Yini Han?Jianzhi Niu?Feizhou Wu

Land-use changes in the small watershed of the Loess Plateau, hilly-gully region,China

Yini Han?Jianzhi Niu?Feizhou Wu

As more and more farmland is converted to forestry,the need for effective decision support regarding the use ofland in the fragile ecologicalenvironmentof the Loess Plateau hilly-gully area.The Luoyugou watershed was chosen as the study area to calculate the single dynamic degree,integrated dynamic degree,and change indexes of land use,as wellas the land-use type transition matrix.This was done by interpreting the TM and SPOT images of the Luoyugou watershed in 1986,1995,and 2004 and making statistical analysis.The results of our statistical analysis show that the conversion of slope farm land to terrace and forest land plays a dominant role in land-use changes in the Luoyugou watershed from 1986 to 2004.The land-use changes are mainly driven by population growth,socio-economic development,consumer spending,and investmentin forestecology.

Loess Plateau hilly-gully regionLuoyugou watershedLand use and cover changesDriving force

Introduction

Land-use/land-coverchange(LUCC)is the study ofglobal land-use change and the core project of the International Geosphere Biosphere Programme(IGBP).In recent years,understanding of LUCC has developed significantly in the international academic community(Mayer and Turner 1996;Luo and Ni2000;Shietal.2002;Lambin and Geist 2001).Land use is a long-term management activity that can reflect the relationship of the area and ratio of the nationaleconomy,the rationality of land use and the traits of production structure(Ren etal.2003).

Both soilerosion and sedimentexportto rivers and lakes are considered to be important environmental problems. The driving factors of land-use change have been the focus of research scholars at home and abroad(Rindfuss et al. 2004;Liao et al.2011;Yang et al.2011;Xie et al.2011; Wu etal.2012;Chen and Xia 2012).With the development ofsocialproductivity,population growth,and expansion of human activities,conflicts between the economic demand and limited land resources are growing significantly,which severely limits the rapid development of social economy. Therefore,the study of land-use structural change is instructive for regional industrial distribution and rational use of land(Baietal.2004;Zhang 1999;Liu etal.2006). Climate,soil and hydrology are natural driving forces, while social driving forces include population change,the economic gap between the rich and the poor,technological development,economic enhancement,and political economic structure(Stern et al.1992;Turner et al.1994).

The scale of land-use/land-cover change(LUCC)study area is shifting from global concerns to specific regions (Wang and Bao 1999;Hao etal.2007).The application of remote sensing and geographic information systems(GIS) makes it possible to establish temporal and spatial databases of land use and land cover(Loveland et al.1991; Ren et al.2007;Yang et al.2013).This kind of data is conducive to in-depth analysis of resource-environment effects and the temporaland spatialvariation rule of LUCC (Chen and Verburg 2000;Holling 1992;Hall 1995).In thepresent study,the Luoyugou watershed was chosen as the study area.It is located in Tianshui City,Gansu Province, and belongs to the third sub-region of the Loess Plateau hilly-gully region.This watershed is characterized by heavy soilerosion,due to the lack of infrastructure to adapt to land-use changes.More scientific insight will help the people of this area to utilize land efficiently and provide evidence for better managementofland resources of entire hilly-gully region.It can also improve the ecologicaland environmental conditions,socio-economic development, living standards,and the region’s overall competitiveness.

Materials and methods

Study area

The Luoyugou watershed is located in northern suburb of Tianshui City.It is a tributary on the left bank of Jihe River,which in turn is a tributary of the Weihe River.The whole valley is under the jurisdiction of Tianshui City.The geographical position is between longitude 105300–105450and latitude 34340–34400.The watershed terrain slopes from northwest to southeast,and the main channel length is 21.8 km.The altitude ofthis watershed is between 1,165.1 and 1,895 m.

Luoyugou watershed belongs to the zone of temperate continental monsoon climate,with an annual average temperature of 10.7C and an annual average rainfall of 533.7 mm.Its annual average runoff modulus is 3.069 9 104/km2,and the annual average erosion modulus is 4,851.6t/km2.By the end of 2005,our survey showed thatLuoyugou watershed had of4,362 totalhouseholdsand the total population was 19,752,of which 10,228 people are working.The employment rate is 51.78%.The annual per-capita income is 1,205 Yuan.The main agricultural products in this area are winter wheat,maize,potato, rapeseed,and flax.Additional products include apples, peppers,and other fruitand vegetables.The main livestock are cattle,horses,donkeys,mules,pigs,sheep,and chickens(Li et al.2006).

Data processing

The projectteam relied on ArcGIS software.First,we took TMimages in 1986 and 1995,and in 2004,we took ETM? image data of the Luoyugou watershed as the basic data resource.Next,we combined the 1:10,000 topographic map to take a geometric correction by quadratic polynomials according to the standard division range.The errorin a control pointis less than one pixel.Second,we established the regional interpretation signal according to the available statistics,topographic maps,various thematic maps and field data.Third,we used human–computer interaction and Gaussian maximum likelihood classifi cation to interpret these three maps of land use types and evaluated its accuracy according to the field experimental data.Finally,we derived the dynamic information of land use and land cover changes by spatialoverlay analysis and generated relevantthematic maps.

Land-use quantitative change analysis method

To establish a more complete description of land-use changes in the Luoyugou watershed,this study looked at quantities and extents.In the first,the contribution rate of change,intensity index of change,single land-use dynamics degree,and integrated land-use dynamics degree were used as indicators.In the second,we introduced the comprehensive index,the volume,and the rate of land-use change.

The contribution rate of change refers to the percentage of land-use change area for certain types over the totalarea of land-use change in the same period;the formula is as follows(Cai 2001):

The intensity index of change indicates the change of a certain land use type ibetween period a and b in unitarea; the formula is as follows(Li et al.2011):

Of the above-mentioned two formulas,the Aimeans the contribution rate of the type i,LTI means the intensity index of type i;Uai,Ubimean the area of type i in the beginning and the end of the study period respectively;LAimeans the area of the whole watershed;T means internal time of b and a(year).

The single land-use dynamics reflect the situation of land-use area change with a certain type and in a certain time.The formula is as follows(Li et al.2003):

The integrated land-use dynamics degree can be used to describe the rate of regional land-use change.The formula is as follows:

where,K is the land-use change dynamics of one type,T is the study period,Ua,Ubmeans quantity of a certain land-use type in the beginning and the end of the study period, LUimeans the land area of type i in the beginning of study period,D LUi-jmeans the absolute land area change of type i(Li etal.2006;Huang 2004).

The land-use degree comprehensive index is a continuous function between 100 and 400,which reflects the degree of the land use.Its formula is as follows(Zhuang and Liu 1997):

where,Ljis the composite index ofland-use degree,Aithe land-use degree classification index of level i,Cithe area rate of the land-use degree classification index of the level i,and n is the land-use degree classification index.

The change of land-use degree within a particular range is a result of changes of various land-use types.The volume of land-use change and the rate of land-use change can revealthe change’s overallleveland its quantitative trends. The formula is as follows(Li et al.2005):

where,Laand Lbmean land-use comprehensive index of time a and b respectively,Aiis the land-use degree classification index of leveli,Ciband Ciamean the area rate of land use of the level i in b time point and a time point respectively,and R is the rate of land-use change.If D Lba[0 or R[1,the land use in this area is developing, otherwise,itis a phase of adjustmentor recession.

Results

Land-use changes analyses of Luoyugou watershed

The Luoyugou watershed is a typical agroforestry watershed.In 1986,the total area of farmland and terrace was 72.57%,while the total area of forest and grass was 19.30%as the Fig.1.In 1995,the total area of farmland and terrace decreased to 62.39%,while the total area of forestand grass increased to 29.37%as in Fig.2.In 2004, the area rate of farmland fell to 12.26%,forest area had grown to 22.30%,which is shown in Fig.3.

Fig.1 Land-use structure map of Luoyugou watershed in 1986

Fig.2 Land-use structure map of Luoyugou watershed in 1995

Fig.3 Land-use structure map of Luoyugou watershed in 2004

As a soil and water conservation practice,the conversion of sloping farmland to terraces(termed as terrace engineering)took place from the middle of the1980s.Terrace,woodland and farmland changed the most in Luoyugou watershed from 1986 to 1995(Tables 1 and 2). Terrace and forest land increased by 2,590.79 ha and 794.11 ha,respectively(Table 1),for a growth rate of 38.44 and 11.78%(Table 2)during the period.Farmland area decreased by 3,333.78 ha with a contribution rate of 49.46%during the period.As the intensity index shows, great land-use changes have taken place in the years from 1986 to 1995,especially in sloping farmland(LTI= -4.565%),terrace(LTI=3.457%),and forest (LTI=1.087%)(Table 2).While in 1995 to 2004,the land use changed slightly:sloping farmland reduced (LTI=-2.371%)while forest(LTI=0.638%)and terrace(LTI=1.752%)increased during 1986–2004 (Table 2).The change from sloping farmland to terrace and forest land was the dominant factor driving land-use change in the Luoyugou watershed from 1986 to 2004.

Table 2 Contribution rate and intensity index of land-use changes in Luoyugou watershed

The single land-use dynamic degree was calculated according to formula 3(Table 3).The land-use type with lower dynamic degree is stable in the study period,for its difficulty to be transferred to other land-use type(Zhang et al.2004).In general,land use in the Luoyugou watershed changed rapidly during 1986–1995:terrace was the most active(27.5%),followed by forest(10.75%),and sloping farmland(7.65%).While residential land and grassland changed slowly,shrub land,and bare land were almost invariable.The period from 1995 to 2004 had a slight change in land-use types relative to 1986–1995.

The calculated integrated land-use dynamic degree of 1986–1995 was 5.22%,higher than that of 1995–2004 (0.62%)and 1986–2004(2.69%).Itshows thatthe range of land-use change in the Luoyugou watershed from 1986 to 1995 was the maximum.

The degree of land-use change in the Luoyugou watershed was calculated according to formulas 5,6,and 7 (Table 4).The rate of land use change degree of 1986–1995,1995–2004 and 1986–2004 was-0.0356, -0.0237,and-0.0586,respectively.Allthree values were below 0,which indicated that land use in the Luoyugou watershed was still in the adjustment phase and would continue to change over time.

The driving forces of land use change of Luoyugou watershed

The raw data in this study consists of 16 socio-economic indicators and a variety ofland-use types(Bare land,Forest land,Shrub land,Grassland,Residential land,Sloping farmland,Terrace)areas of Luoyugou watershed in 1986,1995,and 2004.The detailed raw data are shown in Table 5.

Table 3 Single land-use dynamic degree of Luoyugou watershed(Unit:%)

Table 4 The degree of land-use change in the Luoyugou watershed

Table 5 Socio-economic indexes and land-use area of Luoyugou watershed in 1986,1995,and 2004

Sloping farmland,terrace,and forest land took up the largestarea ofthe watershed,a totalof82.49%.(The other four land-use types in the area were relatively small,while the change degree was very light.)We focused on those three categories to analyze to revealthe causes and internal mechanisms of land-use change in the Luoyugou watershed.

Driving force analysis of sloping farmland

SPSS statisticalanalysis software was used to calculate the correlation coefficient between sloping farmland(Y6)and various socio-economic indices coming from investigation of the Luoyugou watershed and the economic yearbook of Tianshui city,and got the stronger social-economic index by combining it with other relative research results.The chosen indices are X1 total population(people);X4 the proportion of employed population(%);X7 rural total expenditure per capita(Yuan);X8 rural total income per capita(Yuan);X9 the proportion of primary industry output value(%);X12 the proportion of forestry outputvalue(%); and X14 the proportion of fishery output value(%).(The method was also used in driving force analysis of terraces and forest land-use change.)

Principal component analysis was conducted on standardized data in SPSS.The principal components analysis eigen values and variance contribution rates,and the main components of the load are shown in Table 6. The analysis results of sloping farmland are given in Table 7.

Table 6 Eigen values and variance contribution rate

Driving population and economic development

From Table 7,we can see that the X1 total population (people);X4 the proportion of employed population(%); X7 ruraltotalexpenditure per capita(Yuan);X8 ruraltotal income per capita(Yuan);and X9 the proportion of primary industry outputvalue(%)have higher loading in the first two principal components,which indicates a closer relationship between theses four factors and F1.These four factors include population economics,rural total income per capita,and the proportion of primary industry output value factor.So principal component F1 can be named as driving population and economic development.

Driving forestry projects and consumer spending

From Table 7,we can see that farmers’per-capita total expenditures,forestry and fishery production value ratio have larger loading than other factors in principal component F2.The proportion of forestry output is the maximum load factors,which shows that the F2 mainly reflects the role offorestry production value.The totalexpenditure per farmer also has a certain load in principal component F2, so we named the principal component F2 as driving forestry projects and consumer spending.

The first two principal components’cumulative contribution rate has reached 100.00%(Table 6),which indicates the first two principal components are already reflected in all the information.So the following analysis focuses on the effectofthese two principalcomponents on the sloping farmland change.

Driving force analysis of terrace

Driving economic development and people’s living standards

PrincipalComponents factorloading matrix was calculated and given in Table 8.The loading of X1 total population (people);X4 the proportion of employed population(%); X7 rural total expenditure per capita(Yuan);X8 rural total income per capita(Yuan);X9 the proportion of primary industry output value(%);X12 the proportion of forestry output value;and X14 the proportion of fishery output value(%);are higher in the firsttwo principalcomponents F1 and F2.F1 reflects thatthese seven factors are mainly relative with economic development and people’s living standards,and principal component F1 can be named for driving economic development and people’s living standards.

Crop cultivation development driving

From Table 8,we can obviously see that the principal component F2 mainly reflects the loading of the X15(the proportion of planting output value)is 0.994,a high loading.So the principalcomponent F2 can be named crop cultivation development driving.

Driving force analysis of forest land

Population and economic development driving

In Table 9,there are some indices larger than others in principal components F1(Abs[0.8),including X1 total population(people);X5 GDP per capita(Yuan);X6 living land area per capita(m2);X8 rural total income per capita (Yuan);X9 The proportion of primary industry output value(%);and X11 the proportion of tertiary industry output value(%).So the principal component F1 mainly reflects these six factors’information and represents thepopulation and economic aspects.The principalcomponent F1 can be named as driving population and economic development.

Table 7 Principal component loading matrix(sloping farmland)

Table 8 Principal component loading matrix(terrace)

Forest fisheries development and employment consumer driving

From Table 9,we can see that X4 the proportion of employed population(%);X7 rural total expenditure per capita(Yuan);X12 the proportion of forestry output value (%);and X14 the proportion of fishery outputvalue(%)are relatively larger loading factors in the principalcomponent F2,which reflect forestry and fisheries development, employment,and the role of consumption.So the principal component F2 can be named as driving forestand fisheries developmentand employment.

Conclusion and discussion

We quantitatively analyzed the land-use changes in number,degree,and type adjustment in the Luoyugou watershed from 1986 to 2004.The results showed that the sloping farmland declined sharply during from 1986 to 2004 and the rate of sloping farmland changed from59.67%in 1986 to 14.03%in 1995,and to 12.26%in 2004(Table 10).Terrace and forestland increased rapidly and changed from 12.90,9.54%in 1986 to 48.36%,and 19.80%in 1995,to 47.94%,22.30%in 2004 (Table 10).Grassland and shrub land declined slowly. Residential land increased slowly,while sloping farmland mainly changed into terraces and forest land.

Table 10 The area proportion of land-use types in 1986,1995 and 2004

Several projects dealing with sloping farmland regulation,natural forest protection,and grain-to-green—helped the area of terrace and forest increased rapidly while sloping farmland reduced sharply.Population growth turned outto be one of the mostdynamic driving forces inland use change,the result of an increase in demand for residential and infrastructure land.The degree of land-use changes of 1986–1995,1995–2004 and 1986–2004 is -0.0356,-0.0237 and-0.0586 respectively.All three values are less than zero so thatitcan indicate thatland use of Luoyugou watershed is stillin the adjustmentphase and the land use type willcontinue to dynamic changes by time.

We selected 16 socio-economic indexes thatare closely related to land-use change in the Luoyugou watershed and used principal component analysis to analyze the driving force ofsloping land,terrace,and forestland,which are the larger watershed land-use change types.The results showed thatsloping farmland changes are mainly driven by population,economic development and forestry projects and consumption spending.Terraced land use changes are driven by economic development,people’s living standards and planting development.Forestland changes are mainly driven by population,economic development,national forestprojects,employmentand consumption spending.

The study involved comprehensive consideration of the relationship between driving forces and land-use change, which matched the view ofCai(2001)thatthe relation ofland use/land coverand population growth and economic restructuring wasnotnecessarily consistentwith globaltrends.From a global scale,large numbers of forests were converted to farmland asthepopulation grew and thedemand forfood rose. A different situation occurred in the Luoyugou watershed, which illustrated that various driving forces and various indexes should be taken into accountto analyze the relationship between land-use change and driving forces.

Itis noteasy to analyze the processes and driving factors of land-use changes because they are very complex processes and affected by the natural environment,socioeconomic development,and government policy and many other factors(Baietal.2004;Zhang 1999;Liu etal.2006; Chen 1997).From here,we should look forremote-sensing image data from earlier years and more comprehensive socio-economic data of the Luoyugou watershed so thatwe can gain more systemized,more scientific and more comprehensive understanding aboutthe land-use changes of the Luoyugou watershed.

AcknowledgmentsThis research was supported by the National Basic Research Program of China(2007CB407207)and National Natural Science Foundation of China(30800888).

Bai WQ,Yan JZ,Zhang YL(2004)Land use/land cover change and driving forces in the region of upper reaches of the Dadu River. Prog Geogr 23(1):71–78

Cai YL(2001)A study on land use/cover change:the need for a new integrated approach.Geogr Res 20(6):645–652

Chen BM(1997)Studies on land use and land coverchange in China and man’s driving force upon it.Nat Resour 2:31–36

Chen YQ,Verburg Peter H(2000)Multi-scale spatialcharacterization of land use/land cover in China.Sci Geogr Sin 20(3):197–202

Chen HS,Xia B(2012)Analysis of land use change in rapidly urbanized area-a case study of Dongguan city,Guangdong Province.Res Soil Water Conserv 19(1):239–242

Hall CSC(1995)Modelling spatial and temporal patterns of tropical land use change.J Biogeogr 22:753–757

Hao SL,Cao LH,LiBC(2007)Study on driving force ofland use and land cover change on small scale.Res Soil Water Conserv 14(2):228–230

Haung AJ(2004)A analysis on the land use dynamic change and its driving forces in Nanping Municipality.Fujian Geogr 19(3):18–20

Holling CS(1992)Cross 2 scale morphology,geometry and dynamics of ecosystems.Ecol Monogr 62(4):447–502

Hou LS,Cai YL(2004)Substance analysis and progress review of study on land use/cover change.Prog Geogr 23(6):96–104 in Chinese

Hu M(2007)Study on land use change and its driving forces in Lanzhou.Xibei Normal University,Lanzhou,pp 25–35 in Chinese

Lambin EF,Geist HJ(2001)Globalland-use and land-cover change: what have we learned so far?Glob Change News Lett46:27–30

Li TS,Jiao CX,Zhang H etal(2003)Analysis on the land utilization dynamic of Xi’an.J Arid Land Resour Environ 17(4):29–33

Li XB,Lei SP,Huang ML(2005)Analysis on the land use degree change and the driving forces in Jianou City.J China West Normal University(Nat Sci)26(2):223–226

Li ZH,Song GB,Gao JX(2006)The research on the relationship of land use change and pattern and passage effect in LRGR.Chin Sci Bull 51:90–99

Li X,Wu KJ,Cao Q(2011)The study of Tianshui land use change and its ecological service value.J Yellow River 33(4):66–69

Liao YP,Wang JH,Fu R(2011)An analysis on dynamic changes and driving factors of the sandy land of Aba District in Northwest Sichuan Province.Res Soil Water Conserv 8(3):51–54

Liu DW,Song KS,Wang DD,Zhang SP(2006)Dynamic change of land-use patterns in west part of Songnen Plain.Sci Geogr Sin 26(3):277–283

Loveland TR,MerchantJW,Ohlen DO etal(1991)Developmentofa land cover characteristics database for the conterminous US. Photogramm Eng Remote Sens 57:1453–1463

Luo GL(2003)Study of land use change and driving forces with Jiangan County as an example.Southwest Agricultural University,Chongqing,pp 45–56 in Chinese

Luo XH,Ni JR(2000)Progress of study on land use and land cover change.J Apply Found Eng 8(3):262–272 in Chinese

Ma XH(2003)Analysis to land use change and driving forces in suburb of Tongzhou City.Shanxi Normal University,Tongzhou, pp 1–49 in Chinese

Mayer WB,Turner BL II(1996)Land-use/land-cover change: challenges for geographers.Geo J 39(3):237–240

Ren ZY,Zhang YF,Li J(2003)Land use change and ecological security evaluation.Science Publishing Company,Beijing

Ren QL,Ming D,Jian YC etal(2007)Quantification ofthe impactof land-use changes on ecosystem ervices:a case study in Pingbian county,China.Environ Monit Assess 1(28):503–510

Rindfuss RR,Walsh SJ,Turner BL II et al(2004)Developing a science of land change:challenges and methodological issues. Proc Natl Acad Sci 101(39):13976–13981

ShiPJ,Song CQ,Jing GF(2002)Strengthening the study ofland use/ cover change and its impacton eco-environmentalsecurity.Prog Earth Sci 17(2):161–168 in Chinese

Stern PCOR,Young D Druckman(eds)(1992)Globalenvironmental change:understanding the human dimension.National Research Council Report,Washington D.C

Turner II BL,Meyer WB,Skole DL(1994)Global land-use/landcover change:towards an integrated program of study.Ambio 23(1):91–95

Wang XL,Bao YH(1999)Study on the methods of land use dynamic change research.Prog Geogr 18(1):81–87

Wu MF,Ou MH,Liao RH(2012)Analysis on land use change and its driving forces in the developed area:a case study of Guangdong Province.Res Soil Water Conserv 19(1):179–183

Xie F,Shu XB,Liao FQ,Xiao-yu MO(2011)Analysis on land use changes and driving forces in Fuliang County,Jiangxi Province. Res Soil Water Conserv 18(2):213–221

Yang M,Zhang GL,Hou YP(2011)Advances and prospects of the driving force of regionalland use change researches.Geogr Geoinf Sci27(1):95–100

Yang J,Zhuang JY,Zhang JC(2013)Study on the change ofland use in the Xuzhou city based on RS and GIS.J Nanjing For Univ (Natural Sciences Edition)37(2):85–91

Yukio H(2001)Some thoughts on human aspects of LUCC in Japan and China.In:Proceeding of international conference on land use/cover change dynamics—LUCCD,Beijing,17–26

Zhang M(1999)Statisticalanalysis to regionalland use structure and its driving forces.J Nat Resour 4(4):381–384

Zhang Y,Liu GM,Ma HY et al(2004)Changing characteristics of land-use and land-cover in the Heihe River basin,China during 1987–2000.J Glaciol Geocryol 26(6):740–746

Zhuang DF,Liu JY(1997)Study on the model of regional differentiation of land use degree in China.J Nat Resour 12(2):105–111(in Chinese)

5 May 2013/Accepted:17 September 2013/Published online:27 January 2015

The online version is available athttp://www.springerlink.com

Corresponding editor:Chai Ruihai

Key Laboratory of Soil and Water Conservation& Desertification Combating,Beijing Forestry University, Beijing 100083,China

e-mail:nexk@bjfu.edu.cn