Abstract

The effect of different land use types on the physicochemical water quality of a semi-arid coastal basin is examined. From nine (9) sites, a comprehensive sampling campaign was executed during October 2018 to January 2019. One-way ANOVA, Pearson correlation and multiple regression analyses were used to determine the relationship between water quality characteristics and land use types at the sub-basin and buffer-zone scales. The one-way ANOVA test indicated that most of the parameters are significantly different (p < 0.05) among the sampling sites with the exception of pH, total hardness (TH), nitrate-nitrogen (NO3-N), ammonia-nitrogen (NH3-N), dissolved oxygen (DO), chemical oxygen demand (COD) and iron (Fe). Agriculture and built-up/bare land had a positive relationship with turbidity, TSS, conductivity and Fe within 50 m and 150 m buffer zones. Built-up/bare land showed a positive relationship with turbidity, TSS and Cl at the sub-basin scale. Forest cover correlated negatively with water quality although not significantly. Grassland correlated significantly with temperature, Cl and total hardness. Results of the multiple regression analysis indicate that land use types within the riparian buffer zones had greater impact on water quality than at the sub-basin scale. This work provides essential information for land use planners and water managers towards sustainable water resources management.

HIGHLIGHTS

  • The effects of land use variability on water quality of a coastal basin have been examined.

  • High TSS and nitrate concentrations found in the Densu River.

  • Built-up and agricultural fields strongly identified as pollution sources.

  • Land uses at riparian buffer zone have greater influence on water quality.

  • Dynamics of water quality variation is influenced by multi-spatial scales.

INTRODUCTION

On the global scale many rivers are experiencing deteriorating water quality as a result of activities within their catchments. The nature of rivers on a land surface is mostly reflected by the effect of several activities along their pathways. Among other factors, the relationship between landscape characteristics and surface water quality is influenced by urbanisation, population increase and agricultural activities with tendencies of altering the ecological health of water systems (Lintern et al. 2017). Contemporary studies allude to the impacts of land use on water quality (Yu et al. 2016; de Mello et al. 2018; Xu et al. 2019). Generally, built-up/ bare-land and agricultural land uses are mostly associated with organic pollution, heavy metals and nutrients (Lee et al. 2009; Bu et al. 2014; Yadav et al. 2019). On the other hand, forest cover is observed to have an inverse relationship with degraded water quality and is thus deemed to be a good predictor for reducing the deteriorating nature of water quality (Sliva & Williams 2001; Tu 2011; Wang et al. 2014; Gyamfi et al. 2016a). In exploring the link between land uses and water quality, researchers have adopted diverse statistical techniques inclusive of multivariate methods, linear models and redundancy approaches. For example, de Mello et al. (2018) studied the effects of land use and land cover on the quality of low-order streams in southern Brazil using multivariate and linear mixed models. From their work, it was concluded that agricultural and urban areas contribute to water quality degradation while forest cover plays an important role in keeping the water clean. Namugize et al. (2018) used Pearson correlation to determine the effects of land-use and land-cover changes on water quality in the uMngeni river in South Africa. From the results, built-up/urban areas within the sub-catchments positively correlated with NH4, NO3, TSS, and TP, an indication of high pollution levels. Characteristically, the relationship between land use and water quality varies significantly over space and geographic locations mainly due to the different catchment physiognomies and pollution sources (Tu 2011). Moreover, some studies are of the view that land use on riparian-buffer scale influences water quality better than on a catchment scale (Tran et al. 2010; Shi et al. 2017), while others are also of the view that catchment scale better influences the water quality (Ding et al. 2016; Zhang et al. 2019).

The Densu River is one of the most important river systems in Ghana and provides several benefits to humans and the ecosystem. Diverse anthropogenic activities within the basin have caused land use cover changes and have predisposed the basin to a number of pollution challenges (Gyimah et al. 2020). Several studies carried out in the basin have used mathematical models and theories to evaluate the pollutant loadings from point and non-point sources (Karikari & Ansa-Asare 2006; Asante et al. 2008; Fianko et al. 2010; Osei et al. 2010) with none too little focus on the relationship between spatial distributions of land use types and water quality. In the era of Integrated Water Resources Management, it is imperative to consider land use planning in water quality monitoring and compliance frameworks (Mantey et al. 2011; Antwi-Agyakwa 2014). It is imperative therefore to integrate Geographic Information Systems (GIS) and statistical techniques in unravelling the impacts of land use spatial variability on water quality with the view of enhancing water resources management (Mainali & Chang 2018; Kimengich et al. 2019).

To this end, the study seeks to investigate the spatial variation of physicochemical parameters and land use types of the Densu River and subsequently evaluate the complex relationship between land use and water quality parameters at the sub-basin and buffer-zone scales. This study will help explore the impact of land use and landscape pattern on water quality and also inform water resources managers to target appropriate scales for the improvement of water quality.

MATERIALS AND METHODS

Study area

The Densu River Basin (DRB), one of the coastal basins in Ghana, is under focus in this study (Figure 1). The DRB with an estimated drainage area of 2,490 km2 and length of 160 km takes its source from the Atewa forest range and meanders downwards into the Weija Reservoir, and flows finally into the sea. The water resources contribute to the economic livelihood of the people living within the basin. Agriculture is one of the main economic activities within the basin especially in the rural areas. Most people practice both commercial and subsistence farming. The majority of the farmers practice slash and burn as traditional bush fallow, which amounts to deforestation in the basin (WRC 2007). Koforidua and Nsawam are the two most populated urban settings within the DRB with estimated populations of 146,473 and 44,738 respectively, projected from 2010 data (Ghana Statistical Service 2010). The average temperature is about 27 °C with mean annual precipitation of about 1,600 mm–1,700 mm at the upstream and 800 mm–900 mm downstream (WRC 2007). The mean annual runoff of the total basin is about 280 million m3 (WRC 2007). The geological formation of the basin is comprised mainly of upper and middle Birimian rocks and the Togo series. Three major soil groups, namely Togo soils, Granite soils and Birimian soils, are found in the basin (Alfa et al. 2011).

Figure 1

Map of Densu River Basin showing drainage network and sampling sites.

Figure 1

Map of Densu River Basin showing drainage network and sampling sites.

Sampling framework and water quality analysis

Water samples were collected at nine sampling points along the Densu River. These points were selected based on their different land use types within the basin. The river was sampled once every month from October 2018 to January 2019. Duplicate samples were taken at each station, between the hours of 6:00–10:00 am, at a depth of 15–20 cm. Sampling bottles were soaked in nitric acid and rinsed with distilled water before sampling. The samples were analysed for conductivity (Cond), turbidity (Turb), total suspended solids (TSS), pH, total dissolved solids (TDS), temperature (Temp), total hardness (TH), sodium (Na), chloride (Cl), nitrate-nitrogen (NO3-N), phosphate (PO4-P), ammonia-nitrogen (NH3-N), dissolved oxygen (DO), chemical oxygen demand (COD), biological oxygen demand (BOD5) and iron (Fe). Temperature and conductivity were measured in-situ with a WAGTECH (Micro 600) conductivity meter. Measurement of pH was done by using a pH 3000 series multi-parameter (Eutech Instruments, Singapore). Turbidity and TSS were measured with a HACH 5200 turbidimeter. Nitrate-nitrogen was measured using the hydrazine reduction method and PO4-P was determined using the stannous chloride method. Alkalinity was determined using the strong acid titration method. Ammonia-nitrogen was determined using the direct nesslerisation method. Chloride were measured using the argentometric method. The EDTA titration method was used to determine total hardness. COD was determined with the closed tube reflux method. BOD5 was determined using the dilution method by using five day incubation at 20 °C followed by DO determination. DO was determined by using the azide modification of the Winkler method. All measurements followed the standard methods for the examination of water and wastewater (APHA 2012). Quality control measures were done by calibrating all the instruments before use in the field and the laboratory. Blanks and quality control sample analyses were performed on the samples for quality assurance purposes. The laboratory analyses were done in duplicates and the mean values recorded. Ionic error balance was checked from the analytical results and >10% ionic balance error was rejected.

Land cover analysis

Landsat 8 satellite image for 2018 was obtained from the United States Geological Survey (USGS Earth Explorer). Maximum likelihood classification was used for supervised classification. Samples were trained based on their spectral reflectance and field references including Google maps and Bing maps. Five land use types (i.e. agriculture, forest, grassland, built-up/bare land and water bodies) were categorised based on the field references from Google maps and the reconnaissance survey. Anderson's level 1 classification scheme was adopted in this study for the land use classification (Table 1). The land use types were forest, agriculture, grassland, and built up/bare land; water bodies including rivers, lakes and wetlands were also captured in the classification. The overall accuracy of the classified image was 80% with a kappa coefficient of 0.86. Buffer widths of 50 and 150 m were created from the stream network at each water quality monitoring site. The buffer zone widths were selected based on the recommended Design Standards for Riparian Buffer Zones in Ghana. This procedure was performed to compare the influence of land use on water quality both at the sub-basin scale and the various buffer scales. Aster DEM data with 30 m × 30 m resolution was used for the delineation of the Densu Basin. For catchment delineation, the Wang and Liu algorithm (Wang & Liu 2006) was used in identifying and filling surface depressions in the DEM. Subsequently, the processed DEM was used in the Soil and Water Assessment Tool (SWAT) employing the automatic catchment delineator, to extract the sub-catchments of the Densu Basin. Random samples of the various types of land uses were collected through ground-truthing and Google visual observations to aid in the land cover classification and validation process.

Table 1

Land-use classification scheme after Anderson et al. (1976) 

Land cover classDescription
Grassland Herbaceous, shrub and bush and mixed rangeland 
Water Lakes, reservoirs, streams 
Agriculture Crop fields and pastures 
Forest Deciduous, commercial and mixed forest 
Built-up Residential, commercial services, industrial, transportation, communication, mixed urban or built-up lands 
Land cover classDescription
Grassland Herbaceous, shrub and bush and mixed rangeland 
Water Lakes, reservoirs, streams 
Agriculture Crop fields and pastures 
Forest Deciduous, commercial and mixed forest 
Built-up Residential, commercial services, industrial, transportation, communication, mixed urban or built-up lands 

Statistical analysis

One-way analysis of variance (ANOVA) with post hoc test was used to determine the differences in water quality among the sampling sites at a significance level of 0.05. The Kolmogorov–Smirnov goodness-of-fit test (K-S test) was used to examine the characteristics of the distribution inherent in the water quality and land use variables. Non-normal variables were log-transformed to get the distribution closer to normal before the ANOVA analysis was carried out.

The relationship between land use and water quality was performed with multiple regression analysis and Pearson correlation. The correlation analysis developed a correlation matrix between the land use types and water quality in order to determine the type of interaction between them. Moreover, the multiple regression analysis explained the magnitude and influence of the land use (predictor variables) on water quality parameters (response variables). The statistical analysis was performed at two buffer scales (i.e. 50 m and 150 m) and also at the sub-basin level (Figure 2). Stronger positive correlation showed values closer to 1 and those closer to −1 showed a stronger negative correlation between variables. Factor analyses were used to determine the pollution factors affecting the water quality among the sampling sites. According to Liu et al. (2015), factor loading value >0.75 is described as ‘strong’ loading, 0.75–0.50 is ‘moderate’ and 0.50–030 is described as ‘weak’. Water quality was initially standardised with z-scale transformation. Kaiser–Meyer–Olkin (KMO) and Bartlett's sphericity tests were performed to determine the adequacy of the water quality data for the analysis. The analysis was performed using SPSS 25.

Figure 2

Land use classification of the Densu Basin showing the riparian buffer zones and the sub-basins.

Figure 2

Land use classification of the Densu Basin showing the riparian buffer zones and the sub-basins.

RESULTS AND DISCUSSION

Spatial variation of water quality and land use characteristics

Physicochemical parameters of water samples were analysed using descriptive analysis. Mean and standard deviation values at different sampling sites are presented in Table 2. The K-S test of normality is presented in Table 3. Variables with significant test (p < 0.05) indicated that the data were not normally distributed whereas those with less significant test (p > 0.05) showed a normal distribution of the variable. Subsequently, non-normal variables (i.e. pH, forest and grassland) were log-transformed. One-way ANOVA test indicated that most of the parameters are significantly different (p < 0.05) among the sampling sites with the exception of pH, TH, NO3-N, NH3-N, DO, COD and Fe. The correlation matrix from Table 4 indicates high positive correlation between conductivity, turbidity, TSS, PO4-P, temperature, Na, Cl and TDS. These parameters indicate the main contaminants within the basin. High mean values of conductivity, TDS, TH, Na, and Cl were observed at Afuaman, which could be due to the sediment runoff from bare lands and agricultural fields or the leaching of ions from the soil beneath into the water. Lower values were observed at Potroase, which is located upstream the basin. A similar observation at Potroase has been reported by other studies (Karikari & Ansa-Asare 2006; Darko et al. 2013). A high concentration of NO3-N (0.14 ± 0.17 mg/L), turbidity (51.3 ± 30.6 mg/L), TSS (45.3 ± 25.5 mg/L), and minimum DO concentration of 5.9 ± 2.9 mg/L were recorded at Nsawam. The sampling site is located near market communities and highly dense residential and agricultural fields, therefore sewage sludge, chemicals from fertilisers and sediment runoff may explain the high TSS and nitrate leaching into the Densu River. Fianko et al. (2009) reported similar findings at the same area. Also, phosphate showed a strong significant correlation with conductivity, turbidity, TSS, TDS, TH and NO3-N. The observed correlation could be a good indication of nutrient and sediment leaching from agricultural fields and runoff from urban areas. Chemical oygen demand (COD) significantly correlated with TH and PO4-P, which indicates a high contribution of oxidizable organic and inorganic pollutants (Otukune & Biukwu 2005). Pakro recorded high COD concentration (32.3 ± 2.1 mg/L). Additionally, BOD significantly correlated with NO3-N, which indicates the contribution of organic matter and nutrient enrichment from domestic and agricultural sources. Iron (Fe) showed high positive correlation with turbidity, TSS and PO4-P.

Table 2

Descriptive analysis of physicochemical parameters at different sampling sites

WEIJAASHALAJAAFUAMANNSAWAMPAKROMANGOASEAKWADUMKOFORIDUAPOTROASE
pH 8.1 ± 0.3 7.8 ± 0.3 7.6 ± 0.2 8.1 ± 0.5 7.8 ± 0.4 7.8 ± 0.2 7.8 ± 0.3 7.8 ± 0.2 7.7 ± 0.3 
Cond (μS/cm) 305 ± 40.6 410 ± 169 493 ± 243 414 ± 46.2 399 ± 164 372 ± 163 234 ± 29 247 ± 38.5 189 ± 27.5 
Turb (NTU) 15.5 ± 7.5 44.8 ± 24.7 40.8 ± 5.6 51.3 ± 30.6 34.3 ± 11.9 23.5 ± 16.9 19.3 ± 8.6 17.0 ± 9.8 8.3 ± 4.3 
TSS (mg/L) 16.3 ± 8.3 44.5 ± 18.0 37.5 ± 6.6 45.3 ± 25.5 33.3 ± 12.7 27.0 ± 16.2 17.8 ± 6.2 15.3 ± 5.7 9.3 ± 5.4 
Temp (°C) 29.7 ± 0.4 28.3 ± 0.7 29.6 ± 0.7 28.1 ± 1.2 28.6 ± 0.9 27.3 ± 1.6 26.7 ± 1.7 27.6 ± 1.3 25.1 ± 0.5 
TDS (mg/L) 183 ± 24.3 246.0 ± 102 296 ± 146 248 ± 27.7 239 ± 98.4 223. ± 98.3 141 ± 17.3 148 ± 23.1 113 ± 16.5 
TH (mg/L) 92.3 ± 14 123.3 ± 46.4 140 ± 54.3 115 ± 2.2 128 ± 46.3 121 ± 42.5 93.5 ± 8.9 97.4 ± 11.1 69.7 ± 6.6 
Na (mg/L) 30.1 ± 4.1 37.9 ± 15.0 49.5 ± 29.8 36.3 ± 6.8 38.0 ± 16.1 34.3 ± 19.7 16.1 ± 3.9 15.9 ± 0.8 11.7 ± 1.9 
Cl (mg/L) 29.5 ± 7.4 36.1 ± 20.9 52.8 ± 36.4 28.6 ± 8.9 29.8 ± 14.0 26.1 ± 16.1 15.5 ± 1.9 14.8 ± 3.8 12.8 ± 2.7 
NO3-N (mg/L) 0.04 ± 0.03 0.03 ± 0.02 0.03 ± 0.01 0.14 ± 0.17 0.06 ± 0.04 0.07 ± 0.02 0.04 ± 0.03 0.02 ± 0.02 0.04 ± 0.06 
PO4-P (mg/L) 0.1 ± 0.03 0.3 ± 0.04 0.2 ± 0.04 0.3 ± 0.2 0.3 ± 0.05 0.3 ± 0.1 0.1 ± 0.04 0.2 ± 0.03 0.1 ± 0.02 
NH3-N (mg/L) 0.4 ± 0.1 0.4 ± 0.1 0.3 ± 0.2 1.0 ± 0.3 0.5 ± 0.6 0.4 ± 0.3 0.8 ± 0.8 0.3 ± 0.2 0.3 ± 0.2 
DO (mg/L) 7.1 ± 0.2 6.2 ± 1.2 6.3 ± 1.2 5.9 ± 2.9 7.9 ± 0.3 7.7 ± 0.6 6.6 ± 1.1 7.7 ± 1.3 6.0 ± 1.8 
BOD (mg/L) 2.8 ± 1.1 1.0 ± 0.4 2.3 ± 0.5 3.7 ± 2.6 3.2 ± 0.9 3.2 ± 1.9 1.3 ± 1.1 2.9 ± 1.0 1.0 ± 0.7 
COD (mg/L) 22.7 ± 2.6 29.0 ± 7.8 25.9 ± 7.9 28.1 ± 10.2 32.3 ± 2.1 25.6 ± 3.7 19.9 ± 10.9 28.5 ± 13.2 20.3 ± 6.1 
Fe (mg/L) 0.1 ± 0.03 1.8 ± 1.4 1.4 ± 0.9 2.0 ± 1.0 1.6 ± 0.8 1.4 ± 1.1 1.3 ± 0.8 1.0 ± 0.9 0.7 ± 0.3 
WEIJAASHALAJAAFUAMANNSAWAMPAKROMANGOASEAKWADUMKOFORIDUAPOTROASE
pH 8.1 ± 0.3 7.8 ± 0.3 7.6 ± 0.2 8.1 ± 0.5 7.8 ± 0.4 7.8 ± 0.2 7.8 ± 0.3 7.8 ± 0.2 7.7 ± 0.3 
Cond (μS/cm) 305 ± 40.6 410 ± 169 493 ± 243 414 ± 46.2 399 ± 164 372 ± 163 234 ± 29 247 ± 38.5 189 ± 27.5 
Turb (NTU) 15.5 ± 7.5 44.8 ± 24.7 40.8 ± 5.6 51.3 ± 30.6 34.3 ± 11.9 23.5 ± 16.9 19.3 ± 8.6 17.0 ± 9.8 8.3 ± 4.3 
TSS (mg/L) 16.3 ± 8.3 44.5 ± 18.0 37.5 ± 6.6 45.3 ± 25.5 33.3 ± 12.7 27.0 ± 16.2 17.8 ± 6.2 15.3 ± 5.7 9.3 ± 5.4 
Temp (°C) 29.7 ± 0.4 28.3 ± 0.7 29.6 ± 0.7 28.1 ± 1.2 28.6 ± 0.9 27.3 ± 1.6 26.7 ± 1.7 27.6 ± 1.3 25.1 ± 0.5 
TDS (mg/L) 183 ± 24.3 246.0 ± 102 296 ± 146 248 ± 27.7 239 ± 98.4 223. ± 98.3 141 ± 17.3 148 ± 23.1 113 ± 16.5 
TH (mg/L) 92.3 ± 14 123.3 ± 46.4 140 ± 54.3 115 ± 2.2 128 ± 46.3 121 ± 42.5 93.5 ± 8.9 97.4 ± 11.1 69.7 ± 6.6 
Na (mg/L) 30.1 ± 4.1 37.9 ± 15.0 49.5 ± 29.8 36.3 ± 6.8 38.0 ± 16.1 34.3 ± 19.7 16.1 ± 3.9 15.9 ± 0.8 11.7 ± 1.9 
Cl (mg/L) 29.5 ± 7.4 36.1 ± 20.9 52.8 ± 36.4 28.6 ± 8.9 29.8 ± 14.0 26.1 ± 16.1 15.5 ± 1.9 14.8 ± 3.8 12.8 ± 2.7 
NO3-N (mg/L) 0.04 ± 0.03 0.03 ± 0.02 0.03 ± 0.01 0.14 ± 0.17 0.06 ± 0.04 0.07 ± 0.02 0.04 ± 0.03 0.02 ± 0.02 0.04 ± 0.06 
PO4-P (mg/L) 0.1 ± 0.03 0.3 ± 0.04 0.2 ± 0.04 0.3 ± 0.2 0.3 ± 0.05 0.3 ± 0.1 0.1 ± 0.04 0.2 ± 0.03 0.1 ± 0.02 
NH3-N (mg/L) 0.4 ± 0.1 0.4 ± 0.1 0.3 ± 0.2 1.0 ± 0.3 0.5 ± 0.6 0.4 ± 0.3 0.8 ± 0.8 0.3 ± 0.2 0.3 ± 0.2 
DO (mg/L) 7.1 ± 0.2 6.2 ± 1.2 6.3 ± 1.2 5.9 ± 2.9 7.9 ± 0.3 7.7 ± 0.6 6.6 ± 1.1 7.7 ± 1.3 6.0 ± 1.8 
BOD (mg/L) 2.8 ± 1.1 1.0 ± 0.4 2.3 ± 0.5 3.7 ± 2.6 3.2 ± 0.9 3.2 ± 1.9 1.3 ± 1.1 2.9 ± 1.0 1.0 ± 0.7 
COD (mg/L) 22.7 ± 2.6 29.0 ± 7.8 25.9 ± 7.9 28.1 ± 10.2 32.3 ± 2.1 25.6 ± 3.7 19.9 ± 10.9 28.5 ± 13.2 20.3 ± 6.1 
Fe (mg/L) 0.1 ± 0.03 1.8 ± 1.4 1.4 ± 0.9 2.0 ± 1.0 1.6 ± 0.8 1.4 ± 1.1 1.3 ± 0.8 1.0 ± 0.9 0.7 ± 0.3 
Table 3

The Kolmogorov–Smirnov test of normality for water quality parameters and land use characteristics

Water qualityK-S testSig.
pH 0.290 0.028 
Cond 0.178 0.200* 
Turb 0.181 0.200* 
TSS 0.207 0.200* 
Temp 0.113 0.200* 
TDS 0.178 0.200* 
TH 0.166 0.200* 
Na 0.196 0.200* 
Cl 0.199 0.200* 
NO3 0.253 0.101* 
PO4 0.225 0.200* 
NH3 0.257 0.088* 
DO 0.199 0.200* 
BOD 0.223 0.200* 
COD 0.153 0.200* 
Fe 0.185 0.200* 
50 m bufferK-S testSig.
Forest 0.314 0.011 
Grassland 0.519 0.000 
Built-up/bare land 0.253 0.100* 
Agriculture 0.155 0.200* 
150 m bufferK-S testSig.
Forest 0.243 0.002 
Grassland 0.253 0.006 
Built-up/bare land 0.307 0.114* 
Agriculture 0.181 0.200* 
Sub-basinK-S testSig.
Forest 0.336 0.004 
Grassland 0.321 0.008 
Built-up/bare land 0.253 0.100* 
Agriculture 0.155 0.200* 
Water qualityK-S testSig.
pH 0.290 0.028 
Cond 0.178 0.200* 
Turb 0.181 0.200* 
TSS 0.207 0.200* 
Temp 0.113 0.200* 
TDS 0.178 0.200* 
TH 0.166 0.200* 
Na 0.196 0.200* 
Cl 0.199 0.200* 
NO3 0.253 0.101* 
PO4 0.225 0.200* 
NH3 0.257 0.088* 
DO 0.199 0.200* 
BOD 0.223 0.200* 
COD 0.153 0.200* 
Fe 0.185 0.200* 
50 m bufferK-S testSig.
Forest 0.314 0.011 
Grassland 0.519 0.000 
Built-up/bare land 0.253 0.100* 
Agriculture 0.155 0.200* 
150 m bufferK-S testSig.
Forest 0.243 0.002 
Grassland 0.253 0.006 
Built-up/bare land 0.307 0.114* 
Agriculture 0.181 0.200* 
Sub-basinK-S testSig.
Forest 0.336 0.004 
Grassland 0.321 0.008 
Built-up/bare land 0.253 0.100* 
Agriculture 0.155 0.200* 

*indicates normal distribution of data (p > 0.05), bold values indicate non-normal distribution.

Table 4

Correlation analysis of water quality parameters at the Densu River Basin

pHCONDTURBTSSTEMPTDSTHNaClNO3-NPO4-PNH3-NDOBODCODFe
pH                
Cond (μS/cm) 0.015               
Turb (NTU) 0.183 .861**              
TSS (mg/L) 0.146 .889** .987**             
Temp (°C) 0.361 .715* 0.502 0.500            
TDS (mg/L) 0.015 1.000** .861** .889** .715*           
TH (mg/L) − 0.121 .945** .790* .826** 0.657 .945**          
Na (mg/L) 0.031 .986** .786* .824** .768* .986** .915**         
Cl (mg/L) − 0.117 .914** .686* .709* .773* .914** .822** .942**        
NO3-N (mg/L) 0.514 0.299 0.394 0.407 − 0.004 0.299 0.164 0.288 0.031       
PO4-P (mg/L) 0.083 .783* .827** .853** 0.413 .783* .873** .697* 0.497 0.360      
NH3-N (mg/L) 0.625 0.062 0.366 0.307 0.032 0.062 0.039 0.017 − 0.157 0.619 0.316     
DO (mg/L) 0.041 − 0.051 − 0.306 − 0.254 0.215 − 0.051 0.173 − 0.017 − 0.165 − 0.113 0.193 − 0.144    
BOD (mg/L) 0.524 0.393 0.276 0.255 0.473 0.393 0.387 0.381 0.166 0.535 0.470 0.283 0.501   
COD (mg/L) 0.135 0.617 0.639 0.654 0.477 0.617 .695* 0.557 0.379 0.166 .771* 0.030 0.387 0.514  
Fe (mg/L) − 0.074 0.600 .823** .820** 0.044 0.600 .669* 0.475 0.309 0.383 .875** 0.416 − 0.151 0.177 0.598 
pHCONDTURBTSSTEMPTDSTHNaClNO3-NPO4-PNH3-NDOBODCODFe
pH                
Cond (μS/cm) 0.015               
Turb (NTU) 0.183 .861**              
TSS (mg/L) 0.146 .889** .987**             
Temp (°C) 0.361 .715* 0.502 0.500            
TDS (mg/L) 0.015 1.000** .861** .889** .715*           
TH (mg/L) − 0.121 .945** .790* .826** 0.657 .945**          
Na (mg/L) 0.031 .986** .786* .824** .768* .986** .915**         
Cl (mg/L) − 0.117 .914** .686* .709* .773* .914** .822** .942**        
NO3-N (mg/L) 0.514 0.299 0.394 0.407 − 0.004 0.299 0.164 0.288 0.031       
PO4-P (mg/L) 0.083 .783* .827** .853** 0.413 .783* .873** .697* 0.497 0.360      
NH3-N (mg/L) 0.625 0.062 0.366 0.307 0.032 0.062 0.039 0.017 − 0.157 0.619 0.316     
DO (mg/L) 0.041 − 0.051 − 0.306 − 0.254 0.215 − 0.051 0.173 − 0.017 − 0.165 − 0.113 0.193 − 0.144    
BOD (mg/L) 0.524 0.393 0.276 0.255 0.473 0.393 0.387 0.381 0.166 0.535 0.470 0.283 0.501   
COD (mg/L) 0.135 0.617 0.639 0.654 0.477 0.617 .695* 0.557 0.379 0.166 .771* 0.030 0.387 0.514  
Fe (mg/L) − 0.074 0.600 .823** .820** 0.044 0.600 .669* 0.475 0.309 0.383 .875** 0.416 − 0.151 0.177 0.598 

**Correlation is significant at the 0.01 level (2-tailed).

*Correlation is significant at the 0.05 level (2-tailed).

The land use characteristics in Figure 3 indicate the proportion of land use types (i.e. built-up/bare land, agriculture, forest land, grassland and water body) within the 50 m buffer scale, 150 m buffer scale and the sub-basin of the nine sampling sites. Built-up/bare land and agricultural lands were the dominant land use at the downstream parts of the Densu River Basin. Within the 50 m buffer, built-up/bare land occupied the highest percentage area of 51% at Ashalaja. Agricultural lands occupied the highest percentage area of 54% and 47% at Nsawam and Afuaman, respectively. This could contribute to the high concentration of NO3-N, turbidity, and TSS at Nsawam and conductivity, TDS, TH, Na, and Cl at Afuaman, respectively. Potroase and Pakro occupied the highest proportions of 74% and 55% of forest cover within the 50 m buffer. This could contribute to the low water quality values at Potroase. The dominant land use within the 150 m buffer was agricultural land use, which was evidenced at Afuaman (58%), Ashalaja (51%), Nsawam (51%), and Koforidua (45%). This could influence the relatively high turbidity (44.8 ± 24.7 NTU; 51.3 ± 30.6 NTU) and PO4 (0.3 ± 0.04 mg/L; 0.3 ± 0.2 mg/L) concentration at Ashalaja and Nsawam respectively, due to runoff from unconventional farming activities and soil erosion (Amoako et al. 2010). The percentage area of grassland was high at Akwadum (55%) and Mangoase (47%). The dominant land use at the sub-basin was built-up/bare land. Large proportions of built-up/bare land uses were found at Nsawam (85%), Ashalaja (85%), Afuaman (68%) and Weija (68%). This could be due to the increase in population and rate of urbanisation in these areas. Grassland was dominant at Mangoase (87%) and Akwadum (84%) and forest cover was dominant at Potroase (93%) and Pakro (50%).

Figure 3

Land use proportion among the sampling sites within (a) 50 m buffer, (b) 150 m buffer and (c) sub-basin scales of the DRB.

Figure 3

Land use proportion among the sampling sites within (a) 50 m buffer, (b) 150 m buffer and (c) sub-basin scales of the DRB.

Principal Component Analysis (PCA) in Table 5 shows extractions of four factors. Factor 1 accounted for 40% of the total variance in water quality parameters, with strong loadings of conductivity, temperature, total dissolved oxygen, total hardness, sodium and chloride. The concentration of these ions could be due to seawater intrusion from the sea. A similar result was reported by Osei et al. (2010). Factor 2 accounted for 27% of the total variance with strong factor loadings of phosphate (PO4-P) and iron (Fe), which indicates human influence on the water quality, especially agricultural activities. Factor 3 accounted for 14.41% of the total variance in the water quality parameters and depicted high loadings of DO and BOD from domestic waste discharge. Factor 4 accounted for 12.90% of the total variance and showed strong loading of pH and ammonia (NH3-N) from agricultural fields.

Table 5

Factor scores of water quality parameters from principal component analysis

ParametersFactor 1Factor 2Factor 3Factor 4
pH 0.022 − 0.115 0.134 0.934 
Cond (μS/cm) 0.894 0.431 0.057 0.011 
Turb (NTU) 0.669 0.667 − 0.195 0.212 
TSS (mg/L) 0.686 0.681 − 0.149 0.154 
Temp (°C) 0.897 − 0.174 0.256 0.212 
TDS (mg/L) 0.894 0.431 0.057 0.011 
TH (mg/L) 0.794 0.514 0.206 − 0.142 
Na (mg/L) 0.93 0.307 0.085 0.009 
Cl (mg/L) 0.973 0.089 − 0.101 − 0.14 
NO3-N (mg/L) 0.05 0.662 0.496 0.383 
PO4-P (mg/L) 0.371 0.822 0.343 0.022 
NH3-N (mg/L) − 0.113 0.444 − 0.28 0.774 
DO (mg/L) − 0.09 − 0.072 0.938 − 0.137 
BOD (mg/L) 0.266 0.191 0.704 0.523 
COD (mg/L) 0.429 0.585 0.509 − 0.037 
Fe (mg/L) 0.232 0.947 − 0.145 0.017 
Eigenvalue 6.2 4.3 2.3 2.1 
% Variance 40 27 14 13 
Cumulative % 40 66 80 93 
ParametersFactor 1Factor 2Factor 3Factor 4
pH 0.022 − 0.115 0.134 0.934 
Cond (μS/cm) 0.894 0.431 0.057 0.011 
Turb (NTU) 0.669 0.667 − 0.195 0.212 
TSS (mg/L) 0.686 0.681 − 0.149 0.154 
Temp (°C) 0.897 − 0.174 0.256 0.212 
TDS (mg/L) 0.894 0.431 0.057 0.011 
TH (mg/L) 0.794 0.514 0.206 − 0.142 
Na (mg/L) 0.93 0.307 0.085 0.009 
Cl (mg/L) 0.973 0.089 − 0.101 − 0.14 
NO3-N (mg/L) 0.05 0.662 0.496 0.383 
PO4-P (mg/L) 0.371 0.822 0.343 0.022 
NH3-N (mg/L) − 0.113 0.444 − 0.28 0.774 
DO (mg/L) − 0.09 − 0.072 0.938 − 0.137 
BOD (mg/L) 0.266 0.191 0.704 0.523 
COD (mg/L) 0.429 0.585 0.509 − 0.037 
Fe (mg/L) 0.232 0.947 − 0.145 0.017 
Eigenvalue 6.2 4.3 2.3 2.1 
% Variance 40 27 14 13 
Cumulative % 40 66 80 93 

Bold values indicate factor loading >0.75, which is described as ‘strong’ loading.

Relationship and influence of land use on water quality

The relationships between land-use and water quality at the buffer and sub-basin scales are presented in Table 6, using the Pearson correlation. The results explain that built-up/bare lands within the 50 m buffer showed a significant positive correlation with conductivity (r = 0.68, p ≤ 0.05), turbidity (r = 0.76, p ≤ 0.05), TSS (r = 0.72, p ≤ 0.05), TDS (r = 0.68, p ≤ 0.05) and Cl (r = 0.71, p ≤ 0.05). Agriculture showed a significant positive correlation with turbidity (r = 0.83, p ≤ 0.01), TSS (r = 0.77, p ≤ 0.05) and Fe (r = 0.82, p ≤ 0.01). Built-up/ bare land within the 150 m buffer scale had high positive correlation with conductivity, turbidity, TSS, temperature, TDS and total hardness but this was not statistically significant. Agriculture showed significant positive correlation with turbidity (r = 0.81, p ≤ 0.01), TSS (r = 0.77, p ≤ 0.05) and Fe (r = 0.80, p ≤ 0.01). At the sub-basin level, built-up/bare land had a significant positive correlation with turbidity (r = 0.71, p ≤ 0.05), TSS (r = 0.66, p ≤ 0.05), and Cl (r = 0.69, p ≤ 0.05). Forest cover correlated negatively with water quality parameters at the different scales but this was not statistically significant. However, grassland had a significant negative correlation with temperature (r = −0.73, p ≤ 0.05) and total hardness (r = −0.67, p ≤ 0.05).

Table 6

Results of Pearson correlation between land use types and water quality within the buffer zones and sub-basin scale of the sampling sites in the Densu Basin

ForestGrasslandAgricultureBuilt-up/bare land
50 m buffer     
pH − 0.409 − 0.383 − 0.022 − 0.136 
Cond − 0.091 − 0.559 0.581 0.680* 
Turb − 0.185 − 0.503 0.829** 0.758* 
TSS − 0.082 − 0.505 0.772* 0.720* 
Temp − 0.463 0.729* 0.255 0.444 
TDS − 0.091 − 0.559 0.581 0.680* 
TH − 0.042 0.665* 0.601 0.597 
Na − 0.066 − 0.540 0.469 0.605 
Cl − 0.193 − 0.436 0.461 0.710* 
NO3-N 0.241 − 0.035 0.080 − 0.160 
PO4-P − 0.019 − 0.573 0.635 0.488 
NH3-N − 0.293 − 0.357 0.446 − 0.126 
DO 0.195 − 0.414 − 0.406 − 0.480 
BOD − 0.168 − 0.499 − 0.011 − 0.029 
COD 0.131 − 0.496 0.365 0.392 
Fe 0.077 − 0.367 0.820** 0.514 
150 m buffer     
pH − 0.368 − 0.304 − 0.079 0.165 
Cond − 0.316 − 0.589 0.570 0.524 
Turb − 0.337 − 0.605 0.813** 0.596 
TSS − 0.286 − 0.592 0.768** 0.500 
Temp − 0.644 − 0.576 0.264 0.524 
TDS − 0.316 − 0.589 0.570 0.524 
TH − 0.283 − 0.351 0.604 0.528 
Na − 0.297 − 0.637 0.463 0.435 
Cl − 0.403 0.681* 0.474 0.456 
NO3-N 0.133 − 0.266 − 0.029 0.191 
PO4-P − 0.105 − 0.211 0.612 0.438 
NH3-N − 0.113 − 0.187 0.362 0.286 
DO 0.083 0.517 − 0.388 − 0.078 
BOD − 0.171 0.007 − 0.117 0.468 
COD − 0.106 − 0.186 0.396 0.411 
Fe − 0.016 − 0.134 0.798** 0.443 
Sub-basin     
pH − 0.395 − 0.383 − 0.091 0.307 
Cond − 0.221 − 0.559 0.244 0.616 
Turb − 0.290 − 0.503 0.489 0.713* 
TSS − 0.227 − 0.505 0.467 0.663* 
Temp − 0.511 0.729* 0.101 0.587 
TDS − 0.221 − 0.559 0.244 0.616 
TH − 0.196 0.665* 0.388 0.426 
Na − 0.190 − 0.540 0.174 0.602 
Cl − 0.281 − 0.436 0.150 0.693* 
NO3-N 0.039 − 0.107 − 0.100 0.240 
PO4-P- − 0.077 − 0.573 0.413 0.294 
NH3-N − 0.287 − 0.300 0.437 0.344 
DO 0.137 − 0.414 − 0.093 − 0.621 
BOD − 0.165 − 0.499 − 0.289 0.020 
COD 0.043 − 0.496 0.297 0.073 
Fe − 0.046 − 0.367 0.622 0.299 
ForestGrasslandAgricultureBuilt-up/bare land
50 m buffer     
pH − 0.409 − 0.383 − 0.022 − 0.136 
Cond − 0.091 − 0.559 0.581 0.680* 
Turb − 0.185 − 0.503 0.829** 0.758* 
TSS − 0.082 − 0.505 0.772* 0.720* 
Temp − 0.463 0.729* 0.255 0.444 
TDS − 0.091 − 0.559 0.581 0.680* 
TH − 0.042 0.665* 0.601 0.597 
Na − 0.066 − 0.540 0.469 0.605 
Cl − 0.193 − 0.436 0.461 0.710* 
NO3-N 0.241 − 0.035 0.080 − 0.160 
PO4-P − 0.019 − 0.573 0.635 0.488 
NH3-N − 0.293 − 0.357 0.446 − 0.126 
DO 0.195 − 0.414 − 0.406 − 0.480 
BOD − 0.168 − 0.499 − 0.011 − 0.029 
COD 0.131 − 0.496 0.365 0.392 
Fe 0.077 − 0.367 0.820** 0.514 
150 m buffer     
pH − 0.368 − 0.304 − 0.079 0.165 
Cond − 0.316 − 0.589 0.570 0.524 
Turb − 0.337 − 0.605 0.813** 0.596 
TSS − 0.286 − 0.592 0.768** 0.500 
Temp − 0.644 − 0.576 0.264 0.524 
TDS − 0.316 − 0.589 0.570 0.524 
TH − 0.283 − 0.351 0.604 0.528 
Na − 0.297 − 0.637 0.463 0.435 
Cl − 0.403 0.681* 0.474 0.456 
NO3-N 0.133 − 0.266 − 0.029 0.191 
PO4-P − 0.105 − 0.211 0.612 0.438 
NH3-N − 0.113 − 0.187 0.362 0.286 
DO 0.083 0.517 − 0.388 − 0.078 
BOD − 0.171 0.007 − 0.117 0.468 
COD − 0.106 − 0.186 0.396 0.411 
Fe − 0.016 − 0.134 0.798** 0.443 
Sub-basin     
pH − 0.395 − 0.383 − 0.091 0.307 
Cond − 0.221 − 0.559 0.244 0.616 
Turb − 0.290 − 0.503 0.489 0.713* 
TSS − 0.227 − 0.505 0.467 0.663* 
Temp − 0.511 0.729* 0.101 0.587 
TDS − 0.221 − 0.559 0.244 0.616 
TH − 0.196 0.665* 0.388 0.426 
Na − 0.190 − 0.540 0.174 0.602 
Cl − 0.281 − 0.436 0.150 0.693* 
NO3-N 0.039 − 0.107 − 0.100 0.240 
PO4-P- − 0.077 − 0.573 0.413 0.294 
NH3-N − 0.287 − 0.300 0.437 0.344 
DO 0.137 − 0.414 − 0.093 − 0.621 
BOD − 0.165 − 0.499 − 0.289 0.020 
COD 0.043 − 0.496 0.297 0.073 
Fe − 0.046 − 0.367 0.622 0.299 

**indicates significance at p ≤ 0.01 and *indicates significance at p ≤ 0.05.

Table 7

Results of multiple regression of the effect of land use on water quality at the buffer scales and sub-basin

Water quality (responses)Predictors
50 m bufferForestGrasslandAgricultureBuilt-up/bare landR2
Temperature  −   0.76 
Turbidity   0.84 
TSS   0.72 
Fe    0.85 
150 m bufferForestGrasslandAgricultureBuilt-up/bare landR2
Turbidity    0.74 
TSS    0.59 
Fe    0.64 
Cl  −   0.47 
Sub-basinForestGrasslandAgricultureBuilt-up/bare landR2
Temperature  −   0.60 
Turbidity    0.62 
TH  −   0.54 
TSS    0.63 
Cl    0.50 
Water quality (responses)Predictors
50 m bufferForestGrasslandAgricultureBuilt-up/bare landR2
Temperature  −   0.76 
Turbidity   0.84 
TSS   0.72 
Fe    0.85 
150 m bufferForestGrasslandAgricultureBuilt-up/bare landR2
Turbidity    0.74 
TSS    0.59 
Fe    0.64 
Cl  −   0.47 
Sub-basinForestGrasslandAgricultureBuilt-up/bare landR2
Temperature  −   0.60 
Turbidity    0.62 
TH  −   0.54 
TSS    0.63 
Cl    0.50 

(+) indicates positive significant correlation (p < 0.05, p < 0.01) and (−) indicates negative significant correlation, blank space represents no correlation.

Several researchers have reported trends of the impact of urban, agricultural and forest land use on water quality (Sliva & Williams 2001; Ou et al. 2016; Gu et al. 2019; Yadav et al. 2019). For example, Yadav et al. (2019) revealed a positive correlation of total phosphorus (TP) with urban land use at both sub-basin and buffer-zone scale and attributed it to point sources such as treatment plants, domestic sewers, wastewater treatment plants and industries. In this study, built-up and agricultural land appeared to be the most important predictors of water quality variability at both the buffer-zone and sub-basin scales. The positive correlation between built-up/bare land and TDS, conductivity, Cl, turbidity and TSS at both buffer zones and sub-basin are most likely to be ascribed to sediment runoff from construction sites, weathering of rocks and erosion from bare areas (Kusimi 2009). Road salts can be a great contributor to chlorides in receiving waters (Granato et al. 2015). The positive correlation between agricultural lands and turbidity, TSS and Fe at the 50 m and 150 m buffer zones could likely be attributed to the sediment runoff from loose soils on agricultural lands into the river (Lintern et al. 2017; Yadav et al. 2019). Forest and grassland depicted negative correlation with water quality parameters. This indicates that as forest land increases, degraded water quality decreases and vice-versa. According to Sliva & Williams (2001), vegetation can improve water quality deterioration by absorbing nutrients and blocking sediment runoff. Tu (2011) revealed similar results, indicating forest and grasslands as indicators for good water quality. One of the main effects of riparian forests on streams is shading, which influences water temperature and aquatic productivity (Sugimoto et al. 1997; Newton & Ice 2016).

Evaluation of the complex relationship between land use and water quality

There have been several reports on the complex relationship between land use types and water quality. Many researchers have debated whether the influence of land use on water quality at the buffer scale is more important than at the sub-basin level. Sliva & Williams (2001) reported that the correlation at the catchment scale was better than at the 100 m buffer scale. Other studies including Hunsaker & Levine (1995) indicated that 200 m or 400 m buffer strips are considered to better describe the influence of land use on water quality. Johnson et al. (1997) also reported that land use within a 100 m buffer has better influence on water quality than the whole catchment. Moreover, Chang (2008) indicated that the land cover and topographic factors at the riparian-buffer scale better explained the variations in BOD, COD, SS, TN and TP than the whole basin. This study used a multiple regression model to determine the effects of land use types on water quality at both buffer and sub-basin scales. The four land use types were regressed against the water quality parameters with significant correlation, with results summarised in Table 7. Turbidity (R2 = 0.84), TSS (R2 = 0.72), temperature (R2 = 0.76) and Fe (R2 = 0.85) were highly influenced by built-up/bare land and agricultural activities within the 50 m buffer zone. At the 150 m buffer, agriculture showed as a positive contributor to turbidity (R2 = 0.74), TSS (R2 = 0.59) and Fe (R2 = 0.64). Grassland correlated with Cl (R2 = 0.47), while forest cover did not show any significant correlation. At the sub-basin scale, turbidity (R2 = 0.62), TSS (R2 = 0.63) and Cl (R2 = 0.50) were influenced by built-up/bare land. Total hardness (R2 = 0.54) and temperature (R2 = 0.60) were influenced by grassland. Comparing the correlation coefficient (R2) of the response variables (water quality variables) at the different spatial scales, the water quality variable that had the highest multiple regression coefficient was turbidity (R2 = 0.84) and was better correlated with the 50 m buffer zone. Hence it indicates the largest effect on water quality. This corresponds with results from several studies that suggested that land use within riparian buffer zones has greater impact on water quality than at the sub-catchment or catchment scales (Sliva & Williams 2001; Ou et al. 2016; Zhang et al. 2019). The influence of built-up/bare land and agricultural land use within the 50 m buffer indicates that anthropogenic activities contribute more pollution at the buffer-zone scale than at the sub-basin. Therefore, more management strategies should be given at different buffer-zone scales than in a whole-catchment scale approach.

Land use and water management strategies

The Densu River is one of the major rivers in Ghana, fundamental to water supply and food safety. Therefore, land use and water management measures should effectively be promoted to improve the catchment ecology and water quality. This study indicates the influence of both land use and landscape patterns on water quality. The wide distribution of forest cover and grassland at the upstream parts of the basin improved water quality while built-up/bare land and agricultural activities at the midstream and downstream parts led to water quality deterioration. Hydrologically, these land uses influence the degree of runoff generation (Gyamfi et al. 2016b; Kwarteng et al. 2020) which has the propensity to carry contaminants to the Densu River. According to WRC (2007), human activities have overwhelmingly impacted on the ecosystem, and there is total loss of riverine forest, and hence surface water resources are vulnerable to pollution. Damaged forest areas should be restored to support the water treatment capacity in the region. From the result, water quality is affected more greatly at the buffer-zone than at the sub-basin scale, and the primary proportion of land use types affecting water quality are built-up/bare land and agricultural land. These suggest that catchment management should incorporate and implement better land use planning to improve water quality within riparian buffer zones.

CONCLUSION AND RECOMMENDATION

We conclude that the water quality of the Densu River is not only affected by land use types but is also dependent on multi-spatial scales. Land use types within the river basin showed high correlation with water quality at both buffer and sub-basin scales. Turbidity, TSS, conductivity, Cl and TDS were better correlated with the 50 m and 150 m buffer scales due to the agricultural and domestic activities within the riparian buffer zones. Turbidity, TSS and Cl were better correlated with the sub-basin, which could also be from point and non-point sources of pollution such as leachate from landfills, septic system discharges, salt water influx and road construction. Forest was negatively correlated with water quality but not significantly. Grassland correlated significantly with temperature, Cl and total hardness. This indicates that grassland and forest lands are indicators for good water quality by reducing erosion, absorbing nutrients and also improving aquatic productivity. Comparing the influence of the three landscape metrics on water quality and turbidity indicates high correlation coefficients (R2 = 0.84) and (R2 = 0.74) at the 50 m buffer and 150 m buffer zones respectively, which explains that turbidity is highly influenced by activities within the buffer zones rather than at the sub-basin. Therefore, the influence of water quality at the sub-basins is slightly better than that within the riparian buffer zones. We also highlighted that human activities from point and non-point sources resulted in water quality deterioration, using principal component analysis. These findings and methods used could be useful to identify the sources of pollution and improve water management and land use planning within the basin. Due to the landscape effect on water quality, we recommend that land use planners and water managers should demonstrate Best Management Practices at different spatial scales, especially at the riparian zones, rather than in a whole catchment management approach. Moreover agricultural activities, construction and residential areas within 50 and 150 m riparian buffers should be banned and relocated. Further investigations will require natural factors such as population, topography, soil and climate to deeply disclose the reasons for water quality variability within the watered. Also, further work will require both spatial and temporal sampling regime and high-resolution maps for the analysis.

ACKNOWLEDGEMENT

This study was funded by the Regional Water and Environmental Sanitation Centre Kumasi (RWESCK) at the Kwame Nkrumah University of Science and Technology (KNUST), Kumasi, with funding from the Ghana Government through the World Bank under the Africa Centres of Excellence project. The views expressed in this paper do not reflect those of the World Bank, Ghana Government and KNUST. CSIR-Water Research Institute is also duly acknowledged for availing their facilities and supporting the field work as well as the laboratory analysis. Authors are grateful to Mr Victor Mantey for his assistance during the field sampling and laboratory analysis phase of this work.

DATA AVAILABILITY STATEMENT

All relevant data are included in the paper or its Supplementary Information.

REFERENCES

REFERENCES
Alfa
B.
,
Hasholt
B.
,
Jørgensen
N. O.
&
Banoeng-Yakubo
B.
2011
Rainfall and water resources of a coastal basin of Ghana
.
Journal of Hydrologic Engineering
16
(
4
),
316
323
.
Amoako
J.
,
Karikari
A. Y.
,
Ansa-Asare
O. D.
&
Adu-Ofori
E.
2010
Water quality characteristics of Densu River basin in south-east Ghana
.
Water Science and Technology
61
(
6
),
1467
1477
.
Anderson
J. R.
,
Hardy
E. E.
,
Roach
J. T.
&
Witmer
R. E.
1976
.
A Land Use and Land Cover Classification System for Use with Remote Sensor Data. Geological Survey Professional Paper 964. US Government Printing Office
,
Washington, DC, USA
.
Antwi-Agyakwa
K. T.
2014
Assessing the Effect of Land Use Land Cover Change on Weija Catchment
.
Master's thesis
,
Kwame Nkrumah University of Science and Technology
,
Kumasi
,
Ghana
.
APHA
2012
Standard Methods for the Examination of Water and Wastewater
, 22nd edn.
American Public Health Association
,
Washington, DC, USA
.
Asante
K. A.
,
Quarcoopome
T.
&
Amevenku
F. Y. K.
2008
Water quality of the Weija reservoir after 28 years of impoundment
.
West African Journal of Applied Ecology
13
(
1
),
125
131
.
Darko
H. F.
,
Ansa-Asare
O.
&
Paintsil
A.
2013
A number description of Ghanaian water quality – a case study of the Southwestern and Coastal Rivers Systems of Ghana
.
Journal of Environmental Protection
4
,
1318
1327
.
de Mello
K.
,
Valente
R. A.
,
Randhir
T. O.
,
dos Santos
A. C. A.
&
Vettorazzi
C. A.
2018
Effects of land use and land cover on water quality of low-order streams in Southeastern Brazil: watershed versus riparian zone
.
Catena
167
,
130
138
.
Ding
J.
,
Jiang
Y.
,
Liu
Q.
,
Hou
Z.
,
Liao
J.
,
Fu
L.
&
Peng
Q.
2016
Influences of the land use pattern on water quality in low-order streams of the Dongjiang River basin, China: a multi-scale analysis
.
Science of the Total Environment
551–552
,
205
216
.
Fianko
J. R.
,
Osae
S.
&
Achel
D.
2009
Impact of anthropogenic activities on the Densu River in Ghana
.
Water and Environment Journal
23
(
3
),
229
234
.
Fianko
J. R.
,
Lowor
S. T.
,
Donkor
A.
&
Yeboah
P. O.
2010
Nutrient chemistry of the Densu River in Ghana
.
The Environmentalist
30
(
2
),
145
152
.
Ghana Statistical Service
2010
Population and Housing Census Report
.
Ghana Statistical Service, Accra, Ghana
.
Granato
G. E.
,
DeSimone
L. A.
,
Barbaro
J. R.
&
Jeznach
L. C.
2015
Methods for Evaluating Potential Sources of Chloride in Surface Waters and Groundwaters of the Conterminous United States, Open File Report 2015-1080
.
US Geological Survey, Reston, VA, USA
.
Gu
Q.
,
Hu
H.
,
Ma
L.
,
Sheng
L.
,
Yang
S.
,
Zhang
X.
,
Zhang
M.
,
Zheng
K.
&
Chen
L.
2019
Characterizing the spatial variations of the relationship between land use and surface water quality using self-organizing map approach
.
Ecological Indicators
102
,
633
643
.
Gyimah
R. A. A.
,
Gyamfi
C.
,
Anornu
G. K.
,
Karikari
A. Y.
&
Tsyawo
F. W.
2020
Multivariate statistical analysis of water quality of the Densu River, Ghana
.
International Journal of River Basin Management
.
doi: 10.1080/15715124.2020.1803337.
Hunsaker
C. T.
&
Levine
D. A.
1995
Hierarchical approaches to the study of water quality in rivers
.
BioScience
45
(
3
),
193
203
.
Johnson
L.
,
Richards
C.
,
Host
G.
&
Arthur
J.
1997
Landscape influences on water chemistry in Midwestern stream ecosystems
.
Freshwater Biology
37
(
1
),
193
208
.
Karikari
A. Y.
&
Ansa-Asare
O. D.
2006
Physico-chemical and microbial water quality assessment of Densu River of Ghana
.
West African Journal of Applied Ecology
10
(
1
).
doi: 10.4314/wajae.v10i1.45701
.
Kusimi
J. M.
2009
Analysis of sedimentation rates in the Densu River channel: the result of erosion and anthropogenic activities in the Densu Basin
.
West African Journal of Applied Ecology
14
(
1
).
doi: 10.4314/wajae.v14i1.44703
.
Kwarteng
E. A.
,
Gyamfi
C.
,
Anyemedu
F. O. K.
,
Adjei
K. A.
&
Anornu
G. K.
2020
Coupling SWAT and bathymetric data in modelling reservoir catchment hydrology
.
Spatial Information Research
.
doi: 10.1007/s41324-020-00337-7.
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
.
Landscape and Urban Planning
92
(
2
),
80
89
.
Lintern
A.
,
Webb
J. A.
,
Ryu
D.
,
Liu
S.
,
Bende-Michl
U.
,
Waters
D.
,
Leahy
P.
,
Wilson
P.
&
Western
A. W.
2017
Key factors influencing differences in stream water quality across space
.
Wiley Interdisciplinary Reviews: Water
5
(
1
),
e1260
.
Liu
X.
,
Xue
Y.
,
Ji
Y. P.
,
Xu
B. D.
&
Ren
Y. P.
2015
An assessment of water quality in the Yellow River estuary and its adjacent waters based on principal component analysis
.
China Environmental Science
35
(
10
),
3187
3192
.
Mantey
S.
,
Tagoe
N. D.
&
Aduah
M. S.
2011
Detecting areas of vegetation change in the Densu River Basin, Ghana
. In:
River Basin Management VI (C. A. Brebbia, ed.), WIT Press, Southampton, UK
, pp.
37
48
.
Namugize
J. N.
,
Jewitt
G.
&
Graham
M.
2018
Effects of land use and land cover changes on water quality in the uMngeni river catchment, South Africa
.
Physics and Chemistry of the Earth, Parts A/B/C
105
,
247
264
.
Osei
J.
,
Nyame
F. K.
,
Armah
T. K.
,
Osae
S. K.
,
Dampare
S. B.
,
Fianko
J. R.
,
Adomako
D.
&
Bentil
N.
2010
Application of multivariate analysis for identification of pollution sources in the Densu delta wetland in the vicinity of a landfill site in Ghana
.
Journal of Water Resource and Protection
2
(
12
),
1020
1029
.
Otukune
T. V.
&
Biukwu
C. O.
2005
Impact of refinery influent on physico-chemical properties of a water body on Niger Delta
.
Journal of Applied Ecology and Environment Research
3
,
61
72
.
Tran
C. P.
,
Bode
R. W.
,
Smith
A. J.
&
Kleppel
G. S.
2010
Land-use proximity as a basis for assessing stream water quality in New York State (USA)
.
Ecological Indicators
10
,
727
733
.
Wang
G.
,
A, Y., Xu
Z.
&
Zhang
S.
2014
The influence of land use patterns on water quality at multiple spatial scales in a river system
.
Hydrological Processes
28
(
20
),
5259
5272
.
Water Resources Commission (WRC)
2007
Integrated Water Resources Management Plan of the Densu Basin, Component Report
.
Water Resources Commission
,
Accra, Ghana
.
Xu
J.
,
Jin
G.
,
Tang
H.
,
Mo
Y.
,
Wang
Y. G.
&
Li
L.
2019
Response of water quality to land use and sewage outfalls in different seasons
.
Science of The Total Environment
696
,
134014
.
Yadav
S.
,
Babel
M. S.
,
Shrestha
S.
&
Deb
P.
2019
Land use impact on the water quality of large tropical river: Mun River Basin, Thailand
.
Environmental Monitoring and Assessment
191
(
10
),
614
.