This is the accepted version of the article: Wang, Weiqi; Sardans i Galobart, Jordi; Wang, Chun; [et al.]. «Effects of Fertil- ization on Porewater Nutrients, Greenhouse-gas Emissions and Rice Productivity in a Subtropical Paddy Field». Experimental agriculture, Vol. 55, Issue 3 (June 2019), p. 395-411. DOI 10.1017/S0014479718000078 This version is avaible at https://ddd.uab.cat/record/204027 under the terms of the license 1 EFFECTS OF FERTILIZATION ON POREWATER NUTRIENTS, 2 GREENHOUSE-GAS EMISSIONS AND RICE PRODUCTIVITY IN A 3 SUBTROPICAL PADDY FIELD 4 5 By WEIQI WANG1,2*, JORDI SARDANS3,4*, CHUN WANG1,2, CHUAN TONG1,2, 6 QINYANG JI1,2, JOSEP PEÑUELAS3,4 7 1 8 Institute of Geography, Fujian Normal University, Fuzhou 350007, China 2 9 Key Laboratory of Humid Subtropical Eco-geographical Process, Ministry of 10 Education, Fujian Normal University, Fuzhou 350007, China 3 11 CSIC, Global Ecology CREAF- CSIC-UAB, Cerdanyola del Valles, 08193 12 Barcelona, Catalonia, Spain 4 13 CREAF, Cerdanyola del Valles, 08193 Barcelona, Catalonia, Spain 14 15 Short title: Fertilization and greenhouse-gas emissions 16 17 *Corresponding author: 18 Weiqi Wang. Email: wangweiqi15@163.com; Phone: +86-591-83465214; 19 Fax: +86-591-83465397 20 21 SUMMARY 22 23 Suitable fertilization is crucial for the sustainability of rice production and for the 24 potential mitigation of global warming. The effects of fertilization on porewater 25 nutrients and greenhouse-gas fluxes in cropland, however, remain poorly known. We 26 studied the effects of no fertilization (control), standard fertilization and double 27 fertilization on the concentrations of porewater nutrients, greenhouse-gas fluxes and 28 emissions and rice yield in a subtropical paddy in southeastern China. Double 29 fertilization increased dissolved NH4+ in porewater. Mean CO2 and CH4 emissions 30 were 13.5% and 7.4%, 20.4% and 39.5% higher for the standard and double 31 fertilizations than the control, respectively. N2O depositions in soils were 61% and 32 101% greater for the standard and double fertilizations than the control, respectively. 33 The total global warming potentials (GWPs) for all emissions were 14.1% and 10.8% 34 higher for the standard and double fertilizations than the control, respectively, with 35 increasing contribution of CH4 with fertilization and a CO2 contribution > 85%. The 36 total GWPs per unit yield were significantly higher in the standard and double 37 fertilizations than the control by 7.3% and 10.9%, respectively. The two levels of 38 fertilization did not significantly increase rice yield. Prior long-term fertilization in the 39 paddy (about 20 years with annual doses of 95 kg N ha-1, 70 kg P2O5 ha-1 and 70 kg 40 K2O ha-1) may have prevented these fertilizations from increasing the yield. However, 41 fertilizations increased greenhouse-gas emissions. This situation is common in paddy 42 fields in subtropical China, suggesting a saturation of soil nutrients and the necessity 43 to review current fertilization management. These areas likely suffer from 44 unnecessary nutrient leaching and excessive greenhouse-gas emissions. These results 45 provide a scientific basis for continued research to identify an easy and optimal 1 46 fertilization management. 47 Keywords: Paddy field; CH4 flux; N2O flux; CO2 flux; porewater nutrient; fertilizer 48 49 INTRODUCTION 50 51 Greenhouse gases (GHGs), such as carbon dioxide (CO2), methane (CH4) and 52 nitrous oxide (N2O), contribute about 80% to the current global radiative forcing 53 (Myhre et al., 2013). Agricultural activities contribute approximately 20% of the 54 present concentrations of atmospheric GHGs (Hütsch, 2001), especially the emissions 55 of CH4 and N2O from paddy fields (Myhre et al., 2013), so minimizing the release of 56 these GHGs from paddies and thus mitigating their adverse impacts on climate change 57 are of the utmost importance. As a main cereal crop, rice currently feeds more than 58 50% of the global population (Haque et al., 2015) and rice production will need to 59 increase by 40% by the end of 2030 to meet the demand for food from the growing 60 population worldwide (FAO, 2009). 61 Fertilization is important for sustainable rice production (Linquist et al., 2013) 62 and numerous studies have been devoted to the development of suitable practices of 63 fertilizer management for both improving rice yields and mitigating GHG emissions, 64 including the application of fertilizers such as straw mulch (Dossou-Yovo et al., 2016), 65 silicate fertilizer (Wang et al., 2015) and the establishment the most adequate N, P and 66 K fertilizers (Thao et al., 2015). Fertilizers are key for rice productivity, but the 67 amounts needed for maintaining rice yield and minimizing environmental effects are 68 unknown. Chemical fertilizers are necessary to keep the world’s population fed, but 69 their overuse threatens the safety of the soil, water and air (Zhang et al., 2013a). 70 China has become the world's largest consumer of fertilizer, and its fertilizer-use 71 efficiency is much less than half the amendment amount (Cheng and Li, 2007). 72 Determining the amount suitable for supporting rice productivity and reducing 73 nutrient losses is therefore very important (Zhao et al., 2015). 74 Chemical fertilizers, especially N fertilizers, influence the dynamics of GHGs the 75 most (Zhong et al., 2016). Zhao et al. (2015) reported that N2O emissions increased 76 but CH4 emissions decreased with the level of fertilization. Zhong et al. (2016) 77 indicated that the global warming potential (GWP) and rice yields increased with the 78 rate of application of nitrogen fertilizer. In contrast, Zhang et al. (2013a) showed that 79 N fertilization could reduce GHG emissions. Some studies have showed that 80 maintaining soil fertility and crop productivity and at the same time reducing GHGs 81 emissions have several trade off questions to take into account (Bhatia et al., 2005). 82 The substitution of inorganic N fertilizers by organic crop manure residues improves 83 yield and soil health but can increase CH4 emissions (Bhatia et al., 2005; 2012). At 84 medium-long term, the substitution of inorganic by organic residues as fertilizers 85 improve soil aggregate stability, soil water holding capacity and soil microbial activity 86 (Sihi et al, 2016a; 2017) without any serious decrease in yield production (Sihi et al., 87 2012). Thus, establishing suitable fertilization rates to ensure rice yields and reduce 88 GHG emissions is important for field management (Zhong et al., 2016). The control 89 of water inundation management and the time to plant and sowing rice plants (Dari et 2 90 al., 2017) or the use of crop manure and/or urea plus dicyanamide have reduced N2O 91 emissions in paddy soils without reducing yield (Pathak et al., 2002) and increased the 92 recovery efficiency of N added by fertilizers (Pathak 2010). Anyway, fertilization 93 management influences soil properties and this affects GHGs emissions in wetland 94 areas (Davidson and Janssens, 2006), and there is not a general consensus in the 95 adequate fertilization management for an equilibrated soil fertility and rice production 96 without increased GHGs emissions. 97 China has the world's second largest area of rice cultivation, and the associated 98 GHG emissions account for about 40% of the total agricultural GHG emission. In 99 China, 90% of paddies are in the subtropics, such as in Fujian, Jiangxi and Hunan 100 Provinces. Developing effective strategies to increase the cost-effectiveness of rice 101 agriculture, enhancing crop yield and mitigating GHG emissions from paddies in 102 subtropical China to minimize future food shortages and adverse climate change is 103 thus a global objective of national importance. We pursued this objective by: (1) 104 determining the emissions of CO2, CH4 and N2O in response to the application of 105 different amounts of fertilizers in paddy fields; (2) exploring the effect of amendment 106 amount on the concentrations of porewater nutrients; and (3) assessing the impacts of 107 the fertilizer applications on crop productivity. The results obtained in this study will 108 provide a scientific basis for the suitable management of fertilization for rice 109 agriculture and the evaluation of most current fertilization strategies for increasing 110 yield and decreasing GHG emissions. 111 112 MATERIALS AND METHODS 113 114 Study site 115 Our study was conducted at the Wufeng Agronomy Field of the Fujian Academy of 116 Agricultural Sciences in Fujian Province, southeastern China (26.1°N, 119.3°E) 117 (Supplementary Material Fig. S1). A field experiment was carried out during the early 118 paddy season (16 April to 16 July) in 2014. Air temperature and humidity during the 119 study period are shown in Fig. S2. The soil of the paddy was poorly drained, and the 120 proportions of sand, silt and clay particles in the top 15 cm of the soil were 28%, 60% 121 and 12%, respectively (Wang et al., 2015). Other properties of the top 15 cm of soil at 122 the beginning of the experiment were: bulk density, 1.1 g cm-3; pH (1:5 with H2O), 123 6.5; organic carbon (C) content, 18.1 g kg-1; total nitrogen (TN) content, 1.2 g kg-1 124 and total phosphorus (TP) content, 1.1 g kg-1 (Wang et al., 2015). Water level was 125 maintained at 5-7 cm above the soil surface throughout the growing season by an 126 automatic water-level controller, and the paddy was last drained two weeks before 127 harvest. 128 We established triplicate plots (10 × 10 m) for two treatments and a control in 129 which rice seedlings (cultivar Hesheng 10) were inserted to a depth of 5 cm with a 130 spacing of 14 × 28 cm using a rice transplanter. The age of the rice seedlings was 21 131 days when they were transplanted. The two treatments consisted of standard 132 fertilization and fertilization with double amount of the standard fertilization. The 133 control plots were not fertilized. The standard fertilization treatment consisted of three 3 134 applications of N-P2O5-K2O (16-16-16%; Keda Fertilizer Co., Ltd. Jingzhou, China) 135 and urea (46% N) fertilizers. The first application was one day before transplantation 136 at rates of 42 kg N ha-1, 40 kg P2O5 ha-1 and 40 kg K2O ha-1. The second application 137 was broadcasted during tiller initiation (seven days after transplanting (DAT)) at rates 138 of 35 kg N ha-1, 20 kg P2O5 ha-1 and 20 kg K2O ha-1. The third application was 139 broadcasted during panicle initiation (56 DAT) at rates of 18 kg N ha-1, 10 kg P2O5 140 ha-1 and 10 kg K2O ha-1. These doses are the most commonly used in paddy fields in 141 subtropical China (with annual doses of 95 kg N ha-1, 70 kg P2O5 ha-1 and 70 kg K2O 142 ha-1), which constitute about 90% of the paddy fields in China (Wang et al., 2015). 143 The double-fertilization treatment consisted of twice the amounts of the standard 144 fertilization but the same schedule. The plots had previously been fertilized with 145 amounts equal to the standard fertilization during a period of about 20 years (Wang et 146 al., 2015). All control and treatment plots received the same amount of water. The 147 field was plowed to a depth of 15 cm with a moldboard plow and was leveled two 148 days before rice transplantation. 149 150 Measurement of CO2, CH4 and N2O emissions 151 Static closed chambers were used to measure CO2, CH4 and N2O emissions. The 152 chambers were made of PVC and consisted of two parts, an upper transparent 153 compartment (100 cm high, 30 cm wide, 30 cm long) placed on a permanently 154 installed bottom collar (10 cm high, 30 cm wide, 30 cm long). Each chamber had two 155 battery-operated fans to mix the air inside the chamber headspace, an internal 156 thermometer to monitor temperature changes during gas sampling and a gas-sampling 157 port with a neoprene rubber septum at the top of the chamber for collecting gas 158 samples from the headspace. We deployed three replicate chambers in each plot. A 159 wooden boardwalk was built for accessing the plots to minimize the disturbance of 160 the soil during gas sampling. The chambers were installed on soil with plants inside.. 161 Gas flux was measured weekly in all chambers. Gas samples were collected from 162 the chamber headspace using a 100-mL plastic syringe with a three-way stopcock 0, 163 15 and 30 min after deployment of the upper compartment. The samples were 164 immediately transferred to 100-mL air-evacuated aluminum foil bags (Delin Gas 165 Packaging Co., Ltd., Dalian, China) sealed with butyl rubber septa and transported to 166 the laboratory for the analysis of CO2, CH4 and N2O. 167 168 CO2, CH4 and N2O concentrations in the headspace air samples were determined 169 by gas chromatography (Shimadzu GC-2010 and Shimadzu GC-2014, Kyoto, Japan) 170 using a stainless steel Porapak Q column (2 m long, 4 mm outer diameter, 80/100 171 mesh). A methane conversion furnace, flame ionization detector (FID) and electron 172 capture detector (ECD) were used for the determination of the CO2, CH4 and N2O 173 concentrations, respectively. The operating temperatures of the column, injector and 174 detector for the determination of CO2, CH4 and N2O concentrations were adjusted to 175 45, 100 and 280 °C; to 70, 200 and 200 °C and to 70, 200 and 320 °C, respectively. 176 These temperatures were the optimum temperatures for the different parts of the 177 instrument. Helium (99.999% purity) was used as a carrier gas (30 mL min-1), and a 4 178 make-up gas (95% argon and 5% CH4) was used for the ECD. The gas chromatograph 179 was calibrated before and after each set of measurements using 503, 1030 and 2980 180 μL CO2 L-1 in He; 1.01, 7.99 and 50.5 μL CH4 L-1 in He and 0.2, 0.6 and 1.0 μL N2O 181 L-1 in He (CRM/RM Information Center of China) as primary standards. CO2, CH4 182 and N2O fluxes were then calculated as the rate of change in the mass of CO2, CH4 183 and N2O per unit of surface area and per unit of time by using a closed-chamber 184 equation (Ali et al., 2008): 185 F= ((M/V) x (dc/dt)) x H x (273/(273 + T)), where F is the corresponding gas 186 flux (mg/mg m-2 h-1), M is the molecular weight, V is the height of the chamber above 187 water surface (m), and T is the air temperature inside the chamber (ºC). 188 . 189 190 Global warming potential (GWP) 191 GWP is typically estimated using CO2 as the reference gas, and a change in the 192 emission of CH4 or N2O is converted into “CO2-equivalents” (Hou et al., 2012). The 193 GWP for CH4 is 34 (based on a 100-year time horizon and a GWP for CO2 of 1), and 194 the GWP for N2O is 298 (Myhre et al., 2013). The GWP of the combined emission of 195 CH4 and N2O was calculated as (Ahmad et al., 2009): GWP = (cumulative CO2 196 emission × 1 + cumulative CH4 emission × 34 + cumulative N2O emission × 298). 197 198 Collection of soil porewater 199 Porewater was sampled in situ once a week from 16 April to 16 July 2014. Three 200 specially designed PVC tubes (5.0 cm inner diameter) were installed to a depth of 15 cm 201 in each plot. Porewater samples were collected using 50-mL syringes and then separated 202 into two parts: about 10 mL were injected into pre-evacuated 20-mL vials, and the 203 remaining 40 mL were injected into 100-mL sample vials. The samples were stored in a 204 cool insulated box in the field until transported to the laboratory where they were 205 stored at -20 °C until the analysis of nutrients and GHG concentrations. 206 207 Measurement and calculation of dissolved CO2, CH4 and N2O concentrations 208 The sample vials were thawed at room temperature and then vigorously shaken for 209 5 min to equilibrate the CH4 concentrations between the water and the headspace. Gas 210 samples were collected from the headspaces of the vials and analyzed for CO2, CH4 211 and N2O concentrations by gas chromatography (Shimadzu GC-2010 and Shimadzu 212 GC-2014, Kyoto, Japan; see above for more details). 213 The concentrations (C) of CO2, CH4 and N2O dissolved in the water were 214 calculated following Ding et al. (2003): C = (Ch × Vh)/(22.4 × Vp), where Ch is the 215 CO2, CH4 and N2O concentration (μL L-1) in the air samples from the vials, Vh is the 216 volume of air in the vial (mL) and Vp is the volume of the water in the vial (mL). 217 218 Measurement of soil, porewater properties and rice yield 219 Soil temperature, pH, salinity, redox potential (Eh) and water content of the top 15 220 cm of soil were measured in triplicate in situ in each plot on each sampling day. 221 Temperature, pH and Eh were measured with an Eh/pH/Temperature meter (IQ 222 Scientific Instruments, Carlsbad, USA), salinity was measured using a 2265FS EC 223 meter (Spectrum Technologies Inc., Paxinos, USA) and water content was measured 5 224 using a TDR300 meter (Spectrum Field Scout Inc., Aurora, USA). The concentrations 225 of NH4+, NO3-, TN and TP in the porewater were determined using a sequence flow 226 analyzer (San++, SKALAR Corporation production, Breda, The Netherlands). The 227 concentration of dissolved organic carbon (DOC) was determined using a TOC 228 Analyzer (TOC-V CPH, Shimadzu Corporation, Kyoto, Japan) and a filter paper of 60 229 m of pore diameter. Rice yield was determined at the harvesting stage by manual 230 harvest (Wang et al., 2015). 231 232 Statistical analysis 233 Differences in soil properties; CO2, CH4 and N2O emissions; dissolved porewater 234 CO2, CH4 and N2O concentrations and porewater nutrient concentrations among the 235 treatments and control were tested for statistical significance by repeated-measures 236 analyses of variance. The relationships between mean GHG emissions and the soil 237 properties, dissolved porewater GHG concentrations and porewater nutrient 238 concentrations were determined by Pearson correlation analysis. These statistical 239 analyses were performed using SPSS Statistics 18.0 (SPSS Inc., Chicago, USA). We 240 analyzed the effects of multiple soil variables as fixed factors on the production rates 241 of the three GHGs using general linear models with and without spatial correlation. 242 We used linear (lm) and mixed (lme) functions with the “nlme” and “lme4” R 243 packages. We chose the best model for each dependent variable using the Akaike 244 information criterion. We used the MuMIn R package in mixed models to estimate the 245 percentage of variance explained by the model. 246 247 RESULTS 248 249 GHGs dissolved in porewater and emitted from the paddy 250 CO2, CH4 and N2O emissions varied significantly among most sampling dates 251 (P<0.01, Table S1) but the treatments and the interaction of sampling date and 252 treatment were not significant (P>0.05). CO2 flux generally remained low (< 354 mg 253 m-2 h-1) during the first 29 DAT but then increased to a seasonal peak (> 2811 mg m-2 254 h-1) between 29 and 71 DAT (Fig. 1A). The rice was nearly ripe by 71 DAT, with a 255 corresponding decrease in CO2 emissions until harvesting in July. The CH4 emissions 256 were low soon after rice transplantation and peaked by 71 DAT in all treatments (Fig. 257 1B). The paddy was drained after the rice reached maturity, with a corresponding 258 decrease in CH4 emissions until harvesting in July. 259 Mean CO2 emissions were 13.5% and 7.4% higher for the standard and double 260 fertilizations than the control, respectively. Mean CH4 emissions were 20.4% and 39.5% 261 higher for the standard and double fertilizations than the control, respectively. Mean 262 N2O soil depositions were 61% and 101% higher for the standard and double 263 fertilizations than the control, respectively, mostly due to the lower (negative) values 264 at 36 DAT (Fig. 1C), despite no overall effect of the treatments on N2O emission 265 determined by the mixed linear models (Table 1). Dissolved CO2, CH4 and N2O 266 concentrations varied significantly among sampling dates (P<0.01, Table S1). 267 Treatments and the interaction of date and treatment were significant for dissolved 268 CO2 concentration but not for dissolved CH4 and N2O concentrations (P>0.05; Table 269 S1, Fig. 2). 6 270 Mean dissolved CO2 concentration was 11.4% lower for the standard fertilization 271 and 95.0% higher for the double fertilization than the control. Mean dissolved CH4 272 concentrations were 25.8% and 18.9% lower for the standard and double fertilizations 273 than the control, respectively. Mean dissolved N2O concentrations were 21.0% and 274 73.5% higher for the standard and double fertilizations than the control, respectively. 275 The mixed linear models (with plot and time as random factors) showed that 276 standard fertilization increased CO2 emissions and soil Eh and salinity, whereas 277 double fertilization increased CH4 emission, porewater CO2 concentration, soil 278 salinity and porewater NH4+ and DOC concentrations (Table 1). This mixed linear 279 model analysis is more robust to detect the effects of treatments than the previously 280 commented ANOVA. CO2 emission was correlated positively with soil Eh, 281 temperature, water content and porewater NH4+ concentration and negatively with soil 282 pH and porewater TN, TP and DOC concentrations (Table S2). 283 284 Soil and porewater properties of the paddy 285 Soil Eh, salinity and porewater NH4+, TP and DOC concentrations, soil pH, 286 temperature, water content and porewater NO3- and TN concentrations varied 287 significantly among sampling dates (P<0.01; Table S3, Figs. 3 and 4). The interaction 288 of sampling date and treatment had significant effects on soil pH, temperature, water 289 content and porewater NO3- and TN concentrations (P<0.05). Mean soil pH, Eh, 290 salinity, water content and temperature for the standard and double fertilizations 291 differed by <10% from the control. Mean porewater NH4+ concentrations were 114.8% 292 and 213.7% higher for the standard and double fertilizations than the control, 293 respectively. Mean porewater NO3- concentrations were 17.8% and 30.9% higher for 294 the standard and double fertilizations than the control, respectively. Mean porewater 295 TN concentrations were 19.7% and 42.0% higher for the standard and double 296 fertilizations than the control, respectively. Mean porewater TP concentrations were 297 45.7% and 213.3% higher for the standard and double fertilizations than the control, 298 respectively. Finally, mean porewater DOC concentrations were 9.3% and 11.7% 299 higher for the standard and double fertilizations than the control, respectively. 300 301 Relationships among GHG emissions, dissolved GHG concentrations and soil and 302 porewater properties 303 The mixed linear models (with plot and time as random factors) showed that CH4 304 emissions were positively correlated with soil water content (Table S2). N2O emission 305 was correlated positively with porewater total TN concentration and negatively with 306 soil pH. Porewater CO2 concentrations were correlated positively with soil salinity 307 and water content and negatively with porewater TN, TP, NH4+ and DOC 308 concentrations. Porewater CH4 concentrations were positively correlated with soil 309 water content, and porewater N2O concentrations were positively correlated with soil 310 temperature and porewater NH4+ concentrations. 311 Soil relationships varied between the treatments (Tables S4 and S5). Seasonal 312 CO2 emission was positively correlated with soil Eh (P<0.05, Table 2) and 313 temperature (P<0.01, Table S4) in all plots. Seasonal CH4 emission was positively 314 correlated with soil salinity (P<0.01) and water content (P<0.05) in all plots. 7 315 Dissolved CO2 concentration was positively correlated with soil water content 316 (P<0.05), while dissolved N2O concentration was negatively correlated with soil 317 temperature (P<0.05, Table S4) in all plots. 318 Seasonal CO2 emission was positively correlated with CH4 emission (P<0.05, 319 Table S5) and dissolved CO2 concentration (P<0.01). Seasonal CH4 emission was 320 positively correlated with dissolved CH4 concentration (P<0.01) and dissolved CO2 321 concentration was positively correlated with dissolved CH4 concentration (P<0.01). 322 323 Rice productivity and GWP 324 The average rice yield was about 5.2% higher for the standard fertilization and 325 about 1.4% lower for the double fertilization than the control, which were not 326 significantly different (Table 2). Contribution of CO2 to total GWP (> 85%) was 327 higher than CH4 and N2O. The total GWPs for all emissions were 14.1% and 10.8% 328 higher for the standard and double fertilizations than the control, respectively. The 329 total GWPs per unit yield were significantly higher by 7.3% and 10.9% for the 330 standard and double fertilizations than the control, respectively. 331 332 DISCUSSION 333 334 Effects of fertilization on CO2 emissions 335 Mean CO2 emissions were higher for the standard and double fertilizations than the 336 control, for several potential reasons. First, fertilization, such as N fertilization, 337 promotes the deposition of photosynthetically derived C into soil organic carbon 338 pools. Then, soil respiration increases when inputs of active C substrates increase (Ge 339 et al., 2015). Second, fertilizer can provide many nutrients for microbial growth 340 (Inselsbacher et al., 2011), and the increase in microbial activity promotes soil 341 respiration and thus CO2 emission (Adewopo et al., 2015). Third, NH4+ from 342 fertilizers can be oxidized to NO 3- when paddies are drained, increasing soil NO 3- 343 concentration. This NO3- would be reduced when the paddies are reflooded, 344 producing CO2 (Wang et al., 2015). Moreover, NH4+ amendment in our study would 345 be associated with ferric reduction, increasing the production and release of CO 2 346 (Luo et al., 2016). Ferric reduction should also decrease the number of iron plaques 347 (by the higher solubility of Fe2+ than Fe3+) on the rice roots, which would increase the 348 transport of gases throughout the rice plants (Huang et al., 2012). Transport by rice 349 plants is the most important pathway of gas emission to the atmosphere (Wassmann 350 and Aulak, 2000). Decreases in the number of iron plaques will promote root 351 ventilation, so more CO2 is produced and transported through the internal system of 352 interconnected gas lacunae in plants. The positive correlation between soil redox 353 reactions and CO2 emission is consistent with this result. 354 CO2 emission varied seasonally, increasing with rice growth and temperature. 355 Temperature controls CO2 production and emission by not only increasing soil 356 microbial activity (Vogel et al., 2014) or controlling their physiology (aka carbon use 357 efficiency) (Allison et al., 2010; Frey et al., ; Kivlin et al., 2013; Sihi et al., 2017), but 358 also by altering plant respiration (Slot et al., 2013), substrate availability and quality, 359 species composition, water availability and aerobic/anaerobic conditions (Davidson 8 360 and Janssens, 2006; Inglett et al., 2012; Sihi, 2015; Sihi et al., 2016a; 2016b). Higher 361 temperatures increase soil CO2 emissions in subtropical wetlands (Inglett et al., 2012). 362 C quality influences primarily SOM decomposition at low temperatures while at 363 high temperatures nutrient availability controls SOM decomposition in subtropical 364 wetlands (Sihi et al., 2016b). In our study we have obtained consistent results with 365 these previous reports. Soil CO 2 concentration in porewater increased with 366 temperature especially in double fertilized soils (Figs. 3 and 4), showing higher 367 organic matter decomposition with temperature mainly in double N fertilized soils. 368 Given that the major fraction of rice croplands in southeastern Asia corresponds to 369 puddled/wetland conditions, understanding complex interactions in such 370 environments is important for improving our capacity for future projections under 371 warming climate for both natural and agricultural systems. 372 373 Effects of fertilization on CH4 emissions 374 Mean CH4 emissions were 20.4% and 39.5% higher for the standard and double 375 fertilizations than the control, respectively. As stated above, fertilization promotes the 376 deposition of photosynthetically derived C into SOC pools (Ge et al., 2015). Such 377 C can contribute up to 52% of the CH 4 emissions from paddy soils by the exudation 378 of labile organic C from roots to the rhizosphere, which will then produce methane. 379 These results are consistent with the lack of fertilization effects on rice yield. While 380 fertilization can enhance photosynthesis, more photosynthates are allocated to root 381 exudates and this reduces allocation to growth and yield. The other 48% of the CH 4 382 is emitted from old soil C (Minoda et al., 1996), promoting CH4 production and 383 emissions (Minoda et al., 1996). Fertilization, especially N fertilization, will also 384 increase the availability of nutrients, which will promote CH 4 production and 385 emissions from microbes (Naik et al., 2015). Since N fertilizer was provided in NH4+ 386 form, it could have inhibited CH4 oxidation because of the structural resemblance of 387 NH4+ with CH4 promoting enzymatic substrate competition (Gulledge and Schimel, 388 1998). Nevertheless, most studies testing for different methods and substances for N 389 fertilization have observed enhancement of CH4 production and emission (Pathak, 390 2010), which is in agreement with our results. 391 CH4 emission varied seasonally and emissions were low soon after rice 392 transplantation when soil was not strictly anaerobic. The correlation of soil redox 393 reactions with CH4 emission supported this finding (Table S3). CH4 emissions were 394 also low during the final ripening and drainage periods. These results agreed with 395 those by Minamikawa et al. (2014), in which a lowering of the water table decreased 396 the abundance of the methanogenic archaeal population and hence CH4 production 397 and increased the abundance of methanotrophs and thus CH4 oxidation. 398 399 Effects of fertilization on N2O emissions 400 N2O emission was low throughout the growing season, with no obvious pattern 401 of seasonal variation (Fig. X). The paddies in our study region are strongly N limited 402 (Wang et al., 2015), so together with low levels of soil O2, most of the N2O produced 403 is likely reduced to N2 and this would lead to very low emissions or even a net uptake 9 404 of N2O (Zhang et al., 2010). Pulses in ammonium and nitrate availability after a 405 fertilization have been related to N2O production (Pathak et al., 2002). Specific N2O 406 fluxes and the contribution of nitrifying and denitrifying bacteria are controlled 407 mainly by soil moisture (Davidson et al., 1993). However, the results of our study 408 showed that the N added by fertilization has not been sufficient to rise N2O emissions 409 in this paddies with low N concentrations. 410 411 Best management practices to reduce GWP 412 Our results suggested that the application of fertilizer increased the impacts of rice 413 agriculture on climate change, with higher total GWPs per unit yield compared to the 414 controls. The fertilizations did not significantly increase the rice yield but they 415 increased the soil porewater nutrient concentrations, which has the potential risk of 416 nutrient loss, eutrophication and higher costs. Judicious use of fertilization should 417 be reconsidered in a sustainable agriculture and our results provide strong evidence 418 that the current strategy of fertilization in most rice croplands in subtropical China 419 over several years will saturate the soil fertility, increasing the release of nutrients 420 to continental water and favoring CH 4 and CO2 production and emission without 421 increasing rice yield. Our findings suggest that alternating years of standard and 422 low fertilizations could decrease water pollution and mitigate GHG emissions 423 without decreasing rice yield, an issue to be further studied. 424 425 Acknowledgments. The authors would like to thank Hongchang Ren, Xuming Wang 426 and Dongping Zeng for their assistance with field sampling. Funding was provided by 427 the National Science Foundation of China (41571287), the Program for Innovative 428 Research Team at Fujian Normal University (IRTL1205), the European Research 429 Council Synergy grant ERC-SyG-2013-610028 IMBALANCE-P, the Spanish 430 Government grant CGL2013-48074-P and the Catalan Government grant SGR 431 2014-274. 432 433 REFERENCES 434 435 Adewopo, J.B., Silveira, M.L., Xu, S., Gerber, S., Sollenberger, L.E. and Martin, T. 436 (2015). Management intensification effects on autotrophic and heterotrophic soil 437 respiration in subtropical grasslands. Ecological Indicators 56: 6–14. 438 Ahmad, S., Li, C., Dai, G., Zhan, M., Wang, J., Pan, S. and Cao, C. (2009). 439 Greenhouse gas emission from direct seeding paddy field under different rice 440 tillage systems in central China. Soil & Tillage Research 106: 54–61. 441 Allison, S.D., Wallenstein, M.D. and Bradford, M.A. (2010). Soil-carbon response to 442 warming dependent on microbial physiology. Nature Geoscience 3: 336-340. 443 Bhatia, A., Pathak, H., Jain, N., Singh, P.K. and Singh, A.K. (2005). Global warming 444 potential of manure amended soils under rice-wheat system in the Indo-Gangetic 445 plains. Atmospheric Environment 39: 6976-6984. 10 446 Bhatia, A., Pathak, H., Jain, N., Singh, P.K. and Tomer, R. (2012). Greenhouse gas 447 mitigation in rice-wheat system with leaf color chart-based urea application. 448 Environmental Monitoring and Assessment 184: 3095-3107. 449 Cheng, S.H. and Li, J. (2007). Modern Chinese rice. Beijing: Jindun Press. 450 Dari, B., Sihi, D., Bal, S.K. and Kunwar, S. (2017). Performance of direct-seeded rice 451 under various dates of sowing and irrigation regimes in semi-arid region of India. 452 Paddy and Water Environment 15: 395-401. 453 Davidson EA and Janssens IA (2006). Temperature sensitivity of soil carbon 454 decomposition and feedbacks to climate change. Nature 440:165-173. 455 Ding, W.X., Cai, Z.C., Tsuruta, H. and Li, X.P. (2003). Key factors affecting spatial 456 variation of methane emissions from freshwater marshes. Chemosphere 51: 167– 457 173. 458 Dossou-Yovo, E.R., Brüggemann, N., Jesse, N., Huat, J., Ago, E.E. and Agbossou, 459 E.K. (2016). Reducing soil CO2 emission and improving upland rice yield with 460 no-tillage, straw mulch and nitrogen fertilization in northern Benin. Soil & 461 Tillage Research 156: 44–53. 462 FAO [Food and Agricultural Organization of the United Nations]. (2009). 463 OECD-FAO Agricultural Outlook 2011–2030. 464 Frey, S. D., Lee, J., Melillo, J. M. and Six, J. (2013). The temperature response of soil 465 microbial efficiency and its feedback to climate. Nature Climate Change 3: 466 395-398. 467 Ge, T., Liu, C., Yuan, H., Zhao, Z., Wu, X., Zhu, Z., Brookes, P. and Wu, J. (2015). 468 Tracking the photosynthesized carbon input into soil organic carbon pools in 469 a rice soil fertilized with nitrogen. Plant and Soil 392: 17–25. 470 Gulledge, J. and Schimel, J. P. (1998) Low-concentration kinetics of atmospheric 471 CH4 oxidation in soil and mechanism of NH 4+ inhibition. Applied and 472 Environmental Microbiology 64: 4291-4298. 473 Haque, M.M., Kim, S.Y., Ali, M.A. and Kim, P.J. (2015). Contribution of greenhouse 474 gas emissions during cropping and fallow seasons on total global warming 475 potential in mono-rice paddy soils. Plant and Soil 387: 251–264. 476 Hou, H., Peng, S., Xu, J., Yang, S. and Mao, Z. (2012). Seasonal variations of CH4 477 and N2O emissions in response to water management of paddy fields located in 478 Southeast China. Chemosphere 89: 884–892. 479 Huang, Y., Chen, Z. and Liu, W. (2012). Influence of iron plaque and cultivars on 480 antimony uptake by and translocation in rice (Oryza sativa L.) seedlings exposed 481 to Sb (III) or Sb (V). Plant and Soil 352: 41–49. 482 Hütsch, B.W. (2001). Methane oxidation in non-flooded soils as affected by crop 483 production. European Journal of Agronomy 14: 237–260. 484 Inglett K Inglett P Reddy K Osborne T (2012). Temperature sensitivity of greenhouse 485 gas production in wetland soils of different vegetation. Biogeochemistry 486 108:77-90 487 Inselsbacher, E., Wanek, W., Ripka, K., Hackl, E., Sessitsch, A., Strauss, J. and 488 Zechmeister-Boltenstern, S. (2011). Greenhouse gas fluxes respond to different 489 N fertilizer types due to altered plant-soil-microbe interactions. Plant and Soil 11 490 343: 17–35. 491 Kivlin, S.N., Waring, B.G., Averill, C. and Hawkes, C.V. (2013). Tradeoffs in 492 microbial carbon allocation may mediate soil carbon storage in future climates. 493 Frontiers in Microbiology 4: 261. 494 Linquist, B.A., Liu, L., van Kessel, C. and van Groenigen, K.J. (2013). Enhanced 495 efficiency nitrogen fertilizers for rice systems: Meta-analysis of yield and 496 nitrogen uptake. Field Crops Research 154: 246–254. 497 Liu, Y., Yang, M., Wu, Y., Wang, H., Chen, Y. and Wu, W. (2011). Reducing CH4 and 498 CO2 emissions from waterlogged paddy soil with biochar. Journal of Soil 499 Sediments 11: 930–939. 500 Luo, M., Zeng, C.S., Tong, C., Huang, J.F., Chen, K. and Liu, F.Q. (2016). Iron 501 reduction along an inundation gradient in a tidal sedge (Cyperus malaccensis) 502 marsh: the rates, pathways, and contributions to anaerobic organic matter 503 mineralization. Estuaries and Coasts 39: 1679–1693. 504 Minamikawa, K., Fumoto, T., Itoh, M., Hayano, M., Sudo, S. and Yagi, K. (2014). 505 Potential of prolonged midseason drainage for reducing methane emission from 506 rice paddies in Japan: a long-term simulation using the DNDC-Rice model. 507 Biology and Fertility of Soils 50: 879–889. 508 Minoda, T., Kimura, M. and Wada, E. (1996). Photosynthates as dominant source of 509 CH4 and CO2 in soil water and CH4 emitted to the atmosphere from paddy fields. 510 Journal of Geophysical Research 101: 21091–21097. 511 Myhre, G., Shindell, D., Bréon, F.M., Collins, W., Fuglestvedt, J., Huang, J., Koch, D., 512 Lamarque, J.F., Lee, D., Mendoza, B., Nakajima, T., Robock, A., Stephens, G., 513 Takemura, T. and Zhang, H. (2013). Anthropogenic and Natural Radiative 514 Forcing. In: Climate Change 2013: The Physical Science Basis. Contribution of 515 Working Group I to the Fifth Assessment Report of the Intergovernmental Panel 516 on Climate Change [Stocker, T.F., Qin, D., Plattner, G.K., Tignor, M., Allen, S.K., 517 Boschung, J., Nauels, A., Xia, Y., Bex, V. and Midgley, P.M. (eds.)]. Cambridge 518 University Press, Cambridge, United Kingdom and New York, NY, USA. pp 714. 519 Naik, K.P., Krishnamurthy, N., Ramachandra, C., Hareesh, G.R., Jayadeva, H.M. and 520 Mavarkar, N.S. (2015). Effect of sources of nutrients on methane emission, 521 production and water productivity of rice (Oryza sativa L.) under different 522 methods of cultivation. Research on Crops 16: 213–218. 523 Pathak, H. (2010). Mitigating greenhouse gas and nitrogen loss with improved fertilizer 524 management in rice: quantification and economic assessment. Nutrient Cycling in 525 Agroecosystems 87: 443-454. 526 Pathak H, Bhatia A, Prasad S, Singh S, Kumar S, Jain MC and Kumar U (2002). 527 Emission of Nitrous Oxide from Rice-Wheat Systems of Indo-Gangetic Plains of 528 India. Environmental Monitoring and Assessment 77: 163-178. 529 Sihi, D. (2015). Processes and modeling of temperature sensitivity of soil organic 530 matter decomposition in subtropical wetlands (Doctoral dissertation, ProQuest 531 Dissertations Publishing) 532 Sihi, D., Dari, B., Sharma, D.K., Pathak H., Nain L. and Sharma, O.P. (2017a). 533 Evaluation of soil health in organic vs. conventional farming of basmati rice in 12 534 North India. Journal of Plant Nutrition and Soil Science, DOI: 535 10.1002/jpln.201700128. 536 Sihi, D., Inglett, P.W., Gerber, S. and Inglett, K.S. (2017b). Rate of warming affects 537 temperature sensitivity of anaerobic peat decomposition and greenhouse gas 538 production. Global Change Biology, DOI: 10.1111/gcb.13839. 539 Sihi, D., Gerber, S., Inglett, P.W. and Inglett, K.S. (2016a) Comparing models of 540 microbial-substrate interactions and their response to warming. Biogeosciences 541 13:1-2. 542 Sihi D, Inglett PW and Inglett KS (2016b). Carbon quality and nutrient status drive the 543 temperature sensitivity of organic matter decomposition in subtropical peat soils. 544 Biogeochemistry 131(1-2): 103-119. 545 Sihi, O., Sharma, O.K., Pathak, H., Singh, Y.V., Sharma, C.P., Chaudhary, A. and Oari, 546 B. (2012). Effect of organic farming on productivity and quality of basmati rice. 547 Oryza 49: 24-29. 548 Slot, M., Wright, S.J. and Kitajima, K. (2013). Foliar respiration and its 549 temperature sensitivity in trees and lianas: in situ measurements in the upper 550 canopy of a tropical forest. Tree Physiology 33: 505–515. 551 Thao, H.M., Hong, N.H., Tuyen, N.T., Van Toan, N. and Quyen, L.N. (2015). Study 552 on the effect of some of N, P, K Fertilizer Compounds on the Yield and Quality 553 of Bp53 Rice Variety. Journal of Agricultural Technology 11: 2149–2156. 554 Vogel, J.G., Bronson, D., Gower, S.T. and Schuur, E.A.G. (2014). The response of 555 root and microbial respiration to the experimental warming of a boreal black 556 spruce forest. Canadian Journal of Forest Research 44: 986–993. 557 Wang, W., Sardans, J., Lai, D.Y.F., Wang, C., Zeng, C., Tong, C., Liang, Y. and 558 Peñuelas, J. (2015). Effects of steel slag application on greenhouse gas emissions 559 and crop yield over multiple growing seasons in a subtropical paddy field in China. 560 Field Crops Research 171: 146–156. 561 Wassmann, R. and Aulakh, M.S. (2000). The role of rice plants in regulating 562 mechanisms of methane emissions. Biology and Fertility of Soils 31: 20–29. 563 Zhang, W., Dou, Z., He, P., Ju, X., Powlson, D., Chadwick, D., Norse, D., Lu, Y.L., 564 Zhang, Y., Wu, L., Chen, X.P., Cassman, K.G. and Zhang, F.S. (2013a). New 565 technologies reduce greenhouse gas emissions from nitrogenous fertilizer in 566 China. Proceedings of the National Academy of Sciences 110: 8375–8380. 567 Zhang, A., Cui, L., Pan, G., Li, L., Hussain, Q., Zhang, X., Zheng, J. and Crowley, D. 568 (2010). Effect of biochar amendment on yield and methane and nitrous oxide 569 emissions from a rice paddy from Tai Lake plain, China. Agriculture Ecosystems 570 & Environment 139: 469–475. 571 Zhao, Z., Yue, Y., Sha, Z., Li, C., Deng, J., Zhang, H., Gao, M. and Cao, L. (2015). 572 Assessing impacts of alternative fertilizer management practices on both nitrogen 573 loading and greenhouse gas emissions in rice cultivation. Atmospheric 574 Environment 119: 393–401. 575 Zhong, Y., Wang, X., Yang, J., Zhao, X. and Ye, X. (2016). Exploring a suitable 576 nitrogen fertilizer rate to reduce greenhouse gas emissions and ensure rice yields 577 in paddy fields. Science of the Total Environment 565: 420–426. 13 578 Tables 579 Table 1. Results of the linear analysis of the effects of the mixed models, with treatment as a fixed factor, plot and time as random factors on 580 greenhouse-gas emissions and porewater greenhouse-gas concentrations and other soil variables as dependent variables. R2m is the variance 581 explained by the fixed factors, and R2c is the variance explained by the overall model (fixed + random). Statistical significant values are in bold 582 type. 583 Dependent variable Y~ treatment, random=list(~1|plot, ~1|time) R2 Standard fertilization Double fertilization (relative to control) (relative to control) t P t P CO2 emission R2m=0.0036, R2c=0.92 41.8 <0.0001 2.51 0.088 2 2 CH4 emission R m=0.024, R c=0.44 1.15 0.25 2.23 0.028 2 2 N2O emission R m=0.0035, R c=0.49 -0.68 0.5 -0.83 0.41 2 2 Dissolved porewater CO2 concentration R m=0.21, R c=0.54 -0.71 0.48 5.93 <0.0001 2 2 Dissolved porewater CH4 concentration R m=0.0037, R c=0.39 -0.81 0.42 -0.59 0.55 2 2 Dissolved porewater N2O concentration R m=0.017, R c=0.11 0.43 0.67 1.44 0.15 2 2 Soil pH R m=0.00045, R c=0.99 0.15 0.88 -1.89 0.063 2 2 Soil Eh R m=0.0015, R c=0.97 -2.20 0.031 -1.64 0.10 2 2 Soil salinity R m=0.0328, R c=0.66 1.95 0.055 3.33 0.0013 2 2 Soil temperature R m=0.00032, R c=0.98 -0.70 0.49 0.73 0.47 2 2 Soil water content R m=0.0012, R c=0.92 1.06 0.29 1.20 0.24 Porewater NH4+ concentration 2 2 R m=0.043, R c=0.24 1.37 0.17 2.56 0.013 2 2 Porewater NO3- concentration R m=0.0026, R c=0.0026 0.32 0.75 0.55 0.58 2 2 Porewater TN concentration R m=0.018, R c=0.33 0.82 0.41 1.75 0.084 2 2 Porewater TP concentration R m=0.022, R c=0.081 0.34 0.73 1.59 0.12 2 2 Porewater DOC concentration R m=0.0063, R c=0.82 -1.52 0.13 -1.91 0.059 14 584 585 Table 2. Effect of the fertilizations on the global warming potential. Rice yield Global warming potential (kg CO2-eq ha-1) Global warming potential Global warming potential Treatment (Mg ha-1) CO2 CH4 N2O (kg CO2-eq ha-1) (kg CO2-eq Mg-1 yield) Control 7.07±0.40 19 286±2553 2607±937 -27.4±84.9 21 801±3326 3134±616 Standard 7.44±0.35 21 883±1009 3139±340 -44.7±52.6 fertilization 24 872±1147 3363±245 Double 6.97±0.24 20 709±681 3635±722 -55.1±44.5 fertilization 24 160±295 3475±129 586 587 15 588 Figure legends 589 590 Figure 1. Seasonal variation of CO2 (A) and CH4 (B) emissions and N2O (C) fluxes 591 from the control and treatment plots. Error bars indicate one standard error of the mean 592 of triplicate measurements. 593 594 Figure 2. Seasonal variation of dissolved porewater CO2 (A), CH4 (B) and N2O (C) 595 concentrations for the control and treatment plots. Error bars indicate one standard 596 error of the mean of triplicate measurements. 597 598 Figure 3. Seasonal variation of soil pH (A), Eh (B), temperature (C), salinity (D) and 599 water content (E) for the control and treatment plots. Error bars indicate one standard 600 error of the mean of triplicate measurements. 601 602 Figure 4. Seasonal variation of dissolved porewater NH4+ (A), NO3- (B), TN (C), TP 603 (D) and DOC (E) concentrations for the control and treatment plots. Error bars 604 indicate one standard error of the mean of triplicate measurements. 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 16 628 629 630 631 Figure 1 632 633 634 635 636 637 638 639 640 641 642 643 17 644 645 646 647 648 649 650 651 652 653 654 655 Figure 2 656 657 658 18 659 660 Figure 3 661 662 19 663 664 Figure 4 665 20