Comparison of sustainability issues in two sensitive areas of China

Journal of Geochemical Exploration 110 (2011) 15–22

Contents lists available at ScienceDirect

Journal of Geochemical Exploration

j ourna l homepage: www.e lsev ie r.com/ locate / jgeoexp

Households’ willingness to reduce pollution threats in the Poyang Lake region,southern China

Lin Zhen a,⁎, Fen Li b, Heqing Huang a, Oliver Dilly c, Jiyuan Liu a, Yunjie Wei a,d, Li Yang a,d, Xiaochang Cao a,d

a Center for Natural Resources and Environmental Security, Institute of Geographic Sciences and Natural Resources Research, Chinese Academy of Sciences, Beijing 100101, Chinab Beijing Research & Development Centre, Shen Zhen Institute of Building Research, Beijing 100044, Chinac School of Integrated Climate System Sciences, KlimaCampus, Universität Hamburg, Hamburg 20144, Germanyd Graduate University of Chinese Academy of Sciences, Beijing 100049, China

⁎ Corresponding author at: Center for Natural ResourcInstitute of Geographic Sciences and Natural ResourcesSciences, 11A Datun Road, Chaoyang District, Beijing 1008196; fax: +86 10 6485 4230.

E-mail addresses: [email protected], linlinzhen@ya

0375-6742/$ – see front matter © 2011 Elsevier B.V. Aldoi:10.1016/j.gexplo.2011.02.003

a b s t r a c t

Article history:Received 14 May 2010Accepted 7 February 2011Available online 17 February 2011

Keywords:Water pollutionEutrophicationEnvironmental awarenessWillingness to payPoyang LakeSouthern China

Environmental threats to wetland ecosystems are increasing, and these ecosystems are becoming increasinglysensitive to human impacts, leading to deterioration of these already fragile ecosystems. Poyang Lake is thelargest freshwater lake in China and one of the most important wetlands in the world. However, waterpollution and related environmental changes have increasingly drawn the scientific community's attention.The goal of this paper is to provide insights into the environmental threats to the Poyang Lake region asperceived from the households’ perspective, and to investigate their willingness to pay for conservation of thelake's environment. We collected both primary and secondary data through a questionnaire delivered to 270households and analysis of existing water monitoring data. The major threat confronting the Poyang lakeregion is water pollution; water quality, as represented by the total nitrogen (TN) and total phosphorus (TP)concentrations and by chemical oxygen demand (COD), suggests a moderate to severe degree ofeutrophication. The situation has worsened in recent years, particularly due to high TN and TP in agriculturaldrainage water caused by increasingly intensive use of chemical fertilizers by local farmers. Most householdswere willing to pay to mitigate these threats, but the magnitude of the payment was related to a farmer'sdependence on the lake for their production and daily life. The results of our analysis will help managersdevelop more effective environmental management policies.

es and Environmental Security,Research, Chinese Academy of101, China. Tel.: +86 10 6488

hoo.com (L. Zhen).

l rights reserved.

© 2011 Elsevier B.V. All rights reserved.

1. Introduction

Human societies depend on ecosystems and their delivery ofenvironmental services, but through the actions of broader biogeo-chemical cycles, human activities at one place can influenceenvironmental conditions and people elsewhere (Vatn, 2010). Thisproblem becomes severe in wetland ecosystems, where people andthe environment are interconnected through the water resourcesprovided by the ecosystem and used by local residents. This isparticularly true in China's Poyang Lake region. Poyang Lake is thelargest freshwater lake in China and one of the most importantwetlands in the world. It is situated in northern Jiangxi province(Fig. 1), near the southern bank of themiddle and lower reaches of theYangtze River, which accounts for 9% of the Yangtze River basin. Thelake exhibits marked seasonal changes in both area and volume(Wang et al., 2004). Poyang lake is an overflow lake; instead of being

supplied by streams or rivers that flow directly into the lake, itincreases in size when the Yangtze River overflows its banks or duringheavy rainfalls that produce large volumes of surface flow, anddecreases in size due to less water inflow from the Yangtze River,reduced precipitation, deep drainage and evaporation, and bothreceives and releases water in response to seasonal variations. Theshape and water level in the lake vary seasonally, with a 13-mdifference in the water height between the summer rainy season andthe winter dry season. In the study area, mean temperatures rangefrom a low of 4.4 °C in winter to a high of 30.0 °C in summer. Annualprecipitation averages 1387–1795 mm, with 48.2% of this total fallingduring the summer growing season (Wang et al., 2004). Because ofthe fertile surrounding land and rich bio-resources, Poyang Lake is animportant habitat for a number of wild animals, including 159 waterbird species (53% of the total number in China). It is one of the mostimportant overwintering areas for migratory water birds in Asia, andis home to 98% of the global population of Siberian cranes (Grusleucogeranus), 50% of white-naped cranes (Grus vipio), 50% of swangeese (Anser cygnoides), and tens of thousands of egrets, spoonbills,storks, swans, geese, ducks, and shorebirds. Based on the existingresearch (e.g., Li et al., 2009), it is known that the Poyang Lake servesall the main functions of an ecosystem, including provisioning,

Fig. 1. Location of Poyang Lake in China and in Jiangxi Province.Software: ArcGIS. Source: Data Center for Resources and Environmental Sciences, Chinese Academy of Sciences, 2004.

16 L. Zhen et al. / Journal of Geochemical Exploration 110 (2011) 15–22

supporting, and regulating environmental functions, as well asserving cultural and recreational functions for local residents.

Poyang Lake and its surrounding regions comprise two cities andten counties: Nanchang and Jiujiang cities, and Nanchang, Xinjiang,Jinxian, Yongxiu, Dean, Xingzi, Duchang, Hukou, Yugan, and Poyangcounties. These cover a total land area of 202.89 ha and sustain apopulation of 8.86 million. The rural population totals 5.99 million(68% of the total). Farmers practice small-scale farming activities, andcultivate more than ten types of crops in small farm fields, includingrice (double or triple cropping), cotton, sweet potato, beans, peanuts,oil plants, vegetables, and fruits. To ensure a high yield, farmersusually use high inputs of fertilizers and pesticides.

Arable land and water area surrounding the lake are the two maintypes of land use (Fig. 2), and there is usually competition between thetwo types of land use in areas at risk of flooding. The area of arable landper capita totals 0.045 ha, which is only 42.4% of China's average valueand less than the FAO's minimum recommendation (=0.053 ha) tosustain a human life. The most significant land use changes in recent

Fig. 2. Land use patterns in the Poyang Lake region in 2008.Software: ArcGIS. Source: Data Center for Resources and Environmental Sciences,Chinese Academy of Sciences, 2005.

years have been a decreasing farmland area and an increasing waterarea. The farmland area decreased by 19 416 ha between 1997 and2005 (a 4.7% decrease compared with the area of cultivated land in1997), and the total area of bodies of water increased by 14 661 ha (by3.9% of the area in 1997) as a result of a land use policy designed toconvert cultivated land into natural water areas to assist in wetlandconservation (Wang et al., 2004). During this period, the forest areadecreased by 1066 ha (0.2% of the 1997 value) and the grassland areadecreased by 7.5% of the 1997 value. The built-up land increased from59 148 ha in 1997 to 62 957 ha in 2005, a 6.4% increase, due toincreasing urbanization, and the area of unused land increased slightly(by 0.42%) during the same time period.

Due to population growth (a 19.3% increase from 1997 to 2007)and increasingly intensive economic activities, environmental degra-dation has begun to occur. Pollution (including eutrophication) hasintensified due to the discharge of waste water from industrial anddomestic sources (Yang et al., 2011), as well as due to agriculturalchemical inputs (Lv, 1996; Wang et al., 2008). The total nitrogen (TN)and total phosphorus (TP) concentrations averaged 1.06 mg/L and0.067 mg/L, respectively, in the lake water, indicating severeeutrophication (Wang et al., 2008). The region's biodiversity hasalso decreased due to water pollution (Yu and Sun, 2006).

These statistics suggest that overuse of agricultural chemicals is amajor cause of the lake's water pollution problems (e.g., Wang et al.,2008). Researchers from other countries have also explored theimpacts of agricultural chemicals on water quality; for instance, OECD(1982) summarized data from around theworld and reported that thenitrogen and phosphorus loading from communities and from thefertilization of cultivated land and forests have resulted in excessivenutrient inputs to lakes around the world.

Local households are the users and managers of the region'sresources, and Wunder et al. (2008) noted that, at least in theory,systems based on the needs of these households are “muchmore likelyto be efficient” because these people have better knowledge of theresource and their needs than is possible for a central managementagency. Therefore, it is imperative for decision-makers to understandthe perspectives of the local people on environmental changes and theirwillingness to pay (WTP) for conservation of the resources that sustainthem; this knowledge can guide them to improve management of theresources. In the present study, we attempted to provide insights intothe environmental challenges in the Poyang Lake region from thehouseholds’ perspective. To do so, we investigated their willingness topay for conservation of the lake's environment.We used environmental

17L. Zhen et al. / Journal of Geochemical Exploration 110 (2011) 15–22

and socioeconomic research methods based on both primary andsecondary data sources for the entire area.

2. Research methodology

The secondary data included a 1:100 000 scale land use mapproduced from LandSat TM satellite images produced by the Resourceand Environmental Data Center of the Chinese Academy of Sciences(Liu et al., 2002). Other environmental data and documents werecollected from the Lake Management Committee of Jiangxi Province,the Environmental Protection Department of Jiangxi Province, theManagement Committee of the Poyang Lake Nature Reserve, theCounty Bureau for Science and Technology, and the TownshipCommittees of administrative regions surrounding the lake. Theprimary data were derived through household questionnaires, groupdiscussions and staff members of the local institutions.

2.1. Household questionnaire

In this study, we defined a household as a basic family unit in whichtwo or more generations live together to share expenditures andincomes. We developed a structured household questionnaire toidentify the perceptions of farmers aboutmajor environmental changesand their willingness to pay to protect their environment. We surveyedthe participants using a structured interview process based on aquestionnaire thatwe had previously tested to ensure the validity of theresults. Based on our discussions with the abovementioned institutions,we selected representative communities inwhich to conduct the survey.The first selection criterion was that the community must be near thelake (Fig. 3); the average distance to the main body of the lake was1.2 km,with a range of 0.5 to 2.0 km; in addition, the economic activitiesand daily life of the survey participants must depend heavily on thelake's water resources. Our study included a total of nine villages in sixcounties, and we surveyed their residents from 25 March to 11 April2008. We used simple random sampling to select the households, andincluded a total of 270 householders in the final survey. These familiesrepresented more than 40% of the households in the selected villages.

2.2. Group discussions

We used group discussions to clarify any questions raised by theresults of our questionnaire-based survey. In the group discussions,

Fig. 3. Location of the villages surveyed in the Poyang Lake region.Software: ArcGIS. Source: GPS data imported from the Google Earth® 2010 Imagery.

we identified the desires of the participants for actions that wouldresolve the problems they identified. In each group discussion, weattempted to identify shared topics of interest related to resourcemanagement and the participants' expectations and willingness topay for environmental conservation.

2.3. Willingness to pay for environmental conservation

We used the contingent valuation method (CVM) to quantify eachhousehold's willingness to pay (WTP) for environmental conserva-tion. CVM is a kind of stated-preference approach that employs ahypothetical market system to extract WTP or willingness to acceptfor environmental goods (Carson, 2000; Hadker et al., 1997). CVM hasbecome one of the most widely used valuation techniques due to itsflexibility and its ability to estimate total values. In this study, wefollowed a specific design that has been developed over several yearsand studies (e.g., Liu and Zhen, 2007; Spash and Hanley, 1995; Spashet al., 2009).The survey was based on four sections: describing thepurpose of the survey to participants and gathering of basicdemographic data; the participant's awareness of environmentalchanges in Poyang Lake and the surrounding regions; the participant'sWTP to reduce pollution threats in the lake; and identification ofpreferred solutions to the perceived problems.

In principle, the WTP represented the amount a respondent waswilling to pay in exchange for an improvement in environmentalquality (e.g., conservation of water, soil, and waterbirds; reduceddumping of wastes in the lake). The specific WTP was the maximumamount they would be willing to pay (from their annual income) torestore the environment of Poyang Lake and its surrounding area.

Existing methods for estimating WTP include Continuous CVM(represented by open-ended questions) and Discrete CVM (repre-sented by Dichotomous Choice questions). In Continuous CVM, theinterviewees are free to answer the open-ended questions by fillingup the maximum amount they are willing to pay; also, it is easy fordata analysis. The disadvantage is that it is sometimes difficult for theinterviewees to give the appropriate answer when they don't haveenough background information about the research object, or whenthey are actually not sure the maximum amount they are able to offerwhen they have to do so (Loomis andWalsh, 1997). While in DiscreteCV, the interviewees are required to show their willingness bychoosing either “Yes” or “No”, they don't need to indicate the specificamount they are able to pay, which can avoid the problem ofinconsistency between the stated WTP and actual amount being ableto pay (Hoehn and Randall, 1987).

To gain necessary data for both WTP and associated specificamount to pay, we used single bound Dichotomous CVM method. Inthe questionnaire, the interviewees were asked “Are you willing topay for environmental conservation of Poyang lake?”, the answer partinclude “Yes, I do” and “No, I don't pay” (refers to Table 2), the bidvalues were given to those who have agreed to pay to choose, whichare 75, 188, 375, 750, 1125, 1875, and 3750 CNY per ha (1 CNY=USD0.1465). Those bid values were determined based on preliminarysurvey data on income levels and pre-interviews with local peopleand officials.

For the specific analysis of WTP, Probit/Logit mode (Haneman,1984) was applied; the standard form of the model is as following:

Prob = 1− 1 + exp B0−B1 Xð Þ½ �f g−1 ð1Þ

Where,

Prob is probability to pay,B0, B1 are regression coefficients,X is the bid value.

Table 1Basic information on the surveyed households (n=270).

Basic information Group Population Proportion of total (%)

Gender Male 197 73Female 73 27

Age (years) ≤30 12 4.531 to 50 104 38.5≥51 154 57.0

Education None 50 18.4Primary school 106 39.3Middle school 79 29.3High school or better 35 13.0

Primary crop Rice 188 69.8Oil plants 42 15.5Cotton 26 9.6Sweet potato 6 2.2Peanuts and sesame 5 1.7Vegetables 2 0.9Other 1 0.3

Income (CNY/year) ≤10 000 94 34.810 001 to 20 000 81 30.020 001 to 50 000 69 25.6≥50 001 26 9.6

18 L. Zhen et al. / Journal of Geochemical Exploration 110 (2011) 15–22

The relationship between the bid values and the consensus rate(i.e., the proportion of the households who are willing to pay thecorresponding bid value) is shown in the following function:

P = 1= 1 + b0 × bx1� � ð2Þ

Where,

P is the consensus rate,x is the bid value, andb0 and b1 are regression coefficients, where b0=e−B0, b1=eB1.

This is the cumulative probability function, and through differen-tiation of x (dx), we can get probability density function; when t tendsto be infinite, the integral probability density function tends to be one.

Expected WTP of households is calculated using the followingformula, which is derived from the above formula (1) in the conditionof WTP≥0. The data was from close-ended question by giving specificbid values (Park et al., 1991; Sheng et al., 2001).

E xð Þ = ∫ +∞0

xρdx ð3Þ

Where,

ρ is probability density of the households’ willingness to pay,and is expressed as:

ρ = −P′ xð Þ ð4Þ

Where,

P is the consensus rate from above formula (2),P′(x) represents the probability of a WTP value less than x.

Continuous CVM including open-ended question was designed tocollect information of individual household's willingness to pay forconservation of water resources in Poyang Lake region. Theinterviewees were asked “how much are you willing to pay forenvironmental conservation of Poyang Lake?” To answer thisquestion, each individual put his/her maximum affordable amountfor payment. Data is analyzed using the following formula (5) to gettheir expected amount of WTP:

E WTPð Þ = ∑4

i=1pibi ð5Þ

Where,

pi is the proportion of households by income groups,i is number of income groups, which is 4 in this study with

respective income of (CNY)≤10 000, 10 001 to 20 000,20 001 to 50 000, ≥50 001, and

bi is the amount of WTP of each income group (CNY/ha/year).

This individual household based WTP is complementary of theresults derived from single bound Dichotomous CVM (i.e., aboveformula (3)), combination of those two estimation methods is veryhelpful for analyzing WTP of the household.

We used version 16.0 of the SPSS software for Windows (SPSS,Chicago, IL) to analyze the data wherever needed, and Analyze-Regression–Linear Regression was used to analyze relationshipsbetween WTP and householders’ characteristics.

3. Results and discussions

3.1. Basic information on the households and their perceptions

The households averaged 4.8 people, of which most (73%) weremale (Table 1). The per capita farmland area was 0.057 ha. Therespondents ranged from 20 to 67 years old, but most (57.0%) were51 years old or older. The respondents 30 years old or younger onlyaccounted for 4.5% of the total. Most (81.6%) had at least a primaryschool education. Most (69.8%) cultivated at least one rice crop astheir primary vegetable, followed by cultivation of oil plants (15.5%,primarily rapeseed) and cotton (9.6%). The annual income of thehouseholds averaged about 11 027 CNY, with the main incomesources being off-farm work (37% of total income), farming (27%),remittance from the relatives or friends working in the cities (13%),government loans (8%), fishing (8%), animal rearing (5%), andgovernment subsidies to pay for conservation of wetland areas (2%),such as payments to stop farming under land conversion program.

3.2. Household awareness of the causes and impacts ofenvironmental change

Our surveys revealed five major perceived environmental pro-blems in the Poyang Lake area (Table 2): water pollution (65% of therespondents), solid waste pollution (58%), soil loss by water erosion(40%), sedimentation (26%), and flood risks (20%). The factors thatwere perceived to contribute most to these adverse changes wererapid population growth (65% of the respondents), inputs of chemicalfertilizers (64%), disposal of solid domestic wastes (31%), intensivefishing (31), and sand mining from the lake and surrounding areas(20%).

Water quality analyses (Table 3) show that since 2002, pollutionlevels have generally worsened. Chemical oxygen demand (COD)increased by an average of 16.7%, versus 129.4% for NH4

+–N. The lake'swater quality degraded from level III in 2002 to level V in 2007, theworst level since 2002. The survey respondents believed thatpopulation growth and the associated increases in inputs ofagricultural chemicals were the major causes of this water pollution.To increase crop production, farmers have applied increasing amountsof agricultural chemicals; for example, fertilizer use increased by18.3% and pesticide use increased by 66.7% (Table 3). All of thesampled households have applied chemical fertilizers for rice, sweetpotato, cotton, oil crops, and vegetables. The main fertilizers wereurea, diammonium phosphate (DAP), and potassium sulfate (PS).

Table 2Responses to questions designed to reveal the awareness and attitudes of local people.Source: Household surveys, 2008.

Question Responses Proportion ofrespondents (%)

1 2 3 4 5

What negative environmental changes have occurred in thePoyanghu Lake region in recent years?

(1) Water pollution, (2) Soil loss by water erosion, (3) Sedimentation,(4) Solid waste pollution, (5) Flood risk

65 40 26 58 20

What are the main causes of these environmental changes? (1) Population growth, (2) Chemical inputs, (3) Domestic wastes,(4) Intensive fishing, (5) Sand mining

65 64 48 31 20

What conservation measures would you be willing to adopt? (1) Reduce chemical inputs, (2) Convert lakeside fields into natural wetlands,(3) Safe disposal of domestic wastes, (4) Protect waterbirds, (5) Reduce fishing

43 60 38 30 19

Are you willing to pay for environmental conservation of Poyang lake? (1) Yes, I pay (2) No, I don't pay 84 16What would you like your payment to be used for? (1) Clean the lake's water, (2) Safe disposal of domestic wastes,

(3) Hiring local residents as inspectors to prevent future pollution,(4) Develop eco-tourism, (5) No ideas

81 34 78 41 8

Table 4Sources of nitrogen and phosphorus pollution in the water of Poyang Lake.Source: Wang et al. (2008).

19L. Zhen et al. / Journal of Geochemical Exploration 110 (2011) 15–22

These are the most commonly used fertilizers in China due to theirhigh nutrient concentration, low price, and high availability. Thefertilizer combination differed between crops; urea+DAP was usedby 84% of rice farmers, and the remaining rice farmers applied a morebalanced combination (urea+DAP+PS). Vegetable farmers normallyused urea+DAP+PS; this is because the higher economic benefitsderived from vegetable cultivation and sales led farmers to devotespecial care to vegetable production.

Application of fertilizers (estimated based solely on the nitrogencontent) increased from 1.06 t N/ha annually in 2002 to1.22 t N/haannually in 2007, and both amounts are much higher than themaximum amount (0.250 t N/ha) recommended by the Ministry ofEnvironmental Protection (under decree number [2007]195). Pesticideis also intensively used, with the amount increasing from 0.04 t/haannually in 2002 to 0.07 t/ha annually in 2007 (Table 3).

Increasingly intensive use of chemical fertilizers and pesticides hasincreased crop yield very significantly over the past years, however, ithas also brought severe environmental concerns. Of the 130 largelakes in China, more than 60 are seriously polluted, mainly due to theinput of N and P from different sources, and increased application of Nand P fertilizer is believed to be a primary cause of pollution of lakewater by these nutrients (Zhang et al., 1995). A recent investigation ofPoyang Lake (Wang et al., 2008) showed that nitrate pollution of thelake's water due to excessive N fertilization has become a seriousproblem. The maximum TN and TP concentrations increased from0.076 mg/L and 0.684 mg/L in 1988, respectively, to 0.148 and2.38 mg/L in 1996 (Wang et al., 2008). Concentrations of TN and TPin the agricultural drainage water reached 13.47 and 28.63 mg/L,respectively (Table 4), implying extremely severe eutrophicationbased on the criteria in Table 5. The water samples were obtainedduring the second cultivation season for rice, therefore the majorsource of these pollutants is applied fertilizers. Another explanationfor the high P concentration is the release of soluble reactivephosphorus from lake-bottom sediments as a result of sand mining(Wang et al., 2004). High concentrations of NO3

––N (7.35 mg/L) and

Table 3Water quality in Poyang Lake and inputs of agricultural chemicals in the surroundingfarmland (COD, chemical oxygen demand; DOC, dissolved oxygen content).

2002 2003 2004 2005 2006 2007

COD (mg/L)a 1.2 0.9 1.4 0.9 1.6 1.4NH4

+–N (mg/L)a 0.85 1.02 0.92 1.03 0.78 1.95DOC (mg/L)a 7.3 7.64 7.51 6.23 5.8 3.26Water quality levelb III IV III IV III VAnnual fertilizer use (t N/ha)c 1.03 1.06 1.24 1.37 1.34 1.22Annual pesticide use (t/mu)c 0.04 0.04 0.05 0.06 0.06 0.07

a Wang (2007).b http://datacenter.mep.gov.cn.c Bureau of Statistics of Jiangxi (2003–2008).

NH4+–N (5.48 mg/L) were also observed in the groundwater and

urban sewage, respectively.Recent tests of water samples indicate that the TN concentrations

in Poyang Lake ranged between 1.17 and 2.15 mg/L and TP rangedfrom 0.08 to 0.12 mg/L (Wu andWen, 2009), which are both classifiedas severe eutrophication (Table 5).

The COD level has also increased in recent years, reaching amaximum of 1.6 mg/L in 2006 (Table 3), which represents a 33.3%increase from the 2002 level. Based on the criteria in Table 5, thisrepresents a moderate eutrophication level, implying a high risk ofsevere eutrophication in the future, if current chemical use is not to bereduced.

3.3. Household perceptions of environmental conservation

To protect water quality, most of the respondents (60%) werewilling to convert their lakeside farm fields into natural wetlands; thiswas followed by a willingness to reduce chemical use (43%) and tosafely dispose of domestic solid wastes (38%). The conversion offarmland into wetland is one of the central government's majorstrategies to protect the Poyang wetland areas (Tang, 2006). Since1998, China has conducted a series of flood-prevention projects,including this type of land conversion. About 221 000 householdswith a total population of 908 200 moved away from lakeshore areas,and most of the remaining households who still live near the lakehave converted some of their farmland into natural wetland. Thegovernment has invested 341.35 billion CNY in this project, leading tothe conversion of 86000 ha. In December 2005, the State Councilissued a decree entitled “Operationalizing Scientific DevelopmentConcepts and Strengthening Environmental Protection”, in which it

Water sources NO3−–N (mg/L) NH4

+–N (mg/L) TN (mg/L) TP (mg/L)

Agriculture 1.15 0.90 13.47 28.63Urban 0.97 5.48 6.55 1.15Groundwater 7.35 0.28 7.80 0.08

Table 5The criteria used to assess the severity of lake eutrophication in China (Shu et al., 1996).

Eutrophicationlevel

Criteria for evaluation

TN (mg/L) TP (mg/L) COD (mg/L)

Low b0.4 b0.02 b0.3Moderate 0.4 to 1.2 0.02 to 0.05 0.3 to 2.0Severe N1.2 N0.05 N2.0

20 L. Zhen et al. / Journal of Geochemical Exploration 110 (2011) 15–22

recommended that ecological projects such as farmland conversionshould continue.

Under this policy, each household has converted some of theirfarmland into wetland (an average of 0.14 ha per household,equivalent to 0.03 ha per person), accounting for one-third of thetotal land area according to our survey. Our previous researchrevealed a negative correlation between the area of land conversionunder this policy and household income (r=−0.475, pb0.01) in thePoyang Lake region (CCICED, 2007). To compensate farmers for theeconomic loss resulting from their loss of farmland, the governmenthas been providing compensation both in cash and in kind. For eachunit area (ha) of converted farmland, farmers receive a payment of177 CNY/year and 35.5 kg of grain (rice). Farmers are eager to converttheir remaining farmland into wetland because of the governmentcompensation scheme, but workers released from farming activitiesmust have a chance to find off-farm jobs that will let them earnincome to replace the farm income they have lost. For the householdswe surveyed, we found that 37% of household income was derivedfrom working away from the farm (e.g., in cities as migrant workers),the main reasons are insufficient land areas for faming and lowincome from faming.

3.4. Willingness to pay for the conservation of Poyang Lake

Most farmers (84%) were willing to pay for environmentalconservation (Table 2); most wanted their payment to be used forcleaning the lake's water (81%) and hiring inspectors to protect thewater against future pollution (78%), but many also wanted it to beused for developing an eco-tourism industry (41%) and safelydisposing of domestic solid waste through establishing disposingplants (34%). The bid amount was negatively correlated with theproportion of respondents who were willing to pay that amount(Table 6); the greater the required payment, the fewer respondentswere willing to pay owing to low income of the households and onlyabout 8.5% of them had average annual income of 20 001 CNY andabove. Most households (84%) were willing to pay 75 CNY annuallyper ha of farmland, and 58% were willing to pay as much as375 CNY/ha. However, only 11% were willing to pay as much as3750 CNY/ha.

Householders with the highest income were willing to pay themost (2486 CNY/ha annually) for environmental conservation(Table 7), but these householders only accounted for 6.3% of the

Table 6The relationship between the bid amount and the proportion of the households willingto pay that amount for the conservation of Poyang Lake.Source: Field survey, 2008.

Annual bid amount (CNY/ha) 75 188 375 750 1125 1875 3750

Proportion of households willingto pay (%)

84 69 58 41 33 30 11

Table 7Relationship between household income and WTP.

Income group (CNY/year)

≤10 000 10 001 to20 000

20 001 to50 000

≥50 001

WTP (CNY/ha/year) 424 1020 454 2486Proportion of those who arewilling to pay (%)a

35.4 40.6 17.7 6.3

Proportion of total respondents (%) 12.6 14.4 6.3 2.2

Note: Using formula (5), the average amount of WTP is calculated, and the amount is802 CNY/ha/household/year.

a Is ratio of Number of households who have shown their WTP in each income groupto Total number of households who have shown their WTP.

households that werewilling to pay; the groupwith an annual incomeof 10 001 to 20 000 CNY had the next highest WTP (1020 CNY), andaccounted for 40.6% of the households that were willing to pay (thehighest proportion). This can be explained by the fact that in thePoyang Lake region, about one-third of the households rely wholly onfarming activities for their livelihood, and this group has an annualincome ranging between 10 001 and 20 000 CNY. This group is mostaware of the importance of the quality of the lake's water and of thesurrounding environment for their daily life, and they therefore showa relatively high willingness to pay for improvement. For those whoseincome was above 20 001 CNY, they relied mainly on off-farm incomesource working as migrant workers in the cities, and their average off-farm income accounted for more than 65% of the total income, theyrelied less on farm land and therefore had less WTP for the Lake'sconservation.

To calculate the expected amount the households would be willingto pay for the environmental protection of Poyang Lake, we usedlogistic regression analysis. Based on the above data, we plotted thegraph shown in Fig. 4, which shows the relationship between the bidamount and the proportion of households who were willing to paythat amount (P). Using the data from Table 6, we obtained thefollowing regression equation:

P =1

1 + 0:444 × 1:003t F = 23:469; P = 0:005;R2 = 0:824� �

The regression was both close and significant. The relationshipbetween the bid amount and the probability density of the house-holds’ willingness to pay is expressed as follows:

ρ=−P′ tð Þ= 1:332 × 10−3 ×1:003t

1 + 0:444� 1:003t� �2 F=23:469; P=0:005;R2 =0:908

� �

Fig. 4 shows that the consensus rate decreases steadily withincreasing bid value. Most of the farmers were willing to pay between750 and 1125 CNY/year (with a high probability density of nearly8×10−4 for this range of values). The expected amount to pay can beestimated using Eqs. (2)–(4), which predict an average annual perhousehold payment of 956 CNY/ha.

Meanwhile, by using formula (5), the average amount of WTP wascalculated, and the amount is 802 CNY/ha/household/year. This andthe above results could be used as reference values to design aconservation payment scheme and determine the total fundingrequired to promote conservation in the Poyang Lake region.

Fig. 4. Functional relationships between bid amount, consensus rate, and probabilitydensity. Software: Matlab.

Table 8Relationships between WTP and householders’ characteristics.

Parameters Unstandardizedcoefficients (UC)

Std. error t Sig.

Age −0.004 0.003 −1.216 0.229Average annual income −1.438 5.819 −0.247 0.806Production income 0.023 0.012 1.913 0.060Area converted from arableland to lake

25.744 31.207 0.825 0.413

Perception of water qualitya −0.047 45.377 −0.444 0.658Awareness of the lake's functionb 19.999 99.218 0.202 0.841

a 1—know very well; 2—know; 3—know a little bit; 4—not know.b 1—aware; 2—not aware.

21L. Zhen et al. / Journal of Geochemical Exploration 110 (2011) 15–22

Analysis of relationship between WTP and householders’ char-acteristics indicates (Table 8) that the coefficient of productionincome is statistically significant with WTP (Sig=0.06), this isbecause the higher the income from farm land production, the higherreliance on the land and water for farming, therefore, the moreconcern about wetland and the lake's water quality. While othercoefficients such as age, average annual income, area converted fromarable land to lake, perception of water quality, and awareness oflake's function are not statistically significant with WTP, therefore,those characteristics have not affected WTP significantly.

4. Conclusion

Our research used government statistics, household surveys, andfocused discussion groups to identify resident perceptions of theenvironmental threats in the Poyang Lake region of southern China,and household awareness of the causes and impacts of these threats.We also estimated their willingness to pay for conservation of thePoyang Lake region. We obtained the following main conclusionsfrom our analysis:

• Water pollution is the major environmental threat in the PoyangLake region, with TN, TP, COD, NO3

−–N, and NH4+–N concentrations

increasing from 2002 to 2007 and current water quality classified inthe moderate to severe eutrophication levels.

• Increasing overuse of fertilizers and pesticides by local farmers, andparticularly of nitrogen fertilizers, is the main cause of the pollution.Rapid population growth has led to more intense economicactivities and produced more solid wastes and pollutants that aredischarged into the lake and the surrounding environment.

• Most households are willing to pay to conserve the lake throughvarious means, including reduced use of agricultural chemicals andconversion of lakeshore rice fields into natural wetlands. Mostwould also be willing to pay for environmental conservation of thelake region; the average annual payment per household wasestimated to be 956 CNY per ha based on CVM (Haneman, 1984).To cross-check the result, we have designed open-ended question toget individual household based WTP, which was 802 CNY per ha.Therefore, we could conclude that average annual amount of WTPper household is 802–956 CNY/ha, which is nearly 1/10 of thehouseholds’ annual income derived from arable land production.This result is of significance for real policy making because firstly, itindicates a high investment is needed for conservation of PoyangLake's environment. Currently, the household receives 177 CNY/hafrom the government for conservation of the lake's water, which isfar from meeting the demand as mentioned by the local peopleduring the survey. The estimated result is helpful for thegovernment to design a conservation payment scheme and makebudget plan to promote conservation. Secondly, a higherWTP of thelocal people implies their high sense and awareness of environ-mental protection.

• According to descriptive analysis on the numerical data ofquestionnaire, a significant positive correlation was observedbetween income from agricultural production and WTP, implyinga strong relation between households’ dependence on wetland'sland and water resources and their concerns to the Lake'sconservation. The result would provide important information forthe management of the lake by taking into consideration ofperceptions and willingness of the householders who are involvedheavily in farming activities, because they are the real users andmanagers of wetland ecosystem.

• To meet the growing pressure on the region's water resources, ourresults suggest that managers should focus on encouraging moreefficient and balanced use of chemical and organic fertilizers.Developing appropriate payment systems to encourage moreefficient and sustainable use of the region's resources would alsoencourage conservation.

Acknowledgements

We are grateful to the Editor and two reviewers for their veryconstructive comments and suggestions for improvement of themanuscript. This research was funded by the National Key Project forbasic research (973): Ecosystem Services and Ecological Safety of theMajor Terrestrial Ecosystems of China (grant number: 2009CB421106);China–EU Corporation Program of Ministry of Science and Technology(MOST) of China (grant number: 0813); Technical Support Project ofMOST (grant numbers: 2006BAC08B06 and 2008BAK50B05). We aregrateful for the support provided by these organizations.

References

Bureau of Statistics of Jiangxi, 2003–2008. Jiangxi Statistical Yearbook. China StatisticsPress, Beijing. (in Chinese).

Carson, R.T., 2000. Contingent valuation: a user's guide. Environmental Science &Technology 34 (8), 1413–1418.

CCICED, 2007. Eco-compensation mechanisms and policies in China. Science Press,Beijing. (in Chinese).

Hadker, N., Sharma, S., David, A., Muraleedharan, T.R., 1997. Willingness to pay forBorivli National Park: evidence from a contingent valuation. Ecological Economics21 (2), 105–122.

Haneman, W.M., 1984. Welfare evaluations in contingent valuation experiments withdiscrete responses. American Journal of Agricultural Economics 66, 332–341.

Hoehn, J.P., Randall, A., 1987. A satisfactory benefit cost indicator from contingentvaluation. Journal of Environmental Economics and Management 14 (3),1226–1247.

Li, F., Zhen, L., Huang, H.Q., Han, P., Liu, X.L., Jiang, L.G., Wei, Y.J., 2009. Impacts of landuse functional change on WTA and economic compensation for core stakeholders:a case study in Poyang Lake. Resources Science 31 (4), 580–589 (in Chinese).

Liu, X.L., Zhen, L., 2007. Stakeholders’ consumption of ecosystem services andwillingness to accept: a case study in Jinghe Watershed. Resources Sciences 29(4), 103–108 (in Chinese).

Liu, J.Y., Liu, M.L., Zhuang, D.F., Zhang, Z.X., Deng, X.Z., 2002. Spatial pattern analysis ofland use change in China in recent years. Science in China (Series D) 32 (12),1031–1040 (in Chinese).

Loomis, J.B., Walsh, R.G., 1997. Recreation Economic Decisions: Comparing Benefits andCosts (2nd) M. Venture Publishing Inc.

Lv, L.J., 1996. Investigation on Poyang lake water pollution by eutrophication. Journal ofLake Science 8 (3), 241–247 (in Chinese).

OECD, 1982. Eutrophication of Waters. Monitoring, assessment and control. Organisa-tion for Economic Co-operation and Development, Paris.

Park, T., Loomis, J.B., Creel, M., 1991. Confidence intervals for evaluating benefitsestimates from dichotomous choice contingent valuation studies. Land Economics67 (1), 64–73.

Sheng, Z., Xie, S.Q., Pan, C.Y., 2001. 3rd. Probability and StatisticsHigh Education Press,Beijing.

Shu, J.H., Huang, W.Y., Wu, T.G., 1996. Studies on the classification of trophic types ofChina's Lakes. Journal of Lake Sciences 8 (3), 193–200 (in Chinese).

Spash, C.L., Hanley, N., 1995. Preferences, information and biodiversity preservation.Ecological Economics 12 (3), 191–208.

Spash, C.L., Urama, K., Burton, R., Kenyon, W., Shannon, P., Hill, G., 2009. Motives behindwillingness to pay for improving biodiversity in a water ecosystem: economics,ethics and social psychology. Ecological Economics 68 (4), 955–964.

Tang, C.J., 2006. Research on land use structure dynamic change in the wetlandrestoration area—a case study of Poyang Lake area. Jiangxi Normal University,Nanchang. (in Chinese).

Vatn, A., 2010. An institutional analysis of payments for environmental services.Ecological Economics 69 (6), 1245–1252.

22 L. Zhen et al. / Journal of Geochemical Exploration 110 (2011) 15–22

Wang, M.L., 2007. Spatio-temporal distribution of nitrogen and phosphorus and itsgeochemical modeling in Poyang Lake Catchments. Nanchang University, Nanchang.(in Chinese).

Wang, X.H., Fan, Z.W., Cui, L.J., Yan, B.Y., Tan, H.R., 2004. Wetland ecosystem assessmentof Poyang Lake. Science Press, Beijing. (in Chinese).

Wang, M.L., Zhou, W.B., Hu, C.H., 2008. Status of nitrogen and phosphorus in waters ofLake Poyang Basin. Journal of Lake Science 20 (3), 334–338 (in Chinese).

Wu, Y.J., Wen, S.Z., 2009. Nitrogen, phosphorus concentrations and distribution featureof nutrient in waters of Poyanghu Lake. Modern Enterprise Culture 14, 15–16(in Chinese).

Wunder, S., Engel, S., Pagiola, S., 2008. Taking stock: a comparative analysis of paymentsfor environmental services programs in developed and developing countries.Ecological Economics 65 (4), 834–852.

Yang, Z.P., Lu, W.X., Long, Y.Q., B.H., Yang, Q.C., 2011. Assessment of heavy metalscontamination in urban topsoil from Changchun City, China. Journal of Geochem-ical Exploration 108, 27–38.

Yu, Z.Y., Sun, C., 2006. Fishery resources in Poyang Lake. Chinese Fisheries Economics 5,43–45 (in Chinese).

Zhang, W.L., Tian, Z.X., Zhang, N., Li, X.Q., 1995. Investigation of nitrate pollution inground water due to nitrogen fertilization in agriculture in North China. PlantNutrition and Fertilizer Sciences 1 (2), 80–87 (in Chinese).