Quick Search: Article NO. Chinese Title English Title Pin Yin Name Real Name Institution(In Chinese) Institution(In English) Chinese KeyWords English KeyWords Abstract(In Chinese) Abstract(In English) Fund Project(In Chinese) Advanced Search
 寒旱区科学  2018, Vol. 10 Issue (3): 261-270  DOI: 10.3724/SP.J.1226.2018.00261 0

Citation

Tong HL, Shi PJ, Zhang XB, et al. 2018. Characterization of landscape pattern based on land economic niche change: A case study in Ganzhou, Gansu Province, China. Sciences in Cold and Arid Regions, 10(3): 261-270. DOI: 10.3724/SP.J.1226.2018.00261.
[复制英文]

Correspondence to

PeiJi Shi, College of Geography and Environmental Science, Northwest Normal University, 967 Anning East Road, Lanzhou, Gansu 730070, China. E-mail: xbsdspj@163.com

Article History

Accepted: April 8, 2018
Characterization of landscape pattern based on land economic niche change: A case study in Ganzhou, Gansu Province, China
HuaLi Tong , PeiJi Shi , XueBin Zhang , ZaiYan Li
College of Geography and Environmental Science, Northwest Normal University, Lanzhou, Gansu 730070, China
Abstract: Land use change has a profound impact on biodiversity and ecological processes, and is closely related to changes in landscape patterns. This paper introduces the theory and method of land economic niche into landscape ecology, which provides a new method for spatial characterization of urban and rural spatial landscape patterns. Based on this theory, this paper analyzes the landscape pattern of Ganzhou District by using Landsat images as data source in 1995, 2000, 2005, 2010 and 2015. We calculated the land economic niche by applying the niche potential theory. Combined with the theory of landscape ecology, we explored the effects of the land economic niche change on the landscape pattern at a county scale. The results show that economic niche of construction land, watershed and farmland increased during 1995–2015, and grassland declined significantly. The economic niche of farmland, construction land, watershed and grassland show a negative correlation with the number of patches (NP), fragmentation index (FN) and the fractal dimension index (FD), and had a positive correlation with the aggregation index (AI). There was no significant correlation between the forest land economic niche and landscape metrics. The change of land economic niche has a driving effect on the landscape pattern of the county, which can represent the economic development direction of Ganzhou District. The land economic niche is closely related to the landscape type which can directly obtain an economic benefit.
Key words: land economic niche    landscape metrics    landscape pattern    characterization

1 Introduction

Landscape ecology is an interdisciplinary between geography and ecology, which was developed in the 1980s. As the core research contents, spatial statistical characteristics comparison (Li, 2000; Qi et al., 2013 ), landscape metrics (Wu, 2000; Wu et al., 2002 ) and Markov transfer matrix (Feng et al., 2010 ) have become the main methods and indicators for landscape spatial analysis. Landscape metrics are widely used in landscape pattern analysis and dynamic research (Fan and Ding, 2016; Robinson and Weckworth, 2016), which contain highly concentrated landscape information, and simply reflect the landscape structure and spatial allocation (Zhang et al., 2003 ). With the development of GIS and the upgrading of Fragstats (Chen et al., 2002 ), research on landscape metrics continues to develop, such as patch shape index, landscape diversity, fractal dimension, fragmentation, grain degree and other quantitative research landscape indices (Pennock et al., 1994 ; Wu et al., 2002 ; Fu et al., 2011 ). In terms of landscape representation, landscape metrics can reveal landscape patterns such as the evolution and alteration of landscapes (Verburg et al., 1999 ; Freitas et al., 2013 ), but does not reveal the driving and influencing factors (Li et al., 2004 ). Therefore, landscape metrics has become a key issue in understanding the driving and influencing factors that cause landscape changes at a regional scale (Chaplot et al., 2005 ; Diggelen et al., 2005 ; Bajocco et al., 2016 ).

A niche refers to the position and function of a species or population in the ecological environment, that is, the position and interdependence temporally and spatially (Zhang and Xie, 1997; Silvertown, 2004). With the development of niche theory, some scholars have proposed natural, social and economic complex ecosystems to apply niche theory to a wider range of fields (Basille et al., 2008 ; Nascimento et al., 2010 ; Cardoso et al., 2017 ). In recent years, the theory has a close connection with the study of land resources (Hao et al., 2010 ; Ceccarelli et al., 2014 ; Hu et al., 2016 ). Some scholars have constructed the land niche on the basis of the ecological niche theory and model (Zhang et al., 2002 ; Dan and Seifert, 2011). The land niche can reflect space, function and position of each land type in the land use system, and quantitatively express the ability of each land type to occupy or utilize a new habitat. At the spatial level, the land niche is a measurement of the interaction between different geographical units (Qi et al., 2007 ; Wu and Li, 2014; Hu et al., 2016 ). The larger the land niche, the stronger the ability of the regional spatial unit to attract population and material circulation.

The change of land use type driven by land niche has a direct or indirect effect on regional material circulation and energy flow, which is one of the important factors of regional land landscape pattern change (Yu et al., 2015 ). Therefore, it is of great significance to characterize the regional spatial landscape pattern using the land niche theory. Based on the ecological significance of landscape pattern on a county scale, this paper used Ganzhou District as the study area. We selected six landscape indices that characterize landscape quantity, structure and morphological characteristics and human disturbance, which clarifies the trend of land use change and explores the characterization of landscape pattern based on land economic niche.

2 Study area and material 2.1 Study area

Ganzhou District is a county level district, administered by Zhangye, a prefecture-level city, and Ganzhou is the city proper of Zhangye. This county is located in the middle of the Hexi Corridor, between 100°04'E–100°52'E, 38°32'N–39°24'N. It is adjacent to Shandan and Minle counties to the east, Sunan Yugur Autonomous County to the south, Linze County to the west, and Inner Mongolia Autonomous Region to the north. The county administers a national economic and technological development zone, 18 townships, 5 street offices and 245 administrative villages. Heihe River, which traverses the county, is the second largest inland river in China, and forms a well-known oasis known as "Sai Shang Jiang Nan", meaning Lush southern-type fields on the frontier. Ganzhou is 65 km long from east to west and 98 km from north to south. The total area is 3,657 km2.

The climate in Ganzhou District is typical continental arid. There is scarce precipitation and strong evaporation and large daily temperature difference. The north side are Heli Mountain and Longshou Mountain, and south sides is Qilian Mountain, and the terrain tilts from southwest to northeast. Because of soil and geographical differences, Ganzhou District contains different types of landscape, including mountain forest, plain forests, desert shrub grassland, meadow, marsh and wetlands. The Sai Shang Jiang Nan oasis is formed from five major rivers, the Heihe, Suyoukou, Dayekou, Daciyao and Shandan. Annual runoff is 2.46 billion m3.

Ganzhou District is a commodity grain base in Hexi Corridor and a tourist city along the Silk Road (Tong et al., 2015 ; Tong et al., 2018 ). It has a total population of over 513.6 thousand (2015), 48% of whom are urban dwellers. The total GDP of Zhangye City reached 15.675 billion Yuan (approximately 2.58 billion US Dollars) in 2015. Ganzhou District has a balanced industrial structure, with a ratio of primary: secondary: tertiary industry of 22.7:22.2:53.1 in 2015.

 Figure 1 Location of study area
2.2 Data and methodology 2.2.1 Data process

Data for this study include satellite remote sensing images of Landsat TM and ETM, and Ganzhou District Statistical Yearbook for 1995, 2000, 2005, 2010 and 2015 (Ganzhou Statistics Bureau, 2005–2015). Remote sensing data was obtained from the USGS website (http://glovis.usgs.gov), the images were mainly obtained in August and September. Vegetation characteristics are more robust during this period, which is conducive to classification and interpretation. Auxiliary data is from the second national land survey data of Ganzhou District. Based on the landscape classification system, the remote sensing image (30 m) of the study area was supervised by Envi5.0 software. Land use classification was obtained by geometric correction, cloud removal and image enhancement. On the basis of this classification, the land use map, Google Earth high resolution images, and field investigation results of different periods were used to correct interpretation results, and finally the land use data of Ganzhou District were obtained. Through the verification of these results, overall accuracy reached more than 90%, which meet the accuracy requirements of this study. Secondly, interpretation of the land use map was converted into a grid format in Arcmap10.2, using the Fragstasts 4.2 software. Thus, the landscape metrics can be calculated.

According to the classification of land use status (GB/T 21010-2007) and the classification method of land use/cover classification system of remote sensing in China, combined with the landscape characteristics of the study area, Ganzhou District is divided into six landscape types: farm land, forest land, grassland, construction land, watershed and bare land. The watershed includes rivers, lakes and wetlands. The construction land includes urban, rural residential and mining.

2.2.2 Methodology

(1) Land economic niche

Niche is the relative position and function of the biological unit in the process of interacting with the environment in a particular ecosystem (Zhu, 1997). Mainly, it includes two aspects: first, the state of biological units (biomass, the number of individuals and biological units, energy, adaptability, resource share, level of economic development), which is the result of past growth and development, socio-economic development and the interaction with the environment, that is, the niche state. Second is the influence or dominance of biological units, such as the rate of energy, growth, economic growth and material transformation, the ability to occupy a new habitat, and productivity, that is, niche potential. The land ecology niche is a natural, social and economic multidimensional concept, but under the market economy condition, its main function is the economic niche (Qin and Gao, 2004; Wu and Li, 2014). The land economic niche is mainly composed of the comparative economic benefits of land use, and the status, effect and function of the land use ecological element in the process of regional economic sustainable development (Chen et al., 2010 ). The formula for the land economic niche is as follows (Xiao et al., 1997 ):

 ${N_i} = \left( {{S_i} + {A_i}{P_i}} \right)/\sum\nolimits_{j = 1}^n {\left( {{S_i} + {A_i}{P_i}} \right)}{S_i} = {M_i}{X_i}/{R_i}{P_i} = {Y_i}/{R_i}$ (1)

where, Ni is the land economic niche of different land use types; Si is the state of different land use types and Pi is the potential of different land use types; Ai is the dimension conversion coefficient; MiXi is the product value of land area and the unit area income, which is total revenue or the GDP of the ith type land use, Yi is the annual growth rate of the economic income or GDP of the ith type land; Ri is the population attached to the ith type land.

The economic benefits of the land type used in this paper are obtained from the statistical yearbook. The farmland output value is the benefit of agricultural in the year. The forest production value is the forestry benefit and the grassland output value mean the grassland benefit. The watershed output value is the benefit of the aquatic products. The construction land output value is the total benefit of second and tertiary industries.

(2) Landscape metrics

Landscape metrics is the most widely used method for landscape spatial pattern analysis. Based on previous studies, this paper selected the more commonly used and meaningful indicators (Qi et al., 2013 ), including total area of patches (CA), number of patches (NA), the mean patch area (AREA_MN), the aggregation index (AI), the fragmentation index (FN) and the fractal dimension index (FD). The calculation of landscape metrics is calculated by Fragstats 4.2 software. FN is calculated by the formula based on the data calculation of Fragstats 4.2 software. The aforementioned index can represent the characteristics and change of spatial pattern in Ganzhou District from four aspects.

1) Quantity characteristics. CA determines landscape extent and the maximum scale of research and analysis, which is the basis for calculating other indicators. NP is the number of land use type patches that can be used to describe the heterogeneity of the entire landscape. AREA_MN is the average area of a patch type, which represents an average condition. The value change can reflect the rich landscape ecological information, which is the key to reflect landscape heterogeneity.

2) Morphological characteristics. FD is a measurement of complex irregularity, which is between 1 and 2, and can quantitatively describe core area size of a certain type and the meandering/sinuosity of its boundary line. If FD approaches 1, patch geometry tends to be more regular; conversely, patch geometry tends to be irregular. This is mainly to reflect the validity of the complex form of space, which also reflects the impact of human activity on the landscape pattern (Krummel et al., 1987 ).

3) Structure. AI indicates the adjacent relationship between similar types of patch pixels from the landscape level. When there is no common boundary between all the pixels in a certain type, the aggregation index is the lowest. Conversely, the larger the common boundary, the greater the aggregation index (maximum value is 100).

4) Human disturbance. FN is used to describe landscape fragmentation. This refers to the process of a plaque mosaic from single, homogeneous and continuous to complex, heterogeneous and discontinuous individual under natural or human disturbance. The formula is as follows (Li, 1989):

 ${\rm FN}={\rm MPS}(n-1)/A$ (2)

where, MPS is the mean patch area of all kinds of patches in the landscape, n is the total number of patches of the landscape type, and A is the landscape area.

3 Results and analysis 3.1 The change of land economic niche in Ganzhou

The land economic niche reflects the function of land cover in Ganzhou from the economic output value. The landscape pattern can express the spatial distribution of land cover, and the landscape metrics reflects the spatial structural of land cover. The land economic niche combined with landscape pattern and metrics indicate the ecological possession in function, space and structure of land cover. According to the Formula (1), the land economic niche of Ganzhou District from 1995 to 2015 is calculated (Table 1).

Table 1 Land economic ecological niche of landscape from 1995 to 2015 in Ganzhou District

As presented in Table 1, during 1995–2015, niches of farmland, watershed and construction land in Ganzhou show a rising trend. The economic niche of cultivated land rose from 0.11 to 0.14, with an average annual growth rate of 1.2%. The economic niche of watershed increased from 0.07 in 1995 to 0.14 in 2015, with an average annual increase of 3.5%. In these five types of land cover, the annual growth rate of watershed economic niche is the highest. The study area is located in the northwestern arid region whose ecology is very fragile, with most of the output value of water lies in its indirect creation of economic value and protection of the regional ecological elements to guarantee normal operation and survival. Construction land niche rose from 0.34 in 1995 to 0.65 in 2015, with an average annual growth of 3.2%. The economic growth of forest land first dropped then rose to a peak of 0.06 in 2005, then decline again from 2005 to 2015. The overall trend of forest was declining, with an annual decline rate of 4.5%. The decline of grassland was the largest, from 0.43 in 1995 to 0.05 in 2015, with an annual decrease of 10.2%. This indicates that the grassland economic niche in the study area accounted for the most important position before 2000. With rapid urbanization, construction land gradually replaced grassland. The construction economic niche has become the largest since 2000. The construction land output value has the greatest contribution to the economic development of Ganzhou District. This is because China's rapid economic development led to rapid urbanization over the past 20 years.

The general change of land economy ecological niche is as follows. The economic niche of construction land and watershed are basically increasing while grassland is decreasing. The economic niche of farmland shows a slight upward trend while grassland shows a dramatic decreasing trend. Combined with the trend of local land use change, the regional production value and urbanization level, the output value of the first industry in Ganzhou District show a downward trend, the second and third industries developed rapidly, the output value became high, resulting in change in the development model. Second and tertiary industries became the main source of income in Ganzhou District.

3.2 Spatial analysis of landscape pattern change in Ganzhou District

According to the local circumstances and the national "land use classification" (GB/T 21010-2007), the study area is divided into six categories which are farmland, forest, grassland, construction land, watershed, and bare land (Figure 2) through ENVI 5.0. Aiming at the acquired five phases of remote sensing image data, classification was carried out by a combination of visual interpretation and supervised classification. The total accuracy of the five phases of data after image classification is 90.6%, 94.2%, 93.8%, 94.6% and 95.4%, respectively. The Kappa coefficients are 0.8709, 0.9326, 0.9209, 0.9338 and 0.9423, respectively, which meet the requirements of research and provides reliable data guarantee for later comprehensive analysis.

As presented in Figure 2, during the study period, there was an obvious increase of farmland and construction land, while grassland and bare land continuously decreased. The degradation of grassland and conversion to farmland are obvious, which is the embodiment of large-scale agriculture and the result of increased human activity.

 Figure 2 The spatial pattern of land use in Ganzhou District from 1995–2015
3.3 The change of landscape metrics 3.3.1 The change of landscape type

As presented in Table 2, farmland is the main land use type in the study area, followed by grassland, watershed and construction land. Farmland, construction land and watershed areas increased continuously from 1995 to 2015. Among which, the CA of farmland increased from 895.625 km2 to 1,120.500 km2, with an increase of 224.875 km2. NP of farmland decreased, AREA_MN and AI increased, and FN became smaller. CA of construction land increased from 99.94 km2 to 146 km2, with an increase of 46.46 km2. NP of construction land decreased, AREA_MN and AI increased, and FN became smaller. CA of forest land changed slightly, with a reduction of 2.81 km2. CA of the watershed decreased from 1995 to 2005, but increased between 2005 and 2015. This increase is due to local government's implementation of the wetland protection policy to restore wetlands and the building of wetland parks and three artificial lakes in urban areas after 2006. Overall, the water area shows a decreasing trend generally from 135.63 km2 to 123.38 km2, with a decrease of 12.25 km2. The main source of landscape change in Ganzhou District is grassland and bare land. From 1995 to 2015, the grassland decreased from 794.94 km2 to 680.19 km2, a reduction of 114.75 km2. The bare land decreased from 1,647.5 km2 to 1,506.1 km2, a reduction of 141.4 km2. NP of grassland and bare land shows an increasing trend while AREA_MN shows a decreasing trend.

Table 2 Landscape indices of different land use types of Ganzhou District in 1995, 2000, 2005, 2010 and 2015
3.3.2 The change of landscape morphology

The FD absolute value change during 1995–2015 (Table 2) for the watershed is the highest, followed by grassland, farmland, forest and construction land. According to the landscape ecology theory, the complex patch morphology indicates a high degree of naturalization and low self-similarity. The FD of grassland and construction land are relatively low, indicating that the patch of grassland and construction land have high self-similarity, and the patch morphology is regular, which is strongly influenced by human activity. Due to the dynamic change of the landscape's FD, the FD of farmland and construction land decreased from 1.562 to 1.544 and from 1.531 to 1.511, respectively. This decrease demonstrates how human activity consolidated the scope and shape of farmland, and construction land was programmatically expanded during the process of urbanization. The FD of the watershed increased from 1.702 to 1.719 due to local governmental implementation of the wetland restoration and protection policy and the Heihe River Basin program. The FD of forest and grassland shows an increasing trend. The FD of forest and grassland increased from 1.491 to 1.549 and from 1.571 to 1.584, respectively. This increase is due to overgrazing and land degradation, with portions of forest and grassland converted to farmland and construction land.

3.3.3 The change of landscape structure

Landscape structure is the spatial arrangement and composition of landscape components and elements. In landscape ecology, AI indicates the adjacency between pixels of the same type patch at a landscape level. As presented in Table 2, AI of farmland patch is the highest except for bare land, followed by grassland and forest, while watershed and construction land is relatively low. During 1995–2015, the AI of farmland, watershed and construction land show an increasing trend while forest, grassland and bare land show a decreasing trend. The AI of farmland increased from 79.75 to 82.68 (4% increase), watershed increased from 34.22 to 42.42 (23.94% increase), and construction land increased from 25.95 to 36.55 (41% increase). The AI of forest decreased from 60.55 to 55.89 (7.7% decrease), and grassland decreased from 64.59 to 57.99 (10% decrease). Combined with Figure 2, we can see that the AI of farmland, watershed and construction land increased because their total area increased. The reduction of forest and grassland AI are closely related to the reduction of the corresponding total area. Portions of forest land and grassland area were converted into cultivated land and construction land. Also, overgrazing and degradation are the main reasons for the reduction of forest and grassland areas.

3.3.4 The impact of human activity

From the absolute value of landscape FN for Ganzhou District during 1995–2015, construction land is the largest; watershed is second, followed by forest, grassland, farmland and bare land. In landscape ecology, landscape structure of construction land is simple, the degree of homogenization is high, and has the highest land economic niche. Construction land is the most affected by human activity and has the highest FN in the study area. The study area is in an arid region, located in a fragile ecological zone in Northwest China. The total amount of water resources is unstable, and coupled with human interference, wetland degradation is serious, and the FN of watershed is high. The few natural forest is mostly located near water sources. In recent years, sheltered forest area has increased slightly due to China's national policy, but the distribution is heterogeneous and dependent on human activity. The spatial landscape of grassland is complex, heterogeneity is high, and land economic ecology is less than construction land and farmland. The FN of farmland and bare land are relatively low. Farmland is affected by terrain, water, soil quality and human activity. In order to facilitate cultivation, farmland in the study area is mostly large and continuous. The FN of bare land is low because most of the landscape in Ganzhou District is bare rock and desert, where conversion to other land types is difficult, and human activity has little influence on bare land. However, in recent years, numerous improved land and large-scale renovation projects have been started in Ganzhou, such as "Wa Qu Pai Jian" which means "Digging canals and removing alkali Project", and "Farming Plan in Gebi Desert". In order to improve the regional ecological environment, a new urban plan will be implemented, expanding west into barren lands.

3.4 The relationship between land economic niche and landscape metrics

In order to study the relationship between the land economic niche and corresponding landscape metrics, the Pearson correlation coefficient was obtained by using SPSS (Table 3). R1j is the correlation coefficient between the farmland economic niche and landscape metrics. R2j is the correlation coefficient between the forest economic niche and landscape metrics. R3j is the correlation coefficient between the grassland economic niche and landscape metrics. R4j is correlation coefficient between the watershed economic niche and landscape metrics. R5j is the correlation coefficient between the construction land economic niche and landscape metrics.

Table 3 Correlation coefficients between land economic ecological niche and landscape indices

As presented in Table 3, CA, NP, AREA_MN, AI, FN and FD of farmland and grassland are all correlated with the land economic niche. The watershed and forest have no significant correlation between the corresponding landscape metrics and the land economic niche. It can be found that NP and FN are negatively correlated with the land economic niche, and AI is positively correlated with the land economic niche. However, the NP of construction land is positively correlated with the land economic niche. Because the study area belongs to an oasis area, and some urban area is also an important ecological protection area.

4 Discussion

The theory of land economic niche combines natural and human factors, focusing on land use economic benefits, which describe the contribution of different land use to the local economy. Also, regional economic development is an important driving force for the change of landscape pattern. Results of the correlation between the land economic niche and corresponding landscape metrics show that partial landscape metrics has a significant correlation with the land with large direct economic benefits, such as farmland, grassland and construction land. The land economic niche is negatively correlated with NP, FN and FD, and positively correlated with AI. Correlation analysis of the land economic niche and corresponding landscape metrics shows that the land economic niche is only relevant to NP, FN, FD and AI of farmland, grassland and construction land, which are affected by change of the land economic niche. The economic niche of forest and watershed has no significant correlation with the corresponding landscape metrics. On the one hand, watershed and forest in the arid area does not increase more direct economic benefits. Furthermore, watershed and forest are determined by natural properties. Water is limited by weather and human factors, and the growth period of trees is long. In general, the change of land economic niche affects the change in regional landscape patterns.

In agriculture, farmland area increased year by year, NP, FN and FD decreased with the increase of farmland economic niche, and AI increased, which is consistent with the correlation analysis. In the second and third industry, the construction land economic niche increased year by year, CA, NP and AI increased, FN and FD decreased with niche increase, which is consistent with the correlation analysis except for CA. In animal husbandry, the grassland economic niche show a decreasing trend, CA decreased, NP show an increasing trend, and AI show a decreasing trend which is consistent with the conclusions of the correlation analysis. Farmland in the study area has been developed over a large area, and construction land is planned to expand from the original oasis area. Some of the grassland has converted to cultivated land and construction land. Although there is a small amount of farmland and bare land converted to grassland, the grassland area is still shrinking. Thus, there is a serious ecological security threat of the Ganzhou oasis area. The watershed area has decreased, NP, FN and FD increased with the land economic niche decrease, which is consistent with the correlation analysis. Since Ganzhou District is in an arid ecological fragile area in northwest China, the wetland and Heihe River play an important role in maintaining regional ecological balance as the largest water body in the study area.

In this study, although the economic niche of forest and watershed has no significant correlation with the corresponding landscape metrics, the reduction of forest and watershed area will inevitably lead to a decrease of regional water conservation function. Also, the increase and intensive use of farmland will aggravate agricultural water use. The shortage of water resources leads to the inability of anthropogenic development of bare land in the study area, and limited human activity leads to limited economic development. Even cause a series of problems such as ecological deterioration. The local government has made some effort to solve these problems where the watershed area has increased since 2005 and watershed economic niche also increased. Other measures need to be implemented such as converting low-quality farmland to grassland and forest which can resist salt intrusion and drought. Based on the regional farmland red line barren land should be converted into grassland and forest. At the same time, high-quality farmland is strictly restricted for other purposes. For bare lands around the oasis, we can improve the soil and plan grass and forest.

Comparing the economic niche of grassland and construction land in 2010 and 2015, grassland was reduced by 58.3%, and construction land increased by 14%, which indicate the economic benefits of rapid development of second and tertiary industries accounted for a larger proportion. Also, production output of second and tertiary industries will have a certain impact on the agricultural economy. Under the current situation, how to adjust the regional industrial structure is still an important issue.

5 Conclusions

This paper introduced the theory and method of land economic niche into landscape ecology, which provide a new method for spatial characterization of urban and rural spatial landscape patterns. Our case study contributes to the literature and practice of characterization of landscape pattern by calculating land economic niche. Change of land economy niche can mark the direction of economic development of Ganzhou District. The general characteristic of the land economic niche in Ganzhou District is as follows, economic niche of construction land, watershed and farmland increased during 1995–2015, while grassland declined significantly. Combined with local land use change trend, regional GDP and urbanization level, the output value of the primary industry in Ganzhou District shows a decreasing trend. Also, second and third industries developed rapidly and the output value was high, which resulted in the development model change. The area of local construction land continued to grow, and the second and third industries became the major economic income in Ganzhou District.

The change of land economic niche has a driving effect on the change of landscape pattern on a county scale. Economic development and the acceleration of urbanization process results in a change of local landscape type under the influence of human activity. Economic development led to changes in land cover. During 1995–2015, land cover changed greatly in Ganzhou District, where the conversion of grassland and bare land to farmland was obvious. The economic niche of farmland, construction land, watershed and grassland show a negative correlation with NP, FN and FD, and had a positive correlation with AI. There is no significant correlation between the forest land economic niche and landscape metrics. Also, the landscape pattern of forest land changed due to the conversion of cultivated land, construction land, grassland and land degradation. The landscape metrics also changed accordingly.

The land economic niche can characterize the direction of regional human activity. The pattern of landscape changed greatly from 1995 to 2015 in Ganzhou District, and landscape heterogeneity show an increasing trend. Also, land use is developing in the direction of diversification and homogenization. In particular, the landscape structure in Ganzhou District in 2015 was complex. The landscape pattern in the study area is becoming more and more complicated, and the human disturbance has increased. The increasing areas of farmland and construction land are due to human activity. The farmland is developing in a large continuous patch while construction land is planned to expand from the original oasis area programmatically. The watershed economic niche increased year by year while the forest economic niche is small and has decreased. At the same time, the economic niche and area of the grassland decreased, and the degree of fragmentation is high.

Acknowledgments:

This work is funded by National Natural Science Foundation of China (Grant No. 41771130 and Grant No. 41661035). We highly appreciate the editor and the two reviewers for their valuable comments, suggestions and editing, which have greatly helped to improve this manuscript.

References
 Bajocco S, Ceccarelli T, Smiraglia D, et al.. 2016. Modeling the ecological niche of long-term land use changes: the role of biophysical factors. Ecological Indicators, 60: 231-236. DOI:10.1016/j.ecolind.2015.06.034 Basille M, Calenge C, Marboutin E, et al.. 2008. Assessing habitat selection using multivariate statistics: some refinements of the ecological-niche factor analysis. Ecological Modelling, 211(1–2): 233-240. DOI:10.1016/j.ecolmodel.2007.09.006 Ceccarelli T, Bajocco S, Perini L, et al.. 2014. Urbanization and land take of high quality agricultural soils—exploring long-term land use changes and land capability in northern Italy. International Journal of Environmental Research, 8(1): 181-192. Chaplot V, Giboire G, Marchand P, et al.. 2005. Dynamic modelling for linear erosion initiation and development under climate and land-use changes in northern Laos. Catena, 63(2–3): 318-328. DOI:10.1016/j.catena.2005.06.008 Chen W, Xiao D, Li X. 2002. Classification, application, and creation of landscape indices. The journal of applied ecology, 13(1): 121-125. Chen Y, Zhang RZ, Chen B, et al.. 2010. Construction and application on the niche theory of land use. Arid Land Geography, 33(5): 791-801. DOI:10.13826/j.cnki.cn65-1103/x.2010.05.022 Dan LW, Seifert SN. 2011. Ecological niche modeling in Maxent: the importance of model complexity and the performance of model selection criteria. Ecological Applications: A Publication of the Ecological Society of America, 21(2): 335-342. Diggelen RV, Sijtsma FJ, Strijker D, et al.. 2005. Relating land-use intensity and biodiversity at the regional scale. Basic & Applied Ecology, 6(2): 145-159. DOI:10.1016/j.baae.2005.01.004 Fan Q, Ding S. 2016. Landscape pattern changes at a county scale: a case study in Fengqiu, Henan Province, China from 1990 to 2013. Catena, 137: 152-160. DOI:10.1016/j.catena.2015.09.012 Feng Y, Luo G, Zhou D, et al.. 2010. Effects of land use change on landscape pattern of a typical arid watershed in the recent 50 years: A case study on Manas River Watershed in Xinjiang. Acta Ecologica Sinica, 30(16): 4295-4305. Freitas MWDD, Santos JRD, Alves DS. 2013. Land-use and land-cover change processes in the Upper Uruguay Basin: linking environmental and socioeconomic variables. Landscape Ecology, 28(2): 311-327. DOI:10.1007/s10980-012-9838-9 Fu B, Liang D, Lu N. 2011. Landscape ecology: Coupling of pattern, process, and scale. Chinese Geographical Science, 21: 385-391. DOI:10.1007/s11769-011-0480-2 Ganzhou Statistic Bureau, 2015. Ganzhou District Statistical Yearbook. Hao SL, Cao LH, Li BC. 2010. Application of niche theory to land utilization: A case study on Shanghuang in Guyuan. Soil and Water Conservation in China, 3(3): 35-37. Hu M, Zhang J, Liang E, et al.. 2016. Analysis of land utilization ecological niche change in Manas River Watershed over the past 37 years. Journal of Shihezi University, 34(5): 571-575. Krummel JR, Gardner RH, Sugihara G, et al.. 1987. Landscape patterns in a disturbed environment. Oikos, 48(3): 321-324. Li BL. 2000. Fractal geometry applications in description and analysis of patch patterns and patch dynamics. Ecological Modelling, 132(1–2): 33-50. DOI:10.1016/S0304-3800(00)00303-3 Li H, 1989. Spatial Temporal Pattern Analysis of Managed Forest Landscapes. Ph.D. Thesis. Oregon: The Oregon State University. Li X, Bu R, Chang Y, et al.. 2004. The response of landscape metrics against pattern scenarios. Acta Ecologica Sinica, 24: 123-134. Cardoso MB, Ladio AH, Lozada M. 2017. Niche breadth and redundancy: Useful indices to analyse fuelwood use in rural communities. Journal of Arid Environments, 145: 52-59. DOI:10.1016/j.jaridenv.2017.05.007 Nascimento APBD, Ferreira ML, Molina SMG. 2010. Ecological niche theory: non-traditional urban and rural human populations. Journal of Human Ecology, 32(3): 175-182. Pennock DJ, Anderson DW, Jong ED. 1994. Landscape-scale changes in indicators of soil quality due to cultivation in Saskatchewan, Canada. Geoderma, 64(1–2): 1-19. DOI:10.1016/0016-7061(94)90086-8 Qi X, Mao J, Cheng Y. 2007. Study on optimizing strategies of city spaces basing on niche theory: A case in Zengcheng City. Planners, 23(2): 89-91. Qi Y, Wu JG, Li JL, et al.. 2013. Landscape dynamics of medium- and small-sized cities in eastern and western China: a comparative study of pattern and driving forces. Acta Ecologica Sinica, 33(1): 275-285. Qin JC, Gao M. 2004. Economic niche of land use and sustainable development in mountains: A case study of Chongqing. Journal of Mountain Research, 21(6): 702-706. Robinson H, Weckworth B, 2016. Landscape ecology: linking landscape metrics to ecological processes. Snow Leopards. Elsevier Inc. pp. 395–405. Silvertown J. 2004. The ghost of competition past in the phylogeny of island endemic plants. Journal of Ecology, 92(1): 168-173. DOI:10.1111/j.1365-2745.2004.00853 Tong HL, Shi PJ, Zhu GF, et al.. 2015. The characteristics of oasis urban expansion and drive mechanism analysis: a case study on Ganzhou District in Hexi Corridor, China. Sciences in Cold and Arid Regions, 7(3): 282-292. DOI:10.3724/SP.J.1226.2015.00282 Tong H, Shi P, Bao S, et al.. 2018. Optimization of urban land development spatial allocation based on ecology-economy comparative advantage perspective. Journal of Urban Planning and Development, 144(2): 05018006, 1-14. DOI:10.1061/(ASCE)UP.1943-5444.0000444 Verburg PH, Koning GHJD, Kok K, et al.. 1999. A spatial explicit allocation procedure for modelling the pattern of land use change based upon actual land use. Ecological Modelling, 116(1): 45-61. DOI:10.1016/S0304-3800(98)00156-2 Wu JG, 2000. Landscape Ecology: Pattern, Process, Scale and Hierarchy. Beijing: Higher Education Press. Wu J, Shen W, Sun W, et al.. 2002. Empirical patterns of the effects of changing scale on landscape metrics. Landscape Ecology, 17(8): 761-782. DOI:10.1023/A:1022995922992 Wu Q, Li Y. 2014. The representation of landscape pattern to land economic niche change: A case study of Huizhou City in Guangdong Province. Scientia Geographica Sinica, 34(6): 705-710. DOI:10.13249/j.cnki.sgs.2014.06.705 Xiao DN, Bu RC, Li XZ. 1997. Spatial ecology and landscape heterogeneity. Acta Ecologica Sinica, 17(5): 453-461. Yu HY, Zhang F, Wang J, et al.. 2015. Application of land economic ecological niche in landscape pattern analysis at county level: A case study of Jinghe County in Xinjiang, China. Chinese Journal of Applied Ecology, 26(12): 3849-3857. Zhang GM, Xie SC. 1997. Development of niche concept and its perspectives: A review. Chinese Journal of Ecology, 16(6): 46-51. Zhang QJ, Fu BJ, Chen LD. 2003. Several problems about landscape pattern change research. Scientia Geographica Sinica, 23(3): 264-270. Zhang X, Ge XD, Pu LJ, et al.. 2002. A study on the economic niche of land use theory and mechanisms of cropland conservation. Journal of Natural Resources, 17(6): 677-683. DOI:10.11849/zrzyxb.2002.06.004 Zhu CQ. 1997. The niche ecostate-ecorole theory and expansion hypothesis. Acta Ecologica Sinica, 17(3): 324-332.