Spatial-Temporal Changes and Influencing Factors of Ecological Protection Levels in the Middle and Lower Reaches of the Yellow River

Abstract

In recent years, ecological problems in the middle and lower reaches of the Yellow River have been frequent. Therefore, exploring its core influences can advance the implementation of “Ecological Protection and High-quality Development of the Yellow River Basin”. This paper constructs an indicator system based on PSR guidelines, evaluates the ecological protection level of 55 cities in the middle and lower reaches of the Yellow River from 2009 to 2019, and uses correlation analysis with geographically and temporally weighted regression (GTWR) model to explore the spatial distribution characteristics of influencing factors, such as intensity of fertilizer application, amount of agricultural film applied, afforestation area per capita, and green technology innovation level on the ecological protection level. It is found that the overall level of ecological protection has shown a steady increase, but the spatial distribution varies widely. The ecological level increased from 0.2218 to 0.3357, showing a decreasing distribution trend from coastal to inland. Furthermore, it is found that the ecological protection level has a significant positive spatial correlation, mainly for similar clustering. The Global Moran’s I for ecological protection level is greater than 0, and the Moran scatter plot has a high number of cities distributed in the first and third quadrants. There is a heterogeneity in the spatial and temporal distribution of factors influencing the level of ecological protection. Fertilizer application, the agricultural film uses, and afforestation area per capita are mainly negatively affected, while green innovation level has a strong positive effect, and agricultural film use, afforestation area per capita, and green innovation level become the core influencing factor of different regions. Therefore, in the middle and lower reaches of the Yellow River, the ecological protection level should be improved by implementing a regional differentiated development strategy, realizing cross-regional linkages between cities and focusing on differences in core driving factors.

1. Introduction

In 2019, the status of the Yellow River Basin as an important ecological barrier in China was established by the introduction of the strategy “Ecological Protection and High-Quality Development of the Yellow River Basin” [1], indicating that the protection of the Yellow River basin is conducive to the improvement of its ecological environment and the ecological security of China. Then, relevant documents were issued one after another with the aim of implementing the national strategy as soon as possible, emphasizing that protecting the region’s ecological environment is a practical need to prevent and resolve ecological security risks and build China into a beautiful nation [2]. Environmental protection is to protect productivity [3,4,5], especially in the middle and lower reaches where the human–land relationship is complex [6]. Human activities combined with natural disasters have caused severe damage to the ecological environment, such as water pollution [7,8,9], desertification, soil erosion, and extinction of species [10]. Therefore, the ecological environment is highly susceptible to degradation, and ecological fragility has become a prominent problem, and hence improving the ecological environment condition is an urgent requirement. Therefore, studying spatial and temporal evolution of ecological protection level and influencing factors in the middle and lower reaches of the Yellow River is a great practical significance to promote the “Ecological Protection and High-quality Development in the Yellow River basin” strategy as soon as possible.

The foundation of human survival and development is the ecological environment [11], and its condition is related to the prosperity and decline of a country’s civilization and the region’s sustainable development. However, in the past, the country emphasized development rather than protection and developed the economy at the cost of excessive consumption of resources and damage to the environment [12], such as the direct discharge of waste gas and wastewater from heavy chemical industries into the environment [13], which led to serious water pollution and air pollution, and the ecological environment was seriously damaged. With the promotion and implementation of the ecological civilization policy formulated by the state, ecological protection has begun to receive attention. Ecological protection is an ongoing act to control pollution and repair the ecological environment [14], which is reflected in various aspects such as water conservation, harmless waste treatment, and afforestation [15]. The First National Conference on Environmental Protection, held in 1973, was the beginning of environmental protection in China, and the “Certain Provisions on the Protection and Improvement of the Environment (Trial Draft)” adopted by the conference was the prototype of China’s environmental protection law [16]. Xi Jinping has repeatedly stressed that “we should adhere to the priority of protection” [17], pointing out the relationship between development and conservation and indicating the country’s importance attaches to ecological civilization. Therefore, a large number of scholars have established an ecological protection level indicator system based on the PSR Model [18] and used the entropy method, principal component analysis method, and entropy-weighted TOPSIS [19] to measure the ecological protection level. In 2019, the “Ecological Protection and High-quality Development of the Yellow River Basin” was proposed, on this basis, a large number of scholars began to focus on the ecological and economic development on this region, there have been many studies, such as the eco-efficiency, ecological risks, ecology and high-quality development [20], urbanization and ecological protection [21].

Domestic and international studies have mainly focused on low-carbon sustainable development and the relationship between ecology and economy, lacking analysis of the influencing factors of ecological protection and empirical tests. Therefore, to measure the level of ecological protection, this paper first constructs an ecological protection system and then uses the GTWR model to analyze the degree of influence of intensity of fertilizer application, amount of agricultural film applied, afforestation area per capita and the green technology innovation level to the ecological protection level, and Arcgis was used to visualize the spatio-temporal evolution characteristics. It is expected to provide practical reference for promoting the implementation of strategy “Ecological Protection and High-quality Development in the Yellow River basin”.

2. Martials and Methods

2.1. Indicator System

This paper aims to reveal the evolutionary trends of ecological protection level; therefore, the first step is to build an ecological protection system that reflects the ecological condition. Combining the results of previous research and referring to the studies of Cui Panpan [22] and Sun Jiqiong [23], based on the Pressure-State-Response Model, the ecological protection system is constructed from the state level, the pressure level, and the response level (

Table 1. Indicator system of ecological protection level.

First-Level Index	Second-Level Index	Unit	Attribute
state level	Greenery coverage in built-up areas	%	+
	Green space per capita	m2/person	+
pressure level	Industrial wastewater emissions per unit GDP	Tonnes/million RMB	-
	Industrial SO2 emissions per unit GDP	Tonnes/million RMB	-
response level	Ratio of sound disposal of domestic waste	%	+
	Ratio of sewage sent to treatment plants	%	+

).

2.2. Analytical Framework

2.2.1. Ecological Conservation Levels Model

Differences in units and scales between different indicators may have an impact on the evaluation results, so the indicators are standardized by:

$$Y_{\lambda ij}^{*}=\left(x_{\lambda ij}-x_{min}\right)/(x_{max}-x_{min})\hspace{1em}(\mathrm{positive} \mathrm{indicators})$$

(1)

$$Y_{\lambda ij}^{*}=\left(x_{max}-x_{\lambda ij}\right)/(x_{max}-x_{min})\hspace{1em}(\mathrm{negative} \mathrm{indicators})$$

(2)

This paper draws on the practice of Feng Xinghua and others [24] to use the entropy weighting method to decide the index weight, which can reduce the interference of subjective factors, where: h, m, n are the number of years, cities, indicators.

To eliminate the impact of “0” on subsequent calculations after the standardization of the indicators:

$$Y_{\lambda ij}=Y_{\lambda ij}^{*}+{10}^{-4}$$

(3)

Proportion of the indicators:

$$P_{\lambda ij}=Y_{\lambda ij}/{\displaystyle \sum }_{\lambda =1}^h{\displaystyle \sum }_{i=1}^m{Y}_{\lambda ij}$$

(4)

Entropy value of each indicator:

$$E_j=-k{\displaystyle \sum }_{\lambda =1}^h{\displaystyle \sum }_{i=1}^m{P}_{\lambda ij}ln{P}_{\lambda ij}$$

(5)

where $k=1/ln\left(m\times h\right)$.

Redundancy of information entropy:

$$D_j=1-E_j$$

(6)

Weighting of each indicator:

$$W_j=D_j/{\displaystyle \sum }_{j=1}^n{D}_j$$

(7)

Ecological protection level:

$$U_1={\displaystyle \sum }_{j=1}^n{W}_j{Y}_{\lambda ij}$$

(8)

2.2.2. GTWR Model

This paper draws on a study by Li Enkang [25] using GTWR model. GTWR can effectively deal with spatio-temporal non-smoothness by introducing a temporal dimension based on the examination of spatial heterogeneity. The main equations are as follows:

$$Y_i={\alpha }_0\left(x_i,y_i,t_i\right)+{\displaystyle \sum }_r{\alpha }_r\left(x_i,y_i,t_i\right)X_{ir}+{\epsilon }_i$$

(9)

$$d_{ij}^{ST}=\sqrt{\rho \left[{\left(x_i-x_j\right)}^2+{\left(y_i-y_j\right)}^2\right]+\mu {\left(t_i-t_j\right)}^2}$$

(10)

$$W_{ij}^{ST}=exp\left\{-\left(\rho \left[{\left(x_i-x_j\right)}^2+{\left(y_i-y_j\right)}^2\right]+\mu {\left(t_i-t_j\right)}^2/b_{ST}^2\right)\right\}$$

(11)

where $Y_i$ is the explanatory variable; $\left(x_i, y_i, t_i\right)$ is the geographical location and time at which region $i$ is located; ${\alpha }_0\left(x_i, y_i, t_i\right), {\alpha }_r\left(x_i, y_i, t_i\right), X_{ir}$ is the regression constant, regression coefficient and value of independent variable; $i$ and $j$ are different regions; the parameters $\rho$ and $\mu$ are scaling factors measuring spatial and temporal distances; and $b_{ST}$ is the bandwidth of the spatio-temporal weight function. Error term ($\epsilon$) is assumed to be normally distributed at zero mean value and constant variance [26,27].

2.3. Data Sources

In the middle and lower reaches of the Yellow River, there are five provinces (autonomous regions), Shandong, Henan, Inner Mongolia, Shanxi, and Shaanxi respectively, but the city of Laiwu was incorporated into Jinan in 2019 and data for most cities in Inner Mongolia Autonomous Region are seriously missing. Therefore, based on the principles of data availability and feasibility, only 55 cities were selected. The data were collected from the Statistical Yearbook of five provinces (autonomous regions), the China City Statistical Yearbook, the Statistical Bulletin or Information from Statistical Bureaus of 55 cities. For the processing of some of the missing data, linear interpolation was carried out using data from neighboring years

3. Results and Discussion

The above method was used to measure the ecological protection level of 55 cities respectively during the period 2009–2019, and the overall mean values were calculated by average method. In addition, to overcome the errors caused by data fluctuations, a three-year averaging method was used to divide the whole study period into four periods.

3.1. Ecological Protection Level

3.1.1. Chronological Changes

In the middle and lower reaches of the Yellow River, the ecological protection level is lower, all less than 0.5, but shows a slow growth trend at first and then a fast-growing trend. The ecological protection level grew slowly from 2009–2011, the growth rate was only 0.86%, accelerated from 0.256 to 0.320 for 2012–2017, and slowed down from 2018–2019 (Figure 1). Probably because the industrial structure is dominated by primary and secondary industries [28] and has a large number of energy and heavy chemical bases, the long-term rough economy development pattern has led to serious environmental pollution [17,29,30,31]. Therefore, the ecological protection level grew slowly from 2009 to 2011, only from 0.222 to 0.241. In 2012, China gave prominence to the construction of ecological civilization, leading to the comprehensive promotion of the concept and practice of ecological civilization. Government departments have increased their supervision of the environment and implemented stricter environmental regulation policies, which have led to enterprises actively fulfilling their social responsibilities, upgrading waste treatment equipment and reducing pollutant emissions. The air quality has continued to improve, forest ecology has been restored, and water resources have been used more efficiently, resulting in a significant increase in the ecological protection level from 2012 to 2017; the growth rate was 25%. Moreover, many provinces promoted the delineation of the ecological protection red line, and the areas included were prohibited from engaging in industrial and urbanized development, effectively protecting the ecosystem and allowing the ecological environment to be restored; the ecological environment has reached a basic stable state, and the level of ecological protection showed a slow increase in 2018–2019, only from 0.327 to 0.336 (Figure 1).

3.1.2. Spatial Distribution Characteristics

Following Zhang [32], we used ArcGIS10.8 to generate a spatial change map of ecological protection levels in the middle and lower reaches of the Yellow River, then use the natural break point method to classify the ecological protection levels into four levels: higher level, high level, low level, and lower level (Figure 2). The ArcGIS was developed by Environmental Systems Research Institute (the company is headquartered in Redlands, California, USA), and first released in 1999. According to the classification of ecological protection levels, the spatial distribution of ecological protection levels has the following characteristics.

(1)

From coastal to inland areas, the ecological level tends to decrease from high to low. The most cities with high ecological protection level are mainly located in coastal areas, fewer in inland areas. Coastal cities with the region’s good economic advantages, such as Qingdao and Weihai, have launched green financial products and related platforms to support enterprises to engage in environmentally friendly production, prompting them to transform and upgrade, which has led to an improved level of ecological protection. At the same time, inland areas are relatively backward economically, green financial products have not yet been promoted and ecological policies are weakly regulated. Therefore, most high-level cities are distributed in downstream coastal areas, and there is a certain spatial coupling with regions with higher economic development.

(2)

The level of ecological protection varies greatly between provinces. Shandong and Henan are located in the North China plain, with flat terrain, fertile soil, many rivers and lakes [33], it is suitable for various organisms to survive. Therefore, they have a better ecological environment foundation. The industrial structure is transformed more quickly, with less heavy and chemical industries. A series of ecological compensation policies are issued to improve the ecological protection level, so they are in areas with a higher ecological protection level in the four periods and have more cities with a high ecological protection level. The Loess Plateau contains 40% of the Shaanxi Province’s total area and almost the entire Shanxi Province. It has a loose soil structure, which lacks vegetation protection, and is susceptible to serious soil erosion during the summer months when precipitation is concentrated, especially during heavy rainfall [34]. Shaanxi has a large amount of coal, oil and gas resources, and the heavy chemical industries are in dominant position. The large-scale exploitation of resources has led to a large reduction of surface vegetation, coal ash and other wastes have also caused damage to the surface environment and exacerbated soil erosion [35], resulting in serious damage to the ecological environment. Hence, the areas mainly have low ecological protection levels in four periods.

(3)

The number of cities with high ecological protection levels has increased slightly. The study areas have a more prominent conflict between protection and development, so to improve its ecological situation, a series of measures has been enacted by the state, from the National Comprehensive Water Resources Plan (2020–2030) approved in 2010, which included the Yellow River Basin Water Resources Utilisation Plan as an important part, to “Ecological Protection and High-quality Development of the Yellow River Basin” strategy proposed in 2019, the implementation of relevant policies has led to the improvement of the ecological situation, resulting the number of cities with high ecological protection levels rising from 21.8% in the first period to 23.64% in the fourth period.

3.2. Spatial Correlation Analysis of Ecological Protection Level

3.2.1. Global Spatial Autocorrelation Analysis

As shown in

Table 2. Ecological protection level’s Global Moran’s I index and test.

Years	2009	2010	2011	2012	2013	2014	2015	2016	2017	2018	2019
Global Moran’s I	0.205	0.233	0.225	0.233	0.217	0.199	0.2	0.188	0.207	0.151	0.213
z	2.674	2.999	2.897	2.989	2.815	2.59	2.594	2.468	2.689	2.054	2.756
p	0.004	0.001	0.002	0.001	0.002	0.005	0.005	0.007	0.004	0.02	0.003

, which shows the values and tests of Global Moran’s I for ecological protection level, all Global Moran’s I during the study period were greater than 0.1, all p-values were less than 5%, and z-values were more than 1.96, all of which passed the significance test. This indicates that the environmental protection level has a high positive spatial correlation and shows an apparent spatial clustering of similarities.

The Global Moran’s I index for ecological protection level has shown a fluctuating trend from 2009 to 2019. In 2010 and 2012, the Global Moran’s I index was the largest, indicating that spatially concentrated distribution of cities with similar ecological protection levels reached its peak in these two years; after 2013 the Global Moran’s I showed an overall fluctuating trend, indicating a change in the concentration of areas with similar levels of ecological protection.

3.2.2. Local Spatial Autocorrelation Analysis

All global Moran’s I index are greater than 0.1, which only reflect the existence of agglomeration or outliers in the study area, but cannot reflect the characteristics of local spatial agglomeration. Local Moran’s I can compensate for this deficiency and show the local spatial agglomeration through Moran scatter plots.

From the Moran scatter plot (Figure 3), the local spatial autocorrelation index of the ecological protection levels of cities in 2009 and 2019 are mainly located in the first and third quadrants, which belong to the “high-high” and “low-low” agglomeration types, showing significant positive local spatial correlation, and more cities are in the “low-low” type. The Moran scatter plot in 2009 and 2019 shows little change, indicating that the distribution of spatial agglomeration of ecological protection levels is relatively stable, with cities mainly located in the first and third quadrants, indicating that most of them are of the same type of agglomeration.

In 2009 and 2019, the majority of cities belonged to the “low-low” agglomeration type, probably because of the poor ecological condition and the emission of industrial pollutants, which have left many cities on the verge of redlining their air quality, leaving fewer cities with high ecological protection levels. The lack of cross-regional synergistic management policies by the basin government makes it difficult for cities with high ecological protection levels to play as a bellwether and help neighboring cities with low levels improve their ecological protection levels.

In 2009, all cities of “high-high” agglomeration type were located in Shandong Province, and in 2019 the number of cities in this type increased slightly, with several cities in Shanxi, such as Jinzhong, changing from the “high-low” type to this type. This is mainly because Shandong Province has a better natural environment and has issued many documents on the construction of ecological civilization, so the “high-high” agglomeration type is primarily located in Shandong Province.

4. Factors Influencing the Level of Ecological Protection

4.1. Variable Selection

Based on previous research and data availability, four data items, fertilizer application intensity, agricultural film application, level of green technology innovation and per capita afforestation area, were selected as influencing factors for the level of ecological protection (

Table 3. Influencing factors.

Indicators	Specific Indicators	Explanation of Indicators
Intensity of fertilizer application	Fertilizer use/crop area sown	Agricultural pollution and ecological protection
Amount of agricultural film applied	Agricultural plastic film use/crop sown area
Green technology innovation level	Number of green patent applications/grants	Green technology applied to ecological conservation
Afforestation area per capita	Area of planted forests/population	Artificially modified ecosystems

).

(1)

The intensity of fertilizer application. Agricultural surface pollution has become one of the most important sources of water pollution in China, so agricultural pollution can greatly influence ecological protection. To maintain crop yields, farmers use chemical fertilizers to increase soil fertility, but there are no precise standards for the amount of fertilizer applied, so farmers are prone to over-spreading fertilizers, resulting in chemicals substances that are not absorbed by plants entering the soil first and then rivers and groundwater with the seepage of water, causing eutrophication of river and the appearance of various algae, polluting water sources, and affecting the ecosystem.

(2)

Amount of agricultural film applied. Some crops need to be winterized and the use of agrarian film provides higher temperatures and humidity, so some winterized crops in the north require large amounts of film, but agricultural film cannot be reused and its chemical material produces harmful substances when burned, so it is thrown away in large quantities, leading to a lot of white pollution in the countryside, which is not easily decomposed and can be harmful to the ecological environment.

(3)

Green technology innovation level. Green technology innovation generally refers to new methods or technologies that can improve the environment [36,37], which have been studied and discussed by scholars in many aspects, and are believed to be able to mitigate environmental risks [38], improve resource use efficiency [39], reduce pollution rates [40], save energy [41,42], and bring ecological honor [43], so governments and relevant organizations have taken green technological innovation as an important driving force to achieve environmental protection [16].

(4)

Afforestation area per capita. Forestry ecological security is an important part of national ecological security. As a component of terrestrial systems, forests and crops can play a number of roles such as climate regulation and biodiversity enrichment [44], contributing to regional sustainable development and ecological civilization construction [45]. Planted forests can replenish the forestry area reduced by logging operations or related production, maintain or enhance the capacity of forest systems to function, and improve the ecological environment.

4.2. Data Verification

After standardizing the variables, regression analysis was used to check for multicollinearity. To avoid pseudo-regressions in the regression, variables’ variance inflation factors (VIF) need to be excluded once they are greater than 10. The level of ecological protection was chosen as the dependent variable, and the intensity of fertilizer application, the amount of agricultural film applied, the level of green technology innovation and the afforestation per capita were selected as independent variables for empirical testing.

Table 4. Relevant parameters of the GTWR.

Bandwidth	Residual Squares	Sigma	AICc	R2	R2 Adjusted	Spatio-Temporal Distance Ratio
1.1435	1.1314	0.0429	−2130.5430	0.8005	0.7992	0.9365

shows the relevant parameters of GTWR regression results. From the goodness of fit, the corrected R² is 0.7992, indicating that this GTWR regression model can better account for the impact of influencing factors on the ecological protection level.

4.3. Evolution of the Spatial Distribution of Influencing Factors

There is spatial variation in the extent to which each driver influences the level of ecological protection over time. Figure 4 illustrates the spatial distribution of regression coefficients for each influence factor in the GTWR model.

(1)

The impact of fertilizer application intensity on the level of ecological protection is mostly negative, and the areas with a higher level of influence are more concentrated in distribution. The intensity of fertilizer application only had the most significant impact on the level of ecological protection in Hohhot in the first period, the number of cities in the highest-level increased greatly in the second and third periods. In Shandong, northern Shanxi and Yulin in Shaanxi Province, the intensity of fertilizer application showed a negative impact on the level of ecological protection, while in Henan, south-central Shaanxi, and southern Shanxi it showed a positive impact, probably because the reduction of fertilizer advocated by the government has played a role, the small amount of fertilizer applied is absorbed by plants, promoting plant growth and thus allowing the level of ecological protection to be improved, thus showing a positive impact.

(2)

The negative impact of agricultural film application on the level of ecological protection shows an upward trend, with a smaller number of areas with a positive impact. From the first and third periods, the areas where the impact of agricultural film application is negative are mainly located in southern Shaanxi, central Henan, and coastal Shandong. In the second period, Henan turns to be in a positive distribution, but the positive value is smaller. This indicates that a small amount of agricultural film can protect some non-cold-tolerant plants successfully through winter, which can help maintain the regional ecology. However, the high application amount of agricultural film in most areas damaged the local soil and water environment, so overall showed a negative impact.

(3)

The green technology innovation level positively impacts the ecological protection level, but the degree of effects shows a gradual downward trend. The number of cities in high-level influenced by the level of green technology innovation is decreasing. Most cities in Henan Province have reduced from a higher level of influence in the previous two periods to a low level of influence. The green technology innovation level has a high and stable impact on the ecological protection level in the cities of Shanxi Province, with a gradual decline from a higher level in Henan Province, and mainly at a lower level of impact in Shandong and Shaanxi. Innovation is the first driver of green development, and technological innovation plays an increasingly important role in improving environmental quality [46]. The green technologies are difficult to disseminate, complicated to use, and costly, resulting in a low conversion rate and a long period between their approval and practical application to ecological conservation, making the green technology innovation level in this period less effective in enhancing ecological conservation.

(4)

The impact of per capita afforestation on the ecological protection level is primarily negative, and shows the first decreasing and then increasing trend. This is mainly because artificial afforestation requires seedling breeding, which consumes many human and material resources, and the survival rate of transplanted seedlings is improved through fertilization and other technical means, raising local carbon emissions. Moreover, the effect of seedlings on improving the ecological level is limited, and the positive impact of afforestation on the environment needs to wait for the saplings to grow up, so there is a lag in the effect of artificial afforestation on improving ecological protection. The areas where the impact of afforestation per capita is positive are mainly distributed in the south of Shaanxi and Shanxi and the north of Henan, where the Yellow River and its branches especially flow through, and the soil erosion is more serious. To combat soil erosion, the governments of Shaanxi and other regions have attached great importance to tree planting activities, promoting a comprehensive national compulsory tree planting campaign through various forms such as pledging to plant and adopt trees, donating funds for afforestation, network planting, nurturing and management, etc. Moreover, the afforestation area is more extensive, with an increase of 66,700 square kilometers of forest area in Shaanxi alone, and the trees transplanted in earlier years have started to take effect, significantly improving the poor ecological state of the region.

4.4. Distribution of Urban Core Influences

The influence of different drivers on the level of ecological protection varies from period to period in the middle and lower reaches of the Yellow River. So, the influencing factor with the most significant degree of influence in each city at three periods was selected as the core influencing factor. Specific information can be obtained from Figure 5.

(1)

Core influences in the same area are relatively stable. The cities with the application amount of agricultural film as a core influencing factor are fewer, which may be due to the use of agricultural film in more occasional areas and the duration of film use, so the impact on the level of ecological protection is small. However, the use of biodegradable film and the limitations of film recycling have led to the accumulation of film residues in the environment year after year, which affects water infiltration and reduces soil water content, leading to serious consequences such as the salinization of the soil. As a result, the impact of agricultural films on the level of ecological protection increases. The number of cities with a green technology innovation level as core impact factors is more. Still, it shows a decreasing trend, mainly due to the high costs and barriers to green technology innovation’s research and development, which makes enterprises lack the motivation to research and development, leading to a gradually decreasing impact of green technology innovation. Hence, the number of cities with it as a core influence factor decreases more. The number of cities with afforestation area per capita as a core influencing factor is small, but shows an increasing trend, probably because transplanted saplings need time to grow into woods before they can play a role in windbreak and sand-fixation, reduce the carbon dioxide content of the area, which in turn has an impact on the greenhouse effect, causing afforestation area per capita impact on ecological protection level is lagging.

(2)

The core influencing factors are different between regions. The core influencing factor for Shaanxi is mainly the amount of agricultural film applied, for Henan and Shanxi is the green technology innovation level, for Shandong is green technology innovation level and afforestation area per capita. In Shaanxi, the low importance attached to green technologies and less green technology applications make its positive impact on ecological protection limited, but the local area has a wide range of agricultural product bases with a large amount of agricultural film application, therefore, the amount of agricultural film applied becoming the core factor influencing the level of ecological protection. As more and more attention being paid to the Yellow River’s ecological environment, Shaanxi promotes sustainable agricultural development projects. It implements projects such as the comprehensive recycling of agricultural film, so the number of cities in Shaanxi with agricultural film application as a core factor for ecological protection level has been reduced in the third period. Henan, in order to strengthen energy conservation and emission reduction, has carried out green technology-related seminars and built a green technology innovation system. At the same time, Shanxi focused on green transformation, using green technology to promote low carbon emissions and develop clean energy, making the level of green technology innovation a core influencing factor for most cities in Henan and Shanxi provinces; Shandong attaches more importance to green technology and forest resources, and since 2004 Shandong has held several Since 2004, Shandong has held many green technology exhibitions and environmental technology forums to promote the exchange of green technology among municipalities, and has established a forest chief system and actively promoted the greening of the national territory, which has greatly improved the ecological level of each municipality. Shandong attaches more importance to green technology and forest resources, and since 2004 Shandong has held several green technology exhibitions and environmental technology forums to promote the exchange of green technology among cities, and has established a forest chief system and actively promoted the greening of the national territory, which has greatly improved the ecological level of each city. So, the improvement of ecological protection in Shandong cities is mainly dependent on the implementation of policies and is characterized by a two-wheeled core influence of the green technology innovation level and the afforestation area per capita.

5. Conclusions and Policy Implications

5.1. Conclusions

According to the data from 55 cities in the middle and lower reaches of the Yellow River for 2009–2019, this paper studied the spatio-temporal evolution pattern of ecological protection level, analyzes the spatial and temporal characteristics of ecological protection levels through Arcgis mapping, and uses the GTWR model to reveal the influence degree of its influencing factors. Findings of the study can be summarized as:

(1)

The ecological protection level shows a steady upward trend, but the spatial distribution varies greatly. The ecological level increased from 0.222 to 0.336 for 2009–2019, and the growth rate was 51.35%. The growth trend is divided into three stages: slow growth from 2009–2011, rapid growth from 2012–2017, and slower growth from 2018–2019. The level of ecological protection shows a decreasing distribution trend from the coast to the interior. Most cities with high-level ecological protection are located at Shandong and Henan provinces. While most cities from Shaanxi provinces have a low-level ecological protection.

(2)

The ecological protection level shows a significant positive spatial correlation and are mainly clustered in the same category. The index of Global Moran’s I are greater than 0.1, which all pass the significance test, the spatial positive correlation of ecological protection level was significant. The change of Global Moran’s I index is small, indicating that the ecological protection level clustering state is relatively stable between cities. The first and third quadrants of the Moran scatter plot have the majority of cities distributed, showing a clear clustering of same agglomeration. The third quadrant belongs to the “low-low” category and distributes the largest number of cities.

(3)

The factors affecting the ecological protection level differ in spatial and temporal distribution. The ecological protection level is strongly influenced by the green technology innovation level in three periods, mainly by a positive way, while the intensity of fertilizer application, amount of agricultural film applied, and afforestation area per capita mostly show a small and negative influence on the level of ecological protection. The core influencing factor for most cities in Shaanxi is the amount of agricultural film applied, for most cities in Henan and Shanxi provinces is mainly the green technology innovation level, the core influencing factors for most cities in Shandong is the afforestation area per capita and green technology innovation level.

5.2. Policy Implications

Based on the empirical results and main conclusions of this paper, the cities in the middle and lower reaches of the Yellow River have a wide range of resource endowments and need to promote the level of ecological protection according to local conditions. Based on the findings, the study suggests the following policy implications.

Implement regional differentiated development strategies to promote the sustainable development of the ecological environment. According to the ecological condition of the area and relevant policies, the functional areas are divided into prohibited development zones and restricted development zones for differentiated management. Prohibited development zones are based on the core criterion of protecting the ecological environment and prohibit all activities of development and construction. Restricted development zones can engage in activities such as developing the economy based on protecting the ecological regions with low levels of ecological protection, promote industrial optimization and development, while high regions increase ecological compensation. In the middle reaches of the Yellow River, most cities are at lower ecological protection levels, but the coal and natural gas production and other chemical industries in the regional bases of Shaanxi and Shanxi have formed a scale advantage. They are concerned with national energy security, so the relevant enterprises cannot be directly shut down and banned. It is necessary to implement industrial upgrading policies, ensure environmental protection investment, encourage enterprises to improve their sewage and waste gas purification rates by introducing sewage and waste gas treatment systems, and support the environmentally friendly oriented economic growth of the chemical bases to improve the regional ecological environment. The cities located in the lower reaches of the Yellow River have a high level of ecological protection, but are densely populated and under great ecological pressure. It is important to adhere to the concept that ecological protection is the protection of productivity, raise environmental regulation standards, increase spending on scientific research for environmental protection, set up ecological and environmental monitoring platforms, increase funding for ecological, environmental protection and restoration, such as wetlands, to promote a sustainable improvement in the ecological environment.

Realize cross-regional linkages and collaborate to improve ecological protection in the basin. First, following the strategy of “Ecological Protection and Quality Development of the Yellow River basin”, the basin area needs to be united and the provinces need to abandon the idea of regional independence in dealing with environmental pollution and break through administrative boundaries, set targets for ecological protection and enhancement for the whole area, and build a unified institutional mechanism, such as create an ecological environment protection alliance to form a regional layout for synergistic development and complementary advantages. Second, we should jointly organize academic conferences and forums focusing on ecological protection, strengthen the exchange of advanced technology and management experience between regions, promote the free flow of technical personnel, environmental information, and other factors to allocate factors efficiently, and take full advantage of high-level areas as a bellwether and implement ecological protection measures jointly with surrounding low-level areas. To enhance the radiation-driven effect of high-level agglomeration areas and avoid the phenomenon of “hollow” at the junction of high-level agglomeration areas and low-level agglomeration areas.

It is required to focus on differences in core influencing factors to improve ecological protection efficiently. First, the government should strengthen its efforts to educate farmers, reduce their use of traditional agricultural films, implement an incentive system, and promote organic and biodegradable agricultural films to reduce the negative ecological impact of traditional agricultural films. Second, the government should take notice of the green technology, which can efficiently improve the ecological environment, use preferential policies to recruit scientific and technological research and development personnel, increase financial expenditure to support the research and development of green technology innovation level, strengthen its driving ability, and set up a special fund and talent pool for green technology talents, a green innovation research and development institution and a green technology trading market in some cities with better economic condition and innovation environment to promote research and development, introduction and diffusion of green technology innovation levels.