2.1 Google Trends

Google Trends (https://trends.google.com/trends/) [34] is a freely accessible online database that provides time-series data on search volumes for specific keywords entered into Google's search engine. Introduced by Google in 2006, this database has been widely utilized as a research tool across various disciplines, including marketing, economics, public health, and social sciences [35]. Notably, Google Trends is a valuable method for understanding the dynamics of public interest, with applications extending beyond consumer behavior and market trends to include forecasting infectious disease outbreaks and assessing political interest [36,37].

Note that Google Trends is based on an anonymized, categorized, and aggregated sample of actual Google searches. It removes repeated searches and excludes queries generated by Google products and services, including internal searches from AI Mode and AI Overviews. Therefore, AI-generated queries within Google’s own search products are unlikely to directly affect the Trends series. In addition, because the Google Trends time series analyzed in this study covers January 2010 through December 2023, potential distortions from generative-AI search interfaces are unlikely to affect the period analyzed.

In recent years, Google Trends has also been increasingly employed in the fields of biodiversity and environmental conservation [38,39]. For example, Jarić et al. [40,41] used Google Trends to demonstrate that public interest in biodiversity conservation can change rapidly over short periods. Their findings suggest that Google Trends can serve as an effective tool for continuously monitoring consumer awareness in environmental issues.

Prior work has mapped public perceptions of marine issues and the science–policy–society interface largely through cross-national surveys and media-based corpora, but behavior-based, comparable evidence remains limited. Traditional surveys and official statistics are indispensable, yet they are not designed to capture real-time fluctuations in public attention. In contrast, aggregated search behavior offers a complementary, high-frequency indicator that can be processed at scale. Accordingly, we use Google Trends search volume as a proxy for public attention to fishery resources, and we pair it with statistical series as described below.

2.2 Data collection

In this study, search volume data obtained from Google Trends were used as an indicator to measure public interest in the selected fish species. All data used in this study were collected in May 2024. The analysis focused on fish species subject to Total Allowable Catch (TAC) regulations in Japan as of 2024, excluding Pacific bluefin tuna (Thunnus thynnus). The targeted species included Japanese horse mackerel (Trachurus japonicus), Japanese sardine (Sardinops melanostictus), mackerel species (chub mackerel: Scomber japonicus and spotted mackerel: Scomber australasicus), Pacific saury (Cololabis saira), Alaska pollock (Gadus chalcogrammus), Japanese flying squid (Todarodes pacificus), and snow crab (Chionoecetes opilio).

Pacific bluefin tuna was excluded from the analysis due to the difficulty in accurately identifying search terms, as the term “Maguro” (tuna) is commonly used to refer to multiple tuna species, making it challenging to isolate searches specifically related to Thunnus thynnus. The keyword selection process is illustrated below, using Pacific saury as an example.

1. (1). Listing Common Names
2. For each fish species, commonly used names were identified, including their hiragana, katakana, and kanji representations, by referencing sources such as Wikipedia and online dictionaries. For example, in the case of Pacific saury, the terms “さんま (Hiragana),” “サンマ (Katakana)” and “秋刀魚 (Kanji)” were considered.
3. (2). Comparison Using Google Trends
4. Search volumes for all candidate terms were compared to determine the most frequently searched keyword. This comparison was based on the average relative search volume over the entire study period (2010–2023). In cases where average values were close, the term with the highest peak search volume was selected.
5. (3). Validation of Search Results
6. A manual Google search was conducted to verify whether the selected keyword primarily returned information relevant to the target fish species. If unrelated results appeared frequently, the next most searched keyword was considered instead.

For instance, in the case of Pacific saury, the term “Sanma (Hiragana)” was found to generate numerous search results related to a person’s name rather than the fish. As a result, “Sanma (Katakana)” the second most searched term, was adopted as the appropriate keyword. Similar ambiguities were addressed for other species. For example, for Japanese sardine, the search term “Iwashi (Hiragana)” was compared against “Katakuchi-Iwashi” and “Urume-Iwashi,” and since the latter two had negligible search volumes, “Iwashi (Hiragana)” was selected. In the case of Alaska pollock, searches for “Tara (Katakana)” and “Tara (Hiragana)” were found to overlap significantly with “Madara (Pacific cod),” leading to the adoption of “Sukesoudara (Katakana)” as the optimal keyword.

The final keywords selected for each species were as follows:

1. Japanese horse mackerel: “Aji (Hiragana)”
2. Japanese sardine: “Iwashi (Hiragana)”
3. Mackerel species: “Saba (Katakana)”
4. Pacific saury: “Sanma (Katakana)”
5. Alaska pollock: “Sukesoudara (Katakana)”
6. Japanese flying squid: “Surumeika (Katakana)”
7. Snow crab: “Zuwaigani (Katakana)”

The monthly relative search volume (RSV) data for each keyword was obtained from Google Trends, covering the period from January 2010 to December 2023. Google Trends reports relative search volume (RSV), a normalized index scaled from 0 to 100 within the specified time window and region rather than absolute search counts. Data collection was configured to the “Japan” region, “all categories,” and “web search” only. Additionally, the related search queries and prefecture-level search volumes were collected as annual data from 2010 to 2022.

Furthermore, supplementary datasets were obtained for comparative analysis: monthly catch volumes were obtained from the market-distribution statistics portal of the Fisheries Agency of Japan (https://www.market.jafic.or.jp/fKoukaiMain.html) [42]. Consumer Price Index (CPI) data (excluding Alaska pollock and snow crab) was obtained from the Statistics Bureau of Japan (https://www.stat.go.jp/data/cpi/index.htm) [43]. Market transaction prices from the Tokyo Metropolitan Central Wholesale Market were obtained as monthly data from the Tokyo Metropolitan Government website (https://www.shijou.metro.tokyo.lg.jp/torihiki/geppo) [44]. Prefecture-level annual catch volumes were obtained from Ministry of Agriculture, Forestry and Fisheries website (https://www.maff.go.jp/j/tokei/kouhyou/kaimen_gyosei/index.html) [45]. All datasets were collected and analyzed in accordance with the terms of service and usage policies of each data provider. For analysis, all datasets were integrated by species into a single time-series file, and the same file also includes values derived during preprocessing and statistical analyses. In addition, the Tokyo Metropolitan Government has changed the public dissemination format for the Central Wholesale Market statistics, and some historical pages are no longer directly accessible. To ensure reproducibility, we retained the processed monthly price series used in this study as part of the analysis dataset.

2.3 Statistical analysis

All statistical analyses in this study were performed using the SciPy 1.15.1 statistical analysis library in Python 3.12.2, while figures and tables were generated using the Matplotlib 3.8.3 visualization library. To ensure methodological consistency with the Consumer Price Index (CPI), the market price data was indexed by setting the 2023 average price to a value of 100. Subsequently, all time-series datasets were standardized using a z-score transformation (mean of 0 and standard deviation of 1).

2.3.5 Path analysis.

To elucidate the dynamics of public interest in the fish species targeted in this study and identify its driving factors, path analysis was employed. The time-series data used in the analysis were the residual components extracted through STL decomposition. The path analysis model was constructed based on theoretical causal relationships, taking into account the results of previous analyses (Fig 1).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Conceptual path analysis illustrating the hypothesized relationships between variables.

The model posits that fishery catch volume (MC_R) influences market price (MP_R), which in turn affects the Consumer Price Index (CPI_R). Public search volume (RSV_R) is influenced by all three preceding variables.

https://doi.org/10.1371/journal.pone.0342833.g001

Specifically, the model assumed the following relationships:

1. Market price (MP_R) is influenced by fishery catch volume (MC_R).
2. Consumer Price Index (CPI_R) is influenced by market price (MP_R).
3. Search volume (RSV_R) is influenced by fishery catch volume (MC_R), market price (MP_R), and consumer price index (CPI_R).

Each variable was defined as the residual component obtained through data preprocessing, and latent factor analysis was not performed. In the baseline specification, we did not estimate residual covariances; contemporaneous associations among predictors were represented by directed paths. As a robustness check, we additionally allowed a residual covariance between wholesale and retail prices (MP_R ~~ CPI_R) while retaining the pass-through path (CPI_R ~ MP_R); model fit and path estimates were virtually unchanged, so the covariance-free specification was adopted.

The analysis was conducted using the Semopy 2.3.11 library in Python. Estimation was performed using the maximum likelihood method (ML), with Sequential Least Squares Programming (SLSQP) employed as the optimization algorithm. After estimation, the model’s goodness of fit was evaluated using multiple fit indices, including the chi-square test, Comparative Fit Index (CFI), and Root Mean Square Error of Approximation (RMSEA), based on commonly accepted criteria (e.g., CFI > 0.95, RMSEA < 0.08). Additionally, variance inflation factors (VIF) were calculated to examine multicollinearity among variables and verify the model's stability. All VIF values were below 3.0, indicating that multicollinearity was not a significant issue in our model. Finally, path diagrams were initially generated using Graphviz 0.20.3 to determine the layout. The final figures were manually redrawn and formatted based on the Graphviz output; all coefficients and significance annotations were taken directly from the results.

3.1 Seasonal-trend decomposition using LOESS

The results of the time-series analysis revealed that both fisheries catch volume and search volume exhibited seasonal periodicity across all fish species. Additionally, the fish species could be broadly categorized into two groups based on the similarity of their catch volume and search volume patterns:

1. Species with similar seasonal patterns in catch volume and search volume (Japanese horse mackerel and Pacific saury).
2. Species with differing seasonal patterns in catch volume and search volume (Japanese sardine, Japanese flying squid, snow crab, Alaska pollock, and mackerel species).

These findings indicate that the temporal relationship between fishery catch volume and public interest varies across species.

3.2 Temporal dynamics: Price and seasonality as key drivers of public interest

3.3 Spatial dynamics: Regional catches shape local interest

Similar to the time-series analysis, the relationship between prefecture-level catch volume and search volume varied across fish species. For Japanese horse mackerel and Japanese sardine, a weak positive correlation was observed (Figs S6.1 and S6.2 in S1 File). The results suggest that Japanese horse mackerel is more frequently searched in southern prefectures, whereas Japanese sardine is more frequently searched in northern prefectures. This implies that while both species are widely recognized as staple seafood products, consumer interest tends to be higher in regions where they are more frequently caught. For snow crab and Alaska pollock, a strong positive correlation was found (Figs S6.3 and S6.4 in S1 File), indicating that search volume is significantly higher in regions with greater catch volumes. For the remaining three species, no significant correlation was observed, suggesting that they are widely searched across Japan, regardless of regional catch volumes (Figs S6.5-S6.7 in S1 File).

3.4 Analysis of related search keywords

Across all seven fish species, search queries were primarily related to price information and cooking methods. Additionally, distinct cultural, regional, and seasonal factors were reflected in the search keywords for each species (S1-S7 Tables). “Setsubun” recurrently accompanied sardine queries (reflecting the hiiragi-iwashi custom); horse mackerel showed limited keyword diversity with modest regional-branding interest; squid and snow crab terms often overlapped with related species and processed products (and, for crab, strong e-commerce intent); Alaska pollock queries emphasized product names and regional name variants; and mackerel queries centered on processed or recipe-specific items (e.g., canned, miso-simmered). Notably, Pacific saury queries increasingly referenced catch declines and poor harvests, indicating heightened public concern over rapid stock depletion.

3.5 Path analysis

Path analysis demonstrated a strong model fit for Japanese horse mackerel, Japanese sardine, Pacific saury, and mackerel species (Table 1). The estimated relationships are illustrated in Fig 2. The results indicated that market price at the central wholesale market primarily influenced search volume across these species. Additionally, Pacific saury was the only species where fishery catch volume had a significant, albeit small, impact on search volume. For Japanese flying squid, the model fit was lower compared to other species. However, both CPI and market price contributed to search volume, with CPI having a greater influence than market price, suggesting that consumer price fluctuations play a more significant role in search behavior for Japanese flying squid.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Model fit indices for path analysis.

https://doi.org/10.1371/journal.pone.0342833.t001

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. Results of path analysis.

(i) Horse mackerel, (ii) Japanese sardine, (iii) Pacific saury, (iv) Mackerel, (v) Japanese flying squid. Statistically significant paths (p < 0.05) are indicated with higher opacity. Path analysis was applied to monthly data from 2010 to 2023 for all species.

https://doi.org/10.1371/journal.pone.0342833.g002

1. (i). Japanese Horse Mackerel
2. Search volume (RSV_R) exhibited a significant negative association with market price (MP_R) (β = −0.129, p = 0.03), indicating that a decline in market price leads to an increase in search volume. In contrast, the effects of fishery catch volume (MC_R) and consumer price index (CPI_R) were not statistically significant (Fig 2. (i)).
3. (ii). Japanese Sardine
4. Search volume (RSV_R) was significantly influenced by market price (MP_R) (β = −0.093, p = 0.04), indicating that a decrease in market price is associated with an increase in search volume. Meanwhile, fishery catch volume (MC_R) and CPI (CPI_R) had no statistically significant effects (Fig 2. (ii)).
5. (iii). Pacific Saury
6. Search volume (RSV_R) exhibited a significant negative association with market price (MP_R) (β = −0.216, p = 0.02), confirming that a decline in market price is associated with an increase in search volume. Additionally, fishery catch volume (MC_R) had a significant positive effect on search volume (β = 0.133, p = 0.042), suggesting that an increase in catch volume leads to a rise in public interest. Conversely, CPI (CPI_R) did not have a significant effect (Fig 2. (iii)).
7. (iv). Mackerel Species
8. Search volume (RSV_R) was significantly influenced by market price (MP_R) (β = −0.211, p = 0.01), indicating that a decline in market price is associated with an increase in search volume. However, fishery catch volume (MC_R) and CPI (CPI_R) did not exhibit significant effects (Fig 2. (iv)).
9. (v). Japanese Flying Squid
10. Search volume (RSV_R) exhibited significant negative associations with both CPI (CPI_R: β = −0.636, p < 0.001) and market price (MP_R: β = −0.161, p = 0.04), suggesting that a decrease in price is associated with an increase in search volume. However, the effect of fishery catch volume (MC_R) was not statistically significant, indicating a limited contribution to search behavior (Fig 2. (v)). Additionally, the model fit indices were relatively low, suggesting room for improvement in the model’s explanatory power.

3.6 Integrative summary

Taken together, these results position Pacific saury as a distinctive outlier among the seven species. Like Japanese horse mackerel, saury exhibits seasonal synchrony between catch and search peaks, yet it diverges from the others along three dimensions. First, saury is price-positive at the monthly scale (higher search interest when CPI is higher), with CPI often remaining elevated around the September peak when catch, demand, and supply rise simultaneously. Second, in the path analysis, saury is the only species for which fishery catch volume has a significant direct (albeit small) positive effect on search interest, beyond the common negative price effect—indicating that availability itself amplifies public attention. Third, keyword patterns increasingly reference catch declines/poor harvests, signaling stock-status concerns that are less pronounced for other species. Consistent with this picture, prefecture-level coupling between catch and searches—strong for snow crab and Alaska pollock and weak for horse mackerel and sardine—was not evident for saury, suggesting nationally diffuse interest decoupled from local landings. Altogether, saury’s combination of seasonal synchrony, price-positive coupling, catch-driven attention, and stock salience marks it as qualitatively different from species whose search interest is primarily associated with lower prices or localized supply conditions.

4.1 What drives public interest?

4.2 Can pacific saury serve as a flagship species?

Based on the findings above, Pacific saury has emerged as a species that both reflects the recent deterioration of fishery resources and commands broad public attention, making it well suited as a flagship for outreach and management communication. Despite an overall downward trend, saury remains one of the most-searched species (second only to mackerel), and prefecture-level analyses show minimal regional disparity—indicating nationwide recognition rooted in Japan’s fish-eating culture and in its role as an indicator of the sector’s decline. Unlike most species, saury exhibits distinct supply–demand–price dynamics: whereas higher catches typically coincide with lower CPI and increased searches, saury shows its highest CPI and search volume in September, when both supply and demand peak, suggesting that domestically harvested autumn saury is ascribed special seasonal value rather than being treated merely as a low-cost staple. This interpretation aligns with evidence that food-related information seeking is stimulated not only by economic cues but also by health/safety, convenience, cultural/affective meanings, and ethical considerations [47]. Taken together, this combination of national salience, cultural meaning, and economic relevance provides practical leverage for using saury as a flagship species to focus communications, frame stock-status messaging, and time seasonal campaigns that promote sustainable fisheries.

The concept of flagship species has traditionally focused on charismatic megafauna such as tigers, elephants, and pandas [27,32]. More recently, flagship species have been increasingly recognized as strategic marketing tools aimed at influencing consumer behavior and fostering conservation efforts [32,47]. Building on this idea, the concept of “flagship events” extends the flagship-species approach by treating attention-grabbing natural or human-driven events as comparable levers; for example, recurrent seasonal phenomena can function as flagship events that draw audiences into biodiversity challenges and potentially mobilize support [48]. From this perspective, Pacific saury, as both a cultural icon and a species reflecting the challenges of fishery resource management, is well suited for media campaigns promoting sustainable fisheries. Species awareness-day type campaigns can measurably increase online information seeking and, when paired with explicit calls to action, can also strengthen engagement outcomes [49].

Considering these factors, Pacific saury has strong potential to serve as a flagship species for promoting sustainable fisheries in Japan. Its high societal recognition, cultural significance, economic value, and consumer interest in its declining catch volumes make it a compelling candidate for raising awareness of sustainable fisheries. However, sustaining long-term public interest and linking this awareness to policy initiatives require concrete strategic efforts. This is particularly relevant given persistent shortfalls in biodiversity-related media attention often without deliberate framing and amplification [50]. By leveraging the findings of this study and positioning Pacific saury as a flagship species for sustainable fisheries, broader societal awareness and policy-driven initiatives can be effectively advanced.

Unlike many terrestrial flagships that can prompt direct, hands-on stewardship, a pelagic and widely traded food fish is unlikely to mobilize on-site conservation activities. If adopted as a flagship, Pacific saury would instead function through communication, markets, and governance: concentrating seasonal messaging, guiding menus and retail cues toward responsible options, encouraging sustainable consumption choices, reinforcing the perceived value and legitimacy of sustainable fisheries, and enlarging public attention from saury to additional species of concern. Social media should be treated as a central dissemination channel alongside search and news media, given its documented influence on food choice, purchase intention, and consumption behavior [51]. In this sense, the primary contribution lies in shaping awareness, preferences, and support for management measures rather than eliciting direct intervention on the ground.

5.1 Policy and management implications

5.2 Limitations and future work

Despite these limitations, the approach remains low-cost, scalable, and readily transferable. Properly timed to predictable peaks and adapted to local contexts, this methodology can support evidence-based outreach and policy design. Ultimately, integrating societal data into resource management is not just a technical exercise but a necessary step toward fostering a more engaged and informed public, capable of stewarding our shared marine heritage into the future.

Furthermore, a specific limitation of our study is that the path analysis model for Japanese flying squid did not achieve a good fit, unlike the models for other species. This suggests that the simple causal structure we hypothesized may not fully capture the complex drivers of public interest for this particular species. The demand for Japanese flying squid is heavily tied to the processed food industry, for example through dried-squid products, and its price and availability may be influenced by factors beyond domestic catch and consumer prices, such as import volumes and processing-industry trends. Future research—both for Japanese flying squid and more generally across species—should explore alternative model specifications and incorporate additional explanatory variables, including media-coverage intensity; social-media activity and engagement metrics; campaign and event indicators; market-side factors such as import volumes and processing-industry conditions; and, when relevant, indicators of seafood safety and environmental risks such as pollution incidents, contamination advisories, or food-safety concerns. Qualitative approaches, including semi-structured interviews with consumers, retailers and market actors, and fishery stakeholders, can be used to strengthen and contextualize the quantitative findings by clarifying how and why particular species attract public attention, especially in terms of perceived cultural and culinary value, consumption contexts, and their potential positioning as flagship species for sustainability communication.