Abstract

In order to quantify the relationship between human activity intensity and the evolution of stream structure and connectivity, the characteristics of spatial distribution and temporal evolution of stream structure and connectivity in different periods were studied by using river structure parameters and landscape ecology methods. The results indicate the following. (1) The mainstream trend of stream system, with the stream development coefficient decreased by 18.35% and 22.50% for the second- and third-level of stream, is obvious. (2) Human activity intensity has a significant impact on the change of stream structure and connectivity. Under the background of economy domination, the stronger human intervention is, the simpler stream structure will be, and the poorer stream connectivity will be. (3) From the view of time series, it is considered that human activities are the main reasons for affecting the change of stream structure and connectivity in the short term, while the natural factors have little effect. (4) The length of low-level streams is more affected by human activities than that of the high-level streams, but water surface ratio is opposite. This study can provide a reference for coordinating the contradiction between socio-economic development and protection of stream systems.

INTRODUCTION

River system is an important channel for fluvial material and energy transmission. The connectivity and integrity of its structure play a crucial role in maintaining the health of river ecosystem (Bianchi et al. 2013). However, with the acceleration of social and economic development, the stream structure and connectivity are constantly changing under the influence of human activities, resulting in the decline of stream function, storage capacity and self-purification ability (Bergkamp et al. 2000; Zuo et al. 2016a). Currently, the situation of stream structure damage and connectivity weakness are incompatible with the requirement of the national strategy of Ecological Civilization Construction (Lei & Dayong 2011; Wei et al. 2011; Zhang et al. 2016). Therefore, in order to coordinate the contradiction of social economic development and stream protection, it is of great significance to summarize evolution characteristics of stream structure and connectivity and to explore the relationship between human activity intensity and stream structure and connectivity evolution from the perspective of time series and space pattern, respectively.

There has been a deep research on the structure of the stream system. Horton (1945) and Strahler (1952) verified the self-similarity of the natural streams in structural composition, and proposed the famous Horton law and Strahler stream classification. Subsequently, the law of the stream structure has been continuously improved and developed (Tarboton 1996; Yuan et al. 2005). The stream self-similarity theory and fractal theory were widely used in the research of stream morphology (Mandelbrot & Aizenman 1979; Barbera & Rosso 1989; Nikora & Sapozhnikov 1993), after the fractal theory was originally developed by Mandelbrot in the 1970s. On the study of stream connectivity, Western et al. (2001), Pringle (2003), and Freeman et al. (2007) believed that stream connectivity was the movement efficiency of runoff from sources – mainstream – watershed network – outlets, or the transfer rate of organic matter and energy with the water as the medium among the elements of water cycle. Gubiani et al. (2007) pointed out that stream connectivity was the material, energy and organisms conversion in hydrosphere or elements of hydrosphere with water as the medium, from the perspective of biological ecology. Turnbull et al. (2008) considered that the dynamic and function connectivity were the regional mutual connection dynamic properties described by the process. Atkins et al. (2016) assessed the connectivity of groundwater-surface water using chemical tracers in an Australian river system.

Shaying River Basin, which is located in the central part of Henan province and northwest of Anhui province, is a typical watershed with multi-sluices, high pollution and dense population. It is the important guarantee of water for agricultural production in Henan province (national important grain production base). A large number of sluices have been built in the middle and upper reaches, and water pollution is serious in the middle and lower reaches as the influence of industrial, agricultural and urban development along the river (Zuo et al. 2015). Due to the strong interference of human activities, the self-purification capacity of streams decreased. As a result, ecological water demand could not be effectively met, and the stream integrity, water mobility, water quality state and biological diversity have been seriously damaged. In order to explore the influence of human activity intensity on the change of stream structure and connectivity from the perspective of time series, characteristics of stream structure and connectivity evolution in the basin were analyzed based on remote sensing images interpretation with four representative periods of 1965, 1989, 2000 and 2015. On the base of regionalization of human activities, the characteristics of stream structure and connectivity evolution in different regions were summarized.

STUDY AREA AND DATA

Study area

Shaying River, which originates in Mount Song in Henan province (111°56′44″–116°31′07″E, 32°29′24″–34°57′15″N), is the largest tributary of Huai River and the area is 39,075.30 km2 (Figure 1). Across Henan and Anhui provinces, the river flows through more than 40 cities or counties, flowing into the Huai River in Mohekou Town of Yingshang County. The study area is located in the transition zone from warm temperate zone to subtropical zone. The annual average temperature is 14–16 °C, and the average annual precipitation is 753.43 mm. The inter-annual variation of rainfall is large and unevenly distributed during the year with 6–9 months accounting for about 60–80% of the year. There are 14 streams with the catchment area above 1,000 km2, and more than 100 streams with the catchment area between 100 and 1,000 km2 (Zuo et al. 2016b). From upstream to downstream, streams on the right bank of the mainstream are Sha River, Li River, North Ru Rive and Fenquan River, etc.; on the left bank of the mainstream are Qingyi River, Jialu River, Xinyun River, Xincai River, etc. (Table 1). In recent years, industry and agriculture have developed rapidly in the river basin. The light industry dominated by textile industry, is distributed mainly in Zhengzhou, Pingdingshan, Zhoukou and Fuyang. Great efforts are made in emerging industries, and an industrial system has been formed with the mainstay of metallurgy, coal industry, building materials, electricity, textile, paper making, brewing, chemical, pharmaceutical, food processing and cigarettes in Shaying River Basin. These efforts will play an important role in the social and economic development, but it will also cause some impact on the structure and connectivity of the stream system.

Table 1

Geographic attributes of the main tributaries of Shaying River

Stream nameLength/kmWatershed area/km2Flowing through
Sha River 365.03 12,522.39 Lushan Country, Pingdingshan urban areas, Xiangcheng Country, Wuyang Country, Luohe City, Shangshui Country, Zhoukou City 
Li River 162.81 2,707.22 Ye Country, Wuyang Country, Luohe City 
North Ru River 264.30 5,845.65 Song Country, Ruyang Country, Ruzhou City, Xia Country, Baofeng Country, Ye Country, Xiangcheng Country 
Qingyi River 112.23 2,168.12 Xuchang Country, Linying Country, Yanling Country, Xihua Country 
Jialu River 256.88 6,748.38 Xinmi City, Zhongmou Country, Kaifeng City, Weishi Country, Fougou Country, Xihua country, Zhoukou urban areas 
Xinyun River 59.09 1,457.19 Taikang Country, Fugou Country, Xihua Country, Huayang Country 
Xincai River 89.50 1,660.91 Huaiyang Country, Dancheng Country, Shenyang Country 
Fenquan River 241.22 7,758.16 Shangshui Country, Xiangcheng City, Shenqiu Country, Jieshou City, Linquan Country, Fuyang City 
Stream nameLength/kmWatershed area/km2Flowing through
Sha River 365.03 12,522.39 Lushan Country, Pingdingshan urban areas, Xiangcheng Country, Wuyang Country, Luohe City, Shangshui Country, Zhoukou City 
Li River 162.81 2,707.22 Ye Country, Wuyang Country, Luohe City 
North Ru River 264.30 5,845.65 Song Country, Ruyang Country, Ruzhou City, Xia Country, Baofeng Country, Ye Country, Xiangcheng Country 
Qingyi River 112.23 2,168.12 Xuchang Country, Linying Country, Yanling Country, Xihua Country 
Jialu River 256.88 6,748.38 Xinmi City, Zhongmou Country, Kaifeng City, Weishi Country, Fougou Country, Xihua country, Zhoukou urban areas 
Xinyun River 59.09 1,457.19 Taikang Country, Fugou Country, Xihua Country, Huayang Country 
Xincai River 89.50 1,660.91 Huaiyang Country, Dancheng Country, Shenyang Country 
Fenquan River 241.22 7,758.16 Shangshui Country, Xiangcheng City, Shenqiu Country, Jieshou City, Linquan Country, Fuyang City 
Figure 1

Distribution of the main rivers of study area. A: Sha River; B: North Ru River; C: Qingyi River; D: Jialu River; E: Xinyun River; F: Li River; G: Xincai River; H: Fenquan River.

Figure 1

Distribution of the main rivers of study area. A: Sha River; B: North Ru River; C: Qingyi River; D: Jialu River; E: Xinyun River; F: Li River; G: Xincai River; H: Fenquan River.

Data

According to the purpose of the study, the data of this paper are divided into two categories. The first type of data is used to calculate human activity zoning; the second type of data is used to extract the characteristic parameters of the water system. The first type of data includes: digital elevation model (DEM), population and sluice numbers. The data of DEM with spatial resolution of 30 m are obtained from Geospatial Data Cloud: http://www.gscloud.cn/, the data of population come mainly from the Statistical Bureau of Henan Province and Anhui Province in 2015, and the position of sluices are derived from the Hydrological Bureau of the Huaihe River Basin. The second type of data mainly refers to remote sensing images, including four periods (1965, 1987, 2000 and 2015). The data of 1987 come from Landsat 5 satellite digital products, 2000 from Landsat 7 ETM SLC-on satellite digital products, and 2015 from Landsat 8 OLI_TRIS satellite digital products. These remote sensing images were derived from Geospatial Data Cloud and made a series of preprocessing such as radiometric correction, geometric correction, histogram matching, image stitching, image filtering and enhancement. Normalized difference water index (NDWI), as an important basis for stream extraction, was obtained by the superposition calculation of green band and near-infrared band, NDWI = (p(Green)-p(NIR))/(p(Green) + p(NIR)), where p(Green) is green band and p(NIR) is near-infrared band. The stream map of 1965 with no remote sensing data was obtained by reverse method on the base of the stream map of 1987, combined with paper topographic maps, water conservancy annals and so on. The specific steps are as follows: first of all, to determine the quantity and distribution of water conservancy projects construction in the Shaying River Basin during the period between 1965 and 1987 by viewing water conservancy annals and local magazines, and to remove them from the map of 1987; secondly, paper maps close to 1965 were digitized for revising the stream map of 1965.

METHODOLOGY

In order to explore the relationship between human activity intensity and the change of stream system structure and connectivity, the watershed is firstly divided into different levels of human activity regions according to its physiographic characteristics and socio-economic development status. Secondly, the structure and connectivity parameters of stream system of four representative periods (1965, 1987, 2000 and 2015) in different human activity regions were extracted by geo-statistic method. Finally, variation characteristics of stream structure and connectivity were analyzed from the perspective of time and space, respectively. The methods used included human activity index (HAI), and stream structure and connectivity parameters.

HAI

HAI is used for the comprehensive quantitative evaluation of human activity intensity. With the complexity of human activities, variables considered in HAI calculation are increasing. Variables need to be selected according to the actual situation of the basin. Shaying River is the largest tributary of the Huaihe River, and the number of sluices and dams in the Huaihe River Basin is the largest in China (Yan et al. 2009; Zhang et al. 2010). In view of the prominent features of population density, high pollution and a large number of dams and sluices in Shaying River Basin, the sluice and dam factors should not be ignored in the study of the effect of human activities on stream system structure and connectivity. Therefore, taking full account of the actual situation of Shaying River Basin, sluices and dams were selected as one of the main factors in HAI calculation. In addition, due to the unique natural geographical characteristics of the Shaying River Basin, the level of economic and social development of the basin is highly correlated with its topography. The idea of human activity intensity partition in the paper is as follows: firstly, each factor was partitioned separately, and then the partition results of multiple factors were superimposed (natural geography and human activity factors). Natural geography factors were considered as a separate aspect, and their partition results were superimposed with the partition results of HAI. In order to avoid the information overlap, population density and the sluices and dams were selected in the calculation of HAI (Costanza 1997; Lazo 1998; Su et al. 2012).  
formula
(1)
where is the HAI in region i; is the population density index in region i; is the sluice number index in region i; is the weight, , , in this paper.

Stream structural parameters

Stream structural parameters are used to characterize the impact of human activity intensity on stream structure state. Drainage density and water surface ratio are usually used to reflect the changes of stream structure in their numbers; stream development coefficient, stream length and mainstream area-length ratio reflect the changes of stream structure in their law; and fractal dimension of stream network is an important parameter of reflection of stream structure in their complexity (Table 2).

Table 2

Evaluation index system of stream structure and connectivity

Criteria layerIndexesFormulasPhysical meaning
Structure Drainage density   Length development of stream network 
Water surface ratio   Area development of stream network 
Stream length  – Length development of stream network 
Stream development coefficient   Development degree of stream system for different grades 
Mainstream area-length ratio   Non-synchronization of mainstream's area and length development 
Fractal dimension of stream network   Complexity of the stream system 
Connectivity Node connection rate   Ability of each stream node to connect with their surrounding streams 
Stream connectivity degree   Connectivity degree of stream system 
Criteria layerIndexesFormulasPhysical meaning
Structure Drainage density   Length development of stream network 
Water surface ratio   Area development of stream network 
Stream length  – Length development of stream network 
Stream development coefficient   Development degree of stream system for different grades 
Mainstream area-length ratio   Non-synchronization of mainstream's area and length development 
Fractal dimension of stream network   Complexity of the stream system 
Connectivity Node connection rate   Ability of each stream node to connect with their surrounding streams 
Stream connectivity degree   Connectivity degree of stream system 
Drainage density (km/km2, Equation (2)) is the total length of the river per unit area, reflecting the development of stream length (Carlston 1963; Tucker et al. 2001). For the natural watersheds with less influence of human activities, the bigger the value, the greater the basin's cutting strength and the faster its hydrological response to rainfall. On the contrary, it has the opposite effect.  
formula
(2)
where l is the total length of stream system within the basin (km); A is the basin area (km2).
Water surface ratio (%, Equation (3)) is the proportion of total area of rivers and lakes within the basin accounting for the basin area, reflecting the development of stream area. In general, the bigger the value, the stronger the ability for the river system to reduce flood peak and the better for the protection of stream ecological environment (Gao et al. 2014).  
formula
(3)
where is the total area of lakes and rivers within the basin (km2).
Stream development coefficient (km/km, Equation (4)) refers to the ratio of the tributaries length in different grades to the mainstream, reflecting the development of tributaries of different grades. The bigger the value, the greater development of tributaries of different grades is.  
formula
(4)
where is the length of the stream system for the level of w (km); is the length of mainstream (km).
Mainstream area–length ratio (km2/km, Equation (5)) refers to the ratio of mainstream's area to its length, reflecting non-synchronization of mainstream's area and the development of the length. The bigger the value, the stronger the river discharge capacity.  
formula
(5)
where is the area of mainstream (km2).
Fractal dimension of stream network D (−, Equation (6)) indicates the complexity of the river system (Barbera & Rosso 1989; Tarboton et al. 1990). The bigger the value, the more complicated the river system.  
formula
(6)
where D is fractal dimension of stream network (−); is the branching ratio; is the length ratio; ; is the slope of the linear regression between logarithm of or and . Generally speaking, the value of changes from 3 to 5, and the value of from 1.5 to 3.

Connectivity parameters

Stream connectivity is an important reflection of functional state for different types of water bodies (rivers, lakes, wetlands). The connectivity characteristics of river system reflect the interference degree by human activities, as well as the effective degree of river's natural ecological function. Based on the calculation method of river corridor connectivity in landscape ecology, indexes of and are used to indicate the connectivity status of river system in Shaying River Basin (Table 2).

Node connection rate (−, Equation (7)) is an important indicator to measure the ability of each stream node to connect with their surrounding streams. The value of ranges from 0 to 3: when , this indicates that there is no stream network; when , this indicates the formation of a single stream loop; when , this indicates that the structure of stream system is tree-like development; when , this indicates that the connection of stream network is more complex. Stream connectivity degree (−, Equation (8)) is the ratio of the number of river connecting lines and the maximum possible connecting lines. The value of ranges from 0 to 1. The bigger the value, the better the stream connectivity. When , this indicates that the node has no connection line; when , this indicates that each node is connected to each other.  
formula
(7)
 
formula
(8)
where is node connection rate; L is the number of stream chains; N is the number of stream nodes; is stream connectivity degree; is the maximum possible number of connecting lines of stream network.

RESULTS AND DISCUSSION

Regionalization of human activity intensity

With the administrative region as the calculating unit, the HAI was calculated using Equation (1). As topographic factors were closely related to the intensity of human activities, the feature of topographic differentiation was significant in Shaying River Basin (Figure 2(a)). Altitude and slope were selected in the regionalization of human activity intensity. The division results based on altitude and slope were overlaid with the results of HAI. The watershed was divided into three regions, namely low human activity region, moderate human activity region and high human activity region (Figure 2(b)).

Figure 2

DEM and human activity regionalization in Shaying River Basin.

Figure 2

DEM and human activity regionalization in Shaying River Basin.

Region I (Low human activity region): the average altitude is 886.19 m and the average slope 20.43°. The categories of land use are dominated by forest land, accounting for 83.92% of the total area, and there was almost no urban–rural residential land. Due to the higher terrain and less human activities, the natural ecosystems in this region were relatively well preserved.

Region II (Moderate human activity region): The average altitude was 363.13 m and the average slope 8.53°. The categories of land use are dominated by arable land, accounting for 56.60% of the total area, followed by forestland, grassland and urban–rural residential land. Due to the impact of human activities, degradation of water quality and damage of water ecosystem was serious in parts of the river.

Region III (High human activity region): The average altitude was 77.32 m and the average slope 1.03°. Land use categories were dominated by arable land, accounting for 76.24% of the total area, followed by the urban–rural residential land, and the distribution of grassland and woodland are less. Due to the flat terrain and the suitable climate, the region had the remarkable characteristic of dense population and intense human activities. The industry, agriculture and urban development in this region in recent years had caused tremendous pressure on natural ecosystems. Water quality was severely deteriorated, and the water ecosystem was badly damaged in most of the river.

Stream extraction and classification

With the background of major events of China's historical development and its impact on water conservancy development, and the actual situation of Shaying River Basin, four representative periods of 1965, 1987, 2000 and 2015 were selected to study the variation characteristics of stream structure and connectivity in Shaying River Basin. The stream system can basically be considered in a natural state due to the small impact of human activities in 1949. However, China experienced the Great Leap Forward period and set off a reservoir construction boom from 1950 to 1965. Vast majority of sluices and dams were built during this period, especially in the Huaihe River Basin. Therefore, 1965 was a year with special significance. During 1965–1987, China's water conservancy construction was almost at a stagnant stage due to the Cultural Revolution and Reform and Opening. Water conservancy development in Shaying River Basin was also in a stagnant stage, and the effect of human activities on the stream was small. With the economic and social development, human demand for water resources was growing. During 1987–2000, exploitation and utilization of water resources was excessive, and water disasters had occurred frequently. Water resources, water environment, and water ecology had become important issues in China's water conservancy development during this period. Since 2000, development and protection of water resources has been paid more and more attention, and the ideas of comprehensive resource utilization, ecological restoration and environmental protection have been gradually accepted. Based on the above analysis, the four representative periods of 1965, 1987, 2000 and 2015 were selected. For the representative periods with the data of remote sensing image (1987, 2000 and 2015), NDWI was calculated firstly, and then stream system were extracted preliminarily combined with multi-band visual interpretation. In order to ensure that the extracted stream system was consistent with the actual situation, many times of field investigation had been organized by the research team within the river basin. Since there were no remote sensing data for 1965, the data of stream system in this period were obtained by the method of reverse deduction based on stream system of 1987 and other dates such as Water Conservation Annals and Regional Annals. In accordance with above-mentioned ideas and methods, the spatial distribution of the surface stream system for the four representative periods was obtained (Figure 3).

Figure 3

Stream maps of each representative period (1965, 1987, 2000 and 2015) in Shaying River Basin.

Figure 3

Stream maps of each representative period (1965, 1987, 2000 and 2015) in Shaying River Basin.

It can be seen from Figure 3 that the structure of the water system showed a decreasing trend from 1965 to 2015. From the perspective of spatial evolution, the change of stream structure was small in the upper reaches of the basin, and larger in the middle and lower reaches of the basin with a clear downward trend. In order to quantitatively describe the change of the stream structure and connectivity of the basin, parameters of the stream structure and connectivity were calculated. However, the terrain in Shaying River Basin was relatively flat. The stream system did not always match the Horton law, and it was classified according to its size, function and status in this paper. If the stream width was greater than 20 m, it would be divided into the first level of stream (mainstream); if the width between 10 and 20 m, it would be divided into the second level of stream; if the width between 0 and l0 m, it would be divided into the third level of stream.

Change characteristics of stream structure

The change of stream structure has very important influence on the storage capacity of rivers and lakes, and the prevention of flood disaster. On the basis of stream extraction and classification (linear), stream quantity and structure characteristic parameters and corresponding rate of change were calculated by statistical method for analyzing stream evolution characteristics in Shaying River Basin (Table 3).

Table 3

The results of stream quantity and structure characteristic parameters and corresponding rate of change during the representative period

Representative periods (year)
Rate of change (%)
Indexes19651987200020151965–19871987–20002000–2015
Kw Dd 0.21 0.20 0.18 0.17 −6.05 −9.75 −3.16 
Wp/% 0.58 0.53 0.52 0.51 −9.32 −0.66 −3.61 
L 8,101.09 7,599.49 6,899.02 6,512.93 −6.19 −9.22 −5.59 
Second-level 2.18 2.05 1.91 1.78 −6.21 −6.91 −6.68 
Third-level 3.13 3.20 2.58 2.48 2.21 −19.24 −4.15 
RAL 0.10 0.10 0.10 0.10 8.69 −1.33 0.77 
D 1.37 1.17 1.35 1.38 −14.24 14.82 2.33 
Representative periods (year)
Rate of change (%)
Indexes19651987200020151965–19871987–20002000–2015
Kw Dd 0.21 0.20 0.18 0.17 −6.05 −9.75 −3.16 
Wp/% 0.58 0.53 0.52 0.51 −9.32 −0.66 −3.61 
L 8,101.09 7,599.49 6,899.02 6,512.93 −6.19 −9.22 −5.59 
Second-level 2.18 2.05 1.91 1.78 −6.21 −6.91 −6.68 
Third-level 3.13 3.20 2.58 2.48 2.21 −19.24 −4.15 
RAL 0.10 0.10 0.10 0.10 8.69 −1.33 0.77 
D 1.37 1.17 1.35 1.38 −14.24 14.82 2.33 

Quantitative characteristics

The number of stream system tended to decrease from the perspective of change of stream quantity characteristics. Drainage density and water surface ratio showed a decreasing trend (Table 3). Drainage density decreased by 6.05%, and water surface ratio decreased by 9.32% from 1965 to 1987, and the integrated reduction rate was the largest compared with 9.75% and 0.66% from 1987 to 2000, and 3.16% and 3.61% from 2000 to 2015. It was speculated that this trend was related to the historical background of China's social development. China was facing the strategic opportunity for reform and opening during 1965 to 1987. Social development focused on economic construction during this period, resulting in water conservancy development lagging behind. Excessive use of water resources made the surface area of water gradually reduce and part of the rivers even dry up, which led to a significant change of drainage density and surface water ratio. The trend of stream quantitative structure showed that human activities had a negative impact on the change of stream quantitative characteristics under the situation of economy dominated. The stronger the human activity intensity, the more obviously a stream quantitative characteristic tended to reduce.

Figure 4 shows the variation of stream quantity parameters with time in different human activity intensity regions from 1965 to 2015. It can be seen from the figure that the change of stream quantity parameters (drainage density and water surface ratio) with time was small in the low human activity region. In contrast, the change of stream quantity parameters with time was large in the medium and high human activity region. As the stream evolution in low human activity region was mainly driven by natural factors, we believed that the influence of natural factors on stream structure and connectivity was not obvious in a short time. On the contrary, human activities were the main factors influencing the change of stream structure and connectivity in the short term. Figure 4 also shows that the drainage density and water surface ratio of stream networks in moderate and high human activity regions appeared a decreasing trend. The reduction rates of drainage density and water surface ratio were 0.017 and 0.037, which were larger than those of in moderate human activity region (0.015 and 0.019). This further illustrated the relationship between human activity intensity and evolution of stream quantity characteristic from a spatial distribution point of view. The stronger the human activity intensity, the more obvious the change of stream quantitative characteristic in the economy dominated areas.

Figure 4

Changes of stream quantitative characteristics in different human activity regions.

Figure 4

Changes of stream quantitative characteristics in different human activity regions.

Structural characteristics

From the perspective of structure change, the stream structure in Shaying River Basin tended to be simple (Table 3). Stream development coefficient of the second level of stream decreased gradually from 1965 to 2015, and the third level of stream increased first and then decreases with 1987 as the node. The reduction rate of the third level of stream was greater than that of the second level of stream after 1987. The changing trend indicated that the mainstream trend of the stream system was obvious and the low-level streams were more susceptible to human activities. Mainstream area-length ratio showed a slow increasing trend from 1965 to 2015, indicating that the water surface area of mainstream kept getting big. The reason might be that large numbers of water conservancy projects (reservoirs, sluices) had been built in the basin, which led to an increase of stream surface area as reservoir impoundment. Fractal dimension of stream network was an important index to characterize the complexity of stream system. From Table 3 it can be seen that the change of fractal dimension of stream network was complicated with a decreasing trend from 1965 to 1987 and an increasing trend from 1987 to 2015. Among them, the rate of change of fractal dimension was 14.24% from 1965 to 1987, 14.82% from 1987 to 2000, and 2.33% from 2000 to 2015, indicating that the change of complexity of stream system began to decrease after 2000. The reason might be that human activities in the Shaying River Basin were very intense before 2000 at the development stage of engineering water conservancy. Besides, water shortage was serious during this period, and the modification effect of stream system was big for meeting the need of rapid social and economic development, leading to great change of stream complexity. Since 2000, China has entered the development stage of ecological water conservancy with a focus of ecological environment protection. Stream protection was highly valued during this period and its complexity change was reduced.

The length and surface area of first, second and third level of streams were calculated for analyzing the effect of human activity intensity on different levels of stream (Table 4). The length of the third level of stream decreased by 810.56 km from 1965 to 2015 with a reduction rate of 20.89%, and 4.06 km2 for water surface area with a reduction rate of 20.92%; the length of the second level of stream decreased by 501.46 km with a reduction rate of 18.53%, and 8.93 km2 for water surface area with a reduction rate of 14.43%; the length of mainstream decreased by 276.14 km with a reduction rate of 18.21%, and 16.99 km2 for water surface area with a reduction rate of 11.61%. The length and surface change of third level of stream was obviously greater than that of the second and main stream, which illustrated that low-level streams were more susceptible to human activities and the reduction of the river length was mainly caused by the disappearance of the third level of streams. As the proportion of low-level streams in water surface area was small, we believed that the lengths of low-level streams were more affected by human activities than high-level streams, while water surface areas were more easily affected by human activities with high-level streams.

Table 4

The length and area changes for different levels of streams

Representative periodsLength/km
Area/km2
First-levelSecond-levelThird-levelFirst-levelSecond-levelThird-level
1965 1,515.90 2,705.85 3,879.34 146.32 61.86 19.40 
1987 1,216.97 2,490.83 3,891.69 127.68 59.23 19.46 
2000 1,257.18 2,395.25 3,246.59 130.14 58.63 16.23 
2015 1,239.76 2,204.39 3,068.78 129.33 52.93 15.34 
Representative periodsLength/km
Area/km2
First-levelSecond-levelThird-levelFirst-levelSecond-levelThird-level
1965 1,515.90 2,705.85 3,879.34 146.32 61.86 19.40 
1987 1,216.97 2,490.83 3,891.69 127.68 59.23 19.46 
2000 1,257.18 2,395.25 3,246.59 130.14 58.63 16.23 
2015 1,239.76 2,204.39 3,068.78 129.33 52.93 15.34 

In order to analyze the impact of human activities on stream of different levels from perspective of spatial pattern, the changes of stream length and surface area were analyzed in low human activity region, moderate human activity region and high human activity region, respectively (Figure 5). It can be seen that the variation of steam length and water surface area was the most obvious in high human activity region, showing a downward trend, followed by moderate human activity region and then low human activity region (basically remain unchanged). This further illustrated the significant influence of human activity intensity on change of stream structure characteristics. The degree of interference of human activity on stream structure increased with human activity intensity, and this interference made the evolution of stream structure towards the direction of simplification.

Figure 5

Characteristics of stream structure changes in different human activities regions.

Figure 5

Characteristics of stream structure changes in different human activities regions.

Change characteristics of stream connectivity

Stream system, as an important carrier of matter and energy transport, has nature functions as water and sediment transport, river bed shaping, water purification, and social functions as shipping, power generation and irrigation. Stream connectivity plays an important role in raising water mobility, improving stream ecological environment and strengthening the ability of urban flood and waterlogging control. At present, the situation of stream connectivity is usually described by and index, which are important indicators expressing the corridor connectivity in landscape ecology. The calculation results are listed in Table 5.

Table 5

The calculation results of stream connectivity parameters and corresponding rate of change for the representative periods

IndexesRepresentative periods (year)
Rate of change (%)
19651987200020151965–19871987–20002000–2015
 2.12 1.96 1.96 1.94 −7.89 −0.25 −0.92 
 0.71 0.66 0.66 0.65 −7.82 −0.35 −0.85 
IndexesRepresentative periods (year)
Rate of change (%)
19651987200020151965–19871987–20002000–2015
 2.12 1.96 1.96 1.94 −7.89 −0.25 −0.92 
 0.71 0.66 0.66 0.65 −7.82 −0.35 −0.85 

It can be seen from Table 5 that and indexes of 1965 were larger than the other representative periods, indicating that stream connectivity was good during this period. Because people's ability of stream transform was weak in 1965, and streams in the basin basically maintained the original appearance. Indexes of β and γ showed a decreasing trend from 1965 to 2015, and the rate of decrease were the fastest from 1965 to 1987 with a value of 7.89% for and 7.82% for . After 1987, the decreasing speed was reduced. The trend of index indicated that the connectivity between corridor and nodes was weakened. On the one hand, due to the serious shortage of water resources in the Shaying River Basin, a large number of sluice and reservoir projects had been built (Figure 1) since the 1960s for meeting the need of social production development, which blocked the connectivity of water bodies. On the other hand, with the accelerating urbanization process, large numbers of low-level systems disappeared due to landfills in the river basin, resulting in a decrease in the number of low-level stream chains, so the promotion of stream connectivity capability was hampered. The change of water connectivity indexes indicated that the impact of human activities on water connectivity was significant. For Shaying River Basin, the intensity of human activities was inversely related to water connectivity (i.e. water connectivity decreased with the increase of human activity intensity). Since regionalization of human activity did not take into account the integrity of stream system, it was not possible for statistical analysis of changes of and indices in different human activity regions. That is, it was impossible to analyze the impact of human activities on stream connectivity from the perspective of spatial pattern, which would be the focus for further research.

CONCLUSIONS

Summarizing characteristics of stream structure and connectivity evolution and analyzing relationships between human activity intensity and steam structure and connectivity evolution, are helpful to understand the evolution law of stream system under the influence of different human activity intensity. Supported by remote sensing data, the variation characteristics of stream structure and connectivity had been analyzed in the paper according to four representative periods of 1965, 1987, 2000 and 2015. Based on partitions, the relationships between human activity intensity and variation of stream structure and connectivity were discussed. The main conclusions are as follows:

  • (1)

    The stream structure in the river basin tended to be simplified. The second and third levels of stream were reduced by 18.35% and 22.50% from 1965 to 2015. The stream length and water surface area decreased by 1587.82 km and 29.98 km2 from 1965 to 2015. This indicated that the mainstream trend in the basin is obvious.

  • (2)

    The influence of human activities on the length of low-level streams was greater than that of high-level streams, while the water surface area was opposed. The results showed that the length of the third-level streams decreased by 810.22 km from 1965 to 2015, and their area of water surface decreased by 4.06 km2; decreased by 501.46 km for the length and 8.93 km2 for water surface area of the second-level streams; decreased by 276.14 km for the length and 16.99 km2 for water surface area of the first-level streams, indicating that the low-level streams are more likely to disappear than that of the high-level streams under the influence of human activities.

  • (3)

    There was a significant effect of human activity intensity on the structure and connectivity of stream system. In the context of economic development, the stronger the human activity intensity was, the greater the effect on structure and connectivity of the stream system would be.

ACKNOWLEDGEMENTS

This research is supported by National Natural Science Foundation of China (No. 51279183 and 51509222) and Program for Innovative Research Team (in Science and Technology) in University of Henan Province (No. 13IRTSTHN030).

REFERENCES

REFERENCES
Barbera
,
P. L.
&
Rosso
,
R.
1989
On the fractal dimension of stream networks
.
Water Resources Research
25
(
4
),
735
741
.
Bergkamp
,
G.
,
McCartney
,
M.
,
Dugan
,
P.
,
McNeely
,
J.
&
Acreman
,
M.
2000
Dams, ecosystem functions and environmental restoration
.
Thematic Review II
1
,
1
187
.
Bianchi
,
T. S.
,
Allison
,
M. A.
&
Cai
,
W. J.
, (eds)
2013
Biogeochemical Dynamics at Major River-Coastal Interfaces: Linkages with Global Change
.
Cambridge University Press
,
Cambridge, UK
.
Carlston
,
C. W.
1963
Drainage Density and Streamflow
.
US Government Print Office
,
Washington, DC, USA
.
Freeman
,
M. C.
,
Pringle
,
C. M.
&
Jackson
,
C. R.
2007
Hydrologic connectivity and the contribution of stream headwaters to ecological integrity at regional scales1
.
Journal of the American Water Resources Association
43
(
1
),
5
14
.
Gao
,
C.
,
Liu
,
J.
,
Cui
,
H.
,
Wang
,
Z.
&
He
,
S.
2014
Optimized water surface ratio and pervious surface proportion in urbanized riverside areas
.
Environmental Earth Sciences
72
(
2
),
569
576
.
Gubiani
,
E. A.
,
Gomes
,
L. C.
,
Agostinho
,
A. A.
&
Okada
,
E. K.
2007
Persistence of fish populations in the upper Parana River: effects of water regulation by dams
.
Ecology of Freshwater Fish
16
(
2
),
191
197
.
Mandelbrot
,
B. B.
&
Aizenman
,
M.
1979
Fractals: form, chance, and dimension
.
Physics Today
32
(
5
),
65
66
.
Nikora
,
V. I.
&
Sapozhnikov
,
V. B.
1993
River network fractal geometry and its computer-simulation
.
Water Resources Research
29
(
10
),
3569
3575
.
Strahler
,
A. N.
1952
Hypsometric (area-altitude) analysis of erosional topography
.
Geological Society of America Bulletin
63
(
11
),
1117
1142
.
Tarboton
,
D. G.
1996
Fractal river networks, Horton's laws and Tokunaga cyclicity
.
Journal of Hydrology
187
(
1–2
),
105
117
.
Tarboton
,
D. G.
,
Bras
,
R. L.
&
Rodriguez-Iturbe
,
I.
1990
Comment on ‘on the fractal dimension of stream networks’ by Paolo La Barbera and Renzo Rosso
.
Water Resources Research
26
(
9
),
2243
2244
.
Tucker
,
G. E.
,
Catani
,
F.
,
Rinaldo
,
A.
&
Bras
,
R. L.
2001
Statistical analysis of drainage density from digital terrain data
.
Geomorphology
36
(
3
),
187
202
.
Western
,
A. W.
,
Bloschl
,
G.
&
Grayson
,
R. B.
2001
Toward capturing hydrologically significant connectivity in spatial patterns
.
Water Resources Research
37
(
1
),
83
97
.
Yan
,
J.
,
Jun
,
X.
,
Wang
,
G. S.
&
Zhao
,
C. S.
2009
Simulation of environmental change in response to operation of dams in Huaihe Basin
.
Water Science and Engineering
2
(
3
),
27
37
.
Yuan
,
W.
,
Yang
,
K.
,
Tang
,
M.
&
Xu
,
Q. X.
2005
Stream structure characteristics and their impact on storage and flood control capacity in the urbanized plain river network
.
Geographical Research
5
,
008
.
Zhang
,
Y. Y.
,
Jun
,
X.
,
Tao
,
L.
&
Shao
,
Q. X.
2010
Impact of water projects on river flow regimes and water quality in Huai River Basin
.
Water Resources Management
24
(
5
),
889
908
.
Zhang
,
X.
,
Wang
,
Y.
,
Qi
,
Y.
,
Wu
,
J.
,
Liao
,
W.
,
Shui
,
W.
&
Qi
,
H.
2016
Evaluating the trends of China's ecological civilization construction using a novel indicator system
.
Journal of Cleaner Production
133
,
910
923
.
Zuo
,
Q. T.
,
Chen
,
H.
,
Dou
,
M.
,
Zhang
,
Y. Y.
&
Li
,
D. F.
2015
Experimental analysis of the impact of sluice regulation on water quality in the highly polluted Huai River Basin, China
.
Environmental Monitoring and Assessment
187
(
7
),
45001
45015
.
Zuo
,
Q. T.
,
Luo
,
Z. L.
,
Shi
,
Y. Q.
,
Gan
,
R.
,
Liu
,
J.
&
Chen
,
H.
2016b
Main parameters and physiographic characteristics of Shaying River Basin
.
Water Resources and Hydropower Engineering
47
(
12
),
66
72
(in Chinese)
.