The land use within a catchment has great impacts on the water quality of rivers (Huang et al., 2013). The water quality of rivers may degrade due to the changes in the land cover patterns or land use practices within the catchment as human activities increase (Sliva et al., 2001; Ngoye and Machiwa, 2004; Huang et al., 2013).

Comparative studies have found that land use significantly impacts river water quality and that the mechanisms involved can be complex. Human activities such as deforestation, agricultural activities and urbanization generally modify landscape characteristics, alter runoff volume, change water temperature, generate pollution, increase algal production and decrease concentrations of dissolved oxygen in water bodies (Ding et al., 2015). In the course of industrialization, urbanization and agricultural expansion many countries strongly depend on natural resources and this result in land use and land cover change (Lamek et al., 2016).

In Rwanda, surface water is currently polluted by various land use practices such as the use of fertilizers and pesticides in agriculture to improve the yield productivity as the soil is becoming more and more degraded (Christian et al., 2012). These chemicals find their way into surface water through runoff. In the same vein, land use practices such as trampling of stocks, human disturbances, burning of vegetation, increased housing developments associated with urbanization, dumping of untreated effluent in rivers and marshlands, roofing of housing complexes and paving of roads and other access routes, soil excavation processes have devastated vegetation cover to such an extent that the soil surface of areas has become susceptible to erosion (Christian et al., 2012). The impact of intense land use and land use practices in the Albertine Rift region was highlighted at the side of Democratic Republic of Congo (Kimbadi et al., 1999; Bagalwa, 2006). From our knowledge, similar assessment is missing for Sebeya River and its catchment. The Sebeya catchment is part of the Albertine Rift region on Rwanda side, where (rural) population density and land cover/use changes are much higher in various river catchments of Rwanda, pollution is an issue of concern. For instance, the Nyabugogo River carries high loads of nutrients in terms of total nitrogen and phosphorus (Nhapi et al., 2011). The authors concluded that the Nyabugogo River system is heavily polluted and urgent action to control both rural and urban pollution is required. However, a river that is strongly affected is Sebeya River with high loads of sediments and high bacteria counts (Minirena-RNRA, 2015).

Like many other river catchments in developing countries, Sebeya catchment also lacks data on its water quality monitoring. However, a concern has been raised recently about its water quality status in terms of elevated levels of E. Coli, coliform bacteria and other pathogens from untreated sewage, high organic loads, high biological oxygen demand (BOD⁵) and chemical oxygen demand (COD), low dissolved oxygen (DO) concentrations, very high sediment loads and turbidity (W4GR, 2016). Therefore, it is of crucial importance to study and understand how current land use types and their practices have impacted the quality of water in the Sebeya catchment area. The findings could be used for proper informed planning and management decisions as well as promoting integrated water resource management in Rwanda. In order to understand the impacts of current land use on water quality of the Sebeya catchment area, we considered the seasonal variation of various water quality parameters across the different land use/land cover in the study area. Additionally, we examined the relationships between different land use/land cover based on the physiochemical and bacteriological composition of the water samples in the Sebeya catchment area across the wet and dry season, respectively.

Materials and Methods

Study area

Sebeya catchment is a part of the Congo-Kivu catchment positioned in the upper portion of the Congo basin. The catchment has a main river, the Sebeya River which runs48 km, flowing in a north-westerly path from its origin in the highlands of the Congo-Nile divide, at an elevation of 2,660 meters above the sea level, into the catchment outflow at Lake Kivu at an elevation of 1,470 meter above the sea level, in Rubavu town (W4GR, 2018). The catchment has other rivers contributing to the main river. Figure 2.1 shows the Sebeya catchment drainage network, elevation and sub-catchments

Due to its position, the Sebeya catchment is classified by the World Wildlife Fund for Nature (WWF) as ‘Albertine Rift Montane Forests Eco-region’; the eco-region is an area of unique faunal and moderate floral endemism; the region similarly supports the mountain gorilla (Gorilla beringei beringei), which is one of the most appealing gorilla species in Africa (W4GR, 2018).

Sebeya catchment is characterized by short dry season and long rainy season with high rainfall of 1200mm/year and above. The population in the catchment confirmed that within a period of 20 minutes to 3 hours after a heavy rain, floods occur; regions with an altitude higher than 2,000 meters above the sea level and an annual average temperature of around 17°C (W4GR, 2018). Flooding in the catchment naturally occurs in mid flat areas of the steep parts created by rift formation situated mostly in the flat area around Nyundo. Such resulting impact acts as a natural retention buffer for floods. Consequently, resulting in flash flood type which causes property and infrastructure damages. The Sebeya catchment is dominated by agriculture land use followed by forestry and grazing land uses.

Water quality data collection

Samples of water were collected using glass bottles of 0.75 land were stored at 40C before conducting laboratory analysis. Sampling frequency was set to be two times to cover both the rain (in April) and dry seasons (in July) in order to be able to assess all the changes that might occur due to seasonal variations. A total of 24 samples were collected consisting of 12 samples collected in the wet season and 12 samples in the dry season. The water quality parameters considered in the study were temperature, electrical conductivity, total suspended solids, turbidity, pH, total nitrogen, total phosphorus, dissolved oxygen, chemical oxygen demand, biological oxygen demand, and Escherichia coli. These parameters were chosen because they were anecdotally reported to be in high concentration in the Sebeya catchment and needed to be confirmed. IDEXX Quanti-Tray 2000 MPN (most probable number) table, incubation, spectrophotometry, and digestions methods were used to analyse E-coli, BOD⁵, COD, and TN/TP respectively.

Table 1: Pertinent information on sampling locations

Table 2.1 provides information on the sampling locations.

Figure 1: Sebeya catchment elevation, waterways, and sub-catchments. Source: W4GR, 2018

The map of Sebeya catchment area where sample points are collected is shown in Figure 1.

Method of Statistical analysis

The study used Principal Component Analysis (PCA), a statistical technique that uses an orthogonal transformation to convert observations of possibly correlated variables values of linearly uncorrelated variables called principal components. This helps streamline the number of contributing variables for the data analysis and interpretation. This transformation is defined in such a way that the first principal component has the largest possible variance (that is, it accounts for as much of the variability in the data as possible), and each succeeding component in turn has the highest variance possible under the constraint of orthogonality, that is, it is orthogonal to the preceding components. The resulting vectors are uncorrelated orthogonally.

PC’s may be defined in terms of the population (using ∑) or in terms of a sample (using S). Let

Where yj = a1j x1+a2j x2+:::+ap j xp are a sequence of “standardized” linear combinations (SLC’s) of the x’s such that for j ≠ k: i.e.a₁; a₂,…,a_p form an orthonormal set of p-vectors. Equivalently, the (p x p) matrix A formed from the columns {aj} satisfies A^TA = I_p (= AA^T); so by definition is an orthogonal matrix.

We choose a₁ to maximize,

, (1)

subject to . Then, we chose a₂ to maximize,

, (2)

subject to and , which ensures that y₂ will be uncorrelated with y₁. Subsequent PC’s are chosen as the SLC’s that have maximum variance subject to being uncorrelated with previous PC’s.

To find the first PC, we use the Lagrange multiplier technique for finding the maximum of a function f(x) subject to an equality constraint g (x) = 0. We define the Lagrangean function

, (4)

where λ is a Lagrange multiplier. We need a result on vector differentiation.

Result

Let x = (x₁, x₂ ... x_n) and . If b (n x 1) and A (n x n), symmetric, are given constant matrices, then,

1st PC

Differentiating (4) using the results, give

(5)

Showing that a₁ should be chosen to be an eigenvector of ∑; say a₁ = v with Eigen value λ. Suppose the Eigen values of ∑ are ranked in decreasing order λ₁≥λ₂≥λ₃≥,…,≥ λ_p> 0.

(6)

Therefore, in order to maximize Var (y₁), a₁ should be chosen as the eigenvector v1corresponding to the largest Eigen value λ₁ of ∑.

Results and Discussion

Results presentation of the measured water quality parameters

Table 2 and 3 show the laboratory analysis results of the water quality parameters from water samples collected at the 12 sample points in the wet season and in the dry season.

Table 2: Laboratory results of the analysed water quality parameters in the wet season

Note: S: Sample, WQP: Water Quality Parameter, DO: Dissolved Oxygen, TOC : Temperature, Turb: Turbidity, EC: Electrical Conductivity, TSS: Total Suspended Solids, TN: Total Nitrogen, TP: Total Phosphate, BOD: Biochemical Oxygen Demand, COD: Chemical Oxygen Demand, E-coli: Escherichia Coli, NA: Not Applicable. Standard: Rwanda Standard Board (RSB) 2008; WHO (2004) and Wyness et al. (2003)

Table 3: Laboratory results of the analysed water quality parameters in the dry season

Note: S: Sample, WQP: Water Quality Parameter, DO: Dissolved Oxygen, TOC : Temperature, Turb: Turbidity, EC: Electrical Conductivity, TSS: Total Suspended Solids, TN: Total Nitrogen, TP: Total Phosphate, BOD: Biochemical Oxygen Demand, COD: Chemical Oxygen Demand, E-coli: Escherichia Coli, NA: Not Applicable. Standard: Rwanda Standard Board (RSB) 2008; WHO (2004) and Wyness et al. (2003)

Descriptive Evaluation of the Variation water quality parameters at various sample points with the International Standard

In evaluating the level of DO variation in the Sebeya catchment across the different land use with respect to the standard values. The DO values across these catchment is expected to be greater than or equal to (the same as) the DO standard value. That is, the permissible limits for DO value of any water body. Our result shows that all the sampled points DO values were far greater or above the DO standard value with sample point 4, 8 & 2 respectively, having the most significant variation in Wet season among samples taken at Sebeya river headwater. As shown in Table 4 and Figure 2 below.

Figure 2: DO variation in the Sebeya Catchment with respect to Standard value

Table 4: DO (mg/l) variation from the Standard value

Key: VDW: Variation of Dissolved Oxygen from the standard value in wet season; VDD: Variation of Dissolved Oxygen from the standard value in the dry season.

In similar sense, the temperature level variation in the Sebeya catchment across the different land use with respect to the standard values is examined. The temperature values across this catchment are expected to be very close to the standard temperature value. That is, the permissible limits for temperature value for any water body. Our result shows that at all the sampled points, the temperature values fall below the standard temperature value. As shown in Table 5 and Figure 3 below.

Figure 3: Variation of Temperature in the Sebeya Catchment with respect to Standard value.

Table 5: T^oC variation from the Standard value.

Key: VTW: Variation of Temperature from the standard value in wet season; VTD: Variation of Temperature from the standard value in the dry season.

The variation of turbidity in the Sebeya catchment during wet and dry seasons ranged between 0.52 NTU and 3880 NTU. At all the sampled points, turbidity values recorded were far beyond the highest turbidity permissible limits of 5 NTU except at SP1, SP2 and SP3 which were the reference points and located in the livestock grazing land use. Largely, wet season turbidity values were higher compared to dry seasons due to erosion, deforestation, poor road construction and landslides of fragile hills. However, SP8 recorded the highest turbidity value in the dry season than in the wet season. This can be attributed to upstream mining activities, reduced river dilution and low discharge during dry season. Extremely turbid water is unhealthy for household use, is visually unappealing, can choke fish gills, can clog drip irrigation and water treatment equipment and is the reason of nasty taste and odours of surface water (Pullanikkatil et al, 2015). As shown in Table 6 and Figure 4 below.

Figure 4: Variation of Turbidity in the Sebeya catchment with respect to the standard value

Table 6: Turbidity variation from the standard value

Key: VTW: Variation of Turbidity from the standard value in wet season; VTD: Variation of Turbidity from the standard value in the dry season.

The variation pH in the Sebeya catchment ranged between 4.02 and 7.82. Mostly pH was high in the wet season than in the dry season. Most of the sampled points had pH values ranging between 6.5 and 8.5 which is the standard range for pH except SP1, SP2, SP3 and SP8 which had pH values lower than 6.5. These sites were located in forested areas and it was found that streams flowing through forested areas are usually acidic due to decomposition of soil organic matter which releases acids thus lowering pH (Hunchak – Kariouk and Nicholson, 2001; Coulter and Kolka et al., 2004; Kambwiri et al., 2014). On the other hand, SP8 was downstream of tea plantation which is known to grow well in acidic soils thus tea plantation may cause acidification of river especially through tea drainage which is practised in the Sebeya catchment (Kambwiri et al., 2014). The pH is an important variable in water quality assessment as it influences many biological and chemical processes within a water body. Low pH in streams may release toxic heavy metals such as Cd, Co, Cu, Hg, Ni, Pb and Zn which in return eliminate many types of aquatic life through influencing adversely the structure of macro- invertebrate community and species diversity (Kimmel et al.,1985; Abel, 2002). As shown in Figure 5 below.

Figure 5: Variation of pH in the Sebeya catchment with respect to the standard value

EC varied between 37.9 μS/cm -158.6 μS/cm and 37.8 μS/cm -173.4 μS/cm in the wet and dry season respectively. In both seasons, EC values were almost the same with a slight increase at SP6 and SP11 during wet and dry season respectively. However, all the sampled points had an electrical conductivity which were in acceptable range with respect to the Standard value (EC < 1000μS/cm). As shown in Table 7 and Figure 6 below.

Figure 6: Variation of Electrical conductivity in the Sebeya catchment with respect to the standard value

Table 7: Electrical conductivity variation from the standard value

Key: VEW: Variation of electrical conductivity from the standard value in wet season; VED: Variation of electrical conductivity from the standard value in the dry season.

TSS varied between 1 mg/l and 2455 mg/l during the wet and dry seasons in the Sebeya catchment. All the values recorded at sampled locations exceeded standard value of less than 30 mg/l except at SP1, SP2 and SP3 which were the headwaters with no to minimum anthropogenic activities. High TSS values were observed in the wet season than dry season. TSS in the rain season peaked at SP9; this sample was taken after a heavy rain which carried a lot of sediment from the highlands of Karambo sub-catchment. High sediment loads result in increased flood damage, reduced water body capacity via sedimentation, and a rise in water treatment costs (Skinner et al. 1997). As shown in Table 8 and Figure 7 below.

Figure 7: Variation of TSS in the Sebeya catchment with respect to the standard value.

Table 8: TSS variation from the standard value

Key: VTSSW: Variation of TSS from the standard value in wet season; VTSSD: Variation of TSS from the standard value in the dry season

Seasonal variation of total nitrogen in Sebeya catchment was evidently significant. TN varied between 0 mg/l and 51.4 mg/l. All recorded values in wet season were above the standard value of 3 mg/l. On the other hand, dry season recorded values were in acceptable range since 3 mg/l is the highest permissible limits. On one hand SP1, SP2 and SP3 which are headwaters where there is almost no human activities recorded TN values higher than the standard and they are located in grassland/grazing land use. This may be attributed to some vegetation in the grassland may be leguminous in nature and aid in fixing atmospheric nitrogen into the soil which later gets denitrified to inorganic nitrates and with runoff action end up into the river (Kambwiri et al., 2014). On the other hand, the high variation of TN from the standard value at SP6, SP8, SP9 SP11 and SP12 in the wet season can be attributed to the decomposition of livestock wastes, human wastes, plant decomposition and runoff of fertilizers used in agricultural lands as well as the discharge of municipal waste into rivers through runoff. High nitrogen level can encourage rapid growth of algae and other aquatic plants which may lead to eutrophication of rivers. Besides, excessive growth of aquatic organisms, results in water intakes clogging, reduced dissolved oxygen, and poor light penetration to deeper waters (Chambers et al., 2001). As shown in Table 9 and Figure 8 below.

Figure 8: Variation of TN in the Sebeya catchment with respect to the standard value.

Table 9: TN variation from the standard value

Key: VTNW: Variation of TN from the standard value in wet season; VTND: Variation of TN from the standard value in the dry season

Total phosphorus varied between 0.29 mg/l and 3.7 mg/l. Although the high levels of total phosphorus were observed in the wet season than dry season; all the sampled locations were below the highest acceptance value (standard value) of 5 mg/l. As shown in Table 10 and Figure 9 below.

Figure 9: Variation of TP in the Sebeya catchment with respect to the standard value

Table 10: TP variation from the standard value

Key: VTPW: Variation of TP from the standard value in wet season; VTPD: Variation of TP from the standard value in the dry season.

BOD⁵ varied between 8.4 mg/l and 119.7 mg/l. While the recorded values of BOD⁵ in wet season were below the standard value; the BOD⁵ recorded values in dry season were above the standard value. High BOD⁵ at SP9 in the dry season may be caused by sewage discharge, animal waste and industrial effluents discharge into the river. Beside SP9 is located downstream of an urban area (Mahoko city) which is characterized by poor solid and liquid waste disposal leading to inadequate effluent/sewage discharge and poor disposal of animal waste. High BOD⁵ reduces the amount of dissolved oxygen. The decrease in dissolved oxygen in aquatic ecosystems may have adverse effects on many aquatic organisms such as Ephemeroptera (mayflies), Trichoptera (caddisflies), and Plecoptera (stoneflies) which respire with gills or by direct cuticular exchange drop and may be entirely eliminated with oxygen depletion (Abel, 2002). As shown in Table 11 and Figure 10 below.

Figure 10: Variation of BOD in the Sebeya catchment with respect to the standard value

Table 11: BOD variation from the standard value

Key: VBODW: Variation of BOD from the standard value in wet season; VBODD: Variation of BOD from the standard value in the dry season.

COD varied between 6.5 mg/l and 402 mg/l during the wet and the dry seasons. COD recorded values were higher than the standard value of 50 mg/l in both seasons except at SP1, SP2, SP3, SP5 and SP10. High variation from the standard value were observed in the dry season peaking at 402 mg/l at SP9. The reason for this is that the same sample point was recorded to have a high BOD which implies organic matter pollution of the river. COD parameter is related to organic matter and it can be described as the oxygen amount that is required to oxidise all organic matter prone to oxidation by a strong chemical agent such as dichromate (Du Plessis 2014). Same as BOD⁵, high COD leads to dissolved oxygen depletion and affect aquatic life and diversity. As shown in Table 12 and Figure 11 below.

Figure 11: Variation of COD in the Sebeya catchment with respect to the standard value.c

Table 12: COD variation from the standard value

Key: VCODW: Variation of COD from the standard value in wet season; VCODD: Variation of COD from the standard value in the dry season.

E. Coli varied between 0 MPN/100ml and 20630 MPN/100ml. According to the RSB standard, surface water should not have E.coli more than 4 MPN/100ml. All sampled points had E.coli values far beyond the standard value in both rain and dry season. Dry season E.coli values increased remarkably at downstream sites (SP10, SP11 and SP12). This may be attributed to direct discharge of raw or partly untreated sewage from households or industries since they are close to the rivers. The rain season concentrations were very high and that can be related to recent increased run-off in the rain season from agricultural land with manure fertilizers; it can also be caused by leaking septic tank or inappropriate disposal of animal wastes in the urban areas. E.coli peaked at SP6 with 20630 MPN/100ml in the wet season. This site was located downstream of a livestock grazing land where livestock dung may be washed by rainfall; the site was also surrounded by scattered households without proper sewage disposal and human wastes were seen in most of the small routes heading to the sample points. This showed how open defecation around SP6 was prevalent. E.coli recorded values at SP9 and SP11 were also high during the wet season. These sites were located in the centre and downstream of Mahoko city respectively where the Sebeya River has been flooding washing all urban waste ranging from raw sewage to leaked waste from septic tanks. The increase in E.coli during the wet season showed diffuse pollution from runoff from the catchment and points to inadequacies in land management, poor sanitation and waste management in the catchment area. So intense rainfall, runoff and soil erosion carrying manure applied in agricultural lands led to the recorded elevated values of E.coli. It was found that untreated slurry and faeces of grazing animals can convey a variety of bacterial and protozoan pathogens (Hooda et al., 2000). Hooda et al. (2000) further discussed that faecal contamination has been reported in streams draining dairy farms, subsurface runoff from manure applied fields and surface runoff from grazed grasslands. The presence E.coli indicates fresh faecal contamination and may be indicators of disease causing organisms that cause diseases such as intestinal infections, dysentery, hepatitis, typhoid fever, cholera and other illnesses, thus making the water unfit for drinking (Pullanikkatil et al, 2015). As shown in Table 13 and Figure 12 below.

Figure 12: Variation of E. Coli in the Sebeya catchment with respect to the standard value

Table 13: E. Coli variation from the standard value

Key: VELW: Variation of E. Coli from the standard value in wet season; VELD: Variation of E. Coli from the standard value in the dry season.

Variation of the water quality parameters across season.

We conducted the principal component analysis (PCA) to identify the water quality parameters mostly associated with various land use area surrounding the Sebeya water catchment area. Each sample points have mixed land use characteristics as shown in Table (1).

We consider the concentration of the water quality parameters across these sample points during the dry and wet season. This helps us understand the variability of these water quality parameters across the various sample points with respect to the two seasons (Dry & Wet). This was carried out to better assess the concentration of these water quality parameters across season with respect to sample points.

We also considered in the PCA seasonal variation of the parameters concentration across the sample points under consideration. The analysis was then classified into two major categories (Dry and Wet season).

Dry Season: During the dry season, the PCA successfully extracted three components which accounted for 75% total variation (Figure 13). Figure 13 shows the visual extraction of the components with their various Eigen values. Based on Eigen value of 1 and above, three components were extracted.

We subjected the components loading and parameters identification (in respect of most concentrated parameters at various land use) on land use on the first two (2) extracted components which accounted for 62% of the total variation. As shown in Figure 13 and Table 14.

Figure 13: The visual component extraction

Table 14: Important Components extracted

Basically, this led to the plot (Figure 13) which shows the visual concentration of the parameters across each sample points (land use area) around the Sebeya catchment. We however present the detailed loadings of each water quality parameters with respect to the first 6 components extracted. See Table 14 for the statistics.

Meanwhile, considering the visual plot (Figure 13), water quality parameters under concentration are found to be mostly concentrated at sample points 12,9, 4, 8, 10 and 11 (see Table (1) for land use characteristics) based on components 1 and 2 respectively.

Figure 14: Concentration of water quality parameter based on the first two extracted components.

In order to measure the loading power (the percentage accounted for by each parameter to the components loading) of each water quality parameters, we consider Table 3.14. The most pronounced water quality parameters (using 30% and above) with respect to the first component are TSS, BOD, Temperature, Turbidity, TN, and COD respectively.

Table 15: Parameters loadings

Wet Season: During the wet season, the PCA successfully extracted three components which accounted for 83% total variation (Figure 14). Figure 14 shows the visual extraction of the components with their various Eigen values. Based on Eigen value of 1 and above, three components were extracted.

We subjected the components loading and parameters identification (in respect of most concentrated parameters at various land use) on land use on the first two (2) extracted components which accounted for 70% of the total variation. As shown in Figure 15 and Table 16.

Figure 15: The visual component extraction

Table 16: Important Components extracted

Basically, this led to the plot (Figure 14) which shows the visual concentration of the parameters across each sample points (land use area) around the Sebeya catchment. We however present the detailed loadings of each water quality parameters with respect to the first three (3) components extracted. See Table 3.16 for the statistics.

Meanwhile, considering the visual plot (Figure 15), water quality parameters under concentration are found to be mostly concentrated at sample points 11, 6, 12, 9, 8 and 4 (see Table (1) for land use characteristics) based on components 1 and 2 respectively.

Figure 16: Concentration of water quality parameter based on the first two extracted components.

In order to measure the loading power (the percentage accounted for by each parameter to the components loading) of each water quality parameters, we consider Table 3.16 the most pronounced water quality parameters (using 30% and above) with respect to the first component are TSS, Turbidity, EC, TP, COD and E-coli respectively.

Table 17: Parameters loadings

As discussed, our result show that some of the water quality parameters occur in high concentration at the sample points 4, 6, 8, 9, 11 and 12 in both seasons. As shown, turbidity, temperature, TSS, TN, BOD⁵ and COD are more pronounced at the identified sample points during the dry season. For example, high turbidity and TSS at S8 can be attributed to intense mining activities that occur upstream and use the Sebeya River for sieving minerals such as coltan. So low dilution of the Sebeya River in the dry season coupled with those unsustainable mining practices lead to high turbidity and sedimentation of the Sebeya River.

In Wet season, turbidity, EC, TSS, TP, COD and E-coli are more concentrated at the indicated sample points. This can be explained by increased soil erosion and agriculture runoff coupled with landslides which have been happening in the study area due to prolonged intense rainfall. Thus, turbidity, TSS, and COD concentration levels remain very significant at both seasons at the specified sample points.

Considering the characteristics or features of the land use area under which each of the identified sample points is taken, majority of the sample points constitute settlements, mining sites, forestry, croplands and tea plantation. See figure 17 below.

Figure 17: Visual Description of the relevant sample Locations across mixed land use types

The result reveals that these water quality parameters (leading to higher water pollution) is majorly associated to land use area with human settlements and forestry. Our result corroborate literature reports which stated that, settlements land use area has poor waste management which result in direct discharge of sewage into the river; mining and deforestation activities being done in forestry area lead to high turbidity and sedimentation of the Sebeya river and its tributaries (W4GR, 2018). Our results also establish that, land use areas such as cropland and tea plantations also contribute significantly to high concentration of the identified water quality parameters (causing high water pollution).

Conclusion

This study examined the impacts of current land use on water quality of the Sebeya catchment area. This assessment was made possible through the collection and analysis of water quality samples from the Sebeya River and some of its tributaries during the wet and the dry seasons. Most of the water samples taken from the Sebeya River and its major tributaries had high concentrations of the measured parameters which exceeded standard values. Some parameters varied strongly from the standard value more in the wet season than in the dry season.

Most of the measured water quality parameters were concentrated on sample points that are in settled areas around the Sebeya catchment. Waste management, of both liquid and solid in the urban areas of the Sebeya catchment area and periodic water quality monitoring are recommended to assess the level of pollution. Furthermore, all users and relevant stakeholders should take an active role in the conservation of the Sebeya River and its major tributaries in order to reduce and avoid further degradation of the catchment through different land uses.

Acknowledgements: We thank Water for Growth Rwanda and Rwanda Water and Forestry Authority who supported financially the entire field visits and laboratory analysis of water samples that were collected to analyse the impacts of land use on water quality in the Sebeya catchment area, Rwanda.

References

• Abel, P.D. Water pollution biology. Second edition. (2002) Taylor and Francis Ltd. 286pp.

  PubMed│CrossRef│Others

• Bagalwa, M. The impact of land use on water quality of the Lwiro River, Democratic Republic of Congo, Central Africa. (2006) African J Aquatic Sci 31(1): 137-143.

  PubMed│CrossRef│Others

• Chambers, P.A., Guy, M., Roberts, E.E., et al. 2001. Nutrients and their impact on the Canadian environment: Agriculture and agri-food Canada, environment Canada, fisheries and oceans Canada, health Canada and natural resources Canada. 241 pp.

  PubMed│CrossRef│Others

• Christian, D.S., Valentine, M.U., Mardoche, M.B., et al. Water quality monitoring in Rwanda. (2012) Butare, 54pp.

  PubMed│CrossRef│Others

• Coulter, C.B., Kolka, R., Thompson, A. Water quality in agricultural, urban and mixed land use watersheds. (2004) J Am Water Res Assoc JAWRA 40(6): 1593-1601.

  PubMed│CrossRef│Others

• Ding, J., Jiang, Y., Fu, L., et al. Impacts of land use on surface water quality in a subtropical River basin: A case study of the Dongjiang River basin, South-eastern China. (2015) Water 7: 4427-4445.

  PubMed│CrossRef│Others

• Du Plessis, A., Harmse, T., Fethi Ahmed, F. Quantifying and predicting the water quality associated with land cover change: A case study of the Blesbok Spruit catchment, South Africa. (2014) J Water 6: 2946-2968.

  PubMed│CrossRef│Others

• Foley, J.A., Defries, R., Asner, G.P. Global consequences of land use. (2005) Science 309: 570-574.

  PubMed│CrossRef│Others

• Hartman, G. Effects of forest management activities on watershed processes. (2004) Fishes and Forestry-Worldwide Watershed Interactions and Management: 271-302.

  PubMed│CrossRef│Others

• Hooda, P.S., Edwards, A.C., Anderson, H.A., et al. A review of water quality concerns in livestock farming areas. (2000) Sci Total Environ 250: 143-167.

  PubMed│CrossRef│Others

• Huang, J., Zhan, J., Yan, H., et al. Evaluation of the impacts of land use on water quality: A case study in The Chaohu lake basin. (2013) Scientific World J 01-08.

  PubMed│CrossRef│Others

• Hunchak, K., Nicholson, R.S. Watershed contributions to nutrient and other non-point source contaminants to the Barnegat Bay Little Egg Harbour Estuary. (2001) J Coastal Res 32: 28-31.

  PubMed│CrossRef│Others

• Kambwiri, A.M., Changadeya, W., Chimphamba, J., et al. Land use impacts on water quality of Rivers draining from Mulanje Mountain: A case of Ruo River in the southern Malawi. (2014) Malawi J Sci Technol 10(1): 15-31.

  PubMed│CrossRef│Others

• Kimbadi, S., Vandelannoote, A., Deelstra, H., et al. Chemical composition of the small Rivers of the north-western part of Lake Tanganyika. (1999) Hydrobiologia 407: 75-80.

  PubMed│CrossRef│Others

• Kimmel, W.G., Murphey, J.D., Sharpe, W.E., et al. Macro invertebrate community structure and detritus processing rates in two south-western Pennsylvania streams acidified by atmospheric deposition. (1985) Hydrobiologia 124(2): 97-102.

  PubMed│CrossRef│Others

• Lamek, N., Lanhai, L., Alphonse, K., et al. Agricultural impact on environment and counter measures in Rwanda. (2016) African J Agricult Res 11(25): 2205-2212.

  PubMed│CrossRef│Others

• Malone, S.J. Agricultural nonpoint source pollution management: Water quality impacts of Balm road treatment marsh, Hillsborough County, Florida. (2009) 1-108.

  PubMed│CrossRef│Others

• Minirena-RNRA. Rwanda national water resources master plan. (2015) Kigali, 253pp.

  PubMed│CrossRef│Others

• Ngoye, E., Machiwa, J.F. The influence of land-use patterns in the Ruvu River watershed on water quality in the River system. (2004) Physics and Chemistry of the Earth 29(15-18): 1161-1166.

  PubMed│CrossRef│Others

• Nhapi, I., Wali, U.G., Uwonkunda, B.K., et al. Assessment of water pollution levels in the Nyabugogo catchment, Rwanda. (2011) Open Environ Engg J 4: 40-53.

  PubMed│CrossRef│Others

• Pullanikkatil, D., Palamuleni, L.G., Ruhiiga, T.M. Impact of land use on water quality in the Likangala catchment, southern Malawi. (2015) African Journal of Aquatic Science 40(3): 277-286.

  PubMed│CrossRef│Others

• Rwanda Natural Resources Authority 2011-2012. Water quality monitoring in Rwanda. Report I, II, III.

  PubMed│CrossRef│Others

• Skinner, J.A., Lewis, K.A., Bardon, K.S., et al. An overview of the environmental impact of agriculture in the UK. (1997) J Environ Manage 50(2): 111-128.

  PubMed│CrossRef│Others

• Sliva, L., Williams, D.D. Buffer zone versus whole catchment approaches to studying land use impact on River water quality. (2001) Water Res 35(14): 3462-3472.

  PubMed│CrossRef│Others

• Tong, S.T., Chen, W. Modelling the relationship between land use and surface water quality. (2002) J Environ Manage 66: 377-393.

  PubMed│CrossRef│Others

• Uwimana, A., Dam, A.V., Gettel, G., et al. Effects of River discharge and land use and land cover (LULC) on water quality dynamics in Migina catchment, Rwanda. (2016) Environ Manage 60 (3): 496-512.

  PubMed│CrossRef│Others

• Wan, R., Cai, S., Li, H., et al. Inferring land use and land cover impact on stream water quality using a Bayesian hierarchical modelling approach in the Xitiaoxi River watershed, China. (2014) J Environ Manage 133: 1-11.

  PubMed│CrossRef│Others

• Water for growth Rwanda. Catchment plan Sebeya - Characterisation and vision. (2016) Kigali, 52pp.

  PubMed│CrossRef│Others

• Water for Growth Rwanda. Integrated water resources management programme: Catchment Plan Sebeya. 2017-2023. Version 2.0. (2018)

  PubMed│CrossRef│Others

• Zhao, J., Lin, L., Yang, K., et al. Influences of land use on water quality in a reticular river network area: A case study in Shanghai, China. (2015) Landscape Urban Planning 137: 20-29.

  PubMed│CrossRef│Others