Japanese consumer's visual marketing preferences and willingness to pay for rice produced by biodiversity‐friendly farming

A market‐based approach for biodiversity conservation can promote sustainable food production, but little is known about how consumers perceive rice and organisms conserved through biodiversity‐friendly farming (BFF). To address this gap, we conducted a web survey of 600 Japanese respondents to investigate their visual preferences for 20 photographs of organisms and landscapes found in paddy field areas and their willingness to pay (WTP) for BFF rice. Our results showed that paddy field landscapes were most preferred, followed by plants and birds, while insects and an endangered frog were the least preferred. Lowland flat and mountainous terraced paddy landscapes were most often selected as package photographs for BFF rice. Most respondents were willing to pay a premium of 0%–30% for BFF rice. Interest in biodiversity conservation was a common respondent characteristic, and it had a relatively strong influence on image preferences and WTP. Presence of children in certain age groups, sex, age, and experience with living near farmlands or conservation activities also affected the image preference or WTP answers, but to a lesser extent. These findings reveal consumer perceptions of BFF and the importance of effective eco‐labeling strategies for communicating conservation activities and improving promotion of sales to consumers.


| INTRODUCTION
Sustainable agriculture plays a crucial role in conserving natural resources and biodiversity in production landscapes (Bovarnick, 2003).Mosaics and heterogeneity of land use are important for maintaining biodiversity, largely because of the contribution of small-scale agriculture utilizing low input cultivation systems and diverse landraces (Biasi & Brunori, 2015;Moser et al., 2002).However, small agricultural businesses maintaining local biodiversity face significant challenges in operating profitably within a competitive marketplace.To address this issue, biodiversity labeling schemes have emerged as a means of communicating the positive outcomes of biodiversity conservation on farmlands to consumers (Edwards & Laurance, 2012;Lenzen et al., 2012).Although organic agriculture is a representative example of biodiversity-friendly farming (BFF), a meta-analysis revealed yield losses of up to 19.2% (Ponisio et al., 2015), which pose an economic threat to producers.Organic agriculture is recognized as having the potential for expansion due to the ability to charge a price premium despite the impact of increased costs and decreased yields, but the actual worldwide area under organic cultivation is limited to 1% of total cropland (Crowder & Reganold, 2015).Therefore, stable price premiums for produce harvested through BFF practices, which are largely still at the stage of development and proliferation, are crucial to expanding biodiversity conservation efforts.
The most widely cultivated rice, Oryza sativa L., is the staple food of an estimated 3.5 billion people worldwide (Muthayya et al., 2014).Although rice paddies are an important component of global wetlands, their expansion and the inappropriate use or management of water, pesticides, and fertilizers have caused declines in various ecosystem services from wetland areas, including biodiversity (MEA, 2005).Implementing water-saving practices in paddy fields has led to shortened periods of flooding, which undermines the habitat value of wetlands for a globally endangered waterbird species (Herring et al., 2021).As paddy field size has increased due to the agricultural intensification, the vegetated area of levees has decreased, leading to a reduction in the population continuity of natural enemies and biodiversity in paddy field landscapes (Way & Heong, 1994).
As a countermeasure to the decline of biodiversity in the areas of conventional rice cultivation, BFF has been adopted in various regions through implementing practices such as organic farming, integrated pest management, reduction of chemical pesticides and fertilizers, winter flooding, extending the flooded period during cultivation, earthen-ditch installation, and improved regulation and land-use planning (Clements et al., 2010;Herring et al., 2021;Katayama et al., 2020).The effectiveness of labeling to indicate the conservation target species and biodiversity-friendly practices has been confirmed in actual markets (Mameno et al., 2021;Oishi, 2016) and choice experiments (Herring et al., 2022).For example, Mameno et al. (2021) found a price premium of approximately 20% on sales of BFF rice could be realized with eco-labeling.However, a higher price premium is desirable if yield loss, annual yield fluctuations, and additional economic costs for implementing BFF practices are all considered.The conservation of diverse wildlife may make it possible to achieve higher-priced sales in cases where social capital (e.g., connection to farmer friends, local environmental non-governmental organizations, or social clubs) can be utilized to sell environmentally friendly produce at a premium (Nishimura et al., 2012a).
In addition, successful examples of advantageous sales have actually been achieved by combining the conservation of a flagship species of stork, pesticide reduction, and outreach activities involving multiple stakeholders (Uenishi, 2019).Producers who combined certified flagship species conservation with reduction of synthetic pesticide/fertilizer have been able to obtain price premiums of more than 200% (Herring et al., 2022).However, for small-scale producers who lack such social capital or flagship species in the local species pool, these precedents have limited implications in terms of implementing BFF farming from a business model perspective.
Private eco-labeling is a scheme that is easy for small-scale producers to implement.However, consumers are skeptical of food labels if they lack familiarity and trust in the commodities (Sirieix et al., 2013).Previous studies have shown that the proper selection of taxa or landscapes for labeling of BFF products was crucial for effective communication between producers and consumers (Herring et al., 2022;Mameno et al., 2023;Sehra & MacMillan, 2021).Consumers in the capital city of Tokyo preferred images of traditional rural landscapes the most, followed by the black crowned-night heron (Nycticorax nycticorax) and the Japanese pond loach (Misgurnus anguillicaudatus, Sehra & MacMillan, 2021).In a choice experiment of rice labeling, consumers showed a greater preference for birds than for fish (Mameno et al., 2023).These studies indicate that the selection of labels representing biodiversity in the paddy field landscape is crucial to promote purchases of BFF products in Japan.In studies of organic agricultural products, consumer attributes such as household composition, income, gender, and knowledge about organic farming have been shown to strongly influence purchasing behavior or willingness to pay (WTP, Li & Kallas, 2021;McFadden & Huffman, 2017;Yiridoe et al., 2005).Moreover, consumers' attitude about the environment is known to influence the WTP for some organic products (De Magistris & Gracia, 2008;Nguyen et al., 2019).In recent years, the increase in number of urban residents and the resulting loss of people's experiences with nature and biodiversity have been shown to influence the formation of awareness of and attitudes toward biodiversity conservation (Fukano & Soga 2021, Lockwood, 2013).It is conceivable that these influences also affect purchasing behavior and WTP for BFF products.These previous findings imply the necessity to examine the relationship between visual preferences for biodiversity as well as the demographic attributes of consumers to properly target consumers demanding BFF products and contribute to the appropriate use of labeling tailored to different consumer segments.Furthermore, the evaluation of the WTP for BFF rice by consumers with different attributes is crucial information for producers when determining product pricing.
The Japanese archipelago is spread out over a wide area, and local flora and fauna are diverse.However, by implementing BFF practices, some common organisms can be conserved, and these species can be used as indicators to evaluate the outcomes of BFF practices in paddy fields (National Agriculture and Food Research Organization, 2020).If some of those common indicators are viewed positively by consumers, using their images on labels may improve sales and profits, leading to the proliferation of BFF in various areas.In this study, we aimed to evaluate the following by using survey data obtained from a web questionnaire of Japanese respondents: (1) the visual preferences of respondents of multiple landscapes and species conserved by BFF practices and the socio-economic characteristics of respondents affecting the preference patterns; (2) suitable photographs for BFF rice packages; and (3) the relationship between respondent characteristics and willingness to pay (WTP).Based on the results, we discuss the key factors shaping consumer preferences for images of biodiversity in rice paddy landscapes and recommend labeling strategies to promote BFF rice sales.

| Questionnaire design
The respondent's preferences for photographs of landscapes, plants, and animals were collected on a Likert scale from 1 (dislike) to 5 (like).Twenty photographs were selected as examples of traditional paddy field landscapes or indicator species conserved under BFF practices (Table S1, Figure S1; National Agriculture and Food Research Organization, 2020).Following the photograph preparation method proposed by Landov a et al. ( 2018), we placed plants and animals on a white background to avoid potential background effects.An introductory statement provided a common background explanation prior to the questions.The statement explained that the following pictures show landscapes or organisms found in rice paddies cultivated with consideration for the local natural environment during agricultural activities.The 20 photographs were then presented in a randomized order to each respondent without the place or scientific names being shown.The purpose of the questions was to evaluate the respondent's visual preferences for the landscapes and species, which is why we avoided providing their names, which could possibly evoke a response related to their conservation status or cultural value.
In the next question, the respondents were asked to select up to three favorite photographs that would entice them buy rice cultivated to preserve the landscape and species in rice paddies.The option of "there is no suitable photo" was included.In the next query, the respondents were asked whether or not they want to buy biodiversity-friendly rice at different price points.To reduce hypothetical bias in the WTP questionnaire, we added the following statement following the "cheap talk" script approach proposed by Cummings and Taylor (1999): "In surveys that ask for the willingness to pay for a specific product or service in a hypothetical scenario, respondents may sometimes answer that they would be willing to pay a higher price than they actually would.Please consider your actual purchasing behavior when answering the following question." Then, respondents were asked whether or not make the purchase 5 kg of polished rice at the following prices (all in Japanese yen): 2019 (the same price as rice produced with conventional practices), 2221, 2423, 2625, 3029, 3533, and 4038 with the indication of the price premium against conventional rice in each choice.The format of the question offered a binary choice-"buy" or "do not buy."This process was repeated for all prices in ascending order.This approach aimed to uncover instances where respondents displayed a stronger inclination to support BFF, indicating a pronounced preference for a specific higher price premium while exhibiting hesitancy toward lower prices.This format, which allowed respondents to choose "do not buy" for all seven prices, was aimed at introducing realism to the decision-making process, as discussed by Koemle and Yu (2020).Responses indicating "do not buy" for all seven price points were interpreted as representing a value of 0 yen in terms of actual sales.These responses were incorporated into the dataset for the regression analysis.These price thresholds were chosen to be close to those reported from previous studies on biodiversity friendly rice (Mameno et al., 2021;Sehra & MacMillan, 2021).

| Data collection
Data were collected February 16-17, 2023 from a sample of 600 respondents (Table S2) in an Internet questionnaire administered by Rakuten Insight (https://member.insight.rakuten.co.jp/).The respondents were all Rakuten Insight panel members and over 20 years old.Respondent selection was designed to be proportional to the age structure of the male and female population in Japan, divided into six age groups: 20-29, 30-39, 40-49, 50-59, 60-69, and 70+.The number of respondents living in the three major Japanese cities (Tokyo, Aichi, and Osaka) was 190, accounting for 31.7% of the 600 respondents (Table S2).This was significantly higher than the corresponding value of 24.3% of the population in 2022 statistics (https://www.stat.go.jp/data/jinsui/2022np/index.html, p < .001,prop-test).Referring to items asked in questionnaire surveys of previous studies on BFF products (Herring et al., 2022;Sehra & MacMillan, 2021), the following socio-demographic and general information was collected: sex, age, hometown, marital status, education, current occupation, annual income, number and school age of children, number of household members, and rice consumption rate was calculated by dividing the number of rice meals by the total number of meals in a week.To detect the degree of respondent's contact with agricultural environment, we asked whether there were paddy fields or other farmlands within a 1-2 min walk of their current residence.We also asked whether they had lived near paddy fields or farmlands in the past, and if so, for how many years (<1 year, 1-4 years, 5-9 years, 10-19 years, and ≥ 20 years).The respondent's interest in biodiversity conservation was categorized as interested, fairly interested, neutral, not very interested, and uninterested.The respondents were also asked if they had participated in environmental conservation activities after specific examples of conservation activities (e.g., protecting and increasing sea and terrestrial organisms, extermination of alien species, and environmental education and awareness activities) had been explained.

| Evaluation procedure
A pair-wise Wilcoxon test with Holm correction was performed to compare the ordinal strength of preference among the 20 photographs.To extract consumer socioeconomic characteristics that influence photographic preferences from multivariate data involving 600 individuals with diverse attributes, it is crucial to avoid the problem of multiplicity in statistical testing.Multivariate analytical approaches for interview data with Likert-scale scores can be employed to address this concern, as demonstrated by previous studies (e.g., Campos-Silva et al., 2021;Martín-L opez et al., 2007).We employed the Permutational Multivariate Analysis of Variance (PERMANOVA) to examine the correlation between preference score compositions and the respondent's characteristics.In the PERMANOVA and the two regression analyses (Firth's bias-reduced logistic regression and Tobit multiple linear regression), the following fixed-effect linear predictors were included as categorical variables: presence of a child (in different age groups), currently living near farmlands, time spent living near farmlands, interest in biodiversity conservation, experience of conservation activities, income, marital status, last educational experience, and sex.Numerical variables included age and rice consumption rates.Significant variables identified through PERMANOVA were further explored using Nonmetric Multidimensional Scaling (NMDS) to visualize the relationships between photograph preference scores and respondent demographic characteristics in a reduced-dimensional space.If some species points are closer to the centroid of one categorical respondent group than to another, it suggests a stronger association with the first group.If a species point is close to an arrow representing a numerical variable, it suggests a strong association between the photograph preference scores and the variable.To complement the interpretation of NMDS ordination, we calculated the mean preference scores for the 20 photographs for significant variables.
To examine the characteristics of respondents who answered, "there is no applicable photo available" (i.e., none), Firth's bias-reduced logistic regression analysis (Heinze & Schemper, 2002) was performed because perfect separation was detected in the preliminary binomial generalized linear model analysis.We confirmed the variation inflation factor of the fixed effect terms was <5 in this analysis.A post hoc pair-wise proportion test was applied to compare the proportion of answers against significant categorical variables.After the 69 "none" answers were removed, PERMANOVA was performed to examine the relationship between package selection and respondent's characteristics using the presence/absence answers on the preferred package (n = 531).
Because the price threshold was truncated at 0 yen for the all "do not buy" answers and 4038 yen (a 100% price premium), we applied Tobit multiple linear regression analysis to the answered prices as the response variable and the respondents' characteristics as explanatory variables.In this analysis, data from respondents who did not know their income (n = 50) and who did not want to answer the income question (n = 88) were omitted to prioritize the accuracy of income data and its influence on the answered prices.We again confirmed the variation inflation factor of the fixed effect terms was <5 in this analysis.As a post hoc test, a pair-wise Wilcoxon test was performed to compare the ordinal amounts of answered prices among respective consumer attributes that were found to be significant in the Tobit multiple linear regression ( p < .05).
We used violin plots, which are similar to boxplots but more flexible in terms of expression of data density, to visualize WTP against significant categorical variables detected by Tobit multiple linear regression.The analyses were performed with R software, version 4.2.3 (R Development Core Team, 2023) using the following packages: MASS, AER, car, tidyr, dplyr, ggplots2, ggthemes, ggsignif, gridExtra, vegan, logistf, reshape2, and ggpubr.

| Image preferences
Respondent's image preferences varied significantly (Figure 1).Overall, landscape photographs were the most preferred, followed by plants, birds, and Japanese rice fish.Dragonflies and loaches followed, while aquatic insects and the endangered Daruma pond frog were the least preferred group.As shown in Table 1, the PERMA-NOVA results indicated that interest in biodiversity conservation explained the total compositional variance of respondents' visual preference scores most effectively.Additionally, variables such as age, having children in pre-school or high school, current and lifetime rural residency, and sex were significant but less explanatory.According to the NMDS ordination in Figure 2, which had a stress value of 0.128, and a supplementary comparison of the mean preference scores for the 20 photographs (Table S3), diverse photographic preferences were observed, depending on the demographic characteristics of the respondents.A correlation was identified between the level of interest in conservation and the degree of preference for most of the photographs.Elderly individuals, those living near farmlands, and males exhibited a tendency to prefer small animals.Among respondents with preschool children, a tendency to not prefer dragonflies was evident, while households with high school students exhibited a tendency to not favor small animals and birds.

| Preferred biodiversity-friendly rice packaging
Photographs of flat rice fields without trees and terraced rice fields were dominantly selected as the preferred packaging photos (Figure 3), but 69 respondents answered that there were no suitable photos for the package.Firth's regression analysis showed that interest in biodiversity conservation and sex were significant variables (Table S4).There were significantly more "no suitable photo" responses from those who were uninterested in conservation than (fairly) interested respondents (28.6% of uninterested versus 2.7% of fairly interested and 4.6% of interested respondents, p < .05,pairwise prop-test).There were more "no suitable photo" responses from male (14.1%) than female (8.9%).After the 69 "none" answers were excluded, PERMANOVA using data for the package selection of the 531 remaining respondents showed that there was no significant correlation between the respondent characteristics and package selection patterns (p > .05,Table S5).The results indicate a universal package preference trend for landscape labels that is not dependent on consumer demographic characteristics across all generations and genders.

| Willingness to pay
As shown in Table 2, variables that exhibited a significant correlation with the answered price were interest in biodiversity conservation, experience with biodiversity conservation activities, and sex.Female respondents and those who had higher interest in or experience with conservation activities had a significantly higher WTP for biodiversity-friendly rice (Figure 4).Many responses were concentrated between 2019 and 2625 (0%-30% price premium).

| Image preferences and respondent characteristics
Our results showed that there was no tendency for species that have been designated as endangered under the IUCN or domestic laws in Japan to have higher scores (Figure 1, Table S1).It is known that many people around the world have a dislike or fear of insects (e.g., Fukano & Soga, 2021;Lockwood, 2013).In addition, people tend to experience fear and negative emotions toward certain species that are commonly associated with potential harm to human beings, such as snakes, spiders, and insects that can cause venomous bites or stings (Martín-L opez et al., 2007).Therefore, lower preference scores for insects and an endangered Daruma pond frog may be related to people's negative perceptions for those fauna.In contrast, landscape photographs were preferred as compared with specific organisms (Figure 1).Various elements such as the percentage of plant cover, the amount of water, the presence of mountains, and the color contrast are known to influence the visual perceptions of landscapes (Arriaza et al., 2004), but we were T A B L E 1 Results of permutational multivariate analysis of variance on the respondent's visual preference for the 20 landscape or species photographs (n = 600, Bray-Curtis distance index, 999 permutation).limited to only three landscape photographs and are unable to draw conclusions on the effects of these types of differences.Future research is needed to evaluate how consumers perceive the photos of farmland landscapes with such diverse elements.Although the methods and analysis were different, Sehra and MacMillan (2021) also found that landscapes and birds were effective means for raising the price premium for BFF rice, and that the Daruma pond frog was not.Our results, drawn from a broader demographic range in terms of ages and regions, suggest a similar pattern.
The PERMANOVA and NMDS ordination results showed that interest in biodiversity conservation had a relatively strong influence on image preference patterns (Table 2, Figure 2, Table S3).The characteristics of respondents that negatively affect their preferences, such as lower interest in biodiversity, urban living, and younger age, seem to reflect generational differences in the opportunities for interaction with the natural world to some extent.In Japan, over the last half century, there has been a constant migration to urban areas (Ministry of Land, Infrastructure, Transport and Tourism, 2011).As a result, younger generations have less direct experience with nature, leading to what has been called "extinction of experience" (Miller, 2005;Soga & Gaston, 2016).This phenomenon might have also affected how consumers perceive biodiversity in paddy fields and influenced their image preference.

| Packaging strategy
In the analysis of package selection, "none" responses were frequently obtained from male respondents and those who were uninterested in biodiversity (Table S4).Gender differences in esthetic perception have been insufficiently studied (Yang et al., 2018), but our results suggest the importance of marketing strategies that  2) on the ordination results of species.Circle sizes in (a-g) are proportional to the mean preference score of each photograph.recognize gender differences in consumer purchasing behavior of BFF rice.When the "none" answers were removed from the packaging analysis, there were no significant characteristics correlated with package selection patterns (Table S5).This result differs from those of the analysis of preferences for the 20 different photographs, most likely because of differences in the methods of inquiry.In the photo rating query, respondents were asked to rate all photos on a Likert scale, whereas in the package selection, respondents were asked to choose up to three candidate photographs or no candidate photos.Regardless of the degree of preference for a particular photograph (i.e., species), the results reflect a common tendency to prefer landscape photos, especially lowland flat areas without trees and mountainous terraced paddy landscapes (photograph codes b and c in Figure 1), relative to photos of other organisms.Based on these results, it would currently be advantageous to choose landscape photographs to promote BFF rice to an unspecified number of consumers rather than use specific conservation target species from a sales strategy perspective.

| Willingness to pay
Respondent's answers about a price premium for BFF rice were mainly concentrated on a premium of 0%-30% relative to the price of conventional rice (Figure 4).Previous actual market and choice experiment studies have shown a 20%-30% price premium (Herring et al., 2022;Mameno et al., 2021;Policy Research Institute, MAFF, 2010), suggesting our results are reasonable and may reflect the actual purchasing behavior of consumers.Rice grown with one method of organic agriculture, which only used non-chemical fertilizers and did not use pesticides, was able to sustain a price premium of more than 50% in Japan (Policy Research Institute, MAFF, 2010).A study comparing WTP for organic and BFF rice found that consumers were willing to pay a higher price premium for organic products as compared to BFF products (Sehra & MacMillan, 2021).The price premium for organic rice; however, has varied among countries and respondents' income.In many countries, one of the main reasons consumers give for purchasing organic rice is the enhanced safety due to the non-use of F I G U R E 3 Cumulative counts of the photographs of preferred landscape or species for biodiversity-friendly rice packages.The respondents were instructed to select up to three favorite photographs or not to select any photograph if they did not have a favorite photograph (none).The legend of the stacked bar graph is color-coded according to the number of responses for each respondent.
pesticides (e.g., Ara, 2003;Hosseini et al., 2019;Li et al., 2019;Sriwaranun et al., 2015;Wijesinghe & Nazreen, 2020).Other positive motivations have been given for purchasing behavior, such as improved taste, better product quality, environmental considerations, and support for the local economy (Hughner et al., 2007;Sriwaranun et al., 2015).On the other hand, a lack of knowledge of environmental concerns hindered the increase of WTP for organic rice in Indonesia (Grimm et al., 2023).BFF and its products in Japan have not yet established these kinds of positive motivations and prior knowledge for consumers, which probably resulted in the realization of lower price premiums in our study.
T A B L E 2 Analysis of deviance table on the willingness to pay for rice produced by biodiversity-friendly farming (n = 462).Tobit multiple linear regression analysis was performed for the upper (4038 yen) and lower (0 yen) truncated data of respondent's answered price.As shown in Table 2 and Figure 4, stronger interest in biodiversity conservation, experience with conservation activities, and female gender are consumer characteristics that indicate a higher WTP for BFF rice.Although males had a higher WTP for bittern-friendly rice in Australia (Herring et al., 2022), several studies have indicated females are more likely to pay a premium for environmentally friendly products (Lockie et al., 2004;Loureiro et al., 2002;Tobler et al., 2011), which is in accordance with our results.According to expenditure statistics for single-person households in Japan in 2022, males spent 107% more on dining out and spent 19% less on purchasing rice than females (Ministry of Internal Affairs and Communications, 2022).In addition, a comparison of household workload between couples in Japan from 1994 to 2009 revealed that women clearly did a larger amount of housework, including cooking, throughout the survey period (Tsuya et al., 2012).Furthermore, a nationwide survey conducted in 2021 targeting individuals aged 15 and above (https://www.e-stat.go.jp/en/stat-search/ files?page=1&toukei=00200533&tstat=000001158160) indicates that across all age groups, women spend more time per week managing meals.These gender differences in purchasing behaviors and household workload might partly affect consumers' WTP for BFF rice in our study.
In Japan, diverse threatened species have been conserved in regions where BFF has been ongoing for many years, and an approximate 30% price premium for rice has been realized in these cases (Policy Research Institute, MAFF, 2010).Adding a price premium seems to be possible when social capital is available so that farmers are able to share their BFF practices and outcomes on biodiversity within local communities (Nishimura et al., 2012a).To help in the proliferation of BFF products and enhance biodiversity conservation across wider regions where local social capital does not work, it will be necessary to not only nurture interest in conservation but also to foster the types of consumer motivations observed in organic agriculture.For example, combining attributes, such as the use of conservation target species and eco-friendly practices (e.g., reduction of synthetic pesticides/fertilizers), should be an effective promotion measure to motivate consumers to purchase BFF products (Herring et al., 2022;Mameno et al., 2023;Nishimura et al., 2012b).

| Limitations
One limitation of the present study was that the respondents of our questionnaire were sampled only from monitors registered with Rakuten Insight, and 31.7% of the respondents were from three major Japanese cities (Tokyo, Osaka, and Nagoya).In the 2022 census, the population from these three cities accounted for 24.3% of the total population in Japan (https://www.stat.go.jp/data/ jinsui/2022np/index.html).This bias in sampling may have an effect on respondents' prior knowledge of organisms that are not often found in urban areas and may have influenced their responses.Moreover, respondents were limited to Internet users because they were the registered monitors for online surveys.According to data from a 2020 survey of 41,387 people in Japan (https:// www.soumu.go.jp/johotsusintokei/statistics/statistics05.html), more than 90% of people between 20 and 59 years old used the Internet, but only 82.7% of people in their 60s, 59.6% of people in their 70s, and 25.6% of people in their 80s used it.For this reason, data from monitors in their 60s and older are highly skewed toward Internet users.Previous studies have shown significant differences in demographic and behavioral characteristics between Internet user and nonusers, including in household size, education, household income, money spent, and information acquisition behaviors (e.g., Bonn et al., 1998;Shimizu, 2003;Vatsa et al., 2023).These limitations should be addressed in future studies.

| Conclusion
Based on the results of our survey, from the perspective of promoting and expanding the consumption of rice produced with consideration for biodiversity conservation among the public, it was suggested that landscape photographs are currently more effective than featuring individual species when selecting packaging.Moreover, most respondents were willing to pay a premium of 0%-30% for BFF rice, establishing a practical pricing range for producers engaged in BFF.While analytical techniques in species identification are expected to become more widely adopted due to advancements such as environmental DNA metabarcoding (Thomsen & Willerslev, 2015) and AI image identifiers (August et al., 2020), the results of this study raise questions about the effectiveness of a direct promotional approach using species photographs as labels, even when indicator species conserved by BFF implementation can be identified.Incorporating textual information, such as conservation ranks of detected species, conservation effects at the ecosystem level, and reductions in pesticide and chemical fertilizers implemented in BFF will likely be necessary.In the future, it will be important to achieve social change that garners consumer support for the conservation of not only attractive paddy landscapes but also rare and indigenous species to expand BFF.To achieve this, it is necessary to increase opportunities to interact with various organisms through environmental education, agro-tourism, and biodiversity conservation activities that can transform consumer awareness and attitudes.These measures will help to mitigate the impacts of the extinction of experience on the market-based approach to biodiversity conservation.

1
Comparison of visual preference scores among the 20 landscape or species photographs.Mean ± SD (thick horizonal line and vertical bars) was overlaid with violin plots of the Likert scores (1: dislike to 5: like).Different lower-case letters at the top indicate significant differences by the pair-wise Wilcoxon test ( p < .05 with Holm correction).

F
I G U R E 2 Ordination by nonmetric multidimensional scaling (NMDS) based on visual preference composition.The Bray-Curtis distance was applied as a dissimilarity index of visual preference data collected as Likert-scale scores (1: dislike to 5: like).(a) Biplot of age on the ordination results of species.(b-g) Spider plots for each significant category of respondents' characteristics detected by PERMANOVA (see details in Table

F
I G U R E 4 Relationship between answered prices for biodiversity-friendly farming (BFF) rice and significant variables, as detected by Tobit multiple linear regression analysis.Bold horizontal line represents the mean, and the vertical lines extending from it represent the standard deviation.Mean ± SD was overlaid with a violin plot for the respondents' answered prices: (a) conservation interest, (b) experience with conservation activities, and (c) sex.Different lower-case letters at the top indicate significant differences by the Wilcoxon test (p < .05 with Holm correction only for (a)).