Improving water quality in China: Environmental investment paysdividends
Yongqiang Zhou a, b, c, 1, Jianrong Ma d, 1, Yunlin Zhang a, *, Boqiang Qin a, **,Erik Jeppesen c, e, Kun Shi a, Justin D. Brookes f, Robert G.M. Spencer g, Guangwei Zhu a,Guang Gao a
a Taihu Laboratory for Lake Ecosystem Research, State Key Laboratory of Lake Science and Environment, Nanjing Institute of Geography and Limnology,Chinese Academy of Sciences, Nanjing, 210008, Chinab University of Chinese Academy of Sciences, Beijing, 100049, Chinac Sino-Danish Centre for Education and Research, Beijing, 100190, Chinad Key Laboratory of Reservoir Aquatic Environment, Chongqing Institute of Green and Intelligent Technology, Chinese Academy of Sciences, Chongqing,400714, Chinae Department of Bioscience and Arctic Research Centre, Aarhus University, Vejlsøvej 25, DK-8600, Silkeborg, Denmarkf Water Research Centre, Environment Institute, School of Biological Science, University of Adelaide, 5005 Adelaide, Australiag Department of Earth, Ocean and Atmospheric Science, Florida State University, 32306, Tallahassee, FL, USA
a r t i c l e i n f o
Article history:Received 26 November 2016Received in revised form27 March 2017Accepted 9 April 2017Available online 14 April 2017
Keywords:EutrophicationLand use and land cover (LULC)Water qualityGovernment-financed
* Corresponding author. Nanjing Institute of GeogrAcademy of Sciences, 73 East Beijing Road, Nanjing, 2** Corresponding author. Nanjing Institute of GeogrAcademy of Sciences, 73 East Beijing Road, Nanjing, 2
E-mail addresses: [email protected] (Y. Zhang)1 These authors contributed equally to this study.
This study highlights how Chinese economic development detrimentally impacted water quality inrecent decades and how this has been improved by enormous investment in environmental remediationfunded by the Chinese government. To our knowledge, this study is the first to describe the variability ofsurface water quality in inland waters in China, the affecting drivers behind the changes, and how thegovernment-financed conservation actions have impacted water quality. Water quality was found to bepoorest in the North and the Northeast China Plain where there is greater coverage of developed land(cities þ cropland), a higher gross domestic product (GDP), and higher population density. There aresignificant positive relationships between the concentration of the annual mean chemical oxygen de-mand (COD) and the percentage of developed land use (cities þ cropland), GDP, and population densityin the individual watersheds (p < 0.001). During the past decade, following Chinese government-financed investments in environmental restoration and reforestation, the water quality of Chineseinland waters has improved markedly, which is particularly evident from the significant and exponen-tially decreasing GDP-normalized COD and ammonium (NH4
þ-N) concentrations. It is evident that theincreasing GDP in China over the past decade did not occur at the continued expense of its inland waterecosystems. This offers hope for the future, also for other industrializing countries, that with appropriateenvironmental investments a high GDP can be reached and maintained, while simultaneously preservinginland aquatic ecosystems, particularly through management of sewage discharge.
Inland waters such as lakes, reservoirs, streams, rivers, etc,
aphy and Limnology, Chinese10008, China.aphy and Limnology, Chinese10008, China., [email protected] (B. Qin).
provide a wide variety of ecosystem services ranging from potablewater, to sources of food, through to transportation and sites forrecreation (Downing et al., 2006). However, these inland waterecosystems are threatened across the planet by the dual pressure ofanthropogenic activities and climate change (Bragazza et al., 2012;Feng et al., 2008; Williamson et al., 2014). Land use intensificationand urbanization have resulted in increased discharge of waste-water from households, agriculture, and industry, resulting in anelevated risk of point and non-point source pollution (Foley et al.,2005; Zhou et al., 2016). Eutrophication is a common
Y. Zhou et al. / Water Research 118 (2017) 152e159 153
consequence of deteriorating water quality and is a state charac-terized by high nutrient levels, low water transparency, andexcessive growth of algal cells (Paerl et al., 2011a, 2011b; Qin et al.,2015). Blooms of toxic and hypoxia-generating cyanobacteria are anindicator of advanced eutrophication and represent a serious threatto potable water supplies and the ecological sustainability of inlandwater ecosystems (Paerl and Huisman, 2008; Paerl and Otten,2013). Therefore, protection of inland waters is required to safe-guard the quality and safety of the myriad of ecosystem servicesthat these waters provided to consumers (Guo, 2007; Paerl andHuisman, 2008). Furthermore, primary production and land usepractices within a watershed impact greenhouse gas emissionsfrom inlandwaters, as for example organic carbon ismetabolized tocarbon dioxide (CO2) and/or methane (CH4) and nitrogenouscompounds are transformed to nitrous oxides (N2O) (Butman andRaymond, 2011; Davidson et al., 2015; Maberly et al., 2012;Weyhenmeyer et al., 2015). Knowledge of impacts on water qual-ity in inland waters, and the associated drivers, are therefore ofgreat importance to obtain a better understanding of their role inthe global carbon cycle (Bianchi, 2011; Li et al., 2016; Maberly et al.,2012; Piao et al., 2009) and to allow for implementation of mea-sures to protect inland freshwater resources (Williamson et al.,2008, 2009).
The water quality of inland waters is determined by numerousfactors such as land use, hydrologic conditions, and anthropogenicactivities (Kellerman et al., 2014; Kothawala et al., 2014). The effectsof the morphological features of a lake, including depth, surfacearea, and water retention time, on trophic levels have been widelyinvestigated (N~oges et al., 2003; N~oges, 2009), and many studieshave been undertaken to elucidate the relationships betweenwaterquality and watershed land use (Carney, 2009; Kothawala et al.,2014; Liu et al., 2011; Müller et al., 1998; Maberly et al., 2003;Stedmon and Markager, 2005; Taranu and Gregory-Eaves, 2008).However, most studies to date have been conducted in lakes withsimilar land use in their watersheds and with comparable lakemorphometric characteristics, and typically have limited spatialand temporal resolution. Furthermore, the majority of the studieshave focused on direct relationships betweenwater quality and thecorresponding flow rate at different time scales (i.e. daily/weekly/monthly) (Chen et al., 2015; Goldman et al., 2014; Guo et al., 2014;Koch et al., 2013; Stedmon andMarkager, 2005; Striegl et al., 2005).Generally, population density and industrial development havebeen considered important driving factors of declining waterquality in inland waters (Duan et al., 2009). The gross domesticproduct (GDP) offers one measure of development that can beanalyzed. However, the population density and economic statisticaldata presented in the limited number of studies linking thesemetrics to water quality are frequently estimated from informationderived from statistical yearbooks published by various adminis-trative units (Duan et al., 2009; Huang et al., 2014). In most cases,though, administrative units do not encompass the whole water-shed. Further investigation into the direct relationships betweenwater quality, human population density and GDP of individualwatersheds is therefore required to assess the role these metricsmay play in determining the health of inland waters.
In China, eutrophication coupled with environmental degrada-tion has been considered as one of the prices paid for the rapideconomic development occurring since the “reform and openingup-policy” was introduced in the late 1970s (Liu and Diamond,2005; Liu et al., 2013). Importantly, in recent years, measures toremedy eutrophication have been introduced such as the NaturalForest Conservation Program (NFCP) and the Grain to Green Pro-gram (GTGP, also known as the Sloping Land Conversion Programand the Farm to Forest Program), the world's largest government-financed conservation action programs whose implementation
was prompted by severe droughts in 1997 and unprecedentedfloods in 1998 (Liu et al., 2008; Ouyang et al., 2016). Furthermore,today, billions of RMB Chinese Yuan are dedicated to ecosystemrestoration via environmental protection actions (e.g. the estab-lishment of wastewater treatment plants). It is yet to be determinedhow these actions affect water quality and whether they haveeffectively reversed anthropogenically induced water qualitydegradation to any extent.
The objective of this study was to investigate the long-termvariations of water quality in Chinese inland waters and to accessthe factors contributing to changes in water quality. An additionalaim was to elucidate how conservation actions introduced by theChinese government in the past two decades have impacted waterquality. To elucidate the dynamics of water quality in Chineseinland waters, weekly data on water quality for the period January2006 to December 2015 (n ¼ 499) were examined for a total of 145sites spanning the entire country (Fig. 1). These water qualitymeasurements included dissolved oxygen (DO), chemical oxygendemand (COD), and ammonium (NH4
þ-N) at all 145 sites. Data onland use and land cover (LULC), population density, and GDP with1-km spatial resolution and 5-year intervals during the past twodecades were analyzed to assess relationships between waterquality and the hypothesized driving factors. Government led in-vestments in areas such as environmental restoration and refor-estation were further investigated to determine how these actionshave impacted the water quality in Chinese inland waters duringthe past two decades.
2. Materials and methods
2.1. Gauged datasets
Water quality, including DO, COD, and NH4þ-N, was determined
on a weekly basis for a total of 145 monitoring sites in major riversand lakes in China from January 2006 to December 2015 (Fig. 1).The data were collected by the China National EnvironmentalMonitoring Center (data available at http://www.cnemc.cn/publish/107/0594/350/newList_1.html). The main indices forchemical indicators of different water quality levels were catego-rized according to the water quality standards for surface waters inChina (GB3838-2002). Level I: COD �2.0 mg L�1; NH4
� DO < 3.0 mg L�1. Water belonging to levels I to III is potable afterconventional water treatment. The highest level of water quality(i.e. the most polluted) parameters (DO, COD, and NH4
þ-N) set thelevel for the site.
Data on annual water quality levels for >11� 104 km rivers from1999 to 2014 were obtained from the Annual Bulletin of WaterResources published by the Ministry of Water Resources of China(http://www.mwr.gov.cn/zwzc/hygb/szygb/).
2.2. Watershed land use, population density, and GDP data
The watershed boundaries of all the 145 study sites weredelineated using a DEM with a resolution of 90 m (available athttp://srtm.csi.cgiar.org/index.asp) in Q-GIS software (Version2.14.3; available at http://www.qgis.org/en/site/). LULC composi-tions in 1995, 2000, 2005, and 2010 of the individual watershedslocated upstream of each water quality monitoring site were
Fig. 1. Location of sampling sites (n ¼ 145, circles) used in this study with corresponding multi-year (2006e2015) weekly mean trophic levels (aed). The faint red-dotted line inpanel “a” delineates the boundaries upstream of the individual monitoring sites. Also shown are the spatial distribution of multi-year (2006e2015) weekly mean dissolved oxygen(DO; e), the chemical oxygen demand (COD; f), and ammonium (NH4
þ-N; g). The three indices are all given in mg L�1. (For interpretation of the references to colour in this figurelegend, the reader is referred to the web version of this article.)
Y. Zhou et al. / Water Research 118 (2017) 152e159154
extracted from 1 km resolution national land cover data for Chinainterpreted from Landsat TM images (available at http://www.resdc.cn/) using ArcGIS 10.1 software. The LULC classes werefurther grouped into seven categories: forest, grassland, water-bodies, cities, bare soil, sea reclamation, and cropland (Liu andBuhe, 2000; Liu et al., 2003, 2005). Population density and GDPdata from 1995, 2000, 2005, and 2010, with 1 km resolution(available at http://www.resdc.cn/), were extracted for individualwatersheds using ArcGIS 10.1 software.
2.3. Investments in water quality protection actions
Data on annual investments in reforestation from 1999 to 2014were obtained from the website of the State Forestry Bureau ofChina (http://www.forestry.gov.cn/Common/index/62.html). Dataon annual investments in environmental restoration (e.g. externalloading reduction and sediment dredging), environmental protec-tion facilities (e.g. the establishment of sewage treatment plants),urban sewer lines, and industrial pollution restoration from 1999 to2014 were obtained from the website of the National Bureau ofStatistics (http://data.stats.gov.cn/).
2.4. Relationships between water quality and the associated drivingfactors
Data on land use and land cover (LULC), population density, andGDP were only available from 1995, 2000, 2005, and 2010, whereasdata on water quality parameters, including COD, NH4
þ-N, and DO,were available from 2006 to 2015, implying that the time series ofanthropogenic and water quality data only matched for 2010. Inorder to determine if the correlation recorded between theanthropogenic and water quality data in 2010 may also apply toother time periods, we used mean COD, NH4
þ-N, and DO values forthe first eight weeks of 2006 to represent 2005 and correlated thesevalues with LULC, population density, and GDP in 2005. Weassumed the changes in LULC, population density, and GDP in Chinain the first eight weeks of 2006 to be insignificant.
2.5. Statistical analyses
Statistical analyses, including mean values, standard deviation,t-test, and linear correlations, were conducted using MATLABR2012a software. Spatial distribution of water quality-related pa-rameters was determined with ArcGIS 10.1 software. Linear fitting
Y. Zhou et al. / Water Research 118 (2017) 152e159 155
and t-test results with p < 0.05 are reported as significant. Meansare given with plus/minus their standard deviations.
3. Results
3.1. General results on water quality
Water quality for the 145 nation-wide sites was divided into fivecategories: I to V, ranging from good to poor (Fig. 1) as described insection 2.1. Spatially, water quality levels were notably poorer(dominated by conditions poorer than the characteristics stipulatedfor level III) in eastern China than in the other regions. This wasparticularly pronounced for the sites in the North China Plain andthe Northeast China Plain (Fig. 1). Accordingly, markedly higherCOD and NH4
þ-N and low DO were recorded here than in theremaining areas (Fig. 1). In comparison, relatively good waterquality was observed in South China where COD and NH4
þ-N levelswere typically low and DO relatively high. In the study period, highannual mean COD concentrations (7 mg L�1) were recorded during2006e2008 at all the 145 monitoring sites, after which theydecreased significantly to approximately 4e5mg L�1 in 2009e2015(p < 0.001; Fig. 2a and b). Similarly, annual mean NH4
þ-N concen-trations decreased from approximately 1.2 mg L�1 in 2006e2009 toapproximately 0.6 mg L�1 in 2010e2015 (p < 0.001; Fig. 2a and b),and mean annual DO concentrations increased markedly fromapproximately 7 mg L�1 in 2006e2009 to approximately 8 mg L�1
in 2010e2015 (p < 0.001; Fig. 2). Notably, annual mean GDP-normalized COD and NH4
þ-N decreased substantially from 2006 to2015 (p< 0.001; Fig. 2c). High CODwas recorded in DecembereMayand low COD in JuneeNovember in 2006e2009, whereas no sig-nificant inter-annual variation occurred during 2010e2015 (Fig. 2aand b). High concentrations of both NH4
þ-N and DO were recordedin NovembereMarch and lower values were observed inAprileOctober (Fig. 2a).
From 1999 to 2014, the riverine water quality was dominated bylevels II, III, and IV and by conditions poorer than the characteristicsstipulated for level V (Fig. 3a). The percentage of rivers categorizedto levels I, IV, and V has not varied significantly during the pasttwenty years. A notable increasing trend was recorded for level IIfrom <30% in 2004e2006 to >40% in 2013e2014 (Fig. 3a). Theproportion of rivers categorized as level III decreased slightly from
Fig. 2. Long-term (2006e2015) variations of monthly mean DO, COD, and NH4þ-N for all the
three indices. Linear fitting over the time series of the annual mean values of the three indicethe past decade (c). Error bars in panels b and c show the standard deviations of monthly
32% in 1999 to 27% in 2002e2010, followed by a further decrease to25% in 2012e2013 (Fig. 3a). The relative proportion of riverine datashowing a water quality poorer than level V increased from 16% in1999e2002 to 22% in 2003e2007 and then decreased markedly to13% in 2014 (Fig. 3a).
3.2. Investments in environmental restoration
Investments in reforestation and total environmental restora-tion (e.g. external loading reduction) have increased considerablyduring the past twenty years (inflation being adjusted), fromalmost nothing in 1994 to as much as 400 billion RMB (US$: 58billion calculated given an exchange rate of 1 US$ ¼ 6.89 RMB yuanwhich was the rate on 21 March, 2017) and 1000 billion RMB yuan(US$: 145 billion), respectively, in 2014 (Fig. 3b). In comparison,investments in environmental protection facilities (e.g. the estab-lishment of sewage treatment plants) and urban sewer systemsincreased notably from 1999 to 2010 and have remained at 500billion (US$: 73 billion) and 100 billion (US$: 15 billion) RMB yuanyr�1, respectively, since then (Fig. 3b). Investments to address in-dustrial pollution have increased from 20 billion (US$: 3 billion) in1999 to 100 billion (US$: 15 billion) in 2014 (Fig. 3b).
3.3. Land use and land cover (LULC) change, population density,and GDP results
Agricultural cropland dominated LULC in the Sichuan Basin,North China Plain, and on the Northeast China Plain (Fig. 4; Fig. S1).Urbanization covered a high proportion of the land area in easternChina, especially in the large river deltas (Fig. 4; Fig. S1). Tempo-rally, the urban percentage of LULC increased from 1.6% in 1990 to2.0% in 2010. The agricultural cropland percentage of LULCremained constant from 18.6% in 1990 to 18.5% in 1995, thenincreased slightly to 18.9% in 2000, and finally decreased slightly to18.8% in 2010 (Fig. 4; Fig. S1). The forest percentage of LULCincreased slightly from 23.4% in 1990 to 23.7% in 1995, followed bya small decline to 23.3% in 2000e2010 (Fig. 4; Fig. S1). Similarly tocropland, a relatively high population density and GDP wererecorded in eastern China, especially in the Sichuan Basin, theNorth China Plain, and the Northeast China Plain (Fig. 4; Fig. S1).GDP has increased rapidly over the past two decades. In the Huai
145 sites (a). Error bars denote the standard deviations of weekly mean values of thes (b) and non-linear fitting over the time series of GDP-normalized COD or NH4
þ-N overmean values of the three indices.
Fig. 3. Annual mean percentages of individual water quality levels for >11 � 104 km rivers covering the entire country (a). Annual investments in reforestation, environmentalrestoration, environmental protection facilities, urban sewer lines, and land remediation after industrial pollution from 1999 to 2014 (b). Relationships between the summedpercentages of water quality levels I þ II þ III and investments in reforestation (c) and with total investments in environmental restoration (d). Relationships between annual meanCOD concentrations at all the 145 sites shown in Fig. 1 and reforestation investments (e) and total investments in environmental restoration (f). Error bars in panels e and f indicatethe standard deviations of monthly mean COD values.
Y. Zhou et al. / Water Research 118 (2017) 152e159156
and Hai watersheds, for example, GDP increased from, respectively,237� 104 (US$: 34� 104) and 205 � 104 (US$: 30� 104) RMB yuankm�2 in 1995 to 1599 � 104 (US$: 232 � 104) and 1230 � 104 (US$:179 � 104) RMB yuan km�2 in 2010 (Fig. 4; Fig. S1). However,population density has remained stable in the two watersheds forthe past two decades (Fig. S1).
3.4. Relationships between water quality and LULC, populationdensity, and GDP
A significant positive relationship was observed between annualmean COD concentrations and the percentage of cities of LULC inthe upstream areas of all water quality monitoring sites (p < 0.001;Fig. 5a). Similarly, significant positive relationships were foundbetween annual mean NH4
þ-N concentrations and the percentage ofcities of LULC (p < 0.01; Fig. S2). Furthermore, the greater theportion of developed land use (the percentages of cities þ croplandof LULC) in the individual watersheds, the greater was the annualmean COD concentrations (p < 0.001; Fig. 5b). A comparable rela-tionship was recorded between the percentage of cities þ croplandof LULC and annual mean NH4
þ-N concentrations (p < 0.01; Fig. S2).In comparison, no significant relationship was observed betweenannual mean DO and the percentage of cities of LULC or developedland use (cities þ cropland) in the individual watersheds (p > 0.05;Fig. S3).
Significant positive correlations were traced between annualmean COD and population density and between annual mean CODand GDP of the individual watersheds upstream of each waterquality monitoring site (p < 0.001; Fig. 5c and d). In contrast, nosignificant relationship was observed between annual mean DO,NH4
þ-N, and population density or between annual mean DO, NH4þ-
N, and GDP (p > 0.05; Fig. S2; Fig. S3).
3.5. Relationships between water quality and environmentalinvestments
Significant positive relationships were detected between the
summed percentages of water quality levels I þ II þ III in theriverine data and investments in reforestation (r2 ¼ 0.66, p < 0.001;Fig. 3c) and between the summed percentages of water qualitylevels I þ II þ III in the river data and total investments in envi-ronmental restoration (r2 ¼ 0.46, p < 0.001; Fig. 3d). Significantnegative relationships were recorded between annual mean CODconcentrations at all the 145 sampling sites and investments inreforestation (r2 ¼ 0.73, p < 0.001; Fig. 3e) and between annualmean COD concentrations and total investments in environmentalrestoration (r2 ¼ 0.84, p < 0.001; Fig. 3f).
4. Discussion
Our results demonstrate markedly higher COD and NH4þ-N
levels, lower DO concentrations, and a correspondingly moredegraded water quality index in the North China Plain and theNortheast China Plain than in the remaining areas of China. This canprimarily be explained by an elevated proportion of developed landuse (cities þ cropland), higher GDP, and greater population density(Fig. 4; Fig. S1), as highlighted by the relationships between COD,NH4
þ-N, and the percentage of cities of LULC, developed land use(cities þ cropland), GDP, and the population density of the indi-vidual watersheds (Fig. 5; Fig. S2). Land use and land cover canimpact the watershed export of dissolved organic matter andinorganic nutrients (Kellerman et al., 2014; Kothawala et al., 2014;Le et al., 2015; Spencer et al., 2013). This is supported by the rela-tively highermulti-yearmean COD andNH4
þ-N and lowDO found atmetropolitan sites than in peri-urban and rural areas (Fig. 1; Fig. 2).Secondly, the notably lower precipitation in North China than inSouth China (Piao et al., 2010) means that flushing and dilutioneffects are more pronounced in the latter (Paerl and Huisman,2008). In North China, water diversion projects coupled withrelatively high evaporation create protracted periods of drought inspring and early summer, prolonging the water retention time andamplifying the concentration effect of degraded waters in this re-gion (Piao et al., 2010).
An improvement in water quality of Chinese inland waters has,
Fig. 4. Land use and land cover (LULC, left), GDP (in RMB 104 Yuan km�2, middle), and population density (in ca. km�2, right) in China in 1995 (upper panels), 2000 (middle panels),and 2010 (lower panels).
Y. Zhou et al. / Water Research 118 (2017) 152e159 157
however, been observed during the past decade as evidenced bydecreasing annual mean COD and NH4
þ-N concentrations andincreasing DO at the 145 nation-wide monitoring sites (Fig. 2b).This improvement is exemplified in the exponentially decreasingGDP-normalized COD and NH4
þ-N (Fig. 2c). The increasing per-centages of the summed water quality levels I þ II þ III and thedecreasing percentage of a water quality poorer than the charac-teristics stipulated for level V (Fig. 3a) recorded for the>11 � 104 km rivers nation-wide provide further evidence of this.The improved water quality can be attributed to several factors.Firstly, investments in environmental protection, restoration, andreforestation have increased markedly since 1999 (Fig. 3b). Theenhanced investments in wastewater treatment plants haveresulted in an increase in the extent of areas serviced by dual waterdistribution systems and improved sewage treatment capacity(Yang et al., 2006). The apparent relationships between the in-vestments in reforestation and environmental restoration andimproved water quality (Fig. 3cef) highlight that government-
financed environmental actions have markedly improved the wa-ter quality (Liu et al., 2008; Ouyang et al., 2016). Secondly, enhancedwater retention due to damming of reservoirs can reduce thetransport of sediment downstream (Wang et al., 2011, 2015; Yanget al., 2014), and a slightly increasing percentage of forest and areduction of cropland LULC (Fig. 3; Fig. 4) (Hu et al., 2015; Yu et al.,2015), has somewhat limited soil erosion especially in south-western China, the Loess Plateau, and the middle reaches of YellowRiver during the past two decades (Wang et al., 2011, 2015; Yanget al., 2014). This may have contributed to the observed improve-ments in water quality and dissolved oxygen concentrations assuspended sediment has been considered a significant source ofparticulate organic carbon (Bilotta and Brazier, 2008; Galy et al.,2015). More sustainable land use practices may also increase theretention time of polluted or eutrophic water, leading to microbialand photochemical degradation. Finally, expanded wastewatertreatment capability is also likely to have contributed to the waterquality improvement over the past decade (Yan and Xu, 2014). This
Fig. 5. Correlations between chemical oxygen demand (COD) and (a) percentages of cities of LULC, (b) percentages of cities and cropland of LULC, (c) Gross Domestic Product (GDP),and (d) population density. Error bars represent COD standard deviations.
Y. Zhou et al. / Water Research 118 (2017) 152e159158
concurs with the increased investments in environmental resto-ration, especially urban sewer lines and wastewater treatmentplants (Fig. 3).
Our results demonstrate a positive effect of the extensive in-vestments in environmental restoration made by the Chinesegovernment (Fig. 3). Exponentially decreasing GDP-normalizedCOD and NH4
þ-N concentrations (Fig. 2c) in fresh waters haveoccurred despite a marked increase in GDP during the past decade.In the coming decades, external loading reduction, created by, forexample, higher expanded wastewater treatment capacity and theinitiatives embedded in the Grain for Green Project may furtherimprove the water quality in the receiving waterbodies (Jeppesenet al., 2005a, 2007, 2005b; Paerl et al., 2011a; Paerl et al., 2011b;Xu et al., 2010).
5. Conclusions
(1) Our results demonstrate that the increasing GDP in Chinaover the past decade did not occur at the further expense ofits inland water ecosystems. This is particularly evident fromthe significant and exponentially decreasing GDP-normalized COD and ammonium (NH4
þ-N) concentrationsthroughout the country, coinciding with increasinggovernment-financed environmental investments duringthe past two decades.
(2) This offers hope for the future, also for other industrializingcountries, that with significant environmental investments ahigh GDP can be reached and maintained, while simulta-neously restoring and preserving inland aquatic ecosystems,particularly through appropriate management of sewagedischarge.
Acknowledgments
This work was supported by the National Natural ScienceFoundation of China (grants 41621002, 41601537, 41325001, and
41230744) and International Scientific Cooperation Project(2014DFG91780). Erik Jeppesenwas supported by theMARS project(Managing Aquatic ecosystems and water Resources under multi-ple Stress) funded under the 7th EU Framework Programme, Theme6 (Environment including Climate Change), Contract No.: 603378(http://www.mars-project.eu), ‘CLEAR’ (a Villum Kann Center ofExcellence project), and PROGNOS (Predicting in-lake RespOnses tochanGe using Near real time MOdelS e Water Joint ProgrammeInitiative). Wewould like to express our deep thanks to AnneMettePoulsen from Aarhus University for editorial assistance. The dataused in the paper will be available upon request (e-mail: [email protected] or [email protected]). The authors are grateful to theeditor and the anonymous reviewers for their helpful commentsand suggestions.
Appendix A. Supplementary data
Supplementary data related to this article can be found at http://dx.doi.org/10.1016/j.watres.2017.04.035.
References
Bianchi, T.S., 2011. The role of terrestrially derived organic carbon in the coastalocean: a changing paradigm and the priming effect. Proc. Natl. Acad. Sci. 108(49), 19473e19481.
Bilotta, G., Brazier, R., 2008. Understanding the influence of suspended solids onwater quality and aquatic biota. Water Res. 42 (12), 2849e2861.
Bragazza, L., Parisod, J., Buttler, A., Bardgett, R.D., 2012. Biogeochemical plantesoilmicrobe feedback in response to climate warming in peatlands. Nat. Clim.Change 3 (3), 273e277.
Butman, D., Raymond, P.A., 2011. Significant efflux of carbon dioxide from streamsand rivers in the United States. Nat. Geosci. 4 (12), 839e842.
Carney, E., 2009. Relative influence of lake age and watershed land use on trophicstate and water quality of artificial lakes in Kansas. Lake Reserv. Manag. 25 (2),199e207.
Chen, W., Habibul, N., Liu, X.Y., Sheng, G.P., Yu, H.Q., 2015. FTIR and synchronousfluorescence heterospectral two-dimensional correlation analyses on thebinding characteristics of copper onto dissolved organic matter. Environ. Sci.Technol. 49 (4), 2052e2058.
Davidson, T.A., Audet, J., Svenning, J.C., Lauridsen, T.L., Sondergaard, M.,Landkildehus, F., Larsen, S.E., Jeppesen, E., 2015. Eutrophication effects on
Galy, V., Peucker-Ehrenbrink, B., Eglinton, T., 2015. Global carbon export from theterrestrial biosphere controlled by erosion. Nature 521 (7551), 204e207.
Goldman, J.H., Rounds, S.A., Keith, M.K., Sobieszczyk, S., 2014. Investigating organicmatter in Fanno Creek, Oregon, Part 3 of 3: identifying and quantifying sourcesof organic matter to an urban stream. J. Hydrology 519, 3028e3041.
Guo, L., 2007. Doing battle with the green monster of Taihu Lake. Science 317(5842), 1166.
Guo, W., Yang, L., Zhai, W., Chen, W., Osburn, C.L., Huang, X., Li, Y., 2014. Runoff-mediated seasonal oscillation in the dynamics of dissolved organic matter indifferent branches of a large bifurcated estuary-The Changjiang Estuary.J. Geophys. Res. Biogeosciences 119 (5), 776e793.
Hu, Y., Zhang, X., Mao, R., Gong, D., Liu, H., Yang, J., 2015. Modeled responses ofsummer climate to realistic land use/cover changes from the 1980s to the 2000sover eastern China. J. Geophys. Res. Atmos. 120 (1), 167e179.
Huang, C., Wang, X., Yang, H., Li, Y., Wang, Y., Chen, X., Xu, L., 2014. Satellite dataregarding the eutrophication response to human activities in the plateau lakeDianchi in China from 1974 to 2009. Sci. Total Environ. 485e486, 1e11.
Jeppesen, E., Jensen, J.P., Sondergaard, M., Lauridsen, T.L., 2005a. Response of fishand plankton to nutrient loading reduction in eight shallow Danish lakes withspecial emphasis on seasonal dynamics. Freshw. Biol. 50 (10), 1616e1627.
Jeppesen, E., Søndergaard, M., Meerhoff, M., Lauridsen, T.L., Jensen, J.P., 2007.Shallow lake restoration by nutrient loading reductiondsome recent findingsand challenges ahead. Hydrobiologia 584 (1), 239e252.
Jeppesen, E., Sondergaard, M., Jensen, J.P., Havens, K.E., Anneville, O., Carvalho, L.,Coveney, M.F., Deneke, R., Dokulil, M.T., Foy, B.O.B., Gerdeaux, D., Hampton, S.E.,Hilt, S., Kangur, K., Kohler, J.A.N., Lammens, E.H.H.R., Lauridsen, T.L., Manca, M.,Miracle, M.R., Moss, B., Noges, P., Persson, G., Phillips, G., Portielje, R.O.B.,Romo, S., Schelske, C.L., Straile, D., Tatrai, I., Willen, E.V.A., Winder, M., 2005b.Lake responses to reduced nutrient loading - an analysis of contemporary long-term data from 35 case studies. Freshw. Biol. 50 (10), 1747e1771.
Kellerman, A.M., Dittmar, T., Kothawala, D.N., Tranvik, L.J., 2014. Chemodiversity ofdissolved organic matter in lakes driven by climate and hydrology. Nat. Com-mun. 5, 3804.
Koch, J.C., Runkel, R.L., Striegl, R., McKnight, D.M., 2013. Hydrologic controls on thetransport and cycling of carbon and nitrogen in a boreal catchment underlainby continuous permafrost. J. Geophys. Research-Biogeosciences 118 (2),698e712.
Kothawala, D.N., Stedmon, C.A., Muller, R.A., Weyhenmeyer, G.A., Kohler, S.J.,Tranvik, L.J., 2014. Controls of dissolved organic matter quality: evidence from alarge-scale boreal lake survey. Glob. Change Biol. 20 (4), 1101e1114.
Le, C., Lehrter, J.C., Hu, C., Schaeffer, B., MacIntyre, H., Hagy, J.D., Beddick, D.L., 2015.Relation between inherent optical properties and land use and land coveracross Gulf Coast estuaries. Limnol. Oceanogr. 60 (3), 920e933.
Li, B., Gasser, T., Ciais, P., Piao, S., Tao, S., Balkanski, Y., Hauglustaine, D., Boisier, J.P.,Chen, Z., Huang, M., Li, L.Z., Li, Y., Liu, H., Liu, J., Peng, S., Shen, Z., Sun, Z.,Wang, R., Wang, T., Yin, G., Yin, Y., Zeng, H., Zeng, Z., Zhou, F., 2016. Thecontribution of China's emissions to global climate forcing. Nature 531 (7594),357e361.
Liu, J., Buhe, A., 2000. Study on spatial-temporal feature of modern land-use changein China: using remote sensing techniques. Quat. Sci. 20 (3), 229e239.
Liu, J., Diamond, J., 2005. China's environment in a globalizing world. Nature 435(7046), 1179e1186.
Liu, J., Li, S., Ouyang, Z., Tam, C., Chen, X., 2008. Ecological and socioeconomic effectsof China's policies for ecosystem services. Proc. Natl. Acad. Sci. U. S. A. 105 (28),9477e9482.
Liu, J., Liu, M., Tian, H., Zhuang, D., Zhang, Z., Zhang, W., Tang, X., Deng, X., 2005.Spatial and temporal patterns of China's cropland during 1990e2000: ananalysis based on Landsat TM data. Remote Sens. Environ. 98 (4), 442e456.
Liu, J., Liu, M., Zhuang, D., Zhang, Z., Deng, X., 2003. Study on spatial pattern of land-use change in China during 1995e2000. Sci. China Ser. D Earth Sci. 46 (4),373e384.
Liu, W., Zhang, Q., Liu, G., 2011. Effects of watershed land use and lake morphometryon the trophic state of Chinese lakes: implications for eutrophication control.Clean. - Soil, Air, Water 39 (1), 35e42.
Liu, X., Zhang, Y., Han, W., Tang, A., Shen, J., Cui, Z., Vitousek, P., Erisman, J.W.,Goulding, K., Christie, P., Fangmeier, A., Zhang, F., 2013. Enhanced nitrogendeposition over China. Nature 494 (7438), 459e462.
Müller, B., Lotter, A.F., Sturm, M., Ammann, A., 1998. Influence of catchment quality
and altitude on the water and sediment composition of 68 small lakes inCentral Europe. Aquat. Sci. 60 (4), 316e337.
Maberly, S.C., Barker, P.A., Stott, A.W., De Ville, M.M., 2012. Catchment productivitycontrols CO2 emissions from lakes. Nat. Clim. Change 3 (4), 391e394.
Maberly, S.C., King, L., Gibson, C.E., May, L., Jones, R.I., Dent, M.M., Jordan, C., 2003.Linking nutrient limitation and water chemistry in upland lakes to catchmentcharacteristics. Hydrobiologia 506 (1e3), 83e91.
N~oges, P., N~oges, T., Tuvikene, L., Smal, H., Ligeza, S., Kornij�ow, R., Peczula, W.,B�ecares, E., Garcia-Criado, F., Alvarez-Carrera, C., 2003. Factors controllinghydrochemical and trophic state variables in 86 shallow lakes in Europe.Hydrobiologia 506 (1e3), 51e58.
N~oges, T., 2009. Relationships between morphometry, geographic location andwater quality parameters of European lakes. Hydrobiologia 633 (1), 33e43.
Ouyang, Z., Zheng, H., Xiao, Y., Polasky, S., Liu, J., Xu, W., Wang, Q., Zhang, L., Xiao, Y.,Rao, E., 2016. Improvements in ecosystem services from investments in naturalcapital. Science 352 (6292), 1455e1459.
Paerl, H.W., Hall, N.S., Calandrino, E.S., 2011a. Controlling harmful cyanobacterialblooms in a world experiencing anthropogenic and climatic-induced change.Sci. Total Environ. 409 (10), 1739e1745.
Paerl, H.W., Huisman, J., 2008. Climate. Blooms like it hot. Science 320, 57e58.Paerl, H.W., Otten, T.G., 2013. Environmental science. Blooms bite the hand that
Controlling harmful cyanobacterial blooms in a hyper-eutrophic lake (LakeTaihu, China): the need for a dual nutrient (N & P) management strategy. WaterRes. 45 (5), 1973e1983.
Piao, S., Ciais, P., Huang, Y., Shen, Z., Peng, S., Li, J., Zhou, L., Liu, H., Ma, Y., Ding, Y.,Friedlingstein, P., Liu, C., Tan, K., Yu, Y., Zhang, T., Fang, J., 2010. The impacts ofclimate change on water resources and agriculture in China. Nature 467 (7311),43e51.
Piao, S., Fang, J., Ciais, P., Peylin, P., Huang, Y., Sitch, S., Wang, T., 2009. The carbonbalance of terrestrial ecosystems in China. Nature 458 (7241), 1009e1013.
Qin, B., Li, W., Zhu, G., Zhang, Y., Wu, T., Gao, G., 2015. Cyanobacterial bloommanagement through integrated monitoring and forecasting in large shalloweutrophic Lake Taihu (China). J. Hazard. Mater. 287, 356e363.
Spencer, R.G.M., Aiken, G.R., Dornblaser, M.M., Butler, K.D., Holmes, R.M., Fiske, G.,Mann, P.J., Stubbins, A., 2013. Chromophoric dissolved organic matter exportfrom U.S. rivers. Geophys. Res. Lett. 40 (8), 1575e1579.
Stedmon, C.A., Markager, S., 2005. Resolving the variability in dissolved organicmatter fluorescence in a temperate estuary and its catchment using PARAFACanalysis. Limnol. Oceanogr. 50 (2), 686e697.
Striegl, R.G., Aiken, G.R., Dornblaser, M.M., Raymond, P.A., Wickland, K.P., 2005.A decrease in discharge-normalized DOC export by the Yukon River duringsummer through autumn. Geophys. Res. Lett. 32 (21), L21413.
Taranu, Z.E., Gregory-Eaves, I., 2008. Quantifying relationships among phosphorus,agriculture, and lake depth at an inter-regional scale. Ecosystems 11 (5),715e725.
Wang, H., Saito, Y., Zhang, Y., Bi, N., Sun, X., Yang, Z., 2011. Recent changes ofsediment flux to the western Pacific Ocean from major rivers in East andSoutheast Asia. Earth-Science Rev. 108 (1), 80e100.
Wang, S., Fu, B., Piao, S., Lü, Y., Ciais, P., Feng, X., Wang, Y., 2015. Reduced sedimenttransport in the Yellow River due to anthropogenic changes. Nat. Geosci. 9 (1),38e41.
Weyhenmeyer, G.A., Kosten, S., Wallin, M.B., Tranvik, L.J., Jeppesen, E., Roland, F.,2015. Significant fraction of CO2 emissions from boreal lakes derived fromhydrologic inorganic carbon inputs. Nat. Geosci. 8, 933e936.
Williamson, C.E., Brentrup, J.A., Zhang, J., Renwick, W.H., Hargreaves, B.R., Knoll, L.B.,Overholt, E.P., Rose, K.C., 2014. Lakes as sensors in the landscape: optical metricsas scalable sentinel responses to climate change. Limnol. Oceanogr. 59 (3),840e850.
Williamson, C.E., Dodds, W., Kratz, T.K., Palmer, M.A., 2008. Lakes and streams assentinels of environmental change in terrestrial and atmospheric processes.Front. Ecol. Environ. 6 (5), 247e254.
Williamson, C.E., Saros, J.E., Vincent, W.F., Smold, J.P., 2009. Lakes and reservoirs assentinels, integrators, and regulators of climate change. Limnol. Oceanogr. 54(6), 2273e2282.
Xu, H., Paerl, H.W., Qin, B.Q., Zhu, G.W., Gao, G., 2010. Nitrogen and phosphorusinputs control phytoplankton growth in eutrophic Lake Taihu, China. Limnol.Oceanogr. 55 (1), 420e432.
Yan, Y., Xu, J., 2014. Improving winter performance of constructed wetlands forwastewater treatment in northern China: a review. Wetlands 34 (2), 243e253.
Yang, C.P., Shen, Q.S., Chen, H., Zeng, G.M., Zhong, Y.Y., 2006. Dual water distributionsystems in China. In: 8th Annual Water Distribution Systems Analysis Sympo-sium, Cincinnati, Ohio, USA.
Yang, S.L., Milliman, J.D., Xu, K.H., Deng, B., Zhang, X.Y., Luo, X.X., 2014. Downstreamsedimentary and geomorphic impacts of the three Gorges dam on the Yangtzeriver. Earth-Science Rev. 138, 469e486.
Yu, P., Wang, Y., Coles, N., Xiong, W., Xu, L., 2015. Simulation of runoff changescaused by cropland to forest conversion in the upper Yangtze River region, SWChina. PLoS One 10 (7), e0132395.
Zhou, Y., Jeppesen, E., Zhang, Y., Shi, K., Liu, X., Zhu, G., 2016. Dissolved organicmatter fluorescence at wavelength 275/342 nm as a key indicator for detectionof point-source contamination in a large Chinese drinking water lake. Che-mosphere 144, 503e509.
![ExFuse: Enhancing Feature Fusion for Semantic Segmentationstatic.tongtianta.site/paper_pdf/05bd3984-c228-11e8-bbe7-00163e08bb86.pdf · tion”, which derives from [30,1] in super](https://static.documents.pub/doc/80x56/5e7beee63066dd18ec00753d/exfuse-enhancing-feature-fusion-for-semantic-tiona-which-derives-from-301.jpg)
