Effects of climate change and adaptive options for maize–wheat rotations in the North China Plain: a model-based evaluation

Climate change threatens the stability of crop production in the North China Plain (NCP), a major grain-producing region in China. Assessing crop responses to future climate change and the effectiveness of adaptive strategies is essential for sustaining regional food security. We used calibrated DSSAT–CERES models for maize and wheat to simulate phenological development and yield under future climate scenarios. Simulations were conducted for the 2050s and 2080s across multiple Shared Socioeconomic Pathway (SSP) scenarios, incorporating projected changes in temperature, solar radiation, and precipitation. Adaptive measures, including optimized sowing dates and alternative cropping systems, were evaluated. By the 2080s, the NCP is projected to experience pronounced warming, with maximum and minimum temperatures increasing by 2–6 °C and 0.4–3 °C, respectively, alongside increases of 0.3–0.8 MJ m⁻² in solar radiation and 280–430 mm in precipitation. Climate warming advanced anthesis and maturity by up to 12 and 21 days for summer maize and 28 and 14 days for winter wheat, leading to substantial yield reductions (17–82% for maize and 26–36% for wheat). Maize was more sensitive to warming than wheat. Optimized sowing dates increased maize yields by up to 90% and improved maize–wheat rotation yields by 20–35%, while single cropping further increased maize yields by 43–113% and wheat yields by 5–40%. Although climate change is projected to substantially reduce crop yields in the NCP, targeted management and cropping system adjustments can effectively mitigate yield losses and enhance agricultural resilience.

1 Introduction

Climate change poses one of the greatest challenges to global food security by altering crop growth conditions and threatening the stability of agricultural production (Porter et al., 2017). Rising temperatures, shifting precipitation regimes, and increased frequency of extreme events have already influenced crop development and productivity worldwide (Rosenzweig et al., 2014; Asseng et al., 2015). In China, the North China Plain (NCP) is the most important grain-producing region, supporting over one-third of the national wheat and maize production (Liang et al., 2011). However, this double-cropping system is highly vulnerable to climate variability, as it depends on finely tuned seasonal temperature and radiation conditions to sustain yields (Trnka et al., 2014; Ray et al., 2019; Rezaei et al., 2023).

Crop phenology is a sensitive indicator of climate change impacts. Warming trends accelerate crop development, leading to earlier flowering and maturity and a shorter grain-filling period, which ultimately reduces yields (Tao et al., 2013; Liu et al., 2019). Previous studies have demonstrated that a 1°C increase in growing-season temperature can reduce wheat and maize yields by 3–10% depending on growth stage sensitivity (Zhao et al., 2017). While management interventions such as adjusting sowing dates or planting densities have shown potential to mitigate yield losses, their effectiveness within the wheat–maize rotation system remains poorly quantified under future climate scenarios (Abramoff et al., 2023). Furthermore, most existing studies have focused on single crops (Yan et al., 2016), leaving a gap in understanding how entire rotation systems respond to climate change and whether alternative systems, such as single cropping, could provide greater resilience.

Process-based crop models, such as the DSSAT-CERES suite, offer a robust framework for integrating climate, soil, and management information to assess the impacts of climate change on crop growth and yield (Pathak et al., 2006; Djumaniyazova et al., 2010). These models have been widely applied for simulating phenological shifts, yield responses, and adaptive management strategies across diverse agroecosystems (Liu et al., 2017). Yet, comprehensive assessments of the NCP maize–wheat rotation system under the latest Shared Socioeconomic Pathways (SSPs), with explicit consideration of both crop-specific responses and system-level adaptations, remain limited (Xiao et al., 2020).

The objectives of this study are therefore threefold: (i) to quantify the effects of future climate scenarios on maize and wheat phenology and yields in the NCP; (ii) to evaluate the effectiveness of adaptive management measures, including sowing date and planting density adjustments, in mitigating yield losses; and (iii) to assess the potential of alternative cropping systems, such as single maize or single wheat, for sustaining production under different emission pathways. By addressing these objectives, this study provides new insights into the vulnerability and adaptive capacity of the NCP cropping system, offering a scientific basis for developing climate-resilient management strategies and safeguarding regional food security in the coming decades.

2 Materials and methods

2.1 Study site

The experimental site was located in Yueling village, Wangdu County, Hebei province (115.2°E, 38.69°N), which was the typical winter wheat-summer maize rotation area conducted by householders in the North China Plain (Figure 1). On average, the rainfall is 496 mm, mainly from July to September, and the annual average temperature is 12.6°C. The nature of the soil is sandy loam, belonging to a temperate semi-humid continental monsoon climate zone, with sufficient sunlight. Maize is planted in early June, harvested in late September, and then wheat is planted, harvested in early June in the next year. The fields investigated in this study are all winter wheat-summer maize rotation fields. The main wheat variety is ‘Luyuan 502’, with the straw crushed and returned to the field, and the main maize variety is ‘Denghai 605’, which adopts no-tillage and simultaneous sowing with fertilizer. Before sowing, composite soil samples from the 0–20 cm layer were collected using a soil auger. Subsamples were thoroughly mixed, quartered, and brought to the laboratory for analysis. Soil organic matter was determined by the potassium dichromate oxidation method; total nitrogen by the semi-micro Kjeldahl method; available phosphorus by the Olsen extraction; available potassium by NH₄OAc extraction with flame photometry; and soil pH using a glass electrode (1:2.5 soil: water), and the main parameters are provided in Supplementary Tables S2–S5.

Figure 1

Figure 1. Location of the study area in the North China Plain. The left panel shows the administrative boundary of China with the North China Plain highlighted in green. The right panel presents an enlarged view of the North China Plain, including the provinces of Beijing, Tianjin, Hebei, Shandong, Henan, Anhui, and Jiangsu. The red star indicates the experimental site located in Yueling village, Wangdu County, Hebei Province. Major nearby cities are also labelled to provide geographical orientation.

2.2 Experimental management data

We investigated the management measures and machinery use by face-to-face interviewing 50 farmers, including winter wheat and summer maize varieties, sowing date and rate, sowing row spacing and depth, fertilizer formula and amount, fertilizer distance, fertilizer application depth, irrigation date, and so on. To ensure that the collected management data were representative of the local wheat–maize production system, we used a stratified purposive sampling approach to select 50 households from Yueling village. This village is the major grain-producing area within Wangdu County and features a highly uniform soil type, cropping system, and water management practices across farms. The selected households covered a broad range of farm sizes, experience levels, and management intensities, thereby capturing the diversity of practices within the local production environment. Because the village exhibits relatively homogeneous agricultural conditions, interviewing 50 farmers provides a sufficiently robust representation of local-scale management practices. The five-point sampling method was used to collect soil samples from the 0–20 cm arable layer of each farmer’s plot before sowing to determine soil nutrient levels. From 2017 to 2018, household data were used to calibrate and validate cultivar-specific genetic coefficients for both crops. For winter wheat, eight households were used for calibration and sixteen households for validation. For summer maize, sixteen households were used for calibration and twelve households for validation (Supplementary Tables S2–S5). Meanwhile, obtain the phenological phase of maize through field observation. Plant samples were taken at the six-leaf, tasseling, and maturity stages of maize, dried in an oven at 105°C, and then dried to constant weight at 75°C to obtain the biomass for each stage. For winter wheat, aboveground biomass was obtained by collecting plant samples at the anthesis and harvest stages in the householders’ plot. These samples were dried at a temperature of approximately 70°C for 48 h in an oven and weighed. In the study area, farmers rely on surface-water irrigation drawn from nearby rivers. Each irrigation event applies approximately 90 mm of water. The specific irrigation schedule, including the timing and number of irrigation events for both winter wheat and summer maize, was incorporated into DSSAT and is detailed in Supplementary Tables S2–S5.

2.3 Climate data

Historically observed climate data were collected over 58 years (1961–2018) from the China Meteorological Data Network (http://data.cma.cn) at the Xingtai weather station, located adjacent to the experimental site in Hebei Province (114.50° E, 37.07° N). The dataset included daily maximum temperature (Tmax), minimum temperature (Tmin), precipitation (Prec), and sunshine hours (Sh). Daily solar radiation (Rad) was estimated from sunshine hours using the Ångström–Prescott equation.

Future climate projections were obtained from the Coupled Model Intercomparison Project Phase 6 (CMIP6) (O’Neill et al., 2016) of the World Climate Research Program (WCRP) (https://esgf-node.llnl.gov/search/cmip6/). CMIP6 provides enhanced climate projections by combining SSPs with Representative Concentration Pathways (RCPs), thereby capturing interactions between socioeconomic development, energy use, emissions, and land-use change. Among the SSPs, SSP1 and SSP5 represent relatively optimistic development trajectories, with SSP1 emphasizing sustainable growth and SSP5 relying on fossil-fuel–intensive development. In contrast, SSP3 and SSP4 reflect more pessimistic futures, with limited investment in education and health, rapid population growth, and increasing inequality.

Seven emission pathways were considered in this study: SSP1-1.9, SSP1-2.6, SSP2-4.5, SSP3-7.0, SSP4-3.4, SSP4-6.0, and SSP5-8.5. For regional analysis, we selected the CanESM5 global climate model (GCM) developed by the Canadian Centre for Climate Modelling and Analysis (CCCma) at a native resolution of approximately 2.8° × 2.8°. CanESM5 was chosen because of its reliable performance in reproducing historical temperature and precipitation patterns in East Asia and its frequent use in impact assessments across China (Swart et al., 2019).

Daily projections of Tmax, Tmin, precipitation, and solar radiation were extracted and bias-corrected using the quantile–quantile (qq) mapping method. Compared with linear scaling, qq-mapping has the advantage of correcting not only the mean bias but also the distribution of climate variables, particularly improving the representation of extremes (Jakob Themeßl et al., 2011). This is especially important for crop modeling, as crop growth and yield are highly sensitive to extreme weather events such as heat waves and droughts.

The baseline period was defined as 1975–2005, and two future time slices were considered: the 2050s (2020–2050) and the 2080s (2050–2080). These periods were selected to represent near- and mid-century climate conditions and to evaluate potential impacts and adaptation strategies under different emission pathways.

2.4 DSSAT model

This study employed CERES-Wheat and CERES-Maize within the DSSAT models to simulate the regional attainable yields of winter wheat and summer maize based on existing varieties and management techniques. Originally developed by James W. Jones in 1983, the DSSAT model has become one of the most widely used crop models globally (Jones et al., 2003). DSSAT is a biophysical model designed to simulate crop growth and development at the field scale. DSSAT is a process-based crop modeling platform that integrates crop physiological processes with soil water dynamics, nitrogen cycling, and management operations. The CERES models describe phenological development using thermal time accumulation, simulate photosynthesis and biomass formation with radiation-use efficiency approaches, and calculate grain yield based on partitioning coefficients regulated by stress factors.

Model simulations in DSSAT require three major categories of input data. First, the model uses daily weather variables, including maximum and minimum temperature, solar radiation, precipitation, and, when available, additional variables such as wind speed and relative humidity. Second, DSSAT requires detailed soil profile information describing soil texture, hydraulic properties, organic matter content, and nutrient status to initialize water and nutrient balance calculations. Third, field management inputs are needed to specify cultivar genetic coefficients, sowing date and density, fertilization and irrigation schedules, residue management, and tillage operations. Together, these inputs allow the CERES models to simulate crop growth, phenology, biomass accumulation, and yield formation under varying environmental and management conditions. In this study, DSSAT-CERES-Wheat and DSSAT-CERES-Maize were used to simulate the growth processes and yield of wheat and maize grain yield using the baseline and future climate scenarios. In 2017 and 2018, 14 farmers’ data were used to calibrate and verify the genetic parameters of the local main summer maize cultivar— ‘Denghai 605’ and main winter wheat cultivar — ‘Luyuan 502’, respectively. To verify the accuracy of model simulation, we used the root mean square error (RMSE) (Equation 1), normalized root mean square error (nRMSE) (Equation 2), and the absolute relative error (ARE) (Equation 3) to test the agreement between observed and simulated values:

$\begin{array}{ll}RMSE=\sqrt{\frac{{\sum }_{i=1}^n{(y_i-{\widehat{y}}_i)}^2}{n}}& (1)\end{array}$

$\begin{array}{ll}nRMSE=\frac{RMSE}{{\overline{y}}_i}\times 100\% & (2)\end{array}$

$\begin{array}{ll}ARE=\frac{\left|y_i-{\widehat{y}}_i\right|}{{\overline{y}}_i}\times 100\%& (3)\end{array}$

where $y_i$ and ${\widehat{y}}_i$ are the observed and simulated values, respectively. And ${\overline{y}}_i$ are the averages of the observations.

It is generally believed that the smaller the value of nRMSE and ARE, the smaller the simulation error. When nRMSE is<10%, the simulation accuracy is excellent; when 10%< nRMSE<20%, the simulation accuracy is good; when 20%< nRMSE<30%, the simulation accuracy is generally acceptable; and when nRMSE>30%, the simulation accuracy is poor.

2.5 Projected changes in annual rotation yield

In this study, we focused on the projected changes in the annual rotation yield of winter wheat–summer maize systems under optimized management for the 2050s (2020–2050) and 2080s (2050–2080) in the North China Plain. For each simulation year, the annual rotation yield (Y_rot) was calculated as the sum of winter wheat grain yield and summer maize grain yield. The baseline rotation yield (Y_{rot, base}) was defined as the mean annual rotation yield over the historical period (1975–2005), and the future rotation yield (Y_rot,fut​) as the mean value over each future time slice under a given SSP scenario.

The absolute change in annual rotation yield was computed as:

$$\text{Δ}{Y}_{rot}=\text{Δ}{Y}_{rot,fut}-\text{Δ}{Y}_{rot,base}$$

and the relative change was expressed as:

$$\text{Δ}{Y}_{rot}(\%)=\frac{\text{Δ}{Y}_{rot,fut}-\text{Δ}{Y}_{rot,base}}{Y_{rot,base}}\times 100\%$$

These metrics were used to quantify how optimized management can modify the response of the wheat–maize rotation system to future climate change.

3 Results

3.1 Calibration and validation of the DSSAT model by household data

From 2017 to 2018, sixteen households were selected for model parameter calibration, while the remaining twelve households were used for validation (Supplementary Tables S2–S5). For the CERES-Maize model, the calibrated parameter set showed strong performance in simulating anthesis day, biomass, and yield (Table 1). During calibration, the nRMSE for anthesis day was 2.76%, and the ARE was below 5.00%. Validation results showed an nRMSE of 2.75% and an average ARE of 2.03%, both lower than those obtained during calibration. For biomass simulation, the nRMSE was 13.37% and the average ARE was 12.10%, indicating that the simulated biomass was in close agreement with the measured values under the calibrated parameter set. For yield, the nRMSE and ARE values during calibration and validation were 6.77%, 6.28%, 5.91%, and 5.62%, respectively, further confirming the reliability of the model (Supplementary Table S6).

Table 1

Table 1. Calibrated genetics parameters of Denghai 605 in DSSAT-CERES-Maize model.

Similarly, the CERES-Wheat model performed well for the winter wheat variety ‘Luyuan 502’ (Table 2). The model accurately simulated anthesis day (ADAT), with an average nRMSE of 0.60% during calibration and a verified ARE of only 0.43%. Biomass simulation achieved an average nRMSE of 12.60% and an ARE of 10.50%. For yield, the calibration results showed an nRMSE of 5.07% and an average deviation (ARE) of 2.81%. Validation results yielded an nRMSE of 5.84% and an ARE of 5.36%, both within 10% (Supplementary Table S7).

Table 2

Table 2. Calibrated genetics parameters of Luyuan 502 in DSSAT-CERES-Wheat model.

Overall, these results demonstrate that once calibrated, the DSSAT-CERES models effectively capture crop phenology, biomass accumulation, and yield performance for maize and wheat across varying weather conditions and management practices.

3.2 Projected climatic conditions under future climate scenarios

The projected changes in meteorological factors in Wangdu County, Hebei Province, from 2015 to 2080 under different SSP climate scenarios are shown in Figure 2. Although the magnitude of change varies among scenarios, all exhibit an overall upward trend relative to the historical baseline (1975–2005). Under SSP1-1.9, representing a sustainable development pathway, the average maximum temperature is projected to increase by 2.01°C, the minimum temperature by 0.44°C, radiation by 0.84 MJ m^-2, and precipitation by 282.43 mm over the next 60 years. The SSP1-2.6 scenario shows a similar pattern, with slightly higher radiative forcing (stabilizing at 2.6 W m^-2 by 2100), resulting in increases of 2.45°C (maximum temperature), 0.99°C (minimum temperature), 0.79 MJ m^-2 (radiation), and 288.53 mm (precipitation).

Figure 2

Figure 2. The selected GCM projected values of mean maximum temperature (Tmax) (a) and minimum temperature (Tmin) (b), mean solar radiation (c), and total precipitation (d) during Wangdu county in 1975–2100 under SSP126-SSP585. Historically observed value was shown as an orange line.

Under the intermediate SSP2-4.5 scenario (4.5 W m^-2), the maximum and minimum temperatures are projected to rise by 2.69°C and 1.00°C, respectively, with radiation increasing by 0.30 MJ m^-2 and precipitation by 304.10 mm. SSP3-7.0 shows strong warming, with average increases of 2.51°C (maximum temperature) and 2.09°C (minimum temperature), accompanied by a decline in radiation (–0.44 MJ m^-2) and a precipitation increase of 349.60 mm. SSP4-3.4 and SSP4-6.0 yield moderate changes, with maximum temperatures rising by 2.79–2.98°C, minimum temperatures by 1.70–2.09 °C, radiation changing by +0.72 to –0.12 MJ m^-2, and precipitation increasing by 330.69–390.44 mm.

The high-emission SSP5-8.5 scenario projects the most pronounced changes. Compared with the baseline, maximum and minimum temperatures increase by 2.62°C and 2.79 °C, respectively, with radiation increasing by 0.36 MJ m^-2 and precipitation by 425.76 mm. Notably, the timing of warming differs between scenarios: under SSP5-8.5, the annual maximum temperature is projected to rise by 1.5°C by 2026 and 2°C by 2034, whereas under SSP1-1.9, these thresholds are reached in 2036 and 2070, respectively. Similarly, minimum temperatures under SSP5-8.5 reach +1.5°C by 2041 and +2.0°C by 2043, which are 11 and 28 years earlier than under SSP1-1.9. Precipitation under SSP5-8.5 is also substantially higher, averaging 143.3 mm more than under SSP1-1.9.

3.3 Project yield change of summer-maize and winter-wheat

The projected yields of summer maize and winter wheat showed substantial declines relative to the baseline across all SSP scenarios (Figure 3). For summer maize, yields in the 2050s decreased by approximately 20–70% depending on the emission pathway, with the smallest reduction observed under SSP1-1.9 and the largest under higher-emission scenarios (e.g., SSP3-7.0 and SSP5-8.5). By the 2080s, the magnitude of decline became even more pronounced, ranging from about 17% under SSP1-1.9 to more than 80% under SSP5-8.5, suggesting that the negative impact of climate change on maize production will intensify over time.

Figure 3

Figure 3. Boxplots show the distributions of simulated yields (kg ha^-1) for maize and wheat under the baseline climate and six future Shared Socioeconomic Pathway (SSP) scenarios across two time periods. Panels (a) and (b) present maize yields for 2021–2050 and 2050–2080, respectively, while panels (c) and (d) show wheat yields for the same periods. The baseline represents historical climate conditions, and SSP1–1.9, SSP1–2.6, SSP2–4.5, SSP3–7.0, SSP4–3.4, SSP4–6.0, and SSP5–8.5 represent future climate scenarios with increasing radiative forcing. Boxes indicate the interquartile range (25th–75th percentiles), horizontal lines denote medians, whiskers extend to 1.5 times the interquartile range, and dots represent outliers.

Winter wheat exhibited a similar downward trend, though the reductions were comparatively smaller than those of maize. In the 2050s, wheat yields decreased by roughly 28–33% across scenarios, while in the 2080s, the reductions ranged between about 26% and 36%. Although inter-scenario variability was evident, the general pattern indicated that even under more sustainable development pathways, wheat yields would remain significantly lower than baseline levels.

Overall, these results highlight that future climate change is projected to exert stronger yield penalties on summer maize than on winter wheat, and that the severity of yield losses increases with higher emissions and in the later part of the century.

3.4 Changes in crop phenology

Results indicate that future climate change will substantially affect the phenology of summer maize and winter wheat, particularly flowering and maturity periods (Figure 4). In general, both stages advanced across most SSP scenarios in the 2050s and 2080s, reflecting their strong temperature dependence. For summer maize, ensemble projections showed flowering dates advanced by 4–10 days in the 2050s and 4–12 days in the 2080s, while maturity occurred 3–14 days and 1–21 days earlier, respectively. Winter wheat exhibited a similar response, with anthesis advancing by 2–6 days in the 2050s and 14–28 days in the 2080s under SSP1-2.6–SSP5-8.5 scenarios. Correspondingly, maturity occurred 5–14 days earlier in the 2050s and 4–6 days earlier in the 2080s. An exception was observed under the low-emission SSP1-1.9 scenario, where flowering was delayed by ~12 days in the 2050s and ~10 days in the 2080s, although maturity still advanced by ~5 days. Overall, these findings suggest that rising growing-season temperatures are likely to accelerate crop development, leading to earlier flowering and shortened maturity durations under most scenarios.

Figure 4

Figure 4. Projected changes in flowering (ADAP) and maturity (MDAP) dates of wheat (a) and maize (b) in the North China Plain under baseline (1975–2005) and seven CMIP6 climate scenarios (SSP1-1.9, SSP1-2.6, SSP2-4.5, SSP3-7.0, SSP4-3.4, SSP4-6.0, SSP5-8.5). Bars represent the average number of days after planting (DAP) for anthesis and maturity. Numbers above the bars indicate changes relative to the baseline. Both maize and wheat show pronounced shifts in phenology under future climates, with earlier flowering and shortened maturity durations in most scenarios, particularly under high-emission pathways. (a) shows maize, and (b) shows wheat.

3.5 Sowing date and planting density adjustment design for future climate change

The seasonal analysis using the DSSAT-CERES model demonstrated that grain yield was strongly influenced by planting date and planting density under different future climate scenarios in the North China Plain (Figures 5, 6). For summer maize, advancing the sowing date from 10 June to 30 June consistently improved yields across all SSPs. Yield gains ranged from ~40–95% in the 2050s and ~43–98% in the 2080s, with the largest increases observed under high-emission scenarios. In contrast, adjustments to winter wheat sowing dates within the rotation cycle generally had a modest negative effect. Compared with the optimal sowing window of 10–30 October, average wheat yields decreased by ~5–12% in the 2050s and ~9–20% in the 2080s under most SSPs, although a slight yield increase (≈10–13%) was observed under the low-emission SSP1-1.9 scenario (Figure 6).

Figure 5

Figure 5. Simulated yield responses of maize and wheat to different sowing dates under multiple CMIP6 climate scenarios in the North China Plain. SD1–SD5 represent alternative sowing dates at 10-day intervals within the recommended sowing window. Boxplots indicate the distribution of simulated yields for each crop and sowing date, with medians (solid lines), interquartile ranges (boxes), whiskers (1.5× IQR), and dots representing outliers. Results show that sowing date adjustments substantially influence yield outcomes, with optimized sowing dates generally improving maize performance while having mixed effects on wheat. Panels a–g correspond to different climate scenarios: (a) SSP1-1.9; (b) SSP1-2.6; (c) SSP2-4.5; (d) SSP3-7.0; (e) SSP4-3.4; (f) SSP4-6.0; (g) SSP5-8.5.

Figure 6

Figure 6. Simulated yield responses of maize and wheat to different planting densities under multiple CMIP6 climate scenarios in the North China Plain. SD1–SD5 represent alternative sowing dates at 10-day intervals within the recommended sowing window. Boxplots indicate the distribution of simulated yields for each crop and sowing date, with medians (solid lines), interquartile ranges (boxes), whiskers (1.5× IQR), and dots representing outliers. Results show that sowing date adjustments substantially influence yield outcomes, with optimized sowing dates generally improving maize performance while having mixed effects on wheat. Panels a–g correspond to different climate scenarios: (a) SSP1-1.9; (b) SSP1-2.6; (c) SSP2-4.5; (d) SSP3-7.0; (e) SSP4-3.4; (f) SSP4-6.0; (g) SSP5-8.5.

When considering the maize–wheat rotation system as a whole, optimizing sowing date and planting density resulted in substantial increases in annual system productivity. Rotation yields improved by approximately 20–35% in the 2050s and 18–34% in the 2080s across SSPs. These results highlight that although climate change poses a considerable threat to food security, optimizing management practices such as sowing date and density can partially offset yield losses and enhance system resilience, even though a gap remains compared with baseline production (Figure 7).

Figure 7

Figure 7. Projected maize and wheat yields in the North China Plain under multiple CMIP6 climate scenarios. Boxplots show simulated yields for baseline management (red, blue) and optimized sowing date management (orange, green) in the 2050s (left panels) and 2080s (right panels). Each boxplot summarizes the distribution of simulated yields, with medians (horizontal lines), interquartile ranges (boxes), whiskers (1.5× IQR), and outliers (dots). Maize results are shown in the top panels and wheat results in the bottom panels. Scenarios on the x-axis correspond to SSP1-1.9 (ssp119), SSP1-2.6 (ssp126), SSP2-4.5 (ssp245), SSP3-7.0 (ssp370), SSP4-3.4 (ssp434), SSP4-6.0 (ssp460), and SSP5-8.5 (ssp585).

3.6 Changing wheat-wheat and summer-maize rotation to a single crop

Crop phenology is highly sensitive to climate change, particularly temperature, which can either advance or delay the cropping season. In the North China Plain, the conventional wheat–wheat–summer maize rotation may not always maximize the use of thermal and light resources under future climates. Model simulations suggest that replacing the rotation system with a single crop could substantially improve yield performance (Figure 8).

Figure 8

Figure 8. Projected changes in annual rotation yield of winter wheat–summer maize systems in the North China Plain under optimized management measures for the 2050s (2020–2050) and 2080s (2050–2080). Results are shown for seven CMIP6 climate scenarios (SSP1-1.9, SSP1-2.6, SSP2-4.5, SSP3-7.0, SSP4-3.4, SSP4-6.0, SSP5-8.5). Boxplots represent the distribution of simulated yield changes relative to the baseline, with solid lines indicating medians, box edges the interquartile range, whiskers 1.5× IQR, and dots outliers. Optimized management strategies increased rotation yields across all scenarios, though the magnitude of improvement varied with emission pathways and time periods.

For single maize, yields increased markedly relative to the optimized rotation system under most SSP scenarios, with gains of approximately 43–52% in the 2050s and 48–113% in the 2080s (SSP1-2.6–SSP5-8.5). However, under the low-emission SSP1-1.9 scenario, single maize resulted in yield reductions of about 18% in the 2050s and 15% in the 2080s. In contrast, single wheat also showed consistent yield benefits, though to a smaller extent than maize. Relative to the optimized rotation, single wheat yields increased by ~5–25% in the 2050s and ~6–40% in the 2080s across most scenarios.

Overall, these results indicate that shifting from a wheat–maize rotation to single cropping, particularly maize, could enhance yield potential under medium- to high-emission futures, while single wheat provides moderate but stable improvements. Nonetheless, under low-emission scenarios, rotation systems may remain more advantageous than single maize (Figure 9).

Figure 9

Figure 9. Effects of adaptive management strategies on maize and wheat yields under future climate scenarios in the North China Plain. Panels (a) and (b) show maize yields in the 2050s and 2080s, respectively, while panels (c) and (d) show wheat yields for the same periods. Results are presented across multiple Shared Socioeconomic Pathway (SSP) scenarios. For maize, red boxes represent baseline maize management, dark orange boxes represent optimized maize management (opt-maize), and light orange boxes represent summer maize–based cropping systems (summer-maize). For wheat, blue boxes indicate baseline wheat management, green boxes represent optimized wheat management (opt-wheat), and light blue boxes represent winter wheat systems (winter-wheat). Boxplots show the median (horizontal line), interquartile range (box), and 1.5× interquartile range (whiskers), with dots indicating outliers.

4 Discussion

Climate warming is increasingly recognized as a key driver of phenological change in agricultural systems, with substantial implications for crop productivity and regional food security. In the North China Plain (NCP), our model simulations showed consistent advancement of flowering and maturity stages for both maize and wheat under future climate scenarios. This aligns with long-term field records in northern China, where wheat anthesis advanced by approximately 4–6 days per °C increase in mean temperature (Liu et al., 2019). Such shifts shorten the effective grain-filling period, limiting biomass accumulation and yield potential (He et al., 2024). Similar patterns have been reported globally, where a 10-day reduction in wheat grain-filling duration was associated with 5–8% yield loss (Asseng et al., 2015). The projected advancement of maize maturity by up to three weeks under high-emission scenarios, therefore, represents a major constraint on sustaining yields.

The greater vulnerability of maize compared to wheat reflects differences in crop physiology and climate exposure. Maize is a C4 species optimized for temperatures between 25–30°C, but exposure to extreme heat (>35°C) during silking or grain filling severely reduces kernel set. Empirical studies in the NCP found that each 1°C increase in maximum temperature during silking reduced maize yields by 8–10% (Tian et al., 2019). Wheat, by contrast, grows primarily during cooler months and partially benefits from elevated atmospheric CO₂ concentrations, which enhance photosynthesis and water-use efficiency (Van Kesteren et al., 2013). These differences explain why maize yields were projected to decline by as much as 80%, while wheat yields decreased by 25–36% across scenarios. The crop-specific response highlights the importance of differentiated adaptation strategies: managing heat stress for maize versus improving resource efficiency and resilience in wheat (Xiao et al., 2021).

Adaptive management practices, such as sowing date optimization, can alleviate part of the yield decline (Abramoff et al., 2023; Luo et al., 2023). By shifting maize sowing later into June, crop growth aligns better with favorable temperature and radiation conditions, improving yields by up to 90% in some scenarios. However, the double-cropping rotation introduces unavoidable trade-offs: delaying maize sowing shortens the pre-winter growth period of wheat, reducing its yield (Zhang et al., 2023). Similar trade-offs have been documented in field trials, where optimizing maize sowing increased yields by 12–20% but reduced wheat production by 5–15% (Qu et al., 2025). These findings underscore the complexity of optimizing two crops within the same annual rotation under a warming climate, where management gains for one crop often come at the expense of the other.

The exploration of alternative cropping systems offers additional insights. Simulations of single cropping, especially single maize, revealed substantial yield advantages under medium- to high-emission scenarios, with increases of 40–100% compared with the traditional maize–wheat rotation (Meng et al., 2013). Such results reflect the benefit of allocating the entire growing season to a single crop, thereby maximizing resource use and reducing exposure to critical heat stress windows. Single wheat also demonstrated yield improvements, though to a lesser extent (Hochman et al., 2017). However, system-level implications must be carefully considered. Shifting away from rotations may improve individual crop yields but could reduce overall system diversity, affect soil fertility, and alter the balance between staple food supply and feed or industrial demand. Moreover, under low-emission pathways (e.g., SSP1-1.9), traditional rotations remained more favorable, indicating that cropping system transitions must be aligned with broader climate and socio-economic trajectories.

Considerable variability exists among CMIP6 models in simulating regional temperature and precipitation patterns, and this inter-model spread can influence projected crop responses (Eyring et al., 2016). Therefore, the yield impacts reported here should be interpreted with this uncertainty in mind. Future studies should adopt multi-model ensembles to capture the full range of climate projections better and reduce uncertainty in impact assessments. Second, although single cropping showed clear biophysical yield benefits, we did not evaluate its socioeconomic feasibility, which may constrain real-world adoption (Guo et al., 2024). Third, groundwater depletion in the North China Plain was not considered, and irrigation limitations may reduce the effectiveness of some adaptation strategies. These aspects warrant further investigation. Furthermore, this study did not explicitly account for the direct fertilization effect of elevated atmospheric CO₂ concentrations. Although higher CO₂ may enhance photosynthesis and water-use efficiency in C₃ crops such as wheat, its impact on maize is generally limited due to its C₄ physiology (Leakey et al., 2006). Excluding this effect may therefore lead to slightly conservative yield estimates for wheat under future scenarios. Incorporating explicit CO₂ response functions in future modeling efforts would help reduce this source of uncertainty and provide a more comprehensive assessment of climate change impacts. Finally, while management interventions improved productivity under future climates, none fully restored yields to baseline levels (Li et al., 2014). This highlights the limits of field-level adaptation and emphasizes the need for integrative strategies. These may include breeding for heat- and drought-tolerant cultivars, optimizing fertilizer and water management, and enhancing soil organic matter to buffer against climatic extremes (Crain et al., 2018). At the policy level, adaptive cropping system planning should be combined with investments in technology and infrastructure to reduce vulnerability (Duesberg et al., 2014). Together, these measures could help balance food security, resource use, and environmental sustainability in one of China’s most critical agricultural regions.

5 Conclusion

This study provides a comprehensive assessment of maize–wheat rotation systems in the North China Plain under future climate change using the DSSAT-CERES models. Results show that rising temperatures will advance phenology, shorten growth duration, and substantially reduce yields, with maize being more sensitive than wheat (Rezaei et al., 2023). Adaptive management measures, such as adjusting sowing dates and planting densities, can partly offset yield losses, while system-level adaptations like single cropping may further enhance production under high-emission scenarios. However, none of these strategies fully recover baseline yields, underscoring the limits of management interventions alone (Rising and Devineni, 2020; Abramoff et al., 2023; Qu et al., 2025).

Our results further show that the suitability of cropping systems varies across climate scenarios. Under medium- to high-emission pathways (SSP2-4.5 to SSP5-8.5), single maize provides the largest yield gains (40–100%), while single wheat offers moderate but consistent benefits. In contrast, under the low-emission SSP1-1.9 scenario, the wheat–maize rotation remains the more favorable option, as single maize reduces yields. These findings indicate that policymakers should adopt flexible, scenario-specific strategies—maintaining rotations under strong mitigation futures while considering single-crop systems as warming intensifies—to enhance the climate resilience of this key grain-producing region.

Data availability statement

The original contributions presented in the study are included in the article/Supplementary Material. Further inquiries can be directed to the corresponding author.

Author contributions

LM: Writing – original draft. ZL: Writing – review & editing, Data curation. LS: Writing – review & editing. HR: Writing – review & editing. SH: Writing – review & editing. YZ: Writing – review & editing.

Funding

The author(s) declared that financial support was received for this work and/or its publication. The work was supported by the Tianjin Renai College–Tianjin University Faculty Development Fund Collaborative Project (Grant No. FZ231004). The funder had no role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Conflict of interest

The authors declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Generative AI statement

The author(s) declared that generative AI was not used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Supplementary material

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fagro.2025.1736278/full#supplementary-material

References

Abramoff R. Z., Ciais P., Zhu P., Hasegawa T., Wakatsuki H., and Makowski D. (2023). Adaptation strategies strongly reduce the future impacts of climate change on simulated crop yields. Earth Future. 2023, 11. doi: 10.1029/2022EF003190

Crossref Full Text | Google Scholar

Asseng S., Ewert F., Martre P., Rötter R. P., Lobell D. B., Cammarano D., et al. (2015). Rising temperatures reduce global wheat production. Nat. Clim. Change. 5, 143–147. doi: 10.1038/nclimate2470

Crossref Full Text | Google Scholar

Crain J., Mondal S., Rutkoski J., Singh R. P., and Poland J. (2018). Combining high-throughput phenotyping and genomic information to increase prediction and selection accuracy in wheat breeding. Plant Genome. 11, 170043. doi: 10.3835/plantgenome2017.05.0043

PubMed Abstract | Crossref Full Text | Google Scholar

Djumaniyazova Y., Sommer R., Ibragimov N., Ruzimov J., Lamers J., and Vlek P. (2010). Simulating water use and N response of winter wheat in the irrigated floodplains of Northwest Uzbekistan. Field Crops Res. 116, 239–251. doi: 10.1016/j.fcr.2010.01.001

Crossref Full Text | Google Scholar

Duesberg S., Dhubháin Á.N., and O Connor D. (2014). Assessing policy tools for encouraging farm afforestation in Ireland. Land Use Policy. 38, 194–203. doi: 10.1016/j.landusepol.2013.11.001

Crossref Full Text | Google Scholar

Eyring V., Bony S., Meehl G. A., Senior C. A., Stevens B., Stouffer R. J., et al. (2016). Overview of the Coupled Model Intercomparison Project Phase 6 (CMIP6) experimental design and organization. Geosci. Model. Dev. 9, 1937–1958. doi: 10.5194/gmd-9-1937-2016

Crossref Full Text | Google Scholar

Guo X., Zhang P., and Yue Y. (2024). Prediction of global wheat cultivation distribution under climate change and socioeconomic development. Sci. Total Environ. 919, 170481. doi: 10.1016/j.scitotenv.2024.170481

PubMed Abstract | Crossref Full Text | Google Scholar

He Y., Xiong W., Hu P., Huang D., Feurtado J. A., Zhang T., et al. (2024). Climate change enhances stability of wheat-flowering-date. Sci. Total Environ. 917, 170305. doi: 10.1016/j.scitotenv.2024.170305

PubMed Abstract | Crossref Full Text | Google Scholar

Hochman Z., Gobbett D. L., and Horan H. (2017). Climate trends account for stalled wheat yields in Australia since 1990. Glob. Change Biol. 23, 2071–2081. doi: 10.1111/gcb.13604

PubMed Abstract | Crossref Full Text | Google Scholar

Jakob Themeßl M., Gobiet A., and Leuprecht A. (2011). Empirical-statistical downscaling and error correction of daily precipitation from regional climate models. Int. J. Climatol. 31, 1530–1544.

Google Scholar

Leakey A. D. B., Uribelarrea M., Ainsworth E. A., Naidu S. L., Rogers A., Ort D. R., et al. (2006). Photosynthesis, productivity, and yield of maize are not affected by open-air elevation of CO2 concentration in the absence of drought. Plant Physiol. 140, 779–790. doi: 10.1104/pp.105.073957

PubMed Abstract | Crossref Full Text | Google Scholar

Li K., Yang X., Liu Z., Zhang T., Lu S., and Liu Y. (2014). Low yield gap of winter wheat in the North China Plain. Eur. J. Agron. 59, 1–12. doi: 10.1016/j.eja.2014.04.007

Crossref Full Text | Google Scholar

Liang W., Carberry P., Wang G., Lü R., Lü H., and Xia A. (2011). Quantifying the yield gap in wheat–maize cropping systems of the Hebei Plain, China. Field Crops Res. 124, 180–185. doi: 10.1016/j.fcr.2011.07.010

Crossref Full Text | Google Scholar

Liu B., Martre P., Ewert F., Porter J. R., Challinor A. J., Müller C., et al. (2019). Global wheat production with 1.5 and 2.0°C above pre-industrial warming. Glob. Change Biol. 25, 1428–1444. doi: 10.1111/gcb.14542

PubMed Abstract | Crossref Full Text | Google Scholar

Liu G., Hou P., Xie R., Ming B., Wang K., Xu W., et al. (2017). Canopy characteristics of high-yield maize with yield potential of 22.5 Mg ha–1. Field Crops Res. 213, 221–230. doi: 10.1016/j.fcr.2017.08.011

Crossref Full Text | Google Scholar

Luo N., Meng Q., Feng P., Qu Z., Yu Y., Liu D. L., et al. (2023). China can be self-sufficient in maize production by 2030 with optimal crop management. Nat. Commun. 2023, 14. doi: 10.1038/s41467-023-38355-2

PubMed Abstract | Crossref Full Text | Google Scholar

Meng Q., Hou P., Wu L., Chen X., Cui Z., and Zhang F. (2013). Understanding production potentials and yield gaps in intensive maize production in China. Field Crops Res. 143, 91–97. doi: 10.1016/j.fcr.2012.09.023

Crossref Full Text | Google Scholar

O’Neill B. C., Tebaldi C., van Vuuren D. P., Eyring V., Friedlingstein P., Hurtt G., et al. (2016). The scenario model intercomparison project (ScenarioMIP) for CMIP6. Geosci. Model. Dev. 9, 3461–3482. doi: 10.5194/gmd-9-3461-2016

Crossref Full Text | Google Scholar

Pathak H., Li C., Wassmann R., and Ladha J. K. (2006). Simulation of nitrogen balance in rice–wheat systems of the indo-gangetic plains. Soil Sci. Soc Am. J. 70, 1612–1622. doi: 10.2136/sssaj2005.0165

Crossref Full Text | Google Scholar

Porter J. R., Howden M., and Smith P. (2017). Considering agriculture in IPCC assessments. Nat. Clim. Change. 7, 680–683. doi: 10.1038/nclimate3404

Crossref Full Text | Google Scholar

Qu X., Yao W., Ji H., Xu Y., Jia R., Chen X., et al. (2025). Optimizing nitrogen fertilization and irrigation strategies to balance agroecosystem services in the wheat-maize double cropping system: A 21-year field study. Field Crops Res. 322, 109706. doi: 10.1016/j.fcr.2024.109706

Crossref Full Text | Google Scholar

Ray D. K., West P. C., Clark M., Gerber J. S., Prishchepov A. V., and Chatterjee S. (2019). Climate change has likely already affected global food production. PLoS One. 14, e0217148. doi: 10.1371/journal.pone.0217148

PubMed Abstract | Crossref Full Text | Google Scholar

Rezaei E. E., Webber H., Asseng S., Boote K., Durand J. L., Ewert F., et al. (2023). Climate change impacts on crop yields. Nat. Rev. Earth Environ. 4, 831–846. doi: 10.1038/s43017-023-00491-0

Crossref Full Text | Google Scholar

Rising J. and Devineni N. (2020). Crop switching reduces agricultural losses from climate change in the United States by half under RCP 8.5. Nat. Commun. 2020, 11. doi: 10.1038/s41467-020-18725-w

PubMed Abstract | Crossref Full Text | Google Scholar

Rosenzweig C., Elliott J., Deryng D., Ruane A. C., Müller C., Arneth A., et al. (2014). Assessing agricultural risks of climate change in the 21st century in a global gridded crop model intercomparison. Proc. Natl. Acad. Sci. U.S.A 111, 3268–3273. doi: 10.1073/pnas.1222463110

PubMed Abstract | Crossref Full Text | Google Scholar

Tao F., Zhang S., and Zhang Z. (2013). Changes in rice disasters across China in recent decades and the meteorological and agronomic causes. Reg Environ. Change. 13, 743–759. doi: 10.1007/s10113-012-0357-7

Crossref Full Text | Google Scholar

Tian B., Zhu J., Nie Y., Xu C., Meng Q., and Wang P. (2019). Mitigating heat and chilling stress by adjusting the sowing date of maize in the North China Plain. J. Agron. Crop Sci. 205, 77–87. doi: 10.1111/jac.12299

Crossref Full Text | Google Scholar

Trnka M., Rötter R. P., Ruiz-Ramos M., Kersebaum K. C., Olesen J. E., Žalud Z., et al. (2014). Adverse weather conditions for European wheat production will become more frequent with climate change. Nat. Clim. Change. 4, 637–643. doi: 10.1038/nclimate2242

Crossref Full Text | Google Scholar

Van Kesteren B., Hartogensis O. K., van Dinther D., Moene A. F., and De Bruin H. A. R. (2013). Measuring H2O and CO2 fluxes at field scales with scintillometry: Part I – Introduction and validation of four methods. Agric. For. Meteorol. 178-179, 75–87. doi: 10.1016/j.agrformet.2012.09.013

Crossref Full Text | Google Scholar

Xiao D., Liu D. L., Feng P., Wang B., Waters C., Shen Y., et al. (2021). Future climate change impacts on grain yield and groundwater use under different cropping systems in the North China Plain. Agric. Water Manage. 246, 106685. doi: 10.1016/j.agwat.2020.106685

Crossref Full Text | Google Scholar

Xiao D., Liu D. L., Wang B., Feng P., Bai H., and Tang J. (2020). Climate change impact on yields and water use of wheat and maize in the North China Plain under future climate change scenarios. Agric. Water Manage. 238, 106238. doi: 10.1016/j.agwat.2020.106238

Crossref Full Text | Google Scholar

Yan P., Zhang Q., Shuai X.f., Pan J.x., Zhang W.j., Shi J.f., et al. (2016). Interaction between plant density and nitrogen management strategy in improving maize grain yield and nitrogen use efficiency on the North China Plain. J. Agric. Sci. 154, 978–988. doi: 10.1017/S0021859615000854

Crossref Full Text | Google Scholar

Zhang D., Guo Y., Fan Z., Hu X., Hao X., Fang L., et al. (2023). Trade-offs between grain yields and ecological efficiencies in a wheat–maize cropping system using optimized tillage and fertilization management on the North China Plain. Environ. Sci. pollut. Res. 30, 24479–24493. doi: 10.1007/s11356-022-23644-y

PubMed Abstract | Crossref Full Text | Google Scholar

Zhao C., Liu B., Piao S., Wang X., Lobell D. B., Huang Y., et al. (2017). Temperature increase reduces global yields of major crops in four independent estimates. Proc. Natl. Acad. Sci. U.S.A 114, 9326–9331. doi: 10.1073/pnas.1701762114

PubMed Abstract | Crossref Full Text | Google Scholar