (Cooker et al., 2005). 2.4.2.5 Coagulants Coagulation, or clotting of the milk, is the basis of cheese production. Coagulation is brought about by physical and chemical modifications to the constituents of milk and leads to the separation of the solid part of milk, the curd from the liquid part, the whey. Milk coagulation being one of the most important steps in the cheese manufacturing process determines the final cheese properties (García, et al., 2012). The difference in protein matrix degradation as a result of the agents used in the clotting process affects the changes that take place in the yield, University of Ghana http://ugspace.ug.edu.gh 25 the cheese texture (elasticity, fragility, adhesiveness, hardness, gumminess and chewiness) and the development of flavours (especially a bitter taste), through the production of hydrophobic peptides and the hydrolysis of caseins (Macedo et al., 1993).The two types of coagulants used in cheese making are, rennet (Animal and Microbial rennets) and plant or vegetable extracts. In the olden days, many cheeses were made with vegetable or plant coagulant but some are now made with animal or microbial coagulants. Animal and microbial coagulants give more consistent products and are also cheaper and easier to use, thus avoiding the labour intensive and expensive collection of plants (Roseiro et al., 2003) 2.4.2.5.1 Rennet coagulants Rennet is the most common coagulant. The most commonly used rennet contains the enzymes chymosin and pepsin, either as an extract from the abomasums of calf or as the recombinant products (microbial source). The development of products with new sensory and textural features is one of the main areas of innovation in cheese making (García, et al., 2012). Calf rennet, has until lately been the reference product against which alternative products are measured. Adult bovine rennet is the most widely used alternative to calf rennet because it contains the same active enzymes as calf rennet. Bovine rennet has a high pepsin content which gives the product a high sensitivity to pH, and a higher general proteolytic activity. Also lamb ovine and kid-caprine or caprine rennet are very similar to calf or adult bovine rennet, but they are best suited for clotting milk of their own species (Foltman, 1992). Animal rennet is usually mixed with lipases, especially during the manufacture of South Italian cheeses, in which they produce a characteristic flavour. Such products are called rennet paste, and they are made by maceration and drying of stomachs from suckling calves, lambs or kid-caprine, which have their stomachs filled with milk. Thus rennet paste contains University of Ghana http://ugspace.ug.edu.gh 26 a mixture of rennet and lipase (pregastric and possibly gastric) enzymes in an un-standardised ratio (Law and Tamime, 2010). 2.4.2.5.2 Microbial coagulants The most popular microbial coagulants used for cheese making are of fungal origin. Most bacterial proteases identified as milk clotting enzymes are unsuitable due to their high proteolytic activity (Law and Tamime, 2010). The most regularly used microbial coagulants are proteases derived from Rhizomucor miehei, Rhizomucor pusillus and Cryphonectria parasitica, R. Miehei, have been used as a substitute of animal rennet for almost 40 years (Jacob et al., 2011). C. parasitica proteases cleave the Ser104-Phe105 bond in κ-casein, while R. miehei cleaves the Phe105-Met106 bond. Also, the higher heat stability of the derivatives obtained from R. miehei may be due to excessive proteolysis, with a reduced ripening time and bitter cheeses. Coagulants with greater heat stability than calf rennet should be avoided and there should be differences in the coagulation temperature to limit excessive proteolysis (Sousa et al., 2001). 2.4.2.5.3 Fermentation Produced Chymosin FPC is chymosin produced by fermentation of a Genetically Modified Organism (GMO).They contain chymosin similar to the chymosin from animal source, thus they have the same amino acid sequence as chymosin from animal stomach. The main FPC, which contains bovine chymosin B, is now considered to be the ideal milk clotting enzyme against which all other milk-clotting enzymes are measured. The production and application of bovine type FPC has been reviewed by Harboe, (1992a, 1993); Repelius, (1993) and recently, a new production of FPC, equal to camel chymosin, has been manufactured. FPC (camelus) has been found to be an effective coagulant for bovine milk than FPC (bovine), and is University of Ghana http://ugspace.ug.edu.gh 27 characterised by its very high specificity against caseins, which leads to high cheese yields without creating any bitterness (Law and Tamime, 2010). 2.4.2.5.4 Vegetable or Plant Coagulants Cheeses made with vegetable coagulant can be found mainly in the Mediterranean, West African and Southern European countries. Spain and Portugal have the largest variety and production of cheeses using Cynara sp. as the vegetable coagulant. Rennet substitutes of plant origin have been increasingly used to manufacture cheese. Juice extracts from fruits and plants have long been used as milk coagulants (O’Connor, 1993). Application of plant coagulants allows target cheese production, and hence contributes to improve the nutritional input of those populations on whom restrictions are imposed by the use of animal rennets (Gupta & Eskin, 1997). Several plant preparations have been shown to clot milk (Aworth and Muller, 1987; Edwards and Kosikowski, 1983; Padmanabhan et al., 1993; Pozsar et al., 1969; Tamer, 1993) to clot milk; however, the majority proved unsuitable for cheese production due to their excessively proteolytic activity which reduces cheese yield and produce bitter flavours in the final cheese (Lo Piero et al., 2002). Cynara sp. (cardoon) extract has been widely used for centuries for making traditional Portuguese and Spanish ewe’s milk cheeses. Similarly, Calotropis procera (Sodom apple) which grows abundantly in many parts of Africa has been used for traditional cheese making in West African countries, such as Nigeria, the Republic of Benin and Ghana.Other types of plant or vegetable extracts used as cheese coagulants are lemon juice (Adetunji et al., 2007), Carica papaya leaves and sap, the berries of Solanum elaeagnifolium (trompillo or silverleaf nightshade) (Martinez-Ruiz et al., 2013) pineapple (bromelin), castor oil seeds (ricin) and latex of the fig tree and the plant. The types of rennet and coagulants and their characteristics have been reviewed by several authors (Harboe, 1992b; Guinee & Wilkinson, 1992; Garg & Johri, 1994). These extracts are suitable University of Ghana http://ugspace.ug.edu.gh 28 for softer curd cheese which is consumed within a few days. The extracts are not suitable for hard cheese with long maturing periods because of their excessive proteolytic activity which results in bitter flavours in the ripened cheese (O’Connor, 1993).Rennet and coagulants are most efficiently categorised according to their source. Table 2.2 shows the predominant types of coagulant used for cheese making and their active enzyme components. Table 2.1: The most commonly used rennet and coagulants and their enzymes Group Source Examples of rennet and coagulants Active enzyme components Animal Bovine stomachs Ovine stomach Caprine stomach Calf rennet, adult bovine rennet Rennet paste Lamb rennet, ovine rennet Kid-caprine rennet, caprine rennet Bovine chymosin A, B and C, pepsin A and gastriscin The same as above, plus lipase Ovine chymosin and pepsin Caprine chymosin and pepsin microbial Rhizomucor miehei Cryphonectria Parasitica Miehei coagulant type L, TL, XL and XLG/XP Parasitica coagulant Rhizomucor miehei aspartic proteinase Cryphonectria parasitica aspartic Proteinase FPCa Aspergillus niger Kluyveromyces marxianus var. Lactis CHY-MAXTM CHY-MAXTM M Maxiren R _ Bovine chymosin B Camelus chymosin Bovine chymosin B Vegetable Cynara cardunculus Cardoon Cyprosin 1, 2 and 3 and/or cardosin A and B Source: Law and Tammine, 2010 2.4.2.3 Salt Salt (sodium chloride) may be added to some varieties of cheese, the quantity and method of addition depending on the recipe. Salt may be added directly to the milk or curd pieces; it University of Ghana http://ugspace.ug.edu.gh 29 may be rubbed into the finished cheese or the cheese may be immersed in a brine solution (O’Connor, 1993). 2.5.0 Cheese Manufacture Cheese manufacture is one of the classical examples of food preservation, dating from 6000- 7000 BC (Fox, 1993). It preserves the most important constituents of milk (i.e. fat and protein) which are determinant factors of cheese yield, (Banks et al., 1981), as cheese exploits two of the classical principles of food preservation, i.e. lactic acid fermentation and reduction of water activity through removal of water and addition of salt (NaCl) and the establishment of a low redox potential, as a result of bacterial growth which contributes to the storage stability of cheese (Coker et al., 2005). Cheese making remained an art rather than a science until relatively recently. With the gradual acquisition of knowledge on the chemistry and microbiology of milk and cheese, it became possible to direct the changes involved in cheese making into a more controlled fashion (Fox, 1993). A number of developments have taken place which helps the cheese maker to produce a better and more consistent quality cheese. These developments include ; Pasteurisation, Acidification, Coagulation, Synerises, Salting, Ripening 2.5.1 Pasteurisation Milk may be heat treated or pasteurized at 73oC for 15 seconds (O’Connor, 1993). Pasteurisation of the milk kills nearly all the microorganisms present, including the harmful pathogenic bacteria that cause diseases, such as tuberculosis and leptospirosis, and other undesirable microorganisms such as yeasts and coli forms that may alter the cheese characteristics by producing carbon dioxide and undesirable proteolysis (Coker et al., 2005). University of Ghana http://ugspace.ug.edu.gh 30 The milk may be standardised, thus the fat content may be increased or reduced or the casein to fat ratio may be adjusted (O’Connor, 1993). 2.5.2 Acidification One of the basic operations in the manufacture of most cheese varieties is a progressive acidification throughout the manufacturing stage for some varieties during the early stages of ripening, thus acidification commences before and transcends the other manufacturing operations (Fox, 1993). Acidification of the milk is important for the proper release of whey from the cheese curd and to control the growth of many undesirable bacteria (Coker et al., 2005). Acidification controls the growth of many species of non-starter bacteria in cheese, especially pathogenic food poisoning and gas producing microorganisms. Properly made cheese is a very safe product from the public health viewpoint. In addition to producing acid, many starter bacteria produce probiotics that also restrict or inhibit the growth of non starter microorganisms (Fox, 1993). It is usually accomplished by the addition of lactic acid bacteria that convert lactose to lactic acid. Most varieties of cheese cannot be made without the addition of a "starter" which is a culture of carefully selected lactic acid- producing bacteria. The large volumes of starter required for cheese making are made in special bulk starter fermentation pots in which the milk is heat treated to destroy unwanted bacteria, spores and phages and cooled to about 22°C, a temperature suitable for starter growth. The frozen starter is mixed in and fermentation continues for about 6 to 16 hours. The amount of starter required varies for the different cheese varieties (Coker et al., 2005). Good quality starter is required, the type and quantity will be determined by the cheese recipe. For some cheese varieties commercial starter preparations are not used; natural fermentation or whey from the previous lot of cheese made may be used (O’Connor, 1993). University of Ghana http://ugspace.ug.edu.gh 31 2.5.3 Coagulation During coagulation, modifications of the milk protein complex occur under defined conditions of temperature and by action of a coagulant agent, which changes the physical aspect of milk from liquid to a jelly-like mass. Various coagulants are available, lemon juice, plant rennet and proteolytic enzyme such as chymosin (rennin) and proteolytic enzymes from the mould Rhizomucor miehei obtained by biotechnology. These enzymes have an acidic nature, thus they have optimum activity in a slightly acidic environment. Therefore, the action of lactic acid bacteria (LAB) in this phase is crucial as they are required to rapidly release enough lactic acid, to lower the milk pH from 6.7 to about 6.2 which creates an appropriate environment for optimum activity of rennin and to a pH of 4.5 as the processing proceeds, creating an unsuitable environment for unwanted microbes, thus increasing the safety of the end product (Kongo, et al., 2013). A rennet coagulum consists of a continuous matrix of strands of casein micelles, which incorporate fat globules, water, minerals and lactose and in which microorganisms are entrapped (Coker et al., 2005). The coagulants bring about, under defined conditions of temperature, quantity and time to coagulate the milk into a firm jelly- like mass (O’Connor, 1993). 2.5.4 Syneresis Syneresis is the rearrangement of casein molecules, which results in a tightening of the casein network. The end result is that moisture is squeezed out of the casein network. (Law and Tamime, 2010). Syneresis, or shrinking, of the coagulum is largely the result of continuous rennet action. It causes loss of whey, and is accelerated by cutting, stirring, cooking, salting or pressing the curd, as well as the increasing amount of acid produced by the starter, and gradually increases during cheese making. As a result, the cheese curd contracts and moisture is continuously expelled during the cooking stages (Coker et al., 2005). The rate and extent University of Ghana http://ugspace.ug.edu.gh 32 of syneresis are influenced by, milk composition, especially Calcium and casein, pH of the whey, cooking temperature, rate of stirring of the curd-whey mixture and time. The composition of the finished cheese is determined by the extent of syneresis (Fox, 1993). 2.5.5 Salting Salt is added to cheese as a preservative and because it affects the texture and flavour of the final cheese by controlling microbial growth and enzyme activity. The salt can be added either directly to the curd after the whey is run off and before moulding or pressing into shape, or by immersing the shaped cheese block in brine for several days following manufacture (O’Connor, 1993).The level and method of salting have a major influence on pH changes in cheese. The concentration of NaCI in cheese is between ( 0·7--4% and 2-10%) thus salt in the moisture phase is sufficient to halt the growth of starter bacteria. Some varieties, mostly of British origin, are salted by mixing dry salt with the curd towards the end of manufacture and hence the pH of the curd for these varieties must be close to the ultimate value (pH 5·1) at salting (Fox, 1993). Addition of salt to the cut curd draws more whey from the cheese curd and some of the salt diffuses into the curd. The pH of the curd, the contact time and the salt particle size and structure are all important in determining how much salt is absorbed by the curd. Salt is also involved in physical changes in cheese protein solubility and conformation, which influence cheese rheology and texture. Another important function of salt in cheese is as a flavour or a flavour enhancer (Coker et al., 2005). Salt also retards or prevents the growth of bacteria which may cause flavour and other defects in the cheese (O’Connor, 1993). University of Ghana http://ugspace.ug.edu.gh 33 2.5.6 Ripening or Maturation Some cheeses are consumed fresh, however most cheese varieties are not ready for consumption at the end of manufacture and undergo a period of ripening (curing, maturation) which varies from about three weeks to more than two years. The ripening process of cheese is very complex and involves microbiological, biochemical, structural, physical and sensory changes during storage to the curd resulting in the flavour and texture characteristic in the particular variety (Mcsweeny, 2004). Cheese texture softens during ripening as a consequence of hydrolysis of the casein micelle by proteolysis and changes to the water-binding ability of the curd and changes in pH which may cause other changes such as the migration and precipitation of calcium phosphate (Mcsweeny, 2004). It has a major effect on the quality of most cheese varieties with the exception of unripened cheeses including fresh acid curd cheeses (Quark and Cream cheese) and some ingredient cheeses (Law and Tamime, 2010). Cheese ripening involves the primary degradation of milk constituents by glycolysis, lypolysis and proteolysis (Marilley and Casey, 2004). Glycolysis occurs when lactose is metabolised completely to lactic acid and catabolised to form acetic and propionic acids, carbon dioxide, esters and alcohol by the enzymes of the starter cultures or secondary cultures in the milk (Fox et al., 2000). The presence of residual lactose persisting in cheese during maturation is undesirable as it makes the cheese less suitable for lactose intolerant consumers (Lomer et al., 2008), and also because it can be used as a growth substrate by non starter lactic acid bacteria, which can affect the flavour and quality when present in high numbers >108 cfu g−1 (Beresford &Williams, 2004). Fat is a major component in most cheese varieties, with the exception of some low fat fresh acid cheeses, such as Quark and Cottage cheese, and contributes directly and indirectly to rheology, texture, cooking properties and flavour. During ripening lipids are broken down to form free fatty acids, and catabolised to form ketones, lactones and esters by natural milk University of Ghana http://ugspace.ug.edu.gh 34 enzymes (milk lipase) and lipids that are added to enhance flavour in particular cheese varieties. Lipase action is high in raw milk compared to pasteurized milk cheeses. According to (Vlaemynck 1992), pasteurization of milk partially inactivates milk lipase. The optimum pH and temperature of milk lipase is 8.0 to 9.0 and at 35oC to 40oC respectively. A combination of low pH and high salt concentration also inhibit the activity of milk lipase. Though Lipolysis is needed for flavour enhancement, too much of it imparts a rancid flavour which is undesirable in fresh cheeses (Ashaye et al., 2006). Proteolysis also involves the gradual breakdown of proteins (caseins) to form peptides and amino acids by the enzymes of the coagulant (residual chymosin), the natural milk enzymes (peptidases) and the enzymes of the starter culture (Bylund, 1995). Proteolysis in cheese is important for flavour and texture development due to the breakdown of the protein network, decrease in water activity through water binding by carboxyl and amino acid groups and increased pH (Sousa et al., 2001).The sequence of residues of the casein is strongly hydrophobic and confers intact casein with strong self association and aggregation tendencies in the cheese environment. Its cleavage is generally considered to be a major factor contributing to the decrease in the rubberyness of young internal ripened hard cheeses, such as Cheddar, Gouda, and Mozzarella, and their conversion to smooth bodied mature cheeses (Law and Tamime, 2010). Proteolysis also has a major effect on the cooking properties of cheese, increase in proteolysis generally coincide with increases in the levels of protein hydration, free fat and of heat induced flow ability (Guinee, 2003). The degree of stretchability of the melted cheese also increases progressively with proteolysis to a level, which depends on the variety and decreases afterwards (Law and Tamime, 2010). University of Ghana http://ugspace.ug.edu.gh 35 2.5.7 pH of cheese The pH of cheese is a very important physicochemical parameter affecting the texture, flavour and microbiological safety of cheese. The addition of starter cultures for milk fermentation, the de-acidification of some varieties of cheese during maturation and the ability of the curd formed to resist changes in pH determine the final pH of cheese. The main components of cheese that affects the pH of cheese are 0=-the caseins. The degradation of casein produces inorganic phosphates and organic acids and their levels in cheese are influenced by milk composition, curd treatment and its effect on syneresis (Lucey et al., 1992). Lipolysis also increases the acidity of cheese by the production of free fatty acids (Law and Tamime, 2009).Decrease in pH of cheeses during maturation is due to the continued production of lactic acid lactic acid bacteria and the liberation of amino acids such as aspartic and glutamic acids during proteolysis (Sallami et al., 2004). The pH of soft fresh cheeses ranges from approximately 4.1 to 5.4 (ICMSF, 1996), however, the buffer maximum which is around pH of 5.0 is very important in cheese making since the optimum pH for most cheese ranges from 5.0 - 5.2. As the pH of cheese is reduced towards pH 5.0 by lactic acid fermentation, the buffer capacity is also increasing. The effect is to give the cheese maker substantial room for disparity in the rate and amount of acid production. Without milk's built in buffers it would be difficult to produce cheese in the optimum pH range. (http://www.uoguelph.ca/foodscience/cheesemaking-technology/section-b-analytical/process- and-quality-control-proceudures/ph University of Ghana http://ugspace.ug.edu.gh 36 2.6.0 Rheology of cheese 2.6.1 Texture measurement Cheese texture may be defined as a ‘composite sensory attribute’ resulting from a combination of physical properties that are perceived by the senses of touch (including kinaesthesis and mouth-feel), sight and hearing. It can be measured directly using a trained sensory panel; however, owing to the difficulty and cost in assembling sensory panels, they are not routinely used for gauging cheese texture. Instead, cheese texture is generally measured indirectly using rheological techniques (O’Callaghan & Guinee, 2004). The rheology of hard or semi-hard cheese is commonly assessed by compression of a cylindrical or cubic cheese sample between two parallel plates of a texture analyser (Fenelon & Guinee, 2000; Everard et al., 2007c). The cheese sample is placed on a base plate, and is compressed at a fixed rate (typically 20 mm min−1) to a predetermined (e.g. 75% of its original height) by the mobile plate (cross head). The compression may be carried out in one or two cycles (bites). Analysis of the force – displacement or stress – strain curves, often referred to as texture profile analysis, enables the determination of a number of rheological parameters e.g. fracture stress, fracture strain, firmness and springiness, which are related to sensory textural characteristics, such as brittleness, shredability, hardness and chewiness (O’Callaghan & Guinee, 2004; Dimitrelia & Thomareis, 2007). 2.6.2 Colorimetry Colour is an important measure of quality in the food industry because it is considered by consumers to be related to product freshness, ripeness, desirability and food safety (McCraig, 2002; Jeli´nski et al., 2007). Colour measurement instruments, in accordance with the standards developed by the Commission Internationale de l’ ´ Eclairage, transform or filter reflected spectra to produce reproducible colour space coordinates, namely, L* (index of University of Ghana http://ugspace.ug.edu.gh 37 whiteness), a* (index of redness), and b* (index of yellowness) (Commission Internationale de l’ ´ Eclairage, 1986; MacDougall, 2001). Colour measurements are normally carried out in a laboratory based instrument (HunterLab meter or Minolta Chroma meter) but they can also be acquired by online instruments. Owing to ageing effects of light sources and detector systems, regular calibration of colorimetric equipment against colour standards is essential. Colorimetry is used routinely in quality control and product development to assess the colour of curd and cheese. Colour is related to diet of cow, addition of colouring and cheese variety. Recent studies also showed the potential role of colorimetry in assessing ripening of smear- ripened cheese (Dufoss´e et al., 2005; Olson et al., 2006) and for measuring defects, such as browning, during cheese maturation (Carreira et al., 2002). 2.6.3 Classification of cheese There are hundreds of varieties of cheese, but each relies on similar principles of coagulating the proteins in milk to form curds and then separating them from the liquid whey. The percentage of water present in cheese, the microorganism used in ripening, and the length of the maturing period of the cheese differentiates the many types of cheese present today (Coker et al., 2005). Cheeses may be broadly classified into ‘soft’, ‘semi-hard’ and ‘hard’ cheeses (Fellows, 2014).Table 2.1 shows the classification of cheese based on their moisture conten University of Ghana http://ugspace.ug.edu.gh 38 Table 2.2: Classification of cheese Type of cheese Moisture content (%) Fat content (%) Texture Shelf-life Soft cheeses 45–75 <40 Soft, white, spreadable A few days or weeks Semi-hard cheeses 35–45 <35 Firm, crumbly, can be sliced Several weeks Hard cheeses 30–40 <30 Very firm, dense Several months CTA 2014 http://knowledge.cta.int/ Peter Fellows 2.7.0 Food Product Development Product Development is a systematic, commercially oriented research to develop products and processes satisfying a known or suspected consumer need. Product development is a method of industrial research in its own right. It is a combination and application of natural sciences with the social sciences of food science and processing with marketing and consumer science into one type of integrated research whose aim is the development of new products. The most widely referenced normative product development models are those of Booz, Allen and Hamilton Inc. (1982) and Cooper and Kleinschmidt (1986). There are essentially four basic stages in these models for every product development process. These are: product strategy development; product design and development; product commercialization; product launch and post-launch. The food industry appears to be populated with companies that prefer to re-develop existing products (incremental change), rather than create new products (radical change). Because food product development is University of Ghana http://ugspace.ug.edu.gh 39 considered a highly risky venture, the incremental change strategy may be an attempt to increase success rates. Ironically, this apparently ‘safe’ approach perpetuates the problem of high food product failure, since truly innovative products are often more successful for a company (Stewart-Knox & Mitchell, 2003). However, there are some indications that certain factors may improve the number of the success rate in product development. Three important factors that contribute to new product success were cited by Ilori et al., (2000). They were: marketing and managerial synergy, strength of marketing communications and launch effort, and market need, growth and size. These factors emphasize the role of marketing in the product development process. Tetra Pak (2004) found one or more of the following features are typical of new products that succeed in the marketplace. Therefore, these could be used as criteria while screening ideas in the product development process: noticeable advantages for the consumer; the more the better; distinctive details that are important to the consumer; satisfy the consumers’ need for convenience, youth, better diet, less stress, perfect taste and variation; reliable brand; advertising breakthrough. Stewart-Knox & Mitchell (2003) found that understanding consumer needs and expectations and retailer involvement in product development were associated with product success. 2.8.0 Packaging and Preservation of cheese Increasing the shelf life of food using different preservative methods has always been a major concern. Several preservation techniques are available to extend the shelf storage of food products, among which packaging is the most capable. The packaging process undertakes several basic roles such as preventing microbial and chemical quality deterioration and enhancing the handling and marketing for packaged products. Dairy products are an important food group highly suggested by nutritionists and it is one of the most perishable category foods thus extending their shelf life and keeping quality for a long time is important. University of Ghana http://ugspace.ug.edu.gh 40 Consumers are aware of the possible hazards of preservatives therefore technologists and researchers have attempted to introduce new non preservative methods (Coker et al., 2003). One technique is the modified atmosphere packaging (MAP), which changes the natural gas surrounding the product in the package to slow down spoilage. Vacuum packaging is a type of packaging whereby the air within the pack is removed and the pack sealed creating a vacuum around the product. The gaseous atmosphere of the vacuum package is likely to change during storage due to metabolism of the product or microorganisms thus the atmosphere becomes indirectly modified. Vacuum packaging has so far been the most widely used packaging technique for cooked products (Borch et al., 1996, Korkeala et al., 1985 and Samelis et al., 2000). Perishable products have exhibited superior quality under vacuum or MAP storage than under ambient conditions. Recent cheese packaging protects the food from microorganisms and prevents moisture loss. Unripened cheeses are packaged immediately after the curd is collected and must be immediately refrigerated. Ripened cheeses go through various procedures during packaging for preservative reasons. Some ripened cheeses are coated in wax to protect them from mould contamination and to reduce the rate of moisture loss. Cheeses that naturally develop a thick, tightly woven rind, such as Swiss, do not require waxing. A second method of ripened cheese packaging involves applying laminated cellophane films to unwaxed cheese surfaces. The most common packaging film consists of two laminated cellophane sheets and a brown paper overlay or a metal foil wrap. Cheeses that are made and matured in large blocks are usually cut into appropriate size and shrink wrapped in an atmosphere of carbon dioxide, which dissolves into the body of the cheese, or vacuum sealed in a special "top-and-bottom" "webbed" package. The subsequent anaerobic environment prevents mould growth on the cheese surface. Many cheeses, such as the Brie and Camembert are packaged in special aerating wraps and stored in porous boxes Coker et al., 2003). University of Ghana http://ugspace.ug.edu.gh 41 2.8.1 Food Irradiation Food safety is a subject of growing importance to consumers. One reason is the emergence of new types of harmful bacteria or evolving forms of older ones that can cause serious illness. Scientists, regulators and lawmakers, working to determine how best to combat food borne illness, are encouraging the use of technologies that can enhance food safety worldwide, (FDA, 1997). Irradiation can be an effective way to help reduce food-borne hazards and ensure that harmful organisms are not in the foods we buy. During irradiation, foods are exposed briefly to a radiant energy source such as gamma rays or electron beams within a shielded facility. Irradiation is not a substitute for proper food manufacturing and handling procedures. Cheese is an important integral part of diet around the world especially in Europe, America and it is consumed almost three times in a day. Soft fresh cheeses like cottage cheese and ‘wagashie’ have a short shelf life due to the high moisture content which creates a suitable environment for spoilage and pathogenic microorganisms. Food irradiation however is a preservation method that improves the safety and shelf life of food products. It could be used to replace chemical preservatives as well as thermal treatment. It is considered as cold pasteurization of food and does not leave any residue or render the food radioactive. It is permitted in 35 countries worldwide for 40 different food products (Robert, 1998; Loaharanu, 2005 and Thayer, 2005). The use of gamma radiation from a Cobalt 60 source in dairy product is considered as one of the most important peaceful application of nuclear energy (FDA, 1997 and WHO, 2005). There was no hazard caused by irradiation up to 10 kGy which could not cause cancer, genetic mutation or tumours (Mason, 1993; Sofos, 2002; Mehran et al., 2005). Therefore, hospitals use irradiated food for patients with severely impaired immune system (Lee, 1994; FAO, 1998; Leuschner & Boughtflower, 2002; Bernnand, 2006 and Konteles et al., 2009). University of Ghana http://ugspace.ug.edu.gh 42 Bongirwar and Kumta (1967) reported that Cheddar cheese developed off-flavors when irradiated at 0.5 kGy; however, none was detected when the dose was reduced to 0.2 kGy. A dose greater than 1.5 kGy, when applied to Turkish Kashar cheese, not only resulted in off- flavor development but also contributed to colour deterioration (Jones and Jelen, 1988).By decreasing the dose to 1.2 kGy the sensory problems were eliminated and the mold-free shelf life was extended from 12 to 15 days when stored at room temperature. In contrast, non- irradiated cheese became moldy within 3 to 5 days. When combined with refrigeration storage, irradiation increased the shelf-life period of the cheese fivefold. With Gouda cheese, however, no taste difference was reported between irradiated (3.3kGy) and non-irradiated samples (Rosenthal et al., 1983). Among the preservation methods to ensure safety of whey cheeses are irradiation combined with vacuum packaging (Tsiotsias et al., 2002) and using of antimicrobial compounds (Samelis et al., 2000) both of which had been applied to typical Greek whey cheese. 2.9.0 Sensory Evaluation Sensory evaluation has been defined as a scientific discipline used to evoke measure, analyze and interprete reactions to those characteristics of food and materials as they are perceived by the senses of sight, smell, taste, touch and hearing. Sensory characteristics are an important determinant in the choice of the food products by the consumer. Therefore, the measurement of sensory characteristics is an important point for the producer. Sensory analysis is the most direct and thus the most valid way of measuring the organoleptic characteristics of food (Piggott, 1995). Instrumental measurements could replace the sensory analysis only if they have been validated by experiments showing strong correlations between both sets of data and if the predictive value of the instrumental measurements have been demonstrated. Sensory University of Ghana http://ugspace.ug.edu.gh 43 measurements are often described as 'subjective' measurements; this is true as the measuring instrument is a panel of subjects, but does not mean that the subjects are not objective. Subjects in a panel are requested to disregard their personal liking and concentrate on the description of their perceptions. To perform such rneasurements, trained subjects are used, whereas consumers (naive subjects) are used for hedonic measurements (Issanchou and Schlich, 1997). Cheese is assessed and graded by cheese panellists and quality control personnel in industries to ensure that its texture and flavour conform to a generally agreed consensus for a particular variety (van Hekken et al., 2006; Sameen et al., 2008). Grading of cheeses for attributes like appearance, flavour, body and texture or for precise defects like bitterness, mottled appearance on an agreed scale is carried out to determine the grade acceptability of cheese for specific markets. Quality scoring remains the most commonly used type of sensory evaluation in the cheese industry to determine acceptability or rejection on the basis of scores obtained, (Law and Tamime, 2010). 2.9.1 Descriptive Sensory Analysis According to Delahunty & Drake (2004), Descriptive Sensory Evaluation refers to a collection of techniques that discriminate between the sensory attributes of an product including a range of cheeses and to determine quantitative and qualitative description of all the sensory differences that can be identified. The sensory characteristics are defined in terms of a lexicon of agreed attributes assigned by trained consumer panellists. Each attribute is scored on a linear scale and the resultant data are typically presented in the form of spider web diagrams or principal components loading plots, for the purpose of discriminating between cheeses, (Law and Tamime, 2010). University of Ghana http://ugspace.ug.edu.gh 44 Descriptive Data Analysis is mainly used as a tool for development of new cheeses and as a quality control tool, provided that a standard cheese of acceptable quality is available for comparison with other samples. Thus, the degree of excellence of cheeses may be differentiated as determined by consumer acceptance, market research, and ‘difference from a standard determined by consumer panels (Law and Tamime, 2010). The Descriptive Data Analysis is simply a set of attributes or descriptors that a panel has agreed upon that enables them to fully describe the sensory properties of the products being evaluated. Descriptive sensory analysis addresses some of the problems of language use, interpretation and scaling difficulties. To achieve this, a sensory quality program is organized where time and effort is taken to recruit and train panelists. This procedure also helps to obtain reliable data on the product being evaluated. Methods for generating descriptors are classified according to whether the results are qualitative or quantitative even though one could be transformed to another. An example of a qualitative method is Aroma profile. Examples of the quantitative type are Texture Profile, Quantitative Descriptive Analysis, Free Choice Profiling, Spectrum Analysis, Diagnostic Descriptive Analysis (Stone and Sidel, 1993) and Repertory Grid (Gains 1990). After the generation of descriptors, it is necessary to determine which of the descriptors sufficiently describe the product. A descriptive sensory analysis is useful when details of products need to be characterized in both academic and industrial research. 2.9.2 The selection of a descriptive analysis Panel The most important point in sensory measurements is the panel. Thus, care is required in setting up the panel. The first step is to recruit the subjects within or outside the company. In both cases, the first criteria to be taken into account are motivation and availability (Issanchou University of Ghana http://ugspace.ug.edu.gh 45 et al., 1997). All descriptive methods require a panel with some degree of training or orientation. This is achieved by screening and selection of panelist per the project demand. An important issue is the number of assessors required to carry out profiling. In literature, the descriptive panel size varies from eight, the recommended minimum number to 24 (Muir and Hunter, 1991-1992).The most important thing that ensures success in training is the commitment and motivation of the panelist to the project, panelist should attend all training and evaluation sessions. Individual interview can be used to determine the commitment and motivation, and availability can be determined by filling out a time table of available hours per week (Murray et al., 2001). Personality of the panelist is very important in determining the success or failure of sensory panelists. Studies by Piggot and Hunter (1999) showed that elaborative screening procedures did not determine the ability of a panelist to perform well but a concentration and personality test may be the best predictor of a good panelist together with verbal creativity and test of discrimination ability (Murray et al., 2001). 2.9.3 Training of the panel The panel is trained to use a common frame of reference to define the product attributes and their intensity in the product under test. This is done by exposing the panel to the range of products under test. The panel mentally refers to the background information and reference points which serve as a frame of comparison when evaluating products (Munoz and Civille, 1998). Before the training, the panel is allowed to use their own frame of reference to describe the product qualitatively with their own words to describe perceptions and quantitatively by using previous experiences to rate intensities ability (Murray et al., 2001). Trained panelists acquire a common qualitative and quantitative frame of reference that uses a standard language to describe sensory concepts and a common scale. The Panel is advised to rate products based on the term generation and concept formation sessions and not on their University of Ghana http://ugspace.ug.edu.gh 46 own personal experiences (Murray et al., 2001). (ASTM, 1996; ISO 8586-1, 1993; ISO 8586- 2, 1994) recommends the use of a trained or expert panel in performing sensory profiling. This is necessary because training the panel enables the panelists to take on an analytical frame of mind. Conversely, untrained consumers tend to act non-analytically when scoring attributes (Lawless and Heymann, 1998). Free choice profiling however, does not require trained panelists and has been used for cheeses (R´etiveau et al., 2005; Drake et al., 2001), dairy desserts (Gonz´asalez and Costell, 2006), and fresh products. Training procedures to facilitate concept alignment in descriptive analysis should be very extensive, but will depend on the approach or method chosen, the time available and the products under test. 2.9.4 Descriptive Attribute Generation Selecting and defining descriptors for the products being tested is the critical step in a descriptive sensory analysis. A well developed lexicon would help sensory researchers to conduct a precise sensory analysis and compare the results from different sites. The training phase of descriptive analysis starts with the formation of a common language which describes the product attribute very well. Usually using a new panel will help in generating a new sensory language but an experienced panel leader may be included to assist the process (Murray et al., 2001). The panel is usually exposed to a wide range of products in the category under test. Then products are assessed together and the descriptive profile of one product is compared to and in the contest of the other products. This is the most important stage as the product under assessment is very well defined (Murray et al., 2001). Drake et al., (2001, 2002) have developed the descriptive language for Cheddar cheese flavour by using a large representative sample set and validated the lexicon at three different locations. The expanded lexicon for the flavour attributes of French cheeses has been developed by Rétiveau et al., (2005). University of Ghana http://ugspace.ug.edu.gh 47 2.9.5 Quantitative Descriptive Analysis (QDA) Quantitative descriptive analysis was developed in 1970 to correct some perceived problems associated with flavour profile method (FMP) (Stone and Sidel, 1998). Quantitative descriptive analysis is a non technical common language to avoid biasing response behaviour that may occur by giving out a language. It involves correct and non- correct answers (Murray et al., 2001). QDA uses reference standards when a problem with a particular term is noticed and a reference is needed for the subjects (Stone and Sidel, 1993). The panel leader is not allowed to be a participant of the process to prevent bias, unstructured line scale is used to score the intensity of the rated attributes. The panel is trained for approximately 10 to 15 hours to enable them understand the meaning of the attributes. QDA uses relative differences among the product for evaluation and not the absolute differences, thus results of QDA will show that the panellists are calibrated based on the relative differences among the samples. Analysis of Variance is usually used to analyze QDA results and the cobweb or spider diagram is used to graphically represent the data (Murray et al., 2001). 2.9.6 Principal Composite Analysis (PCA) Drake, (2007) defined PCA as a multivariate data compression technique that allows multiple treatments to be graphically displayed as they are differentiated by multiple variables. Principal component analysis forms the basis for multivariate data analysis. PCA has been described by Pearson as finding lines and planes of closest fit to systems of points in space.The most important use of PCA is indeed to represent a multivariate data table as a low-dimensional plane, consisting of 2 to 5 dimensions, such that an overview of the data is obtained. This overview may reveal groups of observations, trends, and outliers. This overview also uncovers the relationships between observations and variables, and among the University of Ghana http://ugspace.ug.edu.gh 48 variables themselves. Statistically, PCA finds lines, planes and hyper planes in a dimensional space that approximate the data as well as possible in the least squares sense (Drake, 2007). It is easy to see that a line or a plane that is the least squares approximation of a set of data points makes the variance of the co-ordinates on the line or plane as large as possible. The technique is mostly used to assess how several products are differentiated by several sensory descriptors. Attributes that are highly positively correlated will lie close to each other (Lawless and Heyman, 1998). 2.10.0 Experimental Design 2.10.1 Response Surface Methodology and Box Behnken Design Response surface methodology is a statistical mathematical method which uses quantitative data in an experimental design to determine and simultaneously solve multivariate equations (Giovanni, 1983). RSM is a collection of mathematical and statistical techniques useful for the modeling and analysis of problems in which a response of interest is influenced by several variables (Montgomery, 2005; Kiran et al., 2007). It aims at building a regression model that is closest to the true regression model based on observation data and the model is empirical, thus the true regression model is usually never known (Montgomery, 2005 and Kiran et al., 2007). The empirical model technique is devoted to the evaluation of the relationship between a set of controlled experimental factors and the observed results (Annadurai and Sheeja, 1998). The main aim of response surface methodology is to optimize the factors that produce the maximum and minimum value of the response. Since the quality characteristics of a production may not be linear with the input variables, a second degree polynomial is used if there is a curvature in the system. Empirical evidence has shown that the higher order polynomial or the quadratic model is enough for the optimum region (Meyers and Montgomery, 2002). The response surface can be used to graphically make judgements about University of Ghana http://ugspace.ug.edu.gh 49 the relationship between explanatory and response variables (Tong, et al., 2011). This helps to know the treatment combinations that give the optimum response (Hinkelmann and Kempthorne, 2007). Applications of the response surface methodology are the development and formulation of new products and the improvement of an existing product. To analyze a process mutually with a response, Y which depends on the input factors X1, X2, …, Xn, the correlation between the response and the input process parameters are described as Y = f (X1, X2, …, Xn) + ε (1) Where f is the real response function, its format being unknown and ε is the residual error which describes the differentiation that can be incorporated by the function f. Because the correlation between the response and the input variables can be described as a surface of the X1, X2…, Xn coordinates in the graphical sense, so the investigation of these relationships is named as the response surface study. The most common designs, i.e. central composite design (CCD) and Box-Behnkendesign (BBD), are the principal response surface methodology that has been widely used in various experiments (Box et al., 1978). 2.10.2 Box Behnken Design The Box Behnken design which is a type of Response Surface Methodology is an independent quadratic design in that it does not contain an embedded factorial or fractional factorial design. It was developed by Box and Behnken in 1980 and it is used to develop second – order response surface models. Some three-level designs which have been proposed by Box and Behnken are formed by combining 2k factorials with incomplete block designs. . The level of one factor is fixed at the center level while combinations of all levels of the other factors are applied (Montgomery, 2005). Box-Behnken which is a spherical and revolving design has been applied in optimisation of chemical and physical processes (Oscar et al., 1999; Muthukumar et al., 2003) because of its reasoning design and excellent outcomes. Box- University of Ghana http://ugspace.ug.edu.gh 50 Behnken design does not contain any points at the vertices of the cubic region created by the upper and lower limits for each variable and results in the reduced number of required experimental runs. This could be advantageous when the points on the corners of the cube represent factor-level combinations that are prohibitively expensive or impossible to test because of physical process constraints (Box and Behnken, 1960; Montgomery, 2005). Table 2.3 shows a three level experiment using the Box Behnken method with the 15 combinations generated by the design. Table 2.3: The Box Behnken Experimental Design Rank A B C 1 -1 -1 0 2 1 -1 0 3 -1 1 0 4 1 1 0 5 -1 0 -1 6 -1 0 1 7 -1 0 1 8 1 0 1 9 0 -1 -1 10 0 1 -1 11 0 -1 1 12 0 1 1 14 0 0 0 15 0 0 0 University of Ghana http://ugspace.ug.edu.gh 51 CHAPTER THREE 3.0 MATERIALS AND METHODS 3.1 Breif field study A brief field study was conducted to identify wagashie production sites and to obtain samples for analysis. 3.2 Sampling of wagashie ‘Wagashie’ samples were obtained from Nima and Ashaiman markets. Six samples were purchased from both areas. Samples were packaged in flexible polyethylene bags, kept in an ice chest containing ice packs and transported to the Food Research Institute (CSIR) for micobiological analyses. 3.3 Microbiological analysis of the market samples and laboratory prepared samples of wagashie To assess the safety of the wagashie samples,enumeration or detection of the following indicator and enteric pathogens were carried out; Aerobic mesophiles, Feacal coliforms, E.coli, Staphylococcus aureus, Salmonella spp., Bacillus cereus, Enterobacteriaceae, Enterococcus, yeast and moulds. 3.3.1 Serial Dilution University of Ghana http://ugspace.ug.edu.gh 52 Ten (10) grams of each sample was added to 90 ml of sterile Saline Peptone Solution (SPS) containing 0.1% peptone and 0.8% NaCl with pH adjusted to 7.0. The sample was homogenized in the stomacher (Lab Blender, Model 4001, Seward Medical, London, England) for 90 s at normal speed to obtain 1:10 dilution. Further dilutions were made to obtain a 10 fold dilution after which 1ml of each dilution was transfered into sterile petri dishes and the appropriate media added. All analyses were done in triplicate. 3.3.2 Enumeration of Aerobic Mesophiles Aerobic mesophiles were enumerated by the pour plate method using Plate Count Agar (Oxoid CM ; Oxoid Ltd Basingstoke,Hamshire,UK). About 10 to 15 ml of the media was poured on the plates, swirled gently and allowed to solidify for 10 min. The plates were incubated at 30 °C for 72 h in accordance with the Nordic Committee on Foods Analysis Method ( NMKL. No. 86, 2006). Plates with 25 to 250 colonies were selected and counted using a colony counter. 3.3.3 Enumeration of Yeast and Moulds Yeast and moulds were enumerated by the pour plate method using Oxytetracycline Glucose Yeast Extract agar (Oxoid CM, ; Oxoid Ltd Basingstoke Hamshire,UK) supplemented with Oxytetracycline to inhibit the growth of bacteria,the pH was adjusted to 7.0 About 10 to 15 ml of the media was poured on the plates, swirled gently and allowed to solidify for 10 min. The plates were incubated at 30 °C for 120 h in accordance to ISO 7954,1987(E) . Colonies were counted after 3,4 and 5 days of incubation and a microscopic examination was done to distinguish colonies of yeast and moulds. University of Ghana http://ugspace.ug.edu.gh 53 3.3.4 Enumeration of Total Coliform Feacal coliforms were enumerated by pour plate method on Trypton Soy Agar (Oxoid CM0131; Oxoid Ltd Basingstoke Hamshire,UK a non-selective agar) at pH 7.3 and overlaid with Violet Red Bile Agar (Oxoid CM0107 ; Oxoid Ltd Basingstoke Hamshire,UK a selective agar,) at pH 7.4 and allowed to cool for 10 min. The plates were incubated in inverted positions at 44 °C for 24 h . Plates with 10 to 100 colonies were selected for counting, colonies with purplish red colonies with 0.5 mm diameter or greater were counted. Colonies were confirmed by selecting 5 suspected colonies and inoculating into Brilliant Green Bile Broth containing Durham tubes (Oxoid CM0329; Oxoid Ltd Basingstoke Hamshire,UK) at a of pH 7.4. The tubes were incubated at 37 °C for 24 h ( in accordance with NMKL No. 44 2004) . Gas production at the bent portion of the Durham tubes indicated a positive reaction. 3.3.5 Enumeration of E. coli E. coli were enumerated by the pour plate method using a non-selective agar; Trypton Soy Agar (Oxoid CM0131 ). About 5 ml of the TSA was transfered into the petri dish at a pH of 7.3 , overlaid with a selective agar; Violet Red Bile Agar (Oxoid CM0107) at a pH of 7.4 to a ratio of 2:1.The inocula and the substrates were mixed thoroughly and the plates were incubated in inverted positions at 44 °C for 24 h. Plates with 10 to 100 suspicious and typical colonies which were dark red with 0.5mm diameter were selected. Colonies were confirmed using EC Broth (Oxoid CM 0469) at a pH of 6.9, followed by Trypton Water (Oxoid CM087) at a pH of 7.5. University of Ghana http://ugspace.ug.edu.gh 54 The tubes were incubated at 44 °C for 24 h (NMKL. No. 125, 2005) and tested for indole reaction. A red colour change was recorded as a positive reaction. 3.3.6 Enumeration of Staphylococcus aureus Enumeration of Staphylococcus aureus was by the spread plate technique on Baird-Parker Agar (BP, Oxoid CM 0275, Hampshire, England.). 0.1ml of the aliquot was innoculated on the surface of the media and spread on the surface with a sterile glass rod. Egg Yolk Tellurite Emulsion (SR54) was added and the plates were incubated at 37oC for 48 h, (NMKL No. 66 4th Ed 2009) . Colonies were confirmed using Rabbit Plasma Serum for coagulate positive test 3.3.7 Enumeration of Bacillus cereus Bacillus cereus was enumerated by the spread plate method on Bacillus cereus agar (Oxoid CM 617 and SR99) supplemented with polymycin B and egg yolk emulsion. 0.1 ml of the aloquot was innoculated on the surface of the media and a sterile rod was used to spread the innoculum on the surface of the media. The plates were incubated at 30 0C for 24 h (NMKL No 67, 2010). Suspected colonies were confirmed on Blood Agar Base (Oxoid CM 617) and microscopy done. 3.3.8 Detection of Salmonella spp. University of Ghana http://ugspace.ug.edu.gh 55 Approximately 25g of the food sample was weighed aseptically into sterile Stomacher bag, 225 ml of Buffered Peptone Water was added and homogenized in a stomacher at medium speed. The sample solution was incubated at 37 °C for 16 h. After incubation, 0.1ml of the sample solution was inoculated into 10 ml of Rappaport-Vasilliadis (RV) broth and incubated in a water bath at 42 °C for 24 h. Following enrichment, in the Rappaport-Vasilliadis (RVS) broth, the culture was streaked on XLD (Xylose Lysine Deoxycholate) in Petri -dishes, and the plates were incubated at 37°C for 24 h. Biochemical and seriological test were done to confirm the presence of Salmonella 3.3.9 Enumeration of Enterococcus Enumeration of Enterococcus was done by the Pour Plate Method on a Tryptone soy agar (TSA) overlaid with Slanetz and Bartley (S&B). About 1 ml of the aliquot was pippeted into the petri dish to cover the base and the medium was added. The plates were incubated at 440C for 48hrs (NMKL No 68. 5th Ed.2011). 3.3.10 Enumeration of Enterobacteriaceae Enterobacteriaceae were enumerated by pour plate method on Voilet Red Bile Glucose Agar (VRBGA) (Oxoid CM0107 ; Oxoid Ltd Basingstoke Hamshire,UK a selective agar,) at pH 7.4 . The plates were incubated in inverted positions at 37 °C for 24 h . Plates with 10 to 100 colonies were selected for counting, colonies with purplish red colonies with 0.5 mm diameter or greater were counted. Colonies were confirmed by selecting 5 suspected colonies and inoculating into Brilliant Green Bile Broth at a pH of 7.4 containing Durham tubes (Oxoid CM0329; Oxoid Ltd Basingstoke Hamshire,UK). University of Ghana http://ugspace.ug.edu.gh 56 The tubes were incubated at 37 °C for 24 h ( in accordance with NMKL No. 144 2004) . gas production at the bent portion of the Durham tubes indicated a positive reaction. 3.4.0 Laboratory preparation of traditional ‘wagashie’ 3.4.1 Preparation of Coagulant Fresh stems of Calotropis procera obtained from the premises of Food Research Institute, Accra were washed thoroughly and cut on a chopping board .The stems were crushed in a mortar with a pestle. The crushed stems were collected in a bowl and weighed. About 500ml of water was added to the stems and a weighed amount of salt was added. The mixture was left to stand for a maximum of 10 minutes and filtered. Figure 3.1 shows the flow diagram of the plant extract preparation. Fresh stems of Calotropis procera (Sodom apple) Wash thoroughly Cut stems into pieces Crush weighed stems in a mortar Add 500mls of water Add salt (NaCl) Leave mixture to stand for 10 minutes Filter with a colander Plant Extract Figure 3.1: Flow diagram for plant extract preparation (coagulant). Collect in a bowl and weigh University of Ghana http://ugspace.ug.edu.gh 57 3.4.2 Preparation of ‘wagashie’ (Traditional method) Two (2) litres of raw cow milk (obtained from Animal Research Institute, Council for Scientific and Industrial Research), was pasteurized at 80oC for 30 min and cooled to a temperature of 40 o C. The coagulant was filtered and added to the milk. The mixture was stirred briefly and cooked under low temperature to enhance coagulation. Curd formation began after 10 min. when there was a clear separation between the curd and the whey, the curd was boiled for 20 min to enhance curd firmness. The curd was poured into a colander to expel the whey and the curds collected were transferred into a muslin cloth where it was tied and pressed with a 6 kg load for further whey drainage for about 10 min. The matted curd was cut into equal shape and size with a mould and packaged prior to sensory evaluation. Figure 3.2 shows the flow diagram for ‘wagashie’ preparation (Traditional method). Raw cow milk Pasteurize milk at 80 ͦ C for 30 min Add plant extract Curd formation Boil curd for 20 min Drain Whey Press curd in a muslin cloth Remove curd and cut Package Figure 3.2 Flow diagram of ‘wagashie’ production process (Traditional method) University of Ghana http://ugspace.ug.edu.gh 58 3.4.3 Modification of the Wagashie process In order to improve on the product quality of the traditional wagashie,various modifications of the production method were carried out and the product evaluated by sensory evaluation and physicochemical analysis. The modifications were as follows; 1. The use of commercial rennet used in the cheese industry tocoagulate the fresh milk instead of the extract of the Sodom apple. 2. Fermentation of fresh cow milk with a starter culture (Chr Hensen 10-12 Boege Alle DK- 2970 Hoersholm-Denmark) prior to coagulation to modify the taste and improve on the safety of wagashie. 3. Optimisation of the quantities of ingredients and processing parameters with respect to the sensory quality of wagashie. 3.4.4 Preparation of ‘wagashie’ using commercial rennet The process for making the improved ‘wagashie’ was similar to the traditional ‘wagashie’ process but the Calotropis procera was replaced with commercial rennet. In the rennet prepation, the milk was pasteurised to 80 ͦ C for 30 min and allowed to cool to about 40 ͦ C before the rennet was added with a stirile pippet and stirred. It was left to stand for 10 mins to enable curd formation and boiled for 20 min to make the curds firm before boilling and draining the curds. Figure 3.3 is the flow diagram showing the process of making the improved ‘wagashie’ with rennet and plant extract as coagulant. University of Ghana http://ugspace.ug.edu.gh 59 Preparation with Rennet Preparation with plant extract Pasteurisation of milk Raw milk Pasteurisation of milk (80 ͦ C,30min) cool to 40 ͦ C cool to 40 ͦ C Fermentation (starter culture) Fermentation (starter culture) Addition coagulant (Rennet) Heat milk to 60 ͦ C Curd formation Addition of coagulant (plant extract) Curd boiling Whey drainage curd formation Curd boiling Curd pressing whey drainage Cutting Curd pressing & cutting Packaging Packaging Figure 3.3: Flow diagram of ‘wagashie’ process with Plant extract and rennet for both non-fermented and fermented preparations with cheese and yoghurt cultures. Non fermented wagashie Non- fermented wagashie University of Ghana http://ugspace.ug.edu.gh 60 3.5.0 Fermentation of fresh cow milk using starter culture As part of the modification process, raw cow milk was fermented and coagulated with Calotropis procera or rennet. The fermentation was done by inoculating the fresh milk with the starter culture at a temperature of 42 ͦ C in a water bath. The procedure for traditional wagashie preparation was then applied. Figure 3.3 shows the flow diagram for wagashie prepared with fermented and non-fermented raw cow milk using plant extract and rennet as the coagulating agents. 3.5.1 Starter culture for the fermentation of raw cow milk 3.5.1.1 Fermentation using Yoghurt culture One (1) litre of fresh cow milk was inoculated with 10 ml of yoghurt culture (obtained from Animal Research Institute, CSIR-Accra) containing Lactobacillus thermophilus and Lactobacillus thremoduric. 3.5.1.2 Fermentation using freeze dried cheese culture. Freeze dried cheese culture obtained from Chr Hensen lab 10-12 Boege Alle DK- 2970 Hoersholm-Denmark was activated by two successive transfers in 9 ml of MRS broth for 24 h and 16 h respectively, incubated at 30oC. The mixture was transferred into sterile centrifuge tubes and centrifuged at 4500 g per minute for 15 minutes. The cultures were washed twice with sterile distilled water. It was then diluted with 9mls of sterile distilled water and homogenised with a votex mixer. Ten or 5 mls of the culture was inoculated into 2 L of milk and allowed to ferment in reciprocal shelves at 42 oC. University of Ghana http://ugspace.ug.edu.gh 61 3.5.1.3 Enumeration of LAB in yoghurt and cheese cultures Enumeration of Lactic acid bacteria was done by the pour plate method using deMan, Rogosa and Sharpe (MRS, Oxoid CM361) agar with pH 6.2. 0.1% cycloheximide supplement was added to suppress yeast growth and Cystein HCL to achieve anaerobic conditions during incubation without having to use Anaerocult A. The plates were incubated anaerobically in an anaerobic jar at 30°C for 120 h. 3.5.2 Fermentation of milk with cheese culture and yoghurt culture The rate of acidification of the yoghurt and cheese cultures were determined by 24 h fermentation of pasteurised raw cow milk with both cultures in a water bath (M 25 LAUDA) set at 420C. Ten (10) ml of the yoghurt culture and 10 ml and 5 ml of the cheese culture was inoculated into 2 L of pasteurized raw cow milk. The initial pH of the milk was recorded and the change in pH of the milk after every hour was recorded for 24 h. 3.5.3 Selection of appropriate pH after 24 h of fermentation After the 24h fermentation, 2 h, 4 h and 6 h were selected for milk fermentation with 10 ml cheese culture. The durations were selected based on their corresponding pH values attained after the 24 h fermentation. A preliminary work was done where wagashie was prepared with milk fermented for the selected periods. An informal sensory evaluation was done by an untrained seven member panel where the taste, texture and colour of the samples were assessed. Wagashie prepared with milk fermented for 4 h was accepted to be the maximum period of fermentation and the minimum was zero h fermentation which was used in the box Behnken design to optimise the wagashie process. The samples assessed were fresh, fried and smoked. University of Ghana http://ugspace.ug.edu.gh 62 3.6.0 Design of Experiment for ‘wagashie’ preparation A three variable Box Behnken Design for response surface methodology was used to study the combined effect of the coagulants (Extract and Rennet), salt (NaCl) and Fermentation time on the responses; texture, colour, taste and overall acceptability of ‘wagashie’ over three levels. The range and levels of the variables optimised are shown in Table 3.1 and Table 3.2 for plant extract used as coagulant and rennet used as coagulant for ‘wagashie’ preparation respectively. The Box-Behnken design is suitable for the exploration of quadratic response surfaces and generates a second degree polynomial model which is then used to optimise a process using a small number of experimental runs. The design requires an experimental number of runs according to: N=K2+K+CP Where k is the factor number which is 3, Cp is the number of replications at the centre point which is 3. The design which was developed using Minitab14 resulted in 15 experimental runs shown in Table 2. The 15 experimental runs were randomized to maximise the effects of the independent variables based on preliminary experiments. In determining the design matrix of the Box Behnken design, the level of one of the factors was fixed at the centre point while combinations of all levels of the other factors were applied (Meyers and Montgomery, 2002). Tables 3.1 and 3.2 show the coded and actual values for the level of the process variables for the plant extract and rennet coagulated wagashie’. University of Ghana http://ugspace.ug.edu.gh 63 Table 3.1: Coded and actual levels of the factors for three levels Box Behnken design for ‘wagashie’ using plant extract as coagulant. Independent Variables Symbols coded and actual values -1 0 +1 Fermentation Time(hr) X1 0 2 4 NaCl (g) X2 14 19 24 Extract weight (g) X3 100 150 200 Table 3.2: Coded and actual levels of the factors for three levels Box Behnken design for ‘wagashie’ using Rennet as coagulant Independent Variables Symbols coded and actual values -1 0 +1 Fermentation Time(hr) X1 0 1 3 NaCl (g) X2 7 10.25 14 Rennet conc. (ml) X3 0.7 5.35 10 University of Ghana http://ugspace.ug.edu.gh 64 3.6.1 Optimisation of the ‘wagashie’ process using Box-Behnken Design (plant extract preparation) The Box Behnken design which is a type of the response surface methodology was used to generate 15 combinations for the process variables fermentation time h, salt (g) and coagulant (plant extract (g) and rennet (ml)). Tables 3.3 and 3.6 shows the 15 combinations for both plant extract and rennet preparations combined by the Box Behnken De sign Table 3.3: The Box Behnken Design matrix of variables (k=3) for the optimisation of ‘wagashie’ coagulated with plant extract Run number X1 X2 X3 Fermentation time(hr) Extract(wt/g) Salt (NaCl)(g) 1 -1 1 0 0 200 19 2 0 -1 1 2 100 24 3 -1 -1 0 0 100 19 4 0 -1 -1 2 100 14 5 -1 0 1 0 150 24 6 1 0 -1 4 150 14 7 0 0 0 2 150 19 8 0 1 -1 2 200 14 9 0 1 1 2 200 24 10 -1 0 -1 0 150 14 11 0 0 0 2 150 19 12 1 0 1 4 150 24 13 1 -1 0 4 100 19 14 0 0 0 2 150 19 15 1 1 0 4 200 19 Table 4.3 shows the 15 combinations of the process variables, fermentation time (h), salt (g) and extract weight 9(g) generated by the Box Behnken design which provided the optimisation quality attributes (taste, texture, colour and overall acceptability) for the sensory evaluation of ‘wagashie’. Table 3.3 shows the coded and the actual values of the process University of Ghana http://ugspace.ug.edu.gh 65 variables that were used for the optimisation process. X1 represents the minimum value or level, X2 represents the middle value or level and X3 represents the maximum value or levels of the process variables which were fermentation time (h), extract weight (g) and weight of salt (g) for the product. The main effect of the sensory characteristics of ‘wagashie’ as a function of fermentation time, extract weight and weight of salt which include colour, taste, texture and overall acceptability is discussed in the results. 3.6.2 Optimisation of the ‘wagashie’ process using Box Behnken Design (commercial rennet preparation) Table 3.4: The Box Behnken Design matrix of variables (k=3) for the optimisation of ‘wagashie’ coagulated with Rennet (ml) Table 4.6 shows the 15 combinations of the process variables, fermentation time (h), salt (g) and rennet (ml) generated by the Box Behnken design which provided the optimised quality attributes for sensory evaluation of ‘wagashie’. Table 3.4 shows the coded and the actual values of the process variables that were used for the optimisation process. X1 represents the Run order X1 X2 X3 Fermentation time(hr) Salt (g) Rennet (ml) 1 1 1 0 3 14.00 5.35 2 -1 1 0 1 14.00 5.35 3 1 -1 0 3 7.00 5.35 4 0 -1 -1 2 7.00 0.70 5 0 0 0 2 10.50 5.35 6 -1 -1 0 1 7.00 5.35 7 0 1 -1 2 14.00 0.70 8 0 0 0 2 10.50 5.35 9 1 0 -1 3 10.50 0.70 10 0 -1 1 2 7.00 10.00 11 0 1 1 2 14.00 10.00 12 0 0 0 2 10.50 5.35 13 -1 0 1 1 10.50 10.00 14 1 0 1 3 10.50 10.00 15 -1 0 -1 1 10.50 0.70 University of Ghana http://ugspace.ug.edu.gh 66 minimum value or level, X2 represents the middle value or level and X3 represents the maximum value or levels of the process variables which were fermentation time (h), rennet concentration (ml) and weight of salt (g). The main effect of the sensory characteristics of ‘wagashie’ as a function of fermentation time, rennet concentration and salt which include colour, taste, texture and overall acceptability is discussed in the results. 3.7.0 Sensory Evaluation Hedonic sensory evaluation and the quantitative descriptive sensory evaluation was used to evaluate the ‘wagashie’ samples. A 9- point hedonic scale was used to rate the acceptability and an unstructured 10cm line scale was used to measure the intensity of the sensory attributes; taste, texture, colour and aroma of the ‘wagashie’ samples. 3.7.1 Hedonic Sensory Evaluation Two hedonic sensory evaluations were carried out. The first hedonic sensory was used to discriminate among the fifteen wagashie samples generated by the box behnken design. The second hedonic sensory was carried out for confirmatory affective testing (consumer preference) of the optimised ‘wagashie’ samples. Fifteen voluntary panellists were selected for the first sensory and twenty voluntary panellists were selected for the second sensory.The selection was done based on their familiarity with cheese or wagashie and were regular consumers of the product. The panellists were to rate the acceptability of the ‘wagashie’ samples. Ten (10) grams of ‘wagashie’ was served to each panellist. The samples served were coded with three digit random numbers and evaluated at room temperature with uniform lighting conditions in a well structured sensory evaluation laboratory at the CSIR-Food Research Institute. Each panel was seated in individual cubicle, water and tissues were served University of Ghana http://ugspace.ug.edu.gh 67 for mouth rinsing and hand cleaning after every evaluation. Evaluation of the wagashie samples was done based on the sensory attributes; colour, taste, texture and overall acceptability using a 9-point hedonic scale of 1 to 9 ;(1 = ‘dislike extremely’, 2 = ‘dislike very much’, 3 = ‘dislike moderately’, 4 = ‘dislike slightly’, 5 = ‘neither like nor dislike’, 6 = ‘like slightly’, 7 = ‘like moderately’, 8 = ‘like very much’, 9 = ‘like extremely’). Fifteen fresh samples were assessed for the first hedonic sensory whiles seven samples (fresh, fried and smoked) were assessed during the affective sensory. 3.7.2 Quantitative Descriptive Analysis 3.7.2.1 Selection of Panellists After optimising the product, a twelve member panel from CSIR-Food Reaserch Institute were selected and trained for a Quantitative Descriptive Sensory Evaluation. 3.7.2.2 Training of Panellists The training for the Quantitative descriptive sensory evalution was done for three days, two hours per day. A breif introduction about the origin, process, consumption of ‘wagashie’ and the purpose for the gathering was given. The panellists developed words (descriptors) which described the sensory attributes; aroma, texure, taste and colour of laboratory- prepared ‘wagashie’ (fermented and non-fermented) and ‘wagashie’ sold on the market (control).The laboratory- prepared ‘wagashie’ samples were fresh, fried and smoked whiles the market ‘wagashie’ samples were fresh and fried. University of Ghana http://ugspace.ug.edu.gh 68 3.7.2.3 Generation of Descriptors for ‘wagashie’ An agreed consensus was reached for the descriptors and were defined by the panellists.The selection of the descriptors were confirmed using reference samples whose sensory attributes (aroma or taste or texture or colour) are closely related or the same as the descriptors developed for ‘wagashie’. The panellists were trained on how to quantify the descriptors on an unstructured 10cm line scale. 3.7.2.4 Assessment of wagashie using QDA The XLSTAT V. 14 was used to develop a Balanced Block Design where the pattern for seving the samples were completely randomised to eliminate any form of bias during the evaluation. The sensory evaluation was done in a well structured sensory evaluation laboratory with individual booths or cubicles and uniform lighting conditions. Eleven of samples served were fresh, fried and smoked samples were served. The fried ‘wagashie’ samples were fried for 2 minutes in sunflower oil at 110 0C, the smoked samples were smoked on a coalpot for 20 minutes under low heat. 10g of each sample was served to each panellist for assessment. The panellists were served with water for mouth rinsing after assessing each sample and tissues for cleaning of the hand to minimise all forms of error. The evaluation was done for three days; six samples were assessed on the first day,two samples on the second day and five samples on the third day. Panelists quantified the attributes ; Taste, Texture, Colour and Aroma on an ustructured 10cm line scale. 3.8.0 Physicochemical Analyses The following physical and chemical properties were determined in the ‘wagashie’ samples; University of Ghana http://ugspace.ug.edu.gh 69 3.8.1 Determination of pH Ten (10) g of each of the ‘wagashie’ sample was weighed into a stomacher bag, 100 mls of distilled water was added. The content was homogenised in a stomacher for 20s and the pH reading was determined by dipping the probe of a calibrated Mettle Toledo pH meter into the sample. 3.8.2 Determination of Total Titratable acidity To 10 g of test portion, water was added to a volume of 105 ml, the mixture was agitated and filtered. 25 ml of the aliquot filtrate which represented 2.5 g of the test portion was measured into a conical flask and titrated against 0.1M NaOH using phenolphthalein as indicator. The results were expressed as lactic acid. 1 ml 0.1 M NaOH = 0.0090 g lactic acid or ml 0.1 M NaOH/100 g. (AOAC, 2006). % acid (wt/wt) = NxV1x Eq wt V2 x 10 Where, N= Normality of titrant, NaOH (meq/ml) = 0.1 V1= Titre (ml) Eq wt = Equivalent weight of Lactic acid (mg/meq) = 90.08 V2 = Volume of Sample (ml) = 25ml University of Ghana http://ugspace.ug.edu.gh 70 3.8.3 Determination of protein Kjeldahl method was used to determine the crude protein content of the samples. About 0.2 g of the sample was weighed and ground on a filter paper. The filter paper was folded and dropped into a digestion tube and concentrated sulphuric acid (15 ml) and a catalyst (Kjeltab) tablet were added. The digester was set at 400 oC to digest the sample. When there was a colour change of the sample from black to green; the digestion process was put to a stop. The sample was then distilled with 80 ml of water and 80 ml of 40% sodium hydroxide in a distillation unit. During distillation the steam generated by a heating apparatus distilled the ammonia and water into a receiving flask containing 25 ml of boric acid. The boric acid and NH3 + H2O formed was titrated with standardised 0.1M HCl which resulted in a change from colourless solution to a pink colour formation at the end point. A blank was run under the same condition as with the sample. Total nitrogen content was then calculated according to the formula: (Titre (of sample) – blank) x concentration of standardised HCl x 14.007 10 x weight of sample The total nitrogen was converted to crude protein by multiplying it with a factor 6.25. 3.8.4 Determination of Fat About 2g of the macerated sample was dried in an oven as described for moisture determination below. The dried sample was then placed in an extraction thimble and stopped with grease-free cotton. Prior to extraction, the round bottom flask was dried, cooled and weighed. The thimble was placed in the extraction chamber and 240 ml of petroleum ether was added to extract the fat. The extraction was done for 15 h at a condensation rate of 5 - 6 University of Ghana http://ugspace.ug.edu.gh 71 drops per second. The fat extracted was then dried in an oven at 103 ± 2 oC for 1 h. The dried fat was then cooled and weighed. A blank was run with the same procedure but without the sample. 3.8.5 Determination of FFA About three drops of phenolphthalein was added to 40 ml of ethanol and neutralised with 0.1 N Sodium hydroxide. About 5 g of the well-mixed sample was added; the mixture was boiled on a hot plate and titrated with 0.1N NaOH. The titre was recorded and the free fatty acid (FFA) was determined according to the calculation below: FFA = Titre x Normality of NaOH (0.1) x Factor of dominant FFA (Oleic) 10 x weight of sample The FFA determined was expressed as oleic acid. 3.8.6 Moisture determination Five grams of well-mixed portion of the sample was weighed on an analytical balance with an error of 0.0001g. This was preceded by heating the metal can made of alumina for 20 minutes at 103 ± 2 oC and cooling in a desiccator to constant weight. The can with the sample was then dried in an oven (Genlab Ltd, England) at 103 ± 2 oC for 4 hours. The drying started at the time the oven attained a temperature of 103± 2 oC. The % moisture was determined at room temperature according to the calculation below. Moisture (%w/w)=[(wt of dish + fresh sample) – (wt of dish + dried sample)]x 100% [(wt of dish + fresh sample) – (wt of dish)] University of Ghana http://ugspace.ug.edu.gh 72 3.8.7 Determination of Ash content of the samples Ash is the inorganic residue obtained by burning off the organic matter of feedstuff at 400- 600 OC in muffle furnace for 4hours. 2g of the sample was weighed into a pre-heated crucible. The crucible was placed into muffle furnace at 400-600 OC for 4 hours or until whitish-grey ash was obtained. The crucible containing the ash was then placed in the desiccator prior to weighing. %Ash = wt. of crucible+ash – wt. of crucible Wt. of sample 3.8.8 Colour measurement of ‘wagashie’ The colour of ‘wagashie’ was determined by using the L a b colour notation system. The equipment used was the Minolta Chroma Meter (Model Cr- 200 Minolta Camera, Japan). The ‘wagashie’ samples were sliced and arranged in a Petri dish and covered. The colour measurement for the processed sample (smoked) was done on the inside and outside of the ‘wagashie’ samples. The colour was measured by placing the equipment on the surface of the petri dish containing the sample. The readings were taken randomly from three spots and the mean reading was calculated. Before measuring, the chroma meter was calibrated with a white tile and checked for recalibration in between measurements, although no modifications were required. Colour values were recorded as L* = darkness/lightness (0= black, 100 = white), a* (–a* = greenness and +a* = redness), and b*(–b* = blueness, +b* = yellowness). 3.8.9 Texture Profile Analysis (TPA) of ‘wagashie’ Texture properties of the ‘wagashie’ samples were determined by a Texture Analyzer TA- XT2 (Stable Micro Systems Ltd., Surrey, UK). The samples were shaped uniformly with a University of Ghana http://ugspace.ug.edu.gh 73 cylindrical cork borer (10 mm in diameter). Fifteen measurements were taken on each ‘wagashie’ sample. TPA parameters measured were hardness, adhesiveness which is the degree of stickiness by mouth feel when chewed five times, springiness (the force with which the sample returns to its original shape or size after partial compression it is also known as elasticity of cheese), gumminess (which is the density during chewing time required to break up a semi solid food until it is suitable for chewing) and chewability (which is the number of chews needed to masticate the sample to a consistency suitable for swallowing). These parameters were measured by the software with a P175 75m compression platern and a cylindrical probe with pre-test speed of 2 mm/sec, test speed 1.5 mm/sec, post-test speed 5 mm/sec, distance 10mm, time 3sec and a contact force of 5.0g. The samples were uniformly shaped with a weight of 1g, width of 10mm and a height of 10mm. Samples were penetrated to 75% of their original height with a constant speed penetration of 5 mm/ sec. 3.9.0 Shelf life study of wagashie 3.9.1 Irradiation and Packaging of wagashie In order to select an appropriate dose for the irradiation of ‘wagashie’, a preliminary study was carried out whereby the improved ‘wagashie’ samples were subjected to four different doses; 1kGy, 2kGy, 3kGy and 4kGy from Cobalt-60 source. These doses were selected based on the purpose of the radiation process which was to extend the shelf life of ‘wagashie’ by decontaminating the product of all microorganisms that may shorten the shelf life of the product. The dose absorbed by the samples per hour was 1.61kGy thus about 2 h and 30 min was used to deliver all four doses. For purpose of preliminary trial, unpasteurised cow milk was used to prepare the ‘wagashie’ samples. Microbiological analysis was carried out after the preparation of the samples. The samples were vacuum packaged with a vacuum University of Ghana http://ugspace.ug.edu.gh 74 packaging machine in high density polyethylene bags obtained from Kasoa in the Central region of Ghana. The gauge of the polyethylene was 150mm by 30mm in length and width respectively before irradiation. After irradiation, the samples were stored under ambient conditions (28 ͦ C for 14 days) and analysed by an informal sensory evaluation. The colour by appearance, the texture by hand feel and the aroma were assessed. After the evaluation, all the samples except the sample irradiated with a dose of 4kGy had developed green and yellow discolouration, off flavour and the very soft texture with whey exudates in the packaging material. The 4kGy was therefore chosen for the continuation of the study. The actual samples for the shelf life study were prepared from pasteurized raw cow milk. The samples were divided into two sets, the irradiated and non irradiated samples, the non irradiated samples served as a control for the irradiated samples 3.9.2 Storage of ‘wagashie’ The vacuum and normal packaged irradiated and non irradiated samples were stored under ambient conditions. Microbiological tests were carried out daily for the fresh samples (shorter shelf life) and weekly for the roasted samples until spoilage was observed in the samples. The pH of the samples were measured and recorded. 3.10 Statistical Analysis Analysis of variance (one way ANOVA) was used to determine the significant differences among sensory parameters, physicochemical properties, the safety assessment of ‘wagashie’,shelf life study and the rheological analysis of ‘wagashie’. Minitab V 17 was used to generate model equations for each sensory parameter which was used to plot three dimensional response surface plots. University of Ghana http://ugspace.ug.edu.gh 75 Microsoft Excel 2007 and XLStat 2014 was used generate the spider plot and the principal composite analysis respectively for the Quantitative Descriptive Analysis. Fresh wagashie Smoked wagashie Soddom apple plant Fried wagashie University of Ghana http://ugspace.ug.edu.gh 76 CHAPTER FOUR 4.0 RESULTS 4.1.0 Market Survey 4.1.1 Microrganisms present in market ‘wagashie’ samples Table 4.1 Shows the mean population of microorganisms detected in market wagashie samples. Significant differences occurred between the count of microorganisms detected in the fried and fresh ‘wagashie’ samples at p<0.05. The highest count of microorganisms was 1.38x109 CFU/g (aerobic mesophiles) which was detected in the fresh sample obtained from Nima. Bacillus cereus was not detected in the fried samples but was detected in low counts in the fresh samples. Yeast and moulds were detected in all samples except the fresh sample obtained from Ashaiman. There was no detection or counts for Salmonella and Staphylococcus aureus respectively. Table 4.1 also shows the effect of deep frying and immediate packaging on the safety of the market wagashie samples. Superscript letters ‘a’ and ‘b’ show the significant differences between the fresh and the fried wagashie samples at p<0.05. Aerobic mesophiles were detected in high count in the fresh sample. Moulds and Enterococcus were detected in low counts whiles Salmonella, Bacillus cereus, Enterobacteriaceae, Stahpyloccocus, yeast and mould, E. coli and coliforms were not enumerated or detected in the fried samples . The fresh samples however contained almost all the pathogens except Salmonella and Staphylococcus. The highest count of microorganisms was 2.4x1010 (aerobic plate count) and was recorded in the fresh sample and the lowest count of microorganism recorded was 1.0x101 (yeast) which was recorded in the fried sample. University of Ghana http://ugspace.ug.edu.gh 77 Table 4.1: Mean microbial count in fresh and fried market ‘wagashie’ samples in g/CFU. Means in the same column with the same letters are not significantly different (p<0.05) Table 4.2: Effect of deep frying and aseptic packaging on the microbial count of fresh ‘wagashie’ obtained from Nima in g/CFU. Means in the same column with the same letters are not significantly different (p<0.05) Sample Aerobic Plate count Yeast Moulds Coliform Bacteria E. coli Entero bacteriaceae Entero coccus B. cereus Salmonell a spp. Staphy lococc us aureus Fresh Nima 1.4x109b 1.1x107a 1x15a 5.6x105b 3.6x104b 3.9x106b 8.1x104a 7.0 x102b Not detected Not detecte d Fresh Ashiama n 3.0x109a 4.7x106b Not detected 2.3x107a 4.1x104a 5.7x106a 1.0x104b 1.1x103a Not detected Not detecte d Fried Nima 8.3x104c 1.3x102c 5.0x10b 2.6x103d 1.0 x10c 1.9x106c 6.0x102c Not detected Not detected Not detecte d Fried Ashiama n 2.9x106c 2.0x10c 1.0x10c 3.7x105c Not detected 5.8x105d 3.3x102c Not detected Not detected Not detecte d Sample Aerobic Plate count Yeast Mould Coliform Bacteria E. coli Entero bacteriaceae Entero coccus B. cereus Salmone lla spp. Staphylo coccus aureus Fresh 2.4x1010b 1.68x107 2.0x105b 1.8x107 1.4x106 2.1x108 7.4x106b 9.0x102 0 0 Fried 5.6x107a 0.0 1.0x10a 0.0 0.0 0 2.0x10a 0 0 0 University of Ghana http://ugspace.ug.edu.gh 78 4.1.2 pH of the market wagashie samples. From fig.4.1, Significant differences occurred in the pH values of the market samples. The highest pH recorded in the samples was 5.56 which was for a fried sample and the lowest pH was 5.09 which was a fresh sample. the fresh samples had low pH values with high count of enteric pathogens and the fried samples recorded high pH values with low counts of enteric pathogens. Figure 4.1: Mean pH of ‘wagashie’ sampled from the market 4.2.0 Fermentation trials for laboratory preparation of ‘wagashie’; Rate of fermentation of raw cow milk with cheese and yoghurt starter cultures 4.2.1 Fermentation with different concentrations of cheese culture To choose the best concentration of the starter culture for the fermentation of milk, this experiment was carried out. Figure 4.2 shows the rate of fermentation of 1 L of pasteurized raw cow milk with two different concentrations of cheese culture 5 ml and 10 ml. From the graph there was not much difference in the rate of fermentation between the 4.8 4.9 5 5.1 5.2 5.3 5.4 5.5 5.6 5.7 Fresh Ashiaman Fresh Nima Fried Ashiaman Fried Nima Aseptic Fried wagashie Fresh wagashie p H Sample University of Ghana http://ugspace.ug.edu.gh 79 two cultures at the end of the 24 h fermentation period, but the rate of pH change for the cheese culture with a concentration of 10 ml was faster after two hours of fermentation than the cheese culture with a concentration of 5 ml. The cheese culture with concentration 10 ml, reduced from a pH of 6.12 to a pH of 3.92 at the end of the 24 h fermentation whiles the pH of the cheese culture with concentration 5 ml, reduced from 6.12 to 4.14 at the end of the fermentation period. Since the rate of fermentation of the cheese culture with concentration of 10 ml was faster than the 5 ml concentration, the 10 ml was selected for the fermentation of raw cow milk for the study. Figure 4.2: The rate of pH change after 24 hours fermentation in a water bath set at 45oC with 5mls and 10mls of cheese culture 0.00 1.00 2.00 3.00 4.00 5.00 6.00 7.00 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 p H o f R aw c o w m ilk p e r 1 h r fe rm e n ta ti o n Fermentation time (hr) pH 1 pH 2 5 mls of cheese culture 10 mls of cheese culture University of Ghana http://ugspace.ug.edu.gh 80 4.2.2 Fermentation with 10 ml of both cheese and yoghurt cultures Figure 4.3 shows the rate of acidification for the cheese and yoghurt cultures at a concentration of 10 ml in 1 L of pasteurized raw cow milk for both cultures after 24 h fermentation. From figure 4.3, pH 1 represents 10 ml of yoghurt culture whiles pH 2 represent 10 ml of cheese culture. With an initial pH of 6.04 ,the pH of the yoghurt culture reduced after 1 h of fermentation whiles the pH of the cheese culture started reducing after 2 h of fermentation. However, there was not much difference in the pH of the milk at the end of the 24 h fermentation for both cultures. The fermentation was carried out under the same temperature condition in a water bath set at 42 oC. This experiment was carried out to know how effective the yoghurt culture ferment milk as compared with the cheese culture so that in case of unavailability of the cheese culture,the yoghurt culture could be substituted. Figure 4.3: The rate of pH change after 24 h fermentation of 2 L of pastuerized fresh cow milk in a water bath set at 45oC with 10 ml of yoghurt culture and 10 ml of cheese culture. 0.00 1.00 2.00 3.00 4.00 5.00 6.00 7.00 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 p H Fermentation time (h) pH 1 pH 2 10 mls of cheese culture 10mls of yoghurt culture University of Ghana http://ugspace.ug.edu.gh 81 4.3.0 Optimisation of the ‘wagashie’ process 4.3.1.0 Using the response surface methodology to optimise the ‘wagashie’ process prepared with plant extract as coagulant. The response surface methodology was used to determine the effects of salt (g), fermentation time (h) and plant extract (g) on the sensory attributes: colour, taste, texture and overall acceptability of ‘wagashie’. The Box Behnken design was used to study the effects of variation in the levels of milk fermentation time (0h-4h), quantity of salt (14g – 24g) and extract weight (100g – 200g) for ‘wagashie’ production on the various responses. The relationships between the various responses are represented by a three dimensional response surface. The percentage variability (R2) obtained were high (close to 100%) for the various responses which indicates a good model which can best be used to describe the effects of the variations on the responses. Results imputed were obtained using a nine point hedonic sensory evaluation. The discussions are based on the trends of the visualised effects and the significant terms in the fitted model. 4.3.1 Texture The analysis of the coefficients for the regression models showed that the independent variables had significant effects except plant extract (g) which had marginal significance on the texture of ‘wagashie’. The fitted regression model (R2) for texture was 99.8% which denotes a good model and can be used to best describe the texture of ‘wagashie’. Figure 4.4 shows the response plots for texture of wagashie showing the effect of fermentation time (h) and salt (g) at a constant extract weight of 150g (7.5%). The response plots shows a significant linear effect on fermentation time and salt, a significant quadratic effect on fermentation time (h) , salt (g) and University of Ghana http://ugspace.ug.edu.gh 82 extract (g) and a significant interaction effect of extract (g) and salt (g) on taste.The linear effects showed that texture decreased or increased continiously in the experimental field, the quadratic effect showed the possibility of an optimal region for salt weight (14g - 24g) and fermentation time (0h-4h) and the interraction showed that the texture of wagashie is affected by fermentation time, exract (g) and salt (g). From figure 4.4, the score for texture increased with decreasing fermentation time (h) and increasing quantity of salt (g) but decreased with increasing fermentation time and increase in the quantity of salt (g). Figure 4.4: Response Surface plot representing the effect of Fermentation time and Salt on the score for Texture when the weight of extract is 150 g. Fitted regression model equation for texture = 8.93544- 0.785552x1+0.01397x2- 0.23233x3 + 0.00239x1x2 + 0.00239x2x3 + 0.017019x12 + 0.00005x22 +0.01653x32 T exture 6.0 6.5 0.0 Fermentation time (hr) 1.5 3.0 7.0 7.5 4.5 22.5 20.0 17.5 15.0 Salt(g) Hold Values Extract(wt/g) 150 University of Ghana http://ugspace.ug.edu.gh 83 4.3.2 Colour Colour gives the first impression of food quality thus it is one of the most important attributes affecting the consumer acceptance of food. The analysis of the coefficient estimates for the regression model showed that the independent variables had no significant effect on the colour of ‘wagashie’. The R2 was 91.7%. There was no significance (>0.05) in the linear and quadratic effects and the interaction of the independent variables on the score for colour. Figure 4.5 shows the response surface plot of the effect of fermentation time (h) and salt (g) on the score for colour when the extract (g) is constant at 150g (7.5%), the response surface showed linear, interactive and quadratic effects of the factors. The linear effect showed that colour decreased or increased continuously in the experimental field. Colour score increased with increase in the quantity of salt (g) and decrease in fermentation time (h).It decreased with increasing fermentation time (h) and increase in the quantity of salt (g). The quadratic effect showed the existence of an optimal level for colour in the experimental field. University of Ghana http://ugspace.ug.edu.gh 84 Figure 4.5: Response Surface plot representing the effect of fermentation time and Salt on the score for Colour when the Extract is 150 g. The fitted regression model for colour = 9.81519- 0.06669x1+0.00519x2-0.17858x3 - 0.00479x1x2 – 0.00100 x2x3 - 0.000099x12 - 0.00007x22 +0.00864x32 4.3.3 Taste Taste is one of the most important attributes of ‘wagashie’ aside texture and it mostly determines the acceptability of the product. Figure 4.6 shows the response plot for taste which showed a linear, quadratic and interaction effects of fermentation time (h) and quantity of salt (g) at a constant extract weight of 150 g representing 7.5%. From the surface plots, the score for taste increased with increase in the quantity of salt (g) and decrease in fermentation time (h). It decreased with increasing fermentation time and decreasing quantity of salt (g). The analysis of the coefficient estimates for the regression model showed that there was a significant linear effect of Colour 6.8 7.0 7.2 0.0 1.5 3.0 Fermentation time (hr) 7.4 4.5 22.5 20.0 17.5 15.0 Salt(g) Hold Values Extract(wt/g) 150 University of Ghana http://ugspace.ug.edu.gh 85 the quantity of salt (g), a marginal significant effect of fermentation time (h) and a non significant linear effect of plant extract (g) on the score for taste.There was a significant linear effect of salt (g), a significant quadratic effect of fermentation time and salt (g) and a significant interaction effect of salt (g) and extract (g) on taste. The fitted regression model had an R2 of 96.7% which depicts a good model and can best describe the taste of ‘wagashie’ prepared with extract as coagulant. Figure 4.6: Response Surface plot representing the effect of Salt and Fermentation time on the score for taste when extract is 150 g. The fitted regression model for colour = 14.1206- 0.4577x1 + 0.0103x2 - 0.8404x3 - 0.0022x1x2 – 0.0011 x2x3 - 0.0769x12 - 0.0000x22 +0.270x32 4.3.4 Overall Acceptability The analysis of the coefficient estimates for the regression model showed that the independent variables had no linear significant effect on the score for overall T aste 6.5 7.0 0.0 Fermentation time (hr) 1.5 3.0 7.5 4.5 22.5 20.0 17.5 15.0 Salt(g) Hold Values Extract(wt/g) 150 University of Ghana http://ugspace.ug.edu.gh 86 acceptability of ‘wagashie’ but showed the quadratic effect of fermentation time (h) as the significant term. From the response plot (fig 4.7),the score for overall acceptability increased with increasing weight of salt (g) and decreasing fermentation time (h) and decreased with increasing fermentation time (h) and increasing quantity of salt (g). In fig 4.7, the linear effect showed that overall acceptability increased or decreased continuously in the experimental field. The quadratic effect showed that there is an optimal level within the fermentation time (h) range and the interactive effect showed that overall acceptability was not affected by the combined effect of the variables. Figure 4.7: Response surface plots representing the effect of salt and fermentation time on the score for Overall acceptability when plant extract weight is 150 g. The fitted regression model for Overall acceptability = 11.0888- 0.530x1 - 0.0065x2 - 0.3693x3 - 0.0099x1x2 – 0.0015 x2x3 - 0.1379x12 - 0.0001x22 +0.0166x32 Overall acceptability 6.5 7.0 0.0 Fermentation time (hr) 1.5 3.0 7.5 4.5 22.5 20.0 17.5 15.0 Salt(g) Hold Values Extract(wt/g) 150 University of Ghana http://ugspace.ug.edu.gh 87 4.3.5 Optimisation process To optimise the process variables (salt, fermentation time and extract) for ‘wagashie’ production with respect to the responses, the contour plots were overlaid. Fig 4.8 shows an overlay of contour plots for the responses, texture, taste, colour and overall acceptability of ‘wagashie’. From fig 4.8, the optimum levels for the preparation of ‘wagashie’ with maximum texture, taste, colour and overall acceptability were 0 h fermentation time, 23g to 24g of salt and 150g of the plant extract. Corresponding to the optimised values for the process variables the predicted values for taste, texture, colour and overall acceptability were 7.38, 7.18, 7.47 and 7.36 respectively at 23g of salt and 0 h fermentation time. For the purpose of the study the optimum conditions were chosen and prepared which was then compared with ‘wagashie’ prepared with milk fermented for 4h with extract weight of 150g and 23g of salt. Using a 9- point hedonic scale, a sensory evaluation was carried out to confirm the predicted values obtained after the optimisation process. The products assessed were in the fresh, fried and smoked for Figure 4.8: Contour plot for texture, taste, colour and overall acceptability of ‘wagashie’ overlaid on one axis of fermentation time (h) and salt (g) Fermentation time (hr) S a l t ( g ) 43210 24 22 20 18 16 14 Hold Values Extract(wt/g) 150 University of Ghana http://ugspace.ug.edu.gh 88 Figure 4.9: Mean pH values for wagashie samples coagulated with plant extract combined by Box Behnken Design Figure 4.9 shows the mean pH values of ‘wagashie’ recorded after preparation. Overlapping error bars indicates a non significant difference at p<0.05 whiles error bars that do not overlap represents a significant difference between the samples. From figure 4.6 the highest pH value was 7.02 and the lowest pH value recorded was 6.1. These samples were non-fermented and fermented samples respectively. 5.4 5.6 5.8 6 6.2 6.4 6.6 6.8 7 7.2 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 p H Run Number University of Ghana http://ugspace.ug.edu.gh 89 Figure 4.10: The TTA of wagashie samples combined by Box Behnken Design using plant extract as cogulant Figure 4.10 shows the Titratable acidity of the ‘wagashie’ samples after preparation. Overlapping error bars indicate a non significant difference at p<0.05 whiles error bars that do not overlap represents a significant difference between the samples. The TTA recorded showed the amount of lactic acid present in the ‘wagashie’ as a result of the lactic acid fermentation of the milk. The TTA values recorded were low ranging from 0.001 to 0.89. 4.4.0 Using the Response Surface methodology to optimise the wagashie process using commercial rennet as coagulant. The response surface methodology was used to determine the effects of salt (g), fermentation time (h) and rennet (ml) on the sensory attributes, colour, taste, texture and overall acceptability of ‘wagashie’. The Box Behnken design was used to study 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 TT A Run Number University of Ghana http://ugspace.ug.edu.gh 90 the effects of variation in the levels of milk fermentation time (0h-4h), weight of salt (7g – 14g) and rennet concentration (0.7ml – 10ml) for ‘wagashie’ production on the various responses. The relationships between the various responses are represented by a three dimensional response surface and an overlaid contour plots. The percentage variability (R2) obtained for the various responses; texture, colour, taste and overall acceptability were 97.06%, 91.49%, 89.53% and 74.67% which indicates a fairly good model may be used to describe the effects of the variations on the responses. Results imputed were obtained from a sensory evaluation using fifteen untrained panellists. The discussions are based on the trends of the visualised effects and the significant terms in the fitted regression model 4.4.1 Texture The analysis of the coefficients for the regression model showed that the independent variable rennet (ml) had no significant effects on the score for texture.However the quantity of salt (g) had a significant effect and fermentation time (h) had a marginal significant effect on the score for the texture of ‘wagashie’. The fitted regression model (R2) for texture was 97.06% which denotes a good model and can be used to describe the texture of ‘wagashie’. Figure 4.11 shows the effect of fermentation time (h) and salt (g) at a constant rennet concentration of 5.35ml (0.27%) on the sensory score for texture of wagashie. From the results of the analysis of variance, there was a significant linear and quadratic effect of the variables on the score for texture and a non significant interaction effect of the variables on the texure score. There was a significant linear and quadratic effect of salt and a marginal significant interaction effect of fermentation time (h) and rennet (ml) on texture .From the response plot University of Ghana http://ugspace.ug.edu.gh 91 (figure 4.11),the score for texture increased with increasing fermentation time (h) and increasing quantity of salt (g) but decreased with increase fermentation time (h) and decrease in the quantity of salt. The linear effects showed that the sensory score for texture decreased or increased continously in the experimental field, the quadratic effect showed possibility of an optimal region for the quantity of salt (11g - 14g) and fermentation time (0h-4h) and the interraction showed that the score for the texture of wagashie was marginally affected by fermentation time (h) and rennet concentration (ml). Generally, the texture of cheese depends on the initial process during which milk acidify (fermentation), duration of ripening and the changing miosture content of the cheese. Figure 4.11: Response Surface plot representing the effect of salt and Fermentation time on the score for Texture when Rennet is 0.27%. The fitted regression model for texture = 5.00 - 0.258 x1 + 0.510 x2 - 0.1844 x3 + 0.0160 x12 - 0.02986 x22 + 0.01168 x32 + 0.0160 x1x2 + 0.02151 x1x3 T exture 5.5 6.0 0.0 1.5 3.0 Fermentation time(hr) 6.5 7.0 14 12 10 Salt (g) 8 4.5 Hold Values Rennet (ml) 5.35 University of Ghana http://ugspace.ug.edu.gh 92 4.4.2 Colour The analysis of the coefficient estimates for the regression model showed that there was a significant linear effect of the variables on taste, however there was no significant effect of fermentation time on the colour of ‘wagashie’ but there was a significant effect of salt and rennet concentration on the score for taste . The fitted regression model R2 was 91.49%.There was no significance in the quadratic and interaction effect of the independent variables on the score for the colour of ‘wagashie’. Figure 4.12 shows the response surface plots of the effect of fermentation time (h) and salt (g) on the score for colour when the rennet concentration is constant at 5.35 ml (0.27%). The response surface showed linear, interactive and quadratic effects on the colour of wagashie. The linear effect showed that colour increased or decreased continuously in the experimental field. From the response surface, colour increased with increasing quantity of salt (g) and fermentation time (h) and decreased with increasing fermentation time (h) and decreasing quantity of salt (g). The quadratic effect showed the existence of an optimal region for colour in the experimental field and the interactive effect showed that colour was not significantly affected by the levels of fermentation time and weight of salt. A significant lack of fit showed that a higher model is needed to explain the effect of the factors on the sensory score for the colour of ‘wagashie’. University of Ghana http://ugspace.ug.edu.gh 93 Figure 4.12: Response Surface plot representing the effect of salt and Fermentation time on the score for colour when Rennet is 0.27%. The fitted regression model for Colour = 7.079 - 0.184 x1+ 0.151 x2 - 0.0660 x3 - 0.0028 x12 - 0.01255 x22 + 0.00214 x32 + 0.01429 x1x2 + 0.00753 x2x3 4.4.3 Taste Fig. 4.13 shows the effects of fermentation time and salt on taste at a constant rennet concentration of 5.35ml representing 0.27%. From the surface plots, the score for taste increased with increase in the quantity of salt and increase in the fermentation time (h). It however, decreased with decrease in fermentation time (h) and increase in the quantity of salt (g). The analysis of the coefficient estimates for the regression model showed that there was no significant linear effect of fermentation time (h) and rennet concentration (ml) on taste but there was a significant linear effect of salt (g) on the score for taste. The quadratic effect and the interaction of the variables Colour 6.8 7.0 8 10 12 Salt (g) 7.2 7.4 14 3.0 1.5 Fermentation time(hr) 0.0 4.5 Hold Values Rennet (ml) 5.35 University of Ghana http://ugspace.ug.edu.gh 94 showed no significant effect on the taste. The fitted regression model, R2 was 89.93%. A significant lack of fit shows that a higher model is needed to better explain the effect of the factors on the taste of ‘wagashie’ prepared with rennet as coagulant. Cheese flavour is regarded as a component of taste and aroma and taste refers to the water soluble fraction which includes peptides, amino acids, organic acids, salts and amines. Figure 4.13: Response Surface plot representing the effect of salt and Fermentation time on the score for Taste when Rennet is 0.27%. The fitted regression model for Taste = -3.11 - 0.211x1 + 1.795x2 - 0.472x3+ 0.0449x12 - 0.0806x22 + 0.0109x32 + 0.0185x1 x2+ 0.0019x1 x3 + 0.0310x2 x3 T aste 6.5 7.0 0.0 Fermentation time(hr) 1.5 3.0 7.5 4.5 12 10 8 14 Salt (g) Hold Values Rennet (ml) 5.35 University of Ghana http://ugspace.ug.edu.gh 95 4.4.4 Overall acceptability The analysis of the coefficient estimates for the regression model showed that the independent variables had no significant effect on the score for overall acceptability of ‘wagashie’. The linear and quadratic effects and the interaction had no significant effect on the score for overall acceptability of ‘wagashie’. From the response plot (fig 4.14) overall acceptability increased or decreased continuously within the experimental field thus overall acceptability increased with increase in the quantity of salt (g) with an increase in the fermentation time (h) and decreased with an increase fermentation time (h) and a decrease in the quantity of salt (g). The fitted regression model R2 for overall acceptability was 74.67% and the adjusted R2 was 29.08%, which shows that the model cannot be used to explain the sensory score for the overall acceptability of ‘wagashie’and that a higher model is needed to best explain. University of Ghana http://ugspace.ug.edu.gh 96 Figure 4.14: Response Surface plot representing the effect of salt and Fermentation time on the score for overall acceptability when Rennet is 0.27%. The fitted regression model for Overall acceptability = 7.27 - 0.339 x1 + 0.093 x2 - 0.085 x3 + 0.0378 x12 - 0.0107 x22 - 0.0208 x32 - 0.0114 x1x2 + 0.0427 x1x3) + 0.0237 x2x3 4.4.5 Optimisation process To optimise the process variables (salt (g), fermentation time (h) and rennet concentration (ml)) for ‘wagashie’ production with respect to the responses, the contour plots were overlaid on one axis of fermentation time (h) and salt (g) at a constant rennet concentatration of 5.35 ml. Fig 4.15 shows an overlay of contour plots for the responses, texture, taste, colour and overall acceptability after the consumer assessment of ‘wagashie’. From Fig 4.15, the optimum levels for the Overall Acceptability 6.5 7.0 0.0 Fermentation time(hr) 1.5 3.0 7.5 4.5 12 10 8 14 Salt (g) Hold Values Rennet (ml) 5.35 University of Ghana http://ugspace.ug.edu.gh 97 preparation of ‘wagashie’ with maximum texture, taste, colour and overall acceptability were 4 h fermentation time, 11 to 14 g of salt and 5.35ml of rennet. The corresponding optimised values of the process variables and the predicted values for taste, texture, colour and overall acceptability were 7.37, 6.68, 7.37 and 7.22 respectively at 11 g of salt and 4 h fermentation time. For the purpose of the study, the optimum conditions were chosen and prepared and compared with ‘wagashie’ prepared with non fermented milk with rennet concentration of 5.35 ml and 11g of salt. Using a 9- point hedonic scale, a sensory evaluation was carried out in order to confirm the predicted values obtained after the optimisation process. The products assessed were in the fresh, freid and smoked forms. Figure 4.15: Contour plot for texture, taste, colour and overall acceptability of ‘wagashie’ overlaid on one axis of fermentation timeand salt at a constant rennet concentration of 5.35ml. Fermentation time(hr) S a l t ( g ) 43210 14 13 12 11 10 9 8 7 Hold Values Rennet (ml) 5.35 University of Ghana http://ugspace.ug.edu.gh 98 Figure 4.16: The mean pH of wagashie samples combined by Box Behnken Design using Rennet as coagulant Figure 4.16 shows the mean values for the pH of the wagashie samples. High pH values were recorded in the samples. Overlapping error bars indicates a non significant difference at p<0.05 whiles error bars that do not overlap represents a significant difference between the samples. The highest pH recorded was as 6.83 and the lowest was recorded as 6.15. Significantly, not much differences were observed in the samples. 5.4 5.6 5.8 6 6.2 6.4 6.6 6.8 7 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 p H Run Number University of Ghana http://ugspace.ug.edu.gh 99 Figure 4.17: The TTA of wagashie samples combined by Box Behnken Design using Rennet as cogulant Table 4.9 shows the TTA of the ‘wagashie’ samples determined after preparation. Overlapping error bars indicates a non significant difference at p<0.05 whiles error bars that do not overlap represents a significant difference between the samples. The TTA values recorded ranged from 0.25 as the minimum and 0.79 as the maximum values recorded in the samples. 0 0.2 0.4 0.6 0.8 1 1.2 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 TT A Run Number University of Ghana http://ugspace.ug.edu.gh 100 4.5.0 Sensory Evaluation 4.5.1 Affective Sensory Evaluation The optimised ‘wagashie’ samples were prepared and processed and an affective sensory test was carried out to confirm the predicted ratings. The results obtained after the consumer preference testing carried out on the wagashie samples are shown in table 4.3. Table 4.3: Mean scores for the confirmatory affective sensory evaluation Sample code Taste Colour Texture Overall acceptability 980 5.95±1.91bc 7.45±0.89a 6.80±1.28a 6.35±1.87ab 591 6.70±1.69ab 7.15±1.31a 6.80±1.39a 6.65±1.42ab 115 7.45±1.23ab 7.15±1.31a 7.55±1.23a 7.65±1.09a 744 7.00±1.84ab 7.00±1.84a 7.20±1.85a 7.00±1.78ab 410 7.81±0.85a 6.77±1.65a 7.10±1.39a 7.73±0.83a 483 6.64±1.81abc 7.73±1.03a 7.05±1.17a 6.86±1.61ab 276 5.13±2.08c 6.96±1.66a 6.81±1.22a 5.73±2.07b Means with the same alphabet are not significantly different from each other at p<0.05. 980: Fermented- Rennet coagulated- fresh 591: Fermented- extract coagulated- fresh 115: Fermented- rennet coagulated- fried 744: Rennet coagulated- fermented smoked 410: Non fermented- extract coagulated - smoked 483: Non fermented -Rennet coagulated- fresh 276: Non fermented -rennet coagulated -smoked Table 4.3 shows the scores for the confirmatory affective test. Scores for taste ranged between 5.1 and 7.8, with Extract coagulated non- fermented smoked ‘wagashie’ being rated significantly higher. The rating for colour and texture however was not significantly different among the various treatments. This observation suggests that, University of Ghana http://ugspace.ug.edu.gh 101 regardless of the treatment, this attribute was equally liked by the panellists. The most liked ‘wagashie’ was the “Extract coagulated non-fermented smoked”. This sample obtained a score of 7.7, which may be interpreted as “like very much” on the 9-point hedonic scale, whereas the least preferred was “Rennet coagulated non- fermented -smoked”. Significantly, there was no difference in the acceptability of the non fermented extract coagulated smoked sample and the fermented rennet coagulated fried sample at p< 0.05.The rennet coagulated smoked sample had a score of 7.00 which was rated as liked very much on the 9-point hedonic scale. The pH and TTA of the ‘wagashie’ samples obtained after preparation are recorded in table 4.4. Significant differences were observed in the samples for pH and TTA values at p<0.05. Low pH values with corresponding high TTA values were recorded in the fermented samples as observed in the fermented- rennet coagulated- fresh and smoked samples; 980 and 744 respectively and the fermented–extract coagulated- fresh sample; 591 ,except the fermented rennet coagulated fried sample; 115. The non fermented samples also recorded high pH values with low TTA values as observed in the non fermented extract coagulated smoked sample; 410, the non fermented – rennet coagulated- fresh sample;483 and the non fermented- rennet coagulated –smoked sample;276. The pH and TTA values recorded for the samples that were used for confirmatory affective testing are shown in Table 4.4. University of Ghana http://ugspace.ug.edu.gh 102 Table 4.4: Mean pH and TTA values of the optimised ‘wagashie’ samples after the confirmatory affective sensory evaluation. Sample Ph TTA 980: Fermented- Rennet coagulated - fresh 5.81±0.01cd 0.95±0.43a 591: Fermented- extract coagulated - fresh 5.73±0.01d 0.88±0.08a 115: Fermented- rennet coagulated - fried 5.87±0.01c 0.34±0.08ab 744: Rennet coagulated- fermented- smoked 5.33± 0.06e 0.79±0.00a 410: Non fermented- extract coagulated - smoked 6.89±0.01b 0.36±0.00ab 483: Non fermented -Rennet coagulated - fresh 7.09± 0.01a 0.11±0.05b 276: Non fermented -rennet coagulated - smoked 7.08± 0.01a 0.34±0.03ab Means with the same alphabet are not significantly different from each other at p<0.05 4.5.2 Quantitative Descriptive Sensory Evaluation Twenty-two descriptive terms were generated for the ‘wagashie’ after which a formal evaluation was carried out. After the formal evaluation, complete data sets were obtained for the 13 judges. Some of the descriptors were considered as desirable and undesirable. The results of the ANOVA for the attribute ratings across the 14 samples for the 13 judges are summarized in Table 4.5.Significant differences occurred between the samples with regard to the texture, taste, colour and aroma. 4.5.2.1 Taste Significant differences occurred between the samples for taste. However no significant differences were observed in the fermented- rennet coagulated- fresh sample; 614, the fermented - extract coagulated –fresh sample; 616, the fermented rennet coagulated - fried sample;421 and the fermented -extract coagulated- fried University of Ghana http://ugspace.ug.edu.gh 103 sample; 706 for sour taste at p<0.05. No significant differences were observed in the samples for bland and cheesy taste. There were significant differences between the Fermented- rennet coagulated- fried sample; 421 and the Non -fermented- rennet coagulated- fresh sample; 110 for milky taste. The fresh market sample; 997 was scored highest for bitter and bland tastes while the fried market sample;246 was scored highest for sour taste. The fermented- rennet coagulated- fried sample; 421 was scored highest for fried egg taste and salty taste. 4.5.2.2 Colour The colour of ‘wagashie’ was assessed for white colour, for the interior part of the product and brown colour, for the exterior part of the product. Significant differences occurred in the scores for colour for the samples.However, there were no significant differences between the whitish colour of the non fermented -rennet coagulated- smoked sample; 203, the fermented- rennet coagulated- fried sample; 961, the fermented- extract coagulated- fried 706 and the non fermented extract coagulated fried; 707. There were also no significant difference in the brownish colour of the fermented rennet coagulated smoked; 417; Non fermented extract coagulated fried sample; 707, Non fermented extract coagulated smoked sample;961 at p<0.05. There were no significant differences in the whitish colour of the fermented and non-fermented rennet and extract coagulated fresh samples, a simmilar trend was observed in the fried samples at p<0.05. For the smoked samples, there were significant differences between the fermented –rennet coagulated smoked sample; 417and the non- fermented rennet and extract- coagulated –smoked samples 203 and 961 respectively for the interior white colour University of Ghana http://ugspace.ug.edu.gh 104 of the product. Colour and flavour attributes are very important cheese acceptance criteria for consumers. 4.5.2.3 Aroma From table 4.5, significant differences occurred in the aroma of the samples however there were no significant differences for cheesy aroma, beefy aroma and fermented cassava dough aroma in the samples. There was a significant difference between the fried market sample 246 and the improved fried samples (Fermented- rennet coagulated- fried sample; 421, non fermented -rennet coagulated- fried sample; 101, fermented -extract coagulated- fried sample; 706, non fermented - extract coagulated –fried sample; 707) for spoilt milk aroma. Significant differences occurred in the samples for milky aroma, yoghurt aroma, fried sweet potato aroma, doughnut aroma, fried ripe plantain aroma and smoked chevon aroma. Aroma is an important component of the sensory property of cheese and is one of the first stimuli to be perceived before consumption. 4.5.2.4 Texture Generally, significant differences were observed in the texture of the samples. Considering the various descriptors for texture, there were no significant differences in the scores for soft texture between the fermented extract coagulated fried sample; 706, the fried market sample; 246 and the non fermented- rennet coagulated- fresh sample; 110. Significant differences were observed in the fermented-rennet coagulated- smoked sample; 417, the fermented-extract coagulated fresh sample; 616, the non fermented extract coagulated fresh sample; 355 and the fresh market sample; 997 for spongy texture. No significant differences were observed in the University of Ghana http://ugspace.ug.edu.gh 105 fermented rennet coagulated smoked sample; 417, the non- fermented – extract coagulated smoked sample; 961, the non- fermented –rennet coagulated-freid sample; 101, the non fermented-extract coagulated- fried sample; 707 and the non fermented- extract coagulated-fresh sample; 355 for smooth texture. The fresh market sample; 997 had the highest score for soft texture,the non fermented-extract coagulated- fried sample;707 had the highest score for spongy texture, the fermented-extract coagulated fresh sample; 616 had the highest score for smooth texture and the non fermented- extract coagulated- fresh sample 355 had the highest score for crumbly texture. University of Ghana http://ugspace.ug.edu.gh 106 Table 4.5: Mean values for wagashie descriptors during the Quantitative Descriptive Analysis Sample Sensory attributes Soft texture Spongy texture Smooth texture Crumbly texture Milky aroma Yoghurt aroma Cheesy aroma Beefy aroma Fried sweet potato aroma Fried ripe plantain aroma Smoked chevon aroma spoilt milk aroma F 110 5.33±2.3abc 6.63±1.7a 4.58±2.8abcd 5.48±2.4a 5.98±2.2a 4.18±3.0abcd 5.34±2.7a 4.25±2.7a 0.93±1.5bcde 0.89±1.6cdef 1.20±1.8bcd 1.26±1.6b 355 3.74±1.9bcde 6.28±1.4a 3.72±3.0cd 6.35±2.0a 6.01±2.2a 4.61±2.7abcd 4.32±2.7a 3.84±3.0a 0.77±1.5cde 0.57±1.0def 1.98±2.7abcd 1.97±2.0ab 101 3.53±2.3cde 6.67±1.6a 3.53±2.7cd 5.56±2.8a 4.40±2.4ab 2.99±2.6cd 3.69±2.5a 2.76±2.6a 3.82±2.5a 3.31±2.9b 1.26±1.5bcd 1.14±1.5b 838 6.52±1.5ab 4.86±2.3ab 7.01±1.5ab 4.46±2.5ab 6.03±1.9ab 5.76±2.2adc 5.09±1.7a 2.77±2.0a 0.20±0.3de 0.18±0.3abcde 1.25±1.7bcd 2.42±1.8ab 616 7.63±1.2a 2.31±2.0b 7.54±1.7a 2.03±2.3b 6.21±1.8a 7.54±0.9a 5.61±2.1a 2.43±2.4a 0.27±0.4de 0.19±0.3ef 0.75±1.4bcd 1.64±1.9ab 614 5.19±2.0abcd 3.88±2.2ab 5.42±2.1abc 5.54±2.2ab 5.03±1.8ab 5.83±1.5abc 4.49±2.3a 1.93±2.5a 0.16±0.2de 0.18±0.2ef 0.15±0.2cd 3.32±3.5ab 997 7.07±1.3a 5.00±1.5ab 6.98±1.6ab 5.84±2.1a 5.56±2.3ab 6.43±2.3ab 3.82±2.7a 1.61±2.4a 0.21±0.4e 0.14±0.2f 0.19±0.3d 4.45±3.2a f 246 5.40±1.9abc 5.98±1.5a 4.80±1.4abcd 5.39±1.8a 4.03±2.4ab 2.40±2.5cd 3.17±2.2a 2.00±2.4a 3.11±2.5abcde 3.61±2.8abc 1.12±1.7bcd 2.54±2.4ab 421 4.03±2.0bcde 5.53±2.5a 3.86±1.6bcd 4.83±1.9ab 3.07±1.5ab 3.09±2.2bcd 2.59±1.9a 1.98±1.7a 4.24±2.2a 4.16±3.0ab 0.75±1.1bcd 0.72±0.9b 706 5.39±1.2abc 6.25±1.5a 5.58±1.9abc 3.63±2.2ab 2.52±1.3b 2.61±2.2cd 2.02±1.8a 1.34±1.3a 3.73±2.6ab 3.73±2.6abcd 0.77±1.5bcd 0.74±0.8b 707 3.32±2.6cde 6.76±2.1a 3.05±2.3cd 5.23±2.5ab 4.56±2.8ab 2.86±2.6cd 4.14±2.7a 2.26±3.2a 3.68±3.0ab 4.37±3.2a 1.42±2.1bcd 1.03±1.1b S 203 2.15±1.9e 6.17±2.0a 2.02±2.0d 5.90±2.1a 4.86±2.3ab 2.99±2.7cd 3.94±2.3a 3.85±3.3a 3.16±3.4abcd 2.42±2.1abcdef 3.05±3.2abc 1.96±2.1ab 961 1.71±1.8e 6.07±2.7a 2.67±2.6cd 5.64±2.3a 5.59±2.3ab 3.23±3.0cd 4.44±1.9a 3.71±3.2a 1.93±2.3abcde 1.63±2.5abcdef 3.58±3.1ab 2.16±2.5ab 417 2.47±1.8de 4.90±2.3ab 3.04±2.2cd 4.43±2.9ab 3.35±2.7ab 1.70±1.9d 2.55±2.2a 2.85±2.7a 1.54±2.4abcde 1.22±1.9bcdef 4.79±3.0a 0.85±1.4b Means in the same column with the same letters are significantly different (p<0.05), F= Fresh, f= fried, S= Smoked Fresh samples: Fermented rennet coagulated; 614, market sample; 997, Non fermented- rennet coagulated; 110, Non fermented- extract coagulated; 355, Fermented- extract coagulated with extra fat of 63g; 838, Fermented- extract coagulated; 616. Fried samples: Fermented -rennet coagulated; 421, fermented -extract coagulated; 706, market sample; 246, Non fermented extract coagulated; 707, non fermented -rennet coagulated; 101 Smoked samples: Fermented rennet coagulated; 417, Non -fermented extract coagulated; 961, Non fermented rennet coagulated; 203 University of Ghana http://ugspace.ug.edu.gh 107 Table 4.5 contd: Table 4.5: Mean values for wagashie descriptors during the Quantitative Descriptive Analysis Means in the same column with the same letters are significantly different (p<0.05), F= Fresh, f= freid, S= smoked Fresh samples: Market wagashie; 997, Fermented rennet coagulated; 614, Fermented- extract coagulated; 616, Fermented- extract coagulated+ 63g of fat; 838, Non- fermented- rennet coagulated; 110, Non- fermented- extract coagulated; 355. Fried samples: Fermented -rennet coagulated; 421, Fermented -extract coagulated; 706, Market wagashie; 246, Non- fermented - extract coagulated; 707, Non- fermented - rennet coagulated; 101 Smoked samples: Fermented- rennet coagulated; 417, Non - fermented extract coagulated; 961, Non- fermented rennet coagulated; 203 Sample Sensory Attributes Fermented cassava dough aroma Doughnut aroma Sour taste Bland Cheesy taste Milky taste Bitter taste Fried egg taste Salty taste Whitish colour Brownish colour F 110 0.40±0.8a 0.54±0.8 ab 1.10±0.1.0b 1.16±1.7b 5.16±2.3a 6.30±1.5a 0.16±0.2c 2.27±2.5abc 1.12±1.4bc 6.56±1.4a 1.58±2.5d 355 0.50±1.1a 0.45±0.8 ab 1.62±2.2b 1.27±1.8b 3.84±2.4a 5.88±2.7ab 0.33±0.5c 3.09±2.9abc 1.25±1.6abc 5.97±1.3ab 0.65±1.9d 101 0.45±0.9a 1.87±1.9 ab 1.30±1.6b 0.59±0.8b 4.74±2.1a 5.14±2.2ab 0.38±0.7c 4.15±2.3a 1.97±2.2abc 6.05±1.4ab 5.14±2.4bc 838 1.14±1.5a 0.22±0.3 ab 3.06±2.5ab 1.57±2.1ab 4.20±2.4a 4.79±1.9ab 0.28±0.3c 0.71±1.0bc 2.29±2.4abc 5.89±2.0ab 0.29±0.7d 614 1.04±1.3a 0.19±0.2 ab 3.13±2.9ab 0.62±0.6b 4.56±2.3a 4.54±1.5ab 0.59±1.2c 0.37±0.3c 4.10±2.8ab 5.20±1.8ab 0.26±0.5d 616 1.44±1.7a 0.28±0.4 ab 3.01±2.4ab 1.80±1.8ab 4.71±2.6a 5.02±2.5ab 0.41±0.41c 0.74±1.1bc 1.72±2.2abc 6.09±2.1ab 0.16±0.4d 997 3.57±3.0a 0.26±0.5 b 2.13±1.9b 2.85±2.8a 3.70±2.8a 3.38±2.9ab 2.07±2.0a 0.48±0.9c 1.03±1.6bc 6.78±1.5a 0.15±0.2d 246 0.60±0.9a 1.19±1.5ab 5.48±2.5a 1.61±2.0b 4.0±2.0a 3.73±2.43ab 2.33±2.1b 2.24±2.2abc 1.99±1.8abc 5.78±1.4ab 7.92±1.9d f 707 0.39±0.9a 2.12±2.0 a 1.13±1.7b 1.03±1.4b 4.26±2.2a 5.08±2.5ab 0.17±0.3c 4.60±2.7a 1.29±1.7abc 5.53±1.5ab 4.84±1.8bc 706 0.79±1.9a 1.73±1. ab 4.05±2.3ab 1.16±1.4b 3.11±2.1a 3.52±2.0ab 1.07±1.5c 2.47±2.3abc 3.26±3.0abc 5.76±1.4ab 6.84±2.0abc 421 0.32±0.4a 2.38±1.6 a 3.49±1.9ab 0.72±0.8b 2.81±1.7a 2.70±1.2b 0.83±1.1c 2.92±2.3abc 4.21±3.0a 5.38±1.4ab 7.23±1.5abs S 203 0.61±1.3a 1.00±1.7 ab 1.57±2.0b 0.89±1.3b 3.67±2.5a 4.75±2.8ab 0.42±0.6c 3.98±2.2ab 1.15±1.2bc1 5.50±1.7ab 4.48±2.3c 961 0.67±1.3a 0.85±1.74 ab 1.92±2.3b 0.89±1.1b 3.67±2.1a 5.06±2.1ab 0.28±0.31c 4.08±2.6a 1.80±2.1abc 4.98±1.6ab 5.49±1.4bc 417 0.66±1.8a 0.60±1.0 ab 1.39±1.7b 2.69±2.6b 2.86±2.3a 3.35±3.1ab 0.24±0.3c 2.38±2.5abc 0.55±0.9c 3.94±1.9b 5.51±2.2bc University of Ghana http://ugspace.ug.edu.gh 108 4.6.0 Principal composite analysis (PCA) In the PCA, all attributes were considered, and it was applied on the mean values of each individual attribute per ‘wagashie’ sample as obtained from quantitative descriptive sensory analysis of aroma, texture, taste and colour of wagashie. Figure 4.8 graphically represents a PCA bi-plot of the attributes and the position of the different ‘wagashie’ samples relative to the attributes that were rated. Figure 4.18 is a graphical representation of the relationship between the sensory attributes and the ‘wagashie’ samples. The total variation of the data for F1 and F2 was 65.97%. Factor F1 accounted for 41.3% of the total variation which shows the highest percentage variation among the samples. It was defined mainly by the attributes, milky taste, milky aroma, cheesy aroma, cheesy taste and yoghurt aroma on the positive side and beefy aroma, smoked chevon aroma, fried egg aroma, crumbly, and spongy on the negative side. F2 accounted for 24.66% which shows the second largest percentage variation among the samples. The non fermented- rennet coagulated- fresh sample; 110, the Non fermented extract coagulated fresh; 355, the fermented extract coagulated sample with extra fat of 63g; 838 and the Fermented extract coagulated; 616 had high intensities for milky taste, milky aroma, cheesy aroma, cheesy taste and yoghurt aroma which denotes a positive correlation, this was represented in the second quadrant. However, they had low intensities for fried sweet potato aroma, doughnut aroma, fried ripe plantain aroma, salty taste and brownish colour because they in opposite quadrants and thus are negatively correlated. University of Ghana http://ugspace.ug.edu.gh 109 Figure 4.18: PCA bi-plot of quantitative descriptive sensory used to describe the sensory attributes of ‘wagashie’ in their fresh, fried and roasted forms. 997246 203 110 355961 707 101 838 614 616 706 421 417 Soft Spongy Smooth Crumbly Milky aroma Yoghurt aroma Cheesy aroma Beefy aroma Fried sweet potato aroma Fried ripe plantain aroma Smoked chevon aroma spoilt milk aroma Fermented cassva dough aroma Doughnut aroma Sour taste Bland Cheesy taste Milky taste Bitter taste Fried egg taste Salty taste Whitish colour Brownish colour -10 -8 -6 -4 -2 0 2 4 6 8 10 -12 -10 -8 -6 -4 -2 0 2 4 6 8 10 12 14 F2 ( 2 4 .6 6 % ) F1 (41.31 %) Biplot (axes F1 and F2: 65.97 %) Non-fermented samples Fermented samples Market samples Fresh samplesFried samples Smoked samples NB:101 and 707 are fried Fresh samples University of Ghana http://ugspace.ug.edu.gh 110 Also, they have similar attributes with the Fermented rennet coagulated; 614 and the market sample; 997 because they were all fresh ‘wagashie’ samples and were on the same side (positive) on the F1 axis. The Fermented rennet coagulated fresh sample; 614 and the fresh market sample; 997 had positive correlation with whitish colour, spoilt milk aroma, fermented cassava dough aroma, soft, smooth, bitter taste, sour taste and bland taste .They however had low intensities for beefy aroma, smoked chevon aroma, fried egg aroma, crumbly, and spongy. The fermented rennet coagulated -smoked sample; 417, the Non -fermented extract coagulated –smoked sample; 961, Non fermented rennet coagulated- smoked sample; 203, Non fermented extract coagulated- fried sample; 707 and the non fermented -rennet coagulated fried sample; 101 had high intensities for beefy aroma, smoked chevon aroma, fried egg aroma, crumbly texture, and spongy texture and have positive correlation. However, they had low intensities for whitish colour, spoilt milk aroma, fermented cassava dough aroma, soft, smooth, bitter taste, sour taste and bland taste because they were on opposite quadrants and were negatively correlated. The fresh market sample; 997, had positive correlation with almost all the undesirable descriptors which were bitter taste, fermented cassava dough aroma, spoilt milk aroma and bland taste. The fermented rennet - coagulated fresh sample; 614 was slightly associated with the undesirable descriptors which were bitter taste, fermented cassava dough aroma, spoilt milk aroma and bland taste because they were in the same quadrant. University of Ghana http://ugspace.ug.edu.gh 111 4.6.1 Cluster Analysis The results achieved by the Hierarchial Agglomerate Cluster Analysis using XLSTAT 2014 for the thirteen ‘wagashie’ samples considering all the sensory variables are presented as a dendrogram in Figure 4.19. Five clusters were formed. Group A (997,838,614 and 616) was made up of fresh fermented samples, both rennet and extract coagulated except sample 997(market sample). Group B was formed by 246,706,421, which were all fried fermented and non- fermented both rennet and extract coagulated samples. Group C was made up of 4 samples (two smoked non fermented rennet and extract coagulated samples; 203, 961 and two fried non -fermented both rennet and extract coagulated samples; 101 and 707). Group D was formed by two samples; which were non fermented rennet coagulated fresh ‘wagashie’ sample, 110 and a non fermented extract coagulated fresh ‘wagashie’ sample, 355. Group E was formed by one sample; 417 representing fermented - rennet coagulated- smoked sample. These results showed that there were differences among the ‘wagashie’ samples in different groups and the sensory data may contain enough information to attain differences in the ‘wagashie’ samples based on the classes established. However, samples that belong to the same class had similar attributes and were not different from each other. The cluster analysis was used to ascertain the similarities and the disimilarities among the samples. University of Ghana http://ugspace.ug.edu.gh 112 Group A Group B Group C Group D Group E Figure 4.19: Dendogram from cluster analysis of ‘wagashie’ samples considering sensory attributes 9 97 6 16 8 38 6 14 2 46 7 06 4 21 1 10 3 55 4 17 7 07 1 01 2 03 9 61 0 50 100 150 200 250 300 D is si m ila ri ty Dendrogram University of Ghana http://ugspace.ug.edu.gh 113 Figure 4.20: Spider plot of the fresh and processed ‘wagashie’ samples after the quantitative descriptive sensory evaluation 0.00 1.00 2.00 3.00 4.00 5.00 6.00 7.00 8.00 Soft Spongy Smooth Crumbly Milky aroma Yoghurt aroma Cheesy aroma Beefy aroma Fried sweet potato aroma Fried ripe plantain aroma Smoked chevon aroma spoilt milk aroma Fermented cassva dough aroma Doughnut aroma Sour taste Bland Cheesy taste Milky taste Bitter taste Fried egg taste Salty taste Whitish colour Brownish colour 997 246 203 110 355 961 707 101 838 614 616 706 421 417 University of Ghana http://ugspace.ug.edu.gh 114 4.6.2 Spider plot Figure 4.20 shows the spider plot of the ‘wagashie’ samples assessed during the quantitative descriptive sensory evaluation which shows the descriptive sensory analysis profile more vividly. Each spoke symbolizes one attribute, and the relative intensity corresponds to that point in which the product line crosses, with the lowest and highest intensities toward the centre point and farthest from the centre, respectively. From the spider plot, the non fermented- extract coagulated- fried sample; 707 had the highest intensity, whiles the rennet coagulated fresh sample; 614 had the lowest mean for fried egg taste. The market fresh sample; 997 had the highest intensity with a mean of 4.45 for spoilt milk aroma. The market sample 997 had the highest intensity with a mean of 6.78 for whitish colour. The fried market sample; 246 had the highest intensity with a mean of 7.92 for brownish colour. The fried market sample; 246 with a mean of 2.33 and the fresh market sample with mean; 2.07 had the highest intensity for bitter taste whiles the non- fermented-rennet coagulated fresh sample; 110 had the lowest intensity for bitter taste with a mean of 0.16. The non- fermented-rennet coagulated fresh sample; 110 had the highest intensity with a mean of 6.30, 5.16 for milky and cheesy tastes, whiles the fermented-rennet coagulated fried sample; 421 with a mean of 2.70, 2.81 had the lowest intensity for milky and cheesy tastes respectively. The fresh market sample; 997 with a mean of 2.85, had the highest intensity for bland taste. The fermented- rennet coagulated fried sample; 421 with a mean of 2.38 had the highest intensity for sour taste. The fermented –rennet coagulated -smoked sample; 417 with a mean of 4.79 had the highest intensity followed by the non-fermented rennet coagulated- smoked sample; 203 with mean 3.05 for smoked chevon aroma. The non-fermented extract coagulated fresh sample; 355 with a mean of 6.35 had the highest intensity University of Ghana http://ugspace.ug.edu.gh 115 for crumbly texture whiles the fermented-extract coagulated fresh sample; 616 with a mean of 2.03 had the lowest intensity for crumbly texture. The fermented-extract coagulated fresh sample; 616 with a mean of 7.54, had the highest intensity smooth texture.The non-fermented rennet coagulated –fresh sample; 110, with a mean of 5.16 and 6.30 had the highest intensities for cheesy and milky tastes respectively. The fermented extract –coagulated fresh sample; 616 with a mean of 5.61 and 7.54 had the highest intensities for cheesy and yoghurt aroma. Significant differences occurred among the samples at p<0.05 in relation to attribute intensities and this is shown by the different superscript alphabet in table 4.5, except fermented cassava dough aroma, cheesy taste, bland taste, beefy aroma and cheesy aroma. 4.7.0 Proximate composition of ‘wagashie’ The chemical analysis of the ‘wagashie’ samples was carried out to determine the nutritive value of the fermented rennet coagulated wagashie and the results can be observed in table 4.15. Table 4.6: Mean values for the chemical composition of ‘wagashie’ Sample Ash g/100g FFAg/100g Protein Fat/100g Moistureg/100g Fermented Smoked wagashie 1.21±0.014c 0.53±0.01a 30.18 ± 0.71a 21.40±0.69b 41.99 ± 0.09c Non-Fermented fresh wagashie 1.95±0.007b 0.21±00b 23.20 ± 0.24c 14.49±1.33c 45.71±0.12b Fermented fresh wagashie 0.95±0.056d 0.16±0.07b 19.28 ± 0.18d 15.48±0.06c 56.10±0.21a Non –fermented Fried wagashie 2.09±0.311a 0.16±0.08b 24.97±0.12b 25.32±0.31a 37.69±0.11d Means with the same superscript alphabets in the same column are not significantly different at p<0.05 From Table 4.6, significant differences were observed in the mean value for ash at p< 0.05, the mean values for ash ranged from 0.95 g to 2.09 g with the non- University of Ghana http://ugspace.ug.edu.gh 116 fermented fried ‘wagashie’ sample rated significantly higher. The mean values for FFA ranged from 0.16g to 0.53g. The mean values for fat ranged from and 14.49 g to 25.32 g . Also the mean values of protein ranged from 19.28 to 30.18 with the fermented smoked wagashie sample significantly rated higher . The mean values for moistureranged from 37.69 to 56.10 with the fermented fresh ‘wagashie’ sample significantly rated higher. No significant differences were observed between the mean value of the fat and FFA in the fermented and non –fermented - fresh ‘wagashie’ sample at p<0.05. The non fermented- fried ‘wagashie’ sample recorded the highest mean value for ash and fat whiles the fermented smoked ‘wagashie’ sample recorded the highest mean value for protein. The fermented fresh sample recorded the highest value for moisture. 4.8.0 Rheology of Improved ‘wagashie’ 4.8.1 Colour The average values of colour parameters for ‘wagashie’ obtained from instrumental measures are shown in Table 4.7. Table 4.7: Mean values with standard deviations for the colour of improved ‘wagashie’ Sample L* a* b* Fermented smoked extract (interior) 79.25±1.5c -7.15±0.3a +15.42±0.1ab Fermented smoked extract (exterior) 57.00±2.8d +2.26±0.5d +17.50±3.2b Fermented smoked rennet (interior) 79.84±0.4bc -6.42±0.2abc +18.42±0.4ab Fermented smoked rennet (exterior) 59.83±1.1d +0.69±0.6e +15.69±0.3b Fermented extract fresh 83.36±0.9ab -5.58±0.2ab +17.50±0.4ab Fermented rennet fresh 84.11±1.2a -5.58±0.2c +19.43±0.7a Non fermented rennet fresh 84.34±0.5a -6.13±0.1bc +19.29±0.6a Means with the same alphabets in the same column are not significantly different from each other at p<0.05 University of Ghana http://ugspace.ug.edu.gh 117 Table 4.7 shows the mean values with standard deviations for the colour of the fresh, fried and smoked ‘wagashie’ samples at p<0.05. Significant differences occurred in the L* value of the samples however, no significant difference was observed between the L* mean value of the fermented and non-fermented rennet coagulated fresh samples at p<0.05. Significant differences were not observed in the exterior colours of the rennet coagulated and extract coagulated smoked samples.The lightness index of the ‘wagashie’ ranged from 84.34 to 57. The higher the L* value, the lighter the colour and the lower the L* value the darker the colour. The mean value for a* colour ranged from -0.69 to +7.15. The interior colour of the smoked samples had positive mean values for a* whiles the interior colour of the fresh samples had negative mean values of a*. Significant differences, thus occurred between the samples for a* value at p< 0.05. Significant differences occurred between the samples for the mean values of b*, however, no significant difference were observed between the interior colours of the extract coagulated - smoked sample, the rennet coagulated -smoked sample and the extract coagulated fresh sample. Also significant differences were not observed in the mean values of b* for the exterior colour of both extract and rennet coagulated smoked samples and between the rennet coagulated fermented and non fermented fresh samples. The mean values for b* ranged from +19.43 to +15.42. 4.8.2 Texture Evaluation of texture by instrumental and sensory analysis is important to new product development.The results for Texture Profile Analysis (TPA) for ‘wagashie’ are compiled in Table 4. 8. Five parameters were measured; hardness, adhesiveness, University of Ghana http://ugspace.ug.edu.gh 118 springiness, gumminess and chewiness. The samples analysed were rennet coagulated fermented and non –fernemented smoked, fried and fresh samples Table 4.8: Mean values for the textural characteristics of ‘wagashie’ Texture Characteristics Sample Springiness (mm) Adhesiveness g/sec Chewiness (N) Gumminess (N) Hardness Hardness (g) Non fermented fresh 0.915 ± 0.02a -1.425 ± 1.97a 2086 ± 106c 1984 ± 352cd 4862 ± 890bc Non fermented fried 0.912 ± 0.03a -1.014 ± 0.35a 6397 ± 896a 6982 ± 1152a 12144 ± 3078a Non fermented smoked 0.909 ± 0.03a -0.790 ± 0.35a 2822 ± 106bc 2729 ± 602bc 4578 ± 1300bc Fermented fried 0.866 ± 0.08a -1.274 ± 0.17a 658 ± 104d 1222 ± 223de 6372 ± 951b Fermented smoked 0.713 ± 0.08b -0.872 ± 0.66a 2930 ± 345b 3475 ± 473b 3931 ± 590c Fermented fresh 0.635 ± 0.06c -0.993 ± 0.64a 516 ± 124d 695 ± 200cd 4164 ± 1218c Mean with the same superscript alphabets in the same column are not sisgnificantly different at p<0.05 From table 4.8, Textural differences were observed in the samples though there was no significant difference between the values for adhesive texture at p>0.05. The mean values for adhesive texture ranged from –0.790 to -1.425. Significant differences were observed between the fermented smoked and fresh samples for springiness at p<0.05. However, no significant difference were observed between the non fermented fresh, fried and smoked samples and the fermented fried sample for University of Ghana http://ugspace.ug.edu.gh 119 springiness at p<0.05, the values for springiness ranged from 0.653 to 0. 915. Significant differences were observed for chewiness at p<0.05, however, no significant differences were observed between the fermented fresh sample and the fermented fried sample at p<0.05. The mean value for chewiness ranged from 516 to 6397. Also significant differences were observed in the samples for gumminess at p<0.05. The mean values for gumminess ranged from 694 to 6982. There was no significant difference between the hard texture of the fermented fresh sample and the fermented smoked sample. The non fermented fried sample recorded the highest value for hardness while the fermented smoked sample recorded the lowest mean value for hard texture. 4.9.0 Safety of improved ‘wagashie’ To ascertain the safety of the improved ‘wagashie’, microbiological tests were carried out on the samples. The tests carried out were similar to the tests carried out on the market ‘wagashie’ samples. Table 4.9 shows the results of the microbiological tests done on the improved ‘wagashie’ samples. Significant differences were observed between the two samples for the count for coliform, E. coli and yeast with the fermented fresh sample rated significantly higher than the fermented smoked sample at p<0.05. Generally, it was observed that the count for microorganisms in the fermented fresh sample was significantly higher than the fermented smoked sample. There were no counts for Salmonella, Staphylococcus aureus, Bacillus cereus and mould in the two samples. University of Ghana http://ugspace.ug.edu.gh 120 Table 4.9: Mean microbial counts in the improved ‘wagashie’ in CFU/g to ascertain the safety of the lobratory prepared samples Means with the same superscript alphabets in the same column are not significantly different at p<0.05 4.10.0 Shelf life of wagashie Table 4.10 shows the mean count for aerobic mesophiles, yeast and mould during the ambient storage of the ‘wagashie’ samples for 5 weeks. Twenty grams samples were analysed immediately after preparation for day 0 and the rest were packaged for irradiation. The microbial count in the irradiated samples reduced after irradiation (Day 3) whiles that for the non-irradiated samples increased. Sample Coli form E. coli Yeast Moul d Aerobic mesophiles Salmonel la Staphyloc occus Bacillus cereus Fermented fresh 3x10a 3x10a 9 x 103a 0 2.9x105 a 0 0 0 Fermented smoked 0b 0b 8x102b 0 0b 0 0 0 University of Ghana http://ugspace.ug.edu.gh 121 From Table 4.10, after the third day, there was a slight increase in the count for aerobic mesophiles and yeast and no count for mould. During the first week, there were counts for mould mostly in both the irradiated and non-irradiated control samples (normal packaged). The non-irradiated vacuum packaged samples however recorded a slight growth for mould whiles the irradiated vacuum packaged samples did not have counts for mould. The counts for aerobc mesophiles and yeast also increased. During the second week of storage growth of aerobic mesophiles, yeast and mould increased in the normal/non- vacuum packaged irradiated and non-irradiated samples (nNS, nNF, nIF, nIS). From the third week to the fifth week of storage, there were no mould growth in the irradiated vacuum packaged samples.The growth of aerobic mesophiles and yeast in the irradiated and non-irradiated vacuum packaged samples increased in the 3rd and 4th weeks. A slight increase was recorded in week 5 with the non-irradiated vacuum packaged fresh sample recording the highest count for aerobic mesophiles, yeast and mould. University of Ghana http://ugspace.ug.edu.gh 122 Table 4.10: Changes in the mean microbial counts in the rennet coagulated fermented fresh and smoked ‘wagashie’ samples as affected by packaging and irradiation for the 5 weeks storage period in CFU/g (from day 0 to week 2) PCA Yeast Mould Sample Day 0 Day 3 Week 1 Week 2 Day 0 Day 3 Week 1 Week 2 Day 0 Day 3 Week 1 Week 2 VIF - 1.0 x 10 1c 1.0 x 102d 2.3 x 103b - 5.0 x 10c 1.0 x 102c 6.0 x 10 2d - - - - nIS - 4.0 x 10 1c 2.5 x 104cd 1.6 x 106ab - 1.5 x 10c 5.0 x 102c 1.1 x 105cd - - 2.5 x 10 2d 1.5 x 10 2b VIS - 1.0 x 10 1c 4.0 x 101d 3.0 x 102b - 0 0 2.5 x 10d - - 0 - nIF - 3.0 x 10 1c 3.7 x 104c 2.8 x 106ab - 2.0 x 10c 2.5 x 102c 3.0 x 106b - - 2.0 x 103c 3.5 x 104b nNF 2.9x102 3.0 x 103a 2.8 x 105a 3.7 x 107a 9.0x103a 8.0 x102a 9.0 x 103a 1.9 x 107a - - 1.0 x 104a 5.0 x 105a nNS - 2.0 x 103b 2.5 x 104b 1.5 x 106ab 8x102b 5.0 x102b 5.0 x 103b 1.9 x 106b - - 1.3 x 103b 5.0 x 104b VNF - 1.0 x 103c 2.0 x 103d 4.3 x 103b - 1.2 x 10c 8.0 x 102c 1.1 x 103d - - 2.0 x 102d - VNS - 2.0 x 103c 1.7 x 103d 1.5 x 103b - 6.0 x 10c 1.3 x 10c 2.4 x 102d - - 1.0 x 102d - Means with the same superscript alphabets in the same column are not significantly different at p<0.05, ‘ –‘ = none Legend: V- Vacuum packaged, n- non- vacuum packaged I- Irradiated, N- Non-irradiated, F- Fresh, S-Smoked University of Ghana http://ugspace.ug.edu.gh 123 Table 4.10 contd.: Changes in the mean microbial count in the rennet coagulated fermented fresh and smoked ‘wagashie’ samples as affected by the packaging material and irradiation for the 5 weeks storage period in CFU/g (week 3 to week 5). PCA Yeast Mould Sample Week 3 Week 4 Week 5 Week 3 Week 4 Week 5 Week 3 Week 4 Week 5 VIF 1.1 x 10 4c 2.3 x 10 5b 9.0 x 10 6b 6.0 x 10 3c 2.0 x 10 4b 1.7 x 10 4c - - - VIS 1.3 x 10 3d 3.0 x10 5b 8.0 x 10 5c 1.2 x 10 3c 4.0 x 10 4b 1.3 x 10 4c - - - VNF 2.5 x 10 6a 2.0 x10 7a 5.0 x 10 7a 8.0 x 10 5a 2.0 x 10 5a 4.0 x 10 6a 1.6 x 10 3a 8.0 x 10 4a 7.0 x 10 4a VNS 1.9 x 10 5b 1.2 x10 5b 4.0 x 10 6b 1.1 x 10 4b 8.0 x 10 4b 2.0 x 10 5b 4.0 x 10 3b 8.0 x 10 3b 1.0 x 10 4b Means with the same alphabets are in the same column are not significantly different at p<0.05, ‘–‘= none Legend: V- Vacuum packaged, I- Irradiated, N- Non-irradiated, F- Fresh, S-Smoked University of Ghana http://ugspace.ug.edu.gh 124 4.10.1 pH of the wagashie samples during the 5 weeks storage period. The pH values recorded for the product in the fresh and smoked forms for the radiated and non-irradiated samples during the 5 weeks storage period are summerised in Table 4.11. Table 4.11: Mean pH values of the ‘wagashie’ samples during the 5 weeks storage period Sample Day 0 Day 3 (after irradiation) Week 1 Week 2 Week 3 Week 4 Week 5 VIF 4.99 bc 4.96e 4.93f 4.94b 4.96c 5.30c nIS 4.93 d 4.94f 4.92f 5.15a 5.15b 5.20d VIS 5.00 b 4.99d 5.03e 4.89b 5.22a 5.63a nIF 4.88 e 4.95ef 4.88 g 4.89b 5.25a 5.58b nNF 4.99 a 4.97c 5.22a 6.44a - - - nNS 4.90b 5.05a 5.15b 5.54c - - - VNF 4.92d 4.92g 6.09c - - - VNS 4.88 e 5.07c 6.28b - - - Means with the same superscript alphabets in the same column are not significantly different at p<0.05 Legend: V- Vacuum packaged, n- normal packaged/non vacuum packaged I- Irradiated, N- Non-irradiated, F- Fresh, S-Smoked, ‘–‘= none From Table 4.11, significant differences occurred in the pH values of the samples at p<0.05. There was a slight increase in pH on the Day 3 (after irradiation). Significant differences were not observed in the pH values recorded for the non-irradiated vacuum packaged smoked sample and the irradiated normal packaged fresh sample University of Ghana http://ugspace.ug.edu.gh 125 and like wise between the non-irradiated- vacuum packaged fresh sample and the normal- packaged –irradiated smoked sample on the 3rd day after storage. On week 1, there was a slight increase in pH for the irradiated vacuum packaged samples and the non-irradiated vacuum packaged samples (IVS, IVF, NVS and NVF), however there was a wide increase in the pH values for the non- vacuum packaged samples (nNF,nNS,nIS, nIF). On week 2, there was an increase in pH values for the non- irradiated and irradiated non vacuum packaged samples. On week 3 there was a slight increase in pH for the irradiated vacuum packaged fresh samples; IVF and the non –irradiated vacuum packed smoked sample; NVS. There was an increase in pH for the irradiated vacumm packaged smoked sample; IVS (from 4.92 in weeks 2 to 5.15 in week 3) and a slight decrease in the non-irradiated vacuum packaged fresh sample; NVF. On week 4, an increase in pH was observed in the non-irradiated vacuum packaged samples; NVS, NVF and slight increase in the irradiated vacuum packaged fresh sample;IVF, however,no increase was observed in the irradiated vacuum packed sample;IVS. On week 5, a general increase in pH was observed in all thesamples. University of Ghana http://ugspace.ug.edu.gh 126 CHAPTER FIVE 5.0 DISCUSSION ‘Wagashie’ a traditional cheese in West Africa usually coagulated with coagulant from plant origin (Calotropis procera) is considered as a type of soft cheese due to its high moisture content of about 50% (Ashaye et al., 2006). Extracts from the succulent leaves and stems of Calotropis procera is used for the coagulation and has been reported to contain rennet enzymes called calotropin that coagulates milk (Belewu and Aina, 2000). Although coagulants from plant sources like Calotropis procera are available for milk coagulation, their excessive proteolytic nature reduces cheese yield and increases the perception of bitter tastes, making its use more difficult for cheese making (Lo Piero et al., 2002 and Roseiro et al., 2003). Also traditional ‘wagashie’ due to its low salt content has a bland taste and a short shelf life of three days after preparation (Ashaye et al., 2006). Thus this study was carried out to develop the existing product by improving the safety, sensory quality, and shelf life. 5.1 Safety of wagashie The safety of market ‘wagashie’ samples was assessed in the laboratory to ascertain its safety on the market. It was then compared to the safety of the improved ‘wagashie’ prepared under laboratory conditions. Various indicator and pathogenic microorganisms including aerobic mesophiles, Yeast and moulds, coliform bacteria, E. coli, Staphylococcus aureus Bacillus cereus, Salmonella spp, Enterococcus, Enterobacteriaceae were assayed. The results of the microbiological analysis of the University of Ghana http://ugspace.ug.edu.gh 127 market wagashie samples indicated that fresh wagashie had high counts of enteric pathogens than fried wagshie samples. However, the level of contamination in both fresh and fried samples as shown in the results of the microbilogical analysis, were above the acceptable limits in accordance with GSB Standards, (1998) and the Standard method for the examination of dairy products, 2001 (Ledenbach and Marshal, 2009). To determine the means of reducing levels of contamination, fresh market ‘wagashie’ samples were fried, aseptically packaged and tested microbiologically in the laboratory. Frying reduced the counts for indicator and pathogenic organisms but could not reduce them to meet acceptable standards per GSB standard 1998 because of the high counts in aerobic mesophiles (>106 CFU/g). The high microbial load found in the wagashie samples investigated in the present study is in agreement with the work by Elkhider et al., (2011) who reported that cheese samples collected from different producers in rural areas of eastern Sudan indicated that the level of hygiene and production methods, source of raw milk and its handling could be the main factors responsible for high microbial loads which affect the quality of cheese. It was however observed during the market survey that the high microbial counts recorded in the market wagashie samples was because the curds were not packaged but rather exposed on metal trays for retailing. Retailers also hand picked ‘wagashie’ curds into flexible polyethylene bags which served as a point of exposure to post production contamination.Tohibu,et al., (2013), also reported that, ’wagashie’ is not packaged after preparation thus producers hand pick the curds into flexible polyethylene films exposing the product to contamination, reducing the shelf life and making the product unsafe for the consumer. The improved ‘wagashie’ samples produced in this study were fermented to improve the safety of the product due to the University of Ghana http://ugspace.ug.edu.gh 128 antimicrobial activity of lactic acid bacteria. According to Adams and Moss, (1999), lactic acid bacteria especially lactobacillus spp. has been found to produce bacteriocin in addition to lactic acid and hydrogen peroxide during their lactic fermentation. Bacteriocin have been found to inhibit a wide range of bacteria including Gram positive and Gram negative food spoilage and pathogenic bacteria such as E. coli (Ogumbanwo et al., 2003). The milk was fermented for 4 h to a pH of 5.06 with10 ml and 20 ml of cheese and yoghurt cultures. The concentration of cells responsible for the fermentation of milk was 24x1010 CFU/ml and 4x1010 CFU/ml for cheese and yoghurt cultures respectively after plating out. The yoghurt culture was used at a point to replace the cheese culture because of the inactivity of the cheese culture which resulted from the inconsistencies in storage conditions. Fermentation reduced the counts for indicator and pathogenic organisms in the improved wagashie compared to the results of the market ‘wagashie’ samples. Smoking eliminated aerobic mesophiles and reduced the yeast count to 103 CFU/g which was within the acceptable limits for microorganisms in food products in accordance with GSB Standard (1998) and the Standard method for the examination of dairy products, (Ledenbach and Marshal, 2009). The improved ‘wagashie’ therefore had lower counts for pathogenic and indicator organisms than market ‘wagashie’ samples. Packaging the wagashie samples after preparation reduced the risk of post- contamination of the product. According to Poças and Pintado, (2010), packaging is increasingly recognized as an important factor in protecting and controlling the quality and safety of cheese, as well as addressing consumer issues. The pH values recorded for the maket samples were generally high, although pH of the fresh samples was relatively lower than the fried samples. This observation University of Ghana http://ugspace.ug.edu.gh 129 accounted for the high microbial load in the market samples. However the pH of the laboratory prepared samples were low and thus the low microbial counts. 5.2 Improving the quality of wagashie 5.2.1 Affective testing To improve on the sensory quality of ‘wagashie’, the process variables involved in the preparation of traditional ‘wagashie’ and the improved ‘wagashie’ were optimised by a three variable Box Behnken design to compare their acceptability, taste, colour and texture with a 9-point hedonic sensory evaluation. Though milk used for the traditional ‘wagashie’ preparation was not fermented, fermentation was included mainly to improve on the taste and safety of the product. The results of the response surface generated by the Box Behnken design as a result of sensory evaluation showed that the panellists preferred the non fermented wagashie better than the fermented wagashie. This is because, the texture , colour, taste and overall acceptability of the traditional ‘wagashie’ was rated higher when fermentation time was 0 h ,the quantity of salt was 23 g and 150g of plant extract was used in the preparation. The predicted scores for texture, colour, taste and overall acceptability at these levels were 7.18, 7.47, 7.38 and 7.36 respectively which was rated as ‘like moderately’on the hedonic scale. The results of the overlaid contour plots for all the attributes; texture, colour, taste and overall acceptability which were overlaid on one axis of fermentation time and salt also confirmed this finding. The panellists generally, rated the fermented sample lower because the bitter aftertaste imparted by the plant extract combined with the sourness imparted by the fermented milk made the product very bitter. Commercial rennet was used to replace the plant extract to eliminate the bitter taste imparted into the product. From the results of the response surface generated by the University of Ghana http://ugspace.ug.edu.gh 130 Box Behnken design for the improved ‘wagashie’ as a result of sensory evaluation, the texture , colour, taste and overall acceptability was rated higher when the milk was fermented for 4 h, the quantity of salt used was 11g and the concentration of commercial rennet used was 5.35ml. the overlaid contour plots generated for the four attributes overlaid on one axis of fermentation time and salt with a constant rennet concentration of 5.35 ml also confirmed the finding stated earlier .The predicted scores for the attributes texture, colour, taste and overall acceptability for the wagashie samples were 6.67, 7.37, 7.37 and 7.22 respectively. The score can be interpreted as ‘like moderately’ on the 9- point hedonic scale except texture which was rated as ‘like slightly’. These results showed that when commercial rennet was used as the coagulant, only a sour taste was perceived. Since rennet does not influence the taste of the product, this was considered as desirable by the panellists as with the case of European cheese. The results from the two preparations showed that, when wagashie was prepared using the extract of Sodom apple, the panel preferred that the product was not fermented at all. However when wagashie was prepared with commercial rennet, the panel preferred that the milk was fermented for 4 h. Also the plant extract coagulated wagashie and the fermented rennet coagulated wagashie were both rated as ‘like moderately’ by the sensory panel. The results of the pH and TTA values recorded for the ‘wagashie’ samples were relatively high as compared to other fresh cheeses. This was because most of the acidity was lost into the whey during draining and pressing of the curd. This is because a sour taste was observed upon tasting the whey. The loss of acidity and increase in pH during draining and pressing also resulted from the insufficient lactic acid produced by the starter culture in the milk during fermentation. This was University of Ghana http://ugspace.ug.edu.gh 131 because of the low concentration of the starter culture introduced for the milk fermentation. The concentration of starter culture was therefore increased to achieve the required pH at draining. This was in line with the finding by Lucey and Fox, (1992) who stated that, curd is cooked to expel moisture at a temperature which normally adversely affects the starter bacteria. The cheese maker must therefore exert judgement to ensure that the desired acid development in the curd is reached at about the same time as the required moisture content. To compensate for seasonal changes in milk composition it is necessary to vary the percentage of innocula of the starter culture to achieve the required acidity at draining. The second hedonic sensory evaluation was carried out to confirm the consumer preference of the optimised ‘wagashie’ prepared with the extract from the Sodom apple plant and commercial rennet. In this evaluation, the non-fermented plant extract coagulated sample was rated highest for taste followed by the rennet coagulated fermented fried sample. However, significant differences were not observed in their acceptability.The non-fermented rennet coagulated-smoked sample was rated significantly low among the samples for taste; this observation confirmed the results of the discriminatory hedonic sensory evaluation used for the product optimisation where the panellists rated the non-fermented rennet coagulated fresh sample low. The optimisation procedure was able to standardize the traditional production method which eliminated the bitter aftertaste imparted by the plant extract. The non fermented extract coagulated sample was rated significantly lower than the other samples for colour. This can be as a result of the green pigmentation imparted in the wagashie by the extract from the sodom apple plant as in agreement with Chikpah et al., (2014) who stated that, the use of the Calotropis plant as coagulant of soya milk resulted in a green colouration and bitter flavours. The University of Ghana http://ugspace.ug.edu.gh 132 fermented rennet coagulated smoked sample scored highest for texture whiles the fermented rennet and extract coagulated fresh samples scored least.It therefore showed that, smoking improved the texture of the fermented product. The pH values of the non-fermented samples used for the confirmatory affective test were significantly higher than the fermented samples. This was because the inoculum was increased (20 ml per L instead of 20ml per 2 L of yoghurt culture) to improve on the safety and the sensory quality of the product. 5.2.2 Quantitative Descriptive Sensory Analysis During the results of the Quantitative Descriptive Sensory evaluation, 22 lexicons were generated for ‘wagashie’ by the 13 panellists used. They described the sensory characteristics of the fresh, fried and smoked forms of the fermented and non- fermented- rennet coagulated wagashie, laboratory prepared fermented and non fermented plant extract coagulated wagashie, and the fresh and fried market ‘wagashie’. PCA was performed to illustrate graphically the correlation ratings given to the different descriptors and to visualise how the ‘wagashie’ samples were related to each other and their relationship with the generated attributes. It was observed that, all the fermented samples were grouped together whiles the non-fermented samples were also grouped together. Also all fresh samples were grouped together in the same quadrant because they had similar attributes. The smoked samples were grouped in one quadrant and the fried samples were also in the same quadrant except the non- fermented- rennet coagulated and the non-fermented –extract coagulated fried samples which were in the same quadrant with the smoked samples This is because, the rating was influenced mostly by fermentation and the processing methods applied . From the groups formed by the cluster analysis, it was observed University of Ghana http://ugspace.ug.edu.gh 133 that, there were not much difference between the samples prepared with commercial rennet and the samples prepared with the extract from Sodom apple. This shows that the coagulant used had little or no influence on the acceptability of the product. The rennet coagulated preparation can therefore replace the plant extract prepation without any diificulty in consumer acceptance. This will save producers from going through the stress involved in the preparation of the plant extract. It will also reduce the bitter effect it imparts in ‘wagashie’ observed in the market samples which is as a result of the excessive proteolytic nature of the Calotropis plant which reduces cheese yield and increases the perception of bitter tastes, making its use more difficult for cheese making (Lo Piero et al., 2002, Roseiro et al., 2003). The mean intensity rating for each attribute was used to create a sensory descriptive analysis profile and the spider plot showed the descriptive sensory analysis profile more vividly (Stone, 1992). The non –fermented extract and rennet coagulated samples were mainly described to have milky aroma, milky taste, cheesy taste, cheesy aroma and yoghurt aroma. The fermented fresh samples (both plant extract and rennet coagulated) were described to have, yoghurt aroma, spilt milk aroma, fermented cassava dough aroma and whitish colour. The rennet coagulated sample was rated more whitish than the extract coagulated sample, was also described to have a sour taste, soft texture and a smooth texture. This is similar to the findings by Winwood, (1983); Koth and Richter, (1989), who reported that the organoleptic properties which typify soft unripened cheeses include, milky, white colour, soft body, smooth texture, a good spreadability, no signs of syneresis on the cheese surface, no dryness or grittiness and a mild to acidic flavour. The fresh market wagashie samples were described to have a bland taste, bitter aftertaste, sour taste, soft texture and smooth texture.They had the highest intensities University of Ghana http://ugspace.ug.edu.gh 134 for bland and bitterafter taste which were considered as undesirable. The smooth texture of the fresh market sample was because it was being stored in its whey at the retailing point. The sourness resulted from the activity of lactic acid bacteria present in the natural microbial flora in the milk. These organisms can induce fermentation in the product when conditions are favourable causing the sourness. It is possible because fresh milk used for the preparation of traditional wagashie is not pasteurized. The smoked samples were described to have smoked chevon aroma because of the smoke that was imparted into the product as done with goat meat.They were described as beefy aroma which resulted from the milk from the cow origin. They were also described as crumbly texture because smoking reduced the moisture content of the product. The fried samples were described to have spongy texture because the samples soaked some of the oil that was used during frying. The fried samples were described to have fried egg taste, fried ripe plantain aroma, doughnut aroma, fried sweet potato aroma. This is because the panel likened the taste and aroma of the fried ‘wagashie’ to be the taste sensations observed after eating egg that is fried and the aroma sensations observed as a result of frying sweet potato, egg, ripe plantain and doughnut respectively. They were described as brownish in colour because of the processing methods applied (frying and smoking). They were also described to have salty taste because of the salt which is in the form of sodium chloride added. 5.3 Chemical analysis of wagashie samples Cheese is a nutrient dense food which provides fat, high quality proteins, oligopeptides, amino acids, vitamins and minerals. The results of the proximate analysis carried out on the samples showed that, the fermented rennet coagulated University of Ghana http://ugspace.ug.edu.gh 135 smoked sample had the highest FFA and protein content. It also had high fat content with reduced moisture.This is because smoking reduced the moisture content by evaporation which caused, the protein and fat to become more concentrated in the cheese. This showed that the fat content in the milk used in making the product was high, thus the current conserns about the intake of high fat and it relation to coronary heart diseases has caused consumers to be concerned about their fat intake. The smoked sample was therefore considered for the rest of the study.The low moisture content recorded in the smoked ‘wagashie’ may prolong its shelf life whereas the high FFA value can cause spoilage to set in by the action of lipolysis. The fried ‘wagashie’ sample had the highest fat content because the oil used in frying was absorbed into the product. It also recorded the lowest moisture content due to moisture loss during frying. The fermented sample had the lowest ash content because of the breakdown of lactose, calcium phosphate and other minerals in the milk by the starter culture introduced for fermentation, excessive breakdown of calcium phosphate affects the texture of the final product. Lucey and Fox,(1992) stated that, the relationship between the rate of moisture (and lactose) removal versus rate of lactic acid production by the lactic acid bacteria to lower the curd pH has profound effects on the characteristics of the final cheese.This was observed in the fermented fresh sample which was brittle in texture. This effect has been reported by Lucey and Fox, (1992), that the rapid and extensive acid production will remove more calcium and phosphate, to produce a brittle cheese with a lower mineral content.The non-fermented sample however had the highest moisture and ash contents. The wagashie samples had high moisture content which is in agreement with a report by Barabano and Rasmussen (1992) and Dave et al., (2003) who found University of Ghana http://ugspace.ug.edu.gh 136 that, cheese made with fermentation produced chymosin contained higher moisture than cheese made from other coagulants. 5.4 Rheology of wagashie 5.4.1 Colour Determination The results for the rheology of the improved wagashie showed that the colour of the wagashie sample were lighter in colour because of their high L* values. This is because milk which is the major ingredient in ‘wagashie’ is light in colour and therefore the product basically takes the colour of the milk. The lower L* values for the outer colour of the smoked and fried samples is mainly due to browning reactions, which are influenced by the distribution of water and the reaction of reducing sugars and amino acids (Kent and Evers 1994). The samples had positive b* values, which was an indication of slightly yellowish colour. The samples also had positive a* value for exterior part of the product which showed a slightly green colour and negative a* values for the interior part of the product which showed a slightly red colouration. 5.4.2 Texture Profile Analysis (TPA) The results for the TPA analysis also showed that texture of the samples were hard and chewy because of the high values recorded. The hardness resulted from the high protein content in the ‘wagashie’ samples. This has been confirmed by Simoes et al.,( 2013) who stated that lower concentration of protein reduced hardness in cheese and and higher concentrations increase hardness. The protein matrix also produces elasticity and it is the main factor responsible for flexibility and recovery after University of Ghana http://ugspace.ug.edu.gh 137 tension.The values for chewiness in the sample were considered high when compared to those reported for cheese of fete type, which is a type of cottage cheese made from buffalo milk (Kumar et al., 2011). Simoes et al., (2013) stated that chewiness value decreased with decrease in the concentration of added cow milk. 5.5 Shelf life of wagashie In order to extend the shelf life of wagashie, the samples were vacuum packaged and irradiated. Tsiotsias et al., (2002) also confirmed that among the preservation methods to ensure safety of whey cheeses are irradiation combined with vacuum packaging. The shelf life of the wagashie samples was carried out for 5 weeks. It was observed after the study that all the control samples which were the non-irradiated and non- vaccum/ normal packaged samples had undergone spoilage at the end of the second week. The spoilage was characterised mainly by profulous mould growth, off odour, moisture exudation, discolouration and bloated packaging as a result of gases produced by yeast and other spoilage organisms in the samples, and rancidity. Poças and Pintado, (2010) reported that, chemical failures associated with excessive water loss or rind decolourization due to excessive exudation, off-flavours, and defatted tastes originate from oxidation (rancidity) or through microbial metabolism (bitterness, rancidity, acidity), and openings or irregular holes formed by uncontrolled microbial fermentation. According to GSB Standard 1998, the minimum count for yeast and mould should be <104 CFU/g, and 1.0 x 106 CFU/g for total plate count. Also the Standard method for the examination of dairy product 2001 (Lebenach and Marshall, 2009), which is the regulatory standard for indicator organisms for European Union and the United States stated that, the maximum acceptable limits for total plate count should be <106, thus, the samples had exceeded University of Ghana http://ugspace.ug.edu.gh 138 the maximum limits and were discarded. The samples that remained for rest of the storage period were the irradiated vacuum packaged freshand smoked samples and the non-irradiated vacuum packaged fresh and smoked samples (IVS, IVF.NVS and NVF). The vacuum packaged non-irradiated samples were able to stay on the shelf for 2 weeks. According to Kreft, (2008), vacuum packaging increased the shelf life of Gouda cheese to 10 weeks. It also increased the shelf life of Parmigiano-Reggiano cheese to 6 months (Severini et al., 1998). Their spoilage was however characterised by bloated packaging as, result of gas production by yeast and other microorganisms and few mould growth. The irradiated vacuum packaged fresh sample also deteriorated on the 3rd week and was characterised by bloated packaging. The bloating of the packaging material was as a result of the anaerobic condition and the activity of yeast created in the packaging material. This condition must therefore be controlled. Poças and Pintado (2010), stated that anaerobic condition inside packaged cheese must be controlled because Clostridium spp. may grow and produce gas due to anaerobic fermentation of lactate. They also stated that, active packaging systems with antimicrobials, as well as packages with adequate oxygen permeability, may solve some of these frequent microbial failures. Though the smoked sample had microbial count which did not exceed the maximum acceptable limits of spoilage or indicator organisms in food with regard to GSA Standard, (1998) and the Standards for the examination of dairy products as mentioned earlier,the product had developed off- odour but no visual defects were observed. This was also observed by Bongirwar and Kumta (1967) who reported that Cheddar cheese developed off-flavours when irradiated at 0.5 kGy. However, according to Chincholle, (1991) Camembert cheese suffered no off-flavour University of Ghana http://ugspace.ug.edu.gh 139 development up to a dose of 3 kGy. No mould growth was observed in the irradiated vacuum packaged samples during the 5 weeks storage period. Jones and Jelen, (1988) observed also that when Turkish Kashar cheese was exposed to 1.2 kGy, a mould free shelf life was obtained for 12 to 15 days of storage. Though ‘wagashie’ is known to be a perishable product with a shelf life 2 to 3 days (Ashaye et al., 2006), its shelf life was extended to a maximum of 3 weeks when preserved by vacuum packaging and 4 kGy of gamma radiation from a Cobalt 60 source. 5.5.1 pH of wagashie samples during the 5 weeks storage period Fresh cheeses have high pH when acid formed during milk fermentation is lost during whey drainage and pressing; the high pH creates a suitable condition in the cheese for the growth of spoilage organisms like Bacillus cereus, yeast and moulds etc. The pH obtained after milk fermentation for ‘wagashie’ preparation was 5.06 and decreased to 4.9 after preparation. The results of the pH of the samples during the 5 weeks storage, showed that the pH values of the samples increased as the microbial count increased. This observation is contrary to the findings by Nobile et al., (2009) who stated that the pH of Ricotta cheese packaged under modified atmosphere increased as the microbial count reduced upon storage. This may be influenced by the type of organisms responsible for the spoilage and the chemical changes that occur in the product.The high pH recorded was as a result of proteolysis which is the gradual breakdown of proteins to form amino acids and peptides which increase alkalinity. This is influenced by the enzymes in the coagulant and the activity of the microorganisms present in the cheese (Bylund, 1995). University of Ghana http://ugspace.ug.edu.gh 140 CHAPTER SIX 6.0 CONCLUSIONS AND RECOMMENDATIONS 6.1 CONCLUSIONS Wagashie sold on the market pose a health hazard to consumers since they generally contain indicator and pathogenic microorganisms which exceed the limits prescribed by GSB and other regulatory bodies. However, the hazards are reduced when the wagashie sold is in a fried form. The traditional process of wagashie production can be improved or industrialised by replacing the use of the plant extract with rennet and other commercial cheese coagulants. This eliminates the bitterness imparted to wagashie and makes it easier to standardise the process. By fermenting the pasteurised milk used to produce wagashie, both the sensory quality and the safety of the product were improved due to the antimicrobial properties of the lactic acid bacteria used for the fermentation and a more pronounced taste of the acidified product. The combined effect of fermentation, smoking and vacuum packaging extended the shelf life of wagashie to 2 week without refrigeration. However when combined with irradiation, the shelf life extended to 3 weeks. . University of Ghana http://ugspace.ug.edu.gh 141 6.2 RECOMMENDATIONS The selection of the optimum packaging system must be considered which should include the fact that cheese is a complex dynamic matrix in which several microbial, physical, and biochemical changes occur during storage. This can be done by incorporating Modified Atmosphere packaging whereby CO2 can be introduced into the vacuum packaging material and the application of low doses of radiation (1kGy to 3 kGy) to further extend the shelf life of ‘wagashie’ whiles maintaining its quality. The chemical analysis as well as sensory evaluation of ‘wagashie’ during storage should be carried out along with microbiological tests to observe their influence on each other. The texture of the fermented ‘wagashie’ sample should be improved by maturing it under controlled conditions to further improve on the product. Studies to improve on the yield of the rennet coagulated ‘wagashie’ should be carried out. University of Ghana http://ugspace.ug.edu.gh 142 REFERENCES Adams M. R., Moss M. O., Food Microbiology. Royal Society of Chemistry, Cambridge. 1999; pp 255 – 74. Adetunji V.O and O.O Babalobi (2011) A comparative Assessment of the Nutritional contents of ‘Wara’ A West African soft cheese using calotropis procera and Cymbopogon citrates as coagulants. African Journal of food, Agriculture, Nutrition and Development. Vol. 41 pp.331–336 Adetunji, V. O., Alonge, D. O, Singh, R. K and Chen J. (2007). Production of wara, West African Soft cheese using lemon juice as a coagulant. Journal of food science and technology, 41:331-336. Alalade OA and JA Adeneye (2006), The effect of storage period on the chemical composition and coliform microflora of wara cheese.International Journal of Dairy Science.Vol. 1(2):pp. 126-130. Annadurai G., R.Y. Sheeja, (1998) “Use of Box-Behnken design of experiments for the adsorption of verofix red using biopolymer”, Bioprocess Engineering, Vol. 18, pp. 463-466 Anon. (2003) Foodborne disease in Australia: incidence, notifications and outbreaks. Annual report of the Oz FoodNet network, 2002. Commun Dis Intell. 27(2):209-243. . Animal Production Directorate (2003). Ghana’s country report on animal genetic resources. MOFA, Ghana. Appiah, F. (2000). Studies on traditional cheese (Woagashie) processing in the Ashanti Region of Ghana. MSc thesis submitted to Food Science Department of KNUST. Ashaye, O.A., Taiwo, O.O., and Adegoke, G.O. (2006). Effect of local preservative (Aframomum danielli) on the chemical and sensory properties of stored warakanshi. African Journal of Agricultural Research Vol. 1 (1). Pp. 010-016. Association of Official Analytical Chemists (AOAC). (2006). Official Methods of the Association of Official Analytical Chemists, Association of Official Analytical Chemists, Inc., Virginia, USA. Pp 69-84. Association of Official Analytical AOAC, (1995d). Determination of ash in cheese (method 935.42). Official Methods of Analysis. Vol. 1. Arlington: AOAC Publishers, pp 59. ASTM Sensory Testing Methods (1996). 2nd Edition Copyright (pp. 58–70). University of Ghana http://ugspace.ug.edu.gh 143 Aworh OC (2008). The Role of Traditional Food Processing Technologies In National Development: theWest African Experience. Chapter 3 from Using Food Science and Technology to Improve Nutrition and Promote National Development, Robertson, G.L. & Lupien, J.R. (Eds), © International Union of Food Science & Technology Aworh, O.C. and Muller, H.G. (1987), “Cheese making properties of vegetable rennet from Sodom apple (Calotropis procera)”, Food Chemistry journal, Vol. 26 pp. 71-79 Aworh, O.C., (2008) The Role of Traditional Food Processing Technologies In National Development: theWest African Experience. Chapter 3 from Using Food Chapter 3 from Science and Technology to Improve Nutrition and Promote National Development, Robertson, G.L. & Lupien, J.R. (Eds), © International Union of Food Science & Technology Aworh, O. C.; Egounlety, M. (1985) Preservation of West African soft cheese by chemical treatment. J. DairyRes, 52, 189-195. Banks J.M., Banks W., Muir D.D., Wilson A.G., (1981). Cheese yield: composition does matter. Dairy Ind. Int. Vol. 46 (5), pp. 15, 17, 19, 21-22. Balciunas E. M., F. Martinez A. C., Todorov S. D., G. de Melo Franco B. D., Converti A. and Oliveira R. P. de Souza (2012), “Novel Biotechnological Applications of Bacteriocins: A Review,” Food Control, Vol. 32, No. 1, 2013, pp. 134-142. Barbano, D.M., Rasmussen, R.R. and Lynch, J.M. (1991) Influence of somatic cell count and milk age on cheese yield. Journal of Dairy Science, 74, 369–388. Barabano D.M., Rasmussen R.R., (1992). Cheese yield performance of fermentation produced chymosin and other milk coagulants. Journal of Dairy Science Vol. 75, pp. 1-12. Basu A, Chaudhury A .K. N. (1991). Preliminary studies on the anti-inflammatory and analgesic activities of Calotropis procera root extract. Journal of Ethnopharmacology, Vol.31: pp. 319-324, Bernnand, C.P. (2006): Food irradiation. Frequently asked questions. University of Reading UK The Food Safety Advisory Centre London. Beresford, T.P. and Williams, A. (2004) The microbiology of cheese ripening. Cheese: Chemistry, Physics and Microbiology, Volume 1: General Aspects (eds P.F. Fox, P.L.H. McSweeney, T.M. Cogan and T.P. Guinee), 3rd Ed, pp. 287–317. University of Ghana http://ugspace.ug.edu.gh 144 Beukes, E.M., Bester, B.H. and Mostert, J.F. (2001) The microbiology of South African traditional fermented milks. International Journal Food Microbiology Vol.63, pp. 189–197. Bongirwar DR. and Kumta US. (1967). Preservation of cheese with combined use of gamma-rays and sorbic acid, International Journal of Applied Radio Isotopes Vol. 18: pp. 133. Borch, E., Kant-Muermans, M.L., and Blixt, Y. (1996). Bacterial spoilage of meat and cured meat products, International Journal of Food Microbiology 33, pp. 103– 120. Box, G.E.P., Hunter, W.G. and Hunter, J.S., (1978). Statistics for Experimenters. (John Wiley & Sons, Inc), pp. 510–539 Box G.E.P., Behnken D.W., (1960) “Some New Three Level Designs for the Study of Quantitative Variables”, Technometrics, Vol. 2, pp. 455-476 Booz-Allen and Hamilton, Inc.. (1982). New product management for the 1980s. New York. pp. 27 Bramley, A.J. & McKinnon, C.H. (1990) The microbiology of raw milk. Dairy Microbiology (ed. R.K. Robinson), Vol. 1, pp. 163–208, Bylund D (1995) Dairy Processing Handbook. Tetra Pak Processing Systems AB, Lund, Sweden. Caplice, E. and Fitgerald, G.F. (1999) Food fermentations: role of micro-organisms in food production and preservation. International Journal Food Microbiology Vol. 50,pp. 131–149. Carreira, A., Dillinger, K., Eliskases-Lechner, F., Loureiro, V., Ginzinger, W. & Rohm, H. (2002)Influence of selected factors on browning of Camembert cheese. Journal of Dairy Research, Vol.69, pp. 281–292. Chikpah S. K., Teye M., Annor J. A. F. and Teye G.A. (2014). Potentials of Sodom apple (Calotropis procera) extract as a coagulant to substitute alum in soy cheese production in Ghana. Elixir International Journal of Food Science, Vol. 79 pp. 30166-30170 Chincholle RC. (1991). Action of the ionization treatment on the soft cheeses made from unpasteurized milk, CR. Acad. Agric. Fr., 77: pp. 26. Codex Alimentarius cheese standard 283-(1978). http://www.codexalimentarius.net. Accessed March 2015 University of Ghana http://ugspace.ug.edu.gh 145 Coker Christina, Craig Honoré, Keith Johnston and Lawrie Creamer (2005), Food Science Section and Cheese and Milk fat Technology Section, New Zealand Dairy Research Institute. Cooke, R. D.; Twiddy, D. R.; Reilly, P. J.(1987) Lactic-acid fermentation as a low- cost means of food preservation in tropical countries. Microbiology Reviews, 46, 369-379. Cooper, R.D., and Kleinschmidt, E.J. (1986). An investigation into the new product process: steps, deficiencies, and impact. Journal of Product Innovation Management. 3(2): pp. 71-85. Commission Internationale de l’ ´ Eclairage (1986) Colorimetry, 2nd edn, Bulletin No. 15.2, Commission Internationale de l’ ´ Eclairage (CIE), Central Bureau Kegelgasse 27A-1030, Vienna. CTA 2014 http://knowledge.cta.int/ Peter Fellows Dave R.I., Sharma P., McMahon D.J.,( 2003). Melt and rheological properties of Mozzarella cheese as affected by starter culture and coagulating enzymes. Lait 83, pp. 61-77. Delahunty, C.M. & Drake, M.A. (2004). Sensory character of cheese and its evaluation. Cheese: Chemistry, Physics and Microbiology, Volume 1: pp. 455–487. De Man, J.C., Rogosa, M. and Sharpe, M.E. (1960) A medium for the cultivation of lactobacilli. Journal of Appllied BacteriologyVol 23, pp130– 135. Dias B. and Weimer B. (1998), “Conversion of Methionine to Thiols by Lactococci, Lactobacilli, and Brevibacteria,” Applied and Environmental Microbiology, Vol. 64, pp. 3320-3326. Dimitreli, G. and Thomareis, A.S. (2007) Texture evaluation of block-type processed cheese as a function of chemical composition and in relation to its apparent viscosity. Journal of Food Engineering, Vol. 79, pp. 1364–1373. Drake, M., A.(2007), Sensory Analysis of Dairy Foods. Journal of Dairy Science Vol. 90: pp. 4925-4937 Drake, M.A., Mcingvale, S.C., Gerard, P.D., Cadwallader, K.R. and Civille, G.V. (2001). Development of a descriptive language for cheddar cheese Journal of food science Vol. 66, pp.1422–1427 University of Ghana http://ugspace.ug.edu.gh 146 Drake, M.A., Gerard, P.D., Wright, S., Cadwallader, K.R. and Civille, G.V. (2002). Cross validation of a sensory language for cheddar cheese. Journal of Sensory studies Vol.17, pp. 215–227. Dubost NJ, Shewfelt RL, Eitenmiller RR. (2003). Consumer acceptability, sensory and instrumental analysis of peanut soy spreads. Journal of Food Quality Vol. 26: pp. 27–42 Dufoss´e, L., Galaup, P., Carlet, E., Flamin, C. & Valla, A. (2005) Spectrocolorimetry in the CIE L*a*b* colour space as useful tool for monitoring the ripening process and the quality of PDO red-smear soft cheeses. Food Research International, Vol. 38, pp. 919–924. Edwards, J. and Kosikowski, F.V. (1983), “Bitter compounds from Cheddar cheese”, Journal of Dairy Science, Vol. 66, pp. 727-734. Elkhider, I.A.E., El Zubeir I.E.M., Basheir A.A., Fadlelmoula A.A.,( 2011), Composition and hygienic quality of Sudanese white cheese produced by small scale in rural area of eastern Sudan. Annuals of Food Science and Technology, Vol. 12(2) pp.186- 192. Elkhider IAE, Ibtisam ElYEl Z, Amani Abd ElMB, Abdelaziz AF (2011). Composition and hygienic quality of sudanese white cheese produced by small scale in rural area of eastern sudan. Ann. Food Sci. Technol. 12:186-192 Everard, C.D., O’Donnell, C.P., O’Callaghan, D.J., Sheehan, E.M., Delahunty, C.M., O’Kennedy, B.T. &Howard,V. (2007c) Prediction of sensory textural properties from rheological analysis for process cheeses varying in emulsifying salt, protein and moisture contents. Journal of the Science of Food and Agriculture, Vol. 87, pp. 641– 650. FAO Food Agriculture Organization (1998): Study Group on High Dose Irradiation. “Food safety.” Weekly Epidemiological Record 16 73(3) pp.9-11. Farkye, N.Y and Vedamuthu, E.R.,( 2002). Microbiology of soft cheeses. In: Robinson, R.K., 2002. Dairy Microbiology Handbook (3rd Edition): Microbiology of soft cheeses: pp 479-480. FDA, Food and Drug Administration (1997): Irradiation in the production, processing and handling of food. 62 Federal Register 232 pp. 64107- 64121 Fellows, P. (2014) Food Processing Technology, Woodhead Publishing, Cambridge, UK. University of Ghana http://ugspace.ug.edu.gh 147 Fernández L., Bhowmik T. and Steele J. (1994), “Characterization of the Lactobacillus helveticus CNRZ32 pepC Gene,” Applied and Environmental Microbiology, Vol. 60, pp. 333-336. Fenelon, M.A. & Guinee, T.P. (2000) Primary proteolysis and textural changes during ripening in Cheddar cheeses manufactured to different fat contents. International Dairy Journal, Vol. 10, pp. 151–158. Foltmann, B.F. (1992) Chymosin: a short review on foetal and neonatal gastric proteases. Scandinavian Journal of Clinical and Laboratory Investigations,Vol. 52, pp. 65–79. Food and Drugs Administration (FDA) (2005) Warns About Soft Cheese. The US Food and drugs Administration. http://www.consumeraffairs.com/news04/2005/fda_cheese.html). Retrieved October 15, 2005 Chris Mercer 2005. Fox P. F. (1993) Cheese: Chemistry, Physics and Microbiology 2nd edn, Vol. 1 and 2. Chapman and Hall,London. Fox P. F., McSweeney P. L. H., Cogan T. M. and Guinee T. P. , (2004) “Cheese: Chemistry, Physics and Microbiology,” Elsevier, 2004. Fox P. F., T. P. Guinee, T. M Cogan and P. L. H. Mc Sweeney, (2000) Cheese rheology and texture, in: Fundamentals of cheese science (Fox P.F. et al., editor: Aspen Publisher Inc,). Fox, P.F. & Brodkorb, A. (2008) The casein micelle: historical aspects, current concepts and significance. International Dairy Journal, Vol. 18, pp. 677–684. . Frank, J. F. (2001). Milk and dairy products. In Doyle, M. P., Beuchat, L. R., & Montville, T. J.(Eds.), Food microbiology: fundamentals and frontiers (2nd Ed.), pp. 111–126. Washington, DC: American Society of Microbiology. Gadaga, T.H., Mutukumira, A.N. and Narvhus, J.A. (2000) Enumeration and identification of yeasts isolated from Zimbabwean Traditional Fermented milk. International Dairy Journal Vol. 10, pp. 459–466. Gains, N.,Thompson, D. M. H.(1990). Contextual evaluation of canned lagers using repertory grid method.International Journal of Food Science and Technology, Vol. 25, pp. 699-705 University of Ghana http://ugspace.ug.edu.gh 148 Garg S. K., Johri, B.N., (1994). Rennet: current trends and future research. Food Review International Vol.10: pp. 313-355. García V. & Rovira S. & Teruel R. & K. Boutoial & J. Rodríguez & I. Roa & M. B. López, 2012.. Effect of vegetable coagulant, microbial coagulant and calf rennet on physicochemical, proteolysis, sensory and texture profiles of fresh goats cheese. Dairy Science and Technology Vol. 92: pp. 691–707 Gatti M., L. C., Rossetti L., Mucchetti G. and Neviani E., (2003) “Biodiversity in Lactobacillus helveticus Strains Present in Natural Whey Starter Used for Parmigiano Reggiano Cheese,” Journal of Applied Microbiology, Vol. 95, pp. 463- 470. Ghana Standards Board Microbiological Standards for Food Samples (GSB, 1998). Giovanni, M. (1983). Response surface methodology and product optimisation. Food Technology, Vol.37 (11), pp. 41-45 Giraffa G., Chanishvili N. and Widyastuti Y. (2010), “Importance of Lactobacilli in Food and Feed Biotechnology,” Research in Microbiology, Vol. 161, No. 6, pp. 480- 487. Government of Ghana (GOG)/ Food and Agricultural Organisation (FAO) (2002): Training programme for the small-scale dairy sector. Government of Ghana/Food and Agricultural Organisation Technical Paper. Gonz´as alez-Tom, L., Costell, E. (2006). Journal of Sensory Studies. Vol. 2 pp.22– 33. Griffiths M. W. and Tellez A. M. (2013), “Lactobacillus helveticus: The Proteolytic System,” Frontiers in Microbiology,Vol. 4, pp. 1-9. Guinee, T.P., Pudja, P.D., Reville, W.J., Harrington, D., Mulholland, E.O., Cotter, M. & Cogan, T.M.(1995) Composition, microstructure and maturation of semi-hard cheeses from high protein ultrafilteredmilk retentates with different levels of denatured whey protein. International Dairy Journal, 5, 543–568. Guinee, T.P. (2003) Role of protein in cheese and cheese products. Advanced Dairy Chemistry, Vol. 1: pp. 1083–1174. Guinee, T.M.,and Wilkinson, M.G. (1992). Rennet coagulation and coagulants in cheese manufacture. Journal of the Society of Dairy Technology, 45, 94–104. Gupta, C.B. and Eskin, N.A. (1997), “Potential use of vegetable rennet in the production of cheese”, Food Technology, Vol. 5 pp. 62-66. Hannon J. A., Kilcawley K. N., Wilkinson M. G., Delahunty C. M. and Beresford T. P. (2007), “Flavor Precursor Development in Cheddar Cheese Due to Lactococcal University of Ghana http://ugspace.ug.edu.gh 149 Starters and the Presence and Lysis of Lactobacillus helveticus,” International Dairy Journal, Vol. 17, pp. 316- 327. Hamann, J. (2003) Definition of the physiological cell count threshold based on changes in milk composition. Mastitis Newsletter 25, Document No. 381, pp. 9–12, International Dairy Federation, Brussels. Harboe, M.K. (1992a) Chymogen, chymosin rennet manufactured by fermentation of Aspergillus niger. Fermentation-Produced Enzymes and Accelerated Ripening in Cheesemaking, Document No. 269, pp. 3–7, International Dairy Federation, Brussels. Helinck S., Bars D. L., Moreau D. and Yvon M. (2004), “Ability of Thermophilic Lactic Acid Bacteria to Produce Aroma Compounds from Amino Acids,” Applied and Environmental Microbiology, Vol. 70, pp. 3855-3681. Hinkelmann, K., and Kempthorne, O. (2007). Design and analysis of experiments: Introduction to experimental design (2nd ed.) New Jersey: John Wiley & Sons, Inc. http://www.uoguelph.ca/foodscience/cheesemaking-technology/section-b- analytical/process-and-quality-control-proceudures/ph (http://www.lactose.com/basic/physiological_properties.html ICMSF "International Committee on Microbiological Specification for foods" (1996) Microorganisms in Foods: Their significance and method of enumeration. 2th Ed. University of Toronto Press, Toronto, Canada, p. 564-790. IDF (1981) Catalogue of Cheeses, Document No. 141, pp. 3–40, International Dairy Federation, Brussels. Ilori, M.O., J.S. Oke & S.A. Sanni (2000), Management of new product development in selected food companies in Nigeria. Technovation, Vol. 20, pp. 333-342. International Livestock Research Institute, (ILRI) (2006). Ethiopia, Addis Ababa, pp. 611-892. E-mail: ILRI-Ethiopia@cgiar.org.Website: www.cgiar.org/ilri/ International Organization for Standardisation, ISO 7959: 1987 (E), Enumeration of yeasts and moulds. International Organization for Standardisation, ISO 6579: 2002 pt 1 and 2, Detection of Salmonella Irvine, F. R. (1961): Woody plants in Ghana. Oxford University Press, London. pp.48-50. University of Ghana http://ugspace.ug.edu.gh 150 Issanchou S, Schlich P, (1997). Sensory analysis: methodological aspects relevant to the study of cheese. Le Lait, 77 (1), pp.5-12. Jacob M, Jaros D, Rohm H (2011) Recent advances in milk clotting activity. International Journal of Dairy Technology Vol. 64: pp.14–33. Jansen, H.G.P. (1990). Consumption of dairy products in northern Nigeria in Dairy Marketing in Sub- Saharan Africa. Proceedings of a Symposium held at ILCA, Addis Ababa, Ethiopia, 26-30 November 1990. International Livestock Centre for Africa (ILCA), Addis Ababa, Ethiopia Jayarao, B.M. and Henning, D.R. (2001) Prevalence of foodborne pathogens in bulk tank milk. Journal-of- Dairy-Science 84(10):2157-2162. Jespersen, L. (2003) Occurrence and taxonomic characteristics of strains of Saccharomyces cerevisiae predominant in African indigenous fermented food and beverages. FEMS Yeast Res 3, pp 191–200. Jelinski, T., Du, C.J., Sun, D.W. & Fornal, J. (2007) Inspection of the distribution and amount of ingredients in pasteurised cheese by computer vision. Journal of Food Engineering, Vol. 83, pp. 3–9. Jones T. H., Jelen P. (1988). Low dose gamma irradiation of camembert, cottage cheese and cottage cheese whey, Milchwissenschaft, 43:233 Joseph, J.K., Akinyosoye, F.A. (1997). Comparative studies on Red Sorghum Extract and Other chemicals as preservatives for West African soft cheese. International. Dairy Journal pp. 193-198. Karbo, N., Bruce, J. and Okantah, S.A. (1998a). A survey on peri-urban dairy in northern Ghana. Paper presented at the 1st Biennial National Agricultural Research Project (NARP) conference held at Accra, Ghana, 16–20 November, 1998. Keenan, T.W. and Maher, I.H. (2006) Intracellular origin of milk fat globules and the nature of the milk fat globule membrane. Advanced Dairy Chemistry, Vol. 2:), pp. 137–171, Kent N. L., Evers AD. (1994). Bread made with gluten substitutes. Technology of Cereals. Oxford, U.K.: Pergamon Press. pp. 215. Kèkè M, Yèhouénou B, Dahouénon E, Dossou J, Sohounhloué DCK (2008). Contribution à l’amélioration de la technologie de fabrication et de conservation du fromage peulh waragashi par injection de Lactobacillus plantarum. Ann. Sci. Agron. Bénin. Vol. 10(1): pp.73-86. Kew, F. (1985): The useful plants of West Tropical Africa, Vol. (1). Families A – D Edition 2 (Ed Burkill, H. M.). Royal Botanical Gardens. Pp.219-222 University of Ghana http://ugspace.ug.edu.gh 151 Khirstova, P. and Tissot, M. (1995). Soda – Anthroquinone pulping of Hibiscus Sabdariffa (Karkadeh) and Calotropis procera from Sudan. Bioresource Technology Vol. 53: pp.67-72. Khuri AI, Cornell JA, (1978). Response surfaces: Design and analysis. New York: John Wiley & Sons Inc. Kiran B., Kaushik A., Kaushik C.P., 2007 “Response surface methodological approach for optimizing removal of Cr (VI) from aqueous solution using immobilized cyanobacterium”, Chemical Engineering Journal, Vol. 126, pp. 147– 153. Kongo J. M. (2013), “Lactic Acid Bacteria as Starter-Cultures for Cheese Processing: Past, Present and Future Developments,”Chapter 1. Korkeala, H., Lindroth, S., Suihko, M., Kuhmonen, A., and Penttilä, P-L. (1985). Microbiological and sensory quality changes in blood pancakes and cooked ring sausage during storage, International Journal of Food Microbiology Vol. 2, pp. 279– 292. Konteles, S.; Sinanoglou VJ; Batrinou A, and Sflomos K. (2009). Effects of gamma- irradiation on Listeria monocytogenes population, colour, texture and sensory properties of Feta cheese during cold storage. Food Microbiology Vol. 26 (2) pp.157- 165. Koth, A.B. and Richter, R.L., 1989. A limited test market evaluation of Quarg sales potential. Cultured Dairy ProductsVol. 24 (3), pp.4-8. Kousta M., Mataragas M., Skandamis P. and Dros-inos E. H., 2010 “Prevalence and Sources of Cheese Contamination with Pathogens at Farm and Processing Levels,” Food Control, Vol. 21, 2010, pp. 805-815. Kreft F. (2008). Packaging of Cheese, Agrotechnology & Food Sciences Group of the Wageningen University and Research Centre. http://www.vers.nl/Projects-Cheese Packaging. Kumar S., Kanawjia S. K., and Khatkar S., (2011) Effect of rate of addition of starter culture on textural characteristics of buffalo milk Feta type cheese during ripening. Journal of Food Science and Technology (on line), pp 1-5. Laemmli UK (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage T4.Nat 227: pp. 680–685 University of Ghana http://ugspace.ug.edu.gh 152 Lartey, B.L., (1975) Some Technological Transfromation and Research needs in Ghana’s Indegenous food industries. UNIDO Joint Consultation on Promotion of Industrial Research and Services in Africa. Lagos 15-22 September 1975, UNIDO. Lawless, H.T and Heyman, H. (1998). Sensory Evaluation of Food: Principles and Practices. Chapman & Hall /International Pub. New York, pp. 819. Law A.Barry and Tamime A.Y. (2010). Technology of cheese making second edition. John Wiley & Sons Ltd, The Atrium, Southern Gate, Chichester, West Sussex, PO19 8SQ, United Kingdom. Ledenbach L. H. and Marshall R. T., (2009). Microbiological Spoilage of Dairy Products. Compendium of the Microbiological Spoilage of Foods and Beverages. Food Microbiology and Food Safety, ISBN: 978-1-4419-0825 pp. 41-67 Lee, P.R. (1994): Irradiation to prevent food borne illness. Journal of Animal Research and Practice Vol. 272 (4) pp.28-91. Leuschner R. and Boughtflower M. (2002): Laboratory scale preparation of soft cheese artificially contaminated with low levels of Escherichia coli O157, Listeria monocytogenes, and Salmonella enterica Serovars Typhimurium, Enteritidis. Journal of Food Protection Vol. 65 (3) pp. 508-514. Loaharanu, P. (2005): Irradiation as a cold pasteurization process of food. Journal of Veterinary and Parasitology Vol. 64 (2) pp.171-182. Lomer, M.C., Parkes, G.C. & Sanderson, J.D. (2008) Review article: lactose intolerance in clinical practice – myths and realities. Alimentary Pharmacology and Therapeutics, Vol 27(2) pp. 93–103. Lo Piero AR, Puglisi I, Petrone G (2002) Characterization of “Lettucine”, a serine- like protease from Lactuca sativa leaves, as a novel enzyme for milk clotting. Journal of Agricutural Food Chemistry Vol. 50: pp.2439–2443 Lucey, J.A. & Fox, P.F. (1992) Rennet coagulation properties of late-lactation milk: effect of pH adjustment, addition of CaCl2, variation in rennet level and blending with mid lactation milk. Irish Journal of Agricultural and Food Research, Vol. 31, pp. 173–184. Lucey, J.A. (2004) Formation, structural properties and rheology of acid-coagulated milk gels. Cheese: Chemistry, Physics and Microbiology, Vol. 1: General Aspects MacDougall, D.B. (2001) Principles of colour measurement for food. Instrumentation and Sensors for the Food Industry (eds E. Kress-Rogers & C.J.B. Brimelow), pp. 63–84, Woodhead Publishing, Cambridge, UK. University of Ghana http://ugspace.ug.edu.gh 153 Macedo, I. Q., Faro, C. J., & Pires, E. V. (1993). Specificity and kinetics of the milk- clotting enzyme from cardoon (Cynara cardunculus L.) toward bovine j-casein. Journal of Agricultural and Food Chemistry, 41, 1537–1540. Mahami, T., Ocloo, F. C. K., Odonkor, S. T., Owulah, C. and Coffie, S. A. (2012). Preliminary study on the influence of moringa seed extracts supplementation on the yield and quality of cottage cheese, Vol. 2(1): pp. 4-8. Martínez-Ruiz N. R., López-Díaz J. A. and Rodrí- guez-Torres K. I., (2010) “Physicochemical Characteristics and Sensory Properties of Asadero Cheese manufactured with Plant Coagulant from Solanum elaeagnifolium. Food and Nutrition Sciences, Vol. 4, pp. 75-81 Marilley, L. and Casey M.G., (2004) Flavours of cheese products: metabolic pathways. Analytical tools and identification of producing strains. International Journal of Food Microbiology Vol. 90: pp. 139-159. Maragkoudakis P. A.,. Miaris C, Rojez P., Manalis N, Magkanari F., Kalantzopoulos G. and Tsakalidou E., (2006 )“Production of Traditional Greek Yoghurt Using Lactobacillus Strains with Probiotic Potential as Starter Adjuncts,” International Dairy Journal, Vol. 16, No. 1, , pp. 52- 60. Mason, J.O. (1993): Food irradiation. Public Health report Vol. 108, No. 3, pp. 402- 410. McCraig, T.N. (2002) Extending the use of visible/near-infrared reflectance spectrophotometers to measure colour of food and agricultural products. Food Research International Vol. 35, pp. 731–736. McMahon, D.J. & Oommen, B.S. (2008) Supramolecular structure of the casein micelle. Journal of Dairy Science, Vol. 91, pp.1709–1721. McSweeney P. L. H. and Sousa M. J.,( 2000 )“Biochemical Pathways for the Production of Flavour Compounds in Cheese during Ripening: A Review,” Lait, Vol. 80, No. 3, pp. 293-324. McSweeney, P.L.H. (2004) Overview of cheese ripening: introduction and overview. Cheese: Chemistry, Physics and Microbiology, Volume 1: General Aspects (eds P.F. Fox, P.L.H. McSweeney, T.M. Cogan and T.P. Guinee), 3rd edn, pp. 348–360. University of Ghana http://ugspace.ug.edu.gh 154 Mehran, N.T.; Tawfeak, Y. and Hewedy, M. (2005): Incidence of pathogens in kareash cheese. Egyptian Journal of Dairy Science Vol. 26 (1) pp. 295-300. Montgomery D.C. (2005), Design and Analysis of Experiments, 6th ed., John Wiley & Sons. Inc, ISBN: 978-0-471-48735-7. Muir, D.D. & Hunter, E.A. (1991/1992) Sensory evaluation of Cheddar cheese: order of tasting and carry-over effects. Food Quality and Preference, Vol. 3, pp. 141–145. Murray, J.M., Delahunty, C.M. & Baxter, I.A (2001). Descriptive sensory analysis: past, present and future. Food Reseach International Vol.34, pp. 461-471. Muehlenkamp-Ulate, M.R., and Warthesen, J.J. (1999). Evaluation of several non- starter lactobacilli for their influence on Cheddar cheese slurry. Journal of Dairy Science Vol. 82, pp. 1370–1378. Muthukumar, M., Mohan, D. and Rajendran, M., (2003) Optimization of mix proportions of mineral aggregates using Box-Behnken design of experiments. Cement and Concrete Composites, Vol. 25: pp. 751–758. Munoz, A.M., and Civille, G.V. (1998). Universal, product and attribute specific scaling and the development of common lexicons in descriptive analysis. Journal of Sensory Studies, Vol.13, pp. 57-75 Meyers, R.H., & Montgomery, D.C (2002). Response surface methodology: Process and product optimisation using designed experiments 92nd Ed.). USA: John Wiley & Sons, Inc. Nobile Del, M. A., Conte, A., Incoronato, A. L. and Panza, O., (2009). Modified atmosphere packaging to improve the microbial stability of Ricotta. African Journal of Microbiology Research Vol. 3 (4) pp. 137-142 Nordic Committee of Food Analysis, Enumeration of Aerobic Mesophiles NMKL No. 86, (2006) Nordic Committee of Food Analysis, Enumeration of Total Coliform NMKL No. 44, (2004) Nordic Committee of Food Analysis, Enumeration of E. coli, NMKL No. 125, (2005). Nordic Committee of Food Analysis, Enumeration of Staphylococcus aureus, NMKL No. 66 4th Ed., (2009). Nordic Committee of Food Analysis, Enumeration of Bacillus cereus, NMKL No. 67, (2010). University of Ghana http://ugspace.ug.edu.gh 155 Nordic Committee of Food Analysis, Enumeration of Enterococcus, NMKL No. 68 5th, (2011). Nordic Committee of Food Analysis, Enumeration of Enterobacteriaceae, NMKL No. 114, (2004). Nsiah-Ababio, R. (1998). Potential contribution of milk sales to farmers’ income. BSc dissertation, Faculty of Agriculture, Kwame Nkrumah University of Science and Technology, Kumasi, Ghana Obodai, M. And Dodd C.E.R (2006) Characterization of dominant microbiota of a Ghanaian fermented milk product, Nyarmie, by culture and non-culture- based methods. Journal of Applied Microbiology Vol. 100 pp1355-1363 O’Brien, B., Mehra, R., Connolly, J.F. & Harrington, D. (1999c) Seasonal variation in the composition of Irish manufacturing and retail milks. 4: Minerals and trace elements. Irish Journal of Agriculture and Food Research, Vol. 38, pp. 87–99. O’Connor C B. (1993)Traditional cheese making manual ILCA (International Livestock Centre for Africa), Addis Ababa, Ethiopia. O’Connor C B and Tripachi BR (1995), Milk processing techniques Fermented Milk. In: Rural Dairy Processing Training Series: Module 3, Int. Livestock Centre for Africa Publications. O’Callaghan, D.J. & Guinee, T.P. (2004) Rheology and texture of cheese. Cheese: Chemistry, Physics and Microbiology, Volume 1: General Aspects Ogunbanwo ST, Sanni AI, Onilude AA. Influence of cultural conditions on the production of bacteriocin by Lactobacillus brevis OG1.African J Africa Journal of Biotechnology Vol 2(7): pp. 179 – 84. Ogundiwin, J.O, Oke, O.L. (1983). Factors affecting the processing of Wara – a Nigerian white cheese. Food Chemistry Vol. 11: pp 1-13. Ogundiwin, J.O. (1978). A study of the traditional manufacturing processes and chemical composition of warankasi, a Nigerian soft white cheese. Nigerian Food Journal Vol. 2: pp 72-78. Okantah, S. A. , Obese, F. Y. , Odooye E. O. K., Gyawu P. and Asante Y.(1999) A survey on processing and marketing characteristics of peri-urban agropastoral dairying in Ghana. Ghana Journal of Agriculture Vol.32, 55. Okantah, S. A. (1990) Traditional milk production and processing on the Accra plains. Cattle Research Network Newsl. No.2 International Livestock Centre for Africa, Addis Ababa, Ethiopia. University of Ghana http://ugspace.ug.edu.gh 156 Olalokun EA, (1996), Milk production in West Africa: objectives and research nisin for the preservation of nono, a Nigerian fermented milk product, Journal of Food Processing and Preservation Vol. 20,pp. 71-78. Oliver S. P., Jayarao B. M. and Almeida R. A.( 2005), “Food-borne Pathogens, Mastitis, Milk Quality, and Dairy Food Safety,” 43rd National Mastitis Council— Annual Meet-ing Proceedings, Orlando, pp. 3-27. Olson, D.W., van Hekken, D.L., Tunick, M.H., Soryal, K.A. & Zeng, S.S. (2006) Effects of aging on functional properties of caprine milk made into Cheddar and Colby-like cheeses. Small Ruminant Research, Vol. 70, 218–227. Olson N. F., (1990) “The Impact of Lactic Acid Bacteria on Cheese Flavor,” FEMS Microbiology Reviews, Vol. 87, No. 1-2, pp. 131-148. Oscar, S.C., Gerardo, S.C., Jose, L.P.H., Mariano, G.R. and Ernesto, F.T., (1999), Effect of substrate composition on the mycelia growth of Pleurotus ostreatus: an analysis by mixture and response surface methodologies. Process Biochemistry, Vol. 35: pp.127–133. Otchere, E.O. and Okantah, S.A (2001) Small holder dairy production and marketing in Ghana. Padmanabhan, S. Chitre, A, and Shastri, N.V. (1993), “Milk clotting protease isolated from Dieffenbachia maculate”, Die Nahrung, Vol. 37, pp. 99-101. Papaioannou, G., Chouliara, I., Karatapanis, A.E., Kontominas, M.G., and Savvaidis, I.N. (2006). Shelf-life of a Greek whey cheese under modified atmosphere packaging. International Dairy Journal Vol 17, Issue 4, pp 358-364. Parrotta, J. A. (2001): Healing plants of Peninsular India. AB International Wallingford, UK. P.944. Pintado, M.E., and Malcata, F.X. (2000b). Optimization of modified atmosphere packaging with respect to physicochemical characteristics of Requeijao. Food Research International, Vol. 33, pp.821–832. Piggott J.R (1995) Designs questions in sensory and consumer science. Food Quaility Pref. Vol. 6, pp.217-220. University of Ghana http://ugspace.ug.edu.gh 157 Piggot, J.R and Hunter,E.A.(1999). Review: Evaluation of assessor performance in sensory analysis.Italian Journal of Food Science. 4(11), pp. 289-303 Poças, M. F. and Pintado, M. (2010) Packaging and shelf life of cheese. Chapter 6 Food Packaging Department, Biotechnology College Portuguese Catholic University Porto, Portugal © 2010 by Taylor and Francis Group, LLC. Pozsar, H.K., Vamos, V.L., Nonn, S.H. and Hegedues, V.E. (1969), “Investigation of milk clotting enzymes from the point of view of cheese production -Selective inhibition of the proteolytic activity of plant proteases”, Elelmiszertudomany, Vol. 3, pp. 55-70. Repelius, C. (1993) Application of Maxiren in cheesemaking. Milk Proteins ‘93 (ed. J. Mathieu), Environment and Natural Resources Research Vol. 2, No. 1 pp. 114– 121 Rétiveau, A., Chambers, D.H. and Esteve, E. (2005). Developing a lexicon for the flavor description of French cheeses. Food Quality Preference Vol 16, pp. 517–527. Robert, T. (1998): Cold Pasteurization of food by irradiation. Nutrition Food Science Vol. 11 (1), pp. 40-44. Rogelj I., Perko B., Francky A., Penca V. and Pungercar J. (2001). Recombinant lamb chymosin as an alternative coagulating enzyme in cheese production. Journal of Dairy Science 84: pp 1020-1026. Roseiro, L.B., Barbosa, M., Ames, J.M. and Wilbey, R.A. (2003), “Cheese making with vegetable coagulants - the use of Cynara L. for the production of ovine milk cheeses”, International Journal of Dairy Technology, Vol. 56 No. 2, pp. 76 - 85. Rosenthal I, Martinot M, Linder P, Juven BJ. (1983). A study of ionizing radiation of dairy products, Milchwissenschaft,38: pp 467. Rossetti L., Fornasari M. E., Gatti M., Lazzi C., Neviani E. and Giraffa G. (2008), “Grana Padano Cheese Whey Starters: Microbial Composition and Strain Distribution,” International Journal of Food Microbiology, Vol. 127, No. 1-2, pp. 168-171. Rynne, N.M., Beresford, T.P., Kelly, A.L. & Guinee, T.P. (2004) Effect of milk pasteurization temperature and in situ whey protein denaturation on the composition, texture and heat-induced functionality of half-fat Cheddar cheese. International Dairy Journal, Vol.14, pp 989–1001. University of Ghana http://ugspace.ug.edu.gh 158 Sallami, L., Kheadr, E. E., Fliss, I., and Vuillemard, C.J. (2004). Impact of autolytic, proteolytic, and nisin-producing adjunct cultures on biochemical and textural properties of Cheddar cheese. Journal of Dairy Science, Vol. 87, pp. 1585–1594. Samelis, J., Kakouri, A., and Rementzis, J. (2000). Selective effect of the product type and the packaging conditions on the species of lactic acid bacteria dominating the spoilage microbial association of cooked meats at 4 °C. Food Microbiology Vol. 17, pp. 329–340. Sameen, A., Anjum, F.M., Huma, N. & Nawaz, H. (2008) Quality evaluation of Mozzarella cheese from different milk sources. Pakistan Journal of Nutrition, Vol. 7, pp. 753–756. Sanni, A.I, Onilude, A.A., Momoh, M.O. (1999). Selection of starters and a starter- mediated novel procedure for production of Wara, a West African soft cheese. International Journal of Food Science and Technology Vol. 34: pp. 325-333. Scott, R. (1986), “Cheese making practice”, 2nd Edition, Elservier Applied Science Publisher, London, pp. 529. Sessou Philippe, Souaïbou Farougou, Paulin Azokpota, Issaka Youssao, Boniface Yèhouenou, Serge Ahounou and Dominique Codjo Koko Sohounhloué (2013). Endogenous methods for preservation of ‘wagashi’, a Beninese traditional cheese. African Journal of Agricultural Research Vol. 8(31), pp. 4254-4261. Severini C., Bressa F., Romani S., Dalla Rosa, M. 1998. Physical and chemical changes in vacumm packaged Parmigiano Reggiano cheese during storage. Journal of Food Quality Vol. 21: pp. 355–367. SFSP–GTZ–MOFA (Ministry of Food and Agriculture). (1998). Survey of milk production in the Techiman District of Brong Ahafo region. Paper presented at the SFSP–GTZ–MOFA workshop in co-operation with the Dairy Science Unit, Animal Science Department, Kwame Nkrumah University of Science and Technology, Kumasi held at Sunyani, Ghana, October 1998. Sharma, P., Singh L., & Gilbaghi, N. (2009). Optimisation of process variables for decolorization of Disperse Yellow 211 by Bacillus subtilis using Box-Behnken design. Journal of Harzadous materials, Vol.164 (2-3), pp. 1024-1029. Shittu, B. O.; Popoola, T. O. S. And Taiwo, O. (2004): Potentials of Calotropis procera leaves for Wastewater treatment. Proceedings of the International Conference on Science and National Development held at University of Agriculture, Abeokuta. pp. 97-101. Sidhu JS, Al-Saqer J, Al-Zenki S. 1997. Comparison of methods for the assessment of the extent of staling in bread. Journal of Food Chemistry Vol. 58: pp 161–7. University of Ghana http://ugspace.ug.edu.gh 159 Simões M. G., Rabelo J. G., Portal R. E., Domingues A. F. N., Oliveira E. B., Fortes F.C. L. L. (2013) , Physicochemical properties of Butter cheese from Marajó manufactured with buffalo milk and cow milk. IOSR Journal of Environmental Science, Toxicology and Food Technology Vol. 5, Issue 3 pp 2319-2402 Slaterry L., Callaghan J. O., Fitgerald G. F., Beresford T. and Ross R. P. (2010), “Invited Review: Lactobacillus helveticus- A Thermophilic Dairy Starter Related to Gut Bacteria,”Journal of Dairy Science, Vol. 93, No. 10, , pp. 4435-4454. Sofos, J.N. (2002): Microbial control in foods, needs and concern. In Control of food borne microorganisms (Juneja, V. and Sofos, J.N. eds.)1st Ed. New York, Academic Press, London, pp.1-13. Sousa, M.J., Y.O. Ard and P.L.H. McSweeney, 2001 Advances in the study of proteolysis during cheese ripening. International Dairy Journal, Vol. 11: 327-345 Steel, R. and Torrie, J. (1982): Introduction to statistics, International 2nd Ed. pp. 200-258 (Eds. H McGrow & E Kogak) USA. Stewart-Knox, B., & Mitchell, P. (2003). What separates the winners from the losers in new food product development? Trends in Food Science and Technology, Vol. 14: pp 58-64. Stiles ME, Holzapfel WH. (1997 Lactic acid bacteria of foods and their current taxonomy. International Journal of Food Microbiology ; 36: 1- 29. Stone, H., and Sidel, J. (1993), Sensory evaluation practices, pp.143–201, Academic Press, New York. Stone H. 1992. Quantitative descriptive analysis (QDA). In: Hootman RC, editor. Manual on descriptive analysis testing for sensory evaluation. ASTM Manual Series MNL 13. Philadelphia, Pa.: ASTM. Stone H, Sidel J. 1998. Sensory evaluation by quantitative descriptive analysis: development, applications, and the future (abstract). IFT Annual Meeting; 20-24 June; Atlanta, Ga. Abstract NR 44-2. Tamer, I.M. (1993), “Identification and partial purification of a novel milk clotting enzyme from Onopordum turcicum”, Biotechnology Letters, Vol. 15, pp. 427-432. Tohibu Sulley Addo, Amankwah Emmanuel and Oduro Ibok (2013). Chemical stability of vacuum packaged West African cheese (Wagashie). Scientific Research and Essays Vol. 8(26), pp. 1212-1218 University of Ghana http://ugspace.ug.edu.gh 160 Tybor, P.T. and Gilson, W. (2003) Dairy Producer’s Guide to Food Safety in Milk Production. Cooperative Extension Service, College of Agriculture and Environmental Science, University of Georgia, Athens, Georgia, USA. Tsiotsias A, Savvaidis I, Vassila A, Kontominas MG, Kotzekidou P (2002). Control of Listeria monocytogenes by low-dose irradiation in combination with refrigeration in the soft whey cheese. Anthotyros Food Microbiology Vol. 19:pp. 117-126. Turkoglu, H., Z.G. Ceylan, K.S. Dayisoylu,( 2003).The microbiological and chemical quality of Orgu cheese produced in Turkey. Pak. Journal of food Nutrition, Vol. 2(2), pp. 92-94 Thayer, D.W. (2005): Food irradiation: benefits and concerns. Journal of Food Quality Vol. 13 (1) pp. 147-169. Tetra Pak (2004), Company Magazine number 89, Lund, Sweden Van Hekken, D.L, Drake, M.A., Molina Corral, F.J., Guerrero Prieto, V.M. and Gardea, A.A. (2006) Mexican Chihuahua cheese: sensory profiles of young cheese. Journal of Dairy Science, Vol. 89, pp. 3729– 3738 Vlaemynck, G. (1992) Study of lypolytic activity of the lipoprotein lipase in lunch cheese of the Gouda type. Milchwissenschaft, Vol. 47, pp. 164–167. Wouters J. T. M., Ayad E. H. E., Hugenholtz J. and Smit G. (2002), “Microbes from Raw Milk for Fermented Dairy Products,” International Dairy Journal, Vol. 12, No. 2-3, pp. 91-109 WHO, World Health Organization (2005): Safety and nutritional adequacy of irradiated food. Geneva, Switzerland. Winwood, J., (1983). Quark production methods: the present and future. Journal of the Society of Dairy Technology 36 (4), pp. 107-109. Widyastuti Y., R. and Febrisiantosa A., (2014) "The Role of Lactic Acid Bacteria in Milk Fermentation," Food and Nutrition Sciences, Vol. 5 No. 4, pp. 435-442. Wu, D., Zhou, J., and Li, Y. (2009) Effect of the sulfidation process on the mechanical properties of a CoMo/Al2 03 hydrotreating catalyst. Chemical Engeneering Science, Vol. 64(2), pp. 198-206. Zacharof M. P. and Lovitt R. W., (2012) “Bacteriocins Produced by Lactic Acid Bacteria,” APCBEE Procedia, Vol. 2, pp. 50-56. University of Ghana http://ugspace.ug.edu.gh 161 APENDDIX Apendix I Principal Component Analysis: Correlation of the discriptors of wagashie Factor loadings: F1 F2 F3 F4 F5 F6 F7 F8 F9 F10 F11 F12 F13 Soft 0.6801 0.5407 0.1734 -0.3253 0.2713 0.0913 -0.1363 0.0176 -0.0122 -0.0073 0.0643 0.0623 -0.0149 Spongy - 0.6882 - 0.1278 0.4562 0.2634 0.3818 -0.1692 0.0812 -0.0061 0.0602 -0.0703 0.1903 -0.0811 0.0152 Smooth 0.7059 0.5022 0.0229 -0.2626 0.2977 0.2236 -0.0649 -0.0513 0.0572 -0.1642 0.0295 -0.0330 0.0484 Crumbly - 0.4493 - 0.1451 0.4723 0.5535 0.0722 -0.3389 0.2922 0.1006 -0.1098 0.0364 -0.0483 0.1230 0.0367 Milky aroma 0.5378 - 0.5508 0.4614 0.2078 -0.1576 0.1899 -0.1225 -0.0987 0.0501 0.1936 0.1440 0.0600 -0.0372 Yoghurt aroma 0.9343 - 0.0302 - 0.0374 -0.2207 0.0131 -0.0686 0.2170 -0.0384 -0.1179 0.0675 0.0572 0.0070 -0.0288 Cheesy aroma 0.5487 - 0.5290 - 0.0498 -0.3360 -0.4147 -0.0081 -0.1177 0.2890 0.1524 0.0438 0.0800 0.0060 0.0529 Beefy aroma 0.0941 - 0.5143 0.1413 -0.1703 0.8014 0.0340 0.0140 0.0498 0.1379 0.0960 -0.0316 0.0330 -0.0366 Fried sweet potato aroma - 0.7147 0.2519 0.4368 -0.2390 -0.1226 0.2208 -0.1101 0.2892 -0.1222 -0.0091 -0.0213 -0.0337 -0.0416 University of Ghana http://ugspace.ug.edu.gh 162 Fried ripe plantain aroma - 0.6786 0.3585 0.3195 -0.3359 -0.0778 0.3276 0.1178 -0.1287 -0.0892 -0.0303 0.1364 0.1534 0.0349 Smoked chevon aroma - 0.4045 - 0.2054 - 0.6330 0.5266 0.1914 0.1588 0.0072 0.1342 -0.0066 0.0876 0.1519 -0.0690 0.0269 spoilt milk aroma 0.5370 0.3776 0.3755 0.5538 -0.1809 0.0052 0.2739 0.0032 -0.0053 -0.0403 0.0549 -0.0924 -0.0192 Fermented cassva dough aroma 0.6868 0.4027 0.0791 0.4596 -0.1541 0.2523 0.1892 0.1217 0.1023 0.0093 0.0043 -0.0024 -0.0112 Doughnut aroma - 0.7379 0.3846 0.3148 -0.3145 0.0293 0.2395 0.0807 0.1430 -0.1024 0.1046 0.0036 -0.0546 -0.0042 Sour taste 0.1144 0.9063 0.2328 0.2045 -0.0054 -0.1330 -0.1493 -0.0767 0.1054 0.0915 0.0001 0.0019 0.0649 Bland 0.1294 0.1498 - 0.4397 0.6650 0.1753 0.5154 0.0224 -0.0411 -0.0782 0.0832 -0.1132 0.0215 0.0198 Cheesy taste 0.1700 - 0.4409 0.7765 0.2321 0.0348 0.1936 0.0323 0.1311 0.1931 -0.1003 -0.1108 0.0552 0.0160 Milky taste 0.2336 - 0.7972 0.4173 -0.2023 0.0054 0.0670 -0.0935 -0.1346 -0.1587 0.1513 -0.0618 -0.0830 0.0683 Bitter taste 0.1205 0.7466 0.3080 0.3783 -0.0621 -0.2247 -0.3433 -0.0355 0.0061 0.1237 -0.0017 0.0215 -0.0288 Fried egg taste - 0.6271 - 0.4913 0.3244 0.1722 -0.2768 0.2169 -0.0320 -0.2864 0.1081 -0.0936 -0.0030 -0.0507 -0.0277 Salty taste - 0.3602 0.4788 - 0.0528 -0.5641 -0.0770 -0.0060 0.4637 -0.0831 0.2300 0.1868 -0.0497 -0.0424 0.0029 Whitish colour 0.6105 0.1866 0.7321 -0.1026 0.1367 -0.0009 -0.0044 0.0033 -0.1183 0.0005 -0.0390 -0.1079 0.0071 Brownish colour - 0.8845 0.3043 0.0642 0.0896 0.0223 0.0216 -0.2943 -0.0008 0.1385 0.0641 -0.0384 -0.0243 0.0019 Factor scores: University of Ghana http://ugspace.ug.edu.gh 163 Observation F1 F2 F3 F4 F5 F6 F7 F8 F9 F10 F11 F12 F13 997 4.8591 3.2630 2.1102 3.5438 -0.3949 0.4483 1.0498 -0.0260 0.0402 -0.1213 -0.0295 -0.1492 0.0290 246 - 1.0383 3.1858 0.9035 0.9275 -0.0486 -1.1839 -1.8244 0.1382 0.2130 0.6393 0.2202 0.4383 -0.0868 203 - 1.4521 - 2.4828 0.6833 1.0232 -0.5336 -0.4594 -0.1020 -0.8118 0.0448 -0.0092 -0.5321 -0.2823 -0.4018 110 1.7131 - 2.0619 1.7021 -0.7035 3.5628 -0.0254 0.2077 0.0871 0.3692 0.0193 -0.2774 0.2463 0.0361 355 0.8085 - 2.1088 0.1521 -0.0877 0.0394 -1.4332 0.5261 -0.4307 -1.3273 0.5417 0.3572 -0.1299 0.1475 961 - 1.7646 - 2.9796 - 0.0350 1.4325 -0.9221 -0.3502 -0.4040 -0.2188 1.1376 0.0668 0.0998 -0.2674 0.3175 707 - 2.4387 - 1.4209 1.7824 0.0336 -0.8061 1.7571 0.0508 -0.6234 -0.3339 -0.4893 0.4124 0.6038 0.0096 101 - 1.9925 - 1.1097 2.2418 -0.6150 -0.6949 0.5296 -0.1646 1.8570 -0.3676 0.0347 -0.1422 -0.2514 -0.0280 838 3.0930 - 0.3293 - 1.0721 -1.6923 0.1862 -0.3681 -0.1853 0.2033 0.4082 -0.5458 0.9600 -0.2708 -0.1928 614 1.7820 - 0.3546 - 1.8955 -1.2591 -1.5304 -1.3479 1.1615 0.4320 0.2531 -0.2846 -0.4458 0.5784 0.0092 616 4.7089 - 0.1553 - 1.6046 -1.9928 -0.7164 1.7101 -1.2057 -0.3084 -0.1426 0.5016 -0.3849 -0.0945 0.0599 706 - 2.0749 3.1787 - 0.1832 -1.0078 0.5295 -0.6218 -0.8564 -0.3751 -0.4469 -1.0309 -0.3826 -0.2661 0.1573 421 - 3.8324 3.3316 0.0597 -2.4658 0.1672 0.5226 1.5713 -0.3698 0.4218 0.6404 0.0878 -0.1949 -0.0152 417 - 2.3712 0.0438 - 4.8447 2.8633 1.1621 0.8222 0.1752 0.4463 -0.2696 0.0374 0.0571 0.0398 -0.0416 University of Ghana http://ugspace.ug.edu.gh 164 Appendix II Generated Attributed by the 13 member panel during the Quantitative Descriptive sensory evaluation of ‘wagashie’ on ‘taste’, ‘texture’, ‘colour’ and ‘aroma’. During the training of the panel members for the descriptive analysis 23 descriptors were generated in a focus group discussion. These descriptors were documented and used to evaluate the 14 samples. Table 4.10 shows the attributes and the generated descriptors by the 13 member panel for the quantitative descriptive sensory evaluation Table 4.5: Generated attributes and descriptors for ‘wagashie’ Attributes Descriptors Taste Sour Bland Cheesy Milky Bitterness Fried egg Salty Texture Soft Smooth Crumbly Spongy Aroma Milky Yoghurt Cheesy Beefy Spoilt milk Fermented cassava dough Doughnut Fried ripe plantain Smoked chevon Fried sweet potato Colour Whitish Brownish University of Ghana http://ugspace.ug.edu.gh 165 Definition of generated attributes Some of the descriptors generated were considered as desirable and undesirable, table 4.6 shows the desirable and undesirable descriptors and their definitions. Table 4.6: The desirable and undesirable descriptors generated by the 14 member panel during the training Attribute Desirable Definition Undesirable Definition Taste Sour Taste sensation associated with fermented milk Bland Tasteless sensation in a food product Cheesy Taste sensation associated with cheese Bitter taste Taste sensation of quinine or caffeine Milky Taste associated with fresh milk Fried egg Taste associated with fried egg Salty Taste produced by a solution of sodium chloride Texture Soft The degree of stickiness in the mouth Smooth The spreadability of the cheese Spongy The texture University of Ghana http://ugspace.ug.edu.gh 166 characteristics of a bouncy cheese Crumbly The extent to which the cheese breaks in the mouth Aroma Milky Aromatics of milk from dairy origin Spoilt milk Aromatics of spoilt milk Yoghurt Aromatics of plain yoghurt Fermented cassava dough Aromatics of fermented cassava dough Cheesy The aromatic sensation of cheese Beefy Smell associated with cow meat Smoked chevon The aromatics associated with grilled goat meat Fried sweet potato The smell associated with fried sweet potato Fried ripe plantain The smell associated with fried ripe plantain Doughnut The smell associated with fried University of Ghana http://ugspace.ug.edu.gh 167 doughnut Colour Whitish Lighter colour Brownish Darker colour Sample Questionnaire for the Acceptability of Wagashie Name:………………………………………………………………………… Date:…………. Please rate the texture, colour, taste and overall acceptability of the samples provided. Indicate your choice with a tick. Please rinse your mouth in-between sample tasting. Texture: Sample code 1st 2nd 3rd 4th 5th 9. Like extremely 8. Like very much 7. Like moderately 6. Like slightly 5. Neither like nor dislike 4. Dislike slightly 3. Dislike moderately 2. Dislike very much 1. Dislike extremely Apenddix III Response Surface Regression: Texture, versus Ferment time, ... University of Ghana http://ugspace.ug.edu.gh 168 Response Surface Regression: Texture versus Ferment time, Extract(wt/g, Salt(g) The analysis was done using uncoded units. Estimated Regression Coefficients for Texture Term Coef SE Coef T P Constant 11.2921 3.15705 3.577 0.016 Ferment time -1.1441 0.38526 -2.970 0.031 Extract(wt/g) 0.0300 0.02148 1.395 0.222 Salt(g) -0.5506 0.24774 -2.222 0.077 Ferment time*Ferment time 0.1143 0.03771 3.031 0.029 Extract(wt/g)*Extract(wt/g) -0.0000 0.00006 -0.749 0.488 Salt(g)*Salt(g) 0.0249 0.00603 4.125 0.009 Ferment time*Extract(wt/g) 0.0043 0.00145 2.985 0.031 Ferment time*Salt(g) -0.0032 0.01449 -0.224 0.831 Extract(wt/g)*Salt(g) -0.0024 0.00058 -4.123 0.009 S = 0.2898 R-Sq = 96.9% R-Sq(adj) = 91.3% Analysis of Variance for Texture Source DF Seq SS Adj SS Adj MS F P Regression 9 13.1486 13.14861 1.460956 17.39 0.003 Linear 3 8.7824 1.45320 0.484400 5.77 0.044 Square 3 2.1857 2.18568 0.728561 8.67 0.020 Interaction 3 2.1805 2.18047 0.726825 8.65 0.020 Residual Error 5 0.4200 0.41997 0.083993 Lack-of-Fit 3 0.4069 0.40690 0.135633 20.76 0.046 Pure Error 2 0.0131 0.01307 0.006533 Total 14 13.5686 Unusual Observations for Texture Obs StdOrder Texture Fit SE Fit Residual St Resid 1 1 4.600 4.918 0.251 -0.318 -2.19 R 13 13 6.870 6.553 0.251 0.317 2.19 R R denotes an observation with a large standardized residual. Response Surface Regression: Colour versus Ferment time, Extract(wt/g), Salt(g) The analysis was done using uncoded units. Estimated Regression Coefficients for Colour Term Coef SE Coef T P Constant 10.2946 2.82625 3.642 0.015 Ferment time -0.1657 0.34489 -0.480 0.651 Extract(wt/g) 0.0045 0.01923 0.237 0.822 Salt(g) -0.2766 0.22178 -1.247 0.268 University of Ghana http://ugspace.ug.edu.gh 169 Ferment time*Ferment time -0.0383 0.03376 -1.136 0.308 Extract(wt/g)*Extract(wt/g) 0.0000 0.00005 0.235 0.824 Salt(g)*Salt(g) 0.0113 0.00540 2.086 0.091 Ferment time*Extract(wt/g) 0.0022 0.00130 1.657 0.158 Ferment time*Salt(g) -0.0065 0.01297 -0.501 0.638 Extract(wt/g)*Salt(g) -0.0010 0.00052 -1.927 0.112 S = 0.2594 R-Sq = 86.5% R-Sq(adj) = 62.2% Analysis of Variance for Colour Source DF Seq SS Adj SS Adj MS F P Regression 9 2.1541 2.1541 0.23934 3.56 0.088 Linear 3 1.2937 0.1353 0.04511 0.67 0.606 Square 3 0.4086 0.4086 0.13619 2.02 0.229 Interaction 3 0.4518 0.4518 0.15060 2.24 0.202 Residual Error 5 0.3366 0.3366 0.06731 Lack-of-Fit 3 0.2299 0.2299 0.07663 1.44 0.435 Pure Error 2 0.1067 0.1067 0.05333 Total 14 2.4906 Response Surface Regression: Taste versus Ferment time, Extract(wt/g), Salt(g) The analysis was done using uncoded units. Estimated Regression Coefficients for Taste Term Coef SE Coef T P Constant 15.5101 4.91673 3.155 0.025 Ferment time -0.5930 0.59999 -0.988 0.368 Extract(wt/g) 0.0192 0.03346 0.573 0.591 Salt(g) -1.0134 0.38583 -2.627 0.047 Ferment time*Ferment time 0.0384 0.05872 0.655 0.542 Extract(wt/g)*Extract(wt/g) -0.0000 0.00009 -0.410 0.699 Salt(g)*Salt(g) 0.0316 0.00940 3.358 0.020 Ferment time*Extract(wt/g) 0.0027 0.00226 1.185 0.289 Ferment time*Salt(g) -0.0018 0.02257 -0.078 0.941 Extract(wt/g)*Salt(g) -0.0011 0.00090 -1.185 0.289 S = 0.4514 R-Sq = 81.0% R-Sq(adj) = 46.8% Analysis of Variance for Taste Source DF Seq SS Adj SS Adj MS F P Regression 9 4.33817 4.33817 0.48202 2.37 0.178 Linear 3 1.33535 1.84830 0.61610 3.02 0.132 Square 3 2.42915 2.42915 0.80972 3.97 0.086 Interaction 3 0.57367 0.57368 0.19123 0.94 0.488 Residual Error 5 1.01860 1.01860 0.20372 University of Ghana http://ugspace.ug.edu.gh 170 Lack-of-Fit 3 0.97000 0.97000 0.32333 13.31 0.071 Pure Error 2 0.04860 0.04860 0.02430 Total 14 5.35677 Unusual Observations for Taste Obs StdOrder Taste Fit SE Fit Residual St Resid 1 1 5.400 5.875 0.391 -0.475 -2.10 R 13 13 6.800 6.325 0.391 0.475 2.10 R R denotes an observation with a large standardized residual. Response Surface Regression: Overall acce versus Ferment time, ... The analysis was done using uncoded units. Estimated Regression Coefficients for Overall acceptability Term Coef SE Coef T P Constant 13.4324 5.04147 2.664 0.045 Ferment time -0.9158 0.61521 -1.489 0.197 Extract(wt/g) 0.0152 0.03430 0.443 0.676 Salt(g) -0.7157 0.39562 -1.809 0.130 Ferment time*Ferment time 0.0611 0.06021 1.015 0.356 Extract(wt/g)*Extract(wt/g) -0.0000 0.00010 -0.240 0.820 Salt(g)*Salt(g) 0.0257 0.00963 2.666 0.045 Ferment time*Extract(wt/g) 0.0053 0.00231 2.312 0.069 Ferment time*Salt(g) -0.0100 0.02314 -0.432 0.684 Extract(wt/g)*Salt(g) -0.0015 0.00093 -1.588 0.173 S = 0.4628 R-Sq = 83.0% R-Sq(adj) = 52.5% Analysis of Variance for Overall acceptability Source DF Seq SS Adj SS Adj MS F P Regression 9 5.2432 5.2432 0.58258 2.72 0.141 Linear 3 1.7995 1.2774 0.42581 1.99 0.234 Square 3 1.7186 1.7186 0.57288 2.67 0.158 Interaction 3 1.7251 1.7251 0.57504 2.68 0.157 Residual Error 5 1.0709 1.0709 0.21419 Lack-of-Fit 3 0.9237 0.9237 0.30789 4.18 0.199 Pure Error 2 0.1473 0.1473 0.07363 Total 14 6.3142 Response Surface Regression: Texture versus Fermentation time, salt(g), Rennet(ml) The following terms cannot be estimated and were removed: University of Ghana http://ugspace.ug.edu.gh 171 salt(g)*Rennet(ml) Method Analysis of Variance Source DF Adj SS Adj MS F-Value P-Value Model 8 1.72837 0.216046 12.38 0.031 Linear 3 0.70253 0.234178 13.42 0.030 Fermentation time 1 0.14981 0.149813 8.59 0.061 salt(g) 1 0.24000 0.240000 13.76 0.034 Rennet(ml) 1 0.02821 0.028213 1.62 0.293 Square 3 0.55729 0.185763 10.65 0.042 Fermentation time*Fermentation time 1 0.00988 0.009882 0.57 0.506 salt(g)*salt(g) 1 0.32120 0.321202 18.41 0.023 Rennet(ml)*Rennet(ml) 1 0.15301 0.153015 8.77 0.059 2-Way Interaction 2 0.18138 0.090688 5.20 0.106 Fermentation time*salt(g) 1 0.02138 0.021376 1.23 0.349 Fermentation time*Rennet(ml) 1 0.16000 0.160000 9.17 0.056 Error 3 0.05233 0.017444 Lack-of-Fit 1 0.01307 0.013067 0.67 0.500 Pure Error 2 0.03927 0.019633 Total 11 1.78070 Model Summary S R-sq R-sq(adj) 0.132077 97.06% 89.22% Coded Coefficients Term Effect Coef SE Coef T-Value P-Value VIF Constant 6.5133 0.0763 85.42 0.000 Fermentation time 0.3533 0.1767 0.0603 2.93 0.061 1.44 salt(g) -0.6000 -0.3000 0.0809 -3.71 0.034 1.84 Rennet(ml) -0.1533 -0.0767 0.0603 -1.27 0.293 1.25 Fermentation time*Fermentation time 0.1283 0.0642 0.0853 0.75 0.506 1.22 salt(g)*salt(g) -0.7317 -0.3658 0.0853 -4.29 0.023 1.22 Rennet(ml)*Rennet(ml) 0.5050 0.2525 0.0853 2.96 0.059 1.25 Fermentation time*salt(g) 0.223 0.112 0.101 1.11 0.349 1.70 Fermentation time*Rennet(ml) 0.4000 0.2000 0.0660 3.03 0.056 1.00 Regression Equation in Uncoded Units Texture = 5.00 - 0.258 Fermentation time + 0.510 salt(g) - 0.1844 Rennet(ml) + 0.0160 Fermentation time*Fermentation time - 0.02986 salt(g)*salt(g) + 0.01168 Rennet(ml)*Rennet(ml) + 0.0160 Fermentation time*salt(g) + 0.02151 Fermentation time*Rennet(ml) Response Surface Regression: Colour versus Fermentation time, salt(g), Rennet(ml) University of Ghana http://ugspace.ug.edu.gh 172 The following terms cannot be estimated and were removed: Fermentation time*Rennet(ml) Method Analysis of Variance Source DF Adj SS Adj MS F-Value P-Value Model 8 0.522550 0.065319 5.37 0.061 Linear 3 0.344981 0.114994 9.46 0.027 Fermentation time 1 0.043802 0.043802 3.60 0.130 salt(g) 1 0.189113 0.189113 15.56 0.017 Rennet(ml) 1 0.148519 0.148519 12.22 0.025 Square 3 0.077544 0.025848 2.13 0.240 Fermentation time*Fermentation time 1 0.000380 0.000380 0.03 0.868 salt(g)*salt(g) 1 0.070917 0.070917 5.83 0.073 Rennet(ml)*Rennet(ml) 1 0.006417 0.006417 0.53 0.508 2-Way Interaction 2 0.100025 0.050012 4.11 0.107 Fermentation time*salt(g) 1 0.040000 0.040000 3.29 0.144 salt(g)*Rennet(ml) 1 0.060025 0.060025 4.94 0.090 Error 4 0.048619 0.012155 Lack-of-Fit 2 0.048619 0.024309 Pure Error 2 0.000000 0.000000 Total 12 0.571169 Model Summary S R-sq R-sq(adj) R-sq(pred) 0.110248 91.49% 74.46% 0.00% Coded Coefficients Term Effect Coef SE Coef T-Value P-Value VIF Constant 7.3300 0.0637 115.16 0.000 Fermentation time -0.1813 -0.0906 0.0477 -1.90 0.130 1.13 salt(g) -0.3075 -0.1537 0.0390 -3.94 0.017 1.00 Rennet(ml) 0.3338 0.1669 0.0477 3.50 0.025 1.13 Fermentation time*Fermentation time -0.0225 -0.0113 0.0637 -0.18 0.868 1.08 salt(g)*salt(g) -0.3075 -0.1538 0.0637 -2.42 0.073 1.03 Rennet(ml)*Rennet(ml) 0.0925 0.0463 0.0637 0.73 0.508 1.08 Fermentation time*salt(g) 0.2000 0.1000 0.0551 1.81 0.144 1.00 salt(g)*Rennet(ml) 0.2450 0.1225 0.0551 2.22 0.090 1.00 Regression Equation in Uncoded Units Colour = 7.079 - 0.184 Fermentation time + 0.151 salt(g) - 0.0660 Rennet(ml) - 0.0028 Fermentation time*Fermentation time - 0.01255 salt(g)*salt(g) + 0.00214 Rennet(ml)*Rennet(ml) + 0.01429 Fermentation time*salt(g) + 0.00753 salt(g)*Rennet(ml) University of Ghana http://ugspace.ug.edu.gh 173 Response Surface Regression: Taste versus Fermentation time, salt(g), Rennet(ml) Analysis of Variance Source DF Adj SS Adj MS F-Value P-Value Model 9 6.48237 0.72026 2.85 0.210 Linear 3 3.37098 1.12366 4.45 0.126 Fermentation time 1 0.56856 0.56856 2.25 0.231 salt(g) 1 3.03882 3.03882 12.03 0.040 Rennet(ml) 1 0.07008 0.07008 0.28 0.635 Square 3 2.29669 0.76556 3.03 0.193 Fermentation time*Fermentation time 1 0.09106 0.09106 0.36 0.591 salt(g)*salt(g) 1 1.87072 1.87072 7.40 0.072 Rennet(ml)*Rennet(ml) 1 0.15759 0.15759 0.62 0.487 2-Way Interaction 3 0.57631 0.19210 0.76 0.586 Fermentation time*salt(g) 1 0.02860 0.02860 0.11 0.759 Fermentation time*Rennet(ml) 1 0.00122 0.00122 0.00 0.949 salt(g)*Rennet(ml) 1 0.43574 0.43574 1.72 0.280 Error 3 0.75793 0.25264 Lack-of-Fit 1 0.09127 0.09127 0.27 0.653 Pure Error 2 0.66667 0.33333 Total 12 7.24031 Model Summary S R-sq R-sq(adj) R-sq(pred) 0.502637 89.53% 58.13% * Coded Coefficients Term Effect Coef SE Coef T-Value P-Value VIF Constant 6.537 0.290 22.52 0.000 Fermentation time 0.688 0.344 0.229 1.50 0.231 1.44 salt(g) 2.135 1.068 0.308 3.47 0.040 2.13 Rennet(ml) -0.242 -0.121 0.229 -0.53 0.635 1.44 Fermentation time*Fermentation time 0.359 0.180 0.299 0.60 0.591 1.14 salt(g)*salt(g) -1.974 -0.987 0.363 -2.72 0.072 1.68 Rennet(ml)*Rennet(ml) 0.473 0.236 0.299 0.79 0.487 1.14 Fermentation time*salt(g) 0.258 0.129 0.384 0.34 0.759 1.71 Fermentation time*Rennet(ml) 0.035 0.017 0.251 0.07 0.949 1.00 salt(g)*Rennet(ml) 1.008 0.504 0.384 1.31 0.280 1.71 Regression Equation in Uncoded Units Taste = -3.11 - 0.211 Fermentation time + 1.795 salt(g) - 0.472 Rennet(ml) + 0.0449 Fermentation time*Fermentation time - 0.0806 salt(g)*salt(g) + 0.0109 Rennet(ml)*Rennet(ml) + 0.0185 Fermentation time*salt(g) + 0.0019 Fermentation time*Rennet(ml) + 0.0310 salt(g)*Rennet(ml) University of Ghana http://ugspace.ug.edu.gh 174 Response Surface Regression: Overall Acceptab versus Fermentation tim, salt(g), Rennet(ml) Analysis of Variance Source DF Adj SS Adj MS F-Value P-Value Model 9 2.57655 0.28628 1.64 0.305 Linear 3 0.40912 0.13637 0.78 0.554 Fermentation time 1 0.20161 0.20161 1.15 0.332 salt(g) 1 0.08000 0.08000 0.46 0.529 Rennet(ml) 1 0.12751 0.12751 0.73 0.432 Square 3 0.91690 0.30563 1.75 0.273 Fermentation time*Fermentation time 1 0.08447 0.08447 0.48 0.518 salt(g)*salt(g) 1 0.06361 0.06361 0.36 0.573 Rennet(ml)*Rennet(ml) 1 0.74354 0.74354 4.25 0.094 2-Way Interaction 3 1.25053 0.41684 2.38 0.185 Fermentation time*salt(g) 1 0.02560 0.02560 0.15 0.718 Fermentation time*Rennet(ml) 1 0.63202 0.63202 3.62 0.116 salt(g)*Rennet(ml) 1 0.59290 0.59290 3.39 0.125 Error 5 0.87402 0.17480 Lack-of-Fit 3 0.70222 0.23407 2.72 0.280 Pure Error 2 0.17180 0.08590 Total 14 3.45057 Model Summary S R-sq R-sq(adj) R-sq(pred) 0.418097 74.67% 29.08% 0.00% Coded Coefficients Term Effect Coef SE Coef T-Value P-Value VIF Constant 7.030 0.241 29.12 0.000 Fermentation time -0.317 -0.159 0.148 -1.07 0.332 1.00 salt(g) -0.200 -0.100 0.148 -0.68 0.529 1.00 Rennet(ml) 0.252 0.126 0.148 0.85 0.432 1.00 Fermentation time*Fermentation time 0.302 0.151 0.218 0.70 0.518 1.01 salt(g)*salt(g) -0.263 -0.131 0.218 -0.60 0.573 1.01 Rennet(ml)*Rennet(ml) -0.898 -0.449 0.218 -2.06 0.094 1.01 Fermentation time*salt(g) -0.160 -0.080 0.209 -0.38 0.718 1.00 Fermentation time*Rennet(ml) 0.795 0.397 0.209 1.90 0.116 1.00 salt(g)*Rennet(ml) 0.770 0.385 0.209 1.84 0.125 1.00 Regression Equation in Uncoded Units Overall Acceptability = 7.27 - 0.339 Fermentation time + 0.093 salt(g) - 0.085 Rennet(ml) + 0.0378 Fermentation time*Fermentation time - 0.0107 salt(g)*salt(g) - 0.0208 Rennet(ml)*Rennet(ml) - 0.0114 Fermentation time*salt(g) + 0.0427 Fermentation time*Rennet(ml) + 0.0237 salt(g)*Rennet(ml) University of Ghana http://ugspace.ug.edu.gh 175 University of Ghana http://ugspace.ug.edu.gh