Introduction

Since the reform and opening up, China has made remarkable progress in agricultural development, ensuring food security and increasing agricultural output. However, the long-term model of high-input, high-pollution, and high-emission extensive agriculture has led to significant issues, as inevitably faced by other developing countries, including soil degradation, agricultural pollution, and slow improvements in product quality1,2. To address these challenges, the government has identified high-quality agricultural development (HQAD) as a breakthrough. As China shifts from a major agricultural producer to an agricultural powerhouse, advancing HQAD has become an inevitable and critical task.

Academic research plays a vital role in advancing social practice. Consequently, HQAD has not only become a key government policy but has also a prominent research topic. The design of a comprehensive and systematic evaluation framework has garnered particular attention, which is essential for accurately evaluating HQAD effectiveness and guiding future development. From a research perspective, previous studies have mainly evaluated HQAD from two perspectives: (1) The economic perspective, which focuses on optimizing the allocation of economic production factors and employs total factor productivity (TFP) as a key indicator of agricultural growth rates and production efficiency1,3. (2) The perspective that views HQAD as an extension of the high-quality development concept. This has led scholars to develop evaluation frameworks based on five dimensions: innovation, coordination, green, openness, and sharing. Regarding research areas, studies have primarily focused on either the national level3,4 or specific regions, such as the Yangtze River Economic Belt in China5,6. In terms of research methods, TOPSIS3,5 and Hierarchical Analysis4 have been utilized.

Despite the development of several evaluation frameworks for HQAD, deficiencies persist. From a research perspective, some indicators are too broad to capture the specific nuances of HQAD4,7 For instance, although the water quality compliance rate is crucial for green agriculture, it is also influenced by activities in other industries, such as services and manufacturing. Additionally, indicators like green space areas fail to reflect the input factors critical to agricultural production5. Regarding research areas, provincial-level data provides an overview of the national HQAD. However, China’s vast territory and regional disparities in natural conditions, resource endowments, and economic development make direct comparisons among provinces problematic. Therefore, using prefecture-level city data, with provinces as the research units, could provide more meaningful insights. Heilongjiang Province presents a unique opportunity to address these research gaps.

As China’s primary commodity grain base and strategic grain reserve region, Heilongjiang Province plays a crucial role in the nation’s overall HQAD process8. It is also a key factor in achieving the country’s goal of becoming an agricultural powerhouse. Therefore, accurately and scientifically measuring the level of HQAD in Heilongjiang Province has become an urgent issue. However, focusing solely on its overall development level is insufficient to capture its developmental trends and spatial distribution. The existence of spatial disparities in HQAD among Heilongjiang’s four regions and its prefecture-level cities, along with identifying their underlying causes, requires further exploration. Although these regions share similarities in resource endowments and policy support, significant differences persist. These disparities may lead to varying developmental paths for HQAD across regions. A detailed analysis of these regional disparities is essential for formulating more tailored and effective regional development policies. Furthermore, while trend analysis can provide some insights into the current state of HQAD in Heilongjiang, it is inadequate for long-term planning. To better guide future strategic planning, accurately predicting future developmental trends is essential.

Based on this, we employ the Entropy Weight-TOPSIS model, the Dagum Gini Coefficient and its decomposition, and the Markov Chain, with prefecture-level cities in Heilongjiang Province as the research units to analyze the level of HQAD, regional disparities, and dynamic evolution from 2013 to 2022. First, grounded in a comprehensive understanding of HQAD and the new development philosophy, we construct an evaluation framework from four dimensions: green agriculture, quality and efficiency improvement of supply, agricultural structural coordination, and shared benefits of development. Second, the cities are integrated into four major regions based on agricultural endowments: the Xing’an Mountains region, the Semi-mountainous Areas of Mudanjiang, the Sanjiang Plain, and the Songnen Plain. The Dagum Gini Coefficient is applied to assess regional disparities in HQAD within Heilongjiang Province and identify the primary sources of spatial disparities. Third, the Markov Chain is employed to simulate and predict the dynamic evolution of HQAD within Heilongjiang Province. Finally, we propose policy recommendations for the Heilongjiang government, other similar regions, and policymakers to achieve comprehensive HQAD, addressing both regional and national challenges.

The potential contributions of this study are as follows: (1) Drawing on existing literature and grounding in the new development philosophy, this study constructs a more scientific evaluation framework for HQAD. This framework addresses the limitation of current indicators, which are too broad to accurately capture HQAD levels. (2) Based on data from prefecture-level cities, this study measures the level of HQAD in Heilongjiang Province. As a critical contributor to national food security, Heilongjiang provides more detailed empirical evidence for HQAD in China. This fills a gap in previous research that relied on provincial data with varying resource endowments, leading to general assessments of national agricultural development. (3) The study employs the Entropy Weight-TOPSIS Model to evaluate HQAD levels and provides a detailed analysis. Compared to traditional TOPSIS methods, this model reduces the impact of subjective factors, thereby enhancing the scientific accuracy of HQAD assessment.

Methodology

Study area

Heilongjiang Province (Fig. 1) is located in northeastern China and comprises 12 prefecture-level cities, including Harbin, Daqing, Hegang, etc. As one of China’s 34 provincial-level administrative regions, the province has a total arable land area of 257.9 million acres, accounting for more than 1/9 of the nation’s total, and ranking first nationwide. Situated in one of the world’s three primary black soil zones, Heilongjiang Province has a typical black soil area of approximately 10.4 million hectares, representing 56.1% of northeast China’s total black soil area. The province’s flat terrain is ideal for large-scale mechanized farming. Additionally, Heilongjiang possesses numerous rivers and a well-developed water system with excellent water quality, making it the most water-rich province in northern China. The cold temperate and temperate continental monsoon climate is highly conducive to crop growth. As a result, Heilongjiang has become one of the most important commodity and strategic reserve grain bases in China, which is crucial for safeguarding national food security. In 2023, Heilongjiang Province achieved a total grain output of 77.88 million tons, maintaining its position as the top grain-producing province in the country for the 14th consecutive year.

Fig. 1
figure 1

Location of the study area. This map is created by the authors with ArcGIS software (version 10.2 ESRI, Redlands, CA, USA, http://www.esri.com/software/arcgis/arcgis-for-desktop).

Full size image

Data

The latest panel data of 12 prefecture-level cities in Heilongjiang Province from 2013 to 2022 are selected to quantify and analyze the HQAD. These data were obtained from the annual Heilongjiang Statistical Yearbook (2014–2023), China City Statistical Yearbook (2014–2023), statistical yearbooks of various cities (2014–2023), and statistical bulletins on the national economic and social development of cities from 2013 to 2022. The data of the Daxing’anling region were not included in the analysis due to lack of information. Linear interpolation was used to fill in individual missing data.

Methods

A systemic integration model is constructed to evaluate and analyze HQAD within Heilongjiang Province. It contains three modules: a multidimensional evaluation model, a disparities analysis model, and an evolution prediction model. The multidimensional evaluation model undertakes the function of quantifying HQAD, the disparities analysis model analyzes regional disparities in HQAD and identifies the primary sources of spatial disparities, and the evolution prediction model predicts the dynamic evolution of HQAD.

Multidimensional evaluation model

The meaning of HQAD. The shift from quantity-driven to quality-driven development is evident in many countries. In China, the concept of “high-quality development” was first introduced at the 19th National Congress of the Communist Party of China in 2017. “High quality” serves as an evaluative standard, defining the type of development to achieve, while “development” emphasizes the goal of advancing to a higher stage. Within this framework, HQAD plays a crucial role in the agriculture9. Although there is no universally accepted definition in the academic literature, the core characteristics of HQAD fundamentally include green agriculture, quality and efficiency improvement of supply, agricultural structural coordination, and shared benefits of development.

  1. (1)

    Green agriculture. Green agriculture represents the dominant form of HQAD. The National Strategic Plan to Boost High-quality Development of Agriculture (2018–2022) emphasizes the acceleration of green agriculture to advance its high-quality development. Therefore, green development principles should be integrated throughout the entire process. This is evident in two key aspects: First, enhancing the conservation and efficient utilization of agricultural resources to address resource and environmental constraints. Second, fostering ecological agriculture, promoting sustainable practices, and reinforcing agro-ecological protection and restoration to achieve harmony between humans and nature.

  2. (2)

    Quality and efficiency improvement of supply. Achieving HQAD involves a process of improving its quality and efficiency. Scholars generally agree that HQAD fundamentally involves good product quality, sound industrial benefits, and high production efficiency10. Specifically, first, improving product quality is crucial. As consumer preferences shift from quantity to quality with socioeconomic advancements11, it becomes necessary to produce high-quality agricultural products to meet market demands. Second, enhancing industrial benefits is crucial. As they dictate income, investment, and competitiveness. Third, increasing production efficiency involves optimizing the ratio of agricultural input to output, maximizing output with minimal costs.

  3. (3)

    Agricultural structural coordination. As the inherent characteristic of HQAD, coordinated agricultural structures development is primarily reflected in agricultural industrial12. This involves the coordinated development of planting structures, variety compositions, sectoral structures, and all connections within the agricultural value chain.

  4. (4)

    Shared benefits of development. The goal of development lies the improvement of people’s livelihoods, with a focus on increasing farmers’ incomes, optimizing rural public services, refining social security systems, and ensuring equal opportunities for all to benefit from development5. Consequently, enhancing rural public service delivery, achieving stable growth of farmer income, narrowing the urban-rural gap, and promoting urban-rural integration are essential for assessing the inclusiveness of development outcomes.

Indicators. Building on existing evaluation frameworks for HQAD and those specific to agricultural modernization in Heilongjiang Province13,14,15,16, and grounded in a deep understanding of High-quality Development, we construct a multidimensional evaluation framework tailored to Heilongjiang Province. Adhering to principles of scientific rigor, systematic approach, forward-looking perspective, and data accessibility, this framework encompasses four dimensions—green agriculture, quality and efficiency improvement of supply, agricultural structural coordination, and shared benefits of development—comprising a total of 19 indicators (Table 1).

  1. (1)

    Green agriculture. As a prerequisite for HQAD, green agriculture is reflected by three indicators. Chemical fertilizer used per unit area reflects the environmental impact of chemical residues in agricultural production. Rural electricity consumption measures the consumption of electrical resources during agricultural production. Water-saving irrigation, an indicator of water conservation efforts, is crucial for ecological protection within agricultural practices.

  2. (2)

    Quality and efficiency improvement of supply. As the essence of HQAD, it is evaluated through eight indicators concerning product quality, industrial benefits, and production efficiency. Product quality is measured by the number of green food products, reflecting quality and sustainability of agricultural production. Industrial benefits are evaluated by the added value of the primary industry, which represents real value beyond material resource consumption. The stable increases in food yield per unit area and comprehensive production capacity of economic crops serve as the foundation for supply quality and efficiency. Agricultural labor productivity and land productivity indicate labor efficiency and output value from production materials, respectively. Agricultural scale and mechanization levels, indicators of agricultural intensification and mechanization, directly influence production efficiency.

    Table 1 Evaluation frameworks of HQAD of Heilongjiang Province.
    Full size table
  3. (3)

    Agricultural structural coordination. The degree of coordination within agricultural structures significantly influences HQAD. This dimension is reflected by five indicators that address agricultural industrial coordination. Agricultural industrial coordination is reflected by crop diversification, particularly the proportion of grain crops within total crop production. Additionally, the ratio of output from agriculture, forestry, animal husbandry and fishery services, the industrial structure adjustment index, and fiscal support for agriculture, and the proportion of the primary industry contribute to evaluating the coordination of agricultural industrial structures.

  4. (4)

    Shared benefits of development. As the fundamental goal of HQAD, the shared benefits of development are reflected by three indicators. Medical and health levels, per capita disposable income of rural residents, and the income ratio between urban and rural residents reflect the level of shared benefits in HQAD. The first two indicators focus on welfare distribution, while the latter highlights the equitable sharing of benefits.

Classification of HQAD levels. Based on the criteria for classifying HQAD levels outlined by Zhang16, the 80%, 100%, and 120% thresholds of Heilongjiang’s overall HQAD level for each year are used as cut-off points. HDij represents the HQAD level of city j in year i, and HQDi represents the overall HQAD level of Heilongjiang Province in year i. Consequently, the HQAD levels in Heilongjiang Province are divided into four categories (Table 2).

Table 2 The classification of HQAD levels of Heilongjiang Province.
Full size table

Entropy weight-TOPSIS model. The Entropy Weight-TOPSIS Model is employed to evaluate the HQAD levels. This method integrates the Entropy Weight method, an objective multi-index comprehensive evaluation method, with the TOPSIS method, a multi-criteria decision analysis tool. The Entropy Weight method determines the weight of each indicator based on the information content it provides17, thereby avoiding the issues of information redundancy and biased outcomes associated with subjective weighting methods like the Analytic Hierarchy Process (AHP) and expert scoring. Meanwhile, the TOPSIS model evaluates the advantages and disadvantages of each sample by measuring its proximity to the ideal solution18. We first employ the Entropy Weight method to calculate the weight of each indicator within the evaluation framework, and then utilize the TOPSIS method to calculate the HQAD level. The specific steps are as follows17:

To express the original evaluation indicator matrix X given m evaluation subjects and n specific evaluation indicators, where xij represents the original numerical data of the j index for the prefecture-level i city, the matrix can be structured as follows:

$$X = \left[ {\begin{array}{*{20}c} {x_{{11}} } & \cdots & {x_{{1n}} } \\ \vdots & \ddots & \vdots \\ {x_{{m1}} } & \cdots & {x_{{mn}} } \\ \end{array} } \right]$$
(1)

First, it needs to be standardized. The formula is as follows:

$${\text{Positive}}\;{\text{attributes}}:y_{{ij}} = {\raise0.7ex\hbox{${x_{{ij}} - \min x_{{ij}} }$} \!\mathord{\left/ {\vphantom {{x_{{ij}} - \min x_{{ij}} } {\max x_{{ij}} - \min x_{{ij}} }}}\right.\kern-\nulldelimiterspace} \!\lower0.7ex\hbox{${\max x_{{ij}} - \min x_{{ij}} }$}}$$
(2)
$${\text{Negative}}\;{\text{attributes}}:y_{{ij}} = {\raise0.7ex\hbox{${\max x_{{ij}} - x_{{ij}} }$} \!\mathord{\left/ {\vphantom {{\max x_{{ij}} - x_{{ij}} } {\max x_{{ij}} - \min x_{{ij}} }}}\right.\kern-\nulldelimiterspace} \!\lower0.7ex\hbox{${\max x_{{ij}} - \min x_{{ij}} }$}}$$
(3)

Among them, yij is the standardized value of the prefecture-level i city of the j index. After standardization, the matrix Y is obtained.

$$Y = \left[ {\begin{array}{*{20}c} {y_{{11}} } & \cdots & {y_{{1n}} } \\ \vdots & \ddots & \vdots \\ {y_{{m1}} } & \cdots & {y_{{mn}} } \\ \end{array} } \right]$$
(4)

Second, the matrix Y is normalized to get the matrix R. The formula is as follows:

$$R_{{ij}} = {\raise0.7ex\hbox{${y_{{ij}} }$} \!\mathord{\left/ {\vphantom {{y_{{ij}} } {\mathop \sum \nolimits_{{i = 1}}^{m} y_{{ij}} }}}\right.\kern-\nulldelimiterspace} \!\lower0.7ex\hbox{${\mathop \sum \nolimits_{{i = 1}}^{m} y_{{ij}} }$}}$$
(5)

Third, the information entropy ej is calculated as follows:

$$e_{j} = - K\mathop \sum \limits_{{i = 1}}^{m} r_{{ij}} \ln r_{{ij}} ,K = {\raise0.7ex\hbox{$1$} \!\mathord{\left/ {\vphantom {1 {\ln m}}}\right.\kern-\nulldelimiterspace} \!\lower0.7ex\hbox{${\ln m}$}}$$
(6)

Fourth, the weight wj of the index is calculated as follows:

$$w_{j} = {\raise0.7ex\hbox{${\left( {1 - e_{j} } \right)}$} \!\mathord{\left/ {\vphantom {{\left( {1 - e_{j} } \right)} {\mathop \sum \nolimits_{{j = 1}}^{n} \left( {1 - e_{j} } \right)}}}\right.\kern-\nulldelimiterspace} \!\lower0.7ex\hbox{${\mathop \sum \nolimits_{{j = 1}}^{n} \left( {1 - e_{j} } \right)}$}}$$
(7)

Fifth, the canonical matrix G is obtained using the vector normalization method, The formula is as follows:

$$G = \left[ {\begin{array}{*{20}c} {g_{{11}} } & \cdots & {g_{{1n}} } \\ \vdots & \ddots & \vdots \\ {g_{{m1}} } & \cdots & {g_{{mn}} } \\ \end{array} } \right]$$
(8)

gij is calculated by the following formula:

$$g_{{ij}} = ~{\raise0.7ex\hbox{${y_{{ij}} }$} \!\mathord{\left/ {\vphantom {{y_{{ij}} } {\sqrt {\mathop \sum \nolimits_{{i = 1}}^{m} y_{{ij}}^{2} } }}}\right.\kern-\nulldelimiterspace} \!\lower0.7ex\hbox{${\sqrt {\mathop \sum \nolimits_{{i = 1}}^{m} y_{{ij}}^{2} } }$}}$$
(9)

Sixth, to make the evaluation results more objective, a standardized analysis matrix Z is constructed according to the index weight wj, and the formula is as follows:

$$Z = ~\left[ {\begin{array}{*{20}c} {w_{1} g_{{11}} } & \cdots & {w_{n} g_{{1n}} } \\ \vdots & \ddots & \vdots \\ {w_{1} g_{{m1}} } & \cdots & {w_{n} g_{{mn}} } \\ \end{array} } \right]$$
(10)

Seventh, the positive ideal solution\(\:{{\:Z}_{j}}^{+}\)and the negative ideal solution \(\:{{Z}_{j}}^{-}\)represent the maximum and minimum values of the j index of the i prefecture-level city. The formula is as follows:

$$Z_{j} ^{ + } = ~\left[ {z_{1}^{ + } ,z_{2}^{ + } , \ldots ,z_{n}^{ + } } \right]\left( {j = 1,2, \ldots ,n} \right)$$
(11)
$$Z_{j} ^{ - } = ~\left[ {z_{1}^{ - } ,z_{2}^{ - } , \ldots ,z_{n}^{ - } } \right]\left( {j = 1,2, \ldots ,n} \right)$$
(12)

Eighth, euclidean distance is used to calculate the distance from the evaluation index of high agricultural quality level in Heilongjiang Province to the positive and negative ideal value. Among them, \(\:{d}_{i}^{+}\:\)is the distance between the j index and the positive ideal solution \(\:{{\:Z}_{j}}^{+}\), and \(\:{d}_{i}^{-}\) is the distance between the j index and the negative ideal solution \(\:{{Z}_{j}}^{-}\). The formula is as follows:

$$d_{i}^{ + } = ~\sqrt {\mathop \sum \limits_{{j = 1}}^{n} \left( {z_{{ij}} - z_{j}^{ + } } \right)^{2} } \;\;\left( {1 \le i \le m,~1 \le j \le n} \right)$$
(13)
$$d_{i}^{ - } = ~\sqrt {\mathop \sum \limits_{{j = 1}}^{n} \left( {z_{{ij}} - z_{j}^{ - } } \right)^{2} } \;\;\left( {1 \le i \le m,~1 \le j \le n} \right)$$
(14)

Finally, Si is the comprehensive evaluation index for the i prefecture-level city of HQAD level in Heilongjiang Province. The formula is as follows:

$$S_{i} = ~{\raise0.7ex\hbox{${d_{{ij}}^{ - } }$} \!\mathord{\left/ {\vphantom {{d_{{ij}}^{ - } } {d_{{ij}}^{ + } + d_{{ij}}^{ - } }}}\right.\kern-\nulldelimiterspace} \!\lower0.7ex\hbox{${d_{{ij}}^{ + } + d_{{ij}}^{ - } }$}}\;\;~\left( {1 \le i \le m} \right)$$
(15)

Disparities analysis model

To deepen the understanding of the sources and magnitude of regional disparities in HQAD within Heilongjiang Province, this study employs the Dagum Gini coefficient and its decomposition19. The formula is as follows19:

$${\text{G}} = {\raise0.7ex\hbox{${\mathop \sum \nolimits_{{{\text{j}} = 1}}^{{\text{k}}} \mathop \sum \nolimits_{{{\text{h}} = 1}}^{{\text{k}}} \mathop \sum \nolimits_{{{\text{i}} = 1}}^{{{\text{n}}_{{\text{j}}} }} \mathop \sum \nolimits_{{{\text{r}} = 1}}^{{{\text{n}}_{{\text{h}}} }} \left| {{\text{y}}_{{{\text{ji}}}} - {\text{y}}_{{{\text{hr}}}} } \right|}$} \!\mathord{\left/ {\vphantom {{\mathop \sum \nolimits_{{{\text{j}} = 1}}^{{\text{k}}} \mathop \sum \nolimits_{{{\text{h}} = 1}}^{{\text{k}}} \mathop \sum \nolimits_{{{\text{i}} = 1}}^{{{\text{n}}_{{\text{j}}} }} \mathop \sum \nolimits_{{{\text{r}} = 1}}^{{{\text{n}}_{{\text{h}}} }} \left| {{\text{y}}_{{{\text{ji}}}} - {\text{y}}_{{{\text{hr}}}} } \right|} {2\mu {\text{n}}^{2} }}}\right.\kern-\nulldelimiterspace} \!\lower0.7ex\hbox{${2\upmu {\text{n}}^{2} }$}}$$
(16)

where k is the number of regional divisions; µ is the provincial average; n is the sum of the number of prefecture-level cities in the region; nj, nh are the number of cities in regions j and h; yji and yhr are the HQAD levels of any prefecture-level cities i and r in regions j and h.

The overall Gini coefficient G is decomposed into the contribution of intra-regional variation Gn, the contribution of inter-regional net differences Gj, and the contribution of hypervariable density Gt, and satisfies:

$${\text{G}} = {\text{G}}_{{\text{n}}} + {\text{G}}_{{\text{j}}} + {\text{G}}_{{\text{t}}}$$
(17)

Evolution prediction model

We employ the Markov chain to construct an evolution prediction model, which can depict the dynamic evolution characteristics of HQAD in Heilongjiang Province. This is achieved by constructing a state transition matrix, specifically as follows20:

$${\text{P}}\left\{ {{\text{X}}\left( {\text{t}} \right) = {\text{j}}|{\text{X}}\left( {{\text{t}} - 1} \right) = {\text{i}}_{{{\text{t}} - 1}} , \ldots {\text{X}}\left( 0 \right) = {\text{i}}_{0} } \right\} = {\text{P}}\left\{ {{\text{X}}\left( {\text{t}} \right) = {\text{j}}|{\text{X}}\left( {{\text{t}} - 1} \right) = {\text{i}}} \right\}$$
(18)

The state of the random variable X at a future time point t is influenced solely by its state at t-1. The transition probability Pij from state i to state j can be estimated using maximum likelihood estimation as \(\:{\text{P}}_{\text{i}\text{j}}=\raisebox{1ex}{${\text{n}}_{\text{i}\text{j}}$}\!\left/\:\!\raisebox{-1ex}{${\text{n}}_{\text{i}}$}\right.\), where nij is the number of times the HQAD level transitions from state i to state j between 2013 and 2022, and ni represents the total number of occurrences of state i. Furthermore, by employing the state transition matrix P formed by the transition probabilities Pij, we can infer the dynamic evolution trend of HQAD in Heilongjiang Province.

Results

Evaluation and analysis of HQAD in Heilongjiang Province

Figure 2 illustrates the overall level of HQAD in Heilongjiang Province from 2013 to 2022, along with levels across four dimensions. As shown in Fig. 2, the HQAD in Heilongjiang Province exhibits the following characteristics:

Fig. 2
figure 2

Heilongjiang Province’s HQAD level and its sub-dimension level from 2013 to 2022.

Full size image
  1. (1)

    The overall level of HQAD in Heilongjiang Province has shown an upward trend. From 2013 to 2022, the HQAD level in Heilongjiang Province increased from 0.160 to 0.396, with an average annual growth rate of 10.595%. This indicates that over the past decade, the overall level of HQAD in Heilongjiang Province has steadily improved, driven by the implementation of national and provincial policies promoting agricultural advancement, as well as local initiatives focused on enhancing agricultural quality.

  2. (2)

    Heilongjiang Province faces significant structural challenges in its pursuit of HQAD. (1) The average level of green agriculture is 0.431, higher than Heilongjiang’s overall HQAD (0.323). This highlights its foundational role. However, the level of green development fluctuated around 0.404 until a breakthrough occurred in 2021. (2) The average level of quality and efficiency improvement in supply (0.349) and shared benefits of development (0.354) are approximately equal to Heilongjiang’s overall HQAD average (0.323). Both indicators were the key areas of transformation in Heilongjiang’s agricultural development and significant drivers of its HQAD before 2017. (3) From 2017 to 2022, agricultural structural coordination became the primary factor affecting HQAD in Heilongjiang Province. Overall, it showed a fluctuating upward trend, increasing from 0.255 in 2013 to 0.309 in 2022, still below the average level for Heilongjiang’s overall HQAD. This suggests insufficient momentum in agricultural structural reforms, which constrain the province’s HQAD. Specifically, a significant increase occurred in 2016, possibly driven by Heilongjiang’s proactive response to the central government’s call in the No. 1 Central Document of 2016, which introduced supply-side structural reforms to address challenges in agriculture, rural areas, and farmers. However, a slight decline occurred from 2019 to 2021, likely exacerbated by the COVID-19 pandemic. During the COVID-19 pandemic, the demand for food safety and self-sufficiency among consumers intensified. Additionally, the pandemic led to a labor force reversal, with some migrant workers choosing to return to rural areas21. This decline highlights vulnerabilities that require government attention, particularly regarding inadequate reform resilience and low capacity for risk resistance.

  3. (3)

    The regional disparities in the HQAD within Heilongjiang Province are substantial but gradually narrowing. As shown in Table 3, in terms of temporal trends, the HQAD levels across all prefecture-level cities have generally shown an upward trend. Regarding developmental trends, seven prefecture-level cities, including Harbin, Qiqihar, Hegang, Shuangyashan, Yichun, Jiamusi, and Suihua have experienced stable and continuous progress in HQAD. Conversely, HQAD in Jixi, Daqing, Qitaihe, Heihe and Mudanjiang has shown unstable fluctuations. Concerning development speed, Hegang experienced the highest annual average growth rate of 18.828%. As for development levels, Jixi, Hegang, Jiamusi, Heihe and Qitaihe remained at a low HQAD level in 2013, whereas all cities had almost reached at least a medium-low level by 2022. This indicates that Heilongjiang’s prefecture-level cities are advancing in unison. From a spatial perspective, Harbin and Qiqihar located in the Songnen Plain, consistently ranked the highest in average HQAD levels from 2013 to 2022. This may be attributed to their strong agricultural foundations as traditional farming regions.

Table 3 The HQAD levels of prefecture-level cities in Heilongjiang Province from 2013 to 2022.
Full size table

Additionally, they benefit from well-developed infrastructure and convenient transportation, which facilitates the circulation of agricultural products. Moreover, these cities serve as major hubs for agricultural science and technology in the province, thereby promoting HQAD. Besides these cities, Mudanjiang also ranks highly in HQAD. This is attributed to its semi-mountainous terrain, which provides diverse land resources for agricultural development. Additionally, focusing on developing specialty agricultural products or organic agriculture has enhanced its added value and market competitiveness. However, Mudanjiang’s HQAD experienced a significant decline in 2018, likely due to an adjustment period in response to national policies aimed at agricultural structural changes. Most of the cities ranked lower are located in the Sanjiang Plain, where the HQAD levels are mostly medium-low or low. These areas are part of former industrial bases. Although abundant in minerals and coal, the government has prioritized the revitalization of old industries. Consequently, the sown areas and agricultural output are relatively low14, resulting in a lower level of HQAD. Notably, despite Heihe’s initially limited agricultural resources, its HQAD has seen rapidly growth since 2018, owing to being the only agricultural green development demonstration zone in Heilongjiang Province.

Regional disparities and their sources in HQAD within Heilongjiang Province

To examine the sources and extent of regional disparities in HQAD levels within Heilongjiang Province from 2013 to 2022, we employed the Dagum Gini coefficient and its decomposition. Prefecture-level cities are integrated into four major regions based on agricultural endowments22: the Songnen Plain (Harbin, Qiqihar, Suihua, Daqing), the Sanjiang Plain (Jixi, Hegang, Shuangyashan, Jiamusi, Qitaihe), the Mudanjiang Semi-mountainous Areas (Mudanjiang), and the Xing’an Mountains Regions (Heihe, Yichun). The intra-regional and inter-regional Gini coefficients within these four regions are calculated separately, thereby measuring the contribution rates of intra-regional Gini coefficients, inter-regional Gini coefficients, and hyper-kernel density to the overall Gini coefficient. Notably, the Semi-mountainous Areas of Mudanjiang, with only one city, are excluded from the calculation of intra-regional disparities.

It is important to note that due to inherent disparities in natural resources and geophysical conditions among the prefecture-level cities in Heilongjiang Province, their developmental priorities vary significantly. For instance, Hegang and Jixi primarily focus on coal mining, while Yichun prioritizes forestry. As these cities do not primarily rely on agriculture, their HQAD levels inevitably lag behind those of agricultural cities. Nevertheless, the assessment of regional disparities and their sources in Heilongjiang’s HQAD is of significant importance. First, promoting high-quality agriculture benefits both agricultural and non-agricultural cities by driving urban transformation and rural revitalization. By analyzing intra- and inter-regional disparities and their evolutionary trends, we can identify the changing patterns of agricultural development disparities among cities, thereby providing tailored strategies to steadily improve HQAD. Second, identifying the primary sources of these disparities allows us to determine the key factors influencing the uneven development of high-quality agriculture in Heilongjiang, ultimately proposing strategies to achieve more balanced HQAD across the province.

Overall disparities in HQAD levels of Heilongjiang Province

Over the past decades, the overall regional disparities in HQAD in Heilongjiang Province have shown a narrowing trend. As depicted in Fig. 3, the overall disparities decreased from 0.201 in 2013 to 0.072 in 2022, indicating an average annual decline rate of 10.849%. Additionally, the overall disparities exhibit a pattern of “sharp decline – fluctuating decline.”

Fig. 3
figure 3

Overall disparities in HQAD levels of Heilongjiang Province from 2013 to 2022.

Full size image

Intra-regional disparities in HQAD levels of Heilongjiang Province

As shown in Fig. 4, the intra-regional disparities in the Songnen Plain generally exhibit an expanding trend, characterized by a pattern of “synchronized progress - sharp rise - sharp decline - sharp rise.” We propose that under various agricultural support policies, prefecture-level cities in this region advanced together from 2013 to 2017, gradually reducing disparities in HQAD. However, Daqing’s economy has long been dominated by the oil industry, resulting in limited investment and attention to agriculture. Therefore, after the high-quality development concept proposed in 2018, Daqing’s HQAD showed insufficient momentum, with a noticeable gap compared to other cities in the Songnen Plain. Although the local government implemented policies such as the “Implementation Opinions on Accelerating Agricultural High-Quality Development” to promote HQAD, the effects remain unstable.

The overall disparity within the Sanjiang Plain exhibits a narrowing trend, with an average annual decrease rate of 21.293%. This trend is driven by the increasing influence of the Sanjiang Plain’s modern agricultural reform pilot zone. It has enabled lower-level cities to catch up with higher-level ones, thus reducing the HQAD gap in the region.

The development disparities among cities within the Xing’an Mountains Regions have gradually narrowed. A plausible explanation is that Heihe, with its early focus on mineral resource and remote geographical location, began pursuing HQAD later than Yichun. Nevertheless, in recent years, Heihe has adhered to the supply-side structural reform in agriculture. Notably, the establishment of the agricultural development demonstration zone in 2018 significantly accelerated agricultural growth, gradually narrowing the gap with Yichun.

Fig. 4
figure 4

Intra-regional disparities in HQAD levels of Heilongjiang Province from 2013 to 2022.

Full size image

Inter-regional disparities in HQAD levels of Heilongjiang Province

As shown in Fig. 5, regional disparities from 2013 to 2022 between Songnen Plain and Mudanjiang Semi-mountainous Areas, Songnen Plain and Xing’an Mountains Regions, Songnen Plain and Sanjiang Plain, Mudanjiang Semi-mountainous Areas and Xing’an Mountains Regions, Mudanjiang Semi-mountainous Areas and Sanjiang Plain, as well as Xing’an Mountains Regions and Sanjiang Plain, generally exhibited a narrowing trend despite brief fluctuations, with annual average decline rates of 14.757%, 14.170%, 11.483%, 9.689%, 14.254%, and 18.719% respectively. This indicates that the Sanjiang Plain and Xing’an Mountains Regions, which have a higher proportion of non-agricultural cities, have seen significant success in agricultural reforms in recent years. These reforms have accelerated HQAD, narrowing the development gap with Songnen Plain and Mudanjiang Plain, where agricultural cities are concentrated. However, fluctuations in inter-regional disparities reveal that agricultural development in each area still lacks consistency and stability in its progression.

In particular, the Gini coefficient between the Songnen Plain and the Semi-mountainous Areas of Mudanjiang consistently remains higher than that between other regions. Songnen Plain and Semi-mountainous Areas of Mudanjiang naturally possess conducive conditions for agricultural development. However, the former primarily focuses on food production, while the latter emphasizes the development of specialty agricultural products or organic agriculture with higher added value and market competitiveness. This has led to disparities between the two regions in areas such as improving the quality and efficiency of agricultural supply, promoting green agriculture, coordinating agricultural structures, and sharing benefits of development. While disparities in agricultural endowments can be mitigated through policy preferences, structural adjustments, and technological innovations, the gap in industrial structures is more challenging to bridge. Consequently, the Gini coefficient between the Songnen Plain and the Semi-mountainous Areas of Mudanjiang is significantly higher than that between other regions.

Fig. 5
figure 5

Inter-regional disparities in HQAD Levels of Heilongjiang Province from 2013 to 2022. (a) The disparity between the Songnen Plain and the other three regions, (b) The disparity between the Semi-mountainous Areas of Mudanjiang and the other three regions, (c) The disparity between the Sanjiang Plain and the other three regions plain, (d) The disparity between the Xing’an Mountains Regions and the other three regions.

Full size image

Regional disparity sources and their contribution rates to HQAD in Heilongjiang

As shown in Fig. 6, from a development trend perspective, the contribution rate of inter-regional disparities to HQAD levels in Heilongjiang Province has decreased from 80.769% in 2013 to 37.481% in 2022. In contrast, the contribution rate of intra-regional disparities has increased, rising from 15.717% in 2013 to 30.065% in 2022. Inter-regional disparities still contribute the most, followed by intra-regional disparities. This is primarily due to the significant differences in natural conditions, economic development levels, and industrial priorities among the four major regions of Heilongjiang: Songnen Plain, Sanjiang Plain, the Semi-mountainous Areas of Mudanjiang, and Xing’an Mountains Regions. These differences result in substantial variations in green agriculture, quality and efficiency improvement of supply, agricultural structural coordination, and the shared benefits of development. However, within the same region, the differences in natural and economic conditions are relatively smaller, contributing less to the overall disparities. Hyperbolic density, reflecting the contribution of overlapping portions among sub-samples to the overall disparity, accounts for a relatively low share in this study, with an average rate of 8.460%. This suggests that the categorization into four major regions effectively distinguishes cities of different types, demonstrating strong rationality.

Fig. 6
figure 6

Regional disparity sources and their contribution rates to HQAD in Heilongjiang Province from 2013 to 2022.

Full size image

Evolution analysis in HQAD levels of Heilongjiang Province

Markov transition matrix probability in HQAD levels of Heilongjiang Province

Given the disparities in the levels of HQAD among regions, it is crucial to answer question of whether HQAD within Heilongjiang Province will decline or improve in the future, the probabilities of different levels evolving into other levels, and the characteristics of the steady-state distribution. Therefore, we employ the traditional Markov chain to examine the internal dynamics and steady-state distribution in HQAD of Heilongjiang Province.

As shown in Table 4, the second row of data indicates that 80.000% of the prefecture-level cities with low-level HQAD (Level I) has progressed to medium-low levels (Level II) by the end of the year. The third row indicates that 80.000% of the prefecture-level cities with medium-low levels maintained their status (Level II) at the end of the year, with 9.231% declining to low levels (Level I), 6.154% advancing to medium-high levels (Level III), and 4.615% reaching high levels (Level IV). The fourth row of data shows that 77.778% of the prefecture-level cities remained at medium-high levels, with 18.519% experiencing a one-level downgrade, 3.703% undergoing a two-level downgrade. Finally, the fifth row indicates that 76.471% of the prefecture-level cities maintained their high levels (Level IV), whereas 11.765% declined by one-level to medium-high levels (Level III), and 5.882% experienced a two-level downgrade to medium-low levels (Level II).

In sum, the distribution and evolution in HQAD of Heilongjiang Province from 2013 to 2022 exhibit the following characteristics: When agricultural development reaches medium-low, medium-high, or high levels, prefecture-level cities generally remain stable, with a low probability of significant cross-tier changes. Most variations occur between adjacent levels.

Table 4 Markov chain transition matrix probability distribution for Heilongjiang’s HQAD levels.
Full size table

Initial and steady state distributions in HQAD of Heilongjiang Province

As shown in Table 5, compared to the initial distribution, the steady-state distribution shows a decrease of 33.117% in the low-level category of HQAD, an increase of 41.326% in the medium-low level, a rise of 5.415% in the medium-high level, and a decline of 13.625% in the high level. This indicates variability in the stability of different levels in HQAD of Heilongjiang Province. Furthermore, the medium-low level accounts for the highest proportion, at 57.993%, while the medium-high level stands at 22.082%, together comprising 80.075% of the total. The high level follows, accounting for 11.375%. These findings suggest that Heilongjiang’s HQAD will converge to a long-term equilibrium state, mainly concentrated at the medium-low and medium-high levels. Moreover, although some prefecture-level cities have previously achieved a high level of high-quality agricultural development, there is a risk of a decline.

Table 5 Initial and steady-state distributions in HQAD of Heilongjiang Province.
Full size table

Discussion and conclusion

Conclusion

Improving agricultural development quality is a key strategy for ensuring national food security and promoting economic growth. We focus on the prefecture-level cities in Heilongjiang Province, constructing a multidimensional evaluation framework to assess the HQAD from four dimensions: green agriculture, quality and efficiency improvement of supply, agricultural structural coordination, and shared benefits of development. Using the Entropy Weight- TOPSIS model, we quantify the HQAD levels in Heilongjiang Province and its prefecture-level cities from 2013 to 2022. Furthermore, by integrating the Dagum Gini Coefficient and its decomposition, alongside the Markov Chain, we explore regional disparities, their sources, and the evolution patterns of Heilongjiang’s HQAD. This research, with its more representative subject, comprehensive analytical framework, and detailed analysis, significantly contributes to the existing literature4,5,7,14.

The conclusions drawn are as follows: First, Heilongjiang Province has experienced steady improvement in overall HQAD from 2013 to 2022, with an average level of 0.323. Despite this, structural challenges persist, and regional disparities remain significant. Within the framework’s dimensions, green agriculture serves as the cornerstone. Before 2017, the primary drivers were quality and efficiency improvement of supply and shared benefits of development. From 2017 to 2022, agricultural structural coordination became the primary factor, though its momentum has been insufficient. At the prefecture level, all cities have shown an upward trend in HQAD, advancing in unison. By 2022, all prefecture-level cities had almost reached at least a medium-low level of HQAD. In terms of spatial distribution, cities in the Songnen Plain, including Harbin, Qiqihar, and Mudanjiang, lead with high or medium-high levels of HQAD. In contrast, most cities in the Sanjiang Plain generally achieve only medium-low or low development levels.

Second, the Dagum Gini Coefficient and its decomposition reveal spatial imbalances in Heilongjiang Province’s HQAD. The overall regional disparity exhibits a diminishing trend. From 2013 to 2022, except for the Songnen Plain, the intra-regional disparities within the Sanjiang Plain and the Xing’an Mountains Regions have narrowed, with their contributions to the overall disparity have continuously increased. In 2013, the Sanjiang Plain showed the largest intra-regional disparity, whereas by 2022, the Songnen Plain took the lead. Inter-regional disparities have also decreased, with the largest gap between the Songnen Plain and the Semi-mountainous Areas of Mudanjiang. Furthermore, inter-regional disparities are the primary source of Heilongjiang Province’s overall disparity in HQAD, although their contribution to the overall disparity has shown a declining trend.

Third, the Markov chain indicates that in the short term, the distribution pattern of Heilongjiang Province’s HQAD remains relatively stable, with most transitions occurring between adjacent stages. When the level of HQAD is categorized as low, medium-low, medium-high, or high, the likelihood of substantial cross-stage transitions in development levels among various prefecture-level cities is relatively low. In the long run, however, Heilongjiang’s prefecture-level cities will trend towards clustering at medium-low and medium-high levels.

Policy implications

The above research conclusions offer valuable policy insights that can assist Heilongjiang and other regions or countries with similar conditions in achieving the goal of HQAD.

First, strengthening agricultural structural coordination is crucial to advancing Heilongjiang’s HQAD. Empirical evidence suggests that from 2017 to 2022, agricultural structural coordination was the primary driver of Heilongjiang’s HQAD, but the momentum has been insufficient. This study suggests that the lack of momentum may be due to the impact of the COVID-19 pandemic. Consequently, the Heilongjiang Provincial Government could increase financial support for agriculture, optimize the allocation of agricultural support funds, prioritize agricultural technology R&D, and enhance crop resilience against to disasters. Finally, while ensuring food security, the agricultural planting structure could be adjusted flexibly to increase the proportion of high-value-added economic crops and specialty agricultural products. Strengthening the integration of agriculture with the secondary and tertiary industries, particularly promoting the development of agricultural product processing and service industries, can form industrial clusters, improving the stability and risk resistance of the agricultural value chain.

Second, promoting coordinated HQAD both within and across regions, tailored to local conditions, is essential. In the Songnen Plain, intra-regional disparities are gradually widening, mainly due to the lack of coordination in Daqing’s agricultural structure. The Daqing government should optimize its agricultural industry, for example, by utilizing oil by-products in agricultural production or developing agricultural tourism based on the oil industry. Nevertheless, the Songnen Plain’s average HQAD level has remained among the highest in Heilongjiang Province over the past decade. Therefore, the provincial government should establish a comprehensive agricultural cooperation mechanism to promote the transfer of advanced technologies and management practices from the Songnen Plain to other regions, maximizing regional synergies and improving overall agricultural development. In the Sanjiang Plain, although the intra-regional disparities have narrowed, the HQAD levels of related cities remain low, due to the scarcity of agricultural resources. These cities should make full use of their unique resources and develop industries tailored to local conditions. For instance, Hegang can use its abundant coal resources to drive HQAD through related industries. Qitaihe, with low supply quality, should focus on expanding its primary industries: grain, livestock products, and processing of green specialty foods, as well as developing moderately scaled operations to improve production efficiency. In the Xing’an Mountains Regions, the intra-regional disparities are gradually narrowing, but Yichun’s momentum remains insufficient. Yichun can learn from Heihe’s development experience, given their similar resource endowments, by promoting green agriculture to enhance HQAD. In the Semi-mountainous Areas of Mudanjiang, despite ranking among the top in Heilongjiang Province in terms of average HQAD over the past decade, there was a sharp decline in 2018. This may have been due to the adaptation period during the agricultural structural adjustment. Since then, the HQAD level has consistently improved. Therefore, this area should continue to utilize its diverse land resources to develop organic agriculture and specialty products to achieve the long-term HQAD goals.

Third, achieving institutional breakthroughs is crucial to advancing prefecture-level cities towards higher levels of HQAD. Research findings indicate that in the long run, Heilongjiang’s prefecture-level cities will cluster towards medium-low and medium-high levels of HQAD. This highlights the necessity for Heilongjiang Province to further elevate its overall HQAD. The provincial government should consider institutional breakthroughs, such as deepening land system reforms, advancing pilot reforms in rural reform zones, fostering new urbanization, accelerating the adoption of advanced agricultural technologies, and optimizing agricultural structures. These measures will be instrumental in achieving a paradigm shift in the levels of HQAD.