Prolonged heavy rainfall and land use drive catchment sediment source dynamics: Appraisal using multiple biotracers

• Sediment source apportionment was evaluated in the Nyando and Sondu-Miriu basins. • Soil unmixing highlights the impact of land degradation on source contributions. • Conservativeness and consensus ranking methods identified the optimum tracers. • These results benefit policymakers developing integrated catchment management plans. Accelerated soil erosion is a major cause of land degradation in East Africa’s agricultural and pastoral landscapes with severe consequences for food, water and livelihood security. In this study, we aimed to provide a tool to support the sustainable management of land and water resources in a region significantly impacted by land degradation. We employed source apportionment methods to quantify the relative contribution of sediment sources within the Nyando and Sondu-Miriu River basins and their subcatchments in the Winam Gulf, Kenya. A total of 237 riverbed sediment samples and 76 composite surface soil samples were collected from the Nyando and Sondu-Miriu River basins. The total elemental concentrations of these samples, determined using ICP-MS/MS, were utilised as geochemical tracer properties. Conservativeness index, consensus ranking and consistent tracer selection methods were then used to identify the optimum unmixing tracers before applying the frequentist unmixing model FingerPro to determine sediment provenance. Sediment source analysis revealed that the Ainamutua and Nyando-Kipchorian subcatchments, areas predominantly affected by land degradation activities such as poor crop management practices and deforestation on steep slopes, contributed 39 ± 4 % and 44 ± 4 %, respectively. In contrast, the Awach Kano and Nyaidho subcatchment, with a higher proportion of tree-cover and lower soil erosion rates, only contributed 17 ± 7 %. In the Sondu-Miriu, the Yurith and Kipsonoi subcatchments contributed 68 ± 5 % and 20 ± 6 %, respectively, due to the predominance of forest encroachment and ridges in the Yurith subcatchment. Additional fingerprinting analysis within each of the Nyando and Sondu-Miriu basins reveals the significance of land use, landform and soil types on source contributions. Quantifying sediment source contributions within large river basins provides essential information for environmental managers and policymakers developing integrated catchment management plans. The results from this study can be used to implement sustainable land use policy focused on soil restoration in the Lake Victoria drainage basin.

<p>Information on the share of river bank erosion to the total sediment load at catchment scale by using the fingerprinting approach is important to address our knowledge of erosion processes to better target soil erosion control measures. In particular, river bank erosion is affected by many factors such as spatial and temporal variables and is difficult to quantify the relationship of the share of bank erosion to catchment size and upland erosion rate without extensive fieldwork and data analysis. Potential tracers including geochemical, fallout radionuclides, bulk and compound-specific stable isotopes, and magnetic properties have been used, often in combination with sediment source apportionment. In this worldwide review, the global dataset for percent share of river bank and surface erosion using fingerprinting approach was collected to establish the significance of catchment size and other physical controls on river bank erosion. Google Scholar and Web of Science were used to review research articles that included river bank/subsurface as one of the sediment sources in the study areas. This database showed that the UK (n = 84), USA (n = 14) and Brazil (n = 10) had the highest number of catchments, followed by Iran (n = 4), Southern Zambia (n = 1), Australia (n = 1), Spain (n = 1), Mongolia (n = 1) and Burkina Faso (n = 1) ranging in size from 0.31 to 15000 km<sup>2</sup>, predominately agriculture. Based on published studies, there is a clear shift of sediment sources from surface erosion to river bank erosion with increasing catchment size. The results show the wide range of relative contributions of surface and river bank sources to the catchment sediment yield around the globe. There are a number of catchments with river bank contribution exceeding 25% and surface contribution exceeding 90% of total sediment loss. This diversity highlights the many factors that influence river bank erosion. In addition to the wide range, sediment source contribution in the range 1-25% from river bank is generally representative around the World. We recommend that long term monitoring of sediment load and surface and river bank sources at nested sites within a catchment are indispensable. Furthermore, limited information on the share of sources often makes it difficult to target mitigation measures reducing sediment loads at the catchment scale.</p><p><strong>Keywords: </strong>Sediment load, catchment size, fingerprinting approach, river bank share</p>

Introduction to the special issue ‘Tracer Applications in Sediment Research’

Applying geochemical and colour properties to quantify sediment sources in a Brazilian semiarid ephemeral river system

Knowledge on sediment behavior in watersheds contributes to properly implement soil and water conservation practices and solve sediment‐related environmental problems. We examined the effects of suspended and bedload sediments and location of potential sediment sources. We used geochemical properties to calculate sediment contributions, with sediment fingerprinting approach, in a watershed (0.98 km2) and in a sub‐watershed (0.39 km2) cropped to commercial eucalyptus along with native forest. The sediment sources evaluated were eucalyptus stand, stream bank along the main channel, and unpaved roads in each watershed. Two size‐fractions of source and sediment samples including fine (<0.063 mm) and coarse (0.063–2 mm) particles were analyzed. Particle size and source location in the watershed are major factors affecting the contribution of sediment sources. The closer the sediment source, the higher the sediment source contribution, especially for coarse particles. Runoff connectivity probably explains large changes in sediment source contributions over very short distances between sampling sites. The stream bank had the greater relative contribution to sediment yield, suggesting that the focal point for erosion control practices should be the channel rather than on the slopes with eucalyptus or riparian vegetation. Future studies should consider uncertainties in grouping sediment sources with a much better signal‐to‐noise ratio.

Improving the quantification of sediment source contributions using different mathematical models and spectral preprocessing techniques for individual or combined spectra of ultraviolet–visible, near- and middle-infrared spectroscopy

Fingerprinting sediment sources in two typical watersheds in the dry-hot valleys of Southwest China: The role of gully and orchard land

Brazil is a leader in the adoption of conservation agriculture practices and technologies. However, the impact of these practices on sediment sources at the catchment scale has not been quantified yet, particularly in grain growing regions, where a conservationist no‐tillage system is implemented to protect soils. To address this knowledge gap, a sediment fingerprinting study based on elemental geochemistry was carried out in a large agricultural catchment (804 km2) of Southern Brazil where no‐tillage practices dominate. A total of 156 soil samples were taken to characterize the three main potential sediment sources: cropland (n = 79), unpaved roads (n = 41), and channel banks (n = 36). Sediment sampling was performed using a time‐integrated sampler (n = 33) and by collecting fine‐bed material (n = 34) at five locations across the catchment. Sediment was also sampled during flood events at the catchment outlet (n = 20). Sediment source contributions were calculated using an optimal suite of geochemical properties and a mixing model. Results showed that although the catchment is not particularly sensitive to soil erosion (i.e., deep clayey soils with gentle slopes), the amount of sediment supplied by cropland to the river network remains very high (up to 1.63 Mg·ha−1·year−1). Sediment fingerprinting results showed that even when no‐tillage is implemented, cropland remains an important source of sediment, supplying up to 70% of the material transiting the Conceição River. Accordingly, the current conservation farming system in this catchment needs to be improved to further reduce soil erosion and sediment yield.

Fingerprinting sediment sources in a typical karst catchment of southwest China

A mixing model to incorporate uncertainty in sediment fingerprinting

Intensive agriculture can impair river water quality. Soil quality monitoring has been used to measure the effect of land use intensification on water quality at a point and field scales but not at the catchment scale. Other farm scale land use pressures, like stocking rate and the value of land, which relate to land use intensity are now publicly available, nationally. We therefore tested whether point scale soil quality measures, together with newly available farm scale land use pressures (land valuation and stocking rate) and existing catchment and climatic characteristics could help predict the behaviour of water quality data across 192 catchments in New Zealand. We used a generalised additive model to make predictions of the change in nitrogen fractions (r2 = 0.65–0.71), phosphorus fractions (r2 = 0.51–0.70), clarity and turbidity (r2 = 0.42–0.46), and E. coli (r2 = 0.35) over 15 years. The state and trend of water quality was strongly related to a refined farm scale land use classification, and to catchment and climatic characteristics (e.g. slope, elevation, and rainfall). Relationships with point scale soil quality measures and the land use pressures were weak. The weak relationship with land use pressures may be caused by using a single snapshot in time (2022), which cannot account for lag times in water quality response but leaves room for additional temporal data to improve predictive power. The weak relationship to soil quality measures was probably caused by limited data points (n = 667 sites) that were unrepresentative of land use, and areas of catchment subject to processes like runoff or leaching. While national soil quality measures might be useful for evaluating environmental risk at the field or farm scale, without a large increase in sampling, they were not relevant at the catchment scale. Additional analyses should be performed to determine how many samples would be needed to detect a change using an environmentally focused soil test that can guide water quality management.

Novel approaches to investigating spatial variability in channel bank total phosphorus at the catchment scale

Compound-specific stable isotope (CSSI) fingerprinting of sediment sources is a recently introduced tool to overcome some limitations of conventional approaches for sediment source apportionment. The technique uses the 13C CSSI signature of plant-derived fatty acids (δ13C-fatty acids) associated with soil minerals as a tracer. This paper provides methodological perspectives to advance the use of CSSI fingerprinting in combination with stable isotope mixing models (SIMMs) to apportion the relative contributions of different sediment sources (i.e. land uses) to sediments. CSSI fingerprinting allows quantitative estimation of the relative contribution of sediment sources within a catchment at a spatio-temporal resolution, taking into account the following approaches. First, application of CSSI fingerprinting techniques to complex catchments presents particular challenges and calls for well-designed sampling strategies and data handling. Hereby, it is essential to balance the effort required for representative sample collection and analyses against the need to accurately quantify the variability within the system. Second, robustness of the CSSI approach depends on the specificity and conservativeness of the δ13C-FA fingerprint. Therefore, saturated long-chain (>20 carbon atoms) FAs, which are biosynthesised exclusively by higher plants and are more stable than the more commonly used short-chain FAs, should be used. Third, given that FA concentrations can vary largely between sources, concentration-dependent SIMMs that are also able to incorporate δ13C-FA variability should be standard operation procedures to correctly assess the contribution of sediment sources via SIMMs. This paper reflects on the use of δ13C-FAs in erosion studies and provides recommendations for its application. We strongly advise the use of saturated long-chain (>20 carbon atoms) FAs as tracers and concentration-dependent Bayesian SIMMs. We anticipate progress in CSSI sediment fingerprinting from two current developments: (i) development of hierarchical Bayesian SIMMs to better address catchment complexity and (ii) incorporation of dual isotope approaches (δ 13C- and δ 2H-FA) to improve estimates of sediment sources.

Community managed forests dominate the catchment sediment cascade in the mid-hills of Nepal: A compound-specific stable isotope analysis

Erosion and loss of soil is a large problem for water quality all over the world. Mitigation measures need to target the main sources of erosion at the right scale. For many areas there is still great uncertainty at the catchment level as to what are the main sources of erosion and how they can be managed. The objective of this article was to investigate the effect of tillage on soil losses at the field and catchment scale. Long-term monitoring data (20 years) from a catchment and a nested field site in the south eastern Norway were used. Results from the Vandsemb field showed that soil losses were on average 940 kg ha−1 yr−1 for years (n=5) when the field was autumn ploughed and 174 kg ha−1 yr−1 for years (n=8) when the field was not tilled in autumn. For the total monitoring period, surface soil losses were 330 kg ha−1 yr−1, whereas sub-surface soil losses were 130 kg ha−1 yr−1. Total phosphorus and dissolved reactive phosphorus (DRP) losses in surface runoff were 460 g ha−1 yr−1 and 150 g ha−1 yr−1, respectively and in sub-surface runoff they were 530 g ha−1 yr−1 and 270 g ha−1 yr−1, respectively. At the catchment scale, no relationship was found between the soil tillage on agricultural fields and the variation in losses of suspended sediments (SS) at the catchment scale. The average losses at Mørdre were for suspended sediments 2140 kg ha−1yr−1, for total phosphorus 2320 g ha−1 yr−1, and for DRP 240 g ha−1 yr−1. Weather conditions, especially during winter, were important for the temporal variation in losses of SS at both sites. Soil losses from the Mørdre catchment were approximately five times higher than from the Vandsemb field. The study findings suggest that changes in soil tillage were not the dominating cause of variations in loss of SS from the catchment, although these changes were important at the field scale.