Catena 75 (2008) 216–222

Contents lists available at ScienceDirect

Catena

journal homepage: www.elsevier.com/locate/catena

Water quality in relation to land use and land cover in the upper Basin,

Siyue Li a,b, Sheng Gu a, Wenzhi Liu a,b, Hongyin Han c, Quanfa Zhang a,⁎ a Center for Watershed Ecology, Wuhan Botanical Garden, the Chinese Academy of Sciences, Wuhan 430074, PR China b Graduate School of the Chinese Academy of Sciences, 100049, PR China c Key Laboratory of Biogeology and Environmental Geology of Ministry of Education, China University of Geosciences, Wuhan 430074, PR China

ARTICLE INFO ABSTRACT

Article history: A total of 42 sampling sites were selected in the riverine network of the upper Han River basin Received 22 January 2008 (approximately 95, 200 km2) of China. Over the time period of 2005–2006, 252 water samples were collected Received in revised form 17 June 2008 and analyzed for physico-chemical variables in order to investigate their spatio-temporal variability in Accepted 19 June 2008 particular the relationship with land use and land cover. Analysis of variance (ANOVA) indicated significant spatial variability in pH, EC, TDS, turbidity, SPM, ORP and nitrogen across the basin. Meanwhile, nitrogen, Keywords: ORP, I and turbidity generally displayed higher values in the rainy season. Correlation analysis and Water quality Mn fi Land use and Land cover regression analysis indicated that water temperature, IMn, and nitrogen were signi cantly related to Han River vegetated coverage, and subwatersheds with higher vegetation cover had relative lower turbidity, SPM, IMn, Physico-chemicals nutrients and TDS. Bare lands had significant influence on nitrogen concentration in the riverine network, implying its large geologic sources in the basin. Percentage of urban area was the predictor for pH and DP,

while agricultural land for SPM and IMn. The research could provide critical information in sustainable land use practice for water resource conservation for the basin. © 2008 Elsevier B.V. All rights reserved.

1. Introduction influenced water nutrients as well (Osborne and Wiley, 1988; Sliva and Williams, 2001). Stream water chemistry is controlled by numerous natural and In 2002, China implemented three-route (East, Middle and West) anthropogenic factors (Ahearn et al., 2005). Their effects on hydro- South to North water Transfer Project (SNWTP) with a capacity of chemistry can either be diffuses (e.g., runoff from urban and crop transferring a total of 44.8 billion m3 of water annually from the cultivation, interflow through organic rich soils) or point pollutants River and its tributaries to drought North China and (e.g., industrial effluents) (Sliva and Williams, 2001; Li et al., 2008). Northwest China (Zhang, 2005). The Han River is the water source Also, watershed characteristics including topography and surficial area of its Middle Route supplying 13.8 billion m3 of water annually to geology can influence surface water quality (Sliva and Williams, 2001). the North China including Tianjin and Beijing City for domestic, In recent years there is a rapid declining availability of usable industrial and irrigation purposes (Li et al., in press). Since late 1980s, freshwater in terms of water quality and quantity due to unsustainable diffuse sources have been dramatically increasing due to intensive land use practices (Ngoye and Machiwa, 2004). anthropogenic activities (Wang et al., 2006) and the Han River has Water quality is generally linked to land use/land cover (LULC) in become one of the most nitrogen-contaminated tributaries of the catchment (Ahearn et al., 2005), and studies have been focusing on Yangtze River (Liu et al., 2003; Li et al., in press) with moderate metal their relationships with water quality variables such as dissolved salts, contamination (Li et al., 2008). The study presents the spatio-temporal suspended solid, and nutrients (Hill, 1981; Allan et al., 1997; Johnson variability of physico-chemicals and the impact of LULC on water et al., 1997; Osborne and Wiley, 1988; Smart et al., 1998; Sliva and quality, and it would help water conservation in the upper Han River Williams, 2001; Turner and Rabalais, 2003; Ahearn et al., 2005). They basin for the interbasin water transfer project. conclude that agricultural land use strongly influences nitrogen (Johnson et al., 1997; Smart et al., 1998; Ahearn et al., 2005), 2. Study area phosphorus (Hill, 1981), and sediments (Allan et al., 1997; Johnson et al., 1997; Ahearn et al., 2005) in stream water. Urban land use has The Han River, the largest tributary of the Yangtze River, originates from Ningqiang County of province, and covers approxi- mately 159×103 km2 with a total length of 1577 km (Yang et al., 1997; – – ⁎ Corresponding author. Tel.: +86 27 87510702; fax: +86 27 87510251. Shen and Liu, 1998). Its upper reaches (31°20' 34°10' N, 106° 112° E; E-mail address: [email protected] (Q. Zhang). 210–3500 m a.s.l) are located in the mountainous region, with

0341-8162/$ – see front matter © 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.catena.2008.06.005 S. Li et al. / Catena 75 (2008) 216–222 217 drainage area of approximately 95.2×103 103 km2 and 925 km long accounting for 70% of the total runoff of the whole basin with large (Shen and Liu, 1998; Fig. 1). The upper basin is located in the north annual and interannual variability (Jin and Guo, 1993; Yang et al., sub-tropic monsoon climatic region. Annual mean precipitation is 1997).The discharge from the Reservoir is about 23.0×109 m3/year 700–1800 mm, of which about 80% falls in the time period from May after the SNWTP, and the peak discharge reaches about 14,000 m3/s in to October (Yang et al., 1997). The annual mean runoff is 41.1×109 m3, 2005 in the past 30 years.

Fig. 1. The location of sampling sites, DEM, river network, and land use and land cover of the upper Han River, China (Zone 1—Laoguan River, Zone 2—Dan River, Zone 3—South of the Mountains, Zone 4—Ziwu River, Zone 5—Hanzhong Plain, Zone 6—North of the Daba Mountains, Zone 7—Ankang Plain, Zone 8—Du River, and Zone 9— Reservoir). 218 S. Li et al. / Catena 75 (2008) 216–222

Table 1 demand (BOD) and potassium permanganate index (IMn)were Description of zones in the upper Han River basin, China analyzed using dilution and seeding method and potassium perman- Zone Sub watershed Watershed characteristics ganate index method, respectively (CSEPB, 2002). Dissolved phos- 1 Laoguan River Intensive human influence including agriculture in buffer phorus (DP) was detected by inductively coupled plasma atomic zone and mineral exploration, belongs to the Dan River emission spectrometer (ICP-AES) (IRIS Intrepid II XSP DUO, USA). basin Quality control procedures, including internal quality control using 2 Dan River The highest bare land type and large landscape variation reference materials and regular participation in interlaboratory 3 South of the Qinling A forested area with lowest urbanization Mountains comparisons, were employed to monitor the validity of the test results. 4 Ziwu River A forest area including coniferous upland with little human influences, showing granite geology 3.2. Land use and land cover analysis 5 Hanzhong Plain The primary agricultural area, and cultivation like rice and maize is carried out 10–100 m from rivers, with higher Landsat Thematic Mapper imagery from 2000 was used to map LULC urbanization (Hanzhong city) in the basin (Shen et al., 2006). The five land cover categories are 6 North of the Daba A primary area for minerals with relative higher Mountains agricultural land area compared to the South of the (1) vegetated land, including coniferous, deciduous forest, mixed co- Qinling Mountains niferous and broad-leaved forest, thickets and herb; (2) agriculture, 7 Ankang Plain The highest agriculture land use area, with intensive including paddy field and dry land; (3) urban, including industrial and cultivation in buffer zones; rivers with dams discharge residential areas; (4) waters, including rivers, wetland and sandy beach; water before flood season 8 Du River The largest catchment in the upper Han River basin with and (5) bare lands, including gravels, bare ground and bare rocks. ArcGIS serial impoundments for hydropower 8.3 Desktop GIS software was used to determine the composition of LULC. 9 Danjiangkou An industrial area (Shiyan city) where industrial effluents Reservoir drain into the rivers, impacting the water quality directly 3.3. Statistical analyses in the Reservoir Because the upper Han River basin has a large land area of 2 The vegetation in the basin is dominated by coniferous, deciduous 95,000 km , and the 42 water sampling sites were located across a forest, mixed coniferous and broad-leaved forest, shrub and herb range of land uses, geology types, and stream orders in the entire upper fi (Shen et al., 2006). Agricultural lands concentrates in low elevation, Han River (Fig. 1). Thus, the upper basin was rst delineated into i.e., Hanzhong plain (zone 5; Fig. 1; Shen et al., 2006). The main subwatersheds using DEM (O'Callaghan and Mark,1984; Gu et al., 2007). crops include maize (Zea mays Linn), wheat (Triticum aestivum Linn), Considering their similarity of land use and land cover (Shen et al., rice (Oryza sativa Linn), cassava (Manihot esculenta), vegetables 2006), water quality, and human activities, the delineated subwater- including radish (Raphanus sativus L.var.radiculus Pers.), cucumber sheds were consequently clustered into 9 zones (Table 1, Fig.1). Analyses (Cucumis sativus L.), tabasco (Capsicum L.) and citrus (Citrus). A very in this study were conducted in the zonal level. small area is utilized for human settlements ranging from small towns Relationships among the considered variables were tested using fi b to villages (Shen et al., 2006), while human population is sparse in Pearson's correlation with statistical signi cance set priori at p 0.05. uplands. Large cities including , Shiyan, Hanzhong and Analysis of variance (ANOVA) was used to compare variations in water fi b Ankang city are located along the Han River corridor. Soil is composed quality under different land uses with signi cance set at p 0.05 fi of yellow brown soil and cinnamon soil (Cai et al., 2000), mapping to (least-signi cance difference, LSD). All the statistical analyses were Ferric Luvisol and Haplic Greyxems respectively by FAO (1971). performed using SPSS 13.0 for windows.

3. Methods 4. Results

3.1. Water sampling and analysis 4.1. Land use and land cover distribution

Six field surveys were conducted in June, August and November Vegetated land was the dominant land cover type in the basin 2005, and April, June and October 2006. Of which, August and (Fig. 2) and it covered from 71.2% (zone 9, Danjiangkou Reservoir) to November in 2005 and October in 2006 are rainy season, and the rest is dry season. A total of 252 grab samples were collected in the 42 sites of the upper Han River basin (Fig. 1). All sampling bottles were soaked for 48 h in 15% nitric acid and subsequently rinsed twice in distilled water prior to use. Water samples were collected at a depth of approximate 10 cm using previously acid-washed high density polyethylene (HDPE) 1 l bottles. The containers were rinsed thrice with sample water on site. A 500 ml subsample was filtered through a previously acid-washed 0.45 µm pore Millipore nitrocellulose membrane filter and the filters were stored in HDPE bottles. Acid-cleaned polyethylene gloves were used while handling all plastic and glass ware. Water temperature (T), dissolved oxygen (DO), pH, oxidation– reduction potential (ORP), electrical conductivity (EC), turbidity, total − dissolved solids (TDS), nitrate–nitrogen (NO3–N) and ammonium– + nitrogen (NH4–N) were determined on site using YSI 6920. The instrument was calibrated at 0 and 100% oxygen saturation before and after usage for DO measurement. The pH sonde was calibrated at 7 and 10, turbidity sonde at 123 and 0, and the nitrogen sondes were both calibrated at 100 mg l− 1 and 1 mg l− 1 before sampling. The membranes Fig. 2. The composition of land use and land cover in the upper Han River basin, China used for filtration were dried at 63 °C to constant mass, and suspended (the area is 4180 km2 for zone 1, 11,300 km2 for zone 2, 15,700 km2 for zone 3, 4030 km2 particulate matter (SPM) was calculated from the difference in the for zone 4, 18900 km2 for zone 5, 9230 for zone 6, 8880 km2 for zone 7, 12,500 km2 for filter paper weights before and after filtering. Biochemical oxygen zone 8 and 9940 km2 for zone 9, respectively). S. Li et al. / Catena 75 (2008) 216–222 219

95.7% (zone 4, Ziwu River) of its respective land area (Fig. 2). Although 4.2. Physical–chemical water quality in the upper Han River basin agricultural widely distributedacrossthebasinbutitmainly concentrated in zones 5–9, comprising from 13.5% to 21.1% of its Spatial and temporal variations of physico-chemicals were land area. Urban occupied less than 1% of the watershed area except in shown in Fig. 3. Nitrogen, ORP, IMn and turbidity generally displayed zone 5 (i.e., Hanzhong Plain) with 1.2%. Bare land had higher higher values in the rainy season, while higher values for T,BOD, percentage in zones 1, 2, and 9 and it ranged from 5.9% to 9.2% (Fig. 2). DP, EC, TDS, SPM, pH and DO in the dry season. Most water quality

Fig. 3. The physical–chemicals (mean±SEM) in the upper Han River basin, China (the different letters indicate statistical difference of total averages among zones at pb0.05; LSD test). 220 S. Li et al. / Catena 75 (2008) 216–222

Fig. 3 ( continued ).

parameters except T,DO,IMn, BOD, and DP showed significant Hanzhong and Shiyan were located. Also, zone 9 displayed the + − spatial variations (p b0.05). The lower values of pH and higher highest NH4–N concentration. EC, TDS and NO3–N had the lowest concentration of DP occurred in zones 5 and 9, where large cities values in zone 4, an area with the highest vegetation cover (Fig. 2).

Table 2 Pearson correlation coefficients between LULC and water quality parameters in the upper Han River basin, China

2 Vegetation Agriculture Urban Waters Bare land Area (km ) Table 3 T −0.747a 0.354 0.118 0.405 0.745a −0.095 Stepwise multiple regression models for physico-chemicals and LULC in the upper Han pH 0.457 −0.394 −0.731a −0.543 −0.055 −0.501 River basin, China EC −0.522 0.256 0.169 −0.059 0.621 0.291 2 2 TDS −0.521 0.255 0.169 −0.059 0.621 0.290 Physico- Independent Regression equations R Adjusted R Turbidity −0.042 0.442 −0.489 −0.166 −0.394 0.175 chemicals variables SPM −0.210 0.633 −0.285 −0.161 −0.359 0.056 T Vegetation 27.780−0.080VEG 0.558a 0.495a DO −0.459 0.572 0.627 −0.021 0.035 −0.186 pH Urban 8.305−0.267URB 0.535a 0.468a ORP 0.239 −0.186 0.317 −0.468 −0.059 −0.232 SPM Agriculture Urban −21.990+4.234AGR−40.181URB 0.740a 0.653a a a a a IMn −0.738 0.788 0.034 0.419 0.163 −0.011 IMn Agriculture 1.836+0.047AGR 0.622 0.568 BOD 0.104 −0.219 −0.215 0.172 0.056 0.189 NH4+–N Waters Bare lands 0.156+0.063WAT+0.022BAR 0.853b 0.804b + a b b − b b NH4–N −0.762 0.213 0.143 0.811 0.810 −0.096 NO3–N Bare lands 1.000+0.155BAR 0.778 0.747 − a b b b NO3–N −0.712 0.219 0.251 0.299 0.882 −0.060 DP Urban 0.001+0.047URB 0.650 0.600 DP −0.531 0.450 0.806b 0.505 0.122 0.542 The water quality parameters without regression models are not listed. a Significance at 0.05 probability level. a Significance at 0.05 probability level. b Significance at 0.01 probability level. b Significance at 0.01 probability level. S. Li et al. / Catena 75 (2008) 216–222 221

4.3. Linking land use/land cover and river water quality which is also reflected by its strong correlations with agriculture (R=0.79, p b0.05) and vegetation (R= −0.74, p b0.05) (Table 2). No Correlation and regression analyses by analyzing the average values of significant correlations between BOD and LULC (Table 2)andpeak the water quality parameters and land use patterns revealed that urban values in different seasons (Fig. 3) indicate the mixed and irregular was significantly correlated to pH and DP, while agriculture to IMn and influences including point source and diffuses (Chen et al., 2000). PH, vegetated land cover was negatively related to T, IMn and nitrogen. Bare corresponding to urban (Table 2; R =0.731, p b0.05), displays lower ground and rocks had strong positive correlation with nitrogen and T values in the Hanzhong plain and Danjiangkou Reservoir, due to (Table 2). Stepwise multiple linear regression demonstrated that no industrial effluents. EC and TDS show lower values in the rainy single land cover type was able to describe the overall water quality, season (Fig. 3), which is primarily due to the dilution by precipitation but most water physico-chemicals could be sufficiently predicted using (Dai, 1997). Their significant difference among zones implies one or two land use/land cover types (Table 3). PH and DP could be the different geology and weathering process across the basin − 2 predicted by urban area, NO3–N by bare lands (R =0.747,pb0.01), IMn by (Zhang et al., 1996). Their lowest constituents in the Ziwu River agriculture (R2 =0.568, pb0.05), SPM by agriculture and urban (R2 =0.653, catchment are associated to its highest vegetation coverage and + 2 pb0.05), and NH4–N by water and bare lands (R =0.804, pb0.01), granite lithology. respectively. The watershed delineation of the study area is variable, so different results about linkages between water quality parameters 5. Discussions and LULC may be obtained in different subwatershed scale. Mean- while, water quality variations in different temporal level increase the The overall concentrations of SPM, nitrogen and DP have increased uncertainties. While total means analysis of water quality variables since 2000 (Chen at al., 2000; Liu et al., 2003), which is associated can give us general information about the impacts of LULC on water with population increase and economic growth in the Han River basin quality. Also, the results can determine the types of LULC controlling (Liu et al., 2003; Wang et al., 2006). However, high vegetated land water quality, and seasonal and spatial patterns of water quality cover (Fig. 2) and low population density in few subwatersheds, such reflect their sources and polluted areas, which is very important for as Laoguan River, south of the Qinling Mountains and Ziwu River, are water quality conservation, i.e., sewage treatment plants construction contributable to the lower concentrations of SPM, nitrate and DP and revegetation. SNWTP will submerge more than 370 km2 land at respectively (Fig. 3; Chen at al., 2000; Liu et al., 2003). low asl and introduce a migration of more than 250,000 people Studies have shown that the percentage of agriculture at (Berkoff, 2003), which will impact land use patterns controlling water watershed scale is a primary predictor for nitrogen and phosphorus quality in the basin (Table 3). For example, the decrease of vegetated

(Hill, 1981; Johnson et al., 1997; Smart et al., 1998; Ferrier et al., 2001; land area and urbanization caused by migration will increase IMn and Ahearn et al., 2005), and there are elevated nitrate concentrations in nutrients. the streams of urbanized areas (Osborne and Wiley, 1988; Sliva and Williams, 2001). This study indicates that nitrogen concentrations 6. Conclusions varies among zones (Fig. 3) and in general have a negative correlation with vegetated coverage perhaps by the reduction of nitrogen loss All the water quality variables demonstrate spatial variations from soil (Table 2; Stark and Hart, 1997). However, bare lands coverage except T, DO, IMn BOD, and DP at subwatershed scale. Of which, is the primary variable to predict nitrogen concentrations (Tables 3; nitrogen, ORP, IMn and turbidity generally display higher contents in Hunsaker and Levine, 1995), that implies a great influence of the rainy season, while higher concentrations for T, BOD, DP, EC, TDS, weathering of bare rocks and gravels on nitrogen contents in the SPM, pH and DO in the dry season. The stretches of the river in riverine network (Holloway et al., 1998; Sliva and Williams, 2001). vegetated areas have lower levels of nutrients compared to areas close In the Han River basin, DP has significant positive correlation with to degraded land use type. Bare lands were significantly correlated to urban and could be predicted using urban area (Tables 2, 3), which is concentrations of nitrogen, and urban to pH and DP, while agriculture consistent with other reports (Osborne and Wiley, 1988; Ferrier et al., to SPM and IMn. 2001; Sliva and Williams, 2001). Meanwhile, relatively higher DP In the upper Han River basin, urban and agriculture are concentrations are observed in zones 5 and 9 where large cities primarily distributed in riparian, therefore, the effects of land use Hanzhong and Shiyan are respectively located (Fig. 3), that also in riparian on water quality and relationships between buffer indicates the impacts of industrial wastes and sewages on P landscape and stream water quality will be the subjects of our concentration. However, the difference in DP among zones is not further study. We expect further land cover changes declining water statistically significant regardless of the difference in land use and quality due to agricultural development and population increase; land cover (Figs. 2, 3), that implies that natural processes may have also, the global climate changes introducing the increase of played an important role in determining DP in the basin (Sliva and temperature and evaporation, the decrease of rainfall, serious soil Williams, 2001). erosion caused by high frequency of flooding and the changes of Urban areas and cultivated lands are primarily located along biochemical cycles, may further offset theses effects. It is recom- the river networks in the Han River basin (Fig. 1), and their mended that efforts to reduce waste discharges into fluvial system impacts on the water quality in streams were expected. Our should be taken in urban (e.g., Danjiangkou Reservoir) and in study indicates that agriculture and urban have dominant control agricultural areas. Additionally, land use in buffer zone in the on SPM and the highest SPM values occurs in the Ankang plain upper Han River basin and planned urban settlements should be (zone 7) where cultivated land occupy 21.1% of the land area (Table 3; emphasized. Fig. 2; Allan et al., 1997; Johnson et al., 1997; Ahearn et al., 2005). Although higher nutrients concentrations are observed in zone 9 Acknowledgements where Siyan city is located (Fig. 3), yet the influences of urban and agriculture on nutrients concentration in streams are relatively The research is supported by “Hundred-talent Project” of the limited in the Han River (Tables 2, 3). Perhaps the small percentage Chinese Academy of Sciences (O629221C01) and the National Key may have caused the underestimation of their impacts on water Technology R&D Program of China (2006BAC10B02). We would like to quality (Fig. 2). thank Sheng Gu, Jia Li, Lianfa Li, Sha Mu and Yiping Wang for their At subwatershed scale, much higher levels in the rainy season assistance with field sampling. We also thank comments and indicate the major source of diffuses for IMn (Fig. 3; et al., 1998), suggestions from two anonymous reviewers. 222 S. Li et al. / Catena 75 (2008) 216–222

References Li, S., Xu, Z., Cheng, X., Zhang, Q., 2008. Dissolved trace elements and heavy metals in the Danjiangkou Reservoir, China. Environmental Geology. doi:10.1007/s00254-007- 1047-5. Ahearn, D.S., Sheibley, R.S., Dahlgren, R.A., Anderson, M., Jonson, J., Tate, K.W., 2005. Ngoye, E., Machiwa, J.F., 2004. The influence of land use patterns in the Ruvu river Land use and land cover influence on water quality in the last free-flowing river watershed on water quality in the river system. Physics and Chemistry of the Earth draining the western Sierra Nevada, California. Journal of Hydrology 313, 234–247. 29, 1161–1166. Allan, J.D., Erickson, D.L., Fay, J., 1997. The influence of catchment land use on stream O'Callaghan, J.F., Mark, D.M., 1984. The extraction of drainage networks from digital integrity across multiple spatial scales. Freshwater Biology 37, 149–161. elevation data. Computer vision. Graphics and Image Processing 28, 323–344. Berkoff, J., 2003. China: The South–North Water Transfer Project—is it justified. Water Osborne, L.L., Wiley, M.J., 1988. Empirical relationships between land-use cover and Policy 5, 1–28. stream water-quality in an agricultural watershed. Journal of Environmental Cai, S., Chen, G., Du, Y., Wu, Y., 2000. Thoughts on sustainable development in the basin Management 26, 9–27. of Hanjiang River. Resources and Environment in the Yangtze Basin 9, 411–418. Sliva, L., Williams, D.D., 2001. Buffer zone versus whole catchment approaches to Chen, J., Gao, X., He, D., Xia, X., 2000. Nitrogen contamination in the Yangtze River studying land use impact on river water quality. Water Research 35, 3462–3472. system, China. Journal of Hazardous Materials A 73, 107–113. Shen, Z., Zhang, Q., Yue, C., Zhao, J., Hu, Z., Lv, N., Tang, Y., 2006. The spatial pattern of Chinese State Environment Protection Bureau (CSEPB), 2002. Water and Wastewater land use/land cover in the water supplying area of the Middle Route of the South to Monitoring Analysis Methods (Press 4). Chinese Environment Science Press, Beijing, North Water Diversion Project. Acta Geographica Sinica 61, 633–644. China. Shen, D., Liu, C., 1998. Effects of different scales of MR-SNWTP on the down stream of Dai, S., 1997. Environment Chemistry. Higher Education Press, Beijing, China, pp. 114–120. the Danjiang Kou reservoir. Acta Geographica Sinica 53, 341–348. FAO/UNESCO., 1971. Soil map of the world 1:5,000,000. UNESCO, Paris, France. Smart, R.P., Soulsby, C., Neal, C., Wade, A., Cresser, M.S., Billett, M.F., Langan, S.J., Ferrier, R.C., Edwards, A.C., Hirst, D., Littlewood, I.G., Watts, C.D., Morris, R., 2001. Water Edwards, A.C., Jarvie, H.P., Owen, R., 1998. Factors regulating the spatial and quality of Scottish rivers: spatial and temporal trends. Science of the Total temporal distribution of solute concentrations in a major river system in NE Environment 265, 327–342. Scotland. The Science of the Total Environment 221, 93–110. Hill, A.R., 1981. Stream phosphorus exports from watersheds with contrasting land uses Stark, J.M., Hart, S.C., 1997. High rates of nitrification and nitrate turnover in undisturbed in southern Ontario. Water Resources Bulletin 17, 627–634. coniferous forests. Nature 385, 61–64. Holloway, J.M., Dahlgren, A., Hansen, B., Casey, W.H., 1998. Contribution of bedrock Turner, R.E., Rabalais, N.N., 2003. Linking landscape and water quality in the Mississippi nitrogen to high nitrate concentrations in stream water. Nature 395, 785–788 River Basin for 200 years. Bioscience 53, 563–572. Letters to Editor. Wang, J., Yan, W., Jia, X., 2006. Modeling the export of point sources of nutrients from Hunsaker, C.T., Levine, D.A., 1995. Hierarchical approaches to the study of water quality the Yangtze River basin and discussing countermeasures. Acta Scientiae Circum- in rivers. BioScience 45, 193–203. stantiae 24, 658–666. Jin, R., Guo, H., 1993. Water resources assessment in the water source areas of the Yang, Y., Zhou, N., Guo, X., Hu, Q., 1997. The hydrology characteristics analysis of Middle Route of the South to North Water Transfer Project and water quantity HanJiang up-streams. Hydrology 54–56. analysis in the Danjiangkou Reservoir. Yangzte River 24, 7–12. Zhang, L., She, Z., Zhang, S., 1996. Study of Chemical Elements in Water Environment. Johnson, L.B., Richards, C., Host, G.E., Arthur, J.W., 1997. Landscape influences on water Chinese Environment Science Press, Beijing, China. chemistry in Midwestern stream ecosystems. Freshwater Biology 37, 193–208. Zhang, Q., 2005. The South-to-North Water Diversion (SNWD) Project. Frontiers in Liu, S., Zhang, J., Chen, H., Wu, Y., Xiong, H., Zhang, Z., 2003. Nutrients in the Changjiang Ecology and the Environment 3, 75–76. and its tributaries. Biogeochemistry 62, 1–18. Zhao, P., Shen, X., Xia, J., Li, Q., Gao, H., Zhang, S., 1998. The Impacts of Silts on Water Li, S., Liu, W., Gu, S., Cheng, X., Xu, Z., Zhang, Q., in press. Spatio-Temporal Dynamics Quality of the and Its Water Pollution Control. Yellow River Water of Nutrients in the Upper Han River Basin, China. Journal of Hazardous Materials. Conservancy Press, Zhengzhou, China. doi:10.1016/j.jhazmat.2008.06.059.