Search

To expand the knowledge base on natural infrastructure for erosion mitigation in the Andes, it is necessary to move beyond case by case empirical studies to comprehensive assessments. This study reviews the state of evidence on the effectiveness of interventions to mitigate soil erosion by water and is based on Andean case studies published in gray and peer-reviewed literature. Based on a systematic review of 118 case studies from the Andes, this study addressed the following research questions. (1) Which erosion indicators allow us to assess the effectiveness of natural infrastructure? (2) What is the overall impact of working with natural infrastructure on on-site and off-site erosion mitigation? (3) Which locations and types of studies are needed to fill critical gaps in knowledge and research? Three major categories of natural infrastructure were considered: restoration and protection of natural vegetation, such as forest or native grasslands, forestation with native or exotic species and implementation of soil and water conservation measures for erosion mitigation. From the suite of physical, chemical and biological indicators commonly used in soil erosion research, two indicators were particularly relevant: soil organic carbon of topsoil and soil loss rates at plot scale. The protection and conservation of natural vegetation has the strongest effect on soil quality, with 3.01±0.893 times higher soil organic carbon content in the topsoil compared to control sites. Soil quality improvements are significant but lower for forestation and soil and water conservation measures. Soil and water conservation measures reduce soil erosion to 62.1g %g ±g 9.2g %, even though erosion mitigation is highest when natural vegetation is maintained. Further research is needed to evaluate whether the reported effectiveness holds during extreme events related to, for example, El Niño-Southern Oscillation. © 2022 Veerle Vanacker et al.

In the Galapagos archipelago, about 96% of the land area has been declared a Protected National Park in 1959. Of the four inhabited islands, Santa Cruz is the most populated, with 15,393 inhabitants in 2010. The non-protected area in Santa Cruz corresponds to the south-central part of the island and the bay area around Puerto Ayora. Over the period 1961-2018, the agricultural land expanded from 6% to 67% of the non-protected land area. In a field-based study around the settlement of Santa Rosa, we monitored hydrometeorological and soil physical and hydrological properties over the period July 2019-December 2021. Six sites were monitored including two replicates per land cover type: (i) native Miconia forest, (ii) agricultural land, and (iii) abandoned farmland with invasive species. The spatiotemporal distribution of rainfall and air temperature over the sites is recorded via one weather station, four rain gauges, air temperature and relative humidity sensors; and the atmospheric input and rainfall were sampled at biweekly basis. After pedological characterization of the six profiles, soil and rock samples were taken per horizon for analysis of elemental chemistry, mineralogy, texture, C/N ratio, and organic matter content. Upslope of the soil profiles, TDR probes measured volumetric soil moisture content, soil electrical conductivity and temperature; and soil water samples were taken using suction lysimeters. Over the monitoring period, the highest rainfall amounts were measured in January (226 to 265 mm), and the lowest in May (20 to 25 mm). Most of the year, the relative air humidity is close to 100% with values dropping to 60% in March. The lowest air temperatures (15 °C) are measured in August, and the highest (29 °C) in March and April. Solar radiation strongly fluctuates from 80 W/m2 during the rainiest month to 220 W/m2 in March. Deeply weathered soils are developed on basaltic parent material and have a depth up to 50 cm. Soils are loose and lack macro-stru

Soil hydraulic properties control the provision of hydrological services. Vegetation and topography influence these properties by altering soil structure and porosity. The underlying mechanisms are not yet fully understood for the high Andean region. In this study, we examined how vegetation and topographic attributes are related to soil hydraulic properties and soil pore structure in young volcanic ash soils, and further correlated them to soil texture, organic carbon, and root characteristics to explain these relationships. In a 0.7 km2 study site located in the Andean páramo of northern Ecuador, we measured soil water retention, saturated hydraulic conductivity, bulk density (BD), and pore size distribution parameters on eight soil profiles with contrasting vegetation types (cushion-forming plants vs. tussock grasses) and topographic positions (summit vs. hillslope). We observed significant differences in soil hydraulic properties and soil pore structure in the uppermost horizons by vegetation type, whereas topography had a minor effect. In the A horizons, we found higher water retention at saturation and field capacity (10%–14%), higher total available water (8%–15%), and higher saturated hydraulic conductivity (4–12 times) under cushion-forming plants compared to tussock grasses. The elevated values under cushion plants were attributed to the presence of larger pores, lower soil BD, and higher soil organic carbon content as a result of coarser root systems. Total available water was generally high (0.34–0.40 cm3 cm−3), and locally not associated with any soil property. The higher water retention in soils under cushion vegetation can enhance soil water storage for plants and the regulation of water flows during prolonged rainfall events. The saturated hydraulic conductivity of the surface horizons is high compared to rainfall intensities resulting in high infiltration capacity; whilst its decline with depth reveals the potential for generation of subsurface stor

Soil hydraulic properties control the provision of hydrological services. Vegetationand topography influence these properties by altering soil structure and porosity. Theunderlying mechanisms are not yet fully understood for the high Andean region. Inthis study, we examined how vegetation and topographic attributes are related to soilhydraulic properties and soil pore structure in young volcanic ash soils, and furthercorrelated them to soil texture, organic carbon, and root characteristics to explainthese relationships. In a 0.7 km2study site located in the Andean páramo of northernEcuador, we measured soil water retention, saturated hydraulic conductivity, bulkdensity (BD), and pore size distribution parameters on eight soil profiles with con-trasting vegetation types (cushion-forming plants vs. tussock grasses) and topo-graphic positions (summit vs. hillslope). We observed significant differences in soilhydraulic properties and soil pore structure in the uppermost horizons by vegetationtype, whereas topography had a minor effect. In the A horizons, we found higherwater retention at saturation and field capacity (10%–14%), higher total availablewater (8%–15%), and higher saturated hydraulic conductivity (4–12 times) undercushion-forming plants compared to tussock grasses. The elevated values under cush-ion plants were attributed to the presence of larger pores, lower soil BD, and highersoil organic carbon content as a result of coarser root systems. Total available waterwas generally high (0.34–0.40 cm3cm 3), and locally not associated with any soilproperty. The higher water retention in soils under cushion vegetation can enhancesoil water storage for plants and the regulation of water flows during prolonged rainfall events. The saturated hydraulic conductivity of the surface horizons is highcompared to rainfall intensities resulting in high infiltration capacity; whilst its declinewith depth reveals the potential for generation of subsurface stormflow, especiallyb

The Andes region is prone to soil erosion because of its steep topographic relief, high spatio-temporal variability in precipitation and heterogeneity in lithological strength. Soil erosion by water is affecting natural and anthropogenic environments through its impacts on water quality and availability, loss of soil nutrients, flood risk, sedimentation in rivers and streams, and damage to civil infrastructure. Sustainable land and water management, referred here as natural infrastructure interventions, aims to avoid, reduce and reverse soil erosion and can provide multiple benefits for the environment, population and livelihoods. In this study, we present a systematic review of peer-reviewed and grey literature involving more than 120 local case-studies from the Andes. Three major categories of natural infrastructure interventions were considered: protective vegetation, soil and water conservation measures, and adaptation measures that regulate the flow and transport of water. The analysis was designed to answer the following research questions: (1) Which soil erosion indicators allow us to assess the effectiveness of natural infrastructure interventions across the Andean range? (2) What is the overall impact of implementing natural infrastructure interventions for on-site and off-site erosion mitigation? The systematic review shows that the effectiveness of protective vegetation on soil erosion mitigation is the most commonly studied characteristic, accounting for more than half of the empirical studies. From the suite of physical, chemical and biological indicators that were commonly used in soil erosion research, our review identified two indicators to be particularly suitable for the analyses of the effectiveness of natural infrastructure interventions: soil organic carbon (SOC) of the topsoil, and soil loss rates at plot scale. The implementation of soil and water conservation measures in areas under traditional agriculture had positive effects on SOC (1.28 to

The high tropical Andes ecosystem, known as páramo, provides important hydrological services to densely populated areas in the Andean region. In order to manage these services sustainably, it is crucial to understand the biotic and abiotic processes that control both water quality and fluxes. Recent research in the páramo highlights a knowledge gap regarding the role played by soil-vegetation interactions in controlling soil-water processes and resulting water and solute fluxes. Here, we determine the hydrological and geochemical fluxes in four soil profiles in the páramo of the Antisana´s water conservation area in northern Ecuador. Water fluxes were measured biweekly with field fluxmeters in the hydrological year Apr/2019- Mar/2020 under two contrasting vegetation types: tussock-like grass (TU) and cushion-forming plants (CU). Soil solution was collected in parallel with wick samplers and suction caps for assessing the concentrations of dissolved cations, anions and organic carbon (DOC). In addition, soil moisture was measured continuously in the upper meter of the soil profile, i.e. first three horizons (A, 2A and 2BC), using water content reflectometers. The vertical water flux in the upper meter of each soil profile was simulated using the 1D HYDRUS model. We carried out a Sobol analysis to identify sensitive soil hydraulic parameters. We then derived water fluxes by inverse modeling, based on the measured soil moisture. We validated the calculated water fluxes using the fluxmeter data. Solute fluxes were estimated by combining the water fluxes and the soil solution compositions. Our preliminary results suggest that water fluxes and DOC concentration vary under different vegetation types. The fluxmeter data from the 2A horizon indicates that the cumulative water flux under TU (2.8 - 5.7 l) was larger than under CU (0.8 – 1.1 l) during the dry season (Aug-Sep and Dec-Jan). However, the opposite trend was observed in the wet season for maximum water fluxes. Moreo

The aim of this research was to analyze the influence of vegetation and topography on the storage of organic carbon and the thickness of the soil in a micro-basin of páramo in the south of Ecuador.A total of 135 sampling sites were selected, which were distributed within the study area. The sites were classified based on the vegetation cover (cushion plants, tussock grasses and native forest) and the topographic position (concave, convex and straight). Altered and unaltered soil samples were collected from the horizons (O, A, AC and AC) at each site. The organic carbon concentration (%C) and bulk density (DA) were determined for the collected soil samples in order to calculate the total carbon storage (TCS). Results showed average TCS values of 27.43 kg/m2 , 34.27 kg/m2 , 41.05 kg/m2 , for tussock grasses, cushion plants and forest vegetation respectively; while for the topographic position the average values varied from 31.06 kg/m2 , 35.16 kg/m2 and 36.53 kg/m2 , for concave, convex and straight, respectively. Forest soils have the greatest accumulation of organic carbon, followed by soils formed under cushion plants and tussok grass covers; although the latter two represent the most important soils for carbon reserves due to their extension on the territory. Furthermore, results revealed that the vegetation cover exerts a greater influence on the soil depth (soil thickness) compared to the effect of topography, especially at the valley bottoms and plain areas covered by cushions plants. Finally, this study has highlighted the importance of carbon storage in the subsoil as an important carbon source and it should be integrated in future research for modelling purposes.

In the Andes region, soil erosion is a major problem, affecting natural and managed environments through its impacts on water quality and availability, loss of soil nutrients, flood risk, sedimentation in rivers and streams, and damage to infrastructure. Sustainable land and water management aiming to avoid, reduce and reverse land degradation can provide multiple benefits for the environment, population and livelihoods. Natural infrastructure, including forests, wetlands, and nature-based solutions can contribute to enhance the availability and quality of water for human consumption, agriculture, power production, and other land use-related ecosystem services. Here, we present a meta-analysis of published studies involving > 100 case-studies from the Andes. In the analyses, we included studies quantifying the effect of one or more elements of natural infrastructure on soil erosion mitigation. Three categories of natural infrastructure were considered: protective vegetation, soil and water conservation measures, and measures that regulate flow and transport of water. The analysis was designed to answer the following questions: (1) How do elements of natural infrastructure compare in terms of relative potential to mitigate soil erosion? (2) To what extent is the effectiveness of natural infrastructure elements dependent on ecosystem characteristics? Local case-studies were classified according to their spatial scale of analysis, experimental design, timespan of measurement, and methodological approach. Studies were spread across seven Andean countries: 1/3 of them were located in Peru, 1/3 in Ecuador, and the remaining part in Colombia, Venezuela, Bolivia, Chile and Argentina. Most studies were conducted at plot (34%) or small catchment scale (30%), followed by landscape analyses (27%) and studies at large catchment scale (9%). The meta-analysis shows that the effectiveness of protective vegetation on soil erosion mitigation is most commonly studied, representing mor

Mountain rivers connect the rugged mountainous terrain of the Andes with the Amazon basin. River incision into bedrock and deep-seated landsliding are the principal mechanisms by which the Andean escarpment is adjusting to regional tectonic forcing. Steep channel gradients in the escarpment zone facilitate upstream propagation of the fluvial knickzone into the Andean Amazon headwaters. Past studies on landscape evolution have often considered knickzone migration to be a steady process, responding to changes in external tectonic and climatic boundary conditions. This study on the Eastern Andean Cordillera in Ecuador uses empirical data on fluvial erosion rates measured on subannual, decadal and millennial time scales to document spatio-temporal variability of river incision and sediment transfer. Our data on the Coca River indicate that the non-steady character of the erosion rates can be attributed to autogenic variability strongly mediated by river bed lithological strength. Beyond illustrating the effect of bed lithology on fluvial erosion rates, our work also shows the limitations of forecasting future erosion rates based on short time series because of non-steady erosion processes in Andean Amazon basins.

Soils play a key role in the provision of vital ecosystem services. Soil functions, that deliver these services, are governed by soil properties. Soil structure is a fundamental property of soils since it controls water, geochemical and biological processes. The soil pore system, one of the main components of soil structure, can be affected by different biological feedbacks. Vegetation can have an impact on soil pore system through changes in pore size distribution and porosity, causing differences in soil hydraulic properties as well as soil-water processes. In high elevation tropical Andean ecosystems (páramos) little is still known about vegetation feedbacks on soil properties. At high elevation páramos (above 4100m), it is possible to find high diversity and co-dominance of plant species over short distances. In these landscapes, cushion plants and tussock grasses dominate alongside shrubs. These vegetation types, adapted to extreme local climatic conditions, are placed on young volcanic soils. We take advantage of this diverse setting, located within Antisana´s water conservation area in the north of Ecuador, by studying soil hydraulic properties and soil pore system in eight soil profiles. We hypothesize that the effect caused by Calamagrostis intermedia (tussock) and Azorella pedunculata (cushion) species on soil pore system and soil hydraulic properties at different horizons will be statistically different. In addition, we explore these effects in relation to other soil's physical properties and root traits. Soil hydraulic properties were determined on the basis of field observed saturated hydraulic conductivity as well as based on water retention contents at saturation (porosity), field capacity and permanent wilting point measured in the laboratory by the multi-step outflow method and the porous membrane pressure cell. Furthermore, water retention curves were fitted to measured data by the bimodal van Genuchten model. Based on these fittings the pore size

Tropical mountain regions are prone to high erosion rates, due to the occurrence of heavy rainfall events and intensely weathered steep terrain. Landslides are a recurrent phenomenon, and often considered as the dominant erosion process on the hillslopes and the main source of sediment. Quantifying the contribution of landslide-derived sediment to the overall sediment load remains a challenge. In this study, we derived catchment-average erosion rates from sediment gauging data and cosmogenic radionuclides (CRN), and examined their reliability and validity for constraining sediment yields in tectonically active regions. Then, we analysed the relationship between catchment-average erosion rates and landslide-derived sediment fluxes. The Pangor catchment, located in the western Andean mountain front, was selected for this study given its exceptionally long time series of hydrometeorological data (1974-2009). When including magnitude-frequency analyses of the sediment yields at the measurement site, the corrected gauging-based sediment yields remain one order of magnitude lower than the CRN-derived erosion rates. The underestimation of catchment-average erosion rates from gauging data points to the difficulty o f e xtrapolating fl ow fr equency an d se diment ra ting da ta in no n-stationary hy drological regimes, and severe undersampling of extreme events. In such conditions, erosion rates derived from cosmogenic radionuclides are a reliable alternative method for the quantification of catchment-average sediment yield. Landslide inventories from remote sensing data (1963-2010) and field measurements of landslide geometries are the input data for the derivation of landslide-derived sediment fluxes. The landslide-related erosion rates of 1688−+326901 and 630−+108300 t.km2.y-1 are similar to the CRN-derived erosion rates, likely indicating that landslides are the main source of sediment in this mountainous catchment. Copyright: © 2020 CSIC This is an open-access article d

Although climate exerts a major control on mineral weathering and soil formation processes, the combined effect of vegetation and topography can influence the rate and extent of chemical weathering at the hillslope scale. In this paper, we examined spatial patterns in volumetric strain and soil weathering extent associated with topographic gradients and vegetation patterns. In a high Andean catchment, we selected 10 soil toposequences on andesitic flows: 5 under tussock grasses, 3 under cushion forming plants, and 2 under native forest. Along each toposequence, one pit was excavated at the shoulder, backslope, and toeslope resulting in 30 soil profiles. Depth‐weighted total soil porosity of the 30 soil profiles averaged 64 ± 6%. The association between volumetric strain and soil organic C indicates that biotic agents can be effective in dilating the regolith during weathering. The young, postglacial volcanic soils were depleted in mono‐divalent and divalent cations, with total mass losses ranging between 793 and 1610 kg/m2 . The accumulation of Al‐humus complexes in the soil matrix plays an essential role in chemical transformation of the nonallophanic soils. Beyond the marginally significant topographic control on chemical weathering extent, our data show highly significant differences in chemical weathering extent between vegetation communities with total mass losses in forest soils being, respectively, 19% and 22% higher than in grasslands and cushion‐forming plants. The vegetation mosaic in alpine ecosystems might therefore provide essential clues to underst

Mountains play an important role in the denudation of continents and transfer erosion and weathering products to lowlands and oceans. The rates at which erosion and weathering processes take place in mountain regions have a substantial impact on the morphology and biogeochemistry of downstream reaches and lowlands. The controlling factors of physical erosion and chemical weathering and the coupling between the two processes are not yet fully understood. In this study, we report physical erosion and chemical weathering rates for five Andean catchments located in the southern Ecuadorian Andes and investigate their mutual interaction. During a 4-year monitoring period, we sampled river water at biweekly intervals, and we analyzed water samples for major ions and suspended solids. We derived the total annual dissolved, suspended sediment, and ionic loads from the flow frequency curves and adjusted rating curves and used the dissolved and suspended sediment yields as proxies for chemical weathering and erosion rates. In the 4-year period of monitoring, chemical weathering exceeds physical erosion in the high Andean catchments. Whereas physical erosion rates do not exceed 30 t km−2 y−1 in the relict glaciated morphology, chemical weathering rates range between 22 and 59 t km−2 y−1 . The variation in chemical weathering is primarily controlled by intrinsic differences in bedrock lithology. Land use has no discernible impact on the weathering rate but leads to a small increase in base cation concentrations because of fertilizer leaching in surface water. When extending our analysis with published data on dissolved and suspended sediment yields from the northern and central Andes, we observe that the river load composition strongly changes in the downstream direction, indicating large heterogeneity of weathering processes and rates within large Andean basins.

With the purpose of studying an integral dynamics between the soil and water factors in the high Andean ecosystems, this research focused on the modeling of water flow and nitrate transport through the sub surface soil solution of the Cuevas microbasin (4.2 km2), in the south of the Ecuadorian Andes. The Hydrus 1D model was used to model the water content and the transport of nitrate through the soil profile in thirty monitoring points distributed in the microbasin under the coverings of páramo, polylepis forest and cushion. The model has been shown to be capable of achieving acceptable adjustments, considering the PBIAS error index, using inverse modeling in the optimization of pedotransference parameters. The results obtained reveal that the soil under the forest cover would be storing a greater percentage of water, up to 10% more than that reached under the coverings of páramo and cushion; On the other hand, nitrate seems to be consumed more under the cushion coverage. The investigation supposes an important advance in the integral study of watersheds; the findings, methods and limitations found in this work could help in the planning of future research within the axis of hydrological and geochemical modeling.

Andean headwater catchments are an important source of freshwater for downstream water users. However, few long-term studies exist on the relative importance of climate change and direct anthropogenic perturbations on flow regimes in these catchments. In this paper, we assess change in streamflow based on long time series of hydrometeorological data (1974–2008) and land cover reconstructions (1963–2009) in the Pangor catchment (282 km2) located in the tropical Andes. Three main land cover change trajectories can be distinguished during the period 1963–2009: (1) expansion of agricultural land by an area equal to 14 % of the catchment area (or 39 km2) in 46 years' time, (2) deforestation of native forests by 11 % (or −31 km2) corresponding to a mean rate of 67 ha yr−1, and (3) afforestation with exotic species in recent years by about 5 % (or 15 km2). Over the time period 1963–2009, about 50 % of the 64 km2 of native forests was cleared and converted to agricultural land. Given the strong temporal variability of precipitation and streamflow data related to El Niño–Southern Oscillation, we use empirical mode decomposition techniques to detrend the time series. The long-term increasing trend in rainfall is remarkably different from the observed changes in streamflow, which exhibit a decreasing trend. Hence, observed changes in streamflow are not the result of long-term change in precipitation but very likely result from anthropogenic disturbances associated with land cover change.

Andean headwater catchments play a pivotal role to supply fresh water for downstream water users. However, few long-term studies exist on the relative importance of climate change and direct anthropogenic perturbations on flow regimes. In this paper, we assess multi-decadal change in freshwater provision based on long time series (1974–2008) of hydrometeorological data and land cover reconstructions for a 282 km2 catchment located in the tropical Andes. Three main land cover change trajectories can be distinguished: (1) rapid decline of native vegetation in montane forest and páramo ecosystems in ∼ 1/5 or 20 % of the catchment area, (2) expansion of agricultural land 10 by 14 % of the catchment area, (3) afforestation of 12 % of native páramo grasslands with exotic tree species in recent years. Given the strong temporal variability of precipitation and streamflow data related to El Niño–Southern Oscillation, we use empirical mode decomposition techniques to detrend the time series. The long-term increasing trend in rainfall is remarkably different from the observed changes in streamflow that exhibit a decreasing trend. Hence, observed changes in streamflow are not the result of long-term climate change but very likely result from direct anthropogenic disturbances after land cover change. Partial water budgets for montane cloud forest and páramo ecosystems suggest that the strongest changes in evaporative water losses are observed in páramo ecosystems, where progressive colonization and afforestation of high alpine grasslands leads to a strong increase in transpiration losses.

This thesis aims to study the effect of vegetation cover and anthropogenic activities on soil formation in a páramo micro-catchment in the upper part of the Yanuncay basin (Cuevas stream). Non-allophonic Andosols are the main soils in the study area where the overgrazing and burnings are the common agricultural practices. Three types of vegetation were identified, grass páramo (Pj), cushion páramo (PHA), and Polylepis (BP) forest. 6 soil catenas were made and 18 soil profiles were studied for the distinct types of vegetation. Results of the statistical analyses show that the availability of cations and Si is low, whereas the availability of Al and Fe is high and these elements are strongly correlated to organic carbon. The upper horizons present values of pH lower to 5, bulk density is lower to 0.9gr/cm³, and the values of the ratio of Alp/Alo are between 0.5 and 1. These findings indicate that the páramo soils of the southern Ecuador are non-allophonic Andosols and the alumínic-organic complexes prevail, while the formation of allophones is limited, therefore these soils are well developed. Our analysis show that there are two patterns between vegetation groups, apparently the grass páramo and Polylepis forest have the same behavior, as they do not show significative differences. On the other hand, the results for the cushion páramo show a distinct behavior in comparison to the other two groups. Land use plays an indirect role but important in soil development, particularly the Pj and Pha are subjected to overgrazing of animals and burnings, these agricultural practices contribute further to disturb the biogeochemical cycle of páramo ecosystems

Rapid land use/-cover change has increasingly transformed the hydrological functioning of tropical Andean ecosystems. The hydrological response to forest cover change strongly depends on the initial state of the ecosystem. Relatively little is known about human-disturbed ecosystems where forest plantations have been established on highly degraded land. In this paper, we analyze the impact of forest change on water and sediment fluxes for a highly degraded Andean catchment. Different pathways of land cover change (1963–2007) are observed in the Jadan catchment, with deforestation taking place in remote uplands and recovery and reforestation in the middle and lower parts where agricultural and bare lands are prevalent. Time series analyses of streamflow and rainfall data (1979/1982–2005/2007) show significant shifts in the distribution of rainfall and flow data. Changes in discharge are not resulting from changes in precipitation, as the direction of change is opposite. The removal of native forest for rangeland or croplands (by −20 km2) is likely to have contributed to the increase in total annual water yield, through an increase in annual baseflow by 25 mm. The observed changes in peakflow are important as the 1st percentile highest flow rates were 54% lower, while the 1st percentile rainfall amounts increased by 52%. The observed decrease in peakflow cannot be explained by clearcut of native forest, but is likely to be related to reforestation of degraded lands as well as spontaneous recovery of vegetation on remaining grazing lands. Over the same time period, a major decrease in specific sediment yields and suspended sediment loads was observed. Although deforestation in the upper parts led to increased landslide activity, this change is not reflected in an increased sediment yield. Small upland rivers are often nearly completely blocked by landslide material, thereby reducing their potential to transport sediment. In contrast, the reduction in estimated erosion i