International Journal of Energy and Water Resources (2021) 5:193–204

https://doi.org/10.1007/s42108-021-00125-5

ORIGINAL ARTICLE

Impact of land‑use/land‑cover dynamics on water quality in the Upper

Lilongwe River basin, Malawi
I. S. Nkwanda1 · G. L. Feyisa2 · F. Zewge1 · R. Makwinja1,3 

Received: 6 February 2021 / Accepted: 31 March 2021 / Published online: 13 April 2021
© Islamic Azad University (IAU) 2021

Abstract
The United Nations 2018 World Water Development report projected that by 2050, nearly 6 billion globe population is likely
to face acute water scarcity. In Africa, about 65% of the population has no access to clean water. In Malawi, increased water
pollution instigated by environmental degradation and land-use/land-cover dynamics (LULCD) is one of the major causes
of clean water scarcity. The situation is worse, especially in the Malawi capital city Lilongwe. This study used 1989, 1994,
2014, and 2019 Landsat images to dissect the link between water quality variation and LULCD in the Upper Lilongwe River
catchment. The water quality data were obtained from Water Quality and Environmental Management Division, Lilongwe
Water Board. The LULCD results showed a decrease in forest land (46.6%), water body (25%), and grass/shrubs/bare land
(11.5%). The built-up area increased by 130% while cultivated land by 35.2%. The study showed an increase in turbidity,
nitrate, phosphate, sulfate, biological oxygen demand (BOD), chemical oxygen demand (COD), total suspended solids (TSS),
electric conductivity (EC), faecal coliform, and total dissolved solids (TDS). Dissolved oxygen (DO) and pH on the other
hand decreased. The LULCD (cultivated and built-up areas) correlated positively with major nutrient loading, EC, faecal
coliform, TDS, BOD, COD, and TSS. The DO and pH negatively correlated with cultivated and built-up classes implying
that LULCD to some extent contributed to the water pollution in the river. Therefore, proper land-use planning and catch-
ment management are required to improve the status of water quality in the river.

Keywords  Acute water scarcity · Environmental degradation · Landsat images · Water pollution

Introduction about 40% of the river systems in the world have been pol-
luted (Bi et al., 2018), and the statistics show the increas-
Water quality deterioration is one of the critical challenges ing trend with extension from tributaries to the mainstream,
facing many countries across the globe (Biswas & Torta- diffusion from regional basins of Sub-Saharan Africa, pen-
jada, 2019; Boretti & Rosa, 2019). The situation is worse in etration from the surface to ground, and spreading from
Africa where rapid population growth in different countries urban to rural areas (Guan & Hubacek, 2008; Jiang, 2009;
is not matched with the available natural resources. With the Kanyika-Mbewe et al., 2020). Human activities such as
increased human population and socio-economic develop- increased deforestation, wide expansion of economic activ-
ment, the global water scarcity is grim, water pollution is so ities, rapid population growth, and increased agricultural
exacerbated, and surface freshwater pollution has become a activities are the major drivers of LULC dynamics (Bone
major public hazard (Bi et al., 2018). Estimates show that et al., 2017; Gondwe et al., 2019; Hu et al., 2008; Showqi
et al., 2014). Many researchers investigating the linkages
* R. Makwinja between LULCD and water quality changes at the African
[email protected] Sub-Saharan regional, national, and local level have pointed
out that a significant relationship between LULCD and water
1
African Center of Excellence for Water Management, Addis quality changes exists (Bu et al., 2014; Chen et al., 2016;
Ababa University, P.O. Box 1176, Addis Ababa, Ethiopia
Du Plessis et al., 2014; Rajaei et al., 2018; Teixeira et al.,
2
Center for Environmental Sciences, Addis Ababa University, 2014; Zhou et al., 2012). Changes in nutrient concentration,
P.O. Box 1176, Addis Ababa, Ethiopia
suspended sediments, frequent floods, changes in baseflow,
3
Senga Bay Fisheries Research Center, P. O. Box 316, Salima, annual mean discharges, seasonal variation in streamflow,
Malawi

13
Vol.:(0123456789)

194 International Journal of Energy and Water Resources (2021) 5:193–204

evapotranspiration, surface runoff generation, soil water to study the status of water quality in this river. Studies such
content, groundwater recharge, water scarcity, catchment as those of Phiri et al. (2005), Dryton (2008), and Nyasulu
degradation, pollution, salinization, low dissolved oxygen, (2008) have pointed out that the status of water quality in
increased silt load, high alkalinities, low pH, and weed inva- the river is currently far below the WHO standard for clean
sion are linked to LULCD (Makwinja et al., 2019; Munthali water. While the Malawi government enforces its legislation
& Murayama, 2011; Pullanikkatil et al., 2016). to control pollution based on these findings, there is also an
Malawi is ranked as one of the countries in the world urgent need to dissect the link between LULCD and water
highly endowed with a variety of freshwater systems (Mak- quality changes that drive water scarcity to stir discussion
winja et al., 2019), including Lake Malawi, Lake Malombe, among policymakers. Therefore, the main objective was to
Lake Chilwa, Shire River, and several networks of river dissect the link between water quality variation and LULCD
systems such as Lilongwe River, Dwangwa, Bua, Linthipe, to provide baseline information for policymakers and other
Ruo, South and North Rukulu, and other small streams stakeholders of Lilongwe city to make an informed decision
which form a backbone of the national economy (Kum- in managing the water resources within the basin and also
ambala, 2010). Despite having this variety of freshwater add to the body of knowledge in the area of sustainable water
systems, Malawi is still ranked among the countries in the and land resource management.
Sub-Saharan region least likely to achieve the Sustainable
Development Goal for safely managed drinking water. The
country requires heavy investment to achieve ‘basic’ drink- Materials and methods
ing water for everyone by 2030 (Hope & Ballon, 2019).
Malawi government report projected that by 2025, Malawi Study area
is likely to experience severe water scarcity (GoM, 2004).
Rapid population growth, climate change, increased poverty, The study was conducted from February to March 2019 in
environmental degradation and unsustainable use of forest the Upper Lilongwe River catchment (Fig. 1). The Upper
land, and poor agricultural activities are the major causes of Lilongwe River catchment stretches from the Dzalanyama
increased water pollution in Malawi (GoM, 2016; Kambwiri forest reserve that forms the boundary between Malawi
et al., 2014; NSO, 2016). Researchers have ranked Malawi and Mozambique to Bwaila South in Lilongwe city and
as one of the highest in terms of forest loss in the SADC lies between Latitudes − 14.3000160 and − 14.0062090 S
region with estimation showing that about 30,000–40,000 and Longitudes 33.4819180 and 33.7578520 E, and has a
hectares of forested land are lost annually due to wide expan- catchment area of approximately 1­ 870km2. Lilongwe River
sion of agricultural land and excessive use of biomass such is composed of five main tributaries such as Likuni, Katete,
as wood and charcoal (Ngwira & Watanabe, 2019). Studies Lisungwe, Nanjiri, and Nathenje, and is the only source of
such as those of Nyasulu, (2008), Kwanjana (2009), and water for Lilongwe city. It is dammed twice to create storage
Chidya et al. (2019) ranked Lilongwe River in the Central for the dry season. The two dams—Kamuzu Dam 1 (KD1)
Region as one of the rivers in Malawi experiencing a seri- and Kamuzu Dam 2 (KD2)—constructed in 1966 and 1989,
ous threat of pollution due to increased human activities. Its respectively, have a combined storage of 24 million m ­ 3,
upper catchment has seriously degraded due to continuous which is barely able to sustain current demand during the
LULC dynamics. It is expected that in the next few decades, dry season. Water is abstracted at an intake point on the river
water pollution will become a serious threat to sustainable about 20 km downstream of KD2 and treated in Treatment
development in the city. Currently, about 90% of industrial, Works 1 (TW1) and 2. The catchment of Upper Lilongwe
domestic, and municipal wastewater are discharged into the River generally experiences a warm tropical climate with
river and with the increased urbanization, effluents from the an average temperature of about 15 °C during the cool dry
wastewater are projected to increase. Nutrient loading is season and 30 °C during the hot dry season, average annual
the most dangerous water quality threat and is often associ- rainfall ranging between 800 and 1000 mm (Sibande et al.,
ated with pathogen loading. Agricultural activities are the 2020), and mean monthly streamflow of − 0.058 m ­ 3 ­s−1 dur-
predominant source of phosphorous and nitrogen and the ing the dry season (August–November) and + 1.432 ­m3 ­s−1
current level of nitrogen and phosphorous pollution from during the wet season (December–April). Lilongwe city
agriculture is expected to increase due to the increased con- population has increased to more than 3000% between 1966
version of forest land to agriculture in the upper catchment and 2008, thus, from 19,425 to 674,448 though the current
of the river. Rapid population growth, wide economic explo- estimate could be higher than this.
sion, and lack of effective wastewater treatment are further
expected to increase pollution in the Lilongwe River. Due to
increased demand for clean water and increased water pol-
lution, there has been a growing interest among researchers

13
International Journal of Energy and Water Resources (2021) 5:193–204 195

Fig. 1  Upper Lilongwe River catchment

Data collection and analysis using the multiple-tube-fermentation-technique. This tech-

nique involves a series of preliminary and confirmatory
The surface water quality data of the Upper Lilongwe tests in which gas bubbles are formed in a small glass
River catchment were obtained from the Water Quality vial by the action of coliform bacteria in a lactose broth
and Environmental Management Division of the Lilongwe medium and a statistical analysis, called the most probable
Water Board. Water quality parameters such as pH, DO, number (MPN), is then made (Schuettpelz, 1969).
BOD, COD, turbidity, faecal coliform, EC, TDS, tempera- The primary and secondary data were acquired for LULC
ture, TSS, nitrate, phosphate, and sulfate were selected. analysis. An exploratory survey was conducted to precede
The selected parameters represented the major physical, ground-truthing and also to understand the socio-economic
chemical, and biological quality of water in the river sys- activities in the study area. Satellite images of the Landsat
tem. Water temperature, pH, electrical conductivity (EC), Thematic map of 1989, 1994, 2014, and 2019 were used.
total dissolved solids (TDS), and dissolved oxygen (DO) All images had a spatial resolution of 30 m which is large
are measured on-site using WTW MultiLine P4 multi- enough to visualize changes in land use from Landsat 5,
parameter probe, model (Multi 3430). The BOD and COD 7, and 8. Other geospatial data such as shapefiles and top-
are measured using Proteus Multi-parameter Water Qual- ographic maps for the delineation of the study area were
ity Sensor (model2018PESC1138f). Turbidity (NTU) and obtained from the Malawi Department of Survey. Image
total suspended solids (TSS) are measured using a turbid- processing operations include geo-referencing, enhance-
ity meter (Model No: TU-2016). Phosphate, nitrate, and ment, and false-color composite for discrimination of fea-
sulfate ( PO−3
4  , NO3  ,
−
4  ) are determined using UV/vis-
SO2− tures were also performed. The land in the study area was
ible spectrophotometer (model T90) following Molybde- classified according to its description such as cultivated
num blue method in conjunction with Ultraviolet Spectro- land, water, forest, grass, shrubs, settlements, and others.
photometry for NO−3 and PO−3 4  , and Turbidimetric method
This was done using spectral signatures. Using prior knowl-
for SO2−4
. The determination of faecal coliform is done edge of the area and secondary data (LULC maps obtained

13

196 International Journal of Energy and Water Resources (2021) 5:193–204

from the department of the survey), each pixel was classi- scientific studies associated with the accumulation of these
fied to the most likely class based on the Maximum Likeli- contaminants in water are scarcely available. This study used
hood algorithm where each pixel is assigned to the spectral nutrient loading (nitrate, phosphate, and sulfate), EC, faecal
class that has the greatest probability density function for coliform, TDS, BOD, COD, TSS, pH, and turbidity as an
the multispectral values of the pixel. The catchment was indicator of the suitability of water in the upper catchment of
classified into five land-use/land-cover classes (Table 1). All Lilongwe River for both rural and urban consumption. The
the five classes presented in Table 1 were verified by field pH of water plays an important role in chemical speciation
observation and through key informant interviews with the and chemical dissolution and precipitation. The negative pH
elderly people who have continuously been residing in the trend in this study suggests increased metal dissolution in
area for more than 50 years and have seen the changes in the the river and a high probability of bioavailability and toxicity
area over the years. The reference data for the 2019 LULC to humans and aquatic biota. Faecal coliform is used as an
map were also collected randomly in the catchment from all indicator of faecal pollution in the water, while turbidity and
the five classes using Garmin handheld GPS. The archived TSS are associated with the presence of viruses, bacteria,
Google earth images were used to extract reference data for and other toxic chemicals such as heavy metals. Van Loon
the 2019 image. The confusion matrix and Kappa coefficient and Duffy (2005) suggested that excessive nitrates ­(NO3−)
were used to assess the accuracy of the LULC maps. can cause many health disorders such as methemoglobine-
mia, gastric cancers, goiter, and hypertension. The presence
Statistical analysis of sulfate in a very high concentration is also associated with
respiratory problems, and in conjunction with sodium and
Geospatial LULC Landsat satellite data were analyzed using magnesium, sulfate can exert a cathartic effect on diges-
GIS and remote-sensing approaches. ArcGIS version 10.7.1 tive tracts. This study showed that the water pH, DO, and
and ENVil 5.4 software were used. The mixed-effects linear BOD in the river have been fluctuating over the past 30 years
regression model was used to explain the extent to which (Fig. 2). The highest BOD was recorded in 2019, while pH
LULC classes influence water quality parameters in the and DO showed a negative trend (Fig. 2a). These negative
study area. The STATA version 12 was used for inferential trends exhibited by the pH and the DO are linked to the
analysis. increased dynamics of forest cover, livestock excrement, and
municipal and domestic effluent discharged directly into the
river, causing depletion of oxygen through biochemical pro-
Results and discussion cesses such as decomposition of organic materials—a condi-
tion that leads to higher BOD and COD (Bakure et al., 2020;
Studies have shown that over 80% of the Malawi rural popu- Bu et al., 2014; Chemura et al., 2020; Ding et al., 2016;
lation depends on unprotected water from rivers, streams, Fentie et al., 2020; Howell et al., 2016; Iloms et al., 2020;
lakes, wetlands, hand-dug wells, springs, and surface water Kannel et al., 2008; Kükrer & Mutlu, 2019; Tahiru et al.,
(Kanyika-Mbewe et al., 2020; Makwinja et al., 2019), though 2020; Zhou et al., 2012). Annual average EC, TDS, tem-
it is very apparent that access to safe clean water is currently perature, and COD also fluctuated positively with both EC
a global concern. In the upper catchment of Lilongwe River, and TSS registering the peak in 2019 while EC in 2014 and
water from these unprotected sources is heavily contami- TDS in both 2014 and 2019, respectively (Fig. 2b). The TDS
nated with agricultural chemicals such as inorganic ferti- throughout the study period was below the World Health
lizers, insecticides, and other human contaminants, though Organisation (WHO) guideline of 50  mg/L. The lowest

Table 1  Description of LULC categories

Land cover Descriptions

Cultivated land This category applies to cultivatable land and cultivated land with all planted crops such as paddy, maize, and other
harvestable crops, and also to agroforestry areas which are not defined as forest land by the national definition
Shrubs/grassland, bare land This category includes pastures, savanna vegetations, and grazing lands, bare lands, rocks, and unmanaged areas
which are not covered by any of the categories. Shrubs were defined as small either closed or scattered trees
Built upland This includes an area used for settlements, permanent concentration buildings, and man-made structures such as
transport infrastructure and activities, ranging from large villages to town scale
Water bodies Water bodies were defined as areas that are naturally covered by water such as wetland, swamps, lakes, rivers,
reservoirs, and others
Forest land This includes general woodland with herbaceous layer, broad-leaved deciduous trees, closed and vegetative cover
with abiotic and biotic forces of its biotope

13
International Journal of Energy and Water Resources (2021) 5:193–204 197

Fig. 2  Water quality parameters 16 14.44

for upper Lilongwe River 450
14
11.65 400 EC(ums/cm)
12 10.5 10.45
9.57 350
10 TDS(mg/L)
9.12 9.32 300
8 7.89 250
5.67 6.98 7.48
6.7 6.89 6.7
6 5.89 200 TEMP (0C)
5.72 5.7
4.67 150
4
100 COD (mg/L)
2
50
0 TSS (mg/l)
1989 1994 2000 2014 2016 2019 0

pH DO BOD

SO4 (mg/l)
Faecal coliform count/100ml
Phosphate (mg/l)
700 26160 30000 Nitrate (mg/l)
600 614 25000
500 20000 4 150
400 433
15000 3
300 100
10000 2
200 195 50
147 2900 5000 1
100 10871.7 281 820 592
034.3 0 0 0
1989 1994 2000 2014 2016 2019 1989 1994 2000 2014 2016 2019

turbidity (108 NTU) was recorded in 1989 and the maximum Malawi’, it was revealed that increased in faecal coliform by
(614 NTU) was recorded in 2019. However, the turbidity 176.1% and Escherichia coli counts by 157% in the down-
values throughout the study period were higher than 0.2NTU stream of urban areas were highly linked to human activities
recommended by WHO guidelines for clean water and the which conforms to the current findings in Upper Lilongwe
Malawi Bureau of Standards (MBS, 2017; WHO, 2008;). River. Nitrate, sulfate, and phosphate occur naturally from
Researchers such as Razali et al. (2018) and Woldeab et al. the weathering process. The study showed a temporal varia-
(2019) explained that increased agricultural activities are tion of nitrate and phosphorous concentration in the river for
highly linked to increased turbidity in the river system. This the past 30 years with the highest concentration recorded in
study showed that cultivated land in the Upper Catchment of 2019 and the lowest in 1989. This increase in concentration
Lilongwe River has been increasing since 1989 and it is very could be attributed to increased sewage effluents and sur-
apparent that such increase influenced the turbidity. Didone face runoff from non-point sources such as agricultural land,
et al. (2014) also suggested that an increase in annual flood- grazing land, settlements, urban areas, and point source such
ing events and poor agricultural activities in the catchment as industry. This is evident in LULCD analysis of cultivated
is highly linked to high turbidity in the river. Towards the land and built-up areas in Table 2. Furthermore, increased
end of the 2019 rainy season, Malawi experienced one of built-up areas and decreasing glass/shrubland (Table  2)
the worst strong winds, heavy rains, and floods induced by were an indicator of a high rate of conversion of forest land
severe weather system (Tropical Depression 11) that formed into settlements and grazing land. Kindu et al. (2013) also
offshore central Mozambique and hit Malawi before loop- reported the same in the Munessa–Shashemene Landscape
ing and tracking back and converting into intense Tropical of the Ethiopian Highlands. Nutrient loading is often associ-
Cyclone Idai (TC Idai) that also seriously affected Mozam- ated with pathogen loading. This study recorded the highest
bique and Zimbabwe. These heavy rains lead to increased (26,160 counts/ml) in 2019 and the lowest (34.3 counts/mL)
turbidity in the Lilongwe River in 2019 as compared to other in 1989 above WHO recommended standard of zero counts
previous years. for portability and all purposes. The findings of this study
Faecal coliform (Fig. 2c), sulfate, nitrate, and phosphate conform to the findings of Phiri et al. (2005), Dryton (2008),
(Fig. 2d) also fluctuated positively and registered the peak in and Nyasulu (2008) who reported that uncontrollable anthro-
2019, respectively. Pullanikkatil et al. (2015) explained that pogenic activities such as lack of proper catchment manage-
faecal coliforms increase with the increase in urbanization ment were responsible for water quality deterioration in the
and agricultural activities. In their studies on ‘impact of land Upper Lilongwe River catchment.
use on water quality in the Likangala catchment, southern

13

198 International Journal of Energy and Water Resources (2021) 5:193–204

Table 2  Area distribution of Land cover type 1989 1999 2014 2019

land-cover classes from 1989
to 2019 Area (Ha) % Area (Ha) % Area (Ha) % Area (Ha) %

Forest 66,727 36.9 54,640 30.2 41,388 22.9 35,594 19.6

Waterbody 14,702 8.1 12,769 7.1 11,828 6.6 10,980 6.1
Grass/shrubs/bare land 29,460 16.3 32,025 17.7 26,506 14.7 26,074 14.4
Built upland 14,164 7.8 17,763 9.8 24,554 13.6 32,696 18.1
Cultivated Land 55,924 30.9 63,780 35.2 76,701 42.4 75,633 41.8
Total 180,977 100 180,977 100 180,977 100 180,977 100

Land‑use and land‑cover change in Upper Lilongwe et al. (2020) in Ghana. Conservation efforts have been done
River catchment in the catchment over the past years to reverse the situation.
Recently, the Malawi government deployed armed forces
The study revealed that forest, waterbody, cultivated, grass/ in the Dzalanyama forest to protect the reserve from over-
shrubs/bare land, and built-up area (Fig. 3) were dominant exploitation by the local population. The initiative was taken
land-use and land-cover classes in the catchment under after noting that the rate of deforestation in the catchment
study. In 1989, about 36.9% of the catchment was predomi- was threatening the Lilongwe River, which is the main sup-
nately covered by forest, 8.1% (14702 ha) by water, grass/ ply of water to the Capital City of Malawi (Ngwira & Wata-
shrubs/bare land covered about 16.3% (29460 ha), built- nabe, 2019). However, some key informants expressed that
up 7.8% (14164 ha), and cultivated land covered 30.9% this initiative has been continuously hampered by rapid pop-
(55924 ha) (Table 2). Figure 3 further shows that in 1989, ulation growth, increased poverty, and a lack of alternative
the woodlands on the Dzalanyama mountain range which is livelihood options among the local population. The Upper
the main source of Lilongwe River were intact and the catch- Catchment of Lilongwe River is one of the areas where the
ment had few places with smaller woodlands. The forest majority of the local population are living under extreme
area was large as compared to the other years. From 1989 to poverty that is difficult to escape and face tough options
1999, the forest land was reduced to about 6.7% (54640 ha), between catchment conservation and their survival (Tobin
7.3% (41388 ha) from 1999 to 2014, and 3.3% from 2014 & Knausenberger, 1998). Poverty has pushed the local popu-
to 2019. The reduction in aerial coverage for grass/shrubs/ lation to desperately depend on forest resources for their
bare land within the period from 1989 to 2019 was about sustenance (Meher, 2009; Pullanikkatil et al., 2020). One of
3,386 ha representing 11.49%. The loss of forest land could the key informants experessed that this initiative has fuelled
be attributed to increased charcoal burning, overgrazing, the conflict between the local population and the govern-
and increased settlements which predominately turned the ment authorities. Other conservation efforts introduced by
area into bare land (Jamu et al. 2003; Gondwe et al., 2019; both government and non-governmental organizations have
Munthali & Murayama, 2011; Pullanikkatil et al., 2016; also faced challenges in their implementation due to the lack
Sibande et al., 2020). The rate of increase in human settle- of transformative policies and economic incentive models
ments in the Upper Lilongwe catchment between 1989 and to stimulate the local population to effectively manage the
2019 was 130%, while cultivated land was 35.24% which catchment.
conforms to the findings of Käyhkö et al. (2011) in Tanzania,
Reed et al. (2015) in Botswana, Chinangwa et al. (2016) in Water quality and LULCD in the Lilongwe River
Malawi, Krishna et al. (2017) in Indonesia, Mmbaga et al. Upper Catchment
(2017) in Tanzania, and Ngwira and Watanabe (2019) in
Malawi. The decrease in forest land was an indicator of In this study, nutrient loading (nitrate, phosphate, and sul-
encroachment in the protected area (Palamuleni et al. 2011; fate), EC, faecal coliform, TDS, BOD, COD, TSS, pH, and
Pullanikkatil, 2015) which eventually affected the quality turbidity were the topmost water quality parameters link-
of water in the river such as total coliform counts as pre- ing up with LULCD. Table 3 shows that turbidity levels
viously reported by Chimwanza et al. (2006) in Likangala increased with a decrease in forest land throughout the study
River, Southern Malawi and Kosamu et al. (2011) in Blan- period and these findings are in line with the findings of
tyre City Malawi. Water bodies also showed some negative Hughes et al. (2012) in New Zealand, Harrison et al. (2013)
patterns over the study period (Table 2) and were reduced in south-eastern Australia, and Wan et al. (2014) in Xitiaoxi
from 14702 to 10980 ha representing a 25% decrease which River, China. Cultivated land also affected turbidity levels.
also conforms to the findings of Makwinja et al. (2012) in An increase in cultivated land was associated with increasing
Southern Malawi, Elias et al. (2019) in Ethiopia, and Tahiru turbidity levels, suggesting that reducing cultivated land and

13
International Journal of Energy and Water Resources (2021) 5:193–204 199

Fig. 3  Land use/land cover classes

increasing forest land in the watershed would consequently of Tampa Bay, Florida. It should be noted that bare land
reduce turbidity. Madriñán et al. (2012) and Schoonard et al. is highly exposed to runoff and surface runoff is typically
(2014) also found that increased turbidity and TSS were associated with increased turbidity in many water bodies
highly linked to agricultural activities within the watershed (G ̈oransson et al., 2013; Hughes et al., 2012; McCarthy

13

200 International Journal of Energy and Water Resources (2021) 5:193–204

Table 3  Mixed-effects Factor Coef Std.Err z p >|z| (95% conf. interval)

regression model for forest
land-use class pH − 12,310.13 0.105 − 5.59 0.000** − 1662.18 − 7994.08
Do 29,390.11 92.32 8.18 0.000** 22,349.28 36,430.94
BOD − 2739.61 52.29 − 4.20 0.000** − 4018.1 − 1461.14
Turbidity − 13.871 8.54 − 1.62 0.105ns − 30.62 2.878
Faecal coliform − 0.283 0.1406 − 2.02 0.044* − 0.559 − 0.008
EC − 252.12 3423.69 21.47 0.000** − 331.39 − 172.86
TDS − 460.571 294.35 − 1.56 0.118ns − 1037.48 116.344
Temp − 582.62 1121.59 − 0.52 0.603ns − 2789.91 1615.67
TSS − 166.84 17.959 − 9.29 0.000** − 202.041 − 131.64
Nitrate − 5763.25 1037.94 − 5.55 0.000** − 7797.58 − 3728.92
Phosphate − 20,746.28 984.89 − 21.06 0.000** − 22,676.64 − 18,815.93
Sulphate − 27,726.68 8456.53 − 3.28 0.000** − 44,201.18 87,136.98

Log likelihood = − 35.244, prob > chi 2 = 0.000, Wald Chi 2 = 216.96, ns indicates not significant, while **
and * indicate significance at 0.01 and 0.05 probability level of confidence

et al., 2018). Therefore, the transformation of forest land nitrate, phosphate and sulfate, faecal coliform, and turbid-
by human activities such as increased settlements and agri- ity displayed a positive relationship which is in line with
cultural activities into bare land could be the precursor for the findings of Wan et al. (2014), Schoonover and Lockaby
a high level of turbidity in the river. This general trend has (2006), and Dai et al. (2017). The positive and negative
also been reported by previous researchers such as Ngoye regression coefficients implied that when settlements and
and Machiwa (2004) in Ruvu River Eastern Tanzania and cultivated land increase, there will be an increase in EC,
Namugize et al (2018) in South Africa. The mixed-effects TDS, TSS, COD, nitrate, phosphate, sulfate, faecal coliform,
regression model for built-up and cultivated land-use classes and turbidity, and decrease in pH and DO and the reverse is
and water quality parameters indicated some levels of inter- true (de Mello et al., 2018; Lawniczak et al., 2016; Roberts
actions (Tables 4 and 5) which conforms to the findings of & Marsh, 1987). The aerial coverage of the water body also
Lee et al. (2009), Tanaka et al. (2016), Mupepele and Dor- decreased throughout the period with increasing TDS, EC,
mann (2017) and de Mello et al. (2018). Cultivated land phosphate, sulfate, and DO. The findings are in line with
and built-up land classes displayed a significant (p < 0.05) what Williams and Melack (1997) reported in the Central
regression coefficient with water quality parameters. The Amazon, Welde and Gebremariam (2017) in the Tekeze
water pH and DO display a negative relationship with built- Dam watershed, Northern Ethiopia, and Makwinja et al
up areas and cultivated land, while EC, TDS, TSS, COD, (2019) in Central Malawi. The regression coefficient of

Table 4  Mixed-effects Factor Coef Std.Err z p >|z| (95% conf. Interval)

regression model for built-up
land-use class pH − 8251.95 2621.79 − 3.16 0.002** 13,390.57 3113.338
Do 2025.421 2771.977 0.73 0.465** − 3407.55 7458.398
BOD 881.596 370.389 2.38 0.000** 8.204 27.226
Turbidity 17.71 4.8523.65 3.65 0.000 8.204 27.226
Faecal coliform 0.349 0.03 11.64 0.000 0.29069 0.4084
EC 107.61 8.637 12.46 0.000** 0.2907 0.4084
TDS 262.52 83.167 3.16 0.020 − 99.51 425.52
Temp 1181.04 1614.11 0.73 0.464 − 1982.56 4344.64
TSS 76.029 7.741 9.82 0.000** 60.8568 91.201
COD 34.9681 9.969 3.51 0.000 15.4279 54.508
Nitrate − 879.035 1173.38 − 0.75 0.454 − 3178.82 1420.75
Phosphate 14,697.15 1113.41 13.2 0.000** 12,514.9 16,879.41
Sulphate 14,697.15 1109 11.34 0.000** 9555 13,550.43

Log likelihood = − 234.456, Wald Chi 2 = 6578.89, Prob > Chi2 = 0.000, ns indicates not significant, while
** and * indicate significance at 0.01 and 0.05 probability level of confidence

13
International Journal of Energy and Water Resources (2021) 5:193–204 201

Table 5  Mixed-effects Factor Coef Std.Err z p >|z| (95% conf. Interval)

regression model for cultivated
land-use class pH − 6508.024 426.71 − 15.25 0.000** − 7344.37 − 5671.67
DO − 2946.58 451.16 − 6.53 0.000** − 3830.846 − 2062.32
BOD 1466.97 1170.05 1.25 0.21 − 826.284 3760.22
Turbidity 14.69 0.32 0.96 0.338 − 15.346 44.742
Faecal coliform 0.0175 0.074 0.24 0.814 − 0.128 0.1632
EC 216.016 21.379 21.379 0.000** 174.11 257.92
TDS 402.62 16.755 24.03 0 369.78 435.466
Temp − 267.158 325.197 − 0.82 0.411 − 904.53 370.216
TSS 218.53 20.97 10.42 0.000** 177.41 259.649
COD − 142.38 27.01 − 5.27 0 − 195.34 − 89.43
Nitrate − 142.386 27.018 − 5.27 0 − 195.34 − 89.431
Phosphate 11,054 587.118 18.83 0.000** 9904.15 12,205.61
Sulphate 21,943.28 3793.55 5.78 0.000** 14,508.06 29,378.5

Log likelihood = − 257.104, Wald Chi 2 = 2434.01, Prob > Chi2 = 0.000, ns indicates not significant, while
** and * indicate significance at 0.01 and 0.05 probability level of confidence

BOD, EC, TDS, EC, TSS, Nitrate, Phosphate, and Sulphate of the river was assessed over 30 years to determine its
in Table 6 showed a significant (p = 0.000 and p = 0.000) state. The general findings of the study showed an increase
negative relationship with water body, while pH and DO in both nutrients (nitrate, phosphate, and sulfate), turbid-
was positive, suggesting that decrease in water body influ- ity, and total coliform in the water. The study demonstrated
enced water pH and DO positively, while BOD, TDS, EC, that LULCD that took place in the Upper Lilongwe River
TSS, nitrate, phosphate, and sulfate were influenced nega- catchment over the past 30 years influenced the water qual-
tively (Elka & Laekemariam, 2020; Lawniczak et al., 2016; ity in the river basin. The water resources in the upper
Shrestha & Basnet, 2018; Yong & Chen, 2002). section of the Lilongwe River are very indispensable for
both urban and rural consumption in the Lilongwe dis-
trict. Therefore, this study recommends that proper land-
Conclusion use planning and catchment management are essential to
improve and sustain the quality of water in the river.
This study assessed the link between the upper catchment
of Lilongwe River LULC classes and the quality of water
in the river basin. The surface water from the upper section

Table 6  Mixed-effects Factor Coef Std.Err z p >|z| (95% conf. Interval)

regression model for waterbody
land-use class pH − 1753.46 46.12 38.02 0.000** 1663.07 1843.86
Do 423.19 48.76 − 8.68 0.000** − 518.77 − 327.62
BOD − 484.12 54.38 − 8.9 0.000** − 590.71 − 377.544
Turbidity 0.769 0.124 1.08 0.280 ns − 0.62722 2.1651
Faecal coliform − 0.036 0.027 − 1.33 0.0.183 − 0.0896 0.01709
EC 26.336 7.834 − 3.36 0.001** − 41.689 − 10.981
TDS 93.29 39.609 − 2.35 0.019 − 171.08 − 15.505
Temp 282.55 153.27 − 1.84 0.060 ns − 582.95 17.855
TSS 17.217 2.4543 − 7.02 0.000** − 22.027 − 12.407
Nitrate − 16.17 6.596 − 1.68 0.093 − 2727.42 207.89
Phosphate 2427.167 464.55 − 5.22 0.000** − 3337.67 − 1516.66
Sulphate 14,911.34 365.72 40.77 0.000** 14,194.53 15,628.15

Log likelihood = − 21.278, Wald Chi 2 = 3131.41, Prob > Chi2 = 0.000, ns indicates not significant, while
** and * indicate significance at 0.01 and 0.05 probability level of confidence

13

202 International Journal of Energy and Water Resources (2021) 5:193–204

Acknowledgements  The authors wish to thank the Lilongwe Water Forestry Research, 2016, 12. https://​doi.​org/​10.​1155/​2016/​58470​
Board Central Laboratory for providing time-series water sample data 68 Article ID 5847068.
and laboratory analysis. Dai, X., Zhou, Y., Ma, W., & Zhou, L. (2017). Influence of spatial vari-
ation in land-use patterns and topography on water quality of the
Funding  This research was funded by the Federal Democratic Republic rivers inflowing to Fuxian Lake, a large deep lake in the plateau
of Ethiopia through the African Centre of Excellence for Water Man- of southwestern China. Ecological Engineering, 99, 417–428.
agement, Addis Ababa University, and Grant number: GSR/7365/11. https://​doi.​org/​10.​1016/j.​ecole​ng.​2016.​11.​011.
de Mello, K., Valente, R. V., Randhir, T. O., dos Santos, A. C. A., &
Vettor, C. A. I. (2018). Effects of land use and land cover on water
Declarations  quality of low-order streams in Southeastern Brazil: Watershed
versus riparian zone. CATENA, 167, 130–138. https://​doi.​org/​10.​
Conflict of interests  The authors of this paper declare that there is no 1016/j.​catena.​2018.​04.​027.
conflict of interest. Didoné, E. J., Minella, J. P. G., Reichert, J. M., Merten, G. H., Dal-
bianco, L., Peixoto Barrros, C. A., & Rafael, R. (2014). Impact
of no-tillage agricultural systems on sediment yield in two large
catchments in Southern Brazil. Journal of Soils and Sediments,
14, 1287–1297. https://​doi.​org/​10.​1007/​s11368-​013-​0844-6.
References Ding, J., Jiang, Y., Liu, Q., Hou, Z., Liao, J., Fu, L., & Peng, Q. (2016).
Influence of the land use pattern on water quality in low-order
Bakure, B., Fikadu, S., & Malu, A. (2020). Analysis of physicochemi- streams of the Dongjiang River basin: A multi-scale analysis. Sci-
cal water quality parameters for streams under agriculture, urban ence of the Total Environment, 51–552, 205–216. https://​doi.​org/​
and forest land-use types: In the case gilgel Gibe catchment South- 10.​1007/​s11368-​013-​0844-6.
west Ethiopia. Applied Water Science, 10(234), 1–8. https://​doi.​ Dryton, J. (2008). Biochemical monitoring of some rivers in Malawi. .
org/​10.​1007/​s13201-​020-​01318-9. Ministry of Water Resources Development.
Bi, W., Weng, B., Yuan, Z., Ye, M., Zhang, C., Zhao, Y., Yan, DXu., Du Plessis, A., Harmse, T., & Ahmed, F. (2014). Quantifying and pre-
& T. . (2018). Evolution characteristics of surface water quality dicting the water quality associated with land cover change: A
due to climate change and LUCC under scenario simulations: A case study of the Blesbok Spruit Catchment South Africa. Water,
case study in the Luanhe River Basin. International Journal of 6(10), 2946–2968.
Environmental Research and Public Health, 15(8), 1724. https://​ Elias, E., Seifu, W., Tesfaye, B., & Girmay, W. (2019). Impact of land
doi.​org/​10.​3390/​ijerp​h1508​1724. use/cover changes on lake ecosystem of Ethiopia central rift val-
Biswas, A., & Tortajada, C. (2019). Water quality management: ley. CogentFood & Agriculture, 5(1), 1595876. https://​doi.​org/​10.​
A globally neglected issue. International Journal of Water 1080/​23311​932.​2019.​15958​76.
Resources Development, 35(6), 913–1916. https://​doi.​org/​10.​ Elka, E., & Laekemariam, F. (2020). Effects of organic nutrient sources
1080/​07900​627.​2019.​16705​06. and NPS fertilizer on the agronomic and economic performance
Bone, R. A., Parks, K. E., Hudson, M. D., Tsirinzeni, M., & Willcock, of Haricot Bean (Phaseolus vulgaris L.) in Southern Ethiopia.
S. (2017). Deforestation since independence: A qualitative assess- Applied and Environmental Soil Science, 2020, 9. https://​doi.​org/​
ment of four decades of land-cover change in Malawi. Southern 10.​1155/​2020/​88535​52 Article ID 8853552.
Forest, 79(4), 269–272. Fentie, S., Jembere, K., Fekadu, E., & Wasie, D. (2020). Land use
Boretti, A., & Rosa, L. (2019). Reassessing the projections of the World and land cover dynamics and properties of soils under different
Water Development Report. Clean Water, 2(15), 1–6. https://​doi.​ land uses in the Tejibara watershed Ethiopia. The Scientific World
org/​10.​1038/​s41545-​019-​0039-9. Journal, 2020(12), 2020. https://​doi.​org/​10.​1155/​2020/​14794​60
Bu, H., Meng, W., Zhang, Y., & Wan, J. (2014). Relationships between Article ID 1479460.
land use patterns and water quality in the Taizi River basin, China. GoM. (2004). National water policy. . Government of Malawi.
Ecological Indicator, 41, 187–197. https://​doi.​org/​10.​1016/j.​ecoli​ GoM. (2016). Malawi climate change management policy. . Govern-
nd.​2014.​02.​003. ment of Malawi.
Chemura, A., Rwasoka, D., Mutanga, O., Dube, T., & Mushore, T. Gondwe, M., Cho, M., Chirwa, P., & Geldenhuyw, C. (2019). Land
(2020). The impact of land-use/land cover changes on the water use land cover change and the comparative impact of co-manage-
balance of the heterogeneous Buzi sub-catchment, Zimbabwe. ment and government-management on the forest cover in Malawi
Remote Sensing Applications: Society and Environment, 18, 1–11. (1999–2018). Journal of Land Use Science, 14(4–6), 281–305.
https://​doi.​org/​10.​1016/j.​rsase.​2020.​100292. https://​doi.​org/​10.​1080/​17474​23X.​2019.​17066​54.
Chen, Q., Mei, K., Dahlgren, R. A., Wang, T., Gong, J., & Zhang, M. Göransson, G., Larson, M., & Bendz, D. (2013). Variation in turbid-
(2016). Impacts of land use and population density on seasonal ity with precipitation and flow in a regulated river system—river
surface water quality using a modified geographically weighted Göta Älv SW Sweden. Hydrology and Earth System Sciences, 17,
regression. Science of the Total Environment, 572, 450–466. 2529–2542. https://​doi.​org/​10.​5194/​hess-​17-​2529-​2013.
https://​doi.​org/​10.​1016/j.​scito​tenv.​2016.​08.​052. Guan, D., & Hubacek, K. (2008). A new and integrated hydro-eco-
Chidya, R., Singano, L., Chitedze, I., & Mourad, K. (2019). Standards nomic accounting and analytical framework for water resources: A
compliance and health implications of bottled water in Malawi. case study for North China. J Environ Manag, 88(4), 1300–1313.
International Journal of Environmental Research and Public https://​doi.​org/​10.​1016/j.​jenvma.​2007.​07.​010.
Health, 16(6), 951. https://​doi.​org/​10.​3390/​ijerp​h1606​0951. Harrison, E., Dyer, F., Wright, D., & Levings, C. (2013). Historical
Chimwanza, B., Mumba, P., Moyo, B., & Kadewa, W. (2006). The land-use influences the long-term stream turbidity response to a
impact of farming on river banks on the water quality of the rivers. wildfire. Environmental Management, 53, 393–400. https://​doi.​
International Journal of Environmental Science and Technology, org/​10.​1007/​s00267-​013-​0193-7.
2, 353–358. https://​doi.​org/​10.​1007/​BF033​25896. Hope, R., & Ballon, P. (2019). Global water policy and local payment
Chinangwa, L., Pullin, A., & Hockley, N. (2016). Livelihoods and wel- choices in rural Africa. NPJ Clean Water, 2, 21. https://​doi.​org/​
fare impacts of forest comanagement. International Journal of 10.​1038/​s41545-​019-​0045-y.

13
International Journal of Energy and Water Resources (2021) 5:193–204 203

Howell, C., Myburgh, P., Lategan, E., & Hoffman, J. (2016). Seasonal masters thesis. Zomba, Malawi: University Of Malawi, Chancel-
variation in the composition of winery wastewater in the breede lor College.
river valley with respect to classical water quality parameters. Lawniczak, E., Zbierska, J., Nowak, B., Achtenberg, K., Grześkowiak,
South African Journal of Enology and Viticulture, 37(1), 31–38. A., & Kana, K. (2016). Impact of agriculture and land use on
Hu, H., Liu, W., & Cao, M. (2008). Impact of Land use and land cover nitrate contamination in groundwater and running waters in cen-
changes on the ecosystem in Menglun, Xishuangbanna, Southwest tral-west Poland. Environmental Monitoring and Assessment, 188,
China. Environmental Monitoring and Assessment, 146, 147–156. 172. https://​doi.​org/​10.​1007/​s10661-​016-​5167-9.
https://​doi.​org/​10.​1007/​s10661-​007-​0067-7. Lee, S., Hwang, S., Lee, S., Hwang, H., & Sung, H. (2009). Landscape
Hughes, A., Quinn, J., & McKergow, L. (2012). Land use influences on ecological approach to the relationships of land use patterns in
suspended sediment yields and event sediment dynamics within watersheds to water quality characteristics. Landscape and Urban
two headwater catchments, Waikato, New Zealand. New Zealand Planning, 92, 80–89. https://​doi.​org/​10.​1016/j.​landu​rbplan.​2009.​
Journal of Marine and Freshwater Research, 46(3), 315–333. 02.​008.
https://​doi.​org/​10.​1080/​00288​330.​2012.​661745. Madriñán, M., Al-Hamdan, M., Rickman, D., & Ye, J. (2012). Rela-
Iloms, E., Ololade, O., Ogola, H., & Selvarajan, R. (2020). Investigat- tionship between watershed land-cover/land-use change and water
ing industrial effluent impact on municipal wastewater treatment turbidity status of Tampa Bay Major Tributaries, Florida, USA.
Plant in Vaal, South Africa. International Journal of Environ- Water, Air, & Soil Pollution, 2012(223), 2093–2109.
mental Research and Public Health. https://​doi.​org/​10.​3390/​ijerp​ Makwinja, R., Chapotera, M., Likongwe, P., Banda, J., & Chijere, A.
h1703​1096. (2012). Location and roles of deep pools in Likangala river during
Jamu, D., Chimphamba, J., & Brummett, R. (2003). Land use and Land 2012 recession period of lake Chilwa Basin. International Journal
cover changes in the LIkangala catchment of the Lake Chilwa of Ecology, 2014, 4. https://​doi.​org/​10.​1155/​2014/​294683 Article
basin, Malawi: Implication for managing a tropical wetland. Afri- ID 294683.
can Journal of Aquatic Science, 28(2), 123–135. https://​doi.​org/​ Makwinja, R., Kosamu, I., & Kaonga, C. (2019). Determinants and val-
10.​2989/​16085​91030​95037​77. ues of willingness to pay for water quality improvement: Insights
Jiang, Y. (2009). China’s water scarcity. J Environ Manag, 90(11), from Chia Lagoon Malawi. Sustainability, 11, 4690. https://​doi.​
3185–3196. https://​doi.​org/​10.​1016/j.​jenvm​an.​2009.​04.​016. org/​10.​3390/​su111​74690.
Kambwiri, A., Changadeya, W., Chimphampha, J., & Tandwe, T. MBS. (2017). Catalog of Malawi Standards. . Malawi Bureau of
(2014). Land use impacts on water quality of Rivers draining from Standards.
Mulanje Mountain: A case of Ruo River in Southern Malawi. McCarthy, M., Muller-Karger, F., Otis, D., & Méndez-Lázaro, P.
Malawi Journal of Science and Technology, 10(1), 15–31. (2018). Impacts of 40 years of land cover change on water quality
Kannel, P., Lee, S., & Lee, Y. (2008). Assessment of spatial-temporal in Tampa Bay Florida. Cogent Geoscience, 4, 1. https://​doi.​org/​
patterns of surface and groundwater qualities and factors influenc- 10.​1080/​23312​041.​2017.​14229​56.
ing management strategy of the ground system in an urban river Meher, R. (2009). Globalization, displacement and the livelihood
corridor of Nepal. Journal of Environmental Management, 86, issues of tribal and agriculture dependent poor people: the case
595–604. https://​doi.​org/​10.​1016/j.​jenvm​an.​2006.​12.​021. of mineral-based industries in India. Journal of Developing Socie-
Kanyika-Mbewe, C., Thole, B., Makwinja, R., & Kaonga, C. ties, 25(4), 457–480. https://​doi.​org/​10.​1177/​01697​96X09​02500​
(2020). Monitoring of carbaryl and cypermethrin concentra- 403.
tions in water and soil in Southern Malawi. Environmental Mmbaga, N., Munishi, L., & Treydte, A. (2017). How dynamics and
Monitoring and Assessment, 192, 595. https://​doi.​org/​10.​1007/​ drivers of land use/land cover change impact elephant conserva-
s10661-​020-​08557-y. tion and agricultural livelihood development in Rombo Tanzania.
Käyhkö, N., Fagerholm, N., Asseid, B., & Mzee, A. (2011). Dynamic Journal of Land Use Science, 12(2–3), 168–181. https://​doi.​org/​
land use and land cover changes and their effect on forest 10.​1080/​17474​23X.​2017.​13133​24.
resources in a coastal village of Matemwe, Zanzibar Tanzania. Munthali, K., & Murayama, Y. (2011). Land use/cover change detec-
Land Use Policy, 28(1), 26–37. https://​doi.​org/​10.​1016/j.​landu​ tion and analysis for Dzalanyama forest reserve, Lilongwe. Proce-
sepol.​2010.​04.​006. dia Social and Behavioral Science, 21, 203–211. https://​doi.​org/​
Kindu, M., Schneider, T., Teketay, D., & Knoke, T. (2013). Land 10.​1016/j.​sbspro.​2011.​07.​035.
Use/land cover change analysis using object-based classification Mupepele, A., & Dormann, C. (2017). Influence of forest harvest on
approach in Munessa-Shashemene landscape of the Ethiopian nitrate concentration in temperate streams—a meta-analysis. For-
Highlands. Remote Sens, 2013(5), 2411–2435. https://​doi.​org/​ ests, 8, 5. https://​doi.​org/​10.​3390/​f8010​005.
10.​3390/​rs505​2411. Namugize, J., Jewitt, G., & Graham, M. (2018). Effects of land use and
Kosamu, I., Mawenda, J., & Mapoma, H. (2011). Water quality changes land cover changes on water quality in the uMngeni river catch-
due to abattoir effluent: A case on Mchesa Stream in Blantyre, ment, South Africa. Physics and Chemistry of the Earth. https://​
Malawi. African Journal of Environmental Science and Technol- doi.​org/​10.​1016/j.​pce.​2018.​03.​013.
ogy, 5(8), 589–594. Ngoye, E., & Machiwa, J. (2004). The influence of land-use patterns
Krishna, V., Kubitza, C., Pascual, U., & Qaim, M. (2017). Land mar- in the Ruvu river watershed on water quality in the river sys-
kets, property rights, and deforestation: Insights from Indonesia. tem. Physics and Chemistry of the Earth A, B, C, 29(15–18),
World Development, 99(2017), 335–349. 1161–1166.
Kükrer, S., & Mutlu, E. (2019). Assessment of surface water quality Ngwira, S., & Watanabe, T. (2019). An analysis of the causes of defor-
using water quality index and multivariate statistical analyses in estation in Malawi: A case of Mwazisi. Land, 8(3), 48. https://d​ oi.​
Saraydüzü Dam Lake, Turkey. Environmental Monitoring and org/​10.​3390/​land8​030048.
Assessment. https://​doi.​org/​10.​1007/​s10661-​019-​7197-6. NSO. (2016). Malawi demographic and health survey 2015–2016. .
Kumambala, P. (2010). Sustainability of water resources develop- National Statistics Office.
ment for Malawi with particular emphasis on North and Central Nyasulu, T. (2008). Assessment of water quality in Urban Rivers—
Malawi.Ph.D. thesis. Glasgow: University of Glasgow. a case study of Lilongwe River in Malawi. . University of
Kwanjana, A., (2009). Assessment of Biochemical Constituents Levels Zimbabwe.
in the Water of the Upper Catchment of Lilonger River—Malawi. Palamuleni, L., Ndomba, P., & Annegarn, H. (2011). Evaluating land
cover change and its impact on hydrological regime in Upper

13

204 International Journal of Energy and Water Resources (2021) 5:193–204

Shire river catchment Malawi. Regional Environmental Change, evaluation of percentage contribution to electrical conductiv-
11(4), 845–855. https://​doi.​org/​10.​1007/​s10113-​011-​0220-2. ity. Journal of Chemistry, 2018, 9. https://​doi.​org/​10.​1155/​2018/​
Phiri, O., Mumba, P., Moyo, B., & Kadewa, W. (2005). Assessment 83696​13 Article ID 8369613.
of the impact of industrial effluents on water quality of receiving Sibande, C., Mtonga, E., Chidya, R. C. G., Kanzunguze, A., & Wand,
rivers in urban areas of Malawi. International Journal of Envi- E. M. (2020). An Assessment of streamflow linkage between land
ronmental Science & Technology, 2, 237–244. https://​doi.​org/​10.​ use or land cover change and in Lilongwe River Basin, Malawi.
1007/​BF033​25882. Malawi Journal of Science and Technology, 12(1), 1–17.
Pullanikkatil, D., Mograbi, P. J., Palamuleni, L., Ruhiiga, T., & Shack- Tahiru, A., Doke, D., & Baatuuwie, B. (2020). Effects of land use and
leton, C. (2020). Unsustainable trade-offs provisioning ecosystem land cover changes on water quality in the Nawuni Catchment of
services in rapidly changing Likangala River catchment in south- the White Volta Basin Northern Region Ghana. Applied Water
ern Malawi. Environment, Development and Sustainability, 22, Science. https://​doi.​org/​10.​1007/​s13201-​020-​01272-6.
1145–1164. https://​doi.​org/​10.​1007/​s10668-​018-​0240-x. Tanaka, O., Teixeira de Souza, A., Moschini, L., & de Oliveira, A.
Pullanikkatil, D., Palamuleni, L., & Ruhiiga, T. (2015). Impact of land (2016). Influence of watershed land use and riparian character-
use on water quality in the Likangala catchment, southern Malawi. istics on biological indicators of stream water quality in south-
African Journal of Aquatic Science, 40(3), 277–286. https://​doi.​ eastern Brazil. Agriculture, Ecosystems and Environment, 216,
org/​10.​2989/​16085​914.​2015.​10777​77. 333–339. https://​doi.​org/​10.​1016/j.​agee.​2015.​10.​016.
Pullanikkatil, D., Palamuleni, L., & Ruhiiga, T. (2016). Assessment Teixeira, Z., Teixeira, H., & Marques, J. (2014). Systematic processes
of land-use change in Likangala River catchment, Malawi: A of land use/land cover change to identify relevant driving forces:
remote sensing and DPSIR approach. Applied Geography, 71, Implications on water quality. Science of the Total Environment,
9–23. https://​doi.​org/​10.​1016/j.​apgeog.​2016.​04.​005. 470–471(2014), 1320–1335. https://​doi.​org/​10.​1016/j.​scito​tenv.​
Rajaei, F., Sari, E. A., Salmanmahiny, A., Randhir, T. O., Delavar, M., 2013.​10.​098.
Dehrooz, R. D., & Bavani, A. M. (2018). Simulating the long-term Tobin, R. J., & Knausenberger, W. I. (1998). Dilemmas of develop-
effect of Hyrcanian forest loss on phosphorous loading at the sub- ment: Burley tobacco, the environment and economic growth in
watershed level. Journal of Arid Land, 10(2018), 457–469. https://​ Malawi. Journal of Southern African Studies, 24(2), 405–424.
doi.​org/​10.​1007/​s40333-​018-​0012-3. https://​doi.​org/​10.​1080/​03057​07980​87085​82.
Razali, A., Ismail, S. N., Awang, S., Praveena, S. M., & Abidin, E. Z. Van Loon, G., & Duffy, S. (2005). Environmental chemistry: a global
(2018). Land-use change in the highland area and its impact on perspective. (2nd ed.). Oxford University Press.
river water quality: A review of case studies in Malaysia. Ecologi- Wan, R., Cai, S., Li, H., Yang, G., Li, Z., & Nie, X. (2014). Inferring
cal Processes. https://​doi.​org/​10.​1186/​s13717-​018-​0126-8. land use and land cover impact on stream water quality using a
Reed, M. S., Stringer, L. C., Dougill, A. J., Perkind, J. S., Atlhopheng, Bayesian hierarchical modeling approach in the Xitiaoxi River
J. R., Mulale, K., & Favretto, N. (2015). Reorienting land degra- Watershed, China. Journal of Environmental Management, 133,
dation towards sustainable land management: Linking sustainable 1–11. https://​doi.​org/​10.​1016/j.​jenvm​an.​2013.​11.​035.
livelihoods with ecosystem services in rangeland systems. Journal Welde, K., & Gebremariam, B. (2017). Effect of land uses land cover
of Environmental Management, 151, 472–485. https://​doi.​org/​10.​ dynamics on the hydrological response of watershed: A case study
1016/j.​jenvm​an.​2014.​11.​010. of Tekeze Dam watershed, northern Ethiopia. International Soil
Roberts, G., & Marsh, T. (1987). The effects of agricultural practices and Water Conservation Research, 5(1), 1–16. https://​doi.​org/​10.​
on the nitrate concentrations in the surface water domestic supply 1016/j.​iswcr.​2017.​03.​002.
sources of western Europe. International Association of Hydro- WHO. (2008). World Health Organisation (WHO) guidelines for drink-
logical Science, 164(1987), 365–380. ing water quality. . World Health Organisation.
Schoonard, C., Collins, J., Paxton, C., & Meindl, C. (2014). Explaining Williams, M., & Melack, J. (1997). Solute export from forested and
the spatial variability of summer rainfall in Pinelas County Florida partially deforested catchments in the Central Amazon. Biogeo-
using ArcGIS. Physical Geography, 35(2), 107–122. https://​doi.​ chemistry, 38(1997), 67–102. https://​doi.​org/​10.​1023/A:​10057​
org/​10.​1080/​02723​646.​2014.​895893. 74431​820.
Schoonover, J., & Lockaby, B. (2006). Land cover impacts on stream Woldeab, B., Ambelu, A., Mereta, S., & Beyene, A. (2019). Effect of
nutrients and fecal coliform in the lower Piedmont of west Geor- watershed land use on tributaries’ water quality in the East Afri-
gia. Journal of Hydrology, 331(3–4), 371–382. https://​doi.​org/​10.​ can Highland. Environmental Monitoring and Assessment. https://​
1016/j.​jhydr​ol.​2006.​05.​031. doi.​org/​10.​1007/​s10661-​018-​7176-3.
Schuettpelz, D. (1969). Fecal and Total Coliform Tests. . Department Yong, S., & Chen, W. (2002). Modeling the relationship between land
of Natural Resources. use and surface water quality. Journal of Environmental Manage-
Showqi, I., Rashid, I., & Romshoo, S. (2014). Land use land cover ment, 66(4), 377–393. https://​doi.​org/​10.​1006/​jema.​2002.​0593.
dynamics as a function of changing demography and hydrol- Zhou, T., Jiangou, W., & Peng, S. (2012). Assessing the effects of
ogy. GeoJournal, 79, 297–307. https:// ​ d oi. ​ o rg/ ​ 1 0. ​ 1 007/​ landscape pattern on river water quality at multiple scales: A
s10708-​013-​9494-x. case study of the Dongjiang River watershed, China. Ecological
Shrestha, A., & Basnet, N. (2018). The correlation and regression Indicators, 23, 166–175. https://​doi.​org/​10.​1016/j.​ecoli​nd.​2012.​
analysis of physicochemical parameters of river water for the 03.​013.

13