Search

To truly understand the hydrologic and erosive processes that occur at the catchment scale regarding land use and soil management changes, intensive monitoring is required over a long period. Variables such as precipitation, flow rate and suspended sediment concentration are the fundamentals needed to estimate sediment yield (SY). However, in order to shed more light on the effects of soil management changes on SY, traditional hydrologic monitoring techniques can be paired with sediment tracing to identify sediment sources. In addition, mathematical models that predict erosion and SY can also help understand the dynamics of erosion and deposition between the hillslope and stream channel. Given the above, this article provides an overview of almost 20 years of monitoring (2002–2021) in the Arvorezinha experimental catchment (1.23 km2) located on the edge of the Brazilian meridional plateau in southern Brazil, which is subject to extreme erosion due to a combination of factors related to intensive agriculture and steep slopes. The catchment was selected as a study site to evaluate the effects of land‐use changes and soil conservation management on hydrology and SY. The text reviews the history of the study, synthesizes the main results and describes the evolution of the monitoring techniques used while giving special attention to the fingerprinting approach. In addition, the article discusses the importance of catchment‐scale studies for teaching, generating technical‐scientific knowledge and fostering collaboration between national and international research groups. [ABSTRACT FROM AUTHOR]

Abstract In areas with extensive soil erosion, there is a need to identify the sources supplying sediment to rivers in order to improve our understanding of erosion processes and to guide the implementation of effective soil management measures. Geochemical elements are often used as tracers in sediment fingerprinting studies. Among them, P has often been used as it may specifically trace sediment originating from cropland. Although P-fractions can provide more detailed information than the total P content, their potential to distinguish between different potential sources of sediments has rarely been tested. The objective of the current research was to evaluate the contribution of sediment sources in a rural catchment (Arvorezinha, 1.23 km2) of Southern Brazil, using different sets of tracers (i.e. P-fractions only, geochemical tracers only, and a combination of both tracers). Three potential sediment source types were identified in the catchment: (i) cropland, (ii) unpaved roads and (iii) stream channel banks. Thirty-one suspended sediment samples were collected in the river draining this catchment during 9 significant floods between 2009 and 2011. Estimates of P-fractions were obtained by Hedley fractionation using anion exchange resin membrane (P resin); 0.5 mol L−1 NaHCO 3 (pH 8.5) (P bic); 0.1 mol L−1 NaOH (P NaOH0.1); 1.0 mol L−1 HCl (P HCl); 0.5 mol L−1 NaOH (P NaOH0.5); and residual P (P residual). Total organic (P org) and inorganic P (P inorg) were obtained by the ignition method. The concentrations of P HCl , P NaOH0.5 and P inorg in suspended sediments were higher than those found in the sources. Accordingly, these fractions were considered not to be conservative and were excluded from further analysis. With the exception of P residual , all P-fractions passed the Kruskal-Wallis H -test and were considered as potential tracers of sediment sources. When the P-fractions were combined with geochemical tracers, the discrimination power was much higher than when using geochemical tracers only. The Mahalanobis distance was 9% higher than that observed in the geochemical approach. The calculated contribution of sediment sources using the three approaches was very similar, and cropland was the dominant source of sediment, with calculated contributions of 56 ± 19, 46 ± 16 and 48 ± 16% for those approaches based on geochemical elements, P-fractions, and geochemical + P-fractions, respectively. These results demonstrate the effectiveness of some P-fractions to trace potential sediment sources in rural catchments, as they have a high discriminant power while remaining conservative during sediment transport in the catchment. Graphical abstract Unlabelled Image Highlights • Combining P-fractions with geochemical tracers increased the discrimination power. • P-fractions alone provided results similar to those obtained with geochemical tracers. • Three P-fractions in suspended sediments were considered non-conservative. • All P-fractions were considered as potential tracers of sediment, except residual P. [ABSTRACT FROM AUTHOR]

Abstract: Sediments produced from eroding cultivated land can cause on‐site and off‐site effects that cause considerable economic and social impacts. Despite the importance of soil conservation practices (SCP) for the control of soil erosion and improvements in soil hydrological functions, limited information is available regarding the effects of SCP on sediment yield (SY) at the catchment scale. This study aimed to investigate the long‐term relationships between SY and land use, soil management, and rainfall in a small catchment. To determine the effects of anthropogenic and climatic factors on SY, rainfall, streamflow, and suspended sediment concentration were monitored at 10‐min intervals for 14 years (2002–2016), and the land use and soil management changes were surveyed annually. Using a statistical procedure to separate the SY effects of climate, land use, and soil management, we observed pronounced temporal effects of land use and soil management changes on SY. During the first 2 years (2002–2004), the land was predominantly cultivated with tobacco under a traditional tillage system (no cover crops and ploughed soil) using animal traction. In that period, the SY reached approximately 400 t·km−2·year−1. From 2005 to 2009, a soil conservation programme introduced conservation tillage and winter cover crops in the catchment area, which lowered the SY to 50 t·km−2·year−1. In the final period (2010–2016), the SCP were partially abandoned by farmers, and reforested areas increased, resulting in an SY of 150 t·km−2·year−1. This study also discusses the factors associated with the failure to continue using SCP, including structural support and farmer attitudes. [ABSTRACT FROM AUTHOR]

One of the challenges of modern agriculture is to adapting to climatic effects, including the capacity of rainy events in causing erosion. The understanding of these phenomena relies on monitoring rain variables that express the magnitude and pattern of erosive agents. Kinetic energy (E) is a fundamental variable to represent the erosivity and to enable the estimation of erosion and sediment yield in mathematical models. In Brazil, there are no direct and continuous measurements of E, and empirical equations to estimate rainfall intensity (I) are used instead. As a result, empirical equations to estimate E from I are used. To assess local behavior between these variables, rain variables (energy, volume and intensity) and associated processes (water flow (Q) and suspended sediment concentration (SSC)) were measured in a small rural catchment. The work was developed in a small rural catchment (1.23 km 2 ) located in southern Brazil where intensive land use with tobacco cultivation has caused high rates of erosion and sediment yield. This study proposes an alternative equation E = f(I) for the study region and provides a comparison with previous equations already proposed by Foster et al. (1981); van Dijk et al. (2002); Wilkinson (1975); Brown and Foster (1987) and USDA-ARS (2013). Furthermore, this work explores the seasonal variations and the differences in the magnitude of the events of these relationships. The measured E values were similar to those estimated by the equations of Foster et al. (1981); USDA-ARS (2013) and van Dijk et al. (2002), whereas Brown and Foster (1987) underestimate the estimation of E. Finally, it was evaluated the predictive ability of the total kinetic energy, 30 min maximum intensity (I 30 ) and total volume of the rainfall (P total ) in explaining the response variables which reflect the hydrology and sediment yield processes considering the Q and SSC. The main results of the work were: i) a new equation to estimate E based on the I measured, its seasonal variation and similarity with the previously known equations and ii) the relationship between E and the hydrological variables at catchment scale. [ABSTRACT FROM AUTHOR]

Purpose: The impact of agriculture on water resources has long been a problem associated with the formation of runoff, the siltation of lakes and reservoirs, and overall depletion of water quality. In Brazil, these problems are mainly related to soil degradation by water erosion. However, studies of catchment-scale erosion are still rare particularly in grain-producing regions which have adopted conservative tillage systems for soil protection. In order to contribute to a better understanding of the impact of conservation agriculture on water resources, this study determined the runoff coefficient and sediment yield for two agricultural catchments. Materials and methods: Hydrological and sedimentological monitoring was conducted in two catchments: the Conceicao catchment is characterized by grain production in weathered soils and a gently sloping landscape, while the Guapore catchment is characterized by heterogeneous soils and topography. Both catchments have problems associated with water erosion. Results and discussion: The magnitudes of annual runoff coefficients and sediment yield were high, even if compared to similar agricultural regions, including a catchment with widespread adoption of no-tillage. The sediment yield was 140 t km year, and the runoff coefficient was 14 % for the Conceicao catchment, while the sediment yield was 270 t km year, and the runoff coefficient was 31 % for the Guapore catchment. The results indicate that problems such as gullies, soil compaction, runoff, floods, siltation, and water quality depletion associated with the misuse of agricultural areas in terms of soil conservation and water use are still evident and important even in regions with widespread adoption of no-tillage systems. Conclusions: The magnitudes of both runoff and sediment yield clearly indicate the need to adopt complementary practices of soil conservation measures, such as mechanical runoff control. [ABSTRACT FROM AUTHOR]

Purpose: Hydrosedimentological modeling is a tool that can be used to understand better important processes occurring at the catchment scale, such as runoff and sediment yield. The aim of this study was to use the Limburg Soil Erosion Model (LISEM) to describe the runoff and sediment yield during rainfall-runoff events in a small rural catchment in southern Brazil. Materials and methods: The study was conducted in the Lajeado Ferreira Creek catchment (drainage area of 1.19 km) where intense land use has caused a negative impact on water resources. Thirteen rainfall-runoff events that occurred in 2010 and 2011, including high-magnitude events, were used to model hydrosedimentological processes. Results and discussion: Results: from the calibration and validation stages indicate that the model had a good performance when representing the hydrograph, including events with greater complexity. The use of a second soil layer in the model increased its efficiency, which is in accordance with the importance of subsurface flow in this catchment and its sensitivity to the physical properties of the soil, which are essential for controlling hydrosedimentological processes at the catchment scale. The simulation of sediment yield was overestimated by the model, constrained by the lack of sensitivity of the model to soil cohesion and the stability of soil aggregates. During the model calibration stage, these parameters had values different from those measured in the field. Conclusions: The LISEM model performed well in representing runoff for events of different magnitudes. The discretization of the physical-hydrologic properties in the soil profile enabled the evaluation of the effect of subsurface impediment layers on water infiltration and runoff. The simulation was less accurate for suspended sediment concentration than for runoff. This indicates the need for further studies to either identify other factors controlling erosion and sediment yield that have not been identified by the model, or identify if the representation of the physical parameters is inadequate, especially the values of soil cohesion and aggregate stability. [ABSTRACT FROM AUTHOR]

Purpose: Information on the effects of eucalyptus forests on hydrosedimentological processes is scarce, particularly at the catchment scale. Monitoring and mathematical modeling are efficient scientific tools used to address the lack of information for natural resource management and the representation and prediction of those processes. This study evaluates the effects of eucalyptus cultivation on hydrosedimentological processes in watersheds and to use the Limburg soil erosion model (LISEM) to represent and predict hydrological processes. Material and methods: The study was conducted in two forested watersheds: the main watershed (94.46 ha) and a nested sub-watershed (38.86 ha), both cultivated with eucalyptus and residual riparian native forest, located in southern Brazil. Hydrosedimentalogical monitoring was conducted from 16th February 2011 to 31st December 2012, and LISEM model calibrations were performed on the bases of six storms events. Results and discussion: The sediment yield for 2011 was 41.6 Mg km and 38.5 Mg km for the watershed and sub-watershed, respectively. An extreme event in 2012 provided greater sediment yield for the sub-watershed (99.8 Mg km) than that for the watershed (51.7 Mg km). Rainfall events with a greater maximum intensity generated rapid discharge and suspended sediment concentration responses in the sub-watershed due to the smaller drainage area and steeper landscape. In the main watershed, the accumulation of flood waves occurred for most events, with less steep hydrographs, and a later occurrence of the discharge peak after that of the sub-watershed. The LISEM adequately reproduced the peak discharge and runoff for the calibrated events; however, the peak time and the shape of the hydrograph were not adequately represented. Conclusions: The hydrosedimentological patterns of the watershed and sub-watershed, both cultivated with eucalyptus, was characterized by sedimentographs preceding hydrographs during rainfall-runoff events where scale effects occur, with maximum discharge and specific sediment yield greater in the watershed than that in the sub-watershed. Empirical models based on hydrologic variables may be used for estimating the suspended sediment concentration and sediment yield. Therefore, LISEM may be used for the prediction of hydrological variables in these forested watersheds. [ABSTRACT FROM AUTHOR]

• Geochemical tracers were not conservative in a catchment with highly weathered soils. • Magnetic and UV tracers had low discriminating potential between land uses. • The low discrimination between sources results in great uncertainties in the results. • Quantifying sediment contributions from more than three sources remains challenging. • Despite limited discrimination, results of all tracers combinations remained consistent. An important step in the sediment source fingerprinting approach is the selection of the appropriate tracing parameters to maximize source discrimination. The use of multiple tracing properties may reduce uncertainties and increase discrimination between sources. Accordingly, this study investigates the discrimination and quantifies the contribution of sediment sources delivering sediment to a river draining a homogeneous subtropical agricultural catchment based on the combination of ultraviolet–visible spectra derived parameters (UV), magnetic (M), and geochemical tracers (GEO). The investigated catchment (Conceição River − 804 km2), located in Southern Brazil, has predominantly deep and strongly weathered Ferralsols. The main land-uses found in the area are cropland (89%), pasture (5%) and forest (5%). A total of 187 samples were collected to characterise the five main sediment sources, including cropland, pastures, unpaved roads, gullies and stream banks. A total of 53 tracers, including 21 geochemical tracers, two magnetic properties and 30 parameters derived from UV spectra, were analysed. Tracers were selected following a three step procedure, including: (i) an interquartile range test, (ii) a Kruskal–Wallis H test, and (iii) a linear discriminant function analysis (LDA). The LDA was performed using six different sets of variables: (i) GEO only; (ii) UV only; (iii) M + UV (MUV); (iv) GEO + UV (GUV); (v) GEO + M (GM) and (vi) GEO + M + UV (GMUV). The selected tracers were introduced into a mass balance mixing model to estimate the source contributions to in-stream sediment by minimizing the sum of square residuals. Most geochemical tracers were considered not conservative by using the interquartile range test in this catchment with highly weathered soils. The GM approach resulted in the highest percentage of samples correctly classified (SCC), with 74%, followed by the approaches with GMUV and GUV, with 73%. Alternative tracers, UV individually or combined with M tracers, correctly classified only 59 and 60% of the samples, respectively. Moreover, they did not provide significant additional discrimination power even when combined with the GEO tracers. The apportionment model resulted in similar source contribution results for all approaches, with the absence of significant difference when comparing the mean source contributions obtained for the entire set of sediment samples (Cropland: 17–23%; Pastures: 24–34%; Unpaved Roads: 3–12%; Stream Banks: 26–31%; Gullies: 14–19%). Due to the strong homogeneity of soil types found in the Conceição catchment, these differences in source contributions remained very low and the results of the mixing model were impacted by the high number of potential sources and the relatively limited quality of the sediment source discrimination. According to the model results, the low discrimination between the potential sediment sources illustrates the difficulties for discriminating land-used based sediment sources, with more than three potential sources, in homogeneous catchments with highly weathered soils (e.g. Ferralsols, Nitisols) under tropical conditions. [ABSTRACT FROM AUTHOR]

Information on the main sources supplying deleterious sediment loads to river systems is needed to improve our understanding of soil erosion processes to better target erosion control measures. In particular, it is important to quantify the respective contributions of surface and subsurface sources to material degrading waterbodies. Radiocesium (137Cs), emitted during thermonuclear bomb testing (~1950–1980) and nuclear accidents provides significant discrimination between surface material exposed to atmospheric fallout and subsurface material sheltered from it. A systematic worldwide review of research articles (n = 123) that used 137Cs to trace sediment sources showed that the United Kingdom (n = 24), Australia (n = 23) and the United States (n = 20) had the highest number of publications utilizing 137Cs in a sediment tracing framework. In contrast, few studies were published for catchments in Africa (n = 9) or South America (n = 2). In the northern hemisphere, positive relationships were evident between 137Cs activities in surface and subsurface sources and the proportion of thermonuclear bomb fallout. However, given the low proportions of fallout received in regions between 0 and 20°N and 0–20°S, the potential application of 137Cs tracing techniques may be limited in this area of the world as well as in agricultural regions with severe soil erosion (i.e. Chinese Loess Plateau and South Africa). In total, 94% of the studies researching surface and subsurface sources that analysed 137Cs as a potential tracing property included this parameter in end-member mixing models. The main challenges for the future of this technique are related to the access to ultra-low background gamma spectrometry facilities and the development of surrogate tracers. Future research should ensure that basic catchment information and details on the sampling design are properly documented to ensure studies are reproducible. Moreover, researchers should provide raw measurement data to help improve our global understanding of 137Cs dynamics in soil erosion research. • The meta-analysis compiled 123 articles considering 137Cs for sediment tracing. • 70% of these studies were conducted in Europe, North America and Australia. • Fallout levels are too low in regions near the Equator and with extensive erosion. • 137Cs was included in 71 to 94% of studies modelling source contributions. • Future research should systematically include details on sampling design. [ABSTRACT FROM AUTHOR]

Most existing erosion models implicitly assume that runoff is generated due to infiltration excess, i.e. Hortonian overland flow. However, at catchment scale the dynamics is more complex because there exist others water sources and process involved, for example, translatory, return, and macropore flow, which can move slowly from soil towards to catchment outlet. Therefore, those must be included in studies of flow dynamics and description of erosion process. Because an error at representation of water sources can lead to errors on erosion process modelling, when slowly movements, are the main contributor at catchment outlet. Thus, we aim to investigate the role played by water sources in the catchment and how it may affect the process-based erosion models performance. The study was conducted in Arvorezinha catchment, Brazil, and has a drainage area of ca. 1.23 km². Currently, the catchment is characterized by the cultivation of tobacco, especially through the minimum cultivation system. Accumulated rainfall is measured in a rain gauge and used as a reference to adjust the measurements of two pluviograph. Monitoring of water discharge (Q) and suspended sediments concentration (SSC) is performed in the outlet of the catchment through a Parshall flume with a limnigraph and turbidimeter, respectively. Using dissolved silicon concentration (DSi) the hydrograph can be separated into two components: i) surface runoff, i.e., fast movement water (new water) and ii) groundwater+soil water, i.e., slow movement water (old water). Twenty-six samples were collected during subsurface+base flow conditions to characterize DSi in groundwater+soil water from 9/7/2011 to 9/7/2012. Sampling during events started on 10/01/2011 and lasted until to 10/26/2013, in total six rainfall events to characterize surface runoff. An aliquot of each sample (sediment + water) was collected and filtered through a 0.22 µm porosity. The determination of DSi content was through an ICP-OES. Modeling of the erosion process (sediment yield-SY) was performed using the Limburg Soil Erosion Model (2.01 version). A manual calibration procedure was used, and its procedure was repeated for both approaches: with and without runoff separation. For the six rainfall events evaluated, 62% of the hydrograph is formed by groundwater+soil water, especially in autumn-winter and with low-medium magnitude (up to Q = 800 L.s-1). Runoff separation by DSi was essential for obtaining the correct estimation of surface runoff volumes. Furthermore, the effect of dilution during the events was observed, i.e. as Q increases, DSi concentration decreases. Hydrossedimentological monitoring quantifies the sediment at the catchment outlet, which is the result of sediment from soil detachment in the catchment, carried by surface runoff and diluted in the flows that move slowly in outlet (subsurface and base flow). On the other hand, erosion process modeling done by the LISEM model is based only on sediment transported by surface runoff. Since the model simulates only surface flow, the dilution effect does not occur and thus SSC and SY are simulated incorrectly. Therefore, it is essential to consider different sources of water on the occurrence of runoff events, because the dynamics of the sediment is directly influenced by those. [ABSTRACT FROM AUTHOR]

The intensification of agricultural activities increases the transfer of phosphorus (P) to surface water bodies causing a wide range of environmental, social and economic problems, such as eutrophication. This paper describes the dynamics of P loss during rainfall events (intra-event) and between events (inter-event) in a rural catchment with intensive soil use in southern Brazil, emphasizing the relationships of the solutes with the hydrosedimentological variables. The information on the magnitude and the pattern of P transport was obtained during eighth significant rainfall events, in addition to the sampling performed during the baseflow period (between events). During the events, we evaluated the transfer dynamics of the total P (TP), dissolved P (DP), particulate P (PP) and suspended sediment concentration (SSC). Phosphorus (dissolved and particulate) is predominantly lost during events. The concentrations of DP and PP were low in samples collected in the period between events, when groundwater flow predominates. There is predominance of subsurface flow for the total volume of DP drained into the catchment. The hysteresis of TP forms a loop in the counterclockwise direction at the moment the soil is exposed by plowing, due to P transfer from the cropland to the water course. The transfer of P occurred predominantly in the particulate form (88%) due to the high affinity of P with mineral colloids. There was predominance of clockwise hysteresis for TP and SSC during the fallow period and at the beginning of soil mobilization, indicating the presence of P and sediment sources closer to the catchment outlet. However, during the development period of the crop, when the soil was more exposed due to plowing, counterclockwise hysteresis was observed for TP, which indicates a predominance of sources farther from the outlet, such as the cropland. Thus, we believe that efforts towards adopting practices that reduce erosion can contribute to a reduction of P transfer to water sources. Consequently controlling the process of eutrophication. [ABSTRACT FROM AUTHOR]

Owing to the rapid expansion of agriculture in South America in recent decades, soil erosion and fine sediment supply to river networks, which lead to deleterious on-site and off-site environmental impacts, are exacerbated in intensively cultivated catchments. Measuring soil inventories of bomb-derived fallout radiocesium (137Cs) bound to fine particles is one of the few techniques available to reconstruct soil redistribution rates and evaluate the sustainability of farming practices over the recent phase of agricultural intensification (1960s–2020). However, information about the spatial distribution of 137Cs fallout across the soils of South America remains scarce, and the published data has not been synthesized at the scale of this subcontinent so far. The objective of the current research is therefore to quantify and map the initial 137Cs fallout at the scale of South America, based on the compilation of published 137Cs inventories, additional measurements conducted on undisturbed soil profiles and digital soil mapping as this baseline information may be useful for a wide range of Earth Science applications. A database of 137Cs inventories at 96 reference sites (i.e. areas without soil erosion nor accumulation) has been compiled for a variety of soil profiles (Argentina = 10, Brazil = 34, Chile = 46, Uruguay = 5, French Guiana = 1) located between 5.3° North latitude and 53° South latitude. The spatial distribution of 137Cs fallout was shown to be highly latitude-dependent, with a maximum in the 30–50° South latitude band. There were higher fallout levels than expected between 20 and 60° South latitude compared to the previous estimations made by UNSCEAR. A partial least square regression approach based on rainfall data and geographical information as covariates was used to create a baseline map of 137Cs fallout in soils of continental South America. This baseline map provides a powerful reference dataset to anticipate the order of magnitude of 137Cs inventories in undisturbed soil profiles collected in Brazil and Southern Chile and for numerous other applications in Earth Sciences. The potential application of the 137Cs inventory technique in countries of South America in general, and in regions with vulnerable ecosystems threatened by the expansion of agricultural activities in particular, is discussed in light of this comprehensive literature review. Furthermore, the regions (i.e. the North-Western part of the continent) where additional samples should be collected in priority to improve this baseline map are outlined. Our results demonstrate that 137Cs inventories are sufficiently high to investigate soil redistribution rates in most of South American countries where detectable levels of 137Cs can be expected to be found in sites exposed to erosion. [ABSTRACT FROM AUTHOR]

• Spectral pre-processing, multivariate method, and spectral ranges were tested. • Quality of model calibration depends mainly on the spectral preprocessing technique. • NIR range alone or combined with other spectral ranges resulted in better models. • The lowest errors were obtained by combining UV–Vis, NIR and MIR spectra. • Savitzky-Golay 1st derivative and normalization were the best spectral pre-processing. In recent years, several sediment fingerprinting studies have used ultraviolet–visible (UV–Vis), near-infrared (NIR) and middle-infrared (MIR) spectroscopy as a low cost, non-destructive and fast alternative to obtain tracer properties to estimate sediment source contributions. For this purpose, partial last square regression (PLSR) has often been used to build predictive parametric models. However, spectra preprocessing and more robust and non-parametric models such as support vector machines (SVM) has gained little attention in these studies. Accordingly, the objectives of the current research were to evaluate (i) the accuracy of two multivariate methods (PLSR and SVM), (ii) the effect of eight spectra preprocessing techniques, and (iii) the effect of using the information contained in the UV–Vis, NIR and MIR regions considered either separately or in combination on sediment source apportionment. The estimated source contribution was then compared with contributions obtained by the conventional fingerprinting approach based on geochemical tracers. This study was carried out in the Arvorezinha catchment (1.23 km2) located in southern Brazil. Forty soil samples were collected in three main potential source (cropland surface, unpaved roads and stream channels) and twenty-nine suspended sediment samples collected at the catchment outlet during nine rainfall-runoff events were used in this study. Both PLSR and SVM models showed a higher accuracy when calibrated and validated with the spectra submitted to spectral processing when compared to the direct use of the raw spectra. The best model results were obtained with PLSR and SVM mathematical models associated with the spectral preprocessing techniques 1st derivative Savitzky-Golay (SGD1), normalization (NOR) and combining NOR + SGD1 in the UV–Vis + NIR + MIR. The lowest errors were observed when the UV–Vis + NIR + MIR bands were combined due to the gain in information and, consequently, the increase in discriminant power achieved by the models. Despite the good accuracy of the models calibrated and validated with the artificial mixtures, significant errors remain when results of source contributions are compared to those obtained with the conventional sediment fingerprinting technique based on geochemical tracers. Nevertheless, the magnitude of the contributions calculated by the spectroscopy and geochemical approaches remains very similar for all sources, especially when using the SVM-UV–Vis + NIR + MIR model. Therefore, spectroscopy proved to be a fast, cheap and accurate technique, offering an alternative to the conventional geochemical approach for discriminating sediment source contributions in agricultural catchments located in subtropical regions. [ABSTRACT FROM AUTHOR]