Ecological Indicators 60 (2016) 202212

Contents lists available at ScienceDirect

Ecological Indicators
journal homepage: www.elsevier.com/locate/ecolind

Effect of land use types on stream water quality under seasonal

variation and topographic characteristics in the Wei River basin, China
Songyan Yu a , Zongxue Xu a,b, , Wei Wu c , Depeng Zuo a,b
a
College of Water Sciences, Beijing Normal University, Xinjiekouwai Street 19, Haidian, Beijing 100875, China
b
Joint Center for Global Change Studies (JCGCS), Beijing 100875, China
c
School of Resources and Environment, University of Jinan, Jinan 250022, China

a r t i c l e i n f o a b s t r a c t

Article history: The effect of topographic characteristics of land uses on stream water quality must be addressed for a
Received 28 June 2014 better understanding of the complex relationship between land use and stream water quality. In this
Received in revised form 17 June 2015 study, Geographic Information System (GIS) and Pearson correlation analysis were used to determine
Accepted 18 June 2015
whether there were relationship between land use types and stream water quality at the sub-basin scale in
Available online 17 July 2015
the Wei River basin, China, during the dry and rainy seasons in 2012. Temporal variation of these relations
was observed, indicating that the relationships between water quality variables and different land uses
Keywords:
were weaker in rainy seasons than that in dry seasons. Compared with other land uses, agriculture
Water quality
Land use
and urban lands had a stronger relationship with water quality variables in both rainy and dry seasons.
Topographic characteristics Topographic characteristics of land use were employed to further analyze these relationships. The results
Seasonal variation showed that seasonal variation also occurred in the complex relationship, and land uses in steeper slopes
Wei River generally had a stronger inuence on stream water quality than those in atter ones. For the riparian zone
of each sampling site, the slope coefcients were weaker than those at the sub-basin scale. Land use type
near stream water was generally a better indicator for the effectiveness of water quality. These results
suggest that the slope and proximity should be taken into account for better land use management.
2015 Elsevier Ltd. All rights reserved.

1. Introduction
Although the signicant impact of land use on stream water
quality has been well documented (Johnson and Gage, 1997; Allen,
2004; Hurley and Mazumder, 2013; Bu et al., 2014; Ye et al., 2014),
further study on the complex association should be considered as
much as possible, particularly on how the association varies tem-
porally and with different topographical characteristics of land use.
As part of the climatic inuence on surface water quality (Clark
et al., 1996; Hunter, 2003), the importance of the temporal scale
on the corresponding relationship should be noted. Dominant fac-
tors of the temporal scale, such as precipitation, temperature and
agricultural activities, vary among seasons, and given their role on
ow convergence process and contaminant inputs into water bod-
ies, researchers have suggested that it is imperative to consider
Corresponding author at: College of Water Sciences, Beijing Normal University,
Xinjiekouwai Street 19, Haidian District, Beijing 100875, China.
Tel.: +86 010 58801136.
E-mail addresses: yusongyan@mail.bnu.edu.cn (S. Yu), zongxuexu@vip.sina.com
(Z. Xu).
seasonal variation when studying the impact of land use on river
water quality (Johnson et al., 1997). Ye (2014) indicated that while
non-point source pollution was predominant in rainy seasons, agri-
cultural and forested land showed stronger association with water
chemistry. Bu (2014) employed statistical and spatial analyses to
investigate the relationship between land use patterns and river
water quality in the Taizi River basin, China, during both dry and
rainy seasons, and found that during dry seasons, point source
pollution was predominant, but in the rainy season water quality
showed mixed pollution from both point and non-point sources.
Many studies have addressed the general relationship between
land use patterns and water quality; however, a full understanding
of the correlation has not been achieved because land use types as
a whole are actually mutual reections of multiple factors, such as
geomorphic characteristics and anthropogenic activities.
In addition, the topographic characteristics (e.g., averaged slope)
of catchments were frequently used to obtain an improved under-
standing of the association between land use and water quality.
Christopher et al. (2010) distinguished the effect of different prox-
imities of far-eld and near-eld land use on water quality and indi-
cated that the near-eld, 200 m buffer on each side of the stream
showed signicant inuences on river health, while the far-eld

http://dx.doi.org/10.1016/j.ecolind.2015.06.029
1470-160X/ 2015 Elsevier Ltd. All rights reserved.
 S. Yu et al. / Ecological Indicators 60 (2016) 202212
buffer did not. Wang (2013) performed an analysis on the corre-
lation between water quality and land use types by incorporating
the average slope of each sub-basin, but the average slope at the
sub-basin level was still of a relatively coarse resolution, while
other factors affecting water quality at the sub-basin scale were
still hidden. Generally, previous studies mainly focused on single
geomorphic factors, and few studies have made attempts to fur-
ther study the geomorphic characteristics at the class level, which
is of a ner resolution, to obtain an improved understanding of the
dynamic complexity of relationship between land use patterns and
water quality.
The objectives of this study were (1) to compare the inuence
of land use on stream water quality in rainy seasons to that in dry
seasons in an attempt to determine which season has a more sig-
nicant impact on the stream water quality in the Wei River basin,
and (2) to incorporate the topographic characteristics of each land
use type at the sub-basin scale to detect how the slope of each
type affected the association. The inuence of the slope on this
type of association could add a new dimension of understanding
to the environmental factors encompassing a watershed drainage
area.
2. Materials and methods description
2.1. Study area description
Wei River, the largest tributary of the Yellow River, originates
north of the Niaoshu Mountain at 3485 m in Gansu Province, China,
and ows across 818 km with 134.3 104 km2 of drainage area. The
study basin consists of three catchmentsthe Wei River catchment
(W, representing 46.1% of the total area), the Jing River catchment
(J, representing 33.85% of the total area) and the Beiluo River catch-
ment (L, representing 20.1% of the total area)and each catchment
has special characteristics (Fig. 1). The Wei River catchment is the
most developed area of the three catchments, which means that
Fig. 1. Study area and distribution of sampling sites. The black solid triangles represent sites in the Wei River basin. The circles represent the main cities in the basin. (Base
map is derived from DEM by extracting borders of the basin in ArcGIS software).
203
it has a large amount of industry and a high percentage of sealed
surfaces; in the Jing River catchment and the Beiluo River catch-
ment, oil extractions, which strongly affect the surrounding
environment, occur (Kang et al., 2008). In addition, this river pro-
vides numerous ecosystem services, including a source of water for
domestic, industrial and agricultural purposes in the Guanzhong
Plain, thereby playing an important role for the development of
economics and society in Western China.
Land use information was obtained from the Resource and
Environmental Sciences Data Center of CAS (Chinese Academy
of Science). Data was provided in a 30 m 30 m resolution and
included information from six land use categories (as shown in
Fig. 2): (1) agriculture land, including paddy elds and dry land; (2)
forestland, including shrub land and sparse woodlot; (3) grassland,
including different coverage types; (4) water body, including rivers,
wetlands and sandy beaches; (5) urban land, including industrial
and residential areas; and (6) barren land, including gravel, bare
ground and bare rocks.
A digital elevation model (DEM) at a 30 m 30 m resolution was
used to delineate the basin area and to extract topographic char-
acteristics, such as the average slope of each land use type at each
sub-basin. The Wei River basin was delineated into 44 sub-basins
based on the locations of sampling sites. Each sampling site was
selected as the outlet point for a discrete sub-basin. Therefore, the
effect of land use composition and its topographic characteristics
(e.g., average slope and distance) on water quality was analyzed at
the sub-basin level.
In addition, to further study the impact of topographic char-
acteristics on the relationship between land use type and water
quality, four categories of slope were extracted from the DEM,
including category I (05 ), category II (515 ), category III (1530 )
and category IV (3068 ), in both sub-basin and riparian scales.
These four types of slope were then taken into account after Pear-
sons correlation analysis on the relationship between land use and
water quality.
204 S. Yu et al. / Ecological Indicators 60 (2016) 202212
Fig. 2. Land use within the Wei River basin. Inset charts display the percentage of different land use types in the whole basin and three divided catchments, where the
abscissa axis of the histogram is marked by A (agriculture land), F (forestland), G (grassland), W (water body), U (urban land), B (barren land).
2.2. Sampling sites and water quality
Water was sampled in four eld surveys along the Wei River
during two periods: the dry season (April in 2012 and 2013) and
the rainy season (October in 2011 and 2012). The mean values were
used for each season in the statistical analysis. Forty-four sampling
sites, comprising 23 sites in the Wei River catchment, 11 sites in the
Jing River catchment, and 10 sites in the Beiluo River catchment,
were sampled in the whole basin during both dry and rainy sea-
sons. To make a balance between cost-effective sampling strategy
and optical representativeness of water quality within the basin,
rstly, for the Wei River catchment, this study focused on the main
stream and included only a few tributaries, as the catchment has an
elongated shape and its tributaries are comparatively short; for the
Jing Catchment and Beiluo Catchment, because of their relatively
small areas, this study paid as much attention on the main stream
as on tributaries (Fig. 1).
Twelve representative parameters were measured, including
pH, dissolved oxygen (DO), electrical conductivity (EC ), total dis-
solved solids (TDS), chloride (Cl ), sulfate (SO4 2 ), permanganate
index (CODMn ), ammoniacal nitrogen (NH3 -N), nitrite nitrogen
(NO2 -N), nitrate nitrogen (NO3 -N), orthophosphate (PO4 3 ), and
total phosphorus (TP). The values of pH, DO, EC, and TDS were
directly measured in situ using a multiparameter water quality
monitoring instrument YSI 85. Other parameters were measured
in the laboratory according to the national standard criterion (GB
3838-2002).
2.3. Statistical analysis
For water quality properties, two statistical features, namely
the average and standard deviation, were calculated at each sub-
basin and the Wei River catchment, Jing River catchment and Beiluo
River catchment, respectively, in dry and rainy seasons. A one-
way analysis of variance test (ANOVA test) was performed on the
log-transformed water quality values to determine whether the
differences between the two seasons were signicant.
Correlations between land use and water quality chem-
istry were tested by using Pearsons correlation with sta-
tistical signicances at the p < 0.05 level (one-tailed) in dry
and rainy seasons, respectively. Before the correlation analy-
sis, KolmogorovSmirnov test (KS test) was used to test the
normal distribution for all of the variables. In addition, Pear-
sons correlation analysis was also conducted for the impact of
incorporating the average slope of land use type within each sub-
basin on the relationship between land use and stream water
quality.
The ANOVA test is a relatively robust procedure with respect to
violations of the normality assumption (Kirk, 1995). Because water
quality data considered is not ordinal, the ANOVA test is more suit-
able than a non-parametric test such as KruskalWallis one-way
analysis of variance. Also, the KS test was chosen because it is one
of the most useful and general non-parametric methods for com-
paring two samples, and testing for the normality of distribution.
KS test was however also chosen for this study because it is sen-
sitive to differences in both location and shape of the empirical
cumulative distribution functions of two samples. Other tests such
as the ShapiroWilk test or AndersonDarling test have their own
disadvantages (Eadie et al., 1971). For instance, the ShapiroWilk
test was known not to work well with many ties (many iden-
tical values). Lastly, Pearsons correlation coefcient was chosen
for this study because it is widely used in the sciences as a mea-
sure of the degree of linear dependence between two variables
(Pearson, 1895). Other similar tests such as Kendalls Tau corre-
lation coefcient were less inuenced by the effects of outliers and
small sample numbers and were only used if small data sets were
strongly inuenced by unusual values (Buck et al., 2004; Hesel and
Hirsch, 1995). These three statistical methods were used to identify
the correlation between water quality variables and environmental
factors.
 S. Yu et al. / Ecological Indicators 60 (2016) 202212 205

3. Results analysis
3.1. Land use distribution of the Wei River basin at the sub-basin
scale
Table 1 shows the distribution of land use for each sub-basin
in 2007. Agricultural land and grassland were the most common
land uses in the basin. Other land uses including urban, water
bodies and barren land were relatively minor components of the
sub-basins (<3% combined for all 44 sub-basins except for W21). It
should also be noted that some sub-basins, such as W21 and W18,
showed a greater proportion of urban land use, reecting an accel-
erating rate of urbanization. As shown in Fig. 1, these sub-basins
are located within the Guanzhong Plain, where several cities are
located, such as Xian and Xianyang. One interesting observation
was that sub-basins with a greater proportion of urban land have a
lesser proportion of grassland but a greater proportion of agricul-
tural land. This observation may be explained by the assumption
that urbanization has decreased the amount of the grassland to
support more living residents in the city. Because more food was
necessary, agricultural land has been retained at a high proportion
(Song et al., 2007).
Agriculture land was the dominant land use type for most sub-
basins, with areas ranging from 2% to 87% (mean 43%). Likewise,
a signicant amount of grassland was also observed in most sub-
basins, ranging from 1.5% to 86.3%. Twenty-six out of 44 sub-basins
had a signicant amount of forestland (i.e., >4.0%), with as much as
76.7% of the total area in one sub-basin (L5). However, forest was
the scarcest land use in the Jing River catchment, while agriculture
was the scarcest land use in the Beiluo River catchment.
3.2. Physicochemical water quality properties in the Wei River
basin
No signicant temporal change was observed in the ANOVA
test (p > 0.05) for the variables measured during the dry and wet
seasons. These twelve parameters could be used for correlation
analysis between land use and water quality. As shown in Fig. 3,
the mean concentrations of EC , DO, TDS TP, Cl , NO2 -N and NH3 -N
were slightly higher in the dry season than that in the rainy season.
Conversely, the mean concentrations of other parameters showed
the opposite result.
The mean concentration and standard deviation of water quality
variables were calculated at different scales, as shown in Table 2.

Table 1
Land use distribution at the sub-basin scale.

Agriculture land (%) Forestland (%) Grassland (%) Water body (%) Urban land (%) Baren land (%)

Wei catchment
W1 43 4 53
W2 52 1 47 1
W3 35 26 38
W4 32 14 55
W5 74 1 25
W6 77 23
W7 75 2 22
W8 72 3 24
W9 59 15 25 1
W10 3 61 36
W11 16 35 50
W12 4 39 57
W13 40 23 37
W14 43 22 31 2 2
W15 44 26 29 1
W16 63 15 22 1
W17 2 69 28
W18 62 21 15 2
W19 37 39 22 1
W20 33 25 42 1
W21 69 4 1 2 23 1
W22 69 11 18 2
W23 72 12 13 1 1

Jing catchment
J1 32 2 65
J2 31 3 66
J3 29 33 38
J4 41 1 58
J5 46 16 37 1
J6 60 3 36 1
J7 66 1 33
J8 77 2 21
J9 55 10 34
J10 37 19 44
J11 83 4 12 1

Beiluo catchment
L1 42 2 57
L2 13 1 86
L3 20 1 78
L4 7 42 51
L5 22 16 61
L6 6 45 49
L7 8 77 15
L8 25 30 44
L9 42 24 33
L10 87 3 8 1 1
206 S. Yu et al. / Ecological Indicators 60 (2016) 202212

Fig. 3. Physicochemical water quality properties during both the rainy and dry season in the Wei River basin.

On examining the effect of scales, different water quality variables
took on different spatial patterns. For example, the mean concen-
trations of nutrient elements such as TP, CODMn , NO2 -N, NO3 -N,
NH3 -N and PO4 3 were 0.77 mg/L, 5.66 mg/L, 3.00 mg/L, 11.91 mg/L,
0.88 mg/L and 1.06 mg/L respectively, and were much higher in the
Wei River catchment than that in other two catchments, indicat-
ing that domestic and industrial sewage and agricultural non-point
source pollutants had strongly negative relationships with water
quality. It is mainly due to two reasons. The rst one is that the
percentage of agricultural lands in the Wei River catchment is more
than that in other two catchments (56.1%, 45.5%, 26.0% within the
Wei River catchment, the Jing River catchment and the Beiluo River
catchment, respectively). This result is in agreement with previous
studies in which agricultural activities showed signicant correla-
tions with nutrient elements in many places such as America, China,
India, Japan and New Zealand (Allen, 2004; Amiri and Kaneyuki,
2009; Buck et al., 2004; Chattopadhyay et al., 2005; Wang, 2013).
The other reason is that there are a few cities (Fig. 1), such as Tian-
shui, Baoji, Xianyang, Xian and Weinan cities, with relatively high
development along the major stream of the Wei River. Human
activities in these nearby cities caused consequent discharge of
domestic and industrial sewage into the river. At the same time,
the concentrations of EC , TDS, and DO in the Jing River catchment
and the Beiluo River catchment were signicantly higher than that
in the Wei River catchment (1182 and 1348 to 664 s/cm, 929 and
1181 to 524 mg/L, 10.9 and 10.9 to 9.1 mg/L, respectively), suggest-
ing the presence of a serious soil erosion issue, which has been
widely mentioned in previous studies (Mo, 2008; Qin, 2009; Xu
et al., 2008). In addition, because of high percentage of grassland,
animal herding is quite common in the two catchments, thus bank
erosion caused by drinking and crossing animals could be an expla-
nation for the higher TDS and DO values. As showed in Table 1, it
may be illogical at the rst glance because grassland covers quite
a large part of this catchment. However it should be noted that
only in summer and early fall do grass thrive, preventing soil ero-
sion, but in other periods of the year, when live grass disappear
from the surface of soil, soil erosion will come back eventually
(Qin, 2009). The chloride (Cl ) and sulfate (SO4 2 ) concentrations
(168.82 mg/L and 664.37 mg/L, respectively) was highest in the Jing
River catchment. Cl is considered to correlate with human activi-
ties, especially with constructed areas, while SO4 2 is derived from
a wealth of sources, such as domestic and industrial sewage, and
minerals in nature (Huang et al., 2012). Therefore, Cl may be asso-
ciated with urban area and agricultural lands, while SO4 2 is more
likely to be correlated with urban areas and soil composition in this
catchment.
3.3. Relationship between land use and water quality and its
temporal effect
The KolmogorovSmirnov goodness of t test suggested that
the rainy season variables such as EC , TDS, CODMn , Cl , PO4 3
and NH3 -N were not distributed normally (p < 0.05); and for the
dry season, EC , TP, Cl , NO2 -N, PO4 3 , SO4 2 , and NH3 -N did
not t the normality of distribution; thus, these variables were
log-transformed for further analysis. The log-transformed results
showed that all of the log-transformed water quality parameters,
except PO4 3 , were distributed normally (p > 0.05), so PO4 3 was
not employed for the following correlation analysis.
As shown in Table 3, the statistical analysis indicated that in
the rainy season several water quality variables were signicantly
correlated with agricultural area, forest area, and grass area, as well
as with the urban and barren areas at a signicance level of >0.95
(i.e., p < 0.05). However, no water quality variable had a signicant

Table 2
The mean concentration and standard deviation of the water quality variables at different scales.

Water quality parameters Sub-basin W catchment J catchment L catchment

Mean SD Mean SD Mean SD Mean SD

pH 7.53 (5.23, 8.96) 0.64 7.33 (5.23, 8.42) 0.67 7.78 (6.84, 8.69) 0.54 7.70 (6.77, 8.96) 0.56
EC (s/cm) 955 (183, 4208) 724 664 (183, 1387) 302 1182 (451, 4208) 893 1348 (218, 3869) 918
DO (mg/L) 10.0 (4.0, 17.7) 2.45 9.1 (4.0, 15.0) 2.0 10.9 (8.2, 17.2) 2.4 10.9 (5.0, 17.7) 2.8
TDS (mg/L) 780 (146, 3295) 584 524 (147, 1222) 245 929 (374, 3094) 647 1181 (404, 3296) 763
TP (mg/L) 0.58 (0.01, 4.73) 0.76 0.77 (0.01, 4.73) 0.78 0.25 (0.02, 0.96) 0.28 0.51 (0.02, 4.19) 0.95
CODMn (mg/L) 4.51 (1.06, 27.88) 3.28 5.66 (2.02, 27.88) 4.01 3.14 (1.53, 5.33) 0.98 3.48 (1.06, 8.08) 2.04
Cl (mg/L) 94.9 (0.6, 827.3) 157 51.9 (0.6, 210.9) 48 168.8 (8.3, 827.3) 195 108.4 (1.8, 792.3) 230
NO2 -N (mg/L) 2.06 (0, 56.07) 6.23 3.00 (0, 56.07) 8.57 1.21 (0, 4.84) 1.26 0.95 (0, 4.71) 1.27
NO3 -N (mg/L) 10.85 (1.59, 42.29) 8.69 11.91 (1.59, 35.60) 7.08 11.26 (3.10, 42.29) 11.22 8.07 (1.60, 36.70) 8.64
PO4 3 (mg/L) 0.70 (0.01, 2.57) 0.78 0.88 (0.01, 2.57) 0.83 0.62 (0.02, 2.15) 0.75 0.40 (0.01, 1.46) 0.61
SO4 2 (mg/L) 483 (5, 1567) 277 371 (5, 795) 163 664 (248, 1412) 296 533 (158, 1567) 342
NH3 -N (mg/L) 0.95 (0.01, 9.55) 1.70 1.06 (0.01, 9.55) 2.07 0.87 (0.01, 3.73) 0.99 0.80 (0.01, 5.64) 1.46
 S. Yu et al. / Ecological Indicators 60 (2016) 202212 207

Table 3
Pearsons correlation coefcients between land use types and water quality chemistry during both the rainy and dry season.

Agriculture land Forestland Grassland Water body Urban land Barren land

Rainy season
pH 0.461* 0.465
EC 0.334 0.396
DO 0.418 0.430* 0.637
TDS 0.300* 0.418
Cl 0.334 0.304
NO2 -N 0.341 0.399 0.404
NO3 -N 0.335 0.312*

Dry season
EC 0.525 0.344 0.436
DO 0.378 0.411
TDS 0.429 0.409
TP 0.404*
CODMn 0.352 0.425 0.471
Cl 0.47 0.326
NO2 -N 0.378 0.544 0.454
NO3 -N 0.306
SO4 2 0.502 0.269*
NH3 -N 0.342 0.406

Only signicant relationships are listed. All p-values are less than 0.05, except where denoted.
*
Denotes signicant relationship at a level of p = 0.051.

correlation with water bodies, as they only had a correlation at a
signicance of >0.90 and <0.95. Although many researchers have
argued that water bodies such as wetlands may have high level
of nitrates and phosphates delivered from point and non-point
sources of contamination, this may be signicantly reduced
because of natural deposition and degradation (Anbumozhi et al.,
2005; Naiman and Decamps, 1997). On the other hand, from the
viewpoint of water quality parameters, all of the parameters,
except TP, CODMn , SO4 2 and NH3 -N, were strongly associated
with one or more land use types.
For the dry season, the statistical analysis presented different
correlation patterns from those in the rainy season, showing that
water quality variables were inuenced more by all types of land
uses. All land use types had a strong relationship with water qual-
ity variables at a signicant level >0.95 (i.e., p < 0.05). On the other
hand, most of the eleven water quality variables that were con-
sidered could be explained by land use variables. EC and NO2 -N
were strongly associated with as many as three land use variables.
However, no land use showed a signicant relationship with the
pH value.
On the whole, it is observed that the relationships between
water quality variables and land use were weaker in the rainy sea-
son than that in the dry season. In addition, compared with other
land use types, agriculture and urban lands had a stronger relation-
ship with water quality variables in both rainy and dry seasons.
3.4. Contribution of topographic characteristics of land use on the
complex relationship
The correlation between land use and water quality was further
analyzed by taking into account the classied slope from land use
to outlets within each sub-basin in this study, as shown in Table 4.
At the sub-basin perspective, for the dry season, total phospho-
rus (TP), permanganate index (CODMn ), chloride (Cl ), nitrogen
(NO3 -N) and sulfate (SO4 2 ) tend to be signicantly associated
with steeper areas within agricultural lands, forestlands and bar-
ren lands. It is mainly because in the dry period, soil surface in
agricultural and barren areas may not be covered by vegetation to
a large extent, so that it is more likely for soil erosion to occur in
steeper areas than that in atter areas. Also Ec and TDS, which has
close correlation with soil particles discharged into rivers, showed
signicant relationships with diverse slope categories of areas
in agricultural lands, forested lands, grasslands and urban lands,
indicating that these four land use types can be referred as sources
rather than sinks for soil particles. For rainy seasons, it presented
a weaker correlation with the topographic characteristics of each
land use. For example, compared to the results in dry seasons,
only CODMn and Cl were correlated with steeper areas within
agricultural lands and forestlands, respectively. Two water quality
variables, Ec and TDS, were still signicantly inuenced by various
slope categories of areas within agricultural lands, forestlands and
grasslands. This could be seen as a further support to the statement
that Ec and TDS were mostly an issue arose from bad management
of natural lands (Allen, 2004).
For the riparian zone of each sampling site, the slope coefcients,
shown in Table 5, were generally less strong than those at the sub-
basin scale. It should also be noted that the inuence of land use was
much more signicant in the atter areas than that in the steeper
ones. For example, within agricultural lands, two water quality vari-
ables, DO and NO2 -N, were strongly associated with areas of slope
category I and II. In addition, within forestland, DO and nutrients,
including TP, NO2 -N and NO3 -N, had a signicant relation to areas
of slope category I and II. The same results were found in grassland
and barren lands as well.
4. Discussion
4.1. Temporal differences in the linkage between land use and
water quality
Correlations between the percentage cover of each land use type
and the water quality variables in both rainy and dry seasons were
analyzed at all sampling sites (Table 4). Our results showed signif-
icant temporal differences between these correlations.
In rainy seasons, with the exception of SO4 2 , NH3 -N, TP and
CODMn , all of the water quality variables were correlated with
at least one land use type. The lack of correlation between land
use and NH3 -N suggested that the loading process of NH3 -N was
impacted by many factors that were not conrmed in wastewater
discharged by domestic and industrial activities, as concluded by
Ahearn (2005). Similar results were obtained by other researchers
(Osborne and Wiley, 1988; Sliva and Williams, 2001; Galbraith and
Burns, 2007; Zhao, 2008).
In addition, agriculture was positively associated with physi-
cochemical and nutrient variables in both rainy and dry seasons,
which indicated that it actually served as a source for pollution in
208 S. Yu et al. / Ecological Indicators 60 (2016) 202212

Table 4
Correlation of the slope categories of each land use type within sub-basins with water quality variables.

Agriculture land Forestland Grassland

I II III IV I II III IV I II III IV

Dry season
EC 0.522 0.59 0.349 0.521 0.403* 0.58 0.451 0.351*
DO 0.334 0.316 0.331
TDS 0.522 0.569 0.306 0.565 0.462 0.549 0.428 0.406
TP 0.267* 0.449 0.546
CODMn 0.414* 0.263 *

Cl 0.322 0.416 0.567 0.466

NO2 -N 0.254*
NO3 -N 0.412 0.278*
SO4 2 0.357 0.504 0.56 0.409

Rainy season
pH 0.279* 0.502
EC 0.400 0.590 0.263* 0.547 0.536 0.557 0.337 0.403
DO 0.273*
TDS 0.399 0.589 0.264* 0.573 0.586 0.556 0.333 0.487
TP 0.264*
CODMn 0.276* 0.455
Cl 0.272* 0.351 0.422* 0.382*
NO2 -N
NO3 -N
SO4 2 0.386

Water bodies Urban land Barren land

I II III IV I II III IV I II III IV

Dry season
EC 0.829 0.331* 0.361*
DO
TDS 0.556 0.464
TP 0.603* 0.683
CODMn 0.352 0.517 0.470 0.428* 0.727
Cl 0.671
NO2 -N 0.463* 0.77
NO3 -N 0.867
SO4 2 0.371

Rainy season
pH 0.431* 0.456
EC 0.535*
DO 0.416
TDS 0.534*
TP 0.883
CODMn 0.667
Cl
NO2 -N 0.32* 0.439 0.661
NO3 -N
SO4 2

Only signicant relationships are listed. All p-values are less than 0.05, except where denoted.
*
Denotes signicant relationship at a level of p = 0.051.

the Wei River basin. The positive relation of urban land with NO2 -
N, DO and NO3 -N should also be noted. Because urban land was
always considered to be an important variable that was associated
with water quality variables (Reimann et al., 2009; Guo et al., 2010),
urban areas tend to have low concentration of DO (Huang et al.,
2013), which may result in impeded nitrication. Accordingly, our
study demonstrated that urban land was negatively associated with
DO.
Furthermore, grass is a detention medium that is helpful in
reducing the amount of surface runoff and sediment yield (Allen,
2004). It is reasonable to anticipate a negative relationship between
grassland and water quality variables. However, our results were
not in agreement with the anticipation; however, positive correla-
tions between EC , TDS and grassland (r = 0.396, 0.418, respectively)
were identied. As mentioned above, a large distribution of grass-
land existed in the study area, but for all sub-basins, the average
distance of grassland to outlets was the second longest (Table 6),
which indicated that sediments were ltered by grassland in the
upstream parts of the basin on the one hand, while the surface
runoff on the other hand still had great potential to pick up soil
particles downstream before reaching the main stream. In addition,
soil structure had become loose due to intensive grazing in recent
years, this was frequently observed during the four eld surveys,
and it might have undermined the effect of grassland in purifying
runoff even in the upstream.
Our results also suggested that forestland use was closely related
to the fewest water quality variables than any other land use types
in rainy seasons according to the Pearsons correlation analysis, and
only a negative correlation between Cl and forestland was found
(r = 0.304). This was mainly because forestland was located mostly
upstream of the watershed, whereas urban and agriculture land
were chiey situated in the riverine area, as shown in Table 6.
In dry seasons, agriculture land use was strongly correlated with
most physicochemical and nutrient variables, suggesting that agri-
cultural land use had negative effects on river water quality because
of intensive fertilization and irrigation during the farming season
(Ngoye and Machiwa, 2004; Lee et al., 2009). For example, accord-
ing to the eld surveys, April, which is the dry season as dened in
 S. Yu et al. / Ecological Indicators 60 (2016) 202212 209

Table 5
Correlation of the slope categories of each land use type within the riparian zone with water quality variables.

Agriculture land Forestland Grassland

I II III IV I II III IV I II III IV

Dry season
pH 0.914*
DO 0.311 0.4 0.605 0.378 0.42 0.941*
TP 0.261* 0.7
CODMn 0.947* 0.528*
Cl 0.513*
NO2 -N 0.314 0.311* 0.688 0.554 0.298*
NO3 -N 0.589
SO4 2 0.56

Rainy season
pH 0.293* 0.313* 0.441
EC 0.513
DO 0.274* 0.532* 0.290*
TDS 0.491*
NO2 -N 0.919*
NO3 -N 0.925* 0.761 0.615
SO4 2 0.941*

Water bodies Urban land Barren land

I II III IV I II III IV I II III IV

Dry season
EC 0.829 0.331* 0.361*
DO
TDS 0.556 0.464
TP 0.603* 0.683
CODMn 0.352 0.517 0.47 0.428* 0.727
Cl 0.671
NO2 -N 0.463* 0.77
NO3 -N 0.867
SO4 2 0.371
NH3 -N 0.356

Rainy season
pH 0.431* 0.456
EC 0.535*
DO 0.416
TDS 0.534*
TP 0.883
CODMn 0.667
NO2 -N 0.32* 0.439 0.661

Only signicant relationships are listed. All p-values are less than 0.05, except where denoted.
*
Denotes signicant relationship at a level of p = 0.051.

this study, is in the farming season. Due to a small amount of pre-
cipitation during this period, irrigation is necessary after sowing.
Because the efciency of water utilization is low in the study area,
it appears likely that a lot of water that was used for irrigation would
return to streams through surface and subsurface runoff (Sliva and
Williams, 2001; Allen, 2004), especially within steep areas. This is
further supported by the general high concentration of most nutri-
ent variables within sub-basins during dry seasons, mainly caused
by the fertilization of agriculture elds in the study area.
In addition, the relatively small occupation of the correlation
coefcient for urban areas suggested that it was agricultural land,
rather than urban land, that was the primary contributor to water
quality during dry seasons. However, urban land still had a stronger
correlation with water quality variables in dry seasons compared to
that in rainy seasons, such as NO2 -N (r = 0.399, 0.454 in rainy and
dry seasons, respectively). Furthermore, this temporal difference
for urban land revealed that the degree of river water quality con-
taminated by possible point sources was relatively high compared
to that in rainy seasons because urban land had been considered
as the proxy for point source by many previous researchers (Karr,
1999; Ahearn et al., 2005; Bu et al., 2014). Ahearn and others (2005)
argued that much of the cover in urban areas was impervious and
the drainage is frequently routed to wastewater treatment plants
and then discharged to local rivers as point sources.
Therefore, the land use type had a more signicant correlation
in dry seasons than that in rainy seasons not only in the correla-
tion coefcients for each water quality characteristic but also in
the degree of inuence on each water quality variable. Therefore,
seasonal variation, to some extent, inuences the linkage between
land use characteristics and water quality variables. It is recom-
mended that seasonal variability is recognized in later research as
an important factor that might affect this type of linkage.

Table 6
Average distance of each land use type within the Wei River basin.

Agriculture land Forestland Grassland Water body Urban land Barren land

Distance (m) 41,295.3 47,441.5 46,315.9 44,585.0 36,088.6 38,552.4

210 S. Yu et al. / Ecological Indicators 60 (2016) 202212
In addition, some important caveats apply to studies of the
relationship between land use and stream water quality. When
conducting a Pearsons correlation analysis on this type of relation-
ship, researchers always took land use type into consideration, and
because all land use sums to 100%, different measures of land use
may predict stream water quality nearly equally well (Herlihy et al.,
1998); therefore, the conclusion that the land use type was the pri-
mary driver of stream water quality must be made with caution. In
addition, forecasting changes in stream water quality in response
to changes in land use type may run the risk that the relationship
would alter over time owing to changes in some specic practices
or the environment itself (Allen, 2004).
4.2. Contribution of the topographic characteristics of land use on
the complex relationship
Topographic characteristics, such as the slope and distance of
each land use type to sub-basins outlets, were identied by many
earlier studies to be important factors affecting the inuence of
land use on stream water quality because they partly determined
collecting routines of pollutants from non-point sources to river
(Wissmar and Swanson, 1990; Wang et al., 1997, 2013).
4.2.1. The effect of slope on the complex relationship
When considering the topographic characteristics, including
average slope and distances to the outlets of each sub-basin, our
results suggested that seasonal variation was also observed in the
complex relationship at the class level. As shown in Table 5, the rela-
tionship in dry seasons was stronger overall than that of the rainy
season for all six land use variables, particularly for forestland and
urban land. It is possible that the signicant correlation of urbaniza-
tion with water quality during dry seasons may be associated with
dominant pollutants emission pathways. As mentioned, urban land
was considered as a proxy for point source (Karr, 1999; Ahearn et al.,
2005; Bu et al., 2014). In rainy seasons, precipitation on impervious
surfaces entering stream water through point sources will dilute
the wastewater delivered from wastewater treatment plants, likely
contributing to a weaker relationship between land use and stream
water quality (Elwood et al., 1983).
In addition, our results showed that a higher slope had more sig-
nicant correlations with water quality variables in both dry and
rainy seasons within sub-basins. Richards and others (1996) argued
that the slope was a fundamental parameter to predict the rates
of water ow across surfaces. Therefore, in terms of the stronger
positive correlation with water quality; with an increasing slope,
greater rates of water ow contribute to soil erosion and to the rates
of particulate matter that picks up pollutants, such as NH3 -N, SO4 2 ,
so that the chances of stream water quality to degrade will increase.
However, for forestland, the slope coefcients had signicantly
negative correlations with water quality variables, suggesting that
a low variability in the slope could act as a sink and not a source.
This may be further supported by the fact that topographic features
are not the dominant impact factors on water quality in forest-
lands within the basin. Anbumozhi et al. (2005) argued that within
the watershed, stream water quality varies due to changes in land
use activities in the contributing areas, which prevailed over topo-
graphic features. Accordingly, in our studies, as shown in Pearsons
correlation analysis, the forest area was negatively associated with
stream water quality. Additionally, atter areas are often connected
with developed areas, such as agricultural and urban land, which
were thought to be positively correlated with water quality vari-
ables. Thus, atter areas may discharge more nutrients than steeper
forested hillsides. However, this result was inconsistent with
conclusions made by previous studies (Wissmar and Swanson,
1990; Wang et al., 1997), which indicated that a decreased
deviation in slope was associated with a decrease in pollutant
concentrations in downstream areas. The reason given was that
lower gradient valleys in a downstream area could allow for longer
periods of surface and subsurface water ow across riparian zones.
The increased contact time may increase the effectiveness of veg-
etation in ltering nutrients from runoff (Wang et al., 2013).
Therefore, much attention should be paid to the inuence of slope
on stream water quality, which is vital for the understanding of the
relationship between land use type and water quality.
4.2.2. The effect of proximity on the complex relationship
Several researchers have addressed the issue whether the aver-
age distance of land use to the outlets of sub-basins is another
reasonable predictor of water quality other than land use types
themselves (Osborne and Wiley, 1988; Hunsaker and Levine, 1995;
Johnson et al., 1997; Sponseller et al., 2001; Woodcock et al., 2006).
Our results suggested that for this particular study area, the land
use type near the stream water was generally a better indicator for
effectiveness in affecting water quality. For example, in our study,
the partial rank of the average distance, from the shortest to the
longest, was agriculture, water bodies, grassland and forestland,
respectively, as was the rank of their correlation coefcients. This
conclusion is in agreement with other previous studies (Phillips
and Hanchar, 1996; Christopher et al., 2010; Wang et al., 2013).
Christopher (2010) found that there was a stronger relationship
between water habitat quality characteristics and land use vari-
ables within the 200-m buffer zone than over the entire watershed,
especially for urbanization.
However, one thing should be kept in mind that many fac-
tors will also inuence the relationship between the stream water
quality and surrounding land use besides proximity, and there
were counterintuitive ndings in our results. Although urban land
was closer to sub-basin outlets than agricultural land, the asso-
ciation of agricultural land with water quality was still stronger
than urban land. Other studies have also shown similar results
with the respect of this topic. For instance, Hunsaker and Levine
(1995) concluded that when land use was classied at the level of
the entire catchment, the relationship between land use and water
quality was conversely stronger than if only a 200 or 400 m buffer
strip was considered. Sliva (2001) found that water quality associ-
ated with catchment scale landscape was slightly better than with
buffer landscape using both the multiple regression analysis and
the redundancy analysis.
It is possible that such mixed results about this topic stem from
the fact that in some instances contaminant loadings can be inde-
pendent of land use proximity (Christopher et al., 2010), rather than
a direct reection of adjacent land use. When there are wastewater
treatment plants or underground drainage networks near samp-
ling sites, particularly within urban land, storm water runoff and
efuent from wastewater treatment plants can be piped past the
adjacent land use directly into streams.
4.3. The synergistic effect of geological factors on water quality
The Loess Plateau, which is famous for server soil erosion world-
wide, is located in the northern part of the Jing River catchment
(JRC) and the Beiluo River catchment (BRC), while the Qinlin Moun-
tains, which is well vegetated, meaning less soil erosion happening,
is situated in the southern Wei River catchment (WRC). Therefore, it
is reasonable to say that soil erosion in the JRC and BRC is more com-
mon and server than that in the WRC (Mo, 2008; Qin, 2009; Xu et al.,
2008). Soil erosion is usually related to higher values of phosphate,
but as shown in Table 2, the PO4 3 values in the BRC and the JRC are
lower than that in the WRC (0.4 and 0.62 to 0.88 mg/L). It might be
explainable if the synergistic effect of geological factors is consid-
ered. The soil erosion does not necessarily cause higher phosphate,
though in most situations it does. As shown in the soil map (Fig. 4),
 S. Yu et al. / Ecological Indicators 60 (2016) 202212
Fig. 4. DEM and Soil types map of the Wei River basin (Zuo designed, Zuo et al., 2015).
loess soil (CS1510) is the most widely distributed soil type in the
JRC and the BRC, and the cinnamon soil (CS1111) in the WRC. Previ-
ous studies indicated that dissolved phosphorus (DP), which is the
main phosphorus formation ushed into rivers, is quite different in
the two different soil types. Loess contains only 2.37 mg/kg DP (Li
et al., 1982, 2007), while cinnamon soil is about 10 mg/kg (http://
baike.baidu.com/view/139602.htm), so that the lower PO4 3 in the
JRC and the BRC compared to the WRC could be explainable with
the different orthophosphate backgrounds.
5. Conclusions
The results of this study demonstrate the relationship between
land use types and stream water quality during both dry and rainy
seasons in 44 sub-basins of the Wei River basin, China. In rainy
seasons, water quality variables were signicantly correlated with
agricultural area, forest area, and grassland area as well as with
urban and barren areas at a signicance level of >0.95. During
dry seasons, the statistical analysis presented different correlation
patterns from those in rainy seasons, showing that water quality
variables were more inuenced by all types of land uses. Therefore,
seasonal variation was observed, indicating that the land use type
had a more signicant correlation in dry seasons than that in rainy
seasons on both water quantity and quality.
The topographic characteristics of land use, such as the average
slope and proximity of land use to stream water, were found to be
important factors affecting the contribution of land use to stream
water quality. Further analysis showed that seasonal variation also
occurred in the complex relationship, and steeper land generally
had a stronger inuence on the stream water quality than atter
land. This study does not only considered the relationship between
land use and water quality, but also the relationship between land
use and seasonal variations on the one hand, and land use and
topographic characteristics on the other hand. Thus, the results of
this study support the recommendation that water quality can be
improved by better land use management.
Acknowledgements
This study was supported by the National Science Foundation
of China (51279005). Dr. Xuwang Yin at Dalian Ocean University
is greatly appreciated for organizing four eld surveys to obtain
stream water quality data, as well as his colleagues for their fol-
lowing work in data analysis and the identication of the samples.
211
References
Ahearn, D.S., Sheibley, R.W., Dahlgren, R.A., Anderson, M., Johnson, J., Tate, K.W.,
2005. Land use and land cover inuence on water quality in the last free-
owing river draining the western Sierra Nevada, California. J. Hydrol. 313,
234247.
Allen, J.D., 2004. The inuence of land use on stream ecosystems. Annu. Rev. Ecol.
Evol. Syst. 35, 257284.
Amiri, B.J., Nakane, K., 2009. Modeling the linkage between river water quality and
landscape metrics in the Chugoku District of Japan. Water Resour. Manage. 23,
931956.
Anbumozhi, V., Radhakrishnan, J., Yamaji, E., 2005. Impact of riparian buffer zones
on water quality and associated management considerations. Ecol. Engine 24,
517523.
Bu, H., Meng, W., Zhang, Y., Wan, J., 2014. Relationships between land use patterns
and water quality in the Taizi River basin, China. Ecol. Indic. 41, 187197.
Buck, O., Niyogi, D.K., Townsend, C.R., 2004. Scale-dependent of land use effects
on water quality of streams in agricultural catchments. Environ. Pollut. 130,
287299.
Chattopadhyay, S., Asa Rani, L., Sangeetha, P.V., 2005. Water quality variations as
linked to landuse pattern: a case study in Chalakudy river basin, Kerala. Curr.
Sci. 89 (12), 21632169.
Christopher, P.T., Robert, W.B., Alexander, J.S., et al., 2010. Land-use proximity as a
basis for assessing stream water quality in New York State (USA). Ecol. Indic. 10,
727733.
Clark, R.M., Geldreich, E.F., Fox, K., Rice, E.W., Johnson, C.H., Goodrich, J.A., Barnick,
A., Abdesaken, F., 1996. Tracking a Salmonella Serovar typhimiirium outbreak
in Gideon Missouri: role of contaminant propagation modeling. J. Water Supply
Res. Technol. Aqua 45 (4), 171183.
Elwood, J.W., Newbold, J.D., ONeil, R.V., Van Winkle, W., 1983. Resource spiraling: an
operational paradigm for analyzing lotic ecosystems. In: Fontaine, T.D., Bartell,
S.M. (Eds.), Dynamics of Lotic Ecosystems. Ann Arbor Science, Ann Arbor, MI, pp.
327.
Eadie, W.T., Drijard, D., James, F.E., Roos, M., Sadoulet, B., 1971. Statistical Methods
in Experimental Physics. North-Holland, Amsterdam, pp. 269271, ISBN 0-444-
10117-9.
Galbraith, L.M., Burns, C.W., 2007. Linking land-use, water body type and water
quality in southern New Zealand. Landsc. Ecol. 22, 231241.
Guo, Q., Ma, K., Yang, L., He, K., 2010. Testing a dynamic complex hypothesis in the
analysis of land use impact on lake water quality. Water Resour. Manage. 24,
13131332.
Herlihy, A., Stoddard, J.L., Johnson, C.B., 1998. The relationship between stream
chemistry and watershed land cover data in the Mid-Atlantic region, USA. Water
Air Soil Pollut. 105, 377386.
Hesel, D.R., Hirsch, R.M., 1995. Statistical Methods in Water Resources. Elsevier
Science.
Huang, J.L., Huang, Y.L., Li, Q.S., Zhou, Z.R., Feng, Y., Zhang, Z.Y., 2012. Preliminary
analysis of spatiotemporal variation of water qualtiy and its inuencing factors
in the Jiulong River Watershed. Environ. Sci. 33 (4), 10981107.
Huang, J.L., Li, Q., Pontius Jr., R.G., Klemas, V., Hong, H., 2013. Detecting the dynamic
linkage between landscape characteristics and water quality in a subtropical
coastal watershed, southeast China. Environ. Manage. 51 (1), 3244.
Hunsaker, C.T., Levine, D.A., 1995. Hierarchical approaches to the study of water
quality in rivers. BioScience 45, 193202.
Hunter, P.R., 2003. Climate change and waterborne and vector-borne disease. J. Appl.
Microbiol. 94(S), 37S46S.
Hurley, T., Mazumder, A., 2013. Spatial scale of land-use impacts on riverine drinking
source water quality. Water Resour. Res. 49, 15911601.
Johnson, L., Richards, C., Host, G., Arthur, J., 1997. Landscape inuences on water
chemistry in Midwestern stream ecosystems. Freshw. Biol. 37, 193208.
212 S. Yu et al. / Ecological Indicators 60 (2016) 202212
Johnson, L.B., Gage, S.H., 1997. Landscape approaches to the analysis of aquatic
ecosystems. Freshw. Biol. 37, 113132.
Kang, Y., Yue, L.P., Xu, Y., Sun, B., 2008. Impact of petroleum exploration on water
environment in the upper reaches of North Luohe River in Northern Shanxi. J.
Catastrophol. 23 (3), 7175 (in Chinese).
Karr, J.R., 1999. Dening and measuring river health. Freshw. Biol. 41 (2), 221
234.
Lee, S.W., Hwang, S.J., Lee, S.B., Hwang, H.S., Sung, H.C., 2009. Landscape ecological
approach to the relationships of land use patterns in watersheds to water quality
characteristics. Landsc. Urban Plan. 92, 8089.
Li, D.X., Wang, M.L., Gao, M.Y., 1982. Phosphorus background values and it fertilizer
effect in Loess. Shanxi J. Agric. Sci. 28, 1719 (in Chinese).
Li, Y.Y., Shao, M.A., Zheng, J.Y., Li, Q.F., Zhang, X.C., 2007. Experimental study on the
impacts of rainfall intensity on phosphorus loss from loessial slope land. Trans.
Chin. Soc. Agric. Eng. 23, 3946 (in Chinese).
Kirk, R.E., 1995. Experimental Design: Procedures for the Behavioral Sciences.
Brooks/Cole, Pacic Grove, CA, USA.
Mo, L., 2008. Runoff cost of Sediment Control of Soil and Water Conservation Prac-
tices with Hydrology Methods in Beiluo River basin. Northwest A&F University
(in Chinese).
Naiman, R.J., Decamps, H., 1997. The ecology of interfaces: Riparian zones. Annu.
Rev. Ecol. Syst. 28, 621658.
Ngoye, E., Machiwa, J.F., 2004. The inuence of land-use patterns in the Ruvu
river watershed on water quality in the river system. Phys. Chem. Earth 29,
11611166.
Osborne, L.L., Wiley, M.J., 1988. Empirical relationships between land use/cover
and stream water quality in an agricultural watershed. J. Environ. Manage. 26,
927.
Pearson, K., 1895. Notes on regression and inheritance in the case of two parents.
Proc. R. Soc. Lond. 58, 240242.
Phillips, P.J., Hanchar, D.W., 1996. Water-Quality Assessment of the Hudson River
Basin in New York and Adjacent States Analysis of Available Nutrient, Pesti-
cide, Volatile Organic Compound and Suspended-Sediment Data. U.S. Geological
Survey Water-Resources Investigations Reports, pp. 96105.
Qin, W., 2009. Characteristics of Soil Erosion and its response to Re-vegetation in
the Upper Reaches of Beiluo River. Beijing Forestry University, pp. 101132 (in
Chinese).
Reimann, C., Finne, T.E., Nordgulen, ., Saether, O.M., Arnoldussen, A., Banks, D.,
2009. The inuence of geology and land-use on inorganic stream water quality
in the Oslo region, Norway. Appl. Geochem. 24, 18621874.
Richards, C., Johnson, L.B., Host, G.E., 1996. Landscape scale inuences on stream
habitats and biota. Can. J. Fish. Aquat. Sci. 53, 295311.
Sliva, L., Williams, D.D., 2001. Buffer zone versus whole catchment approaches
to studying land-use impact on river water quality. Water Res. 35 (14),
34263472.
Song, J.X., Xu, Z.X., Liu, C.M., Li, H.E., 2007. Ecological and environmental stream
ow requirements for the Wei Riverthe largest tributary of the Yellow River.
Hydrol. Process. 21 (8), 10661073 (in Chinese).
Sponseller, R.A., Beneld, E.F., Valett, H.M., 2001. Relationships between land use,
spatial scale and stream macroinvertebrate communities. Freshw. Biol. 46,
14091424.
Wissmar, R.C., Swanson, F.J., 1990. Landscape Disturbances and Lotic Ecotones. The
Ecology and Management of Aquatic-Terrestrial Ecotones. Parthenon Publishing
Group, Carnforth, UK, pp. 6589.
Wang, G.Q., Yinglan, A., Xu, Z.X., Zhang, S.R., 2013. The inuence of land use patterns
on water quality at multiple spatial scales in a river system. Hydrol. Process.,
http://dx.doi.org/10.1002/hyp.10017
Wang, L.W., Lyons, J., Kanehl, P., Gatti, R., 1997. Inuences of watershed landuse on
habitat quality and biotic integrity in Wisconsin streams. Fisheries 22, 612.
Woodcock, T., Mihuc, T., Romanowicz, E., Allen, E., 2006. Land-use effects on
catchment- and patch-scale habitat and macroinvertebrate response in the
Adirondack Uplands. Am. Fish. Soc. Symp. 48, 395411.
Xu, W.M., Li, Y.N., Lian, Z.M., Wang, W.Q., 2008. Countermeasure of biology restora-
tion for degenerated vegetation of the Northern Luohe River Basin. J. Northwest
For. Univ. 23 (5), 5154 (in Chinese).
Ye, Y., He, X., Chen, W., Yao, J., Yu, S., Jia, L., 2014. Seasonal water quality upstream
of Dahuofang Reservoir, China the effects of land use type at various spatial
scales. CLEAN Soil Air Water 42, 14231432.
Zhao, J., 2008. Landscape Pattern Change and Its Environmental Response Across
Multiple Spatial Scales in Tidal Plain. East China Normal University, China, 107
pp. (in Chinese).
Zuo, D.P., Xu, Z.X., Peng, D.Z., Song, J.X., Cheng, L., Wei, S.K., Abbaspour, K., Yang, H.,
2015. Simulating spatiotemporal variability of blue and green water resources
availability with uncertainty analysis. Hydrol. Proc. 29, 19421955.