lnfluences of large-scale farming on carbon emissions from cropping:Evidence from China

Ll Ya-ling ,Yl Fu-jin ,YUAN Chong-jun

1 College of Economics and Management,Anhui Agricultural University,Hefei 230036,P.R.China

2 China Academy for Rural Development,Zhejiang University,Hangzhou 310058,P.R.China

3 School of Public Affairs,Zhejiang University,Hangzhou 310058,P.R.China

4 College of Economics and Management,Nanjing Agricultural University,Nanjing 210095,P.R.China

Abstract Reducing agricultural carbon emissions is important to enable carbon emission peaking by 2030 in China. However,China’s transformation towards large-scale farming brings uncertainties to carbon emission reduction. This study quantifies the carbon emissions from cropping based on life cycle assessment and estimates the effects of farm size on carbon emissions using a fixed effects model. Furthermore,the variations of the carbon emissions from cropping driven by the changes in farm size in future years are projected through scenario analysis. Results demonstrate an inverted U-shaped change in total carbon emission from cropping as farm size increases,which is dominated by the changes in the carbon emission from fertilizer. Projections illustrate that large-scale farming transformation will postpone the peak year of total carbon emission from cropping until 2048 if the change in farm size follows a historical trend,although it is conducive to reducing total carbon emission in the long run. The findings indicate that environmental regulations to reduce fertilizer usages should be strengthened for carbon emission abatement in the early stage of large-scale farming transformation,which are also informative to other developing countries with small farm size.

Keywords: farm size,carbon emission,agricultural transition

1.Introduction

China’s central government has released an ambitious action plan to reach carbon emission peak by 2030 that attracts growing concerns about the agricultural sector’s carbon emission reduction. Existing research showed that global food systems would prevent limiting global warming to 1.5°C above pre-industrial levels by the end of the century (Clarket al.2020). Food systems are projected to contribute 33.5% of the total anthropogenic greenhouse gas (GHG) emissions by covering emissions associated with the activities through the various stages and sectors of the global food system,with China contributing 13.5%and becoming the largest emitting economy in the world(Crippaet al.2021). China’s agricultural GHG emissions include carbon dioxide (CO2),methane (CH4) and nitrous oxide (N2O),which are important components of agricultural carbon emissions (Zhaoet al.2022). Farmingrelated activities are an important source of carbon emissions in the agricultural sector,with agrochemical inputs,rice production and energy use all contributing significantly to carbon emissions (Tian and Chen 2021).Thus,reducing carbon emissions from cropping in China is critical to China’s goal of reaching the carbon emission peak by 2030 and contributes to global carbon reduction.

Crop production management,such as improving tillage practices,soil management,and agriculture techniques,has been widely proposed as a means of reducing carbon emissions (Snyderet al.2009;Cilliset al.2018). However,reducing agricultural carbon emissions in developing countries will be more complicated in the stage of transformation from small-scale to large-scale farming. More than 80% of the world’s farms operate on less than 2 ha of land in developing countries (Fan and Rue 2020). As the world’s developing economy,China’s agricultural production has been dominated by smallholder farms over a long period. Meanwhile,policies promoting agricultural land consolidation and the development of rural professional cooperative associations have aided the expansion of farms over the past 2 decades (Zhanget al.2016). The aging of agricultural labors,migration of rural residents,and rapid urbanization in China also facilitate farm size expansion (Masterset al.2013;D’Amouret al.2017;Wanget al.2021). According to the observations from the Chinese Academy of Agricultural Sciences(CAAS),the rural residents’ per-capita arable land in China has increased by 60% from 2002 to 2017 (Appendix A).Large-scale farming has become an inevitable trend of China’s agricultural development.

The consequent adjustments of production technology from small-scale to large-scale farming,including agrochemical usage,energy inputs,and crop mixes,affect the carbon emissions from cropping (Fig.1). Various mechanisms cause carbon emissions to rise or fall. First,agrochemical usage is likely to decrease with farm scale expansion because of more efficient use and better management (Wuet al.2018;Duanet al.2021;Zhuet al.2022). By contrast,agrochemicals are viewed as laborsaving tools for agricultural production (Rahman 2015),which may increase agrochemical usage with farm size expansion. Moreover,the decline in crop diversity with farm size expansion probably increases pesticide usage to control pests and diseases (Ratnadasset al.2012;Jacquetet al.2022). Second,large-scale farming is conducive to energy use since it improves the accessibility and efficiency of mechanized production (Yilmazet al.2005;Unakitanet al.2010;Qiuet al.2017). The large gap of agricultural mechanization between China and developed countries raises concerns about China’s ability to peak agricultural carbon emissions (Jinet al.2021). However,large-scale farms typically employ advanced machinery and energyeffective management (Wanget al.2017) and further lead to carbon emission reduction when the farm size increases to a certain level. Additionally,the electricity consumption for agricultural irrigations is expected to increase with farm size expansion. Third,large-scale farming changes the crop mixes. Although farmers intend to plant more cash crops to maximize profits (Zeng 2015),the labor shortage is a critical constraint to specialize cash crops’ production as farm size rises (Van den Berget al.2007). Therefore,the uncertainties arising from the transformation to largescale farming must be factored in to reduce China’s agricultural carbon emissions,which is also important to global agricultural carbon abatement.

Fig.1 Conceptual framework.

Previous studies have examined the effects of farm size on agricultural carbon emissions. However,no consensus has emerged resulting from the differences in carbon emission quantification and study regions. Some studies calculated the carbon emissions from six sources related to cropping activities,including the inputs of chemical fertilizers,pesticides,agricultural film,diesel,electricity for irrigation,and plowing based on national provincial level data. The results presented that farm size had a U-shaped relationship with total carbon emission from cropping(Liu and Xiao 2020) and negatively affected the carbon emission per unit of output value (Weiet al.2023). Another part of relevant research applied life cycle assessment(LCA) method,however CO2emissions were excluded,to calculate China’s carbon emissions from cropping using the survey data in five provinces (Jilin,Jiangsu,Sichuan,Shaanxi,and Hebei) in China. The findings indicated that the expansion of farm size significantly reduced carbon emission per unit area,but raised total carbon emission from cropping (Xuet al.2022). Given the incomprehensive carbon emission quantification system and limited data in previous studies,the uncertainties driven by the transformation to large-scale farming on China’s agricultural carbon emissions need to be further investigated.Moreover,whether agricultural carbon emissions will peak by 2030 under the effects of large-scale farming in China remains obscured and deserves more discussions.

This study extends existing research from three perspectives. First,we construct a set of emission factors that fit China’s crop management patterns and climatic conditions to quantify China’s carbon emissions from cropping. A cradle-to-grave LCA method are employed to quantify the carbon emissions associated with cropping activities from raw material extraction to product transportation to storage centers. The adoptions of region-and crop-specific emission factors and cradle-tograve LCA method will improve the accuracy of calculation results. Second,we use the nationwide county-level data from 2002 to 2017 to empirically estimate the effects of farm size on the carbon emissions from cropping. Data improvements will overcome the limitations of capturing the heterogeneous and dynamic effects of farm size on carbon emissions with provincial panel data or survey data of certain regions. Finally,we set up various scenarios of possible changes in farm size to project China’s carbon emissions from cropping in the future based on the estimated effects of farm size on carbon emissions. Therefore,our research fills out the gap of projecting China’s agricultural carbon emission changes under the transformation towards large-scale farming.

The remainder of this study is organized as follows.Section two introduces the data and methods used for quantifying China’s carbon emissions from cropping,estimating the effects of farm size on carbon emissions and projecting carbon emission changes in future years.Section three displays the results of carbon emission quantification,estimated effects of farm size on carbon emissions and carbon emission projection. Section four presents further discussions about the results. The final section concludes our findings and proposes policy implications. The conceptual framework is shown in Fig.1.

2.Data and methods

2.1.Quantification of carbon emissions from cropping

The transformation to large-scale farming affects the carbon emissions both from field and agricultural input manufacture. Therefore,a cradle-to-grave LCA from raw material extraction,to industrial processing,field utilization,product harvest and finally transportation to storage centers provides much more comprehensive ways to quantify the carbon emissions from cropping.The methodology of LCA has been widely adopted in previous studies (West and Marland 2002;Martin-Gorrizet al.2020;Verdiet al.2022). In this study,the carbon emissions from cropping includes (1) CH4emissions from rice paddies;(2) N2O emissions from farmland caused by nitrogen fertilizer application;(3)CO2,CH4and N2O emissions from agricultural machinery operation,groundwater irrigation and agricultural inputs manufacturing. We extend the carbon emissions from the manufacture of agricultural inputs,including not only carbon emissions from their manufacture itself,but also those from their transportation and repair. Technically,we multiply commodity-specific production and input levels by the corresponding emission factors to account for the carbon emissions from cropping. The method of carbon emission quantification in detail is summarized in Appendix B. It should be noted that most of emission factors used in this study capture the differences in crop management patterns and climatic conditions across regions and crops,which can obviously improve the accuracy of China’s agricultural carbon emission quantification.

2.2.Strategy to estimate the effects of farm size on carbon emissions

To identify the effect of farm size on total carbon emission from cropping in China,we apply a fixed effects model using county-level data from 2002 to 2017:

where InEitdenotes the logarithm of countyi’s total carbon emission from cropping in yeart.Farmsizeitis countyi’s average farm size,measured by the percapita arable land in rural areas.Sq_Farmsizeitis the squared term of farm size.Xitis a control variable vector.We control the total planted area,which is expected to positively affect carbon emissions. We also control the share of area affected by disaster out of the total planted area,resulting in the reduction of the carbon emissions from agricultural production. Moreover,the variables associated with agricultural production revenue and cost,including the producer price index for planting products,fertilizers,and pesticides,and average wage,must be controlled because they affect decisions for the usages of agricultural inputs. Variables vary across entities but do not change over time,such as land fertility,can be controlled by using fixed effects regression. In addition,the county-level data are helpful for removing the interferences of micro factors. To control for the variables that vary over time but not across entities,such as policies and technology advancement,we add time dummy variables (Tt) into the fixed effects model.The correlation matrix and descriptive statistics for the variables are shown in Appendices C and D.βs(s=0,1,2,and 3) are the parameters to be estimated.εitis the error term. A bidimensional clustering strategy is employed to capture the serial and spatial correlations of the error terms (Cameronet al.2011). In addition,we add the cubic terms of farm size into eq.(1) and apply the U-test method proposed by Lind and Mehlum (2010) to verify the U-shaped or inverted U-shaped relationships between total carbon emission from cropping and farm size. The Akaike Information Criterion and Bayesian Information Criterion are used to select the most appropriate model to estimate.

Furthermore,we attempt to examine the effects of farm size on the carbon emissions from various sources related to cropping activities. Carbon emissions from rice paddies,fertilizers,pesticides,agricultural machinery,and irrigations are discussed to reveal the mechanism of farm size affecting total carbon emission from cropping. The empirical model specifications are described in Appendix E.

2.3.Scenario designs to project carbon emissions

The trend of large-scale farming resulting from land consolidation,rising labor cost,and agricultural labor migration in China will not be reversed in the short term,but the rate of farm size changes has different possibilities. The projections of carbon emissions are based on the consequent adjustments of agrochemical usage,energy inputs,and crop mixes resulted from the changes in farm size given other factors fixed. Given the effects of farm size on the carbon emissions from cropping,we propose a scenario analysis related to the changes in farm size to project China’s carbon emissions from cropping in future years. First,we use nonparametric statistics to illustrate the changes in farm size based on historical observations. Given the heterogeneous land endowments and agronomic conditions,seven subregions,including East China,South China,Central China,Southwest China,North China,Northwest China,and Northeast China are discussed in this study (Appendix F). We find increasing but nonlinear trends in farm size over years in most regions,whereas the farm size of South China has declined since 2010 because of urban land expansion driven by the rapid urbanization (Xionget al.2012) (Appendix G). Second,we apply polynomial regressions to estimate the changes in farm size over years for every region. The shape of the nonparametric charts determines the degrees of polynomial regression and model forms for each region.Model specifications are described in Appendix H.Furthermore,the bootstrap method estimates inference because of the asymptotic refinement (Beran 1988).Technically,1 000 bootstrap samples are drawn to estimate the confidence interval (CI) for a significance level of 5% (Cameron and Trivedi 2010).

Based on the estimation results about the changes in farm size over time (Appendix I),we construct a set of scenarios to project the carbon emissions from cropping in future years. Specifically,we assume the fitted value,and the upper and lower bound values of the estimators’95% CI as the three possible changes in farm size.Scenario settings are described as follows:

Business as usual (BAU) scenario: We assume that the change in farm size in China follows a historical trend.The estimated coefficients are used to predict farm size in future years.

Fast-change scenario: The farm size in China is assumed to change faster than the historical average trend. The upper bound of the 95% CI for estimated coefficients is used to predict farm size in future years.

Slow-change scenario: We assume that the farm size change in China is slower than the historical average level. The lower bound of the 95% CI for the estimated coefficients is used to predict farm size in future years.

2.4.Data sources

We obtained the nationwide county-specific agricultural production data from the Agricultural Information Institute of Chinese Academy of Agricultural Sciences (Tibet is not included because of data availability). This study discusses 10 types of crops: rice,wheat,maize,soybean,potato,cotton,rapeseed,sugar crops,vegetables,and fruits,accounting for more than 85% of the total sowing area. The agricultural production cost and price data are from theChina Rural Statistical Yearbookand the annualNational Agricultural Product Cost and Benefit Data Compilation. The emission factors adopted in this study are derived from relevant studies (Appendix J).

3.Results

3.1.Carbon emissions from cropping

The chord diagram is widely applied in previous studies to visualize the changes in total carbon emission from cropping over time and sources (Fig.2,the detailed information is shown in Appendix K) (Hazenet al.2019;Laspidouet al.2020;Galanakiset al.2021;Zenget al.2021). We find that China’s total carbon emission from cropping shows a fluctuating upward trend over the years.The total carbon emission increased from 143 Tg carbon equivalent (CE) in 2002 to the highest level of 186 Tg CE in 2015. A slight decline was observed after 2015,which is attributed to the reduction in agrochemical use because of the launch of the Action Plans for Zero Growth of Fertilizer and Pesticide Use in 2020.

Fig.2 Carbon emissions (Tg CE) from cropping over the years. Data source: authors’ calculations.

The carbon emissions from various sources related to cropping activities are further examined. The carbon emissions from crop production and agricultural input manufacturing account for 40 and 60% of the total carbon emission in 2017,respectively. With regard to the carbon emissions from crop production,CH4emission from rice paddies is the largest emitter,with 36 Tg CE in 2017. Farmland N2O emission from nitrogen fertilizer application is the second largest source of emissions,with 20 Tg CE in 2017.In terms of the carbon emissions from agricultural input manufacturing,fertilizer manufacturing is the largest emitter with 85 Tg CE in 2017,followed by pesticide manufacturing with 22 Tg CE. Overall,the carbon emission from fertilizer usage,including its manufacturing and application,dominates the total carbon emission from cropping,accounting for approximately 59% in 2017. CH4emission from rice paddies is another primary emitter,accounting for 20%.The carbon emission from pesticide usages accounts for 12% of the total emissions. However,the carbon emissions from agricultural mechanization,irrigation,and plastic film usage account for small proportions.

3.2.lnfluences of farm size on carbon emissions

The effects of farm size on the carbon emissions from cropping are illustrated in Fig.3. We find that the total carbon emission from cropping has an inverted U-shaped relationship with farm size (the detailed estimation results are reported in Appendix L). The turning point of the inverted U-shaped curve for farm size is 1.66 ha.Specifically,the total carbon emission increases before the farm size reaches the value of 1.66 ha and decreases when farm size exceeds 1.66 ha. Given that the current average farm size in China is 0.24 ha,much lower than the turning point,the carbon emission is expected to increase with the expansion in farm size.

Fig.3 Relation between farm size and total carbon emission from cropping (E). The turning point of farm size is 1.66 ha.The data on carbon emissions have been log-transformed.Each data point represents the average value of farm size within a certain farm size group,with 0.1 ha as the interval(31 farm size groups in total). The bubble size of each data point represents the sample size in every group. Data source:authors’ calculations.

To shed more lights on the inverted U-shaped relationship between farm size and total carbon emission from cropping,we further reveal the effects of farm size on the carbon emissions from various sources. The inverted U-shaped relationships are also presented between farm size and the carbon emissions from fertilizer and machinery use,with turning points for farm sizes at 1.61 and 1.51 ha,respectively (Fig.4-A and B). Given the increasing carbon emissions before approaching the turning point,the substitution effects of the usage of modern agricultural factors,such as fertilizers and diesel,on labor can offer explanation. When the farm size is larger than the turning point,better farming knowledge and management skills with expanding farm size lead to the adoption of modern agricultural technologies and management practices. The adoptions of advanced agricultural machinery and soil testing for formulated fertilization could improve the input efficiency,thereby reducing carbon emissions (Niroula and Thapa 2007;Tanet al.2010;Adamopoulos and Restuccia 2014;Zhuet al.2017;Wuet al.2018).

Fig.4 Relations between farm size and the carbon emissions (E) from various sources. A,fertilizers. B,agricultural machinery.C,rice paddies. D,pesticides. The turning point of farm size in Fig.3-A–C are 1.61,1.51,and 1.29 ha,respectively. The data on carbon emissions have been log-transformed. Each data point represents the average value of farm size within a certain farm size group,with 0.1 ha as the interval (31 farm size groups in total). The bubble size of each data point represents the sample size in every group. Data source: authors’ calculations.

In contrast with the inverted U-shaped relationship between farm size and total carbon emission from cropping,the carbon emission from rice paddies has a U-shaped relationship with farm size,with a turning point for farm size at 1.29 ha (Fig.4-C). Before reaching the turning point,farmers are inclined to grow more cash crops and reduce rice production to maximize profits (Zeng 2015). After reaching the turning point,the increase of labor costs driven by the farm size expansion makes farmers reduce planting labor-intensive cash crops and further increase the cultivation of land-intensive crops,such as rice (Van den Berget al.2007). Furthermore,farm size positively affects carbon emission from pesticide usage (Fig.4-D). This phenomenon can be attributed to the increases of pests resulting from reduced crop diversity as farm size expands (Ratnadasset al.2012;Jacquetet al.2022). The results are also consistent with the findings that more pesticides are used as farm size increases to reduce the risk of yield losses (Gonget al.2016). Pesticide application has limited effects on the total carbon emission with farm size expansion as pesticides account for a small proportion of total carbon emissions.

Overall,the inverted U-shaped change in total carbon emission from cropping is primarily attributed to the similar changes in the carbon emissions from fertilizer and agricultural machinery as farm size increases. Given that fertilizer usages are the greatest emitters,the inverted U-shaped relationship between farm size and total carbon emission is dominated by the inverted U-shaped relationship between farm size and the carbon emission from fertilizers. Although the carbon emissions from rice paddies and pesticides show different change trends,they cannot alter the inverted U-shaped changes in total carbon emission from cropping as farm size rises.

3.3.Projection of carbon emissions from cropping

We show the projected total carbon emission from cropping in China under various farm size change scenarios (Fig.5). To avoid mixed results driven by other factors,we have to assume that the farm size is the only variable that will change at different speeds.The black curve depicts the historical changes in total carbon emission from cropping. The colored curves show the change tendency of total carbon emission from cropping under different scenarios in future years,and the changes are attributed to the fluctuations in agricultural inputs and crop mixes resulting from the farm size variations. In the BAU scenario,total carbon emission will decline until 2029,then increase until 2048,and finally decline thereafter,detailed information is shown in Appendix M. Total carbon emission from cropping in 2048 is 185 Tg CE,which is close to the amount in 2015,indicating that a long plateau from 2015 to 2048 is observed after the peak value of carbon emission emerged in 2015. Thus,our study proposes that the changes in farm size alter the downward trend of total carbon emission from cropping after 2015,causing carbon emission to rise at a certain stage. In the fastchange scenario,total carbon emission will increase until 2039,with the highest carbon emission of 188 Tg CE,and then decrease. The peak year is 9 years later than the target year to peak carbon emission. In the slowchange scenario,an overall downward trend of total carbon emission from cropping is presented after 2015,with a negligible increase from 2050 to 2063. Thus,the inevitable trend of large-scale farming is likely to pose challenges to reach carbon emission peak by 2030 for China’s crop production.

Fig.5 Projections of total carbon emission from cropping in China under various scenarios. The upper limit of farm size is set to 8.9 ha,which is the rural residents’ per-capita arable land in Australia because it has the largest farm size in the world.BAU,business as usual. Data source: authors’ calculations.

We project the carbon emissions from cropping across seven subregions in China (Fig.6). Taking the BAU scenario as an example,we find that carbon emission will peak in 2015 in the Northeast China. The reason is that the Northeast China has the largest average farm size in China,and the farm size in most counties allocates the right to the turning point in Fig.3,resulting in decreasing carbon emission after 2015. The carbon emission in the South China will peak in 2017. The counties with rising farm size dominate an increasing carbon emission in South China before 2017. The farm size in most counties alters to reducing after 2017 and allocates the left to the turning point in Fig.3,resulting in decreasing carbon emissions in South China. The carbon emissions in the Southwest and East China will peak in 2045 and 2059,respectively. Meanwhile,the carbon emissions in the North,Central,and Northwest China will peak even after 2085. Given that the farm size in the Southwest,East,North,Central,and Northwest China shows an increasing trend in future years,the year of peak carbon emission from cropping in these regions is determined by both the initial value and growth rate of their farm size. The explanation is similar to that of the differences in carbon emissions in various scenarios. Small initial value or growth rate of farm size leads to late year of peak carbon emission.

Fig.6 Projections of the year of peak carbon emission across regions in China under various scenarios. A,slow-change scenario.B,business as usual (BAU) scenario. C,fast-change scenario. The yellow shading indicates that the region’s carbon emission peaked in 2015,the orange shading denotes that the region’s carbon emission will peak after 2015 and before 2085,and the red shading means that the region’s carbon emission will peak after 2085. The numbers indicate the year when the region’s carbon emission peaked,and 2085+means that the region’s carbon emission will peak after 2085. Data source: authors’ calculations.

4.Discussion

This study uses the emission factors that fit China’s crop management patterns and climatic conditions to quantify the carbon emissions from cropping based on the LCA principle. The quantification results of total carbon emission from cropping in China is similar to Liet al.’s (2022) calculation due to the adoption of crop-and region-specific emission factors. Total carbon emission from cropping in our study is 178 Tg CE in 2017. The carbon footprint from cropping found in Liet al.(2022)was calculated to be 670 Tg CO2-eq in 2018,which can be converted to 183 Tg CE. The carbon emission from fertilizers is the largest emitter and accounts for over half of total carbon emission,indicating that reducing fertilizers’ usages is the most efficient way to achieve agricultural carbon emission reduction. Additionally,CH4emission from rice paddies is the second largest emitter and accounts for 20% of total carbon emission,which illustrates that previous studies on agricultural carbon emission calculation that ignored CH4emissions from paddy fields seriously underestimated total carbon emission from cropping. The inconsistent methods and materials for agricultural carbon emission quantification lead to the incomparability and irreproducibility of relevant research. Thus,it is crucial to establish a more comprehensive,region-and crop-specific calculation system to quantify agricultural carbon emission in the future.

An inverted U-shaped relationship between farm size and total carbon emission from cropping is demonstrated in this study. The result is in contrast with the finding of the U-shaped changes in total carbon emission from cropping as farm size rises in Liu and Xiao (2020),which is probably attributed to the different methods and emission factors adopted for carbon emission quantification. Nonetheless,Xuet al.’s (2022) finding demonstrated that farm size expansion led to increasing total carbon emission from cropping,which is partially validated by our finding. This study further illustrates the nonlinear relationship between farm size and total carbon emission from cropping. The farm size positively affects total carbon emission from cropping at the current stage in China,which is consistent with the finding by Xuet al.(2022). However,total carbon emission from cropping is expected to decrease with farm size expansion when farm size exceeds 1.66 ha in our study.

Projections show that a long plateau until 2048 is observed after the peak value of carbon emission emerged in 2015 if the change in farm size follows a historical trend. The result is inconsistent with the finding by Jinet al.(2021) that China’s agricultural carbon emission peaked in 2015,which is acceptable only if all conditions remain constant. Our study introduces the uncertainties from large-scale farming transformation to China’s agricultural carbon emission projection,proposing that farm size expansion will pose challenges to agricultural carbon emission reduction in the short run but facilitate carbon emission reduction in the long run.However,this study is limited to project the changes in farm size based on historical trend. More discussions about farm size changes in future years considering exogeneous factors,such as policy interventions and urban land-use planning,are required to improve the accuracy of carbon emission projections.

5.Conclusion and implications

In 2017,China’s total carbon emission from cropping is approximately 178 Tg CE. Fertilizer usage,including its manufacturing and application,is the largest source of carbon emissions,accounting for 59% of total emission from cropping. China’s total carbon emission from cropping has an inverted U-shaped relationship with farm size,with 1.66 ha serving as the turning point,which is dominated by the inverted U-shaped fluctuations in the carbon emission from fertilizers with farm size expansion.Given the current average farm size in China is 0.24 ha,the farm size will positively affect total carbon emission from cropping in the short run. Projections present that the farm size variations will alter the declining tendency in China’s total carbon emissions from cropping after 2015. Total carbon emission will decline until 2029,then increase until 2048,and finally decline thereafter if the change in farm size follows a historical trend. Thus,the production technology adjustments resulted from largescale farming transformation will pose challenges for China’s goal of achieving a carbon emission peak by 2030. However,large-scale farming is conducive to reducing carbon emissions from cropping in the long run.

Our study reveals the threat to achieve a carbon emission peak by 2030 from the large-scale farming transition in China. This preliminary analysis provides directions for further research on agricultural carbon emission reduction. Additionally,the findings could help policymakers and farm operators recognize the uncertainties of agricultural carbon emission reduction due to the large-scale farming transformation,and formulate corresponding policies and measures. Although large-scale farming is conducive to reducing total carbon emission from cropping in the long run,environmental regulations to reduce the usages of agrochemicals,especially fertilizers should be strengthened to decrease carbon emissions in the early stage of transforming to large-scale farming in China. Furthermore,our findings are informative to agricultural carbon emission reduction for other developing countries with small farm size,such as India,Vietnam and Indonesia.

Acknowledgements

The authors gratefully acknowledge the financial support from the Natural Science Foundation of China–Bill &Melinda Gates Foundation Joint Agricultural Research Project (NSFC–BMGF;72261147758),the National Social Science Foundation of China,the China Resource,Environmental and Development Research Institute,Nanjing Agricultural University,China and the Research Funding Project of Anhui Agricultural University,China(rc402108).

Declaration of competing interest

The authors declare that they have no conflict of interest.

Appendicesassociated with this paper are available on https://doi.org/10.1016/j.jia.2023.08.006