Effects of Sand-Harvesting on River Water Quality and Riparian Soil Physico-Chemical Properties ()
1. Introduction
Sand is mined, using various technologies, mainly for construction industry from sources on land and under water [1], especially along the rivers on the riparian lands and the roadsides. Sand-harvesting occurs on small and large- scales, depending on the tools used [2] and the purpose for the harvesting. Small-scale harvesting normally involves simple tools like spades, wheelbarrows, and hoes, while large-scale operations involve machinery [3]. The persons harvesting the sand wade into the rivers with spades and containers, scoop the sand from the riverbeds and the banks into their containers and then carry the sand offshore in heaps ready for selling.
There is a growing demand for sand sourced from rivers [4] particularly in developing countries where rapid socio-economic development causes the construction industry to grow strongly [5]. A report from the United Nation Nations Environment Program [6] estimated that between 32 and 50 billion tons of river sand and gravel are harvested annually worldwide, making the sand mining sector a key contributor to Gross Domestic Product (GDP) [7] across the globe. Kenya, under the social pillar of its development blueprint, Vision 2030, [8] seeks to enhance the mining sector’s contribution to its GDP [9]. In 2015, the mining sector’s share of Kenya’s GDP was 0.8%. But the Kenya government aims to increase this to 10% by the year 2030 through its guiding mining and minerals policies, [10], including [11] and [12]. The legislation and guidelines are geared towards sustainable mining, including low-value extractives like river sand [9], murram and unprocessed gravel among others.
Indiscriminate sand and gravel extraction has placed immense pressure on the environment, especially major rivers, threatening the health of riverine ecosystems [2] [13] with the greatest damage generally more pronounced in small river catchments. The widespread distribution of river sand-harvesting activities continues to degrade river water quality and the surrounding riverine environment. [14] and [5] reported damaged, contamination or drying-up of waterbodies that provide water to communities near mining activities. Sand-harvesting also modifies the physico-chemical composition of river water by influencing chemical parameters including turbidity, TSS, magnesium and iron [15], posing risks to aquatic and human life [16]. Few studies, available publicly, quantify the physical alterations that accompany sand-harvesting and how they are linked to ecological impacts [17].
In Kenya, there is limited information on sand-harvesting activities, outside the sand rich Makueni and lower eastern Counties with limited enforcement of existing policy frameworks [18]. This has resulted in widespread, unregulated, and outright illegal river sand-harvesting activities to meet the construction industry’s high demands for sand. Sand-harvesting occurs extensively in Kakamega County, in the Western part of Kenya. But there is insufficient data on its effects on water quality and the surrounding environment. [19] outlines various mining activities that yield, for instance, about 278,000 tons of sand annually. Such mining is a livelihood support activity for more than 80,000 people in the region and has had negative impacts on the health of rivers like the Shiatsala, Yala and Isiukhu, the largest river network in the region.
The widespread distribution of river sand-harvesting activities continues to degrade river water quality and the surrounding riverine environments [6]. This paper contributes to knowledge by documenting how unsustainable sand harvesting practices pose serious environmental problems to river ecosystems. The resultant impacts from river sand harvesting are a cause for concern even by the UN [6] [17] [20] [21]. This paper also discusses the impact of sand harvesting on river water quality and riparian areas in the given region of study. No such study has ever been undertaken in this particular region. The methodology applied is also new, and will enrich such studies in future. The findings will also be useful to policy makers in water resources management.
It is against this background that this study examined the effects of river sand-harvesting on riverine ecosystems along the two rivers with extensive sand-mining activities in Kakamega County. Emphasis was placed on the effects of the sand-harvesting on water quality and the riparian soil’s physico-chemical properties. The study was carried out on the Shiatsala and Lusumu rivers within Kakamega County, during the dry period December 2020 and wet period June 2020.
2. Materials and Methods
2.1. Study Area
Kakamega County is situated in the western part of Kenya and lies between latitudes 0˚07'0"N and 0˚16'30"N, and longitudes 34˚37'30"E and 34˚49'0"E, an approximate 400 km northwest of Nairobi City. It covers 3034 km2 with Kakamega Town as its administrative headquarters, and its altitude is ranges from 1240 and 2000 m. Several large rivers traverse the county, including Yala, Isiukhu, Lusumu, and Nzoia (Figure 1).
Mean annual precipitation is 1280 mm and rainfall is bi-modally distributed, with the short rains occurring during October to December and long rains from March to May [20] [21]. These established seasons are slowly shifting forward in the recent decades. The short rainy season is characterized by less intense downpours and daily variability, while the long rains is characterized by heavy downpours almost daily. The average rainfall is 500 to 800 mm (short rainy season) and 1000 to 1200 mm (long rain season) [21]. Notably, Kakamega Forest plays a significant role, modulating the environment of the surrounding areas, including precipitation [22]. The dry season runs from December to February. The temperature varies annually with an average temperature range of 10.6˚C to 27.7˚C [23].
2.2. Sampling Site Selection
Eight experimental and two control sites were sampled purposively. The
![]()
Figure 1. Sampling points on rivers traversing Kakamega County. Major rivers are named in red, sampling points labelled in black.
sand-harvesting sites were chosen using reconnaissance data that identified sand-harvesting hotspots along the two rivers. Four sand-harvesting sampling sites were established on each river, between 2 and 15 km apart as shown in Figure 1. The four sites on the Shiatsala River were at Kwa Thomas, Shikunga, Lumanyasi and Shikoti, while those on the Lusumu River were at Mwera, Ndombi, Shikutse and Lwakhupa. The control sites were at Shamberere on the Shiatsala River and Lusumu B on the Lusumu River, where there was no evidence of sand-harvesting.
2.3. Data Collection
Data on eight parameters was collected as shown in Table 1.
2.4. Laboratory Analysis
The parameters were analyzed as tabulated in Table 2.
![]()
Table 2. Methods used during laboratory analysis.
2.5. Sampling Site Selection
The water quality and soil physico-chemical parameters for the sampling sites were compared between periods/seasons. Significant differences between periods/seasons were determined using the paired-sample T-test at 5% confidence level. Significant differences between the control and sand-harvesting sites were investigated using the independent sample T-test at 5% confidence level. One Way ANOVA was used to test for differences between the control and sand-harvesting sampling sites, and significant differences between the sampling sites were investigated with Tukey’s Honestly Significant Difference (HSD) post-hoc tests at 5% confidence level. Statistical analysis was done using Statistical Package for the Social Sciences (SPSS) Version 22.
3. Results and Discussions
3.1. Effects of Sand Harvesting on Turbidity
Turbidity was high and fluctuated during the rainy season compared to the dry season. Figure 3 shows the analysed results. Notably, the sites where sand- harvesting was extensive—Lumanyasi and Shikoti on Shiastala river, and Lwakhupa on the Lusumu river—reported the highest turbidity levels during the rainy season. Turbidity levels increased significantly at all sampled sites during the rainy season (t(9) = −2.679, p = 0.025). This agrees with reports by [31] that rivers tend to have higher turbidity during the rainy season and sand-harvesting influences turbidity among other properties. Similarly, [32] in a study in India on River Periyar, concluded that relatively larger amounts of sand were harvested in the middle parts, raising riverine turbidity levels. Also, [33] attributed high turbidity levels in the reservoir he was studying to settling and resuspension of solids. Hence, this agrees with findings of this study that river sand-harvesting impacts water quality through re-suspension of sediments in the river resulting in temporary increases in turbidity.
However, at sampling sites 1 (Mwera), 2 (Ndombi bridge) and 3 (Shikutse) of Lusumu river, the turbidity levels were almost similar both during the rainy and dry season. This could be because the three sites had ongoing sand harvesting
![]()
Figure 2. 5 m × 5 m zig zag soil sampling quadrat.
![]()
Figure 3. Turbidity in the rainy and dry seasons.
activities during the reconnaissance study period of December 2019, but the sites had been abandoned during the data collection period of 2020. Hence, sand harvesting was not taking place at the time of field visit, both in the rainy and dry season. In addition, it was observed that sampling sites 1 (Mwera), and 2 (Ndombi bridge) were being rehabilitated through planting of trees by members of the Lusumu Water Resources Users Association (WRUA).
3.2. Effects on Total Suspended Solids
The results for TSS rainy during dry season are shown in Table 3 and Figure 4. Sampling sites 3 (Lumanyasi) and 4 (Shikoti) on Shiastala River, and sampling site 4 (Lwakhupa) on Lusumu River reported the highest TSS values that averaged 125 mg·L−1 that agrees with [31] and [34] [35] observation that high TSS levels are associated with activities of sand-harvesting. In addition, high demand for sand with the presence of reliable access road was noted to increase the frequency of sand-harvesting, thus increasing the level of TSS.
Logging activities upstream [34], natural runoff which allow more silt and clay to flow into the river [36] [37] and reduced river channel water levels [33] [35] can also contribute to high TSS level.
3.3. Effect on Soil Moisture Content
The rainy and dry season soil moisture content results are shown in Figure 5. Soil moisture content decreased at the sand-harvesting sampling sites from 22.71% ± 3.61% during the rainy season to 17.56% ± 7.10% during the dry season. A comparison of soil moisture content for the rainy and dry season using paired sample T-test analysis revealed a significant increase in moisture content at all sampling sites during the rainy season (t(9) = −2.566, p = 0.030) from 18.14 ± 6.66 m to 22.3% ± 3.44%; an increase of 4.21% ± 5.18%. Results of the independent sample T-test revealed a statistically insignificant difference in soil moisture content between the control sites and the sand-harvesting sampling sites during the rainy (t(8) = −0.642, p = 0.539) and dry seasons (t(8) = 0.526, p = 0.613). [38]
![]()
Figure 4. Rainy and dry season TSS concentrations.
![]()
Figure 5. Site soil moisture content in %.
![]()
Table 3. Wet and dry season TSS results.
reports that soil disturbance can affect soil moisture content. Also, although [38] found that the number of soil particles that are water-stable correlates negatively with sand content in the soil, this study did not find direct relationship between sand harvesting and soil moisture content.
3.4. Effects on Total Nitrogen
The rainy and dry season soil Total Nitrogen results are shown in Table 4 and Table 5. At the two control sites, the soil total nitrogen content decreased from
![]()
Table 4. Mean values for the soil physico-chemical parameters between dry and rainy season.
![]()
Table 5. Physico-chemical soil parameter values during the dry and rainy seasons.
0.400 ± 0.000 mg·L−1 during the rainy season to 0.205 ± 0.007 mg·L−1 during the dry season. At sampling sites 4 (Shikoti and Lwakhupa) where sand harvesting was reportedly very intense, the total nitrogen content increased from 0.23 mg·L−1 during the rainy season to 0.28 mg·L−1 during the dry season and 0.25 mg·L−1 during the rainy season to 0.26 mg·L−1 during the dry season respectively, while it decreased in the other sand harvesting sites. However, despite the above difference, results of the independent sample T-test revealed no statistically significant difference in total nitrogen content between the control and sand-harvesting sites during either the rainy (t(8) = 0.306, p = 0.767) or dry seasons (t(8) = −2.122, p = 0.067).
The primary impacts of river sand harvesting are the direct removal of vegetation, which alters the rates of nitrogen cycling hence the productivity of the ecosystem [17]. However, though [38] found out that soil disturbance can affect soil physico-chemical properties, this study did not find direct relationship between sand harvesting activities and soil TN.
3.5. Effects on Phosphorus, pH and Organic Carbon
The rainy and dry season soil phosphorus, pH and OC content results are shown in Table 4 and Table 5. Results of the independent sample T-test revealed a statistically insignificant difference in phosphorus content between the control sites and the sand-harvesting sites during both the rainy (t(8) = −1.140, p = 0.287) and dry seasons (t(8) = −1.489, p = 0.175); no statistically significant difference was found between the soil pH of the control and sand-harvesting sampling sites, in either the rainy or dry seasons; and a statistically insignificant difference of soil organic carbon content between the control sites and the sand-harvesting sites during the rainy (t(8) = 1.114, p = 0.298) and dry seasons (t(8) = 0.955, p = 0.511). This is an indication that sand-harvesting does not affect riparian soil P content, pH, OC levels and textural class.
4. Conclusion
River sand-harvesting impacts water quality through re-suspension of sediments in the river resulting in temporary increases in turbidity. This study finds a significant relationship between sand harvesting activities, and river water turbidity and level of TSS. Though other studies found out that soil disturbance can affect soil physico-chemical properties, this study did not find a direct relationship between sand harvesting activities and soil moisture content, TN, P, pH, OC and textural class. Also, the intensely harvested sites showed an increase in soil total N during the dry season, but it decreased at all other sites. The study, therefore, concludes that sand-harvesting does not directly affect the riparian soil moisture content, TN, P, pH, OC or textural class, but reduces productivity of riparian land and puts the riverine ecosystems at risk. Ultimately, there is a need for proper sand mining practices, upscaling of conservation efforts and creation of awareness on the need to sustainably carry out sand harvesting activities.
Acknowledgements
A.I. L. thanks Kakamega County Department of Water, Environment, Natural Resources, and Climate Change. The authors appreciate their institutions support and especially Mr. Emmanuel Mzungu, Mr. Richard Wepukhulu, Dr. Rose Fukwo, Ms. Judith Kavala, Mr. Dennis Oginga and Mr. Stephen Panyako for their assistance during data collection and analyses of the samples collected.