Aquaculture 544 (2021) 737137 Contents lists available at ScienceDirect Aquaculture journal homepage: www.elsevier.com/locate/aquaculture Information impact on consumers' perceptions towards aquaculture: Dismantling the myth about feeds for farmed fish M. Carrassón a, *, 1, A. Soler-Membrives a, 1, 1, M. Constenla a, C. Escobar b, R. Flos c, J.M. Gil b, c, V. Luzón d, F. Piferrer e, L. Reig b, c a Departament de Biologia Animal, de Biologia Vegetal i d'Ecologia, Universitat Autònoma de Barcelona, Cerdanyola del Vallès, 08193 Barcelona, Spain b CREDA-UPC-IRTA, Center for Agro-Food Economics and Development, C/Esteve Terrades 8, Castelldefels, 08660 Barcelona, Spain c Departament d'Enginyeria Agroalimentària i Biotecnologia, Universitat Politècnica de Catalunya, C/Esteve Terrades 8, Castelldefels, 08660 Barcelona, Spain d Departament de Comunicació Audiovisual i Publicitat, Universitat Autònoma de Barcelona, Cerdanyola del Vallès, 08193 Barcelona, Spain e Institut de Ciències del Mar (ICM-CSIC), Pg. Marítim de la Barceloneta 37-49, 08003 Barcelona, Spain A R T I C L E I N F O A B S T R A C T Keywords: Aquaculture products are commonplace in markets around the world. However, despite efforts to minimize the Seafood consumption negative perceptions towards aquaculture, several misbeliefs or myths still persist, and thus globally consumers Aquaculture tend to value wild fish more highly than farmed fish . The lack of information has been shown to be one of the Perception most important causes of this preference, driving buying decisions to be more emotional than rational. The aim Sociodemographic factors Knowledge about aquaculture of this study was to determine whether scientific-supported information contrasting one myth could contribute to Information tools a better perception of farmed products. To that end, consensus on a series of aquaculture-related issues among different scientists, external experts, and aquaculture societies was used to build up the scientific information. This information was provided to 300 Spanish consumers using two different communication tools (150 con­ sumers each tool): an interactive web documentary and a written and printed document, to detect possible differences in the change of consumers' perception. Consumers were asked for their degree of agreement on a set of 14 statements before and after providing the scientific information. A variable collecting the assessment of each of the statements was calculated as the Overall-perception. Possible significant differences between the scores before and after providing the information and for the ‘overall perception’ were analysed separately for each communication tool as well as for the combined sample. Possible relationship between the consumers' perception with the sociodemographic factors, the consumers' knowledge and the fish consumption habits were also assessed. Results show that consumer's perception of aquaculture before the query were moderate (5.6 average in a 0 to 10 scale) but that it increased slightly but significantly and regardless of the communication tool used. Among sociodemographic factors, age and gender were the ones that most influenced consumer's per­ ceptions, being older people those who exhibited a generally more positive opinion towards aquaculture. The effects of consumption habits and knowledge about aquaculture were also the two most explicative factors for change in perception. Importantly, the opinion of consumers with less knowledge about seafood products in general and production methods or consuming only wild fish products, improved after being exposed to the information. These results demonstrate the utility of science- and fact-based communication campaigns to improve the societal perception of aquaculture practices and products, regardless of the tool used to transmit this information. 1. Introduction while aquaculture production has increased significantly (FAO, 2020; Pauly et al., 2003; Pauly et al., 2002). These factors allow consumers to Worldwide demand for fish products is continuously increasing. have access to aquaculture products in the marketplace at reasonable Over the last decade, fish from capture fisheries has remained constant prices. Thus, globally, from 1990 to 2018, there was a rise in * Corresponding author. E-mail address: Maite.Carrasson@uab.cat (M. Carrassón). 1 contributed equally towards this study https://doi.org/10.1016/j.aquaculture.2021.737137 Received 13 December 2020; Received in revised form 24 April 2021; Accepted 3 July 2021 Available online 6 July 2021 0044-8486/© 2021 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). M. Carrassón et al. Aquaculture 544 (2021) 737137 aquaculture production of 527%, and currently aquaculture provides of the process. about half of the total fish supply (FAO, 2020). Aquaculture products are Aquaculture is constantly changing, so when consumers try to find now routinely consumed in Spain and many other countries around the information about its products and processes, it can be quite difficult world (about the 20% of the aquatic consumption products in the Eu­ both to find up-to-date, friendly, and scientifically sound material and, ropean Union come from aquaculture, MAPAMA, 2018) with different at the same time, to avoid an excess of information, not always well species and formats. In fact Spain is an important seafood trader, the first focused or with the appropriate scientific rigor. One interesting remark European Union country producer of aquaculture in terms of biomass, is that, after receiving information, the perceptions and opinions of and the third country in terms of value income, mainly thanks to the consumers can change very quickly depending on the trust in the source production of bivalves (APROMAR, 2020). However, despite the posi­ and, therefore, it is not surprising that opinion changes can arise even tive changes in environmental and welfare aspects of aquaculture, which after a simple discussion between pairs lasting only several minutes is a consequence of the sector evolution and its increasing awareness for (Reig et al., 2019). sustainability issues, a lack of favourable perception still remains among In aquaculture, the emergence of widely held but false beliefs or many consumers (Reig et al., 2019). Indeed, there is an overall lack of ideas (referred to as “myths” in the rest of this paper) has been described awareness regarding aquaculture products and their potential safety and concerning various topics such as feed (Ayvaz, 2017; Hardy, 2005), quality characteristics (Kaimakoudi et al., 2013). genetic manipulation, or the use of therapeutic products (Bergh, 2007). When farmed products are evaluated, they are rarely considered on Some of these myths condition consumer's perception of aquaculture their own since comparison with wild specimens is very often un­ and its products (Claret et al., 2014; Fernandez-Polanco and Luna, 2012; avoidable. In this scenario, there is a clear preference for the wild fish, Reig et al., 2019). In the work of Reig et al. (2019), positive aspects which is more valued, as evidenced for example by their higher price in detected included market issues and the stable quality of farmed prod­ the markets. When comparing wild and farmed fish prices from different ucts. On the other hand, negative perceptions included the lack of suf­ species at the Central Fish Market in Barcelona (MERCABARNA, 2019) ficient information about aquaculture and quality. Animal welfare and the differences are very important. Aspects such as fish size or supply environmental impact issues were not of much concern, although they volume, among others, might partially explain market price differences included some minor positive and negative perceptions. Some of these between wild and farmed fish. Thus, for example, Regnier and Bayr­ negative aspects were identified as erroneous, matching several of the amoglu (2017) found that for wild and farmed European seabass in so-called myths about aquaculture. The need for communication cam­ France there was not a homogeneous or integrated market, suggesting paigns addressing these issues was a relevant conclusion of that study. that consumers are more sensitive to aquaculture processes in high For the present paper, the hypothesis is that providing scientifically valued species, while wild and farmed sea bream markets were partially sound but accessible information might be key to clarify several myths integrated. In the same sense, in Spain, the production method (capture or wrong beliefs about aquaculture. This means taking a step forward vs. aquaculture) was an important factor for consumers when choosing towards start providing solutions for a problem that has already been sea fish (Claret et al., 2014). very well described: the different valuation of the aquaculture products The difference in perception by wild and farm fish has been by the consumer. described by different authors, who demonstrate that consumers tend to The myth chosen to dispel by providing information was one of those value more the wild than the farmed fish (Bacher, 2015; Claret et al., identified as being relevant in the study of Reig et al. (2019) and was 2014; López-Mas et al., 2021; Reig et al., 2019; Vanhonacker et al., related to the feeding of the farmed fish. It was formulated as follows: 2013; Verbeke et al., 2007) even if in a blind tasting panel they cannot the feeding that the fish receive in the fish farms can have detrimental differentiate between both origins (Claret et al., 2016). In general, Eu­ consequences for the quality of the product, the fish and the consumers' ropean consumers weight scientific concerns about environment or health, and the environment. As can be seen, fish feeding encompasses human health related to farmed and wild fish, but they consider the many different elements about aquaculture production, so dispelling latter more natural and thus they have a preference for it (Schlag and this myth through scientific-based information (e.g., Kok et al., 2020) Ystgaard, 2013). Several studies have attempted to determine the at­ can provide positive guidance on several aspects of aquaculture (health, tributes that consumers tend to value in farmed fish. These aspects are environment, sustainability, among others that are at the base of other related to the environmental impact of the production, as well as myths). concern for fish welfare (Aarset et al., 2004; Feucht and Zander, 2016), a These myths can be dismantled by providing valid objective argu­ lower level of pollution, heavy metals, and parasites (Claret et al., 2014; ments through documentation. The importance of information is a López-Mas et al., 2021), better control of the entire production process determinant factor of individual behavior. Rather than simply memo­ and better price and availability (López-Mas et al., 2021), and more rizing the content and answering the questions, the consumers read and stable quality (Reig et al., 2019), among others. According to market understand the relevance of the information and in turn the relevance of statistics (MERCABARNA, 2019) even though consumers prefer wild their decisions on fish consumption. Thus, providing certain scientific fish, the consumption of farmed fish is higher. Therefore, increasing the based information focused on the main benefits of aquaculture or the knowledge of consumers about aquaculture in several aspects such as minimisation of the negative effects is crucial to tackle those myths. safety, health or sustainability, could maintain and increase the positive Moreover, besides providing the right information, it is important to image and consumption of farmed fish itself (Altintzoglou et al., 2010). make it available to all consumers utilizing a friendly language. When Moreover, a set of negative attributes are also assigned to farmed fish considering how this information should be provided, two aspects have (e.g., attributes related to animal welfare, quality and environment to be taken into account. First, the possible influence of the socio- impact; see Reig et al. (2019)), but one common element that frequently economic profile of respondents and, secondly, the influence of the arises is a demand for more information about aquaculture processes communication tool used to provide the information. Factors like age and products. In Reig et al. (2019), the highest concordance in negative (Verbeke et al., 2005), gender, and education (Pieniak et al., 2010; aspects included the lack of sufficient information about aquaculture Verbeke and Vackier, 2005) can influence fish consumption and and the quality of farmed products. Therefore, there is not only a dis­ perception and even affect buying decisions (Kraus et al., 2017). So, it is torted and frequently negative idea about several aspects of aquaculture important to consider such socio-economic features in any consumer but also a perception of needing more information. According to Pieniak perception research. Considering the tool used to provide the informa­ et al. (2007), one of the consequences of this lack of information is that tion, when comparing web-based and paper-based questionnaires in buying decisions can be made more emotionally than rationally. This previous surveys, the first produced information of similar or superior makes these decisions more voluble and fragile, as they do not consider quality than the traditional paper version (Vergnaud et al., 2011). In other objective aspects, like the knowledge of the product or the quality another research investigating three methods, paper-based, computer- 2 M. Carrassón et al. Aquaculture 544 (2021) 737137 based, and web-based (Touvier et al., 2010) few differences were Additionally, two communication tools were implemented to test if observed despite some evidence in favor of paper-based questionnaire there was a difference depending on the format in which information administration. was provided. Based on what has been discussed above, the main objective of this Providing a powerful information tool could add value to the sector, study was to determine whether consumer's perceptions of aquaculture contributing to improving its image among consumers (and increasing and, specifically, on the myth that the feed fish receive can have detri­ the value of its products). This is, furthermore, a key element that can mental consequences for the animal, the consumers, and the environ­ contribute to the sustainable development of aquaculture. ment could be dispelled by providing correct information. To this, we developed and launched a questionnaire in which we assessed con­ 2. Materials and methods sumers' perceptions towards aquaculture processes and products. This was done in two stages: before and after providing them with scientific The whole process included different steps, from selecting and ana­ information about the selected myth, i.e. the feeding of farmed fish. lysing the myth to the final output (Fig. 1). Fig. 1. A flow diagram that summarizes the process of the survey during the different steps and outputs obtained in each step. 3 M. Carrassón et al. Aquaculture 544 (2021) 737137 2.1. Generating the information to dispel the myth: a consensus among Table 1 researchers Survey technical sheet. Universe frame Usual consumers of fish within 20–70 years old The information to be provided to consumers focused on feed for Scope Spain farmed fish, although it also contained other aspects related to aqua­ Sample size 300 consumers culture, such as animal welfare and traceability. First, an extended Sample error ± 5.66% literature review was compiled and written in a scientifically compre­ Confidence level 95% (k = 1.96) hensive language, including all the scientific references (to justify its Sample design Stratified by gender and place of residence Control measures Pilot survey (14 consumers), repeated three times contents). The building up of a first draft included several rounds of Field work dates July and August 2017 discussion within the authors of this study, covering different expertise: aquaculture production, fish nutrition, and health, genetics, economy, and communication. Each expert wrote a first draft of one section of the The procedure was structured in several blocks, in the following document, including all scientific references, and was later circulated order: (A) gathering and recording consumers' knowledge and fish and discussed in face-to-face meetings where suggestions were debated consumption habits, (B) gathering and recording consumers' beliefs and at large. The draft was then rewritten and a consensus was reached when opinions towards aquaculture, (C) display of the communication tool to all members of the team approved it. This document was then sent to one consumers, (D) repetition of bloc B, and (E) gathering and recording expert of each of the following fields and organizations: Aquaculture information on consumers' sociodemographic characteristics. It is Spanish society, European Aquaculture Society, Spanish producers as­ important to note, that once consumers had read the information, they sociation, university scientist, aquaculture research center scientist, feed were not allowed to go back to the questionnaire to check their first set company, food safety). Their knowledge and opinions were taken into of answers (before displaying the information). account and helped to improve the final document's content. Items from blocks A and E are displayed on Table 2. They collect On a second step, the team wrote a synthesis of the previous docu­ sociodemographic, consumption and knowledge data which are gener­ ment choosing proper but easy-to-understand words and eliminating ally gathered in consumer studies. Blocks B and D are shown on Sup­ scientific references as well as footnotes, to simplify the reading. Special plementary Material S2. The 14-items set collect the most relevant care was taken to not change the conceptual content of the message. myths identified from Reig et al. (2019), among others. The 14 state­ Again, a consensus was reached within the authors' team by approval of ments were classified according to their character: personal opinions (O) the document. A list of questions was then decided and the answers about aquaculture and, respondents' opinions about what society (S) written with contents of the synthesis document, maintaining the facts, knows or feels about aquaculture practices and products (Table 3, factor but using a language adapted keeping in mind that the target population “type”). Participants' were asked about their agreement with each was the consumers. The aim was to communicate true, science-based, statement measured by an 11-point Likert scale, where 0 and 10 mean and clear messages that could be conveyed to society. null and full agreement, respectively, with the statement related to the From the final document, two formats were produced, by the com­ selected myth (the feeding for farmed fish). Block C provided the munications expert, each adapted to a different type of communication communication tool. A fraction of its contents are shown in the Sup­ tool: (1) a written and printed document; and (2) an interactive web plementary Material S1 of this manuscript. documentary. Both tools shared the same information and phrasing. The sociodemographic factors (block E of the questionnaire) However, the web documentary allowed the readers to play a more included gender, age, area of residence, yearly gross household income, active role, by including photographic material in a way that made it and family structure. For the statistical analysis, gender was divided into more amiable to read (Supplementary Material S1B). Also, the written (0) male and (1) female. Age was categorized into three groups: (1) 22 to document accounted for the whole information, all in once. Instead, the 33 years old, (2) 34 to 53 years old and, (3) 54 to 72 years old. Age web documentary provided a first highlighted information and the groups were based on current demographic trends in Spain and were reader could ask for further detailed information. The photographic chosen to represent individuals of three different generations. The area material to be included was agreed upon after discussion among the of residence was divided into inland (0) and coastal (1) provinces (ter­ authors of this study. Both documents included the identity of the au­ ritorial divisions in Spain), depending whether the province has a sea thors' affiliation (universities and research centers) to provide context to coast or not. At the time when the survey was carried out, in Spain the respondents when receiving the questionnaire. average household income per capita was ~14,000 euro and average household size was ~2,5 (INE, 2020). This is why we took 35 k euro as 2.2. Empirical application the middle average class in total household income and thus household income was categorized into three groups: (1) low (< 35.000 €), (2) The questionnaire was tested first a total of three times using a pilot medium (~35.000 €) and, (3) high (> 35.000 €). Family structure was sample of 14 different consumers each time and subsequently revised to segmented into three groups: (1) independent youngsters and adults, (2) improve readability and understanding. From our pilot survey, adjust­ single-parent families and couples with children and, (3) couples ments were made to the vocabulary and the structure of some of the without children (Table 2). sentences. Besides, some technical issues from the online questionnaire Fish consumption habits and consumers' knowledge factors (block A also needed to be rearranged. For example, the speed in which the screen of the questionnaire) included: weekly consumption of seafood products would jump from page to page, a clarification of how to return to the (number of times per week that seafood is consumed) [wconsump]; questionnaire after the information was read, among others of the kind. weekly expenditure on seafood products (euros spent per week on sea­ Once it was deemed suitable and polished, it was presented to 300 food) [wexpend]; diversity of species consumed (number of different Spanish consumers (“the sample”) qualified by having purchased and seafood species usually consumed) [divers]; consumption of wild or consumed seafood products at least once in the last 2 months. The farmed fish or both [aqconsump]; consumers' knowledge about the fieldwork was subcontracted to a company specialized in marketing definition of aquaculture [def]; consumers' knowledge about the pro­ research and the sample was stratified by gender and place of residence. duction method of the seafood that they consume [knowofmethod], and; The 300 participants were randomly assigned to one of the two consumers' knowledge about seafood products in general [consknow]. communication tools (150 each) while assuring the stratification within For the statistical analysis, wconsump was divided into (1) once a week, each group. Each group visualized one of the designs (written document (2) 2 times a week, (3) 3 times a week, (4) 4 times a week, and (5) 5 or or web documentary). Table 1 shows a summary of the technical sheet of more times a week. Wexpend was segmented in (1) ≤ 10 €, (2) from 11 to the survey. 20 € and (3) from 21 to 30 € and (4) > 30 €. Divers was classified as (1) 4 M. Carrassón et al. Aquaculture 544 (2021) 737137 Table 2 low (1 to 3 species from a list of 10 species including the most consumed Sociodemographic and Consumption and knowledge characteristics of the in Spain - based on sales statistics from (MERCABARNA, 2019) (2) consumers participating in the present study (corresponding to the blocks E and medium (4 to 7 species), and (3) high (8 to 10 species). Aqconsump was A from the procedure survey). Data as percent. Consumers were grouped into divided into (1) only wild, (2) wild and aquaculture, and (3) only two communication tools: information provided on a written document and aquaculture. Def classified participants into two groups: (0) wrong information provided on an interactive web documentary. * indicates statisti­ response and (1) right response (this information was derived from a cally significant differences Chi-squared test between the two tool. multiple-choice question concerning aquaculture's definition with one Factors Written Web Combined correct definition and three incorrect ones). Knowofmethod and con­ document (n documentary (n (n = 300) sknow were both elicited from a self-assessment 11-point Likert scale = 152) = 148) where 0 indicated no knowledge whatsoever and 11 deep knowledge. SOCIODEMOGRAPHIC Then we classified participants in three discrete classes depending on (BLOCK E) Gender grades: (1) low (from 0 to 3 points), (2) medium (from 4 to 6 points), and Male 46.1 50.7 48.3 (3) high (from 7 to 10 points) (Table 2). Female 53.9 49.3 51.7 Age 2.3. Data analysis 22 to 33 years old 10.5 20.3 15.3 34 to 53 years old 78.3 68.9 73.7 54 to 72 years old 11.2 10.8 11.0 Chi-square tests were applied to determine possible differences in Yearly gross Household sociodemographic and the consumption and knowledge factors between Income* the consumers presented with the two communication tools - (1) in­ Above 35,000€ 40.1 27.0 33.7 formation provided on a written document and, (2) information pro­ Around 35,000€ 32.9 34.5 33.7 Below 35,000€ 27.0 38.5 32.7 vided on an interactive web documentary. A non-parametric Family structure multidimensional scaling (MDS) was applied on square-root trans­ Couples or single parent 68.0 59.9 64.0 formed first response (before being exposed to information) data of 14 with children statements set using the factor “type” (O, S, see Table 3) to visualize the Couples with no children 24.7 29.6 27.1 Independent adults/ 7.3 10.6 8.9 ordination of 14 statements concerning the similarity of the character. youngsters The consumer's responses obtained for the 14 statements before and Area of residence after providing the information are the core of our statistical analysis. Mainland 49.3 48.6 49.0 For the analysis, to make proper comparisons between the 14 different Coastal 50.7 51.4 51.0 statements, some of them needed to be positivized (particularly, state­ CONSUMPTION AND KNOWLEDGE (BLOCK A) ments #1, 4, 5, 6, 7, and 9, shown in Table 3) since these were in fact Weekly consumption of reverse-worded sentences. seafood products Data obtained from the 14 statements sets were tested for normality [wconsump] and homoscedasticity. A variable named ‘Overall-perception’ was Once a week 9.7 9.9 9.8 2 times a week 29.2 37.6 33.3 calculated as the mean of the responses obtained for the 14 statements 3 times a week 23.6 27.7 25.6 sets, before and after providing the information, and provides con­ 4 times a week 22.2 12.8 17.5 sumer's overall perception towards aquaculture. Mean and standard 5 or more times a week 15.3 12.1 13.7 deviation were calculated for each statement and for the Overall Weekly expenditure on perception (both before and after reading the information provided). seafood products [wexpend]* ≤ 10 € 18.5 27.4 22.9 Possible significant differences between the scores before and after from 11 to 20 € 39.7 42.5 41.1 reading the 14 statements and for the Overall perception were analysed from 21 to 30 € 18.5 19.2 18.8 by the paired t-test, for each one of the two communication tools as well > 30 € 23.3 11.0 17.1 as for the combined sample. Furthermore, the change (c) in each one of Diversity of species consumed [divers] the 14 statements for both the two tools as well as the combined sample Low (1 to 3 species) 21.1 29.1 25.0 was calculated as the difference between the response before and after Medium (4 to 7 species) 67.8 60.8 64.3 being presented with information (Change-variables: from c1 to c14, High (8 to 10 species) 11.2 10.1 10.7 Table 4). An Overall-change variable was also calculated as the mean of Consumption of wild or farmed the values obtained for the 14 Change-variables. Possible significant fish or both [aqconsump] Only wild 1.3 2.7 2.0 differences for c1 to c14 and the overall-change variable separately in Both 92.10 87.8 90.0 two sets according to the communication tool provided (web docu­ Only aquaculture 6.6 9.5 8.0 mentary or written document) were analysed by respective t-tests. Consumers' definition of To analyze the main drives for change of opinion and taking into aquaculture [def] Right response 87.5 85.1 86.3 account the structure of the survey, we considered sociodemographic Wrong response 12.5 14.9 13.7 factors and consumers' consumption habits and knowledge factors. Consumers' knowledge about Redundancy analysis (RDA) (van den Wollenberg, 1977) was applied to the production method test the relation between sociodemographic factors, as multiple expli­ [knowofmethod] cative variables, and the 14 statements set before reading the informa­ Low (from 0 to 3 points) 23.7 30.4 27.0 Medium (from 4 to 6 points) 34.2 30.4 32.3 tion, as multiple response variables. The analysis was also applied to the High (form 7 to 10 points) 42.1 39.2 40.7 14 Change-variables. The statistical significance of each of the axes Consumers' knowledge about derived by each analysis was tested with a Monte Carlo permutation test seafood products in general (Hope, 1968). The number of permutations was set at 500. Pearson [consknow] correlations were used to explore the relationship among these vari­ Low (from 0 to 3 points) 18.4 17.6 18.0 Medium (from 4 to 6 points) 42.8 40.5 41.7 ables. Possible effects of the two important explicative factors in the High (form 7 to 10 points) 38.8 41.9 40.3 RDA (which turned out to be age and gender) were tested for the Overall perception and the Overall-change variable, each of the 14 statements (before reading the information), and; the 14 Change-variables (c1 to c14) using General Linear Models (GLM), with post-hoc pairwise 5 M. Carrassón et al. Aquaculture 544 (2021) 737137 Table 3 Influence of information provision on aquaculture perceptions. Mean (X) and standard deviation (SD) of the 14 statements (11-point Likart scale, 0–10) and the Overall perception (mean of the responses obtained for the 14-item sets), before and after reading the provided information. (R) refers to the reverse-worded statements that required to be positivized. Changes were analysed separately for those who read the written document, those who read the web documentary and the combined sample. Statement type indicates whether it concerns about the respondent's own opinion (O) or what they think society (S) knows or feels about aquaculture. Different superscript letters show significant differences (p < 0.05) among before and after reading information (paired T-test). Written document (n = 152) Web documentary (n = 148) Combined (n = 300) Before After Before After Before After Type X SD X SD X SD X SD X SD X SD 1. There are no chemical products in fish feeds that can be O 5.09 2.52 5.96 2.95 5.03a 2.35 6.01b 2.85 5.06 2.44 5.99 2.90 a b a b harmful to our health (R) 2. Aquaculture products are healthy O 6.87 1.85 7.32 1.91 6.86 1.92 7.50 1.74 6.87 1.88 7.41 1.83 a b a b a b 3. Feeds for farmed fish are healthy for the fish O 6.19 2.05 7.11 1.97 6.18 1.88 7.40 1.89 6.18 1.97 7.25 1.94 a b a b a b 4. The quality of farmed fish is unaffected by the feeds that they O 4.63 2.35 5.55 2.86 4.78 2.61 5.91 2.85 4.70 2.48 5.73 2.86 a b a b a b receive (R) 5. Resources are not wasted by the use of wild fish in farmed O 5.19 2.24 6.06 2.66 5.05 2.23 5.78 2.67 5.12 2.23 5.92 2.67 a b a b a b fish feeds (R) a 6. Society trusts aquaculture processes (feeding, medication, S 3.47 1.97 3.45 1.81 3.38 1.96 3.64 1.94 3.43 1.96 3.54 1.88 a a a a a treatments…) (R) 7. Society is informed about aquaculture processes (feeding, S 2.36 2.02 2.59 1.71 2.51 1.97 2.61 a 1.90 2.43 2.00 2.60 1.80 a a a a a medication, treatments…) (R) 8. There are more sanitary controls for farmed fish than for O 6.78 2.15 7.91 1.70 6.47 2.11 7.52 1.86 6.62 2.13 7.72 1.79 a b a b a b wild fish 9. Society tends to accept aquaculture products (R) S 4.28 2.25 3.95 2.29 4.41 2.29 3.76 2.14 4.34 2.27 3.86 2.22 a a a b a b 10. Fish farming is more environmentally friendly than other O 6.01 2.38 7.28 2.07 6.01 2.05 7.24 1.82 6.01 2.22 7.26 1.95 a b a b a b livestock sectors 11. Aquaculture products taste good O 6.47 2.11 7.32 1.85 6.72 1.75 7.41 1.67 6.59 1.94 7.36 1.76 a b a b a b 12. Aquaculture products have affordable prices O 7.01 1.73 7.70 1.48 6.98 1.83 7.28 1.67 7.00 1.78 7.49 1.59 a b a b a b 13. Aquaculture products are quality products O 6.62 1.94 7.48 1.82 6.72 1.87 7.62 1.59 6.67 1.90 7.55 1.71 a b a b a b 14. Aquaculture products provide essential nutrients for our O 7.05 1.89 7.49 2.05 7.19 1.85 7.89 1.65 7.12 1.87 7.69 1.87 a b a b a b health Overall perception – 5.57a 1.03 6.22 1.12 5.59 0.95 6.25 1.05 5.58 0.99 6.24 1.09 b a b a b comparisons. The analyses using the change of opinion were repeated 3.1. Effect of the information provided for the aggregated sample separately depending on the type of communication tool provided. Redundancy analysis (RDA) was also applied to test the relation Before providing the information, the obtained value for the Overall among the 14 Change-variables (c1 to c14 from Table 4) and the con­ perception was 5.58, which shows that the values of consumers' agree­ sumption and knowledge factors. Then, using individuals as replicate ment with the statements were moderate (Table 3). Responses by samples, differences in 14 Change-variables (c1 to c14) and the Overall- statements show that the three most positive considerations about change variable were tested using GLM, followed by post-hoc Student- aquaculture products are: providing essential nutrients, being healthy, Newman-Keuls test. This was done for the three factors that after the and having affordable prices (statements #14, 2, and 12, respectively). RDA turned out to be the most important: aqconsump, knowofmethod and Assessments showing the three lowest scores are those related to re­ consknow, plus the factor “type of communication tool”. spondents' opinions about what society knows or feels about aquacul­ Relative importance of the 14- statements sets (before the informa­ ture: trusting aquaculture processes, being informed about aquaculture tion) and the 14-Change variables were explored separately with processes and tending to accept aquaculture products (statements #6, 7, Random Forests (Strobl et al., 2008) using R Studio 1.1.463. The cforest and 9, respectively). () function was used to apply the method with subsampling without Fig. 2 shows a two-dimensional non-metric Multidimensional Scaling replacement to account for variables with varied scales of measurement (MDS) plot of the 14 statements with superimposed clusters resulting and number of categories. Mean decrease in accuracy was used to order from group average clustering, at similarity levels of 85%. It shows a variables by their importance. The analysis included all the factors used clear separation between respondents' personal opinions about aqua­ in both RDA. culture (O), and those about what society knows or feels in the field of aquaculture (S). Combined sample scores before reading the information 3. Results significantly differed (t-test: t = 49,162, p < 0.001) between personal opinions and society knowledge statements (6.18 ± 0.84 vs 3.40 ± 0.95, Possible differences in the characterization of consumers from the respectively; mean ± SD). two samples were tested. Only significant differences between the two The combined respondents' assessments for the 14 statements tools (written document and web documentary) were found for the (Table 3) showed a moderate but significant increase in agreement factors Yearly Gross Household Income (Х2 = 6.936, p = 0.031) and scores after reading the information for the Overall perception and for wexpend (Х2 = 9.154, p = 0.027) (Table 2). No differences were found 11 of the statements (1, 2, 3, 4, 5, 8, 10, 11, 12, 13 and 14 from Table 3) for the remaining 10 factors considered. (paired t-test, p < 0.05 in all cases). However, the assessment to state­ ment # 9 got significantly lower scores after reading the information while statements 6 and 7 did not change (p > 0.05). The three latter statements, nevertheless, do not relate to the consumer him/herself, but 6 M. Carrassón et al. Aquaculture 544 (2021) 737137 Table 4 Mean (X) and standard deviation (SD) of the change produced between the re­ sponses of 14 statements before and after reading the provided information and the Overall change variable. Changes were analysed separately for those who read the written document, those who read the web documentary and the combined sample. Written Web documentary (n Combined document = 148) sample (n = (n = 152) 300) X SD X SD X SD c1. There are no 0.88 3.04 0.98 3.20 0.93 3.11 chemical products in fish feeds that can be harmful to our health c2. Aquaculture 0.45 2.16 0.64 2.04 0.54 2.10 products are healthy c3. Feeds for farmed 0.92 2.33 1.22 2.14 1.07 2.24 fish are healthy for the fish Fig. 2. Multidimensional scaling (MDS) showing ordination of the first c4. The quality of 0.91 3.34 1.14 3.15 1.03 3.24 response of 14 statements depending of the character or type of the statement. farmed fish is S: statements related to the opinion of what respondents think society knows or unaffected by the feels; O: personal opinion of the respondents over the aquaculture product. feeds that they receive c5. Resources are not 0.87 2.91 0.73 2.85 0.80 2.88 3.2. Effect of the sociodemographic factors wasted by the use of wild fish in farmed Unless otherwise stated, no interaction was found between factors fish feeds c6. Society trusts − 0.03 2.09 0.26 2.48 0.11 2.29 tested (sociodemographic or consumption habits and knowledge) in any aquaculture of the above-described analyses (respective GLMs). processes (feeding, The RDA showed that 79.10% of the variability in consumer's re­ medication, sponses to the 14 statements before being provided with the information treatments…) c7. Society is informed 0.23 2.17 0.11 2.39 0.17 2.28 was explained by age, gender, family structure, yearly gross household about aquaculture income, and area of residence as explanatory variables (p < 0.05) (not processes (feeding, shown). The standardized canonical coefficients of age and gender were medication, the two more important for the two first axes. treatments…) There was a significant relationship between the Overall perception c8. There are more 1.13 2.25 1.05 2.15 1.09 2.20 sanitary controls for before reading the information and age (GLM, p = 0.027), being the farmed fish than for oldest adult group (range: 54–72 years) those which showed a more wild fish positive opinion towards aquaculture. However, only answers to two of c9. Society tends to − 0.33 2.53 − 0.64 2.30 − 0.48 2.42 the 14 statements were significantly influenced by age (GLM, p < 0.05 in accept aquaculture products both cases): statements #3 and 13 (feed safety in farmed fish and the c10. Fish farming is 1.26 2.74 1.24 2.42 1.25 2.58 quality of aquaculture products), with a more positive agreement again more from the oldest adult group. Regarding gender, men's scores were higher environmentally before reading the information in statements # 8 and 14 (more sanitary friendly than other controls for farmed vs wild fish and the essential nutrient provided by livestock sectors c11. Aquaculture 0.85 1.94 0.69 1.91 0.77 1.92 aquaculture products) (GLM, p < 0.05 in both cases). products taste good Regarding the change in perception after reading the scientific in­ c12. Aquaculture 0.68 1.59 0.30 1.97 0.50 1.79 formation, the RDA performed with the 14 Change-variables produced products have similar results (70.56%, p = 0.05), being again age and gender the two affordable prices c13. Aquaculture 0.86 1.79 0.90 1.97 0.88 1.88 more explicative factors in the analysis (Fig. 3). However, the GLMs products are quality indicate that there was a significant change produced after the infor­ products mation only for the gender factor in the overall change and in one c14. Aquaculture 0.44 2.06 0.70 1.72 0.57 1.90 statement (c1: no chemical products in fish feeds harming our health), products provide being women consumers the ones who improve more their opinion essential nutrients for our health (GLM, p < 0.05 in both cases) (Fig. 4A, Gender). Overall perception 0.65 1.00 0.67 1.00 0.66 1.00 3.3. Effect of consumption habits and knowledge factors to what the consumer thinks about how society thinks or feels. For the group that read the written document, statement #9 also did The RDA showed that 65.70% of the variability of the changes pro­ not change significantly (Table 3). These were the only differences duced after the information (Change-variables; c1-c14) was explained detected between the groups exposed to the two different communica­ by the participants' consumption habits and knowledge about seafood tion tools. Furthermore, the Overall-change variable and the 14 Change- and aquaculture as explicative factors (p < 0.05). Particularly, kno­ variables (c1 to c14) were not dependent on the type of communication wofmethod and consknow were the two most explicative factors, together tool received (t-test, p > 0.05 in all cases) (Table 4). with aqconsump (Fig. 5). Interactions between aqconsump and consknow were found for the Overall-change variable and for the statements c1 and c2 (no chemical products in fish feeds harming our health and the healthy of aquaculture 7 M. Carrassón et al. Aquaculture 544 (2021) 737137 only found for the statements c4 (feed does not affect the quality of farmed fish) and c12, pointing out that consumers with less knowledge about the production method [knowofmethod] change more their opinion when reading the information document (Fig. 4B, C). Besides, people with less knowledge about seafood products in general [con­ sknow] changed more their responses in Overall-change, c1, c2, and c12 (Fig. 4D, E, F, G). Similarly, people who consume only wild fish products [aqconsump] improved more their opinion in c1 and c5 (no chemical products in fish feeds harming our health and feed farmed fish does not waste wild fish resources) than those who consume aquaculture prod­ ucts (GLM, p < 0.05 in all cases) (Fig. 4H, I). In any case, there were no differences in the change of the response by the type of communication tool (GLM, p > 0.05 in all cases). In general terms, knowofmethod and consknow were also the factors that contributed more to explaining the variation in the responses before reading the provided information and on the change produced between Fig. 3. Redundancy Analysis (RDA) among the change of opinion in responses the responses before and after reading the information (Random For­ of survey after reading the materials provided (c1-c14) and the sociodemo­ ests). Both variables were the most listed in the top 1st and 2nd positions graphic factors (age, gender, family structure, income, region) as the explica­ at both analyses. tive variables. 4. Discussion products) (from the 14 Change variables). An interaction between kno­ wofmethod and consknow for the statement c12 (the affordable prices of The goal of this study was to test whether providing consumers with aquaculture products) also was detected. In general, the change in each scientifically-based information yet in a friendly manner could response (14 Change variables) did not differ for the factors aqconsump, contribute to an overall better perception of aquaculture. Additionally, knowofmethod, consknow, and type of document read. Differences were we also tested whether the medium by which the information was Fig. 4. Significant change in the level of consumer's agreement (0− 10) after being exposed to the information to specific statements according to different factors. Significant differences in the change of response for the statement c1 (A) between men and women; differences in the change of response for the statement c4 (B) and c12 (C) between consumers with low, medium or high knowledge about the production method [knowofmethod]; differences in the change of response for the Overall change (D), and the statements c1 (E), c2 (F) and c12 (G) between people with low, medium or high knowledge about seafood products [consknow]; dif­ ferences in the change of response for the statements c1 (H) and c5 (I) between people who consume wild, mix or farmed fish products [aqconsump]. Different letters show significant differences in the change-variables among groups. Among the GLM tested, only the ones that have significant relationships with the factors are shown. 8 M. Carrassón et al. Aquaculture 544 (2021) 737137 marketing, and that has been previously observed when exploring consumer's answers to survey statements (Graeff, 2003) or to the opinion on food issues (Fernbach et al., 2019). Such an effect could explain the observation that, regardless of the tool used, statement #9 (“Society tends to accept aquaculture products”), the most generalist and straightforward of the societal type of statements, was the only for which consumers reduced their agreement scores. Many factors influence the public perception of aquaculture pro­ cesses and products that do not only include objective knowledge but also preconceived ideas, attitudes, and beliefs (Bacher, 2015). In our study, among the sociodemographic factors, age and gender were the ones that had more influence on consumer's perceptions. Indeed, age is known to be one of the major factors affecting consumers' perceptions (Verbeke et al., 2005). However, the influence of that factor can vary among studies. Verbeke et al. (2007) noticed that the oldest respondents hold a stronger belief that wild fish are healthier than farmed fish. Instead, we found that Spanish consumers belonging to the oldest age Fig. 5. Redundancy Analysis (RDA) of seven factors of knowledge and con­ class (54–72 years) had an initial higher acceptance of aquaculture when sumption of aquaculture species related to the change produced between the compared to younger age classes. This does not mean that wild seafood responses before and after reading the provided information (c1-c14). Abbre­ is less values compared to farmed fish, but at least elucidates a quite viations for the factors: wconsump: weekly consumption of seafood products; good initial perception from elderns regarding aquaculture. Gender in­ wexpend: weekly expenditure on seafood products; divers: diversity of species consumed; aqconsump: consumption of wild or farmed fish or both; def: con­ fluences on consumer perceptions are widely reported, e.g., on Portu­ sumers' definition of aquaculture; knowofmethod: consumers' knowledge about guese seafood preferences (Cardoso et al., 2013). However, in our study the production method of the seafood that they consume; consknow: consumers' this influence was not conclusive, as men exhibited more positive beliefs knowledge about seafood products in general. than women only for the statements #3 and #14, while women con­ sumers improved more their opinion after reading the scientific presented (written document or web-based tool) could influence con­ information. sumer's possible changes in perception. As expected, our results clearly The results showing lack of major differences depending on the tool show a moderate but significant more positive opinion of consumers used to present the information are in line with results of studies that towards aquaculture and aquaculture products after being presented also tested the influence of the tool when retrieving consumer percep­ with proper information, independently to the tool provided. Some tions, sometimes showing a slight advantage of the web-based ap­ interesting insights provided by this study are discussed below. The proaches (Touvier et al., 2010; Vergnaud et al., 2011). In any case, the performance of a pilot survey, as was done in this paper, is highly lack of differences can be regarded as positive because this allows in advisable to avoid misinterpretations of the questions and adjust the future informative campaigns the possibility to exploit a diversity of language and the semantics, prior the final survey is launched. It also communication tools. It is important to note that the message was the ensures that the order and the flow of the questionnaire is correct. These same for the two communication tools used in the study. The lack of can be taken as the recommendations to researchers or institutions that significant differences between the two tools and a moderate but sig­ aim to contribute to debunking these myths. nificant improvement for the Overall perception of aquaculture after First, in general, there were no differences in any but two of the reading the information was observed, allowed us to validate the mes­ sociodemographic and consumption and knowledge variables between sage used, since it was the same for both communication tools. The fact the participants who used the text document and those that used the that the information provided comes from a group of experts and has web-based tool. The only significant difference was in the Yearly gross been validated by different reputables organization in the aquaculture household income and weekly expenditure on seafood products. Thus, sector may be a key factor responsible for that improvement in con­ about two-thirds of the consumers who read the written document had sumers' opinion, regardless the tool. an income around or above 35,000 € while around two-thirds of those Average scores before providing information to consumers (ranging who used the web documentary had an income around or below that 6.87–7.12 in the Combined sample) indicate a medium-to-high positive figure. This was related to the weekly expenditure on seafood products opinion to the statements. This is quite comparable to what was since consumers who read the written document tended to spend more observed in an study carried out in the West Coast of Sweden. Swedish on seafood than those that were exposed to the web documentary. respondents tended to be favourable though a majority chose neutral However, based on the results of the 14 statements, in which there is an responses to the query “how would you rate your general opinion to­ overall concordance, it seems that this difference in these two factors did ward aquaculture?” (Thomas et al., 2018). Given that Sweden is a not bear any influence on the overall results of our study. country of great and ancient national heritage in aquaculture, the The lower mean score of societal vs. personal statements (mean, SD: dominance of neutral responses (60% versus the 40% of positive re­ 3.40, 0.95 vs 6.18, 0.84, respectively) when assessing somebody else's sponses) is still surprising. However, it is worth noting that very few behavior concerning our own can be explained by consumers' social respondents expressed a bad or very bad opinion. So, accouting that in desirability. This phenomenon occurs because consumers tend to reply Spain consumers had a similar neutral-to-positive opinion as in Sweden, to personal opinions in the way they believe that it is correct or more aquaculture perception rates occupy a good position. Furthermore, agreeable to the researcher, providing their own responses with a other authors have reported a non negative opinion about aquaculture, certain bias (Bonsall and Shires, 2009; Hufnagel and Conca, 1994). It is occurring simultaneously with a more favourable opinion towards the also possible that greater knowledge of a subject, with a more critically wild product, among Spanish consumers (Claret et al., 2016). In fact informed position on it, can increase one's acceptance of the subject Fernández-Polanco and Luna (2010) suggested that in Spain both wild while projecting lower expectations of acceptance by others or that the and cultured seabream share a common image, more related to the opinion on one's knowledge is lower than what they think. This re­ species than to the origin, in such a way that both products become sembles the Dunning-Kruger effect (Kruger and Dunning, 1999) a complementary rather than substitutes. The country's great potential for cognitive bias in psychology that explains the success of content aquaculture reflected to the notable supply of farming marine species may point out that consumers have learned to check for the seafood 9 M. Carrassón et al. Aquaculture 544 (2021) 737137 origin and its traceability (Garza-Gil et al., 2016). Thus they may tend to two different tools for providing the information, indicating that associate some regions origin with a certain quality and food safety. educational campaigns can exploit to their benefit printed documents Still considering that there was not a notable previous prejudice to­ and web-based formats as complementary communication channels to wards aquaculture, and the limitation of the memory effect of the reach potentially larger audiences. We have identified the most impor­ participant when conducting the survey, information given to con­ tant sociodemographic factors to be taken into account. This informa­ sumers did improve their opinion. Regardless of whether the informa­ tion will be particularly useful in the design of future campaigns tailored tion was presented as a text document or was web-based, the results of at, for example, particular gender, age- or income-class groups. We have this study showed an increase in the agreement scores in all but two of determined that consumers with the lowest knowledge of seafood the 14 statements posed in the combined sample. The correlation be­ products and aquaculture improve more their opinion when the infor­ tween increased awareness with change in perceptions has been also mation is provided. Results are also consistent regardless of whether assessed in salmon farming (Wrigley, 2017). These results demonstrate they concern specific questions or when gauged globally. Thus, in this the utility of science-and-fact-based communication campaigns to in­ regard the results of this study are robust. This study has provided in­ crease consumers' awareness and information about food products. sights into the usefulness of information campaigns to enhance the so­ Moreover, potential social changes such as the incentivation of the cietal acceptance of aquaculture practices and products, on the medium sustainable development goals including the Blue Growth Economy to convey such information and also has provided a sort of baseline as a goals may accelerate consumer's opinion changes and gaining more reference point for further studies. adepts for aquaculture products. So, these results could be taken as an experimental indicators of policy-process learning to measure change in Author Contributions perception of aquaculture. The more we know the better we choose. It is interesting to note that we found a significant and inverse Maite Carrassón: Conceptualization, Data curation, Formal analysis, relationship between the initial score and the change in score. Among Investigation, Methodology, Project administration, Supervision, Vali­ consumption and knowledge factors, knowofmethod and consknow were dation, Visualization, Writing - original draft, Writing - review & edit­ the two most explicative factors, together with aqconsump. These results ing. Anna Soler-Membrives: Conceptualization, Data curation, Formal could somehow be anticipated and confirm the link between knowledge analysis, Investigation, Methodology, Supervision, Validation, Visuali­ on a particular food production method and the consequent acceptance zation, Writing - original draft, Writing - review & editing. Maria Con­ of the resulting product irrespective of its actual nature. Claret et al. stenla: Conceptualization, Data curation, Formal analysis, Methodology, (2016) found that people tend to value wild products more when they Validation, Visualization, Writing - review & editing. Cristina Escobar: know their precedence of origin, but perhaps without adequate knowl­ Conceptualization, Data curation, Methodology, Validation, Visualiza­ edge. Therefore, providing adequate information on aquaculture im­ tion, Writing - review & editing. Rosa Flos: Conceptualization, Project proves its image and therefore the acceptance of aquaculture products. administration, Supervision, Data curation, Visualization, Writing - Thus, for example, Bastian et al. (2015) in a survey of Australian wine original draft, Writing - review & editing. Jose M Gil: Conceptualization, consumers found that the use of winemaking additives, even commonly Data curation, Methodology, Validation, Visualization, Writing - review used and legally permitted additives such as tartaric acid, preservatives, & editing. Virginia Luzon: Conceptualization, Validation, Visualization, and tannins were considered far less acceptable by less knowledgeable Writing - review & editing. Francesc Piferrer: Conceptualization, consumers. This is thus relevant for the differences we found in the Investigation, Supervision, Validation, Visualization, Writing - original change of statement #4 (“The quality of the farmed fish in unaffected by draft, Writing - review & editing. Lourdes Reig: Conceptualization, the feeds that they receive”), pointing out that consumers with less Project administration, Supervision, Data curation, Visualization, knowledge about the production method [knowofmethod] change more Writing - original draft, Writing - review & editing. their opinion when reading the information document. Another inter­ esting insight is that consumers with the lowest knowledge on seafood Declaration of Competing Interest products, in general, were the ones that clearly more benefited from being provided with information, as shown in two particularly inter­ The authors declare that they have no known competing financial esting aspects for such type of consumers: informing them about the interests or personal relationships that could have appeared to influence healthiness and affordable prices of aquaculture products, an aspect that the work reported in this paper. has been deemed of crucial importance in previous studies on consumer perceptions on aquaculture (Dey et al., 2005). This indicates that the Acknowledgments lower the initial score the higher the increase in agreement after being exposed to information. In that regard, therefore, it could be said that We are grateful to three anonymous reviewers for providing their the present study already fulfilled the mission of promotional campaigns valuable comments and corrections on the manuscript The authors by at least contributing to the education of consumers with the lowest would like to thank the Reference Research Network in Aquaculture of knowledge of seafood products. Finally, a particularly illustrative the Generalitat of Catalonia (XRAq) for funding this research. Further­ insight is that consumers that usually opt for wild fish improved most more, we would like to thank all the experts from the Aquaculture their agreement with statement #1 (“There are no chemical products in Spanish society, European Aquaculture Society, Spanish producers as­ fish feeds that…”) and #5 (“Resources are not wasted by the use of wild sociation, universities, aquaculture research centers and feed companies fish in farmed fish feeds”) which, again, indicates the usefulness of in­ who participated in the study. formation campaigns to increase consumer's acceptance and thus con­ sumer's choice of new food products and in this particular case educating them on the fact that aquaculture is tending towards sustainability and Appendix A. Supplementary data concern for the environment. These results are in line with those of an study on the perception of salmon farming in Scotland, where it was Supplementary data to this article can be found online at https://doi. shown the priority that people attach to the environmental performance org/10.1016/j.aquaculture.2021.737137. of the salmon aquaculture industry, relative to other objectives (Whit­ marsh and Wattage, 2006). References In summary, we conclude that consumers significantly improve their Aarset, B., Beckmann, S., Bigne, E., Beveridge, M., Bjorndal, T., Bunting, J., perceptions towards aquaculture processes and products when exposed McDonagh, P., Mariojouls, C., Muir, J., Prothero, A., Reisch, L., Smith, A., to scientifically-based information. These results are consistent across Tveteras, R., Young, J., 2004. The European consumers’ understanding and 10 M. Carrassón et al. Aquaculture 544 (2021) 737137 perceptions of the “organic” food regime: the case of aquaculture. Br. Food J. 106 Kraus, A., Annunziata, A., Vecchio, R., 2017. Sociodemographic factors differentiating https://doi.org/10.1108/00070700410516784. the consumer and the motivations for functional food consumption. J. Am. Coll. Altintzoglou, T., Verbeke, W., Vanhonacker, F., Luten, J., 2010. The image of fish from Nutr. 36 https://doi.org/10.1080/07315724.2016.1228489. aquaculture among Europeans: impact of exposure to balanced information. Kruger, J., Dunning, D., 1999. Unskilled and unaware of it: how difficulties in J. Aquat. Food Prod. Technol. 19, 103–119. https://doi.org/10.1080/ recognizing one’s own incompetence lead to inflated self-assessments. J. Pers. Soc. 10498850.2010.492093. Psychol. 77 https://doi.org/10.1037/0022-3514.77.6.1121. APROMAR, 2020. Aquaculture in Spain 2020. Annual Report, 95 pp. López-Mas, L., Claret, A., Reinders, M.J., Banovic, M., Krystallis, A., Guerrero, L., 2021. Ayvaz, Z., 2017. Consumer preference for seafood: the myths and realities, in: fifth Farmed or wild fish? Segmenting European consumers based on their beliefs. international conference sustainable postharvest and food technologies - INOPTEP, Aquaculture 532, 735992. https://doi.org/10.1016/j.aquaculture.2020.735992. April 23-28th 2017, Vrsac, Serbia, 4. MAPAMA, 2018. Estadísticas Pesqueras, p. 302. Bacher, K., 2015. Perceptions and misconceptions of aquaculture. Globefish Res. MERCABARNA, 2019. Mercabarna Central Fish Market Statistics [WWW Document]. Program. FAO 120, 35pp. URL. https://www.mercabarna.es/sala-de-premsa/publicacions/llibres-estadistics/. Bastian, S., Saltman, Y., Johnson, T., Wilkinson, K., 2015. Australian wine Pauly, D., Christensen, V., Guénette, S., Pitcher, T.J., Sumaila, U.R., Walters, C.J., consumers&rsquo; acceptance of and attitudes toward the use of additives in Watson, R., Zeller, D., 2002. Towards sustainability in world fisheries. Nature 418, wine and food production. Int. J. Wine Res. 7, 83. https://doi.org/10.2147/IJWR. 689–695. S90802. Pauly, D., Alder, J., Bennett, E., Christensen, V., Tyedmers, P., Watson, R., 2003. The Bergh, Ø., 2007. The dual myths of the healthy wild fish and the unhealthy farmed fish. future for fisheries. Science (80-. ). https://doi.org/10.1126/science.1088667. Dis. Aquat. Org. https://doi.org/10.3354/dao075159. Pieniak, Z., Verbeke, W., Scholderer, J., Brunsø, K., Olsen, S.O., 2007. European Bonsall, P., Shires, J., 2009. Estimating the robustness of questionnaire results: lessons consumers’ use of and trust in information sources about fish. Food Qual. Prefer. 18, from a mixed-mode survey of expectations for tele-working and road-based business 1050–1063. https://doi.org/10.1016/j.foodqual.2007.05.001. travel. Transportation (Amst). 36, 47–64. https://doi.org/10.1007/s11116-008- Pieniak, Z., Verbeke, W., Scholderer, J., 2010. Health-related beliefs and consumer 9185-5. knowledge as determinants of fish consumption. J. Hum. Nutr. Diet. 23 https://doi. Cardoso, C., Lourenço, H., Costa, S., Gonçalves, S., Nunes, M.L., 2013. Survey into the org/10.1111/j.1365-277X.2010.01045.x. seafood consumption preferences and patterns in the portuguese population. Gender Regnier, E., Bayramoglu, B., 2017. Competition between farmed and wild fish: the and regional variability. Appetite 64, 20–31. https://doi.org/10.1016/j. French sea bass and sea bream markets. Aquac. Econ. Manag. 21, 355–375. https:// appet.2012.12.022. doi.org/10.1080/13657305.2016.1189012. Claret, A., Guerrero, L., Ginés, R., Grau, A., Hernández, M.D., Aguirre, E., Peleteiro, J.B., Reig, L., Escobar, C., Carrassón, M., Constenla, M., Gil, J.M., Padrós, F., Piferrer, F., Fernández-Pato, C., Rodríguez-Rodríguez, C., 2014. Consumer beliefs regarding Flos, R., 2019. Aquaculture perceptions in the Barcelona metropolitan area from fish farmed versus wild fish. Appetite 79, 25–31. https://doi.org/10.1016/j. and seafood wholesalers, fishmongers, and consumers. Aquaculture 510. https://doi. appet.2014.03.031. org/10.1016/j.aquaculture.2019.05.066. Claret, A., Guerrero, L., Gartzia, I., Garcia-Quiroga, M., Ginés, R., 2016. Does information Schlag, A.K., Ystgaard, K., 2013. Europeans and aquaculture: perceived differences affect consumer liking of farmed and wild fish? Aquaculture 454, 157–162. https:// between wild and farmed fish. Br. Food J. 115, 209–222. https://doi.org/10.1108/ doi.org/10.1016/j.aquaculture.2015.12.024. 00070701311302195. Dey, M.M., Rab, M.A., Paraguas, F.J., Piumsombun, S., Bhatta, R., Alam, M.F., Strobl, C., Boulesteix, A.-L., Kneib, T., Augustin, T., Zeileis, A., 2008. Conditional Ahmed, M., 2005. Fish consumption and food security: a disaggregated analysis by variable importance for random forests. BMC Bioinformatics 9, 307. https://doi.org/ types of fish and classes of consumers in selected Asian countries. Aquac. Econ. 10.1186/1471-2105-9-307. Manag. 9 https://doi.org/10.1080/13657300590961537. Thomas, J.B.E., Nordström, J., Risén, E., Malmström, M.E., Gröndahl, F., 2018. The FAO, 2020. The state of the world fisheries and aquaculture. Sustainability in action. Nat. perception of aquaculture on the Swedish west coast. Ambio 47, 398–409. https:// Res. https://doi.org/10.4060/ca9229en. doi.org/10.1007/s13280-017-0945-3. Fernández-Polanco, J., Luna, L., 2010. Analysis of perceptions of quality of wild and Touvier, M., Méjean, C., Kesse-Guyot, E., Pollet, C., Malon, A., Castetbon, K., cultured seabream in Spain. Aquac. Econ. Manag. 14, 43–62. https://doi.org/ Hercberg, S., 2010. Comparison between web-based and paper versions of a self- 10.1080/13657300903566878. administered anthropometric questionnaire. Eur. J. Epidemiol. 25 https://doi.org/ Fernandez-Polanco, J., Luna, L., 2012. FACTORS AFFECTING CONSUMERS’ BELIEFS 10.1007/s10654-010-9433-9. ABOUT AQUACULTURE. Aquac. Econ. Manag. 16, 22–39. https://doi.org/10.1080/ van den Wollenberg, A.L., 1977. Redundancy analysis an alternative for canonical 13657305.2012.649047. correlation analysis. Psychometrika 42. https://doi.org/10.1007/BF02294050. Fernbach, P.M., Light, N., Scott, S.E., Inbar, Y., Rozin, P., 2019. Extreme opponents of Vanhonacker, F., Pieniak, Z., Verbeke, W., 2013. European consumer image of farmed genetically modified foods know the least but think they know the most. Nat. Hum. fish, wild fish, seabass and seabream. Aquac. Int. 21, 1017–1033. https://doi.org/ Behav. https://doi.org/10.1038/s41562-018-0520-3. 10.1007/s10499-012-9609-2. Feucht, Y., Zander, K., 2016. Aquaculture in the German print media. Aquac. Int. https:// Verbeke, W., Vackier, I., 2005. Individual determinants of fish consumption: application doi.org/10.1007/s10499-016-0021-1. of the theory of planned behaviour. Appetite 44, 67–82. https://doi.org/10.1016/j. Garza-Gil, M.D., Vázquez-Rodríguez, M.X., Varela-Lafuente, M.M., 2016. Marine appet.2004.08.006. aquaculture and environment quality as perceived by Spanish consumers. The case Verbeke, W., Sioen, I., Pieniak, Z., Van Camp, J., De Henauw, S., 2005. Consumer of shellfish demand. Mar. Policy 74, 1–5. https://doi.org/10.1016/j. perception versus scientific evidence about health benefits and safety risks from fish marpol.2016.09.011. consumption. Public Health Nutr. 8, 422–429. https://doi.org/10.1079/ Graeff, T.R., 2003. Exploring Consumers’ answers to survey questions: are uninformed PHN2004697. responses truly uninformed? Psychol. Mark. https://doi.org/10.1002/mar.10090. Verbeke, W., Sioen, I., Brunsø, K., Henauw, S., Camp, J., 2007. Consumer perception Hardy, R.W., 2005. Fish meal myths concerning omnivorous farmed fish. Aquac. Mag. versus scientific evidence of farmed and wild fish: exploratory insights from 31, 53–58. Belgium. Aquac. Int. 15, 121–136. https://doi.org/10.1007/s10499-007-9072-7. Hope, A.C.A., 1968. A Simplified Monte Carlo Significance Test Procedure. J. R. Stat. Vergnaud, A.C., Touvier, M., Méjean, C., Kesse-Guyot, E., Pollet, C., Malon, A., Soc. Ser. B 30. https://doi.org/10.1111/j.2517-6161.1968.tb00759.x. Castetbon, K., Hercberg, S., 2011. Agreement between web-based and paper versions Hufnagel, E.M., Conca, C., 1994. User response data: the potential for errors and biases. of a socio-demographic questionnaire in the NutriNet-Santé study. Int. J. Public Inf. Syst. Res. 5 https://doi.org/10.1287/isre.5.1.48. health 56. https://doi.org/10.1007/s00038-011-0257-5. INE, 2020. Continuous Household Survey (CHS). Year 2019. Instituto Nacional de Whitmarsh, D., Wattage, P., 2006. Public attitudes towards the environmental impact of Estadística. www.ine.es/en/. salmon aquaculture in Scotland. Eur. Environ. 16, 108–121. https://doi.org/ Kaimakoudi, E., Polymeros, K., Schinaraki, M.-G., Batzios, C., 2013. Consumers’ attitudes 10.1002/eet.406. towards fisheries products. Procedia Technol. 8, 90–96. https://doi.org/10.1016/j. Wrigley, J., 2017. The “F-word”: awareness and perceptions in fin-fish farming and the protcy.2013.11.013. “F-word”: awareness and perceptions in fin-fish farming and aquaculture policies Kok, B., Malcorps, W., Tlusty, M.F., Eltholth, M.M., Auchterlonie, N.A., Little, D.C., aquaculture policies, WWU graduate school collection. Western Washington Harmsen, R., Newton, R.W., Davies, S.J., 2020. Fish as feed: Using economic University. doi:10.25710/m9d4-cj68. allocation to quantify the Fish In : Fish Out ratio of major fed aquaculture species. Aquaculture 528, 735474. https://doi.org/10.1016/j.aquaculture.2020.735474. 11