Effects of climate factors on wheat and maize under different crop rotations and irrigation strategies in the North China Plain

The North China Plain (NCP) is a crucial agricultural region for grain production in China, primarily focusing on wheat and maize cultivation. However, these crop yields are highly vulnerable to fluctuations in temperature and precipitation, with climate change being a significant factor. This study investigates the impact of climate factors on wheat and maize yields in the NCP under various crop rotations and water supply strategies. Using the Agricultural Production Systems sIMulator crop mechanism model, we evaluated the effects of different crop rotations and water supply strategies on wheat and maize yields. A comprehensive analysis of the simulated results determined the yield variation trends and their correlation and sensitivity to different climate factors. The findings revealed that precipitation levels over the past 40 years showed no significant trend, while radiation levels showed a significant decreasing trend, and annual mean maximum and minimum temperatures displayed significant increasing trends. Furthermore, the study found that irrigation practices and crop rotations substantially impact grain yield in the study area, with average yields ranging from 8105.5 kg ha−1 under rainfed conditions to 13 088.8 kg ha−1 under fully irrigated conditions. There was a statistically significant trend of increasing yields for fully irrigated Monocrop-Wheat and decreasing yields for fully irrigated Monocrop-Maize over the past 40 years. Sensitivity analysis also showed that rational rotation and irrigation can reduce grain yield sensitivity to climate change. In conclusion, the prudent use of rotation and irrigation can enhance food production resilience to climate change. These findings have significant implications for developing strategies to optimize crop yields and adapt to climate change in the NCP while considering trade-offs.


Introduction
The North China Plain (NCP) is a major grain production region in China, where wheat and maize are two of the main crops grown.However, temperature and precipitation significantly influence the yields of these crops, with climate change playing a significant role in these variations.It is important to examine the influence of climate factors on the spatial and temporal variability of wheat and maize yields in the NCP under different crop rotation patterns and water supply strategies.By understanding the impact of these factors, decision-makers can find strategies to optimize crop yields and adapt to climate change in the region.
Several studies have investigated the influence of climate factors on crop yields in the NCP.According to Shao et al (2021), elevating temperature has a mild negative effect on maize yield, while Song et al (2022) found that climate change has had a significant impact on wheat and maize yields at the household and village levels.Adverse weather events such as drought, flooding, and blight can significantly reduce wheat yields by 10%, 6%, and 16%, respectively, due to variations in precipitation (Song et al 2022).These findings underscore the importance of considering precipitation's role in deciding wheat yields and the potential impacts of climate change on agricultural production in the region.Solar dimming has also had a significant effect on maize yield potential in the NCP, with Meng et al (2020) observing a 17% decrease in solar radiation over the maize growing season, which explained 87% of the decrease in yield potential.In a study by Chen et al (2012), it was found that decreasing trends in solar radiation had a significant negative impact on simulated wheat and maize yields in Beijing, while the warming trend had a significant negative impact on simulated maize yield at Beijing, but not on wheat yield at either Beijing or Zhengzhou.The interannual variability of precipitation made its trends less prominent.In addition to climate factors, management practices are also an important influence on yield in a changing environment.Ojeda et al (2021) found that the main drivers of irrigated potato yield variance in Tasmania, Australia were crop management factors, particularly irrigation strategy and planting date, which were influenced by soil type and climate factors such as global solar radiation and temperature.The relative importance of these factors varied depending on the soil type.Changes in climatic factors such as temperature, precipitation and solar radiation significantly affect crop yields in the NCP.In addition to climatic factors, crop management practices such as irrigation strategies and sowing dates can also influence yields, and their relative importance may vary depending on soil type.Different crop rotation patterns can be created by changing the sowing dates and different crop arrangements.Crop rotation has a positive effect on soil fertility and plant nutrition.It can improve the physical environment of the soil, facilitating water infiltration and holding, aeration and root growth (Usharani et al 2019).Soil chemical analysis results suggest that the labile pool of organic matter in soil is more responsive to changes in crop rotation than to mineral fertilizer application, and that agricultural landscapes with different crop rotations may have different impacts on microbial cycling of organic matter in soil waters (Xu et al 2013).Additionally, Crop rotation also increases soil organic matter (SOM) which improves water retention and drought tolerance, as well as decreasing erosion.Additionally, crop rotation affects nutrient management by changing the nutrient requirements of different crops (Deiss et al 2021).A meta-analysis conducted in China found that crop rotation can increase crop yields by an average of 20% compared to continuous monoculture practices.The benefits of crop rotation on grain yield in the NCP depend on the type of pre-crops, ranging from 2% for triticale pre-crops to 27% for grain leguminous precrops, and can result in a 50% increase in profit for Chinese farmers (Zhao et al 2020).
Crop rotation and water management strategies can significantly affect water consumption.Specific crop rotations can supply benefits to subsequent crops and potentially to the environment.Improving soil physical conditions can facilitate water infiltration, retention, aeration, and root growth.The timing and amount of irrigation water applied are crucial decisions that affect profits and crop yields.Studies have shown that diversifying crop rotations can be an effective way to mitigate over-exploitation of groundwater, ensure food security, and increase farmers' income.For example, Yang et al (2015) found that annual average water use efficiency (WUE) increased in a specific rotation for cropping systems studied in the NCP.Gao et al (2022) used modeling analysis to evaluate the water efficiency and impacts on groundwater table depth of different crop rotation systems in Zhangjiakou, China.They found that certain systems such as a 2 year fallow with one single crop and potato-maize and 2 year fallow-potato rotations were more water-efficient.Thilagavathi et al (2021) investigated the impact of three optimization algorithms on agricultural land in Coimbatore, India.Their findings indicate that optimal land allocation can significantly increase agricultural profits.Jiang et al (2021) conducted a field study in the Loess Plateau region of China to examine the effects of different crop rotation sequences on soil water storage, crop yield, water use, and water productivity.Their results suggest that the effects of crop rotation sequences vary by year and treatment, and the treatments with the highest grain yield and WUE are not consistent.Aggarwal et al (2022) developed a decision support tool to optimize crop rotation practices for sustainable agricultural land use.The tool aims to preserve land fertility, maximize profit, minimize pollution, and reduce water usage.They tested the tool by analyzing the suitability of seven major crops in India, and the results suggest that sugarcane and other crops are viable options.
Climate change can significantly impact crop rotation by altering the suitability of crops for a given location and affecting planting and harvesting timing and success.One way climate change can affect crop rotation is by changing growing conditions for different crops.Zhang et al (2021) found that climate change, specifically temperature and humidity changes, significantly affected drought occurrence and crop yield in the NCP during the winter wheat growth season.Temperature showed a notable upward trend that worsened water deficit and drought risk.Li and Lei (2022) reported that climate extremes, such as heat stress, spring frost, meteorological drought, and extreme wet events, had a negligible impact on winter wheat and summer maize growth in the NCP.However, mean climate conditions, particularly temperature, solar radiation, and vapor pressure deficit, had a considerable influence on crop growth.
The Agricultural Production Systems sIMulator (APSIM) is a crop modeling tool that can simulate the impacts of various cropping practices, including crop rotation, on crop yields and other agronomic and environmental outcomes (Le Gal et al 2010).APSIM can evaluate the potential benefits and tradeoffs of different crop rotation strategies under various climate scenarios, including those influenced by climate change.One way to study crop rotation in the context of climate change using APSIM is by simulating the performance of different crops under different climate conditions.For instance, APSIM can estimate how different crops are likely to respond to changes in temperature, precipitation, and other climate variables, and how these responses may vary across regions.This information can help identify crop rotations that are more resilient to the impacts of climate change and potential vulnerabilities in current crop rotation practices.Additionally, APSIM can evaluate the impacts of different cropping practices, such as irrigation and fertilization, on crop yields and other agronomic and environmental outcomes.This information can help mitigate the negative impacts of climate change on crop production and optimize crop rotations to maximize yields and reduce risks.Therefore, the APSIM model is a useful tool for studying the impacts of climate change on crop rotation and finding strategies to adapt to changing growing conditions.
Previous studies indicate that crop rotation patterns and irrigation practices significantly affect wheat and maize yields across the NCP.However, further research is needed to better elucidate the interactions between these factors and pinpoint strategies to improve crop resilience to climate change in the region.The APSIM model can simulate the effects of various cropping systems, including rotations, on yields and other outcomes under projected climate scenarios.This tool enables identifying resilient rotations and evaluating cropping practices to adapt to changing conditions.In this study, we used APSIM simulations and statistical analyses to examine how climate influences wheat and maize yields in the NCP under different rotations and irrigation regimes.Our objectives were to (1) identify yield trends under different crop rotations and irrigation practices, (2) analyze the correlation between climate factors and yield under different crop rotation patterns and irrigation practices, (3) elucidate the drivers of crop yield variability in the NCP and identify strategies to bolster crop resilience to climate change.

Study region
This study's regional crop simulation covered an area located between the eastern Taihang Mountains and the northern Yellow River (112.5 • E-119.5 • E, 34.8 • N-40.5 • N) in the NCP, encompassing the entire plain of Beijing, Tianjin, Hebei, and northwestern Shandong Province (figure 1).This region has a flat topography and an average elevation of 30 m.It has a relatively mild climate, with an annual average temperature of 14 • C and a frost-free period of over 200 d.The area experiences distinct seasonal variations, with warm and rainy summers and cold, dry winters.These meteorological conditions are typical of East Asian monsoon regions and are conducive to cultivating a wide range of crops.NCP covers an area of approximately 1.4 × 10 5 km 2 , which constitutes 8.3% of China's total agricultural land.Among that, 54% is designated as cropland, with 85% of the cropland being irrigated (Pei et al 2015).The soil in the NCP is characterized as loamy, with a bulk density of 1.3 g cm −3 and a pH range of 7.5-7.9 at the 0-20 cm depth.The concentration of SOM and total nitrogen in the top 20 cm of soil is approximately 1.0%-2% and 1.2 g kg −1 , respectively.The grain production in the NCP significantly contributes to China's overall agricultural output, yielding approximately 26.2 million tons of wheat and 26.8 million tons of maize (Zhang et al 2018).These figures account for 20.3% of China's total wheat production and 12.2% of the nation's total maize production.The dominant cropping system in the region is double cropping of winter wheat and summer maize, which accounts for 49.5% of the total area under cropping (figure 1).The cropping patterns map is based on the ChinaCP dataset (Qiu et al 2022).The region has always relied on groundwater for irrigation, ensuring high food production.

Crop model description
This study utilized the APSIM Wheat and Maize model (Keating et al 2003, Holzworth et al 2018), a process-based crop model.The model is constructed based on the Plant Modelling Framework, which is a general model framework for simulating plant phenological and physiological processes in response to environmental conditions and field management practices.The APSIM model was widely applied to crop simulation in the NCP.(Zhang et al 2012, Gaydon et al 2017).
APSIM simulations are generated using crop management, weather, and soil data inputs.Weather data requires daily measurements of precipitation, solar radiation, and maximum and minimum air temperature (at 2.0 m above ground).The weather file named MET requires three parameters, including latitude, TAV (annual average air temperature), and AMP (annual amplitude in mean monthly temperature).Soil profile data parameters include bulk density, organic carbon, soil pH, electrical conductivity, and soil hydraulic parameters (air dry, crop lower limit (LL15), drained upper limit (DUL), and saturated water content (SAT)).Crop management inputs include cultivar, plant density, crop configuration (plant spacing, row spacing), sowing date, fertilization, and irrigation regime.The climate data employed in the dataset includes air maximum temperature, air minimum temperature, precipitation, and radiation grid at a resolution of 0.25 The soil dataset is also at a resolution of 0.25

Scenarios setting
Table 1 shows the modeling scenarios setting.The first column lists the crop rotation patterns, including the crop rotation sequence and species.Winter wheat & Summer maize refer to model simulates of an annual double-cropping system of winter wheat and summer maize; Winter wheat & Summer maize + Monocrop maize refer to that model simulates of the 2 year crop rotation cycle.In the first year, winter wheat and summer maize are planted in a double-cropping system.In the second year, only maize is planted.Monocrop Winter wheat and Monocrop maize refer to model simulations of annual monoculture wheat and monoculture maize, respectively.The second to fourth columns describe the water supply scenarios, which include rainfed, deficit irrigation, and full irrigation.The rainfed scenario relies on natural precipitation as the water supply (water stress factor of photosynthesis about 0.52 for maize, 0.77 for wheat).
The stress factors for maize range from heavy to light from 0 to 1, while the opposite is true for wheat, ranging from 1 to 0. The deficit irrigation scenario supplies water according to the crop water stress factor of photosynthesis about 0.72 for maize and 0.54 for wheat.The full irrigation scenario ensures that the crop is not subjected to water stress with a full irrigation water supply.The Management module of Automatic_irrigation_based_on_water_deficit is used to simulate the full irrigation scenario, while the deficit irrigation scenario is simulated by the Management module of Automatic Irrigation in the APSIM.The Automatic Irrigation module's irrigation amount range was set from 60 to 420 mm with a step of 60 mm.For the full irrigation simulations, irrigation was triggered based on the plant-available water content (PAWC) in the top 0-0.6 m soil layer.Irrigation occurred when PAWC dropped below 80% of capacity, replenishing soil moisture to field capacity.Irrigation was also triggered if more than 5 d elapsed since the previous event, again refilling to field capacity to avoid plant water stress.For the deficit irrigation scenarios, total irrigation amounts followed table 1. Irrigation was triggered at 30% PAWC, applying <60 mm per event, with at least 15 d between events, until the seasonal irrigation limit was reached.
The simulation scenarios were designed by combining 281 weather data and soil grid points, four cropping rotations, and three water supply strategies (auto full irrigation, deficit irrigation, rainfed).The plant density and cultivars were fixed to disentangle the rotation and water supply strategy effects on yield variance from those due to interactions and feedback effects among crop management, soil, and weather conditions.The cultivars' genotype parameters used in this study, as reported by Yan et al (2020), showed exceptional performance.The wheat cultivar used in this study was Kenong199 and the maize hybrid cultivar was Zhengdan958, both of which are commonly grown in NCP.Variety turnover at different years and variety variation in space were not considered in this study.
The sowing in this study was done by automatically selecting the sowing date based on preset conditions rather than a fixed date (table 2).The sowing conditions for wheat and maize were based on the harvest of the previous crop, soil moisture content, and precipitation conditions to determine sowing suitability.Once the conditions met the criteria, sowing would occur.The APSIM model simulates monocrops and rotations in the same way, with both being a succession of soil-centered crops grown over decades years with no soil state initialization in between.The monocrop model has a longer fallow period, which can also be referred to as a rotation of one crop, whereas the wheat-maize double-crop rotation has almost no fallow period.

Data analysis and visualization
Python language was used for data analysis and visualization.The Cartopy library was used for map visualization, and the Matplotlib library was used for bar plots.Statistical analysis was performed using the SciPy and statsmodels libraries.The Mann-Kendall test was used to test the time series trend of the grain yield and climate factors data, while Pearson's correlation was used to test the correlation between the grain yield and climate factors.Two sensitivity indices were used to quantify the contribution of climate factors to grain yield (Wallach et al 2006).The main effect (ME) was used to quantify the effect of individual climate factors on grain yield, and the total effect (TE) was used to quantify the interaction effect of climate factors on grain yield.The ME is calculated by the following formula: where ȲF i represents the mean crop yield across all factors F i , which include soil type, maximum temperature, minimum temperature, rainfall, global solar radiation, rotation and irrigation strategy.We calculated the WUE and irrigation water productivity (IWP) with these equations: where Y denotes the yield, ET denotes the evapotranspiration, Y rainfed denotes the yield under rainfed, Y irrigation denotes the yield under irrigation.

Effects of different crop rotation and water supply strategies on yield, water consumption, WUE, irrigation volume and IWP
The study results indicate that various irrigation practices and crop rotations significantly impact grain yield in the study area (figure 2).On average, the yield of different crop rotation patterns was 8105.5 kg ha −1 The yield difference analysis revealed that irrigation could significantly enhance crop yields in the NCP.
Compared to the rainfed, yields under deficit irrigation were augmented by 28.7%, with the highest increase of 46.6% observed for Monocrop-Wheat.
The least yield increase, however, was noted at 8.8% for the combination of Winter-Wheat & Summer Maize + Monocrop-Maize.Yields under full irrigation were, on average, 66.8% higher than those under rainfed conditions.Monocrop-Wheat reported the highest yield increase of 106.1%, whereas Monocrop-Maize observed the smallest increase at 37.7%.
These results unequivocally suggest that diverse crop rotations respond differently to various irrigation strategies.
Additionally, the irrigation strategy and crop rotation influenced field water consumption, WUE, and IWP (table 3).The maximum average water consumption was 581 mm for 2 year maturity, while the minimum was 277 mm for monocrop wheat.WUE increased with increasing irrigation, with the highest value of 2.7 kg m −3 for 2 year triple cropping and the lowest value of 2.2 kg m −3 for monocrop wheat.The highest WUE for monocrop maize was 3.3 kg m −3 , while the remaining three rotations had a similar WUE of approximately 1.9 kg m −3 .The irrigation strategy also significantly affected WUE, with deficit irrigation being higher than irrigation water efficiency under full irrigation conditions for lower monocrop maize, but lower than irrigation water efficiency under full irrigation for monocrop wheat and 2 year triple crop.However, there was no change in irrigation efficiency under two maturity per year.

Response of yield change trend to different crop rotation and water supply strategies
The Mann-Kendall analysis revealed a statistically significant trend of increasing yields for fully irrigated Monocrop-Wheat over the past 40 years (figure 3).
Conversely, there has been a significant trend of decreasing yields for fully irrigated Monocrop-Maize.
In contrast, the grain yields of the double-cropped Winter-Wheat & Summer Maize (WW&SM) system and the triple-cropped Winter Wheat & Summer Maize + Monocrop-Maize (WW&SM + M) system have shown no significant trends under current climate change conditions.The rate of increase in grain yield for Monocrop-Wheat was 449.0 kg ha −1 per 10 years, while the rate of decrease for Monocrop-Maize was 313.4 kg ha −1 per 10 years.These results suggest that Monocrop-Maize production is struggling to adapt to current climate change trends, whereas Monocrop-Wheat is benefiting from these trends.In contrast, the multi-cropped systems demonstrated resilience to climate change.

Response of the correlation between yield and climate factors to the different crop rotation and water supply strategies
Crop rotation significantly impacted the relationship between yield and radiation and temperature (figure 4).The mean correlation coefficient between yield and radiation was −0.28 for Monocrop-Wheat and 0.22 for Monocrop-Maize, indicating opposing functions of the two crops.The multi-crop rotations were similar to Monocrop-Maize in showing a positive correlation between yield and radiation, with mean correlation coefficients of 0.13 for the WW&SM and 0.10 for the WW&SM + M, and maize system, respectively.The order of correlation strength was Monocrop-Maize > WW&SM + M > WW&SM.
The yield of Monocrop-Maize had a negative correlation with the maximum temperature, with a mean maximum correlation coefficient of 0.19.In contrast, the yield of Monocrop-Wheat was positively correlated with the minimum temperature, with a maximum correlation coefficient of 0.28.The yields of the other crop rotations were less correlated with the maximum temperature.As shown in supplementary figure 3, both minimum and maximum temperatures are trending upwards, while radiation is trending downwards.Based on these trends, it can be inferred that monoculture maize yield will be negatively impacted, while monoculture wheat yield will increase, and the other rotations will be less affected or insensitive.The water supply strategy modifies the relationship between yield and precipitation for the different crop rotations, as well as the relationship between yield and maximum temperature for certain rotations.Irrigation reduces the negative correlation between yield and maximum temperature for Monocrop-Maize and enhances the positive correlation for the other three rotation patterns.Wise use of irrigation will increase the resilience of food production to climate change.

Sensitivity of yield to climatic factors under different crop rotation and water supply strategies
Sensitivity analysis is a technique used to determine how changes in input variables can affect the output of a system.In the context of climate change impact on yield, sensitivity analysis can identify the most important environmental factors, such as temperature, precipitation, and soil properties, for crop yield.By understanding the most influential environmental  factors, decision-makers can develop strategies to mitigate the effects of climate change and increase crop yields.In this study, sensitivity analysis showed that the maximum impact of precipitation on the yield of monocrop wheat was 49.7% and 46.6% under rainfed and deficit irrigation, respectively (figure 5).
Although the effect of precipitation on yield was significantly reduced to 10.2% when sufficient irrigation was applied, the interaction with the other three climatic factors still had the greatest impact on yield (figure 6).The impact of precipitation on yield was weaker for Monocrop-Maize, WW&SM + M,

Discussion
Our study aimed to analyze the effects of climate factors on the spatial and temporal variability of wheat and maize yields in the NCP, a vital grain production region in China where these crops are predominantly cultivated.We investigated the impacts of climate change, crop rotation patterns, and water supply strategies on the yields of wheat and maize in the region.The findings of this study have significant implications for the development of strategies to optimize crop yields and enhance adaptation to climate change in the NCP, contributing to the body of knowledge in the field of agricultural sustainability.

Variations of climate conditions
The varying growing seasons and sowing dates of different crops result in differing climate environments for different crop rotation methods (Ojeda et al 2021).This study shows that different rotations have varying climatic conditions within the study area.Specifically, monocrop wheat exhibits lower levels of precipitation, radiation, and both maximum and minimum temperatures compared to the annual averages.In contrast, monocrop maize has higher levels of precipitation and lower levels of radiation, as well as significantly higher maximum and minimum temperatures.The 2 year triple cropping pattern has intermediate levels of precipitation and radiation, as well as higher maximum and minimum temperatures compared to the annual averages.Finally, the traditional double cropping pattern has the highest utilization rate of precipitation and radiation, as well as temperatures like the annual averages.It is worth noting that the monocrop wheat model indicates that the growing season for wheat is during a period of poor light, temperature, and water conditions, which greatly limit the growth of other crops.On the other hand, the monocrop maize model indicates that the growing season for maize is at a time when light, temperature, and water conditions are favorable, and these conditions are conducive to the growth of a wide range of crops.The traditional double cropping pattern maximizes the utilization of precipitation and radiation, which can optimize grain yields (Xu et al 2021), However, this system may also have the greatest water demand and risk of groundwater depletion (Sun et al 2011).The 2 year triple cropping pattern has intermediate levels of both precipitation and radiation.
The optimal sowing date and density are key agronomic factors influencing wheat yield.Sowing date affects wheat developmental stages like tillering, which in turn impacts final yield.Earlier sowing often increases tillering due to favorable temperatures and adequate moisture during early growth stages (Rasmussen and Thorup-Kristensen 2016).More tillers allow for increased spike number and kernel count per unit area, both drivers of higher grain yield (Tilley et al 2019).However, late-sown wheat experiences reduced tillering from cooler fall temperatures, constraining yield (Yin et al 2020).Sowing density also affects tillering by changing light interception and the availability of nutrients and moisture per plant.Higher densities reduce light penetration into the canopy inside, increasing competition between plants and lowering tiller production.But sparse stands with excess resources cannot capture adequate light.Optimal stand density maximizes fertile tiller number for a given environment and sowing date (Gao et al 2021).Therefore, the ideal wheat sowing density for a specific sowing date balances individual plant tillering capacity with sufficient population for maximum light interception and yield.Later sowing dates require higher seeding rates to compensate for reduced tillering ability (Ren et al 2019).But excessively high densities exacerbate light competition without improving fertile tiller survival.Integrating sowing date, stand density, and tillering responses is necessary to optimize wheat yields.

Yield and WUE under different irrigation strategies and crop rotations
The study results demonstrate that irrigation practices and crop rotations have significant impacts on yield in the study area (figure 2).On average, yields increased with increased irrigation, with the highest yields observed under fully irrigated conditions.Among the different crop rotation patterns, Winter-Wheat & Summer-Maize > Winter-Wheat & Summer Maize + Maize > Monocrop-Maize > Monocrop-Wheat had the highest to lowest yields under the same irrigation conditions.However, it is worth noting that the use of irrigation can also have negative environmental impacts, such as water resource depletion and soil erosion.Decision-makers may need to weigh the potential benefits of increased yields against these negative impacts when deciding on irrigation practices.Additionally, the irrigation strategy and crop rotation influenced field water consumption, with the maximum average water consumption observed for double cropping and the minimum for monocrop wheat.This may also be a factor for farmers to consider when deciding on rotation practices (Yan et al 2020).
The study results show that WUE increases with increasing irrigation, with the highest value observed for 2 year triple cropping and the lowest value for monocrop wheat (table 3).However, the irrigation strategy significantly affected WUE, with deficit irrigation being more efficient than full irrigation for monocrop maize but less efficient than full irrigation for monocrop wheat and 2 year triple crop.This suggests that there may be trade-offs between yield and WUE depending on the irrigation strategy and crop rotation used.Zhao et al (2020) also showed, through meta-analysis, that crop rotation increases crop yields by an average of 20% compared to continuous monoculture.Yu et al (2020) reported that appropriate irrigation practices can improve wheat water-use efficiency by 6.6%, with the best results occurring in areas with less than 200 mm of precipitation and loamy or sandy soil, using border and furrow irrigation methods.These findings suggest that farmers can increase grain yields by carefully considering irrigation practices and crop rotations.
The Mann-Kendall analysis in this study shows a statistically significant trend of increasing yields for fully irrigated Monocrop-Wheat over the past 40 years, with a rate of increase in grain yield of 449.0 kg ha −1 per 10 years.This may be due to the crop benefiting from the current climate change trends.On the other hand, there has been a significant trend of decreasing yields for fully irrigated Monocrop-Maize, with a rate of decrease of 313.4 kg ha −1 per 10 years.These results suggest that Monocrop-Maize may be struggling to adapt to current climate change trends, while Monocrop-Wheat is benefiting from these trends.Xiao et al (2021) simulated results similarly indicated that future climate change would have negative impact on maize yield across all cropping systems but have positive impact on wheat yield under most climate change scenarios.However, it is worth noting that these findings are based on fully irrigated conditions, which are not applicable to all cropping conditions.Additionally, the grain yields of the double-cropped Winter-Wheat & Summer Maize (WW&SM) system and the triplecropped Winter Wheat & Summer Maize + Maize (WW&SM + M) system have shown no significant trends under current climate change conditions.This suggests that these multi-cropped systems may be more resilient to climate change compared to the monocrop systems.Maintaining crop diversity is important for maintaining ecosystem functions related to food production and for supporting crop yields, particularly in the context of ongoing land-use change (Dainese et al 2019).These findings highlight the importance of considering both irrigation practices and crop rotations when analyzing the impacts of climate change on crop yields.While Monocrop-Wheat may be benefiting from current trends, other crops and cropping systems may be more vulnerable or resilient to these trends.

Drivers of yield variance
The study results indicate that crop rotation significantly impacts the relationship between yield and radiation and temperature (figure 4).Monocrop-Wheat had a mean correlation coefficient of −0.28 between yield and radiation, while Monocrop-Maize had a coefficient of 0.22, indicating opposing functions in terms of their relationship with radiation.The multi-crop rotations, including the Winter-Wheat & Summer Maize (WW&SM) and the Winter Wheat & Summer Maize + Maize (WW&SM + M) systems, had a positive correlation between yield and radiation, similar to Monocrop-Maize.In terms of correlation strength, the order was Monocrop-Maize > WW&SM + M > WW&SM.Regarding temperature, Monocrop-Maize had a negative correlation with the maximum temperature, with a mean maximum correlation coefficient of 0.19.In contrast, Monocrop-Wheat had a positive correlation with the minimum temperature, with a maximum correlation coefficient of 0.28.The yields of the other crop rotations were less correlated with the maximum temperature (Bao et al 2022).Supplementary figure 3 shows that both minimum and maximum temperatures are trending upwards, while radiation is trending downwards.Based on these trends, it can be inferred that the yield of monoculture maize will be negatively impacted, while the yield of monoculture wheat may be positively impacted.The other crop rotations may be less affected or insensitive to these trends.Observational data analysis reveals that maize yield has increased over the past 30 years, while radiation levels have decreased (Deng et al 2020).This is likely due to improved water and fertilizer management and advances in cultivars (Ren et al 2021).It is worth noting that the water supply strategy can modify the relationship between yield and precipitation for the different crop rotations, as well as the relationship between yield and maximum temperature for certain rotations.Irrigation not only fulfills crop water demand but also mitigates crop heat stress (Tack et al 2017).Irrigation can reduce the negative correlation between yield and maximum temperature for Monocrop-Maize and enhance the positive correlation for the other three rotation patterns.Li et al (2020) found that 16% of the increased yield from irrigation is due to the cooling effect.These results suggest that judicious irrigation use may bolster the resilience of food production to climate change by alleviating the adverse effects of rising temperatures on some crops.
Sensitivity analysis is a useful tool for understanding how changes in certain input variables can affect the output of a system, such as crop yield.In this study, sensitivity analysis was used to identify which environmental factors, including temperature, precipitation, and soil properties, had the greatest impact on crop yield.The results showed that the maximum impact of precipitation on the yield of monocrop wheat was 49.7% and 46.6% under rainfed and deficit irrigation, respectively (figure 5).Although sufficient irrigation reduced the effect of precipitation on yield to 10.2%, the interaction with the other three climatic factors still had the greatest impact on yield.Conversely, the impact of precipitation on yield was weaker for Monocrop-Maize, Winter-Wheat & Summer Maize + Maize (WW&SM + M), and Winter-Wheat & Summer Maize (WW&SM), at 22.1%, 17.5%, and 19.2%, respectively.This suggests that maize may be more sensitive to changes in precipitation than wheat, and that multi-crop systems may be less sensitive to changes in precipitation than single-crop systems.Although there has not been a notable trend in precipitation in recent decades (supplementary figure 3), research suggests that precipitation is likely to fluctuate more significantly in the future as a result of climate change (Yan et al 2022).Under water stress, the average effect of climatic factors on yield was 65.8%, with Monocrop-Maize having the strongest impact, followed by Monocrop-Wheat, WW&SM + M, and WW&SM.This suggests that wheat is less affected by climate than maize, and multi-crop systems are less affected by climate than single-crop systems.Utilizing crop rotations with diverse species takes advantage of the unique attributes of each crop.For instance, winter wheat can benefit from warmer winters to bolster biomass accumulation (He et al 2020a), counterbalancing maize yield declines under hotter temperatures.
Additionally, wheat employs the C 3 photosynthetic pathway, whereas maize utilizes the C 4 pathway, resulting in distinct responses to climatic factors.It has been proposed that winter wheat, as a C 3 plant, would likely see greater benefits in a future scenario involving an increase in atmospheric CO 2 levels, while the C 4 crop maize would likely see fewer bene Theoretically, maize may benefit more from increased temperatures than wheat.Nevertheless, considering the wheat-growing environment in the NCP, the average wheat season temperature is a mere 8.3 • C, lower than the optimal growth temperature.Consequently, higher future temperatures should prove more favorable for wheat cultivation.The results of the present study show similar findings.Seasonal maximum temperatures for maize may exceed 35 • C, inducing heat stress and potentially reducing yields (Sánchez et al 2014).C 4 plants generally possess greater water-use efficiency than C 3 plants, rendering them more resilient to drought conditions.In areas where water scarcity is anticipated to intensify due to climate change, C 4 crops may outperform C 3 crops.However, precipitation is expected to increase in North China in the future.C 3 plants exhibit a more pronounced stimulation of photosynthesis in response to elevated CO 2 levels compared to C 4 plants, suggesting that increasing atmospheric CO 2 may benefit wheat more than maize.Synthesizing the above factors, it is likely that C 3 plants will gain more benefits from future climate change trends.
Irrigation can mitigate the impact of climate on crop yield.Full irrigation reduced the average effect of climatic factors on yield to 44.1%, indicating a 33.0%reduction in the sensitivity of grain production to climate.Zaveri and Lobell (2019) reported that irrigation has played a significant role in the growth of wheat as a major crop in India and is expected to become even more important as a way to adapt to climate change.Those results showed national wheat yields in the 2000s to be 13% higher than the yields projected without irrigation expansion since 1970.Furthermore, irrigated wheat exhibited approximately 25% less sensitivity to heat relative to purely rainfed conditions.These findings indicate irrigation could serve as an effective strategy to improve crop yields under climate change.However, this will significantly increase the need for irrigation from the point of view of water use (Tijjani et al 2022).

Conclusion
This study examined the influence of climate factors on the spatial and temporal variability of wheat and maize yields in the NCP.Results demonstrate significant spatial and temporal variability in climatic conditions within the study area, with decreasing radiation levels and increasing maximum and minimum temperatures.The highest levels of precipitation are in the north and southeast, while the eastern region experiences higher levels of radiation, and the western and eastern edges tend to have higher temperatures at the same latitude compared to the middle.The study also found that irrigation practices and crop rotations have significant impacts on yield in the study area, with yields increasing on average with increased irrigation, and the highest yields observed under fully irrigated conditions.Additionally, different crop rotation patterns affect crop yields, with the monocrop wheat model having the lowest levels of precipitation, radiation, and temperature, the monocrop maize model having the highest levels of precipitation and maximum temperature, and the traditional double cropping pattern having the highest utilization levels of both precipitation and radiation, which may be optimal for grain production but also have the highest water usage and risk of groundwater depletion.These findings have important implications for the development of strategies to optimize crop yields and adapt to climate change in the NCP while considering the trade-offs.

Data availability statements
The data cannot be made publicly available upon publication because no suitable repository exists for hosting data in this field of study.The data that support the findings of this study are available upon reasonable request from the authors.

Figure 1 .
Figure 1.Map of the study region and the spatial distribution of cropping systems within the area.
ȲF −i represents the mean crop yield across all factors except F i .The ME captures the proportion of crop yield variance explained by each individual factor without considering interactions among them.The value of ME = 1 indicates that the assessed factors fully explain the crop yield variance, while the value less than 1 suggests the presence of residuals, indicating the need for additional factors to explain the variance.The residuals refer to the portion of change in the dependent variable y that is caused by factors other than the effect of the independent variable x.In this study, the residual represents the portion of change in crop yield caused by non-climatic factors.The TE takes into account the interactions of a particular factor with other factors, thus, TE does not include residuals.The mean of ME and TE was calculated across the years 1980-2018.

Figure 2 .
Figure 2. Annual average grain yield of different crop rotations under three water supply strategies.Crop rotations are displayed in rows.The first row is for wheat monoculture (Monocrop Wheat); the second row is for maize monoculture (Monocrop Maize); the third row is for the three-harvest in 2 year pattern, with winter wheat and summer maize in the first year, plus maize monoculture implemented in the second year (Winter Wheat & Summer Maize + Monocrop Maize); and the fourth row is for a double cropping winter wheat and summer maize pattern (Winter Wheat & Summer Maize).Water supply strategies (Rainfed, Deficit irrigation and Full irrigation) are displayed in the first three columns.The last two columns display the difference between deficit irrigation and rainfed (Deficit irrigation-Rainfed), as well as between full irrigation and rainfed (Full irrigation-Rainfed), respectively.

Figure 3 .
Figure 3. Yield change trend of different crop rotation and water supply strategies (ns indicates no significant trend, while * denotes a significant trend at the p < 0.05 level.).

Figure 4 .
Figure 4.The correlation between yield and climate factors under different crop rotation and water supply strategies.
and WW&SM, at 22.1%, 17.5%, and 19.2%, respectively.The average effect of climatic factors on yield under water stress was 65.8%, and the order of the strength of the climatic factors on the yield of different rotations was Monocrop-Maize > Monocrop-Wheat > WW&SM + M > WW&SM, indicating that wheat was less affected by climate than maize, and multi-crop systems were less affected by climate than single-crop systems.The average effect of climatic factors on yield was reduced to 44.1% under full irrigation, indicating that irrigation reduced the sensitivity of grain production to climate by 33.0%.

Figure 5 .
Figure 5.The main effect (ME) capturing the proportion of crop yield variance explained by each individual factor without considering interactions among them.Lowercase letter (a) stands for Monocrop-Wheat, (b) for Monocrop-Maize, (c) for Winter Wheat & Summer Maize + Monocrop-Maize, (d) for Winter Wheat & Summer Maize.

Figure 6 .
Figure 6.The total effect (TE) taking into account the interactions of a given factor with other factors.Lowercase letter (a) stands for Monocrop-Wheat, (b) for Monocrop-Maize, (c) for Winter Wheat & Summer Maize + Monocrop-Maize, (d) for Winter Wheat & Summer Maize.

Table 1 .
Irrigation threshold settings under different crop rotation patterns.
The climate data is from the China meteorological forcing dataset (He et al 2020b), covering the period from 1979 to 2018, while the soil data is sampled in 2010 and obtained from the SoilGrids database (Hengl et al 2017).Saturated hydraulic conduct, Saturation water content (SAT), DUL and lower limit (LL15) is calculated with Saxton or Rawls method: If soil organic carbon stock (SOCS) values are missing for any of the layers, then Saxton method (Saxton et al 1986) is used, if SOCS values are known for all layers of the soil, then Rawls method (Rawls et al 1982) is used.

Table 2 .
Configuration of crop model parameters.

Table 3 .
Evapotranspiration (ET), irrigation amount, water use efficiency (WUE) and irrigation water productivity (IWP) of different crop rotations under different water supply strategies.