Climate Change Impacts on Agroecosystems in China: Processes, Mechanisms and Prospects

Building a more resilient response system to climate change for sustainable development and reducing uncertainty in China’s food markets, requires access to historical research gaps and mapping future research progress for decision making. However, the lack of quantitative and objective analyses to ensure the stability and development of agroecosystems increases the complexity of agro-climatic mechanisms, which leads to uncertainty and undesirable consequences. In this paper, we review the characteristics of climate change in China (1951–2020), reveal the mechanisms of agroecosystem structure in response to climate, and identify challenges and opportunities for future efforts in the context of research progress. The aim is to improve the scientific validity and relevance of future research by clarifying agro-climatic response mechanisms. The results show that surface temperature, precipitation, and frequency of extreme weather events have increased to varying degrees in major agricultural regions of China in 1951–2020. And they have strong geographic variation, which has resulted in droughts in the north and floods in the south. Moreover, climate change has complicated the mechanisms of soil moisture, Net Primary Productivity (NPP), soil carbon pool, and crop pest structure in agroecosystems. This lends to a reduction in soil water holding capacity, NPP, soil carbon content, and the number of natural enemies of diseases and insects, which in turn affects crop yields. However, human interventions can mitigate the deterioration of these factors. We have also realized that the methodology and theory of historical research poses a great challenge to future agroecosystem. Historical and projected climate trends identified current gaps in interdisciplinary integration and multidisciplinary research required to manage diverse spatio-temporal climate change impacts on agroecosystems. Future efforts should highlight integrated management and decision making, multidisciplinary big data coupling, and numerical simulations to ensure sustainable agricultural development, ecological security, and food security in China.


Introduction
The six climate assessment reports issued by the Intergovernmental Panel on Climate Change (IPCC) all in-dicate that global climate change will continue in the foreseeable future, and the contributing factors will gradually become more diverse and complex, such as the impact of human activities (Qin and Thomas, 2014;Fan et al., 2021). Climate changes in temperature, precipitation patterns, and the frequency of extreme weather events, affect the material cycle, energy exchange, and biodiversity in ecosystems (Ding, 1989). Furthermore, clarifying the response of terrestrial ecosystems to climate change has been a core scientific question in global change research (Feng et al., 2016). Therefore, conducting a comprehensive review of current research not only helps to identify the limitations of past studies on land-climate interactions and formulate new research questions, but also provides a clear direction for future research to address these issues. This is an important step towards a scientific understanding of how terrestrial ecosystems respond to climate change, and is also of great significance for achieving sustainable socioeconomic development in the region.
As one of the terrestrial ecosystems, agroecosystems shoulder the crucial responsibility of global food security. And they are subjected to the highest degree of human intervention, rendering them vulnerable to the complexities and fluctuations of climate change (Fuhrer, 2003;Guo and Zhou, 2007). China, the world's thirdlargest cultivable land, has the potential to feed a substantial portion of the global population owing to its diverse climate and environment that can support the growth of most crops under suitable conditions. However, the complexity of its farmland distribution and types has created management challenges. The region's multifaceted geography has resulted in a complex climate, rendering the response of the regional agricultural ecosystem to climate change a daunting task. Furthermore, extreme weather events triggered by climate change have substantially affected the energy and material exchanges within the ecosystem that regulate the agricultural food production and the ecosystem's climate (Kimball, 1983;Cure and Acock, 1986). Agroecosystems are crucial to sustainably produce food, for biodiversity, the carbon cycle, and for soil moisture. However, they are prone to disturbances and are extremely vulnerable to climate change. Current research has tended to focus on the response of agroecosystems to changes in climate factors, and has rarely provided a systematic overview of agro-climatic response mechanisms. This study explores the characteristics of climate change and the structural mechanisms of agroecosystems to fill this gap.
Food security risk has been demonstrated to arise from climate change variations in rainfall, coupled with increases in CO 2 concentration, which can negatively impact crop growth and food production (Rosenberg, 1981;Wang et al., 2022). However, they are highly susceptible to overlooking the impact of human activities during research (Guo, 2015;Zhou et al., 2016). In the systematic analysis of irrigation efficiency in the arid northwest region of China, Fu et al. (2022) found that the farmland ecosystem in the Tarim River Basin is expanding and becoming less dependent on precipitation due to improvements in irrigation measures. It emphasizes the positive impact of agricultural practices on ecosystem sustainability and resilience, providing insight into the potential for managing and mitigating the impact of climate change in vulnerable regions. Thus, human activities can both affect crop growth by influencing climate change and reduce the response to climate by changing agricultural management practices. Based on current research, it has been widely recognized that the response and feedback processes between land and climate systems are critical for understanding the impacts of climate change and less comprehensive evaluation of specific ecosystems such as agroecosystems (Walther, 2010;Grimm et al., 2013). Given the growing significance of human activities in both climate change and its mitigation, this study includes an examination of ways to reduce the impact of climate change on agroecosystems. This is an aspect that has been overlooked by most current research.
Therefore, in this study, we review the research during 1951-2020 on climate change and agroecosystems in China (Hong Kong, Macao and Taiwan of China are not included in this study), focusing on climate forcing, human interventions, mechanisms of impact processes, and agroecosystem components affecting food production. The main objectives are to: 1) review the characteristics of climate change (temperature, precipitation, and extreme weather events) in China; 2) assess the material cycles and mechanisms affecting the state of agroecosystem components in the context of climate change and human interventions; and 3) make recommendations for research needed to effectively and sustainably manage current and future agroecosystems in response to current research questions.

Methodology
This paper presents a comprehensive analysis of the cur-rent state of knowledge on the topic using a combination of literature review and data analysis approaches. To ensure the comprehensiveness of our review, we conducted a thorough search of relevant keywords and phrases on databases such as PubMed, Web of Science, and Scopus. We meticulously reviewed and summarized the main findings and conclusions of each article, and analyzed climate change characteristics and trends using Chinese meteorological statistics. The methodology is rigorous and systematic, providing an objective overview of the current state of knowledge in the field.
The study is structured into four main sections (Fig. 1). Firstly, a comprehensive summary of surface air temperature, precipitation, and frequency of extreme climate events in China from 1951 to 2020 was provided. This includes an overview of the overall trends and spatial distribution, and aims to elaborate on the changes of climate factors in the major agricultural regions of China. Surface air temperature, precipitation, and extreme weather events were selected due to their close relationship to agroecosystems. Secondly, we analyzed the impact of temperature and precipitation on soil moisture, Net Primary Productivity (NPP), soil carbon pool, pests and diseases, to understand the response mechanisms of these factors to climate change. Thirdly, we examined the impact of these factors on crop production in the context of climate change, and evaluated the effectiveness of human intervention in reducing cli-mate stress. Finally, based on the research trends and methodological theories in each section, we propose directions for future research efforts.

Characteristics of Climate Change in China During 1951-2020
According to the IPCC's Sixth Assessment Report (AR6), global surface temperature in the first two decades of the 21st century was 0.99°C warmer than in the period from 1850 to 1900, and in the period from 2011 to 2020, it was 1.09°C warmer. Since the 1950s, increases were recorded in worldwide average precipitation, heat days, heat nights, heat waves, and increased frequency and intensity in extreme climate events (Legg, 2021). Since the middle 20th century, temperature, precipitation, and extreme agricultural climate in China have exceeded global averages (Ding et al., 2006).
Human activities contribute significantly more to surface temperature than natural forcings. From 2010 to 2019, the range of surface temperature warming caused by human activities was 0.8°C-1.3°C, while the range of natural forcings was only -0.1°C to 0.1°C. In addition, human-induced greenhouse gases have been strongly proven to be the main driving factor affecting temperature and precipitation changes (Eyring et al., 2021). A large amount of research also shows that human activities, such as grazing, land cover change, and fossil fuel exploitation, further alter the biogeochemical cycle of terrestrial ecosystems, thus changing the NPP of surface vegetation and increasing research uncertainties (Du et al., 2004;Li and Zhao, 2012;Chen et al., 2013;Ye et al., 2013;Ge et al., 2021).Therefore, under the dual driving forces of natural and human activities, it is crucial to have an accurate understanding of how surface temperature, precipitation, and extreme agricultural climate events are changing, and how they impact agroecosystems.

Surface temperature change in China
Since the 1900s, annual average surface temperature across China warmed at a rate of 0.15°C/10yr, although from 1951 to 2020, it warmed by 0.26°C/10yr, with considerable interdecadal oscillations Xu et al., 2018). Since the turn of the century, annual average surface temperature increase slowed, but summer and winter temperatures tended towards hotter  Fig. 1 Diagram illustrating the main structure of the paper. NPP is Net Primary Productivity summers and colder winters (Li et al., 2015). From 2012 surface temperature warmed due to the influence of the monsoon climate (Zhai et al., 2016).
Regional surface temperatures vary significantly. In the eastern and western regions of China, temperatures seldom deviate from national averages, but it rise faster in the northern than southern regions (Ding et al., 2007;Li et al., 2017). Annual average surface temperature in eight regions of China (North China, Northeast China, East China, Central China, South China, Southwest China, Northwest China, and Qinghai-Tibet Plateau) (Song, 2021) consistently increased from 1961 to 2020, with the fastest warming rate of 0.36°C/10yr in Qinghai-Tibet Plateau and the slowest of 0.17°C/10yr in Southwest China. The warming rate exceeds 0.4°C/10yr in the north-central Qinghai-Tibet Plateau, northwestern Heilongjiang, central and northeastern Inner Mongolia, and parts of Heilongjiang (Song, 2021).
Maximum and minimum surface air temperatures will keep rising until the middle and end of the 21st century Xie et al., 2022). From 1951 to 2020, China's annual average surface maximum temperature warmed at rates of 0.20°C/10yr and the minimum warmed at 0.34°C/10yr, with a particularly pronounced upward trend from 1970 to 2020 (Song, 2021).

Precipitation changes in China
Dynamical changes in the land-atmosphere system's energy balance and water cycle have also caused significant precipitation fluctuations (Yin et al., 2021). Since the turn of the 20th century, an overall increasing trend of extremes was accompanied by interannual and interdecadal variability, particularly in the eastern China (Lu et al., 2014). According to Song (2021), there is no discernible trend change in average annual precipitation in China from 1991 to 2020, but from 1961 to 2020, the trend increased by 5.1 mm/10yr, despite interdecadal variations. Overall, precipitation is primarily high before the 21st century and as air temperature began to rise again in 2012, precipitation began to increase steadily (Song, 2021).
Regional and seasonal differences in China's precipitation arise from dramatic elevation changes, high terrain in the west and low topography in the east, the vast latitudinal range between north and south, and its placement relative to the East Asian monsoon. Across the eight regions of China (defined previously), average an-nual precipitation trends between 1961 and 2020, increased in Qinghai-Tibet Plateau (at 10.4 mm/10yr), Southwest China decreased, and the other regions showed no appreciable changes. Greatest significant increases in precipitation occurred in the eastern Jiangnan, the north-central Qinghai-Tibet Plateau, and the northern and western Xinjiang (Song, 2021). Seasonal changes in precipitation during 1961-2020 are not apparent, apart from minor increases in spring and summer, and a minor decline in autumn and winter, but most apparent in the fall season .
Precipitation days in China since 1961 decreased significantly by 1.9 d/10yr, but the number of days with heavy rainfall rose by 4.0%/10yr (Song, 2021). Between 1960 and 2013, the frequency of days with light and medium rainfall decreased by 3.85%/10yr and 1.17%/10yr, respectively, and heavy rainfall days increased; these trends were closely correlated to annual rainfall days (Ma and Zhou, 2015).

Changes in extreme weather events in China
In China, global warming has increased the frequency and intensity of extreme climate, including extreme hightemperature, low-temperature, extreme precipitation, and drought, with significant impacts on energy and material exchanges in agroecosystems (Zhou and Qian, 2021). Low-temperature extremes decreased significantly in China from 1961 to 2020, while extreme hightemperatures increased since the 1990s. Extreme precipitation increased, and the frequency of regional drought events increased slightly, with clear regional and interdecadal variations (Song, 2021). For instance, the massive floods in 1998 submerged 2.1 × 10 7 ha of land in the Yangtze River Basin, resulting in significant economic losses (Zong and Chen, 2000).
Despite variations across China, extreme heat events generally rose in frequency, with an average increase of 4.4 time/10yr in extreme heat episodes (Qin and Zhai, 2021). Locations of extreme temperatures rose significantly from 2000, and the total between 2001 to 2010 surpassed those from the 1970s and 1980s. Spatially, the frequency of extreme heat events was fairly uniform in the 1980s across the eastern Northwest China, the southern North China in the 1990s, and predominately in the southern region, northern China, and Sichuan Basin after the turn of the century. Analyses of heat event clusters show impact intensities growing in tandem with changes in the statistical and regional incidences of extreme heat events (Qin and Zhai, 2021). Five of the top ten cluster heat events across China between 1951 and 2012, occurred since the turn of the century. Cluster heat events after 2000 were the most significant by frequency, duration, and impact range (Kuang et al., 2014). The biggest heat wave to hit the southern China since 1951 occurred in the summer of 2013, lasting 62 d and affecting 19 provinces, including the Jiangnan, Jianghuai, Jianghan, and Chongqing regions. In addition, 132 meteorological stations recorded maximum temperatures above 40°C (Qin and Zhai, 2021).
According to Qin and Zhai (2021), extreme low-temperature incidents decreased by 9.9 time/10yr between 1961 and 2015, mostly in the 1960s and 1970s. Thereafter, frequency significantly decreased, but the intensity rose. Record-breaking extreme low-temperature incidents mostly occurred in North and Southwest China in the 1980s, the Loop and Southern areas in the 1990s, North and Northeast China in the 21st century.
Precipitation depends on the water-holding capacity of the atmosphere, which in turn is impacted by global warming. Extreme-precipitation frequency increased dramatically across China between 1961 and 2020, increasing by 18 station/10yr (Song, 2021). In the western arid zone, western eastern arid zone, northern Southwest China, southern Central China, and eastern South China, the frequency increased significantly between 1951 and 2014, but decreased in the eastern arid zone, southern Northeast China, northern North China, and southern Southwest China (Gu et al., 2016). From 1961 to 2016, there was a significant increase in the southern China, particularly in the northwest, middle, and lower reaches of the Yangtze River, the southern China, and some areas of Southwest China. In contrast, extreme precipitation decreased in the northern China, some areas of Northeast China, and some areas of Southwest China (He and Zhai, 2018;Wang et al., 2014). Extreme precipitation records increased by 3.7 station/10yr, with an increasing trend from 1961 to 2018, so intensity and spatial scale are increasing. Also, daily precipitation increased nationally, reaching 2-10 mm/10yr in Jiangnan, the central and western South China, and Hainan (Qin and Zhai, 2021). It is worth noting that research has shown a significant increase in the trend of heavy rainfall in China, especially in the middle and lower reaches of the Yangtze River (Guan et al., 2017), which means that the probability of flooding has changed. In addition, according to statistical data (https://data. stats.gov.cn, 1980-2020), the gap between the affected areas by floods and droughts in China is gradually narrowing and showing a significant increasing trend.
Droughts have become more frequent, more widespread, and more severe in China. Since the turn of the 20th century, 185 regional droughts occurred between 1961 and 2020, including 16 extremely severe, 39 severely severe, and 77 moderately severe events. Interdecadal variations included a phase bias from 2003 to 2008, a phased reduction after 2009, and many droughts from the late 1970s to the 1980s, compared to a low number in the 1990s (Song, 2021). Areas most affected were south of the Yellow River Basin and north of the Yangtze River, and the frequency of droughts was higher in the north than the south, and in the east than the west (Han et al., 2019). Droughts are often accompanied by other events. A study on heatwaves from 1960 to 2006 found that, except for the central of China, the frequency of heatwave events increased in most areas, with the greatest increase observed in the Qinghai-Tibet Plateau and the southern coast (Piao et al., 2010). In addition to heat waves, dry-hot wind is extremely prone to occur in North China. Using the Weather Research & Forecasting model (WRF) and Peking University Land Model (PKULM) to reproduce the formation mechanism and crop response mechanism of dry-hot wind, it is found that its formation is closely related to surface temperature and solar radiation, and it is easy to occur under drought and high temperature weather conditions .
Climate change is mainly manifested in surface temperature and precipitation, and is also the main driving factor for changes in various factors of agroecosystems. Warmer temperatures not only affect the water and heat exchange within a geographical area, but also stress the chlorophyll content in crops, reducing photosynthesis (Piao et al., 2010). Agroecosystems are used to ensure human food security, and crop growth depends primarily on water, temperature, and solar radiation, meaning that agroecosystems are closely related to climate change (Porter and Semenov, 2005). Due to its vast territory, different agricultural regions in China have different response characteristics to climate change (Fig. 2). Temperature and precipitation are important factors for judging the types of extreme climate events in a region.
Droughts generally occur in the north, while in the south, floods are more frequent, which also shows changes in the precipitation pattern. The northwest agricultural region is high in altitude with abundant thermal energy resources, but relies heavily on precipitation . In the northeastern agricultural region, the impact of surface temperature increase on wheat yield is greater than that of rice, indicating that climate change is not friendly to rain-fed crops (Tao et al., 2008). Moreover, the trend analysis of precipitation in the northern and southern China shows that the gap between the precipitation trends in the two regions is gradually widening, with the southern agricultural region having a better climate advantage (Piao et al., 2010). However, in recent years, the frequency of extreme weather events in the southern region has been increasing, especially in the middle and lower reaches of the Yangtze River, where extreme events such as heavy rain, heat waves, dry-hot wind, and droughts have become more frequent (Zhai et al., 2005). In the face of the characteristics of various agricultural regions, there is currently no welldeveloped comment on the response mechanisms of agroecosystems to climate change, and most studies fo-cus on individual regions.

Impacts of Climate Change on Agroecosystems
Climate variations in surface temperature, precipitation, and extreme weather events have flow-on impacts to terrestrial ecosystems. These include changes in crop phenology, soil water content, surface carbon cycling, and surface vegetation respiration. The black-soil region of the northeastern China, which accounts for 30% of the country's grain production and 12.9% of China's land area, has experienced changes in plant growth status, soil microorganisms, soil freeze-thaw environment, soil carbon, and nitrogen cycle (Fan et al., 2009;Liu et al., 2018;Hou et al., 2020). Four crucial characteristics of agroecosystems-soil moisture, NPP, soil carbon pool, and pests and diseases (Fig. 3)-should be investigated to increase farmland biodiversity, enhance land use, and safeguard the red line of arable land. Additionally, the impacts of climate change on each of these four factors have the potential to result in immeasurable losses in food production. Therefore, this section explores the effects of including surface temperature and precipitation on food production as well as the extreme weather events based on the structural mechanisms of soil moisture, NPP, soil carbon pool and pests and diseases. And to explore how human activities intervene in agroecosystems in response to climate stress, we have also explored them. However, human activities are highly subjective in nature resulting in a lack of in-depth studies. In addition, gaps in historical research are summarized in relation to the mechanisms of the major components of agroecosystems, and suggestions for future research are provided in the next section.

Impact on soil moisture in agroecosystems
Soil moisture is crucial for crop growth and it is a key ecological indicator of agroecosystem response to climatic conditions, and a crucial component of the water cycle and energy balance in ecosystems (McColl et al., 2017). During the 21st century, research and monitoring technologies, significantly improved soil moisture and climate change monitoring (Yang et al., 2004;Hou and Wulanbater, 2006;Yang, 2009). Previous studies found that surface temperature and precipitation were the most important drivers of soil moisture changes. Anticipated global warming over the next 50 yr will alter ecological and material cycles of agroecosystems through the relationship, illustrated in Fig. 3, between temperature, soil moisture pathways and evaporation (JIn et al., 2019). Research on soils in the Loess Plateau, under the same conditions of soil water content and water absorption, suggest that an increase in temperature will increase soil water potential, and decrease soil water retention capacity; processes involved ( Fig. 3) include soil's pore structure, water surface tension, and density, which affect the soil's water diffusivity, saturated hydraulic conductivity, and unsaturated hydraulic conductivity (Gao and Shao, 2011). Temperature increase affects water evaporation, soil moisture evaporation, and changes soil physical and chemical properties, which then directly and indirectly reduces soil moisture (Fig. 3).
Precipitation increases surface moisture (Zhang et al., 2020b), and a study in the northeastern black soil region shows soil moisture content after the growth season was positively related to rainfall during the growth period (Zou et al., 2011). In areas near rivers, soil moisture is also influenced by river runoff. However, current research has shown that in the major river regions of the northern China, human water intake from rivers has led to a decrease in river flow. This impact is more significant in areas that heavily rely on soil moisture in arid and semi-arid regions (Ren et al., 2002). Topography and environmental factors cause variations in the inverse relationship between precipitation and surface temperature, and vice versa for soil moisture. In North China, soil moisture content at various depths showed a decreasing trend from 2013 to 2019 despite relatively stable precipitation and rising surface temperatures. The correlation was stronger at 10 cm, 20 cm, and 50 cm soil depths than at 100 cm depth . Wang et al. (2009) analyzed data from agricultural and pastoral meteorological stations in Hulun Buir City from 1988 to 2007 and found an insignificant upward trend in annual mean temperature, and a significant decrease in annual precipitation (reduction of 8.275-10.347 mm/yr). Soil moisture content within 0-50 cm decreased sharply at a rate of 3.816-0.723 mm/yr. Extensive areas of farmland experienced soil drought. Additionally, seasonal changes in the temperature-precipitation effect on soil moisture were stronger in summer and autumn, and lower in winter and spring. Soil moisture significantly impacts crop growth, soil salinity, and grain yield in agroecosystems. Reductions in soil moisture have led to ecologically-degraded primary and secondary saline conditions in six provinces/ autonomous regions in China, namely Shaanxi, Gansu, Ningxia, Qinghai, Xinjiang, and Inner Mongolia, representing 69.0% and 9.4% respectively of the country's primary and secondary saline land (Qin et al., 2002). Monitoring of soil moisture and maize seedling growth in Jilin Province during the spring of 2010 and 2011 revealed that for every 1.0% decrease in soil moisture, the relative plant height, relative dry weight, and relative leaf age of seedlings decreased by 5.5%, 5.6%, and 11.0%, respectively. Also, different effects on maize growth were observed with different soil moisture levels (Ma et al., 2014). Under similar other conditions, soil moisture increases of 16.523%-34.533% significantly increased soybean production in Heilongjiang Province from 1988 to .
In the face of the forcing factors of soil moisture, humans have employed methods such as irrigation, artificial rainfall, and greenhouse planting to mitigate the impact of agroecosystems on climate change, but the cost of these methods is enormous (Fischer et al., 2007). Irrigation, by directly replenishing soil moisture, is the primary means of preventing drought and ensuring food production, and improving irrigation methods can reduce agriculture's dependence on precipitation and expand agricultural production areas (Fu et al., 2022). However, facing global warming, most studies have only investigated single factors that affect soil moisture, and lack exploration of the coupling and interdependence among multiple factors. In addition, in the face of extreme weather events and their impact on soil mois-ture, China's agriculture has not yet developed a targeted and sound protection system.

Impact on NPP in agroecosystems
Gross Primary Productivity (GPP) is the total organic matter assimilated by plants through photosynthesis utilizing solar energy per unit of time and space, whereas NPP is the total organic matter that is still present after the plants have consumed it (Zhu et al., 2007b). NPP is crucial in controlling biological processes in ecosystems. It affects carbon cycles, ecosystem productivity, carbon sources and sinks (Field et al., 1998). The total NPP of terrestrial vegetation in China increased by 0.76 Pg C at the end of the 20th century, with climate change-related increases accounting for 0.36 Pg C and other factors for 0.40 Pg C respectively and contributing 48.1% and 51.9% to the overall NPP growth trend (Zhu et al., 2007a).
NPP in agroecosystems are hampered by temperature above or below thresholds for plant photosynthesis, which can result in plant mortality. As shown in Fig. 3, temperature affects the surface accumulation temperature that alters activities of photosynthetic enzymes and photosynthesis. When plants lack water, their stomata close, reducing water evaporation and decreasing CO 2 inhalation, resulting in a lack of carbon raw material and a decrease in the rate of plant photosynthesis. Similarly, when soil moisture levels are excessive, roots become oxygen deficient and plant growth is stunted (Zou, 2014). As a result, temperature and precipitation are only advantageous for plant NPP if they fall within the proper range. Also, atmospheric CO 2 had a positive catalytic effect on C 3 participating in plant photosynthesis, which in turn affects plant organic matter production (Huang, 2003). In 1998, empirical relationships between CO 2 and NPP demonstrated that NPP could only rise by 6.0% (219 Tg C/yr) when CO 2 concentration reached 519 ppmv (Xiao et al., 1998). Models of NPP of agricultural vegetation in China by Wang et al. (2006b) discovered that elevated CO 2 had a catalytic rising-trend effect on NPP, increasing it from 338.7 μmol/mol to 369.5 μmol/mol from 1980 to 2000.
Studies of NPP response to climate change, tend to combine effects of many elements, rather than study individual climate effects, such as temperature, precipitation, and light. For example, experiments on NPP in Shanghai farmland showed that from 1961 to 2006, average annual NPP increased by 64.37 g/m 2 under rising temperature and precipitation of 0.053°C/yr and 3.997 mm/yr, respectively, and its average growth rate was 1.43 g/(m 2 ·yr). Substantial positive correlations were found between temperature, precipitation and NPP. Additionally, plant NPP reacted differently in various places to temperature and precipitation . In the northeastern China's farmland growingseason NPP, temperature and precipitation from 2000 to 2006 showed a negative correlation of NPP with temperature -0.13, and a positive correlation with precipitation of 0.62 (P < 0.05) (Zhu et al., 2010). While these studies have verified that factors such as temperature and precipitation have a significant impact on NPP, they tend to focus on single-factor research, and their verification methods do not fully simulate the relationship between NPP and climate. The use of relevant models can better compensate for these deficiencies. For example, using the Daily Century model to simulate the yield and NPP of major crops in the United States, the results show that using the model for prediction is more accurate than most published results (Zhang et al., 2020a). However, research on model coupling in China is not yet fully developed.

Impact on soil carbon pools in agroecosystems
Soil is a vast carbon reservoir that not only supplies nutrients for plant growth and development, but also can sequester carbon. Thus, soil condition determines national food security, ecosystem balance, the global carbon cycle, and mitigates climate change. Key mechanisms involve the soil carbon pool controlling plant growth and development, effects on physical and chemical characteristics, and biodiversity. Soil carbon stores are however extremely vulnerable to exogenous factors like burning straw and artificial fertilizers. As well, climatic change can alter NPP, external carbon input, and organic carbon decomposition, which in turn alters carbon cycling (Ma et al., 2019).
While both organic and inorganic carbon can be found in soil carbon pools, organic carbon is enriched from diverse carbon inputs (dead plant and animal residues, above-ground apoplast, root stubble and secretion, input of exogenous organic carbon) and outputs (soil respiration, root respiration, chemical oxidation of carbon-containing materials and decomposition of exogenous carbon input to the soil) .
Thus, climate change induces changes in soil microbial activity that affect soil organic carbon content and richness. Three significant impacts of climate change on soil organic carbon are as follows (Fig. 3): 1) increase in ambient temperature decreasing the proportion of soil bacteria; 2) Increase in the proportion of fungus; 3) A rise in the rate at which organic carbon will decompose, resulting in a slight feedback temperature rise and decrease in soil organic carbon. Overall, a rise in atmospheric carbon dioxide increases fungus decomposition of organic carbon, which enhances the greenhouse effect (Jenkinson et al., 1991).
Temperature regulates both the plant conversion rate of soil and atmospheric inorganic carbon to organic carbon, and plant wastes returned to the soil as organic carbon. Photosynthetic enzymes activity is also enhanced. Research on black soils in Northeast China during the freezing season, using simulated warming, showed snow melt increased soil water and decreased both soil and active organic carbon . Precipitation regulates soil moisture and soil respiration with impacts on soil organic carbon. Higher precipitation increases soil moisture, reduces soil aeration, reduces organic carbon susceptibility to mineralization, and enhances the soil's ability to digest plant and animal leftovers. Increased soil moisture hastens photosynthesis and the buildup of soil organic carbon. In contrast, when precipitation reduces, soil porosity increases, aeration and respiration rise, favoring mineralization and decomposition and reduction of soil organic carbon (Jiang et al., 2007). CO 2 levels impact soil microbial activity and plant photosynthesis. Increased CO 2 increases improves photosynthesis, plant biomass, soil organic carbon, and plant dry matter. Increased CO 2 reduces soil organic carbon as it affects the breakdown of organic matter by microorganisms (Van et al., 2014).
Despite considerable variation, temperature and precipitation are the main causes of changes in soil carbon pools. Zhou et al. (2003) studied more than 2000 soil profiles in China, and found significant differences between soil organic carbon, temperature, and precipitation along temperature gradients. When mean annual temperature was ≤ 10°C, correlations of soil organic carbon content with temperature and precipitation were -0.411 and 0.279, respectively, which was also the highest negative association with temperature. Between 10°C and 20°C, the correlation between temperature and precipitation was 0.320, hence temperature had less impact on soil organic carbon response to precipitation. Above 20°C, impacts were less pronounced.
The response of soil carbon pools to climate change is positive, but it is not advisable to use the same approach to measure the response in China's nine major agricultural regions due to their different geographical characteristics and climate differences. However, the use of organic fertilizers has helped mitigate the impact of climate change on soil carbon pools. Triberti et al. (2008) conducted comparative analyses on the carbon sequestration of organic fertilizers and found that they have a positive effect on soil carbon sequestration. In addition, the emission of CO 2 gas can also have a fertilization effect, but this effect on NPP is largely offset by the negative effects of climate change (Lobell et al., 2011). Future research work is necessary to study the combined impacts of climate and human actions such as organic fertilizers, irrigation, and crop kinds.

Impact on farmland pests and diseases in agroecosystems
Crop pests and diseases in China are characterized by a diverse variety, wide extent, and serious reduction in grain production. From 1990-1991, a wheat powdery mildew pandemic in China caused a loss of up to 1.438 × 10 10 kg of wheat production. Despite remediation the following year 7.7 × 10 9 kg was lost, and grain yields reduced by 30%-50% (Huo et al., 2012). At the beginning of the 21st century, yield losses from pest and diseases of rice, maize, wheat, and soybean, increased from 1.054 × 10 10 kg to 1.271 × 10 10 kg. Loss rates increased by 20.58%, and the area affected by pests and diseases exceeded 4.2 × 10 8 ha (Xia, 2010). Studies show that climate change is closely related to the occurrence of biological disasters in agricultural fields.
As the impact mechanism diagram shows (Fig. 3), climate changes affect pests and diseases in four main ways: first, they affect the growth and development of pests and diseases. Increased temperature exacerbates the damage caused by pests and diseases, facilitates the incubation of pathogens, shortens development time, and thus increases the number of pests and diseases per generation (Yang et al., 2002). Warmer winters and extended survival of winter-planted crops, indirectly reduce the mortality, and enhances over-wintering, of pests and diseases, which increase the following year (Wang et al., 2006a). Second, it can affect a variety of pests and diseases. In the 1950s-1970s, there were about 10 pest outbreaks per year in China, 14 in the 1980s, 18 in the 1990s, and an average of about 30 per year by the beginning of the 21st century (Xia, 2010). Third, changes in climate can lead to an increase in the adaptation of pests and diseases to their environment, bringing forward time of occurrence and affecting a wider range of crops. Before the 1970s, wheat powdery mildew in China occurred more severely in the southwest. With warming, the mildew boundaries moved northward. By early 1980s, the boundary expanded, and impacts increased, to the Jianghuai and Huaihua wheat regions. After the mid-1980s, powdery mildew expanded northward to the Huanghuai wheat region, and now powdery mildew has become a frequent disaster in wheat production in more than 20 provinces/autonomous regions/municipalities (Huo et al., 2002). Xiao et al. (2007) studied wheat stripe rust in the Longnan Mountains from 1957-2006 and found that the altitude of affected regions increased approximately 100-300 m, and occurrence was earlier from March to February. From 1961 to 2010, the area affected by agricultural pests and diseases in China increased 5.38 times overall, and by 2.51, 9.78, and 8.66 times for wheat, maize, and rice, respectively (Jiao, 2014). Finally, the natural enemies of pests and diseases are impacted. Carabid beetle (a member of the Coleoptera, suborder Adephaga, family Carabidae) which moves quickly throughout the day on the surface are natural enemies of pests that infest farms and forests. They are also a significant indicator organism of environmental ecological changes because of sensitivity to environmental changes. As the temperature rises, Carabid beetle populations in low latitude black soil farming region of Northeast China declined compared to high latitudes (Ma et al., 2019).
In summary, climate change increases pests and diseases and a decline in natural enemies. The challenge therefore is to find ways to reverse these trends. Regional variations of climate change also led to differences in the regional response of pests and diseases. One study found Southwest China had higher temperature and precipitation and fewer sunshine hours. Longer sunshine hours are less favorable for pest and disease growth and development. Temperature and precipitation have a greater impact on pests and diseases in Northwest China, but different pest and disease species respond differently to temperature and precipitation. For example, wheat stripe rust is more sensitive to temperature, while aphids and cotton bollworm are more sensitive to moisture. Pests and diseases in the northern China are strongly influenced by precipitation, temperature, and insolation, which are inversely proportional to the area of pest and disease occurrence and positively proportional to temperature. Northeast China is strongly influenced by temperature, while warming in winter reduces the mortality of overwintering pest eggs, which in turn increases the chances of pest and disease occurrence the following year . In the southern China, improving the planting methods can help reduce the impact of pests and diseases. For example, Liang et al. (2016) conducted a study on rice planting methods and found that intercropping can effectively suppress pests and diseases in dryland agricultural systems and increase rice yield per unit area. In the face of climate change, declining biodiversity in agroecosystems, and invasive alien species, there is an urgent need to establish a good crop pest monitoring and early warning system. However, the lack of understanding of the characteristics of regional agroecosystems makes it difficult to establish corresponding prevention systems (Wu et al., 2022). Therefore, it is crucial to understand the regional characteristics of pest response to climate and establish monitoring and prevention systems suitable for various agricultural regions to curb the spread of pests and diseases and ensure food security.

Impact on food production in agroecosystems
Greenhouse gases CO 2 , CH 4 , and N 2 O have increased since the industrial revolution by 30%, 145%, and 15%, respectively. These gases raise global temperatures (Zhiqiang and Chengquan, 1999), and significantly impact food production. Yang et al. (2011), projected that, compared to 1950 to 1980, climate warming, from 2011-2040, will move the northern limit of cropping in China northward. Grain yields may increase from cropping changes from mono-to-biannual and bi-to-triannual. From 1961 to 2010, the light-temperature production potential of corn in the black soil area of Northeast Heilongjiang Province increased by 367 kg/ha per decade, while the climatic production potential decreased by 61.5 kg/ha per decade. Temperature and precipitation therefore significantly impact corn production (Sun et al., 2013).
Extreme climate events such as droughts, floods, wind, dry-hot wind, and hail alter regional climatic conditions with impacts on crop growth and development. Food security is compromised as agro-meteorological hazards have immediate, severe, and pervasive effects on food production, despite other multifaceted factors (e.g., light resources, temperature, water environment, soil quality, and agricultural management) . According to the National Statistical Yearbook of China (https://data.stats.gov.cn, 1980China (https://data.stats.gov.cn, -2020, the average annual area affected by meteorological disasters is about 403 821 km 2 , and disaster area is 207 410 km 2 , causing direct economic losses in agriculture of more than 14.5 billion dollars. From 1980 to 2020, the average annual area affected by meteorological disasters in China was 26.22% of the sown area of crops, with loss from droughts and floods at 53.36% and 29.07%, respectively (Fig. 4a).
Droughts and floods are mainly caused by changes in precipitation. Severe water shortage in soil triggers regional drought, causing growth retardation, wilting, flower, fruit drop, and increased risk of plant death. Pu et al. (2008) analyzed the grain yield in the Hedong region of Gansu from 1981 to 2006 and found that during severe drought, regional winter wheat and maize maximum yields reduced by 34%-40% and 20%-37%, respectively. If high temperatures occur on top of this, it will lead to the occurrence of severe heat waves and dry hot winds, affecting the formation of grains and causing a significant reduction in crop yields. In the North China Plain region, the frequency of regional dry hot winds has significantly increased during 2011-2018 , and they mostly occur during the flowering period of winter wheat, seriously affecting the dry weight of wheat. Although sprinkler irrigation can effectively reduce the impact of dry hot winds, the climate conditions at this time increase the difficulty of implementation (Cai et al., 2022). On the other hand, precipitation increases for a short period triggers regional flooding, causing serious waterlogging, crop damage and erosion. Extended water logging can produce toxic substances that poison crop roots.
Other extremes from wind, hailstorms, and low-temperature cold can also affect production that varies by region (Fig. 4b). A study of agroclimatic disaster impacts in China from 1970 to 2014 ranked drought first in area affected, followed by floods, wind, hail, and low-temperature cold damage (Yu et al., 2017). Regional variations showed drought dominates in North China, Northwest China, and Northeast China, with disaster rates of about 10%. Floods dominate in the south-central and northeastern China, with a disaster rate of more than 4%. Wind and hail have a large impact in North and Northwest China, followed by South, North, Southwest, and South-Central China. Low-temperature cold damage is mainly in Northwest China (Yu et al., 2017). Overall, extreme climate impacts pose a serious threat to agricultural production and food security.

Prospects
The impact and feedback mechanisms of climate change on agroecosystems are complex, and we cannot guarantee a complete review of all processes. However, it is beyond doubt that the changes between them can threaten food security. The physical processes of landatmosphere interactions are extremely complicated and constrained by many factors, which leads to the complexity of climate. Therefore, how to prevent the impact of climate change on agroecosystems and how to respond to the resulting harm should be the focus of our future research (Zhou, 2015). Irrigation, planting methods and dates, breeding, fertilizers, and pesticides can indeed effectively address the adverse factors caused by climate change. However, in the process of using these measures, we should also consider factors such as biod-iversity, environmental adaptability, and economics. These measures are based on an understanding of the impact mechanisms. However, China has not yet formed a targeted plan for agricultural research. In addition, facing the impact of climate change on China's agricultural system, using more advanced crop models to reproduce historical changes (Wang et al., 2020) is more accurate and convincing than traditional statistical models, machine learning, and other methods Xue et al., 2007;Pu et al., 2008). This can also provide accurate predictions for future agricultural development. Our review has documented numerous elements affecting the relationships, that vary with time and regions, between farming and climate and impacts on ecosystems of farmland. There is currently less research on the effects of future climate change on farming ecosystems than on the effects of historical climate change. Future research on climate change impacts on China's agroecosystems should concentrate on the following areas: (1) Recognizing the farming ecosystem as a complex system requiring multidisciplinary study of interrelationships between ecosystem organic and inorganic components, processes, and climate change. This should include aspects of climatology, ecology, agricultural science, and global change science. As current research suggests (Jiang et al., 2007;Gao and Shao, 2011;Van et al., 2014a;Zou, 2014), soil carbon storage, NPP, and soil microbiota are all complex problems that cannot be fully addressed by a single discipline. The mechanisms of their interactions are complex, and interdisciplinary collaboration can better clarify their impact mechanisms. Additionally, a multidisciplinary big data platform, facilitated by rapid development in remote sensing and geographic information sciences, is vital to enable studies of the relationship between climate and agroecosystems across the different regions with differing climate, soil, cropping, productivity, pests and diseases.
(2) Strengthen studies of the mechanisms, temporal and regional variations, of climate change impacts on soil moisture, temperature, precipitation, and CO 2 . Temperature effects the activity of soil microbes, plant photosynthesis, and other processes. Precipitation affects soil moisture content. Although multi-factor, long-term observation, and integrated simulation experiments are lacking, and process models are not flawless, existing studies on the impact of single factors on ecosystems is quite in-depth. However multi-factor studies are vital to investigate synergistic, contrasting, and compound impacts. In recent years, dynamic crop models have become the most effective tool for studying the impact of climate change on crop yields and have been proven to be more advantageous than traditional models (Peng et al., 2018;Zhang et al., 2020a;Partridge et al., 2021). It improve the simulation of land-atmosphere interactions, not only by simulating the impact of climate change on crop growth processes and grain yields, but also by including feedback processes of crop growth to the climate (Yu et al., 2022a). However, the application of crop models is mostly concentrated in the northeastern and northern China, and other agricultural production areas lack the application of the model. In addition, most crop models were developed abroad, the localization of model parameters and finding suitable model parameters for major agricultural regions will be an important issue for future research. For regional adaptation management in China, it is urgently necessary to develop a coupled agro-climate system that integrates models of climate, land surface process, crop dynamics, carbon, nitrogen cycles, and remote sensing mapping (Yu et al., 2022b). This system will use a combination of field observations and computer simulations to thoroughly assess, and forecast, the effects of climate change on agroecosystems.
(3) Bolster studies on how farming ecosystems re-spond to climate change. By altering the energy and water balance between the surface and the atmosphere, agricultural practices, and farmland phenology both create feedback that can either slow down or speed up climate change. Some studies have shown that during the crop growth period, changes in physiological structures such as crop canopy height and leaf area index significantly affect the distribution of energy between sensible and latent heat by altering surface roughness and crop evapotranspiration, which in turn affects regional climate patterns . Zhang et al. (2013) found that increasing surface coverage in spring crops in the North China Plain can produce a cooling and moistening effect. In addition, as mentioned in the section on soil carbon pools, the impact of agricultural soil microbes on the atmosphere, such as CO 2 emissions, can have a regional warming effect (Jenkinson et al., 1991). A review of historical research reveals that the lack of certainty in assessments of climate change impacts on agroecosystems is due in part to ignoring the feedback effects of farming ecosystems on climate change. The key to tackling this issue is to create crop dynamics models adapted to China's climate change environment, nested regional farmland variations, and couple these models with appropriately-resolved climate models. The goal is to accurately simulate the relationship between farmland and climate, and reduce the impact of uncertainties.
(4) Enhance efforts to study agricultural climatic adaptation. Growing traditional crops in China has become extremely difficult due to the irreversibility of climate change and the frequency of extreme weather and climatic events in recent years. Crop productivity and quality, crop cultivation methods, and other factors will alter because of climate change. For example, Liu et al. (2013) analyzed meteorological station data and found that advancing the maize planting date could help the crop adapt to climate change, resulting in a 4% increase in crop yield. Additionally, the northern limit and arrangement of China's cropping system will shift northward (Yang et al., 2011). Quantitative analysis of crop adaptation to climate, is however constrained by the enormous number of crop species in China, the complexity of cropping systems and patterns, and from regional variations in climate and farmland practices characteristics.
The climatic conditions in the north and the south of China are very different. In recent years, extreme farmland climate events have mainly manifested as drought in the north and flood in the south (Fig. 2). Therefore, in future research, focus on combining the climate and geographical characteristics of different agricultural regions, and strengthen the research on crop species renewal, planting ratio, farming methods and disaster prevention, so as to improve the adaptability of crops to climate change. The farmland ecosystem should also evolve from a single cultivated farmland ecosystem to an ecologically complex agroecosystem (farmlandgrassland, farmland-wetland, etc.) to increase its ecological stability, improve its resistance to climate change, and bring about production and ecosystem benefits.
(5) Concentrate research on identifying and mitigating the main hazards associated with farming disasters. This requires urgent studies of the interaction mechanisms between hazards, impacts on farmland ecology, and their derived disasters (such as pests and diseases, soil quality changes) (Wu et al., 2022). What monitoring system and methods should be adopted to deal with the imbalance of soil moisture and carbon storage? Although the use of irrigation, fertilizers, and pesticides can alleviate the external constraints on crops (Triberti et al., 2008;Fu et al., 2022), it is necessary to establish a mechanism to prevent and handle the differences in soil conditions, biological species, and precipitation patterns in different agricultural regions in China. As a result, it is essential to improve risk identification for catastrophic climate, pests, and diseases. There is a need for an indicator system for China's farmland climate disasters and pests that is adapted to regional climate changes. The bigdata monitoring system should be designed to enhance the detection, early warning, prediction, and rapid response strategies for farmland climate disasters and pests based on different spatial scales. Risk management strategies should also be established for major farmland disasters suitable for China's agricultural industry.

Conclusions
Ensuring food security and maintaining a normal standard of living for the population is forcing China's agroecosystems to make positive changes in response to shocks caused by climate change. However, the lack of quantitative and objective analyses to ensure the stabil-ity and development of agroecosystems increases the complexity of agro-climate mechanisms, leading to uncertainty and undesirable consequences. In this paper, we review the characteristics of climate change in China during 1951-2020 and the systematic progress made in agroecosystem response to climate change, possible limitations, and propose potential solutions to existing problems.
We find that surface temperature, precipitation, and extreme climate events (such as droughts, floods, and high temperatures) have increased to varying degrees in China, accompanied by significant geographical differences. The processes and structural mechanisms of soil moisture, carbon pools, NPP, and pests in agroecosystems are becoming more complex under the influence of surface temperature and precipitation, which are seriously affecting crop growth and development. The forcing of climatic factors leads to decreases in soil water holding capacity, NPP, soil carbon content, and the number of natural enemies of diseases and insects, which in turn lead to lower crop yields. Through analysis of historical studies, it has been determined that the examination of single-influence climate factors on agroecosystems has hindered the exploration of agroclimatic mechanisms and processes; the utilization of multidisciplinary intersections can aid in clarifying the functions and impacts of various components within the system; and the implementation of robust early warning systems for farmland hazards can effectively mitigate the loss of crop yields. The theoretical lag in historical research methods poses a great challenge to future agroecosystem. Therefore, there is a need to explore disciplinary integration, coupled land-climate process models, and improved agro-climatic early warning systems to better capture the determinants of crop production and enhance analytical capabilities.