Rice paddy-field microbial fuel cells (RPF-MFCs, Figure 1) are sediment-type MFCs that convert soil organic matter into electricity with the aid of rhizosphere microbes [1] [2] . Different from other sediment-type MFCs [3] ,

RPF-MFCs utilize organic matter excreted from rice roots (root exudates [4] ) for electricity generation. RPF- MFCs are expected to be applied to on-site power sources for remote sensors that facilitate smart agriculture. A previous study has shown that rice roots excrete sugars (e.g., glucose) and organic acids (e.g., acetate) that can be converted to electricity by anode-associated microbes [5] . Another study combined phylogenetic and metagenomic approaches for characterizing anode-associated microbes, suggesting that bacteria affiliated with the genus Geobacter (particularly those closely related to G. pelophilus and G. psychrophilus) were involved in electricity generation in RPF-MFCs [6] . Another previous study has examined factors affecting electricity generation in RPF-MFCs and found that cathode modification with platinum catalysts, anode position and external load largely affected power outputs [7] . In addition, it has also been found that graphite-felt cathodes equipped with polystyrene-foam floaters enable relatively stable and high power outputs from RPF-MFCs [2] . Based on these examinations, the maximum power and short-circuit current densities have reached 80 mW∙m⁻² and 550 mA∙m⁻², respectively (based on the projected anode area) [2] . Compared to the maximum power density (6 mW∙m⁻²) reported in the first trial of RPF-MFCs in 2008 [5] , the latest power density (80 mW∙m⁻²) has increased over 10-fold, showing that the RPF-MFC technology has progressed substantially.

Despite the substantial improvement achieved in the recent studies [2] [7] , there still exist unexamined factors that may be important for further increasing electric outputs from RPF-MFCs. In the present work, we attempted to identify which electrode (namely, anode or cathode) is the limiting factor for electricity generation in RPF-MFC. For this purpose, RPF-MFCs were equipped with different sizes of anodes and cathodes, and power outputs were measured during the operation.

In order to examine effects of anode and cathode performances on power outputs from the total RPF-MFC system, RPF-MFCs with different sizes of anodes and cathodes (Table 1) were operated, and their power indices (P_max and Q_max values) were compared (Figure 2). For a reactor-type MFC, polarization analyses using a reference electrode is possible for evaluating electrochemical performances of anodes and cathodes [8] . However, since a reference electrode did not work in soil, we employed the method in which RPF-MFC systems with different sizes of anodes and cathodes were comparatively analyzed. The operation of the RPF-MFCs was initiated when rice-plant seedlings were transplanted and terminated when rice crops were harvested (30 August). During the operation, polarization measurements were conducted several times to examine temporal changes in power outputs (Figure 2).

We found that LE generated larger power than SE (Figure 2(a)). Power densities for LE were however smaller than those for SE (Figure 2(b) and Figure 2(c)), probably because more organics were supplied from rice roots to anode areas close to rice plants than those far from plants. We also found that P_max values for SE and LE were not dependent on operational periods (Figure 2(a)). On the other hand, P_max values for the AL and CL systems were dependent on operational periods; namely, P_max for AL on day 93 was much higher than that on day 20, while vice versa for CL. These data indicate that the anode performance increases during the operation of RPF-MFC, while the cathode performance decreases. It is likely that the anode improvement and cathode deterioration similarly affected the P_max values in the SE and LE systems, resulting in the relatively stable P_maxvalues.

In the present study, the highest Q_max value (approximately 140 mW∙m⁻²) was observed for the AL system on day 93 (Figure 2(b)). This value for the AL system is much higher than that for the SE system, even though these MFC had the same size of anode. In addition, this is the highest Q_max among values (based on the projected area of anodes) so far reported for RPF-MFCs (refer to Ref. 2 for the summary of previous data), suggesting that the large cathode is effective for enhancing the anode-based Q_max value. It is therefore concluded that the cathode needs to be improved for eliciting the maximum capacity of rhizosphere bacteria for generating electricity at the anode of RPF-MFC.

Previous studies have suggested that bacteria affiliated with the genus Geobacter are involved in electron transfer to anodes in RPF-MFC systems [6] and plant-associated MFCs in pot-culture systems [12] [13] . Comparative metagenomics of anode-associated microbiomes showed that bacteria closely related to G. pelophilus and G. psychrophilus are involved in electricity generation [6] . In order to examine if Geobacter bacteria also occurred in anode-associated microbiomes, the present study conducted phylogenetic analyses of 16S rRNA gene fragments PCR-amplified from anode microbiomes. Figure 3(a) presents phylogenetic distributions at the family level of bacteria in bulk soil and those in anode-associated microbiomes in the AL and CL systems. It is shown that, compared to Geobacteraceae bacteria in the bulk soil (<0.1%), those in the AL (2.4%) and CL (1.8%) anode samples increased substantially. Comparative analyses of Geobacteraceae sequences retrieved in the present study and those in the databases indicate that those closely related to G. pelophilus [14] and G. psychrophilus [15] were major members in the AL and CL anode sample (Figure 3(b)), suggesting that these Geobacter bacteria were involved in electricity generation in the RPF-MFC system. This result therefore supports the previous finding that G. pelophilus and G. Psychrophilus relatives play important roles in electricity generation in RPF-MFC.

The present study suggests that the cathode is the limiting factor for electricity generation in RPF-MFC after anode microbiomes are sufficiently enriched. The cathode performance is found to be decreased during the operation, probably owing to deterioration of the cathode catalysts. In order to sustainably improve the cathode performance, future studies will develop methods to effectively utilize biocathodes [16] [17] for RPF-MFC.

We thank Noda Natural Symbiotic Farm Co. for managing the rice paddy field. We also thank Ayako Matsuzawa and Nanako Amano for technical assistance.

NagayoshiUeoka,NaokoSese,MayuSue,AtsushiKouzuma,KazuyaWatanabe, (2016) Sizes of Anode and Cathode Affect Electricity Generation in Rice Paddy-Field Microbial Fuel Cells. Journal of Sustainable Bioenergy Systems,06,10-15. doi: 10.4236/jsbs.2016.61002