Appearance and colour as indicators for the evaluation of intrinsic and extrinsic factors driving consumer preference for fresh meat

The quality of meat derives from a very complex production system that leads to a high variability between fresh meat samples, related to the biological variability of animals, the rearing technique and the slaughtering procedures. These factors affect the quality of meat perceived by consumers. The perception of quality, that includes information, choice of purchase and consumption, defines the overall satisfaction for a product, and the decision of future purchase. Consumers become the jury of the meat quality, so the whole industry wants to meet their needs, in order to induce consumer to buy the products. Understanding the consumer needs is the critical point of meat industry that allows comprehending the meaning of meat quality from the consumer point of view and, thus, the development of a more attractive and high quality product provided with all the information required by the consumers. Appearance and colour are the first aspects that consumers evaluate at the purchase, and they seem to be the main drivers of meat choice. But, since those aspects are subjective it is important to answer those questions: how do consumers perceive appearance and colour? What do appearance and colour communicate to consumers? Furthermore, colour and appearance seem to be indicators of several intrinsic characteristics of meat, i.e. the correct decrease of pH after slaughtering, the type of animals, the freshness and wholesome of meat. So, could consumers appreciate by themselves these properties or should the industry communicate them to consumers? Is it possible to predict the eating quality evaluating appearance and colour of fresh meat? These are the questions, I tried to answer with this work. The PhD project was developed to identify how appearance and colour could affect the judgment of consumer liking and how appearance and colour could be related and used as predictors of eating properties of meat. Two experiments were conducted to understand consumer behaviour, especially the influence of familiarity on liking. Other two experiments were developed to assess the relationship between appearance and eating quality of meat. A synthesis of each experiment is reported below. The objective of the first work was to compare consumer liking and perception of beef quality attributes as a function of their familiarity and involvement with fresh meat. Ninety-three meat consumers were classified on the basis of their familiarity with fresh meats. Socio-demographic differences between the clusters were found to relate to gender and age, and high familiarity (HF) consumers showed higher involvement with meat. HF consumers enjoyed consuming meat, and they associated a symbolic value to it. In addition, their liking ratings were higher than those of low familiarity (LF) consumers for both appearance and taste of three specific types of fresh meat over the course of product shelf-life. The perceived risks associated with meat consumption and product choice were similar between groups. Both consumer segments reported that the most important driver of fresh meat purchase is its appearance, while the role of extrinsic cues differed among the groups. The HF group needed more information when choosing meat. Regardless of familiarity level, liking was consistent with beef appearance as affected by storage, but the prediction of experienced sensory quality lacked consistency when the perceived intrinsic cue variation was not associated with meat freshness. The aim of the second study was to define a quality meat grid system based on butcher’s appearance judgment as a tool to predict beef eating quality of Italian Simmental (IS) beef. Longissimus thoracis muscle (LT) of IS steak between the 8th - 9th ribs were evaluated. First, experts developed a grid system for the evaluation of the steaks quality, then the quality of 29 IS steaks was evaluated. A trained panel performed a quantitative descriptive analysis of LT from the same 29 carcasses. The quality index identified two levels of beef quality: standard and high. Results showed that the differences in terms of quality highlighted by experts were also reflected in the cooked meat, when evaluated by a trained panel. It seems that the developed quality index is a helpful tool to valorise the IS meat, guaranteeing the eating quality of beef. The third experiment aimed at assessing liking and preference for capretto and chevon as a function of consumer familiarity with goat meat. Five meats were produced: traditional milk capretto (MC), heavy summer capretto (HSC), summering, fall and late fall chevon. HSC was the most tender meat, having less cooking losses than both MC and redder chevon types. The instrumental profile corresponded with the appearance and texture attributes perceived by panelists. With aging of kids, meat lost its milk aroma (MC) and sweet taste (HSC) and acquired an increasing intensity of goat flavour and livery notes, partially related to feeding regime and fatty acid profile. Providing heavier carcasses outside the peaks of festival demand is promising for the goat industry, because a niche market preferred chevon over capretto. While the cluster of consumers who were unfamiliar with chevon showed a decrease in pleasantness when tasting chevon, the familiar group reduced their ratings only for meat from the oldest kids. The aim of the forth experiment was to measure the sensory variability of varying types of cattle meat, in terms of age and gender, and to evaluate the prediction power of visible spectrum analysis in defining the sensory proprieties. 24 lots of cattle meats, veal (V), beef from young bull (B) and heifer (H), were considered. Quantitative descriptive analysis (QDA), due by a trained panel, and reflectance spectra of visible space were recorded. The data set was divided into calibration set (18 samples) and validation set (6 samples) for external validation. PLS-1 was performed for each sensory attribute, in order to estimate the best prediction model, considering the coefficient of prediction Q2. In the second step, the performance of the models was evaluated using the validation set, considering R2p, coefficient of determination of prediction, RMSEP, root mean square error of prediction and SEP, standard error of prediction. The visible spectra showed very good prediction capacity (Q2 >0.7) for some sensory attributes: colour hue, beef flavour, whey flavour, coarseness. These results were confirmed in the validation phase. Juiciness also was a predictable attribute, but limitations in the validation phase has been detected. As expected, tenderness is not related to the visible spectra. The full Visible light spectrum is needed in order to obtain a good regression model for flavour attributes and coarseness. Other sensory attributes, as juiciness and colour hue, gave a better prediction model when some specific wavelengths were chosen. Concluding, the differences in consumers’ involvement and familiarity for meat were reflected in their choice, preference and liking, when products with different sensory proprieties were presented. Appearance and colour resulted important attributes that would drive the choice of product. However, more and reliable information are needed by consumer, considering their insecurity at purchase. Information about sensory profile of meat could be a way to differentiate the meat market and could help consumer when they have to buy fresh meat. Expert guide and judgments, as well as visible spectra, are good predictors of the intensity of some sensory attributes.