Introduction

Nature-based solutions (NbS) are defined as “Actions to protect, sustainably manage, and restore natural or modified ecosystems, that address societal challenges effectively and adaptively, simultaneously providing human well-being and biodiversity benefits” (International Union for Conservation of Nature and Natural Resources 2016). NbS have attracted increasing attention as the complexity of societal issues intensifies. A unique flood control method called the paddy field dam (PFD) is one example of an NbS in Japan.

Among the multiple functions of a paddy field, the PFD utilizes the storage function (Ministry of Agriculture, Forestry and Fisheries 2022). In the PFD approach, farmers control the drainage rate and temporarily store rainfall on paddy fields. As a result, the risk of flood damage to residents in downstream areas is reduced. The cooperation of farmers is essential to maximize the effectiveness of this approach.

Several studies have examined the quantitative and economic characteristics of the PFD. Yoshikawa et al. (2009) revealed that peak main river and canal discharge was decreased by 25–29% when PFDs were used, and Miyazu et al. (2013) reported that the value of PFD activity was approximately 112,000 yen/ha/year. In terms of social aspects, Tamura et al. (2018) investigated factors influencing farmer participation in PFD initiatives via questionnaires and interviews, and Oishi et al. (2019) investigated awareness and perceptions of PFDs in rural communities. However, no previous studies have focused on the fact that only people living downstream from a PFD benefit from this system, even though it is primarily the service of farmers in upstream areas who usually receive few benefits (Yoshikawa et al. 2011). Furthermore, no studies have examined differences in perceptions of the PFD between these two groups of stakeholders. This information is necessary for determining how to promote PFD activity.

To address this issue in the present study, we used a questionnaire survey to examine gaps in awareness and expectations regarding the flood control effect of PFDs.

Methods

Study area

This study was conducted in the Fuchu-machi of Toyama city in Japan. The Ida River, which is a tribute of the Jinzu River, flows through this area. Several small tributes of the Ida River, such as the Tsubono and Iso Rivers, flow through the study area (Fig. 1). When the water level of the Ida River rises as the result of heavy rainfall, discharge from these tributes cannot drain into the Ida River. This causes inundation floods both above and below floor level.

Fig. 1
figure 1

Study area and its schematic diagram. Map source: Geospatial Information Authority of Japan

Full size image

In response to this situation, the Fuchu Area Paddy Field Dam Promotion Council was formed in 2012 in the Fuchu area, and the PFD activity has been implemented in the upstream area of the Tsubono and Iso Rivers (Noda et al. 2022). In this area, a subsidy from Toyama city supports the implementation of PFDs, in addition to multifunctional grants from institutions that provide economic incentives to individuals who implement PFDs (Ministry of Agriculture, Forestry and Fisheries 2023). As of 2020, there were 155 members involved in PFD activities, including farmers and farming corporations, with 384.47 ha of paddy fields included in the program (Noda et al. 2022).

Questionnaire survey

We administered a questionnaire survey to the farmers and beneficiaries in the study area. The farmers were all regular members of the JA-Aoba, which is the local agricultural association, and the beneficiaries were members of households in Sasakura and Hayahoshi (Fig. 1). A total of 1273 and 1967 farmers and beneficiaries responded to the survey, respectively. The questionnaires were directly distributed to farmers at the annual JA-Aoba meeting at the end of January 2022 and to beneficiaries at regular meetings of the neighbor council in each area in mid-February 2022. We asked the respondents to return the answer sheets via mail by the end of March 2022. In order to promote better understanding, brief explanatory material on the PFD was distributed together with a questionnaire.

The questionnaire was multiple choice and included questions about demographic characteristics, PFD awareness, and expectations regarding PFD-related reductions in river flow and flood damage during a flood. The demographic information included gender, age, hometown, years in the area, farming activity related to paddy fields, and flood experience. Awareness of the PFD was reported using three response options: “Know well,” “Sounds familiar,” and “Did not know.” Expectations regarding PFD-related reductions in river flow and flood damage during a flood were reported using five response options: “More than 90%,” “About 75%,” “About 50%,” “About 25%,” and “Less than 10%.”

Screening of the responses was conducted prior to data analysis. We removed samples with invalid answers, such as multiple answers to a single question or an unreasonable combination of answers, as determined via cross-tabulation.

After screening, the valid samples were divided into three groups: beneficiaries, farmers who had implemented the PFD, and farmers who had not implemented the PFD.

The data analysis was based on three factors: awareness of the PFD, expectations regarding PFD-related reductions in river flow and flood damage during a flood, and the relationship between awareness and expectations regarding the PFD.

In terms of PFD awareness, we calculated the differences between the actual and expected values using the Chi-square test and then conducted a residual analysis for each factor.

In terms of expectations regarding PFD-related reductions in river flow and flood damage during a flood, differences in the population mean were tested using Welch’s t test. The expectation levels for those who answered “More than 90%” and “Less than 10%” were calculated as 95% and 5%, respectively. Moreover, we compared the expectations given by the questionnaire respondents with predications of the actual values calculated using a simplified estimation method (Yoshikawa 2022). This method estimates the potential of the peak cut rate based on the areal rate of a paddy field in the target catchment.

For the calculation, we subdivided the study area into an upstream area that included the Iso River catchment and a downstream area that included the Tsubono River catchment.

Finally, another cross-tabulation was applied to examine the relationship between the level of awareness and expectations regarding the PFD.

Results and discussion

Questionnaire collection and respondent profiles

The number of responses and the result of demographic characteristics are summarized in Table 1 and Fig. 2. The total number of collected response forms was 918, of which 415 were from beneficiaries and 503 were from farmers. The total number of valid responses was 821, of which 381 were from beneficiaries and 440 were from farmers.

Table 1 Group classification and number of responses
Full size table
Fig. 2
figure 2

Profile of respondents: a gender, b age, c hometown, d years lived in the area (for immigrants), e flood experience

Full size image

Regarding the gender of the respondents, 61.4% of the beneficiaries were males and 38.6% were females. As for the farmers, the majority of respondents, 92.5% were males and 7.5% were females. In terms of age, 0.5% of the beneficiaries were below 19 years, 11.0% were in their 20–30s, 29.4% were in their 40–50s, 48.8% were in their 60–70s, and 10.2% were in their 80s or older. No farmer was aged 19 or younger, 0.5% were in their 20–30s, 13.6% were in their 40–50s, 75.9% were in their 60–70s, and 10.0% were in their 80s or older. The highest proportion of both beneficiaries and farmers were in their 60–70s. Over half of the beneficiaries, 55.4%, were from the study area. Among farmers, the majority, 93.2%, were from the study area, while only 6.8% had migrated from other areas. Concerning these migrants, the largest proportions of both beneficiaries and farmers had already lived in the study area for more than 30 years, 33.7% of beneficiaries and 76.7% of farmers, respectively. Among beneficiaries, 18.9% had lived in the study area for 10–20 years, followed by 17.8% who had lived there for less than 5 years. For farmers, those who had lived in the study area for 10–20 years and those who had lived in the study area for 20–30 years followed with 10.0% each. With respect to flood experience, the largest proportions of both beneficiaries and farmers had never experienced flooding, 77.2% and 89.1%, respectively. Beneficiaries had a slightly higher proportion of those who had experienced flooding compared to farmers, with 15.5% having experienced flooding below floor level and 7.3% having experienced flooding above floor level. As for farmers, 8.0% had experienced flooding under the floor and 3.0% had experienced flooding above the floor.

Regarding the responses from farmers, 146 were from those who had implemented the PFD, and 294 were from those who had not implemented the PFD. Beneficiaries accounted for 46.4% of the total number of valid responses and farmers who had/had not implemented the PFD accounted for 17.8% and 35.8% of the total responses, respectively.

Awareness of the PFD

The responses regarding awareness of the PFD are shown in Fig. 3. Our analysis revealed that 59.6% of the beneficiaries answered that they “Did not know” about the PFD, while 47.7% of the farmers responded that they “Know [the PFD] well.” The Chi-square test and residual analysis indicated that the farmers were significantly more aware of the PFD than the beneficiaries.

Fig. 3
figure 3

Level of awareness of the PFD. Fy farmers who have implemented the PFD, Fn farmers who have not implemented the PFD

Full size image

However, 29.9% of the farmers who had not implemented the PFD, which is equivalent to 20% of the total number of farmers, were still unaware of the existence of the PFD. Although PFD activity has been implemented in our target area for 10 years, the level of awareness, especially for the beneficiaries and farmers who have not implemented the PFD, is still limited.

Table 2 compares the values obtained in the present study with those of previous studies. In comparison with Oishi et al. (2019), the proportion of respondents who answered that they were “aware” (“Well known,” “Little known,” “Sounds familiar”) of the PFD was greater in our study. One reason for this may be that a higher proportion of farmers, who are expected to have a high level of awareness of the PFD, were surveyed in our study (53.6% in our study, 17.1–21.3% in Oishi et al. (2019)). The largest respondent group in both studies comprised those who answered that they had never heard of the PFD.

Table 2 Percentage of responses in each study on the level of awareness (%) and elapsed time from the start of the activity to the conduct of the questionnaire survey (year)
Full size table

A comparison with Tamura et al. (2018), which examined a level of awareness among a group of farmers who could be considered as not implementing PFD, confirmed a high degree of similarity with our study. It should be noted that Tamura et al. (2018) conducted their questionnaire survey during the year when PFD activity was first considered in the area, whereas our study was conducted 10 years after the start of the PFD activity in our target area. These results imply that PFD awareness is not widespread among farmers who have not implemented the PFD, i.e., those who have the potential to implement the PFD.

With regard to the flood experience, significant differences were identified among beneficiaries. Beneficiaries with flood experience (n = 87) were found to have a significantly higher level of awareness of PFD than those without flood experience (n = 294) (p = 0.002). This trend was not observed among farmers.

Expectations regarding the effect of the PFD

Expectations regarding the PFD-related reduction in river discharge and flood damage during a flood are shown in Fig. 4. The beneficiaries had the highest expectations among the three groups for both effects, with an average of 39.9% and 45.7% reduction in river discharge and flood damage, respectively. Farmers who had implemented the PFD had the next highest levels of expectation, and farmers who had not implemented the PFD had the lowest expectations. We found no significant difference between the beneficiaries and the farmers who had implemented the PFD for both variables. However, the farmers who had not implemented the PFD had significantly lower expectations than the other two groups, with an average of 29.0% reduction in river discharge and a 32.7% reduction in flood damage. Farmers as a whole expected a 31.8% reduction in river discharge during floods and a 35.9% reduction in flood damage on average.

Fig. 4
figure 4

Degree of expected effect for the PFD. Each plot represents the mean value of the degree of expectation in each group, and the bar represents the 95% confidence interval of the mother mean. Fy farmers who have implemented the PFD, Fn farmers who have not implemented the PFD. *α = 0.0167

Full size image

The respondents’ experience of flooding was found to have an impact, particularly with regard to their expectations of the reduction in flood damage. Among beneficiaries, those with flood experience (n = 87) expected a 51.4% reduction, significantly higher than the 44.0% expected by those with no flood experience (p = 0.013).

A similar trend was observed among farmers, with those with flood experience (n = 48) expecting 43.2%, significantly higher than the 34.9% expected by those with no flood experience (n = 392) (p = 0.046).

The difference between experienced and non-experienced flood respondents was not significant in terms of their expectations of the effect of reducing peak flows during floods. Additionally, for farmers, no significant difference was detected in whether the presence or absence of flood experience influenced the implementation of the PFD.

Table 3 shows the catchment area, paddy area, and rate of paddy field area in each area. Figure 5 plots the relationship between the paddy field area rate and the river flow peak cutoff rate with additional points highlighting the upstream area, downstream area, and overall target area.

Table 3 Area of each catchment and paddy field area, and percentage of paddy field area
Full size table
Fig. 5
figure 5

Relationship between paddy field area rate and peak flow reduction rate. The solid black line is obtained using the simplified estimation method of Yoshikawa (2022)

Full size image

We found that the paddy field area rate on the downstream side of the study area including the Tsubono River was 27.2%, with an assumed river flow peak cutoff rate of 16.6%. The paddy field area rate on the upstream side of the basin including the Iso River was 59.3%, with an assumed river flow peak cutoff rate of 40.4%. In the study area as a whole, the paddy field area rate was 46.1% and the assumed river flow peak cutoff rate was 30.0%.

After comparing these values with the expectations of the beneficiaries and farmers (as calculated from the questionnaire survey; see Fig. 5), the mean expectation level of the farmers (31.8%) was close to the estimated peak cutoff rate for the entire target area, while the mean expectation level of the beneficiaries (39.9%) was excessive compared with the actual assumed value.

In particular, the peak cutoff rate in the downstream area, where most of the beneficiaries lived, was low at 16.6%. Given that the estimated value was calculated based on the assumption that all of the paddy fields had installed the PFD devices, the real effect in the current situation would likely be smaller than the estimated value. This would increase the gap between the estimated value and the expected value given by the beneficiaries.

Relationship between awareness and expectations regarding the PFD

To clarify the relationship between the level of awareness and expectations, especially the PFD-related reduction in river discharge, we developed a cross-table (Table 4). In addition, for the beneficiaries and farmers as a whole, the data were plotted in a box plot (Fig. 6) to illustrate the trends.

Table 4 Level of expectations of different awareness groups regarding the effectiveness of the PFD in river flow reduction in flood (%)
Full size table
Fig. 6
figure 6

Relationship between awareness and expectation of the PFD. Each plot represents the mean value of the level of expectation in each group, and the bar represents the 95% confidence interval of the mother mean

Full size image

Among the beneficiaries, the expected benefit of the PFD was lowest among those who answered that they “Know [the PFD] well,” at 33.1%. This proportion was significantly lower than that of beneficiaries who reported that they “Did not know” about the PFD. These data indicate that as the level of awareness increases among beneficiaries, the expectations decrease, as confirmed by the data shown in Fig. 6.

Conversely, among the farmers, the expected benefit of the PFD was highest among those who answered that they “Know [the PFD] well.” This value was significantly higher than that of those who “Did not know” about the PFD. These data indicate that among all farmers, as awareness increases, so too do expectations. This trend was also seen in the group of farmers who had not implemented the PFD.

Moreover, the expectations held by the beneficiaries and farmers in the “Know well” category were similar.

These results imply that the gap in the level of expectations between the beneficiaries and farmers could be narrowed by enhancing awareness of the PFD in the study area.

It should be noted, however, that the effects of the PFD are limited. The expectations of the beneficiaries and farmers in the “Know well” category (33.1% and 36.0%, respectively) were higher than the above-mentioned calculated value from the simplified estimation method (30.0%). Thus, to ensure that people obtain a full understanding of the PFD, information about both its limitations and effectiveness should be communicated.

Conclusion

This study examined differences in levels of awareness and expected effects among stakeholders regarding the PFD. We found that farmers had a higher awareness regarding the PFD than beneficiaries. However, about 30% of the farmers who had not implemented the PFD were not aware of its existence, indicating inadequate dissemination of information. With regard to the effects of the PFD, the results showed an opposite trend, with beneficiaries having a higher level of expectation than the farmers. Moreover, farmers who had not implemented the PFD had low expectations. A comparison between these data and the estimated effects showed that the expectations held by the farmers as a whole were almost the same as the estimated effects in the target area; the beneficiaries conversely had excessive expectations. Among the beneficiaries, the level of expectation decreased as the level of awareness increased, and for the farmers, the level of expectation increased as the level of awareness increased.

One challenge facing the PFD activity is the uncertainty among stakeholders. This study clarified the perceptions of the stakeholders and indicated that the gap between farmers and beneficiaries in the expected effects could be narrowed by increasing awareness. Our comparison between the expected and estimated effects suggests that, when spreading awareness, it is necessary to provide accurate information about the effects of the PFD based on the actual situation. Although this study is comparable to previous ones, it was a case study of one region. Therefore, similar studies in other regions are needed to ensure external validity.