Economic Assessment of Food Legumes Breeding in China: Evidence Using a Provincial Level Dataset

Abstract

Advances in crop breeding techniques and economic evaluation are critical to ensuring and improving crop yields and sustainable development. Based on the provincial data on food legumes breeding (FLB) in China from 2001 to 2020, the research and development (R&D) expenditure distribution, FLB contribution rate, and R&D expenditure on FLB were estimated. The economic benefits of output per unit area and R&D expenses were evaluated, and relevant policy suggestions were put forward. The study found that: (i) the R&D expenditure on FLB’s showed a significant upward trend, and the development can be divided into a starting, stable, and rapid growth period. There were significant differences in R&D expenditures across eight provinces of China with relatively high R&D expenditures for FLB; (ii) the R&D expenditure on FLB had a significant lag in the improvement of per mu yield (1 mu = 0.067 hectares). For every 10,000 yuan increase in R&D funding for FLB, the yield per mu will increase by 0.145 kg/mu in the next five years, and the regional spillover effect of breeding costs is significant; and (iii) the marginal revenue of R&D expenditures FLB from 2008 to 2020 is 34.91 yuan, that is, every 1 yuan invested in breeding R&D can bring a short-term marginal revenue of 34.91 yuan and a long-term marginal revenue of 337.23 yuan. Based on the above analysis, some recommendations were proposed and discussed, including further increasing the R&D investment of FLB, strengthening industry–university–research cooperation in breeding, improving the breeding research system, and promoting the multi-dimensional development of FLB industry–university–research services.

1. Introduction

In China, food legumes, also known as edible beans, refer to all kinds of edible legume crops except soybeans and peanuts, the seeds and pods of which are mainly harvested for human consumption or the production of animals and plants [1,2]. According to the United Nations Food and Agriculture Organization statistics, almost all countries and regions in the world are planting edible beans, with an annual planting area of about 89.064 million hectares and a total output of about 88.4 million tons [3,4]. Food legumes are known as a superfood because they are nutrient-dense and high in protein, making them an excellent source of plant protein. Beans are rich in trace elements, free of cholesterol and gluten, low in fat, and an important source of dietary fiber. Food pulses have a place in the “food basket” of the United Nations World Food Programme’s food aid strategy. In addition, nitrogen-fixing bacteria can also be produced in the root nodules of beans, providing nitrogen for plants and improving soil fertility [4,5].

Therefore, consuming pulses not only plays a key role in addressing food security and ensuring a healthy and balanced diet for all, but also contributes to sustainable agriculture and food systems. To this end, the United Nations declared 2016 the International Year of Pulses (IYP), intending to raise public awareness of the nutritional value of pulses. Recognizing the potential role of pulses in advancing the 2030 Agenda for Sustainable Development, particularly the Sustainable Development Goals, the United Nations General Assembly proclaimed 10 February as World Pulses Day [6]. The FAO Director-General pointed out that the sale price of pulses is higher than that of other major crops, providing small farmers with good opportunities to grow cash crops, while also helping to achieve environmental and biodiversity goals [7,8]. China is the fourth largest producer of edible beans after India, Canada, and Myanmar in the world, accounting for about 5.6% of global production [1]. Beans are also traditional medicinal and food homologous crops in China, which play a vital role in improving people’s nutrition, promoting the adjustment of planting structure, and adapting to the challenges of climate change.

In recent years, with the support of scientific research funds such as the National Key Research and Development Programs (NKPs), and through joint research in multiple ecological regions and fields, great progress has been made in the breeding of edible legumes. Accurate identification of agronomic traits of more than 30,000 edible leguminous germplasm resources, screened or screened out a batch of excellent germplasm with significant resistance to beetles, brown spot and bacterial blight, drought resistance, and salt stress resistance resource created. Several gene loci for yield-related traits were mapped or cloned with significant resistance to drought, stress, plant diseases, and pests. Genomics-related research has entered the international frontier.

First, a world high-density haplotype map of common bean was constructed, and the high-quality reference genome of mung bean and the first pan-genome map of mung bean were assembled. Second, the breeding technology was continuously updated. China’s edible bean breeding research work began in the late 1970s, mainly focusing on the purification and rejuvenation of local varieties, systematic breeding, and introduction of foreign excellent varieties. Gradually, crossbreeding replaced systematic breeding and developed into the dominant breeding method. With the development of molecular markers, molecular breeding systems such as anthracnose resistance, mung bean beetle resistance, and a low tannin broad bean were established, which has promoted the improvement of the level of edible bean breeding to some extent.

Third, new varieties were chosen, such as Zhonglv, Jilv, and other mung bean series, Longyun, Pinjinyun and other common bean series, Zhonghong, Jizhong and other small bean series, Yunwan and other edible bean varieties. The breeding of new varieties effectively alleviated the problems of low yield, poor resistance, and poor quality of omnivorous legume varieties, and improved the competitiveness of high-quality provenance of omnivorous legumes. The promotion and application of new legume varieties and the contribution rate of improved edible legume varieties to industrial development have played an important role in the layout of major producing areas and the adjustment of industrial structure.

The continuous advancement of breeding technology is an important driving force for improving crop yields [9]. Studies have found that improvements in breeding techniques can significantly improve the adaptability of crops to biotic (disease and insect pests) and abiotic (soil, terrain, and climate) factors [10], and improve crop economic performance, and nutritional levels [11]. Compared with agrochemical technology and modern technology, the application threshold of farming technology is lower [12,13], which can reduce the input of labor, machinery, and energy in agricultural production. At the same time, compared with traditional pesticide plant protection applications, the selection of insect-resistant varieties has significant advantages in terms of insect-resistant effects and cost savings [14]. In addition, this breeding technique can significantly improve the adaptability of edible beans to drought and heat stress [15,16], and resistance to pests [17], thereby effectively improving crop yield. In recent years, with the promotion of breeding technology, the yield per unit area of edible beans in China has steadily increased, from 102.99 kg/mu in 2002 to 127.52 kg/mu in 2020, an increase of 1.24 times per mu, with an average annual growth of 1.47%. Especially since 2011, the yield of edible beans per mu has increased steadily and improved from 110.53 kg/mu in 2011 to 127.52 kg/mu in 2020, with an annual growth rate of 1.63%.

Currently, with the improvement of people’s living standards and the increasing emphasis on health, people increasingly hope to improve their dietary structure and protein intake by increasing the consumption of beans. Therefore, the consumption of pulses is becoming more and more widespread, and the consumer group is also increasing. The Chinese government has also issued a series of documents to increase support for the soybean industry. For example, in 2019 the Central Committee of the Communist Party of China proposed to actively develop the grain and soybean industry. Therefore, how to use advanced science and technology to improve the breeding level of edible beans will be an important, challenging, and strategic research topic. However, there are few kinds of literature on the economic evaluation of food legumes breeding (FLB) in China. Based on the survey data on the capital input and output of edible legume farming units in China since 2000, the regional distribution of research and development (R&D) expenditures for edible legume farming was systematically analyzed, and the contribution rate and economic benefits of edible legume farming were obtained. Legume cultivation units were estimated. The impact of FLB expenditure on its yield per unit area was researched and developed to make relevant recommendations. It is believed that these achievements will provide an important reference for further improving the breeding level of edible beans in the world including in China.

The remainder of this paper proceeds as follows. Section 2 presents the literature review. Section 3 reports methods and data sources. Section 4 deliberates the empirical results of the study, and the final section presents conclusions and discussions.

2. Literature Review

Increasing cash crop production (food legumes) to increase household well-being has been at the heart of the food policy debate in various developing countries [18,19]. Numerous findings have revealed that crop production could be a real technique to expand household economics and farming expansion [20,21]. For instance, Christiansen et al. [22] study of coffee farmers in Tanzania shows that in the event of health and drought shocks, growers continued economically strong contrasted to other major crops, which had a positive influence on agricultural growth and economic well-being of rural farmers. Kennedy et al. [23] stated that crop-sharing program increases household income in Africa and Southeast Asia Finnis [24]. It was later introduced that in southern India, farmers who grow cash crops received higher financial and social assistance, while in Malawi, households who chose to grow crops also had significantly higher incomes than those who did not grow cash crops [25].

Many documented channels to support crop production can expand the economic income of households. First, cash crop production can be a practical way to increase family farming income [26,27]. Particularly cash crop (staple crops) typically results in higher income per unit of land, counting water, land, skills, and labor input. Second, rising cash crop production to bits helps to diversify household life, thereby further increasing household resilience to economic shocks and other climate-related shocks. For example, several studies have found that crop diversification, such as intercropping and crop rotation, can improve the resilience of family farming production [28,29,30,31]. Third, the crop production assistances also advantage other non-cash-crop farmers through the influence on employment because most crop production is labor-intensive [32,33]. Increased labor demand for high-value crops is likely to raise average wages for non-cash crop growers. In addition, the introduction of cash crop opportunities has shown that households can reduce cash constraints and be able to purchase better crop production inputs [34].

Thus, improving their ability to adapt to yield enhancement techniques, information technology, and agronomic practices [31,35]. These cash profits ultimately provide growers with the opportunity to capitalize and expand farm management, thereby inspiring agricultural innovation and cumulative outcomes [31,36,37]. However, there are reasons to question the positive impact of crop yields and technology adoption on household economic well-being [26,38]. For example, new experience shows that crop production has failed to improve household economic well-being in some developing countries, especially the poorest households, due to high barriers to entry [39]. They found that promoting crop production did little to expand the poorest living standards and that these households were often barred from participating in the production of crops [40]. In China, prior findings on cash crop production have focused on two aspects. The literature focuses on the concept and theoretic basis of growers’ crop production selection and its effective mechanism [41]. Another study on the production of crops focused on how crop production influences household labor sharing, successive household relocation decisions, and other non-economic outcomes such as ecological and environmental results [42,43].

While there are good indications that FLB can be an effective way to promote the economic well-being of families [40,44], it is not clear to what extent and under what circumstances cash crop production can achieve desirable results in micro-households level [25,45]. In addition, farmers’ decisions on edible soy production are increasingly influenced by perceived hazards from environmental change [46,47] and for the production of soybean in China, farmers’ insights have improved due to repeated vicissitudes in temperature, which directly affects farmers’ predictable soybean profits and economic assistance [48]. In response to this heightened awareness of climate-related risks and the observed adverse effects on soybean production, farmers may consider other alternatives and more resilient crops to address this potential negative impact [49]. As assumed by Asrat and Simane [50] and Ojo and Baiyegunhi [51], adaptation to climate change involves a multi-step process in which strong perceptions must be credible and, subsequently, appropriate on-the-spot responses to these changes can be introduced. Approximate findings have detected this association [46,47,52,53,54,55] and decided that household adaptation to climate change behavior is directly associated with its perception [56]. Nevertheless, there is an inadequate investigation of the collective influences of edible bean farming and its effect on commercial and farming growth [57,58] and financial wellbeing [51]. It can be tested from many experiences and lessons, the important role of the FLB economic evaluation survey to increase farmers’ income and China’s agricultural development.

3. Methodology

3.1. Methods

3.1.1. Global Moran’s I Index

The global Moran’s I index is used to examine the degree of spatial aggregation of the spatial sequence ${\left\{x_i\right\}}_{i=1}^n$. Before establishing a spatial econometric model, it is usually necessary to test whether the explanatory variables or the explained variables have spatial autocorrelation through the global Moran’s I index. If the explanatory variable or the explained variable has significant spatial autocorrelation, it means that a spatial econometric model can be established to estimate it; otherwise, other econometric methods should be selected. The calculation formula is as follows:

$$I=\frac{{{\displaystyle \sum }}_{i=1}^n{{\displaystyle \sum }}_j^n{w}_{ij}\left(x_i-\overline{x}\right)\left(x_j-\overline{x}\right)}{{{\displaystyle \sum }}_{i=1}^n{\left(x_i-\overline{x}\right)}^2}$$

(1)

where, $w_{ij}$ is the element at row i and column j in the spatial weight matrix. When the global Moran’s I index is positive, it indicates that there is spatial autocorrelation between high and high values, and between low and low values. When the global Moran’s I index is negative, it indicates a negative spatial autocorrelation between high and low values.

3.1.2. Spatial Panel Model Building

The spatial panel model considers the spatial dependence of the cross-sectional units of panel data, which is helpful to measure interactions among regional economic behaviors and makes the estimated results of the model more accurate. Thus, the adjacency matrix is introduced to build the spatial panel model:

$$\{\begin{array}{c}{y}_{it}=\tau y_{i,t-1}+\rho w_i^{’}{y}_t+x_{it}^{’}\beta +d_i^{’}{X}_t\delta +u_i+{\gamma }_t+{\epsilon }_{it}\\ {\epsilon }_{it}=\lambda m_i^{’}{\epsilon }_t+v_{it}\end{array}$$

(2)

where W is the spatial weight matrix built according to the geographical position of each province, $w_i^{’}$ is row 3 of the spatial weight matrix W, $y_{jt}$ represents the matrix composed of per mu yield of beans, $w_i^{’}={{\displaystyle \sum }}_{j=1}^n{w}_{ij}{y}_{jt}$, $y_{i,t-1}$ is the first-order lag term of the explained variable, $\rho w_i^{’}{y}_t$ and $d_i^{’}{X}_t\delta$ and represents the spatial lag terms of the explained variable and the independent variable respectively, $u_i$ is an individual effect, ${\gamma }_t$ is time effect, and ${\epsilon }_{it}$ is the disturbance term changing with individual and time, with t = 19 and i = 31. When $\lambda$ = 0, this model is called the spatial Durbin model (SDM). When $\lambda$ = 0, it indicates that explained variable has a spatial lag term and the explanatory variable does not. Then, the spatial Durbin model can be simplified into a spatial autoregressive model (SAR). When $\delta +\beta ·\rho =0$, there is no spatial lag term in the explanatory variable and the explained variable, but their random error terms have a spatial effect, and the spatial Durbin model degenerates into a spatial error model (SEM).

3.2. Data Sources

From a global perspective, since the 19th century, scientific breeding has begun to rise, crop breeding has gradually developed into a relatively independent industry, and the degree of commercialization of seeds has been continuously improved. In the 1960s, in India, Pakistan, Mexico, and other regions of the world, the “green revolution” based on seed improvement technology to increase crop yield successfully verified the outstanding contribution of breeding technology to crop yield improvement, and further strengthened the following consensus: the public sector should bear an important responsibility in aquaculture [59,60]. However, cereal crops have been the focus of R&D expenditures in breeding programs, while edible legumes have been lacking in R&D expenditures [61]. At present, large-scale breeding enterprises pay more attention to highly commercialized crops such as corn and soybeans [62], and increasing public financial support plays a key role in maintaining the sustainable and healthy operation of the breeding industry. FLB Systems [63]. Compared with private investment, public investment is not limited to “profit-seeking”, and multi-dimensional considerations are added in terms of resistance to pests and diseases, drought resistance, etc., which will provide support for the comprehensive development of food and legume breeding technology [64].

The funding source of food legume breeding in China is no exception, and public funds are the main source of FLB in China. According to the survey of this study, since 2000, the main source of R&D expenditure on FLB in China is the special fund investment of national and local financial science and technology plans. The national financial projects mainly include Special Scientific Research Projects for Public Welfare Industry (agriculture), China Agriculture Research System of MOF and MARA-Food Legumes, National Key Research and Development Program, National Natural Science Foundation of China, etc. Shaanxi, Hebei, Jiangsu, Qinghai, Jiangsu, Shanxi, Guizhou, Yunnan, Hubei, and other provinces have launched financial programs for food legumes. The Ministry of Agriculture and Rural Affairs has already launched the germplasm resources protection and utilization programs. A few other prefecture-level cities, such as Tangshan in Hebei Province, Linxia in Gansu Province, and Dingxi in Gansu Province, have set up FLB programs. Since 2013, the Chinese Academy of Agricultural Sciences (CAAS) has continuously funded the research and development of FLB through the Innovation Project of the Chinese Academy of Agriculture Science (CAAS) and the Basic Scientific Research Fund of the CAAS. As far as the R&D expenditure of FLB is concerned in this paper, the research period is from 2001 to 2020, and the research space is 31 provinces, autonomous regions, and municipalities in China (except Hong Kong, Macao, and Taiwan in China). The provincial R&D expenditure data of the FLB was refined and analyzed according to when and where the above research plans were implemented.

4. Results of the Study

4.1. R&D Expenditure in FLB Shows a Significant Growth

On the whole, the R&D expenditure on FLB in China showed a significant upward trend. Especially since the establishment of the China Agriculture Research System of MOF and MARA-Food Legumes in 2008, the R&D expenditure on FLB has been strengthened, driving the local financial investment in the research and development of FLB. From 2008 to 2020, the investment in the research and development of FLB increased by 1.96 times in China, with an average annual growth rate of 5.33%. A rapid growth period of the R&D expenditure of FLB was found during 2017–2020, with an average annual amount of RMB 24.58 million yuan and an average annual growth rate of 4.58%. Significant growth in the R&D expenditure of FLB at this stage was caused by obtaining the China Agriculture Research System of MOF and MARA-Food Legumes and the financial support of the National Key Research and Development Program. The annual R&D expenditure of FLB in China had noted differences among provinces, of which eight provinces of Qinghai, Hebei, Yunnan, Heilongjiang, Jiangsu, Shanxi, Shaanxi, and Jilin had their R&D expenses for FLB higher than the national average level.

From 2010 to 2020, there was also a clear regional shift in FLB, and R&D spending, with spending in Northeast China, especially Jilin, Heilongjiang, and Hebei, increasing in 2020 compared to spending in 2010 and 2015. Compared with 2010 and 2015, R&D spending in South China decreased in 2020. R&D spending in the western and central regions remained stable from 2010 to 2020 (see Figure 1).

4.2. Investigation of Contribution of FLB R&D Funds to Yield per Unit Area

The correlation between the R&D expenditure and the per mu yield of FLB was analyzed by applying a fixed effect. As shown in

Table 1. Fixed effect panel data model estimation of average yield per mu of food legumes in 31 provinces during 2002–2020.

Variable	Coefficient	Standard Error	t Value	p Value
L. Average yield per mu of food legumes	0.181	0.0351	5.15	0.000
L3. R&D expenditure	0.141	0.0644	2.19	0.029
L5. R&D expenditure	0.145	0.0623	2.32	0.021
Constant term	95.74	4.757	20.12	0.000

Notes: $R^2=0.143$, and the sample size is 434. L, L3, and L5 denote the lag of 1, 3, and 5 years respectively.

, the R&D expenditure had a significant lag in enhancing the per mu yield of food legumes, and the contribution of the R&D expenditure of food legumes after 5 years of spending to the yield per unit area was slightly higher than that after 3 years of spending, that is to say, R&D expenditure of the former 5th-year increase 10 thousand yuan could increase current year’s yield by 0.145 kg/mu (see Table 1). Moreover, before building the spatial panel model, the spatial correlation of food bean yield per mu in 31 provinces was tested. Test results indicated that a positive spatial auto correlation was recorded among the per mu yield of food legumes in 31 provinces, and the average yield per mu of food legumes was affected by those in neighboring provinces (see

Table 2. Test results of the spatial panel model.

Test	Hausman	Wald-Err	Wald-Lag	LR-Err	LR-Lag
chi2(1) chi2(1)	2.78	7.79	6.63	7.83	6.64
Prob > chi2 Prob > chi2	0.096	0.005	0.010	0.005	0.010

).

Furthermore, to analyze the FLB effect of expenditure on yield, an adjacency matrix, and a spatial panel model were introduced (see Equation (2)). R&D expenditure is the explained variable and yield of FLB is the independent variable. The estimation of the fixed effect space Durbin model display that after five years of expenditure, the R&D expenditure effects a significant positive impact on the per mu yield of beans. The total effects in the spatial econometric model can be divided into direct effects and indirect effects (spatial spillover effects). Wx in

Table 3. SDM estimation of average yield per mu of food legumes in 31 provinces from 2002–2020.

Average Yield per mu of Food Legumes	Coefficient	Standard Error	Z Value	p Value
Main L5. R&D expenditure	0.133	0.0602	2.22	0.027
Wx L5. R&D expenditure	0.253	0.0983	2.58	0.010
Spatial rho	0.118	0.0725	1.63	0.104
Variance sigma2_e	646.7	43.9800	14.71	0.000
Direct effect	0.143	0.0618	2.32	0.020
Indirect effect	0.297	0.1030	2.87	0.004
Total effect	0.440	0.1210	3.65	0.000

Notes: $R^2=0.105$, and the sample size is 434. L5 denotes the lag of 5 years.

is the effect of changes in local R&D spending on production levels in neighboring regions. After the total effect is decomposed, it is found that after five years of investment, the direct effect, spatial spillover effect, and total effect of R&D investment on the average yield of grain and beans are 0.143, 0.297, and 0.440, respectively, reaching 5% and 1% in the significance test. This displays that the increase in R&D funds for FLB has not only had a positive impact on the per mu yield of local legumes but also has a positive effect on the improvement of grain yield per mu.

4.3. Economic Return on the R&D Expenditure of FLB

In order to verify the positive and the long-term effects of R&D expenditure of FLB on the yield of food legumes, three models were built, respectively. The three models are in the format of a fixed-effects panel data regression model. The main advantage of fixed-effects models is that they control for all time-invariant omitted variables. The model was used to estimate the impact of R&D expenditure on grain and pulse production (see

Table 4. Estimation of the yield of food legumes in 31 provinces during 2002–2020.

Variable	Model 1	Model 2	Model 3
L. Food legumes yield	0.9270 ***	0.9310 ***	0.9240 ***
	(0.0122)	(0.0132)	(0.0136)
R&D expenditure	0.0146 ***		0.0126 **
	(0.0045)		(0.0060)
L5. R&D expenditure		0.0179 ***	0.0086
		(0.0052)	(0.0068)
Constant	0.00245	−0.0286	−0.245
	(0.2760)	(0.2630)	(0.2820)
Observations	558	434	434
Number of prov	31	31	31

Notes: ** and *** represent significant levels of 5% and 10% respectively.

). In model 1, the influence of the R&D expenditure of FLB on its yield was only considered. In model 2, the influence of the R&D expenditure of FLB on its yield after 5 years of spending was analyzed. In model 3, the influences of R&D expenditure of FLB on its yield in the current spending year and after 5 years of spending were comprehensively considered. In model 1, the estimated coefficient of R&D expenditure on the yield of food legumes was 0.0146 and reached a level of 1% in the significance test, indicating that the R&D expenditure on FLB had a positive impact on its yield growth. In model 2, the estimated coefficient of R&D expenditure on the yield of food legumes after five years of spending was 0.0179 and reached a level of 1% in the significance test, indicating that the R&D expenditure not only had a current temporary effect on the improvement of food legumes yield but also had a long-term effect, which was consistent with the spatial autocorrelation test based on the global Moran’s I index.

Based on the above parameters, the total, the marginal, and the long-term marginal revenues of R&D expenditure of FLB were further calculated. The marginal revenue of the R&D expenditure of FLB represented the added value of the total revenue brought by the R&D expenditure of RMB 1 yuan. By multiplying the elastic value of R&D expenditure in a certain year by the total output value of food legumes in a certain year in the whole country or a certain region (see

Table 5. Revenue of R&D expenditure of FLB during 2008–2020.

Year	R&D Expenditure of FLB (RMB 10 Thousand Yuan)	Total Revenue (RMB 10 Thousand Yuan)	Marginal Revenue (RMB 10 Thousand Yuan)	Long-Term Marginal Revenue (RMB 10 Thousand Yuan)
2008–2020	63	11020	34.91	337.23

Notes: Average service life defined for agricultural research achievements in this study was 6 years, and long-term marginal revenue of R&D expenditure = marginal revenue × average service life of research achievements × (1 + annual growth rate of R&D expenditure).

), it could be seen that from 2008 to 2020, the marginal revenue of R&D expenditure of FLB was RMB 34.91 yuan, namely RMB 1 yuan invested in the research and development of FLB could bring a marginal revenue of RMB 34.91 yuan, similarly, a study indicated a $28 of benefits to pulse growers for each dollar invested into genetic improvements in Canada [65]. In the long run, the marginal revenue of R&D expenditure for FLB was RMB 337.23 yuan, higher than the marginal revenue (RMB 320 yuan) of the general agricultural R&D expenditure [66].

5. Conclusions and Discussion

Advances in breeding technology are a significant part of food security, and adequate investment is key to sustainable breeding systems. The study found that the absolute value of R&D expenditures on bean breeding in China showed an obvious upward trend, and there were significant differences among provinces. The research and development expenditure of legume breeding in China lags behind the increase of legumes per mu. After five years of spending, R&D spending on grain and bean breeding contributed slightly more to its per mu yield than spending over three years. The increase in research and development funds not only has a positive impact on the local grain yield per mu but also promotes the increase in the per mu grain yield in the surrounding areas. R&D expenditures in FLB also lead to high returns on investment. From 2008 to 2020, the marginal income of bean planting R&D expenditure was 34.91 yuan, and the long-term marginal income of bean planting R&D expenditure was 337.23 yuan, both higher than the agricultural average. R&D spending. Based on the above important findings, to further promote the scientific progress of edible bean breeding in China, the following suggestions were put forward and discussed:

The first suggestion is to further increase the investment in FLB research and development, and continuously strengthen long-term and stable financial support for basic research and key breeding projects of edible legumes. Some scientists divide the development of breeding into four eras, namely the era of farmers’ selection, the era of phenotypic selection, the era of molecular breeding, and the era of big data intelligent design breeding. At present, the world-class seed industry has entered the era of design breeding 4.0, while FLB is still in the stage of 2.0 to 3.0. Therefore, some measures should be taken to promote the transformation of R&D technology to intelligent precision breeding, including: (i) making full use of public, social and financial capital; and (ii) strengthening R&D expenditures of industry–university–research consortia to develop a good breeding foundation and strong strength, and (iii) consolidating the basic theory and core technology of breeding research and development.

Secondly, academia–industry collaboration and their joint breeding research should be strengthened. Enterprises will be the innovation entity of the food legumes seed industry in the future. However, for minor crops like food legumes, the main force of genetic breeding innovation and research accumulation are still in research institutions. The academia–industry collaboration should be encouraged from the angle of policies to: (i) promote cooperation between the seed enterprises and the research institutions including universities; (ii) form regional innovation advantages; (iii) improve the regional promotion level of food legumes varieties; and (iv) form the clustering effect of regional breeding.

Third, more efforts should be made to improve the FLB research system and to promote the multi-dimensional development of breeding industry–university–research-service in order to clarify the innovation goals and varieties of FLB that need to be broken through, keep up with and lead the market demand, and change to the mode of “delicious and easy to use” and “improving quality and efficiency”. The research and development funding and management system of the food, bean, and seed industry should be straightened out, and the implementation of a commercial FLB system of “questions issued by enterprises + answers by scientific research institutions + enterprise evaluation + expert post-event provision” should occur. A “sales service” should be formed at a faster rate. In addition, a joint FLB investment mechanism should be established between the government, financial institutions, and enterprises to promote: (i) the subsidy target from “there is a clear”; (ii) the research and development model changes from “directed research and development” to “independent research and development”.