The typical flavour of goat milk products: technological aspects

The production of goat milk at a world-wide level is 12 million tonnes,1 a large part of which is used for direct consumption. In Europe, however, goat milk is used principally for cheese production. While goat cheeses are consumed as traditional products in Mediterranean countries, the demand for them is on the increase in northern Europe. This new interest could be partially explained by the unique sensorial properties of goat milk products, which are characterized by a specific and typical (‘goat’) flavour. This specific flavour may be undesirable in milk for direct consumption but for cheese production its presence can be much sought-after. The goat flavour originates from the lipid fraction,2 and several aroma compounds (free fatty acids) have been identified (Table 1). Among them, 4-methyloctanoic acid (4-met-C8) and 4-ethyloctanoic acid (4-met-C8) have been found to be principally responsible for the goat flavour, since they are perceived even at very low concentrations and are present at levels higher than their perception threshold.3 The level of these compounds increases during cheese ripening, leading to a flavour improvement.4 These fatty acids, as well as others, are produced by lipolysis, which can be considered at three distinct levels: (i) initial milk lipolysis (spontaneous lipolysis); (ii) lipolysis induced by milking practices, storage and technological operations; and (iii) microbial lipolysis, generated notably by ripening strains. The spontaneous lipolysis is influenced by several genetic and environmental factors,5 which induce important variations of the typical flavour of goat milk at the herd level. At the dairy plant level, the typical flavour of goat milk products should be controlled. The industrial processes applied to goat milk can affect its sensorial properties, and this could have some significant consequences for the manufacture of fresh cheeses. For ripened cheeses, the surface flora has a dominating influence on the development of the typical flavour of goat cheeses.6 This paper presents some of our results concerning: (i) the impact of technological treatments and lipase adjunction on the flavour of goat milk; (ii) the influence of the sensorial properties of goat milk on the typical flavour of fresh cheeses; and (iii) the contribution of yeasts and moulds to the development of the typical flavour of mould-ripened soft goat cheeses. The lipolysis level and the sensorial properties of goat milk (overall, goat and rancid flavours) have been evaluated before and after skimming, cold storage, mild heat treatment and homogenization. The results (Table 2) showed that skimmed goat milk had a very mild flavour compared to full fat raw milk. Heating (65°C, 1 min) led to a slight reduction in the overall and goat flavours, and to a decrease in the level of lipolysis. On the other hand, some treatments such as cold storage and homogenization were found to improve the overall and goat flavours of milk, but this was also accompanied by an increase in the rancid flavour, leading to a flavour defect after homogenization (Table 2). Another illustration of these changes might be given by evaluation of the free fatty acid concentrations in milk after cold storage and homogenization (Fig. 1). The total content of free fatty acids was 40 µg/ml fat in the control milk and it reached 100 and 120 µg/ml fat after cold storage and homogenization, respectively. The level of almost all the free fatty acids increased after these two treatments. However, cold storage induced an important increase in the content of C8 and C10, whereas homogenization led to a preferential increase of the C12 and C16 concentrations. Concentration of free fatty acids of goat milk before and after cold storage and homogenization. Lipids were extracted as described by Le Quéréet al.,4 free fatty acids were obtained according to Needs et al.17 and the analysis was carried out by gas chromatography. Experiments were carried out in order to evaluate the impact of lipase adjunction on the sensorial quality of goat milk. Animal lipases (kid goat, lamb and pig) and microbial lipases (Rhizopus arrhizus, Rhizomucor miehei, Aspergillus niger and Penicillium roqueforti) were added (0.05% and 0.1%) to pasteurized goat milk. The flavour of the milks was then evaluated by trained panellists. No changes were observed in the goat flavour, but an important increase in the rancid flavour was perceived in goat milks treated with the microbial lipases (results not shown). However, Kim Ha and Lindsay7 have shown that microbial lipases produced by Aspergilllus niger, Penicillium roqueforti and Rhizopus arrhizus were able to release relatively large amounts of free 4-et-C8 in goat milk (4.3, 3.4 and 2.1 µg/g fat, respectively). However, the environmental conditions (pH, temperature and water activity) are determining factors for the use of lipases as flavouring agents. In a previous study,8 fresh lactic cheeses were produced with milk having a strong or a weak goat flavour. Fresh cheeses from milk with a strong goat flavour were found to have a stronger goat odour (+23%) and goat flavour (+17%), and a higher free fatty acid content (+40%) than cheeses from weakly goat-flavoured milk. These observations were in agreement with results obtained on white cheese9 and whey cheese.10 However, this study also showed that the sensorial properties of ripened cheeses were more strongly related to the ripening parameters than to the sensorial properties of the raw material.8 Therefore, our investigations on the flavour of ripened cheeses evaluated the influence of the ripening cultures. The development of the flavour of mould-ripened soft goat cheeses is mainly due to the lipolytic activity of the surface flora, underlining the need for a proper selection of the ripening strains. An illustration of such a selection strategy was given in a previous work6 and will be briefly described in this section. In a first step, a rapid screening was carried out on goat milk incubated with 30 different cultures (yeasts and moulds). This survey allowed us to select some strains that could give rise to an increase in the goat flavour, without any appearance of flavour defects. Then, the selected cultures (Penicillium candidum, PC; Geotrichum candidum, GC; Debaryomyces hansenii, DH; Rhodosporidium infirmominiatum, RI) were used for the manufacture of lactic and Camembert-type goat cheeses (mould-ripened soft cheeses). Various combinations of strains were tested, with P. candidum or G. candidum as the primary culture, and the others as associated strains. The obtained results (Fig. 2) showed that G. candidum was the most suitable strain for the development of the goat flavour in cheeses. The use of P. candidum led to less typical products, as well as to the development of flavour defects in lactic cheeses and to mushroom and mouldy notes in Camembert-type cheeses. When these two strains were used as a coculture, the influence of G. candidum was stronger in Camembert-type cheeses and the influence of P. candidum was stronger in lactic cheeses. The use of D. hansenii or R. infimominiatum, associated to P. candidum or G. candidum, did not affect greatly the sensorial properties of goat cheeses. Principal Component Analysis of the sensorial data obtained for soft goat cheeses ripened with various yeasts and moulds (lactic and Camembert type).6 Correlation plot (left) and plot of the cheese samples with the corresponding ripening strains (right). The results summarized in this paper illustrate how the typical flavour of goat milk products can be controlled. Such technological means could be highly useful for producers involved in goat milk transformation, helping them meet emerging and specific consumer’s needs and add value to goat milk products.