Journal of Environmental Management 213 (2018) 309e319

Contents lists available at ScienceDirect

Journal of Environmental Management

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

Research article

Assessing the impacts of land use on downstream water quality using

a hydrologically sensitive area concept
Subhasis Giri a, *, Zeyuan Qiu b, Zhen Zhang c
a
School of Environmental and Biological Sciences, Rutgers University, New Brunswick, NJ 08901, USA
b
Department of Chemistry and Environmental Science, New Jersey Institute of Technology, University Heights, Newark, NJ 07102, USA
c
Data Science and Informatics, DowDuPont, Indianapolis, IN 46268, USA

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

Article history: Understanding the relationship between land use and water quality is essential to improve water quality
Received 27 October 2017 through carefully managing landscape change. This study applies a linear mixed model at both water-
Received in revised form shed and hydrologically sensitive areas (HSAs) scales to assess such a relationship in 28 northcentral New
16 February 2018
Jersey watersheds located in a rapidly urbanizing region in the United States. Two models differ in terms
Accepted 21 February 2018
of the geographic scope used to derive land use matrices that quantify land use conditions. The land use
matrices at the watershed and HSAs scales represent the land use conditions in these watersheds and
their HSAs, respectively. HSAs are the hydrological “hotspots” in a watershed that are prone to runoff
Keywords:
Hydrologically sensitivity areas
generation during storm events. HSAs are derived using a soil topographic index (STI) that predicts
Soil topographic index hydrological sensitivity of a landscape based on a variable source area hydrology concept. The water
Linear mixed model quality indicators in these models are total nitrogen (TN), total phosphorus (TP) and total suspended
Land use matrix solids (TSS) concentrations in streams observed at the watershed outlets. The modeling results suggest
Water quality that presence of low density urban land, agricultural land and wetlands elevate while forest decreases
Northcentral New Jersey TN, TP and/or TSS concentrations in streams. The watershed scale model tends to emphasize the role of
agricultural lands in water quality degradation while the HSA scale model highlights the role of forest in
water quality improvement. This study supports the hypothesis that even though HSAs are relatively
smaller area compared to watershed, still the land uses within HSAs have similar impacts on down-
stream water quality as the land uses in entire watersheds, since both models have negligible differences
in model evaluation parameters. Inclusion of HSAs brings an interesting perspective to understand the
dynamic relationships between land use and water quality.
© 2018 Elsevier Ltd. All rights reserved.

1. Introduction
Rapid urbanization has significantly transformed the landscape
in the U.S. during last four decades (USDA, 2015). New Jersey being
the most densely populated state in the U.S. represents such a
typical landscape transformation. From 1986 to 2012, approxi-
mately 29 percent increase in urban lands was observed in New
Jersey, accompanied by 26.7 percent decrease in agricultural lands,
6.7 percent decrease in forest, and 5.4 percent loss in wetlands
(Lathrop et al., 2016).
Urbanization increases impervious surface area and alters
magnitude, volume, frequency, and timing of high streamflow
* Corresponding author.
E-mail address: subhasis.giri@rutgers.edu (S. Giri).
events (Shuster et al., 2005; Walsh et al., 2005), which causes
streambank erosion, modifies channel morphology, transports
nutrients, metals, pharmaceuticals, and toxic substances to streams
(Fitzpatrick et al., 2005; Hatt et al., 2004; and Kolpin et al., 2002),
and directly or indirectly changes hydrological, biological, and
chemical processes of an aquatic ecosystem (Li and Zhang, 2011; Li
et al., 2009; Yu et al., 2013). Water quality degradation has
prompted an increasing interest in better understanding how land
uses in a landscape affect downstream water quality (Huang et al.,
2013; Li et al., 2009; Pratt and Chang, 2012; Wan et al., 2014; Wilson
and Weng, 2010). There are numerous researches that have
attempted to better understand the effects of land use on water
quality (Giri et al., 2012, 2016; Huang et al., 2013; Li et al., 2009;
Nejadhashemi et al., 2011; Wilson and Weng, 2010). Two most
commonly used approaches are biophysical watershed modeling
and statistical modeling. The biophysical watershed modeling

https://doi.org/10.1016/j.jenvman.2018.02.075
0301-4797/© 2018 Elsevier Ltd. All rights reserved.
310 S. Giri et al. / Journal of Environmental Management 213 (2018) 309e319
involves extensive input data, model calibration, and in-depth
modeling knowledge and has become a specialized expertise in
water resource management. Statistical modeling offers a simpler
alternative to biophysical watershed modeling by directly linking
land use conditions and water quality that can be applied efficiently
to a large region. Most statistical modeling studies use non-spatial
land use matrices such as percentage of urban land and impervious
surface cover in a watershed to determine the effects of land use on
downstream water quality (e.g. Comelo et al., 1996; Johnson et al.,
1997). Such non-spatial landscape indicators assume that each
part of landscape has equal potential to affect stream water quality,
which is overly simplistic and might misunderstand the relation-
ship between land use and water quality (Williams et al., 2005).
Two types of progress have been made to overcome the limi-
tation of such simplistic non-spatial landscape representation. First,
an inverse distance weighted statistical approach was developed to
assess the relationship between land use and water quality. This
approach assigned more weights to land uses closer to a stream
than land uses further away from the stream when evaluating the
impacts of land use on stream water quality (Kennen et al., 2008;
Peterson et al., 2010). Second, critical areas such as riparian zones
of streams were considered to have the most significant impacts on
stream water quality (NRC, 2002). The land uses within the riparian
zones instead of whole watershed were consequently used to
assess the impacts of land use on stream water quality (Baker et al.,
2006; Paringit and Nadaoka, 2003). Although the riparian zone
approach is interesting, it has some drawbacks (Qiu, 2009). First,
there is no uniform way of defining the width of riparian zones.
Second, riparian zone is not a surrogate measure of hydrological
sensitivity considering spatially varying hydrological connectivity
in landscape.
We use a concept called hydrologically sensitive areas (HSAs) to
improve the understanding of hydrological connectivity between
terrestrial landscapes and aquatic stream. HSAs are the areas in a
watershed having higher propensity to generate runoff. This is
consistent with the variable source area (VSA) hydrology concept
where the primary source of runoff is saturated areas in uplands,
whose scale varies depending on storm intensity (Qiu et al., 2014;
Walter et al., 2000). These smaller saturated areas in a watershed
primarily generate and transport pollutants to streams and influ-
ence stream hydrographs (Hewlett, 1982).
Although some studies (e.g. Easton et al., 2008; Heathwaite
et al., 2005; Qiu, 2009; Walter et al., 2000, 2009) have used VSA
hydrology to identify HSAs and target best management practices
(BMPs) to control nonpoint source pollution, however, they were
conducted on agricultural field or small watershed scale with im-
plicit assumption that BMPs within HSAs would be more effective
in improving water quality. No study has attempted to empirically
test and validate whether the high-intensity land uses within HSAs
such as agricultural and urban lands contribute more to water
quality degradation across watersheds. The objectives of this study
is to assess the impact of land use at both HSAs and watershed scale
on water quality using a linear mixed model and to test the hy-
pothesis that even though HSAs represent a small fraction of a
watershed, the land uses within HSAs have the similar impacts on
downstream water quality as the land uses in the watershed.
Validation of such a hypothesis is a critical step to the development
of efficient watershed management strategies for water quality
improvement. For example, watershed manager can strategically
target HSAs for implementing BMPs to enhance their effectiveness
and cost-effectiveness in improving water quality if the hypothesis
is deemed to be true.
2. Materials and methods
2.1. Study area
This study was conducted in 28 watersheds located in the
northcentral New Jersey including Essex, Hunterdon, Mercer,
Middlesex, Monmouth, Morris, Somerset, Sussex, Union, Passaic,
Burlington, and Ocean Counties (Fig. 1). All 28 watersheds were
located in three physiographic regions including Valley and Ridge,
Highlands, and Piedmonts where VSA hydrology is considered to be
a dominant hydrological process for runoff generation (Qiu, 2009;
Walter et al., 2002). There were 68 watersheds with the long-
term water quality monitoring stations in the region, but we only
selected these 28 watersheds having no water transfer and located
entirely in New Jersey.
The watersheds were delineated using the Streamstats, a web-
based watershed delineation tool developed and maintained by
the U.S. Geological Survey. Latitude and longitude of each water
quality monitoring station were submitted through the Streamstats
interface to obtain the watershed boundary. An example of delin-
eated watershed using the Streamstats is presented in Fig. 2.
These watersheds vary in size from 24 square kilometers (km2)
to 2062 km2. These watersheds had experienced varying degrees of
urbanization. The primary urbanization form is the expansion of
low density residential areas, accompanied by losses in agricultural
land, forest, and wetlands (Lathrop et al., 2016). Such changes in
land uses have altered watershed hydrology as well as the physical,
chemical, and biological condition of streams in New Jersey
(Kennen et al., 2008, 2010).
2.2. Soil topographic index
Soil topographic index (STI) is an indicator of hydrological
sensitivity of a landscape and is calculated using following equation
(Giri et al., 2017; Buchanan et al., 2014; Qiu, 2009; Walter et al.,
2002):
 
a
STI ¼ ln (1)
T tanðbÞ
where a is the upslope contributing area per unit contour length
(m), b is the local surface slope (mm1), T is a soil transmissivity
(m2/day) computed as the product of the saturated hydraulic
conductivity (m/day) and the depth to a restrictive layer (m). STI
indicates the likelihood of a point in a watershed to generate runoff
and is used to identify spatial distribution of runoff contributing
 
a
areas in watersheds (Qiu, 2009). Wetness index, ln tanð bÞ , was the
most common form of topographic index (Beven and Kirkby, 1979).
STI extends this topographically based wetness index by consid-
ering soil water storage capacity above a restrictive layer and is
more applicable to the hydrological process in the Northeast in the
U.S. (Buchanan et al., 2014; Qiu, 2009; Walter et al., 2002).
2.2.1. Soil transmissivity
Soil transmissivity was based on soil saturated hydraulic con-
ductivity and soil depth (Buchanan et al., 2014) of topsoil layers in
the soil survey geographic (SSURGO) database downloaded from
the U.S. Department of Agriculture (USDA) Geospatial Gateway. The
information on saturated hydraulic conductivity and the soil depth
in the SSURGO database was extracted using a soil data viewer
developed by USDA Natural Resources Conservation Service. This
saturated hydraulic conductivity for calculating T in Equation (1) is
the geometric mean of the saturated hydraulic conductivity of all
soil layers above a restrictive layer (Qiu, 2009). A correction factor
 S. Giri et al. / Journal of Environmental Management 213 (2018) 309e319 311

Fig. 1. Location of the selected watersheds in northcentral New Jersey.

of 0.000864 was multiplied to the soil transmissivity layer to merged into a single soil transmissivity layer for the region, which
convert the unit into m2/day. The soil transmissivity layer was is finally clipped into 28 soil transmissivity layers for those selected
created for each County. All soil transmissivity layers were then watersheds based on the watershed boundaries.
312 S. Giri et al. / Journal of Environmental Management 213 (2018) 309e319
Fig. 2. The U.S. Geological Survey Streamstats interface for watershed delineation.
2.2.2. Wetness index
The wetness index was based on the light detection and ranging
(LiDAR) digital elevation model (DEM) at a 3-m resolution. The
LiDAR DEM for New Jersey was obtained from New Jersey
Department of Environmental Protection (NJDEP). This fine scale
elevation data better represents the complexity of the terrain
condition in the study area (Valley and Ridge, Highlands, and
Piedmonts). Similar to soil transmissivity layers, separate LiDAR
DEMs were created for each of those 28 watersheds by clipping the
statewide LiDAR DEM using the watershed boundaries. Fill, calcu-
lation of slope, and calculation of catchment area were performed
one-at-a-time to generate the wetness index for each watershed. To
ensure a continuous flow in each grid of LiDAR DEM, the depression
was filled using fill sinks (Planchon and Darboux, 2001). Least
Square Fitted Plane method (Horn, 1981; Costa-Cabral and Burges,
1996) was used to calculate slope of each gird while multiple
triangular flow direction method (Seibert and McGlynn, 2007) was
performed to obtain the catchment area of each grid. The wetness
index was generated using the SAGA geographic information sys-
tem in R. The soil transmissivity was then combined to the wetness
index to create the soil topographic index for each of the 28 wa-
tersheds following Equation (1). Fig. 3 presents an example of how
wetness index and soil transmissivity are combined to develop STI
for one of the watersheds in the study area.
2.3. Hydrologically sensitive areas (HSAs)
HSAs are the areas in a watershed having high potential to
generate runoff (Qiu, 2009). Identification of HSAs is essential to
optimize the use of resources for both stakeholders and policy
makers. Qiu (2009) discussed several ways of delineating HSAs in a
watershed and recommended a simple threshold value approach.
Following the approach, the areas in a watershed having the STI
values greater than a threshold value are considered to be HSAs in
that watershed. This study delineated the HSAs using the STI
threshold value of ten, i.e. the areas having the STI values greater
than or equal to ten were the HSAs. The areas of the delineated HSA
made up approximately 27 percent of the watersheds for 28 wa-
tersheds, which is slightly larger than 20 percent, the potential
saturation areas under typical topographic and climate conditions
suggested by Herron and Hairsine (1998). Different STI threshold
values were assessed in our other watershed studies with signifi-
cant local stakeholder engagement in the region and the threshold
value of ten seemed to delineate HSAs that are empirically verifi-
able and administratively manageable (Qiu, 2015; Qiu et al., 2014).
2.4. Land use matrix
Land use matrices for these watersheds and their HSAs was
extracted from the 2007 land use/cover data maintained by NJDEP
to understand the relationship between land use and water quality.
The 2007 land use/cover data were produced by visually inter-
preting color infrared aerial photography. Through this process,
photo interpreters examine each image and classify the image into
various land use/cover categories based on a modified Anderson
classification system (NJDEP, 2010). The areas of various land use
categories were extracted from the 2007 land use/cover data
(agricultural land, forest, high density urban land, medium density
urban land, low density urban land, rural residential, wetlands, and
water) for these watershed and their HSAs. High density urban land
is located at densely populated urban areas having either high
density single units or multiple units on one eighth to one fifth acre
lots (NJDEP, 2007). Medium density urban land is found in urban/
suburban areas and consists of residential units on greater than one
eighth acre up to half acre lots while low density urban land con-
tains single residential unit on greater than half acre to one acre lots
(NJDEP, 2007). Rural residential consists of single residential unit
between one acre to two acre lots (NJDEP, 2007). Land use matrices
include the percentage of each type of land use in watersheds and
their HSAs. High density urban land and medium density urban
land were then combined into high medium density urban land to
capture the effect of urbanization with higher land use intensity on
 S. Giri et al. / Journal of Environmental Management 213 (2018) 309e319 313

Fig. 3. Spatial distribution of (a)wetness index, (b)soil transmissivity and (c) soil topographic index for a selected watershed.

water quality. landscape dominated by various urban development such as

Table 1 presents the descriptive statistics on the percentages of commercial and industrial complex and high, medium, and low
various land use types in watersheds and their HSAs. Land uses density residential areas. These watersheds had 36 percent of for-
varied significantly across those watersheds from the pristine est, 34 percent urban, 13 percent agriculture, 14 percent wetlands,
landscape dominated by forests and wetlands to the urbanizing and 3 percent water; and their HSAs were made of 30 percent
314 S. Giri et al. / Journal of Environmental Management 213 (2018) 309e319
Table 1
Descriptive statistics of percentages of land uses in watersheds and HSAs.
Land use (%) Total N In Watershed
Mean
Agricultural land 28 13.43
Forest 28 36.11
High medium density urban land 28 20.26
Low density urban land 28 3.91
Rural residential 28 9.72
Water 28 2.08
Wetlands 28 14.49
forest, 28 percent urban, 13 percent agriculture, 22 percent wet-
lands, and 7 percent water, where urban included high, medium,
and low density urban lands and rural residential.
2.5. Water quality monitoring data
The locations of the water quality monitoring stations were
downloaded from NJDEP and the water quality monitoring data
including total nitrogen (TN), total phosphorus (TP) and total sus-
pended solid (TSS) for each station were obtained from National
Water Quality Monitoring Council data portal. Available water
quality monitoring data for each station was downloaded and was
expressed as concentration (mg/l). The downloaded water quality
data were further processed to keep the data only from 2006 to
2008 for each station for statistical analysis as described below to
assess the effects of the 2007 land use conditions on water quality.
The water quality monitoring data in three years (2006e2008; one
prior and one after) instead of only one year (2007) were used to
capture the broad impacts of land use on water quality due to un-
known and uncertain land use and climate conditions. The water
quality at these monitoring stations were sampled 4e5 times a year
depending the available funding in New Jersey. There were 12e15
water quality measurements per site during the three-year period.
2.6. Statistical analysis
A statistical analysis was conducted to understand the rela-
tionship between the land use matrices in the watersheds and the
HSAs to the water quality measurements (TSS, TN, and TP) observed
at the watershed outlets. The land use matrix was a set of predictors
and the water quality measurements were response variables. A set
of predictors (i.e. the land use matrix in 2007) would correspond to
multiple responses (i.e. several water quality measurements
observed from 2006 to 2008) and the number of responses would
also vary across watersheds because different numbers of moni-
toring could be conducted at different monitoring states. Therefore,
unequal repeated measures (the water quality data) were modeled
in R using the following linear mixed model for each of the three
water quality indicators at both watershed and HSA scales:
Yij ¼ bX i þ ui þ εij
where i was the index for watershed and j was the index for the
number of measurements on water quality varying by watershed.
Yij was the observed water quality measurement for watershed i,
i ¼ 1, …, 28, on its jth measurement, j ¼ 1, 2, …, ni as repeated
measures; Xi was a land use matrix, i.e., the percentage of
agricultural land, forest, high medium density urban land, low
density urban land, rural residential, wetlands, and water in
watershed i or its HSAs, b represented the fixed effects of these
predictors, ui represents the random effect due to the unique
In HSAs
Minimum Maximum Mean Minimum Maximum
0.06 34.10 12.98 0.11 33.07
8.19 80.94 30.36 7.37 64.62
1.23 72.63 17.30 1.28 73.15
0.46 9.07 2.77 0.41 6.57
1.96 21.66 7.14 1.63 15.94
0.35 6.43 7.24 0.80 19.73
1.66 31.45 22.19 3.38 42.88
characteristics of watershed i, and εij is the residual. We
assumed that εij ~ N (0, s2) and ui ~ N (0, t2) for this analysis with
t2 capturing the variation due to different watershed charac-
teristics, and s2 measuring the unexplained variation.
A Log transformation of water quality data was performed to
avoid the impacts of the skewness in the response variables in
statistical analysis. The predictors were expressed in percentage
and rural residential is considered as baseline. The parameters in
the model were estimated using lme function in nlme package by
Maximum likelihood method in R. A backward stepwise elimina-
tion of predictors was performed using Akaike information crite-
rion (AIC) based on an initial regression analysis to assess the
individual predictors' influence for each water quality indicator. In
the initial step, all predictors were added in the regression equation
and AIC was calculated. In the next step, one of the predictor was
dropped from the regression equation and a new AIC was calcu-
lated. If the new AIC is smaller than the previous AIC, then the
dropped predictor was removed permanently from the regression
equation. However, if the new AIC is greater than the previous AIC,
the dropped predictor was added back to the regression equation.
This process would repeat until minimum AIC was obtained. Two
additional model evaluation measures, Bayesian information cri-
terion (BIC) and log-likelihood (LogLik), were also used to deter-
mine which land use matrix (within HSAs or whole watershed)
would project a stronger relationship between land use and water
quality. To visualize the data and assess their relationship for
selecting predictors, Spearman's rank correlation coefficient, scat-
terplot, and histogram were also plotted prior to the linear mixed
modeling.
3. Results and discussion
3.1. Visualization of data and correlations
Fig. 4 and Fig. 5 present the Spearman's correlation matrices
among water quality indicators (TN, TP, and TSS) and all land use
categories for the watersheds and their HSAs, respectively. Spear-
man's rank correlation coefficients are presented in the lower
panel, scatter plots are located in the upper panel, and histograms
(2) are presented in the diagonal. The line in the scatter plot represents
the simple linear regression between a pair of two variables. The
correlation between two variables is presented by Spearman's rank
correlation coefficients while the distribution of data is visualized
by histogram. According to Mukaka (2012), the absolute value of
the coefficient below 0.30 indicates negligible correlation, between
0.30 and 0.50 low correlation, between 0.50 and 0.70 moderate
correlation and above 0.70 high correlation.
Fig. 4 indicates there are positive moderate correlation between
TN and both high medium density and low density urban lands,
negative low correlation between TN and forest, and negligible
correlation between TN and other land uses including agricultural
 S. Giri et al. / Journal of Environmental Management 213 (2018) 309e319
Fig. 4. Spearman's Correlation Matrix for water quality indicators and land uses in watersheds: A for agricultural land, B high medium density urban land, C low density urban land,
D water, E wetlands, F forest, and G rural residential; Bold numbers indicate that coefficients are statistically significant at 10 percent level of confidence.
lands, water, wetlands and rural residential at the watershed scale.
The similar relationships between TN and land uses are observed at
the HSA scale except that the correlation between TN and low
density urban land, which is downgraded from moderate to low
(Fig. 5).
The correlation analyses also identify positive moderate corre-
lation between TP and both high medium density and low density
urban lands, negative moderate correlation between TP and forest,
and negligible correlation between TP and other land uses at the
watershed scale (Fig. 4). Similarly, the correlation between TP and
low density urban land is downgraded from moderate to low at the
HSA scale (Fig. 5). There are negligible correlations between TSS
and all land uses at both watershed and HSA scales.
3.2. Relationship between TN and land use matrix
Table 2 presents the fixed effect coefficient of different land uses
on TN concentration in streams along with their p-values and three
model evaluation parameters at both watershed and HSA scales. At
the watershed scale, both agricultural land and low density urban
land have significant and positive impacts on TN concentration. The
results indicate that the presence of low density urban land in
watersheds tends to have greater impacts on TN concentration than
agricultural land. For an example, one percent increase in agricul-
tural land in a watershed would increase TN concentration by 0.263
percent while one percent increase in low density urban land in a
watershed would elevate TN concentration by 0.424 percent. High
medium density urban land and wetlands seem to have positive
impacts on TN concentration, however, such impacts are deemed to
be statistically insignificant.
The HSA scale model drops high medium density urban land
315
and adds forest in its contributing land use matrix. Unlike the high
medium density urban land at the watershed scale that has posi-
tive, but insignificant impact on TN concentration, forest has
negative and insignificant impact on TN concentration in streams at
the HSA scale. Similar to the watershed-scale model, the HSA scale
model shows that agricultural land and low density urban land
have positive and significant impacts, and wetland has positive, but
insignificant impacts on TN concentration (Table 1).
The selection of the land use categories as predictors are eval-
uated by using three evaluation parameters (AIC, BIC, and Loglik) as
discussed above. Smaller AIC and BIC and larger Loglik value
represent better model prediction. As shown in Table 2, the
watershed scale model seems to have better predicting power than
the HSA scale model; however, the differences are generally very
small and mostly negligible.
The significant impacts of agricultural land and low density
urban land on TN concentration are understandable. Nitrogen fer-
tilizer application in agricultural land is often one of the primary
sources to the elevated TN concentration in streams. The low
density urban land is a predominant urban development in New
Jersey (Lathrop et al., 2016). Higher TN concentration from the low
density urban land may be due to fertilizer application in lawn care
and presence of animal waste such as horses in the study area.
Another possible explanation of higher TN concentration from low
density urban land may be due to legacy nitrogen in the form of
biogeochemical nitrogen such as retention of nitrogen in the root
zone from past agricultural activities. As most low density urban
lands in New Jersey were converted from agricultural land. Such
biogeochemical legacy nitrogen might act as a long-term source of
mineralization and leaching to waterways resulting in increased TN
concentration in streams (Van Meter et al., 2016). Tsegaye et al.
316 S. Giri et al. / Journal of Environmental Management 213 (2018) 309e319

Fig. 5. Spearman's Correlation Matrix for water quality indicators and land uses in HSAs: A.1 for agricultural land, B.1 high medium density urban land, C.1 low density urban land,
D.1 water, E.1 wetlands, F.1 forest, and G.1 rural residential; Bold numbers indicate that coefficients are statistically significant at 10 percent level of confidence.

Table 2 Table 3
Land uses and their associated fixed effects and the model evaluation statistics on TN Land uses and their associated fixed effects and the model evaluation statistics on TP
concentration in streams. concentration in streams.

Predictors Watershed Scale HSA Scale Model Predictors Watershed Scale HSA Scale Model
Model Model

b-value p-value b-value p-value b-value p-value b-value p-value

*** ***
Intercept 0.308 0.000 0.304 0.000 Intercept 2.821*** 0.000 2.859*** 0.000
Agricultural land 0.263** 0.017 0.205* 0.085 Agricultural land 0.301* 0.066 0.143 0.434
Low density urban land 0.424*** 0.006 0.336** 0.026 Low density urban land 0.683*** 0.000 0.401* 0.085
High medium density urban land 0.033 0.811 Wetland 0.275** 0.039 0.293* 0.077
Wetland 0.053 0.536 0.090 0.375 Forest 0.272 0.181
Forest 0.108 0.391 Model Evaluation Statistic
Model Evaluation Statistic AIC 729.73 734.65
AIC 349.37 351.01 BIC 753 761.8
BIC 375.21 376.84 Loglik 358.86 360.32
Loglik 167.68 168.50 *** **
significant with at least 1 percent level of confidence; 5 percent level of con-
*** **
significant with at least 1 percent level of confidence; 5 percent level of con- fidence; and * 10 percent level of confidence.
fidence; and * 10 percent level of confidence.

concentration in streams along with their p-value at both water-

(2006) found that agriculture and urban were the primary source of
nitrogen in Wheeler Lake Basin in northern Alabama and southern
Tennessee in U.S. Wilson and Weng (2010) made the similar
observation in Lake Calumet Watershed in the Greater Chicago
region. Tu (2011) found that forest had negative relationship with
nitrogen concentration in the stream in Boston metropolitan area
in eastern Massachusetts. However, none of the previous studies
made further distinction on the impact of different types of urban
lands on TN concentration in streams.
3.3. Relationship between TP and land use matrix
Table 3 presents the fixed effect coefficients of land uses on TP
shed and HSA scales. The watershed scale modeling results suggest
that agricultural land, low density urban land, and wetland have
positive and significant impacts on TP concentration. In a water-
shed, one percent increase in low density urban land would elevate
TP concentration by 0.683 percent while one percent increase in
agricultural land would increase TP concentration by 0.301 percent,
and one percent increase in wetlands also increase TP concentra-
tion by 0.275 percent. The positive and significant contribution of
wetland to TP concentration in streams is counter-intuitive. One
possible explanation is that wetlands can absorb or release nutri-
ents depending on their nutrient saturation status. For example,
Ardon et al. (2010) found that wetland produced greater soluble
 S. Giri et al. / Journal of Environmental Management 213 (2018) 309e319
reactive phosphorus and TP loads to streams than agricultural land
in a Timber Lake restoration project in North Carolina. In the study
area, the nutrient storing capacity of wetland might be surpassed,
resulting in the release of phosphorus into streams.
The HSA scale model shows the similar result as in the water-
shed scale model with a few exceptions. First, agricultural land
shows positive, but insignificant impacts on TP concentration.
Second, forest is added into the contributing land use matrix and
has a negative, but insignificant impact on TP concentration.
This TP modeling results are similar to the results found in other
studies. Pratt and Chang (2012) and Wan et al. (2014) observed that
agricultural land and urban land were the primary contributor to
phosphorus loading to streams in the Portland metropolitan area in
Oregon and Xitiaoxi River Watershed in China, respectively. Tu
(2011) found that the presence of forest significantly decreased
phosphorus concentration in streams around the Boston metro-
politan area in eastern Massachusetts.
Similarly, the watershed scale model shows a better predicting
power than the HSA scale model because of the better model
evaluation statistics, i.e. lesser AIC and BIC, and greater Loglik
values, but the differences in those values are small and negligible
(Table 3).
3.4. Relationship between TSS and land use matrix
The selected land use predictors and their fixed effect on TSS
concentration in streams, and associated p-values are presented in
Table 4. The watershed scale model shows agricultural land, forest,
and high medium density urban land has negative impacts on TSS
concentration in streams. The positive impacts of forest and high
medium density urban land on TSS concentration are significant.
Trees and shrubs in forest generally reduce soil erosion. High per-
centage of impervious surface area in high medium density urban
land density also helps prevent soil erosion. The impact of agri-
cultural lands on TSS concentration in streams is negative, but
insignificant in the study area.
The HSA scale model drops out agricultural land and only in-
cludes forest, and high medium density urban land in its contrib-
uting land use matrix. Like the watershed scale model, the presence
of both forest and high medium density urban land density in HSAs
reduce TSS concentration in streams. The impacts of forest are
significant while the impacts of high medium density urban land
density are insignificant at the HSA scale (Table 4).
Both watershed and HSA modeling results agree that the pres-
ence of forest significantly reduces TSS concentration in streams,
but disagree in every other aspect, which could be due to the
insignificant source contribution from upland in the study area.
First, TSS concentration in streams is generally not significant water
Table 4
Land uses and their associated fixed effects and the model evaluation statistics for
TSS concentration in stream.
Predictors Watershed Scale HSA Scale Model
Model
b-value p-value b-value
Intercept 1.302 ***
0.000 1.303
Agricultural land 0.248 0.108
Forest 0.505*** 0.009 0.254
High medium density urban land 0.446** 0.036 0.161
Model Evaluation Statistic
AIC 657.55 658.42
BIC 678.61 675.97
Loglik 322.77 324.21
***
significant with at least 1 percent level of confidence; **
5 percent level of con-
fidence; and * 10 percent level of confidence.
317
quality issue in the study area (NJDEP, 2014). Second, TSS concen-
tration in streams may primarily be due to the instream processes
rather than overland processes in the study area. Qiu and Wang
(2014) found that more than 60 percent of the sediment load to
streams came from stream banks erosion and streambed sediment
resurfacing during large storm events in Neshanic River Watershed,
one of these 28 selected watersheds in this study. Furthermore,
Nagel et al. (2007) also found that more than 50 percent of sedi-
ment loads in Cayuga and Catskills Watersheds in New York was
due to stream bank erosion where VSA hydrology was the domi-
nant process for runoff formation. The model evaluation statistics
shows the mixed results regarding the predicting power of the two
models. The watershed scale model shows better predicting power
in terms of AIC and Loglik and worse in terms of BIC, but their
differences are small and negligible.
4. Conclusion
Understanding the effect of land use on water quality is vital to
sustainable water resources management. This study assessed the
relationship between land use and water quality in watersheds
located in northcentral New Jersey using a regional linear mixed
model. Specifically, this study tried to identify which types of land
uses contribute to stream water quality degradation. The response
variables were TN, TP, and TSS concentration whereas the pre-
dictors were agricultural land, forest, high medium density urban
land, low density urban land, rural residential, wetlands, and water.
The analysis was performed at both watershed and HSA scales
where HSAs of each watershed were delineated using STI. The areas
in a watershed having STI greater than ten were considered to be
HSAs in that watershed.
Although urbanization is considered to be a major cause of
stream water quality degradation, but no study further delineates
the impacts of different urbanization forms on water quality. This
study first attempts to consider how different types of urban lands
affect water quality. In this study, the urban land was further
classified into high medium density urban land and low density
urban land to assess their impacts on water quality. The results at
both watershed and HSA scales suggested that low density urban
land significantly contributed to elevated TN and TP concentration
in streams. The watershed scale models also implied that high
medium density urban land increased TN concentration, but
significantly reduced sediment loads to streams. The HSA scale
models did not identify the significant impacts of other forms of
urban land uses except the positive and significant impact of low
density urban land on TN and TP. The modeling results also sug-
gested that agricultural land and wetlands increased while forest
reduced TN, TP or TSS concentration in streams with varying levels
of statistical significance.
This is the first study that tries to compare the impacts of land
uses on water quality between the watershed and HSA scales. The
modeling results suggested that the predicting power of linear
mixed models both at HSA and watershed scale were similar
because of negligible difference in three model evaluation statistics.
This supports our hypothesis that even though HSAs represent
p-value much smaller areas compared to watershed, the land uses in HSAs
***
0.000 have similar impacts on downstream water quality as the land uses
in entire watershed do. Furthermore, the HSA scale modeling re-
**
0.020 sults show some interesting insights into the dynamic relationship
0.132
between land use and water quality. For example, the HSA scale
models downplayed the impacts of agriculture on water quality.
Both models identified agricultural land as a significant contributor
to an increase in TN concentration in streams, but the HSA scale
model showed that agricultural land insignificantly contributed to
increase in TP concentration and dropped out agricultural land as a
318 S. Giri et al. / Journal of Environmental Management 213 (2018) 309e319
potential contributor to TSS concentration. The HSA scale model
concluded that forest helped reduce TN, TP and TSS concentration
in streams and emphasized the positive impacts of forest in water
quality improvement.
This study supports the hypothesis that land uses in HSAs have
similar or comparable significant impacts on in-stream water
quality as the land uses in entire watersheds do, since the differ-
ences in model evaluation statistics between the watershed and
HSA scale models are very small and negligible. The findings would
facilitate the scientific state-of-the-art management strategies for
watershed protection. Watershed managers might focus on HSAs in
a watershed rather than the whole watershed to implement BMPs
for improving water quality. For example, implementation of con-
servation practices are required to control nonpoint source pollu-
tion from agriculture, however, it is not economically feasible to
implement conservation practices in all agricultural lands in a
watershed. Alternatively, conservation practices can be imple-
mented in the identified agricultural lands within HSAs to optimize
available resources. Based on this study, policymakers should target
agricultural lands and low density urban land in HSAs to reduce TN
concentration in streams. Additionally, in order to reduce TSS
concentration in streams, forest in HSAs should be strategically
protected. The identified HSAs in a watershed should also be pro-
tected from urban development in order to maintain landscape
integrity in these sensitive areas to maintain healthy watershed
conditions. This study helps address two practical questions faced
by watershed managers: 1)where in a watershed should be tar-
geted for implementing BMPs? 2) which specific types of land uses
such as different types of urban land should be aimed for man-
agement to improve water quality in a watershed? The findings of
this study would help prioritize different parts of a watershed for
conservation effort as well as direct urban development away from
the sensitive portions of a watershed. This will bring tremendous
environmental benefits to a society. In case of New Jersey where
urban sprawl is so profound, the findings can be incorporated into
different municipal land use controls such as steep slope restric-
tion, stream corridor protection, open space, and farmland pres-
ervation to mitigate possible negative consequences of
urbanization on streams (Qiu et al., 2014).
Inclusion of HSAs into assessing the relationships between land
use and water quality invites more scientific inquiries on this
interesting and important research subject. The study itself can be
extended in several different ways in different geographic regions
to enrich the understanding of the relationship between land use
and water quality. First, more watersheds with land use and water
quality monitoring data across different time periods can be
included to further assess the relationship between land use and
water quality. Second, different threshold STI values can be chosen
to delineate HSAs and assess their impacts on the modeling results.
Third, different statistical models such as a Bayesian hierarchical
model instead of a linear mixed model can be used to better un-
derstand the spatio-temporal interaction between land use and
water quality across a region. The better understanding of the
relationship between land use and water quality would help
develop more effective and efficient watershed management
practices and strategies to protect water resources and improve
water quality.
Acknowledgements
This study was partially supported by the USDA National Insti-
tute of Food and Agriculture through an Agriculture and Food
Research Initiative Competitive grant to New Jersey Institute of
Technology (Grant Number 2012-67019-19348).
References
Ardon, M., Montanari, S., Morse, J.L., Doyle, M.W., Bernhardt, E.S., 2010. Phosphorus
export from a restored wetland ecosystem in response to natural and experi-
mental hydrologic fluctuations. J. Geophys. Res. 115, G04031.
Baker, M.E., Weller, D.E., Jordan, T.E., 2006. Improved methods for quantifying po-
tential nutrient interception by riparian buffers. Landsc. Ecol. 21, 1327e1345.
Beven, K.J., Kirkby, M.J., 1979. A physically based, variable contributing area model
of basin hydrology. Hydrol. Sci. Bull. 24 (1), 43e69.
Buchanan, B.P., Fleming, M., Schneider, R.L., Richards, B.K., Archibald, J., Qiu, Z.,
Walter, M.T., 2014. Evaluating topographic wetness indices across central New
York agricultural landscapes. Hydrology Earth Syst. Sci. 18, 3279e3299.
Comelo, R.L., Paul, J.F., August, P.V., Copeland, J., Baker, C., Hale, S.S., Latimer, R.W.,
1996. Relationships between catchment stressors and sediment contamination
in Chesapeake Bay estuaries. Landsc. Ecol. 11, 307e319.
Costa-Cabral, M.C., Burges, S.J., 1996. Digital elevation model networks (DEMON): a
model of flow over hillslopes for computation of contributing and dispersal
areas. Water Resour. Res. 30, 1681e1692.
Easton, Z.M., Walter, M.T., Steenhuis, T.S., 2008. Combined monitoring and
modeling indicate the most effective agricultural best management practices.
J. Environ. Qual. 37, 1798e1809.
Fitzpatrick, F.A., Diebel, M.W., Harris, M.A., Arnold, T.L., Lutz, M.A., Richard, K.D.,
2005. Effects of urbanization on the geomorphology, habitat, hydrology, and
fish index of biotic integrity of streams in the Chicago area, Illinois and Wis-
consin. Am. Fish. Soc. Symposium 47, 87e115.
Giri, S., Nejadhashemi, A.P., Woznicki, S.A., 2012. Evaluation of targeting methods
for implementation of best management practices in the Saginaw River
Watershed. J. Environ. Manag. 103, 24e40.
Giri, S., Qiu, Z., Zhang, Z., 2017. A novel technique for establishing soil topograpghic
index thresholds in defining hydrologically sensitive areas in landscape.
J. Environ. Manag. 200, 391e399.
Giri, S., Qiu, Z., Prato, T., Luo, B., 2016. An Integrated Approach for Targeting Critical
Source Areas to Control Nonpoint Source Pollution in Watersheds. Water Re-
sources Management. https://doi.org/10.1007/s11269-016-1470-z.
Hatt, B.E., Fletcher, T.D., Walsh, C.J., Taylor, S.L., 2004. The influence of urban density
and drainage infrastructure on the concentrations and loads of pollutants in
small streams. Environ. Manag. 34, 112e124.
Heathwaite, A.L., Quinn, P.F., Hewett, C.J.M., 2005. Modelling and managing critical
source areas of diffuse pollution from agricultural land using flow connectivity
simulation. J. Hydrology 304, 446e461.
Herron, N.F., Hairsine, P.B., 1998. A scheme for evaluating the effectiveness of ri-
parian zones in reducing overland flow to streams. Aust. J. Soil Res. 36,
683e698.
Hewlett, J.D., 1982. Principles of Forest Hydrology. University of Georgia Press,
Athens, GA.
Horn, B.K., 1981. Hill shading and the reflectance map. Proc. IEEE 69, 14e47.
Huang, J., Zhan, J., Yan, H., Wu, F., Deng, X., 2013. Evaluation of the impacts of
landuse on water quality: a case study in the Chaohu Lake Basin. Sci. World J. 10,
1e7.
Johnson, L.B., Richards, C., Host, G.E., Arthur, J.W., 1997. Landscape influences on
water chemistry in Midwestern stream ecosystems. Freshw. Biol. 37, 193e208.
Kennen, J.G., Kauffman, L.J., Ayers, M.A., Wolock, D.M., Colarullo, S.J., 2008. Use of an
integrated flow model to estimate ecologically relevant hydrologic character-
istics at stream biomonitoring sites. Ecol. Model. 211, 57e76.
Kennen, J.G., Murray, M.R., Beaulieu, K.M., 2010. Determining hydrologic factors that
influence stream macroinvertebrate assemblages in the northeastern U.S.
Ecohydrology 3, 88e106.
Kolpin, D.W., Furlong, E.T., Meyer, M.T., Thurman, E.M., Zaugg, S.D., Barber, L.B.,
Buxton, H.T., 2002. Pharmaceuticals, hormones, and other organic wastewater
contaminants in U.S. streams, 1999e2000: a national reconnaissance. Environ.
Sci. Technol. 36, 1202e1211.
Lathrop, R.G., Bognar, J.A., Hasse, J.E., 2016. Changing Landscapes in the Garden
State: Land Use Change in NJ 1986-2012. Grant F. Walton Center for Remote
Sensing and Spatial Analysis. Rutgers University, New Brunswick, New Jersey,
08901.
Li, S., Gu, S., Tan, X., Zhang, Q., 2009. Water quality in the upper Han River basin,
China: the impacts of land use/land cover in riparian buffer zone. J. Hazard.
Mater. 165, 317e324.
Li, S., Zhang, Q., 2011. Response of dissolved trace metals to land use/land cover and
their source apportionment using a receptor model in a subtropic river, China.
J. Hazard. Mater. 190, 205e213.
Mukaka, M.M., 2012. Statistics corner: a guide to appropriate use of correlation
coefficient in medical research. Malawi Med. J. 24, 69e71.
Nagel, G.N., Fahey, T.J., Ritchie, J.C., Woodbury, P.B., 2007. Variations in sediment
sources and yields in the Finger Lakes and Catskills regions of New York. Hydrol.
Process. 21, 828e838.
National Resources Council (NRC), 2002. Riparian Areas: Functions and Strategies
for Management. National Academy Press, Washington DC.
Nejadhashemi, A.P., Wardynski, B.J., Munoz, J.D., 2011. Evaluating impact of land use
changes on hydrological responses in the agricultural regions of Michigan and
Wisconsin. Hydrology Earth Syst. Sci. Discuss. 8, 3421e3468.
New Jersey Department of Environmental Protection (NJDEP), 2007. Land Use Land
Cover Classification System: NJDEP Modified Anderson System 2002. Available
at: http://www.state.nj.us/dep/gis/digidownload/metadata/lulc02/
 S. Giri et al. / Journal of Environmental Management 213 (2018) 309e319
anderson2002.html (accessed 20/6/15).
New Jersey Department of Environmental Protection (NJDEP), 2010. NJDEP 2007
Land Use/Land Cover Update. NJDEP, Trenton, New Jersey.
New Jersey Department of Environmental Protection (NJDEP), 2014. Water Quality
Management Planning Program. Available at: http://www.nj.gov/dep/wqmp/
(accessed 20/5/15).
Paringit, E.C., Nadaoka, K., 2003. Sediment yield modeling for small agricultural
catchments: land-cover parameterization based on remote sensing data anal-
ysis. Hydrol. Process. 17, 1845e1866.
Peterson, E.E., Sheldon, F., Darnell, R., Bunn, S.E., Harch, B.D., 2010. A comparison of
spatially explicit landscape representation methods and their relationship to
stream condition. Freshw. Biol. https://doi.org/10.1111/
j.1365e2427.2010.02507.x.
Planchon, O., Darboux, F., 2001. A fast, simple and versatile algorithm to fill the
depressions of digital elevation models. Catena 46, 159e176.
Pratt, B., Chang, H., 2012. Effects of land cover, topography, and built structure on
seasonal water quality at multiple spatial scale. J. Hazard. Mater. 209, 48e58.
Qiu, Z., 2009. Assessing critical source areas in watersheds for conservation buffer
planning and riparian restoration. Environ. Manag. 44, 968e980.
Qiu, Z., Hall, C., Drewes, D., Messinger, G., Prato, T., Hale, K., Van Abs, D., 2014.
Hydrologically sensitive areas, land use controls, and protection of healthy
watersheds. J. Water Resour. Plan. Manag. 04014011, 1e10.
Qiu, Z., Wang, L., 2014. Hydrological and water quality assessment in a suburban
watershed with mixed land uses using the SWAT model. J. Hydrological Eng. 19,
816e827.
Qiu, Z., 2015. Validation of a locally revised topographic index in Central New Jersey,
USA. Water 7, 6616e6633.
Seibert, J., McGlynn, B.L., 2007. A new triangular multiple flow direction algorithm
for computing upslope areas from gridded digital elevation models. Water
Resour. Res. 43, W04501.
Shuster, W.D., Bonta, J., Thurston, H., Warnemuende, E., Smith, D.R., 2005. Impacts
of impervious surface on watershed hydrology: a review. Urban Water J. 2,
263e275.
Tsegaye, T., Sheppard, D., Islam, K.R., Johnson, A., Tadesse, W., Atalay, A., Marzen, L.,
2006. Development of chemical index as a measure of in-stream water quality
in response to landuse and land cover changes. Water Air & Soil Pollut. 174,
319
161e179.
Tu, J., 2011. Spatially varying relationships between land use and water quality
across an urbanization gradient explored by geographically weighted regres-
sion. Appl. Geogr. 31, 376e392.
USDA, 2015. Summary Report: 2007 2012 National National Resources Inventory,
Natural Resources Conservation Service, Washington, DC, and Center for Survey
Statistics and Methodology. Iowa State University, Ames, Iowa. Available at:
http://www.nrcs.usda.gov/technical/NRI/2007/2007_NRI_Summary.pdf.
(Accessed 17 September 2014).
Van Meter, K.J., Basu, N.B., Veenstra, J.J., Burras, C.L., 2016. The nitrogen legacy:
emerging evidence of nitrogen accumulation in anthropogenic landscapes.
Environ. Res. Lett. 11 https://doi.org/10.1088/1748-9326/11/3/035014.
Walsh, C.J., Roy, A.H., Feminella, J.W., Cottingham, P.D., Groffman, P.M.,
Morgan II, R.P., 2005. The urban stream syndrome: current knowledge and the
search for a cure. J. North Am. Benthol. Soc. 24, 706e723.
Walter, D.M., Roy, A.H., Leigh, D.S., 2009. Environmental indicators of macro-
invertibrate and fish assemblage integrity in urbanizing watersheds. Ecol. Indic.
9, 1222e1233.
Walter, M.T., Steenhuis, T.S., Mehta, V.K., Thongs, D., Zion, M., Schneiderman, E.,
2002. Refined conceptualization of TOPMODEL for shallow subsurface flows.
Hydrol. Process. 16, 2041e2046.
Walter, M.T., Walter, M.F., Brooks, E.S., Steenhuis, T.S., Boll, J., Weiler, K., 2000. Hy-
drologically sensitive areas: variable source area hydrologically implications for
water quality risk assessment. J. Soil Water Conservation 3, 277e284.
Wan, R., Cai, S., Li, H., Yang, G., Li, Z., 2014. Inferring land use and land cover impact
on stream water quality using Bayesian hierarchical modeling approach in the
Xitiaoxi river watershed, China. J. Environ. Manag. 133, 1e11.
Williams, M., Hopkinson, C., Rastetter, E., Vallino, J., Claessens, L., 2005. Relation-
ships of land use and stream solute concentrations in the Ipswich River basin,
Northeastern Massachusetts. Water Air & Soil Pollut. 161, 55e74.
Wilson, C., Weng, Q., 2010. Assessing surface water quality and its relation with
urban land cover changes in the Lake Calumet area, Greater Chicago. Environ.
Manag. 45, 1096e1111.
Yu, D., Shi, P., Liu, Y., Xun, B., 2013. Detecting landuse water quality relationships
from the view point of ecological restoration in an urban area. Ecol. Eng. 53,
205e216.