Experimental design and crop establishment
A microcosm experiment was conducted at Yamagata University, Tsuruoka City, Japan, from May to October 2019, with six growth containers (36 cm in height, 30 cm in width, 60 cm in length) to simulate paddy fields of 0.18 m2 in area (see Supplementary Fig. S1). The experiment was laid out in a completely randomized design with three replications of two treatments: (1) rice cropping under CSI and (2) conventional rice cultivation fertilized with mineral fertilisers and irrigated with tap water (Control).
Each container was filled with 32 kg of a paddy soil collected from an experimental field in the university farm and transplanted with four hills of 30-day-old seedlings (Oryza sativa L., cv. Bekoaoba) on 27th May 2019. The experiment was performed in accordance with relevant guidelines and regulations for research involving plants. The experimental soil was classified as loamy soil (air-dried, 20% moisture) with the following basic properties: pH (H2O) of 5.78, electrical conductivity (EC) of 0.09 dS m−1, SOM of 4.9%, and a total N, P, and K of 1.46, 0.88, and 3.17 g kg−1, respectively. The TWW used in the CSI system was collected from a local WWTP and monitored weekly for its basic properties (Table 2) following our previous studies6,7. In brief, pH, EC, and DO of water samples were measured on-site using pH/conductivity and DO portable meters (D-54 and OM-51, HORIBA, Ltd., Kyoto, Japan), whereas TOC and total N were analyzed using a TOC analyzer (TOC-VCSV, Shimadzu Corp., Kyoto, Japan) attached to a total N measuring unit (TNM-1, Shimadzu Corp., Kyoto, Japan). After a standard acid-digestion of water samples6, the concentration of P was measured using a portable colorimeter (DR/890, HATCH, USA), and the concentration of K was measured using an inductively coupled plasma mass spectrometry (ICP-MS ELAN DRCII, PerkinElmer Japan Co., Ltd.). The tap water used in this study was also tested on a regular basis and found to be stable throughout the crop season, with the following properties: pH of 7.8, EC of 0.095 dS m−1, DO, TOC, N, and P of 6.85, 0.49, 0.06, and 0.07 mg L−1, respectively, with K being below the ICP-MS detection limit (< 0.3 mg L−1).
The CSI system has been described thoroughly in our previous study6. Briefly, TWW was supplied to the experimental containers through underground perforated pipes from which TWW infiltrated the soil layers and eventually overflowed out of the simulated paddy fields (see Supplementary Fig. S1). The optimal irrigation rate suggested in our prior investigation was adopted as follows: a flow rate of 25 L m−2 day−1 was implemented during the active tillering and reproductive periods, and a lower rate of 8.3 L m−2 day−1 was implemented at the early tillering and late ripening stages6. Irrigation with TWW was initiated 3 DAT and maintained continuously throughout the crop season. No exogenous fertiliser was applied with CSI treatment. In the Control, rice plants were conventionally supplemented with high doses of mineral fertilisers for basal (160 kg N–P2O5–K2O ha−1) and topdressing fertilisation (100 kg N–K2O ha−1), while daily irrigation was conducted by manually adding tap water to maintain 5 cm of standing water above the soil surface. Consistent with both treatments, irrigation was withheld from 43 to 52 DAT for a mid-season drainage (MSD). The rice plants were harvested on 1st October 2019 and rice yields were assessed as the weight of brown rice at 15% moisture.
CH4 emission measurement
The static closed chamber method31 was used to sample the gas as described in our previous study6. Briefly, before the gas sampling, six transparent static chambers were mounted securely on the growth containers at the water-seal collars (see Supplementary Fig. S1). Gas samples were extracted from the chambers 0, 15, and 30 min after the chamber deployment and immediately transferred to the laboratory for CH4 measurement using a gas chromatograph (GC-2014, Shimadzu Corp., Kyoto, Japan). The gas was collected from around 10:00 a.m. to 11:00 a.m. once a week throughout the crop season, while the sampling frequency was increased to once every 2 days during the MSD period.
Daily fluxes and seasonal cumulative emissions of CH4 were calculated following the standard procedures as described previously31. In order to correlate the cumulative emission of CH4 to the rice production, the yield-scaled emission was calculated using the following equation:
$$ {text{YSE}} = frac{{{text{CE}}}}{{text{Y}}}, $$
where, YSE, CE and Y represents the yield-scaled emission (kg CH4 t−1 yield), the cumulative emission (kg CH4 ha−1) and the yield of brown rice (t ha−1), respectively.
Soil sampling
The paddy soils were sampled at six time-points, which were representative of different soil conditions and growth stages of the rice plants during the crop season: pre-transplantation, maximum tillering, panicle initiation, booting, grain filling, and ripening stages (0, 43, 52, 64, 107, and 126 DAT, respectively). Samples were collected in each container using a soil core (1.5 cm in diameter × 15 cm in depth) from which the upper and lower layers of the paddy soil (0–1 and 12–15 cm, respectively) were collected separately. The respective soil layers from three replicates were mixed to form one composite sample of each treatment, placed in a 50 mL commercial centrifuge tube (VWR SuperClear Ultra-High Performance Freestanding Centrifuge Tubes with Flat Caps, VWR international, USA) and then transferred to the laboratory and immediately stored at − 80 °C.
DNA extraction, PCR, and quantitative PCR assays
Soil DNA was extracted from 24 samples (2 treatments × 2 layers × 6 time points) using 0.25 g of the frozen soils as the input for the DNeasy PowerSoil Kit (QIAGEN, Hilden, Germany) following the manufacturer’s instructions. Next, DNA concentration was determined using Qubit 4 Fluorimeter (Thermo Fisher Scientific, Waltham, Massachusetts, USA) and the total DNA extracts were stored at − 80 °C until further analysis.
The 24 total DNA extracts were all used for Illumina MiSeq 16S rRNA gene sequencing. Amplicon library preparation and Illumina MiSeq sequencing were performed by Fasmac Co., Ltd. (Atsugi, Japan). The universal primers U515F (5′-ACACTCTTTCCCTACACGACGCTCTTCCGATCTGTGCCAGCMGCCGCGGTAA-3′) and U806R (5′-GTGACTGGAGTTCAGACGTGTGCTCTTCCGATCTGGACTACHVGGGTWTCTAAT-3′) were used following a previous study32. They targeted the V4 hypervariable regions of the archaeal and bacterial 16S rRNA genes and were used for the first 16S rRNA gene amplification33. Operational taxonomic units (OTUs) were defined at the sequence similarity level of 97% and a representative sequence from each OTU was assigned to a taxonomic identity using the Quantitative Insights into Microbial Ecology (QIIME) software package34. Compositions of the archaeal and bacterial communities (at the genus level) in the soil samples were determined by classifying the taxa of each OTU using the Greengenes database at Fasmac Co., Ltd., Japan.
In order to evaluate the total load of archaeal and bacterial communities in the soil samples, two real-time PCR (qPCR) assays targeting the 16S rRNA genes of the archaeal and bacterial DNA were performed, respectively, using a CFX96 Touch Real-Time Detection System (Bio-Rad Laboratories, Inc. Hercules, CA, USA). The primers and probes used for the assays were previously developed35,36. The standards (102–106 gene copies) for the qPCR assays of archaea were prepared using the strain NBRC110930 Haloarchaeobius iranensis, while those for the qPCR assay of bacteria were prepared using the strain NBRC3301 Escherichia coli K12. The reaction conditions for the amplification of archaea were as follows: 95 °C for 10 min, followed by 40 cycles at 95 °C for 20 s and 60 °C for 1 min; in contrast, the conditions for bacterial amplification were as follows: 50 °C for 2 min and 95 °C for 10 min, followed by 50 cycles of 95 °C for 15 s and 60 °C for 1 min. The abundance of methanogenic and methanotrophic genera was calculated based on their proportions compared to the quantity of the total archaeal and bacterial communities in the soil samples.
Statistical analysis
Differences in CH4 emissions between the examined treatments were evaluated using Student’s t tests at a significance level of 0.05. The Pearson correlation analysis was carried out between CH4 fluxes and the abundance of methanogenic and methanotrophic communities, using the data derived from the 6 time-points across the treatments. All statistical analyses were performed using IBM SPSS Statistics 24.0.
Source: Ecology - nature.com
