Closing crop yield and efficiency gaps for food security and sustainable agriculture

The world population is estimated to reach 10 billion by 2050. The rise in population coupled with improving living standards will require a doubling of the crop yield to fulfill food demand and livestock feed,which means crop yield will need to increase by 60–110% between 2005 and 2050 (Tilmanet al.2011). However,studies show that the production of major food crops is only increasing at the rate of 0.9–1.6% per year,which is not enough to feed the growing population (Rayet al.2013). Ensuring food security is an important issue of concern all over the world. One way to increase the crop supply is to expand arable land,and the other way is to increase crop yield per area. Foleyet al.(2011) found that global crop production has increased by 28% in the past (during 1985–2005),with an 8% contribution coming from the expansion of crop area,while the remaining 20% contribution coming from increased crop yield. Global scale analysis shows that the yield gaps exhibit different spatial patterns,and the crop yield is rising more and more slowly than before,especially in Europe (Foleyet al.2011).China has more than 20% of the world’s population,but its arable land area is only 7–9% of the world’s total amount(Lamet al.2013). Closing the yield gap and maintaining the increases in yield are very important for our food selfsufficiency and food security.

Many efforts have been made to boost yield increases,among which fertilizer plays a significant role. Fertilizer consumption increased more than nine-fold from 1961 to 2013 (Lu and Tian 2017). However,inefficient nutrient uptake in crops leads to significant nutrient losses to the environment,causing serious environmental consequences(Juet al.2009). From a meta-analysis of 140 published results,Slatteryet al.(2013) found that besides of nutrient supply,CO2concentration and shading also make significant contributions to closing the yield gap. Thus,improving canopy photosynthesis can reduce the yield gap. Unreasonable water inputs or improper planting patterns may also lead to a decrease of natural resource utilization efficiency. To further increase resource utilization efficiency and close efficiency gaps,we need to better manage nutrients and water,match crop varieties with solar radiation and temperature distributions in different regions,and increase crop tolerance to biotic and abiotic stresses.

In China,rice,wheat and maize are the major food crops.Their planting areas are widely distributed and resources such as light,temperature,water,and soil fertility vary in different regions. There are great differences in yield levels in the main production regions. The resource utilization efficiency is relatively low,and the mechanisms for synergetic increases of crop yield and efficiency are still unclear. To solve these problems,we design experiments with treatments which achieve different yield levels and characterize the demand of crops for resources on individual,populational and even regional levels. By using modeling,and physiological and biochemical technology analyses,we quantify the yield and efficiency gaps of the major crops in China,find the determining factors and mechanisms and provide effective management strategies for closing these gaps.

Most of studies in this special issue come from the National Key Research and Development Program of China (2016YFD0300100). During 2016–2020,we collaborated with scientists from 34 institutes and universities to address the challenges and opportunities for closing the crop yield and efficiency gaps to ensure food security and achieve the sustainable development of agriculture in the future. There are 22 papers organized into the following four sections.

Section 1:Using modeling methods to evaluate yield and efficiency gaps

Yield/efficiency gap was defined as the difference between achieved yield/efficiency and potential yield/efficiency (Van Ittersumet al.2013). The achieved yield can be measured directly,while the potential yield of a certain region can be simulated by a model. A proportion of 80% of the potential yield should be considered as the attainable yield (Cassmanet al.2003). Resource use efficiency (e.g.,nitrogen,water and solar radiation) was defined as the amount of output achieved per unit input. Ronget al.(2021) comprehensively review the currently studies on yield and efficiency gaps for the world’s major food crops in recent years. Based on more than 110 published papers,data from FAO and the Global Yield Gap and Water Productivity Atlas,Ronget al.(2021) summarize the concepts,quantitative methods for gap analysis,yield limiting factors,and resource utilization efficiency of wheat,maize and rice.

In China,the high yield maize belt has a wide span from east to west,resulting in a large differences in solar radiation and maize yields in different regions. Liu G Zet al.(2021)use multi-site experimental data,survey data from farmers or the published literature and simulations with the Hybrid-Maize Model to assess the yield gap and try to reduce the yield gap by matching planting density with solar radiation.They find a linear correlation between optimum planting density and accumulated solar radiation. By using this method,they suggest increasing planting density 47.3–84.8% in different regions,in order to increase grain yield by 10.9–20.2% and reduce the yield gaps by 10.4–33.7%.Northeast China (NEC) is one of the major maize production areas in China. Suet al.(2021) analyze spatial distributions and temporal trends of maize potential yield of the actual planted cultivars,and reveal the radiation use efficiency(RUE) and heat resource use efficiency (HUE) in NEC during the past 37 years by using an APSIM-Maize Model.Their results show that using maize cultivars with a longer growing season can reduce the yield and efficiency gaps.

Yield and efficiency gaps are also studied using models in wheat (Yaoet al.2021) and soybean (Nehbandaniet al.2021). A DSSAT-CERES-Wheat Model is used by Yaoet al.(2021) to simulate wheat yield and nitrogen use efficiency (NUE) under different agricultural management treatments at a regional scale. The results show quantitative comparisons of the contributions among different treatments.For reducing the yield gap,the relative contribution rates are nitrogen application>sowing date adjustment>soil management>planting density; for increasing NUE,the relative contribution rates are soil management>sowing date adjustment>planting density>nitrogen application (Yaoet al.2021). A Soybean Simulation Model (SSM-iCrop2) was used together with Geographical Information System (GIS)to determine the potential yield and yield gap of soybean in Golestan,Iran. In this study,the authors use data from 24 weather stations and soil data to calculate yield gap at county and province levels. The results show that there is a large yield gap of soybean and water use efficiency(WUE) which varies the in western and eastern regions. By closing 80% of current yield gap,soybean production in this province could increase by 66% (Nehbandaniet al.2021).

This section also includes two studies which develop new models and new parameters. Although Hybrid-Maize and APSIM models have been widely used in estimating maize potential yield at specific sites,as mentioned above,these models are not suitable for analysis over broad regions (e.g.,national or global scales). The reason is due to the sparse distribution of weather stations on a national or global scale,and limitations on field management information over broad regions. Zhanget al.(2021) develop a process-based and remote sensing driven yield model for maize (PRYM–Maize).Under eight data-model coupling strategies,this new model gives the highest correlation and smallest standard error for estimating maize potential yield,thus providing a scientific basis for facilitating the maize potential yield analysis over the NEC. In addition,Cao Det al.(2021) present an improved regional parametric syntheses approach,and establish a “rice zoning adaptability criteria and dynamic harvest index (RZAC-DHI)”. This new parameter can effectively simulate the rice cultivation area and yield.The rice area can be extracted by RZAC from Moderate Resolution Imaging Spectroradiometer time-series data and phenological information,and yield can be obtained by DHI calculation and net primary productivity. By analyzing the spatial-temporal patterns of the rice cultivation area and yield in NEC during the past 15 years,the results show that rice planting areas are mainly located along the Songhua and Liaohe rivers in China. The rice cultivation area and yield increased by 58 and 90%,respectively,in NEC during the past 15 years. These new methods are useful for showing the dynamic changes of crop yield in a broader scale,which provides a new basis for planting management and agricultural policy making (Cao Det al.2021).

Section 2:The main factors determining yield and efficiency gaps at different levels

Unlike the model simulation results in last section,all the results in this and the following sections are based on actual grain yield/efficiency data from field experiments or farmer’surveys. This section includes four papers:two investigate the main factors determining the maize yield gap in Huang-Huai-Hai and southern China,respectively (Liu Y Eet al.2021; Shaoet al.2021); one determines the causes of differences in rice yields in southern China (Wanget al.2021),and another one analyzes the technical efficiency of rice,maize and wheat across three major grain producing regions in China (Zhouet al.2021).

Crop yield is determined by many different factors,such as ecological environment,cultivar performance,cropping system,production management,and socio-economic factors. Scientists often use different management approaches to achieve different yield levels (e.g.,basic productivity,farmer practices,high yield and high efficiency level,and super high yield level) under the same environmental conditions. Wanget al.(2021) compare the yield and grown phenotypes of 10–30 high yield hybrid rice cultivars which were grown in the upper or middle–lower reaches of the Yangze river in southern China. Different amounts of accumulated temperature and solar radiation in the upper or middle–lower reaches lead to significant changes in tillering duration,seed setting and dry matter accumulation. Results from this work show not only the yield gap as affected by environmental conditions but also provide guidance on cultivar selection and crop management practices for high yield rice production in different regions.Liu Y Eet al.(2021) analyze maize yield gaps at different levels in the Huang-Huai-Hai summer maize region in China. By comparing the priorities of management factors(including planting density,fertilizer,hybrids,and irrigation)for reducing the yield gap,they find that increasing planting density is the key factor for decreasing the yield gap between‘farmer practices level’ and ‘high yield and high efficiency level’; while choosing hybrids with density and lodging tolerance is the key factor for decreasing the yield gap between ‘high yield and high efficiency level’ and ‘super high yield level’. There are about 200–300 million smallholder farmers in China (Cuiet al.2018),and yield gap analysis is more difficult for smallholder farmers due to their diverse cultivars,crop management methods and yield levels. In Jiangsu Province in China,maize is mostly produced by smallholder farmers. Shaoet al.(2021) develop a method consisting of five progressive procedures to analyze maize yield levels,yield gaps,determination factors,and so on. From the results of a questionnaire investigation of 1 023 smallholder farmers,they find that planting density,fertilizer application rate,hurricane-caused lodging,and pests-caused damages are the major causes of yield gaps.Technical efficiency is used to evaluate how well a producer is utilizing the technologies to produce the maximum output(Chavaset al.2005). Regarding the technical efficiency in these smallholder farmers,Wanget al.(2011) compare the spatial variation of technical efficiency in farmers’ fields for rice,wheat and maize production in China. The study is based on a large-scale farm household survey,which covered 1 218 rice farmers,3 566 wheat farmers and 2 111 maize farmers in the main producing areas. Technical efficiency was estimated by stochastic frontier analysis(SFA),which is a one-step approach using the maximum likelihood to estimate the parameters of the production function equation and the technical efficiency model. The analysis results show that technical efficiency is relatively high among all the rice,wheat and maize farmers,and rice farmers have the highest technical efficiency at nearly 0.9 on average. Both local environmental and socio-economic factors significantly affect farmers’ technical efficiency (Zhouet al.2021).

Section 3:Physiological mechanisms for closing yield and efficiency gaps

As mention above,the ecological environment and agronomic management practices affect crop yield significantly. Grain yield is more affected by planting density in maize than other cereals due to its low tillering ability(Sangoiet al.2002). Average maize grain yield per area increased dramatically during the second half of last century,which has been attributed to several factors including high planting densities (Duvick and Cassman 1999). However,high density planting might increase the risk of lodging,but an understanding of the physiological determinants of maize resilience to the population density stress still remains unclear (Sangoiet al.2002). In this section,three papers focus on exploring the physiological mechanisms for maize yield improvement under the high-density condition (Cao Y Jet al.2021; Liu X Met al.2021; Yanget al.2021). Yanget al.(2021) find that the maize cultivars Xianyu 335 (XY335)and Zhengdan 958 (ZD958) respond differently to changes in solar radiation under a planting density of 120 000 plants ha–1. Under high solar radiation conditions,XY335 has a higher yield than ZD958; while under low solar radiation conditions,ZD958 has a higher yield than XY335. The yield gap caused by a decrease of solar radiation is greater in XY335,which can be explained by decreases of dry matter accumulation in the shoot,root and especially in the ear(Yanget al.2021). Another study analyzed the influence of leaf reduction on physiological performance and maize yield in high and low planting density populations. All the leaf reduction treatments reduce growth performance and maize yield under the low planting density conditions. Interestingly,Cao Y Jet al.(2021) find a removal of 1/4 of each leaf length per plant can increase the leaf photosynthetic rate at the grain-filling stage,dry matter accumulation,harvest index,and grain yield under high planting density conditions. In this treatment,root activity and soluble sugar content in the bleeding sap also increased,leading to higher nitrogen uptake and nitrogen accumulation in the grain. These results provide evidence that proper removal of the leaf source could improve maize growth,grain yield and NUE under high planting density conditions (Cao Y Jet al.2021).Liu X Met al.(2021) investigate the effects of chemical regulation and nitrogen fertilizer on lodging and maize yield under high density conditions. Three nitrogen application rates and a plant growth regulator (a mixture of 3% DTA-6 and 27% ethephon) are used in different treatments. The results show that the chemical control treatment reduces lodging percentage but nitrogen application increases lodging percentage by regulating lignin synthesis enzymes and lignin content. The use of an intermediate nitrogen level(200 kg ha–1) and chemical control gives a high grain yield under high density conditions.

This section also includes two studies showing the influences on the crop yield and efficiency from the ecological environment,especially the soil conditions (Guoet al.2021; Maet al.2021). One study uses a PTM-48A photosynthesis monitor to makein situmeasurements of diurnal changes of photosynthetic parameters in fieldgrown super high yield (SH) and high yield (HY) wheat plants together with various soil factors. The results show that soil temperature and soil moisture are better suited for higher rates of leaf photosynthesis under the SH treatment than under the HY treatment at noon (Maet al.2021). The other study involved a two-year field experiment in a sodic saline-alkaline paddy field. The Songnen Plain (Jilin,China)is among the world’s largest areas of land with sodic salinealkaline soil. Rice is a major staple crop for this region,and is often used to mitigate the saline–alkaline soil problem.Five nitrogen application rates and three hill densities are used to analyze rice growth,yield and NUE in saline–alkaline soil conditions. The results show that the highest yield was obtained at 142.61 kg N ha−1with a planting density of 33.3×104ha−1. However,NUE is negatively correlated with grain yield. Thus,the combined changes in the N application rate and hill density are needed to optimize rice yield and nitrogen efficiency in sodic saline–alkaline paddy fields (Guoet al.2021).

Section 4:Effective management strategies for closing yield and efficiency gaps

The application of nitrogen fertilizers has increased crop yields and grain quality. Normally,the high yield is achieved through high nitrogen input. However,many studies have shown that excessive nitrogen supply cannot further increase crop yield. Decreasing the efficiency gaps is as important as decreasing the yield gaps. This section includes six papers showing effective management strategies for closing yield and efficiency gaps.

NUE needs to be increased to reduce residual nitrogen release into the environment. One option for reducing soil residual nitrogen is to match the crop demand with soil nitrogen availability. By using slow-release fertilizer,Li G Het al.(2021) decrease nitrogen supply by 11.1%,while increasing maize grain yield and NUE by 3.2 and 22.2%,respectively. This one-time application of slow release fertilizer provides a simple,efficient and businessfriendly strategy for spring maize production in Jiangsu Province in China. Fuet al.(2021) find that planting at high density with reduced nitrogen input and delayed nitrogen application (DPRN) is also a feasible approach for simultaneously increasing grain yield,NUE and RUE in the double rice cropping system in South China. Here,the delayed nitrogen application matches better with the larger nitrogen demand in the reproductive stage. However,the reduction of nitrogen input also depends on the accumulation levels of N in the soil. With the improvement of mechanization levels,it is also possible to mechanically place N-fertilizer directly besides the seedlings at a (5.5±0.5)-cm soil depth when rice is transplanted. Zhuet al.(2021)compare seven different N-supply treatments and find that mechanized deep placement of mixed urea and controlledrelease urea at transplanting is a highly-efficient cultivation technology that enables increases of rice yield and NUE simultaneously.

Regarding to strategies for environmental factor regulation,three papers in this section provide new approaches to improve the growth conditions for crops from different aspects. Wuet al.(2021) use a new tillage approach to optimize soil properties. Three years of field experiments reveal that deep vertical rotary tillage (DVRT) can reduce the soil bulk density and increase the soil water content at the anthesis stage,leading to increases of wheat yield,NUE and WUE by 22.0,14.5 and 19.0%,respectively. Li J Pet al.(2021) use a micro-sprinkling approach to improve WUE in winter wheat. With two-year field experiments testing various combinations of four irrigation amounts and three nitrogen levels,they find micro-sprinkling with 120 mm of irrigation and a total nitrogen application of 195 kg ha–1can lead to increases in wheat grain yield,WUE and NUE on the North China Plain. Gonget al.(2021) use a combined plant growth regulator application to optimize maize structure for better use of solar radiation at the population scale. In a high density maize population,lodging and shading from neighboring plants are two problems which limit grain yield.EDAH (1.5 mL L–1)+DA-6 (1 mL L–1) treatment can reduce plant height and ear height,and can increase lignin content,thereby leading to higher lodging resistance. In addition,the EDAH+DA-6 treatment decreases leaf area and tilt angle,creating a compact maize plant structure,which is good for photosynthesis efficiency in the high-density maize population. As a result,the EDAH+DA-6 treatment reduces the lodging rate by 6.95% and increases the maize grain yield by 15.51% (Gonget al.2021).

Taken together,these results provided a theoretical basis and technical support for coordinating the high yield with high resource use efficiency in rice,maize and wheat.Although most of the case studies were conducted in China,closing the yield gap and efficiency gap are always hot topics all over the world. In this special issue,we give a comprehensive report showing the progress we have made in China,from model simulation and determination factor analysis to mechanism characterization and management strategy applications. We hope our readers will find some interesting and useful information from this special issue.We thank Dr.Hou Peng and Dr.Li Congfeng,Institute of Crop Sciences,Chinese Academy of Agricultural Sciences,for their valuable comments regarding this manuscript. We thank the Editorial Department and Editor-in-Chief of theJournal of Integrative Agriculturefor their collaboration in organizing this special issue,inviting the peer reviewers and eventually assembling the accepted manuscripts into this piece of work. We are thankful for the funding from the National Key Research and Development Program of China(2016YFD0300100).