Semi-Perennial Rice Sustains Yield and Food Safety Under Heatwaves and Beyond

Semi-Perennial Rice Sustains Yield and Food Safety Under Heatwaves and Beyond | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Semi-Perennial Rice Sustains Yield and Food Safety Under Heatwaves and Beyond Zheng Chen, Sha Zhang, Jing Song, Longhua Wu, Yong-Guan Zhu This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-5397288/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Heatwaves threaten global rice security and challenge the United Nations’ Zero Hunger goals. Semi-perennial rice, with its ratooning ability to regrow from stubble after hot seasons, has shown potential for sustainable production. Here, we investigated the vulnerability and resilience of ratoon cropping system through two-year manipulative experiments of realistic heatwaves and analyzed experimental and global datasets. Our findings reveal that ratoon crops can compensate for up to 100% of heatwave-induced yield losses, maintain nutritional quality, and reduce arsenic (As) contamination by half. These benefits arise from widespread yield trade-offs between main and ratoon crops, and the seasonal mismatch between soil As availability and plant uptake. Ratoon practice can increase annual rice production by 3.9 million Mg in marginal regions, feeding 80 million people while concurrently reducing As exposure by 535 kg in China. This underscores the importance of integrating semi-perennial rice into climate-resilient agriculture. Earth and environmental sciences/Environmental sciences/Environmental impact Scientific community and society/Agriculture Biological sciences/Plant sciences/Plant stress responses/Heat Earth and environmental sciences/Biogeochemistry/Element cycles Figures Figure 1 Figure 2 Figure 3 Figure 4 Main Heatwaves severely impacted Asia's rice production, threatening food security for billions of people 1 , 2 . Short-term extreme temperatures during heatwaves can disrupt reproductive yield formation within hours, causing significant crop failures 2 , 3 . These conditions have stagnated yields in major rice-producing regions 4 and hinder progress toward the United Nations' Sustainable Development Goal of 'Zero Hunger', which aims to improve both yield and nutritional quality (e.g., iron (Fe), zinc (Zn), selenium (Se)) while reducing carcinogenic contaminants like As 5 . The micronutrient deficiencies, or “hidden hunger”, affects about two billion people 6 , and dietary As from rice poses a growing global health threat amid climate change 7 . Despite these challenges, feasible climate-smart solutions remain scarce. A promising solution is transitioning from traditional annual rice cultivation to semi-perennial production, which can desynchronize key crop stages from heatwave periods and increase the multi-cropping index 8 , 9 . Semi-perennial rice, with its ratooning phenotype, regrows from old stubbles, enabling "one sowing, multiple harvests" per year 10 . Although typically grown annually, most rice varieties are inherently semi-perennial, as demonstrated by a screening of 2,000 varieties that produced 0.8-3 Mg ha − 1 ratoon yield 11 . This system offers agronomic potential in radiation-limited regions while reducing input dependency on tillage, fertilizers, and pesticides and saving 3/5 cropping time 12 , 13 . Recent high-yield records of ratoon rice 14 and perennial rice 13 underscore the feasibility of this transition across wider agroclimatic zones. However, the overall potentials of semi-perennial rice responding to heatwaves are not well evaluated yet 15 . Heatwaves disrupt plant carbon metabolism and storage 16 , reducing yield and grain weight while altering the elemental balance and carbon-to-element ratios (Fig. 1 ). Although abrupt extreme stress causes infertility 16 , we hypothesize carbon for seed production would be diverted to compensatory ratooning growth for a second harvest, negatively correlating main yields (Hypothesis 1) 14 , 17 . Such compensation effects have been observed with rice crop losses due to climate disasters like floods and droughts 18 , 19 Furthermore, cooler ratoon seasons could slow soil-plant biogeochemical processes 20 , 21 , reducing nonessential toxic element uptake and translocations (Hypothesis 2) 22 , potentially enhancing yield and food safety in the post-heatwave period. In this study, we comprehensively investigated the responses of ratoon cropping system to heatwaves. In 2021 and 2022, we captured and manipulated unprecedented heatwaves at the seasonal scale for various scenarios, using customized sunlit AmbControl facilities (details in discussions of ref 23 ). Seventy heatwave days during the crop season (Supplementary Fig. 1) and 27 extreme days with daily Tmax exceeding 40.0°C were observed in 2022, marking the driest and hottest summer in East Asia monsoon region for six decades (Supplementary Fig. 2). Contrastingly, 2021 rice production was entirely under rainy season with 51 heatwave days, resulting the wettest hot summer for last two decades. We quantified yield responses, nutritional qualities, and As-related food safety risks under four scenarios, i.e. recent and elevated heatwaves compared to mild and cool climate scenarios (Fig. 1 ). We extended the growth duration to ratoon season to assess the ratoon effect on yield and qualities. Additionally, a global dataset of rice As levels from 5,104 site-year-variety instances and 74 paired main-ratoon grain As surveys in China are compiled to illustrate the agronomic potential of ratoon cropping for healthier rice production under a changing climate. Using four sunlit AmbControl facilities, which replicate ambient temperature dynamics by one-minute resolution, we created different climate scenarios to assess heatwave impacts on ratoon rice in 2021 and 2022 (annul temperature dynamics available in ref 23 ). Rice was grown under various temperature scenarios, including elevated heatwaves (+ 1.5°C), ambient heatwaves (+ 0°C), mild climate (daily Tmax < 30°C in 2021; daily Tmax < 33°C in 2022), and cool climate (-4°C in 2021; fixed 28.9°C in 2022). Theoretically, heatwaves during critical stages (flowering and grain filling) reduce carbon allocation to grain, lowering yield, while excess carbohydrates promote new panicle growth for a second harvest 24 . This carbon loss in grains can increase concentrations of other elements. Heatwaves accelerate soil chemical reactions, raising nutrient and As uptake and contamination risks. Given the human toxicity of inorganic As (iAs) and the phytotoxicity of organic As (oAs), these risks are exacerbated under heatwaves but mitigated under cooler ratoon conditions. Results Main Yield loss and Ratoon Yield Compensation Significant yield losses were recorded during heatwaves, but ratoon rice compensated with improved yields. Under two-year heatwave scenarios, rice variety Yliangyou1 experienced drastic average yield reductions of 61–78% (Fig. 2 a, b). Uncontrolled ambient and open-field observations in two year confirmed similar yield losses, ranging from 69 to 100% (Supplementary Fig. 3). Interestingly, despite comparable mean temperatures between heatwave and cool climate conditions in 2022 (28.2°C vs. 28.7°C), yields under heatwaves was 88.9% lower. Heatwaves also caused significant grain chalkiness, degrading the commercial value of rice (Supplementary Fig. 4), while minimal chalkiness (< 3%) was observed under mild and cool conditions. Chalkiness is the typical phenotype which can be induced by high temperatures 25 . This indicated that extreme temperatures have a more pronounced effect on yields and grain quality than seasonal averages. The ratooning phenotype of semi-perennial rice demonstrated strong resilience to heatwaves, as it compensated summer yield losses with improved growth during cooler periods (Fig. 2 c, Hypothesis 1). This yield compensation was linearly quantified: declines in main crop yields during heatwaves led to proportional increases in ratoon yields. In 2022, ratoon rice maintained an average of 11 grams per plant annually across diverse climate conditions (P > 0.05). This yield consistency underscores the inherent resilience of rice to extreme climate events. A broader dataset for the rice cultivar Yliangyou1 (N = 37 site-year surveys across China, Supplementary Fig. 5) revealed a clear tradeoff pattern (P = 0.015, Fig. 2 c) between main and ratoon crop yields across high-yield sites (N = 15 out of 37). The main-ratoon yield tradeoff pattern were also reported for other site-variety observations 14 . To further assess yield compensation, we examined literature on rice crops experiencing significant main crop yield failures (> 80%) due to heatwaves, floods, droughts, or intentional mowing for forage purposes across China (Fig. 2 d, Supplementary Fig. 6). Ratoon cropping demonstrated substantial resilience to these disturbances, with yield compensations of 51.9 ± 11.7% (± 1 standard deviation, sd ) after heatwaves, 53.7 ± 16.9% after floods, 69.2 ± 13.9% after droughts, and 71.9 ± 29.8% after mowing compared to historical yield records (Fig. 2 d). However, ratoon yield compensation was inhibited by soil As contamination. In a supplementary experiment conducted in 2022, using an As-contaminated soil (47 mg kg⁻¹ total As, tAs), heatwaves resulted in near-total crop failure (Supplementary Fig. 7r). Even under mild and cool climates, both the main and ratoon crops suffered yield reductions due to serious straighthead disease. The dimethylarsinic (DMA) is a common cause to straighthead disease 26 . We verified that more than 83% of grain As was DMA (tAs range: 2,200–3,100 µg kg − 1 ) (Supplementary Fig. 7p,q). Efforts to mitigate grain As uptake through irrigation management in 2022 were moderately successful (tAs < 150 µg kg − 1 , Supplementary Fig. 8). However, main crop yields were still nonexistent under heatwaves, and ratoon yields were also minimal. These findings suggest that while ratoon rice exhibits resilience to extreme climatic conditions, its compensatory capacity is severely constrained by soil As contamination. Heatwave Effect and Ratoon Effect on Grain As Heatwaves increased the concentration of toxic elements, particularly As, in rice, while ratoon rice exhibited lower uptake of these elements. Our unique heatwave control facility enabled the creation of mild baseline climate scenarios, allowing us to quantify the effects of heatwaves and ratooning directly. Over two years, heatwaves increased total As (tAs) concentrations in grain by 28% ( P < 0.001), while ratooning practices reduced them by 37% ( P < 0.001, Fig. 3 c). We focused on As because, under continuously flooded rice cultivation, grain cadmium (Cd) (0.031 ± 0.026 mg kg⁻¹) and lead (Pb) concentrations (0.047 ± 0.033 mg kg⁻¹) across all treatments in both years were far below the United Nations’ food safety standards (0.2 mg kg⁻¹) (Supplementary Figs. 9,10). In contrast, grain As concentrations (0.52 ± 0.18 mg kg⁻¹) frequently exceeded food safety standards (tAs: 0.3 mg kg⁻¹; inorganic As [iAs]: 0.2 mg kg⁻¹). We further analyzed As speciation, differentiating between iAs and organic As (oAs). In 2022, across all treatments and seasons, oAs accounted for 28.5 ± 7.7% of tAs in grains (Fig. 3 b). However, in 2021, the oAs percentage was significantly higher, reaching 60.2 ± 12.4% (Supplementary Figs. 9,10). Notably, in 2021, despite daily Tmax being 5–10°C higher during the grain filling stage (data in ref 23 ), heatwaves did not increase grain iAs concentrations but increase oAs concentrations significantly (Supplementary Figs. 7,9). Mechanisms Heatwave-induced carbon loss and grain weight reduction are common drivers of changes in grain element concentrations. A piecewise linear model identified a turning point at a Tmax of 31.4°C or a Tmean of 24.3°C (Supplementary Fig. 11). Below this temperature, moderate heat increased grain weight ( P = 0.08), whereas excessive heat reduced it. Over two years, heatwaves decreased grain carbon content by 13% and caused a 30% weight loss, contributing to an "enrichment effect", where concentrations of other elements increased (Supplementary Figs. 12, 13). This uniquely resulted in negative correlations between grain weight and concentrations of trace elements, not only As but also Fe, Zn, Se, boron (B), molybdenum (Mo), lead (Pb), nickel (Ni), antimony (Sb), cobalt (Co), cadmium (Cd), and copper (Cu) (Supplementary Figs. 12,13). However, ratoon grain weight has no clear relationships with As concentrations ( P = 0.83 based on linear regression, Supplementary Fig. 14). This suggested that compared to baseline climates, only heat-stressed grain had increased As and rice nutrients. Further findings support Hypothesis 2: cooler temperatures reduce soil As availability, leading to lower grain As concentrations in ratoon crops. We analyzed harvests based on ratoon cropping calendars, revealing that early ratoon yields, which developed during high-temperature conditions, contained higher As concentrations, while second ratoon crops, grown entirely in cooler periods, had lower As levels. Time-series data of grain As concentrations (Fig. 3 b) and dissolved As in soils, with and without plants (Supplementary Fig. 15), demonstrated a decline in As with decreasing temperatures. Across elevated heatwave, ambient heatwave, and mild climate scenarios, iAs levels in ratoon grains often met or fell below the international food standard of 200 µg kg⁻¹, highlighting ratoon cropping's potential to mitigate As accumulation. Interestingly, despite the highest average temperature (28.9°C vs. ambient Tmean < 25.5°C), the fixed cool climate scenario during the second ratoon period resulted in higher grain As concentrations, suggesting that high temperatures are critical for maintaining elevated grain As concentrations. A third mechanism contributing to lower As levels in ratoon crops is reduced in-plant translocation of As during cooler ratoon seasons. Stepwise analyses of nodes, flag leaves, husks, and grains (Supplementary Figs. 7,9,10) confirmed this mechanism, which applies to most elements except magnesium (Mg) and Fe. Magnesium, essential for photosynthesis and carbon allocation, is highly mobile within plants 27 . Ratoon grains accumulated 64% higher Mg compared to main crops (Supplementary Figs. 14, 16,17), likely due to in-plant reallocations, as no additional Mg fertilizer was applied beyond nitrogen at the start of the main crop season. Responses Trajectory of Grain As to Key Drivers We further assessed the effects of grain weight, soil porewater As concentrations, and mean daily Tmax during the corresponding crop seasons on grain As concentrations (Fig. 3 d). Using Shapley Additive Explanations (SHAP) values from the TreeExplainer of a trained XGBoost model, we quantified the individual contributions of each feature relative to the baseline prediction. The XGBoost model explained 86.3% of the variation in grain As concentrations. Grain weight accounted for 21.2% of the variation, soil porewater As concentration for 34%, and temperature (Tmax) for 31.1% (Fig. 3 d). While the temperature effect on grain As followed a non-threshold response, the apparent sensitivity of grain As concentrations to temperature was lower than expected, increasing by 21 µg kg − 1 per 1°C rise. Broader Insights from Global Datasets Grain As concentrations were broadly influenced by seasonal temperature fluctuations. To analyze this trend further, we utilized two external datasets: the first contained paired As concentrations in main and ratoon rice grains from China (N = 74 paired data points, Fig. 4 a), while the second encompassed global grain As concentrations (N = 5104) from various harvest dates, spanning major rice-producing countries, different agronomic zones, water management practices, and 2,451 rice varieties (Fig. 4 b, Supplementary Fig. 18). The China dataset highlighted a consistent discrepancy between main and ratoon crop As concentrations, with ratoon rice exhibiting a 49.2 ± 1.8% reduction in grain As (Fig. 4 a). This reduction aligns closely with our experimental results from heatwave manipulations (Fig. 3 c). In the global dataset, a pronounced seasonal temperature dependence on As uptake was observed, with late-harvested grains showing a 21.9 ± 3% reduction in As under conventional cultivation (Fig. 4 b). This pattern held true when comparing grains from the same cultivars planted at different times, with those harvested in cooler seasons displaying an approximate 45.0% decrease in As concentrations. Discussion Our climate manipulation experiments have demonstrated the severe damage heatwaves inflict on rice systems but also revealed the significant potential of semi-perennial rice to provide additional yields and safer food. These findings are especially relevant for rice-producing regions, where global warming is increasing the frequency of heat extremes, posing serious risks to food security and exacerbating soil As contamination. One critical discovery from comparing heatwave and non-heatwave scenarios is the rapid yield compensation following reproductive failure (Supplementary Fig. 19). While this compensation did not fully recover the original yield potential, it significantly approached it. Unlike typical ratoon practices, where tillers emerge after the main crop matures or is harvested (as in cooler climates, Fig. 3 a), under heatwave conditions, ratoon tillers appeared concurrently with reproductive failure, redirecting surplus carbohydrates toward ratoon yield formation. The molecular mechanisms of ratoon compensation effect remain unclear. However, the response resembles that observed in rice subjected to artificial mowing or topping, where apical dominance is broken, promoting side tiller growth by redistributing plant resources (Fig. 1 d) 17 . Heatwaves appear to trigger a similar redistribution, enabling ratoon growth to partially compensate for main crop yield losses. However, deciphering the full metabolic network is challenging, as ratoon plants systematically differ from main crop rice. For instance, in trials with 429 global varieties, ratoon plants exhibited 50% reductions in plant height, panicle length, leaf size, and grain numbers compared to main crops 28 . Even though, some varieties have shown robust ratooning ability, with ratoon yields comparable to the main rice yield 14 . This suggests that ratooning under heatwaves or other stressors may represent an underappreciated form of resilience in rice systems. Despite the agronomic potential of ratoon rice, soil As contamination can undermine this promise. Yield reductions in elevated-As soils are linked to changes in As speciation, particularly microbial methylation, which can severely damage rice yields 26 . In elevated-As soils, grain oAs was found to be the predominant species (> 80%), which is unusual for Chinese rice (usually ~ 29%) 29 . This highlights the need for controlling soil As methylation, particularly as root exudates can promote methylation (Supplementary Fig. 20), leading to higher As uptake and potential phytotoxicity 30 . Although the exact impact of oAs on yield reductions remains uncertain due to limited data, its influence on food security should not be underestimated. Our findings provide valuable insights into the factors driving As accumulation in rice grains. First, grain weight loss and enhanced in-plant translocations during heatwaves are the primary drivers of increased elemental concentrations, aligning with experimental warming studies. Grain weight loss, however, has rarely been considered in explaining changes in the ionome 22 , 31 . Second, while elevated temperatures increase tAs in grains, this does not always correspond to higher levels of iAs, the primary food safety concern. This stands in contrast to previous reports that suggest each degree of temperature increase doubles the food safety risk from iAs 22 . Our data show that higher temperatures do not consistently enhance As mobility in the soil-plant system. Third, we observed that grain tAs and porewater tAs closely track temperature changes across our experiments, regional studies, and global datasets (Fig. 4 ). The reduced As in ratoon rice is largely attributed to lower soil As availability and reduced in-plant translocation under cooler conditions. This seasonal variability provides a concise framework for assessing food safety risks more accurately. Although ratoon rice tends to accumulate lower levels of As, it also shows reduced accumulation of essential micronutrients such as Fe, Zn, and Se (Extended Figs. 7, 16,17). However, in 2022, the reduction in grain Fe levels was not statistically significant, highlighting the role of physiological regulation in controlling the accumulation of certain essential elements. For other beneficial trace elements, agronomic biofortification may offer a potential solution. Ratoon rice holds substantial promise for addressing food security challenges, particularly in regions vulnerable to climate extremes. There are currently 1.1 million ha for ratoon rice in China. Based on nation-wide yield average of 3.64 Mg ha − 1 in China 32 , this cultivation method could yield an additional 3.9 million Mg currently, potentially increasing to as much as 12 million Mg with using radiation-marginal rice production areas. Assuming the average annual rice consumption per person is approximately 150 kg 33 . This can feed approximately 80 million people. Considering current and potential ratoon yields in China, ratooning could prevent about 535 Kg and potentially 1,700 kg tAs per year, respectively. This also exceed the rice As export (323 kg iAs per year) from United States' rice industry 34 . While our heatwave-manipulative experiments have provided critical insights, field studies under diverse real-world conditions are needed to validate ratoon rice's performance across broader agricultural systems. Further investigation into the molecular pathways governing heat-induced ratooning will be essential to fully unlock its potential, not just in rice but potentially in other crops. Our study demonstrates that ratoon rice can effectively compensate for yield losses caused by heatwaves while reducing food contamination risks from As. These findings suggest that ratoon rice cultivation could play a key role in enhancing food security and promoting the sustainability of agricultural systems. Policymakers and agricultural stakeholders should prioritize ratoon rice as a strategy for adapting rice production to the increasing threats posed by climate extremes. Methods Heatwave Definition Summer heatwaves are defined as periods with three consecutive days where the daily Tmax exceeds 36.2°C for at least one hour each day. This threshold is derived from the 30-year Tmax average (1961–1990) plus an additional 5°C at our study site. Rice and Soil We used a popular rice variety Ylaingyou1 (hybrid), which is moderately sensitive to temperature and for As bioaccumulation 35 . It has a cumulative harvested area of 2.7 million ha in China ( https://www.natesc.org.cn ). This variety has moderate ratooning ability 36 . Two local low-As paddy soils i.e., soil XC (31°29′49.095"N,120°32′49.927"E) and soil YCH (31°28′25.52"N,120°46′18.88"E) and an As-elevated paddy soil SY (30°0' 8.03"N, 120°47'58.14"E) were collected from Tai Lake area near our experimental site. Soil As were extracted by aqua regia and determined by Inductively coupled plasma mass spectrometry (ICP-MS, NexION 350X, PerkinElmer).Soil organic matter and pH and texture were measured by loss-on-ignition method 37 and soil paste method 38 . Soil information is in Supplementary Table 1. Soils were only gently mixed by hand to avoid violent destruction of soil aggregates before experiments. AmbControl Facility We constructed six out-door sunlit AmbControl facilities (Supplementary Fig. 21) designed to capture and manipulate natural heatwaves with ± 5% precision, matching temperature-regulated setups based on ambient temperature fluctuations. Detailed technique notes are described elsewhere 23 . Briefly, AmbControl facilities consist of six independent chambers bathed in natural sunlight, with four dedicated to experimentation and two on standby. Each chamber, measuring 1.7m in height, 1.6m in length, and 1.0m in width, is constructed from a double layer of 1.3 cm thick transparent glass, offering a total growing space of 1.6 square meters. They are mounted on wheels for easy position randomization. Equipped with an autonomous computer-controlled heating and cooling system, the AmbControl chambers fine-tune the interior air temperature in response to real-time ambient conditions. Cold and hot air are delivered above the crop canopy to achieve leaf vibration. A suite of flexibly positioned sensors from BGT Technology (China) tracks canopy temperature, relative humidity, and photosynthetic photon flux density, with manual adjustments for plant growth stages. Central system control and data logging are managed through SCADA software with a user-friendly interface. The AmbControl facility features an integrated Web API for remote regulation and data retrieval. It ensures that plants receive a realistic photosynthetic light profile, with PPFDs ranging from 300 to 2000 µmol m − 2 s − 1 across daylight hours of the crop season. Temperature, humidity, and light data are auto-logged every minute, round the clock. To maintain optimal light entry, the top glass panels are cleaned biweekly, and any humidity disparity between the internal and external AmbControl environments is mitigated through consistent ventilation. Temperature Manipulations Based on the experimental design (Fig. 1 ), we conducted temperature manipulations in 2021 and 2022. Detailed temperature manipulation methods are described in the facility test notes 23 . Briefly, the 2021 experiments involved four climate scenarios: 1. Elevated heatwave scenario by adding 1.5°C to ambient conditions; 2. Ambient heatwave scenario reflecting natural temperature fluctuations; 3. Mild climate scenario consistent to ambient temperatures while excluding heat events above 30°C; and 4. Cool climate scenario by deducting 4°C from ambient temperatures meanwhile excluding heat events above 30°C. In 2022, the experimental design was refined. The ambient and 1.5°C scenarios persisted, but the non-stress condition was revised to a ceiling below 33°C. We also introduced a fixed-temperature protocol at 28.9°C, informed by the prior average air temperatures from June and August in 2021. For both years, to control for sidewall heating effects, the pots were bathed in river sands within larger insulated containers, protecting them from external temperature interference and enabling straightforward position randomizations. Experiments We conducted a two-year experiment in 2021 and 2022 to explore the response of main crop rice and ratoon rice system to different temperature scenarios inside AmbControl facilities (Supplementary Fig. 22). Low-As XC soil (6.7 mg kg − 1 ) and moderate-As YCH soil (13 mg kg − 1 ) was used in 2021 and 2022, respectively. The soil tAs was lower and comparable to the nation average of 12.6 mg kg − 1 in Chinse paddy soils 29 . Four and five replications for each temperature manipulation were used in 2021 and 2022, respectively. In 2022, two unplanted pots under each climate scenario were used to investigate the impact of plants. Rice pots, made of high-density polyvinyl chloride with a diameter of 18 cm and height of 30 cm, were filled with approximately 6.3 kg of moist soil. Urea fertilizer (CON 2 H 4 ) was applied to the soil at a rate of 60 kg nitrogen ha − 1 at the beginning of season. Rice Yliangyou1 was germinated in deionized water and grown in corresponding experimental soil for two weeks. Three vigorous seedlings were transplanted into each pot, and a continuous flooding water management system was designed for all treatments, which were flooded by approximately 0–5 cm throughout the growth period. To maintain consistency with unchambered ambient scenario, a supplementary rainwater collection system (Supplementary Fig. 21) was used, and locally collected rainwater was used for irrigation. The natural precipitation in 2021 fully met the water demand of the entire experiment. However, during the critical growth period in 2022, we did not collect enough rainwater, which resulted in supplementary irrigation with deionized water until the end of the experiment. The randomization process was carried out weekly by rotating square containers and changing chamber locations. During the grain filling of main crop rice, the temperature under fixed temperature scenario was lower than that under other scenarios, while it was higher over the period of ratoon crop season. Supplementary Experiment In 2022, the same elevated-As SY soil was used under ambient climate conditions in a supplementary experiment to investigate whether reduced porewater arsenic through limited irrigation could alleviate arsenic stress and enhance yield compensation. Four pots from previous ambient and < 33°C scenarios were placed in ambient conditions with restricted irrigation to prevent the reductive As release. Crop management was consistent with that used in low-As soils, and grain analysis followed the procedures outlined earlier. Rice Harvest and Analysis The ratooning of Yliangyou1 had two distinct stages. In the first stage, ratoon tillers rapidly grow from the upper node of unfilled panicles. The second stage is the new tillers regenerated from the bottom nodes, which becomes the main yield component of ratoon rice. We harvested 2nd and 3rd ratoon grains for independent analyses in 2022. The first and second harvests only harvest the portion above the flag node. Grains, flag leaf and flag node were harvested manually and immediately freeze-dried for 24h (moisture content at about 10%). Dehusking is carried out manually. Dehusked grains were weighed to quantify yields. Morphological chalky features of grains were recorded. After the 3rd harvest, the aboveground biomass was determined after 24-h freeze-drying. Plant C/N analysis was performed by elemental analyzer (FLASH 2000 CHNS, Thermo Fisher). To extract total contents of elements in plant samples, 0.14–0.2 g whole grain samples or husks, flag leaf and nodes were digested in 5 mL concentrated HNO 3 (70% Ultrapure grade) and 2 mL H 2 O 2 (25% analytical grade) using microwave digestion (CEM Mars 9), with 60-min ramp phase to 60-min at 180°C. Diluted digestion solution was analyzed on ICP-MS. Reference Materials of Rice GSB-1 was used for method recovery. Internal standards (10 µg L − 1 ) of Rh and Ln were used to calibrate instrument signal drift. Enzymatic extraction method was developed to extract As, S, and P in grains using degassed α-amylase mix solution (10% Hexanol, 0.3 mM Ammonium Citrate; 20 mM NH 4 HCO 3 , 1:1 sample/α-Aly by weight, pH 7) under sonification without heating for 60 minutes, followed by heating at 60°C for 60 minutes, and 80°C for 30 minutes. Filtered solution was analyzed immediately by ion chromatography (Dionex ICS-1100, Thermo Scientific; AG/AS23 IonPac column, 4*250 mm, eluent gradient 20 mM NH 4 HCO 3 (pH 10) at a flow rate of 1.0 ml min – 1 ) coupled to ICP-MS in dynamic reaction cell mode (gas flow of O 2 , 1.0 mL min − 1 ). This method and similar ones used in previous studies have shown robust separation of As(III), As(V), DMA, and MMA (Supplementary Fig. 23) 39 – 41 . Arsenic, S, and P were detected simultaneously as AsO ( m / z 91), SO ( m / z 48) and PO ( m/z 47) to avoid interferences of As + with ArCl + ( m / z 75), S + with O 2+ ( m / z , 32). Retention times of the As species were verified by comparison with diluted commercial standards (GBW08666-08670, National Institute Metrology, China) of As(III), As(V), AsB, MMA, and DMA. We also monitor simultaneously AsO+ ( m / z 91), and SO + ( m / z 48) to double check the thiolation states (Supplementary Fig. 23). Mixed standards of DMA and sulfate were used for peak area quantification using external calibration curves. Porewater Sampling and Analysis In 2021, we conducted porewater sampling during the flowering-grain fill stage using homemade porewater samplers. The samples were preserved using 2% nitric acid and analyzed using ICM-MS. During the three grain fillings in 2022, porewater extraction was carried out using a pre-buried microdialysis device in the soil. The principles and extraction methods are described in detail elsewhere 40 . Porewater C/N analysis was immediately performed using a TOC/TON analyzer. A partial porewater sample was preserved using 10mM DTPA and analyzed for As species using IC-ICP-MS immediately. Data Source We aggregated six diverse datasets from an array of sources, including Google Scholar, Web of Science, China National Knowledge Infrastructure (CNKI), and government reports, to support our research. First, we screened reports on rice ratooning under extreme conditions using keywords such as “heatwave*,” “extreme rain*,” “flood*,” and “drought.” These reports included data on yield, geographic coordinates, event dates, and rice varieties. To estimate the ratoon compensation, we further extracted the yield of the same cultivars under optimal management practices from the China Rice Data Center ( https://www.ricedata.cn ). Second, we conducted a systematic review for global yield data on main and ratoon rice crops, using keywords such as "ratoon*" and “rice”. We screened out 1,454 relevant literature, resulting in a comprehensive dataset that encompasses 11,542 site-year-cultivar observations. This dataset included dates of sowing, transplanting, and harvest and geographic coordinates. When explicit location data was missing, it was retrieved either from related studies at the same site or, if only the address was available, via Google Earth Engine for high-resolution location data. Third, to discern the seasonal patterns of As accumulation in rice grains, a global dataset includes rice varieties, water managements, grain As concentrations, harvest days, and monthly air temperatures were compiled. Key words of “rice”, “arsenic”, “geno*”, and/or “gene*”, “sow*”, “planting date”, “season*”, “water manage*”, “alternate wetting and drying” were used to retrieve published studies in field trails. This selective process yielded 11 significant studies, which we used for data compilation, employing WebPlotDigitizer when necessary to convert graphical data into usable figures. Temperature data was sourced from metrological stations or, alternatively, from worldweatheronline.com. The harvest day was deduced from known or estimated days to heading, adding an additional 30 days. Final dataset outlined grain As concentrations from China, the United States, and Bangladesh, which are typical climate zones for rice production. One of the significant dataset was from Pinson et al. 42 that had two strict water management practices i.e., continuously flooding and non-flooded throughout crop seasons for 1700 rice varieties. Forth, a dataset for grain As concentrations in paired main crop and ratoon crop rice was compiled using the keywords of “ratoon*”, “rice”, and “arsenic”. This specialized dataset, extracted from five key publications, provided 74 site-variety records exclusively sourced from prominent ratoon rice areas in China, inclusive of geographical data and harvest dates. Fifth, to determine the critical levels of extreme air temperature and VPD in our study site, daily based air temperatures, RH, and VPD from 1961 to 2019 are retrieved from local meteorological station in Kunshan (31.395083932921473N, 120.99975176170581E) 43 . The 2021 and 2022 data were used from our experimental monitoring described in following sections. The 2020 data was absent. Data Normalization Our data normalization accounted for variations within each cultivar group, adjusting grain As concentrations (Fig. 4 d), onto a 0 to 1 scale for comparability by following equation: $$\:Normalized\:value\:=\frac{\:{y}_{i}\:-\:min\left({y}_{i}\right)}{max\left({y}_{i}\right)\:-\:min\left({y}_{i}\right)}\:$$ , Where this normalization adjusts individual yield values based on each cultivar's grain tAs concentrations, setting the lowest value as 0 and the highest as 1, with all other values scaled proportionately in between. This approach harmonizes the data allowing consistent analytical assessment across different factors. Tailored Piecewise Linear Model. In our methods of plotting the relationships between grain weights and temperatures (Supplementary Fig. 11), and between grain tAs concentration and grain weight (Supplementary Fig. 14), we constructed a tailored piecewise linear model using the \(\:lmfit\) Python library to dissect the data into distinctive linear segments at optimal breakpoints. We initialized parameters with an initial guess by visual inspection and set bounds to refine the search space. Utilizing the Levenberg-Marquardt minimization method, we conducted an optimization process, aiming for convergence with a chi-squared tolerance of 1E-14 and an extremely low x-tolerance value (1E-20), reflecting high precision. The outcome of the minimization was assessed by examining the convergence message and scrutinizing the optimized parameters. The optimal value of the breakpoint was extracted and used to finalize the piecewise linear model with \(\:myPWLF.fit\_guess\) , ensuring an accurate fitting to the observed data. ML and ML Model Explanation We implemented the RF algorithm using the XGBoost library in Python. One key benefit of XGBoost is its skill in striking an optimal balance between accuracy and simplicity. This is achieved using regularized objective functions that prevent overfitting, ensuring robust and reliable predictions. For all models used, we allocated 80% of the data for training the RF model and reserved the remaining data for cross-validation. Additionally, we fine-tuned the following parameters: learning_rate, n_estimators, max_depth, subsample, colsample_bytree, gamma, alpha, and lambda. Hyperparameter optimization was conducted using Optuna, a Python package for automated hyperparameter tuning, for 50 combinations. Throughout this process, we employed the xgb.cv function for 10-fold cross-validation to prevent overfitting and evaluated the performance of each of the 50 models. The Mean Absolute Error (MAE) and Root Mean Squared Error (RMSE) on the test set, rather than the training set, served as the basis for selecting the best model among all evaluated models. To interpret the results of XGBoost models, we employed SHAP values to estimate the effect of each input features. SHAP values offer a method for explaining machine learning models based on information theory, which are widely used in machine learning and ecology 44 . SHAP values decompose a prediction into the sum of contributions from each feature plus the average model prediction. Formally, for a prediction \(\:\widehat{{y}_{i}}\) , the SHAP values \(\:{\varphi\:}_{j}\) for features 𝑗 satisfy: $$\:\widehat{{y}_{i}}={\varphi\:}_{0}+\sum\:_{j=1}^{p}{\varphi\:}_{j}$$ , where \(\:{\varphi\:}_{0}\) is the average prediction of the model across all instances (the base value). SHAP values indicate how much each feature contributes to moving the prediction from the average prediction \(\:{\varphi\:}_{0}\) to the actual prediction \(\:\widehat{{y}_{i}}\) . A positive SHAP value indicates that the feature increases the prediction, while a negative SHAP value indicates that the feature decreases the prediction. As a result, the effect of each feature on model predictions can be additively calculated. The SHAP Python package along with built-in functions from the XGBoost package was used to calculate SHAP values. Statistical Analysis Planned and unpaired comparisons were performed between ambient heatwave scenarios and other climate scenarios using Tukey's honestly significant difference (HSD) test (P ≤ 0.05) in Excels. Normal distribution of data is tested if needed with the Kolmogorov–Smirnov Test. Declarations Data availability All raw data, models, maps are available at https://github.com/xjtluhes/heatwave_semi-perennial_arsenic. Code availability All source code and models are available at https://github.com/xjtluhes/heatwave_semi-perennial_arsenic. Acknowledgments We are deeply grateful to Dr. Johannes M. H. Knops for his insightful discussions during the preparation of this manuscript. We also acknowledge the financial support from the National Science Foundation of China (No. 42477116). Special thanks to our lab members, particularly Qianrui Huangfu, and to our technicians, Xiao Zhou and Xiaoping Xie, for their valuable technical assistance and support in conducting the experiments. Ethics Declarations The authors declare no competing interests. Author contribution statements S.Z. (first author) conceived and designed the experiments, performed the experiments, analyzed the data, contributed materials and analysis tools, and wrote the manuscript under the supervision of the corresponding author, Z.C. secured the funding, conceived and designed the experiments, and contributed materials and analysis tools. J.S., L.H. W., and Y.G. Z. contributed to data analysis and manuscript writing. References Lesk C, Rowhani P, Ramankutty N (2016) Influence of extreme weather disasters on global crop production. Nature 529:84–87 Zhu X et al (2022) Manure amendment can reduce rice yield loss under extreme temperatures. Commun Earth Environ 3:147 Yao Q et al (2024) Molecular mechanisms underlying negative effects of transient heatwaves on crop fertility. Plant Communications Gerber JS et al (2024) Global spatially explicit yield gap time trends reveal regions at risk of future crop yield stagnation. Nat Food 5:125–135 Herrero M et al (2021) Articulating the effect of food systems innovation on the Sustainable Development Goals. Lancet Planet Health 5:e50–e62 Initiative M (2009) Investing in the future: a united call to action on vitamin and mineral deficiencies. Global Rep, 23 Biswas JK, Warke M, Datta R, Sarkar D (2020) Is arsenic in rice a major human health concern? Curr pollution Rep 6:37–42 Cassman KG, Grassini P (2020) A global perspective on sustainable intensification research. Nat Sustain 3:262–268 Henry A (2024) A step forward in breeding for ratooning ability in rice. Molecular Plant Yuan S, Cassman KG, Huang JL, Peng SB, Grassini P (2019) Can ratoon cropping improve resource use efficiencies and profitability of rice in central China? Field Crops Res 234:66–72 Zhang B, Zhu X, Sun X, Zhang G, Bai C (2000) Screening of early indica ratoon rice varieties and research on high-yielding supporting technologies. Jiangxi Agricultural Sci Technol, 10–12 Wang W et al (2020) Ratoon rice technology: A green and resource-efficient way for rice production. Adv Agron 159:135–167 Zhang S et al (2023) Sustained productivity and agronomic potential of perennial rice. Nat Sustain 6:28–38 Xu F et al (2021) The ratoon rice system with high yield and high efficiency in China: Progress, trend of theory and technology. Field Crops Res 272:108282 Breshears DD et al (2021) Underappreciated plant vulnerabilities to heat waves. New Phytol 231:32–39 Fleisher DH, Barnaby JY, Li S, Timlin D (2022) Response of a U.S. rice hybrid variety to high heat at Two CO2 concentrations during anthesis and grainfill. Agric For Meteorol 323:109058 Qi D et al (2024) Mowing and nitrogen management guidelines for superior rice ratoon yields. Field Crops Res 308:109302 Yang DJ, Zhang NH, Lin ZP (2008) Research on the application of high-yield cultivation techniques of floodwater receding ratoon rice. Agricultural Sci Technol Communication, 62–64 (In Chinese) Hu GS, Jiang XM, Chen FL, Ni SJ (2004) Causes of heat damage in middle-season rice and countermeasures. China Rice 10:32 (In Chinese) Wang J et al (2020) Thiolated arsenic species observed in rice paddy pore waters. Nat Geosci 13:282–287 Yao H, Conrad R (2000) Effect of temperature on reduction of iron and production of carbon dioxide and methane in anoxic wetland rice soils. Biol Fertil Soils 32:135–141 Muehe EM, Wang T, Kerl CF, Planer-Friedrich B, Fendorf S (2019) Rice production threatened by coupled stresses of climate and soil arsenic. Nat Commun 10:4985 Zhang S, Chen Z, Zhou X, Wu L (2024) -h. Manipulating nature heatwaves in chambers: bridging controlled and field conditions for accurate warming study in plant and soil systems. agriRxiv Lin F et al (2023) GF14f gene is negatively associated with yield and grain chalkiness under rice ratooning. Front Plant Sci 14:1112146 Nevame A et al (2018) Relationship between high temperature and formation of chalkiness and their effects on quality of rice. BioMed research international. 1653721 (2018) Tang Z et al (2020) Dimethylarsinic acid is the causal agent inducing rice straighthead disease. J Exp Bot 71:5631–5644 Ishfaq M et al (2022) Physiological Essence of Magnesium in Plants and Its Widespread Deficiency in the Farming System of China. Front Plant Sci 13 Hu H, Shiying Z (1987) A study on the correlation between ratoon rice and main crop traits. Jiangxi Agricultural Sci Technol, 3–5 Zhang S et al (2024) Escalating arsenic contamination throughout Chinese soils. Nat Sustain Bhattacharyya P et al (2014) Effect of elevated carbon dioxide and temperature on phosphorus uptake in tropical flooded rice (Oryza sativa L). Eur J Agron 53:28–37 Shimoyanagi R, Abo M, Shiotsu F (2021) Higher temperatures during grain filling affect grain chalkiness and rice nutrient contents. Agronomy 11:1360 Cao Y-x, Zhu J-q, Hou J (2020) Yield gap of ratoon rice and their influence factors in China Abdullah AB, Ito S, Adhana K (2006) in Proceedings for Workshop and Conference on Rice in the World at Stake, Vol. 2 28–43 Shi Y-L, Chen W-Q, Zhu Y-G (2024) Direct, Embedded, and Embodied Trade of Arsenic: 1990–2019. Environmental Science & Technology Feng A-X et al (2020) Screening of rice varieties with low accumulation of heavy metals based on multiple target elements and their absorption and transport characteristics in rice plants. J Agricultural Resour Environ 37:988–1000 Zhu XQ, Deng QC, Li J (2007) Demonstration performance and key technologies of early-season rice-ratoon rice for the super hybrid rice YLiangyou1. China Rice 4:4 Nelson DW, Sommers LE (1996) Total carbon, organic carbon, and organic matter. Methods of soil analysis: Part 3 Chemical methods 5, 961–1010 Miller RO, Kissel DE (2010) Comparison of soil pH methods on soils of North America. Soil Sci Soc Am J 74:310–316 Heitkemper DT, Vela NP, Stewart KR, Westphal CS (2001) Determination of total and speciated arsenic in rice by ion chromatography and inductively coupled plasma mass spectrometry. J Anal At Spectrom 16:299–306 Yuan Z-F et al (2021) Simultaneous measurement of aqueous redox-sensitive elements and their species across the soil-water interface. J Environ Sci 102:1–10 Nawrocka A, Durkalec M, Michalski M, Posyniak A (2022) Simple and reliable determination of total arsenic and its species in seafood by ICP-MS and HPLC-ICP-MS. Food Chem 379:132045 Pinson SRM et al (2015) Worldwide genetic diversity for mineral element concentrations in rice grain. Crop Sci 55:294–311 Qin M et al (2021) Impact of climate change on evaporation paradox in province of Jiangsu in southeastern China. PLoS ONE 16:e0247278 Berdugo M, Gaitán JJ, Delgado-Baquerizo M, Crowther TW, Dakos V (2022) Prevalence and drivers of abrupt vegetation shifts in global drylands. Proceedings of the National Academy of Sciences 119, e2123393119 Additional Declarations There is NO Competing Interest. Supplementary Files SupplementaryInformation.docx Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-5397288","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":380631617,"identity":"bfe14dd6-d9e5-4a8f-a1cd-3a43809c62f7","order_by":0,"name":"Zheng Chen","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA8UlEQVRIiWNgGAWjYDACCcYGMG3AwMD4AEjLMDCwEa+FGYgNeIjQAqWBytkkiNLCL93c+PBLhY28OXt3WsXHtj88/OxtCQw/Krbh1CI552CzscyZNMOdPWe33ZzZZsAj2XPsAGPPmds4tRjcSGyTlmw7nGBwI3fbbV6gFoMb6Q3MjG24tdiDtfwDarn/dlsxUVoMJBLbJD82gGzh3cYM0ZJ2AK8WiTtAvzAcSzPccCZ3s+SMc8YgvyQcxOcX/tntDx/+qLGRNzh+duOHD2VycsAQM3zwowK3FhBg5kEXOYBXPRAw/iCkYhSMglEwCkY2AADZK1jd0szFqQAAAABJRU5ErkJggg==","orcid":"https://orcid.org/0000-0002-1184-3682","institution":"Xi'an Jiaotong-Liverpool University","correspondingAuthor":true,"prefix":"","firstName":"Zheng","middleName":"","lastName":"Chen","suffix":""},{"id":380631620,"identity":"1814c7c0-2a0f-4aaa-95a4-a7fef54ebb49","order_by":1,"name":"Sha Zhang","email":"","orcid":"","institution":"University of Liverpool","correspondingAuthor":false,"prefix":"","firstName":"Sha","middleName":"","lastName":"Zhang","suffix":""},{"id":380631622,"identity":"7b66f10c-5659-4283-af8f-936985acf343","order_by":2,"name":"Jing Song","email":"","orcid":"","institution":"Institute of Soil Science, CAS","correspondingAuthor":false,"prefix":"","firstName":"Jing","middleName":"","lastName":"Song","suffix":""},{"id":380631624,"identity":"9b369267-80f8-414a-b71f-9066bf59d53b","order_by":3,"name":"Longhua Wu","email":"","orcid":"","institution":"Institute of Soil Science","correspondingAuthor":false,"prefix":"","firstName":"Longhua","middleName":"","lastName":"Wu","suffix":""},{"id":380631626,"identity":"916b91e1-f06b-4f44-b71f-342c370f4fcd","order_by":4,"name":"Yong-Guan Zhu","email":"","orcid":"https://orcid.org/0000-0003-3861-8482","institution":"Research Center for Eco-Environmental Sciences, Chinese Academy of Sciences","correspondingAuthor":false,"prefix":"","firstName":"Yong-Guan","middleName":"","lastName":"Zhu","suffix":""}],"badges":[],"createdAt":"2024-11-05 17:15:59","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-5397288/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-5397288/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":69820664,"identity":"4752a76e-8f81-4097-bf9d-81ff36f3077e","added_by":"auto","created_at":"2024-11-25 14:12:28","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":341206,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eExperimental design and a conceptual framework illustrating post-heatwave ratooning responses of semi-perennial rice that can enhance yield resilience and decrease As-related food safety risks.\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"image1.png","url":"https://assets-eu.researchsquare.com/files/rs-5397288/v1/a15b5a0ce43abb732f0e7c06.png"},{"id":69820661,"identity":"337eb24d-49e4-47ce-b7fe-906a10efeee6","added_by":"auto","created_at":"2024-11-25 14:12:28","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":474073,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eYield tradeoffs between main crop and ratoon crop rice linked to post-heatwave responses for yield compensation.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ea, b\u003c/strong\u003e, Annual yield (main + ratoon) in 2021 and 2022, respectively. Data are presented as mean ± \u003cem\u003esd\u003c/em\u003e (N = 5). \u003cstrong\u003ec\u003c/strong\u003e, Linear relationships between main crop yield and ratoon crop yield in 2021 (R\u003csup\u003e2 \u003c/sup\u003e= 0.39, \u003cem\u003eP \u003c/em\u003e=\u003cem\u003e \u003c/em\u003e0.003) and in 2022 (R\u003csup\u003e2\u003c/sup\u003e = 0.52, \u003cem\u003eP\u003c/em\u003e \u0026lt; 0.001). Field yield observations (N =15, R\u003csup\u003e2 \u003c/sup\u003e= 0.37, \u003cem\u003eP \u003c/em\u003e= 0.015) of the rice cultivar \u003cem\u003eYliangyou1\u003c/em\u003e used in this study are also included. \u003cstrong\u003ed\u003c/strong\u003e, Ratoon compensations for the main crop yield losses caused by floods (N = 50), heatwaves (N = 5), droughts (N = 25) and intended mowing (N = 48), respectively.\u003c/p\u003e","description":"","filename":"image2.png","url":"https://assets-eu.researchsquare.com/files/rs-5397288/v1/faf3acaa78537509aebe40c5.png"},{"id":69820663,"identity":"2f58cdd3-0e75-4aa8-a3e3-8042279a8ef2","added_by":"auto","created_at":"2024-11-25 14:12:28","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":761002,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eSeasonal variations in grain As concentrations in main and ratoon rice are controlled by grain weight, porewater chemistry, and temperature.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ea\u003c/strong\u003e. The first ratoon crop was mainly harvested from upper tillers, which rapidly regenerated after crop failure under heatwaves. The 2nd ratoon crop was harvested from basal tillers in the late season. \u003cstrong\u003eb\u003c/strong\u003e. Grain As concentrations in main and ratoon crops from 2022. Total As concentrations include inorganic As species (arsenite [As(III)] and arsenate [As(V)]) and organic As species (monomethylarsonic acid [MMA] and dimethylarsinic acid [DMA]). Error bars denote standard errors (se) (N = 2-7). \u003cstrong\u003ec\u003c/strong\u003e, Combined heatwave effect and ratoon effect on grain As concentrations. \u003cstrong\u003ed\u003c/strong\u003e, XGBOOST machine learning model explaining the variations of total grain As concentration (N = 244) using grain weight, porewater As concentration, and air temperature (Tmax). Shaded areas represent the 95% confidence interval.\u003c/p\u003e","description":"","filename":"image3.png","url":"https://assets-eu.researchsquare.com/files/rs-5397288/v1/dbe05acf6d8438fe92fbf4d3.png"},{"id":69820662,"identity":"b55a7f83-ff98-4da0-b4c4-05dd3f2a5831","added_by":"auto","created_at":"2024-11-25 14:12:28","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":265596,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eSeasonal grain total As concentrations in main, ratoon, and conventional rice crops across varied harvest times.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ea.\u003c/strong\u003e Data synthesis illustrating the distribution of sowing and harvest dates for ratoon rice (N = 11,542 paired records), alongside paired field measurements of As concentrations in grains from the main and ratoon crops (upper Whisker plots, N = 74 paired records). \u003cstrong\u003eb.\u003c/strong\u003e Seasonal patterns of air temperature (monthly mean Tmax) and normalized grain As concentrations (excluding data from Faridpur, Bangladesh). This dataset encompasses 5,104 records from major rice-producing countries (e.g., China) that practice ratoon cropping, covering diverse rice production zones (temperate, subtropical, tropical), water management systems (flooded, non-flooded, alternating wetting and drying), and rice varieties (N = 2,451). The bar plots represent data within the shaded regions, corresponding to the same harvest date ranges as shown in panel \u003cstrong\u003ea\u003c/strong\u003e.\u003c/p\u003e","description":"","filename":"image4.png","url":"https://assets-eu.researchsquare.com/files/rs-5397288/v1/3d97c8491530d8e3b4dbdccb.png"},{"id":73717291,"identity":"86c38af2-0b98-41ed-9b8e-fe85853c106d","added_by":"auto","created_at":"2025-01-14 01:13:39","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":2723268,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-5397288/v1/dd514775-4ea4-43e2-96eb-fc39b0772016.pdf"},{"id":69820666,"identity":"fad2577f-0a02-4594-9e50-466428ace1a6","added_by":"auto","created_at":"2024-11-25 14:12:28","extension":"docx","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":16047355,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryInformation.docx","url":"https://assets-eu.researchsquare.com/files/rs-5397288/v1/7bea7b9579cf1a756d970f4b.docx"}],"financialInterests":"There is \u003cb\u003eNO\u003c/b\u003e Competing Interest.","formattedTitle":"Semi-Perennial Rice Sustains Yield and Food Safety Under Heatwaves and Beyond","fulltext":[{"header":"Main","content":"\u003cp\u003eHeatwaves severely impacted Asia's rice production, threatening food security for billions of people\u003csup\u003e\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e, \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e. Short-term extreme temperatures during heatwaves can disrupt reproductive yield formation within hours, causing significant crop failures\u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e, \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u003c/sup\u003e. These conditions have stagnated yields in major rice-producing regions\u003csup\u003e\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e\u003c/sup\u003e and hinder progress toward the United Nations' Sustainable Development Goal of 'Zero Hunger', which aims to improve both yield and nutritional quality (e.g., iron (Fe), zinc (Zn), selenium (Se)) while reducing carcinogenic contaminants like As\u003csup\u003e\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e\u003c/sup\u003e. The micronutrient deficiencies, or \u0026ldquo;hidden hunger\u0026rdquo;, affects about two billion people\u003csup\u003e\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u003c/sup\u003e, and dietary As from rice poses a growing global health threat amid climate change\u003csup\u003e\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e\u003c/sup\u003e. Despite these challenges, feasible climate-smart solutions remain scarce.\u003c/p\u003e \u003cp\u003eA promising solution is transitioning from traditional annual rice cultivation to semi-perennial production, which can desynchronize key crop stages from heatwave periods and increase the multi-cropping index\u003csup\u003e\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e, \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e\u003c/sup\u003e. Semi-perennial rice, with its ratooning phenotype, regrows from old stubbles, enabling \"one sowing, multiple harvests\" per year\u003csup\u003e\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e\u003c/sup\u003e. Although typically grown annually, most rice varieties are inherently semi-perennial, as demonstrated by a screening of 2,000 varieties that produced 0.8-3 Mg ha\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e ratoon yield\u003csup\u003e\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e\u003c/sup\u003e. This system offers agronomic potential in radiation-limited regions while reducing input dependency on tillage, fertilizers, and pesticides and saving 3/5 cropping time\u003csup\u003e\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e, \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e\u003c/sup\u003e. Recent high-yield records of ratoon rice\u003csup\u003e\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e\u003c/sup\u003e and perennial rice\u003csup\u003e\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e\u003c/sup\u003e underscore the feasibility of this transition across wider agroclimatic zones. However, the overall potentials of semi-perennial rice responding to heatwaves are not well evaluated yet\u003csup\u003e\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eHeatwaves disrupt plant carbon metabolism and storage\u003csup\u003e\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e\u003c/sup\u003e, reducing yield and grain weight while altering the elemental balance and carbon-to-element ratios (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). Although abrupt extreme stress causes infertility\u003csup\u003e\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e\u003c/sup\u003e, we hypothesize carbon for seed production would be diverted to compensatory ratooning growth for a second harvest, negatively correlating main yields (Hypothesis 1)\u003csup\u003e\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e, \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e\u003c/sup\u003e. Such compensation effects have been observed with rice crop losses due to climate disasters like floods and droughts\u003csup\u003e\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e, \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e\u003c/sup\u003e Furthermore, cooler ratoon seasons could slow soil-plant biogeochemical processes\u003csup\u003e\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e, \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e\u003c/sup\u003e, reducing nonessential toxic element uptake and translocations (Hypothesis 2)\u003csup\u003e\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e\u003c/sup\u003e, potentially enhancing yield and food safety in the post-heatwave period.\u003c/p\u003e \u003cp\u003eIn this study, we comprehensively investigated the responses of ratoon cropping system to heatwaves. In 2021 and 2022, we captured and manipulated unprecedented heatwaves at the seasonal scale for various scenarios, using customized sunlit AmbControl facilities (details in discussions of ref\u003csup\u003e\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e\u003c/sup\u003e). Seventy heatwave days during the crop season (Supplementary Fig.\u0026nbsp;1) and 27 extreme days with daily Tmax exceeding 40.0\u0026deg;C were observed in 2022, marking the driest and hottest summer in East Asia monsoon region for six decades (Supplementary Fig.\u0026nbsp;2). Contrastingly, 2021 rice production was entirely under rainy season with 51 heatwave days, resulting the wettest hot summer for last two decades. We quantified yield responses, nutritional qualities, and As-related food safety risks under four scenarios, i.e. recent and elevated heatwaves compared to mild and cool climate scenarios (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). We extended the growth duration to ratoon season to assess the ratoon effect on yield and qualities. Additionally, a global dataset of rice As levels from 5,104 site-year-variety instances and 74 paired main-ratoon grain As surveys in China are compiled to illustrate the agronomic potential of ratoon cropping for healthier rice production under a changing climate.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eUsing four sunlit AmbControl facilities, which replicate ambient temperature dynamics by one-minute resolution, we created different climate scenarios to assess heatwave impacts on ratoon rice in 2021 and 2022 (annul temperature dynamics available in ref\u003csup\u003e\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e\u003c/sup\u003e). Rice was grown under various temperature scenarios, including elevated heatwaves (+\u0026thinsp;1.5\u0026deg;C), ambient heatwaves (+\u0026thinsp;0\u0026deg;C), mild climate (daily Tmax\u0026thinsp;\u0026lt;\u0026thinsp;30\u0026deg;C in 2021; daily Tmax\u0026thinsp;\u0026lt;\u0026thinsp;33\u0026deg;C in 2022), and cool climate (-4\u0026deg;C in 2021; fixed 28.9\u0026deg;C in 2022). Theoretically, heatwaves during critical stages (flowering and grain filling) reduce carbon allocation to grain, lowering yield, while excess carbohydrates promote new panicle growth for a second harvest\u003csup\u003e\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e\u003c/sup\u003e. This carbon loss in grains can increase concentrations of other elements. Heatwaves accelerate soil chemical reactions, raising nutrient and As uptake and contamination risks. Given the human toxicity of inorganic As (iAs) and the phytotoxicity of organic As (oAs), these risks are exacerbated under heatwaves but mitigated under cooler ratoon conditions.\u003c/p\u003e"},{"header":"Results","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e\n \u003ch2\u003eMain Yield loss and Ratoon Yield Compensation\u003c/h2\u003e\n \u003cp\u003eSignificant yield losses were recorded during heatwaves, but ratoon rice compensated with improved yields. Under two-year heatwave scenarios, rice variety \u003cem\u003eYliangyou1\u003c/em\u003e experienced drastic average yield reductions of 61\u0026ndash;78% (Fig. \u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003ea, b). Uncontrolled ambient and open-field observations in two year confirmed similar yield losses, ranging from 69 to 100% (Supplementary Fig.\u0026nbsp;3). Interestingly, despite comparable mean temperatures between heatwave and cool climate conditions in 2022 (28.2\u0026deg;C vs. 28.7\u0026deg;C), yields under heatwaves was 88.9% lower. Heatwaves also caused significant grain chalkiness, degrading the commercial value of rice (Supplementary Fig.\u0026nbsp;4), while minimal chalkiness (\u0026lt;\u0026thinsp;3%) was observed under mild and cool conditions. Chalkiness is the typical phenotype which can be induced by high temperatures\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e25\u003c/span\u003e\u003c/sup\u003e. This indicated that extreme temperatures have a more pronounced effect on yields and grain quality than seasonal averages.\u003c/p\u003e\n \u003cp\u003eThe ratooning phenotype of semi-perennial rice demonstrated strong resilience to heatwaves, as it compensated summer yield losses with improved growth during cooler periods (Fig. \u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003ec, Hypothesis 1). This yield compensation was linearly quantified: declines in main crop yields during heatwaves led to proportional increases in ratoon yields. In 2022, ratoon rice maintained an average of 11 grams per plant annually across diverse climate conditions (P\u0026thinsp;\u0026gt;\u0026thinsp;0.05). This yield consistency underscores the inherent resilience of rice to extreme climate events.\u003c/p\u003e\n \u003cp\u003eA broader dataset for the rice cultivar \u003cem\u003eYliangyou1\u003c/em\u003e (N\u0026thinsp;=\u0026thinsp;37 site-year surveys across China, Supplementary Fig. 5) revealed a clear tradeoff pattern (P\u0026thinsp;=\u0026thinsp;0.015, Fig. \u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003ec) between main and ratoon crop yields across high-yield sites (N\u0026thinsp;=\u0026thinsp;15 out of 37). The main-ratoon yield tradeoff pattern were also reported for other site-variety observations\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e14\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e\n \u003cp\u003eTo further assess yield compensation, we examined literature on rice crops experiencing significant main crop yield failures (\u0026gt;\u0026thinsp;80%) due to heatwaves, floods, droughts, or intentional mowing for forage purposes across China (Fig. \u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003ed, Supplementary Fig. 6). Ratoon cropping demonstrated substantial resilience to these disturbances, with yield compensations of 51.9\u0026thinsp;\u0026plusmn;\u0026thinsp;11.7% (\u0026plusmn;\u0026thinsp;1 standard deviation, \u003cem\u003esd\u003c/em\u003e) after heatwaves, 53.7\u0026thinsp;\u0026plusmn;\u0026thinsp;16.9% after floods, 69.2\u0026thinsp;\u0026plusmn;\u0026thinsp;13.9% after droughts, and 71.9\u0026thinsp;\u0026plusmn;\u0026thinsp;29.8% after mowing compared to historical yield records (Fig. \u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003ed).\u003c/p\u003e\n \u003cp\u003eHowever, ratoon yield compensation was inhibited by soil As contamination. In a supplementary experiment conducted in 2022, using an As-contaminated soil (47 mg kg⁻\u0026sup1; total As, tAs), heatwaves resulted in near-total crop failure (Supplementary Fig.\u0026nbsp;7r). Even under mild and cool climates, both the main and ratoon crops suffered yield reductions due to serious straighthead disease. The dimethylarsinic (DMA) is a common cause to straighthead disease\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e26\u003c/span\u003e\u003c/sup\u003e. We verified that more than 83% of grain As was DMA (tAs range: 2,200\u0026ndash;3,100 \u0026micro;g kg\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e) (Supplementary Fig.\u0026nbsp;7p,q). Efforts to mitigate grain As uptake through irrigation management in 2022 were moderately successful (tAs\u0026thinsp;\u0026lt;\u0026thinsp;150 \u0026micro;g kg\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e, Supplementary Fig. 8). However, main crop yields were still nonexistent under heatwaves, and ratoon yields were also minimal. These findings suggest that while ratoon rice exhibits resilience to extreme climatic conditions, its compensatory capacity is severely constrained by soil As contamination.\u003c/p\u003e\n\u003c/div\u003e\n\u003ch3\u003eHeatwave Effect and Ratoon Effect on Grain As\u003c/h3\u003e\n\u003cp\u003eHeatwaves increased the concentration of toxic elements, particularly As, in rice, while ratoon rice exhibited lower uptake of these elements. Our unique heatwave control facility enabled the creation of mild baseline climate scenarios, allowing us to quantify the effects of heatwaves and ratooning directly. Over two years, heatwaves increased total As (tAs) concentrations in grain by 28% (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.001), while ratooning practices reduced them by 37% (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.001, Fig. \u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003ec).\u003c/p\u003e\n\u003cp\u003eWe focused on As because, under continuously flooded rice cultivation, grain cadmium (Cd) (0.031\u0026thinsp;\u0026plusmn;\u0026thinsp;0.026 mg kg⁻\u0026sup1;) and lead (Pb) concentrations (0.047\u0026thinsp;\u0026plusmn;\u0026thinsp;0.033 mg kg⁻\u0026sup1;) across all treatments in both years were far below the United Nations\u0026rsquo; food safety standards (0.2 mg kg⁻\u0026sup1;) (Supplementary Figs.\u0026nbsp;9,10). In contrast, grain As concentrations (0.52\u0026thinsp;\u0026plusmn;\u0026thinsp;0.18 mg kg⁻\u0026sup1;) frequently exceeded food safety standards (tAs: 0.3 mg kg⁻\u0026sup1;; inorganic As [iAs]: 0.2 mg kg⁻\u0026sup1;).\u003c/p\u003e\n\u003cp\u003eWe further analyzed As speciation, differentiating between iAs and organic As (oAs). In 2022, across all treatments and seasons, oAs accounted for 28.5\u0026thinsp;\u0026plusmn;\u0026thinsp;7.7% of tAs in grains (Fig. \u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003eb). However, in 2021, the oAs percentage was significantly higher, reaching 60.2\u0026thinsp;\u0026plusmn;\u0026thinsp;12.4% (Supplementary Figs.\u0026nbsp;9,10). Notably, in 2021, despite daily Tmax being 5\u0026ndash;10\u0026deg;C higher during the grain filling stage (data in ref\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e23\u003c/span\u003e\u003c/sup\u003e), heatwaves did not increase grain iAs concentrations but increase oAs concentrations significantly (Supplementary Figs.\u0026nbsp;7,9).\u003c/p\u003e\n\u003ch3\u003eMechanisms\u003c/h3\u003e\n\u003cp\u003eHeatwave-induced carbon loss and grain weight reduction are common drivers of changes in grain element concentrations. A piecewise linear model identified a turning point at a Tmax of 31.4\u0026deg;C or a Tmean of 24.3\u0026deg;C (Supplementary Fig.\u0026nbsp;11). Below this temperature, moderate heat increased grain weight (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.08), whereas excessive heat reduced it. Over two years, heatwaves decreased grain carbon content by 13% and caused a 30% weight loss, contributing to an \u0026quot;enrichment effect\u0026quot;, where concentrations of other elements increased (Supplementary Figs.\u0026nbsp;12, 13). This uniquely resulted in negative correlations between grain weight and concentrations of trace elements, not only As but also Fe, Zn, Se, boron (B), molybdenum (Mo), lead (Pb), nickel (Ni), antimony (Sb), cobalt (Co), cadmium (Cd), and copper (Cu) (Supplementary Figs.\u0026nbsp;12,13). However, ratoon grain weight has no clear relationships with As concentrations (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.83 based on linear regression, Supplementary Fig.\u0026nbsp;14). This suggested that compared to baseline climates, only heat-stressed grain had increased As and rice nutrients.\u003c/p\u003e\n\u003cp\u003eFurther findings support Hypothesis 2: cooler temperatures reduce soil As availability, leading to lower grain As concentrations in ratoon crops. We analyzed harvests based on ratoon cropping calendars, revealing that early ratoon yields, which developed during high-temperature conditions, contained higher As concentrations, while second ratoon crops, grown entirely in cooler periods, had lower As levels. Time-series data of grain As concentrations (Fig. \u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003eb) and dissolved As in soils, with and without plants (Supplementary Fig.\u0026nbsp;15), demonstrated a decline in As with decreasing temperatures. Across elevated heatwave, ambient heatwave, and mild climate scenarios, iAs levels in ratoon grains often met or fell below the international food standard of 200 \u0026micro;g kg⁻\u0026sup1;, highlighting ratoon cropping\u0026apos;s potential to mitigate As accumulation. Interestingly, despite the highest average temperature (28.9\u0026deg;C vs. ambient Tmean\u0026thinsp;\u0026lt;\u0026thinsp;25.5\u0026deg;C), the fixed cool climate scenario during the second ratoon period resulted in higher grain As concentrations, suggesting that high temperatures are critical for maintaining elevated grain As concentrations.\u003c/p\u003e\n\u003cp\u003eA third mechanism contributing to lower As levels in ratoon crops is reduced in-plant translocation of As during cooler ratoon seasons. Stepwise analyses of nodes, flag leaves, husks, and grains (Supplementary Figs.\u0026nbsp;7,9,10) confirmed this mechanism, which applies to most elements except magnesium (Mg) and Fe. Magnesium, essential for photosynthesis and carbon allocation, is highly mobile within plants\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e27\u003c/span\u003e\u003c/sup\u003e. Ratoon grains accumulated 64% higher Mg compared to main crops (Supplementary Figs.\u0026nbsp;14, 16,17), likely due to in-plant reallocations, as no additional Mg fertilizer was applied beyond nitrogen at the start of the main crop season.\u003c/p\u003e\n\u003ch3\u003eResponses Trajectory of Grain As to Key Drivers\u003c/h3\u003e\n\u003cp\u003eWe further assessed the effects of grain weight, soil porewater As concentrations, and mean daily Tmax during the corresponding crop seasons on grain As concentrations (Fig. \u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003ed). Using Shapley Additive Explanations (SHAP) values from the TreeExplainer of a trained XGBoost model, we quantified the individual contributions of each feature relative to the baseline prediction. The XGBoost model explained 86.3% of the variation in grain As concentrations. Grain weight accounted for 21.2% of the variation, soil porewater As concentration for 34%, and temperature (Tmax) for 31.1% (Fig. \u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003ed). While the temperature effect on grain As followed a non-threshold response, the apparent sensitivity of grain As concentrations to temperature was lower than expected, increasing by 21 \u0026micro;g kg\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e per 1\u0026deg;C rise.\u003c/p\u003e\n\u003ch3\u003eBroader Insights from Global Datasets\u003c/h3\u003e\n\u003cp\u003eGrain As concentrations were broadly influenced by seasonal temperature fluctuations. To analyze this trend further, we utilized two external datasets: the first contained paired As concentrations in main and ratoon rice grains from China (N\u0026thinsp;=\u0026thinsp;74 paired data points, Fig. \u003cspan class=\"InternalRef\"\u003e4\u003c/span\u003ea), while the second encompassed global grain As concentrations (N\u0026thinsp;=\u0026thinsp;5104) from various harvest dates, spanning major rice-producing countries, different agronomic zones, water management practices, and 2,451 rice varieties (Fig. \u003cspan class=\"InternalRef\"\u003e4\u003c/span\u003eb, Supplementary Fig.\u0026nbsp;18).\u003c/p\u003e\n\u003cp\u003eThe China dataset highlighted a consistent discrepancy between main and ratoon crop As concentrations, with ratoon rice exhibiting a 49.2\u0026thinsp;\u0026plusmn;\u0026thinsp;1.8% reduction in grain As (Fig. \u003cspan class=\"InternalRef\"\u003e4\u003c/span\u003ea). This reduction aligns closely with our experimental results from heatwave manipulations (Fig. \u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003ec). In the global dataset, a pronounced seasonal temperature dependence on As uptake was observed, with late-harvested grains showing a 21.9\u0026thinsp;\u0026plusmn;\u0026thinsp;3% reduction in As under conventional cultivation (Fig. \u003cspan class=\"InternalRef\"\u003e4\u003c/span\u003eb). This pattern held true when comparing grains from the same cultivars planted at different times, with those harvested in cooler seasons displaying an approximate 45.0% decrease in As concentrations.\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eOur climate manipulation experiments have demonstrated the severe damage heatwaves inflict on rice systems but also revealed the significant potential of semi-perennial rice to provide additional yields and safer food. These findings are especially relevant for rice-producing regions, where global warming is increasing the frequency of heat extremes, posing serious risks to food security and exacerbating soil As contamination.\u003c/p\u003e \u003cp\u003eOne critical discovery from comparing heatwave and non-heatwave scenarios is the rapid yield compensation following reproductive failure (Supplementary Fig.\u0026nbsp;19). While this compensation did not fully recover the original yield potential, it significantly approached it. Unlike typical ratoon practices, where tillers emerge after the main crop matures or is harvested (as in cooler climates, Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ea), under heatwave conditions, ratoon tillers appeared concurrently with reproductive failure, redirecting surplus carbohydrates toward ratoon yield formation.\u003c/p\u003e \u003cp\u003eThe molecular mechanisms of ratoon compensation effect remain unclear. However, the response resembles that observed in rice subjected to artificial mowing or topping, where apical dominance is broken, promoting side tiller growth by redistributing plant resources (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ed)\u003csup\u003e\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e\u003c/sup\u003e. Heatwaves appear to trigger a similar redistribution, enabling ratoon growth to partially compensate for main crop yield losses. However, deciphering the full metabolic network is challenging, as ratoon plants systematically differ from main crop rice. For instance, in trials with 429 global varieties, ratoon plants exhibited 50% reductions in plant height, panicle length, leaf size, and grain numbers compared to main crops\u003csup\u003e\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e\u003c/sup\u003e. Even though, some varieties have shown robust ratooning ability, with ratoon yields comparable to the main rice yield\u003csup\u003e\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e\u003c/sup\u003e. This suggests that ratooning under heatwaves or other stressors may represent an underappreciated form of resilience in rice systems.\u003c/p\u003e \u003cp\u003eDespite the agronomic potential of ratoon rice, soil As contamination can undermine this promise. Yield reductions in elevated-As soils are linked to changes in As speciation, particularly microbial methylation, which can severely damage rice yields\u003csup\u003e\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e\u003c/sup\u003e. In elevated-As soils, grain oAs was found to be the predominant species (\u0026gt;\u0026thinsp;80%), which is unusual for Chinese rice (usually\u0026thinsp;~\u0026thinsp;29%)\u003csup\u003e29\u003c/sup\u003e. This highlights the need for controlling soil As methylation, particularly as root exudates can promote methylation (Supplementary Fig.\u0026nbsp;20), leading to higher As uptake and potential phytotoxicity\u003csup\u003e\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e\u003c/sup\u003e. Although the exact impact of oAs on yield reductions remains uncertain due to limited data, its influence on food security should not be underestimated.\u003c/p\u003e \u003cp\u003eOur findings provide valuable insights into the factors driving As accumulation in rice grains. First, grain weight loss and enhanced in-plant translocations during heatwaves are the primary drivers of increased elemental concentrations, aligning with experimental warming studies. Grain weight loss, however, has rarely been considered in explaining changes in the ionome\u003csup\u003e\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e, \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e\u003c/sup\u003e. Second, while elevated temperatures increase tAs in grains, this does not always correspond to higher levels of iAs, the primary food safety concern. This stands in contrast to previous reports that suggest each degree of temperature increase doubles the food safety risk from iAs\u003csup\u003e\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e\u003c/sup\u003e. Our data show that higher temperatures do not consistently enhance As mobility in the soil-plant system. Third, we observed that grain tAs and porewater tAs closely track temperature changes across our experiments, regional studies, and global datasets (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e). The reduced As in ratoon rice is largely attributed to lower soil As availability and reduced in-plant translocation under cooler conditions. This seasonal variability provides a concise framework for assessing food safety risks more accurately.\u003c/p\u003e \u003cp\u003eAlthough ratoon rice tends to accumulate lower levels of As, it also shows reduced accumulation of essential micronutrients such as Fe, Zn, and Se (Extended Figs.\u0026nbsp;7, 16,17). However, in 2022, the reduction in grain Fe levels was not statistically significant, highlighting the role of physiological regulation in controlling the accumulation of certain essential elements. For other beneficial trace elements, agronomic biofortification may offer a potential solution.\u003c/p\u003e \u003cp\u003eRatoon rice holds substantial promise for addressing food security challenges, particularly in regions vulnerable to climate extremes. There are currently 1.1\u0026nbsp;million ha for ratoon rice in China. Based on nation-wide yield average of 3.64 Mg ha\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e in China\u003csup\u003e\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e\u003c/sup\u003e, this cultivation method could yield an additional 3.9\u0026nbsp;million Mg currently, potentially increasing to as much as 12\u0026nbsp;million Mg with using radiation-marginal rice production areas. Assuming the average annual rice consumption per person is approximately 150 kg\u003csup\u003e\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e\u003c/sup\u003e. This can feed approximately 80\u0026nbsp;million people. Considering current and potential ratoon yields in China, ratooning could prevent about 535 Kg and potentially 1,700 kg tAs per year, respectively. This also exceed the rice As export (323 kg iAs per year) from United States' rice industry\u003csup\u003e\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eWhile our heatwave-manipulative experiments have provided critical insights, field studies under diverse real-world conditions are needed to validate ratoon rice's performance across broader agricultural systems. Further investigation into the molecular pathways governing heat-induced ratooning will be essential to fully unlock its potential, not just in rice but potentially in other crops.\u003c/p\u003e \u003cp\u003eOur study demonstrates that ratoon rice can effectively compensate for yield losses caused by heatwaves while reducing food contamination risks from As. These findings suggest that ratoon rice cultivation could play a key role in enhancing food security and promoting the sustainability of agricultural systems. Policymakers and agricultural stakeholders should prioritize ratoon rice as a strategy for adapting rice production to the increasing threats posed by climate extremes.\u003c/p\u003e"},{"header":"Methods","content":"\u003cdiv id=\"Sec10\" class=\"Section2\"\u003e \u003ch2\u003eHeatwave Definition\u003c/h2\u003e \u003cp\u003eSummer heatwaves are defined as periods with three consecutive days where the daily Tmax exceeds 36.2\u0026deg;C for at least one hour each day. This threshold is derived from the 30-year Tmax average (1961\u0026ndash;1990) plus an additional 5\u0026deg;C at our study site.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003ch2\u003eRice and Soil\u003c/h2\u003e \u003cp\u003eWe used a popular rice variety Ylaingyou1 (hybrid), which is moderately sensitive to temperature and for As bioaccumulation\u003csup\u003e\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e\u003c/sup\u003e. It has a cumulative harvested area of 2.7\u0026nbsp;million ha in China (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.natesc.org.cn\u003c/span\u003e\u003cspan address=\"https://www.natesc.org.cn\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e). This variety has moderate ratooning ability\u003csup\u003e\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e\u003c/sup\u003e. Two local low-As paddy soils i.e., soil XC (31\u0026deg;29\u0026prime;49.095\"N,120\u0026deg;32\u0026prime;49.927\"E) and soil YCH (31\u0026deg;28\u0026prime;25.52\"N,120\u0026deg;46\u0026prime;18.88\"E) and an As-elevated paddy soil SY (30\u0026deg;0' 8.03\"N, 120\u0026deg;47'58.14\"E) were collected from Tai Lake area near our experimental site. Soil As were extracted by aqua regia and determined by Inductively coupled plasma mass spectrometry (ICP-MS, NexION 350X, PerkinElmer).Soil organic matter and pH and texture were measured by loss-on-ignition method\u003csup\u003e\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e\u003c/sup\u003e and soil paste method\u003csup\u003e\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e\u003c/sup\u003e. Soil information is in Supplementary Table\u0026nbsp;1. Soils were only gently mixed by hand to avoid violent destruction of soil aggregates before experiments.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec12\" class=\"Section2\"\u003e \u003ch2\u003eAmbControl Facility\u003c/h2\u003e \u003cp\u003eWe constructed six out-door sunlit AmbControl facilities (Supplementary Fig.\u0026nbsp;21) designed to capture and manipulate natural heatwaves with \u0026plusmn;\u0026thinsp;5% precision, matching temperature-regulated setups based on ambient temperature fluctuations. Detailed technique notes are described elsewhere\u003csup\u003e\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e\u003c/sup\u003e. Briefly, AmbControl facilities consist of six independent chambers bathed in natural sunlight, with four dedicated to experimentation and two on standby. Each chamber, measuring 1.7m in height, 1.6m in length, and 1.0m in width, is constructed from a double layer of 1.3 cm thick transparent glass, offering a total growing space of 1.6 square meters. They are mounted on wheels for easy position randomization.\u003c/p\u003e \u003cp\u003eEquipped with an autonomous computer-controlled heating and cooling system, the AmbControl chambers fine-tune the interior air temperature in response to real-time ambient conditions. Cold and hot air are delivered above the crop canopy to achieve leaf vibration. A suite of flexibly positioned sensors from BGT Technology (China) tracks canopy temperature, relative humidity, and photosynthetic photon flux density, with manual adjustments for plant growth stages. Central system control and data logging are managed through SCADA software with a user-friendly interface.\u003c/p\u003e \u003cp\u003eThe AmbControl facility features an integrated Web API for remote regulation and data retrieval. It ensures that plants receive a realistic photosynthetic light profile, with PPFDs ranging from 300 to 2000 \u0026micro;mol m\u003csup\u003e\u0026minus;\u0026thinsp;2\u003c/sup\u003e s\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e across daylight hours of the crop season. Temperature, humidity, and light data are auto-logged every minute, round the clock. To maintain optimal light entry, the top glass panels are cleaned biweekly, and any humidity disparity between the internal and external AmbControl environments is mitigated through consistent ventilation.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec13\" class=\"Section2\"\u003e \u003ch2\u003eTemperature Manipulations\u003c/h2\u003e \u003cp\u003eBased on the experimental design (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e), we conducted temperature manipulations in 2021 and 2022. Detailed temperature manipulation methods are described in the facility test notes\u003csup\u003e\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e\u003c/sup\u003e. Briefly, the 2021 experiments involved four climate scenarios: 1. Elevated heatwave scenario by adding 1.5\u0026deg;C to ambient conditions; 2. Ambient heatwave scenario reflecting natural temperature fluctuations; 3. Mild climate scenario consistent to ambient temperatures while excluding heat events above 30\u0026deg;C; and 4. Cool climate scenario by deducting 4\u0026deg;C from ambient temperatures meanwhile excluding heat events above 30\u0026deg;C. In 2022, the experimental design was refined. The ambient and 1.5\u0026deg;C scenarios persisted, but the non-stress condition was revised to a ceiling below 33\u0026deg;C. We also introduced a fixed-temperature protocol at 28.9\u0026deg;C, informed by the prior average air temperatures from June and August in 2021. For both years, to control for sidewall heating effects, the pots were bathed in river sands within larger insulated containers, protecting them from external temperature interference and enabling straightforward position randomizations.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec14\" class=\"Section2\"\u003e \u003ch2\u003eExperiments\u003c/h2\u003e \u003cp\u003eWe conducted a two-year experiment in 2021 and 2022 to explore the response of main crop rice and ratoon rice system to different temperature scenarios inside AmbControl facilities (Supplementary Fig.\u0026nbsp;22). Low-As XC soil (6.7 mg kg\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e) and moderate-As YCH soil (13 mg kg\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e) was used in 2021 and 2022, respectively. The soil tAs was lower and comparable to the nation average of 12.6 mg kg\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e in Chinse paddy soils\u003csup\u003e\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eFour and five replications for each temperature manipulation were used in 2021 and 2022, respectively. In 2022, two unplanted pots under each climate scenario were used to investigate the impact of plants. Rice pots, made of high-density polyvinyl chloride with a diameter of 18 cm and height of 30 cm, were filled with approximately 6.3 kg of moist soil. Urea fertilizer (CON\u003csub\u003e2\u003c/sub\u003eH\u003csub\u003e4\u003c/sub\u003e) was applied to the soil at a rate of 60 kg nitrogen ha\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e at the beginning of season. Rice \u003cem\u003eYliangyou1\u003c/em\u003e was germinated in deionized water and grown in corresponding experimental soil for two weeks. Three vigorous seedlings were transplanted into each pot, and a continuous flooding water management system was designed for all treatments, which were flooded by approximately 0\u0026ndash;5 cm throughout the growth period. To maintain consistency with unchambered ambient scenario, a supplementary rainwater collection system (Supplementary Fig.\u0026nbsp;21) was used, and locally collected rainwater was used for irrigation. The natural precipitation in 2021 fully met the water demand of the entire experiment. However, during the critical growth period in 2022, we did not collect enough rainwater, which resulted in supplementary irrigation with deionized water until the end of the experiment. The randomization process was carried out weekly by rotating square containers and changing chamber locations. During the grain filling of main crop rice, the temperature under fixed temperature scenario was lower than that under other scenarios, while it was higher over the period of ratoon crop season.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec15\" class=\"Section2\"\u003e \u003ch2\u003eSupplementary Experiment\u003c/h2\u003e \u003cp\u003eIn 2022, the same elevated-As SY soil was used under ambient climate conditions in a supplementary experiment to investigate whether reduced porewater arsenic through limited irrigation could alleviate arsenic stress and enhance yield compensation. Four pots from previous ambient and \u0026lt;\u0026thinsp;33\u0026deg;C scenarios were placed in ambient conditions with restricted irrigation to prevent the reductive As release. Crop management was consistent with that used in low-As soils, and grain analysis followed the procedures outlined earlier.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec16\" class=\"Section2\"\u003e \u003ch2\u003eRice Harvest and Analysis\u003c/h2\u003e \u003cp\u003eThe ratooning of \u003cem\u003eYliangyou1\u003c/em\u003e had two distinct stages. In the first stage, ratoon tillers rapidly grow from the upper node of unfilled panicles. The second stage is the new tillers regenerated from the bottom nodes, which becomes the main yield component of ratoon rice. We harvested 2nd and 3rd ratoon grains for independent analyses in 2022. The first and second harvests only harvest the portion above the flag node. Grains, flag leaf and flag node were harvested manually and immediately freeze-dried for 24h (moisture content at about 10%). Dehusking is carried out manually. Dehusked grains were weighed to quantify yields. Morphological chalky features of grains were recorded. After the 3rd harvest, the aboveground biomass was determined after 24-h freeze-drying.\u003c/p\u003e \u003cp\u003ePlant C/N analysis was performed by elemental analyzer (FLASH 2000 CHNS, Thermo Fisher). To extract total contents of elements in plant samples, 0.14\u0026ndash;0.2 g whole grain samples or husks, flag leaf and nodes were digested in 5 mL concentrated HNO\u003csub\u003e3\u003c/sub\u003e (70% Ultrapure grade) and 2 mL H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2\u003c/sub\u003e (25% analytical grade) using microwave digestion (CEM Mars 9), with 60-min ramp phase to 60-min at 180\u0026deg;C. Diluted digestion solution was analyzed on ICP-MS. Reference Materials of Rice GSB-1 was used for method recovery. Internal standards (10 \u0026micro;g L\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e) of Rh and Ln were used to calibrate instrument signal drift. Enzymatic extraction method was developed to extract As, S, and P in grains using degassed α-amylase mix solution (10% Hexanol, 0.3 mM Ammonium Citrate; 20 mM NH\u003csub\u003e4\u003c/sub\u003eHCO\u003csub\u003e3\u003c/sub\u003e, 1:1 sample/α-Aly by weight, pH 7) under sonification without heating for 60 minutes, followed by heating at 60\u0026deg;C for 60 minutes, and 80\u0026deg;C for 30 minutes. Filtered solution was analyzed immediately by ion chromatography (Dionex ICS-1100, Thermo Scientific; AG/AS23 IonPac column, 4*250 mm, eluent gradient 20 mM NH\u003csub\u003e4\u003c/sub\u003eHCO\u003csub\u003e3\u003c/sub\u003e (pH 10) at a flow rate of 1.0 ml min\u003csup\u003e\u0026ndash;\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u003c/sup\u003e) coupled to ICP-MS in dynamic reaction cell mode (gas flow of O\u003csub\u003e2\u003c/sub\u003e, 1.0 mL min\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e). This method and similar ones used in previous studies have shown robust separation of As(III), As(V), DMA, and MMA (Supplementary Fig.\u0026nbsp;23)\u003csup\u003e\u003cspan additionalcitationids=\"CR40\" citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e\u003c/sup\u003e. Arsenic, S, and P were detected simultaneously as AsO (\u003cem\u003em\u003c/em\u003e/\u003cem\u003ez\u003c/em\u003e 91), SO (\u003cem\u003em\u003c/em\u003e/\u003cem\u003ez\u003c/em\u003e 48) and PO (\u003cem\u003em/z\u003c/em\u003e 47) to avoid interferences of As\u003csup\u003e+\u003c/sup\u003e with ArCl\u003csup\u003e+\u003c/sup\u003e (\u003cem\u003em\u003c/em\u003e/\u003cem\u003ez\u003c/em\u003e 75), S\u003csup\u003e+\u003c/sup\u003e with O\u003csup\u003e2+\u003c/sup\u003e (\u003cem\u003em\u003c/em\u003e/\u003cem\u003ez\u003c/em\u003e, 32). Retention times of the As species were verified by comparison with diluted commercial standards (GBW08666-08670, National Institute Metrology, China) of As(III), As(V), AsB, MMA, and DMA. We also monitor simultaneously AsO+ (\u003cem\u003em\u003c/em\u003e/\u003cem\u003ez\u003c/em\u003e 91), and SO\u003csup\u003e+\u003c/sup\u003e (\u003cem\u003em\u003c/em\u003e/\u003cem\u003ez\u003c/em\u003e 48) to double check the thiolation states (Supplementary Fig.\u0026nbsp;23). Mixed standards of DMA and sulfate were used for peak area quantification using external calibration curves.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec17\" class=\"Section2\"\u003e \u003ch2\u003ePorewater Sampling and Analysis\u003c/h2\u003e \u003cp\u003eIn 2021, we conducted porewater sampling during the flowering-grain fill stage using homemade porewater samplers. The samples were preserved using 2% nitric acid and analyzed using ICM-MS. During the three grain fillings in 2022, porewater extraction was carried out using a pre-buried microdialysis device in the soil. The principles and extraction methods are described in detail elsewhere \u003csup\u003e\u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e\u003c/sup\u003e. Porewater C/N analysis was immediately performed using a TOC/TON analyzer. A partial porewater sample was preserved using 10mM DTPA and analyzed for As species using IC-ICP-MS immediately.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec18\" class=\"Section2\"\u003e \u003ch2\u003eData Source\u003c/h2\u003e \u003cp\u003eWe aggregated six diverse datasets from an array of sources, including Google Scholar, Web of Science, China National Knowledge Infrastructure (CNKI), and government reports, to support our research.\u003c/p\u003e \u003cp\u003eFirst, we screened reports on rice ratooning under extreme conditions using keywords such as \u0026ldquo;heatwave*,\u0026rdquo; \u0026ldquo;extreme rain*,\u0026rdquo; \u0026ldquo;flood*,\u0026rdquo; and \u0026ldquo;drought.\u0026rdquo; These reports included data on yield, geographic coordinates, event dates, and rice varieties. To estimate the ratoon compensation, we further extracted the yield of the same cultivars under optimal management practices from the China Rice Data Center (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.ricedata.cn\u003c/span\u003e\u003cspan address=\"https://www.ricedata.cn\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eSecond, we conducted a systematic review for global yield data on main and ratoon rice crops, using keywords such as \"ratoon*\" and \u0026ldquo;rice\u0026rdquo;. We screened out 1,454 relevant literature, resulting in a comprehensive dataset that encompasses 11,542 site-year-cultivar observations. This dataset included dates of sowing, transplanting, and harvest and geographic coordinates. When explicit location data was missing, it was retrieved either from related studies at the same site or, if only the address was available, via Google Earth Engine for high-resolution location data.\u003c/p\u003e \u003cp\u003eThird, to discern the seasonal patterns of As accumulation in rice grains, a global dataset includes rice varieties, water managements, grain As concentrations, harvest days, and monthly air temperatures were compiled. Key words of \u0026ldquo;rice\u0026rdquo;, \u0026ldquo;arsenic\u0026rdquo;, \u0026ldquo;geno*\u0026rdquo;, and/or \u0026ldquo;gene*\u0026rdquo;, \u0026ldquo;sow*\u0026rdquo;, \u0026ldquo;planting date\u0026rdquo;, \u0026ldquo;season*\u0026rdquo;, \u0026ldquo;water manage*\u0026rdquo;, \u0026ldquo;alternate wetting and drying\u0026rdquo; were used to retrieve published studies in field trails. This selective process yielded 11 significant studies, which we used for data compilation, employing WebPlotDigitizer when necessary to convert graphical data into usable figures. Temperature data was sourced from metrological stations or, alternatively, from worldweatheronline.com. The harvest day was deduced from known or estimated days to heading, adding an additional 30 days. Final dataset outlined grain As concentrations from China, the United States, and Bangladesh, which are typical climate zones for rice production. One of the significant dataset was from Pinson et al.\u003csup\u003e\u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e\u003c/sup\u003e that had two strict water management practices i.e., continuously flooding and non-flooded throughout crop seasons for 1700 rice varieties.\u003c/p\u003e \u003cp\u003eForth, a dataset for grain As concentrations in paired main crop and ratoon crop rice was compiled using the keywords of \u0026ldquo;ratoon*\u0026rdquo;, \u0026ldquo;rice\u0026rdquo;, and \u0026ldquo;arsenic\u0026rdquo;. This specialized dataset, extracted from five key publications, provided 74 site-variety records exclusively sourced from prominent ratoon rice areas in China, inclusive of geographical data and harvest dates.\u003c/p\u003e \u003cp\u003eFifth, to determine the critical levels of extreme air temperature and VPD in our study site, daily based air temperatures, RH, and VPD from 1961 to 2019 are retrieved from local meteorological station in Kunshan (31.395083932921473N, 120.99975176170581E)\u003csup\u003e\u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e43\u003c/span\u003e\u003c/sup\u003e. The 2021 and 2022 data were used from our experimental monitoring described in following sections. The 2020 data was absent.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec19\" class=\"Section2\"\u003e \u003ch2\u003eData Normalization\u003c/h2\u003e \u003cp\u003eOur data normalization accounted for variations within each cultivar group, adjusting grain As concentrations (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ed), onto a 0 to 1 scale for comparability by following equation:\u003cdiv id=\"Equa\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equa\" name=\"EquationSource\"\u003e\n$$\\:Normalized\\:value\\:=\\frac{\\:{y}_{i}\\:-\\:min\\left({y}_{i}\\right)}{max\\left({y}_{i}\\right)\\:-\\:min\\left({y}_{i}\\right)}\\:$$\u003c/div\u003e\u003c/div\u003e,\u003c/p\u003e \u003cp\u003eWhere this normalization adjusts individual yield values based on each cultivar's grain tAs concentrations, setting the lowest value as 0 and the highest as 1, with all other values scaled proportionately in between. This approach harmonizes the data allowing consistent analytical assessment across different factors.\u003c/p\u003e \u003cp\u003e \u003cb\u003eTailored Piecewise Linear Model.\u003c/b\u003e \u003c/p\u003e \u003cp\u003eIn our methods of plotting the relationships between grain weights and temperatures (Supplementary Fig.\u0026nbsp;11), and between grain tAs concentration and grain weight (Supplementary Fig.\u0026nbsp;14), we constructed a tailored piecewise linear model using the \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:lmfit\\)\u003c/span\u003e\u003c/span\u003e Python library to dissect the data into distinctive linear segments at optimal breakpoints. We initialized parameters with an initial guess by visual inspection and set bounds to refine the search space. Utilizing the Levenberg-Marquardt minimization method, we conducted an optimization process, aiming for convergence with a chi-squared tolerance of 1E-14 and an extremely low x-tolerance value (1E-20), reflecting high precision. The outcome of the minimization was assessed by examining the convergence message and scrutinizing the optimized parameters. The optimal value of the breakpoint was extracted and used to finalize the piecewise linear model with \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:myPWLF.fit\\_guess\\)\u003c/span\u003e\u003c/span\u003e, ensuring an accurate fitting to the observed data.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec20\" class=\"Section2\"\u003e \u003ch2\u003eML and ML Model Explanation\u003c/h2\u003e \u003cp\u003eWe implemented the RF algorithm using the XGBoost library in Python. One key benefit of XGBoost is its skill in striking an optimal balance between accuracy and simplicity. This is achieved using regularized objective functions that prevent overfitting, ensuring robust and reliable predictions. For all models used, we allocated 80% of the data for training the RF model and reserved the remaining data for cross-validation. Additionally, we fine-tuned the following parameters: learning_rate, n_estimators, max_depth, subsample, colsample_bytree, gamma, alpha, and lambda. Hyperparameter optimization was conducted using Optuna, a Python package for automated hyperparameter tuning, for 50 combinations. Throughout this process, we employed the xgb.cv function for 10-fold cross-validation to prevent overfitting and evaluated the performance of each of the 50 models. The Mean Absolute Error (MAE) and Root Mean Squared Error (RMSE) on the test set, rather than the training set, served as the basis for selecting the best model among all evaluated models.\u003c/p\u003e \u003cp\u003eTo interpret the results of XGBoost models, we employed SHAP values to estimate the effect of each input features. SHAP values offer a method for explaining machine learning models based on information theory, which are widely used in machine learning and ecology\u003csup\u003e\u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e44\u003c/span\u003e\u003c/sup\u003e. SHAP values decompose a prediction into the sum of contributions from each feature plus the average model prediction. Formally, for a prediction \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\widehat{{y}_{i}}\\)\u003c/span\u003e\u003c/span\u003e, the SHAP values \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\varphi\\:}_{j}\\)\u003c/span\u003e\u003c/span\u003e for features \u0026#119895; satisfy:\u003cdiv id=\"Equb\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equb\" name=\"EquationSource\"\u003e\n$$\\:\\widehat{{y}_{i}}={\\varphi\\:}_{0}+\\sum\\:_{j=1}^{p}{\\varphi\\:}_{j}$$\u003c/div\u003e\u003c/div\u003e,\u003c/p\u003e \u003cp\u003ewhere \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\varphi\\:}_{0}\\)\u003c/span\u003e\u003c/span\u003e is the average prediction of the model across all instances (the base value). SHAP values indicate how much each feature contributes to moving the prediction from the average prediction \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\varphi\\:}_{0}\\)\u003c/span\u003e\u003c/span\u003e to the actual prediction \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\widehat{{y}_{i}}\\)\u003c/span\u003e\u003c/span\u003e. A positive SHAP value indicates that the feature increases the prediction, while a negative SHAP value indicates that the feature decreases the prediction. As a result, the effect of each feature on model predictions can be additively calculated. The SHAP Python package along with built-in functions from the XGBoost package was used to calculate SHAP values.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec21\" class=\"Section2\"\u003e \u003ch2\u003eStatistical Analysis\u003c/h2\u003e \u003cp\u003ePlanned and unpaired comparisons were performed between ambient heatwave scenarios and other climate scenarios using Tukey's honestly significant difference (HSD) test (P\u0026thinsp;\u0026le;\u0026thinsp;0.05) in Excels. Normal distribution of data is tested if needed with the Kolmogorov\u0026ndash;Smirnov Test.\u003c/p\u003e \u003c/div\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eData availability\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAll raw data, models, maps are available at https://github.com/xjtluhes/heatwave_semi-perennial_arsenic.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCode availability\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAll source code and models are available at https://github.com/xjtluhes/heatwave_semi-perennial_arsenic.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAcknowledgments\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe are deeply grateful to Dr. Johannes M. H. Knops for his insightful discussions during the preparation of this manuscript. We also acknowledge the financial support from the National Science Foundation of China (No. 42477116). Special thanks to our lab members, particularly Qianrui Huangfu, and to our technicians, Xiao Zhou and Xiaoping Xie, for their valuable technical assistance and support in conducting the experiments.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eEthics Declarations\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors declare no competing interests.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthor contribution statements\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eS.Z. (first author) conceived and designed the experiments, performed the experiments, analyzed the data, contributed materials and analysis tools, and wrote the manuscript under the supervision of the corresponding author, Z.C. secured the funding, conceived and designed the experiments, and contributed materials and analysis tools. J.S., L.H. W., and Y.G. Z. contributed to data analysis and manuscript writing.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eLesk C, Rowhani P, Ramankutty N (2016) Influence of extreme weather disasters on global crop production. Nature 529:84\u0026ndash;87\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhu X et al (2022) Manure amendment can reduce rice yield loss under extreme temperatures. Commun Earth Environ 3:147\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYao Q et al (2024) Molecular mechanisms underlying negative effects of transient heatwaves on crop fertility. Plant Communications\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGerber JS et al (2024) Global spatially explicit yield gap time trends reveal regions at risk of future crop yield stagnation. Nat Food 5:125\u0026ndash;135\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHerrero M et al (2021) Articulating the effect of food systems innovation on the Sustainable Development Goals. Lancet Planet Health 5:e50\u0026ndash;e62\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eInitiative M (2009) Investing in the future: a united call to action on vitamin and mineral deficiencies. Global Rep, 23\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBiswas JK, Warke M, Datta R, Sarkar D (2020) Is arsenic in rice a major human health concern? Curr pollution Rep 6:37\u0026ndash;42\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCassman KG, Grassini P (2020) A global perspective on sustainable intensification research. Nat Sustain 3:262\u0026ndash;268\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHenry A (2024) A step forward in breeding for ratooning ability in rice. Molecular Plant\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYuan S, Cassman KG, Huang JL, Peng SB, Grassini P (2019) Can ratoon cropping improve resource use efficiencies and profitability of rice in central China? Field Crops Res 234:66\u0026ndash;72\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhang B, Zhu X, Sun X, Zhang G, Bai C (2000) Screening of early indica ratoon rice varieties and research on high-yielding supporting technologies. Jiangxi Agricultural Sci Technol, 10\u0026ndash;12\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWang W et al (2020) Ratoon rice technology: A green and resource-efficient way for rice production. Adv Agron 159:135\u0026ndash;167\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhang S et al (2023) Sustained productivity and agronomic potential of perennial rice. Nat Sustain 6:28\u0026ndash;38\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eXu F et al (2021) The ratoon rice system with high yield and high efficiency in China: Progress, trend of theory and technology. Field Crops Res 272:108282\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBreshears DD et al (2021) Underappreciated plant vulnerabilities to heat waves. New Phytol 231:32\u0026ndash;39\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFleisher DH, Barnaby JY, Li S, Timlin D (2022) Response of a U.S. rice hybrid variety to high heat at Two CO2 concentrations during anthesis and grainfill. Agric For Meteorol 323:109058\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eQi D et al (2024) Mowing and nitrogen management guidelines for superior rice ratoon yields. Field Crops Res 308:109302\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYang DJ, Zhang NH, Lin ZP (2008) Research on the application of high-yield cultivation techniques of floodwater receding ratoon rice. Agricultural Sci Technol Communication, 62\u0026ndash;64 (In Chinese)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHu GS, Jiang XM, Chen FL, Ni SJ (2004) Causes of heat damage in middle-season rice and countermeasures. China Rice 10:32 (In Chinese)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWang J et al (2020) Thiolated arsenic species observed in rice paddy pore waters. Nat Geosci 13:282\u0026ndash;287\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYao H, Conrad R (2000) Effect of temperature on reduction of iron and production of carbon dioxide and methane in anoxic wetland rice soils. Biol Fertil Soils 32:135\u0026ndash;141\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMuehe EM, Wang T, Kerl CF, Planer-Friedrich B, Fendorf S (2019) Rice production threatened by coupled stresses of climate and soil arsenic. Nat Commun 10:4985\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhang S, Chen Z, Zhou X, Wu L (2024) -h. Manipulating nature heatwaves in chambers: bridging controlled and field conditions for accurate warming study in plant and soil systems. agriRxiv\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLin F et al (2023) GF14f gene is negatively associated with yield and grain chalkiness under rice ratooning. Front Plant Sci 14:1112146\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNevame A et al (2018) Relationship between high temperature and formation of chalkiness and their effects on quality of rice. BioMed research international. 1653721 (2018)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTang Z et al (2020) Dimethylarsinic acid is the causal agent inducing rice straighthead disease. J Exp Bot 71:5631\u0026ndash;5644\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eIshfaq M et al (2022) Physiological Essence of Magnesium in Plants and Its Widespread Deficiency in the Farming System of China. Front Plant Sci 13\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHu H, Shiying Z (1987) A study on the correlation between ratoon rice and main crop traits. Jiangxi Agricultural Sci Technol, 3\u0026ndash;5\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhang S et al (2024) Escalating arsenic contamination throughout Chinese soils. Nat Sustain\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBhattacharyya P et al (2014) Effect of elevated carbon dioxide and temperature on phosphorus uptake in tropical flooded rice (Oryza sativa L). Eur J Agron 53:28\u0026ndash;37\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eShimoyanagi R, Abo M, Shiotsu F (2021) Higher temperatures during grain filling affect grain chalkiness and rice nutrient contents. Agronomy 11:1360\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCao Y-x, Zhu J-q, Hou J (2020) Yield gap of ratoon rice and their influence factors in China\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAbdullah AB, Ito S, Adhana K (2006) in Proceedings for Workshop and Conference on Rice in the World at Stake, Vol. 2 28\u0026ndash;43\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eShi Y-L, Chen W-Q, Zhu Y-G (2024) Direct, Embedded, and Embodied Trade of Arsenic: 1990\u0026ndash;2019. Environmental Science \u0026amp; Technology\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFeng A-X et al (2020) Screening of rice varieties with low accumulation of heavy metals based on multiple target elements and their absorption and transport characteristics in rice plants. J Agricultural Resour Environ 37:988\u0026ndash;1000\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhu XQ, Deng QC, Li J (2007) Demonstration performance and key technologies of early-season rice-ratoon rice for the super hybrid rice YLiangyou1. China Rice 4:4\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNelson DW, Sommers LE (1996) Total carbon, organic carbon, and organic matter. Methods of soil analysis: Part 3 Chemical methods 5, 961\u0026ndash;1010\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMiller RO, Kissel DE (2010) Comparison of soil pH methods on soils of North America. Soil Sci Soc Am J 74:310\u0026ndash;316\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHeitkemper DT, Vela NP, Stewart KR, Westphal CS (2001) Determination of total and speciated arsenic in rice by ion chromatography and inductively coupled plasma mass spectrometry. J Anal At Spectrom 16:299\u0026ndash;306\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYuan Z-F et al (2021) Simultaneous measurement of aqueous redox-sensitive elements and their species across the soil-water interface. J Environ Sci 102:1\u0026ndash;10\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNawrocka A, Durkalec M, Michalski M, Posyniak A (2022) Simple and reliable determination of total arsenic and its species in seafood by ICP-MS and HPLC-ICP-MS. Food Chem 379:132045\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePinson SRM et al (2015) Worldwide genetic diversity for mineral element concentrations in rice grain. Crop Sci 55:294\u0026ndash;311\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eQin M et al (2021) Impact of climate change on evaporation paradox in province of Jiangsu in southeastern China. PLoS ONE 16:e0247278\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBerdugo M, Gait\u0026aacute;n JJ, Delgado-Baquerizo M, Crowther TW, Dakos V (2022) Prevalence and drivers of abrupt vegetation shifts in global drylands. Proceedings of the National Academy of Sciences 119, e2123393119\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-5397288/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-5397288/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eHeatwaves threaten global rice security and challenge the United Nations\u0026rsquo; Zero Hunger goals. Semi-perennial rice, with its ratooning ability to regrow from stubble after hot seasons, has shown potential for sustainable production. Here, we investigated the vulnerability and resilience of ratoon cropping system through two-year manipulative experiments of realistic heatwaves and analyzed experimental and global datasets. Our findings reveal that ratoon crops can compensate for up to 100% of heatwave-induced yield losses, maintain nutritional quality, and reduce arsenic (As) contamination by half. These benefits arise from widespread yield trade-offs between main and ratoon crops, and the seasonal mismatch between soil As availability and plant uptake. Ratoon practice can increase annual rice production by 3.9\u0026nbsp;million Mg in marginal regions, feeding 80\u0026nbsp;million people while concurrently reducing As exposure by 535 kg in China. This underscores the importance of integrating semi-perennial rice into climate-resilient agriculture.\u003c/p\u003e","manuscriptTitle":"Semi-Perennial Rice Sustains Yield and Food Safety Under Heatwaves and Beyond","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2024-11-25 14:12:23","doi":"10.21203/rs.3.rs-5397288/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"ac9e9640-e117-4fff-87d8-b07d8f99e704","owner":[],"postedDate":"November 25th, 2024","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[{"id":40536053,"name":"Earth and environmental sciences/Environmental sciences/Environmental impact"},{"id":40536054,"name":"Scientific community and society/Agriculture"},{"id":40536055,"name":"Biological sciences/Plant sciences/Plant stress responses/Heat"},{"id":40536056,"name":"Earth and environmental sciences/Biogeochemistry/Element cycles"}],"tags":[],"updatedAt":"2025-01-14T01:05:30+00:00","versionOfRecord":[],"versionCreatedAt":"2024-11-25 14:12:23","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-5397288","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-5397288","identity":"rs-5397288","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}