Evaluation of groundwater recharge by rice and crop rotation fields in Kumamoto, Japan

Rice farming can have beneficial influences on the surrounding environment by providing multifunctionality in terms of flood mitigation (Masumoto et al. 2006; Yoshikawa et al. 2010), groundwater recharge (Tanaka et al. 2010; Maruyama et al. 2014; Yoshioka et al. 2016), and water purification (Tabuchi & Kuroda 1991; Onishi et al. 2012). The magnitude of multifunctionality is determined by both natural and anthropological factors. For example, Kuroda et al. (2006) reported that paddy fields can act as a nitrogen sink when nitrogen concentration in inflow water is high. Water management practices also influence the function of paddy fields in the environment. For example, intermittent irrigation decreases surface runoff and nutrient loss (Liang et al. 2013). Reuse of drainage water can enhance the function of nitrogen removal in paddy fields (Hama et al. 2011).

Kumamoto in Japan is a large city, which supplies domestic water for one million citizens with natural groundwater. About 6 × 10⁸ m³/year of the groundwater is recharged from permeable areas in the Kumamoto region. In particular, paddy fields (rice farming) make a large contribution to the groundwater recharge because the fields are artificially flooded for about four months. Paddy fields in the middle area of Shira River, which flows through the Kumamoto region, recharge about 0.8 × 10⁸ m³/year of groundwater (Takemori & Ichikawa 2008). However, the amount of groundwater resources in the region is tending to decrease due to the expansion of the urban area and decreasing paddy fields. Therefore, local governments in Kumamoto have tried to increase the amount of groundwater recharge in the agricultural area. In particular, farmers in the middle Shira River area can receive subsidies as part of a measure by the local governments if they flood their fields during the fallow period of upland crops. This measure is considered as a good example of integrated watershed management because the environmental sectors in the local governments collaborate with the agricultural sectors to manage the groundwater resources. However, there are some problems with the measure. First, rice farmers cannot receive any financial return although rice farming makes a large contribution to the groundwater recharge too. On the other hand, water is used excessively or inefficiently to flood the fallow fields because the farmers who cultivate upland crops have no motivation to manage water carefully in the fallow fields. Then, the careless water management causes a shortage of irrigation water in the paddy-field districts, especially in the downstream area. In addition, global climate change may exacerbate the current situation as follows: (i) the water supply will be critical because the number of no rain days tends to increase in Kumamoto; and (ii) water demand in the paddy fields may increase because the rising air temperature leads to the increase of evapotranspiration. In other words, inappropriate measures for the groundwater resources can cause water shortage on the surface (groundwater recharge area).

Previous research (e.g., Yamane et al. 2003; Shimada 2013) has focused on the innovative aspect of the political measures. There is no research that evaluates the negative effect of the measures. The objective of this study is to evaluate the risk of artificial flooding of the fallow fields.

Kumamoto City is located in the center of Kyushu Island, which is one of four major islands and the most western one in Japan. The city and its surrounding cities and towns are referred to as Kumamoto region (Figure 1). The region covers 1,041 km². Shira River flows through the region from east to west. In the upstream of the river watershed, there is an active volcano, Mt. Aso. About 600 km² of impermeable layer forms the groundwater basin under the Kumamoto region. About 6 × 10⁸ m³ of water is annually recharged from the surface land. In particular, a paddy-field district in the middle area of Shira River watershed has significant potential to recharge groundwater because the soils have high permeability. Some paddy plots have high permeability of more than 300 mm/d (Takemori & Ichikawa 2008). However, the area of paddy field in the district is decreasing due to the expansion of the urban and residential area. The number of rice farmers is also decreasing gradually in the middle Shira River area because of the economic situation in Japan. Crop rotation and cultivation of upland crops such as carrots and potatoes in paddy fields is implemented in the district. The decrease of paddy fields and rice farming has resulted in the decrease of the groundwater resources in Kumamoto.

Kumamoto City government has developed an administrative measure to enhance groundwater recharge from the paddy-field district in the middle Shira River area. The governmental office decided to give subsidies (\110,000/ha) to farmers in the crop-rotated paddy fields if they artificially flooded fields for more than one month during the fallow period. This measure increased the flooded areas and may contribute to the recovery of the groundwater resources. The flooded areas and their groundwater recharge are roughly estimated as 500 ha and 1.5 × 10⁷ m³. However, overuse of irrigation water is observed in those areas. This may be because the farmers do not have any motivation to take care of the flooded fields during the fallow period. Then, the careless water management results in a shortage of irrigation water in the downstream area.

The annual mean temperature and rainfall in the middle Shira River area is 16.1 °C and 2,263 mm, respectively (Japan Meteorological Agency 2018). A paddy plot was investigated.

If ΔQ_j > Q_j, it was judged that water shortage occurred in the segment. The location, number, and area of rice or crop rotation fields with water shortages were estimated in each simulation. The location of rice and crop rotation fields was randomly set in each simulation. Ten thousand simulations were conducted under a certain condition. In this study, it is assumed that the frequency of water shortage at each field indicated the drought risk.

Shira River is the main water resource for the paddy-field district. Figure 3 shows the return period of flow volume of the river, which is estimated using observation data of 1993–2016 (Ministry of Land Infrastructure Transport & Tourism 2018). The frequency of low flow volume was higher in August than in June and July. In this study, the probability distribution of the flow volumes of irrigation water was not considered because the flow volumes of irrigation water were not strongly correlated with the river flows. However, the flow volume of irrigation water was stepwisely changed in the simulation. The total volume of irrigation water was estimated from the water level, which was constantly observed at the start point of the irrigation canal in the paddy-field district. Then, the mean value of the flow volumes at the high water level was used as the maximum volume of irrigation water.

Temporal variations in the inflow and outflow in the study plot are shown in Figure 4. Irrigation is normally conducted on non-rainy days. The rate of irrigation water ranged from 20 mm/d to 300 mm/d. In general, the paddy soil is dried for a short term in mid-summer to supply oxygen to the rice roots. After the season, water demand by the paddy soil increases. Accordingly, the maximum rate was 329 mm/d, which was observed in early August.

Surface runoff was observed even on some non-rainy days, suggesting that irrigation was ineffective. This may be due to high rates of irrigation. In recent years, older farmers are tending to excessively irrigate once to save time and effort. Our results clearly suggest that the amount of irrigation water increases due to the careless irrigation. In other words, the amount of irrigation water would be decreased by effective use of rainfall through careful irrigation. However, farmers using public irrigation systems have little incentive to use irrigation water effectively (Yoon et al. 2003; Yoon et al. 2006).

The water balance of the study plot is shown in Table 2. The total amount of irrigation during the observation period was 6,139 mm. The total amount of rainfall during the irrigation period was 940 mm. On the other hand, total amounts of surface runoff and evapotranspiration were 2,977 mm and 396 mm, respectively. Consequently, the total amount of percolation, which was estimated from the water balance, was 3,706 mm.

The percolation rate of the paddy plot was approximately 33 mm/d (=3,706 mm/114 day). This value is higher than that in other low-lying paddy fields, which ranges from 10 to 20 mm/d. Considering that the mean value of evapotranspiration on fine days was 5.5 mm/d, the rate of water loss of the paddy plot, which was calculated as the sum of evapotranspiration and percolation, was roughly 40 mm/d. On the other hand, the mean value of irrigation water during the normal irrigation period was estimated as 54 mm/d. It is indicated that excess irrigation was conducted in the paddy plot. About half of the irrigation water was discharged as runoff water. Although the fields with high percolation rates have large potential to recharge groundwater, farmers have to intake large amounts of irrigation water to keep flooding the fields. Then, if they do not control the water intake carefully, part of the irrigation water will be discharged as surface runoff.

Figure 5 shows the spatial distribution of water shortage risk in the downstream area (about 59 ha) of the paddy-field district under the condition that the proportion of the crop rotation field was 70% and the supply of irrigation water decreased 60% from the maximum level. It is confirmed that the risk of water shortage was high in the downstream area.

The simulation suggested that water shortage would occur if the supply of irrigation water decreased 11% from the maximum available level. The risk curve of water shortage is shown in Figure 6. The vertical axis means the percentage of the total area of paddy plots with water shortage to that of the paddy-field district. The increase of the crop rotation fields increases the risk of water shortage. For example, under the condition that the supply of irrigation water is 60% of the maximum available level, the proportion of the fields with water shortage increases from 4% to 33% when the proportion of the crop rotation field increases from 40% to 100%. Therefore, it is suggested that the artificial flooding of the crop rotation fields may lead to water shortage. However, the situation can be improved by efficient water use, in which surface runoff is reduced.

Figures 7 and 8 show temporal variation in annual mean air temperature in the Kumamoto region and no rain days in the Aso caldera, the upstream area of the Shira River watershed, respectively. The current trends of the local climate are not desirable for us. The rising air temperature may lead to the increase of evapotranspiration in the paddy fields. In the Kumamoto region, the total amount of the potential evapotranspiration in the rice growing season (June–September) is estimated to be approximately 590 mm by the Thornthwaite method. It would be 638 mm if the air temperature increased by 1 °C. In addition, the increase of no rain days may lead to the decrease of the river flow. The climate change does not improve but exacerbates the current situation because the water demand in the paddy fields increases and the water supply from the river decreases.

Although rice is a staple food in Japan, rice cultivation has been tending to decrease due to the economic conditions. On the other hand, crop rotation is increasing in paddy fields in Japan. Our research field is one of the typical cases in Japan. Our research suggests that the artificial flooding of the crop-rotation fields (fallow fields), which is implemented to increase the groundwater resources, can cause water resource problems on the ground surface. Water shortage impedes rice cultivation and may promote crop rotation. In other words, the land use change can be accelerated by inappropriate water use, which may result from the land use change and the local climate change.

The artificial flooding of the crop rotation field during the fallow period has large potential to recharge groundwater. However, our observation reveals that a large amount of irrigation water was used ineffectively. Due to the inefficient water use, the risk of water shortage in the paddy-field district increases when the area of the crop rotation field increases. Surface runoff caused by the over intake of irrigation water should be reduced to achieve a good balance between the rice field (rice cultivation) and the crop-rotation field (groundwater recharge).

We thank Mr Otaguro and Mr Maeda for providing access to the paddy plot for our investigation and for providing information on water management and farming activities in the plot. This study was partly funded by a Grant-in-Aid for Scientific Research from the Japan Society for the Promotion of Science.