1 Acid rain mediated nitrogen and sulfur deposition alters soil 2 nitrogen, phosphorus and carbon fractions in a subtropical paddy 3 4 Chun Wanga,b, Yunying Fangc, Wanli Ana,b, Congsheng Zenga,b, Weiqi Wanga,b,*, Jordi Sardansd,e,*, 5 Marcos Fernández-Martínezd,e, Josep Peñuelasd,e a 6 Institute of Geography, Fujian Normal University, Fuzhou 350007, China b 7 Key Laboratory of Humid Subtropical Eco-geographical Process, Ministry of Education, Fujian 8 Normal University, Fuzhou 350007, China cNSW 9 Department of Primary Industries, Elizabeth Macarthur Agricultural Institute, Menangle, 10 NSW 2568, Australia dCSIC, 11 Global Ecology Unit CREAF-CSIC-UAB, 08913 Bellaterra, Catalonia, Spain eCREAF. 12 08913 Cerdanyola del Vallès, Catalonia, Spain 13 14 *Corresponding authors at: Institute of Geography, Fujian Normal University, Fuzhou 350007, 15 China. 16 E-mail addresses: wangweiqi15@163.com (W. Wang); j.sardans@creaf.uab.cat (J. Sardans) 17 18 19 20 21 22 23 24 25 26 27 28 29 ABSTRACT 1 30 Agricultural ecosystems are globally important sinks of carbon and other nutrient 31 elements. In China, acid rain events affect about 0.62 million km2, representing about 32 6.4% of total land area; however, the impacts of acid rain mediated nitrogen (N) and 33 sulfur (S) depositions on soil carbon and nutrient stocks in paddy soils and implications 34 for yield production under climate change are unclear. We conducted a field experiment 35 during two annual crop seasons to determine the effects of simulated acid rain on soil 36 organic carbon (SOC) fractions and nutrients in a subtropical paddy in China. Acid rain 37 treatments comprised solutions of HNO3 + H2SO4 to simulate N and S deposition at pH 38 levels of 4.5, 3.5, and 2.5. The results showed that content of soil C fractions varied 39 with acid rain pH. Acid rain led to increased SOC content and decreased ratios of soil 40 labile organic carbon (LOC): SOC and dissolved organic carbon DOC: SOC 41 concentration, independently of crop season and growth stage. Soil salinity was 42 positively associated with SOC suggesting that higher levels of salinity inhibit C 43 decomposition favoring SOC accumulation. Treatment effects of acid rain on soil 44 microbial C and N depended on crop growth stage. Concentration of Fe2+ was positively 45 correlated with DOC in early and late paddy soils under acid rain, possibly as a result 46 of Fe2+ retention by DOC, when Fe3+ is reduced to Fe2+. Acid rain led to increases in 47 soil TN, TP, available N, NH4+ concentration, and SOC, and decreases in ratios of DOC: 48 SOC, indicating decreases in soil biological activity and mineralization processes. 49 Increases in dead rice plant biomass under acid rain were consistent with the increase 50 in soil N and P concentrations, due to reduced nutrient uptake, and higher levels of total 51 SOC. 52 53 Keywords: Organic carbon; Nutrient; Stoichiometry; Acid rain; Paddy 54 2 55 1. Introduction 56 Acid rain is a key environmental problem, because acidification influences soil nutrient 57 storage and release processes (Rosi-Marshall et al., 2016), animal growth (Warren et 58 al., 2017; Wei et al., 2017), plant growth (Medeiros et al., 2016), and soil metal leaching 59 (Li et al., 2015). Acid rain has been shown to affect plant photosynthetic and antioxidant 60 activities (Chen et al., 2013), physiological responses (Kováčik et al., 2011), and litter 61 decomposition (Wang et al., 2012), and while a number of studies have investigated 62 effects of simulated acid rain on soil respiration and its drivers in a subtropical mixed 63 conifer and broadleaf forest (Liang et al., 2016), effects on organic carbon fractions 64 remain unclear. 65 Global enhancement of carbon (C) sequestration in agricultural soils is considered 66 a key measure to offset anthropogenic GHG emissions (Wang et al., 2015a, b), 67 particularly in crops, such as rice paddy fields (Pan et al., 2004). Rice is a staple cereal 68 crop for more than 50% of the global population (Haque et al., 2015); however, a 40% 69 increase in its production by the end of 2030 is required to meet the food demands of 70 the growing global population (FAO, 2009). Acid rain enhances photosynthetic 71 parameters in rice, and hence grain yields (Wang et al., 2014a), and it general increases 72 CO2 emission and decreases the N nutrient availability (Ouyang et al., 2008), significant 73 losses of major plant nutrients, such as potassium, calcium, and magnesium (Nawaz et 74 al., 2012), increases recalcitrant organic matter ( Wu et al., 2016). Rice cultivation plays 75 an important role in the mitigation of atmospheric carbon dioxide (CO2) (Lal, 2004), 76 because the low soil organic C (SOC) content in paddy soils facilitates greater potential 77 for sequestration of additional C under suitable management practices than soils with 78 greater levels of SOC content (Pan et al., 2004; Wissing et al., 2011; Wang et al., 2015a), 79 and it affects soil active C, which is the fraction of soil C with high levels of activity, 80 due to impacts on plants and microorganisms and susceptibility to oxidation and 81 decomposition (Kimura et al., 2004; Chen et al., 2010; Wang et al., 2015a). Nitrogen 82 (N) inputs impact C and nutrient stoichiometry and, therefore, nutritional conditions 83 required for microbe growth (Wang et al., 2014b), so the fraction of active C in paddy 84 soil has been used as an indicator of management impacts on microbial and plant 3 85 growth, and soil C dynamics (Purakayastha et al., 2008; Gong et al., 2009; Xu et al., 86 2011a,b). 87 Adequate levels of soil nutrients are essential to support the sustainable production 88 of food (Ford et al., 2016); however, loss of nutrients creates environmental problems, 89 such as eutrophication of water (Smith et al., 1999). Acid rain may affect soil nutrients 90 directly through acidification and input of large amounts of active nutrients, such as N. 91 Effects of acid rain may vary among nutrients, and with intensity and level of acidity; 92 for example, soil NO3- and available P were reduced in tea plantation rhizosphere soil 93 (Hu et al., 2017), while there was little effect on soil C and nutrient status (Chen et al., 94 2015). Effects of acid rain on soil nutrient status in paddy crops are unclear, although it 95 is known that frequent periods of acid rain are associated with high levels of N and 96 sulfur (S) deposition that indicate subsequent changes in soil chemistry, structure, C 97 content, and nutrient status (Dise and Verry, 2001; Krusche et al., 2003; Hu et al., 2018). 98 Soil nutrient status drives ecosystem carbon sequestration and nutrient dynamics 99 (Hessen et al., 2004), as a result of processes associated with litter C input (Wang et al., 100 2016), soil C concentration and storage (Wang et al., 2015c), and release of C from soils 101 (Wang et al., 2014b). Besides nutrient stoichiometry, soil properties, such as iron 102 dynamics (Peng et al., 2015), salinity (Wang et al., 2017a), water content (Wang et al., 103 2013), and pH (Jin and Wang, 2018) affect soil C fractions and nutrient content that are 104 known to vary with rice growth (Wang et al., 2017b). However, effects of acid rain on 105 variation in soil C fractions with rice crop growth stage are unclear. 106 China cultivates the second largest area of rice in the world, where 90% of the 107 paddies are in the subtropics, such as in Fujian, Jiangxi, and Hunan Provinces. The 108 development of strategies to increase the cost-effectiveness of rice agriculture and 109 enhancement of crop yield and carbon sequestration from paddies has been investigated 110 in subtropical China (Wang et al., 2015a, b). In China, acid rain affects an area of about 111 0.62 million km2 (equivalent to about 6.4% of total land area), and this deposition is 112 associated with inputs of N and S (about 24.2 and 21.1% of the total ion equivalent in 113 acid rain, respectively) (Ministry of Ecology and Environment of the People’s Republic 114 of China, 2018) that is commonly occurring in several paddy areas of the country 4 115 (Larssen and Carmichael, 2000). 116 We aimed to determine the impacts of acid rain mediated N and S depositions on 117 soil carbon and nutrient stocks in paddy soils and implications for yield production. We 118 expected to provide a scientific basis for the effective assessment of the impact of acid rain on soil 119 nutrient status and carbon sequestration. We also expected to provide suggestions for better 120 responding to the impact of acid rain in future. 121 122 2. Materials and methods 123 2.1. Study site and experimental design 124 A field experiment was established in 2015 at the Fujian Academy of Agricultural 125 Sciences, Fujian, southeastern China (119.3°E, 26.1°N,) (Fig. S1) during early (16 April 126 to 16 July) and late (25 July to 6 November) paddy seasons. The proportions of sand, 127 silt, and clay in the upper 15 cm of soil were 28, 60, and 12%, respectively, and other 128 physicochemical properties comprised bulk density: 1.1 g cm-3; pH: (1:5 with H2O) 6.5; 129 organic C content: 18.1 g kg-1; total N content: 1.2 g kg-1; and, total P content: 1.1 g kg- 1 130 (Wang et al. 2014c, 2015b). Air temperature and humidity during the study period 131 were shown in Fig. S2. 132 The paddy soils were plowed to a depth of 15 cm using a moldboard plow and 133 then leveled immediately prior to transplantation of rice seedlings of ‘Hesheng 10’ and 134 ‘Qinxiangyou 212’ in the early and late paddy seasons, respectively, using a rice 135 transplanter, to a depth of 5 cm, and with plant and row spacings of 14 and 28 cm, 136 respectively. Fertilizer was applied as NH4-P2O5-K2O (16-16-16; Keda Fertilizer Co., 137 Ltd., Jingzhou, China) and urea (46% N), according to common agronomic practice in 138 the region. Fertilizer was applied 1 d before transplantation (N, P, and K at 42, 40 and 139 40 kg ha-1, respectively); during tiller-initiation, 7 d after transplantation (DAT), (N, P, 140 and K at 35, 20, and 20 kg ha-1, respectively); and, during panicle-initiation at 56 DAT; 141 (N, P, and K at 18, 10, and 10 kg ha-1, respectively). Paddy soils were flooded from 0 142 to 37 DAT, during which time, an automatic water-level controller was used to maintain 143 water level at 5-7 cm above the soil surface. Then, the paddies were drained between 144 37 and 44 DAT; we maintained moist soils between 44 and 77 DAT for the early paddy 5 145 and between 44 and 91 DAT for the late paddy. Paddies were drained 2 weeks before 146 harvest (77 DAT for the early crop, 91 DAT for the late crop), and we harvested the rice 147 crops at 92 DAT and 106 DAT for the early and late crops, respectively. 148 Three acid rain treatments, with pH 4.5, 3.5, and 2.5, were based on the acidity of 149 rain water recorded in natural rain events in the study area (Table S1). Simulated acid 150 rain was added every 7 d as simulated rainfall, as an amount equivalent to the local 151 weekly mean rainfall during the respective paddy crop seasons, using mixed acid 152 solutions of HNO3 and H2SO4 (Wang et al., 2014a) to simulate N and S deposition. 153 Three replicates of experimental and control plots were arranged in a randomized block 154 design, where the 10-m2 plots were separated by 0.5-cm thick, 30-cm high PVC plate 155 (of which 20 cm was inserted into the soil) and a buffer zone of about 2 m. Control 156 plots received non-acidified water. Plots were managed according to local common 157 practice (Zhang et al., 2013; Wang et al., 2015a, b), and wooden bridges were 158 constructed to minimize disturbance to soils during sample collection. 159 160 2.2. Determination of soil and rice properties 161 Soils were sampled (N = 72) during the rice greening, jointing, and mature stages (8, 162 78, and 92 DAT for the early paddy; 8, 64, and 106 for the late paddy) using a core 163 sampler (0.3-m long, with 0.1-m diameter) from the upper 15-cm soil layer. The sample 164 cores were divided into two parts, with one maintained at 4 °C for the measurement of 165 soil microbial biomass carbon (MBC), microbial biomass nitrogen (MBN), available 166 nitrogen (N), available phosphorus (P), and dissolved organic carbon (DOC), and the 167 other was air-dried and finely ground using a ball mill, after all roots and visible plant 168 remains had been removed, for determination of total soil organic carbon (SOC) and 169 labile organic carbon (LOC). 170 DOC was determined by extracting C from the soils using 0.5 mol l-1 of K2SO4 and 171 measuring its concentration using a TOC-V CPH total C analyzer (Shimadzu Scientific 172 Instruments, Kyoto, Japan). MBC was determined using chloroform fumigation and 173 extraction using 0.5 mol l-1 of K2SO4 (Lu, 1999) prior to measurement of C 174 concentration using a TOC-V CPH total C analyzer (Shimadzu Scientific Instruments, 6 175 Kyoto, Japan). MBN was determined using chloroform fumigation and extraction using 176 0.5 mol l-1 of K2SO4 (Lu, 1999) prior to measurement using a sequence flow analyzer 177 (San++, SKALAR Corporation production, Breda, The Netherlands); additional details 178 are provided by Fang et al. (2018). Available P was extracted using the Mehlich method 179 and measured using a sequence flow analyzer (San++, SKALAR Corporation 180 production, Breda, The Netherlands). Available N was calculated as the sum of NO3- 181 and NH4+ in soil that was extracted using 2 mol l-1 of KCl and measured using a 182 sequence flow analyzer (San++, SKALAR Corporation production, Breda, The 183 Netherlands). Total SOC and total N (TN) were determined using an Elementar Vario 184 MAX CN Analyzer (Elementar Scientific Instruments, Hanau, Germany), and soil LOC 185 was extracted using 333 mM of KMnO4 in a digestion method (Xu et al., 2011b) and 186 then determined using a UV-2450 spectrophotometer (Shimadzu Scientific Instruments, 187 Kyoto, Japan). Soil total P (TP) concentration was measured extracted using perchloric- 188 acid digestion and determined using a sequence flow analyzer (San++, SKALAR 189 Corporation production, Breda, The Netherlands). 190 Soil salinity and temperature were measured using a 2265FS EC/temperature 191 meter (Spectrum Technologies Inc., Paxinos, USA), while soil pH (1:5 soil:water ratio) 192 was measured using a Starter 300 pH meter (Ohaus Scientific Instruments, Parsippany, 193 USA), and soil water content was measured using a TDR300 water content meter in 194 situ (Spectrum Technologies Inc., Paxinos, USA). Total active Fe concentration was 195 determined using a colorimetric method, where 1,10-phenanthroline was added to 1 196 mol l-1 HCl fresh-soil extracts to react with ferrous ions (Lu, 1999; Wang et al., 2014c). 197 Ferric Fe concentration was calculated by subtracting concentration of ferrous Fe from 198 that of total active Fe (Wang et al., 2014c). Rice height and grain yield were determined 199 at the harvesting stage directly (Wang et al., 2014c). 200 201 2.3. Statistical analysis 202 We used general mixed models (GLM) in the “nlme” package in R, with the “lme” 203 function (Pinheiro et al., 2016) to analyze the effects of acid rain independently of the 204 season and growth stage, with acid rain season and growth stage as independent fixed 7 205 categorical variables and plot as independent random factors with all soil studied 206 variables as continuous dependent variables. Thereafter, we also analyzed the effects of 207 acid rain within each crop season by similar mixed model but without season as fixed 208 categorical independent variable.Non-normally distributed variables were log- 209 transformed.. We chose the best model for each dependent variable based on the Akaike 210 information criterion, and we used the MuMIn (Barton, 2012) R package to estimate 211 the proportion of variance (%) explained by the mixed models. Tukey’s post hoc tests 212 (at P < 0.05) were used to detect treatment differences using the “multcomp” (Hothorn 213 et al., 2013) R package, with the “glht” function. 214 Associations between soil organic fractions and properties were tested using 215 Pearson correlation analysis and treatment effects were tested using Bonferroni’s post 216 hoc test (at P < 0.05). These analyses were performed using SPSS Statistics 18.0 (SPSS 217 Inc., Chicago, USA). 218 Overall differences in soil traits among the four treatments were tested using a 219 general discriminant analysis (GDA). Discriminant analyses consist of a supervised 220 statistical algorithm that derives an optimal separation between groups established a 221 priori by maximizing between-group variance, while minimizing within-group variance 222 and controlling the effects of a categorical variable (here, season × crop stage) 223 (Raamsdonk et al., 2001). Thus, GDA is a useful tool for the identification of variables 224 that drive greatest differences among groups, and here, it allowed the discrimination of 225 overall differences in soil traits under contrasting levels of acid rain independently of 226 season and crop growth stage. Before conducting these multivariate analyses, we 227 selected the sampling adequacy of individuals and the set of variables by the Barlett’s 228 test of sphericity (<0.05) and the Kaiser-Meyer-Olkin measure (>0.50). We removed 229 the variables with communality values < 0.5 and perform GDA analyses with the 8 230 variables with communality > 0.5. To perform these sampling adequacy analyses we 231 used the package psych (Revelle, 2010). A cross-validation procedure (leave-one-out) was 232 applied to test the adequacy of the discriminant model. GDAs were performed using 233 Statistica 8.0 (StatSoft, Inc. Tulsa, USA). 234 235 3. Results 236 3.1. Effect of acid rain on soil carbon and nutrient concentrations and soil physyco- 237 chemical variables. 238 Acid rain treatments led to greater levels of SOC, TN, soil N-avail, NO3-, TP and SWC, 239 but lower pH, C:N, bacterial otus concentration, and ratios of LOC/SOC and DOC/SOC, 240 regardless of crop season and crop growth stage (Table S2). 241 Analyzing the relationships of acid rain with the studied soil variables within each 242 crop season, we observed between and within crop growth season some differences in 243 several soil variables (Table S3, Fig. 1, Fig. S3). However, in both crop seasons, we 244 observed that acid rain treatment decreased soil pH and increased TN, TP, SOC and N- 245 availability (Table S3, Fig. 1-4). In early paddy season acid rain increased soil N- 246 availability: P-availability ratio, soil water content and salinity and decreased soil C:N 247 ratio (Table S3, Fig. 3 and 4). While in late paddy season acid rain was associated with 248 an increase of soil NO3- and MBC concentrations and with lower LOC:SOC, DOC:SOC 249 and MBC:SOC ratios and soil microbial otus concentrations (Table S3, Fig. 1-3). Soil 250 TN and TP concentrations, and MBC:MBN ratios varied among rice growth stages 251 (Table S3, Fig. 2,3). Ratios of soil C:P, N:P and LOC:available-N varied among rice 252 growth stages in the two crop seasons (Table S3, Fig. 3). 253 Soil available N concentrations in early paddy soils were 45.5% greater in the pH 254 4.5 acid rain treatment than control. We found that soil NO3- concentrations were 255 81.7% greater in the pH 2.5 acid rain treatment than the control; however, NH4+ 256 concentrations in early paddy soils were 49.4, 54.9, and 50.6% lower in the pH 4.5, 257 3.5, and 2.5 acid rain treatments, respectively, than the control. In early paddy soils, 9 258 soil MBN was about 52.2 and 35.5% greater in the pH 4.5 and 2.5 acid rain treatments 259 than the control, while in late paddy soils, it was about 27.4, 33.8, and 38.3% lower in 260 the pH 4.5, 3.5, and 2.5 treatments, respectively, than the control. In early paddy soils, 261 LOC:available N ratios were 22.5% lower in the pH 4.5 acid rain treatment than the 262 control. Levels of soil Fe2+ concentration, salinity, and temperature varied among rice 263 growth stages (Table S3, Fig. 5) in the two crop seasons, as did levels of soil water 264 content and pH (Table S3, Fig. 4). 265 266 3.2. Physico-chemical soil variables (Fe, Salinity, SWC, temperature and pH.) 267 MBC was negatively correlated with DOC in early (r = -0.758, P < 0.01) and late (r = 268 -0.436, P < 0.01) paddy soils (Table S4). Soil TN and NH4+ concentrations were 269 positively correlated with DOC in early (r = 0.504 and 0.472, respectively, P < 0.01) 270 and late (r = 0.341 and 0.749, respectively, P < 0.05) paddy soils, and soil available N 271 was positively correlated with SOC and DOC in early (r = 0.503 and 0.654, respectively, 272 P < 0.01) and late (r = 0.362 and 0.757, respectively, P < 0.05) paddy soils (Table 1). 273 TP was positively associated with SOC in early (r = 0.811, P < 0.01) and late (r = 0.480, 274 P < 0.01) paddy soils (Table 1). Ratio of LOC:available N was negatively correlated 275 with SOC and DOC in early (r = -0.403 and -0.673, respectively, P < 0.05) and late (r 276 = -0.333 and -0.579, respectively, P < 0.05) paddy soils, and the ratio of available 277 N:available P was positively associated with DOC in early (r = 0.673, P < 0.01) and 278 late (r = 0.631, P < 0.01) paddy soils (Table 1). 2+ 279 In general, soil salinity was positively correlated with SOC, and Fe was 280 positively correlated with DOC in the two crop seasons (P < 0.05, Table 1). 281 282 3.3. Effect of acid rain on rice yield and plant traits 283 There were no effects of acid rain treatment on rice yield or plant height in either crop 284 season (P > 0.05). In early paddy soils, rice yields in the control and acid rain treatments 285 pH 4.5, 3.5, and 2.5 were 4.63 ±0.52, 4.12 ±0.42, 4.37 ±0.11, and 4.55 ±0.42 Mg ha-1, 286 respectively, while in late paddy soils, they were 6.73 ±0.77, 5.89 ±0.25, 7.15 ±0.10, 10 287 and 6.25 ±0.21 Mg ha-1, respectively. Simulated acid rain resulted in greater proportions 288 of moribund aboveground biomass at the end of the rice growth period, prior to harvest 289 (data no shown). 290 291 3.4. Overall effects of simulated acid rain treatments on soil traits 292 GDA showed there were acid rain treatment effects on traits of paddy soils compared 293 with untreated control soils, and between acid rain treatment pH 4.5 and the other two 294 acid rain treatments (pH to 3.5 and 2.5); there were no overall differences in traits of 295 soils treated with simulated acid rain at pH 2.5 and 3.5 (Tables 2, 3). The percent of 296 correct predictions were 87.3%. Key drivers of differences in soil traits between the 297 control and acid rain treatments along Root 1 were soil pH and SOC, where lower levels 298 of higher pH and greater levels of SOC were associated with acid rain, while key drivers 299 of differences between the pH 4.5 treatment and the pH 3.5 and 2.5 treatments along 300 Root 2 were SOC, SWC, TP, DOC, Fe3+, and Navai:Pavai ratio (Fig. 6). 301 302 4. Discussion 303 4.1. Effect of acid rain on carbon fractions, nutrient, and C, N, P stoichiometry 304 The results clearly showed that acid rain with N and S inputs as produced in natural 305 events in China clearly increase SOC and soil total N and P. This is associated with a 306 decrease of soil pH and more mortality in aboveground plant biomass. Thus, despite 307 litter production increases following acid rain events, resulting in greater soil inputs of 308 C, N, and P (Wang et al., 2012); the results strongly suggest that litter decomposition 309 rates have decreased due to acid rain observing of LOC:SOC and LOC:SOC ratios and 310 total soil P. This showed a lower rates of SOC conversion to DOC and LOC under acid 311 rain consistent with the decrease of microbial community density. 312 Our finding that acid rain resulted in greater levels of soil TN and available N was 313 consistent with the association between acid rain and N inputs (Rosi-Marshall et al., 314 2016) but also with the association of lower levels of pH and microbe inhibition (Babich 315 and Stotzky, 1982). The very high rise in soil nitrate and a no clear rise in soil 316 ammonium would be consistent with the rise of nitrification activity observed in other 11 317 studies (Chao, 2010), and thus with higher mineralization of N and P and few N, P loss 318 (Wang et al., 2012; Chen et al., 2015). The results also strongly suggest that the drop 319 plant growth and nutrient uptake capacity (Malkanthi et al., 1995; Medeiros et al., 2016) 320 linked to reduced root growth as observed in other studies (Liao and Chen, 1991; 321 Kováčik et al., 2011; Medeiros et al., 2016) have conducted to an accumulation of C, 322 N and P in soils. Thus, simulated acid rain at the level and characteristics as the observed 323 in several cropland areas of China have a general short-time effect reducing soil pH and 324 soil microbe content, associated to a rise in SOC and total N and P in soil associated all 325 them to a drop in plant activity.. 326 327 4.2. Effect of growth stage on soil properties, carbon fractions, nutrients, and C, N, and 328 P stoichiometry 329 Concentrations of soil Fe2+, water content, salinity, and temperature varied among rice 330 growth stages in the two crop seasons. The greater levels of soil active Fe2+ 331 concentrations at the greening growth stage than at the jointing and mature stages were 332 related to greater anoxic soil conditions and, therefore, the reduction of Fe3+ to Fe2+ that 333 occurred under flood conditions (Liu et al., 2013). However we have not observed 334 changes in soil Fe2+ and Fe3+ concentrations the possible Fe3+ reduction by sulfate 335 application, because sulfate-reduced products act as electron donors for the reduction 336 of Fe3+ to Fe2+ (Nicol et al., 2016) as observed in other studies (Fry et al., 1986). 337 In general, we found that total SOC concentrations, soil TN, TP, available N, and 338 available NH4+ concentrations varied among rice growth stages in the two crop seasons, 339 coinciding with the rice crop phenology and with the periods of fertilizer application 340 such as observed in previous studies (Bowatte et al., 2010; Wang et al., 2017b; Song et 341 al., 2018; Wang et al., 2018a,b).. 342 343 4.3. Associations among soil carbon fractions, nutrients and stoichiometry, and soil 344 traits 345 DOC was the active C resource for microbe growth; therefore, greater levels of DOC 346 should have increased microbe growth (Schmidt et al., 1997; Wang et al., 2015a), 12 347 however, we have observed that acid rain whereas increase DOC was associated to a 348 decrease in soil microbe otus number. The decrease in microbes may reduce 349 mineralization of SOC into labile carbon, such as DOC (Kunlanit et al., 2014), as 350 reflected by decrease of microbe otus and the rise of DOC:SOC under acid rain. 351 Soil TN, available N, NH4+ concentration, and available N:available P ratios 352 were positively associated with DOC in the two crop seasons, indicating the study 353 paddy soils were N, rather than P-limited. This is consistent with the positive effects of 354 N fertilization in the paddy soils of this cropland area (Zhang et al., 2009; Xu et al., 355 2010; Wang et al., 2014b; 2018c; Zhu et al., 2016). Overall, our data showed that acid 356 rain decreased soil microbe biomass and SOC mineralization, despite greater 357 availabilities of N and P that were probably related to decreased plant and microbial 358 uptake. Thus, we suggest acid rain decreases biological activity, including 359 mineralization of soil organic matter and plant-microbe uptake, that leads to increases 360 in losses of available forms of nutrients and accumulation of C, N, and P soil recalcitrant 361 fractions. 362 363 5. Conclusions 364 1. Acid rain increased SOC content and decreased ratios of soil LOC: SOC, DOC: SOC 365 and MBC:SOC concentration ratios, whereas total soil N and P concentrations 366 increased, indicating a decrease in soil mineralization and/or lower plant uptake. 367 2. Increased soil availability of N and P (P only in late paddy soils) under acid rain was 368 likely related to diminished plant nutrient up-take, as supported by an increase in the 369 proportion of dead rice plant biomass under acid rain. 370 3. The concentration of Fe2+ was positively correlated with DOC in early and late paddy 371 soils under acid rain, possibly as a result of Fe2+ retention by DOC when Fe3+ is reduced 372 to Fe2+. 373 4. The observed increases in labile soil N and P, together with greater levels of SOC 374 under acid rain strongly suggest that prolonged acid rain on paddy should favor the 375 accumulation of C, N, and P recalcitrant fractions in soils leading to lower levels of soil 376 fertility in paddy soils. 13 377 378 5. Soil TN, available N, NH4+ concentration, and ratio of available N:available P were 379 positively associated with DOC in the two crop seasons, indicating the study site paddy 380 soils were N, rather than P-limited. But the positive role on soil microbe and plant 381 activity and SOC decomposition that can be expected from N deposition accompanying 382 acid rain is counteracted by the negative effects of soil acidification. 383 384 Acknowledgments 385 The authors would like to thank Dengzhou Gao, Miaoying Wang, and Qiuli Zhu for 386 assistance during field sampling. Funding was provided by the National Science 387 Foundation of China (41571287, 31000209), Research Project of Public Institute of 388 Fujian Province (2018R1034-1), Outstanding Young Research Talents in Higher 389 Education of Fujian Province (2017), the European Research Council Synergy grant 390 ERC-SyG-2013-610028 IMBALANCE-P, the Spanish Government grant PID2019- 391 110521GB-I00, and the Catalan Government grant SGR 2017-1005. 392 393 Conflicts of Interest 394 The authors declare no conflicts of interest. 395 396 References 397 Babich, H., Stotzky, G., 1982. Nickel toxicity to microbes: effect of pH and implications 398 for acid rain. Environ. Res. 29(2), 335–350. 399 Barton, K., 2012. MuM. In: Multi-model inference. R package version 1.7.2. 400 http://CRAN. R-283 project.org/package=MuMIn. at http://cran.r-project.org/ 401 package=MuMIn>. 402 Bowatte, S., Jia, Z., Ishihara, R., Nakajima, Y., Asakawa, S., Kimura, M., 2010. 403 Molecular analysis of the ammonia oxidizing bacterial community in the surface 404 soil layer of a japanese paddy field. Soil Sci. Plant Nutr. 52(4), 427–431. 405 Chao, W.L., 2010. Effects of acid rain on the nitrifying activity of copper amended soil. 406 Lett. Appl. Microbiol. 21(1), 23–24. 14 407 Chen, H.L., Zhou, J.M., Xiao, B.H., 2010. Characterization of dissolved organic matter 408 derived from rice straw at different stages of decay. J. Soil Sediment. 10, 915– 409 922. 410 Chen, J., Wang, W.H., Liu, T.W., Wu, F.H., Zheng, H.L., 2013. Photosynthetic and 411 antioxidant responses of Liquidambar formosana and Schima superba seedlings 412 to sulfuric-rich and nitric-rich simulated acid rain. Plant Physiol. Bioch. 64, 41– 413 51. 414 Chen, X., Chen, F., Ye, S., Yu, S., Fang, X., Hu, X., 2015. Responses of rhizosphere 415 nitrogen and phosphorus transformations to different acid rain intensities in a hilly 416 red soil tea plantation. Chin. J. Appl. Ecol. 26(1), 1–8. 417 Dise, N.B., Verry, E.S., 2001. Suppression of peatland methane emission by cumulative 418 sulfate deposition in simulated acid rain. Biogeochemistry 53(2), 143–160. 419 Fang, Y., Singh, B.P., Luo, Y., Boersma, M., Zwieten, L.V., 2018. Biochar carbon 420 dynamics in physically separated fractions and microbial use efficiency in 421 contrasting soils under temperate pastures. Soil Biol. Biochem. 116, 399–409. 422 FAO [Food and Agricultural Organization of the United Nations]., 2009. OECD-FAO 423 Agricultural Outlook 2011–2030. 424 Ford, H., Roberts, A., Jones, L., 2016. Nitrogen and phosphorus co-limitation and 425 grazing moderate nitrogen impacts on plant growth and nutrient cycling in sand 426 dune grassland. Sci. Total Environ. 542, 203–209. 427 Fry, I.V., Lazaroff, N., Packer, L., 1986. Sulfate-dependent iron oxidation by 428 Thiobacillus ferrooxidans: characterization of a new EPR detectable electron 429 transport component on the reducing side of rusticyanin. Arch. Biochem. Biophys. 430 246(2), 650–654. 431 Gong, W., Yan, X., Wang, J., Hu, T., Gong, Y., 2009. Long-term manuring and 432 fertilization effects on soil organic carbon pools under a wheat-maize cropping 433 system in North China Plain. Plant Soil 149, 318–324. 434 Haque, M.M., Kim, S.Y., Ali, M.A., Kim, P.J., 2015. Contribution of greenhouse gas 435 emissions during cropping and fallow seasons on total global warming potential in 436 mono-rice paddy soils. Plant Soil 387, 251–264. 15 437 Hessen, D.O., Ågren, G.I., Anderson, T.R., Elser, J.J., Ruiter, P.C.D., 2004. Carbon 438 sequestration in ecosystems: the role of stoichiometry. Ecology 85(5), 1179–1192. 439 Hothorn, T., Bretz, F., Wesrfall, P., 2013. Package “mulcomp” (WWW document). 440 U.R.L. http://cran.stat.sfu.ca/web/packages/mulcomp/mulcomp.pdf. (accessed 441 19.12.17). 442 Hu, M., Wilson, B.J., Sun, Z., Huang, J., Tong, C., 2018. Effects of nitrogen and 443 sulphate addition on methane oxidation in the marsh soil of a typical subtropical 444 estuary (min river) in China. Chem. Ecol. 34, 610–623. 445 Hu, X.F., Chen, F.S., Wine, M.L., Fang, X.M., 2017. Increasing acidity of rain in 446 subtropical tea plantation alters aluminum and nutrient distributions at the root- 447 soil interface and in plant tissues. Plant Soil 461, 261–274. 448 Jin, S., Wang, H., 2018. Relationships between soil pH and soil carbon in china’s 449 carbonate soils. Fresen. Environ. Bull. 27, 605–611. 450 Kimura, M., Murase, J., Lu, Y., 2004. Carbon cycling in rice field ecosystems in the 451 context of input, decomposition and translocation of organic materials and the 452 fates of their end products (CO2 and CH4). Soil Biol. Biochem. 36, 1399–1416. 453 . 454 Kováčik, J., Klejdus, B., Bačkor, M., Stork, F., Hedbavny, J., 2011. Physiological 455 responses of root-less epiphytic plants to acid rain. Ecotoxicology 20(2), 348–357. 456 Krusche, A.V., de Camargo, P.B., Cerri, C.E., Ballester, M.V., Lara, L.B., Victoria, R.L., 457 Martinelli, L.A., 2003. Acid rain and nitrogen deposition in a sub-tropical 458 watershed (piracicaba): ecosystem consequences. Environ. Pollut. 121(3), 389– 459 399. 460 Kunlanit, B., Vityakon, P., Puttaso, A., Cadisch, G., Rasche, F., 2014. Mechanisms 461 controlling soil organic carbon composition pertaining to microbial decomposition 462 of biochemically contrasting organic residues: evidence from middrifts peak area 463 analysis. Soil Biol. Biochem. 76(1), 100–108. 464 Lal, R., 2004. Soil carbon sequestration impacts on global climate change and food 465 security. Science 304,1623–1627. 16 466 Larssen, T., Carmichael, G.R., 2000. Acid rain and acidification in China: the 467 importance of base cation deposition. Environ. Pollut. 110(1), 89–102. 468 Li, J., Jia, C., Lu, Y., Tang, S., Shim, H., 2015. Multivariate analysis of heavy metal 469 leaching from urban soils following simulated acid rain. Microchem. J. 122, 89–95. 470 Liang, G., Hui, D., Wu, X., Wu, J., Liu, J., Zhou, G., Zhang, D., 2016. Effects of 471 simulated acid rain on soil respiration and its components in a subtropical mixed 472 conifer and broadleaf forest in southern China. Environ Sci.- Proc. Imp. 18(2), 473 246–255. 474 Liao, L., Chen, C., 1991. Effects of simulated acid rain on root growth of Cunninghamia 475 lanceolata and Schima superba saplings planted in acidified soil. J. Environ. Sci. 476 3(2), 45–52. 477 Liu, Y., Li, F.B., Xia, W., Xu, J.M., Yu, X.S., 2013. Association between ferrous iron 478 accumulation and pentachlorophenol degradation at the paddy soil-water 479 interface in the presence of exogenous low-molecular-weight dissolved organic 480 carbon. Chemosphere 91(11), 1547–1555. 481 Lu, R.K., 1999. Analysis methods of soil science and agricultural chemistry. Beijing: 482 Agriculture Science and Technology Press. 483 Malkanthi, D.R.R., Yokoyama, K., Yoshida, T., Moritsugu, M., Matsushita, K., 1995. 484 Effects of low pH and AI on growth and nutrient uptake of several plants. Soil Sci. 485 Plant Nutr. 41, 161–165. 486 Medeiros, J.S., Tomeo, N.J., Hewins, C.R., Rosenthal, D.M., 2016. Fast-growing Acer 487 rubrum differs from slow-growing Quercus alba in leaf, xylem and hydraulic trait 488 coordination responses to simulated acid rain. Tree physiol. 36(8), 1032–1044. 489 Ministry of Ecology and Environment of the People’s Republic of China., 2018. 490 Bulletin on the state of China's ecological environment in 2017. pp 15–16. 491 Nawaz, R., Parkpian, P., Garivait, H., Anurakpongsatorn, P., Delaune, R.D., Jugsujinda, 492 A., 2012. Impacts of acid rain on base cations, aluminum, and acidity development 493 in highly weathered soils of Thailand. Commun. Soil Sci. Plan. 43(10), 1382–1400. 494 Nicol, M.J., Tjandrawan, V., Zhang, S., 2016. Cathodic reduction of iron(III) and 495 copper(II) on various sulfide minerals in chloride solutions. Hydrometallurgy 166, 17 496 113–122. 497 Ouyang, X.J., Zhou, G.Y., Huang, Z.L., Liu, J.X., Zhang, D.Q., Jiong, L.I., 2008. Effect 498 of simulated acid rain on potential carbon and nitrogen mineralization in forest 499 soils. Pedosphere 18(4), 503–514. 500 Pan, G., Li, L., Wu, L., Zhang, X., 2004. Storage and sequestration potential of topsoil 501 organic carbon in China's paddy soils. Global Change Biol. 10, 79–92. 502 Pinheiro, J., Bates, D., DebRoy, S., Sarkar, D., Core, T.R., 2016. nlme: Linear and 503 Nonlinear Mixed Effects Models. R package version 3.1–126, http://CRAN. 504 Purakayastha, T.J., Rudrappa, L., Singh, D., Swarup, A., Bhadraray, S., 2008. Long- 505 term impact of fertilizers on soil organic carbon pools and sequestration rates in 506 maize-wheat-cowpea cropping system. Geoderma 144, 370–378. 507 Raamsdonk, L.M., Teusink, B., Broadhurst, D., Zhang, N.S., Hayes, A., Walsh, M.C., 508 Berden, J.A., Brudle, K.M., Kell, D.K., Rowland, J.J., Westerhoff, H.V., van 509 Dam, K., Oliver, S.G., 2001. A functional genomics strategy that uses 510 metabolome data to reveal the phenotype of silent mutations. Nat. Biotechnol. 19, 511 45–50. 512 Revelle, W., 2010. Package “psych.” Retrieved August 24, 2010, from 513 http://personality‐project.org/r/psych_manual.pdf 514 Rosi-Marshall, E.J., Bernhardt, E.S., Buso, D.C., Driscoll, C.T., Likens, G.E., 2016. 515 Acid rain mitigation experiment shifts a forested watershed from a net sink to a 516 net source of nitrogen. P. Natl. Acad. Sci. USA 113(27), 7580–7583. 517 Schmidt, I.K., Michelsen, A., Jonasson, S., 1997. Effects of labile soil carbon on 518 nutrient partitioning between an arctic graminoid and microbes. Oecologia 112(4), 519 557–565. 520 Smith, V.H., Tilman, G.D., Nekola, J.C., 1999. Eutrophication: impacts of excess 521 nutrient inputs on freshwater, marine, and terrestrial ecosystems. Environ. Pollut. 522 100(1–3), 179–196. 18 523 Song, Y., Song, C., Ren, J., Tan, W., Jin, S., Jiang, L., 2018. Influence of nitrogen 524 additions on litter decomposition, nutrient dynamics, and enzymatic activity of 525 two plant species in a peatland in northeast China. Sci. Total Environ. 625, 640– 526 646. 527 Wang, C., Liu, X.T., Tong, C., Chen, X.X., Chen, Y.Y., Mou, X.J., Wan, S.A., 2017a. 528 Effects of hydrologic and salinity gradients on soil organic carbon composition in 529 Min river estuarine wetland. China Environ. Sci. 37(10), 3919–3928. 530 Wang, L., Wang, W., Zhou, Q., Huang, X., 2014a. Combined effects of lanthanum (III) 531 chloride and acid rain on photosynthetic parameters in rice. Chemosphere 112, 532 355–361. 533 Wang, R.L., Christian, S., Dayan, F.E., Song, Y.Y., Su, Y.J., Zeng, R.S., 2012. 534 Simulated acid rain accelerates litter decomposition and enhances the allelopathic 535 potential of the invasive plant wedelia trilobata (creeping daisy). Weed Sci. 60(3), 536 462–467. 537 Wang, W., Lai, D.Y.F., Wang, C., Pan, T., Zeng, C., 2015a. Effects of rice straw 538 incorporation on active soil organic carbon pools in a subtropical paddy field. Soil 539 Till. Res. 152, 8–16. 540 Wang, W., Lai, D.Y.F., Abid, A.A., Neogi, S., Xu, X., Wang, C., 2018c. Effects of steel 541 slag and biochar incorporation on active soil organic carbon pools in a subtropical 542 paddy field. Agronomy 8, 135. 543 Wang, W., Lai, D.Y.F., Li, S., Kim, P.J., Zeng, C., Li, P., Liang, Y., 2014c. Steel slag 544 amendment reduces methane emission and increases rice productivity in 545 subtropical paddy fields in China. Wetl. Ecol. Manag. 22, 683–691. 546 Wang, W., Sardans, J., Lai, D.Y.F., Wang, C., Zeng, C., Tong, C., Liang, Y., Peñuelas, 547 J., 2015b. Effects of steel slag application on greenhouse gas emissions and crop 548 yield over multiple growing seasons in a subtropical paddy field in China. Field 549 Crop. Res. 171, 146–156. 550 Wang, W., Sardans, J., Tong, C., Wang, C., Ouyang, L., Bartrons, M., Peñuelas, J., 2016. 551 Typhoon enhancement of n and p release from litter and changes in the litter N:P 552 ratio in a subtropical tidal wetland. Environ. Res. Lett. 11(1), 014003. 19 553 Wang, W., Sardans, J., Wang, C., Pan, T., Zeng, C., Lai, D., Bartrons, M., Peñuelas, J., 554 2017b. Straw application strategy to optimize nutrient release in a southeastern 555 China rice cropland. Agronomy 7(4), 84. 556 Wang, W., Sardans, J., Wang, C., Zeng, C., Tong, C., Bartrons, M., Peñuelas, J., 2018b. 557 Steel slag amendment increases nutrient availability and rice yield in a subtropical 558 paddy field in China. Expl. Agric. 54, 842–856. 559 Wang, W., Sardans, J., Zeng, C., Zhong, C., Li, Y., Peñuelas, J., 2014b. Response of soil 560 nutrient concentrations and stoichiometry to increased human disturbance in a 561 subtropical tidal wetland. Geoderma 232–234, 459–470. 562 Wang, W., Wang, C., Sardans, J., Tong, C., Jia, R., Zeng, C., Peñuelas, J., 2015c. Flood 563 regime affects soil stoichiometry and the distribution of the invasive plants in 564 subtropical estuarine wetlands in China. Catena 128, 144–154. 565 Wang, W., Wang, C., Sardans, J., Tong, C., Ouyang, L., Asensio, D., Gargallo-Garriga, 566 A., Peñuelas, J., 2018a. Storage and release of nutrients during litter 567 decomposition for native and invasive species under different flooding intensities 568 in a Chinese wetland. Aquat. Bot. 149, 5–16. 569 Wang, Y., Liu, X., Tang, C., Xu, J., 2013. pH change, carbon and nitrogen 570 mineralization in paddy soils as affected by Chinese milk vetch addition and soil 571 water regime. J. Soil Sediment. 13(4), 654–663. 572 Warren, D.R., Kraft, C., Josephson, D.C., Driscoll, C.T., 2017. Acid rain recovery may 573 help to mitigate the impacts of climate change on thermally sensitive fish in lakes 574 across eastern North America. Global Change Biol. 23(6), 2149–2153. 575 Wei, H., Liu, W., Zhang, J., Qin, Z., 2017. Effects of simulated acid rain on soil fauna 576 community composition and their ecological niches. Environ. Pollut. 220, 460– 577 468. 578 Wissing, L., Kölbl, A., Vogelsang, V., Fu, J., Cao, Z., Kögel-Knabner, I., 2011. Organic 579 carbon accumulation in a 2000-year chronosequence of paddy soil evolution. 580 Catena 87, 376–385. 581 Wu, J., Liang, G., Hui, D., Deng, Q., Xiong, X., Qiu, Q., Liu, J., Chu, G., Zhou, G., 582 Zhang, D., 2016. Prolonged acid rain facilitates soil organic carbon accumulation 20 583 in a mature forest in southern China. Sci. Total Environ. 544, 94–102. 584 Xu, M., Lou, Y., Sun, X., Wang, W., Baniyamuddin, M., Zhao, K., 2011b. Soil organic 585 carbon active fractions as early indicators for total carbon change under straw 586 incorporation. Biol. Fert. Soils 47, 745–752. 587 Xu, R., Wang, J., Zhang, G., Dai, Q., 2010. Changes of microbe and organic matter 588 content in paddy soil applied with straw, manure and nitrogen fertilizer. Acta Ecol. 589 Sin. 21(13), 3584–3590. 590 Xu, S., Shi, X., Zhao, Y., Yu, D., Li, C., Wang, S., Tan, M., Sun, W., 2011a. Carbon 591 sequestration potential of recommended management practices for paddy soils of 592 China, 1980–2050. Geoderma 166, 206–213. 593 Zhang, B., Pang, C., Qin, J., Liu, K., Xu, H., Li, H., 2013. Rice straw incorporation in 594 winter with fertilizer-N application improves soil fertility and reduces global 595 warming potential from a double rice paddy field. Biol. Fert. Soils 49, 1039– 596 1052. 597 Zhang, L.M., Shan, L., Bouman, B.A.M., Xue, C.Y., Wei, F.T., Tao, H.B., Yang, X.G., 598 Wang, H.Q., Zhao, D.L., Dittert, K., 2009. Response of aerobic rice growth and 599 grain yield to n fertilizer at two contrasting sites near Beijing, China. Field Crop. 600 Res. 114(1), 45–53. 601 Zhu, S., Vivanco, J.M., Manter, D.K., 2016. Nitrogen fertilizer rate affects root 602 exudation, the rhizosphere microbiome and nitrogen-use-efficiency of maize. 603 Appl. Soil Ecol. 107, 324–333. 604 605 21 606 Tables 607 Table 1 608 Pearson correlation analysis of association between carbon fractions and environmental factors. Early paddy Late paddy Variable SOC DOC MBC EOC SOC DOC MBC LOC TN 0.575** 0.504** -0.535** -0.031 0.253 0.341* -0.078 -0.338* Available N 0.503** 0.654** -0.676** 0.058 0.362* 0.757** -0.184 -0.217 NO3- -0.301 0.384* -0.200 -0.360* 0.213 -0.271 0.268 -0.062 NH4+ 0.604** 0.472** -0.568** 0.201 0.290 0.749** -0.223 -0.187 Microbial N -0.125 0.417* -0.173 -0.259 -0.126 -0.158 0.423* -0.044 TP 0.811** -0.378* 0.256 0.175 0.480** 0.065 -0.225 -0.112 Available P 0.327 -0.106 -0.332* 0.362* 0.433** 0.202 0.132 0.187 C:N ratio 0.294 -0.742** 0.603** 0.305 0.361* -0.164 0.081 0.283 C:P ratio 0.128 0.422* -0.513** 0.134 0.380* 0.124 0.295 0.056 N:P ratio -0.137 0.747** -0.698** -0.153 -0.086 0.182 0.102 -0.179 LOC:available N -0.403* -0.647** 0.627** 0.304 -0.333* -0.579** 0.256 0.687** LOC:available P -0.094 0.019 0.191 0.445** -0.261 -0.216 0.039 0.847** 22 available N:available P 0.397* 0.673** -0.529** -0.075 0.082 0.631** -0.266 -0.310 MBC:MBN 0.283 -0.620** 0.618** 0.155 -0.012 -0.082 0.107 0.355* Total Fe -0.237 0.428** -0.229 0.010 0.134 0.141 -0.177 0.113 Fe2+ 0.600** 0.441** -0.527** 0.093 -0.015 0.533** -0.325 -0.159 Fe3+ -0.385* 0.320 -0.100 -0.012 0.167 -0.382* 0.130 0.291 Salinity 0.551** 0.159 -0.483** 0.399* 0.338* 0.690** -0.210 -0.291 Water content 0.542** 0.309 -0.539** 0.372* 0.262 0.503** -0.318 -0.022 pH -0.044 0.275 -0.307 0.221 -0.373* 0.205 -0.099 0.056 Temperature -0.244 -0.745** 0.745** -0.084 0.232 0.626** -0.231 -0.046 609 *P < 0.05; **P < 0.01. 610 611 612 613 614 615 616 617 618 619 620 23 621 24 622 Table 2 623 Discriminant analysis of treatment effects on soil traits (independent continuous variables), with 624 season as a categorical independent factor. 625 pH 4.5 pH 3.5 pH 2.5 M = 49.9 M =5 5.4 M = 10.5 Control F = 8.76 F = 9.72 F = 10.0 P < 0.00001 P < 0.00001 P < 0.00001 M = 7.61 M = 14.4 pH 4.5 F = 1.38 F = 2.52 P = 0.17 P = 0.004 M = 11.8 pH 3.5 F = 2.03 P = 0.019 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 25 656 657 Table 3 658 General discriminant analysis (Wilks’ λ and P) of soil variables。 659 Test Value F P pH 0.602 8.80 <0.0001 Fe2+ 0.758 4.25 0.011 Fe3+ 0.823 2.86 0.049 salinity 0.828 2.78 0.049 Water % 0.713 5.378 0.0033 temperature 0.974 0.368 0.78 SOC 0.630 7.828 0.0003 DOC 0.881 1.80 0.16 MBC 0.939 0.871 0.46 LOC/SOC 0.886 1.72 0.18 DOC/SOC 0.874 1.93 0.14 MBC/SOC 0.944 0.794477 0.50 NO3 - 0.972 0.381 0.77 NH4+ 0.948 0.730 0.54 MBN 0.859 2.19 0.10 Total P 0.871 1.98 0.13 Available P 0.827 2.79 0.053 C/N 0.866 2.07 0.12 C/P 0.871 1.98 0.13 N/P 0.904 1.42 0.25 LOC/N-available 0.840 2.53 0.070 LOC/P-available 0.847 2.43 0.081 N-available/P-available 0.950 0.706 0.55 MBC/MBN 0.858 2.20 0.10 timing 0.596 1.52 0.11 660 661 662 663 664 665 666 667 26 668 669 Figure legends 670 671 Fig. 1. Treatment effects on SOC, LOC, DOC, and MBC in early (A, C, E, and G, 672 respectively) and late (B, D, F, and H, respectively) paddies. Data are means ±SE. 673 Different uppercase letters indicate treatment differences among rice growth stages and 674 lowercase letters indicate treatment differences within a rice growth stage (P < 0.05). 675 676 Fig. 2. Treatment effects on total N, available N, NO3-, NH4+, MBN, total P, and 677 available P in early (A, C, E, G, I, K, and M, respectively) and late (B, D, F, H, J, L, 678 and N, respectively) paddies. Data are means ±SE. Different uppercase letters indicate 679 treatment differences among rice growth stages and lowercase letters indicate treatment 680 differences within a rice growth stage (P < 0.05). 681 682 Fig. 3. Treatment effects on ratios of C:N, C:P, N:P, LOC:available N, LOC:available 683 P, available N:available P, and MBC:MBN in early (A, C, E, G, I, K, and M, 684 respectively) and late (B, D, F, H, J, L, and N, respectively) paddies. Data are means 685 ±SE. Different uppercase letters indicate treatment differences among rice growth 686 stages, and lowercase letters indicate treatment differences within a rice growth stage 687 (P < 0.05). 688 689 Fig. 4. Treatment effects on salinity, water content, temperature, and pH in early (A, C, 690 E, and G, respectively) and late (B, D, F, and H, respectively) paddies. Data are means 691 ±SE. Different uppercase letters indicate treatment differences among rice growth 692 stages and lowercase letters indicate treatment differences within a rice growth stage (P 693 < 0.05). 694 695 696 Fig. 5. Treatment effects on total active Fe, Fe2+, and Fe3+ in early (A, C, and E, 697 respectively) and late (B, D, and F, respectively) paddies. Data are means ±SE. Different 27 698 uppercase letters indicate treatment differences among rice growth stages and lowercase 699 letters indicate treatment differences within a rice growth stage (P < 0.05). 700 701 702 Fig. 6. GDA biplot of treatment effects on soil traits by grouping factor (treatment by 703 season). Data are means ±95% CI and the standardized canonical discriminate 704 function coefficients for the first two roots represent the soil variables as independent 705 variables. pH: soil pH; Fe2+: soil Fe2+ concentration; Fe3+: soil Fe2+ concentration; 706 salinity: soil salinity; SWC: soil water content; T: soil temperature; DOC: dissolved 707 organic carbon; SOC: soil organic carbon; MBC: microbial carbon concentration; 708 MBN: microbial nitrogen concentration; NO3-: soil nitrate ion concentration; NH4+: 709 soil ammonium concentration; TP: soil total P concentration; avaiP: soil P plant 710 availability; C:N: soil total C:N ratio; C:P: soil total C:P ratio; N:P: soil total N:P 711 ratio; MBC:MBN: MBC:MBN ratio; LOC:Navai: LOC:available N ratio; LOC:Pavai: 712 LOC:available P ratio; DOC:SOC: DOC:SOC ratio; MBC:SOC: MBC:SOC ratio; 713 Navai:Pavai: Navai:Pavai ratio; and, LOC:SOC: LOC:SOC ratio. 714 715 716 717 718 719 720 721 722 723 724 725 726 727 28 728 30 30 Control pH 4.5 A B SOC concentration (g kg-1) pH 3.5 pH 2.5 SOC concentration (g kg ) -1 24 Bb Bd 24 Bc Aab Aa Aa Aa Aa Ac Ab Aa Aa Aa Ab Ab Aa AaAa Ab Ab Aa Aa Aa Ab 18 18 12 12 6 6 0 0 Turn green Jointing Mature Turn green Jointing Mature 729 9.0 9.0 C D LOC concentration (g kg ) LOC concentration (g kg ) Bb -1 Ab -1 7.5 7.5 Ab Bb Aa Ba Aa Aab 6.0 Aa AaAa Aa Aa Aa Aa Aa 6.0 Aa Aab Aa Aab Aa Aa Aa Aa 4.5 4.5 3.0 3.0 1.5 1.5 0.0 0.0 Turn green Jointing Mature Turn green Jointing Mature 730 90 90 E F DOC concentration (mg kg ) DOC concentration (mg kg ) Cc -1 -1 Cb Bb Bb 72 Ca 72 Ba BbBab Bab Ba Ba Bb Ab Ba Bb Aab 54 Aa AaAa Aa 54 Aa Aa Aab Aa 36 36 18 18 0 0 Turn green Jointing Mature Turn green Jointing Mature 731 200 200 Cc G Bb H MBC concentration (mg kg ) MBC concentration (mg kg ) -1 Bb -1 160 160 Ab Bb Aa Cb Bb Aab Bb Ab Aa Aa Aa Bb Cb Aa 120 120 Ba Aa Ab Ba 80 Aa Aa 80 Aa 40 40 0 0 Turn green Jointing Mature Turn green Jointing Mature 732 733 Fig. 1. 29 4 4 Control pH 4.5 A B Total N concentration (g kg ) -1 Total N concentration (g kg ) pH 3.5 pH 2.5 -1 3 3 Ab Aa Aab Aa Ba Ba Aa Aa Ab Ba Ba Aa Ac Aa Ac Bb Bb Ba Ab 2 Cc 2 Ba Ca Ca Ba 1 1 0 0 Turn green Jointing Mature Turn green Jointing Mature 734 150 150 Available N concentration (mg kg ) Available N concentration (mg kg ) -1 D -1 C 120 Cb Cb 120 Ca Ba 90 90 Bc Bc Bb 60 Ab Ba Aa Ba Aa 60 Bb Aa Ba Ab Aa Aa Aa Aa Aab Ab Aa Aa 30 30 0 0 Turn green Jointing Mature Turn green Jointing Mature 735 Available NO3 concentration (mg kg ) Available NO3 concentration (mg kg ) 40 40 -1 -1 E F Aa 32 Aa 32 Aa 24 Aa 24 Bab Ba 16 16 Aa Aa Bb - - 8 Aa Ba Ba Aa 8 Bb Ba Bb Bb Ac Ba Ab Cd Cc Bd ABb 0 0 Turn green Jointing Mature Turn green Jointing Mature 736 Available NH4 concentration (mg kg ) Available NH4 concentration (mg kg ) 120 120 -1 -1 G H Aa Aa 90 90 Aa Ac Ba Ab Aa 60 Bb Ba 60 Ac Bc Ad ABa Cc Ba Ca Ba Ba Ba Cb CbCb BbBab + + 30 30 0 0 Turn green Jointing Mature Turn green Jointing Mature 737 30 80 80 Cb MBN concentration (mg kg ) I J -1 MBN concentration (mg kg ) -1 Bc 60 60 Ba 40 Ab Aa 40 Ab Ba Ba Bc Bb Bc BaBa Ba Aa Bb Abc 20 Aa Aa 20 Ab Aa Aa Aa Aa 0 0 Turn green Jointing Mature Turn green Jointing Mature 738 3 3 K L Total P concentration (g kg ) Aa -1 Total P concentration (g kg ) ABa Aa -1 Ab Ab AaAa Aa Aa ABab Ac Aa Aa Ba Ab Aa Ba Ba Aa Ba Cab 2 Bb Ba 2 Bb 1 1 0 0 Turn green Jointing Mature Turn green Jointing Mature 739 150 150 Available P concentration (mg kg ) Available P concentration (mg kg ) -1 M N -1 120 Aa 120 Aa Bb Bb 90 90 Bc Ad Ab Bab 60 Ab Bb AaAab 60 Aa Ac Aa Aa Bb Ab Ba Aa Aa Cb Bb Aa 30 30 0 0 Turn green Jointing Mature Turn green Jointing Mature 740 741 742 Fig. 2. 743 744 745 746 31 16 16 Control pH 4.5 A B pH 3.5 pH 2.5 12 Bc Ba Bb Cb 12 Ba Ca Ba Ba Bc C:N ratio Abc C:N ratio Ab Aa AaAa Ba Aa Aa Ab Aa Aa Aa Aa Aab Aa 8 8 4 4 0 0 Turn green Jointing Mature Turn green Jointing Mature 747 16 16 C D 12 12 Ba ABb BcABab Aa AaAa Aab Aab C:P ratio Bb Bab Aa Aa Aa Aa Ab AaAa AaAa ABa Aab C:P ratio Aa Aa 8 8 4 4 0 0 Turn green Jointing Mature Turn green Jointing Mature 748 2.0 2.0 E F 1.6 1.6 Ca N:P ratio BaBaCa Bb Bb Cb N:P ratio 1.2 Ba Bab 1.2 BaBabBab BaBa Ba Ba AaAa Aa Aa Aa AaAa Aa 0.8 0.8 0.4 0.4 0.0 0.0 Turn green Jointing Mature Turn green Jointing Mature 749 250 250 G H Aa LOC:available N ratio 200 LOC:available N ratio 200 Aa Aa Aa Ab Aa Aa 150 Aab 150 Ab Ab Aa ABa Abc ABa Aa Aa 100 Ac 100 Bb Bb Ba Bb Bab 50 BbBb 50 0 0 Turn green Jointing Mature Turn green Jointing Mature 750 32 120 120 I J Aa LOC:available P ratio LOC:available P ratio 90 Aa 90 Aa Aa Aa Aa Aa Aa Aa Aab ABa Aab Aa Aa Aa Aa Aa Aa Aa Abc Aa Bc Bb Aa 60 60 30 30 0 0 Turn green Jointing Mature Turn green Jointing Mature 751 2.0 2.0 K L Available N:available P ratio Available N:available P ratio Aa 1.6 1.6 Ab 1.2 1.2 Acd Ac Aa Aa 0.8 Ba Ba Ba 0.8 Ab Bc Bb Ba Ca Ac Bd Ba Ba Ba ABa Ba Ba 0.4 Ca Ba 0.4 0.0 0.0 Turn green Jointing Mature Turn green Jointing Mature 752 40 40 M N Aa 32 32 MBC:MBN ratio MBC:MBN ratio Aa 24 24 Aab 16 Aa 16 Aab Aab Aa ABa 8 ABb 8 Ba Aab Aa Ba Ba Ba Aa Ba BabBab Bb Bb Bb Bb ABb 0 0 Turn green Jointing Mature Turn green Jointing Mature 753 754 Fig. 3. 755 756 757 758 759 760 761 762 763 33 1.5 1.5 Control pH 4.5 A B pH 3.5 pH 2.5 1.2 Aa Aa 1.2 Salinity (mS cm ) -1 Salinity (mS cm ) -1 Aa Aa 0.9 Aa Aa 0.9 Aab Aab Aab Ab Abc Ba Ba Ba BabBa 0.6 Ba BaBa 0.6 Bc BbBab Bbc Bc 0.3 0.3 0.0 0.0 Turn green Jointing Mature Turn green Jointing Mature 764 90 90 Aa Aa C D Ab Ab Ab Ab AbAab ABab Ba 72 Ba 72 Bc Bbc BbBabBab Bab Ca Water content (%) Water content (%) CbBab Ca Bab Cab 54 Cb 54 36 36 18 18 0 0 Turn green Jointing Mature Turn green Jointing Mature 765 35 35 E Aa Aa Aa Aa F Aa Aa AabAb Ba Ba Bb Bab 28 28 Ba Ba Ba Ba Temperature (℃) Ca Ca Ca Ca Temperature (℃) Ca CabCbCb 21 21 14 14 7 7 0 0 Turn green Jointing Mature Turn green Jointing Mature 766 8 8 G H Aa Aa Aa Aa AbAcAb Ba Ab Ab ABc Ca Aab Bab Bb Ab Ab Ab AbAbBc Ab Bb Cb 6 6 pH 4 4 pH 2 2 0 0 Turn green Jointing Mature Turn green Jointing Mature 767 768 Fig. 4. 769 34 20 20 Control pH 4.5 A B Total Fe concentration (g kg ) -1 Total Fe concentration (g kg ) pH 3.5 pH 2.5 -1 16 16 Ba Ba Ab Aa Aa Ab Aa Aa Aa Aa Ab Aa Aa Aa Aa Aa Aa Aa Aa Aa Aa AaAaAa 12 12 8 8 4 4 0 0 Turn green Jointing Mature Turn green Jointing Mature 770 8 8 Total Fe concentration (g kg ) C D -1 Total Fe concentration (g kg ) -1 Ba Ba 6 6 Aa Ca 4 4 Bc Ab Ba Bb Bb ABa Ac AaAaAa Aa Aa AaAa BbAb Aa Aa AaAa 2+ 2 2+ 2 0 0 Turn green Jointing Mature Turn green Jointing Mature 771 16 16 Total Fe concentration (g kg ) E F -1 Total Fe concentration (g kg ) -1 Aa Ab 12 Aa Aa Aa 12 Ab Ba Aa Ba Ba Ab Ab Aa AaAa Aa Aa Aa BaAa Aa Aab Aab Aa 8 8 3+ 4 3+ 4 0 0 Turn green Jointing Mature Turn green Jointing Mature 772 773 774 775 Fig. 5. 776 35 Loadings -0.3 -0.2 -0.1 0 0.1 0.2 0.3 pH 2.5 SWC 0.2 2 LOC/SOC Root 2 (13.7%) LOC:Pavai 1 Fe2+ LOC:Navai 0.1 avaiP N:P Control Salinity Loadings Scores T C:P 0 NO3- NH + MBN 0 4 C:N MBC/SOC DOC/SOC pH 3.5 MBC MBC/MBN pH -1 -0.1 DOC SOC TP Navai:PavaiFe3+ -2 pH 4.5 -0.2 -6 -4 -2 0 2 4 6 Root 1 (77.9%) Scores 777 778 779 780 781 782 Fig. 6. 783 784 785 786 787 788 36