Segmentation of Food Consumers Based on Their Sustainable Attitude

The proposed study aims to segment consumers based on a sustainable approach to the consumption of food. The shift in consumers’ attitudes towards more balanced food consumption can be one of the sustainability drivers for entire food chains and may result in more sustained energy usage in the whole food chain and implementation of farm to fork strategy to the practice. We considered consumers’ attitudes as a multidimensional construct. Under this assumption, we asked respondents a series of questions related to the cognitive, behavioral, and affective components of an attitude. Data were collected from a market survey run among 433 consumers. We identiﬁed three consumer segments. The “Doers” segment exhibits sustainable behavior to a greater extent than the others. At the same time, they have less knowledge about the concept of food sustainability while the affective dimension was developed on an average level. The “Conscious” segment had well-developed cognitive and affective dimensions (which might indicate their openness to the information about sustainability positive feelings), however, it was not reﬂected in their behavior. Finally, the “Reluctant” segment, did not show a sustainable attitude towards food consumption in any of the analyzed dimensions. Answering the question of how common sustainable attitudes are may help in determining the market potential and in developing product and promotion strategies.


Introduction
It is hard to disagree that to combat climate changes the current consumption patterns must be altered globally [1][2][3][4][5][6][7][8][9]. The process of transformation of consumption plays an important role in facing environmental challenges, both globally and locally, and the sense of responsibility for the society, future generations, and the planet [10][11][12][13][14][15][16][17][18][19][20]. The Sustainable Development Goal (SDG) 12 Ensure sustainable consumption and production patterns defined by the European Commission have become a focal point of action until 2030 (Agenda 2030) [21]. This is reflected in the published strategy Farm to Fork [22,23] and the newest edition of the report "The State of Food Security and Nutrition in the World 2020" [24], where it was underscored that consumption patterns conforming to sustainable food consumption (SFC) will be playing a key role in achieving sustainable development, climate goals and satisfying the needs of the ever-growing population. Such processes are crucial nowadays as food markets are assessed as unbalanced. Production of raw materials and food processing requires a high energy input, creating negative side effects, such as greenhouse gas emission, food losses, and environmental burden.
The socio-economic transformation brought about qualitative and quantitative changes in food consumption. Other than by reducing the quantity of used goods and services, SFC can be also achieved by creating more appropriate consumption patterns, all with respect to the basic life needs and aspirations to improve the lives of current and future get markets by international corporations with macro-criteria (such as economic prosperity, political and cultural system, infrastructure), which started to be used for international market segmentation purposes [65,66] segmentation of groups of potential customers, from abroad.
The identification of consumer segments based on balancing their attitudes towards food consumption is of fundamental importance for the further development of SFC, especially in the practical aspect. Any efforts aiming to alter unsustainable attitudes and behaviors of consumers should be focused on learning and understanding ways to influence consumers to change their behaviors towards more sustainable food, positioning sustainable food, creating effective strategies and information campaigns [7,26,46,56,62,67,68]. The knowledge of consumer profiles allows the creation of a more efficient information and education policy, maintained both by the government and NGOs.
The purpose of this paper is to segment consumers based on a sustainable approach to the consumption of food. The prospects for a more sustainable consumption depend on the ability to improve the innovativeness level of consumers, enterprises, and science and of government and public institutions. Undoubtedly, the creation of a favorable environment that promotes development of consumption sustainability requires a closer cooperation between the academic environment, industry, government, and social organizations. In the context of the topic addressed in this paper, it is important to know consumer attitudes towards sustainable food consumption. The consumer segmentation presented in this paper narrows that gap.

Review of Studies on Consumer Segmentation with Regard to SFC
The review of studies on consumer segmentation in relation to the concept of SFC has been prepared as the result of searching databases of scientific publications. The search covered the articles made accessible from 1990 until March 2020 in Web of Science and Scopus. Publications were selected if the searched phrases appeared in one of the following elements: thematic sections, keywords, titles, and abstracts. The following phrases were selected for searching: "food", "consumption", "segmentation", "segment", followed by "sustainable consumption" and "cluster analysis".
The search of the Scopus base yielded 24 records, while the Web of Science database returned 13 records. When assessing the usefulness of publications using Moher's et al. [69] method, twenty-seven papers published after 2010 were identified (Figure 1).  business, green sustainable science technology, and environmental engineering. Analyzing the findings by country, it was found that a significant part of them were from Europe, especially Italy, Germany, and the Netherlands. The studies within the analyzed area have been performed, among others, by: Wageningen University and Research Centre (4 publications); Parthenope University of Naples (2); Gent University (2); Aarthus University (2); University of Gottingen (2). The articles have been analyzed and evaluated in terms of usefulness by two, independent scientific institutions (UPP and UD). To be included in the project, an article had to be prepared based on original data. The details of every publication have been identified by the first researcher, then verified by the second researcher and presented in the collective list in Table 1. Any publications that failed to satisfy these criteria have been excluded from analysis. FRL segment 1 ("Everyday food providers"); FRL segment 2 ("Food ignoramuses"); FRL segment 3 ("Enthusiastic food consumers"); FRL segment 4 ("Uninvolved food consumers"); FRL segment 5 ("Traditional family oriented food consumers"). Cluster 1 ("Pro-selves"); Cluster 2 ("Average consumers"); Cluster 3 ("Conscious consumers"). 8 Stranieri, S., Baldi, L.   Cluster 1 ("Meat-and vegetable-based meal consumers"); Cluster 2 ("Milk and meat consumers"); Cluster 3 ("Average consumers"); Cluster 4 ("Fruit, vegetable and dairy product consumers"); Cluster 5 ("Bread and bakery product consumers"); Cluster 6 ("No milk and dairy product consumers"). 11 Annunziata  Cluster 1 ("Eco-friendly, intensive farming supports"); Cluster 2 ("Indifferent"); Cluster 3 ("Sustainable farming supports"); Cluster 4 ("Ambiguous").

Socio-Demographics
Socio-demographics have been used in most of the cited works. This type of variable is one of the commonly used variables in consumer behavior studies [7,73,81]; they are Energies 2021, 14, 3179 7 of 28 easy to measure and strongly determine consumer's behaviors, which in turn makes them work well as a basis of segmentation.
Gender of consumers was considered in 21 papers (Table 2). For instance, Gen Z segments of the U.S. sustainable food market differ with respect to gender. In addition, Verain et al. [7] stated that "unsustainers" and "product-oriented" consumers were more often male than female compared with "curtailers". Analyzing two sub-samples characterized by different age: "millennials" (18-35 years) and "non-millennials" (36-88 years) revealed that gender is significant and positive only in the category of "non-millennials", while the income level is significant and positive for "millennials". Sogari et al. [81] who ran a study on the vine market noted that "if in the past wine was seen as mainly a male beverage, in recent years more and more females are becoming wine aficionados". Significant gender differences across segments were shown by Su et al. [89], who analyzed consumer groups based on environmental consciousness. Table 2. Socio-demographic factors used in the analyzed papers.
Education of consumers was considered in 21 of the analyzed papers. Seven articles showed that education significantly diversifies the consumers' behaviors and attitudes [68,70,72,74,75,79,88]. Vanhonacker et al. [68] found that the large majority of the "Conscious", "Active" and "Unwilling" consumers were higher educated, while a more balanced distribution in education level was found among the "Ignorant" and the "Uncertain". Ghvanidze et al. [88] stated that the "Apathetic" segment had high levels of education and income.

Environmental Sustainability
Categories Environmental Sustainability Items Numbers Refer to the Studies List in Table 1 Footprint related Self-evaluation of personal footprint 1 Personal relevance of the ecological footprint 1 Ecological footprint of food consumption (ecological footprint (gha) = quantity consumed per year per person (kg) × ecological footprint intensity (gha/kg)) 10 Certificates Sustainable certification 3 Sustainability labels on products 18 Attitude to using sustainability labels: (1) Propensity to read labels before consuming food products, (2) Degree of knowledge regarding sustainability labels. Two of the discussed publications focused on the selection and consumption of food and the related impact on the environment measured by the "ecological footprint" [68,76]. Vanhonacker et al. [68] show that numerous consumers fail to see the impact of animal production on the natural environment. They also point to the presence of alternative behaviors in relation to conventional meat consumption, for which eating habits and cultural patterns must be adapted. However, the readiness to pay higher prices is significantly lower than the readiness to consume. Mózner [76], however, has noted that consumers who ate more fruit, vegetables, and milk products did not have a smaller ecological footprint in terms of the entire food consumption.
Three publications examined the issue of food certification [70,77,82]. The used certification systems differ in terms of information value, methods of assigning and conducting monitoring activities. The impact of the origin certificate on consumers has been confirmed Energies 2021, 14, 3179 9 of 28 by Bronnmann and Asche [70], who showed that consumers were more willing to pay a higher price and were more eager to make purchases, all in relation to wild fish. Similar results confirming the readiness to pay a higher price have been also obtained by Aprile and Mariani [77], Janßen and Langen [82], and La Lama et al. [79].
The consumers' focus on environmental protection issues have been also considered in the studies by Jakubowska and Radzymińska [87] and by La Lama et al. [79]. Consumers can be divided into ones expressing strong pro-environment attitudes (and health-related values) and ones that do not take environmental aspects into account when selecting food [87]. Similar results have been obtained by Su et al. [89] who stated that the sustainable food market can be segmented according to environmental awareness. Sogari et al. [81] have identified a connection between the selection of a product and the consumer's engagement in the purchasing process and their environmental protection awareness.

Behavioral Factors
Behavioral changes have been used in 19 out of 27 articles [7,26,47,62,68,70,72,74,76,78,[80][81][82][84][85][86]88,91]. An overview of examples of the factors that have been included in the segmentation studies of the cited authors can be found in Table 4. The behavioral factors consisted mostly of buying behavior and general food choice motives/attribute importance by food category sustainability, general/life attitude, consumption habits, occasional behavior, ethical issues, consumer preferences for product information related to environmental issues. Table 4. Behavioral factors used in the analyzed papers. Table 1 Buying behavior   Table 1 Ethical issues Diet, agreement regarding the dietary guideline 7, 15 Produced in full respect of human rights, without exploiting women and children 11

Behavioral Categories Behavioral Items Numbers Refer to the Studies List in
Ethical concern about production processes 24 A few of the issues used for the purposes of establishing consumer segmentation are their behaviors related to purchasing ecological food products and their willingness to pay the price. For example, considering the willingness to pay measures for the respective attributes, it has been noted that the buyers of ecological food were more inclined to pay a premium for ecological products [47,62]. Vanhonacker et al. [68] and Palmieri and Forleo [26] have confirmed that consumers were more inclined to buy products perceived as more sustainable. Consumers exhibiting a more sustainable approach are also more willing to buy innovative products.
An important aspect is also the identification of factors behind the selection of food, with health and quality attributes playing an important role [62]. Risius et al. [85], who segmented fish consumers, emphasized the major importance of the country of origin. Lots of attention was also paid to connecting consumer values with purchasing behaviors. An example could be the attention paid by humans to the natural environment or the wellbeing of animals [93]. Verain et al. [7] found that food involvement and personal norms with regard to healthy and sustainable food positively affect sustainable food choices. Annunziata and Mariani [62] emphasized that the importance of ethical values (i.e., that food was produced with complete observance of human rights or with no abuse to women and children) and environmental values (e.g., that food was produced in an uncontaminated environment, in an eco-friendly manner and with support for local farmers).
The behaviors are related to the level of the consumers' awareness about sustainability, as confirmed by Prokeinová and Paluchová [78] with regard to the younger consumers who choose environmentally friendly and socially acceptable products more willingly than their parents.

Psychographic Factors
Psychographic variables have been used as a basis for segmentation in three of the analyzed papers [7,75,86]. Verain et al. [7] have confirmed that social and personal norms, ability, subjective knowledge, and food involvement significantly differentiate meat consumers' behaviors. For example, "product-oriented consumers differ from each other in that the product-oriented attach more importance to social norms and have a higher ability to judge sustainably produced food, and subjective knowledge on sustainable products". Van Huy et al. [75] have focused, among other aspects, on the attention to healthy food, love of cooking, convenience, and love of local and organic food. Wang and Somogyi [86] have examined the level of acceptability of crustaceans from sustainable production among Chinese consumers. They have found that personal standards significantly influence the purchase intentions of consumers.

Economic Factors
A wide range of economic variables, such as net household income per month, financial status of household, and employment status, were used as segmentation criteria or profiling variables. For example, Verain et al. [7] have identified four consumer segments that differ significantly regarding economic factors, e.g., segments named "Curtailers" have lower incomes compared to the "Sustainers" segment. The study by Van Huy et al. [75] has highlighted a nexus between the identified segments and the level of income of the consumers. Yildirim and Candan [93] have concluded that most green product buyers were at a high level of income. These results show that a more prominent presence of sustainable attitudes can be expected from the wealthier consumers.
The employment status has been included in three articles, and only in one study, significant differences across segments have been found by Lavelle et al. [72], who stated that respondents differ in their uptake of occasional and habitual pro-environmental behaviors. Considerable differences exist between the two behavior groups with regard to key socio-demographic variables, such as employment status and income, as well as residential location and housing tenure.

Affective Factors
Two of the presented articles considered the affective factors. In one of them, Krystallis [91] has taken into account consumers' assessments related to environmental factors, the perception of processing and well-being of animals in the case of processed meat. On the other hand, Hasanzade et al. [84] have analyzed the connection between ethical product attributes (e.g., ethical criteria of animal welfare, environmental protection, and labor and human rights) with the selected elements of behaviors.

Lifestyle and Values Factors
As shown by Aydin and Ünal [93], the lifestyle of consumers is related to the sustainability of consumption behaviors. In the publications of Thøgersen [71] and Van Huy et al., [75] the food-related lifestyle (FRL) model developed by Grunert et al. [94] has been used. For example, Thøgersen [71] investigates how the country of residence and FRL interact in shaping (un)sustainable food consumption patterns. The analysis has revealed that the outcome variables vary significantly across FRL segments. Further, after controlling for FRL, the direct effect of country class is highly significant for meatless suppers and marginally significant for buying organic food but non-significant for eating beef and food innovativeness. Van Huy et al. [75] have stated that the FRL model enables better understanding of how consumers employ food and its culinary aspects to achieve certain values in their lives. The acceptance of a specific lifestyle by consumers plays an important role in creating a "green economy" [95].
Consumer value-related approach represents a study by Hölker et al. [80] which developed consumer segmentation based on the human-animal relationship. The segmentation has been prepared according to specific values, such as animal rights, utilitarianism, new contractarian approach, abolitionism, original anthropocentrism, and anthropocentrism with indirect duties. At the same time, Yildirim and Candan [83] have found that green product buyers could be segmented into subgroups according to their personal values and consumption values. Significant differences regarding personal values, especially related to environmental matters, have been confirmed by Jakubowska and Radzymińska [87] as well.
The presented content proves that there are studies on consumer attitudes and behaviors in relation to SFC. Most of the researchers have focused on capturing individual factors or groups of factors allowing the identification of symptoms of sustainable behav-iors of consumers. The researchers focused on environmental aspects (food choices and practices beneficial to the environment). It is also understood that other factors can also impact attitudes, which are basic determinants of human behavior [96]. Some of the studies have specified the attitude of consumers towards food categories (e.g., organic products), while some other towards products (e.g., fish, seaweed, lettuce, wine). The authors of all studies considered the behavioral component and many studies also accounted for the attitude's affective component [26,[71][72][73][74][75]79,[86][87][88][89]. It is worth noting that accounting for consumer attitudes in three dimensions was an infrequently taken approach, which forms a theoretical basis for the considerations in our article. A similar perspective has also been applied by, for instance, Jakubowska and Radzymińska [87], La Lama et al. [79], Palmieri and Forleo [26], Thøgersen [71], Van Huy et al. [75], Vanhonacker et al. [68].

Methodological Approach
The article is based on results of the authors' own surveys conducted by means of direct interviews. The study aims to identify sustainable consumer behaviors with regard to food consumption.
The general methodological approach is presented in Figure 2. A plan for collecting the data was developed, with the assumption that the goal would be to distinguish consumer segments by the level of sustainability of food consumption. The plan covered: developing a method of selecting consumers for the study and organizing the study itself, namely training interviewers and determining the time frame of the study. Afterward, a questionnaire was developed and put to test, ultimately serving to collect the research material. Once the data had been collected, the raw material was subjected to formal and substantive evaluation. The prepared material was then subjected to clustering by means of a non-hierarchical clustering algorithm. The distinguished segments were then described. The general methodological approach is presented in Figure 2. A plan for collecting the data was developed, with the assumption that the goal would be to distinguish consumer segments by the level of sustainability of food consumption. The plan covered: developing a method of selecting consumers for the study and organizing the study itself, namely training interviewers and determining the time frame of the study. Afterward, a questionnaire was developed and put to test, ultimately serving to collect the research material. Once the data had been collected, the raw material was subjected to formal and substantive evaluation. The prepared material was then subjected to clustering by means of a non-hierarchical clustering algorithm. The distinguished segments were then described.

Selection of Respondents
The respondents selected for the study are adults, over 18 years of age, living in the region of Wielkopolska, Poland. The region was selected because: (1) the inhabitants of economically developed regions usually have higher education levels and higher revenues, etc., which means this concept is more widespread in such regions; (2) new directions in food consumption and nutrition usually spread among larger urban agglomerations, inhabitants of regions developed socially and economically, for whom they are certain, sought for a model of food consumption; (3) Wielkopolska is one of the best economically developed regions in Poland (in terms of gross GDP and growth, foreign capital involvement concentration and investment expenses, unemployment rate, revenues, education, and human capital potential) [97].
The quota sampling method was used for the selection of respondents for the study. Public statistics data served as a basis to determine the sample's structure that reflected the structure of Wielkopolska inhabitants in terms of age and gender.

Selection of Respondents
The respondents selected for the study are adults, over 18 years of age, living in the region of Wielkopolska, Poland. The region was selected because: (1) the inhabitants of economically developed regions usually have higher education levels and higher revenues, etc., which means this concept is more widespread in such regions; (2) new directions in food consumption and nutrition usually spread among larger urban agglomerations, inhabitants of regions developed socially and economically, for whom they are certain, sought for a model of food consumption; (3) Wielkopolska is one of the best economically developed regions in Poland (in terms of gross GDP and growth, foreign capital involve-ment concentration and investment expenses, unemployment rate, revenues, education, and human capital potential) [97].
The quota sampling method was used for the selection of respondents for the study. Public statistics data served as a basis to determine the sample's structure that reflected the structure of Wielkopolska inhabitants in terms of age and gender. The sample was 433 individuals. The social and demographic characteristics of the survey's participants is presented in Table 5. We live like average people-we have enough for daily needs, but must save for major purchases

44.57
We live well-we can afford a lot, without having to save for specific purposes 36.03 We live very well-we can afford certain luxuries 4.62

Questionnaire Development and Data Collection
Coming from the concept of the approach proposed by Breckler [98], which assumes three-dimensionality of attitudes towards SC, the questionnaire used in the survey has been divided into three parts. Each part contained questions related to the cognitive, behavioral, and affective components, respectively. Questions characterizing food consumption sustainability were asked within each area.
The first version of the questionnaire was subjected to pilot tests. The test consisted of 30 personal interviews with individuals who represented the target group. The purpose of the test was to check the questionnaire: in the formal aspect, namely whether the instructions and the formatting of the document are clear and facilitate the efficient work of the interviewers, -in the content aspect, i.e., identification of questions for which there is little variation in answers, making them impossible to use as a basis for segmentation or profiling variables.
After collecting remarks from interviewers who did pilot tests and following the analysis of respondents' answers, the necessary corrections were introduced, mainly related to the formulation of questions. The final questionnaire, used in the study, is the result of the introduction of these changes.
The face-to-face interviews for the study were performed between November and December 2019.

Data Cleaning
The collected material was subjected to post hoc data cleaning approach. The procedure of data cleaning assumes they are verified by the researchers (Exploratory Data Analysis) [99] and consisted of checking outliers in order to identify any errors that would occur during uploading or digitizing the data; checking the presence of any missing values and replacing them with a median, if found; and checking for any duplicated records in the database and removing such duplicates, if found. The cleaned database was subjected to statistical analysis.

Data Analysis-Clustering
We used a k-means cluster analysis (non-hierarchical clustering algorithm), which groups objects on a set of user-selected characteristics. The resulting clusters should exhibit high internal (within-cluster) homogeneity and high external (between-cluster) heterogeneity. Thus, if the classification is successful, objects within clusters will be close together when plotted geometrically and different clusters will be far apart. Cluster analysis was chosen because of the data reduction procedure, which is done objectively by reducing the information from an entire population or sample to information about specific groups, as a large number of observations are meaningless unless classified into manageable groups [100].
Based on selected questions within each of the SC dimensions (cognitive, behavioral, and affective), 3 indices were created, by transforming all variables into dichotomic ones and aggregating sums of answers. There are two types of variables: dichotomic (1/0)-no transformation was done; -ordinal (scales 1-3, 1-5 or 1-7)-top box answers were recoded into 1.
Because of the different number of variables used for each dimension, it was necessary to make sure we worked with roughly equal amplitudes within each index. Cluster analysis is sensitive to different scales, as it uses distance measures, therefore by using a mean and standard deviation values of each section, we standardized them with z-score transformation (z = (x − µ)/σ) [101], to eliminate the impact of larger standard deviation.

Description of Dimensions
As the three-dimensional model of attitudes was employed, each dimension was reflected by the number of questions in the questionnaire (Table 6). For the cognitive dimension, 6 questions (22 variables in the data) were used to measure and to differentiate respondents by their awareness of the sustainable consumption concept. The behavioral dimension covers 4 questions, representing 28 variables in the data. In the affective dimension, we used 2 questions, based on which 17 variables in the data were created.   -"Po stronie natury"/English: "On the nature side byŻywiec Zdrój" (Żywiec Zdrój) -"Zadbaj ośrodowisko"/English: "Take care about the environment" (Carlsberg Polska) -"Danone dlaśrodowiska"/English: Danone for the environment (Danone) -"Tworzenie wspólnej wartości"/"Creating the value together" (Nestle) -"Życie w sposób zrównoważony"/"Sustanaible life" (Unilever) -"Kto nie przyniesie-odpada"/Who will not bring-will be dismissed (Coca-Cola HBC Polska, Tesco Polska and Carlsberg Polska) 1 point if yes (12) The following is a list of various statements and each one of them has potential benefits and inconveniences for you. Consider every approach and decide if it is ultimately BENEFICIAL or DISADVANTAGEOUS for you.
-I buy local products, but spend more time shopping -I support local economy by consuming local products, but pay more for them -I buy food in smaller portions, but use more packaging -By not shopping "to stockpile for later" I lose the benefit of promotional prices for certain food products -I improve the quality of consumed food, but at the same time I reduce the amount of consumed food -Avoiding excessive, disadvantageous consumption requires self-control and self-discipline -I am ready to pay more for higher quality products -

Clustering
When running the analysis, the number of clusters was specified as 2, 3 and 4, to compare them and find the best solution. As the 2-cluster solution gave the result of data division into negative values of cognitive, behavioral, and evaluative indices for one cluster and positive for another one, that does not provide an insightful interpretation. As we needed to choose a better solution from two options, i.e., 3-clusters and 4-clusters, we run a validation procedure and assessed differences of cognitive, behavioral, affective indexes between clusters. Considering that in the 4-cluster solution there are more pairs with an insignificant difference, we decided to choose a 3-cluster solution for further interpretation.
The relative centers of each cluster on a standardized scale, are presented in Table 7, and reflect differences between the clusters. The first cluster has the highest values of behavioral aspect, which means that this group represents the most sustainable pattern of behavior. The second one is the only one with positive values for cognitive aspects, so respondents with the highest awareness level. The third group includes respondents with the lowest values for each aspect. The ANOVA analysis allows determination of the importance of each index with the relative weight. For the 3 indices, F values are large and all of them are significant (p < 0.001), indicating that they have a significant impact on the results for the 3-cluster solution ( Table 8). To validate the analysis, we created cluster membership and compare, whether mean values of each cluster are significantly different within each index. All 3 clusters are significantly different, so we can reject the null hypothesis, that the group means are all equal in all indices: To sum up, there are 3 clusters that describe and differentiate all observations: -Cluster 1-low cognitive, high behavioral, high affective; -Cluster 2-high cognitive, low behavioral, high affective; -Cluster 3-low cognitive, low behavioral, low affective.

Results
As a description of the results, we will first present a synthetic description of segments, followed by the characteristics of their representatives within three attitude dimensions. We identified three segments (clusters) and based on the dominant characteristics named them as: cluster 1-"Doers", cluster 2-"Conscious" and cluster 3-"Reluctant" their characteristics can be presented on the basis of the data in Table 9.  Cluster 1-segment "Doers" This segment included 31% of food consumers. These are action-oriented people, whose consumption-oriented behaviors show a higher number of sustainable behaviors than in the case of other people, despite the fact that their knowledge about sustainable consumption is low and their opinions about this concept are moderately positive. This means that the reasons behind their (sustainable) behaviors may come from other areas, e.g., the socialization or upbringing process, material situation, living environment, etc. This segment shows a significantly higher percentage of people living in rural areas. The members of this segment have diversified incomes, whose distribution is similar to the one in the examined sample. These people are slightly older than the average for the examined group.
Cluster 2-segment "Conscious" This segment included 30% of food consumers. The individuals in this segment have a high value of indexes within the cognitive and affective dimensions of attitudes, but a noticeably lower intensity of sustainable behaviors. These consumers have a better understanding of and a more positive attitude towards the sustainability concept but have not yet introduced a higher number of sustainable behaviors. However, their higher awareness may be considered a good foundation to grow into a fully sustainable attitude. This segment has a higher representation of women. It is also a segment with a major share of young people and the highest share of individuals with higher education in the identified segments (which may determine the higher level of awareness). The group's higher level of income and satisfaction from material status should be also underscored.
Cluster 3-segment "Reluctant" This segment included 39% of food consumers. This group consists of individuals with negative attitudes towards the idea of sustainable consumption, manifested both in the low awareness of concepts, negative opinions of them, and low importance of sustainable behaviors. The group is nearly equally represented by men and women, with age distribution similar to the distribution in the examined group. The representatives of this segment may be encountered in towns of various sizes. The profiling data show that this segment is more often populated by individuals with lower education, whose material situation is poor (they declare lower satisfaction with their incomes and that their incomes are lower).
Segment characteristics in the cognitive sphere Significant differences can be spotted among segments in the cognitive sphere. The individuals in the "Conscious" segment had a high awareness of the terms used for describing sustainable consumption behaviors. In total, 87% of its representatives have met with the term "sustainable consumption". In the case of "Doers" and "Reluctant" segments, this was 36% and 19%, respectively. We can also observe a high activity of the "Conscious" segment individuals in the search for information about food and a high amount of the retained information. The "Reluctant" segment individuals declare looking for information about food in the least degree (17% of the segment's representatives). Additionally, they pay little attention to information about consumption sustainability.
Segment characteristics in the behavioral sphere The readiness to introduce changes in nutrition is mainly declared by individuals in the "Doers" and "Conscious" segments (67% and 70%, respectively)-however, these changes are of a different nature. The representatives of the "Doers" segment refer to the sustainability concept to a higher degree (e.g., increasing the share of fruit and vegetables in their diets, preparing meals by cooking them themselves). They also more frequently declared limiting undesirable behaviors such as eating fried meals and meals with a highfat content, consuming sugar, salt, and stimulants. The "Doers" representatives limited meat consumption to the highest degree. The members of the "Conscious" segment would introduce similar changes, but they were declared with lower frequency. This segment dominates only in terms of introducing healthy and organic products to the diet. The nutritional changes were the least common in the "Reluctant" segment, declared by around 1/3 of the segment's participants. Moreover, they were related to the sustainable consumption principles to a lower degree than in the remaining segments.
The individuals in the "Does" and "Conscious" segments more often declared changes related to sourcing food for their households. The use of large retail facilities (such as supermarkets and hypermarkets) grew among the "Doers". This could be related to the fact that half of the members of this segment live outside towns, which is where structural changes to retail have been happening in the recent years, leading to changes in purchasing models. Different behaviors are represented by the representatives of the "Conscious" and "Reluctant" segments, both very similar to each other in terms of the place of living. The "Conscious" segment was more eager to use smaller, specialized shops and marketplaces. On the other hand, the "Reluctant" segment preferred large-size retailer and increased the importance of online shopping for food.
The "Does" and "Conscious" segments are changing their behaviors related to sourcing food in the direction convergent with sustainability principles, but they are doing it differently. For example, the "Doers" more frequently declare limiting wastage. For the "Conscious" segment, however, making food products autonomously and purchasing low-processed products are more typical.
Segment characteristics in the affective sphere Certain similarities emerge between the "Doers" and the "Conscious" segments, and a major difference in relation to the "Reluctant" segment. The two first segments have a positive attitude towards sustainable consumption principles and appreciate the impact of such behaviors on the local communities, environment, and local economy. Still, the individuals in the "Conscious" segment are firmer in this regard than the "Doers". For example, they have a higher propensity to pay more for organic products or spend extra time on sustainable behaviors, e.g., searching for information on nutrition principles. The individuals in the "Reluctant" segment, on the other hand, assess various signs of sustainable consumption as disadvantageous for them.
Contrary to our expectations, not all consumers are interested in food and sustainable consumption. The "Doers" are interested in food for various reasons, which fail to cover the issue of sustainable consumption. There may be an impression that this topic is disregarded by such individuals. The "Conscious" segment consisted of consumers actively looking for information about food and nutrition, sensitive to the issues of sustainable consumption.
Consumers have varying awareness on consumption sustainability (behavioral aspect), buying and eating behaviors reinforcing responsibility towards the planet and future generations (behavioral aspect), the perception of promotional messages related to the balancing of food product choices, and the subjective evaluation of benefits for consumers resulting from sustainable behaviors (affective aspect). The study is one of numerous papers about profiling food consumers based on a three-element attitude dimension in relation to SFC.

Cognitive Dimension
Our studies have revealed that the information about environmental, economic, and social consequences of excessive consumption are factors that foster the popularization of sustainable attitudes. The force of various media should be used to reduce adverse behaviors [16,75,102] by building SC awareness and affecting consumption behaviors, as supported by e.g., Hasanzade et al. [84], Prokeinová and Paluchová [78], Sogri et al. [81], Wang and Somogyi [86], Verain et al. [73]. A better understanding of the characteristics of consumers via segmentation facilitates the preparation of a more effective communication strategy. Public institutions, food producers, and commerce should take steps leading to a better understanding of food and nutrition, potential benefits for the environment, responsibility for the planet and future generations. The existence of a connection between the type of consumers and susceptibility to messages communicated, for example, via social media has been indicated by Sogari et al. [81] claiming that "the greater the importance the consumer places on the product/process dimension of environmental sustainability, the higher the self-selection in market segments". The researchers also pointed to the need to increase the possibilities of communicating the activities of enterprises in environmental protection. Hasanzade et al. [84] have added to the literary references the differentiation of consumer segments due to behaviors resulting from the way they react to messages, additionally showing the need to carefully select information about products (e.g., the product's ethical character). However, not all studies confirm the need to differentiate communication efforts. According to Verain et al. [73] there exist messages that can have a universal character and reach all consumer groups.

Behavioral Dimension
Labelling products as environmentally friendly and promoting "new food" may be helpful in reinforcing sustainable consumer choices [74,85,103]. The results of our studies show that marking food products as healthy and organic may be important for the "Conscious" segment. This segment is similar to the "Trendsetters" segment that appeared in the study by Van Huy et al. [75].
Nevertheless, our study fails to provide a detailed insight into the issue of a sustainable diet, which is discussed, for example, by Verain et al. [7,73] who have confirmed that consumer segments are differentiated by the approach to health and eating healthy, organic food products. The interest in healthy eating has been also noted by Van Huy et al. [75], who noted that the interest in healthy nutrition is positively related to organic, locally sourced food.
Verain et al. [73] have proved that individuals oriented toward sustainable growth have been limiting the consumption of meat after learning about the benefits this approach provides both to their health and the environment. This is also confirmed by our study since the inclination to eat meat has been used to differentiate the segments. Meat consumption has been limited to the highest degree by the "Doers", closely trailed by the "Conscious" segment.
The results of our study show a relation between the place of residence of consumers and their sustainable attitudes. It seems that people living in rural areas have "natural" sustainable attitudes (mainly in the behavioral dimension), which may be related to the specific characteristic of the rural environment as a place for living and working. La Lama et al. [79] also point to the role played by the place of residence of consumers, additionally pointing to other factors such as income, lifestyle, access to frequently updated information, and telecommunication technologies.
Vermeir et al. [18] have stated that "many consumers express environmental concern but do not consistently act on it. That is, consumer attitudes toward environmental sustainability are mainly positive, but there is a notable gap between favorable disposition and actual purchase of sustainable food products, i.e., the attitude-behavior gap". Similar conclusions can be formulated on the basis of our segmentation-the "Conscious" segment shows a well-developed cognitive and affective dimension of the attitude, with a less developed behavioral dimension.

Affective Dimension
The affective dimension proves the consumers' positive approach towards the concept and principles of sustainable consumption and their commitment to the idea. The results we have obtained show that this is fostered by the consumers' higher material status, although this is not the only condition. The "Conscious" segment is quicker to see health benefits coming from SC than the "Doers". This segment shows similarities to consumers of cluster 3 in the research conducted by Verain et al. [73], who pay close attention to SC attributes. In the segmentation done by La Lama et al. [79] one of the segments is "Skeptical". It consists of consumers with an egocentric approach and little contact with modern food production practices, who are reluctant to pay higher prices for welfarefriendly products. The segment is similar to our "Reluctant" segment, showing a negative attitude towards the sustainability concept. Groups of consumers with a negative attitude towards SC have a significant impact on the development of the market of products covered by this concept. The research conducted by Gerini et al. [47] has shown that consumers may exhibit positive attitudes towards sustainable products, while being reluctant to pay higher prices for them. If most consumers are not willing to pay more for, e.g., organic products, whose production is more costly, such products will not be able to capture a higher share of the market. The growth of the market for sustainable products and services should be supported by good access to information [27,47,104,105]. The factor supporting the growth of this market may include references to consumer ethnocentrism and localism, as discussed, for example, by Van Huy et al. [75] on the basis of studies conducted among consumers in Vietnam, and to personal responsibility [47].

Conclusions
Conclusions in two areas may be formulated on the basis of our work. The first conclusion applies to the review of papers on food consumer segmentation based on the sustainability concept. The second one covers the segmentation and characterization of such segments in the region of Wielkopolska, Poland.
It may be concluded, on the basis of the literature review, that eight categories of variables are used for the purposes of segmentation of consumers with regard to food consumption. These were: environmental sustainability, socio-demographic, psychographic, economic, behavioral, affective factors, lifestyle, and consumers' values. The factors related to the environment and consumption behaviors should be considered the dominating category. The multi-faceted nature of consumer attitudes leads researchers to use a wide range of variables related to lifestyle, values, preferences, and consumers' attitude towards sustainable food consumption and environmental issues. Given the significant diversity of the used factors, it may be also observed that individual papers most usually put emphasis on one of the dimensions of the attitude. The theoretical foundation of the approach to segmentation proposed by us is the concept of a three-element attitude structure, thanks to which factors that may be indicators of sustainability can be approached comprehensively. The results we obtained prove that such an approach is efficient in the segmentation of consumers.
The consumers' segmentation model proposed in that paper contributes to the knowledge about consumers' sustainable behavior and might be used for further research development and as well as by practitioners and consumer policymakers. Identified segments represent a different potential for adoption of sustainable behaviors what implies the necessity of implementing various methods of promoting the idea of sustainability among them.
Two of the identified segments ("Doers" and "Conscious") represent a certain escalation of sustainable attitudes, and the characteristics of these segments show that they may grow in the future. The representatives of the third segment ("Reluctant") are negatively oriented towards the sustainability concept. Taking into account the distribution of the size of individual segments, it may be assumed that around 60% of consumers make up a group that may positively modify their attitudes towards sustainable consumption. Consequently, this group is the target of activities that may be taken to support the development of sustainable attitudes. A universal set of activities promoting sustainable food consumption attitudes should cover activities in two main areas. The first one would be to build consumer awareness in relation to the concepts of sustainability and sustainable consumption, along with the derived benefits. This area should mainly impact the "Doers"-thanks to the reinforcement of the consumers' knowledge it will be possible to achieve the effect of further intensification of behaviors. The second area would be activities incentivizing to take sustainable activities, on the basis of the already developed consumer awareness.
Understanding and recognizing consumers' attitudes and behaviors is useful for industrial practitioners and decision-makers making efforts to transition into more sustainable food systems. Information and communication strategies should be built upon full knowledge about food consumers from a given region, taking into consideration the three-dimensional nature of their attitudes. By adapting the content of messages to the profiles of specific consumer segments, emphasis should be applied to informing about benefits coming from the consumption of sustainable food, in order to motivate the sensitive segments and to raise the consumers' awareness about the benefits stemming from the pursuit of the sustainable food consumption model.

Limitations and Recommendations for Further Research
A conclusion can be drawn from our literary research that the food consumption segmentation proposition presented in the article is one of very few propositions that apply to the concept of making food consumption attitudes sustainable and is based on a three-element attitude concept. Maintaining such an approach in the future will make result comparison possible, leading to a better understanding of the consumers.
Our study is limited by a relatively low sample size and its regional reach, which implicates the need to continues similar studies, but in pan-regional and international scope. It is also worth expanding the future studies with qualitative elements, for a better understanding of mechanisms of shaping sustainable attitudes.
The knowledge of signs of sustainable consumption and factors that determine it is still emerging and needs intensive studies. It would be beneficial to direct such future studies at methods of combating any factors that hinder the popularization of sustainable attitudes. Examples of valuable directions would be the impact of retail structure on making food consumption behaviors sustainable, or the characteristics of one's living environment as a factor determining the sustainability of attitudes. Additionally, depending on the type of food considered by consumers, their attitudes can be differentiated.